AUGUST 1979 The Analyst Vol. 104 No. 1241 Electrochemical Studies of Strongly Chelating Anthraquinone Derivatives G. Ali Qureshi," G. Svehla and M. A. Leonard Department of Analytical Chemistvy, The Queen's University, Belfast, Northern Ireland, BT9 6AG Analytically important, strongly chelating anthraquinones and their deriva- tives were studied by d.c. polarography, cyclic voltammetry and micro- coulometry to investigate their redox characteristics. All 18 substances were reduced in a two-electron reversible or quasi-reversible process in both aqueous and 75% ethanolic solutions. Depending on pH and the medium, single or double polarographic reduction waves appeared, which were diffusion controlled, although in some instances adsorption pre-waves were also observed. This behaviour is similar to the known behaviour of simpler quinone systems.The variation of the IlkoviC coefficient and half-wave potential with pH was studied in detail to investigate the acid - base behaviour of the species involved in the reduction process. As a result, it was possible to describe the reduction mechanism of the anthraquinones involved. A number of new pK values were determined and others confirmed. Attempts to find linear free energy relationships were generally unsuccessful. Keywords : A nthraquinone derivatives ; polarography ; cyclic voltammetry ; half-wave potentials; PK values Strongly chelating anthraquinones are selective and sensitive reagents suitable for the spectrophotometric determination of a number of metallic and non-metallic substances. Among them, alizarin fluorine blue [3-NN-di (carboxymethyl)aminomethyl- 1,2-dihydroxy- anthraquinone] was introduced first by Belcher et aZ.l for the determination of fluoride.Leonard and co-workers2-11 have synthesised and studied a number of new reagents, all related to alizarin fluorine blue, and applied them for various analytical purposes. Because of their quinoidal structure, these substances display interesting electrochemical behaviour, which we undertook to study by polarography and cyclic voltammetry. Some anthra- quinone derivatives (including alizarin) have been studied polarographically in the past ,12--19 but these did not involve most of the compounds studied by ourselves. In our study we tried to avoid buffers with complex-forming characters as far as possible, in order to be able to compare the polarographic behaviour of these anthraquinones with that of their com- plexes.Experimental Reagents Some of them (I-XI and XIV) were commercially available; others (XII, XIII, XV and XVI) were prepared by Mannich condensation of alizarin with the appropriate agent, and XVII and XVIII were prepared earlier by A1 Ani20 in this Department. The purity of each compound was checked by electrophoresis, thin-layer chromatography and by elemental analysis. Compounds which, when analysed, gave more than 3% divergences from the theoretical carbon or hydrogen values were rejected. For the study 5 x lO-*molI-l aqueous solutions were prepared of each compound; sometimes a few drops of 0.1 mol 1-1 sodium hydroxide solution were added to assist rapid dissolution.* Present address: Department of Analytical Chemistry, University of Uppsala, S-75121 Uppsala 1, Sweden. 705 The compounds involved in this study are listed in Table I.706 were insoluble in water at certain pH values. used; their composition will be mentioned in the appropriate section of the text. from analytical-reagen t grade reagents. QURESHI et al. : ELECTROCHEMICAL STUDIES OF Analyst, Vol. 104 Studies were also made in solutions containing 75% V/V ethanol, as some anthraquinones In some instances special solutions were All buffer solutions, supporting electrolytes and maximum suppressors were prepared White-spot nitrogen gas was used for deaeration. TABLE I COMPOUNDS INCLUDED IN THE STUDY No. I I1 111 IV V VI VII VIII IX X XI XI1 XI11 XIV xv XVI XVIl Name of compound Substituents 1-Hydroxyanthraquinone R, == OH 2-H ydrox yanthraquinone R, == OH Alizarin R,, K, = OH 2,3-Dihydroxyanthraquinone R,, K3 = OH Alizarin-3-sulphonic acid Alizarin-5-sulphonic acid 3-Nitroalizarin (alizarin orange) 1,2,3-Trihydroxyanthraquinone 1,2,4-Trihydroxyanthraquinone 1,2,7-Trihydroxyanthraquinone 1,2,5,8-Tetrahydroxyanthraquinone (alizarin red) R,, K, = OH; R, = SOaH R,, R, = OH; R, = S03H R,, SR, = OH; R, = NO, R,, R,, R3 = OH R,, R2, R4 = OH R,, R,, R, = OH (anthragallol) (purpurin) !quinalizarin) Rip R,, R,, R, = OH R,, R, = OH; R3 = CH,-NH-CH,COOH Alizarin-3-methylglycine Alizarin-3-methylsarcozine R,, R, = OH; Ra = CH2-N<~~~coOH Alizarin fluorine blue R,, R, = OH; R3 = CH,-N(CH,COOH), R,, R, = OH; R3 = CH,-N(CH,COOH),; R5 = S03H Alizarin fluorine blue 5-sulphonic Alizarin-3-methyldiethylamine 1,2-Dihydroxy-3-( 2,4-dihydroxy- acid R,, R, = OH; R3 = CH,-N(C,H5), benzeneazo) anthraquinone R,, R, == OH; R3 = -N=N HO XVIII 1 -H ydrox y-2- (2,4-dihydroxy- benzeneaz0)anthraquinone