In the manufacturing process of float glass often atmospheric pressure chemical vapour deposition (APCVD) reactors are integrated on-line for the deposition of functional thin solid films. Such functional films have applications in architectural glass, flat panel displays and solar cells. As glass moves downstream in the process, the thin film is deposited at temperatures between 500 to 700°C. The high temperatures make it difficult to monitor the deposition process and thin film quality control is commonly done at the end of the line or at lower temperatures. A time delay therefore exists between the point of thin film deposition and subsequent quality control, which can lead to large quantities of defective product being produced before faults are detected. It is therefore desirable to monitor in the APCVD reactor for rapid feedback of unexpected deviations from desired process conditions, reaction progress and fault detection. High uniformity of film properties across the substrate are important, but APCVD reactors are often empirically designed and the detailed chemical reaction mechanism is unknown. This leads to inefficient gas flow patterns and precursor utilization as well as difficulties in the design of new reactors. The APCVD deposition of tin oxide from the mono-butyl-tin tri-chloride (MBTC) is an example of such a process. Optical monitoring instruments in-situ and in-line on the APCVD reactor provided rapid feedback about process stability and progress non-invasively. Near infrared diode laser absorption spectroscopy (NIR-LAS) monitored the concentration of the reaction species hydrogen chloride (HCl) in-situ and spatially in the coating zone. A mid-infrared grating absorption spectrometer (IR-GAS) with novel pyro-electric array detector monitored the concentration of precursor entering the coating system simultaneously. In combination these instruments provide the means for rapid process feedback. Fourier transform infrared absorption spectroscopy (FTIR) was used to investigate the unknown decomposition pathway of the precursor to find the yet unknown key tin radical that initiates film growth. Stable species forming during MBTC decomposition over a temperature range of 170 to 760°C were investigated but the tin intermediate remains unknown. Computational fluid dynamics (CFD) is routinely employed in research and industry for the numerical simulation of CVD processes in order to predict reactor flow patterns, deposition rates, chemical species distribution or temperature profiles. Two and three dimensional models with complex geometries and detailed reaction models exist. A three dimensional computational fluid dynamics (CFD) model of the used APCVD reactor was built using the Fluent CFD software. The numerical simulation included a chemical model that predicted qualitatively the chemical species distribution of hydrogen chloride in the gas phase. This was confirmed through comparison with NIR-LAS results. Design shortcomings due to inefficient flow patterns were also identified. In combination the optical tools developed provide the means for safe and efficient manufacturing of thin films in APCVD reactors. CFD simulations can be used to increase precursor utilization and film uniformity in the development of new reactor designs.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:618035 |
| Date | January 2014 |
| Creators | Hehn, Martin Christoph |
| Contributors | Martin, Philip |
| Publisher | University of Manchester |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | https://www.research.manchester.ac.uk/portal/en/theses/diagnostics-and-modelling-of-atmospheric-pressure-chemical-vapour-deposition-reactors(e3a692e7-2b47-4d5d-9d4c-3997a88893f6).html |
Page generated in 0.0018 seconds