Arnold G. Fogg

Direct reductive amperometric determination of nitrate at a copper electrode formed in situ in a capillary-fill sensor device

A method has been developed for determining nitrate amperometrically by direct reduction at a freshly deposited copper electrode surface in a capillary-fill device (CFD). Copper(II) is added to the nitrate sample which is then taken up into the device. The potential of the screen-printed carbon electrode is held at –0.75 V versus the screen-printed silver reference electrode. At this potential, copper is plated onto the carbon electrode forming a freshly prepared copper electrode. At the same time dissolved oxygen is reduced. The potential is then scanned to more negative potentials and the signal at –0.90 V, due to the reduction of the nitrate, is measured. The method for determining nitrate given here is preliminary to the production of CFDs in which chemical reagents, copper sulphate and potassium hydrogen sulphate (used to produce the acidity), are screen-printed or otherwise coated onto the upper plate within the device.

Shapes of normal and reverse flow injection signals: on-line formation of iodine from iodate, iodide and hydrogen ion with detection by visible spectrophotometry

Signal shapes for the on-line formation of iodine from iodate, iodide and hydrogen ion in a single-channel manifold using large-volume slug and large-volume time-based injections have been determined using visible spectrophotometry. These large injection volume studies were made first as a means of understanding the shapes of normal and reverse flow injection signals obtained at more conventional injection volumes (10–100 µl). The signal shapes at large injection volumes were determined for the six possible combinations of the reagents in the two solutions serving as carrier stream and injectate so that one solution contained two reagents and the other solution one or two reagents. Each combination of reagents represents two complementary systems in which the roles of each solution as carrier stream and injectate are reversed. At these large injection volumes each signal consisted of two independent peaks caused by dispersion at the front and rear boundaries of the injected bolus. The signals obtained for the time-based injections for complementary systems were identical in shape and height except that the front peak of one system was identical with the rear peak of its complementary system and vice versa. Clearly, at such large injection volumes the terms normal flow injection (nFl) and reverse flow injection (rFl) have no real meaning, the shape of each independent peak being determined by the composition and relative positions of the two solutions forming the boundary at which the peak is formed. For slug injections, similar shapes were observed but the peak heights were affected markedly by the greater dispersion at the rear boundary which travels further than the front boundary. This comparison of the signals obtained with slug and time-based injections, despite different flow-rates being used for the two modes of injection, clearly shows the effect of the unequal dispersion at the two boundaries in the slug injection method. Examination of the signals obtained with time-based injection, however, clearly indicates that the solution compositions, and their relative positions in the flow stream, also affect the shapes and relative heights of the front and rear peaks. The shapes of all these signals are illustrated. The effect of reducing the slug injection volume stepwise from 2 ml to 100 µl was studied for the IO3–l– < H+ and H+ < IO3–l– systems (< denotes the direction of the boundary shape). This indicated that the shapes and heights of the single peaks observed in the rFl and nFl formation of iodine carried out at the more conventional lower injection volumes are determined by dispersion at the rear and front boundaries of the bolus, respectively. Hence, as the two peaks observed in a large-volume injection merged as the injection volume was decreased, the major peak predominated and became the observed signal. The use of a much smaller injection volume was necessary in rFl than in nFl in order to obtain a single peak.

Shapes of flow injection signals: effect of refractive index on spectrophotometric signals obtained for on-line formation of bromine from bromate, bromide and hydrogen ion in a single-channel manifold using large-volume time-based injections

The shapes of the spectrophotometric signals obtained with a single-channel manifold for large-volume (4 ml) time-based injections for the six possible combinations of the reagents bromate, bromide and nitric acid in the injectate and carrier stream, by which bromine can be formed on-line, have been determined. The injectate and carrier stream were 5.25 × 10–4M in bromate, 0.030 M in bromide and 1 M in nitric acid when these reagents were present. The signals consisted of two separate peaks caused by formation of bromine at the front and rear boundaries of the injected bolus. When both injectate and carrier stream were 1 M in nitric acid (i.e., for the reagent combination H+BrO3–-H+Br–) the two peaks were of equal height, and the signal was virtually the same whichever solution was used as the injectate. In reagent combinations where only one solution contained nitric acid the peaks were different in size, the smaller peak being that produced by the boundary in which the acidic solution was flowing behind the other solution. This difference in size between the front and rear peaks was shown to be caused by refractive index effects. When the refractive indices of the two solutions were matched either by increasing the potassium bromide concentration or by making the non-acidic solution 7% in sodium nitrate, the peaks became equal in size. When the potassium bromide concentration was increased there was an appreciable increase in peak size (about 4-fold): the changes in the amount of bromine formed must be due to kinetic or equilibrium effects. This increase in size did not occur when sodium nitrate was used to balance the refractive indices.

