Electron-transfer component in hydroxyl radical reactions observed by time resolved resonance raman spectroscopy

Article abstract of DOI:10.1021/ja9800838

The existence of an electron-transfer pathway in the reaction of ·OH radical with aromatic molecules in water has been established, for the first time, using time-resolved resonance Raman spectroscopy as a diagnostic tool and p-dimethoxybenzene as a model system. In the currently accepted mechanism, the cation radical is produced by ·OH addition to the ring, followed by loss of OH^-. The present work demonstrates that this process competes with direct electron transfer. A generalized reaction mechanism has been proposed in terms of potential energy diagrams to explain two-step formation of the cation radical. In this reaction mechanism, the electron- transfer component and the rate of OH^- elimination from the ·OH adduct both depend on the ionization potential (IP) of the molecule. The cation radical yield by electron transfer increases from 6% in p-dimethoxybenzene to 30% in p-anisidine and 85% in p-phenylenediamine. For neutral molecules with IP > 8 eV, the ·OH addition is the first step in the chemistry, and for IP < 7 eV, it is the electron transfer. In the intermediate IP range, both processes occur simultaneously.

Full text of DOI:10.1021/ja9800838

J. Am. Chem. Soc. 1998, 120, 4161-4166
4161
Electron-Transfer Component in Hydroxyl Radical Reactions
Observed by Time Resolved Resonance Raman Spectroscopy†
G. N. R. Tripathi
Contribution from the Radiation Laboratory, UniVersity of Notre Dame, Notre Dame, Indiana 46556
ReceiVed January 6, 1998
Abstract: The existence of an electron-transfer pathway in the reaction of •OH radical with aromatic molecules
in water has been established, for the first time, using time-resolved resonance Raman spectroscopy as a
diagnostic tool and p-dimethoxybenzene as a model system. In the currently accepted mechanism, the cation
radical is produced by •OH addition to the ring, followed by loss of OH-. The present work demonstrates that
this process competes with direct electron transfer. A generalized reaction mechanism has been proposed in
terms of potential energy diagrams to explain two-step formation of the cation radical. In this reaction
mechanism, the electron-transfer component and the rate of OH- elimination from the •OH adduct both depend
on the ionization potential (IP) of the molecule. The cation radical yield by electron transfer increases from
6% in p-dimethoxybenzene to 30% in p-anisidine and 85% in p-phenylenediamine. For neutral molecules
•
with IP > 8 eV, the OH addition is the first step in the chemistry, and for IP < 7 eV, it is the electron
transfer. In the intermediate IP range, both processes occur simultaneously.
Introduction
the reaction involves an electron transfer (ET) component, and
what mechanistic implications that would have for the early
chemical events, has never been answered. In the commonly
held reaction mechanism, the •OH radical reacts with aromatic
molecules by addition to the ring.2-9 This behavior contrasts
The oxidation of organic substrates by the •OH radical is one
of the most widely studied reactions because of its central role
in chemistry and biology, organic synthesis, photocatalysis in
aqueous environments, wastewater treatment, and numerous
other chemical processes.1-10 The reaction is used to prepare
radical intermediates in aqueous medium, to investigate their
structure and reactivity by ESR and time-resolved Raman
spectroscopy.9,11 However, the fundamental question of whether
•
with the highly oxidizing nature of the OH radical (E°(•OH/
OH-) ) 1.9 V)12 that should generally favor electron transfer.
The adduct (hydroxycyclohexadienyl) radicals undergo loss of
OH- by reaction with H+ and/or water to form the cation radical
(adduct-mediated electron transfer; AMET). The ET component
in the reaction is difficult to recognize, and can be easily
confused with AMET, for the following reasons: (1) Experi-
mentally, a clear resolution cannot be made between the ET
† The research described herein was supported by the Office of Basic
Energy Sciences, Department of Energy. This is Document No. NDRL-
4044 from the Notre Dame Radiation Laboratory.
(1) See, Biweekly List of Papers on Radiation Chemistry and Photo-
chemistry, pubished by the NDRL Data Center, Radiation Laboratory,
University of Notre Dame, Notre Dame, IN. Farahataziz; Ross, A. B.
Selected Specific Rates of Reactions of Transients from Water in Aqueous
Solution, NSRDS-NBS 59; US Department of Commerce, 1977.
(2) Sonntag, C. v. The Chemical Basis of Radiation Biology; Taylor and
Francis: New York, 1987.
