Journal of Geochemical Exploration, 48 ( 1993 ) 109- 133 109

El sevi er Sci ence Publ i sher s B.V., Ams t e r da m

Lithogeochemical techniques using immobile
elements
W. H. MacLean and T.J. Barrett
Department of Earth and Planetary Sciences, McGill University, 3450 University Street, Montreal,
Quebec H3A 2A 7, Canada
(Recei ved 13 January 1992; accepted after revision 6 April 1993)
ABSTRACT
Elements of demonst rat ed i mmobi l i t y in hydrot hermal alteration and met amorphi sm provi de use-
ful lithogeochemical paramet ers for mi neral expl orat i on in greenstone terranes. They yield precise
identifications of precursor vol cani c rock type and magmat i c affinity, and quant i t at i ve estimates of
mass, vol ume and mineralogical changes. I mmobi l e i ncompat i bl e elements (Zr, Y, Nb, REE) estab-
lish affinities, and moni t or fract i onat i on of compat i bl e elements in tholeiitic and transitional vol cani c
suites. Mass changes in mobi l e component s can be det ermi ned for a homogeneous rock unit or a
cont i nuous vol cani c series. Mass change results also bear directly on the i nt erpret at i on of REE pro-
files in altered rocks. Nor mat i ve calculations allow chemical analyses to be portrayed as alteration
minerals, which are also useful in interpreting oxygen isotope data. Immobi l e element techniques
produce a large number of parameters that can be illustrated on maps and diagrams, and otherwise
used in exploration. Usage of these t echni ques brings deeper insight to the evol ut i on of volcanic stra-
tigraphy and the hydrot hermal processes that formed the alteration zones and ore deposits.
I NTR ODUC TI ON
U nr av el l i ng v ol cani c str ati gr aphy is a maj or task i n mi ner al ex p l or at i on i n
gr eenst one belts. T he mi ner al ogi cal , tex tur al and chemi cal i d ent i t i es o f r ocks
are of t en obscur ed by hy d r ot her mal al t er at i on and me t amor p hi s m, l ead i ng to
i ncor r ect i d ent i f i cat i on o f v ol cani c r ock t y p es ( basal t, and esi t e, r hy ol i te ) and
r ock af f i ni t i es ( e. g. t hol ei i t i c, cal c- al k al i ne) . Cor r ect i d ent i f i cat i ons are par-
ti cul ar l y need ed i n d eep dr i l l i ng ex p l or at i on p r ogr ams wher e dr ill hol es are
necessar i l y wi d el y sp aced ow i ng to dr i l l i ng costs. Accur at e i nt er p r et at i on o f
dr ill cor e d ata i s necessar y not onl y f or out l i ni ng mi ner al i z at i on, but al so f or
d ef i ni t i on o f al t er at i on patter ns, bet ween- hol e l i t hol ogi c cor r el ati on, and es-
t abl i s hment o f v ol c ani c - t e c t oni c setti ng.
A numbe r o f r ecent st ud i es i n ex p l or at i on l i t hogeochemi s t r y hav e str essed
the i d ent i f i cat i on and use o f el ement s that hav e r emai ned i mmobi l e d ur i ng
hy d r ot her mal al t er at i on and me t a mo r p hi s m ar ound or e d ep osi t s ( e. g. ,
MacLean and Kr ani d i ot i s, 1987; MacLean, 1988, 1990; Ri char d s et al. , 1989;
0375- 6742/ 93/ $06. 00 1993 El sevi er Sci ence Publ i sher s B.V. All rights reserved.
1 1 0 W . H . Ma c LEAN AND l' .J. BARRETq
Elliott-Meadows and Appleyard, 1991; Barrett and MacLean, 1991 ). Some
i mmobi l e el ement s are also i ncompat i bl e in igneous processes, and can be
used to moni t or fractionation processes. Combi nat i ons of i mmobi l e- com-
patible and i mmobi l e- i ncompat i bl e el ement s produce per manent and diag-
nostic i mpri nt s of fractionation lines and plots that define magmat i c affini-
ties. Usage of these i mpri nt s leads to i mpr oved lithostratigraphic correlation,
i dent i fi cat i on of altered rock precursors, recognition of alteration haloes, cal-
culation of mass changes, and an i mpr oved underst andi ng of oxygen isotope,
REE and trace el ement distributions.
TECHNIQUES
Mappi ng and core logging followed by petrography and l i t hogeochemi st ry
is t he logical order of investigation, with new i nformat i on on rock type and
affinity gai ned at each step. Good field and core sample i dent i fi cat i on is crit-
ical for pet rographi c-l i t hogeochemi cal sampl i ng which provi des t he frame-
work for subsequent chemi cal calculations.
Large files of l i t hogeochemi cal dat a are amassed i n many expl orat i on pro-
grams i n greenst one belts. Dat a are commonl y t reat ed using well-established
geochemi cal di scri mi nant s, such as SiO2 cont ent , Mg-number, di fferent i at i on
index, alkali-SiO2 and AFM plots, whi ch were created pri mari l y to classify
pristine igneous rocks. Most of t he el ement s i n these di scri mi nant s are ig-
neous-compat i bl e, and are also mobi l e duri ng wat er-rock interaction. They
can become highly dispersed duri ng hydrot hermal alteration at moderat e to
high wat er-rock ratios. Al t hough t hei r mobi l i t y actually makes t hem useful
i n out l i ni ng hydrot hermal alteration zones, it also makes t hem unfi t for i den-
tification of precursor rock type and magmat i c affinity. By contrast, treat-
ment of t he chemi cal data using i mmobi l e el ement t echni ques identifies rock
type and affinity, and provi des a number of extra, firmly based quant i t at i ve
parameters.
Met hods for det er mi ni ng whet her certain el ement s have been i mmobi l e i n
met asomat i sm and hydr ot her mal alteration have been discussed, among oth-
ers, by Gresens ( 1967 ), Babcock ( 1973 ), Finlow-Bates and Stumpfl ( 1981 ),
Grant (1986), Krani di ot i s and MacLean ( 1987) , MacLean ( 1988, 1990)
and Elliott-Meadows and Appl eyard (1991 ). In this article we review t he
met hods empl oyed recently by Krani di ot i s and MacLean ( 1987 ), MacLean
(1988, 1990), and Barrett and MacLean (1991). We also outline t he se-
quence of steps requi red i n a lithological investigation t hat uses i mmobi l e
el ement s t o establish chemost rat i graphi c uni t s i n altered and unal t ered vol-
canic sequences, and t o calculate t he mass changes t hat occur wi t hi n t he al-
t erat i on zones. This mass-based approach finds direct application i n explo-
ration and drilling programs.
LITHOGEOCHEM1CALTECHNIQUES USING IMMOBILE ELEMENTS
Immobile elements
111
El ement s shoul d be t est ed for i mmobi l i t y i n each area of st udy, preferabl y
on an i dent i f i abl e single vol cani c flow t hat can be t r aced laterally t hr ough an
al t er at i on zone. Us i ng t he pr ocedur es out l i ned by MacLean and Kr ani di ot i s
( 1987 ) for a single pr ecur sor syst em, an i mmobi l e el ement pai r i n t hi s sampl e
dat a set will f or m a hi ghl y cor r el at ed ( r - - 0. 90- 0. 99) l i near t r end owi ng t o
mass gai ns and losses of mobi l e c ompone nt s i n t he al t ered par t of t he flow; at
i nf i ni t e mass gain, t he regressed l i ne i deal l y passes t hr ough t he ori gi n (Fig.
1 ). Resul t s of t hi s t est show t hat A1, Ti and t he hi gh-fi el d st rengt h t race ele-
ment s ( H FSE) Zr, Nb, Y and t he REE are i mmobi l e i n al t er at i on zones
a r ound a large numbe r of vol canogeni c massi ve sul phi de ( VMS) deposi t s i n
gr eenst one belts ( MacLean and Kr ani di ot i s, 1987, MacLean, 1988 ). W i t hi n
t hi s gr oup of el ement s, Y and t he l i ght -REE are slightly mobi l e unde r some
condi t i ons of al t erat i on, and r-values are lower. Care mus t be t aken t o excl ude
f r om t hi s t est pai rs of el ement wi t h hi gh chemi cal affi ni t y wi t h one anot her ,
such as K- Rb and Zr - H f .
3 0
Phelps Dodge deposit: . , ~ j , ~ e
25 altered rhyolite J _ . , . ~ "
C~ 1 5
O
_ ~ / j "
I " 't
s S
I ~ . s . .
5 ,,,, / Alteratton hne:
, y * " Pr e c u r s o r y = 0. 034x + 0. 579;
s S
,,, r = 0. 993
0 I I I I l I I
0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 0 6 0 0 7 0 0 8 0 0
Zr ppr n
Fig. 1. Effects of mobile mass gain and loss on immobile AI and Zr during alteration of an
initially homogeneous rhyolite unit (single precursor system) at the Phelps Dodge VMS deposit
(data from MacLean and Kranidiotis, 1987 ).
