1030 Analyst, November, 1979, Vol. 104,pp. 1030-1036 Extraction of Nanogram Amounts of Cadmium and Other Metals from Aqueous Solution Using Hexamethyleneammonium Hexamethylenedithio- carbamate as the Chelating Agent A. Dornemann and H. Kleist Bayer AG, Werk Uerdingen, A C-F Untersuchungslaboratorium, D 4150 Krefeld 11, FHG A group of metals can be extracted from aqueous solution by using hexa- methyleneammonium hexamethylenedithiocarbamate as the chelating agent and a mixture of 2,4-dimethylpentan-3-one and xylene as the organic phase. A description is given of a procedure for the determination of microgram and nanogram amounts of the nine metals silver, bismuth, cadmium, copper, nickel, lead, thallium and zinc in aqueous solution. The influence of a high content of iron and copper on the extraction is described.Keywords : A tomic-absorption spectrometry ; metal extraction ; nanogram avnozGnts of metals ; dithiocarbarnate chelation Liquid - liquid extraction is the preferred process for the enrichment of trace amounts of heavy metals from aqueous solution for subsequent determination by means of atomic- absorption spectrometry. In view of the high selectivity of atomic absorption the main emphasis is on group extraction processes. The organic extract should be suitable for both electrothermal atomisation and atomisation in a flame. With an extract of this kind, in which the individual concentrations of metals generally vary a great deal, it is possible to take the better suited atomisation technique for the determination of the particular met a1 under consideration.Firstly, a group of nine metals, silver, bismuth, cadmium, cobalt, copper, nickel, lead, thallium and zinc, were to be extracted simultaneously, and secondly, cadmium contents in the nanogram per gram range were to be determined in the presence of a large excess of iron. It was found that the extraction processes described in the literature as a preliminary stage to atomic-absorption spectrometry1,2 were not suitable for the two purposes referred to above. The most widely used method involves the use of ammonium tetramethylene dithiocarbamate (also known as ammonium pyrrolidine dithiocarbamate, APDC) as the complexing agent, and 4-methylpentan-2-one (methyl isobutyl ketone, MIBK) as the organic phase. Although MIBK has good combustion properties in the flame, it can lead to problems in electrothermal atomisation. A further disadvantage of MIBK is its high solubility in water.Moreover, the cadmium - APDC complex is insufficiently stable in #an acidic medium. This paper describes some investigations that led to the use of hexamethyleneammonium hexamethylenedithiocarbamate (HMA IIMDC) as the chelating agent and of a mixture of xylene and 2,4-dimethylpentan-3-one (diisopropyl ketone, DIPK) as the organic phase. Further experiments clarified the behaviour of the recommended extraction system towards an excess of iron and copper. This study of the extractability of heavy metals had a dual purpose. Experimental Reagents distilled before use. The inorganic reagents used were of the highest available purity; organic solvents were 4-MethyJfientan-2-one (methyl isobutyl ketone, MIBK) . This compound boils at 117-1 18 "C.2,4-Dimethylpentan-3-one (diiso@ro$yl ketone, DIPK) . This compound boils at 124-125 "C. Xylene. Hexamethyleneammonium hexamethylenedithiocarbamate (HMA HMDC) . To a solution This solvent is a mixture of the various isomers.DORNEMANN AND KLEIST 1031 of 224 ml of distilled hexamethyleneimine (boiling-point 136-138 "C) in 300 ml of xylene, which is being cooled in an ice-bath, add, within 30 min and with constant stirring and cooling, 60 ml of distilled carbon disulphide (boiling-point 46.2 "C). Collect the white crystalline precipitate on a funnel, wash it three times with diethyl ether and then dry it between filter-papers. Caution-Hexamethyleneimine is a severe poison and appropriate precautions should be taken.The preparation of each special reagent used in the procedure is described in