Chemical Physics

arXiv:physics.chem-ph

Atomic and molecular structure, dynamics, spectroscopy, chemical reactions.

Looking for a broader view? This category is part of:

Trending in Chemical Physics

AceFF: A State-of-the-Art Machine Learning Potential for Small Molecules

We introduce AceFF, a pre-trained machine learning interatomic potential (MLIP) optimized for small molecule drug discovery. While MLIPs have emerged as efficient alternatives to Density Functional Theory (DFT), generalizability across diverse chemical spaces remains difficult. AceFF addresses this via a refined TensorNet2 architecture trained on a comprehensive dataset of drug-like compounds. This approach yields a force field that balances high-throughput inference speed with DFT-level accuracy. AceFF fully supports the essential medicinal chemistry elements (H, B, C, N, O, F, Si, P, S, Cl, Br, I) and is explicitly trained to handle charged states. Validation against rigorous benchmarks, including complex torsional energy scans, molecular dynamics trajectories, batched minimizations, and forces and anergy accuracy demonstrates that AceFF establishes a new state-of-the-art for organic molecules. The AceFF-2 model weights and inference code are available at https://huggingface.co/Acellera/AceFF-2.0.

2601.00581

Jan 2026Chemical Physics

Related categories:

Accelerator Physics Atmospheric and Oceanic Physics Applied Physics Atomic and Molecular Clusters Atomic Physics

310 papers

Scalable Machine Learning Force Fields for Macromolecular Systems Through Long-Range Aware Message Passing

Machine learning force fields (MLFFs) have revolutionized molecular simulations by providing quantum mechanical accuracy at the speed of molecular mechanical computations. However, a fundamental reliance of these models on fixed-cutoff architectures limits their applicability to macromolecular systems where long-range interactions dominate. We demonstrate that this locality constraint causes force prediction errors to scale monotonically with system size, revealing a critical architectural bottleneck. To overcome this, we establish the systematically designed MolLR25 ({Mol}ecules with {L}ong-{R}ange effect) benchmark up to 1200 atoms, generated using high-fidelity DFT, and introduce E2Former-LSR, an equivariant transformer that explicitly integrates long-range attention blocks. E2Former-LSR exhibits stable error scaling, achieves superior fidelity in capturing non-covalent decay, and maintains precision on complex protein conformations. Crucially, its efficient design provides up to 30% speedup compared to purely local models. This work validates the necessity of non-local architectures for generalizable MLFFs, enabling high-fidelity molecular dynamics for large-scale chemical and biological systems.

2601.03774Jan 2026

View

A Length-Gauge Origin-Invariant Approach to Vibrational Circular Dichroism Spectra without Gauge-Including Atomic Orbitals

We have extended the origin-invariant length gauge (LG(OI)) approach -- originally developed by Caricato and co-workers for optical rotation (OR) and electronic circular dichroism (ECD) -- to vibrational circular dichroism (VCD). This approach avoids the need for gauge-including atomic orbitals (GIAOs), which are typically required to circumvent the unphysical dependence of the CD rotatory strengths on the arbitrary choice of coordinate origin for length gauge (LG) computations. Benchmark VCD spectra are presented for (P)-hydrogen peroxide, (S)-methyloxirane, (1R, 5R)-α-pinene, and (1R, 4R)-camphor using Hartree-Fock (HF) theory and density functional theory (DFT) methods across a range of basis sets and compared to those obtained from LG, velocity-gauge (VG), and GIAO computations. These analyses show that for VCD the LG(OI) approach does not converge to the basis-set limit as rapidly as the GIAO approach, but does yield similar quality spectra as GIAO for all major VCD peaks for quadruple-zeta-quality basis sets. The LG(OI) and VG VCD spectra are less reliable compared to GIAOs for smaller basis sets.

2601.02343Jan 2026

View

Robust and fragile quantum effects in the transfer kinetics of delocalized excitons between B850 units of LH2 complexes

Aggregates of light harvesting 2 (LH2) complexes form the major exciton-relaying domain in the photosynthetic unit of purple bacteria. Application of a generalized master equation to pairs of the B850 units of LH2 complexes, where excitons predominantly reside, provide quantitative information on how the inter-LH2 exciton transfer depends on the distance, relative rotational angle, and the relative energies of the two LH2s. The distance dependence demonstrates significant enhancement of the rate due to quantum delocalization of excitons, the qualitative nature of which remains robust against the disorder. The angle dependence reflects isotropic nature of exciton transfer, which remains similar for the ensemble of disorder. The variation of the rate on relative excitation energies of LH2 exhibits resonance peaks, which however is fragile as the disorder becomes significant. Overall, the average transfer times between two LH2s are estimated to be in the range of 4 - 25 ps for physically plausible inter-LH2 distances.