R, = OH; R, = -N=N Apparatus All polarographic studies were carried out with the Radiometer PO4 polarograph, equipped with the E65 dropping-mercury electrode (IIME) in conjunction with the drop life timer and current sampler unit DLT1.Cyclic voltammetric measurements were made with an instrument, built in this Depart- ment, consisting of a Hewlett-Packard 3310B function generator, 1201B 100-pV dual-trace oscilloscope and a home-made potentiometric circuit with IR-drop compensation.A Radiometer P985B hanging-mercury electrode served as the working electrode, a platinum foil was used as the indicator electrode and a saturated calomel electrode as the common reference electrode. The three electrodes were mounted in such a way that the referenceA ugust ) 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 707 electrode was shielded from the IR drop between the other two electrodes. In some instances the hanging-mercury electrode was replaced with an amalgamated platinum-pin electrode. Results were evaluated on the spot using the oscilloscope in its storing mode, but whenever the sweep rates allowed, voltammograms were also recorded with a Hewlett- Packard 7035B X - Y recorder. For microcoulometric experiments a Manousek cell was used in conjunction with the PO4 polarograph. Current veys‘szcs time curves (at constant potentials) were obtained with an instrument designed and built in the Department.The stabilised potential source had a tolerance of &l mV and the current veysus time curves were displayed on an oscilloscope and photo- graphed. The currents were measurable with a relative standard deviation of less than 1%) and the system had a response time of a few milliseconds. pH measurements were carried out with an EIL 7030 pH meter with its glass and saturated calomel electrodes. Procedures All polarographic measurements were carried out with a drop lifetime of 2 s with 70% blanking time and a 50-cm mercury column height (corrected for back-pressure), unless otherwise stated.The anthraquinones were mixed with the appropriate buffer and a few drops of 0.1% Triton X solution and water to yield a 2.5 x mol 1-1 concentration in the cell (unless otherwise stated). Solutions were deaerated with a stream of nitrogen for 10 min. Polarograms were recorded with the slowest possible speed between 0 and -1.8 V. The polarograms were evaluated by the “point method”21 for both the half-wave potential and limiting current. The pH of each solution was measured after the polarographic experiment. The cyclic voltammetric experiments were carried out with electrolytes similar to those used for polarography. Sweep rates were between 10 and 1000 mV s-l, with current sensitivities adjusted appropriately. With a 10 mV s-l sweep rate a full-scale sensitivity of 1 pA was usually sufficient.Microcoulometric experiments were carried out on 2-ml solutions, which contained initially 5 x moll-1 of the alizarin together with the supporting electrolyte. The solution was placed in the Manousek cell and the capillary was positioned in such a way that its end was just below the surface of the solution. The deaerated solution was kept under a nitrogen atmosphere and was mixed after each 5 min by passage of a burst of nitrogen. A polaro- gram was then recorded every 30min. The electrolysis was carried out at a constant potential corresponding to a well defined limiting diffusion current. After 20-25y0 con- version the electrolysis was discontinued and the solution replaced with the pure supporting electrolyte to measure the residual current, which was then used as a correction when evaluating the results.This was done both on the basis of Faraday’s law directly, and with Gilbert and Rideal’s logarithmic method. 22 Current versus time curves were measured only if adsorption phenomena were suspected. In such instances the electrocapillary curve was first obtained by counting the number of drops falling naturally from a capillary per minute, at various electrode potentials. From the electrocapillary curve appropriate potential values were then selected, covering both distorted and undistorted regions, at which the current versus time curves were then measured. These curves were evaluated according to the guidelines given by Heyrovsky and K ~ t a . ~ ~ As the main task was to elucidate the electrochemical properties of strongly chelating anthraquinones, we studied compounds XII-XVIII in more detail than the others, which were examined first for the sake of comparison, mainly to investigate substitution effects. Our report will therefore be concentrated on the former group of substances. Results Polarography each compound.solution, single, double or triple waves were obtained for all the substances involved. The polarographic behaviour was investigated within the widest possible pH range for Depending on the medium and the pH of the medium and the pH of the The708 QURESHI et