Está en la página 1de 423

Departamento de Qumica Inorgnica e Ingeniera Qumica

FACULTAD DE CIENCIAS

UNIVERSIDAD DE CRDOBA

APROVECHAMIENTO INTEGRAL DEL RESIDUO DE LA INDUSTRIA DEL ACEITE DE


PALMA (EFB). OBTENCIN DE DERIVADOS DE LAS HEMICELULOSAS, PASTAS
CELULSICAS Y CELULOSA NANOFIBRILAR

INTEGRAL USE OF THE RESIDUE FROM THE PALM OIL INDUSTRY (EFB).
PRODUCTION OF DERIVATIVES OF HEMICELLULOSES, CELLULOSIC PULPS AND
NANOFIBRILLAR CELLULOSE

TESIS DOCTORAL PRESENTADA POR


ANA FERRER CARRERA
CRDOBA, ENERO 2013

TITULO: APROVECHAMIENTO INTEGRAL DEL RESIDUO DE LA INDUSTRIA


DEL ACEITE DE PALMA (EFB). OBTENCIN DE DERIVADOS DE
LAS HEMICELULOSAS, PASTAS CELULSICAS Y CELULOSA
NANOFIBRILAR
AUTOR: ANA FERRER CARRERA
Edita: Servicio de Publicaciones de la Universidad de Crdoba.
Campus de Rabanales
Ctra. Nacional IV, Km. 396 A
14071 Crdoba
www.uco.es/publicaciones
publicaciones@uco.es

APROVECHAMIENTO INTEGRAL DEL RESIDUO DE LA INDUSTRIA DEL ACEITE DE


PALMA (EFB). OBTENCIN DE DERIVADOS DE LAS HEMICELULOSAS, PASTAS
CELULSICAS Y CELULOSA NANOFIBRILAR

LOS DIRECTORES,

Fdo. Luis Jimnez Alcaide


Catedrtico de Universidad del
Departamento de Qumica Inorgnica
e Ingeniera Qumica de la Universidad
de Crdoba

Fdo. Alejandro Rodrguez Pascual


Profesor Titular del Departamento
de Qumica Inorgnica e Ingeniera
Qumica de la Universidad de Crdoba

MEMORIA PRESENTADA PARA OPTAR


AL GRADO DE DOCTOR INTERNACIONAL EN CIENCIAS QUMICAS

LA DOCTORANDA,

Fdo. Ana Ferrer Carrera


Licenciada en Ciencias Qumicas
por la Universidad de Crdoba

TTULO DE LA TESIS: Aprovechamiento integral del residuo de la industria del

aceite de palma (EFB). Obtencin de derivados de las hemicelulosas, pastas


celulsicas y celulosa nanofibrilar

DOCTORANDA: ANA FERRER CARRERA

INFORME RAZONADO DE LOS DIRECTORES DE LA TESIS

(se har mencin a la evolucin y desarrollo de la tesis, as como a trabajos y publicaciones derivados de la misma).

Los abajo firmantes, directores de esta Tesis de Doctorado, hacen constar que la
doctoranda ha realizado el trabajo que presenta como Memoria de Tesis de Doctorado en
los laboratorios de Ingeniera Qumica del Departamento de Qumica Inorgnica e
Ingeniera Qumica de la Universidad de Crdoba (Espaa), as como en los laboratorios
de Ingeniera Qumica del Departamento de Qumica-Fsica e Ingeniera Qumica de la
Universidad de La Corua (Espaa) y del Departamento de Tecnologa de Productos
Forestales de la Universidad de Aalto (Finlandia), donde ha realizado sendas estancias.
La evolucin y desarrollo de la Tesis ha sido la normal y el trabajo resultante rene las
condiciones exigidas para ser presentado como Tesis de Doctorado.
Finalmente hemos de hacer constar que de los datos y resultados obtenidos del trabajo,
se han publicado 8 artculos de investigacin en Revistas Cientficas especializadas de
elevado ndice de impacto. Igualmente, los resultados obtenidos han dado lugar a la
presentacin de 3 comunicaciones en Congresos Cientficos de gran trascendencia.
Por todo ello, se autoriza la presentacin de la Tesis Doctoral.
Crdoba, 7 de Noviembre de 2012
Firma de los directores

Fdo.: Luis Jimnez Alcaide

Fdo.: Alejandro Rodrguez Pascual

Departamento de Qumica Inorgnica e Ingeniera Qumica


FACULTAD DE CIENCIAS

UNIVERSIDAD DE CRDOBA

Luis Jimnez Alcaide, Catedrtico de Universidad, y Alejandro Rodrguez Pascual, Profesor


Titular, del Departamento de Qumica Inorgnica e Ingeniera Qumica de la Universidad
de Crdoba, en calidad de directores de la Tesis Doctoral presentada por la Licenciada en
Ciencias Qumicas, D.a Ana Ferrer Carrera,
INFORMAN:
Que la Memoria de Tesis Doctoral titulada: Aprovechamiento integral del
residuo de la industria del aceite de palma (EFB). Obtencin de derivados de las
hemicelulosas, pastas celulsicas y celulosa nanofibrilar ha sido desarrollada en los
laboratorios del Departamento de Qumica Inorgnica e Ingeniera Qumica de la Facultad
de Ciencias de Crdoba (Universidad de Crdoba, Espaa), del Departamento de QumicaFsica e Ingeniera Qumica (Universidad de La Corua, Espaa) y del Departamento de
Tecnologa de Productos Forestales (Universidad de Aalto, Finlandia). Adems, la citada
Memoria se ha realizado utilizando la metodologa adecuada y haciendo uso de las
tcnicas idneas para culminar los objetivos previstos. Asimismo, la Licenciada Ferrer
Carrera ha mostrado en todo momento una alta dedicacin e inters por el tema de
trabajo, superando con creces los niveles de calidad cientfica exigibles para la obtencin
del Grado de Doctor. Como consecuencia de los trabajos realizados, se han publicado
varios artculos en revistas internacionales indexadas en el Journal Citation Reports (JCR)
y se han presentado numerosas comunicaciones en congresos, tanto nacionales como
internacionales. A nuestro juicio, esta tesis rene todos los requisitos exigidos a este tipo
de trabajos, por lo que autorizamos su presentacin y posterior defensa pblica.
Y para que conste y surta los efectos pertinentes, expiden el presente informe en la
ciudad de Crdoba, a 7 de Noviembre de 2012.

Fdo.: Dr. Luis Jimnez Alcaide

Fdo.: Dr. Alejandro Rodrguez Pascual

Departamento de Qumica Inorgnica e Ingeniera Qumica


FACULTAD DE CIENCIAS

UNIVERSIDAD DE CRDOBA

Dr. Arturo Fco. Chica Prez, Director del Departamento de Qumica Inorgnica e
Ingeniera Qumica, de la Universidad de Crdoba,
INFORMA:
Que la Memoria de Tesis Doctoral titulada: Aprovechamiento integral del
residuo de la industria del aceite de palma (EFB). Obtencin de derivados de las
hemicelulosas, pastas celulsicas y celulosa nanofibrilar ha sido desarrollada en los
laboratorios del Departamento de Qumica Inorgnica e Ingeniera Qumica de la Facultad
de Ciencias de Crdoba (Universidad de Crdoba, Espaa), del Departamento de QumicaFsica e Ingeniera Qumica (Universidad de La Corua, Espaa) y del Departamento de
Tecnologa de Productos Forestales (Universidad de Aalto, Finlandia). Adems, la citada
Memoria se ha realizado utilizando la metodologa adecuada y haciendo uso de las
tcnicas idneas para culminar los objetivos previstos. Asimismo, la Licenciada Ferrer
Carrera ha mostrado en todo momento una alta dedicacin e inters por el tema de
trabajo, superando con creces los niveles de calidad cientfica exigibles para la obtencin
del Grado de Doctor. Como consecuencia, esta Memoria rene todos los requisitos
exigidos a este tipo de trabajos, por lo que autorizo su presentacin y posterior defensa
pblica.

Crdoba, a 7 de Noviembre de 2012.

Fdo.: Dr. Arturo Fco. Chica Prez

Mediante la defensa de esta Memoria se pretende optar a la obtencin de la Mencin de


Doctorado Internacional, habida cuenta de que la doctoranda rene los requisitos
exigidos para tal mencin:

1. Cuenta con los informes favorables de dos doctores pertenecientes a instituciones


de Enseanza Superior de pases europeos distintos de Espaa.
2. Uno de los miembros del tribunal que ha de evaluar la Tesis pertenece a un centro
de enseanza superior de otro pas distinto de Espaa.
3. Parte de la defensa de la Memoria se realizar en la lengua oficial de otro pas que
sea distinta a cualquiera de las lenguas oficiales de Espaa.
4. La doctoranda ha realizado una estancia de Investigacin mnima de 3 meses de
duracin en los laboratorios del Departamento de Tecnologa de Productos
Forestales de la Universidad de Aalto, Espoo, Finlandia, gracias a la concesin de
una ayuda para estancias en el extranjero asociada a una beca de Formacin de
Profesorado Universitario del Ministerio de Ciencia e Innovacin del Gobierno de
Espaa, que ha contribuido a su formacin y permitido desarrollar parte del
trabajo experimental de esta Memoria.

RECONOCIMIENTO DE APOYOS INSTITUCIONALES


El presente trabajo ha sido desarrollado en la Universidad de Crdoba, dentro de
la

lnea

de

investigacin

denominada

APROVECHAMIENTO

DE

MATERIALES

LIGNOCELULSICOS que pertenece al grupo consolidado del Plan Andaluz de


Investigacin, Desarrollo e Innovacin (PAIDI) INGENIERA QUMICA (RNM-271).
El soporte econmico para el desarrollo de las actividades vinculadas al trabajo
experimental proviene de una Beca del Programa de Formacin de Profesorado
Universitario (FPU) del Ministerio de Ciencia e Innovacin con referencia AP2008-00860, y
cuya resolucin tuvo lugar el 8 de Julio de 2009. Asimismo, la financiacin del trabajo
tambin ha sido posible gracias a los Proyectos de Investigacin CTQ2007-65074-C02-01,
TRACE2009-0064, CTQ2010-19844-C02-01 y P10-TEP-6261, liderados por los Directores
de esta Tesis Doctoral.

AGRADECIMIENTOS

La redaccin y defensa de esta Memoria de Tesis Doctoral culmina con el alcance de una meta
personal cuyo comienzo tuvo lugar hace cuatro aos ya. Result ser una oportunidad maravillosa para
adentrarme en el mundo de la Investigacin y sin la cual, no sera hoy lo que soy en cuanto a lo
profesional se refiere. As, hoy intento plasmar todo los resultados que, con gran dedicacin, he obtenido
tras una larga y, al mismo tiempo, fructfera etapa de laboratorio, seguida de un proceso de intenso
aprendizaje de interpretacin y discusin de los mismos. No obstante, en el transcurso de todos estos
aos, jams me he encontrado sola y he recibido apoyo y ayuda de un gran nmero de personas. Se podra
decir que estas personas, de un modo u otro, han contribuido a la realizacin de la misma. Es por ello que
merecen una mencin muy especial: mis directores, mis compaeros y amigos y, por supuesto, mi familia.
Mis directores
Los Directores de este Trabajo de Investigacin, D. Luis Jimnez Alcaide y D. Alejandro
Rodrguez Pascual, a los que deseo expresar mi ms sincero agradecimiento por su constante dedicacin
y, por supuesto, por la confianza que desde el primer momento depositaron en m.
Mis compaeros y amigos
D. Sixto Ruiz Madueo, por la inestimable ayuda recibida, tanto fuera como dentro del
laboratorio, durante la realizacin de este trabajo de investigacin y por los buenos momentos que me ha
brindado gracias a su buen humor.
Los doctores Rafael Snchez Serrano (Rafa choro) y Ana Requejo Silva (mi otra mitad, la famosa
Pili). No s qu decir que no sepis ya, sois increbles y he pasado muy buenos momentos en este
departamento gracias a ustedes. Siempre dispuestos a echarme una mano, a aconsejarme y a sacarme una
sonrisa cuando ms lo necesitaba.
El resto de compaeros del rea de Ingeniera Qumica del Departamento de Qumica Inorgnica e
Ingeniera Qumica de la Universidad de Crdoba. Deseo expresar especialmente mi gratitud a mi Anetta
y mi Carmencita (gracias por vuestra amistad y cario: tan lejos y al mismo tiempo tan cerca); Antoito
(no os preocupis que yo hago tesis); Zoilo y Ftima (compaeros de batalla en el laboratorio); Pepe
(siempre tan optimista y correcto); Inma (siempre dispuesta a resolver cualquier problema logstico); Ins,
Ana Caete, Ada, Manolo Dios, Antonio Serrano, Marisa y Mary. Todos y cada uno de ellos me han
aportado tantas y tantas cosas que resulta prcticamente imposible el poder resumir ahora en un par de
lneas lo que han significado para m. Gracias a todos por cada sonrisa, cada nimo, cada consejo, etc. y
por supuesto por la ayuda recibida en cada momento que la he necesitado.
No me gustara olvidar a todas aquellas personas que me han recibido con los brazos abiertos
durante las estancias de investigacin realizadas en el transcurso de estos aos. En primer lugar, gracias
a D. Alberto de Vega y a D. Pablo Ligero, por su apoyo y por ser unos jefes excepcionales durante mi
estancia en la Facultade de Ciencias, Universidade de A Corua (Espaa). Tambin quera agradecer a
mis compaeros de laboratorio (Tania, Ana, Mara, Estefana, Martn, entre otros) su inestimable apoyo
y cario que recib durante mi estancia en La Corua. Special thanks should be given to Professor
Orlando J. Rojas, for his collaboration and support during and after my research stay in the University

School of Science and Technology, Aalto University (Finland). His scientific advice and friendship are
very important for me. I am also very grateful to Mao for her hospitality and help. Moreover, a special
mention goes to my colleagues Anne, Marcelo, Michael and Carmen for their valuable help and kindness
during my stay in Finland. I am also grateful to the rest of the members of the University School of
Science and Technology (especially to the lab technicians) who were helping me every time that I needed
it. Adems me gustara dar las gracias a mi querida indgena, la doctora Cristina Castro, que se
convirti en mi mano derecha, la persona ms importante para m durante mi estancia en Finlandia y
siempre le estar agradecida por demostrarme tanto en tan poco tiempo. Eres nica.
Tambin me gustara dedicar unas palabras de agradecimiento a personas que estn fuera de mi
mbito profesional, pero que igualmente son imprescindibles para m. Mis amigas de siempre
(especialmente, mi Pequea, Nevado e Irene) y mis amigos de la carrera (Mrquezsi, Mrquezva, Glam,
Kiko, entre muchos otros). A todos los que he nombrado y a los que no he podido nombrar por espacio,
gracias por vuestro apoyo y cario.
Mi familia
Especialmente mis padres, Jos Luis y Ana Rosa; dos pilares fundamentales en mi vida y que sin
sus consejos, apoyo y ayuda no sera hoy lo que soy. Gracias de todo corazn por estar siempre a mi lado
en todo momento y espero algn da corresponderos, al menos la tercera parte, por todo lo que habis
hecho por m.
Mi hermano, Abel; muy distintos, pero siempre dispuestos a ayudarnos en todo lo posible. Siempre
te estar agradecida por tu apoyo incansable tanto en los malos como en los buenos momentos. Del mismo
modo, mi abuela Anita, siempre te recordar como alguien muy especial en mi vida.
Adicionalmente, deseo recordar a una persona que hoy por hoy desgraciadamente ya no est, pero
que vive en m igualmente, mi abuelo Rafael. S por descontado lo orgullossimo que estara de m y la
ilusin que le hubiese hecho verme convertida en Doctora. Le agradezco desde lo ms profundo de mi
alma que me haya querido tanto y que siempre haya estado a mi lado. S que nunca me abandonar.

A todos, sin excepcin, GRACIAS

Son vanas y estn plagadas de errores las ciencias que no han nacido del
experimento, madre de toda certidumbre
Leonardo da Vinci
Hay alguien tan inteligente que aprende de la experiencia de los dems
Voltaire

ndice/Index

1. RESUMEN/ABSTRACT ........................................................................................................ 3
2. OBJETIVOS/AIMS ................................................................................................................ 17
3. INTRODUCCIN .................................................................................................................. 27
3.1. MATERIALES LIGNOCELULSICOS ........................................................................ 27
3.1.1. ESTRUCTURA Y COMPOSICIN DE LOS MATERIALES
LIGNOCELULSICOS ........................................................................................... 28
3.1.1.1. Estructura de la pared celular ........................................................................ 28
3.1.1.2. Componentes estructurales de la pared celular ............................................ 30
3.1.1.2.1. Celulosa .......................................................................................... 31
3.1.1.2.2. Hemicelulosas ................................................................................ 32
3.1.1.2.3. Lignina ............................................................................................ 33
3.1.1.3. Componentes no estructurales de la pared celular ....................................... 36
3.1.2. CULTIVOS LIGNOCELULSICOS ..................................................................... 36
3.1.2.1. Fibras madereras ............................................................................................ 37
3.1.2.2. Fibras alternativas a las madereras ................................................................. 38
3.1.2.2.1. Materias primas alternativas para la produccin de pastas ............ 39
3.1.2.2.2. Caractersticas generales de las materias primas alternativas ........ 42
3.2. BIOREFINERA DE MATERIALES LIGNOCELULSICOS .................................... 44
3.2.1. TRATAMIENTO HIDROTRMICO .................................................................... 49
3.2.1.1. Composicin de las hemicelulosas ................................................................ 50
3.2.1.1.1. Xilanos ............................................................................................ 51
3.2.1.1.2. Mananos ......................................................................................... 51

3.2.1.1.3. Galactanos ...................................................................................... 52


3.2.1.1.4. Glucanos ........................................................................................ 52
3.2.1.1.5. Xiloglucanos .................................................................................. 52
3.2.1.1.6. Sustancias pcticas ......................................................................... 52
3.2.1.2. Despolimerizacin de las hemicelulosas por tratamiento hidrotrmico ..... 53
3.2.1.3. Condiciones de operacin en el tratamiento hidrotrmico ......................... 55
3.2.1.3.1. Temperatura .................................................................................. 55
3.2.1.3.2. Tiempo de operacin .................................................................... 55
3.2.1.3.3. Concentracin de slidos .............................................................. 55
3.2.1.3.4. Tamao de partcula ..................................................................... 56
3.2.1.4. Efectos de los tratamientos hidrotrmicos sobre las hemicelulosas ............ 56
3.2.1.5. Productos obtenidos en el tratamiento hidrotrmico .................................. 58
3.2.1.5.1. Xilooligosacridos ......................................................................... 58
3.2.1.5.2. Medios fermentativos obtenidos mediante la combinacin de
etapas de autohidrlisis-posthidrlisis .......................................... 59
3.2.1.5.3. Sustancias antioxidantes ................................................................ 60
3.2.1.5.4. Productos de la hidrlisis enzimtica del residuo slido
del tratamiento hidrotrmico ....................................................... 61
3.2.1.5.5. Pasta celulsica .............................................................................. 61
3.2.2. PROCESOS DE OBTENCIN DE PASTAS CELULSICAS .......................... 62
3.2.2.1. Consideraciones generales ............................................................................ 62
3.2.2.2. Pasteados qumicos convencionales ............................................................. 63

3.2.2.2.1. Pasteado a la sosa ........................................................................ 64


3.2.2.2.2. Pasteado kraft o al sulfato ........................................................... 68
3.2.2.2.3. Pasteado al sulfito ....................................................................... 69
3.2.2.3. Pasteados con disolventes orgnicos ............................................................. 70
3.2.2.3.1. Pasteado con alcoholes .................................................................. 73
3.2.2.3.2. Pasteado con cidos orgnicos ...................................................... 75
3.2.2.3.3. Pasteado con etanolaminas ............................................................ 75
3.2.3. PROCESOS DE BLANQUEO DE PASTAS CELULSICAS ............................ 76
3.2.3.1. Consideraciones generales ............................................................................. 76
3.2.3.2. Blanqueo con oxgeno ................................................................................... 79
3.2.3.3. Blanqueo con ozono ...................................................................................... 81
3.2.3.4. Blanqueo con perxido de hidrgeno .......................................................... 83
3.2.3.5. Blanqueo con enzimas ................................................................................... 92
3.2.4. CELULOSA NANOFIBRILAR (NFC): COMPOSICIN
Y PRINCIPALES APLICACIONES ........................................................................ 95
3.2.4.1. Produccin de celulosa micro/nanofibrilada ................................................ 97
3.2.4.2. Aplicaciones de la celulosa nanofibrilar ........................................................ 101
3.2.4.2.1. Aplicaciones mdicas ..................................................................... 101
3.2.4.2.2. Refuerzo de composites ................................................................. 101
3.2.4.2.3. Embalaje ......................................................................................... 102
3.3. PALMA ACEITERA ........................................................................................................ 103
3.3.1. GENERALIDADES ................................................................................................... 103

3.3.2. USOS DEL ACEITE DE PALMA ........................................................................... 105


3.3.3. RESIDUO DE LA INDUSTRIA DEL ACEITE DE PALMA ............................. 106
4. MATERIAL Y MTODOS .................................................................................................... 113
4.1. CARACTERIZACIN DE LOS MATERIALES ........................................................... 113
4.1.1. CARACTERIZACIN FSICO-QUMICA DE LA MATERIA PRIMA ........... 113
4.1.2. CARACTERIZACIN QUMICA DE LA FRACCIN SLIDA
PROCEDENTE DEL TRATAMIENTO HIDROTRMICO ............................ 114
4.1.3. CARACTERIZACIN QUMICA DE LA FRACCIN LQUIDA
PROCEDENTE DEL TRATAMIENTO HIDROTRMICO ............................ 114
4.1.4. CARACTERIZACIN FSICO-QUMICA DE LA PASTA ............................... 114
4.1.5. CARACTERIZACIN DE LAS HOJAS DE PAPEL ........................................... 115
4.1.6. CARACTERIZACIN QUMICA DE LAS LEJAS O LICORES
NEGROS DE COCCIN PROCEDENTES DEL PASTEADO ........................ 115
4.1.7. CARACTERIZACIN DE LAS SUSPENSIONES Y PELCULAS
DE CELULOSA NANOFIBRILAR (NFC) ............................................................ 116
4.2. EQUIPOS UTILIZADOS EN LA EXPERIMENTACIN ........................................... 117
4.2.1. MICROSCOPIO PARA MEDIDA DEL TAMAO DE LAS
FIBRAS DE LA MATERIA PRIMA ........................................................................ 117
4.2.2. MOLINO PARA EL DESMENUZADO DE LA MATERIA PRIMA ................ 117
4.2.3. DIGESTOR PARA TRATAMIENTOS HIDROTRMICOS
Y PASTEADOS ......................................................................................................... 118
4.2.4. DESFIBRADOR DE MATERIALES COCIDOS .................................................. 119

4.2.5. SEPARADOR DE FIBRAS E INCOCIDOS .......................................................... 119


4.2.6. SEPARADOR DE FIBRAS Y AGUA ...................................................................... 120
4.2.7. REFINADOR-DESINTEGRADOR DE FIBRAS .................................................. 121
4.2.8. REFINADOR .............................................................................................................. 121
4.2.9. CENTRFUGA PARA ESCURRIDO DE PASTAS ............................................... 122
4.2.10. DESINTEGRADOR DE PASTAS CELULSICAS ........................................... 122
4.2.11. REFINMETRO SHOPPER-RIEGLER .............................................................. 123
4.2.12. FORMADOR DE HOJAS ....................................................................................... 123
4.2.13. PRENSA PARA HOJAS .......................................................................................... 124
4.2.14. SECADOR DE HOJAS ........................................................................................... 124
4.2.15. FLUIDIZADOR A ALTA PRESIN .................................................................... 125
4.2.16. FORMADOR DE PELCULAS DE CELULOSA NANOFIBRILAR (NFC) ... 125
4.2.17. MICROSCOPIO DE FUERZA ATMICA (AFM) ............................................. 126
4.2.18. INSTRUMENTACIN PARA LA MEDIDA DE PROPIEDADES
FSICAS DE LAS HOJAS DE PAPEL Y PELCULAS
DE CELULOSA NANOFIBRILAR (NFC) ........................................................... 126
4.3. PROCEDIMIENTOS ...................................................................................................... 127
4.3.1. TRATAMIENTO HIDROTRMICO DE LA MATERIA PRIMA .................... 127
4.3.2. PROCESO DE PASTEADO .................................................................................... 127
4.3.3. PROCESO DE REFINADO DE LAS PASTAS ..................................................... 128
4.3.4. PROCESO DE BLANQUEO DE LAS PASTAS CON
PERXIDO DE HIDRGENO ............................................................................. 128

4.3.5. FORMACIN DE SUSPENSIONES DE CELULOSA


NANOFIBRILAR (NFC) .......................................................................................... 128
4.3.6. FORMACIN DE HOJAS DE PAPEL Y PELCULAS
DE CELULOSA NANOFIBRILAR (NFC) ............................................................ 129
4.4. DISEO DE EXPERIMENTOS Y MODELOS ............................................................ 130
4.4.1. DISEO FACTORIAL DE EXPERIMENTOS .................................................... 130
4.4.2. MODELO POLINMICO ...................................................................................... 130
4.4.3. MODELO NEUROBORROSO .............................................................................. 131
5. BIBLIOGRAFA/REFERENCES ........................................................................................... 141
6. PUBLICACIONES CIENTFICAS/SCIENTIFIC PUBLICATIONS .................................. 173
6.1. INFLUENCE OF TEMPERATURE, TIME, LIQUID/SOLID
RATIO AND SULPHURIC ACID CONCENTRATION ON THE
HYDROLYSIS OF PALM EMPTY FRUIT BUNCHES ............................................... 173
6.2. PULPING OF EMPTY FRUIT BUNCHES (EFB) FROM THE PALM OIL
INDUSTRY BY FORMIC ACID .................................................................................... 191
6.3. MILOX FRACTIONATION OF EMPTY FRUIT BUNCHES FROM

ELAEIS GUINEENSIS .................................................................................................... 217


6.4. TCF BLEACHING OF SODA-ANTHRAQUINONE AND DIETHANOLAMINE
PULP FROM OIL PALM EMPTY FRUIT BUNCHES ................................................ 241
6.5. MODELING HYDROGEN PEROXIDE BLEACHING OF SODA PULP
FROM OIL-PALM EMPTY FRUIT BUNCHES ........................................................... 257
6.6. A NEURAL FUZZY MODEL APPLIED TO HYDROGEN PEROXIDE

BLEACHING OF NON-WOOD SODA PULPS .......................................................... 283


6.7. BIOBLEACHING OF PULP FROM OIL PALM EMPTY FRUIT BUNCHES
WITH LACCASE AND XYLANASE ............................................................................ 307
6.8. VALORIZATION OF RESIDUAL EMPTY PALM FRUIT BUNCH FIBERS
(EPFBF) BY MICROFLUIDIZATION: PRODUCTION OF
NANOFIBRILLATED CELLULOSE AND EPFBF NANOPAPER ............................ 329
7. CONCLUSIONES/CONCLUSIONS .................................................................................... 355
8. ANEXO/APPENDIX .............................................................................................................. 371
9. PRODUCCIN CIENTFICA/SCIENTIFIC PRODUCTION ........................................... 383

Resumen/Abstract

Resumen/Abstract
RESUMEN

La fabricacin de papel implica dos etapas: obtencin de pasta y formacin de las


hojas. La obtencin de pasta consiste en la transformacin de las materias primas en fibras
celulsicas aisladas en una suspensin acuosa diluida, mediante la separacin de la lignina,
que es el agente que, a modo de cemento, aglomera las fibras celulsicas y fija su posicin.
En la formacin de hojas de papel, la pasta, tratada convenientemente para desarrollar
algunas propiedades caractersticas, se somete a operaciones que tienen por finalidad
extenderla en forma de lminas y eliminar la mayor parte de su humedad (Jimnez, 2005).
Las fibras de celulosa constituyen un componente esencial de los tejidos vegetales,
cuya funcin es la de dar resistencia a los mismos. La celulosa para la fabricacin de papel se
obtiene principalmente de madera (55 %), de otras fibras vegetales denominadas alternativas
(9 %) y de papel recuperado (36 %) (Brown y col., 2000). Pero, en definitiva, aproximadamente un 88-92 % de las materias primas utilizadas en la fabricacin de papel est constituido por maderas frondosas y conferas, correspondiendo el resto a las materias primas
alternativas (Jimnez, 2005). Entre stas ltimas se encuentran los residuos agrcolas, los
residuos agroindustriales (como los procedentes de la industria del aceite de palma o el bagazo de la caa de azcar) y los procedentes de cultivos alternativos a los agroalimentarios
(como tagasaste, leucaena, paulownia, etc.) (Alaejos y col., 2004; Jimnez, 2005).
La produccin de pasta celulsica para papel u otros usos se encuentra en la
actualidad con la necesidad de buscar nuevas materias primas alternativas a las convencionales (maderas), motivada principalmente por aspectos econmicos y medioambientales. En efecto, la utilizacin de madera como principal materia prima est
conduciendo a un problema de abastecimiento, as como a deforestaciones incontroladas de
bosques autctonos y tropicales que constituyen autnticas reservas mundiales, cuya
desaparicin ocasionara serios problemas ecolgicos. Por todo ello, se est estudiando la
posibilidad de utilizar residuos agrcolas, vegetales alternativos a las maderas clsicas (anuales
o no) y residuos de las industrias agroalimentarias, como puede ser el residuo de la industria
del aceite de palma (EFB, del ingls Empty Fruit Bunches), que se estudia en este trabajo.
Las materias primas alternativas son utilizadas en muchos pases industrializados para
producir papeles especiales; sin embargo, en otros pases se convierten en la principal

Resumen/Abstract
materia prima para la fabricacin de papel, como es el caso de China, donde suponen el 60
% de las fibras utilizadas para la produccin de papel (http://www.arbolesornamentales.com).
Estas fibras alternativas presentan un gran potencial de desarrollo para sustituir a las
fibras madereras, a pesar de ciertos inconvenientes como la posible inferior calidad de las
pastas; el elevado coste de estas materias primas para situarlas a la entrada de los digestores,
debido a la caresta de la recogida, transporte y dems manipulaciones; y las producciones
estacionales y de corta duracin, por lo que en breves periodos de tiempo se ha de
recolectar todo el material requerido para un ao de trabajo (Jimnez, 2005).
Por otra parte es digno de resaltar el inters que los consumidores muestran por
adquirir productos que respeten e incluso ayuden a la conservacin del medio ambiente, lo
que les lleva en ocasiones a estar dispuestos a pagar un poco ms a fin de conseguir
productos de los denominados ecolgicos. En este sentido, la utilizacin de este tipo de
materias primas alternativas podra contribuir a preservar y mantener el medio ambiente, al
conseguir disminuir la gran cantidad de madera utilizada como materia prima en la
produccin de pastas para papel.
La bsqueda de tecnologas ms respetuosas con el medio ambiente debe incluir, al
tiempo de usar materias primas alternativas a la madera, la utilizacin de procesos de
pasteado y de blanqueo de las pastas menos contaminantes, y el aprovechamiento de las
otras fracciones de las materias primas distintas de la celulosa. Es posible lograr esto
utilizando pretratamientos hidrotrmicos, que permitan separar hemicelulosas que pueden
dar lugar a diversos productos (aditivos alimentarios, frmacos, azcares, etanol, xilitol,
furfural, etc.); procesos de pasteado que no utilicen azufre como los procesos a la sosa y
los que utilizan disolventes orgnicos para la extraccin de la lignina; y tratamientos de
separacin de los disolventes orgnicos (para reciclarlos en el caso de utilizarse en el
pasteado) y lignina, que puede transformarse en diversos productos qumicos (resinas,
poliuretanos, acrilatos, epxidos, composites, etc.). De esta manera se mejora el rendimiento
y la economa del proceso global y se preserva el medio ambiente, con la consiguiente
repercusin ecolgica favorable.
Aunque desde hace tiempo se conoce la posibilidad de conseguir pastas mediante la
utilizacin de disolventes orgnicos sin azufre, hasta el momento slo se han puesto a punto
estos procesos (denominados organosolv) en plantas piloto (Jimnez, 2005; Lpez y col.,
2006; Muurinen, 2000; Rodrguez y Jimnez, 2008), por lo que se sigue investigando en

Resumen/Abstract
dichos procesos y, tambin, en la mejora de procesos clsicos de pasteado sin azufre,
principalmente en los procesos a la sosa y a la sosa-antraquinona (Jimnez, 2005).
La capacidad de produccin actual de pastas, si bien se ha incrementado en la mayora
de los pases desarrollados en la ltima dcada, no cubre las necesidades de consumo, que
estn aumentando mucho en pases en vas de desarrollo y an en pases desarrollados
(Aspapel, 2008); por lo que se han de construir nuevas plantas, pero utilizando procesos que
impliquen inversiones pequeas, bajos costes de produccin, alta calidad de productos y
conservacin adecuada del medio ambiente, as como un buen aprovechamiento de las
materias primas (operando de forma que se obtengan altos rendimientos) y recurriendo a
materias primas alternativas. Todo ello se consigue con la aplicacin de procesos de
pasteado que utilizan en la coccin reactivos sin azufre, como la sosa y disolventes orgnicos;
con la utilizacin de materias primas alternativas, como EFB; y con la aplicacin de procesos
de blanqueo de las pastas resultantes que empleen agentes blanqueantes sin cloro elemental
o totalmente libres de cloro (blanqueos ECF y TCF).
Por otra parte, cabe mencionarse el inters reciente que ha despertado la produccin
de estructuras nanofibrilares a partir de las pastas celulsicas. Concretamente, la produccin
de celulosa nanofibrilar (NFC, del ingls Nanofibrillar Cellulose) est siendo estudiada en
varios centros de investigacin en todo el mundo, tanto a nivel bsico como industrial. La
produccin de NFC se ha hecho bsicamente a partir de pastas blanqueadas, sin embargo
recientes experiencias han mostrado que la produccin de NFC a partir de pastas crudas
(con elevado contenido en lignina) es factible (Spence y col., 2010); incluso se ha observado
que algunas propiedades fsicas pueden mejorar debido a la presencia de lignina. De
especial inters es la posibilidad de obtener pelculas para empacado con alta capacidad de
barrera de oxgeno (aire) y otros gases y vapores. El uso de pastas de EFB para la
produccin de NFC es novedoso y no se ha encontrado en la bibliografa ninguna referencia
al respecto, pues hasta la fecha la mayora de los esfuerzos se han centrado en la produccin
de NFC a partir de pastas blanqueadas de especies madereras.
La presente Memoria de Tesis de Doctorado es fruto del trabajo realizado en los
laboratorios de Ingeniera Qumica del Departamento de Qumica Inorgnica e Ingeniera
Qumica de la Universidad de Crdoba (Espaa), as como en los laboratorios de Ingeniera
Qumica del Departamento de Qumica-Fsica e Ingeniera Qumica de la Universidad de
La Corua (Espaa) y en los laboratorios del Departamento de Tecnologa de Productos
Forestales de la Universidad de Aalto (Finlandia), donde la doctoranda ha realizado sendas
5

Resumen/Abstract
estancias. La financiacin del trabajo ha sido posible gracias a los Proyectos de Investigacin
CTQ2007-65074-C02-01, TRACE2009-0064, CTQ2010-19844-C02-01 y P10-TEP-6261,
liderados por los Directores de esta Tesis Doctoral, as como a la Beca FPU disfrutada por
la doctoranda los ltimos tres aos.
El objetivo de la Tesis es el aprovechamiento integral del residuo de la industria del
aceite de palma (EFB), para la obtencin de derivados de las hemicelulosas, pastas
celulsicas y celulosa nanofibrilar.
El residuo de la industria del aceite de palma, ms conocido como EFB (Empty Fruit
Bunches), est constituido por el material lignocelulsico que acompaa al fruto de la palma
de aceite (es decir, los racimos sin fruto). El EFB utilizado en este trabajo proviene de la
palma africana (Elaeis guineensis), originaria de frica occidental, de la cual se obtena aceite
hace 5000 aos, especialmente en Guinea Occidental. Despus de los viajes de Coln se
introdujo en Amrica y, en pocas ms recientes, desde Amrica pas a Asia llegando a
Sumatra y Malasia hacia 1900 (Rodrguez y col., 2008). Malasia es el mayor productor de
aceite de palma en la actualidad, con el 51 % de la produccin mundial, lo que constituye un
gran recurso econmico para ese pas. Tambin se est extendiendo el cultivo de la palma
aceitera a pases de frica Occidental (Nigeria, Guinea, Ghana, etc.), Amrica Latina
(Ecuador, Colombia, Honduras, etc.) y Asia (Tailandia) (http://www.wrm.org.uy).
Las plantaciones de palma aceitera comienzan a producir frutos a los 4-5 aos de ser
plantadas, alcanzando su mayor produccin entre los 20 y 30 aos, despus declinan y dejan
de ser rentables, especialmente por la altura a la que se encuentran los frutos. Los racimos
pesan entre 15 y 25 Kg, conteniendo de unos 1000 a 4000 frutos de forma ovalada, de 3 a 5
cm de longitud. Cada hectrea de palma aceitera produce 10 toneladas anuales de frutos, de
los cuales se extraen 3000 Kg de aceite de palma, su principal aplicacin
(http://www.arbolesornamentales.com).
En este trabajo, en primer lugar, se caracteriz qumica y morfolgicamente el EFB,
determinando las solubilidades en agua caliente y en sosa al 1 % y los contenidos de
extrables con etanol-benceno, cenizas, holocelulosa, -celulosa y lignina, as como la
longitud y dimetro de las fibras.
Posteriormente, el EFB se someti a procesos hidrotrmicos, obtenindose dos
fracciones: una lquida, constituida por productos de la degradacin de las hemicelulosas
(monmeros, oligmeros, cidos orgnicos y productos de descomposicin de azcares) y

Resumen/Abstract
otra slida, rica en -celulosa y lignina. Se estudi la influencia de las variables de operacin
en los tratamientos hidrotrmicos del EFB, con y sin catalizador, sobre el rendimiento de la
fraccin slida y la composicin de la fraccin lquida, caracterizada mediante HPLC, con el
fin de determinar la distribucin de los principales componentes en dicha fraccin.
Por otra parte, la fraccin slida procedente del tratamiento hidrotrmico se someti a
distintos pasteados con diferentes reactivos qumicos (sosa-antraquinona, etanol y
etanolamina) con el fin de obtener pastas que puedan utilizarse en la produccin de papel.
Adems, se estudi el pasteado del EFB para la produccin de pastas de papel,
utilizando un proceso clsico o convencional (a la sosa-antraquinona) y otros novedosos
como los que utilizan disolventes orgnicos (etanol, etanolamina, dietanolamina, cido
peroxifrmico y cido frmico).
En los pasteados del EFB con cido frmico (proceso Formosolv) y cido
peroxifrmico (proceso Milox), se estudi la influencia de las variables de operacin sobre
las caractersticas de las pastas y hojas de papel resultantes, con el fin de encontrar las
condiciones ptimas de operacin.
Tambin se llevaron a cabo estudios sobre el blanqueo de pastas del EFB, mediante
procesos TCF (totalmente libres de cloro). Se consider el blanqueo de pastas a la sosaantraquinona de EFB utilizando perxido de hidrgeno, estudiando la influencia de las
variables de operacin sobre las caractersticas de las pastas blanqueadas. Igualmente se
estudi el blanqueo de pastas a la sosa-antraquinona y a la dietanolamina de EFB
mediante la secuencia de blanqueo OZP (oxgeno - ozono - perxido de hidrgeno). Finalmente, tambin se estudi el blanqueo de pastas a la sosa-antraquinona de EFB con el uso
de un pretratamiento enzimtico.
Para terminar, se procedi a la obtencin de celulosa nanofibrilar mediante el proceso
de microfluidizacin, y se determinaron las propiedades qumicas, fsicas y morfolgicas de
las suspensiones acuosas y pelculas de celulosa nanofibrilar formadas, utilizando pastas del
EFB obtenidas con sosa-antraquinona, cido frmico y cido peroxifrmico, respectivamente.

Resumen/Abstract

Resumen/Abstract
ABSTRACT

Papermaking involves two stages: pulping and formation of paper sheets. The pulping
consists of the transformation of the raw materials into isolated cellulosic fibers within a
dilute aqueous suspension. This is possible by separating lignin, which is the agent that
agglomerates the cellulosic fibers and fixes or cements their position. In the formation of
paper sheets, pulp, suitably treated to develop some characteristic properties, is subjected to
operations that are intended to extend the pulp into sheets and eliminate most of its
moisture (Jimnez, 2005).
Cellulose fibers are an essential component of plant tissues, whose function is to give
them resistance. Cellulose for paper manufacturing is mainly obtained from wood (55 %)
from other vegetable fibers known as alternatives (9 %) and from recovered paper (36 %)
(Brown et al., 2000). But, ultimately, about 88-92 % of the raw materials used in the manufacture of paper consists of hardwoods and softwoods, corresponding the rest to alternative
raw materials (Jimnez, 2005). Agricultural residues, agro-industrial residues (such as those
from the palm oil industry or bagasse from sugar cane) and those from the agri-food
alternative crops (such as tagasaste, leucaena, paulownia, etc.) are among the latter (Alaejos et
al., 2004; Jimnez, 2005).
Currently, the production of cellulose pulp for paper or for other uses needs to seek
new alternative raw materials to the conventional ones (wood), driven mainly by economical
and environmental aspects. Indeed, the use of wood as the main raw material is leading to a
supply problem, as well as uncontrolled deforestations of native and tropical forests which
are world reserves. This can cause serious ecological problems in the future. Therefore, the
possibility of using agricultural residues, alternative vegetables to traditional wood (annual or
otherwise) and agri-food industry residues, such as the residue from the palm oil industry
(EFB, Empty Fruit Bunches), is being studied.
Alternative raw materials are used in many industrialized countries to produce
specialty papers; but in other countries they become the main raw material for papermaking,
such as China, where they account for 60 % of the fibers used for the production of paper
(http://www.arbolesornamentales.com).
These alternative fibers have a great potential for development to replace wood fibers,
despite some drawbacks such as a possible lower quality of the pulp; the high cost of raw
9

Resumen/Abstract
materials to place them at the entrance of the digesters, due to the high cost of collection,
transportation and other handling; and seasonal productions with a short duration, so in
short periods of time it has to collect all the material required for a years work (Jimnez,
2005).
Also noteworthy is the interest shown by consumers to buy products that respect and
even help to preserve the environment, which sometimes leads them to be willing to pay a
little more to get so-called ecological products. In this sense, the use of such alternative raw
materials could help to preserve and maintain the environment, in order to reduce the
amount of wood used as raw material in the production of paper pulp.
The search for technologies that respect more the environment should include (not
only the use of alternative raw materials to wood) the use of pulp pulping and bleaching
processes less pollutant and the use of other fractions of raw materials different from
cellulose. This can be achieved by using hydrothermal pretreatments, which allow the
separation of hemicelluloses which may give rise to different products (food additives,
pharmaceuticals, sugars, ethanol, xylitol, furfural, etc.); pulping processes which do not use
sulphur as the processes "soda" and those using organic solvents for the extraction of lignin;
and treatments of separation of the organic solvents (for recycling if used in the pulping
process) and lignin, which can be transformed into various chemicals (resins, polyurethanes,
acrylates, epoxides, composites, etc.). This will improve performance and overall process
economy as well as preserving the environment, resulting in a positive ecological impact.
Although, the possibility of obtaining pulps using organic solvents without sulphur has
long been known, these processes have only been recently initiated in pilot plants (Jimnez,
2005; Lpez et al., 2006; Muurinen, 2000; Rodrguez and Jimnez, 2008); so further
research in these processes continues and also in the improvement of sulphur-free classic
pulping processes, mainly in the processes "soda" and "soda-anthraquinone" (Jimnez, 2005).
The current production capacity of pulps, although it has risen in most developed
countries in the last decade, does not cover consumption needs, which are increasing sharply
in developing countries and even in developed countries (Aspapel, 2008); so that new plants
must be built, but using processes that involve small investment, low production cost, high
quality products and adequate environmental conservation as well as a good utilization of raw
materials (operating so that high yields will be obtained) and using alternative raw materials.
This is achieved by applying pulping processes that use chemicals without sulphur in
cooking, such as soda and organic solvents; with the use of alternative raw materials such as
10

Resumen/Abstract
EFB; and by applying pulp bleaching processes which employ elemental chlorine free
bleaching agents or totally chlorine free (ECF and TCF bleaching).
On the other hand, the recent interest awakened in nanofibrillated structure
production from cellulosic pulps should be mentioned. Specifically, the production of
"Nanofibrillar Cellulose" (NFC) is being studied in several research centers around the world,
both basic and industrial levels. The production of NFC has come primarily from bleached
pulps, however, recent experiences have shown that the production of NFC from raw pulps
(with high lignin content) is feasible (Spence et al., 2010). It has even been observed that
some physical properties can be improved due to the presence of lignin. Of particular
interest is the possibility of obtaining films for packaging with high oxygen barrier capability
(air) and other gases and vapors. The use of EFB pulps for the production of NFC is novel
and no reference in the literature about it has been found, since to date most efforts have
focused on the production of NFC from bleached wood pulps.
This Doctoral Dissertation Report is the work done in the laboratories of Chemical
Engineering, the Department of Inorganic Chemistry and Chemical Engineering, at the
University of Crdoba (Spain), in the laboratories of Chemical Engineering, the Department
of Physical Chemistry and Chemical Engineering, at the University of La Corua (Spain) and
in the laboratories of the Department of Forest Products Technology, at Aalto University
(Finland), where the PhD student has carried out both stays. Funding for the work has been
thanks to the Research Projects CTQ2007-65074-C02-01, TRACE2009-0064, CTQ201019844-C02-01 and P10-TEP-6261, led by the PhD supervisors, as well as the FPU PhD
Scholarship, which the PhD student awarded three years ago.
The objective of this Thesis is the integral use of the residue from the palm oil industry
(EFB) to obtain derivatives of the hemicelluloses, cellulosic pulps and nanofibrillar cellulose.
The residue from the palm oil industry, better known as EFB (Empty Fruit Bunches),
consists of the lignocellulosic material accompanying the oil palm fruit (i.e., bunches without
fruit). The EFB used in this work has been from the Elaeis guineensis (African palm), native
to West Africa, from which oil has been obtained for the last 5000 years, especially in
western Guinea. After the voyages of Columbus it was introduced in America and, more
recently, from America it went to Asia where it was introduced in Sumatra and Malaysia in
1900 (Rodrguez et al., 2008). Malaysia is the largest producer of palm oil today, with 51 %
of the world production, which is a major economic resource for the country. The
cultivation of oil palm is also spreading to West African countries (Nigeria, Guinea, Ghana,
11

Resumen/Abstract
etc.), Latin America (Ecuador, Colombia, Honduras, etc.) and Asia (Thailand)
(http://www.wrm.org.uy).
Oil palm plantations start to bear fruit after 4-5 years of being planted, with the highest
production between 20 and 30 years, then decline and cease to be profitable, especially due
to the height at which the fruits are found. The bunches weigh between 15 and 25 Kg,
containing about 1000 to 4000 oval fruits, from 3 to 5 cm in length. Each hectare of oil palm
produces 10 tons of fruit per year, which is extracted from 3000 Kg of palm oil, which is its
main application (http://www.arbolesornamentales.com).
In this work, firstly, EFB was characterized chemically and morphologically,
determining solubilities in hot water and in 1 % sodium hydroxide and extractable contents
of ethanol-benzene, ash, holocellulose, -cellulose and lignin, as well as the length and
diameter of the fibers.
Subsequently, EFB underwent hydrothermal processes, yielding two fractions: a liquid
one, consisting of products from degradation of the hemicelluloses (monomers, oligomers,
organic acids and sugar decomposition products) and a solid phase-rich in -cellulose and
lignin. The influence of the operating variables in the EFB hydrothermal treatments, with
and without a catalyst, on the yield of the solid fraction and the composition of the liquid
fraction, characterized by HPLC, to determine the distribution of the main components in
this fraction, was studied.
On the other hand, the solid fraction from the hydrothermal treatment was subjected
to various pulping processes with different chemical reagents (soda-anthraquinone, ethanol
and ethanolamine) in order to obtain pulps which can be used in the production of paper.
In addition, the EFB pulping process for the production of paper pulps, using a
classical or conventional process ("soda-anthraquinone") and other innovative processes as
those using organic solvents (ethanol, ethanolamine, diethanolamine, peroxyformic acid and
formic acid) was studied.
In the EFB pulping processes with formic acid (Formosolv process) and peroxyformic
acid (Milox process), the influence of the operating variables on the characteristics of the
resulting pulps and paper sheets in order to find the optimum operating conditions, was
studied.
Moreover, studies related to the EFB pulp bleaching were carried out, using TCF
(totally chlorine free) processes. "Soda-anthraquinone" pulp bleaching of the EFB using
12

Resumen/Abstract
hydrogen peroxide was considered. The influence of the operating variables in the
characteristics of the bleached pulps, was studied. Also, the soda-anthraquinone" and
"diethanolamine" pulp bleaching processes of the EFB were studied, using the OZP
bleaching sequence (oxygen - ozone - hydrogen peroxide). Finally, the "soda-anthraquinone"
EFB pulp bleaching, using an enzymatic pre-treatment, was also studied.
Finally, nanofibrillar cellulose was obtained by the microfluidization process, and the
chemical, physical and morphological properties of aqueous suspensions and films of
nanofibrillar cellulose formed by using EFB pulps obtained with soda-anthraquinone, formic
acid and peroxyformic acid, respectively, were determined.

13

Resumen/Abstract

14

Objetivos/Aims

Objetivos/Aims
OBJETIVOS

El objetivo general del presente Trabajo es el aprovechamiento integral del residuo de


la industria del aceite de palma (EFB, es decir, los racimos sin fruto de la palma de aceite),
siguiendo el esquema de la Biorefinera, con la idea de aprovechar sus principales constituyentes.
En este trabajo se abordarn los siguientes puntos:
1. La caracterizacin fsico-qumica del EFB, determinando sus principales
constituyentes, as como la longitud y dimetro de las fibras.
2. El estudio de la influencia de las variables de operacin en el tratamiento
hidrotrmico del EFB, con y sin catalizador, sobre el rendimiento de la
fraccin slida y la composicin de la fraccin lquida.
3. El estudio de la influencia de las variables de operacin en la obtencin de
pastas celulsicas del EFB, utilizando disolventes orgnicos, sobre las caractersticas de las pastas resultantes, con el fin de encontrar las condiciones ptimas de operacin.
4. Blanqueo de pastas del EFB mediante procesos totalmente libres de cloro,
estudiando la influencia de las variables de operacin sobre las caractersticas
de las pastas blanqueadas.
5. Obtencin de celulosa nanofibrilar, y determinacin de propiedades qumicas,
fsicas y morfolgicas de las suspensiones acuosas y de las pelculas de celulosa
nanofibrilar formadas, utilizando diferentes pastas del EFB.
De manera ms concreta, el plan de trabajo propuesto es el que sigue:
I.

Caracterizacin del EFB, determinando las solubilidades en agua caliente y en


sosa al 1 % y los contenidos de extrables con etanol-benceno, cenizas, holocelulosa, -celulosa y lignina, as como la longitud y dimetro de las fibras.

II.

Estudio de la influencia de las variables de operacin del tratamiento


hidrotrmico del EFB sin catalizador (temperatura, tiempo y relacin lquido/slido) y con catalizador (temperatura, tiempo y concentracin de cido

17

Objetivos/Aims
sulfrico), sobre el rendimiento de la fraccin slida y la distribucin de los
principales componentes en la fraccin lquida.
III.

Caracterizacin de la fraccin lquida obtenida en el punto II, determinando


los contenidos de azcares y cido actico, mediante la utilizacin del HPLC.
Determinacin del rendimiento de la fraccin slida obtenida en el punto II
mediante gravimetra.

IV.

Estudio del pasteado de la fraccin slida del tratamiento hidrotrmico del


EFB y comparacin de las propiedades de las pastas y hojas de papel con las
correspondientes a pastas y hojas de papel obtenidas mediante pasteado del
EFB.

V.

Estudio de la influencia de las variables de operacin en la obtencin de pastas


celulsicas del EFB con disolventes orgnicos (tiempo, concentraciones de
reactivos), sobre las caractersticas de las pastas crudas.

VI.

Caracterizacin de las pastas obtenidas en el punto V, determinando su


rendimiento, viscosidad, nmero kappa y grado de blancura.

VII.

Caracterizacin qumica de la leja o licor negro de coccin procedente del


pasteado considerado como ptimo obtenido en el punto V.

VIII.

Estudio del blanqueo de pastas a la sosa-antraquinona y a la dietanolamina


del EFB mediante la secuencia de blanqueo OZP (oxgeno - ozono - perxido
de hidrgeno).

IX.

Estudio de la influencia de las variables de operacin en el blanqueo con


perxido de hidrgeno de pasta a la sosa-antraquinona del EFB, sobre las
caractersticas de las pastas blanqueadas.

X.

Estudio del blanqueo TCF de pasta a la sosa-antraquinona del EFB


mediante el uso de un pretratamiento enzimtico con lacasa y/o xilanasa, con
el fin de estudiar la evolucin de la blancura y de las propiedades colorimtricas de la pasta celulsica obtenida.

XI.

Caracterizacin de las pastas blanqueadas y de sus hojas de papel (segn


proceda) obtenidas en los puntos VIII, IX y X, determinando sus propiedades
fsicas y/o qumicas.

18

Objetivos/Aims
XII.

Evaluacin de propiedades iniciales de las fibras a usar para la produccin de


celulosa nanofibrilar (NFC): fibras de EFB y fibras de pastas de EFB obtenidas
con sosa-antraquinona, cido frmico, y cido peroxifrmico, respectivamente.

XIII.

Obtencin de celulosa nanofibrilar mediante el proceso de microfluidizacin,


y caracterizacin de propiedades fsicas, qumicas y morfolgicas de las
suspensiones acuosas y de las pelculas de celulosa nanofibrilar resultantes,
usando pastas de EFB obtenidas con sosa-antraquinona, cido frmico y cido
peroxifrmico, respectivamente.

XIV.

Evaluacin de propiedades fsicas, qumicas, mecnicas y pticas de las


pelculas de NFC obtenidas a partir de pastas de EFB.

Las innovaciones que este trabajo presenta se relacionan en los siguientes puntos:

La utilizacin de EFB en procesos de pasteado como materia prima alternativa


a las convencionales utilizadas, contribuyendo a suplir la falta de materias
primas madereras que en la actualidad se est produciendo.

Fraccionamiento hidrotrmico de EFB, con la finalidad de aprovechar la


fraccin hemicelulsica.

Estudio del uso de diferentes reactivos y disolventes para el proceso de


pasteado de EFB, incluyendo sosa y disolventes orgnicos, menos lesivos para
el medio ambiente que los mtodos clsicos convencionales que utilizan
azufre.

Eliminacin de grandes cantidades de residuos agroalimentarios, que slo


aportan costes a las industrias agroalimentarias y deterioros en los sistemas
ecolgicos (malos olores, plagas, etc.), al tiempo que se evitan deforestaciones
al utilizar una materia prima distinta de las maderas clsicas.

Disminucin de las deforestaciones y replantaciones incontroladas, que


pueden producir desequilibrios en los sistemas ecolgicos.

Se aumentar la economa de la produccin de pastas para papel, por el mejor


aprovechamiento de las materias primas mediante la consecucin de
subproductos, y las menores inversiones necesarias para este tipo de procesos
de pasteado.

19

Objetivos/Aims
El inters del trabajo es generalizado para las empresas relacionadas con la fabricacin
de pastas y de papel y para las industrias agroalimentarias, y radica principalmente en los
siguientes factores:

Investigacin de las posibilidades de utilizacin de materias primas alternativas


para la fabricacin de papel de impresin y escritura (blancos).

Contribuir al ahorro de madera (principal materia prima del papel) y


compensar su dficit en la Unin Europea (la madera ocupa el segundo lugar
en las importaciones europeas, detrs del petrleo).

Mejora de la calidad de los efluentes de los procesos de coccin de las


materias primas, al utilizar procesos libres de azufre, como los organosolv o el
proceso a la sosa.

Mejora de la calidad de los efluentes de blanqueo de las pastas, al utilizar


procesos de blanqueo totalmente libres de cloro.

Posibilidad de extrapolar los resultados obtenidos a otras materias primas


madereras y alternativas.

Obtencin de papel competitivo con los tradicionales.

Disminucin de las importaciones de pastas celulsicas y de papel, que en las


ltimas dcadas han aumentado enormemente.

Ahorro de capital para la implantacin de fbricas de pastas a escala industrial.

Menores gastos derivados de la conservacin del medio ambiente.

Aumento de la rentabilidad de las industrias agroalimentarias, al considerar el


EFB como subproducto, en lugar de como residuo.

20

Objetivos/Aims
AIMS

The general aim of this Work is the integral use of the palm oil industry residue (EFB,
i.e. bunches without fruit of oil palm), following the Biorefinery scheme, with the idea of
utilizing its own constituents.
This work will address the following points:
1. The physical and chemical characterization of the EFB, identifying its main
constituents, as well as the length and diameter of the fibers.
2. The study of the influence of the operating variables in the EFB hydrothermal
treatment, with and without a catalyst, on the yield of the solid fraction and the
composition of the liquid fraction.
3. The study of the influence of the operating variables in obtaining EFB
cellulosic pulps, using organic solvents, on the characteristics of the resulting
pulps, in order to find the optimum operating conditions.
4. The EFB pulp bleaching using totally chlorine free processes, studying the
influence of the operating variables on the characteristics of the bleached
pulps.
5. The obtaining nanofibrillar cellulose, and the determination of chemical,
physical and morphological properties of the aqueous suspensions and films of
nanofibrillar cellulose formed by using different pulps from EFB.
More specifically, the proposed work plan is as follows:
I.

The characterization of the EFB, determining solubilities in hot water and in 1


% sodium hydroxide and extractable contents of ethanol-benzene, ash,
holocellulose, -cellulose and lignin, as well as the length and diameter of the
fibers.

II.

The study of the influence of the operating variables in the EFB hydrothermal
treatment without a catalyst (temperature, time, and liquid/solid ratio) and with
a catalyst (temperature, time and sulphuric acid concentration), on the yield of
the solid fraction and the distribution of the main components in the liquid
fraction.

21

Objetivos/Aims
III.

The characterization of the liquid fraction obtained in II, determining the


contents of sugars and acetic acid, by the use of HPLC. The determination of
the yield of the solid fraction obtained in II, by a gravimetric method.

IV.

The study of the pulping process of the solid fraction of the EFB hydrothermal
treatment and the comparison of the properties of the pulps and paper sheets
with those for pulps and paper sheets obtained by EFB pulping process.

V.

The study of the influence of the operating variables to obtain EFB cellulosic
pulps with organic solvents (time and reagent concentrations), on the
characteristics of raw pulps.

VI.

The characterization of pulps obtained in V, determining their yield, viscosity,


kappa number and brightness.

VII.

The chemical characterization of liquor or black liquor from the optimum


pulping process obtained in V.

VIII.

The study of "soda-anthraquinone" and "diethanolamine" pulp bleaching of


EFB by the OZP bleaching sequence (oxygen - ozone - hydrogen peroxide).

IX.

The study of the influence of the operating variables in "soda-anthraquinone"


pulp bleaching of EFB with hydrogen peroxide, on the characteristics of
bleached pulps.

X.

The study of "soda-anthraquinone" pulp TCF bleaching of EFB, using an


enzymatic pretreatment with laccase and/or xylanase, in order to assess the
evolution of brightness and colorimetric properties of the cellulosic pulp
obtained.

XI.

The characterization of bleached pulps and their paper sheets (as appropriate)
obtained in VIII, IX and X, determining their physical and/or chemical
properties.

XII.

The evaluation of initial properties of the fibers used for the production of
nanofibrillar cellulose (NFC): EFB fibers and fibers from EFB pulps obtained
with soda-anthraquinone, formic acid, and peroxyformic acid, respectively.

XIII.

The obtaining of nanofibrillar cellulose by the microfluidization process, and


the characterization of physical, chemical and morphological properties of the
aqueous suspensions and films of nanofibrillar cellulose formed, by using EFB

22

Objetivos/Aims
pulps obtained with soda-anthraquinone, formic acid and peroxyformic acid,
respectively.
XIV.

The evaluation of the physical, chemical, mechanical and optical properties of


the films of NFC obtained from EFB pulps.

This work presents the following innovations:

The use of EFB in pulping processes as an alternative raw material to the


conventional ones used, helping to overcome the lack of wood raw materials
which is currently happening.

The EFB hydrothermal fractionation, in order to exploit the hemicellulose


fraction.

The study of the use of different reagents and solvents for the pulping process
of EFB, including soda and organic solvents, less harmful to the environment
than conventional classic methods which use sulphur.

The removal of large quantities of agri-food residues, which increase only agrifood industry costs and damage to ecological systems (odors, pests, etc.), while
deforestation is avoided by using a different raw material from classic woods.

The decreasing uncontrolled deforestation and replanting, which can cause


imbalance in ecological systems.

The improvement of the economics of paper pulp production, for the better
use of raw materials by obtaining byproducts, and lower investment required
for this type of pulping processes.

The work is of widespread interest to companies involved in the manufacture of pulps


and paper and to agri-food industries, and this interest is mainly based on the following
factors:

The investigation of the potential use of alternative raw materials for the
manufacture of printing and writing (white) paper.

The contribution of saving wood (main raw material of paper) and offset its
deficiency in the European Union (the wood is second in European imports,
after oil).

23

Objetivos/Aims

The improvement of effluent quality from cooking processes of raw materials,


using sulphur-free processes, such as the organosolv or the "soda" process.

The improvement of effluent quality from bleaching of pulps by using totally


chlorine free bleaching processes.

To have the possibility of extrapolating the results obtained to other wood raw
materials as well as alternative raw materials.

The obtaining of competitive paper with traditional one.

To decrease imports of cellulosic pulps and paper, which in recent decades


have greatly increased.

To reduce costs for future investment to implement pulp factories on an


industrial scale.

Lower environmental conservation costs.

To increase profitability of agri-food industries, considering the EFB as a


byproduct, rather than a residue.

24

Introduccin

Introduccin
3.1. MATERIALES LIGNOCELULSICOS

Los materiales de origen vegetal, que estn formados principalmente por polisacridos
(celulosa y hemicelulosas) y lignina, reciben el nombre de materiales lignocelulsicos. La
formacin de biomasa vegetal se lleva a cabo mediante la fotosntesis, que con el concurso
de la luz solar produce molculas de alto contenido energtico bajo la forma de energa
qumica.
Los materiales lignocelulsicos pueden clasificarse, atendiendo a su procedencia, en
(Caparrs, 2007):

Materiales de origen forestal, generados en los bosques, representando el 70 %


del total de los materiales lignocelulsicos.

Materiales de origen agrcola, procedentes de los residuos de las cosechas,


como pajas de cereales, podas de frutales, etc.

Materiales de origen industrial, obtenidos de desechos de algunas industrias,


como las cscaras de semillas, las mondas de frutas, el bagazo de la caa de
azcar, los racimos vacos de fruto de la palma aceitera, serrines, etc.

Materiales de origen urbano, constituidos por materiales como papeles o


cartones de composicin celulsica, que suponen ms de la mitad del peso de
las basuras.

Los materiales lignocelulsicos ms utilizados en la industria papelera son los


procedentes de las especies madereras, suponiendo el 88-92 % de las materias primas usadas
para este fin; se pueden clasificar, en funcin de sus caractersticas estructurales,
composicin qumica y propiedades mecnicas, de la siguiente manera (Roman, 2011):

Maderas frondosas o duras (hardwoods) de hoja caduca, como las


angiospermas, que pueden ser de crecimiento lento, como el roble (Quercus

robur), o de crecimiento rpido, como el eucalipto (Eucalyptus globulus).

Maderas conferas o blandas (softwoods), propias de zonas fras y templadas,


de hoja perenne, que corresponden a las gimnospermas, como el pino (Pinus

pinaster) o el abeto (Abies alba).

27

Introduccin
En la actualidad, el dficit de estas especies madereras est centrando la investigacin
en la bsqueda y aplicacin de materiales lignocelulsicos alternativos como materias primas
en la industria de la pasta celulsica y papel en diversos pases (Jimnez y col., 2002). Entre
estos materiales destacan los:

Residuos agrcolas, como pajas de cereales, podas de frutales, etc.

Residuos de industrias agroalimentarias, como el bagazo de la caa de azcar y


los racimos sin fruto de la palma aceitera.

Especies no arbreas, como el lino, yute, camo, etc.

Especies arbreas, como la paulownia, la madera de tagasaste, la madera de


podas de diferentes rboles, etc.

3.1.1. ESTRUCTURA Y COMPOSICIN DE LOS MATERIALES LIGNOCELULSICOS


Los materiales lignocelulsicos se caracterizan por la presencia de paredes celulares,
constituidas principalmente por una serie de capas coaxiales de microfibrillas de celulosa
(esqueleto) dispersas en una matriz amorfa de hemicelulosas y lignina, que en su conjunto
representan el 80-90 % del peso total (Holtzapple, 1993).
Adems de estos constituyentes mayoritarios, los materiales lignocelulsicos contienen
pequeas cantidades de otros constituyentes polimricos (peptinas, almidn y protenas)
junto a otros compuestos de bajo peso molecular (extractos y compuestos inorgnicos), que
forman parte del lumen y, por tanto, no afectan a la estructura de la pared celular (Garca,
2007). En la Figura 3.1 se presentan los componentes principales de los materiales lignocelulsicos.

3.1.1.1. Estructura de la pared celular


La composicin qumica, anatoma y propiedades fsicas, varan ampliamente entre los
diferentes materiales lignocelulsicos; siendo sus principales componentes, de gran inters
desde un punto de vista papelero, los polmeros que constituyen las paredes celulares de las
clulas vegetales; stas permanecen unidas entre s a travs de sus paredes celulares por la
lmina media, capa de pectinas de calcio y magnesio, que sirven de unin entre dichas
paredes celulares (Alen, 2000).
28

Introduccin

Figura 3.1. Componentes principales de los materiales lignocelulsicos.

Cuando estas clulas estn en perodo de crecimiento presentan una pared primaria
compuesta principalmente por microfibrillas de celulosa (9-25 %), por una matriz de hemicelulosas (25-50 %), por pectinas (10-35 %) y por protenas (10 %).
A medida que la clula va creciendo, aparece la pared secundaria, que se va lignificando desde el exterior y est compuesta por microfibrillas de celulosa dispuestas de
manera ordenada, siendo ms densa que la pared primaria.
Las clulas que sufren lignificacin lo hacen a partir de la lmina media hacia el
interior, depositndose la lignina que va remplazando al agua. La pequea cantidad de agua
que queda despus de la lignificacin puede ser desplazada por la deposicin de algunos
minerales, como carbonato clcico o slice, y taninos (Barcel y col., 2009; Garca, 2007).
Las funciones fundamentales de la pared celular son las de proteger el interior de la
clula, proporcionarle rigidez y actuar como mediadora en las relaciones de la clula con su
entorno (absorcin, transpiracin, translocacin, secrecin, etc.). En la Figura 3.2 se muestra
la estructura de la pared celular.

29

Introduccin

Figura 3.2. Estructura de la pared celular.

3.1.1.2. Componentes estructurales de la pared celular


Dentro de los componentes estructurales de la pared celular se encuentran la lignina y
los polisacridos, siendo los ms abundantes la celulosa y las hemicelulosas. En la Figura 3.3
se presenta la distribucin de los componentes mayoritarios a travs de la pared celular.

Lignina

Hemicelulosas

Celulosa

Figura 3.3. Distribucin de los componentes mayoritarios a travs de la pared celular.


30

Introduccin
3.1.1.2.1. Celulosa
La celulosa constituye la fraccin mayoritaria de los materiales lignocelulsicos, con
una proporcin del 40-50 % del peso seco en las especies madereras y del 25-40 % en los
residuos agrcolas y agroalimentarios. Es un polmero lineal cuya unidad monomrica es la
D-glucosa, que se une a travs de un enlace glucosdico en la configuracin . El enlace se
produce entre el tomo 1 de carbono de la -D-glucosa y el tomo 4 de la siguiente
molcula de glucosa (Otero, 1988). En la Figura 3.4 se muestra la estructura de la celulosa.

Figura 3.4. Estructura de la celulosa.

La celulosa tiene un grado de polimerizacin promedio de 10000 en la pared


secundaria, y en la pared primaria oscila entre 2000 y 4000 (Siau, 1984). Los enlaces
covalentes dentro y entre las unidades de glucosa dan como resultado una molcula recta y
rgida, con alta resistencia a la tensin. Adems, las molculas de celulosa tienden a formar
puentes de hidrgeno intra e intermoleculares (Figura 3.5), con un efecto profundo sobre la
morfologa, rigidez, orientacin, resistencia y reactividad de las cadenas celulsicas.

Figura 3.5. Modelo de molculas de celulosa unidas por puentes de hidrgeno.


31

Introduccin
La celulosa es insoluble en la mayora de los disolventes, incluyendo lcalis fuertes, y
es difcil aislarla de los materiales lignocelulsicos en forma pura, debido a que se encuentra
ntimamente ligada con la lignina y las hemicelulosas (Pettersen, 1984). En la estructura de la
celulosa se alternan zonas ordenadas, cristalinas, con zonas ms desordenadas, amorfas. Las
zonas que presentan cristalinidad elevada, que constituyen el 55-75 % de la fibra de celulosa,
son ms estables, difciles de penetrar por disolventes y reactivos, comunican rigidez y
presentan estabilidad dimensional y elevada resistencia a la solvatacin, a la traccin y al
alargamiento. Por el contrario, las zonas amorfas son ms accesibles y ms susceptibles a
todas las reacciones qumicas, y favorecen el hinchamiento, el alargamiento y la flexibilidad
de la fibra. Las dimensiones de las microfibrillas no son uniformes, sino que varan segn su
origen y su posicin dentro de la pared celular; las microfibrillas, en combinacin con los
restantes materiales de la matriz, suministran la rigidez y la resistencia necesarias para que los
vegetales crezcan erguidos (Garca, 2007) (Figura 3.6).

A. Vista longitudinal/transversal de varias fibras en el


tejido maderero.
B. Porcin de la capa S2, que muestra macrofibrillas.

C. Porcin de una macrofibrilla, compuesta de agregados


de microfibrillas incrustadas en un material no celulsico.
D. Seccin transversal de una microfibrilla individual
compuesta de agregados de molculas de celulosa.
E. Regin de una microfibrilla donde las cadenas
celulsicas muestran un alto orden.
F. Organizacin de las molculas de celulosa. Sabiendo
que la unidad bsica que se repite es la celobiosa.
G. Unidad de celobiosa: dos unidades de glucosa
conectadas por un tomo de oxgeno.

Figura 3.6. Representacin esquemtica de la estructura microscpica y submicroscpica de la


celulosa.

3.1.1.2.2. Hemicelulosas
Las hemicelulosas son polisacridos ms complejos que la celulosa, qumicamente
heterogneos, constituidos por combinaciones de monosacridos ramificados, amorfos y de
32

Introduccin
menor grado de polimerizacin (150-200); entre ellos destacan las pentosas (D-xilosa, Larabinosa), las hexosas (D-glucosa, D-manosa y D-galactosa), los cidos urnicos (cido Dglucournico, cido 4-O-metil-D-glucournico y cido D-galactournico) y las 6-deoxi
hexosas (L-ramnosa y L-fucosa) (Figura 3.7).
En las maderas conferas abundan los hexosanos (manosa y galactosa), predominando
los galactoglucomananos y los glucomananos, polmeros lineales de glucosa y manosa unidos
por enlaces -1,4, con predominio de las manosas, que no presentan ramificaciones ni
grupos sustituyentes laterales.
En las maderas frondosas predominan los pentosanos, como la xilosa, donde se
observan glucuronoxilanos (Sjstrm, 1993), aunque tambin glucomananos, que poseen
mayor cantidad de manosa que los de las maderas conferas, presentando dos tipos de
sustituyentes: los grupos O-acetil y la galactosa, por lo que stos tambin se denominan
galactomananos (Fengel y Wegener, 1989).
Las hemicelulosas actan como matriz de soporte para las microfibrillas de celulosa, y
constituyen el 25-35 % de ella, si bien este porcentaje vara sensiblemente entre las especies
vegetales.
Las hemicelulosas, con estructura amorfa y ramificada, ejercen de almacn de
sustancias de reserva en la pared celular, teniendo una funcin reguladora, estructural y, por
tanto, de control de la expansin celular. Adems, facilitan las uniones qumicas entre la
celulosa y la lignina, al unirse formando puentes de hidrgeno con los grupos -CH2OH de

las cadenas de celulosa y los oxgenos glucosdicos de las hemicelulosas.

3.1.1.2.3. Lignina
La lignina es el polmero ms abundante en las especies vegetales, detrs de la
celulosa, estando presente en un 25-33 % en las maderas conferas y en un 18-34 % en las
maderas frondosas (Aitken y col., 1988). Aunque la concentracin de lignina es elevada en la
lmina media (80 %), la mayor parte de la lignina de la madera (60-75 %) se encuentra en la
pared celular.
La lignina se define como un heteropolmero aromtico tridimensional con estructura
irregular formado por unidades de fenilpropano, unidas por diferentes tipos de enlaces, que
varan entre las especies vegetales, situacin en la planta, naturaleza, edad y condiciones de
crecimiento de las clulas (Fergus y Goring, 1970) (Figura 3.8).
33

Introduccin

Figura 3.7. Estructura de las hemicelulosas y sus derivados.

Los monmeros que la conforman son los denominados alcoholes p-hidroxicinamlicos, que incluyen el alcohol p-cumarlico que da lugar a las unidades de phidroxifenilo (unidades H) y no presenta ningn sustituyente; el alcohol coniferlico que da
lugar a las unidades guayacilo (unidades G), presentando un grupo metoxilo en la posicin 3
del anillo aromtico; y el alcohol sinaplico que da lugar a las unidades siringilo (unidades S),
presentando dos grupos metoxilo en las posiciones 3 y 5 del anillo (Figura 3.9).

34

Introduccin

Figura 3.8. Estructura de la lignina.

La lignina de frondosas est constituida principalmente por unidades guayacilo (56 %),
que derivan del alcohol coniferlico, y por unidades siringilo (40 %) que derivan del alcohol
sinaplico (Lin y Dence, 1992), mientras que la lignina de conferas est constituida por
unidades guayacilo (80 %) (Jung y Fahey, 1983). Las unidades guayacilo (G) tienen un nico
grupo metoxilo y la posicin C-5 est libre, pudiendo formar enlaces carbono-carbono; por
tanto, aquellas ligninas con mayor cantidad de unidades G tienen una estructura que se
degrada con mayor dificultad. Las hemicelulosas estn asociadas a la lignina a travs de las
unidades de arabinosa, xilosa y galactosa, por enlaces de tipo glucosdico, ter benclico y
ster benclico, formando complejos. La lignina es la sustancia que le confiere rigidez a las
paredes celulares, ofreciendo proteccin a la madera contra la degradacin microbiana y
permitiendo reducir el nivel de higroscopicidad (Siau, 1984).

Figura 3.9. Unidades estructurales de la lignina.


35

Introduccin
3.1.1.3. Componentes no estructurales de la pared celular
Estos componentes representan un 0-10 %, aunque puede llegar en algunos casos al
40 % en peso seco de los materiales lignocelulsicos. Estos componentes no influyen en la
estructura morfolgica de las clulas, pero tienen influencia en el procesado de los
materiales lignocelulsicos (Caparrs, 2007; Roman, 2011), siendo los ms importantes los
siguientes:
Agua. Procede de los fluidos biolgicos de las plantas, del carcter
higroscpico de los materiales lignocelulsicos, as como de las condiciones
ambientales. Representa un 0,08-0,10 % del peso seco de los materiales lignocelulsicos.
Cenizas. Suelen ser sales inorgnicas como cloruros, sulfatos, silicatos,
compuestos de calcio, potasio y magnesio, y en menor medida de fsforo,
manganeso, hierro, cobre, zinc, sodio o cloro. Representan en torno a 0-15 %
del peso seco de los materiales lignocelulsicos.
Protenas. Suelen estar presentes en pequeas cantidades, siendo las
principales las extensinas, lectinas, protenas arabino-galactano, protenas ricas
en glicina y protenas ricas en prolina.
Extrables. Son restos de los fluidos biolgicos de los vegetales de composicin
heterognea, que se extraen fcilmente con distintos disolventes orgnicos.
Sustancias pcticas. Son polmeros ricos en cido D-galactournico con una
cadena lineal altamente ramificada, siendo los principales el homogalactouronano, los ramnogalactouronano I y II y el arabinogalactano. Suelen
estar presentes en porcentajes menores al 1 %.

3.1.2. CULTIVOS LIGNOCELULSICOS


Los cultivos lignocelulsicos incluyen especies tanto de origen forestal como agrcola y
poseen un gran inters industrial. Entre los principales usos de los cultivos lignocelulsicos
se encuentra la produccin de pasta de celulosa. Por otro lado, estos cultivos presentan un
gran potencial como materia prima en el contexto de las futuras biorefineras para la
produccin de biocombustibles y otros productos de inters, como alternativa al petrleo.

36

Introduccin
La principal fuente de fibra de celulosa virgen utilizada actualmente en la fabricacin
de pasta de celulosa la constituyen los cultivos de fibras madereras, mientras que las fibras
alternativas o no madereras se utilizan en menor proporcin.
La amplia disponibilidad y concentracin de madera en zonas de fcil acceso, su
elevado contenido en fibras, el bajo coste de manejo y transporte y la facilidad de
almacenamiento, as como su estabilidad y comportamiento durante el proceso de obtencin de celulosa, han sido los desencadenantes de su gran utilizacin en la industria de la
pasta y papel. Sin embargo, existe en la actualidad un renovado inters por el uso de
materias primas alternativas, debido, entre otras razones, a la gran disponibilidad de algunas
de ellas, como los residuos agrcolas y los de industrias agroalimentarias. Estos materiales
alternativos constituyen una fuente abundante de fibras en determinadas zonas geogrficas,
principalmente en pases en vas de desarrollo. La gran variedad de caractersticas,
dimensiones de sus fibras y composicin qumica de estos materiales les confieren un gran
potencial como materias primas (Garca, 2007).

3.1.2.1. Fibras madereras


Las fibras madereras provienen de especies vegetales que desarrollan un tronco donde
se acumulan preferentemente las mejores fibras. Las conferas constituyen el primer cultivo
forestal a escala mundial para la obtencin de pasta de papel, seguidas de cerca por las
frondosas.
Las conferas presentan fibras largas (de 3 a 5 mm) y son ptimas para la fabricacin
de papeles de elevada resistencia mecnica. En trminos econmicos generales son ms
valiosas que las frondosas, ya que sus troncos son ms largos y rectos, su madera es
uniforme, ligera y blanda, por lo que es ms fcil de trabajar, y presentan una mayor
proporcin de elementos fibrosos adecuados para la mayora de las calidades papeleras
(Garca, 2007; Lpez y col., 2010b). Las principales conferas usadas para la fabricacin de
pasta de papel son el pino y la pcea.
La madera de frondosas es ms dura, con fibras cortas (entre 0,75 y 2 mm) y dan lugar
a pastas menos uniformes que las de conferas. El papel fabricado con maderas frondosas es
ms dbil que el de conferas, pero su superficie es ms lisa y, por lo tanto, es mejor para
papel de escritura. Otra de las ventajas es que el crecimiento de las especies de frondosas es

37

Introduccin
ms rpido que el de las conferas. Las principales frondosas utilizadas en el sector papelero
son el eucalipto, el chopo y el abedul (Lpez y col., 2010b).

3.1.2.2. Fibras alternativas a las madereras


Las fibras alternativas constituyen una excelente materia prima para la produccin de
pasta de celulosa. Uno de los hechos apremiantes que conducen a la utilizacin de estas
materias primas es la cada vez mayor produccin de pasta y papel, que da lugar a la escasez
de materias primas convencionales y a la desforestacin de los suelos con los problemas
ecolgicos que ello conlleva (Lpez y col., 2010b).
En general, las fibras alternativas tienen una estructura menos densa y ms porosa, lo
que implica menores requerimientos de energa y de productos qumicos para la separacin
de las fibras durante la produccin de pasta de papel. Adems, presentan ciclos de
crecimiento ms cortos, alcanzando la madurez ms rpidamente que las especies madereras
y en muchos casos los rendimientos en materia prima y en pasta obtenidos son mayores
(Tabla 3.1).
Algunas pastas de fibras largas alternativas tienen propiedades superiores a las mejores
pastas de conferas, pues son extremadamente resistentes. El principal inconveniente de este
tipo de materias primas es que la mayora de ellas slo estn disponibles en ciertas pocas
del ao.
Las fibras alternativas se pueden clasificar en tres categoras:

Fibras procedentes de tallos de diversas plantas, como lino, camo, kenaf y


yute, y de hojas, como abac y sisal.

Fibras de residuos agrcolas y agroalimentarios, como la paja de cereales, tallos


de maz, paja de arroz, bagazo de la caa de azcar o los racimos vacos de
fruto de la palma aceitera.

Fibras de vegetales silvestres, como bamb o hierba elefante.

Actualmente, todos estos tipos de fibras representan una alternativa para la produccin
de pasta de celulosa en pases con baja disponibilidad de madera y en los que abundan estos
materiales alternativos. As, el uso de estas fibras para la produccin de pasta de celulosa ha
ido aumentando, especialmente en los pases en vas de desarrollo, como India, China y
algunos pases latinoamericanos.
38

Introduccin
Especies

Rendimiento materia seca (t/ha)

Rendimiento Pasta (t/ha)

Trigo

2,5

1,1

Avena

1,6

0,7

Centeno

2,2

1,1

Arroz

3,0

1,2

Caa de azcar (bagazo)

9,0

4,2

Bamb

4,0

1,6

Miscanthus sinensis

12,0

5,7

Canary grass

6,0

3,0

Caa comn

9,0

4,3

Kenaf

15,0

6,5

Camo

12,0

6,7

Abedul

3,4

1,7

Eucalipto

15,0

7,4

Confera escandinava

1,5

0,7

Confera de crecimiento rpido

8,6

4,0

Tabla 3.1. Rendimiento promedio de algunas materias primas.

En los pases desarrollados, las fibras alternativas se usan principalmente para la


produccin de papeles especiales. En Espaa existen varias empresas que fabrican pasta de
papel a partir de este tipo de fibras, destacando la empresa CELESA, que utiliza fibras
liberianas (del tallo) de lino, camo y yute, y fibras de hojas de sisal y abac para fabricar
pasta de celulosa para papeles especiales de distintas caractersticas, tales como papeles para
cigarrillos, filtros especiales o papeles dielctricos; y la empresa Ecopapel, que produce pasta
de paja de cereales para la fabricacin de diferentes tipos de envases (cajas para frutas,
cartones de huevos, recipientes sanitarios, bandejas para diferentes alimentos, etc.).

3.1.2.2.1. Materias primas alternativas para la produccin de pastas


Desde hace poco ms de un siglo, el empleo de madera para la fabricacin de papel
ha permitido que las cantidades elaboradas de este producto crecieran progresivamente con
39

Introduccin
el transcurso de los aos, llegando a un consumo de madera de la misma magnitud que el
de petrleo (Jimnez, 2005; Lpez y col., 2010b). Ello ha conducido a un problema de
abastecimiento de madera, que se agudiza con el tiempo.
Por esta razn, muchas de las investigaciones llevadas a cabo en los ltimos tiempos se
han centrado en la bsqueda de nuevas materias primas, para evitar, de esa forma, las
deforestaciones incontroladas que presentan problemas ecolgicos serios en los ecosistemas.
De esta manera se ha llegado al estudio de diversos materiales, como los residuos agrcolas y
agroalimentarios y vegetales alternativos a los conseguidos en cultivos agroalimentarios.
Desde la dcada de los setenta del pasado siglo, la produccin de pastas de materias
primas alternativas ha pasado desde aproximadamente un 7 % a casi un 12 % del total de la
pasta producida, creciendo a un ritmo 2-3 veces superior al de las pastas de madera (Alaejos
y col., 2004; Eroglu y Tutus, 2004; Jimnez, 2005; Lpez y col., 2010b; Simula, 2002).
La utilizacin de residuos agrcolas y agroalimentarios y de vegetales de cultivos
alternativos a los agroalimentarios, parece una buena alternativa a las materias primas
madereras, ya que pueden dar lugar a excelentes papeles de propiedades especiales, y a
veces constituyen la nica fuente de materias primas en algunas zonas geogrficas.
Por otra parte, es sabido que los consumidores estn cada vez ms interesados en
disponer de papeles obtenidos mediante tecnologas limpias, o en los procedentes de fibras
recicladas o de materiales alternativos; as la utilizacin de residuos agrcolas y agroalimentarios y de vegetales alternativos podra contribuir a preservar y mantener el medio
ambiente, al conseguir disminuir la gran cantidad de madera utilizada como materia prima
en la produccin de pastas para papel (Jimnez, 2005; Lpez y col., 2010b).
Por otra parte, dado que el consumo de papel va paralelo al nivel de vida de los pases,
que va creciendo en todo el orbe, es previsible que este aumento, en mayor o menor grado,
se mantenga en el futuro.
El uso de materiales lignocelulsicos alternativos para la produccin de pasta y papel,
viene avalado por el consumo creciente de estos productos; en efecto, a pesar de que el
desarrollo de Internet y las nuevas tecnologas de informacin pronosticaban un descenso en
el consumo de papel, el crecimiento de ste fue mayor del 10 % en los ltimos veinte aos
del pasado siglo (Lichtenthaler y Peters, 2004). El 30 % del papel que se consume en la
actualidad se emplea en funciones asociadas a nuevas tendencias de la informacin que no
existan hace diez aos (Rodrguez y col., 2009). Adems, mientras que el 90 % de las
40

Introduccin
materias primas para la produccin de papel lo constituyen las maderas de frondosas y
conferas, en los ltimos 10 aos el incremento en la produccin mundial de pasta de
madera ha sido aproximadamente del 9 %, en contraste con el 23 % o ms que aument la
produccin de pastas de materiales alternativos para la fabricacin de papel (Anuario FAO,
2012).
En los ltimos aos ha habido una mayor conciencia en el uso de materias primas
alternativas, lo que permite suplir la escasez de materias primas madereras, as como evitar
las desforestaciones y replantaciones incontroladas, que dan lugar a graves problemas
ecolgicos y al cambio climtico.
La produccin pastero-papelera absorbe una gran parte de la creciente explotacin
maderera mundial. En 1983, alrededor del 35 % de la produccin comercial de madera en
todo el mundo se destin a fabricar papel o productos anlogos, y este porcentaje creci
hasta el 50 % en el ao 2010. En los Estados Unidos, la presin ecologista para la reduccin
de las talas de rboles podra impulsar un fuerte desarrollo de materias primas alternativas
en la fabricacin de pasta y papel (Jimnez y Rodrguez, 2009; Lpez y col., 2010b).
En Espaa, la principal materia prima para la industria celulsica es el eucalipto, pero
hoy en da la opinin pblica y, ms en concreto, las corrientes ecologistas no ven con muy
buenos ojos las plantaciones de estas especies de crecimiento rpido.
Todo ello, junto a otros factores, ha llevado a la disminucin de la superficie ocupada
por plantaciones de eucaliptos y, consecuentemente a la menor produccin de madera de
esta especie. Este problema se est subsanando, al menos temporalmente, con el
espectacular incremento de las importaciones de madera procedente de pases
sudamericanos. Frente a este panorama, una alternativa posible est en el incremento de
produccin de pasta a partir de materiales no convencionales o alternativos.
Dentro de estas fuentes de materias primas no convencionales se incluyen las materias
primas no leosas, tanto procedentes de plantas silvestres como de plantaciones, as como
tambin los residuos agrcolas generados en labores agrcolas y forestales, y los residuos
industriales de carcter lignocelulsico, como los procedentes de industrias agroalimentarias
o de las industrias de transformacin de la madera.
En la actualidad, las fibras alternativas representan slo una pequea parte de la
materia prima utilizada en la fabricacin de papel y cartn a escala mundial. En algunos

41

Introduccin
pases, sin embargo, estas fibras se utilizan ampliamente, llegando a representar importantes
porcentajes respecto a su produccin total de pasta celulsica.
Las regiones ms productoras de pastas celulsicas a partir de materiales alternativos
son las de Asia y el Pacfico. En particular, China e India. Por el contrario, en Amrica del
Norte, Amrica Latina, Europa y frica el uso de fibras alternativas ha sido relativamente
limitado.
Segn algunos estudios realizados por Jimnez (2005), Lpez y col. (2010b) y Phillips
(1998) es previsible que en los prximos aos se siga produciendo un incremento en el
consumo mundial de pasta de papel procedente de especies vegetales distintas a las
madereras.

