Analyst, June, 1968, Vol. 93, pp. 375-382 375 Studies of the Separation of Trace Metals by the Manganese Dioxide “Collection” Method Part 111.” The Behaviour of Copper and Zinc. Further Studies of the Behaviour of Lead and Tin: The Determination of Lead and Tin in Coppepbase Alloys BY C . M. PYBURN AND G. F. REYNOLDS? (Loughborough University of Technology, Loughborough, Leicestershire) Studies of the collection of copper and zinc by co-precipitation with manganese dioxide are reported. The collection of copper is shown to be small under all conditions studied; that of zinc is negligible. Further studies of the collection of lead and tin are presented, and a method for their separate and simultaneous determination is described. The virtual non-collection of copper and zinc is used as a basis for the application of these methods to the determination of lead and tin in copper-base alloys.The collection of lead and tin is made from a 0-008 M acidic solution containing ethanol to hold the tin in solution. After dissolution of the precipitate, tin and lead are determined together by polarography in acidic chloride solution. Lead is then determined alone by polarography in alkaline mannitol solution and the tin content found by difference. Results are presented and some aspects discussed. THE use of manganese dioxide, produced by the reaction of manganese(I1) ion with potassium permanganate, as a “collector” for the co-precipitation of metals was first described by Blumentha1,l who used it for antimony. He indicated that the procedure could also be applied to bismuth and tin.Kallman and Prestira2 showed that the precipitation of bismuth was incomplete in acidic solutions of concentration greater than 0.07 N. Previous work by one of us has shown that antimony and tin are collected over a wide range of acid concen- trations. Similar results have been obtained by Babko and Shtoka10.~ Several analytical applications have been reported, for example, the determination of antimony with Rhodamine B by MacNulty and Woollard,4 antimony in copper by Park and Lewis,s antimony and bismuth in anode copper by Yamazaki,B arsenic and tin in lead by Luke7 and antimony in cast iron by Rooney.8 No previous comprehensive study has, however, been made of the elements to which the procedure can be applied, or the con- centration range or conditions in which it is efficient for each species.The present work was undertaken to provide a comprehensive study, as mentioned above. Previous papers in the series have established conditions and concentration ranges for the “collection” of antimony, bismuth and tin,g and for tin and 1ead.lO In the present work, studies of copper and zinc have been made and further information on the behaviour and determination of lead and tin is presented. EXPERIMENTAL The “collection” procedure previously describedg Treat the sample solution, in a 400-ml beaker, with sufficient nitric acid solution to make the volume up to 200m1, add 5ml of 5 per cent. manganese(I1) sulphate solution and heat to boiling, then add 2-5 ml of 1.25 per cent. potassium permanganate solution, dropwise, and continue boiling for a further 2 minutes.Allow the solution to stand a t a was used in all experiments. * For details of Parts I and I1 of this series, see reference list, p. 382. t Present address: Shandon Scientific Co. Ltd., 65 Pound Lane, Willesden, London, N.W.10. 0 SAC and the authors376 PYBURN AND REYNOLDS: STUDIES OF THE SEPARATION OF TRACE METALS [Analyst, Vol. 93 temperature of 70” C for about 30 minutes and then filter off the manganese dioxide pre- cipitate with an 11-cm, Whatman No. 40 filter-paper. Wash the precipitate three times with 10-ml portions of the nitric acid solution used for the initial dilution; the last traces of precipitate are transferred from the beaker to the filter-paper by the first two washings.Two methods for the dissolution of the manganese dioxide have previously been des- cribed10; that in which a mixture of nitric acid and hydrogen peroxide is used was adopted for this work, as follows. Dissolve the manganese dioxide precipitate completely by allowing about 1 O m l of a hot mixture of 1.2 M nitric acid and 100-volume hydrogen peroxide (20 + 1) to run down the sides of the filter-paper. Wash the filter-paper with a further 15ml of the nitric