Development of Ce IV and Mn II Oscillometry: A New Estimation Method Based on Peak Potential and Oscillation Period of Chemical Oscillations

AU J.T. 7(2): 49-58 (Oct. 2003) Development of Ce IV and Mn II Oscillometry: A New Estimation Method Based on Peak Potential and Oscillation Period of Chemical Oscillations David T. Win, Ammie Ohn* and Myint Myint Si* Faculty of Science and Technology, Assumption University Bangkok, Thailand Abstract The linear relationship of oscillator concentration and oscillation characteristic (peak potential or oscillation period) in the bromate / oscillator / acetylacetone system of Belousov-Zhabotinsky chemical oscillations was examined and used as a basis for a new novel method for estimation of Ce IV and Mn II cation oscillators. The oscillation period (tos) based version was termed tos oscillometry and the peak potential (pp) based version was called pp oscillometry. Keywords: Cation estimation method, Ce IV and Mn II, oscillators, bromate, acetylacetone, chemical oscillation, Belousov - Zhabotinsky systems, peak potentials, oscillation characteristics, oscillation period. 1. Introduction The main objective of this work was to demonstrate the possibility of using peak potential (pp) and oscillation period (tos) values that accompany chemical oscillations for estimation of cations, such as Ce (IV) and Mn (II), which can act as oscillators. It is a novel alternative analytical estimation method. Belousov-Zhabotinsky oscillations are the most popular of the chemical oscillation systems. They consist of bromate, an organic component containing an active methylene group such as acetylacetone, and an oscillator which flip-flops or oscillate between the reduced state [example Ce (III)] and oxidized state [example Ce (IV)]. Ce (IV), Mn (II) and Fe (phen) 3 2+ (Ferroin) are some commonly used oscillators (Win and Win 1985). Continuous alternate oxidation and reduction, caused by the following two reactions occurring repeatedly in turn, one * Department of Chemistry, University of Yangon, Yangon, Myanmar after the other, result in oscillations, as shown in the rough diagram below. Oxidized Oscillator Reduced Oscillator Ce (IV) Ce (III) Mn (VII) Mn (II) Fe (phen) 3 3+ Fe (phen) 3 2+ (a) Bromate acting on reduced oscillator producing the oxidized oscillator (Oxidation). (b) Acetylacetone acting on oxidized oscillator producing the reduced oscillator (Reduction). The above rough diagram shows only the end states and serves to visualize the main part (the oscillation part) of chemical oscillations, and is therefore not a reaction mechanism scheme. The actual reaction mechanism is complex. The use of oscillation characteristics in cation estimation, rather than the study of the chemical oscillation process itself, is the goal of this study. Hence, elucidation of reaction mechanism is not contemplated in this study. Chemical oscillations are accompanied by potential changes, which may be followed

via an appropriate electrode system, such as potential variations are the oscillation traces(win et al. 2002a) The initial reaction at the start, before onset of oscillations, would be oxidation for reduced oscillators like Mn (II) and Fe 2+ (phen) 3 or reduction for oxidized oscillators like Ce (IV). These reactions drive the chemical system towards a condition where the oscillator potential was positioned within the balance-point region, a potential range between bromate and acetylacetone potentials (Than 2001). Oscillations start when this was achieved. Time needed for onset of oscillations called induction period (ip), the peak height called peak potential (pp), the time between two consecutive peaks called oscillation period (tos) and the time from start of oscillations to the finish of oscillations called the total oscillation time (tot), are the major oscillation characteristics. If these are linearly related to oscillator