Advanced Computational Platform for Electron Transport in Electronic Devices and Biological Sensors (ACRAB)
Prof. Stefano Sanvito, School of Physics and CRANN, Trinity College Dublin
The total solar energy absorbed on earth in one year is about twice as abundant as the total that will be ever extracted from all non-renewable sources combined (petrol, natural gas, nuclear etc.). One hour of solar irradiation surpasses the current yearly world energy consumption. Interestingly even gas and oil are the ultimate products of solar energy conversion through the plants life cycle. Gas and oil however have formed over millions of years. Can we now harvest efficiently this enormous amount of energy at the speed required by our society in a sustainable and environmentally friendly way? This is where computational science can help.
A newly established research collaboration between the Centre for Research on Adaptive Nanostructures and Nanodevices (CRANN) at Trinity College (Ireland) and the King Abdullah University of Science and Technology (KAUST, Saudi Arabia) has the potential of solving some of the technological issues on the way to massive solar energy production. The joint CRANN-KAUST project aims to understand the microscopic origin of the present limitations of solar energy conversion. This will pave the way for a new generation of photovoltaic devices, more efficient and cheaper than those currently available. Such an advance will be achieved by means of sophisticated high-end computational simulations based on fully quantum mechanical theory. The CRANN-KAUST project however does not stop here. The computational tools developed in Prof. Sanvito's group at CRANN are in fact so versatile that a much broader range of devices can be simulated. Thus, together with photovoltaic cells, also novel bio-sensors and chemical sensors will be investigated. The project, sponsored by KAUST, is worth 1,500,000 US dollars, it will involve 2 senior researchers and three PhD students both at KAUST and TCD, and it will avail of the computational power of the KAUST supercomputer Sasheen.
Simulations of a polymeric DNA molecule transmigrating across a silicon nitrate nanopore, which has been functionalized to accommodate two carbon nanotubes. As the DNA strand passes through the pore the electronic tunnel current across the carbon nanotube is measured. This depends on the particular basis occupying the nanotube, so that the device can be used as a tool for DNA sequencing
The computational core of the project is the Smeagol code (www.smeagol.tcd.ie). Smeagol was created about 5 years ago to fill a wide gap in the area materials modelling, namely that of electron transport simulations for nano-devices. Intriguingly in almost all modern electronic devices (transistors, read-head for hard disk drives, sensors, etc.) the signal to detect is an electrical current. When the device has nanometer (one billionths of a meter) size, the current is no longer a classical object and needs to be calculated by using quantum mechanics. Furthermore one needs to include in the calculation a precise description of the device at the atomic level since at such a small length scale a single atom can significantly change the properties of the device itself.
Over the last few years Smeagol has developed considerably from its initial core thanks to a number of Science Foundation of Ireland sponsored projects and collaboration with the Trinity Centre for High Performance Computing. At present calculations for devices comprising an excess of 10,000 atoms are possible but the next development will reach the 100,000 mark. This is roughly the total number of atoms forming the active region of the next generation transistors. As such it represents a milestone where computer simulations will be able to predict the behavior of the real transistors, which form the computer itself. How this goal will be achieved?
Most of the development effort will be carried out within the CRANN-KAUST collaboration. Prof. Stefano Sanvito and Prof. Udo Schwingenschlögl will lead the CRANN and KAUST teams respectively, while Dr. Ivan Rungger, also in CRANN, will overlook the technical aspects. Furthermore collaboration with TCHPC and the KAUST supercomputer team will be in place. The research programme develops three main aspects: 1) to make the Smeagol code scalable to large computational infrastructures so that systems comprising in excess of 100,000 atoms can be tackled, 2) to include in Smeagol tools capable of dealing with strong electron-electron correlations so that effects such as Coulomb blockade, Kondo and electron transport in photo-excited systems will be investigated, 3) to couple Smeagol with molecular dynamics to study electron transport in solutions.
For the first time we see the possibility of simulating real devices, including in the simulations all the atoms that compose them. This is revolution with respect to the past and it is exciting to be part of such an adventure !!
Potential drop profile across a devices made from gold electrodes and a spin-crossover molecule. Spin crossover molecules are compounds, which can change their spin states with an external stimulus. They are emerging as potential candidates for new combined memory and logic devices.