Accurate oxidation potentials of benzene and biphenyl derivatives via electron-transfer equilibria and transient kinetics

Article abstract of DOI:10.1021/jo9011267

(Graph Presented) Nanosecond transient absorption methods were used to determine accurate oxidation potentials (E_ox) in acetonitrile for benzene and a number of its alkyl-substituted derivatives. E_ox values were obtained from a combination of equilibrium electron-transfer measurements and electron-transfer kinetics of radical cations produced from pairs of benzene and biphenyl derivatives, with one member of the pair acting as a reference. Using a redox-ladder approach, thermodynamic oxidation potentials were determined for 21 benzene and biphenyl derivatives. Of particular interest, E_ox values of 2.48 ± 0.03 and 2.26 ± 0.02 V vs SCE were obtained for benzene and toluene, respectively. Because of a significant increase in solvent stabilization of the radical cations with decreasing alkyl substitution, the difference between ionization and oxidation potentials of benzene is ～0.5 eV larger than that of hexamethylbenzene. Oxidation potentials of the biphenyl derivatives show an excellent correlation with substituent σ⁺ values, which allows E_ox predictions for other biphenyl derivatives. Significant dimer radical cation formation was observed in several cases and equilibrium constants for dimerization were determined. Methodologies are described for determining accurate electrontransfer equilibrium constants even when dimer radical cations are formed. Additional equilibrium measurements in trifluoroacetic acid, methylene chloride, and ethyl acetate demonstrated that solvation differences can substantially alter and even reverse relative E_ox values.

Full text of DOI:10.1021/jo9011267

pubs.acs.org/joc
Accurate Oxidation Potentials of Benzene and Biphenyl Derivatives via
Electron-Transfer Equilibria and Transient Kinetics
Paul B. Merkel,* Pu Luo, Joseph P. Dinnocenzo,* and Samir Farid*
Department of Chemistry and the Center for Photoinduced Charge Transfer, University of Rochester,
Rochester, New York 14627-0216
pmerkel@rochester.rr.com; jpd@chem.rochester.edu; farid@chem.rochester.edu
Received May 27, 2009
Nanosecond transient absorption methods were used to determine accurate oxidation potentials
(
E ) in acetonitrile for benzene and a number of its alkyl-substituted derivatives. E values were
ox ox
obtained from a combination of equilibrium electron-transfer measurements and electron-transfer
kinetics of radical cations produced from pairs of benzene and biphenyl derivatives, with one member
of the pair acting as a reference. Using a redox-ladder approach, thermodynamic oxidation potentials
were determined for 21 benzene and biphenyl derivatives. Of particular interest, Eox values of 2.48 (
0
.03 and 2.26 ( 0.02 V vs SCE were obtained for benzene and toluene, respectively. Because of a
significant increase in solvent stabilization of the radical cations with decreasing alkyl substitution,
the difference between ionization and oxidation potentials of benzene is ∼0.5 eV larger than that of
hexamethylbenzene. Oxidation potentials of the biphenyl derivatives show an excellent correlation
þ
with substituent σ values, which allows E predictions for other biphenyl derivatives. Significant
ox
dimer radical cation formation was observed in several cases and equilibrium constants for
dimerization were determined. Methodologies are described for determining accurate electron-
transfer equilibrium constants even when dimer radical cations are formed. Additional equilibrium
measurements in trifluoroacetic acid, methylene chloride, and ethyl acetate demonstrated that
solvation differences can substantially alter and even reverse relative Eox values.
1
,2
for electron-transfer reactions. They are also useful for
1
. Introduction
3
estimating bond dissociation energies and acidity constants
Reliable redox potentials for aromatic hydrocarbons are
fundamental thermodynamic properties that have a wide
range of utility. For example, they are important for devel-
oping a detailed understanding of the energetics and kinetics
(2) Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem. Soc. 1990,
12, 4290.
(3) (a) Okamoto, A.; Snow, M. S.; Arnold, D. R. Tetrahedron 1986, 42,
175. (b) Popielarz, R.; Arnold, D. R J. Am. Chem. Soc. 1990, 112, 3068.
1
6
(
1) (a) Beletskaya, I. P.; Makhon’kov, D. I. Russ. Chem. Rev. (Engl.
(c) Maslak, P.; Vallombroso, T. M.; Chapman, W. H.; Narvaez, J. N. Angew.
Chem., Int. Ed. Engl. 1994, 33, 73. (d) Wayner, D. D.; Parker, V. D. Acc.
Chem. Res. 1993, 26, 287.
Transl.) 1981, 50, 534. (b) Chanon, M.; Tobe, M. L Angew. Chem., Int. Ed.
Engl. 1982, 21, 1. (c) Julliard, M.; Chanon, M. Chem. Rev. 1983, 83, 425.
DOI: 10.1021/jo9011267
r 2009 American Chemical Society
Published on Web 07/09/2009
J. Org. Chem. 2009, 74, 5163–5173 5163

Merkel et al.
JOCFeatured Article
of radical ions. In addition, they provide critical bench-
mark values for testing recent quantum chemical approaches
3
d
where Keq is the electron-transfer equilibrium constant and
k1 and k are the forward and reverse electron-transfer rate
-1
4
for predicting redox potentials in solution.
constants, respectively. Under conditions where the decay of
•
þ •þ
and BP are slow relative to k [BP] and k [B], an
1 -1
For compounds that form reactive radical cations upon
one-electron oxidation, the most common cyclic voltam-
metric (CV) methods for determining oxidation potentials
often fail to provide reversible voltammograms. In part for
this reason, a wide range of oxidation potentials can fre-
quently be found in the literature. For example, literature
oxidation potentials for the fundamental hydrocarbon ben-
B
equilibrium is effectively established (see the Supporting
Information) involving the radical cations and their unox-
idized forms. The difference in the oxidation potentials of BP
and B, E (BP) - E (B), can be determined from eq 2. This
analysis makes the conventional assumption that concentra-
tions can be used in place of activity coefficients. Thus, the
ΔEox values derived from the redox equilibrium measure-
ments are strictly differences in formal potentials, rather
than standard potentials. We note, however, that the neglect
of activity coefficients should be inconsequential in the
examples studied herein because the ionic strength of the
medium is particularly low (∼1 mM) and because the activity
coefficients of the species on both sides of the equilibrium in
eq 1a are expected to be quite similar.
ox
ox
5
zene vary from 2.26 to 2.82 V vs SCE in acetonitrile -a range
of nearly 13 kcal/mol! The range of literature oxidation
potentials for toluene (1.98-2.40 V) is only marginally
5
a-c,5k,5m,6
better.
and fast scan potentiostats allows redox potentials to be
measured for radical ions with lifetimes in the nanosecond
In principle, the use of microelectrodes
7
8
range. In practice, slow heterogeneous electron transfer at
the electrode for molecules like benzene and toluene do not
permit the full advantage of fast scan CV techniques to be
realized.
Herein, we utilize nanosecond transient absorption spec-
9
troscopy to monitor the electron-transfer equilibrium
established between the radical cation of a reference com-
pound of known oxidation potential and an electron donor
of unknown oxidation potential to determine thermodyna-
mically meaningful oxidation potentials for a range of
biphenyl (BP) and benzene derivatives (B), including ben-
zene and toluene. The methodology is illustrated for the case
in which a BP derivative serves as the reference compound
and a B derivative is the compound whose oxidation poten-
tial is to be measured by using the redox equilibrium in eq 1,
k1
BP þ B þ h BP þ B
•
•þ
ð1aÞ
ð1bÞ
k-1
,10
Keq ¼ k1=k-1
EoxðBPÞ -EoxðBÞ ¼ -RT ln Keq ¼ -0:0255 ln Keq ð2Þ
Although the redox equilibrium method has the virtue that
it can provide oxidation potentials with extraordinary accu-
racy and precision, it also has some limitations. First, as
mentioned above, the method requires that the decay of
intermediate radical cations be slow relative to the electron
exchange in eq 1a. Second, reliable equilibrium constants can
only be determined for donor and reference compounds
whose oxidation potentials are relatively similar. For exam-
(
4) For recent attempts to calculate oxidation potentials, see: (a) Winget,
P.; Weber, E. J.; Cramer, C. J.; Truhlar, D. G. Phys. Chem. Chem. Phys. 2000,
, 1231. (b) Baik, M.-H.; Friesner, R. A. J. Phys. Chem. A 2002, 106, 7407. (c)
Han, Y.-K.; Jung, J.; Cho, J.-J.; Kim, H.-J. Chem. Phys. Lett. 2003, 368, 601.
d) Fu, Y.; Liu, L.; Yu, H.-Z.; Wang, Y.-M.; Guo, Q.-X. J. Am. Chem. Soc.
