Determination Of Zirconium And Hafnium

Journal of Radioanalytical Chemistry VoL 5 (1970) 51--60

S I M U L T A N E O U S D E T E R M I N A T I O N OF Z I R C O N I U M A N D H A F N I U M I N S T A N D A R D R O C K S BY N E U T R O N ACTIVATION ANALYSIS T. V.

REBAGAY,W. D. EHMANN

Department of Chemistry, University of Kentucky, Lexington, Kentucky 40506 (USA)

(Received September 15, 1969)

A sensitive analytical technique has been developed to determine zirconium and hafnium at the 1 ppm and 0.01 ppm levels, respectively, in natural silicate matrices. The technique is based on the use of a Ge(Li) detector for high-resolution ),-ray spectrometry following irradiation with thermal neutrons. Simultaneous separation of both elements is achieved by addition of only zirconium carrier and use of a strongly basic anion-exchange resin. Hafnium was shown to follow zirconium throughout the chemical procedures. The chemical yield of the separation procedure was determined for each sample by use of an automated fast-neutron activation analysis system, which obviated the need to convert the separated zirconium to a conventional gravimetric weighing form. The method has been applied to the analysis of a variety of standard rocks and related natural materials.

Introduction A major problem in the use of neutron activation for the determination of trace elements in complex matrices is the interference of one photopeak with another in the 7-ray spectra measured with conventional NaI(T1) detectors. This is most evident in the determination of zirconium in natural materials. In the thermal-neutron activation determination of this element, the radionuclide used as zirconium indicator is either 97Zr (T = 17.0 hrs) or 95Zr (T = 65.5 d). The principal means of production of these radionuclides are the reactions: 9~Zr(n, ~)97Zr and 94Zr(n, T)D5Zr. A number of potential sources of primary and spectral interferences are known and are reviewed in more detail by REBAGAY. 1 Although the yield of 95Zr is several magnitudes lower than that of 97Zr, it becomes the most logical choice for the indicator radionuclide when transportation of the irradiated specimens to a distant laboratory is required. 95Zr which has two 7-ray lines of equal intensity (724 keV and 756 keV; LEDERERet al. 2) decays to 95Nb (T -- 35.0 d), with a 7-ray line (765 keV) lying very close to one of the 95Zr lines. It is obvious that resolution of these lines could not be achieved in a NaI(T1) scintillation spectrum. However, a lithium-drifted germanium detector that has a very much better energy-resolving power ( 3 - 5 keV F W H M ) than the widely-used NaI(T1) detector (50 keV F W H M ) makes it possible to discriminate between the close-lying lines of 95Zr and its daughter, 95Nb. J. RadioanaL Chem. 5 (1970)

52

T . V . R E B A G A Y , W. D. E H M A N N : D E T E R M I N A T I O N OF Z I R C O N I U M A N D H A F N I U M

Hafnium in natural materials is determined by thermal-neutron activation analysis through the reaction, aS~ 7 ) 1 8 1 H f . Possible primary and spectral interferences are also discussed in REBAGAY.1 Several fairly extensive reviews dealing with the separation of zirconium and hafnium have appeared in the literature (STEINBERG,z VINAROV4), but, up to the present, even the best of these methods are comparatively complex, owing to the difficulties associated with the problems of separating closely chemically-related elements. The strong adsorption of Zr(IV) and Hf(IV) from dilute sulfuric acid and/or ammonium sulfate media by strongly basic anion resins, has been demonstrated and made the basis of their separation from scandium, the rare earths, and other elements (HAGUE and MACHLAN, 5 MACHLAN and HAGUE,6 HAMAGUCHI et al., 7 DANIELSSON,8 STRELOWand BOTHMAg).The radiochemical separation procedure outlined below has been shown to isolate zirconium and hafnium from the bulk of interfering activities in thermal-neutron-irradiated rocks and meteorites with yields adequate for Ge(Li) detection.

