Electroreduction of Cobalt-Amino Peroxs ComplexesPart 1.-The Reduction of Oxygen in the CO(II) + Ammonia System-/-BY ARMAND BETTELHEIM,* M. FARAGGI, 1. HODARA AND J. MANASSEN:Atomic Energy Commission, Nuclear Research Centre-Negev,P.O.B. 9001, Beer-Sheva, IsraelReceived 19th March, 1976The electroreduction of O2 in Co(@ + NH3 solutions has been investigated using the rotatingdisc electrode technique. It was found that the electroactive species, in the potential range 0 to- 1 V against s.c.e., are O2 and the peroxo complex [(NH3)5Co-02-Co(NH3)5]+4. The electro-activity of the complex was shown by the appearance of a new cathodic wave and by the linearincrease of the limiting current of this wave with the peroxo complex concentration. It is suggestedthat four electrons are involved in the reduction of the complex on a Pt rotating disc electrode.Its diffusion coefficient was calculated to be 8 .6 ~ cm2 s-I. The decomposition of the peroxocomplex to form a mononuclear complex of Co(n1) was a slow process (k = 3 . 0 ~ s-l).Synthetic reversible oxygen carrying chelates have been of interest as modelcompounds in the study of the reversible oxygenation mechanisms involved in naturaloxygen carriers, e.g., the haemoglobins and haemocyanins.It is well known that many CO(II) complexes take up molecular oxygen readilyin aqueous solutions to give binuclear peroxo c~mplexes.l-~ The best knowncomplex is [(NH3)5C~-02-C~(NH3)5]f4. The reversibility of the reaction withoxygen and the rate of oxygen uptake have been investigated by numerous authorsin the past using spectroscopic, potentioinetric and kinetic methods.3*This work deals with the electroreduction of oxygen in the presence of the CO(II) +ammonia system, using the rotating disc electrode (r.d.e.) technique.EXPERIMENTALU7ater was triple distilled.Merck pro analysis Co(N03)2, NH4N03 and NH3 wereused without further purification. All other materials were of analytical grade. Matheson“ extra dry ” oxygen and ‘‘ high purity ” nitrogen were used for the saturation of thesolutions.The solutions were prepared by mixing 6.3 mol dm-3 NH3 and 2 rnol dm-3 NH4N03 atpH 10.3 with 0-10 mmol dm-3 CO(NO,)~ shortly before the experiments. Saturation of thesolutions by the appropriate gas was carried out by continuous bubbling for at least 30 min.through a trap of aqueous ammonia (6.3 mol dm-3).Two types of solutions were prepared :Solutions A; these were saturated with oxygen before the addition of the Co(u) ion and nofurther oxygen was introduced afterwards.Solutions B ; oxygen saturation was carried out after the addition of the Co(u) ion and wascontinued for a further 30 min.In the type A solutions, the total concentration of oxygen is constant and equal to thesolubility of oxygen in the 6.3 rnol dm-3 NH3 and 2 mol dm-3 NH4N03 solutions :(1) [OJ+ [CO-O~-CQ~ = constant (solutions A).7 Based on A.B.3 Ph.D.Thesis submitted to the Weizmann Institute of Science, Rehovoth, Israel.1 Present address : The Weizmann Institute, Rehovoth, Israel.14144 ELECTROREDUCTION OF O2 I N C ~ ( I I ) + N H ,In the type B solutions, it is assumed that the Concentration of the free oxygen [not boundto Co(n) complex] is constant, its value being determined by its solubility in the 4.3 mol dm-3NH3 and 2 mol dm-3 NH4N03 solutions :(2)The peroxo coixentration in solutions B increases as the Co(rr) ion concentration is increased.The maximum concentration of the peroxo complex is half the concentration of the addedCO(II) ion.The rotating disc electrode (r.d.e.) consisted of a platinum cylinder pressed into teAon(supplied by Pine Tnst.Co.) The area of the electrode was 0.458 cm2. The electrodc wasrotated at 340r.p.m. by means of a Pir-Rotator (Pine Tnst. Go., serial 35"). The counterelectrode was a platinum wire, separated from the solution by a porous glass sinter.Thereference electrode was a saturated calomel electrode (s.c.e.) made according to Meitesand Thomas." All the potentials were measured against this electrode. Pretreatmentof the Ft-r.d.e. was according to Bockris and CO-workers.'Potentiostatic current against potential curves were obtained by applying to the cell apotential sweep of 0.45 V min-l using a potentiostat (both the potential ramp generator andthe poteiitiostat