Electron transfer reactions taking place at the surface of dye-sensitized semiconductors are key processes in dye-sensitized solar cells (DSSCs). After light absorption, the excited state of a dye injects an electron into a wide-bandgap semiconductor, usually titanium dioxyde, TiO2. The formed oxidized dye can then be intercepted by a redox mediator, typically iodide, before charge recombination between the injected electron and the oxidized dye occurs. This mediator transports the positive charge to a counter-electrode. A quenching of the dye excited state by the redox mediator might prevent efficient electron injection. To develop an optimal DSSC device, it is important to keep the absorbed photon-to-current efficiency as quantitative as possible. Therefore, every reaction has to be ideally placed along the temporal evolution after primary step of photon absorption by the dye. A good kinetic scheme of reactions, including undesirable charge annihilation loss pathways needs to be established. The different parameters affecting all of these series of reaction have to be well understood. The aim of this dissertation is to investigate these key reactions. Chapter 1 reviews the theoretical background necessary for understanding the following experimental results. Chapter 2 summarizes the experimental tools that are used throughout this work as well as the main advances that have been implemented to the existing pump-probe spectrometer. In chapter 3, a potential annihilation reaction is investigated, namely the reductive quenching of the dye. This reaction leads to the formation of a reduced dye that is not found to inject an electron into TiO2. This reaction occurs under two specific conditions. It requires an overloading of dye on the semiconductor surface, leading to the formation of aggregates that are not properly electronically coupled to the TiO2. It also occurs only when a high concentration of iodide redox mediator is used, typically larger than 1 M. This reaction is also observed on a dye|Al2O3 system. Alumina is an insulator and precludes electron injection, allowing the observation of the reduced dye formation, which takes place within the first picoseconds after light absorption. These findings are relevant for the development of new electrolytes that will have to suppress this reaction pathway. Chapter 4 contains a study on the dye ground state regeneration dynamics. When the dye is adsorbed as a monolayer onto the semiconductor surface, the oxidized dye interception is efficient over a broad range of iodide concentrations. When using ruthenium bipyridyl complexes as sensitizer, two different situations are encountered. The first one reveals an associative mechanism, with a rate of regeneration that reaches a maximum after saturation of the active sites of the dye. The second mechanism presents a repulsive behavior. When the dye structure renders more difficult the approach of iodide, the response of the regeneration dynamics to the bulk iodide concentration implies a sub-linear response. The use of four different organic donor-bridge-acceptor (D-π-A) dyes revealed fast interception rates, dependant on the π-bridge structure. This study allows for a better understanding of the dye cation interception reaction and might help in the design of new organic dyes, revealing the importance of the access of iodide to the sensitizer. Chapter 5 presents a wide study of an organic dye, having a D-π-A structure. All the timescales are investigated, from the early femtoseconds to the millisecond for a complete map of the reactions that take place at the TiO2 surface. Electron injection is found to be surprisingly slow with a τ = 1.7 ps. The dye cation thus formed is intercepted by an iodide-based electrolyte with 99 % efficiency, confirming the potential of these dyes in the development of DSSCs. The last chapter opens the discussion on future projects for the continuation of this research.