We report the static and dynamical properties of liquid water at the level of second-order Møller-Plesset perturbation theory (MP2) with classical and quantum nuclear dynamics using a neural network potential. We examined the temperature-dependent radial distribution functions, diffusion, and vibrational dynamics. MP2 theory predicts over-structured liquid water as well as a lower diffusion coefficient at ambient conditions compared to experiments, which may be attributed to the incomplete basis set. A better agreement with experimental structural properties and the diffusion constant are observed at an elevated temperature of 340 K from our simulations. Although the high-level electronic structure calculations are expensive, training a neural network potential requires only a few thousand frames. This approach shows great potential, requiring modest human effort, and is straightforwardly extensible to other simple liquids.

The fixation of ${\rm CO}_2$ is a promising carbon-neutral approach and drawing lots of attention in the past decades. Among lots of fixation of ${\rm CO}_2$ methods, the cycloaddition of ${\rm CO}_2$ into epoxides (CCE) reaction is important because it could generate high value-added products. Recently, employing chalcogen bonding (ChB) catalysts for the CCE reaction has been proposed and offers significant advantages of low price, environmental friendliness, and ease of recycling. In this work, the ChB catalyzed the CCE reaction is investigated by high level theoretical calculations. The reaction can be divided into three subprocesses: ring-opening, nucleophilic addition, and formation of cyclic carbonate. Both Se- and Te-based ChB catalysts have shown the potential for catalyzing the CCE reaction, the overall catalytic performance of Te-based catalysts was superior to the Se-based ones. Overall, the most positive electrostatic potentials of the ChB catalysts, electrostatic term of the ChB binding energies, electron density at the ChB critical points, and electron density difference play important roles in the ChB catalyzed CCE reaction. This work elucidated the ChB catalyzed mechanism of the CCE reaction and provided theoretical guidance for the future development of efficient ChB catalysts for the CCE reaction.

Density functional theory (DFT) calculations were performed to elucidate the reaction mechanism of the Pd-catalyzed carbonylation of propargylic alcohol (1), leading to the efficient synthesis of cyclohexyl $α$-methylene-$β$-lactone (2). Our study revealed that the reaction proceeds through a four-step pathway: alkyne migration and insertion, CO insertion, HCl-assisted hydrogen transfer, and final C–O annulation. Notably, the final C–O annulation step was identified as the rate-determining step (RDS) of the overall catalysis, with a free energy barrier of 25.7 kcal/mol (i.e., $\mathbf{IM4}→\mathbf{TS4}).$ Additionally, we uncovered the critical role of the HCl during the reaction pathway, a demonstrating that it acts as a co-catalyst, proton shuttle, and hydrogen bond donor/acceptor. NBO, EDA-NOCV, and HIGM analyses further revealed that the remarkable stability of the transition state $\mathbf{TS3}$ in the presence of HCl primarily arises from strong electrostatic attraction and orbital interaction energies between the two interacting fragments. These mechanistic insights provide valuable insight and guidance for the rational design of new Pd-catalyzed transformations.

The integration of both rigid and flexible components holds great potential to significantly enhance the overall performance of organic electronic devices. Non-covalent interactions are frequently harnessed to augment the planar conjugation of polymers, consequently elevating the rigidity of these polymers. However, the influence of the dihedral angle distortion between donor and acceptor units, which is induced by the flexibility inherent in donor-acceptor copolymers, on charge transport remains poorly understood. In this study, we systematically investigate intra-chain charge transport parameters and charge mobility for the 3,6-bis(thiophen-2-yl) diketopyrrolopyrrole (DPPT) conjugated with various donor moieties. Combining with density functional theory (DFT) and the Su-Schrieffer-Heeger (SSH) model, we find that when the non-covalent interactions between the donor and acceptor units are enhanced (as exemplified by DPPT-FT and DPPT-BO), the coupling between electrons and low-frequency vibrations is significantly suppressed. Simultaneously, the intra-chain electronic coupling increases owing to substantial orbital overlap. Surface hopping simulations are utilized to study the charge transport properties. For DPPT-T, DPPT-FT, and DPPT-BT, weaker molecular rigidity and disordered chain packing lead to thermally activated hopping transport (low electronic coupling and high reorganization energy). In contrast, the enhanced structural rigidity of DPPT-BO facilitates charge delocalization, leading to an initial improvement in carrier mobility under low-temperature conditions, and thermal fluctuation effects induce a band-like behavior at high temperature.

