- Email: [email protected]
The determination of sc, y, nd, ce and la in silicate rocks by a combined cation exchange-spectrochemical method
ANALYTICA
THE BY
DETERMINATION A COMBINED
OF CATION
CHIMICA
ACTA
SC, Y, Nd, Ce AND
355
La IN
SILICATE
EXCHANGE-SPECTROCHEMICAL
R. A. EDGE Dcpurfmerrl O/Gcochemislry.
AND
METHOD
L. H. AHRENS
University
(Rcccivcd
ROCKS
August
of Cape
Town
(SoutJr
Africa)
5th, rgG~)
The application of combined ion cschange enrichment and c1.c. arc spectrochemical methods for determining trace clemcnts which are not detectable by direct excitation in rocks, soils and meteorites has been discussed by EDGE et aZ.1, BROOKS et aZ.2 and AHRENS et aZ.3*4. Anion exchange enrichment has been used for estimating Cd, Zn, Sn, Bi and Tl (observations have also been maclc on Nb, MO, In, A6 and the platinum metals)znfi.fl. Cation eschange techniques 3.7eRhave been applied to tin, molybdenum, caesium and some rare earths and to a few elements detectable by direct d.c. arc excitation cg. scandium, gallium and rubidium. EDGE et al.1 have briefly outlined some of these investigations and AHKENS et al. 3 have described in detail the estimation of caesium in chondritic meteorites and basic rocks. In the present paper investigations are described on the combined use of cation exchange enrichment and spectrochemical analysis for determining SC, Y, Nd, Ce and La in common silicate rocks (granites and diabascs). Although combined cation exchange-spectrochemical analysis procedures have been employed for estimating traces of rare earths in zirconium metal” and uranium compounds’0 and trace amounts of yttrium in bioto@cal material+‘, no workers appear to have employed this technique for estimating rare earths in silicate rocks. EXPERIMENTAL
Afifiaratrrs
and reagents
Standard spectrographic equipment was used and is described below together with the excitation conditions: Spectrogaph: Hilger E49z large quartz and glass. External optics: Hilger Eg58 lens focusscd on slit; step sector (2 : I ratio). Slit length: II cm. Slit width: 0.001 mm. Wavelength range: 3800-5300 A. Glass optics. Photographic plate : Kodak 103-0. Electrodes: Upper electrode (- ve), Champion Ship carbon 0.5 cm diam. rod. Lower clcctrode (+ve), National carbon regular grade graphite, 2.4 mm int. diam. x 3 mm depth. Current: 7 A d.c. Esposure: to completion, usually about 60 sec. Sample: 6 lV HCl effluent resiclue and 5 rng 5% ZrOz-C mix. Photographic processing: 46 min at 20 in I
lint in those basic rock rare earth conccntratcv whcrc .-1ual. CJlirrr. rich,
26:( 1962)
355-362
R. A. EDGE, Ia. H. AHRENS
35’5
Ion exchange columns of length 38 cm and internal diameter 1.7 cm were used. The resin was the strongly acid Dowex 50 8X. 200-400 mesh, H-form. The following reagents and materials were also used: A.R. HCl (s.g. 1.18); A.R. HE (40%), A.R. HC104 (60%) ; deionised water; “specpure” YnOa, NdzOn, La203 and ZrOz and National Carbon S.P-2 graphite powder. Spectrochemical tests on the purity of the liquid reagents and the cation exchanger did not show detectable SC 4246.8, Y 4398.0, La 4333.7, Nd 4303.6 and Cc 4296.7. Volumes of reagents larger than those used in the actual analysis were taken to dry- _’ ness and examined spectrochemically. The resins were examined by arcing the ash from 5 g of resin.
Standards Natural silicate standards (granite G-I and diabase W-I (AHRDNS AND FLIXSCHER~~)) were used exclusively and were carried through the same dissolution, column and spectrographic procedures as the rock samples. The values of SC, Y, Ce, Nd and La used to establish the working curves are given in Table I.
AMOUNTS
FOR
I’HI~I’ARATION
(p&n.,
Slurtdnrd
G-I grunitc
W-I diabasc
-
--.--
OIr WORKING CC
CURVKS Nd
La
IP.P*rn.)
(p.p.*ll.)
21
600
IO0
150
30
50
50
30
-_-
(P.P.1n.l
-
Wherever possible the rccommencled values of AHIUZNS AND FLEISCHER~~ were employed. However, BERMAN’S~~ values for yttrium (30 p.p.m.) in W-I and neodymium (IOO p.p.m.) in G-I had to be used together with the ones recommended by AHRBNS’AND FLEISCHEIX~~ for yttrium in G-I and neodymium in W-I, to give -45O Y and Nd working curves. Since the rccommendcd, as well as other values did not give satisfactory z-point working curves for cerium and lanthanum, one-point working curves were prepared from the standard rocks G-I and W-I for determining these elements in granitic ancl basic rocks respectively. A one-point working curve was prepared from G-I for determining scandium in granitic rocks. The SC lines were too black for measurement in the rare earth concentrate prepared from W-I. Since scandium may be reaclily determinccl in basic rocks by direct d.c. arc excitation of the rock powder14 no attempt was made to reduce the sensitivity of the spectrographic procedure dcvcloped for analysing rare earth concentrates.
