ANALYST, AUGUST 1986, VOL. 111 879 Application of the Reductive Flow Injection Amperometric Determination of Iodine at a Glassy Carbon Electrode to the lodimetric Determination of Hypochlorite and Hydrogen Peroxide Antoine Y. Chamsi and Arnold G. Fogg Chemistry Department, Loughborough University of Technology, Loughborough, Leicestershire LEI1 3TU, UK Dissolved molecular oxygen is insignificantly reduced at a glassy carbon electrode that is being held at -0.2 V vs. SCE for the determination of iodine. The tendency for molecular oxygen to oxidise iodide to iodine in acidic solutions, however, is a serious problem in the development of flow injection amperometric methods based on the use of iodine reactions in which iodine is monitored at -0.2 V. In the work described here, iodine (I2), has been determined at concentrations above 6 x 10-6 M in neutral, unsparged potassium iodide solution by injection through PTFE transmission tubing into a nitrogen-sparged 0.4% potassium iodide eluent in a flow injection system.By sparging the sample solution with nitrogen for 30 s before injection and shielding the PTFE transmission tubing with nitrogen, it was possible to determine iodine at concentrations above 1 X 10-7 M, with no significant loss of iodine being caused by the sparging. A nitrogen-sparged eluent, consisting of 0.04% potassium iodide in 0.3% acetic acid solution, remains free of iodine for several hours when stored under nitrogen. Hypochlorite was determined directly on-line in the range 0.1 x 10-4-2.5 x 10-4 M by injecting nitrogen-sparged sample solutions into this eluent.The reaction is fast and complete on-line. Hydrogen peroxide was determined on-line by injecting nitrogen-sparged sample solutions, 2.5 M in sulphuric acid, into a nitrogen-sparged 5% potassium iodide eluent. The reaction of hydrogen peroxide with iodide was incomplete on-line. Iodine formed in the molybdate-catalysed and the uncatalysed reactions of hydrogen peroxide with iodide off-line was determined by injection into a 1 M sulphuric acid eluent. Keywords : Flow injection analysis; amperometric detection; iodine determination; h ypochlorite determination; hydrogen peroxide determination Iodine titrations are useful in analytical chemistry because of the relatively low standard reduction potential of the 12 - 21- couple (0.54 V vs.NHE). This low reduction potential means that iodine reacts more selectively than many other oxidising agents and also that iodide is oxidised to iodine by a wide range of oxidising agents. The couple is reversible and very sharp end-points are possible in titrations with thiosulphate using visible means of detecting iodine, starch indicator or amperometric or biamperometric end-point detection. In iodimetric methods, the reduced forms of couples with lower reduction potentials are determined by monitoring the amount of iodine used to oxidise them, whereas oxidised forms of couples with higher standard reduction potentials are determined by the amount of iodine produced on reaction with iodide. The possibility of carrying out iodimetric reactions on-line using a flow injection technique with amperometric detection of iodine at a glassy carbon electrode has been investigated.Previous work in this laboratory has included on-line bromina- tion reactions and hypochlorite has been determined by injection into acidic bromide solutions.* Lee and Pollard3 have developed a spectrophotometric flow injection method of determining the iodine values of fatty acids. Miller et ~ 1 . ~ have described two flow injection methods using continuous and stopped-flow spectrophotometric detection of iodine; in their work, exclusion of air from the flow system was essential in order to avoid oxidation of iodide by oxygen. Experimental The flow injection analysis was carried out in a single-channel system similar to that described previously.5 The eluent flow was produced by means of a Metrohm pressure bottle (EA l l O l ) , which has a maximum nitrogen pressure limit of about 0.8 bar.The sample (a fixed volume of about 100 pl) was injected with a low-pressure Rheodyne valve (5020) connected