Experimental Sample and flux-monitor preparation and irradiation Sources and descriptions of the samples used in this study are given by REBAGAY.1 Additional details of the preparation of the zirconium and hafnium comparative standards, referred to as flux monitors, are also given by REBAGAY.x Only the main points of the experimental technique will be restated here. Aliquants of about 0.1 to 0.5 g of the powdered samples, which have been ovendried at 110 ~ for a day, were placed in clean high-purity quartz vials and heatsealed. Suitable aliquots of the standard solutions of zirconium and hafnium were transferred into similar quartz vials, containing approximately 50 mg of Johnson, Matthey 'Spec-pure' SiO2. The vials containing the zirconium or hafnium standard solutions were oven-dried for one day, and heat-sealed. Two flux monitors for each element were included in each irradiation run. Flux monitors prepared in this manner were found to yield reproducible specific activities (_+. 1-2~o). All vials containing the samples and flux monitors were wrapped in aluminum foil, and placed in a standard aluminum irradiation can. They were then shipped to Argonne National Laboratory for irradiation for 7 to 10 days at fluxes in the range 1 to 3 9 1013 n. cm -2 9 sec -1 in the CP-5, heavy-watermoderated reactor, at a position having a slow-neutron to fast-neutron ratio (ns/nf) of approximately 100/1. After irradiation, the samples were allowed to decay for 2 to 3 days before shipment to our laboratories.

The radiochemical-separation procedure The radiochemical procedure for simultaneous determination of zirconium and hafnium presented here is based in part on the procedures of HAMAGUCHI et al. 7 and ELWELL and WooD. z~ 3-. Radioanal.

Chem. 5 (1970)

T. V. R E B A G A Y , W. D. E H M A N N : D E T E R M I N A T I O N OF Z I R C O N I U M A N D H A F N I U M

53

Samples: Selected meteoritic phases, tektites, and terrestrial rocks. Irradiation time: 7to 10 days at approximately 1 to 3 9 1013n 9 cm -2 9 sec -1. Working time: Approximately 12 hrs per sample set ( 8 - 10 samples). Chemical yield: 30 to 80 ~ . Procedure: 1. Place about one gram of Na202, and a weighed amount of zirconium carrier (30 to 50 mg ZrO2) in a 50-ml nickel crucible. Open the vial containing the irradiated sample, and empty its contents into the crucible. Rinse the vial with about 5 ml of 12N HC1 and place the rinse solution in a small beaker. Wash the vial twice with 5 ml portions of 6N HC1 solution, and add the washings to the solution in the small beaker. Cover the mixture in the crucible with more Na202 ( ~ 2 g) and heat the crucible gently over a Meeker burner for about 10 min to insure complete fusion of the mixture. A dull-red fusion melt is produced. Cool the melt by blowing a stream of air across the sides of the crucible. 2. Dissolve the cold melt in about 50 ml of distilled water, and 50 ml of 12N HC1 in a 400-ml plastic beaker. This must be done with care to avoid sputtering due to excess Na202. Add the washings of the vial to the mixture in the plastic beaker. Stir the mixture with a Teflon rod, and then heat the solution under an infrared lamp to expel any remaining hydrogen peroxide. Wash the crucible with about 10 ml of 6N HC1, and add the washings to the solution. 3. Precipitate the hydroxides with excess ammonia. Centrifuge, and wash the precipitate with about 50 ml of dilute ammonia (1 part conc. NH4OH + 10 parts of distilled water). Dissolve the precipitate in 5 ml of concentrated H2SO4, and evaporate to complete dryness. Quantitative depolymerization of zirconium and hafnium ions is assured by taking the solution to fumes of sulfuric acid (STRELOW and BOTNMAg). 4. Leach the residue with about 25 ml of 0.1M (NH4)2SO4-0.025M H2SO4 mixed solution. Centrifuge, and collect the supernatant. Wash the residue thrice with the mixed solution, using 25 ml portions for each wash. Combine the washings with the supernatant. 5. Prepare the anion-exchange assembly as follows: (a) Construct the column from polyethylene tubing of 5/8" inside diameter, and 30 cm long. Use polyethylene shavings to support the resin. Fit the column with a 200-ml polyethylene dropping bottle which is connected to a plastic joint. To facilitate the entry of the influent solution, cut the bottle at the bottom with a knife. (b) Prepare the resin bed by slurrying about 8 grams of Dowex 1 - X 8 (in the chloride form) resin into the column with water. Wash the resin with 3M H2SO4 until the effluent gives a negative test with AgNO 3. Then wash it thoroughly with water until the effluent gives a negative test with BaC12. Finally, pre-equilibrate the resin with 0.1M (NH4)2SO4 - 0 . 0 2 5 M H2SO4 mixed solution by passing several column volumes of this solution through the column. Pass the solution from step 4 through the column at a flow rate of I ml/min. 6. Wash the column with about 200 ml of the mixed solution, and then with 50 ml of 1M (NH4)~SO4-0.025M H2SO 4 solution at a flow rate of 1.5 ml/min. 7. Elute Zr(IV) and Hf(IV) with 100 ml of 4M HCI. To the eluate, add a slight excess of ammonium hydroxide to precipitate the hydroxides of zirconium and 9". Radioanal.

