Cadmium telluride (CdTe) photovoltaic technology stands out as the most cost-effective thin-film photovoltaic technology due to its low manufacturing costs and high module efficiency. The incorporation of selenium (Se) into CdTe, forming a Cd(Se,Te) alloy absorber, has significantly improved the record power conversion efficiency (PCE), rising from 16.5% in 2011 to over 22% in 2016. This improvement stems from the reduced bandgap of the Cd(Se,Te) absorber (less than 1.4 eV), enabling better light absorption at longer wavelengths and thus increasing the short-circuit current density (Jsc). While Jsc improved, open-circuit voltage (Voc) suffered, hindering overall PCE gains. Historically, only First Solar Inc. achieved high PCEs significantly exceeding those of CdTe solar cells. Recent progress involved replacing the commercial SnO₂ buffer layer with zinc magnesium oxide (ZMO), leading to Cd(Se,Te) solar cells with PCEs up to 19%. ZMO’s higher conduction band minimum (CBM) creates a CBM spike at the ZMO/Cd(Se,Te) interface, accumulating electrons and reducing nonradiative recombination. However, ZMO’s low electron conductivity, difficulty in doping, moisture sensitivity, and reproducibility issues remain significant challenges. The commercial SnO₂ buffer, known for its stability and conductivity, offers a more reliable alternative. This research explores methods to reduce recombination at the front interface, enabling efficient Cd(Se,Te) solar cell fabrication using the readily available SnO₂ buffer.

Bandgap gradients have proven effective in reducing non-radiative recombination and improving Vocs in other thin-film solar cells like Cu(In,Ga)Se₂ and Cu(Zn,Sn)Se₂. These gradients are typically created by introducing homovalent elements, such as varying the Ga/In ratio in Cu(In,Ga)Se₂ or incorporating Ag in Cu(In,Ga)Se₂ and Cu(Zn,Sn)Se₂. For Cd(Se,Te) cells, incorporating a CdS layer seemed promising; however, previous attempts resulted in the formation of a photo-inactive wurtzite Cd(S,Se) layer with a detrimental heterointerface, leading to increased recombination and lower performance. This paper addresses this prior challenge by exploring an alternative approach to create a photoactive bandgap gradient.

This study introduces a novel approach to create a photoactive region with a desirable bandgap gradient in Cd(Se,Te) thin-film solar cells without forming detrimental heterointerfaces. The key innovation lies in incorporating oxygenated CdS and CdSe layers (Cd(O,S), Cd(O,Se)) before depositing the CdTe absorber layer. CdCl₂ treatment then forms a penternary cadmium chalcogenide, Cd(O,S,Se,Te), with the same zinc blend structure as the absorber. This differs from previous attempts using pure CdS and CdSe layers, which resulted in a photo-inactive wurtzite Cd(S,Se) layer. The Cd(O,S,Se,Te) region possesses a wider bandgap than the surrounding Cd(Se,Te), creating a beneficial bandgap gradient that reduces hole density and introduces a small CBM spike. 1D SCAPS simulations were performed to model the impact of this gradient on device performance, showing significant reduction in nonradiative recombination and thus improved Voc and PCE. Time-resolved photoluminescence (TRPL), photoluminescence quantum yield (PLQY), cross-sectional scanning transmission electron microscopy (STEM), energy dispersive X-ray spectroscopy (EDS), and time-of-flight secondary ion mass spectroscopy (TOF SIMS) were employed to characterize the material properties and verify the effects of the compositional gradient. Control devices without the oxygenated layers were fabricated for comparison. The impact of bandgap and trap density on device performance was also explored through simulations.

SCAPS simulations predicted a significant improvement in Voc and PCE due to the reduced nonradiative recombination caused by the introduced Cd(O,S,Se,Te) region with a bandgap gradient. Experimental results strongly supported the simulations. Steady-state PL measurements revealed a much higher PL intensity in the target devices with the Cd(O,S,Se,Te) region compared to control devices when using a shorter wavelength excitation beam (405 nm), indicating reduced recombination at the interface. TRPL measurements showed significantly longer carrier lifetimes in the target devices, confirming the suppressed recombination. PLQY measurements further supported the reduction in recombination. The champion Cd(Se,Te) solar cell with the Cd(O,S,Se,Te) region achieved a remarkable PCE of 20.03%, with a Voc of 0.863 V, a Jsc of 29.2 mA cm⁻², and a FF of 79.5%. This is a significant improvement compared to the control devices (PCE of 18.3%, Voc of 0.834 V), demonstrating the effectiveness of the introduced compositional gradient in enhancing device performance while using a commercial SnO₂ buffer layer. The enhanced reproducibility of the target devices highlights the advantages of using the commercially available and stable SnO2 buffer. Cross-sectional HAADF-STEM and EDS mapping confirmed the formation of the penternary Cd(O,S,Se,Te) region without the formation of any detrimental heterointerfaces. TOF SIMS provided detailed compositional profiles of the region, which were used to estimate the bandgap of the penternary material.

The results demonstrate that the introduction of a photoactive Cd(O,S,Se,Te) region with a bandgap gradient near the front junction significantly enhances the performance of Cd(Se,Te) thin-film solar cells, addressing the long-standing challenge of achieving high Vocs with a commercial SnO₂ buffer layer. The suppression of nonradiative recombination at the front interface, confirmed through various characterization techniques, is the primary mechanism responsible for the observed improvement. This approach offers a practical and scalable method to improve the efficiency of Cd(Se,Te) solar cells while maintaining the advantages of using a commercially available and robust SnO₂ buffer. The findings hold significant implications for advancing the cost-effectiveness and performance of thin-film solar technology. The strategy of using oxygenated precursors to facilitate the formation of a desired graded composition inside the absorber is also highly promising for improving other types of thin film solar cells.

This work successfully demonstrates the creation of a bandgap gradient in Cd(Se,Te) thin-film solar cells using oxygenated CdS and CdSe layers. This approach led to a champion device with a 20.03% PCE, a significant advancement in the field. The enhanced performance is attributed to the reduced nonradiative recombination at the front interface. This method provides a pathway to highly efficient Cd(Se,Te) solar cells using commercially available SnO₂ buffer layers, offering improved stability and scalability compared to ZMO-based devices. Future research could explore optimizing the oxygen concentration and layer thicknesses to further enhance performance, as well as investigating the applicability of this technique to other thin-film solar cell technologies.

The study primarily focused on the front interface optimization. While bulk recombination was also improved to some extent, future work could investigate further bulk optimizations. The precise control over the oxygen concentration during deposition remains a challenge that could affect the reproducibility of the results. More detailed simulations and modeling may help further elucidate the impact of the specific compositional gradient and band alignment on device performance. The study's findings are based on laboratory-scale devices, and further research is needed to assess the scalability and long-term stability of these devices in real-world manufacturing conditions.