SEPTEMBER, 1973 THE ANALYST Vol. 98, No. 11 70 Mass and Charge Transfer Kinetics and Coulometric Current Efficiencies Part IX.* An Examination of the Titanium(1V) - Titanium(II1) System and the Effects of Ultratrace Impurities in Sulphuric Acid BY E. BISHOP AND P. H. HITCHCOCK-/- (Chemistry Depavtment, University of Exeter, Stocker Road, Exeter, EX4 4QD) Titanium(II1) has found considerable use as a coulometric intermediate, and is the strongest reductant that can be generated with good efficiency in aqueous media. At such high concentrations, a surfactant impurity becomes adsorbed on the working electrode and very seriously reduces the speed of the charge-transfer process and therefore the current efficiency. A method of purification of the sulphuric acid by electrosorption is shown to give a charge-transfer rate constant in excess of 10-5 1 cm-2 s-l, but charcoal column purification is without effect.Both 7 and 10 M sulphuric acid media were examined, but because the mass-transfer rate is decreased by the higher viscosity of the 10 M acid there is little difference in current efficiency. Addition of iron(II1) or iron(I1) hinders the adsorption of the impurity and the deactivation of the electrode, but offers no special benefit. Electrode kinetic parameters are reported for various treated and untreated media, and current efficiencies for the generation of titanium(II1) are computed. The behaviour in perchloric acid is re-interpreted. The medium is restricted to strong sulphuric acid. TITANIUM(III) was first introduced as a coulometric aniperostatic intermediate by Arthur and Donahuel in 1952.It was one of the earliest intermediates to receive scientific rather than empirical e~aluation,~J and, although the literature is not e~tensive,l-~~ titanium(II1) shares with iron(I1) the greatest number of applications among reductant intermediates. On this account, and because it is the most powerful reductant that can be satisfactorily generated in aqueous media, a thorough investigation has been made both of the electrode kinetics and generation current efficiency of the titanium(1V) - titanium(II1) system and of the genera- tion medium. The system has been used in the determination of iron(II1) d i r e ~ t l y , ~ v ~ - ~ s ~ s ~ ~ and of other materials via iron (I1 I) ,8316917 9 21 ruthenium( IV) ,24 925 iridium (IV) , 25 molybdenum( VI) ,4 s5 s2' uranium(V1) ,11914926 phosphate via molybdenum(V1) ,13 chromium(V1) in standardisation,20 vanadium(V),8~11s15~21 titanium metal and its compounds,15~16*21 selenium( IV),4 tell~rium(IV),~s~ tellurium(VI),G 9-quinonedioximeg and organic dyestuff s.17~~8 Gold plated on platinum was originally used as the electrode material1 ; copper and copper amalgam,23 titanium metal,15 and mercury14J9,20 have been used, but platinum is the most popular material.3~6~10~11~15-17~21 Hydrochloric acid,l~59~~ citric acid,l4 and admixtures of hydrofluoric7 or phosphoric22 acid with sulphuric acid have been used as the medium, but sulphuric acid at high concentra- tion2~3~6,8~10,11~26 has met with most approval.Lingane and Kennedy2 found the charge-transfer process to be slow at the dropping- mercury electrode in perchloric acid of all concentrations and in dilute sulphuric and hydro- chloric acids, as well as in a 4.0 M solution of sodium sulphate in 0.2 M sulphuric acid, but fast in a 4.0 M solution of sodium hydrogen sulphate in 0.2 M sulphuric acid and in high concentrations of sulphuric and hydrochloric acids.The fast charge transfer was ascribed to complexation (ligand exchange with water molecules), and specifically to hydrogen sulphate ions in sulphuric acid media. The process is assisted also by the increase in conditional potential of the system in such media. Habashy28y29 confirmed these findings, but found that the titanium(1V) wave was not well separated from the hydrogen-ion wave at platinum. The conditional potential shifts faster than the background reduction wave with complexation, Present address: Ever Ready Co.