6555
Oxidation of Imidazolepentacyanoiron(11) by Hydrogen Peroxide Mark L. Bowers, Dan Kovacs, and Rex E. Shepherd* Contributionfrom the Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260. Received March 17, 1977
Abstract: The oxidations of ( C N ) s F e ( i m ~ ) ~and - ( C N ) ~ F ~ ( N - C H ~by ~ ~H202 Z ) ~proceed by dissociation of the corresponding imidazole and rate-limited substitution of H202. Hydroxyl radical formed in the initial one-electron reduction of H202 is scavenged competitively by ( c N ) ~ F e ( i m z - R ) ~H202, -, or organic components (RH). H2EDTA2- is inhibitory; tert-butyl alcohol, imidazole, and acrylonitrile are kinetically equivalent to HO. itself. Substitution of the imidazoles on (CN)sFeH203was followed by competition with 2-methylpyrazine. Values of the second-order substitution rate constants are 410 f 30 (imz), 418 f 20 (N-CH3imz), 323 & 6 M-I s-I (2CH3pz) a t 25.0 OC, p = 0.10, phosphate buffer. The dissociation of the ( C N ) s F e ( i ~ n z - R ) ~complexes occurs with AH* = 21.5 f 0.5 kcal/mol, AS* = + I f 2 eu, k298 = (2.20 f 0.21) X s-I (imz); AH* = 19.5 f 0.5 kcal/mol, AS* = - 5 f 2 eu, k298 = (3.24 f 0.22) X s-I (N-CH3imz). The stoichiometry for the reduction of H202 by ( C N ) s F e ( i ~ n z - R ) ~is- 0.50 a t limiting initial time. The Fe(II1) product complexes, or species in equilibrium with them, catalyze the decomposition of H202 into 02. The rate of decomposition is dependent on [H202I2 and [ ( C N ) s F e ( i m ~ - R ) ~ with - l kd = 8.4 X lo-* M-2 s-l, p = 0.10, T = 25.0 OC, in the presence of 0.01 M H2EDTA2-.
The coordination of histidine as a n axial ligand for iron in the cytochromes and heme proteins promotes an interest in the coordination of imidazole with low-spin Fe(I1) and Fe(II1) We recently characterized N-3 bound imidazole and histidine complexes of (CN)5Fe3- and (CN)5Fe2moieties5 This report describes the kinetics of the oxidation of ( C N ) ~ F e ( i m z ) ~and - ( c N ) ~ F e ( N - c H $ m z ) ~by - HzO2. Initially, we had hoped that a study of this system might provide evidence for imidazyl or Fe(II1) imidazyl intermediates similar to those proposed by Wang as radical precursors in the ATP synthesis within mitochondria and related model systems.6 The data obtained for the oxidation of ( C N ) ~ F e ( i m z ) ~ by H202 support a dissociative mechanism. Loss of imidazole from ( C N ) s F e ( i m ~ ) ~and - substitution of H202 into the inner coordination sphere of (CN)sFeH203- are rate determining. Our results are in harmony with the general observation that reduction of H102 by metal centers having labile coordination sites is about IO4 faster than when H202 is restricted to an outer-sphere role.' The thermally accessible dissociation of imidazole from ( C N ) ~ F e ( i m z ) ~ provides catalysis via (CN)5FeH2O3- for reaction with H202. The observed thermal pathway is similar to the path observed in the oxidation of Fe(CN)64- by H202 which is activated by photodissociation of CN-.8 The hydroxyl radical which is formed either reacts with a second mole of (CN)jFe(imz))- or it may be intercepted by other organic molecules present in the reaction medium. The resultant organic radicals may be separated into several classes on the basis of their reactivity toward the Fe(I1) and Fe(II1) species in the system. The differentiation parallels the observations of Walling for the reactivity of organic radicals produced by Fenton's reagent and the HzOrlFe(EDTA)2- ~ y s t e mA. ~substantial catalase activity has been observed for the ( C N ) 5 F e ( i m ~ ) ~product. -
Experimental Section (CN)~Fe(imz)~and (cN)~Fe(N-CH3imz)~Solutions. The (CN)sFeL3- solutions were prepared immediately before use by combining weighed amounts of the desired imidazole, dissolved in Ar or N2 purged buffer, with weighted samples of Na3[(CN)5FeNH31.3H20.S All manipulations of the Fe(I1) imidazole complex solutions were carried out using syringe or inert atmosphere techniques. Platinum or Teflon needles were used to eliminate contact with catalytic sources of iron. Samples were protected by subdued light with aluminum foil wrapped vessels. (CN)sFe(imz)2- Solutions. (CN)sFe(imz)2- was prepared for kinetic and equilibrium studies by oxidation of ( C N ) ~ F e ( i m z ) ~with - HzO2. The (CN)sFe(imz)2- product was separated from excess H202 and
