Analyst, June, 1979, Vol. 104, pp. 531-537 531 Polarography of Green S F. E. Powell Department of Science and Food Technology, Grimsby College of Technology, Hurnberside, DN34 5BQ N u n s Corner, Grimsby, South The food dye Green S, 4- [4-dimethylammoniocyclohexa-2,5-dienylidene- (4-dimethylaminophenyl)methyl]-3-hydroxynaphthalene-2,7-disulphonic acid, monosodium salt, is reduced a t the dropping-mercury electrode from 50% ethanolic solutions with the total consumption of two electrons. Polarograms follow theoretical predictions in the pH range 2.7-8.75. The reduction mechanism involves two electron transfer steps that are sufficiently differenti- ated at higher pH for separate waves to appear. Keywords : Green S ; food dye ; polarograplay Carbonium ions of the triarylmethane type are reducible at the dropping-mercury electrode (D.M.E.), and the parent species, triphenylmethane, has been studied in various media.For example, in liquid sulphur dioxide, the cation undergoes single-electron reduction to the radical, which dimerisesl : .. . . - - (1) #&+ + e + . . .. .. .. - * (2) 2$3C' -+ dimer .. In protonating solvents, a second reduction step is possible : Thus, in methanesulphonic acid2 and sulphuric acid3 media, this step is in competition with the dimerisation. Dyes based on this structure have also been examined polarographically.4-7 The methane carbon atom is the electroactive centre of these molecules also and reduction occurs through the steps described above. Variety of behaviour is afforded by the nature of the attached aryl groups.Strongly electron-attracting aryl groups lower the electron density of the electroactive centre, which is then readily reduced in a two-electron wave. However, two distinct single-electron waves are observed with strongly donating aryl groups. The mechanism is also sensitive to acid - base eq~ilibria.~ Within this framework, a study has been made of the dye Green S (C.I. 44090), the mono- sodium salt of 4-[4-dimethylammoniocyclohexa-2,5-dieny~dene-(4-~methylaminophenyl)- methyl]-3-hydroxynaphthalene-2,7-disulphonic acid. As the dye is a permitted food colour, the information so obtained should provide a rational basis for the design of polarographic procedures for its analytical determination in foodstuffs. Experimental For polarography, 50% aqueous ethanol solvent was used to eliminate possible adsorption effects.4 The ethanol was of analytical-reagent grade and was free from impurity waves.Conventional McIlvaine and Clark - Lubs buffer components were used and the solutions were made up to a consistent ionic strength of 0.15 by the addition of potassium chloride when necessarya8 Solutions also contained 0.01 yo of methylcellulose as maximum suppressor ; protein surfactants were avoided because of the possibility of binding with the dyestuff .g Solutions were filtered and well de-aerated before polarography. A water-jacketed electrochemical cell was employed, consisting of a working D.M.E., auxiliary electrode of area 19.6 cm2 and a saturated calomel electrode (S.C.E.) as reference. Polarograms were recorded with a Heath EUW-401 polarograph used in conjunction with a Servoscribe RE51 1 recorder and EIL 38B external millivoltmeter.Data for logarithmic A commercial sample of Green S was recrystallised from ethanol in low yield.532 POWELL : POLAROGRAPHY Analyst, Vol. 104 analysis of the waves were obtained manually. Corrections of the faradaic currents for residual current and mercury column heads (h) for the back pressure were made. Plateau currents were compared at the h value of 61.!)cm, where the mercury flow-rate, m, was 1.62 mg s-l. Drop times (t) were determined indirectly from the current versus time traces displayed on a Telequipment S51 B oscilloscope, direct visual observation being impossible because the dye absorbs strongly in the visible region at 635nm.In all solutions t was 2.9-3.0 s in the potential range of interest. All measurements were made at 25 & 0.2 "C. Results and Discussion The dye is reduced in a single wave up to pH 7.3, where wave splitting becomes discernible. At pH 8.75 the division has developed to an extent that distinct waves can be recognised. Representative polarograms are shown in Fig. 1. Fig. 1. Representative polarograms. A, pH 4.2, E , = -0.600 V, depolariser concentration c = 1.17 x C, pH 7.3, E , = -0.754V, c = 1.03 x m; and D, pH 8.75, E+ (first wave) = -0.'700 V, E , (second wave) = 1 0 - 3 ~ ; B, PH 5.7, E , = -omov, c = 1.16 x 1 0 - 3 ~ ; -0.932 v, G = 1.13 X M. Limiting Currents Well defined