Next-generation kesterite thin-film solar cells: development, characterization and modeling

Abstract

In order to cope with the ever-growing energy demand while cutting down greenhouse gas emissions, solar power must be deployed at the largest extent possible. This requires the widespread installation of photovoltaic technologies beyond utility-scale to comply with the greater integrability and adaptability needed in emerging markets. Thin-film solar cells based on chalcogenide compounds hold highly encouraging promises for such applications, due to their versatile architecture and tuneable properties. Among them, kesterite materials with the Cu2ZnSn(S,Se)4 crystalline structure provide an interesting alternative based on Earth-abundant elements and compatible with low-cost processes, making them a promising solution with reduced carbon footprint and cost. Yet, their performance must be increased to align their competitiveness with commercial technologies. Reaching this objective entails to resolve their still too significant deficit in open-circuit voltage, which is the main culprit for their limited efficiency around 15% for the highest-performing devices. The physical origin of these losses is manifold: control of the kesterite phase and composition in a uniform fashion, regulation of its crystalline quality to favour low disorder, monitoring of the growth environment and conditions enabling weakly defective materials and interfaces, as well as preserving all these aspects when tuning the absorber bandgap for relevant applications, among others. This thesis aims at tackling these challenges on different levels. First of all, it is demonstrated through an extensive review of the literature that germanium (Ge) alloying constitutes a promising strategy to enhance the quality of kesterite thin films by overcoming intrinsic electronic limitations related to tin (Sn) while promoting superior structural morphology. On top of this, this approach also allows to control the absorber bandgap, which contributes to a greater versatility of the kesterite compounds. Still, most studies have focused on low to moderate proportions of Ge corresponding to a narrow bandgap range not yet matching the specifications of emerging applications. This highlights the need to pursue the exploration of the broad Ge compositional domain. The importance of the kesterite deposition process is also detailed, showing the greater potential of solution-based routes allowing finer control of composition and phases at the lab scale, thus providing conditions for high efficiency. In comparison, samples from the presented starting baseline, deposited via sequential physical processes, exhibit limited performance with symptoms of low opto-electronic kesterite quality and device non-ideality. Following this, a molecular ink chemical route in ambient environment is developed to allow flexible Ge alloying in high-quality single-phase thinfilm kesterite absorbers. Remarkably, this updated baseline enables to tune the kesterite bandgap without compromising material quality. In particular, narrow band tails associated to mitigated open-circuit voltage radiative losses are ensured for a broad bandgap range in which the device performance potential is therefrom augmented. Resolving this prerequisite for photovoltaic absorbers is however only one of the main steps in the development of a solar cell technology. Indeed, non-radiative losses leading to low minority carrier lifetime remain an even more critical challenge to be tackled. It originates from the numerous intrinsic point defects in the kesterite lattice, the exact nature, location and dominance of which remain partly undetermined. It is therefore essential to pinpoint the underlying mechanism responsible for defect-assisted recombination and the associated performance losses. This is especially relevant for the next-generation devices based on multinary-alloyed solution-processed kesterite absorbers which are presently leading the way towards future efficiency breakthroughs. Eventually, based on the study of temperature- and light intensitydependent current-voltage measurements confronting analytical models and SCAPS-1D simulations, the closer-to-ideal behaviour of state-of-theart kesterite devices is demonstrated as a consequence of weaker band tailing and bandgap and potential fluctuations. The close agreement with the single diode formalism combined with the observed dark and light reconciliation allow to gauge the various contributions to the open-circuit voltage deficit. The dominant loss mechanism is hypothesized to be a defectrich kesterite layer at the interface with the buffer, highlighting the diode ideality factor and saturation current density as the primary causes of restrained performance. Light is also shed on carrier trapping-detrapping via shallow defect states as the mechanism behind shunt leakage currents, also contributing to lower efficiency. The applicability of this whole analysis extends across various samples, which emphasizes its potential to support further enhancements of next-generation kesterite solar cells.