were built at the electronics laboratory of the Nuclear Research Centre-Negev).Coulometric measurements at a controlled potential were made with the aid of a voltageto pulse rate converter unit [Elron-model A(PRC I-B)] and an electron scaler (Elron-modelPolarographic experiments were performed with a Radiometer-PoIariter Type PO 4d.Excess of dissolved oxygen was removed by bubbling nitrogen for 20 min before recordingpolarograms.The dropping mercury electrode (d.m.e.) was a capillary electrode (suppliedby Sargent and Co.). The capillary characteristics were In = 1.799 mg s-l, t = 4.69 sand in3 t i = 1.91 mg3 s-3 under a mercury head of 850 nrn at a potential of -0.5 V aginsts.c.e.fO,] = constant (solutions B).N r s- 1 4-PI.All measurements were performed at 25°C (& 0.l"C).RESULTS AND DISCUSSIONIn arninoniacal solutions, the deca-amine complex [(NH3)5Co-0,-Co(NH3)5]+4is most stable in the pH range 10-12 and high ammonia concentration ([NH3]-7 moldm-3).3* This is presumably the range of pH in which the aquo-pcnta-amine CO(II) coinplex predominates in the absence of oxygen.The significance ofthis is clear, in view of the equilibrium~ C Q ( N H ~ ) ~ ( H ~ O ) + ~ + 0 2 + [(NH~)~CO-O~-CO(NH~)~]+~ + 2H28. (3)It is of interest to know whether oxygen is the only electroactive species in thepotential range 0 to - 1 V in the above equilibrium.Fig. 1 describes typical current against voltage curves measured in oxygensaturated ammoniacal solutions. Curve C is a voltammogram of oxygen reductionin the NH3/NH4N03 base electrolyte saturated with oxygen. The reduction ofoxygen on a Pt electrode in the above ammoniacal solutions is thus characterized bya single wave with a value of E3 = -0.24 17. In the presence of Co(n), two waves(waves 1 and 11) are observed.The first wave (wave I) has a similar value of E, asbefore and the half wave potential value of the second wave (wave 11) is -0.58 V.Voltammograms Al, A, and A3 (fig. 1) describing reduction in type A solutions[saturated with oxygen before adding Co(r1) ion] indicate that the total limiting currentdecreases as the concentration of CO(II) ion is increased. The opposite effect isobserved in typz B solutions [saturated with oxygen after the Co(11) addition] : thetotal limiting current increases as the Co(~r) ion concentration is increased (curves B1,13, and B3).The single wave of cathodic reduction of oxygen in 6.3 mol d n r 3 NH, anA . BETTELHEIM, M . FARAGGI, I . HODARA AND J . MANASSEN 1452 rnol d n ~ - ~ NH4N03 solutions is explained as a reduction involving the transfer of 4electrons.This was shown by a coulometric collecting charge at a controlledpotential of -0.6 V. The charge collected under these conditions was 2.8 C ascompared with the theoretical value of 0.73 C per electron. Similar results in acidsolutions were reported by Bockris and co-workers.'It should be noted that wave I appears in solution with and without the presenceof Co(rr) ion. Its potential is independent of the cation presence. This seems toindicate that this wave is related to the reduction of free oxygen.The dependence of the limiting current of wave I was plotted aga,inst the CO(II)ion coilcentration (fig. 2). In this figure, it is clearly demonstrated that, whcreas thet , ,-0.3 -0.6 -0.3voltage against s.c.e.FIG.1.---R.d.e. voltainniograms ( w = 340 r.p.m.) of 6.3 mol d ~ n - ~ NH3 and 2 mol dm-3 NH4N03oxygen saturated solutions. Curve C, no Co(r1j ion present; curves Al, A2 and A3, type Asolutions with 2, 4 and 8 mmol dm-3 Co(NO3j2 respectively ; curves B1, B2 and B3, type €3solutions with 2, 4 and 8 minol dm-3 CO(NO,)~ respectively.0.60 4 8[CO(IOl /mMFIG. 2.-The dependence of the limiting current of wave I on CO(NO,)~ concentration in 6.3 niol d1n-3NH3 and 2 mol dm-3 NH4N03 solutions : curve A1, type A air saturated solutions ; curve AZ,type A oxygen saturated solutions ; curve B1, type B air saturated solutions ; curve BZ, type B oxygensaturated solutions146 ELECTROREDUCTION OF 0 2 IN CO(II)+NH~limiting current in type A solutions decreases with the total Co(n) ion concentration,in type B solutions it has apparently a constant value.This strengthens the suggestionthat the first wave (wave I) is related to the free oxygen reduction. In the type Asolutions, the overall oxygen concentration is constant (and equal to the solubilityof oxygen in 6.3 mol dm-3 NH3 and 2 mol dm-3 NH4N03), therefore the productionof the peroxo complex decreases the free oxygen concentration and thus decreases thelimiting current. In the type B solutions, the free oxygen concentration is constantand thus one would expect a constant limiting current for the reduction of free oxygen.