One of the key issues in manufacturing of CdTe solar cells is the formation of a stable and low-barrier ohmic contact to the CdTe layer. When typical metals are deposited onto the CdTe surface, a barrier is formed at the interface due to an unfavorable energy-band alignment. The alignment is determined by the work-function of the metal, interface states, and metal-induced gap states at the interface, which leads to pinning of the CdTe Fermi-level close to the middle of its energy gap. The main subject of this PhD work was the fundamental investigation of the parameters involved in formation of reliable ohmic back contacts to the CdTe layer. For this purpose, a multifunction vacuum chamber, which was equipped with a plasma ion source and multiple sputtering targets and PVD sources, was integrated with the DAISY-SOL. Solar cells were prepared by sequential deposition of the CdS and CdTe layers onto TCO-coated glass substrates by close-space-sublimation (CSS). Then, the cells were completed by a CdCl2 treatment followed by deposition of the back-contact layer(s). After any layer deposition or treatment, the sample could be transferred to an Escalab spectrometer for XPS/UPS characterization without breaking the ultra-high vacuum environment. The goals of this project were to eliminate the use of copper (due to stability concerns), and to prepare back contact without etching. In order to meet these goals, the CdCl2 activation had to be performed in a non-oxidizing atmosphere. Therefore, an in-situ CdCl2 activation reactor was developed and integrated to the system to perform the treatment in vacuum. This facility, besides its benefits in fabrication of solar cells in an all-dry process, enabled us to study effects of the CdCl2 activation on CdTe and CdS layers with photoelectron spectroscopy (PES). Moreover, since during the activation no oxide layer was formed, it was possible to study the electrical properties of the CdTe solar cells by conductive AFM (c-AFM) with nanoscale resolution. Effect of typical treatments, such as CdCl2 activation and chemical etching, which usually must be performed before the back-contact deposition, were also investigated by this technique. The electrical properties of the completed solar cells were characterized by current-voltage, external quantum efficiency (EQE), and capacitance-voltage (C-V) measurements. The surface morphology and structure of the layers were investigated by AFM, SEM and XRD. The resistivity and carrier concentration of some thin films were also investigated by the four-point probe and Hall-effect measurements, respectively. Three different back contact approaches were investigated in this work: The first approach was application of an intermediate layer (or a combined layer structure) between the CdTe and metallic back contact. Several interlayers based on ZnTe, Sb2Te3, Sb and Te were applied and the important formation parameters of each contact were optimized. Cell efficiencies as high as the efficiencies obtained with our standard wet process were obtained. This was achieved without any need for chemical etching or washing processes, and of course, without copper. Based on our previous investigations, nitrogen doped ZnTe (ZnTe:N) was a good candidate to be used as the back-contact interlayer; because it could be doped p-type to high levels to realize ohmic contact and a negligible valence-band offset was measured at the CdTe/ZnTe interface. However, solar cells with ZnTe:N/Au back contact had rather low efficiencies, which could be related to the degradation of the cells during deposition of the ZnTe:N layers. As an alternative, antimony was selected to be used as the p-type dopant in ZnTe. After trying several methods, a 4-step procedure was developed and highly p-type antimony doped ZnTe (ZnTe:Sb) thin films with conductivity of 31 S/cm (resistivity of 0.03 Ω.cm) were successfully deposited. Although the contact between ZnTe:Sb and the metallization layer (Mo or Au) was ohmic, the efficiency of solar cells prepared with this interlayer was also low. Another approach was Sb-doping of the CdTe surface to form a tunneling contact. For this purpose, an efficient method was developed, which was similar to the Sb doping of ZnTe layers with some modifications. Solar cells with Sb-doped CdTe surface showed a very low series resistance (3-5 Ω/cm2) and efficiency of the cells was improved by 20%, as compared with our standard cell. With this method, solar cells with efficiency of 12.8% and fill factor as high as 69.3 % with Voc and Jsc of 770 mV and 24 mA/cm2 were produced. This was our record efficiency at the time when this project was performed. The last approach was investigation of reduced molybdenum oxide (MoOx) as the back-contact interlayer. MoOx layers were deposited by RF reactive sputtering and PVD methods. By changing the deposition parameters followed by XPS and UPS characterization, the optimum condition for depositing conductive MoOx layers with work-function larger than 6 eV were determined. CdTe solar cells prepared with MoOx back contact interlayer had better performance compared with the cells without this layer, but their performance was not higher than those prepared with our standard back contact (i.e. NP-etching followed by gold sputtering). Three sets of interface experiments were performed to study the contact between the CdTe surface and MoOx layers deposited by sputtering and PVD methods. The interface investigations were performed by stepwise deposition of MoOx layers onto the CdTe surface followed by XPS and UPS measurement after each step. The resulting energy band diagrams showed that the MoOx interlayers cannot provide a low resistance contact to the CdTe, despite the high work-function and good conductivity of these layers. It was also found that, the work-function difference was compensated by a large interface dipole and the Fermi-level of the CdTe was always pinned at 0.9 eV above the valence band. Therefore, in order to achieve a low resistance back contact with MoOx buffer layer, the CdTe surface must be highly doped to facilitate carriers tunneling, otherwise a large barrier height is formed at the interface. Most of the copper-free back contacts investigated in this work did not show any degradation after at least one-year storage in the lab.