Abstract Wide-bandgap perovskites (1.6-1.8 eV) are promising top sub-cells for tandem and indoor photovoltaics, but high bromine content induces halide segregation and accelerated degradation. In this work, we systematically investigate three bulk modifiers: Ti3C2Tx (2D flakes), Methylammonium chloride (crystallization additive), and the self-assembled monolayer 4-(pyridin-4-yl)triphenylamine in CsFAPb(I-Br)3 (1.72 eV) p-i-n solar cells. Notably, Ti3C2Tx modified devices deliver a champion power conversion efficiency of 18.84 % for 1.72 eV wide-bandgap perovskites, maximizing efficiency through optimized energy-level alignment and reduced series resistance. To address operational stability, we identified two distinct ion-migration-driven degradation pathways. Local ion interactions at the hole-transport interface causes void formation and photocurrent loss, which is effectively suppressed by methylammonium chloride. In contrast, ion accumulation at extraction interfaces primarily degrades the fill factor. By immobilizing mobile ions, the 4-(pyridin-4-yl)triphenylamine self-assembling monolayer prevents void formation, improves fill factor thermal resilience, and achieves an 80 % operational lifetime of over 800 hours under continuous stress (ISOS-L-2, 65 °C). These findings demonstrate that managing ionic dynamics and passivating traps are complementary strategies for stabilizing wide-bandgap perovskite solar cells.
Abstract In situ transmission electron microscopy has significantly advanced catalysis research by enabling high-resolution, real-time investigation of catalytic materials under realistic operating conditions, overcoming the fundamental limitations of conventional ex situ characterization. This review systematically summarizes the state of the art in in situ transmission electron microscopy for catalysis. Representative technological innovations in in situ gas and liquid transmission electron microscopy have enabled direct observation of previously elusive dynamic phenomena, including surface reconstruction, metal-support interaction, and active site evolution, thereby providing mechanistic insights at high resolution. The integration of complementary spectroscopic techniques and emerging analytical methods has further empowered the correlation of structural dynamics with catalytic performance. Despite these advances, fundamental challenges persist in temporal resolution limits, electron-beam-induced perturbations, and high-dimensional data interpretation. By establishing direct connections between microscopic catalyst dynamics and macroscopic catalytic performance, in situ transmission electron microscopy has become highly valuable for developing high-performance catalysts for clean energy applications, including water electrolysis, fuel cells, and CO2 reduction. This review aims to bridge the gap between fundamental characterization and rational catalyst design, offering actionable insights for the development of high-efficiency catalytic systems.
Abstract Artificial intelligence has widely optimised material screening, reaction tuning and unit-level digital twins in chemical and energy industries, facilitating clean processes. Yet existing tools fail to support high-level strategic decisions on product portfolios and cross-site production networks. Fragmented cross-departmental data and static planning models delay agile responses to volatile feedstock, demand and logistics conditions. Drawing on knowledge-graph process twin frameworks, this commentary argues industrial artificial intelligence should evolve from isolated process optimisation to enterprise-wide portfolio-network intelligence. Machine-readable unified knowledge bases integrating products, assets, supply chains and carbon constraints underpin scalable business digital twins, enabling consistent, traceable cross-functional optimisation. Such infrastructure empowers artificial intelligence to evaluate technical feasibility alongside profitability, asset utilisation and carbon performance under uncertainty, guiding optimal product selection and site allocation. This paper concludes that structured corporate knowledge infrastructure, rather than standalone algorithms, constitutes the long-term foundation for artificial intelligence-driven holistic business decision-making across chemical value chains.
Abstract Accurate estimation of the state of health for lithium-ion batteries is crucial for battery management systems. To address the conflict between deep feature extraction and physical priors in single-path models, and the inability of fixed aging parameters to adapt to the dynamic degradation of lithium-ion batteries under real electric vehicle operating conditions, this paper proposes a dual-path physics-informed neural network for state-of-health prediction. First, the peak and area are extracted as aging features from the incremental capacity curves. Second, a dual-path parallel network architecture is constructed, comprising a feature attention-gated residual network prediction main branch and a dynamic parameter branch. Third, the universal differential equation paradigm is introduced to transform traditional fixed-parameter physical aging equations into a dynamic partial differential system driven by real-time health characteristics. Finally, the method is verified on the public Center for Advanced Life Cycle Engineering single-cell battery and self-assembled LBP battery pack dataset. Experimental outcomes demonstrate that the dual-path physics-informed neural network substantially enhances state-of-health estimation performance, with average mean absolute errors as low as 0.6920% and 0.2937%, respectively.
Abstract Legacy utility boilers often provide limited access to operational data and lack integrated measurement systems, while their combustion settings may have deviated from the optimal operating point established during commissioning. To address this information-limited condition and evaluate the thermal consequences of changes in combustion on the water-steam side, this study suggests a measurement-constrained integrated workflow to optimize combustion control using the minimum available boiler operating data. The integrated workflow couples a proposed zonal heat-generation model with a water-steam model, uses steam-side measurements to evaluate inferred zone heat absorption and representative thermal states, links the resulting thermal information to the environmental-emissions model as well as efficiency calculation, and then optimizes excess air and fuel-air control coefficients. The proposed framework was evaluated and validated under steady-state conditions using operational data from a 250 MW power plant, through direct validation studies on the heat generation side and indirect validation studies on the water-steam side. At nominal load, the selected multi-objective operating point reduced the NOx level from 375 ppm to 330 ppm, while maintaining efficiency; the computed +0.0143% change shows no efficiency penalty. The results show that the proposed integrated workflow can support steady-state NOx–efficiency optimization with limited operational data and provide a reusable workflow architecture for other boiler applications subject to appropriate target-system configuration and validation.