concentrations, then an analytical method for cation estimation may be established (Si 1992). Some cations such as Ce (IV) and Mn (II) can function directly as oscillators, whereas some cations can function only when complexed with ligands. For example, Fe (II) functions as an oscillator only when combined with 1,10-phenanthroline as 2+ ferroin complex, Fe(phen) 3. Oscillation characteristics may also be linearly related to bromate and acetylacetone concentrations. Thus, oscillometry may also be used for estimation of bromate and acetylacetone, or any organic compound with active methylene CH 2 groups (Ohn 1992). Continued interest in the B-Z system is seen in the recent publications where alternative electrodes were examined (Win, et al. 2002a); three-dimensional oscillation models were developed (Field and Gyoergi 1991, 1992); theoretical calculations were made (Martin 2001); a paper on non-linear chemical dynamics was published (Epstein and Pojman 2000) and Fe (II) Oscillometry was developed (Win, et al. 2002b). Moreover, the B-Z system had been used for demonstrating chemical oscillations SCE-Pt or SCE-Cu. The temporal electrode (Shakhashiri 1985; Field 1972; Pojman, et al. 1994). The three main chemical oscillation models are the Lotka-Volterra (Lotka 1920), Brusselator - proposed by Pregogine and his collaborators at the Free University of Brussels (Glandsdorff and Prigogine 1971) and Oregonator (Field and Noyes 1974) models that use sophisticated mathematics. For example, the Brusselator uses limited cycles. Review papers have summarized the findings (Noyes 1989). Chemical oscillations in nature are responsible for the ubiquitous appearance of periodicities in biological systems. Examples are the buzz in the circadia and its temperature dependence, circadian rhythms such as the sleep cycle, breathing, and the heart beat (Cervellati, et al. 1998). Examples of recent interest in B-Z system are apparent in some papers by Cervellati and co-workers (Cervellati, et al. 1998; Cervellati and Mongiorgi 1998; Cervellati, et al. 1999) and conferences such as Oscillations & Dynamic Instabilities In Chemical Systems held during 28 July 2 August 2002 at Queen s College, Oxford, UK (Queen s College 2002). The present work involves development of a method for the estimation of Ce (IV) and Mn (II), based on chemical oscillation characteristics. From the above review of published work, it is evident that use of chemical oscillation characteristics for analytical estimations has not been contemplated or attempted before. Thus a new word Oscillometry has to be coined*. Estimations using peak potentials (pp) are named pp oscillometry and those using oscillation periods (tos) are called tos oscillometry. Of the oscillation characteristics, peak potential (pp) and oscillation period (tos) are fairly reproducible (Win, et al. 2002b). Hence, they are chosen as parameters to be measured for use in the analytical estimation of Ce (IV) and Mn (II). Measurement of these parameters is well established, but employing the measured results for cation estimation is the novel aspect of this work. 50

* It must not be confused with Oscillimetry where electrical oscillations are used for analytical estimations. 2. Experimental The equipment, experimental set-up and procedure were as described previously (Win, et al. 2002a). The bromate / oscillator Ce (IV) or Mn (II) / acetylacetone system was examined. Potassium bromate, acetylacetone, cerium (IV) sulphate, manganese (II) sulphate, 1.5M sulphuric (VI) acid and 1,10- phenanthroline was of reagent grade (BDH) and was used without further purification. Standard solutions of the above reagents were prepared by dissolving accurately weighed amounts in exactly measured volumes of 1.5 M sulphuric (VI) acid, except for ferroin solutions which were prepared with deionized water. A Corning ph meter was used in conjunction with platinum Pt and standard