Browse Articles

Severe adverse clinical impacts are predicted by an early high positive fluid balance in patients with severe acute pancreatitis

Activation of methane by U+ studied by guided ion beam tandem mass spectrometry and quantum chemistry

Reaction pathways of all products formed in the U+ + CH4 (CD4) reaction were explored as a function of kinetic energy using guided ion beam tandem mass spectrometry and quantum chemical calculations. UH+, UC+, UCH+, UCH2+, and UCH3+ (and their perdeuterated analogues) are formed in endothermic reactions. In both systems, the UCH2+ (UCD2+) dehydrogenated product was the dominant product in the low-energy region, whereas the UH+ (UD+) hydride product became predominant at high energies. The kinetic energy behavior of the various products is consistent with a common intermediate of H–U+–CH3 (D–U+–CD3). The kinetic energy dependence of all product cross sections was modeled to obtain experimental bond dissociation energies at 0 K (in eV): D0 (U+–H) = 2.42 ± 0.10, D0 (U+–C) = 3.95 ± 0.12, D0 (U+–CH) = 4.91 ± 0.09, D0 (U+–CH2) = 4.11 ± 0.04, and D0 (U+–CH3) = 2.41 ± 0.09. Quantum chemical calculations using the UCCSD(T) and UB3LYP approaches with the cc-pwCVXZ-PP basis set with MDF-60 pseudopotential for U+ and the aug-cc-pCVXZ and aug-cc-pVXZ (X = T, Q) basis set for carbon and hydrogen, respectively, validate the experimental bond dissociation energies and outline the potential energy surface for all reactions observed. In addition, spin–orbit corrections of the bond energies for all products were calculated at a CASSCF-CASPT2-RASSI level.

Design of a hybrid quantum machine learning architecture and analysis of quantum noise effects

A combined experimental and theoretical study on electron induced fragmentation of methyl acetate, a model compound for side chain fragmentation and decarboxylation as pathways to main chain scission of polymethyl methacrylate as EUV lithography resist material

With extreme ultraviolet lithography (EUVL) being established alongside the conventional deep UVL in high-volume semiconductor manufacturing processes comes a transition from the use of non-ionizing radiation to the use of ionizing radiation in the lithographic process. Correspondingly, the chemistry in the solubility switching of the resist materials changes from photochemistry to electron-induced chemistry and may thus resemble the resist chemistry induced in electron beam lithography rather than the photochemistry governing deep ultraviolet lithography. This in turn calls for rethinking of the resist formulations, a better understanding of the respective electron induced chemistry, and eventually tailoring it to provide high performance EUVL formulations. In the current study, we take a step in this direction and revisit electron induced fragmentation of methyl acetate as the simplest model compound for the functional side group of polymethyl methacrylate (PMMA), a high-performance main chain scission resist material in electron beam lithography. Appearance energies for individual fragmentation reactions in dissociative ionization (DI) in the gas phase are determined, and quantum chemical calculations are conducted to elucidate the underlying reactions. The results are discussed in context to previous work on dissociative ionization and dissociative electron attachment (DEA) of methyl acetate, and quantum chemical calculations are used to explore the thermo-chemistry of decarboxylation as a path to main chain scission of PMMA through both DI and DEA when this resist material is exposed to EUV radiation.

Hierarchical multi-agent reinforcement learning for retrieval-augmented industrial document question answering

Abstract Multimodal industrial documents–such as operation manuals, circuit diagrams, and parameter tables–contain domain knowledge distributed across text, images, and document layout. However, most existing retrieval-augmented generation (RAG) frameworks rely on static retrieval and fusion policies with fixed modality weights and uniform retrieval depth, making them less adaptable to diverse query intents and dynamic cross-modal dependencies. As a result, they often retrieve incomplete evidence and yield suboptimal reasoning in complex long-document scenarios. To address these challenges, we propose MARL-RAGDoc, a hierarchical multi-agent reinforcement learning framework for multimodal retrieval-augmented reasoning. A high-level coordinator agent dynamically allocates modality weights and retrieval depth based on query characteristics, while specialized text, image, and table agents perform fine-grained evidence selection within their respective candidate pools. A collaborative reasoning module integrates the retrieved evidence and provides hierarchical reward signals to continuously optimize retrieval policies. Experimental results on multiple multimodal document benchmarks demonstrate that MARL-RAGDoc consistently outperforms baselines in both retrieval accuracy and reasoning performance, while remaining computationally efficient. Our code and dataset are publicly available at  https://github.com/Yihong-Q/MARL-RAGDoc .

