- Email: [email protected]
Analyte preconcentration and separation from small volumes by electrodeposition for electrothermal atomic absorption spectroscopy
0039-9140/93 s6.00+0.00
Tahnro,Vol. 40,No.12,~~. 18294831, 1993 Printedin Great Britain. All rights reserved
Copyright@J l993PcrgamonRess Ltd
ANALYTE PRECONCENTRATION AND SEPARATION FROM SMALL VOLUMES BY ELECTRODEPOSITION FOR ELECTROTHERMAL ATOMIC ABSORPTION SPECTROSCOPY JAROSLAVP.MATOUSEK* and
H. KIPTONJ. POWELL
Department of Chemistry, University of Canterbury, Private Bag 4800, Christchurch, New Zealand (Received 23 December 1992. Revised 23 April 1993. Accepted 17 May 1993) Summary-Electrodeposition of Pb from 50 pl volumes of O.lM KCl solution was studied by electrolysis at uncontrolled potentials of 4.0-6.0 V on a pyrolytic graphite platform cathode. Deposition etliciency was evaluated as a function of time by ASV measurements on aliquots of the electrolysed solution. Quantitative separation of the analyte from the matrix was achieved in a relatively short time, aided by convective stirring of the sample achieved through gas evolution. Thus, the feasibility of rapid electrodeposition directly in a pyrolytic graphite-coated furnace has been demonstrated, allowing construction of an automated electrodeposition-ekctrothermal atomic absorption spectroscopic system.
Electrolysis in an attractive method for both preconcentration and separation of analyte from interfering matrix prior to electrothermal atomic absorption spectroscopy (ETAAS). While alkali metal halides cause most serious interferences in ETAAS, they are ideal media for electrochemical techniques. For this reason, a number of studies and systems have been reported which advantageously combine the ability of both techniques. These have included (i) electrodeposition onto Hg-coated graphite furnaces from a circulating solution,‘-3 (ii) electrodeposition (at 3-5 V)’ or adsorption5 on a tungsten wire followed by its insertion into a cold furnace (or a pre-heated furnace”), (iii) electrodeposition in a well in a porous carbon rod’ and (iv) electrodeposition onto a Hg-coated pyrolytic graphite platform.* Circuitry has involved three-electrode systems’.3*7 and simple two-electrode constant potential systems.* The electrolytic method can achieve useful improvements in sensitivity, compared with conventional ETAAS, such as a 15-fold enhancement by 300 set electrolytic deposition on a tungsten wire from an unstirred solution4 and a 20- to 50-fold enhancement by 60 set adsorption from a stirred solution5 More importantly it also achieves separation of analyte from a matrix *Author for correspondence. Qn leave from Department of Analytical Chemistry, University of New South Wales, P.O. Box 1, Kensington, NSW 2033, Australia.
which may be difficult to remove by thermal pretreatment and may be a source of both chemical and spectral interferences in ETAAS, e.g. NaCl in seawater or blood digests. In addition, controlled potential electrolysis may achieve speciation by plating only the labile species of a metal. ‘s3When coupled with ETAAS the electrode position technique is also applicable to metals which have low sensitivity by ASV (such as Ni, Mn, Cr and Co) due to their irreversible reduction.’ However, despite these advantages, the recovery of the analyte from the sample is generally low and slow. If quantitative deposition is required, prohibitively long electrolysis times are involved.’ If, on the other hand, short deposition times are used only a small fraction of the analyte is recovered. For example, Fairless and Bard’ reported an efficiency of 1% in plating Cu from 5 ~1 stationary solution on a porous carbon rod (60 set at 0.18 V). In addition to slow analyte accumulation all of the above approaches suffer from common problems that include contamination, complicated apparatus (especially for the recirculating technique) and invariably poor reproducibility. A possible solution to such problems is automation of the combined electrodeposition-ETAAS technique. In situ deposition as performed by Fairless and Bard’ with subsequent removal of the exhausted electrolyte and washing of the deposit would lend itself well
1829
1830
JARO~LAV P. MATOU~EK and H. KIPTONJ. POWELL