1 12 W H . Ma c LEAN AND T. J. BARRETT
Incompatible elements
A second r equi r ement for our procedures is t he presence of highly i ncom-
patible t race el ement s (bul k di st ri but i on coefficients Dso]id/melt < 0 . 1 ) to
moni t or degree of fract i onat i on and magmat i c affinity. A pai r of i ncompat i -
ble el ement s i n a vol cani c sampl e set produces a l i near magmat i c enr i chment
t r end t hat i nvari abl y passes t hrough t he origin (Fig. 2). Incompat i bl e ele-
ment enr i chment is model l ed by t he Rayleigh fract i onat i on equation:
Ci=Co FD-I ( 1)
where Co and Ci are concent rat i ons of an i ncompat i bl e el ement in initial and
fract i onat ed magmas, respectively; F is t he fraction of r emai ni ng magma; and
D is t he bul k ( s ol i d/ mel t ) di st ri but i on coefficient. W hen D is small (esti-
mat ed at ~ 0.05 for Zr ) , t he el ement can be used t o moni t or t he fract i onat i on
process.
The H FSE are highly i ncompat i bl e duri ng fract i onat i on i n t he range of ba-
salt and basaltic-andesite in most t hol ei i t i c and calc-alkaline suites, but, in
our experience, they remai n i ncompat i bl e in andesite, dacite and rhyolite only
in suites of t hol ei i t i c to t ransi t i onal (mi l dl y calc-alkaline) affinity. The thol-
ei i t i c-t ransi t i onal suites are low in K (generally <0. 5 wt% K20) and frac-
t i onat e t o sodic rhyolite, with albite as t he sole feldspar phenocryst phase. In
1 0 0
8O
6O
~. .
4 0
2 0
0 4 0 0
Host vol cani cs at the Ansil mine:
f ract i onat i on t rend gi ven
by least al tered sampl es
Zr/Y = 5.0
o
O0 y - 5 54 + 0. 201x r - 0 95
i ! !
1 0 0 2OO 3 0 0
Zr ppm
Fig. 2. Incompat i bl e Y and Zr enri chment in vol cani c st rat i graphy at the Ansil mine, Nor anda
( dat a from Barrett et al., 1991 a) .
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS 1 13
E
120
100'
80,
60'
40
20
0 100 200 300 400 500
Zr ppm
Fig. 3. Schemat i c Y- Zr pl ot used for de t e r mi na t i on of magmat i c affi ni t y.
t he maj ori t y of fract i onat ed calc-alkaline suites, where K-fel dspar is also a
phenocryst phase i n dacites and rhyolites, t he H FSE are mor e compat i bl e,
and D values are vari abl e and may be >t I. The H FSE shoul d be t est ed for
i ncompat i bi l i t y i n a rock suite, using plots such as Y- Zr or La- Yb, before
cont i nui ng wi t h these procedures. H owever, calc-alkaline suites can still be
t reat ed using i mmobi l e compat i bl e el ement s, as out l i ned below.
Some ratios bet ween t he H FSE change with rock affinity, and t hereby pro-
vi de di scri mi nant s unaffect ed by alteration. The change is part i cul arl y large
and di st i nct i ve for Zr / Y ratios: these range from ~ 3 to 5 for tholeiites, t o ~ 7
to ~ 30 for calc-alkaline rocks (Fig. 3) (cf. Lesher et al., 1986). Whereas
plots of Zr versus Y are l i near for tholeiites, and exhibit steady i ncompat i bl e
enr i chment f r om basalt t o rhyolite, those for a si mi l ar range of calc-alkaline
rocks pr oduce poorl y correl at ed t rends, wi t h depl et i ons caused by variable
compat i bi l i t y of bot h Zr and Y. Rel at i ve depl et i on is, however, greater for Y,
so t hat t he Zr / Y val ue is hi gher for calc-alkaline rocks. We designate samples
with i nt ermedi at e Zr / Y values ( ~ 5-7 ) "t ransi t i onal "; they have linear H FSE
enr i chment t rends, but slightly st eeper REE profiles and slightly hi gher A1-
cont ent s t han tholeiites.
Magmat i c affinity can be conf i r med using changes i n La/ Yb and Zr / Yb
ratios, as t he heavy REE are geochemi cal l y si mi l ar t o Y. Slopes of Zr versus
Nb plots coi nci de i n some t hol ei i t i c t o mi l dl y calc-alkaline suites, but diverge
i n ot hers (Barret t et al., 1991 a,b, 1992 ). The Zr / Nb ratio is distinctly hi gher
1 14 W. H . MacLEAN AND T.J. BARRETT
in anorogeni c (ri ft -rel at ed) t han i n orogeni c (subduct i on-rel at ed) alkaline
rocks, and provi des a reliable di scr i mi nant for magmat i c suites and t ect oni c
setting.
Other, less abundant H FSE (e.g., Hf, Ta, W, Th ) can also be used to deter-
mi ne rock affinity, but, because of t hei r low abundances ( ~ 0.1 - 10 ppm ) and
high analytical costs, t hey are not as suitable as those descri bed above. In our
experience, Zr is t he most useful el ement to moni t or fract i onat i on; it is im-
mobi l e, highly i ncompat i bl e in t he rock series not ed above, and sufficiently
abundant ( ~ 50-1000 ppm) for precise chemi cal analyses. The Zr / Y ratio is
t he most conveni ent and least expensive di scr i mi nant for t hol ei i t i c and calc-
alkaline rocks, and can be appl i ed on a single sampl e basis; however, affinities
shoul d be conf i r med by ot her tests on a port i on of t he sample set.
CH EMI CAL ANALYSES
Chemi cal analyses for t he H FSE must be precise and accurat e in or der to
distinguish t he small differences bet ween fract i onat i on t rends and el ement
ratios of i ndi vi dual samples. In our experi ence, only t he dat a obt ai ned for Zr,
Y and Nb by X-ray fl uorescence using pressed rock powder pellets are pres-
ently suitable for establishment of magmat i c affinity and fractionation trends.
Wi t h this t echni que t here is nei t her t he sampl e di l ut i on probl em i nher ent in
borat e bead preparat i on, nor t he probl em of i ncompl et e solution in acid
digestion met hods. Good qual i t y analyses for REE and less abundant H FSE
are obt ai ned by i nst r ument al neut r on act i vat i on analyses ( I NAA) , or a com-
bi nat i on of i nduced coupl ed pl asma and mass spect romet ry ( I CP- MS)
t echni ques.
SINGLE PRECURSOR
Al t erat i on of a chemi cal l y homogeneous rock produces a single precursor
system. A plot of an i mmobi l e el ement pai r i n such a uni t forms an "altera-
t i on l i ne" t hat is l i near and passes t hrough t he origin (Fig. 1 ). For i mmobi l e-
i ncompat i bl e el ement pairs, alteration and fract i onat i on lines coi nci de for rock
types of t he same magmat i c affinity (Fig. 4a) . W hen t he pairs are compat i -
bl e- i ncompat i bl e, or bot h are compat i bl e, a separate al t erat i on line is formed
for each chemi cal l y di st i nct homogeneous rock uni t (Fig. 4b ). Single precur-
sor al t erat i on lines are t herefore very useful i n di st i ngui shi ng homogeneous
vol cani c uni t s and correl at i ng t hem bet ween drill holes and t hrough altera-
t i on zones.
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS
100
8 0
E 8o
Q .
>" 40
20
A n s i l v o l c a n i c r o c k s :
o l eas t a l t e r e d . /
_. a l t e r e d .
. . ~ , f p
o , f ~ . , " . " o
e l 0 o

O o0" ~1~" ' ' u " "
~ , ' , : ,

i El
I I I
0 100 200 300 400
Z r p p m
115
30
S c h e m a t i c a l t e r a t i o n l i n e s
Two r hvol i t e uni t s A and B
2 s / ~ / ~
Pr ec ur s or B , r
2 0 I . i r e ' ' ` "
O ~ D" , t L /
, o / " - j " I
/ A ~ . . , " Pr e c u r s o r A
5 . : 2 : , , 2 : : / "
~. b
0 I I I I I I I
0 100 200 300 400
Z r p p m
Fig. 4. ( a) Y - Z r pl ot of least al tered and al tered samples f r om the Ansi ] mi ne (see Fig. 2).
I mmobi l e- i ncompat i bl e element pairs i n least altered and altered samples plot along the same
regression line. ( b) Alteration lines for i mmobi l e compat i bl e-i ncompat i bl e element pair can
identify two di st i nct rhyolite units, A and B, wi t hi n a mi ne stratigraphy.
116 W. H . Ma c LEAN AN[ ) T.,I BARRETT
MULTIPLE PRECURSORS
Fractionation trends
Once an H FSE is shown t o be hi ghl y i ncompat i bl e and i mmobi l e, it can be
used i nst ead o f t he mobi l e moni t or s (e. g SiO2, Mg- number , di f f er ent i at i on
i ndex) whi ch suppl y onl y s emi - quant i t at i ve est i mat es of degree of i gneous
f r act i onat i on. The H FSE are quant i t at i ve moni t or s because of t hei r hi gh
i ncompat i bi l i t y.
