Chalcopyrite-based solar cells, specifically Cu(In,Ga)Se₂ (CIGS) and its silver-containing variant (Ag,Cu)(In,Ga)Se₂ (ACIGS), have shown significant progress in efficiency, reaching a record of 23.35% in 2019. However, further improvements have proven challenging. This research aims to address this challenge by exploring novel approaches to material composition and grading to optimize carrier collection and minimize recombination losses. The incorporation of silver (Ag) into the CIGS absorber has been shown to enhance grain growth, reduce structural disorder, and potentially improve the electronic properties of the material. A key focus is on the optimization of the gallium (Ga) concentration profile within the absorber layer, which significantly influences the bandgap energy and, therefore, the open-circuit voltage (Voc). Previous studies have utilized 'notch' or 'V-shaped' Ga profiles, but this work explores the impact of a 'hockey stick'-like profile. The post-deposition treatment (PDT) using RbF is another crucial aspect, known to passivate surface defects and enhance device performance. This study investigates the combined effect of these modifications on device efficiency, aiming to provide valuable insights for future advancements in chalcopyrite solar cell technology. The ultimate goal is to push the efficiency beyond 23.35%, aiming for values exceeding 25%. This is particularly relevant given the potential of chalcopyrite solar cells as a cost-effective and high-efficiency alternative to silicon-based solar cells.

Over the past decade, the efficiency of chalcopyrite solar cells has increased from about 20% to 23%, largely due to the incorporation of heavy alkali elements such as potassium, cesium, and rubidium into the absorber film and at its interfaces. The addition of silver to CIGS has been shown to facilitate grain growth by lowering the melting temperature and enhancing reaction rates. Silver partially replaces copper in the chalcopyrite lattice, impacting the structural disorder, interdiffusion of Ga and In, and the conduction and valence band edges. A record efficiency of 23.35% was achieved by Solar Frontier in 2019 using a sequential sulfurization after selenization approach, leading to a gradient in Ga concentration towards the Mo back contact, creating a back surface field. To further enhance Voc, sulfur was added at the absorber surface, gradually decreasing the valence band minimum and increasing the bandgap towards the buffer layer. Other research groups, such as ZSW, have utilized a three-stage co-evaporation method and a 'notch' GGI profile to achieve efficiencies of 22.6% in 2016. These studies highlight the ongoing efforts to optimize the absorber composition and interface properties to improve efficiency, setting the stage for the present work.

The champion solar cell in this study consists of a soda lime glass/Mo/NaF-precursor/ACIGS/RbF-PDT/CdS/i-ZnO/ZnO:Al/MgF2 stack. A 290-nm-thick Mo back contact was sputter-deposited and coated with a 10-nm NaF precursor layer. A modified three-stage co-evaporation process was used to grow the 2.0-2.1 μm thick ACIGS film at a maximum temperature of 530 °C. The final absorber composition was GGI of 0.28, AAC of 0.19, and [I]/[III] = 0.84, as measured by XRF. A RbF PDT was applied by depositing 3-5 nm of RbF at 350 °C, followed by a ~25-nm thick CdS buffer layer grown via CBD at 60 °C. The i-ZnO and ZnO:Al layers were sputtered, and an aluminum grid was deposited. A MgF2 anti-reflection coating was applied. Electro-optical characterization involved JV and EQE measurements after 24 hours of light soaking. The JV data were fitted using a one-diode model. PL measurements were conducted using a spectrophotometer equipped with an integrating sphere. The ERE was calculated using the integrated emission and scattering yields. Capacitance-voltage profiling was performed to determine doping profiles. Material characterization involved XRF, SEM, STEM-EDS, nano-XRF, and GDOES analyses to determine the composition, morphology, and elemental distribution throughout the device structure.

The champion device achieved a certified efficiency of 23.64% (23.75% in-house). The ideality factor (n=1.30) suggests reduced recombination in the space charge region. EQE analysis revealed parasitic absorption losses in the CdS buffer and window layers, highlighting potential for improvement. SEM images showed large ACIGS grains (>1 μm). Nano-XRF and STEM-EDS analysis revealed a 'hockey stick'-like Ga profile and relatively constant GGI in the upper half of the absorber, minimizing bandgap fluctuations. Rb was found agglomerated at interfaces and grain boundaries. The RbF PDT resulted in the formation of Rb-rich patches (possibly RbInSe₂) at the CdS/ACIGS interface. STEM-EDS line scans across the heterojunction revealed a thin (<5 nm) RbInSe₂ layer and possible Cd diffusion into this layer. Analysis of the back contact region showed Ga and In variations between grains, with Rb in most grain boundaries and Cu depletion alongside Ag enrichment. The external radiative efficiency (ERE) was measured at 1.6%, the highest reported for a chalcopyrite absorber. Comparison with state-of-the-art Si heterojunction solar cells showed comparable Voc deficits but lower FF and Jsc losses in the Si cell. The efficiency reached 72% of the SQ limit, indicating potential for further enhancement.

The achieved 23.64% efficiency demonstrates the effectiveness of the combined strategies implemented in this study. The 'hockey stick' Ga profile successfully minimized bandgap fluctuations, resulting in reduced Voc losses. The high silver concentration contributed to improved absorber quality. The RbF PDT passivated the surface and grain boundaries, further reducing recombination losses. The high ERE value suggests that further reduction of non-radiative recombination is key to achieve higher efficiency. The identification of parasitic absorption losses highlights opportunities for improvement through alternative buffer and window layers. These findings provide valuable insights for future research on chalcopyrite solar cells. The observed Rb-In-Se phase, though potentially not entirely beneficial, highlights the complex interplay between alkali treatments and absorber properties.

This study presents a 23.64% efficient chalcopyrite solar cell achieved through high-concentration silver alloying, a ‘hockey stick’ gallium profile, and RbF post-deposition treatment. The results underscore the importance of minimizing bandgap fluctuations and passivating grain boundaries and interfaces. Future work should focus on reducing parasitic absorption in the window and buffer layers and further improving the absorber quality to reach ERE values significantly exceeding 1.6% to achieve efficiencies beyond 25%.

While the study provides comprehensive characterization, some limitations remain. The interpretation of Rb-rich patches and their impact on performance requires further investigation. The lateral resolution of some techniques may have limited the detailed analysis of compositional variations within individual grains. The analysis of grain boundaries could benefit from higher-resolution techniques and more extensive statistical analysis. Some observed compositional changes might be influenced by the sample preparation techniques used for electron microscopy.