Physchem

The solvent-accessible surface area (SASA) of amino acid residues is a crucial parameter for protein structure analysis; however, precise computational methods such as FreeSASA are computationally expensive. As an alternative, empirical approximations based on residue interaction network (RIN) graphs can offer high speed while maintaining acceptable accuracy. In this study, we propose and validate three empirical functions for estimating relative SASA—approx_sasa, surface_score, and exp_sasa—using node degree as the sole argument. We present a comparative analysis of two graph construction approaches: the classical Cα-graph (8 Å threshold) and the heavy-atom graph (HAG, 5.0 Å threshold). Parameters were calibrated on a dataset of 509 protein structures (128,794 residues) using the true relative SASA calculated by the FreeSASA library. An extended set of 11 topological features was also developed and validated. Ensemble models (Random Forest, XGBoost) achieved a best performance of MAE = 0.057 ± 0.033 and Pearson r = 0.915 ± 0.080 on HAG, outperforming graph neural networks (GCN, GAT, GraphSAGE) in this setting. The empirical formulas demonstrate extreme computational efficiency (0.008 ms per structure), ~26,000× faster than FreeSASA, making them suitable for large-scale pipelines requiring both speed and interpretability. Random Forest on HAG is recommended for applications requiring maximum accuracy, while GraphSAGE on HAG is a viable deep learning alternative. Full article

Greywater treatment is an essential component of sustainable water management strategies, particularly in regions experiencing water scarcity. This study presents a quantitative comparative evaluation of three conventional inorganic coagulants—aluminum sulfate (Al₂(SO₄)₃), ferric sulfate (Fe₂(SO₄)₃), and ferric chloride (FeCl₃)—for domestic greywater treatment using standardized jar test procedures. The initial turbidity of the greywater was 186 ± 5 NTU. Coagulant performance was assessed based on final turbidity, removal efficiency, and pH variation, with dosages expressed as mg of active metal per liter to ensure comparability. Ferric chloride exhibited the highest turbidity removal efficiency, achieving a maximum removal efficiency of 96.2 ± 0.8% at 68.8 mg Fe/L and reducing turbidity to 7 ± 1 NTU. Ferric sulfate achieved a maximum removal efficiency of 92.5 ± 1.2%, while aluminum sulfate reached 84.9 ± 1.4% under the evaluated conditions. Statistical analysis confirmed significant differences among coagulants at intermediate and high dose ranges (p < 0.05). The results indicate that ferric chloride provides the most effective clarification performance for the tested greywater matrix, while maintaining final pH values within a suitable range for non-potable reuse. This study contributes to coagulant selection for decentralized greywater treatment systems through standardized, quantitative comparison under identical operational conditions. Full article

Low-molecular-weight lychee polyphenol (Lyp), a standardized oligomeric polyphenol derived from Litchi chinensis fruit, possesses potent antioxidant activity but is susceptible to physicochemical instability, which may limit its practical application. In this study, freeze-dried systems of Lyp with α-, β-, and γ-cyclodextrins (CDs) were prepared to investigate the effects of cyclodextrins on the physicochemical properties and antioxidant activity of Lyp. Powder X-ray diffraction, thermogravimetric analysis, Fourier transform infrared spectroscopy, near-infrared spectroscopy, and solution-state NMR were employed to characterize the prepared systems. Freeze-drying produced amorphous Lyp/CD systems accompanied by enhanced thermal stability and changes in the hydrogen-bonding environment compared with the corresponding physical mixtures. Solution-state ¹H NMR and NOESY analyses suggested molecular association between Lyp constituents and cyclodextrins, with the βCD system exhibiting the most pronounced spectral changes. Consistent with these physicochemical findings, the freeze-dried βCD system showed the greatest enhancement of DPPH radical scavenging activity among the three cyclodextrins examined. These results suggest that βCD provides the most favorable molecular environment for Lyp, leading to improved physicochemical characteristics and antioxidant performance. The present findings demonstrate that freeze-drying combined with β-cyclodextrin is an effective strategy for improving the physicochemical performance and antioxidant functionality of low-molecular-weight lychee polyphenol. Full article

