Spectrophotometric determination of tungsten in ores and steel by chloroform extraction of the tungsten-thiocyanate-diantipyrylmethane complex

Tolanta, Vol

22. pp 837-841

Pergamon

Press, 1975 Prmted

m Great Bntain

SPECTROPHOTOMETRIC DETERMINATION OF TUNGSTEN IN ORES AND STEEL BY CHLOROFORM EXTRACTION OF THE TUNGSTEN-THIOCYANATE-DIANTIPYRYLMETHANE COMPLEX ELSIEM.

DONALDSON

Mineral Sciences Division, Mines Branch, Department of Energy, Mines and Resources, Ottawa, Canada

(Received 12 February 1975. Accepted 19 March 1975)

Summary-A method for determining up to about 6% of tungsten in ores and mill products is described. It is based on the extraction of the yellow tungsten(Vtthiocyanate+liantipyrylmethane ionassociation complex into chloroform from a 2.4M sulphuric acid-7.8M hydrochloric acid medium containing ammonium hydrogen fluoride as masking agent for niobium. The molar absorptivity of the complex is 1510 l.mole-‘.mm-’ at 404nm, the wavelength of maximum absorption. Moderate amounts of molybdenum and selenium may be present in the sample solution without causing appreciable error in the result. Interference from large amounts is avoided by separating these elements from tungsten by chloroform extraction of their xanthate complexes. Large amounts of copper interfere during the extraction of tungsten because of the precipitation of cuprous thiocyanate. Common ions,

including uranium, vanadium, cobalt, titanium, arsenic and tellurium, do not interfere. The proposed method is also applicable to steel.

Recently, a more specific and sensitive thiocyanate The preparation and characterization of certified method was described for the determination of reference ores is a continuing facet of the Canadian Certified Reference Materials Project. As part of this tungsten in stee15-’ and steel-making materials.8 It project, the author was asked to participate in the involves chloroform extraction of the neutral ion-asinterlaboratory programme for the certification of sociation complex formed between tungsten(V)-thiothree tungsten ores CT-l, BH-1 and TLG-1. In pre- cyanate and tetraphenylarsonium chloride from an 8M hydrochloric acid medium and subsequent photoliminary work, a modification’ of a spectrophotometric thiocyanate method developed by Freund et metric measurement of the extract. Of the interfering al.’ was investigated. In this method, tungsten is elements mentioned above, only niobium and moderate amounts of molybdenum interfere. Interference reduced to the quinquevalent state with stannous chloride, in a 3.6M sulphuric acid-58M hydrochloric from niobium is obviated by washing the tungsten acid medium, and the absorbance of the anionic tungextract with ammonium hydrogen fluoride solusten(V)-thiocyanate complex is measured directly in tion’-’ or by extracting tungsten from a fluoride an aqueous medium’ or after extraction of the com- medium.* Because of its relative specificity, this plex into an organic solvent (isopropyl ether or n- method was considered in the present work. However, amyl alcohol).’ However, this method did not yield preliminary experiments carried out with tetraphenylconsistent results for the ores when measurement was arsonium chloride and a widely used analogous commade both in aqueous media and after extraction of pound, diantipyryhnethane,‘*” showed that the latter the complex. It is known that thiocyanate methods functioned equally well for formation of an ion-asbased on the formation of the anionic tungsten comsociation compound with the anionic tungsten-thioplex are subject to interference from coloured ions cyanate complex. Moreover, the tungsten-thioand from molybdenum, vanadium, niobium, titanium cyanate-diantipyrylmethane complex is more soluble and cobalt when measurement is made in an aqueous in chloroform than the corresponding tetraphenylarmedium, and to even greater interference from molybsonium chloride complex. Consequently, this reagent denum, vanadium, niobium and titanium after extracwas utilized in the present work, which describes the tion of the complex. 3*4Consequently, a method was successful determination of tungsten in ores, mill prosought that would be more specific and reliable and ducts and steel. Interference from large amounts of would be applicable to the determination of both molybdenum is avoided by separating it from small and moderate amounts of tungsten in ores. tungsten by chloroform extraction of its purple-red 837

