化学
催化作用
钴
氨生产
纳米颗粒
镧
氮化物
多相催化
活化能
金属
化学工程
屏障激活
氢
纳米技术
分子动力学
氨
产量(工程)
封装(网络)
氮气
无机化学
协同催化
金属有机骨架
反应机理
计算化学
过渡金属
组合化学
化学物理
化学稳定性
反应中间体
作者
Jie Luo,Junyi Yang,Chuwei Zhu,Yi Huang,Jin-Xun Liu,Wei-Xue Li
摘要
Developing catalysts for ammonia synthesis under mild conditions is of paramount importance but remains a grand challenge, primarily due to the trade-off between efficient dinitrogen activation and hydrogen poisoning. Nitride-supported metal catalysts exhibit remarkably low-temperature activity, yet the atomic origin of their synergy remains a subject of intense debate. Building on our recently established generalized theories of strong metal–metal interaction (SMMI) and encapsulation, we combine ab initio atomistic thermodynamics and machine-learning potential-driven molecular dynamics simulations to systematically construct a unified mechanistic framework for these highly active systems. Using lanthanum nitride (LaN)-supported cobalt (Co) nanoparticles as a prototypical model, we reveal that, driven by SMMI and facilitated by nitrogen vacancies, the nitride support spontaneously restructures to encapsulate metal nanoparticles via the formation of subnitride overlayers. This thermodynamically stable encapsulated architecture constructs contiguous perimeter sites─involving both low-valent La cations and electron-rich cobalt metal─that fundamentally alter the catalytic landscape, while intrinsically securing the antisintering stability of small Co nanoparticles. First-principles calculations reveal that these perimeter sites cooperatively facilitate N2 activation while simultaneously suppressing H-poisoning. Consequently, microkinetic simulations yield an apparent activation energy of 50 kJ mol–1, in good agreement with experimental values. By naturally accounting for a diverse array of experimental observations, this work establishes SMMI-driven encapsulation as a robust physical framework for understanding and designing nitride-based catalysts for ammonia synthesis under mild conditions.
科研通智能强力驱动
Strongly Powered by AbleSci AI