In this study, the Boscovich curve of 1763 is reinterpreted as a mean-field potential for interacting particles in condensed matter. In a dense many-body system, each particle experiences an effective potential arising from the average distribution of all the others. This mean-field potential, which exhibits alternating maxima (energy barriers) and minima (coordination shells), thereby reducing the complexity of the N-body problem to an effective two-body radial problem, with the correlation distance r as the key variable. The relationship between the PMF and the radial distribution function g(r) is given by the Kirkwood equation U_B(r) = −kT ln g(r), which provides a multi-well potential in condensed matter. Furthermore, the system is described by the Fisher density functional equation for the correlation amplitudes, −2kT ∇²ψ(r) + U_B(r)ψ(r) = μψ(r) whose eigenvalues μ_i correspond to potential levels and whose eigenfunctions ψ_i are the correlation amplitudes of the coordination shell structure. Based on the multi-well potential picture, the oscillatory behavior of U_B(r) is modeled analytically by a weighted sum of Lennard-Jones potentials, modulated by sigmoid functions. The parameters—well depths, widths, and coordination distances—are assigned on the basis of known structural properties of the system, derived either from experimental data or from geometric models such as FCC or HCP lattices. The radial distribution function is then reconstructed as a linear combination of the squared eigenfunctions obtained from the Fisher equation. The resulting discrete eigenvalue spectrum provides a spectral interpretation of the shell structure of condensed matter, wherein the complexity of many-body interactions is encoded in a hierarchy of correlation modes, each associated with a specific coordination shell. Unlike classical DFT—which relies on approximate excess free-energy functionals—and Ornstein–Zernike theory—which requires closure approximations—our approach provides a direct spectral interpretation of the coordination shell structure through the eigenvalue spectrum of the Fisher equation, where the PMF acts as the effective potential and the radial distribution function is reconstructed as a combination of squared eigenfunctions. The method is validated for liquid argon and FCC lattices and establishes a historical connection with Boscovich’s curve as a statistical potential. Full article

Graphene oxide (GO) and reduced graphene oxide (rGO) are widely studied two-dimensional carbon nanomaterials for optoelectronic devices. Because the oxygen content and degree of reduction govern the electronic structure of GO-derived films, controlling the reduction process is essential for tailoring their properties. Laser-induced reduction provides a tunable, contact-free route to transparent and conductive graphene-based layers. In this work, 80 nm spray-coated GO layers were reduced using a KrF excimer laser (248 nm, 20 ns) at a fluence of 20 mJ cm⁻², while systematically varying the number of laser pulses (LP) from 1 to 1000. We tuned the degree of GO reduction by stepwise increasing the number of LP and followed the resulting changes in surface composition using X-ray photoelectron spectroscopy (XPS) and near-edge X-ray absorption fine-structure (NEXAFS) spectroscopy. The surface composition evolves non-monotonically with the number of laser pulses, revealing a multi-step reduction mechanism. At low laser doses, epoxide groups are preferentially removed or converted, generating a more disordered distribution of hydroxyl-containing sites on the GO sheets. At intermediate laser doses, oxygen-containing groups are depleted, and sp² conjugation is restored. After extended irradiation in air, however, oxygenated surface species partially re-form. Conductivity measurements show that the sheet resistance reaches a minimum at approximately 300 LP, consistent with efficient chemical reduction and recovery of the conjugated carbon network. These results provide molecular-level guidelines for optimizing laser-induced GO reduction toward graphene-based transparent conductive layers. Full article

