Molecular solar thermal storage materials
Energy production from so-called renewables often relies on fluctuating sources such as sun and wind, the availability of which are typically poorly matched with actual demand. As a result short and intermediate energy storage solutions are needed. In this regard, molecular solar thermal (MOST) storage provides attractive options for large-scale energy storage. For the method to be robust, one requires a reversible chemical reaction, which i) can be photo-induced by sunlight and ii) is reversible via a catalytic reaction. The norbornadiene/quadricyclane (NQ) system satisfies these criteria and is a prime candidate for this technology. Optical, catalytic, and thermodynamic properties of this system can be manipulated over a wide range by addition of appropriate functional groups. We are currently investigating these aspects both experimentally and computationally. Supported by our previous allocation we have recently published three computational papers on this topic [see e.g., our paper highlighted on the cover of ChemSusChem, http://tinyurl.com/hzpo4ns]. 1) Computational design During the next years, we plan to extend our efforts in the area of computational design of molecular compounds with properties optimized for MOST applications. Our database currently contains only 64 different compounds (corresponding to about 600 conformers). In order to provide a basis for a more exhaustive exploration, we plan to considerable expand this database. We will then employ machine learning techniques to sample the database in order identify candidate compounds for further computational and experimental exploration. In this context, we also intend to explore alternative design strategies based on integration of multiple N/Q units via polymerization and ring forming reactions. For this part of the project we will primarily rely on density functional theory (DFT) as implemented in the NWChem package, which is already available on beskow. NWChem itself parallelizes well. In addition, we will have to employ optimization algorithms that require multiple images to be run in parallel (such as nudged elastic band). 2) Efficient back-conversion A key aspect with to regard to the implementation of MOST systems in practical applications concerns the efficient back-conversion via appropriate catalytic reactions. In this context, we plan to investigate the energy landscape of the NQ reaction on transition metal and transition metal oxide surfaces as well as nano-clusters and study the associated optical properties. To this end, we will employ both density functional theory (DFT) as well as time dependent DFT (TD-DFT) calculations. For many of these calculations periodic boundary conditions and plane-wave basis sets are more appropriate than the localized basis sets to be used in part 1 of this project. Here, we will therefore rather rely on codes such as VASP and GPAW, both of which exhibit very good scalability and parallel performance.