2
ple, a ΔEox of only 200 mV corresponds to K >2000, which
eq
is generally too difficult to accurately determine. As we will
show, when these conditions are not met the electron-trans-
fer equilibrium constant can instead be reliably determined
by kinetic approaches that directly measure the rate con-
(
2
2
005, 127, 7227. (e) Yu, A.; Liu, Y.; Li, Z.; Cheng, J.-P. J. Phys. Chem. A
007, 111, 9978. (f) Singh, N. K.; Shaik, M. S.; O’Malley, P. J.; Popelier, P. L.
A. Org. Biomol. Chem. 2007, 5, 1739. (g) Riahi, S.; Norouzi, P.; Bayandori
Moghaddam, A.; Ganjali, M. R.; Karimipour, G. R.; Sharghi, H. Chem.
Phys. 2007, 337, 33. (h) Alizadeh, K.; Seyyedi, S.; Shamsipur, M. Pol. J.
Chem. 2008, 82, 1449.
1
1
stants for electron exchange (k and k ).
1
-1
(
5) (a) Lund, H. Acta Chem. Scand. 1957, 11, 1323. (b) Loveland, J. W.;
2
. Results
We begin by describing the spectral characterization of
Dimeler, G. R. Anal. Chem. 1961, 33, 1196. (c) Pysh, E. S.; Yang, N. C. J. Am.
Chem. Soc. 1963, 85, 2124. (d) Neikam, W. C.; Desmond, M. M. J. Am.
Chem. Soc. 1964, 86, 4811. (e) Gough, T. A.; Peover, M. E. Proceedings of the
3
rd International Polarography Congress, Southhampton, 1965; Macmillan:
benzene and biphenyl radical cations in terms of their
extinction coefficients at their absorption maxima (λmax).
The spectral data are then used to determine ΔEox for
electron-transfer equilibria between different electron do-
nors and their radical cations. Finally, for cases where the
equilibrium method is difficult to apply, transient kinetics
are used to determine the individual rate constants for
electron-transfer equilibration and, thereby, Keq and ΔEox
values.
London, 1966; p 1017. (f) Hansen, R. L.; Toren, P. E.; Young, R. H. J. Phys.
Chem. 1966, 70, 1653. (g) Hoytink, G. J. Disc. Faraday Soc. 1968, 45, 14. (h)
Osa, T.; Yildiz, A.; Kuwana, T. J. Am. Chem. Soc. 1969, 91, 3994. (i) Parker, V. D.
J. Am. Chem. Soc. 1976, 98, 98. (j) Tanimoto, I.; Kushioka, K.; Kitagawa, T.;
Maruyama, K. Bull. Chem. Soc. Jpn. 1979, 52, 3586. (k) Howell, J. O.;
Goncalves, J. M.; Amatore, C.; Klasinc, L.; Wightman, R. M.; Kochi, J. K. J.
Am. Chem. Soc. 1984, 106, 3968. (l) Butin, K. P.; Moiseeva, A. A.; Magdesieva,
T. V.; Sergeeva, E. V.; Rozenberg, V. I.; Kharitonov, V. G. Russ. Chem. Bull. 1994,
4
3, 783. (m) Fukuzumi, S.; Ohkubo, K.; Suenobu, T.; Kato, K.; Fujitsuka, M.; Ito,
O. J. Am. Chem. Soc. 2001, 123, 8459.
6) Neikam, W. C.; Dimeler, G. R.; Desmond, M. M. J. Electrochem.
Soc. 1964, 111, 1190.
7) (a) Wightman, R. M.; Wipf, D. O. Electroanal. Chem. 1989, 15, 267.
b) Andrieux, C. P.; Hapiot, P.; Saveant, J.-M. Chem. Rev. 1990, 90, 723. (c)
(
2
.1. Radical Cation Spectra and Extinction Coefficients.
(
•
þ
(
Radical cation extinction coefficients (ε ) were obtained for
most electron donors by comparing transient absorption
spectra obtained by pulsed laser (7 ns, 343 nm) irradiation
of air-saturated acetonitrile solutions containing 1 mM
Heinze, J. Angew. Chem., Int. Ed. Engl. 1993, 32, 1268. (d) Amatore, C.;
Bouret, Y.; Maisonhaute, E.; Abruna, H. D.; Goldsmith, J. I. C. R. Chim.
2
003, 6, 99.
(8) Amatore, C.; Maisonhaute, E.; Simmoneau, G. J. Electroanal. Chem.
000, 486, 141.
2
þ
N-methylquinolinium hexafluorophosphate (NMQ ), to-
luene (0.5 M), and either a BP or a B derivative (∼10 mM).
(
9) For reviews, see: (a) Steenken, S. Landolt-B o€ rnstein 1985, 13e, 147. (b)
Wardman, P. J. Phys. Chem. Ref. Data 1989, 18, 1637. (c) Stanbury, D. M.
In General Aspects of the Chemistry of Radicals; Alfassi, Z. B., Ed.; Wiley: New
York, 1999; Chapter 11.
(
10) Guirado, G.; Fleming, C. N.; Lingenfelter, T. G.; Williams, M. L.;
(11) Gould, I. R.; Wosinska, Z. M.; Farid, S. Photochem. Photobiol.
2006, 82, 104.
Zuilhof, H.; Dinnocenzo, J. P. J. Am. Chem. Soc. 2004, 126, 14086.
5
164 J. Org. Chem. Vol. 74, No. 15, 2009

Merkel et al.
JOCFeatured Article
Although this system for generating radical cation spectra
1
0
has been described in detail elsewhere, we highlight its
essential features here for clarity. Pulsed laser excitation of
þ
1
þ
NMQ produces its singlet-excited state ( NMQ *), which
is rapidly intercepted by toluene to efficiently produce
•
þ
•
toluene and the N-methylquinolinyl radical (NMQ ) by
•
photoinduced electron transfer. NMQ is rapidly scavenged
by dioxygen in <200 ns leading to O2 , which does not have
interfering absorptions in the visible region where the radical
•-
•
þ
cations strongly absorb. Interception of toluene by the BP
or B derivatives produces their radical ions with a concen-
tration essentially independent of the donor. At the moder-
ate laser powers used in these experiments there was typically
minimal second-order decay of the radical ions at 200 ns,
such that the transient spectral maxima provided a direct
FIGURE 1. Radical cation absorption spectra of biphenyl (BP6),
methyl 4-biphenylcarboxylate (BP3), p-xylene (B10), and chloro-
benzene (B2) generated by laser pulse excitation (7 ns, 343 nm) of
•
þ
measure of the relative extinction coefficients of the BP and
•
þ
B
mined by using biphenyl radical cation (BP6 ) as reference
derivatives. Absolute extinction coefficients were deter-
•
þ
þ
0.01 M solutions of donor (except 0.1 M for B2) with 1 mM NMQ
in air-saturated acetonitrile. Solutions of BP3, BP6, and B10 also
contained 0.5 M toluene codonor.
•
þ
-1
-1
2
(
ε =14500 M cm at 668 nm in acetonitrile).
were generated by pulsed laser excitation of air-saturated
þ
acetonitrile solutions of NMQ containing 0.05-2.0 M
•
toluene. Following reaction of NMQ with O , the radical
2
•
þ
•þ
cations B and B2 are the only species that absorb in the
visible region of the transient absorption spectra. The in-
creases in optical density that occur at wavelengths >600 nm
with higher concentrations of toluene reflect the formation
The method for determining radical cation extinction
coefficients described above is suitable only for electron
donors whose oxidation potentials are significantly lower
•
þ 12
•þ
of B2
. At these wavelengths essentially only B absorbs.
