Lithium-ion battery (LIB) thermal runaway generates characteristic gases, including H2, CO, C2H4 and C2H2, necessitating the development of efficient sensing materials for early-warning safety monitoring. Herein, trinuclear Cu3 and Zn3 cluster-functionalized graphitic carbon nitride (g-CN) monolayers were designed as sensing platforms, and their adsorption mechanisms and sensing performances toward thermal runaway gases were systematically investigated by first-principles calculations. Both Cu3 and Zn3 clusters were found to be stably anchored on the g-CN substrate through strong metal–N coordination interactions, while effectively regulating the electronic structure of g-CN and enhancing its surface activity. Compared with pristine g-CN exhibiting weak physisorption, Cu3 modification induces strong chemisorption toward all investigated gases through Cu 3d orbital-mediated interactions and significant charge transfer with C2H2 showing the strongest adsorption interaction. In contrast, Zn3-g-CN exhibits only moderate chemisorption toward C2H2 owing to the electronically stable d10 configuration of Zn. Theoretical sensing evaluation demonstrates that Cu3-g-CN enables rapid recovery (10.71 s) for H2 detection at elevated temperature, whereas Zn3-g-CN exhibits favorable reversible sensing performance toward C2H2, achieving a recovery time of 62.69 s at 348 K. This work highlights the decisive role of metal cluster electronic configuration in regulating adsorption strength, charge transfer, and sensing reversibility, providing insights into the rational design of g-CN-based materials for LIB thermal runaway monitoring.
The predictive performance of molecular machine learning models depends on how molecules are featurized, yet fair comparisons across methods are rare because studies often vary architectures and hyperparameters. We benchmark five geometry-based molecular featurization techniques that operate on three-dimensional atomic coordinates, i.e., Coulomb Matrix, Bag of Bonds, Atom-Centered Symmetry Functions (ACSF), Smooth Overlap of Atomic Positions (SOAP), and Weighted Views, across five benchmark datasets (ESOL, FreeSolv, BBBP, BACE, ClinTox). While Weighted Views employs a unique weighted aggregation architecture, we controlled for model capacity by adjusting network widths so that all methods had roughly the same number of trainable parameters. All representations are evaluated under stratified k-fold cross-validation with three independent seeds, and results are reported as the mean and standard deviation across the resulting test folds.
Perovskites being an advanced class of materials gained considerable attention in diverse fields because of their compositional flexibility and tunable physicochemical properties. Herein, first-principles calculations within the framework of density functional theory (DFT) were carried out to explore the structural, electronic, optical, elastic, thermodynamic behaviors, and electron density distribution of cubic TlXF3 (X = Zn, Cd, and Hg) fluoro-perovskites in the Pm-3 m phase. Negative formation energy values confirm their energetically favorable. The band structure calculated using the HSE06 hybrid functional reveal direct band gaps of 4.120 eV for TlZnF3 and 4.021 eV for TlCdF3, whereas TlHgF3 exhibits indirect bandgap of 1.450 eV, associated with compositional dependent variations in the Tl-6p, Hg-5d, and F-2p orbital contributions near the band edges, supported by DOS analysis. Optical results reveal strong absorption coefficients and enhanced dielectric response of TlHgF3 exhibits distinctive optoelectronic competence. Elastic constants (C11, C12, and C44) meet Born stability criteria, further confirm their mechanical stability, while the Pugh's and Poisson's ratio predict ductile characteristics. Additionally, Debye temperatures (θD), melting points, and Grüneisen parameters substantiate significance thermal stability and lattice anharmonicity. These findings highlight the potential of studied fluoroperovskites for distinct technological applications, with TlZnF3 and TlCdF3 as a promising candidate for UV optoelectronic devices, while TlHgF3 favorable for photovoltaic applications.
The anticancer prodrug treosulfan undergoes nonenzymatic conversion to the monoepoxide ((2S,3S)-1,2-Epoxybutane-3,4-diol 4-methanesulfonat (EBDM)) and subsequently to the diepoxide ((2S,3S)-1,2:3,4-Diepoxybutane (DEB)), both of which alkylate DNA. This work reports a dispersion-corrected density functional theory investigation, at the ωB97XD/6–311 + G(d,p) level with Solvation Model based on Density (SMD implicit solvation), of the adsorption of treosulfan and its three active metabolites (EBDM01, EBDM02, DEB) on pristine C20 fullerene in the gas phase and aqueous solution. The structurally distinct adsorption configurations were optimised and characterised by their energetic, structural, electronic, and spectroscopic properties. All adsorption energies are negative, and the drug molecules remain intact upon adsorption according to trivial bond perturbation, and vibrational frequency shifts, confirming reversible physisorption. The most stable aqueous-phase configurations exhibit adsorption energies of −9.21, −7.43 and − 4.39 kcal mol−1 (and corresponding enthalpies of −7.66, −5.88 and − 2.80 kcal mol−1) for treosulfan cfg1, EBDM02 cfg4 and DEB cfg1, respectively. The adsorption enthalpy weakens monotonically along the prodrug activation pathway, establishing a thermodynamic gradient consistent with activation-triggered drug release in which the prodrug is preferentially retained by the carrier while the ultimate alkylating metabolite is the most readily released. The aqueous phase qualitatively reorders the EBDM configuration ranking relative to the gas phase, validating the necessity of dual-phase modelling for biologically relevant conclusions. These results provide a computational basis for the design of fullerene-based carriers targeting treosulfan and its activation products.
In this work, we present a systematic investigation of the ground- and excited-states of six anionic sulfonated azo dyes with broad applications in the textile industry: Allura Red, Amaranth, Carmoisine, Ponceau, Red 2G, and Sunset Yellow. Ground-state properties were evaluated using density functional theory (DFT) at the ωB97X-D4/ma-def2-SVP level of theory. To identify the most suitable functional for excited-state calculations and to assess vertical excitation energies to low-lying excited-states, we tested a series of hybrid, meta-hybrid, and range-separated hybrid functionals from Jacob’s ladder using the Tamm-Dancoff approximation (TDA) and the ma-def2-SVP basis set and compared to results from RI-CC2 and STEOM-DLPNO-CCSD approaches. Excited-state properties were then further examined using TDA at the CAM-B3LYP//ωB97X-D4/ma-def2-SVP level of theory. Additionally, the quantum theory of atoms in molecules (QTAIM) was applied to explore density-related properties in both ground and excited-states. We found, perhaps not surprisingly, that excitation weakens the N=N bond in both the enol and keto forms of the dyes, but with variations depending on the molecule, conformer, and electronic state.