ligand by means of anion exchange on AG-4X resin in either the CIor Sod2- form. The product was isolated immediately from the resin phase by solutions of NaCl at the desired ionic strengths. Samples were stored at 5 OC in thedark with known amounts of free ligand imidazole readded to suppress the conversion of (CN)sFe(imz)*- into Fe2(CN) lo4- dimeric ions.I0-Il Catalase activity studies were carried out in black-painted volumetric flasks, additionally wrapped in aluminum foil and suspended in a constant temperature bath at 25.0 OC. H 2 0 2 Analysis. The H202 was titrated by standard (NH4)2Ce(N03)6 solution in H2S04. H202, determined on aliquots for the catalase and stoichiometry experiments, was separated from other components of the reaction medium by passage through AG-4X resin in the S042- form. The aliquot with H2O was titrated by Ce(1V). Rapid Scan Spectra. The oxidation of ( C N ) ~ F e ( i m z ) ~by- H202 was examined from 250 to 550 n m by a Vidicon rapid scanning spectrophotometer.I2 Absorbance data as a function of wavelength and time were stored for recall in a PDPl2 computer. Wavelength and intensity calibrations were carried out on each run by storage of the spectrum of holmium oxide glass as a reference. Stopped-Flow Kinetic Data. Data for the oxidation of (CN)sF e ( i m ~ ) ~were - obtained using a Durrum D-I I O stopped-flow with appropriate temperature control equipment. The photomultiplier output was converted to absorbance and displayed on a Tektronix storage oscilloscope. The data record was obtained by photographing the scope. Reactions lasting longer than could be easily displayed over 50 s on the oscilloscope were recorded with a Houston 2000 X-Y recorder with a calibrated time base. Results obtained in this manner were in agreement with preliminary data obtained on a computerinterfaced stopped-flow instrument.I2 The final absorbance was checked against the Beer's law calculated ( t 1.15 X IO3 M-I cm-' for (CN)sFe(imz)2- at 403 n m ) value, based on the weight of Na3[ Fe(CN)sNH3].3H20 used to prepare ( C N ) ~ F e ( i m z ) ~ - . UV-Visible Spectra. Spectra were obtained on a Varian-Cary I I8 spectrophotometer with a thermostated sample compartment. Dissociation kinetics of imidazole and N-methylimidazole from (CN)SFe3- were followed at 450 nm by scavenging (CN)sFeH?O3- with 2-methylpyra~ine.~ Equilibrium studies with (CN)sFe(imz)2- and NCS- were monitored at the 590-nm maximum of (CN)sFeSCN3where t 2700 M-I cm-1.13,14 Reactants for the dissociation experiments and equilibrium studies were flushed with Nz and temperature equilibrated prior to mixing. The mixed solution was transferred in a syringe operation to a spectrophotometric cell which was capped with a septum, preflushed with N2, and equilibrated at the same temperature of the cell compartment. Buffers and pH Measurements. The pH of the reaction solutions was monitored with an Orion 601 digital pH meter standardized with commercial buffers. Buffer solutions were prepared from analytical weights of known sodium buffer salts. Kinetic results were reproducible from different sources of buffer salts. Because the decomposition of H2Oz is very sensitive to iron and
Bowers, Kovacs, Shepherd
/
Oxidation of Imidazolepentacyanoiron(l1)
6556
v. 20
OO
40
GO
i2&m~],
Figure 1. Rapid scan spcctr;i: curvcs ( I o w c ~10 t highcst)
I .O. 5.0, 20.0, 40.0, and 5 6 0 I; absorbance 0.00 to 0.50: wavolength 279 lo 539 nm; maxima at 294, 357. 403 nm; [(CN)#e(im7j3-] = 9.70 X IO-‘ M: [HiOz]=2.OflX 1 0 ~ z M : p = 0 . 1 0 . T = 2 5 0 C . at
Figure 3. lnhibihon of the (CN)sFe ihz’-/H202 reaction by added imidazole: [(CN)sFe(it~k)~-] = 2.55 X i O V M. [H2021 = 0.103 M. pH 7.0. p = 0.10, T = 25.0 ‘C, [imz] = freebase ligand concentration calculated with pK. = 6.95, pH measuremenlS. and total analytical concentration of imidazole.
y7F Q
H202 in the system. The kinetic scheme may be additionally altered by the presence of radical scavengers if the resultant radical (Re) is reducing for ( C N ) ~ F e ( i m z ) ~The - . effect of the organic radical scavengers will be described in a later section.
time
.I 07
scheme I
Fipre 2. Oscilloscopic data recording of (CN)SFe imzl-/HzOI reaction at 403 nm: (a) [(CN)~Fe(imz)~-] = 2.55 X IO-‘ M. [HlOz] = 0.103 M. pH 7.00 phosphate bhffer. p = 0.1. time base = 2.00 s/div. [imz] = [Fe(ll)],.; (b) same conditions as (a) except that [imz] = 9.90 X IO-’ M [Fedl)]tot and time base = 5.00 s/div.