2
•
þ
•þ
than that of toluene. For donors with E values close to or
The absorption spectra of B and B2 are similar near 430
nm, and the extinction coefficients of the two species are
nearly the same in this region. By fitting the changes in the
ox
greater than that of toluene, ε• values were obtained with-
out toluene by comparing transient absorption spectra of
solutions with sufficient concentration of B or BP to effi-
þ
•
þ
•þ
absorption at 750 nm to the expression [B ]/[B ]=K [B],
2
D
1
þ
•þ
ciently quench NMQ *. Repetition of these experiments in
the presence of a low concentration of biphenyl (∼1 mM)
which assumes negligible B absorption at this wavelength,
a KD value of 5 was determined for toluene in acetonitrile.
KD values for five other benzene derivatives that formed
dimer radical cations were determined in a similar manner
(Table 2). Negligible radical cation dimerization was ob-
served for the other benzene derivatives or for any of the
biphenyl derivatives used in this study.
•
þ
•þ
•þ
þ
permitted conversion of B or BP to biphenyl prior to
•
significant radical cation decay. The absolute ε values of
•
þ
•þ
•þ
B
and BP could then be determined from the known ε
•
þ
for biphenyl , as described above. Typical radical cation
absorption spectra are shown in Figure 1. A summary of the
extinction coefficients and spectral maxima for all of the
radical cations studied are provided in the second and third
columns of Table 1.
When toluene was used as a codonor at high concentra-
tions, spectral evidence was found for complex formation
between toluene and the radical cations of three benzene
•
þ
2
.2. Dimer Radical Cation Formation. During the course of
derivatives (toluene/B ) according to eq 4. Estimates for the
mixed complex formation constants (K ) were determined in
our experiments, we noted that the radical cation equilibria
for some of the benzene derivatives were complicated by the
formation of dimer radical cations (B2 ) when the neutrals
were used in high (g0.05 M) concentrations. At these con-
centrations, the equilibration in eq 3 must be taken into
account for the accurate determination of Eox values.
C
an analogous manner to the KD values and are listed in
Table 2.
1
2
•þ
KC
toluene þ B þ h toluene=B
•
•þ
ð4Þ
Fortunately, the magnitudes of the KD and K values in
C
Table 2 are small enough that equilibria (3) and (4) do not
significantly interfere with the determination of equilibrium
oxidation potentials for the vast majority of substrates
studied here. As will be shown, only in the case of electron-
transfer equilibria involving benzene did dimer radical
cation formation need to be explicitly considered for the
determination of accurate oxidation potentials.
2.3. Eox Values via Radical Cation Equilibria. Eox values
were obtained by measuring electron-transfer equilibrium
constants of reaction 1a for each benzene derivative (B1-
B13) vs one or more biphenyl derivatives (BP1-BP8).
A redox ladder was created with these compounds using
KD
•
þ
•þ
B þ B h B2
ð3Þ
It was possible to obtain the equilibrium constant (K ) for
D
•
þ
B2 formation from measurements of transient absorption
spectra of the donor radical cations as a function of donor
concentration. For example, shown in Figure 2 are the
spectral changes for solutions containing toluene radical
cation as a function of toluene concentration. These spectra
(12) Bockman, T. M.; Karpinski, Z. J.; Sankararaman, S.; Kochi, J. K. J.
Am. Chem. Soc. 1992, 114, 1970.
J. Org. Chem. Vol. 74, No. 15, 2009 5165

Merkel et al.
JOCFeatured Article
TABLE 1. Radical Cation Spectral Data and Oxidation Potentials of Benzene and Biphenyl Derivatives in Acetonitrile
ε• (M cm
þ
-1
-1
)
λ
max (nm)
Eox (V vs SCE)
donor
a
benzene derivatives
benzene
B1
B2
B3
B4
B5
B6
B7
B8
2200
5100
2900
2900
3200
2800
4700
3200
3800
3900
4000
3300
3750
4900
432
473
448
429
448
431
470
475
498
437
447
460
464
454
2.48 ( 0.03
2.46 ( 0.03
2.28 ( 0.02
2.26 ( 0.02
2.10 ( 0.01
2.09 ( 0.01
2.06 ( 0.01
2.05 ( 0.01
2.04 ( 0.01
2.01 ( 0.01
1.905 ( 0.01
1.825 ( 0.01
1.82 ( 0.01
chlorobenzene
tert-butylbenzene
toluene
m-xylene
o-xylene
1,4-di-tert-butylbenzene
mesitylene
1,3,5-tri-tert-butylbenzene
p-xylene
1,2,4-trimethylbenzene
1,2,3,4-tetramethylbenzene
1,2,3,5-tetramethylbenzene
durene
B9
B10
B11
B12
B13
B14
b
1.75
biphenyl derivatives
0
BP1
BP2
BP3
BP4
BP5
BP6
BP7
BP8
a
4-CO
4-CO
4-CO
3-CO
4-Br, 4 -Br
biphenyl
4-OCOMe
4-Me
2
2
2
2
Bu, 4 -CF
3
13800, 33500
14500, 42200
11500, 28000
12700, 25300
11000, 37000
14500, 28000
15200, 31000
14400, 32300
692, 394
708, 406
682, 395
666, 385
704, 423
668, 383
672, 399
670, 393
2.35 ( 0.02
2.28 ( 0.02
2.10 ( 0.01
2.06 ( 0.01
1.98 ( 0.01
0
Bu, 4 -CO
Me
2
Bu
H
0
c
d
1.95
1.81 ( 0.01
1.795 ( 0.01
For comparison, reversible oxidation potentials in acetonitrile have been reported for other methyl-substituted benzenes: pentamethylbenzene (1.69)
13 b
c
and hexamethylbenzene (1.58 V vs SCE), respectively. From ref 13. From ref 2. From ref 10.
d
TABLE 2. Equilibrium Constants for Radical Cation Dimerization, K
D
,
and for Complexation with Toluene, K , in Acetonitrile at Ambient Tem-
C
perature
-1 a
)
-1
a
benzene derivative
K
D
(M
K
C
(M toluene)
benzene
tert-butylbenzene
toluene
12
2
5
m-xylene
o-xylene
p-xylene
5
4
2
2
1.5
0.5
a
Estimated error, ∼20%.
indistinguishable from the initial concentrations. The
low concentrations of radical cations render potential bimo-
lecular coupling reactions insignificant relative to their first-
order decay, which proceeds with a rate constant of ∼2 ꢀ
FIGURE 2. Transient radical cation absorption spectra (molar
absorptivity units) measured 300 ns after 343 nm laser excitation
5
-1
10 s (vide infra). Although absolute radical cation extinc-
tion coefficients are provided in Table 1, the determinations
of Keq depends only on relative extinction coefficients, which
were determined with higher precision (5-10%) in the
equilibration experiments. It is worth noting that even an
uncertainty of 10% in the relative extinction coefficients
corresponds to an error in ΔEox of only 0.002 V. In general,
transient absorption spectra were obtained at successive
delay times to ascertain that the ratio of absorptions of
þ
of air-saturated acetonitrile solutions containing 1 mM NMQ and
0
.05, 0.50, and 2.0 M toluene.
biphenyl (BP6) and durene (B14) as redox anchors. The
oxidation potential of BP6 was previously determined from
1
0
redox equilibration experiments and, in turn, rests on
the carefully measured value for E (B14) determined by
ox
Amatore and Lefrou using ultrafast scan cyclic voltammetry
1
3
•þ
•þ
and ultramicroelectrodes. Appropriate concentrations of
donors were used so that equilibration among donors and
their radical cations was much faster than radical ion decay,
which typically required several microseconds, thus allowing
an equilibrium to be effectively established in solutions
containing the two electron-donors.
B
and BP remained constant and that equilibrium was
achieved.
