• Energy Research

AbstractRecent breakthroughs in Cu(In,Ga)Se2 (CIGS) thin film solar cell energy conversion efficiency are related to the application of a potassium fluoride post‐deposition treatment (KF‐PDT) to the completed absorber. Using X‐ray photoelectron spectroscopy and Raman scattering, we compare CIGS layers prior and after the KF‐PDT in the case of a deterioration and an improvement of the solar cells photovoltaic performance. The purpose is to study and model the modification of the surface in both cases and address some of the required characteristics of the absorber, grown on soda lime glass by 3‐stage process, in order to take advantage of the treatment. We show that, in both cases, KF‐PDT induces the formation of GaF3, which is removed during the subsequent chemical bath deposition of CdS, explaining the Ga depleted absorber surface, already reported in literature. However, the presence or not of an ordered defect compound (ODC), correlated with the third stage duration during the CIGS growth, is shown to be crucial in the modifications of the surface induced by the treatment. When an ODC is present prior the treatment, KF‐PDT leads to the formation of a surface layer of In2Se3 containing K, and the photovoltaic performance of completed solar cells are improved. When no ODC is present prior KF‐PDT, no trace of K is found at the absorber surface after the treatment, copper (Cu) segregates into detrimental CuxSe phases, high amount of elemental Se is formed, and the photovoltaic performance are lowered. The role of the ODC during the KF‐PDT is finally discussed.

AbstractRecent breakthroughs in Cu(In,Ga)Se2 (CIGS) thin film solar cell energy conversion efficiency are related to the application of a potassium fluoride post‐deposition treatment (KF‐PDT) to the completed absorber. Using X‐ray photoelectron spectroscopy and Raman scattering, we compare CIGS layers prior and after the KF‐PDT in the case of a deterioration and an improvement of the solar cells photovoltaic performance. The purpose is to study and model the modification of the surface in both cases and address some of the required characteristics of the absorber, grown on soda lime glass by 3‐stage process, in order to take advantage of the treatment. We show that, in both cases, KF‐PDT induces the formation of GaF3, which is removed during the subsequent chemical bath deposition of CdS, explaining the Ga depleted absorber surface, already reported in literature. However, the presence or not of an ordered defect compound (ODC), correlated with the third stage duration during the CIGS growth, is shown to be crucial in the modifications of the surface induced by the treatment. When an ODC is present prior the treatment, KF‐PDT leads to the formation of a surface layer of In2Se3 containing K, and the photovoltaic performance of completed solar cells are improved. When no ODC is present prior KF‐PDT, no trace of K is found at the absorber surface after the treatment, copper (Cu) segregates into detrimental CuxSe phases, high amount of elemental Se is formed, and the photovoltaic performance are lowered. The role of the ODC during the KF‐PDT is finally discussed.

AbstractChemical bath deposited (CBD)Zn(S,O,OH) is among the alternatives to (CBD)CdS buffer layers in Cu(In,Ga)Se2(CIGSe)‐based devices. Nevertheless, the performances reached by devices buffered with (CBD)Zn(S,O,OH) vary strongly from one sample to another and from one laboratory to another, indicating that parameters of minority impact with (CBD)CdS‐buffered devices have major influence when buffered with (CBD)Zn(S,O,OH). Moreover, the literature reports, but not systematically, the requirement of substituting the standard resistive intrinsic ZnO by (Zn,Mg)O and/or soaking the devices in ultraviolet‐containing light in order to reach optimal device operation. The present study investigates the impact of the three following parameters on the optoelectronic behavior of the Cu(In,Ga)Se2/(CBD)Zn(S,O,OH)/i‐ZnO‐based solar cells: (i) CIGSe surface composition; (ii) (CBD)Zn(S,O,OH) layer thickness; and (iii) i‐ZnO layer resistivity. The first conclusion of this study is that all of these parameters are observed to influence the electrical metastabilities of the devices. The second conclusion is that the light soaking time needed to achieve optimal photovoltaic parameters is decreased by (i) using absorbers with Cu content close to stoichiometry, (ii) increasing the buffer layer thickness, and (iii) increasing the resistivity of i‐ZnO. By optimizing these trends, stable and highly efficient Zn(S,O,OH)‐buffered CIGSe solar cells have been fabricated. Copyright © 2014 John Wiley & Sons, Ltd.