It is i mpor t a nt t o est abl i sh t he f r act i onat i on lines as accur at el y as possible,
as t hey f or m t he basi s o f t he pr ecur sor i dent i f i cat i on and t he mass change
cal cul at i ons. Pl ut oni c rocks, such as t onal i t es and gabbros, can be i ncl uded in
t he sampl e set i f fi el d r el at i ons i ndi cat e t hat t hey are synvol cani c. St rongl y
f el dspar - por phyr i t i c rocks shoul d be excl uded f r om Al -beari ng pl ot s, owi ng
t o pl agi ocl ase accumul at i on and r esul t i ng hi gh A1 cont ent s. Ther e is al ways a
cer t ai n a mo u n t o f scat t er i n t he l east al t er ed sampl e dat a used t o fit t he frac-
t i onat i on lines. W i t h few dat a at t he st art of expl or at i on, t he l i ne fits are ap-
2.0
,,e
~ Schemati c thol ei i ti c
e e fracti onati on trend
/
' . . . . . . Al terati on l i nes
0 1.o.
,4 "
X
0.5" / o . . . . - ' " x ~ IRhyoliteJ
, ' 1 I " " _ . . , , . . o O - . - ~ ' ~
J ~ - ' " ~0._ . 0- - ' 0" 0
, . . - _ . . . . . . . . . . a
0 . 0 " " " ' ' - - . , . , ,
0 1 0 0 2 0 0 3 0 0 4 0 0
Z r p p m
Fig. 5. Idealized fractionation curves for immobile elements in a tholeiitic volcanic suite, with
superposed chemical data for altered rocks. The two well defined alteration lines represent ho-
mogeneous units within the stratigraphy. Their intersection with the fractionation curves iden-
tifies the units as an andesite and a rhyolite. Other altered samples on these plots have each a
distinct alteration line emanating from the origin. (a) TiO2-Zr: The fraetionation curve ex-
hibits strong Ti-enrichment in the basalts (paralleling typical iron-enrichment in tholeiites),
followed by strong Ti depletion from andesite to rhyolite.
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS 117
o ~
O3
o
<:
20
15
10
t'
0
I "
/ Q o .
I g a ~ L I I t l o s
0 X i s ~ , . s O
" ~ ' ~ A n d e s l t e L _ ~ ~.,- o
/ O X / s --
I S S s ~ O
, , Z, ", . , , ' " . . . 4 " ~ . . 9. ~ " ~" ~ ' b
s ~,I
/ , / / : , , , ~ Schemat i c thol ei i ti c
" f ract i onat i on t rend
i ~,s~o ,S
'~,,.'," - ..... Al t erat i on l i nes
r
I I I
100 2 0 0 3 0 0 4 0 0
Z r p p m
2 0
15
~ I0
I
o
#
/o
I R h y o " t e l ~ . , . ~ D a c i t e ] ' ~ / , . , , , , 4
s b ,,,, q,,"
f s s S
s S
s / $6 S
s S s o s S
~ / / ' / /,m x ' ' $ ' e "
, S , X ~ - - . s J . . . f - . . ~
t / ~ d A n d e s l t e e ~
C
5 Jo /
I I I I I
T
/ / / . . . ' / "
I
Schemat i c thol ei i ti c
/ , , ; - ' " / f ract i onat i on t rend
t s s s .
~,~'.. . . . . . . Alterat0on hnes
0'
0. 0 0. 5 1. 0 1. 5 2. 0
TiO2 %
Fig. 5. (b) AI203-Zr: The fract i onat i on curve is near linear. Plagioclase accumul at i on may pro-
duce anomalously high A1 values in some basalts and andesites. (c) A1203-TiO2: The fraction-
ation curve exhibits a sharp turn owing to Ti -enri chment in the basalts and Ti-depletion from
andesite to rhyolite. This may lead to two intersections by alteration lines. However, this is
normally not a probl em as this plot is used for calc-alkaline rocks, where basalts are not common
(and in any case show low Ti enri chment ).
1 1 8 W.H. MacLEAN AND T.J. BARRETT
pr oxi mat e, but even t hese are super i or t o havi ng no gui di ng f r act i onat i on
t r ends at all. As expl or at i on proceeds, and mor e least al t ered sampl e dat a be-
come available, t he l i ne fits shoul d i mpr ove.
In t hol ei i t i c vol cani c suites, t he shapes of i deal Fe O- Zr and Ti O2- Zr (Fig.
5a) f r act i onat i on curves exhi bi t i ni t i al st rong enr i chment s (D << 1 ) in ba-
salts; t hi s is t he t ypi cal i r on- enr i chment t r end for rocks of t hol ei i t i c affinity.
The e nr i c hme nt is fol l owed by a change- over t o curves wi t h concave deple-
t i on ( D> 1 ) f r om andesi t e t o rhyol i t e. Mg O- Zr and Ca O- Zr f r act i onat i on
curves are concave ( D> 1 ) over t he full range of basal t t o rhyol i t e. By con-
trast, AI 203- Zr is near l i near wi t h a slightly negat i ve sl ope ( D~ 1 ) (Fig. 5b ).
Si O2- Zr is near l i near f r om basal t t o daci t e wi t h a posi t i ve sl ope ( D< 1 ), but
fl at t ens of f (D ~ 1 ) i n t he rhyol i t es due t o quar t z pr eci pi t at i on.
Curve f i t t i ng
Usi ng least al t ered sampl es, chemi cal dat a for compat i bl e el ement s exhib-
i t i ng concave depl et i on- t ype f r act i onat i on can be fit usi ng exponent i al or sec-
ond- or der pol ynomi al curves. Bot h curves c ommonl y gi ve si mi l ar dat a fits,
as expressed by r val ues for t he cur ve fit. A pol ynomi al shoul d be fit t o t he
i mmobi l e el ement profi l e ( Al 203- Zr or Ti O2- Zr ) used t o est abl i sh t he pre-
cur sor Zr moni t or val ues for al t er ed sampl es (see bel ow) . W i t h t hi s f or m of
equat i on, it is si mpl er t o cal cul at e t he i nt er sect i on of f r act i onat i on and alter-
at i on lines.
Ti O2- Zr and Fe O- Zr dat a for least al t er ed basal t s can be fit wi t h a separat e
set of curves. Fr act i onat i on t r ends for t hol ei i t i c rocks rise st eepl y in t he basal t
t o basal t i c- andesi t e range, but follow a depl et i on t r end f r om andesi t e t o rhy-
olite. In t he basal t t o basal t i c-andesi t e range, t he t r ends are sub-parallel t o
al t er at i on lines emanat i ng f r om t he ori gi n and, t herefore, not useful in mass
change calculations. I n t hi s composi t i onal range, it is si mpl er t o use t he A1203-
Zr f r act i onat i on cur ve t o cal cul at e mass changes, as al t er at i on and fract i ona-
t i on lines i nt ersect at hi gh angles.
Multiple fractionation curves
In some vol cani c sequences, t wo or mor e f r act i onat i on t r ends can be recog-
ni sed on a Ti O2- Zr pl ot , even t hough separat e t r ends are not appar ent in
pl ot s of ot her maj or oxi des versus Zr. Usual l y, t he separat e t r ends are also
evi dent i n Y- Zr and ot her H FSE pl ot s, and i ndi cat e t he presence of mor e
t han one ma gma t ype as t he source of t he vol cani c rocks i n t he st rat i graphy.
I n all such cases, t he sampl e set shoul d be di vi ded first i nt o subset s based on
t hese t rends. It is for t he H FSE pl ot s, i n part i cul ar, t hat hi gh qual i t y chemi cal
analyses for Zr, Y and Nb are r equi r ed, especi al l y i n t he basal t range wher e
t hei r cont ent s are low and t he f r act i onat i on l i nes converge.
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS 1 19
W hen al t ered samples are superi mposed on t he Ti O2-Zr and Al203-Zr plots,
t hei r l ocat i ons must be consi der ed i n rel at i on t o t he appropri at e fract i onat i on
curve, or mass change cal cul at i ons will be i n error. Thi s emphasi zes t he im-
por t ance of separat i ng t he al t ered sampl es i nt o t hei r pr oper affinity groups.
Anomalous Zr and other HFSE behavior
In some t hol ei i t i c-t ransi t i onal suites, Zr becomes moder at el y depl et ed i n
t he rhyolites, while t he ot her H FSE most l y r emai n i ncompat i bl e ( Kr ani di -
otis, 1985 ). This is usually det ect ed on t he Y- Zr plot, al t hough t he scatter of
dat a coul d be i nt er pr et ed as ei t her Zr loss or Y enr i chment . The probl em is
resol ved using plots of Nb - Zr and Y- Nb, whi ch isolate t he depl et ed compo-
nent . Zr-depl et ed rhyolites are locally present at t he Phel ps Dodge ( Kr ani di -
otis, 1985 ) and Al der mac (Barret t et al., 199 l b) VMS deposits.
ALTE R ATI ON LI NE S
In plots of i mmobi l e compat i bl e- i ncompat i bl e el ement pairs, samples t hat
initially lay along a fract i onat i on curve may be di l ut ed or concent r at ed by
mobi l e mass change. All such sampl e poi nt s will be di spl aced from t he frac-
t i onat i on curve and each will creat e a l i near "al t er at i on l i ne" t hat passes
t hrough t he origin ( MacLean, 1990). The precursor composi t i on for each
sampl e lies at t he i nt ersect i on of t he al t erat i on and fract i onat i on lines i n each
plot using an i mmobi l e compat i bl e- i ncompat i bl e el ement pair.
