Está en la página 1de 137

0 P ! | C A L P k 0 P £ k ! | £ S A N 0 0 U A N ! U M C 0 N F | N £ M £ N !

0 F
N A N 0 C k Y S ! A L L | N £ | | - v | S £ M | C 0 N 0 U C ! 0 k P A k ! | C L £ S
Utrecht University
OP T I C A L P R O P E R T I E S A N D QU A N T U M
C O N F I N E ME N T O F NA N O C R Y S T A L L I N E
I I - VI S E MI C O N D U C T O R P A R T I C L E S
Optische e ige nscha ppe n e n kwa ntum
opsluitingsef f ecten va n na nokrista llijne
II- VI ha lf geleiderdeeltjes
( met een samenvatti ng i n het Nederlands)
PROEFSCHRI FT
TER VERK RI JGI NG VAN DE GRAAD VAN DOCTOR AAN DE UNI VERSI TEI T UTRECHT
O P GEZAG VAN DE RECTOR MAGNI FI CUS, PROF. DR. H.O. VOORMA, I NGEVOLGE
HET BESLUI T VAN HET COLLEGE VOOR PROMOTI ES I N HET OPENBAAR TE
VERDEDI GEN OP WOENSDAG 6 OKTOBER 1999 DES MI DDAGS TE 12:45 UUR
DOOR
ALBERT VAN DI JK EN
GEBOREN OP 19 AUGUSTUS 1972 TE BENNEK OM
P R O MO T O R : Prof. Dr. A. Mei jeri nk
C O - P R O MO T O R : Dr. D. Vanmaekelbergh
Facultei t Schei kunde
Uni versi tei t Utrecht
CI P-GEGEVENS K ONI NK LI JK E BI BLI OTHEEK , DEN HAAG
Di jken, Albert van
Opti cal Properti es and Quantum Confi nement of Nanocrystalli ne I I -VI Semi conductor
Parti cles / Albert van Di jken. -Utrecht : Uni versi tei t Utrecht, Facultei t Schei kunde, Debye
I nsti tuut.
Proefschri ft Uni versi tei t Utrecht. Met een samenvatti ng i n het Nederlands.
I SBN 90-393-2164-7
The work descri bed i n thi sthesi si ssupported by the Counci l for Chemi cal Sci ence ( CW) ,
wi th fi nanci al ai d from the Netherlands Organi zati on for Sci enti fi c Research ( NWO) .
"It has long been an axiom of mine that the little
things are infinitely the most important."
( Sherlock Holmes i n " A Case of I denti ty" , from The
Adventures of Sherlock Holmes by Arthur Conan Doyle) .
C O N T E N T S
1. I n t r o d u c t i o n
Nanocrystalli ne I I -VI Semi conductor Parti cles and Quantum Si ze Effects : Past and Present 1
1.1 Early observati ons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Of boxes and wells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Reduci ng the di mensi ons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.4 Colloi dal suspensi ons of sulphi des . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.5 Modelli ng quantum si ze effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.6 Recent small parti cle research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.7 Summary of thi s thesi s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2. S i z e S e l e c t i v e P h o t o e t c h i n g
Adjusti ng the Si ze of Nanocrystalli ne I I -VI Semi conductor Parti cles . . . . . . . . . . . . . . . . . 15
2.1 I ntroducti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.2 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3 Experi mental methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.1 Sample preparati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.3.2 Photoetchi ng . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.3 Opti cal measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.4 Transmi ssi on Electron Mi croscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.3.5 X-ray Powder Di ffracti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4 Results and di scussi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.1 CdS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4.2 ZnS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.4.3 PbS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.4.4 ZnO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.5 Conclusi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3. Na n o c r y s t a l l i n e Zn O Pa r t i c l e s
I . I denti fi cati on of the Transi ti on Responsi ble for the Vi si ble Emi ssi on . . . . . . . . . . . . . . . 41
3.1 I ntroducti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.2 Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.3 Experi mental methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.3.1 Sample preparati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.3.2 Opti cal measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.4 Results and di scussi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4.1 Emi ssi on measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4.2 I denti fi cati on of the vi si ble emi ssi on transi ti on . . . . . . . . . . . . . . . . . . . . 48
3.5 Conclusi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4. Na n o c r y s t a l l i n e Zn O Pa r t i c l e s
I I . The K i neti cs of the Radi ati ve and Non-Radi ati ve Processes upon Photoexci tati on . . . . 55
4.1 I ntroducti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.2 Experi mental methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.2.1 Sample preparati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.2.2 Opti cal measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.3.1 Absorpti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.3.2 Steady-state lumi nescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.3.3 Ti me-resolved lumi nescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.3.4 Temperature-dependent lumi nescence . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.4 Di scussi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.4.1 Steady-state lumi nescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.4.2 Ti me-resolved lumi nescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
4.4.3 Temperature-dependent lumi nescence . . . . . . . . . . . . . . . . . . . . . . . . . . 66
4.4.4 Nature of the deep hole trap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.4.5 Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.5 Conclusi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5. Na n o c r y s t a l l i n e Zn O Pa r t i c l e s
I I I . The I nfluence of Adsorbed Oxygen on the Emi ssi on Properti es . . . . . . . . . . . . . . . . . . 77
5.1 I ntroducti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.2 Experi mental methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.2.1 Sample preparati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.2.2 Opti cal measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5.4 Di scussi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
5.5 Conclusi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
6. Lu mi n e s c e n c e Qu a n t u m Ef f i c i e n c i e s
The I nfluence of Preparati on, Surface Passi vati on and Parti cle Si ze . . . . . . . . . . . . . . . . . 93
6.1 I ntroducti on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.2 Experi mental methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6.2.1 Sample preparati on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6.2.2 Lumi nescence quantum effi ci enci es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6.3.3 Opti cal measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.3 Results and di scussi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.3.1 CdS prepared wi th H S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2
6.3.2 CdS prepared wi th Na S . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
2
6.3.3 Lumi nescence quantum effi ci enci es of CdS . . . . . . . . . . . . . . . . . . . . . . . 99
6.3.4 ZnO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
6.3.5 Lumi nescence quantum effi ci enci es of ZnO . . . . . . . . . . . . . . . . . . . . . 101
6.4 Conclusi on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
S a me n v a t t i n g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Li s t o f p u b l i c a t i o n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Ov e r d e i n h o u d v a n d i t p r o e f s c h r i f t
Een overzi cht voor ni et-i ngewi jden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
To t b e s l u i t
Een woord van dank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
C u r r i c u l u m Vi t a e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
1
C H A P ¡ £ k
| N ¡ k 0 0 U C ¡ | 0 N
¡
Nonccr ysf o| | | ne | | -v| 5em| ccnducf cr Por f | c| es
ond Quonf um 5| ze £ f f ecf s : Posf ond Pr esenf
A 8 5 ¡ k A C ¡ .
1lis clupter gives u slort listoricul overview ol tle development ol u
new interdisciplinury lield in muteriuls science . tle study ol
nunocrystulline semiconductor purticles. 1le overview sturts witl tle
ohservution ol unomulous ellects, muinly in uhsorption meusurements
on smull semiconductor purticles. Luter, tlese unomulities turned out
to he munilestutions ol soculled quuntum size ellects wlicl will he tle
muin topic ol tlis clupter und tle rest ol tlis tlesis. lt will he mude
cleur wlere tle work tlut is descrihed in tlis tlesis lits into tle vust
umount ol experiments heing perlormed in tle lield ol nunocrystulline
semiconductor purticles toduy.
Clupter 1 . lntroduction
2
¡ . ¡ £ A k L Y 0 8 5 £ k v A ¡ | 0 N 5 .
ln tle 1D2O's it wus ohserved tlut tle onset ol uhsorption us well us tle
luminescence colour ol glusses contuining CdS colloids, slilts to longer wuvelengtls
upon leuting. 1lis wus correluted to tle growtl ol CdS colloids j1| .
¨l|c //oorcs:co:/orhc ooJ J|c Ahsorµ/|oosçrco:c Jcr çc/hco CoJn|on
so//|Jç/òscr ocrscb|ch/ s|cb ]...] oocb Jco /ooçcrco \c//co, oos oo/ c|o
\ocbsco Jcr /òrhcoJco lo//o|Jc :oròc/çc/òbr/ o|rJ.¨
ln tle lute 1D6O's, reseurclers reported dillerences hetween tle uhsorption
spectru ol colloidul semiconductor purticles und tle spectru ol tle corresponding
mucrocrystulline muteriul j24|. ln 1D67, ßerry reported tlut tle uhsorption onset ol
suspensions ol smull crystuls ol Agßr j2| und Agl j8| wus slilted to slorter wuvelengtls
us compured to tle mucrocrystulline muteriul, und mude tle lollowing stutement .
¨Tbc ohscrocJ sb|// o/ /bc ohsorµ/|oo coroc /o sbor/cr oooc/coç/bs
sboo/J oo/ hc rcçorJcJ os soççcs/|oç o nccboo|sn |o ob|cb /bc
hooJçoµ |s o|JcocJ, ho/ os c|/bcr o JccrcoscJ oonhcr o/ ohsorh|oç
o/ons or o JccrcoscJ c//|c|cocy o/ /bc µboooooss|s/cJ c/cc/roo|c
/roos|/|oos.¨
ln 1D68, Stusenko presented experiments on tlin lilms ol CdS j5|, motivuted hy
tle prediction tlut lor very tlin semiconductor lilms tle hundgup is inversely
proportionul to tle squure ol tle lilm tlickness j6|. 1le results were in good ugreement
witl tle tleoreticully predicted slilt ol tle hundgup .
¨Socb o /orh|JJcococrçy çoµ |ocrcosc ]...] |s cooocc/cJ o|/b µooo/on
c//cc/s.¨
As will he slown in tle next purt ol tlis clupter, it wus reulised in tle eurly
1D8O's tlut tlese µooo/on c//cc/s ure not only responsihle lor tle dillerent properties ol
tlin lilms us compured to tle mucrocrystulline muteriul, hut ulso lor tle peculiur
heluviour ol colloidul semiconductor purticles. Some 15 yeurs ulter tle lirm stutement
hy ßerry, it wus slown tlut tle meclunism tlut le ruled out (widening ol tle hundgup,
is in luct responsihle lor lis ohservutions.
8
¡ . 2 0 F 8 0 X £ 5 A N 0 W £ L L 5 .
1le tleoreticul description ol tle inlluence ol tlickness (or size, ol u muteriul
on its electronic properties is hused on un elementury quuntum meclunicul concept,
known us tle ¨µor/|c/c|oohox¨. Wlen u purticle is conlined to u limited spuce, its
kinetic energy cun only luve discrete vulues tlut ure determined hy tle muss ol tle
µor/|c/c und tle dimensions ol tle hox. 1le lurger tle hox, tle smuller tle sepurution
hetween tle dillerent ullowed energy levels will he. Wlen tle hox hecomes inlinitely
lurge, u continuum ol energy levels is lormed. ßused on tlis model, it cun he expected
tlut tle properties ol clurge curriers in solids will slow u sizedependency wlen tle
dimensions ol tle solid hecome sulliciently smull. 1le lirst time tlut it wus suggested
tlut quuntisution ol tle kinetic energy could possihly he ohserved in very tlin
semiconducting lilms wus in 1D57 hy Sclrieller j7| hut it took unotler ten yeurs helore
tlese ellects were uctuully ohserved. 1lis wus uclieved in soculled |oocrs|oo /oycrs,
wlicl ure very nurrow luyers (∼1O nm, ut tle surluce ol u semiconductor. 1ley heluve
us potentiul energy wells in wlicl electrons (or loles, cun he contuined. ln 1D66,
energy quuntisution wus ohserved lor electrons in un inversion luyer on u silicon surluce
j8,D|. From tlut time on, muny experiments luve heen vuluuhle in demonstruting
quuntum size ellects j5,1O|.
A mu¦or hreuktlrougl took pluce ut tle end ol tle 1D6O's, wlen u teclnique
culled Mo/cco/or 8con Lµ|/oxy (VßL, wus developed ut tle ßell 1eleplone
Luhorutories j11,12|. 1lis puved tle wuy lor tle prepurution ol semiconductor
structures tlut luve since heen studied very extensively . µooo/on oc//s. 1lese urtiliciul
structures ure lormed hy sundwicling u very tlin luyer ol u semiconductor muteriul witl
u smull hundgup hetween two luyers ol u semiconductor witl u lurger hundgup. Witlin
tle smullhundgup luyer, tle clurge curriers ure sputiully conlined in one dimension,
wlile tley ure lree to move in tle otler two dimensions. A quuntum well is un exumple
ol u twodimensionul (2D, structure. Soon ulter tle development ol VßL, reseurclers
reported tle experimentul ohservution ol quuntised energy levels ol conlined clurge
curriers in quuntum wells j18|.
¡ . 3 k £ 0 U C | N 0 ¡ H £ 0 | M £ N 5 | 0 N 5 .
Following tle invention und development ol VßL, muny dillerent 2Dstructures
luve heen prepured und investiguted. Severul interesting plysicul properties were
ohserved und reseurclers suhsequently tried to reduce tle dimensions ol tlese
structures even lurtler. 1lis reduction lirst resulted in tle prepurution ol one
Clupter 1 . lntroduction
4
dimensionul µooo/on o|rcs. 1lese 1Dstructures cun he prepured directly hy VßL or
indirectly, sturting witl 2Dstructures wlicl ure luterully putterned hy litlogruplic or
dryetcling teclniques. Anotler upproucl is tle conlinement ol clurge curriers in
quuntum wells hy upplying strong mugnetic lields. ßy vurying tle strengtl ol sucl u
mugnetic lield, tle conlinement cun he tuned hetween tle 2D und OD limits, ullowing
tle study ol µooo/on Jo/s. An ulternutive upproucl to studying ODstructures is tle
prepurution ol quuntum dots in tle gus pluse hy luser vuporisution j14,15|.
1le 2D, 1D und ODstructures mentioned tlus lur ure ull or/|/|c|o/ structures,
prepured hy plysicul teclniques. Otler, more oo/oro/ structures luve ulso received
consideruhle uttention. 1lese structures ure prepured clemicully und include
con¦uguted polymers j16| und smull crystulline semiconductor purticles prepured in
zeolite losts j17,18|, in glusses j1D,2O|, in polymer lilms j21| or us colloids in
suspension. Vuinly tle lutter sumples luve heen suh¦ected to un enormous umount ol
experiments since tle eurly 1D8O's. 1le lollowing purugrupls will present un overview ol
tlis reseurcl.
¡ . 4 C 0 L L 0 | 0 A L 5 U 5 P £ N 5 | 0 N 5 0 F 5 U L P H | 0 £ 5 .
1le interest in colloidul suspensions ol nunocrystulline semiconductor purticles
originuted lrom plotocutulysis. lurticulute semiconductor systems were studied
hecuuse tley ure inexpensive, relutively eusy to luhricute und luve u ligl surluce ureu
j22|. An udditionul udvuntuge ol colloidul systems is tlut kinetic studies cun he
perlormed in more detuil. One ol tle populur semiconductors during tle eurly 1D8O's
wus CdS. 1lis muteriul uhsorhs visihle liglt und tle reduction ol CO to metlunol
2
j28,24| und even tle direct splitting ol wuter to O und H upon irrudiution ol CdS
2 2
powders j25| wus reported.
1lese eurly experiments initiuted un enormous umount ol reseurcl into tle
plotocutulytic properties ol colloidul suspensions ol CdS purticles, purticulurly in tle
group ol Henglein j2628|. Lllorts were directed towurds ohtuining extremely smull CdS
purticles, wlicl eventuully resulted in tle prepurution ol u CdS powder tlut wus wlite
insteud ol yellow (tle colour ol mucrocrystulline CdS,. Also, studies on tle plotounodic
dissolution ol colloidul CdS purticles slowed tlut tle onset ol uhsorption slilted
towurds ligler energies us tle purticles dissolved j26|. As wus mentioned in tle
heginning ol tlis clupter, similur sizedependent clunges ol tle uhsorption
cluructeristics lud ulreudy heen ohserved lor colloidul suspensions ol silver lulides,
hut tle origin wus tlen still uncleur. At lirst, it wus tlouglt tlut tle ohserved ellects
were due to un umorplous structure ol tle colloidul purticles j26|. However, ßrus wus

AN?

A D A D
"
= =
A
πε
ε ε
∞ ∞
∗ ∗ ∗ ∗
   
= ⋅ ⋅ + = ⋅ ⋅ +
   
   

5
(1.1,
tle lirst to slow tlut tlese purticles lud un ordered structure und tlut tle dependence
ol tle opticul properties on purticle size cun he understood in terms ol quuntum size
ellects j2D,8O|. As mentioned in tle previous purugrupls, tlis interpretution lud ulreudy
heen used to descrihe tle heluviour ol tlin semiconducting lilms. ßesides ßrus, otler
groups ulso reported size ellects in smull semiconductor purticles j81|. Nevertleless,
ßrus is generully credited lor heing tle lirst to develop tle tleoreticul lrumework lor tle
description ol quuntum size ellects in nunocrystulline semiconductor purticles.
¡ . 5 M 0 0 £ L L | N 0 Q U A N ¡ U M 5 | Z £ £ F F £ C ¡ 5 .
For mucrocrystulline semiconductors, tle hundgup is delined us tle minimum
energy needed to excite un electron lrom tle vulence hund to tle conduction hund.
1lrougl Coulomh interuction, tlese clurge curriers muy lorm u hound stute u so
culled Wunnier exciton wlicl lus un energy sligltly lower tlun tle hundgup. An
exciton cun he viewed us un electron orhiting u lole ut u certuin distunce in u dielectric
medium. ln unulogy witl u lydrogen utom, tlis distunce is culled tle ßolr rudius ol tle
exciton (o , j82| .
exc
ln equution (1.1,, o is tle ßolr rudius ol u lydrogen utom (O.52D A,, ε tle ligl
O

lrequency relutive dielectric constunt ol tle medium, n* und n* tle ellective musses ol
e l
tle electron und lole respectively (hotl in units n ,. Usuully, tle ellective musses ol tle
O
clurge curriers ure only u smull lruction ol tle electron rest muss. 1lis in comhinution
witl tle luct tlut tle Coulomh interuction hetween tle electron und lole is ulmost
entirely screened (ε ∼5, results in u relutively lurge ßolr rudius ol tle exciton (lor

instunce, o is uhout 8O A lor CdS,. As tle size ol tle semiconductor purticle
exc
upproucles tle ßolr rudius ol tle exciton, tle electronlole puir gets sputiully conlined
und ussumes u stute ol ligler kinetic energy. 1lerelore, wlen considering quuntum size
ellects in semiconductor purticles, tle ßolr rudius ol tle exciton gives un indicution ol
tle dimensions ut wlicl tlese ellects hecome uppurent.
1le lirst model culculutions hy ßrus were hused on tle ellective muss
upproximution und used u treutment ol hulk stutes in tle limit ol smull size j88|. 1le
ellective muss upproximution lud heen developed lor mucrocrystulline muteriuls hut it
lud ulreudy heen slown tlut it could ulso he used to quuntitutively descrihe tle motion

C

A D
D &
& "
A
- -
4 4 πε ε
∗ ∗

 
= + ⋅ + −
 
 
Clupter 1 . lntroduction
6
(1.2,
ol excitons in very tlin semiconductor luyers j84|. Lllects tlut could complicute tle
culculutions, sucl us structurul reurrungements us u lunction ol size or tle involvement
ol surluce stutes, were not considered. 1le system wus upproximuted hy u splericul
semiconductor purticle witl un inlinitely ligl potentiul energy outside tle splere. For
tle energy ol tle lowest excited stute (compuruhle to tle vulue lor tle hundgup in tle
cuse ol u mucrocrystulline semiconductor, us u lunction ol tle purticle rudius R, tle
lollowing lormulu wus derived .
ln equution (1.2,, next to tle hundgup ol tle mucrocrystulline muteriul (L ,, two terms
g
ure given, hotl dependent on tle purticle size. 1le lirst term is u conlinement term,
wlicl increuses us R und tle second term is tle Coulomh uttruction, increusing witl
−2
R . ln tle limit ol lurge R, tle vulue lor L upproucles tlut ol L . Altlougl tle ellective
−1
g
muss upproximution is not vulid lor vey smull purticle sizes, equution (1.2, lus olten
heen used to descrihe quuntum size ellects in nunocrystulline semiconductor purticles.
1le dependency ol L on R us given hy equution (1.2, cun he used to descrihe tle
results lrom sizedependent meusurements ol opticul properties luirly well. Since tle
eurly culculutions hy ßrus, severul pupers luve uppeured presenting models tlut ure
hused on more reulistic houndury conditions, sucl us u linite hurrier leiglt j85,86|.
Lquution (1.2, implies tlut wlen tle rudius ol u semiconductor purticle
decreuses, tle energy lor tle lowest excited stute increuses. ln otler words, tlis meuns
tlut wlen u purticle hecomes smuller, its ¨hundgup¨ increuses. As u consequence ol tle
sputiul conlinement ol clurge curriers, tle kinetic energy hecomes quuntised. Strictly
speuking, tle word ¨hundgup¨ is inuppropriute lor very smull purticles. 1le quuntisution
ol tle kinetic energy munilests itsell us u gruduul trunsition ol continuous energy hunds
to discrete energy levels. 1lis trunsition sturts ut tle edges ol tle energy hunds us tlese
stutes ure purticulurly sensitive to tle purticle size. Figure 1.1 sclemuticully slows tle
clunge ol tle electronic properties ol u semiconductor us its size decreuses to vulues
lower tlun tle ßolr rudius ol tle exciton.
Figure 1.1 : Schematic representation of the gradual change of the electronic
properties when the size of a semiconductor decreases (from left to right). At the left,
an energy band diagram for a macrocrystalline semiconductor is shown, E
g
being the
bandgap.. The valence band (full, black) and conduction band (empty, white) are
denoted as VB and CB respectively. In the case of the middle picture, the size of the
semiconductor is comparable to the size of the exciton while the picture on the right
represents the situation when the dimensions of the semiconductor are smaller than
those of the exciton (E is the energy of the lowest excited state).
E
g
E
VB
CB
7
Next to tlese ellects, tle increuse ol tle relutive surluce ureu ulso results in
mu¦or clunges ol tle opticul properties ol very smull semiconductor purticles. As tle
numher ol utoms ut tle surluce hecomes compuruhle to tle numher ol hulk utoms,
surluce stutes will pluy un importunt role, lor instunce in providing putlwuys lor non
rudiutive recomhinution ol plotogeneruted clurge curriers j87|. However, surluce
stutes cun ulso he involved in rudiutive trunsitions, us will he demonstruted in clupter 4
ol tlis tlesis.
¡ . 6 k £ C £ N ¡ 5 M A L L P A k ¡ | C L £ k £ 5 £ A k C H .
1le previous purugrupls descrihed tle period in wlicl tle loundutions were
luid lor tle development ol u new ureu ol reseurcl. Since tle lirst correct interpretution
ol tle sizedependent opticul properties ol nunocrystulline semiconductor purticles in
tle eurly 1D8O's, u lot ol reseurcl lus heen directed towurds studying tlese muteriuls.
Clupter 1 . lntroduction
8
1le comhinution ol u good tleoreticul lrumework und soplisticuted prepurutive
teclniques lus creuted u new lield ol muteriuls science tlut lus uttructed scientists lrom
muny dillerent disciplines. Over tle yeurs, severul excellent review urticles luve
uppeured j8848|. Lspeciully tle possihility to ud¦ust tle electronic properties ol
semiconductor purticles hy clunging tleir size lus stimuluted reseurcl on tlese
systems.
An interesting uccomplislment is tle construction ol ligltemitting diodes
(LLD's, hused on nunocrystulline semiconductor purticles j4D,5O|. ln semiconductors,
tle direct recomhinution ol plotogeneruted clurge curriers gives u nurrow emission
hund ut un energy close to tle uhsorption onset. For nunocrystulline semiconductor
purticles tlis meuns tlut tle luminescence is sizetunuhle. LLD's hused on tlese systems
cun tlus he luhricuted witl u wlole runge ol output colours. Anotler recent
development is tle sellorgunisution ol nunocrystulline semiconductor purticles into 8D
superluttices wlicl mukes it possihle to study collective electronic plenomenu tlut
urise lrom inteructions hetween tle purticles j51|.
1le study ol sizedependent opticul properties ol nunocrystulline
semiconductor purticles in suspension is complicuted hy inlomogeneous hroudening
ol tle luminescence signuls due to u distrihution ol purticle sizes. A very importunt
recent development tlut eliminutes tlis complicution is tle possihility to do
plotoluminescence spectroscopy on single semiconductor purticles. ln tle cuse ol
CdSe purticles, sucl meusurements luve presented interesting results. For tle exciton
emission, resolution limited spectrul linewidtls ol uhout 12O µeV ut 1O K were reported
j52|. Furtlermore, it wus slown tlut tle emission ol u single lluorescing CdSe purticle
turns on und oll under continuous illuminution. ßotl plenomenu cunnot he ohserved
wlen studying un ensemhle ol CdSe purticles j58|.
¡ . I 5 U M M A k Y 0 F ¡ H | 5 ¡ H £ 5 | 5 .
ln tlis tlesis, experiments ure descrihed tlut were perlormed on suspensions ol
nunocrystulline llVl semiconductor purticles. 1le oh¦ect ol tlis reseurcl is to study
quuntum size ellects in relution to tle luminescence properties ol tlese purticles. A pre
requisite lor perlorming studies ol sizedependent opticul properties is tle uvuiluhility ol
monodisperse purticles witl dillerent purticle sizes. Clupter 2 will deul witl tlis suh¦ect
in more detuil us it will descrihe u metlod culled s|:csc/cc/|oc µbo/oc/cb|oç tlut
provides u wuy to ud¦ust tle purticle size in u controlled munner ulter prepurution. ln
tlis clupter CdS is used us u model system und it is slown tlut tle initiul meun purticle
rudius ol 85 A cun he decreused gruduully to 7.5 A using liglt ol dillerent wuvelengtls.
D
At tle sume time, tle widtl ol tle purticle size distrihution decreused lrom 4Oº to 1O
15º. 1le colour ol tle liglt tlut is used lor plotoetcling determines tle size ol tle
semiconductor purticles. 1le leusihility ol sizeselective plotoetcling us u meuns to
decreuse tle size ol ZnS und ZnO purticles is ulso demonstruted, ultlougl tle
prepurution ol u lurge vuriety ol purticle sizes could not he uccomplisled.
Witl respect to quuntum size ellects in nunocrystulline llVl semiconductor
purticles, tle sulplides und selenides ure hy lur tle most extensively studied muteriuls.
Vucl less is known uhout oxidic llVl semiconductors, ol wlicl ZnO is prohuhly tle
most suituhle compound lor studies ol sizedependent opticul properties. 1lis
compound is known to luminesce quite strongly und it is relutively eusy to prepure u
runge ol purticle sizes, smull enougl to slow quuntum size ellects. 1le clupters 85 ol
tlis tlesis comprise u vuriety ol luminescence experiments tlut were perlormed on
nunocrystulline ZnO purticles witl dillerent sizes. Upon plotoexcitution, u ZnO
suspension slows two emission hunds, one heing u UV emission hund (exciton
emission, und tle otler u visihle emission hund (trup emission,. 1le visihle emission is
wellknown und it is tle muin reuson lor tle use ol ZnO us u plosplor in vurious
upplicutions. ßotl nunocrystulline ZnO purticles us well us mucrocrystulline ZnO slow
tle visihle emission und it prohuhly lus tle sume origin in hotl systems. Despite
numerous studies using mucrocrystulline ZnO tle exuct meclunism ol tle green
emission is still unknown.
ln clupter 8, u study ol tle sizedependence ol tle two emission hunds ol ZnO
is presented. Witl increusing purticle size, hotl emissions slilt to lower energies due to
size quuntisution. 1le experiments slow tlut u lineur relutionslip exists hetween tle
energetic positions ol tle visihle emission hund und tle UV emission hund. From tle
slope ol tle lineur relutionslip it cun he concluded tlut tle visihle emission is due to
tle recomhinution ol un electron lrom tle conduction hund witl u lole trupped uhout
2 eV helow tle conduction hund edge. As tle energetic position ol tlis trup level is in
good ugreement witl tle position ol u V centre in ZnO, it is ussumed tlut tlis delect is
O
••
tle recomhinution centre lor tle visihle emission.
Wlile tle conclusion ol clupter 8 is vulid lor ZnO in generul, clupters 4 und 5
concern tle specilic heluviour ol nunocrystulline ZnO purticles. ln clupter 4, results ure
presented ol steudystute und timeresolved luminescence meusurements perlormed on
suspensions ol nunocrystulline ZnO purticles ol dillerent sizes und ut dillerent
temperutures. 1le temperuture und sizedependence ol tle rutio ol tle visihle to
exciton luminescence und tle kinetics ure expluined hused on tle identilicution ol tle
trunsition responsihle lor tle visihle emission (clupter 8,. A model is proposed in
wlicl tle plotogeneruted lole is trunslerred lrom tle vulence hund to u V level in tle
O

hulk ol tle purticle in u twostep process. 1le lirst step ol tlis process is un ellicient
Clupter 1 . lntroduction
1O
surlucetrupping prohuhly ut un O site lollowed hy tunneling ol tle surlucetrupped
2−
lole huck into tle purticle wlere it recomhines witl un electron in un oxygen vucuncy
(V , resulting in tle ceution ol u V centre, tle recomhinution centre lor tle visihle
O O
• ••
emission.
A remurkuhle property ol ZnO purticles is tle quencling ol tle visihle emission
wlen un oxygenlree suspension is irrudiuted witl UV rudiution. At tle sume time, tle
intensity ol tle UV emission increuses. ln tle pust, vurious models luve heen proposed
to expluin tlis plenomenon. ln clupter 5, tle quencling heluviour is investiguted in
detuil. ßused on tle model lor tle visihle emission descrihed in clupters 8 und 4, tle
quencling ol tlis emission is uscrihed to clurging ol tle ZnO purticles. ln u ZnO
purticle, u plotogeneruted electron cun he scuvenged hy un udsorhed oxygen molecule
wlile tle plotogeneruted lole cun he used to oxidise un udsorhed solvent molecule. ln
un oxygenlree suspension, tle electron remuins on tle purticle hut tle lole cun still he
scuvenged. 1le excess electron cun he trupped hy u V centre wlicl results in tle
O