between the pressure bottle (which contained the eluent) and a laboratory-built detector cell by means of 0.58-mm bore PTFE tubing. The detector cell only holds the glassy carbon electrode, the eluent being introduced to the stationary electrode in a wall-jet configuration. The detector cell was partly immersed in 0.1 M potassium chloride solution. The platinum counter and saturated calomel reference electrodes were placed in this electrolyte to complete the three-electrode system. A newly polished Metrohm glassy carbon disc electrode (EA286,3 mm diameter) was used.For the determination of iodine, the potential of the glassy carbon electrode was held at -0.2 V vs. SCE by means of a PAR 174A polarographic analyser (Princeton Applied Research). Values of currents were recorded on a Tarkan 600 y - t recorder. Eluents were sparged with nitrogen [which had been passed through a vanadium(I1) scrubber] in the pressure vessel before use; nitrogen pressure systems are ideal for maintaining air-free eluents in the reservoir. The values of background currents (the base lines from which current signals were measured) were not normally noted as they did not affect measurements at the levels of iodine monitored here. The background current at -0.2 V vs. SCE that was obtained for the nitrogen-sparged 0.4% potassium iodide solution was 10 nA.Other solutions were sparged where indicated in the text. For determinations of 12 at levels lower than 6 X 10-6 M, sample solutions were sparged with nitrogen for 30 s. At 1 2 levels lower than 1 x 10-6 M, nitrogen sparging of the sample solution was carried out for 30 s directly in the syringe by a method similar to that of Lloyd6; the PTFE transmission tubing was contained within PVC tubing of larger bore and nitrogen was passed through the space between the two tubes. In the applications described in this paper, it would not normally be necessary to work at such low concentrations of iodine.880 ANALYST, AUGUST 1986, VOL. 111 Eluents and Reagents Standard iodine (12) solution, 0.05 M.Dissolve 2g of potassium iodide in about 4 ml of water in a glass-stoppered 100-ml calibrated flask and add, through a small dry filter funnel, about 1.27 g of iodine (analytical-reagent grade). Insert the stopper and agitate the flask until all the iodine has dissolved. Allow the solution to reach room temperature and dilute to 100 ml with water. Store the solution in a cool dark place in a well stoppered glass bottle. Prepare more dilute standard solutions by diluting this solution with 0.4% potassium iodide solution. It is not necessary to de-oxygenate the water used to prepare this solution. The solution was standardised titrimetrically with standard sodium thiosulphate solution7 before use. Flow Injection Determination of Iodine A 0.4% nitrogen-sparged solution of potassium iodide was chosen as the eluent for this work.A 1-m length of 0.58-mm bore transmission tubing and a flow-rate of 6.0 ml min-1 were used. Currents obtained at various fixed potentials for the injection of nitrogen-sparged and unsparged solutions of the same composition as the eluent and for a 30 s-sparged 1.1 x 10-5 M solution of I2 in 0.4% potassium iodide solution are given in Table 1. The effect of not sparging the injected solution clearly becomes more significant at potentials more negative than -0.2 V. At -0.2 V, however, the increase in current when sparging of the injected solution is not carried out is less than 1% of the signal for the 1.1 x 10-5 M I2 solution. No loss of signal by contamination of the electrode was observed during over 100 injections of iodine at this level and the coefficient of variation was less than 1%.Calibration graphs were rectilinear over the range 1 X 10-7-1 x 10-3 M 12. The further precautions to exclude air described above were taken in determining 12 levels below 6 x 10-6 M. Currents at -0.2 V vs. SCE, corrected for the blank at the 4 X 10-7, 1 x 10-6 and 1 x 10-5 M I2 levels, were 24.8, 59.2 and 600 nA, respectively. Standard sodium hypochlorite solution, about 0.05 M. Dilute 20 ml of sodium hypochlorite solution (about 10% available chlorine) and 25 ml of 2 M potassium hydroxide solution to 500 ml with water. This solution was freshly prepared and standardised by reaction with iodide and titration of the liberated iodine with standard sodium thiosulphate solution as required.’ Standard hydrogen peroxide solution, about 0.07 M.Dilute 10 ml of 20 volume hydrogen peroxide solution to 250 ml with water. This solution was freshly prepared and standardised iodimetrically as required, allowing 15 min for reaction with iodide before titrating the iodine formed with standard sodium thiosulphate solution.7 Potassium iodide (0.04%) in 0.3 M acetic acid solution (eluent). Add 17.25 ml of glacial acetic acid to about 500 ml of water, add 0.4 g of potassium iodide dissolved in water and dilute to 1 1. This eluent was freshly prepared as required and was sparged with nitrogen for at least 15 min before use. On-line Determination of Hypochlorite Commercial hypochlorite products are usually alkaline, as hypochlorite is more stable under these conditions.In this study, hypochlorite was injected as solutions 0.1 M in potas- sium hydroxide; these solutions were sparged with nitrogen for 30 s before injection. Optimisation studies indicated that a transmission tube 1 m in length and with a 0.58 mm bore and a flow-rate of 7-9 ml min-1 should be used and the lower flow-rate of 7 ml min-1 was adopted. Currents associated with the electrochemical reduction at various measurement poten- Table 1. Effects of measurement potential on the values of the current signals obtained for the injection of 100 pl of nitrogen-sparged and unsparged blanks with the same composition as the eluent and of a 1.1 X 10-5 M solution of I2 in 0.4% potassium iodide solution (that had been sparged with nitrogen gas for 30 s) into a nitrogen-sparged 0.4% potassium iodide eluent Potential vs.SCE/V . . . . . . 0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5 Signal for unsparged blank/nA . . 4 4 4 4 6 9 14 Signal for sparged blank/nA . . 6 6 8 10 18 58 182 Signal for sparged iodine solution/nA . , . . . . . . 292 456 578 664 680 680 685 Table 2. Effects of measurement potential on the values of current signals obtained for the injection of 100 p1 of 5.75 x lops M hypochlorite in nitrogen-sparged 0.1 M potassium hydroxide into sparged eluents of various acidities all containing 0.4% potassium iodide. The signals obtained with sparged blanks are also given: these values have been subtracted from the values given for injection of hypochlorite PotentialiV 0.2 Signal in 0.1 M potassium hydroxide*/pA .. . . . . <O. 00 1 Blank/pA . . . . . . . . <0.001 Signal in 0.1 M dipotassium hydrogen phosphate*/pA . . 0.016 Blank/pA . . . . . . . . 0.004 Signal in 0.1 M potassium dihydrogen phosphate*/pA . . 0.12 Blank/pA . . . . . . . . 0.024 Signal in 0.3 M acetic acid*/pA . . . . . . . . 0.94 Blank/pA . . . . . . . . 0.010 Signal in 0.1 M sulphuric acid*/pA . . . . . . . . 1.84 Blank/pA . . . . , . . . 0.014 * Containing 0.4% potassium iodide. ~~ 0.1 <0.001 <0.001 0.008 0.006 0.84 0.024 1.65 0.010 2.60 0.004 0 <0.001 <0.001 0.032 0.008 1.60 0.028 2.55 0.008 2.95 0.004 -0.1 0.002 <0.001 0.076 0.008 2.43 0.028 2.94 0.004 3.04 0.004 -0.2 0.006 <0.001 0.108 0.012 2.46 0.024 3.10 0.004 3.14 0.004 -0.3 0.010 0.002 0.148 0.016 2.74 0.016 3.10 0.006 3.20 0.005 -0.4 0.012 0.010 0.108 0.076 2.74 0.012 3.08 0.010 3.20 0.007 -0.5 0.036 0.020 0.056 0.140 2.88 0.008 3.05 0.018 3.10 0.012ANALYST, AUGUST 1986, VOL.111 881 tials of the iodine formed were obtained for the injection of 100 pl of 5.75 X 10-5 M hypochlorite in nitrogen-sparged 0.1 M potassium hydroxide solution into eluents of various acidities, all of which contained 0.4% potassium iodide and had been sparged with nitrogen for at least 15 min before use. These currents and those obtained for the injection of sparged blanks are given in Table 2. From the values of these currents, it is apparent that a reasonably high acidity is required for the full formation of iodine although the signals obtained with 0.3 M acetic acid were not significantly lower than those with 0.1 M sulphuric acid.Because the formation of iodine by the reaction of iodide with dissolved molecular oxygen is much slower in acetic acid solution than in sulphuric acid solution, the use of the 0.3 M acetic acid eluent containing potassium iodide was preferred. The effect of the potassium iodide concentration of the eluent was next studied. The current signals obtained at -0.2 V vs. SCE for identical injections of nitrogen-sparged hypochlorite into sparged 0.3% acetic acid eluent containing 0, 0.004, 0.04, 0.4 and 4% potassium iodide were 0.22, 4.16, 4.04,3.76 and 3.04 PA, respectively. The reduction in current at higher potassium iodide concentrations probably arises owing to the increased viscosity and density of the solutions. A potassium iodide concentration of 0.04% was adopted here to ensure an excess of iodide at the higher levels of hypochlorite determined.Calibration graphs were shown to be rectilinear Table 3. Values of current signals at -0.2 V vs. SCE obtained for the injection of 100 p1 of 6.4 X M hydrogen peroxide in various concentrations of nitrogen-sparged sulphuric acid into sparged 1 YO potassium iodide eluent. Blank signal for 5 M sulphuric acid = 0.17 PA; others lower but not recorded. Blanks not subtracted Sulphuric acid concentrationh . . 0.01 0.1 0.5 1 2.5 5 Signal/yA . . . . . . 0.76 1.14 2.04 3.68 8.64 13.9 Table 4. Values of current signals (FA) at -0.2 V vs. SCE obtained for the injection of 100 pl of 6.4 x M hydrogen peroxide in nitrogen-sparged 1 and 2.5 M sulphuric acid solutions into sparged eluents consisting of solutions of various concentrations of potassium iodide.Blank signal for 5% potassium iodide/2.5 M sulphuric acid eluent = 0.16 FA. Blanks not subtracted Current/pA Potassium iodide concentration/% m/V . . 0.002 0.01 0.1 1 5 Sulphuricacideluent (1 M) . . 0.16 0.19 0.34 3.68 9.6 Sulphuric acid eluent (2.5 M) 0.17 0.19 0.76 8.68 16 over the range 0.13 x 10-4-2.6 X 10-4 M and the coefficient of variation (five determinations) was less than 1% at the 0.65 x 10-4 M level. On-line Determination of Hydrogen Peroxide Optimisation studies indicated that a 0.58 mm bore trans- mission tube of 3 m length and a flow-rate of 1 ml min-1 were satisfactory and these values were therefore adopted. The formation of iodine was incomplete in this on-line determi- nation of hydrogen peroxide; however, increasingly greater amounts of iodine were formed when higher concentrations of sulphuric acid and iodide were used.This is illustrated in Tables 3 and 4, which show the currents obtained when 100-pl volumes of 6.4 x 10-4 M hydrogen peroxide as sparged solutions with different concentrations of sulphuric acid were injected into a 1% potassium iodide eluent which had been sparged for 15 min. The tables also show the currents obtained when hydrogen peroxide was injected as 1 and 2.5 M sulphuric acid solutions into sparged eluents consisting of various concentrations of potassium iodide. Calibration graphs were obtained for the injection of hydrogen peroxide in sparged 2.5 M sulphuric acid solution into a 5% potassium iodide eluent, which had been sparged for 15 min.These were found to be rectilinear over the range 0.06 x 10-3-6 X 10-3 M (Table 5) with coefficients of variation (five determinations) typically less than 1%. Flow Injection Determination of Hydrogen Peroxide with Off-line Formation of Iodine As the formation of iodine was determined to be only about 50% complete in the on-line determination of hydrogen peroxide, some laboratories may prefer to use flow injection amperometry only to replace the step involving titration of the iodine formed in an off-line reaction of iodide and hydrogen peroxide. This simple application of the flow injection amperometric determination of iodine is illustrated here. The titration methods given by Vogel7 were adapted for use with FIA.A transmission coil length of 1 m (0.58 mm bore) and a flow-rate of 6.0 ml min-1 were used. A volume of 2.5 ml of 20% potassium iodide solution and and 0-5 ml of sample solution (0.05 x 10-2-6 X 10-2 M in hydrogen peroxide). were added to 40 ml of 1 M sulphuric acid contained in a 50-ml calibrated flask. After diluting with 1 M