In plots wi t h mobi l e el ement s versus Zr (e.g. SiO2-Zr, FeO- Zr , CaO- Zr ,
et c. ), t he mobi l e el ement s do not behave i n an orderl y fashion. Thei r dis-
pl acement f r om t he fract i onat i on t r end is r andom, and t hei r precursor com-
positions cannot be det er mi ned directly from t hei r l ocat i on relative t o t he
fract i onat i on curve. Inst ead, t he established precursor Zr-val ue ( f r om A1203-
Zr or Ti O2- Zr ) is subst i t ut ed i nt o t he mobi l e el ement - Zr fract i onat i on curve
equat i ons t o solve for t he precursor cont ent of each mobi l e el ement . A
spreadsheet comput er pr ogr am is conveni ent for these calculations.
The high angle of i nt ersect i on of t he al t erat i on and fract i onat i on lines for
i mmobi l e compat i bl e- i ncompat i bl e el ement pairs provi des a very useful tool
for i dent i fyi ng di st i nct i ve rock uni t s i n al t erat i on zones (Fig. 5b). As t he
number of samples i n t he dat a set is bui l t up wi t h cont i nued expl orat i on, al-
t erat i on lines for t hese uni t s become mor e evi dent . I ndi vi dual al t erat i on lines
i dent i fy di st i nct i ve rock uni t s and provi de a powerful tool for lithostrati-
graphi c correl at i on.
AI-Ti plots f or calc-alkaline rocks
In calc-alkaline rock suites, wher e t he H FSE commonl y are compat i bl e and
do not moni t or fract i onat i on well, an al t ernat i ve pr ocedur e is adopt ed t hat
120 W. H . MacLEAN AND T. L BARRETT
utilizes i mmobi l e, compat i bl e el ement s. Et her i dge et al. ( 1983 ) and Ri char ds
et al. ( 1989 ) as s umed t hat Al, Ti and Zr were i mmobi l e dur i ng al t er at i on and
me t a mor phi s m, and used t hese el ement s t o est i mat e mass changes in rock
syst ems t hey as s umed t o be homoge ne ous (i.e. single pr ecur sor ) . Berni er and
MacLean ( 1989) and Li aghat and MacLean ( 1992) empl oyed Al 203-Ti O2
pl ot s t o est abl i sh t he i mmobi l i t y of A1 and Ti i n single homogeneous precur-
sor uni t s, and t hen used t hese el ement s t o const r uct f r act i onat i on t r ends for
mul t i pl e pr ecur sor syst ems in associ at ed t hol ei i t i c and calc-alkaline suites.
On a pl ot of A1203-TiO2 (Fig. 5c), vol cani c rocks show Ti enr i chment in
t he basal t i c range of f r act i onat i on, whi ch is fol l owed by Ti depl et i on from
andesi t e t o rhyol i t e. Thi s pr oduces a shar p reversal in t he f r act i onat i on line.
As a result, st rongl y al t ered basal t s are not easily di st i ngui shed f r om ot her
l i t hol ogi es i n t hi s plot. Thi s is general l y not a seri ous pr obl em for calc-alkaline
vol cani c rocks whi ch are most l y i n t he andes i t e- r hyol i t e range. It woul d be
t r oubl es ome i n t he bi modal t hol ei i t i c vol cani c suites, but t hi s pl ot generally
is not used for such rocks. Cal c-al kal i ne rocks may have mor e scat t er in Al
al ong t he A1203-TiO2 f r act i onat i on line t han t hol ei i t i c- t r ansi t i onal rocks,
whi ch pr oduces larger er r or i n mass change cal cul at i ons.
MASS CHANGES
Mass changes can be cal cul at ed for each mobi l e el ement based on t he dilu-
t i on or concent r at i on of an i mmobi l e c ompone nt ( MacLean and Kr ani di ot i s,
1987 ). The cal cul at i ons have some si mi l ari t i es t o t hose of Gr esens (1967 )
and t he modi f i cat i ons by Babcock ( 1973 ) and Gr ant ( 1986 ), who used var-
i ous el ement s t o est i mat e mass and vol ume changes i n single pr ecur sor sys-
t ems. Regardl ess of t he me t hod empl oyed t o est abl i sh i mmobi l i t y, cert ai n ele-
ment s can be used as t he basi s for mass change cal cul at i ons i n bot h single and
mul t i pl e pr ecur sor al t er at i on systems. I n t he i mmobi l e el ement pr ocedur es
we use ( out l i ned bel ow) , vol ume changes do not ent er i nt o mass change cal-
cul at i ons. Vol ume changes can, however , be cal cul at ed f r om mass change
results.
The mass of a sampl e aft er al t er at i on ( r econst r uct ed compos i t i on, R. C. ) is
cal cul at ed f r om t he i ni t i al chemi cal analysis on t he basis of 100 mass uni t s
( mass uni t s and wt% val ues are i nt er changeabl e ) of pr ecur sor rocks, usi ng Zr
as t he i mmobi l e moni t or :
R. C. = Zrprecu . . . . /Zraltered X O/ocomponentaltered ( 2)
Tot al mass change and mass changes for each el ement are cal cul at ed as:
Mass Change = R. C. - Pr ecur sor Compos i t i on ( 3 )
I mmobi l e c ompone nt s i n t he chemi cal analysis of t he al t ered sampl e in Fig.
LI TH OGEOCH EMI CAL TECH NI QUES USI NG IMMOBILE ELEMENTS 121
6 hav e appar entl y been d ecr eased by one- t hi r d r elative to t he pr ecur sor anal-
ysis. T he r econstr ucted c omp os i t i on af ter al ter ati on, wi t h t he i mmobi l e com-
p onent s no w r estor ed to thei r i ni ti al abund ances , is 150 uni t s. Thi s y i el d s a
mobi l e mas s gai n o f 50 uni t s ( i . e. , 50 wt %) .
In a mul t i p l e pr ecur sor sy stem, each alter ed samp l e is related to its partic-
ular pr ecur sor as part o f t he mass change cal cul ati on. In smal l data sets, error
is i nt r od uced i nt o t he cal cul at i on i f the i gneous f r act i onat i on cur v e is not wel l
d ef i ned , or i f a si ngl e pr ecur sor is taken as r epr esentati v e of a gr oup o f sam-
pl es that i ni ti al l y had sl i ght comp os i t i onal v ar i ati on ( i . e. , r anged al ong a nar-
r ow sect i on of t he f r act i onat i on cur v e) . Ne w data sup p l i ed by ongoi ng ex p l o-
ration ref ine the cur ve equati ons, shar pen the comp osi t i onal ranges of v ol cani c
f or mat i ons, and d ef i ne t he al ter ati on l i nes i n greater detail.
Increases i n mass most l y result f r om si l i ci f i cati on or car boni ti z ati on as v oi d
i : : i : i , : ; ! i : : / i ~ i i : i :
A nalys is of A nalys is of
unal teredrock al teredrock i : i : i i i i i i i i :
(precursor) : ! i : : i i : ! : i i : : : i i i i : i i
1 0 0 . . . . . . . . . . . . . . . . . - i i ; i ~ , . : i ~
u n i t s ! i i i i i ! i i i i i l i l i ! i ! i ! i ! ~ ! i ! i ! i . . . . . . . . . . . i : : : i i i i i i ! i ! i ! ! ! : :
! i i ! ! ~ i i ! i i i i ! i i i i i i i i ! i l i ! i : i ~ i i i . . . . . . . : i i i i i ! i :
i i i i i i i i i i i i i ! i ! i i i i i i i i i i i ! ! i : : ~ : : . . . . i
i i i i i i i i i i i ! i i i i l i l i i i i i ! i : . . . . . . . . . . . .
i i ! i i i i i ~ ! i ! i i ! i i i i i i i i i : : : ~ : : : !
i i i i i i i i i ! i i i i i i i i ! i i i i i i i ~ i : ~ : : : : : : :
iiiiiiiiiii! ! iii! ii! iiiiiiii M o b i l e ::~:: ~::!:
~:~i~i~:~ components iiiiiii!!i~i!!!!!!i!i!!ii . . . .
: i i i i i i i : i i i i i i i i i i : i i i i ! i i i i . . . . . . : :
i ! i i i i i i i i i i i i i ! i l i i i i ! ! i i i ; i i i i i i i i i i i ! i i i i i : : : : :
iiiiiiiii!iiiiiiiiii~!!iiiii ~ : : : . . . . . : : : ~
~ iii, ! ~ iiiiiiiii~ , ~ iiiii~ iiiii
Reconstructed
composition
1
MASS
GA IN
" ~ . ~ ' 1 " , , ~ ' 5 ~ . . . . . . . - - "
, n ~ . ~ O ~ l ~ c o m p o n e n t s
, ; , , | - - - - - ~
. . . . . . . . . .