Neural network quantum states represent a powerful approach for solving electronic structures in strongly correlated molecular and material systems. For a neural network ansatz to be accurate, it must effectively learn the phase of a complex wave function. In this work, we demonstrate several different network structures as the phase network for a Transformer-based neural network quantum state implementation. We compare the accuracy of ground state energies, the number of parameters, and computational time across several small molecules. Furthermore, we propose a phase network setup that combines cross attention and multilayer perceptron structures, with the number of parameters remaining constant across different systems. Such an architecture may help reduce computational costs and enable transfer learning to larger quantum chemical systems.

With the rapid development of A-D-A non-fullerene acceptors, organic solar cells (OSCs) have made significant progress. However, compared to inorganic and perovskite solar cells, the energy loss in OSCs remains relatively large. In this Perspective, we summarize our recent computational advances in elucidating the mechanisms of energy loss in OSCs at the molecular level. We highlight strategies to minimize voltage loss during charge generation, suppress triplet-channel recombination, and reduce non-radiative voltage loss by modulating both molecular and aggregation structures.

Simulating non-Markovian quantum dissipative dynamics remains a major challenge in theoretical and computational chemistry. While traditional numerical methods such as hierarchical equations of motion are numerically exact, they suffer from prohibitive computational costs when modeling systems with complex environmental couplings or strong non-Markovian effects. To address this limitation, we propose a deep learning framework based on two-dimensional convolutional neural networks (2D-CNN) for efficiently predicting long-time quantum dissipative dynamics using only short-time trajectory data. Our approach incorporates a 1D-to-2D feature reconstruction strategy, which transforms 1D time-series data into 2D images, and a multi-timescale fusion network to resolve complex dynamical features. We validate the framework on two paradigmatic cases — dissipative relaxation in a two-level system and Rabi oscillations in a dissipative spin system — achieving prediction mean absolute errors of $10^{-3}$ and $10^{-2},$ respectively. The results highlight the effectiveness of our 2D-CNN approach in capturing long-time temporal correlations, providing a computationally efficient pathway for simulating quantum dynamics in realistic open systems.

This work employed DFT calculations to elucidate the mechanism of water oxidation reaction catalyzed by a mononuclear pyridine-based copper complex, which was reported to be a homogeneous water oxidation catalyst in a pH=11.84 buffer solution. The coordination of one water molecule with the ${\rm Cu}^{{\rm II}}$ center leads to the generation of the ${\rm Cu}^{{\rm II}}-{\rm OH}_2 (1-{\rm H}_2{\rm O}).$ The active species $({\rm Cu}^{{\rm IV}}={\rm O, 3})$ is generated after two subsequent proton-coupled electron transfer processes from $1-{\rm H}_2{\rm O}.$ $3$ triggers the O-O bond formation via water nucleophilic attack mechanism. The triplet ${\rm O}_2$ can be released after following two oxidation processes. The formation of the O-O bond is the rate-determining step for the catalytic cycle associated with a total barrier of 19.3 kcal/mol.

Machine learning force fields (MLFFs) offer a promising balance between quantum mechanical (QM) accuracy and molecular mechanics efficiency. While MLFFs have shown strong performance in modeling short-range interactions and reproducing potential energy surfaces, their ability to capture long-range cooperative effects remains underexplored. In this study, we assess the ability of three MLFF models — ANI, MACE-OFF, and Orb — to reproduce cooperative interactions arising from environmental induction and dispersion, which are essential for many biomolecular processes. Using a recently proposed framework, we quantify hydrogen bond (H-bond) cooperativity in N-methylacetamide polymers. Our results show that all MLFFs capture cooperativity to some extent, with MACE-OFF yielding the closest agreement with QM data. These findings highlight the importance of evaluating many-body effects in MLFFs and suggest that H-bond cooperativity can serve as a useful benchmark for improving their physical fidelity.

A simple model based on the two-electron two-orbital textbook problem is presented and used to analyze pairwise interatomic interactions in metal-ligand bonding. In particular the two types of covalency discussed during the last decade for actinide-ligand interactions, overlap/interaction driven and energy-near-degeneracy driven covalency, as well as their influence on the bond strengths and interatomic charge build-up are discussed. The hydration complexes ${\rm M}({\rm H}_2{\rm O})^{4+}_n$ of selected tetravalent lanthanide and actinide ions are used to probe the performance of the model for an analysis of calculations as well as for predictions.