The transition toward low-carbon cement production requires alternative fuels with improved environmental performance and resource efficiency. Refuse-Derived Fuel (RDF), produced from pre-treated non-recyclable waste streams, represents a promising alternative fuel; however, its heterogeneous and varying composition presents challenges for stable combustion and process optimization. In this study, a comprehensive physicochemical characterization of RDF was performed and compared with pet coke, a conventional, fossil cement kiln fuel. The analysis included manual sorting, particle size distribution, elemental characterization, Scanning Electron Microscopy coupled with Energy dispersive X-ray spectroscopy (SEM/EDS), X-ray Diffraction (XRD), thermogravimetric and differential scanning calorimetry (TGA/DSC), and Higher Heating Value (HHV) determination. The RDF sample exhibited a heterogeneous polymeric-mineral composition dominated by plastics, paper, textiles, and inorganic fractions. TGA revealed a broad multi-stage thermal degradation profile, while calorimetry indicated HHV of 19.2 ± 0.3 MJ/kg and ash content of 11.6 wt.%, compared with 34.57 ± 0.11 MJ/kg and 1.97 wt.% for pet coke, respectively. SEM/EDS and XRD analyses confirmed the coexistence of polymeric and mineral phases in RDF, whereas pet coke exhibited a predominantly carbonaceous and homogeneous composition. The generated dataset supports computational fluid dynamics (CFD)-based cement kiln combustion models and RDF utilization for fossil fuel substitution in the cement industry. Full article

The increasing prevalence of heavy metal contamination in aqueous environments poses severe risks to both human health and the ecosystem. To address this environmental concern, this study focused on the utilization of a nature-based silver clay nanocomposite (NAgC) for the removal of cadmium (Cd(II)) and chromium (Cr(VI)) ions from aqueous media. The NAgC was synthesized through the reduction of silver ions in a clay matrix by employing the leaf of Manihot esculenta extract as the reducing agent. The NAgC was characterized using Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), scanning electron microscopy (SEM) and transmission electron microscopy (TEM), Brunauer, Emmett, and Teller (BET) surface area as well as energy-dispersive X-ray spectroscopy (EDX). The adsorption behavior of the NAgC in batch experiments under varying pH, contact time, initial concentration, and adsorbent dosage were investigated for the removal of cadmium and chromium from aqueous media. The optimum adsorption of cadmium ions was 157.87 mg/g, achieved at an initial concentration of 20 mg/L at 1 g dosage at pH 7 for 60 min, while that of chromium was 137.91 mg/g at a metal concentration of 20 mg/L at 1 g dosage at pH 7 for 60 min. The Freundlich isotherm model adequately explained the adsorption data of both Cd(II) and Cr(VI). However, the pseudo-first order (PFO) adequately interpreted the kinetic data for both Cd(II) and Cr(VI). Full article

Messenger RNA–lipid nanoparticles (mRNA-LNPs) have emerged as a transformative platform for nucleic acid therapeutics, yet their complex four-component lipid architecture comprising ionizable lipids, PEG–lipids, helper phospholipids, and cholesterol presents substantial analytical challenges for quality control and regulatory compliance. This review presents, for the first time, a critical evaluation of liquid chromatography–mass spectrometry (LC-MS) strategies specifically tailored to quantify all four lipid classes and their degradation products within mRNA-LNP formulations. Unlike prior general lipidomics reviews, we provide a comparative assessment of orthogonal LC modalities including reversed-phase ultra-high-performance liquid chromatography, hydrophilic-interaction liquid chromatography, ion-pairing reversed-phase LC, and high-performance liquid chromatography charged aerosol detection with explicit performance metrics (sensitivity, linearity, and run time). We further integrate emerging analytical frontiers—single-particle analysis, degradation product profiling (e.g., oxysterols and reactive electrophiles), and regulatory frameworks (ICH Q2(R1), Analytical Quality by Design)—to offer a practical guide for method selection. This review’s uniqueness lies in its systematic, application-focused comparison of LC-MS workflows addressing lipid-specific vulnerabilities, matrix effects, and stability-indicating parameters, filling a critical gap between analytical chemistry and mRNA-LNP product development. Full article

