Atom probe tomography (APT) has been used to study the distribution of Na atoms in polycrystalline Cu(In,Ga)Se-2 (CIGSe) thin films. APT, which allows separate investigations of grain boundaries and grain interiors chemistry, shows the presence of inter- and intra-granular Na segregations. It is highlighted that these segregations are found associated to Cu-depletion and In-enrichment. The segregation of Na to crystalline point defects and extended ones is finally discussed regarding its impact on the electrical properties of CIGSe layers. (C) 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

The reason why so-called wide-bandgap CuIn1−xGaxSe2 (CIGSe with x > 0.4) based solar cells show hindered performance compared with theoretical expectations is still a matter of debate. In the present Letter, atom probe tomography studies of CuIn1−xGaxSe2 polycrystalline thin films with x varying from 0 to 1 are reported. These investigations confirm that the grain boundaries (GBs) of low gallium containing (x < 0.4) CIGSe layers are Cu-depleted compared with grains interior (GI). In contrast, it is observed that the GBs of widest band gap CIGSe films (x > 0.8) are Cu-enriched compared with GI. For intermediate gallium contents (0.4 < x < 0.8), both types of GBs are detected. This threshold value of 0.4 surprisingly coincides with solar cells output voltage deviation from theoretical expectations, which suggests modifications of GBs properties could participate in the loss of photovoltaic performance.

We report on the characterization of the chemical and electronic interface structure of the heavily intermixed In 2 S 3 /Cu(In, Ga)Se 2 (CIGSe) interface. Discussing our findings inferred from direct and inverse photoemission as well as soft x-ray emission spectroscopy, we particularly focus on the impact of the interfacial intermixing/interdiffusion processes on the electronic band alignment at the interface. Furthermore, we present deposition-temperature-dependent data on the (In, Al) 2 S 3 /CIGSe interface. We find that the chemical processes at the buffer/absorber interface are thermally activated with a threshold temperature of approximately 200°C.

The goal of this work is to classify and systematically investigate the N1 Deep Level Transient Spectroscopy (DLTS) response characteristic for Cu(In,Ga)Se2-based heterojunctions. Up to four different peaks giving rise to the N1 signal are observed. It is shown that the contribution of the components strongly depends on the metastable state of the sample (different kind of light soaking and reverse bias treatment) and measurement conditions. In the second part of the paper the unusual features of the N1 response, such as an enormously high signal level, are discussed. It is argued, basing on the correlations between capacitance–voltage curves and DLTS spectra, that some of the properties might be due to intrinsic InCu DX centers.

The CIGSe/In2S3 interface is known to be highly diffuse because of the migration of Cu from the CIGSe into the In2S3. Most of the analytical techniques allowing the determination of composition profiles throughout this interface involve ion etching either during the samples preparation or during data acquisitions. In the present work, we have explored the potential of the Raman scattering for the characterization of such interfaces. This technique is non destructive and provides information on both the composition and the structure of the materials that are probed. Three CIGSe/In2S3 structures have been investigated; the parameter varying being the substrate temperature during the In2S3 deposition. For the first time we could demonstrate that at high temperature, the CuInS2 Cu–Au phase is formed at the CIGSe/In2S3 interface. Furthermore, the thickness of the ordered defect compound at the CIGSe surface increases with the deposition temperature. All of the new knowledge collected during this work shows the relevance of using the Raman scattering technique for the characterization of the CIGSe/In2S3 interface.

Atom Probe Tomography (APT) technique is the only nano-scale-sensitive analytic tool allowing 3D chemical analysis with atomic scale resolution. For long restricted to conductive samples, implementation of ultra fast laser pulsing extend now the field of applications to the analysis of semiconductor materials. In the present study, high efficiency Cu(In, Ga)Se 2 (CIGSe) thin films have been investigated by APT in order to solve interrogations about grain boundaries (GBs) composition. The analyzed CIGSe layers have been grown by co-evaporation on Mo-coated soda-lime glass substrates following the standard 3-stage process and the atom probe tips prepared using a focused ion beam (FIB) equipment. In order to ensure the presence of GB in the small APT investigated volume, location and misorientation of GBs have been determined by electron backscattering scanning diffraction (EBSD) and one GB interface placed close to the edge of the tip. From APT analyses, spatial distribution of CIGSe elements can be imaged at atomic scale; particular attention has been devoted to the composition profiles at the vicinity of the CIGSe GB interface. New results are compared with usual CIGSe GB passivation models.

