Themed collection Most popular 2024 physical chemistry articles

A deep understanding of electrode–electrolyte interfaces requires the development of modelling protocols spanning from the local microscale to system-level macroscopic sizes which can be validated by comparison with high-quality experimental results.

In this perspective, the progress in ion-regulated organic RTP materials and described the roles of ions, including ion–π interactions, electrostatic interactions, and coordinate interactions, have been summarized.

Each step in model development affects its accuracy.

We review the current capabilities of local electron correlation methods up to the gold standard CCSD(T) level, which are now capable of routinely reaching large systems of practical interest in p- and d-block, as well as surface and biochemistry.

AI and automation are revolutionizing catalyst discovery, shifting from manual methods to high-throughput digital approaches, enhanced by large language models.

Self-interaction error leads to runaway error accumulation when density functional theory is used in conjunction with the many-body expansion.

This study delves into the magnetic response of core electrons and their influence on the global magnetic response of planar and three-dimensional systems containing heavy elements, employing the removing valence electron (RVE) approximation.

A generalized and extensible machine-learned molecular mechanics force field trained on over 1.1 million QC data applicable for drug discovery applications. Figure reproduced from the arXiv:201001196 preprint under the arXiv non-exclusive license.

Fast room-temperature phosphorescence with high quantum yields up to 38% in solution from metal-free organic 1,2-diketones is reported, along with the mechanism and molecular design principles governing the fast phosphorescence.

Extracting knowledge from complex chemical texts is essential for both experimental and computational chemists. Fine-tuned large language models (LLMs) can serve as flexible and effective extractors for automated data acquisition.

nach0 is a novel multi-domain and multi-task language model pre-trained on unlabeled text from scientific literature, patents, and molecule strings to incorporate a range of chemical and linguistic knowledge.

DiffBindFR, a diffusion model based flexible full-atom protein–ligand docking tool, demonstrates its superior docking and side-chain refinement accuracy with reliable physical plausibility.

Autonomous process optimization (APO) is a technology that has recently found utility in a multitude of process optimization challenges.

This work outlines a simple synthetic strategy providing formylated stiff-stilbenes. The photoisomerization of formylated stiff-stilbenes could be fully controlled using visible light and is accompanied by a high photostationary state distribution.

Detailed DFT and ab initio calculations unveil the correlation between spin–phonon vibrations and blocking temperature to provide design clues to improve single-ion magnet characteristics in the pseudo-D_5h family of Dy(III) SMMs.

Simulations reveal that both charge and exciton delocalisation can significantly improve the efficiency of charge generation in organic photovoltaics and explain the failure of classical hopping approaches.

This study incorporates Minnesota density functionals into the current knowledge of describing the structural and dynamical properties of liquid water through ab initio molecular dynamics, with M06-2X(-D3) functionals showing the most promise.

The origin of the electrophilicity of a series of cyclohexanones and benzaldehydes is investigated using the activation strain model and quantitative Kohn–Sham molecular orbital (MO) theory.

The air–water interface is not the site for H₂O₂(aq) formation; instead, it takes place at the solid–water interface.

Despite theoretical difficulties, we herein for the first time demonstrate an effective concept for the synthesis of an orange-red multiple resonance (MR) emitter centered on a pyridine ring via stereo effects.

PoseBusters assesses molecular poses using steric and energetic criteria. We find that classical protein-ligand docking tools currently still outperform deep learning-based methods.

Hydrogen atom transfer (HAT) reactions, as they occur in many biological systems, are here predicted by machine learning.

By using resonance Raman spectroscopy and temperature-dependent pulse EPR spectroscopy, we show that bond stretching vibrational modes > 200 cm⁻¹ drive spin relaxation in planar and ruffled copper porphyrins.

An artificial molecular design workflow using a genetic algorithm with molecular strings and artificial neural networks applied to organic molecules with inverted singlet-triplet gaps finds 1000 candidates with appreciable fluorescence rates.

A machine learning model, trained on a large COSMO-RS dataset, enables accurate and rapid predictions of solvation effects on reaction rates for diverse reactions and solvents only based on atom-mapped reaction SMILES and solvent SMILES.

This study presents a modular autonomous workflow for solid-state chemistry comprising three separate robots, allowing automated powder X-ray diffraction to underpin crystalline materials discovery.

Here we propose CarsiDock, a deep learning-guided docking approach that leverages large-scale pre-training of millions of docking complexes for protein–ligand binding pose generation.

Symmetry breaking charge transfer followed by charge separation with appreciable lifetimes has been demonstrated in a newly synthesized bisstyrylBODIPY dimer in polar solvents using pump-probe and other techniques.