Reaction of phenols with the 2,2-diphenyl-1-picrylhydrazyl radical. Kinetics and DFT calculations applied to determine ArO-H bond dissociation enthalpies and reaction mechanism

Article abstract of DOI:10.1021/jo8016555

(Figure Presented) The formal H-atom abstraction by the 2,2-diphenyl-1-picrylhydrazyl (dpph^?) radical from 27 phenols and two unsaturated hydrocarbons has been investigated by a combination of kinetic measurements in apolar solvents and density functional theory (DFT). The computed minimum energy structure of dpph^? shows that the access to its divalent N is strongly hindered by an ortho H atom on each of the phenyl rings and by the o-NO₂ groups of the picryl ring. Remarkably small Arrhenius pre-exponential factors for the phenols [range (1.3-19) × 10⁵ M^-1 s^-1] are attributed to steric effects. Indeed, the entropy barrier accounts for up to ca. 70% of the free-energy barrier to reaction. Nevertheless, rate differences for different phenols are largely due to differences in the activation energy, E_a,1 (range 2 to 10 kcal/mol). In phenols, electronic effects of the substituents and intramolecular H-bonds have a large influence on the activation energies and on the ArO-H BDEs. There is a linear Evans-Polanyi relationship between E _a,1 and the ArO-H BDEs: E_a,1/kcal x mol^-1 = 0.918 BDE(ArO-H)/kcal x mol^-1 - 70.273. The proportionality constant, 0.918, is large and implies a "late" or "product-like" transition state (TS), a conclusion that is congruent with the small deuterium kinetic isotope effects (range 1.3-3.3). This Evans-Polanyi relationship, though questionable on theoretical grounds, has profitably been used to estimate several ArO-H BDEs. Experimental ArO-H BDEs are generally in good agreement with the DFT calculations. Significant deviations between experimental and DFT calculated ArO-H BDEs were found, however, when an intramolecular H-bond to the O^? center was present in the phenoxyl radical, e.g., in ortho semiquinone radicals. In these cases, the coupled cluster with single and double excitations correlated wave function technique with complete basis set extrapolation gave excellent results. The TSs for the reactions of dpph ^? with phenol, 3- and 4-methoxyphenol, and 1,4-cyclohexadiene were also computed. Surprisingly, these TS structures for the phenols show that the reactions cannot be described as occurring exclusively by either a HAT or a PCET mechanism, while with 1,4-cyclohexadiene the PCET character in the reaction coordinate is much better defined and shows a strong π-π stacking interaction between the incipient cyclohexadienyl radical and a phenyl ring of the dpph^? radical.

Full text of DOI:10.1021/jo8016555

Reaction of Phenols with the 2,2-Diphenyl-1-picrylhydrazyl Radical.
Kinetics and DFT Calculations Applied To Determine ArO-H Bond
Dissociation Enthalpies and Reaction Mechanism
Mario C. Foti,*,† Carmelo Daquino,† Iain D. Mackie,‡ Gino A. DiLabio,*,‡ and
K. U. Ingold§
Istituto di Chimica Biomolecolare del CNR, Via del Santuario 110, I-95028 ValVerde (CT) Italy, National
Institute for Nanotechnology, National Research Council of Canada 11421 Saskatchewan DriVe, Edmonton, AB,
Canada T6G 2M9, and National Research Council of Canada, 100 Sussex DriVe, Ottawa, ON, Canada K1A 0R6
mario.foti@icb.cnr.it; gino.dilabio@nrc-cnrc.gc.ca
ReceiVed August 12, 2008
The formal H-atom abstraction by the 2,2-diphenyl-1-picrylhydrazyl (dpph•) radical from 27 phenols and
two unsaturated hydrocarbons has been investigated by a combination of kinetic measurements in apolar solvents
and density functional theory (DFT). The computed minimum energy structure of dpph• shows that the access
to its divalent N is strongly hindered by an ortho H atom on each of the phenyl rings and by the o-NO2 groups
of the picryl ring. Remarkably small Arrhenius pre-exponential factors for the phenols [range (1.3-19) × 105
M-1 s-1] are attributed to steric effects. Indeed, the entropy barrier accounts for up to ca. 70% of the free-
energy barrier to reaction. Nevertheless, rate differences for different phenols are largely due to differences in
the activation energy, Ea,1 (range 2 to 10 kcal/mol). In phenols, electronic effects of the substituents and
intramolecular H-bonds have a large influence on the activation energies and on the ArO-H BDEs. There is
a linear Evans-Polanyi relationship between Ea,1 and the ArO-H BDEs: Ea,1/kcal × mol-1 ) 0.918 BDE(ArO-
H)/kcal × mol-1 - 70.273. The proportionality constant, 0.918, is large and implies a “late” or “product-like”
transition state (TS), a conclusion that is congruent with the small deuterium kinetic isotope effects (range
1.3-3.3). This Evans-Polanyi relationship, though questionable on theoretical grounds, has profitably been
used to estimate several ArO-H BDEs. Experimental ArO-H BDEs are generally in good agreement with the
DFT calculations. Significant deviations between experimental and DFT calculated ArO-H BDEs were found,
however, when an intramolecular H-bond to the O• center was present in the phenoxyl radical, e.g., in ortho
semiquinone radicals. In these cases, the coupled cluster with single and double excitations correlated wave
function technique with complete basis set extrapolation gave excellent results. The TSs for the reactions of
dpph• with phenol, 3- and 4-methoxyphenol, and 1,4-cyclohexadiene were also computed. Surprisingly, these
TS structures for the phenols show that the reactions cannot be described as occurring exclusively by either
a HAT or a PCET mechanism, while with 1,4-cyclohexadiene the PCET character in the reaction coordinate
is much better defined and shows a strong π-π stacking interaction between the incipient cyclohexadienyl
radical and a phenyl ring of the dpph• radical.
Introduction
hydrogen-atom abstractions from carbon,2 nitrogen,3 sulfur,4 and
oxygen,5,6 particularly from phenols.5 This radical is popular7
The nitrogen-centered radical 2,2-diphenyl-1-picrylhydrazyl
(dpph•)1 has been extensively employed in kinetic studies of
‡ National Institute for Nanotechnology.
§ National Research Council of Canada.
† Istituto di Chimica Biomolecolare del CNR.
9270 J. Org. Chem. 2008, 73, 9270–9282
10.1021/jo8016555 CCC: $40.75 2008 American Chemical Society
Published on Web 11/08/2008