Typical equilibrium spectra are illustrated in Figure 3 for
•
þ
biphenyl (BP6) and p-xylene (B10). Spectra of BP6 and
•
þ
B10 are given by traces a and b, respectively. Experimen-
tally measured equilibrium mixtures of the two radical cation
species at two different donor concentrations are shown in
spectra c and d. These spectra can be fit quite well by
appropriate combinations of the pure radical cation spectra
a and b, as shown by plots e and f. The ratios of the radical
cations comprising plots e and f together with the asso-
ciated concentrations of BP6 and B10 provide an experi-
The electron-transfer equilibrium constants (K ) were
eq
•
þ
•þ
calculated from [BP][B ]/[B][BP ], and ΔE was calcu-
ox
lated from eq 2. At low pulse energies, the concentrations of
-5
the radical cations (e10 M) are very small compared
to those of the neutral species [B] and [BP], which are
(
13) Amatore, C.; Lefrou, C. J. Electroanal. Chem. 1992, 325, 239.
mental equilibrium constant (K ) of 11-12. This yields
eq
5
166 J. Org. Chem. Vol. 74, No. 15, 2009

Merkel et al.
JOCFeatured Article
FIGURE 3. Transient radical cation absorption spectra measured
FIGURE 4. Transient absorption of the radical cation of methyl
5
trile solutions containing 1 mM NMQ and 0.5 M toluene with
00 ns after 343 nm laser excitation of air-saturated acetoni-
4
time, following 343 nm laser excitation of air-saturated acetonitrile
-biphenylcarboxylate, BP3, monitored at 680 nm, as a function of
þ
(
a) 0.01 M biphenyl (BP6), (b) 0.05 M p-xylene (B10), (c) 0.002 M
þ
solutions containing 1 mM NMQ and 40 mM BP3 with (a) no
added donor, (b) 2.6 mM 1,4-di-tert-butylbenzene (B7), and
BP6 þ 0.05 M B10, and (d) 0.004 M BP6 þ 0.05 M B10. Curve (e) is a
fit of (c) with 33% of (a) plus 67% of (b), and curve (f) is a fit of
(
(
c) 2 mM p-xylene (B10). Trifluoroacetic anhydride (0.04 M) was
d) with 47% of (a) plus 53% of (b), which yield Keq=11-12 for
added to scavenge traces of water in acetonitrile and increase the
radical cation lifetimes for these experiments.
eq 1 and ΔEox (BP6 - B10) = -0.06 V.
where
ΔEox (BP6 - B10) of -0.06 V which, coupled with E (BP6)
ox
a
b
a
b
=
1.95, gives E (B10) = 2.01 vs SCE. Analogous experi-
ðε • þ
Þ
ðε • þ
Þ
ðε • þ
Þ
Þ
ox
BP
BP
B
x ¼
y ¼
z ¼
ments were performed with different concentrations of BP6
and B10 without toluene as a codonor. In all cases, the Eox
value for B10 was 2.01 V vs SCE within experimental error
b
b
ðεB• þ
Þ
ðεB• þ
Þ
ðε • þ
B
In some cases, it was also possible to measure Keq and
Eox values by comparing transient absorption spectra of two
different BP derivatives or two B derivatives, providing that
their radical cation spectra were sufficiently different. One
example is the pair consisting of BP5 and BP6, for which the
absorption maximum of the strong radical cation bands in
the near UV and blue are 423 and 383 nm, respectively. From
the equilibration of BP5 and BP6 and their radical cations a
(
(0.01 V). These latter experiments demonstrate that the
•
þ
toluene/p-xylene complex described above does not detec-
tably affect the determination of the electron-transfer equi-
librium constant.
Additional examples of the determination of Eox values by
the fitting of full radical cation spectra of pairs of donors
under equilibrium conditions are provided in the Supporting
Information. While this method has the advantage that the
entire radical cation spectra are used to determine Eox,
equally accurate values could be obtained from the ratio
ΔE (BP5 - BP6) of 0.03 V was obtained.
ox
In several cases, both the full spectrum fitting and the two-
wavelength methods were used to determine Eox values.
When this was done, the agreement between the methods
was excellent. Oxidation potentials vs SCE in acetonitrile at
295 K obtained by one or both of these methods are provided
in the last column of Table 1. In many cases, more than one
reference donor was used to obtain Eox of the target donor
(see the Supporting Information for a detailed redox ladder).
As expected, the error estimates in Eox naturally increase the
further removed they are from the redox anchors, BP6 and
B14, due to propagation of errors.
(
rium at two wavelengths, a and b, according to eqs 5a and
5
R) of optical densities of pairs of radical cations in equilib-
14,15
b
Ideally, the wavelengths are chosen to yield minimal
values (preferably zero) for the extinction coefficient ratios y
and z. As illustrated by the spectra in Figures 1 and 3, donor
pairs consisting of a biphenyl derivative (BP) and a benzene
derivative (B) are particularly well suited to such determina-
tions due to the large differences in their radical cation
absorption maxima. Because the radical cations of all BP
derivatives have two prominent absorption peaks (Figure 1),
Keq and Eox values were determined using each of the BP
absorption maxima as wavelength a in the above calculations
As also shown in Table 1, tert-butylbenzene and 1,4-
di-tert-butylbenzene were found to have slightly higher
oxidation potentials in acetonitrile than their methyl-
substituted analogues toluene and p-xylene. Interestingly,
literature oxidation potentials determined by cyclic voltam-
to provide a check on calculated values. In general, the E
•
ox
þ
values obtained using the two different BP absorption
maxima agreed to within 0.005 V.
metry (CV) at slow scan rate in CH CN and at high scan rate
3
a
b
(200-2000 V/s) in CF CO H show the opposite trend in
R ¼ ðOD =OD Þ
ð5aÞ
3
2
5
k
Eox values. That the tert-butyl-substituted benzene deriva-
•
þ
tives have higher oxidation potentials in CH CN is addition-
3
½
B ꢁ
x -yR
R -z
¼
ð5bÞ
•
þ
ally confirmed for 1,4-di-tert-butylbenzene (B7) vs p-xylene
(B10) by the data shown in Figure 4. In this experiment,
½
BP ꢁ
•
þ
BP3 is generated and allowed to undergo equilibration
with added B7 or B10 by monitoring the absorption of
(
(
14) Meisel, D.; Neta, P. J. Am. Chem. Soc. 1975, 97, 5198.
15) For an analogous approach used for equilibration between two
excited triplet states, see: Merkel, P. B.; Dinnocenzo, J. P. J. Photochem. and
Photobiol. A, Chem. 2008, 193, 110.
•þ
BP3 at 680 nm, where the benzene radical cations do not
absorb. Shown in trace a is the absorbance of BP3 with no
•þ
J. Org. Chem. Vol. 74, No. 15, 2009 5167

Merkel et al.
JOCFeatured Article
added donor. When 2.6 mM of B7 is added (trace b), the
SCHEME 1
•þ
initial absorbance of BP3 rapidly decreases, with conse-
•
þ
quent formation of B7 , until equilibrium is reached after
0.2 μs. Importantly, addition of a lower concentration
2.0 mM) of p-xylene (B10) leads to a larger decrease in
∼
(
•
þ
the absorption of BP3 , indicating a greater formation
þ
•
of B10 . These data clearly demonstrate that E (B7) >
ox
E (B10). A quantitative analysis gives equilibrium con-
ox
stants of 5 ( 0.7 for B7/BP3 and of 39 for B10/BP3, which
correspond to an oxidation potential for B7 that is 0.05 V
higher than B10 in acetonitrile.
To determine if differential solvation could substantially
alter relative oxidation potentials, analogous experiments
were performed with BP3, B7, and B10 in trifluoroacetic acid
containing 7 vol % trifluoroacetic anhydride-the same
5
k
solvent system used in previous fast scan CV experiments.
Interestingly, we found that trifluoracetic acid indeed alters
the relative oxidation potentials of B7 and B10, with B7
being easier to oxidize than B10 by 0.02 V in CF CO H,
3
2
0
d
other trace impurities. When λ ≈ k ≈ k , the difference in
which agrees well with that obtained by fast scan CV (0.03 V).
Additional comparative equilibration experiments with B7,
B10, and BP3 were carried out in methylene chloride and
in ethyl acetate. It was found that the Eox of B7 is 0.01 V
higher than that of B10 in methylene chloride and 0.06 V
higher in ethyl acetate, which underscores the need to con-
sider varying solvation effects when comparing oxidation
potential data.
