J . Phys. Chem. 1985, 89,122-128
722
on values of m/n,AC,/n, and AV/n are shown to be important, especially when such values are compared at the cmc.
Financial assistance from the Department of Chemistry, Brigham Young University, is also acknowledged.
This work is contribution No' 268 from the thermodynamic research laboratory at the National Institute for Petroleum and Energy Research, formerly the Bartlesville Energy Technology Center, Department of Energy, where the research .was partially funded by the Enhanced Oil Recovery Program, and in cooperation with the Associated Western Universities, Inc.
Registry No. C,N(C),Br, 1943-1 1-9; C,,N(C),Br, 2082-84-0; CIONC], 143-09-9; C,,NBr, 32509-44-7; C,oNI, 60734-65-8; CIoNN0,, 60834-13- 1 ; C8S04Na, 142-31-4; C 8 ~ ~ , N 5324-84-5; a, C,so4Na, 142-87-0; CI,$03Na, 13419-61-9; C12S0,Na, 2386-53-0; C7COONa, 1984-06-1; C9COONa, 1002-62-6; C~COOK,13040- 18-1; C ~ C O O R ~ , 52509-45-2; CllCOONa,629-25-4; C,,COOK, 10124-65-9; CllCOORb, 26121-34-6; ClIC0OCs, 14912-89-1.
Radiolytically Induced One-Electron Reduction of Methylvlologen In Aqueous Solution. Reactivity of EDTA Radicals toward Methylviologen' Quinto G. Mdazzani,*2 Istituto di Fotochimica e Radiazioni d'Alta Energia, Consiglio Nazionale delle Ricerche, 401 26 Bologna, Italy
Margherita Venturi, Istituto di Scienze Chimiche, Facolth di Farmacia, Universith Universitii di Bologna, 401 26 Bologna, Italy
and Morton Z. Hoffman* Department of Chemistry, Boston University, Boston, Massachusetts 02215 (Received: August 8, 1984; In Final Form: November 1, 1984)
The reaction of radiolytically generated .OH, .O-, and H. with EDTA produces EDTA radicals via electron and H-atom transfer. At pH 12.5, the EDTA radicals, which are believed to be localized on the carbon atom a to the carboxylate group and formed directly via H abstraction or indirectly via deprotonation of the carbon atom adjacent to the nitrogen radical site in the oxidized amine moiety, reduce MV2+ to MV+. quantitatively with k = 2.8 X lo9 M-I s-l . At pH 8.3, MV'. is generated nearly quantitatively; -90% of the EDTA radicals react directly with MV2+ (k = 7.6 X lo8 M-' s-l) while the remainder undergo [MV2+]-independentunimolecular conversion (k = 2.0 X 104 s-l) to form reducing species. At pH 4.7, at least 30% of the EDTA radicals reduces MV2+directly ( k = 8.5 X lo6 M-' s-') while at least 50% of the radicals converts (k = 2.3 X 104 s-I) to reducing equivalents. In all cases, the G values of MV+. formation are initially at a maximum; continued irradiation causes the yield of further MV+. to decrease due to back-reaction with EDTA radicals. The relevance of the results to the photochemical model system containing Ru(bpy)32+,MV2+,and EDTA is explored.
Introduction In recent years, ethylenediaminetetraacetate (EDTA) has been widely used as a sacrificial electron donor in model systems that are capable of promoting the photosensitized reduction of water to H2.3 As is shown in Scheme I, the role of EDTA in these systems is to scavenge the oxidized form of the photosensitizer (PS+)in order to prevent its rapid reaction with the reduced form of the electron relay species (R-) that results from the electrontransfer quenching of the excited photosensitizer (PS*). As a result of the removal of PS+, annihilation of high-energy species R- can be avoided, permitting it to react with a suitable redox catalyst generating dihydrogen. The most popular and well-studied model system is the one containing Ru(bpy):+ (bpy = 2,2'-bipyridine) as the photosensitizer, methylviologen dication (1, l'-dimethyL4,4'-bipyridinium ion; MV2+) as the electron relay, and EDTA as the sacrificial donor. Studiese6 of the mechanism of formation of the reduced (1) Research supported in part by Consiglio Nazionale delle Ricerche of Italy and in part by the Office of Basic Energy Sciences,Division of Chemical Sciences, U S . Department of Energy. (2) Visiting Scholar, Boston University, 1984. (3) (a) 'Photochemical Conversion and Storage of Solar Energy"; Connolly, J. S. Ed.; Academic Press: London, 1981; (b) 'Photochemical Conversion and Storage of Solar Energy"; Rabani, J., Ed.; Weizmann Science Press: Jerusalem, 1982; (c) 'Photogeneration of Hydrogen"; Hamman, A., West, M.A., Eds.; Academic Press: London, 1982; (d) "Energy Resources Through Photochemistry and Catalysis"; Griitzel, M., Ed.; Academic Press: New York, 1983.