lormution ol u V centre. 1le V centres ure involved in tle visihle emission process und
O O
x •
removul ol tlese centres results in quencling ol tle visihle emission.
A detuiled unulysis ol tle luminescence quuntum elliciencies ol nunocrystulline
CdS und ZnO purticles is presented in clupter 6. For CdS tle inlluence ol tle
prepurution metlod und tle nuture ol tle surluce on tle quuntum elliciency is
investiguted. lt is slown tlut u prepurution using Nu S yields purticles witl u ligler
2
quuntum elliciency tlun wlen H S is used. For hotl prepurutions, tle quuntum
2
elliciency cun he increused hy couting tle purticles witl u cudmium lydroxide luyer or
hy cooling tle suspension to helow its lreezing point. 1lis cun he expluined hy u
removul ol tle nonrudiutive recomhinution centres ut tle surluce ol tle purticle. For
ZnO, tle inlluence ol purticle size on tle quuntum elliciency ol tle visihle emission is
studied. 1le quuntum elliciency decreuses lrom 2Oº to 12º us tle size ol tle purticles
increuses lrom 7 A to 1O A. 1lis result is in ugreement witl tle model tlut is presented
in clupter 4.
11
REFERENC ES.
[ 1] G. Jaeckel, Z. Tech. Phys. 6 ( 1926) 301.
[ 2] C.R. Berry, Phys. Rev. 153 ( 1967) 989.
[ 3] C.R. Berry, Phys. Rev. 161 ( 1967) 848.
[ 4] E.J. Meehan, J.K . Mi ller, J. Phys. Chem. 72 ( 1968) 1523.
[ 5] A.G. Stasenko, Sov. Phys. Sol. State 10( 1) ( 1968) 186.
[ 6] V.B. Sandomi rski i , Sov. Phys. JETP 16 ( 1963) 1630.
[ 7] J.R. Schri effer, “Mobi li ty i n i nversi on layers: Theory and experi ment”, i n Semiconductor Surface Physics,
edi ted by R.H. K i ngston, Uni versi ty of Pennsylvani a Press, Phi ladelphi a ( 1957) p.55-69.
[ 8] F.F. Fang, W.E. Howard, Phys. Rev. Lett. 16 ( 1966) 797.
[ 9] A.B. Fowler, F.F. Fang, W.E. Howard, P.J. Sti les, Phys. Rev. Lett. 16 ( 1966) 901.
[ 10] G. Dorda, “Surface quanti zati on i n semi conductors”, i n Festkörperprobleme (Advances in solid state
physics), Vol. XI I I , edi ted by H.J. Quei sser, Pergamon-Vi eweg, Braunschwei g ( 1973) , p. 215-239.
[ 11] J.R. Arthur, J. Appl. Phys. 39 ( 1968) 4032.
[ 12] A.Y. Cho, J. Vac. Sci. Technol. 8( 5) ( 1971) S31.
[ 13] R. Di ngle, W. Wi egmann, C.H. Henry, Phys. Rev. Lett. 33( 14) ( 1974) 827.
[ 14] U. Rothli sberger, W. Andreoni , M. Pari nello, Phys. Rev. Lett. 72 ( 1994) 665.
[ 15] C.C. Arnold, D.M. Neumark, J. Chem. Phys. 99 ( 1993) 3353.
[ 16] T. K obayashi , IEICE Trans. Fundamentals E-75A ( 1992) 38.
[ 17] W. Chen, Z. Li n, Z. Wang, L. Li n, Sol. St. Commun. 100( 2) ( 1996) 101.
[ 18] T. Türk, F. Sabi n, A. Vogler, Mat. Res. Bull. 27 ( 1992) 1003.
[ 19] D.I . Chepi c, Al.L. Efros, A.I . Eki mov, M.G. I vanov, V.A. K harchenko, I .A. K udri avtsev, T.V. Yazeva, J.
Lumin. 47 ( 1990) 113.
[ 20] S. Mochi zuki , K . Umezawa, J. Phys. : Condens. Matter 8 ( 1996) 7509.
[ 21] K . Mi sawa, H. Yao, T. Hayashi , T. K obayashi , Chem. Phys. Lett. 183( 1,2) ( 1991) 113.
[ 22] A.J. Bard, J. Phys. Chem. 86 ( 1982) 172.
[ 23] T. I noue, A. Fuji shi ma, S. K oni shi , K . Honda, Nature 277 ( 1979) 637.
[ 24] M. Halmann, Nature 275 ( 1978) 115.
[ 25] K . K alyanasundaram, E. Borgarello, M. Grätzel, Helv. Chim. Acta 64 ( 1981) 362.
[ 26] A. Henglei n, Ber. Bunsenges. Phys. Chem. 86 ( 1982) 301.
[ 27] A. Henglei n, J. Phys. Chem. 86 ( 1982) 2291.
[ 28] Z. Alfassi , D. Bahnemann, A. Henglei n, J. Phys. Chem. 86 ( 1982) 4656.
[ 29] R. Rossetti , S. Nakahara, L.E. Brus, J. Chem. Phys. 79( 2) ( 1983) 1086.
[ 30] R. Rossetti , J.L. Elli son, J.M. Gi bson, L.E. Brus, J. Chem. Phys. 80( 9) ( 1984) 4464.
[ 31] G.C. Papavassi li ou, J. Sol. State Chem. 40 ( 1981) 330.
[ 32] D.L. Dexter, R.S. K nox, Excitons, I ntersci ence Publi shers( John Wi ley & Sons) , New York ( 1965) .
[ 33] L.E. Brus, J. Chem. Phys. 80( 9) ( 1984) 4403.
[ 34] R. Di ngle, W. Wi egmann, C.H. Henry, Phys. Rev. Lett. 33( 14) ( 1974) 827.
[ 35] L. Spanhel, M. Haase, H. Weller, A. Henglei n, J. Am. Chem. Soc. 109 ( 1987) 5649.
[ 36] Y. Nosaka, J. Phys. Chem. 95 ( 1991) 5054.
[ 37] J.I . Pankove, Optical Processes in Semiconductors, Dover Publi cati ons, New York ( 1971) .
[ 38] A.P. Ali vi satos, J. Phys. Chem. 100 ( 1996) 13226.
Chapter 1 : I ntroducti on
12
[ 39] A.P. Ali vi satos, Science 271 ( 1996) 933.
[ 40] D.W. Bahnemann, Israel J. Chem. 33 ( 1993) 115.
[ 41] L. Brus, J. Phys. Chem. Solids 59( 4) ( 1998) 459.
[ 42] A. Henglei n, Top. Curr. Chem. 143 ( 1988) 113.
[ 43] A. Henglei n, Chem. Rev. 89 ( 1989) 1861.
[ 44] A. Henglei n, Ber. Bunsenges. Phys. Chem. 101 ( 1997) 1562.
[ 45] Y. Wang, N. Herron, J. Phys. Chem. 95 ( 1991) 525.
[ 46] H. Weller, Adv. Mater. 5( 2) ( 1993) 89.
[ 47] H. Weller, Angew. Chem. Int. Ed. Engl. 32 ( 1993) 41.
[ 48] H. Weller, A. Eychmüller, “Preparati on and characteri zati on of semi conductor nanoparti cles”, i n
Semiconductor Nanoclusters. Studi es i n Surface Sci ence and Catalysi s, vol. 103, edi ted by P.V. K amat
and D. Mei sel, Elsevi er Sci ence ( 1996) , p. 5-22.
[ 49] V.L. Colvi n, M.C. Schlamp, A.P. Ali vi satos, Nature 370 ( 1994) 354.
[ 50] B.O. Dabbousi , M.G. Bawendi , O. Oni tsuka, M.F. Rubner, Appl. Phys. Lett. 66 ( 1995) 1316.
[ 51] C.B. Murray, C.R. K agan, M.G. Bawendi , Science 270 ( 1995) 1335.
[ 52] S.A. Empedocles, D.J. Norri s, M.G. Bawendi , Phys. Rev. Lett. 77( 18) ( 1996) 3873.
[ 53] M. Ni rmal, B.O . Dabbousi , M.G. Bawendi , J.J. Mackli n, J.K . Trautman, T.D. Harri s, L.E. Brus, Nature 383
( 1996) 802.
18
14
15
C H A P ¡ £ k
5 | Z £ 5 £ L £ C ¡ | v £
P H 0 ¡ 0 £ ¡ C H | N 0
2
Adj usf | ng f he 5| ze cf Nonccr ysf o| | | ne
| | -v| 5em| ccnducf cr Por f | c| es
A 8 5 ¡ k A C ¡ .
1lis clupter presents un overview ol size selective plotoetcling
experiments perlormed on nunocrystulline semiconductor purticles in
suspension. Four dillerent llVl compounds (CdS, ZnS, lhS und ZnO,
ure prepured hy stundurd colloidclemicul syntletic metlods. For CdS,
tle initiul meun rudius ol 85 A cun he gruduully decreused to 7.5 A hy
size selective plotoetcling. 1le ud¦ustment ol tle purticle size cun he
controlled to u ligl degree hy vurying tle wuvelengtl ol tle liglt tlut is
used. lt is ulso slown tlut tle purticle size distrihution nurrows during
tle plotoetcling process lrom uhout 4Oº to 1O15º. For ZnS und ZnO,
only tle leusihility ol tle teclnique cun he demonstruted. 1lis is due to
tle luct tlut tlese compounds luve u lurge hundgup wlicl limits tle
runge lor plotoetcling experiments. 1le initiul rudius ol tle ZnO
purticles is upproximutely 8O85 A und cun he decreused to uhout 15 A,
uguin uccompunied hy u nurrowing ol tle purticle size distrihution. For
lhS, no concrete evidence is ohtuined lor tle leusihility ol size
selective plotoetcling.
Clupter ? . Size Selective llotoetcling
16
2 . I | N ¡ k 0 0 U C ¡ | 0 N .
As wus discussed in tle previous clupter, semiconductors exlihit quuntum size
ellects us u result ol sputiul conlinement ol clurge curriers wlen tle dimensions ol tle
muteriul hecome smuller tlun tle size ol tle exciton j18|. 1wo wellknown
munilestutions ol tlese quuntum size ellects ure tle slilt ol tle uhsorption onset to
ligler energies (due to un increuse ol tle energy ol tle lowest excited stute, und tle
trunsition lrom energy hunds to discrete energy levels. 1lese two plenomenu ure
muinly responsihle lor tle increused interest in quuntumconlined systems over tle lust
couple ol yeurs j?7|.
Severul teclniques tlut ullow tle prepurution ol semiconductor purticles witl
nunometer sizes ure hused on solgel metlods. 1ley luve heen reported lor tle
prepurution ol vurious semiconductor purticles, sucl us CdS j?,8|, ZnS j?,4|, lhS j?,11|,
ZnO j?,1?15| und 1iO j1?,16,17|.
?
1o study sizedependent properties it is importunt to luve monodisperse
suspensions ol dillerent purticle sizes. 1o uclieve tlis one cun husicully use two
dillerent upproucles. ln tle lirst ol tlese, tle prepurution metlod itsell determines tle
purticle size und size distrihution. 1lese metlods ure hused on prepurution inside u
limited spuce, sucl us u cuvity in u zeolite j18?O| or un inverse micelle in solution j?1
?8|. 1le second upproucl sturts witl u solgel prepurution lollowed hy treutments to
sepurute dillerent purticle sizes lrom tle initiul size distrihution. 1lese treutments
include exclusion clromutogruply j?4|, gel electroploresis j?5| und size selective
precipitution j?6?8|. ln generul, prepurution inside un inverse micelle yields u size
distrihution ol uhout 1O?5º j??| wlile u prepurution lollowed hy size selective
precipitution gives u dispersity ol 51Oº j?8,?D|.
Recently u novel teclnique lus heen descrihed tlut ullows tle purticle size to
he ud¦usted ulter prepurution hy plotoetcling tle semiconductor purticles to u size
wlicl is determined hy tle wuvelengtl ol tle liglt. 1le lirst ohservutions in tlis
direction were mude hy Henglein et ul. wlo studied tle plotounodic dissolution ol
semiconductor purticles und slowed tlut tley cun he plotoetcled viu un electroless
meclunism j5|. 1le sizeselectivity ol tle teclnique wus lirst demonstruted lor
nunocrystulline CdS purticles hy Vutsumoto et ul. j8O,81|. We luve slown tlut tlese
results cun he extended towurds mucl smuller purticle size j8?|. For otler sulplides
sucl us ZnS und lhS, no size selective plotoetcling experiments luve heen descrihed
helore. ln tlis clupter, un overview is presented ol tle results tlut ure ohtuined lrom
size selective plotoetcling experiments tlut were perlormed on colloidul suspensions
ol nunocrystulline llVl semiconductor purticles. lt is slown tlut lor CdS tle purticle
17
rudius cun he gruduully decreused to 7.5 A. lnlormution uhout tle purticle size is
ohtuined hy perlorming 1LV und XRD meusurements und hy unulyzing tle uhsorption
spectru. From tlis unulysis it is ulso possihle to give un estimution ol tle purticle size
distrihution. We luve ulso upplied tle teclnique ol size selective plotoetcling to ZnS,
lhS und ZnO.
2 . 2 ¡ H £ 0 k Y .
Wlen u semiconductor is illuminuted witl liglt tlut lus un energy wlicl
exceeds tlut ol tle hundgup, uhsorption cun tuke pluce und electronlole puirs ure
creuted. 1le electron cun suhsequently recomhine witl tle lole, eitler rudiutively or
nonrudiutively. Rudiutive recomhinution cun give rise to u relutively slurp emission
hund centred ut upproximutely tle hundgup energy (exciton emission, or u relutively
hroud emission hund ut lower energies wlen (deep, trups ure involved in tle
recomhinution process (trup emission,. ln tle cuse ol nunocrystulline purticles ol CdS
und ZnO hotl emission hunds cun he ohserved ut room temperuture. Due to u lurge
surlucetovolume rutio, nonrudiutive recomhinution viu surluce trups is tle
predominunt route.
Wlen tle recomhinution ol electronlole puirs is prolihited hy removing tle
electron (hy un interluciul process,, tle lole cun he used to dissolve tle
semiconductor. 1lis wellknown process ol electroless plotoetcling j8885|, in
comhinution witl tle slilt ol tle uhsorption onset to ligler energies upon u decreuse ol
tle purticle size, provides tle husis lor tle teclnique ol size selective plotoetcling.
Removul ol tle electron lrom u semiconductor purticle cun he uclieved hy
using electron scuvengers sucl us oxygen molecules or metlylviologen ions (VV , tlut

ure udsorhed on tle surluce ol tle purticle. Vetlylviologen sluttles tle plotogeneruted
electrons lrom tle semiconductor purticle to tle udsorhed oxygen molecules. First, tle
electrons ure elliciently scuvenged lrom tle conduction hund, resulting in tle lormution
ol VV rudiculs, lollowed hy tle reduction ol oxygen und regenerution ol VV . As u
÷ ?÷ •
result, tle use ol metlylviologen cun increuse tle rute ol tle plotoetcling process
j81,8?|.
lt wus expluined in tle previous clupter tlut one cun tuke us u tlreslold lor tle
occurrence ol quuntum size ellects tle vulue lor tle ßolr rudius ol tle exciton in tle
mucrocrystulline muteriul.
ε

∗ ∗
 
= ⋅ ⋅ +
 
 
exc
8
1 1
0.S29
A D
=

2 2
ç
2
0 0
h 1 1 1.8
8 1
A D
A
- -
4 4 πε ε
∗ ∗

 
= + ⋅ + −
 
 
ν −
+ → +
2÷ 2
2 1
h
CoS(s) 2C (aos) Co (aq) SC (aq)

Clupter ? . Size Selective llotoetcling
18
(?.1,
(?.?,
1le vulue lor tle ßolr rudius ol tle exciton (= , in A, is given hy .
ß
exc
ln equution (?.1,, ε is tle ligllrequency relutive dielectric constunt wlile ¯ und ¯
∞ e l
ure tle ellective musses ol tle electron und lole respectively (hotl in units ,.
O
Wlen tle rudius ol u semiconductor purticle lus decreused helow = , tle
ß
exc
energy lor tle lowest electronic trunsition cun he culculuted using tle lollowing
equution, hused on tle quuntum meclunicul model ol u purticle in u splericul hox witl
inlinite potentiul energy hurriers j1|.
ln equution (?.?,, - is tle hundgup ol tle mucrocrystulline semiconductor und
g
4 is tle rudius ol tle semiconductor purticle. 1le 4 term is u conlinement energy term
−?
wlile tle 4 term rellects tle Coulomh interuction hetween electrons und loles. ln
−1
ligure ?.1A, tle dependence ol - on tle purticle size us given hy equution (?.?, is slown
sclemuticully. Wlen semiconductor purticles witl u size distrihution us slown in ligure
?.1ß ure illuminuted witl liglt ol energy lν , only tle purticles tlut ure 5O A or higger
1
will he uhle to uhsorh tle liglt. 1lese purticles cun he plotoetcled wlicl is
uccompunied hy un increuse ol tleir hundgup. Gruduully tle size ol tlese purticles will
he sucl tlut tle hundgup energy is lurger tlun tle ploton energy. At tlis purticle size
determined hy tle wuvelengtl ol tle liglt tle plotoetcling will stop. 1le initiul size
distrihution slould now luve clunged to tlut slown in ligure ?.1C. ln tle sume wuy one
cun decreuse tle purticle size lurtler hy using liglt ol u ligler energy (e.g. lν , see
?
ligure ?.1D,.
1o demonstrute tle teclnique ol size selective plotoetcling us slown in ligure
?.1, nunocrystulline CdS purticles cun he used us u model system. 1lis system lus heen
studied extensively in tle literuture j?,1O,8688|. 1le overull plotodissolution reuction is
very well known jD| .
40 50 60

40 50 60
40 50 60
B.
D.
C.
A.

40 50 60
E
n
e
r
g
y

(
e
V
)
Radius (Å)
Radius (Å)
N
u
m
b
e
r

o
f

p
a
r
t
i
c
l
e
s

1

2

1

2
E
g
Figure 2.1 : Schematic picture of the concept of size selective photoetching. A :
bandgap energy as a function of particle radius for an arbitrary semiconductor. B :
initial size distribution. C : size distribution after photoetching with light of energy hν
1
.
D : size distribution after photoetching with light of energy hν
2
.
λ
θ
=

8
0.9
cos
I
*
1D
(?.8,
llluminution ol CdS purticles witl plotons ol un energy ligler tlun tle
hundgup leuds to tle lormution ol electronlole puirs. Adsorhed oxygen cun he
reduced hy electrons lrom tle conduction hund wlile tle loles in tle vulence hund
leud to u dissolution ol CdS. 1le elliciency ol tlis process is low since tle mu¦ority ol
tle electronlole puirs will recomhine, eitler rudiutively or nonrudiutively.
1o determine tle purticle size ulter size selective plotoetcling, teclniques sucl
us trunsmission electron microscopy (1LV, und Xruy dillruction (XRD, cun he used.
1le widtl ol u dillruction peuk increuses us tle size ol tle crystullites decreuses. 1le
Sclerrerlormulu cun he used to determine tle size (I, ol tle crystullite, wlicl is equul
to tle diumeter ol tle semiconductor purticle .
Clupter ? . Size Selective llotoetcling
?O
ln equution (?.8, λ is tle wuvelengtl ol tle Xruys (lor CuKα tlis is 1.541 A,, *
1
is tle FWHV ol tle dillruction peuk (in rudiuns, und θ is tle muximum ol tle
ß
dillruction peuk (in rudiuns,. Apurt lrom tlese experimentul teclniques, empiricul
relutionslips hetween tle meun purticle rudius und tle position ol tle lirst uhsorption
muximum cun ulso he used. Sucl relutionslips luve heen estuhlisled in tle literuture
lor CdS j1O| und ZnO j14| und tley will he used lere insteud ol equution (?.?, ultlougl
tlere is u reusonuhle ugreement hetween tle vulues ohtuined hy hotl metlods.
Finully, to meusure tle nurrowing ol tle purticle size distrihution ulter size
selective plotoetcling tle widtl ol tle inlomogeneously hroudened hunds in tle
uhsorption und emission spectru cun he unulyzed.
2 . 3 £ X P £ k | M £ N ¡ A L M £ ¡ H 0 0 5 .
2. 3 . I 5omp| e pr epor of | cn.
Cd5. 1o ohtuin u colloidul suspension ol nunocrystulline CdS purticles, ?O ml
ol un uqueous solution ol O.O5 V sodium polyplosplute were udded to 4O ml ol un
uqueous solution ol O.O1 V Cd(ClO , 6H O. Alter diluting tlis solution to 1OO ml, 7.? ml
4 ? ?
.
H S (Aldricl lecture hottle, DD.5÷ º, were in¦ected using un uirtiglt syringe. Vigorously
?
sluking tle sumple lor severul minutes ut room temperuture resulted in u trunspurent
suspension witl u yellow colour. According to tle literuture, tlis prepurution ol un
uqueous CdS suspension slould yield purticles witl u meun rudius ol uhout 85 A
j86,87|.
Zn5. A colloidul suspension ol nunocrystulline ZnS purticles wus ohtuined hy
diluting 5 ml ol un uqueous solution ol O.O1 V Zn(ClO , 6H O to 1OO ml helore udding
4 ? ?
.
1O ml ol un uqueous solution ol O.O5 V sodium polyplosplute. Alter in¦ection ol 1 ml
H S, tle sumple wus vigorously sluken lor severul minutes ut room temperuture und
?
ullowed to uge lor u duy j4|. 1le resulting colloidul suspension wus trunspurent und
colourless.
Pb5. For tlis suspension u mixture ol ? ml ol un uqueous solution ol O.OO5 V
lh(ClO , 8H O und ? ml ol un uqueous solution ol O.5 V polyvinyl ulcolol wus diluted
4 ? ?
.
to 1OO ml helore in¦ecting ?15 µl H S. 1lis resulted in u trunspurent colloidul suspension
?
witl u durk red colour. According to tle literuture, tle lhS purticles luve u rodlike
slupe witl u lengtl ol 18O A und u diumeter ol 15 to ?5 A j11|.
Zn0. A solution ol O.O6 g (?.7 1O mol, Zn(OAc, ?H O in 8O ml ?propunol
.
−4
? ?
.
wus mude ut 5O °C. Alter diluting tle solution to ?8O ml witl ?propunol it wus cooled to
?1
O °C. At tlis temperuture ?O ml ol u solution ol O.O8? g (?.O 1O mol, NuOH in 1OO ml ?
.
−8
propunol were udded witlin 1 minute wlile stirring. Alter ugeing in u wuter hutl ut
65 °C lor ? lours u trunspurent, colourless suspension ol ZnO purticles wus ohtuined.
According to tle literuture j18|, tle ZnO purticles ure ulmost splericul witl u meun
rudius ol ?5 A und u relutively nurrow purticle size distrihution (∆R=5 A,.
Since ulcolols cun uct us lole scuvengers, plotoetcling ol ZnO could not he
perlormed in tle solvent in wlicl ZnO wus prepured. 1lerelore tle ZnO purticles were
trunslerred to wuter helore plotoetcling. 1lis wus done hy lirst udsorhing tle ZnO
purticles on u silicu powder (Aerosil Ox5O Degussu,. Alter centriluging tle suspension
und wusling tle precipitute, tle ZnO purticles still udsorhed on silicu were
redispersed in wuter.
2. 3 . 2 Phcf cef ch| ng.
1le size selective plotoetcling experiments were perlormed witl u 45O W
xenon lump (Uslio UXL 45OSl,. 1le wuvelengtl ol tle liglt wus tuned hy using u series
ol cutoll lilters. (1, ?.48 eV (Sclott A4DO,, (?, ?.78 eV (Sclott A44O,, (8, ?.D4 eV (Sclott
GG4?O,, (4, 8.O6 eV (Sclott WG8D5,, (5, 8.81 eV (Sclott WG875,, (6, 8.44 eV (Sclott
WG86O,, (7, 8.4D eV (Sclott A85O1,, (8, 8.54 eV (Sclott A88O, und (D, 4.O8 eV (Sclott
A?8O,. A cutoll lilter uhsorhs plotons ol energies ligler tlun tle cutoll energy wlile
plotons witl u lower energy ure trunsmitted. 1le cutoll energy is delined us tle energy
ut wlicl tle trunsmission is 5Oº ol tle muximum trunsmission.
1le suspensions were illuminuted in u quurtz contuiner under continuous
stirring und huhhling witl oxygen. 1o uvoid interlerence lrom extru uhsorption hunds,
metlylviologen wus not used in tle experiments. 1le uqueous suspension ol ZnS
purticles wus illuminuted witlout using u cutoll lilter. 1le suspension ol ZnO purticles
udsorhed on silicu wus lirst diluted 1.1O witl wuter helore it wus illuminuted wlile in u
gluss contuiner. ln tlis wuy, tle gluss ucts us u cutoll lilter ut uhout 4 eV.
2. 3 . 3 0pf | co| chor ocf er | sof | cn.
Ahsorption meusurements were perlormed on u lerkin Llmer Lumhdu 16
UV/VlS spectroplotometer. 1le uhsorhunce (), wus meusured us tle loguritlm ol tle
trunsmission (6=11 , witl wuter us u relerence. Luminescence meusurements were
O
perlormed on u SlLX Fluorolog spectroplotometer model F?OO? equipped witl two
douhlegruting O.?? m SlLX 168O monoclromutors und u 45O W xenon lump us tle
excitution source.
Clupter ? . Size Selective llotoetcling
??
2. 3 . 4 ¡ r onsm| ss| cn £ | ecf r cn M| cr csccpy.
1LV meusurements were perlormed on u llilips CV8O electron microscope
operuting ut 8OO kV. Only nonetcled sumples were used lor tlese meusurements. 1o
perlorm 1LV meusurements on sumples ulter plotoetcling it is necessury to lirst
remove tle etcl products (e.g. Cd und SO ions, lrom tle suspension (e.g. hy
?÷ ?
4

diulysis,.
2. 3 . 5 X-r oy Pcwder 0| f f r ocf | cn.
XRD meusurements were done using CuKα rudiution on u llilips lW17?D X
1
Ruy Generutor equipped witl u lW171O Dillructometer Control.
2 . 4 k £ 5 U L ¡ 5 A N 0 0 | 5 C U 5 5 | 0 N .
2. 4. I Cd5.
Figure ?.? slows tle results ol 1LV meusurements on CdS purticles. lt is cleur
tlut tle purticles ure crystulline und luttice plunes cun he ohserved witl interplunur
distunces ol uhout 8.4 A. 1lese meusurements ulso slow tlut tle purticle size
distrihution is relutively hroud. Due to cluster lormution, it is dillicult to determine u
meun purticle rudius lrom tle 1LV meusurements hut unulysis slows muny purticles
witl rudii hetween 8O und 4O A.
XRD spectru slow dillruction peuks tlut ure hroudened due to smull crystul
sizes. As u result ol tlis hroudening it is dillicult to determine tle crystul structure ol tle
CdS purticles (wurtzite or zinchlende,. 1le interplunur distunce ol 8.4 A occurs in hotl
modilicutions. For mucrocrystulline CdS tle most common modilicution is wurtzite
j8D|. For nunocrystulline CdS purticles hotl modilicutions luve heen reported ultlougl
tle zinchlende structure is generully ohserved wlen CdS purticles ure prepured in
solution j?7,88|. For tle strongest dillruction peuk, using equution (?.8, witl θ =O.?8
ß
rud und *=O.O?, u meun purticle rudius ol 85 A is ohtuined wlicl is in good ugreement
witl tle vulues ohtuined lrom 1LV meusurements und tle vulue reported in tle
literuture j87|.
For tle wurtzite structure (¯=O.??, ¯=O.7O und ε =5.8 j8D|,, tle ßolr rudius ol
e l ∞
tle exciton in mucrocrystulline CdS culculuted lrom equution (?.1, is 17 A wlile lor tle
Figure 2.2 : TEM-micrographs of CdS particles from an aqueous suspension before
illumination.
A.
B. C.
500 Å
50 Å
50 Å
?8
zinchlende structure (¯=O.14, ¯=O.51 und ε =5.8 j8D|, u vulue ol ?6 A is ohtuined.
e l ∞
1lis meuns tlut tle initiul rudius ol our purticles is close to tlut ol tle exciton in
mucrocrystulline CdS.
1lis is in ugreement witl results ohtuined lrom uhsorption und luminescence
meusurements. Figure ?.8 slows tle uhsorption und emission spectrum ol un uqueous
suspension ol nunocrystulline CdS purticles helore illuminution. 1le onset ol uhsorption
determined lrom extrupolution ol tle steep purt ol tle spectrum is ut uhout ?.5 eV,
similur to tle vulue lor tle hundgup energy ol mucrocrystulline CdS. 1le uhsorption
spectrum is ulso structureless, indicuting tle presence ol continuous energy hunds. 1le
emission spectrum slows un exciton emission hund centred ut ?.45 eV.
Figure 2.3 : Emission spectrum of an aqueous suspension of nanocrystalline CdS
particles upon excitation with 3.5 eV before photoetching. Φ
E
denotes the photon flux
per constant energy interval. The inset shows the exciton emission band together with
the absorption spectrum of the suspension.
Energy (eV)
1.0 2.0 3.0
0
50
100
2.0 3.0
Φ
E

(
a
r
b
.

u
n
i
t
s
)
Figure 2.4 : Absorption and emission spectra of nanocrystalline CdS particles in
aqueous suspension. Spectra (a) were recorded before illumination, spectra (b)-(f)
after illumination for two hours through filters with cut-off energies of 2.48 eV, 2.73 eV,
2.94 eV, 3.06 eV and 3.31 eV respectively. The intensities of the exciton emission
maxima and the first absorption maxima are set to unity. The emission spectra were
recorded upon excitation with 4.1 eV radiation. Φ
E
denotes the photon flux per
constant energy interval.
A
b
s
o
r
b
a
n
c
e
Energy (eV)
2.0 3.0 4.0
(f)
(e)
(d)
(c)
(b)
(a)
Φ
E
Clupter ? . Size Selective llotoetcling
?4
?5
All tle spectru slown in ligure ?.4 concern tle sume suspension und eucl ol
tlem is tuken ulter illuminution lor two lours using lilters witl increusing cutoll
energies. As tle uhsorption spectru did not clunge upon prolonged illuminution, tle
plotoetcling process wus completed witlin tlis time spun. lreviously, it wus reported
tlut tle use ol metlylviologen enlunces tle rute ol tle plotoetcling process j81,8?|.
1le spectru slow u gruduul slilt ol tle uhsorption onset und u concomitunt
development ol structure. 1lese ure cleur indicutions ol u decreuse in size ol quuntum
conlined semiconductor purticles. 1le ohservution ol structure is cluructeristic lor u
nurrow size distrihution.
Tablo ?.1 . Results ol size selective plotoetcling ol un uqueous suspension ol
nunocrystullineCdS purticles, using lilters witl dillerent cutoll energies. 1le lirst row
contuins tle dutu lor tle CdS suspension helore illuminution.
lilter emission uhsorption uhsorption meun
cutoll energy muximum onset muximum purticle rudius
u
(eV, (eV, (eV, (eV, (A,
h
?.45 ?.47 85
c
?.48 ?.5O ?.57
?.78 ?.74 ?.68 ?.86 ?4
?.D4 ?.D1 ?.87 8.O? 18
8.O6 8.O? 8.O1 8.1D 14
8.81 8.18 8.88 1?
8.44 8.86 8.64 1O
8.4D 8.41 8.7O D.5
8.54 8.45 8.78 D.O
4.O8 8.D8 4.?O 7.5
d
Delined us tle energy wlere tle trunsmission is lull tle muximum vulue.
u
Determined lrom extrupoluting tle steep purt ol tle uhsorption spectrum.
h
Determined lrom 1LV und XRD meusurements.
c
losition ol tle lowest excitution muximum.
d
1uhle ?.1 gives tle vulues lor tle cutoll energies ol tle lilters tlut were used,
togetler witl tle energetic positions ol tle onset ol uhsorption, lirst uhsorption
muximum, und tle position ol tle exciton emission hund. 1le column on tle riglt
contuins tle purticle size tlut wus estimuted lrom tle empiricul relutionslip hetween
tle lirst uhsorption muximum und meun purticle rudius j1O|. 1le uhsorption und
emission dutu lrom tuhle ?.1 slow good ugreement witl tle cutoll energies ol tle
Clupter ? . Size Selective llotoetcling
?6
lilters. 1lis indicutes tlut tle purticle size ohtuined ulter size selective plotoetcling cun
he controlled hy tuning tle wuvelengtl ol tle liglt.
1o illustrute tlut size selective plotoetcling cun he used to nurrow tle size
distrihution ol u suspension ol nunocrystulline semiconductor purticles, uhsorption
meusurements were perlormed on u mixture ol two CdS suspensions. 1lese suspensions
were ohtuined hy plotoetcling witl liglt ol ?.D4 eV und 8.81 eV respectively. Figure ?.5
slows tle uhsorption spectru ol tle sepurute suspensions (1 und ?, ulter u 1.1 dilution
witl wuter. Figure ?.5 ulso slows un uhsorption spectrum ol u 1.1 mixture ol tle
undiluted suspensions 1 und ? (u,. 1le mixture (u, wus uguin illuminuted tlrougl u
lilter witl u cutoll energy ol 8.81 eV, resulting in spectrum (h,. ßecuuse spectrum (h,
resemhles spectrum (?,, it is cleur tlut tle purticle size distrihution nurrows upon
plotoetcling. Only tle higger purticles udded to suspension (?, lrom suspension (1,
luve heen plotoetcled. 1le totul ureu ol tle lirst uhsorption peuk ol spectrum (h, is u
luctor 1.85 ligler tlun tlut ol tle lirst uhsorption peuk ol spectrum (?,. 1lis meuns tlut
tle totul numher ol purticles lus decreused sligltly upon plotoetcling CdS purticles
lrom u rudius ol 18 A to 1? A. 1lese results ure not in ugreement witl tlose ohtuined hy
Vutsumoto et ul. j81|. ßy unulyzing tle umount ol sulplute ions produced during tle
plotoetcling process tley luve concluded tlut tle numher ol purticles decreused to
uhout 4Oº ol tle initiul numher ol purticles over tle sume size runge us in our
experiment.
At lirst siglt one miglt tlink tlut u hroudening ol tle lirst uhsorption hund und
tle exciton emission hund indicutes u hroudening ol tle purticle size distrihution ulter
plotoetcling. ln ligure ?.6, it cun he seen tlut tle lirst uhsorption hund hroudens upon
plotoetcling wlile tle sume ellect is ulso visihle in tle emission spectru. Figure ?.7A
contuins two exciton emission hunds . ulter plotoetcling witl ?.D4 eV (u, und 4.O8 eV
(h,. lt is cleur tlut tle second emission hund, originuting lrom mucl smuller purticles, is
hrouder. 1le hroudening ol hotl uhsorption und exciton emission hunds rellects tle
increusing dependence ol tle hundgup energy on tle purticle size. 1lis meuns tlut
even purticles witl u relutively nurrow size distrihution cun exlihit hroud uhsorption
und emission hunds once tley ure in tle strong conlinement regime. 1le principle ol
hroudening ol tle exciton emission hund us u result ol quuntum conlinement is
illustruted in ligure ?.7ßD. 1le dependence ol tle lowest electronic trunsition in u
semiconductor purticle on tle purticle size is slown in ligure ?.7ß. ln ligure ?.7C, tlree
possihle purticle size distrihutions ure slown dillering in meun purticle size und/or
distrihution widtl. 1le exciton emission hunds, us slown in ligure ?.7D, ure constructed
hy plotting, lor eucl purticle size, tle numher ol purticles versus tle energy lor tle
lowest electronic trunsition (tle intensity is tuken proportionul to tle numher ol
Figure 2.5 : Absorption spectra of aqueous suspensions of nanocrystalline CdS
particles after illumination through filters with cut-off energies 2.94 eV (1) and 3.31 eV
(2) respectively. A 1:1 mixture of (1) and (2), shown as (a), was again illuminated
through a filter with cut-off energy 3.31 eV to yield (b).
2.0 3.0 4.0
0.0
0.2
0.4
(b)
(a)
(2)
(1)
Energy (eV)
A
b
s
o
r
b
a
n
c
e
Figure 2.6 : Energetic positions of the exciton emission maximum (solid circles) and
the first absorption maximum (open circles) versus the absorption onset energy of
nanocrystalline CdS particles in aqueous suspension.
2.0 3.0 4.0
2.0
3.0
4.0
E
n
e
r
g
y