838

ELSIEM. DONALDSON

xanthate complex,‘1-‘4 and from niobium ing with ammonium hydrogen fluoride.

by mask-

EXPERIMENTAL Apparatus

Funnels for filtering the extracts were made from broken 20-ml pipettes by cutting the bulb in half. Reagents Standard tungsten solution, 25 us/ml. Dissolve 0.8973 g of sodium tungstate dihydrate in water and dilute to 500ml. Dilute 5ml of this stock solution to 2OOml with water. Prepare fresh as needed. diantipyrylmethane, 1% solution in Zoo/, hydrochloric acid.

Dissolve 0.5a of 4.4’-methvlenediantinvrene in 25 ml of water contaiiing l&l of concentrated~hydrochloric acid and dilute to 50ml with water. Prepare fresh every 2 days. Stannous chloride, 45% w/v solution. Dissolve 22.5 g of stannous chloride dihydrate in concentrated hydrochloric acid and dilute to 5Oml with the same acid. Prepare fresh as required. Potassium thiocyanate, 20% w/v solution. Prepare fresh every seven days. Potassium ethyl xanthate, 200/, w/v solution. Prepare fresh as required. Ammonium hydrogen jluoride, 10% w/v solution. Tartaric acid, 7.5% and SPA w/v solutions. Sodium hydroxide, 50% w/v solution. Hydrochloric acid, 8M. Chloroform, analytical reagent grade, containing

1% of thioglycollic acid. Prepare fresh as required. Pure chloroform is also needed. Procedures Calibration curve. By burette, add 1, 2, 4, 6, 8 and 12ml of standard 25 &ml tungsten solution to six 125~ml Erlenmeyer flasks and dilute each solution to approximately 15ml with water. Add 15ml of water to a seventh flask; this constitutes the blank. Add 1Oml of concentrated sulphuric acid, 20ml of concentrated hydrochloric acid and 5 ml of 45% stannous chloride solution to each flask, mixing thoroughly after each addition, then place the flask in a boiling water-bath for 30 min. Remove the flask and cool the solution to lG15” in an ice-bath. Transfer each solution to a 125-ml polypropylene separatory funnel, marked at approximately 75m1, and wash the flask three times with cold concentrated hydrochloric acid contained in a plastic wash-bottle. Add the washings to the funnel and dilute the resulting solution to the mark with cold concentrated hydrochloric acid. Add 10 ml of 10% ammonium hydrogen fluoride solution, 1Oml of 20% potassium thiocyanate solution (Note 1) and 2ml of 1% hiantipyryhnethane solution to each funnel, stopper and mix thorouahlv. Add 10 ml of chloroform containing 1% thioglycollic acid, stopper tightly and shake for 2-m& Allow several min for the layers to separate, then drain the chloroform layer into a 60-ml glass separatory funnel. Extract twice more. by shaking for 2 min, using 05 ml of diantipyrylmethane solution each time, and 5- and 3-ml portions of chloroform, respectively (Note 2). Wash the aqueous phase by shaking it for 30 set with 3 ml of chloroform. Combine the extracts, add 1Oml of 8M hydrochloric acid and shake for approximately 30 sec. Allow several min for the layers to separate, then filter the chloroform extract through a thick wad of cotton-wool into a dry 25-ml volumetric flask. Wash the aqueous layer twice with 2-3-ml portions of chloroform containing thioglycollic acid, filter the washings into the volumetric flask and dilute to volume with chloroform containing thioglycollic acid