3.1.2.2.2. Caractersticas generales de las materias primas alternativas


Los vegetales alternativos tienen menos lignina que las maderas convencionales, por lo
que suelen ser ms fcil deslignificarlos. En cuanto al tamao de sus fibras, en la Tabla 3.2
se presentan las dimensiones de fibras de diferentes especies alternativas, comparndolas
con las de algunas especies de madera (Atchison y McGovern, 1993).
Los datos muestran la gran variabilidad de las caractersticas de estas fibras. Muchas de
ellas son similares a las fibras cortas de maderas frondosas, mientras que otras son tan largas
que es conveniente acortarlas para que sirvan mejor para producir papel.
En general, el dimetro de las fibras alternativas es pequeo, siendo la textura de la
pasta resultante menos densa. Las dimensiones de la fibra dan una idea de su buena utilidad
potencial para la fabricacin de pasta y papel. De hecho, desde el punto de vista tcnico,
puede producirse cualquier calidad de papel utilizando una combinacin adecuada de fibras
vegetales alternativas.

42

Introduccin
Fibras madereras convencionales

Longitud media
(mm)

Dimetro medio
(micras)

Conferas de zona templada

2,7-4,6

32-43

Frondosas de zona templada

0,7-1,6

20-40

Frondosas tropicales diversas

0,7-3,0

20-40

Eucalipto

0,7-1,3

20-30

Longitud media
(mm)

Dimetro medio
(micras)

Abac (camo de Manila)

6,0

24

Bagazo de caa de azcar (desmedulado)

1,0-1,5

20

Bamb

2,7-4,0

15

Tallos de maz y de sorgo (desmedulado)

1,0-1,5

20

Fibra de algodn

25

20

Tallos de algodn

0,6-0,8

20-30

Crotalario (camo sun)

3,7

25

Esparto

1,5

12

Paja de lino

30

20

Camo

20

22

Yute

2,5

20

Fibra blanda de kenaf

2,6

20

Fibra medular de kenaf

0,6

30

Carrizo

1,0-1,8

10-20

Paja de arroz

0,5-1,0

8-10

Cardo

1,3

19

Sisal

3,0

20

Paja de trigo

1,5

15

Fibras alternativas o no madereras

Tabla 3.2. Dimensiones de las fibras de plantas no leosas.

43

Introduccin
3.2. BIOREFINERA DE MATERIALES LIGNOCELULSICOS

El carcter no renovable de los combustibles fsiles hace insostenible el modelo de


desarrollo econmico actual. Por esta razn y ante la necesidad de fuentes de energa
renovables, nace el desarrollo de la denominada Biorefinera, que sigue un modelo similar
al de una refinera petrolfera. Se basa en la utilizacin de materiales lignocelulsicos como
materia prima de carcter renovable para obtener combustibles y diversos productos
qumicos. En la Figura 3.10 se presenta una esquematizacin del concepto de Biorefinera.
El origen del concepto de Biorefinera es relativamente reciente; data del ao 1981 y
fue acuado en la dcada de los 90 por C. A. Abbas, de la Archers Daniel Midlan Company
(Illinois), como un neologismo en el que el concepto de Forest Refinery fue definido
como aquellos materiales lignocelulsicos en forma de madera que pueden ser
fraccionados y convertidos en otros productos de valor aadido (Myerly y col., 1981). Cabe
mencionar que este trmino surge de la explotacin y fraccionamiento de cereales, maz y
soja.

Figura 3.10. Esquematizacin del concepto de Biorefinera.

Segn el Departamento de Energa de EE.UU., una biorefinera es un concepto


general de una planta industrial de procesado de biomasa lignocelulsica, donde se
convierte y se extrae un amplio espectro de productos de alto valor aadido. El Laboratorio
de Energa Renovable Nacional de EE.UU. la define como una instalacin que integra los
44

Introduccin
procesos de conversin de biomasa para producir combustibles, energa y productos
qumicos. El concepto de biorefinera de la biomasa es similar al de las refineras de
petrleo de hoy, que producen a partir de l mltiples combustibles y productos derivados.
El Centro de Investigacin Energtica de los Pases Bajos (Energy Research Center of the
Netherlands, ECN) define la biorefinera como aquellas instalaciones en las que tiene lugar
el fraccionamiento de los materiales lignocelulsicos en diferentes componentes, que
pueden dirigirse al mercado directamente o ser transformados previamente mediante
diferentes tratamientos.
Las biorefineras de aceites vegetales (Cherubini, 2010) y de polisacridos de la
biomasa (Kamm y Kamm, 2004) son las ms utilizadas segn su grado de integracin y de la
mayor o menor produccin de biocombustibles, bioproductos y productos intermedios para
la sntesis de otros derivados; se clasifican en biorefineras de primera, segunda y tercera
generacin (Kamm y col., 2006). La situacin actual de las biorefineras es muy variada.
Mientras que EE.UU. es el nico pas que ha implantado una poltica integral I+D+i
relacionada con el desarrollo de una bioeconoma, en la Unin Europea la utilizacin de la
biomasa est exclusivamente centrada en la obtencin de energa y biocombustibles.
Actualmente, en EE.UU., se producen unas 100 millones de toneladas de productos
qumicos, entre productos qumicos de base, compuestos intermedios, especialidades y
productos de qumica fina. De todos ellos, slo el 10 % proceden de la biomasa (Office of
the Biomass Program, 2005). Esta situacin puede extenderse en mayor o menor grado,
pero con magnitudes de similar orden porcentual, al resto de los pases industrializados. Los
compuestos qumicos orgnicos representan el mercado ms directo y mayor para los
bioproductos basados en unidades bsicas similares (building blocks, productos qumicos
de base) (Lovins, 2004; Office of the Biomass Program, 2005; Werpy y Petersen, 2004). Por
lo tanto, existe un margen muy amplio para, mediante un apropiado programa de desarrollo
socio-econmico-tecnolgico, desplazar parcialmente al petrleo de nuestra economa, en
una cantidad porcentualmente significativa, como materia prima para la obtencin de
productos qumicos, sustituyndolo por la biomasa.
En este sentido, en los ltimos 10 aos, en algunos pases, como EE.UU.
(http://www1.eere.energy.gov/biomass/pdfs/final_2006_vision.pdf;

http://www.bioproducts-

bioenergy.gov/pdfs/FinalBiomassRoadmap.pdf; Lovins, 2004; Office of the Biomass Program, 2005; Werpy y Petersen, 2004), Canad (Archambault, 2004) y Alemania (Nitsch y
col., 2000; Pflaum y Rettweiler, 2005), se han incrementado exponencialmente los estudios,
45

Introduccin
iniciativas y proyectos tecnolgicos relacionados con el uso de biomasa como materia prima,
tanto para la produccin de bioenerga como de bioproductos.
En EE.UU. se prev que para el ao 2020 se incrementen al menos en un 25 % las
materias primas basadas en compuestos qumicos orgnicos a base de carbono y en un 10 %
el uso de combustibles lquidos (en comparacin con los niveles de 1994), que sern
producidos por mtodos biotecnolgicos. Esto significa que ms del 90 % del consumo de
los productos qumicos orgnicos en los EE.UU. y hasta el 50 % de las necesidades de
combustible lquido seran productos de base biotecnolgica (National Research Council,
2000). El comit Tcnico Asesor de la Biomasa (BTAC) de los EE.UU. (en el que los
principales representantes de las empresas industriales, como Dow Chemical, El Du Pont de
Nemours, LLC Dow Cargill y Genencor Internacional, y los productores de maz y las
asociaciones de defensa de Recursos Naturales del Consejo estn involucrados), que acta
como asesor del gobierno de los EE.UU., ha hecho un plan detallado con pasos hacia los
objetivos de 2030 en lo que respecta a la bioenerga, los biocombustibles y bioproductos
(Tabla 3.3) (Pflaum y Rettweiler, 2005).

Cuota de produccin a partir de biomasa

2010

2020

2030

Generacin de energa

4%

5%

5%

Biocombustibles para transporte

4%

10 %

20 %

Bioproductos

12 %

18 %

25 %

Tabla 3.3. Perspectivas de futuro para las tecnologas biomsicas.

Los estados miembros de la UE se han unido para definir las directrices nacionales
(Directiva 2003/30/EC) hacia una cantidad mnima de biocarburantes y otros combustibles
renovables (con un valor de referencia de 2 % para 2005 y 5,75 % para el ao 2010,
calculada sobre la base del contenido energtico de toda la gasolina y el gasleo para el
transporte). En la Tabla 3.4 se resumen estos objetivos de la UE en materia de
establecimiento de energas renovables y biocombustibles (Kamm y Kamm, 2007; Nitsch y
col., 2000). Como se observa, en la actualidad no existen directrices para productos de base
biotecnolgica en la Unin Europea. La Directiva de los biocombustibles ya incluye
etanol, metanol, ter metlico, hidrgeno y la pirolisis de la biomasa que son las lneas
fundamentales de productos de base biolgica del futuro de la industria qumica.

46

Introduccin
Ao

2010

2020

2030

Bioenerga

12,50

20

26

Biocombustibles

5,75

20

20

Productos qumicos de la biomasa

8-10

Tabla 3.4. Objetivos mnimos de la UE con respecto a la obtencin de bioenerga y bioproductos.

Los datos indican la cuota (%) sobre el total de la demanda de energa, combustibles
para transporte y productos qumicos. Por lo tanto, actualmente en Europa, al contrario que
en EE.UU., no existe ninguna poltica especficamente dirigida al desarrollo de bioproductos
diferentes de los biocombustibles, ya que todava no se han fijado objetivos concretos, ni una
agenda de trabajo a largo plazo. Sin embargo, se ha establecido unas perspectivas de
mercado en las que se incluyen tambin a los biocombustibles (Commission of the european
communties, 2007). En el ao 2005 las perspectivas de mercado en la Unin Europea para
los productos que se obtienen de la biomasa fueron de 77 billones de euros, de los cuales el
7 % fue de la industria qumica. En el ao 2010 las perspectivas fueron de 125 millones de
euros, de los que el 10 % fue de la industria qumica.
De manera general se puede concluir que el objetivo primordial de una biorefinera se
basa en el uso de la biomasa o materiales lignocelulsicos como materia prima para producir
multitud de productos mediante procesos basados en la combinacin de biotecnologa y
tecnologa qumica. La biomasa vegetal utilizada puede ser de origen agrcola o forestal.
Ante el panorama mundial actual, que presenta escasez de recursos y energa, as
como una elevada problemtica medioambiental, se propone, como una fuente sostenible
de materias primas, la utilizacin de materiales lignocelulsicos, siendo sus tres
componentes qumicos principales (celulosa, hemicelulosas y lignina) susceptibles de ser
separados siguiendo este esquema de biorefinera.
El esquema de biorefinera se clasifica segn la materia prima utilizada en (Kamm y
Kamm, 2007):

Biorefinera lignocelulsica, que utiliza materias primas procedentes de las


maderas y de los residuos agroindustriales y agrcolas.

Biorefinera de cultivo integral, que emplea como materia prima cereales


como el maz.
47

Introduccin

Biorefinera en verde, basada en la utilizacin de cultivos vegetales


inmaduros, como la hierba o los tallos de los cereales.

Biorefinera de dos plataformas, una para el procesado de azcares y otra para


el gas de sntesis.

En la actualidad, cobra especial importancia la biorefinera de materiales


lignocelulsicos, tambin denominada de segunda generacin, al usar biomasa vegetal o
materiales lignocelulsicos de bajo coste y pequea importancia comercial. Al emplear
como materia prima la biomasa vegetal se pretende aprovechar todas las fracciones que
procedan de la misma, obteniendo combustibles (como bioetanol o biodiesel) y diversos
productos qumicos de inters comercial (productos para cosmtica, alimentacin, farmacia,
polifenoles, etc.), producindose una valorizacin mxima de la biomasa vegetal y una
minimizacin en la produccin de residuos (Figura 3.11). Este tipo de biorefinera, basada
en el aprovechamiento de los materiales lignocelulsicos, se propone como una tecnologa
interesante de gran aplicacin en el futuro, ya que adems del bajo coste de adquisicin de
las materias primas, ofrece la produccin de una amplia variedad de productos.
Actualmente, las dificultades que se presentan en este tipo de tecnologa son debidas a
inconvenientes tecnolgicos y al deficiente aprovechamiento de la lignina.

Figura 3.11. Obtencin de productos de alto valor aadido a partir de una Biorefinera de materiales
lignocelulsicos.
48

Introduccin
3.2.1. TRATAMIENTO HIDROTRMICO
Tradicionalmente, la fraccin ms abundante de los materiales lignocelulsicos y en la
que se han centrado todos los esfuerzos de investigacin ha sido la celulosa,
desaprovechndose dos fracciones muy importantes presentes en los materiales
lignocelulsicos, como son las hemicelulosas y la lignina.

Figura 3.12. Esquematizacin del tratamiento hidrotrmico.

Esto sugiere la conveniencia de integrar lneas de produccin de subproductos en la


fabricacin de pasta celulsica, al objeto de aprovechar las fracciones constituidas por
hemicelulosas y lignina.
Diferentes procesos que se enmarcan dentro del concepto del aprovechamiento
fraccionado de los materiales lignocelulsicos han sido desarrollados, fundamentalmente
para maderas de conferas y frondosas (Jimnez, 2005; Rijkens, 1984). Sin embargo, en el
caso de los materiales alternativos a los convencionales para la fabricacin de pasta para
papel, como los procedentes de las actividades agrcolas y agroindustriales, el
fraccionamiento adquiere especial inters por cuanto supondra un beneficio ecolgico al
eliminar fuentes de contaminacin, al tiempo que los dotara de un valor aadido que no
tienen como tales residuos. As, tras unas operaciones de acondicionamiento de las materias
primas (descortezado, desmedulado, astillado, etc.), un modo de alcanzar el aprovechamiento integral de las mismas puede consistir en la utilizacin de tratamientos
hidrotrmicos o de autohidrlisis, que permiten separar las hemicelulosas del resto de los
componentes de las materias primas, para obtener productos de alto valor aadido (aditivos
alimentarios, frmacos, etanol, xilitol, furfural, etc.) (Figura 3.12).

49

Introduccin
Posteriormente, la fraccin slida procedente del tratamiento hidrotrmico se somete
a procesos de pasteado adecuados, con la finalidad de separar la celulosa (empleada para la
fabricacin de papel) de la lignina, que puede emplearse para la produccin de diversos
productos con elevado valor comercial, como resinas, poliuretanos, acrilatos, epxidos,
composites, etc.
En los tratamientos hidrotrmicos, la biomasa lignocelulsica es tratada con agua a
elevada temperatura, entre 150 y 250 C, sufriendo las hemicelulosas reacciones de hidrlisis
que son catalizadas por los protones generados en la autoionizacin del agua. Los grupos
acetilo presentes en las hemicelulosas en forma de steres (xilanos en los residuos agrcolas y
agroalimentarios) son degradados, liberndose stos en forma de cido actico (McGinnis y
Shafizadeh, 1990; Rijkens, 1984).
En lo que sigue se estudia la composicin de las hemicelulosas y el proceso
hidrotrmico mediante el cual se despolimerizan, atendiendo a los fundamentos de la
autohidrlisis que tiene lugar y a las condiciones de operacin requeridas. Se terminar con
una exposicin de la utilizacin comercial de los productos de la despolimerizacin de las
hemicelulosas.

3.2.1.1. Composicin de las hemicelulosas


Las hemicelulosas, trmino utilizado para aludir a los polisacridos de la pared celular
extrables con lcalis acuosos, incluyen a todos los polisacridos estructurales salvo la
celulosa (Holtzapple, 1993; Rojas, 1997). Su funcin en la pared celular es servir de puente
de unin entre la celulosa y la lignina. Las hemicelulosas se encuentran en proporciones del
20 al 30 % variando mucho de un material a otro, siendo mayor en los materiales
alternativos que en los madereros que presentan mayores contenidos de celulosa y lignina.
Las caractersticas principales de las hemicelulosas, por las que pueden ser separadas
del resto de los componentes, son su alta velocidad de hidrlisis en medio cido, as como
su elevada solubilidad en lcalis; estas caractersticas las distinguen de la celulosa y de la
lignina, que destaca por su resistencia a la oxidacin. Pero tambin existen otras diferencias
importantes entre las hemicelulosas y la celulosa, como el grado de polimerizacin (200-300
en las hemicelulosas frente a los 500-1500 de la celulosa); la heterogeneidad (las hemicelulosas son unos heteropolmeros formados por distintos azcares, mientras que la
celulosa est constituida por homopolmeros formados por unidades de glucosa); la
50

Introduccin
estructura amorfa de las hemicelulosas frente a la alta cristalinidad de la celulosa; y la
presencia de ramificaciones, ya que mientras que la celulosa es un polmero lineal las
hemicelulosas son polmeros ramificados.
Los distintos heteropolmeros que estn presentes en la estructura de las
hemicelulosas reciben sus nombres de los monmeros mayoritarios que los constituyen
(xilosa, manosa, galactosa, arabinosa, grupos acetilo y cidos urnicos). Para el caso de
maderas conferas, los heteropolmeros predominantes son los mananos, mientras los
xilanos representan el grupo mayoritario en las maderas frondosas (Jimnez, 2005). En lo
que sigue se consideran los principales heteropolmeros de las hemicelulosas.

3.2.1.1.1. Xilanos
Despus de la celulosa, los xilanos son los polisacridos ms abundantes en los
vegetales. Los xilanos son los heteropolmeros mayoritarios en las maderas de frondosas y
en los materiales alternativos, mientras que en las maderas de conferas su presencia es
menor, aunque en porcentajes apreciables. En las maderas de conferas aparece como
arabinofuranosa, que estabilizan la molcula frente a las bases. Su grado de polimerizacin
es aproximadamente 100. En cuanto a las maderas de frondosas, las ramificaciones del
esqueleto estn formadas por cido metilglucournico o por grupos acetilo, con trazas de
arabinosa. Este es un polmero ms largo, con un grado de polimerizacin de
aproximadamente 190 unidades. El enlace del cido metilglucournico es mucho ms
resistente a la accin de los cidos que los enlaces de la xilosa o de los grupos acetilo, que se
degradan ms fcilmente.

3.2.1.1.2. Mananos
Los mananos representan el grupo de heteropolmeros predominantes en las maderas
de conferas, apareciendo en pequeas proporciones en las maderas de frondosas y en los
materiales alternativos. En las maderas de frondosas aparecen cadenas formadas por manosa
y glucosa, sin ramificaciones, mientras que en maderas de conferas y materiales alternativos
se observa el mismo esqueleto pero con ramificaciones de una galactosa, muy susceptibles a
la hidrlisis mediante cidos. En estos dos ltimos materiales, al aumentar el nmero de
unidades de galactosa aumenta la solubilidad en agua y disminuye la solubilidad en lcalis.

51

Introduccin
3.2.1.1.3. Galactanos
Los galactanos estn presentes en los materiales lignocelulsicos en menor proporcin
que los xilanos y mananos, y son muy solubles en agua. En las maderas de conferas estos
heteropolmeros presentan una estructura muy ramificada, con residuos laterales de
arabinosa y/o cidos urnicos o cadenas de varias unidades que comienzan por galactosa.
Aparecen tambin en las maderas de frondosas, con ramnosa en las ramificaciones,
pudiendo encontrarse en menores proporciones xilosa, cido glucournico y cido
galactournico.

3.2.1.1.4. Glucanos
Los glucanos son compuestos constituidos mayoritariamente por unidades de glucosa.
El heteropolmero laricina est presente en las maderas de conferas y algunos materiales
alternativos en porcentajes del 2-4 %, con un grado de polimerizacin de 170-205,
ramificado y con enlaces -1,3. La laricina presenta trazas de cidos glucournico y
galactournico.

3.2.1.1.5. Xiloglucanos
Los xiloglucanos se encuentran en pequeas proporciones en las maderas de
conferas. Es un heteropolmero de glucosa y xilosa que puede presentar trazas de arabinosa,
galactosa, fucosa, ramnosa y grupos acetilo. Probablemente desempeen su funcin durante
el crecimiento del vegetal.

3.2.1.1.6. Sustancias pcticas


Estas sustancias constituyen un conjunto de heteropolmeros caracterizado por un
esqueleto formado por unidades del cido galactournico. Se encuentran en la pared
primaria y sobre todo en la laminilla media. Son ms abundantes cuando la clula es joven.
En las maderas estn presentes en porcentajes menores del 1 % y pueden presentar monmeros de ramnosa formando parte del esqueleto. Debido a su estructura, tienen la
propiedad de formar geles de alta viscosidad en disoluciones acuosas, pudiendo ser
utilizados como espesantes.

52

Introduccin
3.2.1.2. Despolimerizacin de las hemicelulosas por tratamiento hidrotrmico
Existen varios tipos de procesos hidrotrmicos, todos ellos basados en la autohidrlisis
de las hemicelulosas, que presentan diferencias en el modo de operacin. Adems del
simple proceso hidrotrmico propiamente dicho (en el que se somete la materia prima a
elevadas temperaturas en medio acuoso), es muy importante el proceso de explosin de
vapor, que consiste en que una vez producido el proceso de autohidrlisis la mezcla se
somete a una descompresin sbita para provocar la vaporizacin del agua contenida en las
fibras y la consiguiente desagregacin de la matriz celulsica.
Tambin son conocidos los procesos: RASH (Biermann y col., 1984), Masonite
(Carrasco y Roy, 1992), IOTECH (Marchessault y col., 1983; Rijkens, 1984), Siropulper
(Wallis y Wearne, 1985), Stake (NDiaye y col., 1996; Rubio y col., 1994; Taylor y Yu,
1995); todos ellos variantes del proceso de explosin de vapor.
El tratamiento hidrotrmico consiste, como se ha mencionado anteriormente, en tratar
las materias primas lignocelulsicas acondicionadas con agua a temperatura elevada, entre
150-250 C. Bajo estas condiciones se generan protones a partir del agua que actan como
catalizadores de la hidrlisis de las hemicelulosas, atacando a los grupos acetilo, que se
liberan en forma de cido actico. Este cido acta como el verdadero catalizador del
proceso ya que su contribucin a la generacin de protones es de 1700 a 1000000 de veces
mayor que la del agua, por lo que se puede despreciar la contribucin de los protones
acuosos al proceso de hidrlisis una vez que se ha empezado a generar el cido actico
(Carrasco, 1989; Heitz y col., 1986).
Durante el tratamiento hidrotrmico se produce la solubilizacin total o parcial de las
hemicelulosas y su conversin en oligmeros o monmeros, que pueden tener muy diversas
aplicaciones (Aoyama, 1996; Aoyama y col., 1995; Lpez-Alegret, 1996; Modler, 1994;
Saska y Ozer, 1995).
La eliminacin de las hemicelulosas al ser degradadas, redistribuye la lignina sobre la
superficie de las fibras y deja expuesta la superficie de la celulosa, aumentando la
accesibilidad de la misma a una hidrlisis enzimtica (Lamptey y col., 1985).
La presencia de pequeas cantidades de diversos cidos, como el cido sulfrico, hace
disminuir los valores de algunas variables de operacin durante el proceso, como el tiempo
y la temperatura, adems de disminuir la produccin de compuestos inhibidores de la
fermentacin y contribuir a la eliminacin completa de las hemicelulosas.
53

Introduccin
En este tipo de procesos aparecen otras reacciones de menor importancia que dan
lugar a la formacin de furfural a partir de las pentosas, de 5-hidroximetil-2-furfural (HMF) a
partir de las hexosas (Baugh y col., 1988), y de anhdrido carbnico por descomposicin de
los grupos carboxilos presentes en los cidos urnicos (Carrasco y col., 1987); as mismo
tiene lugar la condensacin de algunas molculas inestables que intervienen como medio de
reaccin, la descomposicin de productos como el furfural, sensibles a la concentracin de
cido (Baugh y McCarty, 1988), la descomposicin de HMF a cido frmico y levulnico, y
las reacciones de condensacin con la lignina (Aoyama y col., 1995; Muzzy y col., 1983).
Los procesos de autohidrlisis presentan las siguientes ventajas con respecto a la
hidrlisis cida:

Simplicidad del proceso por ausencia de reactivos qumicos distintos del agua,
minimizando el impacto ambiental del proceso (Overend y col., 1987).

Mnima generacin de productos de degradacin, al operar en condiciones


ms suaves que en otros tratamientos, obtenindose diversos oligosacridos
que pueden ser utilizados para distintos fines comerciales.

Mayor facilidad en los posteriores procesos de separacin de la lignina y la


celulosa, al provocar este tratamiento alteraciones fsico-qumicas al eliminar
las hemicelulosas que actan de puente entre la celulosa y la lignina, y al
despolimerizar la lignina.

Ausencia de lodos resultantes de la neutralizacin del cido y de su


recirculacin, simplificando la economa del proceso.

Posibilidad de someter la suspensin resultante a una descompresin sbita


(explosin de vapor) causando una desagregacin de la matriz lignocelulsica
que resulta favorable para un posterior procesamiento a efectos de obtencin
de pastas celulsicas.

Menores problemas de contaminacin y corrosin, por el moderado pH al


que se opera (Abatzoglou y col., 1992).

54

Introduccin
3.2.1.3. Condiciones de operacin en el tratamiento hidrotrmico
Los tratamientos hidrotrmicos pueden llevarse a cabo en un intervalo de condiciones
de operacin muy amplio. Las variables ms influyentes en este tipo de tratamientos son la
temperatura, tiempo, concentracin de slidos y tamao de partcula.

3.2.1.3.1. Temperatura
Suele fijarse como lmite inferior de referencia una temperatura de 100 C (Abatzoglou y col., 1992), dado que por debajo de esta temperatura no se producen procesos de
autohidrlisis (Tortosa y col., 1995). En cuanto al lmite superior se ha operado a
temperaturas de hasta 300 C (Tortosa y col., 1995). Atendiendo a criterios de selectividad
del proceso, para favorecer las reacciones de despolimerizacin de las hemicelulosas, es
preciso establecer el rango de operacin de temperaturas que convenga. As, para un
proceso hidrotrmico en medio cido dbil se advierte que los enlaces ter de la lignina
comienzan a romperse a 160-180 C, mientras que las reacciones de despolimerizacin de la
celulosa se inician a temperaturas superiores a 210 C (Heitz y col., 1986; Rubio y col.,
1994).

3.2.1.3.2. Tiempo de operacin


El tiempo de operacin puede variar desde segundos hasta varias horas. En
tratamientos a temperaturas relativamente elevadas suele ser importante el tiempo de
calentamiento (rampa de calentamiento) necesario para alcanzar la temperatura fijada de
operacin (Carrasco y Roy, 1992). En algunos casos, el tiempo de calentamiento puede ser
muy superior al de operacin o reaccin, mientras que el tiempo de enfriamiento suele ser
comparativamente pequeo (velocidad de enfriamiento de 30 C/min) y no influye apreciablemente en el proceso (Martnez y col., 1995,1997).

3.2.1.3.3. Concentracin de slidos


La concentracin de slidos se expresa habitualmente mediante la relacin
lquido/slido, parmetro que relaciona la cantidad de agua que se aade por unidad de
masa de slido (sobre base libre de humedad). En la bibliografa se encuentran valores entre
2 y 40 g de agua/g de material lignocelulsico, aunque lo habitual es alrededor de 10 g/g. Un
55

Introduccin
valor elevado de la relacin lquido/slido beneficia la extraccin de las hemicelulosas
pero es muy costoso, mientras que valores bajos son econmicamente ms favorables en
principio, aunque pueden provocar problemas de limitaciones de difusin y transporte de
los productos de reaccin (Jimnez, 2005).

3.2.1.3.4. Tamao de partcula


Un tamao de partcula pequeo facilita el tratamiento hidrotrmico, pero aumenta
considerablemente el gasto energtico en la molienda. Alternativamente, un tamao grande
puede dar lugar a limitaciones de difusin. Lo usual en los trabajos de laboratorio son
tamaos de partcula en el intervalo 0,5-10 mm, mientras que a escala industrial o de planta
piloto son normales tamaos de varios centmetros. El estudio de la influencia del tamao
de partcula sobre el proceso ha sido considerado por algunos autores (Ballesteros y col.,
2000; Brownell y col., 1986).

3.2.1.4. Efectos de los tratamientos hidrotrmicos sobre las hemicelulosas


Durante los tratamientos hidrotrmicos el cido actico crea un medio con valores de
pH comprendidos entre 3 y 4 (Heitz y col., 1991; Kaar y col., 1998; Tortosa y col., 1995;
Weil y col. 1998), en el cual se despolimerizan las hemicelulosas, generndose una mezcla
de oligmeros de azcares, monosacridos y, dependiendo de las condiciones de operacin,
productos de descomposicin de stos. En condiciones adecuadas la suma de oligosacridos
y monosacridos (xilosa, glucosa, arabinosa, etc.) procedentes de las hemicelulosas oscila
entre el 60 y el 85 % del total de las hemicelulosas iniciales (Aoyama y col., 1995; Ballerini y
col., 1994; Bouchard y col., 1991; Conner y Lorenz, 1986; Lamptey y col., 1985). La
proporcin de oligosacridos que aparecen en los licores hidrotrmicos es superior a la de
monosacridos, pudiendo representar ms de la mitad de las hemicelulosas iniciales
(Aoyama y col., 1995; Aoyama y Seki, 1994,1999; Bouchard y col., 1991; Burtscher y col.,
1987; Conner y Lorenz, 1986; Saska y Ozer, 1995).
La explicacin de que la recuperacin de oligmeros de azcares y monosacridos en
la fase lquida no sea cuantitativa respecto de la fraccin solubilizada, se debe a la
participacin de reacciones de degradacin, que conducen a furfural y HMF, muy sensibles
a las condiciones de operacin seleccionadas, pudindose encontrar porcentajes muy

56

Introduccin
variados de stos para un mismo material (Bouchard y col., 1991; Heitz y col., 1991;
Lamptey y col., 1985; Muzzy y col., 1983; Tortosa y col., 1995).
Los tratamientos hidrotrmicos en condiciones relativamente suaves de operacin no
provocan alteraciones de consideracin sobre la celulosa, para lo que se necesita alcanzar
temperaturas superiores a 260 C (Bonn y col., 1983), ya que las caractersticas fsicoqumicas de la celulosa, especialmente su estructura cristalina, hacen que su hidrlisis sea
ms difcil que la de las hemicelulosas. Este tratamiento es capaz de conseguir una
solubilizacin de las hemicelulosas entorno al 90 %, prdidas en la fraccin celulsica
alrededor del 10 % y un grado de deslignificacin del 10-20 %.
Los efectos de los tratamientos hidrotrmicos sobre la celulosa incluyen un
incremento del rea superficial, que aumenta con la severidad del tratamiento (funcin de la
temperatura y el tiempo conjuntamente), ya que la eliminacin de las hemicelulosas y parte
de la lignina aumenta el volumen de los poros; una disminucin del grado de
polimerizacin, que aumenta con la severidad del tratamiento; alteraciones en la
cristalinidad, debido a la solubilizacin selectiva de la celulosa amorfa; y por ltimo,
alteraciones en la microestructura, debido a que la estructura de la biomasa lignocelulsica a
nivel celular se ve fuertemente afectada por los procesos hidrotrmicos, ya que stos causan
la rotura de la laminilla media y separan la pared primaria de las clulas, provocando que a
una mayor severidad del tratamiento las clulas se separen entre s, afectando a las partes
interiores de stas, y la lignina se deposite en forma de partculas sobre las clulas (Roman,
2011).
La lignina puede sufrir reacciones de degradacin y de repolimerizacin durante la
autohidrlisis. La cantidad de lignina solubilizada depende de la condiciones de operacin y
de la materia prima empleada. La solubilizacin de la lignina supone la rotura de los enlaces
lignina-hidratos de carbono, que corresponden fundamentalmente a enlaces entre la lignina
y las hemicelulosas.
Las unidades siringilo (presentes en mayor proporcin en maderas de frondosas y en
materiales alternativos) se solubilizan ms fcilmente en medio acuoso (Holtzapple, 1993).
Para una temperatura dada, la fraccin de lignina solubilizada aumenta con el tiempo y
despus disminuye por efecto de la repolimerizacin. Se ha propuesto un mecanismo de
dos reacciones en serie para la lignina: en la primera etapa se solubilizan rpidamente
fragmentos de lignina de bajo peso molecular y alta reactividad, por rotura de los enlaces
lignina-hidratos de carbono; en el segundo paso, ms lento, pueden ocurrir reacciones de
57

Introduccin
repolimerizacin de la lignina, en presencia de los cidos orgnicos liberados durante la
autohidrlisis, formando productos de condensacin insolubles.
Los azcares hemicelulsicos y/o sus productos de degradacin (como el furfural)
tambin pueden reaccionar con la lignina. La formacin de lignina insoluble (tambin
llamada pseudolignina) por condensacin aumenta el contenido en lignina de Klason
(fraccin insoluble en H2SO4 al 72 %) de los residuos slidos resultantes de los tratamientos

hidrotrmicos. Esta lignina presenta un menor peso molecular que la original, sin diferencias
estructurales apreciables entre ambas (Roman, 2011).
As, la fraccin slida procedente del proceso hidrotrmico es susceptible de ser ms
fcilmente deslignificada posteriormente, debido a la destruccin de la matriz lignocelulsica, por la eliminacin de las hemicelulosas.

3.2.1.5. Productos obtenidos en el tratamiento hidrotrmico


Los tratamientos hidrotrmicos tienen diferentes objetivos, entre los que destacan: los
enfocados a la produccin de oligosacridos, fundamentalmente xilooligosacridos; aqullos
cuyo objetivo es la produccin de medios fermentativos obtenidos mediante la combinacin
de etapas de autohidrlisis/posthidrlisis; los encauzados a la produccin de sustancias
antioxidantes; los tratamientos cuyo objetivo es la obtencin de un residuo slido que sea
apto para la hidrlisis enzimtica; y por ltimo, los encaminados a la obtencin de una
fraccin slida susceptible de ser pasteada.

3.2.1.5.1. Xilooligosacridos
Los productos de degradacin obtenidos en los tratamientos hidrotrmicos son en su
mayora xilooligosacridos (XO) (oligosacridos compuestos fundamentalmente por unidades de xilosa). Estos xilooligosacridos son obtenidos en la industria a partir de la biomasa
lignocelulsica, aunque aparecen tambin en otro tipo de materiales, como fruta, leche o
miel.
Las frondosas y las materias primas alternativas son los materiales lignocelulsicos
tpicos para la produccin de xilooligosacridos, ya que estn constituidos por elevadas
proporciones de xilano. Es destacable su utilizacin en diversos campos de aplicacin, como
en la agricultura (en estimuladores y aceleradores del crecimiento); en alimentacin animal
58

Introduccin
(domsticos y peces); en el campo farmacutico (para prevenir enfermedades
gastrointestinales y dolencias como la osteoporosis, otitis, prurito cutneo y problemas de
cabellos y piel); y en alimentacin humana (en ingredientes de alimentos reforzados,
alimentos especiales (antiobesidad) o alimentos simbiticos (con microrganismos y
principios alimentarios con efecto positivo para la salud)) (Campbell y col., 1997; Domnguez y col., 2004; Jeong y col., 1998; Nabarlatz, 2006; Suwa y col., 1999; Toshio y col., 1990;
Vzquez y col., 2006).
La purificacin de las disoluciones de xilooligosacridos puede seguir diferentes
estrategias, intentando eliminar los compuestos no deseados y consiguiendo el grado de
polimerizacin requerido. Los tratamientos fisicoqumicos necesarios dependen del grado
de pureza deseado.
Para separar los XO con el deseado intervalo de grado de polimerizacin se pueden
utilizar tcnicas basadas en el uso de membranas (Gulln y col., 2008), que permiten un
fraccionamiento de los XO y la eliminacin de los compuestos que no estn formados por
cadenas de azcares.
Otros mtodos de purificacin de los licores tienen lugar utilizando tcnicas de
adsorcin, con el uso de adsorbentes como el carbn activo, bentonita, tierra de diatomeas,
hidrxido de aluminio, xido de aluminio, titanio, slice y materiales porosos sintticos.
Tambin son utilizadas resinas de intercambio inico; se llevaron a cabo estudios con
resinas cambiadoras de cationes, de aniones y mezclas de ambos tipos (Roman, 2011).

3.2.1.5.2. Medios fermentativos obtenidos mediante la combinacin de etapas de


autohidrlisis-posthidrlisis
Una primera etapa de autohidrlisis permite la solubilizacin de prcticamente la
totalidad de las hemicelulosas a oligosacridos y monosacridos. La recuperacin de las
hemicelulosas como monmeros y oligmeros poda variar entre un 55-99 % del total,
siendo bastante habituales valores entre 80-85 % (Garrote y col., 2001), observando que al
aadir cido sulfrico a los licores procedentes del tratamiento hidrotrmico se producan
conversiones de los oligosacridos a monosacridos con rendimientos entre 55 y 99 %
(Allen y col., 1996).
Una aplicacin interesante de estos licores, debido a su elevado contenido en xilosa, es
la obtencin, por fermentacin, de productos de inters industrial como el xilitol (Rivas y
59

Introduccin
col., 2002). A esta segunda etapa de hidrlisis de los oligosacridos a monosacridos (para su
utilizacin en el metabolismo de los microrganismos), normalmente se le denomina
posthidrlisis. En este tratamiento se suele utilizar cido sulfrico en concentraciones en
peso de 3-7 %, durante tiempos que van desde 2 h a 100 C hasta 5-60 min a 121 C. Las
ventajas de estos tratamientos de hidrlisis cida frente a la hidrlisis enzimtica, es que la
velocidad de reaccin es mucho mayor.
Tras las etapas secuenciales de autohidrlisis-posthidrlisis se pueden obtener
disoluciones ricas en xilosa, que pueden ser metabolizadas por los microrganismos, pero
tambin aparecen compuestos que inhiben el crecimiento de los mismos, como HMF,
furfural, cido actico, etc., siendo necesario aplicar tratamientos fsico-qumicos para la
detoxificacin de estos licores, logrando as una fermentacin eficiente de los hidrolizados.
Se ha encontrado una relacin directa entre la severidad del tratamiento hidrotrmico
y el carcter inhibidor de los hidrolizados (Larsson y col., 1999), siendo necesario operar
bajo condiciones suaves de operacin para limitar la deshidratacin de los azcares a
furfural y HMF, y los compuestos de degradacin de stos y la liberacin de compuestos
derivados de los fragmentos de la lignina solubilizados.

3.2.1.5.3. Sustancias antioxidantes


Los procesos hidrotrmicos solubilizan parte de la lignina que pasa a la fase lquida,
junto a los oligmeros, azcares, extractos, productos de degradacin, cido actico, etc. Si
los licores procedentes de los tratamientos hidrotrmicos se someten a procesos de
extraccin (para purificar los xilooligosacridos o para eliminar compuestos inhibidores de la
fermentacin), se obtienen extractos con alto contenido de compuestos fenlicos, que
pueden utilizarse como agentes antioxidantes o antimicrobianos (Paraj y col., 2008), dotando a la fraccin fenlica de una aplicacin con valor econmico.
La mayora de los compuestos antioxidantes de origen natural provienen de extractos
de materiales de origen vegetal y presentan naturaleza polifenlica. Otras actividades
fisiolgicas de los antioxidantes naturales incluyen actividades antibacteriana, antiviral,
antimutagnica, antialergnica, anticarcinognica, inhibidora del incremento de la presin
arterial, antiulcerosa y anticariognica (Caparrs, 2007).
Estudios sobre la posible toxicidad, baja eficacia y elevado coste de fabricacin de
alguno de los antioxidantes sintticos, muestran las ventajas de los antioxidantes de origen
60

Introduccin
natural. La capacidad antioxidante de los compuestos fenlicos es comparable con la de los
antioxidantes sintticos y adems pueden tener otros efectos beneficiosos adicionales.

3.2.1.5.4. Productos de la hidrlisis enzimtica del residuo slido del tratamiento


hidrotrmico
La presencia de las hemicelulosas y la lignina bloquean el acceso de las enzimas a los
enlaces glucosdicos de la celulosa de los materiales lignocelulsicos. Pero si stos se
someten a tratamiento hidrotrmico se produce una degradacin estructural, permitiendo
mejorar la accesibilidad de las enzimas hacia la celulosa y aumentando la susceptibilidad de
la celulosa a la hidrlisis, as como la facilidad de eliminacin de la lignina. De este modo,
los tratamientos hidrotrmicos se presentan como un pretratamiento adecuado para la
hidrlisis enzimtica, al eliminar las hemicelulosas y parte de la lignina.
Al estar compuesto el residuo slido resultante del tratamiento hidrotrmico
fundamentalmente por celulosa y lignina, la hidrlisis enzimtica transforma la celulosa en
glucosa (dejando un residuo de lignina); tambin puede realizarse un tratamiento intermedio
de deslignificacin, solubilizando la lignina, y realizando la hidrlisis enzimtica sobre el
slido enriquecido en celulosa (Requejo y col., 2011).
En el proceso de autohidrlisis-hidrlisis enzimtica se consiguen rendimientos
(medidos como celulosa convertida en glucosa respecto a la materia prima) del 70 al 100 %
(Cara y col., 2007; Prez y col., 2008).

3.2.1.5.5. Pasta celulsica


Un tratamiento hidrotrmico moderado (con bajos valores de la temperatura y de los
tiempos de operacin) puede dar lugar a una fraccin slida constituida principalmente por
celulosa y lignina. Esta fraccin ha sido alterada de tal forma que puede someterse a un
proceso de pasteado no excesivamente enrgico para dar lugar a pasta de papel con
propiedades aceptables y en algunos casos mejores que las correspondientes a la pasta
obtenida directamente de la materia prima original (Caparrs y col., 2007,2008a,b).

61

Introduccin
3.2.2. PROCESOS DE OBTENCIN DE PASTAS CELULSICAS
3.2.2.1. Consideraciones generales
El pasteado de las materias primas consiste en la separacin de las fibras celulsicas de
la lignina que las une. De manera general existen dos formas de liberar las fibras: mecnica y
qumica, que dan lugar a las denominadas pastas mecnicas y qumicas, respectivamente.
En el pasteado mecnico se utilizan equipos especiales que, mediante energa
mecnica, liberan las fibras slo parcialmente, sin la intervencin de reactivos qumicos; a
veces, sin embargo, se utiliza vapor de agua o reactivos qumicos diluidos, con la finalidad de
reblandecer la lignina, pero no de eliminarla. Las pastas mecnicas tienen rendimientos
elevados. Existen diversos tipos de pasteado mecnico, dando lugar a las pastas: mecnicas
de molino o piedra, mecnicas de refinador, termomecnicas, quimicomecnicas y
termoquimicomecnicas (Jimnez, 2005; Libby, 1976; Lpez y col., 2010b).
Por el contrario, en el pasteado qumico se utilizan compuestos qumicos para disolver
o degradar la lignina hasta un punto en que las fibras pueden separarse sin apenas energa
mecnica; normalmente la descompresin o soplado de una lejiadora o digestor donde se
verifica la disolucin o degradacin de la lignina, o los chorros de agua de dilucin en las
tinas de soplado, son suficientes para separar las fibras. El principal inconveniente de las
pastas qumicas es la gran cantidad de materia prima que se pierde, ya que el peso de fibra
obtenido supone slo el 35-55 % de la materia prima utilizada. Entre estas pastas se consideran las pastas alcalinas (como las kraft o al sulfato y a la sosa), las pastas cidas (como
la pasta al sulfito) y las pastas obtenidas con disolventes orgnicos (las organosolv) (Cartagena, 1994; Jimnez y Maestre, 1997; Libby, 1976; Lpez y col., 2006; Rodrguez y Jimnez,
2008).
Para obtener rendimientos ms elevados que los conseguidos en las pastas qumicas,
se necesita que el ataque a la lignina sea ms moderado y ello se compensa con un
tratamiento fsico en un equipo especial para la fibrilacin; a estos procesos, que constan
de las operaciones de coccin y fibrilacin, que consumen reactivos qumicos y utilizan
medios mecnicos, se les llama intermedios o semiqumicos. Una gran variedad de procesos
se incluyen en esta categora, estando unos ms cerca de las operaciones qumicas y otros de
las mecnicas. Las operaciones o tratamientos qumicos en los procesos semiqumicos
difieren de los practicados en las fbricas de pastas mecnicas (quimicomecnicas y

62

Introduccin
quimicotermomecnicas) en que mientras que en stos se pretende slo reblandecer la
lignina, en los procesos semiqumicos se desea tambin eliminar parte de ella.
A modo de resumen, una clasificacin muy corriente de las pastas es segn el
rendimiento conseguido: mecnicas con un rendimiento superior al 80 %, qumicas con un
rendimiento inferior al 60 % y semiqumicas con rendimientos intermedios (60-80 %); de
todos modos se ha de tener en cuenta que estos lmites son flexibles pues se encuentran
pastas semiqumicas con rendimientos desde el 50 % hasta el 90 % (Jimnez, 2005; Libby,
1976; Lpez y col., 2010b).
Es comn en todos los procesos de pasteado, llevar a cabo etapas de
acondicionamiento de la materia prima; as, los troncos de madera o los materiales
alternativos deben sufrir etapas de descortezado y/o desmedulado, de astillado o desmenuzado o molido, y de seleccin de los tamaos ms adecuados, normalmente entre 10 y 30
mm de longitud y entre 2 y 8 mm de espesor.

3.2.2.2. Pasteados qumicos convencionales


Los procesos ms utilizados para la fabricacin de pastas celulsicas son los procesos
qumicos, con una capacidad de produccin de 130,01 millones de toneladas en el ao 2011
a nivel mundial (Anuario FAO, 2012). Dentro de estos procesos el ms habitual es el
pasteado al sulfato o kraft, seguido en importancia por los pasteados al sulfito y a la
sosa, existiendo pequeas cantidades de pasta qumica producidas por otros procesos como
pueden ser los que utilizan disolventes orgnicos.
La materia prima ms habitual en los procesos de pasteado qumico es la madera,
destacando la de conferas para pastas sin blanquear y la de frondosas para pastas
blanqueadas, con un consumo muy superior al de los vegetales alternativos o no madereros.
El principal inconveniente que presentan algunos vegetales alternativos es el elevado
contenido de cenizas, que origina graves problemas en los circuitos de recuperacin de las
lejas de estos procesos.
El objetivo general de los procesos qumicos de obtencin de pastas celulsicas es
degradar y disolver la lignina, dejando la mayor parte de los carbohidratos intactos en las
fibras. En la prctica, los mtodos de obtencin de pastas qumicas son capaces de eliminar
gran parte de la lignina, pero tambin degradan una cierta cantidad de hemicelulosas y

63

Introduccin
celulosa, por lo que el rendimiento en pasta es bajo en relacin con el pasteado mecnico,
usualmente sobre el 40-60 % de la materia prima original (Libby, 1976).

3.2.2.2.1. Pasteado a la sosa


En los procesos qumicos de pasteado se producen efluentes contaminantes,
particularmente cuando se utilizan compuestos de azufre como agentes deslignificantes. Con
el fin de evitar este problema se ha considerado la posibilidad de usar disolventes orgnicos
libres de azufre para eliminar la lignina y separar las fibras celulsicas; pero aunque esta
alternativa es conocida desde hace tiempo, este tipo de procesos slo se ha llevado a cabo a
escala de planta piloto (Hergert, 1998; Jimnez y Maestre, 1997; Lpez y col., 2006;
Muurinen, 2000; Rodrguez y Jimnez, 2008). Por todo ello, la situacin actual hace que se
preste atencin al ms antiguo proceso de pasteado conocido, como es el del pasteado a la
sosa, que consiste en someter las materias primas, troceadas y acondicionadas, a una
digestin con una concentracin dada de hidrxido sdico, a una temperatura y tiempo de
coccin determinados, segn la calidad de las pastas que se deseen obtener y las
caractersticas de las materias primas utilizadas (Gullichsen y Paulapuro, 2000).
Este proceso de pasteado da lugar a pastas qumicas o semiqumicas, segn que las
condiciones de operacin sean ms o menos enrgicas. En el proceso, adems de la pasta,
caracterizada por la presencia predominante de fibras cortas, fcilmente blanqueable y con
abundantes pentosanos, aparecen unas lejas negras o residuales, que contienen productos
de degradacin de la lignina y de hidrlisis de la celulosa, constituidos por sales sdicas de
cidos orgnicos complejos, lignofenolatos y el exceso de sosa. Estas lejas han de someterse
a un proceso de depuracin-recuperacin para ser reutilizadas en el proceso de coccin,
tanto por motivos econmicos como medioambientales, ya que provocaran una gran
contaminacin en las corrientes de aguas naturales donde se vertiesen.
El proceso a la sosa es el ms antiguo y simple de todos los procesos de pasteado, y
puede utilizarse tanto para maderas de conferas como de frondosas (Cartagena, 1994;
Libby, 1976), as como para multitud de materias primas alternativas a las convencionales
como los residuos agrcolas (paja de cereales, podas de olivo, tallos de girasol, sarmientos de
vid, tallos de algodonera, etc.) y agroalimentarios (bagazo de la caa de azcar, racimos
vacos de fruto de la palma aceitera, etc.) y otros vegetales anuales o no (tagasaste, leucaena,
abac, kenaf, etc.) (Jimnez, 2005).

64

Introduccin
Si las condiciones de operacin son lo suficientemente enrgicas puede considerarse a
este proceso como qumico, eliminndose en l gran parte de la lignina de las materias
primas y degradndose tambin parte de las hemicelulosas y algo de celulosa, por lo que el
rendimiento en estas pastas es bajo, sobre un 40-60 % respecto a la materia prima de partida
(Libby, 1976). Si las condiciones de operacin son ms suaves se obtiene una pasta semiqumica, con un aprovechamiento mayor de la materia prima.
El proceso de pasteado a la sosa consta de una serie de operaciones que son
comunes a los distintos sistemas de pasteado qumicos o semiqumicos. De manera general,
el proceso global de obtencin de pasta celulsica a la sosa consta de varias secciones o
reas, destacando las de preparacin de la materia prima, pasteado propiamente dicho de la
materia prima acondicionada, recuperacin y depuracin de las lejas residuales, y blanqueo
de las pastas si lo exigen las aplicaciones de stas. Cada una de estas reas o secciones del
proceso pueden agrupar diferentes operaciones. As, por ejemplo, en la preparacin de la
materia prima se realiza un descortezado para el caso de plantas leosas o un desmedulado
para vegetales con mdula, un troceado o reduccin de tamao para producir astillas, copos
o materia prima desmenuzada, una limpieza o depuracin para eliminar materiales extraos,
etc. (Casey, 1990; Colom, 2005; Jimnez, 2005; Lpez y col., 2010b); en la seccin de
pasteado pueden integrarse las operaciones de impregnacin de la materia prima, coccin o
deslignificacin para separar lignina, lavado de la fraccin slida resultante de la coccin y
escurrido de la misma para eliminar el fluido utilizado en el lavado; de igual modo, las
secciones de recuperacin y depuracin de lejas y de blanqueo de las pastas estn
constituidas por diferentes operaciones. De manera simplificada, en la Figura 3.13 se
presenta el diagrama de flujo del proceso de pasteado a la sosa.
El proceso de pasteado a la sosa se ha empleado para la obtencin de pastas de
diversos materiales alternativos a las materias primas convencionales. A guisa de ejemplos se
seala que se ha aplicado, a veces en trabajos recientes, a materiales tan diversos como:

Genus Quercus, paja de arroz, yute, guayule, plantas acuticas como Eichhornia crassipes,
Scircus grossus y Penicum repens, kenaf, Albizia procera, Anthocephalus cadamba, paja de
trigo, podas de olivo, tallos de girasol, sarmientos de vid, tallos de algodonera, tallos de
sorgo, algas (Typhus spp), Sterculia villosa Rosb., hojas de vegetales, Cyperus alternifolius,
residuos de la industria del tequila, Hesperaloe funifera, paja de cebada, papiro, palma de
aceite, entre otros (Jimnez, 2005; Lpez y col., 2010b).

65

Introduccin

Materia prima

Descortezado

CO2

y/o desmedulado

Astillado

Calcinacin

Depuracin de la

Impregnacin y

CaO

CO3Ca

materia prima

Leja Blanca
NaOH

Caustificacin

Agua

Coccin
CO3Na2
Separacin de

Combustin

incocidos
Vapor
Escurrido y
lavado

Evaporacin
Leja Negra

Pasta cruda

Proceso de
blanqueo

Pasta blanqueada

Figura 3.13. Proceso de pasteado a la sosa.


66

CO3Na2
Suplementario

Introduccin
Se ha comprobado que la adicin de pequeas cantidades de ciertos compuestos
orgnicos que actan como catalizadores (hidracina, hidroxilamina, antraquinona, etc.)
aumentan la velocidad y selectividad del proceso de deslignificacin a la sosa, tanto para
materias primas convencionales (frondosas y conferas), como para otras alternativas a las
anteriores (Jimnez, 2005; Lpez, 1991; Lpez y col., 2010a). El principal compuesto de
este tipo es la antraquinona.
El proceso a la sosa-antraquinona (sosa-AQ) es un proceso relativamente reciente
propuesto por Holton (1977); consiste en una variante del proceso a la sosa, en la que se
utiliza antraquinona como catalizador durante la operacin de coccin.
El mecanismo de actuacin de la antraquinona en el proceso a la sosa-AQ es bien
conocido: acta como un catalizador redox para las reacciones que se producen durante la
coccin (Figura 3.14) (Fleming y col., 1979; Kubes y col., 1978).

Reaccin redox de antraquinona

Oxidacin de aldehdo por antraquinona


Figura 3.14. Reacciones de la antraquinona.

La solubilidad de la antraquinona se debe a la reduccin; los electrones de los grupos


aldehdos de los carbohidratos de las fibras se transfieren a la molcula de antraquinona y
los grupos aldehdos se transforman en grupos carboxilos (Figura 3.14). Esta transformacin
hace que los carbohidratos se estabilicen producindose un aumento del rendimiento y
evitando que se produzca el proceso de peeling (Lwendahl y Samuelson, 1978).

67

Introduccin
Debido a la transferencia de electrones, la antraquinona pasa a su forma reducida,
antrahidroquinona. El mecanismo de aceleracin de la deslignificacin con antraquinona
implica el ataque de la antrahidroquinona a las estructuras transitorias y reactivas de la
lignina (Landucci, 1980).
El uso de la antraquinona como catalizador para acelerar el proceso de
deslignificacin no debe superar la cantidad del 0,1 % sobre base referida al material
lignocelulsico seco; esta cifra es el mximo nivel aceptado que la US Food and Drug
Administration permite para la produccin de papel para usos alimentarios (Barba, 2002).
Goyal y Ray (1989) compararon los procesos a la sosa y a la sosa-AQ. La adicin
de antraquinona da lugar a un aumento del rendimiento y a unas disminuciones del tiempo
de coccin y del nmero kappa de la pasta. Muchos de estos cambios se acentuaban con un
aumento de la carga alcalina.
Cao (1996) y Cao y Zhong-Zheng (1996) encontraron que aadiendo un 1 % de antraquinona al licor de coccin se reduca un 34 % el consumo de lcali, y el nmero kappa disminua 20 unidades. Se contrastaron estos resultados aadiendo un 3 % de perxido de
hidrgeno y entonces el consumo de lcali aument en un 25 %, no afectndose la deslignificacin; la antraquinona resulta ser ms selectiva para la deslignificacin.

3.2.2.2.2. Pasteado kraft o al sulfato


Las pastas que se obtienen mediante este procedimiento se suelen llamar kraft (fuerte)
si se usan para papeles crudos, o al sulfato si van a recibir un posterior blanqueo, si bien
ambas denominaciones se utilizan indistintamente.
La denominacin de proceso al sulfato es debida a que es el sulfato sdico el
reactivo que se repone, aunque el verdadero agente que acta durante la reaccin es el
sulfuro que se genera en el tratamiento de las lejas residuales.
El proceso puede dividirse en dos partes: la primera es la obtencin de la pasta y la
segunda la recuperacin de los productos qumicos usados.
La obtencin de la pasta consta de las siguientes etapas:
1. Las astillas se llevan a la lejiadora donde se cuecen con leja blanca (disolucin
de hidrxido sdico y sulfuro sdico), controlando la relacin lquido/slido.

68

Introduccin
2. La coccin se efecta durante el tiempo establecido, bajo las condiciones de
presin adecuadas (7 a 8 Kg/cm2, durante 2 a 4 horas).

3. La leja negra o residual y la pasta se separan por filtracin. La pasta se lava y la


leja negra se enva a la fase de recuperacin de reactivos.
4. La pasta una vez lavada pasa a la etapa de blanqueo o a la planta de fabricacin
de papel crudo.
En las cocciones kraft, una parte de la leja negra puede recircularse para su utilizacin
como leja de coccin. En algunos casos, de un 40 a un 60 % de leja de coccin puede estar
constituida por leja negra sin que ello afecte al rendimiento de la coccin ni a las
caractersticas de las pastas obtenidas. As se vuelven a utilizar parte de los reactivos sin que
medie una evaporacin costosa, se favorece la penetracin de la leja en las astillas y se utiliza
la energa calorfica de la leja negra.
En la fase de recuperacin de productos, los compuestos orgnicos disueltos en la leja
negra se utilizan para producir energa, reduciendo al mismo tiempo la tasa de efluentes
contaminantes. Las etapas que tienen lugar en la recuperacin son las siguientes:
1. Concentracin de la leja en evaporadores de mltiple efecto.
2. Pulverizacin de la leja concentrada en el horno, donde el carbono reduce el
sulfato sdico a sulfuro sdico.
3. Los slidos fundidos se descargan y se disuelven en agua, dando lugar a la leja
verde.
4. La leja verde se enva a la etapa de caustificacin, donde el carbonato sdico
reacciona con el xido clcico para formar hidrxido sdico.

3.2.2.2.3. Pasteado al sulfito


Este proceso utiliza SO2 y sales amnicas, alcalinas o clcicas como reactivo. Resulta

ms ventajoso que el mtodo a la sosa porque se obtienen rendimientos ms altos y una


elevada blancura de la pasta de celulosa.
Entre los inconvenientes cabe destacar: su adecuacin a un nmero limitado de
maderas, la baja resistencia mecnica de las pastas y su poca resistencia al envejecimiento

69

Introduccin
(Henriksen y col., 2004). El proceso est en franco retroceso tanto por razones tcnicas
como ambientales (Jimnez, 2005).

3.2.2.3. Pasteados con disolventes orgnicos


Los procesos de pasteado qumicos y semiqumicos tradicionales se caracterizan por la
generacin de grandes cantidades de lejas residuales de elevado poder contaminante,
principalmente si contienen compuestos de azufre. Por otra parte, cada vez es ms alto el
nivel de concienciacin sobre la conservacin del medio ambiente por parte de las
sociedades desarrolladas, hecho que debe compaginarse con la mayor demanda de pastas
para papel y cartn que existe en el mercado, por lo que resulta interesante enfocar los
esfuerzos investigadores tanto en la utilizacin de nuevas materias primas (las alternativas a
las tradicionales), como en el estudio y desarrollo de nuevos procesos de pasteado, como los
que utilizan reactivos menos contaminantes y fcilmente recuperables (los disolventes
orgnicos); estos procesos denominados genricamente procesos organosolv, implican,
adems, inversiones pequeas en plantas industriales de reducida produccin, con bajos
costes de produccin y alta calidad de los productos, as como un buen aprovechamiento de
las materias primas, consiguiendo elevados rendimientos para las pastas y la aparicin de
subproductos, con una tendencia al aprovechamiento integral de las materias primas.
Los pasteados llevados a cabo en medio orgnico (organosolv) se definen como
mtodos alternativos para la obtencin de pasta de celulosa y consisten en la deslignificacin
por accin de un disolvente (acetona, etanol, metanol, cido actico, cido frmico, etc.) en
presencia o ausencia de catalizadores (como cidos, antraquinona o sales), consiguiendo una
deslignificacin selectiva y una lignina poco alterada qumicamente.
Las primeras referencias cientficas que se conocen acerca de la deslignificacin con
disolventes orgnicos datan de 1893, en que Kason utiliza etanol y cido clorhdrico;
despus, ya en el siglo pasado, durante los aos treinta son de inters los trabajos de
Aronovsky y Grotner y los de Kleinert y Tayenthal; posteriormente, en los aos cincuenta se
encuentran los trabajos de Brounstein, y hasta los setenta los de Kleinert (Aziz y Sarkanen,
1989; Johansson y col., 1987). Hasta esta fecha, estos procesos no cuestionaron la preponderancia de los procesos qumicos tradicionales.
A lo largo de la dcada de los setenta se empez a pensar cmo solucionar los
inconvenientes de los procesos clsicos (olores desagradables, bajos rendimientos, elevada
70

Introduccin
contaminacin, pastas difciles de blanquear, grandes inversiones, elevados consumos de
agua, energa, reactivos y materias primas, etc.), modificando los procesos, primero, y
proponiendo procesos sin azufre, despus, pero se encontraron nuevas dificultades como la
difcil recuperacin de los reactivos y los problemas de contaminacin (Cox y Worstor,
1971).
En la dcada de los ochenta se revisan los procesos que utilizan disolventes orgnicos,
como alternativos a los procesos clsicos. La principal ventaja de los procesos con
disolventes orgnicos era que aprovechaban ntegramente las materias primas; a veces se
consideraba a estos procesos ms que como productores de pastas, productores de celulosa
hidrolizable, polmeros fenlicos de lignina y azcares (Muurinen, 2000; Oliet, 1998;
Varshney y Patel, 1988).
En la dcada de los noventa el proceso kraft no tiene competidores dentro de los
procesos qumicos de pasteado, pero los problemas medioambientales y las elevadas
inversiones necesarias llevan a la conclusin de que sera conveniente buscar nuevos
procesos de produccin de pastas. stos pueden ser los procesos que emplean disolventes
orgnicos, que adems siguen la filosofa actual de que la materia prima no slo es fuente de
fibras celulsicas, sino que se ha de aprovechar ntegramente; es la denominada biorefinera
de la madera, por similitud al fraccionamiento del crudo del petrleo (Judt, 1990).
En el momento actual y desde finales de la ltima dcada del pasado siglo se ha de
resaltar la aparicin de numerosos trabajos que tratan de la deslignificacin de materias
primas (tanto clsicas como alternativas) mediante procesos que utilizan disolventes
orgnicos (Hergert, 1998; Lpez y col., 2006; Muurinen, 2000; Neves y Neves, 1998; Oliet,
1998; Rodrguez y Jimnez, 2008).
Estos nuevos mtodos de deslignificacin, que utilizan reactivos menos contaminantes
y ms fcilmente recuperables como los disolventes orgnicos (Zhao y col., 2009), comenzaron a tener inters al considerarse tecnologas limpias por presentar numerosas ventajas
frente a los procesos tradicionales, entre las que destacan la reduccin de emisiones
(principalmente de compuestos de azufre), as como los bajos costes de produccin e
inversiones econmicas por la necesidad de una menor capacidad de produccin, unos
rendimientos favorables del proceso al presentar la qumica del proceso una mayor
selectividad, y una fcil recuperacin por destilacin de los disolventes orgnicos (Hergert,
1998; Jimnez, 2005). Aunque tambin presentan algunos inconvenientes, debido al elevado
coste de los disolventes orgnicos, siendo de gran importancia su recuperacin; la volatilidad
71

Introduccin
de los mismos; y que los slidos obtenidos deben ser lavados con una disolucin de aguadisolvente orgnico para evitar la precipitacin de la lignina disuelta.
Los disolventes empleados se pueden clasificar entre los de bajo punto de ebullicin
(metanol, etanol, acetona, acetato de etilo, alcohol benclico, etc.) que son recuperados
fcilmente por destilacin, y los de elevado punto de ebullicin (trietilenglicol, etilenglicol,
aminas, etc.) que tienen la ventaja de operar a presiones moderadas (Kleinert, 1974). Los
disolventes orgnicos de elevado punto de ebullicin tienen la ventaja de que al trabajar a
bajas presiones se pueden aprovechar las instalaciones convencionales ya existentes para los
procesos kraft, al sulfito y a la sosa, con el consiguiente ahorro econmico, lo que no
puede hacerse con los alcoholes y otros disolventes orgnicos de bajo peso molecular
(Sarkanen, 1991; Uraki y Sano, 1999).
De manera general, las mezclas disolvente-agua proporcionan mayores deslignificaciones que los disolventes puros, aunque sean menos selectivas y tengan lugar prdidas
importantes de hemicelulosas (Paszner y Cho, 1989). Pese a que la lignina presente mayor
solubilidad en los disolventes puros es necesaria la presencia de agua para forzar la
eliminacin de los grupos acetilo del material lignocelulsico y, por lo tanto, la acidificacin
del medio por disociacin del cido liberado. Por ello, los procesos que utilizan elevadas
concentraciones de alcohol, para ofrecer una mayor selectividad de deslignificacin, suelen
necesitar de un aporte adicional de cidos u otros aditivos.

Figura 3.15. Tratamiento de deslignificacin mediante disolventes orgnicos.

Para terminar, se ha de sealar que actualmente la aplicacin de cualquier proceso de


pasteado que utiliza disolventes orgnicos implica la obtencin de pastas, lignina, azcares y
72

Introduccin
otros productos, buscando el aprovechamiento integral de las materias primas (Abad y col.,
1997; Botello y col., 1999a,b; Hergert, 1998; Lehnen y col., 2005; Montan y col., 1998;
Muurinen, 2000; Oliet, 1998).
Dado que los alcoholes son los disolventes ms utilizados en los procesos de pasteado,
llegndose incluso a implantar plantas piloto y comerciales en algunos casos, en este trabajo
se tratan por separado estos disolventes orgnicos de otros cuyas investigaciones no han sido
tan extensas, como en el caso de los cidos orgnicos y aminas. En la Figura 3.15 se presenta
un diagrama del tratamiento de deslignificacin mediante disolventes orgnicos.

3.2.2.3.1. Pasteado con alcoholes


Los procesos de pasteado que utilizan alcoholes han sido los ms estudiados,
destacando algunos que se han desarrollado a nivel de planta piloto o a escala comercial,
como los procesos Alcell, MD-Organocell y ASAE, que utilizan etanol, y los procesos
Organocell y ASAM, que utilizan metanol (Jimnez y Maestre, 1997; Kleinert, 1974; Lpez
y col., 2006; Muurinen, 2000). Las principales limitaciones de estos procesos radican en que
cuando se aplican a maderas conferas no se obtienen papeles de calidad y en que las
propiedades de las pastas de maderas frondosas son variables, segn la especie considerada,
lo que puede ser causa de problemas en el caso de mezclas de maderas: abundantes
rechazos, elevado consumo de energa para el refinado, alto coste de blanqueo y bajo
rendimiento (Lpez y col., 2006).
Dentro de los procesos que utilizan alcoholes, pueden considerarse los que no utilizan
catalizadores, como el proceso Alcell (etanol al 50 %v) (Cronlund y Powers, 1992; Lpez y
col., 2006; Ni y Van Heiningen, 1996). Estos procesos sin catalizador pueden realmente
considerarse como autocatalizados, siendo el cido que se libera como consecuencia de la
hidrlisis de las hemicelulosas el que acta como catalizador en las reacciones de
deslignificacin; en efecto, durante la coccin disminuye el pH debido a la liberacin de
cido actico por desacetilacin de las materias primas (Gilarrahz y col., 1998) y de cido
frmico por degradacin de los polisacridos (Gosselink y col., 1995), aunque la cantidad de
cido frmico liberada a la leja durante la coccin es muy pequea (Girard y Van
Heiningen, 1997).
Los procesos autocatalizados con etanol son compatibles con el aprovechamiento de
subproductos como ligninas, azcares, furfural, cido actico, etc. (Oliet, 1998; Paszner y
73

Introduccin
Cho, 1989), considerando que la viabilidad econmica de estos procesos requiere la
comercializacin de los subproductos (Axelsson y col., 2004).
Otros procesos con alcoholes necesitan catalizadores, como el proceso patentado por
Paszner y Cho (1989) que utiliza, junto con el alcohol, sales de magnesio como catalizador.
Para la paja de trigo se ha utilizado un proceso alcalino con etanol, con un pretratamiento
con cido sulfrico, consiguiendo pastas con buenas propiedades de resistencia
(Papatheophanus y col., 1995; Sun y col., 1997). Otro proceso con catalizador es el que
emplea el sistema bicarbonato sdico-alcohol-agua (Lpez y col., 2006).
Tambin se ha utilizado sales alcalinotrreas como catalizadores (proceso NAEM)
(Lohrasebi y Paszner, 2001; Paszner y Cho, 1989), as como el oxgeno molecular en
disoluciones de alcoholes (metanol, etanol y butanol) (Deinko y col., 1992). Se ha de
destacar tambin la iniciativa de Zhang y col. (1998) que estudian el proceso organosolv con
etanol y ozono.
Los procesos que utilizan alcoholes, y en concreto el etanol, se estn usando cada vez
ms en el tratamiento de materias primas alternativas a las convencionales, como residuos
agrcolas y agroindustriales y plantas anuales leosas y herbceas. Estos materiales
normalmente son ms fciles de pastear que los materiales madereros, sin embargo, y
dependiendo de los materiales utilizados, se pueden tener graves problemas causados por el
alto contenido de slice que presentan algunos de estos materiales, afectando tanto a la
calidad de las pastas como a la recuperacin de los disolventes. Se han llevado a cabo
diferentes estudios sobre el pasteado con alcoholes de Cynara cardunculus (Oliet y col.,
2005), caa de azcar (Fernndez, 1996), leucaena (Lpez y col., 2010a), podas de tagasaste
(Lpez y col., 2004), etc.
Al residuo de la industria del aceite de palma (Elaeis guineensis), muy abundante en
Malasia (51 % de la produccin mundial) (Hunt y col., 2002), se le ha aplicado el proceso
al etanol (Aziz y col., 2002). Asimismo se estudi la adicin de catalizadores como el cido
clorhdrico y la sosa. Los catalizadores cidos mejoran el rendimiento de las pastas. Sin la
presencia de cido, una temperatura de 165 C no permite la coccin del material en su
totalidad, apareciendo gran cantidad de materias fibrosas intactas. Sin embargo, los anlisis
qumicos de las pastas obtenidas con cido muestran una peor calidad, con mayores
contenidos de lignina y ms bajos de holocelulosas y -celulosa. Tambin Suleman y Yusoff
(1997) han estudiado este residuo aplicando procesos organosolv.

74

Introduccin
3.2.2.3.2. Pasteado con cidos orgnicos
Pasteado Formosolv
El proceso Formosolv utiliza cido frmico y es muy antiguo, ya que se conoce desde
el 1917, donde se deslignific madera y paja de cereales con este cido. Este proceso ofrece
ventajas, con respecto a otros que utilizan otros disolventes orgnicos, como el bajo coste del
proceso, por ser barato el disolvente, y que opera a temperatura y presin bajas (Erismann y
col., 1994). Se ha estudiado para diversos tipos de maderas como Eucalyptus globulus y

Eucalyptus grandis (Baeza y col., 1991). Para el caso del Eucalyptus grandis se encuentra que
las mejores condiciones son cuando se utiliza una concentracin de cido frmico del 92 %,
clorhdrico al 0,22 %, un tiempo de 90 minutos a 90 C y, mejor an, con reflujo. Se
comprueba que la presencia de agua beneficia las fibras y el rendimiento, perdindose en el
proceso, adems de lignina, muchas hemicelulosas y poca -celulosa (Erismann y col.,
1994).

Pasteado Milox
Una variante del proceso anterior es el proceso Milox, que utiliza cido peroxifrmico
formado espontneamente al mezclar cido frmico (80 %) y perxido de hidrgeno (Aziz y
Sarkanen, 1989; Poppius-Levlin y col., 1991; Ruggiero y col., 1998; Seisto y Laine, 1999).
Este proceso es muy adecuado para deslignificar abedul mediante tres etapas: tratamiento
con cido peroxifrmico a 80 C, tratamiento con cido frmico a 100 C y un blanqueo con
perxido de hidrgeno para conseguir pasta con una blancura de 90 % ISO. Las propiedades mecnicas de las pastas son aproximadamente iguales a las pastas kraft. Para el
caso de maderas conferas los resultados son peores. El principal problema de este proceso
es que al recuperar el disolvente, por destilacin, se forma un azetropo del cido frmico
con el agua, cuya concentracin (78 %) no es suficiente para el proceso (Aziz y Sarkanen,
1989). El proceso Milox tambin se ha estudiado para la deslignificacin de abeto
(Cimpoesu y Obrocea, 1995) y materias primas alternativas (Seisto y Poppius-Levlin, 1996).

3.2.2.3.3. Pasteado con etanolaminas


Los primeros estudios de pasteado con etanolaminas se remontan a finales de los
setenta del pasado siglo. En ellos, Wallis (1978) las aplican a maderas de pino y de eucalipto,
encontrando que la monoetanolamina es ms efectiva que la dietanolamina y la
75

Introduccin
trietanolamina; se consiguen rendimientos un 11-16 % mayores que con el pasteado kraft, y
las propiedades de resistencia de las pastas son similares a las de las pastas kraft. Tambin
Wallis (1980) estudia el pasteado de Pinus elliottii con etanolamina, obteniendo pastas con
propiedades de resistencia similares a las kraft, pero con mayor rendimiento. Otros trabajos
sobre etanolaminas son el de Kubes y col. (1978) que aplican este mtodo a la pcea, y el de
Guha y col. (1987) que lo aplican a la paja de arroz. Por otra parte, al aadir etanolamina a
las lejas de cocciones alcalinas se incrementa la velocidad de degradacin de la lignina
(Rodrguez y Jimnez, 2008). Es importante sealar, por ltimo, los trabajos ms recientemente realizados con aminas sobre distintas materias primas como haya, pcea, podas de
olivo, yute y algodn (Claus y col., 2005; Jahan, 2001; Jahan y col., 2001; Jahan y Farouqui,
2000,2003; Jimnez y col., 2002,2004).

3.2.3. PROCESOS DE BLANQUEO DE PASTAS CELULSICAS


3.2.3.1. Consideraciones generales
El blanqueo de las pastas celulsicas se lleva a cabo para la eliminacin de algunos
constituyentes que aportan color a la pasta cruda y/o para la modificacin de otros,
utilizando generalmente reactivos qumicos en una o ms etapas, y procurando degradar lo
menos posible las fibras celulsicas. Las operaciones que se efectan dependen del carcter
de la pasta cruda y de las propiedades deseadas para los productos finales (Dence y Reeve,
1996; Garca y Vidal, 1984; Loras, 1990). Por tanto, el proceso de blanqueo trata de
incrementar la blancura de la pasta, por eliminacin o modificacin de las principales
sustancias absorbentes de la luz, como la lignina y sus productos de degradacin, resinas,
iones metlicos, hidratos de carbono no celulsicos y otras impurezas coloreadas, pero
siempre manteniendo el compromiso entre el grado de blancura y la prdida de
propiedades mecnicas debido a la degradacin de las fibras de celulosa.
Las principales sustancias absorbentes de luz en las pastas son la lignina y las resinas,
por lo que para blanquear una pasta estas sustancias deben ser transformadas qumicamente
en estado slido, para disminuir sus caractersticas de absorcin de luz, o ser oxidadas,
reducidas o hidrolizadas, para hacerlas solubles en soluciones acuosas y poder as ser
eliminadas de las pastas.
En las pastas mecnicas y semiqumicas, la eliminacin de lignina y resinas conduce a
una prdida de rendimiento elevada, por lo que resulta interesante blanquear sin disolver la
76

Introduccin
lignina, empleando perxidos, hidrosulfitos y ozono, que no eliminan la lignina, sino que
transforman sus grupos cromforos, empleando normalmente una etapa de blanqueo, no
consiguiendo en la mayora de los casos un grado de blancura superior al 70-80 %.
Para las pastas qumicas y algunas semiqumicas, la eliminacin de lignina es necesaria
si el grado de blancura deseado es ms elevado, provocando una prdida de rendimiento
entre el 3-10 % o ms. Este mtodo se lleva a cabo en varias etapas, donde se pretende degradar la lignina y posteriormente neutralizar y disolver sus productos de degradacin,
buscando alcanzar el grado de blancura deseado en las ltimas etapas, tras las etapas de
oxidacin (Dence y Reeve, 1996; Loras, 1990).
En el proceso de blanqueo se adicionan reactivos qumicos en una o ms etapas,
donde adems de aumentar el grado de blancura se modifican otras caractersticas de la
pasta, como su capacidad de absorcin de agua, el grado de polimerizacin, la composicin
de hidratos de carbono y la viscosidad; la lignina se solubiliza utilizando procesos de
oxidacin, reduccin o hidrlisis, mientras que la eliminacin de carbohidratos y resinas se
efecta con etapas alcalinas, y la viscosidad es modificada en la etapas de oxidacin. En la
primera etapa de una secuencia de blanqueo se utilizan los reactivos qumicos que
contribuyen a la deslignificacin de las pastas, mientras que en las ltimas etapas se utilizan
agentes que eliminen el color residual.
En el blanqueo interesa obtener una blancura estable, es decir, que no se pierda con el
tiempo (evitando el envejecimiento o amarilleamiento); para ello se debe operar de forma
que no se introduzcan en la pasta grupos carbonilo (aldehdos o cetonas) y carboxilo
(principalmente en las hemicelulosas).
Los principales agentes de blanqueo son (Garca y col., 1985):

Oxidantes: cloro, hipocloritos de calcio y de sodio, dixido de cloro, clorito


sdico, clorato sdico, perxidos de hidrgeno y de sodio, oxgeno y ozono.

Reductores: hidrosulfitos de sodio y zinc, borohidruros de sodio, bisulfito


sdico, anhdrido sulfuroso y tiosulfato sdico.

Cualquier agente blanqueante reacciona con la pasta de dos maneras:


a) Con la lignina y otros constituyentes no celulsicos de la pasta, para satisfacer
la demanda de blanqueo requerida.
b) Con los componentes celulsicos (celulosa y hemicelulosas), degradndolos.
77

Introduccin
Las velocidades relativas de estos dos grupos de reacciones determinan el grado de
blancura obtenido y la degradacin de la pasta en el blanqueo. En pastas para papel el
objetivo es encontrar agentes blanqueantes que satisfagan la demanda de blanqueo de los
componentes no celulsicos con un mnimo de ataque a los carbohidratos. Para un
determinado agente blanqueante el propsito es encontrar un ptimo para las condiciones
de operacin: tiempo de reaccin, temperatura, pH, consistencia y porcentaje de reactivo
sobre pasta seca.
Durante mucho tiempo se utiliz el cloro en forma de hipoclorito. Posteriormente se
pusieron a punto blanqueos ms eficaces con diversas etapas: CEH (cloro - extraccin
alcalina - hipoclorito), CEHEH (cloro - extraccin alcalina - hipoclorito - extraccin alcalina hipoclorito), etc. Despus, con la incorporacin del dixido de cloro (D) a las secuencias de
blanqueo, se permite blanquear con ms facilidad. Ms tarde, el perxido de hidrgeno (P)
permite sustituir al hipoclorito o reducir el dixido de cloro, disminuyendo as el impacto
medioambiental.
La necesidad de reducir la contaminacin de las fbricas de pastas ha llevado al
estudio de nuevas secuencias de blanqueo. Las investigaciones se han orientado en cinco
direcciones principales (Brown y Abbot, 1995; Bryant y col., 1993; Dence y Reeve, 1996;
Garca y Vidal, 1984; Guerra y Ferraz, 2005; Loras, 1990; Mielisch y col., 1996; Petlicki y
col., 2005):

Utilizacin de secuencias clsicas con reciclado mximo de los efluentes,


tendiendo a la consecucin de fabricaciones sin efluentes.

Utilizacin de secuencias de blanqueo que no incluyan compuestos clorados


en las primeras etapas, cuyos efluentes sin cloruros pueden ser reciclados en
los circuitos de la instalacin fabril.

Nuevos procesos de blanqueo, utilizando reactivos sin cloro elemental (ECF).

Nuevos procesos de blanqueo totalmente libres de cloro (TCF).

Blanqueo biolgico.

En la utilizacin de secuencias de blanqueo sin compuestos clorados (con cloro libre),


las primeras etapas se llevan a cabo principalmente con la pretensin de sustituir, por lo
menos parcialmente, las etapas CE (cloro - extraccin alcalina), causantes del 80 % de la contaminacin. El blanqueo en las etapas posteriores se hace con reactivos convencionales sin
un consumo excesivo de dixido de cloro en la primera etapa. Las investigaciones se han
78

Introduccin
dirigido hacia la utilizacin, en la primera etapa, adems de dixido de cloro, de oxgeno,
ozono, perxido de hidrgeno, dixido de nitrgeno, etc. Si la sustitucin de las etapas
cloradas se hace por componentes sin cloro, los efluentes no portarn cloruros y se pueden
reciclar, para el lavado de la pasta cruda por ejemplo.
Los nuevos procesos de blanqueo sin cloro elemental consisten en la sustitucin total
de las etapas cloradas por compuestos como el dixido de cloro (sin cloro elemental),
independientemente de que se utilicen adems otros agentes de blanqueo totalmente libres
de cloro, como el oxgeno, perxido de hidrgeno, etc.
La utilizacin de procesos totalmente libres de cloro (TCF) emplea reactivos como el
oxgeno, perxido de hidrgeno y ozono, principalmente. Estos tipos de blanqueo son los
que se estudian en este trabajo.

3.2.3.2. Blanqueo con oxgeno


El oxgeno es un reactivo de blanqueo ampliamente utilizado en la produccin de
pastas celulsicas para papel. Su aplicacin a escala industrial fue lenta debido a sus bajos
resultados respecto a la degradabilidad de los carbohidratos y a las propiedades de resistencia de las pastas. Con la aplicacin de compuestos de magnesio y silicatos como
protectores de los carbohidratos, se ha implantado este blanqueo a nivel comercial, como
etapa previa a una secuencia convencional (Jimnez, 2005).
La baja solubilidad del oxgeno en las lejas exige su utilizacin a presin elevada, con
el incremento en el coste de las instalaciones que esto conlleva; sin embargo, los efluentes de
blanqueo con oxgeno no contienen compuestos organoclorados ni cloruros, siendo fciles
de depurar o reciclar, sin que presenten problemas de corrosin en los circuitos de
recuperacin.
En cuanto a las variables ms importantes del proceso de blanqueo con oxgeno,
destacan (Garca y Vidal, 1984; Vidal y Colom, 1986):
o Alcalinidad. Necesaria para alcanzar una velocidad de reaccin aceptable sin
que la presin requerida para la solubilizacin del oxgeno sea excesiva. El
lcali que normalmente suele utilizarse es el hidrxido sdico y su
concentracin vara dependiendo del tipo de pasta, grado de blancura deseado
y consistencia. A mayor concentracin de lcali utilizada, la deslignificacin y la
blancura se incrementan, aunque el grado de polimerizacin de la celulosa
79

Introduccin
disminuye, an en presencia de protectores, as como las propiedades de
resistencia de las pastas.
o Presin. No son necesarias presiones elevadas y el consumo de oxgeno est
relacionado con la reduccin del nmero kappa deseado y con la
concentracin de hidrxido sdico empleada.
o Temperatura. Esta variable est directamente relacionada con la deslignificacin y la degradacin de los carbohidratos, debiendo estar comprendida entre
80 y 140 C.
o Tiempo de reaccin. Su valor oscila entre 15 y 240 minutos a escala de laboratorio, no debiendo exceder de los 60 minutos a escala industrial.
o Consistencia. Con un incremento en esta variable se reduce la transferencia del
oxgeno hacia el licor, facilitando su transferencia desde la fase gaseosa al
interior de las fibras y aumentando, como consecuencia, la deslignificacin. El
incremento excesivo en la consistencia aumenta el consumo de hidrxido
sdico, siendo oportuno operar con una consistencia media del 8-12 %, que
garantice una velocidad de deslignificacin adecuada.
o Aditivos. Tienen como funcin principal proteger las fibras celulsicas contra
la oxidacin o la despolimerizacin alcalina sufrida durante el blanqueo;
siendo las sales inorgnicas de magnesio y los silicatos los primeros en
utilizarse como protectores al retardar la degradacin de la celulosa (Gaspar y
col., 2003); su aplicacin disminuye el consumo de lcali durante el blanqueo
si estas sales se encuentran en presencia de iones frrico. Tambin se utilizan,
como protectores, compuestos orgnicos entre los que destacan el tetraacetato
de etilendiamina (EDTA), el pentaacetato de dietilentriamina (DTPA), la
trietilenamina, compuestos organofosforados, etc.
En el mecanismo de deslignificacin con oxgeno slo intervienen los grupos fenlicos
de la lignina, originados mediante los enlaces ter. Diversos estudios sobre modelos de
lignina sugieren que la reaccin del oxgeno con la lignina comienza por la transferencia de
electrones a grupos fenolato para formar radicales fenoxi, estabilizados por resonancia, no
siendo clara la naturaleza de las reacciones de oxidacin posteriores. Las reacciones del
oxgeno con la lignina son complejas por la gran diversidad de tipos de lignina que existen y
por la gran variedad de especies reactivas oxidantes a que puede dar lugar el oxgeno,
80

Introduccin
producindose una hidrlisis bsica, en la que los iones hidroxilo degradan la lignina y la
solubilizan formando grupos fenlicos. Tambin se producen reacciones mediante especies
inicas como los aniones perxido de hidrgeno, que participan en la eliminacin de
estructuras coloreadas como las quinonas; y las reacciones por radicales (Jimnez, 2005;
Rodrguez, 2002).
La degradacin de los carbohidratos durante el blanqueo se debe en gran medida a
reacciones en cadena originadas por radicales libres, cuyas velocidades estn fuertemente
influenciadas por la presencia de perxido de hidrgeno y de iones metlicos de transicin.
Una extraccin alcalina de la pasta cruda, previa al blanqueo con oxgeno, deslignifica
un 25 % ms que el simple blanqueo con oxgeno. Este pretratamiento influye en las
propiedades de las pastas, de forma que son algo ms difciles de refinar y tienen un ligero
mayor ndice de desgarro y una ligera menor longitud de ruptura; para unas propiedades
fsicas determinadas, la pasta extrada con lcali tiene un nmero kappa ms bajo (Chen y
col., 1995).

3.2.3.3. Blanqueo con ozono


La utilizacin del ozono ha sido estudiada como agente blanqueante en la produccin
de pastas, y en la depuracin de lejas residuales de los procesos de pasteado y de blanqueo
(Jimnez y col., 1996,1997).
A pesar de conocerse su capacidad deslignificante desde hace mucho tiempo, no es
hasta 1968 cuando se descubre su efecto favorable sobre la calidad de las pastas mecnicas y,
por tanto, numerosos estudios son realizados para evaluar las variaciones de las propiedades
de resistencia y de otras caractersticas de las pastas cuando son tratadas con ozono.
El ozono es un oxidante de la mayor parte de los constituyentes de las pastas, por lo
que el propsito de muchos estudios es la identificacin de las condiciones de operacin
para minimizar las reacciones con los carbohidratos, que causan una prdida importante de
las propiedades de resistencia y de viscosidad de las pastas.
El poder blanqueante del ozono es relativamente pequeo para las pastas de maderas
conferas, a pesar de ser un fuerte oxidante. Puede producir incluso disminucin de la
blancura tras el tratamiento, debido a la reduccin del coeficiente de dispersin de luz
causado por un fortalecimiento de los enlaces entre fibras. Por el contrario, las pastas de
maderas frondosas s se pueden blanquear bien con ozono.
81

Introduccin
El blanqueo con ozono no slo no produce compuestos organoclorados en los
efluentes, sino que adems stos no son corrosivos y pueden reciclarse completamente.
Respecto al oxgeno, el ozono es menos contaminante, ya que el color de los efluentes es un
60 % menor y la DQO es apreciablemente ms baja, si bien la DBO es del mismo orden
(Bokstrom y Tuomi, 2001).
El ozono ha sido utilizado para el blanqueo de pastas, solo o en combinacin con
otros agentes de blanqueo en mltiples etapas, con la idea de remplazar las etapas de
deslignificacin que usan cloro.
Diversos compuestos orgnicos (alcoholes, azcares, aminas, etc.) e inorgnicos (sales,
cidos, hidrxidos, etc.) son utilizados en etapas previas a la del ozono o como aditivos en
dicha etapa; stos tienen la finalidad de proteger la celulosa y conducir la reaccin del ozono
slo hacia la destruccin de la lignina. Algunos aditivos orgnicos, como el metanol, ayudan
a preservar la viscosidad durante la deslignificacin con ozono; en algunos casos, como el
del nitrobenceno se incrementa la velocidad de reaccin, sin incrementar la selectividad en
el ataque de la lignina.
La utilizacin de dos o ms etapas de ozono, separadas por etapas de lavado, reduce el
consumo de ozono requerido, siendo menor la prdida de viscosidad y mejorando las
propiedades de resistencia de las pastas. Los resultados de blanqueo son ms eficaces
cuando en la primera etapa de ozono se utiliza la mayor parte del ozono total de un
tratamiento con dos etapas de ozono (65 %) (Byrd y col., 1992).
La mejor selectividad del ozono hacia la lignina ocurre cuando la pasta cruda posee un
menor contenido de sta, lo que indica la necesidad de una fuerte deslignificacin previa al
blanqueo.
La incorporacin de una etapa de ozono a una secuencia ECF o TCF consigue que se
consuman menos reactivos de blanqueo; sin embargo, los valores de ndice de traccin,
ndice de estallido, porosidad y opacidad son menores. La carga de ozono y la consistencia
durante el blanqueo afectan mucho al consumo de reactivos y a la calidad de la pasta final.
Los tratamientos con ozono afectan a las propiedades de resistencia de las pastas,
mejorando la flexibilidad de las fibras, hacindolas ms compresibles, aumentando su
capacidad de enlace mecnico, e incrementando la capacidad de enlace qumico por
formacin de grupos carbonilo y carboxilo, que aumentan el nmero de puentes de
hidrgeno entre las fibras. El incremento del enlace mecnico consigue una mayor
82

Introduccin
resistencia en hmedo inicial de las pastas, as como unos valores mayores para la densidad
aparente, resistencia a la traccin e ndice de desgarro. El incremento del enlace qumico
contribuye a unos mayores valores para los ndices de traccin y de desgarro (Jimnez,
2005).
Las propiedades de resistencia de las pastas mecnicas obtenidas con ozono mejoran
en general; sin embargo, los efectos sobre la blancura, estabilidad de la blancura y opacidad
son negativos, por lo que estas pastas no son adecuadas para papeles de impresin.
Las propiedades de resistencia de las pastas no se afectan cuando una etapa de ozono
se incluye en una secuencia de blanqueo convencional, pero la viscosidad disminuye para un
determinado valor de ndice de estallido. Adems, el uso de una consistencia baja favorece
unos mejores resultados de las propiedades de resistencia frente al uso de una alta
consistencia.
El ozono reacciona con las estructuras fenlicas de la lignina rompiendo los anillos
fenlicos y produciendo compuestos con cadenas cidas o steres. Tambin puede actuar
sobre las cadenas laterales o sobre los ncleos aromticos de la lignina. Existen varias
reacciones indeseables del ozono con los polisacridos (dando lugar a prdidas de
viscosidad) que se clasifican en dos grupos principales: las reacciones de rotura de puentes
glucosdicos y las de oxidacin de grupos hidroxilo a carbonilo.
Las reacciones del ozono con las fibras celulsicas transcurren con ms velocidad
cuando se realizan en fase gaseosa. En estas condiciones las etapas controlantes de la
velocidad de reaccin son las de difusin gaseosa en la pelcula que rodea las fibras y la
adsorcin del reactivo en la superficie de las fibras. Una agitacin adecuada disminuye la
resistencia que ofrece la pelcula gaseosa al decrecer su espesor.
Los radicales hidroxilo generados durante el blanqueo con ozono son los responsables
principales de la degradacin de los carbohidratos (Chirat y Lachenal, 1997).

