Abstract :
[en] With a record power conversion efficiency of 23.35% and a low carbon footprint, Cu(In,Ga)Se2 remains as one of the most suitable solar energy materials to assist in the mitigation of the climate crisis we are currently facing. The progress seen in the last decade of Cu(In,Ga)Se2 advancement, has been made possible by the development of postdeposition treatments (PDTs) with heavy alkali metals. PDTs are known to affect both surface and bulk properties of the absorber, resulting in an improvement of the solar cell parameters open-circuit voltage, short-circuit current density and fill factor. Even though the beneficial effects of PDTs are not questioned, the underlying mechanisms responsible for the improvement, mainly the one related to the open-circuit voltage, are still under discussion. Although such improvement has been suggested to arise from a suppression of bulk recombination, the complex interplay between alkali metals and grain boundaries has complicated the labour to discern what exactly in the bulk material is profiting the most from the PDTs. In this regard, the development of this thesis aims at investigating the effects of PDTs on the bulk properties of Cu(In,Ga)Se2 single crystals, i.e., to study the effects of alkali metals in the absence of grain boundaries. Most of the presented analyses are based on photoluminescence, since this technique allows to get access to relevant information for solar cells such as the quasi-Fermi level splitting and the density of tail states directly from the absorber layer, and without the need of complete devices.
This work is a cumulative thesis of three scientific publications obtained from the results of the different studies carried out. Each publication aims at answering important questions related to the intrinsic properties of Cu(In,Ga)Se2 and the effects of PDTs. The first publication presents a thorough investigation on the effects of a single heavy alkali metal species on the optoelectronic properties of Cu(In,Ga)Se2. In the case of polycrystalline absorbers, the effects of potassium PDTs in the absence of sodium have been previously attributed to the passivation of grain boundaries and donor-like defects. The obtained results, however, suggest that potassium incorporated from a PDT can act as a dopant in the absence of grain boundaries and yield an improvement in quasi-Fermi level splitting of up to 30 meV in Cu-poor CuInSe2, where a type inversion from N-to-P is triggered upon potassium incorporation. This observation led to the second paper, where a closer look was taken to how the carrier concentration and electrical conductivity of alkali-free Cu-poor CuInSe2 is affected by the incorporation of gallium in the solid solution Cu(In,Ga)Se2. The results obtained suggest that the N-type character of CuInSe2 can remain as such until the gallium content reaches the critical concentration of 15-19%, where the N-to-P transition occurs. A model based on the trends in formation energies of donor and acceptor-like defects is presented to explain the experimental results. The conclusions drawn in this paper shed light on why CuGaSe2 cannot be doped N-type like CuInSe2.
Since a decreased density of tail states as a result of reduced band bending at grain boundaries had been previously pointed out as the mechanism behind the improvement of the open-circuit voltage after postdeposition treatments, the third publication focusses on how compositional variations and alkali incorporation affect the density of tail states of Cu(In,Ga)Se2 single crystals. The results presented in this paper suggest that increasing the copper and reducing the gallium content leads to the reduction of tail states. Furthermore, it is observed that tail states in single crystals are similarly affected by the addition of alkali metals as in the case of polycrystalline absorbers, which demonstrates that tail states arise from grain interior properties and that the role of grain boundaries is not as relevant as it was thought. Finally, an analysis of the voltage losses in high-efficiency polycrystalline and single crystalline solar cells, suggested that the doping effect caused by the alkalis affects the density of tail states through the reduction of electrostatic potential fluctuations, which are reduced due to a decrease in the degree of compensation. By taking the effect of doping on tail sates into account, the entirety of the VOC losses in Cu(In,Ga)Se2 is described.
The findings presented in this thesis explain the link between tail states and open circuit voltage losses and demonstrate that the effects of alkali metals in Cu(In,Ga)Se2 go beyond grain boundary passivation. The results presented shed light on the understanding of tail states, VOC losses and the intrinsic properties of Cu(In,Ga)Se2, which is a fundamental step in this technology towards the development of more efficient devices.