al. : ELECTROCHEMICAL STUDIES OF Autalyst, Vol. 104 single waves were proved to be diffusion controlled. The double waves consisted either of two diffusion-controlled waves or of an adsorption pre-wave followed by a diffusion wave.The two diffusion-controlled waves usually appeared at higher pH values, and they were well separated (see Figs. 1 and 2). The adsorption pre-wave appeared at lower pH values and usually disappeared, or became less pronounced as the pH was increased, although in some instances it reappeared again at high pH. values. This adsorption pre-wave always appeared close to the main diffusion wave. The triple waves always consisted of an adsorption pre-wave followed by two diff usion-controlled waves. Triple waves appeared only at high pH values (see Fig. 3). a 3 ov ov ov ov -0.2 v -0.2 v (pH 5.0) (6.0) (7.0) (8.0) (10.1) (12.1) Fig. 1. Polarograms of 2.6 x mol 1-l IX at various pH values in aqueous medium.The polarograms of VIII in Fig. 3 display such waves at and above pH 4.6; in more acidic solutions the two waves consist of an adsorption pre-wave and a diffusion-controlled wave. The over-all heights of the waves were found to be independent of the pH for most o the compounds. Fig. 4 shows the variation of the Ilkovie K coefficients (in amps per mole) for these anthraquinones In certain instances (e.g., I, 111, VI, XI11 and XVI) there was a significant drop in th coefficient as the pH of the solution increased and in other instances (V, XIII, XIV an1 again XVI) significant minima were observed on the K venus pH curve, which cannot b attributed to inconsistencies in the evaluation of the polarograms. The K coefficient depended, as expected, mainly on the medium-in aqueous solutions they averaged betwee 1 and 4 x 10-3 A mol-l and in 75% ethanol between 4 and 7 x 10-3 A mol-1.In some instances, however, significant variations were observed. 0.5 9 2 0 t t t + m - t t t t t -0.1 v-0.1 v-0.1 v -0.3V -0.4V -0.45V -0.5V -0.6V (pH 1.5) (2.2) (3.1) (4.3) (5.6) (7.0) (9.1) (11.1) Fig. 2. medium. Polarograms of 2.6 x lo-* mol 1-l XV at various pH values in aqueousAugust, 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 709 9 1.0 2 1 t t t t t t t +'0.2v.- ov ov ov ov ov ov ov Fig. 3. Polarograms of 3 x mol 1-1 VIII at various pH values in 76% ethanol. n XIV(W) 3.5 x 10-3 ____t I I I I I I I 0 2 4 6 8 10 12 PH I Fig. 4. Variation of the Ilkovit K coefficient with pH for selected anthraquinone derivatives in aqueous (W) and 75% ethanol (alc) solutions.710 QURESHI et al.: ELECTROCHEMICAL STUDIES OF . Analyst, VoZ, 104 The variation of half-wave potentials with pH was investigated with the greatest care, as these relationships provide valuable information on the acid - base equilibria in which the oxidised and reduced species are involved and, in an indirect way, also on the reduction process itself. Figs. 5 and 6 refer to aqueous solutions and Fig. 7 to alcoholic solutions. From the theoretical point of view the half-wave potentials themselves are not so important, but rather the slopes of the Ei versus pH plots, as well as those pH values where these slopes change. The expected values for these slopes are simple multiples of the Nernstian pre- logarithmic constant of 60 mV pH-l with factors of 0, 0.5, 1 and 1.5, that is 0, 30, 60 and 90 mV pH-l.Hence, the lines fitting the experimental points on these figures were drawn deliberately with such slopes-as is clear from the figures themselves, such lines do fit the experimental points reasonably well, although in some instances least-squares fits would produce slopes of different values. In the single instance of XI1 the experimental points could not be joined with a continuous set of lines. I 3 5 7 9 11 13 PH Fig. 5. Variation of the half-wave potentials of compounds 11-IX with pH in aqueous medium. The nature of diffusion-controlled waves has been verified by measurements taken at various mercury column heights and plotting the currents as the function of the square root of the (corrected) height.Such an operation was undertaken for each compound at several pH values, covering all sections of the E, versus pH curves. In addition to the linearity of these plots, we had various indirect evidence for diffusion control; thus, the absence ofAugust, 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 711 irregularities on the electrocapillary curves and the shape of the instantaneous current vwms time curves indicated clearly that diffusion currents were involved. Finally, waves were accepted to be diffusion controlled only when the current vucysza concentration plots proved the expected proportionality. xv -0.95 V ( X ) -0.9 v (XI) -0.9 v (XI I ) -0.7 V (XIII) -0.8 V ( X W -0.8 V (XV) - - c-- c-- - -0.85 V (XVI) t-- -0.9 v (XVII) - -0.9 v (XVIII) - 1 3 5 7 9 11 13 PH Fig.6. Variation of the half-wave potentials of compounds X-XVIII with pH in aqueous medium. The adsorption waves