(Note 3). Determine the absorbance of the blank and each of the first four tungsten extracts, at 404nm, against a reference solution of chloroform containing thioglycollic acid, using 20-mm cells. Determine the absorbance of the blank and each of the last five tungsten extracts in a similar manner, using lo-mm cells. Correct the absorbance value obtained for each tungsten-thiocyanate-diantipyrylmethane extract by subtracting that obtained for the blank. Plot pg of tungsten vs. absorbance for each series of measurements. Ores and mill products. Depending on the expected tungsten content, transfer 0.2-lg of powdered sample to a 50-ml Vycor crucible, add 5 g of fused sodium bisulphate (Note 4) and mix. Cover the crucible and fuse the mixture over a low flame for 34 min to ensure the complete decomposition of tungsten minerals. Allow the melt to cool for 3-4 min, then transfer the crucible and cover to a covered 400-ml beaker containing 100 ml of 7.5% tartaric acid solution. Heat gently until the dissolution of the melt is complete, remove the crucible and cover after washing them thoroughly with 7.5% tartaric acid solution, and evanorate the solution to 140-150 ml. Filter the hot solution *hatman No. 40 paper) into a 200-ml volumetric flask and wash the beaker, paper and residue thoroughly with 7.5% tartaric acid solution. Discard the paper and residue. Cool the filtrate to room temperature and dilute to volume with 7.5% tartaric acid solution. Run a blank determination through the whole procedure. Transfer a suitable aliquot (up to 10 ml) of both the blank and sample solutions, containing not more than 0.25mg of molybdenum, to 125-ml Erlenmeyer flasks. Dilute to approximately 15 ml with water and proceed with the determination of tungsten as described above. If the aliquot taken for analysis contains more than 0.25 mg of molybdenum, transfer identical aliquots of the blank and sample solutions to 60-ml separatory funnels, add 4.5 ml of 8M hydrochloric acid and dilute to approximately 25 ml with water. Add 2ml of 20% potassium ethyl xanthate solution, mix thoroughly, then add 1Oml of chloroform and shake for 1 min. Allow the layers to separate, then drain off and discard the chloroform layer. Repeat the extraction, using 0.3-0.5 ml of xanthate solution and 5 ml of chloroform each time, until the chloroform layer is colourless (Note 5). Transfer the aqueous layer to a 125-ml Erlenmeyer flask and heat gently to remove residual chloroform. Evaporate the solution to approximately 15ml, then proceed with the reduction and subsequent determination of tungsten as described above. Steel. Transfer 0.2-l g of sample to a 400-ml beaker, add 20ml of concentrated hydrochloric acid, cover and heat gently until the sample has dissolved. Add 5 ml of concentrated nitric acid, heat until the destruction of carbides is complete, then remove the cover, add 3 or 4 drops of concentrated hydrofluoric acid and evaporate the solution to dryness to remove nitric acid. Add 5 ml of concentrated hydrochloric acid and approximately 25 ml of water to the residue and heat to dissolve the salts. Add 30ml of 50% tartaric acid solution, cool the solution to room temperature and carefully add 25ml of 50% sodium hydroxide solution. Cool the resulting solution to room temperature, transfer to a 200-ml volumetric flask and dilute to volume with water (Note 6). Run a blank determination. If the separation of molybdenum is not necessary (i.e., less than 0.25mg), transfer suitable aliquots of both the blank and sample solutions (Note 7) to 125-ml Erlenmeyer flasks and proceed with the reduction and subsequent determination of tungsten. If the separation of molybdenum is necessary, transfer suitable aliquots of both the blank and sample solutions to 60-ml separatory funnels, add 5.5 ml of 8M hydrochloric acid, dilute to 25 ml with water and proceed with the xanthate-chloroform extraction (Note 8) and subsequent determination of tungsten as described above.

Spectrophotometric

determination

Notes 1. Sodium thiocyanate cannot be used in place of potassium thiocyanate, because a dense white precipitate of sodium sulphate forms in the solution. 2. Because of the high acid content of the solution, some salts may precipitate during extraction; this does not interfere with the extraction of tungsten. 3. If the blank or tungsten extracts are slightly opalescent, or become opalescent on standing, filter a suitable portion of the extract through two dry Whatman No. 42 filter papers before the spectrophotometric measurement. 4. Potassium pyrosulphate is not recommended for fusion of the sample because of the insolubility of potassium tartrate, which crystallizes from the final solution on cooling and standing. This can cause low results for tungsten because of occlusion. 5. A three-stage extraction with a total volume of 4ml of 20% potassium ethyl xanthate solution is sufficient for the separation of 5 mg of molybdenum (also 5 mg of arsenic, selenium or tellurium). 6. If a precipitate of hydrous oxides is present, allow