Reaction of Phenols with the 2,2-Diphenyl-1-picrylhydrazyl Radical
for such studies because it is monomeric in solution and air
stable,1 is commercially available, and is strongly colored. This
last property allows the course of reaction to be monitored using
conventional UV-vis spectroscopy. Phenols (ArOH) have been
favored substrates for study, frequently because their reactions
with dpph• (reaction 1) are presumed to be similar to those
important in their antioxidant reactions with peroxyl radicals,
ROO• (reaction 2). However, the thermodynamics of H-atom
abstraction and the steric requirements of these two radicals
are very different. Indeed, peroxyl radicals usually react with
any particular phenol some 3 orders of magnitude more rapidly
than does dpph•.5a The divalent nitrogen of dpph• is sterically
shielded (vide infra), which reduces this radical’s reactivity and,
in H-atom abstractions from phenols, produces unusually small
unpaired electron (i.e., the radical center), a pretransition state
(pre-TS), hydrogen-bonded complex is formed between the OH
and a lone pair on Y•. The proton is then transferred from its
two bonding electrons to the radical’s lone pair with the
accompanying electron moving from the 2p lone pair on the
phenol’s oxygen atom to the radical’s singly occupied molecular
orbital (SOMO). The formation of a complex with a hydrogen
bond (HB) between ArOH and Y• causes the reaction to be
entropically disfavored (∆SHB ≈ -11 cal/mol ·K).14 However,
formation of a HB complex also has effects that strongly favor
reaction, whether by the HAT or PCET mechanism. Thus, the
formation of the pre-TS HB complex causes the O atom in
ArOH and the Y• radical center to approach each other more
closely than they would if a HB was not formed. Since the
intersecting ArO-H and Y-H Morse-like potential energy curves
have their minima closer together, the difference in enthalpy
between the pre-TS HB complex and the TS is smaller than
the corresponding enthalpy difference for the same reaction
occurring without HB formation. Furthermore, with a pre-TS
HB complex, the barrier to reaction will not only be lower, it
will also be narrower than in the absence of HB formation. The
narrower barrier will enhance the reaction rate by enhancing
quantum-mechanical tunnelling of the proton and electron.
Finally, proton transfer in the HB complex will be a first-order
process and, as such, will generally have an enhanced pre-
exponential factor compared with a second-order reaction.15
One question we address in the present work, using a
combination of experimental kinetic work and theoretical
calculations, is: Does H-atom abstraction by dpph• from
phenols, reaction 1, occur by the HAT or PCET mechanism?
From the experimental side, this question relates only to
reactions carried out in alkane solvents where neither the SPLET
nor the ETPT mechanism can occur. Both the HAT and PCET
mechanisms will be disfavored by steric congestion around the
divalent nitrogen atom. HAT and PCET are also likely to be
disfavored by the fact that there are other hydrogen bond
accepting groups in dpph• (e.g., the three nitro groups) to which
the phenol might preferentially hydrogen bond, particularly since
the electron withdrawing picryl group will make the divalent
nitrogen electron-deficient. The results of DFT calculations on
the transition state for reaction 1 provide fundamental insights
into the HAT vs PCET reaction mechanism.
Arrhenius factors,5a range 4.5 × 103-1.9 × 106 M-1 s-1
,
whereas “normal” values of A for H-atom abstractions are
generally considered to be about 108.5 M-1 s-1.8 The N-H bond
dissociation enthalpy (BDE) of dpph-H is ∼78.9 kcal/mol,5a
and for most phenols, reaction 1 will be endothermic and slower
than its reverse (reaction -1), whereas reaction 2 will be
exothermic for many phenols9 and faster than its reverse
(reaction -2). The entropy changes in reactions 1 and 2 are
negligible; therefore, ∆G1 ≈ ∆H1 and ∆G2 ≈ ∆H2.5a,10
ArOH + dpph• a ArO• + dpph-H
ArOH + ROO• a ArO• + ROOH
(1)
(2)
Formal H-atom abstractions from phenols, ArOH + Y• a
ArO• + YH, are now recognized to proceed by four different
mechanisms:11 hydrogen-atom transfer (HAT), proton-coupled
electron-transfer (PCET),12 sequential proton-loss electron-
transfer (SPLET),5c-e and electron-transfer proton-loss (ET-PT)
with the proton being initially transferred to a solvent molecule
to which the phenol is hydrogen-bonded.13 In the HAT
mechanism, the proton together with one of its two bonding
electrons are transferred to Y•. In the PCET mechanism of
phenol oxidation by a Y• radical that possesses one, or more,
lone pairs of electrons on the atom that (formally) bears the
(1) Goldschmidt, S.; Renn, K. Ber. Chem. 1922, 55B, 628–643. Walter, R. I.
J. Am. Chem. Soc. 1966, 88, 1930–1937.
(2) Valgimigli, L.; Ingold, K. U.; Lusztyk, J. J. Org. Chem. 1996, 61, 7947–
7950. Pitts, J. N.; Schuck, E. A.; Wan, J. K. S. J. Am. Chem. Soc. 1964, 86,
296–297. Kubo, J. Ind. Eng. Chem. Res. 1998, 37, 4492–4500. (d) Baciocchi,
E.; Calcagni, A.; Lanzalunga, O. J. Org. Chem. 2008, 73, 4110–4115.
(3) Mulder, P.; Litwinienko, G.; Shuqiong, L.; MacLean, P. D.; Barclay,
L. R. C.; Ingold, K. U. Chem. Res. Toxicol. 2006, 19, 79–85. McGowan, J. C.;
Powell, T.; Raw, R. J. Chem. Soc. 1959, 3103–3110.
Our earlier experimental work5a has been significantly
expanded. Arrhenius parameters for hydrogen atom abstraction
from 27 phenols (see Chart 1) in saturated hydrocarbon solvents
are now reported. These activation enthalpies were used to
estimate ArO-H bond dissociation enthalpies (BDEs) for many
of the phenols and these ArO-H BDEs are shown to be in good
agreement with literature values and with the results of DFT
calculations.
(4) Russell, K. E. J. Phys. Chem. 1954, 58, 437–439. Ewald, A. H. Trans.
Faraday Soc. 1959, 55, 792–797.
(5) (a) Foti, M. C.; Daquino, C. Chem. Commun. 2006, 3252–3254. (b) Foti,
M. C.; Daquino, C.; Geraci, C. J. Org. Chem. 2004, 69, 2309–2314. (c)
Litwinienko, G.; Ingold, K. U. J. Org. Chem. 2003, 68, 3433–3438. (d)
Litwinienko, G.; Ingold, K. U. J. Org. Chem. 2004, 69, 5888–5896. (e)
Litwinienko, G.; Ingold, K. U. J. Org. Chem. 2005, 70, 8982–8990.
(6) Astolfi, P.; Greci, L.; Paul, T.; Ingold, K. U. J. Chem. Soc., Perkin Trans.
2 2001, 1631–1633.
Results
(7) SciFinder currently gives 6811 references with the keyword “dpph”.
(8) Benson, S. W. Thermochemical Kinetics, 2nd ed.; Wiley: New York,
1976.
(9) The BDE of ROO-H is relatively constant, 86-88 kcal/mol.
(10) (a) Lucarini, M.; Pedulli, G. F.; Cipollone, M. J. Org. Chem. 1994, 59,
5063–5070. (b) Lucarini, M.; Pedrielli, P.; Peduli, G. F.; Cabiddu, S.; Fattuoni,
C. J. Org. Chem. 1996, 61, 9259–9263.
(11) Litwinienko, G.; Ingold, K. U. Acc. Chem. Res. 2007, 40, 222–230.
(12) (a) Mayer, J. M.; Hrovat, D. A.; Thomas, J. L.; Borden, W. T. J. Am.
Chem. Soc. 2002, 124, 11142–11147. (b) DiLabio, G. A.; Johnson, E. R J. Am.
Chem. Soc. 2007, 129, 6199–6203v. (c) Tishchenko, O.; Truhlar, D. G.;
Ceulemans, A.; Nguyen, M. T. J. Am. Chem. Soc. 2008, 130, 7000–7010.
(13) Galian, R. E.; Litwinienko, G.; Perez-Prieto, J.; Ingold, K. U. J. Am.
Chem. Soc. 2007, 129, 9280–9281.
Experimental Study of the Kinetics of Reaction 1. The
dpph• radical absorbs strongly in the visible part of the spectrum
and in apolar solvents, λmax ≈ 512 nm and ꢀmax ≈ 12000 M-1
(14) de Heer, M. I.; Korth, H-G.; Mulder, P. J. Org. Chem. 1999, 64, 6969–
6975.
(15) Although the rate constant expression derived from Transition State
Theory does not contain any explicit term related to a pre-TS complex, the pre-
TS complex does influence the activation energy through a tunnelling factor.
See: Uc, V. H.; Alvarez-Idaboy, J. R.; Galano, A.; Garcia-Cruz, I.; Vivier-Bunge,
A. J. Phys. Chem. A 2006, 110, 10155–10162, and cited references. Donahue,
N. M Chem. ReV. 2003, 103, 4593–4604.
J. Org. Chem. Vol. 73, No. 23, 2008 9271

Foti et al.
CHART 1. Structures of 27 Phenols and Two Hydrocarbons Employed in the Current Kinetic Studies
cm-1. Conventional UV-vis spectroscopy was used to monitor
the rates of phenol + dpph• reactions at several temperatures
from 5 to 70 °C using a thermostatted ((0.2 °C) stopped-flow
spectrophotometer. In most cases, the solvent was neat cyclohexane
or n-hexane but for a few phenols a small amount of ethyl acetate
or CH2Cl2 (<2% by volume in the final solution) was added to
facilit+ate dissolution and, for solubility reasons, for two phenols
the solvent had to be neat methylene chloride. All solutions were
purged with N2 prior to kinetic measurements. Reactant concentra-
tions were corrected for thermal expansion of the solvent.
The concentration of dpph• was generally in the range
20-100 µM. Although the phenol concentrations were larger
than the dpph• concentrations (often as much as 10-50 times
greater) in many cases the loss of dpph• did not follow first-
order kinetics over the entire course of the reaction. This
behavior arises because reaction 1 is reversible. It is followed
by two irreversible reactions, 3 and 4:
(16) Denisov, E. T.; Khudyakov, I. V. Chem. ReV. 1987, 87, 1313–1357.
(17) Snelgrove, D. W.; Lusztyk, J.; Banks, J. T.; Mulder, P.; Ingold, K. U.
J. Am. Chem. Soc. 2001, 123, 469–477.
9272 J. Org. Chem. Vol. 73, No. 23, 2008

