This website requires JavaScript.

Unveiling the hidden reaction kinetic network of carbon in water with unsupervised machine learning

Chu LiYuan YaoDing Pan
Feb 2024
The dissolution of CO$_2$ in water followed by the subsequent hydrolysis reactions is of great importance to the global carbon cycle, and carbon capture and storage. Despite enormous previous studies, the reaction pathways are still not fully understood at the atomistic scale. Here, we combined \textit{ab initio} molecular dynamics simulations with Markov state models to elucidate the reaction mechanisms and kinetics of CO$_2$ in supercritical water both in the bulk and nanoconfined states. The integration of unsupervised machine learning with first-principles data allows us to automatically identify complex reaction coordinates and pathways instead of \textit{a priori} human speculation. The pyrocarbonate anion (C$_2$O$_5^{2-}$(aq)) was previously hypothesized to have a fleeting existence in water; however our study reveals that it is a crucial reaction intermediate in the nanoconfined solutions. We found that the extreme confinement can enhance the stability of C$_2$O$_5^{2-}$(aq) and even the pyrocarbonic acid (H$_2$C$_2$O$_5$(aq)), which was unknown in water. The unexpected appearance of pyrocarbonates is related to the superionic behavior of the confined solutions. Our study highlights the importance of large oxocarbons in aqueous carbon reactions, with great implications for the deep carbon cycle and the sequestration of CO$_2$.
发布时间 · 被引用数 · 默认排序
发布时间 · 被引用数 · 默认排序