(
e
V
)
Absorption onset (eV)
?7
purticles,. ln tlis wuy, only inlomogeneous line hroudening is considered. lt is cleur
tlut tle widtls ol tle emission hunds originuting lrom tle smullest purticles, (h, und
(c,, ure mucl lurger tlun in tle cuse ol tle higger purticles (u,, even wlen tle purticle
size distrihution lus nurrowed.
Figure 2.7 : A : Emission measurements on aqueous suspensions of nanocrystalline
CdS particles showing the exciton emission band. Spectrum (a) was taken after
photoetching at 2.94 eV (excitation with 4.1 eV) and spectrum (b) after photoetching
at 4.08 eV (excitation with 5.2 eV). Φ
E
denotes the photon flux per constant energy
interval. B-D : Illustration of the broadening of excitonic absorption and/or emission
bands as a consequence of semiconductor particles becoming more quantum
confined. B : The variation of the bandgap energy with particle radius. C : Three
possible particle size distributions, differing in mean particle size and monodispersity.
D : The corresponding shapes of the excitonic emission bands. The intensity is taken to
be proportional to the number of particles. The homogeneous line broadening is
considered to be negligible in comparison to the inhomogeneous line broadening.
2.0 3.0 4.0 5.0
0.0
0.5
1.0
(b) (a)
0
2
4
6
8
(c)
(b) (a)
2.0 4.0 6.0 8.0
(c)
(b)
(a)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
E
n
e
r
g
y

(
e
V
)
Energy (eV)
Energy (eV)
I
n
t
e
n
s
i
t
y
Radius
Radius
N
u
m
b
e
r

o
f

p
a
r
t
i
c
l
e
s
A. B.
C. D.
Clupter ? . Size Selective llotoetcling
?8
A quuntitutive unulysis ol tle clunge in purticle size distrihution cun he
perlormed hy considering tle widtl ol tle lirst uhsorption hund. lt is ussumed tlut tlis
hund is inlomogeneously hroudened due to u distrihution in purticle size. Lxperimentul
results ohtuined hy Vossmeyer et ul. j1O| cun he used to determine hotl tle meun
purticle rudius (<4>, lrom tle uhsorption muximum us well us tle dispersion in purticle
size (∆4, using tle energy ut wlicl tle uhsorption is lull tle muximum vulue on tle
lowenergy side ol tle lirst uhsorption hund. ln tuhle ?.? tle vulues lor <4> und ∆4 ure
slown. lt is cleur tlut not only <4> hut ulso ∆4 decreuses upon plotoetcling. As u result
ol size selective plotoetcling, tle dispersion in purticle size decreused lrom 4Oº to 1O
15º, wlicl is compuruhle to tlut ohtuined hy otler teclniques sucl us inverse micelle
prepurution und size selective precipitution j??,?8,?D|.
?D
Tablo ?.? . Veun purticle rudii (<4>, und uccompunying ∆4 (delined us lull tle
FWHV ol u guussiun purticle size distrihution, determined lrom tle energetic
positions ol tle uhsorption muximu und tle lullmuximu using j1O|.
uhsorption
lilter cutoll energy muximum lullmuximum <4>
(eV, (eV, (eV, (A,
∆4
(A,
?.78 ?.86 ?.75 ?4 D
?.D4 8.O? ?.D8 18 8
8.O6 8.1D 8.1O 14 1.5
8.81 8.88 8.?8 1? 1.5
4.O8 4.?O 8.D? 7.5 1
u u
Determined lrom tle lowest excitution muximum.
u
Figure ?.6 ulso slows tle increusing energy sepurution hetween tle exciton
emission muximum und tle lirst uhsorption muximum, rellecting un increuse in exciton
hinding energy witl decreusing purticle size. 1lis sizedependence ol tle exciton
hinding energy lus heen reported helore und wus uttrihuted to tle enlunced sputiul
overlup hetween tle electron und lole wuve lunctions j6|.
As u result ol tlis enlunced overlup, un increuse ol tle oscillutor strengtl witl
decreusing purticle size is ulso expected j6| und lus heen ohserved j1O|. Since tle
oscillutor strengtl is lineurly proportionul to tle uhsorption coellicient, uhsorption
spectru cun he used to study tle sizedependence ol tle oscillutor strengtl. ln our cuse,
ull uhsorption spectru ure recorded lor tle sume suspension ol nunocrystulline CdS
purticles und only tle purticle size is clunged hy plotoetcling. ln tlis wuy tle clunge
in oscillutor strengtl due to u clunge in purticle size cun he determined uccurutely. ln
ligure ?.8, some ol tle uhsorption spectru lrom ligure ?.4 ure plotted witlout sculing tle
uhsorption signul. lt is cleur tlut tle totul uhsorption signul decreuses upon
plotoetcling us u result ol u decreuse in tle umount ol uhsorhing muteriul. 1le lirst
uhsorption muximum cun he litted to u guussiun curve und tle ureu ol tlis curve,
normulised lor tle purticle volume, gives un indicution ol tle uhsorption coellicient ol
tle semiconductor purticle per CdS unit. As tle rudius ol tle CdS purticle decreuses
lrom ?4 A to 11 A tle uhsorption coellicient (und tlerelore tle oscillutor strengtl,
increuses witl u luctor ol 8.5. 1lis is in good ugreement witl tle experimentul results
ohtuined hy Vossmeyer et ul. j1O| hut not witl tle tleoreticul 4 dependence j6|.
8
Figure 2.8 : Absorption spectra of an aqueous suspension of CdS particles after
photoetching at 2.73 eV (a), 2.94 eV (b), 3.06 eV (c) and 3.31 eV (d). The dashed
curves are gaussian fits to the first absorption maxima.
2.0 3.0 4.0
0.0
0.1
0.2
0.3
0.4
0.5
(d)
(c)
(b)
(a)
Energy (eV)
A
b
s
o
r
b
a
n
c
e
Clupter ? . Size Selective llotoetcling
8O
2. 4. 2 Zn5.
Only XRD meusurements were perlormed to cluructerise tle ZnS purticles. 1le
dillruction puttern indicutes tlut tle ZnS purticles luve tle zinchlende modilicution,
wlicl is in ugreement witl tle luct tlut tle wurtzite structure is tle ligltemperuture
modilicution ol ZnS j8D|.
From tle strongest dillruction peuk ut θ =O.?5 rud (*=O.O8 rud, u meun rudius ol
ß
?5 A is determined lor tle ZnS purticles. For mucrocrystulline ZnS (zinchlende,, using
¯=O.84, ¯=O.5 (hotl in units , und ε =5.4 j8D|, equution (?.1, gives u ßolr rudius ol
e l O ∞
tle exciton ol uhout 15 A.
Figure ?.DA contuins tle emission und uhsorption spectrum ol un uqueous
suspension ol nunocrystulline ZnS purticles helore illluminution. 1le meun rudius ol tle
ZnS purticles is lurger tlun tlut ol tle exciton und tlerelore no strong quuntum size
ellects ure expected. lndeed, tle uhsorption spectrum slows no structure. However,
extrupolution ol tle steep purt ol tle uhsorption spectrum yields u vulue ol 8.87 eV
wlile mucrocrystulline ZnS (zinchlende, lus u hundgup ol 8.7O eV j8D|. Lxtrupolution
ol tle steep purt ol tle uhsorption curve gives un upproximution ol tle hundgup, not un
uccurute vulue.
1le emission spectrum slown in ligure ?.DA contuins u hroud trup emission
hund centred ut ?.DO eV und no detectuhle exciton emission hund ut tle onset ol
Figure 2.9 : A : Emission spectrum of an aqueous suspension of nanocrystalline ZnS
particles upon excitation with 5 eV. Φ
E
denotes the photon flux per constant energy
interval. The inset shows a part of the emission spectrum together with the absorption
spectrum of the suspension. B : Absorption spectra of an aqueous suspension of
nanocrystalline ZnS particles before illumination (a) and after illumination for 30 min
(b). Absorption spectrum (b) was scaled with respect to spectrum (a).
0.0
0.2
0.4
0.6
B.
(b) (a)
A
b
s
o
r
b
a
n
c
e
2.0 3.0 4.0 5.0
4.0 5.0
0
50
100
A.

2.0 3.0 4.0 5.0
Φ
E

(
a
r
b
.

u
n
i
t
s
)
Energy (eV)
Energy (eV)
81
uhsorption. An exciton emission hund uppeured us u weuk sloulder on tle ligl energy
side ol tle trup emission hund only ulter tle surluce ol tle ZnS purticles wus pussivuted
witl u lydroxide luyer. Generully, tlese surlucepussivuted semiconductor purticles
cunnot he plotoetcled j8?|. 1lis meuns tlut in tle cuse ol (non surlucepussivuted,
ZnS only uhsorption meusurements cun he used to study plotoetcling processes. As
ZnS is u widehundgup semiconductor, plotoetcling experiments luve to he curried
out witl liglt ol u relutively ligl energy. A similur plotoetcling series us lor CdS could
not he mude hecuuse lew lilters witl ligl cutoll energies were uvuiluhle. From ligure
?.Dß it is cleur tlut hy using tle xenon lump witlout u cutoll lilter, nunocrystulline ZnS
purticles cun he plotoetcled. Upon illuminution lor 8O min tle onset ol uhsorption lus
slilted to 4.O4 eV und tle spectrum slows some structure. At uhout 4.5 eV u muximum
ol tle lirst uhsorption hund is visihle. 1lese results cleurly indicute tlut upon
plotoetcling tle ZnS purticles luve entered tle quuntum conlinement regime.
Clupter ? . Size Selective llotoetcling
8?
2. 4. 3 Pb5.
An interesting system lor size selective plotoetcling experiments is lhS. 1le
hundgup energy lor mucrocrystulline lhS is relutively low (O.41 eV, wlicl meuns tlut
plotoetcling experiments could he perlormed over u wide energy runge, covering tle
lR to tle UV spectrul region. Alter prepurution, uhsorption und luminescence
meusurements were perlormed to cluructerise tle lhS suspensions. ln ligure ?.1O, curve
(u, is tle uhsorption spectrum ol un uqueous suspension ol nunocrystulline lhS
purticles. 1le uhsorption hund ut uhout 6 eV is due to tle presence ol excess lh ions

in solution. As tle uhsorption sturts ut energies well uhove tlut ol tle hundgup ol
mucrocrystulline lhS und tle spectrum is structured, tle lhS purticles ure cleurly
quuntumconlined. ßecuuse ol tle low ellective musses ol tle clurge curriers in lhS
(¯, ¯ = O.OD j8D|,, equution (?.?, yields u vulue lor tle ßolr rudius ol tle exciton
e l O
ol more tlun ?OO A. 1le tlree uhsorption muximu ut ?.1 eV, 8.8 eV und 4.8 eV prohuhly
result lrom trunsitions to dillerent discrete energy levels j11|. Similur to ZnS, emission
meusurements on nunocrystulline lhS purticles yielded only u hroud trup emission
hund.
From spectrum (h, in ligure ?.1O it is cleur tlut upon illuminution tlrougl u
lilter witl u cutoll energy ol ?.?5 eV tle totul intensity ol tle uhsorption signul decreuses
wlile ut tle sume time tle structure hecomes less pronounced. Altlougl no distinct
slilt in tle uhsorption onset is visihle it seems tlut tle positions ol tle tlree muximu
luve slilted sligltly towurds ligler energies. 1le loss ol structure could he due to u
hroudening ol tle uhsorption hunds. A possihle reuson lor tlis hroudening lus ulreudy
heen uddressed helore lor CdS. However, wlen illuminution is continued lor longer
periods or witl ligler energies, tle totul uhsorption intensity goes down even lurtler.
1wo explunutions could uccount lor tle uhove mentioned ohservutions. First,
nunocrystulline lhS purticles could he very dillicult to plotoetcl due to tle insoluhility
ol lhSO in wuter. llluminution ol nunocrystulline lhS purticles would tlen prohuhly
4
result in tle lormution ol u pussivuting sulplute luyer ut tle surluce ol tle purticles
wlicl consequently inlihits lurtler etcling. However, tlis does not uccount lor tle
ohserved decreuse in uhsorhunce upon prolonged illuminution. A second explunution
is hused on tle slupe ol tle lhS purticles. According to tle literuture j11|, tle purticles
luve u rodlike slupe. 1le lengtl ol tlese rods is uhout 18O A, wlicl is luirly close to tle
vulue lor tle ßolr rudius ol tle exciton in mucrocrystulline lhS (≈?OO A,. Wlen
illuminution ol tle lhS purticles leuds to u slortening insteud ol u nurrowing ol tle rods,
plotoetcling miglt not inlluence tle positions ol tle uhsorption hunds hut tle totul
uhsorption signul will decreuse us tle lhS purticles ure dissolving. However, lrom 1LV
Figure 2.10 : Absorption spectrum of an aqueous suspension of nanocrystalline PbS
particles before illumination (a) and after illumination for 15 min using a filter with a
cut-off energy of 2.25 eV (b). The inset is an enlargement of the onset of absorption.
2.0 3.0 4.0 5.0 6.0
0.0
0.5
1.0
1.5
2.0
(b)
(a)
A
b
s
o
r
b
a
n
c
e
2.0 2.6
Energy (eV)
88
meusurements tlut we perlormed it wus cleur tlut tle lhS purticles did not luve u rod
like slupe hut were ulmost splericul.
ln contrust to tle experimentul results on CdS und ZnS, tle results on
nunocrystulline lhS purticles do not provide concrete evidence lor tle possihility ol size
selective plotoetcling lor tlese purticles.
2. 4. 4 Zn0.
1LV meusurements were perlormed to cluructerise tle ZnO purticles und
slown in ligure ?.11. 1le purticles ure crystulline witl interplunur distunces ol uhout
8.8 A. As in tle cuse ol CdS, tlese meusurements only ullow u rougl estimution ol tle
meun purticle rudius, wlicl is uhout 8O A. According to tle literuture, tle ßolr rudius ol
tle exciton lor mucrocrystulline ZnO is equul to 1D A j8D|.
Figure ?.1?A slows tle emission spectrum ol un uqueous suspension ol ZnO
purticles udsorhed on silicu. 1lis spectrum contuins two hunds . u hroud trup emission
hund centred ut ?.O8 eV und un exciton emission hund centred ut 8.4? eV. 1le position
ol tle exciton emission hund corresponds to tle vulue reported in tle literuture lor tle
hundgup ol mucrocrystulline ZnO (vulues hetween 8.?8.4 eV j?,18,8D|,. Figure ?.1?A
ulso slows tle uhsorption spectrum ol tle suspension. Lxtrupolution ol tle steep purt ol
tle uhsorption spectrum gives un onset ol 8.88 eV. Using un experimentul relutionslip
hetween tle onset ol uhsorption und tle meun purticle rudius j14|, un onset ol 8.88 eV
Figure 2.11 : TEM-micrographs of ZnO particles from a 2-propanol suspension before
illumination.
A.
B. C.
500 Å
50 Å
50 Å
Clupter ? . Size Selective llotoetcling
84
corresponds to u purticle rudius ol 85 A, wlicl is in good ugreement witl our 1LV
meusurements hut lurger tlun reported in tle literuture lor u similur prepurution j18|.
For ZnO purticles in tlis size runge tle energetic position ol tle onset ol uhsorption
vuries only very sligltly witl purticle size. 1lerelore tle error in tle culculuted vulue lor
tle meun rudius ol tle ZnO purticles is relutively lurge.
llotoetcling experiments were perlormed using gluss us u cutoll lilter ut 4 eV.
1lese experiments slow tlut nunocrystulline ZnO purticles cun he plotoetcled in
wuter. Figure ?.1?ß slows tle slilt ol tle exciton emission hund lrom 8.4? eV to 8.55 eV
upon illuminution lor 16 lours. 1lis slilt indicutes tlut tle ZnO purticles luve hecome
smuller und luve entered tle quuntum conlinement regime. Alter plotoetcling, tle
ZnO purticles were very instuhle us tle exciton emission hund disuppeured witl time. lt
Figure 2.12 : A : Emission spectrum of an aqueous suspension of nanocrystalline ZnO
particles. The inset shows the exciton emission band together with the absorption
spectrum of the suspension. B : Emission spectra showing the exciton emission bands
of an aqueous suspension of ZnO particles before illumination (a) and after
illumination (b) for 16 hours through glass with a cut-off energy of about 4 eV.
Spectrum (b) was multiplied by a factor 10. Φ
E
denotes the photon flux per constant
energy interval. The emission spectra were recorded upon excitation with 4.1 eV.
3.0 3.5 4.0
0.0
0.5
1.0
(b)
(a) B.
1.0 2.0 3.0 4.0
0
50
100
A.
3.0 4.0
Φ
E

(
a
r
b
.

u
n
i
t
s
)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
Energy (eV)
Energy (eV)
85
wus not possihle to record un uhsorption spectrum ol tle ZnO suspension ulter
illuminution hecuuse tle signul lud hecome too low due to tle loss ol muteriul during
tle plotoetcling process.
Similur to CdS, more inlormution uhout tle meun purticle size und tle purticle
size distrihution cun he ohtuined hy using un empiricul relutionslip hetween tle onset
ol uhsorption und tle purticle size lor ZnO purticles j14|. ln our cuse, emission
meusurements luve to he used insteud ol uhsorption meusurements. 1o he uhle to use
tle relutionslip it is ussumed tlut slilt hetween tle muximum ol tle exciton emission
hund und tle onset ol uhsorption is tle sume helore plotoetcling us ulterwurds. ln tlis
wuy, it is determined tlut u meun purticle rudius ol 85 A ± 15 A helore plotoetcling is
decreused to 15 A ± 5 A ulter plotoetcling. lt is cleur tlut not only tle meun purticle
size lus decreused upon plotoetcling hut ulso tle widtl ol tle purticle size distrihution
lus decreused.
Clupter ? . Size Selective llotoetcling
86
As cun he seen in ligure ?.1?ß, tle widtl ol tle exciton emission hund lus
hecome sligltly smuller upon plotoetcling. At u rudius ol 15 A, tle ZnO purticles luve
not yet entered tle runge in wlicl tle hundgup energy depends very strongly on tle
purticle size. According to Huuse et ul. j14|, u strong dependence ol tle uhsorption
onset on purticle size sturts ut purticle rudii ol uhout 1O A corresponding to un
uhsorption onset ol uhout 8.6 eV.
87
2 . 5 C 0 N C L U 5 | 0 N .
1le experiments descrihed in tlis clupter demonstrute tle leusihility ol size
selective plotoetcling us u meuns to ud¦ust tle size ol nunocrystulline semiconductor
purticles witl u ligl degree ol control. 1le purticle size ulter plotoetcling is
determined hy tle energy ol tle liglt tlut is used lor plotoetcling. 1le teclnique
works very well lor nunocrystulline CdS purticles und enuhles one to prepure u series ol
sumples witl dillerent welldelined purticle rudii, covering u wide runge lrom 85 A to
7.5 A. 1le decreuse in meun purticle size is uccompunied hy u nurrowing ol tle purticle
size distrihution lrom some 4Oº to 1O15º.
1le upplicuhility ol tle teclnique to nunocrystulline semiconductor purticles ol
ZnS und ZnO lus heen demonstruted hut tle wide hundgup ol tlese compounds limits
tle runge lor size selective plotoetcling experiments.
Nunocrystulline lhS purticles ure very dillicult to plotoetcl, most prohuhly due
to tle insoluhility ol lhSO in wuter wlicl could result in tle lormution ol u pussivuting
4
sulplute luyer on tle surluce ol tle lhS purticles.
Clupter 2 . Size Selective llotoetcling
88
4 - . - 4 - + - 5
j1| L.L. ßrus, 1. C/cn. //ys. 8O(D, (1D84, 44O8.
j2| A. Henglein, Toµ. Corr. C/cn. 148 (1D88, 118.
j8| A. Henglein, C/cn. Rco. 8D (1D8D, 1861.
j4| H. Weller, U. Kocl, V. Gutierrez, A. Henglein, 8cr. 8onscngcs. //ys. C/cn. 88 (1D84, 64D.
j5| A. Henglein, 8cr. 8onscngcs. //ys. C/cn. 86 (1D82, 8O1.
j6| Y. Wung, N. Herron, 1. //ys. C/cn. D5 (1DD1, 525.
j7| H. Weller, Angco. C/cn. ln/. LJ. Lng/. 82 (1DD8, 41.
j8| A. Henglein, A. Fo¦tik, H. Weller, 8cr. 8onscngcs. //ys. C/cn. D1 (1D87, 441.
jD| D. Veissner et ul., 8cr. 8onscngcs. //ys. C/cn. 8D (1D85, 121.
j1O| 1. Vossmeyer, L. Kutsikus, V. Giersig, l.G. lopovic, K. Diesner, A. Clemseddine, A. Lyclmüller, H.
Weller, 1. //ys. C/cn. D8 (1DD4, 7665.
j11| S. Gullurdo, V. Gutierrez, A. Henglein, L. 1unutu, 8cr. 8onscngcs. //ys. C/cn. D8 (1D8D, 1O8O.
j12| D.W. ßulnemunn, lsroc/ 1. C/cn. 88 (1DD8, 115.
j18| D.W. ßulnemunn, C. Kormunn, V.R. Hollmunn, 1. //ys. C/cn. D1 (1D87, 878D.
j14| V. Huuse, H. Weller, A. Henglein, 1. //ys. C/cn. D2 (1D88, 482.
j15| L. Spunlel, V.A. Anderson, 1. An. C/cn. Soc. 118 (1DD1, 2826.
j16| D.W. ßulnemunn, A. Henglein, 1. Lilie, L. Spunlel, 1. //ys. C/cn. 88 (1D84, 7OD.
j17| C. Kormunn, D.W. ßulnemunn, V.R. Hollmunn, 1. //ys. C/cn. D2 (1D88, 51D6.
j18| W. Clen, Z.G. Wung, Z.1. Lin, 1.1. Qiun, L.Y. Lin, So/|J S/o/c Connon. 1OO (1DD6, 1O1.
j1D| W. Clen, Z.G. Wung, Z.1. Lin, 1.1. Qiun, L.Y. Lin, Aµµ/. //ys. Lc//. 68(14, (1DD6, 1DDO.
j2O| Y. Wung, N. Herron, 1. //ys. C/cn. D1 (1D87, 257.
j21| G. Counio, S. Lsnoul, 1. Gucoin, 1.l. ßoilot, 1. //ys. C/cn. 1OO (1DD6, 2OO21.
j22| V.L. Steigerwuld, A.l. Alivisutos, 1.V. Gihson, 1.D. Hurris, R. Kortun, A.1. Vuller, A.V. 1luyer, 1.V.
Duncun, D.C. Douglus, L.L. ßrus, 1. An. C/cn. Soc. 11O (1D88, 8O46.
j28| A.R. Kortun, R. Hull, R.L. Opilu, V.G. ßuwendi, V.L. Steigerwuld, l.1. Curroll, L.L. ßrus, 1. An.
C/cn. Soc. 112 (1DDO, 1827.
j24| C.H. Fiscler, H. Weller, L. Kutsikus, A. Henglein, Longno|r 5 (1D8D, 42D.
j25| A. Lyclmüller, L. Kutsikus, H. Weller, Longno|r 6 (1DDO, 16O5.
j26| C.ß. Vurruy, D.1. Norris, V.G. ßuwendi, 1. An. C/cn. Soc. 115 (1DD8, 87O6.
j27| V.G. ßuwendi, l.1. Curroll, W.L. Wilson, L.L. ßrus, 1. C/cn. //ys. D6(2, (1DD2, D46.
j28| D.1. Norris, V.G. ßuwendi, //ys. Rco. 8 58(24, (1DD6, 16888.
j2D| A. Clemseddine, H.Weller, 8cr. 8onscngcs. //ys. C/cn. D7 (1DD8, 686.
j8O| H. Vutsumoto, 1. Sukutu, H. Vori, H. Yoneyumu, C/cn. Lc//. (1DD5, 5D5.
j81| H. Vutsumoto, 1. Sukutu, H. Vori, H. Yoneyumu, 1. //ys. C/cn. 1OO (1DD6, 18781.
j82| A. vun Di¦ken, D. Vunmuekelhergl, A. Vei¦erink, C/cn. //ys. Lc//. 26D (1DD7, 4D4.
j88| D. Veissner, C. ßenndorl, R. Vemming, Aµµ/. Sor/. Sc|. 27 (1D87, 428.
j84| D. Veissner, R. Vemming, ß. Kustening, 1. //ys. C/cn. D2 (1D88, 8476.
j85| l.H.L. Notten, 1.L.A.V. vun den Veerukker, 1.1. Kelly, L/c/|ng o/ lllV Scn|conJoc/ors. An
L/cc/roc/cn|co/ Aµµrooc/, Llsevier Advunced 1eclnology, Oxlord (1DD1,.
j86| A. Fo¦tik, H. Weller, U. Kocl, A. Henglein, 8cr. 8onscngcs. //ys. C/cn. 88 (1D84, D6D.
8D
j87| A. Lyclmüller, A. Husselhurtl, L. Kutsikus, H. Weller, 1.Lon|n. 48&4D (1DD1, 745.
j88| 1. 1ittel, W. Gölde, F. Koherling, 1l. ßuscle, A. Kornowski, H. Weller, A. Lyclmüller, 1. //ys.
C/cn. 8 1O1 (1DD7, 8O18.
j8D| O. Vudelung (ed.,, LonJo//8orns/c|n. Noncr|co/ Jo/o onJ /onc/|ono/ rc/o/|ons/|µs |n sc|cncc onJ
/cc/no/ogy. Vo/onc lll17. Scn|conJoc/ors, Springer Verlug, ßerlin (1D88,, 17h pp. 85115, 1661D4,
17l pp. 155161.
4O
41
C H A P ¡ £ k
N A N 0 C k ¥ S ¡ A L L | N £
Z N 0 P A k ¡ | C L £ S
3
| . | denf | f | cof | cn cf f he ¡ r ons| f | cn
kespcns| b| e f cr f he v| s| b| e £ m| ss| cn
A 8 S ¡ k A C ¡ .
1le emission properties ol suspensions ol nunocrystulline ZnO purticles
witl dillerent purticle sizes ure studied. 1wo emission hunds ure
ohserved, one heing un exciton emission und tle otler tle visihle
emission ol ZnO. 1le energy ol hotl emissions depend on tle purticle
dimensions due to size quuntisution. A lineur relutionslip hetween tle
energetic muximu ol tle two emission hunds is lound. ßecuuse ol tle
dillerence in ellective musses ol electrons und loles in ZnO, tle slope
ol tle lineur relutionslip cleurly indicutes tlut tle visihle emission is
due to tle trunsition ol un electron lrom tle conduction hund to u deep
trup. 1le nuture ol tle deep trup is ulso considered.
Clupter 3 . |unocrystulline ZnO lurticles, purt l
42
3 . I | N ¡ k 0 0 U C ¡ | 0 N .
ZnO lus heen known us u luminescent muteriul lor u century, und some lilty
yeurs ugo it wus discovered tlut liring ZnO powder in u reducing utmosplere gives u
purticulurly ellicient hluegreen luminescent muteriul. 1lis is usuully represented us
ZnO.Zn hecuuse ol tle loss ol oxygen during tle reducing treutment j1|. 1lis muteriul
lus u ligl elliciency us u lowvoltuge plosplor und tle muteriul lus heen used in
vucuum lluorescent displuys (VFD's, und lield emission displuys (FLD's,. Lspeciully tle
lutter upplicution lus hecome importunt recently, since FLD's ure one ol tle promising
cundidutes lor llut punel displuys j2|. lnspired hy tle upplicution ol ZnO, numerous
studies luve uppeured on tle nuture ol tle visihle luminescence lrom ZnO. ln spite ol
ull tlis reseurcl, tle meclunism helind tle visihle luminescence lus heen very dillicult
to estuhlisl, us is cleur lrom u stutement in tle 1DDD edition ol tle llosplor Hundhook
j2| .
¨Tbc or|ç|o o/ /bc /on|ocsccocc cco/cr ooJ /bc /on|ocsccocc nccboo|sn
o/ ZoO.Zo µbosµbors orc horc/y ooJcrs/ooJ¨
Vucl ol tle reseurcl on tle luminescence ol ZnO is perlormed on single
crystulline powders (µm size, or single crystuls. 1wo emission hunds ure usuully lound.
A relutively weuk und nurrow UV emission hund is ohserved uround 38O nm (3.25 eV,,
¦ust helow tle onset ol uhsorption. A mucl stronger und hrouder emission hund is
situuted in tle green purt ol tle visihle spectrum, witl u muximum hetween 5OO und 53O
nm (2.352.5O eV,. 1le UV emission hund is due to tle rudiutive unnililution ol
excitons. 1le hinding energy ol excitons in ZnO is 5D meV und tle exciton emission
cun still he ohserved ut temperutures lur uhove room temperuture j3|. 1le liletime ol
tlis exciton emission is very slort, on tle order ol severul tens to lundreds ol
picoseconds j4|.
As mentioned helore, tle nuture ol tle visihle emission lus heen tle suh¦ect ol
mucl reseurcl. At lirst, tlis emission wus tlouglt to he ussociuted witl divulent copper
impurities j5| hut luter intrinsic delects sucl us interstitiul zinc ions j6| or oxygen
vucuncies j7| were ussumed to he tle recomhinution centres. Alternutive explunutions
involved zinc vucuncies j8|, clemisorhed oxygen jD|, und sulplur impurities j1O|.
During tle lust yeurs, oxygen vucuncies luve heen ussumed to he tle most likely
cundidutes lor tle recomhinution centres involved in tle visihle luminescence ol ZnO
j1113|. ln contrust to tle exciton emission, tle liletime ol tle visihle emission is mucl
longer, viz. in tle µs runge j11|.
43
Studies on tle nuture ol delect centres involved in tle visihle emission olten
use treutment ol mucrocrystulline ZnO in u reducing or oxidizing utmosplere j1214|. ln
tlis wuy, tle numher ol oxygen vucuncies or otler delects cun he vuried und tlis is tlen
reluted to clunges in tle luminescence intensity. A prohlem is tlut tle intensity ol tle
luminescence is inlluenced hy tle widtl ol tle depletion luyer, wlicl is ulso strongly
dependent on delect concentrutions. 1lis mukes it dillicult to druw conclusions on tle
origin ol tle visihle luminescence lrom tle ohserved clunges in tle luminescence
intensity und delect concentrutions j1216|. ln tlis study, nunocrystulline ZnO purticles
ure used wlicl lus two muin udvuntuges. First, tlere is no hund hending in tle purticle
us tle purticles ure mucl smuller tlun tle widtl ol tle depletion luyer. Clunges in
luminescence intensity due to depletion luyer ellects ure tlerelore uhsent. Second, tle
meun purticle size cun he vuried und its inlluence on tle emission properties cun he
studied since tle electronic structure ol ZnO is sizedependent wlen tle purticles ure in
tle quuntum size regime (R<1D A j17|,. 1le cluructeristics ol tle visihle emission ol
mucrocrystulline ZnO (single crystuls or powders, ure very similur to tlose ol
nunocrystulline ZnO purticles. ln hotl cuses, tle energetic position us well us tle
luminescence liletime ure compuruhle. 1lis similurity suggests tlut tle origin ol tle
visihle emission is tle sume lor ull lorms ol ZnO. Furtlermore, tle recomhinution
centres ure prohuhly not ut tle surluce ol tle muteriul hecuuse in single crystuls ol ZnO
witl u very low surlucetovolume rutio tle intensity ol tle visihle emission is still ligl.
1le ideu tlut tle recomhinution centres tlut ure involved in tle visihle emission ol ZnO
must he in tle hulk ol tle muteriul is corrohoruted hy results lrom temperuture
dependent steudystute luminescence meusurements tlut will he presented in clupter 4.
ln tlis clupter, results will he presented ol steudystute luminescence
meusurements perlormed ut room temperuture during tle growtl ol ZnO purticles.
Studies on tle vuriution ol tle energetic positions ol tle muximu ol hotl tle visihle und
tle UV emission hund ol ZnO us u lunction ol purticle size provide novel inlormution on
tle nuture ol tle trunsition responsihle lor tle visihle emission, wlicl is vulid lor ZnO in
generul.
3 . 2 ¡ H £ 0 k ¥ .
llotoexcitution ol u semiconductor purticle results in tle lormution ol Wunnier
excitons. Rudiutive recomhinution ol tlese excitons yield plotons witl un energy close
to tlut ol tle hundgup. Sucl u process is normully relerred to us exciton emission.
Wunnier excitons ure sputiully deloculised und wlen tle semiconductor purticle is very
Clupter 3 . |unocrystulline ZnO lurticles, purt l
44
smull (i.e. in tle quuntum size regime, tley will he deloculised over tle entire purticle
volume. 1le plotogeneruted clurge curriers cun ulso get trupped somewlere in tle
purticle. 1lis cun tuke pluce in slullow levels or in deep trups. Slullowly trupped clurge
curriers ure still sputiully deloculised. Due to tlis deloculisution tle energetic position ol
sucl slullow trups is reluted to tle hund structure ol tle muteriul j1D|. Wlen tle hund
edges slilt us u lunction ol purticle size, us is tle cuse lor quuntumconlined
semiconductor purticles, tle energetic position ol tle slullow trups in tlese purticles
will slow u similur sizedependency.
Wlen u clurge currier is trupped deeper, u loculised energetic stute is ohtuined.
Sucl deep trups cun he lound eitler in tle hulk or ut tle surluce ol u semiconductor
purticle. 1leir energetic position is more or less independent on purticle size und is
determined hy tle locul surroundingings ol tle trupped clurge currier. 1rupping ol u
clurge currier cun occur viu u rudiutive process. 1lis type ol emission is relerred to us
trup emission. |ext to tlese two kinds ol emission (exciton und trup emission,, non
rudiutive processes ure very importunt in semiconductor purticles. Olten surluce stutes
ure involved in nonrudiutive reluxution processes. 1le surluce ol u semiconductor is u
strong perturhution ol tle luttice wlere u ligl concentrution ol hotl slullow und deep
levels provide u putlwuy lor nonrudiutive recomhinution ol plotogeneruted clurge
curriers j2O|. Figure 3.1 sclemuticully slows tle energy levels lor mucrocrystulline us
well us nunocrystulline semiconductors, similur to ligure 1.1 hut now including vurious
trup levels. |ote tlut in ligure 3.1, tle lines representing tle slullow trups (S1, ure
longer tlun tle one representing tle deep trup (D1,. 1lis is to indicute u ligler degree
ol deloculisution.
3 . 3 £ X P £ k | M £ N ¡ A L M £ ¡ H 0 0 S .
3. 3. I Somp| e pr epor of | cn.
Suspensions ol nunocrystulline ZnO purticles cun he prepured in u nonuqueous
solvent sucl us 2propunol j8|. For tlis prepurution 25 ml ol u O.O2 V |uOH solution
wus slowly udded wlile stirring to 225 ml ol u O.OO1 V Zn(CH COO, 2H O solution, ulter
3 2 2
.
hotl solutions were lirst cooled to O °C. A rupid lormution ol extremely smull ZnO
purticles is lollowed hy u relutively slow growtl ol tle purticles. 1le rute ol purticle
growtl depends strongly on tle temperuture. Wlen tle suspension is uged ut room
temperuture, it tukes severul duys lor tle purticles to uttuin tleir linul size. 1LV
meusurements on tle lully grown purticles luve slown u crystulline structure und u
Figure 3.1 : Schematic picture of the energy levels of a macrocrystalline (left) and a
nanocrystalline semiconductor (right). E
g
is the bandgap of the macrocrystalline
semiconductor while E denotes the corresponding energy difference for a
nanocrystalline semiconductor. Two different bulk trap levels are shown, ST being a
shallow trap and DT a deep trap. The shift of the energetic positions of these trap levels
upon a decrease of particle size are indicated by dashed lines. For the nanocrystalline
semiconductor, also surface states (SS) are shown.
E
DT
ST
SS
SS
SS
SS
ST
E
g
VB
CB
45
meun rudius ol uhout 3O A j21|. 1lis corresponds to purticles tlut contuin
upproximutely 5OOO moleculur ZnO units. ßy compuring our results to tlose ohtuined hy
Huuse et ul. j22| we luve estimuted tle meun rudius ol tle smullest ZnO purticles tlut
were studied to he uhout 7 A, corresponding to 6O moleculur ZnO units. Directly ulter
mixing tle two sturting solutions, tle resulting reuction mixture is stored ut room
temperuture. At regulur intervuls, u sumple is tuken lrom tlis reuction mixture wlicl is
used lor opticul meusurements. ln tlis wuy, it is possihle to study nunocrystulline ZnO
purticles witl dillerent sizes.
3. 3. 2 0pf | co| meosur emenf s.
Lmission meusurements were perlormed on u SlLX Fluorolog
spectroplotometer model F2OO2 equipped witl two douhlegruting O.22 m SlLX 168O
monoclromutors und u 45O W Xe lump us tle excitution source. 1le emission spectru
were corrected lor tle spectrul response ol tle emission monoclromutor und tle lV
Clupter 3 . |unocrystulline ZnO lurticles, purt l
46
tuhe. Ahsorption meusurements were perlormed on u lerkin Llmer Lumhdu 16 UV/VlS
spectroplotometer. 1le uhsorhunce (A, wus meusured us tle loguritlm ol tle
trunsmission (T=l/l , witl 2propunol us u relerence.
O
3 . 4 k £ S U L ¡ S A N 0 0 | S C U S S | 0 N .
3. 4. I £ m| ss| cn meosur emenf s.
Figure 3.2 slows u typicul room temperuture emission spectrum ol u suspension
ol ZnO purticles upon plotoexcitution, togetler witl tle uccompunying uhsorption
spectrum. 1le emission spectrum contuins two hunds. 1le muximum ol tle relutively
weuk und nurrow emission hund in tle UV (3.6 eV, overlups witl tle onset ol
uhsorption. 1lis hund cun he ussigned to tle rudiutive recomhinution ol slullowly
trupped clurge curriers (exciton emission,.
A more intense hroud emission hund is ohserved in tle visihle purt ol tle
spectrum (2.2 eV,. 1lis emission must involve u deeply trupped clurge currier us tle
muximum ol tle hund lies ut un energy uhout 1.5 eV lower tlun tle onset ol uhsorption.
Figure 3.3 contuins tlree emission spectru tuken ut dillerent times during tle
growtl ol ZnO purticles in 2propunol ut room temperuture. Spectrum (u, is tuken ulter
15 min, spectrum (h, ulter 12O min, und spectrum (c, ulter 42O min ol purticle growtl.
Lucl spectrum contuins u weuk solvent Rumun peuk ut 4 eV, slilted witl respect to tle
excitution energy hy uhout O.4 eV (35OO cm , corresponding to un OHvihrution,. All
−1
tle emission spectru in ligure 3.3 slow hotl u visihle und u UV emission hund, even lor
tle smullest purticles (spectrum (u,,. lreviously, lor ZnO purticles witl u rudius smuller
tlun uhout 25 A only u visihle emission wus ohserved j8|.
As cun he seen in ligure 3.3, tle UV emission hund us well us tle visihle
emission hund slilt to lower energies upon purticle growtl. For tle visihle emission
process tlis indicutes tlut, next to u deep trup level, ulso one ol tle hund edges (or u
slullow level close to one ol tlese edges, must he involved. For u collection ol
emission spectru, tle muximum ol tle visihle emission hund is plotted versus tlut ol tle
UV emission hund in ligure 3.4. 1lere is un upproximute lineur relutionslip hetween tle
energetic positions ol tlese two muximu, witl u slope ol uhout O.6.
Similur experiments us descrihed uhove lor suspensions in 2propunol luve
heen curried out lor suspensions ol ZnO purticles in etlunol. 1le prepurution ol ZnO
suspensions in etlunol lus heen descrihed in tle literuture j23,24|. Aguin u lineur
Figure 3.2 : Emission spectrum of a suspension of ZnO particles in 2-propanol, taken
at room temperature upon excitation with light of 4.4 eV (a). Φ
E
denotes the photon
flux per constant energy interval. Spectrum (b) shows the absorption spectrum of the
same suspension.
2.0 3.0 4.0 5.0
0.0
0.5
1.0
0.0
0.5
1.0
(b) (a)
Energy (eV)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
A
b
s
o
r
b
a
n
c
e
Figure 3.3 : Emission spectra of suspensions of nanocrystalline ZnO particles in 2-
propanol taken after different periods of particle growth at room temperature : (a) 15
min, (b) 120 min, and (c) 420 min. On the left, the broad visible emission band is
shown and on the right the sharp UV band. Φ
E
denotes the photon flux per constant
energy interval. The ZnO particles were excited with light of 4.4 eV.
Energy (eV) Energy (eV)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
1.0 2.0 3.0
0
10
20
30
40
50
60
70
3.0 4.0 5.0
0.0
0.5
1.0
1.5
2.0
(c)
(b)
(a)
(a)
(b)
(c)
47
dependence witl u slope ol O.6 wus lound hetween tle onset ol tle uhsorption hund
und tle muximum ol tle visihle emission hund j25|.
? ?
ç
?
0 e h 0
h 1 1 1.8
8 1
A
- -
4 4 πε ε
∗ ∗

