This thesis investigates amorphous (a-Si:H) and microcrystalline (μc-Si:H) solar cells deposited by very high frequency plasma enhanced chemical vapor deposition (VHF-PECVD) in the n-i-p or substrate configuration. It focuses on processes that allow the use of non transparent and flexible substrates such as plastic foil with Tg < 180°C like poly-ethylene-naphtalate (PEN). In the first part of the work, we concentrate on the light trapping properties of a variety of device configurations. One original test structure consists of n-i-p solar cells deposited directly on glass covered with low pressure chemical vapour deposition (LP-CVD) ZnO. For this device, silver is deposited below the LP-CVD ZnO or white paint is applied at the back of the substrate as back reflector. This avoids the parasitic plasmonic absorptions in the back reflectors, which are observed for conventional rough metallic back contacts. Furthermore, the size and morphology of the LP-CVD ZnO is varied and the relation between the substrate morphology and the short circuit current density (Jsc) is experimentally explored. As a result, the Jsc can be increased by 23% for a-Si:H and 28% for μc-Si:H solar cells compared to the case of flat substrate and the role of the size and shape can be clearly separated. We also explore the optical behavior of single and multi-junction devices prepared with different back and front contacts. The back contact consists either of a 2D periodic grid with moderate slope, or of LP-CVD ZnO with random pyramids of various sizes. The front contacts are either a 70 nm thick, nominally flat ITO or a rough 2 microns thick LP-CVD ZnO. We observe that, for a-Si:H, the cell performance is critically dependent on the combination of thin flat or thick rough front TCOs and the back contact. Indeed, for a-Si:H, a thick LP-CVD ZnO front contact provides more light trapping on the 2D periodic substrate. The Jsc relatively increases by 7 % with LP-CVD ZnO compared to ITO. Then, we study the influence of the thick and thin TCOs in conjunction with thick absorbers like triple junction or mc-Si:H solar cells. Because of the different nature of the optical systems, thick (> 1 micron) against thin (<0.3 micron) absorber layer, the antireflection effect of ITO becomes more effective and the structure with the flat TCO provides as much light trapping as the rough LP-CVD ZnO. Finally, the conformality of the layers is investigated and guidelines are given to understand the effectiveness of the light trapping in devices deposited on periodic gratings. In the second part, we quantitatively describe the effect of continually varying the substrate morphology for the device in the n-i-p configuration on open-circuit voltage (Voc) and fill factor (FF). Transmission electron microscope (TEM) observations show that V shape morphology creates nano- cracks and reduces the Voc and FF of the solar cells. Hence, we investigate cell designs and processes that avoid Voc and FF losses. For a-Si:H solar cells, we introduce an amorphous silicon carbide n-layer (n-SiC), a buffer layer at the n/i interface, and show that the new cell design yields high Voc and FF on both flat and textured substrates, contrary to the usual microcrystalline silicon n-doped layer. Finally, the beneficial effect of our optical and electrical findings is used to fabricate a-Si:H solar cell with an initial efficiency of 8.8 % and stabilized efficiency of 7% on plastic foil. We find that for our reduced temperature processes windows, the light-induced degradation of a-Si:H solar cells depends strongly on the thickness of the absorber layer. Indeed, the relative efficiency degradation is reduced from 27% to 17% for 400 nm and 200 thick cells, respectively. This degradation can be further lowered to 15 % in a-Si/a-Si tandem structure, and still using a total 300 nm thick absorber layer. For μc-Si:H solar cells, we introduce a buffer layer with a higher amorphous fraction between the n-doped and intrinsic layer. Our study reveals that the buffer layer limits the formation of voids and porous areas (nano-cracks), which promotes oxygen diffusion in the μc-Si:H material. Therefore, this layer mitigates the Voc and FF losses which enhances the performance of the μc- Si:H solar cell. By applying our findings, we make μc-Si:H solar cells with an efficiency of 8.7% on plastic foil for an only 1.2 m absorber layer thickness. The micromorph solar cell (stack of amorphous and microcrystalline cells) concept is the key for achieving high efficiency stabilized thin film silicon solar cells. We present results with and without an intermediate reflector. In particular, we introduce an original device structure that allows a better control of the layer growth and of the light in-coupling into the two sub-cell components. It is based on an asymmetric intermediate reflector (AIR), which increases the effective thickness of the a-Si:H by a factor of more than three. Hence, the a-Si:H thickness reduction diminishes the light-induced degradation, and micromorph tandem cells with 11.2 % initial and 9.8% stabilized efficiencies (1000h, 50°C, 100mW/cm2) are achieved on plastic foil. The stabilized Jsc of the n-i-p tandem solar cells is close to 12 mA/cm2, which offers the possibility for the low Tg flexible substrate technology to compete with state of the art stabilized thin film silicon devices. Based on the results obtained here, a further optimisation of the ITO/p and p-i interfaces, should allow it to be possible to exceed 12% stabilized efficiency on low Tg plastic substrate for micromorph tandems cells.