•
and AMET processes, unless the OH addition (diffusion-
controlled) and OH- elimination occur at drastically different
time scales. Obviously, the H+-catalyzed OH- loss (generally
diffusion controlled) must be avoided by making measurements
•
in basic solutions. (2) The OH adducts at the different ring
(3) Tripathi, G. N. R.; Y. Su. J. Am. Chem. Soc. 1996, 118, 2235 and
references therein.
sites decay at different rates, and not necessarily into the cation
radical. Complications arise due to the presence of several
transient species, with overlapping absorption, making identi-
fication and kinetic monitoring of the individual species quite
difficult. To establish the ET component in the reaction, it is
extremely important to select model systems in which the
number of possible •OH adducts is small, and to use a structure-
sensitive technique, such as time-resolved resonance Raman
spectroscopy, for transient identification and kinetic monitoring.
(4) Viera, A. J. S. C.; Candeias, L. P.; Steenken, S. J. Chim. Phys. 1993,
90, 881. Viera, A. J. S. C.; Steenken, S. J. Am. Chem. Soc. 1990, 112,
6918; 1989, 109, 1987. Wu, Z.; Ahmad, R.; Armstrong, D. A. Int. J. Radiat.
Phys. Chem. 1984, 23, 251. Elliot, A. J.; McEachern, R. J.; Armstrong, D.
A. J. Phys. Chem. 1981, 85, 68. Clement, J. R.; Lin, W. S.; Armstrong, D.
A. Radiat. Res. 1977, 72, 427.
(5) Neta, P. AdV. Phys. Org. Chem. 1976, 12, 2. Hapiot, P.; Pinson, J.;
Francesch, C.; Mhamdi, F.; Rolando, C.; Neta, P. J. Phys. Chem. 1994, 98,
2641. Neta, P.; Madhavan, V.; Zemel, H.; Fessenden, R. W. J. Am. Chem.
Soc. 1977, 99, 163. Neta, P.; Fessenden, R. W. J. Phys. Chem. 1974, 78,
523.
(6) Steenken, S. Top. Curr. Chem. 1996, 177, 125. Steenken, S. J. Chem.
Soc., Faraday Trans. 1 1987, 83, 113 and references therein.
(7) Jagannadham, V.; Steenken, S. J. Am. Chem. Soc. 1985, 107, 6818.
Jagannadham, V.; Steenken, S. J. Am. Chem. Soc. 1984, 106, 6542. Hazra,
D. K.; Steenken, S. J. Am. Chem. Soc. 1983, 105, 4380. Steenken, S.;
Raghavan, N. V. J. Phys. Chem. 1979, 83, 3101.
(8) Walling, C.; Camaioni, D. M.; Kim, S. S. J. Am. Chem. Soc. 1978,
100, 1603. Walling, C.; Camaioni, D. M. J. Am. Chem. Soc. 1975, 97, 1603.
Walling, C.; Johnson, R. A. J. Am. Chem. Soc. 1975, 97, 363 and references
therein.
(10) Bard, A. J.; Fox, M. A. Acc. Chem. Res. 1995, 28, 141. Fox, M. A.
Sol. Energy Mater. 1995, 38, 381. Fox, M. A. In Photocatalytic Purification
and Treatment of Water and Air; Ollis, D. F., Al-Ekabi, Eds.; Elsevier:
Amsterdam, 1995, Vol. 3, p 163. Goldstein, S.; Czapski, G.; Rabani, J. J.
Phys. Chem. 1994, 98, 6586. Brandys, M.; Sassoon, R. E.; Rabani, J. J.
Phys. Chem. 1987, 91, 953. Jaeger, C. D.; Bard, A. J. Phys. Chem. 1979,
83, 3146. Draper, R. B.; Fox, M. A.; Pelizzetti, E.; Serpone, N. J. Phys.
Chem. 1989, 93, 1938 and references therein.
(11) Tripathi, G. N. R. In AdVances in Spectroscopy, Vol. 18, Time-
resolVed Spectroscopy; Clark, R. J. H., Hester, R. E., Eds.; John Wiley &
Sons: New York, 1989; pp 157-218.
(9) Nonhebel, D. C.; Walton, J. C. Free-radical Chemistry: Structure
and Mechanism; Cambridge University Press: New York, 1974.
(12) Wardman, P. J. Phys. Chem. Ref. Data 1989, 18, 1637.
