Machine learning many-body potentials for colloidal systems

Abstract

Simulations of colloidal suspensions consisting of mesoscopic particles and smaller species such as ions or depletants are computationally challenging as different length and time scales are involved. Here, we introduce a machine learning (ML) approach in which the degrees of freedom of the microscopic species are integrated out and the mesoscopic particles interact with effective many-body potentials, which we fit as a function of all colloid coordinates with a set of symmetry functions. We apply this approach to a colloid-polymer mixture. Remarkably, the ML potentials can be assumed to be effectively state-independent and can be used in direct-coexistence simulations. We show that our ML method reduces the computational cost by several orders of magnitude compared to a numerical evaluation and accurately describes the phase behavior and structure, even for state points where the effective potential is largely determined by many-body contributions.

References in corpus (6)

• Physics-inspired structural representations for molecules and materials
• Beware of density dependent pair potentials
• Capillary Waves in a Colloid-Polymer Interface
• The fluid-fluid interface in a model colloid-polymer mixture: Application of grand canonical Monte Carlo to asymmetric binary mixtures
• Critical behavior in colloid-polymer mixtures: theory and simulation
• Modeling of many-body interactions between elastic spheres through symmetry functions

Cited by in corpus (13)

• Neural functional theory for inhomogeneous fluids: Fundamentals and applications
• Soft Metamaterials: Adaptation and Intelligence
• Perspective: How to overcome dynamical density functional theory
• Machine-learning effective many-body potentials for anisotropic particles using orientation-dependent symmetry functions
• Why neural functionals suit statistical mechanics
• Hyper-density functional theory of soft matter
• A deep learning approach to the measurement of long-lived memory kernels from Generalised Langevin Dynamics
• Active crystallization from power functional theory
• A neural network potential with self-trained atomic fingerprints: a test with the mW water potential
• Machine learning many-body potentials for charged colloids in primitive 1:1 electrolytes
• Diagrammatics for the Inverse Problem in Spin Systems and Simple Liquids
• Determining the chemical potential via universal density functional learning
• Machine-Learned Many-Body Potentials for Charged Colloids reveal Gas-Liquid Spinodal Instabilities only in the strong-coupling regime of Primitive Models