(CN)5FeH203-
+
copper impurities in salts. the reproducibility implies [hat the salts contained an acceptable minimum of contamination: A fey eaily experiments illustrated that presence of metallic irhputilieswill dramatically and erratically alter the observed rate of reaction between . of organic chelating agents (CN jsFe(imzj’- and H 2 0 ~ Addition which passes abstractable hydrogen atoms was judged io be an inapproptlate modeofcontrolon the basisofthe reactibityofthesereagents with HO..”.’6,20
Result! and Discussion Oxidation of (CN)sFe(imzy- by HzOz. Figure 1 illustrates the rapid scan spectrum of the reaction between H202 and (CNlsFe(imz)’- at I .O, 5.0,ZO.O. 40.0, and 56.0 s after mixing. No chemical intermediates a t substantial concentrations are detectable. A typical oscilloscopic data record of the reaction monitored at 403 nm is shown in Figure 2. An inhibition period is observed early in the reaction. A pseudo-first-order period follows the inhibition. The inhibition period is shortened with increasing [imidazole] at constant pH. The rate is also s u p pressed with increasing imidazole. The first-order rate constant obtained from the region after inhibition is k a M . The reciprocal of R o w is shown in Figure 3 as a function of imidazole at pH 7.0. These data indicate a linear proportionality with l / k o b d tending toward zero in the absence of imidazole. The data which we will present are consistent with the mechanism shown by Scheme I. The rate of formation of (CN)sFe(imz)2is strongly dependent on the ampetition of HO- for Fe(I1) and Journal of the American ChemicalSociefy 1 99:20
+ H2Oz A (CN)5FeH202+HO.+OH-
(CN)5FeH202-
+ (CN)sFe(irrlz)’-
.-k3
L
(CN)3FeH2O3-
k-i
HO-
(2)
+ (CN)sFe(im@-
(4)
-OH-
HO- + Hi02
A H2O + HO2.
k6
(5)
+ H202
(6)
k, + (CN)5FeH202- + imz k-i
(7)
2HOi
H3O+
(3)
k,
+ (CN)sFe(imz)’-
(CN)5Fe imz2-
+ (CN)SFe(imz)2-
0 2
IIK. + imz F=+ Himz+
(8)
The kinetic scheme shown by Scheme I predicts a rate iaw given by eq 9 undet the following assumptions: (a) steps 3,4, 5.6,and 8 are rapid Fompared to steps 1.2, add 7, (b) that the amounts of (CN)sFeH203: and (CN)sFeH2O2- are always very small, (c) that other radicals generated from HO. that are oxidizing for (CN)sFe(imz)’- may be represented by the kinetically equivalent step 4,’’ and (d) that the radical rea&tions between HO. and H02- are similar to those studied with
Sepfember 28.1977
6557 Table 1. Rate Constants at 25.0 "C for Scheme I
Rate
k,
Ref
(2.20 0.21) x 10-3 S - I 410 f 30 M-I s - I 107 f 1 M-] s-I -104 M-1 s-I -1011 M-1 s-1 (1.2-4.5) X IO7 M-I s-] 7.6 X I O 5 M-I s-I 5 2 x 10-4 s - ~ -50 M-' s-I
This work This work 17 18, 19
20, this work, 21
30-
20 22
This work, 28 This work, 28
25-
?-.
-M
Table 11. Hydroxyl Radical Scavenger Effects on
(CN)~Fe(imz)~-/Hz02 System kno, M-' H 0.