2
d
time constants, λ - λ , is equal to k [BP] þ k [B], to a very
1
2
1
-1
good approximation (eq 6f). In practice, the individual
values of k and k could be obtained via eq 6f from
1
-1
experiments where [B] and [BP] were varied. Because λ was
more than an order of magnitude greater than λ small
variations in λ had a negligible effect on the determination
1
2
2
of λ1.
2
.4. Eox Values via Electron-Transfer Kinetics. The redox
0
λ1 -λ2 ≈ k1½BPꢁ þ k-1½Bꢁ when kd ≈ kd
ð6fÞ
equilibrium method described above is not suitable for the
determination of accurate oxidation potentials when the
oxidation potential difference between two compounds is
large and/or the decay of their radical cations is competitive
with the electron exchange process. In these cases, the
electron-transfer equilibrium constants can instead be con-
veniently determined by transient kinetics. As we will
show, kinetic methods can also be successfully applied to
reactions where dimer radical cation formation (eq 3) affects
the electron-transfer equilibrium.
In addition to pseudo-first-order decay by reaction with
impurities, the decay of radical cations can also occur by
return electron transfer (via O2 in this case)-a second-
order process. This latter contribution was deliberately
minimized by use of low laser power. The small, unavoidable
contribution from this reaction that remained was approxi-
mated as an additional first-order decay. Occasionally, the
decay does not completely go to a zero baseline, which was
accommodated in the fitting procedure by the constant c in
•
-
1
1
1
1
We first describe application of kinetic methods for de-
termining Keq using the simplified kinetic analysis shown in
Scheme 1, where the decay rate constants for the radical
3
eq 6. In all cases, the contribution from c , if any, was <1%
3
of the maximum optical density.
The kinetic procedure described above utilizes the reaction
0
cations k and k are included in the analysis and dimer
d
d
radical cation formation is unimportant. The forward and
reverse electron-transfer rate constants (k and k ) that
time constants λ and λ to determine k and k and,
1 2 1 -1
thereby, the equilibrium constant for electron transfer
1
-1
define the electron-transfer equilibrium can be obtained
from fitting the changes in optical density (ΔOD) as a
function of time (t) according to the well-known kinetic
(K ). We additionally determined K by use of the pre-
eq
exponential factor c in eq 6a, which provides a simple
1
eq
relationship between the relative concentrations of the two
radical cations. Consider the case of a B/BP system in which
B has a higher Eox than that of BP and [B] . [BP]. Under
these conditions, the one-electron oxidation of the substrates
is initiated almost entirely by interception of the excited
sensitizer by B. If the kinetics are monitored at a wavelength
where εBP•þ > εB•þ, a growth in the optical density is
observed followed by decay of the radical cations. The
relationship between the optical density and equilibrium
1
6
expressions in eq 6.
0
When k ≈ k it is easy to show from eq 6c that λ ≈ k ≈
d
d
2
d
0
k and is, therefore, independent of the concentration ratio
of the reactants, B and BP. For the systems described here,
d
we found λ to be nearly constant and first order (1.8 ( 0.3 ꢀ
2
5
0 s ) at low laser power, suggesting that decay of the
-1
1
radical cations was primarily through reaction with water or
constant is then given by eq 7, where OD is the optical
0
(
16) (a) Birks, J. B. Photophysics of Aromatic Molecules; Wiley-Inter-
science: New York, 1970. (b) Birks, J. B.; Dyson, D. J.; Munro, I. H. Proc. R. Soc.
London 1963, 275, 575.
density at t=0 in the absence of BP, OD corresponds to the
¥
maximum attainable optical density when all initially pro-
(
17) As defined in the text, the ratio of the optical densities (OD
equal to ratio of the extinction coefficients (εBP•þ/ε •þ) at the analyzing
wavelength. Thus, from this ratio and the experimentally determined OD
the nominator in eq 7, (OD - OD ), can be calculated. Alternatively, and
more accurately, (OD /OD ) is obtained from the reciprocal of the intercept
of a plot of 1/(OD - OD ) vs [B]/[BP]. Cf. ref 14.
¥ 0
/OD ) is
•
þ •þ
duced B is intercepted by BP to form BP , and OD is the
•
B
þ
•þ
0
optical density from the equilibrium mixture of B and BP
in the absence of decay. This OD and OD are given by
¥
0
¥
c1, eq 6a. The reciprocal of the slope of a plot of the optical
¥
0
0
5
168 J. Org. Chem. Vol. 74, No. 15, 2009

Merkel et al.
JOCFeatured Article
FIGURE 6. Plot of the difference between the time constants
(
in acetonitrile vs [BP] at two benzene concentrations.
λ
1
-λ
2
), which corresponds to (k
1
[BP] þ k-2[B]), for benzene/BP2
SCHEME 2
•
þ
corresponding reactions involving B2 (k and k ) can be
2
-2
obtained from experiments in which [B] is kept constant and
BP] varied and vice versa. This is illustrated in Figure 6 for
FIGURE 5. Determination of the electron-transfer equilibrium
constant involving toluene and methyl 4-biphenylcarboxylate
[
(
0
BP3). The toluene concentration was 0.05 M (unfilled circles) or
.1 M (filled circles). Top: plot of (λ - λ ) (see text) vs the sum of the
the equilibrium involving the radical cations of benzene and
the biphenyldicarboxylate derivative BP2. As described
above, (λ - λ ) corresponds to k [BP] þ k [B], here
1
2
electron-transfer exchange rates (Scheme 1) using the best-fit rate
constants given in the figure. Bottom: plot of the optical density
function of eq 7 vs the reactants ratio.
1
2
1
-1
measured at two different concentrations of [B]. Because
in each data set [B] was kept constant, the slope of the
fitting line is equal to the effective rate constant (weighted
•
þ
density function of eq 7 vs [B]/[BP] is equal to the equilibrium
1
constant.
average of k and k ) to form BP . That the slopes of the
1
2
7
fitting lines at [B]=0.05 and 0.4 M are essentially the same
9
-1 -1
•þ
•þ
), while the ratio of B /B changes from
2
OD¥ -OD0
OD -OD0
1 ½Bꢁ
(7.5 ꢀ 10 M
s
¼
1þ
ð7Þ
62:38 to 17:83, respectively, shows that k must be very
1
similar to k . This result is somewhat surprising because
the equilibrium constant for dimer formation (K =12 M ,
D
Keq ½BPꢁ
The results from determining the equilibrium constants
2
-
1
•
þ
using both the time constants (λ1 and λ ) and the pre-
exponential (c ) methods are illustrated in Figure 5 for the
case of toluene (B4) and the biphenyl carboxylate derivative
BP3. As shown in Figure 5, the equilibrium constants
determined from both kinetic methods are in excellent
agreement (K =330 and 350, respectively) and correspond
eq
to ΔEox values that differ by <0.002 V. That good fits are
obtained at two different toluene concentrations (0.05 and
0
radical cation does not detectably affect the determination of
K . As described below, this contrasts with experiments to
eq
determine the oxidation potential of benzene, where dimer
radical cation formation must be taken into account.
The effects of dimer radical cation formation during elec-
tron-transfer equilibration can be accounted for by Scheme 2,
where the decay rate constants of the radical cations (k and
d
k ) discussed above are not shown for clarity.
d
Table 2) indicates that B
•
is 60 meV lower in energy than
and k steps, both being
2
2
þ
B
þ B. Apparently the k
1
1
2
exothermic, are not significantly affected by this difference
in driving force.
Next, the concentration of [BP] was held constant and that
of [B] was varied, with the data being analyzed using the
reaction time constants and the optical densities vs [B] as
described above. If there were no reverse electron transfer
involving two benzene molecules, i.e., negligible k-2, both
plots would be linear, which is clearly not the case (Figure 7).
The upward curvature of both plots indicates that the appar-
ent equilibrium constant decreases with increasing concentra-
.1 M) demonstrates that the formation of toluene dimer
tion of [B] due to an increasing contribution from k 2
.