acid- hydrogen peroxide solution and then with a small volume of water. Combine the solution and washings in a 400-ml beaker, then dilute to about 100 ml and boil gently to remove the bulk of the residual hydrogen peroxide and allow to cool. Bubble sulphur dioxide gas slowly through the solution to convert the last traces of hydrogen peroxide into sulphuric acid.Boil the solution to remove excess of sulphur dioxide, cool, transfer to a 250-ml calibrated flask and make up to volume with water. All determinations reported in this paper were made by d.c. or a.c. polarography. All polarography was carried out on a Cambridge pen-recording polarograph, in conjunction with which a Cambridge univector unit was used to provide a.c. facilities. All potentials quoted are on the European sign convention. According to this convention, the potential of the saturated calomel electrode is taken as +0.246 volt versas the normal hydrogen electrode. “COLLECTION” OF COPPER- A series of solutions, each containing 20 mg of copper and with initial acid concentrations within the range 1.2 to 0408 M, was prepared and subjected to the “collection” procedure described above.The manganese dioxide was re-dissolved with nitric acid and hydrogen peroxide, and made up to 250 ml in calibrated flasks, as already described. Twenty-five millilitre portions of each solution were placed in 50-ml calibrated flasks and made up to volume with 0.2 M sodium potassium tartrate. Aliquots were placed in polaro- graphic cells, de-oxygenated by passage of nitrogen gas for 5 minutes and subjected to d.c. and ax. polarography. Well shaped steps and sharp ax. peaks were obtained, with half- wave potentials -0.24 volt versus mercury pool. Concentrations of copper recovered were calculated by reference to a calibration graph. 8 Applied potential (volts versus mercury pool) Fig.1. Polarography of copper in 0.1 M sodium potassium tartrate fter collection on manganese dioxide: (a) d.c. polarographic step: (b) ax. polarographic wave A typical copper step and an a.c. wave are shown in Fig. 1. Copper recoveries are presented in Table I. The results in Table I show that the most efficient collection took place in a solution 0*008 M in acid. Further experiments were, therefore, made at this acid concentration, to study the efficiency of “collection” as a function of initial copper concentration. A series of solutions was prepared as before, but with an initial nitric acid concentration of 0.008 M andJune, 19681 BY THE MANGANESE DIOXIDE “COLLECTION” METHOD. PART 111 TABLE I EFFECT OF NITRIC ACID CONCENTRATION ON THE RECOVERY OF COPPER 377 Acid concentration, 1.2 1-2 0.12 0.12 0.06 0-06 0.03 0-03 0.015 0.015 0.008 0.008 0.004 0.004 0.002 0.002 M Copper * Added, Found, mg mg 20 0-088 20 0.080 20 0,088 20 0.088 20 0.095 20 0-096 20 0.080 20 0.10 20 0.075 20 0.075 20 0.60 20 0.48 20 0.18 20 0.18 20 0.18 20 0.076 Recovery, per cent.0-43 0.40 0.43 0.43 0.48 0.48 0.40 0.50 0-38 0-38 2.50 2.38 0.95 0.95 0.88 0.38 with various copper concentrations covering the range 5 to 100mg. These solutions were treated and polarographed exactly as already described. Copper recoveries are given in Table 11. TABLE I1 RECOVERY OF COPPER Copper & Recovery, Added, mg Found, mg per cent. 6 0.055 1.1 5 0.055 1.1 10 0.08 0.8 10 0.125 1.25 15 0.125 0.83 15 0.10 0.67 20 0.50 2.5 20 0.475 2.38 25 0-325 1.3 25 0.14 0.56 50 0.22 0.44 50 0.206 0.4 1 100 0-225 0.226 100 0.17 0.17 “COLLECTION” OF ZINC- A series of solutions, each containing 20 mg of zinc and with initial acid concentrations covering the range 1.2 to 0-008 M, was prepared and treated exactly as described for copper.It was found that polarography of these solutions yielded highly variable results because of the appearance of a step with a half-wave potential displaced from that expected for zinc. These experiments were repeated with various base electrolytes suitable for the determination of zinc, with similar results. Blank determinations were, therefore, made in the absence of zinc, and from the results obtained it was deduced that the interfering step was caused by residual manganese ion. Consideration was, therefore, given to the use of a base electrolyte that would remove the manganese interference by complex formation, which has no effect on the zinc half-wave potential.This