Fabrication and characterization of photogalvanic quantum wells

Modern photovoltaic devices based on semiconductors rely on the aid of a p-n junction to convert light into an output current. This, while enabling the production of electrical energy from the light, prevents the voltage to be greater than the material bandgap, limiting the maximum efficiency to be less than 33.16% (Shockley-Queisser limit). Polar crystals recently raised as an intriguing possible alternative since those materials bring a strong inversion symmetry breaking, allowing for the spontaneous separation of photo-excited carriers without the need of a space charge field localized at the junction.This could provide a way to overcome the Shockley-Queisser limit. These materials however up to now exhibit poor experimental performances whose cause is mainly attributed to the strong coupling of the carriers with the lattice that turns them into low-mobility polarons. It is argued that a reduction of the crystal size to the nanoscale may forbid self-trapping of charge carriers due to a synergy between quantum effects and short interaction lengths, thus boosting the conversion efficiency to exploitable levels. In this thesis we aim to start an experiment to study the photo-galvanic properties of a prototypical polar material, lithium niobate, LiNbO_3, when the size of the system is decreased at the nanoscale. We attacked the problem both at theoretical and experimental level to enable the fabrication of lithium niobate nano-crystalline films sandwiched between a suitable electrode geometry. In order to do so the project is articulated in three different parts: computational study of the photoemitted charge thermalization length, production of the samples and characterization. The computational study is aimed toward a seizure of the photoemitted ``hot'' carrier energy spectrum and thermalization distance. The electronic band structure has been calculated via computational DFT methods leading also to an estimate of the phonon frequencies and other physical properties of lithium niobate. By the mean of a mean field approximation the relaxation time for electron-phonon scattering in the crystal has been computed at 3e-5 seconds and it has been used in a Monte-Carlo simulation to predict the behaviour of the photogalvanic current. A first estimate non present in literature of the thermalization length in lithium niobate was calculated and a good accordance with experimental data was found. This length averages around 1 nanometer and provide an estimation of the critical size of the film below which the generation efficiency should display the expected enhancement. The sample production has been performed by means of rf sputtering deposition. To ensure the growth of proper film with excellent crystallinity we used a substrate that could act both as an electrode for the solar cell and as an epitaxial frame for the growth of LiNbO_3. Our research lead us to GaN: this material has already been proved as a viable epitaxial-ready substrate for LiNbO_3 while also being a semiconductor. Besides our research, investigating the fabrication and the electrical properties of such a system is therefore of great importance in view of the emerging topic of semiconductor/ferroelectric hybrid structures. Finally, the produced samples have been characterized by means of HRXRD, AFM, and optical absorption spectroscopy to obtain a full picture of the sample structure and composition as a function of the deposition parameters. A characterization routine has been established for future works although further studies are needed in particular to explain the measured mismatch of our films.