Liquid Interface as a Model System for Studying

M. Cristina Lagunas , Debbie S. Silvester , Leigh Aldous , Richard G. Compton ... Trevor J. Davies , A. Christopher Garner , Stephen G. Davies , Richa...
0 downloads 0 Views 199KB Size
Download PDF
3202

J. Phys. Chem. B 1997, 101, 3202-3208

Liquid/Liquid Interface as a Model System for Studying Electrochemical Catalysis in Microemulsions. Reduction of trans-1,2-Dibromocyclohexane with Vitamin B12 Yuanhua Shao and Michael V. Mirkin* Department of Chemistry and Biochemistry, Queens College-CUNY, Flushing, New York 11367

James F. Rusling* Department of Chemistry (U-60), UniVersity of Connecticut, Storrs, Connecticut 06269-4060 ReceiVed: January 16, 1997X

A complex electrochemical catalytic reaction was investigated at the interface between water and benzonitrile as a model for interfacial chemistry in microemulsions. Structures similar to those between oil-water microphases in microemulsions were created by using the interface between two immiscible electrolyte solutions (ITIES). While interfacial area in a microemulsion can be uncertain, the ITIES is well-defined and can be used to evaluate relevant heterogeneous interfacial kinetics. The reaction between the Co(I) form of vitamin B12 generated electrochemically in the water phase, and trans-1,2-dibromocyclohexane (DBCH) in benzonitrile was probed directly at the ITIES by scanning electrochemical microscopy (SECM). Apparent heterogeneous rate constants for the interfacial reaction were extracted from SECM current-distance curves. Influences of reactant concentration, potential drop across the ITIES, and adsorbed surfactants were investigated. Results suggest that the kinetics of reduction of DBCH by B12 Co(I) are more complex at a liquid/liquid interface than that of the simple second-order rate-limiting process in a homogeneous organic solvent.

Introduction The use of novel solvents can significantly broaden the diapason of chemical systems and processes accessible to electrochemical measurements. For example, the introduction of nonaqueous solvents and later supercritical fluids has provided a great expansion of potential windows of working electrodes. Electrochemistry in microemulsions, i.e., microheterogeneous mixtures of oil, water, and surfactant, appears attractive for electrochemical synthesis and other applications.1 Two obvious advantages of microemulsions over conventional nonaqueous solvents include lower toxicity and the possibility of conducting reactions involving species of very different polarities. The latter advantage can be fully realized in bicontinuous microemulsions in which oil and water are present as continuous microphases separated by a monolayer of surfactant.2 The polar and nonpolar reactants present in two phases may be intimately mixed, and often the apparent rate of the interfacial reaction is not greatly limited by phase transfer.3 Electrochemical catalysis and syntheses in bicontinuous microemulsions have been reported.3,4 A typical problem in kinetic studies of electrochemical processes in microemulsions is a poorly defined interfacial area. This parameter is, however, important because the rate of a heterogeneous reaction is proportional to it. In this paper, the organic solvent/water interface is employed as a model experimental system for studies of electrochemical catalysis in microemulsions. The interface between water and organic solvent with an adsorbed surfactant layer should possess essentially the same structure as a bicontinuous microemulsion prepared from the same components. But, unlike an emulsion, the area of the interface between two immiscible electrolyte solutions (ITIES) is well-known and the potential drop across the interface, ∆owφ, can be quantitatively controlled and varied by changing the ratio of concentrations of the potentialdetermining ion in the two liquid phases.5,6 X

Abstract published in AdVance ACS Abstracts, March 15, 1997.

S1089-5647(97)00230-7 CCC: $14.00

It was shown earlier6,7 that kinetics of the charge transfer at the ITIES can be probed directly by scanning electrochemical microscopy (SECM).8 In a typical SECM/ITIES experiment, a tip ultramicroelectrode (UME) with a radius a is placed in an upper liquid layer (e.g., aqueous phase) containing the oxidized form of the redox species, O1. When the tip is held at a negative potential, O1 reacts at the tip surface to produce the reduced form of the species, R1. When the tip approaches the ITIES, the mediator can be regenerated at the interface via the bimolecular redox reaction between R1 in the aqueous phase (w) and O2 in the organic phase (o) k

R1(w) + O2(o) 98 O1(w) + R2(o)

(1)

and the tip current, iT, increases with a decrease in the tipITIES separation, d (positive feedback). The kinetics of reaction 1 can be evaluated from the tip current-distance (or approach) curve. If no regeneration of O1 occurs, the ITIES blocks mediator diffusion to the tip, so iT decreases at smaller d (negative feedback). While conventional studies of the ITIES have been carried out at externally biased polarizable ITIES, in SECM measurements, a nonpolarizable ITIES is poised by the concentrations of the potential-determining ions, providing a constant driving force for the electron transfer (ET) process. This setup essentially eliminates such experimental problems as iR-drop and charging current and allows quantitative studies of the kinetics and mechanism of the interfacial reaction. The work described herein is concerned with catalytic reduction of trans-1,2-dibromocyclohexane (DBCH) by B12s (the Co(I) form of vitamin B12). This process has been studied in various media, and its apparent rate in bicontinuous microemulsions was comparable to those in homogeneous solvents.3a,9,10 A schematic diagram of the SECM experiment is given in Figure 1. Processes occurring at the tip and the ITIES under our experimental conditions using a pH 2.5 water phase are shown in Scheme 1.3a,9,10 © 1997 American Chemical Society