2601.03291Jan 2026

View

AI-Guided Discovery of Novel Ionic Liquid Solvents for Industrial CO2 Capture

We present an AI-driven approach to discover compounds with optimal properties for CO2 capture from flue gas-refinery emissions' primary source. Focusing on ionic liquids (ILs) as alternatives to traditional amine-based solvents, we successfully identify new IL candidates with high working capacity, manageable viscosity, favorable regeneration energy, and viable synthetic routes. Our approach follows a five-stage pipeline. First, we generate IL candidates by pairing available cation and anion molecules, then predict temperature- and pressure-dependent CO2 solubility and viscosity using a GNN-based molecular property prediction model. Next, we convert solubility to working capacity and regeneration energy via Van't Hoff modeling, and then find the best set of candidates using Pareto optimization, before finally filtering those based on feasible synthesis routes. We identify 36 feasible candidates that could enable 5-10% OPEX savings and up to 10% CAPEX reductions through lower regeneration energy requirements and reduced corrosivity-offering a novel carbon-capture strategy for refineries moving forward.

2601.03284Jan 2026

View

AceFF: A State-of-the-Art Machine Learning Potential for Small Molecules

2601.00581Jan 2026

View

Interpretable Machine Learning for Quantum-Informed Property Predictions in Artificial Sensing Materials

Digital sensing faces challenges in developing sustainable methods to extend the applicability of customized e-noses to complex body odor volatilome (BOV). To address this challenge, we developed MORE-ML, a computational framework that integrates quantum-mechanical (QM) property data of e-nose molecular building blocks with machine learning (ML) methods to predict sensing-relevant properties. Within this framework, we expanded our previous dataset, MORE-Q, to MORE-QX by sampling a larger conformational space of interactions between BOV molecules and mucin-derived receptors. This dataset provides extensive electronic binding features (BFs) computed upon BOV adsorption. Analysis of MORE-QX property space revealed weak correlations between QM properties of building blocks and resulting BFs. Leveraging this observation, we defined electronic descriptors of building blocks as inputs for tree-based ML models to predict BFs. Benchmarking showed CatBoost models outperform alternatives, especially in transferability to unseen compounds. Explainable AI methods further highlighted which QM properties most influence BF predictions. Collectively, MORE-ML combines QM insights with ML to provide mechanistic understanding and rational design principles for molecular receptors in BOV sensing. This approach establishes a foundation for advancing artificial sensing materials capable of analyzing complex odor mixtures, bridging the gap between molecular-level computations and practical e-nose applications.

2601.00503Jan 2026

View

QSAR-Guided Generative Framework for the Discovery of Synthetically Viable Odorants

The discovery of novel odorant molecules is key for the fragrance and flavor industries, yet efficiently navigating the vast chemical space to identify structures with desirable olfactory properties remains a significant challenge. Generative artificial intelligence offers a promising approach for \textit{de novo} molecular design but typically requires large sets of molecules to learn from. To address this problem, we present a framework combining a variational autoencoder (VAE) with a quantitative structure-activity relationship (QSAR) model to generate novel odorants from limited training sets of odor molecules. The self-supervised learning capabilities of the VAE allow it to learn SMILES grammar from ChemBL database, while its training objective is augmented with a loss term derived from an external QSAR model to structure the latent representation according to odor probability. While the VAE demonstrated high internal consistency in learning the QSAR supervision signal, validation against an external, unseen ground truth dataset (Unique Good Scents) confirms the model generates syntactically valid structures (100\% validity achieved via rejection sampling) and 94.8\% unique structures. The latent space is effectively structured by odor likelihood, evidenced by a Fréchet ChemNet Distance (FCD) of $\approx$ 6.96 between generated molecules and known odorants, compared to $\approx$ 21.6 for the ChemBL baseline. Structural analysis via Bemis-Murcko scaffolds reveals that 74.4\% of candidates possess novel core frameworks distinct from the training data, indicating the model performs extensive chemical space exploration beyond simple derivatization of known odorants. Generated candidates display physicochemical properties ....