calomel SCE electrodes. The Pt electrode was connected to the indicator terminal and SCE was connected to the reference terminal. A water-jacketed glass reactor cell was placed in a temperature-controlled thermostated water bath. The reactants were placed in separate glass vessels and placed in the above water bath and allowed to reach equilibrium temperature. A total of 20.00 ml potassium bromate solution was introduced into the reactor cell. Nitrogen gas was bubbled through at a fixed rate of 25 ml per minute. The SCE/Pt electrode pair was inserted. Another 20.00 ml of acetylacetone solution followed by 10.00 ml of oscillator solution (cerium (IV) sulphate or manganese (II) sulphate) was introduced. The stop clock was started as soon as the addition of oscillator solution was completed. The oscillations, expressed as temporal potential variations of the Pt indicator electrode in a SCE-Pt electrode pair, were fed to a computer and recorded by a data logging software called Tattletale, which allowed electrode potential sampling at desired set time intervals. The oscillation traces were then displayed via MS Excel. Oscillation periods (tos) and peak potentials (pp) at varied Ce (IV) concentrations were determined from the above oscillation traces. Then tos or pp versus [Ce (IV)] curves were drawn and checked for linearity. If non-linear curves were obtained, logarithm values were tried. The linear curves so obtained were taken as working curves. Standard test samples were prepared and their oscillation traces were obtained. Corresponding tos and pp were found, and the experimental concentrations were read off from the appropriate working curve. The prepared concentrations were taken as the true values. The experimentally determined concentrations (experimental concentrations) were results obtained from the experimental working curves, and would be different from the prepared values. The differences were the errors. Deviations between the prepared and experimental concentrations were found and relative errors were calculated to determine accuracy. To test for precision, the experiment was repeated three times using a particular unknown test sample. Statistical processing (mean, standard deviation, coefficient of variation CV and relative standard deviation RSD) was performed on the experimental results. The average concentration value was shown with its precision at 95% confidence level. The procedure was repeated with Mn (II) oscillator. The method was called oscillometry. 3. Results and Discussion The oscillation period (tos) based method was referred to as tos oscillometry, and the peak potential (pp) based method was called pp oscillometry. tos oscillometry consisted of the following sequence: (i) Setting up the working curve: tos measurement of standards, determining the linear range, drawing working curves [tos or log (tos) versus cation concentration] (ii) Determination of experimental concentrations: Preparation of test samples 51

with known concentrations, tos measurement, reading the concentration from the working curve. (iii) Determination of accuracy: Finding the relative error from the difference between the prepared and experimentally determined concentrations. (iv) Determination of precision: Triple repetition of tos measurement and reading the concentration from the working curve (using the same prepared sample) followed by statistical processing of the three experimentally determined results so obtained. pp oscillometry sequence was along the same lines, pp was used in place of tos. 3.1 The Oscillation Traces Fig. 1 was an example of an oscillation trace, using Mn (II) oscillator. Such traces were used to determine peak potentials and oscillation periods. Regions where oscillations were most stabilized were chosen for such determinations. The values were averaged. Identifying the stabilized region was most crucial as it could affect