Fi g. 6. Illustration o f mass change cal cul ati ons f or an altered rock sampl e, based on 100 uni ts o f
precursor. T he amount s o f i mmobi l e el ement s i n the chemi cal anal ysi s o f the altered rock ap-
pear to have decreased, but s i mp l y hav e been d i l ut ed by ad d ed mobi l e mass. Si nce the i mmobi l e
el ement s hav e not lost mass, they mus t be restored to their i ni ti al content, wi t h the other c o m-
p o n e n t s restored by the same factor. Thi s pr oduces the reconstructed comp os i t i on ( R. C. ) . Mass
change f or each comp onent is the di f f er ence between R. C. and the precursor comp osi t i on.
122 W . H . Ma c LEAN AND T. J. BARRETT
o~
O
q
o
Z
4 -
O
O
6
4
2
0
- 2
- 4
- 6
- 8
- 1 0
Ma s s changes in host vol cani cs, Cor bet mi ne
I Flavrian andesite
o N o r t h w e s t r h y o l i t e
+ Rhyolite dykes in FA
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
o ~
~ e e F e l d s p a r
o~, breakdown Silica
~ i _ . o l o - l e a c h i n g
a
i I i
2 O - 1 0 1 0 2 0 0
A F e O + M g O
4
<~ 2
- 2
M a s s c h a n g e s i n f e l s i c h o s t v o l c a n i c s , H o r n e m i n e
T o t a l s e r l c l t l z a t l o n
o f p l a g l o c l a s e In r h y o l i t e
,~
!
' S e r l c l t e
t
i
, + s i l i c a
i
" e ~ e y e e
S e r l c i t e I
~ O0 0
+ c h l o r l t ~ " ~ 0 d ~ e / ~ 4 r .
, % . . . / .
, ~ . , - / - " .
', ~ S l l l C i f i c a t l o n
~ , . . : . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
' F
b No a l t e r a t i o n
i I i I l
4 0 - 2 0 0 2 0 4 0 6 0
AS i 0 2 ( %)
8 0
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS 123
fillings ( amygdal es , vei nl et s, a nd ma t r i x t o br ecci as, t uf t s a nd s edi ment s ). I n
t hese case, t he over al l vol ume o f t he vol cani c uni t s gener al l y is r et ai ned. Mass
loss is c o mmo n l y due t o l eachi ng o f Si, Na a nd Ca dur i ng chl or i t i zat i on a nd
ser i ci t i zat i on, but can i nvol ve Fe, Mn, Mg or K loss as well. The loss is recog-
ni s ed by i ncr eas ed pr opor t i ons o f i ner t mi ner al s such as chl or i t e a nd seri ci t e,
shr i nkage cracks, a nd slight ar cuat e depr essi ons o f t he or e- f oot wal l cont act
over al t er at i on pi pes. Mass loss ma y be di f f i cul t t o vi sual l y r ecogni se on a
met er - scal e (pi l l ows, f l ows) . For exampl e, a 10% vol ume loss pr oduces onl y
a 3% decr ease i n t he r adi us o f a spher i cal basal t pillow.
Some mobi l e c ompone nt s (Si, K, Ca ) r e mo v e d f r om zones of mas s loss ar e
t r ans f er r ed t o t he out er par t s of al t er at i on zones, s ome cont r i but e t o t he or e-
body ( Fe, S, base met al s ) a nd mi ner al i zed cher t beds (Si, Fe, S) , a nd ot her s
( Ca, Na ) ar e wi del y di sper sed i n seawat er . Muc h of t he al t er at i on i n pi pes,
par t i cul ar l y i n r hyol i t e, r epr es ent s al t er at i on o f vol cani c glass, whi ch is ver y
r eact i ve ( t he sol ubi l i t y o f silica i n t he a mor phous st at e is mu c h hi gher t ha n
f or qua r t z ) . De pl e t i on of silica i n t he cor es of al t er at i on pi pes resul t s f r om
t he hi gh sol ubi l i t y o f a mor phous silica, as well as f r om i ncr eases i n t emper a-
t ur e a nd wa t e r / r o c k rat i os.
Mass changes can be r eadi l y i l l ust r at ed on bi nar y a nd t e r na r y plots. A use-
ful di agr am f or moni t or i ng al t er at i on is A ( Fe O + MgO ) vs. A ( Ca O + Na 2 0 ) ,
whi ch shows i ncr easi ng chl or i t i zat i on a nd Fe + Mg a ddi t i on as t he f el dspar
c o mp o n e n t o f t he r ock is des t r oyed (Fig. 7a ). A pl ot o f ASiO2 vs. AK20 (Fig.
7b ) di st i ngui shes ser i ci t i zed r ocks wi t h silica addi t i on f r om t hose wi t h silica
loss. I n bot h t ypes o f pl ot s, silica loss is i mpor t a nt as a gui de t o hydr ot he r ma l
l eachi ng, whi ch occur s i n t he mos t i nt ense par t s o f t he hydr ot he r ma l syst em.
Cal cul at ed mas s changes can be e xa mi ne d al ong i ndi vi dual dri l l hol es, or con-
t our e d i n ver t i cal a nd pl an sect i ons i f suf f i ci ent dat a ar e avai l abl e. The mass
changes al so r eveal di f f er ences wi t hi n br oa d al t er at i on zones t hat ar e mi ner -
al ogi caUy si mi l ar (e.g. ser i ci t i zed zones ) . Tr e nds i n mas s changes acr oss l i t h-
ol ogi cal uni t s or on a pr ope r t y scal e ar e t hus usef ul i n f ocusi ng expl or at i on
pr ogr ams.
ALTERATION MINERALOGY
Abunda nc e s o f al t er at i on mi ner al s i n dri l l cor e pr ovi de a si mpl e a nd di r ect
me a ns t o por t r a y t he ext ent a nd i nt ensi t y of hydr ot he r ma l al t er at i on. It is,
Fig. 7. Plots of mass changes for selected elements. (b) Correlation of mass changes in K20 and
SiO2 in the alteration at the Home mine (data from MacLean and Hoy, 1991 ). The dominant
alteration was a sericitization accompanied by silicification. Small zones of chlorite-sericite
alteration, accompanied by leaching of silica, occur adjacent to the Cu-bearing VMS ores. (a)
Large losses of CaO + Na20 mass during chloritization of andesite in the footwall of the Corbet
mine are accompanied by major additions of FeO + MgO mass.
124 W . H . MacLEAN AND T J , BARRETT
Horne volcanics
Qu a r t z
, ~ Quartz
/ ' , . . ' k g a i ~
~ t ." 2', \
a
0 2 0 4 0 6 0 8 0 1 0 0
S e r i c i t e A l b i t e
Mo n t b r a v r h y o l i t e
b
o3
c
O
Qu a r t z
altered
/ , , , , , , , f " e \ " l \ ~ atelyaltered
~" e e " " \ o \ S t r o n g l y a l t e r e d
S e r i c i t i t e e " , / ~ k ~ c h . l O r i t i t e
i ~ i " ~ i 0 0 -" ~" I ~ ' ~ w l ~ =
2 0 4 0 6 0 8 0 1 0 0
S e r i c i t e C h l o r i t e
LI TH OGEOCH EMI CAL TECH NIQUES USI NG IMMOBILE ELEMENTS 12 5
however , di ffi cul t and t i me c ons umi ng t o meas ur e t he modal pr opor t i ons of
al t er at i on mi ner al s i n t hi n- sect i on, or by expensi ve me t hods such as X-ray
di f f r act i on or X-ray i magi ng. An equal l y good est i mat e can be ma de by cal-
cul at i ng nor ma t i ve al t er at i on mi ner al abundances f r om bul k rock chemi cal
dat a.
I n pri nci pl e, a mi ner al assembl age por t r ays t he chemi cal compos i t i on of a
rock f or med unde r specific physi o- chemi cal condi t i ons. Al t er at i on ar ound
VMS deposi t s f or ms unde r low pr essur e ( seaf l oor ) greenschi st facies condi -
t i ons ( 2 0 0 - 4 5 0 C) , pr oduci ng al t er at i on mi ner al assembl ages consi st i ng
mai nl y of quart z, chl ori t e, sericite, albite, epi dot e, calcite and sul phi de. We
use t he Ni ggl i -Bart h cat i on nor ma t i ve cal cul at i on pr ocedur e (Bart h, 1962),
whi ch is readi l y adapt abl e t o a c omput e r spreadsheet , and r epr esent s modal
per cent age mor e closely t han a wei ght per cent nor mat i ve cal cul at i on. Car-
bonat e and sul phi de are c omput e d when CO2 and S analyses are available, or
are assi gned as residuals. Al umi no- si l i cat e clays, talc, and ot her mi ner al s are
c omput e d when appr opr i at e. Compa r i s on of nor mat i ve results wi t h modal
est i mat es i ndi cat es t he me t hod is reliable. The nor mat i ve dat a are t hen used
t o const r uct t er nar y al t er at i on mi ner al pl ot s (Fig. 8a, b) , or t o pr oduce con-
t our ed al t er at i on mi ner al i nt ensi t y maps. Nor ma t i ve mi ner al abundances are
also useful i n t he i nt er pr et at i on of oxygen i sot ope dat a ( MacLean and H oy,
1991, Barret t and MacLean, 199 ! ).
