1082 Analyst September 1983 Vol. 108 PP. 1082-1085 Voltammetric Study of the Mercury Dissolution Reaction Mechanism at Solid Electrodes Paul Kiekens," Marc Mertens Charles Laire and Edward Temmerman Labovatory for Analytical Chemistry Ghent University Krijgslaan 28 1 S-12 B-9000 Ghent Belgium Stripping voltammetry with collection a t a rotating platinum ring - glassy carbon disc electrode was used for the elucidation of the mechanism of the electrochemical dissolution of a mercury film from a solid electrode. In complexing electrolytes the dissolution of mercury gives rise to the formation of divalent mercury ions. In non-complexing electrolytes the initial product of the electrodissolution of metallic mercury is also mercury(I1) which reacts with as yet unoxidised mercury atoms present on the electrode surface to give mercury(1) ions.Consequently mercury(1) ions are formed as a result of a reproportionation reaction following the electrochemical oxidation of mercury to mercury(I1). Keywords Stripping voltammetry ; rotating ring-disc electrode ; mercury dissolution kinetics ; reproportionation Mercury has widespread use in electroanalytical chemistry. In stripping voltammetry the mercury film on solid electrodes offers great advantages and has been used in many studies. However the electrodissolution kinetics of mercury itself from a solid electrode have been much less commented on in the literature. In particular no unambiguous answer can be found concerning the problem of the valence of the mercury ions formed during mercury electroionisation.The oxidation may produce either mercury(I) or mercury(I1) or both ions. This study was undertaken to try to elucidate this problem and to obtain an insight into the kinetics of mercury electrodissolution. We used mainly the rotating ring-disc technique.l Mercury(I1) ions are reduced and deposited on the disc of the electrode followed by stripping of the deposit using a linearly varying potential. A fraction (No) of the material stripped at the disc is transported to the surrounding ring electrode where a selective detection is possible at a fixed potential. As the potential of the ring is constant almost no capacitive current is flowing and the base line which is not influenced by the scan rate at the disc is essentially flat. The collection efficiency No,2 represents the fraction of the amount of electroactive species produced at the disc that theoretically reacts at the ring.Part of the reaction product at the disc escapes into the bulk of the solution during the crossing of the gap between ring and disc. No can also be calculated from the dimensions of the ring-disc electrode. Experimental Reagents Experiments were performed in solutions of hydrochloric perchloric sulphuric and nitric acids in solutions of sodium perchlorate sodium thiocyanate sodium sulphate and potassium nitrate at various concentrations and pH values and in 0.1 M acetate buffer solution (pH 4.66). Solutions were prepared from analytical-reagent grade reagents (Merck) and water freshly generated by a Milli-Q system (Millipore Inc.).A standard 0.1 M mercury(I1) solution was prepared from mercury(I1) nitrate and stabilised by adding a suitable amount of concentrated nitric acid (Merck Suprapur) to obtain a final acid concentration of 0.1 M. Cell solutions were carefully de-aerated with nitrogen containing less than 1 p.p.m. of oxygen. Instrumentation A home-made ring-disc electrode with a glassy carbon (Tokai Electrode Mfg. Co. Tokyo) disc and a platinum ring was used. The disc radius and inner and outer ring radii were 0.225, 0.246 and 0.268 cm respectively. This gives a theoretical collection efficiency No of 0.21. * Research Assistant of the NFSR (Belgium) KIEKENS MERTENS LAIRE AND TEMMERMAN 1083 The collection efficiency determined with the copper(I1) - copper(1) system in 0.5 M potassium chloride solution was 0.22.3 A saturated calomel electrode (S.C.E.) served as the reference electrode.A water-bath allowed the temperature to be held constant at 25 & 0.1 "C. The cell assembly apparatus and other equipment have been described el~ewhere.