arXiv subjects

János Daru

Publications and source records attributed to János Daru.

4 recordsLinked to original sources

arXiv2026

Systematic incorporation of nuclear quantum effects into atomistic simulations by smoothed trajectory analysis

Nuclear quantum effects (NQEs) play an essential role in many atomistic systems, yet their explicit inclusion in molecular simulations remains challenging. Path-integral molecular dynamics (PIMD) provides a rigorous framework for incorporating NQEs, but its practical applicability is often limited by the slow and strongly system-dependent convergence with respect to the number of beads. Here we introduce path-integral generalized smoothed trajectory analysis (PIGSTA), a post-processing framework for the systematic incorporation of NQEs into atomistic simulations, using either classical or path-integral molecular dynamics trajectories. By applying analytically defined convolution kernels to simulation trajectories, PIGSTA corrects the frequency-dependent discretization error associated with a finite number of beads, without modifying the underlying dynamics. For harmonic systems, PIGSTA recovers the exact quantum-mechanical limit at any bead number, whereas standard PIMD becomes exact only in the infinite-bead limit. More generally, the method significantly improves the convergence of thermodynamic and structural observables at finite bead numbers and provides an internal, reference-free diagnostic of bead-number convergence based on the consistency of energy and force estimators. We assess PIGSTA for ambient liquid water and for the Zundel cation at ultralow temperature, representing a particularly demanding case for bead-number convergence. In both systems, PIGSTA reproduces the converged PIMD limit and enables physically consistent results at reduced bead numbers when convergence criteria are satisfied. Owing to its post-processing nature and negligible additional computational cost, PIGSTA offers a practical and broadly applicable approach for incorporating NQEs into atomistic simulations.

physics.chem-ph↗

arXiv2025

Short-range $Δ$-Machine Learning: A cost-efficient strategy to transfer chemical accuracy to condensed phase systems

DFT-based machine-learning potentials (MLPs) are now routinely trained for condensed-phase systems, but surpassing DFT accuracy remains challenging due to the cost or unavailability of periodic reference calculations. Our previous work (PRL 2022, 129, 226001) demonstrated that high-accuracy periodic MLPs can be trained within the CCMD framework using extended yet finite reference calculations. Here, we introduce short-range $Δ$-Machine Learning (sr$Δ$ML), which builds on periodic MLPs while accurately reproducing the observables of the high-level method.

physics.chem-ph↗

arXiv2024

Nuclear Quantum Effects in Liquid Water Are Negligible for Structure but Significant for Dynamics

Isotopic substitution, which can be realized both in experiment and computer simulations, is a direct approach to assess the role of nuclear quantum effects on the structure and dynamics of matter. Yet, the impact of nuclear quantum effects on the structure of liquid water as probed in experiment by comparing normal to heavy water has remained controversial. To settle this issue, we employ a highly accurate machine-learned high-dimensional neural network potential to perform converged coupled cluster-quality path integral simulations of liquid H$_2$O versus D$_2$O at ambient conditions. We find substantial H/D quantum effects on the rotational and translational dynamics of water, in close agreement with the experimental benchmarks. However, in stark contrast to the role for dynamics, H/D quantum effects turn out to be unexpectedly small, on the order of 1/1000 Å, on both intramolecular and H-bonding structure of water. The most probable structure of water remains nearly unaffected by nuclear quantum effects, but effects on fluctuations away from average are appreciable, rendering H$_2$O substantially more "liquid" than D$_2$O.

physics.chem-ph↗

arXiv2024

Random sampling versus active learning algorithms for machine learning potentials of quantum liquid water

Training accurate machine learning potentials requires electronic structure data comprehensively covering the configurational space of the system of interest. As the construction of this data is computationally demanding, many schemes for identifying the most important structures have been proposed. Here, we compare the performance of high-dimensional neural network potentials (HDNNPs) for quantum liquid water at ambient conditions trained to data sets constructed using random sampling as well as various flavors of active learning based on query by committee. Contrary to the common understanding of active learning, we find that for a given data set size, random sampling leads to smaller test errors for structures not included in the training process. In our analysis we show that this can be related to small energy offsets caused by a bias in structures added in active learning, which can be overcome by using instead energy correlations as an error measure that is invariant to such shifts. Still, all HDNNPs yield very similar and accurate structural properties of quantum liquid water, which demonstrates the robustness of the training procedure with respect to the training set construction algorithm even when trained to as few as 200 structures. However, we find that for active learning based on preliminary potentials, a reasonable initial data set is important to avoid an unnecessary extension of the covered configuration space to less relevant regions.

physics.chem-ph↗