2512.23080Dec 2025

View

Hamiltonian-Informed Point Group Symmetry-Respecting Ansatz for Variational Quantum Eigensolver

Solving molecular energy levels via the Variational Quantum Eigensolver (VQE) algorithm represents one of the most promising applications for demonstrating practically meaningful quantum advantage in the noisy intermediate-scale quantum (NISQ) era. To strike a balance between ansatz complexity and computational stability in VQE calculations, we propose the HiUCCSD, a novel symmetry-respecting ansatz engineered from the intrinsic information of the Hamiltonian. We theoretically prove the effectiveness of HiUCCSD within the scope of Abelian point groups. Furthermore, we compare the performance of HiUCCSD and the established SymUCCSD via VQE and Adaptive Derivative-Assembled Pseudo-Trotter (ADAPT)-VQE numerical experiments on ten molecules with distinct point groups. The results show that HiUCCSD achieves equivalent performance to SymUCCSD for Abelian point group molecules, while avoiding the potential performance failure of SymUCCSD in the case of non-Abelian point group molecules. Across the studied molecular systems, HiUCCSD cuts the parameter count by 18%-83% for VQE and reduces the excitation operator pool size by 27%-84% for ADAPT-VQE, as compared with the UCCSD ansatz. With enhanced robustness and broader applicability, HiUCCSD offers a new ansatz option for advancing large-scale molecular VQE implementation.

2512.21087Dec 2025

View

Topological resolution of conical intersection seams and the coupled cluster bifurcation via mixed Hodge modules

The rigorous description of Conical Intersections (CIs) remains the central challenge of non-adiabatic quantum chemistry. While the ``Yarkony Seam'' -- the $(3N-8)$-dimensional manifold of degeneracy -- is well-understood geometrically, its accurate characterization by high-level electronic structure methods is plagued by numerical instabilities. Specifically, standard Coupled Cluster (CC) theory suffers from root bifurcations near Ground State CIs, rendering the ``Gold Standard'' of chemistry inapplicable where it is needed most. Here, we present \textbf{QuMorpheus}, an open-source computational package that resolves these singularities by implementing a topological framework based on Dissipative Mixed Hodge Modules (DMHM) [P. Saurabh, arXiv:2512.19487 (2025)]. By algorithmically mapping the CC polynomial equations to a spectral sheaf, we compute the exact Monodromy ($μ$) invariants of the intersection. We demonstrate that this automated algebraic geometry approach correctly identifies the physical ground state topology in the Köhn-Tajti model and resolves the intersection seams of realistic chemical systems, including Ethylene and the Chloronium ion ($\mathrm{H_2Cl^+}$). Furthermore, we apply QuMorpheus to the photoisomerization of Previtamin D, proving that the experimentally observed Woodward-Hoffmann selection rules are a direct consequence of a topological ``Monodromy Wall'' ($μ=1, γ=π$) rather than purely energetic barriers. This establishes a general software solution to the ``Yarkony Problem,'' enabling the robust, automated mapping of global intersection seams in complex molecular systems. The topological stability of these intersections allows for the control protocols discussed in Ref.[P. Saurabh, Submitted to Phys. Rev. X (2025)].

2512.20414Dec 2025

View

Finite-Temperature Thermally-Assisted-Occupation Density Functional Theory, Ab Initio Molecular Dynamics, and Quantum Mechanics/Molecular Mechanics Methods

