Optimisation and Power Scaling of Annular Transverse Laser Modes
Optimisation and Power Scaling of Annular Transverse Laser Modes
Laguerre-Gaussian transverse laser modes can have properties of a ‘vortex’ phase front that carries orbital angular momentum with each photon, or a spatially variant ‘vector’ polarisation state; both result in an annular intensity profile. These modes have been demonstrated to possess a range of advantages over those conventionally used in applications such as laser material processing, but generating high quality annular modes at significant output powers is difficult. To begin, vortex and vector annular modes are mathematically defined and their properties are described. Generation methods and applications of these annular modes are reviewed to date. A radially polarised vector LG01 mode source and a vortex phase LG01 mode source were constructed using a combination of previously demonstrated techniques. These modes were optimised for modal purity for use in other experiments. The radially polarised source was measured to have a high polarisation extinction ratio of 100:1, with an output power of 5.9 W. The vortex mode source had an output power of 210 mW. The lack of a quantified metric of beam quality suitable for higher-order transverse modes motivated the development of a convolutional neural network (named BeamNet) able to predict the modal composition of an intensity profile in real-time using only a camera and a computer. BeamNet was able to predict modal compositions in just 49 µs per image with an error rate of 5.2% when trained on 9 possible modes. BeamNet was initially demonstrated as an alignment aid on a Yb:YAG laser, then used as part of an automated feedback loop to lock the otherwise variable modal composition of a Ho:YAG laser to a fundamental mode or radially polarised LG01 mode. Spatially variant waveplates previously had optical losses that were too high for efficient intracavity generation of radially polarised modes. The demonstration of a refined low-loss waveplate in the cavity of a Nd:YVO4 laser and the subsequent generation of a radially polarised mode is presented. The additional cavity loss of the spatially variant structure is measured as just 0.4%. Finally, the amplification of the developed annular mode sources was explored in a thermally guiding fibre rod. This is a new hybrid amplifier architecture with excellent thermal handling and very large mode areas that shows promise for scaling the powers of annular modes. A radially polarised mode and a vortex phase LG01 mode are successfully amplified. The modes and their properties were well preserved after amplification, with a maximum achieved gain of 7.7 dB.
University of Southampton
Jefferson-Brain, Thomas
8bce2a02-37a4-4277-a8cb-0c40bde57837
Jefferson-Brain, Thomas
8bce2a02-37a4-4277-a8cb-0c40bde57837
Clarkson, William
3b060f63-a303-4fa5-ad50-95f166df1ba2
Jefferson-Brain, Thomas
(2021)
Optimisation and Power Scaling of Annular Transverse Laser Modes.
University of Southampton, Doctoral Thesis, 135pp.
Record type:
Thesis
(Doctoral)
Abstract
Laguerre-Gaussian transverse laser modes can have properties of a ‘vortex’ phase front that carries orbital angular momentum with each photon, or a spatially variant ‘vector’ polarisation state; both result in an annular intensity profile. These modes have been demonstrated to possess a range of advantages over those conventionally used in applications such as laser material processing, but generating high quality annular modes at significant output powers is difficult. To begin, vortex and vector annular modes are mathematically defined and their properties are described. Generation methods and applications of these annular modes are reviewed to date. A radially polarised vector LG01 mode source and a vortex phase LG01 mode source were constructed using a combination of previously demonstrated techniques. These modes were optimised for modal purity for use in other experiments. The radially polarised source was measured to have a high polarisation extinction ratio of 100:1, with an output power of 5.9 W. The vortex mode source had an output power of 210 mW. The lack of a quantified metric of beam quality suitable for higher-order transverse modes motivated the development of a convolutional neural network (named BeamNet) able to predict the modal composition of an intensity profile in real-time using only a camera and a computer. BeamNet was able to predict modal compositions in just 49 µs per image with an error rate of 5.2% when trained on 9 possible modes. BeamNet was initially demonstrated as an alignment aid on a Yb:YAG laser, then used as part of an automated feedback loop to lock the otherwise variable modal composition of a Ho:YAG laser to a fundamental mode or radially polarised LG01 mode. Spatially variant waveplates previously had optical losses that were too high for efficient intracavity generation of radially polarised modes. The demonstration of a refined low-loss waveplate in the cavity of a Nd:YVO4 laser and the subsequent generation of a radially polarised mode is presented. The additional cavity loss of the spatially variant structure is measured as just 0.4%. Finally, the amplification of the developed annular mode sources was explored in a thermally guiding fibre rod. This is a new hybrid amplifier architecture with excellent thermal handling and very large mode areas that shows promise for scaling the powers of annular modes. A radially polarised mode and a vortex phase LG01 mode are successfully amplified. The modes and their properties were well preserved after amplification, with a maximum achieved gain of 7.7 dB.
Text
Jefferson-Brain,_T_L_soton_e-thesis for Award
- Version of Record
Text
PTD_Thesis_Jefferson-Brain-SIGNED
Restricted to Repository staff only
More information
Submitted date: 14 July 2021
Identifiers
Local EPrints ID: 456756
URI: http://eprints.soton.ac.uk/id/eprint/456756
PURE UUID: 82d53a8c-76a6-4a47-9574-306ada7d5bc6
Catalogue record
Date deposited: 10 May 2022 16:54
Last modified: 18 Oct 2024 16:53
Export record
Contributors
Author:
Thomas Jefferson-Brain
Thesis advisor:
William Clarkson
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