Reduction of DBCH with Vitamin B12

J. Phys. Chem. B, Vol. 101, No. 16, 1997 3203

Figure 1. Measurement of the kinetics of the interfacial reaction between DBCH in benzonitrile and Co(I)L in water with the SECM operating in the feedback mode. Electroneutrality was maintained by transfer of perchlorate ions across the interface.

SCHEME 1 [Co(II)L-H]+ + e- a Co(I)L-H

(tip)

(2)

2Co(I)L-H + RBr2 f 2[Co(II)L-H]+ + 2Br- + olefin (ITIES) (3a) or alternatively, eq 1 followed by Co(I)L-H + RBr2 f BrCo(III)L + Br- + olefin + H+ (ITIES) (3b) At this pH, vitamin B12r (i.e., the Co(II) form of vitamin B12) has a protonated benzimidazole side chain, and we show it as [Co(II)L-H]+. This species is reduced at the tip electrode at a diffusion-controlled rate. The neutral product Co(I)L-H formed at the electrode reacts with DBCH (RBr2) by either a multistep radical process (eq 3a)3,9,10 or a concerted E2 elimination as in eq 3b. These two processes are practically indistinguishable from a kinetic point of view. Possible ratedetermining steps (rds) in both alternatives feature the bimolecular reaction of Co(I)L-H with RBr2. The elementary rds is as shown in eq 3b. For eq 3a it is thought to be

Co(I)L-H + RBr2 f [Co(II)L-H]+ + RBr• + Br- (3c) Reactions 2 and 3 together represent electrochemical catalysis of the reduction of DBCH, with a second (or pseudo-first) order chemical reaction following electrode reaction 2. When such a process is studied by cyclic voltammetry, a fairly complicated analysis based on digital simulation is required to extract kinetic parameters from experimental curves. In contrast, under SECM conditions, reaction 3 is mathematically equivalent to kinetically controlled regeneration of the mediator at a solid substrate and can be characterized by the value of an effective heterogeneous rate constant. This greatly simplifies the analysis of the SECM data and provides a viable model for the same interfacial reaction in microemulsions. Reaction 3 injects negative charge into the organic phase that is compensated in our system by ion transfer (IT) of ClO4- from benzonitrile to water. At low concentrations of this common ion (e.g., less than 10 mM),7 the IT may become the ratedetermining step. To avoid this complication, the concentration of ClO4- in benzonitrile was kept high (0.1 M) in all kinetic experiments. Unlike previous SECM studies at the ITIES6,7 where the regeneration of a redox mediator occurred via simple outer sphere bimolecular ET, reaction 3 is a complex, inner sphere multistep process.3,9,10 Thus, additional analysis was carried out to investigate the nature of the rate-limiting step at the ITIES.

Figure 2. Voltammograms of vitamin B12 at an 11-µm-diameter carbon microdisk obtained before (A) and after (B) pre-electrolysis. Solution contained 1 mM vitamin B12, 0.1 M NaClO4, and 0.2 M phosphate buffer. Sweep rate was 100 mV/s. Both solutions were degassed with nitrogen before measurements.

Experimental Section Chemicals. Benzonitrile (99.9%, HPLC grade) and (()trans-1,2-dibromocyclohexane (DBCH, 99%) were from Aldrich. Tetra-n-hexylammonium perchlorate (THAClO4, from Alfa, West Hill, MA) was recrystallized twice by using 1:9 (v: v) ether/ethyl acetate before use. All other chemicals were ACS reagent grade. A pH 2.5 phosphate buffer was prepared with the ionic strength of 0.2 M. At this pH, the Co(II) complex exists in aqueous solution in its “base-off” form, which can be reversibly reduced.10,11 All aqueous solutions were prepared from deionized water (Milli-Q, Millipore Corp.). All organic solutions were freshly prepared and purged by Argon for ca. 30 min before measurements. Vitamin B12a obtained from Sigma Chemical Co. as hydroxocob(III)alamin hydrochloride (crystalline, ca. 98%), which was a complex of Co(III). This species can undergo two separate one-electron reductions to produce Co(I) to react with DBCH. However, the complications caused by the presence of three different Co complexes in the system and the incompletely reversible electrochemistry of the Co(III)/Co(II) couple forced us to prereduce vitamin B12a before SECM experiments. The bulk electrolysis was carried out at constant potential (E ) -0.5 V vs Ag/AgCl) using a BAS 100B electrochemical workstation (Bioanalytical Systems, West Lafayette, IN). A three-electrode cell was placed inside the nitrogen-filled glovebox. Two spectroscopic carbon rods (Ultracarbon Co.) separated by an agar salt bridge filled with buffer solution served as a cathode and an anode, and a Ag/AgCl electrode was used as a reference. The continuous stream of nitrogen was passed through the cell during the electrolysis in order to remove traces of oxygen and provide adequate stirring. The electrolysis was stopped after about 1-2 h, when the current decreased to less than 5% of the initial value, and the solution containing vitamin B12r was transferred directly to the SECM cell inside the glovebox. Figure 2 shows two cyclic voltammograms of vitamin B12 obtained before and after pre-electrolysis at an 11-µm-diameter