Scheme I
4
h
EDTA+
EDTA
@
1
methylviologen radical cation (MV'.) have indicated that the radical resulting from the oxidation of the EDTA by R ~ ( b p y ) ~ ~ + (EDTA,+) can be involved in processes that eventually lead to the generation of a second equivalent of MV'., or to the oxidation of the MV+. back to MV2+. Because of the importance of understanding in detail all the processes occurring in this important model system, we have investigated extensively the following aspects of the reaction in aqueous solution: the stability of MV+. as a function of pH,' the stoichiometry of the platinum-catalyzed formation of H2 from (4) Keller, P.; Moradpour, A.; Amouyal, E.; Kagan, H. B. Nouo. J . Chim. 1980,4, 377-384.
(5) Miller, D.; McLendon, G. Inorg. Chem. 1981, 20, 950-953. (6) Amouyal, E.; Zidler, B. Isr. J . Chem. 1982, 22, 117-124. (7) Venturi, M.; Mulazzani, Q.G.; Hoffmann, M. Z. Radiat. Phys. Chem. 1984, 229-236.
0022-3654/85/2089-0722$01.50/00 1985 American Chemical Society
One-Electron Reduction of Methylviologen in Aqueous Solution
The Journal of Physical Chemistry, Vol. 89, No. 4, 1985 123
MV+S,~ the formation of photoactive charge-transfer complexes between MV2+and EDTA (as well as other sacrificial donors)?JO and the quantum yield of MV+. formation as a function of solution medium.".l2 In this paper we present the results of our detailed study of the interaction of radiolytically generated free radicals derived from EDTA with MV2+as a function of pH; the fact that MV+. is formed from this interaction has already been reported.13
Generation of EDTA Radicals. The radiolysis of aqueous solutions generates ea;, .OH radicals, H- atoms, and molecular H2 and H 2 0 2according to eq 1 where the numbers in parentheses
Experimental Section Materials. Methylviologen dichloride (Aldrich) was recrystallized from water by addition of acetone, collected on a sintered filter, and dried over CaC1, in a vacuum desiccator. Disodium ethylenediaminetetraacetate (Merck) was used as received. 2Propanol was purified by the method of Baxendale and Wardman.I4 Distilled water was further purified by distillation from acidic dichromate and from alkaline permanganate and then fractionated in an all-silica apparatus. The pH of the solutions was adjusted with H2SO4 or NaOH (Merck, Suprapur). Nitrous oxide was purified by passage through a column of NaOH pellets and by trap-to-trap distillation. Solutions were saturated with purified N 2 0 or were degassed by standard vacuum line techniques. Procedures. Continuous y-radiolyses were carried out at room temperature on 10-25-mL samples contained in silica or Pyrex vessels that were provided with silica optical cells on a side arm when required for spectrophotometric observations; spectra were recorded with a Perkin-Elmer Model 555 spectrophotometer. The radiation source was a 6oCo Gammacell (Atomic Energy of Canada, Ltd.) with a dose rate of 1.1 krd m i d as determined with the Fricke dosimeter taking G(Fe3+) = 15.5, where G(X) represents the number of molecules of species X formed or destroyed per 100 eV of energy absorbed by the solution. Pulse radiolyses with optical absorption detection were performed using the 12-MeV linear accelerator at the FRAE Institute of C.N.R., B01ogna.l~ The solution, contained in a Spectrosil cell of 2-cm optical path placed in front of the exit window of the accelerator, was protected from the analyzing light source until necessary by a shutter and appropriate cutoff filters. The dose delivered by each pulse generated -10-5-10-6 M of transient species and was determined by using an 02-saturated 0.1 M KSCN solution and taking Gc = 2.15 X lo4 at 500 nm1.I6 Analyses. The concentration of MV+. generated by the radiation was determined spectrophotometrically by taking e602 = 1.37 X lo4 M-' cm-l.I7 The yields of gaseous products were determined by standard vacuum line techniques and a DAN1 3200 gas chromatograph equipped with Poropack or molecular sieve columns. In this work, when NzO-saturated solutions were irradiated, relatively small amounts of COz were produced in the presence of a large excess of N 2 0 . In order to separate CO, from N 2 0 in neutral or slightly acidic solutions, the gas was removed from the irradiation vessel by means of an automatic Toepler pump and passed through a liquid N 2 trap where C 0 2 and N 2 0condensed. These gases were subsequently transferred into a vessel containing 10 mL of frozen (liquid N2) degassed 0.1 M NaOH. Upon thawing, the solution was pumped to remove the N 2 0 ; CO, was subsequently liberated by the addition of a slight excess of acid through a mercury valve.