~ Electrode Pre-treatment The glassy carbon ring-disc electrode system was polished in the usual way the final polish-ing being effected with 0.05-pm alumina on Buehler microcloth to a mirror-like finish. Ultra-sonic vibration served to remove any alumina or other impurities (e.g. platinum from the ring electrode) that might adhere to the disc electrode surface. The electrochemical pre-treatment of the disc consisted of a continuous cycling of the poten-tial between fixed values in the working-electrolyte at a scan rate of 0.1 V s-l until a repro-ducible and very low background was obtained.The ring was pre-treated in the same way. During this pre-treatment nitrogen was passed through the cell solution. After the de-aeration and electrode activation period a nitrogen atmosphere was maintained above the cell solution. Electrolysis Procedure From an initial disc potential that was positive enough to avoid any mercury(I1) reduction, pre-electrolysis was started by switching the potential of the disc to the desired deposition potential Edep. Before the expiration of the electrolysis a potential that was positive enough to collect (oxidise) mercury(1) ions that might leave the disc electrode during mercury electro-dissolution was applied to the ring electrode. After deposition the disc potential was scanned linearly in the anodic direction and the mercury dissolution curve (disc) and collection curve (ring) were recorded.Results and Discussion A survey of literature data5-ls about mercury electrodissolution indicates that mercury(I1) ions are mainly formed in complexing electrolytes while mercury(1) ions are found in non-complexing electrolytes. Some data are summarised in Table I. One exception can be noticed according to Combet and Dozol electrodissolution of mercury in perchloric and nitric acid at a glassy carbon electrode produces only divalent mercury ions.12 According to Brainina and Neiman,14 a linear dependence is expected between the anodic peak potential E, of the metal stripping curve and the logarithm of the potential scanning rate v.Such a relationship is found for mercury ionisation in various electrolytes. Fig. 1 illustrates the linear behaviour between E and logv for 0.1 M sodium thiocyanate + 0.01 M perchloric acid and for 0.1 M perchloric acid. From the slopes14 of these plots the values of /3n (/3 = anodic transfer coefficient; n = number of electrons involved) were found to be 1.44 TABLE I ELECTRODISSOLUTION PRODUCTS OF MERCURY AT VARIOUS ELECTRODES IN DIVERSE ELECTROLYTES5-13 Electrode Platinum Glassy carbon Platinum Gold . . Platinum Mercury . . Graphite Graphite Glassy carbon Mercury . . ,. * . Electrolyte KNO, HClO, HClO, HNO, H2S04 KSCN KSCN KSCN HClO, HNO, NaCN H2S04 HClO, Reference 5 6 7 8 7 9 10 11 12 13 *Ed <0.6V.1 Ed >0.85 v 1084 Analyst Vol. 108 and 1.74 respectively. Other values obtained are 1.31 in 1 M sulphuric acid and 1.69 in 0.1 M acetate buffer solution (pH = 4.66). These results suggest that the valency of the mercury ions formed during electro-oxidation must be two independent of the electrolyte used to obtain a reasonable value for /3.15 However ring-disc experiments in perchloric and sulphuric acid solutions demonstrated the existence of monovalent mercury ions during electrodissolution. This finding is also in accordance with the experiments by Allen and .Johnson.6 The mercury(1) ions produced a t the disc may react at a sufficiently positive potential at the ring to give mercury(II) or they may also react at a negative ring potential to give metallic mercury.This is shown in Fig 2. The two ring collection peaks at 1.4 V (oxidation) and -0.2 V (reduction) are almost equal, indicating that the deposited mercury is mainly stripped as mercury(1). Of course this is in contradiction with the results obtained from simple semi-logarithmic analysis of the mercury ionisation peak at the disc as mentioned before. KIEKENS et al. VOLTAMMETRY OF THE MERCURY 0.47 4 0.46 c! tn 6 0.45 z 0.44 0.43 $ -2.0 0.04 0.03 4 tn c! 0.02 0.01 $ 0.00 Fig. 1. Dependence of the anodic stripping peak potential ED of mercury dissolution on the logarithm of the electrode potential scanning rate v in (A) 0.1 M perchloric acid and (B) 0.1 M sodium thiocyanate + 0.01 M per-chloric acid.Concentration of mercury(I1) is 2 ELM. 0.6 0.5 0.4 0.3 EJJ VS. S. C. E. Fig. 2. Current - potential graphs for electrodissolution of mercury at a glassy-carbon disc and simultaneous collection of stripped mercury ions at the platinum ring of a rotating ring-disc electrode in 0.1 M perchloric acid. Concentration of mercury-(11) = 10 ELM; deposition time = 1 min; rotation speed of the electrode = 188 rad s-l; E d e p = -1.0 V against the S.C.E.; v = 0.1 V s-1; Er = ring potential; and Ed = disc potential. These conflicting experimental results can be explained by accepting the occurrence of a special type of chemical reaction following mercury electrodissolution Le. reproportionation. The first formed mercury(I1) ions are supposed to react with as yet unoxidised mercury atoms still present on the electrode surface leading to the formation of mercury(1) ions.This explanation is acceptable as the equilibrium constant K for the reaction Hg(0) + 2Hg(I) or KgZ2+ indicates that monovalent mercury ions are stabilised (K = 166).16 These