S0002-7863(98)00083-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 04/11/1998

4162 J. Am. Chem. Soc., Vol. 120, No. 17, 1998
Tripathi
yields of radicals per 100 eV of energy absorbed) on the 100 ns time
In para disubstituted benzenes, there are four equivalent ring
sites and two equivalent ipso positions. Therefore, only two
types of OH adduct and, at most, four transients, including
scale.17 The eaq reacts with N2O, converting into the OH radical in
less than 5 ns. The H• radical is a minor product of radiolysis that
reacts with aromatic molecules by addition, but does not form the cation
-
•
•
radical products of the •OH adducts, are possible in the AMET
mechanism. Therefore, they are expected to be satisfactory
model systems. The radicals derived from benzene derivatives
in water generally have strong absorption in the near-UV and
visible regions, which is essential for high detection sensitivity
in resonance Raman methods.11 In the model systems, the
radicals produced initially by the ET as well as the AMET
mechanism must be observable, so that it can be shown
kinetically that the adduct radical is not the precursor of the
cation radical. Unfortunately, not many para disubstituted
benzenes satisfy this criterion. To illustrate this point, we refer
to a recent examination of the •OH oxidation of p-diaminoben-
zene (p-phenylenediamine) in basic aqueous solutions by time-
resolved resonance Raman spectroscopy.13 In this case, the •OH
adduct at the ipso position may produce the p-aminophenoxyl
radical by NH3 elimination. However, this product radical was
not observed. The initial transient absorption spectrum that was
radical.17 The oxidizing species SO4 was prepared by the reaction
-•
-
-
2-
2-
of eaq with persulfate dianion (eaq + S2O8 f SO4 + SO4-•) in
N2-saturated solutions.18 In that case, the OH radical was scavenged
•
by adding an excess of tert-butyl alcohol (t-BuOH) in solution. The
radiation yield of SO4-• was taken to be half the yield of the •OH radical
in N2O-saturated water. The SO4-• radical generally reacts with organic
substrates by electron transfer.18 Comparison of the initial radical cation
yield in the reaction of •OH radical with that of SO4-•, after
•
compensation for the radiation yields, gives the fraction of the OH
radical involved in electron transfer. If this fraction is small, identifica-
tion of the cation radical, estimation of its yield, and determination of
its growth kinetics are almost impossible by the optical absorption
method due to overlapping absorption by other radicals present in much
larger concentrations. Therefore, the time-resolved resonance Raman
method was used for detection.11
The pulse radiolysis time-resolved Raman spectroscopic technique
used in this Laboratory for the study of short-lived chemical intermedi-
ates in solution has been described in detail elsewhere.11,19 In brief, a
Van de Graaff accelerator is used to generate 2 MeV, 100 ns electron
pulses for radiolysis. A tunable excimer laser pulse pumped dye laser
(10 ns) system is used to probe Raman scattering, and an optical
multichannel analyzer (OMA) accompanied by an intensified gated (20
ns) diode array is used for detection. The accelerator and laser are
operated at a repetition rate of 7.5 Hz, which allows efficient signal
averaging. Raman band positions are measured with reference to the
known Raman bands of common solvents such as ethanol, and are
accurate to within (2 cm-1 for sharp bands and (5 cm-5 for broad
and shoulder bands.
•
attributed to the OH adduct in the literature,14 in support of
AMET, was found to be that of the cation radical.13 The cation
radical forms in two steps, with the major fraction (85 ( 5%)
at a diffusion-controlled rate, and the minor (15 ( 5%) with a
period of ∼50 ns. Thus, ET appears to be the dominant reaction
pathway in this aromatic diamine. However, the adduct radical,
expected to appear simultaneously with 85% of the cation
radicals (ET) and produce 15% of the cation radicals on decay
(AMET), absorbs too weakly to be identified.13 While the cation
radical formation in two steps cannot be explained by AMET
•
The initial OH concentration in our experimental conditions was
•
∼10-5 M. The Raman spectra were generally excited off-resonance
and the scattering center was carefully positioned close to the bottom
of the Raman cell (within 0.4 mm), to minimize attenuation of the
probe laser and scattered signals. The experiments were mostly
performed in neutral and basic solutions, so that H+-catalyzed OH-
loss from the adduct radicals would not contribute to the cation radical
signals.
alone, the production of the OH adduct and cation radicals at
the same rate would be straightforward and convincing experi-
mental evidence for the existence of ET.15
In our search for appropriate model systems, we find that
the existence of an ET component can be unambiguously
demonstrated in the reaction of •OH radical with p-dimethoxy-
benzene (DMB) in water. This molecule does not undergo
structural modifications due to protonation/deprotonation in
moderately acidic and basic solutions, a problem associated with
most hydroxy and amino compounds. Its radical cation may
react with base,16 but even at a diffusion-controlled rate (∼1010
M-1 s-1) of the reaction, a significant decrease in the cation
radical yield on the submicrosecond time scale can occur only
at pH >10. The Raman evidence of a fast and a slow
component in the formation kinetics of cation radical in the •OH
oxidation of p-methoxyaniline (p-anisidine) is also presented.