(G.B.) Ltd., Central Research Laboratory, St. Ann's Road, * For Part VIII of this series, see p. 563; for Part X, see p. 635. London, N15 3TJ. @ SAC and the authors. 625626 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, Vol. 98 and so high sulphuric or hydrochloric acid concentrations are required. Even so, the titanium and hydrogen-ion waves merge to the extent of leaving the limiting current plateau. ill defined, so that high titanium(1V) concentrations are necessary in order to attain an acceptable current efficiency. Phosphoric acid fails to retain titanium( IV) in solution at concentrations above 0.01 M, and so is useless.If strong oxidants such as cerium(1V) or manganese(VI1) are to be determined, they will attack concentrated hydrochloric acid. The choice is therefore restricted to sulphuric acid media. EXPERIMENTAL Apparatus, reagents and general procedures have been described earlier.30s31 The only reasonably pure compound that is readily available is AnalaR potassium oxodioxalatotitanate(1V) ; other materials are liable to contain appreciable amounts of iron, silicon and vanadium among their impurities, and these three elements are particularly difficult to remove, so making the materials that contain them unsuitable for use in kinetic studies. A high-purity, but inert, titanium(1V) oxide was kindly supplied by Laporte Indus- tries Ltd., who quoted levels of iron(II1) of less than 10 p.p,m.and vanadium(V) of less than 5 p.p.m.; it is believed that the material was prepared by vapour-phase hydrolysis of purified tetra-n-butyl titanate(1V). Examination proved that the two compounds mentioned were of a satisfactory purity for use as starting materials. Titanium(1V) chloride-The titanium( IV) oxide was too inert to be dissolved in sulphuric acid, and so was converted into titanium(1V) chloride by Schreyer's method,32 which involves reaction with ccacca'a'or'-hexachloro-m-xylene (Fluka A.G.) . The product was distilled twice before further use. TitaniGm oxodi$wrchZorate-The titanium(1V) chloride was heated with 62 per cent. perchloric acid at a mole ratio of 1 : 2 at 30 to 40 "C and 15 mm of mercury so as to remove hydrogen chloride, and the product recrystallised from the minimum amount of 10 per cent.perchloric acid. The product is reported33 to be TiO(ClO,),.xH,O, where x = 6 to 7. Alterna- tively, hydrated titanium( IV) oxide was precipitated by hydrolysis of titanium(1V) chloride or potassium oxodioxalatotitanate(1V) by adding 50 per cent. ammonia solution to a saturated solution of the titanium salt. The hydrated titanium(1V) oxide was readily separated by centrifugation, and was washed with water until free from ammonia, then dissolved in 62 per cent. perchloric acid. This solution was used directly, but the product could be crystallised out by evaporation at 30 to 40 "C and 15 mm of mercury. Titarti~m(1V) su@hate-Freshly prepared hydrated titanium( IV) oxide was dissolved in concentrated sulphuric acid and the solution diluted to give a 0-5 M concentration of titan- ium(1V) in 10 M sulphuric acid.The solution was standardised by reduction with liquid zinc amalgam, separating the amalgam, running the reduced solution into an excess of iron(II1) solution and titrating the iron( 11) produced against standard potassium dichromate with sodium diphenylamine-4-sulphonate as indicator, or against potassium permanganate with a correction being made for the blank. Titanizcm(II1) s~l$kate-Titanium(III) sulphate was made by determinate reduction of standardised titanium(1V) sulphate with liquid zinc amalgam, and was stored over the liquid amalgam. PRE-TREATMENT OF THE WORKING ELECTRODE- Normally, treatment (c) (see Part VIIsl) was used: immerse the platinum electrode in fresh aqua regia at 50 to 60 "C for 2 minutes, anodise it in concentrated hydrochloric