[peroxo] x 104/mol dm-3FIG. 3.--The dependence of the total limiting current on peroxo concentration for type B, air (curveBJ and oxygen (curve B,) saturated solutions.TABLE EQUILIBRIUM CONCENTRATIONS OF FREE OXYGEN AND OF THE PEROXO COMPLEX INTYPE A AND B SOLUTIONS CONTAINING 6.3 mol dm-3 NH3 AND 2 mol dm--3 NH4N03.type A solutions type B solutionsair oxygen air oxygen[Coltot.1021, beroxol 1021, [peroxol [Od, tperoxol [Od, Iperoxol/mmol dm-3 X lO-4/mol dm-3 x lO-4/mol dm-3 x lO-4/mol dm-3 x lO-4/mol dm-30 1.50 0 6.00 0 1.50 0 6.00 02 0.92 0.58 4.14 1.86 1.50 0.85 6.00 2.354 0.38 1.12 2.16 3.54 1.50 2.95 6.00 6.956 0.23 1.27 1.07 4.93 1.50 5.90 6.00 12.506.00 18.65 8 0.16 1.34 0.40 5.60 1.50 9.50In fig. 3, the total limiting current (wave I+wave 11) was plotted against theequilibrium concentrations of the peroxo complex. These concentrations, given intable 1, were calculated on the basis of Simplicio and Wilkin’s data for the equilibriumgiven in reaction (3) and :CO(NH~)~(H~O)+~ +NH3 + Co(NH3); +H,O (4)2Co(NH,): + 0 2 + [(NH~)~CO-O~-CO(NH~)~]+~ +2NH3. (5)The equilibrium constants of reactions (3), (4) and (5) are 1.6 x lo6 dm6 mok2,0.25 dm3 mol-1 and 2.5 x lo7, respectively.’ In view of the small concentrationchanges in the equilibrium components during the experiment, we assume that thereis no shift in the above equilibria, the voltage sweep being fast enough to preventdepletion of electroactive species by electrolysisA .BETTELHEIM, M. FARAGGI, I . HODARA AND J . MANASSEN 147Wave I1 appears only in solutions containing CO(II) ions. This fact together withthe linear correlation between the overall limiting current and the equilibrium peroxoconcentration suggests that wave I1 characterizes the reduction of the peroxocomplex. The extrapolated value ([peroxo] -+ 0) of the limiting current is the valueexpected for wave I in the absence of Co(11) ion (fig.1, curve C). The values obtainedin fig. 3 for air and oxygen saturated solutions are in agreement (within the experi-mental error) with the limiting current found for wave I in the above solutions.The limiting current for the r.d.e. system is given by :8iL = 0.62 nE;ACD2!3v-1i6~112where iL is the limiting current (expressed in A), n is the number of electrons, Fis theFaraday constant (in C mol-l), A is the electrode area (in cm2), C is the bulk con-centration (in mol ~ r n - ~ ) , D is the diffusion coefficient (in cm2 s-'), v is the kinematicviscosity (in cm2 s-l) and cr) is the rotating velocity (in rad s-l).From the aboveequation, it is expected that parallel linear curves will be obtained for air and oxygentype B saturated solutions (curves B1 and B2 respectively), the ratio iJC being aconstant value for a given electroactive species and a constant speed of rotation.For oxygen, the calculated value of the ratio iL/C (iL is taken from r.d.e. volt-ammograms and C is taken from table 1) is 0.88 A dm3 mol-l. For the peroxocomplex, this value calculated from the slope of the paraller linear curves B1 and B2(fig. 3) is 0.42 A dm3 mol-l. The experimental value of the ratio :nperoxoD&roxo = 0.48.As no, and Do, are known (no, = 4 and Doz = 2.6 x s-'),nperoxoD~eroxo = 1.68 xThe number of electrons involved in the reduction of the peroxo complex(NH3)5Co-0,-Co(NH3)5 could not be determined by our techniques (because offree oxygen reduction interference) as was done for the peroxo complexes of CO(II)ion and ethylenediamine and triethylenetramine as ligand~.