3204 J. Phys. Chem. B, Vol. 101, No. 16, 1997

Shao et al. tip was biased at a potential where the reduction of vitamin B12r was diffusion-controlled. The approach curves were obtained by moving the tip toward the ITIES and recording current (iT) as a function of the distance between the tip and interface (d). All experiments were conducted under dimmedlight conditions to avoid photocleavage of in-situ-generated alkyl-cobalt complexes.3a The coordinate of the ITIES (d ) 0) was determined from the sharp increase (or decrease, if the bottom phase contained no redox species) in tip current that occurred when the tip touched the ITIES. Results and Discussion

Figure 3. Schematic diagram of the SECM cell. The use of the sessile drop-shaped bottom layer allows one to form a stable ITIES when the density of the organic solvent is less than or equal to the density of aqueous solution.

carbon tip. As expected, the wave corresponding to the reduction of vitamin B12a to vitamin B12r disappeared after preelectrolysis leaving only the B12r f B12s (i.e., Co(II)L f Co(II)L) wave with E1/2 close to -0.74 V vs SCE reported for pH 2.5.11 Electrodes and Electrochemical Cells. SECM tips (11-µm diameter) were prepared by sealing carbon fibers (11-µm diameter, Amoco Performance Products, Greenville, SC) in glass as described previously12 and polished with 0.05-µm alumina on felt (Buehler, Ltd., Lake Bluff, IL) before each experiment. The arrangement of the electrochemical cell used is shown in Figure 3. A three-electrode setup was employed with a Pt wire (0.5-mm diameter) as the counter electrode and a Ag/AgCl reference electrode. All electrodes were placed in the aqueous solution (top layer). It was surprisingly hard to find an organic solvent for these experiments. The density of a suitable solvent should be higher than that of water (in order for it to form the bottom layer in the cell), and it should not react with the strongly reducing Co(I)L species. One of a few candidates was benzonitrile (BN). However, very similar densities of BN (1.01 g/cm3 at 25 °C) and water made the ITIES very unstable. To overcome this problem, the organic phase was placed inside a glass tube of 1-mm inner and 5-mm outer diameter (Figure 3). A sessile drop of BN formed at the end of the tube and surrounded by aqueous solution was used as the substrate for SECM measurements. Before measurements, the tip electrode was positioned above the highest point of the drop. This structure was found to be stable even with the density of a solvent inside the tube somewhat lower that the density of surrounding solution. This cell was mounted on a vibrationfree table (Newport Corp., Fountain Valley, CA) and shielded in a Faraday cage. The top phase contained 0.1 M NaClO4 and 1 mM vitamin B12r in the pH 2.5 phosphate buffer, and the bottom phase contained 0.1 M THAClO4 and 1-100 mM DBCH benzonitrile solution. Perchlorate was the only ion common to both phases for all experiments. The ratio of bulk concentrations of ClO4in aqueous and organic phase, [ClO4-]w/[ClO4-]o, determined the potential drop across the ITIES. SECM Apparatus and Procedure. The SECM apparatus was similar to that described previously13 except for a CE-6000 controller (Burleigh Instruments, Fischers, NY) used to control three inchworm motors and a newer version of software generously provided by Prof. D. O. Wipf (Mississippi State University). The glovebox containing the SECM assembly was purged with purified nitrogen before measurements. The x and y axis inchworm motors were used to position the tip electrode in the aqueous phase above the top of the organic drop. The

Vitamin B12 as an SECM Mediator. Unlike previous SECM studies of the ITIES where redox mediators were simple one-electron redox couples, the electrochemistry of vitamin B12 is rather complicated.11 From Figure 2, one can see that a wellshaped steady state voltammogram for reduction of vitamin B12r ([Co(II)L-H]+) can be obtained only after nearly complete removal of oxygen and pre-electrolysis of the Co(III) form. Using an equation for the tip current at an infinite-substrate separation,

iT,∞ ) 4nFaDc

(4)

