arXiv subjects

Toon Verstraelen

Publications and source records attributed to Toon Verstraelen.

12 recordsLinked to original sources

arXiv2026

Reliable Viscosity Calculation from High-Pressure Equilibrium Molecular Dynamics: Case Study of 2,2,4-Trimethylhexane

Viscosity is a fundamental property of liquid lubricants, yet it is challenging to determine accurately, especially at high pressures. Although equilibrium molecular dynamics (EMD) simulations are a promising alternative to resource-intensive experiments, practical challenges remain in assessing the sufficiency of simulation time and in controlling uncertainties in the Green-Kubo formalism due to the finite amount of trajectory data. In this work, we extend the STable AutoCorrelation Integral Estimator (STACIE), a recently developed algorithm for estimating transport properties. First, we introduce the Lorentz model to estimate the viscosity and the exponential correlation time from the low-frequency power spectrum of deviatoric pressure fluctuations. Second, we show how to supplement the three conventional off-diagonal elements of the pressure tensor ($P_{xy}$, $P_{yz}$ and $P_{zx}$) with two additional uncorrelated deviatoric pressure components for shear viscosity calculations. Using these improvements, we apply STACIE to calculate the shear viscosity of 2,2,4-trimethylhexane from EMD simulations. We demonstrate STACIE's capability to reliably calculate viscosity under high-pressure conditions, offering a robust and automated solution with validated uncertainty quantification. Our results, when compared to the outcomes of the 10th International Fluid Properties Simulation Challenge, underscore the need for long EMD simulations. Large deviations from experimental viscosities in previous works were primarily due to insufficient simulation times and ad hoc post-processing choices, rather than the limitations of the force fields used. Unlike previous studies, our viscosity estimates agree well with experimental results (relative error < 6%) up to the highest pressure of 1 GPa, highlighting the improved reliability and accuracy of STACIE's systematic approach to viscosity predictions.

physics.comp-ph↗

arXiv2025

STable AutoCorrelation Integral Estimator (STACIE): Robust and accurate transport properties from molecular dynamics simulations

STACIE (STable AutoCorrelation Integral Estimator) is a novel algorithm and Python package that delivers robust, uncertainty-aware estimates of autocorrelation integrals from time-correlated data. While its primary application is deriving transport properties from equilibrium molecular dynamics simulations, STACIE is equally applicable to time-correlated data in other scientific fields. A key feature of STACIE is its ability to provide robust and accurate estimates without requiring manual adjustment of hyperparameters. Additionally, one can follow a simple protocol to prepare sufficient simulation data to achieve a desired relative error of the transport property. We demonstrate its application by estimating the ionic electrical conductivity of a NaCl-water electrolyte solution. We also present a massive synthetic benchmark dataset to rigorously validate STACIE, comprising 15360 sets of time-correlated inputs generated with diverse covariance kernels with known autocorrelation integrals. STACIE is open source and available on GitHub and PyPI, with comprehensive documentation and examples.

physics.comp-ph↗

arXiv2024

A Simple Molecular Model for Hydrated Silicate Ionic Liquids, a Realistic Zeolite Precursor

Despite the widespread use of zeolites in chemical industry, their formation process is not fully understood due to the complex and heterogeneous structure of traditional synthesis media. Hydrated silicate ionic liquids (HSILs) have been proposed as an alternative. They are truly homogeneous and transparent mixtures with low viscosity, facilitating experimental characterization. Interestingly, their homogeneous nature and simple speciation brings realistic molecular models of a zeolite growth liquid within reach for the first time. In this work, a simple molecular model is developed that gives insight into the crucial role of the alkali cations (sodium, potassium, rubidium and cesium). Thereby molecular dynamics simulations are combined with experimental measurements to demonstrate that the HSIL liquid structure strongly depends on the charge density and concentration of the alkali cation. As the water content increases, it transitions from a glassy network with fast ion exchange to an aqueous solution containing long-lasting, solvated ion pairs. Furthermore, simulations reveal that the cation is capable of bringing several silicate monomers together in the glassy network, displaying perfect orientations for condensation reactions that underlie zeolite formation. This work is an important step towards the development of molecular models that can fully describe the early nucleation process of zeolites in combination with experiments.

physics.chem-ph↗

arXiv2024

Managing Expectations and Imbalanced Training Data in Reactive Force Field Development: an Application to Water Adsorption on Alumina