-
Both sets of data in Figure 7 were fitted globally according to
9
-1 -1
Scheme 2 using k =k =7.5 ꢀ 10 M
s , as established from
the data shown in Figure 6. Values for the two remaining
1
2
0
parameters, k and k , were varied to obtain the best fits. The
-1 -2
6
-1 -1
Differences in the electron-transfer rate constants be-
tween the monomeric species (k and k ) and those of the
curves shown in Figure 7 are based on k =3.4 ꢀ 10 M
s
-1
7
-2 -1
and k-2=1 ꢀ 10 M
s , providing Keq=k /k-1=2200 for
1
1
-1
J. Org. Chem. Vol. 74, No. 15, 2009 5169

Merkel et al.
JOCFeatured Article
FIGURE 7. Plots of a function of the electron-transfer rates, eq 6f
(
top), and of the optical densities, eq 7 (bottom), for benzene/BP2 in
acetonitrile as a function of benzene concentration at a constant
BP2] of 0.15 mM.
[
FIGURE 8. Kinetic data for equilibration involving the radical
cations of benzene and BP1 in acetonitrile. Top: plot of a function
of electron-transfer rates, eq 6f, vs [BP1] at two different concentra-
tions of benzene. Bottom: plot of a function of optical densities,
eq 7, vs [benzene]/[BP1].
the benzene/BP2 equilibrium constant. The corresponding
value of ΔEox is 0.20 V, which leads to an oxidation potential
for benzene of 2.48 V vs SCE.
We note that the larger value for k 2 relative to k reflects
•
-
the lower oxidation potential for 2 B f B than B f B . The
-1
relationships. For example, numerous attempts have been
previously made to correlate the oxidation potentials of
substituted benzene derivatives with gas-phase ionization
þ
•þ
2
•
þ
one-electron reduction of BP by two molecules of B to form
•
þ
•þ
5
c,6,18
B2 can be attributed to reaction of a (BP /B) cage encounter
pair with a second molecule of benzene or to encounters
potentials to determine the effects of solvation.
Using
the oxidation potentials in Table 1, such a comparison
can now be made with confidence in the reliability of the
Eox values. Shown in the top part of Figure 9 is a plot of
oxidation potentials for the benzene derivatives in Table 1 vs
•
þ
between BP and two molecules of B in close proximity.
The equilibrium involving the radical cations of benzene
and the biphenyl derivative BP1 was also analyzed using the
kinetic approaches described above for the benzene/BP2
system. Because of the smaller Eox gap for the benzene/
the corresponding adiabatic ionization potentials (IP ),
a
1
9
where the latter are available. The plot leads to several
important conclusions. First, the correlation for benzene and
7
-1 -1
BP1 pair, the rate constant k (∼3.1 ꢀ 10 M
s ) is nearly
-
1
an order of magnitude larger than that of benzene/BP2,
which precluded analyses at high benzene concentrations.
Nevertheless, as shown in Figure 8, k is still equal to k (5.8ꢀ
(18) For representative examples, see: (a) Miller, L. L.; Nordblom, G. D.;
1
2
Mayeda, E. A. J. Org. Chem. 1972, 37, 916. (b) Nelsen, S. F.; Peacock, V.;
Weisman, G. R. J. Am. Chem. Soc. 1976, 98, 5269. (c) Loutfy, R. O. J. Chem.
Phys. 1977, 66, 4781. (d) Nelsen, S. F. Isr. J. Chem. 1979, 18, 45. (e) Gassman,
P. G.; Mullins, M. J.; Richtmeier, S.; Dixon, D. A. J. Am. Chem. Soc. 1979,
9
-1 -1
1
0 M
s ) and k affects the equilibrium between the
-2
radical cations. The best-fit for benzene/BP1 electron-trans-
fer equilibrium constant is ∼190, which leads to ΔE
=
ox
1
01, 5793. (f) Klinger, R. J.; Kochi, J. K. J. Am. Chem. Soc. 1980, 102, 4790.
0
.13 V and a benzene oxidation potential of 2.48 V vs SCE.
Because of the necessitated limits on data acquisition at high
benzene concentrations, k was not as well determined as in
(g) Kanno, T.; Oguchi, T.; Sakuragi, H.; Tokumaru, K.; Kobayashi, T. Denki
Kagaku 1981, 49, 134. (h) Eberson, L. Adv. Phys. Org. Chem. 1982, 18, 79. (i)
Masnovi, J. M.; Seddon, E. A.; Kochi, J. K. Can. J. Chem. 1984, 62, 2552. (j)
Mochida, K.; Itani, A.; Yokoyama, M.; Tsuchiya, T.; Worley, S. D.; Kochi,
J. K. Bull. Chem. Soc. Jpn. 1985, 58, 2149. (k) Colonna, M.; Greci, L.; Poloni,
M.; Marrosu, G.; Trazza, A.; Colonna, F. P.; Distefano, G. J. Chem. Soc.,
Perkin Trans. 2 1986, 1229. (l) Anxolabehere, E.; Hapiot, P.; Sav ꢀe ant, J.-M.
J. Electroanal. Chem. 1990, 282, 275. (m) Nau, W. M.; Adam, W.; Klapstein,
D.; Sahin, C.; Walter, H. J. Org. Chem. 1997, 62, 5128.
-
1
the benzene/BP2 case. Nonetheless, the oxidation potential
of benzene derived from this experiment is identical to that
obtained from the benzene/BP2 equilibrations.
(
19) Lias, S. G. Ionization Energy Evaluation. In NIST Chemistry
3
. Discussion
WebBook, NIST Standard Reference Database Number 69; Linstrom, P. J.,
Mallard, W. G., Eds.; National Institute of Standards and Technology:
Gaithersburg, MD, http://webbook.nist.gov. See the Supporting Informa-
tion for ionization potential values.
The accurate oxidation potentials provided by this work
can be used to critically test a variety of structure-property
5
170 J. Org. Chem. Vol. 74, No. 15, 2009

Merkel et al.
JOCFeatured Article
acetonitrile. This effect extends to biphenyl, which has (IP -
a
E )=6.21 ( 0.13 eV, comparable to that of naphthalene. It
ox
is apparent from the data in Figure 9 that the lower solvation
energies associated with oxidation of these two-ring com-
pounds is comparable to the effect of four to six methyl
groups on the benzene ring.
As shown in the top of Figure 9, tert-butylbenzene (B3)
and 1,4-di-tert-butylbenzene (B7) (triangles) both have high-
er oxidation potentials than the line through the methyl-
substituted benzene data points, with the deviation being
twice as large for B7 (0.14 V) as for B3 (0.07 V). These results
can be attributed to steric inhibition of solvation of the
radical cations of the tert-butyl derivatives. Interestingly,
Eox for chlorobenzene (B2) also shows a positive deviation of
0
.03 V relative to the methyl-substituted benzenes. As the
steric size of a chlorine substituent is smaller than a methyl
2
2
group, this result demonstrates that the nature of the
substituents, not simply their size and number, can also play
a role in the differential solvent stabilization of radical
cations.
The biphenyl derivatives for which oxidation potentials
were determined in this study contain a range of electron-
donating and electron-withdrawing substituents. This pro-
vided an opportunity to determine if the oxidation potentials
could be correlated with Hammett-type substituent con-
stants. Prior correlations between oxidation potentials
þ
23
and Hammett/Brown σ values have been observed for
a variety of electron donors, including triarylamines,
24
2
aryl sulfides, selenides and tellurides, vinyl ethers and
5
2
thioethers, styrenes, and arylmethyl radicals. As shown
6
27
28
in Figure 10, there is an excellent correlation between E and
ox
þ
the sum of the σ values for the biphenyl derivatives, which
should be helpful in estimating oxidation potentials of other
FIGURE 9. (Top) Plot of oxidation potentials for benzene and
methyl-substituted benzenes (circles), tert-butyl-substituted ben-
zenes (triangles), and chlorobenzene (square) in acetonitrile vs their
biphenyl derivatives. Not surprisingly, the dependence of
þ
adiabatic ionization potentials (IP ). (Bottom) Plot of IP
a
methyl-substituted benzenes. Interpolated lines are drawn through
the circles.
a
- Eox for
Eox on Σσ is quite strong, with a slope of 9.0 kcal/mol, which
is equivalent to F=-6.7.
