The aim of this project was to assess the feasibility of using UV spectroscopy for a simple detection system for formaldehyde gas in the indoor environment. Formaldehyde gas is hazardous to human health causing irritation of the eyes, nose and throat, headaches, limited pulmonary function and is a potential carcinogen. Formaldehyde derivatives are used in plywood and fibre board, carpeting, fabrics and some paints. The gas can be emitted from these materials and can build up in the indoor environment. Current methods for detecting formaldehyde gas that are simple, reliable and inexpensive are limited. A literature study of chemicals common to the indoor environment was carried out, and their UV absorption spectra compared to that of formaldehyde. 85 substances and substance groups were considered, 11 of which had absorption spectra that overlapped with the formaldehyde UV absorption band. A region was found between 320 and 360nm with very little spectral interference. Given the number of gases considered, this was a surprising result. Formaldehyde has several strong absorption peaks between regions of very low absorption, allowing for low resolution detection using a single LED source. Two prototype detection systems were developed. The first used a UV LED light source and used a beam splitter to provide one detection channel and one reference channel. The channels used narrow band (laser-line) optical filters. It was thoroughly optimised for noise performance, giving a best case limit of detection of 4.2ppm, limited by source fluctuation and shot noise, and LED thermal drift. Future developments could include a temperature controller inside the casing of the LED, or a multi- pass gas cell to increase sensitivity. With only two channels, the two filter system was susceptible to spectral interference from nitrogen dioxide and nitric acid. The second prototype system was developed using a novel method of passing un-collimated light through a laser-line filter to produce multiple wavelength channels with an angular spread. The principle was validated using two-channel detection with a limit of detection of 6ppm.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:613526 |
| Date | January 2014 |
| Creators | Davenport, John |
| Contributors | Hodgkinson, Jane; Tatam, Ralph P. |
| Publisher | Cranfield University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://dspace.lib.cranfield.ac.uk/handle/1826/8504 |
Page generated in 0.0017 seconds