the suspension to settle before an aliquot is taken for the tungsten determination. 7. If the sample is a high-tungsten steel, further dilution will be necessary before an aliquot is taken for the tungsten determination. In this case, transfer a suitable aliquot of the blank and sample solutions to lOO-ml volumetric flasks, add 5ml of 50% sodium hydroxide solution, dilute to volume with 75% tartaric acid solution, then proceed as described. 8. If the sample contains nickel, vanadium or cobalt, a colourless extract will not be obtained after the complete separation of molybdenum. These elements are partly coextracted as xanthates which continue to colour the extract. RESULTS

Extraction of the tungsten( V~thiocyanate-diantipyrylmethane complex

Previous investigators 5*6 found that an approximately 7M or more hydrochloric acid medium is necessary for the efficient extraction of tungsten-thiocyanate ion-association complexes into chloroform, following the reduction of tungsten in concentrated (12M) hydrochloric acid media with stannous chloride’ or with a mixture of stannous and titanous chlorides6 Because the reduction of tungsten in a more dilute acid medium was considered more suitable for analytical purposes, preliminary experiments were carried out to determine the feasibility of extracting the tungsten-thiocyanate-diantipyryhnethane complex from a sulphuric-hydrochloric acid medium, following the reduction of tungsten with stannous chloride in a 3.6M sulphuric acid-58M hydrochloric acid medium, according to the method described by Freund et al2 In these tests, thioglycollic acid, which is soluble in chloroform, rather than quinol dissolved in ethyl alcohol,‘j was added to the chloroform to reduce interfering organic peroxides that would reoxidize the tungsten complex. The results of these tests, which were carried out in a fluoride medium, showed that up to at least 3OOpg of tungsten could be quantitatively extracted, in three stages, from an approximately 8M hydrochloric acid medium containing the volume of sulphuric acid recommended

of tungsten

839

for reduction by Freund et a1.,2 with 2, 05 and 0.5ml of 1% diantipyryhnethane solution. However, the extracts become turbid almost immediately after filtration and dilution to volume with chloroform containing thioglycollic acid. Ethyl alcohol could not be used to clarify the extracts because of the instability of the complex in chlorofo=thyl alcohol media. It was subsequently found that turbidity can be avoided by washing the extracts with 8M hydrochloric acid. Beer’s law is obeyed over the range investigated. The absorbance of the complex remains constant for at least 24 hr. The molar absorptivity of the complex is 1510 l.mole-‘.mm-’ at 404 nm, the wavelength of maximum absorption. Reduction of tungsten

Although Freund et aL2 claimed that in 3.6M sulphuric acid-5.8M hydrochloric acid media a 5-min heating period at loo” was sufficient for the complete reduction of tungsten in pure tungsten solutions, later investigators’ found that 3&60 min may be required when an appreciable amount of phosphate is present in the sample solution. In the present work, approximately 30 min are required for solutions containing tartaric acid and matrix elements. Separation of molybdenum by extraction of its xanthate complex

The interference of molybdenum in the determination of tungsten by the thiocyanate method is known to be greater in the presence of iron because of an interelement effect.1*6*7To reduce this effect, iron can be separated from molybdenum by extraction of its chloro-complex into methyl isobutyl ketone7 or isopropyl ether’ before the formation of the tungstenthiocyanate complex. Both molybdenum and tungsten can also be separated from iron by chloroform extraction of their cc-benzoinoxime complexes.7 In the present work, it was considered that the complete removal of molybdenum from tungsten would result in a more reliable method, and published data1’-14 on the extraction of molybdenum(V) xanthate indicated that this method might provide a simple and effective means of separating moderately large amounts of molybdenum from tungsten. Preliminary experiments with tungsten solutions (1000 pg) containing tartaric acid showed that at least 1Omg of molybdenum, depending on the amount of potassium ethyl xanthate employed, can be readily extracted into chloroform, in three successive stages, from 0+3M hydrochloric acid. Extraction at lower and higher acidities was not investigated. Complete recovery of the added tungsten was obtained in all the tests and analysis of the final solutions showed that less than 3Opg of molybdenum remained in the aqueous phase after extraction. Subsequent work showed that, in the range of acid concentration (l+2M) chosen, cobalt, nickel and vanadium are partly extracted and arsenic, selenium and tellurium are completely extracted as the xanthates. Copper forms