sulphuric acid, the flask was firmly stoppered and the solution allowed to stand for 15 min in order to allow the iodine to form fully before injecting the solution into nitrogen-sparged 1 M sulphuric acid eluent. The same reaction was carried out adding 2 drops of 3% ammonium molybdate tetrahydrate to the reaction solution as a catalyst. In this latter instance, the iodine forms rapidly and the injection is made immediately on dilution.Table 5. Values of current signals at -0.2 V vs. SCE obtained for the on-line determination of hydrogen peroxide by injection in sparged 2.5 M sulphuric acid solution into sparged 5% potassium iodide eluent. Blank not subtracted Hydrogen peroxide concentration ( X 1 0 - 5 M ) . . . . . . . . o 6.2 18.6 43.4 62.2 434 622 Signal/pA . . , . . . . . 0.16 1.68 4.72 11.4 15.7 113 154 Table 6. Values of current signals at -0.2 V v s . SCE obtained for the off-line determination of hydrogen peroxide using the uncatalysed reaction and the reaction catalysed by ammonium molybdate Hydrogen peroxide concentration Signal for uncatalysed Signal for catalysed ( X 10-4 M) . . . . . . . . o 0.031 0.093 0.23 0.62 1.00 1.86 2.0 4.34 6.00 18.6 62.2 * * * * 4.40 * 9.20 * 28.8 92 300 reaction/pA .. . . . . . . 0.1 reaction/wA . . . . . . . . 0.08 0.19 0.46 1.10 300 * 8.70 * 19.6 28.7 88.0 296 * Not determined.ANALYST, AUGUST 1986, VOL. 111 This might be expected to be a more satisfactory method but, as the molybdate also catalyses the oxidation of iodide to iodine, it is necessary to de-oxygenate sample and reagent solutions when using this catalysed method.’ Currents obtained for the construction of a calibration graph are given in Table 6. These show good rectilinearity and coefficients of variation (five determinations) were typically less than 1%. Discussion Methods are given for the on-line iodimetric determination of hypochlorite and hydrogen peroxide using flow injection analysis, with amperometric detection by means of a glassy carbon electrode held at -0.2 V vs.SCE to monitor the iodine formed. These illustrate the possibilities of carrying out iodimetric determinations on-line in FIA systems. Compari- son of the signals with those obtained by the direct injection of standard iodine solutions under the same conditions indicate that complete reaction of hypochlorite with iodide occurs on-line. Hence maximum sensitivity and reliability is assured. Highly reproducible signals are also obtained for the on-line determination of hydrogen peroxide, but the reaction of hydrogen peroxide and iodide is incomplete. In using this latter method, it is important to ensure that other constituents of the sample are not catalysing the reaction. The possibility of determining by FIA the iodine formed by the complete reaction of iodide and hydrogen peroxide off-line has been demonstrated. A major disadvantage of this off-line approach is that the oxidation by dissolved molecular oxygen of iodide to iodine before injection is readily catalysed by both acidic solutions and light. The use of fully on-line reactions largely overcomes this difficulty and is also more convenient. The work described in this paper is being extended to the development of iodimetric methods in which reproducible amounts of iodine are produced on-line by the acidification of iodate - iodide solution and in which the amount of iodine required to oxidise the determinand is monitored. A.Y.C. thanks the Lebanese University for leave of absence and the Lebanese Government for financial support. References 1. 2. 3. 4. 5. 6. 7. Fogg, A. G., Ali, M. A . , and Abdalla, M. A., Analyst, 1983, 108, 840. Fogg, A. G., Chamsi, A. Y . , Barros, A. A . , and Cabral, J. O., Analyst, 1984, 109, 901. Lee, C. C., and Pollard, B. D., Anal. Chim. Acta, 1984, 158, 157. Miller, K. G., Pacey, G. E., and Gordon, G., Anal. Chem., 1985, 57, 734. Fogg, A. G., and Summan, A. M., Analyst, 1984, 109, 1029. Lloyd, J. B. F., J. Chromatogr., 1983, 256, 323. Vogel, A. I., “A Textbook of Quantitative Inorganic Analy- sis,” Fourth Edition, Longmans, London, 1978, p. 381. Paper A51370 Received October 16th, 1985 Accepted March 3rd, 1986