Micellar Catalysis of Metal-Complex Formation Kinetics of the Reaction between Ni" and Pyridine-2-azo-p-dimethylaniline (PADA) in the Presence of Sodium Dodecylsulphate Micelles ; a Model System for the Study of Metal Ion Reactivity at Charged Interfaces. BY ALAN D. JAMES~ AND BRIAN H. ROBINSON" Chemical Laboratory, University of Kent at Canterbury, Kent CT2 7NU Received 20th December, 1976 The kinetics of the reaction between Ni2+(aq) and the bidentate ligand pyridine-2-azo-p-dimethyl- aniline have been investigated in sodium dodecylsulphate micellar solutions. Considerable catalysis effects (up to lo3) are observed on the rate of complexation. The rate dependence on pH and micelle concentration can be quantitatively explained, the former in terms of the surface potential of the micelle.The reaction occurs in the region of the micelle surface, over the detergent concerrtration range studied, and rate constants appropriate to a surface reaction are derived. The catalysis effect resides in the concentrative effect of the micelle surface on the reagents, since the rate of water loss from Niz+ at the surface of the micelle is little changed from that in bulk water. The effect of added electrolyte on the micelle-catalysed rate has also been investigated. . Much of the interest in micelle-catalysed reactions arises from structural similarities between micelles and globular enzymes, and to parallels between micellar and enzyma- tic catalysis. Although micellar catalysis of organic reactions in aqueous media' has been much studied in recent years,l* relatively little work has been reported involving inorganic reactions. In biological systems, ligand exchange with metal ions often occurs at interfaces, so that the reactivity of metal ions towards ligand exchange in such an environment and the effects of the interface on the overall rate processes are of fundamental importance from both a biochemical and physicochemical viewpoint. A convenient specific reaction for study is that between Ni" and the bidentate nitrogen ligand pyridine-2-azo-p-dimethylaniline (PADA), (1).This reaction has been previously studied in aqueous solution,3* aqueous-alcohol mixtures and in dipolar aprotic solvents.6 The rate-limiting step of the reaction in water is the loss of water from the inner coordination sphere of Ni2+, following rapid formation of an .outer-sphere complex, 3* (Subsequent ring-closure is generally rapid in aqueous solution.) The rate of loss of water is identified with the rate of solvent-exchange p Present address : Akzo Chemie U.K. Ltd, Stockpit Road, Kirkby Industrial Estate, Liverpool L33 7TH. 10A. D. JAMES AND B. H. ROBINSON 11 (ke3 as measured by n.m.r. methods. This mechanism is known as the Eigen- Wilkins mechani~m.~ N kex N fast outer-sphere k b inner-sphere N N Kos [Ni(H20),I2+ + N) + [(H20),NiOH2N)]2+ + [(H20)4Ni )I2+. complex slow complex For [Ni2+] 9 [PADA], and KO, Q [Ni'+]-', scheme (1) leads to eqn (2) kobs = Kosk,[Ni2+] + k b (2) where is the equilibrium constant for outer-sphere complex formation, ke, is the first-order rate constant for water loss, and kobs is the observed first-order rate con- stant for metal complex formation (= +).In this paper, we report factors which influence the kinetics of the reaction between the aquonickel ion and PADA in the presence of micellar sodium dodecyl (lauryl) sulphate. The reaction is related to the incorporation of metal ions into porphyrins, which has been studied both in micelles * and microemulsions.g EXPERIMENTAL Ni(NO&. 