- Email: [email protected]
Determination of bismuth in ores, concentrates and non-ferrous alloys by atomic-absorption spectrophotometry after separation by diethyldithiocarbamate extraction or iron collection
Talama, Vol. 26, pp. 1119 to 1123 Pergamon Press Ltd 1979. Printed in Great Britain
DETERMINATION OF BISMUTH IN ORES, CONCENTRATES A N D N O N - F E R R O U S ALLOYS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY AFTER SEPARATION BY DIETHYLDITHIOCARBAMATE EXTRACTION OR IRON COLLECTION Etsm M. DONALDSON Mineral Sciences Laboratories, Canada Centre for Mineral and Energy Technology, Department of Energy, Mines and Resources, Ottawa, Canada (Received 24 April 1979. Accepted 16 May 1979)
Summary--Two simple, reliable and moderately rapid atomic-absorption methods for determining trace and minor amounts of bismuth in copper, nickel, molybdenum, lead and zinc concentrates and ores, and in non-ferrous alloys, are described. These methods involve the separation of bismuth from matrix elements either by chloroform extraction of its diethyldithiocarbamate (DDTC) complex, at pH 11.5-12.0, from a sodium hydroxide medium containing citric acid, tartaric acid, EDTA and potassium cyanide as complexing agents, or by co-precipitation with hydrous ferric oxide from an ammoniacal medium. Bismuth is ultimately determined, at 223.1 nm, after evaporation of the extract to dryness in the presence of nitric and perchloric acids and dissolution of the salts in 20% v/v hydrochloric acid, or by dissolution of the hydrous oxide precipitate with the same acid solution, respectively. Results obtained by both methods are compared with those obtained spectrophotometrieally by the iodide method after the separation of bismuth by DDTC and xanthate extractions.
A current facet of the Canadian Certified Reference Materials Project, sponsored by CANMET, is the certification of zinc, lead and copper sulphide concentrates, CZN-1, CPB-1 and CCU-1, for a number of minor and trace constituents as well as the principal metals. Recently, as part of this project, a spectrophotometric method was developed for determining bismuth, at the #g/g-level, in diverse sulphide ores and concentrates) This method involves the preliminary separation of bismuth by chloroform extraction of its diethyldithiocarbamate (DDTC) complex from a strongly alkaline sodium hydroxide medium containing citric acid, tartaric acid, EDTA and potassium cyanide as complexing agents. This is followed by extraction of bismuth as the xanthate and subsequent determination as the iodide. In this work, it was suggested that moderate amounts of bismuth could probably also be readily determined by atomic-absorption spectrophotometry after separation by the relatively specific DDTC extraction step. The need for a reliable atomic-absorption method for determining _<200 #g/g of bismuth in sulphide concentrates was apparent from the high results obtained in the CANMET laboratories for bismuth in CZN-1. It was also evident from the wide range of values (19-120 #g/g) obtained by atomic-absorption methods by other laboratories during the interlaboratory certification programme. Most of these methods involved neither preliminary separation nor preconcentration steps, Crown Copyrights reserved.
nor matrix matching in the calibration solutions. This paper describes the application of the DDTC extraction-atomic-absorption finish to ores, concentrates and non-ferrous alloys. It also describes a simpler and more rapid atomic-absorption method based on the separation of bismuth by co-precipitation with hydrous ferric oxide. Results obtained by both methods are compared with those obtained previously by the iodide method) EXPERIMENTAL
Apparatus
A Varian Techtron Model AA6 spectrophotometer equipped with a 10-cm laminar-flow, air-acetylene burner, and the conditions recommended by the manufacturer were used for the determination of bismuth. Reagents Standard bismuth solution, lO00 pxj/ml. Dissolve 0.5000 g of pure bismuth metal in 20 ml of concentrated nitric acid, cool and dilute the solution to 500 ml with water. Prepare a 100-#g/ml solution by diluting 25 ml of this stock solution to 250 ml with water. Prepare the diluted solution fresh as required. Citric acid-tartaric acid solution, 25% of each acid. EDTA, disodium salt-sodium hydroxide solution, 12% of each. Sodium hydroxide, 50% solution. Potassium cyanide, 20% solution. Prepare fresh as required. Sodium DDTC, 1% solution. Prepare fresh as required. required. Iron(III) sulphate solution (1 ml -- 10 rag of iron). Dissolve 25 g of ferric sulphate monohydrate in hot water
1119
1120
ELSIE M. DONALDSON
containing 5 ml of concentrated sulphuric acid, cool and dilute to 500 ml with water.
Ammonia solution 10% v/v. Hydrochloric acid, 20% v/v. Sulphuric acid, 50% v/v. Nitric acid 50% v/v. Chloroform. Analytical-reagent grade. Calibration solutions Add 20 ml of concentrated hydrochloric acid to eight 100-ml standard flasks; then, by burette, add to the first seven flasks 0.5, l, 2, 3, 5, 7.5 and l0 ml, respectively, of the standard 100-#g/ml bismuth solution. The contents of the last flask constitute the zero calibration solution. If bismuth is separated by co-precipitation with hydrous ferric oxide, add 10 ml of iron(III) sulphate solution to each flask. Dilute each solution to volume with water and mix (Note 1).
bepar,,,o)t oJ bismuth by extraction of its DDTC complex Ores and concentrates. Transfer 0.2-0.5 g of powdered sample (Note 2), containing up to approximately 1 mg of bismuth, to a 60-ml zirconium crucible. Add 3 g of sodium peroxide and mix thoroughly (Note 3). Cautiously fuse the mixture over an open flame and maintain it in the molten state for ~ 30 sec to ensure complete decomposition. Allow the melt to cool, then transfer the crucible to a covered 400-ml Teflon beaker containing ~ 80 ml of water and 25 ml of 50% sulphuric acid. When the melt has dissolved, remove the crucible after washing it thoroughly with water, then cover the beaker (Note 4) and evaporate the solution to ~ 75 ml. Remove the cover, add 5 ml of concentrated hydrofluoric acid and evaporate the solution to fumes of sulphur trioxide to remove silica and hydrogen peroxide. Cool, wash down the sides of the beaker with water, evaporate the solution until ~ 10 ml of sulphuric acid remain (Note 5), then cool to room temperature. Add ~40 m l o f water and 5 or 10 g of sodium chloride (Note 6) to the resulting solution and heat gently to dissolve the sodium salts and lead sulphate. Add 20 ml of 25% citric acid-25% tartaric acid solution, mix (Note 7) and add 50 ml of 12% EDTA-12% sodium hydroxide soluuon. Using a pH-meter if necessary (Note 8), or a Small piece of red litmus paper added to the solution, make the solution alkaline (pH 7-11) with 50% sodium hydroxide solution. Cool the solution to room temperature in a water-bath, then adjust the pH to 11.5--12.0 with 50% sodium hydroxide solution. Transfer the resulting solution to a 250-ml separatory funnel, add 30 ml of freshly prepared 20% potassium cyanide solution and mix thoroughly. Add 5 ml of freshly prepared 1% sodium DDTC solution, mix, then add 10 ml of chloroform (Note 9), stopper and shake for 1 min. Allow several min for the layers to separate, then drain the chloroform phase into a 150-ml beaker. Extract the aqueous phase twice more, in a similar manner, with 10and 5-ml portions of chloroform, then wash it by shaking it for ~ 30 sec with 5 ml of chloroform. Add 10 ml of 50% nitric acid to the combined extracts, heat in a hot water-bath to remove the chloroform, then cover the beaker and add I0 ml of concentrated perchloric acid. Evaporate the solution to fumes of perchloric acid and continue fuming for ~ 15 min to ensure the complete destruction of organic material. Remove the cover, wash down the sides of the beaker with water and evaporate the solution to dryness. Add sufficient concentrated hydrochloric acid for the concentration in the final solution to be approximately 20% by volume, warm gently to dissolve -the salts, transfer the solution to a standard flask of appropriate size (25-100 ml), dilute to volume with water and mix. Measure the absorbance of the resulting solution, at 223.1 nrn, in an oxidizing air-acetylene flame (Note 10). Determine the bismuth content of the solution by relating
the resulting value to those obtained concurrently for calibration solutions of slightly higher and lower bismuth concentrations. Lead-, tin- and copper-base alloys. Depending on the expected bismuth content, transfer 0.2-0.5 g of sample to a 400-ml meaker, cover and add 20 ml of 50% nitric acid. When the dissolution of the sample is complete, add 20 ml of 50% sulphuric acid and heat until the evolution of oxides of nitrogen ceases. Remove the cover, wash down the sides of the beaker with water and evaporate the solution until copious fumes of sulphur trioxide are evolved. Cool to room temperature, add ~40 ml of water and, depending on the amount of lead sulphate present, 5 or 10 g of sodium chloride (Note 6). Heat to dissolve the salts, then proceed with the addition of citric acid-tartaric acid and EDTA-sodium hydroxide solutions, the pH adjustment, the extraction of bismuth DDTC and the subsequent determination of bismuth as described above.
Separation of bismuth by co-precipitation with hydrous ferric oxide Following the decomposition of ores and mill products and copper-base alloys (Notes 11-13) as described above, and the ultimate evaporation of the solution to fumes of sulphur trioxide, cool and add ~ 100 ml of water. Add 5 ml of concentrated hydrochloric acid and, if necessary, sufficient iron(Ill) sulphate solution for at least 100 mg of iron to be present. Cover and heat to dissolve the soluble salts. Add sufficient concentrated ammonia solution to precipitate iron as the hydrous oxide, then add 5 ml in excess and boil the Solution to coagulate the precipitate. Allow this to settle, then filter hot (Whatman No. 40 paper) and wash the beaker twice and the paper and precipitate three times with 10% ammonia solution. Discard the filtrate and washings and place a 100-ml standard flask under the funnel. Wash down the sides of the beaker with 40 ml of 20% hydrochloric acid and add the resulting solution to the funnel containing the paper a n d precipitate. Wash the beaker twice with 20% hydrochloric acid, added from a plastic wash-bottle, and add the washings to the funnel. Wash the paper three times with the acid solution. Discard the paper. Dilute the resulting solution to volume with 20% hydrochloric acid, mix and determine the bismuth content of the solution as described above, but by comparison with absorbance values obtained for calibration solutions containing approximately the same concentration of iron(Ill).