to automated electrodeposition-ETAAS performed inside a conventional graphite furnace. However, the very inefficient analyte recovery which results from use of controlled potential electrolysis makes this approach unattractive for practical use. Invariably, controlled potential deposition has been used when combining electrochemical separation and preconcentration with atomic spectroscopic techniques. Even for stirred electrolyte the deposition times are inordinately long.’ This rather rigid adherence to controlled potential electrolysis (which is probably perceived as a well proven electrochemical approach) is not justified for techniques having the selectivity of ETAAS. Unless speciation is an objective, there is no need for controlled potential work. A necessary prerequisite for automation of the electrodeposition-ETAAS technique is fast and efficient separation and preconcentration of the analyte elements from volumes compatible with the capacity of conventional graphite furnaces (usually in the range 20-100 ~1). In order to evaluate if this can be achieved on the time scale similar to conventional graphite furnace analysis, we have studied electrodeposition from 50 ~1 volumes by applying relatively high voltages to sustain high electrolysis currents and to enhance the deposition rate further by convection created by gas evolution at both electrodes. EXPERIMENTAL
To establish conditions for quantitative deposition, electrolysis was effected on samples deposited in the well of a Perkin-Elmer “V’ pyrolytic graphite platform.” The platform with the “V’‘-shaped well 0.7 mm deep was 15 mm long, 4 mm wide and 1 mm thick. It served as a cathode, with a 0.5 mm dia. Pt wire anode mounted centrally 1 mm above the platform surface. In later experiments, the Pt anode was mounted on a micrometer attachment to allow accurate setting of the anode-cathode separation and a 0.5 mm dia. Ag/AgCl reference electrode was added to monitor the cathode potential. The electrolyte used consisted of O.lM KC1 and 0.02M acetate buffer (pH 4.7) containing 0.02,O. 10 or 0.50 mg/l Pb. Each electrolysis was effected on a 50 ~1 sample for a fixed time using a constant potential power supply (4.0, 5.0 or 6.0 V). To determine concentration of the ana-
lyte element remaining in the electrolysed solution, a 25 ~1 aliquot was taken using a micropipette and transferred from the platform to a PAR 303 polarographic cell containing 5 ml Milli-Q water and 20 ~1 Aristar HNO,. The solution was analysed by differential pulse ASV using a PAR 174A polarograph. ASV conditions were: N, flush, 9 min; stir, fast; drop size, medium; deposition time, 6 min; E,, = -0.7 V; scan rate = 5 mV/sec; modulation, 25 mV. After washing with 3% HNO, (electrodeposited Pb is dissolved readily by dilute HNO$ and drying the pyrolytic graphite platform was ready for a new electrolysis. Since measurements of Pb concentrations down to sub-ppb levels were involved in the solutions remaining after electrolysis (and dilution in the polarographic cell), all experiments were performed in a class 100 clean room.
RESULTS AND DISCUSSION
In this study, we have simulated the problem of in situ deposition inside a graphite furnace by using a pyrolytic graphite platform and depositing from small volumes compatible with the size of the graphite furnace in ETAAS. Electrolysis was effected with uncontrolled potential set at 4.0, 5.0 or 6.0 V. The efficiency of the deposition was evaluated as a function of time from the analyte concentration remaining after electrolysis. This was determined by differential pulse ASV measurements on aliquots of the electrolysed solution. The results presented in Fig. 1 show that the deposition efficiency increases significantly with the applied voltage. The increase in the depo-
Fig. 1. The effect of applied voltage on efficiency of electrodeposition of 0.1 mg/l Pb from O.IM KCI. Applied voltages used: (0) 4, (0) 4 with Hg co-deposition, (A) 5 and (0) 6 V.
Analyte preconoentration
1831
and separation
electrolysis currents. It should be noted that similar problems, interpreted as being caused by intercalation, have been experienced when pyrolytic graphite platforms are used in ETAAS.” CONCLUSIONS
OO-’
Time, min.