6Hz0, Na2HP04, NaH2P04 and NaOH were AnalaR grade. PADA was supplied by Sigma Chemicals and 4-dimethylanilinoazobenzene (dimethyl yellow, 2) by Hopkin and Williams. Sodium lauryl sulphate (SLS) was B.D.H. specially pure grade (> 99 %) and was used without further purification. Potassium polyvinylsulphate was supplied by Eastman Kodak, Triton X-100 (iso-octylphenoxypolyethoxylethanol containing approximately 10 mols of ethylene oxide) by B.D.H., sodium methylsulphate by Hopkin and Williams and tetraethylammonium chloride by Aldrich Chemicals.mol dm-3) and PADA solution mol dm-3) were mixed in a small-volume stopped-flow apparatus designed and built in this laboratory. Additional reagents (SLS, buffer, inert salt) were added to the PADA solution before mixing with the Ni2+ solution. Identical results were obtained when micellar solutions containing nickel were mixed with the same concentration of micelles containing PADA, i.e., the micelle concentration was not changed on mixing. However, when PADA was not equilibrated with micellar SLS before mixing with the Ni2+ solution, i.e., when the Ni2++ micelle solution was mixed with PADA, non- exponential traces were obtained.This is consistent with our earlier observations, which show that dye absorption can be a relatively slow process, requiring several hundred milli- seconds for completion.1° All the data reported here refer to solutions in which PADA was equilibrated with SLS before mixing with NiZ+ and where the final concentration of SLS after mixing is always greater than the critical micelle concentration (c.m.c.). First order rate constants (/cobs) or relaxation times (7) were obtained from changes in transmittance at 460 or 550 nm, and the values quoted are the average of five runs under identical conditions. When mi2+] > [PADA], good exponentials were obtained and the derived average first-order rate constants (7-l) are accurate to The pK, values of PADA and dimethyl yellow in the presence and absence of SLS micelles were obtained from spectrophotometric pH-titrations on a Unicam SP8000 spectro- photometer ; the accuracy is estimated as k0.05 units. Sodium hydrogen phosphate buffers were preferred to collidine or lutidine buffers, because the latter may bind hydrophobically to the micelle surface region and interact more strongly with Ni2+ in that state.Binding of phosphate to Ni2+ at the concentration of buffer em- ployed (5 x mol dm-3) is negligible. The pH of the mixed solution was checked using a Radiometer pH meter 26 after collecting the effluent from the stopped-flow drain tube. For the kinetic runs, equal volumes of Ni(N03)2 solution (> 5 x 5 %.RESULTS The visible absorption spectra of PADA, PADAH+, and NiPADA2+ are shown in fig. 1. There is a slight bathochromic shift ( - 3 nm) of the PADAH+ peak on binding12 MICELLAR CATALYSIS OF METAL COMPLEX FORMATION to the micelle, but the spectrum of PADA is hardly changed in the presence of SLS micelles. When Ni(NO& solutions were mixed with PADA solutions containing SLS at concentrations (after mixing) greater than the c.m.c., there was an exponential forma- tion of NiPADA2+, which could be detected from measurements at both 460 and 550nm. The magnitude and direction of the relaxation amplitudes at both wave- lengths were as expected from the spectra in fig. 1. Hence at high pH, there was a decrease in absorbance at 460 nm and an increase at 550 nm, whereas at low pH there was a decrease in absorbance at both wavelengths, with the amplitude at 460 nrn decreasing sharply as the pH was lowered.450 500 550 600 A/nm FIG. 1.-Spectra of PADA (A), PADAH+ (B) and Ni(PADA)2+ (C) in aqueous micellar solution. ([SLS] = 0.025 mol dm-3) In the absence of surfactant, the rate of reaction decreases as the pH is reduced from 8.0. At constant concentration of Ni2+(aq), the pH profile [fig. 2(a)] shows an inflexion around a pH of 4.5, which is close to the spectrophotometrically measured PKa of PADAH+ in bulk water. In the presence of micellar SLS, there is a consider- able rate enhancement [fig. 2(b),(c)]. The pH profiles are of similar appearance, but the inflexions occur at higher pH, in the region of 6.2-6.8, depending on the con- centration of SLS.The pKa of PADAH+ in 0.025moldm-3 SLS was measured independently by spectrophotometric titration and found to be 6.55, in agreement with that obtained kinetically (6.70) at 298.2 K. At fixed concentrations