was achieved by the use of M potassium hydroxide and 0.2 M mannitol.11 This completely removed manganese interference in the region of the zinc half-wave potential, which was -1.2 volts versus mercury pool. The results obtained with a fresh series of solutions, under these conditions, showed that only traces of zinc were “collected” over the whole range of acid concentration studied. The largest trace, however, occurred with solutions of initial acid concentration 1-2 M.378 PYBURN AND REYNOLDS: STUDIES OF THE SEPARATION OF TRACE METALS [Analyst, Vol. 93 Further experiments were carried out with a series of solutions with zinc concentrations in the range of 2 to 1OOmg in 1.2 M acid.No measurable amount of zinc was recovered, even at the highest zinc concentration studied. “COLLECTION” OF LEAD AND TIN- Previous workg has shown that tin is quantitatively “collected” from solutions of acid concentration 1.2 M, but that the efficiency of collection fell to 76 per cent. when the acid concentration was reduced to 0.008 M . ~ O It has also been shownlO that lead is not “collected” from solutions 1-2 M in acid, but that good recoveries are obtained from 0.008 M acid. In view of the low recoveries of copper and zinc reported above, it appeared that procedures could be developed for the determination of tin and lead in copper-base alloys. It was evident that straightforward procedures were feasible for the determination of tin in the presence, or absence, of lead, and for the determination of lead in the absence of tin’ DETERMINATION OF TIN- “Collection” was carried out, as already described, from a copper - zinc solution containing 1.2 M nitric acid.In view of previous results,1° dissolution of the precipitate with a mixture of nitric acid and hydrogen peroxide was not used, as this has been shown to cause oxidation of the tin to tin(1V) oxide. Instead, the filter-paper and precipitate were placed in a beaker and treated with 3 ml of 18 M sulphuric acid and 5 ml of 16 M nitric acid. The filter-paper was then destroyed by repeated evaporation, further amounts of nitric acid being added as necessary. After oxidation of all carbon to give a clear solution, careful evaporation, almost to dryness, was carried out.The residue was allowed to cool, 15ml of 36 per cent. hydrochloric acid were added and the solution diluted to about 50 ml. After heating for a few minutes, the residue dissolved to give a clear solution, which was cooled and transferred to a 250-ml calibrated flask. The flask was made up to volume with washings from the beaker. Fifty millilitres of the solution were transferred to a 100-ml calibrated flask, treated with 10 ml of 36 per cent. hydrochloric acid, 10 ml of N potassium chloride and 10 ml of 5 N sodium hydroxide before being finally made up to volume with water. An aliquot was transferred to a polarographic cell, de-oxygenated by passage of nitrogen for 5 minutes and polarographed for tin.The tin concentration was calculated by reference to a calibration graph. A series of standard copper - zinc solutions containing tin and lead were treated as Tin steps of good shape were obtained, the step heights of which were unaffected by described above. the presence of lead in the original solution. Recoveries are presented in Table 111. TABLE I11 RECOVERY OF TIN FROM SOLUTIONS CONTAINING 0.6 g OF COPPER AND 0.4 g OF ZINC Lead Added, mg 0 0 5 5 5 10 10 10 Tin 1 Added, mg Found, mg 2.6 2.4 2.5 ,2* 5 6.0 4-95 6-0 5-0 10.0 10.2 5.0 4.9 5-0 4.9 10.0 10.0 DETERMINATION OF LEAD (TIN ABSENT)- The “collection” and dissolution procedure used was exactly the same as that used for the studies of copper and zinc. The final solution, after dissolution of the precipitate with the nitric acid - hydrogen peroxide mixture, was made up to 250 ml in a calibrated flask.Fifty millilitres of this solution were transferred to a 100-ml calibrated flask, treated with 10ml of N potassium chloride and made up to volume with water.June, 19681 BY THE MANGANESE DIOXIDE “COLLECTION” METHOD, PART 111 379 An aliquot of the solution was transferred to a polarographic cell, de-oxygenated by passage of nitrogen for 5 minutes and polarographed for lead. The lead concentration was calculated by reference to a calibration graph. A series of copper - zinc solutions, containing lead, was treated