Combining black silicon (BS), a nanostructured silicon containing highly roughened surface morphology with plasmonic materials, is becoming an attractive approach for greatly enhancing light-matter interactions with promising applications of sensing and light harvesting. However, precisely describing the optical response of a heavily decorated BS structure is still challenging due to the increasing complexity in surface morphology and plasmon hybridization. Here, we propose and fully characterize BS-based multistacked nanostructures with randomly distributed nanoparticles on the highly roughened nonflat surface. We demonstrate a realistic 3D modeling methodology based on parametrized scanning electron microscopy images that provides high-precision morphology details, successfully linking the theoretical analysis with experimental optical response of the complex nanostructures. Far-field calculations very nicely reproduce experimental reflectance spectra, revealing the dependency of light trapping on the thickness of the conformal reflector and the atop nanoparticle size. Near-field analysis clearly identifies three types of stochastic "hotspots". Their contribution to the overall field enhancement is shown to be very much sensitive to the nanoscale surface morphology. The simulated near-field property is then used to examine the measured surface-enhanced Raman scattering (SERS) response on the multistacked structures. The present modeling approach combined with spectroscopic characterizations is expected to offer a powerful tool for the precise description of the optical response of other large-scale highly disordered realistic 3D systems.
Keywords: SERS; black silicon; hotspots; random complex nanostructures; realistic 3D modeling; surface plasmons.