|Current Workshop | Overview||Back | Home | Search ||
DFT-based Multilayer Methods for Nanoscale Systems
The field of multiscale modeling is ever growing and broadening as researchers in both academia and industry are looking to gain deeper insight and better predictive power for real-life applications in life sciences, chemistry and material science.
This workshop brought together scientists from diverse fields (quantum chemistry, computational physics and molecular dynamics) to discuss how to further develop multiscale methods. They thereby focused on multilayer methods containing at least one layer treated with a DFT-based electronic structure method.
Particular issues that we were seeking to address:
- coupling strategies for QM/QM and QM/MM simulations in biochemical applications as well as materials science
- dealing accurately with the boundaries between the layers
- specialized polarizable force fields for multi-layer methods
The workshop brought together researchers from different application fields, in particular biochemistry and materials science. As they noted themselves, they develop and use similar methods and techniques, and thus face comparable problems, but they hardly interact with each other because they don't visit the same conferences or publish in the same journals. So one of the main conclusions was that there is a lot to be gained by knowledge transfer between these two seemingly disparate fields. Another issue that was raised a few times was the lack of good benchmark systems for testing developments before they are applied to other real-life systems.
A few specialists from sub-fields (adaptive QM/MM in biochemistry; subsystem DFT/QM-in-QM methods; development of polarizable force field) found each other at the first time and have made (further) plans for follow-up specialized conferences. SCM will host a small scale workshop in February 2014 where one of the topics that was coined will be tackled for implementation in a scientific code (QM-in-QM, or FDE, for periodic systems). For the longer term, plans have been made to assimilate different methodologies in adaptive QM/MM in a similar flexible environment that can be linked to different QM and MM codes.
An issue that came up over and over again is the need for so-called polarizable force fields in QM/MM applications with appropriate treatment of the boundaries (to eliminate ghost forces). This issue awareness was raised and fully subsumed by the participants, but unfortunately no concrete plans or follow-ups have been construed during the conference. The topic itself, how to generate transferable, general polarizable force fields warrants a focused workshop to make headway.
The format of the Lorentz Center@Snellius workshop worked well. The long presentations with explicit room for interruptions and discussions rather than monologues stimulated participation and allowed to gain real insights and to come to new ideas. Consequently, all discussions were more or less done at the end of the day were we scheduled brain storm sessions, which were less effective. Another format that we tried may have worked better there (e.g. specific assignments). We decided to give everyone homework for the last day to write down three take-home action points on the blackboard, which worked brilliantly to activate everyone and also to re-iterate common interests and cross-over points.