Dissolution
procedure
A sample (I g) of rock powder (- 120+5) was moistened with water in a platinum basin and treated with 15 ml of hydrofluoric acid and I ml of perchloric acid. After slow evaporation, IO ml of hyclrofluoric acid and I ml of perchloric acid were added to the dish. Evaporation was continued until the evolution of copious perchloric fumes had
DETERMINATION
OF SC, Y,
Nd, Ce
AND
La
IN ROCKS
357
ceased. Ten ml of 2 N hydrochloric acid was added to the dish and warmed for 2 min, during which period the contents of the dish were stirred. The hydrochloric acid was decanted into a so-ml polythene bottle. The residue was treated with IO ml of hydrofluoric acid and I ml of perchloric acid and slowly taken to dryness. When cool, 6 ml of 2 N hydrochloric acid was added. The contents were warmed and stirred (2 min) and decanted into the polythene bottle containing the first decantations. Any undecomposed material remaining was then fumed with I ml of perchloric acid, The resultant residue usually dissolved completely in 2-3 ml of z N hydrochloric acid. On completion of the dissolution procedure, the platinum dish was washed out with z ml of 2 N hydrochloric acid. The washings were added to the contents of the polytherm bottle.
~e~c~~~#~ent of n cotmmt pocedweJov
cottcentvatin,n rnrc em&s &z cmwton
sdicute rocks
Previous investigations 16 to establish the sequence in which major and trace constituents of common silicate rocks (granite and diabnsc for example) moved through cation eschange columns on elution with various concentrations of hydrochloric acid (2,3 and G N) showed that rare earths, together with barium and strontium, appeared later than any of the major constit~ients in the elution sequences (Table II). ‘I’:\
sr:gursNcls
f31,IS
(M;rjnr mtwtitircrlts _---._..-_._ --I IICf Clf~KCLI. -_-_.--.e.--^.-2 *v ‘I‘i t.i \: si ZY I’b %n Sn
II
c?P ~LI,U1’IOS WiTH Ii’t’t>HcxHf.OltIC
hCll>
iw~ uncicrlirtcci)
....l.‘.___-~-.---.~~~~.._.~____l .~‘vpw”c‘* _L1._..^. .-.__.._._. --.--e__e-. ___-l_____-.-13c
N:r -._Fc --
ME GZ -
Xi <:o
Ii Kb
Cs
Ga C’r
t’:r -
--..-“---BL -__.-- .__,._.._“_ Sr
\’
i3n
l,a
It is evident from Table XI that II suitable enrichment of rare earths may be achieved as follows: the abundant elements arc eluted with a concentration of hydrochloric acid which does not readily desorb the rare earths; the rare earths are then eluted with a stronger concentration of hydrochloric acid and by evaporation of the effluent, concentrated to an amount of material sufficieatl}? small to arc sp~ctro~raphically~
3513
R. A.
EDGE,
I,. H. AHRENS
Choice of chant concentrution Investigations were carrictl out using r-g Samples of Cape Granite (EucE el al.1) and W-I, CLcolumn size of 38 x 1.7 cm ancl slow elution ( -10 ml/h) with 3 N hyclrochloric acid. To facilitate the tletection of rare earths in the effluent fractions collected, acid x mg each of YaO.,3 N&Os an<1 LagOn were dissolvecl in the 2 N hydrocllloric solution of tlecomposccl silicate material Ixforc sorption on the cation exchange column. Since rare earths are clutccl in or&l of increasing ionic radii (Yb+:’ (0.56 A) to La+;% (x.14 A) rcspcctivcly) by all concentrations of hyclrochloric acid from Dowex 50 columnsl”, Y+:l (0.92 A), togcthcr with Nd+: (1.04 A) and Lx+: (x.14 A) servecl to locate the position of the r;trc ealth group as well as SC+:’ (o.81 A) in the cfflucnt. During the cllltion, zo-3o-ml fractions were t:tlccn by means of a fraction collector, cvaporatcd to clx yncss ancl nrccrl to completion at 3 A iLS &scriber1 by JZl)c;lSet al.1 and E:DGEIG. Scmiqunntitutive clution curves wcrc constructccl by plotting relative intensities of rare earth lines US. effluent volume. A scmicluantitativc measure of the concentration of a fat-c earth in the effluent fractions was obtainecl by visually estimating the relative intensity of a suitable spcctrttm lint of the clement using a 7-stepped spectrum line as a source of refcrencc. A typical scmi~~~~nntitativct clution curve is shown in Fig. I.
It must bc cniphasizctl that the clution curve of Fig. 1 shows only the relative clution positions of the various clcmcnts. ‘I’hc absolute positions may vary bctwecn apparently similar bntchcs of resin and it is csscntinl to construct a scparatc clution curve for each batch. Although the above conclitions allow a suitable margin between tlic elution of calcium and the appearance of yttrium in tkc effluent, the volume of eluant required for the removal of the rare earths ( r~xooo ml) must lx rctlucccl for cffcctivc analytical use. This can be done either by reducing the column length or by employing a more suitable cluant concentration. Since a 38 x 1.7 column had nlrcncly been adopted for the cacsium enrichment procedure (SW AHRDNS ct al.“) and as it was desired to use tllc same sample for both cacsium ant1 tlic rat-c earths, the seconcl possibility ivas prefer reel. -4 ltd.