acid for 30 s at 100 mA cm-2, cathodise it in 0.1 M sulphuric acid for 5 minutes at 100 mA cm-2 and wash it thoroughly with water.PURIFICATION OF ELECTROLYTES- (a) Electrolytic method-A portion of 7 or 10 M sulphuric acid was prepared by dilution of concentrated AnalaR or Aristar sulphuric acid in an all-glass cell that carried a clamped cover bearing ground-glass joints lubricated with distilled water. Through the joints were inserted the electrodes, a nitrogen bubbler [the nitrogen was purified by passage through chromium(I1) chloride solution and water] and a siphon tube. The solution was stirred PREPARATION OF TITANIUM COMPOUNDS-September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIRNCIES. PART IX 627 magnetically, and the working cathode was a 100-cm2 piece of platinum black.Electrolysis was conducted at 0.3 V until the current, which was initially about 50 mA, decreased to about 0.1 mA. The time required was variable, but was often as long as 6 hours. The solution was then transferred into the clean voltammetric cell by closing the nitrogen outlet, thus causing the siphon to fill. The first few portions, breaking the syphon through a tap at its highest point, were used to rinse out the cell and then discarded. (b) Column adsorption-Sugar charcoal was prepared by the action of concentrated sulphuric acid on domestic granulated sugar. The washed product was ground in a mortar and packed into a glass column to a depth of 30 cm.The column was mounted over the voltammetric cell, and 7 or 10 M sulphuric acid allowed to percolate at about 5 ml min-l through the column into the cell. A second column filled with commercial granular activated charcoal was also used. RESULTS AND DISCUSSION Before the work had progressed very far, it became evident that the working platinum electrode was being deactivated by an impurity in the system, the presence and identity of which could not be detected other than by its effect in deactivating the electrodes. It was traced to the sulphuric acid, which is used at high concentrations, and was present, although at a much lower level, in Aristar as well as in AnalaR sulphuric acid. No mention of this was made in earlier work, and, while electrode pre-treatments were given attention, no report of purification of the electrolyte has appeared.It must be assumed, therefore, that all earlier work was carried out with untreated supporting electrolytes, and it is appro- priate to discuss the behaviour of the system in such conditions. UNTREATED ANALAR SULPHURIC ACID MEDIA- A treated electrode in 7.0 M sulphuric acid gave curve 1 in Fig. 1, with a suspicion of an adsorption pre-wave. A re-treated electrode gave curve 2 in Fig. 1 after addition of a mixture of titanium(1V) and titanium(II1). Repeated scanning without intervening pre- Electrode potential versus S.H.E./V Fig. 1. Voltammograms of titanium(1V) - titanium(II1) in untreated 7.0 M sulphuric acid: 1, 7-0 M sulphuric acid, freshly activated electrode; 2, 7.0 M sulphuric acid + 2.39 x lo-$ M titanium(II1) + 1.91 x M titanium(IV), freshly activated electrode; and 3, as for curve 2 but after completing seven cycles of cathodic and anodic scanning.Ramp speed 70 mV min-l628 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, VOl. 98 treatment of the electrode showed a drastic reduction in charge-transfer speed, and after six cycles gave the curves 3 in Fig. 1; further scans produced no further change. Removal of the electrode and re-treatment before scanning gave a curve identical with curve 2. The change is therefore in the electrode and not in the solution. The same de-activation occurred when a freshly activated electrode was immersed in the titanium - sulphuric acid solution on open circuit.Deactivation is not therefore due to the passage of current. Some activity could be restored by immersing the electrode for 25 minutes in 1.0 Rii sulphuric acid, but full activation