The design of new solar cells is a very active research field, with several device architectures having been proposed and extensively studied, like luminescent solar concentrators (LSCs) or dye-sensitized solar cells (DSSCs). Regardless of technical considerations, a critical aspect to reach high performance lies in the choice of the chemical composition. For instance, LSCs are made of one or more polymeric matrices containing chromophores or luminescent nanoparticles, which absorb the sunlight and reemit photons with wavelengths preferably toward the infrared/near-infrared region to be collected by photovoltaic cells on the edges of the matrices. Hence, to reach good efficiency, the chromophores must have high optical stability, a strong quantum yield and exhibit large Stokes shifts to limit self-absorption phenomena. Dyes used in DSSCs must be able to absorb sunlight as well, but also be compatible with the adsorbent structure for the electron/charge transfer processes to be effective. Because of the vast choice of existing or new systems at disposal, a purely experimental investigation to find the most adapted materials can be extremely expensive and time consuming. Computational chemistry can assist the selection process by predicting the performance of potential candidates in order to choose the most promising individuals (virtual screening) and identify their strengths and weaknesses to further improve their efficiency (rational design).
In order to be able to provide a reliable support to experimental investigations, one must carefully balance the computational cost and the expected quality of the results. Because of the size of the systems of interest for those technologies—the chromophores are typically composed of several dozens of atoms—density functional theory (DFT) and its time-dependent extension (TD-DFT) are generally used for the electronic structure calculations and the absorption (and emission) spectra are simulated from simple electronic transitions. However, DFT refers to a family of exchange-correlation functionals of varying quality, cost and applicability. Furthermore, simulating the optical properties of chromophores as a single or multiple purely electronic transitions can be too simplistic and misleading, due to the absence of important features present in the spectra, with the seemingly good agreement with experiment often arising from errors compensating neglected effects. Consequently, a top-down approach, starting from the most accurate feasible models, potentially applied first on smaller model systems is better since it provides a way to assess the impact of necessary approximations, hence the reliability of the overall simulation.
A key objective of this research activity is improving the efficiency of currently available manufactured solar cells, either by supporting the interpretation of the processes occurring in natural photosystems, or by looking for new, superior alternatives to replace existing solutions. The fabrication of more effective devices would reduce the dependency on non-renewable sources of energy, in particular fossil fuels, and have positive impacts on the environment and for a sustainable economy. Some of our active projects on this subject are:
For the design of molecules to use in DSSCs and LSCs, the use of internal coordinates-based approaches can be particularly appealing. Indeed, by reducing the coupling between different coordinates, it is possible to isolate large amplitude motions, like structural deformations upon electronic transitions, from the rest of the system. This is a very promising way to treat molecular systems showing some limited degree of flexibility, like some organic chromophores used in LSCs. Indeed, geometry changes upon electronic excitation seems to favor larger Stokes shift, hence limiting the risk of self-absorption, making such molecules interesting candidates for this technology.
A large part of this activity is supported by our virtual spectrometer to simulate vibrationally resolved electronic spectra, with possible account of chirality and resonances effects. As a matter of fact, chiral molecules, such as metal complexes, can have interesting properties as dyes, and standard, non-chiral and chiral spectroscopies are usually combined to fully characterize them. Another interesting spectroscopy in this respect is Resonance Raman (RR), which is based on the selective enhancement of Raman bands connected to the part of the system where the electronic transition is located. RR is widely used in the characterization of dyes due to its ability to isolate a specific molecule in a complex environment or a region within a large system.