of Ni2+(aq) and pH, z-l (or kobs) decreased as the con- centration of micellar surfactant increased. The rate of reaction is, therefore, greatest just above the cmc of SLS. 7-l also increases in an approximately linear way with increase in pi2+] at a particular pH and SLS concentration. These effects are shown in fig. 3.A. D. JAMES A N D B. H. ROBINSON 13 Non-micelle-forming sodium methylsulphate (0.5 mol dm-3), which was added as a possible complexing agent for Ni2+, and potassium polyvinylsulphate, a linear polyelectrolyte, have little catalytic effect on the rate.Triton X-100, an uncharged surfactant with a hydroxyl terminal head-group causes a decrease in rate (table 1). b 1.5 1 .o 0.5 3 4 5 6 7 PH FIG. Z(u).-pH profile for the reaction in the absence of surfactant. [Ni2+] = mol dm-3, T = 298.2 K. [The solid line is a theoretical line calculated from eqn (lo)]. Addition of up to 0.1 mol dm-3 NaCl causes a modest decrease in the catalysed rate for an SLS concentration of 0.025 mol dm-3, but a much greater decrease is obtained with tetraethylammonium chloride (fig. 5) at the same surfactant concentration. An activation enthalpy of 52.3 (f4) kJ mol-' was obtained from the temperature dependence of 2-l over the range 20-40°C for the complex formation reaction at 30 20 10 4 5 6 7 8 9 150 50 FIG.P H 2(b),(c).-pH profile for the reaction in the presence of micelles. [Ni2+] = mol dm-3, T = 298.2 K. (b) = 0.025 rnol d r 3 SLS, (c) = 0.10 mol dm-j SLS.14 MICELLAR CATALYSIS OF METAL COMPLEX FORMATION [SLS] = 0.1 mol dm-3, [Ni2+] = mol dm-3, and pH = 7.5-8.5. (The c.m.c. of SLS does not change significantly over this temperature range.) This compares with a value of 55.2( & 4) kJ mol-1 measured in bulk solution for the same pro~ess.~ The equilibrium constant for complex formation ( K l ) in the presence of micelles at pH - 8 can be measured from the absorbance at 460 and 550 nm, and the total concentrations of Ni2+ and PADA. However, Kl is so high that, with the condition [SLS]/mol dm-3 FIG. 3.-Dependence of the reaction rate on [SLS].(T = 298.2 K.) (a) [Ni2+] = rnol dnr3, pH = 7.5-7.7; (b) [Ni2+] = 3.3 x rnol dm-3, pH = 7.3-7.5 ; (c) [Ni*+] = mol dm-’, pH = 8.0. [Ni2+] 9 [PADA], the absorbance changes are negligibly small in 1 cm optical cells. If we use similar concentrations of the two reactants, the results are best rationalised by allowing for formation of a bis-complex with equilibrium constant K2. A best fit to the equilibrium data is obtained when Kl - K2. Therefore, under the condi- tions of the kinetic runs, when mi2+] 9 [PADA], no bis-complex will be formed. TABLE 1 .-EFFECT OF ADDITIVES ON THE RATE OF COMPLEX FORMATION AT 298.2 K. [PADA] = 2 x mol dm-3 additive none 0.5 mol dm-3/sodium methyl sulphate 0.5 mol dm-3/sodium methyl sulphate 0.01 mol dm-3/polyvinylsulphate anion 0.01 mol dm-3/polyvinylsulphate anion 1 % w/w Triton X-100 1 % w/w Triton X-100 [NiZ+]/mol dm-3 10-3 10-3 5~ 10-3 10-3 5~ 10-3 10-3 5 x 10-3 T-l/S-* 1.4 1.1 5.2 0.6 3.6 0.25 0.9 Qualitatively, we find that the “ apparent ” equilibrium constant Kl ( = [NiPADA2+]/[Ni2+][PADA]) for the reaction in the presence of micelles decreases with increasing SLS concentration above the c.ni.c.tending towards the bulk value at high SLS concentrations. The equilibrium therefore provides a very sensitive indicator of the c.m.c. at - 8 x mol dm-3. Kl - lo5 dm3 mol-l when [SLS] = 0.2 molA. D. JAMES AND B. H. ROBINSON 15 dm-3, compared with a value of 1.3 x lo4 dm3 mol-l in the absence of surfactant (data at 298 IS). MECHANISM OF THE REACTION The pH profile [fig.2(a)] is consistent with the following kinetic scheme : k i Ni2++PADA + NiPADA2+ k- 1 k2 K,: Jr H+ (3) Ka Jt' H+ Ni2+ + PADAH+ + NiPADAH3+. k- 7 Proton-transfer is rapid in the presence of buffer. We also have the condition [N~"]T 9 [PADAIT, where T indicates the total or weighed-in concentration per dm3 of solvent. We define : [NiPADA2 '1 [H ' ] [NiPADAH3+] * and KL = [ PADA] [ H'] [ PADAH'] K , = - (4) Then the general relaxation expression ( 5 ) can be derived It is useful to derive experimental conditions. but is much less reactive from first principles the kinetic equations relevant to our Below pH 8.0, PADAH+ becomes kinetically significant, towards Ni2+ than PADA, so we have : kl k- 1 Ni2++PADA + NiPADA2+ PADAH+ Ka 11 H+ d[NiPADA2+]/dt = kl [Ni2+]CpADA] - k-1[NiPADA2+] [PADAIT = [PADA] + [PADAH+] + [NiPADA2+].