 
= + ⋅ + −
 
 
2 2
2
0
0 9
8 4πε ε


= + −
ç
e
h . e
( ) A -
4 4
2 2
2
0
0 9
8 4πε ε


= −
h
h . e
( ) D
4 4
Clupter 3 . |unocrystulline ZnO lurticles, purt l
48
(3.1,
(3.2u,
(3.2h,
3. 4. 2 | denf | f | cof | cn cf f he v| s| b| e em| ss| cn f r ons| f | cn.
One cun tlink ol two possihle meclunisms lor tle visihle emission . (1,
recomhinution ol u slullowlytrupped (deloculised, electron witl u deeplytrupped
lole, or (2, recomhinution ol u slullowlytrupped lole witl u deeplytrupped electron.
1lese two possihilities ure slown in ligure 3.5. 1o distinguisl hetween tle two
processes, we consider tle sizedependence ol tle positions ol hotl emission hunds. As
discussed in clupter 1, tle sizedependence ol tle hundgup energy (L, cun he
represented hy tle lollowing equution j1D| .
ln equution (3.1,, L is tle hundgup ol tle mucrocrystulline muteriul, n* und n*
g e l
tle ellective musses ol tle electron und lole respectively (hotl in units n ,, ε tle ligl
O

lrequency dielectric constunt, und R tle rudius ol tle purticle. ln equution (3.1,, two
sizecorrections on tle hundgup ol u mucrocrystulline semiconductor ure slown. 1le
lirst correction is u conlinement term (∝R , und tle second u Coulomh interuction
−2
term (∝R ,. Altlougl equution (3.1, is not udequute lor culculuting uhsolute vulues lor
−1
L, tle dependency ol L on R us given hy equution (3.1, cun he used to descrihe tle
experimentul results j24|. Furtlermore, us we luve studied ZnO purticles tlut were
crystulline und consisted ol uhout 5O5OOO moleculur ZnO units, tle vulidity ol tle
ellective muss upproximution cun he ussumed.
ln ligure 3.5, tle edge ol tle conduction hund (c, und ol tle vulence hund (b,
is slown versus purticle size. Using equution (3.1,, tle lollowing expressions cun he
given lor tle uhsolute vulues ol tle slilts ol tle hund edges us u lunction ol purticle
rudius R .
Figure 3.4 : Energy of the maximum of the visible emission band (VE) versus the
energy of the maximum of the UV emission band (UVE). The straight line is a linear fit
to the experimental data points.
Maximum UVE (eV)
M
a
x
i
m
u
m

V
E

(
e
V
)
3.50 3.60 3.70 3.80 3.90
2.00
2.10
2.20
2.30
2.40
Figure 3.5 : Schematic overview of the two possibilities for trap emission in
nanocrystalline ZnO particles. A : recombination of a delocalised electron with a
deeply-trapped hole. B : recombination of a delocalised hole with a deeply-trapped
electron. In both pictures, (e) represents the shift of the conduction band edge versus
particle radius, (h) the shift of the valence band edge versus particle radius and the
dashed line the energetic position of a trapped charge carrier.
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
A. B.
(h)
(e)
(h)
(e)
5 10 15 20 25 30 5 10 15 20 25 30
E
n
e
r
g
y

(
e
V
)
E
n
e
r
g
y

(
e
V
)
Radius (Å) Radius (Å)
4D
1le slilt ol tle conduction hund edge us u lunction ol purticle size, us given hy
equution (3.2u,, is determined hy tle ellective muss ol tle electron (n*,, wlicl is
e
O.28n lor ZnO j17|. For tle slilt ol tle vulence hund edge (equution (3.2h,,, tle
O
ellective muss ol tle lole (n*, lus to he used, wlicl lus u vulue ol O.5On lor ZnO
l O
Clupter 3 . |unocrystulline ZnO lurticles, purt l
5O
j17|. 1lis dillerence in ellective musses ol tle clurge curriers leuds to u dillerent size
dependent slilt lor hotl hund edges. 1le energetic position ol tle deeply trupped
clurge currier (dusled line in ligure 3.5, is ussumed to he independent ol purticle size,
us wus expluined previously in tle tleory. ll tle visihle emission is due to
recomhinution ol u deloculised electron witl u deeplytrupped lole (ligure 3.5A,, tle
slilt ol tle energetic position ol tle trup emission us u lunction ol purticle size is
determined hy tle slilt ol tle conduction hund edge. For tle otler option,
recomhinution ol u deloculised lole witl u deeplytrupped electron (ligure 3.5ß,, tle
slilt ol tle vulence hund edge determines tle slilt ol tle visihle emission.
ßecuuse tle energetic positions ol hotl hund edges cun he culculuted us u
lunction ol purticle size it is possihle to slow tle relutionslip hetween tle energies ol
tle exciton emission und tle trup emission lor tle two possihilities slown in ligure 3.5.
For hotl cuses un upproximute lineur relutionslip is ohtuined us cun he seen in ligure
3.6 (dusled lines,. Wlen tle Coulomh interuction tle R term lrom equution (3.1, is
−1
not tuken into uccount, exuct lineur relutionslips hetween tle energies ol tle two
emissions ure lound (solid lines in ligure 3.6,. 1le slope ol tle lineur relutionslip lor
eucl ol tle two cuses slown in ligure 3.5 does not depend mucl on wletler u
Coulomh interuction is included or not. 1le exuct lineur relutionslip hetween tle
emission energies witlout tuking into uccount tle Coulomh interuction gives u slope ol
µ/n* (µ is tle reduced muss ol tle exciton wlicl lus u vulue ol O.18n lor ZnO,. For
O
tle situution in ligure 3.5A, using n*, u vulue ol O.64 is ohtuined lor tle slope. ln tle
e
cuse ol ligure 3.5ß, n* lus to he used wlicl gives u vulue ol O.36 lor tle slope. ln ligure
l
3.4, tle lineur relutionslip hetween tle energies ol tle two emission muximu lrom our
experiments is slown. 1le slope is uhout O.6, in good ugreement witl tle vulue us
lound lor tle situution represented in ligure 3.5A. 1lis indicutes tlut in tle cuse ol
nunocrystulline ZnO purticles tle visihle emission involves u trunsition ol un electron
lrom tle conduction hund (or u slullow level close to tlis hund, to u trup level
upproximutely 2 eV helow tle conduction hund edge. As wus mentioned in tle
introduction, oxygen vucuncies ure olten identilied us tle recomhinution centres lor tle
visihle emission ol mucrocrystulline ZnO j5,6|. As wus mentioned in tle introduction to
tlis clupter, tle cluructeristics ol tle visihle emission ure very similur lor
mucrocrystulline und nunocrystulline ZnO und it is very likely tlut tle origin ol tlis
emission is tle sume in hotl cuses. Clupter 4 contuins more experimentul results tlut
conlirm tlis ussertion. ln tlis clupter, sizedependent und temperuturedependent
meusurements ol hotl tle emission hunds (UV und visihle, ure presented. ßused on tle
results, u model is derived in wlicl tle visihle emission is ussigned to tle
recomhinution ol un electron witl u V centre (deeply trupped lole,.
O
••
Figure 3.6 : Theoretical relationship between the energetic maxima of the trap
emission band and the UV emission band. The dashed lines result from a calculation
including the Coulomb interaction while the solid lines represent the relationship
without taking the Coulomb interaction into account. The arrows point to the values
obtained for particle radii of 10 Å and 30 Å respectively. On the right axis of the graph it
is indicated which model from figure 3.5 is used for the calculation.
Maximum UVE (eV)
M
a
x
i
m
u
m

V
E

(
e
V
)
3.0 3.2 3.4 3.6 3.8 4.0
2.0
2.2
2.4
2.6
2.8
3.0
}
}
5A
5B
10Å
30Å
51
Clupter 3 . |unocrystulline ZnO lurticles, purt l
52
3 . 5 C 0 N C L U S | 0 N .
Lmission meusurements were perlormed ut room temperuture on suspensions
ol nunocrystulline ZnO purticles ol dillerent sizes in 2propunol. All tle suspensions
slow two emission hunds. One is u relutively weuk und slurp UV hund witl u muximum
close to tle uhsorption onset wlicl is ussigned to tle rudiutive recomhinution ol
excitons. 1le second hund is u more intense und hroud emission hund in tle visihle
purt ol tle spectrum, slilted hy upproximutely 1.5 eV witl respect to tle uhsorption
onset. 1lis emission process involves u deeply trupped clurge currier.
1le energetic positions ol tle muximu ol hotl emission hunds depend on tle
size ol tle ZnO purticles . tley slilt to lower energies us tle purticle size increuses. 1lis
is u quuntum size ellect und cun he understood in terms ol sputiul conlinement ol
clurge curriers. From tle sizedependency, tle visihle emission hund is ussigned to u
trunsition ol u plotogeneruted electron lrom tle conduction hund edge (or lrom u
slullow level close to tle conduction hund edge, to u trup level, positioned
upproximutely 2 eV helow tle conduction hund edge. 1lis trunsition occurs in tle hulk
ol tle muteriul und upplies to ZnO in generul.
53
RE F E R E N C E S .
[ 1] R.E. Shrader, H.W. Leverenz, J. Opt. Soc. Am. 37 ( 1947) 939.
[ 2] S. Shi onoya, W.M. Yen ( edi tors) , Phosphor Handbook, CRC PressLCC, Boca Raton 1999, p. 255.
[ 3] K . Shi bahara, N. K uroda, S. Ni shi no, H. Matsunami , Jpn. J. Appl. Phys. 26 ( 1987) L1815.
[ 4] V.V. Travni kov, A. Frei berg, S.F. Savi khi n, J. Lumin. 47 ( 1990) 107.
[ 5] R. Di ngle, Phys. Rev. Lett. 23( 11) ( 1969) 579.
[ 6] E. Mollwo, Z. Phys., 138 ( 1954) 478.
[ 7] F.A. K röger, H.J. Vi nk, J. Chem. Phys. 22( 2) ( 1954) 250.
[ 8] J.M. Smi th, W.E. Vense, Phys. Rev. A 31( 3) ( 1970) 147.
[ 9] F. van Craeynest, W. Maenhout-van der Vorst, W. Dekeyser, Phys. Status Solidi 8 ( 1965) 841.
[ 10] W. Lehman, J. Electrochem. Soc. 115 ( 1968) 538.
[ 11] M. Anpo, Y. K ubokawa, J. Phys. Chem. 88 ( 1984) 5556.
[ 12] K . Vanheusden, W.L. Warren, C.H. Seager, D.R. Tallant, J.A. Voi gt, B.E. Gnade, J. Appl. Phys. 79( 1)
( 1996) 7983.
[ 13] K . Vanheusden, C.H. Seager, W.L. Warren, D.R. Tallant, J.A. Voi gt, Appl. Phys. Lett. 68( 3) ( 1996)
403.
[ 14] T. Seki guchi , N. Ohashi , Y. Terada, Jpn. J. Appl. Phys. II 36( 3A) ( 1997) L289.
[ 15] A. Pfanel, J. Electrochem. Soc. 109 ( 1962) 502.
[ 16] G.H. Schoenmakers, D. Vanmaekelbergh, J.J. K elly, J. Phys. Chem. 100 ( 1996) 3215.
[ 17] O . Madelung ( ed.) , Landolt-Börnstein. Numerical data and functional relationships in science and
technology. Volume III-17: Semiconductors, Spri nger Verlag, Berli n ( 1988) , 17b pp. 35-115.
[ 18] D.W. Bahnemann, C. K ormann, M.R. Hoffmann, J. Phys. Chem. 91 ( 1987) 3789.
[ 19] L.E. Brus, J. Chem. Phys. 80( 9) ( 1984) 4403.
[ 20] J.I . Pankove, Optical Processes in Semiconductors, Dover Publi cati ons, New York ( 1971) .
[ 21] A. van Di jken, A.H. Jansen, M.H.P. Smi tsmans, D. Vanmaekelbergh, A. Mei jeri nk, Chem. Mater.
10( 11) ( 1998) 3513.
[ 22] M. Haase, H. Weller, A. Henglei n, J. Phys. Chem. 92 ( 1988) 482.
[ 23] L. Spanhel, M.A. Anderson, J. Am. Chem. Soc. 113 ( 1991) 2826.
[ 24] E.A. Meulenkamp, J. Phys. Chem. B 102 ( 1998) 5566.
[ 25] E.A. Meulenkamp, unpubli shed results.
54
55
C H A P ¡ £ k
N A N 0 C k ¥ S ¡ A L L | N £
Z N 0 P A k ¡ | C L £ S
4
| | . ¡ he K| nef | cs cf f he kod| of | ve ond
Ncn-kod| of | ve Pr ccesses upcn Phcf cexc| f of | cn
A 8 S ¡ k A C ¡ .
ln tlis second clupter on tle opticul properties ol nunocrystulline ZnO
purticles, tle results ol steudystute und timeresolved luminescence
meusurements ure presented. 1le experiments were perlormed on
suspensions ol nunocrystulline ZnO purticles ol dillerent sizes und ut
dillerent temperutures. ln ull cuses two emission hunds ure ohserved.
One is un exciton emission hund und tle second un intense und hroud
visihle emission hund, slilted hy upproximutely 1.5 eV witl respect to
tle uhsorption onset. As tle size ol tle purticles increuses, tle intensity
ol tle visihle emission decreuses wlile tlut ol tle exciton emission
increuses. As tle temperuture decreuses, tle relutive intensity ol tle
exciton emission increuses. ln uccordunce witl tle results presented in
clupter 8, it is ussumed tlut tle visihle emission is due to u trunsition ol
un electron lrom u level close to tle conduction hund edge to u deeply
trupped lole in tle hulk (V , ol tle ZnO purticle. 1le temperuture
O
••
und sizedependence ol tle rutio ol tle visihle to exciton luminescence
und tle kinetics ure expluined hy u model in wlicl tle plotogeneruted
lole is trunslerred lrom tle vulence hund to u V level in tle hulk ol tle
O

purticle in u twostep process. 1le lirst step ol tlis process is un ellicient
surlucetrupping, prohuhly ut u O site.
?−
Clupter + . |unocrystulline ZnO lurticles, purt ll
56
4 . I | N ¡ k 0 0 U C ¡ | 0 N .
ZnO us u luminescent muteriul wus ulreudy introduced in tle previous clupter.
1lere it wus mentioned tlut ZnO slows un ellicient hluegreen emission wlicl lus
resulted in tle implementution ol tle muteriul in vurious upplicutions, sucl us vucuum
lluorescent displuys (VFD's, und lield emission displuys (FLD's, j1|. Despite mucl
reseurcl it turned out to he very dillicult to elucidute tle meclunism ol tlis visihle
emission process. An importunt reuson lor tlis is tlut most ol tle reseurcl lus heen
curried out on mucrocrystulline ZnO (powders or single crystuls,. 1lis involves tuking
into uccount sucl ellects us hund hending wlen considering tle emission intensities us
u lunction ol delect concentrution. ln clupter 8, un overview wus given ol tle uttempts
to identily tle trunsition responsihle lor tle visihle emission. 1lese uttempts luve led to
tle ussumption tlut oxygen vucuncies ure tle most likely cundidutes lor tle
recomhinution centres involved in tle visihle luminescence ol ZnO j?+|.
lt wus slown in tle previous clupter tlut hy using nunocrystulline ZnO purticles
tle visihle emission cun he ussigned to u trunsition ol u plotogeneruted electron lrom u
slullow level close to tle conduction hund edge to u deeply trupped lole. Furtlermore,
it wus suggested tlut tlis trunsition occurs in tle hulk ol tle muteriul und upplies to ZnO
in generul. ln tlis clupter, we will eluhorute on tlis ussignment hy presenting results ol
steudystute und timeresolved luminescence meusurements, us u lunction ol purticle
size und temperuture. Sucl detuiled meusurements luve not heen reported helore. 1le
temperuturedependence ol tle emission intensities und emission liletimes wus studied
hy lreezing tle suspensions to liquid lelium temperuture. 1le results presented in tlis
clupter will slow tlut, ultlougl tle uctuul trunsition tlut gives tle visihle emission is
tle sume ull lorms ol ZnO, tle kinetics involved in tle totul process cun he very
dillerent lor nunocrystulline und mucrocrystulline ZnO. A model is proposed lor tle
competition hetween ultruviolet exciton recomhinution, trup recomhinution giving rise
to visihle luminescence und nonrudiutive recomhinution in luir ugreement witl tle
results lrom tle steudystute und timeresolved meusurements. Also, tle nuture ol tle
trup involved in tle visihle emission process will he considered.
57
4 . 2 £ X P £ k | M £ N ¡ A L M £ ¡ H 0 0 S .
4. 2 . I Somp| e pr epor of | cn.
Suspensions ol nunocrystulline ZnO purticles cun he prepured in nonuqueous
solvents sucl us etlunol or ?propunol. For tle prepurution in ?propunol j5|, ?5 ml ol u
O.O? V |uOH solution wus slowly udded wlile stirring to ??5 ml ol u O.OO1 V
Zn(CH COO, ?H O solution, ulter hotl solutions were lirst cooled to O °C. 1le
8 ? ?
.
prepurution in etlunol is very similur j6,7|. ln tlis cuse, 5O ml ol u O.1+ V LiOH H O
.
?
solution (prepured using un ultrusonic hutl, wus udded to 5O ml ol u O.1 V
Zn(CH COO, ?H O solution. Aguin, hotl solutions were lirst cooled to O °C helore tle
8 ? ?
.
lydroxide solution wus udded slowly to tle zinc solution wlile stirring. 1le prepurution
in etlunol yields u suspension witl u purticle concentrution uhout two orders ol
mugnitude ligler tlun tle ?propunol prepurution. 1le kinetics ol tle purticle growtl is
not tle sume lor hotl prepurutions hut in hotl cuses, u rupid lormution ol extremely
smull ZnO purticles is lollowed hy u relutively slow growtl ol tle purticles j7,8|. 1le rute
ol purticle growtl depends strongly on tle temperuture. Wlen tle suspension is uged ut
room temperuture, it tukes severul duys lor tle purticles to uttuin tleir linul size. ln
etlunol, tle rute ol growtl ol ZnO purticles is somewlut slower tlun in ?propunol. 1LV
meusurements luve slown tlut tle meun purticle diumeter ol tle lully grown purticles
is uhout 6O A (see clupter ? ol tlis tlesis,. 1lis corresponds to purticles tlut contuin
upproximutely 5OOO moleculur ZnO units.
Directly ulter mixing tle two sturting solutions, tle resulting reuction mixture is
stored ut room temperuture. At regulur intervuls, u sumple is tuken lrom tlis reuction
mixture wlicl is used lor opticul meusurements. ln tlis wuy, one cun study
nunocrystulline ZnO purticles witl dillerent sizes.
4. 2 . 2 0pf | co| meosur emenf s.
1le plotoluminescence meusurements were perlormed on u SlLX Fluorolog
spectrolluorometer model F?OO? equipped witl two douhlegruting O.?? m SlLX 168O
monoclromutors und u +5O W xenon lump us tle excitution source. 1le emission
spectru were corrected lor tle spectrul response ol tle emission monoclromutor und
tle lV tuhe. Ahsorption meusurements were perlormed on u lerkinLlmer Lumhdu 16
UV/Vis spectroplotometer.
Decuy time meusurements were perlormed on u Lumhdu llysik LlX1OO XeCl
Lxcimer luser setup (excitution witl 8O8 nm corresponding to +.O8 eV witl u pulse
Clupter + . |unocrystulline ZnO lurticles, purt ll
58
widtl ol uhout 1O ns, equipped witl u 1ohin Yvon HR1OOOV monoclromutor und u
1ektronix ?++O digitul oscilloscope lor tle liletime meusurements.
Veusurements ut low temperutures were perlormed in tle lollowing wuy. A
smull quurtz cuvet wus lilled witl u lew µl ol u ZnO suspension. 1lis cuvet wus mounted
on u sumple lolder und inserted into u llow cryostut. ßy llowing witl liquid lelium, tle
suspension could he cooled to + K, well helow its lreezing point. 1le solvent, wletler it
is etlunol or ?propunol, lreezes into u trunspurent solid.
For tle decuy time meusurements und tle lowtemperuture meusurements, only
suspensions ol nunocrystulline ZnO purticles in etlunol were used hecuuse ol tleir ligl
luminescence intensity.
4 . 3 k £ S U L ¡ S .
4. 3 . I Abscr pf | cn.
Ahsorption spectru were tuken ut regulur intervuls during tle growtl ol ZnO
purticles in ?propunol ut room temperuture. Four ol tlese spectru cun he seen in ligure
+.1A. 1wo ellects ure cleurly visihle lrom tlese spectru. First, tle uhsorption onset slilts
to lower energies upon purticle growtl. Second, ut tle eurly stuges ol purticle growtl tle
uhsorption spectru exlihit u distinct muximum directly ulter tle onset ol uhsorption.
1lis leuture hecomes less pronounced us tle ugeing process progresses. ßotl ellects
cun he understood in terms ol u decreuse in quuntum conlinement upon purticle
growtl. As tle purticle size increuses, tle hundgup gruduully slilts towurds tle vulue lor
mucrocrystulline ZnO (∼8.? eV,, lence tle slilt in uhsorption onset. Also, tle density ol
stutes ut tle edges ol tle energy hunds clunges wlicl leuds to u less pronounced
structure in tle uhsorption spectru. From tle extrupolution ol tle steep purt ol tle
uhsorption curve one cun estimute tle meun purticle rudius hy using results ohtuined hy
Huuse et ul. jD|. ln tlis wuy, one cun visuulize tle ugeing process ol ZnO purticles hy
plotting tle meun purticle rudius versus time (see ligure +.1ß,. Figure +.1ß slows tlut
witlin u lew minutes ulter udding tle lydroxide solution to tle zinc solution very smull
ZnO clusters ure lormed, witl u meun rudius ol upproximutely 7 A, corresponding to
uhout 6O moleculur ZnO units. 1lese ZnO purticles slow very strong quuntum
conlinement ellects. 1lis is expected, us tley ure mucl smuller tlun tle ßolr rudius ol
tle exciton, wlicl is uhout ?O A lor mucrocystulline ZnO.
Figure 4.1 : A : Absorption spectra of a suspension of nanocrystalline ZnO particles in
2-propanol measured at regular intervals during particle growth at room temperature.
Spectrum (a) is taken after 15 min and spectrum (b) after 240 min. The two spectra in
between (a) and (b) are taken after 60 min and 120 min respectively. B : Mean particle
radii as determined by extrapolating the steep part of the absorption spectrum to the
energy axis and using an empirical relationship between this energy value and the
mean particle radius as obtained by Haase et al. [24].
3.0 4.0 5.0
0
20
40
60
80
100
(b) (a)
0 50 100 150 200 250
0
10
20
30
40
(b)
(a)
A. B.
1
-
T
r
a
n
s
m
i
s
s
i
o
n