C/rim.
A cfcr,
26
(1962)
35_=j-362
”
DETERhlINATION
OF SC, Y,
Nd,
Cc! AND La
IS ROCKS
359
DIAMOND et nl.lo showed that the masimum elution rate of rare earths from a Dowes 50 column was obtained with 6 N hydrochloric acid. CABELL~~ found the optimum hydrochloric acid concentration for eluting scandium from cation exchange coIumns to be 67 N. Investigations were nest carried out with 1-g Cape Granite and 1-6 W-I samples and elution with 3T6 N hydrochloric acid. After the sorption step the column was ,,,elutcd with 330 ml of 3 N hydrochloric acid at 15 ml/h. Only the last Go ml of this cf-
Ttlr. 3-f~ IV l-tcl
JSLUTIOS
01’
KAKK
15AK’TIIS
FROM
A
MICSH
3s X
1.7
Cl11
-.-_-
-__
-.-.-
_.__.. --
--__.-.--
---
Chanaccl from 3 h’ I-1Cl ciiition to G A\’ I-ICI clution at (ml) : l’lO\v rXtc USd thr~~U~h~Jllt 6 s bl(‘l clution (ml/Ii) : \‘<>I. of cfflucnt fractions collcctcd for monitoring purposes (ml) : (.:iL cltrtion cnnlplctc (ml) : Y hrcnks through (ml) :
S pcolt niiixiilillili (nil): Nd peak masimum (ml) Ia
141
peak maximum clution COllll)lCtC
-_--
lllcilSllWt1
8.x.
ZOO-‘+00
Il’.r ._._ __
330 30
20
3::”
350”
.) IO”
*#SO” 5X0”
:
7 10” 310”
:
.1”H” 4 4 5 ”
575” 720” HOOn
_-__._---
.-..-.
frtrni tlio start
50
20
-..-
Is \‘c~lilnlc
Srcrlrtlr .-_------
330
(ml) : (nil)
Ol’l~O\VIL.S
cupc
.sa,,qh ---
COLUXIS
KiSSIS
of tlic 3 .V I I(:1 clution.
fluent was monitorccl. The column was thcu clutecl with 5oo ml of 6 N l~yclrochloric acid at a flow rate of 20 ml/h. 2o-ml fractions were collcctccl ant1 monitored in the usual way. Semiclunntitative clution curves were constructed and Table III was prelxwccl from the results obtainccl. A typical semiquantitative clution curve is shown in Fig. 2. Since this scheme appreciably rc~ducctl tlic volume of eluant necessary to clutc the rare catths, similar column conditions to those outlined in ‘Table III coulcl be usctl for concentrating of t-arc cnrths from common silicate rocks.
Ba
300 Fig.
2. Location
of rare
400 cnrths
600
600 700 Volume of effluent (ml 1
in tlw cfflucnt
from
the
BOO
3-G 1V H.Cl clution
900 of I g Cnpc Crmnitc.
A~rcl. Ckinr. Acta,
2G
(~c,Ga)
355-3Ga
I<. A.
360
EDGE.
L. H. AHRENS
Recovery studies Recovery stuclics for the elution of pg quantities of rare earths from cation exchange and GORDON columns were not carriecl out. SCHUBEI~T et al. 18, HETTEL AND FASSELO et ~1.1” obtained quantitative rccoverics of tract amounts of rare earths by elution with 67 N hydrochloric acid clution from cation cschange columns. Sfwctrogva~hic fwocedwc for determining SC, Y, Nd, Ce and La A spcctrochcmicnl proccclurc utilising zirconium as an internal stanclatd was dtivelopccl for detclmining rare earth elements in the 6 N hydrochloric acid effluent rcsicluc. Zirconium has approximately the same volatility as the rare earths, it posscsscs groups of lines which arc conveniently placed with respect to the most sensitive t-arc cart11 lines and it is ribsent in the G N hydrochloric acid cfflucnt residue. Zirconium was added as a SOA, %rOz-graphite mix. Since the amount of residue obtainccl from the evaporation of the effluent was very small (approximately S-10 mg), tlic %rOz-graphite mis aided in collecting the residue and served as a matrix for subsequent spectrographic analysis. Although rare earths wcrc present in the effluent residue as volatile chlorides, whereas zirconium was present as the involatile oxide, volatilisation tests at 7A sliowccl that the presence of carbon powder rcclucccl sclcctivc clistillation and ensured a nearly complctc distillation of the zirconium. Full cletails of the spectrographic procedure arc given at the start of tlic Ikperimcntal Section. The analytical precision for rare earth elements csprcsscd as a rclativc deviation (C) is given in Tal>le IV.
._. __.-.--_. I:‘lomwl
._..__ - ._...__ . I\dul
I’ve druid
SC
Nd La Cc
Proccdwc
=
for dctcmtiniq
(~ti~r~clarrl
l
.