Powder X-ray diffraction was used to validate the crystal structure of iron metal powder and to demonstrate a transparent workflow for extracting crystallographic and microstructural information from a simple laboratory dataset. The diffraction profile collected over the angular range of approximately 36° to 95° contains three dominant reflections located at 2θ = 44.850°, 65.218°, and 82.540°. Conversion of the peak positions into d-spacings using Bragg’s law gives values of approximately 2.020 Å, 1.430 Å, and 1.170 Å. The squared-sine ratios, when referenced to the first reflection, follow the sequence of 1:2:3, which is characteristic of the allowed reflections of a body-centered cubic lattice when multiplied by the first allowed value of N = h² + k² + l² = 2. Therefore, the peaks are assigned to the (110), (200), and (211) reflections of α-Fe. The extracted lattice constants are 2.853, 2.860, and 2.865 Å, yielding an average value of 2.859 Å, which is close to the accepted room-temperature value of approximately 2.866 Å for α-Fe. Peak-width calculations based on the observed, instrument-uncorrected FWHM values are included only as illustrative apparent line-broadening indicators. Because an external instrumental standard was not measured under identical conditions, no quantitative coherent-domain size or microstrain is claimed. The Williamson–Hall treatment is therefore used only to demonstrate the sensitivity of size-strain interpretation to peak breadth, profile selection, and the limited number of available reflections. The intensity hierarchy was analyzed using the body-centered cubic-structure factor, reflection multiplicity, Fe atomic form factor, and Lorentz polarization correction. The comparison between calculated and experimental intensities shows stronger disagreement than the lattice-parameter analysis, illustrating the greater sensitivity of intensity analysis to preferred orientation, specimen preparation, peak-profile selection, and background treatment. The work provides a publication-style reconstruction of an iron powder XRD experiment, connecting peak fitting, indexing, lattice-parameter determination, crystallite size estimation, size-strain analysis, and intensity interpretation in a single critical framework. Full article

Drug release from long-acting intravitreal implants is governed by the coupled effects of diffusion, hydrolysis-driven erosion, and progressive shrinkage of the polymeric matrix. To capture these mechanisms, we solve the diffusion equation in a cylindrical domain whose radius decreases according to the hydrolytic degradation kinetics of PLGA, which follow a pseudo-first-order behaviour in aqueous excess. The resulting formulation combines a modal Bessel expansion with an erosion-controlled time transformation, allowing the evolving geometry and the attenuation of the diffusion modes to be incorporated in a fully mechanistic manner. Within this framework, the shrinkage parameter p quantifies the rate of erosion-induced geometric evolution and enables an accurate reconstruction of the experimental dexamethasone release profile. The solution reproduces both the initial fast-release phase and the extended depletion tail from which the characteristic times $t_{0.50}=264.3\mathrm{h}$ and $t_{0.90}=996.5\mathrm{h}$ are extracted, providing compact and physically meaningful indicators of the transition between early and late kinetic regimes. Overall, the approach offers a robust and interpretable description of drug release from shrinking polymeric systems and is directly applicable to the design of long-acting intravitreal therapies. Full article

This study examines interactions in aqueous electrolyte solutions using the equations of the Scatchard and Friedman models. The six mixing parameters of the Scatchard model, b_AB⁽⁰¹⁾; b_AB⁽⁰²⁾; b_AB⁽⁰³⁾; b_AB⁽¹²⁾; b_AB⁽¹³⁾ and b_AB⁽²³⁾, were obtained from the literature and estimated by processing experimental results measured by the isopiestic method for osmotic coefficients of three-component systems: {yKCl + (1 − y)K₂HPO₄} (aq), {yKBr + (1 − y)K₂HPO₄} (aq), {yKNO₃ + (1 − y)K₂HPO₄} (aq), {yK₂SO₄ + (1 − y)K₂HPO₄} (aq), {yKH₂PO₄ + (1 − y)K₂HPO₄} (aq) and {yNaH₂PO₄ + (1 − y) K₂HPO₄} (aq) at 298.15 K. The Friedman parameters were calculated from the adopted Scatchard parameters as functions of ionic strength. The effects of pair, triplet, and quadruplet interactions on the excess Gibbs energy of mixing were analyzed, and the total Gibbs energy of the solutions was determined. In the systems {yKCl + (1 − y)K₂HPO₄} (aq), {yKBr + (1 − y)K₂HPO₄} (aq), {yKNO₃ + (1 − y)K₂HPO₄} (aq), interactions between different anions of the same charge predominate. Triplet interactions dominate in the system {yK₂SO₄ + (1 − y)K₂HPO₄} (aq). The strongest contributions of triplet and quadruplet interactions are observed in the {yKH₂PO₄ + (1 − y)K₂HPO₄} (aq) system, whereas pair interactions between the same ion pairs are dominant in the {yNa₂HPO₄ + (1 − y)K₂HPO₄} (aq) system. Full article