(7) (8) (9) (10) (1 1) Therefore PADA] = ([PADA],- [NiPADA2+])/(1 + K: '[H+]). 7-l = [Ni2+lT(kl/(l +KL1[H+])} +k-,. 2-1 = k , [Ni2+IT + Ll. It follows, on integration for constant [Ni2+] and [H+], that : At high pH when [H+] < Ka, At low pH when [H+] 9 K,, the complete expression (5) must be used. Eqn (10) gives a reasonable fit to the kinetic data obtained in the absence of SLS [fig. 2(a)], with pKa = 4.5, k-l = 0.1 s-l and kl = 1.25 x lo3 dm3 mol-1 s-l. This latter value is in excellent agreement with the previously reported value of (1.35 kO.11) x lo3 dm3 mol-1 s - , . ~ The above kinetic treatment (5)-(10) can be also applied ta the reaction in the presence of micellar SLS, if appropriate values of the Ni2+ concentration and Ka are employed.In micellar solutions, there is an apparent increase in the PKa of PADA of - 2 units, as measured by spectrophotometric titration, because PADA and PADAHf16 MICELLAR CATALYSIS OF METAL COMPLEX FORMATION are located close to the surface of the micelle. The surface PKa of the ligand, (pK,),, is related to the bulk pK, by eqn (12), originally due to Hartley l2 At 298.2 K, (pKa)s-pKa = $159.2. (12) In eqn (12), II/ is the surface potential of the micelle in mV. Values of $, from eqn (12), for PADA and the structurally similar dye dimethyl yellow (2) in 0.025 mol dm-3 SLS are in good agreement. Values of $ obtained from pH-titrations of dimethyl TABLE 2.-vALUES OF “APPARENT” ACID DISSOCIATION CONSTANTS OF DIMETHYL YELLOW AND PADA IN MICELLAR SOLUTIONS.T = 298.2 K. [SLS]/mol dm-3 pL(dimethy1 yellow) PKs(PADA) - 3.35 - 4.5 0.025 5.42 6.55 0.125 5.46 6.59 0.050 5.15 6.28 a 0.10 5.02 6.15 a 0.20 4.85 5.98 a 0 Obtained by comparison with data for dimethyl yellow. yellow had already been measured l3 and were, therefore, used in eqn (12) to derive (PK,)~ values for PADA in the presence of various concentrations of SLS and additives (tables 2 and 3). Values of $ obtained using dimethyl yellow are in satisfactory agreement with those obtained from electrophoretic measurements l4 (table 3) ; this strongly suggests that the major influence on (PKa), is due to the increased hydrogen ion concentration close to the micelle surface, rather than an effect on the intrinsic TABLE 3 . v A L U E S OF “HE SURFACE POTENTIAL AND AREA PER HEAD GROUP AT DIFFERENT SALT CONCENTRATIONS ~aCl]/mol dm-3 ~t4NCl]/mol dm-3 w/(mV)O tp/(mV)b .4/10-16 cm2 C 0 0 0.025 0.050 0.075 0.10 0 0.0125 0.025 0.050 0.10 124 149 61 96 99 58 89 88 55(0.5) 86 82 54(0.2) 84 80 53 63 27 12 6 a From eqn (12) using @K& of dimethyl yellow.b From the electrophoretic data in ref. (14) by interpolation. C Area per head group from ref. (20) ; head group area in closed packed monolayer = 45 x cm2 ref. (21). PKa of the dye, arising from the change in en~ir0nment.l~ In the absence of added inert salt, [H+Is N 10O[H+lb, where [H+], refers to the hydrogen ion concentration at the micelle surface. For the reaction in the presence of micelles, (pK,), must be used in eqn (5) and (10), where [H+] in these equations still refers to the bulk concentration (measured using a glass electrode).The use of these values of ( p Q S is consistent with the inflexion point determined from the plot of 7-l against pH in the presence of micelles [fig 2(b) and (c)].A. I). JAMES AND B. H. ROBINSON 17 The other important consideration in micellar systems is how to express [Ni2+IT. We suppose, in the absence of added NaCl, that all the Ni2+ is contained within the surface region of the micelle, predominantly in the outer Helmholtz layer (Stern layer), but with an intact solvation shell, i.e., within a distance of N lo-’ cm from the surface. It is, therefore, the surface concentration of Ni2+ which is kinetic- ally important. A simple way to express this concentration is by the dimensionless quantity [Ni2+IT/(C-c.m.c.) where Cis the weighed-in SLS concentration and c.m.c.