Building a strong CP2K user community

Author: Iain Bethune
Posted: 4 Mar 2016 | 12:42

CP2K logoLast week we held the 3rd annual CP2K users group meeting down at Kings College London.  Amazingly, we are already half-way through the 5-year 'CP2K-UK' project - the EPSRC-funded community support effort that I'm leading - how time flies!  It was great to see these meetings continue to go from strength to strength.  This year we had over 50 people there on the day from around the UK but also a significant proportion from overseas too!  While the primary aim of our activities is to support the UK research community, if we have a wider impact that's of course a bonus.  

There were several aspects of the meeting which I thought reflect the progress we have made as the 'CP2K community':

Citation rate of MOFs

Firstly, all of the talks led to discussions of some sort or another.  Sanliang Ling's talk on Hybrid Functionals in CP2K was on the programme as this was the number one issue that users wanted to hear about, and there were lots of practical questions about what can be done and how.  I had invited Francois-Xavier Coudert from CNRS (a high profile CP2K user in the area of porous materials, also a free-software advocate and gfortran contributor!) to speak, and not only did he talk about some successes, but also open challenges looking for suggestions - as well as providing 'evidence' that the only topic that will be published on by 2020 will be MOFs (see picture to right!)

Second, it was a pleasure to have Tobi Krämer from Heriot-Watt talking about his experiences using CP2K for modelling Rhodium-alkane sigma-complexes.  Tobi is part of the Macgregor group, who were the first group we did a training day with as part of the CP2K-UK project.  Within the space of 24 months, from previously using gas-phase codes such as Gaussian they have learned how to do condensed-phase DFT calculations with CP2K, have built a close collaboration with CP2K developers in Zurich, and with the help of an ARCHER RAP grant just published their first paper using CP2K.  Despite Tobi's protestations, he is now the de facto expert on doing reliable vibrational analysis calculations with CP2K!

80000 atoms DFT, high accuracy settings. Aggregated nanoparticles in explicit solution Relevant for 3rd generation solar cellsFinally, feedback from participants said there were two reasons why these meetings are really useful - hearing from the CP2K developers what is new and upcoming, but also making connections with other users.  This is a real sign that the community is growing to a point where it no longer needs to rely solely on the support of the CP2K developer group, but is becoming self-sustaining - which was the reason for setting up the CP2K-UK project in the first place!

Following the user group meeting, we had discussions with the other project partners about what we will do in the next year of project - we have plenty of exciting plans, not least with additional funding from two eCSE projects, and new collaborations with other code developers, but I think I'll leave those for another post...

Image (above left): 80000 atoms DFT with CP2K, high accuracy settings. Aggregated nanoparticles in explicit solution. Relevant for 3rd generation solar cells