were investigated by several methods. Firstly, polarograms were obtained at different concentrations. Plotting the wave height against concentration a curve typical for adsorption currents was obtained (Fig. 8 ) . Temperature effects were studied. As can be seen from Fig. 9, the over-all wave height increased with temperature, but the adsorption pre-wave disappeared gradually as the temperature was increased. Instantaneous current veysus time curves were also studied whenever an adsorption wave was suspected. The oscillograms obtained at various potentials of the DME always showed distortions near to the half-wave potential of the adsorption wave (Fig.10). Finally, electrocapillary curves were obtained in the presence of the depolariser. Plotting the average drop lifetime as a function of the potential of the DME, characteristic incisions occurred at the appropriate potentials, whereas in the absence of the depolariser the curve was smooth at these potentials (Fig. 11). The cyclic voltammetric response of those substances which displayed adsorption waves was also interesting. In the first cycle, distortions were observed approximately at the expected potentials, but these disappeared when a second cycle was imposed immediately712 QURESHI et al. : ELECTROCHEMICAL STUDIES OF Analyst, Vol. 104 Fig. 7. Variation of the half-wave potentials of selected anthraquinones with pH in 76% ethanol.I I 0 1 2 c x 104/moi I-' Fig. 8. Variation of the height of the total wave (id) and of the adsorption pre-wave (ia) with con- centration of XV at pH 4.30 in aqueous medium.A ugust, 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 713 after the first. These adsorption waves displayed irreversible or, at very low scan rates, quasi-irreversible behaviour, to some extent contrary to expectation^.^^ This unusual behaviour would merit a more detailed investigation and electrochemists with more theoretical outlooks than that of the present authors might find this a rewarding area of research. t t t t t t t -0.56 V +H- 0.1 v Temperature dependence of the shape and height of the polarographic wave of a 2.6 x lo-* moll-' solution of XV at pH 6.6 in aqueous medium.Fig. 9. Adsorption pre-waves were noted in connection with compounds 111, VI, VII, IX, X, XI, XII, XIII, XIV, XV, XVI and XVIII. As indicated before, their appearance is dependent on the experimental conditions, of which pH, concentration and temperature are the most important. Further details on these adsorption pre-waves are available in Qureshi's thesis.25 All purely diffusion-controlled, single or double waves were subjected to logarithmic analysis in order to calculate the value of the pre-logarithmic factor in the Nernst equation. To our surprise, but in good agreement with results obtained on simpler quinone - hydro- quinone systems,26 the number of electrons (for a coefficient u = 1) was almost always found to be nearly unity, even with the single waves, where a value of 2 would be expected.This behaviour will be dealt with under Discussion. Microcoulometry To obtain a reliable value for the number of electrons taken up by a single molecule of the anthraquinones, we decided to carry out microcoulometric experiments. As the time H -0.1v -0.2 v- -0.3 V 4s ww w -0.35 V -0.4 V -0.5 V Fig. 10. Instantaneous current versus time curves of 2 x 10-4 moll-' solutions of XV at pH 4.30 in aqueous medium at different D.M.E. potentials.7 14 QURESHI et at. : ELECTROCHEMICAL STUDIES OF Analyst, ~ O L ? . 104 0 -0.5 -1.0 -1.5 EIV Fig. 11. Electrocapillary curve of mercury in a 10-l mol 1-1 acetate buffer (pH 4.30) in (A) the absence and (B) the presence of 2 x 10-4 moll-’ XV. duration of a single experiment is considerable, we omitted compounds VII and IX, carried out the measurement with most other compounds in neutral medium only, and only with the four key compounds, XIII-XVI, did we carry out three runs at three different pH values, involving acidic, neutral and basic media.With the three compounds displaying two separate reduction waves (VIII, XVII and XVIII), the pH was adjusted to 9, where the two waves are well established. A separate run was carried out for each wave. All results were evaluated both by graphical integration of the current versus time curves directly through Faraday’s law, and also by using Gilbert and Rideal’s equation.22 The two methods gave almost exactly the same results for the number of electrons: the difference between the corresponding figures was less than 0.5% in all instances.We feel that by evaluating results in both ways the figures are more reliable. The method based directly on Faraday’s law eliminates the apparent weakness of the method of Gilbert and Ridea1,22 which takes into account only the initial and final current, but disregards the over-all shape of the current verszts time curve. All of the compounds with single waves gave values between 1.97 and 2.30 for the number