AbstractChemical bath deposited (CBD)Zn(S,O,OH) is among the alternatives to (CBD)CdS buffer layers in Cu(In,Ga)Se2(CIGSe)‐based devices. Nevertheless, the performances reached by devices buffered with (CBD)Zn(S,O,OH) vary strongly from one sample to another and from one laboratory to another, indicating that parameters of minority impact with (CBD)CdS‐buffered devices have major influence when buffered with (CBD)Zn(S,O,OH). Moreover, the literature reports, but not systematically, the requirement of substituting the standard resistive intrinsic ZnO by (Zn,Mg)O and/or soaking the devices in ultraviolet‐containing light in order to reach optimal device operation. The present study investigates the impact of the three following parameters on the optoelectronic behavior of the Cu(In,Ga)Se2/(CBD)Zn(S,O,OH)/i‐ZnO‐based solar cells: (i) CIGSe surface composition; (ii) (CBD)Zn(S,O,OH) layer thickness; and (iii) i‐ZnO layer resistivity. The first conclusion of this study is that all of these parameters are observed to influence the electrical metastabilities of the devices. The second conclusion is that the light soaking time needed to achieve optimal photovoltaic parameters is decreased by (i) using absorbers with Cu content close to stoichiometry, (ii) increasing the buffer layer thickness, and (iii) increasing the resistivity of i‐ZnO. By optimizing these trends, stable and highly efficient Zn(S,O,OH)‐buffered CIGSe solar cells have been fabricated. Copyright © 2014 John Wiley & Sons, Ltd.

An alternative to conventional Cu(In,Ga)Se2 module structure is proposed and experimentally investigated. This alternative module structure, which consists in applying metallic buses to connect monolithically adjacent cells in series, is likely to offer the opportunity of minimizing both optical and electrical losses observed in conventional module structure compared to small area cells. The fabrication process of such alternative modules is presented. The performances achieved are discussed in comparison with a standard small-area-cell elaborated simultaneously. Despite slightly lower output voltage per cell, the alternative module structure demonstrates an efficiency of 17.2% (with 81% fill factor), against 16.4% (with 75% fill factor) for the standard cell. This promising result opens new routes to decrease the gap observed between small-area-cells and industrial modules.

An alternative to conventional Cu(In,Ga)Se2 module structure is proposed and experimentally investigated. This alternative module structure, which consists in applying metallic buses to connect monolithically adjacent cells in series, is likely to offer the opportunity of minimizing both optical and electrical losses observed in conventional module structure compared to small area cells. The fabrication process of such alternative modules is presented. The performances achieved are discussed in comparison with a standard small-area-cell elaborated simultaneously. Despite slightly lower output voltage per cell, the alternative module structure demonstrates an efficiency of 17.2% (with 81% fill factor), against 16.4% (with 75% fill factor) for the standard cell. This promising result opens new routes to decrease the gap observed between small-area-cells and industrial modules.

Abstract The structural and optical properties of NaxCu1−xIn5S8 powders, with x = 0 , 0.25, 0.5, 0.75 and 1, are determined. It is shown that all of the samples have the same crystalline structure, which indicates the existence of a solid solution over the whole range of x (i.e. 0 ⩽ x ⩽ 1 ). The increase of the optical band gap of these compounds is found to be linear between x = 0 ( E g = 1.5 0 eV ) and x = 0.7 5 (1.85 eV), whereas for x = 1 (i.e. NaIn5S8) Eg is found to be 2.40 eV, which is much higher than the value expected from the linear slope. Such an evolution shows that the copper and the sodium weigh differently on the optical properties of the material. In thin film solar cells with indium sulfide buffer layer, NaxCu1−xIn5S8 compounds are formed at the Cu(In,Ga)Se2/In2S3 interface. The impact of their properties, here determined, on these thin film solar cells is discussed.