Abstract:We present novel non-parametric representation math for local pseudopotentials (PP) based on Gaussian Process Regression (GPR). Local pseudopotentials are needed for materials simulations using Orbital-Free Density Functional Theory (OF-DFT) to reduce computational cost and to allow kinetic energy functional (KEF) application only to the valence density. Moreover, local PPs are important for the development of accurate KEFs for OF-DFT as they are only available for a limited number of elements. We optimize local PPs of tin (Sn) using GP regression to reproduce the experimental lattice constants of {\alpha}- and \b{eta}-Sn, the energy difference between these two phases as well as their electronic structure and charge density distributions, which are obtained with Kohn-Sham Density Functional Theory employing semi-local PPs. The use of a non-parametric GPR-based PP representation avoids difficulties associated with the use of parametrized functions and has the potential to construct an optimal local PP independent of prior assumptions. The GPR-based Sn local PP results in well-reproduced bulk properties of {\alpha}- and \b{eta}-tin, and electronic valence densities similar to those obtained with semi-local PP.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to construct accurate tin (Sn) local pseudopotentials (LPP) in material simulations, especially when using Orbital - Free Density Functional Theory (OF - DFT). Specifically, the goals of the paper are: 1. **Develop a new non - parametric representation method**: Based on Gaussian Process Regression (GPR), propose a new non - parametric local pseudopotential representation method. This method can avoid the difficulties brought by using parametric functions and has the potential to construct optimal local pseudopotentials without relying on prior assumptions. 2. **Optimize the local pseudopotential of tin**: Optimize the local pseudopotential of tin through GPR so that it can reproduce the experimental lattice constants of α - tin and β - tin, the energy difference between the two phases, as well as their electronic structures and charge density distributions. These properties are obtained through Kohn - Sham Density Functional Theory (KS - DFT) using semi - local pseudopotentials. 3. **Improve the accuracy of OF - DFT**: Through the optimized local pseudopotential, improve the accuracy of OF - DFT in simulating the physical and chemical properties of tin and its alloys, especially when dealing with large - scale systems. This helps to accelerate further technological progress based on microscopic and mesoscopic tin systems. 4. **Solve the limitations of existing methods**: Existing local pseudopotential methods usually rely on parametric functions, which may limit the realization of the best local pseudopotentials. In addition, most modern pseudopotentials are designed for KS - DFT and are not suitable for OF - DFT. Therefore, the paper aims to overcome these limitations through machine - learning methods and construct accurate local pseudopotentials suitable for OF - DFT. In summary, this paper aims to improve the accuracy and efficiency of OF - DFT in simulating tin and its alloys by introducing the GPR method to optimize the local pseudopotential of tin, thereby promoting research progress in the field of materials science.

Non-parametric Local Pseudopotentials with Machine Learning: a Tin Pseudopotential Built Using Gaussian Process Regression

Local Volume Effects in the Generalized Pseudopotential Theory

Modeling the high-pressure solid and liquid phases of tin from deep potentials with ab initio accuracy

First-principle optimal local pseudopotentials construction via optimized effective potential method

Greater Transferability and Accuracy of Norm-conserving Pseudopotentials using Nonlinear Core Corrections

In Situ Pseudopotentials for Electronic Structure Theory

High-Quality Local Pseudopotentials for Metals

Neural network potential for studying the thermal conductivity of Sn

A machine-learned kinetic energy model for light weight metals and compounds of group III-V elements

Nonlocal vs Local Pseudopotentials Affect Kinetic Energy Kernels in Orbital-Free DFT

Sparse Gaussian Process Potentials: Application to Lithium Diffusivity in Superionic Conducting Solid Electrolytes

Transferable local pseudopotentials for magnesium, aluminum and silicon

Polyvalent Machine-Learned Potential for Cobalt: from Bulk to Nanoparticles

Efficient non-parametric fitting of potential energy surfaces for polyatomic molecules with Gaussian processes

Development of a machine learning finite-range nonlocal density functional

Suitability of available interatomic potentials for Sn to model its 2D allotropes

Locality Error Free Effective Core Potentials for 3d Transition Metal Elements Developed for the Diffusion Monte Carlo Method

Workhorse semilocal density functional for condensed matter physics and quantum chemistry

Accelerating finite-temperature Kohn-Sham density functional theory with deep neural networks

Enhancing High-Fidelity Neural Network Potentials through Low-Fidelity Sampling