Chem. 5 (1970)

54 T. V. R E B A G A Y , W. D. E H M A N N : D E T E R M I N A T I O N OF Z I R C O N I U M A N D H A F N I U M

hafnium. Centrifuge, and discard the supernatant. Dissolve the precipitate in 30 ml of 6M HC1; add 25 ml of 15~ by weight aqueous mandelic acid solution and digest the mixture at 8 0 - 8 5 ~ for about 30 min. Filter the precipitate, using a separable glass filter column which is fitted with a No. 42 Whatman filter paper. Discard the filtrate. 8. Without removing the filter paper containing the precipitate from the filtration assembly, dissolve the precipitate with about 50 ml of dilute ammonia (1 part conc. NH~OH + 4 parts distilled water). Slight suction may aid the flow of the solution. To the solution, add 25 ml of 15~ mandelic acid, and a few drops of 1 ~ methyl red indicator. Neutralize the solution with concentrated HC1, and then add 10 ml of the concentrated HC1 in excess. Digest the solution at 8 0 - 85 ~ 9. Filter the precipitates of mixed tetramandelates using a separable filter. Wash the precipitate with 50 ml of a solution that is 5 ~ by weight mandelic acid + 2 ~ by volume hydrochloric acid, then with 30 ml of 95 ~ ethanol, and finally with 20 ml of ethyl ether. I0. Mount the precipitate into a No. 00 gelatin capsule, and then package in a 2-dram snap-top polyethylene vial. Center the capsule in the vial with polyethylene spacers to insure uniform geometry. 11. Count the activity of the 724 keV photopeak of 95Zr, and the activity of the 482 keV photopeak of 18IHf, by using a lithium-drifted germanium detector, coupled to a multichannel pulse-height analyzer. 12. Determine the yield of the zirconium carrier non-destructively by using an automated fast-neutron activation analysis system via the reaction: 9~ 2n)SgmZr, or spectrophotometrically. Briefly, the steps involved in the chemical yield determination by activation may be summarized as follows: (a) subtraction of the initial activity of the sample prior to fast-neutron irradiation by analyzer subtraction operation; (b) irradiation of the sample with 14 MeV neutrons for a period of about 3 min; (c) counting in the analyzer addition mode the activity of 89mZr induced in the sample; (d) removal from the spectrum by spectrum stripping of 0.511 MeV annihilation radiation due to aZN resulting from the irradiation of the carbon in the vial, and of the tetramandelates of the sample, by recoil protons; (e) irradiation of zirconium comparators consisting of 99.8~ ZrO 2 for the same length of time, and under the same conditions as the sample (one comparator is usually irradiated prior to that of the sample, and another one just after the irradiation of the sample); (f) measurement of the induced sg'~Zr activity, and the subtraction of the 13N activity; (g) comparison of the integrated activity of the SgmZr of the sample with that of the zirconium comparator, to obtain the yield. The method described above for the chemical yield determination was compared with spectrophotometric methods of HAHN and W E B E R 11 and VAN SANTENet al. le for zirconium, and the chemical-yield values obtained by the various methods generally agreed to within + 5 ~.