(Cj
J4
Y
lb L
._.--._.-.-. ion
-_.tlcvintiotl/nvcr;ljic
‘3 ‘4 ‘3 0
- _.-fountl) - loo.
SC, Y, Cc, Nd and La in
wtuvton
silicate
rocks
The 2 N hyclrochloric acid solution from the HF/HCLO,I dissolution of I g of silicate material is soalcccl into the top of a cation cschangc column at a flow rate not exceeding 0.25 ml/min. Rcforc the sorption step the column is wnshcd with 2-3 column volumes of 3 N hydrochloric acicl. Elution is commcncecl with 330 ml of 3 N hyclrochloric acid at a flow rate of 15 ml/h. On completion of this clution, the column is clutecl with 700 ml of 6 N hydrochloric acid at a flow rate of 20 ml/h. ‘l’hc o-60 ml fraction is collected separately and discardecl. The main 6 N HCl fraction is evaporated to do-50 ml and the residual solution is transferrecl to an go-ml porcelain evaporating basin containing 5 mg of 5% ZrO2graphite mis and evaporated carefully to dryness. The resultant residue is loaded into
DETERnIINATION
OF SC, Y, Nd,
Ce AND
La
IN ROCKS
361
x 3 mm depth electrode and arced under the conclitions a 2.4-mm internal diam. described above. Working curves may be prepared from the standard rocks G-I and W-I, which were carried through the same column and spectrochemical procedure as the rock samples. Standards and rock samples were analysed in duplicate. Blank tests on the reagents were carried through all the steps of the concentration procedure. DISCUSSION
The above combinecl cation-eschange enrichment and spcctrochcmical analysis procedure has been employed to estimate SC, Y, Nd, Ce and La in 13 South African granite rocks and Y, Nd, Cc and La in 6 South African basic rocks by E~ca AND AHREKS~ tion
ranges
who
discuss
which
were
aspects
of the geochemistry
observed
are
Granile roch SC
Y
Cc Sri I.il Ihsic
l?oc/ts Y Cc Nd 1-n
indicated
of the rare
in Table
earths.
‘The conccntra-
V.
I .5-l I s*o-55 *3s-17CJa 17-141’ 1()--I 77’
‘3-33 3547 24-44 17-30
0 The concentration ranges arc’ large ant1 a one-point working curve woultl not normally bc usccl over such a lnrgc conccntrntiun range. The reasons for our doing so hcrc arc given on p. 356. ‘L’hc difficulty would of course 1x2rcmovctl if satisfactory standards were avnilablc.
The less abundant rare earths are not detectable by mcnns of the procedure outlined above but it should be possible to determine most of the rare earth group by taking into solution larger quantities of rock. In the above concentration procedure the rare earths were accompanied by small quantities of calcium, magnesium and aluminium. If larger amounts of material were processed, further concentration of the 6 N hydrochloric acid effluent would have to be carried out by reaclsorption dn a smaller cation eschange column followed by removal of the impurities by clution with dilute hydrochloric acid. The more strongly adsorbed rare earths are then cluted with 6 N acid.
The combined LISCof cation cxchangc cnrichmcnt and spcctrochcmical :m:rlysis for the dctcrmination of rare earths in common silicate rocks is dcscrilxd. Iiirrc earth clcmcntN arc: more strongly adsorbed by cation cxchangc resins than the abundant clcmcnts, hence the latter can bc elutcd with R concentration of acid which dots not clcsorb the rare earths. The rare earths are then clutcd And.
Chim.
Ada,
2d
(1962) 355-362
12. A.
362
EDGE,
L. H. AHRENS
by stronycr hydrochloric acid and ttw cfftucnt is cvaporatcd to an amount of material sufficicntty This J~rw_x2clurC attowxxt tllc dctcrmination of Sc, Y, NcJ, CC and La small to arc SJ~cCt~cj~r~~J~tiicatlp. in 13 Soutt1 r\fricnn-@anitC rpx.ks and Y, ~Vtl, CC anct I.0 in G Soutti African basic rocks.
- ..
RJ%[JMt.:: 1,~s al;tcurs tJrcJtMJSC!n~ unc nidttwtlc cl’onalysc rarcs (SC, Y, Kd, (:c ct: Id) dans Its mincrais d’ions ct Sliition ir J’wiclc chlort~yclricluc.
sJ~cctrochixl,icl~tc XJh!S siticat&,
(a arc) S&_llW’aticJn
JXJUr
LLU
Ic
dosagc
lncJyL!n
dcs
tcrrcv
d’&tlangeUr
,
%US/\iL1MISNI~,\SSI;N~; ~hXtlrC!ttJlln~ ctnL!r SJl~JC~r~J~tl~~lJS~tl~n t)iC CC uiitl Lir) in SiJiJ~;rt~nit1crnJicn. ~~Jll~ll~lVS~~lVS~t1~~S.