= 8 x mol dm-3 at 25°C in the absence of added electrolyte.16 (This expression is obviously inapplicable for SLS concentrations close to the c.m.c.) Then it follows that, as the micelle concentration is increased, the nickel ion will be diluted on the surface, and the variation of 2-l with surfactant concentration can be explained (fig. 3). 150 100 vl ?, L \ $4 50 FIG. 4.-PIot of T- I / I 0.02 0.04 Y 0.06 against [Ni2+]~/(C-c.m.c.){ 1 + (K&-’[H+ }T= Eqn (10) then becomes (13), for the micelle-catalysed reaction. 2-l = k’lCNi”’I,/((C-c.m.c.)(I +(K,)F ‘[H+])) + k’_ 1. 98.2 K. Fig. 4 shows a plot of 2-l against [Ni2+],/(C-c.m.c.){l +(Ka)F1[H+J)) over a wide range of H+, Ni2+ and SLS concentrations, from which we find ki = 3000 S-l.The fact that a reasonable linear plot is obtained is excellent evidence in support of the proposed mechanism. From the dependence of r1 on mi2+] at low pH, and using the value of k i obtained at high pH, k$[= k,/(C-c.m.c.)] can be obtained as 11 s-l. In principle, kL1, the rate constant for the reverse reaction on the micelle surface, could be obtained as the intercept as [Ni2+] 3 0 at high pH [eqn (1 l)], but in practice the reaction goes to completion when [Ni2+] % [PADA]. An intercept of 0.4 s-l was obtained at pH 4.2 but it may contain contributions from both kLl and kL2, depending on the value of (pK& [eqn (5)]. The mechanism predicts that a limiting rate will be reached when the surface becomes saturated with nickel, since no diffusion is then required for reaction.The kinetics are then determined by some first-order process which might be water loss from the inner coordination sphere of nickel (kex), or ring-closure to form the bidentate complex, if this were sterically-hindered on the surface. Alternatively the slowest18 MICELLAR CATALYSIS OF METAL COMPLEX FORMATION process might be diffusion of PADA from an unreactive position in the micelle interior to the micelle surface. A refinement of eqn (13) is to express the surface concentration of Ni2+ in terms of the restricted volume available to the ion close to the micelle surface. If the ion is confined to within a distance r cm from the surface (but in the solvent) and A is the surface area (cm2)/surfactant head group at the surface : molar.[Ni2'IT. 1000 (C - c.m.c.)N,,Ar [Ni2'Is = ~ Thus we might expect that the rate would be affected by changes in structural pro- perties of the micelle, e.g., by the effect of temperature and ionic strength on the head group area. DISCUSSION The plot shown in fig. 4 provides good evidence that the reaction occurs in the region of the micelle surface. It is clearly the surface concentrations of H+, Ni2+ and PADA which are kinetically important. A similar value of z-l is obtained at a particular value of [Ni2+]/(C-c.m.c.) at both low (0.0125 mol dm-3) and high (0.1 mol dm-3) SLS concentrations, which implies that all the Ni2+ (and PADA) is bound to the micelle. Otherwise, as the SLS concentration is increased, more binding of Ni2+ would result, resulting in an increase in 7-l at a particular value of [Ni2+]/ (C-c.m.c.).This conclusion is supported by n.m.r. studies on Mn2+ binding to SLS micelle~,~~ and electrophoretic studies on SLS in the presence of various divalent metal i0ns.I The enhanced rate of reaction between Ni2+ and PADA in the presence of micellar SLS may be due to : (i) An effect on the frequency of water loss (kex) from the inner coordination sphere of Ni2+. (ii) The concentrative effect of the micelle on the reactants. (iii) An increased interaction in the outer-sphere complex on the surface of the micelle. By analogy with the