S-1
Scavenger Imidazole
(ref)
H 2EDTA2-
5 X IO9 (20) 4.7 X IO8 (20) 3 X IO9 (23) 2.8 X IO9
H202
2 X IO7
(histidine) tert- Butyl
alcohol Acrylonitrile
(24)
HO. mode of reaction
Effect on (CN)SFe(imz)3-/ H?O?svstem
H abstraction or ring addition H abstraction
None
Addition
None
H abstraction
Suppression
H abstraction
Suppression
None Figure 4. Inhibition of the (CN)~Fe(N-CH3imz)~-/H202reaction by added N-methylimidazole: conditions identical with those of Figure 3, [N-CH3imz]fcalculated with pK, = 7 .
complex are shown in Figure 4 with a slope of 6.37 X IO3 M-l S.
these species and Fe(CN)64-/3- by radiolysis methods. The coefficient A may be taken as the fraction of productive events forming (CN)5Fe(imz)2- from the competition of (CN)5F e ( i m ~ ) ~and - H202 for HO.. If ks[HzO2] > ] kd[Fe(II)]av (1 A ) 1.0. [Fe(II)]av is defined as the effective Fe(I1) level throughout the run, [Fe(II)]o/2. [Fe(II)]av = 7.0 X M in these studies. Some collected rate constants are shown in Table I. The value for k3 is estimated from the exchange of Fe(CN)63- and ( c N ) ~ F e ( p y ) ~l 9- or the Fe(CN)64-/(CN)5FeH202- exchange.I8 These estimates are consistent with the redox potential of ( C N ) ~ F e ( i m z ) ~ - /within ~0.005 V of the Fe(CN)64-/3- couple5 and the rapid oxidation of (CN)5F e ( i m ~ ) ~by- F ~ ( C N ) Gobserved ~in our laboratory. Steps involving the oxidation of ( C N ) ~ F e ( i m z ) ~by - HOz. were omitted on the basis of the absence of the oxidation of Fe(CN)64- by H02..34The rate constants in Table I show that assumptions (a) and (b) are met. Assumptions (c) and (d) are in agreement with our observations.
+
-
+
- -
The slope of the line defined by the data in Figure 3 is 5.44 M-' s. A rearrangement of eq 12 shows that Scheme I is consistent with the equilibrium induced inhibition of kobsd (eq 13) a t constant [ H202]. At [ H2021 = 0.10 M a fractionation factor of 1.8 is observed and the rate appears to be pseudo-first-order in [H202]. Data for the N-methylimidazole X IO3
Since the N-methylimidazole complex lacks a readily abstractable hydrogen, the reactivities of this complex, (CN)5F e ( N - c H 3 i m ~ ) ~should -, more nearly satisfy condition (d). Combining Kf = (1.29 f 0.18) X I O 5 M-I for (CN)SFe(NC H 3 i m ~ ) ~with - Davies' value of kz allows calculation of ( 1 A ) = 1.8 f 0.2, which is in very good agreement with k4 = 1.10 X 1011M-I s-l, k5 = 2 X IO7 M-l s-l. With thesevalues of k4 and k5, (1 A ) equals 1.79. Assuming that ( 1 A ) = 1.8 the data from Figure 3 imply that Kf for ( C N ) ~ F e ( i m z ) ~ is 1.05 X IO5 M-I, which is in reasonable agreement with the kinetically determined value of (1.86 f 0.18) X IO5. The agreement of t h d i n e t i c constants using Davies' value for kz implies that dissociation of imidazole is followed by substitution controlled reduction of H2Oz by (CN)sFeH2O3- in the H * O ~ / ( C N ) ~ F ~ (reaction. ~~Z)~HO* Radical Scavengers. The (CN)5Fe(im~)~-/HzOz system has been studied in the presence of H2EDTA2-, imidazole, tert-butyl alcohol, and acrylonitrile. The effect of addition of these species is shown in Table 11. Because two of the necessary components (imidazole and H202) in the reaction medium are themselves capable of HO. scavenging it is important to consider their effects first. Imidazole inhibition in Figure 3 shows no tendency toward upward curvature in the concentration range of 2.0 X to 3.0 X M [imzItot. If formation of imidazyl radicals were competitive for H202 scavenging in a manner that netted a reducing radical for (CN)SFe(imz)2- or a dimer diamagnetic product, increasing the imidazole concentration should enhance the suppression of the rate by either of these paths in addition to the equilibrium suppression shown by Scheme I. The absence of any marked upward trend in the data with increasing [imz] implies that if imidazyl radicals are formed, these species are kinetically equivalent to step 4. (Under conditions where [imz]: [H202]:[tert-butyl alcohol] were 2 X 10-2:1:1 no additional inhibition could be detected; when [H202]/[tert-butyl alcohol] = 1 .O, tert-butyl alcohol has a kinetic factor advantage of about
+
Bowers, Kouacs, Shepherd
+
+
/ Oxidation of Imidazolepentacyanoiron(II)
6558 Table 111. Effect of H,EDTA2- on (CN)sFe(im~)~Oxidation0
I 02[imz]
1 O2[H2EDTAI2-
I .02 I .02 I .03
102kob,& s-’
0.00 I .oo 5.00 2.50 0.50 0.50
1.03 1.03 I.Olh
HO-
3.68, 3.61 2.55 2.55 2.55 2.55
N-Methy-
20.) The absence of an experimentally discernible difference in rate allows the inference that the radical of tert-butyl alcohol acts only by oxidizing Fe(I1). H2EDTA2- added to the reaction medium causes a significant retardation in the rate. A sample at high imidazole (-5.8 X M in free base) which shows almost no inhibition period is strongly inhibited by addition of H2EDTA2-. The inhibition period is returned upon addition of H2EDTA2- at the same [imz]. Maximum rate suppression is already achieved at 5.0 X M H2EDTA2- as shown in Table 111. The value of kobsd at 5.0 X M H2EDTA2- is 0.68 f 0.01 of its value in the absence of H2EDTA2- for both the ( C N ) s F e ( i m ~ ) ~ and ( C N ) F ~ ( N - C H ~ ~ oxidations ~ Z ) ~ - by H202. Walling has proposed that HO. extracts hydrogen from the ethylenediamine backbone of H2EDTA2-. The resultant radical is a reducing radical for Fe(II1) (eq 14, 15).24Maximum suppression of the rate is already achieved by the addition of 5.0 X lom3 M H2EDTA2-. OH
OH
Ip-
N
I HO . + H-C-H I H-C-H
-
+
I
Fe(II1)
+ R.
