Dataset for "Silicon-rich silicon nitride: tailoring material properties for photonic integrated circuits dataset"

Dataset for thesis entitled "Silicon-rich silicon nitride: tailoring material properties for photonic integrated circuits" (2024).

Abstract for thesis:
Silicon-rich silicon nitride is a promising hybrid material platform for use in photonic integrated circuits. During the deposition process the silicon content of the bulk material can be altered, allowing the material properties of the new hybrid material to be tailored between those of silicon and silicon nitride. Customisation of the material platform between the advantages of high-index contrast systems and low-index contrast systems is therefore possible. Furthermore, the bandgap energy and nonlinear Kerr index are related to the refractive index. Hence, optimisation of these properties can be accomplished via the silicon content of the material platform. In this work a series of silicon-rich silicon nitride platforms with refractive indices from the stochiometic 1.97 to 2.95 at a wavelength of 1550 nm have been developed and their linear optical properties and bond structures analysed as the silicon content increases. Propagation losses between 3.27 dB/cm and 6.67 dB/cm were measured on waveguides fabricated from 500 nm thick films. Examination of the bandgap reveals a power law relationship between the bandgap energy and the refractive index, which agrees well with literature values for silicon and silicon nitride. Log-log plot analysis of this relationship reveals a transition point between two linear regimes at a refractive index of approximately 2.5 at 1550 nm. This transition point matches with the onset of two-photon absorption reported in the literature. The underlying cause the change in optical properties is attributed to a transition from a fully amorphous structure of the silicon nitride to one where silicon nanoclusters are present. Furthermore, by dispersion engineering of waveguide geometry it proved possible to design waveguide samples capable of sustaining the nonlinear phenomenon of supercontinuum generation using the same waveguide dimensions. Using these samples, it was possible to confirm that increasing silicon content also increases the dispersion. Supercontinua as broad as 800 nm -wider than the broadest state of the art measurement for ultra silicon-rich silicon nitride with equivalent refractive indices - were obtained using a 100 fs pulsed excitation, with a centre wavelength of 1560 nm. By optimising the waveguide geometry for each silicon-rich silicon nitride platform, larger supercontinua is anticipated to improve upon these measurements. These results are of particular importance when designing new chip-scale devices. The thesis is embargoed until 31 July 2025.

This dataset contains:

All data collected during doctoral study. Note that this includes data not used in the thesis. Data is organised into folders based on the equipment used to collect the data. Shortcuts to Matlab scripts used to generate figures can be found in the 'Figure shortcuts' folder.

Folder structure:
Bandgap Measurements - Spectrophotometer data
Bragg Gratings - Refractive index measurements taken using the Fabry-Perot setup & the Bragg grating simulations
Ellipsometry - Ellipsometer data
FTIR - Fourier transform infrared spectroscopy data
Propagation Loss Measurements - Propagation loss data taken using the Fabry-Perot setup
RTA - Rapid Thermal Anneal data from ellipsometry & FTIR
Selectivity - Interferometry data taken during sample etching
Simulations - Simulations of waveguides
Supercontinuum Generation - Supercontinuum generation data taken using free space laser setup

Collected data was processed using Matlab scripts to analyse the data.

Software used: Matlab 2023a (.m, .mat) (peakfit.m), Woollam CompleteEase (.mod, .se) (ellipsometry data), Bruker OPUS (.0)(FTIR), Excel, (.xlsx, .csv), Notepad (.txt), Lumerical Mode Solutions (.lsf,.lms)

Data from CompleteEase, OPUS and Mode Solutions were converted into .csv, .txt and .mat files for analysis in Matlab. Figure generators can be found in 'Figure shortcuts'.

Dataset embargoed until 31 July 2025.