the appro- priate section. Apparatus burner for the air - acetylene mixture and a deuterium background compensator. Perkin-Elmer atomic-absorption spectrometer. Perkin-Elmer graphite furnace, Model HGA500. The Model 420 is used, with a three-slot Results A series of experiments was performed to find the best conditions for the pre-concentration procedure, investigating the chelating agent, the organic solvent and organic metal standard solutions, which should enable a check of the completeness of extraction. The influence of higher concentrations of total extractable metals was also investigated. The results of these experiments are reported below.Chelating Agent The chelating agent should permit complete extraction over the widest possible ranges of pH and concentration. Attempts to extract cadmium at the 100 pg 1-1 level, and at a pH of 2.8, with APDC - MIBK gave poor results. The conditions of the extraction are: sample volume, 100 ml, containing 10 pg of cadmium; addition of 2.5 ml of APDC solution (1 g of APDC in 100 ml of water); pH adjusted to 2.8 after the addition of APDC; addition of 10 ml of MIBK before extraction. As the mixing time of the organic and aqueous phases was increased, the extraction yield diminished.3 After shaking for 1 min, about 75% of the cadmium was recovered; after shaking for 10 min, only about 8% was recovered. Complete extraction is possible only for concentrations of cadmium below 10 pg l-1.4 The rate of decomposition of the initially formed cadmium - APDC chelate is influenced by the pH, the total salt content of the aqueous phase, the material of the vessel and the exposure to light.Complete extraction of cadmium at the 100 pg 1-1 level is possible at pH 4, but these higher pH values restrict the possibilities of selective isolation of cadmium. Another derivative of dithiocarbamic acid, hexamethyleneammonium hexamethylene- dithiocarbamate (HMA HMDC),"s proved to be more suitable for the extraction of cadmium from an acidic medium. Extraction at pH 2, including shaking for up to lOmin, resulted in a complete cadmium yield. Twenty micrograms of cadmium can be completely extracted from 100 ml of an aqueous solution of sodium chloride (10 g 1-1) with 5 mg of HMA HMDC and 10ml of MIBK in the pH range 1-10, the pH in the aqueous phase being measured after the addition of the HMA HMDC solution.Extraction from acidic solution is of particular importance. Organic Phase Volume stability In order to be able to check the process of enrichment, a clearly defined enrichment factor is desirable. This requires that the added volume of the organic phase remains stable during the course of the extraction. In this respect MIBK as the organic phase has a number of dis- advantages. Its solubility in water at room temperature is about 2% V / V , which is relatively high.9 When 500 ml of water were shaken with 10 ml of MIBK, the organic phase disappeared almost completely.The solubility is greatly dependent on the total salt content of the aqueous solution (Table I). As a result, additional variables are introduced into the analytical procedure, and this must be taken into consideration when the enrichment factor exceeds 5. 4-MethylPentan-2-one.1032 DORNEMANN AND KLEIST : DITHIOCARBAMATE EXTRACTION OF NANOGRAM Analyst, VoZ. 104 TABLE I VOLUME CHANGE OF ORGANIC PHASE Aqueous phase 100m1, pH 7 . . . . .. 100 ml HC1, pH 2 .. .. 600 ml H,O, pH 7 . . . . 100 ml H,O, pH 7 . . .. 100 ml NaCl solution, 4 g NaCl . . 100 ml NaCl solution, 10 g NaCl 100 ml HCl, pH 2 100 ml NaCl solution, 4 g 'NaCl ' 100 ml NaCl solution, 10 g NaCl 100 ml H,O, pH 7 . . .. 