arXiv subjects

Cesare Malosso

Publications and source records attributed to Cesare Malosso.

10 recordsLinked to original sources

arXiv2026

Transferable machine learning of excited-state dynamics with extremal pooling

Photochemical processes govern phenomena ranging from solar energy conversion and atmospheric chemistry to vision and photosynthesis. Accurate simulation of these processes requires modeling excited-state potential energy surfaces, often involving chemical reactions, tasks that remain computationally prohibitive for extended systems and long timescales using traditional \textit{ab initio} methods. Machine learning interatomic potentials have revolutionized ground-state simulations, but their extension to excited states faces fundamental challenges: standard architectures assume energy extensivity, an assumption that fails for excited states. Here, we present a size-intensive machine-learning framework for excited-state dynamics based on \textit{extremal pooling} of predicted atomic HOMO and LUMO contributions. Trained exclusively on excitations energies and forces, the architecture learns interpretable atomic-level contributions that encode physical information on the extent of electron localization. We demonstrate this framework on the photoexcited solvated electron in liquid water, a paradigmatic problem in radiation chemistry leading to competing pathways involving both hydrogen-atom dissociation and proton-coupled electron transfer. The model not only reproduces the relevant chain of reactions and product species that form during excitation, but also allows one to explicitly study the dynamics of the solvated electron in quantitative agreement with previously reported Restricted Open-Shell Kohn-Sham calculations, while enabling excited-state simulations of periodic systems at length and time scales inaccessible to the reference electronic-structure method. This work establishes a general strategy for machine learning-driven excited-state dynamics applicable to diverse photochemical systems, from molecular chromophores in solution to extended condensed-phase systems.

physics.chem-ph↗

arXiv2026

Beyond the Virial Expansion: Microscopic Origins of Partial Molar Volumes in LiCl Solutions

Although electrolyte density measurements have been reported for over a century, employing them to obtain accurate partial molar volume (PMV) profiles as a function of salt concentration has remained elusive. Obtaining such curves requires precise density measurements combined with a proper treatment of the associated virial expansion. In this work, we obtain PMV profiles for aqueous LiCl solutions. The resulting data enable the development of highly accurate force fields for Li$^+$ and Cl$^-$ ions, revealing a clear progression from isolated ions to ion pairs and ultimately to higher-order chain and ring structures. Because ion clustering emerges from complex, nonlocal interactions, it cannot be easily mapped onto specific virial terms. Instead, a direct structural and volumetric interpretation can be achieved by partitioning molecular dynamic (MD) simulation snapshots into three-dimensional polyhedral regions associated with individual salt ions and water molecules. The corresponding ionic and water volumes from this treatment quantitatively reproduce the experimental PMV curve. The results demonstrate that the PMV for salt increases (while that of water decreases) up to 6.7 M. Above this concentration, the direction reverses as three- and four-body interactions become prominent. Complementary multivariate curve resolution (MCR) Raman spectroscopy and density functional theory (DFT) calculations elucidate the molecular-level details of water electrostriction, which also persists up to 6.7 M. Significantly, the PMV data can be correlated with key thermodynamic properties, including the osmotic coefficient and the eutectic point. The procedures established here provide a general framework for modeling electrolyte solutions and enable the development of a new generation of accurate force fields for aqueous ions.

physics.chem-ph↗

arXiv2026

Comparing the latent features of universal machine-learning interatomic potentials

The past few years have seen the development of ``universal'' machine-learning interatomic potentials (uMLIPs) capable of approximating the ground-state potential energy surface across a wide range of chemical structures and compositions with reasonable accuracy. While these models differ in the architecture and the dataset used, they share the ability to compress a staggering amount of chemical information into descriptive latent features. Herein, we systematically analyze what the different uMLIPs have learned by quantitatively assessing the relative information content of their latent features with feature reconstruction errors, and observing how the trends are affected by the choice of training set and training protocol. We find that uMLIPs encode the chemical space in significantly distinct ways, with substantial cross-model feature reconstruction errors. When variants of the same model architecture are considered, trends become dependent on the dataset, target, and training protocol of choice. We also observe that fine-tuning of a uMLIP retains a strong pre-training bias in the latent features. Finally, we discuss how atom-level features, which are directly output by MLIPs, can be compressed into global structure-level features via concatenation of progressive cumulants, each adding significantly new information about the variability across the atomic environments within a given system.