by the above procedure. Good steps were obtained with half-wave potentials of about -0.5 volt versus mercury pool.Recoveries are presented in Table IV. TABLE IV RECOVERY OF LEAD FROM SOLUTIONS CONTAINING 0.6 g OF COPPER AND 0-4 g OF ZINC Lead P Added, m g Found, mg 2 2.04 2 2.00 5 4-85 5 4.90 10 9.90 10 10-10 DETERMINATION OF LEAD AND TIN- In order to attempt to develop a method for the simultaneous determination of lead and tin, with a single “collection,” a re-investigation of the behaviour of these two elements in 0.008 M acid was carried out. This arose as the result of observations that a white pre- cipitate formed on boiling tin solutions at this acidity, in the presence of manganese sulphate, before the addition of potassium permanganate. The formation of this precipitate was confirmed and it was removed and examined polarographically. It was found to contain about 77 per cent.of the added tin. It was evident from the above results that the “collection” of tin from 0 - 0 0 8 ~ acid, reported by Reynolds and Tyler,lo was not a true manganese dioxide collection, as the bulk of the tin was precipitated before the formation of the collector. Methods were, therefore, sought to retain the tin in solution at this acid concentration, Many complexing reagents were tried without success, but it was finally found that the presence of ethanol prevented tin precipitation. Furthermore, the ethanol did not react with the potassium permanganate and did not interfere with the “collection” process. The addition of 50 ml of ethanol to the initial solution was shown to be sufficient, and tin recoveries from these 0.008 M acidic solutions were almost 100 per cent.In the presence of ethanol, the nitric acid - hydrogen peroxide method for dissolution of the precipitate was satisfactory. A common procedure for the “collection” and dissolution of lead and tin could, therefore, be used. The procedure for simultaneous polarographic determination of lead and tin in copper- base alloys was based on a method described by Lingane.12 In this method, polarography is performed in acidic and in alkaline solutions. In the former, both lead and tin are reduced at similar half-wave potentials to yield a combined step, whereas in the latter, a lead step only is obtained. A series of “collections” from lead - tin solutions was made, as already described, but with the addition of 50ml of ethanol to the initial solution, after adjustment of the acid concentration to 0*008 M.After dissolving the precipitate and making up the volume in a 250-ml calibrated flask, two 50-ml aliquots were withdrawn and placed in 100-ml calibrated flasks. One of these was treated with 10 ml of 11 M hydrochloric acid, 10 ml of M potassium chloride and 10 ml of 5 M potassium hydroxide. The solution was then made up to volume with water. The other aliquot was treated with 10 ml of 5 M potassium hydroxide and made up to volume with water. Polarography of solutions prepared as above yielded combined lead - tin steps from the acidic aliquot and steps for lead alone from the alkaline solution. The latter, however, suffered interference from reduction of the residual manganese present. This series was, therefore, repeated with addition of 1 g of mannitol to each lead determination before making up to volume (see under “Collection” of zincll) .This eliminated the interference.380 PYBURN AND REYNOLDS: STUDIES OF THE SEPARATION OF TRACE METALS [AfidySt, VOl. 93 Initial recoveries were promising, but an unacceptably high percentage of erratic results occurred. This difficulty was overcome by increasing the amount of potassium permanganate added in the “collection” procedure, from the usual 2-5 ml of 1-25 per cent. solution to 5 ml of 1-25 per cent. solution. It was found, however, that this increase in permanganate con- centration caused manganese interference with the lead step in alkaline solution to recur and it was necessary to increase the amount of mannitol added to 3 g.Recoveries of lead and tin obtained from the series of solutions treated by the final procedure are given in Table V. TABLE V RECOVERIES OF LEAD AND TIN Lead & Added, mg Found, m g 5 4.95 5 4.65 5 4.70 5 4.96 5 4.50 10 9.95 10 9.80 Tin Added, mg 0 2.6 3.0 7.5 10.0 7.5 10.0 Found,. mg 0 2.6 2.86 7.73 10.0 7.80 10.20 In view of the satisfactory nature