behaviour of Mn2+,17 ke, is not expected to be greatly en- hanced by adsorption on the micelle surface. The similarity of the measured AH:, values in the presence and absence of SLS also supports this conclusion. Data available for rates of water exchange about divalent metal ions l9 show that the variation in rate is controlled by AH;.Although k,, may vary from 90 to 8.0 x lo9 s-l on going from V2+ to Cr2+, the factor affecting the rate of exchange is always AH,',, AS& being more or less constant and close to zero. For Ni2+(aq), k,, = 3.0 x lo4 s-l, AH& = 46.8 kJ mol-1 and AS,", = 4.2 J K-l mol-l. Since water loss is rate limiting in aqueous solution (rather than ring-closure), the similarities in the activation enthalpies suggest that water loss is also rate limiting on the surface. From eqn (13) and (14), we can derive [expression (15)] a second-order rate con- stant kl for the surface reaction which is directly comparable with that in bulk solution.kl = k;N,Jr. Using eqn (15) and taking ki = 3000 s-l, A = 60 x cm2 2o (table 3) and r = cm, we obtain kl = 1.1 x lo3 dm3 mol-1 s-1 for complex formation on the surface, similar to that measured in bulk solution of 1.25 x lo3 dm3 mol-1 s-l. This result very strongly suggests that catalysis is largely due to a concentrative effect of the micelle surface. Unless (k& and (KO& change in a compensatory way (unlikely for unrelated processes), we must conclude that the water exchange rate for nickel (II)A. D. JAMES AND B. H. ROBINSON 19 at the micelle surface has a similar value to that in the bulk solvent. At constant [Ni2+IT, therefore, we can write : total volume of reaction medium volume available for Ni2+ at the surface' - - (7- l)SLS ( z - ~ ) H ~ O Since it has now been established that the reaction is not catalysed by lowering of the activation energy barrier for the reaction, but only by concentrating the reagents in a small part of the total reaction volume in the region of the micelle surface, it is convenient to express the rate constant for the complexation reaction ki in the manner applicable to a surface reaction.The surface concentration of Ni2+ expressed as (moles of Ni2+)/(area of micelle surface available) is given by : [Ni2+lB [Ni2']: = mol cm-2 (C - c.m.c.)BAN,v where B refers to bulk concentrations in mol (dm3 of solution)-l. solvent, and surface concentrations are used in defining K& ; Then, for [Hf] < (Ka)s, using eqn (13) and (15): For the reaction at the surface, if the same mechanism operates as in the bulk (17) k; = KLsk:x cm2 mol-' s - k; = k;N,,A 1.1 x 1013 cm2 mol-l s-l.For outer-sphere complex formation between Ni2+ and PADA on the surface of the micelle, assuming that the PADA is tightly bound to the micelle, we can easily calculate (by analogy with the Fuoss Equation) that K& = (4nNava2/2) exp ( -AGint/RT) cm2 mol-1 where a is the distance of closest approach of the reactants on the surface (Ni2+ solvated) and AGint represents the difference in free energy between the nickel ion interacting with a surface site containing PADA compared with a surface made up of negatively-charged sulphate head groups. If we assume a = 4 x lo-' cm, and AGint = +4.0 kJ mol-l (since the outersphere site is less favoured on coulombic arguments) then Kis = 1.2 x lo9 cm2 mol-' and kix - lo4 s-' which is again in very close agreement with that measured in bulk water. This general conclusion might be expected since the loss of water is controlled more by ion-dipole and ligand-field interactions than by structural effects in the solvent near to the mimlle surface.Further experiments are now being carried out to determine to what extent this conclusion can be generalised. The effect of electrolyte (NaCl and (C,H,),NCI) is to cause a decrease in the catalysed rate (fig. 5). Addition of NaCl decreases the surface potential and changes the shape of the micelle from spherical to ellipsoidal so that the surface area per head group (A) is decreased (table 3).20 The