ReaxFF is a computationally efficient model for reactive molecular dynamics simulations, which has been applied to a wide variety of chemical systems. When ReaxFF parameters are not yet available for a chemistry of interest, they must be (re)optimized, for which one defines a set of training data that the new ReaxFF parameters should reproduce. ReaxFF training sets typically contain diverse properties with different units, some of which are more abundant (by orders of magnitude) than others. To find the best parameters, one conventionally minimizes a weighted sum of squared errors over all data in the training set. One of the challenges in such numerical optimizations is to assign weights so that the optimized parameters represent a good compromise between all the requirements defined in the training set. This work introduces a new loss function, called Balanced Loss, and a workflow that replaces weight assignment with a more manageable procedure. The training data is divided into categories with corresponding "tolerances", i.e. acceptable root-mean-square errors for the categories, which define the expectations for the optimized ReaxFF parameters. Through the Log-Sum-Exp form of Balanced Loss, the parameter optimization is also a validation of one's expectations, providing meaningful feedback that can be used to reconfigure the tolerances if needed. The new methodology is demonstrated with a non-trivial parameterization of ReaxFF for water adsorption on alumina. This results in a new force field that reproduces both rare and frequent properties of a validation set not used for training. We also demonstrate the robustness of the new force field with a molecular dynamics simulation of water desorption from a $γ$-Al$_2$O$_3$ slab model.

physics.chem-ph↗

arXiv2023

Sensitivity analysis for ReaxFF reparameterization using the Hilbert-Schmidt independence criterion

We apply a global sensitivity method, the Hilbert-Schmidt independence criterion (HSIC), to the reparameterization of a Zn/S/H ReaxFF force field to identify the most appropriate parameters for reparameterization. Parameter selection remains a challenge in this context as high dimensional optimizations are prone to overfitting and take a long time, but selecting too few parameters leads to poor quality force fields. We show that the HSIC correctly and quickly identifies the most sensitive parameters, and that optimizations done using a small number of sensitive parameters outperform those done using a higher dimensional reasonable-user parameter selection. Optimizations using only sensitive parameters: 1) converge faster, 2) have loss values comparable to those found with the naive selection, 3) have similar accuracy in validation tests, and 4) do not suffer from problems of overfitting. We demonstrate that an HSIC global sensitivity is a cheap optimization pre-processing step that has both qualitative and quantitative benefits which can substantially simplify and speedup ReaxFF reparameterizations.

physics.chem-ph↗

arXiv2022

A new framework for frequency-dependent polarizable force fields

A frequency-dependent extension of the polarizable force field ``Atom-Condensed Kohn-Sham density functional theory approximated to the second-order'' (ACKS2) [J. Chem. Phys. 141, 194114 (2014)] is proposed, referred to as ACKS2$ω$. The method enables theoretical predictions of dynamical response properties of finite systems after a partitioning of the frequency-dependent molecular response function. Parameters in this model are computed simply as expectation values of an electronic wavefunction, and the hardness matrix is entirely reused from ACKS2 as an adiabatic approximation is used. A numerical validation shows that accurate models can already be obtained with atomic monopoles and dipoles. Absorption spectra of 42 organic and inorganic molecular monomers are evaluated using ACKS2$ω$, and our results agree well with the time-dependent DFT calculations. Also for the calculation of $C_6$ dispersion coefficients, ACKS2$ω$ closely reproduces its TDDFT reference. When parameters for ACKS2$ω$ are derived from a PBE/aug-cc-pVDZ ground state, it reproduces experimental values for 903 organic and inorganic intermolecular pairs with an MAPE of 3.84\%. Our results confirm that ACKS2$ω$ offers a solid connection between the quantum-mechanical description of frequency-dependent response and computationally efficient force-field models.

physics.chem-ph↗

arXiv2021

Modeling electronic response properties with an explicit-electron machine learning potential

Explicit-electron force fields introduce electrons or electron pairs as semi-classical particles in force fields or empirical potentials, which are suitable for molecular dynamics simulations. Even though semi-classical electrons are a drastic simplification compared to a quantum-mechanical electronic wavefunction, they still retain a relatively detailed electronic model compared to conventional polarizable and reactive force fields. The ability of explicit-electron models to describe chemical reactions and electronic response properties has already been demonstrated, yet the description of short-range interactions for a broad range of chemical systems remains challenging. In this work, we present the electron machine learning potential (eMLP), a new explicit electron force field where the short-range interactions are modeled with machine learning. The electron pair particles will be located at well-defined positions, derived from localized molecular orbitals or Wannier centers, naturally imposing the correct dielectric and piezoelectric behavior of the system. The eMLP is benchmarked on two newly constructed datasets: eQM7, a extension of the QM7 dataset for small molecules, and a dataset for the crystalline $β$-glycine. It is shown that the eMLP can predict dipole moments, polarizabilities and IR-spectra of unseen molecules with high precision. Furthermore, a variety of response properties, e.g. stiffness or piezoelectric constants, can be accurately reproduced.

physics.chem-ph↗

arXiv2021

Improving the Silicon Interactions of GFN-xTB

A general-purpose Density Functional Tight Binding method, the GFN-xTB model is gaining increased popularity in accurate simulations that are out of scope for conventional ab initio formalisms. We show that in its original GFN1-xTB parametrization, organosilicon compounds are described poorly. This issue is addressed by re-fitting the model's silicon parameters to a data set of ten thousand reference compounds, geometry-optimized with the revPBE functional. The resulting GFN1-xTB-Si parametrization shows improved accuracy in the prediction of system energies, nuclear forces and geometries and should be considered for all applications of the GFN-xTB Hamiltonian to systems that contain silicon.