3.2.3.4. Blanqueo con perxido de hidrgeno


El perxido de hidrgeno puede utilizarse como agente de blanqueo de pastas, solo o
en secuencias ECF (libres de cloro elemental) o TCF (totalmente libres de cloro) (Dence y
Reeve, 1996; Garca y Vidal, 1984; Loras, 1990). Cuando las condiciones de operacin son
suaves el perxido de hidrgeno puede carecer de efecto deslignificante, hacindolo
atractivo para el blanqueo de pastas de alto rendimiento (mecnicas y semiqumicas); en
83

Introduccin
cambio, cuando las condiciones de operacin son ms severas acta como deslignificante,
pudindose aplicar a pastas qumicas, consiguiendo pastas de blancura elevada, muy estables
y poco degradadas.
Las ventajas del empleo del perxido en la industria papelera son su fcil
manipulacin y aplicacin, gran versatilidad, poca volatilidad, fcil transporte y almacenamiento y la naturaleza relativamente no txica de sus productos de reaccin, siendo su
principal problema su fcil descomposicin, por lo que debe aplicarse en condiciones
alcalinas controladas.
El perxido puede aplicarse en la ltima etapa de una secuencia de blanqueo, para
aumentar la blancura final y la estabilidad de las pastas; en una etapa de extraccin alcalina,
para disminuir el amarilleamiento de las pastas provocado en esta etapa, reducir el consumo
de dixido de cloro en las ltimas etapas de blanqueo y mejorar la calidad de los efluentes,
principalmente el color; y en la primera etapa de una secuencia de blanqueo, para reducir o
eliminar la cloracin convencional, con la consiguiente disminucin de carga contaminante
de los efluentes (Dence y Reeve, 1996; Garca y Vidal, 1984; Loras, 1990).
Las variables del proceso que se deben de tener en consideracin son las que siguen:
o Alcalinidad. Dependiendo de las pastas y de las condiciones de tratamiento
con perxido, el pH inicial ha de estar comprendido entre 10 y 12 para un
blanqueo ptimo, llegndose al final a valores comprendidos entre 8,5 y 10. El
pH ha de ser elevado si la consistencia y la temperatura son bajas; siendo ms
alto para las pastas qumicas que para las de alto rendimiento. Tambin el pH
elevado favorece la formacin de carbaniones, sustratos de las reacciones de
oxidacin. Por otra parte, la alcalinidad y la temperatura elevadas aceleran la
descomposicin del perxido, hecho que coincide con una disminucin de la
viscosidad de las pastas, principalmente si se utilizan bajos niveles de aditivos.
La reduccin del nmero kappa y de la viscosidad de las pastas blanqueadas
con perxido depende principalmente de la carga de perxido aplicada, y en
menor medida de la temperatura y de la carga de lcali (Garca y Vidal, 1984;
Jimnez, 2005).
o Consumo de perxido. El consumo de perxido depende, adems del tipo de
pasta, de la carga de sosa, temperatura, pH inicial y consistencia de la pasta
(Garca y Vidal, 1984.). Se ha formulado un modelo para predecir el consumo

84

Introduccin
de perxido durante el blanqueo de pastas mecnicas, con dos sumandos: el
correspondiente al perxido consumido en la reaccin con la pasta y el que
representa el perxido descompuesto mediante la catlisis por metales (Brown
y Abbot, 1995). Para este tipo de pastas mecnicas, un buen grado de
blanqueo se consigue con una concentracin de perxido baja si se opera con
un pH inicial no demasiado elevado y alta consistencia, no influyendo
prcticamente la temperatura. Aumentos de la concentracin de perxido del
2 al 6 % hacen rebajar el pH inicial ptimo para obtener un buen blanqueo.
Para una deslignificacin del 30-40 %, la concentracin de perxido debe ser
de 0,5-1 %, con una carga de sosa del 2-3 %, una consistencia superior al 10 %,
a una temperatura de 80-100 C durante un tiempo de 1-2 horas. Por otra
parte, es conveniente que quede cierta cantidad de perxido residual, que es
funcin de la concentracin de perxido inicial, consistencia de la pasta,
temperatura y tiempo; usualmente se deja una dcima parte del aadido
inicialmente (Garca y Vidal, 1984; Lpez y col., 2010b).
o Consistencia. La consistencia de las pastas durante el blanqueo con perxido
debe ser alta, 10-20 %, para que se eliminen iones metlicos junto con el agua
(Garca y Vidal, 1984), a la vez que permite que las concentraciones locales de
perxido permanezcan por encima de la concentracin crtica (0,016 M); por
debajo de ella se favorece la oxidacin, mediante el oxgeno disuelto en el
agua, de iones Mn (II) y Mn (III) a MnO2, compuesto que ocasiona prdida

de blancura en las pastas (Petlicki y col., 2005; Wekesa y Ni, 2003a). En pastas
termomecnicas (TMP) se demuestra, sin embargo, que valores altos de
consistencia (superior a 10 %) reducen la blancura final hasta en 15 puntos
(Guerra y Ferraz, 2005).

o Temperatura y tiempo. La temperatura de blanqueo para el caso de pastas


kraft, cuando se blanquea en mltiples etapas, debe estar comprendida entre
60 y 80 C, durante un tiempo de 2 a 4 horas, aunque tambin se ha sealado
que la temperatura debe ser de 40 a 60 C y el tiempo de 1 a 3 horas (Garca y
Vidal, 1984). La cada de la viscosidad de las pastas durante el blanqueo con
perxido es menos significativa si la temperatura es baja y se emplean aditivos
(Garca y Vidal, 1984).

85

Introduccin
La evolucin de la deslignificacin que provoca el perxido con el tiempo
muestra una rpida variacin al comienzo, que se incrementa a altas
temperaturas y mayores concentraciones iniciales de perxido, atenundose
despus hasta alcanzar un valor prcticamente estacionario, que depende
bsicamente de la dosis aplicada de perxido. Las temperaturas elevadas
disminuyen el rendimiento y el grado de polimerizacin de los carbohidratos;
si bien no ejercen una influencia significativa en el grado de deslignificacin
final alcanzado por la pasta, permiten conseguirlo sin embargo en menor
tiempo (Garca y Vidal, 1984). Un tiempo de tratamiento demasiado largo,
bajo condiciones que impidan la presencia de perxido residual en la leja de
blanqueo, puede dar lugar a una reversin de la blancura por accin del lcali
(Garca y Vidal, 1984; Petlicki y col., 2005).
o Aditivos. Para evitar la dismutacin del perxido (con formacin de oxgeno),
que perjudica las fibras celulsicas y puede estabilizar algunos grupos
cromforos en las pastas deslignificadas, se han de eliminar ciertos iones
metlicos y enzimas que catalizan la dismutacin. Ello se consigue mediante
aditivos como agentes tampones y estabilizantes. La adicin de aditivos
estabilizantes hace ms selectiva la deslignificacin, consiguiendo mayor grado
de eliminacin de la lignina (Mielisch y col., 1996). A igualdad de nmero
kappa y sin afectar significativamente al rendimiento, un aumento de la dosis
de estabilizante proporciona pastas con un mayor grado de blancura y
viscosidad, a niveles constantes de las dems variables (Garca y Vidal, 1984).
De los aditivos ms conocidos, destacan los silicatos con pequeas cantidades
de Mg (II), que forman una suspensin coloidal que absorbe los iones
metlicos (Bryant y col., 1993; Garca y Vidal, 1984). El efecto estabilizante de
los silicatos es probablemente debido a la desactivacin de partculas coloidales
finamente dispersas de hidrxidos metlicos y xidos hidratados que
descomponen el perxido en agua y oxgeno (Gierer y col., 1993). Los silicatos
adicionados no evitan apreciablemente la formacin de radicales hidroxilo que
tienen un papel relevante en el blanqueo con perxido (Hobbs y Abbot,
1994b), con un efecto positivo en la blancura final debido a la rotura de la
matriz rgida de la lignina, lo que facilita la penetracin de los agentes de
blanqueo (Gierer y col., 1993). El efecto favorable de las sales de magnesio en

86

Introduccin
el blanqueo con perxido en medio bsico puede resultar de la accin del Mg
(II) que estabiliza el perxido, de la accin del hidrxido magnsico que adsorbe los metales pesados, de la formacin de complejos Mg-Fe-cido aldnico (o cido sacarnico), y de la formacin de complejos del Mg (II) con las
formas oxidadas de las unidades glucopiranasa, bajo la forma de enodiol,
estabilizando el enlace glucosdico e impidiendo as la despolimerizacin
(Garca y Vidal, 1984). Se ha demostrado que la mayor parte de la descomposicin del perxido usado en etapas de blanqueo de pastas qumicas se debe
a reacciones con los metales de transicin presentes en la pasta, en lugar de los
presentes en la fase lquida (Lapierre y col., 2003). Tambin se ha estudiado la
estabilizacin de iones Fe (II) y Mn (II) utilizando precipitados de magnesio y
algn polmero aninico (Liden y Ohman, 1997).
La eliminacin de metales es particularmente importante cuando el blanqueo
con perxido se produce en la primera etapa de una secuencia de blanqueo,
necesitando entonces pretratar la pasta con un cido. Cuando el blanqueo se
realiza en otras etapas, los tratamientos realizados en etapas anteriores
normalmente han eliminado la mayor parte de los metales, por lo que los
aditivos para su eliminacin son menos necesarios (Garca y Vidal, 1984). Un
pretratamiento cido puede remplazar el tratamiento con un secuestrante en el
blanqueo de pastas con perxido sin causar disminucin de blancura y
viscosidad de las pastas, pero es necesario optimizar el pH, el tiempo y la
temperatura y adicionar iones de magnesio en la etapa de perxido (Bouchard
y col., 1995; Lapierre y col., 1997; Qiu y col., 2005).
La adicin de nitrato alumnico durante el blanqueo de pastas de eucalipto con
perxido favorece el proceso, debido, al parecer, a que los iones de aluminio
modifican el comportamiento cataltico de los metales de transicin presentes
en la pasta, especialmente del manganeso, reduciendo el consumo de perxido
y aumentado la blancura final (Hobbs y Abbot, 1994a).
Tambin se ha propuesto la alternativa de eliminacin de iones metlicos
mediante pretratamientos con cido, utilizando el cambio inico (Bryant y
Edwards, 1996).
Otros aditivos pueden ser algunos compuestos orgnicos como el tetraacetato
de etilendiamina (EDTA), el cido N-hidroxietilen diamina triactico
87

Introduccin
(HEDTA), el pentaacetato de dietilentriamina (DTPA), el triacetato de nitrilo
(NTA), trietanolamina y otros como los poliacrilatos, organofosforados y
polilactonas (Garca y Vidal, 1984).
Tambin se ha utilizado el aditivo 1,5 diacetil-2,4 dioxohexahidro-1,3,4 triacina
(DAHT), encontrndose las condiciones ptimas cuando la concentracin es
de 0,2-0,6 % respecto a la celulosa seca, a un pH de 11 y 70 C de temperatura
(Lliskovic y Kules, 1992). Igualmente se han conseguido buenos resultados
cuando se utiliza dietilenotriaminapentametileno-cido fosfnico (DTPMPA)
si se mantiene el pH entre 7 y 10 (Hassan, 2003; Prasakis y col., 1996).
Ms recientemente se ha estudiado la adicin de dietilenotriaminametilenocido fosfnico, DTMPA, en el blanqueo de pasta kraft de pino con perxido
y oxgeno molecular (Kadla y col., 1999), as como el efecto del EDTA en el
blanqueo de pastas kraft de maderas conferas durante el blanqueo con
perxido, con objeto de eliminar iones de manganeso, hierro y cobre (Brelid y
Friberg, 1998); igualmente se ha estudiado el efecto de la presencia de DTPA
y/o Mg (II) durante el blanqueo de pastas kraft de maderas conferas (Gartner
y Gellerstedt, 2000). Otra posibilidad sera el uso de zeolitas como
secuestrantes de metales de transicin para inhibir la descomposicin del
perxido; principalmente se ha estudiado su uso en el blanqueo de pasta
mecnica (TMP) y pasta qumica kraft (Dyhr y Sterte, 1998,1999; Finnegin y
col., 1998). En pastas mecnicas se obtienen resultados de blancura similares a
los obtenidos cuando el secuestrante es el DTPA (Leduc y col., 2005).
Tambin se ha estudiado la utilizacin de boro-aminas en el blanqueo de
pastas con perxido, encontrando que son ms efectivas que el borohidruro
sdico y que el hidrosulfito sdico (Pedneault y col., 1999; Pellerin y col.,
2000). Han aparecido trabajos que utilizan un pretratamiento con DTPA para
eliminar los complejos que forman los taninos con el hierro y que son
causantes de oscurecimiento de pastas mecnicas de pino y abeto (Friman y
col., 2004), as como para inhibir la actividad cataltica de Mn (II) y Mn (III); se
comprueba que existe mayor efectividad del DTPA sobre el Mn (II), por lo
que se aconseja su uso combinado con borohidruro sdico, en condiciones
bsicas, que reduce el Mn (III) a Mn (II) (Qiu y col., 2003); buscando esta

88

Introduccin
reduccin tambin se puede eliminar el aire/oxgeno disuelto purgando el
sistema con nitrgeno, evitando as la formacin de Mn (III) (Ni y Qiu, 2003).
Otra posibilidad para disminuir la actividad cataltica de los iones de Mn es el
uso de silicatos de sodio en presencia de precipitados de carbonato clcico
(Wekesa y Ni, 2003b). Un estudio riguroso de las reacciones que tienen lugar
entre los iones de manganeso y el perxido, se puede encontrar en el trabajo
de Wekesa y Ni (2003a).
Las principales reacciones del perxido son las que tienen lugar con la lignina y con
los carbohidratos de las pastas (Dence y Reeve, 1996; Garca y Vidal, 1984; Jimnez, 2005;
Loras, 1990).
El efecto blanqueante del perxido se atribuye a la accin oxidante del HOO-,

procedente de la disociacin del perxido, sobre los grupos carbonilo conjugados (p- y oquininas, coniferil aldehdos) para producir aldehdos y cidos carboxlicos incoloros. La
concentracin de HOO- aumenta con la temperatura y con el pH, por lo que el blanqueo se
debe realizar en condiciones alcalinas.

La cintica del blanqueo de pasta mecnica con perxido de hidrgeno viene dada
por:
-dCc/dt = K (H2O2)1,0 (OH-)0,3 Cc5

donde Cc es la concentracin de cromforos, por lo que slo depende de las concentraciones de stos y de la de perxido, del pH y de la temperatura, no influyendo
prcticamente los aditivos como silicatos y sales de magnesio.
Un modelo cintico que explica el blanqueo con perxido en medio bsico supone
que se establece un equilibrio cuando la velocidad de eliminacin de grupos cromforos se
iguala a la de formacin de los mismos. Este modelo justifica las relaciones observadas entre
el grado de blancura y el tiempo, as como la respuesta de la pasta cuando la concentracin
de perxido cambia (Ginting y Abbot, 1993).
Tambin se ha propuesto otro modelo cintico para el blanqueo de pastas mecnicas,
encontrndose que el orden de reaccin del anin perxido dentro de la pared de las fibras
es menor que la unidad, lo que puede atribuirse a la adsorcin de perxido sobre la
superficie de los poros de las fibras (Been, 1995).

89

Introduccin
Para el caso de pastas qumicas se ha llegado a la conclusin de que el mecanismo del
blanqueo es ms complejo que para las pastas mecnicas, no pudindose explicar por una
simple ecuacin cintica, lo que puede ser debido a que la lignina ha sido ms o menos
modificada. Se comprueba que en las dos primeras horas de blanqueo es cuando existe un
mayor consumo de perxido, produciendo un gran aumento de la blancura y una pequea
despolimerizacin, que aumenta ms despus de este tiempo (Garca y Vidal, 1984).
Funazukuri y col. (1995) proponen un modelo simple en el que la velocidad de blanqueo se
expresa como una cintica de primer orden con una energa de activacin de 29,4 KJ/mol.
Tambin se ha propuesto un modelo de dos etapas para el blanqueo con perxido en
una secuencia TCF para pastas obtenidas con el proceso Alcell (Ni y col., 1997): la primera
muy rpida, y la segunda, basada en la destruccin de grupos cromforos, responde a la
expresin:
-dC/dt = 1,41 1010 exp(-69,610/(RT)) (HOO-)0,5(C-Cf)2

donde C significa grupos cromforos y t tiempo.

Para el caso del Eucalyptus grandis se ha estudiado la influencia de la concentracin

de perxido y la concentracin de hidrxido sdico en el blanqueo, en dos etapas, de pasta


quimicomecnica (Area y col., 1997). Se comprueba que el consumo de perxido en la
primera etapa vara linealmente con la carga inicial; la blancura y la resistencia mecnica de
las pastas incrementan linealmente con la carga inicial de hidrxido sdico. En la segunda
etapa la blancura vara con el cuadrado de la carga inicial de perxido.
El efecto del blanqueo puede atribuirse a la reaccin del perxido con algunos grupos
carbonilo y grupos fenlicos libres de la lignina. Tambin se ha propuesto que el mecanismo
de reacciones de la lignina de conferas y el perxido incluya desmetilacin, formacin de
quinonas, rotura de cadenas laterales y apertura de anillo, segn una serie de reacciones
secuenciales o competitivas que individualmente pueden tener una influencia adversa o
favorable sobre el color de la pasta. Parece ser que slo los grupos fenlicos no eterificados
o los monmeros con cadenas laterales con grupos carbonilo, son susceptibles de ser
atacados por el perxido; por tanto, poca lignina se degrada y se solubiliza, por lo que el
rendimiento disminuye muy poco. Para el caso de la lignina de frondosas, slo los ncleos
siringilo no eterificados reaccionan con el perxido, aunque ms fcilmente que los grupos
guayacilo correspondientes de las conferas (Garca y Vidal, 1984).

90

Introduccin
Se ha confirmado que junto al perxido acta tambin el oxgeno (formado por
dismutacin del primero) sobre la lignina, por lo que debido a la accin simultnea o
secuencial de ambos oxidantes se dificulta la interpretacin qumica del proceso, complicado
adems por las reacciones no oxidantes promovidas por el lcali, que inician los procesos de
oxidacin y cooperan o compiten con ellos. La accin del oxgeno origina una
fragmentacin y/o formacin de estructuras cromofricas conjugadas, mientras que la accin
del perxido implica la eliminacin de grupos cromforos sin una apreciable degradacin.
Para el caso de pastas qumicas y semiqumicas las reacciones entre los grupos
cromforos de la lignina y el perxido son posiblemente similares a los de las pastas
mecnicas, pero los efectos observados estn fuertemente influenciados por el tipo y
condiciones de la coccin.
En presencia de iones metlicos, el perxido y los hidroperxidos formados en las
reacciones de oxidacin de la lignina y de los carbohidratos producen una serie de radicales
libres que actan como poderosos oxidantes capaces de degradar celulosa y lignina, por lo
que el perxido se comporta como un agente deslignificante. Para esta deslignificacin no se
precisan los silicatos ni otros estabilizantes, slo sulfato magnsico para proteger la celulosa
de la despolimerizacin. Los iones metlicos ms importantes causantes de la formacin de
radicales son los de hierro, cobre, cobalto, nquel y manganeso. Segn la bibliografa parece
ser que afectan de forma diferente a la degradacin de los carbohidratos y a la deslignificacin. El cobre acta indiscriminadamente sobre las velocidades de deslignificacin y
de degradacin de los carbohidratos, el hierro y el cobalto degradan la celulosa sin acelerar
la deslignificacin y el manganeso cataliza especialmente la degradacin de la lignina (Garca
y Vidal, 1984).
En el blanqueo de pastas mecnicas bajo condiciones suaves las reacciones de los
carbohidratos son insignificantes. En las pastas qumicas son algo ms intensas, pero no
severas, como se deduce de la poca prdida de viscosidad de las pastas. Mediante oxidacin
radical de grupos terminales reductores de los carbohidratos por el oxgeno se evitan
reacciones de peeling, lo que causa pocas prdidas de rendimiento. La despolimerizacin
por rotura de la cadena del polisacrido se debe en gran parte a la accin de las especies
radicales sobre la glucosa de las cadenas celulsicas. La cada del grado de polimerizacin de
la celulosa resulta particularmente perjudicial para las caractersticas fsico-mecnicas de las
pastas despus del blanqueo. La utilizacin de un agente protector permite remediarlo en
gran medida (Garca y Vidal, 1984).
91

Introduccin
Ms recientemente se han estudiado las reacciones del perxido durante el blanqueo
de pastas con los grupos cromforos de la lignina (Pan y col., 2000a,b).
Para finalizar, se ha de sealar que tambin se han realizado diferentes estudios
cinticos relacionados con el blanqueo con perxido de pastas obtenidas mediante el
proceso Alcell (Schroeter, 1991), con el blanqueo de pastas kraft-antraquinona de pino con
perxido en presencia de un catalizador complejo de manganeso (Chen y col., 2003; Cui y
col., 2000), y con el blanqueo de pastas mecnicas con perxido (Xu, 2000a,b).

3.2.3.5. Blanqueo con enzimas


Con la implantacin de la biotecnologa se busca el desarrollo de tecnologas limpias
que optimicen los procesos tradicionales de pasteado y blanqueo en la industria papelera,
logrando un menor consumo de reactivos y una generacin de efluentes con menor carga
contaminante.
En la actualidad, la aplicacin de enzimas en los tratamientos de blanqueo aporta
numerosas ventajas, tanto ecolgicas como tcnicas, al disminuir la contaminacin por un
menor consumo de reactivos clorados en secuencias ECF, que mejora la economa del
proceso; y al aumentar la selectividad hacia la lignina en secuencias TCF, que mejora la
calidad de las pastas al obtenerse mayores blancuras.
Las enzimas son catalizadores proteicos producidos a partir de microrganismos vivos,
que aceleran la velocidad de las reacciones qumicas y son altamente especficos del tipo de
reaccin que catalizan. Sus aplicaciones en el campo de la biotecnologa de la biomasa
lignocelulsica, se basan en la utilizacin de estas enzimas en distintas partes del proceso
para degradar o eliminar un compuesto concreto. La aplicacin directa de las enzimas
mejora el proceso debido a su alta especificidad, evitando la degradacin de otros
compuestos. Las enzimas se clasifican en funcin del polmero lignocelulsico al que
ataquen en: celulasas, hemicelulasas y ligninasas o enzimas ligninolticas.
El objetivo actual del tratamiento enzimtico no es sustituir al blanqueo qumico, sino
reducir su uso para poder aprovechar todas las ventajas que aportan estos tratamientos, sin
afectar al grado de blancura y a las propiedades mecnicas de las pastas celulsicas (Martn,
2011).
Varias razones han motivado la aplicacin de enzimas en el proceso de blanqueo. Por
un lado, existen razones medioambientales, como la reduccin de la carga contaminante
92

Introduccin
derivada de la disminucin en el consumo de reactivos clorados (en secuencias ECF). Por
otro lado, las razones tcnicas incluyen el aumento de selectividad hacia la lignina (en
secuencias TCF), la economa de reactivos de blanqueo y la mejora de la calidad de las
pastas al obtenerse mayores blancuras. Sin embargo, la eficacia de los tratamientos
enzimticos es, hoy da, limitada y el objetivo actual no es sustituir completamente el
blanqueo qumico, sino reducir su uso para limitar la contaminacin hasta niveles asumibles
y reducir el coste de reactivos qumicos para hacer rentable la inversin. Al mismo tiempo,
se pretende que la blancura y las resistencias mecnicas de las pastas no se resientan.
La aplicacin directa de microrganismos tales como los hongos ligninolticos que se
emplean en el biopulpeo, tendra consecuencias indeseables en el bioblanqueo, como la
degradacin de la celulosa o la contaminacin de las pastas con sustancias que acompaan o
son producidas por los hongos. Por ello, la bsqueda, seleccin y posible modificacin de
enzimas implicadas en la degradacin de la lignocelulosa para adecuarlas a las condiciones
industriales del blanqueo de pastas, representa uno de los campos de investigacin con
mayor potencial en esta industria. El coste de las enzimas se ha ido reduciendo a medida
que el bioblanqueo abra nuevas expectativas y no parece suponer un impedimento para el
futuro.
El empleo de enzimas ligninolticas, como las peroxidasas y lacasas, fue adquiriendo
mayor importancia en el bioblanqueo; las lacasas son enzimas que usan oxgeno molecular
como sustrato y lo reducen hasta agua; son enzimas del tipo fenoloxidasa, que contienen
cuatro tomos de cobre en su estructura, uno de los cuales est implicado en la oxidacin de
sustratos, mientras el resto participan en el transporte de electrones hasta el oxgeno. El
elevado peso molecular de estas enzimas slo les permite oxidar compuestos fenlicos. Sin
embargo, en presencia de mediadores redox son capaces de oxidar unidades no fenlicas,
siendo estos mediadores (compuestos de bajo peso molecular) capaces de difundir mejor en
la fibra y reaccionar con las fracciones de lignina a las que la enzima no tiene acceso debido
a su elevado peso molecular. Esta cooperacin de la enzima y los mediadores es lo que se
denomina sistema lacasa-mediador, en el cual la enzima oxida un mediador qumico que, a
su vez, oxida la lignina. Desde su descripcin, los sistemas lacasa-mediador han sido
investigados con xito para la deslignificacin y blanqueo y una disminucin de la carga
contaminante de los efluentes, mientras que las propiedades mecnicas de las pastas se
mantienen o mejoran (Bourbonnais y Paice, 1996; Camarero y col., 2004; Eugenio y col.,
2010; Ibarra y col., 2006; Moldes y col., 2008). Sin embargo, el proceso an no se ha

93

Introduccin
implantado a escala industrial probablemente debido a las dificultades derivadas del uso del
mediador y a la falta de optimizacin del proceso. El elevado coste y los posibles problemas
de toxicidad de algunos mediadores han dirigido las investigaciones a la bsqueda de
mediadores alternativos, incluyendo algunos metabolitos fngicos as como productos
resultantes de la degradacin de la lignina como se ha indicado anteriormente (Caas y
Camarero, 2010; Fillat y col., 2010; Moldes y col., 2008).
Pese a que las xilanasas no degradan la lignina, stas fueron las primeras enzimas
utilizadas en el blanqueo de pasta kraft (Viikari y col., 1986) y son hoy da las nicas que se
aplican a nivel industrial. La aplicacin de las xilanasas, previamente al blanqueo, ha
demostrado que es posible alcanzar el nivel deseado de blancura con un ahorro de reactivos
de blanqueo, reduciendo as la contaminacin por cloroligninas. En aos posteriores, otros
trabajos confirmaron la capacidad de las xilanasas para mejorar los efectos de las etapas de
blanqueo (Paice y col., 1988; Ragauskas y col., 1994; Senior y Hamilton, 1992; Viikari y col.,
1994). Las xilanasas atacan la cadena principal de las hemicelulosas, y las -xilosidasas
hidrolizan los xilooligosacridos a xilosa; otras, como las -D-glucurosidasas, rompen las
ramificaciones, y las -L-arabinosidasas hidrolizan los enlaces entre los xilanos y los cidos
acticos y fenlicos. Adems de hidrolizar las hemicelulosas, las hemicelulasas favorecen la
hidrlisis de la celulosa y potencian una posterior deslignificacin debido a la rotura de
enlaces lignina-carbohidratos, a la eliminacin de xilanos reprecipitados que dificultan el
acceso de los reactivos de blanqueo y/o a la eliminacin de regiones de hemicelulosas con
alto contenido en cidos hexenurnicos. El mecanismo de actuacin de las xilanasas ha sido
descrito por diversos autores; de acuerdo a Kantelinen y col. (1993), las xilanasas eliminan
los xilanos precipitados, depositados sobre las fibras y que dificultan el acceso de los
reactivos de blanqueo (Shah y col., 2000); otra explicacin, de acuerdo a Wong y col.
(1997), supone que las hemicelulosas estn unidas a la lignina por enlaces covalentes y que la
accin de la enzima libera estos compuestos permitiendo su extraccin del medio.
Los principales inconvenientes de estos tratamientos se basan en el elevado coste de
los mediadores sintticos y los posibles problemas de toxicidad de algunos de ellos, lo que
dirige nuevas investigaciones hacia la bsqueda de mediadores naturales alternativos como
metabolitos fngicos o productos resultantes de la degradacin de la lignina (vanilina, cido
p-cumrico, etc.).

94

Introduccin
3.2.4. CELULOSA NANOFIBRILAR (NFC): COMPOSICIN Y PRINCIPALES APLICACIONES
El campo emergente de la nanociencia estudia la manipulacin y caracterizacin de
objetos cuyas dimensiones se encuentran entre los diez hasta los cientos de nanmetros
aproximadamente. La nanotecnologa es una disciplina ingenieril que enfoca la nanociencia
al hecho de crear productos o procesos basados en uno o mltiples componentes integrados
a escala nanomtrica (Ozin y col., 2008; Whitesides, 2005).
La celulosa es un polmero natural esencial en los procesos de pasteado y fabricacin
de papel y es la macromolcula ms abundante que existe. Aproximadamente 7,51010

toneladas de celulosa son utilizadas cada ao para la produccin de biomasa celulsica, que
es la fuente renovable ms grande del planeta (Spence, 2011).
La molcula de celulosa (Figura 3.4) consiste en un homopolmero lineal compuesto
de (1,4)--glucopiranosa. El dmero, llamado celobiosa (Figura 3.16), es la unidad de
repeticin de la celulosa. El grado de polimerizacin se determina por el nmero de
unidades de glucosa anhidra, y vara dependiendo del origen de la celulosa. Alrededor de 30
unidades de glucosa anhidra confieren al polmero su estructura caracterstica y las
propiedades de la celulosa (Nehls y col., 1994) pero el grado de polimerizacin puede ser
mayor de 15000 monmeros.

Figura 3.16. Estructura de la celobiosa.

La celulosa nanofibrilar (NFC) se prepara a partir de fibras madereras y alternativas


usando tratamientos mecnicos o una combinacin de procesos qumicos y enzimticos. La
NFC es un material prometedor con una serie de propiedades extraordinarias (Turbak y
col., 1983). En la celulosa nanofibrilar, las fibras celulsicas son desintegradas en sus
unidades subestructurales y microfibras, contando con longitudes dentro de la escala
micromtrica y anchuras que van desde los diez a los cientos de nanmetros (Figura 3.17).
Gracias a la baja relacin existente entre su pequeo dimetro y su larga longitud

95

Introduccin
(anchura/longitud), las nanofibras tienen un rea superficial especfica mayor que la de las
fibras presentes en la celulosa de partida.

Figura 3.17. Desarrollo de fibras celulsicas a partir de molculas de celulosa.

El mayor desafo propuesto para el uso de NFC est asociado con la gran polaridad de
la celulosa. Los numerosos grupos hidroxilo presentes en la celulosa forman fuertes puentes
de hidrgeno entre cadenas de celulosa y a su vez con el agua. En suspensiones acuosas, los
grupos hidroxilo facilitan la dispersin de fibrillas, pero en disolventes no polares,
promueven la formacin de agregados, y, en consecuencia, se pierden los beneficios que
existen al trabajar en la escala nanomtrica. Para evitar la formacin de agregados entre
fibrillas, la NFC debe ser almacenada como suspensin acuosa con menos de un 2 % en
peso del contenido de slidos, lo cual limita muy severamente su utilizacin a escala
industrial (Berglund, 2005).
Como cabe esperar, la nanocelulosa obtenida a partir de pasta maderera es
significativamente ms fuerte que la celulosa presente en la pasta inicial. Esto es debido a
que las microfibrillas individuales presentan mayor resistencia a la traccin, al tiempo que se
incrementa el rea superficial especfica (Taniguchi y Okamura, 1998). Este incremento en
el rea superficial especfica se traduce en la creacin de enlaces ms fuertes y una mayor
cantidad de puentes de hidrgeno. Se cree tambin que la NFC presenta un mecanismo
conocido con el nombre de crack-stopping que se traduce en una mejora de las
propiedades mecnicas (Nakagaito y Yano, 2006). Adems, las propiedades reolgicas de la
NFC permiten su uso como espesantes o estabilizantes de los alimentos, pintura, cosmtica
y suspensiones o emulsiones farmacuticas (Andresen y Stenius, 2007; Turbak y col., 1983).
96

Introduccin
Debido a su abundancia, biodegradabilidad y naturaleza renovable, as como a su
extraordinaria estructura y propiedades fsicas, la NFC ha captado la atencin debido a la
posibilidad de usarla como refuerzo en composites, revestimientos, pelculas, membranas y
en el embalaje de materiales. Adems es un nanomaterial natural que proporciona un
amplio rango de posibilidades para obtener mejores propiedades en diferentes productos
finales. Es ampliamente reconocido que las nanofibras difieren en trminos de comportamiento y funcionalidad respecto a las fibras de celulosa. Debido a sus caractersticas, muchos
de los esfuerzos en investigacin han sido enfocados al estudio de aplicaciones y
propiedades de la NFC. Adems, en la actualidad, la conciencia medioambiental y el
impulso social en el uso de productos sostenibles y respetuosos con el medio ambiente, nos
hace esperar que se incremente el uso de NFC como alternativa a los materiales derivados
del petrleo.

3.2.4.1. Produccin de celulosa micro/nanofibrilada


La celulosa microfibrilar (MFC) fue producida por primera vez en el ao 1983 por
Turbak y colaboradores usando celulosa purificada obtenida a partir de pasta maderera y
con la ayuda de un homogeneizador a alta presin. Sin embargo, la celulosa microfibrilar
puede ser producida usando distintos recursos como pueden ser los residuos de la
agricultura, los procedentes del mundo animal o a partir de la celulosa bacteriana.
La homogenizacin a alta presin ha sido aplicada a muchos tipos de fuentes de
celulosa como por ejemplo la remolacha azucarera, la patata y el pltano. En estos
materiales, las microfibrillas son ms fciles de liberar de la matriz de fibras que las fibrillas
presentes en la madera, requiriendo la desintegracin menos energa. Sin embargo, estos
materiales son heterogneos y estn compuestos de grandes aglomeraciones de fibras
(Dinand y col., 1996,1999; Dufresne y col., 2000; Heux y col., 2000; Lowys y col., 2001;
Zuluaga y col., 2007).
Las fibras celulsicas pueden ser tambin desintegradas en subestructuras mediante el
uso de una vigorosa hidrlisis cida. En la hidrlisis cida las regiones amorfas de
microfibrillas de celulosa son disueltas dando lugar a un material cristalino residual, que
comparado con la NFC tiene una baja relacin anchura/longitud de fibras (Battista y col.,
1956; De Souza y Borsali, 2004; Dong y col., 1996; Revol y col., 1992,1994).

97

Introduccin
Estas fibras semejantes a cristalitos son denominadas microcristales de celulosa,
tambin conocidos con el nombre de whiskers or nanorods (Figura 3.18) Los microcristales de celulosa no presentan las mismas propiedades de tipo gel que presentan las
nanofibrillas y eso es debido a su forma tubular, teniendo una ms baja relacin
anchura/longitud y una superestructura cristalina.
Como se mencion anteriormente, adems de las fibras madereras, las suspensiones
de nanocristales de celulosa han sido preparadas a partir de una gran variedad de fuentes de
celulosa como algodn, celulosa bacteriana y animales marinos. Por ejemplo, la celulosa
bacteriana (Figura 3.19) es producida por Acetobacter xylinum y se caracteriza por tener
fibras de celulosa individuales y largas con una relacin anchura/longitud ms grande que la
presentada por las nanofibras de celulosa obtenidas a partir de madera. La celulosa
bacteriana se presenta en forma de gel y como un material homogneo, y no contiene
lignina, pectinas y hemicelulosas. Las aplicaciones de la celulosa bacteriana son, por
ejemplo, en la fabricacin de nanocomposites (Yano y Nakahara, 2002) y materiales
biomdicos como los apsitos para heridas, soporte de huesos en injertos y estructuras de
tejidos de cartlago y vasos sanguneos (Bodin y col., 2007; Klemm y col., 2001; Svensson y
col., 2005).
(a)

(b)

Figura 3.18. Imgenes de AFM de a) nanocristales de celulosa; b) celulosa nanofibrilar.

El algodn ha sido usado para la produccin de estructuras de celulosa de tamao


nanomtrico, aprovechando su relativamente bajo contenido en hemicelulosas y lignina, en
comparacin con la madera. En cuanto a la celulosa animal, se ha convertido en el primer
material estudiado debido a su relativa pureza, as como a su potencial para producir
whiskers celulsicos prcticamente libres de defectos.

98

Introduccin

Figura 3.19. Imgenes de celulosa bacteriana.

La Figura 3.20 resume las diferentes rutas que se pueden seguir para obtener
nanofibras de celulosa (Pkk y col., 2007). En el mtodo 1, una fuerte hidrlisis cida
conduce a agregados de suspensiones coloidales, altamente cristalinos, y con una baja
relacin anchura/longitud de fibras. Adems de la hidrlisis, la sonicacin (generalmente,
ultrasonidos) debera destruir los agregados de fibrillas para dar lugar a fibrillas de celulosa.
La celulosa microfibrilar se prepara mediante la desintegracin, aplicando fuerzas de ruptura
muy fuertes (mtodo 2), que producen fibrillas conectadas de manera inherente y
enredadas, as como agregados de fibrillas, generndose redes con gran fortaleza mecnica.
El mtodo 3 incorpora la hidrlisis enzimtica, que produce un mezcla entre fibrillas de
celulosa I (de alrededor de 5 nm de espesor) y agregados de fibrillas (de alrededor de 10-20
nm de espesor).
Se han empleado diferentes mtodos mecnicos eficientes para liberar fibrillas de
celulosa de las paredes celulares de los vegetales, como son la homogenizacin,
microfluidizacin, micro-molienda y ruptura criognica. De acuerdo a Spence y col. (2010),
el problema principal que existe cuando se procesan pastas madereras para producir
MFC/NFC con el homogeneizador y el microfluidizador (fluidizador a alta presin), es la
longitud de fibra, que causa que el instrumento se obture debido a que las fibrillas se
enredan. Con el fin de reducir el tamao de fibra y/o para pre-defibrilar las fibras, se han
usado pretratamientos mecnicos, qumicos y enzimticos, lo que se traduce en una
reduccin de la frecuencia con la que se obtura el instrumento. Este tipo de pretratamientos
tambin eliminan la pared celular primaria, en la que las microfibrillas se organizan de
manera aleatoria; y es en la pared celular secundaria donde ms eficientemente las fibras se
localizan de forma ms organizada para su posterior procesamiento (Montanari y col.,
2005).
99

Introduccin

Figura 3.20. Rutas diferentes para obtener celulosa nanofibrilar.

Generalmente, los pretratamientos promueven el proceso de delaminacin, dando


lugar a un consumo de energa ms bajo, mejor estabilidad del proceso y fibrillas ms
pequeas y uniformes. Los pretratamientos qumicos incluyen la hidrlisis cida controlada,
hidrlisis e hinchamiento alcalino, as como modificaciones qumicas de la superficie de la
celulosa. En cuanto a los tratamientos enzimticos, Pkk y col. (2007) demostraron que
bajas concentraciones (0,02 %) de enduclanasa facilitaba la fibrilacin. Adems, la carboximetilacin y la oxidacin usando TEMPO (catalizador usado en presencia de exceso de leja
(NaClO)) como mediador fueron introducidas para mejorar el procesamiento de diversas
maneras, y todas ellas reducen el consumo de energa en el procesamiento.
Como alternativas a la reduccin, por va mecnica, del tamao de la fibra y al
pretratamiento, se encuentran los discos refinadores, molinos PFI, corte manual y las
batidoras Valley (Andresen y Stenius, 2007; Henriksson y col., 2008; Herrick y col., 1983;
Iwamoto y col., 2005; Nakagaito y Yano, 2006; Turbak y col., 1983). Los procesos de
refinado mecnico tienden a aumentar el tamao de los nanoporos entre las fibras
celulsicas (Berthold y Salmn, 1996) pero al mismo tiempo daan la estructura de la
microfibrilla mediante la reduccin de la masa molar y el grado de cristalinidad (Henriksson

100

Introduccin
y col., 2007); adems, este proceso puede ser muy energtico. El refinado adems tiende a
aumentar la capacidad que muestran las paredes de las fibras para enlazarse con el agua,
como por ejemplo el valor de retencin de agua (WRV). El refinado a niveles convencionales da lugar a un incremento en la flexibilidad de las fibras y tambin hace que stas
estn ms preparadas para el colapso, dndole forma a las fibras (Hubbe y col., 2008).
Con respecto al uso del pretratamiento qumico, las hidrlisis cida, alcalina y enzimtica son procesos comunes a los utilizados en la industria papelera y maderera, sin
embargo, otros mtodos como la oxidacin TEMPO no son usados industrialmente y se
presentan como una tecnologa emergente.

3.2.4.2. Aplicaciones de la celulosa nanofibrilar


3.2.4.2.1. Aplicaciones mdicas
La biocompatibilidad, no toxicidad y biodegradabilidad de los nanomateriales
celulsicos son propiedades importantes en las aplicaciones bioqumicas y biomdicas.
Todas estas caractersticas hacen de las microfibrillas celulsicas un material prometedor
para la nanotecnologa. Varios investigadores han mencionado el uso de elementos
nanocelulsicos como estructuras en las cuales un hueso o tejido puede desarrollarse ms
fcilmente. Adems, la comunidad biomdica ha comenzado a explotar las cualidades
hidroflicas de la celulosa con el fin de desarrollar hidrogeles. stos son materiales idneos
para aplicaciones mdicas y farmacuticas como pueden ser la administracin de frmacos,
estructuras para tejidos, sensores, vlvulas, etc. (Hubbe y col., 2008).

3.2.4.2.2. Refuerzo de composites


Las muy buenas propiedades de resistencia, estabilidad trmica y otras propiedades
intrnsecas de las fibrillas (bajo coste, salubridad, renovabilidad) as como el conveniente
procesado (flexibilidad durante la fabricacin) (Berglund, 2005; Kamel, 2007; Nakagaito y
Yano, 2006) hacen que las MFC/NFC sean idneas para su uso como refuerzo en
nanocomposites (Berglund, 2005; Hubbe y col., 2008) y productos de papel (Turbak y col.,
1983). Entre las aplicaciones de las MFC/NFC pueden incluirse: componentes estructurales
en coches, cajas de embalaje, juguetes, almacenamiento de latas y muebles de exterior
(Kokta y col., 1983), as como materiales biomdicos. El uso de la celulosa en la fabricacin

101

Introduccin
de nanocomposites aumenta la resistencia, haciendo posible una reduccin sustancial en el
peso con el mismo rendimiento.

3.2.4.2.3. Embalaje
Otra aplicacin, conocida con el nombre de embalaje de alimentos, ha sido
investigada durante los ltimos aos, principalmente debido a la creciente demanda en la
seguridad, relacin coste/eficiencia y problemas medioambientales del producto, as como el
inters de los consumidores. Para aumentar el rendimiento del embalaje de materiales de
forma que se pueda hacer frente a todo este tipo de demandas, estn siendo desarrollados y
optimizados materiales de embalaje modificados, con el fin de que puedan ser potencialmente comerciables. Hoy en da estos materiales son extensamente producidos usando
plsticos sintticos derivados del petrleo, pero debido a la conciencia medioambiental, los
materiales derivados de fuentes renovables estn siendo fuertemente investigados como
potenciales sustitutos (Spence y col., 2010). Estos materiales deben proporcionar proteccin
para los productos con el fin de conseguir una vida satisfactoria y duradera a los mismos
niveles que aqullos obtenidos usando materiales derivados del petrleo (Rhim, 2007; Rhim
y Ng, 2007).
Las fibras celulsicas, como papel y cartn, han sido tradicionalmente usadas en
embalaje para una amplia variedad de alimentos como productos alimentarios secos,
congelados o lquidos, bebidas e incluso alimentos frescos (Kirwan y Strawbridge, 2003).
MFC y NFC han emergido recientemente como un material potencial de embalaje ya que
muestran propiedades de barrera y mecnicas muy competitivas. Las pelculas formadas a
partir de dicho material han mostrado bajas velocidades de transmisin de oxgeno, 17
mL/m2da, comparables al embalaje sinttico procedente del alcohol-vinil-etileno polister.

Para pelculas de MFC, la porosidad, que es una propiedad a tener en cuenta para el

embalaje, es modificable mediante el secado de diferentes disolventes, creando caractersticas que pueden proporcionar una ventaja sobre los plsticos.

102

Introduccin
3.3. PALMA ACEITERA

3.3.1. GENERALIDADES
Existen tres especies de palma aceitera: la Africana (Elaeis guineensis), la Nol o
Americana (Elaeis oleifera) y la Corozo Colorada (Elaeis odera) (Rodrguez y col., 2008). La
palma aceitera es un cultivo propio de regiones tropicales hmedas, caracterizadas por altas
temperaturas, abundante insolacin y suficiente humedad. Este cultivo se adapta a una
amplia variedad de suelos, aunque se desarrolla mejor en suelos profundos, bien drenados,
frtiles y con abundante materia orgnica. La especie utilizada en este trabajo es la palma
africana, originaria de frica occidental, de la regin del Golfo de Guinea, extendindose
aproximadamente hasta los 15 de latitud norte y sur. Desde tiempos remotos, la planta crece de forma silvestre, siendo su fruto utilizado para la extraccin de aceite para el consumo
humano, especialmente en Guinea Occidental. Despus de los viajes de Coln se introduce
en Amrica y, en pocas ms recientes, a Asia, llegando a Sumatra y Malasia hacia 1900.
Indonesia y Malasia son los mayores productores de aceite de palma con el 85 % de la
produccin mundial, lo que constituye un gran recurso econmico para estos pases.

Figura 3.21. Plantacin de palma.

103

Introduccin
Tambin se est extendiendo el cultivo de la palma (Figuras 3.21 y 3.22) a pases de
frica Occidental (Nigeria, Ghana, etc.), Amrica Latina (Ecuador, Venezuela, Brasil,
Colombia, Honduras, etc.) y Asia (Tailandia).
Debido a su buen rendimiento por hectrea, sus bajos costes de produccin y los
mltiples usos de su aceite, la palma aceitera se ha convertido en la principal fuente de aceite
vegetal del planeta, por delante de la soja, con 37 millones de toneladas producidas en el
ao 2007 (31 % de la produccin mundial de aceite comestible) (United States Department
of Agriculture (FAS), 2010).
Aunque todava existen explotaciones tradicionales en zonas de frica, la palma
aceitera es bsicamente un cultivo industrial orientado a la exportacin (el 72 % de la
produccin mundial). El principal importador mundial de aceite de palma es China, con 5,5
millones de toneladas/ao; la Unin Europea importa 4,1 millones de toneladas/ao, la
India 4,1 millones de toneladas/ao y Pakistn, 2,0 millones de toneladas/ao (United States
Department of Agriculture (FAS), 2010).

Figura 3.22. Planta joven de palma aceitera.

En la actualidad, las compaas productoras de aceite lo suministran a un amplio


abanico de clientes: refinadoras, minoristas, industrias agroalimentarias, y plantas industriales
de agrocombustibles.
Los racimos de fruto (Figura 3.23) se presentan entre el tronco y la hoja y en forma
acorazonada con una dimensin promedio de 50 centmetros de largo por 35 de ancho. El
fruto es una drupa ssil cuya forma vara desde casi esfrica a ovoide o alargada, con
tamaos de 2 a 5 cm de dimetro.

104

Introduccin

Figura 3.23. Racimo de la palma aceitera.

3.3.2. USOS DEL ACEITE DE PALMA


El aceite de palma tiene mltiples usos, siendo el principal como aceite comestible
para la elaboracin de frituras, aunque tambin se ha popularizado como materia prima
para margarinas y productos de confitera. Igualmente se utiliza frecuentemente en la
fabricacin de jabones, cosmticos de alta calidad y detergentes. Otro uso importante es en
la elaboracin de grasas lubricantes y secadores metlicos, destinados a la produccin de
pinturas, barnices y tintas. Adems en la medicina se ha comprobado su alta eficiencia en el
control y curacin de enfermedades relacionadas con los lpidos (colesterol), el cncer y la
trombosis arterial. En la industria siderrgica se utiliza para la produccin, en fro, de
laminados de acero y en la industria de pigmentos se emplea como materia prima, debido a
su alto contenido de compuestos carotenoides.
El aceite de palma tiene una aplicacin importante en la produccin de
agrocombustibles de primera generacin. Estos combustibles son fabricados en su mayora a
partir de productos alimenticios como soja, colza, maz, girasol o caa de azcar. La
demanda de agrocombustibles se centra principalmente en EE.UU. y la Unin Europea
(UE); en la ltima se tiene el objetivo de incorporar, antes de 2020, un 10 % de agrocombustibles al total de combustibles utilizados en el transporte. En Espaa, el sector del
transporte supone un 39 % del consumo energtico total, uno de los consumos ms altos en

105

Introduccin
la UE (IDAE, 2006), lo que hace pensar que Espaa va a jugar un papel importante en la
importacin de materias primas para agrocombustibles.
Por estas razones, est previsto que la demanda de aceite de palma se duplique hasta
2030 y se triplique hasta 2050. Ello es debido a que la palma aceitera es la planta oleaginosa
ms productiva del planeta, lo que la convierte en la ms interesante para la produccin de
agrodisel. Su rendimiento es 10 veces superior al de la soja (CIRAD, 2007). Segn las
especies y las plantaciones, el rendimiento medio de aceite de palma crudo es 3,5
toneladas/hectrea/ao, y para la fabricacin de agrodisel, 1 tonelada de aceite es
equivalente a 0,9 toneladas de diesel (Pioch y Vaitilingom, 2005).

3.3.3. RESIDUO DE LA INDUSTRIA DEL ACEITE DE PALMA


El residuo de la industria del aceite de palma, ms conocido como EFB (Empty Fruit
Bunches, racimos sin fruto) est constituido por el material lignocelulsico que acompaa al
fruto de la palma de aceite.
Las plantaciones de palma comienzan a producir frutos a los 4-5 aos de ser
plantadas, alcanzando su mayor produccin entre los 20 y 30 aos, despus declinan y dejan
de ser rentables, especialmente por la altura a la que se encuentran los frutos. Los racimos
pesan entre 15-25 Kg, conteniendo de unos 1000 a 4000 frutos. Cada hectrea de palma
aceitera produce 10 toneladas anuales de frutos, de los cuales se extraen 3000 Kg de aceite
de palma, su principal aplicacin.
En cuanto a la utilizacin del residuo de la industria del aceite de palma (EFB) en la
produccin de pastas y papel, existe una bibliografa bastante amplia.
El primer trabajo que se ha encontrado sobre el pasteado de EFB data de 1977. En l,
Muthurajah y Peh utilizan el proceso kraft con una concentracin de lcali activo del 16 %,
durante 3 horas a 160 C, obteniendo una pasta con un rendimiento del 56 % y un nmero
kappa de 16,9. Las propiedades qumicas de las fibras de EFB son similares a las de
maderas frondosas, excepto en el contenido de pentosanos que es ms elevado. Las
propiedades fisicomecnicas de las hojas de papel son menores que las correspondientes al
papel de pasta al sulfito neutro de una especie de maderera frondosa.
El proceso kraft tambin ha sido estudiado por otros autores. Khoo y Lee (1991) lo
aplican a EFB, cscaras del fruto y hojas de la palma. La pasta de EFB obtenida tiene un

106

Introduccin
rendimiento bueno (56 %) y proporciona hojas de papel con moderada resistencia y alto
ndice de alargamiento (8 %).
Ibrahim (2002) compara la composicin de pastas de EFB obtenidas mediante los
procesos: kraft, kraft-antraquinona, a la sosa y a la sosa-antraquinona. La pasta a la
sosa es la que tiene los mayores contenidos de lignina y holocelulosa; la pasta a la sosaantraquinona la que proporciona mayores valores para el contenido de -celulosa y para la
viscosidad; y la pasta kraft-antraquinona es la de mayor contenido de glucosa y los ms bajos
contenidos de arabinosa, manosa y galactosa.
Tambin el proceso kraft-antraquinona fue estudiado introduciendo un pretratamiento con sosa a baja concentracin, producindose una disminucin del contenido
residual de lignina en la pasta. Se concluye tambin que pequeas cantidades de
antraquinona aadidas mejoran significativamente el rendimiento de la pasta, pero ste
disminuye al incrementar la concentracin de antraquinona. La pasta conseguida es fcil de
refinar y tiene una blancura superior al 40 % (Akamatsu y col., 1987a).
El proceso a la sosa ha sido aplicado tambin a EFB por Law y Jiang (2001),
obteniendo una pasta que se compara a la de lamo tembln obtenida bajo idnticas condiciones. Las fibras de EFB tienen ms espesor de pared celular y mayor rigidez que las de
lamo tembln. Las fibras de EFB tienen unas elevadas solubilidades en agua caliente y en
sosa al 1 % y un elevado valor del contenido de cenizas. Las pastas de EFB se blanquean
con perxido de hidrgeno mejor que las de lamo tembln, y dan lugar a hojas con menor
ndice de traccin, pero mayor alargamiento e ndice de desgarro. La pasta de EFB puede
ser mezclada con la de maderas conferas para producir cartn y varios grados de papel de
impresin y escritura.
Otro trabajo (Daud y col., 2004b) estudia, mediante la utilizacin de un diseo
experimental de composicin central, la influencia de las condiciones de operacin
(temperatura, tiempo y concentracin de lcali) sobre las propiedades de la pasta de EFB
(rendimiento, nmero kappa e ndices de traccin y desgarro). Se consiguen rendimientos
de 30-45 %. Se considera que los valores de 160 C, 60-120 minutos y 20-30 % de lcali son
suficientes para el adecuado pasteado.
Daud y col. (1998) llevan a cabo el pasteado de EFB con sosa, carbonato sdico y
sulfito sdico, siendo el primer proceso el ms eficiente, con un rendimiento del 50-60 % y

107

Introduccin
buenas propiedades de resistencia. Mezclada la pasta obtenida con pasta termomecnica de
conferas (1:1) producen papel para peridicos con aceptables propiedades.
El proceso a la sosa-antraquinona tambin ha sido aplicado a EFB para conseguir
pasta semiqumica para sustituir a la pasta de madera en la fabricacin de corrugado-medio
para cartn (Roliadi y Pasaribu, 2004). Se utilizan concentraciones de sosa de 4-8 %,
concentraciones de antraquinona de hasta 0,15 %, relacin lquido/slido de 5:5, temperatura de 135 C y tiempo de 3 horas.
La obtencin de pasta termomecnica de EFB ha sido estudiada por Akamatsu y col.
(1987b), utilizando un pretratamiento con sosa al 3 % a 50 C, requiriendo poca energa en
el refinador, en comparacin con el bagazo y otras fibras no madereras tropicales. Sus hojas
de papel poseen comparativamente altos valores para las propiedades de resistencia, a causa
de la gran cantidad de fibras largas que posee la pasta.
Kamishima y col. (1988) comparan la pasta termomecnica de las hojas de palma de
aceite con la de EFB, observando que la primera tiene menos blancura.
Tambin se ha estudiado la obtencin de pasta mecnica de EFB con un tratamiento
de perxido alcalino, con varios tratamientos de acondicionamiento previo qumico y
mecnico (Daud y col., 2004a; Ghazali y col., 2006).
Tambin Khoo y Lee (1991) obtienen pasta termomecnica de EFB con un rendimiento bajo (76 %), pero con mejor resistencia al desgarro y ligeramente menor ndice de
traccin que la correspondiente de maderas conferas.
Mediante un pretratamiento de hidrlisis se obtiene pasta para disolver utilizando sosa
(Daud y col., 2005). Se encontraron los siguientes valores: rendimiento 31,2 %, nmero
kappa 6,0, -celulosa 96,9 %, viscosidad 16,1 cp y cenizas 0,15 % para condiciones de
operacin ptimas: 161 C, 100 minutos y 26,1 % de lcali.
Procesos organosolv para la obtencin de pasta de EFB tambin han sido estudiados.
Suleman y Yusoff (1997) utilizan etanol con EFB, cscaras del fruto y hojas de palma de
aceite, con cido clorhdrico como catalizador, que afecta considerablemente al
rendimiento, cantidad de astillas, nmero kappa y propiedades de las hojas de papel
formadas. Se obtiene un satisfactorio rendimiento y calidad de la pasta cuando se opera
durante 120 minutos, con una relacin lquido/slido de 12, 180 C y un 60 % de etanol,
con o sin catalizador.

108

Introduccin
Empleando mezclas de etanol-agua (1:1), a 165-180 C durante 2 horas con una
relacin lquido/slido de 10, y con o sin un 10 % de cido clorhdrico 1 N y de sosa 1,15 M
como catalizadores, se estudia la produccin de pasta de EFB (Aziz y col., 2002). Se
consigue un mximo rendimiento (57 %) cuando se opera a 170 C durante 2 horas sin
adicin de catalizadores, mientras que el rendimiento es mnimo (45 %) cuando se opera a
175 C durante 2 horas con catalizador cido.
Finalmente, tambin se ha aplicado el proceso organosolv, libre de azufre, IDE
modificado: HIDE (Quader y Lnnberg, 2004). El proceso IDE utiliza tres etapas:
impregnacin de materia prima con carbonato sdico concentrado, coccin con una
disolucin de etanol, y extraccin con una mezcla de etanol-agua. El proceso modificado
HIDE consiste en comenzar con un ligero medio cido con el fin de conseguir un alto
rendimiento de celulosa para propsitos qumicos.
Tambin se ha investigado el biopasteado aplicado a EFB, utilizando un hongo blanco
(K14), que hace disminuir ms rpidamente el contenido de lignina residual que los hongos

Phanerochaete chrysosporium y Coriolus versicolor. Debido al tratamiento con K14 el


ndice de desgarro de las hojas de papel formadas se incrementa un 16 % y el nmero kappa
decrece un 4,2 %, comparado con un pasteado sin tratamiento con hongo (Goenadi y col.,
1998).
Para terminar se citan dos trabajos que blanquean pastas kraft de EFB mediante
mtodos TCF (totalmente libres de cloro): el primero (Tanaka y col., 2004) blanquea con la
secuencia OZP, con la intercalacin entre las etapas O y Z de una etapa con cido sulfrico.
El segundo (Leh y col., 2004) blanquea con la secuencia OZP pastas de EFB a la sosa, a
la sosa-antraquinona, kraft y kraft-antraquinona, considerando el efecto del blanqueo sobre
la viscosidad y el nmero kappa de las pastas blanqueadas. Las pastas a la sosaantraquinona y kraft-antraquinona blanqueadas alcanzan el 80 % de blancura, manteniendo
la viscosidad por encima de 15 cp.

109

Introduccin

110

Material y Mtodos

Material y Mtodos
En este captulo se describen brevemente y de manera concisa los equipos y
procedimientos utilizados en la caracterizacin de los materiales (materia prima y
productos), el tratamiento hidrotrmico, el pasteado, el blanqueo de las pastas y la
produccin de celulosa nanofibrilar, as como el diseo de experimentos utilizado y los
modelos que relacionan las variables dependientes con las de operacin en los diferentes
procesos considerados.

4.1. CARACTERIZACIN DE LOS MATERIALES

En lo que sigue se mencionan los mtodos para la caracterizacin de las materias


primas, fracciones slida y lquida procedentes del tratamiento hidrotrmico, pastas y hojas
de papel, lejas o licores negros de coccin procedentes del pasteado y suspensiones acuosas
y pelculas de celulosa nanofibrilar.

4.1.1. CARACTERIZACIN FSICO-QUMICA DE LA MATERIA PRIMA


Los mtodos de preparacin o acondicionamiento de las muestras y de anlisis de las
mismas han sido los siguientes:

Cenizas: Norma TAPPI T 211.

Solubilidad en agua caliente: Norma TAPPI T 257.

Solubilidad en sosa al 1 %: Norma TAPPI T 212.

Extrables con etanol o etanol-benceno: Norma TAPPI T 204.

Lignina: Norma TAPPI T 222.

Holocelulosa: Mtodo de Wise y col. (1946) y Norma TAPPI T 9.

-celulosa: Norma TAPPI T 203.

Hemicelulosa: mediante la diferencia entre los contenidos de holocelulosa y celulosa.

Tamao de las fibras: mediante el uso de un microscopio de proyeccin.

113

Material y Mtodos
4.1.2. CARACTERIZACIN QUMICA DE LA FRACCIN SLIDA PROCEDENTE
DEL TRATAMIENTO HIDROTRMICO
Para determinar el rendimiento en slidos de la fraccin slida procedente del
tratamiento hidrotrmico, es necesario filtrar y lavar dicha fraccin slida una vez ha tenido
lugar el tratamiento hidrotrmico correspondiente. Tras el filtrado y lavado de la misma, se
determina el rendimiento en slidos mediante gravimetra.

4.1.3. CARACTERIZACIN QUMICA DE LA FRACCIN LQUIDA PROCEDENTE DEL TRATAMIENTO HIDROTRMICO


Para determinar los contenidos de glucosa, xilosa, arabinosa y cido actico en la
fraccin lquida procedente del tratamiento hidrotrmico, se toma una cierta cantidad de
sta y se hidroliza con cido sulfrico al 4 % en peso, durante 20 minutos a 2 atmsferas de
presin (121 C). Las muestras se filtran con filtros de 0,45 m y se analizan mediante
HPLC (Agilent 1200 series, con columnas Biorad Aminex HPX-87H; elucin con 0,01 M
H2SO4 y con caudal de 0,6 mL/min; detector de ndice refractivo) para determinar el

incremento en la concentracin de monosacridos y cido actico, causado por la


posthidrlisis, que est relacionado con el contenido de oligmeros y sus grados de
sustitucin por los grupos acetilo.

4.1.4. CARACTERIZACIN FSICO-QUMICA DE LA PASTA


Los mtodos de caracterizacin fsico-qumica de las pastas han sido los siguientes:

Rendimiento: mediante gravimetra.

Nmero kappa: Normas UNE 57-034, ISO 302 y TAPPI T 236.

Blancura: Norma ISO 3688.

Holocelulosa; -celulosa; lignina; extrables con etanol; cenizas; hemicelulosa:


Mtodo de Wise y col. (1946) y Norma TAPPI T 9; Norma TAPPI T 203;
Norma TAPPI T 222; Norma TAPPI T 204; Norma TAPPI T 211; mediante
la diferencia entre los contenidos de holocelulosa y -celulosa.

Viscosidad: Normas UNE 57-039-92, ISO 5351-1, TAPPI T 230 y TAPPI T


254.

114

Material y Mtodos

Grado de polimerizacin: Mtodo de Gruber y Gruber (1981).

Grado de desgote o refinado: Normas TAPPI T 248 e ISO 5267-1:1999.

cidos hexenurnicos: Mtodo de Gellerstedt y Li (1996).

4.1.5. CARACTERIZACIN DE LAS HOJAS DE PAPEL


Se consideran los siguientes mtodos para la caracterizacin de las hojas de papel:

Preparacin y acondicionamiento de las hojas de papel: Normas UNE 57-042,


TAPPI T 272 e ISO 5269-2.

Espesor: Norma TAPPI T 411.

Gramaje: Norma TAPPI T 410.

Permeabilidad en aire: Norma ISO 5636-3.

ndice de traccin, longitud de ruptura y alargamiento: Normas UNE 57-054,


TAPPI T 494 e ISO 5270.

ndice de estallido: Normas UNE 57-058 y TAPPI T 403.

ndice de desgarro: Normas UNE 57-033, TAPPI T 414 e ISO 5270.

Blancura: Normas UNE 57-062 y TAPPI T 452.

Reflectancia azul difusa (para medida de Blancura): Norma TAPPI T 525.

Coordenadas CIE L*a*b* y CIE L*C*: Norma TAPPI T 527.

ngulo de contacto: Norma TAPPI T 558.

Envejecimiento acelerado: Norma UNE 57-092-4.

4.1.6. CARACTERIZACIN QUMICA DE LAS LEJAS O LICORES NEGROS DE


COCCIN PROCEDENTES DEL PASTEADO
El licor negro procedente del pasteado se diluye y agita con agua destilada hasta
conseguir una relacin volumtrica de 7:1, con el fin de precipitar la lignina y obtener
muestras vlidas para su anlisis cromatogrfico.

115

Material y Mtodos
Los slidos precipitados se lavan de manera repetida con agua destilada hasta llegar a
la neutralidad, de manera que se puedan extraer la mayor cantidad posible de carbohidratos.
La determinacin cuantitativa de monosacridos se lleva a cabo con cromatografa lquida.
Las muestras se diluyen con agua destilada hasta una concentracin determinable de
azcares, se neutralizan con carbonato brico y se filtran a travs de membranas de 0,20
0,45 m (segn proceda).
La determinacin individual de monosacridos y furfural se lleva a cabo usando un
HPLC Series 1500 Waters equipado con columnas Biorad Aminex HPX-87P y
termostatizadas a 85 C. El detector de ndice refractivo se encuentra a una temperatura de
30 C. El furfural se determina mediante un detector de UV-VIS a una de 254 nm. Como
fase mvil se emplea agua y el flujo es de 0,6 mL/min.

4.1.7. CARACTERIZACIN DE LAS SUSPENSIONES Y PELCULAS DE CELULOSA NANOFIBRILAR (NFC)


Se consideran los siguientes procedimientos:

116

Nmero kappa: Norma SCAN-C 1:00.

Viscosidad: Norma ISO 5351.

Grado de polimerizacin: Mtodo de Gruber y Gruber (1981).

Espesor: Norma TAPPI T 411.

Gramaje: Norma TAPPI T 410.

Permeabilidad en aire: Norma ISO 5636-3.

Propiedades mecnicas: Norma EN ISO 291:2008.

ngulo de contacto: Norma TAPPI T 558.

Material y Mtodos
4.2. EQUIPOS UTILIZADOS EN LA EXPERIMENTACIN

Se describen brevemente diversos equipos utilizados en la caracterizacin de las


materias primas, tratamiento hidrotrmico, pasteado, blanqueo de las pastas, formacin de
hojas de papel y produccin de celulosa nanofibrilar.

4.2.1. MICROSCOPIO PARA MEDIDA DEL TAMAO DE LAS FIBRAS DE LA


MATERIA PRIMA
Se utiliza un microscopio con visor lateral, que refleja las imgenes de las fibras sobre
un papel blanco (Figura 4.1).

Figura 4.1. Microscopio para medida de longitud de fibra de la materia prima.

Tras conseguir la individualizacin de las fibras mediante el empleo de una disolucin


de Schulzt (200 mL de agua, 80 mL de cido ntrico concentrado y 4 g de clorato potsico),
se ponen unas gotas de la suspensin obtenida en un portaobjetos y se someten a un barrido
en el microscopio para observar y medir todas las fibras presentes.

4.2.2. MOLINO PARA EL DESMENUZADO DE LA MATERIA PRIMA


Se utiliza, en la mayora de los casos, un molino-cortador convencional (Figura 4.2)
que consigue tamaos de haces de fibras de unos 1 a 5 cm de longitud, rechazando una
fraccin de finos (menores que 1 cm) y gruesos (mayores que 5 cm).

117

Material y Mtodos

Figura 4.2. Molino de la materia prima.

4.2.3. DIGESTOR PARA TRATAMIENTOS HIDROTRMICOS Y PASTEADOS


Consiste en un reactor cilndrico discontinuo de 15 L de capacidad, recubierto de una
camisa en cuyo interior hay un conjunto de resistencias elctricas para la calefaccin. Est
diseado para soportar hasta 15 Kg/cm2 de presin. El conjunto est calorifugado adecuadamente.

El equipo anterior est unido, mediante un eje rotativo, a la unidad de control, donde
existe un motor, que controla la agitacin del reactor, y los dispositivos de medida y control
de la temperatura de calefaccin. Completa el conjunto una vlvula de seguridad, acoplada
al reactor, y una vlvula convencional de expansin (Figura 4.3).

Figura 4.3. Digestor para tratamientos hidrotrmicos y pasteados.

118

Material y Mtodos
Los procesos de coccin o pasteado de la materia prima que se realizan a presin
atmosfrica se llevan a cabo en matraces de fondo redondo preferentemente de 250 mL. En
estos procesos es necesario que exista reflujo total y un buen sistema de agitacin.

4.2.4. DESFIBRADOR DE MATERIALES COCIDOS


Consiste en un recipiente cilndrico de 30 L de capacidad provisto de una hlice
batidora que es movida, por la parte inferior, gracias a un motor que gira a 1200 rpm, y un
tubo de vaciado al que se le acopla una vlvula de drenaje (Figura 4.4).
La pared interior del recipiente est provista de cortacorrientes rectangulares
especiales. Debido a la forma de la hlice y de las cortacorrientes, la desintegracin de los
materiales celulsicos es muy eficaz, no ocasionando dao a las fibras de celulosa ni
realizando un refinado de la pasta.

Figura 4.4. Desfibrador de materiales cocidos.

4.2.5. SEPARADOR DE FIBRAS E INCOCIDOS


Se utiliza un equipo constituido por un recipiente cilndrico de 40 cm de dimetro y
20 L de capacidad, cuya base inferior va provista de un filtro de placa metlica circular con
ranuras longitudinales rectas de 0,4-1,0 mm de espesor, distribuidas paralelamente en dos
secciones semicirculares. La parte inferior del equipo lo forma un sistema de tubos que
puede conectarse a vaco o simplemente al desage (Figura 4.5).

119

Material y Mtodos

Figura 4.5. Separador de fibras e incocidos.

El material cocido y desfibrado se vierte en la parte superior del cilindro, aadiendo


abundante agua y manteniendo una agitacin adecuada para evitar el taponamiento del filtro.
Durante la operacin, las fibras pasan al filtrado mientras que los incocidos y otras
impurezas se retienen sobre el filtro.

4.2.6. SEPARADOR DE FIBRAS Y AGUA


Las fibras procedentes del desfibrador de materiales cocidos (Figura 4.4) y del
separador de fibras e incocidos (Figura 4.5), portando gran cantidad de agua, se vierten en el
separador de fibras y agua (Figura 4.6). ste consta de un tamiz fino (luz de malla 0,14
0,16 mm, segn proceda), sobre el que quedan las fibras, mientras que el agua pasa al
desage. Tambin se utiliza este separador de fibras para lavar las pastas que se obtienen tras
los tratamientos hidrotrmicos y de pasteado de las materias primas.

Figura 4.6. Separador de fibras y agua.


120

Material y Mtodos
4.2.7. REFINADOR-DESINTEGRADOR DE FIBRAS
Para la desintegracin mecnica o disgregacin de las fibras se utiliza un refinador de
discos de laboratorio Sprout-Bauer (Figura 4.7). Esta unidad puede utilizarse adems para
refinado de pastas, la preparacin de la materia prima y el desfibrado de materiales cocidos.
El sistema de discos refinadores es accionado mediante un motor de 25 CV. La alimentacin puede realizarse manualmente, usando una manivela que acta sobre un tornillo sin
fin, o mediante un motor elctrico.

Figura 4.7. Sprout-Bauer.

4.2.8. REFINADOR
El refinado de las pastas crudas se llev a cabo preferentemente en un molino refinador PFI de la firma Metrotec, dotado de sistemas adecuados para la medida del nmero
de vueltas, a que se somete el material a refinar, y del consumo de energa necesaria para el
refinado (Figura 4.8).

Figura 4.8. Refinador PFI.


121

Material y Mtodos
4.2.9. CENTRFUGA PARA ESCURRIDO DE PASTAS
Para disminuir la humedad de las pastas se utiliza una centrfuga Willians (Figura 4.9).
Consta de una cesta cilndrica en cuyas paredes laterales se alberga una tela metlica que
retendr la pasta, dejando pasar el lquido clarificado. La cesta est unida a un motor de
3450 rpm mediante un acoplamiento cnico. Completa el sistema un temporizador
automtico, que controla la duracin del proceso, y un freno para disminuir el tiempo de
parada.

Figura 4.9. Centrfuga.

4.2.10. DESINTEGRADOR DE PASTAS CELULSICAS


La disgregacin de las pastas secas para la preparacin de pastas de papel
previamente a los ensayos de caracterizacin o a la formacin de hojas de papel, se lleva a
cabo en un mezclador, donde se pone en suspensin la pasta mediante la adicin de agua
(Figura 4.10).

Figura 4.10. Desintegrador.


122

Material y Mtodos
Consta de un vaso y un eje rotor, de dimensiones estandarizadas segn la Norma
UNE-57-026, y un motor que hace girar el rotor correctamente. Con este diseo se intenta
no alterar el estado dimensional de las fibras celulsicas.

4.2.11. REFINMETRO SHOPPER-RIEGLER


La medida del grado de refinado o de desgote de las pastas se lleva a cabo en un
refinmetro Shopper-Riegler, siguiendo la Norma TAPPI T 248. Consta de una cmara de
desgote provista de una tela metlica, cono de cierre y un embudo colocado en un soporte
adecuado (Figura 4.11). El fundamento del mtodo consiste en que un volumen conocido
de suspensin de pasta en agua, desgota, a travs de la torta de pasta formada durante el
ensayo sobre la tela metlica, en el embudo. El filtrado es recogido por una salida lateral en
una probeta especial, graduada en unidades SR. Para una descarga de 1 L le corresponde 0
SR, mientras que cuando la descarga sea de 0 L, le corresponde 100 SR.

Figura 4.11. Refinmetro Shopper-Riegler.

4.2.12. FORMADOR DE HOJAS


Est compuesto por una tolva superior de metacrilato transparente, abatible para la
extraccin de hojas, y una tela metlica montada sobre un disco taladrado, que hace de
soporte de las hojas (Figura 4.12). Completa la instalacin un sistema de aspiracin

123

Material y Mtodos
formando sifn y un depsito para el desage. Permite la elaboracin de hojas de 215 mm
de dimetro.

Figura 4.12. Formador de hojas.

4.2.13. PRENSA PARA HOJAS


Se utiliza una prensa modelo PR.03-402 (Figura 4.13), con el fin de reducir la
humedad de las hojas de ensayo de papel fabricadas en el formador de hojas.

Figura 4.13. Prensa para hojas y aros secadores.

4.2.14. SECADOR DE HOJAS


Consta de un juego de placas y aros secadores, donde se intercalan las hojas de papel
que han de secarse, permaneciendo all el tiempo suficiente para reducir la humedad
aproximadamente al 10 % (Figura 4.13).

124

Material y Mtodos
4.2.15. FLUIDIZADOR A ALTA PRESIN
La obtencin de suspensiones acuosas de celulosa nanofibrilar (NFC) se realiza a
travs del proceso conocido como microfluidizacin, y para ello se utiliza un fluidizador a
alta presin (microfluidizador) (Figura 4.14).

Figura 4.14. Microfluidizador.

4.2.16. FORMADOR DE PELCULAS DE CELULOSA NANOFIBRILAR (NFC)


La formacin de pelculas de celulosa nanofibrilar se lleva a cabo en un equipo de
sobrepresin (Figura 4.15).

Figura 4.15. Equipo de sobrepresin para la formacin de pelculas de NFC.

125

Material y Mtodos
4.2.17. MICROSCOPIO DE FUERZA ATMICA (AFM)
El estudio de la morfologa, rugosidad, espesor y distribucin material de las
dispersiones y pelculas de celulosa nanofibrilar se realiza en un microscopio de fuerza
atmica (Figura 4.16).

Figura 4.16. Microscopio de fuerza atmica.

4.2.18. INSTRUMENTACIN PARA LA MEDIDA DE PROPIEDADES FSICAS DE


LAS HOJAS DE PAPEL Y PELCULAS DE CELULOSA NANOFIBRILAR (NFC).
En la Figura 4.17 se muestran los instrumentos necesarios para la medida de las
distintas propiedades fsicas, tanto de las hojas de papel como de las pelculas de celulosa
nanofibrilar (NFC).

Figura 4.17. (De izquierda a derecha). Mquina universal para medida del ndice de traccin,
longitud de ruptura y alargamiento; equipo para medida del ndice de estallido; equipo para medida
del ndice de desgarro; equipo para medida de la blancura.

126

Material y Mtodos
4.3. PROCEDIMIENTOS

4.3.1. TRATAMIENTO HIDROTRMICO DE LA MATERIA PRIMA


Se introduce la materia prima en el reactor de la Figura 4.3 y se le aade la cantidad
de agua necesaria para conseguir la adecuada relacin lquido/slido. Despus se procede al
calentamiento de la mezcla hasta una temperatura dada, continuando el calentamiento
durante un tiempo determinado. A continuacin se descomprime el reactor y se separan las
fracciones lquida y slida, para su posterior tratamiento o anlisis.

4.3.2. PROCESO DE PASTEADO


En el reactor de la Figura 4.3 (proceso de pasteado a presin superior a la atmosfrica)
se introduce la materia prima molida y tamizada (EFB o fraccin slida procedente del
tratamiento hidrotrmico del EFB), aadiendo a continuacin los reactivos de coccin y el
agua necesarios para conseguir el hidromdulo o relacin lquido/slido deseado. Despus
de la coccin, a la temperatura y tiempo fijados, los materiales cocidos se filtran y lavan con
agua para eliminar cualquier resto de leja impregnada en ellos (Figura 4.6) y, aadiendo una
cierta cantidad de agua, se pasan por el desfibrador de materiales cocidos o hmedo (1200
rpm, 30 min) (Figura 4.4). Posteriormente, las pastas se refinan-desfibran en el refinador
Sprout-Bauer (Figura 4.7). Las pastas as obtenidas pueden utilizarse directamente para la
formacin de hojas o ser blanqueadas, o pueden secarse para posteriores ensayos. A
continuacin, se hacen pasar las pastas a travs del separador de fibras e incocidos (Figura
4.5). Para el secado de las pastas, se someten a un escurrido en una centrfuga (Figura 4.9) y
se secan a temperatura ambiente hasta una humedad aproximada del 10 %.
Si el proceso de pasteado se lleva a cabo a presin atmosfrica, se utilizan matraces de
fondo redondo acoplados a sistemas de reflujo. En ellos se introduce la materia prima
molida y tamizada, junto con los reactivos de coccin y el agua necesarios para conseguir el
hidromdulo o relacin lquido/slido deseado. Se calienta la mezcla hasta alcanzar la
temperatura deseada, bien podra ser la temperatura de ebullicin, y es en ese momento
cuando se aade el catalizador del proceso. La mezcla se mantiene a reflujo total con
agitacin durante el tiempo que sea necesario. Despus de la coccin, las pastas se separan
del lquido de coccin por filtracin y los slidos se lavan con disoluciones concentradas del
reactivo apropiado para evitar la deposicin de la lignina disuelta en la pasta. Finalmente, las
127

Material y Mtodos
pastas se lavan con agua destilada hasta llegar a la neutralidad y se dejan secar a temperatura
ambiente.

4.3.3. PROCESO DE REFINADO DE LAS PASTAS


Para el refinado de las pastas en el molino PFI (Figura 4.8) se procede como sigue: se
toman 30 g de pasta, midiendo su humedad, y se les aaden 2 L de agua, desintegrando a
continuacin a 10000 rpm (Figura 4.10). Despus se filtra para eliminar la mayor cantidad
de agua posible y la pasta escurrida se lleva a un vaso tarado, donde se pesa y se le aade
agua hasta alcanzar un peso de 300 g. A continuacin, se adhiere la pasta a las paredes del
vaso del PFI, girando el crter para homogeneizar la pasta y facilitar su adherencia.
Posteriormente se descienden las palas y se encaja la tapa del vaso. Se programa el nmero
de vueltas y se pone en marcha el equipo. Cuando finaliza la operacin se retira la pasta y se
lava con 0,5 L de agua para liberar la pasta retenida en las partes del PFI.

4.3.4. PROCESO DE BLANQUEO DE LAS PASTAS CON PERXIDO DE


HIDRGENO
Para el blanqueo de las pastas con perxido de hidrgeno, se introducen stas junto
con los reactivos (perxido de hidrgeno, oxgeno, silicato de sodio y sulfato de magnesio)
en un reactor de 5 L de capacidad y, tras ajustar a una consistencia del 5 %, se opera
manteniendo la temperatura a 105 C durante un tiempo determinado.
Una vez terminado el proceso se separa la leja de la pasta mediante filtracin y se lava
esta ltima para posteriormente secarla y homogeneizarla, antes de proceder a su
caracterizacin o a la formacin de hojas de papel.

4.3.5. FORMACIN DE SUSPENSIONES DE CELULOSA NANOFIBRILAR (NFC)


La celulosa nanofibrilar (NFC) se obtiene a travs del proceso de microfluidizacin.
Para llevar a cabo este proceso se necesitan suspensiones de pasta en agua destilada con una
consistencia del 1,8 %. Las muestras se hacen pasar por el equipo microfuidizador (Figura
4.14), alrededor de 5 veces y son sometidas a un bombeo para incrementar la presin hasta
unos 55 MPa. Este equipo opera a presin y flujo constantes y cuenta con una cmara de
interaccin que permite desfibrilar las fibras presentes en la pasta a travs de fuerzas de
128

Material y Mtodos
colisin e impactos contra las paredes del mismo. A travs de este proceso, las fibras se
rompen generando estructuras de tamao nanomtrico. La temperatura no es controlada en
este proceso, pero temporalmente debe detenerse el proceso, especialmente cuando la
temperatura alcanza un valor de 90 C, puesto que en estos casos, puede obturarse la
bomba. La temperatura idnea de trabajo es de 45 C. Las suspensiones de NFC que
resultan de este proceso deben almacenarse en una cmara acondicionada a 4 C.

4.3.6. FORMACIN DE HOJAS DE PAPEL Y PELCULAS DE CELULOSA


NANOFIBRILAR (NFC)
Para la formacin de las hojas de papel, las pastas se tratan en el desintegrador (Figura
4.10) durante 30 minutos, aadindoles el agua necesaria para alcanzar una consistencia del
1,5 %, segn la Norma UNE 57-026.
Posteriormente, la pasta desintegrada se lleva al formador de hojas (Figura 4.12),
obtenindose una hoja circular hmeda, que se despega de la malla del formador mediante
papel de filtro. A continuacin la hoja se prensa (Figura 4.13) para eliminar humedad y
finalmente se lleva al aro secador (Figura 4.13) hasta alcanzar la humedad de equilibrio
correspondiente a la humedad relativa del ambiente.
Para la formacin de pelculas de NFC es necesario diluir las suspensiones de NFC al
1 % de consistencia en agua destilada y agitar la mezcla durante 12 horas a temperatura
ambiente. Despus se introduce una alcuota de la mezcla en el equipo de sobrepresin
(Figura 4.15), y se somete a una presin de 2,5 bar y un tiempo de drenaje de 30 minutos.
Para conseguir la eliminacin del exceso de agua presente en las suspensiones de NFC y no
eliminar con ella las fibrillas, se utiliza un filtro a modo de pantalla. Para obtener pelculas de
NFC totalmente planas y lisas, se emplea una prensa fra, en la que las pelculas de NFC se
someten a una presin de 4,1 bar durante 4 minutos. A continuacin, las pelculas de NFC
hmedas se secan, utilizando para ello dos lminas de papel de filtro, y se someten a una
temperatura de 100 C y una presin de 0,18 bar durante 2 horas en una prensa. Finalmente, las pelculas de NFC se almacenan en una cmara acondicionada (23 C y 50 % de
humedad relativa del ambiente).

129

Material y Mtodos
4.4. DISEO DE EXPERIMENTOS Y MODELOS

4.4.1. DISEO FACTORIAL DE EXPERIMENTOS


Se utiliza un diseo factorial de experimentos, de composicin central, para estudiar la
influencia de las variables de operacin sobre las variables dependientes, propias de las
fracciones resultantes de los tratamientos hidrotrmicos y de las pastas obtenidas en los
procesos de pasteado y blanqueo.
El modelo ensayado experimenta con una serie de puntos (experimentos) alrededor
de un punto central (experimento central). Este diseo satisface los requerimientos generales
de que todos los parmetros del modelo matemtico pueden ser estimados sin un nmero
excesivo de experimentos (Montgomery, 1991).
El diseo utilizado est definido por tres parmetros: nmero de variables (k);
constante (p), que toma los valores 0 (para k<5) y 1 (para k5); y nmero de puntos

centrales (nc). Estos parmetros originan tres grupos de puntos:

2k-p puntos que constituyen un diseo factorial


2k puntos axiales

nc puntos centrales

El nmero total de puntos (experimentos) vendr dado por la expresin:


n = 2k+2k+nc

[1]

donde k es el nmero de variables independientes.


4.4.2. MODELO POLINMICO
El ajuste de los datos experimentales de diversos ensayos hidrotrmicos, de pasteado o
de blanqueo, se realiza utilizando un modelo polinmico, con la idea de encontrar
relaciones entre las variables respuesta o dependientes con las variables de operacin o
independientes.
Los datos experimentales pueden ensayarse con el siguiente modelo polinmico de
segundo orden:

130

Material y Mtodos

donde:

Y = a0 + ni=1 bi Xni + ni=1 ci X2ni + ni=1;j=1 dij Xni Xnj


Xn = 2 X

X-X

mx-Xmn

(i<j)

[2]

[3]

siendo Y la variable respuesta o dependiente (caractersticas o propiedades de las fracciones


resultantes de los tratamientos hidrotrmicos y de las pastas obtenidas en los pasteados y

blanqueos), Xn las variables independientes (valores normalizados de las variables de

operacin de los procesos hidrotrmicos y de pasteado y blanqueo), y a0, bi, ci y dij son

constantes caractersticas desconocidas, que se estiman de los ajustes de los datos


experimentales.

Las variables independientes se normalizan de -1 a +1 de acuerdo con la ecuacin [3],


con el fin de facilitar la comparacin directa de los coeficientes y comprender mejor los
efectos de las variables independientes; esta normalizacin tambin mejora la estimacin de
los coeficientes de regresin, reduciendo las interrelaciones entre los trminos lineales y
cuadrticos.
Utilizando datos experimentales para cada una de las variables dependientes de los
procesos considerados, para los distintos experimentos del diseo experimental, y
procediendo con ellos a un anlisis de regresin mltiple, utilizando el programa BMDP
(Dixon, 1992), considerando todos los trminos de la ecuacin [2] y eliminando aquellos
trminos cuyos valores de F de Snedecor sean menores que 2-4, utilizando el mtodo de
stepwise (Draper y Smith, 1981), se encuentran ecuaciones que relacionan las distintas
variables dependientes con las independientes.