plateau regions were observed on polarograms of the dye in each of the buffer solutions.A linear dependence of the mean limiting current (iL) on depolariser concentra- tion (C) at six values between 0.2 and 1.2 mol EL-^ was established in each instance. These lines passed through the origin and their respective gradients ( i L / C ) are given in Table I. Further, iL was found to be directly proportion,al to h* when the mercury head was varied between 32 and 62 cm (Table I). These results show that limiting currents are diffusion con trolled. TABLE I EXPERIMENTAL RESULTS ir./C/ Mean deviation I l El4 PH w-4 m3 mol-1 of iLh-f,% WA m* mol-1 mg-P s* n v vs. S.C.E. 2.7 2.04 0.4 1.23 1.9 -0.495 3.7 1.98 1.1 1.20 1.9 -0.582 4.2 1.94 1.1 ~. ~ 4.6 2.00 1.7 5.7 2.06 0.9 6.3 2.02 0.4 7.3 2.20 0.9 7.95 2.06 1.1 8.75 1.10 0.8 (1st wave) (total wave) 2.18 1.2 * 1st wave.f 2nd wave, 1.17 1.21 1.24 1.22 1.33 1.24 0.66 1.32 1.8 -0.600 1.9 -0.617 1.9 -0.690 1.9 -0.724 2.1 -0.754 1.9 -0.706* - 0.8741 1.0 -0.700* 2.0 -0.932t E F / dEt/dPH/ 0: Vvs. S.C.E. mV 0.35 +o.oin - 0.37 -0.291 -87 0.37 -0.350 -36 0.37 -0.410 -42.5 - - - - - - - - 0.874 - 1.00 - 0 0.76 -0.932 -73 P 2.02 0.99 - 0.85 - - - - - 0 0.94June, 1979 OF GREEN S 533 The diffusion current constant ( I ) , viz., iLICmW, of the total wave was found to be essentially uniform over the pH range studied (Table I), indicating a common stoicheiometry for the reduction process. According to the IlkoviE equation, this term can be identified with the product 607nD*, where n is the number of electrons involved in the reduction process. The adoption of a diffusion coefficient (D) value of 1.14 x 10- cm2 s-l, as deter- mined for the related dye Brilliant Green in 50% ethanol by a porous diaphragm m e t h ~ d , ~ produces the n values shown in Table I.Evidently, the complete reduction is a two-electron process, but at pH 8.75 the reduction takes place in two single-electron steps, each of which is diffusion controlled. Attempts to extend the range to higher pH values resulted in colourless solutions, probably owing to the formation of the electroinactive carbinol.6 Experimental Wave Forms At pH 8.75 the difference between half-wave potentials (E,) of the split waves exceeds 200 mV, permitting separate logarithmic analysis (Fig. 2) ; from the slopes of these plots it appears that the first wave corresponds to a reversible single-electron transfer but that the second electron is taken up irreversibly in the subsequent wave with a transfer coefficient (a) of 0.76.Comparison of the E* values at pH 8.75 and pH 7.95 (Table I) shows that the second but not the first wave is pH dependent. The number of protons (9) involved in the rate-determining step of the more negative irreversible process evaluated fromlO suggests the uptake of a single proton (Table I). Fig. 2. at pH 8.75. and ( b ) , second wave slope 79 mV. Logarithmic analysis of waves (a), First wave slope 59 mV; Hence, the first wave can be ascribed to process (1) and the second to process (3). Dimerisa- tion of the free radical does not take place, as the difference between E, for the separate waves does not change with depolariser concentration.ll On decreasing the pH, the half-wave potential of the irreversible wave is displaced to more positive potentials, overtaking that of the reversible wave, resulting in a merger.The appropriate logarithmic analyses are shown in Figs. 3 and 5 and the data are given in Table I.534 POWELL : POLAROGRAPHY Analyst, Vol. 104 Theoretical Wave Forms Mizutani et aZ.12 have presented a theoretical treatment of the reduction scheme below for an expanding plane electrode : O + e + Z 2 + e + R (heterogeneous rate constant = k ) This treatment can be readily adapted to the proposed mechanism for Green S with the recognition that here k is a pseudo-first-order rate constant dependent on the hydrogen-ion concentration : k = k, [H-t]' . . .... .. * * (5) p protons being involved in the transition state of the rate-determining step and k , is the true heterogeneous rate constant. Provided that the half-wave potential of the irreversible step, E r , is sufficiently more positive than that of the first, reversible step, E r , then a fused wave results, with i.e., the logarithmic plot is linear. pH 4.6. taken as 0.4 in this pH range. This situation arises in the reduction of Green S up to From the experimental slopes (Fig. 3), the a value of the irreversible step can be Further, the composite E t can be written as12 -1.0 -0.5 0 0.5 1.0 1.5 I L -- I Fig. 3. Logarithmic analysis