Nanophotonic waveguides for chip-scale raman spectroscopy: Theoretical considerations

Abstract

Highly evanescent nanophotonic waveguides enable extremely efficient Raman spectroscopy in chip-scale photonic integrated circuits due to the continuous excitation and collection of Raman scattering along the entire waveguide length. Such waveguides can be used for detection and identification of condensed-phase analytes, or, if functionalized by a sorbent as a top-cladding, can be used to detect trace concentrations of chemical species. The scattering efficiency is modified in guided-mode structures compared to unconfined, micro-Raman geometries. Here, we describe the theoretical framework for understanding the Raman scattering efficiency in nanophotonic waveguides, and compare these calculations to our measurements of trace gases in hypersorbent-clad silicon nitride waveguides.

Waveguide-enhanced Raman spectroscopy using visible laser excitation

“…The spectra obtained with a 633 nm laser are approximately 4× stronger than those obtained with the 785 nm laser at small wavenumbers, and over 10× stronger at longer wavenumbers. These signal differences are due to a number of factors, including waveguide loss, edge coupling, trench insertion loss, WERS efficiency 16 and the wavelength dependence of the Raman cross-section. An accounting of the specific contribution of each of these factors will be the subject of a future publication.…”

Waveguide-Enhanced Raman Spectroscopy (WERS): An Emerging Chip-Based Tool for Chemical and Biological Sensing

“…To assess the theoretical performance of WERS for biosensing, Jun et al [ 45 ], Dhakal et al [ 46 ], and Stievater et al [ 22 ] first developed a theoretical framework for the excitation, collection, and enhancement of Raman signal with Si 3 N 4 waveguides, before testing the theory experimentally. The emitting molecule is modeled as an oscillating dipole near the waveguide surface.…”

“…The emitting molecule is modeled as an oscillating dipole near the waveguide surface. In the discussion below, we present the key aspect of WERS theory, drawing from [ 22 , 45 , 46 ].…”

“…Although the results obtained were found to have a significant standard deviation, the experimental data were consistent with the equation. Likewise, work from the Stievater lab demonstrated good correlation between theory and experiment, but also some variability due to variation in waveguide fabrication and other factors [ 22 ]. Later, Dhakal et al [ 48 ] validated the equation for both TE and TM mode using waveguides with different lengths, widths, designs (rectangular and slot waveguides), and refractive indices (Si 3 N 4 , TiO 2 , and silicon).…”

“…In contrast to SERS, where Raman scattering only comes from the enhancing hotspots, in WERS, analytes can interact with the evanescent field over the entire length of the waveguide, and the waveguide also collects and conducts the Raman scattering to the spectrometer. In theory, waveguides provide dozens-fold enhancement of Raman signal per cm [ 22 ]. For example, a direct experimental comparison of WERS with a standard confocal Raman microscope reported by Dhakal et al [ 23 ] achieved more than four-orders-of-magnitude higher spontaneous Raman signal for WERS.…”