Determination of ten rare-earth elements and yttrium in silicate rocks by ICP—AES without separation and preconcentration

Tahra, Vol 38, No 3, pp

0039-9140/91$300+000

329-332, 1991

Pergamon Press

Pnnted m Great Bntam

plc

DETERMINATION OF TEN RARE-EARTH ELEMENTS AND YTTRIUM IN SILICATE ROCKS BY ICP-AES WITHOUT SEPARATION AND PRECONCE~~TION M. S. RATHI, P. P.

KHANNA*

and PULOK KUMAR

MUKHEI~JBE

Spectra Laboratory, Wa&a Institute of HmAayan Geology, 33 General Mahadeo Smgh Road, Dehra Dun 248 001. In&a (Recewed 20 Ucto&er 1989 Rewed 25 J&y 1990. Accepted 19 September 1990)

Summary-Lanthanum, cenum, neodymium, samarium, europium, gadohmum, dysprosnun, erbmm, ytterbium, lutetmm and yttnum have been determined m 8 international rock standards by inductively coupled plasma atormt: emzssion spectrometry (ICP-AES) wrthout pnor ion-exchange separation and preconcentration The results for La, Ce, Nd, Eu, Dy, Yb and Y were m good agreement with the reported values, whereas those for Sm, Gd, Er and Lu were less accurate. However, the results for Sm, Gd, Er and Lu can also be used for stu&es of petrogenesls

Rare-earth element (REE) geochemistry has been established as one of the most powerful tools for petrogenetic studies of rocks. REEs (except for Ce) are comp~atively inert towards secondary geological processes, and hence their relative abundances even m altered rocks, give valuable informatron about the origin of rocks.’ High-resolution ICP-AES has become a wrdely used and popular technique for measurement of major and trace elements, including REEs, 3% in analytical geochemistry. However, preparation of samples for ICP-AES determination of the REEs in rock samples generally mvolves separation of REE fracttons by cationexchange, followed by concentratron by evaporation.%* This procedure IS quite tedious and thus hinders the routme productron of REE data. The atm of the present lnvestigatlon was to find whether the separation and preconcentration steps can be avoided by calibrating the instrument with standards having a chemical matrix similar to that of the samples. EXPERIMENTAL

Dried samples (0.5 g) were digested wrth a hydro~uo~~-perchlo~~ acid mixture in open PTFE beakers. The digestion was repeated 5 or 6 times with evaporation to perchloric acid fumes after each, in order to ensure complete dissolution, and the final solution was made up accurately to 50 ml. The instr~entation used *Author for correspondence

was a Jobin Yvon model JY 70 Plus ICP, with a l-m focal length Czerny-Turner, holographic grating, 3600 grooves/mm, and a Memhard concentric glass nebuliier “type C”. The RF generator was a 40.68 MHz single-phase unit, with power stable to better than 0.01%. The gas flow-rates (l./min) were: outer (coolant) 12; intermediate (au~lia~) 0; central (carrier -t sheath) 0.55. A set of four rock standards (Syenite STM-1, Quartz latite QLO-1, Diabase W-2, supplied by U.S.G.S. and Granite GS-N, supplied by A.N.R.T., France) and a blank were used for calibratton m order to cover a wide compositional range of geologrcal materials. Six rock standards (Basalt BE-N, from A.N.R.T., Basalt BHVO- 1, Mica schist SDC-1, Granite G-2, Rhyolite RGM-1 and Granodiorite GSP-1 from U.S.G.S.) were treated as unknown samples and analysed for REE and Y. A validation test was done with a different set of standards for cahbranon (SDC-1, BE-N, BHVO-1, W-2, and Gabbro MRG-1, supplied by C.C.R.M.P., Canada) and a blank, for measurement of REEs and Y in QLO- 1, GS-N, G-2, RGM-1 and GSP-1. The results are summarized in Table 1. RESULTS AND DISCUSSION

The concentrations of the REEs and Y in the reference rock standards were taken from the latest compilation by GovindarajuP Some of the REE values for these reference rocks are not 329

80 (82) 136 (152) 64 (70) 37(36) 62(64)* 25(1 8) 13 (12) 18 6 (9.5)* 2 l(2 5)’ 0 4 (0.24) 27 (30)

398 852 413 765 430.358 381967 353 171 328 937 359 260 335 a48 331271 261 542 371030

1

21 7 (15 8) 41(39) 32 2 (25 2) 237(206) 5 5 (5 2) 2 68 (2 02) 96(62) 178(64) 24(24) 0.39 (0 29) 27 7 (27.6)

(A)

BHVO42 (42) 92 (93) 45 (40) 2 l(1 71) 66(67) 3 3 (4) 12(82) 8(72) 3 1 (4 l)* 0 29 (0 S3)* 30 (40)

(A)