Reaction of Phenols with the 2,2-Diphenyl-1-picrylhydrazyl Radical
ArO•+dpph• f products
ArO•+ArO• f dimer or disproportionation products (4)
(3)
Rate constants, k1, determined as indicated above, for the
reaction of dpph• with phenols 1-27 and hydrocarbons 28 and
29 at 25 °C in an alkane solvent are given in Table 1. The
temperature dependence of k1 was determined in cyclohexane
from 5 to 70 °C. All Arrhenius plots showed good linearity (r2
> 0.98). The derived pre-exponential factors, A1, and activation
energies, Ea1, are included in Table 1. These two parameters
yield the enthalpy and entropy of activation via eqs II and III21
The desired second-order rate constants, k1, were determined
using various strategies. In those cases where the loss of dpph•
followed first-order kinetics over the entire course of reaction,
values of k1 were obtained from the slope of plots of the
obserVed (pseudofirst-order) rate constants vs [ArOH], kobs
)
k1[ArOH] + k0. In all other cases, values of k1 were determined
either from the initial rate of loss of dpph• or by kinetic
modeling. Modeling involved the fitting of simulated curves of
dpph• loss, calculated using reactions 1, 3, and 4, to the
experimental curves over the entire course of the reaction. These
simulations were done with the following assumptions: (i) The
value of the rate constant, k3, was e A1 (the A-factor5a of
reaction 1). (ii) k4 ≈ 108-109 M-1 s-1 (in most cases).16 (iii)
In a few cases, it was necessary to introduce into the modeling
scheme the slow dissociation of the C-C or C-O phenoxyl
radical dimer, i.e., the reverse of reaction 4, typically k-4 < 50
∆Hq1)Ea,1-RT
∆Sq1 ) R[ln A1 + ln(h ⁄ kBT) - 1]
(II)
(III)
where the temperature to be used is the midpoint of the
experimental range, i.e., T≈310 K; RT ≈ 0.62 kcal/mol.
Deuterium Kinetic Isotope Effects. Cyclohexane solutions
containing dpph• together with phenols 1, 2, 3, 9 (the solution
of 9 also contained <2% by volume of CH2Cl2), 10, and 23
were shaken with a few drops of D2O. Complete deuterium
exchange occured within a few minutes.23 The clear solutions
were used to determine k1(D) values at ∼298 K. Values of k1(H)
were also obtained using similar solutions containing a few drops
of H2O. The ratios, k1(H)/k1(D), are given in Table 2.
s-1
k1 and k-1
.
16 Modeling provided optimized values of the rate constants,
.
The poor solubility of phenols 13 and 22 in cyclohexane and
n-hexane necessitated measurement of the kinetics in neat
CH2Cl2. In these two systems, the absolute concentrations of
the phenols were comparatively low, ca. 200 µM (though still
Theoretical Calculations.24
The dpph• Radical. The energy-optimized25-27 structure of
dpph• is shown in Figure 1, and its Cartesian coordinates are
provided in the Supporting Information (SI). The steric crowding
in dpph• is evident from parts a and b of Figure 1: an ortho-
hydrogen atom on each of the two phenyl groups shield the
formal N• radical center (hereafter, simply N•) from above and
below the C-N•-N plane [R(N•-HC))2.73, 2.50 Å], and a
nitro group hinders access to the N• lone pair [R(ONO-N•))
2.77 Å]. This shielding is crucially important for the reactivity
and stability of dpph• because it prevents direct access to the
unsatisfied valence of N•. Figure 1b shows the singly occupied
molecular orbital (SOMO) of dpph•. The iso-surface shows that
while much of the unpaired electron is shared between the two
N atoms, the spin is also well-delocalized into the ring moieties.
Bond Dissociation Enthalpies of Substituted Phenols.
Density functional theory (DFT)-based approaches are able to
higher than [dpph•]). The measured second-order rate constants
-1
in CH2Cl2 at 298 K, kC1 H Cl , were 3000 ( 18 and 297 ( 8 M
s-1 for 13 and 22, respectively. Since CH2Cl2 is a hydrogen
bond acceptor (HBA), these two k1 values will be lower than
in a (non-HBA) saturated hydrocarbon,11,17 and therefore, they
are not directly comparable with the k1 values for the other
phenols. Fortunately, the magnitude of the CH2Cl2-induced rate
reduction for each phenol will be the same whether reaction 1
occurs primarily by the HAT or by the PCET mechanism.11
The rate reductions arise because only that (small) fraction of
phenol molecules that are not hydrogen bonded to a CH2Cl2
molecule can react with an attacking dpph• radical.11,17 The
required room temperature rate constant in an alkane solvent,
k1, can be reliably calculated from the measured room temper-
2
2
11,17
S
ature rate constant, k1 , in a neat HBA solvent, S, via eq I.
(21) Moore, J. W.; Pearson, R. G. Kinetics and Mechanism, 3rd ed.; Wiley:
New York, 1981.
log k1 ) log kS1 + 8.3R2Hꢀ2H
(I)
(22) (a) Mulder, P.; Korth, H.-G.; Pratt, D. A.; DiLabio, G. A; Valgimigli,
L.; Pedulli, G. F.; Ingold, K. U J. Phys. Chem. A 2005, 109, 2647–2655. (b)
Brigati, G.; Lucarini, M.; Mugnaini, V.; Pedulli, G. F. J. Org. Chem. 2002, 67,
4828–4832. (c) Lucarini, M.; Pedulli, G. F.; Guerra, M. Chem.sEur. J. 2004,
10, 933–939. (d) Wayner, D. D. M.; Lusztyk, J. E.; Page, D.; Ingold, K. U.;
Mulder, P.; Laarhoven, L. J. J.; Aldrich, H. S. J. Am. Chem. Soc. 1995, 117,
8737–8744. (e) Wayner, D. D. M.; Lusztyk, J.; Ingold, K. U.; Mulder, P. J.
Org. Chem. 1996, 61, 6430–6433.
H
H
In this equation, R2 and ꢀ2 quantify the hydrogen-bond
donor (HBD) activity of the phenol reactant18 and the HBA
activity of S,19 respectively. From the 298 K rate constants for
the reaction of dpph• with catechol 19 (R2 ) 0.726)20 in
H
cyclohexane (1900 M-1 s-1) and in CH2Cl2 (85 M-1 s-1), the
most appropriate value of ꢀ2H for CH2Cl2 was calculated (eq I)
(23) See, e.g.: Gardner, D. V.; Howard, J. A.; Ingold, K. U. Can. J. Chem.
1964, 42, 2847–2851, and references cited therein.
H
to be ca. 0.2. The R2 values of 13 and 22 have not been
(24) All calculations were performed using: Gaussian 03, Revision C.02:
Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant,
J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi,
M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara,
M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda,
Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross,
J. B.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma,
K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford,
S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi,
I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian, Inc., Pittsburgh PA, 2003.
(25) Using B3LYP26,27/6-31G(d).
H
reported. An R2 ≈ 0.3 was estimated for 13 from the value
for 2-methoxyphenol (R2 ) 0.24-0.26)20 with the inclusion
H
H
of a statistical factor of 2. An R2 ≈ 0.7 was estimated for 22
because such a value is intermediate between the values for
20
H
H
catechol (R2 ) 0.726) and 3,5-di-tert-butylcatechol (R2
)
0.67).20 On the basis of these estimates of R2 and ꢀ2H, the k1
values for 13 and 22 were calculated to be ∼9400 and ∼4300
M-1 s-1, respectively, in an alkane solvent.
H
(18) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Duce, P. P.; Morris, J. J.;
Taylor, P. J. J. Chem. Soc., Perkin Trans. 2 1989, 699–711.
(19) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Morris, J. J.; Taylor, P. J.
J. Chem. Soc., Perkin Trans. 2 1990, 521–529.
(20) Foti, M. C.; Barclay, L. R. C.; Ingold, K. U. J. Am. Chem. Soc. 2002,
124, 12881–12888.
(26) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(27) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785–789.
J. Org. Chem. Vol. 73, No. 23, 2008 9273