required the electrochemical treatment. At least part of the deactivation is due to adsorption and not to recrystallisation of the platinum surface. Repetitive anodisa- tion and cathodisation [method (b), see Part VI131] at 100 mA c M 2 in 1.0 M sulphuric acid produced full activity as well as the stripping method given above; there was no difference between a cathodic and an anodic finish, because any oxide that remained on the platinum surface would immediately be reduced chemically on immersion in solutions that contain titanium(II1). The hysteresis between the forward and reverse voltage sweeps in curves 2 and 3 is also symptomatic of ad~orption.~~ Adsorption slows the titanium charge-transfer process, moves the titanium reduction wave so that it merges with the hydrogen-ion wave and seriously attenuates the current efficiency for the generation of titanium( 111).Adsorption deactivation probably accounts for the poor definition of the titanium waves recorded by Lingane and Kennedy3; their polarographic waves2 were not affected in this way because the electrode surf ace was continuously being renewed. PURIFIED SULPHURIC ACID MEDIA- With a freshly pre-treated electrode, very little hysteresis occurred in electrolytically puri- fied sulphuric acid, and the curves did not change appreciably when scans were repeated several times over a period of several hours without intervening re-activation of the electrode.The adsorbable impurities were therefore present in the sulphuric acid or the water30 that was used for dilution. The water was ruled out by purification of concentrated sulphuric acid by method (a) and then dilution with untreated water, when curves as good as those in purified 7-0 M sulphuric acid were obtained. -0.2 1 I I I I I 1 I I 0 5 0.4 0.3 0.2 0.1 Electrode potential versus S.H.E./V Fig. 2. Voltammograms of titanium(1V) - titanium(II1) in 7.0 M sulphuric acid purified by electrosorption: 1, treated 7.0 M sulphuric acid; 2, treated 7-0 M sulphuric acid + 1.0 x M titanium(1V); and 3, treated 7.0 M sulphuric acid + 1.08 x M titanium(1V) + 1.02 x M titanium(II1).Electrode activated before each cycle. Ramp speed 70 mV min-'September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES. PART I X 629 Column treatment failed to remove the surfactant impurity, and contaminated the solution with ionic impurities in one instance. The method was therefore abandoned. Fresh stock titanium(1V) and titanium(II1) solutions in electrolytically purified sulphuric acid were prepared. Fig. 2 shows scans of purified 7.0 M sulphuric acid (curve 1) with a very low background current over a large range of potentials, and of titanium(1V) alone (curve 2) and mixed with titanium(II1) (curve 3). Curves 2 and 3 show remarkably little hysteresis, which is a good indication that the platinum surface did not change appreciably over the potential range scanned.Repeated scans showed excellent reproducibility (better than 1 per cent.) for both titanium and background waves. Similar behaviour occurred in purified 1 0 ~ sulphuric acid, as shown in Fig. 3. The curves in Figs. 1 to 3 were analysed for the kinetic parameters listed in Table I by the diffusion-corrected Tafel plots of Lewartowic~~~ and also, for Fig. 1, by pattern theory.35 The difference between the treated and untreated supporting electrolyte is dramatic. 