No abstract

A mechanism is proposed that reproduces in simulations the oscillations seen during the methylene-blue-catalyzed reduction of 02 by HS" in a continuous-flow stirred tank reactor (CSTR). It contains 14 reactions and is based on experiments and simulations of simpler reactions, including the reduction of MB+ by HS" in the presence and absence of 02 and the reactions of H202 and 02 with HS". All experiments on component reactions as well as the CSTR oscillations can be simulated by the same set of reactions and rate constants. The major dynamic feature of the mechanism is competition for MB' by the oxidizing agents 02 and H202 and the reducing agents HS" and HS*.

plexed to either the solvent or the electrode surface (or both). There is no evidence from the PMV data to suggest that surface effects are responsible for the long lifetime of the intermediate. It is, however, possible that the estimated lifetime is that of the nitrilium ion formed by reaction of acetonitrile with the diphenylmethyl cation provided that the nitrilium ion is electroactive (presumably by dissociative reduction to form the radical and acetonitrile).It is possible to estimate the potential for the dissociative reduction of the nitrilium ion. Steenken and McClelland showed that the formation of this species is reversible.* Under conditions where it is formed almost exclusively, the major product is actually diphenylmethanol with the Ritter product always a minor component of the mixture. The equilibrium constant for this process must be >1000,35 giving a free energy change of <-4 kcal mol-' (<0.18 eV). Saveant and co-workers have suggested that the overpotential for a dissociative electron-transfer reaction will be essentially equivalent to the enthalpy change for the bond dissociation of the bond in the oxidized form of the electroactive specie^.'^ Using this reasoning, the reduction of the nitrilium ion is expected near 0.17 V vs SCE. A mechanism consistent with all of the experimental data is shown in Scheme I. (35) Steenken, S., private communication. (36) Saveant, J. M. J . Am. Chem. Soc. 1987, 109, 6788. ConclusionsImplicit finitedifference simulation of photomodulation voltammograms can be used to obtain lifetimes of radicals and product ions as well as values for the heterogeneous electron-transfer rate and Eo for redox reactions of radicals. Our experimental setup restricts lifetime measurement to an upper limit of ca. 300 ps (due to the limiting time response of the electrochemical cell). The effects of changes in the kinetic parameters can be summarized as follows:(1) At high potentials (above the diffusion limit) the phase of the electrochemical signal depends only on the lifetime of the radical.(2) Changes in the lifetime of the ion affect the quadrature signals more than those signals that are in phase with the excitation lamp. The effect of changes in ion lifetime, however, is to shift the position of E,/* but not to change the shape of the wave (Le., the breadth of the wave does not change).(3) Decreases in the heterogeneous electron-transfer rate (k,) increase the breadth and ElFor the oxidation of the dipienylmethyl radical the data suggest that nitrilium ion, which is formed in a reaction between the carbocation and acetonitrile, will either undergo a dissociative electron transfer to form the radical or slowly react to form products with a rate that depends on the concentration of water in the solvent system.of the voltammetric waves.The dynamical behavior of the methylene blue-oxygen-sulfide oscillator in a continuous flow stirred tank reactor is investigated. At low flow rates oscillations of period one emerge at a supercritical Hopf bifurcation, whereas a subcritical Hopf bifur...

Oscillations are observed in a CSTR during the methylene blue catalyzed oxidation of HS" by 02 at natural pH (=12). Oscillations in the concentration of the oxidized form of methylene blue can be observed spectrophotometrically, and oscillations in [02] can be observed amperometrically. Oscillations appear in a relatively narrow but significant range of flow rates and are inhibited by H202, presumably a product of the reduction of 02 by HS". A mechanism is proposed based on radical-chain processes and using inhibition as well as autocatalysis as the nonlinearities leading to oscillation. Removal of 02 by HS' is catalyzed by methylene blue, which is not itself reduced by HS" until almost all 02 is removed. The reduction of 02 leads 1 to H202 whose concentration may be high when the methylene blue is finally reduced. The oxidation of the reduced form of methylene blue by 02 is inhibited by H202 or some other product of the reduction of HS" by 02 or MB"1". Thus methylene blue remains in its reduced state until this inhibiting species is removed either by the flow or by reaction, probably with HS", or until the flow increases [02] to the point where a possibly autocatalytic radical-chain oxidation of the reduced form of methylene blue may begin. When methylene blue is reoxidized, 02 has been replenished, and the cycle repeats. Experiments also were carried out in which the oscillator was perturbed by varying the CSTR pumping rate sinusoidally. In all cases the response of the oscillator to perturbation was periodic or quasi-periodic. Regions of entrainment separated by regions of quasi-periodicity were found. AH peaks in the Fourier spectra of sinusoidally perturbed oscillators showing quasi-periodicity could be identified as linear combinations of the perturbing frequency and the natural frequency of the oscillator.

No abstract