600 ml H,O, pH 7 . . .. Organic solvent taken MIBK MIBK MIBK MIBK MIBK DIPK DIPK DIPK DIPK 70% DIPK- 30% xylene 30% xylene 70% DIPK - Volume of organic phase added/ml 10 10 10 10 10 10 10 10 10 10 10 Volume of organic phase after equilibration/ ml 8.0 8.0 8.6 8.8 t0.5 9.2 9.2 9.4 9.6 >9.8 >9 Decrease in volume of organic phase, % 20 20 14 12 > 95 8 8 6 4 <2 < 10 2,4-DimethyZpentan-3-one.In comparison with MIBK, DIPK represents a considerable improvement. Its solubility in water at room temperature is only 0.5% V/V.99 Vari- ations in the total salt content of the solution have only a slight effect on solubility (Table I). A further improvement in volume stability can be achieved with a mixture of 70% V/V of DIPK and 30% V/V of xylene. When 100ml of water were shaken with 10ml of this solvent mixture no change in the volume of the organic phase could be detected with the usual measuring vessels used for analytical work.When 500ml of water were shaken with 10ml of this mixture the reduction in volume of the organic phase was less than 1 ml, this volume decrease being partially caused by the fact that small droplets of the organic phase adhered to the walls of the vessel, or remained suspended in the aqueous phase. Thus, the solvent mixture DIPK - xylene permits enrichment factors of up to 50; no additional steps are necessary, and data on the salt content of the aqueous phase are not required. Mixture of 70% 2,4-dimethylpentany3-one and 30% xylene (V/V). Sensitivity and behaviouur during atomisation Flame atomisation of a given concentration of metal in the three organic phases investi- gated resulted in the same absorbance.This absorbance is about 3.5 times greater than the absorbance of aqueous metal solutions with the same nebuliser - burner assembly. In electrothermal atomisation, the solvent mixture DIPK - xylene evaporates, leaving no residue. MIBK extracts, however, tend towards resinification and coke formation upon heating; this effect impairs the reproducibility of the atomisation process, particularly for elements, such as cadmium, which have low atomisation temperatures. Metal Standard Solutions in the Solvent Mixture Pureearation Standard solutions of metals made up in the extraction solvent mixture are necessary in order to check the yield of extract from synthetic solutions and to establish calibration graphs. Oil-soluble standards are available commercially for only some of the metals covered by the group-extraction procedure.ll Tests carried out with a commercial oil- soluble standard for cadmium (cyclohexanylbutyric acid cadmium salt) were unsatisfactory. The following process proved to be suitable for use with all metals covered by the group extraction procedure.The starting solution used is a standard solution in nitric acid of the metal, to which some citric acid and about a ten-fold volume of formic acid is added. To this mixture is added about the same volume of DIPK. The result is a clear, homo- geneous solution. Part of this solution is then diluted by a factor of about 1000 with the extraction solvent mixture containing HMA HMDC. In this way organic standard solu- tions are obtained, the compositions of which largely correspond to those of the solutionsNovember, 1979 AMOUNTS OF CADMIUM AND OTHER METALS FROM AQUEOUS SOLUTION obtained from the extraction procedure.Working procedure. 1033 Details of their preparation are given under Stability If kept in a laboratory without protection against the light, organic metal standard solutions will remain stable for at least 1 d. If kept in a refrigerator at 278 K, they are stable for several weeks. The metal-containing extracts obtained by means of the solvent mixture DIPK - xylene exhibit the same stability as the organic metal standard solutions when the extracts are pipetted off from the aqueous layer after phase separation. Influence of Extractable Metals Inplzence of iron In addition to the metals mentioned in the group extraction process, the following metals can also be wholly or partially extracted into the organic solvent mixture: arsenic, iron, gallium, germanium, indium, manganese, antimony and tin.The alkali metals, the alkaline-earth metals, aluminium, selenium, yttrium, the lanthanides, titanium, zirconium and hafnium remain in the aqueous phase. When examining samples with greater contents of extractable metals, an important question is up to what total content of extractable