SDC-1 30 5 (27) 62 (54 6) 21(26) 1 62 (143) 4 8 (3 8) 2 3 (2 32) 7 (4 88) 3 (4 7)* 06(23) 0 2 (0.37) 32 (24)

QLO- 1 (B) 74 5(75) 148 (135) 50 S(50) 19(1 7) 4 (3)* 1 4 (1 5) 92t77) 3 (5 2) 07(1 7)* 0.21 (0 22) 22 (19)

(8)

GS-N

(A) 84 155 54 I8 32 09 109 46 14 0.12 100

(B) 93 (89) 187(160) 55 (55) 18(1 4) 3 3(24) 0 8(0 8) 102(72) 2 6(4 3)* 051(092)* 0 12 (0 ll)* 13(11)

G-2

:: 24 83 33 08 0 14 22

2s 56 27

- (4

@I 27 (24) 59 (47) 24 (19) lO(O66) 47(48) 2 3 (2 6) 6 8 (4 3) 17(37) 0 3 (2 6) 0 14(041) 31(25)

RGM-1

(A)

2g2

2: f 75 19

162 374 159 31 68

@I

2: : ;;6?) 44(12 1) 0.7 (2 7)* 0 2 (0 214)* 33 (26)

184 (184) 457 (399) 188 (196) 3.1(2.33) 7 l(5 5)

GSP- 1

:i Er LU Y

?i

La Ce Nd EU

Element

54 (2) 1 8 (29) 3 2 (33) 09(13) 0.9(51) 46(7) 44(52) 0 (9) 10 (9)

84 (6) 155 (3)

(A)

G-2

7451) 148 (10) 505(l) 1.9(12) 4 (33) 14(7) 9.2 (20) 3 (42) 0.7 (59)

2!!i)

(B)

KS0

85 5 (4) 156 (3) 53 (4) 1 31(6) 2 22 (8) 0 61(24) 722(03) 4 07 (5) NR 0 08 (27) NR

(0

0::(14) NR

73 (3) 144 (7) 50 0 (0) 1.66 (2) 3 l(3) 14(7) 6 9 (10) 45fll)

OS-N (A) Z!(6) 159(19) 3 l(33) 6 8 (24) 16 (6) %6(9) 7 5 (38) 1.9 (30) 02 (7) 23 (12)

162(12)

(CY 177 (4) 431(8) 197 (0 5) 2 91(25) 6 17 (12) 1 33 (22) 25 8 (2) 11 7(3) 2 56 (5) 0 17 (21) 232(11)

GSP-1 (B)

U(l) 7 (46) 3 (36) 0 6 (74) 02(46) 30 (25)

30 5 (13) 62 (14) 2?: (4) 1.62(13) 4 8 (26)

(C)6 26 4 (2) 52 l(5) 21 2 (19) 1 3 (9) 3 57 (6) 2 29 (1) 4 21(14) 3 58 (24) 2 13 (7) 0 38 (3) 22 8 (5)

QLG-1

sz(O) 92(l) 45(13) 2 1(23) c?_+(2) 3 3(18) 12 (46) 8lll) 3.1 (24) 0 29 (45) 32 (20)

(A)

xx-1 CC)’

408(z) 1 65 (4) 6 3 (6) 3 41(15) 8 l(l) 74(3) 3.46116) 047(11) 32 3 (19)

41 2 (2) 87 6 (6)

&i(9) 37 (3) 62(3) 2 5 (39) 13(8) 18 6(96) 2 1(16) 0 4 (67) 27 (10)

80 (2) 136(11)

(A)

KS 82 87 (1) 152 08 (0) 63 21 (10) 3 71(3) 6.27 1 77 (2) 1185(l) lO.Ol(5) 2 49 (0 4) 0 26 (8) 30.87 (3)

BE-N

Table 2 Compartson of REE and Y concentrations &g/g) m 6 rock standards, measured with and wlthout use of a separation techmque [calibration sets as for Table 1, N R meaus not reported, (C) mduates the results reported m the reference grven for use of the separation method, percentage devratzous from recommended9 values are gtven m parentheses]

& Yb Sm Gd Er LU Y

Ce Nd

Ia

BE-N (A)

Lme, nm

(recommended values9 are grven m parentheses)

1 Results @g/g) for analysts wltb (A) STM-1, QLG-1, W-Z and GS-N as cahbratmn standards and (B) SDC-1, BE-N, BRVO-1, W-2 and MRG-1 as cahbrafion standards

Element

Table

i !z 0,

3:

Determmatlon

of rare-earth elements by ICP-AES

331

Table 3 Slopes and Intercepts of the cahbration graphs obtained with the two sets of standards Set (B)

