The theoretical and practical aspects of the sensitivity that can be achieved in a flow injection manifold with on-line chemical reactions are investigated with a view to decreasing the detection limits. Optimization of the common manifold configurations, normal single line, double line and reverse single line, was carried out by varying the parameters of flow rate(s), injected volume, reagent concentration and manifold length. The reaction of iron(II) with 1,10-phenanthroline was used for the comparison. Ascorbic acid was added to both sample and reagent to avoid side reactions of 1,10-phenanthroline and iron(III). Results showed that manifold and confluence point geometry affected peak height. A sensitivity comparison, using a 21 point calibration, showed that the manifold sensitivity is similar but that the reverse single line $>$ double line $>$ normal single line. The double line manifold has the best detection limit (10 ng mL$\sp{-1})$ due to the absence of a refractive index effect. Experiments were performed to look at the use of solid molybdate reagent columns for the determination of phosphate. Solid columns without any support were found to be very unstable. Stability was improved by the addition of 5 $\mu$m silica but it was still not reproducible enough. Precipitation of barium molybdate within a cation exchange resin did produce better reproducibility but the complexity added made it impractical. Membrane reagent introduction for the determination of phosphate and chloride was investigated. Nafion cation exchange tubing was used for the introduction of acid and hydrazinium ion for phosphate, and acid, iron(III) and thiocyanate and mercury(II) for chloride. Molybdate was introduced for phosphate determination using a supported liquid membrane of tri-butyl phosphate in n-heptane. Sensitivity was improved for the phosphate determination but not for the chloride determination. The use of an ammonia permeation tube was studied for calibration in flow injection. Nessler's reagent was used for the determination. Release rate temperature variation and day to day reproducibility were investigated. Analysis of a sample of pond water was achieved using the permeation tube for on-line, zero dilution, standard additions calibration.
Identifer | oai:union.ndltd.org:UMASS/oai:scholarworks.umass.edu:dissertations-8943 |
Date | 01 January 1994 |
Creators | Chalk, Stuart James |
Publisher | ScholarWorks@UMass Amherst |
Source Sets | University of Massachusetts, Amherst |
Language | English |
Detected Language | English |
Type | text |
Source | Doctoral Dissertations Available from Proquest |
Page generated in 0.0015 seconds