where c and D are the concentration and diffusion coefficient of the electroactive species, one can calculate the diffusion coefficient of vitamin B12r in aqueous solution of 2 × 10-6 cm2/ s. This is consistent with previously published values.9-11 The next step was to check that the SECM approach curves obtained with Co(II)L/Co(I)L are in agreement with the theory for a reversible one-electron mediator. Such curves were obtained using a solid substrate (Figure 4, curves A and B). With the positively biased glassy-carbon substrate, the reoxidation of Co(I) produced a high positive feedback (curve 4A). In contrast, no regeneration of the mediator occurred at an insulating mica substrate (curve 4B). Both curves are in good agreement with the theory14 (solid lines). When the ITIES served as the substrate, the diffusion-controlled positive feedback current was observed with pure DBCH used as the bottom phase (curve 4C) and a pure negative feedback with BN containing no DBCH (curve 4D). The last result shows that the reaction between Co(I)L and BN is negligibly slow and no appreciable adsorption of Co(I)L occurs at the ITIES under our experimental conditions. Heterogeneous Kinetics at the ITIES. The overall process shown in Figure 1 consists of four stages:7 mediator diffusion between the tip and the ITIES, interfacial reaction 3a or b, diffusion of DBCH in BN, and charge compensation by IT. In principle, any of these stages can be rate limiting, but in our experiments the concentration of ClO4- in BN was always sufficiently high to exclude the possibility of ion transfer limitations. From Figure 5 one can see that the normalized tip current is significantly lower than the diffusion limit given the upper dashed curve when the concentration of DBCH in BN (cDBCH) is less than about 20 mM. Thus, for lower values of cDBCH the process is limited either by the rate of the interfacial reaction or by transport of DBCH in BN. The latter possibility can be ruled out. The lower limit for diffusion-limiting current of DBCH in BN is7

id g 4nFaDDBCHcDBCH

(5)

where DDBCH is the diffusion coefficient of DBCH in BN. The diffusion coefficient of DBCH in DMF (DDBCH,DMF) is 2 × 10-5 cm2/s.3a One can estimate the diffusion coefficient of DBCH in BN using Walden’s rule:

Reduction of DBCH with Vitamin B12

J. Phys. Chem. B, Vol. 101, No. 16, 1997 3205

Figure 5. SECM current-distance curves for an 11-µm-diameter C tip UME in aqueous solution approaching the water/benzonitrile interface. The aqueous solution contained 1 mM vitamin B12r, 0.1 M NaClO4, and 0.2 M phosphate buffer. The tip potential was -0.9 V vs Ag/AgCl. The BN solution contained 0.1 M THAClO4 and (1) 2, (2) 5, (3) 10, or (4) 20 mM DBCH. Squares are experimental points; the solid line is a theoretical fit obtained from eqs 8a and 8b. See Figure 6 for rate constant values. The dashed curves represent theory for pure positive (upper curve) and pure negative (lower curve) feedback.

Since in all experiments discussed below cDBCH g cB12, the diffusion transport of DBCH in BN was not rate limiting, and one should presume that redox reaction 3 controlled the rate of the overall process. This conclusion is consistent with a marked increase in the normalized tip current at higher concentrations of DBCH in BN (Figure 5). The first-order effective heterogeneous rate constant for this reaction can be extracted from the iT-d curves using eqs 8:7

ITk ) ISk(1 - ITins/ITc) + ITins

(8a)

ISk ) 0.78377/L(1 + 1/Λ) + [0.68 + 0.3315 exp(-1.0672/L)]/[1 + F(L,Λ)] (8b)

Figure 4. Diffusion-controlled SECM approach curves obtained with vitamin B12r serving as a mediator. The tip potential was held at -0.8 to -1.0 V vs Ag/AgCl, corresponding to the plateau current of vitamin B12r reduction. The aqueous solution contained 1 mM vitamin B12r, 0.1 M NaClO4, and 0.2 M phosphate buffer. (A) The glassy-carbon substrate was biased at +0.5 V vs Ag/AgCl. (B) Mica served as an insulating substrate. The organic phase was (C) 1.25 mM THAClO4 in DBCH and (D) 0.1 M THAClO4 in BN. The tip was scanned at 0.5 µm/s. Solid curves in each picture represent the theory for diffusioncontrolled positive (A and C) and negative (B and D) feedback.14

DDBCHηBN ) DDBCH,DMFηDMF

(6)

where ηBN ) 1.2 cP and ηDMF ) 0.8 cP are viscosities of BN and DMF at 25 °C, respectively.15 The substitution of these values into eq 6 yields DDBCH ) 1.33 × 10-5 cm2/s. Thus, the ratio of id to iT,∞ is

id/iT,∞ g DDBCHcDBCH/DB12cB12 ) 6.5cDBCH/cB12

(7)

where DB12 and cB12 are the diffusion coefficient and concentration of vitamin B12 in aqueous solution, respectively.