Foti et al.
TABLE 1. Rate Constants k1 and k-1 (M-1 s-1, 298 K), Arrhenius Parameters A1 (M-1 s-1), and Ea,1 (kcal/mol) for the Reactions of Phenols
1-27 and Hydrocarbons 28 and 29 with dpph• in an Alkane Solvent (Experimental/Estimateda ArO-H BDEs (kcal/mol) in Solution at 298 K
(1 M Standard State), Error ( 1 kcal/mol, and Calculated ArO-H BDEs in the Gas Phase at 298 K (1 atm Standard State))b
c
d
e
no.
substituents
k1
k-1 /104
A1 /105
Ea1
OH BDEa,b,f
sourcea
1
2
3
4
5
6
7
8
none
2-OMe
0.10
0.92
50
238
3
180
155
1.4
16.0
1.7
2.9
19.0
1.7
2.9
4.1
16.0
6.6
6.0
11.0
12
2.3
2.3
1.5
1.3
1.5
3.5
7.3
2.9
1.7
7.3
5.9
1.6
1.7
2.0
0.045
4.6
4.1
9.8
7.2
4.9
5.3
6.5
4.4
4.7
8.3
3.6
3.8
3.25
5.1
2.0
3.2
3.3
4.5
5.6
3.9
3.5
4.2
5.8
3.2
2.5
4.85
6.3
6.2
5
86.3 (87.8)
85.1 (85.5)
82.1 (81.7)
82.0g (82.2)
83.9 (84.2)
81.0 (80.2)
81.5 (81.1)
85.7 (86.6)
80.4 (80.0)
80.6 (80.6)
80.0 (80.2)
82.1 (83.0)
e78.7i (79.1)
80.0 (78.9)
78.5 (78.5)
82.1 (80.6)
83.4 (82.1)
80.8 (79.6)
80.7 (77.2)
81.2 (80.7)
83.5 (84.7)
80.0 (75.7)
77.2 (75.3)
81.6 (81.4)
83.4 (83.2)
83.0 (82.3)
80.1 (78.2)
(74.1)
22a
5a
10b
22e, 10b, AR
AR
AR
14
K
eq X
eq X
eq X
eq X
eq X
eq X
4
1.1
2,6-di-MeO
4-MeO
2-MeO-4-CH2CH2OH
2,6-di-MeO-4-Me
2,4-di-MeO
3-MeO
3
0.73
16
9
3,4-di-MeO
sesamol
1800
935
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
2-MeO-4-OH
3-OH-4-MeO
2,5-di-MeO-4-OH
3-OH-4,6-di-MeO
ubiquinol-0
ubiquinol-0 ether
ubiquinol-0 ether
2,4-di-MeO-6-Me
2-OH
sinapic acid
ferulic acid
CAPE
PMHC
2,4,6-tri-Me
2,6-di-Me
4400
200
9400h
999
990
68
11
482
1900
226
0.5
1
1.8
5a
5a
eq X
22c
5a
0.7
1.9
10
5a
4300h
8415
47
eq X
10b
10b
10b
AR
22a
4.3
5
0.91
0.00100l
0.00156l
2,6-di-Me-4-Cl
2,4,6-tri-tBu
11.8
11.5
(74.0)
a The experimental BDEs, when available, are taken from the literature (references indicated); otherwise they have been estimated with the substituent
additivity rules5a,22b (AR) or from Ea,1 (eq X), or from the kinetics (K) for reactions 1 and -1 (as described for 8 in the text). b BDEs computed by
density functional theory are given in parentheses. See text for additional details. The standard states in which the reported ArO-H BDEs are obtained
differ, viz. 1 M for experimental and 1 atm for calculated BDEs; to convert between the two standard states, BDE(1 atm) ) BDE(1 M) + 0.4 kcal/mol
at 298 K. c Reproducibility was better than ( 10%. d Values obtained by simulation (see text). e A factors for 5, 6, 8, 17, 20, 21, and 22 (given in
italics) were assumed to be equal to those for 2, 3, 1, 15, 3, 2, and 19, respectively. f The experimental OH BDEs determined via the electron
paramagnetic resonance, radical buffer technique (refs 10 and 22b, c) were adjusted downward by 1.1 kcal/mol as indicated in ref 22a. g This is the
average of two experimental values (81.3 and 81.8 kcal/mol) and an AR estimation (82.8 kcal/mol). h Rate constant calculated for an alkane solvent
from value measured in CH2Cl2 (see text). i Reaction 1 with this phenol is likely to be thermoneutral or slightly exothermic and thus the ArO-H BDE
obtained by eq X is most likely overestimated (see text). l Rate constant based on initial rates. The kinetics (in the absence of dioxygen) were
complicated by autocatalysis (see the Experimental Section).
TABLE 2. Deuterium Kinetic Isotope Effects for Reaction 1 in
Cyclohexane at ca. 298 K and Reaction Enthalpies ∆H1 (kcal/mol)
those determined by EPR and other methods. The absolute value
computed for the PhO-H BDE is 87.8 kcal/mol, which is in
good agreement with the critically evaluated “best” value of
86.7 ( 0.7 kcal/mol.22a The average deviation between the
calculated and experimental O-H BDEs in Table 1 is ca. 1 kcal/
mol. Note that the estimated error on the measured BDEs is (
1 kcal/mol. A similarly good result is obtained for the N-H BDE
of dpph-H, viz. calculated 79.6 kcal/mol vs measured 78.9 (
0.5 kcal/mol.5a
The largest errors in the calculated BDEs are associated with
catechol 19 and caffeic acid phenethyl ester (CAPE), 22, for
which the O-H BDEs are computed to be, respectively, 3.5
and 4.3 kcal/mol lower than the measured values. A careful
examination of the molecule and radical stabilization enthalp-
ies30 for these two systems reveals that the computed increase
in the strength of the intramolecular HB on going from the
phenol to the phenoxyl radical has probably been overestimated.
This overestimation is probably connected with the well-known
“delocalization problem” in DFT31 because many of the different
functionals we used to explore this problem also gave O-H
b
H
ArOH
k1(H)/k1(D)a
∆H1
phenol (1)
2.0
3.3
3.3
1.4
1.3
1.9
7.4
6.2
3.2
1.1
1.6
2-methoxyphenol (2)
2,6-dimethoxyphenol (3)
3,4-dimethoxyphenol (9)
sesamol (10)
PMHC (23)
-1.7
a These ratios are the average of three different measurements. The
relative error is estimated to be approximately ( 20%. b The N-H BDE
of dpph-H is 78.9 ( 0.5 kcal/mol (see ref 5a).
compute reasonably accurate BDEs in many cases. In this work,
we have calculated the O-H BDEs for the phenols listed in
Table 1 using a DFT-based approach outlined previously by
one of us.28 The method works quite well for many applications
but, like most other DFT methods, is subject to deficiencies
related to improper modeling of van der Waals interactions and
overstabilization from electron delocalization. Nevertheless, the
DFT-calculated O-H BDEs are in very good agreement with
(29) Perdew, J. P. Phys. ReV. B 1986, 33, 8822–882X.
(30) Pratt, D. A.; DiLabio, G. A.; Mulder, P.; Ingold, K. U. Acc. Chem. Res.
2004, 37, 334–340.
(28) Using the approach outlined in: Johnson, E. R.; Clarkin, O. J.; DiLabio,
G. A J. Phys. Chem. A 2003, 46, 9953–9963, which is based on the B326P8629
method.
(31) Johnson, E. R.; Becke, A. D. J. Chem. Phys. 2008, 128, 124105.
9274 J. Org. Chem. Vol. 73, No. 23, 2008

Reaction of Phenols with the 2,2-Diphenyl-1-picrylhydrazyl Radical
FIGURE 1. Perspective images of the calculated structure of dpph•. The constituent atoms are colored as follows: oxygen ) red, nitrogen ) blue,
carbon ) gray, hydrogen ) white. (a) The view is approximately perpendicular to the plane defined by the hydrazyl group N atoms and the C1 of
the picryl ring. (b) The view is approximately in the plane defined by the hydrazyl group N atoms and the C1 of the picryl ring. The singly occupied
molecular orbital (SOMO) is also shown, with the red and green colors representing the two phases of the orbital.
BDEs in 19 and 22 that were lower than the experimental values.
We note that the O-H BDEs obtained from the DFT treatment
of molecules (not both molecules and radicals) containing
intramolecular HBs, for example, 2, 12-18, 20, and 21, are in
rather good agreement with experiment.
The O-H BDEs in catechol (and probably all substituted
phenols) can be calculated accurately using the coupled cluster
with single and double excitations (CCSD) correlated wave
function approach, with correlation consistent basis sets and with
basis set extrapolation.32 This is the case because the CCSD
method does not suffer from the delocalization problems that
plague most DFT methods. Unfortunately, calculations of this
type are extremely time-consuming. However, we have previ-
ously shown that a locally dense basis set (LDBS) approach
can be used to reduce the computational requirements associated
with such calculations.33 In the LDBS approach, regions of
FIGURE 2. Structure of the HB complex of dpph• and p-methox-
chemical importance within a molecule or radical are treated
yphenol to the N• lone pair. This is the prereaction complex formed
using larger basis sets and regions of lesser chemical importance
prior to H atom transfer from the phenol to dpph•. The deviation of
are treated using smaller basis sets.34 This LDBS technique gives
the H-bond from linearity, defined by θ(O-H-N•), is ca. 26.3°. The
extent to which the nascent transferring H atom deviates from residing
an O-H BDE for phenol of 87.2 kcal/mol,35 in excellent
agreement with the measured value given in Table 1 and with
the “best” value of 86.7 ( 0.7 kcal/mol. 22a The CCSD/LDBS
O-H BDE for catechol is 80.3 kcal/mol, which is 6.9 kcal/mol
less than phenol. This ∆BDE value is in excellent accord with
our previous assessment of 7.0 ( 0.2 kcal/mol36 and, within
the error bars, of the experimental ∆BDE value of 5.6 kcal/
mol obtained from Table 1.37
Hydrogen Bonding to dpph. Additional calculations were
performed on a few conformations of HB complexes between
dpph• and p-methoxyphenol38 in order to provide some insight
into the nature of dpph•/ArOH prereaction complexes. An
in the plane defined by the picryl C1 and the hydrazyl N atoms,
ꢁ(H-C-N•-N), is 12.6°. See the Supporting Information for additional
details.
exhaustive search of conformational space was not performed.
However, some important structures were obtained, and a
selection of these are collected in the Supporting Information.
The calculations support our expectations that the three NO2
groups of dpph• are stronger HB acceptors than the lone-pair
on N•.
For the prereaction complex (see Figure 2), the calculations
predict that the OH group of the phenol remains essentially in
the plane of the phenol’s ring upon HB formation, viz.,
ꢁ(H-O-C-C) ) 6.3°. The HB angle, θ(O-H-N•), deviates
from the ideal value of 180° by 26.3 because of steric effects.
The extent to which the nascent transferring H atom deviates
from being in the plane defined by the picryl C1 and the
hydrazyl N atoms, i.e. ꢁ(H-C-N•-N), is 12.6°. Also, the
phenolic ring is, for steric reasons, oriented roughly perpen-
dicular to the plane defined by C-N•-N.39 Similar geometric
parameters were found for prereaction complexes involving 1
and 8 (see the Supporting Information).
(32) We used the basis set extrapolation method suggested in: Martin, J. M. L.
Chem. Phys. Lett. 1996, 259, 669–678.
(33) See, for example: DiLabio, G. A. J. Phys. Chem. A 1999, 51, 11414–
11424. Johnson, E. R.; McKay, D. J. J.; DiLabio, G. A. Chem. Phys. Lett. 2007,
435, 201–207.
(34) In this work, LDBS calculations of the O-H BDEs in phenol and
catechol, viz. C6H5OH f C6H5O• + H• and C6H4OHOH f C6H4OHO• + H•,
used aug-cc-pVDZ and aug-cc-pVTZ basis sets on the bold atoms and cc-pVDZ
basis sets on the remaining atoms.
(35) We note that these calculations are not thorough with respect to basis
set size, basis set extrapolation, or electron correlation. However, more extensive
treatments may not yield results in better agreement with the best available
experimental O-H BDE data. This is exemplified to some extent in: Costa
Cabral, B. J.; Canuto, S. Chem. Phys. Lett. 2005, 406, 300–305, and commented
upon in ref 36.
(36) DiLabio, G. A.; Mulder, P. Chem. Phys. Lett. 2006, 417, 566–569.
(37) It is worthwhile pointing out that these calculations can be further
simplified by using extrapolated MP4(SDQ) energies to correct the CCSD/aug-
cc-pVDZ LDBS energies. Additional information is provided in the Supporting
Information.
(38) Using B3LYP/6-311++G(2d,2p)//B3LYP/6-31G(d) as implemented in
ref 24.
It is very important to point out that we were unable to obtain
a reasonable, bound structure involving HB formation with the
(39) For reference, the “ideal” PCET (HAT) pre-reaction complex structure
would have θ(O-H-N•) ) 180° (180°), ꢁ(H-O-C-C) ) 0° (90°), and ꢁ(H-
C-N•-N) ) 0° (90°), and the phenolic ring co-planar with (perpendicular to)
the C-N•-N plane.
J. Org. Chem. Vol. 73, No. 23, 2008 9275

Foti et al.
FIGURE 3. Calculated transition state structures for hydrogen atom exchange between phenol and dpph•. The view is in the C-N•-N plane with
the phenol ring in the foreground. The constituent atoms are colored as follows: oxygen ) red, nitrogen ) blue, carbon ) gray, hydrogen ) white.
To help guide the eye, bonds between the O, H, and N atoms involved in the H-transfer have been drawn. Key structural information associated
with the atoms at the atom transfer center are given below. The degree to which the OH group is rotated out of the plane of the phenol ring is
indicated by ꢁ(H-O-C-C) and the extent to which the transferring H-atom deviates from coplanarity with the picryl C1 and the hydrazyl N
atoms is indicated by ꢁ(H-C-N•-N). For phenol, (1), TS: R(O-H) ) 1.25 Å, R(H-N•) ) 1.22 Å, θ(O-H-N•) ) 161.7°, ꢁ(H-O-C-C)) 43.8°,
ꢁ(H-C-N•-N)) 40.1°. Related TS structures not shown in this figure: m-methoxyphenol, 8, R(O-H) ) 1.23 Å, R(H-N•) ) 1.24 Å, θ(O-H-N•)
) 162.9°, ꢁ(H-O-C-C)) 44.4°, ꢁ(H-C-N•-N)) 39.6°. p-Methoxyphenol, 4, R(O-H) ) 1.19 Å, R(H-N•) ) 1.22 Å, θ(O-H-N•) ) 163.9°,
ꢁ(H-O-C-C)) 40.9°, ꢁ(H-C-N•-N)) 38.8°. TS molecular orbitals, with relative phases indicated by the red and green colors: (a) Singly
occupied molecular orbital (SOMO). The nominally singly occupied p-type orbital on N• is particularly evident in the magnified view. The component
of the orbital on dpph• is very similar to the SOMO on isolated dpph•; see Figure 1b. (b) A low-energy, doubly occupied molecular orbital
(HOMO-6) showing the lone-pair orbital on N•.
unpaired electron on the N• in dpph•. The formation of a
prereaction complex that would faVor direct hydrogen atom
transfer from the OH group of a phenol to the orbital on the
hydrazyl moiety of dpph• that holds the unpaired electron would
appear to be highly disfaVored.
Transition-State Structures and Kinetics for the
Reaction of dpph• with Three Phenols. The calculated40
transition-state (TS) structure for the reaction of dpph• with 1
is shown in Figure 3 along with some structural parameters for
the TSs involving 1, 4, and 8. For each of these reactions, only
a single TS structure was explored. It is expected that other,
higher energy pathways are also possible for these reactions.
Kinetic parameters associated with the reactions were obtained
from transition-state theory in a manner described previously
and incorporate quantum-mechanical tunnelling.12b
is expected based on the O-H BDEs of the corresponding
phenols, viz. 1 > 8 > 4, see also Table 1. Steric hindrance
causes θ(O-H-N•) in the TSs to be ca. 16 to 18° away from
linearity,39 as it does in the H-bonded prereaction complexes,
see Figure 2. To some extent, steric hindrance in the TS is
responsible for ꢁ(H-O-C-C) dihedral angles that are 40.9 -
44.4° (“ideal” 0° for PCET, 90° for HAT).39 Interestingly, values
of ꢁ(H-C-N•-N), for which the “ideal” value is 0° for PCET
and 90° for HAT,39 track the O-H BDE values in the phenols,
viz. 1 (40.1°) > 8 (39.6°) > 4 (38.8°). Thus, in H-atom exchange
between phenols and dpph• the TS structures deviate consider-
ably from the “ideal” PCET geometry that would presumably
be obtained in the absence of steric hindrance.39
Figure 3 also displays the singly occupied molecular orbital
(SOMO) and a low-lying doubly occupied molecular orbital (six
orbitals below highest-occupied molecular orbital (HOMO), i.e.
HOMO-6) for the TS structure of the reaction of 1 with dpph•,
which are representative of the orbitals of TSs involving other
phenols. A fuller appreciation of the structure of the orbitals
shown in Figure 3 can be obtained by viewing the animations
given in the SI. The orbitals reveal that the transferring H atom
interacts with both the singly occupied p-type orbital on N• (see
magnified area in Figure 3a) and with the lone-pair orbital on
the N• (see magnified area in Figure 3b).
The trends of the O-H bond lengths in the TS structures
involving 1, 8, and 4 (see Figure 3 caption) follow that which
(40) The calculations involved the use of the LDBS approach in conjunction
with the B3LYP DFT method for structure optimizations. Additional details of
the LDBS partitioning are given in the Supporting Information. TSs were
optimized and validated by animating the single negative vibration mode that
connects reactants to products. Following the geometry optimizations, we
performed single-point energy calculations at the B3LYP/6-311++G(2d,2p) level
of theory. We used this approach to study HAT/PCET in some prototypical
systems in reference 12b. However, in the present work, B3LYP tends to give
Ea,1 values that are too large, likely because the TS structures are very sterically
crowded. We found better agreement between calculated and experimental Ea,1
values in this work by computing single-point energies with B971/6-
311++G(2d,2p). The better agreement is most like due to the fact that B971
can, to some extent, predict dispersion binding in some systems.41
The calculated Ea,1 values40 associated with the reactions of
1, 4, and 8 are in very good agreement with those obtained
from experiment (see Table 1), viz. calculated (experimental)
9276 J. Org. Chem. Vol. 73, No. 23, 2008