Notes 1. The calibration solutions should be prepared fresh every week because they are not stable on prolonged standing. 2. Larger samples should not be used unless the volume of EDTA-sodium hydroxide solution that is employed subsequently for masking purposes is increased correspondingly. 3. If the sample contains little, or no, acid-insoluble material, it can be decomposed with acids in a Teflon beaker as follows. Add 10 ml of 20% v/v bromine-carbon tetrachloride solution, cover and add 15 ml of concentrated nitric acid. Allow the solution to stand for ~ 15 min, then heat gently to remove the bromine and carbon tetrachloride. Add 20 ml of 50% sulphuric acid, heat gently until the evolution of oxides of nitrogen ceases, then remove the cover and add 5 ml of concentrated hydrofluoric acid. Evaporate the solution to fumes of sulphur trioxide, then proceed as described. 4. The solution should be kept almost completely covered during the initial evaporation, to avoid loss by spray. 5. Low results will be obtained if the solution is evaporated to dryness. If this occurs, add 20 ml each of 50%
Determination of bismuth sulphuric acid and water, heat to dissolve the salts, then evaporate the solution to fumes of sulphur trioxide and proceed as described. 6. Approximately 10 g of sodium chloride should be used if more than about 250 mg of lead is present. It can be omitted if lead sulphate is absent. 7. If the subsequent DDTC extraction cannot be performed the same day, allow the solution to stand overnight at this point. Addition of the EDTA-sodium hydroxide solution is not harmful except that the EDTA will precipitate from the acidic solution during prolonged standing. 8. A pH-meter is.only necessary for highly coloured copper and nickel solutions of low iron content. If an appreciable amount of iron is present, add 50% sodium hydroxide solution until the solution changes colour because of formation of the reddish-brown iron(llI)-EDTA complex. 9. Carbon tetrachloride instead of chloroform is not recommended for the extraction of bismuth from solutions containing an appreciable amount of lead. Lead DDTC, which is less soluble in this solvent, precipitates in the organic phase and may interfere mechanically with the extraction of bismuth. 10. Scale expansion (~ 2.5--5-fold) is recommended for the determination of approximately 1 #g/ml or less of bismuth. 11. This method is not recommended for samples containing more than 50 mg of aluminium, antimony or tin or more than 200 mg of lead. 12. Samples containing more than 1 mg of bismuth can be taken if the final solution is diluted to an appropriate volume with 20~ hydrochloric acid and if the calibration solutions contain approximately the same concentration of iron(Ill). 13. Up to 1 g of sample can be taken if the iron content does not exceed approximately 20~o and if large amounts of other elements that form insoluble hydrous oxides (Note 11) are absent. Approximately 4 g of sodium peroxide should be used for the decomposition of 1 g of sample. RESULTS
Atomic-absorption finish after separation of bismuth by extraction of its DDTC complex It was shown previously~ that lead is partly coextracted into chloroform as the DDTC complex, at pH 11.5--12.0, from a sodium hydroxide medium containing citric acid, tartaric acid, EDTA and potassium cyanide as complexing agents. Tests, in which the extract was ultimately treated with nitric and perchloric acids as described in the proposed method, showed that approximately 14--17 nag of lead are coextracted, at the 300-mg level, when the recommended amount of sodium DDTC (50 mg) is used. However, up to at least 2000 #g/ml can be present in the final solution taken for analysis without interfering in the determination of bismuth by atomic-absorption spectrophotometry under the proposed conditions. The use of less sodium DDTC to decrease the co-extraction of lead is not recommended. It reduces the rate of complex formation and may result in incomplete extraction of bismuth. Thallium(III) is completely coextracted as the DDTC complex but up to at least 50 #g/ml will not interfere.
Atomic-absorption finish after separation of bismuth by co-precipitation with hydrous ferric oxide Co-precipitation procedures, involving the hydrous oxides of lanthanum,2-5 titanium6 and zirconium,7
! 121
followed by an atomic-absorption finish, have been used for the separation and determination of bismuth and other elements in sulphide concentrates,4,s various metals2'7 and leach solutions: Iron(Ill) has also been recommended for co-precipitation purposes, s'9 and has been used previously by the author for the preliminary separation of arsenic ~° and tellurium~ from matrix elements in copper, nickel, molybdenum and zinc concentrates before their extraction as xanthates and subsequent spectrophotometric determination. Consequently, it was considered that this separation step, in conjunction with an atomic-absorption finish, should also provide a rapid, simple and reliable method for determining bismuth in diverse ores and concentrates. The use of the other co-precipitants mentioned above was not considered necessary or desirable, where only bismuth is concerned, because most sulphide ores and concentrates usually contain sufficient iron to co-precipitate small amounts of bismuth completely. Furthermore, large amounts of these elements, in conjunction with a large amount of iron, not only increase the metal ion content of the final solution, but also produce a bulky precipitate that is slow to filter and dissolve. Tests, in which bismuth was separated by co-precipitation with 100 mg of iron(III), as described previously,l o.1 ~ followed by dissolution of the precipitate in 20% hydrochloric acid, yielded complete recovery of 20- and 1000-/ag amounts of bismuth. Further tests showed that up to 50 mg (or 500 pg/ml in the final solution) of manganese(II), antimony(V), zirconium, aluminium and tin(IV), which also form insoluble hydrous oxides, will not interfere either in the coprecipitation or in the subsequent determination o f bismuth by atomic-absorption spectrophotometry. Larger amounts of tin and antimony cause low results for bismuth because of the slow and incomplete dissolution of the precipitate. A larger amount of aluminium results in a solution that passes very slowly through the filter paper. It also causes slightly high results for bismuth. Up to at least 2000 pg/ml of lead and iron(Ill), 500 pg/ml of nickel, copper(II), sodium and zinc, 300 #g/ml o f molybdenum(VI) and arsenic(V), and 50/ag/ml of indium and thallium(Ill) can be present in the final solution, without interfering in the determin~/tion of bismuth. More than 2000 pg/ml of lead may result in the precipitation of lead chloride in the solution.
Applications To test the reliability of the proposed methods, they were applied to the analyses of the CCRMP zinc, lead and copper concentrates, CZN-1, CPB-1 and CCU-I, and to two CCRMP reference ores, MP-1 and PR-I, that have been certified for bismuth. The methods, where applicable, were also applied to certified reference copper-, tin- and lead-base alloys. The results of these analyses are given in Table 1. Except as indicated in the footnotes to Table 1, 0.5-g samples were taken in these tests.
1122
ELsm M. DONALDSON
¢'-1
-
o.
e.
~
=
~.
e
~.
II_lll
.
g.
o=
r. q l $ ~ e o ....
~.~.~.e.
0
r~
Z
o
2 eL
e~
0
8
•~
8
e~
o.
8 2 ,..
~
cq
g
0
0 e.,
0 e~
8 =
..~
~ ~ ' ~
.o
o~
z
.~
-~
.~
~
g-~,, ~ ~.
Z
o
[-
0
•-
°
~
°
~L
~
~
.1"-4
~
-
-
~ v ~
~
v
zzzzzz==
~2 =
=_o
..~
Determination of bismuth DISCUSSION Table 1 shows that the results obtained for the CCRMP reference ores and concentrates and for the National Bureau of Standards and British Chemical Standards non-ferrous alloys (where applicable) by the two methods, which involve completely different separation procedures, are in excellent agreement with each other, with those obtained previously by the author using the DDTC-xanthate extractioni o d i d e method, 1 and with the respective certified values. Unfortunately, a consensus value is not available for the CCRMP copper concentrate, CCU-1, because it was not analysed for bismuth during the interlaboratory certification programme. Furthermore, only a consensus value is available for CZN-1 because it could not be certified for bismuth. The precision of the results obtained for CZN-1 by the three methods suggests strongly that the wide range of values obtained in the interlaboratory certification programme is due to uncompensated interelement and/or matrix effects. These effects, as shown by the results obtained by one laboratory (0.0024%), can be minimized by simulating the composition of the sample in the calibration solutions. However, the inherent disadvantage in this method is that it is only applicable to samples in which the approximate content of the predominant matrix elements is known. Similarly, although matrix effects can also be eliminated, to some extent, by using the standard additions method, the results obtained by another participating laboratory (0.0092%), and work by other investigators, t2 has shown that this method is not always reliable. In the proposed method based on the separation of bismuth by extraction of its DDTC complex, the addition of lead, before the extraction, to eliminate interference from molybdenum and zinc, as described previously, ~ is not necessary. These elements only cause low (unexplained) results for bismuth if it is subsequently stripped from the extract by shaking it with 12M hydrochloric acid. Earlier, ~ the author stated that the probable lower limit of this method is about 0.01%. However, this estimate was not based on the use of scale expansion for the determination of bismuth. If an expanded scale is used, the method
1123
is suitable for samples containing ~ 0.0005% or more of bismuth. If desired, this lower limit can be decreased to ~0.0002% if the salts ultimately obtained are dissolved in 2 ml of concentrated hydrochloric acid and the resultant solution is diluted to 10 ml. A method involving the extraction of an analogous bismuth complex, formed with ammonium pyrrolidinedithiocarbamate, into methyl isobutyl ketone, at pH ~ 10.5, from an EDTA-potassium cyanide-citrate medium, followed by the direct atomic-absorption determination of bismuth in the extract has been reported. 13 However, although it has been shown that this method yields excellent results for bismuth in non-ferrous alloys, the proposed method is more suitable for routine work because of the greater stability and ease of preparation of aqueous calibration solutions. The method based on co-precipitation of bismuth with iron(Ill) is suitable for samples containing ~0.001% or more of bismuth if a 1-g sample is taken. This lower limit can also be decreased to ~ 0.0005% if the solution obtained after dissolution of the precipitate is concentrated by evaporation and ultimately diluted to 50 ml. This method is not recommended for samples of high aluminium, tin, antimony and/or lead contents. Both methods are also applicable to nickel and molybdenum ores and concentrates. REFERENCES
1. E. M. Donaldson, Talanta, 1978, 25, 131. 2. W. Reichel and B. G. Bleakley, Anal. Chem., 1974, 46, 59. 3. M. Thompson, B. Pahlavanpour, S. J. Walton and G. F. Kirkbright, Analyst, 1978, 103, 705. 4. E. A. Jones, Natl. Inst. Metallurgy, Johannesburg, Rept., No. 1787, 1976. 5. R. V. D. Robert, ibid., No. 1838, 1976. 6. S. J. Royal and K. Dixon, ibid., No. 1795, 1976. 7. M. Malusecka and H. Jedrzejewska, Chem. Anal. (Warsaw), 1976, 21, 575. 8. ldem, ibid., 1976, 21, 585. 9. J. D. Mullen, Talanta, 1976, 23, 846. 10. E. M. Donaldson, ibid., 1977, 24, 105. 11. Idem, ibid., 1976, 23, 823. 12. J. W. Hosking, K. R. Oliver and B. T. Sturman, Anal. Chem., 1979, 51, 307. 13. H. K. Y. Lau, H. A. Droll and P. F. Lott, Anal. Chim. Acta, 1971, 56, 7.