Set (A) Element

Slope

Intercept

Slope

Intercept

La Ce Nd Eu DY Yb Sm Gd Er Lu Y

0093 0 18 0069 0 0032 0 0078 0 0026 0 017 0 0037 3 6 x 1O-6 0 0032 0.0028

-95 -244 -198 -12 -33 -05 -102 -19 00 -12 00

0 023 0 052 0 039 ‘00027 0 0072 0 0023 0012 0 0076 2 3 x 10-5 0 0027 0.0038

-62 -202 -144 -08 -24 -03 -69 -07

well characterized and are proposed only for information (they are marked with an asterisk in Table 1). A comparison of our results (Table 2) with those obtained by use of a separation techmqueH,‘O*” and also with the consensus values,’ shows that those for La, Ce, Nd, Eu, Dy, Yb, and Y are in good agreement. Agreement for Sm, Gd, Er and Lu is rather poor, but most of the differences are within the acceptable limits of error for petrogenetic purposes. Some of our measurements (underlined in Table 2) are in better agreement with the reported values than are those obtained by using a separation technique. The values found for Er were always lower than the reported values, whereas those for Sm were always high. The accuracy of the results for these two elements was better for basic rocks than for granitic rocks. The reason for this is unclear, but is possibly spectral interference by elements such as Ba, Sr, Zr, etc. which are generally present in gramtic

Table 4 MaJor component composlhon Standard

BO,

Al@,

GS-N

65 80

14 67

65 55

(Gramte) QLO-1 (Quartz latlte) STM-1 (Syemte) gtbase) SDC- 1 (Mica s&St) BE-N (Basalt) BHVO- 1 (Basalt) MRG-1 (Gabbro)

-12

rocks at levels as much 10-100 times those in basic ones. The cation-exchange methods do not give complete separation of the REEs from the other rock components, some elements, such as Ba, Ca, Sr, Ti and Zr, being incorporated m the REE fraction.3*7 In our method, however, rock standards were used instead of salt standards for calibration, and as these are chosen so that their matrices are similar to those of the samples, there is some degree of compensation of chemical and spectral interferences. The results obtained for G-2, RGM-1 and GSP-1 differed, however, according to which set of calibration standards was used. All the calibrations were found to satisfy a first order fit. The slopes and intercepts obtained for both calibration sets are given m Table 3. The major component composition of the standards is given m Table 4. In summary, it may be concluded that separation and preconcentration of REEs is not essential for their determination in sihcate rock

(%) of the standards

used

for cahbration

Fe,G,

Fe0

MnO

MgO

Na,O

K,O

TlO,

P,O,

1.92

1.65

0 056

2 30

2 50

3 77

4 63

0 68

0 28

16 18

102

2 97

0093

100

3 17

420

360

0 624

0.254

59 64

1839

2 87

2 09

0 22

0101

109

8 94

4 28

0 135

0 158

52 44

15 35

1 52

831

0 163

6 37

10 87

2 14

0 627

106

0 131

65 85

1575

2 62

3 93

0 114

169

140

2 05

3 28

101

0 158

38 20

10 07

5 34

6 74

0 20

13 15

13 87

3 18

1 39

261

105

49 94

13 80

2 82

8 58

0 168

7 23

1140

2 26

0 52

271

0 273

39 09

8 46

8.36

8 66

0 17

13 55

14 71

0 74

0 18

3 77

0 08

CaO

332

M S. R~rm et al

samples. It is difficult to determine Pr, Tb, Ho and Tm by this method, but the rest of the REEs and Y can be measured with reasonable accuracy in a IO-mg/ml rock solution by using high-resolution ICP-AES. The accuracy achieved is sufficient for the purpose of geological interpretation. The results could probably be further improved by using calibration standards that all closely match the samples in chemical composition.

REFERENCES

2 3. 4 5 6. I

authors are grateful to Dr. V. C Thakur, Director, W I H.G., for provtdmg all faculties for this study The Department of Science and Technology (Government of India) is thanked for financial support for procurement of the ICP umt, under the Sohd Earth Scheme. Professor A K. Smha IS thanked for hts encouragement throughout the work Acknowledgements-The

8 9 10 11.

P. Henderson (ed.), Rare Earth Element Geochemutry, p. 510 Elsevter, Amsterdam, 1984 S. E. Church, Geostand. Newsl, 1981, 5, 133. J N Walsh, F. Buckley and J Barker, Chem Geof., 1981, 33, 141. K. GovmdraJu and G Mevelle, Spctrochun Acta, 1983, jsB, 1447 I Jarvts and K. E. Jarvts, Chem Geol., 1985, 53, 335 K E. Jarvts and I. Jarvts, Geosfand NewsI., 1988, 12, 1 J G. Crock, F E. Ltchte, G. 0 bddle and C L Beech, Talanta, 1986, 33, 601 I. Roelandts and G Mrchel, Geostand. Newsf , 1986,10, 135. K. GovmdaraJu, rbrd., 1989, 13, 1. I Roelandts, Chem. Geol., 1988, 67, 171 K GovmdraJu and G. Mevelle, J. Anal At. Spectrom 1987, 2, 615.