J. RadioanaL Chem. 5 (1970)

T. Vo REBAGAY, W. D. E H M A N N

D E T E R M I N A T I O N OF Z I R C O N I U M A N D H A F N I U M

55

Counting procedures The counting system used in most of this work for measuring the abundances of zirconium and hafnium was an ORTEC 8 1 0 1 - 10 Ge(Li) detector, and its associated electronic accessories. The sensitive volume is 10 cm a, with a drift depth of 7 mm. Some of the later data were obtained using a Nuclear Diode 35 cm 3 Ge(Li) detector. A broad gain adjustment is provided by a linear amplifier (ORTEC Model 410 linear amplifier), and the expansion of a particular region o f interest in a spectrum is done by an ORTEC Model 411 biased amplifier.

0~33 . ~ "~, ~ k

.o 181tj,~ 0 346 I

151 . ~f 0.482

" 9s~r ~-~'

95Nb 0":65 '

z ~D

.im-

Energg, MeV

Fig. 1. 4096-channel )'-ray spectra of (A) zirconium and hafnium extracted from an irradiated rock sample, (B) processed zirconium flux monitor A 400-channel pulse-height analyzer was used most commonly in this work, although a 4096-channel analyzer was used occasionally. Figs 1 and 2 are typical spectra obtained following radiochemical separation and counting by a Ge(Li) detector coupled to a 4096-channel pulse-height analyzer in the region 0 to 1 MeV. The energy calibration of the pulse-height scale was done by means of known 7-rays emitted by standard sources. An outstanding feature of the spectra shown in Figs 1 and 2 is the excellent separation of the 724 keV and 756 keV photopeaks of 9SZr from the 765 keV photopeak of its daughter, 9SNb. The absence of interfering radionuclides in the region of interest is also evident. Most of the photopeaks due to 181Hflie below the 482 keV level (Fig. 2). The background radiation in the low-energy region is slightly increased by the Compton contributions of high-energy y-rays. How]. Radioanal. Chem. 5 (1970}

56

T. V. REBAGAY,

W. D. EHMANN:

DETERMINATION

OF

ZIRCONIUM

AND

HAFNIUM

ever, the 482-keV photopeak of 181Hf stands out distinctly above baseline radiations, and the magnitudes of the baseline on both sides of the photopeaks are almost the same, making delineation of the baseline simple. This peak is also free from spectral interferences. The radiochemical purity of the processed samples was established in three ways, by: (1) precise photopeak-energy calibration, (2) redetermination of experimental abundances and peak ratios following decay for an extended time, and (3) chemical recycling. Details may be found in REBAGAY.1 181HF 0133 A

~ .

.

,

181Hf 1~1Hs;.b 0 3 4 6 0.482

I

95Nb 0.765

/0756,

(3

C)_ ~o

o o

Energy, N e V

Fig. 2. 4096-channel ~-ray spectra of (A) zirconium and hafnium extracted from an irradiated rock sample, (B) processed hafnium flux monitor

Results and discussion

The analytical results for zirconium and hafnium in a variety of natural materials are summarized in Table 1. The corresponding Zr/Hf ratios are also included in the Table. Estimation o f errors

The small sample sizes (0.1 to 0.5 g) used in this work require that special care must be taken to obtain representative samples for activation. This requirement is very difficult to attain, due to the inhomogeneity of many natural materiJ. Radioanal.

Chem. 5 (1970)

T. V. REBAGAY, W. D. EHMANN: DETERMINATION OF ZIRCONIUM AND HAFNIUM 57 Table 1 Zirconium and hafnium abundances in some standard rocks and related natural materials ppm*

Hf, ppm*

A. Standard rocks DTS 1, Std. Dunite PCC-1, Std. Peridotite W - l , Std. Diabase BCR-1, Std. Basalt A G V 1 , Std. Andesite GSP 1, Std. Granodiorite G - l , Std. Granite G-2, Std. Granite GA, Std. Granite

1.2, 1.5 [1.4] 5.3, 11 [8.1] 87, 119, 123 [110] 161, 164, 226 [1841 213 624, 666 [645] 198, 205, 255 [219] 378, 397, 403 [393] 198, 235 [217]