Recently, thermally-assisted-occupation density functional theory (TAO-DFT) [J.-D. Chai, J. Chem. Phys. 136, 154104 (2012)] has been demonstrated to be an efficient and accurate electronic structure method for studying the ground-state properties of large multi-reference (MR) systems at absolute zero. To explore the thermal equilibrium properties of large MR systems at finite electronic temperatures, in the present work, we propose the finite-temperature (FT) extension of TAO-DFT, denoted as FT-TAO-DFT. Besides, to unlock the dynamical information of large MR systems at finite temperatures, FT-TAO-DFT is combined with ab initio molecular dynamics, leading to FT-TAO-AIMD. In addition, we also develop FT-TAO-DFT-based quantum mechanics/molecular mechanics (QM/MM), denoted as FT-TAO-QM/MM, to provide a cost-effective description of the thermal equilibrium properties of a QM subsystem with MR character embedded in an MM environment at finite temperatures. Moreover, the FT-TAO-DFT, FT-TAO-AIMD, and FT-TAO-QM/MM methods are employed to explore the radical nature and infrared (IR) spectra of n-acenes (n = 2--6), consisting of n linearly fused benzene rings, in vacuum and in an argon (Ar) matrix at finite temperatures. According to our calculations, for n-acenes at 1000 K or below, the electronic temperature effects on the radical nature and IR spectra are very minor, while the nuclear temperature effects on these properties are noticeable. For n-acene in an Ar matrx at absolute zero, the Ar matrix has minimal impact on the radical nature of n-acene, while the co-deposition procedure of n-acene and Ar atoms may affect the IR spectrum of n-acene.

2512.20313Dec 2025

View

Pushing the limits of one-dimensional NMR spectroscopy for automated structure elucidation using artificial intelligence

One-dimensional NMR spectroscopy is one of the most widely used techniques for the characterization of organic compounds and natural products. For molecules with up to 36 non-hydrogen atoms, the number of possible structures has been estimated to range from $10^{20} - 10^{60}$. The task of determining the structure (formula and connectivity) of a molecule of this size using only its one-dimensional $^1$H and/or $^{13}$C NMR spectrum, i.e. de novo structure generation, thus appears completely intractable. Here we show how it is possible to achieve this task for systems with up to 40 non-hydrogen atoms across the full elemental coverage typically encountered in organic chemistry (C, N, O, H, P, S, Si, B, and the halogens) using a deep learning framework, thus covering a vast portion of the drug-like chemical space. Leveraging insights from natural language processing, we show that our transformer-based architecture predicts the correct molecule with 55.2% accuracy within the first 15 predictions using only the $^1$H and $^{13}$C NMR spectra, thus overcoming the combinatorial growth of the chemical space while also being extensible to experimental data via fine-tuning.

2512.18531Dec 2025

View

Super-Poissonian Squeezed Light in the Ground State of Strongly Coupled Light-matter Systems

Strong light-matter coupling enables hybrid states in which photonic and electronic degrees of freedom become correlated even in the ground state. While many-body effects in long-range dispersion interactions are known to reshape electronic properties under such conditions, their impact on quantum-optical observables remains largely unexplored. Here, we address this problem using quantum electrodynamical density-functional theory (QEDFT) combined with the recently developed photon-many-body dispersion (pMBD) functional, which can capture higher-order electron-photon correlations and multi-photon processes. We compute ground-state photonic observables including photon number fluctuations, second-order correlations, and quadrature variances, and find squeezing and super-Poissonian photon statistics emerging from light-matter interactions in the strong coupling regime. Our results demonstrate that capturing the full hierarchy of many-body, electron-photon and multi-photon correlations is essential for a consistent description of quantum-optical properties in strongly coupled molecular systems, establishing QEDFT as a first-principles framework for predicting nonclassical photonic features in the ground state of complex systems.

2512.18242Dec 2025

View

Benchmarking the Impact of Active Space Selection on the VQE Pipeline for Quantum Drug Discovery

Quantum computers promise scalable treatments of electronic structure, yet applying variational quantum eigensolvers (VQE) on realistic drug-like molecules remains constrained by the performance limitations of near-term quantum hardwares. A key strategy for addressing this challenge which effectively leverages current Noisy Intermediate-Scale Quantum (NISQ) hardwares yet remains under-benchmarked is active space selection. We introduce a benchmark that heuristically proposes criteria based on chemically grounded metrics to classify the suitability of a molecule for using quantum computing and then quantifies the impact of active space choices across the VQE pipeline for quantum drug discovery. The suite covers several representative drug-like molecules (e.g., lovastatin, oseltamivir, morphine) and uses chemically motivated active spaces. Our VQE evaluations employ both simulation and quantum processing unit (QPU) execution using unitary coupled-cluster with singles and doubles (UCCSD) and hardware-efficient ansatz (HEA). We adopt a more comprehensive evaluation, including chemistry metrics and architecture-centric metrics. For accuracy, we compare them with classical quantum chemistry methods. This work establishes the first systematic benchmark for active space driven VQE and lays the groundwork for future hardware-algorithm co-design studies in quantum drug discovery.