H ydr ot he r ma l al t er at i on mi ner al s do not readi l y i nt er act or recrystallize
dur i ng subsequent regi onal greenschi st facies me t a mor phi s m, whi ch occurs
unde r si mi l ar t emper at ur e condi t i ons. Bul k chemi st r y, ot her t han loss of vol-
atiles (wat er, CO2, S), does not appear t o be not i ceabl y changed by amphi -
bol i t e facies me t a mor phi s m, whi ch conver t s al t er at i on mi ner al ogy t o cordi er-
i t e - a nt hophyl l i t e - ga r ne t - bi ot i t e assembl ages ( Ri ver i n and H odgson, 1980;
Ber ni er et al., 1987; Berni er and MacLean, 1989 ). Nor ma t i ve al t er at i on mi n-
erals pr ovi de a conveni ent means t o compar e al t er at i on zones i n greenschi st
facies and hi gher grade me t a mor phi c t erranes.
NORMATIVE MINERAL DIAGRAMS
H ydr ot he r ma l al t er at i on of sodi c rhyol i t e at t he H o me Mi ne is i l l ust rat ed
i n Fig. 8a. As al t er at i on progressed, Na and Ca were l eached f r om al bi t i c feld-
spar and rhyol i t e glass, and t he A1-Si r esi due c ombi ne d wi t h Fe + Mg and K
f r om t he r ock and hydr ot her mal fl ui d t o f or m chl ori t e and sericite, respec-
Fig. 8. Triangular plots of cation normative alteration minerals. (a) Quartz-albite-sericite: The
diagram illustrates the sericitization and silicification of rhyolite in the alteration zone at the
Home mine (from MacLean and Hoy, 1991 ). (b) Quartz-chiorite-sericite: Chlorite and seri-
cite, formed by alteration of plagioclase components in Montbray rhyolite, were subsequently
concentrated by wholescale leaching of silica (from Barrett and MacLean, 1991 ).
126
o~
c-
O
~
a~
D
<
50
40
30
20
10
W.H. MacLEAN AND T.J. BARRETT
Hor ne vol cani ~ ro~ks
Seri ci t e a
Ch
Precursor r
i i !
0 40 6O 8O
Si cat i on %
Quartz
m
w
O0
0
, m
I00{
8 0 '
60
401
20
Quartz
h
Mo n t b r a v r h v o l i t e s
b
ee ~ Weakly to moderately
0 . . . . . . . . ~" ~ altered rhyolites
Pr ecur sor - . . . ~ , ~
. ~ "
hl ori t e
a l t e r e d r h y o l i t e s
i i i !
0 10 20 30 40 50
( Fe+Mg) cat i on %
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS 127
tively. Because A1 was i mmobi l e, t he abundance of chlorite and sericite (bot h
Al-bearing) was l i mi t ed t o t he initial Al -cont ent of t he rhyolite. The least al-
t ered rhyolite samples plot close t o t he al bi t e-chl ori t e tieline, but alteration
has shi ft ed ot her samples towards t he seri ci t e-chl ori t e boundary, with a ma-
j ori t y of t hem near t he sericite apex. A plot of quar t z- ser i ci t e- chl or i t e reveals
t hat t he domi nant alteration assemblage is quartz-sericite (MacLean and Hoy,
1991).
In t he early stage of al t erat i on of albite t o chlorite and sericite at low wat er /
rock ratios, rhyolite retains most of its initial quartz cont ent ( ~ 50%). Wi t h
i ncreased alteration, this assemblage can be reduced to sericite and chlorite,
as occurs i n alteration zones i n t he Mont bray rhyolite near Nor anda (Barrett
and MacLean, 1991 ). Resi dual concent rat i on of these two mi neral s is caused
by wholescale hydr ot her mal di ssol ut i on of quartz at high wat er / r ock ratios
(Fig. 8b). Extremely al t ered rhyolites plot along t he chl ori t e-seri ci t e tieline.
Chlorite and sericite are stable under t he alteration condi t i ons of VMS de-
posits i n greenst one belts (Ri veri n and H odgson, 1981 ), whereas AI-Si clay
mi neral s are not. Ot her stable phases i ncl ude rutile, zircon and REE-bearing
minerals.
BINARY CATI ON DI AGRAMS
Alteration of volcanic rocks is also well port rayed on binary cation dia-
grams. Quartz, albite, chlorite, sericite and ot her mi neral s can be pl ot t ed as
mi neral formul ae along with cat i on percent values of el ement s i n t he altered
rocks. Samples of silicified rhyol i t e at t he H ome mi ne plot bet ween t he pre-
cursor and quartz on an A1-Si di agram (Fig. 9a). Rhyolites affected by strong
silica-leaching mi grat e t owards t he seri ci t e-chl ori t e tieline.
The S i - ( Fe +Mg ) di agram (Fig. 9b) is useful i n revealing quartz loss and
t he inertness of chlorite and sericite. Initial alteration of t he Mont bray rhyo-
lite spread samples along an approxi mat el y const ant Si t rend, l i mi t ed by t he
quart z-seri ci t e and quar t z- chl or i t e tielines. Loss of Si t hen shifted t he sam-
ples t o t he seri ci t e-chl ori t e t i el i ne where t hey formed stable residues.
VOLUME CH ANGES
Generally we exami ne alteration i n t erms of absolute changes i n the amount
of mass t ransferred i nt o or out of a rock by hydr ot her mal fluids. This is mor e
Fig. 9. Pl ot s of percent ages of cat i ons in bul k rock chemi cal analyses. Mi neral formul ae are also
shown in t erms of cat i on component s. ( a) A1 vs. Si pl ot i l l ust rat i ng al t er at i on of rhyol i t e at the
H ome mine. Si l i ci fi ed sampl es pl ot bet ween precursor and quartz. The ser i ci t e- chl or i t e t i el i ne
forms a bar r i er to furt her al t er at i on ( dat a from MacLean and H oy, 1991 ) ( b) Si vs. ( Fe + Mg)
pl ot i l l ust rat i ng al t er at i on of t he Mont br ay rhyol i t e. Sericite and chl ori t e f or med by al t erat i on
of plagioclase are concent rat ed along the seri ci t e-chl ori t e tieline by wholesale leaching of quartz.
128 W.H. MacLEAN ANI)1 .J. BARRETT
i mpor t a nt t han vol ume changes as a par amet er t o exami ne and i nt er pr et hy-
dr ot her mal processes. Of course, vol ume will usual l y change as mass of t he
rock changes, but an ext ra st ep is r equi r ed t o cal cul at e t hi s variable.
Changes in rock vol ume owi ng t o al t er at i on can be est i mat ed f r om mea-
sured pr ecur sor and al t ered rock densi t i es and t he meas ur ed rat i os of an im-
mobi l e el ement , e.g., Zrprecursor / Zral t ered ( Shr i ver and MacLean, 1993 ). Rock
densi t i es and vol ume changes can also be est i mat ed closely from nor mat i ve
mi ner al cont ent s of pr ecur sor s and r econst i t ut ed sampl es, and mi ner al den-
sities. Absol ut e vol ume of a r ock uni t may not change as a result of voi d filling
or voi d f or mat i on, al t hough t he vol umes of i ndi vi dual mi neral s change. Many
al t er at i on mi ner al s have fi xed or nar r ow ranges in densi t y (quart z, albite,
sericite, epi dot e) , but chl ori t e has a wi der range because of Fe - Mg subst i t u-
t i on ( chl or i t e densi t i es can be es t i mat ed f r om mi cr opr obe mi ner al analyses ).
Vol ume loss in st rongl y chl or i t i zed rhyol i t e at t he Phel ps Dodge VMS de-
posi t was as muc h as 60% ( MacLean and Kr ani di ot i s, 1987). By cont rast , a
10% vol ume i ncrease was es t i mat ed for t he bul k of t he H or ne mi ne al t erat i on
zone, wher e si l i ci fi cat i on was wi despr ead ( MacLean and H oy, 1991 ).
EFFECT ON <5180 VALUES
Mass changes also affect t he i nt er pr et at i on of bul k c~80 dat a for al t ered
rocks ( MacLean and H oy, 1991; Barret t and MacLean, 1991 ). Not all whol e
rock c$180 var i at i on is caused by changes i n t emper at ur e and ~t 80 of t he hy-
dr ot her mal fluid. For exampl e, t he bul k c$180 of an al t ered rock c ompos e d of
quar t z (~ 180 ~ 10%o) and chl ori t e (c5180 ~ 20/00 ) will be l ower ed by l eachi ng
of quar t z at cons t ant T and ~t 80 ~ 0%0 ( f l ui d) (see MacLean and H oy, 1991
for cal cul at i on pr ocedur es ) . Compl et e l eachi ng of t he quar t z under t hese con-
di t i ons woul d pr oduce a chl or i t i t e r ock wi t h a bul k ~ 180 ~ 2%0. Similarly, sil-
icification increases bul k c~180 values, wi t h t he hi gher values pr oduced at lower
al t erat i on t emper at ur es. As a resul t of t hese effects, c$180 val ues of al t ered
rocks shoul d be assessed in light of mass changes as well as t emper at ur e and
fl ui d 6180 values.