Results and Discussion
The optical absorption spectra and rate parameters available
•
from an earlier study on the OH reaction with p-dimethoxy-
benzene20 were used to define the chemical conditions under
which cation radical formation by electron transfer (ET) could
be distinguished from that by adduct-mediated electron transfer
(AMET). p-Dimethoxybenzene is not very soluble in water,
and only 3 × 10-4 M of it could be dissolved in solution. The
•
reaction occurs with a rate constant of 7 × 109 M-1 s-1
,
It is shown that the AMET mechanism for the OH oxidation
equivalent to a rate of 2.1 × 106 s-1 for the substrate
is exclusive only for aromatic molecules with ionization
potentials greater than 8 eV. For molecules with lower
ionization potential, competition between ET and AMET
mechanisms must be considered.
concentration used in the present work. Initially, a transient
•
absorption with λmax at 300 nm, attributed to the OH adduct
(DMB-OH•), appears. In acidic solutions, the adduct radical
decays, by the reaction with H+, at a rate constant of 2 × 108
M-1 s-1, into a transient species with λmax in the 460 nm region,
assigned to the cation radical (DMB+•). The bimolecular
Experimental Procedure
The •OH radical was produced by pulse radiolysis of N2O-saturated
-
•
•
water. The radiolysis of oxygen-free water produces eaq (2.6), OH
reactions of the OH adduct occur at a rate constant (2k) of
(2.7), and H• (0.6) radicals (numbers in parentheses are G values, i.e.,
∼109 M-1 s-1, corresponding to an initial decay rate of ∼5 ×
103 s-1 for the radical concentration of ∼5 × 10-6 M-1, used
(13) Sun, Q.; Tripathi, G. N. R.; Schuler, R. H. J. Phys. Chem. 1990,
94, 2216.
(14) Rao, P. S.; Hayon, E. J. Phys. Chem. 1975, 79, 1063.
(15) The observed rates of adduct and cation radicals should be the same
and equal to the rate of •OH encounter with the molecule (diffusion-
controlled), if they are produced simultaneously. In this work, we are
concerned only with the observed rates. The second-order rate constant for
a product radical can be derived by multiplying the rate constant for the
•OH encounter with the measured yield of the radical.
(17) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
Phys. Chem. Ref. Data 1988, 17, 513.
(18) Neta, P.; Huie, R. E.; Ross, A. B. J. Phys. Chem. Ref. Data 1988,
17, 1027.
(19) Tripathi, G. N. R. In Multichannel Image Detectors II; Talmi, Y.,
Ed.; ACS Symp. Ser. No. 236 American Chemical Society, Washington,
DC, 1983; p 171.
(20) O’Neill, P.; Steenken, S.; Schulte-Frohlinde, D. J. Phys. Chem. 1975,
79, 2773.
(16) Tripathi, G. N. R. Chem. Phys. Lett. 1992, 199, 409. Holcman, J.;
Sehested, K. J. Phys. Chem. 1976, 80, 1642.

Electron Transfer Component in Hydroxyl Radical Reactions
J. Am. Chem. Soc., Vol. 120, No. 17, 1998 4163
Figure 2. Formation kinetics of the p-dimethoxybenzene cation radical
Figure 1. Transient Raman spectra obtained by excitation at (a) 465
nm, 1.5 µs after the electron pulse, and (b) 420 nm, 45 µs after the
electron pulse on pulse radiolysis of an N2O-saturated aqueous solution
of p-dimethoxybenzene (pH 7, phosphate buffer).
•
(DMB+•; 6% yield) on OH oxidation of p-dimethoxybenzene (3 ×
10-4 M, pH 7, N2O saturated). The solid circles (b) show the formation
-•
kinetics of DMB+• produced by SO4 oxidation (100% yield). The
•
-•
reaction rates of OH and SO4 are comparable.
in the optical absorption work.20 At pH >6, the rate of cation
radical formation by the reaction of adduct radicals with H+,
i.e., <2 × 102 s-1, is much slower than the bimolecular reaction
rate (∼5 × 103 s-1). The formation of the cation radical by
water-catalyzed loss of OH- from DMB-OH• was also not
without phosphate buffer in solution, but with reduced intensity,
as expected due to the rapid loss of cation radical by reaction
with OH-.16
The straightforward kinetic Raman observation of DMB+•,
•
in the DMB + OH reaction in water, at a rate comparable to
seen. Therefore, the rate was estimated to be <103 s-1 20
Thus,
.
