INSPIRE Grant Report: Dr. Artem Fediai
Published on in NEWS
The INSPIRE Grants Programme supports cfaed members who want to pass a research stay abroad. By the help of this programme, Dr. Artem Fediai (Carbon Path) stayed at King’s College in London from beginning of February 2016 until the beginning of April. His aim was to work on the integration of the ab initio quantum transport models developed within Carbon path with the free quantum chemistry simulation package CP2K. Read his short report about the British adventure:
The campus I was visiting lies in the heart of London, the city of Westminster. In terms of physics, King’s College ancient traditions. It is famous, for instance, by James Maxwell who worked there and Peter Higgs, who was a King’s student. I was visiting the group of Prof. Lev Kantorovich, a condense matter physicist, who originally represents a famous Soviet school of physics. My main collaborator was Dr. Lianheng Tong who received his doctoral degree in physics from Cambridge University being at a bachelor level rather a pure mathematician. This combination of these qualifications of my hosts suited well the goal we were aimed at.
During my work in the Carbon Path of cfaed I have developed and implemented numerically a method allowing parameter-free quantitative simulations of the extended CNT-metal contacts, which was my task within the Carbon Path. It is partially based on the CP2K simulation tool which is extremely well parallelized, but my own codes were disintegrated from the CP2K, and were only slightly parallelized (24 cores maximum). This makes it impossible to perform the nonequilibrium simulation of the CNT-metal contacts due to prohibitively large simulation time. None of other quantum chemistry code has this theoretical approach implemented. Therefore, I decided to make it myself using the support of the CP2K developers, who knows the structure of the code. That was the reason I have applied for this visit.
During my stay in King’s we have developed the structure of the input-output data, add a new simulation type to the cp2k and developed the most critical numerical algorithms as well as parallelization strategies. Before I left, we have created the shared repository for our scripts in github which will allow us to work and contribute to this project together until it is completed. After that, we can use CP2K to perform unique and high-performance atomistic simulations of the carbon-based devices, developed within the Carbon Path.
Dr. Artem Fediai