4.4.3. MODELO NEUROBORROSO


La modelizacin borrosa, basada en la pionera idea de Zadeh (1965), es una poderosa
herramienta para describir el comportamiento no lineal de sistemas complejos. Otra
poderosa herramienta para modelar estos sistemas complejos son las redes neuronales, cuya
caracterstica ms importante es su habilidad para alcanzar un mapa exacto del
comportamiento no lineal a travs de datos entrada-salida (variable independiente-variable
dependiente) sin tener un conocimiento exacto de las relaciones funcionales entre ellos
(Emmanouilides y Petrou, 1997). La conjuncin de sistemas borrosos y redes neuronales

131

Material y Mtodos
combina las ventajas de ambos sistemas y ofrecen una herramienta ms poderosa para
modelar. Se usan las redes neuronales como herramienta en sistemas borrosos.
La modelizacin de sistemas de dinmica no lineal cuyo ajuste es complicado requiere
adoptar configuraciones complejas. Sin embargo, dadas las caractersticas de los sistemas
objeto de estudio (hidrotrmicos, pasteados y blanqueos), se ha escogido un modelo
neuroborroso de configuracin simple para predecir el comportamiento de las propiedades
de las pastas obtenidas en funcin de las variables de operacin en los procesos
hidrotrmicos, de pasteado o blanqueo.
Un conjunto borroso es aqul en que la transicin entre la pertenencia y la no
pertenencia a un determinado atributo es gradual y no abrupta. Como ejemplo puede
tomarse la estatura de una persona: la lgica convencional establecera una altura lmite en la
que desde dicha altura hacia arriba el sujeto sera alto y en caso contrario bajo; sin embargo,
la lgica borrosa establece que existe una regin intermedia entre ser alto o bajo, dndole a
cada altura en esa regin intermedia una funcin de pertenencia (o grado de afinidad) con
respecto al atributo que se est considerando. Se podra ilustrar este ejemplo mediante la
Figura 4.18.
La modelizacin borrosa consiste en la aplicacin de un conjunto de teoras borrosas
para la representacin de las principales caractersticas de un sistema.
Los principios de estas teoras son:

Consideracin de variables lingsticas, que sustituyen a las variables


numricas.

Existencia de relaciones simples entre las variables, mediante declaraciones


condicionales borrosas o reglas borrosas.

Existencia de relaciones complejas entre las variables, mediante algoritmos


borrosos.

132

Material y Mtodos
Concepto clsico
Grado de
pertenencia
Alto

Bajo
Altura

1.80

Concepto borroso
Grado de
pertenencia

Alto

Bajo
1.50

2.20

Altura

Figura 4.18. Transicin entre la pertenencia y la no pertenencia a un determinado atributo, segn la


lgica convencional y la lgica neuroborrosa.

Una variable lingstica es aquella cuyos valores son frases o palabras (que toman
variables borrosas como sus valores). Por ejemplo, si se considera la altura como una
variable lingstica, podra haber diferentes trminos para describirla: alto, bajo, medio, etc.
(sustituyendo a un valor numrico de altura).
Las declaraciones condicionales borrosas son las relaciones entre las variables
lingsticas, y constituyen la produccin de reglas. La estructura de estas reglas es:
SIENTONCES En nuestro caso sera, por ejemplo: SI la temperatura (T) es baja, el
tiempo (t) bajo y la concentracin (C) baja, ENTONCES el sistema se comportara
linealmente respecto a estas variables (si se considera un modelo lineal).

133

Material y Mtodos
Se podran hacer combinaciones (reglas borrosas o localizadores) con los distintos
valores altos-bajos de las tres variables (T: temperatura, t: tiempo, C: concentracin de un
reactivo):
1) 1-1-1: T baja, t corto, C baja

5) 1-2-2: T baja, t largo, C alta

2) 1-1-2: T baja, t corto, C alta

6) 2-1-2: T alta, t corto, C alta

3) 1-2-1: T baja, t largo, C baja

7) 2-2-1: T alta, t largo, C baja

4) 2-1-1: T alta, t corto, C baja

8) 2-2-2: T alta, t largo, C alta

La funcin matemtica que representa un modelo borroso tiene la siguiente forma:


lm 1 =
y l [in 1 Fil ( xi , il )
=
y=
e
lm 1[=
in 1 Fil ( xi , il )
=

[4]

donde:

ye, es el valor estimado de la propiedad que se quiere modelar.


m, el nmero de reglas borrosas.

n, el nmero de variables independientes.


yl, el consecuente de una regla borrosa.

Fil(xi,il), la funcin de pertenencia de las variables de entrada dentro de su


universo de discurso.

Se podra considerar que el ncleo de un sistema borroso lo constituyen una base de


reglas borrosas R formada por un conjunto de reglas:
R = {R1, R2, R3,Rm}

[5]

donde cada una de estas reglas tiene la forma:

Ri: si X1 es F1l, X2 es F2l y Xn es Fnl entonces yl es Gl

donde Fil y Gl representan los trminos lingsticos de entrada y salida.

Como se coment anteriormente, la funcin de pertenencia es un valor indicativo del

grado de afinidad de un valor a cierto grupo (de un atributo). Dentro de una variable se
puede definir el nmero de grupos que se quiera, de forma que a mayor nmero de grupos
ms definido estar el sistema pero ms complicado de interpretar ser.

134

Material y Mtodos
Como grupo se entiende al conjunto de valores de una variable que pueden
catalogarse de igual forma.
De entre las diversas funciones de pertenencia para ser introducidas en el estimador
del modelo borroso, se puede considerar la lineal, con la frmula general:

Para valor bajo de la variable X1: F1 = 1-(1/(Xmx-Xmn))(X-Xmn)

Para valor alto de la variable X2: F2 = (1/(Xmx-Xmn))(X-Xmn)

donde X es el valor absoluto de la variable considerada, Xmx el valor mximo o alto de la

variable y Xmn el valor mnimo o bajo de la misma.

El consecuente de las reglas borrosas aporta la influencia de la variable de salida sobre

el sistema de inferencia borroso. El consecuente es un polinomio dependiente de las


variables de entrada Xi, y puede ser cualquier funcin no lineal, ms o menos simple. En

nuestro caso se considera el consecuente ms simple, el singletn, que constituye un


parmetro constante.
De esta forma, el modelo neuroborroso queda representado por:
Ye =

a1R1+a2R2+a7R7+a8R8
R1+R2+R7+R8

[6]

donde, si las variables independientes son la temperatura (T), el tiempo (t) y la


concentracin de reactivo (C):
R1 = T1t1C1

R5 = T1t2C2

R3 = T1t2C1

R7 = T2t2C1

R6 = T2t1C2

R2 = T1t1C2

siendo:

R8 = T2t2C2

R4 = T2t1C1
T1 = 1- (T
t1 = 1- (t

(T-Tmn)

ma x-Tmn)

(t-tmn)

ma x-tmn)

C1 = 1- (C

(C-Cmn)

ma x-Cmn)

T2 = (T

t2 = (t

(t-tmn)

ma x-tmn)

C2 = (C

(T-Tmn)

ma x-Tmn)

(C-Cmn)

ma x-Cmn)

y a1, a2, y a8 son los parmetros constantes correspondientes al consecuente aplicado.


135

Material y Mtodos
En las funciones de pertenencia se observa que stas valen 1 para valores altos, y 0
para valores bajos.
A veces pueden considerarse otras funciones de pertenencia ms complejas con la
finalidad de conseguir modelos ms precisos y con mayores valores de los coeficientes de
regresin (R2). En nuestro caso de tres variables independientes, puede introducirse en una

de ellas un nivel ms, de forma que en sta habra entonces tres niveles (bajo, medio y alto),
en lugar de los dos niveles para las otras dos variables (alto y bajo). De esta manera se
tendrn entonces 12 reglas borrosas: combinaciones de los tres niveles de la variable
independiente para la que se han elegido, con los otros dos niveles de las otras dos variables
independientes.
Si cuando se utilizan tres niveles para una denominada variable independiente se
utiliza una funcin de pertenencia de tipo gaussiana:
f = exp (-0.5(X-Xc/L)2)

[7]

donde X es el valor absoluto de la variable considerada, Xc el valor central de la variable y L

el ancho de la distribucin gaussiana de la variable en cuestin, la ecuacin matemtica del

modelo borroso tiene 12 trminos en el numerador (en lugar de 8), donde, si la T es la


variable con tres niveles, las reglas borrosas seran:
R1 = T1t1C1

R2 = T1t1C3

R5 = T1t3C3

R6 = T3t1C3

R9 = T2t1C1

R10 = T2t1C3

donde 1 es el nivel bajo, 2 el medio y 3 el alto, y:


T1 = exp (-0.5(T-Tmn/L)2)
t1 = 1- (t

R4 = T3t1C1

R7 = T3t3C1

R8 = T3t3C3

R11 = T2t3C1

R12 = T2t3C3

T2 = exp (-0.5(T-Tmedia/L)2) T3 = exp (-0.5(T-Tmx/L)2)


(t-tmn)

ma x-tmn)

C1 = 1- (C

R3 = T1t3C1

(C-Cmn)

ma x-Cmn)

t2 = (t

(t-tmn)

ma x-tmn)

C2 = (C

(C-Cmn)

ma x-Cmn)

Para la estimacin de los parmetros o valores constantes de la ecuacin se utiliza la


herramienta informtica ANFIS Edit (Adaptative Neural Fuzzy Inference System). Una red
equivalente de un sistema borroso puede representarse de la forma que se presenta en la
Figura 4.19.

136

Material y Mtodos
En el esquema mostrado abajo se observa que existen Xn variables de entrada, de

forma que se le asignan M funciones de pertenencia a cada variable; dichas funciones de


pertenencia se combinan entre ellas M veces dando lugar al nmero de reglas borrosas del

modelo; a cada regla se le aplica el consecuente que se elija o desborrosificador para que
integrando todos ellos se tenga la modelizacin del sistema.

Figura 4.19. Estructura de un sistema neuroborroso.

137

Material y Mtodos

138

Bibliografa/References

Bibliografa/References
REFERENCIAS BIBLIOGRFICAS

Abad, S., Alonso, J. L., Santos, V., Paraj, J. C. (1997). Furfural from wood in catalyzed
acetic acid media. A mathematical assessment. Bioresource Technology, 62, 3, 115122.
Abatzoglou, N., Chornet, E., Belkacemi, K., Overend, R. P. (1992). Phenomenological
kinetics of complex systems: the development of a generalized severity parameter and
its application to lignocellulosics fractionation. Chemical Engineering and Science, 47,
1109-1122.
Aitken, Y., Cadel, F., Voillot, C. (1988). Constituants fibreux des ptes, papiers et cartons.
Pratique de lanalyse, 1re d. Grenoble: Centre Technique du Papier; St Martin
d'Heres: Ecole Francaise de Papeterie et des Industries Graphiques.
Akamatsu, I., Husin, M. B., Kamishima, H., Hassan, A. H. (1987a). Industrial utilization of
oil palm (Elaeis guineensis) by-products. I. Kraft-anthraquinone pulping of oil palm
empty fruit bunches. Cellulose Chemistry and Technology, 21, 1, 67-75.
Akamatsu, I., Kobayashi, Y., Kamishima, H., Hassan, K., Yusoff, M. N., Husin, M., Hassan,
A. H. (1987b). Industrial utilization of oil palm by-products. II. Thermomechanical
pulping of empty fruit bunches. Cellulose Chemistry and Technology, 21, 2, 191-197.
Alaejos, J., Lpez, F., Eugenio, M. E., Fernndez, M. (2004). Non-wood raw materials for
pulp and paper making. A review. Afinidad, 61, 513, 400-410.
Alen, R. (2000). Structure and chemical composition of wood, Cap. I. Papermaking Science
and Technology. Vol. 3: Forest Products Chemistry. Finnish Paper Engineers
Association. Helsinki.
Allen, S. G., Kam, L. C., Zemann, A. J., Antal, M. J. (1996). Fractionation of sugar cane with
hot, compressed, liquid water. Industrial Engineering Chemistry Research, 35, 27092715.
Andresen, M., Stenius, P. (2007). Water-in-oil emulsions stabilized by hydrophobized
microfibrillated cellulose. Journal of Dispersion Science and Technology, 28, 837-844.
Anuario FAO (2012). http://faostat.fao.org (Acceded November 2012).

141

Bibliografa/References
Aoyama, M. (1996). Steaming of bamboo grass. II. Characterization of solubilized
hemicellulose and enzymatic digestibility of water-extracted residue. Cellulose
Chemistry and Technology, 30, 385-393.
Aoyama, M., Seki, K. (1994). Chemical characterization of solubilized xylan from steamed
bamboo grass. Holz Roh Werkst., 52, 388.
Aoyama, M., Seki, K. (1999). Acid catalysed steaming for solubilisation of bamboo grass
xylan. Bioresource Technology, 69, 91-94.
Aoyama, M., Seki, K., Saito, N. (1995). Solubilization of bamboo grass xylan by steaming
treatment. Holzforschung, 49, 193-196.
Archambault, E. (2004). Canadian R&D Biostrategy: Towards a Canadian R&D Strategy for
Bioproducts and Bioprocesses. Science-Metrix preparado para el National Research
Council of Canada.
Area, M. C., Barboza, O. M., Valade, J. L. (1997). Bleaching of Eucalyptus grandis
chemimechanical pulps. TAPPI Journal, 80, 3, 141-145.
Aspapel (Informe estadstico) (2008). http://www.aspapel.es (Acceded September 2012).
Atchison, J. E., McGovern, J. N. (1993). History of paper and the importance of non-wood
plant fibres. En F. Hamilton y B. Leopold, eds. Paper and paper manufacture. Vol. 3.
Secondary fibres and non-wood pulping, p. 3. Atlanta, GA, Estados Unidos, TAPPI
Press.
Axelsson, P., Ek, M., Teder, A. (2004). Influence of alkali profiling in birch kraft pulping on
QPQP bleachability. Nordic Pulp and Paper Research Journal, 19, 1, 37-43.
Aziz, A. A., Husin, M., Mokhtar, A. (2002). Preparation of cellulose from oil palm empty
fruit bunches via ethanol digestion: effect of acid and alkali catalysts. Journal of Oil
Palm Research, 14, 91, 9-14.
Aziz, S., Sarkanen, K. V. (1989). Organosolv pulping - A review. TAPPI Journal, 72, 3, 169175.
Baeza, J., Urizar, S., Erismann, N. M., Freer, J., Schmidt, E., Durn, N. (1991). Organosolv
pulping V: Formic acid delignification of Eucalyptus globulus and Eucalyptus grandis.
Bioresource Technology, 37, 1, 1-6.

142

Bibliografa/References
Ballerini, D., Desmarquest, J. P., Pourqui, J., Nativel, F., Rebeller, M. (1994). Ethanol
production from lignocellulosics: Large scale experimentation and economics.
Bioresource Technology, 50, 17-23.
Ballesteros, I., Oliva, J. M., Navarro, A. A., Gonzlez, A., Carrasco, J., Ballesteros, M.
(2000). Effect of chip size on steam explosion pretreatment of softwood. Applied
Biochemistry and Biotechnology, 84-86, 97-110.
Barba, C. (2002). Sntesis de carboximetilcelulosa (CMC) a partir de pastas de plantas
anuales. Tesis Doctoral. Tarragona.
Barcel, B. J., Nicols, G., Sabater, B., Snchez-Tams, R. (2009). Fisiologa vegetal.
(Primera edicin 1980 y las 11 siguientes hasta 2009). Ed. Pirmide, Madrid. 566 pp.
and ISBN: 978-84-368-1525-2.
Battista, O. A., Coppick, S., Howsmon, J. A., Morehead, F. F., Sisson, W. A. (1956). Leveloff degree of polymerization. Relation to polyphase structure of cellulose fibers.
Journal of Industrial and Engineering Chemistry, 48, 333-335.
Baugh, K. D., Levy, J. A., McCarty, P. L. (1988). Thermochemical pretreatment of
lignocellulose to enhance methane fermentation: II. Evaluation and application of
pretreatment model. Biotechnology and Bioengineering, 31, 62-70.
Baugh, K. D., McCarty, P. L. (1988). Thermochemical pretreatment of lignocellulose to
enhance methane fermentation: I. Monosaccharide and furfurals hydrothermal
decomposition and product formation rates. Biotechnology and Bioengineering, 31,
50-61.
Been, J. (1995). Bleaching. A novel approach to kinetic modeling of the hydrogen peroxide
brightening of mechanical pulp. TAPPI Journal, 78, 8, 144-152.
Berglund, L., (2005). Chapter 26: Cellulose-based nanocomposites. In Natural Fibers,
biopolymers, and biocomposites (pp. 807-832). FL, USA: CRC Press, Boca Raton.
Berthold, J., Salmn, L. (1996). Effects of mechanical and chemical treatments on the pore
size distribution in wood pulps examined by inverse size exclusion chromatography
(ISEC). Nordpap (Oslo), Vol. 670, SCAN forsk-rapport.
Biermann, C. J., Schultz, T. P., McGinnis, G. D. (1984). Rapid steam hydrolysis/extraction
of mixed hardwoods as a biomass pretreatment. Journal of Wood Chemistry and
Technology, 4, 111-128.
143

Bibliografa/References
Bodin, A., Backdahl, H., Risberg, B., Gatenholm, P. (2007). Nanocellulose as a scaffold for
tissue engineered blood vessels. Tissue Engineering, 13, 885-885.
Bokstrom, M., Tuomi, A. (2001). Interest in applying ozone growing. Paperi Ja Puu-Paper
and Timber, 83, 2, 124-127.
Bonn, G., Concin, R., Bobleter, O. (1983). Hydrothermolysis - A new process for the
utilization of biomass. Wood Science and Technology, 17, 195-202.
Botello, J. I., Gilarranz, M. A., Rodrguez, F., Oliet, M. (1999a). Preliminary study on
products distribution in alcohol pulping of Eucalyptus globulus. Journal of Chemistry
Technology and Biotechnology, 74, 2, 141-148.
Botello, J. I., Gilarranz, M. A., Rodrguez, F., Oliet, M. (1999b). Recovery of solvent and
byproducts from organosolv black liquor. Separation Science and Technology, 34, 12,
2431-2445.
Bouchard, J., Nguyen, T. S., Chornet, E., Overend, R. P. (1991). Analytical methodology for
biomass pretreatment. Part 2: Characterization of the filtrates and cumulative product
distribution as a function of treatment severity. Bioresource Technology, 36, 121-131.
Bouchard, J., Nugent, H. M., Berry, R. M. (1995). A comparison between acid treatment
and chelation prior hydrogen-peroxide bleaching of kraft pulps. Journal of Pulp and
Paper Science, 21, 6, 203-208.
Bourbonnais, R., Paice, M. G. (1996). Enzymatic delignification of kraft pulp using laccase
and a mediator. TAPPI Journal, 79, 199-204.
Brelid, H., Friberg, T. (1998). TCF bleaching of coniferas kraft pulp. 5. Investigation of
EDTA-metal complexes in chelation filtrates. Nordic Pulp and Paper Research
Journal, 13, 2, 112-118.
Brown, D. G., Abbot, J. (1995). Effects of metal-ions and stabilizers on peroxide
decomposition during bleaching. Journal of Wood Chemistry and Technology, 15, 85111.
Brown, L. R., Flavin, C., French, H. (2000). La situacin del mundo 2000. Editorial Icaria,
Barcelona (ISBN: 84-7426-472-3).
Brownell, H. H., Yu, E. K. C., Saddler, J. N. (1986). Steam-explosion pretreatment of wood:
effect of chip size, acid, moisture content and pressure drop. Biotechnology and
Bioengineering, 28, 792-801.
144

Bibliografa/References
Bryant, P. S., Edwards, L. L. (1996). Cation exchange of metals on Kraft pulp. Journal of
Pulp and Paper Science, 22, 1, 37-42.
Bryant, P. S., Robarge, K., Edwards, L. L. (1993). Transition-metal profiles in open and
closed Kraft fiber lines. TAPPI Journal, 76, 10, 148-159.
Burtscher, E., Bobleter, O., Schwald, W., Concin, R., Binder, H. (1987). Chromatographic
analysis of biomass reaction products produced by hydrothermolysis of poplar wood.
Journal of Chromatography, 390, 2, 401-412.
Byrd, J. M. V., Grazel, J. S., Singh, R. P. (1992). Delignification and bleaching of chemical
pulps with ozone: a literature review. TAPPI Journal, 75, 3, 207-213.
Camarero, S., Garca, O., Vidal, T., Colom, J., del Rio, J. C., Gutirrez, A., Gras, J. M.,
Monje, R., Martnez, M. J., Martnez, A. T. (2004). Efficient bleaching of non-wood
high-quality paper pulp using laccase-mediator system. Enzyme and Microbial
Technology, 35, 2-3, 113-120.
Campbell, J. M., Fahey, G. C., DeMichele, S. J., Garleb, K. A. (1997). Metabolic
characteristics of healthy adult males as affected by ingestion of a liquid nutritional
formula containing fish oil, oligosaccharides, gum arabic and antioxidant vitamins.
Food Chemistry Toxicology, 35, 1165-1176.
Caas, A. I., Camarero, S. (2010). Laccases and their natural mediators: Biotechnological
tools for sustainable eco-friendly processes. Biotechnology Advances, 28, 6, 694-705.
Cao, B. (1996). Comparative study of soda ash-anthraquinone pulping of wheat straw.
Pulping Conference, 387-392. Proceedings TAPPI. TAPPI Press.
Cao, B., Zhong-Zheng, L. (1996). The effects of hydrogen peroxide and anthraquinone on
soda ash pulping of wheat straw. Holzforschung, 50, 62-68.
Caparrs, S. (2007). Fraccionamiento integral de vegetales no alimentarios para la obtencin
de pasta celulsica y subproductos. Tesis Doctoral. Escuela Politcnica Superior.
Huelva.
Caparrs, S., Ariza, J., Lpez, F., Daz, M. J. (2007). Optimizing cellulosic paper obtained
from Arundo Donax L. under hydrothermal treatment. Journal of Industrial and
Engineering Chemistry, 13, 3, 465-473.

145

Bibliografa/References
Caparrs, S., Ariza, J., Lpez, F., Nacimiento, J. A., Garrote, G., Jimnez, L. (2008a).
Hydrothermal treatment and ethanol pulping of sunflower stalks. Bioresource
Technology, 99, 5, 1368-1372.
Caparrs, S., Daz, M. J., Ariza, J., Lpez, F., Jimnez, L. (2008b). New perspectives for

Paulownia fortunei L. Valorisation of the autohydrolysis and pulping processes.


Bioresource Technology, 99, 741-749.
Cara, C., Romero, I., Oliva, J. M., Sez, F., Castro, E. (2007). Liquid hot water pretreatment
of olive tree pruning residues. Applied Biochemistry and Biotechnology, 137-140, 112, 379-394.
Carrasco, F. (1989). Fundamentos del fraccionamiento de la biomasa. Afinidad, 46, 425429.
Carrasco, F., Chornet, E., Overend, R. P., Heitz, M. (1987). Fractionnement de deux bois
tropicaux (Eucalyptus et Wapa) par traitment thermomecanique en phase aqueuse.
Partie II: Characteristiques chimiques des rsidus et considerations cinetiques sur la
solubilisation des hemicelluloses. Canadian Journal Chemical Engineering, 65, 71-77.
Carrasco, F., Roy, C. (1992). Kinetic study of dilute-acid prehydrolysis of xylan-containing
biomass. Wood Science and Technology, 26, 189-208.
Cartagena, M. C. (1994). La Biosfera IV: el rbol muerto como fuente de materia prima. In
Vian Ortuo, A. (ed). Introduccin a la Qumica Industrial. Revert, Barcelona,
Espaa.
Casey, J. P. (1990). Pulpa y papel. Noriega-Limusa. Mxico.
Chen, C. L., Capanema, E. A., Gracz, H. S. (2003). Comparative studies on the
delignification of pine kraft-anthraquinone pulp with hydrogen peroxide by binucleus
Mn (IV) complex catalysis. Journal of Agricultural and Food Chemistry, 51, 21, 62236232.
Chen, J. Y., Shimizu, Y., Takai, M., Hayashi, J. (1995). Effect of kraft cooking conditions on
the efficiency of oxygen delignification. I. The structural characteristics of kraft and
residual lignins. Mokuzai gakkaishi, 41, 8, 758-766.
Cherubini, F. (2010). The biorefinery concept: Using biomass instead of oil for producing
energy and chemicals. Energy Conversion and Management, 51, 7, 1412-1421.

146

Bibliografa/References
Chirat, C., Lachenal, D. (1997). Effect of hydroxyl radicals on cellulose and pulp and their
occurrence during ozone bleaching. Holzforschung, 51, 2, 147-154.
Cimpoesu, G., Obrocea, P. (1995). Research on the delignification of spruce wood by
organic peroxyacids. Cellulose Chemistry and Technology, 29, 4, 487-496.
CIRAD (2007). Dossier sobre palma de aceite.
Claus, I., Kordsachia, O., Schrder, N., Karstens, T. (2005). Monoethanolamine (MEA)
pulping of beech and spruce wood for production of dissolving pulp. Holzforschung,
58, 6, 573-580.
Colom, J. F. (2005). Introduccin a la fabricacin de pastas y papel. Evolucin histrica.
Importancia y situacin de la industria papelera en el contexto industrial. En:
Introduccin a la historia papelera. Iniciacin al conocimiento del papel. Servicio de
Publicaciones de la ETSII. Terrassa. Barcelona.
Commission of the european communities (2007). Communication from the commission to
the council, the european parliament, the european economic and social committee
and the committee of the regions. A lead market initiative for Europe. COM, 860 final
SEC(2007)1729.
Conner, A. H., Lorenz, L. F. (1986). Kinetic modeling of hardwood prehydrolysis. Part III:
Water and dilute acetic acid prehydrolysis of southern red oak. Wood and Fiber
Science, 18, 2, 248-263.
Cox, L. A., Worstor, H. E. (1971). An assessment of some sulfur free chemical pulping
process. TAPPI Journal, 54, 11, 1890-1892.
Cronlund, M., Powers, J. (1992). Bleaching of Alcell organosolv pulps using conventional
and nonchlorine bleaching sequences. TAPPI Journal, 75, 6, 184-194.
Cui, Y., Puthson, P., Chen, C. L., Gratzl, J. S., Kirkman, A. G. (2000). Kinetic study on
delignification of kraft-AQ pine pulp with hydrogen peroxide catalyzed by Mn (IV)Me4DTNE. Holzforschung, 54, 4, 413-419.
Daud, W. R. W., Law, K. N., Ghazali, A. (2004a). Alkaline peroxide mechanical pulping of
oil palm lignocellulosics. Part 1. Effects of chemical and mechanical pretreatments on
empty fruit bunch (EFB) chemical composition. JIRCAS Working Report, 39, 82-85.
Daud, W. R. W., Law, K. N., Valade, J. L. (1998). Chemical pulping of oil palm empty fruit
bunches. Cellulose Chemistry and Technology, 32, 133-143.
147

Bibliografa/References
Daud, W. R. W., Leh, C. P., Zainuddin, Z., Tanaka, R. (2005). Optimisation of soda
pulping variables for preparation of dissolving pulps from oil palm fibre.
Holzforschung, 57, 1, 106-113.
Daud, W. R. W., Zainuddin, Z., Lee, L. K. (2004b). Influence of pulping variables on the
properties of Elaeis guineensis soda pulp as evaluated by response surface
methodology. Wood Science and Technology, 38, 3, 191-205.
De Souza, M. M., Borsali, R. (2004). Rodlike cellulose microcrystals: structure, properties,
and applications. Macromolecular Rapid Communications, 25, 7, 771-787.
Deinko, I. P., Makarova, O. V., Zarubin, M. Y. (1992). Delignification of wood by oxygen in
low-molecular-weight alcoholic media. TAPPI Journal, 75, 9, 136-140.
Dence, C. W., Reeve, D. W. (1996). Pulp bleaching: principles and practice. TAPPI Press,
Atlanta, GA, 4, 1, 215-239.
Dinand, E. Chanzy, H., Vignon, M. R. (1996). Parenchymal cell cellulose from sugar beet
pulp: preparation and properties. Cellulose, 3, 183-188.
Dinand, E., Chanzy, H., Vignon, M. R. (1999). Suspensions of cellulose microfibrils from
sugar beet pulp. Food Hydrocolloids, 13, 3, 275-283.
Dixon, W. J. (1992). BMDP Statistical software manual. University of California Press,
(ISBN: 9780520081406) (178 pp.).
Domnguez, H., Paraj, J. C., Cruz, J. M., Garrote, G. (2004). Production of
xylooligosaccharides by autohydrolysis of lignocellulosic materials. Trends in Food
Science and Technology, 15, 3, 115-120.
Dong, X. M., Kimura, T., Revol, J. F., Gray, D. G. (1996). Effects of ionic strength on the
isotropic-chiral nematic phase transition of suspensions of cellulose crystallites.
Langmuir, 12, 8, 2076-2082.
Draper, N., Smith, H. (1981). Applied Regression Analysis. Wiley. New York.
Dufresne A., Dupeyre, D., Vignon, M. R. (2000). Cellulose microfibrils from potato tuber
cells: Processing and characterization of starch-cellulose microfibril composites.
Journal of Applied Polymer Science, 76, 14, 2080-2092.
Dyhr, K., Sterte, J. (1998). Use of zeolites in hydrogen peroxide bleaching of pulp. Nordic
Pulp and Paper Research Journal, 23, 4, 257-262.

148

Bibliografa/References
Dyhr, K., Sterte, J. (1999). Effects of zeolite addition on the manganese catalyzed
decomposition of hydrogen peroxide. Journal of Porous Materials, 6, 4, 275-282.
Emmanouilides, C., Petrou, L. (1997). Identification and control of anaerobic digesters using
adaptive, on-line trained neural networks. Computers and Chemical Engineering, 21,
1, 113-143.
Erismann, N. M., Freer, J., Baeza, J., Durn, N. (1994). Organosolv pulping-VII:
Delignification selectivity of formic acid pulping of Eucalyptus grandis. Bioresource
Technology, 47, 3, 247-256.
Eroglu, H., Tutus, A. (2004). The history and current status of the paper and paperboard
industry in Turkey. APPITA Journal, 57, 4, 268-273.
Eugenio, M. E., Santos, S. M., Carbajo, J. M., Martn, J. A., Martn-Sampedro, R., Gonzlez,
A. E., Villar, J. C. (2010). Kraft pulp biobleaching using an extracellular enzymatic
fluid produced by Pycnoporus sanguineus. Bioresource Technology, 101, 6, 18661870.
Fengel, D., Wegener, G. (1989). Wood, chemistry, ultrastructure, reactions. De Gruyter.
612. Berlin.
Fergus, B. J., Goring, D. A. I. (1970). The location of guaiacyl and syringyl lignins in birch
xylem tissue. Holzforschung, 24, 113-117.
Fernndez, N. (1996). Desarrollo de pulpas y papeles de bagazo. Revista ICIDCA sobre los
derivados de la caa de azcar, 30, 114-129.
Fillat, A., Colom, J. F., Vidal, T. (2010). A new approach to the biobleaching of flax pulp
with laccase using natural mediators. Bioresource Technology, 101, 11, 4104-4110.
Finnegin, D., Stack, K. R., Dunn, L. (1998). Peroxide bleaching using zeolites. Part I:
peroxide decomposition. APPITA Journal, 51, 3, 219-223.
Fleming, B. I., Kubes, G. J., McLeod, J. M., Bolker, H. I. (1979). Soda pulping with
nitrogenous redox catalysts. TAPPI Journal, 62, 6, 57-59.
Friman, L., Hglund, H., Hgberg, H. E., Agnemo, R. (2004). Tannin-iron impregnated
thermomechanical pulp - Part II: Bleachability and brightness reversion. Nordic Pulp
and Paper Research Journal, 19, 4, 525-531.

149

Bibliografa/References
Funazukuri, T., Takahashi, H., Miyajima, K., Wakao, N. (1995). Kraft pulp bleaching with
hydrogen peroxide in a batch reactor. Kagaku kgaku ronbunsh, 21, 5, 866-870.
Garca, J. A. (2007). Fibras papeleras. Ediciones UPC. Barcelona. Espaa.
Garca, J. A., Vidal, T. (1984). Blanqueo de pastas en la industria papelera. Publicacin de la
Universidad Politcnica de Catalua. Barcelona.
Garca, J. A., Vidal, T., Colom, J. F. (1985). Caractersticas de los efluentes de blanqueo.
Ingeniera qumica, 4, 71-82.
Garrote, G., Domnguez, H., Paraj, J. C. (2001). Manufacture of xylose-based,
fermentation media from corncobs by posthydrolysis of autohydrolysis liquors.
Applied Biochemistry and Biotechnology, 95, 3, 195-207.
Gartner, A., Gellerstedt, G. (2000). Oxidation of residual lignin with alkaline hydrogen
peroxide. Part I: changes in hydrophilic groups. Journal of Pulp and Paper Science,
26, 12, 448-454.
Gaspar, A., Evtuguin, D. V., Pascoal Neto, C. (2003). Oxygen bleaching of kraft pulp
catalysed by Mn (III)-substituted polyoxometalates. Applied Catalysis A: General 239,
157-168.
Ghazali, A., Wanrosli, W. D., Law, K. N. (2006). Alkaline peroxide mechanical pulping
(APMP) of oil palm lignocellulosics. Part II: Empty fruit bunch (EFB) responses to
pretreatments. APPITA Journal, 59, 1, 65-70.
Gierer, J., Jansbo, K., Reitberger, T. (1993). Formation of hydroxyl radicals from hydrogen
peroxide and its effects on bleaching of mechanical pulp. Journal of Wood Chemistry
and Technology, 13, 4, 561-581.
Gilarrahz, M. A., Oliet, M., Rodrguez, F., Tijero, J. (1998). Ethanol-water pulping: cooking
variables optimization. The Canadian Journal of Chemical Engineering, 76, 2, 253260.
Ginting, Y. A., Abbot, J. (1993). Testing of equilibrium kinetics models for peroxide
bleaching of Pinus radiata TMP. Journal of Pulp and Paper Science, 19, 4, 143-151.
Girard, R., Van Heiningen, A. R. P. (1997). Yield determination and mass balance closure
of batch ethanol-based solvent cookings. Pulp and Paper Canada, 98, 8, 65-68.

150

Bibliografa/References
Goenadi, D. H., Away, Y., Suharyanto, P., Panji, T., Watanabe, T., Kuwahara, M. (1998).
Bio-pulping of empty fruit bunches of oil palm by white rot fungi isolated from tropical
plantation. International Conference on Biotechnology in the Pulp and Paper
Industry, 7th, Vancouver, B.C., June 16-19, B, 49-51.

Gosselink, R. J. A., Dam, J. E. G., Zomers, F. H. A. (1995). Combined HPLC analysis of


organic acids and furans formed during organosolv pulping of fiber hemp. Journal of
Wood Chemistry and Technology, 15, 1, 1-25.
Goyal, S. K., Ray, A. K. (1989). Economic comparison of soda and soda AQ pulping
processes for cereal straws. Pulping Conference: [proceedings], 775-780.
Guerra A., Ferraz, A. (2005). Evaluation of bleachability and brightness stability of
thermomechanical pulp from Eucalyptus grandis. Papel -Sao Paulo-, 66, 8, 74-79.
Guha, S. R. D., Sirmokadan, N. N., Goudar, T. R., Rohini, D. (1987). Pulping rice straw by
pollution free organosolv process. IPPTA, 24, 1, 24-26.
Gullichsen, J., Paulapuro, H. (2000). Chemical Pulping. Vol. 6A y 6B. Fapet Oy. Finlandia.
Gulln, P., Gonzlez-Muoz, M. J., Domnguez, H., Paraj, J. C. (2008). Membrane
processing of liquors from Eucalyptus globulus autohydrolysis. Journal of Food
Engineering, 87, 2, 257-265.
Hassan, E. B. M. (2003). Oxygen-peroxide (Eop) delignification and peroxide (P) bleaching
of bagasse kraft pulp. Journal of Scientific and Industrial Research, 62, 7, 651-744.
Heitz, M., Capek-Mnard, E., Koeberle, P. G., Gagn, J., Chornet, E., Overend, R. P.,
Taylor, J. D., Yu, E. (1991). Fractionation of Populus tremuloides at the pilot plant
scale: optimization of steam pretreatment conditions using the STAKE II technology.
Bioresource Technology, 35, 1, 23-32.
Heitz, M., Carrasco, F., Rubio, M., Chauvette, G., Chornet, E., Jaulin, L., Overend, R. P.
(1986). Generalized correlations for the aqueous liquefaction of lignocellulosics.
Canadian Journal Chemical Engineering, 64, 647-650.
Henriksen, H. C., Naes, T., Rodbotten, R., Aastveit, A. (2004). Prediction of important
sulphite pulp properties from near infrared spectra: a feasibility study and comparison
of methods. Journal of Near Infrared Spectroscopy, 12, 5, 279-286.
Henriksson, M., Berglund, L. A., Isaksson, P., Lindstrm, T., Nishino, T. (2008). Cellulose
nanopaper structures of high toughness. Biomacromolecules, 9, 6, 1579-1585.
151

Bibliografa/References
Henriksson, M., Henriksson, G., Berglund, L. A., Lindstrm, T. (2007). An
environmentally friendly method for enzyme-assisted preparation of microfibrillated
cellulose (MFC) nanofibers. European Polymer Journal, 43, 8, 3434-3441.
Hergert, H. L. (1998). Developments in organosolv pulping. An overview. Enviromentally
friendly technologies for the pulp and paper industry.
Herrick, F. W., Casebier, R. L., Hamilton, J. K., Sandberg, K. R. (1983). Microfibrillated
cellulose: morphology and accessibility. Journal of Applied Polymer Science: Applied
Polymer Symposium, 37, 797-813.
Heux, L., Chauve, G., Bonini, C. (2000). Nonflocculating and chiral-nematic self-ordering of
cellulose microcrystals suspensions in nonpolar solvents. Langmuir, 16, 21, 8210-8212.
Hobbs, G. C., Abbot, J. (1994a). The effect of aluminum on the two-stage peroxide
bleaching of Eucalyptus regnans stone groundwood pulp. Journal of Pulp and Paper
Science, 20, 2, 50-54.
Hobbs, G. C., Abbot, J. (1994b). The role of the hydroxyl radical in peroxide bleaching
processes. Journal of Wood Chemistry and Technology, 14, 2, 195-225.
Holton, H. (1977). Soda additive softwood pulping: a major new process. Pulp and Paper
Canada, 78, 10, 19-24.
Holtzapple, M. T. (1993) Hemicelluloses. En: Encyclopaedia of food science, food
technology and nutrition. Ed. Academic Press. Londres, 2324-2334.
http://www.arbolesornamentales.com (Fundacin Ecologa y Desarrollo (2001). Zaragoza.
Gnero de Palmas) (Acceded September 2012).
http://www.bioproducts-bioenergy.gov/pdfs/FinalBiomassRoadmap.pdf

(BTAC

(2002).

Roadmap for Biomass Technologies in the United States. Biomass Technical


Advisory Committee. Washington, DC. Biomass Research and Development
Initiative) (Acceded September 2012).
http://www.wrm.org.uy (Acceded September 2012).
http://www1.eere.energy.gov/biomass/pdfs/final_2006_vision.pdf (BRDI (2006). Vision for
bioenergy and biobased products in the United States. Biomass Research and
Development Initiative) (Acceded September 2012).

152

Bibliografa/References
Hubbe, M. A., Rojas, O. J., Lucia, L. A., Sain, M. (2008). Cellulosic nanocomposites: a
review. BioResources, 3, 3, 929-980.
Hunt, R., Causton, D. R., Shipley, B., Askew, A. (2002). A modern tool for classical plant
growth analysis. Ann Bot., 90, 485-488.
Ibarra, D., Camarero, S., Romero, J., Martnez, M. J., Martnez, A. T. (2006). Integrating
laccasemediator treatment into an industrial-type sequence for totally chlorine-free
bleaching of eucalypt kraft pulp. Journal of Chemical Technology and Biotechnology,
81, 7, 1159-1165.
Ibrahim, R. (2002). Chemical composition of alkaline pulps from oil palm empty fruit
bunches. Oil Palm Bulletin, 44, 19-24.
IDAE (2006). Manuales de energas renovables: Biocarburantes en el transporte.
Iwamoto, S., Nakagaito, A. N., Yano, H., Nogi, M. (2005). Optically transparent composites
reinforced with plant fiber-based nanofibers. Applied Physics A: Materials Science &
Processing, 81, 1109-1112.
Jahan, M. S. (2001). Soda-amine pulping of cotton stalk. Pulping Conference, United States,
1175-1183.
Jahan, M. S., Farouqui, F. I. (2000). Pulping of whole jute plant (Corchorus capsularis) by
sodaamine process. Holzforschung, 54, 6, 625-630.
Jahan, M. S., Farouqui, F. I. (2003). Kinetics of jute pulping by soda-amine processes.
Cellulose Chemistry and Technology, 36, 3-4, 357-366.
Jahan, M., Farouqui, F., Abdullah, I. S. (2001). Pulping of jute with amines. Cellulose
Chemistry and Technology, 35, 1-2, 177-187.
Jeong, K. J., Park, I. Y., Kim, M. S., Kim, S. C. (1998). High-level expression of an
endoxylanase gene from Bacillus sp. in Bacillus subtilis DB104 for the production of
xylobiose from xylan. Applied Microbiology and Biotechnology, 50, 113-118.
Jimnez, L. (2005). Pastas celulsicas de materias primas alternativas a las convencionales.
Editorial Grficas Sol (ISBN: 84-96302-10-5).
Jimnez, L., Maestre, F. (1997). Disolventes orgnicos para la obtencin de pastas de
celulosa. Revisin bibliogrfica. Afinidad, 54, 467, 45-50.

153

Bibliografa/References
Jimnez, L., Navarro, E., de la Torre, M. J., Prez, I. (1996). El ozono en el blanqueo de
pastas celulsicas para papel. Primera Parte. El Papel (ISSN 1134-8062), 59, 11, 5056.
Jimnez, L., Navarro, E., de la Torre, M. J., Prez, I. (1997). El ozono en el blanqueo de
pastas celulsicas para papel. Segunda Parte. El Papel (ISSN 1134-8062), 60, 12, 3134.
Jimnez, L., Rodrguez, A. (2009). Fabricacin de papel a partir de materias primas
alternativas a las convencionales. Ed. Grficas Sol. (SE-7355-2008).
Jimnez, L., Rodrguez, A., Calero, A. M., Eugenio, M. E. (2004). Use of ethanolaminesoda-water mixtures for pulping olive wood trimmings. Chemical Engineering
Research and Design, 82, 8, 1037-1042.
Jimnez, L., Villar, J. C., Rodrguez, A., Jimnez, R., Calero, A. M. (2002). Influencia de las
variables del proceso de pasteado de podas de olivo con etanolamina y sosa sobre la
composicin de las pastas. Afinidad, 59, 500, 399-408.
Johansson, A., Aaltonen, O., Ylinen, P. (1987). Organosolv pulping: methods and pulp
properties. Biomass, 13, 1, 45-65.
Judt, M. (1990). Recent developments in pulp production and their suitability for use in
industrial process in developing countries. Cellulose Sources Exploitation, 81-88.
Jung, H. J., Fahey, G. C. Jr. (1983). Effects of phenolic monomers on rat performance and
metabolism. Journal of Nutrition, 113, 3, 546-556.
Kaar, W. E., Gutirrez, C. V., Kinoshita, C. M. (1998). Steam explosion of sugarcane
bagasse as a pretreatment for conversion to ethanol. Biomass and Bioenergy, 14, 3,
277-287.
Kadla, J. F., Chang, H. M., Jameel, H. (1999). The reactions of lignins with high
temperature hydrogen peroxide. Part 2. The oxidation of kraft lignin. Holzforschung,
53, 3, 277-284.
Kamel, S. (2007). Nanotechnology and its applications in lignocellulosic composites, a mini
review. eXPRESS Polymer Letters, 1, 9, 546575.
Kamishima, H., Matsuo, R., Fukuoka, S., Akamatsu, I., Kobayashi, Y., Husin, M., Miswar,
O., Haji, H. (1988). Industrial utilization of oil palm by-products. III.

154

Bibliografa/References
Thermomechanical pulping of oil palm fronds. Cellulose Chemistry and Technology,
22, 3, 351-361.
Kamm, B., Gruber, P. R., Kamm, M. (2006). Biorefineries - industrial processes and
products: status quo and future directions, (ISBN: 978-3527310272).
Kamm, B., Kamm, M. (2004). Principles of biorefineries. Applied Microbiology and
Biotechnology, 64, 2, 137-145.
Kamm, B., Kamm, M. (2007). Biorefineries-multi product processes. White Biotechnology
(Advances in Biochemical Engineering/Biotechnology), 105, 175-204.
Kantelinen, A., Hortling, B., Sundquist, J., Linko, M., Viikari, L. (1993). Proposed
mechanism of the enzymatic bleaching of kraft pulp with xylanases. Holzforschung, 47,
4, 318-324.
Khoo, K. C., Lee, T. W. (1991). Pulp and paper from the oil palm. APPITA Journal, 44, 6,
385-388.
Kirwan, M., Strawbridge, J. (2003). Plastics in food packaging. In: Coles, R., McDowell, D.,
Kirwan, M. J., (eds.). Food Packaging Technology, CRC Press LLC, Boca Raton. (pp.
174-240).
Kleinert, T. N. (1974). Organosolv pulping with aqueous alcohol. TAPPI Journal, 57, 8, 99102.
Klemm, D., Schumann, D., Udhardt, U., Marsch, S. (2001). Bacterial synthesized cellulose artificial blood vessels for microsurgery. Progress in Polymer Science, 26, 9, 15611603.
Kokta, B. V., Chen, R., Daneault, C., Valade, J. L. (1983). Use of wood fibers in
thermoplastic composites. Polymer Composites, 4, 4, 229232.
Kubes, G. J., Fleming, B. I., MacLeod, J. M., Bolker, H. I. (1978). Sulfur free
delignification. Redox potentials and pulping rates of soda amine liquors. TAPPI
Journal, 61, 8, 46-50.
Lamptey, J., Robinson, C. W., Moo-Young, M. (1985). Enhanced enzymatic hydrolysis of
lignocellulosic

biomass

pretreated

by

low-pressure

steam

autohydrolysis.

Biotechnology Letters, 7, 531-536.

155

Bibliografa/References
Landucci, L. L. (1980). Quinones in alkaline pulping. Characterization of an
anthrahydroquinone-quinone methide intermediate. TAPPI Journal, 63, 7, 95-99.
Lapierre, L., Berry, R., Bouchard, J. (2003). The effect of magnesium ions and chelants on
peroxide bleaching. Holzforschung, 57, 6, 627-633.
Lapierre, L., Paleologou, M., Berry, R., Bouchard, J. (1997). The limits of metal removal
from kraft pulp by acid treatment. Journal of Pulp and Paper Science, 23, 11, 539-542.
Larsson, S., Palmqvist, E., Hahn-Hgerdal, B., Tengborg, C., Stenberg, K., Zacchi, G.,
Nilvebrant, N. O. (1999). The generation of fermentation inhibitors during dilute acid
hydrolysis of softwood. Enzyme and Microbial Technology, 24, 3-4, 151-159.
Law, K. N., Jiang, X. (2001). Comparative papermaking properties of oil-palm empty fruit
bunch. TAPPI Journal, 84, 1, 1-13.
Leduc, C., Rouaix, S., Turcotte, F., Daneault, C. (2005). Use of zeolites for the peroxide
bleaching of mechanical pulp. Pulp and Paper Canada, 106, 4, 51-55.
Leh, C. P., Tanaka, R., Ikeda, T., Wan Rosli, W. D., Magara, K., Hosoya, S. (2004). Totally
chlorine-free bleaching of oil palm empty bunch (EFB) chemical pulps. JIRCAS
Working Report, 39, 95-99.
Lehnen, R., Saake, B., Nimz, H. H. (2005). Furfural and hydroxymethylfurfural as byproducts of FORMACELL pulping. Holzforschung, 55, 2, 199-204.
Libby, C. E. (1976). Ciencia y tecnologa sobre pulpa y papel. Tomo II. Papel. 3 Edicin
Ca. Editorial Continental sa Mxico-Espaa-Argentina-Chile. Venezuela. MARCO,
PABLO.
Lichtenthaler, F. W., Peters, S. (2004). Carbohydrates as green raw materials for the
chemical industry. Comptes Rendus Chimie, 7, 6590.
Liden, J., Ohman, L. O. (1997). Redox stabilization of iron and manganese in the +11
oxidation state by magnesium precipitates and some anionic polymers. Implications
for the use of oxygen-based bleaching chemicals. Journal of Pulp and Paper Science,
23, 5, 193-199.
Lin, S. Y., Dence, C. W. (1992). Methods in lignin chemistry. Springer-Verlag, Berlin.
Lliskovic, N., Kules, M. (1992). Cellulose peroxide bleaching with a new activator. Cellulose
Chemistry and Technology, 26, 4, 437-441.

156

Bibliografa/References
Lohrasebi, A. H., Paszner, L. (2001). Predictive modeling of high pressure naem-catalyzed
(HP-Alpulp) organosolv pulping of spruce. TAPPI Journal, 84, 4, 1-17.
Lpez, F., (1991). Contribucin al estudio de la obtencin de pastas celulsicas a partir de
residuos agrcolas. Tesis Doctoral. Universidad de Crdoba.
Lpez, F., Alfaro, A., Garca, M. M., Daz, M. J., Calero, A. M., Ariza, J. (2004). Pulp and
paper from tagasaste (Chamaecytisus proliferus L. F. ssp palmensis). Chemical
Engineering Research and Design, 82, 8, 1029-1036.
Lpez, F., Garca, J. C., Prez, A., Garca, M. M., Feria, M. J., Tapias, R. (2010a). Leucaena

diversifolia a new raw material for paper production by soda-ethanol pulping process.
Chemical Engineering Research and Design, 88, 1-9.
Lpez, F., Jimnez, L., Garca, J. C., Feria, M. J. (2010b). Ciencia y tecnologa en el
procesado de materias primas para pasta celulsica y papel. (ISBN: 978-84-92944-132).
Lpez, F., Jimnez, L., Rodrguez, A. (2006). Los alcoholes como disolventes orgnicos para
la obtencin de pastas celulsicas. Afinidad, 63, 523, 174-182.
Lpez-Alegret, M. (1996). Fructooligosacridos. Alimentaria, 34, 21-23.
Loras, V. (1990). Blanqueo. En: Pulpa y Papel. Noriega-Limusa. Mxico.
Lovins, A. B. (2004). Winning the oil endgame: innovation for profits, jobs and security.
Rocky Mountain Institute (309 pp.).
Lwendahl, L., Samuelson, O. (1978). Carbohydrate stabilization with anthraquinone during
alkaline pulping. Polymer Bulletin, 1, 205-210.
Lowys, M. P., Desbrieres, J., Rinaudo, M. (2001). Rheological characterization of cellulosic
microfibril suspensions. Role of polymeric additives. Food Hydrocolloids, 15, 1, 2532.
Marchessault, R. H., Malhatra, S. L., Jones, A. Y., Perovic, A. (1983). The wood explosion
process: characterization and uses of lignin/cellulose products. In: Soltes J. (ed) Wood
and Agricultural Residues. Academic Press, New York London, 401-413.
Martn, R. (2011). Integracin del proceso kraft de obtencin de pasta de celulosa en el
esquema de una Biorrefinera. Tesis Doctoral. Universidad Complutense de Madrid.

157

Bibliografa/References
Martnez, J. M., Granado, J. M., Montan, D., Salvad, J., Farriol, X. (1995). Fractionation
of residual lignocellulosics by dilute-acid prehydrolysis and alkaline extraction:
Application to almond shells. Bioresource Technology, 52, 1, 59-67.
Martnez, J. M., Reguant, J., Montero, M. A., Montan, D., Salvad, J., Farriol, X. (1997).
Hydrolytic pretreatment of softwood and almond shells. Degree of polymerization and
enzymatic digestibility of the cellulose fraction. Industrial and Engineering Chemistry
Research, 36, 3, 688-696.
McGinnis, G. D., Shafizadeh, F. (1990). Celulosa y hemicelulosa. En: Pulpa y papel.
Qumica y tecnologa qumica. Ed. Wiley and Sons, 29-70.
Mielisch, M. J., Pratt, R., Kordsachia, O. (1996). Katalysierte Wasserstoffperoidbleiche. Das
Papier, 17, V16-V23.
Modler, H. W. (1994). Bifidogenic factors: sources, metabolism and applications.
International Dairy Journal, 4, 5, 383-407.
Moldes, D., Daz, M., Tzanov, T., Vidal, T. (2008). Comparative study of the efficiency of
synthetic and natural mediators in laccase-assisted bleaching of eucalyptus kraft pulp.
Bioresource Technology, 99, 17, 7959-7965.
Montanari, S., Roumani, M., Heux, L., Vignon, M. R. (2005). Topochemistry of
carboxylated cellulose nanocrystals resulting from TEMPO-Mediated Oxidation.
Macromolecules, 38, 5, 1665-1671.
Montan, D., Farriol, X., Salvad, J., Jollez, P., Chornet, E. (1998). Fractionation of wheat
straw by steam-explosion pretreatment and alkali delignification. Cellulose pulp and
byproducts from hemicellulose and lignin. Journal of Wood Chemistry and
Technology, 18, 2, 171-191.
Montgomery, D. C. (1991). Diseo y anlisis de experimentos. Grupo Editorial
Iberoamericana. Mexico.
Muthurajah, R. N., Peh, T. B. (1977). Manufacture of paper pulps from oil palm empty
bunch waste. Proceedings Malaysian International Symposium Palm Oil Process, 147157.
Muurinen, E. (2000). Organosolv pulping. A review and distillation study related to
peroxiacid pulping. Doctoral Thesis. University of Oulu, Finland.

158

Bibliografa/References
Muzzy, J. D., Roberts, R. S., Fieber, C. A., Faass, G. S., Mann, T. M. (1983). Pretreatment
of hardwood by continuous steam hydrolysis. En: Wood and agricultural residues. Ed.
Academic Press, Inc., 351-368. New York.
Myerly, R. C., Nicholson, M. D., Katzen, R., Taylor, J. M. (1981). The forest refinery.
Chemtech, 11, 186-192.
NDiaye, S., Rigal, L., Larocque, P., Vidal, P. F. (1996). Extraction of hemicelluloses from
poplar, Populus tremuloides, using an extruder-type twin-screw reactor: a feasibility
study. Bioresource Technology, 57, 1, 61-67.
Nabarlatz, D. A. (2006). Autohydrolysis of agricultural by-products for the production of
xylo-oligosaccharides. Doctoral Thesis. Universitat Rovira.
Nakagaito, A. N., Yano, H., (2006). Nanocomposites based on cellulose microfibril. ACS
Symposium Series, 938 (Cellulose Nanocomposites), 11, 151-168.
National Research Council (2000). Biobased industrial products, priorities for research and
commercialization. National Academic Press, Washington D. C. ISBN 0-309-05392-7.
Nehls, I., Wagenknecht, W., Philipp, B., Stscherbina, D. (1994). Characterization of
cellulose and cellulose derivatives in solution by high resolution 13C-NMR
spectroscopy. Progress in Polymer Science, 19, 1, 29-78.
Neves, F. L., Neves, J. M. (1998). Organosolv pulping: A review. Papel, 59, 8, 48-52.
Ni, Y., Dixon, C., Ooi, T. (1997). Kinetics of hydrogen peroxide bleaching of ALCELL(R)
derived pulp. The Canadian Journal of Chemical Engineering, 75, 1, 48-52.

Ni, Y., Qiu, Z. (2003). Methods to decrease manganese-induced decomposition of peroxide.


APPITA Journal, 56, 5, 355-358.
Ni, Y., Van Heiningen, A. R. P. (1996). Lignin removal from Alcell pulp by washing with
ethanol and water. TAPPI Journal, 79, 3, 239-243.
Nitsch, J., Fischedick, M., Allnoch, N., Baumert, M., Langni, O., Nast, M., Stai, F.,
Staude, U. (2000). Klimaschutz durch Nutzung erneuerbarer Energien. In:
Klimaschutz durch Nutzung erneuerbarer Energien Berichte/Umweltbundesamt.
Erich Schmidt Verlag GmbH & Co., Berlin (ISBN 3-503-05888-5).
Office of the Biomass Program, EERE/DOE (2005). Multi-Year Program Plan 2007-2012.
U.S. Department of Energy.

159

Bibliografa/References
Oliet, M. (1998). Estudio sobre la deslignificacin de Eucalyptus globulus con etanol/agua
como medio de coccin. Tesis Doctoral. Universidad Complutense de Madrid.
Oliet, M., Gilarranz, M. A., Domnguez, J. C., Alonso, M. V., Rodrguez, F. (2005). Ethanolbased pulping from Cynara cardunculus L. Journal of Chemical Technology and
Biotechnology, 80, 7, 746-753.
Otero, M. (1988). Composio quimica dos materiais lignocelulsicos. Celulose e Paper.
Vol. I: Tecnologa de Fabricacin de Pasta Celulsica, 45-106. Instituto de Pesquisas
Tecnolgicas del Estado de Sao Paulo, S.A.
Overend, R. P., Chornet, E., Gascoigne, J. A. (1987). Fractionation of lignocellulosics by
steam-aqueous pretreatments [and discussion]. Philosophical Transactions of the
Royal Society London A, 1561, 321, 523-536.
Ozin, G. A., Arsenault, A. C., Cademartiri, L. (2008). Nanochemistry: a chemical approach
to nanomaterials. RSC Publishing (ISBN: 978-1847558954).
Pkk, M., Ankerfors, M., Kosonen, H., Nyknen, A., Ahola, S., Osterberg, M.,
Ruokolainen, J., Laine, J., Larsson, P. T., Ikkala, O., Lindstrm, T. (2007). Enzymatic
hydrolysis combined with mechanical shearing and high-pressure homogenization for
nanoscale cellulose fibrils and strong gels. Biomacromolecules, 8, 6, 19341941.
Paice, M. G., Bernier, R. Jr., Jurasek, L. (1988). Viscosity-enhancing bleaching of hardwood
kraft pulp with xylanase from a cloned gene. Biotechnology and Bioengineering, 32, 2,
235-239.
Pan, G. X., Spencer, L., Leary, G. J. (2000a). A comparative study on reactions of hydrogen
peroxide and peracetic acid with lignin chromophores. Part 1. The reaction of
coniferaldehyde model compounds. Holzforschung, 54, 2, 144-152.
Pan, G. X., Spencer, L., Leary, G. J. (2000b). A comparative study on reactions of hydrogen
peroxide and peracetic acid with lignin chromophores. Part 2. The reaction of
stilbene-type model compounds. Holzforschung, 54, 2, 153-158.
Papatheophanus, M. G., Koullas, D. P., Koukios, E. G. (1995). Alkaline pulping of
prehydrolysed wheat straw in aqueous-organic solvent systems at low temperatures.
Cellulose Chemistry and Technology, 29, 1, 29-40.
Paraj, J. C., Domnguez, H., Moure, A., Daz-Reinoso, B., Conde, E., Soto, M. L., Conde,
M. J., Gonzlez-Lpez, N. (2008). Recovery of phenolic antioxidants released during
160

Bibliografa/References
hydrolytic treatments of agricultural and forest residues. Electronic Journal of
Environmental, Agricultural and Food Chemistry, 7, 8, 3243-3249.
Paszner, L., Cho, H. J. (1989). Organosolv pulping: acidic catalysis options and their effect
on fiber quality and delignification. TAPPI Journal, 72, 2, 135-142.
Pedneault, C., Pellerin, C., Robert, S. (1999). Amine boranes as new reductive bleaching
chemicals on softwood pulp single stage and multistage processes. TAPPI Journal, 82,
2, 110-114.
Pellerin, C., Pedneault, C., Robert, S. (2000). Bleaching chemistry of amine boranes with
lignin model compounds. Journal of Pulp and Paper Science, 26, 12, 436-440.
Prez, J. A., Ballesteros, I., Ballesteros, M., Sez, F., Negro, M. J., Manzanares, P. (2008).
Optimizing liquid hot water pretreatment conditions to enhance sugar recovery from
wheat straw for fuel-ethanol production. Fuel, 87, 17-18, 3640-3647.
Petlicki, J., van de Ven, T. G. M., Palusova, D. (2005). Physicochemical aspects of catalytic
decomposition of hydrogen peroxide by manganese compounds. Industrial and
Engineering Chemistry Research, 44, 7, 2002-2010.
Pettersen, R. C. (1984). The chemical composition of wood. American Chemical Society.
Washington, D.C., 57-126.
Pflaum, H., Rettweiler, M. (2005). Innovations- und Netzwerkmanagement: Das Netzwerk
PLANET. Chemie Ingenieur Technik, 77, 15351541 (doi: 10.1002/cite.200500124).
Phillips, F. H. (1998). Evaluacin y elaboracin de los recursos de fibras para la fabricacin
de pasta. Depsito de documentos de la FAO.
Pioch, D., Vaitilingom, G. (2005). Palm oil and derivatives: fuels or potential fuels?
Olagineux, Corps Gras, Lipides, 12, 2, 161-169.
Poppius-Levlin, K., Mustonen, R., Huovila, T., Sundquist, J. (1991). MILOX pulping with
acetic acid/peroxyacetic acid. Paperi Ja Puu-Paper Timber, 73, 2, 154-158.
Prasakis, J., Sain, M., Daneault, C. (1996). Metal management improves peroxide bleaching
of TMP. TAPPI Journal, 79, 10, 161-166.
Qiu, Y. G., Cai, L. S., Fang, G. R. (2005). Pressured hydrogen peroxide bleaching of sodaAQ bagasse pulp. Pulp and Paper, 24, 6, 5-9.

161

Bibliografa/References
Qiu, Z., Ni, Y., Yang, S. (2003). Using DTPA to decrease manganese-induced peroxide
decomposition. Journal of Wood Chemistry and Technology, 23, 1, 1-11.
Quader, M. M. A., Lnnberg, B. (2004). Pulp and chemical cellulose obtained with oil palm
empty fruit bunches by a modified IDE process. JIRCAS Working Report, 39, 47-53.
Ragauskas, A. J., Poll, K. M., Cesternino, A. J. (1994). Effects of xylanase pretreatment
procedures on nonchlorine bleaching. Enzyme and Microbial Technology, 16, 6, 492495.
Requejo, A., Peleteiro, S., Rodrguez, A., Garrote, G., Paraj, J. C. (2011). Secondgeneration bioethanol from residual woody biomass. Energy Fuels, 25, 4803-4810.
Revol, J. F., Bradford, H., Giasson, J., Marchessault, R. H., Gray, D. G. (1992). Helicoidal
self-ordering of cellulose microfibrils in aqueous suspension. International Journal of
Biological Macromolecules, 14, 3, 170-172.
Revol, J. F., Godbout, L., Dong, X. M., Gray, D. G., Chanzy, H., Maret, G. (1994). Chiral
nematic suspensions of cellulose crystallites; phase separation and magnetic field
orientation. Liquid Crystals, 16, 1, 127-134.
Rhim, J. W. (2007). Potential use of biopolymer-based nanocomposite films in food
packaging applications. Food Science and Biotechnology, 16, 5, 691-709.
Rhim, J. W., Ng, P. K. (2007). Natural biopolymer-based nanocomposite films for
packaging applications. Critical Reviews in Food Science and Nutrition, 47, 4, 411-433.
Rijkens, B. A. (1984). Hydrolysis processes for lignocellulosic material. CECD Workshop
Cellulose Programme. Brunscheweig.
Rivas, B., Domnguez, J. M., Domnguez, H., Paraj, J. C. (2002). Bioconversion of
posthydrolysed autohydrolysis liquors: an alternative for xylitol production from corn
cobs. Enzyme and Microbial Technology, 31, 4, 431-438.
Rodrguez, A. (2002). Obtencin de pastas celulsicas de podas de olivo mediante mtodos
menos contaminantes. Tesis Doctoral. Universidad de Crdoba.
Rodrguez, A., Jimnez, L. (2008). Pulping with organic solvents other tan alcohols.
Afinidad, 535, 188-196.
Rodrguez, A., Moral, A., Serrano, L., Jimnez, L. (2009). Residuos de la industria del aceite
de palma para fabricar papel. Ingeniera Qumica, 468, 58-65.

162

Bibliografa/References
Rodrguez, A., Serrano, L., Moral, A., Prez, A., Jimnez, L. (2008). Use of high-boiling
point organic solvents for pulping oil palm empty fruit bunches. Bioresource
Technology, 99, 1743-1749.
Rojas, E. (1997). La fibra diettica. Estudio bioqumico y aplicaciones clnicas. Alimentacin,
Nutricin y Salud, 4, 3, 59-72.
Roliadi, H., Pasaribu, R. A. (2004). The possible utilization of empty oil-palm bunches as a
substitute of pulp wood for the manufacture of corrugating-medium paperboard using
semi-chemical soda-anthraquinone pulping. APPITA annual Conference and
Exhibition, 58th, 2, 335-341.

Roman, A. (2011). Contribucin al estudio de procesos para la obtencin de bioetanol de


segunda generacin. Tesis Doctoral. Universidad de Vigo.
Rubio, M., Tortosa, J. F., Gmez, D., Miana, A., Soler, A. (1994). El fraccionamiento de
los

materiales

vegetales

mediante

procesos

combinados

de

autohidrlisis-

organosolvlisis. Afinidad, 51, 95-102.


Ruggiero, R., Machado, A. E. H., Prez, D. S., Grelier, S., Nourmamode, A., Castellan, A.
(1998). Bleached chemical pulp from Eucalyptus grandis wood produced by
peroxyformic acid pulping and photochemical bleaching. Holzforschung, 52, 3, 325332.
Sarkanen, K. V. (1991). Chemistry of solvent pulping. TAPPI Journal, 73, 10, 215-219.
Saska, M., Ozer, E. (1995). Aqueous extraction of sugarcane bagasse hemicellulose and
production of xylose syrup. Biotechnology and Bioengineering, 45, 6, 517-523.
Schroeter, M. C. (1991). Possible lignin reactions in the organocell pulping process. TAPPI
Journal, 74, 10, 197-200.
Seisto, A., Laine, J. (1999). Surface chemical composition of birch Milox pulp. Paperi Ja
Puu-Paper and Timber, 81, 1, 53-57.
Seisto, A., Poppius-Levlin, K. (1996). Peroxyformic acid pulping of non wood plants by the
milox method. Part I: Pulping and bleaching. TAPPI Journal, 80, 9, 215-221.
Senior, D. J., Hamilton, J. (1992). Use of xylanases to decrease the formation of AOX in
kraft pulp bleaching. Journal of Pulp and Paper Science, 18, 5, 165-168.

163

Bibliografa/References
Shah, A. K., Cooper, D., Adolphson, R., Eriksson, K. E. L. (2000). Xylanase treatment of
oxygen-bleached hardwood kraft pulp at high temperature and alkaline pH levels gives
substantial savings in bleaching chemicals. Journal of Pulp and Paper Science, 26, 1, 811.
Siau, J. F. (1984). Transport processes in wood. Springer Series in Wood Science. SpringerVerlag. New York (ISBN: 978-3642692154).
Simula, M. (2002). Fibre supply: The Achilles hell of the global pulp and paper industry.
Paperi Ja Puu-Paper and Timber, 84, 1, 16-19.
Sjstrm, E. (1993). Wood chemistry - fundamentals and applications. Academic Press, San
Diego, California.
Spence, K. L., (2011). Processing and properties of microfibrillated cellulose. Doctoral
Dissertation. North Carolina State University.
Spence, K. L., Venditti, R. A., Rojas, O. J., Habibi, Y., Pawlak, J. J. (2010). The effect of
chemical composition on microfibrillar cellulose films from wood pulps: water
interactions and physical properties for packaging applications. Cellulose 17, 835-848.
Suleman, Y. H., Yusoff, M. N. M. (1997). Ethanol-water pulping of oil palm fibers using
hydrochloric acid as a catalyst and test paper characterization. Mokuzai Gakkaishi, 43,
12, 1016-1021.
Sun, R. C., Lawther, J. M., Banks, W. B. (1997). Physico-chemical characterization of
organosolv lignins from wheat straw. Cellulose Chemistry and Technology, 31, 3-4,
199-212.
Suwa, Y., Koga, K., Fujikawa, S., Okazaki, M., Irie, T., Nakada, T. (1999). Bifidobacterium
bifidum proliferation promoting composition containing xylooligosaccharide. Patente
de Estados Unidos n US5939309.
Svensson, A., Nicklasson, E., Harrah, T., Panilaitis, B., Kaplan, D. L., Brittberg, M.,
Gatenholm, P. (2005). Bacterial cellulose as a potential scaffold for tissue engineering
of cartilage. Biomaterials, 26, 4, 419-431.
Tanaka, R., Wan Rosli, W. D., Magara, K., Ikeda, T., Hosoya, S. (2004). Chlorine-free
bleaching of kraft pulp from oil palm empty fruit bunches. JARQ (Japan International
Research Center for Agricultural Sciences and University Sains Malaysia), 38, 4, 275279.
164

Bibliografa/References
Taniguchi, T., Okamura, K. (1998). New films produced from microfibrillated natural
fibres. Polymer International, 47, 3, 291-294.
Taylor, J. D., Yu, E. K. C. (1995). Continuous steam explosion. Chemtech, 25, 2, 38-41.
Tortosa, J. F., Rubio, M., Soler, A. (1995). Autohidrlisis de tallo de maz en suspensin
acuosa. Afinidad, 52, 305-311.
Toshio, I., Noriyoshi, I., Toshiaki, K., Toshiyuki, N., Kunimasa, K. (1990). Production of
xylobiose. Patente de Japn n JP2119790.
Turbak, A. F., Snyder, F. W., Sandberg, K. R. (1983). Microfibrillated cellulose, a new
cellulose product: properties, uses, and commercial potential. Journal of Applied
Polymer Science: Applied Polymer Symposium, 37, 815-827.
United States Department of Agriculture (FAS) (2010). Lower sunseed and rapeseed crops
have little impact on global vegetable oil market. Oilseeds: world markets and trade.
Uraki, Y., Sano, Y. (1999). Polyhydric alcohol pulping at atmospheric pressure: an effective
method for organosolv pulping of softwoods. Holzforschung, 53, 4, 411-415.
Varshney, A. K., Patel, D. (1988). Biomass delignification: organosolv approach. Journal of
Scientific and Industrial Research, 47, 6, 315-319.
Vzquez, M. J., Alonso, J. L., Domnguez, H., Paraj, J. C. (2006). Enhancing the potential
of oligosaccharides from corncob autohydrolysis as prebiotic food ingredients.
Industrial Crops and Products, 24, 2, 152-159.
Vidal, T., Colom, J. F. (1986). Extraccin alcalina oxidativa. Utilizacin de oxgeno.
Ingeniera qumica, 18, 205, 57-67.
Viikari, L., Kantelinen, A., Sundquist, J., Linko, M. (1994). Xylanases in bleaching: from an
idea to the industry. FEMS Microbiology Reviews, 13, 2-3, 335-350.
Viikari, L., Ranua, M., Kantelinen, A., Sundquist, J., Linko, M. (1986). Bleaching with
enzymes. Proceedings of Biotechnology in the Pulp and Paper Industry, 67-69.
Wallis, A. F. A. (1978). Wood pulping with mono, di and triethanolamine. APPITA
Journal, 31, 6, 443-448.
Wallis, A. F. A. (1980). Wood pulping with monoethanolamine in pressure vessels.
APPITA Journal, 33, 5, 351-355.

165

Bibliografa/References
Wallis, A. F. A., Wearne, R. H. (1985). Fractionation of the polymeric components of
hardwoods by autohydrolysis-explosion-extraction. APPITA Journal, 38, 6, 432-437.
Weil, J., Brewer, M., Hendrickson, R., Sarikaya, A., Ladisch, M. R. (1998). Continuous pH
monitoring during pretreatment of yellow poplar wood sawdust by pressure cooking in
water. Applied Biochemistry and Biotechnology, 70-72, 1, 99-111.
Wekesa, M., Ni, Y. (2003a). Further understanding of the chemistry of manganese-induced
peroxide decomposition. Canadian Journal of Chemical Engineering, 81, 5, 968-972.
Wekesa, M., Ni, Y. (2003b). Stabilization of peroxide systems by silicate and calcium
carbonate and its application to bleaching of recycled fibres. Pulp and Paper Canada,
104, 12, 85-87.
Werpy, T., Petersen, G. (2004). Top value added chemicals from biomass. Vol. I. results of
screening for potential candidates from sugars and synthesis gas. Eds. The pacific
northwest national laboratory and the national renewable energy laboratory, U.S.
N.DOE/GO-102004-1992.
Whitesides, G. M., (2005). Nanoscience, nanotechnology and chemistry. Nanoscience and
Chemistry, 1, 2, 172-179.
Wong, K. K. Y., Ed, D. J., Saddler, J. N., Allison, R. W. (1997). Mechanisms of xylanase
aided bleaching of kraft pulp. Part 2: Target substrates. APPITA Journal, 50, 6, 509518.
Xu, E. C. (2000a). H2O2 bleaching of mechanical pulps. Part 1: Kinetics and mechanism.
Journal of Pulp and Paper Science, 26, 10, 367-376.

Xu, E. C. (2000b). H2O2 bleaching of mechanical pulps. Part 2: pH and temperature.


Journal of Pulp and Paper Science, 26, 11, 407-413.

Yano, H., Nakahara, S. (2002). High strength material using cellulose microfibrils. Patente
de Estados Unidos n 7378149.
Zadeh, L. A. (1965). Fuzzy sets. Inf. Control, 8, 338353.
Zhang, X. Z., Kang, G. J., Ni, Y., Van Heiningen, A. R. P., Mislankar, A., Darabie, A.,
Reeve, D. W. (1998). Initial delignification and cellulose degradation of conventional
and ethanol-assisted ozonation. Journal of Wood Chemistry and Technology, 18, 2,
129-157.

166

Bibliografa/References
Zhao, X., Cheng, K., Liu, D. (2009). Organosolv pretreatment of lignocellulosic biomass for
enzymatic hydrolysis. Applied Microbiology and Biotechnology, 82, 5, 815-827.
Zuluaga, R., Putaux, J. L., Restrepo, A., Mondragn, I., Gan, P. (2007). Cellulose
microfibrils from banana farming residues: isolation and characterization. Cellulose,
14, 6, 585-592.

167

Bibliografa/References

168

Publicaciones Cientficas/Scientific Publications

Informacin y criterios de calidad de las Publicaciones


La Normativa Reguladora de los Estudios de Doctorado especifica en el
Captulo II en su Art. 24, que las tesis doctorales realizadas como compendio
de publicaciones deben estar al menos constituidas por un mnimo de tres
artculos publicados o aceptados en revistas incluidas en los tres primeros
cuartiles de la relacin de revistas del mbito de la especialidad y referenciadas
en la ltima relacin publicada por el Journal Citation Reports (SCI y/o
SSCI). Es por esta razn, que se incluye este apartado relativo a la calidad de
las publicaciones cientficas indexadas, proporcionando el ttulo y los datos del
manuscrito, los autores y los datos del Journal Citation Reports del ao de
publicacin del manuscrito (en ciertos casos, del ltimo ao disponible).

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 1
INFLUENCE OF TEMPERATURE, TIME, LIQUID/SOLID RATIO AND
SULPHURIC ACID CONCENTRATION ON THE HYDROLYSIS OF PALM
EMPTY FRUIT BUNCHES
Ana Ferrer, Ana Requejo, Alejandro Rodrguez, Luis Jimnez
Bioresource Technology, 2012, 10.1016/j.biortech.2012.10.081
ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de Agricultural
Engineering segn el Journal Citation Reports de 2011)

173

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)

174

Publicaciones Cientficas/Scientific Publications


Bioresource Technology (10.1016/j.biortech.2012.10.081)

INFLUENCE OF TEMPERATURE, TIME, LIQUID/SOLID RATIO AND SULPHURIC


ACID CONCENTRATION ON THE HYDROLYSIS OF PALM EMPTY FRUIT BUNCHES
Ana Ferrer, Ana Requejo, Alejandro Rodrguez, Luis Jimnez

Chemical Engineering Department, Campus of Rabanales, Building Marie Curie (C-3), University of
Crdoba, 14071 Crdoba, Spain
Received 17 April 2012; Received in revised form 15 October 2012; Accepted 16 October 2012

ABSTRACT
The influence of temperature (150-190 C), time (0-20 min), liquid/solid-ratio (6-8)
and sulphuric acid concentration (0.1-0.5 %), on the hydrolysis of palm empty fruit bunches
(EFB) was studied and the liquid and solid fractions were analyzed. Polynomial models were
found to reproduce the experimental results with errors less than 15 % in most of the cases
(except for xylose concentration).
Operating conditions of 190 C for 15 min at a liquid/solid ratio of 6 and a sulphuric
acid concentration of 0.1 % resulted in the production of 3.12, 4.0, 2.35 and 2.28 g/L of
glucose, xylose, arabinose and acetic acid, respectively, starting with 1000 g of EFB. The
yield was 67.96 %.
Soda-anthraquinone, ethanol and ethanolamine pulping processes of the solid fraction
provided pulps with brightness values of 63.24 %, 28.78 %, 48.76 %, respectively, but with
poor resistance properties (6.57-8.54 Nm/g for the tensile index, 0.38-0.44 KN/g for the
burst index and 0.96-1.02 mNm2/g for the tear index). Therefore it is advisable to use the
pulps for specialty papers or for bioethanol production.

KEYWORDS: Empty Fruit Bunches; Hydrolysis; Hemicelluloses; Cellulose derivatives;


Pulp; Bioethanol.

175

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


1. INTRODUCTION
Non-wood plants constitute an effective alternative to wood for production of cellulose
derivatives (pulp for paper, dissolving pulp, sugars by hydrolysis, bioethanol by hydrolysis of
cellulose and later fermentation) in countries with limited forest resources. A promising nonwood raw material is Empty Fruit Bunches (EFB), a residue of the palm oil industry (Ferrer
et al., 2011; Jimnez et al., 2009a,b; Rodrguez et al., 2008).
Conventional methods for producing cellulose also remove hemicelluloses and lignin.
Such components can be recovered by using various procedures commonly associated with
the concept of bio-refinery (Garrote et al., 2007a; Kadam et al., 2008; Pan et al., 2005;
Requejo et al., 2012). The bio-refinery or fractionation of lignocellulosic materials as agrifood residues is especially interesting since it endows an added value and provides an
environmental benefit (Garrote et al., 2007a; Kadam et al., 2008; Requejo et al., 2012;
Sakaki et al., 2006; Snchez et al., 2011; Sasaki et al., 2003).
One fractionation procedure involves the hydrolysis of lignocellulosic materials and
pulping of the resulting solid fraction. Hydrolysis with water at a high temperature produces
an

aqueous

fraction

essentially

containing

hemicellulosic

sugars

consisting

of

monosaccharides (xylose, glucose, arabinose) and xylo-oligosaccharides (Alfaro et al., 2009;


Caparrs et al., 2007, 2008a,b; Yez et al., 2009), which can be further hydrolyzed and
fermented to obtain various products (Boussarsar et al., 2009; Dogaris et al., 2009; Sakaki et
al., 2006; Vzquez et al., 2007). The solid fraction composed largely of lignin and cellulose is
potentially amenable to pulping for paper production (Caparrs et al., 2007, 2008a,b;
Requejo et al., 2012; Snchez et al., 2011), or for other uses such as the production of
bioethanol by hydrolysis of the pulp and subsequent fermentation of glucose (Qing et al.,
2011; Requejo et al., 2011, 2012; Zhang and Lynd, 2010).
Hydrolysis can be conducted over wide ranges of operating conditions (Garrote et al.,
2007b; Sundqvist et al., 2006; Vegas et al., 2008). In a weakly acidic medium, ether bonds in
lignin are broken above 160-180 C and cellulose starts to depolymerize above 210 C.
Process time can also vary from a few seconds to several hours. The liquid/solid ratio can be
set at values from 2 to 40 g water/g material, but usually falls within the range of 8-10 g/g.
The pH has a strong influence on cellulose degradation. Particle size is also influential and
usually ranges from 0.5 to 10 mm in laboratory tests. Pulping processes applied to non-wood
materials, mainly used soda, soda-anthraquinone and various organic solvents (Caparrs et

176

Publicaciones Cientficas/Scientific Publications


al., 2008b; Jimnez et al., 2007, 2009a; Ligero et al., 2008; Lpez et al., 2006; Rodrguez and
Jimnez, 2008; Ziaie-Shirkolaee et al., 2008).
In the present study, the influence of operating variables on the composition of the
liquid and solid fractions was studied. The solid fraction was obtained by sodaanthraquinone, ethanol and ethanolamine pulping processes and the properties of the pulp
and paper sheets obtained were compared with those of pulp and paper sheets obtained
from the raw material without prior hydrothermal treatment.

2. MATERIALS AND METHODS


2.1. Raw material
EFB from Elaeis guineensis was supplied by Straw Pulping Engineering S.L.
(Zaragoza, Spain) and its holocellulose, -cellulose, lignin, ethanol-benzene extractive and
ash contents were determined according to TAPPI standards T-9 m-54, T-203 OS-61, T222, T-204 and T-211, respectively. Fiber length was determined biometrically, under a
Visopan projection microscope, after microcooking the raw material with 10 % soda at 80 C
for 1 h and subsequent staining with 1 % safranin.

2.2 Hydrolysis
EFB and water required to obtain an appropriate liquid/solid ratio were placed in a 15L batch reactor that was heated by an outer jacket containing electrical wires. The reactor
contents were stirred by rotating the reaction vessel via a motor connected through a rotary
axle to a control unit that also measured and controlled pressure and temperature. Once the
mixture was heated to the selected temperature for the pre-set time, the reactor was
depressurized, and the liquid and solid fractions were separated. The temperature profiles
for the heating and cooling stages corresponded to standard operational conditions (Requejo
et al., 2012).

2.3 Characterization of the fractions of the hydrolysis process


Glucose, xylose, arabinose and acetic acid contents of the liquid fraction were
determined as follows.

177

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


An amount of 10-20 g of liquid fraction was placed in a 100-mL ISO bottle and
supplied with sulphuric acid to a 4 %. Then, the bottle was autoclaved at 121 C and 2 atm
for 20 min, cooled to room temperature with water, and its contents were analyzed by HPLC
(Refractive Index Detector, Aminex HPX-87H column eluted with 0.01 M H2SO4, 0.6

mL/min flow rate). The solid yield from the solid fraction was determined gravimetrically.

2.4 Pulping process


The raw material or hydrolyzed raw material was cooked with soda-anthraquinone,
ethanol or ethanolamine in a 15-L batch reactor (Table 1). The cooked material was
unloaded into a washer to remove residual cooking liquor and fiberized in a disintegrator at
1200 rpm for 30 min, after which the pulp was beaten in a Sprout-Bauer refiner and passed
through a screen of 0.16 mm mesh size to remove uncooked particles.
Table 1. Pulping conditions of EFB and solid fraction from EFB hydrolyzed.
Raw material and
pulping

Denomination

Temperature,
C

Time,
min

Reactive, %

Liquid/solid
ratio, w/w

EFB and soda-AQ

EFB-SA

170

30

15 % soda, 1 % AQ

10

EFB hydrolyzed
and soda-AQ

EFB-H-SA

170

30

15 % soda, 1 % AQ

10

EFB and ethanol

EFB-E

170

30

60 % ethanol

10

EFB-H-E

170

30

60 % ethanol

10

EFB-EA

170

30

60 % ethanolamine

10

EFB-H-EA

170

30

60 % ethanolamine

10

EFB hydrolyzed
and ethanol
EFB and
ethanolamine
EFB hydrolyzed
and ethanolamine

Hydrothermal treatment conditions: 190 C; 15 min; liquid/solid ratio of 6; sulphuric acid


concentration of 0.1 %.

2.5 Characterization of pulp and paper sheets


Pulping yield was determined gravimetrically and viscosity and kappa number were
measured in accordance with UNE 57-039 and UNE 57-034, respectively. Paper sheets were
obtained by using an Enjo-F39.71 sheet former and characterized in terms of tensile index,
burst index, tear index and brightness in accordance with UNE standards (57-054, 57-058,
57-033 and 57-062, respectively).

178

Publicaciones Cientficas/Scientific Publications


2.6 Experimental design
The experimental design consisted of a series of points (tests) around a central
composition point (central test) that were used to estimate the quadratic terms of a
polynomial model (Montgomery, 1991). The total number of tests required for the three
independent variables studied in the hydrolysis process [viz. temperature (T), time (t) and
liquid/solid ratio (R) or sulphuric acid concentration (S)] was 15.
The values of the independent variables were normalized by using the following
equation to facilitate direct comparison of coefficients and expose the individual effects of
the independent variables on each dependent variable:
Xn =

)
2(X-X

X max -X min

where Xn is the normalized value of T, t, R or S; X is the absolute experimental value of the

variable concerned;
X is the mean of the extreme values of X; and Xmax and Xmin are its

maximum and minimum value, respectively.

Experimental data were fitted to a second-order polynomial, which relates each


dependent variable (solid yield, glucose, xylose, arabinose and acetic acid) with the
operational variables (temperature, time and liquid/solid ratio or sulphuric acid
concentration).

3. RESULTS AND DISCUSSION


3.1 Raw material
The chemical composition of EFB used is 67.0 % holocellulose; 41.9 % -cellulose;
24.5 % lignin; 1.2 % ethanol-benzene extractives and 3.2 % ash, and the average fiber length
of EFB is 0.53 mm.

3.2 Hydrolysis without a catalyst


The operational values, as well as the respective normalized values are shown in Table
2. The choice of the values of operating variables was done considering the values used for:
paulownia (Caparrs et al., 2008a), corn stalks (Tortosa et al., 1995), wheat straw (Kubikova,
1996), vine shoots (Jimnez et al., 2006), arundo donax (Caparrs et al., 2007), sunflower
stalks (Caparrs et al., 2008b), legumes (Alfaro et al., 2009), Sebatia grandiflora (Yez et al.,
179

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


2009), bagasse from sugar cane (Boussarsar, 2009), rice straw (Rodrguez et al., 2009),

Hesperaloe funifera (Snchez et al., 2011) and olive prunings (Requejo et al., 2012). Table 2
also gives the results of the tests as well as the averages of three determinations of dependent
variables related to the liquid and solid fractions obtained in each test.
Table 2. Values of operational variables, and experimental values of composition of the liquid and
solid fractions obtained by hydrolysis of EFB without a catalyst.
T,
C
170

Operational variables
t,
R,
XT
Xt
min w/w
10
7
0
0

190

20

150

20

190

20

150

Dependent variables
Glucose, Xylose, Arabinose,
g/L
g/L
g/L
0.89
0.82
1.08

Solid
yield, %
78.21

68.21

1.79

2.48

1.74

2.55

-1

81.03

0.80

0.44

1.00

0.96

-1

68.75

2.14

3.99

2.62

3.31

20

-1

-1

84.02

1.34

1.23

1.30

1.52

190

-1

72.65

1.06

0.95

1.23

1.41

150

-1

-1

84.45

0.25

0.18

0.30

0.56

190

-1

-1

78.10

1.33

1.40

1.28

1.57

150

-1

-1

-1

86.00

0.60

0.20

0.74

0.67

170

20

77.10

1.35

2.11

1.32

1.66

170

-1

82.53

0.70

0.36

0.56

0.99

170

10

76.21

0.83

0.82

0.91

1.28

170

10

-1

79.40

0.97

0.98

1.12

1.43

190

10

73.43

1.55

2.31

1.56

2.05

150

10

-1

83.20

0.66

0.43

0.80

0.98

XR

Acetic
acid, g/L
1.32

T, t and R = temperature, time and liquid/solid ratio; XT, Xt and XR = Normalized values of
temperature, time and liquid/solid ratio. The experimental values of dependent variables are the
average of three determinations, whose deviations with respect to the mean value are under 5 %.

A multiple regression analysis of the experimental results with selection of the


statistically significant terms in the polynomial model used (viz. those having a Snedecor Fvalue greater than 4 and a Student t-value greater than 2) provided the following equations:
YI = 78.22 1.05 XTXt 1.37 XR 2.46 Xt 5.76 XT

[1]

XY = 1.02 + 0.34 XT2 0.23 XtXR 0.29 XR + 0.35 XTXt + 0.72 Xt + 0.87 XT

[3]

AA = 1.34 0.13 XtXR + 0.22 XT2 0.17 XR + 0.20 XTXt + 0.48 Xt + 0.62 XT

[5]

GL = 0.95 + 0.20 XT2 0.17 XR + 0.35 Xt + 0.42 XT

AR = 1.00 + 0.26 XT2 0.19 XR + 0.39 Xt + 0.43 XT

180

[2]

[4]

Publicaciones Cientficas/Scientific Publications


where YI is the solid fraction yield, GL the glucose concentration, XY the xylose
concentration, AR the arabinose concentration, AA acetic acid concentration, and XT, Xt and

XR the normalized values of temperature, process time and liquid/solid ratio, respectively.