mercury(1) ions (possibly in a dimeric state) reach the surrounding ring electrode where they can be oxidised or reduced. This proposed reaction mechanism for mercury dissolution also explains the fact that the stripping peaks in thiocyanate solutions or generally in complexing electrolytes are abou September 1983 DISSOLUTION REACTION MECHANISM AT SOLID ELECTRODES 1085 twice the size of the dissolution peaks obtained in perchloric acid or generally in non-complex-ing electrolytes.In non-complexing electrolytes half of the mercury deposit may be stripped from the electrode surface in a non-electrochemical way because of the reproportionation reaction. In a complexing electrolyte our experiments indicate that the newly formed mercury(I1) ions at the electrode surface are involved in a subsequent chemical reaction, stabilising the mercury(I1) ions as complex ions. Investigations in an electrolyte solution of constant ionic strength containing x M sodium thiocyanate + (2-x) M sodium perchlorate + 0.01 M perchloric acid suggest that Hg(SCN) is the primary product of mercury electro-dissolution (0.1 < x < 2). This co-ordination number of 3 was derived from the dependence of the anodic peak potential on the thiocyanate concentration.ll The co-ordination number in the bulk of the solution is reported to be 4.lopl7 The oxidising properties of complexed mercury(I1) species towards mercury atoms can be calculated to be considerably lowered in comparison with the action of no or weakly complexed mercury(I1) ions.In complexing electrolytes the absence of monovalent mercury ions is also confirmed by ring-disc experiments. At the ring only a reduction signal of stripped divalent mercury ions can be obtained at sufficiently negative potentials. Finally we can conclude that the complexing properties of the electrolyte towards mercury ions are of considerable importance during mercury electrodissolution. This is also true for the concentration of the complexing agent. This was observed in hydrochloric acid solutions at various concentrations.At a low hydrochloric acid concentration e.g. 0.01 M both an oxidation and a reduction peak of mercury(1) ions produced at the disc can be obtained a t the ring. With increasing chloride ion concentration both the ionisation peak at the disc and the reduction peak at the ring grow while the oxidation peak at the ring diminishes. This means that more and more mercury(I1) ions are produced instead of mercury(1) ions and that reproportionation becomes less important because of the increased formation of mercury( 11) chloride complexes. At a hydrochloric acid concentration exceeding 2 M the mercury dis-solution peak reaches its maximum value and an oxidation peak at the ring can no longer be obtained.Hitherto only electrodissolution experiments a t a glassy carbon electrode have been de-scribed. However electro-oxidation measurements at a rotating ring-disc electrode with a disc of gold or iridium1* confirm the foregoing results obtained at a glassy carbon electrode. In non-complexing electrolytes the mercury ionisation again seems to be influenced by a repro-portionation reaction. This indicates the exclusive formation of divalent mercury ions at the disc. M.M. and Ch.L. thank the IWONL and P.K. the NFSR for financial support. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. References Albery W. J. and Hitchman M. L. “Ring-disc Electrodes,” Clarendon Press Oxford 1971. Albery W. J. and Bruckenstein S. Trans.Faraday SOC. 1966 62 1920. Napp D. T. Johnson D. C. and Bruckenstein S. Anal. Chem. 1967 39 481. Kiekens P. Verbeeck R. M. H. Donche H. and ‘I’enimerman E. J . Electroanal. Cltem. 1983 147, Hartley A. M. Hiebert A. G. and Cox A. J. J . Electroanal. Chem. 1968 17 81. Allen R. E. and Johnson D. C. Talanta 1973 20 799. Hassan M. Z. Untereker D. F. and Bruckenstein S. J . Eleclroanal. Chem. 1973 42 161. Lindstrom T. R. and Johnson D. C. Anal. Chem. 1981 53 1855. Eluard A. and TrBmillon B. J . Electroanal. Chem. 1967 13 208. Kartushinskaya A. I. Stromberg A. G. and Kolpakova N. A. Elektrokhimiya 1971 7 1243. Brainina Kh. Z. and Neiman E. Ya. Zh. Anal. Khim. 1971 26 875. Combet S. and Dozol M. Electrochim. A d a 1979 24 1283. Kirowa-Eisner E. Talmor D. and Osteryoung J. Anal. Chem. 1981 53 581. Brainina Kh. Z. “Stripping Voltammetry in Chemical Analysis,” John Wiley New York 1974. Verplaetse H. Kiekens P. Temmerman E. and Verbeek F. J . Electvoanal. Chem. 1980 115 235. Cotton F. A. and Wilkinson G. “Advanced Inorganic Chemistry,” Third Edition Interscience, Murray R. W. and Gross D. I. Anal. Chem. 1966 38 392. Laire Ch. unpublished results. 235. New York 1972 p. 509. Received February 25th 1983 Accepted March 18th 198