decrease in $ (due to an increase in the extent of Na+ binding) will lead to desorption of Ni2+, resulting in a decrease in the rate of reaction as observed.However, the accompanying decrease in the surface area per head group, (table 3), would lead to an increase in 7-l of up to 15 %. The effect of NaCl on 2-l is a balance of these opposing factors.20 MICELLAR CATALYSIS OF METAL COMPLEX FORMATION The addition of (C2H&NC1 causes a much larger decrease in the rate of the catalysed reaction than NaCl. (C2H&NC1 also has a much larger effect on the sur- face potential, (table 3), which implies that desorption of the nickel ion is the major factor influencing the rate. 150 7 100 z?. -I I 50 0 \1 ' P 0'8, -.- - '. 1 I I I 0.02 0.04 0.06 0.08 [added salt]/rnol dm-3 FIG.5.-Effect of salt concentration on T-l. [Niz+] = loe3 moI dm-3, [SLS] = 0.025 mol dm-3, T = 298.2 K. ( l a ) Added NaCl; (16) addded NaCl corrected for NiZ+ dilution on the surface arising from a decrease in cmc. (2) Added (C2H5)4NC1 ; (26) added (CZH5)+NCI, corrected for the decrease in c.m.c. The lack of catalysis shown by sodium methylsulphate confirms that alkyl sulphates have no catalytic abilityper se, but that micelles are necessary. The failure of poly- vinylsulphate and Triton X-100 to catalyse the reaction shows that a linear charged surface or a hydrophobic environment are not alone able to concentrate the Ni2+ and PADA reactants. The small decrease in rate caused by addition of these reagents probably reflects a partitioning of one reagent (Ni2+ in the case of polyvinylsulphate, and PADA with Triton X-100) but not the other, out of the bulk solvent.When two ionic species are used as reactants, as in the case of the reaction between Co(NH&C12+ and Hg2+ in the presence of anionic micelles and polyvinylsulphonate,22 large catalytic effects have been observed in both cases. We are indebted to Mr. K. J. A. Hargreaves, who carried out preliminary measure- ments on this system. We thank the S.R.C. for a fellowship (to A. D. J.) and grants for equipment associated with this work. We acknowledge also valuable discussions with Dr. W. Knoche and his colleagues (by a NATO Travel Grant), Dr. J. Holzwarth and Dr. W. J. Albery. E. H. Cordes, Reaction IGnetics in Micelles (Plenum Press, New York, 1973). J. H. and E. J. Fendler, Catalysis in Micellar and Macromolecular Systems (Academic Press, New York, 1975). R. G. Wilkins, Inorg. Cheni., 1964, 3, 520. M. A. Cobb and D. N. Hague, J.C.S. Faruhy I, 1972, 68,932.A. D. JAMES AND B. H. ROBINSON 21 E. F. Caldin and P. Godfrey, J.C.S. Faraday I, 1974, 70, 2260. H. P. Bemetto and Z. Sabet Imani, J.C.S. Faraday I, 1975,71, 1143. Series, No. 49, (Amer. Chem. SOC., Washington, D.C., 1965), p. 55. K. Letts and R. A. MacKay, Inorg. Chem., 1975,14,2990. ’ M. Eigen and R. G. Wilkins, Mechanisms of Inorganic Reactions, ed. R. F. Gould, Adv. Chem. * M. B. Lowe and L. N. Phillips, Nature, 1961,190,202. lo B. H. Robinson, N. C. White and C. Mateo, Adv. Molecular Relaxation Processes, 1975,7,321. l1 M. Eigen and L. DeMaeyer, Technique of Organic Chemistry, ed. S . L. Friess, E. S. Lewis and l2 G. S. Hartley and J. W. Roe, Trans. Faraday SOC., 1940,36, 107. l3 A. D. James, B. H. Robinson and W. Knoche, to be published. l4 D. Stigter, J. Colloid Interface Sci., 1967, 23, 379. P. Mukerjee and K. Banerjee, J. Phys. Chem., 1964, 68,3567. l6 P. Mukerjee and K. J. Mysels, Critical Micelle Concentrations of Aqueous Surfactant Systems (National Bureau of Standards, Washington, 1971). J. Oakes, J.C.S. Faradhy 11, 1973, 69, 1321. A. Weissberger (Interscience, New York, 1963), vol. 8, part 2, p. 913. l8 G. S. Hartley and C. S. Samis, Trans. Faraday Soc., 1938,34, 1288. l9 H. P. Bennetto and E. F. Caldin, J. Chem. SOC. A,, 1971, 2200. 2o D. Stigter, J. Colloid Interface Sci., 1974, 47,473. 21 F. Reiss-Husson and V. Luzzati, J. Phys. Chern., 1964,68, 3504. 22 J.-R. Cho and H. Morawetz, J. Amer. Chem. SOC., 1972, 94, 375. (PAPER 6/2303)