0 5 0.4 0.3 0 2 0.1 Electrode potential versus S.H.E./V Fig. 3. Voltammogram of titanium(1V) - titanium(II1) in 10.0 M sulphuric acid purified by electrosorption: 1, treated 10.0 M sulphuric acid; and 2, treated 10.0 M sulphuric acid + 2.1 x M titanium(1V) + 3.4 x 10-3 M titanium(II1).Electrode activated before each scan. Ramp speed 70 mV min-l When a titanium solution in purified sulphuric acid was left in the cell for 1 day, some gain in impurity occurred, as shown by increased hysteresis and slowing of the charge-transfer process. Repetition of the experiments in a completely sealed all-glass double cell showed detectable accession of impurities in 1 day. In view of other finding~,~'39~7 desorption of some substance from glass or platinum may be the source of the returning impurity, although the possibility that a trace amount of oil was not removed from the cylinder nitrogen in the scrubbers cannot be excluded. After several experiments involving contamination, the 1.0 M sulphuric acid used in the anodisation - cathodisation activation of the electrode was found to lose some effectiveness, in contrast to a fresh electrolyte.Thereafter, the electrolyte was discarded after each activation by either method. A solution of 10 M sulphuric acid was prepared by dilution of Aristar sulphuric acid taken from a freshly opened bottle. Background and titanium(1V) reduction waves are shownin Fig. 4. There was some hysteresis and repeated scans, without reactivating the electrode, showed a slowing down of the charge-transfer process, but full activity was restored by the normal electrode treatment methods. The screw-cap of the bottle was sealed with white wax and, although every care was taken in manipulation, it is possible that a trace amount of wax might have contaminated the acid. When purified by the electrosorption method, Aristar sulphuric acid gave the same behaviour as purified AnalaR sulphuric acid, and even in all-glass apparatus re-contamination was detected after about 20 hours.630 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [Analyst, VOl.98 TABLE I KINETIC PARAMETERS FOR THE TITANIUM(IV) - TITANIUM(III) SYSTEM AT ACTIVATED Titanium concentration range 0.15 x to 1.3 x M PLATINUM ELECTRODES kmass ox/ kmass red/ Medium K / 1 cm-2 s-l a B E,’/V 1 cm-2 s-l 1 cm-z s-l 7-0 M H2S0,, pre- 2 1-4 x 0-58 to 0-69 0.57 to 0.65 0.215 1.9 X 1-9 X 10.0 M H,SO,, pre- 2 1.1 x 0.58 to 0.66 0.57 to 0.63 0-307 1-05 x loAs 0-99 x lop6 electrolysed electrol ysed treated (b) 10.1 x lo-* 0.615 7-0 M H,SO,, un- (u) 6-7 x lo-* 0.728 ::ii8} 0.215 * 1.9 x 10-6 7.0 M H,SO,pZus iron, (c) >, 0.56 0.57 0.215 1.9 x 1.9 x [Fe]/[Ti] = 1 ( d ) 5 x 10-7 0.6 1 0.46 0.215 * 1.9 x 10-6 (a) Aged electrode, negative-going scan.(b) Aged electrode, positive-going scan. (c) (d) Electrode immersed in solution for 30 minutes after activation. Electrode activated, used several times, then left in solution for 12 hours. * Limiting current region ill defined. It is notable that while the impurities exerted a gross effect on the shape of the titanium wave, very little change was observed in the hydrogen-ion reduction wave. UNTREATED SULPHURIC ACID MEDIA CONTAINING ADDED IRON- Voltammetric scans of a mixture of iron and titanium in untreated sulphuric acid showed that the iron mitigated the effect of impurities on the titanium wave. Fig. 5 shows the behaviour of an approximately equimolar solution of titanium and iron in the oxidation states 0.5 0.4 0.3 0 2 0.1 Electrode potential versus S.H.E./V Fig.4. Voltammograms of titanium(1V) in 10.0 M untreated Aristar sulphuric acid: 1, 10.0 M Aristar sulphuric acid, electrode freshly activated; 2, 10.0 M Aristar sulphuric acid + 2.3 x M titanium(IV), electrode used to record three cycles of background, then used for this scan without reactivation ; 3, recorded immediately after curve 2 without electrode activation; and 4, elec- trode activated, conditions as for curve 2. This figure is an actual copy of an original X - Y recorder chart, showing the noise typical of such plots Ramp speed 70 mV min-l.September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES.PART I X 63 I TiIv, FeIII and FeII. A freshly activated electrode used for three scans without reactivation gave titanium(1V) reduction waves that differed by only 5 per cent.; the iron(II1) wave was consistent within 1-5 per cent. The hysteresis was much less than in the absence of iron. After leaving the electrode in the solution on open circuit