of the above results, a method was described for the determination of lead and tin in copper-base alloys. METHOD Place 1 g of the copper-base alloy sample (as drillings or turnings) in a 400-ml beaker and treat it with the minimum volume of 2 N nitric acid. Warm gently to complete dissolution, neutralise to litmus and cool. Add 200 ml of 0.008 M nitric acid, 5 ml of 5 per cent. man- ganese(I1) sulphate solution and 50 ml of ethanol.Heat to boiling, add 5 ml of 1.25 per cent. potassium permanganate solution, dropwise, and continue boiling for a further 2 minutes. Allow the solution to stand at 70” C for 30 minutes and then filter off the manganese dioxide precipitate through an 1 l-cm, Whatman No. 40 filter-paper. Wash the precipitate three times with 0408 M nitric acid solution, the first two washings being used to transfer the last traces of precipitate from the beaker to the filter-paper. Dissolve the manganese dioxide precipitate completely by allowing about 10ml of a hot mixture of 1.2 N nitric acid and 100-volume hydrogen peroxide (20 + 1) to run down the sides of the filter-paper. Wash the filter-paper with a further 15 ml of the nitric acid - hydrogen peroxide solution and then with a small volume of water.Combine the solution and washings in a 400-ml beaker. Boil gently to remove the bulk of the residual hydrogen peroxide. Allow to cool and bubble sulphur dioxide gas slowly through the solution to convert the last traces of hydrogen peroxide into sulphuric acid. Cool, transfer it into a 250-ml calibrated flask and make up to volume with water. DETERMINATION OF LEAD plus TIN- Place 50 ml of the prepared solution in a 100-ml calibrated flask, add 10 ml of 11 N hydrochloric acid, 10 ml of M potassium chloride and 10 ml of 5 M potassium hydroxide, then make up to volume with water. Transfer an aliquot to a polarographic cell, de-oxygenate by passage of nitrogen gas for 5 minutes and record a polarogram, versus the mercury pool as anode.Measure the height of the combined lead - tin step, which occurs at a half-wave potential of about -0-5 volt versus mercury pool. DETERMINATION OF LEAD- Place 50 ml of the prepared solution in a 100-ml calibrated flask and add 10 ml of 5 M potassium hydroxide and 3 g of mannitol. Transfer an aliquot to a polarographic cell, de-oxygenate by passage of nitrogen gas for 5 minutes and record a polarogram versus the mercury pool as anode. Measure the height Make up to volume with water.June, 19681 381 of the lead step that occurs at a half-wave potential of about -0.75 volt versus mercury pool. Calculate the lead content of the sample from a calibration graph prepared with standard solutions, or alloys, subjected to the full procedure.DETERMINATION OF TIN- Subtract the height of the lead step from the height of the combined lead - tin step. Calculate the tin content of the sample from a calibration graph prepared with lead-free standard solutions, or alloys, subjected to the full procedure. BY THE MANGANESE DIOXIDE “COLLECTION” METHOD. PART 111 TRIAL OF THE METHOD A series of solutions was prepared, each containing 0.6 g of copper and 0.4 g of zinc, and treated with varying amounts of standard lead and tin solutions. These solutions were subjected to the full procedure, as described above. The results obtained are given in Table VI. TABLE VI RECOVERY OF LEAD AND TIN FROM SOLUTIONS CONTAINING 0.6g OF COPPER AND 0-4g OF ZINC Lead & Added, mg Found, mg A 6.6 6.3 A 6.6 6.9 B 6-0 4.7 B 6.0 4.8 C 7.6 7.46 C 7.6 7.4 D 4.0 4.0 D 4.0 3.9 Tin t-,--7 Added, mg Found, mg 5.3 6-26 6.3 6.3 10.3 10.3 10.3 10.3 6.3 5.3 6.3 6.2 9.4 9.4 9.4 9.35 In view of the satisfactory results above, several determinations were made with samples of special copper-base alloys, the tin and lead contents of which had been determined gravi- metrically. The results of these analyses are presented in Table VII.TABLE VII RECOVERY OF LEAD AND TIN FROM COPPER-BASE ALLOYS Tin & 7 Present, Found, Present, Alloy per cent. per cent. per cent. Special alloy HC (1) 0- 6 0.62 0-26 (2) 0.6 0.6 1 0.26 Special alloy CA* (1) 4.45 4-40 2.2 (2) 4-45 4.35 2.2 Special alloy TU (1) 0.36 0.33 1.0 (2) 0.35 0.35 1.0 Special alloy TEN* (1) 2-05 2-00 3.5 (2) 2.06 2.00 3.5 0-6 g of sample used, instead of 1 g as specified.Lead - Found, per cent. 0.24 0.24 2-05 2.10 1.06 0-96 3.40 