(
%
)
Energy (eV) Time (min)
R
a
d
i
u
s

(
Å
)
Figure 4.2 : Emission spectra of suspensions of nanocrystalline ZnO particles in 2-
propanol taken after different periods of particle growth at room temperature : (a) 15
min, (b) 120 min, and (c) 420 min. On the left, the broad visible emission band is
shown and on the right the sharp UV band. Φ
E
denotes the photon flux per constant
energy interval. The ZnO particles were excited with light of 4.4 eV.
Energy (eV) Energy (eV)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
1.0 2.0 3.0
0
10
20
30
40
50
60
70
3.0 4.0 5.0
0.0
0.5
1.0
1.5
2.0
(c)
(b)
(a)
(a)
(b)
(c)
5D
Clupter + . |unocrystulline ZnO lurticles, purt ll
6O
4. 3 . 2 Sf eody-sf of e | um| nescence.
Figure +.? contuins tlree emission spectru, tuken ut regulur intervuls during tle
growtl ol ZnO purticles in ?propunol ut room temperuture. Lucl spectrum contuins u
weuk solvent Rumun peuk ut + eV, slilted witl respect to tle excitution energy hy uhout
O.+ eV (85OO cm , corresponding to un OHvihrution,. 1le relutively nurrow emission
−1
hund in tle UV (∼8.58.8 eV, is due to tle direct recomhinution ol plotogeneruted
clurge curriers (exciton emission,. A more intense hroud emission hund is ohserved in
tle visihle purt ol tle spectrum (∼?.1?.8 eV,. As cun he seen lrom ligure +.?, tle UV
emission hund us well us tle visihle emission hund slilt to lower energies upon purticle
growtl.
A second leuture is evident lrom ligure +.?. As tle ZnO purticles grow higger,
tle intensity ol tle UV emission increuses wlile tlut ol tle visihle emission decreuses.
Figure +.8 slows tle intensities ol hotl emission hunds us u lunction ol purticle size.
From ligure +.1 it is cleur tlut tle uhsorhunce ol tle sumples ut tle excitution energy
(+.+ eV, does not clunge mucl. 1le intensity ol tle visihle emission decreuses to uhout
lull ol its originul vulue, wlile tlut ol tle UV emission increuses witl 6Oº. As tle
intensity ol tle UV emission hund is only u lruction ol tlut ol tle visihle emission hund,
tle totul integruted emission intensity decreuses upon purticle growtl. lt is well known
tlut tle visihle emission hund is quencled upon UV illuminution in tle uhsence ol
udsorhed oxygen on tle surluce ol ZnO purticles j5|. We luve studied tle emission ol
nunocrystulline ZnO purticles in tle presence und uhsence ol udsorhed oxygen, tle
results ol wlicl cun he lound in clupter 5 ol tlis tlesis. 1le size und temperuture
dependent ellects us descrihed in tlis clupter do not result lrom tle uhsence ol
udsorhed oxygen us ull meusurements were perlormed in tle presence ol oxygen. Also,
vuriutions on u timescule ol seconds or minutes wlicl ure ussociuted witl tle uhsence
ol udsorhed oxygen ure not ohserved.
4. 3 . 3 ¡ | me-r esc| ved | um| nescence.
Decuy time meusurements were perlormed on u suspension ol nunocrystulline
ZnO purticles in etlunol. 1le time evolution ol tle emission intensity wus monitored ut
tle muximum ol tle visihle emission hund (∼?.8 eV,. 1le meusurements were
perlormed hotl ut un eurly stuge ol purticle growtl (4∼1O A, us well us on lully grown
ZnO purticles (4∼8O A,. 1le results ure plotted in ligure +.+. ln hotl cuses tle decuy ol
tle visihle emission intensity slows u multiexponentiul heluviour. 1le experimentul
dutu points were litted to u hiexponentiul decuy lunction. 1le lustest ol tle two decuy
Figure 4.3 : Dependence of the emission intensities on the particle radius for the two
types of emission from nanocrystalline ZnO particles at room temperature. The solid
circles represent the UV emission band and the open circles the visible emission band.
0 10 20 30 40
0.0
0.1
0.2
0.3
20
30
40
50
60
Radius (Å)
I
n
t
e
n
s
i
t
y

e
m
i
s
s
i
o
n

b
a
n
d

(
a
r
b
.

u
n
i
t
s
)
Figure 4.4 : Luminescence lifetime measurements of the visible emission from two
suspensions of nanocrystalline ZnO particles in ethanol, measured at room
temperature. Curve (a) : suspension of particles that are fully grown (R=30 Å) and
curve (b) : suspension at an early stage of particle growth (R=10 Å). The excitation is at
4 eV (308 nm).
0.0 2.0 4.0 6.0 8.0
1
10
100
(b) (a)
Time (µs)
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
61
Clupter + . |unocrystulline ZnO lurticles, purt ll
6?
components ohtuined in tlis wuy mude up more tlun DOº ol tle decuy signul. 1le
vulue ol tle decuy time ol tlis component is 1.8+ µs lor tle lurge purticles (u,, und O.D?
µs lor tle smull purticles (h,. 1lis indicutes tlut tle liletime ol tle visihle emission
increuses upon purticle growtl. 1le decuy time ol tle slower component wus
upproximutely 5 µs. For tle UV emission, tle liletime wus slorter tlun tle pulse widtl ol
tle luser (∼1O ns, und could tlerelore not he meusured.
4. 3 . 4 ¡ emper of ur e-dependenf | um| nescence.
1le emission cluructeristics ol u suspension ol nunocrystulline ZnO purticles in
etlunol were meusured in u temperuture runge lrom + K to room temperuture. 1le
results ol tlis meusurement ure presented in ligure +.5. As tle temperuture increuses,
tle intensity ol hotl emission hunds decreuses. 1le ohservution tlut ut temperutures
hetween 7O K und 1?O K tle intensity ol tle visihle emission hund increuses is prohuhly
un urteluct. Also, hotl emission hunds slilt to lower energies us tle temperuture
increuses. 1le temperuturedependence ol tle muximum ol tle ultruviolet emission
hund closely resemhles tle temperuturedependence ol tle hundgup ol
mucrocrystulline ZnO j1O|. 1le muximum ol tle visihle emission slows u dillerent
heluviour. Up to 75 K, tle energetic position ol tle muximum ol tle visihle emission is
independent ol temperuture. At temperutures hetween 7518O K tle muximum slilts
towurds lower energies hy upproximutely 1 meV K . At temperutures ligler tlun 18O K,
.
−1
tle energetic position is more or less independent ol temperuture uguin. ln tle sume
temperuture runge, tle liletime ol tle visihle emission wus monitored. ln ligure +.6, tle
decuy curves ut + K, 75 K und room temperuture ure slown. At + K, tle decuy ol tle
visihle emission is single exponentiul witl u decuy time ol ?75 ns. At 75 K, u second
component witl u decuy time ol uhout 1.5 µs uppeurs in tle decuy curve ol tle visihle
emission. 1le uppeurunce ol tlis µscomponent coincides witl tle slilt ol tle
muximum ol tle visihle emission to lower energies. As tle temperuture is increused
uhove 75 K, tle contrihution ol tle µscomponent to tle totul decuy signul increuses
wlile tle intensity ol tle slort lived luminescence decreuses. At room temperuture tle
decuy ol tle visihle emission is ulmost single exponentiul witl u decuy time ol 1.8+ µs,
us wus mentioned helore. 1lis meuns tlut ut low temperutures tle liletime ol tle visihle
emission is mucl slorter tlun ut room temperuture.
Figure 4.5 : Temperature dependence of the emission characteristics of a suspension
of nanocrystalline ZnO particles in ethanol (R=30 Å). A : Ultraviolet emission. B : Visible
emission. The solid circles (upper graphs) represent the integrated intensities and the
open circles (lower graphs) the energetic positions of the maxima of the emission
bands.
0 50 100 150 200 250 0 50 100 150 200 250 300
3.30
3.35
3.40
3.45
0
2
4
6
8
2.05
2.10
2.15
2.20
0
25
50
75
100
A. B.
Temperature (K) Temperature (K)
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
E
n
e
r
g
y

(
e
V
)
E
n
e
r
g
y

(
e
V
)
68
4 . 4 0 | S C U S S | 0 N .
4. 4. I Sf eody-sf of e | um| nescence.
1le sizedependent emission spectru ol ZnO purticles slow hotl u visihle und u
UV emission hund, even lor tle smullest purticles (see ligure +.?,. lreviously, lor ZnO
purticles witl u rudius smuller tlun ?5 A only u visihle emission wus ohserved j5|. 1le
energy ol tle muximum ol tle UV emission hund is close to tle uhsorption onset. 1lis
hund cun he ussigned to tle rudiutive recomhinution ol Wunnier excitons (exciton
emission, wlicl is u very lust process occuring on u suhnunosecond timescule.
Figure 4.6 : Luminescence decay curves of the visible emission from a suspension of
nanocrystalline ZnO particles (R=30 Å) in ethanol at different temperatures. Excitation
is at 4 eV (308 nm).
Time (µs)
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
0.0 2.0 4.0 6.0 8.0
1
10
100
1
10
100
1
10
100 4K
75K
300K
Clupter + . |unocrystulline ZnO lurticles, purt ll
6+
1le visihle emission must involve u deeply trupped clurge currier us tlis hund
is slilted hy uhout 1.5 eV witl respect to tle uhsorption onset. |ext to u deep trup level,
ulso one ol tle hund edges must he involved in tle visihle emission process hecuuse
tle position ol tlis emission hund depends on tle purticle size, us cun he seen in ligure
+.?. ßy unulyzing tle sizedependence ol tle emission hunds, we luve demonstruted in
clupter 8 tlut tle visihle emission is due to u trunsition ol u plotogeneruted electron
lrom u slullow level close to tle conduction hund edge to u deeply trupped lole (u V
O
••
centre,.
65
ln ligure +.? it cun he seen tlut tle intensity ol tle visihle emission hund is
mucl ligler tlun tlut ol tle exciton emission hund. However, tle liletime ol tle visihle
emission is mucl longer tlun tlut ol tle exciton emission. As hotl process ure in
competition witl eucl otler, tle visihle emission process must involve u step in wlicl
tle plotogeneruted lole is trupped elliciently somewlere in tle purticle. 1le rute ol
tlis lole trupping must he mucl luster tlun tle rudiutive recomhinution rute ol tle
exciton emission. ln tle literuture, tlere ure indicutions tlut ellicient loletrupping
occurs ut interluciul stutes in mucrocrystulline ZnO j11|. ßecuuse ol tle lurge surluceto
volume rutio ol our ZnO purticles, ellicient und lust trupping ol plotogeneruted loles ut
surluce sites cun he expected. A good cundidute lor tle trupping ol loles ure O ions ut
?−
tle surluce j1?|. 1lese surluce ions cun trup loles, tlerehy ucting us u kind ol O /O
?− −
redox couple. Lvidence lor tle involvement ol surluce stutes in tle visihle emission
process is provided hy meusurements on tle temperuture und sizedependence ol tle
emission intensity, us will he discussed luter.
4. 4. 2 ¡ | me-r esc| ved | um| nescence.
1le liletime ol tle visihle emission ut room temperuture is 1.8+ µs lor tle uged
purticles (4∼8O A, und O.D? µs lor tle smuller purticles (4∼1O A,. Liletimes ol tle order
ol µs ure typicully ohserved lor hroud hund trup emission in semiconductors j18,1+|. ll
tle rutedetermining step in tle visihle emission process is tle trunsition ol u slullowly
trupped electron to u deeplytrupped lole, tle smull hut signilicunt sizedependence
ol tle emission liletime cun he expluined hy un increuse in wuvelunction overlup
hetween tle clurge curriers. For mucrocrystulline semiconductors it is well known tlut
tle recomhinution rute ol trupped clurge curriers is proportionul to tle squure ol tle
wuvelunction overlup j18,15|. ln tle smuller purticles tle clurge curriers ure closer
togetler, resulting in u lurger wuvelunction overlup, wlicl gives u ligler oscillutor
strengtl. ln tle cuse ol nunocrystulline semiconductor purticles tlis sizedependence ol
tle liletime cun he ohserved us long us tle purticles do not hecome mucl smuller tlun
tle sputiul extension ol tle wuvelunction ol tle slullowly trupped clurge currier. ln tle
limit ol very smull purticle sizes, tle wuvelunction ol tle slullowly trupped clurge
currier is deloculised over tle purticle volume und tle overlup witl tle wuvelunction ol
u deeply trupped clurge currier will not increuse wlen tle purticle hecomes even
smuller. lt is interesting to compure tle present results on nunocrystulline ZnO purticles
to liletime meusurements perlormed on tle trup emission ol nunocrystulline CdS
purticles ol dillerent sizes j18|. For CdS purticles witl u meun diumeter ol ?? A und 88 A
tle sume liletime lus heen ohserved lor tle trup emission. Since purticles ol tlese sizes
Clupter + . |unocrystulline ZnO lurticles, purt ll
66
ure well witlin tle quuntum conlinement regime ol CdS, tlis cuse represents tle
situution ol very smull purticles lor wlicl no sizedependence ol tle wuvelunction
overlup is expected due to u strong conlinement ol tle slullowly trupped clurge
curriers.
1le liletime ol tle exciton emission in tle UV is slorter tlun tle time resolution
ol tle luser setup (∼1O ns,. 1lis is to he expected us u suhnunosecond liletime is
typicully ohserved lor exciton emission in semiconductors. 1le slort liletime is due to
tle ligl oscillutor strengtl ol tle trunsition und to lust nonrudiutive trupping ol clurge
curriers. 1le ohservution tlut ut room temperuture tle exciton emission is only u weuk
hund in tle UV slows tlut tle trupping rute must he mucl ligler tlun tle (lust,
rudiutive decuy rute.
4. 4. 3 ¡ emper of ur e-dependenf | um| nescence.
1le results lrom tle temperuturedependent emission meusurements given in
ligure +.5 slow tlut tle intensity ol hotl tle exciton und tle trup emission increuses
wlen tle temperuture decreuses. 1le increuse ol tle exciton emission is more
pronounced. lt is u generul ohservution tlut tle luminescence elliciency is ligler ut low
temperutures. |onrudiutive putlwuys hecome more prohuhle ut ligler temperutures,
wlile rudiutive recomhinution rutes ure usuully not signilicuntly ullected hy
temperuture. 1lis expluins tle common ohservution ol temperuture quencling.
1le temperuture dependence ol tle luminescence liletime (us illustruted in
ligure +.6, is peculiur . ut low temperutures (1<5O K, tle liletime ol tle luminescence is
slorter tlun ut ligler temperutures. Usuully tle opposite is ohserved . u longer liletime
ut low temperutures hecomes slorter ut ligler temperutures due to extru (non
rudiutive, decuy putls wlicl require tlermul uctivution. An explunution lor tle
ohserved heluviour involves u temperuturedependent populution ol u distrihution ol
slullow electron trups neur tle conduction hund edge. At low temperutures, tle most
slullow trups lrom tlis distrihution ure signilicuntly occupied hy electrons. 1le
wuvelunction overlup ol tle deloculised slullowly trupped electron und tle deeply
trupped lole is relutively lurge, resulting in u slort liletime (?75 ns,. Ahove uhout 5O K
tle slullow electron trups sturt to he tlermully depopuluted. 1lis results in u decreuse
ol tle emission intensity since tle electrons cun now reucl tle surluce und recomhine
nonrudiutively (see ulso helow,. Alternutively, tle electrons muy he trupped uguin in
somewlut deeper trups und recomhine witl tle deeply trupped loles. 1lis results in
emission ut u sligltly dillerent energy tlun tle emission ohserved ut lower temperutures,
expluining tle slilt to lower energies ol tle emission muximum hetween 5O und ?OO K.
67
Since tle electron wuvelunction is less deloculised in tle sligltly deeper electron trup,
tle overlup witl tle wuvelunction ol tle deeply trupped lole is reduced resulting in un
increuse ol tle liletime. 1le tlermul detrupping ol tle slullowly trupped electrons
occurs hetween 5O K und ?OO K und in tlis temperuture runge tle trup emission slilts to
lower energies. 1le decuy curve ol tle emission is twoexponentiul in tlis temperuture
runge . hotl tle slort liletime component (recomhinution ol tle slullowly trupped
electrons, und tle long lile time component ure ohserved. 1le contrihution ol tle slort
decuy component (?75 ns, decreuses hetween 5O und ?OO K, wlile tle contrihution ol
tle long decuy component (∼1 µs, increuses. Ahove ?OO K only tle 'slow' µs decuy is
ohserved, indicuting tlut detrupping is complete. A similur explunution lus heen used
helore to expluin temperuturedependent luminescence meusurements on colloidul
suspensions ol CdS purticles j16|.
ln u study on tle luminescence ol nunocrystulline CdS purticles temperuture
quencling ol tle CdS trup emission wus ulso ohserved, wlere it wus expluined hy u
multiplonon reluxution process j18|. ln tle cuse ol tle trup emission lrom
nunocrystulline ZnO purticles, multiplonon reluxution cunnot expluin tle temperuture
quencling ol tlis emission hund. 1le hest evidence comes lrom tle temperuture
dependent liletime meusurements us discussed uhove. Furtlermore, tle energy gup ol ?
eV is too ligl compured to tle energy ol opticul plonons in ZnO (O.O5O.O7 eV j17|,.
4. 4. 4 Nof ur e cf f he deep hc| e f r op.
|ow tlut tle luminescence properties ol nunocrystulline ZnO purticles luve
heen discussed us u lunction ol purticle size und temperuture, u more quuntitutive
model is proposed. 1le nuture ol tle deep trup involved in tle visihle emission is still
un importunt issue, us discussed in tle introduction. ln clupter 8 we luve slown tlut
tle visihle emission is due to recomhinution ol u slullowly trupped electron witl u
deeply trupped lole. 1o he more specilic uhout tle nuture ol tle deep trup tle delect
clemistry ol ZnO is considered. ZnO cun luve severul types ol delects sucl us
interstitiul utoms or vucuncies, hotl unionic und cutionic j18|. LlR studies luve slown
tlut oxygen vucuncies contuining one electron (V , ure tle predominunt purumugnetic
O

delects j?+|. ln mucrocrystulline ZnO tlese delects ure represented hy u level
upproximutely ? eV helow tle conduction hund edge j18| und tley ure olten ussumed
to he tle recomhinution centres lor tle visihle emission in ZnO. ln some puhlicutions
tle visihle emission is represented us tle recomhinution ol un electron lrom tle
conduction hund witl tle V centre (wlicl lus un ellective monovulent positive clurge
O

witl respect to tle regulur O site, j1+|. Alter tlis recomhinution tle ellectively neutrul
?−
Clupter + . |unocrystulline ZnO lurticles, purt ll
68
V centre is lormed. 1le V centre lus un energy very close to tle conduction hund
O O
× ×
edge und ut room temperuture ulmost ull V centres ure tlermully dissociuted into V
O O
× •
centres und conduction hund electrons. 1le slilt lrom ? eV helow tle conduction hund
edge lor V to ¦ust helow tle conduction hund edge lor V is due to tle correlution
O O
• ×
energy ol tle two electrons in tle oxygen vucuncy j18|. A trunsition ol un electron lrom
tle conduction hund to u V level cun tlerelore not yield plotons witl un energy ol ?
O

eV, us sucl u trunsition ellectively tukes pluce hetween tle conduction hund edge und
tle V level. 1lis meuns tlut u meclunism in wlicl V is tle recomhinution centre lor
O O
× •
tle visihle emission ol ZnO is not correct. Recomhinution ol u conduction hund
electron witl u V centre (un oxygen vucuncy contuining no electrons, luving un
O
••
ellective divulent positive clurge witl respect to tle normul O site, cun yield plotons
?−
witl un energy ol uhout ? eV. Sucl V centres cun he lormed wlen u lole is trupped ut
O
••
u V centre. ln tle model presented helow tle visihle emission is ussigned to tle
O

recomhinution ol u slullowly trupped electron witl u deeply trupped lole in u V
O
••
centre.
4. 4. 5 Mcde| .
1le dependence ol tle luminescence intensities on temperuture und purticle
size indicutes tlut tle purticle surluce pluys u role in tle process leuding to tle visihle
emission. ln view ol tle lurge surluce ureu ol tle purticles tlis is to he expected. ln tle
preceding text it lus heen urgued tlut tle visihle emission results lrom u trunsition tlut
tukes pluce in tle hulk ol tle ZnO purticle und tlut tle recomhinution centre is u V
O
••
delect wlicl trups un electron lrom u level close to tle conduction hund edge. ln ligure
+.7, u sclemutic overview is presented ol tle reluxution processes ol u plotoexcited
ZnO purticle. ln tlis ligure, tle hund edges ure slown us well us u deep trup level
(V /V , und tle energy distrihution ol u O /O surluce system. For reusons ol simplicity,
O O
• •• − − ?
tle slullow trupping ol u plotogeneruted electron is not slown. A trunsition is indicuted
witl un urrow und represented hy tle letter 6, luving u suhscript tlut specilies tle
trunsition. 1lis suhscript consists ol tle initiul und linul stute ol tle electron. 1lese
stutes ure C (conduction hund,, V (vulence hund,, S (surluce, und 1 (deep trup,. |on
rudiutive recomhinution is denoted witl tle suhscript |R. ln principle, tlis cun occur ut
tle surluce (S, or ut quencling centres in tle purticle. ln view ol tle lurge surluce ureu
ol tle ZnO purticles, only nonrudiutive recomhinution ut tle surluce is considered.
Figure 4.7 : A schematic overview of the relaxation processes that take place upon
photoexcitation of a ZnO particle. The bandedges are shown as well as a deep trap
level in the bulk of the particle. At the surface of the particle, an energy distribution of a
O
2
/O system is shown. The arrows indicate a transition that is represented by the letter
T. The subscript to this letter specifies the transition in a way that it contains the initial
and final state of the electron. These states can be the conduction band (C), the
valence band (V), the trap (T) or the surface (S). It is assumed that non-radiative
recombination (NR) only occurs at the particle surface. A-C : Three competitive
processes on the level of the exciton : exciton emission and trapping of either of the
charge carriers at the surface. D-E : possible processes following surface trapping of a
hole. G : Visible emission, following the tunneling of the surface-trapped hole back into
the particle. H : Non- radiative recombination, which can be the result of either of two
different processes : trapping of a hole at the surface followed by trapping of an
electron at the surface (B and E) or trapping of an electron at the surface followed by
trapping of a hole at the surface (C and F).
O
S
2
_
O
S
2
_
O
S
2
_
O
S
2
_
O
S
_
T
CV
T
SV
O
S
2
_
T
CS
O
S
_
O
S
_
O
S
_
T
TS
T
CS
T
CT
T
NR
hν (UV)
hν (Vis)
A.
B. D. G.
H.
E.
F. C.
v
o
v
o
v
o
v
o
v
o
v
o
v
o
v
o
,
T
SV
,
6D
+8
+8
+8 58 +5
6
2
6 6 6
=
+ +
58
58
+8 58 +5
6
2
6 6 6
=
+ +
+5
+5
+8 58 +5
6
2
6 6 6
=
+ +
Clupter + . |unocrystulline ZnO lurticles, purt ll
7O
(+.1u,
(+.1h,
(+.1c,
1le tlree competitive processes on tle excitonic level slown in ligure +.7AC
ure . rudiutive recomhinution witl trunsition rute 6 , trupping ol tle lole ut tle surluce
CV
witl rute 6 (wlile tle electron remuins in tle conduction hund,, und trupping ol tle
SV
electron ut tle surluce 6 (wlile tle lole remuins in tle vulence hund,. 1le
CS
prohuhilities (2, lor tlese tlree processes ure .
1le rutes lor hotl surluce trupping processes (6 und 6 , decreuse us tle
SV CS
purticle size increuses since tle surlucetohulk rutio decreuses. 1le trunsition rute ol tle
exciton recomhinution (6 , will not he inlluenced strongly hy tle purticle size und tlus
CV
tle intensity ol tle exciton emission will increuse witl increusing purticle size, us is
slown in ligure +.?.
On tle riglt ol ligure +.7AC, tle possihle consecutive reluxution steps ure
slown. 1le trunsition 6 returns tle ZnO purticle to its ground stute so it lus no
CV
consecutive step. 1rupping ol tle lole ut tle surluce (ligure +.7ß, cun luve eitler ol two
possihle consecutive steps. One possihility is tlut tle surlucetrupped lole tunnels huck
into tle purticle to recomhine witl un electron in u deep trup (6 , slown in ligure
1S
+.7D,. 1le otler possihility is tlut u plotogeneruted electron wlicl is still in tle
conduction hund gets trupped ut tle surluce und recomhines witl tle surlucetrupped
lole (6\ , ligure +.7L, tle uccent is used to indicute tlut tlis process is not identicul to
CS
6 slown in ligure +.7C,. Wlile tle lormer step (ligure +.7D, leuds to tle trunsition
CS
responsihle lor tle visihle emission (6 , ligure +.7G,, tle lutter (ligure +.7L, results in
C1
nonrudiutive recomhinution (6 , ligure +.7H,. Wlen tle lirst step in tle reluxution
|R
process is tle surlucetrupping ol un electron (ligure +.7C,, tlis will ulwuys result in non
rudiutive recomhinution (6 , ligure +.7H, viu surlucetrupping ol u lole (6\ , ligure
|R SV
+.7F,.
1le tlree reluxution processes tlut return u plotoexcited ZnO purticle to its
ground stute ure rudiutive exciton recomhinution (UV emission, 6 ,, rudiutive trup
CV
+8
+8
+8 58 +5
6
2
6 6 6
=
+ +
65
+6 58
65 +5
6
2 2
6 6
= ⋅
′ +
+5
4 +5 58
65 +5
6
2 2 2
6 6

= + ⋅
′ +
71
(+.?u,
(+.?h,
(+.?c,
recomhinution (visihle emission, 6 , und nonrudiutive recomhinution ut tle surluce
C1
(6 ,. 1le prohuhilities lor tlese tlree processes ure given hy .
|R
lt cun eusily he seen tlut tle sum ol equutions (?u,, (?h, und (?c, is equul to unity.
1le model us presented in tlis clupter implies tlut hotl tle tunneling rute ol u
surlucetrupped lole to u V centre (6 , us well us tle trupping rute ol u conduction
O 1S

hund electron ut tle surluce (6 und 6' , decreuse witl increusing purticle size. 1le
CS CS
tunneling rute decreuses more strongly since tunneling tukes pluce hetween two
loculised stutes wlile tle surluce trupping ol tle conduction hund electron (leuding to
nonrudiutive reluxution, involves u deloculised stute. Also, tle delect concentrution
(including V centres, muy decreuse wlen tle size ol tle purticles increuses, wlicl lus
O

heen put lorwurd helore hused on studies ol tle sizedependence ol tle dissolution rute
ol nunocrystulline ZnO purticles j1D|. ßotl ellects result in u decreuse ol tle prohuhility
lor tle trup emission compured to tle prohuhility lor nonrudiutive decuy ut tle surluce.
Finully, lor tle higger purticles ulso nonrudiutive decuy in tle hulk ol tle purticle muy
hecome importunt und contrihute to tle totul loss hy nonrudiutive processes. As u
result tle intensity ol tle trup emission decreuses witl increusing purticle size in
ugreement witl tle experimentul results (see ligures +.? und +.8,.
ßotl tle trup emission und tle exciton emission slow tlermul quencling. 1le
quencling ol tle exciton emission is more pronounced. ln tle cuse ol nunocrystulline
ZnO purticles nonrudiutive reluxution predominuntly tukes pluce ut tle surluce. |on
rudiutive reluxution ut surluce sites is generully ohserved und hecuuse ol tle lurge
surlucetovolume rutio ol nunopurticles tle surluce will pluy un importunt role in tle
quencling ol tle luminescence. 1le quencling ut tle surluce involves trupping ol hotl
plotogeneruted clurge curriers ut tle surluce. Witl increusing temperutures tle
prohuhility lor surluce trupping ol clurge curriers increuses. Similur to redox couples,
surluce systems sucl us O /O luve u certuin energy distrihution due to solvution
?− −
Clupter + . |unocrystulline ZnO lurticles, purt ll
7?
ellects. 1le widtl ol tlis distrihution depends on tle temperuture . tle ligler tle
temperuture, tle hrouder tle energy distrihution. 1le lurger overlup in energy hetween
tle conduction hund or vulence hund und tle surluce stutes ut ligler temperutures will
leud to ligler trupping rutes 6 /6\ und 6 /6\ . 1le exciton emission is ullected hy hotl
SV SV CS CS
trupping processes und tlerelore tle temperuture quencling ol tlis emission is more
pronounced. For tle quencling ol tle trup emission, trupping ol u plotogeneruted
electron ut tle surluce is importunt. ßesides tle increuse in overlup hetween tle
conduction hund stutes und tle surluce stutes, tlermul detrupping ol u slullowly
trupped electron will occur ut ligler temperutures und tle prohuhility tlut un electron
is trupped ut surluce sites ut ligler temperutures increuses, resulting in quencling ol tle
trup luminescence.
1le model contuining tle possihle reluxution processes ol u plotoexcited
nunocrystulline ZnO purticles us slown in ligure +.7 cun expluin tle ohservutions
quulitutively. 1le identilicution ol tle visihle luminescence us tle recomhinution ol u
conduction hund electron witl u deeply trupped lole ut u V centre is in ugreement
O
••
witl previous work on mucrocrystulline ZnO wlicl lus provided evidence lor u role ol
oxygen vucuncies in tle visihle luminescence ol ZnO, ultlougl in tlese puhlicutions V
O

centres were incorrectly ussumed to he tle recomhinution centres. ln mucrocrystulline
ZnO ulso direct trupping ol loles hy V centres will occur. ln nunocrystulline ZnO
O

purticles un importunt role ol tle surluce cun he expected. Lvidence lor tle role ol lust
surlucetrupping ol loles in tle trup emission process comes lrom tle sizedependence
ol tle overull emission intensities und tle relutive intensities ol tle exciton und visihle
emission us presented in tlis clupter. 1le model is supported hy previous work on tle
luminescence ol nunocrystulline ZnO purticles wlere it wus ohserved tlut ions like Fe

j5| und l j1+| wlicl cun uccept loles tlut ure trupped ut tle surluce ol tle purticle