where ITc, ITk, and ITins represent the normalized tip currents for diffusion-controlled regeneration of a redox mediator, finite substrate kinetics, and insulating substrate (i.e., no mediator regeneration), respectively, at a normalized tip-substrate separation, L ) d/a. ISk is the kinetically controlled substrate current; Λ ) kfd/D, where kf is the apparent heterogeneous rate constant (cm/s), and D is the diffusion coefficient of the reduced mediator in the top phase; and F(L,Λ) ) (11 + 7.3Λ)/Λ/(110 - 40L). These currents are normalized by the tip current at an infinite tip-substrate separation, iT,∞. The values of ITc and ITins are tabulated for a wide range of L,14 and analytical approximations are also available.7 Good agreement between theory (solid line) and experimental data (symbols) was achieved for all curves in Figure 5 using only one adjustable parameter, Λ. We calculated the effective heterogeneous rate constant, kf ) ΛD/d, for different values of cDBCH. The linear dependence of kf vs cDBCH (Figure 6) is similar to those measured previously for two cases of simple interfacial ET at the ITIES.6,7 The deviations from linearity are observed at cDBCH g 20 mM as kf approaches the diffusion limit. From the slope of the straight line in Figure 6 one can calculate the effective bimolecular rate constant, k ) kf/cDBCH ) 3.0 M-1 cm s-1. The comparison of this value to that measured previously by cyclic voltammetry in a bicontinuous microemulsion is not straightforward. The latter quantity, 2.4 × 105, was expressed in M-1 s-1 and considered to be an effective homogeneous rate constant.3a This allowed easy comparison of the rates in microemulsions and homogeneous solvents. However, despite very intimate mixing of components

3206 J. Phys. Chem. B, Vol. 101, No. 16, 1997

Shao et al.

∆owφ ) ∆owφ°ClO4- - 0.059 log

Figure 6. Dependence of the effective heterogeneous rate constant on concentration of DBCH in BN. The kf values were used to fit the approach curves in Figure 5 with a ) 5.5 µm and DB12 ) 2.0 × 10-6 cm2/s. Concentration dependence of kf levels off at higher cDBCH because the rate of the process approaches the diffusion limit.

in a bicontinuous microemulsion, the catalytic reaction remains heterogeneous with polar and nonpolar reactants residing in different phases. A conventional way to compare homogeneous and heterogeneous rates is by multiplying the latter by the thickness of the reaction layer.16 For a sharp interface this thickness is often assumed to be 10-8 cm.16 With this assumption, the value of k appears to be about 3 orders of magnitude higher that the previously measured rate constant. There are several possible reasons for this discrepancy: (i) In a microheterogeneous microemulsion the area of contact between aqueous and organic reactants is still smaller than in a truly homogeneous biphasic system to which a conversion factor of 10-8 cm is applicable. (ii) The aqueous and organic phases in ref 3 did not contain a common ion; therefore, the potential drop across the interface was undefined, and the kinetics could be significantly different. (iii) The rates measured at microelectrodes, which are essentially free from complications associated with iR-drop and charging current, are often much higher than those obtained with conventional size electrodes.17 The possibility of the surfactant effect on the rate is discussed below. Kinetic Analysis of the Interfacial Reaction. The rate of the uncomplicated heterogeneous bimolecular ET at the ITIES can be expressed as6,7

iET ) nFAkfctop

(9)

where A is the interfacial area, ctop is the concentration of the mediator in the top phase, and

kf ) const cbot exp(-∆Gq/RT)

(10)

where cbot is the concentration of the mediator in the bottom phase and ∆Gq is the free energy barrier (J/mol). For lower overvoltages, a Butler-Volmer-type approximation can be used,6,18,19

∆Gq ) -RF(∆E° + ∆owφ)

(11)

where ∆E° is the difference between standard potentials of two redox couples, F is the Faraday, R is the electrochemical transfer coefficient, and ∆owφ is the potential drop across the ITIES. The interfacial potential drop is a Nernstian function of the ratio of common ion concentration in water and the organic phase: 5,6

[ClO4-]w [ClO4-]o

(12)