Reaction of Phenols with the 2,2-Diphenyl-1-picrylhydrazyl Radical
FIGURE 4. Calculated transition state structure for hydrogen atom exchange between 1,4-cyclohexadiene and dpph•. The view is in the C-N•-N
plane with one phenyl ring of the dpph• in the foreground on the right. The constituent atoms are colored as follows: oxygen ) red, nitrogen )
blue, carbon ) gray, hydrogen ) white. To help guide the eye, bonds between the O, H, and N atoms involved in the H transfer are drawn in. Key
structural information associated with the atoms at the atom transfer center are as follows: R(HC-H) ) 1.35 Å, R(H-N•) ) 1.36 Å, θ(HC-H-N•)
) 175.7°. The extent to which the transferring H atom deviates from coplanarity with the picryl C1 and the hydrazyl N atoms, ꢁ(H-C-N•-N),
is 22.8°. TS molecular orbitals, with relative phases indicated by the red and green colors: (a) Singly occupied molecular orbital (SOMO) showing
the nominally singly occupied p-type orbital on N•. The component of the orbital on dpph• is very similar to the SOMO on isolated dpph•; see
Figure 1b. (b) A low energy, doubly occupied molecular orbital (HOMO-8) showing the lone-pair orbital on N•.
Ea,1(1) ) 9.5 (9.8), Ea,1 (4) ) 4.7 (5.3), and Ea,1 (8) ) 8.3 (8.3)
kcal/mol.42 The calculated rate constants, k1, are also in
reasonably good agreement with measured values (see Table
1), viz. calculated (experimental) k1(1) ) 0.44 (0.10), k1(4) )
730 (238), and k1(8) ) 17.0 (1.4) M-1 s-1 at 298 K and maintain
the correct trend. The calculated Arrhenius pre-exponential
factors, A1(1) ) 25, A1(4) ) 30, and A1(8) ) 130 × 105 M-1
s-1 at 298 K are in less satisfactory agreement with the
experimental values of (16-19) × 105 M-1 s-1; see Table 1.
The calculated TS structure for the reaction of 1,4-cyclo-
hexadiene, 28, with dpph• is shown in Figure 4, and its structural
parameters are given in the figure caption. There are important
differences between the TS structure shown in Figure 4 and
the TS structures involving phenols. Specifically, the bond
lengths at the TS center are ca. 0.1 Å longer than those in
phenols, and the cyclohexadiene molecular plane is oriented
perpendicular to the direction of H atom transfer. In addition,
the out-of-plane angle, ꢁ(CH-C-N•-N), is 22.8°, much closer
to the “ideal” value of 0° for PCET39 than the analogous dihedral
angles in the phenols.
Figure 4 also displays the SOMO and a low-lying doubly
occupied molecular orbital (HOMO-8) for the TS structure of
the reaction of 28 with dpph•. Animations of the orbitals are
also provided in the SI. As is the case for the phenol + dpph•
TS, the orbitals show that the transferring H atom interacts with
both the singly occupied p-type orbital on N• (see magnified
area in Figure 4a) and with the lone-pair orbital on the N• (see
magnified area in Figure 4b).
(41) There are a large number of papers on this general subject. One of our
recent works in this area compares the binding in van der Waals complexes as
calculated by several DFT methods to that obtained by second-order Møller-
Plesset perturbation theory (MP2, a low-level, correlated wavefunction method).
See: Johnson, E. R.; DiLabio, G. A. Chem. Phys. Lett. 2006, 419, 333–339.
(42) Analogous data obtained using B3LYP/6-311++G(2d,2p) overestimate
Ea,1 by 1.8 to 2.3 kcal/mol but the trends of the calculated values nicely follow
that established by the measured Ea,1 and O-H BDE. The discrepancies between
the B3LYP-calculated and measured Ea,1’s are partially due to the fact that
B3LYP, as most DFT methods, overestimates steric repulsion.41
The kinetic parameters calculated for the reaction of 1,4-
cyclohexadiene with dpph• are in very poor agreement with
the measured values presented in Table 1. For example, Ea is
calculated to be 19.8 kcal/mol which is more than 8 kcal/mol
higher than the experimental value. A goodly portion of this
discrepancy can probably be attributed to the dispersion problem
J. Org. Chem. Vol. 73, No. 23, 2008 9277

Foti et al.
vary by about 5 orders of magnitude in response to variations
in Ea,1 as a consequence of the effects of the ring substituents.
Our DFT calculations indicate that the reaction of phenols
with dpph• must occur by one pathway and via a single TS
structure. The mechanism for these reactions cannot be described
as either PCET or HAT, but as some combination of the two.43
This is consistent with the results of our previous computational
studies on prototypical H-atom exchange reactions.12b The
geometries and the nature of the HOMOs and SOMOs of the
TS structure are not “ideal” for PCET or for HAT, see Figure
3. The phenolic proton may be transferred to the lone pair on
the N• of the dpph• radical with the accompanying electron
presumably coming from a lone pair on the phenolic oxygen
atom (i.e., PCET). However, the good orbital overlaps required
to allow facile electron transfer appear to be absent. Similarly,
the phenolic proton may also be transferred with one of its
bonding electrons to the unpaired electron on the N• of the dpph•
(i.e., HAT), but both the nuclear and electronic structures of
the TSs are far from ideal for this to be the dominant mechanism.
The A-factors for phenols are significantly lower than the
“normal” A-factor for a bimolecular H-atom transfer, evaluated
to be ∼108.5(0.5 M-1 s-1.8 Since the N• of dpph• is sterically
very well shielded, the entropy requirements in approaching the
TS structure are expected to be particularly unfavorable. This
must be the main cause of the observed low A-factors. The
A-factors for reaction of dpph• with the two 1,4-cyclohexa-
FIGURE 5. Plots of log k1/M-1 s-1 at 298 K vs Ea1 for nine phenols
with “free” OH groups (blue circles) and for ten phenols with an
intramolecular H-bonded OH group plus four 2,6-dimethylphenols (both
as red squares): the slopes of the best-fitted straight-lines were set equal
to their theoretical value of 1/2.303RT ) 0.733 mol/kcal while the
intercepts were left to float: log k1 ) 6.080 - 0.733Ea1 (r2 ) 0.986)
and log k1 ) 5.342 - 0.733Ea1 (r2 ) 0.975), respectively. The intercepts
yield, as the most representative A1 values for these two groups of
phenols, (1.2 ( 0.5) × 106 and (2.2 ( 0.8) × 105 M-1 s-1, respectively.
in DFT which is likely to be particularly bad because the “face-
on” orientation of the 1,4-cyclohexadiene (in contrast to the
“edge-on” orientation of phenols) will involve strong steric
repulsions with the ring moieties of dpph•.
dienes, 28 and 29, are 4.6 × 105 and 4.1 × 105 M-1 s-1
,
respectively (Table 1), which are even smaller than the A-factors
for phenols with “free” OH groups. For these two hydrocarbons,
no HB formation can occur (though there might be some weakly
held, dispersion-bound complexes or nonreactive charge-transfer
complexes). The small A-factors for 28 and 29 must be due,
primarily, to steric hindrance to reaction. In this connection,
the A-factors for H-atom transfers between two oxygen atoms,
reaction 5, rather than between oxygen and nitrogen, reaction
1,
Discussion
Kinetics and Mechanism of Reaction of Phenols with
the dpph• Radical. Table 1 shows that the A-factors for phenols
with both ortho-positions free are higher than the A-factors for
phenols having substituents at the ortho-positions. The lowest
value, 4.5 × 103 M-1 s-1, was measured for 2,4,6-tri-tert-
butylphenol, 27, and corresponds to an activation entropy of
-43.9 cal/mol·K. The highest value, 1.9 × 106 M-1 s-1, was
q
YO• + XOH f YOH + XO•
(5)
found with 4-methoxyphenol, 4, and corresponds to ∆S1
)
-31.9 cal/mol·K. The A-factors for phenols with “free” OH’s
6
are scattered around 1.2 × 10 M-1 s-1 (∆S1 ) -32.9 cal/
mol·K), whereas for intramolecularly hydrogen-bonded and 2,6-
dimethylphenols the A-factors are around 2.2 × 105 M-1 s-1
q
decrease dramatically with increased steric protection of the
reactive sites.45 For example, for YO• ) Me3COO•, A ) 107.2
M-1 s-1, Ea ) 5.2 kcal/mol for XOH ) phenol,46 but for
XOH ) 2,6-di-tert-butyl-4-methylphenol A is only 104.6 M-1
s-1 and Ea ) 0.8 kcal/mol.47 These results show that steric
hindrance to H-atom transfers can result in an unfavorable
activation entropy rather than an unfavorable activation
enthalpy (as is commonly supposed). There is no reason to
q
(∆S1 ) -36.1 cal/mol•K), see Figure 5. These two A-factor
domains, and the even smaller A1 factor for 27, demonstrate
that the activation entropy and, hence, the probability of H-atom
transfer, decreases substantially with increasing steric protection
of the phenol’s OH group. Indeed, the A1 factors are sufficiently
distinct between phenols with “free” OH’s and phenols with
an OH that forms an intramolecular HB or lies between two
methyl groups that the reactive OH group in polyphenols can
often be identified. For example, 11 and 12 have “high” A1
values, implying that their more reactive OH’s are those that
are not hydrogen bonded.
(43) Katarina44 has used a lower level of theory to deduce the mechanisms
of the reactions of a number of phenols with dpph•. The distinction between the
HAT and PCET processes was made on the basis of the difference in the
transferred H-atom’s charge in the TS relative to the starting phenol. A decrease
in this charge in the TS was taken to indicate that the HAT pathway was
dominant, while an increase in charge in the TS was taken to indicate a dominant
PCET mechanism. However, the mechanistic choices that were made on this
basis are difficult to accept. Phenols with quite similar structures, some with
electron-donating and others with electron-withdrawing substituents were put
in the HAT category while others were put in the PCET category. For example,
the HAT mechanism was deduced for phenol, 4-tert-butylphenol, 2-methox-
yphenol, and 4-dimethylaminophenol, but the PCET mechanism was proposed
for 4-methylphenol, 4-methoxyphenol, and 4-chlorophenol, see Table 3 in ref
44.
By using eqs II and III and the values of Ea,1 and A1 from
Table 1, it can be shown that the entropic contribution to the
free energy barrier to reaction is greater than the enthalpic
q
q
contribution, i.e., -T∆S1 > ∆H1 . Indeed, entropy can con-
tribute more than 70% to the free energy barrier. Nevertheless,
q
and despite the fact that ∆H1 accounts for only about 30% of
(44) Katarina, N. M. THEOCHEM 2007, 818, 141–150.
the barrier, Figure 5 demonstrates that, for a family of phenols
with no (or with rather similar) steric protection of their OH
groups, the rate constants for reaction 1 are under enthalpic
control. Indeed, for the phenols with “free” OH groups, k1 values
(45) For a summary, see: DiLabio, G. A.; Ingold, K. U. J. Am. Chem. Soc.
2005, 127, 6693–6699.
(46) Chenier, J. H. B.; Furimsky, E.; Howard, J. A. Can. J. Chem. 1974, 52,
3682–3688.
(47) Howard, J. A.; Furimsky, E. Can. J. Chem. 1973, 51, 3738–3745.
9278 J. Org. Chem. Vol. 73, No. 23, 2008

