Inverse design of drying-induced assembly of multicomponent colloidal-particle films using surrogate models

Area of Science:

• Materials Science
• Soft Matter Physics
• Computational Science

Background:

• Colloidal films assembled by drying are difficult to engineer due to sensitivity to particle properties and processing conditions.
• Developing predictive models for colloidal assembly is crucial for materials design.

Purpose of the Study:

• To develop and test an inverse-design strategy using surrogate modeling for engineering colloidal film structures.
• To identify optimal particle and processing parameters for achieving target film morphologies.

Main Methods:

• Simulated drying of a two-component hard-sphere colloidal suspension using Brownian dynamics.
• Employed surrogate models, including Gaussian Process Regression (GPR) and Chebyshev polynomial interpolation, trained on a loss function derived from simulated film structures.
• Utilized the trained surrogate models for approximation and optimization tasks.

Main Results:

• Surrogate models proved effective for approximating film structures and optimizing design parameters with limited data.
• Gaussian Process Regression models generally offered higher accuracy compared to polynomial interpolants.
• Polynomial interpolants provided a more computationally efficient alternative.

Conclusions:

• The developed inverse-design strategy is a powerful tool for engineering colloidal materials with targeted structures.
• This approach offers a generalizable framework for designing nonequilibrium assembly processes.
• The findings have broad implications for the fabrication of advanced functional materials from colloidal suspensions.