Determination of trace amounts of carcinogenic substances: adsorptive stripping voltammetry of 1-[4′-(phenylazo)phenyl]-3,3-dimethyltriazene at a hanging mercury drop electrode

1-[4′-(Phenylazo)phenyl]-3,3-dimethyltriazene can be determined in the range from 1 × 10–5 to 2 × 10–7 mol dm–3 by direct current or differential-pulse voltammetry at a hanging mercury drop electrode. Direct current adsorptive stripping voltammetry can be used in the range from 2 × 10–8 to 10 × 10–8 mol dm–3 and differential-pulse adsorptive stripping voltammetry in the range from 1 × 10–8 to 2 × 10–11 mol dm–3. The extremely high sensitivity in the latter instance was achieved by using 200-fold diluted Britton–Robinson buffer as a base electrolyte together with a long accumulation time (up to 30 min at 1 × 10–11 mol dm–3). The relative standard deviation (seven determinations at 2 × 10–11 mol dm–3) was 8.6%.

Differential-pulse adsorptive stripping voltammetric determination of sodium cromoglycate at a hanging mercury drop electrode

Sodium cromoglycate is strongly adsorbed onto mercury, and can be determined readily at the 1 × 10–10 mol dm–3 level, after adsorption onto a hanging mercury drop electrode from pH 4.5 acetate buffer solution. The adsorbed cromoglycate anion was reduced at –0.87 V versus Ag—AgCl, after accumulation at –0.6 V. Slightly higher signals can be obtained by accumulating at potentials more positive than the potential of zero charge (–0.5 V). Accumulation at –0.6 V, however, eliminates interference from more easily reduced adsorbates, and the analysis time is decreased. At the 0.5 × 10–7 and 1 × 10 –7 mol dm–3 levels, peak height increased rectilinearly with accumulation time up to 5 and 3 min, respectively, and peaks of useful size were observed at accumulation times as short as 15 s. Relative standard deviations at the 1 × 10–7 mol dm–3 level were typically <4%(five determinations). No significant interference was observed at this level from 1 mg dm–3 levels of several surfactants and 1 × 10–6 mol dm–3 levels of several heavy metal ions. Human serum albumin at the 1.6 × 10–7 mol dm–3 level caused the height of the peak to be reduced to 25% of its original size. The method was tested satisfactorily by analysing four formulation solutions of the drug.

Adsorptive stripping voltammetric behaviour of copper(II) at a hanging mercury drop electrode in the presence of excess of imidazole

In the presence of excess of imidazole (1.0 × 10–3 mol dm–3), copper(II), at pH 8.5, adsorbs at a hanging mercury drop electrode to give two adsorptive stripping voltammetric peaks at –0.36 and –0.46 V. The peak at –0.36 V is only present at accumulation potentials more negative than –0.05 V versus Ag—AgCl: it increases in height as the accumulation potential becomes more negative up to and beyond –0.6 V, the voltammetric sweep being started at –0.20 V. This peak appears to be due to the adsorption of polymeric [CuII(lm)2] or its reduced copper(I) form. The peak at –0.46 V is only present at high imidazole concentrations (>5.0 × 10–4 mol dm–3): the accumulation is uniform from 0.0 to –0.36 V but is negligible at potentials more negative than –0.46 V. This peak appears to be due to adsorption of [Cu(lm)4]2+. On cycling between 0.0 and –0.6 V this latter complex is converted into the polymeric complex and only the peak at –0.36 V remains. Copper(II) can be determined by using the peak at –0.46 V after accumulation at 0.0 V, or at –0.36 V after accumulation at –0.6 V. The latter method is more sensitive: the detection limit is about 2.0 × 10–9 mol dm–3 after accumulation for 3 min.