In this work, we investigate the influence of gallium content on the defects properties of co-evaporated CuInxGa1−xSe2 by sub gap modulated photocurrent spectroscopy and admittance spectroscopy techniques. A series of CuInxGa1−xSe2 based solar cells with different gallium content in the range from 0% to 33%, and with the same CdS buffer layer have been investigated. On one hand, photocurrent spectroscopy results show 2 types of defects named D1 and D2, and on the other hand, admittance spectroscopy results exhibit only one type of defect. I–V curves show that one of the two defects probed by photocurrent is responsible of the dominant recombination mechanisms next to the heterointerface, between the absorber and the buffer layers. Moreover, I–V curves under AM 1.5 conditions show that the cell with no probed D2 defect presents the best photovoltaic performances.

The present contribution deals with the rather longstanding issue of the preferential orientation of Cu(In,Ga)Se-2 polycrystalline thin films. We investigate both the influence of the growth process parameters and that of the presence of Na on the competition between [112] and [220] orientations. The influence of the presence of Na is studied through the comparison of CIGSe layers co-evaporated on our laboratory standard Mo-coated soda lime glass (SLG/Mo) and on substrates with a sodium diffusion barrier (SLG/barrier/Mo); the process dependence of the orientation is evaluated through the comparison of films grown by the standard bithermal three-stage (400-630 degrees C) and the derived isothermal three-stage process (620 degrees C). For all the process/substrate combinations, the properties of the films (preferential orientation, grain size and morphology) have been determined at key steps of the growth. From these experimental results, it can be concluded that, as already suggested in the literature, the final layer orientation is strongly related to the texturation of the (In,Ga)(2)Se-3 precursor; however, the amount of Na available when the film becomes Cu-rich (recrystallization at the end of the 2nd-stage) can also strongly impact the film orientation. Such a phenomenon is herein interpreted by mean of the grain boundary migration model of recrystallization. In agreement with this new interpretation of the experimental data, processes have been designed in order to grow [220] textured CIGSe layers. Copyright (C) 2011 John Wiley & Sons, Ltd.

Atomic scale chemistry of polycrystalline Cu(In,Ga)Se2 (CIGSe) thin film has been characterized at key points of the 3-stage process using atom probe tomography. 3D atom distributions have been reconstructed when the layer is Cu-poor ([Cu]/([Ga] + [In]) < 1), Cu-rich ([Cu]/([Ga] + [In]) > 1), and at the end of the process. Particular attention has been devoted to grain boundary composition and Na atomic distribution within the CIGSe layer. Significant variation of composition is highlighted during the growing process, providing fundamental information helping the understanding of high efficiency CIGSe formation.

The present paper aims at stating when and why small grains transform to large grains during Cu(In,Ga)Se2 (CIGSe) film growth following three-step processes. Experimental observations revealed that such recrystallization is achieved when the nominal composition of the films is close to a 1:1:2 stoichiometry. A new model based on grain boundary migration theory is proposed in order to establish a causal relationship between such a composition threshold and grain boundary motion yielding large grain formation. This model is related to some of the experimental observations related to CIGSe layer growth that have previously been difficult to explain.

The chemical structure of the interface between a nominal In2S3 buffer and a Cu(In,Ga)Se2 (CIGSe) thin-film solar cell absorber was investigated by soft x-ray photoelectron and emission spectroscopy. We find a heavily intermixed, complex interface structure, in which Cu diffuses into (and Na through) the buffer layer, while the CIGSe absorber surface/interface region is partially sulfurized. Based on our spectroscopic analysis, a comprehensive picture of the chemical interface structure is proposed.

The amount of copper excess provided during the Cu(In,Ga)Se2 (CIGSe) 3-stage co-evaporation process is among the most operator subjective. In the present paper the influence of this parameter on the properties of the CIGSe films as well as on the behaviour of the related solar cells is investigated. It is observed that both the In/Ga lateral intermixing and the grain size are enhanced when the excess of copper is increased. Contrary to what could be expected, these changes only weakly affect the performance of the solar cells. Increasing the copper excess also yields a rougher CIGSe morphology. This latter evolution is observed to be the most important factor influencing the device behaviour. Through accurate analysis of quantum efficiency, it is concluded that, in the case of the standard cell structure, there exists a threshold in copper excess, beyond which the cell performance is significantly reduced.