The polynomial models derived for the different dependent variables pertaining to the liquid
and solid fractions were similar to those previously obtained for eucalyptus, tagasaste,
sunflower stalks, vine shoots, Hesperaloe funifera and olive prunings (Caparrs et al.,
2008b; Garrote et al., 2007b; Jimnez et al., 2006; Lpez et al., 2004; Requejo et al., 2012;
Snchez et al., 2011). The estimated values provided by the polynomial model reproduced
the experimental results for the different dependent variables with errors less than 3, 15, 16
and 14 % for solid fraction yield, glucose concentration, arabinose concentration and acetic
acid concentration, respectively. Equation 3, corresponding to the xylose concentration,
reproduces the experimental results with an error of about 24 % in the majority of cases
(except experiments 7 and 13, Table 2). By using equations from [2] to [5] it is possible to
identify the values of the independent variables leading to maximum possible values of the
dependent variables in the liquid fraction of hydrolysis process. Maximum values of the
dependent variables (Table 3) were found using high temperatures, long processing times
and low liquid/solid ratios. In contrast, to ensure high solid yields low values of all
operational variables were required.
Table 3. Maximum values of the dependent variables and corresponding levels of the operating
variables required to obtain them in the hydrolysis process without a catalyst. The values of the
dependent variables, and their deviations (in brackets) from their maximum levels as obtained under
the stated conditions in each case (AD), are also given.
Maximun values
of dependent
variables

XT, Xt, XR to
obtain maximum
values of
dependent
variables

Yield, %

86.76

-1, -1, -1

Glucose, g/L

2.09

+1,+1,-1

Xylose, g/L

3.82

+1,+1,-1

Arabinose, g/L

2.27

+1,+1,-1

Acetic acid, g/L

2.35

+1,+1,-1

Dependent
variables

Values of dependent variables and


deviations from their maximum levels
Case A
73.83
14.90 %
1.74
16.75 %
2.52
34.03 %
1.88
17.18 %
2.35
25.63 %

Case B
72.08
16.92 %
1.92
8.13 %
3.17
17.02 %
2.08
8.37 %
2.76
12.66 %

Case C
71.20
17.93 %
2.00
4.31 %
3.50
8.38 %
2.17
4.41 %
2.96
25.97 %

Case D
73.65
15.11 %
1.72
17.70 %
2.76
27.75 %
1.85
18.50 %
2.48
21.52 %

Case A: 190 C, 10 min, 6 liquid/solid ratio. Case B: 190 C, 15 min, 6 liquid/solid ratio. Case C: 190
C, 17.5 min, 6 liquid/solid ratio. Case D: 185 C, 15 min, 6 liquid/solid ratio.

181

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


By analyzing eq [1] and Fig. 1 and 2 (graphical representation of the equation), the
temperature was identified as the operational variable that more strongly influenced solid
fraction yield, and liquid/solid ratio had the least influence. Similar conclusions are to the
other dependent variables.
To find the best possible compromise between efficient use of the raw material (viz. a
good yield for the solid fraction) and high values for the composition of the liquid fraction,
eqs [1] to [5] were used to simulate cases A-D in Table 3.
The aim was to obtain near-optimal values for the composition of the liquid fraction
and for the solid fraction yield while saving energy, reagents and facilities as far as possible
(i.e. by using a temperature, time and liquid/solid ratio below their respective maximum).
As can be seen from Table 3, obtaining high glucose, xylose, arabinose and acetic acid
concentrations required using a high temperature in combination with a medium treatment
time and low liquid/solid ratio; this will provide a saving in facilities as it avoids using the
longest possible time and the highest liquid/solid ratio.
In contrast, obtaining a good solid fraction yield required using a low temperature. For
cases B and D, solid fraction yield and sugar contents did not deviate much from their
maximum values.

Figure 1. Variation of the yield with the temperature and time.

182

Publicaciones Cientficas/Scientific Publications

Figure 2. Variation of the yield with the time and liquid/solid ratio.

3.3 Hydrolysis with an acid catalyst


Table 4 shows the operating conditions of the 15 experiments in the experimental
factorial design applied to EFB hydrolysis using sulphuric acid as a catalyst, and with a
liquid/solid ratio of 6.
Results obtained for the characterization of the liquid fraction and solid yield are also
shown in Table 4. As can be seen in the EFB hydrolysis without a catalyst, the polynomial
model was not adequate to reproduce the experimental values of the xylose concentration.
For the remaining dependent variables the equations are:
YIs = 77.78 0.72 XS 1.29 XTXt 2.21 XT2 2.48 Xt 6.45 XT

[6]

ARs = 1.41 + 0.09 XTXt + 0.16 XS + 0.31 Xt + 0.90 XT

[8]

GLs = 2.67 0.06 XTXt + 0.05 XS + 0.11 XS2 0.12 X2 0.16 XT2 + 0.18 Xt + 0.52 XT

[7]

AAS = 1.40 + 0.15 XS + 0.34 Xt + 0.86 XT

[9]

where YIs is the solid fraction yield, GLs is the glucose concentration, ARs the arabinose
concentration, AAs the acetic acid concentration, and XT, Xt and XS the normalized values of

the temperature, process time and sulphuric acid concentration, respectively. These
equations reproduced the experimental values for the solid yield and glucose concentration
with errors less than 2 % and 4 %, respectively, and the experimental values of arabinose and
acetic acid concentrations with errors less than 15 %, in the 87 % of the cases.

183

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


Table 4. Values of operational variables and experimental values of composition of the liquid and
solid fractions obtained by hydrolysis of EFB with a catalyst.
Operational variables

Dependent variables
Glucose, Xylose, Arabinose,
g/L
g/L
g/L
2.69
2.11
1.49

T, C

t, min

S, %

XT

Xt

XS

170

10

0.3

Solid
yield, %
79.12

Acetic
acid, g/L
1.52

190

20

0.5

64.30

3.11

5.95

2.96

2.64

150

20

0.5

-1

80.34

2.32

0.38

0.77

0.89

190

20

0.1

-1

67.10

3.13

4.66

2.46

2.63

150

20

0.1

-1

-1

81.00

2.08

0.44

0.64

0.65

190

0.5

-1

72.40

2.98

3.43

2.00

2.11

150

0.5

-1

-1

83.00

1.82

0.18

0.40

0.34

190

0.1

-1

-1

74.00

2.84

3.22

1.84

1.65

150

0.1

-1

-1

-1

83.00

1.72

0.18

0.16

0.18

170

20

0.3

74.20

2.78

2.88

1.78

1.92

170

0.3

-1

79.30

2.31

0.48

1.08

1.10

170

10

0.5

77.10

2.80

2.26

1.78

1.69

170

10

0.1

-1

79.20

2.74

0.98

1.24

1.08

190

10

0.3

67.80

3.02

3.58

2.17

2.07

150

10

0.3

-1

82.76

1.98

0.24

0.43

0.48

T, t and S = temperature, time and sulphuric acid concentration; XT, Xt and XS = Normalized values
of temperature, time and sulphuric acid concentration. The experimental values of dependent
variables are the average of three determinations, whose deviations with respect to the mean value are
under 5 %.

High values of glucose, xylose, arabinose and acetic acid concentrations can be
achieved with high values of the operating variables, whereas a high solid fraction yield
requires low values of operating variables. The most influential variable regarding the
composition of the liquid fraction and the yield of the solid fraction was temperature and the
least influential was the sulphuric acid concentration.
Table 5, as in the case of the hydrolysis without a catalyst, cases B and D are the most
suitable for obtaining high values of sugar concentrations (provide values for the dependent
variables that deviate relatively little from the maximum values, to lower values than the
tested maximum operating variables).
The concentration of xylose increases with increasing temperature, time and sulphuric
acid concentration (Table 4). The most influential variable for xylose concentration was the
temperature and the least influential one was the sulphuric acid concentration.

184

Publicaciones Cientficas/Scientific Publications


At 190 C for 15 min, a liquid/solid ratio of 6 and a sulphuric acid concentration of 0.1
% (Case B), the sum of the contents of glucose, xylose, arabinose and acetic acid was 11.75
g/L, representing a recovery of only 16.8 % of the hemicelluloses contained in the EFB. This
low utilization of hemicelluloses is due to the decomposition of these constituents, resulting
in volatile products, and to the production of oligosaccharides that were not analyzed. Thus,
when using sulphuric acid as a catalyst the maximum yields decreased from 86.76 % to 83.93
%, the maximum level of glucose increased from 2.09 % to 3.19 %, the maximum level of
xylose increased from 3.82 % to 4.66 %, and the maximum level of arabinose increased from
2.27 % to 2.87 %.
Table 5. Maximum values of the dependent variables and corresponding levels of the operating
variables required to obtain them in the hydrolysis process with a catalyst. The values of the
dependent variables, and their deviations (in brackets) from their maximum levels as obtained under
the stated conditions in each case (A-D), are also given.
Maximun values
of dependent
variables

XT, Xt, XS to
obtain maximum
values of
dependent
variables

Yield, %

83.93

Glucose, g/L

Dependent
variables

Values of dependent variables and


deviations from their maximum levels
Case A

Case B

Case C

Case D

-1, -1, -1

69.84
16.79 %

67.96
19.03 %

67.01
20.16 %

70.70
15.76 %

3.19

+1,+1,+1

3.09
3.13 %

3.12
2.19 %

3.11
2.51 %

3.07
3.76 %

Arabinose, g/L

2.87

+1,+1,+1

2.15
25.09 %

2.35
18.12 %

2.45
14.63 %

2.11
26.48 %

Acetic acid, g/L

2.75

+1,+1,+1

2.11
23.27 %

2.28
17.09 %

2.37
13.82 %

2.07
24.73 %

Case A: 190 C, 10 min, 6 liquid/solid ratio, 0.1 % sulphuric acid. Case B: 190 C, 15 min, 6
liquid/solid ratio, 0.1 % sulphuric acid. Case C: 190 C, 17.5 min, 6 liquid/solid ratio, 0.1 % sulphuric
acid. Case D: 185 C, 15 min, 6 liquid/solid ratio, 0.1 % sulphuric acid.

These results suggest that it is appropriate to use sulphuric acid, because the sugar
contents of the liquid fraction were higher than those obtained without a catalyst, but the
yield of the solid fraction decreased and thus the possible production of cellulosic pulp or
ethanol (Requejo et al., 2012).

3.4 Hydrolyzed EFB and EFB pulping processes


Table 1 shows the operational conditions of the pulping processes and Table 6
summarizes the properties of the pulps and paper sheets produced from the hydrolyzed
185

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


EFB, as well as those obtained by pulping of untreated EFB, using the soda-anthraquinone,
ethanol and ethanolamine pulping processes.
Soda-anthraquinone pulping process provided pulps and paper sheets with better
properties than ethanol and ethanolamine pulping processes; the worst process seems to be
the one using ethanol.
Kappa number and brightness of the pulps improved when pulping was preceded by
hydrolysis of the EFB, but viscosity of the pulps and the resistance properties of the paper
sheets decreased when pulping of the hydrolyzed EFB was performed.
Since the use of hydrolyzed EFB pulp is not suitable for the production of paper
sheets with suitable stretch properties, it may be more convenient to use this pulp for the
production of specialty papers or for the production of bioethanol by the hydrolysis of the
pulp for sugars and fermentation for bioethanol.
Table 6. Comparison of the properties of pulp and paper sheets obtained from untreated EFB
pulping processes and from the solid fraction of the hydrolysis treatment of EFB pulping processes.
Burst I.,
KN/g

869

Tensile I.,
Nm/g
11.98

0.68

Tear I.,
mNm2/g
1.41

19.78

413

10.18

0.58

1.91

23.02

32.53

778

11.40

0.56

1.65

EFB-H-SA

5.52

63.24

581

8.54

0.44

1.02

EFB-H-E

95.14

28.78

393

6.57

0.38

1.02

EFB-H-EA

14.80

48.76

625

7.75

0.42

0.96

Process

Kappa
number

Brightness,
%

Viscosity,
mL/g

EFB-SA

12.51

36.69

EFB-E

121.14

EFB-EA

The experimental values of the dependent variables are the average of five determinations, whose
deviations with respect to the mean value are under 5 to 6 %.

4. CONCLUSIONS
Polynomial equations to reproduce the experimental results of the solid fraction yield
and liquid fraction sugar contents from hydrolysis of EFB, with errors less than 15 % in most
of the cases, were obtained.
Pulps from soda-anthraquinone, ethanol and ethanolamine pulping processes of the
solid fraction of the hydrolysis treatment provided pulps with high brightness values, but
worse resistance properties compared to those obtained by pulping EFB. Therefore, it is
advisable to use the pulps for specialty papers or bioethanol production.

186

Publicaciones Cientficas/Scientific Publications


ACKNOWLEDGEMENTS
The authors are grateful to Ecopapel, S.L. (cija, Sevilla, Spain) for their support, and
to Spains DGICyT for funding this research within the framework of the Projects CTQ2010-19844-C02-01, TEP-6261 and TRA-2009-0064.

REFERENCES
Alfaro, A., Rivera, A., Prez, A., Yez, R., Garca, J.C., Lpez, F. 2009. Integral
valorization of two legumes by authohydrolysis and organosolv delignification. Bioresource
Technol., 100, 1, 440-445.
Boussarsar, H., Rog, B., Mathlouthi, M. 2009. Optimization of sugarcane bagasse
conversion by hydrothermal treatment for the recovery of xylose. Bioresource Technol.,
100, 24, 6537-6542.
Caparrs, S., Ariza, J., Lpez, F., Daz, M.J. 2007. Optimizing cellulosic paper
obtained from Arundo donax under hydrothermal treatment. J. Ind. Eng. Chem., 13, 3, 465473.
Caparrs, S., Daz, M.J., Ariza, J., Lpez, F., Jimnez, L. 2008a. New perspectives for

Paulownia fortunei L. Valorization of the autohydrolysis and pulping processes. Bioresource


Technol., 99, 4, 741-749.
Caparrs, S., Ariza, J., Lpez, F., Nacimiento, J.A., Garrote, G., Jimnez, L. 2008b.
Hydrothermal treatment and ethanol pulping of sunflower stalks. Bioresource Technol., 99,
5, 1368-1372.
Dogaris, I., Karapati, S., Mamma, D., Kalogeris, E., Kekos, D. 2009. Hydrothermal
processing and enzymatic hydrolysis of sorghum bagasse for fermentable carbohydrates
production. Bioresource Technol., 100, 24, 6543-6549.
Ferrer, A., Vega, A., Rodrguez, A., Ligero, P., Jimnez, L. 2011. Milox fractionation
of empty fruit bunches from Elaeis guineensis. Bioresource Technol., 102, 9755-9762.
Garrote, G., Falque, E., Domnguez, H., Paraj, J.C. 2007a. Autohydrolysis of
agricultural residues: study of reaction byproducts. Bioresource Technol., 98, 1951-1957.

187

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


Garrote, G., Kabel, M.A., Schols, H.A., Falque, E., Domnguez, H., Paraj, J.C.
2007b. Effects of Eucalyptus globulus wood autohydrolysis conditions on the reaction
products. J. Agr. Food Chem., 55, 22, 9006-9013.
Jimnez, L., Angulo, V., de la Torre, M.J., Ramos, E., Caparrs, S., Prez, A. 2006.
Simulation of the hydrothermal pretreatment of vine shoots for pulping. Afinidad, 63, 524,
277-285.
Jimnez, L., Prez, A., de la Torre, M.J., Moral, A., Serrano, L. 2007.
Characterization of vine shoots, cotton stalks, Leucaena leucocephala and Chamaecytisus

proliferus and of their ethyleneglycol pulps. Bioresource Technol., 98, 3487-3490.


Jimnez, L., Serrano, L., Rodrguez, A., Snchez, R. 2009a. Soda-AQ pulping of palm
oil empty fruit bunches and beating of the resulting pulp. Bioresource Technol., 100, 12621267.
Jimnez, L., Serrano, L., Rodrguez, A., Ferrer, A. 2009b. TCF bleaching of sodaanthraquinone and diethanolamine pulp from oil palm empty fruit bunches. Bioresource
Technol., 100, 1478-1481.
Kadam, K.L., Chin, C.Y., Brown, L.W. 2008. Flexible biorefinery for producing
fermentation sugars, lignin and pulp from corn stover. J. Ind. Microbiol. Biotech., 35, 5,
331-341.
Kubikova, J., Zemann, A., Krkosva, P., Bobleter, O. 1996. Hydrothermal
pretreatment of wheat straw for the production of pulp and paper. Tappi J., 79, 7, 163-169.
Ligero, P., Villaverde, J.J., Vega, A., Bao, M. 2008. Pulping cardoon (Cynara

cardunculus) with peroxiformic acid (MILOX) in one single stage. Bioresource Technol.,
99, 5687-5693.
Lpez, F., Alfaro, A., Garca, M.M., Daz, M.J., Calero, A.M., Ariza, J. 2004. Pulp
and paper from tagasaste (Chamaecytisus proliferus L.F. ssp. palmensis). Chemical
Engineering Research and Design, 82, A8, 1029-1036.
Montgomery, D.C. 1991. Diseo y anlisis de experimentos. Grupo Editorial
Iberoamericana. Mexico.
Pan, X., Arato., C., Gilkes, N., Gregg, D., Mabee, W., Pye, K., Xiao, Z., Zhang, X.,
Saddler, J. 2005. Biorefining of softwoods using ethanol organosolv pulping: Preliminary

188

Publicaciones Cientficas/Scientific Publications


evaluation of process streams for manufacture of fuel-grade ethanol and co-products.
Biotechnol. Bioeng., 90, 4, 473-481.
Qing, Q., Wyman, C.E. 2011. Hydrolysis of different chain length xylooligomers by
cellulose and hemicellulose. Bioresource Technol., 102, 1359-1366.
Requejo, A., Peleteiro, S., Garrote, G., Rodrguez, A., Jimnez, L. 2012. Biorefinery
of olive tree pruning using various processes. Bioresource Technol., 111, 301-307.
Requejo, A., Peleteiro, S., Rodrguez, A., Garrote, G., Paraj, J.C. 2011. Secondgeneration bioethanol from residual woody biomass. Energy and Fuels, 25, 4803-4810.
Rodrguez, A., Moral, A., Snchez, R., Requejo, A., Jimnez, L. 2009. Influence of
variables in the hydrothermal treatment of rice straw on the composition of the resulting
fractions. Bioresource Technol., 100, 4863-4866.
Rodrguez, A., Jimnez, L. 2008. Pulping with organic solvents other than alcohols.
Afinidad, 65, 535, 188-196.
Rodrguez, A., Serrano, L., Moral, A., Prez, A., Jimnez, L. 2008. Use of high-boiling
point organic solvents for pulping oil palm empty fruit bunches. Bioresource Technol., 99,
1743-1749.
Sakaki, T., Yamada, N., Kumagai, S., Uehara, R., Genta, M. 2006. Hydrothermolysis
of agricultural waste. AIChE Annual Meeting Conference Proceeding, San Francisco, CA.
United States, 11, 12-17.
Snchez, R., Rodrguez, A., Navarro, E., Requejo, A., Jimnez, L. 2011. Integrated
utilization of the main components of Hesperaloe funifera. Biochem. Eng. J., 56, 130-136.
Sasaki, M., Adschiri, T., Arai, K. 2003. Fraction of sugarcane bagasse by hydrothermal
treatment. Bioresource Technol., 86, 3, 301-304.
Sundqvist, B., Westermark, U., Erikkson, G. 2006. Cellulose degradation during
hydrothermal treatment of birch wood (Betula pubescens Ehrh.). Cell. Chem. Technol., 40,
3-4, 217-221.
Tortosa, J.F., Rubio, M., Soler, A. 1995. Autohidrlisis de tallo de maz en suspensin
acuosa. Afinidad, 52, 305-311.

189

A. Ferrer et al./Bioresource Technology (10.1016/j.biortech.2012.10.081)


Vzquez, M.J., Alonso, J.L., Domnguez, H., Paraj, J.C. 2007. Production and
refining of soluble products from Eucalyptus globulus glucoronoxylan. Collection of
Czechoslovak Chemical Communications, 72, 3, 307-320.
Vegas, R., Kabel, M., Schols, H.A., Alonso, J.L., Paraj, J.C. 2008. Hydrothermal
processing of rice husks: effects of severity of product distribution. J. Chem. Technol.
Biotechnol., 83, 7, 965-972.
Yez, R., Garrote, G., Dez, M.J. 2009. Valorization of leguminous specie, Sebatia

grandiflora, by means of hydrothermal fractionation. Bioresource Technol., 100, 24, 65146523.


Zhang, J., Lynd, L.R. 2010. Ethanol production from paper sludge by simultaneous
saccharification and co-fermentation using recombinant xylose-fermenting microorganisms.
Biotechnol. Bioeng., 107, 235-244.
Ziaie-Shirkolaee, Y., Mohammadi-Rovshandeh, J., Rezayati-Charani, P., Khajeheian,
M.B. 2008. Influence of dimethyl formamide pulping of wheat straw on cellulose
degradation and comparison with kraft process. Bioresource Technol., 99, 3568-3578.

190

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 2
PULPING OF EMPTY FRUIT BUNCHES (EFB) FROM THE PALM OIL
INDUSTRY BY FORMIC ACID
Ana Ferrer, Alberto Vega, Pablo Ligero and Alejandro Rodrguez
BioResources, 2011, 6 (4) 42824301
ndice de impacto: 1.328 (Cuarta posicin de 21 revistas en el rea de Materials Science,
paper & wood segn el Journal Citation Reports de 2011)

191

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301

192

Publicaciones Cientficas/Scientific Publications


BioResources 6 (4) 42824301
PEER-REVIEWED ARTICLE

bioresources.com

PULPING OF EMPTY FRUIT BUNCHES (EFB) FROM THE PALM OIL INDUSTRY BY
FORMIC ACID
Ana Ferrera, Alberto Vegab, Pablo Ligerob and Alejandro Rodrgueza

a Chemical

Engineering Department, Campus of Rabanales, 14071, University of Crdoba, Crdoba,


Spain
b Department of Physical Chemistry and Chemical Engineering, University of A Corua, Spain

Article submitted: July 4, 2011; Peer review completed: August 10, 2011; Revised version accepted: Sept. 5, 2011;
Published: Sept. 8, 2011

ABSTRACT
Empty Fruit Bunches (EFB) from palm oil were characterized. The holocellulose (67
%), -cellulose (41.9 %) and lignin (24.5 %) contents are similar to wood materials, and
various non-wood materials, but the fiber length is shorter (0.53 mm). The influence of
operational variables in the EFB pulping [formic acid (75-95 %), hydrochloric acid (0.050.15 %) and time (30-150 min)], on the yield, kappa number, viscosity and brightness of the
pulps was studied. By using a factorial design, equations that reproduced the experimental
results for the dependent variables, errors less than 10 % were obtained. These equations
could be used to find suitable conditions, so that operating with not too high values of
operating variables (with minor costs of capital and operation), pulps could be obtained with
acceptable properties. In this way, a cellulosic pulp with a 42.3 % of yield, 22.7 % of
brightness and a 512 mL/g of viscosity was obtained under the following conditions: 92.5 %
of formic acid, 0.075 % of hydrochloric acid and a time of 60 min. A pulp (31.1 of kappa
number and 606 mL/g of viscosity) was bleached by EPabOPoP sequence, achieving a
brightness of 69.4 %, and a loss of viscosity and yield of 34.8 % and 17.7 % respectively. The
residual liquor from the pulping with formic acid 95 %, hydrochloric acid 0.05 % and 30
min, provides a liquor with 18.2 % of residual lignin, 4.1 % of glucose, 9.8 % of xylose and
1.2 % of arabinose, all on dry basis weight of original material.

KEYWORDS: Empty Fruit Bunches (EFB); Biorefinery; Formic acid; Pulp; Lignin; Sugars.
193

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


1. INTRODUCTION
Bio-renewable resources have been important in all cultures as a source of products
and energy during the evolution of humanity. Subsequently, the discovery of huge deposits
of petroleum and the development of engineering and industry led the petrochemical
industry to become dominant in the production of energy and chemicals. After the first oil
crisis the risk of dependence on fossil fuels, the need for revision of these energy models and
the interest in the search for traditional sources based on renewable resources were revealed.
This, coupled with increasingly important social and legal pressures to implement
transformation processes more environmentally acceptable, has made the vision of plant
biomass as a raw material become more interesting. Among the many forms of plant
biomass available, lignocellulosic wastes from certain processing industries can play an
important role in combustion and the subsequent production of energy or in chemical
processing for the isolation of its fibers, which can be used for paper making, cellulose
derivatives and other chemicals; this fractionation of the raw materials contributes to the
development of an industrial concept known as the biorefinery (Rodrguez et al. 2010).
An interesting biomass is Empty Fruit Bunches (EFB), which consists of certain
lignocellulosic material residue from the palm oil industry (Wan Rosli and Law 2011). The
first published work that concerning applications of EFB was that of Muthurajah and Peh
(1977), who used the kraft process with a concentration of active alkali of 16 %, for 3 hours
at 160 C, obtaining a pulp with a yield of 56 % and a kappa number of 16.9. The chemical
properties of EFB fibers are similar to those of hardwood, except for the pentosan content
which is higher. Subsequently, the kraft process has also been studied by other authors
(Akamatsu et al. 1987a, Ibrahim 2002, Khoo and Lee 1991).
Ibrahim (2002) compared the composition of EFB pulp obtained by the kraft, kraftanthraquinone, soda and soda-anthraquinone processes; the pulp obtained with the soda
process had the highest lignin, holocellulose and -cellulose contents and a higher viscosity.
The soda process has also been studied by Law and Jiang (2001), producing fibers with more
wall thickness, greater rigidity, higher solubilities in hot water and 1 %-soda as well as higher
ash content. These pulps were bleached with hydrogen peroxide more easily than those of
aspen; their paper sheets had a lower tensile index, but a greater stretch and tear index than
those of aspen. Moreover, Daud et al. (1998) pulped the EFB with soda, sodium carbonate
and sodium sulfite processes; it was found that the soda process provided the best pulp.

194

Publicaciones Cientficas/Scientific Publications


Wanrosli et al. (2004) studied, using a central composite experimental design, the
influence of operating conditions (temperature, time and alkali concentration) on the
properties of EFB pulps (yield, kappa number, tensile and tear index). They obtained pulp
yields in the range of 30 to 45 %. It was considered that the values of 160 C, 60 to 120
minutes and 20 to 30 % of alkali are sufficient for the proper pulping.
The semi-chemical pulp for board by soda-anthraquinone process (Roliadi and
Pasaribu 2004) has also been studied. The study of thermomechanical pulp has also been
made by several researchers (Akamatsu et al. 1987b, Daud et al. 2005, Ghazali et al. 2006,
Kamishima et al. 1988, Khoo and Lee 1991).
Organosolv processes have also been studied. Pulping was done with ethanol (Aziz et
al. 2002, Suleman and Yusoff 1997) and with a modified IDE process (ethanol containing
soda) (Quader and Lonnberg 2005). Finally, biopulping has been investigated using a white
fungus K14 (Goenadi et al. 1998).
The kraft process for obtaining cellulose pulp remains, due to its favorable
characteristics, predominant in the global paper industry. It is well known that this method
has certain implicit environmental problems arising from the use of inorganic sulphur agents
in cooking and chlorinated derivatives in bleaching. Although there have been significant
improvements in the overall process (extended delignification, use of oxygen bleaching
agents, etc.) researching alternative methods of delignification has continued intensively.
Among these alternatives, the organosolv processes, where delignification is carried out
with the aid of an organic solvent, are among the most investigated. Organosolv processes
can be understood from a conceptual point of view as being different from sulphate or
sulphite cooking. Thus, they can be designed as fractionation treatments rather than simple
delignification (Dapia et al. 2002), so it is possible to try to extract the main polymeric
components of plant tissues in the least degraded form. The possibility to work under milder
conditions, in some cases, is a factor that facilitates the fractionation. Low investment costs
for a new plant, less pollution problems, easy solvent recovery, low prices of some organic
agents used, and the possibility of cooking at atmospheric pressure and low temperatures,
among others, are often cited as additional advantages (Rodrguez et al. 2010). All this has
produced the investigation of these processes in recent years, both with traditional substrates
and new plant vegetables.

195

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


The use of short-chain organic acids (mainly formic and acetic) has emerged as an
attractive and feasible alternative for the delignification of lignocellulosic materials
(Shahriarinour et al. 2011).
After the pioneering work of Pauly in 1917 and the studies of Freudenberg et al. and
Wright and Hibbert in the decade of the 1930s (Ligero et al. 2008, 2010), formic acid has
attracted the attention of researchers. Formic acid has been proved to be a good
delignification agent with a wide variety of starting materials. Thus, it has been used for
pulping different woody species such as eucalyptus (Baeza et al. 1991a, Erismann et al. 1994,
Ligero et al. 2008), pine (Baeza et al. 1991b) and beech (Dapa et al. 2002). It has also
shown its effectiveness in treating a wide variety of non-woody materials such as wheat straw
(Jimnez et al. 1998), rice straw (Lam et al. 2001), Miscanthus (Caridad et al. 2004, Ligero et
al. 2010), and corn cobs (Zhan et al. 2010), among other new plant vegetables.
Formic acid delignification uses the properties of this compound to act simultaneously
as an organic solvent and as an acid. In fact, a significant amount of materials can be
dissolved at relatively low processing temperatures (Baeza et al. 1991a, Ligero et al. 2010,
Zhan et al. 2010), even working without mineral acid added as a catalyst (Jimnez et al.
1998). Formic acid acts through a chemical reaction by partial depolymerization of lignin
and hemicelluloses, producing oligomers that are soluble in the liquid mixture. Cellulose
and liquor are separated by filtration, and the dissolved lignin can be recovered easily by the
modification of the ionic strength of the liquor (water dilution). Furthermore, the distillation
of the liquor enables the recovery of formic acid, leaving a fraction rich in sugars that can be
fermented for alcohols.
When the formic acid concentration is high (e.g., > 80 % by weight) the organosolv
liquors have a great delignification power (Baeza et al. 1991a, Jahan et al. 2007, Jimnez et
al. 1998, Ligero et al. 2008, Zhan et al. 2010). But, increasing formic acid proportion above
80 % damages the cellulose (especially when hydrochloric acid is used as a catalyst) which is
reflected in lower viscosity values. Thus, it is necessary to reach a compromise between the
values of the formic acid concentration, hydrochloric acid and treatment time, so controlling
these variables is of primary importance.
This paper reports the characterization of EFB (holocellulose, -cellulose, lignin, hotwater solubles, 1 %-NaOH solubles, ethanol-benzene extractives, ash and fibers length
distribution). On the other hand, the influence of operational variables in the formic acid
pulping of EFB (viz. formic acid concentration, hydrochloric acid concentration and
196

Publicaciones Cientficas/Scientific Publications


processing time) on the dependent variables of pulps obtained (yield, kappa number,
viscosity and brightness) was studied, in order to find mathematical models that can simulate
the pulping process and find the optimal operating conditions, while achieving some
acceptable values for the properties of the resulting pulp. A pulp was also bleached by the
sequence EPabOPoP (alkali extraction - peracetic in basic medium - oxygen - hydrogen
peroxide with oxygen - hydrogen peroxide). Finally, the residual liquor was characterized by
determining the lignin and fermentable sugar contents.

2. EXPERIMENTAL
Raw material
In this work, EFB has been obtained from African palm (Elaeis guineensis) and
provided by Straw Pulping Engineering S.L. of Zaragoza (Spain) company.

Raw material characterization


Following drying at room temperature, the raw material was cold ground in a Retsch
SM 2000 mill to avoid alterations in its components. The ground product was sieved and the
0.25-0.40 mm fraction (sieves No. 60 and 40 in the Tyler series) was saved for analysis.
Chemical properties of EFB were determined in accordance with the respective
TAPPI standards for the different components, namely: T-9m-54 for holocellulose, T-222
for lignin, T-203 0S-61 for -cellulose, T-257 for hot-water solubles, T-212 for 1 %-NaOH
solubles, T-204 for ethanol-benzene extractives and T-211 for ash. Fiber length was
determined biometrically, under a Visopan projection microscope with 10x objective for
100x magnification, after microcooking the raw material with 10 % of soda at 80 C for 1 h
and subsequent staining the fibers with 1 % safranin. 100 fibers were measured.

Pulping
Mixtures of shredded EFB, water and formic acid (proportions ranging from 75 to 95
% by weight of liquor) were heated to the boiling point (approximately 108 C) in glass Pyrex
flasks. Hydrochloric acid was added (0.05-0.15 % by weight of liquor) when boiling started
(zero time), and mixtures were maintained at total reflux with stirring for different times (30150 min) at atmospheric pressure. After the reaction, the pulps were separated by filtration
197

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


and the solids washed with concentrated formic acid solutions (80 %w) in order to avoid the
deposition of the dissolved lignin on the pulp. Four washings were performed using 2.5, 2.5,
5 and 5 volumes of formic acid solution with respect to the initial dry weight EFB. Finally the
pulps were treated with water until neutrality and let to dry at room temperature (Ligero et
al. 2010).

Pulp characterization
For all experiments the main parameters defining delignification and pulping were
measured as follows: pulp yield after oven drying of a pulp aliquot to constant weight, kappa
number as per TAPPI T236, and intrinsic viscosity as per TAPPI T230. To measure the
ISO brightness, paper sheets from pulps were formed according to TAPPI method T272
and diffuse blue reflectance measured as per TAPPI T525 in a Photovolt 577 reflectometer.

Experimental design
A second order factorial design of experiments was used (Montgomery 1991) which
consisted of a central experiment (in the center of a cube) and several additional points
(additional experiments lying at the cube vertices and side centers). Experimental data were
fitted to the following second-order polynomial:
Ye= a0 + a1XF + a2XH + a3XT + a11XF2 + a12XF XH + a13XF XT + a22XH2 + a23XH XT + a33XT2

(1)

where Ye denotes the response variables [viz., yield (YI), kappa number (KN), viscosity (VI)

or brightness (BR)]; XF, XH and XT are the normalized values of the operational variables
(formic acid concentration (F), hydrochloric acid concentration (H) and processing time (T),
respectively); and a0 to a33 are constants.

The values of the operational variables were normalized to values from -1 to +1 by

using the following expression,


Xn =

)
2(X-X

X max -X min

(2)

where Xn is the normalized value of F, H or T; X is the actual experimental value of the

is the mean of Xmax and Xmin; and Xmax and Xmin are the maximum
variable concerned; X
and minimum value, respectively, of such a variable.

198

Publicaciones Cientficas/Scientific Publications


The normalized values for the independent variables in the 15 experiments conducted
are given in Table 1.
Table 1. Values of operational variables and experimental values of the properties of pulps obtained
by formic acid pulping of EFB.

Normalized values
of operational variables
(XF, XH, XT)
-1,-1,-1

+1,-1,-1

43.6

31.1

606

22.3

-1,+1,-1

44.6

35.3

458

18.0

+1,+1,-1

43.1

33.9

468

22.2

-1,-1,+1

39.7

23.6

389

28.8

+1,-1,+1

40.5

31.7

479

22.1

-1,+1,+1

39.1

23.4

354

26.5

+1,+1,+1

38.5

26.3

425

25.6

-1,0,0

40.0

30.4

402

23.4

10

+1,0,0

41.5

35.3

462

19.7

11

0,-1,0

41.3

24.2

465

27.1

12

0,+1,0

39.3

18.7

420

29.3

13

0,0,-1

44.1

32.7

498

20.0

14

0,0,+1

37.3

26.2

387

25.2

15

0,0,0

39.9

27.9

431

25.0

16

-0.5,-0.5,-0.5

43.4

29.9

461

21.9

17

+0.5,+0.5,+0.5

36.8

26.0

430

26.4

Experiment

Yield, %

Kappa
number

Viscosity,
mL/g

Brightness,
%

46.4

45.1

455

17.7

Characterization of residual liquor


After the fractionation/pulping process, the lignin was precipitated by treating one
volume of black liquor with seven volumes of water and then stirring (Ligero et al. 2010).
The precipitated solids were washed repeatedly with water until neutrality in order to remove
the maximum amount of carbohydrates possible. The quantitative determination of
monosaccharides was performed using liquid chromatography. The samples were diluted
with water to a suitable concentration of sugars, neutralized with barium carbonate, and
filtered through a 0.45 m membrane. A Waters 1500 Series HPLC equipped with a Biorad
Aminex HPX-87P column thermostated to 85 C enabled the determination of individual
monosaccharides (refractive index detector at 30 C) and furfural (UV/VIS wavelength
detector at 254 nm). HPLC-grade water was used as mobile phase at a flow of 0.6 mL/min.
199

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


3. RESULTS AND DISCUSSION
Analysis of EFB
Table 2 shows the results of the chemical analysis of EFB, various agricultural residues
(olive prunings, wheat straw, sorghum stalks, rice straw, sugarcane bagasse, vine shoots and
cotton stalks (Jimnez et al. 1990, 1993, 2006)), alternative raw materials (Jimnez et al.
2005, 2006), and hardwoods and softwoods (Alonso 1976). This table compares similar
characteristics among hardwoods, softwoods and EFB (Alonso 1976). A comparison of the
data for EFB (Table 2) with those of the other raw materials (Table 2) revealed the following:
a) The content in hot water solubles of EFB is similar to those of bagasse and
cotton stalks, but higher than those of the other alternative raw materials except paulownia and Prosopis julyflora -, pine and eucalyptus wood.
b) The content in 1 %-NaOH solubles of EFB is lower than those of olive
prunings, bagasse, cotton stalks, but higher than those of the alternative raw
materials, and pine and eucalyptus wood, and similar to those of the other
agricultural residues studied.
c) The content in ethanol-benzene extractives of EFB is similar to those of rice
straw, bagasse, cotton stalks, and pine and eucalyptus wood, but lower than
those of the other agricultural residues and the alternative raw materials.
d) The ash content of EFB is higher than those of olive prunings, bagasse and
cotton stalks, similar to those of the other agricultural residues, and much
higher than those of pine and eucalyptus wood.
e) The holocellulose content of EFB is higher than those of olive prunings and
sorghum stalks, similar to those of vine shoots, and lower than those of the
other agricultural residues (wheat straw, sunflower stalks and cotton stalks). It is
also higher than those of phragmites and prosopis, lower than those of the
other raw materials, and between those of pine and eucalyptus wood.
f) The lignin content of EFB is higher than those of the agricultural residues,
alternative raw materials and eucalyptus wood, but lower than that of pine
wood.

200

Publicaciones Cientficas/Scientific Publications


Table 2. Chemical properties of various agricultural residues, alternative raw materials and hardwoods and softwoods.
Analysis(%)
EFB

Leucaena diversifolia (Jimnez et al. 2006)


Leucaena colinsi (Jimnez et al. 2006)
Leucaena leucocephala (Honduras) (Jimnez et al. 2006)
Leucaena leucocephala (India) (Jimnez et al. 2006)
Tagasaste(Australia) (Jimnez et al. 2006)
Tagasaste (Nueva Zelanda) (Jimnez et al. 2006)
Tagasaste (La Palma) (Jimnez et al. 2006)
Retama monosperma (Jimnez et al. 2005)
Phragmites (Jimnez et al. 2005)
Arundo donax (Jimnez et al. 2005)
Prosopis julyflora (Jimnez et al. 2005)
Prosopis alba (Jimnez et al. 2005)
Paulownia fortunei (Jimnez et al. 2005)
Olive prunings (Jimnez et al. 1990)
Wheat straw (Jimnez et al. 1990)
Sunflower stalks (Jimnez et al. 1990)
Sorghum stalks (Jimnez et al. 1993)
Rice straw (Jimnez et al. 2005b)
Sugarcane bagasse (Jimnez et al. 2005)
Vine shoots (Jimnez et al. 2006)
Cotton stalks (Jimnez et al. 2006)

Pinus Pinaster
Pine varieties (11 varieties)

Eucalyptus globulus
Eucalyptus varieties (20 varieties)

Hot water
solubles

1 %-NaOH
solubles

4.03
3.24
4.30
5.01
3.98
2.96
2.99
2.41
3.84
5.38
4.73
6.49
4.67
9.6
8.16
12.27
22.72
21.7
16.57
4.4
16.09
3.33
1.99
2.04
2.84
4.19

20.22
17.38
20.02
20.26
18.44
15.55
16.15
16.62
16.93
34.77
26.80
22.56
20.86
31.5
30.04
43.58
47.81
45.58
46.94
33.92
39.21
20.34
7.98
10.33
12.42
14.69

Ethanolbenzene
extractives
1.17
4.44
4.64
6.01
4.64
2.17
3.43
3.30
5.03
6.36
7.30
5.30
4.65
5.50
10.36
4.01
4.07
7.99
1.4
1.73
4.87
1.42
2.57
1.29
1.15
2.09

Ash

Holocellulose

-cellulose

3.2

66.97
77.88
80.79
74.11
75.92
82.16
75.36
76.47
71.76
64.16
70.20
62.77
63.56
70.7
61.47
76.2
71.76
65.93
70.6
80.2
67.14
72.86
69.59
67.60
80.17
77.21

41.90
40.10
43.77
41.21
44.43
47.65
43.59
44.99
42.75
39.76
40.46
36.55
41.55
37.40
35.67
39.72
42.1
41.5

25.8
36.48
29.66
24.43

41.14
58.48
55.92

26
14.38
13.67

52.79

27.68

1.36
6.49
7.9
4.85
15.39
2.1
3.49
2.17
0.54
0.33
0.57
0.57

Hemicellulose

Lignin
24.45
19.09
17.04
19.39
21.43
15.71
14.84
14.10
21.50
23.66
22.34
20.60
19.27
22.4
19.71
17.28
13.44
15.64
25.23
19.8
20.27
21.45
26.22
28.83
19.96
25.54

201

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


EFB fibers length
Figure 1 shows a photograph of the EFB fibers. The maximum, average and minimum
values of fiber lengths were 1.48, 0.53 and 0.27 mm, respectively, and the maximum,
medium and minimum thicknesses or widths were 26, 14 and 8 m. Figure 2 shows the
length distribution curve, showing that the fibers having a size of 0.50 to 0.70 mm are more
abundant.
The average length of the EFB fibers (0.53 mm) is lower than those of other raw
materials: wheat straw (1.14 mm), sorghum stalks (1.32 mm), olive prunings (1.03 mm),
sunflower stalks (1.30 mm), vine shoots (0.79 mm), cotton stalks (1.03 mm), Eucalyptus

globulus (1.05 mm) and Pinus pinaster (2.50 mm).

Fibers, %

Figure 1. Microphotography of EFB fibers.

45
40
35
30
25
20
15
10
5
0

0,2

0,4

0,6

0,8

Fibers length (mm)

1,2

1,4

1,6

Figure 2. Distribution curve of the length of the fibers of the EFB.

202

1,8

Publicaciones Cientficas/Scientific Publications


Pulping models
Preliminary experiments of trial and error were carried out, based on the results of
other researchers on different raw materials, in order to define the ranges of operating
variables. In this way the following ranges were chosen: formic acid concentration from 75 to
95 %, hydrochloric acid concentration, which usually acts as a catalyst, from 0.05 to 0.15 %,
and the processing time from 30 to 150 minutes. Constant conditions were maintained with
respect to operating temperature (108 C) and the liquid/solid ratio (10:1).
Table 1 shows experimental values of the pulp properties, which differed by less than
5-10 % from their means as obtained in triplicate measures. The experimental results were
fitted to a polynomial model by multiple regression using the software BMDP. The terms
showing a Snedecor F-value greater than 2.5 and a Student t-value greater than 1.5 were

deemed statistically significant. Table 3 gives the coefficients of the different terms in the
equations, as well as the highest p and the lowest Student t-values for the terms.
Table 3. Polynomial models for the properties of pulp obtained by formic acid pulping of EFB.
(Values of the constants in the polynomial equations).
Dependent
variable

Values of p and tStudent

Values of the constants in the polynomial equations


a0

a1

a2

a3

a11

a12

a13

a22

a33

p<

>t

Yield

40.0

-0.7

-2.7

1.0

0.6

0.9

0.09

1.9

Kappa
number

26.7

-1.8

-4.7

6.4

3.3

-5.0

3.0

0.13

1.6

Viscosity

436

38

-27

-45

-20

16

0.14

1.6

Brightness

24.8

2.8

-3.2

-2.1

3.4

-2.2

0.04

2.4

The predictions of the previous equations reproduced the experimental results for the
dependent variables with errors less than 3 % for yield, 10 % for kappa number, 6 % for
viscosity and 10 % for brightness of pulps, in most of the cases (Table 4).
The proposed models were validated by conducting two pulping experiments (entries
16 and 17 in Table 1). The errors made in predicting pulp properties by using the
polynomial models were quite small (Table 4). This testified to the accuracy of such models.
The values of the operational variables providing the best pulp properties (yield, kappa
number, viscosity and brightness) were identified by using multiple non-linear programming
as implemented in the method of More and Toraldo (1989). Table 5 shows the optimum
203

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


values of the dependent variables and those of the operational variables required to obtain
them.
Table 4. Values of the dependent variables as estimated with the polynomial models and deviations
from their experimental counterparts (in brackets).
Experiment

Yield, %

Kappa number

Viscosity, mL/g

Brightness, %

45.9 (1.08)

40.9 (9.31)

466 (2.42)

17.9 (1.13)

44.7 (2.52)

34.3 (10.29)

582 (3.96)

22.1 (0.90)

44.5 (0.22)

37.3 (5.67)

452 (1.31)

17.9 (0.56)

43.3 (0.46)

30.7 (9.44)

488 (4.27)

22.1 (0.45)

39.3 (1.01)

24.9 (5.51)

376 (3.34)

27.7 (3.82)

40.5 (0.00)

31.5 (0.63)

492 (2.71)

23.5 (6.33)

37.9 (3.07)

21.3 (8.97)

362 (2.26)

27.7 (4.53)

39.1 (1.56)

27.9 (6.08)

398 (6.35)

23.5 (8.20)

41.0 (2.50)

33.1 (8.88)

398 (1.00)

21.6 (7.69)

10

41.0 (1.20)

33.1 (6.23)

474 (2.60)

21.6 (9.64)

11

40.7 (1.45)

23.5 (2.89)

479 (3.01)

28.2 (4.06)

12

39.3 (0.00)

19.9 (6.42)

425 (1.19)

28.2 (3.75)

13

43.6 (1.13)

34.4 (5.20)

481 (3.41)

19.8 (1.00)

14

38.2 (2.41)

25.0 (4.58)

391 (1.03)

25.4 (0.79)

15

40.0 (0.25)

26.7 (4.30)

436 (1.16)

24.8 (0.80)

16

42.3 (2.53)

31.9 (6.69)

452 (1.95)

22.4 (2.28)

17

38.9 (5.71)

25.4 (2.31)

418 (2.79)

25.2 (4.55)

Table 5. Optimal properties in pulp obtained by formic acid pulping of EFB.

Dependent variable

204

Optimum (maximum or
minimum*) value of the
dependent variable

Values of the operational variables


required to obtain the optimum
values of the dependent variables
XF

XH

XT

Yield, %

45.9

-1

-1

-1

Kappa number

17.8*

-0.23

+1

+0.90

Viscosity, mL/g

582

+1

-1

-1

Brightness, %

29.3

-0.25

+1

+0.75

Publicaciones Cientficas/Scientific Publications


The polynomial equations allowed the identification of the more influencing
operational variables on the pulp properties. The maximum variations in the dependent
variables with changes in the operational variables over the studied range were obtained by
altering one independent variable at a time while keeping all others constant. The results are
shown in Table 6 together with the maximum percent differences in the dependent variables
from their optimum values over the studied variation ranges.
Table 6. Maximum changes in the dependent variables with changes in one operational variable with
others held constant (the percent differences from the changes are given in brackets).
Equation

Variation with the operational variable


Formic acid
Hydrochloric acid
Processing time

Yield, %

1.6 (3.48 %)

1.4 (3.05 %)

6.6 (14.37 %)

Kappa number

9.7 (54.49 %)

6.9 (38.76 %)

10.9 (61.24 %)

Viscosity, mL/g

116 (19.93 %)

94 (16.15 %)

90 (15.46 %)

Brightness, %

5.0 (17.06 %)

3.4 (11.60 %)

6.8 (23.21 %)

Figures 3 to 10 also show the changes of the dependent variables to vary the operating
variables. The maximum yield (45.9 %) corresponded to low values of operating variables.
The processing time influences pulp yield more significantly than the concentration of
formic acid (Fig. 3), while the latter variable has a greater effect than the concentration of
hydrochloric acid (Fig. 4).

Figure 3. Variation of pulp yield with processing time and formic acid concentration, at low
hydrochloric acid concentration (XH = -1).
205

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301

Figure 4. Variation of pulp yield with formic acid concentration and hydrochloric acid concentration,
at short processing time (XT = -1).

The kappa number decreased more sharply with time than with the formic acid
concentration (Fig. 5), indicating that the hydrochloric acid concentration was the least
influential variable (Fig. 6). The minimum kappa number was obtained when the
hydrochloric acid concentration was high, the formic acid concentration near the average
value and close to the maximum time considered.

Figure 5. Variation of pulp kappa number with processing time and formic acid concentration, at
high hydrochloric acid concentration (XH = +1).

206

Publicaciones Cientficas/Scientific Publications

Figure 6. Variation of pulp kappa number with formic acid concentration and hydrochloric acid
concentration at long processing time (XT = +0.90).

The viscosity was maximum (582 mL/g) when operating with low levels of
hydrochloric acid concentration and processing time and a high concentration of formic
acid. The formic acid concentration had more influence on the viscosity than the
hydrochloric acid concentration (Fig. 7), while time influenced the viscosity a little less than
the concentration of hydrochloric acid (Fig. 8).

Figure 7. Variation of viscosity with formic acid concentration and hydrochloric acid concentration at
short processing time (XT = -1).

207

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301

Figure 8. Variation of viscosity with hydrochloric acid concentration and processing time, at high
formic acid concentration (XF = +1).

The processing time had more influence on the brightness than the concentration of
formic acid (Fig. 9), and the concentration of hydrochloric acid was the least influencing
variable again (Fig. 10). The maximum brightness was achieved for an average-high time, a
high concentration of hydrochloric acid and formic acid concentration close to the average
used.

Figure 9. Variation of brightness with processing time and formic acid concentration, at high
hydrochloric acid concentration (XH = +1).

208

Publicaciones Cientficas/Scientific Publications

Figure 10. Variation of brightness with formic acid concentration and hydrochloric acid
concentration, at medium-long processing time (XT = +0.75).

Optimum operating conditions


Considering the "Yield/Kappa" ratio (which is desirable to have at a high value),
adjusting the experimental data to a polynomial model gives the equation:
Yield
=1.53 + 0.07 XH + 0.12 XT 0.31 XF2 0.12 XF XT + 0.34 XH2 0.18 XT2
Kappa
(p < 0.14; t-Student > 1.6)

(3)

The values of the operational variables providing the maximum "Yield/Kappa" ratio
were identified by using multiple non-linear programming as implemented in the method of
More and Toraldo (1989). The maximum value (1.94) is found when the formic acid
concentration is medium (85 %), hydrochloric acid concentration is high (0.15 %) and the
processing time is 90 minutes. Figure 11 shows the change of the "Yield/Kappa ratio with
formic acid concentration and processing time, for a high concentration of hydrochloric acid.
Considering the "Viscosity/Kappa" ratio (which is also desirable to be high), adjusting
the experimental data to Eq. (1) gives the equation:
Viscosity
=16.58 + 0.98 XF 3.71 XF2 1.32 XF XT + 3.97 XH2 1.87 XT2
Kappa
(p < 0.17; t-Student > 1.5)

(4)

209

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


Operating in a similar way as equation (3), equation (4) gives the maximum value of
the "Viscosity/Kappa" ratio of 20.62, for operating conditions close to those found in the case
of the "Yield/Kappa ratio. Figure 12 shows the changes in the "Viscosity/Kappa" ratio by
varying the formic acid concentration and the processing time, for a high hydrochloric acid
concentration.
Chemicals, energy and immobilized capital can be saved in industrial facilities by using
lower formic acid and hydrochloric acid concentrations and shorter processing times, than
considered in the previous paragraph. The data of Tables 3 and 4 can be used to select
values of the operational variables providing near-optimal pulp properties while saving
chemicals, energy and immobilized capital by using lower values of operational variables.
One combination leading to near-optimal properties with reduced costs with properties is
using a processing time of 60 min, a formic acid concentration of 92.5 % and a hydrochloric
acid concentration of 0.075 %. Operating under these conditions the following values for the
dependent variables were obtained: yield 42.25 %, brightness 22.7 %, and viscosity 512
mL/g; these values deviate by 8.0 %, 22.5 % and 12.0 %, respectively of the maximum values
of yield, brightness and viscosity, respectively. It is found that when operating under these
conditions the values "Yield/Kappa" ratio and "Viscosity/Kappa" ratio of 1.68 and 17.60 are
achieved, respectively; these values are far from the top surfaces presented in Figs. 11 and
12.

Figure 11. Variation of Yield/Kappa ratio with formic acid concentration and processing time, at
high hydrochloric acid concentration.

210

Publicaciones Cientficas/Scientific Publications

Figure 12. Variation of Viscosity/Kappa ratio with formic acid concentration and processing time,
at high hydrochloric acid concentration.

Bleaching
In order to obtain a bleached pulp with minimal degradation is suitable to select an
unbleached pulp with higher viscosity (Exp. 2). Applying the sequence EPabOPoP (alkali
extraction - peracetic in basic medium - oxygen - hydrogen peroxide with oxygen - hydrogen
peroxide) gave the experimental results shown in Table 7. As can be seen, the pulp yield is
decreasing with subsequent bleaching stages used near the end of the value of 35.9 %,
representing a 17.7 % of loss in the performance of the original pulp. The viscosity of the
pulps decreased significantly especially in Pab and O stages, reaching the final stage with a
loss of viscosity in the complete sequence of 34.8 %. Finally the brightness rose to 69.4 %.
Table 7. Bleaching of pulp from EFB.
Properties

Original Pulp

Stage E

Stage Pab

Stage O

Stage Po

Stage P

Yield, %

43.6

41.6

40.7

37.3

36.5

35.9

Kappa number

31.1

16.2

13.0

10.1

7.8

5.6

Viscosity, mL/g

606

616

443

422

408

395

Brightness, %

22.3

37.0

42.0

50.6

50.2

69.4

Characterization of residual liquor


The residual liquor from Experiment 2 is considered (which gives rise to a pulp with
high viscosity). Proceeding as specified in section Characterization of residual liquor (in
Experimental), 18.2 % of the lignin of original dry raw material may be separated. The liquid
211

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


fraction obtained after the removal of lignin contained glucose 4.1 %, 9.8 % of xylose and 1.2
% of arabinose, based on the original dry raw material.
Separated lignin may have several applications or it may be converted into fuel gas
through gasification and pyrolysis operations (Snchez et al. 2010). On the other hand, one
of them can ferment sugars derived from the pulping and use them for the production of
alcohols (Garrote et al. 2007a,b, Rodrguez et al. 2010).

4. CONCLUSIONS
1. A comparison of the chemical properties of EFB and various agricultural
residues (olive prunings, wheat straw, sunflower stalks, sorghum stalks, rice
straw, sugarcane bagasse, vine shoots and cotton stalks), alternative raw
materials (leucaena, tagasaste, bridal broom, phragmites, giant reed, prosopis
and paulownia) and hardwoods (eucalyptus) and softwoods (pine) revealed that
EFB is an alternative source of cellulose for producing cellulosic pulp.
2. Organosolv delignification of EFB could be effectively achieved with the formic
acid-hydrochloric acid-water system, at atmospheric pressure.
3. Using polynomial equations, the properties of pulps (yield, kappa number,
viscosity and brightness) have been estimated, with errors less than 10 % in
most of the cases, as a function of operating variables (formic acid and
hydrochloric acid concentrations and processing time).
4. The search for a compromise between pulp properties and chemicals and
energy led to the following operating conditions: concentrations of formic acid
and hydrochloric acid of 92.5 % and 0.075 %, respectively, and processing time
of 60 min. Operating under these conditions provided a pulp with acceptable
values of its properties (42.3 % of yield, 22.7 % of brightness and viscosity of
512 mL/g).
5. A pulp (31.1 of kappa number and 606 mL/g of viscosity) was bleached by
EPabOPoP sequence, achieving a brightness of 69.4 %, and a loss of viscosity
and yield of 34.8 % and 17.7 %, respectively.

212

Publicaciones Cientficas/Scientific Publications


6. The residual liquor was separated in 18.2 % of lignin, 4.1 % of glucose, 9.8 %
of xylose and 1.2 % of arabinose, all on dry basis weight of original raw
material.

ACKNOWLEDGEMENTS
The authors are grateful to Spains DGICyT for funding this research within the
framework of the Projects TRA2009-0064 and CTQ2010-19844-C02-01.

REFERENCES
Akamatsu, I., Husin, M.B., Kamishima, H. and Hassan, A.H. (1987a). Industrial
utilization of oil palm (Elaeis guineensis) by-products. I. Kraft-anthraquinone pulping of oil
palm empty fruit bunches. Cell Chem Technol. 21, 1, 67-75.
Akamatsu, I., Kobayashi, Y., Kamishima, H., Kamarudin, H., Yusoff, M., Husin, M.
and Haji, H. (1987b). Industrial utilization of oil palm by-products. II. Thermomechanical
pulping of empty fruit bunches. Cell Chem Technol. 21, 2, 191-197.
Alonso, L. (1976). Anlisis qumico de diferentes especies forestales. INIA
Ministerio de Agricultura. Madrid, Spain.
Aziz, A.A., Husin, M. and Mokhtar, A. (2002). Preparation of cellulose from oil
palm empty fruit bunches via ethanol digestion: effect of acid and alkali catalysts. J Oil Palm
Res. 14, 1, 9-14.
Baeza, J., Fernndez, A.M., Freer, J., Pedreros, A., Schmidt, E. and Durn, A.N.
(1991b). The influence of formic acid delignification on the enzymatic hydrolysis of Pinus

radiata D. don sawdust. Appl Biochem Biotech. 31, 273-282.


Baeza, J., Urizar, S., Erismann, N.M., Freer, J., Schmidt, E. and Durn, N. (1991a).
Organosolv pulping-V: Formic acid delignification of Eucalyptus globulus and Eucalyptus

grandis. Bioresource Technol. 37, 1-6.


Caridad, R., Ligero, P., Vega, A. and Bao, M. (2004). Formic acid delignification of

Miscanthus sinensis. Cell Chem Technol. 39, 3-4, 235-244.


Dapa, S., Santos, V. and Paraj, J.C. (2002). Study of formic acid as an agent for
biomass fractionation. Biomass Bioenerg. 22, 213-221.
213

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


Daud, W.R.W., Law, K.N. and Valade, J.L. (1998). Chemical pulping of oil palm
empty fruit bunches. Cell Chem Technol. 32, 1-2, 133-143.
Daud, W.R.W., Law, K.N. and Ghazali, A. (2005). Alkaline peroxide mechanical
pulping of oil palm lignocellulosics. Part 1. Effects of chemical and mechanical
pretreatments on empty fruit bunch (EFB) chemical composition. JIRCAS Working
Report 39, 82-85.
Erismann, N.M., Freer, J., Baeza, J. and Durn, N. (1994). Organosolv pulping-VII:
Delignification selectivity of formic acid pulping of Eucalyptus grandis. Bioresource
Technol. 47, 247-256.
Garrote, G., Falque, E., Domnguez, H. and Paraj, J.C. (2007a). Autohydrolysis of
agricultural residues: study of reaction byproducts. Bioresource Technol. 98, 1951-1957.
Garrote, G., Kabel, M.A., Schols, H.A., Falque, E., Domnguez, H. and Paraj, J.C.
(2007b). Effects of Eucalyptus globulus wood autohydrolysis conditions on the reaction
products. J Agr Food Chem. 55, 22, 9006-9013.
Ghazali, A., Wanrosli, W.D. and Law, K.N. (2006). Alkaline peroxide mechanical
pulping (APMP) of oil palm lignocellulosics. Part 2. Empty fruit bunch (EFB) responses to
pretreatments. Appita J. 59, 1, 65-70.
Goenadi, D.H., Away, Y., Suharyanto, P., Panji, T., Watanabe, T. and Kuwahara, M.
(1998). Bio-pulping of empty fruit bunches of oil palm by white rot fungi isolated from
tropical plantation. International Conference on Biotechnology in the Pulp and Paper
Industry, 7th, Vancouver, B.C., June 16-19, B, 49-51.

Ibrahim, R. (2002). Chemical composition of alkaline pulps from oil palm empty

fruit bunches. Oil Palm Bulletin 44, 19-24.


Jahan, M.S., Chowdhury, D.A.N. and Islam, M.K. (2007). Atmospheric formic acid
pulping and TCF bleaching of dhaincha (Sesbania aculeata), kash (Saccharum spontaneum)
and banana stem (Musa Cavendish). Ind Crop Prod. 26, 324-331.
Jimnez, L., de la Torre, M.J., Maestre, F., Ferrer, J.L. and Prez, I. (1998).
Delignification of wheat straw by use of low-molecular-weight organic acids. Holzforschung
52, 191-196.
Jimnez, L., Lpez, F. and Martnez, C. (1993). Paper from sorghum stalks.
Holzforschung 47, 529-533.
214

Publicaciones Cientficas/Scientific Publications


Jimnez, L., Prez, A., de la Torre, M.J., Moral, A. and Serrano, L. (2006).
Characterization of vine shoots, cotton stalks, Leucaena leucocephala and Chamaecytisus

proliferus, and of their ethyleneglycol pulps. Bioresource Technol. 98, 3487-3490.


Jimnez, L., Rodrguez, A., Ferrer, J.L., Prez, A. and Angulo, V. (2005). Paulownia,
a fast-growing plant, as a raw material for paper manufacturing. Afinidad 62, 516, 100-105.
Jimnez, L., Snchez, I. and Lpez, F. (1990). Characterization of Spanish
agricultural residues with a view to obtaining cellulose pulp. Tappi J. 73, 8, 173-176.
Kamishima, H., Matsuo, R., Fukuoka, S., Akamatsu, I., Kobayashi, Y., Husin, M.,
Miswar, O. and Haji, H. (1988). Industrial utilization of oil palm by-products. III.
Thermomechanical pulping of oil palm fronds. Cell Chem Technol. 22, 3, 351-361.
Khoo, K.C. and Lee, T.W. (1991). Pulp and paper from oil palm. Appita J. 44, 6,
385-388.
Lam, H.Q., Le Bigot, Y., Delmas, M. and Avignon, G. (2001). Formic acid pulping
of rice straw. Ind Crop Prod. 14, 65-71.
Law, K.N. and Jiang, X. (2001). Comparative papermaking properties of oil-palm
empty fruit bunches. Tappi J. 84, 1, 95.
Ligero, P., Villaverde, J.J. and Vega, A. (2010). Delignification of Miscanthus x

Giganteus by the Milox process. Bioresource Technol. 101, 3183-3193.


Ligero, P., Villaverde, J.J., Vega, A. and Bao, M. (2008). Delignification of Eucalyptus

globulus saplings in two organosolv systems (formic and acetic acid). Preliminary analysis of
dissolved lignins. Ind Crop Prod. 27, 110-117.
Malaysian Palm Oil Promotion Council, http://www.mpopc.org (acceded January
2011).
Montgomery, D.C. (1991). Design and analysis of experiments. Grupo Editorial
Iberoamericana, Mexico, 303.
More, A. and Toraldo, A. (1989). Algorithms for Bound Constrained Quadratic
Programming Problems. Numerische Mathematik 55, 377-400.
Muthurajah, R.N. and Peh, T.B. (1977). Manufacture of paper pulps from oil palm
empty bunch waste. Proceedings Malaysian International Symposium Palm Oil Process,
147-157.

215

Ferrer et al. (2011). Biorefinery of empty fruit/BioResources 6 (4) 4282-4301


Quader, M.M.A. and Lonnberg, B. (2005). Pulp and chemical cellulose obtained
with oil palm empty fruit bunches by a modified IDE process. JIRCAS Working Report 39,
47-53.
Rodrguez, A., Rosal, A. and Jimnez, L. (2010). Biorefinery of agricultural residues
by fractionation of their components. Afinidad, 67, 545, 14-19.
Roliadi, H. and Pasaribu, R.A. (2004). The possible utilization of empty oil-palm
bunches as a substitute of pulp wood for the manufacture of corrugating-medium
paperboard using semi-chemical soda-anthraquinone pulping. APPITA annual Conference
and Exhibition, 58th, 2, 335-341.

Snchez, R., Rodrguez, A., Navarro, E., Conesa, J.A. and Jimnez, L. (2010). Use of

Hesperaloe funifera for the production of paper and extraction of lignin for synthesis and
fuel gases. Biomass Bionerg. 34, 1471-1480.
Shaahriarinour, M., Wahab, M.N.A., Mustafa, S., Mohamad, R. and Ariff, A.B.
(2011). Effect of various pretreatments of oil palm empty fruit bunches fibres for
subsequent use as substrate on the performance of cellulose production by Aspergillus

terreus. BioResources 6, 1, 291-307.


Suleman, Y.H. and Yusoff, M. (1997). Ethanol-water pulping of oil palm fibers using
hydrochloric acid as catalyst and test paper characterization. Mokuzai Gakkaishi 43, 12,
1016-1021.
Wanrosli, W.D., Zainuddin, Z. and Lee, L.K. (2004). Influence of pulping variables
on the properties of Elaeis guineensis soda pulp as evaluated by response surface
methodology. Wood Sci Technol. 38, 3, 191-205.
Wanrosli, W.D. and Kwei-Nam, L. (2011). Oil palm fibers as papermaking material:
potentials and challenges. BioResources 6, 1, 901-917.
Zhan, M., Qi, W., Liu, R., Su, R., Wu, S. and He, Z. (2010). Fractionating
lignocellulose by formic acid: characterization of major components. Biomass Bioenerg. 34,
525-532.

216

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 3
MILOX FRACTIONATION OF EMPTY FRUIT BUNCHES FROM ELAEIS

GUINEENSIS
Ana Ferrer, Alberto Vega, Alejandro Rodrguez, Pablo Ligero, Luis Jimnez
Bioresource Technology, 2011, 102, 97559762
ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de Agricultural
Engineering segn el Journal Citation Reports de 2011)

217

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762

218

Publicaciones Cientficas/Scientific Publications


Bioresource Technology 102 (2011) 97559762

MILOX FRACTIONATION OF EMPTY FRUIT BUNCHES FROM ELAEIS GUINEENSIS


Ana Ferrera, Alberto Vegab, Alejandro Rodrgueza, Pablo Ligerob, Luis Jimneza
b

a Department

of Chemical Engineering, University of Crdoba, Spain


Department of Physical Chemical and Chemical Engineering, University of A Corua, Spain

Received 4 April 2011; Received in revised form 5 July 2011; Accepted 13 July 2011; Available online 28 July 2011

ABSTRACT
Milox pulping of EFB was used to obtain pulps. In the first Milox stage, the influences
of operating variables on pulp properties were studied and polynomial and neural fuzzy
models that reproduced the experimental results with errors less than 10 %, in most of the
cases, were developed. Operating variables were found (93 wt% of formic acid, 3 wt% of
hydrogen peroxide, and 165 min) that yielded acceptable pulp properties (40.5 % of yield,
50.3 % of brightness and 608 mL/g of viscosity) at reasonable chemical and energy costs.
The second stage was studied by subjecting the liquors of the previously optimized first stage
to different treatment times. This time should be 30 min or less, to avoid a negative effect on
viscosity. The residual liquor from the first Milox stage contained virtually no precipitable
lignin and only low amounts of sugars (wt%): glucose 0.71, xylose 4.22, galactose 1.19,
mannose 0.33, all on dry basis weight of original raw material.

KEYWORDS: Empty Fruit Bunches (EFB); Pulping; Milox process; Bio-refinery; Sugars.

1. INTRODUCTION
Since it has become difcult to meet the demand for pulp for paper and other used
with classic raw materials such as hard- and softwoods, alternative sources such as non-woody
materials, fast growing plants, and residues from agricultural and food and forest industries
219

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


are being explored (Jimnez et al., 2006). An interesting non-woody material is Empty Fruit
Bunches (EFB), the residual lignocellulosic material from the palm oil industry. Pulps have
been produced from EFB by alkaline processes involving the kraft method (Ibrahim, 2002;
Jimnez et al., 2009a), as well as the soda process (Law and Jiang, 2001). Daud et al. (1998)
obtained pulps with soda, sodium carbonate and sodium sulfite processes and found that the
rst process was the most efcient. Ibrahim (2002) compared the composition of EFB pulps
obtained after kraft, kraft-anthraquinone, soda and soda-anthraquinone processes, and
determined that the soda process yielded the highest viscosity and contents of lignin,
holocellulose and -cellulose.
Wanrosli et al. (2004) employed a central composite experimental design to study the
inuence of temperature, time and alkali concentration on the yield, kappa number, tensile
index and tear index of EFB pulps nding that a relatively low temperature (about 160 C)

within the limits of pulping time (60-120 min) and of alkali change between 20 % and 30 % is
generally sufcient. The rening of EFB cellulose pulps obtained by soda-anthraquinone has
been investigated using an experimental design. It was found that operating at 15 % soda,
170 C, 70 min and 2400 number of PFI rening revolutions is possible to obtain cellulose
pulps with properties that differ less than 12 % of their optimal values (59.63 Nm/g tensile

index, 4.17 KN/g burst index and 7.20 mNm2/g tear index), and a Shopper-Riegler value of

47.5 SR. These values are acceptable to obtain paper sheets (Jimnez et al., 2009a). In
order to use the EFB to obtain corrugating medium-grade paper board, Roliadi and Pasaribu
(2004) carried out laboratory scale experiments with a semi-synthetic pulp sodaanthraquinone (4-10 % soda, 0-0.15 % anthraquinone, 135 C and 3 h). Daud et al. (2005)

and Ghazali et al. (2006) have also used the EFB in the alkaline peroxide mechanical
pulping process (APMP), and found that the chemical and mechanical pretreatments
removed 95 % manganese, 45 % calcium, 37 % extractives, 81 % ash and 84 % silica initially
present in the lignocellulosic material.
Organosolv processes involving ethanol (Aziz et al., 2002; Suleman and Yusoff, 1997),
the modied IDE process (Quader and Lonnberg, 2005), or high-boiling point organic
solvents (Rodrguez et al., 2008) have also been employed, and biopulping with the white rot
fungus K14 (isolated from tropical plantation) decreased lignin content more rapidly than
with Phanerochaete chrysosporium or Colorius versicolor (Goenadi et al., 1998). Total
chlorine free (TCF) bleaching of soda-anthraquinone and diethanolamine pulps from EFB
was studied by Jimnez et al. (2009b) who found that diethanolamine pulp produced better
220

Publicaciones Cientficas/Scientific Publications


paper than soda-anthraquinone pulp when an OZP (oxygen - ozone - peroxide) bleaching
sequence was employed.
Organosolv methods for pulping of lignocellulosic materials are an attractive alternative
to the kraft process (Lpez et al, 2006; Muurinen, 2000; Rodrguez and Jimnez, 2008) since
they lead to fewer environmental problems. Many organosolv methods can be carried out at
lower temperatures than the kraft process and therefore are less energy-consuming.
Organosolv pulping methods, when performed under appropriate conditions, allow a high
rate of raw material recovery through the separation of lignin from hemicellulose and
cellulose and can therefore also be considered as fractionation processes (Huijgen et al.,
2010; Pan et al., 2010; Villaverde et al., 2009).
A large number of organic compounds have been tested to achieve delignication of
plant materials including alcohols, amines, ketones, phenols, and aldehydes, but also formic
and acetic acids (Lpez et al., 2006; Mkinen and Lepplahti, 2009; Muurinen, 2000;
Rodrguez and Jimnez, 2008).
When hydrogen peroxide is added to formic acid solutions, peroxyformic acid is
formed (H2O2 + HCOOH HCOOOH + H2O). Peroxyformic acid is a very strong

oxidant and one of the main active reagents in Milox organosolv pulping; however, its

stability is low, as it easily decomposes through a series of acid catalyzed reactions (De
Filippis et al., 2009). It must therefore be prepared in situ for the delignication to be able to
compete successfully with the decomposition acidic reactions, such as irreversible acidcatalyzed decomposition of peroxyformic acid:
HCOOOH CO2 + H2O

2HCOOOH 2HCOOH + O2

HCOOOH + H2O2 HCOOH + H2O + O2

and the acid-catalyzed decomposition of hydrogen peroxide.


2H2O2 2H2O + O2

Milox pulping consists of two or three stages of alternating formic and peroxyformic
acid treatments. Usually, a peroxyformic stage is used at the beginning, at temperatures
ranging from 60 to 80 C. In the second stage, the temperature is raised and the reaction

continues at total reux without additional peroxide. In the third stage, more hydrogen
peroxide is added to the liquor and the temperature is lowered to levels similar to those of
221

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


the rst stage. Milox treatment has been successfully used with a variety of raw materials,
including agricultural plants (Ligero et al., 2010; Seisto and Poppius, 1995), beech (Dapa et
al., 2000), birch (Abad et al., 2000a) and eucalyptus (Abad et al., 2000b).
Milox delignication is based on the electrophilic attack on lignin by the hydroxonium
ion (HO+) which results in the cleavage of barylether bonds, oxidative demethylation, ring

opening, epoxidation, ring hydroxylation and epoxidation (Gierer, 1982). In addition to

lignin, most of the extractives and hemicelluloses are dissolved. Since cellulose bers can be
damaged, it is necessary to nd conditions that allow maximum delignication with
minimum damage to cellulose.
Pulping processes have been simulated extensively using polynomial models, even in
the case of EFB soda-anthraquinone (Jimnez et al., 2009a). Neural fuzzy models have also
been applied to the pulping of some materials, but not EFB. Fuzzy modeling, based on the
pioneering idea of Zadeh (1965), has been successfully applied to the simulation and control
of fermentation processes and anaerobic digestion systems (Tay and Zhang, 1999). Another
powerful tool for modeling complex nonlinear systems is neural network (NN) technology
(Works, 1989). The integration of fuzzy systems and NN can combine the merits of both
systems and offer a more powerful tool for modeling.
The current study investigated the Milox fractionation of EFB in order to obtain
cellulose enriched pulp and a liquor containing the dissolved lignin and degradation
products of the hemicelluloses. The inuences of operating variables in the rst stage of
Milox process were studied by simulating the process through the polynomial and neural
fuzzy models, in order to nd optimal operating conditions to obtain the least degraded
cellulose pulp. Subsequently, the selected pulp was subjected to a second stage of
delignication and the residual liquor was characterized by determining its sugar content.

2. EXPERIMENTAL
2.1. Raw material
EFB from a Malaysian oil palm plant was supplied by Straw Pulping Engineering S.L.
(Zaragoza, Spain). The chemical properties of EFB were determined in accordance with the
respective TAPPI standards for the different components, namely: T-222 for lignin, T-203
OS-61 for -cellulose, T-9m-54 for holocellulose, T-204 for ethanol-benzene extractives and

222

Publicaciones Cientficas/Scientific Publications


T-211 for ash. The analysis shows the following results: 67.0 % of holocellulose, 41.9 % of cellulose, 24.5 % of lignin, 1.2 % of extractives and 3.2 % of ash.
Fiber length was determined under a Visopan projection microscope with 10x
objective for 100x magnication, after microcooking the raw material with 10 % soda at 80
C for 1 h and subsequent staining the bers with 1 % safranin (Snchez et al., 2010). It is

found to be 0.53 mm, and the thickness or width 14 m.

2.2. Pulping
Milox fractionations were performed in 250 cm3 round bottom asks at atmospheric

pressure. In the rst stage selected quantities of EFB, formic acid and water were mixed and
heated to 80 C. Once the target temperature was reached, calculated amounts of hydrogen

peroxide were added to the liquors (zero time) and the reactions were maintained at constant
temperature. Selected reacting mixtures from the rst stage were submitted to a second stage
by raising the temperature to its normal boiling point and holding for different times. The
pulps were ltered, washed four times with 85 % of formic acid in proportions of 2.5, 2.5,
5.0 and 5.0 (mL/g) with respect to the initial dry weight of EFB, and repeatedly with distilled
water until the washings were neutral.

2.3. Pulp characterization


Pulp yield was measured after oven drying of a pulp aliquot to constant weight. The
kappa number and intrinsic viscosity were determined according to TAPPI T236 and
TAPPI T230, respectively. To measure ISO brightness, handsheets from pulps were formed
according to standard TAPPI T272 and diffuse blue reectance was measured according to
TAPPI T525 in a Photovolt 577 reectometer.

2.4. Characterization of liquor


The liquor from the best fractionation assay was diluted with water in a volumetric
ratio of 7:1 to precipitate lignin and to obtain samples for chromatographic analyses.
Monosaccharides were analyzed after neutralization and ltration through a 0.20 m
membranes on a Waters 1500 Series HPLC equipped with a Biorad Aminex HPX-87P
column thermostated to 85 C (refractive index detector at 30 C). Furfural was detected

223

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


with a UV/VIS wavelength detector at 254 nm. HPLC-grade water was used as mobile phase
at a ow of 0.6 mL/min.

2.5. Experimental design


To quantify the effects of variables a 2n factorial design (Montgomery, 1991) was used,

consisting of a central experiment (in the center of a cube) and several additional points
(additional experiments lying at the cube vertices and side centers).

2.5.1. Polynomial modeling


Experimental data were tted to the following second-order polynomial:
Ye= a0 + a1XF + a2XP + a3XT + a12XF XP + a13XF XT + a23XP XT + a11XF2 + a22XP2 + a33XT2

(1)

where Ye denotes the response variables [viz. yield (YI), kappa number (KN), viscosity (VI)
or brightness (BR)]; XF, XP and XT are the normalized values of the operational variables

(formic acid concentration (F), hydrogen peroxide concentration (P) and processing time
(T), respectively); and a0 to a33 are constants.

The values of the operational variables were normalized to values from -1 to +1 by

using the following equation:


Xn =

)
2(X-X

X max -X min

(2)

where Xn is the normalized value of F, P or T; X is the actual experimental value of the


variable concerned;
X is the mean of Xmax and Xmin; and Xmax and Xmin are the maximum
and minimum value, respectively, of such a variable. The normalized values for the

independent variables in the 15 experiments conducted are given in Table 1.

2.5.2. Neural fuzzy modeling


The integration of fuzzy systems and neural networks combines the advantages of the
two systems and provides an especially powerful modeling tool, viz. the neural fuzzy system,
which uses neural networks as tools in fuzzy systems (Jang et al., 1997).
The variation of pulp properties as a function of the operational variables of the
bleaching process can be predicted by using the following expression (Jang et al., 1997):
224

Publicaciones Cientficas/Scientific Publications


lm 1 =
y l [in 1 Fil ( xi , il )
y=
=
e
lm 1[=
in 1 Fil ( xi , il )
=

(3)

ye is the estimated value of the property to be modeled (output variable), m the number of

fuzzy rules, n the number of independent variables, yl the defuzzier of the fuzzy rule and
Fil(xi,il) the membership-function of the independent variables within its range.

In this paper, the term =1 Fil(xi,il) of Eq. (3) will be denoted by Rl which is known

as a fuzzy rule. In this sense, Rl is dened by the product of n membership functions (one

per independent variable). Then, Eq. (3) could be expressed as:


Ye=

'
m
l=1 Y R l

m
l=1 R l

(4)

With three operational variables -our case- the use of a singleton defuzzifier (a constant
parameter: ci) and a linear membership function for the independent variables allowed Eq.
(4) to be simplied to:

Ye=

8l=1 Cl R l
8l=1 R l

(5)

Linear membership functions are selected for independent variables with two levels.
The mathematical equations which respond to linear membership functions are:
(X-X low )

(6)

(X-X high )

(7)

X1 =Xlow = 1- X
X2 =Xhigh = 1-

high - Xlow

X high - X low

X1 and X2 denoting the values of F, P or T, for low and high levels; Xhigh and Xlow the

extreme values of the variable; and X the absolute value of F, P or T.

With three independent variables, the following eight fuzzy rules can be established

according to the extreme (high and low) values of such variables:


R1: low F, low P and low T = F1 P1 T1

R2: low F, low P and high T = F1 P1 T2


...

R7: low F, high P and high T = F1 P2 T2

R8: high F, high P and high T = F2 P2 T2

225

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


With a Gaussian membership function with three levels (low, medium and high) for
one of the variables and a linear membership function with two levels (low and high) for the
other two, Eq. (5) would include 12 terms in the numerator and 12 in the denominator. The
Gaussian membership function would be of the form:
Xi = exp [-0.5 (

X-Xc 2
) ]
L

(8)

where Xi denotes the values of F, P and T, for low, medium and high levels; X is the absolute
value of the variable concerned; Xc its minimum, central or maximum value; and L the width

of its Gaussian distribution.

The parameters and constants in the previous equation were estimated by using the
ANFIS (Adaptative Neural Fuzzy Inference System) Edit tool in the Matlab 6.5 software
suite.

3. RESULTS AND DISCUSSION


3.1. First-stage MILOX process
Based on the results of other researchers for different raw materials (Ligero et al.,
2010) and in order to limit the intervals of operating variables for the first-stage of the Milox
process, a series of preliminary experiments were carried out and the following ranges were
chosen: formic acid concentration, 80-95 wt%; hydrogen peroxide concentration, 2-4 wt%,
and reaction time, 60-180 min. The liquid/solid ratio was xed at the value 10:1 by weight.
Table 1 shows the experimental values of the pulp properties, which differed by less
than 5-10 % from their means as obtained in triplicate measurements.

3.1.1. Polynomial modeling


The experimental results were tted to polynomial models by multiple regression
using the software BMDP. The terms possessing a Snedecor F-value greater than 2.5 and a
Student t-value greater than 1.7 were deemed statistically signicant. Equations obtained, as

well as the highest p and the lowest Student t-values for the terms, are the following:
YI= 42.6 1.8XF 1.5XP 2.1XT 1.4XF XT + 2.8XP2 (p<0.13; t>1.7)
KN= 9.4 2.2XF 5.5XP 3.0XT + 2.5XP XT + 4.1XP2
226

(p<0.04; t>2.3)

(9)
(10)

Publicaciones Cientficas/Scientific Publications


VI= 579 10XP + 13XT + 20 XF2 13XF XP + 15XF XT (p<0.03; t>2.6)

(11)

BR= 46.6 + 2.2XF + 6.4XP + 2.8 XT (p<0.07; t>2.1)

(12)

Table 1. Values of operational variables and experimental values of the properties of pulp obtained
by first stage of Milox pulping process of EFB.
Normalized values of operational
variables

Experiment

Yield, %

Kappa

Viscosity,

Brightness,

number

mL/g

XF

XP

XT

-1

-1

-1

56.6

31.5

601

32.0

+1

-1

-1

45.3

19.4

585

38.7

-1

+1

-1

46.0

10.3

605

47.0

+1

+1

-1

43.4

7.0

549

54.5

-1

-1

+1

45.0

16.4

596

41.2

+1

-1

+1

42.7

11.8

653

44.4

-1

+1

+1

42.0

8.3

605

49.1

+1

+1

+1

41.7

6.3

599

57.9

-1

42.9

8.2

580

51.2

10

+1

41.4

8.7

613

47.4

11

-1

44.8

15.3

600

40.3

12

+1

46.4

8.1

577

51.7

13

-1

43.5

13.6

571

41.7

14

+1

42.1

9.0

587

49.2

15

43.2

7.4

560

52.9

16

-0.5

-0.5

-0.5

46.7

14.5

588

40.5

17

+0.5

+0.5

+0.5

41.7

5.9

584

51.3

The previous equations predict the experimental results of the dependent variables
with errors less than 8 % for the yield, 3 % for the viscosity and 13 % for the brightness
(Table 4). In contrast, for the kappa number, the calculated equation reproduces the
experimental results with errors less than 15 % only in the 47 % of the cases. The models
were validated by conducting two additional pulping experiments (entries 16 and 17 in Table
1). The errors in predicting pulp properties by using the polynomial models were quite small
(Table 4), and thus confirms the accuracy of the models.

227

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


The values of the operational variables providing the best pulp properties (yield,
viscosity and brightness) were identified by using a multiple non-linear programming
method. Table 2 shows the optimum values of the dependent variables and those of the
operational variables required to obtain them.
The polynomial equations for yield, viscosity and brightness of the pulps allow
identifying the operational variables that more markedly influencing the pulp properties. The
maximum variations in the dependent variables with changes in the operational variables
over the studied range were obtained by altering one independent variable at a time while
keeping all others constant (with their optimum values, Table 2). The results are shown in
Table 2 together with the maximum percent differences in the dependent variables from
their optimum values over the studied variation ranges. Figs. 1-4 also reflect the changes
presented in Table 2.
As shown in Table 2 and Fig. 1, the processing time had a greater influence on pulp
yield than the formic acid concentration. Similarly, the hydrogen peroxide concentration had
less influence than formic acid concentration (Table 2, Fig. 2 and Eq. (9)). This is explained
by the different magnitude of the coefcients of the yield equation.
A similar explanation can be given for the remaining dependent variables. The
viscosity also varied more with processing time (Table 2, Fig. 3 and Eq. (11)) and less with
the hydrogen peroxide concentration (Table 2, Fig. 4 and Eq. (11)); the brightness was more
inuenced by the hydrogen peroxide concentration than by the processing time, and the
formic acid concentration was the least inuential variable (Table 2 and Eq. (12)).
The Yield/Kappa ratio is desirable to have a maximum value. This is achieved when
the raw material has a high yield and/or when the pulp obtained has a high brightness (low
kappa number). By tting the experimental data of Yield/Kappa ratio (Table 1) to a
polynomial model gives the equation:
Yield
=4.90 + 0.43 XF + 1.47 XP + 0.47 XT 0.83 XT2
Kappa
(p < 0.05; t-Student > 2.3)

(13)

The values of the operational variables providing the maximum Yield/Kappa ratio
were identified by using a multiple non-linear programming method. The maximum value
(6.87) is achieved when the formic acid concentration is the highest studied (95 %), hydrogen
peroxide concentration is the highest studied (4 %) and the processing time is 137 min. Fig. 5
228

Publicaciones Cientficas/Scientific Publications


shows the change of the Yield/ Kappa ratio with hydrogen peroxide concentration and
processing time, for a 95 % formic acid concentration.
Table 2. Optimal values of properties in the pulp obtained after the rst stage of the EFB Milox
processing, and maximum changes in the dependent variables with changes in one operational
variable maintaining others constant (the percent differences from the changes are given in brackets).
Dependent variable
Yield

Viscosity

Brightness

Optimum (maximum) value of the dependent variable

52.5 %

650 mL/g

58.0 %

Normalized value of formic required to obtain the optimum value


of dependent variables

-1

+1

+1

Normalized value of peroxide required to obtain the optimum value


of dependent variables

-1

-1

+1

Normalized value of time required to obtain the optimum value of


dependent variables

-1

+1

+1

Variation of formic with the operational variable

6.4(12.3 %)

58(8.9 %)

4.4(7.6 %)

Variation of peroxide with the operational variable

4.5(8.6 %)

46(7.0 %)

12.8(5.6 %)

Variation of time with the operational variable

7.0(13.4 %)

56(9.0 %)

5.6(9.7 %)

Fig. 1. Pulp yields in response to processing time and formic acid concentration, at low hydrogen
peroxide concentration (XP = -1).

229

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762

Fig. 2. Pulp yields in response to formic acid and hydrogen peroxide concentrations at short
processing time (XT = -1).

Fig. 3. Viscosities in response to processing time and formic acid concentration at low hydrogen
peroxide concentration (XP = -1).

Fig. 4. Viscosities in response to formic acid concentration and hydrogen peroxide concentration at
long processing time (XT = +1).
230

Publicaciones Cientficas/Scientific Publications

Fig. 5. Yield/Kappa ratio in response to hydrogen peroxide concentration and processing time at
high formic acid concentration (XF = +1).

The Viscosity/Kappa ratio is also desirable to have a high value. Therefore, the
fibers have a high degree of polymerization (high viscosity) and/or a high brightness (low
kappa number). By fitting the experimental data of Viscosity/Kappa ratio (Table 1) to a
polynomial model gives the equation:
Viscosity
=65.46 + 7.17 XF + 19.63 XP + 9.65 XT 10.14 XT2
Kappa
(p < 0.04; t-Student > 2.4)

(14)

Operating in a similar way as Eq. (13) and Eq. (14), gives a maximum value of the
Viscosity/Kappa ratio of 94.56, for operating conditions close to those found in the case of
the Yield/Kappa ratio, high concentrations of formic acid and hydrogen peroxide and 149
min. Fig. 6 shows the changes in the Viscosity/Kappa ratio by varying the hydrogen
peroxide concentration and the processing time, for a 95 % formic acid concentration.