Elimination of double peaks in the iodimetric flow injection visible spectrophotometric determination of sulphite using a single-channel manifold

A flow injection method of determining sulphite with amperometric monitoring of iodine using a single-channel manifold in which iodine is formed in the reverse flow injection (rFl) manner and reacts with sulphite dispersing in the normal flow injection (nFl) manner has been adapted for use with visible spectrophotometry. The carrier stream consists of an alkaline solution containing iodate and an excess of iodide: injections of acid and then of acidified sulphite are made. The decrease in the iodine signal (measured at 352 nm) in the presence of sulphite is proportional to the sulphite concentration in the injectate. The alkalinity of the carrier stream was adjusted to reduce the signal widths and to prevent the appearance of double peaks. A rectilinear decrease in signal size (down to ca. 10% of the signal size in the absence of sulphite) was obtained in the range 1 × 10–5–7 × 10–4M sulphite using a single-channel manifold consisting of 3 m of 0.8 mm bore transmission tubing with a flow-rate of 5 ml min–1 and an injection volume of 15 µl, when the carrier stream was 6.7 × 10–6M in iodate, 6.7 × 10–2M in iodide and 3.5 × 10–3M in sodium hydroxide, and the sample solution was 0.1 M in hydrochloric acid.

Reductive reverse flow injection amperometric determination of nitrate at a platinum electrode after on-line reduction to nitrosyl chloride in concentrated sulphuric acid medium containing chloride

A simple method has been developed for the determination of nitrate using flow injection with amperometric detection. The nitrate sample, which is made 2 M in hydrochloric acid, is the carrier stream, and into this is injected a small volume (25 µl) of concentrated sulphuric acid. The nitrosyl chloride and chlorine formed are detected as a combined reduction signal at a platinum electrode held at +0.70 V versus Ag-AgCl. As formation of nitrosyl chloride only occurs in the initial stages of dispersion of the sulphuric acid into the sample solution (i.e., at high concentrations of sulphuric acid), incorporation of a 40-cm long single bead string reactor as the transmission tubing between the injection valve and the detector, which reduces this dispersion, increases the size of the signal and provides conditions in which gas (mainly hydrogen chloride) formed on-line, which otherwise results in a noisy base line, is re-dissolved before reaching the detector. Calibration graphs were rectilinear over the range from 9 × 10–5M(the detection limit) to 5 × 10–3M. Coefficients of variation were typically less than 4%. Other common constituents of hydroponic fluids did not interfere.

Cathodic stripping voltammetric determination of pentamidine isethionate at a hanging mercury drop electrode

Differential-pulse cathodic stripping voltammetry was used for the determination of trace amounts of pentamidine isethionate at a hanging mercury drop electrode using its reduction peak at –1.57 V in 0.2 mol l–1 sodium hydroxide. The optimum accumulation potential and accumulation time were –1.1 V and up to 180 s, respectively. Linear calibration graphs were obtained up to 1 × 10–6 mol l–1: the limit of detection was calculated to be 3.0 × 10–10 mol l–1. The effect of various components of urine on the voltammetric response was studied, and albumin, creatinine and uric acid caused interference in the method. The direct determination of the drug (>1 × 10–7 mol l–1) in urine can be effected using a high dilution of the sample.

Electrochemical reduction at mercury electrodes and differential-pulse polarographic determination of pentamidine isethionate

Reduction processes are observed for pentamidine isethionate at a dropping mercury electrode above pH 7: the reduction potential is independent of pH below about pH 10. Below pH 10, adsorption of the reduced species is observed, whereas above pH 10 there is a large contribution owing to adsorption of pentamidine isethionate. For the polarographic determination of pentamidine isethionate, a pH of 8–9 is recommended with the addition of Triton X-100 as a maximum suppressor. Pentamidine isethionate can be determined by differential-pulse polarography down to about 5 × 10–6 mol l–1.