The focus in the literature has been on defect modulation through chemistry to enhance the PCE of these technologies at the first processing step. In the case of CIGSSe, alkali doping and alloying with silver in the fabrication of precursor films have shown the highest PCE. Initially, dopants were incorporated through thermal evaporation or wet chemistry, and recently, the trend has shifted towards direct addition into the ink before casting thin films, a method that has boosted performance. Moreover, this chalcopyrite is a non-stoichiometric compound that allows for different ratios of copper while maintaining the same amount of indium and gallium in the lattice. The ratio of copper has been used to control the grain growth, grain boundary chemistry, shallow donor-acceptor pair defects (pair of low energy defects), and secondary phases in CIGSSe films. All these strategies have been implemented in the first step to produce crystalline or amorphous precursor films containing Cu-In-Ga-S, Cu-In-Ga-S-Se, or Cu-In-Ga-Se, achieved through different solvents and precursor chemistries. However, few efforts have been made to control the microstructure.
A different story happens in the second step concerning selenization. In a widely accepted method since 2009, a graphite box filled with a precursor film and elemental selenium is heated in a tubular or rapid thermal processing furnace to temperatures up to ~600 °C. The precursor film is typically deposited using sulfur-based chemistry and is either crystalline or amorphous. Under selenization, the film undergoes recrystallization via a liquid selenium flux that condensates during the transient heating step. The liquid selenium readily dissolves the available copper, forming a Cu-Se secondary phase at the beginning of the selenization process. TEM images of the cross-section of the selenized thin films, coupled with STEM-EDX and diffraction patterns, have revealed that those secondary phases are still present in the bulk. Furthermore, Raman and SEM-EDX analysis detect Cu-Se at the surface, even under copper-poor compositions. For kesterite CZTSSe, in situ energy-dispersive XRD has shown the formation of these secondary phases, while TEM images have revealed a residual secondary ZnSe phase in the bulk. Moreover, depending on the film thickness and the solvent chemistry, a fine-grain layer is formed as an impurity sink primarily for excess selenium, copper selenide species, and carbon. This layer is sandwiched between the coarse grains and the substrate. All these impurities act as defects that diminish optoelectronic properties and hinder charge transport. Nevertheless, when the samples are not selenized under sufficient liquid selenium medium throughout the whole annealing process, thus preventing adequate defect passivation and grain growth, the performance of the devices is low. Therefore, the process of enhancing performance also inherently limits the threshold these technologies could reach.
In this work, the synthesis of highly-packed CIGSSe quantum dots films using amine-thiol chemistry and direct conversion toward the selenide-rich phase during selenization have been employed to minimize microscopic and macroscopic inhomogeneities. The methods involve solvent and ligand engineering to enhance optoelectronic properties and minimize the carbon impurities of the quantum dot thin films. In addition, a strategy that combines direct synthesis with colloidal synthesis of quantum dots enables the insertion of desired phases like Cu-Se at the interface. With a mixture of alkali dopants, these routes produce high-efficiency devices up to 15.9% PCE compared to those reported in the literature. Nevertheless, employing the accepted method of using a simple graphite box containing the precursor film and the selenium source introduces variability in the experiments due to inhomogeneous selenium condensation and macroscopic recrystallization, affecting their reproducibility. A modified graphite box design allows for independent control of the sample and the selenium temperature during the initial heating stage, thereby favoring anion exchange in the lattice converting the sulfide to selenide via selenium vapors and bypassing the formation of secondary phases. This new design is key in the enhancement of the macrostructure and the composition at the nanoscale, ultimately reducing the defects. Addressing these challenges in the two-step solution process of chalcogenide materials will enable the fabrication of high-performance thin-film photovoltaic devices and contribute to the decarbonization of the energy sector.