Templated TiO2 mesoporous films used as highly efficient photoelectrode in Dye-Sensitized Solar Cells

The present thesis is dedicated to the synthesis and characterization of the TiO2 semiconductor layer used as photoelectrode in dye-sensitized solar cells (DSSCs), with the aim to improve their photovoltaic efficiencies.

DSSCs have been reported by O’Regan and Grätzel in the early nineties as a very promising alternative to conventional silicon devices. Main benefits of these cells are their low cost and their mild manufacturing process. In most of the specific literature, DSSCs are made of TiO2 films prepared by doctor-blade or screen-printing of anatase nanoparticles paste. However due to the random organization of the nanoparticles, pore accessibility by the dye and electrolyte could be incomplete and some anatase crystallites could be not connected impeding electron transfer.

The strategy adopted to improve the films properties and thus PV efficiencies involves a surfactant-assisted process allowing the preparation of highly porous layers with well-ordered and accessibles pores as well as improved crystallites connectivity.
The main goal of this work is to increase the film surface area and perfectly control the mesostructure in terms of thickness, pore size, pore organization and pore accessibility in order to maximize the adsorption of active dye and the electrolyte infiltration inside the porous network. Special attention was paid to the tuning of the experimental settings such as the relative humidity conditions, the withdrawal speed and the choices of substrate and surfactant.
Moreover, for DSSCs applications, TiO2 film has to be crystallized in form of anatase. Perfect balance between high crystallinity and mesostructure preservation was studied in order to enhance the cells efficiencies.

Besides, templated films challenge is to obtain thick layers. Indeed, monolayer films are only a few hundred nanometers thick. To increase the film thickness and thus the quantity of active material, a multilayer process was tuned.
Special effort was paid to overcome the surface area limitation induced by the repeated thermal treatments applied during multilayer process. We propose an alternative thermal treatment in order to limit the mesostructure degradation. We also define the maximum crystal size compatible with the preservation of the mesoarchitecture initially induced by templating.
Thick films up to 4 µm were prepared from this multilayer process and show excellent efficiency in combination with N-719 dye (6.1%) when compared to values reported in the literature. Such mesostructured templated films were compared in terms of photovoltaic performances with TiO2 nanoparticles films, generally used in DSSCs.

In a second part, as the goal of this thesis is to improve the current nanoparticles-based DSSCs and prove the viability of the templating alternative, a comparison of the long-term stability of both technologies was performed.
To our knowledge, long-term stability of templated DSSCs has never been reported at this time. However, in case of templated films, the surface area is highly improved and the negative effects of thermal stress, light soaking and UV exposure could be heightened.
Due to their higher active interface, templated films are more sensitive than nanoparticles samples to UV illumination, what can be easily solved by the use of a UV filter. However, they are as stable as nanoparticles samples under visible light soaking (UV filtered) and under thermal stress.
In addition, cells were characterized by electrochemical impedance spectroscopy (EIS). Templated cells show lower transfer resistance, as well as longer electron lifetime compared to nanoparticles DSSCs.
Using templated films in DSSCs is therefore really promising because higher conversion efficiencies are reached without any increase in cells degradation.

Finally, stability limitation encountered by DSSCs are mostly related to the use of liquid electrolytes, which can leak out the cell. Solid-state hole transporting materials are investigated in order to overcome this issue.
However, in solid-state DSSCs, TiO2 films thickness is limited to a few microns allowing the adsorption of a limited amount of photoactive dye and thus leading to a poor light harvesting. Moreover, solid-state DSSCs are characterized by incomplete electrolyte filling, impeding the dye regeneration. Both limitations further lead to low photovoltaic efficiencies.
Due to the surface area improvement as well as the perfect control of the pore organization and the pore size, the templating strategy was investigated to overcome light harvesting and pore filling limitations.
Templated films were prepared from different structuring agents. They show an efficient electrolyte infiltration and a two times higher dye loading compared to nanoparticles layers.
Corresponding photovoltaic performances in liquid-state and solid-state DSSCs have also been evaluated. While templating allows improving the liquid-state cells efficiencies, we cannot conclude for solid-state DSSCs due to device assembly issue and/or bad contacts between the electrodes and the sample holder during the I-V measurements.

We hope that the achievements of this thesis brought a significant contribution to the field of DSSCs. Indeed, the templating strategy is proved to improve the liquid-sate cells efficiency. However, the assembly of solid-state devices and subsequent I-V tests have to be investigated further.
Besides, new pathways are envisaged for interesting future work in both fundamental and applied research fields, such as the synthesis of templated films with hierarchical porosity or scale-up and industrialization of the templated devices.