840

EISE M.

~NALD~N

Table 2. Recovery of tungsten from synthetic tungsten ore samples

Table 1. Effect of diverse ions on the extraction of tungsten (200 Pg) Dwerse ion taken,

W found,

Dwerse ton taken.

W found.

me

M

mg

PS

200 213 220 213 206 200 19st 19st 60 128 f53 187 198 2w 197 202

Fejlii) 50 Mow) 2 Mo(Vi) 05 + Fe(II1) 50 Mo(V1) 035 + Fe(lI1) 50 Mo(V1) 030 + Fe(III) 50 Mo(V1) 0 25 + Fe(III) 50 Mo(VI) 5 Mo(V1) 5 + FefIII) SO S$IV) 10% Sew) 2 SqVl I Se(IV) a5 S&IV) 0.25 SHIV) 5 Co(H) 10 V(V) 10

W found, %

196 196 197 196 199 198 199 196 200 205 I96 197 202 199 197 196

U(w) 10 Nt0’) 10 Zn(lI) 20 Cu(l1) 10 N@I) 10 Bi(III) 10 Sb(V) 10 A&III) IO Mn(II) 10 Tt(IV) 10 zrfrvt 10 Cd(U) 10 Cr(VX) 10 As(III) 10’ Te(IV) IO* P(V) 10

*Removed, in elemental state, by filtration glass-wool before extraction of tungsten. t After x~that~hioroform extraction.

Sample” CT-1 s&e&e ore BH-1 woifram1te ore TLC&l Scheeltte ore

Standard devlatlon,

1056 (I 019-1~078)

0 020

0 409 (0.3811)444)

0.018

@077 (0069-0~086)

OGO6

%

0.119 0.329 0.579 1.079 2.079 4079

0.172 0.326 0.584 1.064 2.040 4.064

0.168 0.322 0.572 1.056 2.064 4+BO

Applications

The proposed method was applied to the analysis of a series of synthetic mixtures of a scheelite ore to which 0.5% of molybdenum was added and in which the tungsten varied from 0.10 to 4.00%. The standard tungsten solution was added to the samples just before dissolution of the sodium bisulphate melt. It was also applied to standard reference steel samples and to the three tungsten ores, CT-l, BH-1 and TLG1, currently undergoing certification for tungsten. For each of these ores, tungsten was determined according to the directives of the Mines Branch’s Canadian Certified Reference Materials Project, i.e., by using five subsamples from each of two bottles. The results of these analyses are given in Tables 2-4. DISCUSSION

Table 2 shows that the results obtained for the synthetic scheelite ore samples, containing @5X, added molybdenum, agree favourably with the total calculated amount of tungsten present, both when tun~ten was determined in the presence of molybdenum and after its separation by xanthate-chloroform extraction. The results obtained (Table 3) for the reference ores are in reasonably good agreement with the aver-

of tun~ten Laboratory

Average value and range. 46 w

After separation of MO

the extraction of ~n~ten. Large amount of copper interfere during extraction because of the precipitation of cuprous thiocyanate.