•
that of OH adduct, provides unambiguous experimental evi-
one cannot observe the cation radical, on the microsecond and
shorter time scale, by the AMET process, at pH >6. In basis
solutions, DMB+• may react with OH-.16 However, even if
the reaction is diffusion controlled, its rate can be comparable
to or faster than the rate of •OH reaction with DMB (2.1 × 106
s-1) only at pH >10. Therefore, if DMB+• is produced in the
reaction, it should be observable, with the yield decreasing
rapidly as pH exceeds 10.
dence of the electron-transfer component in the reaction.15 As
discussed earlier, the formation of DMB+• on the 100 ns time
scale at neutral pH cannot be explained by the reaction of
DMB-OH• with H+, as this process, even at a diffusion-
controlled rate, would require 100 ms or longer time. The
water-induced dissociation of the adducts into the radical cation
must be ruled out, as the process is too slow (k < 103 s-1) to
contribute on the 100 ns time scale. The reaction of DMB-
OH• with buffer cannot be a contributing factor, as (1) buffer
concentration (1 to 10 mM) had no effect on the cation radical
yield and (2) weak cation radical signals were observed even
at pH 11, 100 ns after the electron pulse, without phosphate
buffer in solution. Most importantly, any imaginable reaction
From the above analysis, it is clear that if there is an electron-
transfer component in the reaction, the most suitable pH range
for its evaluation would be between 7 and 9.
Raman Observation of the DMB+• Cation Radical. N2O-
saturated aqueous solutions containing 3 × 10-4 M DMB at
pH 7 ( 0.2 (phosphate buffer) were subjected to pulse radiolysis.
The formation kinetics of DMB+• was monitored by time-
resolved resonance Raman spectroscopy with an excitation
wavelength of 465 nm (λmax of DMB+• at 460 nm). Contrary
to the expectations for the AMET mechanism, we observe a
transient Raman spectrum of DMB+•, with prominent bands at
∼1640 and ∼1375 cm-1 (Figure 1a). Identification of the
spectrum as that of DMB+• follows directly from comparison
with spectra obtained in chemical oxidation reactions.21 The
Raman signals evolved on the 100 ns time scale (Figure 2).
The growth half-period was measured as 300 ((50) ns, which
corresponds to a rate constant of 7.7 ((1.3) × 109 M-1 s-1 (6
× 109 M-1 s-1 by fitting the data with first-order kinetics; see
Figure 2). This rate constant matches well, within experimental
•
of the OH adduct that can give cation radical (AMET) on the
100 ns time scale will lead to its complete disappearance within
a few microseconds, and in no case will the cation radical evolve
•
at the same rate as the OH adduct.
The yield of DMB+• produced by electron transfer (ET) from
•
DMB to OH was determined by comparison with the yield
(Raman signal intensity) obtained when DMB was oxidized by
radiolytically produced SO4-• in a N2-saturated solution (DMB
+ SO4 f DMB+• + SO42-). The spectra were scaled for
-•
•
-•
the radiation yields of OH and SO4 radicals. We find that
the DMB+• cation radical yield by ET is only 6((2)%.
Therefore, it is not surprising that the cation radical formation
by ET was not detected by transient absorption and that only
the adduct absorption which does not decay even on the 100
µs time scale, was observed.20 The observed rate for the 6%
DMB+• production in the •OH reaction is compared, in Figure
•
uncertainty, with the rate constant for the formation of OH
adduct to DMB (7 × 109 M-1 s-1), available in the literature.20
The rate of DMB+• production was found to be independent of
the buffer concentration, but depended linearly on the DMB
concentration in solution. When an excess of t-BuOH was
2, with the formation rate for 100% DMB+• in the SO4
-•
oxidation (electron transfer),20 with 3 × 10-4 M of DMB in
solution at pH 7. It can be seen that both rates are diffusion
controlled and comparable. Thus, we have conclusive experi-
mental evidence that •OH oxidation can involve electron transfer.
•
added to the solution, to scavenge OH radicals, the DMB+•
Raman signals disappeared. The transient Raman signals of
DMB+• were observable even at pH 11, 100 ns after the pulse,
•
In the OH oxidation of p-phenylenediamine where the adduct
absorption is not observed, the yield of the cation radical (85%)
produced at a diffusion-controlled rate can be safely attributed
(21) Ernstbrunner, E. E.; Girling, R. B.; Grossman, W. E. L.; Hester, R.
E. J. Chem. Soc., Perkin Trans. 2 1978, 177.

4164 J. Am. Chem. Soc., Vol. 120, No. 17, 1998
Tripathi
to ET, and the remaining 15% (∼50 ns) to AMET.13 In this
respect, p-dimethoxybenzene and p-phenylenediamine are deemed
to represent two extreme cases. It would be logical to expect
p-methoxyaniline (p-anisidine) to display an intermediate
behavior. We will discuss, later on, the cation radical formation
•
on OH oxidation of p-anisidine in water.
Raman Observation of the p-Methoxyphenoxyl Radical.
