Improving thin-film solar cells through optical modeling
Thin-film solar cells convert solar energy into electrical energy, but require less material and lower temperatures than their conventional counterparts. Typically, thin-film solar cells comprise semi-transparent layers with thicknesses measured in microns and nanometers. Also, they have textured interfaces that scatter incoming light, prolonging the optical path. This improves light trapping and optical absorption by the layers, ultimately increasing the overall efficiency of solar-to-electric energy conversion.
The texture of rough interfaces is random, making it difficult to to describe and model the light scattering. Neither scalar nor vector scattering theories, nor equations derived from them, can be directly used. These equations must be calibrated in order to yield realistic results when applied to thin-film solar cells.
We developed a one-dimensional model for optical analysis of thin-film solar cells.1 Realistic complex refractive indexes were used to describe the optical properties of the layers, and both specular (non-scattered) and scattered light at the textured interfaces were taken into account. The specular light was analyzed in terms of electromagnetic waves, including interference in the simulated layers, while a ray-tracing method was used for scattered light.
The scattered light at the textured interfaces was characterized by its the scattering level (‘haze’), and angular distribution. Haze at the internal interfaces was determined from scalar scattering theory equations calibrated by measurements on solar cell textured substrates.
The optical model was implemented in a user-friendly simulator named SunShine, and applied to different types of thin-film solar cells, such as silicon solar cells (amorphous,2 microcrystalline,3 micromorph,4 and HIT cells), Cu(In,GA)Se2-based solar cells,5 and various amorphous silicon photodetectors.
The SunShine model was then used to explore ways to improve the efficiency of the micromorph tandem solar cell, which consists of an amorphous silicon (a-Si:H) top cell and a microcrystalline silicon (μc-SI:H) bottom cell.4 The scatter levels were calibrated by the optical parameters of realistic state-of-the-art micromorph solar cells, which typically achieve efficiency of around 10% (see Figure 1). The simulated optimized solar cell has a short-circuit current potential above 16mA/cm2, according to simulations, which suggests potential efficiency exceeding 15%.
Based on the consistency between our simulations and the observed cell performance, several modifications were made to increase the simulated efficiency. These included improved back-reflector reflectance, and decreased supporting-layer absorptance. Also implemented were tweaks to the interface scattering levels, broadened scattered-light angular distribution (Lambertian), and optimized antireflective coatings on the front glass substrate. A further modification provided wavelength-selective properties for the intermediate reflector between the top and bottom cells (see Figure 2).
In another application, SunShine was used to design a highly reflective back reflector. High optical reflectance is essential at the back side of a solar cell for efficiently reflecting light back to the absorber layer. Back reflectors made from textured metals such as aluminium and silver are typically used for this purpose. However, metal reflectors suffer from plasmon absorption, which deteriorates their reflectivity properties. Dielectric-based reflectors may overcome this problem.
We simulated the properties of a non-metal back reflector using a one-dimensional photonic-crystal-like structures with high reflectivity.6 Based on thin films—which can be deposited at low-temperatures and are compatible with the deposition processes of thin-film solar cells—these structures were modeled as four paired layers of intrinsic amorphous silicon (i-a-Si:H) and silicon dioxide (SiO2) with thicknesses of 55nm and 135nm, respectively (see Figure 3). Nearly 100% reflectance was achieved in the wavelength range of interest for silicon micromorph solar cells (see Figure 4).
Optical modeling is a useful tool for optimization and improvements of thin-film silicon solar cells. It enables insight into internal optics and provides basis for further optimizations of the structures.
The authors thank Miro Zeman from the Delft Institute of Microelectronics and Submicron Technology for useful discussions on modeling thin-film solar cells, as well as experimental support.