of electrons involved. The values for the double waves were in all instances higher than 1 (for each wave), but never exceeded 1.42. Fig. 12 shows the results obtained for XV. Through the points (which, theoretically, should I 0 120 240 flmin Fig. 12. Current veysus time curve obtained during the microcoulometric reduction of 2 ml of 5 x 10-4 M XV solution.August, 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 715 follow an exponential function) a straight line was drawn and the graph was integrated as a trapezium.From this integration, a value of 2.100 is obtainable for the number of electrons, while Gilbert and Rideal’s method gives 2.105. Cyclic Voltammetry The purpose of the cyclic voltammetric measurements was to find out whether the reduc- tion of these compounds is reversible or not. Up to a sweep rate of 10mVs-1 (which is 2-10 times higher than those used in the polarographic measurements) the reduction was either reversible or quasi-reversible, whereas, as pointed out, the adsorption was quasi- reversible at very low scan rates, or irreversible. In Fig.13 three cyclic voltammograms are reproduced that show these different responses. All of these curves were obtained with a 10 mV s-l sweep rate in buffered aqueous solutions at pH 10. Fig. 13(a) shows the cyclic voltammogram of 2.5 x lo4 mol 1-1 of XIV. The cathodic (positive) and anodic (negative) peak currents are virtually identical; the peak separation - is 33 mV, not far from the theoretical value of 30mV. Hence, one can say that the reduction of XIV is reversible under such conditions. A 10“ mol 1-1 solution of VI [Fig. 13(b)] displays quasi- reversible reduction, the cathodic and anodic peak currents differing by more than 10% and the peak separation being more than double the theoretical value. The adsorption peak, obtained in the first cathodic half-cycle, is barely distinguishable as a faint shoulder on the anodic half-cycle.The adsorption peak, obtained in the first cathodic half-cycle in a 3 x 10-4mo11-1 solution of VIII [Fig. 13(c)], does not show up on the first anodic half- cycle, indicating that the adsorption, under these circumstances, is irreversible. However, the two reduction peaks are reproduced with much reduced definition on the anodic half- cycle, so the reduction of VIII is still a quasi-reversible process. Reduction peak potentials [ED(c1], contrasted with the polarographic half-wave potentials (E*), together with a note on reversibility, are shown in Table 11. t o . l 4 2 0 -0.1 -0.6 -0.7 -0.8 +0.4 0 -0.4 ‘%,d(a) I I I -0.5 -0.7 -0.9 E N to. 1 0 -0.1 -0.3 -0.5 -0.7 -0.9 Fig. 13. Cyclic voltammograms of various anthraquinone derivatives.For explanation, see text. Discussion The purpose of this discussion is to explain the redox and acid - base processes in which the examined aiithraquinones are involved, and to correlate these with the experimental facts. Quinones (q), when reduced by taking up a single electron, form semiquinones (sq), while the uptake of a further electron results in the formation of hydroquinones (hq). In all of the redox processes, therefore, these three states of oxidation are expected to be involved.716 QURESHI et al. : ELECTROCHEMICAL STUDIES OF TABLE I1 CYCLIC VOLTAMMETRIC REDUCTION PEAK POTENTIALS r~,,,,] AND POLAROGRAPHIC Analyst, VoZ. 104 HALF-WAVE POTENTIALS (E,) AT P H 10 OF ANTHRAQUINONES Experimental conditions given in the text.E,(dIV E&/V f - t r \ Compound q/hq 4% sq/hq q/hq qlsq sq/hq I* -0.77 -- 0.74 I1 -0.78 -0.73 I11 -0.85 --0.82 I V -0.86 --0.81 V -0.78 -0.74 V I -0.71 --0.67 V I I -0.60 -0.87 -0.66 -0.83 V I I I -0.60 -0.76 -0.67 -0.72 IX -0.61 -0.81 -0.68 -0.78 X -0.87 -0.83 XI -0.84 --0.81 XI1 -0.82 --0.79 XI11 -0.81 --0.77 XIV -0.77 - 0.73 xv -0.78 - 0.74 X V I -0.79 -- 0.76 XVII -0.77 -0.86 -0.73 -0.83 XVIII -0.76 -0.85 -0.73 -0.81 * In 76% ethanolic solution. t R = reversible; Q = quasi-reversible. Reductiont R R Q Q Q R Q Q aR Q Q QR R R Q Q Each oxidation state, however, can attain different degrees of protonation; some of the substances involved in the study may contain as many as seven protons, which could, under certain circumstances, dissociate.Thus, the uptake of each electron may (or may not) be accompanied by the uptake of one or more protons. Fig. 14 shows all of the possible oxidation and protonation states of the substances involved; boxes drawn with solid lines represent those states which were experimentally verified (the existence of the other states, in extremely acidic or alkaline media, cannot be ruled out, however). Capital, lower-case and Greek letters are used to denote different states of oxidation and protonation, and will be applied later to describe the processes involved. Let us consider for example the reduction of the quinone H5q (c). This may take place in one step (Cy reaction) : at the potential E&t (which is close to the observed Eiqlhq half-wave potential). As a result, a