the precision of both tos and pp values. The identification was especially difficult if the oscillation trace was sloping, as in the sample trace shown above. Unlike Mn (II) oscillator, Ce (IV) oscillator gave oscillation traces that were generally stable and frequently horizontal. Thus, the stabilized region was easily identified. Peak potentials were stable, until towards the end when the oscillations died off, so the pp could be obtained accurately. However, the peak spacing varied and consequently tos were not very accurate. Some oscillation traces had interrupted oscillation series. For example, in the trace shown in Fig. 1, there was a silent gap between 310 s and 560 s, where oscillations had stopped and the electrode potential was constant around 960 mv. Then the oscillations carried on again. In such cases, the most stable group of oscillations should be chosen to determine the tos and pp values. 3.2 The Cerium (IV) Estimation 3.2.1 tos oscillometry: Varied concentrations of Ce (IV) ranging from 3.6 x 10-4 M to 1.82 x 10-3 M were used to obtain oscillation traces and the corresponding tos and pp were determined. The results were shown in Table 1. The tos versus Ce (IV) concentration relationship was non-linear. However, the log tos versus Ce (IV) concentration curve was linear within the range 4 x 10-4 and 9.6 x 10-4 M, as shown in Fig. 2. This linear region was used as a working curve. The linear region was quite narrow, but test solutions of roughly known concentrations could be diluted appropriately and estimated. Since the dilution factors were known, the original concentrations could be calculated accurately. To test for precision, the procedure was repeated three times by using a test sample with Ce (IV) concentration of 6.0 x 10-4 M. tos of 19 s, 19 s and 18 s were obtained. log tos values were calculated and concentrations were read off from the working curve, Fig. 2, giving experimental estimations of 6.3 x 10-4 M, 6.3 x 10-4 M and 6.6 x 10-4 M. The mean was 6.4 x 10-4 M and the standard deviation was 1.7 x 10-5 M. The coefficient of variation CV and relative standard deviation RSD were 2.7 % and 27 ppt respectively. The average result was therefore calculated as 6.4 ± 0.43 x 10-4 M at 95% confidence level. This value of 0.43 in 6.4 reflected a 6.7% precision at 95% confidence level (0.05% significance level). To test for accuracy, three test samples of Ammonium Cerium (IV) sulphate with concentrations of 6.0 x 10-4 M, 7.8 x 10-4 M and 9.7 x 10-4 M were prepared and used as standards or unknowns. Oscillation traces were obtained, tos were determined and log tos were calculated. These log tos values were used to read Ce (IV) concentrations from the working curve, Fig. 2, and accuracy was checked against the actual concentrations used in preparing the samples. The results were shown in Table 2. 52

The first sample had a tos of 19 s and gave an estimation of 6.3 x 10-4 M, a relative error of 5%. The second sample had a tos of 14 s and gave 7.5 x 10-4 M, a relative error of 4%. The third sample had a tos of 9.8 s and gave 9.3 x 10-4 M, a relative error of 4%. The average relative error was therefore 4.3%, reflecting an accuracy that was not the best. In summary, the low precision 6.7 % and the 4.3% accuracy indicated that tos Ce (IV) oscillometry was suitable only for technical estimations. 