MOBI LI TY OF TH E REE
The REE are c ommonl y i ncompat i bl e i n i gneous processes, part i cul arl y in
t hol ei i t i c- t r ansi t i onal suites, but onl y t he heavy REE ( Lu and Yb) are consis-
t ent l y i mmobi l e dur i ng hydr ot her mal al t erat i on (MacLean, 1988 ). The heavy
REE pr oduce l i near al t er at i on lines wi t h Zr and ot her i mmobi l e el ement s, but
t he l i ght -REE dat a f r equent l y scatter, i ndi cat i ng some mobi l i t y. Such mobi l -
ity may cause changes i n t he slopes of chondr i t e- nor mal i zed REE profiles (e.g.
LaN/YbN r at i os) , and t hus i nt r oduce errors i n t he det er mi nat i on of mag-
mat i c affinity. For t hi s r eason it is essent i al t o first demons t r at e i mmobi l i t y
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS 129
of all t he REE i n a suite of al t ered rocks before using t he slopes of t he REE
pat t erns t o obt ai n pet rogenet i c i nformat i on. This is done by pl ot t i ng each
REE against an i mmobi l e el ement such as Zr.
If t he REE have been i mmobi l e, t he addi t i on of mass t o a rock causes REE
profiles i n a single precursor al t erat i on system t o shift vertically down on
chondr i t e- nor mal i zed diagrams, whereas t he removal of mass causes REE
profiles t o shift vertically upwards. I f t he REE also have been mobi l e, it is
difficult t o distinguish these vari ous effects. To eval uat e these changes, t he
profiles are rest ored ( nor mal i zed) to t he original Lu (or Yb) value of t he
precursor. In some al t erat i on zones, as at Phel ps Dodge (MacLean, 1988)
and Ansil (Barret t et al., 1991a), this produces a fan-shaped array of profiles
wi t h Lu as t he hinge. La is t he most mobi l e of t he t ri val ent REE, wi t h t he
ot hers decreasi ng uni f or ml y i n mobi l i t y t owards i mmobi l e Lu (MacLean,
1988 ). A fan of profiles results from REE mobility, whi ch ranges from uni-
form addi t i on of t he lighter REE relative t o t he precursor, to uni f or m deple-
t i on of t he lighter REE.
Mobile REE depl et i on is encount ered in t he most intensely chloritized parts
of al t erat i on zones bel ow VMS deposits (Campbel l et al., 1984; MacLean,
1988, Barrett et al., 1991 ). H ydr ot her mal fluid t emperat ures are highest ( 300-
400C) i n these zones, whi ch were subject ed to silica depl et i on and com-
monl y cont ai n stringer-copper mi neral i zat i on. REE addi t i on is most l y in
nearby zones of sericite al t erat i on and silicification, where fluids were cooler
and preci pi t at ed REE minerals. Light-REE depl et i on anomal i es encount er ed
i n al t ered rock duri ng mi neral expl orat i on are i ndi cat i ve of cent ers of high
t emper at ur e hydr ot her mal fluid flow. REE anomal i es of ei t her type are im-
port ant to document from an expl orat i on vi ewpoi nt .
DISCUSSION
A hypot het i cal exampl e of a sequence of greenst one volcanic rocks rangi ng
f r om basalt t o rhyol i t e bei ng expl ored for VMS deposits is illustrated i n Fig.
10 to highlight rout i ne procedures used for lithological i dent i fi cat i ons and t he
i mpr i nt s of al t erat i on pr ovi ded by i mmobi l e el ement s. One drill hole encoun-
ters relatively unal t er ed (least al t ered) stratigraphy, and from sampl i ng data,
t he vol cani c uni t s are di vi ded i nt o two vol cani c affinity groups, one t hol ei i t i c
and t he ot her mi l dl y calc-alkaline, using a Y- Zr plot. The mai n lithologies
( bas al t - andes i t e- daci t e- r hyol i t e) are i dent i fi ed for each affinity type using
least al t ered samples and Ti OE-Zr and A12Oa-Zr plots; fract i onat i on lines are
fitted to each plot.
Geochemi cal dat a for t he least al t ered rocks (Fig. l 0) reveal t hat a tholei-
itic basalt to rhyol i t e sequence is overl ai n by mi l dl y calc-alkaline rhyol i t e t o
basalt units. Rhyol i t e- r hyol i t e cont act s of t wo types are present: bet ween
t hol ei i t i c rhyolites A and B (B is mor e fract i onat ed t han A) , and t hol ei i t i c
130 W.H. MacLEAN AND T.J. BARRETT
Chemost rat i graphy based on
l east al t ered sampl es from
surf ace and dri l l i ng t ransect s
Co n t a c t of I nt e r e f
Mi l d l y
TIO2%
1"5" 1 B a s ~ A n d Dac
i
Z r 500
c a l c - a l k a l i n e v o l c a n i c s e q u e n c e
AI203% Y ppm
1 5 - ~ ' X o o ~ . ~ y ~ ~
= 7 '
Zr 500 Zr 500 ppm
T h o l e i i t i c v o l c a n i c s e q u e n c e
TIO2% AI203% Y" ppm
1. 5- 15-
, J ,
Zr 500 Zr 500 Zr 500 ppm
Basal t to andesi t e I I Rhy ol i t e
Synvol cani c fel si c
Daci t e ~ i nt r usi on ( t hol ei i t i c)
~,, Fr act i onat i on t r ends
. . . . . . . . . . . . . Al t er at i on l i nes
St rongl y al t ered rocks
near massi ve sul fi de lens
] TiO2
:-_ - ~
,O~ Rhy A
F - ~ ' I
Zr 500
Zr 500
Al t e r e d mi l dl y c a l c - a l k a l i n e v o l c a n i c s e q u e n c e
TIO2% AI203% Y ppm
15 15
/ ~"s .And / ' , ' ' ? ~
l ' , " ~l l " " ~=. RhyC l ' , ~ '~ 'a'k\1
Zr 500 Zr 500 Zr 500 ppm
Al t e r e d t hol e i l t i c v o l c a n i c s e q u e n c e
TIO2% AI203/o Y ppm ~o O
Zr 5(~0 Zr 500 Zr 500 ppm
( ~ ) - ( ~) Rhyol i t e uni ts
LITHOGEOCHEMICAL TECHNIQUES USING IMMOBILE ELEMENTS 131
rhyolite B and mi l dl y calc-alkaline rhyolite C. Both types of contact are sites
of felsic-hosted VMS deposits i n t he Nor anda district, Quebec: t he former for
t he Al dermac ores (Barret t et al., 199 l b) , and t he latter for t he Mobrun and
Del bri dge ores (Barrett et al., 1991, 1993). A number of VMS deposits at
Nor anda also occur at andesi t e-rhyol i t e contacts (e.g., Ansil, Norbec, Millen-
bach: Gi bson and Watkinson, 1990; Chart rand and Cattalani, 1990). Usi ng
t he l i t hogeochemi cal fract i onat i on-al t erat i on relations out l i ned earlier, lith-
ological contacts can be traced from least altered stratigraphy t hrough altera-
t i on zones, and correlated across faults.
The possibility of a genetic relationship bet ween t he volcanic rocks and
i nt rusi ve bodi es (Fig. 10 ) also can be established with maj or (Gol di e, 1978 )
and trace (Paradi s et al., 1988) elements, using t he fractionation relations
and el ement al ratios descri bed above. A co-magmat i c i nt rusi on in t he upper
levels of a volcanic pile will provi de an i mpor t ant source of heat to dri ve
hydr ot her mal circulation.
It is possible to i dent i fy t he precursors lithologies of volcanic rocks that
have been intensely hydrot hermal l y altered and mi neral i zed t o the poi nt of
bei ng unrecognizable. The i mmobi l e el ement i mpr i nt not only yields t he pre-
cise rock type, but also t he magmat i c affinity. In t he Nor anda VMS deposits,
t he paths of alteration can be followed by mass change and nor mat i ve altera-
t i on mi neral plots. For example, silicification and sericitization are domi nant
processes i n rhyolite at t he H or ne mi ne (Figs. 7a, 8a, 9a; MacLean and Hoy,
1991; Barrett et al., 1991a); at t he Corbet mi ne plagioclase breakdown i n
andesi t e is combi ned with chl ori t i zat i on and silica leaching (Fig. 7b; Barrett
et al., 1993 ); and in t he Mont bray rhyolite ext reme mass loss, chloritization
and residual bui dup ofseri ci t e occurs i n mi neral i zed zones (Fig. 8b, 9b; Bar-
rett and MacLean, 1991 ).
SUMMARY
I mmobi l e el ement s provi de new l i t hogeochemi cal tools for mi neral explo-
ration. I mmobi l e HFSE are empl oyed to distinguish bet ween tholeiitic and
Fig. 10. Schematic example of lithogechemical procedures applied to an exploration drilling
program. In the upper figure, a drill hole intersects relatively unaltered volcanic stratigraphy.