htd.tlUCtC! f\btrcniiUllg
%LIr t%%tillllnlln~ VfJn ~1llCtWWl
dcr %_!tb3X!11 t
(SC,
Y,
Ntt,
J-IilfC CinCs
THE BY
DETERMINATION A COMBINED
OF CATION
CHIMICA
ACTA
SC, Y, Nd, Ce AND
355
La IN
SILICATE
EXCHANGE-SPECTROCHEMICAL
R. A. EDGE Dcpurfmerrl O/Gcochemislry.
AND
METHOD
L. H. AHRENS
University
(Rcccivcd
ROCKS
August
of Cape
Town
(SoutJr
Africa)
5th, rgG~)
The application of combined ion cschange enrichment and c1.c. arc spectrochemical methods for determining trace clemcnts which are not detectable by direct excitation in rocks, soils and meteorites has been discussed by EDGE et aZ.1, BROOKS et aZ.2 and AHRENS et aZ.3*4. Anion exchange enrichment has been used for estimating Cd, Zn, Sn, Bi and Tl (observations have also been maclc on Nb, MO, In, A6 and the platinum metals)znfi.fl. Cation eschange techniques 3.7eRhave been applied to tin, molybdenum, caesium and some rare earths and to a few elements detectable by direct d.c. arc excitation cg. scandium, gallium and rubidium. EDGE et al.1 have briefly outlined some of these investigations and AHKENS et al. 3 have described in detail the estimation of caesium in chondritic meteorites and basic rocks. In the present paper investigations are described on the combined use of cation exchange enrichment and spectrochemical analysis for determining SC, Y, Nd, Ce and La in common silicate rocks (granites and diabascs). Although combined cation exchange-spectrochemical analysis procedures have been employed for estimating traces of rare earths in zirconium metal” and uranium compounds’0 and trace amounts of yttrium in bioto@cal material+‘, no workers appear to have employed this technique for estimating rare earths in silicate rocks. EXPERIMENTAL
Afifiaratrrs
and reagents
Standard spectrographic equipment was used and is described below together with the excitation conditions: Spectrogaph: Hilger E49z large quartz and glass. External optics: Hilger Eg58 lens focusscd on slit; step sector (2 : I ratio). Slit length: II cm. Slit width: 0.001 mm. Wavelength range: 3800-5300 A. Glass optics. Photographic plate : Kodak 103-0. Electrodes: Upper electrode (- ve), Champion Ship carbon 0.5 cm diam. rod. Lower clcctrode (+ve), National carbon regular grade graphite, 2.4 mm int. diam. x 3 mm depth. Current: 7 A d.c. Esposure: to completion, usually about 60 sec. Sample: 6 lV HCl effluent resiclue and 5 rng 5% ZrOz-C mix. Photographic processing: 46 min at 20 in I
lint in those basic rock rare earth conccntratcv whcrc .-1ual. CJlirrr. rich,
26:( 1962)
355-362
R. A. EDGE, Ia. H. AHRENS
35’5
Ion exchange columns of length 38 cm and internal diameter 1.7 cm were used. The resin was the strongly acid Dowex 50 8X. 200-400 mesh, H-form. The following reagents and materials were also used: A.R. HCl (s.g. 1.18); A.R. HE (40%), A.R. HC104 (60%) ; deionised water; “specpure” YnOa, NdzOn, La203 and ZrOz and National Carbon S.P-2 graphite powder. Spectrochemical tests on the purity of the liquid reagents and the cation exchanger did not show detectable SC 4246.8, Y 4398.0, La 4333.7, Nd 4303.6 and Cc 4296.7. Volumes of reagents larger than those used in the actual analysis were taken to dry- _’ ness and examined spectrochemically. The resins were examined by arcing the ash from 5 g of resin.
Standards Natural silicate standards (granite G-I and diabase W-I (AHRDNS AND FLIXSCHER~~)) were used exclusively and were carried through the same dissolution, column and spectrographic procedures as the rock samples. The values of SC, Y, Ce, Nd and La used to establish the working curves are given in Table I.
AMOUNTS
FOR
I’HI~I’ARATION
(p&n.,
Slurtdnrd
G-I grunitc
W-I diabasc
-
--.--
OIr WORKING CC
CURVKS Nd
La
IP.P*rn.)
(p.p.*ll.)
21
600
IO0
150
30
50
50
30
-_-
(P.P.1n.l
-
Wherever possible the rccommencled values of AHIUZNS AND FLEISCHER~~ were employed. However, BERMAN’S~~ values for yttrium (30 p.p.m.) in W-I and neodymium (IOO p.p.m.) in G-I had to be used together with the ones recommended by AHRBNS’AND FLEISCHEIX~~ for yttrium in G-I and neodymium in W-I, to give -45O Y and Nd working curves. Since the rccommendcd, as well as other values did not give satisfactory z-point working curves for cerium and lanthanum, one-point working curves were prepared from the standard rocks G-I and W-I for determining these elements in granitic ancl basic rocks respectively. A one-point working curve was prepared from G-I for determining scandium in granitic rocks. The SC lines were too black for measurement in the rare earth concentrate prepared from W-I. Since scandium may be reaclily determinccl in basic rocks by direct d.c. arc excitation of the rock powder14 no attempt was made to reduce the sensitivity of the spectrographic procedure dcvcloped for analysing rare earth concentrates.