Most research on Nanobubble Water (NBW) for lignocellulose conversion has focused on anaerobic digestion, with relatively few studies examining direct pretreatment methods such as liquefaction. In this study, we explored the potential of NBW as a pretreatment method for bagasse meal liquefaction. Bagasse meal was treated with oxygen, nitrogen, and carbon dioxide NBW. Changes in component composition ratio, crystallinity, total crystallinity index (TCI), lateral order index (LOI), pyrolysis peak temperature (T_max), and apparent activation energy (E_a′) were investigated. For the liquefaction process, changes in residue content, apparent liquefaction reaction rate constant (k′), average molecular weight, hydroxyl value, functional group information, and T_max of the liquefied residue were examined. Results showed a 2.0–3.0% decrease in cellulose and a 1.0–4.0% decrease in lignin. Crystallinity increased by 5.8–12%, TCI decreased by 6.7–13%, and LOI increased by 4.7–9.8%. T_max decreased by 1.6–3.5 °C, and E_a′ decreased by 1.9–2.8%, both reaching their lowest values with carbon dioxide NBW. At this time, the residue content decreased by 6.3–19%, and k′ increased by 50%. These findings indicate that NBW pretreatment is a promising approach for liquefying bagasse meal under laboratory conditions. Future investigations will be directed towards its scalability and economic viability. Full article

Propagation anisotropy in one-dimensional conductors is commonly interpreted within the framework of Tomonaga–Luttinger liquid theory, where electron–electron interactions lead to distinct channel-dependent excitation velocities. In this work, we investigate a complementary mechanism in which propagation anisotropy arises from structured substrate modulation. We develop a quantitative two-channel effective-medium transport model incorporating position-dependent dielectric and magnetic coupling terms within an effective spinor Hamiltonian. Under an adiabatic envelope approximation, the coupled spinor dynamics are reduced to an effective scalar propagation equation suitable for numerical simulation. The model predicts that spatial modulation of substrate response can generate measurable channel-dependent velocity splitting, wave-packet deformation, and propagation delay. Numerical simulations show that dielectric modulation, magnetic modulation, relative phase shifts, and moderate disorder influence transport anisotropy in distinct and tunable ways. For experimentally realistic parameter ranges, the predicted propagation delay lies in the picosecond regime over micrometer-scale transport distances. Comparison with conventional Tomonaga–Luttinger liquid theory suggests that substrate-induced effects may coexist with intrinsic many-body interactions and contribute appreciably to observed transport behavior. The proposed framework provides a quantitative phenomenological tool for analyzing substrate-controlled anisotropic transport and offers experimentally testable predictions for low-dimensional quantum systems. Full article

