Removing 65 Years of Approximation in Rotating Ring Disk Electrode Theory with Physics-Informed Neural Networks

Area of Science:

• Electrochemistry
• Catalysis
• Material Science

Background:

• The Rotating Ring Disk Electrode (RRDE) is a crucial tool in electrochemistry, catalysis, and material science.
• Collection efficiency ($$) is a key parameter derived from RRDE measurements, offering insights into reaction mechanisms and kinetics.
• Accurate theoretical prediction of $$ is challenging due to complex mass transport equations and approximations in existing models.

Purpose of the Study:

• To develop a novel method for accurately predicting the collection efficiency ($$) of RRDEs.
• To address the limitations of conventional approximations in calculating $$.
• To investigate the impact of edge effects and velocity profiles on $$.

Main Methods:

• Employed a physics-informed neural network (PINN) to solve the complete convective diffusion mass transport equation.
• Validated the PINN approach against theoretical predictions and experimental data.
• Compared PINN results with traditional analytical models.

Main Results:

• The PINN accurately predicts $$ by solving the full convective diffusion equation.
• Identified and quantified the influence of previously neglected edge effects and velocity corrections on $$.
• Provided a guideline for the applicability of conventional approximations in $$ calculations.

Conclusions:

• Physics-informed neural networks offer a superior approach to calculating RRDE collection efficiency compared to traditional methods.
• Understanding edge effects and velocity corrections is crucial for precise electrochemical analysis.
• This work advances the theoretical foundation for RRDE applications in various scientific fields.