for 12 hours, curve 2 in Fig. 5 was recorded. Had there been no iron present, the curve would have been similar to curve 3 in Fig. 1. If the mole ratio of iron to titanium decreased much below 0-5, the deactivation of the electrode accelerated. At a constant mole ratio of iron to titanium, deactivation of the electrodes was retarded to the same extent whether the solution contained iron(I1) iron(II1) and titanium(IV), or iron(II), titanium(1V) and titanium(II1).The effect of adding iron was not, therefore, that of increasing the zero-current potential in the medium to a more positive potential at which adsorption did not occur. If an electrode that had been deactivated by exposure to an impure titanium - sulphuric acid solution was transferred to an equimolar titanium - iron solution, the observed rate of charge transfer was very markedly decreased. It would therefore seem that the function of the iron is to slow down the adsorption of impurities on the electrode, and not to speed up the reduction of titanium(1V) at a fouled electrode. Electrode potential versus S.H.E./V Fig. 6. Voltammograms of titanium(1V) and iron(II1) - iron(I1) in untreated 7.0 M sulphuric acid: 1, 7.0 M AnalaR sulphuric acid + 10.1 x M titaniuin(1V) + 4-5 X M iron(II1) + 3.4 x 10-3 M iron(II), activated electrode; and 2, electrode left in solution for 12 hours on open circuit, then this curve recorded without further activation PERCHLORIC ACID MEDIA- As perchlorate is the least cornplexing of anions, it was hoped to be able to determine the characteristics of the purely solvated oxotitanium(1V) and titanium(II1) ions in perchloric acid.Scans were performed in 1, 3, 7 and 10 M perchloric acid solutions of oxotitanium(1V) perchlorate, but no titanium reduction wave was found. The experiments were repeated in perchloric acid purified by the electrosorption process, and with electrodes pretreated by all meth0ds,~1 but still no wave was obtained. This result seemed to support Lingane and Kennedy’s contention2 that the titanium couple was “completely irreversible” at all con- centrations of perchloric acid.More exactly, it indicated that the conditional potential of the purely solvated ions and the charge-transfer rate parameters were such that hydrogen ion was reduced in preference to titanium(1V). However, treatment of oxotitanium(1V) perchlorate in perchloric acid with liquid zinc amalgam showed a reddish violet coloured layer in contact with the amalgam, which dis- appeared on shaking, and tests showed that the solution, which was initially free from632 BISHOP AND HITCHCOCK: MASS AND CHARGE TRANSFER [ArtdySt, Vol. 98 chloride, now contained chloride. The titanium(II1) was reacting directly with perchlorate and reducing it to chloride.Titanium(II1) is therefore unstable in perchloric acid and so anodic scanning of such solutions would give misleading results that represent the sum of a homogeneous and a heterogeneous reaction. The homogeneous kinetics were later investi- gated,38 and prompted mechanistic di~coveries.~g THE ELECTRODE KINETIC PARAMETERS- Voltammetry in sulphuric acid purified by electrosorption showed that the limiting cathodic current was proportional to the titanium(1V) concentration and the anodic limiting current was proportional to the titanium( 111) concentration. Essin40 found the same relation- ship at a mercury-jet electrode, but Khomutov and Aboimov41 reported that the limiting cathodic current was not proportional to the titanium(1V) concentration, but they did not purify the electrolyte and it seems likely that electrode poisoning was responsible.Delahay and Tra~htenberg~~ studied the system at a mercury electrode and examined the effects of adsorption of cyclohexanol, n-pentanol and thymol on the charge-transfer rate in 1.0 M tartaric acid. They concluded that a monolayer of these substances on the electrode decreased the value of k from 5 x 1 cm-2 s-l