Tests carried out with moderate amounts of colouted ions, common ions and ions that are known to interfere in tungsten thiocyanate colourimetric methods by forming coloured extractable thiocyanate complexes, showed that, for the quantity of each ion tested (Table l), none, except molybdenum and selenium, interfered in the extraction and subsequent determination of tungsten by the proposed method. However, up to approximately O-25mg of molybdenum and selenium can be present in the aliquot taken for extraction without causing appreciable error. It is not known how selenium interferes but interference from larger amounts can be avoided by separating it from tungsten by xanthate-chloroform extraction. Although arsenic and tellurium (also selenium) are reduced to the elemental state during the reduction of tungsten, interference from these precipitated elements during the extraction procedure can be avoided by removing the precipitates by filtration of the solution through glass-wool before

At

of MO

through

Effect of diverse ions

Laboratory

In presence

%

Ten determinations of tungsten in the ore (TLG-1) by the proposed method gave an average result of 0.079% (cf: Table 3).

a bright yellow precipitate, which is insoluble in chloroform and remains above the chloroform layer, but is decomposed completely during evaporation and boiling of the aqueous layer to destroy excess of xanthate.

Table 3. ~terminatlon

Total W present,

Average value and range, %W

in standard reference ores

Bt

Laboratory

Standard deviation,

Average value and range,

Ths

Ct Standard dwatlon,

work

Average value and range, %W

Standard devtatwn. %

%

%W

%

1018 (@98-i 04)

0023

I.038 (1.01-1~05)

OGl5

1041 (1.02.&I 056)

Ow9

0,416 (Q412-0420)

OGO4

0.439 (0 42-O 46)

@OiO

a410 (0404-@414)

0.003

0.084 (0 077-0089)

0003

0079 (0~07‘&0~082)

0,003

* The molybdenum contents of CT-I, BH-1 and TLC-1 are 0.032, 0.023 and OO02°/0,respectively. t Tungsten determined by a thiocyanate method involving measurement of the anionic complex in aqueous media. The average value is the arithmetic mean of 10 values.

Spectrophotometric Table 4. Determination

841

of tungsten

of tungsten in N.B.S. and B.C.S. steels

Nominal composition, %

Sample NBS-5Oa Chromium-tungsten steel NBS-SOB Tungsten-chromiumvanadium steel NBS 1OlE Chromium-nickel steel NBS-123a Chromium-nickel steel (niobium-bearing) NBS-123b Niobium-tantalum stabilized stainless steel NBS-134A Molybdenum-tungsten high-speed steel NBS-153 Cobalt-molybdenumtungsten steel NBS-155 Chromium-tungsten steel BCS-220/l Tungstenmolybdenum high-speed steel BCS-246 Niobium-molybdenum 18/12 stainless steel BCS-271 Mild steel BCS-273 Mild steel BCS-281 Low-tungsten steel BCS-282 Low-tungsten steel

determination

MO, Certified value and range, % W %

W found, %

0009 18.25 (18.16-18.34)

18.2

0401 18.05 (17.95-18.14)

17.9

0.3 Mn, 0.5 Si, 3.5 Cr, 1.0 V, 0.1 Cu, 0.1 Ni, 0.04 As 0.3 Mn, 0.3 Si, 0.1 Cu, 0.1 Ni, 4.1 Cr, 1.0 V, 0.04 As 1.8 Mn, 0.4 Si, 0.4 Cu, 9.5 Ni, 18.0 Cr, 004 V. 0.2 Co 0.8 Nb, 0.04 V, 0.5 Si, 18.1 Cr

0.426 0.056

0.054t

0.12 0.11*

0.108

0.8 Nb, 0.2 Ta, 0.5 Si, 0.05 V

0.17 0.18*

0.182

0.2 Mn, 0.3 Si, 0.1 Cu, 0.1 Ni, 3.7 Cr, 1.3 V 0.2 Mn, 0.2 Si, 0.1 Cu, O-1 Ni, 4.1 Cr, 2.0 V, 8.5 Co 1.2 &In, 0.3 Si, 0.1 Cu, 0.1 Ni, 0.5 Cr, 0.02 V 5.1 Cr, 2.1 V, 01 Co, 0.2 Si, 0.3 Mn, 0.2 Ni, 0.2 Cu, 0.03 As 0.8 Nb, 18.8 Cr, 12.1 Ni, 0.1 cu 0.1 Cr, 0.1 Sn 0.1 Cr, 0.2 Cu, 0.1 Sn, 0.05 V 0.1 Si, 0.1 Mn 0.1 Si, 0.1 Mn, 0.1 Cr, 0.02 V