Fig. 2. The effect of Pb concentration on efficiency of electrodeposition from O.lM KC1 electrolysed at 6 V. Pb concentrations used (0) 0.02, (0) 0.10 and (A) 0.50 mg/l.
sition rate with applied voltage is related to convective stirring of the sample which is achieved through adequate evolution of Clz, O2 and H,; gas evolution is apparent once the current exceeds 5 mA. Thus, deposition was complete in 4 min at 6.0 V and 90% complete in 12 min at 4.0 V. Co-deposition of Pb with Hg (from 0.0002M Hg’+ ) did not significantly increase plating efficiency (Fig. 1). Our experiments further confirmed that the deposition rate was independent of the Pb concentration as shown in Fig. 2 for 0.02, 0.1 and 0.5 mg/l Pb. This indicates that the rate of deposition is controlled by the degree of convection achieved through the liberation of H,, 0, and Cl,. In order to characterize the pyrolytic graphite platform electrodeposition system, separate experiments were performed in which cathode potentials were measured relative to a Ag/AgCl reference electrode. With the anode-cathode separation set at 0.63 mm by a micrometer attachment, the cathode potential was established as - 1.23, -1.52 and -1.87 V at E,,,,i, = 4.0, 5.0 and 6.0 V, respectively. The current depended on the anode-cathode separation but was approximately 0.3, 2.0 and 4.5-6.0 mA at 4.0, 5.0 and 6.0 V, respectively. The pyrolytic graphite platforms used in these experiments showed only limited lifetime. Exfoliation of pyrolytic graphite layers was observed to increase gradually with number of electrolyses performed, however this did not appear to affect the deposition rate for the first 50 electrolyses. The problem appears to be directly related to the number of pyrolytic graphite layers exposed in the manufacture of the platform used.‘O The electrolyte gradually forces its way between the layers, causing exfoliation, especially at higher
The present report establishes that a batch electrolysis from 50 ~1 solution volumes can achieve quantitative deposition of the analyte element in relatively short times when conducted at a potential which can effect convective mixing by evolution of HZ, 0, and Cl,. In recent publications, Sioda er .1.“,” have suggested and experimentally verified that there are limits to electrolytic preconcentration caused by equilibrium concentrations of cations remaining in solution. It is possible to conclude that even though such limits apply to controlled potential work, they do not measurably affect the outcome of electrolyses performed under the conditions of our experiments. Based on the principles established here, a recent paper by Matousek and Grey” has described instrumental modifications to allow automated in situ electrodeposition of analytes in the conventional pyrolytic graphite-coated furnace. The modified system consists of a GBC GF 2000 graphite furnace system equipped with a PAL autosampler and is capable of performing automated sample loading, electrolysis, withdrawal of the electrolyte, washing and chemical pretreatment of the electrodeposited metal. REFERENCES 1. G. E. Batley and J. P. Matousek, Anal. Chem., 1977,49, 2031. 2. G. Vollard, P. Tschiipel and G. Tiilg, Anal. Chim. Acta, 1977, 90, 15. 3. G. E. Batley and J. P. Matousek, Anal. Chem., 1980,52, 1570. 4. E. J. Czobik and J. P. Matousek, Specrrochim. Acta, 1980, 3SB, 741. 5. Y. Hoshino, T. Utsunomiya and K. Fukui, Chem. Left., 1976, 9, 947. 6. J. P. Matousek, Ph.D. Thesis, The University of New
South Wales, 1978. 7. C. Fairless and A. J. Bard, Anal. L&t., 1972, 5, 433. 8. J. Shiowatana and J. P. Matousek, Talanfa, 1991, 38, 375. 9. H. Matusiewicz, J. Fish and T. Malinski, Annl. Chem., 1987, 59, 2264.
10. G. R. Camrick and B. K. Lumas, Atom. Spectrosc., 1984, 5, 135. 11. R. E. Sioda, Anal. Chem., 1988, 60, 1177. 12. A. Ciszewski, J. R. Fish, T. Malinski and R. E. Sioda, ibid., 1989, 61, 856. 13. J. P. Matousek and R. Grey, Proc. 27th Coil. Specrrosc. Ink, Paper No. B-6.4, Bergen, 1991.