at a clean electrode by several orders of magnitude, although there was little effect on the charge-transfer coefficients. This conclusion agrees with the results in Table I. The Lewartowicz plots for purified sulphuric acid give exchange currents that are very close to the limiting currents, which can therefore be regarded as being least possible values.The value of the charge-transfer rate constant, k , must also be regarded as being a least possible value. However, even if the true rate constant were considerably higher than 10-5 1 cm-2 s-1, very little change in the computed voltammograms would occur, because this value gives voltammograms that deviate only slightly from the infinitely fast mass transfer controlled curves. The tabulated values can be used with confidence in computing current efficiencies. Essin40 found that the sum of the charge-transfer coefficients, u + 8, was less than unity. In pure sulphuric acid, although hysteresis is slight, the sum is greater than unity. This may be because the generated species is not TiIIIL6, but a binuclear complex with a charge-transfer spectrum,39 which could be I-- l t 5 - 2 n L -1 or, as the hydrogen sulphate ion appearslto play a key r61e,2 83-33 In both formulations, an electron-tunnelling mechanism can be expected.THE GENERATION CURRENT EFFICIENCY FOR TITANIUM (111)- The background, reduction of hydrogen ion, rate parameters appear to be little affected by the condition of the electrode, and computer curve-fitting methods3O gave the values in Table 11. From these and the values in Table I, voltammograms were computed by VOLTAM- METRY 9,43 and gave an excellent fit with the experimental curves. There is therefore no potential dependence in k or a. Current efficiency curves were computed for 0.6 M titanium(1V)September, 19731 KINETICS AND COULOMETRIC CURRENT EFFICIENCIES. PART IX TABLE I1 RATE PARAMETERS FOR HYDROGEN EVOLUTION Medium [H3O+I /M k/l(a+aE) mol-(z+aH)Cm-2 s-1 a 7~ H,SO,, untreated .. .. 7 1.85 x 10-lo 0.5 7~ H,SO,, pre-electrolysed . . 7 1.85 x 0.5 1 0 ~ H,S04, pre-electrolysed . . 10 1.85 x 0.5 633 in 7.0 M sulphuric acid.3 Curves 1 and 2 in Fig. 6 show the situation for untreated and purified sulphuric acid, respectively. For a current efficiency loss of 1 p.p.m., the maximum usable current density predicted is 9.5 x A cm-2 for untreated and 5.92 x 10v2 A cm-2 for puri- fied sulphuric acid, and for a loss of 0.1 per cent. the values are 1.94 x and 0.109 A cM2, respectively. Lingane and Kennedy's value for a 0.1 per cent. loss in current efficiency was 3 x 10-3 A cm-2. However, they took no steps to purify the sulphuric acid, and experiment shows that different batches of sulphuric acid have different surfactant impurity levels.I I I I I I I I 0.02 , 0.06 0.10 0-14 0.18 Current density/A cm-* Fig. 6. Computed loss of current efficiency in the generation of titanium(II1) : 1, untreated 7.0 M sulphuric acid medium; 2, 7.0 M sulphuric acid cleaned by electrosorption; and 3, 10.0 M sulphuric acid cleaned by electrosorption. Parameter values as in Tables I and 11, with Sx = cm, [TiIvl~ = 0.6 M and [TP]B = 0 ; a for curve 1 = 0.56, otherwise 0.58 The generation efficiency in clean 10 M sulphuric acid is shown by curve 3 in Fig. 6. This is of interest, because the conditional potential in this medium (Table I) is more positive by 56 mV, yet the computed current efficiencies are about the same for 7 and 10 M sulphuric acid, because the mass-transfer rate constants are decreased by about 50 per cent.by the higher viscosity of the 10 M sulphuric acid and the effects nullify each other. It does appear that the addition of iron is of some use, and might be applied in the determination of vana- dium(V) by adding iron(III), or uranium(V1) by adding i r ~ n ( I I ) . ~ ~ CONCLUSIONS Voltammetry at platinum electrodes has shown that the purest sulphuric acids contain enough adsorbable impurities at the