Login (DCU Staff Only)
Login (DCU Staff Only)

DORAS | DCU Research Repository

Explore open access research and scholarly works from DCU

Advanced Search

Multivariable process diagnostics for plasma systems

Coates, David (2021) Multivariable process diagnostics for plasma systems. Master of Engineering thesis, Dublin City University.

Abstract
Plasma processing has played a crucial and pivotal role in a number of industries, but probably one of the more significant has been the semiconductor sector. With plasma etching and surface treatment being key steps in wafer manufacturing, modest improvements in methods, methodologies, processes or control at the low level have very significant outcomes in the overall production of semiconductors and their varied uses. This work takes a multi-pronged approach from design and analysis of recently developed radio emission spectra capture to system level data acquisition over multi-sensors (which includes novel sensor approaches) for insertion into a hierarchical time series database structure, built upon a novel message-based network bus subsystem. Early work focused on the development of an optical viewport to obtain a planar view of the plasma. This was followed by the development of the non-invasive E and B-field radio probe technique for plasma monitoring RES (Radio Emission Spectroscopy), leading to the development of two branches of the sensor: the time-domain (TRES) and the frequency-domain (FRES). The two techniques can be applied in parallel to the commonly acquired data, but the processing techniques are somewhat different. The investigation of arcing within the chamber, which is very much a transient event, lends itself to TRES processing. TRES processing is by the application of novel near field B- and E-field analysis to detect these time-variant events where the frequency variation is small. The continuous operation of the plasma can be observed by applying digital signal processing (DSP) techniques to the FRES data, where the main measurement is frequency variant and the time resolution is large. Our approach is to use non-invasive measurement techniques from a range of sensors both in real time and in analytical modes. The analytical mode can be instantiated as an analysis technique, field-deployed, and as such can trigger a direct response to the process. The non-invasive sensors can be used as standalone to monitor operating processes and with simple thresholding can produce notifications of significant events. The work on the development of the novel FRES and the TRES sensor systems led to the investigation of the underlying scientific principles for the emission of such waves when driven by a single frequency. This FRES and TRES sensor subsystem has also been a commercial success with the technology being taken up by major semiconductor manufacturers for the determination of plasma conditions while processing wafers. The combination of novel non-invasive sensors, the real time alarm notification, the analysis over the broad-spectrum data and the instantiation of analytic blocks permits a very powerful and cost-effective tool in the control, maintainability and accounting of the production of semiconductor devices in a modern plant.
Metadata
Item Type:Thesis (Master of Engineering)
Date of Award:November 2021
Refereed:No
Supervisor(s):McNally, Patrick J. and Turner, Miles M.
Uncontrolled Keywords:Radio Emission Spectroscopy
Subjects:Engineering > Signal processing
Engineering > Systems engineering
Engineering > Electronic engineering
Physical Sciences > Spectrum analysis
Physical Sciences > Plasma processing
Physical Sciences > Plasmas
DCU Faculties and Centres:DCU Faculties and Schools > Faculty of Engineering and Computing > School of Electronic Engineering
Use License:This item is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 3.0 License. View License
ID Code:26060
Deposited On:29 Oct 2021 15:05 by Patrick Mcnally . Last Modified 29 Oct 2021 15:10
Documents

Full text available as:

[thumbnail of HardBound_Ver8.pdf] PDF - Archive staff only. This file is embargoed until 1 October 2025 - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
13MB
Downloads

Downloads

Downloads per month over past year

Archive Staff Only: edit this record