The application of physics-informed neural networks to hydrodynamic voltammetry

Area of Science:

• Electrochemistry
• Computational Science
• Chemical Engineering

Background:

• Flowing electrochemical systems enhance sensitivity for electro-analysis and enable steady-state measurements.
• Traditional simulation methods like finite difference or finite element can be complex to implement.

Purpose of the Study:

• To explore Physics-Informed Neural Networks (PINNs) as a simplified alternative for simulating electrochemical systems with flow.
• To predict the impact of flow dynamics and electrode geometry on observed currents in channel electrodes.

Main Methods:

• Utilized Physics-Informed Neural Networks (PINNs) for electrochemical simulations.
• Applied PINNs to model transport-limited currents, material transport between electrodes, and reactions preceding electrode reactions in channel flow.

Main Results:

• PINNs demonstrated quantitative agreement with existing solutions in known scenarios.
• The study successfully predicted the behavior of a complex preceding chemical reaction at a microchannel electrode, an unexplored area.

Conclusions:

• PINNs provide a viable, potentially simpler, and easier-to-implement alternative to traditional numerical methods for electrochemical flow problems.
• This approach shows promise for advancing fundamental studies and electro-analysis in flowing systems, including novel reaction mechanisms.