species can this method be applied? Details of the influence of iron on the group extraction are discussed elsewhere.12 The iron content of the aqueous solution can be as high as 25 mg 1-1 without having any detri- mental effect on the group extraction of the nine elements silver, bismuth, cadmium, cobalt, copper, nickel, lead, thallium and zinc.Interferences caused by a greater content of iron (up to 250mg1-l) can be excluded by adding ammonium fluoride (5 g1-l) or by pre- extraction of the iron from 2 N hydrochloric acid solution with N-nitroso-N-phenylhydroxyl- amine. I@uence of copper Copper was investigated as an example of a metal that forms particularly stable com- pounds with dithiocarbamates.13-15 Table I1 shows the results of experiments on the extraction of the elements covered in the group extraction process with aqueous solutions of increasing copper content. The extraction carried out under the conditions set out ,in the working procedure was repeated up to four times in order that the recovery of the metals should be as complete as possible, even at higher copper concentrations.Copper contents of up to 25 mg 1-1 had only a slightly detrimental effect on the extraction of the other metals. With copper contents of 50mg1-1 the first extract contained, in addition to part of the original copper, only some of the silver, cobalt and nickel used (see Table 11); the other metals were found in the subsequent extracts. This result is somewhat surprising in view of the fact that, for the stoicheiometric conversion of the metals with the chelating agent, the capacity of the extraction solution was not exhausted in the first extrac- tion. When the starting solution had a higher copper content, a constant amount of metal (about 40mg) was extracted per extraction step from the aqueous starting solution.The highest copper concentration of 150mg1-1 led to the deposition of a greater amount of precipitate, which adhered to the wall of the vessel and was not completely dissolved by the subsequent extractions. For the purposes of practical analysis, it follows from these experiments that the total concentration of metals to be extracted must not exceed 25 mg 1-1 if the metals are to be completely recovered in a single extraction. Procedure for the Determination of Nine Metals in Water7 Method The weakly acidic aqueous solution (acid concentration 0.5 M) is brought into a pH range of between 2 and 3 by adding a formate buffer solution. After chelating with HMA HMDC, the elements to be determined are extracted with a DIPK - xylene mixture.The metal content of the extract is measured by atomic-absorption spectrometry with an1034 DORNEMANN AND KLEIST : DITHIOCARBAMATE EXTRACTION OF NANOGRAM Analyst, vo,!. 104 TABLE I1 INFLUENCE OF COPPER ON THE EXTRACTION Metal content of initial aqueous solution, 25 pg 1-1 of each metal; each extraction performed according to the standard procedure. Copper concentration added/mg 1-1 0 5 12.5 25 37.5 60 100 150 Extraction run 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 4 Recovery of heavy metals, yo I A 8 Ag Bi >95 >95 >95 >95 - - - - - - - - >95 >96 - - - - > 96 91 6 88 86 11 9 63 29 95 6 44 - 47 79 7 11 23 62 - 12 25 55 - - - - - - - - - co > 95 - - > 96 - - > 95 - - >96 >95 - - - - 61 35 30 61 5 20 20 20 31 - Ni > 95 - - > 95 - - > 95 - A > 95 > 95 - - - - 45 50 29 47 16 10 15 41 - - cu - - - > 96 1 > 96 1 >95 1 > 95 1 76 12 1 39 38 1 26 26 26 1 - - - - air - acetylene flame or an electrothermal atomiser.Metal standard solutions, prepared from an aqueous stock solution, in the DIPK - xylene extraction mixture make it possible to check the completeness of extraction from aqueous standard solutions. Reagents This solution contains 1000 mg 1-1 of silver, bismuth, cadmium, cobalt, copper, nickel, lead, thallium and zinc in water. Dissolve 1000 mg of each of the metals to be determined in nitric acid, sp. gr. 1.41, using gentle heat in order to effect com- plete dissolution. Transfer the solution into a 1-1 calibrated flask, and