A spray pyrolysis route to the undoped ZnO layer of Cu(In,Ga)(S,Se)2 solar cells

The undoped ZnO in the window layer of Cu(In,Ga)(S,Se)₂ solar cells is deposited here by a non-vacuum spray pyrolysis method yielding solar cells of similar efficiencies to cells prepared with the standard sputtered layer. The spray pyrolysis method allows the undoped ZnO to be deposited in sequence on the same apparatus or in-line directly after deposition of the spray-‘ILGAR’ (ion layer gas reaction) In₂S₃ buffer layer. Thus it is highly suitable for manufacturing, not only is spray pyrolysis economical, avoiding the need for two target sputtering systems, but for monolithic integrated modules the second ‘P2’ scribe, carried out to interconnect cells, can be undertaken before sputtering permitting the highly doped ZnO:Al front contact to directly connect to the rear Mo back contact.     

Effect of copper-deficiency on multi-stage co-evaporated Cu(In,Ga)S2 absorber layers and solar cells

Solar cell absorber films of Cu(In,Ga)S₂ have been fabricated by multi-stage co-evaporation resulting in compositional ratios [Cu]/([In] + [Ga]) = 0.93-0.99 and [Ga]/([In] + [Ga]) = 0.15. Intentional doping is provided by sodium supplied from NaF precursor layers of different thicknesses. Phases, structure and morphology of the resulting films are investigated by X-ray diffraction (XRD) and scanning electron microscopy. The XRD patterns show CuIn₅S₈ thiospinel formation predominantly at the surface in order to accommodate decreasing Cu content. Correlated with the CuIn₅S₈ formation, a Ga-enrichment of the chalcopyrite phase is seen at the surface. Since no CuS layer is present on the as-deposited films, functioning solar cells with CdS buffer and ZnO window layers were fabricated without KCN etch. The open-circuit voltage of solar cells correlates with the copper content and with the amount of sodium supplied. The highest efficiency cell (open-circuit voltage 738 mV, short-circuit current 19.3 mA/cm², fill factor 65%, efficiency 9.3%) is based on the absorber with the least Cu deficiency, [Cu]/([In] + [Ga]) = 0.99. The activation energy of the diode saturation current density of such a cell is extracted from temperature- and illumination-dependent current-voltage measurements. A value of 1.04 eV, less than the band gap, suggests the heterojunction interface as the dominant recombination zone, just as in cells based on Cu-rich grown Cu(In,Ga)S₂.     

A simulation study of the effect of the diverse valence-band offset and the electronic activity at the grain boundaries on the performance of polycrystalline Cu(In,Ga)Se2 solar cells

The paper presents a two-dimensional simulation study of a polycrystalline Cu(In,Ga)Se₂ (CIGS) solar cell with various shapes of grains inside the CIGS absorber layer. The grain boundaries (GBs) with a diverse valence-band offset (VBO) and the density of defect states (N_tA) are considered so as to evaluate their effects on the performance of the CIGS cell. The numerical simulations show that a CIGS cell with column-like grains can achieve a high conversion efficiency (η), while the η of a CIGS cell with diamond-like grains is low if the VBO at the GBs exceeds 0.4 eV. The VBO at which the η of the CIGS cell with diamond-like grains peaks is found at 0.20-0.27 eV. A favorable VBO mainly depends on the shape of the grains, but it also depends on the N_tA. The simulations of the CIGS cells in the substrate and superstrate configurations showed that their performances change if the VBO is varied. This result also implies that the configuration of the CIGS cell is important and the substrate configuration with larger grains in the space-charge region has a considerable advantage if the VBO ranges from 0 eV to 0.2 eV.