DETERMINATION OF BISMUTH IN ORES, CONCENTRATES A N D N O N - F E R R O U S ALLOYS BY ATOMIC-ABSORPTION SPECTROPHOTOMETRY AFTER SEPARATION BY DIETHYLDITHIOCARBAMATE EXTRACTION OR IRON COLLECTION Etsm M. DONALDSON Mineral Sciences Laboratories, Canada Centre for Mineral and Energy Technology, Department of Energy, Mines and Resources, Ottawa, Canada (Received 24 April 1979. Accepted 16 May 1979)
Summary--Two simple, reliable and moderately rapid atomic-absorption methods for determining trace and minor amounts of bismuth in copper, nickel, molybdenum, lead and zinc concentrates and ores, and in non-ferrous alloys, are described. These methods involve the separation of bismuth from matrix elements either by chloroform extraction of its diethyldithiocarbamate (DDTC) complex, at pH 11.5-12.0, from a sodium hydroxide medium containing citric acid, tartaric acid, EDTA and potassium cyanide as complexing agents, or by co-precipitation with hydrous ferric oxide from an ammoniacal medium. Bismuth is ultimately determined, at 223.1 nm, after evaporation of the extract to dryness in the presence of nitric and perchloric acids and dissolution of the salts in 20% v/v hydrochloric acid, or by dissolution of the hydrous oxide precipitate with the same acid solution, respectively. Results obtained by both methods are compared with those obtained spectrophotometrieally by the iodide method after the separation of bismuth by DDTC and xanthate extractions.
A current facet of the Canadian Certified Reference Materials Project, sponsored by CANMET, is the certification of zinc, lead and copper sulphide concentrates, CZN-1, CPB-1 and CCU-1, for a number of minor and trace constituents as well as the principal metals. Recently, as part of this project, a spectrophotometric method was developed for determining bismuth, at the #g/g-level, in diverse sulphide ores and concentrates) This method involves the preliminary separation of bismuth by chloroform extraction of its diethyldithiocarbamate (DDTC) complex from a strongly alkaline sodium hydroxide medium containing citric acid, tartaric acid, EDTA and potassium cyanide as complexing agents. This is followed by extraction of bismuth as the xanthate and subsequent determination as the iodide. In this work, it was suggested that moderate amounts of bismuth could probably also be readily determined by atomic-absorption spectrophotometry after separation by the relatively specific DDTC extraction step. The need for a reliable atomic-absorption method for determining _<200 #g/g of bismuth in sulphide concentrates was apparent from the high results obtained in the CANMET laboratories for bismuth in CZN-1. It was also evident from the wide range of values (19-120 #g/g) obtained by atomic-absorption methods by other laboratories during the interlaboratory certification programme. Most of these methods involved neither preliminary separation nor preconcentration steps, Crown Copyrights reserved.
nor matrix matching in the calibration solutions. This paper describes the application of the DDTC extraction-atomic-absorption finish to ores, concentrates and non-ferrous alloys. It also describes a simpler and more rapid atomic-absorption method based on the separation of bismuth by co-precipitation with hydrous ferric oxide. Results obtained by both methods are compared with those obtained previously by the iodide method) EXPERIMENTAL
Apparatus
A Varian Techtron Model AA6 spectrophotometer equipped with a 10-cm laminar-flow, air-acetylene burner, and the conditions recommended by the manufacturer were used for the determination of bismuth. Reagents Standard bismuth solution, lO00 pxj/ml. Dissolve 0.5000 g of pure bismuth metal in 20 ml of concentrated nitric acid, cool and dilute the solution to 500 ml with water. Prepare a 100-#g/ml solution by diluting 25 ml of this stock solution to 250 ml with water. Prepare the diluted solution fresh as required. Citric acid-tartaric acid solution, 25% of each acid. EDTA, disodium salt-sodium hydroxide solution, 12% of each. Sodium hydroxide, 50% solution. Potassium cyanide, 20% solution. Prepare fresh as required. Sodium DDTC, 1% solution. Prepare fresh as required. required. Iron(III) sulphate solution (1 ml -- 10 rag of iron). Dissolve 25 g of ferric sulphate monohydrate in hot water
1119
1120
ELSIE M. DONALDSON
containing 5 ml of concentrated sulphuric acid, cool and dilute to 500 ml with water.
Ammonia solution 10% v/v. Hydrochloric acid, 20% v/v. Sulphuric acid, 50% v/v. Nitric acid 50% v/v. Chloroform. Analytical-reagent grade. Calibration solutions Add 20 ml of concentrated hydrochloric acid to eight 100-ml standard flasks; then, by burette, add to the first seven flasks 0.5, l, 2, 3, 5, 7.5 and l0 ml, respectively, of the standard 100-#g/ml bismuth solution. The contents of the last flask constitute the zero calibration solution. If bismuth is separated by co-precipitation with hydrous ferric oxide, add 10 ml of iron(III) sulphate solution to each flask. Dilute each solution to volume with water and mix (Note 1).