Interchain coupling and vibrational mode analysis of polytetrafluoroethylene using machine-learned potentials

We investigate the temperature-dependent vibrational properties of crystalline polytetrafluoroethylene (PTFE) using molecular dynamics simulations powered by a neural-network potential that explicitly incorporates long-range van der Waals (vdW) interactions. Our simulations reveal a systematic red shift in three vibrational bands (800–700, 680–640, and 385–360 cm−1) as temperature increases. To elucidate the microscopic origin of these shifts, we perform phonon calculations under distinct structural scenarios, including helical unwinding, helix reversal defects, and controlled expansion of in-plane lattice constants. Only the increase in interchain distance reproduces the observed shifts, indicating that these modes are sensitive to intermolecular coupling. Eigenmode analysis shows that these redshifting bands are dominated by symmetric CF2 stretching motions with transverse fluorine displacements, which directly modulate interchain separation. In contrast, torsional, bending, and asymmetric stretching modes exhibit negligible frequency changes. These findings demonstrate that a specific class of vibrational modes serves as a microscopic probe of intermolecular interactions in PTFE and underscore the importance of incorporating long-range vdW corrections into machine-learned potentials for accurate vibrational modeling of polymeric and molecular crystals.

Polyphasic identification (MALDI-TOF + ITS) of mucosal yeasts in hybrid marmosets from Rio de Janeiro

Abstract Callithrix comprises primates popularly known as marmosets. In the city of Rio de Janeiro, the occurrence of a hybrid form of invasive species prevails. These animals, treated here as Callithrix spp., host several microorganisms in their microbiota, some of which can be pathogenic for humans. The aim of this study to describe culture-dependent yeast microbiota of the oral, rectal, and vaginal mucosae of hybrid marmosets ( Callithrix spp.) inhabiting an urban–forest interface in the Atlantic Forest of Rio de Janeiro. Oral, rectal and vaginal samples were collected from 12 individuals during the winter of 2022. Animals were apparently healthy. The microbial agents obtained by culture isolation were identified to species level by polyphasic taxonomy using the MALDI-TOF MS and partial sequence of the internal transcribed spacer region (ITS1-5.8 S-ITS2) of ribosomal. A total of 26 fungal isolates were obtained. The most isolated species in the study was Candida parapsilosis , and the least frequent yeast were of genus Pichia sp., Trichosporon sp., and Torulaspora sp. Fungal infections in wild animals, depending on the causal agent, can be extremely pathogenic and contagious not only among animals, but also among humans, therefore fungal identification in these animals is important for future perspective.

Vibronic coupling of competing internal conversion and intersystem crossing in xanthone

We theoretically investigated vibronic coupling responsible for nonradiative transitions, i.e., internal conversion (IC) and intersystem crossing (ISC), in xanthone. The nonradiative decay pathway of aromatic ketones is often debated because of their fast ISC. Xanthone in the gas phase follows a pathway that obeys El-Sayed’s rule, namely, IC from the 1ππ* to 1nπ* states and ISC from the 1nπ* to 3ππ* states, of which a simple pathway is adequate for analyzing vibronic structures. We employed an expression for the nonradiative rate constant based on Fermi’s golden rule within the mixed-spin crude adiabatic approximation, which has the advantage that both IC and ISC can be considered as equally vibronically induced transitions. Our calculations showed that the IC from the 1ππ* to 1nπ* state was faster than ISC channels because of stronger vibronic coupling and less favorable spin–orbit (SO) coupling to nearby triplets. In addition, the ISC from 1nπ* to 3ππ* was faster than that from 1nπ* to 3nπ* because of the large structural displacement and small energy gap. This study can provide guidelines for determining whether IC or ISC dominates, depending on the balance between vibronic coupling, SO coupling, and the energy gap, thereby informing molecular design for controlled nonradiative decay. ISC can dominate its IC counterpart in a xanthone derivative by tuning the singlet–triplet energy gap and/or the SO coupling.

