The development of semitransparent CdS/CdTe ultrathin solar cells has been delayed as a result of the activation annealing to which the device must be subjected, which may involve problems such as the sublimation of ultrathin films and the diffusion of Cd and S at the interface. In this work, CdS/CdTe ultrathin devices on soda-lime glass/SnO2:F/ZnO substrates were obtained by RF magnetron sputtering. CdS/CdTe ultrathin heterostructures were obtained with the following thicknesses for the CdS thin film: 70, 110, and 135 nm. The CdTe thickness film was kept constant at 620 nm. Subsequently, activation annealing with CdCl2 was carried out at 400 °C. Surface characterization was performed by scanning electron microscopy, which indicated that the CdCl2 annealing tripled the CdTe thin films’ grain size. Raman characterization showed that CdS thin films deposited by RF sputtering present the first, the second, and the third longitudinal optical modes, indicating the good crystallinity of the CdS thin films. The study showed that the photovoltaic properties of the CdS/CdTe ultrathin devices improved as the CdS thicknesses decreased.
This work investigates a non-polluting alternative for activation treatment, consisting of using a saturated solution of MgCl2 in methanol. For this research, three types of solar cells were analyzed: without activation treatment, activation with CdCl2 deposited by sublimation in a closed space, and activation with a saturated solution of MgCl2 by immersion. The solar cells were characterized using several techniques: scanning electron microscopy images were analyzed to show how the grains of the CdTe layer changed after the activation treatment; using the data obtained through a solar simulator, improvements were observed in Voc, Jsc, and FF; efficiency η increased from 4.4% without activation to 11.92% when MgCl2 is used and to 7.42% with CdCl2; these results were contrasted through numerical simulation based on SCAPS-1D software. We propose that the CdS/CdTe heterojunction becomes thinner during activation due to diffusion between these films; in this way, the film thickness and doping concentration depend on the activation treatment. Finally, we demonstrate that the device can be activated with MgCl2, achieving a maximum efficiency of 11.92%; with values of 760 mV, 27.02 mA/cm2 and 0.58 of the Voc, Jsc, and FF respectively.