-
H20
-
ea; (2.8), .OH (2.8), H- (0.6), Hz (0.45), H 2 0 2(0.8) (1)
represent the G values for the species in the absence of scavenging by solutes. In the presence of solutes that scavenge these species, and at high concentrations of scavengers where reaction within the spurs occur to an increased extent, different G values may result. In NzO-saturated solution (25 mM), eaq-is converted to .OH according to reaction 2; in acidic solution, eaq- is converted to H. according to reaction 3; in highly alkaline solution, .OH is converted to eo- by reaction 4. ea,
+ N20
k = 8.7
-
X
W
.OH
lo9 M-l
eaq- + H+
k = 2.2
X
s-l
-
1O1O M-'
*OH i~
+ N, + OH-
9 0 -
pKa = 11.9
(2)
(ref 18)
H-
(3) (ref 18)
s-I
+ H+
(4)
(ref 19)
EDTA reacts rapidly with H. ( k = 6.5 X lo7 M-l s-l a t PH l)zoand with .OH in a pH-dependent manner ( k = 4.0 X lo8and 6.2 X lo9 M-' s1 at pH 4 and 10.6, respecti~ely);~"-~~ Sellersz4 evaluates k(.O- + EDTA4-) as -1 X lo8 M-' s-'. Thus, in N,O-saturated solutions containing EDTA, the primary water radicals are rapidly converted to EDTA-derived radicals. In an analogous manner, the replacement of EDTA with 2-propanol causes (CH3)&OH to be generated via reactions 5 and 6 in 85% yield; the remaining part of the reaction produces the 0 radical.2s *OH + (CH3)zCHOH
k = 1.3
X
lo9 M-'
H- + (CH3)zCHOH
k = 5.0
-+
(CH3)zCOH s-I
(5)
(ref 26)
(CH,)$OH
X lo7 M-' s-l
+ H2O + Hz
(6)
(ref 27)
Results
Continuous Irradiation. The y radiolysis of N20-saturated solutions containing 0.1 1 mM MV2+and 0.1 M EDTA at pH 12.5 results in the formation of the characteristic absorption of MV+. (Figure 1) which is stable for a very long period of time in the absence of air. At wavelengths below the clean isosbestic point at 292 nm,a decrease in the absorption due to MVZ+is discerned. The inset of Figure 1 shows that MV+. builds up quite linearly with increased radiation dose (irradiation time) during the initial (18) Anbar, M.; Bambenek, M.; Ross, A. B. Natl. Stand. Ref. Data Ser.
(8) Venturi, M.; Mulazzani, Q.G.; Hoffman, M. Z. J . Phys. Chem. 1984, 88. 912-918. - -- - - - (9) Hoffman, M. Z.; Prasad, D. R.; Jones G., 11.; Malba, V. J . Am. Chem. SOC.1983, 105,6360-6362. (10) Prasad, D. R.; Hoffman, M. Z. J. Phys. Chem. 1984,88,5660-5665. (11) Mandal, K.; Hoffman, M. Z. J. Phys. Chem. 1984, 88, 185-187. (12) Mandal, K.; Hoffman, M. Z. J . Phys. Chem. 1984,88,5632-5639. (1 3) NenadoviE, M. T.; MiEiE, 0. I.; AdziE, R.R. J . Chem. SOC.,Faraday Trans. 1 1982, 78, 1065-1069. (14) Baxendale, J. H.; Wardman, P. J. Chem. SOC.,Faraday Trans. I 1973, 69, 584-594. (1 5 ) Hutton, A,; Roffi, G.; Martelli, A. Quad. Area Ric. Emilia-Romagna 1974, 5, 67-14. (16) Fielden, E. M.; Holm, N. W. In "Manual on Radiation Dosimetry"; Holm, N. W., Berry, R. J., Eds.; Marcel Dekker: New York, 1970; p 289. (17) Watanabe, T.; Honda, K. J. Phys. Chem. 1982, 86, 2617-2619.