Stability enhancement via speed adaptation and efficiency improvement for induction machine

A molecular density functional theory of aqueous electrolytic solution

We propose a generalization of molecular density functional theory to describe inhomogeneous solvent mixtures, to model electrolytic solutions. Two electrolytic models are presented, both within the HNC approximation. The first one is a two-component mixture representing a primitive-like model of sodium chloride, where the solvent is described as a dielectric continuum. This popular model has the advantage of simplicity, as the ion densities solely depend on spatial coordinates. In addition, we develop a realistic three-component electrolyte model, in which water solvent is described by a third density field that depends on both spatial and orientational coordinates. The proposed methodology and its tridimensional implementation (three spatial coordinates and three Euler angles) are validated by comparing the solvation properties of a sodium cation with the predictions of integral equation theory solved in 1D (one intermolecular distance and five Euler angles), showing near-perfect agreement. This methodology enables the study of solvation properties of solutes of arbitrary shapes in electrolytic solutions, as demonstrated with the prototypical N-methyl acetamide molecule immersed in both electrolytic solution models.

Accelerating supercritical pharmaceutical formulation via interpretable data-driven prediction of drug solubility

Abstract Drug solubility in supercritical carbon dioxide (SC-CO 2 ) plays a pivotal role in the development of particle engineering, drug loading, and solvent-free pharmaceutical formulations. However, experimental solubility determination in supercritical systems remains costly, time-consuming, and compound-specific. In this study, an interpretable data-driven framework is proposed to support pharmaceutical formulation scientists by accurately predicting drug solubility in SC-CO 2 while elucidating the governing physicochemical factors. Multiple machine learning regressors, including Extreme Gradient Boosting and Support Vector Regression, were developed and further integrated into an ensemble strategy to enhance robustness and generalizability. Model performance was systematically optimized using bio-inspired metaheuristic algorithms, enabling efficient hyperparameter selection across complex, nonlinear search spaces. Beyond predictive accuracy, model interpretability was emphasized through sensitivity-based and amplitude-based feature analyses, revealing the dominant molecular descriptors and process conditions influencing solubility behavior. The results demonstrate that the proposed framework not only improves solubility prediction accuracy but also provides mechanistic insights relevant to drug selection, formulation feasibility, and supercritical processing design. This work establishes a practical computational tool for accelerating pharmaceutical development pipelines involving supercritical fluid technologies.

Hierarchical relaxation and the microscopic origin of fast Li+ ions transport in Li7La3Zr2O12

Superionic conductors maintain the structural order of crystals while allowing ions within them to move with liquid-like mobility. Li7La3Zr2O12 (LLZO) is a representative example with high thermal stability, a wide electrochemical window, and fast Li+ ion transport. Despite its technological importance, the microscopic origin of its superionic behavior remains insufficiently understood, particularly the role of collective ion motion. In this work, we employ large-scale molecular dynamics simulations based on a deep neural-network derived potential to investigate the structural and dynamical evolution of undoped LLZO across a broad temperature range. The simulations reveal that several dynamical properties of Li+ ions in LLZO resemble those of glass-forming liquids. A characteristic temperature near the Tammann temperature marks the point at which Li+ ion vibrations deviate from harmonic behavior and cooperative hopping begins to emerge, a change accompanied by enhanced dynamic heterogeneity, as reflected in an increase in the Debye–Waller parameter and a peak in the non-Gaussian parameter. By identifying string-like cooperative motion, we establish a direct link between local vibrational processes, structural relaxation, and long-range ion transport. Furthermore, analysis of the vibrational density of states reveals that the excess low-frequency modes originate from mobile Li+ ions and are closely linked to the onset of cooperative dynamics.