single wave with the uptake of two electrons is obtained. The reduction may, of course, take place in two steps.First the semiquinone is formed (Cc reaction) : (2) H5q + e-- f H,sq- .. .. .. to which a one-electron wave of Ei91sq half-wave potential corresponds. by a second reduction step (cy reaction) : This is followed .. .. (3) H5sq- + e-- + H,hq” . . .. causing the occurrence of a second one-electron wave with Eaeqpg half-wave potential. Whether under given experimental circumstances a single two-electron wave or two single one-electron waves are observed, depends on the stability of the semiquinone. Combining equilibria (1)-(3), one can describe the disproportionation (or dismutation) of the semi- quinone with the equilibriumAugust, 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 717 Double waves 7 Ewsq .-\ f r---------- 1 I I I H7q2+ 1 1 A I r---------- 1 I I H6 9' I I I B I r------ ----i I I I q5- I 1 H I 7 r---------- 1 I I I H 7 sq+ I I a I r-------- -1 I I H3sq3- I I e I -1 r-- ------ I 1 H*sq4-- I I f I r--------- 1 Hsq5- I I 9 I r--------- 1 I I I h sq6- I I I H3 hq4- EII 1 r -.- ---- --- I 1 I Hhq6- I 77 I I r - - - - - - - - - 1 I I I hq7- I I v I L - _ _ _ _ - _ _ _ J f I EXqlhc, Single waves Fig.14. Reduction and protonation of anthraquinones.718 QURESHI et al. : ELECTROCHEMICAL STUDIES OF Analyst, Vole 104 The stability of the semiquinone can be measured through the value of the semiquinone formation constant, K,,, defined from reaction (4) as Heyrovsky and Kuta26 applied such considerations when describing the equation of the polarographic waves of quinones, and came to the conclusion that: (a) if K,, = 0 a single wave with a slope corresponding to a two-electron reduction is obtained; (b) if K,, > 16 two separate waves are obtained, each with a slope corresponding to a single-electron uptake ; and (c) if 0 < Ksq < 16 a single wave is obtained, with a slope corresponding to anything between These considerations explain the curious fact, mentioned under Results, that in many instances single waves were observed, for which the logarithmic analysis gave for the number of electrons a value close to 1, while microcoulometry always resulted in a number close to 2. The reduction of quinones may or may not be accompanied by the uptake (and in some instances by the release) of one or more protons.Although individual polarographic waves do not reveal such processes, the evaluation of the half-wave potential versus pH curves makes it possible to find out about them. To see this clearly let us again examine the reduction of a quinone (H5q). Let us suppose that Ks, = 0, so that there is no semiquinone formation; the reduction can therefore be described as and 2 electrons per molecule. * * (6) .. .. .. H5q + 2e- :+ H,hq2- Depending on the pH of the medium, the quinone may undergo stepwise dissociation. our purpose it will be sufficient to consider the first two steps only: For H5q f H+ -+ H,q- .. .. * . (7) .. and .. .. * * (8) H,q- + H+ + H,q2- .. with the dissociation equilibrium constants .. .. * ' (9) ..and Further, the hydroquinone formed in process (6) may undergo stepwise protonation. sidering the first two of these steps, we can have Con- . . (11) H,hq2- + Hi- + H6hq- .. . . .. and H6hq- + Hr + H,hq .. .. .. . . (12) With the symbols used in Fig. 14 these processes can be characterised with the following dissociation equilibrium constants :Azlgust, 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 719 .. * . .. . . (13) [H+l [H5hq2-1 CH6W-l K P Y = and The oxidation - reduction potential of the system described in reaction (6) can be expressed as x log [H5hq2-l ~ .. .. . . (15) 0.059 2 [H&l E = Eqlhq - - where E+q is the standard oxidation - reduction potential. The two concentrations included in the logarithmic term are in general not known, but can be calculated from the mass balance equations : and where chq and cq are the analytical concentrations of the hydroquinone and quinone species, respectively.Combining equations (9), (10) and (13)-(17) we can express the oxidation - reduction potential as Provided that the diffusion coefficients of the oxidised and reduced forms are equal, the half-wave potential is measured at the point where the analytical concentrations of chq and cq are equal. Thus, from equation (18) we can express the dependence of half-wave potentials on the hydrogen-ion concentration with the equation The experimentally obtained E, v e n m pH curves should conform to this equation. The correlation of theory with experimental results is simple if one considers firstly such extreme cases where equation (19) can be drastically simplified.For example, in very acidic solutions we have and also Applying such simplifications in equation (19) we arrive at the expression X log Kap KO, - 0.059pH . . .. . . (22) 0.059 2 E+ = EQpq - - AS long as the conditions described in the inequalities of (20) and (21) prevail, we therefore720 expect the decrease in the half-wave potential with pH with a 60-mV slope. words, under these conditions the reaction QURESHI et al. : ELECTROCHEMICAL STUDIES OF Analyst, VoZ. 104 In other H,q + 2H+ + 2e- + H,hq . . .. .. . . (23) (using the symbols of Fig. 14, a c 3 a process) is taking place. If these conditions are not more valid, one can describe others that fit the case and with an analogous treatment one can calculate the expected slope of the E+ versus pH curves.For the model chosen one can combine altogether nine different conditions (some of which are less likely to prevail in practice than others), which are shown in Table 111. TABLE I11 SLOPES OF 224 - PH CURVES AND REDUCTION PATHS FOR THE TWO-ELECTRON SINGLE-WAVE REDUCTION OF A QUINONE 1st condition increasing pH According to such considerations, the E, versus pH curves should consist of linear sections with slopes of 0, -30, -60 or -90mVpH-l (the occurrence of a -120mVpH-1 slope being most unlikely), for two-electron, single-wave processes. The experimental results, as shown in Figs. 5, 6 and 7, do in fact correspond to such predictions. It can also be shown that the pH values corresponding to the intersections of the linear portions with different slopes are in fact equal to the appropriate pK values of the substances involved in the process.In a corresponding manner, one can involve other species with different degrees of protonation i n such considerations, and carry out calculations in a similar fashion. Also, it is easy to extend such considerations to each of the double waves (when KBs > 16) with one- electron uptake. Based on such considerations, we evaluated au of the available 224 versyszcs pH curves and, taking into account all other experimental evidence and data obtained from the literature, have summarised our findings in Table IV. Our aim was to find (a) which species and how many electrons are involved in the electrochemical reduction and (b) what are the pK values of the quinones, semiquinones and hydroquinones involved.Table IV also contains information on the E+ versus pH curves themselves and all pK values available in the literature. The symbols and abbreviations are the same as those used in Fig. 14. It is TABLE IV REDUCTION OF ANTHRAQUINONE DERIVATIVES Parameters of the El vs. pH line pK values Reduction mechanism gxperimentally \ Compound Wave pHrange V V pH-' (cf., Fig. 14) obtained Taken from literature (a) Aqueous solutions- I1 qlhq 5.7-11.9 -0.125 -0.060 C; -+ a 111 q/hq 7.2-11.9 -0.220 -0.060 C * a ~ K C D = 8.35 6.9-8.35 -0.160 -0.060 C -., a IV qbq { 8.35-12.1 +0.096 -0.090 D -.,aAugust, 1979 STRONGLY CHELATING ANTHRAQUINONE DERIVATIVES 721 Compound V VI VII VIII IX X XI XI1 XI11 XIV xv XVI XVII XVIII pH range 3 .6 - ~ 6 -6-mll -1 1-1 3.2 3.5-6.0 6.0-6.9 6.9-9.1 9.1-12.4 6.0-13.0 6.0-13.0 7.7-11.9 7.7-11.9 4.0-7.5 7.5-1 3.1 7.5-12.2 7.4-12.9 6.6-12.2 2.4-4.0 4.9-6.25 8.8-11.8 1.8-5.5 5.5-11.2 11.2-12.5 6.25-8.8 2.8-5.5 (0.75) 5.5 (or 7.5)-8.0 8.0-10.7 10.7-12.8 1.6-2.5 2.5-5.6 5.6-8.2 8.2-9.7 9.7-11.4 2.2-4.67 4.67-10.65 10.65-11.6 11.6-12.9 7.1-8.6 8.6-12.4 7.1-8.4 8.4-11.9 7.0-8.4 8.4-11.9 7.0-8.9 8.9-11.9 (b) Solutions containing 75% ethanol- 3.4-9.2 'I1 ¶lhq { 9.2-11.6 IV ¶Ihq { 7.95-1'1.7 6 9-7 95 V qlhq 5.1-1 3.2 2.6-11.9 vll* $h9q 2.6-11.9 xvlll $zq 4.9-12.1 5.6-12.1 TABLE IV (continued) Parameters of the E ) vs. pH line Reduction Intercept/ Slope/ mechanism V V pH-' (cf., Fig. 14) - 0.140 - 0.060 - 0.160 -0.060 -0.340 -0.030 - 0.065 - 0.060 + 0.045 - 0.060 - 0.225 - 0.060 + 0.025 - 0.060 - 0.120 - 0.060 + 0.245 - 0.090 4-0.020 -0.060 - 0.185 - 0.060 -0.235 -0.060 - 0.195 - 0.060 - 0.070 - 0.060 -0.720 0 - 0.190 - 0.060 - 0.165 - 0.060 -0.845 0 -0.185 -0.060 - 0.430 - 0.030 - 0.110 - 0.060 - 0.230 - 0.060 - 0.400 - 0.030 -0.155 - 0.060 - 0.445 - 0.030 - 0.155 - 0.060 -0.500 -0.030 -0.640 0 - 0.125 - 0.060 -0.735 0 -0.230 -0.060 -0.620 0 - 0.120 - 0.060 -0.740 0 - 0.210 - 0.060 -0.145 -0.060 -0.435 -0.030 -0.155 - 0.060 - 0.130 - 0.090 - 0.240 - 0.060 -+ 0.045 - 0.090 -0.185 -0.060 - 0.190 - 0.060 - 0.130 - 0.060 -0.130 -0.060 -0.085 -0.060 - 0.225 - 0.060 pK values obtained Taken from literature r A 1 Experimentally PKuCj = 6 ~ K C D = 6.67''; 5.60:' pKfiy= 11 ~ K D S = 9.1 P K U ~ = 6.0 PKfjyW PKCD ~ K D E = 11.0P PKDE = 10.7827; 11.16 *@ ~ K C D = 6.91'l p K q ~ pKpy% 4.5 ~ K D E ~ ~ K E P W 8.80 pKYsw ~ K O D pKaD = 6.25'0 ~ K F O = 12.2"' PKDE = 11.2 pK,D = 5.50"; 5.12" PK&Z PKy8 3 P K 8 ~ N PKDE ~ K E F = 13.1OZ8 P K D ~ " PKDE pKvs = 8.0 )I< I E = 11.25a8; 11.20*' pKcn = 2.40"; 4.894 p K ~ 8 = 5.54"; 7.55' pKEF = 10.7 PKEB = 10.07"; 10.43' pKao = 11.98"; 11.19' pKpYw PKDE PKEF = 6.6 pKys w pK6€= PKEF ~ K F O = 10.O1O; ~ K O H = 12.31° pKna = 0.5"; pKaD = 1.41° ~ K D E = 2.5"; p K ~ a = 5.810 pKy8 = 11.6 PK+W ~ K C D PKDE = 10.65** pKpy= ~ K D E ~ K C D = ~ K D E = 8.6 pKcd = 8.4 ~ K C D = 4.67a8 ~ K C D = pKm = 