Limitations in Electrochemical Determination of Mass-Transport Parameters: Implications for Quantification of Electrode Kinetics Using Data Optimisation Methods

Abstract

Voltammetric quantification of the electrode kinetics for the quasi-reversible reaction requires detailed experiment–theory comparisons. Ideally, predicted data derived from the theoretical model are fitted to the experimental data by adjusting the reversible potential (E0), heterogeneous electron transfer rate constant at E0 (k0), and charge transfer coefficient α, with mass-transport and other parameters exactly known. However, parameters relevant to mass transport that include electrode area (A), diffusion coefficient (D), and concentration (c), are usually subject to some uncertainty. Herein, we examine the consequences of having different combinations of errors present in A, D, and c in the estimation of E0, k0, and α on the basis of the a.c. (alternating current) voltammetric experiment–theory comparisons facilitated by the use of a computer-assisted parameter optimisation algorithm. In most cases, experimentally reasonable errors (<10 %) in the mass-transport parameters do not introduce significant errors in recovered E0, k0, and α values. However, a pernicious situation may emerge when a slight overestimation of A, D or c is included in the model and results in erroneous identification of a reversible redox process as a quasi-reversible one with a report of apparently quantifiable kinetic parameters k0 and α.

Integration of Heuristic and Automated Parametrization of Three Unresolved Two‐Electron Surface‐Confined Polyoxometalate Reduction Processes by AC Voltammetry

“…), then the significance of the outcome of a multi-parameter fitting exercise is likely to remain equivocal. [6,[19][20][21] Furthermore, a range of other difficulties can be encountered, and thus data optimisation methods are far from flawless, and still require heuristic input, as will become apparent in the present study.…”

“…In particular, if the data set available is inadequate as often is the case when using DC cyclic voltammetry (insufficient data points or range of scan rates etc. ), then the significance of the outcome of a multi‐parameter fitting exercise is likely to remain equivocal ,. Furthermore, a range of other difficulties can be encountered, and thus data optimisation methods are far from flawless, and still require heuristic input, as will become apparent in the present study.…”

A Comparison of Bayesian Inference Strategies for Parameterisation of Large Amplitude AC Voltammetry Derived from Total Current and Fourier Transformed Versions

“…Use of the Nimrod toolkit also has been extended to other mechanisms with caution advised in multi parameter parameterisation to the significant possibility of over fitting the simulation. [39][40][41][42] The present work expands the range of AC data formats that can be used to present experimental data and subjected to data optimisation analysis. The use of Bayesian framework provides access to uncertainties in parameter values reported.…”

“…using the Nimrod toolkit [38] to compare experimental and simulated AC voltammetric data. Use of the Nimrod toolkit also has been extended to other mechanisms with caution advised in multi parameter parameterisation to the significant possibility of over fitting the simulation [39–42] . The present work expands the range of AC data formats that can be used to present experimental data and subjected to data optimisation analysis.…”

“…The use of heuristic method also enables the parameter search range used in the automated parameterisation process to be restricted, hence save computing time. [27,39,44] Accordingly, parameterisation of the FTAC voltammetric data was first conducted heuristically on one of the ten data sets. The results obtained are summarised in Table 5 and the details of the analysis are given below.…”