Convergence of stochastic-extended Lagrangian molecular dynamics method for polarizable force field simulation

Dong An,Sara Y. Cheng,Teresa Head-Gordon,Lin Lin,Jianfeng Lu

DOI: https://doi.org/10.1016/j.jcp.2021.110338

IF: 4.645

2021-08-01

Journal of Computational Physics

Abstract:<p>Extended Lagrangian molecular dynamics (XLMD) is a general method for performing molecular dynamics simulations using quantum and classical many-body potentials. Recently several new XLMD schemes have been proposed and tested on several classes of many-body polarization models such as induced dipoles or Drude charges, by creating an auxiliary set of these same degrees of freedom that are reversibly integrated through time. This gives rise to a singularly perturbed Hamiltonian system that provides a good approximation to the time evolution of the real mutual polarization field. To further improve upon the accuracy of the XLMD dynamics in the context of classical polarizable force field simulation, and to potentially extend it to other many-body potentials, we introduce a stochastic modification which leads to a set of singularly perturbed Langevin equations with degenerate noise. We prove that the resulting Stochastic-XLMD converges to the accurate dynamics, and the convergence rate is both sharp and is independent of the accuracy of the initial polarization field. We carefully study the scaling of the damping factor and numerical noise for efficient numerical simulation for Stochastic-XLMD, and we demonstrate the effectiveness of the method for water molecules described by a polarizable force field.</p>

physics, mathematical,computer science, interdisciplinary applications

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the convergence analysis of the extended Lagrangian molecular dynamics method (XLMD) in classical polarization force - field simulations. Specifically, the authors introduce an extended Lagrangian molecular dynamics method with a stochastic heat bath (Stochastic - XLMD) to improve the accuracy and stability of the traditional XLMD method when dealing with polarization force fields. By introducing an annealing term and noise, Stochastic - XLMD can more accurately simulate the dynamic behavior of atoms without depending on the initial conditions.

### Background and Motivation

In molecular dynamics simulations, especially in classical molecular dynamics using polarization force fields, a linear or nonlinear system needs to be repeatedly solved to determine certain hidden variables, such as induced dipoles or Drude charges. The traditional extended Lagrangian molecular dynamics method (XLMD) improves computational efficiency by introducing auxiliary variables to avoid self - consistent iteration at each step. However, this method still has the problem of insufficient accuracy in some cases, especially when the initial conditions are inaccurate.

### Solution

To overcome these limitations, the authors propose the Stochastic - XLMD method. This method makes the system stable over a long period of time and can converge to the correct dynamic behavior more quickly by introducing an annealing term and noise. Specifically, the equations of the Stochastic - XLMD method can be expressed as:

\[
\begin{aligned}
\ddot{r}_\xi &= F(r_\xi)-\frac{\partial Q}{\partial r}(r_\xi, x_\xi), \\
\epsilon \ddot{x}_\xi &= -\frac{\partial Q}{\partial x}(r_\xi, x_\xi)-\sqrt{\epsilon}\gamma \dot{x}_\xi+\sqrt{2\gamma T \epsilon^{1/4}}\dot{W},
\end{aligned}
\]

where:
- \(r_\xi\) represents the atomic position,
- \(x_\xi\) represents the hidden variable (such as the induced dipole),
- \(\epsilon\) is a small parameter that controls the fast and slow time scales of the system,
- \(\gamma\) is the artificial friction parameter,
- \(T\) is the artificial temperature of the hidden variable,
- \(\dot{W}\) is white noise.

### Main Contributions

1. **Convergence Proof**: The authors prove that in the polarization force - field model, when \(\epsilon\) and \(T\) approach 0, the atomic dynamics of the Stochastic - XLMD method converges to the exact dynamics. Specifically, they prove the following error bound:

\[
E\left\{\sup_{0\leq t\leq t_f}|r_\xi(t)-r^\star(t)|\vee|p_\xi(t)-p^\star(t)|\right\}\leq C\left(\epsilon^{1/2}+\epsilon^{1/4}T^{1/2}+T\right),
\]

where \(C\) is a constant independent of \(\epsilon\) and \(T\).

2. **Independence from Initial Conditions**: An important feature of the Stochastic - XLMD method is that its convergence rate does not depend on the initial conditions of the hidden variables. This is because the introduced annealing term and noise can make the system reach an equilibrium state within a finite time, thereby reducing the influence of the initial error.

3. **Numerical Verification**: The authors verify the validity of the theoretical results through numerical experiments, including simulations of water molecules and some simple model systems. The experimental results show that the Stochastic - XLMD method has high accuracy and stability in practical applications.

### Conclusion

This paper solves the problems of the traditional XLMD method in dealing with polarization force fields by introducing the Stochastic - XLMD method.