0.01, 0.01 [0.01] 0.02, 0.03 [0.031 2.8, 2.9, 3.3 [3.01 2.8 4.1 16, 18 [171 3.8, 4.9 [4.2] 5.8 4.0, 5.0 [4.51

m

B. Tektites Mingenew, W. A. Australite Port Campbell, Vic. Australite

325 306

4.2 3.4

77 90

27

0.81

33

62

1.4

44

Zr,

Specimen

C. Other terrestrial materials Eclogite (Roberts Victor Mine, South Africa) Basalt (Mid-Atlantic Ridge, Station 20)

Zr (ppm)** Hf (ppm)

37 66 52 38 52 68 48

*Mean abundance for replicate analyses is given in square brackets **Not listed where precision is poor, or Hf content is near detection limit

als. To m i n i m i z e e r r o r s due to sampling, replicate analyses o n each s a m p l e are desirable. I n s o m e cases, the n u m b e r o f d e t e r m i n a t i o n s t h a t could be p e r f o r m e d on a given s a m p l e was d i c t a t e d b y t h e availability o f the sample. T h e p o s s i b l e errors in the d e t e r m i n a t i o n o f z i r c o n i u m a n d h a f n i u m , using the technique described above, m a y be s u m m a r i z e d as follows: (1) G r a v i m e t r i c a n d v o l u m e t r i c errors in the p r e p a r a t i o n o f the samples a n d flux m o n i t o r s - c o m p a r e d to o t h e r sources o f error, these m a y be r e g a r d e d as negligible. R e p l i c a t e flux m o n i t o r s yielded specific activities that, in general, a g r e e d to within + 1 - 2 ~ . (2) E r r o r s due to c o u n t i n g statistics a n d p h o t o p e a k i n t e g r a t i o n o f counts d e t e c t e d b y the Ge(Li) d e t e c t o r c o u n t i n g system - all p r o c e s s e d samples were c o u n t e d until at least 5000 counts were a c c u m u l a t e d a b o v e the baseline in t h e ~-ray p h o t o p e a k . F o r s o m e very l o w - a b u n d a n c e samples, this entailed 1 to 2 days o f counting. By careful d e l i n e a t i o n o f the baseline on either side o f the p h o t o p e a k , t h e e r r o r m a y be m i n i m i z e d to _+ 3 ~o. (3) E r r o r s a s s o c i a t e d with t h e yield d e t e r m i n a t i o n - these errors w o u l d be expected to b e r a n d o m . S o m e factors affecting precision in 14 M e V n e u t r o n J.

Radioanal.

Chem. 5 (1970)

58

T.V.

REBAGAY,

W. D . E H M A N N :

DETERMINATION

OF ZIRCONIUM

AND HAFNIUM

activation have been summarized by EHMANN. 13 Several determinations of the chemical yield of the same sample were usually conducted, and a conservative estimate of the error due to this part of the procedure is ___5 - 1 0 ~ . (4) Errors due to sample inhomogeneity - these are difficult to ascertain and may lead to the poor precision shown for samples such as PCC-1. ; For most of the data presented here, error limits of +__10-15~ are felt to be realistic, based on consideration of all potential sources of error. Comparison of data Most activation analysis procedures involving zirconium and hafnium determinations in rocks and meteorites that have appeared in the literature require individual determination of these elements, that is, zirconium is separated from hafnium. Other methods, such as X-ray fluorescence and emission spectrographic techniques, could determine only zirconium, due to the very low hafnium levels in most rocks and meteorites. The neutron activation procedure for these elements described here has the advantage of determining both elements simultaneously in the same sample, without the necessity of separating one from the other. It is seen in Tables 2 and 3 that the zirconium and hafnium values in the U.S. Geological Survey standard rocks obtained by numerous workers are generally in good agreement with the values presented here. Our values are within a +__10 deviation range of the values selected by FLANAGAN,14 based on all values reported in the literature. Zirconium and hafnium in some tektites The zirconium and hafnium abundances in two tektites are listed in Table]l. No published data on abundances of these elements in these particular specimens are available for comparison. However, TAYLOR'S15 data on some australites strewn at approximately the same location as these australites show a close agreement with those reported here for Mingenew and Port Campbell australites. Zirconium and hafnium in an eclogite and an oceanic basalt Data for an eclogite from the Roberts Victor Mine, and a basalt from the Mid-Atlantic Ridge, are also given in Table 1. The eclogite contains 27 ppm Zr and 0.81 ppm Hf, with a corresponding Z r / H f ratio of 33. The basalt from the Mid-Atlantic Ridge gives 62 ppm Zr and 1.4 ppm Hf, yielding a Zr/Hf ratio of 44. These samples are of possible upper-mantle origin. It is believed that the primitive composition of the upper mantle lies between those of the peridotites and basalt, but closer to the peridotites (RINGWOOD16). RINGWOOO16 estimates the primitive composition of the upper mantle as equivalent to a mixture which is approximately three parts of peridotite and one part of basalt. It seems that the eclogite analyzed here matches this theoretical composition. The Mid-Atlantic Ridge basalt, however, resembles very closely the zirconium and hafnium concentration of calcium-rich achondrites. Data on the distriJ. RadioanaL Chem. 5 (1970)

