In recent years, significant progress has been made in applying Artificial intelligence approaches, such as Machine learning and Bayesian inference, to obtain mechanistic information in voltammetric studies. Many methods developed have been applied satisfactorily to noisy synthetic data, but with limited success in analysing experimental data. The experimental data is often difficult to fit due to non-uniqueness of the fitted simulations or deviations of the experimental data from the theoretical model.<br/>So far, the most widely used data analysis algorithms used in voltammetry employ general optimisation methods and Bayesian Inference. However, it is difficult to use these approaches to automatically identify probable mechanisms and quantify all parameters involved. Recent research by the Monash Electrochemistry Group, has found that the weakness of each individual algorithm could be compensated by using a chain of algorithms. Consequently, the recently developed Artificial intelligence platform offers more accurate identification of experimental data than an individual algorithm. In particular, it has been found that a combination of these algorithms with a multivariate Deep neural network for general classification, an unsupervised clustering algorithm for feature extraction of training data, a Deep neural network for classifying the featured data and a Greedy algorithm for parameter selection has led to a more robust method for qualitative and quantitative analysis of experimental data. In this talk, this Artificial intelligence platform will be presented and possible implications for voltammetric data analysis using Artificial intelligence will be described with implications for analytical electrochemistry.

Development of an artificial intelligence platform for voltammetric mechanistic studies

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.