high concentrations used to reduce the charge-transfer rate of the titanium(1V) - titanium(II1) system by several orders of magnitude compared with the values in sulphuric acid purified by electrosorption. Such impurities seriously impair the current efficiency of the generation of titanium(II1). In perchloric acid, the immediately following chemical step of reduction of perchlorate to chloride makes the generation of titanium(II1) impossible in this medium.The non-unity value of a + /3 encourages the view that a binuclear complex of titanium(1V) and titanium(II1) is the electrode reaction product.634 BISHOP AND HITCHCOCK purity 1. 2. 3. 4. 5. 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. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. We are deeply grateful to Imperial Chemical Industries Limited for a grant of research funds extending over 3 years. We also thank Laporte Industries Ltd. for the gift of high- titanium (IV) oxide. REFERENCES Arthur, P., and Donahue, J. F., Analyt.Chem., 1952, 24, 1612. Lingane, J . J., and Kennedy, J. H., Analytica Chim. Acta, 1956, 15, 294. __- , Ibid., 1956, 15, 465. Agasyan, L. B., Nikolaeva, E. R., and Agasyan, P. K., Zh. Analit. Khim., 1967, 22, 904. Binder, E., Goldstein, G., Lagrange, I?., and Schwing, J.-P., Bull. SOC. Chim. Fr., 1965, 2807. Chavdarova, R., and Sheltanov, Ch., C. R. Acad. Bulg. Sci., 1967, 20, 565. Clemency, C. V., and Hagner, A. F., Analyt. Chem., 1961, 33, 888. &ta, F., and Ji;, Y . C., Rev. ChirPz. Acad. Pop. Rom., 1962, 7, 123. Dobychin, S. I., and Kozulya, A. P., Zh. Analit. Khim., 1962, 17, 148. Hitchcock, P. H., Ph.D. Thesis, University of Exeter, 1969. Kennedy, J. H., and Lingane, J . J., Analytica Chim. Acta, 1958, 18, 240. KuCerovskS;, Z., PPibyl, M., and SiSka, M., Chem.Listy, 1965, 59, 604. Lagrange, P., and Schwing, J., Bull. SOC. Chim. Fr., 1965, 2811. Lingane, J . J., and Iwamoto, R. T., Analytica Chim. Acta, 1955, 13, 465. Malmstadt, H. V., and Roberts, C . B., Analyt. Chem., 1955, 27, 741. -- , Ibid., 1956, 28, 1884. Nikolaeva, E. R., Agasyan, P. K., Terenova, K. Kh., and Boikova, S. I., Vest. Mosk. GO~. Univ., Papier, J., Hkvue MLtall., Paris, 1954, 51, 723. Parsons, J . S., and Seaman, W., Analyt. Chem., 1955, 27, 210. Roberts, C. B., Diss. Abstr., 1956, 16, 1798. Sheitanov, Ch., Chavdarova, R., and Konstantinova, M., C. R. Acad. Bulg. Sci., 1966, 19, 1147. SlovAk, Z., and Pfibyl, M., 2. analyt. Chem., 1967, 228, 266. Stenina, N. I., and Agasyan, P. K., Zh. Analit. Khim., 1966, 21, 965. Stenina, N. I., Agasyan, P. K., and Berentsveig, G. A., Ibid., 1967, 22, 91. Takeuchi, T., Yoshimori, T., and Kato, T., Bunseki Kagaku, 1963, 12, 840. Yen, H.-Y., and Liu, Y. H., K’o Hsueh T’ung Pao, 1966, 17, 279. Habashy, G. M., Colln Czech. Chem. Commun., 1960, 25, 3166. -, 2. anorg. Chew., 1960, 306, 312. Bishop, E., and Hitchcock, P. H., Analyst, 1973, 98, 465. Schreyer, R. C., J . Amer. Chem. S O ~ . , 1958, 80, 3483. Krishnan, V., and Patel, C. C., Chem. & Ind., 1961, 321. James, S. D., Electrochim. Acta, 1967, 12, 939. Bishop, E., Analyst, 1972, 97, 761. Bockris, J. O’M., and Huq, A. K. M. S., Proc. 22. SOC., A , 1956, 237, 277. Warner, T. B., Schuldiner, S., and Piersma, B. J., J . Electrochem. Soc., 1967, 114, 1120. Bishop, E., and Evans, N., Talanta, 1970, 17, 1125. East, G. A., and Bishop, E., Proc. SOC. Analyt. Chem., 1972, 9, 186. Essin, 0. A., Acta Phys.-chim. URSS, 1940, 13, 429. Khomutov, N. E., and Aboimov, M. A., Trudy Mosk. Khim.-tekhnol. Inst., 1961, 32, 156. Delahay, P., and Trachtenberg, I., J - Anzer. Chem. Soc., 1958, 80, 2094. Bishop, E., Chemia Analit., 1972, 17, 511. Karp, S., and Meites, L., J . Amer. Chem. SOC., 1962, 84, 906. NOTE-References 30, 31, 36 and 43 are to Parts V, VII, 111 and I of this series, respectively. J , Ibid., 1956, 28, 1412. -- Khim., 1968, 23, 73. , , Ibid., 1973, 98, 553. -- Received February lst, 1973 Accepted ALIarck 27th, 1973