The approach curves in Figure 5 were obtained with [ClO4-]w ) [ClO4-]o ) 0.1 M, corresponding to ∆owφ ) ∆owφ°ClO4-. The standard Galvani potential difference for ClO4-, ∆owφ°ClO4-, can be determined only by using some extra-thermodynamic assumption. However, the free energy of perchorate transfer between water and organic solvents is quite small, and |∆owφ°ClO4-| should be less than 100 mV (e.g., the value of 50 mV was used in ref 20). In contrast, the difference between standard potentials of DBCH and Co(II)L/Co(I)L suggests a driving force > 900 mV.21 Thus the kf for reaction 3 should be very large unless the pre-exponential term in eq 10 is very small. On the other hand, the slow step of the rate-limiting reaction 3 does not have to be the bimolecular electron transfer at the interface. This hypothesis can be checked by analysis of the dependencies of kf on ∆owφ and cB12. According to eqs 10-12, ln kf should be a linear function of the ratio [ClO4]w/[ClO4]o. Such Tafel-type dependencies were observed experimentally for different ET reactions at the ITIES.6,19 However, one can see from Table 1 that the rate of reaction 3 remained essentially constant within the limit of experimental error when the concentration ratio was changed by the factor 100 (this corresponds to a ca. 120-mV change in ∆owφ). This suggests that the interfacial ET may not be the rate-determining step. For a simple ET, the effective bimolecular rate constant, k, of the ET at the ITIES should be independent of the mediator concentration in the top phase.6,7 According to eq 9, the electron transfer current iET, is proportional to cB12, and thus the normalized quantity iET/iT,∞ expressing the rate of ET is independent of cB12. This should be true for an even more general situation, i.e., the rate-limiting step being of first order in Co(I)L. The experimental dependence of k on cB12 is given in Table 2. Log k decreases by about 1.5 units as the vitamin B12 concentration is increased from 0.2 to 5 mM. This suggests that the reaction at the ITIES does not fully conform to a simple second-order kinetic model. Observed catalytic current ratios vs scan rate on conventional electrodes in studies of the same reaction in a bicontinuous microemulsion of didodecyldimethylammonium bromide, oil, and water were in good agreement with a second-order model.3a However, re-examination of the influence of vitamin B12 showed a decrease of 0.5 in apparent log k for an increase of vitamin B12 concentration from 0.4 to 2 mM in this microemulsion. A similar effect was seen when the reaction was catalyzed by Co(salen).22a These observations are qualitatively similar to those at the ITIES (Table 2). On the other hand, the electrochemically determined k is independent of vitamin B12 concentration in a purely homogeneous solvent such as DMF,22b consistent with a simple secondorder rate law. Data for several catalysts (0.4-2 mM catalyst) in the DDAB microemulsion along with data for homogeneous DMF in ref 3a for a series of Co(I) mediators gave a linear log k vs E°′CoII/CoI regression line with correlation coefficient 0.94.22a The conclusion is that catalyst concentration is less important than the mediator formal potential, which is the main factor governing log k. The above discussion suggests that the kinetics of reaction of Co(I)L-H with DBCH (Scheme 1) at a liquid/liquid interface may be more complex than a simple second-order process. As vitamin B12 concentration increases, the system becomes less efficient at producing equivalent amounts of Co(I)L-H at the

Reduction of DBCH with Vitamin B12

J. Phys. Chem. B, Vol. 101, No. 16, 1997 3207

TABLE 1: Effect of the Interfacial Potential Drop (∆woφ ) ∆woφ°ClO4--0.06 log([ClO4-]w/[ClO4-]o) on the Rate of DBCH Reduction [ClO4-]w/[ClO4-]o

∆owφ - ∆owφ°ClO4-, mV

kf, cm/s

0.1 1 10

60 0 -60

3.1 × 10-2 3.0 × 10-2 3.3 × 10-2

TABLE 2: Effect of Vitamin B12r Concentration on the Rate of DBCH Reduction cB12, mM

k, M-1 cm s-1

0.2 1 5

13 3 0.4

TABLE 3: Influence of Surfactant Concentration on the Pseudo-First-Order Rate Constant (cB12 ) 1 mM, cDBCH ) 10 mM) surfactant concentration, µM 0 5 10 100

kf, cm s-1 DDAB

DHP

0.030

0.030 0.020 0.012 0.008

0.032 0.032

interfacial reaction site, or kinetic control begins to shift to a process that is not second order. Catalytic DBCH reduction occurs by a multistage inner sphere electron transfer pathway, and results suggest additional complexity introduced at a liquid/ liquid interface. While the previous conclusion that the main influence on the rate of this reaction depends on the formal potential of vitamin B12 remains valid,3a there appear to be relevant secondary kinetic factors at such interfaces. One of these factors may be mixed kinetic control, i.e., a chemical event other than electron transfer, such as bond breaking or formation, which contributes to reaction kinetics at an interface. Such reactions may also be controlled by the formal potential of the catalyst. For example, a step such as reaction of a Co(I) complexes with an alkyl bromide giving an alkyl-cobalt intermediate can be viewed as an inner sphere electron transfer in the sense that bond breaking and formation are concerted with the transfer of a single electron.3a,23 However, the exact reason for the catalyst concentration effect remains elusive. Influence of Surfactant Adsorbed at the ITIES. In microemulsions a monolayer of surfactant is adsorbed at the liquid/liquid interface. Thus, we probed the effect of both anionic (dihexadecyl phosphate, DHP) and cationic (didodecyldimethylammonium bromide, DDAB) surfactants on the kinetics of DBCH reduction. No appreciable change in kf was detected upon addition of up to 100 µM of DDAB to BN (Table 3). A 3-fold decrease in kf was found at 10 µM DHP in contrast with the observation of nearly total blocking of the interfacial ET by a monolayer of the adsorbed long-chain phospholipids.24 A recent SECM study19 also showed that the rate of interfacial ET between aqueous Ru(CN)64- and the oxidized form of zinc porphyrin in benzene (∆E° for these species is small) decreases markedly in the presence of a surfactant film, and a pronounced decrease in kf was observed even with a submonolayer coverage. However, when Ru(CN)64- was replaced with Fe(CN)64- (∆E° > 0.5 V) the ET remained fast even at the ITIES completely covered with the lipid monolayer.19 This is consistent with our observations. The reaction between Co(I)L and DBCH was fast even in the presence of adsorbed surfactant, and the rate-limiting step was only slightly influenced by adsorption. A more detailed study of this phenomenon including the dependence of adsorption on the interfacial potential drop is currently in progress.