Dissolution
procedure
A sample (I g) of rock powder (- 120+5) was moistened with water in a platinum basin and treated with 15 ml of hydrofluoric acid and I ml of perchloric acid. After slow evaporation, IO ml of hyclrofluoric acid and I ml of perchloric acid were added to the dish. Evaporation was continued until the evolution of copious perchloric fumes had
DETERMINATION
OF SC, Y,
Nd, Ce
AND
La
IN ROCKS
357
ceased. Ten ml of 2 N hydrochloric acid was added to the dish and warmed for 2 min, during which period the contents of the dish were stirred. The hydrochloric acid was decanted into a so-ml polythene bottle. The residue was treated with IO ml of hydrofluoric acid and I ml of perchloric acid and slowly taken to dryness. When cool, 6 ml of 2 N hydrochloric acid was added. The contents were warmed and stirred (2 min) and decanted into the polythene bottle containing the first decantations. Any undecomposed material remaining was then fumed with I ml of perchloric acid, The resultant residue usually dissolved completely in 2-3 ml of z N hydrochloric acid. On completion of the dissolution procedure, the platinum dish was washed out with z ml of 2 N hydrochloric acid. The washings were added to the contents of the polytherm bottle.
~e~c~~~#~ent of n cotmmt pocedweJov
cottcentvatin,n rnrc em&s &z cmwton
sdicute rocks
Previous investigations 16 to establish the sequence in which major and trace constituents of common silicate rocks (granite and diabnsc for example) moved through cation eschange columns on elution with various concentrations of hydrochloric acid (2,3 and G N) showed that rare earths, together with barium and strontium, appeared later than any of the major constit~ients in the elution sequences (Table II). ‘I’:\
sr:gursNcls
f31,IS
(M;rjnr mtwtitircrlts _---._..-_._ --I IICf Clf~KCLI. -_-_.--.e.--^.-2 *v ‘I‘i t.i \: si ZY I’b %n Sn
II
c?P ~LI,U1’IOS WiTH Ii’t’t>HcxHf.OltIC
hCll>
iw~ uncicrlirtcci)
....l.‘.___-~-.---.~~~~.._.~____l .~‘vpw”c‘* _L1._..^. .-.__.._._. --.--e__e-. ___-l_____-.-13c
N:r -._Fc --
ME GZ -
Xi <:o
Ii Kb
Cs
Ga C’r
t’:r -
--..-“---BL -__.-- .__,._.._“_ Sr
\’
i3n
l,a
It is evident from Table XI that II suitable enrichment of rare earths may be achieved as follows: the abundant elements arc eluted with a concentration of hydrochloric acid which does not readily desorb the rare earths; the rare earths are then eluted with a stronger concentration of hydrochloric acid and by evaporation of the effluent, concentrated to an amount of material sufficieatl}? small to arc sp~ctro~raphically~
3513
R. A.
EDGE,
I,. H. AHRENS
Choice of chant concentrution Investigations were carrictl out using r-g Samples of Cape Granite (EucE el al.1) and W-I, CLcolumn size of 38 x 1.7 cm ancl slow elution ( -10 ml/h) with 3 N hyclrochloric acid. To facilitate the tletection of rare earths in the effluent fractions collected, acid x mg each of YaO.,3 N&Os an<1 LagOn were dissolvecl in the 2 N hydrocllloric solution of tlecomposccl silicate material Ixforc sorption on the cation exchange column. Since rare earths are clutccl in or&l of increasing ionic radii (Yb+:’ (0.56 A) to La+;% (x.14 A) rcspcctivcly) by all concentrations of hyclrochloric acid from Dowex 50 columnsl”, Y+:l (0.92 A), togcthcr with Nd+: (1.04 A) and Lx+: (x.14 A) servecl to locate the position of the r;trc ealth group as well as SC+:’ (o.81 A) in the cfflucnt. During the cllltion, zo-3o-ml fractions were t:tlccn by means of a fraction collector, cvaporatcd to clx yncss ancl nrccrl to completion at 3 A iLS &scriber1 by JZl)c;lSet al.1 and E:DGEIG. Scmiqunntitutive clution curves wcrc constructccl by plotting relative intensities of rare earth lines US. effluent volume. A scmicluantitativc measure of the concentration of a fat-c earth in the effluent fractions was obtainecl by visually estimating the relative intensity of a suitable spcctrttm lint of the clement using a 7-stepped spectrum line as a source of refcrencc. A typical scmi~~~~nntitativct clution curve is shown in Fig. I.
It must bc cniphasizctl that the clution curve of Fig. 1 shows only the relative clution positions of the various clcmcnts. ‘I’hc absolute positions may vary bctwecn apparently similar bntchcs of resin and it is csscntinl to construct a scparatc clution curve for each batch. Although the above conclitions allow a suitable margin between tlic elution of calcium and the appearance of yttrium in tkc effluent, the volume of eluant required for the removal of the rare earths ( r~xooo ml) must lx rctlucccl for cffcctivc analytical use. This can be done either by reducing the column length or by employing a more suitable cluant concentration. Since a 38 x 1.7 column had nlrcncly been adopted for the cacsium enrichment procedure (SW AHRDNS ct al.“) and as it was desired to use tllc same sample for both cacsium ant1 tlic rat-c earths, the seconcl possibility ivas prefer reel. -4 ltd.