3.2.2 pp oscillometry: As seen in the working curve, Fig. 3, peak potentials were found to be linearly related to Ce (IV) concentrations. To test for precision, the procedure was repeated three times using a test sample with Ce (IV) concentration of 6.0 x 10-4 M, the same concentration as used for a similar test in the above tos oscillometry. Peak potentials of 22 mv, 21 mv and 20.5 mv were obtained, giving experimental results of 6.0 x 10-4 M, 5.7 x 10-4 M and 5.4 x 10-4 M, respectively. The mean was 5.7 x 10-4 M and the standard deviation was 3.0 x 10-5 M. The coefficient of variation CV and relative standard deviation RSD were 5.2 % and 52 ppt, respectively. The average was therefore calculated as 5.7 ± 0.76 x 10-4 M at 95% confidence level. The precision was thus about 13% at 95% confidence level (0.05% significance level). As before, three unknown samples of ammonium cerium (IV) sulphate with concentrations of 9.0 x 10-4 M, 9.9 x 10-4 M and 12.0 x 10-4 M were used to test for accuracy. As shown in table 2, the estimations were 9.1 x 10-4 M, 10.1 x 10-4 M and 12.1 x 10-4 M, with respective relative errors of 1%, 2% and 0.8%, gave an average relative error of 1.3%, indicating a high accuracy. In summary, the high accuracy of 1.3% showed that pp Ce (IV) oscillometry is suitable for routine analytical estimations of Ce (IV). However, the rather low precision of 13% at 95% confidence level, showed that extra care was needed in performing the determinations. The Student t-test was performed on the pp Ce (IV) oscillometry and tos Ce (IV) oscillometry results. The three experimental results from the pp oscillometry precision test (6.0 x 10-4 M, 5.7 x 10-4 M and 5.8 x 10-4 M) were used as base data and the average experimental value from the tos oscillometry precision test (6.4 x 10-4 M) was used as additional data. A t-value of 4.377 was obtained. The degree of freedom (DF) without the additional data was 2 and with the suspect data was 3. The combined degree of freedom was therefore 5. The critical t- value for the 0.05 significance level (95% confidence level) at DF = 5 was 2.571. The calculated value of 4.377, was 1.7 times over the critical value of 2.571. Thus, the tos result was statistically different from the pp results. The summarized findings for tos oscillometry and pp oscillometry : the standard deviations, CV and RSD were 1.7 x 10-5 M, 2.7 % and 27 ppt for Ce (IV) tos oscillometry, reflecting a precision of 6.7 %; and the corresponding values for Ce (IV) pp oscillometry were 3.0 x 10-5 M, 5.2 % and 52 ppt, reflecting a precision of 13%. The data therefore indicated that pp oscillometry was not as precise as tos oscillometry. In Section 3.1 it was observed that peak potentials could be obtained accurately whereas tos measurements could not be as accurate. Hence, pp oscillometry should have a better accuracy relative to tos oscillometry. This was substantiated by a smaller relative error of 1.3% for pp oscillometry as compared to 4.3% for tos oscillometry. 3.3 The Manganese II Estimation 3.3.1 pp oscillometry: Varied manganese (II) sulphate concentrations ranging from 5.6 x 10-3 M to 8.0 x 10-2 M were used to obtain oscillation traces and thence the corresponding tos and pp. The results were shown in table 3. The peak potential versus Mn (II) salt concentration curve, figure 4, being linear up to 4.8 x 10-2 M, was used as 53

the working curve, limited to the range 5.6 x 10-3 M to 4.8 x 10-2 M. Three samples of manganese (II) sulphate with concentrations of 2.0 x 10-2 M, 3.0 x 10-2 M and 4.5 x 10-2 M were prepared and used as unknowns. The estimated concentrations, read from the working curve, Fig. 4, were 1.9 x 10-2 M, 3.1 x 10-2 M and 4.7 x 10-2 M, respectively, and are shown in Table 2. The respective relative errors were 4%, 3% and 4% gave an average error of 3.6% and reflected an accuracy that was unable to match the accuracy (1.3%) of pp Ce (IV) oscillometry. As before, to test for precision the procedure was repeated three times, using a constant manganese (II) sulphate concentration of 2.0 x 10-2 M. Peak potentials of 16 mv, 18 mv and 17 mv were obtained, giving estimated concentrations of 1.6 x 10-2 M, 2.0 x 10-2 M and 1.8 x 10-2 M, respectively, as read from Fig. 4. The mean was 1.8 x 10-2 M and the standard deviation was 2.0 x 10-3 M. The coefficient of variation CV and relative standard deviation RSD were 11% and 111 ppt, respectively. The average was therefore calculated as 