Native point defects in undoped samples of CuInSe2 (CIS) have been identified by a multiple-edge refinement of the extended x-ray absorption fine structure of the copper, indium and selenium absorption edges. Ab initio theoretical models for the pure compound and for various defect structures were constructed and carrier-type statistics were predicted by simultaneously fitting multiple absorption sites to these models. As expected, a model of our measurements based on pure compounds with no defects does not yield a good fit to the data. We find that a best fit requires a significant population of defects. Preliminary quantitative analysis suggests a 15 % vacancy in Cu, a 2-12% population of Cu-Se anti-sites and 15% In-Se anti-sites.

The opto-electronic properties of CuInSe2 and related compounds depend on their defect chemistry in a way that is far from being understood and in which ab initio calculations could help by providing new insights as shown previously. Ab initio calculations of energy and electronic structure of various intrinsic (including defect pairs) and extrinsic (including potential dopants such as Zn) point defects have been performed in the chalcopyrite semiconductors CuInSe2, some of them being computed for the first time by advanced ab initio techniques. The method used is based on the density functional theory within the framework of pseudo-potentials and plane waves basis set. The results are discussed in view of the existing data, models and calculations.

The presence of deep defect levels in thin film solar cells can highly affect the characteristics of the photovoltaic energy conversion. Therefore, knowledge of the origin and nature of these defects is desirable.

Deep level transient spectroscopy (DLTS) was performed on a series of CdS/CdTe thin film solar cells which were activated in vacuum or air ambient. Temperature scans between 5 and 320K revealed semi-shallow and mid-gap majority traps. These mid-gap traps were also investigated using isothermal DLTS (region 250 to 330K) where the temperature is kept constant and the rate window is varied. This way the mid-gap traps can be characterized completely. Using electrical injection DLTS and optical DLTS minority traps could be detected. Electrical injection DLTS showed a single defect in the bulk of the air activated cells, while optical DLTS revealed the presence of defects close to the CdS/CdTe interface in both types of samples. The nature and origin of these defects are unknown.

CuInSe2 and Cu(In, Ga)Se2 precursor layers have been prepared by electrodeposition, with morphologies suitable for device completion. These precursor films were transformed into photovoltaic quality films after thermal annealing without any post-additional vacuum deposition process. Depending on the preparation parameters annealed films with different band gaps between 1eV and 1.5 eV have been prepared. The dependence of resulting solar cell parameters has been investigated. The best efficiency achieved is about 10,2 % for a band gap of 1.45 eV. This device presents an open circuit voltage value of 740 mV, in agreement with the higher band gap value. Device characterisations (current-voltage, capacitance-voltage and spectral response analysis) have been performed. Admittance spectroscopy at room temperature indicates the presence of two acceptor traps at 0.3 and 0.43 eV from the valance band with density of the order of 2. 1017 cm-3 eV-1.

Large area Cu(In, Ga)(Se, S)2 thin films (CIGSSe) for solar modules are processed by rapid thermal annealing of stacked elemental layers. A pilot line for 60×90 cm2 absorbers and 30×30cm2 modules is now up and running. Modules with efficiencies at 12% are fabricated with excellent structural and electrical uniformity. For the optimization of electrical performance the selenization and sulfurization process is analyzed by x-ray fluorescence (XRFA), x-ray photo-electron spectroscopy (XPS), secondary ion mass spectroscopy (SIMS) and current voltage (I-V) measurements. Sequentially processed CIGSSe films show characteristic depth distribution profiles of gallium and sulfur. Based upon our previously published model obtained by In-Situ XRD analysis the Ga profile can be explained by the strongly inhibited formation of binary gallium-selenides. The sulfur incorporation is investigated by comparing XRFA data and SIMS profiles of samples from different sulfurization recipes. A characteristic dependence of S on the Cu/(In+Ga) ratio is observed. Two different mechanisms for sulfur incorporation are proposed. It will be shown that the accumulation of sulfur towards the back is predominantly due to the incorporation into an intermediate molybdenum sulfo-selenide layer. Device characterization shows that the Ga and S profiles lead to a favorable double band gap grading structure.

Modeling of two-junction tandem devices shows that for optimal device performance, the bandgap of the top cell should be around 1.6-1.8 eV. CdZnTe alloys can be tailored to yield bandgaps in the desired range. In this study, three approaches were used to fabricate these films. The CdTe and ZnTe films were deposited by close-spaced sublimation (CSS) and radio-frequency sputtering (RFS) techniques respectively. In the first approach, we used mixed powders of CdTe and ZnTe as the source for film deposition by CSS. Even for the ZnTe/CdTe (95:5 ratio) source material, the deposited films were entirely CdTe due to higher vapor pressure of CdTe. In the second approach, we used pre-alloyed CdZnTe powders (CERAC, Inc.) as the source. Due to the lower sticking coefficient of Zn, even for the source composition of 75% Zn, these films contained very low quantities of Zn (∼5%). We tried unsuccessfully to increase the Zn content in the films by confining Zn vapor by enclosing the region between the source and substrate, reducing the substrate temperature to 400°C, and adjusting the source/substrate distance. Finally, we used thin-film couples consisting of 300-nm-thick CdTe deposited by CSS and 300-nm-thick ZnTe deposited by RFS; the samples were then heat-treated in cadmium chloride vapor. Compositional analysis of the samples showed extensive interdiffusion of Cd and Zn for the annealed samples. We will present the data on the various stack configurations of CdTe and ZnTe, the effect of different post-deposition anneals, the effect of oxygen on the interdiffusion and alloy formation and its possible correlation to the device performance degradation.

