A fast and rigorous numerical tool to measure length-scale artifacts in molecular simulations
Benedikt M. Reible, Nils Liebreich, Carsten Hartmann, Luigi Delle Site · 2025
Read A fast and rigorous numerical tool to measure length-scale artifacts in molecular simulations by Benedikt M. Reible, Nils Liebreich, Carsten Hartmann, Luigi Delle Site (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-01442.txt.
Introduction
Benedikt Reible, Nils Liebreich, Carsten Hartmann, and Luigi Delle Site build a fast, exact tool for spotting when a molecular simulation box is too small. Splitting a modeled system into independent subsystems changes its free energy, and the two sided Bogoliubov inequality bounds that cost exactly, for both classical and quantum many body systems.
That theorem yields a quality factor measuring how consistent a given box size is. For systems with two body interactions and a known radial distribution function, the factor needs only two six dimensional integrals. The authors present an algorithm for computing it. Its results agree with published findings from simulations run at different box sizes. That gives simulators a solid check for length scale errors.
Catalog record
| CREATOR | Benedikt M. Reible, Nils Liebreich, Carsten Hartmann, Luigi Delle Site |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 63 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-01442
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A fast and rigorous numerical tool to measure length-scale artifacts in molecular simulations Benedikt M. Reible, Nils Liebreich, Carsten Hartmann, Luigi Delle Site Abstract The two-sided Bogoliubov inequality for classical and quantum many-body systems is a theorem that provides rigorous bounds on the free-energy cost of partitioning a given system into two or more independent subsystems. This theorem motivates the definition of a quality factor which directly quantifies the degree of statistical-mechanical consistency achieved by a given simulation box size. A major technical merit of the theorem is that, for systems with two-body interactions and a known radial distribution function, the quality factor can be computed by evaluating just two six-dimensional integrals. In this work, we present a numerical algorithm for computing the quality factor and demonstrate its consistency with respect to results in the literature obtained from simulations performed at different box sizes. I introduction The method of molecular simulation has undoubtedly been highly successful in the study of complex molecular systems Frenkel and Smit 2002 ; Tuckerman 2023 , yet some fundamental questions remain open. For both technical and conceptual reasons, the optimal choice of the system’s size is a major concern in any simulation: it should be large enough in order for the computational representation of the system to reflect physical reality closely, but also small enough to avoid high computational costs associated with large simulations. The inability of a simulation to capture the key physical features of a realistic system fully due to a limited system size is referred to as finite-size effects Salacuse et al. 1996 . In first instance, the use of periodic boundary conditions in molecular simulations alleviates, in part, the problem of finite-size effects. However, if the size of the unit cell is not sufficient to represent the essential features of the bulk of a substance, then its numerical representation as a collection of copies of the unit cell interacting with each other may even amplify the artificial character of the results: since an individual cell does not faithfully represent the local features of the true systems, also the interaction between different cells is not realistic even at a larger scale beyond the unit cell. A discussion of the methods and techniques for handling the problem of finite-size effects in the field of molecular simulation can be found in Ref. Reible et al. 2025a and the references therein. In the present work, we will treat an alternative approach to standard techniques. This method has been developed by some of us in recent years and is based on first principles of statistical mechanics and, in particular, the free energy F as the central quantity. The latter forms the bridge between the microscopic particle ensemble and macroscopic observables. Specifically, it regulates the system’s behavior and drives the first-principle derivation of any thermodynamic property ( Landau and Lifshitz 1980 , p. 48) , ( Huang 1991 , pp. 22 f.) . The corresponding method for determining the optimal size of a simulation box is based on computing upper and lower bounds for the free energy cost \Delta F associated with