Organic Dyes
One major challenge for engineering new sensitizers is to increase their response in the red of near-IR region of the solar spectrum and to show an increased molar extinction coefficient. While an extended light-harvesting is required in general for efficient sensitizers in both liquid and solid state DSC, the demand for highly absorbing dyes becomes dramatic in solid state devices, where a semiconductor film of reduced thickness is employed for maximum efficiency. Organic dye molecules are attractive materials in this respect because of their potential high molar extinction coefficient, adjustable absorption spectral response, and environmental compatibility.
For push-pull organic dyes, on the other hand, the reliable calculation of excitation energies still represents an open issue as a definite and effective computational approach has not yet been defined. As a matter of fact, while TDDFT employing conventional exchange-correlation (x-c) functionals yields quite accurate valence excitation energies both for organic and inorganic molecules, large underestimations are obtained for excited states with a significant long-range CT character and in the case of molecules with spatially-extended π systems. Nevertheless a proper choice of the x-c functional and an accurate evaluation of the possible phenomena taking place in solution, such as protonation/deprotonation of the dye, explicit solvent-dye interactions etc., allow us to predict the absorption spectra of the organic sensitizers within 0.1-0.2 eV of accuracy.
We have studied with theoretical calulations some selected organic dyes. Namely the JK1/JK2 dyes, the squaraine dyes and the so-called D5/D7/D9/D11 series of dyes. JK1/JK2 and the D5/D11 series are prototypical dyes designed according to a D-π-A architecture, characterized by a donor moiety (a tertiary amine), a spacer (thiophene) and san acceptor moiety (cyanoacrylic acid). In the squaraine case, an asymmetry is introduced in the molecule to provide vectorial charge transfer towards the carboxylic function.
The description of the spectroscopic properties of full organic dyes turned out to be rather problematic at TDDFT level. Moreover, the considerably large size of the most efficient organic dyes (~100 atoms) and the need of taking into account the interactions with the environment, rules out the possibility of using high-level correlated wavefunction-based methods, such as Multireference Perturbation Theories (MRPTs) or high-order linear response (LR) or equation-of-motion (EOM) Coupled Cluster, which are in principle unaffected by the CT problem.
The strategy we followed to overcome this problem was to assess the reliability of various TDDFT approaches against high-correlated wavefunction-based methods, performing benchmark calculations on small prototype dyes, representative of the most efficient organic sensitizers employed in DSSCs. Therefore, upon calibration in gas phase of the TDDFT methodology, one can safely move to the description of larger systems in solution, achieving a meaningful interpretation of the experimental data with accuracy comparable to that of high correlated ab initio approaches.
References:
Pastore, M.; Fantacci, S.; De Angelis, F. J. Phys. Chem. C 2010, 114, 22742-22750.
Pastore, M.; Mosconi, E.; De Angelis, F.; Graetzel, M. J. Phys. Chem. C 2010, 114, 7205-7212.
Abbotto, A.; Leandri, V.; Manfredi, N.; De Angelis, F.; Pastore, M.; Yum, J-H.; Nazeeruddin, M. K.; Graetzel, M. Eur. J. Org. Chem. 2011, 6195–6205.