Loss Analysis for Thin-Film Solar Cells via Transfer Matrix and Electrical Finite Element Method

In order to understand and improve a solar cell’s power conversion efficiency (PCE), it is essential to identify and quantify loss mechanisms and thus guide experimental device optimization. We provide this functionality via loss analyses determined from computer-aided modeling and numerical device simulations. Since electrical and optical effects influence each other within solar cells, an isolated parameter variation is often not sufficient for maximizing cell performance. Therefore, a holistic perspective of the entire cell is developed within our digital twin of the solar device. The modeling of the digital twin is achieved by an interplay of an optical transfer-matrix method and an electrical finite element method.<br/>Within this work we identify and quantify all losses from semiconductor to cell level within a thin-film solar cell and allocate them to their corresponding loss mechanisms. We predict characteristic IV curves with high precision and the losses can be extracted retrospectively from the results of the simulation. The losses divide into optical and electrical losses. The optical loss mechanisms are split into grid shading, reflection, and parasitic and incomplete absorption and are quantitatively calculated by the transfer-matrix method. The electrical losses are divided into local MPP mismatches, under-grid reverse currents, and ohmic losses and are determined via the results of the electrical finite element method.<br/>We demonstrate the capability of our spatially resolved loss analysis on thin-film copper indium gallium diselenide (CIGS) cells with a thickness variation for the aluminum doped zinc oxide (ZAO) front contact. Optical complex refractive data and electrical specific resistivity data has been experimentally measured and used as simulation input. Hereby, spatially resolved voltage distributions across the cell are calculated via solving the discrete Poisson equation at each finite node. By varying the operating voltage, IV curves are forecast for different ZAO layer thicknesses. Their coefficients of determination in comparison with experimentally measured data is larger than 0.98 and hence, the values for short circuit current, fill factor, and open circuit voltage are precisely predicted. The consequent loss analysis on the basis of numerical device simulations crucially relies on this accurate IV curve modeling. We extract all loss mechanisms directly and quantitatively from both the optical transfer-matrix and electrical finite element simulation. For thick ZAO layers, optical parasitic absorption within the ZAO layer is by far the dominating loss mechanism, whereas for thin ZAO layers, local MPP mismatches and ohmic losses play a crucial role. Reflection losses are present at each ZAO thickness. The simulated optical interference as a function of ZAO thickness has been experimentally verified by reflectance measurements. Finally, the interplay of optical and electrical simulations within the device simulation leads to a forecast cell PCE. We experimentally confirmed these simulations by IV measurements of ten different ZAO layer thicknesses between 40nm and 1000nm.<br/>The presented loss analysis enables to detect and quantify weak points of thin-film solar cells by spatially resolved identification of optical and electrical loss mechanisms. This novel method opens the door for targeted laboratory work on loss reduction and paves the way towards further improvements in thin-film solar cell efficiencies.