1.8 ± 0.5 x 10-2 M at 95% confidence level, a low precision of 28%. In summary, the low precision of 28% and an accuracy of 3.6% showed that Mn (II) oscillometry could be used for rough technical estimations only. 3.4 Comments on Non-Linearity of Some pp (or tos) versus [Oscillator] Curves As observed above, some oscillation characteristics versus oscillator concentration curves were non-linear. It was thought that the keto-enol tautomerism of acetylacetone might be the cause. Of the two tautomers only the keto form had a methylene group and could participate in the oscillation mechanism. An equilibrium mixture consisted of 15% keto and 85% enol. Of the two steps, oxidation and reduction, which occurred alternately in chemical oscillations, acetylacetone would take part in the reduction step of the oscillation (reaction with the oxidized form of the oscillator, example Ce (IV) and stop reacting at the oxidation step (reaction with the oxidized form of the oscillator, example Ce (IV). Thus, the depletion of the keto form would not be smooth it would be a stop and go process. Consequently, the keto feed from the keto-enol equilibrium would be nonuniform and lead to non-uniform oscillation characteristics and produce non-linearity. Another possible contribution to nonlinearity was the acetylacetone-oscillator cation complexation equilibrium. Enol Keto Ce (IV) / Ce (III) Bromate More and more enol form had to be converted into the keto form as oscillations progressed, and if the equilibrium could not feed keto fast enough, the keto form would be depleted and the oscillations would stop. Then equilibrium could build up the keto form to appropriate levels and restart the oscillations. This would account for the observed interrupted oscillations in Fig. 1. 4. Conclusions It was demonstrated that pp or tos measurements may be used for cation estimation. The developed method, called oscillometry, provides a means for estimating very low concentrations of Ce (IV) and Mn (II). The standard deviations, CV and RSD for pp oscillometry were 3.0 x 10-5 M, 5.2%, 52 ppt for Ce (IV) pp oscillometry, indicated a 13% precision. The accuracy as shown by the average relative error was 1.3%. Hence, pp oscillometry was suitable for routine Ce (IV) analytical estimations. The corresponding values for Ce (IV) tos oscillometry were 1.7 x 10-5 M, 2.7%, 27 ppt, reflecting a precision of 6.7%. The accuracy was 4.3%. This was less than that of the high 1.3% accuracy for Ce (IV) pp oscillometry. Thus Ce (IV) tos oscillometry could not match up to Ce (IV) pp oscillometry in accuracy, and it should be limited to technical estimation of Ce (IV). The better accuracy in pp oscillometr as compared to tos oscillometry for Ce (IV) 54

estimations was in accordance with the observation that peak potentials could be obtained accurately whereas tos could not be as accurate. The corresponding values for Mn (II) pp oscillometry were 2.0 x 10-3 M, 11% and 111 ppt, and gave a precision of 28%. This precision was unable to match the higher precisions, 13% of Ce (IV) pp oscillometry and 6.7% of Ce (IV) tos oscillometry. Also the accuracy shown by the average relative error, for Mn (II) pp oscillometry was 3.6%. This was comparable to that of Ce (IV) tos oscillometry, 4.3%, but three fold less than that of Ce (IV) pp oscillometry, 1.3%. Thus, Mn (II) pp oscillometry was suitable only for rough technical estimation of Mn (II). Acknowledgement The authors are grateful to Assumption University, Bangkok, Thailand and Yangon University, Yangon, Myanmar, for giving them the opportunity to carry out this research. References Cervellati, R.; Fetto, P.; and Dalbagni, G.1998. Teaching non-linear kinetics in the lab. Education in Chemistry (London) 35: 50-2. Cervellati, R.; Benini, R.; and Fetto, P. 1998. Experimental and mechanistic study of the bromomalonic acid - bromate oscillating system catalyzed by [Fe(phen) 3 ] 2+. Int. J. Chem. Kinet. 