Fig. 6. Viscosity/Kappa ratio in response to hydrogen peroxide concentration and processing time
at high formic acid concentration (XF = +1).

231

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


The data in Table 2 can be used to select values of the operational variables providing
near-optimal pulp properties while saving chemicals, energy and investment for industrial
installation by using lower values of operational variables. One combination leading to nearoptimal properties with reduced costs is to use a processing time of 165 min, a formic acid
concentration of 93 % and a hydrogen peroxide concentration of 3 %. Operating under
these conditions the following values for the dependent variables were obtained: yield 40.5
%, brightness 50.3 %, and viscosity 608 mL/g; these values deviate by 22.7 %, 13.3 % and 6.5
%, respectively from the maximum values of yield, brightness and viscosity, respectively. It is
found that when operating under these conditions, Yield/Kappa and Viscosity/Kappa
ratios of 5.10 and 72.23 are achieved, respectively. These values are quite different from the
top surfaces presented in Figs. 5 and 6.

3.1.2. Neural fuzzy modeling


Table 3 shows the values of the neural fuzzy modeling constants, ci, as obtained by

using the yield, kappa number, viscosity, and brightness data of Table 1.

Table 3. Values of the constants ci in the neural fuzzy model for the pulp properties.
Operational variables

Rule
1
2
3
4
5
6
7
8
9
10
11
12
9
10
11
12
R2
232

XF

XP

XT

Yield, %

-1
-1
-1
+1
-1
+1
+1
+1
-1
-1
+1
+1
-1
-1
+1
+1

-1
-1
+1
-1
+1
-1
+1
+1
0
0
0
0
-1
+1
-1
+1

-1
+1
-1
-1
+1
+1
-1
+1
-1
+1
-1
+1
0
0
0
0

44.4
43.0
46.6
44.7
42.5
42.0
44.0
42.2
43.0
42.0
41.7
41.3

0.93

Dependent variables
Kappa
Viscosity,
number
mL/g
32.2
603
16.5
597
9.9
606
1.9
585
8.2
606
1.1
655
6.3
546
6.0
597
557
567
591
609
12.0
5.0
1.3
6.6
0.54

0.76

Brightness,
%
30.4
40.8
45.9
37.8
48.7
44.5
54.2
58.4

45.8
57.1
40.5
52.0
0.94

Publicaciones Cientficas/Scientific Publications


In the cases of yield and viscosity, linear membership functions at two levels (low and
high) have been used for the variables, formic acid concentration and processing time and a
Gaussian membership function at three levels (low, medium and high) for the hydrogen
peroxide concentration variable. For kappa number and brightness, a Gaussian membership
function for the variable processing time and a linear membership function for the other two
variables have been used. Table 4 gives the estimated values of the dependent variables as
calculated with the models, as well as their deviations from the experimental values for the
yield, viscosity and brightness. The differences were all small (9 %, 6 % and 9 % yield,
viscosity and brightness, respectively), so the neural fuzzy modeling method provided
accurate predictions for the pulping process of EFB with the Milox process. However, in the
case of the kappa number, as happened with the polynomial model, the neural fuzzy model
does not adequately reproduce the experimental values, since only 67 % of the cases have
errors less than 15 %.
Table 4. Values of the dependent variables as estimated with the polynomial and neural fuzzy models
and deviations from their experimental counterparts (in brackets).
Exp.

Yield, %

Viscosity, mL/g

Brightness, %

Polynomial

Neural fuzzy

Polynomial

Neural fuzzy

Polynomial

Neural fuzzy

52.2 (7.77)

61.6 (8.83)

698 (0.50)

603 (0.33)

35.2 (10.00)

30.4 (5.00)

45.8 (1.10)

45.9 (1.32)

593 (1.37)

551 (5.81)

39.7 (2.58)

37.8 (2.33)

49.2 (6.96)

48.5 (5.43)

604 (0.17)

604 (0.17)

47.9 (1.91)

45.9 (2.34)

42.8 (1.38)

43.9 (1.15)

547 (0.36)

548 (0.18)

52.4 (3.85)

54.1 (0.73)

45.1 (0.22)

44.3 (1.56)

593 (0.50)

596 (0.00)

40.8 (0.97)

41.1 (0.24)

44.4 (3.98)

42.0 (1.64)

650 (0.46)

653 (0.00)

45.3 (2.03)

44.2 (0.45)

42.2 (0.48)

42.5 (1.19)

599 (0.99)

604 (0.17)

53.5 (8.96)

49.2 (0.20)

41.4 (0.72)

42.2 (1.20)

604 (0.83)

598 (0.17)

58.0 (0.17)

58.0 (0.17)

44.4 (3.50)

43.2 (0.70)

599 (3.28)

567 (2.24)

44.4 (13.28)

50.3 (1.76)

10

40.8 (1.45)

41.7 (0.72)

599 (2.28)

600 (2.12)

48.9 (3.16)

46.5 (1.90)

11

46.9 (4.69)

46.7 (4.24)

589 (1.83)

609 (1.50)

40.3 (0.00)

42.6 (5.71)

12

43.9 (5.39)

43.8 (5.60)

569 (1.39)

588 (1.91)

53.0 (2.51)

54.2 (4.84)

13

44.8 (2.99)

43.2 (0.69)

566 (0.88)

576 (0.88)

43.8 (5.04)

42.5 (1.92)

14

40.5 (3.80)

41.8 (0.71)

592 (0.85)

592 (0.85)

49.4 (0.41)

48.1 (2.24)

15

42.6 (1.39)

42.5 (1.62)

579 (3.39)

584 (4.29)

46.6 (11.91)

48.4 (8.51)

16

46.4 (0.64)

47.2 (1.07)

583 (0.85)

587 (0.17)

40.9 (0.99)

41.7 (2.96)

17

41.0 (1.68)

42.3 (1.44)

586 (0.34)

592 (1.37)

52.3 (1.95)

51.6 (0.58)

233

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


The proposed models were validated by conducting two pulping experiments (Table
1). The values of the dependent variables provided by the neural fuzzy models used and the
errors in the predictions are shown in Table 4. The predicted values deviate slightly from
their experimental counterparts, which conrm the accuracy of the neural fuzzy model used.
By using the polynomial and neural fuzzy models, the errors in the predictions of the
experimental results of yield, viscosity and brightness were low and very similar, so both
models are equally good.
The neural fuzzy models used involve roughly the same number of parameters as
second-order polynomial models. This allows simple neural fuzzy models to be employed
with experiments involving only 15 tests - the typical number in polynomial models.

3.2. Second-stage MILOX process


With the operating conditions previously dened for the rst stage (formic acid
concentration 95 %, hydrogen peroxide concentration 4 % and processing time 137-149
min) maximum values of the ratios Yield/Kappa and Viscosity/Kappa were obtained, but
the viscosity of pulp was lower than the predicted maximum (650 mL/g) by roughly 10 %, in
both cases. This maximum corresponds to the highest levels of formic acid concentration
and treatment time, and to the lowest level of peroxide. Thus, in order to preserve the pulp
viscosity as much as possible, the second stage was performed at high levels of time (180
min) and formic acid concentration (95 %), as predicted by the models, but limiting the
hydrogen peroxide load, which had a negative effect on viscosity, to its lowest level (2 %).
Table 5. Pulp properties after different treatment times on the second stage Milox pulping of EFB.
Time, min

Yield, %

Kappa number

Viscosity, mL/g

Brightness, %

42.0

8.3

605

40.1

30

42.4

8.6

530

46.9

60

41.5

5.4

464

49.5

90

41.1

5.0

438

53.4

120

40.6

4.6

425

53.8

First stage operating conditions: formic acid 95 %, hydrogen peroxide 2 % and processing time 180
min.

234

Publicaciones Cientficas/Scientific Publications


Since the second stage is operated by simply increasing the reaction temperature, its
behavior is reduced to monitoring the process variables with the treatment time. Therefore,
four cookings were conducted under identical conditions to those mentioned in the
preceding paragraph (rst stage). Each one of these reaction mixtures was then submitted to
a second stage at different durations: 30, 60, 90 and 120 min.
Table 5 shows that the second Milox stage severely affected the pulp viscosity while the
reduction in kappa number and the increase in brightness obtained did not justify treatments
longer than 30 min. Even a single stage might be appropriate since it produced a pulp with a
kappa number value low enough to begin a bleaching sequence, while viscosity was
signicantly higher.

3.3. Characterization of residual liquor


The residual liquor from the rst stage of experiment 6 (which yielded a pulp with
high viscosity) was considered in order to attempt to precipitate the dissolved lignin. Dilution
with different proportions of water did not produce a precipitate.
This can be explained by the high oxidizing power of liquor that would have produced
not only a simple delignication, but an oxidative depolymerization, which would have
degraded the lignin oligomers to polar, highly oxidized water soluble compounds.
As shown in Table 6, the monosaccharides content in solution was very low, suggesting
degradative processes, beyond the dehydration to furfural and hydroxymethylfurfural, since
these two compounds were very minor in black liquors.
Table 6. Characterization of liquor.
mg/L in liquor

% Based on original dry raw material

Glucose

290.7

0.71

Xylose

1720.9

4.22

Galactose

484.4

1.19

Mannose

133.1

0.33

Furfural

332.7

0.82

Cooking experimental variables: L/S = 10; formic acid percentage = 95; hydrogen peroxide
percentage = 2. 1st stage: 180 min, 80 C. 2nd stage: 30 min.
235

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


4. CONCLUSIONS
Polynomial and neural fuzzy models predict the pulp properties in the rst stage Milox
pulping with errors less than 10 % in most of the cases. EFB pulping with 93 % formic acid,
3 % peroxide for 165 min to save energy and chemicals and results in a pulp with acceptable
properties.
Second stage Milox pulping must be very short (30 min or less) to avoid a negative
effect on viscosity.
The black liquor from Milox EFB cooking contained no precipitable lignin and low
amounts of fermentable monosaccharides, probably due to the high oxidative power of the
reaction mixture. The optimization of the process yielded a pulp with excellent kappa
number, but if precipitable lignin has to be used as well, a different approach should be
implemented based on the utilization of lower hydrogen peroxide concentrations.

ACKNOWLEDGEMENTS
The authors are grateful to Spains DGICyT for funding this research within the
framework of the Projects TRA2009-0064 and CTQ2010-19844-C02-01.

REFERENCES
Abad, S., Santos, V., Paraj, J.C., 2000a. Formic acid-peroxyformic acid pulping of
aspen wood: an optimization study. Holzforschung 54, 544-552.
Abad, S., Santos, V., Paraj, J.C., 2000b. Evaluation of Eucalyptus globulus wood
processing in media made up of formic acid, water, and hydrogen peroxide for dissolving
pulp production. Ind. Eng. Chem. Res. 40, 413-419.
Aziz, A.A., Husin, M., Mokhtar, A., 2002. Preparation of cellulose from oil palm
empty fruit bunches via ethanol digestion: effect of acid and alkali catalysts. J. Oil Palm Res.
14 (1), 9-14.
Dapa, S., Santos, V., Paraj, J.C., 2000. Formic acid-peroxyformic acid pulping of

Fagus sylvatica. J. Wood Chem. Technol. 20, 395-413.


Daud, W.R.W., Law, K.N., Valade, J.L., 1998. Chemical pulping of oil palm empty
fruit bunches. Cell Chem. Technol. 32 (1-2), 133-143.
236

Publicaciones Cientficas/Scientific Publications


Daud, W.R.W., Law, K.N., Ghazali, A., 2005. Alkaline peroxide mechanical pulping
of oil palm lignocellulosics. Part 1. Effects of chemical and mechanical pretreatments on
empty fruit bunches (EFB) chemical composition. JIRCAS Working Rep. 39, 82-85.
De Filippis, P., Scarsella, M., Verdone, N., 2009. Peroxyformic acid formation: a
kinetic study. Ind. Eng. Chem. Res. 48, 1372-1375.
Ghazali, A., Wanrosli, W.D., Law, K.N., 2006. Alkaline peroxide mechanical pulping
(APMP) of oil palm lignocellulosics. Part 2. Empty fruit bunches (EFB) responses to
pretreatments. Appita J. 59 (1), 65-70.
Gierer, J., 1982. The chemistry of delignication. A general concept. Part 2.
Holzforschung 36, 43-51.
Goenadi, D.H., Away, Y., Suharyanto, P., Panji, T., Watanabe, T., Kuwahara, M.,
1998. Bio-pulping of empty fruit bunches of oil palm by white rot fungi isolated from
tropical plantation. International Conference on Biotechnology in the Pulp and Paper
Industry, 7th, Vancouver, BC, June 16-19, B, pp. 49-51.

Huijgen, W.J.J., Reith, J.H., den Uil, H., 2010. Pretreatment and fractionation of

wheat straw by an acetone-based organosolv process. Ind. Eng. Chem. Res. 49, 1013210140.
Ibrahim, R., 2002. Chemical composition of alkaline pulps from oil palm empty fruit
bunches. Oil Palm Bull. 44, 19-24.
Jang, J.S.R., Sun, C.T., Mizutani, E. 1997. Neuro-fuzzy and Soft Computing: A
Computational Approach to Learning and Machine Intelligence. Prentice Hall.
Jimnez, L., Prez, A., Rodrguez, A., de la Torre, M.J., 2006. New raw materials and
pulping processes for production of pulp and paper. Anidad 63 (525), 362-369.
Jimnez, L., Serrano, L., Rodrguez, A., Ferrer, A., 2009a. TCF bleaching of sodaanthraquinone and diethanolamine pulp from oil palm empty fruit bunches. Bioresour.
Technol. 100, 1478-1481.
Jimnez, L., Serrano, L., Rodrguez, A., Snchez, R., 2009b. Soda-anthraquinone
pulping of palm oil empty fruit bunches and beating of the resulting pulp. Bioresour.
Technol. 100, 1262-1267.
Law, K.N., Jiang, X., 2001. Comparative papermaking properties of oil-palm empty
fruit bunch. Tappi J. 84 (1), 95.
237

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762


Ligero, P., Vega, A., Villaverde, J.J., 2010. Delignication of Miscanthus x giganteus by
the Milox process. Bioresour. Technol. 101, 3188-3193.
Lpez, F., Alfaro, A., Jimnez, L., Rodrguez, A., 2006. Alcohols as organic solvents
for obtainment of cellulose pulp. Anidad 63 (523), 174-182.
Mkinen, T., Lepplahti, J., 2009. Review of the Finnish BioRene - new biomass
products program. Environ. Prog. Sustainable Energy 28, 470-474.
Montgomery, D.C., 1991. Design and Analysis of Experiments. Grupo Editorial
Iberoamericana, Mexico, 303.
Muurinen, E., 2000. Organosolv Pulping. A Review and Distillation Study Related to
Peroxyacid

Pulping.

Oulu,

Finland.

<http://herkules.oulu./isbn9514256611/

isbn9514256611.pdf>
Pan, X., Gilkes, N., Kadla, J., Pye, K., Saka, S., Gregg, D., Ehara, K., Xie, D., Lam,
D., Saddle, J., 2010. Bioconversion of hybrid poplar to ethanol and co-products using an
organosolv fractionation process: optimization of process yields. Biotechnol. Bioeng. 94,
851-861.
Quader, M.M.A., Lonnberg, B., 2005. Pulp and chemical cellulose obtained with oil
palm empty fruit bunches by a modied IDE process. JIRCAS Working Rep. 39, 47-53.
Rodrguez, A., Jimnez, L., 2008. Pulping with organic solvents other than alcohols.
Anidad 65 (535), 188-196.
Rodrguez, A., Serrano, L., Moral, A., Prez, A., Jimnez, L., 2008. Use of highboiling point organic solvents for pulping oil palm empty fruit bunches. Bioresour. Technol.
99, 1743-1749.
Roliadi, H., Pasaribu, R.A., 2004. The possible utilization of empty oil-palm bunches
as a substitute of pulp wood for the manufacture of corrugating-medium paperboard using
semi-chemical soda-anthraquinone pulping. APPITA annual Conference and Exhibition,
58th, vol. 2, pp. 335-341.

Snchez, R., Rodrguez, A., Requejo, A., Ferrer, A., Navarro, E., 2010. Soda pulp and

fuel gases synthesis from Hesperaloe funifera. Bioresour. Technol. 101, 7032-7040.

238

Publicaciones Cientficas/Scientific Publications


Seisto, A., Poppius, K., 1995. Milox pulping of agricultural plants. In: Proceedings of
8th International Symposium on Wood and Pulping Chemistry, June 6-9, Helsinki, Finland,

vol. 1, pp. 407-414.

Suleman, Y.H., Yusoff, M., 1997. Ethanol-water pulping of oil palm bers using
hydrochloric acid as catalyst and test paper characterization. Mokuzai Gakkaishi 43 (12),
1016-1021.
Tay, J.H., Zhang, X., 1999. Neural fuzzy modeling of anaerobic biological wastewater
treatment systems. J. Environ. Eng. 125 (12), 1149-1154.
Villaverde, J.J., Ligero, P., Vega, A., 2009. Formic and acetic acid as agents for a
cleaner fractionation of Miscanthus x giganteus. J. Clean Prod. 18, 395-401.
Wanrosli, W.D., Zainuddin, Z., Lee, L.K., 2004. Inuence of pulping variables on the
properties of Elaeis guineensis soda pulp as evaluated by response surface methodology.
Wood Sci. Technol. 38 (3), 191-205.
Works, G.A., 1989. Neural network basics. In: Proc. AUTOFACT 89, pp. 29-1-29-9.
Zadeh, L.A., 1965. Fuzzy sets. Inf. Control 8, 338-353.

239

A. Ferrer et al./Bioresource Technology 102 (2011) 9755-9762

240

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 4
TCF BLEACHING OF SODA-ANTHRAQUINONE AND DIETHANOLAMINE
PULP FROM OIL PALM EMPTY FRUIT BUNCHES
L. Jimnez, L. Serrano, A. Rodrguez, A. Ferrer
Bioresource Technology, 2009, 100, 14781481
ndice de impacto: 4.253 (Primera posicin de 11 revistas en el rea de Agricultural
Engineering segn el Journal Citation Reports de 2009)

241

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481

242

Publicaciones Cientficas/Scientific Publications


Bioresource Technology 100 (2009) 14781481

TCF BLEACHING OF SODA-ANTHRAQUINONE AND DIETHANOLAMINE PULP


FROM OIL PALM EMPTY FRUIT BUNCHES
L. Jimnez, L. Serrano, A. Rodrguez, A. Ferrer

Department of Chemical Engineering, University of Crdoba, Spain

Received 11 June 2008; Received in revised form 6 August 2008; Accepted 7 August 2008;
Available online 21 September 2008
Short Communication

ABSTRACT
The AOpAZRP bleaching sequence (A, an acid treatment; Op, an oxygen and
peroxide stage; Z, an ozone stage; R, a reductive treatment and P, a peroxide stage) has been
applied to oil palm Empty Fruit Bunches (EFB) soda-anthraquinone and diethanolamine
pulps. On similar kappa numbers for the two types of pulp (14.2 and 17.3), paper sheets
from unbleached soda-anthraquinone pulp exhibited higher values in tensile index (25.8
Nm/g), stretch (2.35 %), burst index (1.69 KN/g), tear index (0.50 mNm2/g) and brightness

(60.6 %) than those obtained for the paper sheets from unbleached diethanolamine pulp;

but the latter type of pulp exhibited higher viscosity (659 mL/g) than the other one. Upon
bleaching with the AOpAZRP sequence, diethanolamine pulp exhibited a higher viscosity
(783 mL/g), and the properties of the paper sheets were close to or even better than those
from the soda-anthraquinone pulp, namely: 22.2 vs. 20.4 Nm/g in tensile index, 1.30 vs.
1.42 KN/g in burst index, 0.71 vs. 0.70 mNm2/g in tear index and 71.3 % vs. 77.5 % in

brightness. Therefore, the properties of paper sheets from diethanolamine pulp evolved
more favorably during bleaching than those of paper sheets from soda-anthraquinone pulp.

KEYWORDS: Empty Fruit Bunches (EFB); Soda-anthraquinone; Diethanolamine; Pulping;


TCF bleaching.

243

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481


1. INTRODUCTION
Paper consumption has grown substantially over the past few decades; thus, the gure
for 2005, world consumption of paper was 352 million tons, a 31.8 % higher than that for
1990 (Annual FAO, 2005). Consumption of raw material pulping has also grown in parallel.
Paper continues to be obtained largely (90-92 %) from hardwood or softwood and, to a
much smaller extent, from non-wood raw materials such as agricultural and agri-food
residues (e.g., oil palm empty fruit bunches) or annual and non-annual plants providing large
amounts of biomass (e.g., kenaf, sorghum). Wood pulp production increased by about 4 %
over the period 1999-2003; by contrast, non-wood pulp production rose by 10 % in response
to the need for alternative raw materials, a need which has arisen from the growing concern
with environmental (e.g., deforestation) and economic problems (some countries lack wood
but abound with non-wood resources) (Alaejos et al., 2004).
One especially attractive non-wood pulping material is provided by oil palm empty
fruit bunches (EFB), which constitutes an effective source of lignocellulosic bers as
conrmed by various researchers. Thus Ibrahim (2002), compared the composition of kraft,
kraft-anthraquinone, soda and soda-anthraquinone pulps from EFB and found the soda
variant to exhibit the best properties. The soda process has also been used to obtain
chemical and semichemical pulps from EFB (Law and Jiang, 2001; Roliadi and Pasaribu,
2004; Wan Rosli et al., 2004). EFB have also been used in organosolv (Aziz et al., 2002;
Quader and Lonnberg, 2005; Rodrguez et al., 2008) and mechanical pulping processes
(Daud et al., 2005; Ghazali et al., 2006).
However, only three references to the bleaching of EFB pulp appear to exist. In them,
kraft and soda pulps from EFB were bleached with totally chlorine free (TCF) sequences. In
one (Tanaka et al., 2004), the pulp was subjected to an OZP (oxygen - ozone - hydrogen
peroxide) sequence with a sulphuric acid treatment inserted between the O and Z steps. In
other (Leh et al., 2005), soda, soda-anthraquinone, kraft and kraft-anthraquinone pulps from
EFB were subjected to an OZP sequence in order to assess the effect of the bleaching
treatment on viscosity and kappa number; soda-anthraquinone and kraft-anthraquinone
pulps exhibited 80 % of brightness and while retaining a viscosity above 15 cp. And the last
one (Leh et al., 2008) optimizes delignication with oxygen in TCF pulp production.
In this work, changes in the properties of soda-anthraquinone and diethanolamine
pulps from EFB (yield, kappa number, beating grade and viscosity) and paper sheets made
from them (tensile index, stretch, burst index, tear index and brightness) were examined and
244

Publicaciones Cientficas/Scientific Publications


compared after each step of an AOpAZRP bleaching sequence (A, an acid treatment; Op,
an oxygen and hydrogen peroxide stage; Z, an ozone stage; R, a reductive treatment and P,
an hydrogen peroxide stage).
The soda-anthraquinone pulping process was chosen on the grounds of its usually
providing quality pulp (Ibrahim, 2002; Law and Jiang, 2001), using no sulphur which avoids
the release of pollutants and causing bad odours and requiring only modest amounts of raw
material to be protable which is compatible with the low non-wood production of some
regions in the world.
Organosolv pulping processes, which use no sulphur-containing reagents either, have
only recently been implemented on the pilot plant or industrial scale (Lpez et al., 2006;
Rodrguez and Jimnez, in press), largely in response to the need to respond to the new
economic and, especially, environmental challenges. Their greatest shortcoming is that they
usually require high pressures, and hence special equipment, which raises operational costs.
This makes the use of solvents with a high boiling point (e.g., diethanolamine) a potentially
interesting alternative, because of installation is less expensive when operating to smaller
pressures (Rodrguez et al., 2008; Rodrguez and Jimnez, in press).
On the other hand, the restrictive legislation of the developed countries with respect to
the use of chlorine compounds in the bleaching of pulps, by means ECF process (elemental
chlorine free), it is leading to the scientic community to the use of TCF bleaching processes
(totally chlorine free). For that reason, in the last years, TCF bleaching sequences are
studying and applying to pulps from different wood and non-wood raw materials: Eucalyptus

globulus (Freire et al., 2006; Roncero et al., 2005; Shatalov and Pereira, 2008a), ax
(Sigoillot et al., 2005), Arundo donax L. (Abrantes et al., 2007; Shatalov and Pereira, 2007,
2008b), Asclepios syriaca (Spiridon, 2007) and wheat straw (Hedjazi et al., 2008).

2. EXPERIMENTAL
2.1. Raw material
There are three major species of oil palm, namely: Elaeis guineensis, E. oleifera and E.

odera. We used E. guineensis, which is known as African palm because it originated in West
Africa, where it has been used to obtain oil for more than ve thousand years (particularly in
Western Guinea). After Columbus travels, African palm was introduced in America, from
where it reached the Asian continent (Sumatra and Malaysia) in ca. 1900. Malaysia is
245

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481


currently the greatest world producer of palm oil, with 51 % of the world output; therefore,
palm oil constitutes a great asset for this country. Cultivation of the plant is currently
expanding in Western Africa (Nigeria, Guinea, Ghana), Latin America (Ecuador, Colombia,
Honduras) and Asia (Thailand) (Rodrguez et al., 2008).
Palm plants give their rst fruits 4-5 years after planting; production peaks after 20-30
years and then declines to eventually unprotable levels particularly as result of the
remaining fruits being too high to collect. Palm bunches weigh 15-25 Kg and contain 10004000 oval-shaped fruits of 3-5 cm of length each one. Each hectare of oil palm produces an
average of 10 tons of fruits per year from which an amount of ca. 3000 Kg of oil is obtained.
The process leaves substantial amounts of empty fruit bunches (EFB) (Rodrguez et al.,
2008).

2.2. Analysis of the raw material


The EFB was characterized as follows: holocellulose by the Wise et al. method (1946);
and -cellulose, lignin, ash and ethanol-benzene extractives according to T-203 OS-61, T222, T-211 and T-204 (TAPPI Standards, 2007), respectively.
Oil palm empty fruit bunches contain 67.0 % holocellulose, 41.9 % -cellulose, 24.5 %
lignin, 1.2 % ethanol-benzene extractives and 3.2 % ash.

2.3. Pulping and pulp characterization


Pulp was obtained by using a 15-L batch cylindrical reactor that was heated by means
of electrical wires and linked through a rotary axle to ensure proper agitation to a control
unit including a motor actuating the reactor and the required instruments for measurement
and control of the pressure and temperature.
The raw material was cooked (with soda-anthraquinone or diethanolamine) in the
reactor. Next, the cooked material was berized in a wet disintegrator at 1200 rpm for 30
min and the screenings were separated by sieving through a screen of 0.16 mm mesh size.
The pulp obtained was beaten in a Sprout-Bauer rener.
The conditions used in the soda-anthraquinone and diethanolamine processes (Table
1), were chosen in such a way as to provide similar kappa numbers for both types of pulp.

246

Publicaciones Cientficas/Scientific Publications


The viscosity, kappa number and beating grade (Shopper-Riegler value) of the pulp
were determined according to T-254, T-236 cm-85 and T-248 (TAPPI Standards, 2007),
respectively. Pulp yield was determined by weighing, after removing the uncooked materials.
Table 1. Operating conditions used to obtain EFB pulp for subsequent A1OpA2ZRP bleaching
sequence.
Pulp

Temperature, C

Time, min

Sodaanthraquinone

170

60

Diethanolamine

180

150

Reagent
concentration, %
15 % soda; 0.5 %
anthraquinone
80 %
diethanolamine

Liquid/solid ratio
4
6

Table 2. Operating conditions used to bleach EFB pulp with an A1OpA2ZRP sequence.
Variables

Stages of sequence AOpAZRP

Op

Consistency (%)

1.5

10

40

10

10

Temperature (C)

25

100

25

25

25

85

Time (min)

30

60

30

60

150

Oxalic acid (%)

Oxygen pressure (Kg/cm2)

4
6

MgSO4 (%)

0.5

0.2

H2O2 (%)

NaOH (%)

1.8

1.8

DTPA (%)

0.1

0.5

Oxygen flow (L/h)

180

Ozone (g/L)

35

NaBH4 (%)

Na2CO3 (%)

2
1

2.4. Bleaching
The two types of EFB pulps were bleached with an AOpAZRP sequence (Serrano,
2008). The operating conditions used to implement this bleaching sequence are summarized
in Table 2.

247

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481


2.5. Paper sheets formation and characterization
Paper sheets were prepared with an ENJO-F-39.71 sheet machine according to the
UNE-57-042 standard.
The tensile index, stretch, burst index, tear index and brightness of paper sheets were
determined according to T-494 om-96, T-494, T-403 om-97, T-414 om-98 and T-452
(TAPPI Standards, 2007), respectively.

3. RESULTS AND DISCUSSION


Table 3 shows the experimental values of yield, kappa number, beating grade
(Shopper-Riegler value) and viscosity for bleached soda-anthraquinone and diethanolamine
pulps, and the tensile index, stretch, burst index, tear index and brightness of paper sheets
obtained from them.
As can be seen, diethanolamine pulp exhibited higher yield losses during the AOp
sequence than soda-anthraquinone pulp, namely: 12.3 % points with respect to unbleached
pulp and 5.8 % points with respect to the raw material vs. 10.8 % and 4.2 % points,
respectively for the soda-anthraquinone pulp. The yield loss over the AZR sequence was also
higher for diethanolamine pulp, which lost 6.8 % and 2.8 % points with respect to
unbleached pulp and the raw material, vs. 2.3 % and 0.9 % points, respectively, for sodaanthraquinone pulp; in any case, the yield losses in this bleaching sequence were lower than
those observed in the previous one (AOp). Finally, diethanolamine pulp also exhibited
higher yield losses in the P step; the loss amounted to 6.0 % and 2.7 % points relative to
unbleached pulp and the raw material, vs. 3.2 % and 1.2 % points, respectively, for sodaanthraquinone pulp; as can be seen, the yield loss in this step exceeded the cumulative loss
in the previous ones (AZR) for soda-anthraquinone pulp, but not for diethanolamine pulp.
The total yield with respect to unbleached pulp over the A1OpA2ZRP sequence was lower
for diethanolamine pulp (74.9 %) than for soda-anthraquinone pulp (83.7 %). However,

because the pulping yield was higher for the latter one, the raw material was more efciently
used with the diethanolamine process than with the soda-anthraquinone process; this
reected in the nal yield with respect to the raw material, which was 33.9 % vs. 32.5 %.

248

Publicaciones Cientficas/Scientific Publications


Table 3. Experimental results in the A1OpA2ZRP bleaching sequence applied to the soda-anthraquinone and diethanolamine pulps.
Yield on
Pulp

Treatment

unbleached
pulp, %

Sodaanthraquinone

Diethanolamine

Yield on
raw
material,
%

ShopperRiegler
value, SR

Kappa

Viscosity,

number

mL/g

Tensile
index,
Nm/g

Stretch,
%

Burst

Tear

index,

index,

KN/g

mNm2/g

Brightness,
%

Untreated

100

38.8

16.0

14.24

605

25.8

2.35

1.69

0.50

60.6

A1Op

89.2

34.6

8.0

8.51

675

14.6

1.21

0.77

0.66

71.0

A1OpA2ZR

86.9

33.7

9.0

3.32

676

17.3

1.46

1.27

0.65

75.1

A1OpA2ZRP

83.7

32.5

10.0

2.82

653

20.4

2.58

1.42

0.70

77.5

Untreated

100

45.2

12.0

17.26

659

13.8

1.66

0.74

0.27

25.9

A1Op

87.7

39.4

10.0

7.43

808

15.7

1.48

0.25

0.68

64.5

80.9

36.6

10.0

6.09

796

13.8

1.00

0.30

0.64

68.0

A1OpA2ZRP

74.9

33.9

11.5

5.54

783

22.2

1.31

1.30

0.71

71.3

A1OpA2ZR

249

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481


Table 4. Comparison of viscosity values and physical properties of various types of bleached pulps.

Type of bleached pulp

Viscosity,
mL/g

Tensile

Burst

Tear

index,

index,

index,

Nm/g

KN/g

mNm2/g

Brightness,
%

EFB diethanolamine pulp bleached with AOpAZRP

783

22.18

1.30

7.10

71.3

EFB soda-anthraquinone pulp bleached with AOpAZRP

653

20.36

1.42

7.00

77.5

EFB kraft pulp bleached with OAZP (Tanaka et al., 2004)

63.9

5.05

11.7

73.9

Wheat straw AS/AQ pulp bleached with OQ(OP)P (Hedjazi et al., 2008)

755

70.0

4.20

5.60

83.6

A. donax L. ethanol-soda pulp bleached with AZERQP (Shatalov and Pereira, 2008b)

988

15.85

1.32

9.16

83.7

A. donax L. ASAM pulp bleached with AZERQP (Shatalov and Pereira, 2008b)

1009

15.28

1.30

9.02

79.4

A. donax L. organocell pulp bleached with AZERQP (Shatalov and Pereira, 2008b)

789

14.26

1.27

8.54

83.4

A. donax L. kraft pulp bleached with AZERQP (Shatalov and Pereira, 2008b)

852

8.22

1.11

8.84

79.2

A. syriaca pulp bleached with XEPP (X=Cartazyme EG) (Spiridon, 2007)

71.73

0.056

6.39

75.51

E. globulus kraft pulp bleached with XQPPP (X=Ecopulp) (Shatalov and Pereira, 2008a)

789

8.67

0.37

4.08

86.4

E. globulus kraft pulp bleached with XQPPP (X=Pulpzyme) (Shatalov and Pereira, 2008a)

661

8.98

0.40

4.26

85.7

A, acid treatment; O, oxygen; P, hydrogen peroxide; Z, ozone; R, reductive treatment; Q, chelating treatment; E, alkaline extraction; X, enzymatic treatment.

250

Publicaciones Cientficas/Scientific Publications


The beating grade changed similarly in both types of bleached pulps in all steps. Thus,
it decreased in the initial bleaching steps (AOp), remained constant or rose to the starting
values in the next stages (AZR) and regained such values in the last step (P). In any case, it
never reached the values for unbleached pulps.
As regards kappa number, soda-anthraquinone pulp exhibited successive decreases of
5.73, 5.19 and 0.5 units after AOp, AZR and P stages, respectively, and so did
diethanolamine pulp (9.83, 1.34 and 0.55 units, respectively). The decrease was therefore
especially marked in the rst two bleaching steps.
The viscosity was invariably higher for diethanolamine pulp. The AOp sequence
increased viscosity in both types of pulps; the AZR sequence maintained it in the case of the
soda-anthraquinone pulp and slightly decreased it in the case of the diethanolamine pulp;
and the P step reduced it in both pulps. However, the viscosity values after the complete
bleaching sequence always exceeded those of their unbleached pulps.
The tensile index of the paper sheets from soda-anthraquinone pulp was strongly
decreased by the AOp sequence and recovered after the AZR sequence and the P step, but
it was always lower than that for its unbleached pulp. In diethanolamine pulp, the AOp
sequence increased the tensile index; the AZR sequence decreased it and the P step
increased it again. Therefore, the paper sheets of the diethanolamine pulp exhibited a nal
tensile index exceeding that of paper sheets from its unbleached pulp.
The stretch of the paper sheets was decreased by the AOp sequence. The stretch of
the soda-anthraquinone paper sheets increased from AOp to AZR and P, whereas that for
diethanolamine paper sheets decreased from AOp to AZR and was then recovered after P.
In any case, the stretch values for soda-anthraquinone paper sheets exceeded those for
diethanolamine paper sheets. The nal stretch for paper sheets from completely bleached
soda-anthraquinone pulp was higher than that for its unbleached pulp; by contrast, that for
paper sheets from completely bleached diethanolamine pulp was somewhat low relative to
paper sheets for its unbleached pulp.
The burst index of the paper sheets was invariably higher for soda-anthraquinone pulp;
however, while this index for paper sheets from completely bleached pulp of this type was
low relative to unbleached pulp, the opposite was true for diethanolamine pulp.
The tear index of paper sheets from soda-anthraquinone and diethanolamine pulps
increased while the bleaching sequence progressed (except in the AZR sequence). In both
251

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481


types of pulps, the tear index was decreased by AZR and increased by both AOp and P
sequences. Also, the nal values were similar for both pulps.
Finally, brightness increased over the successive steps of the AOpAZRP sequence.
Although it reached higher levels for soda-anthraquinone pulp, the unbleached pulp of this
type was brighter than its diethanolamine counterpart. In fact, although the nal brightness
of soda-anthraquinone paper sheets was higher than that for diethanolamine paper sheets
(77.5 % vs. 71.3 %), the overall increase during the bleaching sequence was more marked in
the latter one than in the first one (175,29 % vs. 27.89 %).
The most efcient bleaching subsequence for both soda-anthraquinone and
diethanolamine pulps was AOp, which raised brightness to 17.16 % and 149,03 %,
respectively, in comparison with the brightness values presented by the untreated materials.
Table 4 compared the results of viscosity of pulps and the physical properties of paper
sheets obtained in this work, with those provided by other authors for the bleaching of EFB
and other non-wood raw materials, as well as eucalyptus wood.
Values of the physical properties of the paper sheets corresponding to EFB kraft pulp
bleached with OAZP sequence, are higher than those obtained in this work; it can be due to
the different pulping methods used, as well as the different bleaching and rened conditions
used. Leh et al. (2008) have bleached with oxygen a pre-hydrolysed-soda pulp of EFB,
obtaining a pulp with a kappa number of 2.4 and a brightness of 67.1 %, lower values than
those obtained in this work.
For non-wood bleached pulps with TCF sequences, the following thing can be
indicated, with respect to pulps studied here. For alkaline sulphite-anthraquinone pulp of
wheat straw, bleached with OQ(OP)P sequence, a lower viscosity is found compared with
that for diethanolamine pulp studied in this work; however, it has higher values for the
physical properties of the paper sheets. For A. donax L. pulps obtained with different
organosolv processes (ethanol-soda, ASAM and Organocell) and with kraft process, values of
viscosity, tear index and brightness are found higher than those obtained here for the EFB;
but for tensile index is lower, and for burst index the same value or lower; all these variations
must be explained by the different nature of the materials, as well as the different conditions
used in pulping and bleaching. Finally for A. syriaca, a higher value for tensile index of
bleached pulps is found, a lower ones for burst index and tear index, and an intermediate

252

Publicaciones Cientficas/Scientific Publications


value for brightness, compared with pulps of EFB studied in this work; these variances could
be explained by using the same reasons indicated for the A. donax L.
When comparing the results of pulps of EFB studied with those of E. globulus, it is
observed that viscosities are similar, but brightness is higher and the properties of resistance
are lower for eucalyptus pulps; like in the previous cases, these variances could be explained
by the difference that exists between both materials, and in addition these materials have
been processed by a different form.

4. CONCLUSIONS
The soda-anthraquinone and diethanolamine pulps from bleached EFB with the
AOpAZRP sequence provide viscosity values related to the bleached pulps and physical
property values related to the paper sheets similar to or higher than the values presented by
the pulps obtained from other non-wood raw materials and bleached eucalyptus with TCF
sequences.
Diethanolamine pulp after bleaching with an AOpAZRP sequence exhibited a higher
viscosity (783 mL/g), and the properties of the paper sheets were close to or even better than
those of the paper sheets from bleached soda-anthraquinone pulp (viz. 22.2 vs. 20.4 Nm/g
for tensile index, 1.30 vs. 1.42 KN/g for burst index, 0.71 vs. 0.70 mNm2/g for tear index
and 71.3 vs. 77.5 % for the brightness).

ACKNOWLEDGEMENTS
The authors are grateful to Ecopapel, S.L. (cija, Seville, Spain) and ENCE (Huelva,
Spain) for their support, to Spains DGICyT for funding this research within the framework
of the Project CTQ2007-65074-C02-01, and to the Ramn y Cajal programme (Spains
Ministry of Education and Science) for additional funding.

REFERENCES
Abrantes, S., Amaral, E., Costa, A.P., Shatalov, A.A., Duarte, A.P., 2007. Hydrogen
peroxide bleaching of A. donax L. kraft-anthraquinone pulp. Effect of a chelating stage.
Industrial Crops and Products 25, 288-293.

253

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481


Alaejos, J., Lpez, F., Eugenio, M.E., 2004. Non-wood raw materials for pulp and
paper making. A review. Anidad 61 (513), 400-410.
Annual FAO, 2005.
Aziz, A.A., Husin, M., Mokhtar, A., 2002. Preparation of cellulose from oil palm
empty fruit bunches via ethanol digestion: effect of acid and alkali catalysts. Journal of Oil
Palm Research 14 (1), 9-14.
Daud, W.R.W., Law, K.N., Ghazali, A., 2005. Alkaline peroxide mechanical pulping
of oil palm lignocellulosic. Part 1. Effects of chemical and mechanical pretreatments on
empty fruit bunches (EFB) chemical composition. JIRCAS Working Report 39, 82-85.
Freire, C.S.R., Silvestre, A.J.D., Neto, C.P., Evtuguin, D.V., 2006. Effect of oxygen,
ozone and hydrogen peroxide bleaching stages on the contents and composition of
extractives of Eucalyptus globulus kraft pulps. Bioresource Technology 97, 420-428.
Ghazali, A., Wan Rosli, W.D., Law, K.N., 2006. Alkaline peroxide mechanical
pulping (APMP) of oil palm lignocellulosic. Part 2. Empty fruit bunches (EFB) responses to
pretreatments. APPITA Journal 59 (1), 65-70.
Hedjazi, S., Kordsachia, O., Patt, R., 2008. Alkaline sulte-anthraquinone (AS/AQ)
pulping of wheat straw and totally chlorine free (TCF) bleaching of pulps. Industrial Crops
and Products (doi:10.1016/j.indcrop.2008.03.013).
Ibrahim, R., 2002. Chemical composition of alkaline pulps from oil palm empty fruit
bunches. Oil Palm Bulletin 44, 19-24.
Law, K.N., Jiang, X., 2001. Comparative papermaking properties of oil-palm empty
fruit bunches. TAPPI Journal 84 (1), 95.
Leh, C.P., Tanaka, R., Ikeda, T., Wan Rosli, W.D., Magara, K., Hosoya, S., 2005.
Totally chlorine free bleaching of oil palm empty bunch (EFB) chemical pulps. JIRCAS
Working Report 39, 95-99.
Leh, C.P., Wan Rosli, W.D., Zainuddin, Z., Tanaka, R., 2008. Optimization of
oxygen delignication in production of totally chlorine free cellulose pulps from oil palm
empty fruit bunches fibre. Industrial Crops and Products 28 (3), 260-267.
Lpez, F., Alfaro, A., Jimnez, L., Rodrguez, A., 2006. Alcohols as organic solvents
for the obtainment of cellulose pulp. Afinidad 63 (523), 174-182.

254

Publicaciones Cientficas/Scientific Publications


Quader, M.M.A., Lonnberg, B., 2005. Pulp and chemical cellulose obtained with oil
palm empty fruit bunches by a modied IDE process. JIRCAS Working Report 39, 47-53.
Rodrguez, A., Jimnez, L., in press. Pulping with organic solvents other than alcohols.
Afinidad.
Rodrguez, A., Serrano, L., Moral, A., Prez, A., Jimnez, L., 2008. Use of highboiling point organic solvents for pulping oil palm empty fruit bunches. Bioresource
Technology 99, 1743-1749.
Roliadi, H., Pasaribu, R.A., 2004. The possible utilization of empty oil-palm bunches
as a substitute of pulp wood for the manufacture of corrugating-medium paperboard using
semi-chemical soda-anthraquinone pulping. In: Proceedings of the Fifty-Eighth APPITA
Annual Conference and Exhibition 2, 335-341.
Roncero, M.B., Torres, A.L., Colom, J.F., Vidal, T., 2005. The effect of xylanases on
lignocellulosic components during the bleaching of wood pulps. Bioresource Technology
96, 21-30.
Shatalov, A.A., Pereira, H., 2007. Polysaccharide degradation during ozone-based
TCF bleaching of non-wood organosolv pulps. Carbohydrate Polymers 67, 275-281.
Shatalov, A.A., Pereira, H., 2008a. Effect of xylanases on peroxide bleachability of
eucalypt (E. globulus) kraft pulp. Biochemical Engineering Journal 40, 19-26.
Shatalov, A.A., Pereira, H., 2008b. A. donax L. reed: new perspectives for pulping and
bleaching. 5. Ozone-based TCF bleaching of organosolv pulps. Bioresource Technology 99,
472-478.
Serrano, L., 2008. Empty fruit bunches residues as alternative raw material to obtain
paper. Doctoral Thesis, University of Cordoba, Spain.
Sigoillot, C., Camarero, S., Vidal, T., Record, E., Asther, M., Prez-Boada, M.,
Martnez, M.J., Sigoillot, J.C., Asther, M., Colom, J.F., Martnez, A., 2005. Comparison of
different fungal enzymes for bleaching high-quality paper pulps. Journal of Biotechnology
115, 333-343.
Spiridon, I., 2007. Modifications of A. syriaca fibers for paper production. Industrial
Crops and Products 26, 265-269.
Tanaka, R., Wan Rosli, W.D., Magara, K., Ikeda, T., Hosoya, S., 2004. Chlorine free
bleaching of Kraft pulp from oil palm empty fruit bunches. JARQ 38 (4), 275-279.
255

L. Jimnez et al./Bioresource Technology 100 (2009) 1478-1481


TAPPI Standards, 2007. TAPPI Test Methods, Atlanta.
Wan Rosli, W.D., Zainuddin, Z., Lee, L.K., 2004. Influence of pulping variables on
the properties of E. guineensis soda pulp as evaluated by response surface methodology.
Wood Science and Technology 38 (3), 191-205.
Wise, L.E., Murphy, M., D Adieco, A., 1946. Chlorite holo-cellulose its fractionation
and beating on summative wood analysis and on studies on the hemicellulose. Paper Trade
Journal 122 (2), 35.

256

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 5
MODELING HYDROGEN PEROXIDE BLEACHING OF SODA PULP FROM OILPALM EMPTY FRUIT BUNCHES
Ana Ferrer, Antonio Rosal, Cristina Valls, Blanca Roncero and Alejandro Rodrguez
BioResources, 2011, 6 (2) 12981307
ndice de impacto: 1.328 (Cuarta posicin de 21 revistas en el rea de Materials Science,
paper & wood segn el Journal Citation Reports de 2011)

257

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307

258

Publicaciones Cientficas/Scientific Publications


BioResources 6 (2) 12981307
PEER-REVIEWED ARTICLE

bioresources.com

MODELING HYDROGEN PEROXIDE BLEACHING OF SODA PULP FROM OIL-PALM


EMPTY FRUIT BUNCHES
Ana Ferrera, Antonio Rosalb, Cristina Vallsc, Blanca Ronceroc and Alejandro
Rodrgueza

a Chemical

Engineering Department, University of Crdoba, Spain


Molecular Biology and Biochemical Engineering Department, University Pablo de Olavide, Seville,
Spain
c Textile and Paper Engineering Department, Technology University of Catalua, Spain

Article submitted: December 17, 2010; Peer review completed: February 19, 2011; Revised version received and
accepted: March 1, 2011; Published: March 3, 2011

ABSTRACT
The influence of the variables soda (0.5-3.0 %), hydrogen peroxide (1.0-6.0 %) and
time (1-5 h) in the bleaching of soda pulp of Empty Fruit Bunches (EFB) from oil palm, on
the properties of bleached pulps, was studied. Polynomial and neural fuzzy models
reproduced the results of brightness, kappa number, and viscosity of the pulps with errors
less than 10 %.
By the simulation of the bleaching of pulp, using the polynomial and neural fuzzy
models, it was possible to find optimal values of operating variables, so that the properties of
bleached pulps differed only slightly from their best values and yet it was possible to save
chemical reagents, energy, and plant size, operating with lower values of operating variables.
Thus, operating with 1.13 % soda concentration, 2.25 % hydrogen peroxide concentration
for 3 hours, a pulp with a brightness of 75.1 % (8.1 % below the maximum) and a viscosity of
740 mL/g (10.4 % less than the maximum value), was obtained.

KEYWORDS: Empty Fruit Bunches (EFB); Bleaching; Hydrogen peroxide; Polynomial


model; Neural fuzzy model.

259

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


1. INTRODUCTION
In recent decades the increasing production of pulp for paper and other uses has led
to a problem of inadequate supply of the classic raw materials (mainly hardwoods and
softwoods) in some countries. For that reason there has been a growing emphasis on the use
of alternative raw materials, mainly non-wood waste such as agricultural and food industry
residues, forest residues and fast-growing plants other than conventional woods (Jimnez et
al. 2006).
One of the interesting non-wood resources is Empty Fruit Bunches (EFB), which
consists of the lignocellulosic material residue from the palm oil industry. The value of these
resources has been confirmed by several studies by various researchers.
Several authors have studied the kraft process applied to these residues, of which the
results presented by Ibrahim (2002) are the most relevant and recent. That publication
compares the composition of EFB pulp obtained by the kraft, kraft-anthraquinone, soda,
and soda-anthraquinone processes; the pulp obtained with the soda process had the highest
content of lignin, holocellulose, and -cellulose, as well as a higher viscosity.
The soda process has also been studied by Law and Jiang (2001), producing fibers with
more wall thickness, greater rigidity, higher solubilities in hot water and 1 %-soda, and higher
ash content. These pulps were bleached with hydrogen peroxide more easily than those of
aspen; their paper sheets had lower tensile index, but greater stretch and tear index than
those of aspen.
Moreover, Daud et al. (1998) pulped the EFB with soda, sodium carbonate and
sodium sulfite processes, finding the first process to be the most efficient. Wanrosli et al.
(2004), using a central composite experimental design, studied the influence of operating
conditions (temperature, time, and alkali concentration) on the properties of EFB pulps
(yield, kappa number, tensile index, and tear index). They obtained pulp yields in the range
30 to 45 %. It was considered that the values of 160 C, 60 to 120 minutes, and 20 to 30 % of
alkali are sufficient for the proper pulping.
Jimnez et al. (2009a) also have studied the pulping of EFB with soda-anthraquinone,
considering the influence of pulping variables and subsequent beating. Semi-chemical
pulping for board by the soda-anthraquinone process also has been studied (Roliadi and
Pasaribu 2004).

260

Publicaciones Cientficas/Scientific Publications


Studies of thermomechanical pulp have also been made by several researchers,
highlighting the results of Ghazali et al. (2006). Organosolv processes have also been studied:
ethanol (Aziz et al. 2002), by modified IDE process (Quader and Lonnberg 2005), and highboiling point organic solvents (Rodrguez et al. 2008a). Finally, biopulping has been
investigated using a white fungus K14 (Goenadi et al. 1998).
The pulp bleaching plant, which is the most contaminating section in the paper
manufacturing process, has gone through a number of changes intended to alleviate its
adverse environmental impact.
The need to reduce or eliminate the formation of organochlorinated compounds of
high toxicity during the bleaching processes has led to the emergence of new products in the
market, such as ECF (Elemental Chlorine Free) and TCF (Totally Chlorine Free) pulps
(Ramos et al. 2008). TCF pulping processes avoid the formation of highly toxic
organochlorine compounds (AOX) during bleaching.
Usually, TCF sequences include oxygen, hydrogen peroxide, and ozone-based stages
(Freire et al. 2006, Lpez et al. 2003, Pedrola et al. 2004, Roncero et al. 2003, Shatalov and
Pereira 2005, Torres et al. 2004, Villaverde et al. 2009). Recently, enzyme stages involving
xylanases or the laccase-mediator system have provided very promising results in pulp
bleaching sequences (Valls and Roncero 2009, Valls et al. 2009, 2010a,b).
In the pulping of different materials, several authors have studied the influence of
operating variables on the characteristics of the pulps obtained by the use of polynomial and
neural fuzzy models. Polynomial models have also been used successfully for the study of
operating variables in the bleaching of pulps (Jimnez et al. 2008, Lpez et al. 2001, 2002,
2003, Pedrola et al. 2004, Valls and Roncero 2009, Valls et al. 2009, 2010b), but not studied
in cases in which neural fuzzy models were used.
The aim of this work was to obtain a bleached EFB pulp with the best application
conditions of the pressurized hydrogen peroxide stage (Po). A Po stage was performed after
a WAOQ sequence with EFB pulp (WA, acid washing; O, oxygen delignification; Q,

chelating stage). For this reason this Po stage was optimized following a sequential statistical
plan of three variables (soda concentration, hydrogen peroxide concentration, and
processing time). The results obtained were analyzed according to polynomial and neural
fuzzy models.

261

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


2. EXPERIMENTAL
Raw material
The research employed EFB from Malaysian oil palm (supplied by Palmor Corp.
SDN BHD). Malaysia supplies 51 % of the worlds production of this oil (Malaysian Palm
Oil Promotion Council). Each hectare of oil palm produces an average of 10 tons of fruits
per year, and that amount of fruit yields 3000 Kg of oil (the main product)
(http://www.arbolesornamentales.com).
The chemical properties of EFB were determined in accordance with the respective
TAPPI standards for the different components, namely: T-222 for lignin, T-203 0S-61 for cellulose, T-9 m-54 for holocellulose, T-204 for ethanol-benzene extractives, and T-211 for
ash. The analysis showed the following results: 24.5 % lignin, 41.9 % -cellulose, 67.0 %
holocellulose, 1.2 % extractives, and 3.2 % ash.
The average fiber length of EFB, determined by using a Visopan projection
microscope, was found to be 0.53 mm, and the thickness or width 14 m.

Pulping
Pulp was obtained by using a cylindrical 15-L batch reactor that was heated by means
of electrical wires and attached to a rotary axle (to ensure proper agitation) to a control unit
including a motor activating the reactor and the required instruments for measurement and
control of pressure and temperature.
The raw material was cooked in the reactor, using 15 % of soda, 170 C, a liquid/solid
ratio of 10, and a processing time of 30 minutes, obtaining a pulp yield of 48 %. These
operating conditions were selected based on the results of other authors (Daud et al. 1998,
Jimnez et al. 2009b, Law and Jiang 2001). Next, the cooked material was fiberized in a wet
disintegrator at 1200 rpm for 30 min, and the screenings were separated by sieving through a
screen of 0.14 mm mesh size. The pulp obtained was beaten in a Sprout-Bauer refiner.

Bleaching
The initial pulp properties of EFB were 38.4 % ISO brightness, kappa number 9.7,
and 797 mLg-1 viscosity. Before applying the hydrogen peroxide stage (Po), the pulp was
262

Publicaciones Cientficas/Scientific Publications


washed in acidic medium (WA) and an oxygen delignification stage (O) was performed

followed by a chelating stage (Q) (WAOQ sequence).

The pressurized peroxide bleaching stage (Po) (conditions selected based on previous

works (Aracri and Vidal 2010, Moldes et al. 2010)) was carried out with 25 g o.d.p (over
dried pulp) in a 5-L reactor at 0.6 MPa O2, with 0.5 % o.d.p of Na2SiO3, 0.2 % o.d.p of
MgSO4, at 5 % consistency, at 105 C and 60 rpm with a blade stirrer. The soda
concentration, hydrogen peroxide concentration and processing time were the three
variables of the experimental design, and were varied over the following ranges: 0.5-3 %
o.d.p for soda concentration, 1-6 % o.d.p for hydrogen peroxide, and 1-5 h for processing
time. After the Po stage, the liquors were recovered for pH measurement and the pulp was
efficiently washed for characterization.

Pulp properties
Treated pulp samples were characterized in terms of kappa number, brightness, and
viscosity according to ISO-302, ISO-3688, and ISO-5351-1, respectively. The kappa number
was measured twice, and four measurements of brightness were obtained in order to
calculate a relative standard deviation, which was found to be 0.1 for both properties. The
viscosity was determined 4 times with a standard deviation of 5.

Experimental design
The factorial design used (Montgomery 1991) consisted in a central experiment (in the
center of a cube) and several additional points (additional experiments being at the cube
vertices and side centers).

Polynomial model
Experimental data were fitted to the following second-order polynomial,
Ye= a0 + a1XS + a2XP + a3XT + a11XS2 + a12XS XP + a13XS XT + a22XP2 + a23XP XT + a33XT2 (1)

where Ye denotes the response variables [viz. kappa number (KN), viscosity (VI) or

brightness (BR)]; XS, XP, and XT are the normalized values of the operational variables (soda

263

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


concentration, hydrogen peroxide concentration and time, respectively), and a0 to a23 are
constants.

The values of the operational variables were normalized to values from -1 to +1 by


using the following expression,
Xn =

)
2(X-X

X max -X min

(2)

where Xn is the normalized value of soda concentration (S), hydrogen peroxide (P), or time

(T); X is the absolute experimental value of the variable concerned;


X is the mean of Xmax
and Xmin; and Xmax and Xmin are the maximum and minimum value, respectively, of such a
variable.

The normalized values for the independent variables in the 15 experiments conducted
are given in Table 1.
Table 1. Normalized values of operational variables and experimental values of the properties of
bleached pulp.
Experiment
Soda XS, Peroxide XP, Time XT

264

Brightness,
% ISO

Viscosity,
mL/g

Kappa number

1. -1,-1,-1

63.7

814

4.86

2. -1,+1,-1

72.7

678

3.27

3. +1,+1,-1

77.0

637

3.15

4. +1,-1,-1

71.4

771

3.42

5. -1,-1,+1

71.5

774

3.57

6. +1,-1,+1

75.3

695

3.13

7. -1,+1,+1

80.1

580

1.14

8. 0,0,0

77.7

671

3.00

9. +1,+1,+1

78.9

573

2.95

10. +1,0,0

78.2

641

2.74

11. 0,0,+1

79.3

670

2.70

12. -1,0,0

76.0

731

3.05

13. 0,-1,0

74.0

765

3.23

14. 0,0,-1

74.4

703

3.60

15. 0,+1,0

81.5

637

2.81

16. -0.5,-0.5,-0.5

72.7

745

3.83

17. +0.5,+0.5,+0.5

79.5

626

2.97

Publicaciones Cientficas/Scientific Publications


Neural fuzzy model
The integration of fuzzy systems and neural networks combines the advantages of the
two systems and provides an especially powerful modeling tool, viz. the neural fuzzy system,
which uses neural networks as tools in fuzzy systems (Jang et al. 1997).
Thus, the variation of pulp properties as a function of the operational variables of the
bleaching process can be predicted by using the following expression,
y [ ( x , )
[ ( x , )

m
l
n
l
l
=l 1 =i 1 Fi
i
i
e
m
n
l
l
=l 1 =i 1 Fi
i
i

y =

(3)

where ye is the estimated value of the property to be modeled, m the number of fuzzy rules

applied, n the number of independent variables, yl the defuzzifier of a fuzzy rule and

lFi(xi,il) the membership function of the independent variables within its range.

In this paper, [=1 lFi (xi,li)] will be denoted by Rl which is known as a fuzzy rule.

In this sense, Rl is defined by the product of n membership functions (one per independent
variable). Then, equation (3) could be expressed as:
Ye =

'
m
l=1 Y Rl

m
l=1 Rl

(4)

With three operational variables (our case), the use of a Singleton defuzzifier (a
constant parameter: cl) and a linear membership function for the independent variables

allows equation (4) to be simplified to:

Ye =

8l=1 Cl Rl
8l=1 Rl

(5)

Linear membership functions are selected for independent variables with two levels.
The mathematical equations which respond to linear membership functions are,
(X-X low )

(6)

(X-X high )

(7)

X1 =Xlow = 1- X
X2 =Xhigh = 1-

high - Xlow

X high - X low

X1 and X2 denoting the values of S, P, and T for low and high levels; Xhigh and Xlow the

extreme values of the variable; and X the absolute value of S, P, or T.

265

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


With three independent variables one can establish the following eight fuzzy rules
according to the extreme (high and low) values of such variables:
R1: low S, low P and low T = S1 P1 T1

R2: low S, low P and high T = S1 P1 T2


...

R7: low S, high P and high T = S1 P2 T2

R8: high S, high P and high T = S2 P2 T2

With a Gaussian membership function with three levels (low, medium, and high) for
one of the variables and a linear membership function with two levels (low and high) for the
other two, equation (5) would include 12 terms in the numerator and 12 in the denominator.
The Gaussian membership function would be of the following type,
Xi = exp [-0.5 (

X-XC 2
) ]
L

(8)

where Xi denotes the values of S, P, and T, for low, medium, and high levels; X is the

absolute value of the variable concerned; Xc its minimum, central or maximum value; and L

the width of its Gaussian distribution.

The parameters and constants in the previous equation were estimated by using the
ANFIS (Adaptative Neural Fuzzy Inference System) Edit tool in the Matlab 6.5 software
suite.

3. RESULTS AND DISCUSSION


Based on the results of other researchers for pulps from other materials (Lpez et al.
2002, 2003, Pedrola et al. 2004), and after a set of preliminary experiments, the following
ranges were used for the operational variables during the hydrogen peroxide bleaching of
EFB: 0.5 to 3 % o.d.p for soda concentration, 1 to 6 % o.d.p for hydrogen peroxide
concentration, and 1 to 5 h for processing time.
Table 1 shows the experimental values of the bleached pulp properties of the EFB.
During the hydrogen peroxide stage, the brightness increased while the kappa number and
viscosity decreased, depending on the application conditions. In fact, the bleaching effect of

266

Publicaciones Cientficas/Scientific Publications


hydrogen peroxide has been attributed to its ability to react with various coloured carbonylcontaining structures in lignin (Ramos et al. 2008).

Polynomial model
Table 2 lists the coefficients of the different terms in the polynomial equations, as well
as the R2 and the lowest Student t-values for the terms.

The predictions obtained with the previous equations reproduced the experimental

results for the dependent variables with errors less than 2 % for brightness, 4 % for viscosity,
and 10 % for kappa number in 80 % of the cases (always below 16 %) (Table 3).
Table 2. Polynomial models for the properties of bleached pulp (values of the constants in the
polynomial equations).
Dependent
variable

a0

a1

a0

a3

a11

a22

a33

a12

a13

t*

R2

Brightness

78.6

1.7

3.4

2.6

-1.7

-1.1

-2.0

-1.1

-1.2

2.3

0.99

Viscosity

689

-26

-71

-31

1.5

0.96

Kappa number

3.11

-0.49

-0.48

0.45

0.37

4.2

0.92

* Student t-value for the least significant term

The proposed models were validated by conducting two pulping experiments (entries
16 and 17 in Table 1). The errors made in predicting pulp properties by using the
polynomial models were quite low (see Table 3). This testifies to the accuracy of such
models.
The values of the operational variables providing the best bleached pulp properties
(kappa number, viscosity, and brightness) were identified by using multiple non-linear
programming.
Table 4 shows the optimum values of the dependent variables and those of the
operational variables required to obtain them. In all the cases low or medium soda
concentration are required.
Comparing the maximum values of the results of brightness (81.7 %), kappa number
(1.32), and viscosity (826 mL/g) with those obtained by Jimnez et al. (2009b) for EFB pulps
267

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


obtained by soda-anthraquinone and diethanolamine pulping processes, bleached by
A1OpA2ZRP sequence (A, an acid treatment; Op, an oxygen and peroxide stage; Z, an

ozone stage; R, a reductive treatment, and P, a peroxide stage), it may be seen that in both
cases, the kappa number was higher (2.8 and 5.5), the brightness lower (77.5 % and 71.3 %),
and the viscosity lower (653 mL/g and 783 mL/g).
These discrepancies can be explained by more severe treatment of the pulps in
applying the compared sequence, as well as the different nature of the unbleached pulps
(pulping with diethanolamine).
Lpez et al. (2002 and 2003) applied an experimental design of peroxide bleaching to
kraft pulp from olive prunings, finding that the minimum values of the kappa number of the
bleached pulp and the maximum values of the brightness and the viscosity were achieved
when operating at 5 % peroxide and 210 min.
Moreover, Pedrola et al. (2004) used an experimental design in the P stage of the
sequence OXZP (O, an oxygen stage; X, an enzymatic treatment; Z, an ozone stage, and P, a
hydrogen peroxide stage) applied to eucalyptus kraft pulp, with 4 independent variables:
hydrogen peroxide concentration, soda concentration, processing time and temperature.
The models obtained in the cited research predict values for brightness and viscosity
ranging from 84.6 % to 90.5 % and 890 mL/g to 919 mL/g, respectively, with the best results
when the values for processing time and peroxide concentration were the highest and the
temperature and soda concentration used were the mean values, with the peroxide
concentration being the most influential variable.

268

Publicaciones Cientficas/Scientific Publications


Table 3. Values of the dependent variables as estimated with polynomial and neural fuzzy models, and deviations (in %) from their experimental counterparts (in
brackets).
Brightness, %

Polynomial model
Viscosity, mL/g

Brightness, %

Neural fuzzy model


Viscosity, mL/g

Kappa number

Kappa number

63.8 (0.16)

826 (1.47)

4.90 (0.82)

64.0 (0.47)

820 (0.74)

4.97 (2.26)

72.8 (0.14)

666 (1.77)

3.02 (7.65)

73.1 (0.55)

684 (0.88)

3.38 (3.36)

77.4 (0.78)

632 (0.78)

3.18 (0.95)

77.4 (0.52)

642 (0.78)

3.22 (2.22)

71.8 (0.56)

756 (1.95)

3.26 (4.68)

71.8 (0.56)

776 (0.65)

3.49 (2.05)

71.4 (0.14)

764 (1.29)

3.20 (10.36)

71.9 (0.56)

775 (0.13)

3.68 (3.08)

74.6 (0.93)

694 (0.14)

3.04 (2.88)

75.7 (0.53)

696 (0.14)

3.20 (2.24)

80.4 (0.37)

604 (4.14)

1.32 (15.79)

80.5 (0.50)

580 (0.00)

1.25 (9.65)

78.6 (1.16)

689 (2.68)

3.11 (3.67)

77.3 (0.51)

674 (0.45)

3.20 (6.67)

79.2 (0.38)

570 (0.52)

2.96 (0.34)

79.3 (0.51)

574 (0.17)

3.02 (2.37)

10

78.6 (0.51)

663 (3.43)

3.11 (13.50)

79.2 (1.28)

643 (0.31)

2.77 (1.09)

11

79.2 (0.13)

658 (1.79)

2.63 (2.59)

76.8 (3.15)

656 (2.09)

2.94 (8.89)

12

75.2 (1.05)

715 (2.19)

3.11 (1.97)

74.6 (1.84)

722 (1.23)

3.32 (8.85)

13

74.1 (0.14)

760 (0.52)

3.60 (11.46)

71.5 (3.38)

730 (4.58)

3.47 (7.43)

14

74.0 (0.54)

720 (2.42)

3.59 (0.28)

75.4 (1.34)

706 (0.43)

3.63 (0.83)

15

80.9 (0.74)

618 (2.98)

2.62 (6.76)

80.1 (1.72)

627 (1.57)

3.23 (14.95)

16

73.0 (0.41)

755 (1.34)

3.80 (0.79)

72.6 (0.14)

729 (2.15)

3.76 (4.70)

17

80.7 (0.16)

627 (0.16)

2.83 (4.95)

78.1 (1.76)

638 (1.92)

3.23 (8.75)

Exp.

269

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


Table 4. Optimal properties in bleached pulp.
Dependent
variable

Optimum (maximum or
minimum*) value of the
dependent variable

Values of the operational variables required to


obtain the optimum values of the dependent
variables
XS

XP

XT

Brightness, %

81.7

-0.05(1.69 %)

1(6 %)

0.68(4.36 h)

Viscosity, mL/g

826

-1(0.5 %)

-1(1 %)

-1(1 h)

Kappa number

1.32*

-1(0.5 %)

1(6 %)

1(5 h)

Reported results only partially coincide with those obtained in this work, which may be
due to the different nature of the treated materials (prunings of olive and eucalyptus wood
being compared with a non-wood material: EFB), which had undergone different pulping
procedures (kraft pulping being compared with soda pulping).

Figure 1. Variation of brightness of bleached pulps of EFB with soda concentration and processing
time, for a hydrogen peroxide concentration of 6 %.

The polynomial equations (Table 2) allow one to identify the operational variables that
most markedly influencing the bleached pulp properties. The maximum variations in the
dependent variables with changes in the operational variables over the studied range were
obtained by altering one independent variable at a time while keeping all others constant; the
results are shown in Table 5 together with the maximum percentage differences in the
dependent variables from their optimum values over the studied variation ranges.

270

Publicaciones Cientficas/Scientific Publications

Figure 2. Variation of brightness of bleached pulps of EFB with hydrogen peroxide concentration
and processing time, for a soda concentration of 1.69 %.

85

80

X3 -tim
e (h)

1
0

X3 -tim
e (h)

(%

(%

ox
id

X
2 -P
er

60

65

70

X
2 -P
er
ox
id

75

Kappa number

(b)

O)
Brightness (%IS

(a)

(c)

Figure 3. Surfaces estimated for brightness (a), kappa number (b) and viscosity (c) at different soda
concentrations (0.5 %, 1.75 % and 3 %).

271

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307

Figure 4. Representation of pH vs. brightness of all the experiences performed at different soda
concentrations.

Figure 5. Variation of Viscosity/Kappa number ratio with the hydrogen peroxide concentration and
processing time, for a soda concentration of 0.5 %.

As shown in Table 5 and Figs. 1 and 2, the most influential operational variable on the
brightness of bleached pulps is the hydrogen peroxide concentration, being the soda
concentration less influential. These results are consistent with those obtained by Pedrola et
al. (2004) for the case of eucalyptus kraft pulp bleaching, but not with those obtained for
kraft pulp from olive prunings (Lpez et al. 2002, 2003), where the most influential variable
was processing time. Based on Table 5 and similar to Figs. 1 and 2, it follows that for the
viscosity, also the hydrogen peroxide concentration was the most influential variable, being
processing time less influential; these results are consistent with those of Lpez et al. (2002,
2003) and those of Pedrola et al. (2004). It was also found that, for kappa number, the most
272

Publicaciones Cientficas/Scientific Publications


influential variable was the hydrogen peroxide concentration, coinciding with the results
obtained for the bleaching of kraft olive pruning pulp (Lpez et al. 2002, 2003) and kraft
eucalyptus (Pedrola et al. 2004), and the soda concentration variable was the least influential,
in agreement with the results of Pedrola et al. (2004).
Table 5. Maximum changes in the dependent variables with the changes in one operational variable
(constant value of the others).
Variation with the operational variable

Dependent
variable

Soda (0.5 to 3.0 %)

Peroxide (1.0 to 1.6 %)

Time (1 to 5 h)

Brightness, %

1.9(2.33 %)

6.9(8.45 %)

5.5(6.73 %)

Viscosity, mL/g

70(8.47 %)

160(19.37 %)

62(7.51 %)

Kappa number

1.64(124.24 %)

1.88(142.42 %)

1.70(128.79 %)

The percent differences from the changes are given in parenthesis.

The increase of soda concentration from 0.5 to 3 % caused the brightness to increase
(Fig. 3-a). However, the effect of soda concentration on pulp brightness was affected by
peroxide and the processing time, since at low values of these variables the effect was more
appreciable. At a high processing time with high soda concentration, the variation of pulp
brightness decreased. According to Dence and Reeve (1996), for each peroxide
concentration, exists an optimum soda concentration. At a soda concentration above the
optimum, the effectiveness of hydrogen peroxide is reduced and brightness reversion is
observed. On the other hand, the hydrogen peroxide concentration and the processing time
both increased brightness, with their effects being more apparent at low soda concentration.
Kappa number (Fig. 3-b) was strongly decreased by peroxide and the processing time
at low soda concentration. However, this effect was not appreciable at medium and high
soda concentrations. In general, the increase of soda concentration had a detrimental effect
on kappa number. Only at a low peroxide concentration and a short processing time did the
increase of soda concentration decrease kappa number.
Finally, the pulp viscosity (Fig. 3-c) was negatively affected by all the variables, being the
hydrogen peroxide concentration the most influential independent variable. In contrast to
kappa number and brightness, no interactions between variables were found. The decrease
of viscosity was produced because cellulose may undergo depolymerization by reaction with
hydroxyl radicals (Dence and Reeve 1996).
In general, it was observed that an increase of soda concentration had a detrimental
effect on all the properties. As expected, the pH of the effluent was strongly influenced by
273

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


the soda concentration (Fig. 4). At high soda concentration (XS=+1; 3 %) the pH oscillated

between 11.3 and 12.0; at medium soda concentration (XS=0; 1.75 %) the pH oscillated
between 9.8 to 11.4; and at low soda concentration (XS=-1; 0.5 %) the pH was around 9.0.

The pH around 10.5 has been judged to be the most appropriate for an optimal bleaching of
a P stage (Ramos et al. 2009). Surprisingly at low soda concentration and maximum peroxide
concentration and processing time (experiment 7 in Table 1) the pH strongly diminished
(3.75). Under these conditions, the soda was consumed by the inorganic compounds present
in the pulp and it was not possible to maintain the pH. Moreover, in this set of tests a high
value of brightness (80.1 % ISO) and a low value of kappa number (1.14) were obtained.
However, the viscosity was significantly affected (580 mLg-1). In fact, in Dence and Reeve
(1996) it is reported that in acidic media there is a greater viscosity drop produced by
hydrogen peroxide.

Neural fuzzy modeling


Table 6 shows the values of the constants in the neural fuzzy models, ci, as obtained by

using the brightness and viscosity values of Table 1, which was constructed with a Gaussian
membership function for the variable time processing and a linear membership function for
the other two (soda concentration and hydrogen peroxide concentration).
Table 6 also shows the values of the constants in the neural fuzzy models, as obtained
by using the kappa number values of Table 1. This was constructed with a Gaussian
membership function for the variable soda concentration and a linear membership function
for the other two (peroxide concentration and processing time).
The predictions obtained with the previous models reproduced the experimental
results for the dependent variables with errors less than 3 % for the brightness, 5 % for the
viscosity, and 10 % for the kappa number in 93 % of the cases (always below 15 %) (Table 3).
The constants obtained in the neural fuzzy model make it possible to see the influence
of operating variables separately on the characteristics of the pulps. Thus, at low soda and
hydrogen peroxide concentrations, increasing the time processing would increase the
brightness (from 62.9 % to 71.4 %) as per rules 1 and 2 in Table 6. Also, increasing the
hydrogen peroxide concentration at a low soda concentration and time would increase the
brightness from 62.9 % to 73.1 % as per rules 1 and 3.

274

Publicaciones Cientficas/Scientific Publications


Table 6. Values of the constants ci in the neural fuzzy model for the bleached pulp properties.

Soda (%),
peroxide (%),
time (hours)
0.5,1.0,1

0.5,1.0,5

71.4

780

3.76

0.5,6.0,1

73.1

680

3.34

3.0,1.0,1

70.8

786

3.54

0.5,6.0,5

81.2

571

1.11

3.0,1.0,5

75.1

702

3.25

3.0,6.0,1

77.3

641

3.17

3.0,6.0,5

79.4

569

2.99

1.75,1.0,1

2.85

10

1.75,1.0,5

2.41

11

1.75,6.0,1

4.00

12

1.75,6.0,5

3.51

0.5,1.0,3

77.1

678

10

0.5,6.0,3

72.4

766

11

3.0,1.0,3

83.2

575

12

3.0,6.0,3

77.9

666

0.92

0.96

0.95

Rule

R2

Brightness,
%

Viscosity,
mL/g

Kappa
number

62.9

827

5.11

As in the case of polynomial model, the neural fuzzy model could be validated by the
use of experiments 16 and 17 (Table 1). It is found that by using the neural fuzzy model, the
errors in the prediction of the results of the experiments 16 and 17 are minor (Table 3),
which confirms the model tested.
As can be seen, these errors committed in the simulation model using the neural fuzzy
model were similar to those found by applying the polynomial model, even somewhat lower
in the case of kappa number. In view of this, it follows that both models (polynomial and
neural fuzzy) are suitable for simulating the process of bleaching of EFB pulp, and therefore
may be useful to determine the optimal operating conditions. Similar matches in the
application of polynomial and neural fuzzy models have been found for the pulping of
different materials using different processes: vine shoots with ethanolamine (Jimnez et al.
2007), paulownia with ethanol (Caparrs et al. 2008), and Leucaena leucocephala and

Chamecytisus proliferus, vine shoots and cotton stalks, with ethylene glycol (Rodrguez et al.
2008b).

275

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


Optimum operating conditions
First it seems reasonable to operate under such conditions so as to obtain a bleached
pulp with low kappa number and high levels of brightness and viscosity. Such conditions
may be characterized as having a high value of the ratio "viscosity over kappa number".
Adjusting the experimental data of (Viscosity/Kappa number) to a second degree
polynomial, the following equation can be used:
Viscosity
=237 - 26 xS + 27 xP + 39 xT - 48 XS XP - 45 XS XT + 31 XP XT
Kappa number
(R2 = 0.84; t = 2.0)

(9)

Figure 5 shows the changes in the "Viscosity/Kappa number" ratio in terms of two
operating variables (peroxide concentration and processing time), keeping constant the third
independent variable (soda concentration). As shown, the highest "Viscosity/Kappa number"
ratio (453) corresponded to when one is operating with high levels of peroxide concentration
and time and low concentration of sodium hydroxide. However, under these operating
conditions, although they get good values for the kappa number and brightness, viscosity has
a value far below the maximum (27 % less). Therefore, it is more convenient to find the
optimum operating conditions through simulation, using the polynomial and neural fuzzy
models.
According to the experimental design carried out, the soda concentration has to be
diminished, since it has a detrimental effect on all the properties. In addition, the hydrogen
peroxide has to be kept at a minimum due to its negative effect on the pulp viscosity.
Thus, by applying the polynomial models to various combinations of values of the
operational variables, one can identify those providing acceptable bleached pulp properties
(viz. values close to the optimum ones of Table 3) while saving chemical reagents, energy,
and challenges related to capital investments and facilities, through the use of lower
concentrations of soda and peroxide as well as shorter processing times than those required
to obtain the optimum bleached pulp properties. One such combination uses a soda
concentration of 1.13 % and a hydrogen peroxide of 2.25 % for a time of 3 hours. Under
these conditions it was possible to obtain a brightness of 75.1 % (8.1 % below the maximum)
and a viscosity of 740 mL/g (10.4 % less than the maximum value). Under these conditions it
is possible to obtain a value of the ratio Viscosity/Kappa number of 213, well below the
maximum, 453.
276

Publicaciones Cientficas/Scientific Publications


This brightness value (75.1 %) is higher than that found by Law and Jiang (2001): 64.1
% for when soda pulp bleaching of EFB with 5 % hydrogen peroxide, 4.5 % soda, at 70 C
and 1 h are considered. The difference found may be due to different operating conditions
considered in both cases. Comparing the brightness and viscosity values obtained in this
work with those reported by Jimnez et al. (2009b) to bleach soda pulp and diethanolamine
pulp of EFB (653 mL/g and 783 mL/g and 77.5 % and 71.3 %, respectively), shows that the
obtained values are intermediate between those reported by these authors. This can be
explained by different pulping and bleaching processes used.

4. CONCLUSIONS
1. Based on the R2 values they provided (greater than 0.92 in all cases), the
proposed polynomial and neural fuzzy models can be used to accurately
predict bleached pulp properties as a function of the operating conditions.
Both models reproduced with similar accuracy (errors of 10 %) the properties
of bleached pulps of EFB.
2. Neural fuzzy models allow one to assess the influence of each operational
variable on pulp properties. This allows one to compare any two rules
involving identical levels of two variables and differing in the third in order to
make reliable predictions by determining the influence of the third variable on
each pulp property.
3. Polynomial models can additionally be used to establish the influence of each
operational variable on bleached pulp properties (Table 5), albeit not in a
direct manner as with constants ci in the neural fuzzy models.

4. By simulating the process of bleaching pulp EFB, with the polynomial and
neural fuzzy models, it is possible to find optimal values of operating variables,
so that the properties of bleached pulps differ slightly from their best values
and instead will save chemical reagents, energy and plant size, operating with
lower values of operating variables. Thus, operating with a soda concentration
of 1.13 %, a hydrogen peroxide concentration of 2.25 %, for a time of 3 hours,
it was possible to obtain a pulp with a brightness of 75.1 % (8.1 % below the
maximum) and a viscosity of 740 mL/g (10.4 % less than the maximum value).

277

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


ACKNOWLEDGEMENTS
The authors are grateful to Spains DGICyT and MICINN for funding this research
within the framework of the Projects CTQ2007-65074-C02-01, TRA2009-0064,
FUNCICEL-CTQ2009-12904 and ENZFIBRECELL-CTQ2010-20238-CO3-01.

REFERENCES
Aracri, E., and Vidal, T. (2011). Xylanase- and laccase-aided hexenuronic acids and
lignin removal from specialty sisal fibres, Carbohydr. Polym. 83, 1355-1362.
Aziz, A. A., Husin, M., and Mokhtar, A. (2002). Preparation of cellulose from oil
palm empty fruit bunches via ethanol digestion: Effect of acid and alkali catalysts, J. Oil
Palm Research 14, 1, 9-14.
Caparrs, S., Daz, M. J., Ariza, J., Lpez, F., and Jimnez, L. (2008). New
perspectives for Paulownia fortuni L. valorization of the autohydrolysis and pulping
processes, Bioresource Technol. 99, 741-749.
Daud, W. R. W., Law, K. N., and Valade, J. L. (1998). Chemical pulping of oil palm
empty fruit bunches, Cell. Chem. Technol. 32(1-2), 133-143.
Dence, C. W., and Reeve, D. W. (1996). Pulp Bleaching. Principles and Practice,
TAPPI Press, Atlanta, USA.
Freire, C., Silvestre, A., Neto, C., and Evtuguin, D. (2006). Effect of oxygen, ozone
and hydrogen peroxide bleaching stages on the contents and composition of extractives of

Eucalyptus globulus kraft pulps, Bioresource Technol. 97, 420-428.


Ghazali, A., Wanrosli, W. D., and Law, K. N. (2006). Alkaline peroxide mechanical
pulping (APMP) of oil palm lignocellulosics. Part 2. Empty fruit bunch (EFB) responses to
pretreatments, Appita J. 59, 1, 65-70.
Goenadi, D. H., Away, Y., Suharyanto, P., Panji, T., Watanabe, T., and Kuwahara, M.
(1998). Bio-pulping of empty fruit bunches of oil palm by white rot fungi isolated from
tropical plantation, International Conference on Biotechnology in the Pulp and Paper
Industry, 7th, Vancouver, B.C., June 16-19, B, 49-51.

http://www.arbolesornamentales.com, acceded October 2010.

278

Publicaciones Cientficas/Scientific Publications


Ibrahim, R. (2002). Chemical composition of alkaline pulps from oil palm empty
fruit bunches, Oil Palm Bulletin 44, 19-24.
Jang, J. S. R., Sun, C. T., and Mizutani, E. (1997). Neuro-Fuzzy and Soft Computing:
A Computational Approach to Learning and Machine Intelligence, Prentice Hall.
Jimnez, L., Angulo, V., Caparrs, S., and Ariza, J. (2007). Comparison of
polynomial and neural fuzzy models as applied to the ethanolamine pulping of vine shoots,
Bioresource Technol, 98, 3440-3448.
Jimnez, L., Prez, A., Rodrguez, A., and De la Torre, M. J. (2006). New raw
materials and pulping processes for production of pulp and paper, Afinidad 63, 525, 362369.
Jimnez, L., Ramos, E., De la Torre, M. J., Prez, I., and Ferrer, J. L. (2008).
Bleaching of soda pulp of fibres of Musa textilis nee (abaca) with peracetic acid,
Bioresource Technol. 99, 1474-1480.
Jimnez, L., Serrano, L., Rodrguez, A., and Snchez, R. (2009a). Sodaanthraquinone pulping of palm oil empty fruit bunches and beating of the resulting pulps,
Bioresource Technol. 100, 1262-1267.
Jimnez, L., Serrano, L., Rodrguez, A., and Ferrer, A. (2009b). TCF bleaching of
soda-anthraquinone and diethanolamine pulp from oil palm empty fruit bunches,
Bioresource Technol. 100, 1478-1481.
Law, K. N., and Jiang, X. (2001). Comparative papermaking properties of oil-palm
empty fruit bunch, Tappi J. 84, 1, 95.
Lpez, F., Ariza, J., Eugenio, M. E., Daz, J., Prez, I., and Jimnez, L. (2001).
Pulping and bleaching of pulp from olive tree residues, Process Biochem. 37, 1-7.
Lpez, F., Daz, M. J., Eugenio, M. E., Ariza, J., Rodrguez, A., and Jimnez, L.
(2003). Optimization of hydrogen peroxide in totally chlorine free bleaching of cellulose
pulp from olive tree residues, Bioresource Technol. 87, 255-261.
Lpez, F., Eugenio, M. E., Daz, M. J., Prez, I., and Jimnez, L. (2002). Bleaching
of olive tree residues pulp with peracetic acid and comparative study with hydrogen
peroxide, Ind. Eng. Chem. Res. 41, 3518-3525.
Malaysian Palm Oil Promotion Council (MPOPC), http://www.mpopc.org acceded
October 2010.
279

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307


Moldes, D., Cadena, E.M., and Vidal, T. (2010). Biobleaching of eucalypt kraft pulp
with a two laccase-mediator stages sequence, Bioresource Technol. 101, 6924-6929.
Montgomery, D. C. (1991). Design and Analysis of Experiments, Grupo Editorial
Iberoamericana, Mexico, Mexico, 303.
Pedrola, J., Roncero, M. B., Colom, J. F., Vidal, T., and Torres, A. L. (2004).
Application of an experimental design to modeling the hydrogen peroxide stage in TCF
bleaching of eucalypt pulp, Appita J. 57, 141-145.
Quader, M. M. A., and Lonnberg, B. (2005). Pulp and chemical cellulose obtained
with oil palm empty fruit bunches by a modified IDE process, JIRCAS Working Report 39,
47-53.
Roliadi, H., and Pasaribu, R. A. (2004). The possible utilization of empty oil-palm
bunches as a substitute of pulp wood for the manufacture of corrugating-medium
paperboard using semi-chemical soda-anthraquinone pulping, APPITA Annual Conference
and Exhibition, 58th, 2, 335-341.

Ramos, E., Calatrava, S. F., and Jimnez, L. (2008). Bleaching with hydrogen

peroxide. A review, Afinidad 65, 537, 366-373.


Rodrguez, A., Serrano, L., Moral, A., Prez, A., and Jimnez, L. (2008a). Use of
high-boiling point organic solvents for pulping oil palm empty fruit bunches, Bioresource
Technol. 99, 1743-1749.
Rodrguez, A., Prez, A., De la Torre, M. J., Ramos, E., and Jimnez, L. (2008b).
Neural fuzzy model applied to ethylene-glycol pulping of non-wood raw materials,
Bioresource Technol. 99, 965-974.
Roncero, M. B., Torres, A. L., Colom, J. F., and Vidal, T. (2003). TCF bleaching of
wheat straw pulp using ozone and xylanase. Part A: Paper quality assessment, Bioresource
Technol. 87, 305-314.
Shatalov, A. A., and Pereira, H. (2005). Arundo donax L. reed: New perspectives for
pulping and bleaching. Part 4. Peroxide bleaching of organosolv pulps, Bioresource
Technol. 96, 865-872.
Torres, A. L., Roncero, M. B., Colom, J. F., Martnez, J. A., and Vidal, T. (2004).
Application of an experimental design to modeling of ozone bleaching stage in TCF
processes, Ozone Sci. Eng. 26, 443-451.
280

Publicaciones Cientficas/Scientific Publications


Valls, C., Molina, S., Vidal, T., Del Ro, J. C., Colom, S. F., Martnez, A. T.,
Gutirrez, A., and Roncero, M. B. (2009). Influence of operation conditions on laccasemediator removed of sterols from eucalypt pulp, Process Biochem. 44, 1032-1038.
Valls, C., and Roncero, M. B. (2009). Using both xylanase and laccase enzymes for
pulp bleaching, Bioresource Technol. 100, 2032-2039.
Valls, C., Vidal, T., Gallardo, O., Daz, P., Pastor, F. I. J., and Roncero, M. B.
(2010a). Obtaining low-HexA-content cellulose from eucalypt fibres: Which glycosil
hydrolase family is more efficient?, Carbohydr. Polym. 80, 154-160.
Valls, C., Vidal, T., and Roncero, M. B. (2010b). Boosting the effect of a laccasemediator system by using a xylanase stage in pulp bleaching, J. Hazard. Mater. 177, 586592.
Villaverde, J. J., Ligero, P., and Vega, A. (2009). Bleaching Miscanthus x giganteus
acetosolv pulps with hydrogen peroxide/acetic acid. Part 1: Behaviour in aqueous alkaline
media, Bioresource Technol. 100, 4731-4735.
Wanrosli, W. D., Zainuddin, Z., and Lee, L. K. (2004). Influence of pulping
variables on the properties of Elaeis guineensis soda pulp as evaluated by response surface
methodology, Wood Sci. Technol. 38, 3, 191-205.

