2nd TYC Workshop on Energy Materials: Charge Transfer for Energy Applications
King's College London, United Kingdom
Reducing our dependency on fossil fuels is an urgent and essential step in ensuringenergy security as well as preventing climate change. This requires the design anddevelopment of new materials for alternative and sustainable means of energy generation, storage, distribution and supply. As theoretically minded scientists, we are particularly interested in how theory and computation can contribute to this ambitious goal. It becomes increasingly clear that modelling will play an important role in therational design of materials with desired properties. However, in many areas ofcomputational materials research, the predictive power of existing methods isstill limited. In this respect, the recent surge of funding in the Energy sectorprovides a great opportunity for computational scientists to improve existing methods,and to develop entirely new approaches that have the potential to significantlyreduce the gap between experiments and numerical predictions.
Computational energy materials research is one of the cornerstones of theThomas-Young Centre (TYC, http://www.thomasyoungcentre.org/). The TYC is aLondon-wide interdisciplinary community of research groups working to addresschallenges of society and industry through the theory and simulation of materials.With about 80 participating groups, and ambitious programmes of events, the TYC is asupportive community for research students and young researchers and a source of new collaborations for visiting scientists. In September 2010, the TYC organised thefirst workshop on Energy Materials that brought together internationally leadingcomputational and experimental scientists, across a broad area of energy materialsresearch. Recent advances in photo-induced energy conversion, hydrogen and energystorage, electrochemistry and fuel cells were presented. A collection ofpapers presented at this workshop was published in a themed issue inPhys. Chem. Chem. Phys. this year (vol. 13 (17), pp 7602-7719).
Our vision is to establish the TYC energy workshop as a fixture in the diariesof computational physicists and chemists that share our interest in energy relatedproblems. To achieve this goal, we have created an international scientificadvisory committee, comprised of about 10 individuals who are all leaders in computational or experimental materials research. We plan to organize a focused workshop for 2012 that will showcase the very best of research in the area of charge transfer in energy materials, with contributions from leading computational and experimental research groups from Europe, the US and elsewhere. The number of participants will be about 100-150 and we will strongly encourage participation from industry and charities (e.g. carbon trust). Sponsorship by CECAM would help us cover some of the costs for administrative work, travel and lodging expenses of invited speakers.
Charge transfer underlies the function of many devices ranging from transistors to light-emitting diodes, and from solar to biofuel cells. It is often the limiting factor for the performance of these devices. As such it is absolutely vital to understand the mechanism, thermodynamics and kinetics of this important physical process on a molecular level. Although the time and length scales of the charge transfer that occurs in these devices may be rather different, the inherent dynamics may be similar and indeed be described by similar theoretical approaches. This workshop aims to cross-fertilize areas where charge transfer plays a crucial role. To this end we plan to have 4 sessions on charge transfer within and between (1) organic semiconducting materials (2) dye molecules and semiconductors (3) inorganic oxide materials and (4) biological molecules and inorganic substrates.
The application of density functional methods to charge transfer reactions is not straightforward. The reason is that common exchange-correlation functionals are of limited use for this task due to their tendency to erroneously delocalize electrons, thus prohibiting an accurate modelling of charge transfer states.
This deficiency termed electron self-interaction or delocalization error is intrinsic to GGA and hybrid density functionals unless special care is taken in their parametrization.
In parallel to the development of functionals with minimal self-interaction error, a number of correction schemes have been proposed to minimize the delocalization error of existing and computationally inexpensive density functionals including for instance DFT+U (see e.g. a recent application ), penalty density functionals, constrained DFT and tight binding self consistent charge DFT methods. Future developments and improvements of such schemes are vital for the realistic simulation of charge transfer in energy materials.
Once charge transfer states are characterized with electronic structure methods, one is left with the question of how to convert this information into charge transfer rates that can be compared with experiment. It turns out that approaches based on Marcus or related theories are remarkably successful in many situations. Here charge transfer is considered as a hopping process with hopping rates determined by only three parameter, all of which can be obtained by suitable DFT methods (see above).
Recently, Marcus rates have been used in kinetic Monte Carlo simulation for calculation of electron mobilities in extended organic semiconducting materials (e.g. fullerenes or conjugated polymers), often in noteably good agreement with experiment[5,6,7].
A similar approach has been used in order to understand reliability issues in microelectronic devices, which are related to ET from silicon substrate into gate-oxide layer. Of particular interest are of course situations where the Marcus picture may not apply, for instance in the regime of ultrafast charge transfer reactions, e.g. charge transport in DNA, or when the fluctuations of the environment are non-Gaussian, e.g. charge transfer coupled to chemical reactions. We hope that all these issues will be discussed, both from the computational and experimental perspective.
Francesca Baletto ( Kings College London ) - Organiser
Jochen Blumberger ( University College London ) - Organiser
Alexander L. Shluger ( University College London ) - Organiser