Here is an announcement to whet appetite of DFT practitioners looking for more accurate tools to study strongly correlated systems. At the moment the only tool available in Materials Studio is CASTEP DFT+U implementation which is just an approximate step towards a better description of, for example, f-electron materials. Now thanks to the research collaboration between the group of Dr Cedric Weber at King's College London and CASTEP developers in York and Durham there is a prototype implementation of various flavors of DMFT in CASTEP. The paper published in August 2018 in Physical Review B (https://doi.org/10.1103/PhysRevB.98.075129, preprint at https://arxiv.org/pdf/1804.06805.pdf) gives details of the implementation as well as the first results produced with this code.

DMFT, as expected, improves description of lattice parameters and bulk modulus for f-electron systems such as SmTe, various phases of Ce or cerium sesquioxide. The underlying reason is a stronger localization of f-states as predicted by the DMFT treatment. Another effect of capturing many-body effects on the DMFT level is a more accurate description of atomic forces illustrated by DFT vs DMFT comparison in cerium sesquioxide.

There is still a lot of work to be done to bring this development to Materials Studio users, but this is an extremely encouraging first glimpse of future CASTEP applications "beyond DFT".

Victor Milman, BIOVIA Science Council