281

Ferrer et al. (2011). Modeling of H2O2 bleaching/BioResources 6 (2) 1298-1307

282

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 6
A NEURAL FUZZY MODEL APPLIED TO HYDROGEN PEROXIDE BLEACHING
OF NON-WOOD SODA PULPS
A. Rosal, C. Valls, A. Ferrer and A. Rodrguez
Cellulose Chemistry and Technology, 2012, 46 (12), 105114
ndice de impacto: 0.550 (Dcima posicin de 21 revistas en el rea de Materials Science,
paper & wood segn el Journal Citation Reports de 2011)

283

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)

284

Publicaciones Cientficas/Scientific Publications


Cellulose Chem. Technol., 46 (12), 105114 (2012)

A NEURAL FUZZY MODEL APPLIED TO HYDROGEN PEROXIDE BLEACHING OF


NON-WOOD SODA PULPS
A. Rosal, C. Valls*, A. Ferrer** and A. Rodrguez**

Molecular Biology and Biochemical Engineering Department, University Pablo of Olavide, Seville, Spain
*Textile and Paper Engineering Department, Technology University of Catalua, Spain
**Chemical Engineering Department, University of Crdoba, Spain
Received September 16, 2011

ABSTRACT
A neural fuzzy model was used to examine the influence of pulp bleaching variables of
empty fruit bunches from oil palm (EFB) and Hesperaloe funifera, such as soda
concentration (0.5-3 %), hydrogen peroxide concentration (1-10 %) and processing time (1-5
h), on kappa number, brightness and viscosity. The experimental results are reproduced with
errors below 10 % and 15 % for EFB and H. funifera, respectively.
Bleaching pulp simulation permits to obtain optimal values of the operating variables,
so that the properties of bleached pulps will only slightly differ from their best values, while
the lower values of the operating variables will save chemical reagents, energy and plant size.
Thus, if applying 0.5 % soda and 3 % peroxide for 3 h, it is possible to obtain a pulp with a
brightness of 74.9 % and a viscosity of 716 mL/g, for EFB pulp, and 63.3 % and 584 mL/g,
respectively, for H. funifera pulp.

KEYWORDS: Non-wood material; Pulp; Bleaching; Hydrogen peroxide; Neural fuzzy


model.

1. INTRODUCTION
The pulp bleaching plant, which is the most contaminated sector of the paper
manufacturing process, has gone through a number of changes intended to alleviate its
adverse impact. The need to reduce or eliminate the formation of highly toxic
organochlorinated compounds, during bleaching processes, led to the emergence of new
285

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


products on the market, such as ECF (Elemental Chlorine Free) and TCF (Totally Chlorine
Free) pulps1. TCF pulping processes avoid the formation of highly toxic organochlorine
compounds (AOX) during bleaching, by sequences including oxygen-, hydrogen peroxide-

and ozone-based stages2-8. Recently, the enzyme stages involving xylanases or laccase-

mediator systems provided very promising results in pulp bleaching sequences9-12.

Several authors have studied the influence of the operating variables - during pulping

of different materials - on the characteristics of pulps obtained by polynomial and neural


fuzzy models.
The use of polynomial models has also been successfully employed in the study of the
operating variables involved in pulp bleaching2,5,9,10,12-15, but never in the case in which
neural fuzzy models were used.

On the other hand, in recent decades, in some countries, the increasing production of
pulp for paper and other uses raised the problem of the supply of classical raw materials
(mainly hardwoods and softwoods). That is why, a growing emphasis was put on the use of
alternative raw materials, mainly non-wood wastes, such as agricultural and food industry
residues, forest residues and fast-growing plants other than conventional wood16. Alternative

raw materials may be the empty fruit bunches from oil palm (EFB) and Hesperaloe funifera.

In the present study, EFB from Malaysia, a country producing 51 % of the world
production of oil17-19, was used. Each hectare of oil palm produces an average of 10 tons of

fruits per year, which give about 3000 Kg of oil (as the main product)20. EFB is a
lignocellulosic material residue from the palm oil industry. Oil palm fiber is a non-

hazardous, biodegradable material, extracted from oil palm empty fruit bunches (EFB)
through decortication. The fibers can be processed into various dimensional grades to suit
specific applications, such as ceramic and brick manufacturing, thermoplastic filler, erosion
control, soil stabilization, horticulture, paper production, livestock care and compost.
EFB is an interesting raw material, as confirmed by several studies. Numerous authors
have studied the kraft process applied to these residues, the recent results presented by
Ibrahim21 being the most relevant ones. He compared the composition of EFB pulps

obtained by kraft, kraft-anthraquinone, soda and soda-anthraquinone processes; the pulp

obtained by the soda process had the highest content of lignin, holocellulose and -cellulose,
and higher viscosity, compared to pulps obtained by other processes. Law and Jiang22

studied the soda process for producing fibers with higher wall thickness and rigidity, higher
286

Publicaciones Cientficas/Scientific Publications


solubilities in hot water and 1 % soda and higher ash content. These pulps were bleached
with hydrogen peroxide more easily than those of aspen; the paper sheets made from such
pulps had lower tensile index, but higher stretch and tear index than those of aspen.
Moreover, Daud et al.23 pulped EFB by soda, sodium carbonate and sodium sulfite

processes, the first one appearing as the most efficient. With a central composite
experimental design, Wanrosli et al.24 studied the influence of the operating conditions
(temperature, time and alkali concentration) on the properties of EFB pulps (yield, kappa

number and tensile and tear index), obtaining pulp yields in the 30-45 % range. The values
of 160 C, 60-120 min and 20-30 % alkali were considered sufficient for a proper pulping.
Jimnez et al.25 also studied EFB pulping with soda-anthraquinone, considering the

influence of the pulping variables and of subsequent beating. Semichemical pulping for

board by soda-anthraquinone processes was also investigated26. Thermomechanical pulp was

studied by several researchers, the results of Ghazali et al.27 being worth mentioning.
Organosolv processes have also been studied: ethanol28, by modified IDE process29 and

high-boiling point organic solvents30. Finally, biopulping has been investigated31 using a
white fungus, K14.

Hesperaloe funifera is a plant of the Agavaceae family, up to 80 cm tall and 1.0-1.2 m


wide, with long leaves, up to 5 cm wide and 2-3 cm thick. All species of its genus originated
in Mexico and its neighbouring USA regions, where they have mainly ornamental purposes.

H. funifera has very modest irrigation requirements, as it uses the acid metabolism of
Crassulaceans (CAM) for photosynthesis. Its plants fix carbon dioxide and transpire water
more intensely at night than during the day; also, because their coefficient of transpiration is
lower at night, they use water in a highly efficient manner. Based on these properties, H.

funifera might be an effective cellulose raw material in arid zones32. In several regions of

Spain, the cultivation of this species could be particularly interesting. High-density

plantations (27000 per hectare) can yield approximately 20 tons of dry biomass per
hectare33. These crop yields can be increased by a careful control of plant flowering and by

the use of higher planting densities32.

Although the fiber morphology of H. funifera plants is especially suitable for making

cellulose pulp34, little research in this direction appears to have been conducted. In the few

investigations made, the material was subjected to alkaline sulphite-anthraquinone or


mechanical pulping32,35, the resulting paper sheets having very high tensile, burst and tear

indices, which makes them highly suitable for making special paper. More recently, H.

287

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


funifera has been pulped by different processes: soda-anthraquinone36, ethanolamine,

ethylenglycol and diethylenglycol37 and diethanolamine38,39. Rodrguez et al.40 compared

simulations of diethanolamine pulping using polynomial and neural fuzzy models. Similarly,

H. funifera has been used to advantage hemicelluloses, by authohydrolysis, and lignin, by


acidification separation of the pulping waste liquor41. So far, no studies dealing with
bleaching of H. funifera pulp have been reported in the literature of the field.

The aim of this work has been to obtain bleached pulps from EFB and H. funifera
with the best application conditions of the Po stage. A pressurized hydrogen peroxide stage
(Po) was performed after a WAOQ sequence (WA - acid washing; O - oxygen delignification;

Q - chelating stage). For this reason, this Po stage was optimized following a sequential
statistical plan of three variables (soda concentration, hydrogen peroxide concentration and
processing time). The results obtained were analyzed according to the neural fuzzy model.

2. EXPERIMENTAL
Characterization of raw materials
The chemical properties of EFB were determined according to the respective TAPPI
standards, for different components, namely, T-222 for lignin, T-203 0S-61 for -cellulose,
T-9 m-54 for holocellulose, T-204 for ethanol-benzene extractives and T-211 for ash, the
following results being obtained: 24.5 % lignin, 41.9 % -cellulose, 67.0 % holocellulose, 1.2
% extractives and 3.2 % ash. The average fiber length of EFB, determined on a Visopan
projection microscope, was of 0.53 mm.
The samples of H. funifera, for educational and research purposes, were kindly
supplied by the Hesperaloe Project research team at the University of Arizona. The contents
of holocellulose, lignin, -cellulose, ethanol-benzene extractives, and ash of the raw material
were the following: 76.5 %, 7.3 %, 40.9 %, 4.0 %, and 5.9 %, respectively, and fiber length
was of 4.19 mm.

Pulping
The pulps were obtained in an electrically heated 15-L batch cylindrical reactor,
coupled to a control unit through a rotary axle, to ensure proper agitation, including a motor

288

Publicaciones Cientficas/Scientific Publications


actuating the reactor and the required instruments for measurement and control of both
pressure and temperature.
The EFB was cooked in the reactor, using 15 % soda, 170 C, a liquid/solid ratio of 10
and a processing time of 30 min, at a pulp yield of 48.0 %. When operating with 15 % soda,
at 170 C, a liquid/solid ratio of 8, for 40 min, H. funifera provides a yield of 41.5 %. The
operating conditions were selected from the results of other authors22,23,25. Further on, the
cooked material was fiberized in a wet disintegrator at 1200 rpm for 30 min and the

screenings were separated by sieving through a 0.14 mm mesh screen. The pulp obtained
was beaten in a Sprout-Bauer refiner.

Bleaching
The initial pulp properties of EFB were 38.4 % ISO brightness and 9.7 kappa number
for EFB, and 31.2 % and 13.6, respectively, for the pulp of H. funifera. Before applying the
hydrogen peroxide stage (Po), the pulps were washed in an acidic medium (WA), and an
oxygen delignification stage (O) was performed, followed by a chelating stage (Q) (WAOQ
sequence).

The pressurized peroxide bleaching stage (Po) was carried out with 25 g o.d.p. (over
dried pulp) in a 5-L reactor, at 0.6 MPa O2, with 0.5 % o.d.p. of Na2SiO3, 0.2 % o.d.p. of

MgSO4, and 5 % consistency, at 105 C and 60 rpm. The operating sequence was selected
on the basis of previous experimental results2-8,42.

Soda concentration, hydrogen peroxide concentration and processing time were the

three variables of the experimental design, varying over the following ranges: 0.5-3 % o.d.p.
for soda concentration, 1-6 % o.d.p. for hydrogen peroxide and 1-5 h processing time for the
EFB pulp, and 0.5-3 %, 1-10 % and 1-5 h, respectively, for the H. funifera pulp. After the Po
stage, the liquors were recovered for pH measurement and the pulp was efficiently washed
for characterization.

Pulp properties
The treated pulp samples were characterized in terms of kappa number, brightness
and viscosity, according to ISO 302, ISO 3688 and ISO 5351-1, respectively. Kappa number

289

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


was measured twice and brightness four times, in order to calculate standard deviation, which
was found to be 0.1 % for both properties.

Experimental design
The factorial design used43 consisted of a central experiment (in the center of a cube)

and several additional experiments (additional points lying at cube vertices and side centers).

With the three operating variables considered (soda concentration - S, hydrogen peroxide
concentration - P, and bleaching time - T), the total number of experiments was 15.
The integration of fuzzy systems and neural networks combines the advantages of the
two systems and provides an especially powerful modeling tool, viz. the neural fuzzy system,
which uses neural networks as tools in fuzzy systems44. The variation of pulp properties as a

function of the operational variables of the bleaching process can be predicted by the
following expression44:

Ye=

m
l=1 Cl Rl
m
l=1 Rl

(1)

where Ye is the estimated value of the property to be modeled (dependent variable), cl - a

constant parameter and Rl - a fuzzy rule. With the three independent variables, one can
establish the following eight fuzzy rules, according to the extreme (high and low) values of
such variables (X1i - low and X2i - high):

R1: low S, low P and low T: R1 = X1S . X1P . X1T

R2: low S, low P and high T: R2 = X1S . X1P . X2T

R7: low S, high P and high T: R7 = X1S . X2P . X2T

R8: high S, high P and high T: R8 = X2S . X2P . X2T

The low (X1i) and high (X2i) values are obtained with the following equations (linear

membership function):

(X i -Xlowi )

(2)

(X i -X highi )

(3)

X1i =1- X
X2i =1290

highi -X lowi

X highi -X lowi

Publicaciones Cientficas/Scientific Publications


where Xhighi and Xlowi denote the extreme values of the operational variable and Xi - the
absolute value of S, P or T.

With three levels (low, medium and high) for one of the variables and a linear
membership function with two levels (low and high) for the other two, Eq. (1) would include
12 terms in the numerator and 12 in the denominator. Values Xij (low - X1i, medium - X2i,

and high - X3i of the operational variables S, P and T) are obtained with the following

equation (Gaussian membership function):

Xji = exp [-0.5 (

X i -Xci 2
) ]
Li

(4)

where Xi denotes the absolute value of the variable concerned; Xci is its minimum, central or
maximum value; and Li - the width of its Gaussian distribution.

The constant parameters of Eq. (1) were estimated with the ANFIS (Adaptative Neural

Fuzzy Inference System) Edit tool in the Matlab 6.5 software suite.

3. RESULTS AND DISCUSSION


Based on the results of other researchers on pulps obtained from different
materials2,5,14, and after previous tentative experiments, the ranges reported in the

experimental part were applied for the operational variables during hydrogen peroxide
bleaching of EFB and Hesperaloe funifera pulps.
Table 1 shows the experimental values of the bleached pulps properties of EFB and

Hesperaloe funifera. During the hydrogen peroxide stage, brightness increased, while kappa
number and viscosity decreased, as depending on the application conditions.
In fact, the bleaching effect of hydrogen peroxide has been attributed to its ability to
react with various colored carbonyl-containing structures in lignin1.

291

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


Table 1. Normalized values of operational variables and experimental values of bleached pulp
properties.

Hesperaloe funifera

EFB

Exp.

Soda;
peroxide;
time

Brightness
%

Viscosity
mL/g

Kappa
number

Brightness
%

Viscosity
mL/g

Kappa
number

-1; -1; -1

63.7

814

4.86

51.4

619

5.74

-1; +1; -1

72.7

678

3.27

60.9

502

4.93

+1; +1; -1

77.0

637

3.15

61.7

460

7.79

+1; -1; -1

71.4

771

3.42

53.9

620

6.14

-1; -1; +1

71.5

774

3.57

60.0

623

5.25

+1; -1; +1

75.3

695

3.13

57.9

582

5.66

-1; +1; +1

80.1

580

1.14

68.8

451

1.54

0; 0; 0

77.7

671

3.00

62.4

543

5.20

+1; +1; +1

78.9

573

2.95

66.6

456

4.26

10

+1; 0; 0

78.2

641

2.74

62.9

645

5.54

11

0; 0; +1

79.3

670

2.70

63.4

556

4.34

12

-1; 0; 0

76.0

731

3.05

65.4

532

4.05

13

0; -1; 0

74.0

765

3.23

56.0

669

5.41

14

0; 0; -1

74.4

703

3.60

61.2

596

6.52

15

0; +1; 0

81.5

637

2.81

66.4

548

3.99

16

-0.5; -0.5; -0.5

72.7

745

3.83

60.8

620

5.15

17

+0.5; +0.5; +0.5

79.5

626

2.97

66.0

540

4.63

*The values of the operational variables were normalized to values from -1 to +1 by using the
following equation:
)
2(X-X
Xn=
X max -Xmin

where Xn is the normalized value of S, P and T; X is the actual experimental value of the variable
concerned, and the mean of Xmax and Xmin; Xmax and Xmin are the maximum and minimum value,
respectively, of such a variable.

Bleaching of EFB pulp


Table 2 lists the values of the constant in the neural fuzzy model, ci, obtained from the

brightness and viscosity values listed in Table 1, obtained by a Gaussian membership

function, for the time processing variable, and a linear membership function for the other
two (soda concentration and hydrogen peroxide concentration).

292

Publicaciones Cientficas/Scientific Publications


Also, Table 2 shows the values of the constants in the neural fuzzy model, obtained
from the kappa number values given in Table 1, obtained by a Gaussian membership
function for the soda concentration variable, and a linear membership function for the other
two (peroxide concentration and processing time).
The predictions obtained with the previous models reproduced the experimental
results for the dependent variables with errors below 3 % for brightness, 5 % for viscosity and
10 % for kappa number, respectively, in 93 % of the cases (always below 15 %) (Table 3).
The proposed models were validated in two pulping experiments (entries 16 and 17 in
Table 1). The errors made in predicting pulp properties by means of neural fuzzy models
were quite low (Table 3), which testifies for their accuracy.
Therefore, the neural fuzzy system is suitable for simulating the process of EFB pulp
bleaching. Similar matches in the application of neural fuzzy models were proposed in
pulping of various materials by different processes: vine shoots with ethanolamine45,
paulownia with ethanol46, and Leucaena leucocephala, Chamecytisus proliferus, vine shoots

and cotton stalks with ethylene glycol47.

Lpez et al.2,14, who applied an experimental design of peroxide bleaching to kraft

pulp from olive pruning, by using polynomial models, found out that the minimum kappa

number values of the bleached pulp and the maximum brightness and viscosity values are
achieved when operating at 5 % peroxide, for 210 min.
Moreover, Pedrola et al.5 used an experimental design and polynomial models to

describe the P stage of the OXZP sequence (where O stands for the oxygen stage, X - for the
enzymatic treatment, Z - for the ozone stage, and P - for the hydrogen peroxide stage)
applied to eucalyptus kraft pulp, with 4 independent variables: hydrogen peroxide
concentration, soda concentration, processing time and temperature.
The models obtained in the cited research predict brightness and viscosity values
ranging from 84.6 to 90.5 % and from 890 to 919 mL/g, respectively, the best results being
attained when processing time and peroxide concentration were the highest and temperature
and soda concentration recorded mean values, peroxide concentration being found as the
most influential variable.

293

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


Table 2. Values of the constant in the neural fuzzy model for bleached pulp properties.

Hesperaloe funifera

EFB

Exp.

Soda;
peroxide;
time

Brightness
%

Viscosity
mL/g

Kappa
number

Brightness
%

Viscosity
mL/g

Kappa
number

-1; -1; -1

62.9

827

5.11

49.9

605

5.60

-1; -1; +1

71.4

780

3.76

59.4

626

5.14

-1; +1; -1

73.1

680

3.34

60.4

477

3.18

+1; -1; -1

70.8

786

3.54

52.5

614

6.04

-1; +1; +1

81.2

571

1.11

69.1

435

1.63

+1; -1; +1

75.1

702

3.25

57.1

587

5.59

+1; +1; -1

77.3

641

3.17

61.2

437

8.23

+1; +1; +1

79.4

569

2.99

66.7

449

4.54

0; -1; -1

2.85

685

10

0; -1; +1

2.41

565

11

0; +1; -1

4.00

644

12

0; +1; +1

3.51

523

-1; 0; -1

4.96

10

-1; 0; +1

4.02

11

+1; 0; -1

4.69

12

+1; 0; +1

3.73

-1; -1; 0

77.1

678

64.7

10

-1; +1; 0

72.4

766

61.8

11

+1; -1; 0

83.2

575

65.9

12

+1; +1; 0

77.9

666

66.7

R2

0.92

0.96

0.95

0.80

0.88

0.85

*Normalized values of the operational variables

The reported results coincide only partially with those obtained in this work, which
may be due to the different nature of the treated materials (olive prunings and eucalyptus
wood being compared with a non-wood - EFB) and to the different pulping procedures
performed (kraft pulping versus soda pulping).

294

Publicaciones Cientficas/Scientific Publications


Table 3. Values of the dependent variables as estimated with a neural fuzzy model, and deviations (in
%) from their experimental counterparts (in brackets).

Hesperaloe funifera

EFB
Exp.

Brightness %

Viscosity
mL/g

Kappa
number

Brightness %

Viscosity
mL/g

Kappa
number

64.0(0.47)

820(0.74)

4.97(2.26)

50.9(0.97)

614(0.81)

5.56(3.14)

73.1(0.55)

684(0.88)

3.38(3.36)

56.3(7.55)

549(9.36)

5.16(4.67)

77.4(0.52)

642(0.78)

3.22(2.22)

57.8(6.32)

533(15.87)

8.02(2.95)

71.8(0.56)

776(0.65)

3.49(2.05)

53.5(0.74)

622(0.32)

5.96(2.93)

71.9(0.56)

775(0.13)

3.68(3.08)

59.8(0.33)

618(0.80)

5.07(3.43)

75.7(0.53)

696(0.14)

3.20(2.24)

57.7(0.35)

584(0.34)

5.48(3.18)

80.5(0.50)

580(0.00)

1.25(9.65)

64.7(5.96)

522(15.74)

1.77(14.94)

77.3(0.51)

674(0.45)

3.20(6.67)

63.8(2.24)

610(12.34)

4.45(14.42)

79.3(0.51)

574(0.17)

3.02(2.37)

62.5(6.16)

514(12.72)

4.49(5.40)

10

79.2(1.28)

643(0.31)

2.77(1.09)

64.3(2.23)

624(3.26)

5.52(0.36)

11

76.8(3.15)

656(2.09)

2.94(8.89)

61.2(3.47)

554(0.36)

3.92(9.68)

12

74.6(1.84)

722(1.23)

3.32(8.85)

63.3(3.21)

576(8.27)

4.48(10.62)

13

71.5(3.38)

730(4.58)

3.47(7.43)

54.6(2.50)

666(0.49)

4.98(7.95)

14

75.4(1.34)

706(0.43)

3.63(0.83)

63.3(3.43)

596(0.00)

4.86(25.46)

15

80.1(1.72)

627(1.57)

3.23(14.95)

63.4(3.16)

563(2.74)

4.42(10.78)

16

72.6(0.14)

729(2.15)

3.76(4.70)

58.9(3.13)

615(0.81)

4.76(9.40)

17

78.1(1.76)

638(1.92)

3.23(8.75)

64.1(2.88)

530(1.85)

4.73(2.16)

Figure 1 shows how kappa number varies with the concentrations of soda and
peroxide over short, medium and long time intervals. It is observed that the minimum value
of kappa number corresponds to the pulp bleached at high levels of peroxide concentration
and processing time variables, and at a low value of soda concentration.

295

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


Short time
Medium time
Long time

Figure 1. Variation of kappa number of EFB bleached pulps with soda concentration and hydrogen
peroxide concentration, for short, medium and long processing times.

Figure 2 shows that viscosity was maximum when bleaching was performed at low
values of the operating variables. Finally, Figure 3 shows that brightness was the highest when
operating at high peroxide concentration and low soda concentration, over long time
intervals.
Short time
Medium time
Long time

Figure 2. Variation of viscosity of EFB bleached pulps with soda concentration and hydrogen
peroxide concentration, for short, medium and long processing times.

296

Publicaciones Cientficas/Scientific Publications


Low soda concentration
Medium soda concentration
High soda concentration

Figure 3. Variation of brightness of EFB bleached pulps with hydrogen peroxide concentration and
processing time, for low, medium and high soda concentrations.

According to Figure 1 (illustrating the variations of kappa number with soda


concentration and time, and also with peroxide concentration and time, at low, medium and
high values of the other independent variables), the most influential operational variables on
the kappa number of bleached pulps were hydrogen peroxide concentration and time, which
coincides with the results obtained for the bleaching of kraft olive pruning pulp2,14 and kraft
eucalyptus5, while soda concentration was the least influential variable, which coincides with
the results of Pedrola et al.5 kappa number (Fig. 1) strongly decreased when increasing

peroxide and processing time at low soda concentration. However, this effect was less
notable at medium and high soda concentration values. Generally, the increase in soda
concentration had a detrimental effect on kappa number.
Similarly, as shown in Figure 2, there follows that, in the case of viscosity, hydrogen
peroxide concentration is again the most influential variable, the least influential ones being
processing time and soda concentration; these results are consistent with those of Lpez et
al.2,14 and Pedrola et al.5 Pulp viscosity (Fig. 2) was negatively affected by all variables, the

strongest effect being produced by hydrogen peroxide concentration. The decrease in


viscosity occurred because cellulose may undergo depolymerization by the reaction with
hydroxyl radicals45.

Another observation is that, for brightness (Fig. 3), the most influential operational

variable is hydrogen peroxide concentration, while the least influential one is soda
concentration. These results are consistent with those obtained by Pedrola et al.5 for
297

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


bleaching of eucalyptus kraft pulp, but not with those obtained for kraft pulp from olive
pruning2,14, where the most influential variable was found to be processing time. Generally,

the increase of soda concentration from 0.5 to 3 % increases brightness, as well (Fig. 3).

However, the effect of soda concentration on pulp brightness was affected by peroxide
concentration and processing time, since, at low values of these variables, the effect was more
noticeable. At high processing time and high soda concentration, pulp brightness decreased
(Fig. 3). According to Dence and Reeve48, for each peroxide concentration, there exists an

optimum soda concentration. At soda concentrations above the optimum, the effectiveness

of hydrogen peroxide is reduced and brightness reversion is observed. On the other hand,
hydrogen peroxide concentration and processing time increased brightness, their effect being
more visible at low soda concentration.
A comparison between the maximum values of the results for brightness (80.5 %), and
viscosity (820 mL/g) and the minimum value of kappa number (1.25) with those obtained by
Jimnez et al.49 for EFB pulps, obtained by soda-anthraquinone and diethanolamine
processes, bleached by the A1OpA2ZRP sequence (A, an acid treatment; Op, an oxygen

and peroxide stage; Z, an ozone stage; R, a reductive treatment, and P, a peroxide stage),

permits the observation that, in both cases, kappa number is higher (2.8 and 5.5), brightness
is lower (77.5 % and 71.3 %), and viscosity is also lower (653 mL/g and 783 mL/g); these
discrepancies can be explained by the more severe treatment applied to the pulps in the
compared sequence, and also by the different nature of the unbleached pulps (pulping with
diethanolamine).

Bleaching of Hesperaloe funifera pulp


Table 2 lists the values of constant ci in the neural fuzzy model. For the kappa

number, a Gaussian membership function was used for hydrogen peroxide concentration,
and a linear membership function for the other two (soda concentration and processing
time) while, for viscosity, a Gaussian membership function has been used for the soda
concentration variable and, for brightness, a Gaussian membership function for the
processing time variable.
The predictions obtained with the previous models reproduced the experimental
results for the dependent variables with errors below 8 % for brightness, 15 % for viscosity
and 15 %, respectively, for kappa number (in more than 90 % of the cases) (Table 3).
298

Publicaciones Cientficas/Scientific Publications


As in the case of EFB pulp, experiments 16 and 17 (Table 1) permit to validate the
neural fuzzy model. By the utilization of the neural fuzzy model, the errors possibly
occurring in the prediction of the results of experiments 16 and 17 are reduced (Table 3).
The analysis of Figures 4 to 6 permits similar conclusions as in the case of EFB pulp
bleaching. Thus, Figure 4 shows that the minimum value of kappa number corresponds to
the H. funifera pulp bleached at high peroxide concentration, long processing time and low
value of soda concentration. Figure 5 shows that viscosity is maximum during bleaching
performed at low values of peroxide concentration and processing time, and at the mean
value of soda concentration.
Finally, Figure 6 shows that brightness is the highest when operating at high levels of
peroxide concentration and time and at low soda concentration. Of the dependent variables
considered (kappa number, viscosity and brightness), the most influential variable is
peroxide concentration, and the least - soda concentration.
Short time
Medium time
Long time

Figure 4. Variation of kappa number of H. funifera bleached pulps with soda concentration and
hydrogen peroxide concentration, for short, medium and long processing times.

The maximum values for brightness (68.6 %) and viscosity (680 mL/g) and the
minimum value for kappa number (1.63) are lower than those found for EFB pulp:
brightness - 80.5 %, viscosity - 820 mL/g and kappa number - 1.25.

299

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


Low soda concentration
Medium soda concentration
High concentration

Figure 5. Variation of viscosity of H. funifera bleached pulps with hydrogen peroxide concentration
and processing time, for low, medium and high soda concentrations.
Short time
Medium time
Long time

Figure 6. Variation of brightness of H. funifera bleached pulps with soda concentration and hydrogen
peroxide concentration, for short, medium and long processing times.

Optimum operating conditions


EFB
According to the experimental design carried out, soda concentration has to be
diminished, since it has a detrimental effect upon all properties.
In addition, the hydrogen peroxide concentration should not be high, because of its
negative effect on pulp viscosity.
Thus, by applying fuzzy models to various combinations of values of the operational
variables, one can identify those providing acceptable bleached pulp properties (viz. values
300

Publicaciones Cientficas/Scientific Publications


close to the optimum ones: brightness - 80.5 %, viscosity - 820 mL/g and kappa number 1.25) while saving chemical reagents, energy and the industrial resources for capital
investments in facilities, by using lower concentrations of soda and peroxide and shorter
processing times than those required to obtain optimum bleached pulp properties. One such
combination uses a soda concentration of 0.5 % and a hydrogen peroxide concentration of 3
%, for a time of 3 h. Under such conditions, a brightness value of 74.9 % (7.0 % below the
maximum) and a viscosity value of 716 mL/g (12.7 % less than the maximum value) were
attained. This value of brightness (74.9 %) is higher than that found by Law and Jiang22,

namely 64.1 % for bleaching EFB soda pulp with 5 % hydrogen peroxide, 4.5 % soda, at 70
C, for 1 h. The difference found may be due to the different operating conditions
considered in these cases. A comparison between the brightness and viscosity values
obtained in the present work and those reported by Jimnez et al.49 for bleaching EFB soda

pulp and diethanolamine pulp (653 mL/g and 783 mL/g, 77.5 % and 71.3 %, respectively)
shows that the obtained values are intermediary ones. This can be explained by the different
pulping and bleaching processes used.

Hesperaloe funifera
As in the case of EFB, by applying fuzzy models for H. funifera pulp to various
combinations of values of the operational variables, one can identify those providing
acceptable properties for bleached pulps (viz., values close to the optimum ones: brightness 68.6 %, viscosity - 680 mL/g and kappa number - 1.63), while saving chemical reagents,
energy and the industrial resources for capital investments in facilities, by the application of
lower soda and peroxide concentrations and shorter processing times than those required to
obtain optimum properties of the bleached pulps. One of these combinations uses a soda
concentration of 0.5 %, and a hydrogen peroxide concentration of 3 %, for 3 h. Under these
conditions - similar to those used for EFB pulp - a brightness of 63.3 % (7.7 % below the
maximum) and a viscosity of 584 mL/g (14.1 % less than the maximum value) could be
attained.

4. CONCLUSIONS
The neural fuzzy models applied for bleaching pulps from EFB and H. funifera are
suitable to predict the brightness and viscosity values of bleached pulps, the possible errors
recorded being, in the worst case, below 10 % for EFB pulp and 15 % for H. funifera pulp.

301

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


By simulating the bleaching of EFB and H. funifera pulps by the neural fuzzy models,
the optimal values for the operating variables may be established, so that the properties of
bleached pulps will differ only slightly from their best values; besides, chemical reagents,
energy and plant size will be saved, once operating at lower values of the operating variables.
Thus, operating at a soda concentration of 0.5 %, a hydrogen peroxide concentration of 3 %
for 3 h, a pulp with a brightness of 74.9 % (7.0 % below the maximum) and a viscosity of 716
mL/g (12.7 % less than the maximum value) may be obtained for the EFB pulp, and of 63.3
% (7.7 % below the maximum) and of 584 mL/g (14.1 % less than the maximum value)
respectively, for the H. funifera pulp.

ACKNOWLEDGEMENTS
The authors are grateful to the DGICyT and MICINN of Spain for funding this
research within the framework of the Projects CTQ2007-65074-C02-01, TRA2009-0064,
CTQ2010-19844-C02-01, CTQ2009-12904 and CTQ2010-20238-CO3-01.

REFERENCES
(1)

E. Ramos, S. F. Calatrava and L. Jimnez, Afinidad, 65, 366 (2008).

(2)

F. Lpez, M. J. Daz, M. E. Eugenio, J. Ariza, A. Rodrguez and L. Jimnez,


Bioresource Technol., 87, 255 (2003).

(3)

M. B. Roncero, A. L. Torres, J. F. Colom and T. Vidal, Bioresource Technol., 87,


305 (2003).

(4)

A. L. Torres, M. B. Roncero, J. F. Colom, J. A. Martnez and T. Vidal, Ozone Sci.


Eng., 26, 443 (2004).

(5)

J. Pedrola, M. B. Roncero, J. F. Colom, T. Vidal and L. Torres, Appita J., 57, 141
(2004).

(6)

A. A. Shatalov and H. Pereira, Bioresource Technol., 96, 865 (2005).

(7)

C. Freire, A. Silvestre, C. Neto and D. Evtuguin, Bioresource Technol., 97, 420


(2006).

(8)

J. J. Villaverde, P. Ligero and A. Vega, Bioresource Technol., 100, 4731 (2009).

(9)

C. Valls and M. B. Roncero, Bioresource Technol., 100, 2032 (2009).

302

Publicaciones Cientficas/Scientific Publications


(10) C. Valls, S. Molina, T. Vidal, J. C. Del Ro, J. F. Colom, A. T. Martnez, A. Gutirrez
and M. B. Roncero, Process Biochem., 44, 1032 (2009).
(11) C. Valls, T. Vidal, O. Gallardo, P. Daz, F. I. J. Pastor and M. B. Roncero, Carbohydr.
Polym., 80, 154 (2010).
(12) C. Valls, T. Vidal and M. B. Roncero, J. Hazard. Mater., 177, 586 (2010).
(13) F. Lpez, J. Ariza, M. E. Eugenio, J. Daz, I. Prez and L. Jimnez, Process Biochem.,
37, 1 (2001).
(14) F. Lpez, M. E. Eugenio, M. J. Daz, I. Prez and L. Jimnez, Ind. Eng. Chem. Res.,
41, 3518 (2002).
(15) L. Jimnez, E. Ramos, M. J. de la Torre, I. Prez and J. L. Ferrer, Bioresource
Technol., 99, 1474 (2008).
(16) L. Jimnez, A. Prez, A. Rodrguez and M. J. de la Torre, Afinidad, 63, 362 (2006).
(17) Movimiento mundial por los bosques tropicales, Oil palm of comestics to biodiesel,
Novib and Rel-UITA 2006, 10. Available at
http://www.wrm.org.uy/plantaciones/material/PalmaAceitera
(18) L. K. Chiew and Z. Rahman, J. Oil Palm Research, 14, 1 (2009).
(19) H. D. Rozman, P. P. Lim, A. Abusamah, R. N. Kumar, H. Ismail and Z. A. Mohd.
Ishak, Int. J. Polym. Mater., 44, 179 (1999).
(20) J. M. Snchez, Gua de las plantas ornamentales, Ediciones Mundi-Prensa, 2001,
Avalaible at http://www.arbolesornamentales.es/index.htm
(21) R. Ibrahim, Oil Palm Bulletin, 44, 19 (2002).
(22) K. N. Law and X. Jiang, Tappi J., 84, 95 (2001).
(23) W. R. W. Daud, K. N. Law and J. L. Valade, Cellulose Chem. Technol., 32, 133
(1998).
(24) W. D. Wanrosli, Z. Zainuddin and L. K. Lee, Wood Sci. Technol., 38, 191 (2004).
(25) L. Jimnez, L. Serrano, A. Rodrguez and R. Snchez, Bioresource Technol., 100,
1262 (2009).
(26) H. Roliadi and R. A. Pasaribu, in Procs. 58th APPITA Annual Conference and
Exhibition, Camberra, April 19-21, 2004, pp. 335-341.

303

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)


(27) A. Ghazali, W. D. Wanrosli and K. N. Law, Appita J., 59, 65 (2006).
(28) A. A. Aziz, M. Husin and A. Mokhtar, J. Oil Palm Research, 14, 9 (2002).
(29) M. M. A. Quader and B. Lonnberg, JIRCAS Working Report, 39, 47 (2005).
(30) A. Rodrguez, L. Serrano, A. Moral, A. Prez and L. Jimnez, Bioresource Technol.,
99, 1743 (2008).
(31) D. H. Goenadi, Y. Away, P. Suharyanto, T. Panji, T. Watanabe and M. Kuwahara, in
Procs. 7th International Conference on Biotechnology in the Pulp and Paper Industry,

Vancouver, June 16-19, 1998, pp. 37-44.

(32) S. P. McLaughlin, Econ. Bot., 54, 192 (2000).


(33) A. Wong and S. P. McLaughlin, in Procs. Tappi Pulping Conference, Boston,
November 6-9, 2000, pp. 1-14.
(34) S. P. McLaughlin, J. Arid. Environ., 55, 143 (2003).
(35) M. Fairbank and R. Detrick, Tappi J., 83, 66 (2000).
(36) R. Snchez, A. Rodrguez, A. Requejo, A. Ferrer and E. Navarro, Bioresource
Technol., 101, 7032 (2010).
(37) R. Snchez, A. Rodrguez, E. Navarro, J. A. Conesa and L. Jimnez, Biomass
Bioenerg., 34, 1471 (2010).
(38) R. Snchez, A. Ferrer, L. Serrano, A. Toledano, J. Labidi and A. Rodrguez,
Bioresources, 6, 3 (2010).
(39) R. Snchez, A. Rodrguez, A. Requejo, A. Garca and L. Jimnez, Cellulose Chem.
Technol., 44, 327 (2010).
(40) A. Rodrguez, A. Snchez, A. Ferrer and A. Requejo, Chem. Eng. Res. Des., 89, 648
(2011).
(41) R. Snchez, A. Rodrguez, J. C. Garca, A. Rosal and L. Jimnez, Bioresource
Technol., 102, 1308 (2011).
(42) A. Ferrer, A. Rosal, C. Valls, B. Roncero and A. Rodrguez, Bioresources, 6, 1298
(2011).
(43) D. C. Montgomery, in Grupo Editorial Iberoamericana, Mexico, 1991, 303.

304

Publicaciones Cientficas/Scientific Publications


(44) J. S. R. Jang, C. T. Sun and E. Mizutani, in Neuro-Fuzzy and Soft Computing: A
Computational Approach to Learning and Machine Intelligence, Prentice Hall,
Chandigarth, India, 1997, pp. 1-607.
(45) L. Jimnez, V. Angulo, S. Caparrs and J. Ariza, Bioresource Technol., 98, 3440
(2007).
(46) S. Caparrs, M. J. Daz, J. Ariza, F. Lpez and L. Jimnez, Bioresource Technol., 99,
741 (2008).
(47) A. Rodrguez, A. Prez, M. J. De la Torre, E. Ramos and L. Jimnez, Bioresource
Technol., 99, 965 (2008).
(48) C. W. Dence and D. W. Reeve, Pulp Bleaching. Principles and Practice, TAPPI
Press, Marietta, USA, 1996.
(49) L. Jimnez, L. Serrano, A. Rodrguez and A. Ferrer, Bioresource Technol., 100, 1478
(2009).

305

A. Rosal et al./Cellulose Chem. Technol. 46 (1-2), 105-114 (2012)

306

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 7
BIOBLEACHING OF PULP FROM OIL PALM EMPTY FRUIT BUNCHES WITH
LACCASE AND XYLANASE
R. Martn-Sampedro, A. Rodrguez, A. Ferrer, L.L. Garca-Fuentevilla, M.E. Eugenio
Bioresource Technology, 2012, 110, 371378
ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de Agricultural
Engineering segn el Journal Citation Reports de 2011)

307

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378

308

Publicaciones Cientficas/Scientific Publications


Bioresource Technology 110 (2012) 371378

BIOBLEACHING OF PULP FROM OIL PALM EMPTY FRUIT BUNCHES WITH


LACCASE AND XYLANASE
R. Martn-Sampedroa, A. Rodrguezb, A. Ferrerb, L.L. Garca-Fuentevillaa, M.E. Eugenioa
Instituto Nacional de Investigacin y Tecnologa Agraria y Alimentaria INIA, Carretera de la Corua,
km 7.5, 28040 Madrid, Spain
b Departamento de Ingeniera Qumica, Universidad de Crdoba, Campus de Rabanales, Edicio Marie
Curie, 14071 Crdoba, Spain
a

Received 21 September 2011; Received in revised form 17 January 2012; Accepted 19 January 2012; Available
online 3 February 2012

ABSTRACT
Laccase and xylanase were tested for their suitability for biobleaching of sodaanthraquinone pulp from oil palm empty fruit bunches (EFB). An enzymatic stage with
xylanase (X) and/or laccase (L) was incorporated before the alkaline extraction stage (E) and
the hydrogen peroxide bleaching stage (P). Compared with controls, the LEP sequence
resulted in an improvement of optical properties (brightness and colorimetric properties)
and a reduction of the kappa number. When xylanase and laccase were used jointly, no
improvement was detected, however, when the xylanase application preceded the laccase
stage, the benecial effects of laccase were boosted. Thus, the nal XLEP bleached pulp
showed a kappa number of 5.4 and a brightness of 60.5 % ISO, although the hydrogen
peroxide consumption increased (77.0 % vs. 64.5 % and 73.8 % for EP and LEP
respectively). Finally, after subjecting the bleached pulps to accelerated ageing, the best
optical properties were observed in the XLEP pulp.

KEYWORDS: Oil palm empty fruit bunches (EFB); Alternative raw materials; Xylanase;
Laccase; Biobleaching.

309

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


1. INTRODUCTION
Currently 6.40 x 108 m3 of wood are used each year to manufacture pulp and paper

worldwide (FAO, 2011). Societys awareness of the importance of environmental problems


demands an effort by the paper industry to utilize new, fast-growing high-quality species and
to implement new and less polluting processes for cooking of raw materials and bleaching of
cellulose pulps. For this reason, there is growing interest in using alternative feedstock, such
as non-woody materials, fast growing plants, and residues from agricultural, food and forest
industries (Jimnez et al., 2006). An interesting non-woody material is empty fruit bunches
(EFB), which is the residual lignocellulosic material from the palm oil industry. Malaysia is
currently the worlds largest producer of palm oil, supplying 51 % of the worlds output.
Cultivation of oil palms is currently expanding in Western Africa (Nigeria, Guinea, Ghana),
Latin America (Ecuador, Colombia, Honduras) and Asia (Thailand) (Rodrguez et al.,
2008).
Several authors have used EFB as a raw material for the production of cellulose pulps.
Ibrahim (2002) and Jimnez et al. (2009a) used the kraft process, while Daud et al. (1998)
and Law and Jiang (2001) obtained pulps by treating EFB with soda, sodium carbonate and
sodium sulte. Ibrahim (2002) compared the composition of EFB pulps obtained with kraft,
kraft-anthraquinone, soda and soda-anthraquinone processes and found that the soda
process yielded the highest content of lignin, holocellulose and -cellulose plus the highest
viscosity.
The semichemical pulping for cardboard by a soda-anthraquinone process has been
studied by Roliadi and Pasaribu (2004). Several researchers have tested the production of
thermomechanical pulp, most notably Daud et al. (2005). Furthermore, organosolv pulping
with ethanol, with the modied IDE process and with high-boiling point organic solvents has
been investigated (Quader and Lonnberg, 2005; Rodrguez et al., 2008). Finally, the
bleachability of different EFB pulps has been evaluated (Jimnez et al., 2009b; Law and
Jiang, 2001). All these reports proved that EFB was a suitable alternative raw material for the
production of pulp and paper with acceptable properties.
The use of enzymes to treat pulps before applying standard bleaching sequences is
considered a valuable alternative to chemical treatments. These enzymes could help to
address the environmental concerns and the low selectivity inherent to the application of the
elemental chlorine free (ECF) and totally chlorine free (TCF) bleaching sequences. Laccases
have been the most used enzymes for this purpose. The interest of laccases lies in their
310

Publicaciones Cientficas/Scientific Publications


ability to oxidize a wide variety of phenolic compounds (Thurston, 1994), requiring only
molecular oxygen. Moreover, the oxidative ability of laccases can be extended to nonphenolic structures through the action of some low molecular weight compounds, the
mediators. Once oxidized by the laccases, the mediators become stable radicals which may
continue oxidizing other compounds that were not used directly as substrates by the enzyme
(Bourbonnais and Paice, 1992). The application of both laccases and mediators is
commonly known as the laccase-mediator system (LMS). Moreover, xylanases have been
applied to biobleaching on account of their ability to enhance the bleaching effect of
chemical reagents, thereby affording substantial savings and, more importantly considerably
diminishing the production of pollutants during bleaching (Valls and Roncero, 2009).
There are numerous reports on biobleaching using xylanases or laccases (Chauhan et
al., 2006; Eugenio et al., 2011). However, few studies address the topic of using both
enzymes in sequence or jointly (Kapoor et al., 2007; Oksanen et al., 1997; Valls and
Roncero, 2009) and no studies have utilized these enzymes on EFB pulps.
Consequently, the main objective of this study was to evaluate the feasibility of using
commercial xylanase and laccase, in sequence or jointly, as a treatment prior to a standard
TCF bleaching sequence acting on soda-anthraquinone pulp from EFB. To this end,
delignication degree, viscosity, brightness, CIE L*a*b* and CIE L*C* color coordinates as
well as mechanical strength properties were analyzed. Also, the consumption of hydrogen
peroxide was determined. Finally, the optical properties were determined before and after
an accelerated ageing process of the resulting paper sheets.

2. EXPERIMENTAL
2.1. Chemicals and raw material
2,2-Azino-bis-3-ethylbenzthiazoline-6-sulfonate (ABTS) was purchased from Roche
(Madrid, Spain). Acetosyringone (ASG), NaOH and H2O2 were reagent-grade and obtained

from Merck (Barcelona, Spain) or Sigma-Aldrich (Madrid, Spain). Novozym 51003 (laccase)
and Pulpzyme HC (xylanase), produced by submerged fermentation of genetically modied

Aspergillus sp. and Bacillus sp. respectively, were donated by Novozymes (Bagsvaerd,
Denmark).
The Empty Fruit Bunches (EFB) was supplied by Straw Pulping Engineering S.L.
(Zaragoza, Spain). The chemical composition of this raw material was 1.2 % extractives, 3.2
311

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


% ash, 24.5 % lignin, 41.9 % -cellulose, 67.0 % holocellulose and 25.1 % hemicelluloses
(Ferrer et al., 2011).

2.2. Pulp production


EFB pulp was obtained using a 15-L batch cylindrical reactor that was heated by means
of electrical wires. The reactor was equipped with a rotating axle to ensure proper agitation.
The reactor was connected to a control unit for instruments required for measuring and
controlling pressure and temperature. The raw material was cooked with sodaanthraquinone in the reactor under the following conditions: 170 C, 15 % NaOH, 1 % AQ,
30 min and 10:1 liquid/solid ratio (Daud et al., 1998; Ferrer et al., 2011; Law and Jiang,
2001). The cooked material was berized in a wet disintegrator at 1200 rpm for 30 min and
the particles were separated by sieving through a screen of 0.16 mm mesh.

2.3. Enzymatic pre-treatments


EFB pulp was subjected to a joint laccase/xylanase (XL) or a xylanase treatment
followed by a laccase treatment (X + L).
Experiments were performed using 50 g of EFB pulp, which was introduced into the
reactors with the commercial laccase and chemicals. The mixture was intensively mixed
before adding oxygen at a pressure of 6 Kg per cm2 and submerging the reactors in a

thermostatic bath. Consistency, incubation time, laccase dose, and mediator concentration
were xed during the laccase treatment at 10 %, 2 h, 17.5 U/g over dried pulp (o.d.p), and

0.05 mmol/g o.d.p acetosyringone (4-hydroxy-3,5-dimethoxyacetophenone), respectively (1


U of laccase is dened as the amount of laccase required to convert 1 mol/min of ABTS to
its cationic radical in 0.1 M citrate buffer, pH 5.2 at 24 C). Temperature and pH were set at
45 C and pH 6.5, as per product specication sheet. A few drops of 0.05 % (v/v) Tween 80
were added to all assays in order to improve the interaction between enzyme and substrate.
In those instances when xylanase was added to test its inuence during the laccase treatment,
the concentration used was 20 AXU/g o.d.p. When the xylanase treatment was carried out as
a step prior to the laccase treatment, conditions were as follows: consistency, 10 %;
incubation time, 2 h; 20 AXU/g o.d.p; temperature, 45 C; and pH, 6.5 (as per the product
specication sheet).

312

Publicaciones Cientficas/Scientific Publications


Controls were included in the experimental design as follows:
i.

EP: all enzymatic treatments were omitted (designated as conventional


bleaching).

ii. LcEP: bleaching carried out in the absence of both laccase and mediator, but
in the presence of the rest of components used in the laccase treatment at pH
6.5, and O2 (6 Kg/cm2); and kept at 45 C for 2 h. The LcEP control is
equivalent to the (X + L)cEP control, so this second control was not assayed.

iii. XcLcEP: bleaching carried out with two enzymatic treatments without the
presence of the corresponding enzymes, but in the presence of the other
components used in both enzymatic treatments at pH 6.5; O2 (6 Kg/cm2, only
in Lc stage); and kept at 45 C for 2 h.

2.4. Alkaline extraction and hydrogen peroxide treatments


After the enzymatic pre-treatment, pulps were washed with distilled water until
reaching neutral pH, air dried at room temperature, and subjected to an alkaline extraction
under the following conditions: NaOH, 1.5 % o.d.p; consistency, 5 %; temperature, 90 C;
and treatment time, 120 min. Thereafter, pulps were washed with distillated water again and
a hydrogen peroxide bleaching stage was applied under the following conditions: H2O2, 3 %

o.d.p; NaOH, 1.5 % o.d.p; DTPA, 1 % o.d.p; MgSO4-7H2O, 0.2 % o.d.p; consistency, 5 %;
temperature, 90 C; and treatment time, 90 min. Residual hydrogen peroxide was analyzed

in the bleaching efuent by standard iodometric titration.

2.5. Pulp and paper characterization


Treated pulps were characterized according to the corresponding standards in terms of
their kappa number (UNE 57-034), hexenuronic acid (HexA) content (Gellerstedt and Li,
1996) and viscosity (UNE 57-039-92). Chemical composition was also determined: ethanol
extractives (T204 cm-07), lignin (T222 om-88), holocellulose (T9 m-54), -cellulose (T203
OS-61). Hemicellulose content was estimated as the difference between holocellulose and cellulose contents.
Handsheets were obtained from the treated pulps in accordance with ISO UNE-EN
ISO 5269-2 and characterized in terms of tensile index, stretch and tear index (UNE-EN
313

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


ISO 5270). Brightness (UNE 57-062) and CIE L*a*b* and CIE L*C* coordinates (T527)
were also determined in the handsheets using a spectrophotometer ELREPHO 070
(Lorentze and Wettre).

2.6. Accelerated ageing


The bleached and the unbleached handsheets were subjected to accelerate ageing to
analyze changes in their optical properties. This accelerated ageing was carried out in a
climatic test cabinet CTS (model C-20/250/S) and consisted in a moist heat treatment at 80
C and 65 % relative humidity for 6 days, as specied by the standard UNE 57-092-4. After
accelerated ageing, pulps were characterized in terms of brightness and CIE L*a*b* and
CIE L*C* coordinates, also in keeping with the aforementioned standards.

3. RESULTS AND DISCUSSION


Initially, the kappa number, viscosity and HexA content for the EFB pulp obtained in
this study were 11.8, 896 mL/g, and 21.5 mol/g, respectively. The chemical composition of
this unbleached pulp is shown in Table 1. Subsequently, several bleaching sequences, as
described above, were applied to the EFB pulp, proceeding afterwards to the
characterization of the bleached pulps and handsheets.
Table 1. Chemical composition of EFB unbleached and bleached pulps (%).
Unbleached

LcEPa

LEPa

(X+L)EPa

XcLcEPa

XLEPa

EPa

Ethanol
extractives

0.9

0.5

0.4

0.3

0.5

0.2

0.4

Lignin

2.3

1.3

1.2

1.0

1.3

0.9

2.3

Holocellulose

96.7

98.9

98.7

98.8

99.1

98.3

98.8

-Cellulose

84.2

85.1

84.5

85.4

85.5

85.9

84.8

Hemicelluloses

12.5

13.8

14.2

13.4

13.6

12.4

14.0

aBleaching

314

sequences:

LcEP: Laccase control treatment + alkaline extraction + hydrogen peroxide bleaching

LEP: Laccase treatment + alkaline extraction + hydrogen peroxide bleaching

(X + L)EP: Laccase and xylanase treatment + alkaline extraction + hydrogen peroxide


bleaching

Publicaciones Cientficas/Scientific Publications

XcLcEP: Xylanase control treatment + laccase control treatment + alkaline extraction


+ hydrogen peroxide bleaching

XLEP: Xylanase treatment + laccase treatment + alkaline extraction + hydrogen


peroxide bleaching

EP: Alkaline extraction + hydrogen peroxide bleaching

3.1. Pulp characterization


Fig. 1 shows the kappa numbers (a) and HexA contents (b) along the studied bleaching
sequences and the kappa number and HexA content of the unbleached pulp.

Fig. 1. Kappa number (a) and hexenuronic acid (HexA) content (b) obtained in each stage in the
different bleaching sequences. Values correspond to averages of three repetitions and error bars
represent standard deviations.

In Fig. 1a, a decrease in the kappa number can be observed after the application of all
bleaching sequences. These results are similar to those found by other authors (Eugenio et
al., 2010, 2011; Valls and Roncero, 2009). However, it is important to point out that,
although each stage in the biobleaching sequence led to a reduction in the kappa number,

315

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


the biggest reductions took place in all cases at the P stage. Also noteworthy is the fact that
sequences which used enzymatic treatments (LEP, (X + L)EP and XLEP) resulted in more
pronounced reductions in the kappa number after each stage, compared to their control
treatments or conventional bleaching (EP). Also in Fig. 1, it should be noted that the kappa
number increased from the value observed in the initial pulp (unbleached) when laccase or
xylanase were used individually as a rst stage in the bleaching. The kappa number increase
after the xylanase stage has been explained by Gonalves et al. (2008). These authors have
reported that Pulpzyme (the commercial xylanase used in this study) can change the pulp
composition, enriching its content of cellulose and lignin. In fact, bleached pulp treated with
xylanase (XLEP) showed the lowest hemicelluloses content (Table 1). Therefore, the kappa
number increase after X stage can be associated with the relative greater lignin content due to
the removal of hemicelluloses. The increase in the kappa number after laccase treatment can
be explained by the mediator (acetosyrignone) coupling chemically onto the ber (Cadena et
al., 2010, 2011) and, thus contributing to the kappa number determination. However, when
xylanase acted before laccase, no kappa number increase after the laccase treatment was
observed. A plausible explanation for this nding is that xylanase removes those parts of the
bers that mediators normally stick to, preventing bond formation and, therefore, decreasing
the kappa number. According to Cadena et al. (2011), HexA is the coupling partner of the
mediator, and therefore, the removal of HexA caused by xylanase treatment (Fig. 1b) would
reduce the possibility of bonding between mediator and bers. From this result it can be
concluded that a xylanase treatment favors a subsequent laccase treatment as reported by
others (Surma-Slusarska and Leks-Stepien, 2001; Valls and Roncero, 2009). Consistent
with these observations, when xylanase and the LMS were applied sequentially, a reduction
of both the kappa number and the HexA content at the end of the bleaching sequence was
found, relative to the experiment in which only the LMS was applied (kappa number of 5.4
vs. 6.1 and HexA content of 15.9 vs. 17.7 mol/g, respectively).
The simultaneous action of laccase and xylanase did not result in a greater kappa
number; however the nal kappa number, measured at the end of the whole bleaching
sequence, was similar to that obtained when only LMS was applied. This result is consistent
with that observed by Oksanen et al. (1997) who showed that applying these enzymes jointly
in a single step was not effective due to the inactivation of xylanase. However, when the
xylanase and laccase treatments were applied in sequence, these authors found that
bleaching was enhanced. This inactivation of xylanase when it was applied jointly with LMS

316

Publicaciones Cientficas/Scientific Publications


is supported by the higher hemicelluloses content observed in (X + L)EP pulp compared to
XLEP (13.4 % vs. 12.4 %, respectively). Moreover, a lower removal of HexA was observed
after (X + L) treatment, compared to that found after the sequential application of both
enzymes XL (Fig. 1b). These last results lead to the conclusion that, although some
enzymatic treatments caused the kappa number to increase from its initial value, at the end
of the corresponding bleaching sequences, there was a greater reduction compared to
controls or the EP treatment. This can be explained by changes induced by these enzymes
rendering lignin easier to remove in the subsequent chemical bleaching. It can also be
concluded that, a xylanase treatment helps the action of a subsequent laccase treatment,
leading to the biggest reductions in the nal kappa number (53.9 % compared to 48.4 % for
XLEP and LEP experiments, respectively) and the nal HexA content (25.9 % vs. 17.6 %
for XLEP and LEP, respectively).
The viscosity is an indirect indicator of the length of the polysaccharide chains present
in cellulose pulps (UNE 57-039-92), and, as Fig. 2 shows, after all the enzymatic treatments
were assayed, no big difference in viscosity was observed. The same result has been reported
by Oudia et al. (2008), who proved that high biodelignication can be achieved without
decreasing pulp viscosity. Conversely, other authors (Gonalves et al., 2008; Roncero et al.,
2005) have found differences in viscosity after a xylanase treatment, suggesting the
elimination of hemicelluloses of low molar mass as a plausible explanation. In addition, it
can be observed in Fig. 2 that, at the end of the bleaching sequences, viscosities are slightly
reduced with hydrogen peroxide bleaching the most inuential stage as previously reported
by Martn-Sampedro et al. (2011a). These results are consistent with the greatest reduction in
kappa number observed after the hydrogen peroxide stage, where it is likely that some
carbohydrates were degraded. Fig. 3a shows that brightness increased in all bleaching
sequences. The hydrogen peroxide stage led to the highest increase in all cases consistent
with the kappa number results.
Whenever an enzymatic treatment was included in the sequence, the resulting
brightness was higher than that found in the control or EP experiment (conventional
bleaching). However, some enzymatic steps (L and (X + L)) resulted in a reduction in
brightness relative to the baseline level (unbleached pulp). This has also been observed by
Eugenio et al. (2011) and can be attributed to the formation of new chromophoric groups
during these enzymatic stages. Based on our results, the coupling of the mediator onto HexA
is very probably responsible for this nding, as this reduction in brightness after the
317

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


enzymatic treatment was not observed when only xylanase acted on the initial pulp (XLEP
experiment).

Fig. 2. Viscosity of bleached pulp after each stage of the different bleaching sequences. Values
correspond to averages of three repetitions and error bars represent standard deviations.

Fig. 3. Optical properties of the bleached pulps: (a) brightness measured in each stage of the different
bleaching sequences, and (b) CIE L*a*b* and (c) CIE L*C* color coordinates for all experiments at
the end of the bleaching sequences. Values correspond to averages of three repetitions and error bars
represent standard deviations.

318

Publicaciones Cientficas/Scientific Publications


The XLEP treatment showed the highest brightness at the end of the sequence,
producing two and three points more than the LEP and the (X + L)EP experiments,
respectively. Therefore, it can be concluded that the xylanase treatment helped the action of
the laccase treatment when both enzymes were used separately. Similar results have been
described by Oksanen et al. (1997) and Surma-Slusarska and Leks-Stepien (2001). These
studies suggested that, by eliminating some of the xylan on the fiber surface, xylanase
pretreatment favors enzyme and/or chemical access, therefore, affording better pulp
bleaching.
CIE L*a*b* and CIE L*C* color coordinates (Fig. 3b-c) were measured after the P
stage in all bleached pulps. Results were compared in order to evaluate color changes
undergone during the biobleaching and conventional bleaching processes. Paper color is an
important parameter in many different applications; therefore, a numerical definition of
color is essential for good quality control and for a good customer-producer relationship.
Furthermore, good white base colors are necessary when products are to be dyed, printed,
or otherwise colored.
Fig. 3b shows lower a* and b* coordinates in all biobleached pulps, relative to the
conventionally bleached pulp (EP experiment). These results are consistent with those
reported by other authors (Eugenio et al., 2011; Fillat et al., 2012). However, as opposed to
the results obtained with Eucalyptus globulus pulps (Eugenio et al., 2011) and ax pulps
(Fillat et al., 2012), EFB pulps led to a negative a* value (closer to green than to red)
probably as a result of differences between raw materials. Therefore, a decrease in this value
for EFB pulps meant an increase in the absolute value, which could be a disadvantage if the
final value turns out to be significantly different from the ideal neutral white (value 0).
Remarkably, the XLEP treatment produced pulp with the lowest b* coordinates.
Fig. 3c shows the CIE L*C* color coordinates of pulps after the hydrogen peroxide
stage. All biobleached and control pulps showed increased L* and decreased C* (i.e.
increased lightness and decreased color) relative to the conventional bleaching (EP
experiment). Compared with their respective controls without enzyme addition, an increase
in lightness was noted in all cases.
However, only when xylanase and laccase were applied in sequence, biobleached
pulps showed a decrease in color relative to their controls. Fillat and Roncero (2010) did not
find a reduction in color (C* coordinate) using only laccase in a LE bleaching of ax pulp.

319

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


Our results show that the application of xylanase and laccase in sequence (XLEP) provided
pulp with the highest L* and the lowest C* values of all bleached pulps. Therefore, just like
with the brightness results, xylanase and laccase acting in sequence were better than when
combined in one single step. Moreover, the use of xylanase jointly with the laccase did not
show any advantage in color coordinates with respect to laccase treatment alone.

3.2. Hydrogen peroxide consumption


Hydrogen peroxide consumptions in the hydrogen peroxide bleaching stages are
shown in Fig. 4. Noticeably, experiments carried out with enzymatic treatments consumed
more hydrogen peroxide than their controls, although achieving more delignication and
brightness.
Moreover, while conventional bleaching (EP treatment) consumed 64.5 % of the initial
hydrogen peroxide load, the introduction of enzymatic treatments increased consumption to
72 % in the most favorable case ((X + L)EP treatment).
These results are inconsistent with those reported by Eugenio et al. (2010, 2011);
however, it is important to point out that these authors used a different pulp (Eucalyptus

globulus kraft pulp). In addition, kappa numbers measured in our biobleaching sequences
are lower than those reported by these authors. It is well known fact that, as the
delignication process proceeds, residual lignin becomes much harder to eliminate in a
conventional bleaching sequence.

Fig. 4. Hydrogen peroxide consumption for all bleaching sequences assayed.

320

Publicaciones Cientficas/Scientific Publications


Table 2. Influence of the accelerated ageing treatment in brightness of EFB bleached pulps.
Brightness (% ISO)
Originala

Agedb

Reduction

LcEPc

56.0

52.9

3.1

LEPc

59.0

55.4

3.6

(X+L)EPc

57.8

54.1

3.7

XcLcEPc

57.2

53.5

3.7

XLEPc

60.5

56.3

4.3

EPc

54.9

52.6

2.3

LcEP: Laccase control treatment + alkaline extraction + hydrogen peroxide bleaching

LEP: Laccase treatment + alkaline extraction + hydrogen peroxide bleaching

(X + L)EP: Laccase and xylanase treatment + alkaline extraction + hydrogen peroxide


bleaching

XcLcEP: Xylanase control treatment + laccase control treatment + alkaline extraction


+ hydrogen peroxide bleaching

XLEP: Xylanase treatment + laccase treatment + alkaline extraction + hydrogen


peroxide bleaching

EP: Alkaline extraction + hydrogen peroxide bleaching


a Before
b
c

the accelerated ageing treatment.

After the accelerated ageing treatment.

Bleaching sequences.

This could explain why there was a greater consumption of hydrogen peroxide in our
experiments.
After comparing the levels of hydrogen peroxide consumption of all the enzymatic
bleaching assays tested in our study, it becomes apparent that xylanase and laccase combined
in one single step consume less hydrogen peroxide than when working in sequence.
However, as mentioned above, the XLEP sequence provided pulps with a lower kappa
number and higher brightness. Therefore, if pulps with the same properties were to be
compared, hydrogen peroxide consumption in the XLEP sequence would probably be
lower than in the (X + L)EP sequence.

321

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


3.3. Handsheet characterization
Handsheets were formed from all bleached pulps obtained with the different bleaching
sequences and their mechanical properties were determined. Stretch, tensile index and tear
index are shown in Fig. 5.

Fig. 5. Mechanical properties of handsheets from pulps obtained at the end of all the bleaching
sequences assayed: (a) stretch (%), (b) tensile index and (c) tear index. Values correspond to averages
of three repetitions and error bars represent standard deviations.

The different pulps had similar properties; however, some variations were evident in
the tensile index. This index is related to the linking capacity between bers, so the lower
values obtained for pulps subjected to xylanase treatments (XLEP and (X + L)EP) could be
due to hemicelluloses (xylans) being lost during this process (Table 1). Martn-Sampedro et
al. (2011b) reported similar results with the application of a hemicellulose extraction method
prior to pulping. An even lower tensile index was observed with conventional bleaching (EP),
suggesting that an enzymatic treatment could increase the mechanical properties of the
bleached pulps (Ibarra et al., 2006; Moldes et al., 2008).
The tear index was similar for all bleached pulps, which suggests a similar degree of
ber degradation (similar viscosity, see Fig. 1b), consistent with what San Martin et al. (1995)
322

Publicaciones Cientficas/Scientific Publications


have reported after hemicellulose extraction. Herpol et al. (2002) also found similar results
and reported that the enzymatic treatments involved no signicant degradation of the
cellulose matrix, which suggests a specic effect of xylanase and laccase.
It should also be noted that most of the enzymatic treatments yielded better
mechanical properties than conventional bleaching (EP experiments) or those measured at
baseline in the initial pulp (unbleached). This has also been the observation of Kapoor et al.
(2007) who reported an increase in the tear index of up to 5.8 % compared to the
unbleached pulp after xylanase addition at various stages of pulp bleaching.
From all these results, it can be concluded that EFB provides bleached pulps with
acceptable mechanical properties, compared to other raw materials (Alaejos et al., 2004;
Snchez et al., 2011), and with a low kappa number and high brightness, after a simple
enzymatic-EP bleaching sequence. In addition, there is supporting evidence that the
enzymatic treatment improved the mechanical and optical properties of the pulps and
reduced the lignin content; the most efcient biobleaching sequence is that in which xylanase
and laccase were used sequentially.

3.4. Accelerated ageing


Accelerated ageing of bleached pulps was conducted in order to evaluate the effect of
the enzymatic treatment on the stability of optical properties. Table 2 and Fig. 6 show
changes in brightness and color coordinates, respectively, after accelerated ageing.
Table 2 proves that the enzymatic treatments resulted in greater brightness reduction
during ageing relative to controls and conventionally bleached pulp (EP). Other authors have
reported that optical properties of LMS biotreated pulps were more stable (i.e. smaller
brightness reduction) during accelerated ageing, compared to controls (Cadena et al., 2010;
Martn-Sampedro et al., 2011a). These authors attributed the smaller brightness reduction to
a greater removal of hexenuronic acid in the enzymatic treatment. However, these
researchers used eucalyptus kraft pulp with a higher hexeneuronic acid content than EFB
pulps (40 vs. 21.5 mol/g for unbleached pulps). This nding could be the reason why the
enzymatic treatment did not increase the optical stability of EFB soda pulps during
accelerated ageing. Nevertheless, EFB pulps treated with enzymes showed a higher
brightness value than controls even after accelerated ageing, with the highest brightness
observed in the bleached pulp treated with xylanase and laccase in sequence.
323

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378

Fig.6. Inuence of the accelerated ageing treatment in (a) CIE L*a*b* and (b) CIE L*C* color
coordinates of the bleached pulps. Black and gray symbols represent pulps before and after the
accelerating ageing, respectively. Values correspond to averages of three repetitions.

Fig. 6 shows the CIE L*a*b* and CIE L*C* color coordinates of the bleached pulps
before and after the accelerating ageing. Fig. 6a shows how the a* coordinate shifted to the
right (from green to red) and the b* coordinate increased (from blue to yellow) as a result of
accelerated ageing. Similar results have been found using a bacterial laccase as rst stage in a
LEP bleaching sequence of eucalyptus pulp (Eugenio et al., 2011). Fig. 6b shows the
evolution of lightness (L*) and color (C*) coordinates. An increase in C* and a reduction in
L* were observed after accelerated ageing, as reported by Eugenio et al. (2011). Control
pulps (without enzyme treatment) and conventionally bleached pulp showed higher C* and
lower L* than pulps treated with enzymes. When xylanase was applied jointly with laccase
((X + L)EP), the color properties were slightly better compared with LEP. However, the
sequential application of these enzymes resulted in the lowest C* value and the highest L*
value after ageing.

4. CONCLUSIONS
Oil palm EFB can be used as raw material for the production of soda-anthraquinone
biobleached pulps with competitive optical and mechanical properties. A sequential xylanase
and laccase treatment was the most efcient pretreatment for biobleaching, since it provides
pulps with lower kappa numbers, higher brightness and better colorimetric properties even
after accelerated ageing. Therefore, this two steps enzymatic pretreatment could be a good
option to reduce the environmental problems and increase the selectivity of the conventional
ECF and TCF sequences of EFB.
324

Publicaciones Cientficas/Scientific Publications


ACKNOWLEDGEMENTS
The authors wish to thank the Spanish Ministry of Science and Innovation for funding
this study via Project CTQ 2009-11268, and the FPI fellowship (BES2007-14528).

REFERENCES
Alaejos, J., Lpez, F., Eugenio, M.E., Fernndez, M., 2004. Non-wood raw materials
for pulp and paper making. A review Afinidad LXI, 513, 400410.
Bourbonnais, R., Paice, M.G., 1992. Demethylation and delignication of kraft pulp
by Trametes versicolor laccase in the presence of 2,20-azinobis-(3-ethylbenzthiazole-6sulfonate). Appl. Microbiol. Biot. 36 (6), 823-827.
Cadena, E.M., Vidal, T., Torres, A.L., 2010. Influence of the hexenuronic acid
content on rening and ageing in eucalyptus TCF pulp. Bioresource Technol. 101 (10),
3554-3560.
Cadena, E.M., Du, X., Gellerstedt, G., Li, J., Fillat, A., Garca-Ubasart, J., Vidal, T.,
Colom, J.F., 2011. On hexenuronic acid (HexA) removal and mediator coupling to pulp
ber in the laccase/mediator treatment. Bioresource Technol. 102 (4), 3911-3917.
Chauhan, S., Choudhury, B., Singh, S.N., Ghosh, P., 2006. Application of xylanase
enzyme of Bacillus coagulans as a prebleaching agent on non-woody pulps. Process
Biochem. 41, 226-231.
Daud, W.R.W., Law, K.N., Valade, J.L., 1998. Chemical pulping of oil palm empty
fruit bunches. Cell. Chem. Technol. 32 (1-2), 133-143.
Daud, W.R.W., Law, K.N., Ghazali, A., 2005. Alkaline peroxide mechanical pulping
of oil palm lignocellulosics Part 1. Effects of chemical and mechanical pretreatments on
empty fruit bunch (EFB) chemical composition. JIRCAS Working Report 39, 82-85.
Eugenio, M.E., Santos, S.M., Carbajo, J.M., Martn, J.A., Martn-Sampedro, R.,
Gonzlez, A.E., Villar, J.C., 2010. Kraft pulp biobleaching using an extracellular enzymatic
uid produced by Pycnoporus sanguineus. Bioresource Technol. 101 (6), 1866-1870.
Eugenio, M.E., Hernndez, M., Moya, R., Martn-Sampedro, R., Villar, J.C., Arias,
M.E., 2011. Evaluation of a new laccase produced by Streptomyces ipomoea on
biobleaching and ageing of kraft pulps. BioResources 6 (3), 3231-3241.
325

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


FAO.

Food

and

Agriculture

Organization

of

the

United

Nations.

(http://www.faostat.fao.org). Accessed on September 2011.


Ferrer, A., Rosal, A., Valls, C., Roncero, B., Rodrguez, A., 2011. Modeling hydrogen
peroxide bleaching of soda pulp from oil-palm empty fruit bunches. BioResources. 6 (2),
1298-1307.
Fillat, U., Roncero, M.B., 2010. Flax biobleaching by laccase mediator system at
atmospheric pressure. Afinidad 67 (548), 254-261.
Fillat, A., Gallardo, O., Vidal, T., Pastor, F.I.J., Daz, P., Roncero, M.B., 2012.
Enzymatic grafting of natural phenols to ax fibres: development of antimicrobial properties.
Carbohyd. Polym. 87, 146-152.
Gellerstedt, G., Li, J., 1996. An HPLC method for the quantitative determination of
hexeneuronic acid groups in chemical pulps. Carbohydr. Res. 294, 41-51.
Gonalves, A.R., Moriya, R.Y., Oliveira, L.R.M., Saad, M.B.W., 2008. Xylanase
recycling for the economical biobleaching of sugarcane bagasse and straw pulps. Enzyme
Microb. Technol. 43 (2), 157-163.
Herpol, I., Jeller, H., Fang, G., Petit-Conil, M., Bourbonnais, R., Robert, J.L.,
Asther, M., Sigoillot, J.C., 2002. Efcient enzymatic delignication of wheat straw pulp by
sequential xylanase-laccase mediator treatment. J. Pulp Pap. Sci. 28 (3), 67-71.
Ibarra, D., Camarero, S., Romero, J., Martnez, M.J., Martnez, A.T., 2006.
Integrating laccase-mediator treatment into an industrial type sequence for totally chlorine
free bleaching of eucalypt Kraft pulp. J. Chem. Technol. Biotechnol. 87 (7), 1159-1165.
Ibrahim, R., 2002. Chemical composition of alkaline pulps from oil palm empty fruit
bunches. Oil Palm Bulletin 44, 19-24.
Jimnez, L., Prez, A., Rodrguez, A., de la Torre, M.J., 2006. New raw materials and
pulping processes for production of pulp and paper. Afinidad 63 (525), 362-369.
Jimnez, L., Serrano, L., Rodrguez, A., Snchez, R., 2009a. Soda-anthraquinone
pulping of palm oil empty fruit bunches and beating of the resulting pulp. Bioresource
Technol. 100 (3), 1262-1267.
Jimnez, L., Serrano, L., Rodrguez, A., Ferrer, A., 2009b. TCF bleaching of sodaanthraquinone and diethanolamine pulp from oil palm empty fruit bunches. Bioresource
Technol. 100 (3), 1478-1481.
326

Publicaciones Cientficas/Scientific Publications


Kapoor, M., Kapoor, R.K., Kuhad, R.C., 2007. Differential and synergistic effects of
xylanase and laccase mediator system (LMS) in bleaching of soda and waste pulps. J. Appl.
Microbiol. 103 (2), 305-317.
Law, K.N., Jiang, X., 2001. Comparative papermaking properties of oil-palm empty
fruit bunch. Tappi J. 84 (1), 95.
Martn-Sampedro, R., Eugenio, M.E., Villar, J.C., 2011a. Biobleaching of Eucalyptus

globulus kraft pulps: comparison between pulps obtained from exploded and non-exploded
chips. Bioresource Technol. 102 (6), 4530-4535.
Martn-Sampedro, R., Eugenio, M.E., Revilla, E., Martn, J.A., Villar, J.C., 2011b.
Integration of kraft pulping on a forest biorefinery by the addition of a steam explosion
pretreatment. BioResources 6 (1), 513-528.
Moldes, D., Daz, M., Tzanov, T., Vidal, T., 2008. Comparative study of the efficiency
of synthetic and natural mediators in laccase-assisted bleaching of eucalyptus Kraft pulp.
Bioresource Technol. 99 (17), 7959-7965.
Oksanen, T., Amann, M., Candussio, A., Buchert, J., Viikari, L., 1997. Role of
xylanase in laccase-mediator delignification of kraft pulps. Proceedings of 9th International

Symposium on Wood and Pulping Chemistry (ISWPC) (Montreal) 83, 1-4.

Oudia, A., Queiroz, J., Simoes, R., 2008. Potential and limitation of Trametes

versicolor laccase on biodegradation of Eucalyptus globulus and Pinus pinaster kraft pulp.
Enzyme Microb. Tech. 43 (2), 144-148.
Quader, M.M.A., Lonnberg, B., 2005. Pulp and chemical cellulose obtained with oil
palm empty fruit bunches by a modied IDE process. JIRCAS Working Report 39, 47-53.
Rodrguez, A., Serrano, L., Moral, A., Prez, A., Jimnez, L., 2008. Use of highboiling point organic solvents for pulping oil palm empty fruit bunches. Bioresource
Technol. 99 (6), 1743-1749.
Roliadi, H., Pasaribu, R.A., 2004. The possible utilization of empty oil-palm bunches
as a substitute of pulp wood for the manufacture of corrugating-medium paperboard using
semi-chemical soda-anthraquinone pulping. APPITA annual Conference and Exhibition
(58th) 2, 335-341.

327

R. Martn-Sampedro et al./Bioresource Technology 110 (2012) 371-378


Roncero, M.B., Torres, A.L., Colom, J.F., Vidal, T., 2005. The effect of xylanase on
lignocellulosic components the bleaching of wood pulps. Bioresource Technol. 96 (1), 2130.
San Martn, R., Prez, D., Briones, R., 1995. Simultaneous production of ethanol and
kraft pulp from pine (Pinus radiata) using steam explosion. Bioresource Technol. 53 (3),
217-223.
Snchez, R., Ferrer, A., Serrano, L., Toledano, A., Labidi, J., Rodrguez, A., 2011.

Hesperaloe funifera as a raw material for integral utilization of its components. BioResources
6 (1), 0003-0021.
Surma-Slusarska, B., Leks-Stepien, J., 2001. TCF bleaching of kraft pulps with
laccase and xylanase. J. Wood Chem. Technol. 21 (4), 361-370.
Thurston, C.F., 1994. The structure and function of fungal laccases. Microbiol-SGM
140, 19-26.
Valls, C., Roncero, B., 2009. Using both xylanase and laccase enzymes for pulp
bleaching. Bioresource Technol. 100 (6), 2032-2039.

328

Publicaciones Cientficas/Scientific Publications

PUBLICACIN 8
VALORIZATION OF RESIDUAL EMPTY PALM FRUIT BUNCH FIBERS (EPFBF)
BY MICROFLUIDIZATION: PRODUCTION OF NANOFIBRILLATED
CELLULOSE AND EPFBF NANOPAPER
Ana Ferrer, Ilari Filpponen, Alejandro Rodrguez, Janne Laine and Orlando J. Rojas
Bioresource Technology, 2012, 125, 249255
ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de Agricultural
Engineering segn el Journal Citation Reports de 2011)

329

Ferrer et al./Bioresource Technology 125 (2012) 249-255

330

Publicaciones Cientficas/Scientific Publications


Bioresource Technology 125 (2012) 249255

VALORIZATION OF RESIDUAL EMPTY PALM FRUIT BUNCH FIBERS (EPFBF) BY


MICROFLUIDIZATION: PRODUCTION OF NANOFIBRILLATED CELLULOSE AND
EPFBF NANOPAPER
Ana Ferrera, Ilari Filpponenb, Alejandro Rodrgueza, Janne Laineb and Orlando J.
Rojasb,c
bSchool

aDepartment

of Chemical Engineering, University of Crdoba, Spain

of Science and Technology, Department of Forest Products Technology, Aalto University,

cDepartment

Finland, 00076 Aalto

of Forest Biomaterials, North Carolina State University, Raleigh, NC 27695-8005

Received 4 July 2012; Received in revised form 25 August 2012; Accepted 27 August 2012; Available online 5
September 2012

ABSTRACT
Different cellulose pulps were produced from sulphur-free chemical treatments of
Empty Palm Fruit Bunch Fiber (EPFBF), a by-product from palm oil processing. The pulps
were microfluidized for deconstruction into nanofibrillated cellulose (NFC) and nanopaper
was manufactured by using an overpressure device. The morphological and structural
features of the obtained NFCs were characterized via atomic force and scanning electron
microscopies. The physical properties as well as the interactions with water of sheets from
three different pulps were compared with those of nanopaper obtained from the
corresponding NFC. Distinctive chemical and morphological characteristics and ensuing
nanopaper properties were generated by the EPFBF fibers. The NFC grades obtained
compared favorably with associated materials typically produced from bleached wood fibers.
Lower water absorption, higher tensile strengths (107 to 137 MPa) and elastic modulus (1218 GPa) were measured, which opens the possibility for valorization of such widely available
bioresource.

331

Ferrer et al./Bioresource Technology 125 (2012) 249-255


KEYWORDS: Empty Palm Fruit Bunch Fibers (EPFBF); Nanofibrillated cellulose (NFC);
Film strength; Water interactions; Nanopaper.

1. INTRODUCTION
The marked increase in the use of alternative, non-wood fibers results from the need
to cut costs and avoid negative environmental effects. In fact, non-wood raw materials
constitute the sole useful source of cellulose fiber in some regions of the world (Rodrguez et
al., 2008) and provide several interesting advantages (Rodrguez et al., 2008): (a) allows wood
raw materials to be spared for uses where they are most required; (b) reduces wood and
cellulose fiber imports in countries with a shortage of wood and, (c) satisfies the increasing
demands for paper-grade fibers from green processes.
The use of Empty Palm Fruit Bunch Fiber (EPFBF), bagasse and rice straw for
papermaking and production of composite panels is a common practice in several countries
without large supplies of wood resources (English et al., 1997; Young, 1997). EPFBF, in
particular, comprises residual fibers from the palm (Elaeis guineensis) oil processing.
Malaysia is the largest palm oil producer (51 % of the worldwide production), and constitutes
an important economic resource for this country. Cultures are also extending to countries in
Western Africa (Nigeria, Guinea, Ghana, etc.), South America (Ecuador, Colombia,
Honduras, etc.) and Asia (Thailand). The global production of oil palm has risen from 16
million tons in 2007 to almost 21 million tons in 2010 (FAO, 2010). Palm plants start
fruiting 4-5 years after planting. Fruit bunches usually weigh 15-25 Kg and contain 1000-4000
oval-shaped, 3-5 cm long fruits. Fruit production peaks at 20-30 years, after which they
decline and become unprofitable (especially because their fruits are too high to collect).
Each hectare of oil palm produces an average of 10 tons fruits per year, which give about
3000 Kg of palm oil as the main product (Malaysian Palm Oil Council, 2010). A significant
amount of residual EPFBF can be used as a source of cellulosic fibers if separated by
pulping processes. In this regard and owing to environmental pressures alternative, sulphurfree pulping methods can be considered (Rodrguez et al., 2008). Therefore, in this work
soda-anthraquinone (NaOH-AQ) digestion was applied to yield high quality fibers from
EPFBF (Ibrahim, 2002; Law and Jiang, 2001). Also, two organosolv processes (Formosolv
and Milox) were used on the basis of expected advantages that include (Rodrguez et al.,
2008): (a) reduced small- and mid-scale production costs relative to kraft processes, (b) facile

332

Publicaciones Cientficas/Scientific Publications


and efficient recovery of solvents and by-products; (c) relative low water, energy and
chemicals usage.
Within the idea of harnessing the concept of biorefinery, fractionation of the
byproducts of the palm oil industry may open a promising pathway for comprehensive use of
residual biomass. It should be noted that cellulose is the most important component of
EPFBF, which justify efforts to generate fermentable sugars (Kim et al., 2012). In the context
of the present investigation, fibers from empty palm fruit bunches are proposed as a
resource for obtaining nanofibrillated cellulose (NFC). Because of their wide abundance,
their biodegradability and its renewable nature as well as their unique structural and physical
aspects, NFC has recently garnered much attention for use as reinforcement in composites
and in coatings, films, membranes, packaging materials, etc. It is also a natural nanomaterial
that provides a large range of possibilities to obtain superior properties in different endproducts. For example, it is known that NFC has unique properties and functionalities
compared to macro-scale fibers. Moreover, environmental concerns drive the interest on
sustainable and eco-friendly products, including NFC as an alternative to materials derived
from petroleum or non-renewable sources.
Cellulose nano- and micro- fibers were isolated from banana rachis (Zuluaga et al.,
2007) by using chemical and mechanical treatments; such agricultural residue, which is rich
in cellulose, has attracted interest due to its potential use as a reinforcing component in
composites (Faria et al., 2006; Gan et al., 2004). It is worth noting that that NFC
production has been primarily from bleached pulps. However, interest in the production of
NFC from unbleached fibers has been reported (Ferrer et al., 2012; Spence et al., 2010a,b).
Some physical properties have been observed to improve with the presence of residual cell
wall components (lignin and hemicelluloses). Of particular interest is the possibility of
packaging films with high oxygen barrier capability as well as other gases and vapors.
In this work EPFBF is proposed as a novel and suitable resource to produce NFC.
Thus, different grades of cellulose nanofibrils were isolated from unbleached EPFBF and
their properties as well as those of respective nanopaper were compared with those from
conventional sources. The morphological and structural features of isolated nanofibrils and
reference pulp handsheets were analyzed by atomic force microscopy (AFM) as well as
scanning electron microscopy (SEM). The tensile index, elastic modulus and % elongation of
handsheets and nanopaper with similar grammage were determined. Complementary
examination of specific surface area, water absorbency and rates of capillary absorption are
333

Ferrer et al./Bioresource Technology 125 (2012) 249-255


reported. The results are discussed in light of the properties of NFC from conventional
sources, namely bleached wood pulps.

2. MATERIAL AND METHODS


2.1. EPFBF, EPFBF pulps and handsheets
EPFBF from a Malaysian oil palm plant was supplied by Straw Pulping Engineering
S.L. (Zaragoza, Spain). The chemical composition of EPFBF was determined following
TAPPI standards T-222 for lignin, T-203 for -cellulose, T-204 for ethanol extractives and
T-211 for ash. Holocellulose content was determined by the Wise et al. method (1946). The
analysis showed a composition that included 67 % holocellulose, 41.9 % -cellulose, 24.5 %
lignin, 1.2 % extractives and 3.2 % ash.
Average fiber length and thickness of microcooked EPFBF (10 % soda, 80 C, 1 h)
were found to be 0.53 mm and 14 m, respectively as determined from stainning (1 %
saffranin) and observation under a Visopan projection microscope (10x objective for 100x
magnification) (Snchez et al., 2010).
Three different EPFBF pulp grades were obtained after the following sulphur-free
chemical treatments: (a) soda-anthraquinone (NaOH-AQ): 15 % NaOH; 1 % AQ; 30 min;
170 C; H=10; (b) formic and hydrochloric acids (FoOH, Formosolv): 92.5 % FoOH; 0.075
% HCl; 60 min; 100 C; H=10 and, (c) formic acid and hydrogen peroxide (Milox): 53 %
FoOH; 3 % H2O2; 165 min; 80 C; H=10, where the H is the hydromodule (liquid-solid

ratio). The short notation N, F and M is used to refer to the respective process.

The operation conditions used were selected based on results from previous work
(Jimnez et al., 2009; Ferrer et al., 2011a,b). The pulps were obtained by using a 15-L
rotating batch cylindrical reactor connected to a heater. The cooked, unbleached fibers were
refined in a Sprout-Bauer device and the pulps obtained were screened (0.16 mm mesh
size).
Pulp chemical composition was determined using the standard methods mentioned
above in the case of EPFBF fibers. Hemicellulose content was estimated as the difference
between holocellulose and -cellulose contents. Paper handsheets were produced with an
ENJO-F-39.71 sheet machine according to the UNE-57-042 standard.

334

Publicaciones Cientficas/Scientific Publications


2.2. Isolation of cellulose nanofibrils (NFC)
2.2.1. Mechanical pre-treatment
The NaOH-AQ (N), FoOH-Formosolv (F) and Milox (M) pulps in their never-dried
form were first dispersed in de-ionized water. Afterwards, they were treated (ion exchanged)
under stirring for 20 min with a hydrochloric acid solution 1 M and washed with de-ionized
water two times using a Bchner filter funnel. Following, the fibers were suspended in
NaHCO3 solution 1 M in order to convert the carboxyl groups to their sodium form.

Finally, the fibers were washed with de-ionized water and drained on a Bchner filter funnel
until a filtrate conductivity of less than 20 S/cm was measured.
A refining treatment before microfluidization was necessary in order to enhanced fiber
accessibility and processing efficiency. Therefore, the EPFBF pulps were refined for 20000
revolutions in a PFI-mill according to ISO 5264-2:2002. The drainability of the refined pulps
was determined using the Schopper-Riegler method (SR) according to ISO 5267-1:1999.
The refining process resulted in fibers with enhanced internal and external fibrillation. In
addition, fiber flow was improved and clogging avoided during fluidization. Noting that SR
was not used here as a parameter for property prediction but to control the fluidization
process, it was determined that 90 SR or above was suitable for an effective fluidization
process.