In this paper, it is shown that (In1−xAlx)2S3 thin films can be grown through the co-evaporation of elemental indium, aluminum and sulfur. It is nevertheless observed that the introduction of aluminum within the indium sulfide thin films hinders the crystallites size and even yields almost amorphous films when x is 0.2. The investigations of the optical properties of the films reveal that contrary to what could be expected, the band gap increase is low; the highest values measured do not exceed 2.2eV. However, as suggested by X-ray photoelectron spectroscopy measurements, such widening most probably affects the lower conduction band states.

The present contribution deals with the influence of the copper concentration in Cu(In,Ga)Se2 (CIGSe) on the solar cells based on CIGSe/(PVD)In2S3 and CIGSe/(CBD)CdS. We find that, depending on the buffer layer, the optimum open circuit voltage (Voc) is not reached for the same copper concentration. The values of Voc for the CIGSe/(CBD)CdS solar cells are higher when the copper content is very close to stoichiometry (25%), whereas, the Voc values for CIGSe/(PVD)In2S3 solar cells attain their maximum for lower copper contents. On the other hand, contrary to the case of the (CBD)CdS buffer, the Jsc is strongly hindered for the (PVD)In2S3 buffered cells when the copper content is lowered. The study has been made for different absorber gallium contents and the evolution is coherent with the presence of a cliff at the CIGSe/(PVD)In2S3 interface.

Al doped ZnO (ZnO:Al) is commonly used as front contact in Cu(In,Ga)Se2 based thin film solar cells, and for this application is often deposited by RF magnetron sputtering from a ceramic target. In the present study, we use a configuration in which the substrates are immobile under the sputter target and we find the ZnO:Al film properties to depend strongly on the substrate position relative to the target erosion area (or “racetrack”). The zone immediately under the racetrack has higher resistivity correlated to a lesser crystallinity. The differences between these zones (inside the racetrack, under the racetrack, outside the racetrack) are studied as a function of the substrate: bare soda-lime glass (SLG), (CBD)CdS coated SLG, undoped ZnO coated SLG. It is shown that the evolution of the ZnO:Al films non homogeneity is substrate dependent. The impact on complete CIGSe/CdS/ZnO/ZnO:Al devices is measured and show that the influence of the substrates has to be taking in account; the device prediction made from models using the ZnO:Al properties measured on bare glass are inaccurate.

Co-evaporated Cu(In,Ga)Se-2 (CIGSe) based solar cells with Physical Vapour Deposited (PVD) Indium Sulphide (In2S3) as buffer layer have been studied by admittance spectroscopy and current-voltage characteristics measurements. The results have been compared to those obtained with a reference CBD-CdS/CIGSe device. In darkness, the OVD-In2S3 buffer layer devices exhibit higher densities of trapping defects and low values of shunt resistance. However, under illumination we have observed an important improvement of the In2S3/CIGSe electronic transport properties. This behavior seems to be linked to the presence of a metastable defect with activation energy of 0.3 eV. (C) 2008 Elsevier B.V. All rights reserved.

In the present study the optical properties of co-evaporated indium sulfide thin films are investigated. Before being optically characterized, the composition as well as the crystalline properties of the film have been checked with the help of energy dispersive spectroscopy (EDX) and X-Ray diffraction (XRD) analyses. The optical absorption coefficient alpha of this indium sulfide film has been deduced from reflectivity R(lambda) and transmission T(lambda) measurements. The fit of the curve representing alpha(hv) suggests that the beta-In2S3 has an indirect bandgap of 2.01 eV. Density functional theory (DFT) calculations are performed on this indium sulfide compound, using TB-LMTO code. Through these band structure investigations, an indirect bandgap is predicted as observed experimentally. The top of the valence band is mainly formed by the orbitals of the sulfur atoms. This observation suggests that an over or under stoichiometry in sulfur may affect both the nature and the width of the indium sulfide bandgap. (C) 2008 Elsevier B.V. All rights reserved.