-
OH Fe(I1)
+
I/”
+N
II
(15)
F-”
H-C-H
I
a/
‘0-
The good agreement of the slopes defined by Figures 3 and 4 and the relationship of eq 8 shows that the suppression caused by free imidazoles in the reaction medium is due to the reversal of reaction 1 rather than a redox inhibition as found for H2EDTA2-. If imidazyl radicals were formed and if these radicals were reducing for Fe(III), there would be a positive intercept shown in Figures 3 and 4. The absence of an intercept other than zero is in agreement with the concept that imidazyl radicals are only oxidizing for Fe( 11) and therefore kinetically equivalent to HO.. These observations are consistent with Wang’s postulates concerning the reactivities of imidazyl species.6 Acrylonitrile is also a scavenger for HO. as shown in eq 16. The resultant radical must be rapidly reducible by (CN)5Journal of the American Chemical Society
/ 99:20 /
+
‘i tl
HO-C-$-C=N
k
(16)
I I
H
F e ( i m ~ ) ~The - . rate obtained in the presence of [acrylonitrile] = 6.0 X M reproduced the rate obtained for the same mixture in the buffer alone within f l % : [ ( C N ) ~ F e ( i m z ) ~ - ] i = 4.90 X [imzItot= 1.42 X p = 0.10, T = 25.0 “C, pH phosphate 6.86 (kobsd = 5.78 X s-I). Hydrogen Ion Dependence. Assignment of step 2 in Scheme I as the rate-limiting step is in agreement with the absence of a hydrogen ion term in eq 8. The oxidation of Fe(CN)5H203by H202 is independent of pH from 5.0 to 9.0.’’ Substitution of HO2- is slower than H202 on (CN)5FeH203-. At higher pH values Davies reports a drop in the rate consistent with
1
+-[H30+1 A a
where k2’ is the analogous redox process with HO2- as step 2 for H202. Ka for H202 at 25.0 “ C is 2.15 X M.25Our attempts to study the ( C N ) ~ F e ( i m z ) ~system under conditions where any appreciable amount of H02- is present gave complicated results. At the higher pH values (>9.5) the initial rates are strongly suppressed in agreement with the data of Davies for the lower reactivity of H02-. The experimental difficulties originate in comparable affinities of OH- and imidazole for the Fe(II1) oxidation state in the product (equilibria 18 and 19). 2(CN)5FeOH3- + 2 0 H -
(14)
I
C , -N \
u
( C N ) ~ F e ( i m z ) ~ - OH- + (CN)5FeOH3-
I
.C-H H-y-H
,C=C
u
+
bo
N H,O
,
H‘
2.23b
[H202]= 0.100, p = 0.10 phosphate, T = 25.0 ‘C. limidazole complex same conditions.
