Crystal structure prediction tool a ‘significant and thought-provoking advance’

In 1988 Nature’s editor John Maddox laid down the gauntlet to computational chemists when he said that ‘One of the continuing scandals in the physical sciences is that it remains impossible to predict the structure of even the simplest crystalline solids from a knowledge of their composition.’ In the ensuing decades, great strides have been taken to crack that problem.¹ Now, Matthew Rosseinsky, Paul Spirakis and their colleagues at the University of Liverpool, UK, claim to be able to reliably find the lowest-energy (ground-state) structure of a range of crystalline compounds given only their stoichiometric formulae.²

Being able to predict crystal structures would be a tremendous tool for discovering useful new materials. Although many methods now exist for such prediction, one of the big challenges is confidence that the estimated energies of predicted structures have identified the global optimum. The algorithm devised by the Liverpool team now offers ‘optimality guarantees’ for the structures it identifies.