dilute to the mark with water.The final acid concentration of stock solution I should be about 0.1-0.5 M. Metal nitrates can also be used to prepare this stock solution, provided that their metal content is established by an independent means. If stock solution I contains silver, it must not contain chloride ions; solutions containing lead must be free from chloride and sulphate ions. Prepare aqueous standard solutions by dilution of aqueous stock solution I with 0.1 M nitric acid. This solution contains 50 mg 1-1 of silver, bismuth, cadmium, cobalt, copper, nickel, lead, thallium and zinc in organic solvent. Pipette 5 ml of aqueous stock solution I into a dry, 100-ml calibrated flask. Add 50 ml of formic acid and 0.25 g of citric acid monohydrate. Prepare organic metal standard solutions by dilution of organic stock solution I1 with the extraction solution.The use of dry, 25-ml calibrated flasks and plastic disposable pipettes (0.1-1.0 ml) is recommended. Organic metal standard solutions should be kept in a cool, dark place. Formate bufler solution. Dissolve 268 g of formic acid and 14 g of citric acid monohydrate in about 350 ml of water. Add slowly, with constant cooling and stirring, 243 g of sodium hydroxide. To this mixture add 50 mg of m-cresol purple and dilute the solution to 1 1 with water. Wash this solution twice with 50 ml of extraction solution in order to remove trace amounts of extractable metals. Aqueous stock solution I . Aqueous metal standard solutions. Organic stock solution I I . Adjust to the mark with 2,4-dimethylpentan-3-one.Organic metal standard solutions.November, 1979 AMOUNTS OF CADMIUM AND OTHER METALS FROM AQUEOUS SOLUTION 1.7 g of HMA HMDC in 75 ml of xylene, heating gently if necessary. to the mark with 2,4-dimethylpentan-3-one and keep it in a cool, dark place. HMA HMDC solution, 0.2 M in methanol. of HMA HMDC in methanol, heating gently if necessary. ture and adjust it to the mark with methanol. 1035 Dissolve, in a dry, 250-ml calibrated flask, Adjust the solution In a dry, 100-ml calibrated flask, dissolve 5.5 g Cool the solution to room tempera- Extraction solution, 0.025 M HMA HMDC. VeriJication of complete extraction Before analysing the actual samples, prepare at least three aqueous metal standard solutions, the concentrations of which correspond to the expected concentration ranges of the metals to be determined. Run these aqueous standard solutions through the method described under Working procedure.Then prepare organic standard solutions, the con- centrations of which correspond to the aqueous standard solutions prepared. Extraction is complete if the organic extracts of the aqueous standard solutions give the same absorbance values as the corresponding organic metal standard solutions. Working procedure When complete extraction has been verified, produce calibration graphs by measuring the absorbances of suitable organic standard solutions. In the procedure a 400-ml sample of water is treated with 20 ml of extraction solution, corresponding to an enrichment factor of 20. Other volumes, up to a ratio of the volume of the aqueous phase to the volume of the organic phase of 50: 1, can be used.Measure 400 ml of the water sample in a graduated cylinder and transfer into a 500-ml calibrated flask. Add 20ml of formate buffer solution; the colour of the indicator should be a pure yellow. If a red colour appears, add an additional 20 ml of formate buffer solution. Next add 2 ml of HMA HMDC solution in methanol and shake the flask vigorously. Wait for about 5 min, then add 20 ml of extraction solution and again shake the flask vigorously for at least 3 min. Wait for about 10 min in order to allow the layers to separate. Then carefully add water until the organic layer is completely in the neck of the flask. Adjust the absorbance reading of the atomic-absorption spectrometer to zero while aspirating the extraction solution.Next, aspirate the organic layers of each prepared sample, the organic layer of a blank treated in