We have also observed formation of the p-methoxyphenoxyl
•
radical (MPhO•) on the OH oxidation of DMB. The Raman
excitation in this case was at 420 nm (λmax of MPhO• at 415
nm).11 This transient is formed at a slower rate than the DMB+•
cation radical. The spectrum of the species recorded 45 µs after
the electron pulse, after subtraction of the DMB+• Raman
signals, is shown in Figure 1b. The Raman signals (bands at
1609 and 1518 cm-l)11,22 of this transient disappear on addition
of t-BuOH to the solution. No Raman signal is observed on
radiolysis of solutions containing 3 × 10-4 M DMB and 0.1 M
•
Figure 3. Formation kinetics of the cation radical produced on OH
•
oxidation of p-anisidine (2 × 10-3M, pH 7, N2O saturated) in water.
NaN3 (N3 radical oxidizes p-methoxyphenol but not DMB),
showing that the MPhO• radical originates from DMB and not
p-methoxyphenoxyl radical were not detected. We estimate that
at least 30% of the OH radicals produce the cation radical by
from an impurity of p-methoxyphenol in the sample. The yield
•
•
of MPhO• radical in the reaction of OH with DMB was
the first step, a yield approximately five times that in p-
dimethoxybenzene (6%) and about one-third that in p-phenyl-
enediamine (85%).25 At pH >7, the rate of cation radical
formation by the second step was found to be less than 103 s-1
in p-dimethoxybenzene and ∼2 × 107 s-1 in p-phenylenedi-
amine.13 In p-anisidine, this rate is about 9 × 105 s-1. Thus,
with an increase in the ET component in the reaction, the rate
of cation radical formation by AMET also increases. This is
an important observation that will be explained by the reaction
mechanism that we discuss in the following subsection.
determined by comparison with the yield observed in the
reaction of N3• radical with p-methoxyphenol in basic aqueous
•
solutions.11,22 The N3 radical is produced by the electron-
•
-
transfer reaction of OH radical with N3 (0.1 M). Therefore,
the radiation yield was taken to be the same as that of •OH. We
•
find that 12((2)% of the OH radicals are consumed in the
formation of MPhO•, in the OH oxidation of DMB in near
•
neutral solutions. Since this radical is formed at a slower rate
than DMB+• (microsecond time scale), it cannot be considered
•
the initial reaction product. Its logical precursor is the OH
adduct at the ipso position which converts into phenoxyl radical
by methanol elimination.23 It appears that about 94% of the
•OH radicals produced in the radiolysis of water add to the DMB
ring, in the ratio of 1:8 for ipso to nonipso positions.
Reaction Mechanism
The following reaction scheme, inclusive of the ET and
AMET steps, accounts for the experimental observations made
to date.
In contrast to p-dimethoxybenzene, the p-aminophenoxyl
radical was not observed in the •OH oxidation of p-phenylene-
diamine. This is readily understandable, as only a small fraction
M + •OH
[M+OH–
MOH]cage
MOH•
•
(15%) of the OH radicals add to the ring.13
M+• + OH–
ET
M+• + OH–
AMET
The •OH Oxidation of p-Anisidine. As pointed out earlier,
p-dimethoxybenzene and p-phenylenediamine represent two
extreme situations with respect to ET vs AMET mechanisms,
and p-anisidine is a logical intermediate case. However, the
number of •OH adducts and their transient radical products likely
to be produced doubles due to molecular asymmetry. We focus
here only on the growth kinetics of the cation radical, monitored
at its characteristic resonance Raman bands (1624, 1519, and
1400 cm-1).24 An N2O-saturated aqueous solution containing
2 × 10-3 M p-methoxyaniline at pH 7 (substrate acts as buffer)
was subjected to pulse radiolysis. The temporal evolution of
the Raman bands (excitation in resonance with λmax of the cation
radical at 445 nm) is depicted in Figure 3. It is clearly evident
from Figure 3 that the cation radical is formed in two steps.
The initial step is responsible for about 38% of the cation
radicals produced in the reaction. Both steps combined account
The two-step formation of the cation radical (M+•) in the above
reaction scheme, if analyzed in terms of potential energy
diagrams, provides a clear insight into the early chemical events
and leads to valuable guidance for identifying the chemical
systems in which electron transfer can be observed.
•
Let us consider, for the sake of simplicity, that M and OH
act as atomic species that bond on encounter. The interaction
potential at various interatomic distances can be totally, or in
part, covalent (MOH) or ionic (M+OH-), depending on the
relative energies of the corresponding electronic configurations,
•
with respect to the energy of M + OH (Figure 4). In the
covalent configuration MOH, the C-OH bond (typically 1.5
Å)26 can form due to orbital overlap between an unpaired
electron on O and a ring p(π)-electron, at the expense of the
ring CC (π) bonds. The energy difference between the CO (σ)
•
for 80% of the OH radicals (by comparison with the signals
•
observed by N3 oxidation). About 10% p-aminophenoxyl
radical11 is also produced in the reaction, but Raman signals of
(25) Because of molecular asymmetry, p-anisidine is not an acceptable
model for establishing ET, as one can postulate an extremely short-lived
adduct precursor for the diffusion-controlled component (38%) in the cation
radical formation. Experimentally, only one •OH adduct with absorption in
the 330 nm region is seen which decays at a rate of 9 × 105 s-1. The data
on p-anisidine should be viewed in the light of results on p-dimethoxy-
benzene and p-phenylenediamine.