physics.chem-ph↗

arXiv2026

High-quality, high-information datasets for universal atomistic machine learning

The quality, consistency, and information content of training data is often what determines the practical value of machine-learning models for atomistic simulations. Yet, many widely used electronic-structure databases are assembled having materials screening as primary goal rather than robust force-field learning, are limited in their scope to a specific class of chemical compounds, and/or employ inconsistent DFT functionals and settings. Here we introduce MAD-1.6, a highly curated dataset designed explicitly for training broadly applicable atomistic models across the periodic table at high levels of theory. MAD-1.6 extends the MAD dataset with targeted enrichment strategies that improve the coverage of chemical space to 102 elements while keeping the total number of configurations compact. All structures are computed with a single, standardized all-electron DFT workflow using the r$^2$SCAN meta-GGA functional and consistent convergence settings, ensuring uniformity across chemically heterogeneous systems. The dataset encompasses molecules, clusters, bulk crystals, surfaces, and low-dimensional structures, and its quality and consistency are further enhanced by outlier removal using uncertainty quantification. We demonstrate the high accuracy that can be achieved with the proposed dataset by training PET-MAD-1.6, a generally applicable r$^2$SCAN interatomic potential that covers 102 elements in the periodic table and achieves exceptional levels of benchmark accuracy and stability in challenging simulation protocols.

cond-mat.mtrl-sci↗

arXiv2026

The Photochemical Birth of the Hydrated Electron in Liquid Water

The photophysics and photochemistry associated with irradiating UV light in liquid water is central to numerous physical, chemical and biological processes. One of the key events involved in this process is the generation of the hydrated electron. Despite long study from both experimental and theoretical fronts, a unified understanding of the underlying mechanisms associated with the generation of the solvated electron have remained elusive. Here, using excited-state molecular dynamics simulations of condensed phase photoexcited liquid water, we unravel the key sequence of chemical events leading to the creation of the hydrated electron on the excited state. The process begins through the excitation localized mostly on specific topological defects in the hydrogen-bond network of water which is subsequently followed by two main reaction pathways. The first, leads to the creation of a hydrogen atom culminating in non-radiative decay back to the ground-state within 100 femtoseconds. The second involves a proton coupled electron transfer, giving rise to the formation of the hydronium ion, hydroxyl radical and the hydrated excess electron on the excited-state. This process is facilitated by ultrafast coupled rotational and translational motions of water molecules leading to the formation of water mediated ion-radical pairs in the network. These species can survive on the picosecond timescale and ultimately modulate the emission of visible photons. All in all, our findings provide fresh perspectives into the interpretation of several independent time-dependent spectroscopies measured over the last decades, paving the way for new directions on both theoretical and experimental fronts.

physics.chem-ph↗

arXiv2025

Dynamical Heterogeneity in Supercooled Water and its Spectroscopic Fingerprints

A growing body of theoretical and experimental evidence strongly supports the existence of a second liquid-liquid critical point (LLCP) in deeply supercooled water leading to the co-existence of two phases: a high-and low-density liquid (HDL and LDL). While the thermodynamics associated with this putative LLCP has been well characterised through numerical simulations, the dynamical properties of these two phases close to the critical point remain much less understood. In this work, we investigate their dynamical and spectroscopic features using machine-learning interatomic potentials (MLIPs). Dynamical analyses using the van-Hove correlation function, reveal that LDL exhibits very sluggish and heterogeneous molecular mobility, in contrast to the faster and more homogeneous dynamics of HDL. Infrared absorption (IR) spectra further show clear vibrational distinctions between LDL and HDL, in particular in the far IR region between 400 - 1000 cm-1. Together, these findings provide new dynamical fingerprints that clarify the microscopic behavior of supercooled water and offer valuable guidance for experimental efforts aimed at detecting the long-sought liquid-liquid transition.