+
H
+ imz
+ Fe2(CN)lo4-
(18)
(19)
(CN)5FeOH2-, ( C N ) 5 F e ( i m ~ ) ~ -and , Fe2(CN) to4- were present in the equilibrium reaction mixture in amounts dependent upon the total free imidazole present, reaction time, and the pH. The initial period of the reaction is obscured by the inhibition period and there are problems presented by the 0 2 evolution due to the formation of (CN)5FeOH3- in the presence of H202. For these reasons, studies at lower hydrogen ion conditions were not pursued further. No attempt was made to explore the pH range below 6.0 because the competition of H3O+ for imidazole compared to (CN)5Fe3- is favorable in solutions below pH 6.0.5 Reaction Stoichiometry. The evolution of 0 2 is observed in standing solutions of ( C N ) ~ F e ( i m z ) ~in- the presence of H202. Table IV illustrates four attempts to define the ratio of H202: ( C N ) ~ F e ( i m z ) ~In - . the presence of a ninefold excess of free ligand, designed to prevent formation of p-cyano bridged Fe(I1) and Fe(II1) species, the amount of H202 consumed was always greater than 0.500 and dependent on the temperature and reaction conditions. Experiments were also conducted in the presence of 0.01 M H2EDTA2- to scavenge trace metals in buffer salts in order to reduce the rate of H202 thermal decomposition. The data extrapolated to zero time gave a value of 0.51 f 0.10 for [H202] consumed: [ ( C N ) ~ F e ( i m z ) ~ - ] formed. The limiting stoichiometry for oxidation of ( C N ) ~ F e ( i m z ) ~by - H202 is 2.1. The product, (CN)5,Fe(imz)2-, and species at equilibrium with ( C N ) ~ F e ( i m z ) ~ produce a substantial catalase activity of 0 2 . The catalyzed decomposition of H202 is responsible for values in Table IV exceeding the 0.500 predicted stoichiometry. The data presented in Figure 5 for the catalytic decomposition of H202 at pH 6.86 may be given by the rate law -d[H2021 = kd[H202]2[(CN)5Fe(imz)2-] (20) dt
September 28, 1977
6559 Table IV. H202 Consumed in Reduction of (CN)SFe imz3Run 1 A0 IB IC
ID IE IF IC 2Ah
2B 2E 3AC 3B 4Ad 4B
Reaction time 5 min I5 min 30 min 45 min 1.0 h 1.5 h 2.0 h 1.0 h 3.0 h 5.0 h 5.0 h 8.5 h 3.0 h 6.0 h
[(CN)SFe imz2-] x 102
% Fe total as (CN)SFe imz2-
1.13 1.28 1.37 1.39 1.43 1.45 I .49 1.38 1.52 1.55 1.81 1.91 1.72 1.81
72.9 82.6 88.4 89.1 92.2 93.5 96.1 89.0 98.1 100.0 90.7 95.5 86.5 91.0 1.55 x 3.90 x 1.56 X 24
M
301
0
A -
C
50
1
1
1
100
t i r e (hrs)
Figure 5. Catalase activity of ( C N ) s F e ( i m ~ ) ~0- :, [(CN)sFe(i~nz)~-] = [ti11 = 4.86 x 10-3 M; m, [ I I I ] = 4.86 x 10-3 M; 0 ,[ I I I I = 2.38 x 10-3 M, [tert-butyl alcohol] = 1.14 X M; A, [ I l l ] = 0; i n each [H202]o = 0.1 14 M, [EDTA] = 0.01 M, = 0.075 phosphate/EDTA, T = 25.0 O C , protected from light.
Catalase activity experiments were conducted in the dark at 25.0 OC, p = 0.10, in phosphate buffer. Analysis for H202 was carried out by titration with Ce(1V) after separation of H202 by anion exchange as described in the Experimental Section. Figure 5 shows the second-order behavior for the decomposition of H202 at two different concentrations of (CN)5Fe(imz)2- and in the presence and absence of tert-butyl alcohol. Walling has shown that the catalytic decomposition of H202 by F e ( H ~ 0 ) 6 ~proceeds + via a radical chain pathway which is suppressed by HO. scavengers such as acetone and tert- butyl alcohol.24 The specific third-order decomposition of H202 rate, kd, is 8.4 X lo-* M-2 s-l in the absence of tert-butyl alcohol. Measurements of the H202 decomposition rate are corrected for all decomposition pathways determined for the blank. With tert-butyl alcohol added at