--.
(US.Natl. Bur. Stand.) 1973, No. 43. (19) Rabani, J.; Matheson, M. S. J . Phys. Chem. 1966, 70, 761-769. (20) Neta, P.; Schuler, R. H. Radiat. Res. 1971, 47, 612-627. (21) Lati, J.; Meyerstein, D. J . Chem. SOC.,Dalton Trans. 1978, 1105-1 118. (22) Kraljic, I. In "The Chemistry of Ionization and Excitation"; Johnson, G. R. A., Scholes, G., Eds.; Taylor and Francis: London, 1967; p 277. (23) Kachanova, Z. P.; Kozlav, Y. N. Rum. J . Phys. Chem. 1973, 47, 1190-1191. (24) Sellers, R. M., private communication. (25) Asmus, K. D.;Mockel, H.; Henglein, A. J . Phys. Chem. 1973, 77, 1218-1 221. (26) Farhataziz; Ross, A. B. Natl. Stand. ReJ Data Ser. (US.Natl. Bur. Stand.) 1977. No. 59. (27j Anbar, M.; Farhataziz; Ross, A. B. Natl. Stand. ReJ Data Ser. (US. Natl. Bur. Stand.) 1975, No. 51.
724
The Journal of Physical Chemistry, Vol. 89, No. 4, 1985
I \
t 2). Further irradiation causes G(MVt.) to diminish (Figure 2) until at a delivered dose of -20 krd, corresponding to the apparent conversion of -85% of the MV2+ initially present to MV+., the
r
Mulazzani et al.
One-Electron Reduction of Methylviologen in Aqueous Solution
The Journal of Physical Chemistry, Vol. 89, No. 4 , 1985 125
" " ' I 200
A,""
500
600
Figure 3. Spectral changes observed in the pulse radiolysis of NzOsaturated solutions of 0.01 M EDTA at pH 4.7: optical path = 2 cm; 30 ps (0); 180 ps dose/pulse = 1.67 krd. Time after pulse: 2 ps (0); (0).
//
pti 12.5
Dose. k rad
Figure 5. Absorbance at 602 nm as a function of radiation dose from the pulse radiolysis of NzO-saturatedsolutions (pH 12.5) containing 0.50 mM MV2+and 0.1 M EDTA (0);0.01 M EDTA (0);0.1 M 2-propanol (0); 0.01 M 2-propanol (m). Optical path = 2 cm.
TABLE I: C(MV+.) as a Function of Irradiation Dose and [MV2+] from the Pulse Radiolysis of NZO-Saturated Solutions Containing 0.1 M EDTA at pH 4.7 IMV2+1.mM dose. krd G,(MV+.P G,(MV+.)b 0.10 0.10 0.10 0.30 0.30 0.30 0.50 0.50 0.50 1.o 1.o 1.0
Dose. k rad
Figure 4. Absorbance at 602 nm as a function of radiation dose and pH
from the pulse radiolysis of N20-saturatedsolutions containing 0.50 mM MVZ+and 0.1 M EDTA. Optical path = 2 cm. forms via clean, uncomplicated kinetics that are psuedo first order in MVZ+([MVz+] = 0.5-1.0 mM) and independent of absorbed radiation dose (0.18-1.8 krd); k = (2.8 f 0.2) X lo9 M-'s-'. The absorbance due to MV+. was stable after the pulse and increased linearly with increasing radiation dose (Figure 4) corresponding to G(MV+.) = 7.2. In comparison (Figure 5), when 0.1 M 2propanol replaced the EDTA under the same conditions, G(MV+.) = 6.3, independent of irradiation dose; in the presence of 0.01 M EDTA or 0.01 M 2-propanol, the dose-independent G(MV+.) values were 6.8 and 6.0, respectively. At pH 8.3, less MV+. was produced for the same absorbed dose (Figure 4). At least 90% of the total absorption developed at 602 nm fitted clean pseudo-first-order (in MVZ+)kinetics with k = (7.6 f 0.6) X lo8 M-' s-I. By the use of low pulse doses, the presence of a slower first-order component, producing 10% of the total absorption, was discerned. The rate constant for this process (k = 2.0 X 104 s-') was unaffected by changes in [MVz+] (0.5-1.0 mM). Further decrease of solution pH to 4.7 resulted in a further decrease in the total yield of MV+. (Figure 4). The formation of the absorption at 602 nm occurred in two distinct steps that