Molten salts are increasingly regarded as promising fluids for high-temperature heat transfer, thermal energy storage, and advanced reaction processes, including concentrated solar power (CSP), molten salt oxidation (MSO), and next-generation nuclear reactors. Among these materials, the ternary eutectic mixture Li₂CO₃–Na₂CO₃–K₂CO₃ (32.12–33.36–34.52 wt%) has emerged as a leading candidate due to its wide operating temperature range and favourable thermodynamic properties. Despite its relevance, substantial inconsistencies and gaps remain in the available thermophysical property data, posing challenges for reliable design, modelling, and industrial deployment. This work revisits the Li₂CO₃–Na₂CO₃–K₂CO₃ eutectic through a critical assessment of the literature from its reported melting point at 670 K (397 °C) up to approximately 1200 K (927 °C). Using a methodology inspired by IUPAC-supported strategies previously applied to common liquids such as water and hydrocarbons, we examine the quantity, quality, and coherence of existing measurements. Reference correlations are proposed only where the data are sufficiently robust to justify them. The analysis highlights a pressing need for more accurate and comprehensive measurements—particularly for heat capacity, thermal conductivity, and viscosity—to enable the development of reliable standard reference correlations. Brief recommendations are given on the measurement methods that should be used in high-temperature measurements, namely for heat capacity, viscosity, and thermal conductivity. Reliable thermophysical property data for (LiNaK)₂CO₃ remain limited and inconsistent, despite its relevance for high-temperature energy applications. Density data are comparatively robust, but heat capacity, thermal conductivity, and viscosity still require high-accuracy measurements at elevated temperatures. Addressing these data deficiencies is essential for advancing the safe and efficient use of molten carbonates in high-temperature energy technologies. Full article

Cucurbiturils (CB[n]) are rigid glycoluril-based macrocycles possessing well-defined hydrophobic cavities capable of forming stable host–guest complexes in water. Owing to these properties, CB[n]-containing supramolecular polymer gels have attracted increasing attention as functional composite materials in modern materials science. This review summarizes recent progress in the development of cucurbituril-based supramolecular gels, with particular attention to synthetic approaches, network design, and emerging applications. Both conventional acid-catalyzed methods and more sustainable synthetic strategies for cucurbituril preparation and functionalization are discussed. We further consider the role of CB[n] macrocycles as reversible crosslinking units in polymer networks and analyze how host–guest interactions influence the mechanical properties, self-healing behavior, and stimuli responsiveness of the resulting materials. Recent applications in biomedical engineering, soft electronics, and environmental remediation are also highlighted, demonstrating how molecular-level supramolecular interactions can determine the macroscopic performance of these composite systems. The review concludes with perspectives on scalable synthesis, processing integration, and future directions in supramolecular composite materials. Full article

Dye removal from aqueous solutions remains a major global environmental challenge. Among the various remediation techniques, adsorption using natural materials has gained significant attention. In this study, the adsorption of methylene blue (MB) by a natural volcanic rock (VR) adsorbent—collected from the Cerro Blanco volcano in Divisaderos, Sonora, Mexico—was investigated, and the process efficiency was evaluated at different temperatures. The comprehensive characterization revealed a rough and irregular porous surface via SEM, while the EDS elemental data and the CIPW normative calculations identified the material as a silica-saturated tholeiitic basalt, primarily composed of bytownite (${An}_{71})$ and pyroxenes. This petrological classification was cross-validated by XRD and FTIR spectra, which exhibited vibrational modes characteristic of mafic silicate. The surface analysis via the BET method indicated a specific surface area of 12 m²·g⁻¹, while a BJH analysis indicated a mesoporous structure (average pore diameter of 3.75 nm), and a Type IV isotherm with H3-type hysteresis, suggesting narrow, slit-shaped pores. Batch adsorption experiments demonstrated an exceptional removal efficiency of 99.99% for 50 mg·L⁻¹ MB within only 30 min. The equilibrium data and the adsorption kinetics followed the Langmuir isotherm and a pseudo-second-order model, respectively. Cytotoxicity assays confirmed the VR is biosafe. The combination of high removal efficiency, low cost, and environmental safety positions this material as high-potential adsorbent for sustainable water remediation processes. Full article