a level of [H202]o/[tert-butyl alcohol]^ = IO, a 7.6% decrease is observed in kd. The small magnitude of the effect exerted by tert-butyl alcohol is consistent with competitive scavenging of HO. by H2EDTA2- and tert-butyl alcohol. A similar effect of H2EDTA2- and tert-butyl alcohol is observed in the catalase activity of Fe(EDTA)-.24 A rate law for the catalytic decomposition of HzO2 by an Fe(II1) complex is generally other than a [H202I2 dependence unless the hydroxyl radical chain is strongly retarded by HO- radical scavengers.l5.l6 Effect of Added (CN)sFe(imz)*-. An attempt was made to determine the effect of the Fe(II1) product on the rate shown by the curves in Figure 6. The maximum enhancement is a function both of added ( C N ) ~ F e ( i m z ) ~and - the amount of
[HzOz] cons at t [(CN)=,Fe
0.61 0.91 1.19 1.24 1.53 1.67 1.81 1.43 I .65 1.70 1.24 1.92 1.80 2.25 b
a 1.55 X 4.12 X 1.53 X 24
[ Fe(CN)simz3-], [H202li [imzltot Temp, OC
102[H2021 consumed
0.54 0.7 1 0.87 0.89 1.07 1.15 1.22 1.04 1.09 1.10 0.68 1 .oo
1.05 1.24 d 1.99 X 1 0-2 M 4.27 X IO-2 0.184 22
C
2.00 x M 4.20 X lo-’ 0.184 18
M
imz2-],
free imidazole in the reaction medium (Table V). Increasing [Fe(III)]o by a factor of nearly 10 at constant imidazole shows a rate increase in runs B and D of about 30%. However, in experiments C and D at twice the concentration of [Fe(III)]o and a 1.26 increase in imidazole shows a decrease of 20% in rate of the ( C N ) ~ F e ( i m z ) ~oxidation by H202. The results are in qualitative agreement with Scheme I. Increasing the concentration of ( C N ) ~ F e ( i m z ) ~increases the available pool of (CN)5FeH202-. The cycle shown by Scheme I1 provides an alternate source of HO2- and HO. which may contribute to the oxidation of ( C N ) ~ F e ( i m z ) ~ The - . effect of the suppression caused by imidazole via equilibria 1 and 7 in Scheme I is evident for runs C and D. I n the absence of added (CN)5Fe(imz)2- increasing free imidazole 1.38-fold suppresses the rate by 10%(runs E and A). It should be noted that all of the data presented in Figure 6 and Table V were obtained in the presence of H2EDTA2-. H2EDTA2- was added to assure chelation of any Fe(H20)63+ which might be generated from thermal decomposition of the iron complexes.26A sequence similar to Scheme I1 has been given by Barb, Baxendal, George, and H a r g r a ~ for e ~ ~the ferric ion catalyzed decomposition of H202. Scheme 11
(CN)5FeH202(CN)5FeH203-
+ H202
+ H202
HO2. t (CN)5FeH202-
-+ -
(CN)5FeH203t H O y t H3O+
(21)
(CN)5FeH202HO.
(22)
+
+ OH-
H30+ (CN)5FeH203-
+0 2
(23)
+ (CN)~Fe(imz)~- (CN)~Fe(imz)~+ 0 2 t H+ HO. + ( C N ) ~ F e ( i m z ) ~ - (CN)SFe(imz)2-
(24)
t HO-
(25)
H02.
-
H202 Dependence. Equations 1 1 and 12 predict a nonlinear dependence on [H202]; at low [H202] where (1 A ) approximates 2.0 a linear dependence in [H202] is anticipated. This effect is observed for all of the data in Figure 6 at the lower concentrations (10.03 M). At [HzO,] > 0.10 M a slight suppression is observed. At the high H202 limit eq 12 predicts a linear dependence in [H202] at one-half the slope of the low
Bowers, Kouacs, Shepherd
+
/ Oxidation of Imidazolepentacyanoiron(II)
6560 Table V. Effect of Added (CN)~Fe(imz)l- and Imidazole on the Oxidation of ( C N ) s F e ( i m ~ ) ~ - " 1O4[Fe(l1I)]~
[Fe(Wlo/ [Fe(II)lo
103[imz]f,,,
102kli,it
Run
1.17 4.98 4.00 5.02 0.86
3.0 5.5 9.0 1.2 4.4
A
0.00 0.863 4.66 9.33 0.00
0.000 1.223 6.05 12.10 0.00
B C D E
25.0 "C, p = 0.10 phosphate, [H2EDTA2-] = 0.01 M, [Fe(CN)simz3-]o = 1.30 X M, pH 6.86.
Table VI. Temperature Dependence of Imidazole Dissociation' Ligand
O o y .