-
0.18 0.56 1.64 0.18 0.56 1.64 0.18 0.56 1.64 0.18 0.56 1.64
2.2 1.7 1.4 3.4 2.7 2.1 3.4 2.9 2.4 3.9 3.7 3.3
1.8 1.1 0.3 1.4 0.9 0.6 1.6 1.3 0.7 1.8 1.3 0.8
"Yield of faster component. bG,,,(MV+.) - GI(MV+.); yield of slower component. G,,,(MV+.) is obtained from final plateau of MV+ formation. could be kinetically separated; the yields of MV+. from the fast and slow components are designated GI(MVeq) and Gz(MV+.), 'respectively. The values of G(MV+-) shown in Table I include the yields of MV+. from the reaction of e,; with MV2+ (k = 8.4 X 1O'O M-' s-1)28 in competition with the scavenging of e,; by N20; as [MV2+] is increased, the extent to which e,< escapes reaction with N 2 0 increases. Specifically, the yield of MV+. (G value) deriving from e, is approximately 0.8, 0.45, 0.3, and 0.1 for [MV2+] = 1.0, 0.50, 0.30, and 0.10 mM, respectively. Table I shows that values of both G1(MV+-)and G2(MV+-) decrease with increasing radiation dose. The value of Gz(MV+.) appears to be independent of [MV2+]at low (0.18 krd) radiation dose. Both the fast and slow components of the formation of MV+obeyed first-order kinetics ( k , and k2, respectively). As shown ~~
~
(28) Farrington, J. A.; Ebert, M.; Land, E. J.; Fletcher, K. Biockim. Biophys. Acra 1973, 314, 372-381.
726 The Journal of Physical Chemistry, Vol. 89, No. 4, 1985
Mulazzani et al.
TABLE II: Effect of [MV*+] on the Rate Constants for the Formation of MV+. from the Pulse Radiolysis of N20-Saturated Solutions Containing 0.1 M EDTA at pH 4.7
kl,* s-I
[MV2+],m M
0.10 0.30 0.50 1 .o
k2/IMV2+l, M-l s-l
k2,bs-I
(2.3 i 0.3) (1.7 f 0.3) (2.2 f 0.2) (2.9 f 0.3) av = (2.3 i 0.5)
X lo4 X lo4
(7.8 f 1.5) X IO2 (3.2 0.3j x 103 (4.3 f 0.4) x 103 (7.0 f 0.8) x 103
x 104 X lo4 X IO4
7.8 X 10.7 X 8.6 X 7.0 X av = (8.5 f
IO6 lo6 IO6 IO6 1.6) X lo6
"Radiation dose = 0.18 and 0.56 krd. bRadiation dose = 0.18, 0.56, and 1.64 krd.
in Table 11, the value of kl ((2.3 f 0.5) X lo4 s-' for 0.18- and 0.56-krd radiation doses) was virtually independent of [MV2+] (0.1-1.0 mM); the value of kl was also independent of [EDTA] (0.01-0.1 M). On the other hand, the values of k2 were proportional to [MV2+]; the mean value of the second-order rate constant for the slow process was (8.5 f 1.6) X lo6 M-I S-I. At pH 3.8 in the presence of 0.1 M EDTA and pH 3.3 in the presence of 0.01 M EDTA, the yields of MV+. are reduced further still (Figure 4); the [EDTA] at pH 3.3 was limited by its solubility. The absorbances shown are, in every case, the maximum generated.