Reaction of Phenols with the 2,2-Diphenyl-1-picrylhydrazyl Radical
and N• atoms closer to one another than they would be in the
absence of an HB complex.
(iii) A smaller atomic radius for oxygen than for carbon.
Evans-Polanyi Relationships. For a series of closely related
reactions, the correlation, VIII, between Ea and reaction enthalpy
for exothermic free-radical reactions was first reported by Evans
and Polanyi48a and was further developed by Semenov.48b
Ea ) R∆H0 + ꢀ (0 < R < 1)
(VIII)
The value of R is generally considered to be an indication of
the position of the transition state along the reaction coordinate.49
For exoergic reactions, R < 0.5, and the transition state is
“reactant-like” or “early”. For endoergic reactions, 0.5 < R <
1, and the transition state is “product-like” or “late”. Reaction
1 is endothermic for the majority of the phenols studied in the
present work. If the Evans-Polanyi relation were to be accepted
for these endothermic reactions, eq VIII could be rewritten as
FIGURE 6. Plot of activation energy for reaction 1 vs ArO-H BDE
for those phenols shown in Chart 1 for which reliable BDEs were
available (see Table 1). Blue circles are for 16 of phenols 1-26. Phenols
15 and 23 were considered outliers and were excluded from the solid
correlation line: Ea,1 ) 0.918(O-H BDE) - 70.27 (r2 ) 0.95, eq X).
The red square is for 2,4,6-tritertbutylphenol, 27.
Ea,1 ) RBDE(ArO-H) + constant
(IX)
This is because ∆H0 ) BDE(ArO-H) - BDE(dpph-H) )
expect that the same would not hold true for the reaction of
1
dpph• with phenols.
BDE(ArO-H) -78.9 kcal/mol.
Among phenols 1-27 there are 16 for which reliable ArO-H
BDEs are available. Many are from experimental BDE mea-
surements, and the rest (i.e., 5, 6, and 26) were obtained using
one5a,22b of the numerous additivity rules that describe the effect
of substituents on ArO-H BDEs;5a,10,22,30,50,51 see Table 1. The
ArO-H BDE can also be calculated from the kinetics of reaction
1. For example, 3-methoxyphenol, 8, has k1) 1.4 M-1 s-1 and
k-1 ) 1.6 × 105 M-1 s-1 at 298 K (Table 1). Since the ∆S1 is
negligible5a and the BDE(dpph-H) ) 78.9 kcal/mol,5a the
ArO-H BDE for 8 ) 78.9 - 0.59ln(1.4/1.6 × 105) ) 85.7 kcal/
mol. This BDE is in excellent agreement with a value of 86.2
kcal/mol (1 M standard state) obtained from DFT calculations
on 8; see Table 1 and ref 52.
Kinetics
and
Mechanism
of
Reaction
of
1,4-Cyclohexadiene, 28, with the dpph• Radical. It is com-
monly assumed that H-atom transfers from carbon to an
oxidizing radical will occur purely by the HAT mechanism
because the carbon atom from which the H-atom is derived has
no lone pairs of electrons. This assumption has recently been
challenged by one of us.12b A DFT study of the toluene/benzyl
radical self-exchange reaction showed that π-π stacking
interactions permitted the PCET mechanism to operate, in
addition to HAT, in this C-H---•C f C•---H-C self-
exchange.12b A PCET mechanism also appears to be operating,
to some extent, in the reaction of 1,4-cyclohexadiene with the
dpph• radical. The transition-state structure for this reaction has
the C-H bond pointing roughly toward the lone pair on the N•
of the dpph• radical, see Figure 4, and to a lesser extent toward
the orbital containing the unpaired electron. Additional images
and animations of the TS SOMO orbital show significant π-π
stacking interactions between the incipient cyclohexadienyl
radical and a phenyl group on the dpph•. Thus, the strong
implication is that it is a proton that is transferred to the dpph•
which, in turn, implies that the accompanying electron is drawn
from the π-electron system of the hydrocarbon, rather than from
the C:H bond.
The available ArO-H BDEs range from 77 to ca. 86 kcal/
mol and are plotted against the measured activation energies
from Table 1 (range 2-10 kcal/mol) in Figure 6. Provided the
points for ubiquinol-0 (15), 2,2,5,7,8-pentamethyl-6-hydroxy-
chroman (23), and 2,4,6-tritert-butylphenol (27) are excluded,
(48) (a) Evans, M. G.; Polanyi, M. Trans. Faraday Soc. 1936, 32, 1333–
1360. (b) Semenov, N. Some Problems of Chemical Kinetics and ReactiVity;
Pergamon Press: New York, 1958; Vol. 1.
(49) Fersht, A. R. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 14338–14342.
(50) (a) Mulder, P.; Saastad, O. W.; Griller, D. J. Am. Chem. Soc. 1988,
110, 4090–4092. (b) Jonsson, M.; Lind, J.; Erikson, T. E.; Merenyi, G. J. Chem.
Soc., Perkin Trans. 2 1993, 1567–1568.
(51) Many of the original ArO-H bond enthalpies10, 14, 22 were determined
in benzene and contain an additional contribution from the enthalpy of the
intermolecular H-bond between the phenol and benzene. This quantity is only
∼1 kcal/mol22d for phenol itself and other phenols with “free” OH groups. It is
negligible for hindered and intramolecularly H-bonded phenols.22e Measured
ArO-H BDEs in the liquid-phase must also contain the enthalpy of solvation of
the H• atom, ∆Hsol(H•), which is ∼2.0 kcal/mol22d (almost independent of
solvent). However, only a limited number of liquid-phase ArO-H BDEs reported
in the literature contains this term.14,22e In those cases where H• atom solvation
was obviously contained in the published ArO-H BDE these values were
corrected for ∆Hsol(H•).
Again we attribute the small A-factors, mentioned above, for
the reaction of dpph• with 28 and 29 to steric protection of the
N• of dpph• which greatly reduces the probability of reaction
during radical/molecule encounters. It is noteworthy that the
two heavy atoms between which the H-atom (proton) is
transferred are separated in the TS by a much greater distance
during the reaction of dpph• with 1,4-cyclohexadiene, 28,
R(C-H-N•) ) 2.71 Å (see Figure 4) than they are during its
reaction with phenols, e.g., R(O-H-N•) ) 2.47 Å for 4-meth-
oxyphenol (see Figure 3b). We attribute these large differences
in heavy atom/heavy atom separation in the two TSs to the
following factors:
(i) Much larger steric repulsion in the cyclohexadiene reaction
because of its “face-on” approach to the lone pair on N•
compared with the phenol reactions with their less sterically
demanding “edge-on” approach to the lone pair on N•.
(ii) Formation of a pre-TS HB complex between the phenol’s
OH group and the lone pair on N• (Figure 2) will “draw” the O
(52) Foti, M. C.; Daquino, C.; DiLabio, G. A.; Ingold, K. U. J. Org. Chem.
2008, 73, 2408–2411. Foti, M. C.; Daquino, C.; DiLabio, G. A.; Ingold, K. U.
J. Org. Chem. 2008, 73, 7440.
(53) The ArO-H BDEs for 15 and 23 are known with considerable accuracy
(<(1 kcal/mol),14,22a-c which suggests that these phenols are outliers because
of “unusual” activation enthalpies. Indeed, the Ea,1 values for 15 and 23 are in
the range for diffusion in nonviscous solvents54 and have such small magnitudes
that significant errors may be associated with their determination over the
relatively small temperature range employed. Moreover, the reactions with dpph•
of these two phenols are probably slightly exothermic and it seems highly
improbable that exo- and endothermic reactions would fit the same Ea,1 vs ArO-H
BDE correlation (see text).
J. Org. Chem. Vol. 73, No. 23, 2008 9279