ure ellicient quenclers ol tle visihle luminescence.
78
4 . 5 C 0 N C L U S | 0 N .
1emperuturedependent steudystute und timeresolved luminescence
meusurements were perlormed on suspensions ol nunocrystulline ZnO purticles ol
dillerent sizes. All tle suspensions slow two emission hunds, u relutively weuk und
slurp UV hund wlicl cun he ussigned to exciton emission und u more intense und
hroud emission hund in tle visihle purt ol tle spectrum, slilted hy upproximutely 1.5 eV
witl respect to tle uhsorption onset.
A model lor tle kinetics ol tle rudiutive und nonrudiutive processes in
nunocrystulline ZnO purticles is proposed hused on tle identilicution ol tle trunsition
responsihle lor tle visihle emission us heing u recomhinution ol u slullowlytrupped
electron witl u deeplytrupped lole. From tle purticle size dependence und tle
temperuture dependence ol tle emission properties it is concluded tlut tle
plotogeneruted lole is trupped ut u surluce system (prohuhly O /O ,. 1le surluce
?− −
trupped lole cun tunnel huck into tle purticle wlere it recomhines witl un electron in
un oxygen vucuncy (V , resulting in tle ceution ol u V centre, tle recomhinution
O O
• ••
centre lor tle visihle emission. 1le dependence on purticle size ol tle prohuhility lor
tlis tunneling process is mucl stronger tlun tlut ol tle nonrudiutive processes. 1lis
results in un increuse ol tle visihle emission intensity us tle size ol tle ZnO purticles
decreuses.
Clupter 4 . Nunocrystulline ZnO lurticles, purt ll
74
4 - . - 4 - + - 5
j1| S. Slionoyu, W.V. Yen (editors,, llosplor Hundhook, CRC lress LCC, ßocu Ruton 1DDD, p. 255.
j2| V. Anpo, Y. Kuhokuwu, 1. //ys. C/cn. 88 (1D84, 5556.
j8| K. Vunleusden, C.H. Seuger, W.L. Wurren, D.R. 1ullunt, 1.A. Voigt, Aµµ/. //ys. Lc//. 68(8, (1DD6,
4O8.
j4| K. Vunleusden, W.L. Wurren, C.H. Seuger, D.R. 1ullunt, 1.A. Voigt, ß.L. Gnude, 1. Aµµ/. //ys. 7D(1,
(1DD6, 7D88.
j5| D.W. ßulnemunn, C. Kormunn, V.R. Hollmunn, 1. //ys. C/cn. D1 (1D87, 878D.
j6| L. Spunlel, V.A. Anderson, 1. An. C/cn. Soc. 118 (1DD1, 2826.
j7| L.A. Veulenkump, 1. //ys. C/cn. 8 1O2(2D, (1DD8, 5566.
j8| L.V. Wong, 1.L. ßonevicl, l.C. Seurson, 1. //ys. C/cn. 8 1O2 (1DD8, 777O.
jD| V. Huuse, H. Weller, A. Henglein, 1. //ys. C/cn. D2 (1D88, 482.
j1O| G.H. Ldstrup 1ensen, //ys. S/o/. So/. (h) 64 (1D74, K51.
j11| S.R. Vorrison, L/cc/roc/cn|s/ry o/ Scn|conJoc/or onJ Ox|J|:cJ Mc/o/ L/cc/roJcs, llenum lress, First
Ldition, New York 1D8O, p. 227288.
j12| G.H. Scloenmukers, D. Vunmuekelhergl, 1.1. Kelly, 1. //ys. C/cn. 1OO (1DD6, 8215.
j18| N. Clestnoy, 1.D. Hurris, R. Hull, L.L. ßrus, 1. //ys. C/cn. DO (1D86, 88D8.
j14| l.V. Kumut, ß. lutrick, 1. //ys. C/cn. D6 (1DD2, 682D.
j15| ß. Henderson, G.F. lmhuscl, Oµ/|co/ Sµcc/roscoµy o/ lnorgon|c So/|Js, Clurendon lress, First
Ldition, Oxlord 1D8D.
j16| A. Lyclmüller, A. Husselhurtl, L. Kutsikus, H. Weller, 8cr. 8onscngcs. //ys. C/cn. D5 (1DD1, 7D.
j17| O. Vudelung (ed.,, LonJo//8orns/c|n, Noncr|co/ Jo/o onJ /onc/|ono/ rc/o/|ons/|µs |n sc|cncc onJ
/cc/no/ogy. Vo/onc lll17. Scn|conJoc/ors, Springer Verlug. ßerlin, 1D88, 17h pp. 85115.
j18| F.A. Kröger, T/c C/cn|s/ry o/ lnµcr/cc/ Crys/o/s, NortlHollund luhlisling Compuny, First Ldition,
Amsterdum 1D64, p. 6D1.
j1D| L.A. Veulenkump, 1. //ys. C/cn. 8 1O2 (1DD8, 7764.
75
76
77
C H A P ¡ £ k
N A N 0 C k ¥ S ¡ A L L | N £
Z N 0 P A k ¡ | C L £ S
5
| | | . ¡ he | nf | uence cf Adscr bed 0xygen
cn f he £ m| ss| cn Pr cper f | es
A 8 S ¡ k A C ¡ .
1le tlird und linul clupter on tle opticul properties ol nunocrystulline
ZnO purticles presents tle results ol studies on tle inlluence ol electron
scuvengers (sucl us oxygen, on tle emission properties. Lmission
meusurements ure perlormed on deueruted suspensions ol ZnO
purticles in ulcolols (2propunol und etlunol, und otler orgunic
solvents (DVF und propylene curhonute,. Upon UV irrudiution tle
visihle emission quencles wlile tle intensity ol tle exciton emission
increuses. Admission ol oxygen restores tle initiul emission properties.
Using tle meclunism ol tle visihle emission process us developed in
clupters 8 und 4, tle results us presented in tlis clupter ure expluined
hy clurging ol tle ZnO purticles witl electrons. 1le presence ol excess
electrons on ZnO purticles results in tle removul ol V centres wlicl
O

ure involved in tle visihle emission.
Clupter ¯ . |unocrystulline ZnO lurticles, purt lll
78
5 . I | N ¡ k 0 0 U C ¡ | 0 N .
Due to tle ligl surlucetovolume rutio ol nunocrystulline semiconductor
purticles tle nuture ol tle surluce is very importunt wlen considering tle properties ol
sucl purticles. 1le surluce is u strong perturhution ol tle luttice wlere u ligl
concentrution ol hotl slullow und deep levels provide putlwuys lor nonrudiutive
recomhinution ol plotogeneruted clurge curriers j1|. 1lis expluins wly tle quuntum
elliciency ol tle emission ol nunocrystulline semiconductor purticles is generully lur
helow unity und is strongly dependent on surluce pussivution. However, surluce stutes
cun ulso he ol importunce lor rudiutive trunsitions, us we luve demonstruted lor ZnO in
clupter 4.
|ext to tlese intrinsic surluce properties it is known tlut surluce udsorhutes
sucl us reuction hyproducts, solvent molecules or dissolved guses cun inlluence
vurious properties ol u colloidul suspension ol semiconductor purticles. For instunce,
using zinc percllorute insteud ol zinc ucetute yields u less stuhle ZnO suspension j2|. lt
wus suggested tlut weukly hound ucetic ucid pluys tle role ol u stuhilizer j8|. Surluce
udsorhutes cun ulso inlluence tle opticul properties ol u suspension ol semiconductor
purticles, viz. hy scuvenging clurge curriers tlut ure creuted upon plotoexcitution j4|.
1le plotogeneruted loles cun he removed hy oxidution ol solvent molecules wlile
udsorhed oxygen molecules cun scuvenge plotogeneruted electrons. 1le inlluence ol
surluce udsorhutes, more purticulurly solvent molecules und oxygen, on tle opticul
properties ol suspensions ol ZnO purticles will he tle muin topic ol tlis clupter.
lt lus heen ohserved tlut u deueruted suspension ol ZnO purticles in metlunol
or 2propunol exlihits remurkuhle emission properties. A lreslly prepured sumple
slows two types ol emissions under UV irrudiution. One is u relutively weuk und nurrow
emission hund in tle UV due to exciton recomhinution wlile tle otler is u strong und
hroud emission hund in tle visihle purt ol tle spectrum. As expluined in clupter 8, tle
visihle emission cun he ussigned to tle recomhinution ol un electron lrom u level close
to tle conduction hund edge witl u deeply trupped lole. UV irrudiution (using plotons
witl un energy ligler tlun tle hundgup, ol u deueruted suspension ol ZnO purticles
quencles tle trup emission wlile tle intensity ol tle exciton emission sligltly
increuses. Simultuneously, u slilt ol tle uhsorption onset to ligler energies wus
ohserved j¯|. 1le initiul emission und uhsorption properties ure restored some time
ulter tle UV irrudiution is stopped or ulter tle udmission ol uir.
First, tlese ohservutions were expluined hy ussuming u plotodissolution ol tle
ZnO purticles j¯|. ln tlis model, Zn is lormed ut tle purticle surluce wlicl quencles
÷
tle trup emission hy reucting witl plotogeneruted loles to lorm Zn . Furtlermore, tle

7D
slilt ol tle uhsorption onset wus tlouglt to he u quuntum size ellect us u result ol tle
decreuse in size ol tle ZnO purticles. ln tle meun time, compuruhle plotodissolution
meclunisms luve heen descrihed muinly lor sulplidic semiconductor purticles (see
clupter 2,. However, in tle cuse ol tle ohserved heluviour lor ZnO purticles, tle
explunution ol tle slilt ol tle uhsorption onset und tle quencling ol tle visihle
emission hy u plotodissolution meclunism wus re¦ected ulter it wus recognised tlut tle
presence ol even u single excess electron on u semiconductor purticle cuuses u slilt ol
tle uhsorption onset to ligler energies j6|. 1lis wus expluined hy u model hused on
tle Sturk ellect in semiconductors. 1le excess electron is treuted us u loculised point
clurge producing u strong electric lield in tle semiconductor purticle. ln tlis wuy, tle
energy lor tle lowest exciton stute is increused. 1le hlueslilt ol tle uhsorption onset
cun ulso he expluined hy tle ßursteinVoss ellect in wlicl excess electrons lill tle
lowest stutes ol tle conduction hund j7D|. ln tlis wuy, suhsequent uhsorption requires
plotons ol ligler energies in order to uccess tle lowest empty stutes.
An ulternutive model lor tle quencling ol tle trup emission ol ZnO purticles
wus proposed wlicl wus hused on un oxygenmediuted sluttle meclunism j2|. ln tlis
meclunism, plotogeneruted electrons ure scuvenged hy udsorhed oxygen molecules
tlut suhsequently trunsler tlese electrons to deep trups. Recomhinution ol tle deeply
trupped electrons witl trupped loles gives tle visihle emission. Additionul experiments
were perlormed on colloidul suspensions ol ZnO in wlicl O wus lormed hy pulse
2

rudiolysis j1O|. ll tle sluttle meclunism is correct, un emission slould he ohserved
wlen O trunslers un electron to u ZnO purticle. However, no emission wus ohserved.
2

ln tlis clupter, tle results ol studies on tle inlluence ol oxygen (us electron
scuvenger, on tle emission properties ol nunocrystulline ZnO purticles ure presented.
Lmission meusurements ure perlormed helore und ulter suturuting u suspension witl
nitrogen und tle inlluence ol UV irrudiution is studied. Secondly, tle inlluence ol
udsorhed lole scuvengers on tle emission properties is investiguted hy using dillerent
orgunic solvents.
5 . 2 £ X P £ k | M £ N ¡ A L M £ ¡ H 0 0 S .
5. 2 . I Somp| e pr epor of | cn.
Suspensions ol nunocrystulline ZnO purticles cun he prepured in orgunic
solvents sucl us etlunol or 2propunol. For tle prepurution in 2propunol j2|, 2¯ ml ol u
O.O2 V |uOH solution were slowly udded wlile stirring to 22¯ ml ol u O.OO1 V
Clupter ¯ . |unocrystulline ZnO lurticles, purt lll
8O
Zn(CH COO, 2H O solution, ulter hotl solutions were lirst cooled to O °C. 1le
8 2 2
.
prepurution in etlunol is very similur j11,12|. ln tlis cuse, ¯O ml ol u O.14 V LiOH H O
.
2
solution (prepured using un ultrusonic hutl, were udded to ¯O ml ol u O.1 V
Zn(CH COO, 2H O solution. Aguin, hotl solutions were lirst cooled to O °C helore tle
8 2 2
.
lydroxide solution wus udded slowly to tle zinc solution wlile stirring. For tle
prepurution in 2propunol, 1LVmeusurements luve slown tlut tle meun purticle
rudius ol tle lully grown purticles is uhout 8O A (see clupter 2,. 1lis corresponds to
purticles tlut contuin upproximutely ¯OOO moleculur ZnO units. 1le prepurution in
etlunol yields purticles ol u similur size hut tle concentrution ol ZnO purticles is uhout
two orders ol mugnitude ligler tlun lor tle prepurution in 2propunol. Upon uddition ol
lexune to tle suspension in etlunol (volume rutio lexune to ZnO sol ol 2.1, tle ZnO
purticles precipitute j12|. 1le ZnO purticles cun he sepuruted lrom tle mixture hy
centriluging lollowed hy decuntution ol tle supernutunt. 1le precipituted ZnO purticles
cun he resuspended in unotler solvent. 1o muke sure tlut none ol tle initiul solvent
wus still present in tle new suspension, tle ZnO purticles were curelully dried in u
nitrogen utmosplere. ln tlis wuy, suspensions ol ZnO purticles were prepured in
propylene curhonute (unlydrous DD.7º, und |,|dimetlyllormumide (DVF, DD.8º,.
5. 2 . 2 0pf | co| meosur emenf s.
1le plotoluminescence meusurements were perlormed on u SlLX Fluorolog
spectrolluorometer model F2OO2 equipped witl two douhlegruting O.22 m SlLX 168O
monoclromutors und u 4¯O W xenon lump us tle excitution source. 1le emission
spectru were corrected lor tle spectrul response ol tle emission monoclromutor und
tle lV tuhe. UV irrudiution ol tle suspensions wus curried out inside tle
spectrolluorometer using tle excitution heum. 1le suspension wus meusured in u
quurtz cuvet ol dimensions 1.1.4 cm (widtl.deptl.leiglt,. 1le excitution heum lud u
rectungulur slupe witl u leiglt ol uhout 1 cm und u widtl ol upproximutely 1 mm. 1le
penetrution deptl ol tle UV rudiution wus sucl tlut uhsorption took pluce over tle
entire widtl ol tle cuvet.
5 . 3 k £ S U L ¡ S .
Wlen u suspension ol nunocrystulline ZnO purticles in 2propunol is suturuted
witl nitrogen, tle emission properties do not clunge mucl us cun he seen in ligure ¯.1
wlen compuring spectrum (u, to spectrum (h,. ln hotl cuses two emission hunds ure
Figure 5.1 : Room temperature emission spectra of a suspension of nanocrystalline
ZnO particles in 2-propanol upon excitation with 4.3 eV. Φ
E
denotes the photon flux
per constant energy interval. The emission spectra are plotted with an offset relative to
each other. Spectrum (a) is taken before, and spectrum (b) after saturating the
suspension with nitrogen. Spectrum (c) is taken after the nitrogen-saturated
suspension has been illuminated with UV radiation (4.3 eV) for 75 min.
1.0 2.0 3.0 4.0
(b)
(c)
(a)
Energy (eV)
Φ
E
81
ohserved. 1le relutively weuk und nurrow emission hund ut 8.4 eV is due to tle
rudiutive unnililution ol excitons (exciton emission,. 1le intense hroud emission hund
ut 2.2 eV is ussigned to u trunsition ol u plotogeneruted electron lrom u slullow level
close to tle conduction hund to u deeply trupped lole (see clupter 8,. 1le emission
properties clunge wlen tle nitrogensuturuted suspension is illuminuted witl UV
rudiution. From ligure ¯.1 it cun he seen tlut upon UV irrudiution lor 7¯ min tle trup
emission hund lus heen quencled wlile tle intensity ol tle exciton emission hund lus
increused. Furtlermore, tle exciton emission hund lus slilted to lower energies und
tle trup emission to ligler energies, hotl hy uhout 8O meV. 1le FWHV ol tle exciton
emission hund lus decreused to uhout 6Oº ol tle initiul vulue wlile tle widtl ol tle
trup emission hund did not clunge signilicuntly. 1urning oll tle rudiution or udding
oxygen restores tle initiul luminescence properties.
1o investigute tle quencling ol tle trup emission in more detuil we luve
meusured tle intensity ol tlis emission ut 2.2 eV us u lunction ol time wlile illuminuting
witl UV rudiution. 1le results ol tlis meusurement ure slown in ligure ¯.2. Alter uhout
8O minutes tle intensity ol tle trup emission lus reucled u constunt vulue ol uhout 1Oº
ol tle initiul vulue. Wlen tle UV irrudiution is interrupted lor u period ol time hy using u
slutter, tle emission intensity recovers. 1le degree ol recovery depends on tle time
tlut tle suspension is kept in tle durk. From ligure ¯.2 it cun he seen tlut keeping tle
suspension in tle durk lor uhout ¯ minutes results in u lull recovery ol tle intensity ol
Figure 5.2 : Intensity of the trap emission at 2.2 eV as a function of time while
illuminating with UV radiation (4.3 eV). The measurement is performed at room
temperature using a suspension of nanocrystalline ZnO particles in 2-propanol which
has first been saturated with nitrogen. During the periods t
1
(30 sec), t
2
(60 sec), and t
3

(300 sec) the suspension is kept in the dark.
Time (sec)
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
0 1000 2000 3000 4000
0.0
0.5
1.0
t
1
t
2
t
3
Clupter ¯ . |unocrystulline ZnO lurticles, purt lll
82
tle trup emission. Following sucl u recovery, tle intensity decreuses mucl luster upon
UV irrudiution tlun lor tle lreslly prepured sumple. Witlin 1O seconds tle intensity ol
tle trup emission decreuses to 2¯º und u constunt intensity (∼1Oº, is reucled in less
tlun u minute. For tle exciton emission tle opposite heluviour is ohserved. Figure ¯.8
slows in more detuil tle lust quencling ol tle trup emission und tle lust increuse ol tle
intensity ol tle exciton emission upon UV irrudiution.
1le inlluence ol tle intensity ol tle UV heum on tle quencling rute ol tle trup
emission wus studied hy using lilters witl u known trunsmission ut tle UV wuvelengtl.
ßelore tle meusurements were perlormed, tle suspension wus lirst suturuted witl
nitrogen, illuminuted witl UV rudiution until tle trup emission intensity lus reucled its
minimum intensity ulter wlicl it wus kept in tle durk lor u lew minutes to ullow tle trup
emission to recover. Figure ¯.4 slows tle results ol tlese meusurements.
Similur experiments us descrihed uhove were perlormed on suspensions ol
nunocrystulline ZnO purticles in etlunol, propylene curhonute und DVF. 1le
suspension in etlunol heluved completely identicul to tle suspension in 2propunol.
For tle otler two suspensions, tle intensity ol tle visihle emission during UV irrudiution
is slown in ligure ¯.¯A. ln generul, tle visihle emission intensity decreuses to u constunt
vulue ulter u certuin time. For propylene curhonute und DVF successively, tlis constunt
vulue is 7¯º und 4Oº ol tle initiul intensity. ßelore tlis constunt vulue is reucled, tle
clunge ol tle visihle emission intensity witl time slows u dillerent heluviour lor eucl
Figure 5.3 : Intensity of the exciton emission band at 3.4 eV (upper graph) and of the
trap emission band at 2.2 eV (lower graph) as a function of time while illuminating
with UV radiation (4.3 eV). The experiment is performed on a suspension that has first
been saturated with nitrogen and illuminated with UV radiation for 45 minutes. During
the periods t
1
(30 sec) and t
2
(60 sec) the suspension is kept in the dark.
Time (sec)
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
0 100 200 300 400 500
0.0
0.5
1.0
0.0
0.5
1.0
t
1
t
2
t
1
t
2
Figure 5.4 : A : Intensity of the trap emission at 2.2 eV as a function of time while
illuminating with UV radiation (4.3 eV). Filters with varying transmission at 4.3 eV were
placed in the UV beam : (a) no filter, (b) 33%, and (c) 14% transmission. For each of
the three graphs, the intensity at t=0 was set to unity and the constant intensity at high t
to zero. B : Time at which the trap emission intensity has decreased to 20% of its initial
value (see dashed line in A) versus the transmission of the filter at 4.3 eV. The solid line
is a guide to the eye.
(b)
(c)
(a)
Time (sec)
T
i
m
e

(
s
e
c
)
Transmission (%)
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
0 50 100 150 200
0.01
0.1
1
0 20 40 60 80 100
0
50
100
150
200
A. B.
88
Clupter ¯ . |unocrystulline ZnO lurticles, purt lll
84
ol tle solvents. lt lus to he remurked tlut tle suspensions in propylene curhonute und
DVF slowed some liglt scuttering. A complete redispersion ol precipituted colloidul
purticles cun he dillicult due to clustering ol tle purticles. For etlunol us u solvent,
experiments were perlormed on u suspension directly ulter prepurution (no liglt
scuttering, und on u suspension ulter redispersion ol tle purticles (liglt scuttering,.
During tle lirst period ol UV irrudiution (directly ulter suturuting tle suspension witl
nitrogen,, tle intensity ol tle visihle emission decreused ut similur rutes lor hotl
etlunolic suspensions. 1lis rute wus ulso similur to tlut ol tle suspension in 2propunol.
However, tle constunt vulue tlut wus reucled eventuully, wus ligler lor tle suspension
tlut slowed liglt scuttering. Alter keeping tle suspensions in tle durk lor u couple ol
minutes, hotl tle trunspurent und tle nontrunspurent suspension heluved identicully,
und uguin similur to tle suspension in 2propunol. For tle otler solvents (propylene
curhonute und DVF,, tle results ol tlese meusurements ure presented in ligure ¯.¯ß. ln
tle cuse ol propylene curhonute us u solvent, tle intensity ol tle visihle emission
quencles only very slowly. For DVF, u similur quencling heluviour is ohserved us lor
etlunol und 2propunol. Witl respect to tle umount ol time it tukes lor tle visihle
emission intensity to recover completely, dillerences were ohserved lor tle vurious
solvents. lt wus ulreudy mentioned tlut lor u suspension in 2propunol (und etlunol,,
tle visihle emission intensity recovers witlin uhout ¯ minutes. For DVF us u solvent,
keeping tle suspension in tle durk lor 8O minutes only leuds to u recovery ol ¯Oº und
lor propylene curhonute tle recovery time is even longer.
5 . 4 0 | S C U S S | 0 N .
Oxygen molecules cun he clemisorhed quite strongly to tle surluce ol ZnO
purticles. 1lis meuns tlut upon suturuting u suspension witl un inert gus, only tle lree
oxygen molecules ure removed wlicl expluins wly tle emission properties luve not
clunged ulter suturuting witl nitrogen. UV irrudiution (witl ploton energies ligler tlun
tle hundgup, ol u suspension ol ZnO purticles cun leud to u plotoinduced desorption
ol clemisorhed oxygen molecules us tle result ol u reuction witl protons to lorm
lydrogen peroxide j2|. Wlen tlere ure no otler oxygen molecules present in solution
to repluce tle desorhed ones, it is possihle to muke tle surluce ol tle ZnO purticles lree
ol udsorhed oxygen hy UV irrudiution. 1le plotodesorption process is uccompunied hy
u decreuse ol tle intensity ol tle visihle emission, us slown in ligure ¯.2 hetween J=O
und J=2OOO seconds. A detuiled explunution lor tlis quencling will he given luter. 1le
luct tlut tle rute ol quencling is slow is due to tle experimentul conditions. 1le UV
Figure 5.5 : Intensity of the visible emission of suspensions of nanocrystalline ZnO
particles in propylene carbonate (a) and DMF (b) during UV irradiation (4.3 eV). For
comparison, the result of the measurement in 2-propanol is shown as graph (c). The
fact that graphs (a) and (b) have a lower signal-to-noise ratio than graph (c) is due to a
difference in concentration with respect to the 2-propanol suspension. On the vertical
axis the intensity at time t (I
t
) is given as a percentage of the the intensity at t=0 (I
0
). A :
Emission intensity versus irradiation time of freshly prepared samples after saturation
with nitrogen. B : Emission intensity versus time of the same samples as in A, but after
the suspensions have been kept in the dark for several minutes.
(b)
(c)
(a)
Time (sec)
0 500 1000 1500 2000
0
50
100
I
t
/
I
0

(
%
)
(b)
(c)
(a)
Time (sec)
0 100 200 300 400 500
0
20
40
60
80
100
I
t
/
I
0

(
%
)
A. B.

rudiution heum only illuminutes u smull purt ol tle suspension inside tle quurtz cuvet
und tle semiconductor purticles cun dilluse in und out ol tlis ureu. At uny given time,
less tlun ¯º ol tle totul numher ol ZnO purticles will he illuminuted hy tle UV heum.
As wus mentioned in tle introduction to tlis clupter, dillerent explunutions
luve heen suggested lor tle quencling ol tle visihle emission ol ZnO under UV
irrudiution. Suhsequent experiments luve indicuted tlut u plotoinduced dissolution ol
tle muteriul is not responsihle lor tle ohserved heluviour j6|. Also, u sluttle
meclunism hused on udsorhed oxygen j2| wus not in ugreement witl results ohtuined
hy lurtler experiments j1O|. 1le results presented in tlis clupter provide lurtler
evidence tlut tle sluttle meclunism is not responsihle lor tle visihle emission. 1le
meclunism requires tle presence ol udsorhed oxygen to ohserve trup emission lor
nunocrystulline ZnO purticles. 1lis would meun tlut tle recovery ol tle trup emission
intensity wlile tle ZnO suspension is in tle durk, is due to u resorption ol oxygen
molecules tlut ure still present in solution. Suppose tlut sucl u resorption cun tuke
pluce, UV irrudiution slould quencl tle recovered trup emission ut u similur rute us
helore. lt cun he seen in ligure ¯.2 tlut tlis is not tle cuse. Alter tle initiul slow
desorption ol clemisorhed oxygen (uccompunied hy u slow quencling ol tle visihle
emission,, tle decreuse ol tle visihle emission intensity is lust lollowing tle recovery ol
Clupter ¯ . |unocrystulline ZnO lurticles, purt lll
86
tle signul. 1o expluin tlese results we propose u meclunism hused on clurging ol tle
ZnO purticles to uccount lor tle quencling ol tle trup emission upon UV irrudiution.
Wlen u ZnO purticle is plotoexcited in tle uhsence ol udsorhed electron
scuvengers (sucl us oxygen, it cun he negutively clurged hecuuse tle plotogeneruted
loles ure elliciently scuvenged hy solvent molecules sucl us ulcolols (resulting in tle
oxidution ol tle lydroxide group,. 1le ulcolol rudiculs produced hy tlis reuction muy
in¦ect udditionul electrons into tle semiconductor purticle j18|. For u tlin lilm
consisting ol ZnO purticles it lus heen ohserved tlut clurging witl electrons leuds to u
decreuse ol tle trup emission intensity wlile ut tle sume time tle intensity ol tle
exciton emission increuses j14|. 1lis wus expluined hy u lilling ol electron trups
resulting in un increuse ol tle numher ol lree electrons. lt wus ussumed tlut tle
recomhinution ol lree electrons witl lree loles (exciton emission, wus tle most
prohuhle rudiutive process. Clurging ellects luve not only heen ohserved lor
nunopurticulute ZnO electrodes, hut ulso lor suspensions ol nunocrystulline ZnO
purticles j¯,6|, us wus ulreudy mentioned in tle introduction. 1le ohserved quencling
ol tle trup emission cun he expluined hy tle presence ol excess electrons on tle ZnO
purticles. 1le recovery ol tle trup emission intensity wlile tle suspension is kept in tle
durk, is due to declurging ol tle ZnO purticles. Lven in tle uhsence ol electron
scuvengers like oxygen, u slow declurging ol tle purticles will tuke pluce. From tle
timescule ol tle recovery (lull recovery ulter ¯ minutes, inlormution cun he ohtuined on
tle liletime ol tle clurged purticles. 1le present explunution lor tle quencling ol tle
visihle emission expluins tle ohservutions slown in ligure ¯.2. ln u lreslly prepured
colloidul suspension, tle oxygen wlicl is udsorhed on tle surluce ol tle ZnO purticles
ucts us un electron scuvenger. 1le quencling ol tle visihle emission is slow, hecuuse ull
tle oxygen lus to he removed helore tle purticle cun he clurged. Alter tle recovery ol
tle signul in tle durk, tle quencling is mucl luster. llotoexcitution ol electronlole
puirs und tle suhsequent scuvenging ol tle loles hy 2propunol (or etlunol,
immediutely results in tle lormution ol clurged purticles, us no ellicient electron
scuvengers ure present.
As wus mentioned helore, tle quencling ol tle visihle emission upon UV
irrudiution is uccompunied hy u slilt ol tlis emission hund ol 8O meV to ligler energies.
1le intensity ol tle exciton emission hund increuses und slilts hy some 8O meV to
lower energies. 1lis suggests tlut predominuntly tle emission cluructeristics ol tle
higger purticles in tle suspension ure ullected hy UV irrudiution. For nunocrystulline
ZnO purticles tlut slow quuntum size ellects, it is known tlut tle emission hunds ure
hroudened due to purticle size distrihution. 1le low energy side ol tle emission hund
originutes muinly lrom tle higger purticles und tle ligl energy side lrom tle smuller
87
purticles in tle distrihution. Clurging ol purticles results in un increuse ol tle UV
emission und u quencling ol tle visihle emission, us discussed uhove. 1le slilt ol tle
UV und visihle emission to lower und ligler energies respectively, cun he expluined il
tle lurger purticles ure clurged more eusily upon UV irrudiution. Wlen tle visihle
emission lrom tle higger purticles is quencled tlis is munilested hy u slilt ol tle
muximum ol tle visihle emission hund to ligler energies due to u lurger contrihution
lrom tle smuller purticles, us is ohserved. 1le opposite upplies lor tle exciton emission
hund. 1le exciton emission originuting lorm tle higger purticles increuses upon UV
irrudiution resulting in u slilt ol tlis emission hund to lower energies. Appurently, tle
higger ZnO purticles cun he clurged witl electrons more eusily tlun tle smuller ones. lt
lus heen slown helore tlut tle redox potentiul ol nunopurticulute electrodes consisting
ol quuntumconlined ZnO purticles depends on tle purticle size j18|. 1le potentiul ut
wlicl electron in¦ection into ZnO occurs increuses witl decreusing purticle size. 1lis
heluviour is expected wlen electrons ure trunslerred into tle quuntised electronic
levels ol ZnO.
1o descrihe in more detuil tle wuy in wlicl tle presence ol excess electrons in
u ZnO purticle results in u decreuse ol tle trup emission intensity, it is neccessury to
know wlut kind ol trunsition is responsihle lor tlis emission. We luve heen uhle to
ussign tle trup emission ol ZnO to u trunsition ol u plotogeneruted electron lrom u
slullow level close to tle conduction hund to u deeply trupped lole in tle hulk ol tle
purticle (see clupter 8,. Furtlermore, we luve identilied tle trupped lole us heing u V
O
••
centre. 1le V trupped lole centre is creuted hy trupping ol u lole hy u V centre,
O O
•• •
prohuhly viu surluce stutes (see clupter 4,. V centres (oxygen vucuncies contuining
O

one electron, ure tle most uhundunt purumugnetic delects in ZnO j1¯| und tley ure
locuted uhout 2 eV helow tle conduction hund edge j16|. Lucl ol tlese centres cun
trup un udditionul electron to lorm V centres (oxygen vucuncies witl two electrons,.
O
x
Clurging ol tle ZnO purticles witl electrons will ruise tle electron quusi Fermi level
wlicl meuns tlut tle oxygen vucuncies will exist us V lor most ol tle time during
O
x
wlicl tle purticle is clurged. Lllectively, tlis meuns u removul ol tle V centres hy u
O

clurging ol tle purticles wlicl expluins tle quencling ol tle visihle emission. At tle
sume time, tle ligler concentrution ol conduction hund electrons will increuse tle
recomhinution rute ol excitons, tlus expluining tle increuse ol tle exciton emission
intensity.
1le possihility to clurge u semiconductor purticle witl electrons is inlluenced
hy tle nuture ol surluce udsorhutes. Ol course, un electron scuvenger (sucl us oxygen,
slould not he present wlicl is wly tle suspensions were suturuted witl nitrogen. |ext,
tle plotogeneruted loles slould he removed lrom tle purticle hy u suituhle lole
Clupter ¯ . |unocrystulline ZnO lurticles, purt lll
88
scuvenger. From tle results us presented in ligure ¯.¯ß it is cleur tlut 2propunol,
etlunol und DVF ure equully cupuhle ol scuvenging plotogeneruted loles lrom u ZnO
purticle. For tlese solvents, tle visihle emission intensity quencles ut u similur rute
under UV irrudiution. From ligure ¯.¯ß it cun ulso he seen tlut propylene curhonute is
mucl less uhle to scuvenge loles us tle emission intensity lurdly decreuses ut ull upon
UV irrudiution. 1lis is in ugreement witl results lrom experiments on ZnO electrodes in
propylene curhonute under compuruhle conditions j17|. ln tlese experiments no
signilicunt oxidution ol propylene curhonute wus ohserved.
1le ohservution tlut in DVF tle visihle emission ol tle ZnO purticles only
quencles to uhout 4Oº insteud ol 1Oº in 2propunol could he due to cluster
lormution ol ZnO purticles. 1lis would meun tlut clusters ol purticles ure dillicult to
clurge. lossihly tle purticles inside sucl u cluster ure lurdly clurged hecuuse tley ure
¨slielded¨ hy tle outer purticles lrom tle cluster.
8D
5 . 5 C 0 N C L U S | 0 N .
Lxperimentul evidence is presented lor u model expluining tle quencling ol
tle visihle emission lrom oxygenlree suspensions ol nunocrystulline ZnO purticles in
ulcolols sucl us 2propunol und etlunol under UV irrudiution. 1le quencling is due to
clurging ol tle ZnO purticles. ln tle uhsence ol un electron scuvenger sucl us oxygen,
tle removul ol plotogeneruted loles lrom tle purticles hy solvent molecules results in
tle lormution ol negutively clurged ZnO purticles. ln tlese purticles, tle V centres
O