2.2.2. Mechanical high-shear disintegration via microfluidization


Nanofibrillated cellulose from EPFBF pulps was obtained by microfluidization.
Typically, 1.8 % fiber suspension in water was processed in a high-pressure microfluidizer
(Microfluidizer M-110 P, Microfluidics Corp., 2010). Samples were passed five times
through an intensifier pump that increased the pressure, followed by an interaction chamber
which defibrillated the fibers by shear forces and impacts against the channel walls and
colliding streams. Through this process, fibers were broken up into nano-sized structures
forming slurries of nanofibrillated cellulose. The microfluidizer operated at a constant shear
rate and the operating pressure was maintained at 55 MPa. The temperature was not
controlled but fluidization was temporarily ceased when the temperature of the stock
reached approximately 90 C, to prevent pump cavitation. Processing then recommenced
when the samples had cooled to approximately 45 C. The resulting NFC suspensions were
collected and stored at 4 C in a cold room until needed.
335

Ferrer et al./Bioresource Technology 125 (2012) 249-255


2.2.3. Manufacture of nanopaper (NFC films)
The respective NFC suspensions were diluted to 1 % of consistency in deionized
water. The mixture was then stirred at room temperature overnight. A portion of the slurry
was slowly poured into an over-pressurized device (2.5 bar pressure difference) to produce
nanopaper with basis weight of about 60 g/m2 after drying. A screen was used as a filter to

remove excess water but not the fibrils. The produced webs were placed in a press (4.1 bar,
4 min) to make them flat and smooth. Following, the wet webs were dried between two
blotting papers at high temperature and pressure (hot-pressing) (100 C, 0.18 bar, 2 h). This

last step, besides drying the films, also promoted the formation of inter-fiber bonds because
of the simultaneous applications of heat and pressure treatment. The dried films or
nanopaper samples were stored in a conditioned room at 23 C and 50 % ambient relative
humidity until further analysis.

2.3. Characterization of EPFBF handsheets, NFC suspensions and nanopaper


2.3.1. Kappa number, viscosity and degree of polymerization
The kappa number (KN) before and after microfluidization was calculated by
measuring the consumption of potassium permanganate according to ISO 302 and SCAN-C
1:00 standards, respectively. As a reference, the residual lignin content of the samples can be
roughly estimated by multiplying KN by 0.15 if the contribution of hexenuronic acids is
neglected.
The limiting viscosity number of all NFC films was determined after ISO 5351
standard. The NFC films were dried for 24 h at 50 C. A portion of the NFC was then
dissolved in a cupriethylenediamine (CED) solution. The flow time through a reference
length of a capillary viscometer was determined for the neat CED. The average degree of
polymerization (DP) of the nanofibrillated cellulose samples was calculated from () using
the Staudinger-Mark-Houwink equation []=KDPa (Gruber and Gruber, 1981) where the

constants K and a are dependent on the polymer-solvent system used and the DP of the

cellulose. The measurements were performed on cellulosic fibers dissolved in 0.5 M CED
according to SCAN-CM 15:99 after a slight modification (Henriksson et al., 2007) since the
re-dispersion of dried films was very limited and thus the samples were weighed in the wet
state. The effect of water present in the solution was accounted for in standard solutions with

336

Publicaciones Cientficas/Scientific Publications


0.5 M CED. K and a were assumed to be 2.28 mL/g and 0.76, respectively (for DP > 950)
(Henriksson et al., 2007).

2.3.2. NFC and morphology of EPFBF nanopaper


AFM imaging was performed to assess the morphology, roughness, thickness and
material distribution of dispersed NFC and respective nanopaper. Small pieces (5x5 mm2) of

NFC films or NFC dispersions dried on silica surfaces were mounted on an aluminum
sample holders and examined with a Nanoscope IIIa multimode scanning probe microscope
from Digital Instruments Inc., Santa Barbara, CA. Scanning was performed in tapping mode
in air using silicon cantilevers (NSCl5/AIBS) delivered by MicroMasch, Tallinn, Estonia.
The drive frequency of the cantilever was about 300-360 KHz. The scanned areas (1x1, 3x3,
5x5 and 10x10 m2) were imaged in at least five different areas on the respective sample. No

image processing except flattening was made. Images were taken in height mode, where the
deflection of the cantilever is directly used to measure the z position, and in phase mode,
where the phase shift of the cantilever is used to determine differences in material
composition. The imaging was performed at 30-40 % relative humidity at 25 C. In addition,
scanning electron microscopy was used to observe the microstructure and the transversal
morphology of NFC films. Cross sections of fractured films (liquid nitrogen) were imaged
with a JEOL JSM-6400F (JEOL, Peabody, MA, USA) SEM operating at 5 KV and a
working distance of 20 mm.

2.3.3. Physical properties of the EPFBF handsheets and nanopaper


Congo Red, CR (Merck, C.I.: 22120; chemical formula: C32H22N6Na2O6S2, pKa of

5.5 and molecular weight of 696.7 g/mol) was used to determine the specific surface area of
NFC in aqueous suspensions. CR is a direct benzidine-based dye with two napthionic acid
molecules (Cotoruelo et al., 2010): 1-naphthalenesulfonic acid, 3, 3-(4, 4-biphenylenebis
(azo)) bis (4- amino-) disodium salt. In calculating the specific surface area the CR molecule
was assumed as a planar structure in anionic form with 2.62 x 0.74 x 0.43 nm dimensions
(Lorenc-Grabowska and Gryglewicz, 2007).
CR adsorption experiments were conducted as described by Inglesby and Zeronian
(2002). Fiber and NFC suspensions from EPFBF pulps were dialyzed (10-12 K MWCO
tubing) against a 0.1 M phosphate buffer (pH=6) for several days with multiple changes of
337

Ferrer et al./Bioresource Technology 125 (2012) 249-255


the buffer solution. A 0.0266 % (w/v) NFC suspension for each pulp in the buffer was dyed
with various concentrations of CR (5-15 % of the weight of substrate) at 60 C for 24 h. A
small amount of NaCl (0.004 wt%) was added to each sample at the start of the experiment
to neutralize charged surface sites on the cellulose nanofibrils. After 24 h, the samples were
centrifuged for 10 min, the supernatant was drawn off, and its absorbance was measured at
488 nm with a UV-2550 Shimadzu spectrophotometer. All runs were carried out at the least
by duplicate, with the reported values being the averages. The maximal adsorption of CR on
the fiber and the NFC suspensions was calculated using an equation derived from monolayer
adsorption model (Goodrich and Winter, 2007) assuming a Langmuir-type isotherm
(Yoshida et al., 1989). Thus, the maximum amount of CR adsorbed onto cellulose Amax

(mg/g) was calculated by using Eq. (1) (Ougiya et al., 1998).


[E]

=K
[A]

ad Amax

[E]

+A

max

(1)

where [E] (mg/mL) is the free dye concentration at equilibrium, [A] (mg dye/mg substrate) is

the amount of dye adsorbed by the substrate, and Kad is the adsorption equilibrium constant.

The values for Amax were extracted from the reciprocal of the slope from the linear

equations in a plot of free dye concentration versus free dye concentration/dye on fiber. The
adsorbed molecular area per dye molecule has been estimated by molecular modeling by
Woodcock et al. (1995) who indicated that direct dyes dock to cellulose with the longitudinal
axis parallel to the polymer chain. Thus, by using Amax and the adsorbed surface area of CR
molecule the specific surface area (SSA) of cellulose fibrils was quantified (Eq. (2), (Ougiya
et al., 1998)):
SSA=

Amax NSA
MW1021

(2)

where N is Avogadros constant, SA and MW are the CR molecule surface area (1.73 nm2)
and molecular weight (696.7 g/mol), respectively.

The respective EPFBF fibers and NFC suspensions were used to produce handsheets
and nanopaper (NFC films). Handsheet and nanopaper thickness were determined using
TAPPI Method T411 by means of a Lorentzen & Wettre Micrometer 51 instrument. The
basis weight was determined using TAPPI standard T410, and the apparent handsheet or
nanopaper density was calculated using the apparent thickness and measured basis weight.
Air permeability was measured using a Lorentzen & Wettre SE 114-Bendtsen instrument,
according to ISO 5636-3 and porosity of the sample was calculated form its density (Eq. (3))

338

Publicaciones Cientficas/Scientific Publications


by assuming the density for the cellulose to be 1500 Kg/m3 (Sehaqui et al., 2011). The ISO

brightness (%) of handsheets and respective nanopaper was measured using a Lorentzen &

Wettre SE 070R Elrepho spectrophotometer.

Porosity (%)=[1- sample ]100

(3)

cellulose

2.3.4. Strength properties


Paper handsheet mechanical properties (Fmax, Strength, Stretch, Work and Stiffness)

were measured using a tensile instrument (Lorentze & Wettre Alwetron TH1) according to
ISO 5270. The tests were performed using a clamp length of 100 mm, 12 mm/min
elongation rate, 125 N maximum force and 1.87 KN/m maximum strength. In order to
measure the mechanical properties of nanopaper, universal tensile tests were carried out at
23 C and 50 % relative humidity (RH), in accordance with EN ISO 291:2008, using MTS
400/M Vertical Tensile Tester equipped with a 200 N load cell. Specimens strips of 5 mm
wide and 75 mm length were stamped out from nanopaper samples according to ISO 527,
using a custom made cutting die. The tests were performed with a cross-head speed of 5
mm/min. Tensile index, the tensile strength divided by the basis weight, are reported to
account for variations in film basis weight that could skew the tensile strength results. The
average and standard deviation of five measurements are reported.

2.3.5. Interactions with water


In order to measure the Relative Water Absorption Capacity (RWAC), the respective
EPFBF handsheets and nanopaper samples were conditioned for 60 min in a stove (105 C)
and after this time they were placed inside a desiccator for 30 min. The samples were
immersed in deionized water for 35 h and weighed after 1, 5, 10, 20, 60, 120, 1020 and 2100
min using a top pan balance. Care was taken to remove excess water using absorbent paper
before weighing the samples. Relative water absorption capacity (% RWAC) was determined
using the Eq. (4):
% RWAC=100

Wt0 -Wt
Wt0

(4)

where Wt0 and Wt are the weights of the samples before and after immersion, respectively.
339

Ferrer et al./Bioresource Technology 125 (2012) 249-255


In addition, the initial and dynamic water contact angle (WCA) of the handsheets and
nanopaper samples were determined using an optical contact angle meter (KSV instruments
LTD, CAM 200) on both the top (air) and bottom (screen) sides of the sample, according to
T558 om-97.
Water wicking rates were measured using handsheet and nanopaper strips (3 mm wide
and 30-35 mm length) with a Sigma 70 surface tensiometer (KSV instrument LTCl). The
samples were first wetted to a set depth to ensure that there was complete wetting (zero
contact angle) and the change in pulling force was recorded by using a constant height.
Specifically, the given sample was positioned at a constant zero depth of immersion
relative to the contact line on the water surface. The measured pulling force was used to
calculate the extent of water update.

3. RESULTS AND DISCUSSION


3.1. EPFBF pulp chemical composition, fiber and nanofiber morphology
For ease of discussion, the different EPFBF pulps employed were labeled as indicated
in Table 1; the specimens indicated in the tables are listed in order of increasing pulping
yield (or lignin content). The average and standard deviation of five measurements are
reported. The main chemical characteristics are also provided in Table 1. -Cellulose
contents are found to be relatively high for the NaOH-AQ pulp while the lignin and
extractive contents are low (2.3 % and 2.5 %, respectively). The FoOH pulp has a lignin
content of ca. 9 %. Similar values are measured for ethanol extractives. The hemicellulose
content for Milox pulp is approximately 20 %.
Table 1. EPFBF pulps chemical composition in weight percent. Short notation N, M and F are used
for NaOH-AQ, Milox and FoOH, respectively.
Pulp type

-cellulose

Hemicellulose

Holocellulose

Lignin

Ethanol
extractives

Ash

NaOH-AQ: N

81.8

15.9

97.7

2.3

2.5

1.0

Milox: M

63.4

22.1

85.5

6.2

5.1

1.6

FoOH: F

75.8

6.2

82

8.6

1.7

As previously shown in the case of wood-derived pulps (Spence et al., 2010a),


pretreatment reduces fiber length and fibrils are more easily observed on their surfaces. The
morphology of the fibers/fibrils before and after microfluidization of the EPFBF pulps was
340

Publicaciones Cientficas/Scientific Publications


accessed by optical microscopy (images not shown). After microfluidization, no fibers were
observed but only small cell wall elements. All pulps samples showed similar changes in
morphology upon processing; microfluidization produced a more homogeneous fibril size
distribution.
Fibril webs were prepared from aqueous NFC suspensions, which experienced strong
interfibril interactions upon drying (hydrogen bonding, etc.) and thus produced stiff and
strong self-standing films (nanopaper). Additionally, fiber network samples were obtained
from free drying of small volumes of low concentration NFC suspensions deposited on flat
silica supports. The surface morphology of NaOH-AQ, Milox and FoOH fibers in
handsheets were observed with an AFM (images not shown). Nanofibrils were clearly
identified and the orientation was observed to be random-in-plane. The refined fiber surface
in the handsheets was smooth; at this length scale it is difficult to discern one fiber form
another. A very tight, dense structure was observed in films produced from the NFC
obtained after microfluidization. In addition, a net reduction of fiber diameter was observed.
Whole fibers were no longer observed. Instead, a network with high fibril density, complex
orientation and distribution was formed; they are expected to have a large surface area.
Some few nanofibril agglomerates or bundles were present. Overall, refining and
microfluidization of EPFBF produced changes similar to those observed in the case of wood
pulps and other fiber sources (Zimmermann et al., 2005), i.e., they are subject to distinctive
fiber cell wall deconstruction and size reduction, down to the nanoscale.
It is difficult to draw relationships between the properties of the EPFBF fibers and
respective NFC due to the widely different chemical composition and process condition
used. However, it is interesting to observe that no major differences exist in terms of
coarseness of the fibrils in NFC for all pulp precursor types (or lignin content as indicated by
the kappa numbers, Table 2). However, all fibers/fibril systems obtained from the FoOH
pulps presented higher roughnesses compared to the other EPFBF counterparts (Table 3).
AFM RMS roughness of NFC dispersions obtained after microfluidization (by using same
conditions) of unbleached and fully-bleached birch pulps deposited on silica were found to
be ca. 8 and 17 nm, respectively (scan size of 5x5 m2). The fibrils generated from EPFBF

display similar or slightly smaller values (11-16 nm RMS). Therefore, there is indication that
nanofibrils obtained from EPFBF using same processing conditions are similar or finer than
those obtained from conventional wood fibers.

341

Ferrer et al./Bioresource Technology 125 (2012) 249-255


The microfluidization defibrillated the fibers by shear forces yielding a gel-like
suspension. A layered structure of the respective NFC films (cross section SEM images, not
shown) was observed and they were more or less compact depending on the type of
precursor pulp. The nanofibrils are apparently deposited flatly in swirled conformation
during filtration. In general, the lateral dimensions of the NFC nanofibrils were in the range
of 10-30 nm. The NaOH-AQ NFC films were denser, while the Milox NFC films were
more disorganized. The lignin content in the film did not appear to have any apparent effect
on the density or morphology.

3.2 Characteristics of EPFBF pulps, handsheets and nanopaper


The handsheets and nanopaper samples were characterized in terms of kappa number
(KN), viscosity, density, brightness, basis weight, thickness, air permeability, porosity and
degree of polymerization (DP) (see Table 2).
It is noted that compared to the values for the original pulps, the KN for NFC
suspensions were not significantly altered after microfluidization. However, this process
produced a non-negligible effect on the viscosity. The average molecular weight of cellulose,
as estimated from the average intrinsic viscosity, indicates the extent of chain degradation
due to chemical or mechanical treatment (Lapierre et al., 2006).
Table 2 shows the intrinsic viscosity as well as the calculated degree of polymerization
(DP) of the cellulose in fibers/fibrils in handsheets and nanopaper samples. In general, the
viscosity values were lower for cellulose in nanopaper than in handsheets as a result of the
high shear homogenization used to produce NFC. The DP strongly correlates with the
aspect ratio of the fibrils. For example, larger fibrils with constant diameter have a higher DP
and better network forming capability. Fibrillation leads to a decrease in viscosities or DP,
typically between 15 and 63 % (Zimmermann et al., 2010). The drop is larger for cellulose in
pulps with a high viscosity (such as NaOH-AQ pulp, whose drop is around 285 units); pulps
that are already more degraded due to harsh chemical/mechanical pre-treatment (as Milox
and FoOH pulps) show lower reductions (105 and 247 units, respectively). As expected, the
highest viscosities were measured for the fibers in handsheets. The viscosity of the Milox
system was very low (364 and 259 mL/g for the fibers in handsheets and NFC films,
respectively); this is due to the respective pulping treatment, which in this case involved
hydrogen peroxide that induced extensive degradation of cellulose chains.

342

Publicaciones Cientficas/Scientific Publications


Table 2. Mean values ( standard deviation) of brightness, density, basis weight, thickness, air permeability, porosity, viscosity, kappa number (KN) and degree of
polymerization (DP) for fibers or nanofibers as well as the respective handsheets and nanopaper produced from different EPFBF sources (NaOH-AQ (N), Milox
(M) and FoOH (F)). The average and standard deviation of five measurements are reported.

N
M
F

Handsheet
Nanopaper
Handsheet
Nanopaper
Handsheet
Nanopaper

Kappa
number (KN)
12.8 0.2
13.8 0.3
39.8 0.2
41.3 0.1
52.8 0.5
54.2 0.8

Basis weight
(g/m2)
59.4 0.1
58.8 0.2
63.3 0.3
60.1 0.3
56.7 0.2
51.4 0.2

Thickness
(m)
61.9 0.1
60.6 0.1
68.5 0.1
64.9 0.1
62.4 0.1
54.5 0.1

Density
(g/cm3)
0.96 0.01
0.97 0.11
0.924 0.001
0.926 0.102
0.908 0.099
0.943 0.012

Porosity
(%)
36 0.2
35.3 0.1
38.4 0.1
38.3 0.3
39.5 0.1
37.1 0.2

Air permeability
(mL/min)
1.15 0.47
1.03 0.29
1.32 1.64
1.07 0.37
2.15 0.65
0.68 0.32

Viscosity
(mL/g)
958.8 47
675.5 3.3
364.1 4.5
258.7 9.0
675.8 11.6
428.3 11.5

DP
2835 183
1767 11
750 12
489 23
1747 22
971 34

Brightness
(ISO B %)
22.3 0.4
17.4 0.9
21.2 0.2
15 1
9.9 0.1
9.5 0.2

Table 3. Mean values ( standard deviation) of Relative Water Absorption Capacity (RWAC), Water Contact Angle (WCA) and rates of capillary water
absorption (rate of water uptake) for handsheets and nanopaper samples from respective pulp sources. AFM average surface roughness ( standard deviation) of
EPFBF fiber handsheets, nanopaper and NFC dispersions deposited on silica (scan size 5x5 m2). Mean values ( standard deviation) of Specific Surface Area
(SSA) from Congo Red dye adsorption of pulps and NFC suspensions from EPFBF pulps. The average and standard deviation of five measurements were
reported.

N
M
F

Handsheet/Pulp suspension
Nanopaper/NFC suspension
Fibril dispersion on silica
Handsheet/Pulp suspension
Nanopaper/NFC suspension
Fibril dispersion on silica
Handsheet/Pulp suspension
Nanopaper/NFC suspension
Fibril dispersion on silica

AFMs RMS roughness (nm)


41.7 0.1
23.7 0.2
10.7 0.1
45.1 0.2
24.3 0.1
11.1 0.01
85.0 0.3
28.4 0.3
15.8 0.1

RWAC (%)
158.4 7.0
129.4 5.2
98.6 6.0
81.2 5.0
78.3 2.5
65.6 3.5
-

WCA ()
54 5
50 5
67 5
55 5
79 5
52 3
-

Rate of water uptake (s-1)


10 2
10 1
10 1
10 2
20 0
10 1
-

SSA (m2/g)
30 4
42 10
55 3
77 2
29 2
100 1
-

343

Ferrer et al./Bioresource Technology 125 (2012) 249-255


The viscosities of the fibrillated materials are between 676 and 259 mL/g
(corresponding DPs of 1770 and 490); these values are much higher than those reported for
other sources, e.g., 100 mL/g for fibrillated beech wood pulp (Zimmermann et al., 2010).
The DP values are also higher than that reported for fibrillated wheat straw pulp, 674
(Zimmermann et al., 2010).
According to the results included in Table 2, the thickness of the nanopaper obtained
after microfluidization is reduced, as expected from the changes in film density. In general,
samples that contained higher residual lignin (and as a result other cell wall components)
underwent a higher reduction in film thickness upon refining and microfluidization (e.g., 2, 6
and 13 % reduction in thickness for the precursor N, M and F fibers, respectively). Paper
handsheet densities (0.91-0.96 g/cm3) were slightly lower than those for the nanopaper

samples (0.93-0.97 g/cm3). The sharp density increase after microfluidization of the FoOH

pulp (from 0.91 to 0.94 g/cm3) can be explained by the handsheet formation since compared

to the other pulp sources, a larger amount of fiber is lost through the wire. The nanopaper
samples obtained in this study have higher densities than those reported elsewhere (0.86 and
0.78 g/cm3 obtained from bleached softwood and unbleached softwood, respectively), which
is mainly explained by differences in fiber morphology, type of filtration and pressing

operation (Spence et al., 2010b). Densities of 0.99 and 0.71 g/cm3 were measured for

nanopaper samples obtained following the same conditions used in the present work but
from unbleached and fully-bleached birch pulps, respectively.
As observed from the data in Table 2, the porosity of EPFBF nanopaper is lower than
that of fiber handsheets, as expected. This is the result of the smoother and tighter texture of
nanofibrillated cellulose, which can be appreciated in Table 3. Nanopaper samples of
considerable porosity can be obtained but the pores and gaps within the nanofibrils are
much smaller in size and number. There is an indication that the presence of residual lignin
and other cell wall components induces finer fibrils upon mechanical shearing and therefore
more effective compaction during pressing and drying (Ferrer et al., 2012). The values of %
porosity obtained in this study (range between 35 % and 39 %) are in line with those reported for homogenized bagasse and rice straw, 31 % and 28 %, respectively (Hassan et al.,
2012).
The brightness was evaluated as a relevant parameter in packaging and other grades.
The handsheets and nanopaper samples containing more lignin (Milox and FoOH), possess
a significantly lower brightness compared to samples from NaOH-AQ.
344

Publicaciones Cientficas/Scientific Publications


Upon microfluidization the respective nanopaper samples were found to be darker
(lower brightness) compared to the handsheets. This suggests that the lignin in the
nanofibrils may be redistributed or dispersed on their surface. This is expected to affect the
opacity and scattering coefficients: samples that scatter less light (less opaque) are likely to
appear less bright (Spence et al., 2010b).

3.3 Specific Surface Area (SSA)


The specific surface area (SSA) for fiber and NFC suspensions was calculated by using
the Langmuir equilibrium constant Kad (Eq. (2)) after Congo red adsorption (see Table 3).

The average and standard deviation of five measurements are reported. A significant SSA
increase is observed when evaluating the nanofibrils relative to the respective precursor pulps
fibers. In addition, the results indicate Congo red SSA values of EPFBF-based NFC (42-100
m2/g) lower than the figures reported for fibrils obtained from homogenization of bleached
softwood, unbleached hardwood and unbleached softwood (195, 202 and 161 m2/g,
respectively) but higher than those from thermomechanical and bleached hardwood pulps

(29 and 68 m2/g, respectively) (Spence et al., 2010b). Overall, the highest SSA values are

found for NFC and pulps prepared from FoOH (Table 3). According to our findings and

those from other authors (Spence et al., 2010a, 2011) the SSA increases with pulp yield (also
lignin content), due to their finer fibril dimensions.

3.4 Interactions with water


The relative water absorption capacity (RWAC) more clearly reflects the differences in
the hydrophilicity of the fibers. RWAC decreases upon microfluidization of the EPFBF
fibers: a % reduction in RWCA of 18, 17 and 16 % is observed for NaOH-AQ, Milox and
FoOH systems (Table 3). This observation can be explained by the significantly more
compact, less porous nanopaper samples compared to the handsheet counterparts, which
prevents water penetration. In addition, it is observed that nanopaper samples with higher
lignin content produce lower RWAC values (compare the RWAC of NFC films from
NaOH-AQ, Milox and FoOH).
The water contact angle is also reported in Table 3. Since the residual lignin is less
hydrophilic than cellulose, it is expected that the initial water contact angle increases with
lignin content. This is indeed the case as can be observed for EPFBF handsheets. In the case
345

Ferrer et al./Bioresource Technology 125 (2012) 249-255


of nanopaper samples the WCA values are similar (50-55 ) for all systems within the experimental deviation. In this case the surface roughness associated with the solid-liquid-gas
interfaces of the macroporous surface that plays an important role (Tamai and Aratani,
1972).
The rates of capillary water absorption or wicking of EPFBF handsheets and
nanopaper strips (30-35 mm x 3 mm x 40-45 m) were measured for the three different pulp
precursors after contact with deionized water for 5 h. Table 3 provides an overview of the
rates of water uptake by each system. The average and standard deviation of five
measurements are reported. According to the results, differences in the rates of water uptake
are observed only when the amount of lignin in the system is high, as in the case of FoOH
samples. The rate of capillary water absorption for EPFBF nanopaper it similar for the
systems studied (10 s-1). Furthermore, the measured values for water uptake by EPFBF

nanopaper are slightly lower than those obtained by using similar procedures from birch
pulps (Ferrer et al., accepted for publication). Overall, the results suggest that changes in
interactions with water are related to differences in film pore structure and surface energy.

3.5 Strength properties of EPFBF handsheets and nanopaper


The similar KN values measure for EPFBF fibers and the respective nanofibrils (NFC)
indicate that the fluidization process did not alter lignin content. However, the characteristic
dimensions of fibers/fibrils used for preparation of handsheets is quite different from that
used to manufacture nanopaper, thus the properties are expected to be quite different. Thus,
this work reports on the strength properties of resulting handsheets and nanopaper samples
for comparison purposes (Tables 4 and 5, respectively).
Table 4. Mean values ( standard deviation) of handsheet mechanical properties: Maximun force
(Fmax), Strength, Stretch, Work and Stiffness.
NaOH-AQ

Milox

FoOH

Fmax (N)

76.2 2.9

55.7 2.9

35.3 3.2

Strength (KNm/Kg)

86.7 3.3

61.8 3.2

47.6 4.3

Stretch (%)

4.2 0.5

3.2 0.42

1.33 0.3

Work (J/Kg)

2797 438

1488.7 222

417 131

Stiffness (MNm/Kg)

8.6 0.2

6.9 0.4

6.32 0.4

346

Publicaciones Cientficas/Scientific Publications


The tensile strength indicates the maximum load that a material can support without
breaking and the tensile index is the relation between the tensile strength and the basis
weight of the film. This parameter allows comparing characteristics of samples with slightly
different basis weights. Tensile energy absorption (TEA) is defined as the area under the
load-elongation curve and it is related to the surface area of the sample between the grips.
This result is typically used to characterize the energy absorbing capacity of the sample
before breaking. The results were normalized using the basis weight of the handsheets and
nanopaper samples, which allowed better comparison of the energy absorption. The
modulus of elasticity (or Youngs Modulus) is a measure of the stiffness of the material, but it
only applies in the linear region of the load-elongation curve.
Table 5. Mean values ( standard deviation) of nanopaper mechanical properties: Tensile Strength,
Tensile Index, Breaking Strain, Elastic Modulus, Yield Strength, Yield Strain and Tensile Energy
Absorbed.
NaOH-AQ

Milox

FoOH

Tensile Strength, MPa

137 7

124 12

106 9

Tensile Index, KN/g

2330 121

2063 199

2062 158

Breaking Strain, %

0.96 0.16

0.99 0.12

1.23 0.11

Elastic Modulus, GPa

17.9 1.2

14.2 0.8

11.5 0.5

Yield Strength, MPa

67 18

66 16

53 10

Yield Strain, %

0.43 0.09

0.47 0.08

0.47 0.09

Tensile Energy Absorbed, KJ/m2

36 9

44 10

41 10

Tensile Energy Absorbed, J/g

612 144

732 152

798 167

Table 4 includes the data of force to rupture (or maximum force, Fmax) (N), strength

(KNm/Kg), stretch (%), work (J/Kg) and stiffness (MNm/Kg) of respective paper

handsheets. The measured mechanical performance of these materials is quite interesting.


As can be observed, the handsheet results are much better in the case of the NaOH-AQ
sample.
However, it is difficult to provide any reason for this observation because the
difference in chemical composition of the EPFBF pulps. Nevertheless, it is noted that pulps
with higher lignin content resulted in lower mechanical properties, as expected due to the
interference of lignin with hydrogen bonding which results in a weaker material.
After microfluidization, however, this trend does not apply universally (Table 5). It is
noted that nanopaper samples with higher lignin content does not undergo significant loss in

347

Ferrer et al./Bioresource Technology 125 (2012) 249-255


mechanical properties; on the contrary, a slight increase was observed in some of the
parameters that were measured (for example, the break elongation, breaking strain, yield
strains and the energy to break). Nanopaper samples, expectedly, have different properties
to handsheets and, in most cases the normalized mechanical properties are improved. The
tensile strength obtained for the nanopaper samples from EPFBF pulps (107-137 MPa) is
with no exception higher than the values corresponding to those obtained from birch pulps
by using similar manufacture procedures, ca. 100 MPa (Ferrer et al., 2012). Similar
observation applies to the elastic modulus (12-18 GPa for nanopaper from EPFBF pulps
compared to 9-11 GPa for nanopaper from birch pulps).

4. CONCLUSIONS
We described the production of cellulose nanofibrils extracted from Empty Palm Fruit
Bunch Fibers (EPFBF) obtained after chemical and mechanical treatments. The Soda,
Milox and Formosolv pulping followed by mechanical microfluidization allowed the
production of suspensions of individualized, long nanofibers. In all cases the deconstruction
of EPFBF into nanofibrillated cellulose (NFC) resulted in a decrease in water absorption by
their films, which is an important property for applications such as packaging.
The NFC from EPFBF pulps have properties that compare favorably against typical
NFC obtained from bleached wood fibers and thus represent a value-added material from
this widely available residual bioresource.
Finally, it is demonstrated that EPFBF is a suitable raw material for the production of
nanopaper in high performance applications with structural and strength properties similar
to those reported from wood fibers.

ACKNOWLEDGEMENTS
This work has been performed as a part of the Lignocell Project (Value-added
materials and functional structures from lignocellulosic). We would like to acknowledge
VTT Technical Research Centre of Finland for supplying technical support. Ana Ferrer is
grateful to Spains DGICyT, MICINN and CEIC for funding this research within the
framework of the Projects CTQ2010-19844-C02-01 and P10-TEP-6261.

348

Publicaciones Cientficas/Scientific Publications


REFERENCES
Cotoruelo, L., Marqus, M., Daz, F., Rodrguez-Mirasol, J., Rodrguez, J., Cordero,
T., 2010. Equilibrium and kinetic study of Congo red adsorption onto lignin-based activated
carbons. Transport Porous Med. 83, 573-590.
English, B., Chow, P., Bajwa, D., 1997. Processing into composites. In: Rowell MR,
Young RA, Rowell JK (Eds). Paper and composites from agro-based resources. CRC Lewis
Publishers, New York, pp. 269-297.
FAO

(Food

and

Agriculture

Organization

of

the

United

Nations),

http://www.faostat.fao.org/. Accessed June, 2010.


Faria, H., Cordoeiro, N., Belgacem, M., Dufresne, A., 2006. Dwarf Cavendish as a
source of natural fibers in poly(propylene)-based composites. Macromol. Mater. Eng. 291,
16-26.
Ferrer, A., Vega, A., Rodrguez, A., Ligero, P., Jimnez, L., 2011a. Milox fractionation
of empty fruit bunches from Elaeis guineensis. Bioresour. Technol. 102, 9755-9762.
Ferrer, A., Vega, A., Ligero, P., Rodrguez, A., 2011b. Pulping of empty fruit bunches
(EFB) from the palm oil industry by formic acid. Bioresources 6, 4282-4301.
Ferrer, A., Quintana, E., Filpponen, I., Solala, I., Vidal, T., Rodrguez, A., Laine, J.,
Rojas, O.J., 2012. Effect of Residual Lignin and Heteropolysaccharides in Nanofibrillar
Cellulose and Nanopaper from Wood Fibers. Cellulose, accepted.
Gan, P., Zuluaga, R., Vlez, J.M., Mondragn, I., 2004. Biological natural retting for
determining the hierarchical structuration of banana fibers. Macromol. Biosci. 4, 978-983.
Goodrich, J., Winter, W., 2007. -Chitin nanocrystal prepared from shrimp shells and
their specific surface area measurement. Biomacromolecules 8, 252-257.
Gruber,

E.,

Gruber,

R.,

1981.

Viskosimetrische

Bestimmung

des

Polymerisationsgrades von cellulose. Das Papier 35, 133-141.


Hassan, M., Mathew, A., Hassan, E., El-Wakil, N., Oksman, K., 2012. Nanofibers
from bagasse and rice straw: process optimization and properties. Wood Sci. Technol. 46,
193-205.

349

Ferrer et al./Bioresource Technology 125 (2012) 249-255


Henriksson, M., Henriksson, G., Berglund, L., Lindstrm, T., 2007. An
environmentally friendly method for enzyme-assisted preparation of microfibrillated
cellulose (MFC) nanofibers. Eur. Polym. J. 43, 3434-3441.
Ibrahim, R., 2002. Chemical composition of alkaline pulps from oil palm empty fruit
bunches. Oil Palm Bulletin 44, 19-24.
Inglesby, M., Zeronian, S., 2002. Direct dyes as molecular sensors to characterize
cellulose substrates. Cellulose 9, 19-29.
Jimnez, L., Serrano, L., Rodrguez, A., Snchez, R., 2009. Soda-anthraquinone
pulping of palm oil empty fruit bunches and beating of the resulting pulps. Bioresour.
Technol. 100, 1262-1267.
Kim, S., Park, J.M., Seo, J-W., Kim, C.H., 2012. Sequential acid-/alkali-pretreatment
of empty palm fruit bunch fiber. Bioresour. Technol. 109, 229-233.
Lapierre, L., Bouchard, J., Berry, R., 2006. On the relationship between fibre length,
cellulose chain length and pulp viscosity of a softwood sulfite pulp. Holzforschung 60, 372377.
Law, K., Jiang, X., 2001. Comparative papermaking properties of oil-palm empty fruit
bunch. Tappi J. 84 (1) 95.
Lorenc-Grabowska, E., Gryglewicz, G., 2007. Adsorption characteristics of congo red
on coal-based mesoporous activated carbon. Dyes Pigm. 74, 34-40.
Malaysian Palm Oil Council (MPOC), http://www.mpoc.org.my. Accessed June, 2010.
Ougiya, H., Hioki, N., Watanabe, K., Morinaga, Y., Yoshinaga, F., Samejima, M.,
1998. Relationship between the physical properties and surface area of cellulose derived
from adsorbates of various molecular sizes. Biosci. Biotechnol. Biochem. 62, 1880-1884.
Rodrguez, A., Serrano, L., Moral, A., Prez, A., Jimnez, L., 2008. Use of highboiling point organic solvents for pulping oil palm empty fruit bunches. Bioresour. Technol.
99, 1743-1749.
Snchez, R., Rodrguez, A., Requejo, A., Ferrer, A., Navarro, E., 2010. Soda pulp and
fuel gases synthesis from Hesperaloe funifera. Bioresour. Technol. 101, 7032-7040.
Sehaqui, H., Allais, M., Zhou, Q., Berglund, L., 2011. Wood cellulose biocomposites
with fibrous structures at micro-and nanoscale. Compos. Sci. Technol. 71, 382-387.

350

Publicaciones Cientficas/Scientific Publications


Spence, K., Venditti, R., Habibi, Y., Rojas, O., Pawlak, J., 2010a. The effect of
chemical composition on microfibrillar cellulose films from wood pulps: mechanical
processing and physical properties. Bioresour. Technol. 101, 5961-5968.
Spence, K.L., Venditti, R.A., Rojas, O.J., Habibi, Y., Pawlak, J.P., 2010b. The effect
of chemical composition on microfibrillar cellulose films from wood pulps: water
interactions and physical properties for packaging applications. Cellulose 17, 835-848.
Spence, K.L., Venditti, R.A., Rojas, O.J., Habibi, Y., Pawlak, J.P., 2011. A
comparative study of energy consumption and physical properties of microfibrillated
cellulose produced by different processing methods. Cellulose 18, 1097-1111.
Tamai, Y., Aratani, K., 1972. Experimental study of the relation between contact angle
and surface roughness. J. Phys. Chem. 76, 3267-3271.
Wise, L., Murphy, M., DAdieco, A., 1946. Chlorite holocellulose its fractionation
and beating on summative wood analysis and on studies on the hemicellulose. Paper Trade
Journal 122 (2), 35.
Woodcock, S., Henrissat, B., Sugiyama, J., 1995. Docking of Congo red to the surface
of crystalline cellulose using molecular mechanics. Biopolymers 36, 201-210.
Yoshida, H., Kataoka, T., Maekawa, M., Nango, M., 1989. Surface diffusion of direct
dyes in porous cellulose membranes. Chem. Eng. J. 41(1), B1-B9.
Young, R., 1997. Processing of agro-based resources into pulp and paper. In: Rowell
MR, Young RA, Rowell JK (Eds.). Paper and composites from agro-based resources. CRC
Lewis Publishers, New York, pp. 137-245.
Zimmermann, T., Phler, E., Schwaller, P., 2005. Mechanical and morphological
properties of cellulose fibril reinforced nanocomposites. Adv. Eng. Mater. 7, 1156-1161.
Zimmermann, T., Bordeanu, N., Strub, E., 2010. Properties of nanofibrillated
cellulose from different raw materials and its reinforcement potential. Carbohydr. Polym. 79,
1086-1093.
Zuluaga, R., Putaux, J., Restrepo, A., Mondragn, I., Gan, P., 2007. Cellulose
microfibrils from banana farming residues: isolation and characterization. Cellulose 15, 585592.

351

Ferrer et al./Bioresource Technology 125 (2012) 249-255

352

Conclusiones/Conclusions

Conclusiones/Conclusions
CONCLUSIONES

A lo largo de esta Memoria se han presentado diferentes metodologas que permiten


el aprovechamiento integral de un residuo agroalimentario muy abundante: el residuo de la
industria del aceite de palma. De los resultados experimentales obtenidos y de su
tratamiento y discusin, se pueden extraer las siguientes conclusiones originales:

1. CARACTERIZACIN DE LA MATERIA PRIMA


De la comparacin de la caracterizacin qumica del EFB (4,03 % de solubles en agua
caliente, 20,22 % de solubles en sosa al 1 %, 1,20 % de extrables con etanol-benceno o con
etanol, 3,20 % de cenizas, 67 % de holocelulosa, 41,9 % de -celulosa, 25,1 % de
hemicelulosa y 24,5 % de lignina) con la de residuos agrcolas (podas de olivo, paja de trigo,
tallos de girasol, tallos de sorgo, paja de arroz, bagazo de caa de azcar, sarmientos de vid y
tallos de algodonera), vegetales alternativos a los convencionales (leucaena, tagasaste, retama,
phragmites, arundo, prosopis y paulownia) y maderas de conferas (pino) y frondosas
(eucalipto), puede afirmarse que el EFB constituye una fuente viable y alternativa de celulosa
para la obtencin de pastas celulsicas y papel.
La longitud de las fibras del EFB es pequea (0,53 mm) comparada con la de los
residuos agrcolas, vegetales alternativos y maderas de pino y eucalipto, todas ellas en torno o
por encima de 1 mm.

2. TRATAMIENTOS HIDROTRMICOS
Del estudio de la influencia de las variables del tratamiento hidrotrmico del EFB, sin
catalizador cido, sobre la composicin de la fraccin lquida y el rendimiento de la fraccin
slida resultantes, mediante un diseo factorial de experimentos y la utilizacin de un
modelo polinmico, se encuentran ecuaciones que relacionan el rendimiento de la fraccin
slida y los contenidos de glucosa, arabinosa y cido actico de la fraccin lquida, en
funcin de la temperatura, tiempo y relacin lquido/slido, con errores menores del 15 %,
siendo menores del 10 % en la mayor parte de los casos. La ecuacin encontrada para el
contenido de xilosa no reproduce adecuadamente los resultados experimentales. Para todas

355

Conclusiones/Conclusions
las variables dependientes consideradas, la variable de operacin ms influyente es la
temperatura y la menos influyente la relacin lquido/slido.
Al estudiar la influencia de las variables del tratamiento hidrotrmico del EFB,
catalizado con cido sulfrico, sobre las variables dependientes relacionadas con las
fracciones lquida y slida resultantes, considerando un diseo factorial de experimentos y la
utilizacin de un modelo polinmico, se encuentran ecuaciones que relacionan el
rendimiento de la fraccin slida y los contenidos de glucosa, arabinosa y cido actico de la
fraccin lquida, en funcin de la temperatura, tiempo y concentracin de cido sulfrico,
con errores menores del 15 %. La ecuacin encontrada para el contenido de xilosa no
reproduce adecuadamente los resultados experimentales. Para todas las variables
dependientes consideradas, la variable de operacin ms influyente es la temperatura y la
menos influyente la concentracin de cido catalizador.
Al comparar los valores de las variables dependientes relacionadas con las fracciones
slida y lquida de los tratamientos hidrotrmicos del EFB, sin y con catalizador, se
encuentra que al emplear catalizador el rendimiento mximo decrece un 3,26 %, mientras
que los contenidos de glucosa, xilosa y arabinosa se incrementan un 52,63 %, 22 % y 26,43
%, respectivamente.
Mediante simulacin del tratamiento hidrotrmico con catalizador cido, se encuentra
que operando a 190 C, con una relacin lquido/slido de 6 y una concentracin de cido
sulfrico del 0,1 %, durante 15 minutos, se obtienen valores relativamente altos, respecto a
los mximos, de glucosa (3,12 g/L), xilosa (4,00 g/L), arabinosa (2,35 g/L) y cido actico
(2,28 g/L), ahorrando capital al no operar con mximos valores de las variables de
operacin, y alcanzando un rendimiento de slidos del 68 %.

3.

PASTEADOS

DE

LA

FRACCIN

SLIDA

DEL

TRATAMIENTO

HIDROTRMICO DEL EFB, Y DEL EFB ORIGINAL


Al comparar los valores de las propiedades de las pastas y de las hojas de papel
correspondientes de diversas pastas del EFB y de la fraccin slida de su tratamiento
hidrotrmico, utilizando sosa-antraquinona, etanol y etanolamina, respectivamente, se
comprueba que los valores del nmero kappa y de la blancura de las pastas mejoran cuando
el pasteado ha sido precedido de un tratamiento hidrotrmico, para los tres procesos de
pasteado considerados, pero los valores de la viscosidad de las pastas y de las propiedades
356

Conclusiones/Conclusions
de resistencia de las hojas de papel (ndices de traccin, estallido y desgarro) disminuyen.
Por ello, cuando el EFB se somete a un tratamiento hidrotrmico, con la idea de aprovechar
la fraccin hemicelulsica, la pasta obtenida de la fraccin slida de dicho tratamiento es
ms conveniente emplearla en la fabricacin de papeles especiales o en la produccin de
bioetanol, mediante hidrlisis enzimtica y fermentacin simultneas.

4. PASTEADOS ORGANOSOLV
4.1. Proceso de pasteado Formosolv
Utilizando un diseo factorial de experimentos para el estudio de la influencia de las
variables del proceso de pasteado del EFB con cido frmico a presin atmosfrica, sobre el
rendimiento, nmero kappa, viscosidad y blancura de las pastas obtenidas, se encuentran
modelos polinmicos que reproducen las variables dependientes consideradas en funcin
del tiempo y las concentraciones de los cidos frmico y clorhdrico (catalizador), con
errores menores del 10 %.
Utilizando estos modelos polinmicos pueden encontrarse unas condiciones de
operacin de compromiso (92,5 % de cido frmico, 0,075 % de cido clorhdrico y 60
minutos), de manera que se obtengan unos aceptables valores de las propiedades de las
pastas (42,3 % de rendimiento, 22,7 % de blancura y 512 mL/g de viscosidad) y se ahorren
productos qumicos y energa, respecto a operar de manera ms enrgica.

4.2. Proceso de pasteado Milox


En el pasteado del EFB mediante el proceso Milox se ha estudiado la influencia de las
condiciones de operacin en la primera etapa del proceso sobre las caractersticas de las
pastas obtenidas, utilizando un diseo factorial de experimentos y modelos polinmicos y
neuroborrosos, encontrando ecuaciones que reproducen los valores experimentales de
rendimiento, nmero kappa, viscosidad y blancura de las pastas, en funcin de las
concentraciones de cido frmico y perxido de hidrgeno y el tiempo de proceso, con
errores menores del 10 %.
Mediante simulacin del proceso Milox, utilizando los modelos polinmicos y
neuroborrosos encontrados, se encuentra que operando con un 93 % de cido frmico y 3
% de perxido de hidrgeno durante 165 minutos, se obtiene una pasta con buenas
357

Conclusiones/Conclusions
propiedades (40,5 % de rendimiento, 50,3 % de blancura y 608 mL/g de viscosidad),
ahorrando al mismo tiempo reactivos y energa, con respecto a operar bajo condiciones ms
severas.
Una segunda etapa del proceso Milox debe ser muy corta (30 minutos o menos) para
evitar un efecto negativo en la viscosidad.

5. CARACTERIZACIN DEL LICOR RESIDUAL DE PASTEADOS ORGANOSOLV


El licor residual del pasteado del EFB (95 % de cido frmico, 0,05 % de cido
clorhdrico y 30 minutos) tiene un contenido de lignina del 18,2 %, de glucosa del 4,1 %, de
xilosa del 9,8 % y de arabinosa del 1,2 %, sobre base seca con respecto a la materia prima
original.
El licor residual del proceso Milox no contiene lignina precipitable, y contiene bajas
cantidades de monosacridos fermentables (0,71 % de glucosa, 4,22 % de xilosa, 1,19 % de
galactosa, 0,33 % de manosa y 0,82 % de furfural), sobre base seca de materia prima original.
Ello es debido probablemente al alto poder oxidativo de la mezcla de reaccin.

6. BLANQUEO DE PASTAS CELULSICAS


6.1. Blanqueo con la secuencia AOpAZRP
Del estudio del blanqueo de pastas a la sosa-antraquinona y a la dietanolamina del
EFB con la secuencia AOpAZRP (tratamiento cido - oxgeno con perxido de hidrgeno tratamiento cido ozono - tratamiento reductor - perxido de hidrgeno) se obtienen
pastas blanqueadas con valores de la viscosidad y de las propiedades fsicas de las hojas de
papel (ndice de traccin, alargamiento, ndice de estallido, ndice de desgarro y blancura)
similares o mayores que las de pastas blanqueadas de otros materiales no madereros y de
eucalipto con secuencias TCF (totalmente libres de cloro).
La pasta del EFB a la dietanolamina despus del blanqueo con la secuencia
AOpAZRP exhibe una viscosidad ms alta (783 mL/g) y propiedades de las hojas de papel
cercanas o incluso mejores que las presentadas por las hojas de papel de la pasta a la sosaantraquinona blanqueada con la misma secuencia (22,2 frente a 20,4 Nm/g para el ndice
de traccin, 1,30 frente a 1,42 KN/g para el ndice de estallido, 0,71 frente a 0,70 mNm2/g
para el ndice de desgarro y 71,3 frente a 77,5 % en blancura).
358

Conclusiones/Conclusions
6.2. Blanqueo con la secuencia EPabOPoP
Una pasta del EFB obtenida con cido frmico (95 % de cido frmico, 0,05 % de
cido clorhdrico y 30 minutos), con un nmero kappa de 31,1 y una viscosidad de 606
mL/g, se blanque usando la secuencia EPabOPoP (extraccin alcalina - peractico en
medio bsico oxgeno - perxido de hidrgeno con oxgeno - perxido de hidrgeno)
consiguiendo una blancura del 69,4 %, y unas prdidas de viscosidad y rendimiento del 34,8
% y 17,7 %, respectivamente.

6.3. Blanqueo con perxido de hidrgeno


Del estudio de la influencia de las condiciones de blanqueo con perxido de
hidrgeno de la pasta a la sosa-antraquinona del EFB, sobre las propiedades de las pastas
blanqueadas, siguiendo un diseo factorial de experimentos y modelos polinmicos y
neuroborrosos, se comprueba que se reproducen los valores de viscosidad, blancura y
nmero kappa en funcin de la concentraciones de perxido de hidrgeno y sosa y tiempo
de reaccin, con errores menores del 10 %.
Mediante simulacin del proceso de blanqueo de pastas del EFB, con los modelos
polinmicos y neuronales encontrados, es posible determinar los valores ptimos de las
variables de operacin, de forma que las propiedades de las pastas blanqueadas difieran
ligeramente de sus mejores valores, y al mismo tiempo se consigan ahorros de reactivos
qumicos, energa y tamao de la planta, operando con valores bajos de las variables
operacionales. As, operando con una concentracin de sosa de 1,13 %, una concentracin
de perxido de hidrgeno de 2,25 % y con un tiempo de 3 horas, es posible obtener una
pasta con una blancura de 75,1 % (8,1 % por debajo del mximo) y una viscosidad de 740
mL/g (10,4 % menos que el valor mximo).
De un estudio comparativo del blanqueo de pastas a la sosa-antraquinona del EFB y
de Hesperaloe funifera con perxido de hidrgeno, se concluye que, utilizando un diseo
factorial de experimentos, la aplicacin de modelos neuroborrosos conduce a ecuaciones
que predicen los valores de viscosidad y blancura de las pastas blanqueadas con perxido de
hidrgeno, en funcin de las variables de operacin (concentracin de perxido de
hidrgeno, concentracin de sosa y tiempo de reaccin), con errores menores del 10 % para
el caso del blanqueo de la pasta del EFB, y del 15 % para la pasta de Hesperaloe funifera.

359

Conclusiones/Conclusions
Mediante simulacin del blanqueo con perxido de hidrgeno, utilizando modelos
neuroborrosos, se comprueba que operando con una concentracin de sosa del 0,5 %, una
concentracin de perxido de hidrgeno del 3 %, y un tiempo de 3 horas, se puede obtener
una pasta del EFB con una blancura del 74,9 % (7,0 % por debajo del mximo) y una
viscosidad de 716 mL/g (12,7 % menos que el valor mximo); y una pasta de Hesperaloe

funifera con una blancura del 63,3 % (7,7 % por debajo del mximo) y una viscosidad de
584 mL/g (14,1 % menos que el valor mximo).

6.4. Blanqueo enzimtico


De un estudio del bioblanqueo de pasta a la sosa-antraquinona del EFB utilizando
etapas con las enzimas xilanasa y lacasa, se encuentra que se obtienen pastas con
propiedades mecnicas y pticas muy competitivas. De las secuencias XEP, LEP, (X+L)EP y
XLEP (X, xilanasa; L, lacasa; E, extraccin alcalina; P, perxido de hidrgeno), la ltima fue
la ms idnea, por lo que el pretratamiento ms eficiente para el bioblanqueo es cuando las
enzimas se utilizan consecutivamente, pues proporcionan pastas con valores ms bajos de
nmero kappa (5,4) y ms elevados para la blancura (60,5 % ISO). Por otra parte, este
pretratamiento tambin da lugar a pastas con mejores propiedades colorimtricas, incluso
despus del envejecimiento acelerado. Por tanto, estas dos etapas de pretratamiento
enzimtico podran ser una buena opcin para reducir los problemas medioambientales y
aumentar la selectividad de las secuencias ECF (libres de cloro elemental) y TCF (totalmente
libres de cloro) aplicadas a pastas del EFB.

7. PRODUCCIN DE CELULOSA NANOFIBRILAR


Se ha estudiado la produccin de celulosa nanofibrilar a partir de fibras procedentes
del EFB, mediante la utilizacin de tratamientos qumicos y mecnicos.
Los pasteados a la sosa-antraquinona, Milox y Formosolv seguidos de una
microfluidizacin mecnica permitieron la produccin de suspensiones de nanofibras largas
e individualizadas. En todos los casos, la deconstruccin de las fibras procedentes del EFB y
su paso a celulosa nanofibrilar (NFC) se tradujo en un descenso de la absorcin de agua por
parte de las pelculas de celulosa nanofibrilar formadas, lo que constituye una propiedad
importante para aplicaciones como el empacado.

360

Conclusiones/Conclusions
La celulosa nanofibrilar obtenida de pastas de fibras del EFB muestra propiedades
mejores comparadas con las que presenta la celulosa nanofibrilar obtenida usando fibras
madereras blanqueadas, y por tanto se presenta como un material con un alto valor aadido
debido a que es un residuo agroindustrial ampliamente disponible.
Finalmente, se demuestra que las fibras del EFB son una idnea materia prima para la
produccin de nanopapel con un gran nmero de aplicaciones de elevada trascendencia y
con propiedades estructurales y de fuerza similares a las encontradas para fibras madereras.

361

Conclusiones/Conclusions

362

Conclusiones/Conclusions
CONCLUSIONS

Throughout this Report, different methodologies, which allow the integral use of an
abundant agri-food residue: the residue from the palm oil industry, have been presented.
The following original conclusions can be drawn from the experimental results and their
treatment and discussion:

1. RAW MATERIAL CHARACTERIZATION


By making a comparison of the chemical characterization of EFB (4.03 % of hot water
solubles, 20.22 % of 1 % sodium hydroxide solubles, 1.20 % of ethanol-benzene or ethanol
extractives, 3.20 % of ash, 67 % of holocellulose, 41.9 % of -cellulose, 25.1 % of
hemicellulose and 24.5 % of lignin) with agricultural residues (olive prunings, wheat straw,
sunflower stalks, sorghum stalks, rice straw, bagasse sugar cane, vine shoots and cotton
stalks), vegetables alternative to conventional ones (leucaena, tagasaste, broom, phragmites,
arundo, prosopis and paulownia) and softwoods (pine) and hardwood (eucalyptus), it can be
concluded that the EFB is a viable alternative source of cellulose for the production of
cellulosic pulps and paper.
The fiber length of EFB is short (0.53 mm) compared to agricultural residues,
alternative vegetables as well as pine and eucalyptus woods, all of which are around or above
1 mm.

2. HYDROTHERMAL TREATMENTS
From the study of the influence of the variables of the hydrothermal treatment of EFB,
without an acid catalyst, on the composition of the liquid fraction and the yield of the solid
fraction obtained, using a factorial design of experiments and the use of a polynomial model,
equations, which relate the yield of the solid fraction and the contents of glucose, arabinose
and acetic acid from the liquid fraction, are found depending on the temperature, time and
liquid/solid ratio, with errors less than 15 %, being less than 10 % in most cases. The
equation found for the xylose content does not reproduce experimental results adequately.
For all considered dependent variables, the most influential operation variable is the
temperature and the least influential is the liquid/solid ratio.
363

Conclusiones/Conclusions
By studying the influence of variables of the hydrothermal treatment of EFB, catalyzed
with sulphuric acid, on the dependent variables related to the resulting liquid and solid
fractions, considering a factorial design of experiments and the use of a polynomial model,
equations, which relate the yield of the solid fraction and the contents of glucose, arabinose
and acetic acid from the liquid fraction, are found depending on the temperature, time and
sulphuric acid concentration, with errors less than 15 %. The equation found for the xylose
content does not reproduce experimental results adequately. For all considered dependent
variables, the most influential operation variable is the temperature and the least influential is
the acid catalyst concentration.
By comparing the values of the dependent variables related to the solid and liquid
fractions of EFB hydrothermal treatments, with and without a catalyst, it is found that using a
catalyst the maximum yield decreases 3.26 %, whereas the contents of glucose, xylose and
arabinose increased 52.63 %, 22 % and 26.43 %, respectively.
By simulating the hydrothermal treatment with an acid catalyst, it is found that
operating at 190 C, with a liquid/solid ratio of 6 and a sulphuric acid concentration of 0.1 %
for 15 minutes, relatively high values are obtained, with respect to the maximum ones, of
glucose (3.12 g/L), xylose (4.00 g/L), arabinose (2.35 g/L) and acetic acid (2.28 g/L), saving
capital by not operating with maximum values of operating variables, and achieving a solid
yield of 68 %.

3.

PULPING

PROCESSES

OF

THE

SOLID

FRACTION

OF

THE

HYDROTHERMAL TREATMENT OF EFB, AND ORIGINAL EFB


By comparing the values of the properties of the pulps and paper sheets corresponding
to different EFB pulps and the solid fraction of its hydrothermal treatment, using sodaanthraquinone, ethanol and ethanolamine, respectively, it is found that the kappa number
and brightness values of the pulps are improved when the pulping process has been
preceded by a hydrothermal treatment for the three pulping processes considered, but the
viscosity values of the pulps and the strength properties of the paper sheets (tensile, burst and
tear indexes) decrease. Therefore, when the EFB is subjected to a hydrothermal treatment,
with the idea of using the hemicellulosic fraction, the pulp obtained from the solid fraction of
such treatment is more conveniently used in the manufacture of specialty papers or in the
production of bioethanol by simultaneous enzymatic hydrolysis and fermentation.

364

Conclusiones/Conclusions
4. ORGANOSOLV PULPING PROCESSES
4.1. Formosolv pulping process
By using a factorial design of experiments to study the influence of the variables of the
pulping process of EFB with formic acid at atmospheric pressure, on the yield, kappa
number, viscosity and brightness of the pulps obtained, polynomial models that reproduce
considered dependent variables versus time and concentrations of formic and hydrochloric
(catalyst) acids are found with errors less than 10 %.
By using these polynomial models, operating conditions of engagement (92.5 % formic
acid, 0.075 % hydrochloric acid and 60 minutes) can be found, so as to obtain acceptable
values of the properties of pulps (42.3 % of yield, 22.7 % of brightness and 512 mL/g of
viscosity) and save chemicals and energy, compared to operating in a more energetic way.

4.2. Milox pulping process


In the pulping process of EFB by the Milox process, the influence of the operating
conditions in the first stage of the process on the characteristics of the pulps obtained has
been studied, by using a factorial design of experiments and polynomial as well as neural
fuzzy models, finding equations that reproduce the experimental values of the yield, kappa
number, viscosity and brightness of the pulps, depending on the concentrations of formic
acid and hydrogen peroxide and the process time, with errors less than 10 %.
By simulating the Milox process, using the polynomial and neural fuzzy models, it is
found that operating with 93 % of formic acid and 3 % of hydrogen peroxide for 165
minutes, a pulp with good properties (40.5 % of yield, 50.3 % of brightness and 608 mL/g of
viscosity) is obtained, saving energy and chemicals at the same time, compared to operating
under more severe conditions.
A second stage of the Milox process must be very short (30 minutes or less) to avoid a
negative effect on viscosity.

365

Conclusiones/Conclusions
5. CHARACTERIZATION OF THE RESIDUAL LIQUOR OF THE ORGANOSOLV
PULPING PROCESSES
The residual liquor of the pulping process of EFB (95 % formic acid, 0.05 %
hydrochloric acid and 30 minutes) has a lignin content of 18.2 %, 4.1 % of glucose, 9.8 % of
xylose and 1.2 % of arabinose, on a dry basis with respect to the original raw material.
The Milox process residual liquor contains no precipitable lignin, and contains low
amounts of fermentable monosaccharides (0.71 % of glucose, 4.22 % of xylose, 1.19 % of
galactose, 0.33 % of mannose and 0.82 % of furfural) on a dry basis of the original raw
material. This is probably due to the high oxidative power of the reaction mixture.

6. CELLULOSIC PULP BLEACHING


6.1. AOpAZRP bleaching sequence
From the study of pulp bleaching soda-anthraquinone and diethanolamine of EFB
with the sequence AOpAZRP (acid treatment - oxygen treatment with hydrogen peroxide acid treatment - ozone - reducing treatment - hydrogen peroxide) bleached pulps with
viscosity values and physical properties of the paper sheets (tensile, burst and tear indexes,
stretch and brightness) similar or higher than those of bleached pulps of other non-wood
materials and eucalyptus with sequences TCF (totally chlorine free), are obtained.
Diethanolamine EFB pulp after bleaching with the AOpAZRP sequence exhibits a
higher viscosity (783 mL/g) and properties of the paper sheets near or even better than those
presented by the paper sheets of the soda-anthraquinone pulp, bleached with the same
sequence (22.2 versus 20.4 Nm/g for tensile index, 1.30 versus 1.42 KN/g for the burst
index, 0.71 versus 0.70 mNm2/g for tear index and 71.3 versus 77.5 % of brightness).
6.2. EPabOPoP bleaching sequence
An EFB pulp obtained with formic acid (95 % formic acid, 0.05 % hydrochloric acid
and 30 minutes), with a kappa number of 31.1 and a viscosity of 606 mL/g, was bleached
using the sequence EPabOPoP (alkaline extraction - peracetic in basic medium - oxygen hydrogen peroxide with oxygen - hydrogen peroxide) obtaining a brightness of 69.4 %, and a
loss of viscosity and yield of 34.8 % and 17.7 %, respectively.

366

Conclusiones/Conclusions
6.3. Hydrogen peroxide bleaching
From the study of the influence of the bleaching conditions with hydrogen peroxide of
the pulp soda-anthraquinone from EFB, on the properties of the bleached pulps, following
a factorial design of experiments and polynomial and neural fuzzy models, it is found that
the values of viscosity, kappa number and brightness are reproduced depending on the
hydrogen peroxide and sodium hydroxide concentrations and reaction time, with errors less
than 10 %.
By simulating the pulp bleaching process of EFB, with polynomial and neural fuzzy
models found, it is possible to determine the optimum values of the operating variables, so
that the properties of the bleached pulps differ slightly from its best values, and at the same
time savings in chemicals, energy and size of the pilot plant are achieved by operating with
low values of operational variables. Thus, operating with a soda concentration of 1.13 %, a
hydrogen peroxide concentration of 2.25 % and with a time of 3 hours, it is possible to
obtain a pulp with a brightness of 75.1 % (8.1 % below the maximum) and a viscosity of 740
mL/g (10.4 % less than the maximum value).
In a comparative study of soda-anthraquinone pulp bleaching from EFB and

Hesperaloe funifera with hydrogen peroxide, it is concluded that, using a factorial design of
experiments, the application of neural fuzzy models leads to equations that predict values of
viscosity and brightness of the bleached pulps with hydrogen peroxide, in function of
operating variables (hydrogen peroxide and soda concentrations and reaction time), with
errors less than 10 % in the case of EFB pulp bleaching, and 15 % for Hesperaloe funifera
pulp.
By simulating the hydrogen peroxide bleaching, using neural fuzzy models, it is found
that operating with a soda concentration of 0.5 %, a hydrogen peroxide concentration of 3 %
and a time of 3 hours, a EFB pulp can be obtained with a brightness of 74.9 % (7.0 % below
the maximum) and a viscosity of 716 mL/g (12.7 % less than the maximum value); and a

Hesperaloe funifera pulp with a brightness of 63.3 % (7.7 % below the maximum) and a
viscosity of 584 mL/g (14.1 % less than the maximum value).

6.4. Enzymatic Bleaching


From a study of the soda-anthraquinone pulp biobleaching of EFB using stages with
xylanase and laccase enzymes, it is found that pulps with very competitive mechanical and
367

Conclusiones/Conclusions
optical properties are obtained. From the XEP, LEP, (X + L)EP and XLEP (X, xylanase; L,
laccase; E, alkaline extraction; P, hydrogen peroxide) sequences, the latter was the most
suitable, therefore the most efficient pretreatment for the biobleaching is when enzymes are
used consecutively, as they provide pulps with lower kappa number values (5.4) and higher
for the brightness (60.5 % ISO). Furthermore, this pretreatment also leads to pulps with
better colorimetric properties, even after accelerated ageing. Therefore, these two enzymatic
pretreatment steps might be a good option to reduce environmental problems and increase
the selectivity of the sequences ECF (elemental chlorine free) and TCF (totally chlorine free)
applied to EFB pulps.

7. NANOFIBRILLAR CELLULOSE PRODUCTION


The production of nanofibrillar cellulose from EFB fibers, using chemical and
mechanical treatments, has been studied.
The soda-anthraquinone, Milox and Formosolv pulping processes followed by a
mechanical microfluidization, allowed the production of suspensions of long, individualized
nanofibers. In all cases, the deconstruction of the EFB fibers and its passage to nanofibrillar
cellulose (NFC) resulted in a decrease in water absorption by the nanofibrillar cellulose films
formed, which constitutes an important property for applications such as packaging.
Nanofibrillar cellulose obtained from the fiber pulps of EFB shows better properties
compared with those presented for the nanofibrillar cellulose obtained using bleached wood
fibers, and therefore it is presented as a material with a high added value because it is a
widely available agro-industrial residue.
Finally, it is shown that the EFB fibers are a suitable raw material for the nanopaper
production with a large number of applications of high importance and with structural and
strength properties similar to those found for wood fibers.

368

Anexo/Appendix

Anexo/Appendix
NDICE DE FIGURAS

Figura 3.1. Componentes principales de los materiales lignocelulsicos ............................ 29


Figura 3.2. Estructura de la pared celular ............................................................................. 30
Figura 3.3. Distribucin de los componentes mayoritarios a travs de la
pared celular ........................................................................................................ 30
Figura 3.4. Estructura de la celulosa ..................................................................................... 31
Figura 3.5. Modelo de molculas de celulosa unidas por puentes de hidrgeno ............... 31
Figura 3.6. Representacin esquemtica de la estructura microscpica y
submicroscpica de la celulosa ........................................................................... 32
Figura 3.7. Estructura de las hemicelulosas y sus derivados ................................................ 34
Figura 3.8. Estructura de la lignina ........................................................................................ 35
Figura 3.9. Unidades estructurales de la lignina ................................................................... 35
Figura 3.10. Esquematizacin del concepto de Biorefinera ............................................... 44
Figura 3.11. Obtencin de productos de alto valor aadido a partir de
una Biorefinera de materiales lignocelulsicos ............................................... 48
Figura 3.12. Esquematizacin del tratamiento hidrotrmico ............................................... 49
Figura 3.13. Proceso de pasteado a la sosa ....................................................................... 66
Figura 3.14. Reacciones de la antraquinona ......................................................................... 67
Figura 3.15. Tratamiento de deslignificacin mediante disolventes orgnicos ................... 72
Figura 3.16. Estructura de la celobiosa ................................................................................. 95
Figura 3.17. Desarrollo de fibras celulsicas a partir de molculas de celulosa ................. 96
Figura 3.18. Imgenes de AFM de a) nanocristales de celulosa;
b) celulosa nanofibrilar ..................................................................................... 98
Figura 3.19. Imgenes de celulosa bacteriana ...................................................................... 99
Figura 3.20. Rutas diferentes para obtener celulosa nanofibrilar ........................................ 100
371

Anexo/Appendix
Figura 3.21. Plantacin de palma .......................................................................................... 103
Figura 3.22. Planta joven de palma aceitera ......................................................................... 104
Figura 3.23. Racimo de la palma aceitera ............................................................................. 105
Figura 4.1. Microscopio para medida de longitud de fibra de la materia prima ................ 117
Figura 4.2. Molino de la materia prima ................................................................................ 118
Figura 4.3. Digestor para tratamientos hidrotrmicos y pasteados ...................................... 118
Figura 4.4. Desfibrador de materiales cocidos ..................................................................... 119
Figura 4.5. Separador de fibras e incocidos .......................................................................... 120
Figura 4.6. Separador de fibras y agua .................................................................................. 120
Figura 4.7. Sprout-Bauer ....................................................................................................... 121
Figura 4.8. Refinador PFI ...................................................................................................... 121
Figura 4.9. Centrfuga ............................................................................................................ 122
Figura 4.10. Desintegrador .................................................................................................... 122
Figura 4.11. Refinmetro Shopper-Riegler .......................................................................... 123
Figura 4.12. Formador de hojas ............................................................................................ 124
Figura 4.13. Prensa para hojas y aros secadores ................................................................... 124
Figura 4.14. Microfluidizador ............................................................................................... 125
Figura 4.15. Equipo de sobrepresin para la formacin de pelculas de NFC .................... 125
Figura 4.16. Microscopio de fuerza atmica ........................................................................ 126
Figura 4.17. (De izquierda a derecha). Mquina universal para medida del ndice
de traccin, longitud de ruptura y alargamiento; equipo para
medida del ndice de estallido; equipo para medida del
ndice de desgarro; equipo para medida de la blancura ................................... 126
Figura 4.18. Transicin entre la pertenencia y la no pertenencia a un determinado
atributo, segn la lgica convencional y la lgica neuroborrosa ..................... 133
Figura 4.19. Estructura de un sistema neuroborroso ........................................................... 137
372

Anexo/Appendix
NDICE DE TABLAS

Tabla 3.1. Rendimiento promedio de algunas materias primas .......................................... 39


Tabla 3.2. Dimensiones de las fibras de plantas no leosas ................................................ 43
Tabla 3.3. Perspectivas de futuro para las tecnologas biomsicas ...................................... 46
Tabla 3.4. Objetivos mnimos de la UE con respecto a la obtencin
de bioenerga y bioproductos ............................................................................. 47

373

Anexo/Appendix
NOMENCLATURA

(X + L)EP

Xylanase and Laccase treatment + alkaline extraction + hydrogen


peroxide bleaching

Mean of the extreme values of X

[A]

Amount of dye adsorbed by the substrate (mg dye/mg substrate)

[E]

Free dye concentration at equilibrium (mg/mL)

SR

ShopperRiegler value

a, b, c, d

Constants (constantes)

AA

Acetic acid concentration

AAs

Acetic acid concentration when sulphuric acid is used as a catalyst

ABTS

2,2-Azino-bis-3-ethylbenzthiazoline-6-sulfonate

AFM

Atomic force microscopy (microscopio de fuerza atmica)

ANFIS

Adaptative Neural Fuzzy Inference System

AOpAZRP

A, acid treatment; Op, oxygen and peroxide stage; Z, ozone stage; R,


reductive treatment; P, peroxide stage

374

AOX

Organochlorine compounds

APMP

Alkaline Peroxide Mechanical Pulping process

AQ

Anthraquinone (antraquinona)

AR

Arabinose concentration

ARs

Arabinose concentration when sulphuric acid is used as a catalyst

ASG

Acetosyringone

BR

Brightness

BTAC

Comit Tcnico Asesor de la Biomasa

Grupos cromforos / Concentracin

CAM

Acid metabolism of Crassulaceans

Cc

Concentracin de cromforos

Anexo/Appendix
CE

Cloro - extraccin alcalina

CED

Cupriethylenediamine

CEH

Cloro - extraccin alcalina - hipoclorito

CEHEH

Cloro - extraccin alcalina - hipoclorito - extraccin alcalina hipoclorito

ci

Singleton defuzzier; constant parameter; neural fuzzy modeling


constant

CR

Congo Red

Dixido de cloro

DAHT

1,5diacetil-2,4dioxohexahidro-1,3,4triacina

DBO

Demanda bioqumica de oxgeno

DP

Degree of polymerization

DQO

Demanda qumica de oxgeno

DTMPA

Dietilenotriaminametileno-cido fosfnico

DTPA

Pentaacetato de dietilentriamina

DTPMPA

Dietilenotriaminapentametileno-cido fosfnico

Alkaline extraction stage

ECF

Elementary Chlorine Free (libre de cloro elemental)

ECN

Energy Research Center of the Netherlands

EDTA

Tetraacetato de etilendiamina

EFB

Empty fruit bunches

EFB-E

EFB and ethanol

EFB-EA

EFB and ethanolamine

EFB-H-E

EFB hydrolyzed and ethanol

EFB-H-EA

EFB hydrolyzed and ethanolamine

EFB-H-SA

EFB hydrolyzed and soda-AQ

EFB-SA

EFB and soda-AQ


375

Anexo/Appendix
EP

Alkaline extraction + hydrogen peroxide bleaching

EPabOPoP

Alkali extraction - peracetic in basic medium oxygen - hydrogen


peroxide with oxygen - hydrogen peroxide

EPFBF

Empty Palm Fruit Bunch Fiber

Formic acid concentration / Formosolv (FoOH) process

Fil

Trmino lingstico de entrada

Fmax

Maximum force

GL

Glucose concentration

Gl

Trmino lingstico de salida

GLs

Glucose concentration when sulphuric acid is used as a catalyst

Hydrochloric acid concentration / Hydromodule (liquid-solid ratio)

HEDTA

cido N-hidroxietilen diamina triactico

HexA

Hexenuronic acid

HIDE process

Hydrochloric acid ethanol containing soda process

HMF

5-hidroximetil-2-furfural

HO+

Hydroxonium ion

IDE process

Ethanol containing soda process

JCR

Journal Citation Reports

Nmero de variables

Kad

Adsorption equilibrium constant

KN

Kappa number

Width of the Gaussian distribution (ancho de la distribucin


gaussiana de la variable en cuestin) / Laccase

L/S

Liquid/solid ratio

LcEP

Laccase control treatment + alkaline extraction + hydrogen peroxide


bleaching

376

Anexo/Appendix
LEP

Laccase treatment + alkaline extraction + hydrogen peroxide


bleaching

LMS

Laccase-mediator system

Milox (Formic acid and hydrogen peroxide) process

Number of fuzzy rules (nmero de reglas borrosas)

MFC

Microfibrillar Cellulose (celulosa microfibrilar)

MW

Molecular weight (696.7 g/mol)

Number of independent variables (nmero de variables independientes)

Soda-anthraquinone (NaOH-AQ) process / Avogadros constant

nc

Nmero de puntos centrales

NFC

Nanofibrillar Cellulose (celulosa nanofibrilar)

NN

Neural Network

NTA

Triacetato de nitrilo

o.d.p.

Over dried pulp

OXZP

O, oxygen stage; X, enzymatic treatment; Z, ozone stage; P,


hydrogen peroxide stage

OZP

Oxygen - ozone - hydrogen peroxide

Hydrogen peroxide concentration (concentracin de perxido de


hidrgeno)

Constant

Po

Hydrogen peroxide stage

Reglas borrosas / liquid/solid ratio (relacin lquido/slido)

R2

Coeficiente de regresin

RH

Relative Humidity

Rl

Fuzzy rule (=1 Fil(xi,il))

RWCA

Relative Water Absorption Capacity


377

Anexo/Appendix
S

Sulphuric acid concentration / Soda concentration

SA

Molecular surface area (1.73 nm2)

SEM

Scanning Electron Microscopy

SSA

Specific Surface Area

Processing time (tiempo) / Student t-value

Temperature / Processing time

TCF

Totally Chlorine Free (totalmente libre de cloro)

TEMPO

Catalizador usado en presencia de exceso de leja (NaClO)

TMP

Pasta termomecnica

UE

Unin Europea

VI

Viscosity

WAOQ

WA, acid washing; O, oxygen delignification; Q, chelating stage

WCA

Water Contact Angle

WRV

Water Retention Value (valor de retencin de agua)

Wt

Weight of the sample after immersion

Wt0

Weight of the sample before immersion

Absolute experimental value of the variable concerned / Xylanase

X+L

Xylanase treatment followed by a laccase treatment

X1

Value of the variable for low level

X2

Value of the variable for high level

XC

Minimum, central or maximum value of variable (valor mnimo,


central o mximo de la variable)

XcLcEP

Xylanase control treatment + laccase control treatment + alkaline


extraction + hydrogen peroxide bleaching

378

XF

Normalized value of formic acid concentration

XH

Normalized value of hydrochloric acid concentration

Anexo/Appendix
Xhigh

Extreme value of the variable (the highest one)

Xi

Value of the operating variable for low, medium and high levels
(variable de entrada)

XL

Laccase / Xylanase

XLEP

Xylanase treatment + laccase treatment + alkaline extraction +


hydrogen peroxide bleaching

Xlow

Extreme value of the variable (the lowest one)

Xmax

Maximum value of X

Xmin

Minimum value of X

Xn

Normalized value of the independent variable (valor normalizado de


la variable independiente) / Variable de entrada

XO

Xilooligosacridos

XP

Normalized value of hydrogen peroxide concentration

XR

Normalized value of liquid/solid ratio

XS

Normalized value of sulphuric acid concentration

XT

Normalized value of temperature / Normalized value of process


time

Xt

Normalized value of process time

XY

Xylose concentration

Ye

Estimated value of the property to be modeled (valor estimado de la


propiedad que se quiere modelar)

YI

Solid fraction yield / Yield

YIs

Solid fraction yield when sulphuric acid is used as a catalyst

Yl

Defuzzifier of the fuzzy rule (consecuente de la regla borrosa)

Limiting viscosity

Density

379

Anexo/Appendix

380

Produccin Cientfica/Scientific Production

Produccin cientfica/Scientific production


ARTCULOS DE INVESTIGACIN/RESEARCH ARTICLES

1. Pulp and paper from vine shoots: Neural fuzzy modeling of ethylene glycol pulping.

L. Jimnez, V. Angulo, A. Rodrguez, R. Snchez, A. Ferrer.

Bioresource Technology, 100(2009), 756-762.

ndice de impacto: 4.253 (Primera posicin de 11 revistas en el rea de

Agricultural Engineering segn el Journal Citation Reports de 2009).


2. TCF bleaching of soda-anthraquinone and diethanolamine pulp from oil palm empty
fruit bunches.

L. Jimnez, L. Serrano, A. Rodrguez, A. Ferrer.

Bioresource Technology, 100(2009), 1478-1481.

ndice de impacto: 4.253 (Primera posicin de 11 revistas en el rea de

Agricultural Engineering segn el Journal Citation Reports de 2009).


3. Feasibility of rice straw as a raw material for the production of soda cellulose pulp.

Alejandro Rodrguez, Rafael Snchez, Ana Requejo, Ana Ferrer.

Journal of Cleaner Production, 18(2010), 1084-1091.

ndice de impacto: 2.430 (Dcima posicin de 45 revistas en el rea de

Engineering and Environmental segn el Journal Citation Reports de 2010).


4. Soda pulp and fuel gases synthesis from Hesperaloe funifera.

Snchez Rafael, Rodrguez Alejandro, Requejo Ana, Ferrer Ana, Navarro


Enrique.

Bioresource Technology, 101(2010), 7032-7040.

ndice de impacto: 4.365 (Primera posicin de 12 revistas en el rea de

Agricultural Engineering segn el Journal Citation Reports de 2010).


5. Simulation of Hesperaloe funifera diethanolamine pulping by polynomial and neural
fuzzy models.

Alejandro Rodrguez, Rafael Snchez, Ana Ferrer, Ana Requejo.


383

Produccin cientfica/Scientific production

Chemical Engineering Research and Design, 89(2011), 648-656.

ndice de impacto: 1.968 (Trigsima novena posicin de 133 revistas en el rea


de Engineering and Chemical segn el Journal Citation Reports de 2011).

6. Hesperaloe funifera as a raw material for integral utilization of its components.

Rafael Snchez, Ana Ferrer, Luis Serrano, Ana Toledano, Jalel Labidi, and
Alejandro Rodrguez.

BioResources, 6(1)(2011), 0003-0021.

ndice de impacto: 1.328 (Cuarta posicin de 21 revistas en el rea de

Materials Science, paper & wood segn el Journal Citation Reports de 2011).
7. Modeling hydrogen peroxide bleaching of soda pulp from oil-palm empty fruit
bunches.

Ana Ferrer, Antonio Rosal, Cristina Valls, Blanca Roncero, Alejandro


Rodrguez.

BioResources, 6(2)(2011), 1298-1307.

ndice de impacto: 1.328 (Cuarta posicin de 21 revistas en el rea de

Materials Science, paper & wood segn el Journal Citation Reports de 2011).
8. Study of 3 TCF bleaching processes applied to soda-AQ Hesperaloe funifera pulp.

Alejandro Rodrguez, Ana Requejo, Zoilo Gonzlez and Ana Ferrer.

Cellulose Chemistry and Technology, 45(5-6)(2011), 371-379.

ndice de impacto: 0.550 (Dcima posicin de 21 revistas en el rea de

Materials Science, paper & wood segn el Journal Citation Reports de 2011).
9. Milox fractionation of empty fruit bunches from Elaeis guineensis.

Ana Ferrer, Alberto Vega, Alejandro Rodrguez, Pablo Ligero, Luis Jimnez.

Bioresource Technology, 102(2011), 9755-9762.

ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de

Agricultural Engineering segn el Journal Citation Reports de 2011).


10. Pulping of empty fruit bunches (EFB) from the palm oil industry by formic acid.

384

Produccin cientfica/Scientific production

Ana Ferrer, Alberto Vega, Pablo Ligero and Alejandro Rodrguez.

BioResources, 6(4)(2011), 4282-4301.

ndice de impacto: 1.328 (Cuarta posicin de 21 revistas en el rea de

Materials Science, paper & wood segn el Journal Citation Reports de 2011).
11. A neural fuzzy model applied to hydrogen peroxide bleaching of non-wood soda
pulps.

A. Rosal, C. Valls, A. Ferrer and A. Rodrguez.

Cellulose Chemistry and Technology, 46 (1-2)(2012), 105-114.

ndice de impacto: 0.550 (Dcima posicin de 21 revistas en el rea de

Materials Science, paper & wood segn el Journal Citation Reports de 2011).
12. Biobleaching of pulp from oil palm empty fruit bunches with laccase and xylanase.

R. Martn-Sampedro, A. Rodrguez, A. Ferrer, L.L. Garca-Fuentevilla, M.E.


Eugenio.

Bioresource Technology, 110(2012), 371-378.

ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de

Agricultural Engineering segn el Journal Citation Reports de 2011).


13. Valorization of residual Empty Palm Fruit Bunch Fibers (EPFBF) by
microfluidization: Production of nanofibrillated cellulose and EPFBF nanopaper.

Ana Ferrer, Ilari Filpponen, Alejandro Rodrguez, Janne Laine, Orlando J.


Rojas.

Bioresource Technology, 125(2012), 249255.

ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de

Agricultural Engineering segn el Journal Citation Reports de 2011).


14. Effect of residual lignin and heteropolysaccharides in nanofibrillar cellulose and
nanopaper from wood fibers.

Ana Ferrer, Elisabet Quintana, Ilari Filpponen, Iina Solala, Teresa Vidal,
Alejandro Rodrguez, Janne Laine, Orlando J. Rojas.

Cellulose, 2012 (DOI 10.1007/s10570-012-9788-z).


385

Produccin cientfica/Scientific production

ndice de impacto: 3.600 (Primera posicin de 21 revistas en el rea de

Materials Science, paper & wood segn el Journal Citation Reports de 2011).
15. Influence of temperature, time, liquid/solid ratio and sulphuric acid concentration on
the hydrolysis of palm empty fruit bunches.

Ana Ferrer, Ana Requejo, Alejandro Rodrguez, Luis Jimnez.

Bioresource Technology, 2012 (10.1016/j.biortech.2012.10.081).

ndice de impacto: 4.980 (Primera posicin de 12 revistas en el rea de

Agricultural Engineering segn el Journal Citation Reports de 2011).

386

Produccin cientfica/Scientific production


PARTICIPACIN EN CONGRESOS/CONGRESS CONTRIBUTIONS

1. 15th International Symposium on Wood, Fiber and Pulping Chemistry (Oslo,


Noruega; Junio 2009).

Diethanolamine pulp from solid residue of the hydrothermal treatment of rice


straw.

Rodrguez, Alejandro; Moral, Ana; Serrano, Luis; Ferrer, Ana; Jimnez, Luis.

Comunicacin en forma de pster.

2. 11th European Workshop on Lignocellulosics and Pulp (Hamburgo, Alemania;


Agosto, 2010).

Biorefinery of agricultural residues by fractionation of their components.

A. Rodrguez, R. Snchez, A. Ferrer, E. Navarro, A. Rosal, L. Jimnez.

Comunicacin en forma de pster.

3. VI Congreso Iberoamericano de Investigacin en Celulosa y Papel (Lisboa, Portugal;


Octubre 2010).

Integral utilization of rice straw by using hydrothermal treatment and pulping


process.

Alejandro Rodrguez; Ana Ferrer; Ana Requejo; Luis Jimnez; Antonio Rosal.

Comunicacin en forma de pster.

4. VI Congreso Iberoamericano de Investigacin en Celulosa y Papel (Lisboa, Portugal;


Octubre 2010).

Hesperaloe funifera for pulp and synthesis and fuel gases from the removed
lignin.

Luis Jimnez, Rafael Snchez, Alejandro Rodrguez, Enrique Navarro, Ana


Ferrer.

Comunicacin en forma de pster.

387

Produccin cientfica/Scientific production


5. VI Congreso Iberoamericano de Investigacin en Celulosa y Papel (Lisboa, Portugal;
Octubre 2010).

Characterization by chemical and thermogravimetric analysis of Hesperaloe

funifera.

Alejandro Rodrguez, Rafael Snchez, Ana Requejo, Ana Ferrer, Jos L.


Ferrer, Luis Jimnez.

Comunicacin en forma de pster.

6. Jornadas Doctorales Doctores para la innovacin (Crdoba, Espaa; Junio 2011).

Fabricacin de papel a partir de materias primas alternativas a las


convencionales.

Ana Ferrer Carrera.

Comunicacin oral.

7. 243rd American Chemical Society National Meeting (San Diego, California; Marzo
2012).

Effect of lignin in the properties of nanofibrillated cellulose (NFC) from birch


pulp.

A. Ferrer, E. Quintana, I. Filpponen, J. Laine, T. Vidal, L. Jimnez, A.


Rodrguez, O.J. Rojas.

Comunicacin oral.

8. I Congreso Cientfico de Investigadores en Formacin en Agroalimentacin de la


eidA3 y II Congreso Cientfico de Investigadores en Formacin de la Universidad de
Crdoba (Crdoba, Espaa; Mayo 2012).

Effects of residual lignin on the properties of nanofibrillated cellulose and their


films.

Ana Ferrer Carrera.

Comunicacin oral.

9. 5th International Conference on Sustainable Energy & Environmental Protection


(Dubln, Irlanda; Junio 2012).

388

Produccin cientfica/Scientific production

Use by combustion of residual fractions from non-wood materials used in


paper pulp production.

A. Ferrer, Z. Gonzlez, A. Requejo, F. Vargas and A. Rodrguez.

Comunicacin en forma de pster.

10. ANQUE International Congress of Chemical Engineering "Innovating for the future"
(Sevilla, Espaa; Junio 2012).

Lignin and cellulose nanofibers and nanoparticles from pulping streams:


developments and applications.

Mariko Ago, Joo Silveira, Laura Taajamaa, Elisabeth Quintana, Ana Ferrer,
Ilari Filpponen, Joseph E. Jakes, Eero Kontturi, Teresa Vidal, Luis Jimnez,
Alejandro Rodrguez, Edison Bittencourt, Janne Laine, Orlando J. Rojas.

Comunicacin oral.

11. ANQUE International Congress of Chemical Engineering "Innovating for the future"
(Sevilla, Espaa; Junio 2012).

Organosolv formic acid pulping of empty fruit bunches (EFB) from palm oil
industry.

A. Ferrer, F. Vargas, A. Requejo, L. Jimnez and A. Rodrguez.

Comunicacin en forma de pster.

12. ANQUE International Congress of Chemical Engineering "Innovating for the future"
(Sevilla, Espaa; Junio 2012).

Residual liquor from pulping of Hesperaloe funifera for fuel gases.

F. Vargas, Z. Gonzlez, A. Ferrer, A. Rodrguez and L. Jimnez.

Comunicacin en forma de pster.

13. ANQUE International Congress of Chemical Engineering "Innovating for the future"
(Sevilla, Espaa; Junio 2012).

Ecological packaging from cereal straws cellulosic pulps.

F. Vargas, Z. Gonzlez, A. Requejo, A. Ferrer, R. Snchez and A. Rodrguez.

Comunicacin en forma de pster.


389

Produccin cientfica/Scientific production


14. 12th European Workshop on Lignocellulosics and Pulp (Espoo, Finlandia; Agosto
2012).

Use from olive tree prunings by hydrothermal and combustion processes.

A. Requejo, Z. Gonzlez, A. Ferrer, A. Rodrguez and L. Jimnez.

Comunicacin en forma de pster.

15. ABTCP 2012 - 45th Pulp and Paper International Congress and Exhibition. VII

IberoAmerican Congress on Pulp and Paper Research (So Paulo, Brasil; Octubre
2012).

Fractionation of empty fruit bunches from palm oil by milox process.

Ana Ferrer, Alberto Vega, Ana Requejo, Alejandro Rodrguez, Luis Jimnez,
Jos Luis Ferrer.

Comunicacin en forma de pster.

16. ABTCP 2012 - 45th Pulp and Paper International Congress and Exhibition. VII

IberoAmerican Congress on Pulp and Paper Research (So Paulo, Brasil; Octubre
2012).

Waste agricultural of banana tree used by soda-anthraquinone and combustion


processes.

Antonio Rosal, Ftima Vargas, Zoilo Gonzlez, Alejandro Rodrguez, Ana


Ferrer, Luis Jimnez, Jos Luis Ferrer.

390

Comunicacin oral.