(22) Tripathi, G. N. R.; Schuler, R. H. J. Phys. Chem. 1987, 91, 5881.
(23) Raghavan, N. V.; Steenken, S. J. Am. Chem. Soc. 1980, 102, 3495.
Steenken, S.; O’Neill, P. J. Phys. Chem. 1977, 81, 505. O’Neill, P.; Schulte-
Frohlinde, D.; Steenken, S. J. Chem. Soc., Faraday Discuss. 1977, 63, 141.
O’Neill, P.; Steenken, S. Ber. Bunsen-Ges. Phys. Chem. 1977, 81, 550.
(24) (24) Sun, Q.; Tripathi, G. N. R.; Schuler, R. H. J. Phys. Chem.
1990, 94, 6273.
(26) Pauling, L. The Nature of the Chemical Bond; Cornell University
Press: Ithaca, New York, 1960.

Electron Transfer Component in Hydroxyl Radical Reactions
J. Am. Chem. Soc., Vol. 120, No. 17, 1998 4165
out to be 8.4 eV - IP(M) (EA of •OH taken as 1.8 eV).27 For
IP(M) > 8.4 eV (back reaction faster than forward reaction),
+•
Maq will be difficult to observe in basic solutions, even if
produced exclusively by ET. The free energy difference
•
+•
-
between Maq + OHaq and Maq + OHaq increases by 0.83
eV on OH- protonation (0.059 × pKa of water). Therefore,
+•
for IP(M) > 9.2 eV, Maq will be difficult to observe even in
neutral and acidic aqueous solutions. The numerical values used
here are very approximate. However, this simple thermody-
namical analysis explains why the H+-catalyzed loss of OH-
•
from the OH adducts of molecules with high IP(M), such as
benzene, is not observed.6 The effect of IP(M) in converting
the initial chemical step from •OH addition to electron transfer
can be readily visualized by raising or lowering the ionic
potential energy curve C in Figure 4.
While the above description of the reaction mechanism is
illustrative, it is too simplistic, and of little quantitative
significance in predicting the ET component. In particular, there
can be several isomers of MOH and the M+OH- configurations,
each contributing to the overall reaction process. Fortunately,
the molecular systems for which the ET components are
discussed here represent low (6%), middle (30%), and high
(85%) points of IP(M) dependence. This makes it is possible
to derive some extremely important conclusions, without
recourse to the details of the reaction model. For p-dimethoxy-
benzene, IP(M) ) 7.8 eV.28 Therefore, it can be safely assumed
that for IP(M) > 8 eV, which is the case for many benzene
derivatives, the •OH reaction will lead to adduct formation only.
On the other hand, for p-phenylenediamine, IP(M) ) 7.1 eV,28
which implies that for IP(M) < 7 eV, the reaction must be
considered an electron-transfer reaction.
•
Figure 4. Direct electron transfer (ET) and OH adduct mediated
electron transfer (AMET processes explained by qualitative potential
energy diagrams (see text)).
and CC (π) bonds is known to be small (∼20 kcal/mol or 0.87
eV).26 Therefore, the MOH potential well cannot be very deep
(<1 eV), irrespective of the nature of M. The difference
between the ionization potential (IP in eV) of M and the electron
affinity (EA) of OH (1.83 eV)27 gives the asymptotic energy
•
•
difference between M+• + OH- and M + OH (Figure 4B,C).
If the electrostatic binding energy of M+OH- becomes greater
than IP(M)-EA(•OH), the OH + M encounter can lead to the
•
ionic (M+OH-) configuration (Figure 4B). It should be noted
that the Coulombic binding energy of oppositely charged centers,
separated by <2.5 Å, is greater than 5 eV, and IP(M) ∼ 7 eV
is not uncommon in organic molecules.28 The Coulombic
potential energy curves, in general, are steeper than the covalent
potential energy curves, except for M and •OH separations close
to the MOH bonding distances.26 In nonpolar solvents, the ionic
configuration, M+OH-, will eventually terminate into the
covalent configuration, on dissipation of the excess energy as
thermal energy.29,30 However, if the M+OH- dipole is created
in a polar solvent like water (dielectric constant, ꢀ, ∼80), the
ionic components can solvate and separate rapidly (<100 fs)29,30
(Figure 4A,B). Experimentally, M+• will be observed to appear
The property of the OH adducts to undergo loss of OH- in
•
water (AMET) can also be understood in terms of IP(M).32 The
covalent configuration MOH is unlikely to undergo ionic
dissociation under the influence of solvent dielectric field.