Tahnro,Vol. 40,No.12,~~. 18294831, 1993 Printedin Great Britain. All rights reserved
Copyright@J l993PcrgamonRess Ltd
ANALYTE PRECONCENTRATION AND SEPARATION FROM SMALL VOLUMES BY ELECTRODEPOSITION FOR ELECTROTHERMAL ATOMIC ABSORPTION SPECTROSCOPY JAROSLAVP.MATOUSEK* and
H. KIPTONJ. POWELL
Department of Chemistry, University of Canterbury, Private Bag 4800, Christchurch, New Zealand (Received 23 December 1992. Revised 23 April 1993. Accepted 17 May 1993) Summary-Electrodeposition of Pb from 50 pl volumes of O.lM KCl solution was studied by electrolysis at uncontrolled potentials of 4.0-6.0 V on a pyrolytic graphite platform cathode. Deposition etliciency was evaluated as a function of time by ASV measurements on aliquots of the electrolysed solution. Quantitative separation of the analyte from the matrix was achieved in a relatively short time, aided by convective stirring of the sample achieved through gas evolution. Thus, the feasibility of rapid electrodeposition directly in a pyrolytic graphite-coated furnace has been demonstrated, allowing construction of an automated electrodeposition-ekctrothermal atomic absorption spectroscopic system.
Electrolysis in an attractive method for both preconcentration and separation of analyte from interfering matrix prior to electrothermal atomic absorption spectroscopy (ETAAS). While alkali metal halides cause most serious interferences in ETAAS, they are ideal media for electrochemical techniques. For this reason, a number of studies and systems have been reported which advantageously combine the ability of both techniques. These have included (i) electrodeposition onto Hg-coated graphite furnaces from a circulating solution,‘-3 (ii) electrodeposition (at 3-5 V)’ or adsorption5 on a tungsten wire followed by its insertion into a cold furnace (or a pre-heated furnace”), (iii) electrodeposition in a well in a porous carbon rod’ and (iv) electrodeposition onto a Hg-coated pyrolytic graphite platform.* Circuitry has involved three-electrode systems’.3*7 and simple two-electrode constant potential systems.* The electrolytic method can achieve useful improvements in sensitivity, compared with conventional ETAAS, such as a 15-fold enhancement by 300 set electrolytic deposition on a tungsten wire from an unstirred solution4 and a 20- to 50-fold enhancement by 60 set adsorption from a stirred solution5 More importantly it also achieves separation of analyte from a matrix *Author for correspondence. Qn leave from Department of Analytical Chemistry, University of New South Wales, P.O. Box 1, Kensington, NSW 2033, Australia.
which may be difficult to remove by thermal pretreatment and may be a source of both chemical and spectral interferences in ETAAS, e.g. NaCl in seawater or blood digests. In addition, controlled potential electrolysis may achieve speciation by plating only the labile species of a metal. ‘s3When coupled with ETAAS the electrode position technique is also applicable to metals which have low sensitivity by ASV (such as Ni, Mn, Cr and Co) due to their irreversible reduction.’ However, despite these advantages, the recovery of the analyte from the sample is generally low and slow. If quantitative deposition is required, prohibitively long electrolysis times are involved.’ If, on the other hand, short deposition times are used only a small fraction of the analyte is recovered. For example, Fairless and Bard’ reported an efficiency of 1% in plating Cu from 5 ~1 stationary solution on a porous carbon rod (60 set at 0.18 V). In addition to slow analyte accumulation all of the above approaches suffer from common problems that include contamination, complicated apparatus (especially for the recirculating technique) and invariably poor reproducibility. A possible solution to such problems is automation of the combined electrodeposition-ETAAS technique. In situ deposition as performed by Fairless and Bard’ with subsequent removal of the exhausted electrolyte and washing of the deposit would lend itself well
1829
1830
JARO~LAV P. MATOU~EK and H. KIPTONJ. POWELL
