English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

The high computational cost of evaluating atomic interactions recentlymotivated the development of computationally inexpensive kinetic models, whichcan be parametrized from MD simulations of complex chemistry of thousands ofspecies or other processes and accelerate the prediction of the chemicalevolution by up to four order of magnitude. Such models go beyond the commonlyemployed potential energy surface fitting methods in that they are aimed purelyat describing kinetic effects. So far, such kinetic models utilize moleculardescriptions of reactions and have been constrained to only reproduce moleculespreviously observed in MD simulations. Therefore, these descriptions fail topredict the reactivity of unobserved molecules, for example in the case oflarge molecules or solids. Here we propose a new approach for the extraction ofreaction mechanisms and reaction rates from MD simulations, namely the use ofatomic-level features. Using the complex chemical network of hydrocarbonpyrolysis as example, it is demonstrated that kinetic models built using atomicfeatures are able to explore chemical reaction pathways never observed in theMD simulations used to parametrize them. Atomic-level features are shown toconstruct reaction mechanisms and estimate reaction rates of unknown molecularspecies from elementary atomic events. Through comparisons of the model abilityto extrapolate to longer simulation timescales and different chemicalcompositions than the ones used for parameterization, it is demonstrated thatkinetic models employing atomic features retain the same level of accuracy andtransferability as the use of features based on molecular species, while beingmore compact and parametrized with less data. We also find that atomic featurescan better describe the formation of large molecules enabling the simultaneousdescription of small molecules and condensed phases.

Atomic-Level Features for Kinetic Monte Carlo Models of Complex Chemistry from Molecular Dynamics Simulations