T. V. R E B A G A Y ,

W, D . E H M A N N :

DETERMINATION

OF ZIRCONIUM

AND HAFNIUM

59

Table 2 C o m p a r i s o n of zirconium a b u n d a n c e s (ppm) in s t a n d a r d silicate rocks i

W--1

G--1

G--2

GSP-1

110 100"

219 210"

393

645

91

217

316' 334 273 290 312 4OO 40O 37O 315 390 35O 25O 37O 320 328 275 340 340 3O5 260 28O 393

544* 654 323 538 613 650 650 610 640 555 270 - -

685 585 585 410 580 550 559 460 500 679

I

DTS--1 ]BCR-1

AGV 1) PCC--1

213 227* 186 207 222 198 -250 240 250 315 225 1 7 5

240 210 238 235 240 230 203 280 200 280

l

8.1

1.4 - -

-< 10 <20 <20 --< 2 <10 28 <30

- -

<10 < 20 <20 - -

-< 2 < 10 20 <30 - -

- -

<30 < 5 --< 4 < 10 < 5 -30 < 5

<30 < 5 - -

< 4 < 10 < 5 30 < 5

Reference

This work

184 185" 144 183 200 158 172 200 170 200 275 110 210 145 180 188 165 180 170 178 260 200 197

FLEISCHER t7 F L A N A G A N 14 B I N G H A M and C a o o o s ~s MAYS m S U T T O N e t a l . 2~ BERMAN 21 R O O K E 22 H O S K I N G 23

SARMA a n d DAS-04 D U T R A 25 V O L L R A T H 26 A L E Y e7 GORDON et all s L E R I C H E ea C A R M I C H A E L 36 POWERS al

SIEGERS a n d W O L F F 32 P R I C E 3a S U H R 3~

CARMICHAELet aL as+ C H A M P 36 M A J M U N D A R 37 P A R K E R as

* Compilation values + Average value of wet analysis and X-ray data Table 3 C o m p a r i s o n of h a f n i u m a b u n d a n c e s in U,S. Geological Survey s t a n d a r d rocks Abundance, ppm

Specimen BROOKS 39

Dunite (DTS-1) Peridotite ( P C C - I ) Diabase ( W - l ) Basalt ( B C R - 1 ) Andesite ( A G V - 1 ) Granodiorite (GSP-1) Granite ( G - I ) Granite (G-2)

0,01 0.08 2.2 5.4 6.5 16.7 5.8 7.7

GORDON et al. ~s

3.0 4.7 5.4 15 7.8

Compilation GREENLAND4~

0.03 0.05 1.4 3.3 3.7 11 4.5 6.1

o f F L A N A G A N 14

0.03 0.06 - -

4.4 5.1 12.4 - -

7.5

This work

0.01 0.03 3.0 2.8 4.1 17 4.2 5.8

Y. Radioanal. Chem. 5 (1970)

60 T. V. REBAGAY, W. D. EHMANN: DETERMINATION OF ZIRCONIUM AND HAFNIUM bution of zirconium and hafnium will b e p u b l i s h e d e l s e w h e r e .

in meteorites obtained

using this technique

This work has been supported in part by the U.S. Atomic Energy Commission Contract AT--(40--1)--2670.