However, the ionic configuration M+OH- can be thermally
accessed from the adduct (MOH) ground state (Figure 4). The
adduct radical will dissociate into solvated fragments M+ and
OH-, at a rate of kd exp(-∆E/RT), where kd is the solvent-
induced rate of dissociation in the M+OH- configuration, and
∆E is the energy gap between MOH and M+OH-. This
situation is very similar to the thermal dissociation of structurally
stable radicals in solution for which a two-state model has been
proposed recently.3 As is evident from the energy diagrams in
Figure 4, ∆E depends on IP(M). The ET is high for low IP(M),
which also means low ∆E, and, therefore, a faster rate of OH-
elimination in AMET. Thus, both steps of cation radical
formation are intimately related.
•
at the rate of OH encounter with M, i.e., at a diffusion-
controlled rate.
The hydration free energies (∆G°) of OH- and benzene-
size cations are estimated as -4 (91 kcal mol-1 31
and -2.6
)
eV (60 ( 8 kcal mol-1),27 respectively, which gives the free
energy difference between M+• + OH- and Maq+ + OHaq- as
6.6 eV (Figure 4A,C). If the hydration energies of neutral •OH
(-0.2 eV) and M (-0.1 eV)27 are ignored, the free energy
•
+•
-
difference between Maq + OHaq and Maq + OHaq comes
Conclusion
(27) Pearson, R. G. J. Am. Chem. Soc. 1986, 108, 6109.
(28) Rosenstock, H. M.; Draxl, K.; Steiner, B. W.; Herron, J. T. J. Phys.
Chem. Ref. Data 1977, 6, Suppl. 1.
A direct electron transfer path in the hydroxyl radical reaction
with aromatic molecules in water has been experimentally
demonstrated in selected model systems. This result has far-
reaching consequences for understanding the early chemical
events in one of the most widely studied reactions in chemistry.
The extent of electron transfer is directly related to the ionization
potential of the aromatic molecule. A potential energy model
for the reaction, consistent with the observations, has been
advanced.
(29) For solvation time of medium size molecular ions in nonpolar
solvents (55-90 ps), see: Lin, Y.; Jonah, C. J. Phys. Chem. 1992, 96,
10119. The solvation time decreases by more than two orders of magnitude
in acetonitrile (∼0.3 ps). The trends indicate that the combined rate of M+
and OH- hydration should be much faster. Professor Barbara (private
discussion) suggests a time scale of <50 fs for a major portion of the
hydration energy (∼4 eV) of OH-.
(30) For vibrational relaxation time (a few picoseconds), see: Harris,
A. L.; Brown, J. K.; Harris, C. B. Annu. ReV. Phys. Chem. 1988, 39, 341.
Benjamin, I.; Barbara, P. F.; Bradley, J. G.; Hynes, J. R. J. Phys. Chem.
1995, 99, 7557. Walhout, P. K.; Alfano, J. C.; Khalid, A. M. T.; Barbara,
P. F. J. Phys. Chem. 1995, 99, 7568. Dieter, P. J. In Radiationless Processes;
Fong, F. K., Ed.; Spring-Verlag: New York, 1976; p 169.
It is not uncommon to see a small but fast component in the
growth kinetics of cation radical in various systems, monitored
(31) Friedman, H. L.; Krishnan, C. V. In Water: A ComprehensiVe
Treatise; Franks, F., Ed.; Plenum Press: New York, 1973; Vol. 3.
(32) Holcman, J.; Sehested, K. Nukleonika 1979, 24, 887.

4166 J. Am. Chem. Soc., Vol. 120, No. 17, 1998
Tripathi
by optical absorption.33 The contribution is often ignored, or
interpreted without supporting evidence. Structure-sensitive
techniques, such as pulse radiolysis time-resolved Raman
spectroscopy, are essential for probing the nature and origin of
the individual components in the reaction. However, the
knowledge that the hydroxyl radical does involve an electron
transfer, and that this contribution depends on the ionization
potential of the molecule in question, can be extremely valuable
in interpreting the kinetic and spectral data obtained by the
commonly used transient absorption method.
Acknowledgment. The author wishes to thank Dr. John
Bentley and Professor Paul Barbara for illuminating discussions
on the reaction mechanism and ionic solvation.
(33) Qin, L.; Tripathi, G. N. R.; Schuler, R. H. Z. Naturforsch. Teil A
1985, 40, 1026 (a small ET component (4%) was noticed in the transient
•
absorption study of the OH oxidation of aniline (IP 7.7 eV) in water).
JA9800838