It is worth noting that the higher Eox values of tert-
butylbenzene and 1,4-di-tert-butylbenzene relative to to-
luene and p-xylene discussed above are consistent with the
its methyl-substituted derivatives (circles) shows distinct
downward curvature, demonstrating a nonlinear increase
in stabilization of the radical cations by the solvent (acetoni-
trile) as the benzenes are decreasingly substituted. This trend
is more clearly illustrated in the bottom part
of Figure 9, where a plot of (IP -E ) vs number of methyl
þ
less negative σp value of a tert-butyl group (-0.26) com-
2
3
pared to a methyl group (-0.31). However, that the methyl
derivatives have higher adiabatic ionization potentials re-
veals that the lower oxidation potentials of the methyl-
substituted analogues is not simply due to the greater
“electron-releasing” power of a methyl group, but rather
due to the greater solvent-stabilization of cations that carry
the smaller alkyl group. Solvation also plays a clear role in
the reversal of the oxidation potentials for the pairs toluene
(B4)/tert-butylbenzene(B3) and p-xylene(B10)/1,4-di-tert-
butylbenzene(B7) in acetonitrile (Table 1) vs trifluoroacetic
a
ox
groups shows that the difference in solvation energies is
nearly 0.5 eV greater for benzene than hexamethylbenzene.
These results can be explained by decreased solvation with
methyl substitution caused by increased charge delocaliza-
5
k,20
tion
and steric effects. Charge delocalization presumably
also plays a role in the difference of (IP - E ) between
a
ox
2
1
naphthalene (6.36 eV) and benzene (6.76 eV). The 0.4 eV
energy difference reflects the decrease in solvent stabiliza-
tion for oxidation of two-ring vs one-ring compounds in
(
(
22) Meyer, A. J. Chem. Soc., Perkin Trans. 2 1986, 1567.
23) Hansch, C.; Leo, A.; Hoekman, D. Exploring QSAR, Hydrophobic,
(20) (a) Wayner, D. D. M.; Sim, B. A.; Dannenberg, J. J. J. Org. Chem.
991, 56, 4853. (b) Jonsson, M.; Houmam, A.; Jocys, G; Wayner, D. D. M.
Electronic and Steric Constants; American Chemical Society: Washington,
DC, 1995.
1
J. Chem. Soc., Perkin Trans. 2 1999, 425. (c) Svith, H.; Jensen, H.; Almstedt,
J.; Andersson, P.; Lundb €a ck, T.; Daasbjerg, K; Jonsson, M. J. Phys. Chem. A
(24) Dapperheld, S.; Steckhan, E.; Brinkhus, K.-H. G.; Esch, T. Chem.
Ber. 1991, 124, 2557.
2
004, 108, 4805.
21) The equilibrium constants between the radical cations of naphtha-
(25) Engman, L.; Persson, J.; Andersson, C. M.; Bergulnd, M. J. Chem.
Soc., Perkin Trans. 2 1992, 1309.
(
lene and durene as well as between naphthalene and 1,2,3,5-tetramethylben-
zene in acetonitrile were determined previously to be 2.4 and 0.24,
respectively ( Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem.
Soc. 1990, 112, 4290). Based on the oxidation potential of durene (1.75 V vs
SCE) and 1,2,3,5-tetramethylbenzene (1.82 V) an oxidation potential of 1.78
(26) Bauld, N. L.; Alpin, J. T.; Yueh, W.; Endo, S.; Loving, A. J. Phys.
Org. Chem. 1998, 11, 15.
(27) Che, C.-M.; Li, C.-K.; Tang, W.-T.; Yu, W.-Y. J. Chem. Soc., Dalton
Trans. 1992, 3153.
(28) Sim, B. A.; Milne, P. H.; Griller, D.; Wayner, D. D. M. J. Am. Chem.
Soc. 1990, 112, 6635.
(
0.01 V is calculated for naphthalene.
J. Org. Chem. Vol. 74, No. 15, 2009 5171

Merkel et al.
JOCFeatured Article
While electron-exchange equilibrium could be directly ob-
served in some cases, kinetic methods were found to be
complementary when either ΔEox between the substrates
undergoing electron transfer was too large or when dimer
radical cation formation complicated the equilibrium pro-
cesses. Equilibrium constants for dimer radical cation for-
mation were also determined, as well as aborption maxima
and extinction coefficients for the radical cations studied.
The increase in the oxidation potentials of benzene deri-
vatives with decreasing alkyl substitution was found to be
smaller than the corresponding increase in their gas phase
ionization potentials. This result is consistent with increasing
solvent stabilization of radical cations with decreasing alkyl
substitution on the benzene ring. In addition, differential
solvent effects were found to lead to reversal of relative
oxidation potentials in several cases.
þ
FIGURE 10. Plot of Eox vs Σσ for the biphenyl derivatives.
Biphenyls with meta (3 and 3 ) substituents are indicated by squares
0
0
þ
23
The transient absorption methods described herein should
be of general use for the determination of thermodynami-
cally meaningful oxidation potentials. These, in turn, will be
valuable for critically testing a variety of thermodynamic and
thermokinetic correlations. Further work along these lines is
in progress.
and with para (4 and 4 ) substituents by circles. σ values rela-
tive to the biphenyl are as follows: p-Me=-0.31, m-Me=-0.07,
p-Br= 0.15, p-CO
p-CF =0.61, and OCOMe=-0.19. Literature values for Eox of
,3 -dimethylbiphenyl and 4,4 -dimethylbiphenyl were used.
2 2 2
Me =0.49, p-CO Bu= 0.49, m-CO H= 0.32,
3
1
0
3
acid. In acetonitrile, E (B4) is less than E (B3) by 0.02 V
ox
ox
and E (B10) is less than E (B7) by 0.05 V (Table 1). In
ox
ox
Experimental Section
contrast, in CF CO H the relative oxidation potentials are
3
2
Materials. Acetonitrile (99.93þ% Baker, HPLC grade, 5 ppm
reversed; with E (B4)>E (B3) by 0.08 V and E (B10)>
ox
ox
ox
5
E (B7) by 0.03 V. These data clearly demonstrate the risks
k
H O), trifluoroacetic anhydride (>99%), trifluoroacetic acid
2
(
(
ox
99.5þ%), methylene chloride (99.9þ%), and ethyl acetate
of assuming the transferability of even relative oxidation
potentials from one solvent to another. It is worth emphasiz-
ing that such a conclusion would be difficult to make with
confidence in the absence of reliable oxidation potentials.
Additionally, with regard to solvation effects, the similar-
ity of the oxidation potential differences between B7 and B10
in acetonitrile (dielectric constant=36) and ethyl acetate
99.8þ%) were used as received. N-Methylquinolinium hexa-
þ 10
fluorophosphate (NMQ ) was recrystallized from methanol.
Solid benzene and biphenyl derivatives were purified by recrys-
tallization. Most liquids were fractionally distilled. Some high-
purity benzene derivatives were on hand, including 99.99%
toluene (EMD OmniSolv) and vacuum-sealed ampules formerly
available from the American Petroleum Institute. Particular care
was taken to purify benzene because traces of toluene or other
more easily oxidized derivatives could competitively form radical
cations under experimental conditions. Benzene (99.96%, EMD
OmniSolv) was first purified by several cycles of partial (∼80%)
freezing with removal of the unfrozen liquid, followed by frac-
tional distillation from calcium hydride. The resulting material
contained e1 ppm of toluene and other benzene derivatives as
determined by gas chromatographic analysis.
(dielectric constant=6) implies that solvent dielectric con-
stant is not a major differentiating factor. The trend observed
for the four solvents studied does suggest a possible correla-
tion with solvent β, a measure of hydrogen bond accepting
2
9
ability. Reduced steric hindrance and/or positive charge
localization on the methyl hydrogen atoms of B10 due to
hyperconjugation may result in greater solvation stabiliza-
tion in hydrogen bond accepting solvents such as acetonitrile
and ethyl acetate. For the low-β solvents, the lowering of
the relative Eox of B10 in methylene chloride (n = 1.424)
compared to trifluoroacetic acid (n = 1.285) suggests that
radical cations with greater positive charge localization
may also undergo greater solvation stabilization in higher
polarizability solvents.