40 ~
103/T,K-l
25.0 30.0 32.8 35.0 40.0 25.0 30.0 35.0 40.0
3.354 3.299 3.268 3.245 3.193 3.354 3.299 3.245 3.193
imza
/,
2.0-
Temp, OC
80 ,
[H202] X 1O2
,
12' 0
160 c
N-CH3imzb
Figure 6. Effect of added (CN)~Fe(imz)*-and imidazole on the (CN)5F e ( i m ~ ) ~ - / H 2 0reaction: 2 conditions as described in Table V , run A m, run B 0, run C 0 , run D 0,run E 0 ;solid lines are for illustrative purposes only.
limit domain. The effect of the slight suppression may be interpreted as all HO-species being scavenged by H202 (eq 5). The level of HOT produced is then sufficient to produce a slight inhibition by steps 23, 24, and 3. Since the reduction of Fe(II1) species by HOz- was not considered in the derivation of eq 12, it is no longer valid above [H202] = 0.10 M where [H202]0/[(CN)sFe(imz)~-]0 2 lo3. Under these conditions assuming an estimate of 1.1 X lo4 M-' S - I for kq, virtually all HO. is scavenged by H202 forming HOT. Steps equivalent to (23) and (24) will increasingly retard the rate of production of (CN)SFe(imz)2- as larger fraction of inhibitory events occur at high [H202]. Dissociation of (CN)SFe(imzp- and (cN)~Fe(N-cH$mz)~-. The value of kl was measured at pH 6.86, p = 0.10 (phosphate), under Ar in thermostated cells. The dissociation step was monitored by scavenging of the (CN)5FeH203- by a large excess of 2-methylpyrazine. Absorbance-time data were taken at 450 nm, the maxima of (CN)~Fe(2cH3pz)~-. The rate was found to be independent of the scavenging ligand at [2CH3pz] = 0.0 100, 0.025 1, 0.0501, and 0.100 M. The reaction is dissociative and follows the sequence in Scheme 111.
103k, s-I 2.03 3.45 4.90 6.23 11.1 3.24 5.85 9.48 16.3
-Ln(k/T) 1 1.90
11.38 11.05 10.81 10.25 11.43 10.86 10.39 9.86
Average of two experiments. a Average of three experiments. [2CH3pz] = 0.100 M, [ ( C N ) ~ F e ( i m z - R ) ~ - ]=o 1.40 X M, p = 0.10 (phosphate), pH 6.86, [imz-RItot = 1.0 X IO-* M.
Table VII. Activation Parameters for Imidazole Dissociation Lieand
1 0 3 k , ~ s~- l.
AH+, kcal/ mol
AS,+ call mol K
Imidazole I-Methvlimidazole
2.20 f 0.21 3.24 f 0.22
21.5 f 0.5 19.1 f 0.5
+I f2 -5 f 2
X s-' at 25.0 OC. The temperature-dependent data are presented in Table VI for the ( C N ) ~ F e ( i m z ) ~and - the analogous 1-methylimidazole complex. Activation parameters obtained from the Eyring rate theory are summarized in Table VII. The 1-methylimidazole complex exhibits a lower AH* by 2.4 kcal/mol and a more negative AS* by 6 units (Table VII). We propose that the difference in the activation parameters originates in a cage effect of solvent water for ( C N ) ~ F e ( i m z ) ~and - ( C N ) ~ F e ( N - c H 3 i m z ) ~ -Imidazole . is more stable in solution than 1-methylimidazole because the hydrogen atom on the pyrrole nitrogen can hydrogen bond to surrounding "cage" water molecules. The methyl group of N-methylimidazole cannot. The established polarity of the N H of the pyrrole moiety makes imidazole a slightly better T acScheme 111 ceptor. The dissociation of imidazole costs more in AH* to ki overcome the T interaction. The same factors control the more ( C N ) s F e ( i m ~ ) ~ - H2O (CN)5FeH203- imz k-i negative aS*observed for removal of 1-methylimidazole from its (CN)5Fe3- complex. Imidazole is less disruptive of solvent (26) structure in transferring from the caged, more solvated, k27 (CN)5FeH2O3- 2CH3pz e( C N ) ~ F ~ ( ~ C H ~ ~ Z ) complex ~into the bulk solvent. The 1-methylimidazole costs k-21 more in solvent reordering in its dissociation which is observed H 2 0 (27) as a more negative M*. Formation of (cN)~Fe(N-CH3imz)~-. The formation of the Using the steady-state approximation, the rate for ligand 1 -methylimidazole complex was studied by the competition exchange is given by technique with 2-methylpyrazine as the indicator ligand at - d [ ( C N ) ~ F e ( i m z ) ~ - ]constant concentration. The observed rate constant for - klk27[pz] k-lk-z7[imz] the formation of the kinetically distributed mixture of dt [ (CN)5Fe(i m ~ ) ~ - ] k-I [imzl + kz-i[pzI ( c N ) ~ F e ( 2 C H 3 p z ) ~and - ( c N ) ~ F e ( N - C H 3 i m z ) products ~(28) obeys Under conditions where klk27[pz] >> k-\k-zY[imz] and kobsd = ~ ~ c H , ~ ~ [ ~ C HkN-CH,imzIN-CH3imZ] ~PZI (29) k-l[imz]
