Wet-prepared thin films of Cu2MnSnS4: structural study and photovoltaic performances

Cu2MnSnS4 (CMTS) is a promising candidate for application in thin film for photovoltaics (PV), thanks to sustainability and low cost of its components. Nevertheless, reported efficiencies are still poor, with a current world-record of 1.13% for a solar cell with CMTS absorber layer made by sputtering and sulfurization [1]. This work reports about a simple and low-cost synthesis procedure, starting from a solution containing all the precursors, without the need of hazardous external sulfurizing agents. CMTS thin films were made by blade coating of the solution on molybdenum substrate and followed by an annealing at 550°C under argon atmosphere, without the aid of external sulphurating agents. After annealing, samples were quenched at RT, but some selected samples underwent a slow cooling inside the furnace. Some selected CMTS thin films were subjected to HCl etching (3% HCl, at 75°C for 10 minutes). Prototype devices were made with the architecture Mo/CMTS/CdS/i-ZnO/AZO/Al grid: CdS was deposited by chemical bath deposition, i-ZnO and AZO by sputtering and Al grid was thermally evaporated. Some selected devices were subjected to a thermal treatment at 260°C for 10 minutes on titanium hotplate. After measurements of photovoltaic performances, upper layers were removed and CMTS thin films were extensively studied employing the GI-XRD technique to determine the lattice constants as well as microstructural properties like microstrain and size of coherent scattering domains. These results of the evaluation of GI-XRD data and photovoltaic performances were correlated to precursor composition, cooling conditions after annealing and post deposition treatments (i.e. HCl etching of the absorber before CdS deposition and thermal treatment of the complete device). All CMTS thin films analyzed by GI-XRD exhibited stannite structure, as expected [2], with a very good crystalline quality. The impact of metal ratio in the solution of precursors was studied, keeping copper poor conditions, and varying manganese quantity ([Mn]/[Sn]=1 (sample CMTS-1), [Mn]/[Sn]= 1.13 (sample CMTS-2) and [Mn]/[Sn]=1.27 (sample CMTS-3)). It was found through GI-XRD that CMTS-1 was the the composition displaying more narrow peaks and no hints of secondary phases, while CMTS-2 and CMTS-3 exhibited a feature in the diffraction pattern indicating the existence of stacking faults. Moreover, CMTS-3 could contain traces of MnS. CMTS-1 displayed also less microstrain and bigger coherent scattering domain size with respect to the other thin film. HCl etching of the bare absorber before CdS deposition has the effect of reducing microstrain (except for CMTS-2) and to enlarge coherent scattering domain size. Also stacking faults in CMTS-2 and CMTS-3 were reduced by HCl etching. In CMTS-1 also a reduction of lattice constant c was observed. Thermal treatment of the complete device (at 260°C for 10 minutes) has the effect to increase the microstrain and to elongate the unit cell along the crystallographic c-axis. These two phenomena are coherent with cadmium diffusion from CdS to the absorber substituting manganese and making the p-n junction “inside” the absorber layer, as already proposed for CIGS [3]. Combined effect of etching and thermal treatment resulted in intermediated microstrain and in a unit cell cell elongation along the crystallographic c-axis. CMTS thin films which underwent a slow cooling after annealing showed no difference in GI-XRD analysis, with comparable lattice constants, microstrain and domains size; but slow cooled samples, subjected to the thermal treatment of complete devices, showed less increase in microstrain with respect to the quenched ones. Results of PV measurements are coherent with findings from GI-XRD investigations and those regarding CMTS-1 are discussed in detail since their efficiencies are higher respect to the others. Thermal treatment of the complete device has the most pronounced effect enabling the functioning of the solar cell, with a switch from below 0.1% to 0.24%, suggesting again cadmium diffusion in the absorber. Beneficial effect of ageing was interestingly confirmed [1]: after 105 days, the cell with 0.24% efficiency reached 0.69%. Also, HCl etching has an important effect, but only after thermal treatment, with a jump from 0.24% to 0.60% (i.e. efficiency more than doubled). Combination of thermal treatment, HCl etching and ageing of 40 days were exploited in the campion device, that reached 0.92% efficiency, the current record for wet-synthesized CMTS [4]. Devices based on slow-cooled samples showed 0.40% efficiency even without thermal treatment, reaching 0.50% after that. [1] V. Trifiletti, L. Frioni, G. Tseberlidis, E. Vitiello, M. Danilson, M. Grossberg, M. Acciarri, S. Binetti, S. Marchionna, Manganese-substituted kesterite thin-films for earth-abundant photovoltaic applications, Solar Energy Materials and Solar Cells 254 (2023) 112247. https://doi.org/https://doi.org/10.1016/j.solmat.2023.112247. [2] G. Gurieva, S. Niedenzu, N. Siminel, A. Franz, S. Schorr, The kesterite–stannite structural transition as a way to avoid Cu/Zn disorder in kesterites: the exemplary case of the Cu2(Zn,Mn)SnSe4, Faraday Discuss 239 (2022) 51–69. https://doi.org/10.1039/D2FD00042C. [3] A. Koprek, O. Cojocaru-Miredin, R. Wuerz, C. Freysoldt, B. Gault, D. Raabe, Cd and Impurity Redistribution at the CdS/CIGS Interface after Annealing of CIGS-Based Solar Cells Resolved by Atom Probe Tomography, IEEE J Photovolt 7 (2017) 313–321. https://doi.org/10.1109/JPHOTOV.2016.2629841. [4] F. Butrichi, V. Trifiletti, G. Tseberlidis, B.E.G. Colombo, F. Taglietti, M. Rancan, L. Armelao, S. Binetti, Wet synthesis of Cu2MnSnS4 thin films for photovoltaics: Oxidation control and CdS impact on device performances, Solar Energy Materials and Solar Cells 272 (2024). https://doi.org/10.1016/j.solmat.2024.112924.