to automated electrodeposition-ETAAS performed inside a conventional graphite furnace. However, the very inefficient analyte recovery which results from use of controlled potential electrolysis makes this approach unattractive for practical use. Invariably, controlled potential deposition has been used when combining electrochemical separation and preconcentration with atomic spectroscopic techniques. Even for stirred electrolyte the deposition times are inordinately long.’ This rather rigid adherence to controlled potential electrolysis (which is probably perceived as a well proven electrochemical approach) is not justified for techniques having the selectivity of ETAAS. Unless speciation is an objective, there is no need for controlled potential work. A necessary prerequisite for automation of the electrodeposition-ETAAS technique is fast and efficient separation and preconcentration of the analyte elements from volumes compatible with the capacity of conventional graphite furnaces (usually in the range 20-100 ~1). In order to evaluate if this can be achieved on the time scale similar to conventional graphite furnace analysis, we have studied electrodeposition from 50 ~1 volumes by applying relatively high voltages to sustain high electrolysis currents and to enhance the deposition rate further by convection created by gas evolution at both electrodes. EXPERIMENTAL
To establish conditions for quantitative deposition, electrolysis was effected on samples deposited in the well of a Perkin-Elmer “V’ pyrolytic graphite platform.” The platform with the “V’‘-shaped well 0.7 mm deep was 15 mm long, 4 mm wide and 1 mm thick. It served as a cathode, with a 0.5 mm dia. Pt wire anode mounted centrally 1 mm above the platform surface. In later experiments, the Pt anode was mounted on a micrometer attachment to allow accurate setting of the anode-cathode separation and a 0.5 mm dia. Ag/AgCl reference electrode was added to monitor the cathode potential. The electrolyte used consisted of O.lM KC1 and 0.02M acetate buffer (pH 4.7) containing 0.02,O. 10 or 0.50 mg/l Pb. Each electrolysis was effected on a 50 ~1 sample for a fixed time using a constant potential power supply (4.0, 5.0 or 6.0 V). To determine concentration of the ana-
lyte element remaining in the electrolysed solution, a 25 ~1 aliquot was taken using a micropipette and transferred from the platform to a PAR 303 polarographic cell containing 5 ml Milli-Q water and 20 ~1 Aristar HNO,. The solution was analysed by differential pulse ASV using a PAR 174A polarograph. ASV conditions were: N, flush, 9 min; stir, fast; drop size, medium; deposition time, 6 min; E,, = -0.7 V; scan rate = 5 mV/sec; modulation, 25 mV. After washing with 3% HNO, (electrodeposited Pb is dissolved readily by dilute HNO$ and drying the pyrolytic graphite platform was ready for a new electrolysis. Since measurements of Pb concentrations down to sub-ppb levels were involved in the solutions remaining after electrolysis (and dilution in the polarographic cell), all experiments were performed in a class 100 clean room.
RESULTS AND DISCUSSION
In this study, we have simulated the problem of in situ deposition inside a graphite furnace by using a pyrolytic graphite platform and depositing from small volumes compatible with the size of the graphite furnace in ETAAS. Electrolysis was effected with uncontrolled potential set at 4.0, 5.0 or 6.0 V. The efficiency of the deposition was evaluated as a function of time from the analyte concentration remaining after electrolysis. This was determined by differential pulse ASV measurements on aliquots of the electrolysed solution. The results presented in Fig. 1 show that the deposition efficiency increases significantly with the applied voltage. The increase in the depo-
Fig. 1. The effect of applied voltage on efficiency of electrodeposition of 0.1 mg/l Pb from O.IM KCI. Applied voltages used: (0) 4, (0) 4 with Hg co-deposition, (A) 5 and (0) 6 V.
Analyte preconoentration
1831
and separation
electrolysis currents. It should be noted that similar problems, interpreted as being caused by intercalation, have been experienced when pyrolytic graphite platforms are used in ETAAS.” CONCLUSIONS
OO-’
Time, min.