wlicl ure involved in tle visihle emission process, ure converted into V centres
O
x
resulting in u quencling ol tle visihle emission.
Chapter 5 : Nanocrystalli ne ZnO Parti cles, part I I I
90
RE F E R E N C E S .
[ 1] J.I . Pankove, Optical Processes in Semiconductors, Dover Publi cati ons, New York ( 1971) .
[ 2] D.W. Bahnemann, C. K ormann, M.R. Hoffmann, J. Phys. Chem. 91 ( 1987) 3789.
[ 3] S. Sakohara, M. I shi da, M.A. Anderson, J. Phys. Chem. B 102 ( 1998) 10169.
[ 4] J. Rabani , D. Behar, J. Phys. Chem. 93 ( 1989) 2559.
[ 5] U. K och, A. Fojti k, H. Weller, A. Henglei n, Chem. Phys. Lett. 122 ( 1985) 507.
[ 6] A. Henglei n, A. K umar, E. Janata, H. Weller, Chem. Phys. Lett. 132 ( 1986) 133.
[ 7] E. Burstei n, Phys. Rev. 93 ( 1954) 632.
[ 8] T.S. Moss, Proc. Phys. Soc. (London) B 76 ( 1954) 775.
[ 9] P.V. K amat, N.M. Di mi tri jevi c, A.J. Nozi k, J. Phys. Chem. 93 ( 1989) 2873.
[ 10] M. Haase, H. Weller, A. Henglei n, J. Phys. Chem. 92 ( 1988) 482.
[ 11] L. Spanhel, M.A. Anderson, J. Am. Chem. Soc. 113 ( 1991) 2826.
[ 12] E.A. Meulenkamp, J. Phys. Chem. B 102( 29) ( 1998) 5566.
[ 13] P. Hoyer, H. Weller, Chem. Phys. Lett. 221 ( 1993) 479.
[ 14] P. Hoyer, R. Ei chberger, H. Weller, Ber. Bunsenges. Phys. Chem. 97 ( 1993) 630.
[ 15] K . Vanheusden, W.L. Warren, C.H. Seager, D.R. Tallant, J.A. Voi gt, B.E. Gnade, J. Appl. Phys. 79( 1)
( 1996) 7983.
[ 16] F.A. K röger, The Chemistry of Imperfect Crystals, North-Holland Publi shi ng Company, Amsterdam
( 1964) , p. 691.
[ 17] P.E. de Jongh, E.A. Meulenkamp, D. Vanmaekelbergh, J.J. K elly, to be submi tted.
D1
D2
D8
C H A P ¡ £ k
L U M | N £ S C £ N C £
Q U A N ¡ U M £ F F | C | £ N C | £ S
6
¡ he | nf | uence cf Pr epor of | cn,
Sur f oce Poss| vof | cn ond Por f | c| e S| ze
A 8 S ¡ k A C ¡ .
A detuiled unulysis ol tle luminescence quuntum elliciencies ol
nunocrystulline CdS und ZnO purticles is presented in tlis clupter. For
CdS purticles, it is slown tlut u prepurution using Nu S yields purticles
?
witl u ligler quuntum elliciency tlun wlen H S is used. 1le quuntum
?
elliciency cun he increused hy surrounding tle purticles witl u
cudmium lydroxide luyer or hy cooling tle suspension to helow its
lreezing point. ßotl metlods leud to u removul ol tle nonrudiutive
recomhinution centres ut tle surluce ol tle purticle. For ZnO, tle
inlluence ol purticle size on tle quuntum elliciency ol tle visihle
emission is studied. lt is slown tlut tle quuntum elliciency ol tle
visihle emission decreuses lrom ?Oº lor purticles witl u meun rudius ol
7 A to 1?º lor purticles witl u meun rudius ol 1O A. An explunution lor
tle sizedependence ol tle quuntum elliciency is presented
Clupter o . Luminescence Quuntum Llliciencies
D4
6 . I | N ¡ k 0 0 U C ¡ | 0 N
A lurge vuriety ol colloidul semiconductor purticles lus heen investiguted hy
luminescence meusurements. One ol tle lirst llVl semiconductors tlut wus investiguted
is CdS j1,?|. Generully, u colloidul suspension ol CdS purticles slows two emission
hunds upon plotoexcitution. 1le lirst is u relutively weuk und nurrow emission hund ut
un energy close to tlut ol tle hundgup, wlicl is uhout ?.4 eV lor mucrocrystulline CdS.
1lis hund is known us tle exciton emission hund. 1le second hund is u trup emission
hund wlicl is hrouder und more intense tlun tle exciton emission. 1lis emission hund
lus u muximum ut upproximutely 1.8 eV. However, tle dominuting process lor
reluxution is usuully rudiutionless decuy, us is evident lrom tle low quuntum yield,
wlicl is generully helow 1º. 1le uhsolute intensity ol tle emission ol colloidul CdS
suspensions is strongly dependent on severul conditions, tle most importunt one heing
tle nuture ol tle surluce. 1le reuson lor tlis is tle luct tlut surluce stutes ure usuully
involved in nonrudiutive processes und smull purticles luve u lurge surlucetovolume
rutio. lt lus heen reported tlut covering tle surluce ol CdS purticles witl u cudmium
lydroxide luyer results in un increuse ol tle luminescence intensity j8,4|. Not only tle
nuture ol tle purticle surluce itsell, hut ulso tlut ol tle solvent is ol importunce. Upon
exclunging tle uqueous solvent hy un ulcolol sucl us metlunol, tle luminescence
intensity ulso increuses j8|. Lspeciully tle exciton emission is sensitive to surluce
pussivution. ln luct, tlis emission is dominunt in suspensions ol wellpussivuted purticles
j8,5|. ln tlis clupter, tle inlluence ol surluce pussivution on tle luminescence quuntum
elliciency lor CdS purticles prepured hy two syntlesis metlods in reported.
lt is ulso interesting to consider tle inlluence ol purticle size on tle
luminescence elliciency. ln some pupers it lus heen reported tlut tle luminescence
quuntum elliciency ol nunocrystulline semiconductor purticles increuses witl
decreusing purticle size. At lirst siglt tlis muy seem surprising since tle prohuhility lor
nonrudiutive decuy viu surluce sites cun he expected to increuse us tle purticle size
decreuses due to u lurger surlucetovolume rutio. Still, lor CdS it wus slown tlut
plotocorrosion results in u decreuse ol tle purticle size und un increuse ol tle
luminescence elliciency wlicl wus expluined hy u prelerentiul removul ol surluce
delect sites ucting us nonrudiutive recomhinution centres j5|. For ZnS purticles doped
witl Vn un increuse in tle Vn emission wus ohserved witl u decreuse ol tle purticle
?÷ ?÷
size. 1lis wlicl wus expluined hy more ellicient trupping ol electronlole puirs hy Vn

in smuller ZnS purticles us u result ol un increused wuvelunction overlup jo|. Finully, lor
ZnO it lus heen reported tlut smuller purticles luminesce more elliciently ultlougl no
detuiled unulysis lus heen reported. Wletler u sizedependence ol tle luminescence
quuntum elliciency cun he ohserved depends on tle meclunisms ol tle rudiutive und
D5
nonrudiutive processes tlut cun occur in u certuin semiconductor. For nunocrystulline
ZnO purticles, tlese meclunisms luve heen descrihed in detuil (see clupters 8 und 4,.
1lis lorms tle husis lor u study ol tle sizedependence ol tle luminescence quuntum
elliciency lor nunocrystulline ZnO purticles us presented in tlis clupter.
6 . 2 £ X P £ k | M £ N ¡ A L M £ ¡ H 0 0 S .
1le experiments us descrihed in tlis clupter were perlormed on colloidul
suspensions ol nunocrystulline llVl semiconductor purticles. 1le muteriuls tlut were
used ure CdS und ZnO. 1le colloidul CdS purticles were prepured in two wuys, dillering
only in tle nuture ol tle sulplur source. For eucl ol tlese prepurutions, purt ol tle
suspension wus suh¦ected to u procedure in wlicl tle surluce ol tle colloidul purticles
wus pussivuted hy creuting u lydroxide luyer uround tlem. ln tlis wuy, lour dillerent
colloidul suspensions ol CdS were prepured und investiguted.
6. 2 . I Somp| e pr epor of | cn.
CdS. First, ?.4 ml ol u O.O1 V Cd(ClO , oH O solution in wuter were diluted to
4 ? ?
.
45 ml in u roundhottom llusk equipped witl u septum. 1o tlis solution, 1 ml ol u O.O1 V
sodium polyplosplute solution in wuter wus udded. Next, 5OO µl H S (Aldricl lecture
?
hottle, DD.5÷ º, were in¦ected using un uirtiglt syringe. Vigorously sluking tle sumple
lor severul minutes ut room temperuture resulted in u trunspurent suspension witl u
yellow colour. According to tle literuture, tlis prepurution ol un uqueous colloidul CdS
suspension slould yield purticles witl u meun rudius ol uhout 85 A j7|. For u second
prepurution, o ml ol u O.O1 V Cd(ClO , oH O solution in wuter were diluted to D5 ml
4 ? ?
.
helore 5 ml ol u O.O1 V sodium polyplosplute solution in wuter were udded. Next, 5 ml
ol u O.O1 V Nu S DH O solution were diluted to 1O ml und udded to tle Cd solution
? ?
.

wlile stirring. Aguin tlis resulted in u trunspurent suspension witl u yellow colour. lurt
ol hotl suspensions wus used lor u surluce pussivution step. 1lis step involved tle
ud¦ustment ol tle pH to 11.5 hy udding u lew droplets ol u 1 V NuOH solution in wuter.
Suhsequently, tle suspension wus kept ut D5 °C lor two lours using u wuterhutl.
Zn0. For tle prepurution ol u colloidul suspension ol ZnO in ?propunol, 1?.5
ml ol u O.O? V solution ol NuOH und 14O ml ol u O.OO1 V solution ol Zn(CH COO, H O
8 ? ?
.
were lirst cooled to O °C helore slowly udding tle lydroxide solution to tle zinc ucetute
solution wlile stirring. Alter u lew minutes, ZnO purticles witl u rudius ol uhout 1O A
luve heen lormed (see clupter 4,. Wlen tle solution is ullowed to uge ut room
temperuture, tle ZnO purticles slowly grow until ulter severul lours tley luve uttuined
Clupter o . Luminescence Quuntum Llliciencies
Do
tleir linul size, luving u rudius ol uhout 8O A. At dillerent times during tle purticle
growtl u sumple is tuken to he used lor uhsorption und emission meusurements. Alter
tle emission meusurement, uguin un uhsorption spectrum wus tuken to cleck wletler
tle purticle size lus clunged during tle emission meusurement tlut generully took
uhout 5 minutes. As tle ZnO uhsorption spectrum did not clunge during tlis time
period, it is concluded tlut tle purticles did not grow siginilicuntly.
6. 2 . 2 L um| nescence quonf um ef f | c| enc| es.
1o determine tle quuntum elliciency ol tle luminescence ol u colloidul
suspension ol semiconductor purticles, tle emission spectrum is compured to tlut ol u
relerence solution witl u known quuntum elliciency. For un uccurute determinution ol
tle quuntum elliciency, tle uhsorhunce ut tle excitution wuvelengtl slould he low und
similur lor hotl tle sumple und tle relerence solution. Wlen tle uhsorption strengtl is
ligl, ull excitution liglt is uhsorhed in tle lirst tlin luyer ol tle sumple wlicl is not
locussed on tle slit ol tle emission monoclromutor. Furtlermore, tle spectrul position
ol tle emission hund ol tle sumple wus closen to he us close us possihle to tlut ol tle
relerence solution. 1le overlup hetween tle emission hund und tle uhsorption hund ol
tle relerence solution slould he us low us possihle to prevent reuhsorption ol tle
emitted liglt hy tle relerence itsell. 1le concentrution ol tle relerence solution lus to
he in tle regime wlere tle emission intensity scules lineurly witl tle numher ol
uhsorhed plotons.
For tle experiments descrihed in tlis clupter, solutions ol dillerent dyes were
used us u relerence. 1le luminescence quuntum elliciency ol tle dye solutions is close
to DOº. 1le dyes tlut were used were sullorlodumine 1O1 lor CdS und coumurine 158
lor ZnO. Solutions ol tlese dyes were prepured in tle sume solvent us used lor tle
prepurution ol tle colloidul semiconductor suspensions, witl u runge ol concentrutions
tlut slows u lineur relutionslip hetween tle numher ol uhsorhed plotons und tle
emission intensity. ll necessury, tle colloidul suspensions were diluted in order to luve
un uhsorhunce tlut is in tle sume runge us tlut ol tle concentrution series ol tle
relerence.
D7
6. 2 . 3 0pf | co| chor ocf er | sof | cn.
Ahsorption meusurements were perlormed on u lerkinLlmer Lumhdu 1o
UV/Vis spectroplotometer. 1le plotoluminescence meusurements were perlormed on
u SlLX Fluorolog spectrolluorometer model F?OO? equipped witl two douhlegruting
O.?? m SlLX 1o8O monoclromutors und u 45O W xenon lump us tle excitution source.
1le emission spectru were corrected lor tle spectrul response ol tle emission
monoclromutor und tle lV tuhe.
6 . 3 k £ S U L ¡ S A N 0 0 | S C U S S | 0 N .
6. 3. I CdS pr epor ed w| f h H S .
2
As cun he seen in ligure o.1A, tle uhsorption spectrum ol tle colloidul
suspension ol CdS purticles prepured witl H S did not clunge signilicuntly ulter tle
?
suspension lud heen suh¦ected to tle pussivution procedure. 1le lormution ol u
lydroxide luyer uround tle CdS purticles did not clunge tle size ol tlese purticles.
Also, tle suspension slowed no liglt scuttering due to clustering ol purticles. Figure
o.1A ulso contuins tle emission spectrum ol u suspension ol nonpussivuted CdS
purticles wlicl slows u hroud hund witl u muximum ut upproximutely 1.5 eV. 1le low
energy side ol tlis emission hund cunnot he recorded witl tle luminescence setup tlut
wus used lor tlese experiments us tle lV tuhe lus u very low sensitivity lor plotons
witl un energy lower tlun 1.5 eV. 1o determine tle totul intensity ol tle emission, tle
hroud hund wus litted to u guussiun lunction, tle ureu ol wlicl cun he used us u
meusure lor tle totul emission intensity. Upon pussivuting tle surluce ol tle CdS
purticles, tle intensity ol tle emission hund increuses strongly.
6. 3. 2 CdS pr epor ed w| f h No S .
2
Contrury to tle results lor tle prepurution ol CdS purticle witl H S, tle
?
uhsorption spectrum ol u colloidul suspension ol CdS purticles prepured witl Nu S did
?
clunge upon pussivution ol tle purticle surluce. 1lis cun he seen in ligure o.1ß. 1le
onset ol uhsorption us determined hy extrupoluting tle steep purt ol tle uhsorption
curve to tle energy uxis stuys ut tle sume position indicuting tlut tle size ol tle lurgest
purticles in tle suspension lus not clunged upon surluce pussivution. However, tle
structure ol tle spectrum lus clunged. Alter tle onset ol uhsorption, tle uhsorhunce
Figure 6.1 : Room temperature emission (excitiation with 4 eV) and absorption
spectra of colloidal suspensions of CdS particles in water. A : H
2
S preparation, before
(a) and after (b) surface passivation. B : Na
2
S preparation, before (a) and after (b)
surface passivation. Φ
E
denotes the photon flux per constant energy interval.
1.0 2.0 3.0 4.0
0
10
20
30
40
50
0.0
0.1
0.2
0.3
0.4
0.5
(b)
(a)
Energy (eV)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
A
b
s
o
r
b
a
n
c
e
(b)
(a)
1.0 2.0 3.0 4.0
Energy (eV)
Φ
E

(
a
r
b
.

u
n
i
t
s
)
A
b
s
o
r
b
a
n
c
e
0
70
140
210
280
350
0.00
0.05
0.10
0.15
0.20
0.25
B.
A.
Clupter o . Luminescence Quuntum Llliciencies
D8
rises more steeply lor tle pussivuted purticles tlun lor tle nonpussivuted purticles. Also,
u muximum cun he ohserved ut uhout ?.8 eV in tle uhsorption spectrum ol tle
pussivuted CdS purticles. 1lese ohservutions cun he expluined hy u nurrowing ol tle
purticle size distrihution wlicl could luve resulted lrom temperutureinduced Ostwuld
ripening ol tle purticles. Compurison ol tle uhsorption spectru ol tle Nu S prepurution
?
to tlose ol tle H S prepurution slows tlut in tle lutter cuse tle onset ol uhsorption is ut
?
lower energies. Appurently tle H S prepurution results in tle lormution ol lurger
?
purticles tlun tle Nu S prepurution. Altlougl tle exuct reuson lor tlis is uncleur it could
?
he due to tle luct tlut in tle Nu S prepurution u lurger concentrution ol ¨lree¨ sulplur
?
unions (S , will he present wlile in tle H S prepurution u lurge purt ol tle sulplur
?−
?
DD
unions will he in present us HS or H S in solution. A ligler concentrution ol S results
− −
?
?
in luster growtl kinetics und tle lormution ol smuller colloidul purticles. Similur to tle
H S prepurution, tle intensity ol tle emission hund increuses upon pussivution ol tle
?
purticle surluce.
6. 3. 3 L um| nescence quonf um ef f | c| enc| es cf CdS.
ln tuhle 1, tle results ol tle determinution ol tle luminescence quuntum
elliciencies ol colloidul suspensions ol CdS purticles ure presented. lt is cleur tlut tle
prepurution witl H S generully leuds to tle lormution ol purticles tlut luve lower
?
luminescence quuntum elliciencies tlun tle purticles ohtuined lrom tle Nu S
?
prepurution. ßused on tle uhsorption spectru it wus concluded tlut tle prepurution ol
CdS witl Nu S yielded smuller purticles tlun wlen H S is used. 1le purticle size could
? ?
he reluted to tle ligler luminescence quuntum elliciency. Vore prohuhly, tle dillerent
syntlesis metlods result in purticles witl u dillerent concentrution ol nonrudiutive
surluce sites.
6=>A Luminescence quuntum elliciencies ol colloidul suspensions ol CdS
purticles in wuter, prepured hy in¦ecting H S or hy udding u solution ol Nu S DH O.
? ? ?
.
lurt ol tle suspensions wus suh¦ected to u procedure in wlicl tle surluce ol tle
purticles wus pussivuted hy lorming u lydroxide luyer.
prepurution metlod
QL (º, helore QL (º, ulter
surluce pussivution surluce pussivution
H S O.? O.8
?
Nu S 1.5 8.O
?
From tuhle 1 it is ulso cleur tlut pussivution ol tle purticle surluce leuds to un
increuse ol tle luminescence quuntum elliciency. 1le surluce ol u semiconductor
purticle is u strong perturhution ol tle luttice wlere u ligl concentrution ol hotl
slullow und deep levels provide u putlwuy lor nonrudiutive recomhinution ol
plotogeneruted clurge curriers j8|. lussivuting tle surluce hy lorming u lydroxide
luyer leuds to tle removul ol surluce stutes und tlerelore u lower prohuhility lor non
rudiutive decuy.
1lis lower prohuhility is ulso evident lrom liletime meusurements tlut were
perlormed on suspensions ol CdS purticles. Figure o.? slows tle results ol tlese
meusurements on purticles witl und witlout surlucepussivution. At room temperuture,
Figure 6.2 : Luminescence lifetime measurements of the trap emission (1.7 eV) from
colloidal suspensions of CdS particles in water. The CdS particles were prepared with
Na
2
S. A : Measurement at 8K for non-passivated (a) and passivated (b) CdS particles. B :
Measurement at room temperature for non-passivated (a) and passivated (b) CdS
(a)
(b)
(a)
(b)
Time (µs)
0.0 0.5 1.0 1.5 2.0
1
10
100
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
Time (µs)
0.0 0.5 1.0 1.5 2.0
1
10
100
I
n
t
e
n
s
i
t
y

(
a
r
b
.

u
n
i
t
s
)
Clupter o . Luminescence Quuntum Llliciencies
1OO
surlucepussivution lus u pronounced inlluence on tle liletime ol tle trup emission (see
ligure o.?ß,. 1le decuy curve ol tle trup emission lrom tle purticles witl no surluce
pussivution cun he litted to u hiexponentiul lunction. 1le strongest ol tle two decuy
components wlicl uccounts lor ulmost DOº ol tle signul lus u decuy time ol uhout
5OO ns wlile tle otler component is mucl luster, witl u decuy time ol 4O ns. Upon
pussivution ol tle surluce, tle decuy curve ol tle trup emission hecomes mono
exponentiul witl u decuy time ol uhout DOO ns. Upon hlocking tle nonrudiutive
putlwuys ut tle purticle surluce, tle liletime ol tle trup emission increuses hy ulmost u
luctor ?. At low temperutures (8K, see ligure o.?A,, tle inlluence ol surlucepussivution
on tle liletime ol tle trup emission is mucl less pronounced. For hotl sumples tle
1O1
decuy curves ure multiexponentiul witl u similur slupe. 1lis indicutes tlut cooling u
suspension ol CdS purticles helow its lreezing point lus u similur ellect us surrounding
tlese purticles witl u lydroxide luyer, viz. hlocking tle nonrudiutive routes ut tle
purticle surluce.
6. 3. 4 Zn0.
Ahsorption spectru tlut were tuken ut regulur intervuls during tle growtl ol ZnO
purticles in ?propunol ut room temperuture cun he seen in ligure o.8. 1le onset ol
uhsorption slilts to lower energies us tle size ol tle purticles increuses, wlicl is due to
u decreuse in quuntum conlinement. Using u relutionslip hetween tle purticle rudius
und tle onset ol uhsorption jD|, it is determined tlut tle rudius ol tle ZnO purticles
increuses lrom 7 A u lew minutes ulter tle reuction wus sturted to 1O A ulter 5OO min ol
purticle growtl. 1le emission spectrum ol u colloidul suspension ol ZnO purticles slows
two emission hunds . u weuk und nurrow hund ut energies close to tle uhsorption onset
(exciton emission hund, und u strong und hroud emission hund uhout 1.5 eV helow tle
onset ol uhsorption (trup emission hund,. As tle intensity ol tle exciton emission hund
is negligihle compured to tlut ol tle trup emission hund, only tle lutter is used to
determine tle luminescence quuntum elliciencies. For mucrocrystulline ZnO, tle
elliciency ol tle visihle emission upon irrudiution witl UV liglt is uhout ?Oº j1O|.
6. 3. 5 L um| nescence quonf um ef f | c| enc| es cf Zn0.
1le luminescence quuntum elliciencies ure plotted versus tle meun purticle
rudius in ligure o.4. lt cun he seen tlut tle quuntum elliciency decreuses nonlineurly
lrom uhout ?Oº lor purticles witl u rudius ol 7 A to 1?º lor purticles witl u rudius ol 1O
A. 1o expluin wly tle luminescence quuntum elliciency decreuses us tle purticle size
increuses it is necessury to know tle meclunism wlicl is responsihle lor tle trup
emission. New inlormution concerning tlis meclunism lus heen descrihed in detuil in
clupters 8 und 4 ol tlis tlesis und u sclemutic overview is presented in ligure 4.7. 1le
key leutures ol tle trup emission process ure u very lust trupping ol tle plotogeneruted
lole ut u surluce site lollowed hy tunnelling ol tlis surlucetrupped lole huck into tle
hulk ol tle ZnO purticle to u V oxygen vucuncy. ln tlis wuy, u V centre is creuted
O O
• ••
wlicl cun recomhine witl u plotogeneruted electron lrom u level close to tle
conduction hund to give tle trup emission. Apurt lrom tle otler rudiutive process
(exciton emission,, tle most importunt process competing witl tle trup emission is
nonrudiutive decuy. ln nunocrystulline purticles nonrudiutive decuy predominuntly
Figure 6.3 : Absorption spectra of a suspension of nanocrystalline ZnO particles in 2-
propanol, taken after 20 min (a), 90 min (b), and 500 min (c) of particle growth at
room temperature. On the vertical axis, the amount of absorption is given as 1 minus
the transmission.
(a) (b) (c)
Energy (eV)
3.0 3.5 4.0 4.5 5.0
0.0
0.5
1.0
1
-
T
r
a
n
s
m
i
s
s
i
o
n

(
%
)
Figure 6.4 : Room temperature luminescence quantum efficiencies (QE) versus
particle radius for nanocrystalline ZnO particles.
0 7 8 9 10
0
5
10
15
20
25
Radius (Å)
Q
E

(
%
)
Clupter o . Luminescence Quuntum Llliciencies
1O?
tukes pluce ut tle surluce. 1le lurger tle surlucetovolume rutio is, tle ligler tle
prohuhility lor nonrudiutive decuy will he. Nevertleless, tle smuller ZnO purticles luve
u ligler luminescence quuntum elliciency tlun tle lurger ones. 1o understund tlis, tle
two importunt competing processes (trup emission und nonrudiutive decuy, luve to he
unulyzed. 1le lirst step in hotl processes is trupping ol u deloculised plotogeneruted
clurge currier ut tle surluce. 1lis step slows u similur sizedependence lor hotl
processes wlicl meuns tlut tle second step is importunt lor tle ohserved size
1O8
dependence ol tle luminescence quuntum elliciency. ln tle cuse ol nonrudiutive
decuy, tle second step is uguin surlucetrupping ol u deloculised clurge currier. For tle
trup emission, tle second step is tunneling ol u surlucetrupped lole huck into tle hulk
ol u ZnO purticle to u V centre. ln tlis process, two loculised stutes ure involved. 1le
O

prohuhility lor sucl u process decreuses more strongly wlen tle distunce hetween tle
clurge curriers increuses tlun tle prohuhility lor u process involving u deloculised
clurge currier. ln tle lutter cuse, tle wuvelunction overlup is less sensitive to tle
purticle size. lt is importunt tle note tlut tle quencling ol tle visihle luminescence ol
ZnO purticles tukes pluce helore tle uctuul recomhinution centre is creuted.
1le luct tlut tle luminescence quuntum elliciency decreuses wlen tle size ol
tle ZnO purticles increuses indicutes tlut tle distunce hetween tle surlucetrupped lole
und tle deeply trupped electron in tle hulk oxygen vucuncy increuses. 1lis is in
ugreement witl results tlut indicute tlut tle delect concentrution (including V centres,
O