~ However, assumingnperoxo = 1 and 2, the value of Dperoxo obtained is greater or equal to that of Do2.It is rather unlikely that the peroxo molecule, which has a greater size than theoxygen molecule, would have a similar diffusion coefficient.Assuming nperoxo = 4,then the value of Dperoxo is 8.6 x cm2 s-l. This value is to be compared with thevalue for the superoxo complex of CO(III) estimated in the literature to be 6.0 xcm2 s-l.l0The small value of the diffusion coefficient of the peroxo complex can explain thedecrease in the total limiting current with CO(II) ion addition in type A solutions :the increase in wave I1 (caused by the increase of peroxo concentration) is smallerthan the decrease of wave I (caused by the decrease of free oxygen concentration).Hence, there is a net decrease in total limiting current (curves A', A, and A3 infig.1). In type B solutions, the peroxo concentrations are high and the oxygenconcentration is constant when CO(II) ion is added. Hence, a total increase oflimiting current is observed (curves B1, B2 and B3 in fig. 1).THE RATE CONSTANT OF THE DECOMPOSITION OF THE PEROXO COMPLEXIt is known that CO(II) complexes lose their properties as oxygen carriers by aside reaction ; the peroxo binuclear complex dissociates to yield a mononuclearCO(III) complex :39(6) LCO-O,-COL + 2H+ + 2CO"'L + H202148 ELECTROREDUCTION OF O2 I N CO(II)+NH~The unknown decomposition rate of (NH3)sC~-02-C~(NH3)5 has been investi-gated using the different polarographic behaviour of the peroxo and the CO(III)amino complexes.When an oxygen saturated ammoniacal solution containing Co(11) ion is bubbledwith nitrogen, oxygen is expelled.The equilibrium, represented by reactions (3) and(5), is shifted towards the dissociation of the peroxo complex. The Co(11) complexesare not reduced in the potential range investigated (0 to - 1.0 V). The CO(III)complex(es) obtained via reaction (6) is the only electroactive species which could bereduced at the d.m.e.in the above potential range. Fig. 4 describes a typicalpolarogram of the CO(III) complex (E4 = -0.35 V). The rate of decomposition ofthe peroxo complex was determined by plotting log iL (iL being the limiting currentof the polarographic wave) against time of exposure to oxygen atmosphere (fig. 5).2 0.4-I IIvoltage against s.c.e.F~G. 4.--D.m.e. polarograms of Co(rrr) + NH3 complex produced after hh (curve A), 71 h (curve B)from the decomposition of [(NH3)5C~-02-C~(NH3)51'1.'\I I ItlhFIG. 5.-The time dependence of the polarographic limiting current for the Co(m)+ aiiiinonia complex.The reaction is first order with respect to the Co(n11) complex concentration, with arate constant of 3.0 x In contrast to the rapid oxygenation reaction whichyields the peroxo complex, reactions (3) and (5) with k3 = 2.1 x lo4 dm3 mo1-l s-I,kV3 = 69 s-l and k s - 1 x lo3 dm3 inol-1 s - I , ~ the decomposition of the oxygenatedcomplex to yield Co(rr1) is a slow process.Therefore, the decomposition reactiondoes not interfere in the measurements related to equilibrium (3).s-lA . BETTELHEIM, M. FARAGGI, 1. HODARA AND J . MANASSEN 149CONCLUSIONThe use of the complex [(NH3)5Co-02-Co(NH3)5]+4 increases the limitingcurrent of oxygen reduction without showing any decrease in overpotential. Thisresult can be explained as an increasing effect of the oxygen dissolving capacity of theammoniacal Co(n) solution. A similar effect has been described by Dinkevich andKsenzhek l2 in the Co(11) + histidine system.R. G. Wilkins, Bioinorg. Chem., 1970, 100, 111.k. H. Vogt, H. M. Faigenbaum and S. E. Wiberly, Chem. Rev., 1963, 63, 269.A. G. Sykes and A. J. Weil, Inorganic Reaction Mechanisms, Progress in Inorganic Chemistry,ed. J. 0. Edwards (Interscience, N.Y. 1970), vol. 13, p. 1.E. Bayer and P. Schretzmann, Structure and Bonding, ed. C . K. Jsrgensen (Springer Verlag,Berlin, 1967), vol. 2, p. 181.J. Simplicio and R. G. Wilkins, J. Amer. Chem. Suc., 1969, 91, 1325.L. Meites and H. C. Thomas, Advanced Analytical Chemistry (McGraw Hill, N.Y., 1958), p. 34.A. Damjanovic, M. A. Genshaw and J. O’M. Bockris, J. Electrochem. Soc., 1967,114,466.H. R. Thirsk and J. A. Harrison, A Guide to the Study of Electrode Kinetics (Academic Press,London, 1972), p. 83.A. Bettelheim, M. Faraggi, I. Hodara and J. Manassen, J.C.S. Faraday I, 1977,73,150.(Amer. Chem. SOC., Wash., D.C., 1968), no. 83, p. 79.F. E. Dinkevich and 0. S. Ksenzhek, Trudy Ukrain. Respub. Konf. Elektrokinr., 1973, 1, 39.lo M. Anbar and E. J. Hart, Radiation Chemistry, Advances in Chemistry Series, ed. R. F. Gouldl 1 M. S. Michailidis and R. B. Martin, J. Amer. Chem. Soc., 1969, 91,4683.(PAPER 6/529