30: 291-300. Cervellati, R.; and Mongiorgi, B. 1998. Inhibition of chemical oscillations by bromide ion in the Briggs-Rauscher Reaction. Int. J. Chem. Kinet. 30: 641-6. Cervellati, R.; Cavasino, F.P.; Lombardo, R.; and Turco Liveri M.L., 1999. Micellar effects on the kinetics of cerium (IV) oxidation and the cerium (IV)-catalyzed Belousov-Zhabotinsky Reaction with methyl-ethyl- or benzylmalonic acid. J. Phys. Chem. B. 103: 4285-91. Epstein, I. R.; and Pojman, J. A. 2000. An introduction to nonlinear chemical dynamics: Oscillations, waves, patterns, and chaos. J. Chem. Educ. 77: 450-7. Field, R.J. 1972. A reaction periodic in time and space. J. Chem. Ed. 49: 308-13. Field, R.J.; and Noyes, R.M. 1974. The Oregonator model. J. Chem. Phys. 60:1877-92. Field, R.J.; and Gyoergi, L. 1991. Simple models of deterministic chaos in Belousov - Zhabotinsky reaction. J. Phys. Chem. 95: 6594-602. Field, R.J.; and Gyoergi, L. 1992. A threevariable model of deterministic chaos in the Belousov-Zhabotinsky reaction. Letters to Nature 355: 808-10. Glandsdorff, P.; and Prigogine, I. 1971. Thermodynamic Theory of Structure, Stability and Fluctuation. Wiley, Englewood Cliff, NJ, USA. Lotka, A.J. 1920. Chemical oscillation model. J. Amer. Chem. Soc. 42: 1595-1602. Martin S.T. 2001. Employing complex kinetic diagrams. Math. J. 8: 1-20. Noyes, R.M.1989. Chemical oscillations A review. J. Chem. Ed. 66: 190-9. Ohn, A. 1992. BZ oscillimetry - oxidant estimation. M.Sc. thesis, Yangon Univ., Yangon, Myanmar. Pojman, J.A.; Craven, R.; and Leard, D.C. 1994. Chemical oscillations. J. Chem. Ed. 71: 84-7. Queen s College. 2002. http:// www.grc.uri.edu/programs/2002/osc.htm Shakhashiri, B.Z. 1985. Chemical Demonstrations: A Handbook for Teachers. Univ. of Wisconsin, Madison, WI, USA. Si, M.M. 1992. BZ oscillimetry - oscillator estimation. M.Sc. thesis, Yangon Univ., Yangon, Myanmar. Than, M.M.; Aye, M.M.; and. Win, D.T. 2001. Concentration limits for chemical oscillations. AU J.T. 5: 9-12. Win, D. T.; and Win S. 1985. Chemical Oscillations. CRT Project Report I. Univ. of New South Wales, Sydney, Australia. Win, D.T.; Than, M.M.; and Aye, M.M. 2002a. An alternative electrode system for monitoring Belousov-Zhabotinsky chemical oscillations. AU J.T. 5: 124-8. 55

Win, D.T.; Ohn, A.; and Si, M.M. 2002b. Development of Fe (II) PP oscillometry: A technical estimation method based on peak potentials of chemical oscillations. AU J.T. 6: 1-4. Table 1. Oscillation characteristics at varied cerium IV concentrations [Ce (IV)] x 10000 tos ± 4/ s log tos pp ± 3 / mv 3.6 31 1.49 18 4.8 25 1.4 20 6 20 1.3 22 7.2 15 1.18 24 8.4 11 1.05 26 9.6 7 0.95 28 11.1 7 0.93 31 12 9 0.95 35 18.2 11 1.04 36 Table 2. Accuracy of estimated concentrations Oscillometry type M x 10000 prepared M x 10000 estimated Accuracy [relative error] / % Ce IV tos 6.0 6.2 3 Ce IV tos 7.8 8.2 5 Ce IV tos 9.7 9.9 2 Ce IV pp 9.0 9.1 1 Ce IV pp 9.9 10.1 2 Ce IV pp 12.0 12.1 0.8 Mn II pp 200 190 4 Mn II pp 300 310 3.3 Mn II pp 450 470 4.4 Table 3. Oscillation characteristics at varied Mn (II) concentrations [Mn II] x 1000 tos ± 6 / s pp ± 7 / mv 5.6 10 3 16 12 16 32 13 25 48 22 32 64 25 33 80 23 32 56

Oscillation Trace: [Mn II] = 0.096 M 1060 1040 emf/mv 1020 1000 980 960 0 30 60 90 120 150 180 210 240 270 300 400 570 600 630 660 690 Time/s emf/mv Fig. 1. A sample oscillation trace log tos 1.6 1.5 1.4 1.3 1.2 1.1 1 0.9 0.8 log tos versus [Ce (IV)] 3.6 4.8 6 7.2 8.4 9.6 11.1 12 18.2 [Ce IV] / M x 10000 log tos Fig. 2. log tos versus [Ce (IV)] relationship 57

Peak Potential versus [Ce (IV)] pp / mv 45 40 35 30 25 20 15 10 5 0 3.6 4.8 6 7.2 8.4 9.6 11.1 12 18.2 [Ce IV] / M x 10000 pp? 3 / mv Fig. 3. Peak potential versus [Ce (IV)] relationship Peak Potential versus [Mn (II)] 50 40 pp / mv 30 20 10 0 5.6-10 16 32 48 64 80 [Mn II] / M x 1000 pp? 7 / mv Fig. 4. Peak Potential versus [MnSO 4 ] relationship 58