解耦(概率)
堆积
光催化
瓶颈
材料科学
化学物理
共价键
吸附
化学
可转让性
限制
纳米技术
计算机科学
明细余额
背景(考古学)
不完美的
工作(物理)
多孔性
生物系统
功率(物理)
过氧化氢
积分器
不对称
极限(数学)
偏移量(计算机科学)
计算
计算化学
水溶液
作者
Jiehua Ding,Zhendong Luo,Chenxu Wang,Jianyu Han,Nengchao Luo,Junju Mu,Qianrong Fang,Xukai Zhou
摘要
ABSTRACT Deconvoluting the intertwined dynamic bottlenecks of photogenerated charge and reaction substrates remains a fundamental bottleneck in heterogeneous photocatalysis due to the lack of reasonable and effective tools. This study presents a quantitative, interpretable framework to deconvolve the entangled factors limiting solar‐driven photocatalysis. Using Au(I)‐based covalent metal‐organic framework (CMOF‐Au) isomers with controlled stacking (AA, AB, and ABC) as a model platform for hydrogen peroxide photosynthesis, we applied a Z‑score and LOFO (Leave‐One‐Feature‐Out) importance analysis. This approach revealed that the dominant rate‐determining factor shifts with stacking: bulk charge transfer governs performance in AA stacking, whereas surface adsorption limits AB and ABC stacking. The method successfully quantifies the relative contribution weights of key functions (charge dynamics, interfacial transfer, surface adsorption, and mass transport) and demonstrates high predictive power across diverse photocatalysts, including covalent organic frameworks, porous aromatic frameworks, C 3 N 4 , and overall water‐splitting systems. By uncovering counterintuitive mechanisms, such as Simpson's paradox, our work establishes a generalizable strategy for decoupling multivariate factors, moving photocatalyst design from empirical tuning toward a predictive, function‐aware paradigm.
科研通智能强力驱动
Strongly Powered by AbleSci AI