Discussion EDTA Radicals. The reaction of .OH (and H.) with EDTA is b e l i e ~ e dto~ ~be, mainly ~~ H abstraction from the CH2 group CY to the carboxylate group, although Lati and Meyerstein21 have obtained some evidence for additional sites of attack of -OH, specifically, electron transfer from one of the carboxylate or amine groups and H abstraction from the ethylene bridge. As a matter of fact, it had been suggested earlier" that this latter process represents the principal pathway of the reaction of EDTA with .OH. It is likely that the yields of the various radicals are dependent on the pH of the solution inasmuch as EDTA exhibits pK, values32of 0.0, 1S,2.0, 2.7, 6.1, and 10.2. In the discussion that follows, the term "EDTA" will be used to designate all the forms of that substance irrespective of their states of protonation. In a similar manner, the term 'EDTA radical" will be used generically to indicate the products of the interaction of -OH with EDTA; radicals I-IV are the result of H-atom abstraction and R-C
Hp
-c02- N /cH2 'c H2- c02-
-02C-CH2\ N-CH2 -02c -CH',
EDTA
E, R-
R H -COP-
CHz-N,
R-CH2-Nt
./
CH2-CO2-
\CH2-C02-
I1
I R-CH2-N
/ \
I11
CH2-CO2-
CH2-C02*
c H 2 -co,-
C,
H2-CO
2-
R-CH-N \CH2-c02IV
electron transfer from EDTA. The formulas are shown with deprotonated carboxylate and amine group for the sake of clarity. Any discussion of the interaction of EDTA radicals with MV2+ must take place in the context of the pH-dependent complexation that occurs between MV2+and EDTA.9*10,33The equilibrium constant for the labile formation of a 1:l complex is 1.3, 13, and 18 at p H 4.7, 8.0-10.0, and 11.0, respectivel;I0 in the presence of 0.1 M EDTA the percentage of MV2+, originally present at a concentration of 1 mM, that is complexed at these pH values (29) Norman, R. 0. C.; Pritchett, R. J. J. Chem. Soc. B 1967, 378-384. (30) Bhattacharyya, S. N.; Kundu, K. P. Int. J. Radiur. Phys. Chem. 1972, 4, 31-41. (31) Walling, C. Acr. Chem. Res. 1975, 8 125-131. (32) 'Critical Stability Constants"; Martell, A. E., Smith, R. M.; Eds.: Plenum Press: New York, 1974; Vol. 1, p 204. (33) Kuczynski, J. P.; Milosavijevic, B. H.; Lappin, A. G.; Thomas, J. K. Chem. Phys. Lett. 1984, 104, 149-152.
is 12, 56, and 64%, respectively. These extents of complexation are not changed to any degree if somewhat lower concentrations of MV2+ are initially present. While it is probable that complexation has an effect on the values of the rate constants of steps in the radiolytic reduction of MV2+because of the change in the overall ionic charge it causes, neither complexation nor contact ion pairing will be explicitly shown in the reaction schemes in this paper. Reduction of Methylviologen by EDTA Radicals. p H 12.5. In highly alkaline solution, where EDTA exists as a totally deprotonated species with a charge of 4-, the EDTA radical reduces MV2+ to MV+. quantitatively. In fact, G(MV+.) is 15% larger when 0.1 or 0.01 M EDTA is used as the -OH scavenger than when the same concentrations of 2-propanol are employed, a feature attributed to the inability of the @ radical from 2-propanol to effect the reduction of MV2+. G(MV+.) is increased by -6% when [EDTA] (or [2-propanol]) is increased from 0.01 to 0.1 M, a result of the increased scavenging of radicals within the radiation spurs. The quantitative formation of MV+. at pH 12.5 occurs via a single, fast ( k = 2.8 X lo9 M-I s-I ), uncomplicated bimolecular reaction between MV2+and the products of the interaction between -OH/H. and EDTA suggesting that either a single EDTA radical is involved or that the reactivities of different radicals are the same. The fact that a quantitative yield of G(MV+.) is obtained in the presence of EDTA at pH 12.5 suggests that radiolytically generated H202(G = 0.8) is destroyed by a sequence of reactions (MV+- H202 MV2+ + -OH OH- followed by the scavenging of -OH by EDTA and the reduction of MV2+by EDTA radicals) similar to those we have suggested* to occur in alkaline solutions containing 2-propanol. However, others have obthat MV+. is scavenged by H202with a complex mechanism and that the reaction in the presence of alcohols and excess H202does not proceed with the production of .OH under pulse radiolysis conditions. Under our conditions of continuous radiolysis where no excess H202is added, the system is MV+.-rich rather than H202-rich which can account for the difference in the observed behavior. It appears that radical I is the major species reactive toward MV2+ at pH 12.5. Radical 11, which is identified as EDTA,,' in the photochemical system! can undergo deprotonation of the carbon atom adjacent to the nitrogen radical site to yield radical I (reaction 7). Electron transfer from a carboxylate group of EDTA to form radical I11 must be regarded as being only a minor process, inasmuch as the interaction of .OH with oxalate ion, where there are no H abstraction or amine sites, is rather slow ( k = 1
-
-
+
X
lo7 M-I
s-I).*~
+
-
O n e c a n visualize radical IV undergoing a n
internal rearrangement involving intramolecular H-atom transfer from the carbon CY to the carboxylate group to the carbon radical site on the ethylene bridge to yield radical I (reaction 8). The quantitative reduction of MV2+ can be described by reaction 9. The ultimate fate of the two-electron oxidized form of EDTA would be the formation of hydrolysis products such a s ethylenediaminetriacetate, formaldehyde, and glyoxylate. As Figure 2 illustrates, MV+. builds up quantitatively upon continuous radiolysis at pH 12.5 during the initial portion of the reaction with the rate of further formation decreasing as [MV'.] (34) Levey, G.; Ebbesen, T. W. J. Phys. Chem. 1983, 87, 829-832. (35) Brandeis, M.; Nahor, G. S.; Rabani, J. J . Phys. Chem. 1984, 88, 1615-1623.