2512.18203Dec 2025

View

Signatures of coherent energy transfer and exciton delocalization in time-resolved optical cross correlations

We investigate how optical second-order cross correlations witness the quantum features of a prototype donor-acceptor light-harvesting unit. By considering a pair of detuned two-level emitters electronically coupled and incoherently driven to a non-equilibrium steady-state, we gain insight into how electronic quantum properties such as exciton eigenstate delocalization, coherent energy transfer and steady-state electronic coherence, are manifested in the joint probability of emission or optical second-order cross correlation. Specifically, we show that the frequency associated with oscillations present in time-resolved second-order cross correlation functions quantifies not only the time scale of coherent energy transfer but also the degree of delocalization of the exciton eigenstates. Furthermore, we show that time-resolved cross correlations directly witness steady-state electronic coherence. Our work strengthens the idea that measurements of the intensity quantum cross correlations can provide distinctive signatures of the quantum behavior of biophysical emitters.

2512.17741Dec 2025

View

2512.16882

Cartesian-nj: Extending e3nn to Irreducible Cartesian Tensor Product and Contracion

Equivariant atomistic machine learning models have brought substantial gains in both extrapolation capability and predictive accuracy. Depending on the basis of the space, two distinct types of irreducible representations are utilized. From architectures built upon spherical tensors (STs) to more recent formulations employing irreducible Cartesian tensors (ICTs), STs have remained dominant owing to their compactness, elegance, and theoretical completeness. Nevertheless, questions have persisted regarding whether ST constructions are the only viable design principle, motivating continued development of Cartesian networks. In this work, we introduce the Cartesian-3j and Cartesian-nj symbol, which serve as direct analogues of the Wigner-3j and Wigner-nj symbol defined for tensor coupling. These coefficients enable the combination of any two ICTs into a new ICT. Building on this foundation, we extend e3nn to support irreducible Cartesian tensor product, and we release the resulting Python package as cartnn. Within this framework, we implement Cartesian counterparts of MACE, NequIP, and Allegro, allowing the first systematic comparison of Cartesian and spherical models to assess whether Cartesian formulations may offer advantages under specific conditions. Using TACE as a representative example, we further examine whether architectures constructed from irreducible Cartesian tensor product and contraction(ICTP and ICTC) are conceptually well-founded in Cartesian space and whether opportunities remain for improving their design.

2512.16882Dec 2025

View

Multi-messenger tracking of coherence loss during bond breaking

Coupled electronic and nuclear motions govern chemical reactions, yet disentangling their interplay during bond rupture remains challenging. Here we follow the light-induced fragmentation of Br$_2$ using a coincidence-based multi-messenger approach. A UV pulse prepares the dissociative state, and strong-field ionization probes the evolving system. Coincident measurement of three-dimensional photoion and photoelectron momenta provides real-time access to both the instantaneous internuclear separation and the accompanying reorganization of the electronic structure, allowing us to determine the timescale of bond breaking. We find that electronic rearrangement concludes well before the nuclei reach the bond-breaking distance, revealing a hierarchy imposed by electron-nuclear coupling. Supported by semiclassical modelling, the results show that the stretched Br$_2$ molecule behaves as a two-centre interferometer in which the loss of coherence between atomic centres encodes the coupled evolution of electrons and nuclei. Our work establishes a general framework for imaging ultrafast electron-nuclear dynamics in molecules.

2512.16132Dec 2025

View

Tree Tensor Networks Methods for Efficient Calculation of Molecular Vibrational Spectra

We develop and employ general Tree Tensor Networks (TTNs) to compute the vibrational spectra for two model systems: a set of 64-dimensional coupled oscillators and acetonitrile. We explore various tree architectures, ranging from the simple linear structure of Matrix Product States (MPS), to trees where only the leaf nodes carry a physical leg -- as seen in the underlying ansatz of the Multilayer Multiconfiguration Time-Dependent Hartree (ML-MCTDH) method -- and further to more general trees in which all nodes are allowed to possess a physical leg. In addition, we implement Locally Optimal Block Preconditioned Conjugate Gradient (LOBPCG) methods and Inverse Iteration methods as eigensolvers. By means of comprehensive benchmarking of runtime and accuracy, we demonstrate that sub-wavenumber accuracy in vibrational spectra is achievable with all TTN structures. MPS and three-legged tree tensor network states (T3NS) have similar runtimes, whereas leaf-only trees require significantly more time. All numerical simulations were performed using PyTreeNet, a Python package designed for flexible tensor network computations.