C/rim.
A cfcr,
26
(1962)
35_=j-362
”
DETERhlINATION
OF SC, Y,
Nd,
Cc! AND La
IS ROCKS
359
DIAMOND et nl.lo showed that the masimum elution rate of rare earths from a Dowes 50 column was obtained with 6 N hydrochloric acid. CABELL~~ found the optimum hydrochloric acid concentration for eluting scandium from cation exchange coIumns to be 67 N. Investigations were nest carried out with 1-g Cape Granite and 1-6 W-I samples and elution with 3T6 N hydrochloric acid. After the sorption step the column was ,,,elutcd with 330 ml of 3 N hydrochloric acid at 15 ml/h. Only the last Go ml of this cf-
Ttlr. 3-f~ IV l-tcl
JSLUTIOS
01’
KAKK
15AK’TIIS
FROM
A
MICSH
3s X
1.7
Cl11
-.-_-
-__
-.-.-
_.__.. --
--__.-.--
---
Chanaccl from 3 h’ I-1Cl ciiition to G A\’ I-ICI clution at (ml) : l’lO\v rXtc USd thr~~U~h~Jllt 6 s bl(‘l clution (ml/Ii) : \‘<>I. of cfflucnt fractions collcctcd for monitoring purposes (ml) : (.:iL cltrtion cnnlplctc (ml) : Y hrcnks through (ml) :
S pcolt niiixiilillili (nil): Nd peak masimum (ml) Ia
141
peak maximum clution COllll)lCtC
-_--
lllcilSllWt1
8.x.
ZOO-‘+00
Il’.r ._._ __
330 30
20
3::”
350”
.) IO”
*#SO” 5X0”
:
7 10” 310”
:
.1”H” 4 4 5 ”
575” 720” HOOn
_-__._---
.-..-.
frtrni tlio start
50
20
-..-
Is \‘c~lilnlc
Srcrlrtlr .-_------
330
(ml) : (nil)
Ol’l~O\VIL.S
cupc
.sa,,qh ---
COLUXIS
KiSSIS
of tlic 3 .V I I(:1 clution.
fluent was monitorccl. The column was thcu clutecl with 5oo ml of 6 N l~yclrochloric acid at a flow rate of 20 ml/h. 2o-ml fractions were collcctccl ant1 monitored in the usual way. Semiclunntitative clution curves were constructed and Table III was prelxwccl from the results obtainccl. A typical semiquantitative clution curve is shown in Fig. 2. Since this scheme appreciably rc~ducctl tlic volume of eluant necessary to clutc the rare catths, similar column conditions to those outlined in ‘Table III coulcl be usctl for concentrating of t-arc cnrths from common silicate rocks.
Ba
300 Fig.
2. Location
of rare
400 cnrths
600
600 700 Volume of effluent (ml 1
in tlw cfflucnt
from
the
BOO
3-G 1V H.Cl clution
900 of I g Cnpc Crmnitc.
A~rcl. Ckinr. Acta,
2G
(~c,Ga)
355-3Ga
I<. A.
360
EDGE.
L. H. AHRENS
Recovery studies Recovery stuclics for the elution of pg quantities of rare earths from cation exchange and GORDON columns were not carriecl out. SCHUBEI~T et al. 18, HETTEL AND FASSELO et ~1.1” obtained quantitative rccoverics of tract amounts of rare earths by elution with 67 N hydrochloric acid clution from cation cschange columns. Sfwctrogva~hic fwocedwc for determining SC, Y, Nd, Ce and La A spcctrochcmicnl proccclurc utilising zirconium as an internal stanclatd was dtivelopccl for detclmining rare earth elements in the 6 N hydrochloric acid effluent rcsicluc. Zirconium has approximately the same volatility as the rare earths, it posscsscs groups of lines which arc conveniently placed with respect to the most sensitive t-arc cart11 lines and it is ribsent in the G N hydrochloric acid cfflucnt residue. Zirconium was added as a SOA, %rOz-graphite mix. Since the amount of residue obtainccl from the evaporation of the effluent was very small (approximately S-10 mg), tlic %rOz-graphite mis aided in collecting the residue and served as a matrix for subsequent spectrographic analysis. Although rare earths wcrc present in the effluent residue as volatile chlorides, whereas zirconium was present as the involatile oxide, volatilisation tests at 7A sliowccl that the presence of carbon powder rcclucccl sclcctivc clistillation and ensured a nearly complctc distillation of the zirconium. Full cletails of the spectrographic procedure arc given at the start of tlic Ikperimcntal Section. The analytical precision for rare earth elements csprcsscd as a rclativc deviation (C) is given in Tal>le IV.
._. __.-.--_. I:‘lomwl
._..__ - ._...__ . I\dul
I’ve druid
SC
Nd La Cc
Proccdwc
=
for dctcmtiniq
(~ti~r~clarrl
l
.
