Sreegopal Banerjee College

The catalytic hydrogenation of ethylene (C2H4) on unsupported titanium oxide clusters (TixOy) was investigated using density functional theory to examine the roles of oxygen deficiency, cluster size, and Pt doping. Oxygen-deficient Ti3O5 + generate localized centers that enhance H2 activation and C & boxH;C hydrogenation, with Ti3O5 + showing the lowest intrinsic barrier (2.09 eV). Reactivity improves with cluster size up to Ti4O7 + and then saturates. Single-atom Pt doping forms a bifunctional Pt-O interface that simplifies the mechanism and significantly lowers barriers. PtTi3O5 follows a cooperative Pt-H/O-H pathway, while PtTi4O7 exhibits a metal-centered route with barriers as low as 0.60-0.81 eV. These results demonstrate that oxygen vacancies and Pt doping effectively tune Ti-O clusters toward noble-metal-like hydrogenation activity.

Pneumonia produces $\approx \mathbf{2. 5 M}$ demises year after year; computerized chest X-ray (CXR) screening can associate radiologist unavailability in restrained-resource situations. We offer CXR-AttentionNet, an attention-augmented transfer-learning structure assimilating ResNet50V2, a Convolutional Block Attention Module (CBAM), two-step tweaking, and Youden's J threshold adjustment. Thoroughly authenticated via 5-fold stratified crossvalidation on the entire Kermany standard (5,857 imageries), the prototype attains average accuracy $97.85 \% (\pm 0.19 \%)$, AUC $0.9958 (\pm 0.0014)$, and macro F1 0.97. Grad-CAM records reliably focusing respiratory association, assisting experimental explainability. The contemporary work shows statistically lower accuracy by using traditional deep learning model and default 0.5 conclusion threshold neglects the asymmetric cost of unnoticed pneumonia. Most of the designs neglect spatial attention and well-organized explainability. CXR-AttentionNet moderates the ignored gaps. By this study amalgamating the required supports simultaneously on the above issues the Proposed Model establishes that a CBAM amplified ResNet50V2 by means of cautious dual-phase refinement simplifies reliably along with small variance transversely all folds letting the CNN highlight diagnostically appropriate lung sections prior to global pooling with precision and meticulous adaptation. Concurrently the model generates pictorial explainability by Grad-CAM to strengthen medical confidence while producing extraordinary investigative accuracy with interpretability in a regeneratable Cross-Validation pipeline.

Our work provides a detailed mechanistic pathway for the oxygen reduction reaction during the discharging process and the potentiality of a C 2 N monolayer as the cathode catalyst in non-aqueous Li–air batteries.

The photocatalytic nitrogen reduction reaction (NRR) is a promising approach for green and sustainable ammonia (NH3) production under ambient conditions. However, the design of highly efficient photocatalysts remains a significant challenge owing to the inertness of N2 molecules, complex reaction kinetics, and substantial energy barriers. In this work, we investigate the catalytic mechanism and real-time photocarrier dynamics of NRR on pristine and defect-engineered orthorhombic boron nitride (o-B2N2), a metal-free and environmentally friendly 2D semiconductor. Using density functional theory (DFT) and time-dependent ab initio nonadiabatic molecular dynamics (NAMD) simulations, we systematically studied the electronic structure, optical absorption, free-energy profile, and dynamics of charge carrier recombination of both pristine and vacancy-defective o-B2N2. Our results demonstrate that the introduction of a double vacancy (DV-BN) into o-B2N2 significantly widens the band gap, prolongs the electron-hole recombination time from 2.18 ns (pristine) to 3.15 ns, and markedly improves the photocatalytic activity toward NRR. These findings demonstrate the potential of o-B2N2 as an efficient, low-cost photocatalyst for sustainable ammonia production.

Altmetrics is an interesting complementary tool that assesses the productivity of individual authors, journals, institutions, disciplines, and geographic regions in the socio-academic web space. It provides a more comprehensive and extensive purview of how research is disseminated through the academic community in a multifaceted way. This research aims to assess the research impact of the top 20 NIRF-ranking research institutions in India using altmetrics. The research data have been collected from four different databases, namely Scopus, Altmetric.com, Mendeley.com, and Unpaywall.org, through REST/API calls using a data carpentry tool called OpenRefine. The observed data is analyzed using computational methods, and categorical OA altmetrics advantages (COAAA) are introduced to investigate OA (dis)advantages categorically. Findings reflect that 29.51% of articles have been covered in altmetrics and 96.06% of publications in Mendeley. The overall coverage of altmetrics events still needs to be improved and indicates an increasing trend compared to the past. The significant presence of events, like Twitter, Blogs, News, and Facebook, across the institutions has been observed. Moreover, the coverage of OA articles in altmetrics is considerably higher than that of non-OA articles. Among OA categories, gold OA has the most coverage in altmetrics. All Indian Institute of Medical Science Delhi (AIIMS-D) has held the first position in terms of coverage, while the Indian Institute of Technology Hyderabad (IIT-H) ranked top by correlation (rho) values among those institutions. Almost positive correlations are observed between most of the variables. The overall research results suggest discipline- and institute-specific considerations are pivotal when evaluating institutions' productivity using altmetrics.