physics.chem-ph↗

arXiv2021

ParAMS: Parameter Optimization for Atomistic and Molecular Simulations

This work introduces ParAMS -- a versatile Python package that aims to make parameterization workflows in computational chemistry and physics more accessible, transparent and reproducible. We demonstrate how ParAMS facilitates the parameter optimization for potential energy surface (PES) models, which can otherwise be a tedious specialist task. Because of the package's modular structure, various functionality can be easily combined to implement a diversity of parameter optimization protocols. For example, the choice of PES model and the parameter optimization algorithm can be selected independently. An illustration of ParAMS' strengths is provided in two case studies: i) a density functional-based tight binding (DFTB) repulsive potential for the inorganic ionic crystal ZnO, and ii) a ReaxFF force field for the simulation of organic disulfides.

physics.chem-ph↗

arXiv2021

Fanpy: A Python Library for Prototyping Multideterminant Methods in Ab Initio Quantum Chemistry

Fanpy is a free and open-source Python library for developing and testing multideterminant wavefunctions and related ab initio methods in electronic structure theory. The main use of Fanpy is to quickly prototype new methods by making it easier to transfer the mathematical conception of a new wavefunction ansätze to a working implementation. Fanpy uses the framework of our recently introduced Flexible Ansatz for N-electron Configuration Interaction (FANCI), where multideterminant wavefunctions are represented by their overlaps with Slater determinants of orthonormal spin-orbitals. In the simplest case, a new wavefunction ansatz can be implemented by simply writing a function for evaluating its overlap with an arbitrary Slater determinant. Fanpy is modular in both implementation and theory: the wavefunction model, the system's Hamiltonian, and the choice of objective function are all independent modules. This modular structure makes it easy for users to mix and match different methods and for developers to quickly try new ideas. Fanpy is written purely in Python with standard dependencies, making it accessible for most operating systems; it adheres to principles of modern software development, including comprehensive documentation, extensive testing, and continuous integration and delivery protocols. This article is considered to be the official release notes for the Fanpy library.

physics.chem-ph↗

arXiv2020

Accurate transferable polarization model derived from the monomer electron density

Force field have for decades proven to be an indispensable tool for molecular simulations which are out of reach for ab-initio methods. Recent efforts to improve the accuracy of these simulations have focused on the inclusion of many-body interactions in force fields. In this regard, we propose a transferable inducible dipole model which requires only the monomer electron density as input, without the need for atom type specific parameters. Slater dipoles are introduced, the widths of which are derived from the ab-initio monomer density. An additional exchange-repulsion interaction is introduced in our model, originating from the overlap of the delocalized dipoles with other dipoles and the ground state electron density. This interaction has previously been neglected in point dipole models, as the lack of spatial extent of the dipoles prevents the inclusion of an overlap term. The inclusion of this interaction is shown to significantly improve the prediction of three-body energies. Our model is incorporated in a previously proposed non-covalent force field and is benchmarked on interaction energies of dimers contained in the hsg and hbc6 datasets. Furthermore, we demonstrate the transferability of our model to the condensed phase of water, and to the interaction of CO$_2$ and H$_2$O molecules with the ZIF-8 metal-organic framework. The inherent transferability of our model makes it widely applicable to systems like the aforementioned metal-organic frameworks, where no specifically fitted parameters for polarization models are available in the literature.

physics.chem-ph↗

arXiv2016

Minimal Basis Iterative Stockholder: Atoms in Molecules for Force-Field Development

Atomic partial charges appear in the Coulomb term of many force-field models and can be derived from electronic structure calculations with a myriad of atoms-in-molecules (AIM) methods. More advanced models have also been proposed, using the distributed nature of the electron cloud and atomic multipoles. In this work, an electrostatic force field is defined through a concise approximation of the electron density, for which the Coulomb interaction is trivially evaluated. This approximate "pro-density" is expanded in a minimal basis of atom-centered s-type Slater density functions, whose parameters are optimized by minimizing the Kullback-Leibler divergence of the pro-density from a reference electron density, e.g. obtained from an electronic structure calculation. The proposed method, Minimal Basis Iterative Stockholder (MBIS), is a variant of the Hirshfeld AIM method but it can also be used as a density-fitting technique. An iterative algorithm to refine the pro-density is easily implemented with a linear-scaling computational cost, enabling applications to supramolecular systems. The benefits of the MBIS method are demonstrated with systematic applications to molecular databases and extended models of condensed phases. A comparison to 14 other AIM methods shows its effectiveness when modeling electrostatic interactions. MBIS is also suitable for rescaling atomic polarizabilities in the Tkatchenko-Sheffler scheme for dispersion interactions.

physics.chem-ph↗