Diamond micromachining of photonic materials: dynamics and the limits of precision
Diamond micromachining of photonic materials: dynamics and the limits of precision
This thesis investigates the ultra-precision diamond micromachining of brittle photonic materials, focusing on optimising machining processes to achieve high-quality surfaces for applications in optical and quantum devices. It explores the dynamics of physical machining in materials such as lithium niobate, silica, and silicon, which are crucial in photonics due to their optical properties. By physically machining within the ductile regime, chip-free surfaces were achieved, offering nanometre-scale roughness and high uniformity, without requiring costly post-processing. Ridge waveguides, optical facets, and alignment structures were fabricated with a focus on minimising surface roughness, form errors, and defects. Surface roughness measurements for ridge waveguide sidewalls reached as low as 0.16 nm when using a 100 µm square-profiled blade. The impact of feed rate on blade wear rate and resulting surface roughness was investigated over a range of 0.1 mm s¯¹ to 10 mm s¯¹, revealing a critical shift at 2 mm s¯¹ in blade wear rate and a similar shift in surface roughnesses at 4 mm s¯¹ in lithium niobate. A comparison of blade wear, surface finish, and waveguide performance provides insights into deterministic machining.
A novel confocal probe-based scanner has been developed to profile the long-term variation of ridge waveguide uniformity, indicating an approximate 1.2 µm curvature across all of the ridge waveguides fabricated. A comparison of ridge waveguides fabricated with 300 µm and 100 µm wide dicing blades was performed, resulting in a significant improvement in optical conversion efficiency — an increase of 226% for the ridges machined with a 300 µm wide blade over a 100 µm wide blade. Plotting the accumulative width variation using the confocal probe traces reveals that the ridges machined with the 300 µm wide blade were significantly more consistent in their widths, confirming the cause of the improved optical conversion efficiencies.
The development of a novel 3D passive alignment structure based around machined U-Grooves and optical fibres has been proposed; however, due to inconsistent machine stepper backlash, practical success has been limited. V-Grooves and U-Grooves have been compared from a general coupling alignment perspective, with U-Grooves resulting in a 30% improvement in fibre-to-fibre alignment over an unsupported gap of 3 mm compared to V-Grooves. This has resulted in improvement of coupling efficiency from 97% to 99%.
University of Southampton
D'Souza, Matthew Peter
9e6e1f39-1d57-4848-b4ac-5aafd7ff10df
2025
D'Souza, Matthew Peter
9e6e1f39-1d57-4848-b4ac-5aafd7ff10df
Gates, James
b71e31a1-8caa-477e-8556-b64f6cae0dc2
Gow, Paul
193394b1-fe2d-41de-a9aa-6de7e5925b18
Gawith, Corin
926665c0-84c7-4a1d-ae19-ee6d7d14c43e
Smith, Peter G.R.
8979668a-8b7a-4838-9a74-1a7cfc6665f6
D'Souza, Matthew Peter
(2025)
Diamond micromachining of photonic materials: dynamics and the limits of precision.
University of Southampton, Doctoral Thesis, 132pp.
Record type:
Thesis
(Doctoral)
Abstract
This thesis investigates the ultra-precision diamond micromachining of brittle photonic materials, focusing on optimising machining processes to achieve high-quality surfaces for applications in optical and quantum devices. It explores the dynamics of physical machining in materials such as lithium niobate, silica, and silicon, which are crucial in photonics due to their optical properties. By physically machining within the ductile regime, chip-free surfaces were achieved, offering nanometre-scale roughness and high uniformity, without requiring costly post-processing. Ridge waveguides, optical facets, and alignment structures were fabricated with a focus on minimising surface roughness, form errors, and defects. Surface roughness measurements for ridge waveguide sidewalls reached as low as 0.16 nm when using a 100 µm square-profiled blade. The impact of feed rate on blade wear rate and resulting surface roughness was investigated over a range of 0.1 mm s¯¹ to 10 mm s¯¹, revealing a critical shift at 2 mm s¯¹ in blade wear rate and a similar shift in surface roughnesses at 4 mm s¯¹ in lithium niobate. A comparison of blade wear, surface finish, and waveguide performance provides insights into deterministic machining.
A novel confocal probe-based scanner has been developed to profile the long-term variation of ridge waveguide uniformity, indicating an approximate 1.2 µm curvature across all of the ridge waveguides fabricated. A comparison of ridge waveguides fabricated with 300 µm and 100 µm wide dicing blades was performed, resulting in a significant improvement in optical conversion efficiency — an increase of 226% for the ridges machined with a 300 µm wide blade over a 100 µm wide blade. Plotting the accumulative width variation using the confocal probe traces reveals that the ridges machined with the 300 µm wide blade were significantly more consistent in their widths, confirming the cause of the improved optical conversion efficiencies.
The development of a novel 3D passive alignment structure based around machined U-Grooves and optical fibres has been proposed; however, due to inconsistent machine stepper backlash, practical success has been limited. V-Grooves and U-Grooves have been compared from a general coupling alignment perspective, with U-Grooves resulting in a 30% improvement in fibre-to-fibre alignment over an unsupported gap of 3 mm compared to V-Grooves. This has resulted in improvement of coupling efficiency from 97% to 99%.
Text
Southampton_PhD_Thesis_Archival
- Author's Original
Text
Final-thesis-submission-Examination-Mr-Matthew-D-Souza
Restricted to Repository staff only
More information
Published date: 2025
Identifiers
Local EPrints ID: 502057
URI: http://eprints.soton.ac.uk/id/eprint/502057
PURE UUID: 7af7809c-e619-4c53-bf69-2bc1e46b3a83
Catalogue record
Date deposited: 16 Jun 2025 16:32
Last modified: 11 Sep 2025 03:18
Export record
Contributors
Author:
Matthew Peter D'Souza
Thesis advisor:
James Gates
Thesis advisor:
Paul Gow
Thesis advisor:
Corin Gawith
Thesis advisor:
Peter G.R. Smith
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics