J. Phys. Chem. 1981, 85, 2870-2873
2870
Charge Transport by Electron Exchange Cross Reaction in Cyclic Voltammetry of IrC16S--Fe(CN)63- Mixtures Trapped in Polycationic Films on Electrodes J. Facci and Royce W. Murray’ Kenan Laboratories of Chemistry, University of North Carolina, Chapel Hill, North Carolina 27514 (Received: June 5, 1987; In Final Form: August 7, 1981)
Fe(CN)2- and IrCl2- can be simultaneously partitioned from 2 M LiCl (pH 2.8) solutions into a polycationic copolymer film of vinylpyridine and (ymethacryloxypropy1)trimethoxysilane coated on a Pt electrode. When charge transport (Dd) through the film is slow, the rate of oxidation of Fe(CN)64-becomes accelerated at potentials which oxidize IrCb3-, and the rate of reduction of IrCb2-becomes enhanced at potentials which reduce Fe(CN)6“. This effect is explained in terms of the electron exchange cross reaction between IrC162-and Fe(CN)64-being sufficiently fast relative to physical diffusion rates as to enhance the rate of charge transport in the film. Electron exchange cross reactions are in contrast not important in charge transport rates for electrochemical reactions of redox mixtures in solutions at naked electrodes.
Generally, cyclic voltammograms of mixtures of dissolved redox substances a t naked electrodes are summations of the individual cyclic voltammograms of the separate redox components. For example, in the positive going potential sweep in the voltammogram of a IrCI6”Fe(CN)$- solution (Figure lA), the IrCh“ oxidation wave a t +0.69 V vs. SCE is unaffected by the preceding Fe(CN),& oxidation wave at +0.20 V; its current simply adds to the “diffusion tail” of the Fe(CN)6” oxidation wave. In other words, oxidation of IrC163-provokes no noticeable change in the rate of transport of electrochemical charge to the electrode for oxidation of Fe(CN):-, even though the electron exchange cross reaction IrC1,2- (xl)
+ Fe(CN):-
(xz)
-+ km
IrC1e3- (xl) Fe(CN)63-(xz) (1) is thermodynamically favored and could act to transport charge from distance xz from the electrode to the closer distance xl. Dahmsz and R ~ f f 3 have - ~ discussed “electron-transfer diffusion” in the context of electron self-exchange reactions of a single redox species at various solution concentrations. Extension of their reasoning and theory to redox mixtures explains Figure lA, in terms of the physical diffusion rates of Fe(CN):- and IrC1,3- in solutions (typical Dsoh 5 X lo4 cm2/s) being much faster than the rate at which electrochemical charge is transported from xz to x1 by reaction 1. Thus takin x 2 - x1 as the electron-transfer contact distance (ca. 9 ), k,,, as6 1.2 X lo6 M-l s-l, and 1mM solution concentrations, the rate of charge transport diffusion can be estimated5 as Dtransfer kcrs(x2- ~ ~ ) ~ . l r C / 4 cm2/s (2) which indeed is much less than typical Dsoln. The IrCl:--Fe(CN):mixture in Figure 1A illustrates what is universally true (but seldom expressed) for redox couple solution mixtures; electron exchange cross reactions (and electron self-exchange reactions) are not important
-
1
-
-
~~
~
~
(1) Bard, A. J.; Faulkner, L. R. “Electrochemical Methods”; Wiley:
New York, 1980; p 232. (2) Dahms, H. J . Phys. Chem. 1968, 72, 362. (3) Ruff, I. Electrochim. Acta 1970, 15, 1059. (4) Ruff, I.; Korosi-Odor, I. Inorg. Chem. 1970, 9, 186. (5) Ruff, I.; Friedrich, V. J. J . Phys. Chem. 1971, 75, 3297, 3303. (6) Gordon, B. M.; Williams, L. L.; Sutin, N. J. Am. Chem. SOC. 1961, 83, 2061. 0022-3654/81/2085-2870$01.25/0
considerations in electrochemical charge transport. It is evident that if electron self-exchange (kex)and cross reaction (k,) rates were somehow greatly increased, and/or if physical diffusion rates were somehow greatly decreased, electron exchange reactions could play a more significant transport role. In thin film electrode coatings of polymers with pendant, electroactive sites (e.g., polyvinylferrocene), the relative immobility of the fixed redox sites causes electron selfexchange between neighbor oxidized and reduced sites to be the charge transport mechanism in the redox chemistry of these f i l m ~ . ~We 3 ~ and others have demonstrated that the transport of electrochemical charge in such polymer films nonetheless follows a diffusion and its rate can be expressed as a diffusion constant D, (cm2/s). This has been recently confirmedI8 via an electron-hopping theoretical model. Measured values of D,, to cmz/s) are much lower than physical diffusion constants in solutions, and are further diminished by polymer cross linking.12 We have recently S ~ O W I that I ~ ~charge ~ transport in thin film electrode coatings of polycationic polymers containing electrostatically t r a p ~ e d , ~highly l - ~ ~ charged redox anions, i.e., Fe(CN):- and IrClt-, also obeys a diffusion law. The polymeric film we have employedz4is a copolymer of vi~~~
~
~
(7) Kaufman, F. B.; Engler, E. M. J . Am. Chem. SOC. 1979,101, 547. (8) Kaufman, F. B.; Schroeder, A. H.; Engler, E. M.; Kramer, S. R.; Chambers, J. Q.J . Am. Chem. SOC.1980, 102, 483. (9) Nowak, R. J.; Schultz, F. A.; Umaiia, M.; Lam, R.; Murray, R. W. Anal. Chem. 1980,52, 315. (10) Daum, P.; Murray, R. W. J. Electroanal. Chem. 1979, 103, 289. (11) Daum, P.; Lenhard, J. R.; Rolison, D. R.; Murray, R. W. J. Am. Chem. SOC.1980, 102,4649. (12) Nakahama, S. Unpublished results. (13) Daum, P.; Murray, R. W. J . Phys. Chem. 1981, 85, 389. (14) Oyama, N.; Anson, F. C. J . Electrochem. SOC.1980, 127, 640. (15) Peerce, P. J.; Bard, A. J. J . Electroanal. Chem. 1980, 114, 89. (16) Rubenstein, I.; Bard, A. J., preprint. (17) Shigehara, K.; Oyama, N.; Anson, F. C. J. Am. Chem. Soc. 1981, 103, 2552. (18) Laviron, E. J. Electroanal. Chem. 1980, 112, 1. (19) Kuo, K. N.; Murray, R. W. J . Electroanal. Chem. In press. (20) Facci, J.; Murray, R. W. J. Electroanal. Chem. In press. (21) Oyama, N.; Anson, F. C. J. Electrochem. SOC.1980, 127, 247. (22) Oyama, N.; Anson, F. C. Anal. Chem. 1980, 52, 1192. (23) Shigehara, K.; Oyama, N.; Anson, F. C. Inorg. Chem. 1981, 20, 518.
0 1981 American Chemical Society
The Journal of Physical Chemistty, Vol. 85, No.
Letters
20, 1981 2871
TABLE I: Charge Transport Diffusion Coefficients‘ f o r Oxidation of IrClG3-and Reduction of F e ( C N ) 6 3 -Ions Trapped i n MPS-PVPyH’ Films i n Contact with DH 2.8. 2 M LiCl CIrC&’ - > b
D c t , cm’ls
CFe(CN)6 ’ - 9
M
1.8 1.9 3.2 1.6
0.011 0.10 0.46 1.0 0.035 0.18 0.40 0.24 0.30 0.34 0 0 0.080 0.24 0.40 0.56 0.47
0.09 0.15 0.30 0.57 0.040 0.059 0.058 0.036 0.043 0.069
V Fe(CN: ) -”-
Figure 1. Curve A: cycllc voltammetry at 0.1 V/s of a pH 2.8, 2 M LiCl solution contalnlng Fe(CN),> (Eo’&” = +0.25 V vs. SCE) and IrCI2- (€OfsM = 4-0.69 V), at naked Pt. Curves B: a sequence of cyclic voltammograms observed at 0.1 V/s on a coated PUMPSPVPyH’ polycatlonic electrode during simultaneous partitioning of Fe= -I-0.69 V) (CN)$- ( € O f a M = +0.20 V vs. SCE) and IrClz- (Eofsurf from a dilute mixture of these ions in pH 2.8, 2 M LiCI. The curves represent the typical appearance of the IrC162-’3- and Fe(CN)63-’4surface waves when the polycationic film contains low concentrations of these ions. S = 350 uA/cm2.
Ir
Fe
X
x x 10-9c X
10-’Oc
2.6 x 2.2 x 10.~ 2.8 x 102.3 x
3.8 X 2.2 x 3.3 x 10-9c na na na
1.8 X 10-8esf
1.0 x l o + e,f na 4.4 x 1 0 - ~ ~na, f na 4.0 x 10-9e 3.0 x e 3 f na na 6.1 X
a F r o m slope of chronoamperometric i vs. t-’” p l o t as in Calculated f r o m coverage ( m o l / c m i ) measured ref 20. b y potential s t e p c o u l o m e t r y , divided b y 7000 A film Same electhickness (dry weight). Data f r o m ref 20. trode used f o r this series of measurements. e Same electrode used for this series of measurements. f Cottrell plots f o r these experiments are s h o w n in Figure 2.
1
nylpyridine and (y-methacryloxypropyl)trimethoxysilane, MPS-PVPy. A methanol solution containing sufficient CH3
+CHz-CH-
bo
I
C H e - C H e 2
94
L
O
I OCH~CH,CH~SI(OCH~ 13
MPS-PVPy
copolymer to prepare a film 7000 A thick (dry weight, contains 5.1 X lo-’ mol/cm2 pyridine sites) is applied as a droplet to a superficially oxidized26Pt/PtO disk electrode, slowly evaporated (forming surface PtOSi-bonds), and exposed to moist HC1 vapor to hydrolytically cross link the polymer via the silane function, whereupon the film becomes quite insoluble and adherent to the electrode surface. The Pt/MPS-PVPy films efficiently partition Fe(CN)t- and IrC163-from acidic (2 M LiC1, pH 2.8) solutions, and continue to display their cyclic voltammetric waves after the electrode is transferred to redox ion-free 2 M LiC1, pH 2.8 supporting electrolyte. By adjusting the concentration of or of Fe(CN)63-in the partitioning bath, the quantities incorporated at equilibrium into the Pt/MPS-PVPyH+ film can be systematically variedm from 8 X 10-lo to 800 X 10-lo mol/cm2 or from 0.011 to 1.0 M concentrations.26 (24) Facci, J.; Murray, R. W. Unpublished results. (25) Lenhard, J. R.; Murray, R. W. J. Electroanal. Chem. 1977, 78, 195.
Figure 2. Cottrell plots of the equation i = nFAD,’’* CIr/a1’2t1’2 for current-time curves resulting from the chronoamperometric oxidation of IrCIz- in Pt/MPS-PVPyH’ polycationic film electrodes in pH 2.8, 2 M LiCl where the films contain only IrCle3-at C I r = 0.055 M (curve a) and mixtures of IrCIz- and Fe(CN)z- where C I r is approximately constant at 0.046 f 0.009 M and C F Bvaries from 0.08, to 0.24, to 0.56 M (curves b, c, d, respectively).
The charge transport rate, Dot,measured by chronoamperometry,” for oxidation of IrC163-and for reduction of Fe(CN)63-depends on the concentration of the redox ion in the film and is rather slow above ca. 0.3 M, as shown by exemplary data20 in Table I. It is believed that,20p27 similar to charge transport in fixed site redox polymers,s”Jl transport of electrochemical charge in Pt/MPS-PVPyH+ (26) The quantities, mol/cm2, of IrCleS-or Fe(CN)6S-partitioned are determined from charges under their (slow potential scan) cyclic voltammetric waves or from potential step coulometry. Film concentrations are calculated on the basis of unit density, dry weight thickness, 7000 A, of the polymer film. (27) As discussed elsewhere,20 whether the decrease in D, at high concentration represents a (coulombic repulsion induced) decrease in the electron self-exchange rate or a (electrostatic cross-linking induced) decrease in the rate of short-range diffusion motion is difficult to judge and indeed both factors may be involved. Translating D, into k,, values depends on distinguishing between these two rate-determining steps.
2872
The Journal of Physical Chemisfry, Vol. 85, No. 20, 1981
T
. . i. ; i.
Letters
B
A
.................
X-
Flgure 4. Schematic concentration distance profiles of IrClz- (O), IrCI,*- (m), and Fe(CN):(-) in Pt/MPS-PVPyH’ films as the electrode potential is scanned positively from 0 V. Panel A: curves and 4-0.35V vs. SCE. Panel 6: a-d, profiles at 0,+0.15, ,Ewkta_, profiles at +0.61 V where IrCI, oxidation has begun.
Flgure 3. Steady-state cyclic voltammograms for high Concentration mixture of Fe(CN):and IrCI,” in polycationic Pt/MPS-PVPyH+ electrode in pH 2.8, 2 M LiCI. Curve A: potential limits adjusted so as to scan (0.1 Vls, S = 300 pA/cm2) the redox wave of only one of the two couples, left-hand curve is IrClg2-13-, right hand curve is Fe(CN),”lc. Curve B scan both waves at 0.05 V/s, S = 300 pA/cm2. Curve C: scan both waves at 0.005 V/s, S = 30 bA/cm2. “Prepeaks” indicated by an asterisk. Vertical lines are at Eo’,&
films containing trapped Fe(CN):- or IrC163- involves electron self-exchange reactions between neighbor IrC16% and IrC162-ions and between neighbor Fe(CN):- and Fe(CN)64-ions, with intervening short-range diffusive motions. Rates of the electron exchange and diffusion transport processes are more nearly equal in the film than in aqueous medium. Mixtures of these ions in Pt/MPSPVPyH+ films thus provided an appealing test for participation of the electron exchange cross reaction, reaction 1 in charge transport. In the present paper, we extended the Dct measurements to the oxidation of IrC1:in Pt/MPS-PVPyH+ films equilibrated with various IrC&3--Fe(CN)63-mixtures in pH 2.8, 2 M LiC1. The satisfactorily linear Cottrell plots obtained are illustrated in Figure 2; the slopes yield D,, (Table I) as we have explained before.11,28 The results show that charge transport becomes slow in these films irrespective of the concentration of IrC1:- as long as the total Cyp + CFe(cN) 3- concentrationzgexceeds ca. 0.3 M. This similarity to the hehavior of the individual ions shows the concentration dependence of D,, in the mixture is purely an electrostatic effect operating by electrostatic cross linking and/or Coulombic repul~ion.~’ Cyclic voltammetry of a Pt/MPS-PVPyH+ electrode equilibrated with a pH 2.8,2 M LiCl solution of IrCh3- and Fe(CN)63-so as to incorporate a high concentration of both (28) The slo e of Figure 2 is nFAD,,’IaC/T1/*,where C is the concentration of in the film. (29) That Fe(CN)6“ can act as surrogate for IrCb3- in terms of the latter’s Dd concentration dependency demonstrates the electrostatic character of the effect but does not decide the rate-determining step issue.2l
redox ions (0.30 and 0.34 M, respectively) is shown in Figure 3. If the electrode potential is scanned over only the IrClS2-I3-or Fe(CN):-/4- voltammetric regions, the well-formed waves observed have symmetrical “tails” symptomaticg of diffusional control of the current by the rate of charge transport in the film on the time scale of the potential sweep. Charge transport for the individual waves involves only I r c l ~ ~ - and / ~ - Fe(CN)63-/4-electron self-exchange and short-range diffusion. The “tails” of the waves represent continuing, slow charge transport from the parts of the film most remote from the electrode.30 Steady-state voltammetry where the potential scan encompasses both electrode reactions is quite different as shown in parts B and C of Figure 3. In Figure 3B, a t u = 50 mV/s, the anodic peak current for the IrC163-oxidation wave and the cathodic peak current for the Fe(CN)$reduction wave are much larger than their cathodic and anodic partners, respectively. At a slower potential scan rate, Figure 3C, the peak currents for IrCb3- oxidation at +0.69 V and for Fe(CN):- reduction a t +0.20 V are more nearly equal to their cathodic and anodic counterparts, respectively, but both are preceded by a sharp prepeak (see asterisk). It is evident that in these experiments additional oxidation of Fe(CN)6k occurs concurrently with the IrCh3oxidation wave, and additional IrC12- reduction occurs concurrently with the Fe(CN):- wave. In other words, the onset of IrC1:- oxidation accelerates charge transport for Fe(CN)d oxidation, and the onset of Fe(CN)$- reduction accelerates charge transport for IrCb2- reduction, causing increased current flow in both cases, giving rise to the starred peaks specifically in Figure 3C and the additional but unresolved currents in Figure 3B. Events in parts B and C of Figure 3 for a positive-going potential sweep are further explained by the schematic concentration distance diagrams of Figure 4. Concentration profiles for Fe(CN)64- (-) in the film as the electrode potential is swept through the Fe(CN)64-oxidation wave are represented in Figure 4A. Halting the potential scan at potential d (say +0.35 V) would eventually lead to complete oxidation of Fe(CN)t-, but slowly, since the concentration gradient of Fe(CN)$- sites which drives diffusion and electron self-exchange to continue the Fe(CN)6k oxidation is not very steep. Scanning to potentials (30) Similar voltammetry is observed if the two ions are separately partitioned a t high concentration into a Pt/MPS-PVPyH’ film.
J. Phys. Chem. 1981, 85,2873-2877
oxidizing IrC163- (Figure 4B), on the other hand, opens reaction 1 as a new pathway for oxidation of Fe(CN)$-. Transport of charge for Fe(CN)$- oxidation by reaction 1 is a t this point more efficient31 than transport by Fe(CN),” diffusion or by Fe(CN)63-/4-electron self-exchange because a t the prepeak potential in Figure 3C, the concentration of IrCl$ sites produced near the electrode and near the electrode are both the dCIrcls?/dx gradient large relative to those of Fe(CN),4- as illustrated in Figure 4B. Not all of the remaining Fe(CN)64-necessarily becomes oxidized via reaction l , but more is oxidized within the time span of the potential sweep than would be oxidized in the absence of the IrC1,3- oxidation reaction. Results in parts B and C of Figure 3 directly illustrate the occurrence and importance of electron exchange reactions in charge transport in these films. It is important to realize that, in thin films of redox materials, the effects in parts B and C of Figure 3 depend on charge transport diffusional control of the currents. In contrast, with a thinner film or (low concentration) fast charge transport, which promotes fast equilibration of redox sites in the film with the electrode potential during a potential scan, a Fe(CN),’ oxidation rate enhancement cannot be observed at the IrCb3- oxidation potential since the Fe(CN)&-in the film is already completely oxidized. This voltammetric behavior is illustrated in Figure 1B for a low concentration mixture of IrCb3- and Fe(CN)63-,where the reactions occur independently, in the same manner as they do in solutions (Figure 1A). Finally, note that the appearance of the (asterisk) prepeaks in the voltammograms of Figure 3C is qualitatively reminiscent of prepeaks in voltammetry of spatially segregated bilayers of redox polymer^.^^^^^ In bilayers, electron exchange cross reactions between inner and outer film layers cause sharp prepeaks; D, between the electrode and the outer film layer is (effectively) zero due to the spatial (.e-)
(31) Transport rate enhancements as large as ca. 38% can be anticipated from eq 14 and 17 of ref 2. (32) Abrufia, H. D.; Denisevich, P.; Umai5a, M.; Meyer, T. J.; Murray, R. W. J. Am. Chem. SOC.1981,103, 1. (33) Denisevich, P.; William, K. W.; Murray. R. W. J. Am. Chem. SOC. 1981,103, 4727.
2873
arrangement. Special attention was accordingly paid in the present study to the possibility of IrCb3- and Fe(CN)6* somehow spatially segregating themselves in the MPSPVPyH+ film, since this would compromise the charge transport-based interpretation of Figure 3 which assumes uniform mixing of the two ions. Pertinent observation are as follows: (a) The manner of mixed film preparation (simultaneous partitioning of the two ions) is not, by itself, conducive to layer making for Fe(CN)63-and IrCb3-. Indeed, Figure 1B shows various stages of the partitioning, and the waves for IrC&2-/3-and Fe(CN),*l4- increase together, in proportion to the relative ion concentrations in the partitioning bath. (b) If a film containing only IrCb3- is contacted with a Fe(CN):- solution, IrCe3- is expelled from and Fe(CN)$incorporated into the film to just the extent anticipated by the knownz4partition coefficients of the two ions. (c) X-ray photoelectron spectroscopy was carried out at near normal (87O), intermediate (35O), and grazing (go) electron emission angles, on films containing both IrC1:and Fe(CN)e3-. The ratio of intensities of the Ir 4f and Fe 2p bands did not change with emission angle being 0.40, 0.35, and 0.38, respectively. Since the grazing angle favors surface atoms more than does the near normal angle, this observation shows that no gradient exists in the relative Fe(CN)63-and IrC163-populations a t the outermost film surface. The above results do not rule out spontaneously formed, randomly located, “pockets” of the redox ions, some containing only one ion and some containing only the other. We think this is in fact an implausible eventuality, and are accordingly inclined to accept the idea that the IrClaand Fe(CN)63-in these films are homogeneously mixed. We maintain therefore that the phenomena of Figure 3 demonstrate, for the first time, a role of electron exchange cross reactions (reaction 1)in charge transport in redox mixtures. Acknowledgment. This research was supported in part by a grant from the National Science Foundation. The authors are indebted to Professor A. J. Bard for pointing out the prior work of Ruff. This is paper 40 in a series on chemically modified electrodes.
Studies of H and 0 Atom Reactions by OH Infrared Chemiluminescence B. S. Agrawalla, A. S. Manocha, and D. W. Setser” Department of Chemistry, Kansas State University, Manhattan, Kansas 66506 (Received: June 24, 198 1; In Final Form: August 17, 198 I )
The vibrational energy disposal to OH and relative rate constants have been measured for the 0 + HI and GeH, reactions and for the H + NOz and CIOz reactions. The experiments were done in a flowing afterglow apparatus which gives arrested vibrational distributions as shown by comparisons of the OH(u) distribution from H + NOz with other data in the literature. The energy disposal pattern for the 0 atom reactions closely resembles that for F (or C1) atom reactions. Comparison of the relative emission intensities from H + NOz and H + Clz and the accepted rate constants for the reactions permits a selection to be made for the better vibrational OH u’--* u’- 1 Einstein coefficients. Emission from NO was observed from the H + NOz reaction confirming predictions that NO is involved in the energy disposal. Introduction Infrared chemiluminescence studies of HF, DF, HC1, and DC1 from H, F, and C1 atom reactions have been extensively investigated in our lab by means of the cold-wallarrested vibrational-rotational relaxation techniqueliz and
the flowing-afterglow-arrested vibrational relaxation technique.3,4 Recently, the flowing-afterglow technique (1) K. Tamagake, D. W. Setser, and J. P. Sung, J. Chem. Phys., 73, 2203 (1980).
0022-3654/81/2085-2873$01.25/00 1981 American Chemical Society