cond-mat.soft↗

arXiv2024

Evidence of ferroelectric features in low-density supercooled water from ab initio deep neural-network simulations

Over the last decade, an increasing body of evidence has emerged, supporting the existence of a metastable liquid-liquid critical point in supercooled water, whereby two distinct liquid phases of different densities coexist. Analysing long molecular dynamics simulations performed using deep neural-network force fields trained to accurate quantum mechanical data, we demonstrate that the low-density liquid phase displays a strong propensity toward spontaneous polarization, as witnessed by large and long-lived collective dipole fluctuations. Our findings suggest that the dynamical stability of the low-density phase, and hence the transition from high-density to low-density liquid, is triggered by a collective process involving an accumulation of rotational angular jumps, which could ignite large dipole fluctuations. This dynamical transition involves subtle changes in the electronic polarizability of water molecules which affects their rotational mobility within the two phases. These findings hold the potential for catalyzing new activity in the search for dielectric-based probes of the putative second critical point.

cond-mat.soft↗

arXiv2024

Aqueous Solution Chemistry In Silico and the Role of Data Driven Approaches

The use of computer simulations to study the properties of aqueous systems is, today more than ever, an active area of research. In this context, during the last decade there has been a tremendous growth in the use of data-driven approaches to develop more accurate potentials for water as well as to characterize its complexity in chemical and biological contexts. We highlight the progress, giving a historical context, on the path to the development of many-body and reactive potentials to model aqueous chemistry, including the role of machine learning strategies. We focus specifically on conceptual and methodological challenges along the way in performing simulations that seek to tackle problems in modeling the chemistry of aqueous solutions. In conclusion, we summarize our perspectives on the use and integration of advanced data-science techniques to provide chemical insights in physical chemistry and how this will influence computer simulations of aqueous systems in the future.

physics.chem-ph↗

arXiv2023

Quadriexciton binding energy in electron-hole bilayers

Excitonic condensation and superfluidity have recently received a renewed attention, due to the fabrication of bilayer systems in which electrons and hole are spatially separated and form stable pairs known as indirect excitons. Dichalcogenides- and graphene- based bilayers are nowadays built and investigated, giving access to systems with (i) only spin degeneracy, (ii) spin and valley degeneracy. Simulation studies performed in the last decade at $T=0$ for simple, model electron-hole bilayers, as function of inter-layer distance and in-layer carrier density, have revealed in case (i) the formation of biexcitons in a tiny region of parameter space and in case (ii) the formation of stable compounds made of 4 electrons and 4 holes (quadriexcitons) in a sizable region of parameter space. Of some interest is the relation of the properties of isolated biexcitons (quadriexcitons) and those of their finite density counterpart. In fact, the isolated biexciton has been repeatedly studied in the last years with simulations and other techniques. No simulations, instead, are available to our knowledge for the isolated quadriexciton, for which we present here results of the first quantum Monte Carlo (QMC) study. Stability with respect to the dissociation into biexcitons, and the pair correlations with varying the inter-layer distance $d$ are discussed.

cond-mat.str-el↗

arXiv2022

Viscosity in water from first-principles and deep-neural-network simulations

We report on an extensive study of the viscosity of liquid water at near-ambient conditions, performed within the Green-Kubo theory of linear response and equilibrium ab initio molecular dynamics (AIMD), based on density-functional theory (DFT). In order to cope with the long simulation times necessary to achieve an acceptable statistical accuracy, our ab initio approach is enhanced with deep-neural-network potentials (NNP). This approach is first validated against AIMD results, obtained by using the Perdew-Burke-Ernzerhof (PBE) exchange-correlation functional and paying careful attention to crucial, yet often overlooked, aspects of the statistical data analysis. Then, we train a second NNP to a dataset generated from the Strongly Constrained and Appropriately Normed (SCAN) functional. Once the error resulting from the imperfect prediction of the melting line is offset by referring the simulated temperature to the theoretical melting one, our SCAN predictions of the shear viscosity of water are in very good agreement with experiments.

cond-mat.mtrl-sci↗