bepar,,,o)t oJ bismuth by extraction of its DDTC complex Ores and concentrates. Transfer 0.2-0.5 g of powdered sample (Note 2), containing up to approximately 1 mg of bismuth, to a 60-ml zirconium crucible. Add 3 g of sodium peroxide and mix thoroughly (Note 3). Cautiously fuse the mixture over an open flame and maintain it in the molten state for ~ 30 sec to ensure complete decomposition. Allow the melt to cool, then transfer the crucible to a covered 400-ml Teflon beaker containing ~ 80 ml of water and 25 ml of 50% sulphuric acid. When the melt has dissolved, remove the crucible after washing it thoroughly with water, then cover the beaker (Note 4) and evaporate the solution to ~ 75 ml. Remove the cover, add 5 ml of concentrated hydrofluoric acid and evaporate the solution to fumes of sulphur trioxide to remove silica and hydrogen peroxide. Cool, wash down the sides of the beaker with water, evaporate the solution until ~ 10 ml of sulphuric acid remain (Note 5), then cool to room temperature. Add ~40 m l o f water and 5 or 10 g of sodium chloride (Note 6) to the resulting solution and heat gently to dissolve the sodium salts and lead sulphate. Add 20 ml of 25% citric acid-25% tartaric acid solution, mix (Note 7) and add 50 ml of 12% EDTA-12% sodium hydroxide soluuon. Using a pH-meter if necessary (Note 8), or a Small piece of red litmus paper added to the solution, make the solution alkaline (pH 7-11) with 50% sodium hydroxide solution. Cool the solution to room temperature in a water-bath, then adjust the pH to 11.5--12.0 with 50% sodium hydroxide solution. Transfer the resulting solution to a 250-ml separatory funnel, add 30 ml of freshly prepared 20% potassium cyanide solution and mix thoroughly. Add 5 ml of freshly prepared 1% sodium DDTC solution, mix, then add 10 ml of chloroform (Note 9), stopper and shake for 1 min. Allow several min for the layers to separate, then drain the chloroform phase into a 150-ml beaker. Extract the aqueous phase twice more, in a similar manner, with 10and 5-ml portions of chloroform, then wash it by shaking it for ~ 30 sec with 5 ml of chloroform. Add 10 ml of 50% nitric acid to the combined extracts, heat in a hot water-bath to remove the chloroform, then cover the beaker and add I0 ml of concentrated perchloric acid. Evaporate the solution to fumes of perchloric acid and continue fuming for ~ 15 min to ensure the complete destruction of organic material. Remove the cover, wash down the sides of the beaker with water and evaporate the solution to dryness. Add sufficient concentrated hydrochloric acid for the concentration in the final solution to be approximately 20% by volume, warm gently to dissolve -the salts, transfer the solution to a standard flask of appropriate size (25-100 ml), dilute to volume with water and mix. Measure the absorbance of the resulting solution, at 223.1 nrn, in an oxidizing air-acetylene flame (Note 10). Determine the bismuth content of the solution by relating
the resulting value to those obtained concurrently for calibration solutions of slightly higher and lower bismuth concentrations. Lead-, tin- and copper-base alloys. Depending on the expected bismuth content, transfer 0.2-0.5 g of sample to a 400-ml meaker, cover and add 20 ml of 50% nitric acid. When the dissolution of the sample is complete, add 20 ml of 50% sulphuric acid and heat until the evolution of oxides of nitrogen ceases. Remove the cover, wash down the sides of the beaker with water and evaporate the solution until copious fumes of sulphur trioxide are evolved. Cool to room temperature, add ~40 ml of water and, depending on the amount of lead sulphate present, 5 or 10 g of sodium chloride (Note 6). Heat to dissolve the salts, then proceed with the addition of citric acid-tartaric acid and EDTA-sodium hydroxide solutions, the pH adjustment, the extraction of bismuth DDTC and the subsequent determination of bismuth as described above.
Separation of bismuth by co-precipitation with hydrous ferric oxide Following the decomposition of ores and mill products and copper-base alloys (Notes 11-13) as described above, and the ultimate evaporation of the solution to fumes of sulphur trioxide, cool and add ~ 100 ml of water. Add 5 ml of concentrated hydrochloric acid and, if necessary, sufficient iron(Ill) sulphate solution for at least 100 mg of iron to be present. Cover and heat to dissolve the soluble salts. Add sufficient concentrated ammonia solution to precipitate iron as the hydrous oxide, then add 5 ml in excess and boil the Solution to coagulate the precipitate. Allow this to settle, then filter hot (Whatman No. 40 paper) and wash the beaker twice and the paper and precipitate three times with 10% ammonia solution. Discard the filtrate and washings and place a 100-ml standard flask under the funnel. Wash down the sides of the beaker with 40 ml of 20% hydrochloric acid and add the resulting solution to the funnel containing the paper a n d precipitate. Wash the beaker twice with 20% hydrochloric acid, added from a plastic wash-bottle, and add the washings to the funnel. Wash the paper three times with the acid solution. Discard the paper. Dilute the resulting solution to volume with 20% hydrochloric acid, mix and determine the bismuth content of the solution as described above, but by comparison with absorbance values obtained for calibration solutions containing approximately the same concentration of iron(Ill).