2512.15875Dec 2025

View

Learning continuous SOC-dependent thermal decomposition kinetics for Li-ion cathodes using KA-CRNNs

Thermal runaway in lithium-ion batteries is strongly influenced by the state of charge (SOC). Existing predictive models typically infer scalar kinetic parameters at a full SOC or a few discrete SOC levels, preventing them from capturing the continuous SOC dependence that governs exothermic behavior during abuse conditions. To address this, we apply the Kolmogorov-Arnold Chemical Reaction Neural Network (KA-CRNN) framework to learn continuous and realistic SOC-dependent exothermic cathode-electrolyte interactions. We apply a physics-encoded KA-CRNN to learn SOC-dependent kinetic parameters for cathode-electrolyte decomposition directly from differential scanning calorimetry (DSC) data. A mechanistically informed reaction pathway is embedded into the network architecture, enabling the activation energies, pre-exponential factors, enthalpies, and related parameters to be represented as continuous and fully interpretable functions of the SOC. The framework is demonstrated for NCA, NM, and NMA cathodes, yielding models that reproduce DSC heat-release features across all SOCs and provide interpretable insight into SOC-dependent oxygen-release and phase-transformation mechanisms. This approach establishes a foundation for extending kinetic parameter dependencies to additional environmental and electrochemical variables, supporting more accurate and interpretable thermal-runaway prediction and monitoring.

2512.15628Dec 2025

View

NMIRacle: Multi-modal Generative Molecular Elucidation from IR and NMR Spectra

Molecular structure elucidation from spectroscopic data is a long-standing challenge in Chemistry, traditionally requiring expert interpretation. We introduce NMIRacle, a two-stage generative framework that builds upon recent paradigms in AI-driven spectroscopy with minimal assumptions. In the first stage, NMIRacle learns to reconstruct molecular structures from count-aware fragment encodings, which capture both fragment identities and their occurrences. In the second stage, a spectral encoder maps input spectroscopic measurements (IR, 1H-NMR, 13C-NMR) into a latent embedding that conditions the pre-trained generator. This formulation bridges fragment-level chemical modeling with spectral evidence, yielding accurate molecular predictions. Empirical results show that NMIRacle outperforms existing baselines on molecular elucidation, while maintaining robust performance across increasing levels of molecular complexity.

2512.19733Dec 2025

View

Sample-based quantum diagonalization as parallel fragment solver for the localized active space self-consistent field method

Accurately and efficiently describing strongly correlated electronic systems is a central challenge in quantum computational chemistry, with classical and quantum computers. The localized active space self-consistent field method (LASSCF) uses a product of fragment active spaces as a variational space, with the Schrödinger equation solved exactly in each fragment and the fragment active-space orbitals defined in a self-consistent manner. LASSCF is accurate for systems with strong intra-fragment and weak inter-fragment correlation, and its computational cost is combinatorial with respect to the size of the individual fragment active spaces, rather than their product. However, exactly solving the Schrödinger equation in each fragment remains a substantial bottleneck. Here, we address the possibility of solving the fragment active space Schrödinger equation with approximate methods, particularly sample-based quantum diagonalization (SQD). SQD is a technique that uses a quantum computer to sample configurations from a chemically motivated quantum circuit and a classical computer to mitigate errors and solve the Schrödinger equation in a subspace of the configuration space. We apply the proposed method, LASSQD, to the [Fe(H$_2$O)$_4$]$_2$bpym$^{4+}$ compound and the [Fe$^{\mathrm{III}}$Fe$^{\mathrm{III}}$Fe$^{\mathrm{II}}$($μ$$_3$-O)-(HCOO)$_6$] complex for calculating the intermediate-spin ground state energies. We observe that LASSQD can tackle fragment sizes intractable by LASSCF, achieves within 1kcal/mol agreement to LASSCF, and delivers results that are competitive with alternative classical methods to solve the Schrödinger equation, and thus can be used as a starting point for a perturbative treatment (LASSQD-PDFT) to recover correlation external to the active space.

2512.14936Dec 2025

View

Page 1 of 16