the separation of a large system into two (or more) independent subsystems, and it is expressed in a rigorous theorem, the two-sided Bogoliubov inequality Delle Site et al. 2017 ; Reible et al. 2022 . The quantity \Delta F corresponds to the interface energy when an ideal surface divides the system into two independent parts, and hence the crucial observation is the following: if the interface energy can be neglected compared to some reference energy of the system (e.g., the total potential energy), then it follows that the smaller subsystem still captures the features of the bulk of the substance, thus the size of the original system is certainly sufficient for a satisfactory representation of the bulk. Studies of prototype systems such as interacting quantum gases have shown the validity of the approach Reible et al. 2023 ; Delle Site and Hartmann 2024a . For systems characterized by two-body potentials and documented radial distribution functions (e.g., from numerical data), calculating the upper and lower bound for the interface energy \Delta F is enormously simplified as this task reduces to the straightforward numerical evaluation of six dimensional integrals. In this study, we will implement the numerical procedure for the calculation of such integrals and apply it to systems of Lennard-Jones particles. Such systems have been treated in the literature, and their finite size-effects have been assessed by expensive simulations performed at different sizes of the simulation box. We will show that the results of our approach lead to the same conclusions as those based on the simulation study; such a validation qualifies our method as numerically efficient and physically rigorous. The paper is organized as it follows. In Section II we will introduce the relevant theoretical background on the two-sided Bogoliubov inequality and the finite-size effects criterion based on it. Section III will discuss the special case of systems with two-body potentials and the corresponding simplifications in the criterion; in particular, the integrals that have to be evaluated for it will be given. In Section IV , we shall introduce four different numerical methods for evaluating these integrals. Finally, in Section V we will discuss a particular physical system from the literature on which we will test our finite-size effects criterion, showing various numerical data obtained via the four integration methods to substantiate its effectiveness. II Two-sided Bogoliubov inequality and quality factor II.1 Two-sided Bogoliubov inequality The two-sided Bogoliubov inequality gives an upper and lower bound for the interface free energy \Delta F , that is, the cost of partitioning a system of particles into two (or more) non-interacting subsystems; for simplicity, we only discuss the case of two subsystems which is the most relevant one. Specifically, we consider M particles confined to a spatial region \Omega\subset\mathbb{R}^{3} , described by a probability density function f (or: density operator in the quantum-mechanical case). Suppose that \Omega is divided into two disjoint subregions \Omega_{1},\Omega_{2}\subset\Omega , with s and M-s particles and probability densities f_{1},f_{2} , respectively. Furthermore, assume that the full system is described by a Hamiltonian (function in the classical case; operator in the quantum case) of the form H=H_{0}+U , where H_{0}=H_{1}+H_{2} is the Hamiltonian for the two independent subsystems, and U governs the interaction between \Omega_{1} and \Omega_{2} . In thermal equilibrium at inverse temperature \beta , the full system is described by the density f=Z^{-1}\,\mathrm{e}^{-\beta H} with Z=\int_{\Omega_{1}}\int_{\Omega_{2}}\mathrm{e}^{-\beta H}\mathop{}\!\mathrm{d}\mathbf{r}^{\prime}\mathop{}\!\mathrm{d}\mathbf{r} (trace in the quantum case), and the two independent subsystems are described by the joint probability density f_{0}=f_{1}\cdot f_{2} (tensor product in the quantum case), where f_{i}=Z_{i}^{-1}\,\mathrm{e}^{-\beta H_{i}} with Z_{i}=\int_{\Omega_{i}}\mathrm{e}^{-\beta H_{i}}\mathop{}\!\mathrm{d}\mathbf{r} , i\in\{{1,2}\} . The interface free energy \Delta F is now defined as the relative free energy between f and f_{0} : \Delta F\coloneqq-\beta^{-1}\log\left(\frac{Z}{Z_{0}}\right)\,. Computing \Delta F by traditional free energy calculation methods such as thermodynamic perturbation or particle insertion can be cumbersome, which warrants computationally efficient, yet precise estimates of \Delta F . An upper and a lower bound for \Delta F is expressed by the following theorem, proved for classical systems in Delle Site et al. 2017 and for quantum systems in Reible et al. 2022 . Theorem 1 (Two-sided Bogoliubov inequality)