The low-resistive p- and n-type transparent conductors are necessary for the efficient photovoltaic (PV) solar cells. We review a new valence control method of co-doping with doping Ga- (or In-, or Alsubstitutional impurity, and Li-interstitial impurity) donor and N-acceptor at the same time for the fabrication of a low-resistive p-type ZnO based upon the ab initio calculation. We compare our materials design to fabricate a low resistive p -type ZnO with the recent successful co-doping experiments. The Delafossite structure of CuAlO2 has great potentiality far p-type transparent conducting oxide to apply the high efficient photovoltaic solar-cells combined with n-type transparent conducting oxides such as ITO (indium tin oxides), SnO2 or ZnO. We have calculated the electronic structure, impurity levels, and formation energy of Cu-vacancy, Al-vacancy, Be, Mg, Ca acceptors at the Al-site, and Be, Mg, Ca donors at the Cu-site. Calculated acceptor energy levels are as follows; Cu-vacancy (-300meV from the VBM [valence band maximum]), Al-vacancy (671 meV from the VBM), Be-acceptor (85 meV from the VBM), Mg-acceptor (200 meV from the VBM), and Ca-acceptor (960 meV from the VBM). We find that the Mg impurity at the Cu-site is more stable than the Al-site. Therefore, Mg impurity acts as donor at the Cu-site in the CuAlO2. This is the reason why Mg doping reduces the p-type conductivity in CuAlO2. We propose the following valence control method for the fabrication of p-type CuAlO2; we should dope the high concentration of Cu-vacancy in order to form the impurity band with reducing the Cu vapor pressure during the PLD or MBE crystal growth method, or we dope the Mg or Be acceptors at the Al-site with reducing the Al vapor pressure and increasing the Cu vapor pressure during the thermal non-equilibrium crystal growth method such as MBE or MOCVD. We compare our materials design with the available experimental data. Finally, we propose the co-doping by 2PS(Se)+InCu+VCu for the fabrication of p-type CuInS2 or CuInSe2 based upon ab initio electronic structure calculation.

Epitaxial CuGaSe2 films were grown on GaAs substrates under Cu-excess conditions to obtain stoichiometric compositions. The films were annealed in Ar, Sex or O2 ambients with or without a Cu or Cu-Se cap layer with the intention of changing the intrinsic defect concentrations. Samples were evaluated using low-temperature photoluminescence (PL) measurements. Annealing of the samples dramatically changed the PL spectra indicating that not only interdiffusion had occurred, but defect species and populations were changed. Comprehensive consideration of the changes led to the conclusion that the emissions at 1.62 eV, 1.66 eV and in the range from 1.2 to 1.4 eV are related to specific defects of Se vacancies, Cu vacancy-Se vacancy complexes and interstitial Cu, respectively.

The formation of CuInSe2-CuInS2 alloy films from chalcogenisation of different precursors was investigated by in-situ energy dispersive X-ray diffraction (EDXRD). A sequential synthesis procedure was used. Copper and indium (Cu/In = 1.8) were sputtered on molybdenum coated soda-lime glass and selenium was introduced as a layer of elemental selenium or as a In2Se3 layer. Such prepared precursor films were then sulfurized in elemental sulfur vapor. The effects of the selenium precursors and the influence of sulfurization conditions on the resulting absorber films composition and properties were investigated. It is shown that from different precursors, chalcopyrite phases are formed via significantly different formation pathways which are investigated in detail.

CISCuT - CIS on Cu-Tape - has been established as a new thin film technology, in which Cu/In/S based solar cells are continuously fabricated on a Cu-tape in a series of consecutive roll-to-roll processes. Flexible modules encapsulated into polymer foils are assembled by interconnecting stripes of this tape cell in a roof-tile manner in an automated assembly line. J-V characteristics that show efficiencies of 9 % so far and that have been certified by the Frauenhofer Institut Solare Energiesysteme are presented. Stability tests over more than 8000 hours show promising results. Best results obtained from cells and from modules that have automatically been assembled to different areas are compared. This comparison gives insight into how the different roll-to-roll processes including the assembling of the module influence the overall quality of the module. Latest results including the unique mechanical properties of CISCuT-based solar cells and modules are highlighted. An outlook on the future prospects of this exciting technology is given.