Notes 1. The calibration solutions should be prepared fresh every week because they are not stable on prolonged standing. 2. Larger samples should not be used unless the volume of EDTA-sodium hydroxide solution that is employed subsequently for masking purposes is increased correspondingly. 3. If the sample contains little, or no, acid-insoluble material, it can be decomposed with acids in a Teflon beaker as follows. Add 10 ml of 20% v/v bromine-carbon tetrachloride solution, cover and add 15 ml of concentrated nitric acid. Allow the solution to stand for ~ 15 min, then heat gently to remove the bromine and carbon tetrachloride. Add 20 ml of 50% sulphuric acid, heat gently until the evolution of oxides of nitrogen ceases, then remove the cover and add 5 ml of concentrated hydrofluoric acid. Evaporate the solution to fumes of sulphur trioxide, then proceed as described. 4. The solution should be kept almost completely covered during the initial evaporation, to avoid loss by spray. 5. Low results will be obtained if the solution is evaporated to dryness. If this occurs, add 20 ml each of 50%
Determination of bismuth sulphuric acid and water, heat to dissolve the salts, then evaporate the solution to fumes of sulphur trioxide and proceed as described. 6. Approximately 10 g of sodium chloride should be used if more than about 250 mg of lead is present. It can be omitted if lead sulphate is absent. 7. If the subsequent DDTC extraction cannot be performed the same day, allow the solution to stand overnight at this point. Addition of the EDTA-sodium hydroxide solution is not harmful except that the EDTA will precipitate from the acidic solution during prolonged standing. 8. A pH-meter is.only necessary for highly coloured copper and nickel solutions of low iron content. If an appreciable amount of iron is present, add 50% sodium hydroxide solution until the solution changes colour because of formation of the reddish-brown iron(llI)-EDTA complex. 9. Carbon tetrachloride instead of chloroform is not recommended for the extraction of bismuth from solutions containing an appreciable amount of lead. Lead DDTC, which is less soluble in this solvent, precipitates in the organic phase and may interfere mechanically with the extraction of bismuth. 10. Scale expansion (~ 2.5--5-fold) is recommended for the determination of approximately 1 #g/ml or less of bismuth. 11. This method is not recommended for samples containing more than 50 mg of aluminium, antimony or tin or more than 200 mg of lead. 12. Samples containing more than 1 mg of bismuth can be taken if the final solution is diluted to an appropriate volume with 20~ hydrochloric acid and if the calibration solutions contain approximately the same concentration of iron(Ill). 13. Up to 1 g of sample can be taken if the iron content does not exceed approximately 20~o and if large amounts of other elements that form insoluble hydrous oxides (Note 11) are absent. Approximately 4 g of sodium peroxide should be used for the decomposition of 1 g of sample. RESULTS
Atomic-absorption finish after separation of bismuth by extraction of its DDTC complex It was shown previously~ that lead is partly coextracted into chloroform as the DDTC complex, at pH 11.5--12.0, from a sodium hydroxide medium containing citric acid, tartaric acid, EDTA and potassium cyanide as complexing agents. Tests, in which the extract was ultimately treated with nitric and perchloric acids as described in the proposed method, showed that approximately 14--17 nag of lead are coextracted, at the 300-mg level, when the recommended amount of sodium DDTC (50 mg) is used. However, up to at least 2000 #g/ml can be present in the final solution taken for analysis without interfering in the determination of bismuth by atomic-absorption spectrophotometry under the proposed conditions. The use of less sodium DDTC to decrease the co-extraction of lead is not recommended. It reduces the rate of complex formation and may result in incomplete extraction of bismuth. Thallium(III) is completely coextracted as the DDTC complex but up to at least 50 #g/ml will not interfere.
Atomic-absorption finish after separation of bismuth by co-precipitation with hydrous ferric oxide Co-precipitation procedures, involving the hydrous oxides of lanthanum,2-5 titanium6 and zirconium,7
! 121
followed by an atomic-absorption finish, have been used for the separation and determination of bismuth and other elements in sulphide concentrates,4,s various metals2'7 and leach solutions: Iron(Ill) has also been recommended for co-precipitation purposes, s'9 and has been used previously by the author for the preliminary separation of arsenic ~° and tellurium~ from matrix elements in copper, nickel, molybdenum and zinc concentrates before their extraction as xanthates and subsequent spectrophotometric determination. Consequently, it was considered that this separation step, in conjunction with an atomic-absorption finish, should also provide a rapid, simple and reliable method for determining bismuth in diverse ores and concentrates. The use of the other co-precipitants mentioned above was not considered necessary or desirable, where only bismuth is concerned, because most sulphide ores and concentrates usually contain sufficient iron to co-precipitate small amounts of bismuth completely. Furthermore, large amounts of these elements, in conjunction with a large amount of iron, not only increase the metal ion content of the final solution, but also produce a bulky precipitate that is slow to filter and dissolve. Tests, in which bismuth was separated by co-precipitation with 100 mg of iron(III), as described previously,l o.1 ~ followed by dissolution of the precipitate in 20% hydrochloric acid, yielded complete recovery of 20- and 1000-/ag amounts of bismuth. Further tests showed that up to 50 mg (or 500 pg/ml in the final solution) of manganese(II), antimony(V), zirconium, aluminium and tin(IV), which also form insoluble hydrous oxides, will not interfere either in the coprecipitation or in the subsequent determination o f bismuth by atomic-absorption spectrophotometry. Larger amounts of tin and antimony cause low results for bismuth because of the slow and incomplete dissolution of the precipitate. A larger amount of aluminium results in a solution that passes very slowly through the filter paper. It also causes slightly high results for bismuth. Up to at least 2000 pg/ml of lead and iron(Ill), 500 pg/ml of nickel, copper(II), sodium and zinc, 300 #g/ml o f molybdenum(VI) and arsenic(V), and 50/ag/ml of indium and thallium(Ill) can be present in the final solution, without interfering in the determin~/tion of bismuth. More than 2000 pg/ml of lead may result in the precipitation of lead chloride in the solution.