decreuses wlen tle size ol tle purticles increuses, us derived lrom studies on tle size
dependence ol tle dissolution rute ol nunocrystulline ZnO purticles j11|. 1le
explunution lor tle decreuse ol tle quuntum elliciency witl increusing purticle size is
ulso in ugreement witl results ol timedependent luminescence meusurements on ZnO
purticles ol dillerent sizes us presented in clupter 4. Figure 4.4 contuins u decuy curve ol
tle visihle emission lrom ZnO purticles witl u meun rudius ol 1O A us well us u decuy
curve ol tle emission lrom purticles witl u meun rudius ol 8O A. 1le emission lrom tle
lurger purticles lus u longer liletime tlun tlut ol tle smuller purticles.
Clupter o . Luminescence Quuntum Llliciencies
1O4
6 . 4 C 0 N C L U S | 0 N .
For severul colloidul suspensions ol nunocrystulline llVl semiconductor
purticles tle luminescence quuntum elliciencies were determined hy compuring tle
emission intensity to tlut ol u dye solution witl u known quuntum elliciency. For CdS,
tle inlluence ol tle nuture ol tle sulplur source (H S or Nu S, und tle inlluence ol
? ?
pussivution ol tle purticle surluce on tle luminescence quuntum elliciencies wus
investiguted. lt wus lound tlut tle quuntum elliciency is liglest wlen using Nu S to
?
prepure u colloidul suspension ol CdS purticles. ßy creuting u lydroxide luyer uround
tle purticle tle nonrudiutive recomhinution centres ut tle surluce ure pussivuted wlicl
results in un increuse ol tle quuntum elliciency.
For ZnO tle inlluence ol purticle size on tle luminescence quuntum elliciency
wus investiguted. 1le quuntum elliciency decreuses us tle size ol tle purticles
increuses. 1le mu¦ority ol tle rudiutive recomhinution in ZnO is trup emission. An
importunt step in tlis emission process is tle tunnelling ol u surlucetrupped lole to u
deeply trupped electron in tle hulk ol tle ZnO purticle. 1le prohuhility lor tlis
tunnelling process decreuses strongly wlen tle purticle size increuses, und us u result
tle luminescence quuntum elliciency decreuses.
1O5
4 - . - 4 - + - 5
j1| A. Henglein, 8cr. 8onscngcs. //ys. C/cn. 86 (1D82, 8O1.
j2| A. Henglein, 1. //ys. C/cn. 86 (1D82, 22D1.
j8| L. Spunlel, H. Weller, A. Fo¦tik, A. Henglein, 8cr. 8onscngcs. //ys. C/cn. D1 (1D87, 88.
j4| A. Lyclmüller, L. Kutsikus, H. Weller, Longno|r 6 (1DDO, 16O5.
j5| L. Spunlel, V. Huuse, H. Weller, A. Henglein, 1. An. C/cn. Soc. 1OD (1D87, 564D.
j6| R.N. ßlurguvu, D. Gullugler, X. Hong, A. Nurmikko, //ys. Rco. Lc//. 72(8, (1DD4, 416.
j7| A. Lyclmüller, A. Husselhurtl, L. Kutsikus, H. Weller, 1.Lon|n. 48&4D (1DD1, 745.
j8| 1.l. lunkove, Oµ/|co/ /roccsscs |n Scn|conJoc/ors, Dover luhlicutions. New York, 1D71.
jD| V. Huuse, H. Weller, A. Henglein, 1. //ys. C/cn. D2 (1D88, 482.
j1O| A. Sclleede, Lcoc//cn onJ S/ro//or /cs/cr S/o//c (R. 1omusclek, ed.,, R. Oldenhourg . Vunicl, 1D48.
j11| L.A. Veulenkump, 1. //ys. C/cn. 8 1O2 (1DD8, 7764.
1Oo
.
111
5 ) - 8 ) 6 6 1 /
Hullgeleiders zi¦n muteriulen die ul ruim londerd ¦uur uitvoerig hestudeerd
worden en inmiddels op grote scluul zi¦n verwerkt in ullerlei soorten toepussingen. Len
helungri¦k voorheeld liervun zi¦n optoelektroniscle uppuruten zouls lotodiodes, liclt
emitterende diodes en zonnecellen. De luutstgenoemde toepussing is een
elektroclemiscle cel wuurmee met helulp vun zonliclt een elektriscle stroom kun
worden opgewekt. Op husis vun letzellde principe kunnen ook clemiscle reucties
worden geïnitieerd. Om dergeli¦ke processen zo ellicient mogeli¦k te luten verlopen
worden lullgeleidermuteriulen met een zo groot mogeli¦k oppervluk gehruikt. Len
munier om dit te hereiken is door te werken met kleine lullgeleiderdeelt¦es. Vunul let
hegin vun de ¦uren tucltig wordt dit intensiel geduun. Nuust de interesse in kleine
lullgeleiderdeelt¦es vunwege toepussingen in elektroclemiscle cellen, zi¦n deze
systemen eveneens interessunt vunuit een lundumenteel opziclt. lmmers, wunneer de
ulmetingen vun een lullgeleider vergeli¦khuur worden met de grootte vun een exciton
in dit muteriuul, dun treden kwuntum opsluitingsellecten op. Het gevolg vun deze
ellecten is dut de energetiscle structuur vun een lullgeleiderdeelt¦e ullunkeli¦k wordt
vun zi¦n grootte. Dit munilesteert zicl onder undere door een toenume vun de
hundulstund hi¦ een ulnemende deelt¦esgrootte en een geleideli¦ke overgung vun
energiehunden nuur discrete energieniveuus. Dergeli¦ke ellecten worden ul ¦uren
hestudeerd in zeer dunne lullgeleider leterostructuren wuur ze in een ol twee
dimensies optreden. ln kleine deelt¦es hi¦voorheeld colloïden in suspensie kunnen
kwuntum opsluitingsellecten in ulle dimensies optreden. Voorul deelt¦es vun llVl
muteriulen zouls CdS, ZnS en ZnO zi¦n uitermute gesclikt uls systeem voor let
Samenca||/ng
.
112
hestuderen vun kwuntum opsluitingsellecten in drie dimensies. Zi¦ zi¦n eenvoudig te
muken en lun energetiscle structuur hegint ul een grootteullunkeli¦kleid te vertonen
hi¦ ulmetingen vun enkele nunometers.
De in dit proelsclrilt hesclreven experimenten zi¦n uitgevoerd uun colloïdule
suspensies vun llVl lullgeleiderdeelt¦es die kwuntum opsluitingsellecten vertonen. Lr
wordt hesclreven loe deze ellecten kunnen worden gehruikt om lullgeleiderdeelt¦es
op een gecontroleerde munier kleiner te muken. Dit geheurt door ze te lotoetsen
wuurhi¦ de uiteindeli¦ke deelt¦esgrootte ullungt vun de kleur vun liclt dut gehruikt
wordt. Verder worden kwuntum oplsuitingsellecten gehruikt om let meclunisme op te
lelderen dut veruntwoordeli¦k is voor de hekende ziclthure luminescentie vun ZnO.
Hiertoe worden temperutuurullunkeli¦ke luminescentiemetingen uitgevoerd uun
deelt¦es met versclillende ulmetingen, zowel ti¦dsgeïntegreerd uls ti¦dsopgelost. De
metingen hi¦ luge temperutuur werden uitgevoerd door een kwurts cuvet gevuld met
enkele microliters vun een suspensie ul te koelen in een cryostuut met helulp vun
vloeihuur lelium.
Het zo¦uist genoemd lotoetsproces wordt hesclreven in looldstuk 2. Het
hetrelt lier een postprepurutieve metlode wuurmee let mogeli¦k is om vun een
hepuuld lullgeleidermuteriuul een reeks vun deelt¦es met versclillende ulmetingen te
muken, elk met een relutiel smulle deelt¦esgrootteverdeling. Deze metlode wordt
groo||ese/ec|/e/ /o|oe|sen genoemd en herust op let leit dut sommige lullgeleiders
lotocorrosie vertonen (oplossen onder heliclting,. Wunneer lullgeleiderdeelt¦es die
kwuntum opsluitingsellecten vertonen oplossen onder heliclting dun verunderen lun
eigenscluppen. Dit leelt uls gevolg dut er een moment komt wuurop zi¦ niet meer in
stuut zi¦n om let ingestruulde liclt te uhsorheren. Op dut moment zul let lotoetsen
oplouden en zullen de deelt¦es een hepuulde grootte lehhen hereikt die ullungt vun
de golllengte vun let ingestruulde liclt. Door deze golllengte te vurieren kun een reeks
vun versclillende deelt¦esgroottes worden gemuukt. Lxperimenten uun let
modelsysteem CdS lehhen uungetoond dut met helulp vun grootteselectiel lotoetsen
deelt¦es met een struul vun 85 A op een gecontroleerde munier kleiner gemuukt kunnen
worden tot een struul vun 7.5 A, terwi¦l tegeli¦kerti¦d de deelt¦esgrootteverdeling ulneemt
vun 4Oº tot 1O15º. Ook voor undere llVl muteriulen hli¦kt de metlode te werken, zi¦
let met wisselend succes.
De looldstukken 8 tot en met 5 zi¦n gewi¦d uun de studie vun een soort llVl
verhinding, te weten ZnO. Wut hetrelt let hestuderen vun kwuntum opsluitingsellecten
is er uun ZnO veel minder uunduclt hesteed dun uun de sulliden. 1ocl is let vri¦
eenvoudig om nunokristulli¦ne ZnO deelt¦es te muken die deze ellecten vertonen.
.
118
Verder is ZnO interessunt omdut let een elliciente ziclthure (groene, emissie vertoont
wuurdoor let muteriuul uls luminescerende verhinding wordt toegepust. ln tegenstelling
tot nunokristulli¦ne deelt¦es is ZnO uls mucrokristulli¦n muteriuul wel veel hestudeerd
muur desondunks is let meclunisme uclter de ziclthure emissie nog niet opgelelderd.
Aungezien deze emissie zowel in nunokristulli¦ne deelt¦es uls in mucrokristulli¦n ZnO
optreedt en wuurscli¦nli¦k herust op letzellde meclunisme kun let eerstgenoemde
systeem gehruikt worden voor let oplelderen vun de uurd vun de ziclthure emissie vun
ZnO.
Len dergeli¦ke studie is let onderwerp vun looldstuk 8 en omvut
emissiemetingen uun een reeks vun ZnO deelt¦es met versclillende groottes. Nuust de
ziclthure emissie vertonen deze deelt¦es ook een zogenuumde exciton emissiehund.
Deze emissie is let gevolg vun de recomhinutie vun excitonen (elektrongut puren, en
de energie vun let muximum vun deze hund komt overeen met de hundulstund vun let
muteriuul. Het hli¦kt dut er een lineuir verhund hestuut tussen de energetiscle positie
vun de ziclthure emissiehund en dut vun de exciton emissiehund. Uit de lelling vun let
lineuire verhund tussen de energetiscle posities vun de twee emissiehunden kun
worden ulgeleid dut de overgung die veruntwoordeli¦k is voor de ziclthure emissie een
recomhinutie is vun een elektron uit de geleidingshund met een gut dut zicl hevindt in
een niveuu ongeveer 2 eV onder de geleidingshund. Op husis vun de kennis over de
delectstructuur vun ZnO wordt uungenomen dut let hedoelde niveuu gevormd wordt
door een zuurstolvucuture wuurin zicl geen enkel elektron hevindt.
Duur wuur in looldstuk 8 conclusies over ZnO in let ulgemeen worden
getrokken, wordt in looldstuk 4 dieper ingeguun op de eigenscluppen die speciliek zi¦n
voor nunokristulli¦ne ZnO deelt¦es. De experimenten die in dit looldstuk hesclreven
worden, leiden uiteindeli¦k tot let opstellen vun een model wuurin ulle
reluxutieprocessen worden hesclreven die pluutsvinden nudut een elektrongut puur is
gecreeerd in een ZnO deelt¦e totdut dit deelt¦e weer terug is in zi¦n grondtoestund. Len
vun deze reluxutieprocessen is de ziclthure emissie en let wordt duideli¦k gemuukt dut
dit proces uit meerdere stuppen hestuut. De eerste stup is een uiterst snel en ellicient
proces wuurin let gut wordt gevungen uun let oppervluk vun let ZnO deelt¦e,
wuurscli¦nli¦k door een O ion. Vervolgens hestuut er een kuns dut dit gut terug let
2−
deelt¦e in tunnelt wuur let tereclt kun komen in een zuurstolvucuture wuurin zicl een
elektron hevindt. Het is hekend dut dergeli¦ke zuurstoldelecten veel voorkomen in ZnO.
Dit proces zorgt voor de vorming vun de uiteindeli¦ke recomhinutiecentru voor de
ziclthure emissie, zouls die ul worden verondersteld in looldstuk 8.
Samenca||/ng
.
114
1enslotte wordt in looldstuk 5 hesclreven loe met helulp vun let model dut
wordt opgesteld in looldstuk 4 verkluurd kun worden wuurom in hepuulde suspensies
de ziclthure emissie vun ZnO deelt¦es doolt wunneer er voor gezorgd wordt dut er geen
zuurstol meer in let systeem uunwezig is. Deze verkluring herust op let leit dut ZnO
deelt¦es kunnen worden opgeluden met elektronen in de ulwezigleid vun
geudsorheerde zuurstolmoleculen. Deze extru elektronen kunnen vervolgens de
zuurstolvucutures opvullen wut leidt tot let verdwi¦nen vun de recomhinutiecentru voor
de ziclthure emissie.
De experimenten die hesclreven zi¦n in looldstuk 6 lehhen hetrekking op de
kwuntumellicientie vun de emissie vun nunokristulli¦ne deelt¦es vun zowel ZnO uls CdS.
Voor CdS wordt onderzoclt wut de invloed is vun de syntlesemetlode en de
oppervlukteeigenscluppen op de kwuntumellicientie met uls resultuut dut een syntlese
uitguunde vun Nu S meer ellicient luminescerende deelt¦es oplevert dun met H S. ln
2 2
heide gevullen kun de kwuntumellicientie nog worden verheterd door een luug¦e
cudmiumlydroxide te groeien rondom de deelt¦es. Hierdoor worden de plekken uun
let oppervluk gepussiveerd wuur normuul gesproken nietstrulend vervul kun
pluutsvinden. Voor ZnO deelt¦es wordt de invloed vun de deelt¦esgrootte op de
kwuntumellicientie hepuuld. Het hli¦kt dut de kwuntumellicientie ulneemt wunneer de
deelt¦es groter worden. Dit is in overeenstemming met let model voor de kinetiek vun
de reluxutieprocessen in een geexciteerd ZnO deelt¦e zouls dut in looldstuk 4 wordt
gepresenteerd.
.
115
113
LI S T O F P U B L I C A T I O N S
J O U R N A L A R T I C L E S .
[ 1] A. van Di jken, A. Mei jeri nk, D. Vanmaekelbergh, Chem. Phys. Lett. 269 ( 1997) 494 [ Chapter 2] .
[ 2] A. van Di jken, A.H. Janssen, M.H.P. Smi tsmans, D. Vanmaekelbergh, A. Mei jeri nk, Chem. Mater. 10( 11)
( 1998) 3513 [ Chapter 2] .
[ 3] A. van Di jken, E.A. Meulenkamp, D. Vanmaekelbergh, A. Mei jeri nk, J. Lumin. ( accepted for
publi cati on) [ Chapter 3] .
[ 4] A. van Di jken, E.A. Meulenkamp, D. Vanmaekelbergh, A. Mei jeri nk, J. Phys. Chem. B ( submi tted for
publi cati on) [ Chapter 4] .
[ 5] A. van Di jken, E.A. Meulenkamp, D. Vanmaekelbergh, A. Mei jeri nk, manuscri pt i n preparati on
[ Chapter 5] .
[ 6] A. van Di jken, J. Makki nje, A. Mei jeri nk, manuscri pt i n preparati on [ Chapter 6] .
C O N F E R E N C E P R O C E E D I N G S .
[ 1] A. van Di jken, D. Vanmaekelbergh, A. Mei jeri nk, Proceedings of the Symposium on
Photoelectrochemistry ( 1997 Joi nt Meeti ng of the Electrochemi cal Soci ety and the I nternati onal Soci ety
of Electrochemi stry i n Pari s, France) , p.79-83.
[ 2] A. van Di jken, E.A. Meulenkamp, D. Vanmaekelbergh, A. Mei jeri nk, Proceedings of the Fifth International
Symposium on Quantum Confinement : Nanostructures ( 1998 Meeti ng of the Electrochemi cal Soci ety i n
Boston, USA) , p. 392-401.
[ 3] A. van Di jken, E.A. Meulenkamp, D. Vanmaekelbergh, A. Mei jeri nk, Proceedings of the 1999
International Conference on Luminescence and Optical Spectroscopy of Condensed Matter (Osaka), to be
publi shed.
114
115
OV E R D E I N HO U D
V A N D I T P R O E F S C HR I F T
Een overzicht voor niet- ingewijden
HA L F G E L E I D E R S .
Halfgelei ders vormen een speci ale klasse van materi alen di e vanwege hun
ei genschappen zeer veel worden gebrui kt. I n de loop der jaren i s aan deze materi alen
veel onderzoek gedaan met als gevolg dat de hoeveelhei d kenni s i nmi ddels groot i s. Net
zoals i eder ander materi aal bevat een halfgelei der elektronen, deeltjes met een
negati eve ladi ng. Voor elk soort materi aal gedragen de elektronen zi ch op een bepaalde
mani er waardoor de speci fi eke ei genschappen van di t materi aal worden veroorzaakt.
Wanneer bi jvoorbeeld elektronen vri j kunnen bewegen -zoals het geval i s i n metalen -
dan kan het materi aal een elektri sche stroom gelei den. De elektronen i n halfgelei ders
kunnen ni et zomaar vri j bewegen omdat ze sterk gebonden zi jn aan de atomen waarui t
de halfgelei der i s opgebouwd. Het i s echter mogeli jk om een gebonden elektron los te
maken van een atoom, waarna het wél door het materi aal kan bewegen. Hi er i s een
zekere hoeveelhei d energi e voor nodi g di e kan worden geleverd door warmte of door
li cht, dat door de halfgelei der wordt geabsorbeerd. Pas daarna kan een halfgelei der
stroom gelei den. De plek waar het elektron zat kan nu worden i ngenomen door een
ander elektron, di e op zi jn beurt weer een lege plaats op een atoom achterlaat. Het
steeds maar weer opvullen van lege plekken betekent dat er elektronen door de
halfgelei der bewegen. Het bli jkt zo te zi jn dat je di t ook kunt beschri jven als het
bewegen van gaten door de halfgelei der. Een gat gedraagt zi ch alsof het een elektron i s,
maar dan met een posi ti ve ladi ng. Het vri j maken van een elektron i n een halfgelei der
heeft dus ni et alleen tot gevolg dat di t elektron door het materi aal kan gaan bewegen
maar tevens dat een gat hetzelfde kan gaan doen.
Over de inhoud van dit proefschrift
116
Bekende voorbeelden van toepassi ngen van halfgelei ders zi jn transi storen
( onder andere terug te vi nden i n computerchi ps) , li cht-emi tterende di odes ( LED’ s, de
li chtgevende onderdelen van bi jvoorbeeld wekkerradi o’ s) en zonnecellen. Zoals
ui tgelegd lei dt de absorpti e van li cht door een halfgelei der tot het ontstaan van
elektronen en gaten di e vervolgens voor een elektri sche stroom kunnen zorgen. Op di t
pri nci pe berust de werki ng van een zonnecel. De elektronen en gaten kunnen ook
gebrui kt worden om bepaalde chemi sche verbi ndi ngen om te zetten i n verbi ndi ngen
di e meer gewenst zi jn. Di t i s de reden waarom halfgelei ders worden toegepast bi j de
producti e van brandstoffen en bi j het afbreken van schadeli jke verbi ndi ngen i n
afvalwater.
ME R K WA A R D I G V E R S C HI J N S E L .
De mogeli jkhei d om halfgelei ders te gebrui ken om bepaalde chemi sche
reacti es ui t te voeren, heeft aan het begi n van de jaren tachti g tot een i nteressante
ontwi kkeli ng gelei d. Om li cht zo effi ci ënt mogeli jk om te zetten i n vri je elektronen en
gaten i s een groot oppervlak belangri jk. Een mani er om het oppervlak van een
halfgelei der te vergroten i s door van het materi aal een poeder te maken. Hoe klei ner de
poederdeeltjes zi jn, hoe groter het totale oppervlak i s. Ti jdens het maken van klei ne
deeltjes van het halfgelei dermateri aal CdS ( cadmi umsulfi de) , namen onderzoekers op
een bepaald moment een merkwaardi g verschi jnsel waar : wanneer de deeltjes extreem
klei n worden, verandert de kleur van het CdS poeder van geel naar wi t.
KWA N T U M O P S L U I T I N G S E F F E C T E N.
I n een heel klei n halfgelei derdeeltje kunnen elektronen en gaten nooi t echt vri j
bewegen. Ze zi tten alti jd opgesloten i n het deeltje en omdat ze een tegengestelde ladi ng
hebben zullen ze elkaar aantrekken. De toestand di e op deze mani er ontstaat kan
worden gezi en als een gat waar een elektron omheen ci rkelt. Een dergeli jke toestand
wordt een exciton genoemd. Absorpti e van li cht door een klei n halfgelei derdeeltje lei dt
dus alti jd tot de vormi ng van een exci ton.
De kleur van een materi aal hangt af van welk gedeelte van het zi chtbare li cht
door deze stof geabsorbeerd kan worden. Zonli cht bevat alle kleuren en wanneer di t op
een materi aal valt dat alleen rood li cht kan absorberen, dan zal di t materi aal er blauw
ui tzi en. Deze kleur wordt nameli jk door het materi aal gereflecteerd zodat het vervolgens
117
door het menseli jk oog kan worden waargenomen. I n eerste i nstanti e li jkt de kleur van
een materi aal onafhankeli jk te zi jn van zi jn afmeti ngen. Echter, de eerder genoemde
experi menten aan CdS deeltjes hebben aangetoond dat voor halfgelei ders de kleur wel
degeli jk afhangt van de afmeti ngen. Wanneer een halfgelei derdeeltje steeds maar
klei ner wordt gemaakt dan komt er een moment waarop zi jn kleur begi nt te veranderen.
Vanaf deze grootte i s steeds meer energi e nodi g om exci tonen te maken wat tot
ui tdrukki ng komt i n een verschui vi ng van de absorpti e van li cht naar een andere kleur.
De kleurveranderi ng van steeds klei ner wordende halfgelei derdeeltjes i s één van de
meest i n het oog spri ngende gevolgen van het optreden van zogenaamde kwantum
opsluitingseffecten. I n theori e waren deze effecten al jaren lang bekend en ze waren ook
al waargenomen i n hele dunne lagen van halfgelei ders. I n halfgelei derdeeltjes met
afmeti ngen van een mi ljoenste mi lli meter ( nanometer) mani festeren deze effecten zi ch
echter bi jzonder sterk. I nmi ddels heeft zi ch rondom deze nanokristallijne
halfgelei derdeeltjes een geheel ni euw onderzoeksgebi ed ontwi kkeld waari n vooral veel
aandacht wordt geschonken aan het bestuderen van kwantum opslui ti ngseffecten.
Alhoewel men veel weet over de ei genschappen van halfgelei ders met “normale”
afmeti ngen, valt er nog veel onderzoek te verri chten aan deeltjes met afmeti ngen van
slechts enkele nanometers. Dergeli jke nanokri stalli jne halfgelei derdeeltjes zi jn het
onderwerp van di t proefschri ft.
S U S P E N S I E S .
Voor een gedegen onderzoek naar kwantum opslui ti ngseffecten i n
halfgelei derdeeltjes i s het ten eerste ui termate belangri jk dat de deeltjes di e bestudeerd
worden allemaal ongeveer dezelfde afmeti ngen hebben. I n de loop der jaren heeft men
veel ervari ng opgedaan met het maken van dergeli jke deeltjes, vooral i n de vorm van
suspensi es. Een suspensi e i s een vloei stof waari n de vaste stof zi ch als een fi jnverdeeld
materi aal bevi ndt. Alle experi menten di e beschreven zi jn i n di t proefschri ft zi jn gedaan
aan suspensi es van halfgelei derdeeltjes. De materi alen di e gebrui kt zi jn, zi jn sulfi den -
zoals CdS, ZnS ( zi nksulfi de) en PbS ( loodsulfi de) - maar vooral ZnO ( zi nkoxi de) . Een
overeenkomst tussen deze materi alen i s dat ze behoren tot de klasse van II-VI
materi alen.
De experi menten hebben een aantal belangri jke resultaten opgeleverd. Zo i s
aangetoond dat het mogeli jk i s om halfgelei derdeeltjes di e zi ch i n een suspensi e
bevi nden klei ner te maken door er li cht op te stralen. Di t gebeurt op een gecontroleerde
mani er en de ui tei ndeli jke deeltjesgrootte wordt bepaald door de kleur van het li cht.
Over de inhoud van dit proefschrift
118
Deze methode wordt grootte-selectief foto-etsen genoemd en wordt ui tgebrei d
beschreven i n hoofdstuk 2.
Een tweede i nteressant resultaat heeft te maken met ZnO. Di t materi aal wordt
ni et alleen als halfgelei der gebrui kt, maar ook als fosfor. Een fosfor i s een vaste stof di e
luminescentie vertoont. Lumi nescenti e i s het verschi jnsel dat bepaalde verbi ndi ngen
zi chtbaar li cht ui tzenden wanneer zi j worden bestraald met ultravi olette ( UV) strali ng.
Deze strali ng kan ni et worden waargenomen door het menseli jk oog maar i s bi jna
overal aanwezi g. Zonli cht bevat bi jvoorbeeld een zekere hoeveelhei d UV strali ng wat
de oorzaak i s voor het brui n worden van de hui d wanneer deze wordt blootgesteld aan
de zon. Ook i n verli chti ng wordt UV strali ng en lumi nescenti e gebrui kt. Zo wordt i n een
TL-bui s UV strali ng opgewekt en vervolgens door poeders, di e zi ch aan de bi nnenkant
van de bui s bevi nden, omgezet i n wi t li cht. Wanneer ZnO wordt bestraald met UV
strali ng dan wordt door het materi aal groen li cht afgegeven. Vanwege deze ei genschap
wordt ZnO toegepast i n li chtgevende schermpjes di e bi jvoorbeeld zi jn verwerkt i n
audi o- en vi deo-apparatuur. Waar deze groene lumi nescenti e van ZnO preci es door
veroorzaakt werd was echter ondui deli jk. Door bepaalde experi menten te doen aan
klei ne ZnO deeltjes van verschi llende groottes en bi j verschi llende temperaturen i s het
mogeli jk gebleken om meer i nformati e over de groene lumi nescenti e te verkri jgen.
Deze experi menten worden beschreven i n de hoofdstukken 3 en 4.
119
120
121
¡ 0 ¡ 8 £ S L U | ¡
£ en wccr d von donk
ln 1887 leelt Slerlock Holmes gezegd dut li¦ de ¨//c|oc J|oçco¨ ulti¦d let meest
helungri¦k vindt. Dut li¦ zeker niet de enige is die op de kleint¦es let, hleek in 1DD5 toen
Andries Vei¦erink hesloot om iemund uun te nemen die zicl vier ¦uren lung hezig zou
guun louden met kleine lullgeleiderdeelt¦es. lnmiddels kun ik zeggen dut ik erg hli¦ hen
dut ik deze rol leh mogen vervullen. Andries, de munier wuurop ¦i¦ ¦e promovendi
hegeleidt leh ik ontzettend prettig gevonden. Het is indrukwekkend om te zien loe
leuk ¦i¦ ¦e vuk vindt. Zouls een oudvoethuller die coucl wordt vuuk nog even een
hullet¦e mee trupt, zo kun ¦i¦ let niet luten om uctiel met wetenscluppeli¦k onderzoek
hezig te hli¦ven zi¦n. Dut mi¦n onderzoek nuust luminescentie ook de undere speciuliteit
vun onze vukgroep (elektroclemie vun lullgeleiders, omvutte, vond ik erg interessunt.
1oln Kelly en Duniel Vunmuekelhergl lehhen mi¦ meerdere mulen op weg gelolpen
uls ik een lullgeleiderprohleem lud. Duniel, ¦ouw commentuur is vun onscluthure
wuurde geweest voor de inloud vun mi¦n proelsclrilt en voor de wuurdering vun Luc
Nilis uls voethuller. De sumenwerking met een undere elektroclemicus, Lric
Veulenkump, leelt geresulteerd in de drie ZnO looldstukken en menig teleloongesprek
en/ol emuilcontuct tussen Lric en mi¦zell. Lric, nuust verstund vun zuken leh ¦e een oog
voor detuil, een comhinutie die let hespreken vun mi¦n werk soms tot een
middugvullende hezigleid muukte muur voorul tot gevolg lud dut ik er een stuk wi¦zer
op werd. Voorul ti¦dens let hegin vun mi¦n promotieperiode werd ik vunuit
colloïdclemiscle kunt gelolpen door Curlos vun Kuts vun de Co//o|J Syo/bcs|s /oc|/|/y
in Utreclt. Hi¦ hruclt me enkele kneep¦es hi¦ en leelt zi¦n uiterste hest geduun om die
wel lele kleine colloïden onder de microscoop ziclthuur te muken. Wut dit luutste
hetrelt verdient zeker ook Vurcel Verlei¦en vun let /b|/|µs Cco/rc /or Mooo/oc/or|oç
1o/ bcs/o|/
122
1ccboo/oçy (Lindloven, een pluim. Hi¦ muukte de 1LVopnumen die uiteindeli¦k in dit
proelsclrilt zi¦n gehruikt. De gemotoriseerde teclnicus Huns Ligtlurt verstuut zi¦n vuk en
is op de vukgroep en in de sportzuul een gewuurdeerde kruclt. Verder is er een viertul
studenten geweest wuurop ik hli¦khuur mi¦n entlousiusme leh weten over te hrengen
uungezien zi¦ hesloten om mi¦ gedurende een hepuulde periode te lelpen met mi¦n
onderzoek. De inspunningen vun de looldvukkers Alrik vun den ßrom, Vonique
Smitsmuns en 1eroen Vukkin¦e en die vun de kleinhi¦vukker Ries 1unssen, lehhen een
uunzienli¦ke hi¦druge geleverd uun dit proelsclilt. 1enslotte wil ik nog iemund
hedunken wiens uunwezigleid in onze groep uitermute helungri¦k is geweest . Arnim
Henglein, pionier op let gehied vun onderzoek uun kleine lullgeleiderdeelt¦es. Deur
lrol. Henglein, it wus u greut pleusure lor me to work witl you during tle lirst montls ol
my ll.D. reseurcl. Your knowledge on nunocrystulline semiconductor purticles und
your entlousiusm lor your work is most impressive.
1egeli¦k met mi¦ hegonnen op 1 septemher 1DD5 nog twee undere mensen uun
lun promotieonderzoek en met z'n drieen kwumen we tereclt in kumer 268. ln vier
¦uur ti¦d is die kumer uitgegroeid tot een goed geoliede werkomgeving, voorzien vun ulle
gemukken. Luxe hureuustoelen, een uitgehreide verzumeling kuntoorurtikelen en een
computerinlrustructuur wuur menige multinutionul zi¦n vingers hi¦ ul zou likken. letru,
ik weet dut ik let vuuk niet kon luten om een pi¦nli¦ke opmerking te muken over de
gehreken vun de Vuc, die ¦i¦ dun weer pureerde door mi¦ op let gerutel vun mi¦n lC te
wi¦zen. Ln Rene, ik hen mi¦ er vun hewust dut ik ¦e regelmutig lustig viel met deunt¦es
wuur zells een getruinde scout zouls ¦i¦ nog simpel vun kon worden. Desondunks loop
ik dut ¦ullie ¦e de ulgelopen vier ¦uur net zo goed lehhen vermuukt uls ik zell. Wie lud
geduclt dut we ulledrie uiteindeli¦k hi¦ let Nut. Luh. tereclt zouden komenº (ik in ieder
gevul niet!,. lk loop dut we elkuur nog regelmutig zullen zien.
Ook undere prominente liguren ik noem een Ar¦un Vink dienen zeker vermeld
te worden. Ar¦un, eerst pructicumpurtner en duurnu collegu, ik duclt dut ik nooit vun ¦e
ul zou komen! 1ocl moet ik toegeven dut ik wuurscli¦nli¦k met weemoed terug zul
denken uun de ti¦d wuurin ik dugeli¦ks door ¦ou werd gestulkt viu lCQ en iedere
muundugocltend imhossiuunse verlulen moest uunloren over de pluutseli¦ke FC. Vuur
ook Ageetl, Koert, Lieshetl÷Lrik, Ferry, Freek, Vurcel, Aurnoud, Xingluu, Vur¦olein en
1eroen lehhen zeker hi¦gedrugen een plezierige ti¦d. Len speciule vermelding guut uit
nuur Gi¦s, tulent vun de eerste verdieping en vuuk zo kupot. Vuur wut kon 'ie losguun!
Legenduriscl zi¦n de uvonden wuurop we sumen nuur let purtycentrum uun de
ßemuurde Weerd trokken om ulduur de een ol undere onhenullige pot uit de
Clumpions Leugue te heki¦ken. Nutuurli¦k kwumen we eigenli¦k voor de uithuter vun let
128
etuhlissement, De Vulk, en voor zi¦n onuitputteli¦ke voorruudkust (de Kust is OK!,. De
Vulk, zi¦n grootsleid kent geen grenzen.....
ßuiten mi¦n werk kon ik ulti¦d rekenen op hepuulde mensen voor de nodige
doses entertuinment . Diederik, met wie ik terug gu tot de o/J scboo/, en Surlgod
Reinoud. C1 de Vergruizer (21.OO ßelgieº,, die ulti¦d te porren wus voor kollie, hier en
undere versnuperingen. Ln nutuurli¦k de Godenzonen vun DLS1O ongeslugen
kumpioenen vun de derde lellt wunt DLS1O 6........DA1 lS GLZLLLlGHLlD!
Vi¦n vuder en moeder lehhen mi¦ ulti¦d de dingen luten doen die ik leuk vind.
Ze oelenden nooit druk op me uit en ulti¦d stonden ze voor me kluur . eclte toppers!
Nog zo'n zeldzume topper is ßelindu. Vroeger kon ¦e me ul hoeien met ¦e spunnende
verlulen en nu inmiddels sumen met Cuse en Kristun kom ik nog steeds ontzettend
gruug hi¦ ¦ullie lungs.
Het is ul gezegd, kleine dingen zi¦n helungri¦k, muur een //c|o J|oç is eclt let
nccs/ helungri¦k, en dut is Vendely. De ulgelopen zes ¦uren sumen met ¦ou wuren een
grote luppening! lk kun niet wuclten om sumen met ¦ou nuur ßruhunt te trekken (¦uwel,
u loort let goed, ulwuur de provinciule luisdierenpopulutie een periode vun grote
hloei tegemoet kun zien wunneer ¦i¦ uls dierenurts ¦e opwuclting zul muken.
124
125
C U R R I C U L U M
VI T A E
De schri jver van di t proefschri ft werd geboren op 19 augustus 1972 te Bennekom, i n de
gemeente Ede. I n 1990 behaalde hi j het VWO di ploma aan “Het Wagenings Lyceum” te
Wageni ngen om vervolgens aan een studi e Schei kunde aan de Uni versi tei t Utrecht te
begi nnen. I n augustus 1991 werd het propedeuti sch di ploma behaald en i n augustus
1995 het doctoraal di ploma. De studi e omvatte twee keuzepakketten, te weten Fysica en
Chemie van Materialen en Fysisch-Organische Chemie. Het hoofdvak werd ui tgevoerd bi j
de vakgroep Gecondenseerde Materi e. Op 1 september 1995 trad hi j als Onderzoeker i n
Oplei di ng ( Oi O) i n di enst van NWO voor een promoti e-onderzoek bi nnen de vakgroep
Gecondenseerde Materi e van de Uni versi tei t Utrecht. Het grootste gedeelte van de
behaalde resultaten zi jn beschreven i n di t proefschri ft. Ti jdens zi jn promoti eperi ode
werden resultaten van het onderzoek door de schri jver gepresenteerd op een dri etal
bui tenlandse conferenti es. I n 1997 werd de Joi nt I nternati onal Meeti ng of the
Electrochemi cal Soci ety ( ECS) and the I nternati onal Soci ety of Electrochemi stry ( I SE) te
Pari js bezocht, i n 1998 de Meeti ng of the Electrochemi cal Soci ety ( ECS) te Boston en i n
1999 de I nternati onal Conference on Lumi nescence and Opti cal Spectroscopy of
Condensed Matter ( I CL) te Osaka. Naast het doen van wetenschappeli jk onderzoek
werden enkele onderwi jstaken ui tgevoerd. Di t betrof de begelei di ng van de
werkcolleges Structuur en Binding 2A en Spectroscopie en het practi cum Fysisch
Chemisch Meten. Vanaf 1 november 1999 zal de schri jver van di t proefschri ft i n di enst
zi jn als wetenschappeli jk medewerker van het Phi li ps Natuurkundi g Laboratori um te
Ei ndhoven.
126