Biphenyl derivatives were commercially available except for
0
n-butyl 4 -(trifluoromethyl)biphenyl-4-carboxylate (BP1), which
was prepared as described below.
0
Synthesis of 4 -Trifluoromethylbiphenyl-4-carboxylic Acid. A
0
mixture of 1.35 g of 4 -carboxyphenylboronic acid (8.13 mmol),
0
0
.12 g of triphenylphospine (0.47 mmol), 1.1 mL of 4 -bromo-
benzotrifluoride (7.75 mmol), 0.0345 g of palladium acetate
0.16 mmol), and 1.6 g of sodium carbonate (15.5 mmol) in
0 mL of a 1:1 acetone/water was refluxed under a nitrogen
atmosphere for 5 h. After being cooled to room temperature, the
reaction mixture was filtered through a pad of silica gel and
partially concentrated in vacuo. The residue was filtered to
remove palladium black, and the filtrate was acidified with
HCl. The resulting white solid was collected, washed with water,
(
3
4
. Conclusion
Accurate oxidation potentials have been determined in
acetonitrile for a large number of biphenyl and benzene
derivatives, including benzene and toluene, which have
not been possible to measure by electrochemical methods.
The oxidation potentials were determined by measuring
equilibrium constants for electron exchange between neutral
electron donors and their corresponding radical cations
using several nanosecond transient absorption methods.
and dried (1.84 g, 89%). The crude product was used without
1
further purification: H NMR (500 MHz DMSO-d ) δ 13.08 (br,
6
1.0 H), 8.07 (d, J=15 Hz, 2.1 H), 7.98 (d, J=12.5 Hz, 1.9 H), 7.87
(d, 2.0 H), 7.85 (d, 2.0 H).
3
0
0
Synthesis of n-Butyl 4 -(Trifluoromethyl)biphenyl-4-carboxy-
0
late (BP1). A mixture of 1.84 g of 4 -trifluoromethylbiphenyl-
4
-carboxylic acid, 30 mL of n-butanol, 0.5 mL of sulfuric acid,
(
29) Kamlet, M. J.; Abboud, J-L. M.; Abraham, M. H.; Taft, R. W. J.
Org. Chem. 1983, 48, 2887 (β values: ethyl acetate=0.45; acetonitrile=0.31;
methylene chloride=0.0; trifluoroacetic acid ∼ 0.0).
(30) Korolev, D. N.; Bumagin, N. A. Tetrahedron Lett. 2005, 46, 5751.
5
172 J. Org. Chem. Vol. 74, No. 15, 2009

Merkel et al.
JOCFeatured Article
þ
˚
and 3 A molecular sieves was refluxed under a nitrogen atmo-
1 mM NMQ sensitizer along with 0.5 M toluene in air-
saturated acetonitrile. For the measurement of Eox values of
toluene and electron donors with higher or slightly lower
Eox values, a codonor was not used, and donor concentrations
were selected that were sufficiently high for rapid electron
sphere for 16 h. The mixture was filtered while hot, and the sieves
were rinsed with diethyl ether. The filtrate was successively washed
with water and 5% aq NaHCO , dried over anhydrous MgSO ,
and concentrated in vacuo to give a pale yellow solid. Purification
by column chromatography (SiO , 99:1 hexanes/EtOAc) followed
by two recrystallizations from 3:1 2-propanol/water gave white
3
4
1
þ
2
transfer to NMQ *.
1
þ
The energy of NMQ * was estimated to be 3.53 eV from the
intersection of absorption and fluorescence spectra. A laser
photolysis equilibration procedure was used to determine the
1
plates (0.87 g, 35%; mp 54-55 ꢀC): H NMR (400 MHz CDCl ) δ
3
8
4
.20 (d, J=4.2 Hz, 1.9 H), 7.78 (s, 3.9 H), 7.72 (d, J = 4 Hz, 2.0 H),
.42 (t, J=6.8Hz, 2.0 H), 1.85 (quin, 2.1H), 1.57 (sext, 2.0 H), 1.06
þ
31
reductionpotentialfor NMQ inacetonitrile (-0.91 V vs SCE).
1
3
1
þ
(t, J = 8.2 Hz, 3.0 H); C NMR (125 MHz CDCl , with
assignments based on DEPT-135 and HSQC spectra, and a gated
decoupling experiment to obtain accurate peak integrals) δ 166.3
Thus, the reduction potential of NMQ * is ∼2.62 V vs SCE,
sufficient to oxidize all donors used in this study, including
benzene.
3
0
0
(
1
CdO), 143.9 (C1 or C1 ), 143.6 (C1 or C1 ), 130.2 (C3/5 þ C4),
In some cases, small amounts (0.5-1.0% by volume) of
trifluoroacetic anhydride were added to solutions to scavenge
traces of water that could otherwise decrease radical cation
lifetimes. This resulted in significant lifetime increases (e.g., to
25 μs) for many radical cations and greatly simplified some
equilibrium and kinetic measurements. Minimal lifetime
changes with added trifluoroacetic anhydride were observed
for benzene and toluene radical cations whose decay may be
controlled by reaction with solvent and return electron transfer
from superoxide. It was also found that addition of small
amounts (∼1.0% by volume) of trifluoroacetic acid produced
similar increases in the radical cation lifetimes and their decays
became strictly first order, even at relatively high laser powers.
0 0 0
30.1 (C4 , q, J =32.4 Hz), 127.6 (C2 /6 ), 127.2 (C2/6), 125.8
CF
0
0
(
C3 /5 , q, JCF=3.5 Hz), 124.15 (CF
0.8, 19.3, 13.8, CH CH CH CH ]. Anal. Calcd for C18
C, 67.07; H, 5.32. Found: C, 67.04; H, 5.28.
3
, q, JCF=270 Hz), [65.01,
3
2
2
2
3
17 3 2
H F O :
Laser Flash Photolysis Measurements. Radical cation absorp-
tion spectra and transient kinetics measurements were per-
formed using a nanosecond laser flash photolysis apparatus.
A Lambda Physik Lextra 50 XeCl excimer laser was used to
pump a Lambda Physik 3002 dye laser, providing approxi-
mately 7 ns pulses with an energy of about 2 mJ. Most
measurements were carried out using 343 nm excitation ob-
tained with p-terphenyl as the laser dye. Transient absorptions
were monitored at 90ꢀ to the laser excitation using pulsed xenon
lamps, timing shutters, a monochromator, and a photomulti-
plier tube for kinetic measurements or a diode array detector for
obtaining transient absorption spectra. For kinetic analyses
the signal from the photomultiplier tube was directed into a
Tektronix TDS 620 digitizing oscilloscope and then to a com-
puter for viewing, storage, and analysis. Beam energies were
typically attenuated to less than 1 mJ per pulse to reduce second
order radical ion decay and minimize photodegradation. Data
were averaged over 20-40 pulses and experiments were carried
out at ambient temperature (∼295 K).
•
-
This observation may be due to protonation of O
2
under
these conditions, thus reducing second-order return-electron
transfer. Radical ion decay is further discussed in the Support-
ing Information.
Acknowledgment. Research was supported by a grant
from the National Science Foundation (CHE-0749919).
Supporting Information Available: Benzene radical cation
absorption spectra as a function of [benzene]; analysis of radical
cation equilibria; fitting of spectra of equilibrated radical cat-
ions; Keq relationship ladder showing the pairwise comparisons
used to obtain E values; ionization potentials; rate constants for
ox
radical decay. H and C NMR spectra of BP1. This material is
available free of charge via the Internet at http://pubs.acs.org.
In addition to the electron donors whose radical cation
equilibria were to be measured, solutions usually contained
1
13
(31) Merkel, P. B.; Luo, P.; Dinnocenzo, J. P.; Farid, S. Manuscript in
preparation.
J. Org. Chem. Vol. 74, No. 15, 2009 5173