the same way as the samples and at least three organic standard solutions per element. In addition, aspirate mixed solvent (30% V/V xylene and 70% V/V DIPK) between each sample and between each standard in order to prevent clogging of the nebuliser. Finally, subtract the blank reading from the observed absorbance of each sample to obtain the true absorbance value of the sample. Calculation linear graph paper, of the absorbances of standards of the respective metal. Determine the concentration of each metal in the extract of each sample from plots, on Calculate the TABLE I11 EXTRACTION OF CADMIUM Aqueous solution I 1 Concentration/ Volume/ P.lg I-' ml 0.1 20 0.2 20 0.5 800 1 800 2.5 400 3 800 6 800 10 400 Extraction solution volume/ 1 1 16 16 20 16 16 20 ml Ratio of phases 20: 1 20: 1 50: 1 50: 1 20: 1 50: 1 50: 1 20: 1 Recovery, oh > 95 >95 > 95 > 95 > 95 > 95 > 95 > 95 Measurement technique* Graphite furnace Graphite furnace Flame Flame Flame Flame Flame Flame *Measurement : Perkin-Elmer, Model 420, with either graphite furnace HGASOO accessory, or an air - acetylene flame.1036 DORNEMANN AND KLEIST metal concentration of the samples by dividing the concentration values of the sample extracts by the enrichment factor of the extraction.Application Of the numerous check analyses that have been carried out, only those dealing with the extraction of small amounts of cadmium are reported.They are tabulated in Table 111. In all instances, recoveries were better than 95%. An inter-laboratory comparative test organised by DIN (German Institute for Standardisation) involved the determination of five metals in a sample of water. The starting solution was drinking water, which had been spiked with five heavy metals; ten laboratories participated in this test. The results are given in Table IV. Application of the Dixon testis for a 5% significance level led to TABLE IV INTER-LABORATORY COMPARATIVE TEST, DIN, OCTOBER 1978 Element Cd c o c u Ni Pb Number of laboratories . . . . 10 10 10 10 10 Outliers .. .. .. .. 1 Arithmetic mean of concentrations - 1 - - found/pg 1-1 . . * . .. . . 2.8 16 126 20 28 Absolutelpg 1-1 .. .. . . 0.19 1.6 10 1.7 3.5 Relative (variance), yo . . . . 6.7 9.9 7.9 8.6 12.4 Standard deviation the rejection as outliers of one cadmium value and one nickel value. these outliers the variance for lead was 12.4% and for the other metals below 10%. After elimination of 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. References Kirkbright, G. F., and Sargent, M., “Atomic Absorption and Fluorescence Spectroscopy,” Academic Dean, J. A., and Rains, T. C., Editors, “Flame Emission and Atomic Absorption Spectrometry,” Dornemann, A., Kleist, H., and Goergens, W.. 2. Analyt. Chem., 1977, 284, 97. Department of the Environment, “Cadmium in Potable Waters by Atomic Absorption Spectro- Thorn, G. D., and Ludwig, R. A., “The Dithiocarbamates and Related Compounds,” Elsevier Busev, A. I., Byrko, V. M., Tereschtschenko, A. P., Novikova, N. N., Naidina, V. P., and Terentev, Dornemann, A., and Kleist, H., 2. Analyt. Chem., 1978, 291, 349. Zolotov, Y . A., Analyst, 1978, 103, 56. Ginnings, P. A., Plonk, D., and Carter, E., J. Am. Chem. SOC., 1940, 62, 1923. Saylor, J. H., Baxt, V. J., and Gross, P. M., J . Am. Chem. SOG., 1942, 64, 2742. Dean, J . A., and Rains, T. C., in Dean, J. A., and Rains, T. C., Editors, “Flame Emission and Dornemann, A., and Kleist, H . , 2. Analyt. Cham., 1979, 295, 116. Wickbold, R., 2. Analyt. Chem., 1956, 152, 259. Eckert, G., 2. Analyt. Chem., 1957, 155, 23. Bode, H., and Tusche, K . J., 2. Analyt. Chem., 1957, 157, 414. Dixon, W. J., Ann. Math. Statist., 1951, 22, 68. Press, London, 1974, pp. 491-497. Volume 3, Marcel Dekker, New York, 1975, pp. 635-638. photometry,” Tentative method, HM Stationery Office, London, 1976. Publishing Company, Amsterdam, 1962. P. B., J . Analyt. Chem. USSR, 1970, 25, 665. Atomic Absorption Spectrometry,” Volume 2, Marcel Dekker, New York, 1971, pp. 327-339. Received March 2nd, 1979 Accepted June 4th. 1979