Communications Chemistry最新文献

Mn-site doping in spinel LiMn₂O₄ (LMO) mitigates Mn³⁺-induced Jahn-Teller distortion. However, this strategy faces inherent trade-offs. Specifically, low-valent doping weakens oxygen bonding, while high-valent doping increases Mn³⁺ content. To overcome these limitations, this work proposes dual-lanthanide (La³⁺/Ce³⁺) co-doping. Through sol-gel synthesis, LiLa_0.1Ce_0.1Mn_1.8O₄ (LLCMO) achieves synergistic performance enhancements. Particularly, La reduces Mn³⁺ content to 43.13%, suppressing lattice distortion and widening Li+ diffusion pathways via its large ionic radius. Concurrently, Ce (in a mixed Ce³⁺/Ce⁴⁺ state) enhances charge delocalization, lowering electron transfer barriers and boosting conductivity. Critically, La-Ce cooperation mitigates Mn dissolution while stabilizing the spinel framework. Consequently, LLCMO exhibits a 3.2-fold higher Li+ diffusion coefficient than pristine LMO. Furthermore, it delivers 111.2 mAh g^-1 at 0.5 C with 90.9% retention after 100 cycles, and remarkably retains 76.0 mAh g^-1 after 1000 cycles even at 10 C. Thus, this dual-doping strategy establishes a generalizable design principle for enhancing stability/kinetics in diverse cathodes via a synergistic division-of-labor mechanism.

Efficient photocatalytic hydrogen production from water splitting remains a critical challenge for sustainable energy solutions. Here, we report dual Ni(OH)₂ (NO)/Ni_xP (NP) cocatalysts photo-deposited and ZnS decorated ZnIn₂S₄ photocatalyst (NOP/ZIS-Z). It exhibits efficient photocatalytic hydrogen production (PHP) with the rate of 5.46 mmol·g^-1·h^-1 and 420 nm quantum yield of 55.2% in triethanolamine (TEOA) system upon Xe lamp visible light combined with excellent stability. Impressively, its PHP rate reaches 0.54 mmol·g^-1·h^-1 in pure water under natural sunlight, showing tremendous practical potentials. The synergistic mechanism among ZnS, NO, and NP was revealed: (i) ZnS could transfer electron from ZIS and facilitates charge carrier separation, (ii) NP acts as reduction cocatalysts for proton reduction, (iii) NO functions as oxidation cocatalysts to trap holes for sacrificial reagent/water oxidation. Our work highlights coinstantaneous enhancing photocatalytic reduction and oxidation half-reaction by loading dual cocatalysts onto the heterojunctions.

Atomistic simulations are essential in chemistry and materials science but remain challenging to run due to the expert knowledge required for the setup, execution, and validation stages of these calculations. We present ChemGraph, an agentic framework powered by artificial intelligence and state-of-the-art simulation tools to streamline and automate computational chemistry and materials science workflows. ChemGraph leverages graph neural network-based foundation models for accurate yet computationally efficient calculations and large language models (LLMs) for natural language understanding, task planning, and scientific reasoning to provide an intuitive and interactive interface. We evaluate ChemGraph across 13 benchmark tasks and demonstrate that smaller LLMs (GPT-4o-mini, Claude-3.5-haiku, Qwen-2.5-14B) perform well on simple workflows, while more complex tasks benefit from using larger models. Importantly, we show that decomposing complex tasks into smaller subtasks through a multi-agent framework enables GPT-4o to reach perfect accuracy and smaller LLMs to match or exceed single-agent GPT-4o's performance in these benchmarks.

Ordered membrane nanodomains colloquially known as "lipid rafts" have many proposed cellular functions. However, pharmacological tools to modulate protein affinity for rafts and to manipulate raft formation are currently lacking. We screened 24,000 small molecules for compounds that impact the raft affinity of a known raft-preferring model protein, peripheral myelin protein 22 (PMP22), in giant plasma membrane vesicles (GPMVs). Hits were tested against another model raft protein, MAL, and also tested for their impact on raft stability. We identified three chemically distinct tools for manipulating lipid rafts. Two compounds were found to destabilize ordered domains (VU0607402 and VU0519975) while a third (primaquine diphosphate) increased PMP22 partitioning and stabilized ordered domains. While discovered in a PMP22-focused screen, all three were seen to modulate raft formation in a protein-independent manner by altering lipid-lipid interactions and membrane fluidity. Acute treatment of live cells with the raft destabilizing compound, VU0607402 was seen to modulate TRPM8 channel function, highlighting the utility of this compound in live-cell experiments for dissecting the role that membrane order and fluidity play in cell signaling. These compounds provide pharmacological tools for probing lipid raft properties and function in biophysical experiments and in living cells.