8.35 200 (1.97-2.05)

2.03t

8.38 1.58 (1.54-1.61)

1.55t

0.039 0.517 (0.508-0.526)

0.517

5.20 6.86 (6.78-7.00)

6.78

2.89 022 (0.19-023)

0.223t

0.19, 0.045 0.02 0.02

0.016 0.287 0.696 1.29

* N.B.S. provisional result. t Molybdenum removed by xanthate-chloroform

0.01, (0~0130019) 0.2g0 (0.271-0.282) 0.70 (0.6&0.73) 1.30 (1.28-1.32)

extraction.

age results hitherto reported in the interlaboratory programme of certification. Also, those obtained for the National Bureau of Standards and British Chemical Standards samples of steel are in good agreement with the certified values. The precision of the results for the ores CT-l, BH-1 and TLG-1 (Table 3) is superior, in most cases, to that for sets of results obtained by using thiocyanate methods based on measurement of the anionic tungsten complex. The proposed method is suitable for samples containing up to approximately 6% of tungiten but material containing larger amounts can also IX analysed with reasonable accuracy. It is more sensitive than the thiocyanate methods mentioned above, considerably more selective, and reasonably specific as far as common ions are concerned. Acknowledgement-The author thanks P. Lanthier for determining the molybdenum contents of the standard reference ores. REFERENCES

1 G. H. Faye, R. J. Guest and R. C. McAdam, The Determination of Tungsten in Ores, Concentrates and Steels, Dept. of Energy, Mines and Resources, Mines Branch, Mineral Sciences and Extraction Metallurgy Divisions, Ottawa, Technical Bulletin TB 37, May, 1962.

2. H. Freund, M. L. Wright and R. K. Brookshier, Anal. Chem., 1951, 23, 781. 3. A. G. Fogg, D. R. Marriott and D. T. Burns, Analyst, 1970, 95, 848. 4. W. T. Elwell and D. F. Wood, Analytical Chemistry of Molybdenum and YRmgsten,p. 103. Pergamon, New York, 1971. 5. H. E. Affsprung and J. W. Murphy, Anal. Chim. Acta, 1964, 30, 501. 6. A. G. Fogg, D. R. Marriott and D. T. Burns, Analyst, 1970, 95, 854. 7. A. G. Fogg, T. J. Jarvis, D. R. Marriott and D. T. Burns, ibid., 1971. 96. 475. 8. R. Kajiyama, K. ‘Ichihashi and K. Ichikawa, Bunseki Kagaku, 1969,18,1500; Chem. Abstr., 1970,72, 96387s. 9. E. M. Donaldson and V. H. E. Rolko, Determination of Cobalt and Zinc in Nickel Metal by Atomic-Absorption Spectrophotometry after Separation by Simultaneous Chloroform Extraction of their ThiocyanateDiantipyrylmethane Complexes, Dept. of Energy, Mines and Resources, Mines Branch, Mineral Sciences Divi1o, sion, Ottawa, Technical Bulletin TB 93, August, 1967. V. P. Zhivopistsev, Zavodsk. Lab., 1965, 31, 1043. 11. F. Pavelka and A. Laghi, Mikrochem. Mikrochim. Actu, 1943, 31, 138. 12. W. Geihnan and H. Bode, Z. Anal. Chem., 1948, luI, 495. 13. A. T. Pilipenko and G. I. Gridchina, Tru. Komis. Analit. Khim., Akad. Nauk S.S.S.R., 1951, 3, 178; Chem. Abstr., 1953, 47, 2648b. 14. Shu-Chaun Liang and Pao-Yun Hsu, Hua Hsiieh Hsiieh Pao, 1956, 22, 171; Chem. Abstr., 1958, 52, 6062d.