One-Electron Reduction of Methylviologen in Aqueous Solution CH2R-
CH2-N
Cop-
t/
'CH2-
.
COT
/CH2-Co2-
R-CH-N,
CH2-CO2-
-
,tH-C02CH-CO2-
-
R-CHp-N\
t MVpt
-
H -COO-
~
R-CHz-N,
t Ht (7)
R-CH2-N\
CH2- C02-
/
EH-CO2-
(8) CH2-COf
CH
-Copt MV'.
R-CH2-Nid
(9)
\c H2 -cot increases. This decrease in G(MV+.) with increasing dose and, for the same absorbed dose, decreasing [MV2+] can be explained simply in terms of a competition between reactions 9 and 10. Unfortunately, the products of reaction 10 are not known; we do know that MVO, which is generated upon the reaction of MV+. with e,- and (CHJ2CO- in alkaline solution,' is not formed here.
R-
CHp-N
/
6H-COZ-
t MVt*
\CHp-COz-
-
C , H2 \
--COP +
CHp-COz-
(10)
products
CO2-q t MV2'
+pt
R-CH2-N
-
At pH 8.3, the initial yields of MV+. are nearly quantitative; when low doses are used and back-reactions of MV+. are minimized, the yield of radical I11 appears to be 10%of the total yield of reducing radicals. The remainder of the reducing radicals is made up of radical I formed directly or via radicals I1 and IV. At pH 4.7, the G value of formation of radical I11 is -3 if the values reported in Table I as G1(MV+.) for low doses are corrected for the contribution of e,- to the formation of MV+.; the maximum G value of formation of radical I from G2(MV+-)is -2. The maximum value of G(MV+-) from continuous radiolysis (-4) is consistent with the total yields of radicals I and I11 (-5) when consideration is made of the rapid oxidation of MV+. by the EDTA radicals. The extent of back-reaction increases with decreasing pH (Figure 4) because the ability of radical I to act as a reductant toward MVZ+decreases; at the same time, the ability of radical I to act as an oxidant toward MV+. increases (reaction 13).
-
R-CH2-N
,CH-COpC'
H2- CO;
t MV'.
-
R-CH2-N
p H 8.3 and 4.7. In mildly alkaline solution, EDTA exists essentially as a monoprotonated species. This protonation effects a number of changes in the interaction of EDTA with the electrophilic .OH radical. Electron transfer will be suppressed, and the reduction of the inductive effect of the amine and/or carboxylate groups will render the abstraction of a hydrogen atom from the a-carbon less favored. As a result, there will be a decrease in the probability of electron transfer and hydrogen abstraction processes that form radicals I and 11, with a concomitant increase in the probability of formation of radicals I11 and IV. Further decrease in pH to 4.7 results in EDTA being predominantly in the diprotonated form. Here, the yields of radicals I and 11, relative to I11 and IV, will diminish further. At pH 8.3 and 4.7, the formation of MV+. occurs via two kinetically separated first-order processes: one with a rate constant that is proportional to [MV2+] and strongly dependent on pH, and one in which the contribution to the total yield of MV+. increases with decreasing pH with a rate constant that is independent of [MV2+], [EDTA], and pH. It is clear that a t least two radicals cause the reduction of MV2+in slightly alkaline and acidic solutions: Radical I is identified with the former process for which k9 = 7.6 X lo8 M-I s-' at pH 8.3. As the pH is lowered to 4.7, further protonation of the amine or carboxylate groups adjacent to the radical site of radical I occurs causing it to become a weaker reducing agent toward MV2+; k = 8.5 X lo6 M-ls-l. The value of 1.4 X lo9 M-' s-l previously reported13 for the reduction of MV2+ by EDTA radicals at pH 5 could not be reproduced and no explanation is offered for the large discrepancy. The pH- and [MV*+]-independent path to MV+. requires a reducing radical which is formed in a unimolecular reaction (k = (2.0-2.3) X 104 d ) , which reduces MVZ+at a rate at least ten times faster than the rate of reaction 9, and for which the yields increases with decreasing pH. These requirements rule out radical I, formed from Radicals I1 and IV (reactions 7 and 8), as this reducing agent. One possible candidate is the C o p radical that may form from radical I11 via reaction 1 1 . The reaction of Cop with MV2+ (reaction 12) is known to be rapid (k = 1.6 X 1O'O M-I s-~).,~We pointed out above that radical I11 is not expected to be produced to a significant extent in alkaline solution where EDTA is fully deprotonated. However, in mildly alkaline and acidic solution, the formation of radical I11 via H-atom abstraction from a protonated site might compete with the other .OH reactions. R-CHp-N
The Journal of Physical Chemistry, Vol. 89, No. 4, 1985 121
cg.
