论文和技术资料

Black silicon in Figure 1 is a silicon surface engineered with micro- and nanostructures that exhibit strong optical absorption across a broad spectral range due to the light-trapping effect[1]. This unique optical property has made black silicon a promising material for various applications such as photodetection[1,2] and photovoltaics[3]. Accurately modeling black silicon is essential but challenging due to its inherently heterogeneous sub-wavelength surface features[4]. The finite element method (FEM) is particularly beneficial for modeling such complex surfaces thanks to its adaptive mesh. In this work, we use FEM in COMSOL Multiphysics to model black silicon and study the effect of nano-coatings on its surface, aiming to improve and extend its absorption spectrum.

Figure 2 illustrates a simulation model of black silicon, with its surface profile extracted from the cross-sectional SEM image shown in Figure 1. Assuming the profile length was significantly larger than the nano-features of black silicon, Floquet periodic boundary conditions were applied to the left and right sides of the model instead of perfectly matched layers (PMLs). PMLs were used at the top and bottom of the model and were covered with scattering boundary conditions. To monitor power flow, a reflection port was placed between the excitation port and the top PML, while a transmission port was located above the bottom PML. These ports were used to monitor the reflected power Pr and transmitted power Pt, respectively. Reflectance R and transmittance T were calculated as R=Pr/P0 ,T=Pt/P0 , where P0 is the incident power. The absorbance A of black silicon was determined as A=Pa/P0 , where Pa is the ohmic-loss power dissipated within the black silicon. The refractive index of silicon was from Green’s study[5] in the material library. The model was meshed using physics-controlled meshing with default parameters.

The model was initially validated using polished silicon, as shown in Figure 3(a), exhibiting excellent agreement between the simulation results and experimental data. Similarly, Figure 3(b) presents a comparison of reflectance for black silicon, demonstrating good consistency between the simulation and experiment. This validated model was further used to study optical behaviors of black silicon with nano-coatings. We studied various dielectric and metallic nano-coatings with different thicknesses and found that multiple thin dielectric and metallic layers can enhance the broadband absorption of black silicon.

This work demonstrates the effectiveness of FEM-based modeling for accurate simulation of complex black silicon structures and their optical responses.