References 1. T. V. REBAGAY, Ph.D. Dissertation, University of Kentucky, 1969. 2. C. M. LEDERER, J. M. HOLLANDER, I. PERLMAN, Table of 2sotopes. 6th ed. John Wiley, New York, 1967. 3. E. P. STEINBERG,The Radiochemistry of Zirconium and Hafnium, N A S - - N S - - 3 0 1 1 , 1960. 4. I. V. VINAROV,Russian Chem. Rev., 36 (1967) 522. 5. J. L. HAGUE, L. A. MACHLAN, J. Res. Natl. Bur. Stand. 65A (1961) 75. 6. L. A. MACHEAN, J. L. HAGUE, J. Res. Natl. Bur. Stand. 66A (1962) 517. 7. H. HAMAGUCHI, A. OUCHI, T. SHIMUZU, N. ONUMA, R. KURODA, Anal. Chem., 36 (1964) 2304. 8. L. DANIELSSON, Acta Chem. Scand., 19 (1965) 670. 9. F. W. E. STRELOW, C. J. C. BOTHMA, Anal. Chem., 39 (1967) 595. 10. W. T. ELWELL, D. F. WOOD, Analysis of the New Metals. Pergamon, London, 1966. 11. R. B. HAHN, L. WEBER, Anal. Chem., 28 (1956) 414. 12. R. T. VAN SANTEN, J. H. SCHLEWITZ,C. H. TOY, Anal. Chirn. Acta, 33 (1965) 593. 13. W. D. EHMANN, Fortschritte dee Chemischen Forschung, 14 (1970) 49. 14. F. FEANAGAN, Geoehim. Cosmochirn. Acta, 33 (1969) 81. 15. S. R. TAYLOR, Geochim. Cosrnochim. Acta, 30 (1966) 1121. 16. A. E. RINGWOOD, Advance in E a r t h Science. Ed. P. M. HURLEY, M. 2. T. Press, Camblidge, Mfissachusetts, 1966. 17. M. FLEISCHER, Geochirn. Cosmochim. Acta, 33 (1969) 65. 18. E. BINGHAM, A. A. CHODOS, Private communication, as quoted in FLANAGAN.t4 19. R. E. MAYS, Private communication, as quoted in FLANAGAN.la 20. A. L. SUTTON, JR., H. G. NEIMAN, J. HAEETY, Private communication, as quoted in FLANAGAN.14 21. S. BERMAN, Private communication, as quoted in FLANAGAN.14 22. J. M. ROOKE, Private communication, as quoted in FLANAGAN.14 23. D. E. B. HOSKIN% Private communication, as quoted in FLANAGAN.la 24. B. DAS SARMA, H. B. DAS, Private communication, as quoted in FLANAGAN.14 25. C. V. DUTRA, Private communication, as quoted in FLANAGAN.14 26. J. VOLLRATH, Private communication, as quoted in FLANAGAN.la 27. A. ALLY, Private communication, as quoted in FLANAGAN.14 28. G. E. GORDON, K. RANDLE, G. G. GOLES, J. B. CORLIS, M. H. BEESON, S. S. OXLEY, Geochirn. Cosmochirn. Acta, 32 (1968) 369. 29. H. H. LERICHE, Private communication, as quoted in FLANAGAN.14 30. 2. S. E. CARMICHAEL, Private communication, as quoted in FLANAGAN.14 31. G. M. POWER, Private communication, as quoted in FLANAGAN.14 32. A. SIEGERS, P. WOLFE, Private communication, as quoted in FLANAGAN? ~ 33. N. B. PRICE, Private communication, as quoted in FLANAGAN.14 34. N. H. SUHR, Private communication, as quoted in FLANAGAN?4 35. I. S. E. CARMICHAEL, J. HAMVEL, R. N. JACK, Chem. Geol., 3 (1968) 59. 36. W. H. CI.IAMV, Private communication, as quoted in FLANAGAN.la 37. H. H. MAJMUNDAR, Private communication, as quoted in FLANAGAN.14 38. A. PARKER, Chem. Geol., 4 (1969) 445. 39. C. K. BROOKS, Radiochim. Acta, 9 (1968) 157. 40. L. P. GREENLAND, Anal. Chirn. Acta, 42 (1968) 365. J. Radioanal. Chem. 5 (1970)