A novel mesoporous XAD-based adsorbent (A336/XAD-7) was produced by impregnating the ionic liquid A336 into the pores of XAD-7 resin and used to separate tin(IV) and cadmium(II) from high-level liquid waste (HLLW). The as-produced material was characterized by SEM-EDS, TG-DSC, and N₂ adsorption–desorption isotherms, which revealed a well-developed open pore structure, high loading capacity, and large specific surface area. Adsorption performance analysis showed that in 4 M HCl solution, the experimental saturated adsorption capacity qexp of A336/XAD-7 for Sn(IV) and Cd(II) were 39.51 mg/g and 34.18 mg/g, respectively, with equilibrium reached within 120 min. Among ten coexisting metal ions (Sn⁴⁺, Cd²⁺, Co²⁺, Ni²⁺, Cu²⁺, Eu³⁺, Y³⁺, Ca²⁺, Mg²⁺, Al³⁺) in HLLW, A336/XAD-7 exhibited excellent selectivity for Sn(IV) under high acidity, with a separation factor (SF_Sn/others) of 13.13. Column experiments further evaluated the dynamic separation of Sn(IV) from simulated HLLW using A336/XAD-7, achieving an enrichment factor greater than 7. XPS spectra indicated that the adsorption mechanism involved anion exchange between A336/XAD-7 and the complex anions SnCl₆²⁻ and CdCl₄²⁻. This work demonstrates the application potential of A336/XAD-7 for HLLW treatment and provides valuable guidance for the efficient separation of other metal ions. Full article

Accurate estimation of molecule–nanotube interaction energies is critical for the computational screening of carbon-based materials for hydrogen storage; however, density functional theory (DFT) calculations remain computationally expensive for extensive configurational sampling. In this work, we develop a multilayer perceptron (MLP) surrogate model to predict H₂–CNT interaction energies, represented by $E_{ads}$, for H₂ interactions with a chiral (2,1) carbon nanotube. A curated dataset comprising 696 configurations was generated using DMol³ (BIOVIA Materials Studio), varying intermolecular distance, molecular orientation, and interaction site across three regions: internal cavity, edges, and external surface. The proposed MLP architecture (64–32–1) incorporates GELU activation functions, L2 regularization, and dropout to improve generalization. The model achieves coefficients of determination in the range R² = 0.90–0.96 across all interaction regions, with particularly strong performance at the nanotube edges (R² = 0.9358, MSE = 0.046 eV²), as well as on the external surface (R² = 0.9625, MSE = 0.574 eV²) and within the internal cavity (R² = 0.9051, MSE = 1.506 eV²). The original $E_{ads}$ distribution had a mean of 4.0955 eV and a sample standard deviation of 4.3189 eV. The elevated energy values observed in the internal cavity (up to 12 eV) are consistent with steric repulsion induced by geometric confinement rather than predictive artifacts. The trained MLP showed close agreement with DFT-derived trends, enabling exploration of interaction-energy landscapes spanning both attractive and repulsive regimes. These results indicate that MLP-based models trained on diverse configurational datasets provide a computationally efficient alternative for screening carbon nanostructures in hydrogen storage applications, without substantially compromising accuracy relative to first-principles methods. Full article

In this work, an ultrasonic relaxation spectroscopic study of aqueous ammonium Fe(II) sulfate, aqueous ammonium Fe(III) sulfate and the corresponding ternary system has been undertaken. A variety of acoustic parameters including relaxation frequency, relaxation amplitude and speed of sound were determined as a function of solution concentration. In addition, the adiabatic compressibility and the molar volume change during the ionic association in ammonium Fe(II) sulfate and ammonium Fe(III) sulfate aqueous solutions were also estimated from the acoustic data. This approach facilitated a comprehensive characterization of the three systems across different concentrations. In the two binary systems, the presence of an ion association mechanism was identified involving the divalent and trivalent iron ions, with the sulfate anions, respectively. Furthermore, in the ternary system, an internal sphere oxidation–reduction mechanism occurred between the divalent and trivalent iron ions. All ions within each solution play an active role in shaping the structure of water molecules, owing to the prevailing kosmotropic characteristics specific to each solution. The results are examined within the context of the current phenomenological understanding in the field. Full article