Deep learning models have become fundamental tools in drug design. In particular, large language models trained on biochemical sequences learn feature vectors that guide drug discovery through virtual screening. However, such models do not capture the molecular interactions important for binding affinity and specificity. Therefore, there is a need to merge representations from distinct biological modalities to effectively represent molecular complexes. We present an overview of the methods to combine molecular representations and propose that future work should develop biochemical foundation models that jointly encode diverse molecular modalities. Specifically, learning to merge the representations from internal layers of domain specific biological language models could improve generalizability in the context of interaction prediction. We demonstrate that 'composing' biochemical language models performs similar or better than standard methods representing molecular interactions despite having significantly fewer features. We also discuss recent methods for interpreting and democratizing large language models that could aid the development of interaction aware foundation models for biology. Finally, we present a vision for future research that allows for predicting the evolution of molecular interactions across biophysical contexts.

Photothermal therapy (PTT) has emerged as a promising strategy for treating solid tumors and topical infections by converting the incident light energy into localized heat using photothermal agents. Among these, gold nanoparticles (GNPs) are particularly attractive due to their strong surface plasmon resonance, tunable surface chemistry, biocompatibility and scalability. However, their limited biodegradability and inefficient clearance remain significant translational challenges. In this study, we have developed gold-coated calcium peroxide nanoparticles (CPAu-NPs) that offer dual advantages, enhanced photothermal conversion and intrinsic reactive oxygen species generation. The self-release of oxygen and hydrogen peroxide from CPAu-NPs addresses tumor hypoxia, a key barrier to effective therapy. To further augment therapeutic efficacy, we incorporated Sorafenib, a multi-kinase inhibitor known to induce ferroptosis and inhibit tumor progression in melanoma, a cancer type marked by dysregulated iron metabolism and vulnerability to ferroptosis. This combinatorial approach disrupts critical survival pathways while promoting lipid peroxidation, potentially overcoming resistance to standard treatments. Additionally, we explored the antifungal potential of this system, recognizing the increased susceptibility of immunocompromised cancer patients to fungal infections. Our results suggest that CPAu-NPs, in combination with Sorafenib, provide a multifunctional theranostic platform capable of targeting melanoma cells, modulating the tumor microenvironment, and addressing opportunistic fungal infections.

The presence of amino acids in comets and meteorites has long suggested that prebiotic molecules may have formed in space and contributed to the origins of life on Earth. Glycine, the simplest amino acid, has been identified in several extraterrestrial environments, although its detection in the interstellar medium, including prestellar cores and protostellar regions, remains elusive. Here, we investigate a novel catalytic pathway for glycine formation on silicate grains during relatively warm (> 150 K) stages of star formation. Using atomistic simulations, the feasibility of a Strecker-type synthesis and a direct neutral mechanism involving reactivity between formaldehyde, carbon monoxide and ammonia on forsterite surfaces, the major constituent of interstellar dust, is assessed. Results show that the Strecker pathway is limited by high activation barriers, whereas the proposed direct mechanism proceeds through low-energy surface-stabilized intermediates leading to spontaneous formation of glycine in a single-barrier exoergic process. Additionally, glycine strongly adsorbs onto the mineral surface and is unlikely to desorb under warm conditions. A vibrational analysis reveals that glycine formed through this pathway exhibits spectrally distinct features, including suppression and shifting of characteristic bands, which may account for its persistent non-detection in astronomical observations.

Accurate modeling of conical intersections is crucial in nonadiabatic molecular dynamics, as these features govern processes such as radiationless transitions and photochemical reactions. Conventional electronic structure methods, including Hartree-Fock, density functional theory, and their time-dependent extensions, struggle in this regime. Due to their single reference nature and separate treatment of ground and excited states, they fail to capture ground state intersections. Multiconfigurational approaches overcome these limitations, but at a prohibitive computational cost. In this work, we propose a modified Hartree-Fock framework, referred to as Convex Hartree-Fock, that optimizes the reference within a tailored subspace by removing projections along selected Hessian eigenvectors. The ground and excited states are then obtained through subsequent Hamiltonian diagonalization. We validate the approach across several test cases and benchmark its performance against time-dependent Hartree-Fock within the Tamm-Dancoff approximation.

Machine learning (ML) and artificial intelligence (AI) techniques are transforming the way chemical reactions are studied today. Datasets from high-throughput experimentation (HTE) are generated to better understand the reaction conditions crucial for outcomes such as yields and selectivities. However, it is often overlooked that datasets from such designed experiments possess a specific structure, which can be captured by a statistical model. Ignoring these data structures when applying ML/AI algorithms can result in misleading conclusions. In contrast, leveraging knowledge about the data-generating process yields reliable, interpretable, and comprehensive insights into reaction mechanisms. A particularly complex dataset is available for the Buchwald-Hartwig amination. Using this dataset, a statistical model for such HTE-generated chemical data is introduced, and a parameter estimation algorithm is developed. Based on the estimated model, new insights into the Buchwald-Hartwig amination are discussed. Our approach is applicable to a wide range of HTE-generated data for chemical reactions and beyond.