The microstructure distribution of multiphase porous media has important effects on its macroscopic properties. In this paper, we propose a framework for multiphase reconstruction based on entropy st...

Nature Communications, Published online: 04 December 2025; doi:10.1038/s41467-025-67020-zAqueous zinc metal batteries often face performance decay under thermal environments, originating from side reactions caused by aqueous electrolytes. Here, authors report an organic-inorganic-H2O multiphase aqueous soggy-sand electrolyte that reconstructs hydrogen-bond networks, eliminates parasitic reactions, and promotes uniform Zn deposition at elevated temperature (140 oC).

European Radiology -

Nature Communications, Published online: 04 December 2025; doi:10.1038/s41467-025-67020-zAqueous zinc metal batteries often face performance decay under thermal environments, originating from side reactions caused by aqueous electrolytes. Here, authors report an organic-inorganic-H2O multiphase aqueous soggy-sand electrolyte that reconstructs hydrogen-bond networks, eliminates parasitic reactions, and promotes uniform Zn deposition at elevated temperature (140 oC).

**Abstract:** This paper introduces a novel framework for predicting and ultimately controlling instability in multiphase flows within nanoporous media, a critical challenge in enhanced oil recovery, carbon sequestration, and microfluidic devices. The presented methodology, termed Hybrid Temporal Graph Neural Network with Physics-Informed Reinforcement Learning (HTGNN-PI-RL), integrates the representation power of Graph Neural Networks with the constraints of underlying physics through reinforcement learning, achieving a 15% improvement in stability prediction accuracy compared to traditional computational fluid dynamics (CFD) simulations.

Ecology and Evolution, Volume 15, Issue 12, December 2025.

**Abstract:** This paper proposes a novel methodology, Dynamic Adaptive Resonance Theory (DART), for the real-time stabilization of multiphase flow instabilities within nano-porous media. Leveraging established computational fluid dynamics (CFD) and machine learning techniques, DART dynamically adjusts local permeability and interfacial tensions to suppress instability onset, enhancing oil recovery and chemical transport efficiency. Existing approaches often rely on computationally expensive simulations or require predefined control strategies; DART offers a real-time, adaptive solution with estimated 15% improvement in multiphase flow stabilization compared to passive control methods.

This study used computational fluid dynamics (CFD) modeling to develop prototype almond harvester designs to reduce dust emissions and power demand while maintaining the quality of the harvested product. This work evaluated particle collect...

Computational fluid dynamics (CFD) modeling was evaluated for a physical prototype nut harvester as a method to determine model effectiveness for predicting dust emissions and reduced power demand in terms of pressure drop. Particle-laden a...

**Randomly Selected Sub-Field:** 촉매 반응에서 용매의 결정화 및 자기 조립 제어 (Controlled Crystallization and Self-Assembly of Solvents in Catalytic Reactions) **Generated Research Topic:** **"Kinetic Orchestration of Multiphase Catalysis: A Data-Driven Approach to Solvent-Mediated Crystal Nucleation and Self-Assembly for Enhanced Reaction Selectivity."** ## Kinetics Orchestration of Multiphase Catalysis: A Data-Driven Approach to Solvent-Mediated Crystal Nucleation and Self-Assembly for Enhanced Reaction Selectivity…