(Cj
J4
Y
lb L
._.--._.-.-. ion
-_.tlcvintiotl/nvcr;ljic
‘3 ‘4 ‘3 0
- _.-fountl) - loo.
SC, Y, Cc, Nd and La in
wtuvton
silicate
rocks
The 2 N hyclrochloric acid solution from the HF/HCLO,I dissolution of I g of silicate material is soalcccl into the top of a cation cschangc column at a flow rate not exceeding 0.25 ml/min. Rcforc the sorption step the column is wnshcd with 2-3 column volumes of 3 N hydrochloric acicl. Elution is commcncecl with 330 ml of 3 N hyclrochloric acid at a flow rate of 15 ml/h. On completion of this clution, the column is clutecl with 700 ml of 6 N hydrochloric acid at a flow rate of 20 ml/h. ‘l’hc o-60 ml fraction is collected separately and discardecl. The main 6 N HCl fraction is evaporated to do-50 ml and the residual solution is transferrecl to an go-ml porcelain evaporating basin containing 5 mg of 5% ZrO2graphite mis and evaporated carefully to dryness. The resultant residue is loaded into
DETERnIINATION
OF SC, Y, Nd,
Ce AND
La
IN ROCKS
361
x 3 mm depth electrode and arced under the conclitions a 2.4-mm internal diam. described above. Working curves may be prepared from the standard rocks G-I and W-I, which were carried through the same column and spectrochemical procedure as the rock samples. Standards and rock samples were analysed in duplicate. Blank tests on the reagents were carried through all the steps of the concentration procedure. DISCUSSION
The above combinecl cation-eschange enrichment and spcctrochcmical analysis procedure has been employed to estimate SC, Y, Nd, Ce and La in 13 South African granite rocks and Y, Nd, Cc and La in 6 South African basic rocks by E~ca AND AHREKS~ tion
ranges
who
discuss
which
were
aspects
of the geochemistry
observed
are
Granile roch SC
Y
Cc Sri I.il Ihsic
l?oc/ts Y Cc Nd 1-n
indicated
of the rare
in Table
earths.
‘The conccntra-
V.
I .5-l I s*o-55 *3s-17CJa 17-141’ 1()--I 77’
‘3-33 3547 24-44 17-30
0 The concentration ranges arc’ large ant1 a one-point working curve woultl not normally bc usccl over such a lnrgc conccntrntiun range. The reasons for our doing so hcrc arc given on p. 356. ‘L’hc difficulty would of course 1x2rcmovctl if satisfactory standards were avnilablc.
The less abundant rare earths are not detectable by mcnns of the procedure outlined above but it should be possible to determine most of the rare earth group by taking into solution larger quantities of rock. In the above concentration procedure the rare earths were accompanied by small quantities of calcium, magnesium and aluminium. If larger amounts of material were processed, further concentration of the 6 N hydrochloric acid effluent would have to be carried out by reaclsorption dn a smaller cation eschange column followed by removal of the impurities by clution with dilute hydrochloric acid. The more strongly adsorbed rare earths are then cluted with 6 N acid.
The combined LISCof cation cxchangc cnrichmcnt and spcctrochcmical :m:rlysis for the dctcrmination of rare earths in common silicate rocks is dcscrilxd. Iiirrc earth clcmcntN arc: more strongly adsorbed by cation cxchangc resins than the abundant clcmcnts, hence the latter can bc elutcd with R concentration of acid which dots not clcsorb the rare earths. The rare earths are then clutcd And.
Chim.
Ada,
2d
(1962) 355-362
12. A.
362
EDGE,
L. H. AHRENS
by stronycr hydrochloric acid and ttw cfftucnt is cvaporatcd to an amount of material sufficicntty This J~rw_x2clurC attowxxt tllc dctcrmination of Sc, Y, NcJ, CC and La small to arc SJ~cCt~cj~r~~J~tiicatlp. in 13 Soutt1 r\fricnn-@anitC rpx.ks and Y, ~Vtl, CC anct I.0 in G Soutti African basic rocks.
- ..
RJ%[JMt.:: 1,~s al;tcurs tJrcJtMJSC!n~ unc nidttwtlc cl’onalysc rarcs (SC, Y, Kd, (:c ct: Id) dans Its mincrais d’ions ct Sliition ir J’wiclc chlort~yclricluc.
sJ~cctrochixl,icl~tc XJh!S siticat&,
(a arc) S&_llW’aticJn
JXJUr
LLU
Ic
dosagc
lncJyL!n
dcs
tcrrcv
d’&tlangeUr
,
%US/\iL1MISNI~,\SSI;N~; ~hXtlrC!ttJlln~ ctnL!r SJl~JC~r~J~tl~~lJS~tl~n t)iC CC uiitl Lir) in SiJiJ~;rt~nit1crnJicn. ~~Jll~ll~lVS~~lVS~t1~~S.
htd.tlUCtC! f\btrcniiUllg
%LIr t%%tillllnlln~ VfJn ~1llCtWWl
dcr %_!tb3X!11 t
(SC,
Y,
Ntt,
J-IilfC CinCs