Applications To test the reliability of the proposed methods, they were applied to the analyses of the CCRMP zinc, lead and copper concentrates, CZN-1, CPB-1 and CCU-I, and to two CCRMP reference ores, MP-1 and PR-I, that have been certified for bismuth. The methods, where applicable, were also applied to certified reference copper-, tin- and lead-base alloys. The results of these analyses are given in Table 1. Except as indicated in the footnotes to Table 1, 0.5-g samples were taken in these tests.
1122
ELsm M. DONALDSON
¢'-1
-
o.
e.
~
=
~.
e
~.
II_lll
.
g.
o=
r. q l $ ~ e o ....
~.~.~.e.
0
r~
Z
o
2 eL
e~
0
8
•~
8
e~
o.
8 2 ,..
~
cq
g
0
0 e.,
0 e~
8 =
..~
~ ~ ' ~
.o
o~
z
.~
-~
.~
~
g-~,, ~ ~.
Z
o
[-
0
•-
°
~
°
~L
~
~
.1"-4
~
-
-
~ v ~
~
v
zzzzzz==
~2 =
=_o
..~
Determination of bismuth DISCUSSION Table 1 shows that the results obtained for the CCRMP reference ores and concentrates and for the National Bureau of Standards and British Chemical Standards non-ferrous alloys (where applicable) by the two methods, which involve completely different separation procedures, are in excellent agreement with each other, with those obtained previously by the author using the DDTC-xanthate extractioni o d i d e method, 1 and with the respective certified values. Unfortunately, a consensus value is not available for the CCRMP copper concentrate, CCU-1, because it was not analysed for bismuth during the interlaboratory certification programme. Furthermore, only a consensus value is available for CZN-1 because it could not be certified for bismuth. The precision of the results obtained for CZN-1 by the three methods suggests strongly that the wide range of values obtained in the interlaboratory certification programme is due to uncompensated interelement and/or matrix effects. These effects, as shown by the results obtained by one laboratory (0.0024%), can be minimized by simulating the composition of the sample in the calibration solutions. However, the inherent disadvantage in this method is that it is only applicable to samples in which the approximate content of the predominant matrix elements is known. Similarly, although matrix effects can also be eliminated, to some extent, by using the standard additions method, the results obtained by another participating laboratory (0.0092%), and work by other investigators, t2 has shown that this method is not always reliable. In the proposed method based on the separation of bismuth by extraction of its DDTC complex, the addition of lead, before the extraction, to eliminate interference from molybdenum and zinc, as described previously, ~ is not necessary. These elements only cause low (unexplained) results for bismuth if it is subsequently stripped from the extract by shaking it with 12M hydrochloric acid. Earlier, ~ the author stated that the probable lower limit of this method is about 0.01%. However, this estimate was not based on the use of scale expansion for the determination of bismuth. If an expanded scale is used, the method
1123
is suitable for samples containing ~ 0.0005% or more of bismuth. If desired, this lower limit can be decreased to ~0.0002% if the salts ultimately obtained are dissolved in 2 ml of concentrated hydrochloric acid and the resultant solution is diluted to 10 ml. A method involving the extraction of an analogous bismuth complex, formed with ammonium pyrrolidinedithiocarbamate, into methyl isobutyl ketone, at pH ~ 10.5, from an EDTA-potassium cyanide-citrate medium, followed by the direct atomic-absorption determination of bismuth in the extract has been reported. 13 However, although it has been shown that this method yields excellent results for bismuth in non-ferrous alloys, the proposed method is more suitable for routine work because of the greater stability and ease of preparation of aqueous calibration solutions. The method based on co-precipitation of bismuth with iron(Ill) is suitable for samples containing ~0.001% or more of bismuth if a 1-g sample is taken. This lower limit can also be decreased to ~ 0.0005% if the solution obtained after dissolution of the precipitate is concentrated by evaporation and ultimately diluted to 50 ml. This method is not recommended for samples of high aluminium, tin, antimony and/or lead contents. Both methods are also applicable to nickel and molybdenum ores and concentrates. REFERENCES
1. E. M. Donaldson, Talanta, 1978, 25, 131. 2. W. Reichel and B. G. Bleakley, Anal. Chem., 1974, 46, 59. 3. M. Thompson, B. Pahlavanpour, S. J. Walton and G. F. Kirkbright, Analyst, 1978, 103, 705. 4. E. A. Jones, Natl. Inst. Metallurgy, Johannesburg, Rept., No. 1787, 1976. 5. R. V. D. Robert, ibid., No. 1838, 1976. 6. S. J. Royal and K. Dixon, ibid., No. 1795, 1976. 7. M. Malusecka and H. Jedrzejewska, Chem. Anal. (Warsaw), 1976, 21, 575. 8. ldem, ibid., 1976, 21, 585. 9. J. D. Mullen, Talanta, 1976, 23, 846. 10. E. M. Donaldson, ibid., 1977, 24, 105. 11. Idem, ibid., 1976, 23, 823. 12. J. W. Hosking, K. R. Oliver and B. T. Sturman, Anal. Chem., 1979, 51, 307. 13. H. K. Y. Lau, H. A. Droll and P. F. Lott, Anal. Chim. Acta, 1971, 56, 7.