Figure 7. Catalytic reduction of DBCH in a thin layer of aqueous solution trapped by the tip inside BN. The tip was scanned away from the ITIES at a 5 µm/s. The tip current dropped abruptly when the tip was moved from the BN solution into the aqueous phase. The aqueous solution contained 1 mM vitamin B12r, 0.1 M NaClO4, and 0.2 M phosphate buffer. The BN was 0.1 M in THAClO4 and 20 mM in DBCH.

Microelectrode “Modification” with Vitamin B12. In Figure 7 one can see the current vs position dependence for a UME tip that was moved from aqueous solution containing vitamin B12r into a BN solution of DBCH and then withdrawn back into the aqueous phase (the graph represents the withdrawal of the tip from BN). One should notice a fairly high and stable current flowing to the tip as long as it is in contact with the organic solution. This current is about 10 times higher than the tip current recorded after the tip is moved back into the aqueous phase. In contrast, essentially no current could be detected at the tip brought directly into BN solution and biased at the same potential. The tip current in Figure 7 is clearly due to the same catalytic reduction of DBCH described by eqs 2 and 3. This process occurs in a thin nano- to micrometers layer of aqueous electrolyte trapped inside the organic phase after the tip pushed into the BN surface.7 The tip/water/BN sandwich works as a microreactor where a tiny amount of Co(II)L/Co(I)L catalyst mediates the reduction of a considerable amount of DBCH over some period of time. The considerable stability of such a reactor is due to very low solubility of vitamin B12 in BN that prevents it from being extracted into the organic phase. One may wonder about the abrupt change in the tip current observed in Figure 7 instead of the regular, smooth currentdistance curve that is usually recorded when the tip is moved away from the ITIES. This can be explained by stretching and eventual breaking of a meniscus of BN that follows the retreating tip. As long as the meniscus connects the tip with the bulk BN, the tip current is determined by the thickness of the trapped water layer and independent of tip position. After the meniscus disappears, iT immediately drops to the iT,∞ value. Conclusions The ITIES can be used as a model system to study catalytic electrochemical reactions in microemulsions. The interface between two immiscible liquids with a monolayer of adsorbed surfactant is of the same nature as the boundary between microphases in a bicontinuous microemulsion. The latter interface is not, however, directly accessible to electrochemical measurements. A better control of the ITIES achieved by using the SECM allowed us to investigate dependencies of the rate of electrochemical catalytic reduction of DBCH by vitamin B12 on various experimental parameters, i.e., reactant concentrations, interfacial potential drop, and the presence of adsorbed surfactant. The overall rate of the interfacial reaction was independent of the interfacial potential drop, and the apparent second-order rate constant k depended on the concentration of vitamin B12, but not of DBCH. The rate of DBCH reduction was unaffected by adsorbed DDAB, but decreased somewhat in the presence

3208 J. Phys. Chem. B, Vol. 101, No. 16, 1997 of adsorbed DHP. These results suggest that the interfacial reaction is not fully described by a simple second-order rate law, as it is in homogeneous media. Thus, SECM-ITIES was able to uncover details of interfacial complexity in a reaction that may not be obvious from conventional voltammetric studies in microemulsions.3a,9 Several questions regarding the influence of the interface remain. However, in this first study of such a complex catalytic process at the ITIES, it is clear that interfacial factors exert some influence on reaction kinetics and/or pathway. We have described an experimental approach to the analysis of complex mechanisms that can be realized with SECM-ITIES. Analysis of a complex catalytic reaction mechanism was reduced to measurements of the effective heterogeneous rate constant. The chemical catalytic step (eq 3) following electrochemical reduction of [Co(II)L-H]+ takes place in a heterogeneous fashion at a water/organic interface, similar to the situation with the same reactants in a microemulsion. The experimental arrangement described allows the utilization of a wide variety of organic solvents in such experiments. The ITIES can serve as a controllable source (or sink) of reagents and charges. Utilizing preferential solubility of different reactive species in water or the organic phase, analysis of catalytic electrochemical reactions at liquid/liquid interfaces might be simplified by independently probing successive steps in a complex pathway. Acknowledgment. The support by the Donors of the Petroleum Research Fund administered by the American Chemical Society (M.V.M.) and the National Science Foundation (CTS-9306961 and CTS-9632391, J.F.R.) is gratefully acknowledged. We thank David Wipf for providing the CE 6000 software package for data acquisition and his valuable help with the SECM construction, Bob Wurman for building the electronic part of the SECM instrument, Michael Tsionsky for useful discussions of the cell design for SECM/ITIES experiments, and De Ling-Zhou for helpful discussions of vitamin B12 catalysis. References and Notes (1) For reviews see: (a) Rusling, J. F. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1994; Vol. 18, p 1. (b) Rusling, J. F. In Modern Aspects of Electrochemistry; Bockris, J. O’M., Conway, B. E., White, R. E., Eds.; Plenum Press: New York, 1994; Vol. 26, p 44.

