Thermophotovoltaic devices convert heat to electricity by absorbing infra-red photons with low energies. In order to do so semiconductors materials with low band-gap are needed. The crystalline Germanium (cGe) is an interesting low-cost alternative to III-V semiconductors and this thesis is focused on selective contacts on this material. In order to have a hole selective contact with a surface recombination velocity (Seff < 500 cm=s), a contact resistance < 0:1 Ohms·cm2 and a contact selectivity (S > 4:2) different TMO materials were studied, Molybdenum oxide (MoOx), Vanadium oxide (V Ox) and Tungsten oxide (WOx). Best results that fulfilled the mentioned requirements can be achieved with a TMO of MoOx under temperature treatment. Also, the effect of adding a 3 nm layer of amorphous silicon (a-Si:H) between the substrate and the TMO is studied, which results in and improvement in the Seff and an increase in contact resistivity. This leads to a trade-off, which is not worth for the a-Si:H thickness used in this study. During the process, also some simulations were run to find the optimum Indium Tin Oxide (ITO) thickness and reflectance for a cGe thermophotovoltaic cell depending on the TMO. It Turns out that the importance of the ITO thickness and the TMO is not that relevant in terms of photocurrent, and what really matters is having a ITO + TMO thicknes of around 100 nm with a refractive index (n = 2). The expected external quantum effciency (EQE), internal quantum effciency (IQE) as a function of the Seff is also simulated, clearly showing the importance of having a good Seff at the front surface of the device. Finally, the experimental results obtained for the different hole selective contacts under study were introduced into the simulations to calculate their potential effciency in the final device. An effciency of 5% is reached for the TMO of MoOx, which is still far away from the maximum theoretical reachable value of 12%, but is closer to the highest reported one which is 7.9%.