Samples are first di vi ded using Zr / Y ratios i nt o an upper mildly calc-alkaline group and a lower
tholeiitic group (magmatic affinities should be confirmed with further trace element and REE
dat a). Fract i onat i on curves in Ti O2-Zr and A1203-Zr space are then established for each mag-
matic group. Core is generally sampled every 10-20 m. In the lower figure, a drill hole pene-
trates a strongly altered (light gray) part of the same stratigraphy in the vicinity of a VMS
deposit (black lens). Mass changes due to alteration displace samples of each volcanic uni t from
the fractionation curves along distinct alteration lines. Three rhyolite types ( A- C) are present,
as in the unaltered sequence, but the three rhyolites are more readily recognized in the altered
rocks. The felsic i nt rusi on in this example is of tholeiitic affinity and would be related to the
lower volcanic sequence.
132 W.H. MacLEAN AND TJ BARRETT
cal c- al kal i ne af f i ni t i es i n f resh and altered rock sui tes, and to moni t or frac-
t i onat i on of the comp at i bl e el ement s. The HFSE, AI and Ti f or m the basi s f or
quanti tati v e est i mat i on of mobi l e el ement mass changes. Severely altered rocks
that hav e l ost pr i mar y textur al, mi ner al ogi cal and chemi cal i denti ty can be
conf i d ent l y cl assi f i ed as to i ni ti al rock type and magmat i c af f i ni ty.
Maj or el ement chemi cal data r equir ed f or the cal cul at i ons and pl otti ng pro-
cedur es are av ai l abl e f r om r outi ne bul k rock chemi cal anal y ses, gener ally
av ai l abl e i n mi ner al ex p l or at i on pr ogr ams. Howev er , pr ecise and accurate
trace el ement anal y ses are necessar y f or the est abl i shment of rock af f i ni ti es,
and f r act i onat i on and al ter ati on l i nes. The l i t hogeochemi cal par ameter s pro-
d uced by these t echni ques can be p l otted on ex p l or at i on maps, cr oss- secti ons
and d i scr i mi nat i on di agr ams. The ex act i d ent i f i cat i on of rock ty pes and rock
af f i ni ti es al l ows cor r el ati on of v ol cani c uni ts bet ween drill hol es and across
z ones af f ected by hy d r ot her mal al ter ati on and f aulting. Thi s i nf or mat i on gen-
erates a comp r ehens i v e i nsi ght to v ol cani c l i thostr ati gr aphy, hy d r ot her mal
al ter ati on and or e- f or mi ng pr ocesses i n gr eenst one belts.
REFERENCES
Babcock, R.S., 1973. Comput at i onal models of" metasomatic processes. Lithos, 6: 279-290.
Barrett, T,J. and MacLean, W.H., 1991. Chemical, mass and oxygen isotope changes during
extreme hydrothermal alteration of an Archean rhyolite, Noranda, Quebec. Econ. Geol., 86:
406-414.
Barrett, T,J., MacLean, W.H., Cattalani, S., Hoy, k. and Riverin, G., 1991a. Massixe sulfide
deposits of the Noranda area, Quebec. Ill. The Ansil mine. Can. J. Earth Sci., 28: 1699-
1730.
Barrett, T.J., Cattalani, Chartrand. F., and Jones, P., 1991b. Massive sulfide deposits of the
Noranda area, Quebec. lI. The Aldermac mine. (,'an. J. Earth Sci., 28:1301 -1327.
Barrett, T.,1., Cattalani, S., Hoy, k., Riopcl, J. and Lafleur, P.-J., 1992. Massive sulfide deposits
of the Noranda area, Quebec. IV. The Mobrun mine. Can. J. Earth Sci., 29: 1349-1374.
Barrett, T.J., Cattalani, S. and MacLean, W.H., 1993. Volcanic lithogeochemistry and alteration
at the Delbridge massive sulfide deposit, Noranda, Quebec. In: E i . Hoffman and D.W.
Moore (Editors), Deep Exploration Using Lilhogeochemistry. J. Expl. Geochem., 48: 135-
173.
Barth, T.F.W., 1962. Theoretical Petrology. Wiley, New York, 416 pp.
Bernier, L.R. and MacLean, W.H., 1989. Auriferous chert, banded iron-formation and related
volcanogenic hydrothermat alteration, Atik Lake, Manitoba. Can. J. Earth Sci., 26: 2676-
2690.
Bernier, L.R., Pouliot, G. and MacLean, W.H, 1987. Geology and metamorphism of the Mon-
t auban North Gold Zone: A metamorphosed polymetallic exhalative deposit, Grenville
Province, Quebec. Econ. Geol., 82: 2076-2090.
Campbell, I.H., Lesher, C.M., Coad, P., Franklin, J.M., Gorton, M.P. and Thurston, P.C., t984.
Rare-earth element mobility in alteration pipes below massive Cu-Zn-sulfide deposits. Chem.
Geol., 45: 181-202.
Chartrand, F. and Cattalani, S., 1990. Massive sulfide deposits in Northwestern Quebec. In: M.
LI TH OGEOCH EMI CAL TECH NI QUES USI NG IMMOBILE ELEMENTS 133
Rive, P. Verpaelst, Y. Gagnon, J.M. Lulin, G. Riverin and A. Simard (Editors), The North-
west Quebec Polymetallic Belt. Can. Inst. Min. Metall. Spec. Vol., 43: 77-91.
Elliott-Meadows, S.R. and Appleyard, E.C., 1991. The alteration geochemistry and petrology of
the Lar Cu-Zn deposit, Lynn Lake area, Manitoba, Canada. Econ. Geol., 86: 486-505.
Etheridge, M.A., Wall, V.J., and Vernon, R.H., 1983. The role of fluid phase during regional
metamorphism. J. Metamorph. Geol., 1: 205-226.
Finlow-Bates, T. and Stumpfl, E.F., 1981. The behavior of the so-called immobile elements in
hydrothermally altered rocks associated with volcanogenic submarine-exhalative ore depos-
its. Miner. Deposita, 16: 319-328.
Gibson, H.L. and Watkinson, D.H., 1990. Volcanogenic massive sulphide deposits of the Nor-
anda Cauldron and Shield Volcano, Quebec. In: M. Rive, P. Verpaelst, Y. Gagnon, J.M.
Lulin, G. Riverin and A. Simard (Editors), The Northwest Quebec Polymetallic Belt. Can.
Inst. Min. Metall. Spec. Vol., 43:119-132.
Goldie, R.J., 1978. Consanguineous Archaean intrusive and extrusive rocks, Noranda, Quebec:
chemical similarities and differences. Precambrian Res., 9:275-287.
Grant, J.A., 1986. The isocon diagram - - a simple solution to Gresens' equation for metaso-
matic alteration. Econ. Geol., 81: 1976-1982.
Gresens, R.L., 1967. Composition-volume relationships of metasomatism. Chem. Geol., 2: 47-
55.
Kranidiotis, P., 1985. Geology, geochemistry and hydrothermal alteration at the Phelps Dodge
massive sulphide deposit, Matagami, Quebec. M.Sc. thesis, McGill University, Montreal,
166 pp.
Kranidiotis, P. and MacLean, W.H., 1987. Systematics of chlorite alteration at the Phelps Dodge
massive sulfide deposit, Matagami, Quebec. Econ. Geol., 82:1898-1911.
Lesher, C.M., Goodwin, A.M., Campbell, I.H. and Gorton, M.P., 1986. Trace element geo-
chemistry of ore-associated and barren, felsic metavolcanic rocks in the Superior province,
Canada. Can. J. Earth Sci., 23: 222-237.
Liaghat, S. and MacLean, W.H., 1992. The Key Tuffite, Matagami Mining District: origin of
the tuff components and mass changes. Explor. Min. Geol., 1:197-207.
MacLean, W.H., 1988. Rare earth element mobility at constant inter-REE ratios in the altera-
tion zone at the Phelps Dodge massive sulfide deposit, Matagami, Quebec. Miner. Deposita,
23: 231-238.
MacLean, W.H., 1990. Mass changes in altered rock series. Miner. Deposita, 25: 44-49.
MacLean, W.H. and Kranidiotis, P., 1987. Immobile elements as monitors of mass transfer in
hydrothermal alteration: Phelps Dodge massive sulfide deposit, Matagami, Quebec. Econ.
Geol., 82: 951-962.
MacLean, W.H. and Hoy, L.D., 1991. Geochemistry of hydrothermally altered rocks at the
Horne mine, Noranda, Quebec. Econ. Geol., 86: 506-528.
Paradis, S., Ludden, J. and Gelinas, L., 1988. Evidence for contrasting compositional spectra in
comagmatic intrusive and intrusive rocks of the late Archean Blake River Group, Abitibi,
Quebec. Can. J. Earth Sci., 25: 134-144.
Richards, H.G., Cann, J.R. and Jensenius, J., 1989. Mineralogical zonation and metasomatism
of the alteration pipes of Cyprus sulfide deposits. Econ. Geol., 84:91-115.
Riverin, G. and Hodgson, C.J., 1980. Wallrock alteration at the Millenbach Cu-Zn mine, Nor-
anda, Quebec. Econ. Geol., 75: 424-444.
Shriver, N.A. and MacLean, W.H., 1993. Mass, volume and chemical changes in the alteration
zone at the Norbec Mine, Noranda, Quebec. Miner. Deposita, in press.