Abstract: The mechanism by which mixing noise is produced in flow injection analysis with thermal lens detection is discussed. An He - Ne laser was used to measure the homogeneity of a mixture and the measurements used to determine the relative mixing efficiency of several reactors. Results were used to improve mixing efficiencies in packed-bead reactors. Mixing noise is predominant when mixing solution with differences in composition of 1 mM or higher. The main contribution to mixing noise comes from changes in probe beam direction produced by refraction through irregular liquid - liquid interphases. Packed bead reactors proved much more efficient than open-tube coiled reactors. With suitable reactor design it is possible to reduce mixing noise to below the noise levels from other sources.

Abstract: The use of acid - base indicators to monitor pH changes during flow injection analyzes was investigated. The flow injection system incorporated an injection valve, a flow cell (18 µL), a coil reactor made from a PTFE tube and 0.5 mm i.d. PTFE tubing; a single-bead string reactor following the coil reactor was used for experiments involving bovine serum albumin. A pH meter and a fast-scanning diode-array spectrophotometer linked to a computer completed the assembly. To illustrate the system, in addition to HCl - NaOH systems containing pH indicators, experiments were carried out with a sample flow solution of aqueous bovine serum albumin (25 µM) at several pH values containing 10 µM-thymol blue; the carrier solution contained phthalaldehyde - N-acetyl-L-cysteine - H3BO3 - borate buffer of pH 9.5. The flow rate was 1.9 mL min-1. The pH and the absorbance were continuously monitored, one spectrum per s being recorded. The method allows effective monitoring of pH gradients within the spectrophotometric cell at the same time as the flow injection analysis. In the presence of an acid-base indicator and using fast diode-array spectrophotometric scans, pH gradients in the spectrophotometric cell can be measured at the same time as any other colorimetric flow injection (FI) determination is performed. The effects of refractive index changes, adsorption-desorption processes at the tube walls and association of the indicator species with the system to be investigated are considered. Effective correction for the first two effects is demonstrated. Indicators with different mol. structures and polarities gave similar pH gradients, suggesting adequate accuracy for most potential applications. Application to the detection and correction of systematic and random errors in pH-sensitive FI procedures is demonstrated using the o-phthalaldehyde-N-acetyl-L- cysteine method for bovine serum albumin. In the presence of an acid-base indicator and using fast diode-array spectrophotometric scans, pH gradients in the spectrophotometric cell can be measured at the same time as any other colorimetric flow injection (FI) determination is performed. The effects of refractive index changes, adsorption-desorption processes at the tube walls and association of the indicator species with the system to be investigated are considered. Effective correction for the first two effects is demonstrated. Indicators with different molecular structures and polarities gave similar pH gradients, suggesting adequate accuracy for most potential applications. Application to the detection and correction of systematic and random errors in pH-sensitive FI procedures is demonstrated using the o-phthalaldehyde-N-acetyl-L-cysteine method for bovine serum albumin.

Abstract: The determination of the thyreostatics 2-thiouracil, its derivatives (4-methyl-2-thiouracil, 4-propyl-2-thiouracil and 4-phenyl-2-thiouracil) and methimazole in manufactured dried animal feed by micellar electrokinetic chromatography (MEKC) is described. A 99±5% extraction yield at the 20 mg g 21 level (n = 8) was achieved by shaking the milled fodder with methanol-1 m NaOH (80 + 20). Aliquots of the supernatant were injected in a 75 mm 3 33.5 cm uncoated silica capillary using pressure; separation was performed at 23 degC with 15 kV (positive polarity) in a background electrolyte (BGE) containing 40 mm sodium dihydrogenphosphate, 50 mm sodium dodecyl sulfate and 15 mm Tween 20 at pH 9. When the surfactants were added to the BGE, all the thyreostatics were well resolved and the fodder extracts showed lower backgrounds. The peaks appeared within the 2.25-5.2 min range with efficiencies in the 2.5 3 10 4 -8 3 10 4 range; methimazole appeared in the vicinity of the electroosmotic migration time. Calibration curves were linear within the studied range (20-200 mg mL 21 , r 2 >0.998). Limits of detection in the extracts of spiked fodder samples ranged from 0.25 to 0.4 mg mL 21 , which corresponded to 0.6-1.0 mg of drug per gram of fodder. Peak area repeatabilities were about 4% at the 20 mg mL 21 level.