Fig. 2. The effect of Pb concentration on efficiency of electrodeposition from O.lM KC1 electrolysed at 6 V. Pb concentrations used (0) 0.02, (0) 0.10 and (A) 0.50 mg/l.
sition rate with applied voltage is related to convective stirring of the sample which is achieved through adequate evolution of Clz, O2 and H,; gas evolution is apparent once the current exceeds 5 mA. Thus, deposition was complete in 4 min at 6.0 V and 90% complete in 12 min at 4.0 V. Co-deposition of Pb with Hg (from 0.0002M Hg’+ ) did not significantly increase plating efficiency (Fig. 1). Our experiments further confirmed that the deposition rate was independent of the Pb concentration as shown in Fig. 2 for 0.02, 0.1 and 0.5 mg/l Pb. This indicates that the rate of deposition is controlled by the degree of convection achieved through the liberation of H,, 0, and Cl,. In order to characterize the pyrolytic graphite platform electrodeposition system, separate experiments were performed in which cathode potentials were measured relative to a Ag/AgCl reference electrode. With the anode-cathode separation set at 0.63 mm by a micrometer attachment, the cathode potential was established as - 1.23, -1.52 and -1.87 V at E,,,,i, = 4.0, 5.0 and 6.0 V, respectively. The current depended on the anode-cathode separation but was approximately 0.3, 2.0 and 4.5-6.0 mA at 4.0, 5.0 and 6.0 V, respectively. The pyrolytic graphite platforms used in these experiments showed only limited lifetime. Exfoliation of pyrolytic graphite layers was observed to increase gradually with number of electrolyses performed, however this did not appear to affect the deposition rate for the first 50 electrolyses. The problem appears to be directly related to the number of pyrolytic graphite layers exposed in the manufacture of the platform used.‘O The electrolyte gradually forces its way between the layers, causing exfoliation, especially at higher
The present report establishes that a batch electrolysis from 50 ~1 solution volumes can achieve quantitative deposition of the analyte element in relatively short times when conducted at a potential which can effect convective mixing by evolution of HZ, 0, and Cl,. In recent publications, Sioda er .1.“,” have suggested and experimentally verified that there are limits to electrolytic preconcentration caused by equilibrium concentrations of cations remaining in solution. It is possible to conclude that even though such limits apply to controlled potential work, they do not measurably affect the outcome of electrolyses performed under the conditions of our experiments. Based on the principles established here, a recent paper by Matousek and Grey” has described instrumental modifications to allow automated in situ electrodeposition of analytes in the conventional pyrolytic graphite-coated furnace. The modified system consists of a GBC GF 2000 graphite furnace system equipped with a PAL autosampler and is capable of performing automated sample loading, electrolysis, withdrawal of the electrolyte, washing and chemical pretreatment of the electrodeposited metal. REFERENCES 1. G. E. Batley and J. P. Matousek, Anal. Chem., 1977,49, 2031. 2. G. Vollard, P. Tschiipel and G. Tiilg, Anal. Chim. Acta, 1977, 90, 15. 3. G. E. Batley and J. P. Matousek, Anal. Chem., 1980,52, 1570. 4. E. J. Czobik and J. P. Matousek, Specrrochim. Acta, 1980, 3SB, 741. 5. Y. Hoshino, T. Utsunomiya and K. Fukui, Chem. Left., 1976, 9, 947. 6. J. P. Matousek, Ph.D. Thesis, The University of New
South Wales, 1978. 7. C. Fairless and A. J. Bard, Anal. L&t., 1972, 5, 433. 8. J. Shiowatana and J. P. Matousek, Talanfa, 1991, 38, 375. 9. H. Matusiewicz, J. Fish and T. Malinski, Annl. Chem., 1987, 59, 2264.
10. G. R. Camrick and B. K. Lumas, Atom. Spectrosc., 1984, 5, 135. 11. R. E. Sioda, Anal. Chem., 1988, 60, 1177. 12. A. Ciszewski, J. R. Fish, T. Malinski and R. E. Sioda, ibid., 1989, 61, 856. 13. J. P. Matousek and R. Grey, Proc. 27th Coil. Specrrosc. Ink, Paper No. B-6.4, Bergen, 1991.