3.46 DISCUSSION The results presented in Tables I and I1 show that copper cannot be efficiently collected on manganese dioxide. It has also been shown that collection of zinc is negligible under all conditions studied. The optimum acid concentration for “collection” of copper was found to be 0.008 M, but even under these conditions the recovery of copper was only 2-5 per cent. of 20mg added. In general, an increase in copper concentration tended to increase the amount of copper recovered, but the percentage efficiency of the “collection” decreased sharply.382 PYBURN AND REYNOLDS Mannitol in alkaline solution is a valuable complexing reagent for manganese, as it forms a species that yields an anodic wave in the region of -0.2 volt versus mercury pool.Inter- ference by manganese in the region of 1 volt is therefore eliminated.ll This species also provides a valuable means of determining manganese in the presence of other species whose polarographic step would normally precede that of manganese. The inefficient “collection” of copper and zinc has been turned to advantage by the development of procedures for the determination of lead and tin in copper-base alloys. Tables I11 and IV show that good results are obtained for tin in the presence, or absence, of lead and for lead in “tin-free” alloys. The investigation of the tin precipitation, in the course of developing the method for the simultaneous determination of lead and tin, has provided interesting new information.It is evident that a species must be in true solution for effective “collection” to take place. Otherwise, presumably, the particle size is too large for efficient adsorption to occur on the manganese dioxide surface. Further fundamental studies are desirable. I t would appear, also, that “collection” in the form of simple ion species is desirable, as complexing reagents did not enhance the efficiency of collection, although they assisted in retaining the tin in solution. The mode of action of the ethanol is not fully understood. The method for the determination of lead and tin in copper-base alloys is shown to be reproducible within the range studied (Table VI) and to be applicable to alloys with a variety of tin and lead contents.In this method, a direct correction for the contribution of lead to the combined lead - tin step obtained in acidic solution is made by reference to the height of the lead step obtained in alkaline solution. This is not completely valid, as the diffusion coefficients, and hence the diffusion-current constants, will not be the same in the two media. However, Lingane has shown13 that the ratio of diffusion-current constants in acidic and alkaline solution is 1 : 1.036, so that the above assumption may be made without serious error. The accuracy of the results will, of course, be enhanced if the appropriate correction is made. The chloride content of the base electrolyte used for the determination of tin (1-2 N) is less than that generally recommended. A concentration of 5 to 6 N is more usual. This concentration was chosen to avoid possible loss of lead as chloride when both elements are present together. The results (Tables 111, V and VI) indicate that the present chloride concentration is adequate to give the chloro-complex of tin(1V) that is considered desirable for satisfactory polarography. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. REFERENCES Blumenthal, H., 2 . analyf. Chem., 1928, 74, 33. Kallman, S., and Prestira, F., Ind. Engng Chem. Analyt. Edn, 1941, 13, 8. Babko, A. F., and Shtokalo, M. I., Zav. Lab., 1955, 21, 767. MacNulty, B. J., and Woollard, L. D., Analytica Chim. Acta, 1955, 13, 64. Park, B., and Lewis, E. J., Ind. Engng Chem. Analyf. Edn, 1933, 5 , 182. Yamazaki, Y., Bunseki Kagaku, 1957, 82, 619. Luke, C. L., Ind. Engng Chem. Analyt. Edn, 1943, 15, 626. Rooney, R. C., Analyst, 1957, 82, 619. Ogden, D., and Reynolds] G. F., Ibid., 1964, 89, 538. Reynolds, G. F., and Tyler, F. S., Ibid., 1964, 89, 579. Reynolds, G. F., and Shalgosky, H. I., Analyfica Chim. Acta, 1954, 10, 273. Lingane, J. J., Ind. Engng Chem. Analyt. Edn, 1946, 18, 429. -, “Electroanalytical Chemistry,” Interscience Publishers Inc., New York and London, 1953, p. 318. NOTE-References 9 and 10 are to Parts I and I1 of this series, respectively. Received November 17th. 1967