Research on short-term prediction method of photovoltaic power based on HPO-VMD-BiLSTM

Enhanced second-harmonic generation from WS2/ReSe2 heterostructure

Van der Waals stacking presents new opportunities for nonlinear optics with its remarkable tunability and scalability. However, the fundamental role of interlayer interactions in modifying the overall nonlinear optical susceptibilities remains elusive. In this paper, we report anisotropic enhancement of second-harmonic generation (SHG) from a WS2/ReSe2 heterobilayer, where the individual composite layers possess distinctive crystal phases. We investigate polarization-resolved response and twist-angle dependence in SHG and reveal that band alignment alone is insufficient to explain the observed anisotropy in the modified SHG response. Spectral shifts in excitonic features highlight band renormalization, supporting the role of hybridization between the two layers. Furthermore, SHG enhancement is highly anisotropic and can even be suppressed in some orientations, suggesting possible intensity-borrowing mechanisms within the heterostructure. Our work demonstrates the ability to tune both the intensity and polarization dependence of nonlinear optical responses with van der Waals stacking of distinctive crystal phases.

Quantitative analysis of the effects of air pollution and urbanization on the rate of allergy and chronic obstructive pulmonary disease (COPD)

Random walks and the electronic structure of graphene

Results from the mathematical literature on random walks reveal a closed-form analytical expression for the π-energy and bond number of graphene in the simplest tight-binding model and its Hartree–Fock Hubbard extension. Closed-form expressions follow for all π spectral moments of graphene. Bond numbers of carbon and boron nitride (BN) zigzag nanotubes are found as finite sums, with graphene and hexagonal boron nitride sheets as asymptotes.

An urban bryophyte hotspot in an industrial city: the case of Ostrava Zoo (Czech Republic)

On the second-order functional structure of the chemical potential in conceptual density functional theory

The second-order functional structure of the chemical potential is examined within conceptual density functional theory by treating μ[N, v(r)] as a functional of the electron number and the external potential. In close analogy with the nonperturbative functional expansion of the total energy introduced by Liu and Parr, the chemical potential admits an explicit functional form up to second order that does not rely on a Taylor expansion about a reference state. This representation organizes established first- and second-order response descriptors into a unified response-theoretical framework, where the linear terms recover the standard hardness and Fukui function contributions, while quadratic terms encode nonlinear charge effects and nonlocal response through the hyperhardness, charge sensitivity, and response kernel. The resulting formulation clarifies the hierarchical organization of response properties associated with the chemical potential and provides a complementary perspective on electronegativity equalization within conceptual density functional theory.

A perturbation-recovery generative autoencoder for heterogeneous graphs with attributes missing

Universal scaling laws in melting thermodynamics of gold nanoparticles: Insights from machine learning molecular dynamics

Understanding the melting behavior at the nanoscale regime serves a fundamental role in both the scientific community and industrial applications. In particular, the melting of nanoparticles (NPs) exhibits behaviors that differ qualitatively from bulk materials due to pronounced size-dependent properties and surface/volume ratio effects, but a unified theoretical understanding remains elusive. Here, by developing a machine-learning interatomic potential applicable across diverse local atomic environments and wide temperature ranges, we systematically investigate the melting thermodynamics of Au NPs spanning from small clusters (102 atoms) to large NPs (105 atoms) through a series of nanosecond-long molecular dynamics simulations. A complete solid–liquid phase diagram of NPs across 1–14 nm diameters is presented, clearly distinguishing the unique surface premelting behavior and complete melting. The size-dependent melting curve follows the Gibbs–Thomson relationship. More importantly, we demonstrate that the melting entropy changes in nanoparticle systems substantially deviate from the empirical Richard’s rule and its generalized form valid for bulk elemental systems. Moreover, we found that all the components of melting entropy follow the same scaling law, based on which we derived a thermodynamic correlation between the NP system and its bulk values. These results bridge the thermodynamic description from the single-atom limit to bulk materials, providing a unique insight for understanding and predicting nanoscale melting thermodynamics.