Machine learning potentials for complex aqueous systems made simple
Researchers from the University of Cambridge, University College London, Imperial College London and Charles University in Prague have developed a powerful machine learning based procedure for molecular simulations of complex systems.
The results, reported in the journal PNAS, open the door to the straightforward exploration of many technologically relevant systems by computer simulations.
Understanding complex materials, in particular those with solid-liquid interfaces, such as water on surfaces or under confinement, is a key challenge for technological and scientific progress in order to tackle some of the fundamental issues of our time such as climate change or clean water supply. Although established simulation approaches have been able to provide important atomistic insight into such systems, they come with intrinsic limitations.
In the published work, it is shown how these limitations can be overcome in a simple and automated machine learning procedure to provide accurate models of interactions, as illustrated for a variety of complex aqueous systems.