of experimental . . Log10 7- waves in pH range 2.7-6.3. Hence, E$' values can be calculated in this pH range, utilising the experimental ments and taking EY as -0.700 V, as established at pH 8.75.measure- Differentiating equation (7) with respect to pH and utilising equation (4) givesJune, 1979 OF GREEN S 536 from which $ can be calculated (Table I). Of course, identical results are obtained by direct application of equation (4) to the calculated E r values of the second step. At pH 4.6 and 4.2 a single proton is involved in the rate-determining step of the irreversible process and the over-all reduction mechanism proposed in alkaline solution remains applicable. However, at pH 3.7 and below, a second proton is consumed, presumably in the protonation of an amine group, whose basicity in the intermediate free radical is enhanced over that in the parent carbonium ion because of a reduction in the positive charge density.Mizutani et al. derived the current - potential relation of the consecutive electron transfer (ee) mechanism in closed forrn [equation (19) in their paper12]. This relationship was employed to explore the evolution of the waveforms with pH on a theoretical basis, utilising the following parameter values as guided by the experimental results: D = 10-6cm2s-1, t = 3s, a = 0.4, E r = -0.30 V, Er = -0.70V and a total mean limiting current of 2 PA. The experimental curve found at pH 3.7 could be reconstructed theoretically with close agreement with the observed E+ and logarithmic slope, by assuming k = 6.5 x cm s-l (Fig. 4). I I I I -1.0 -0.5 0 0.5 1 .O - . Log,o - Fig. 4. Logarithmic analysis of theoretical waves in pH range 2.7-6.3.A, pH = 2.7, EF = 0 V, slope = 42 mV; B, pH = 3.7, Eim = -0.3 V, slope = 43 mV; C, pH = 4.3, Ef" = -0.4 V, slope = 42 mV; D, pH = 5.65, Eim = -0.6 V; and E, pH = 6.3, EP = -0.7 V. Theoretical curves at other acidic pH values were generated by duplicate calculations modifying only the parameter E r in accordance with equation (4) and taking p = 2 between pH 2.7 and 3.7 and j5 = 1 elsewhere. This variation in Ep contains implicitly the necessary change in k with pH, as the parameters are related bylo dE$' 59 d loglok (mvl - = - x - .. dpH o! dpH .. .. (9) If required, the inferred k at each pH can be obtained from the basis value at pH 3.7 using equation (5). The self-consistent current - potential curves so obtained are plotted in logarithmic form at 10-mV intervals in Fig.4. On comparison with the experimental data536 POWELL POLAROGRAPHY Analyst, Vol. 104 in Fig. 3, good agreement is observed between slopes and operational E+ values at low pH. Also, the development of a two-segment logarithmic plot with increasing pH is reproduced. In alkaline solutions, the larger a value foundl indicates facilitation of the transfer of the second electron. Consequently, in the theoretilcal calculation at pH 7.3, a has been given the value 0.7, and E f extended to -0.80V in accordance with equation (4) with a in transition from 0.4 to 0.7. These changes again permit a close fit between the theoretical and experimental waveforms (Fig. 5 ) . -1.0 -0.5 0 0.5 1 .o Fig. 5. Logarithmic analysis of wave at pH 7.3. Open circles denote experimental results, full line represents Mizutani et al.relation, with parameter values as in text. Conclusion The following reduction scheme forms the foundation of a consistent theoretical interpreta- tion of the polarographic behaviour of Green S in the pH range of interest:June, 1979 OF GREEN S 537 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Elving, P. J., and Markowitz, J. M., J . Phys. Chem., 1961, 65, 686. Wawzonek, S., Berkey, R., and Thomson, D., J . Electrochem. Soc., 1956, 103, 513. Plesch, P. H., and SestAkova, I., J . Chem. SOC. (B), 1970, 87. Kaye, R. 0.. and Stonehill, H. I., J . Ckern. Soc., 1952, 3231. Ramaiah, N. A., and Katiyar, S. S., Curr. Sci., 1961, 30, 175. Kemula, W., and Axt-Zak, A., Roczn. Chem., 1962, 36, 737; 1963, 37, 113. Bengtsson, G., Acta Chem. Scand., 1966, 20, 1176; 1967, 21, 1138 and 2544; 1968, 22, 1241; 1969, Elving, P. J., Markowitz, J. M., and Rosenthal, I., Analyt. Chem., 1956, 28, 1179. Gurr, F., “Synthetic Dyes in Biology, Medicine and Chemistry,” Academic Press, New York, 1971, Elving, P. J., Pure Appl. Chem., 1963, 7 , 423. Pozdeeva, A. A., and Zhandov, S. I., PYOG. 3rd Int. Polarogr. Congr., Southampton, 1964, 2, 781. Mizutani, F., Sato, N., and Sekine, T., Denki Kagaku, 1978, 46, 247. 23, 435, 448 and 455. p. 404. Received December 7th, 1978 Accepted January 19th, 1979