8.4 pKcd = 8.9 ~ K C D = 8.5*O pK,p = 9.8 ~ K C D = 10.8" ~ K O D = 9.6 pKcn = 9.2 pKan = 7.86P' ~ K D E - 1427 ~ K C D = 7.95='; ~ K D E = 13.31'' p K q W pKan worthwhile pointing out that the experimentally obtained, "polarographic" pK values agree well with the values taken from the literature, which were mostly obtained by spectro- photometry.The agreement is very good for the aqueous solutions of XIII, XIV, XV and722 QURESHI, SVEHLA AND LEONARD XVIII and is acceptable even with the alcoholic solution of 111. Only with the aqueous solution of VI do the data differ substantially. In only these instances do we possess parallel polarographic and spectrophotometric data. When evaluating these results we also tried to find linear free-energy relationships among certain derivatives. It must be said that the original Hammett30 equation cannot be applied here (nor are the appropriate reaction and substituent constants available), and none of the more recent e q u a t i o n ~ ~ l , ~ ~ would really be valid.Still, it seemed to be reasonable to attempt tofind some correlation between the half-wave potentials (at a fixed pH) and the various pK values available. When doing so we hoped to extend the available polarographic free- energy correlation data for a n t h r a q u i n o n e ~ . ~ ~ , ~ ~ ~ ~ ~ We constructed separate diagrams for the P K ; ~ and pK,, values (only these are available in significant numbers) using the data in Table IV. For half-wave potential values we used the intercept figures (which correspond to the half-wave potential values extrapolated to pH 0). Although a number of points fall on a well defined straight line, as expected, we have not published such plots, because in both instances there were a number of unexpected deviations, which we were unable to explain.Also, there are parallel, sometimes greatly differing, values for the same pK value available in the literature, and a choice among them cannot easily be made. Although one can select a number of points that would show a reasonable straight-line relationship, such a selection would be arbitrary and not scientifically justified. A more detailed account of these attempts is available in a thesis.25 1. 2. 3. 4. 6. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. References Relcher, R., Leonard, M. A., and West, T. S., J . Chem. SOC., 1959, 2390. Leonard, M. A., in Johnson, W. C., Editor, “Organic Reagents for Metals and Certain Radicals,” Volume 2, Hopkin and Williams, Chadwell Heath, 1964, p.1. Leonard, M. A., and Nagi, F. I., Chem. Commun., 1968, 20, 1224. Leonard, M. A., and Nagi, F. I., Tuluntu, 1969, 16, 1108. Leonard, M. A., and Nagi, F. I., Analyt. Lett., 1969, 2, 15. Laird, C. K., and Leonard, M. A., Talanta, 1970, 17, 173. Al-Ani, K., and Leonard, M. A., Analyst, 1970, 95, 1039. Al-Ani, K., and Leonard, M. A., Proc. SOC. Analyt. Chem., 1971, 8, 190. Leonard, M. A., and Murray, G. T., Analyst, 1974, 99, 645. Leonard, M. A., Analyst, 1975, 100, 275. Deane, S. F., and Leonard, M. A., Analyst, 197?, 102, 340. Furman, N. H., and Stone, K. G., J . Am. Chem. SOC., 1948, 70, 3056. Wiles, L. A., J . Chem. SOC., 1952, 1368. Edsberg, R. I., Eichlin, D., and Garis, J. J., Arzalyt. Chem., 1953, 25, 798. Starka, L., Vystrcil, A., and Starkova, B., Chem. Listy, 1957, 51, 1440. Broadbent, A. D., and Sommermann, E. F., J . Chem. Suc. (B), 1967, 376. Broadbent, A. D., and Sommermann. E. F., J . Chern. SOC. (B), 1968, 519. Gill, R., and Stonehill, H. I., J . Chem. SOC., 1952, 1845. Gill, R., and Stonehill, H. I., J . Chem. SOC., 1952, 1857. Al-Ani, K., Ph.D. Thesis, Queen’s University, Belfast, 1971. Willard, H. H., Merritt, L. L., Jr., and Dean, J. A., “Instrumental Methods of Analysis,” Fourth Gilbert, G. A., and Rideal, E. K., Trans. Farallay Soc., 1951, 47, 369. HeFovsky, J., and Kuta, J ., “Principles of Polarography,” Czechoslovak Academy of Sciences, Prague, and Academic Press, New York, 1966, pp. 311 et seq. Brown, E. R., and Large, R. F., in Weissberger, A., and Rossiter, B. W., Editors, “Physical Methods of Chemistry,” Volume 1, Part ILA, “Electrochemical Methods,” Wiley, New York, 1971, pp. 502 et seq. Qureshi, G. A., Ph.D. Thesis, Queen’s University, Belfast, 1975. Heyrovsky, J., and Kuta, J ., “Principles Of Polarography,” Czechoslovak Academy of Sciences, Prague, and Academic Press, New York, 1966, p. 181. Radcliffe, Bl., Ph.D. Thesis, Queen’s University, Belfast, 1973. Murray, G. T., Ph.D. Thesis, Queen’s University, Belfast, 1973. Ingman, F., Talanta, 1973, 20, 135. Hammett, L. P., J . Am. Chem. Soc., 1937, 59, 96. Jaffe, H. H., Chem. Rev., 1953, 53, 191. Wells, P. R., Chern. Rev., 1963, 63, 171. Crawford, R. J., Levine, S., Elofson, R. M., and Sandin, R. B., J . Am. Chem. SOC., 1957, 79, 3153. Wiles, L. A., J. Chem. SOC., 1952, 1958. Edition, Van Nostrand, New York, 1965, p. 692. Received June 3rd, 1977 Amended October 19th, 1978 Accepted January 30th, 1979