| || || Sharika, Shalveen.|
| || || Investigation of a fibre-optic chemical sensor for in-situ measurement of pH in soils|
Institution: University of the South Pacific.
Call No.: pac In Process
Copyright:This thesis may NOT be copied without the authors written permission.
Abstract: An investigation was carried out on developing a pH sensor for soils based on fiber optic technology. The design is based on immobilizing a pH sensitive indicator dye on an inert substrate and coupling these with optical fibers that allow pH to be monitored optically. In the first part of the study, common acid-base indicator dyes were screened to identify those which would give appropriate optical signals for the pH range commonly encountered in soils. Soils generally have a pH value in the range of about 5 to 8. The optical signals of pH indicators when present in solution form were obtained to select the appropriate indicators to be utilized for sensor development. The five selected indicators were bromocresol green (BCG), bromocresol purple (BCP), bromothymol blue (BTB), methyl red (MR) and phenol red (PR). The next phase of the project dealt with immobilization of these indicators on selected inert substrates or supports. The supports chosen were three different Amberlite resins: XAD-2, XAD-4 and XAD-7. The five indicators were immobilized on the three resins to give a set of fifteen indicator/resin combinations. For sensor development, the best indicator/resin combination is one that is most stable and most sensitive. One of the requirements for stability is that the indicator does not readily leach from the resin and to establish this, leaching studies were carried out for all indicator/resin combinations. Resins with immobilized indicators were packed in mini-columns and eluted with a cycle of acidic and basic buffers, and distilled water. Leaching of the relevant indicator was monitored optically by measuring absorbance of the leachate at the characteristic absorption wavelength of the indicator. Leaching tests showed XAD-2 to be the best binding substrate as lowest amounts of indicators leached from this resin. XAD-7 was the most unsuccessful membrane in binding with all the indicators. Amongst the indicators, methyl red and phenol red leached out the most from all the resins. Certain indicators such as methyl red from XAD-4, methyl red, bromocresol purple and phenol red from XAD-7 leached out completely. Tests to establish optical signal characteristics after immobilization showed certain indicators giving very poor or no response at all with change in pH. These were almost all indicators adsorbed on XAD-2, except bromocresol green. The pKa values of the indicator in solution form appear to change from the pKa after immobilization. This gives a variation in the response range of the indicator after immobilization. The indicators which had a shift in the pH range of response after immobilization were bromothymol blue on XAD-4 and XAD-7 and phenol red on XAD-4 and XAD-7. Based on tests on stability, reproducibility, response time and pH range; bromocresol green on XAD-2, XAD-4 and XAD-7 were selected as the most appropriate combinations for pH sensor development. The next stage focused on testing performance in soils of pH 5 and 8. The pH of a sub sample of a soil with an existing pH value of 5 was adjusted to 8 by addition of the required amount of calcium carbonate and the original and treated soils were used in the study. One of the most critical factors in sensor performance in soils is the soil moisture content so a series of samples at both pH 5 and pH 8 were prepared containing different amounts of moisture. The soil: water (w/v) ratios used were 1:2.5, 1:1.5, 1:1 and 1:0.5. The performance of the sensor in detecting changes in soil pH was determined by allowing the sensor to first equilibrate in a pH 5 buffer solution and then placing it in a soil sample of pH 8 of given moisture content to determine its response characteristics. For the reverse cycle, it was allowed to equilibrate in pH 8 buffer solution first and then subjected to a pH 5 soil. The time required for the sensor to reach an equilibrium signal in the soil was recorded as the response time and this showed a strong dependence on moisture content.