Shao et al. (2) (a) Evans, D. F.; Mitchell, D. J.; Ninham, B. W. J. Phys. Chem. 1991, 90, 2817. (b) Bourrel, M.; Schechter, R. S. Microemulsions and Related Systems; Marcel Dekker: New York, 1988. (3) For examples see: (a) Zhou, D.-L.; Gao, J.; Rusling, J. F. J. Am. Chem. Soc. 1995, 117, 1127. (b) Zhou, D.-L.; Carrero, H.; Rusling, J. F. Langmuir 1996, 12, 3067. (c) Kamau, G. N.; Rusling, J. F. Langmuir 1996, 12, 2645. (d) Gao, J.; Rusling, J. F.; Zhou, D.-L. J. Org. Chem., 1996, 61, 5972. (4) Kamau, G. N.; Hu, N.; Rusling, J. F. Langmuir 1992, 8, 1042. (5) (a) Girault, H. H.; Schiffrin, D. J. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1989; Vol. 15, p 1. (b) Senda, M.; Kakiuchi, T.; Osakai, T. Electrochim. Acta 1991, 36, 253. (c) Girault, H. H. In Modern Aspects of Electrochemistry; Bockris, J. O’M., Conway, B. E., White, R. E., Eds.; Plenum Press: New York, 1993; Vol. 25, p 1. (6) Tsionsky, M.; Bard, A. J.; Mirkin, M. V. J. Phys. Chem., in press. (7) Wei, C.; Mirkin, M. V.; Bard, A. J. J. Phys. Chem. 1995, 99, 16033. (8) For recent review of the SECM see: (a) Bard, A. J.; Fan, F.-R. F.; Mirkin, M. V. In Electroanalytical Chemistry, Bard, A. J., Ed.; Marcel Dekker: New York, 1994; Vol. 18, p 243. (b) Mirkin, M. V. Anal. Chem. 1996, 68, 177A. (9) Owlia, A.; Wang, Z.; Rusling, J. F. J. Am. Chem. Soc. 1989, 111, 5091. (10) Connors, T. F.; Arena, J. V.; Rusling, J. F. J. Phys. Chem. 1988, 92, 2810. (11) Lexa, D.; Saveant, J. M. Acc. Chem. Res. 1983, 16, 235. (12) Bard, A. J.; Fan, F.-R. F.; Kwak, J.; Lev, O. Anal. Chem. 1989, 61, 1794. (13) Wipf, D. O.; Bard, A. J. J. Electrochem. Soc. 1991, 138, 469. (14) Kwak, J.; Bard, A. J. Anal. Chem. 1989, 61, 1221. (15) Riddick, J. A.; Bunger, W. B., Eds. Organic SolVents; Interscience Wiley: New York, 1970. (16) Feldberg, S. W. J. Electroanal. Chem. 1986, 198, 1. (17) (a) Montenegro, M. I. In Research in Chemical Kinetics; Compton, R. G., Hancock, G., Eds.; Elsevier Science B. V.: Amsterdam, 1994; Vol. 2, p 1. (b) Amatore, C. In Physical Electrochemistry: Principles, Methods, and Applications; Rubinstein, I., Ed.; Marcel Dekker: New York, 1995; p 131. (18) (a) Marcus, R. A. J. Phys. Chem. 1990, 94, 1050. (b) Marcus, R. A. J. Phys. Chem. 1990, 94, 4152; addendum, J. Phys. Chem. 1990, 94, 7742. (c) Marcus, R. A. J. Phys. Chem. 1991, 95, 2010; addendum, J. Phys. Chem. 1995, 99, 5742. (19) Tsionsky, M.; Bard, A. J.; Mirkin, M. V. Unpublished results. (20) Alemu, H.; Solomon, T. J. Electroanal. Chem. 1989, 261, 297. (21) The driving force may be largest for eq 3b, which contains a contribution from the free energy of complexation of vitamin B12a with Br-. For a detailed discussion involving other inner sphere catalysts, see: Lexa, D.; Saveant, J. M.; Schafer, H. J.; Su, K.-B.; Vering, B.; Wang, D. L. J. Am. Chem. Soc. 1990, 112, 6162. (22) (a) Rusling, J. F.; Zhou, D.-L. J. Electroanal. Chem., submitted. (b) Zhou, D.-L. Unpublished results. (23) Saveant, J.-M. AdV. Phys. Org. Chem. 1990, 26, 1-130. (24) Cheng, Y.; Schiffrin, D. J. J. Chem. Soc., Faraday Trans. 1994, 90, 2517.