Research on Thickness Measurement of Perovskite Coatings Based on Energy-Dispersive X-ray Intensity Ratio

[Background] Perovskite solar cells are solar cells that use perovskite-type metal halide semiconductors as light-absorbing materials. The perovskite crystal coating serves as the core functional layer, and the uniformity of its thickness will affect various performances of the cell and even determine the long-term stability of the cell's operation in a humid and hot environment. [Purpose] This study proposes a thickness measurement method based on the combination of Monte Carlo simulation and X-ray intensity ratio in response to the demand for rapid non-destructive testing of the coating thickness of perovskite solar cells. [Methods] The Monte Carlo model of perovskite coating was established by XMI-MSIM software, with a Rh target X-ray tube as the excitation source. The standard curve of X-ray intensity ratio to thickness was generated, and its reliability was verified by combining the experimental data of energy dispersive X-ray fluorescence (EDXRF). [Results] The Monte Carlo model of the perovskite coating was established using XMI-MSIM software with an Rh-target X-ray tube as the excitation source. The simulated structure included a glass substrate, SnO₂ layer, and FAPbI₃ coating (270–900 nm). X-ray fluorescence spectra were simulated to obtain net peak intensities of elements such as I, Ca, Sn, and Si. Four types of X-ray intensity ratios were calculated: I-Lβ₁/I-Lα, I-Lα/Ca-Kα, Ca-Kα/Si-Kα, and Sn-Lα/Si-Kα. These ratios were fitted against coating thickness to establish calibration curves. For validation, EDXRF experiments were conducted on perovskite samples with the same layer structure using a Mo-target X-ray tube and a silicon drift detector. Experimental intensity ratios were compared with simulated values, and coating thicknesses were inversely predicted using the calibration curves. The method's accuracy was evaluated by calculating relative errors between simulated and experimental ratios, as well as between predicted and actual thicknesses.[Conclusions] In the experimental verification, the relative error between the simulated and experimental data of the I-Lβ1/I-Lα element group was less than 3.5%, and the relative error between the inversion thickness and the actual thickness was less than 15%. The effect was the best among all element groups, confirming the applicability of the proposed thickness measurement method in the thickness measurement of multi-layer complex coatings.