(11)
H*
CHz-COz/
t MV2* (13)
'CH2-COC
The G values for the gaseous products determined a t p H 4.7 for N20-saturated solutions of EDTA in the absence and presence of MV2+ provide some insight into the reactions of the EDTA radicals. In the first place, the value of G(N2) obtained in the presence of EDTA (3.1) is the same as that obtained in the radiolysis of N,O-saturated water; clearly, 0.1 M EDTA does not affect the scavenging of e,, by NzO. The decrease in G(N2) to 2.7 in the presence of 0.50 mM MV2+ is due to the competitive scavenging of e, - by MV2+. G(H2) = 0.9 in the presence and absence of MV2$; this value results from the radiolytic yield of H2 (C = 0.3 in N,O-saturated solutions)36 and the scavenging of radiation-generated H atoms (G = 0.6) by EDTA. The value of G(C02) = 4.8 in the absence of MVZ+indicates that decarboxylation of the EDTA radicals is a major process occurring during their natural decay. In the presence of 0.50 mM MV2+ and 0.1 M EDTA at p H 4.7, G(C02) for a total dose of 50 krd is decreased to 1.8. Thus, it appears that MV2+ and/or MV+. intercepts EDTA radicals to a significant extent before they can undergo decarboxylation. Relevance to the Photochemical System. The reductive scavenging by EDTA of Ru(bpy),,+, generated in the oxidative quenching of *Ru(bpy)?+, almost certainly occurs via electron transfer from the amine group, thereby generating radical 11. The rate constant for the reaction is p H dependent with k decreasing with increasing solution a ~ i d i t yconsistent ,~ with the lowering of the reducing ability of EDTA as it becomes protonated. There is, however, disagreement in the l i t e r a t ~ r eas ~ , to ~ the absolute value of k. The results reported here show that the conversion of radical I1 to I and the subsequent reduction of a second equivalent of M V + is quantitative in alkaline solution. Of greater interest is the behavior of the system at pH 4.7; the yield of H2 from the interaction of colloidal Pt with MV+- is a maximum at pH ~ 5 . 6Unfortunately, it is not possible from the results reported here to make the same analysis for acidic solutions due to uncertainties in the yield of radical 11, the rate of conversion of I1 to I, and the degree of competitive decarboxylation of the EDTA radicals. In our recent study12of the photochemistry of degassed solutions (pH 3.7-11.0) containing 0.10 mM Ru(bpy):+, 0.020 M MV2+, and 0.10 M EDTA, we found that MV'. was initially generated linearly with increasing irradiation time, indicating that sufficient EDTA was present to effect reduction of Ru(bpy),,+ under all p H conditions and that MV+. was not scavenged by EDTA radicals. The initial quantum yield of MV+. formation was 0.22 at pH 7.0-11.0,0.20 at pH 5.6,0.14 at pH 4.7, and 0.030 at pH 3.7. The decrease in the value of the quantum yield in increasingly
\CHz-CO;
MVt* t
CO2
(12)
(36) Mahlman, H.A. J . Phys. Chem. 1966,70,3983-3987.
728
The Journal of Physical Chemistry, Vol. 89, No. 4, 1985
acidic solution can be attributed to the introduction of competing acid-catalyzed modes of decay of radicals I and 11, and to a decrease in the efficiency of release of MV+. from the solvent cage in which it is generated in the electron transfer quenching reaction.
Mulazzani et al. The results reported here and in the photochemical studyL2emphasize that the optimum conditions for the generation of MV+. in photochemical model systems are high concentrations of Ru(bpy),*+, MV2+, and EDTA in neutral or alkaline solution.