Foti et al.
as can easily be justified, 53-55 the plot of ArO-H BDE versus
Ea,1 gives a rather satisfactory linear relation for most of the
remaining phenols (see Figure 6), the equation being
Many previously unknown ArO-H BDEs have been deter-
mined by detailed kinetic studies of their reactions with dpph•
over a 65 °C range of temperatures in cyclohexane. The
activation energies of these reactions correlate linearly with the
heats of reaction, ∆H1, and with the ArO-H BDEs, as predicted
by the Evans-Polanyi relationship. However, the validity of this
relationship over a large range of ∆H, or ArO-H BDEs, is
questioned because the position and energy of the TS along
the reaction coordinate will vary with changes in ∆H. Despite
this, over a limited range of ∆H the Evans-Polanyi relationship
can still be useful.
Ea,1(kcal ⁄ mol) ) 0.918BDE(ArO-H) ⁄ kcal ⁄ mol -
70.27 (r2 ) 0.947) (X)
However, we consider it probable that there is not a truly
linear Evans-Polanyi relation for reaction 1, nor for any other
family of reactions, endothermic or exothermic, though there
may appear to be linearity if the range of Ea and ∆H0 are
small.49,57 This is because the Evans-Polanyi constant, R, will
tend to 1 for strongly endothermic reactions and will decline to
0 for exothermic reactions when Ea approaches Ediffusion. The
TS of reaction 1 moves along the reaction coordinate according
to the sign and value of ∆H1.58 Reaction 1 is endothermic for
most phenols in Chart 1 and a “product-like” TS is therefore to
be expected, in agreement with the Hammond postulate.58 The
small deuterium kinetic isotope effect, (see Table 2) is also
consistentwithaproduct-likeTS.59 Accordingly,theEvans-Polanyi
constant, R, is found to be relatively large, being ∼0.9 (see eq
X).
Despite our misgivings about the theoretical validity of the
Evans-Polanyi relation, Figure 6 leaves no doubt that for
phenols having ArO-H BDEs in the range 80-87 kcal/mol, their
BDEs can be correlated with the measured Ea,1 values via eq
X. We have therefore employed eq X to estimate all the
unknown ArO-H BDEs in the 1-26 set of phenols; see Table
1. We evaluate from Figure 2 that the level of accuracy for
these BDEs is probably better than (1 kcal/mol.
Experimental Section
Synthesis of 7 and 14. 2,4-Dimethoxybenzaldehyde (1 g, 6
mmol) and m-chloroperbenzoic acid (mCPBA) (1.56 g, 9 mmol)
were dissolved in 20 mL of anhydrous dichloromethane, and the
solution was refluxed for 16 h. The solvent was then removed and
the residue dissolved in ethyl acetate. The organic layer was washed
with aqueous NaHCO3 and with saturated brine and finally dried
over Na2SO4. Removal of the solvent afforded the crude formate
esters (as shown in the reaction below) of 7 and 14 which were
then hydrolyzed (under nitrogen) in methanol/10% aqueous KOH
(1:2 v/v) at room temperature for 1 h. The solution was then
acidified with HCl 1 N and extracted (3 × 20 mL) with ethyl
acetate. The organic layer was washed with water and saturated
brine and finally dried over Na2SO4. The solvent was removed to
give 880 mg of a crude residue which was purified by column
chromatography on silica gel with cyclohexane/ethyl acetate (90:
10 v/v) as eluent to give: 671 mg (72%) of 2,4-dimethoxyphenol
(7) as a yellow oil and 26.1 mg (3%) of 2,4-dimethoxy-5-
hydroxyphenol (14) as a white powder. Formation of minor
quantities of 14 and 2,6-dimethoxy-3-hydroxyphenol with the above
procedure is due to peracid oxidation of the aromatic ring.60 The
ESI-MS spectrum of 7, in the negative ion-mode, showed a peak
Conclusion
Theoretical calculations have revealed that the reaction of
phenols with dpph• occurs via a single pathway by a mechanism
that has both HAT and PCET character. These same calculations
reveal that the mechanism for the reaction of dpph• with 1,4-
cyclohexadiene also has a substantial degree of PCET character.
In this reaction, the mechanism can be partially described as a
process in which the C-H hydrogen is transferred as a proton
to dpph• with the accompanying electron coming from the
π-electron system of the diene. Thus, the PCET mechanism of
H-atom transfer is more widespread than has been generally
recognized (see also ref 12b).
1
at 153 m/z [M - H]-. H NMR of 7 (CDCl3): δ ) 3.77 (s, 3H),
3.87 (s, 3H), 5.26 (s, 1H), 6.40 (dd, J ) 8.6, 2.7 Hz, 1H), 6.50 (d,
J ) 2.7 Hz, 1H), 6.84 (d, J ) 8.6 Hz, 1H). 13C NMR (CDCl3): δ
) 55.6 (OCH3), 55.7 (OCH3), 3 × CH at 99.3, 104.2, 113.9; 3
quaternary C’s at 139.6, 146.9, 153.4. The EI-MS of 14 showed
1
peaks at m/z, 170 [M]+, 155 and 127. H NMR of 14 (CDCl3): δ
) 3.85 (s, 6H), 5.32 (s, 2H), 6.55 (s, 1H), 6.61 (s, 1H).
(54) The activation energy for diffusion in cyclohexane is estimated to be
approx. 2.6 kcal/mol This value was obtained from the self-diffusion coefficients
for liquid cyclohexane reported in: Jonas, J.; Hasha, D.; Huang, S. G. J. Phys.
Chem. 1980, 84, 109–112, using the equation D/T ) D0 exp(-ED/RT).
(55) For 2,4,6-tri-tert-butylphenol, 27, the ArO-H BDE is 80.1 kcal/mol22a
and Ea,1 ) 5.0 (( 0.5) kcal/mol (see Table 1), values that produce a point that
falls well above the solid straight-line in Figure 2. This “deviation” is not due
to some sterically-induced enhancement of Ea,1 because Ea,1 for 27 is equal, within
experimental error, to Ea,1 for 2,4,6-trimethylphenol, 24 (see Table 1), and 24
lies essentially on the solid line in Figure 6. The “deviation” of 27 arises because
steric repulsion between the OH group and the o-tert-butyl group to which it
points weakens its ArO-H BDE bond (compared with the OH bond in 24, for
which the ring substituents’ electronic effects must be similar, see Table 1) by
about 1.5 kcal/mol.56
(56) (a) Ingold, K. U. Can. J. Chem. 1960, 38, 1092–1098. (b) Ingold, K. U.;
Taylor, D. R. Can. J. Chem. 1961, 39, 471–480. (c) Ingold, K. U.; Taylor, D. R.
Can. J. Chem. 1961, 39, 481–487. (d) Ingold, K. U. Can. J. Chem. 1962, 40,
111–121.
(57) Cohen, A. O.; Marcus, R. A. J. Phys. Chem. 1968, 72, 4249–4256.
Marcus, R. A. J. Phys. Chem. 1968, 72, 891–899.
(58) Hammond, G. S. J. Am. Chem. Soc. 1955, 77, 334–338.
Synthesis of 12. Isovanillin (5 mmol, 760 mg) was added to a
mixture of CH2Cl2 (10 mL), H2O2 (30%) (11 mmol, 1.3 mL), and
SeO2 (0.4 mmol 44 mg). The reaction mixture was allowed to react
under stirring at room temperature for 16 h and then filtered. The
filtrate was added to 20 mL of CH2Cl2, the organic layer was
washed with 10 mL of 10% NaHSO3, followed by 10 mL of 10%
Na2CO3 and 10 mL of saturated brine, and was finally dried over
Na2SO4. The solvent was removed and the residue hydrolyzed
(59) DKIEs are largest for thermodynamically symmetrical TSs, (e.g., identity
reactions, X• + XH f XH + X•) and are smaller for reactions having either an
“early” or a “late” TS; see, e.g.: Pryor, W. A.; Kneipp, K. G. J. Am. Chem. Soc.
1971, 93, 5584–5586. Melander, L.; Saunders, W. H. Reaction Rates of Isotopic
Molecules; Wiley: New York, 1980; Chapter 5.
(60) Bjørsvik, H.-R.; Occhipinti, G.; Gambarotti, C.; Cerasino, L.; Jensen,
V. R. J. Org. Chem. 2005, 70, 7290–7296.
9280 J. Org. Chem. Vol. 73, No. 23, 2008

Reaction of Phenols with the 2,2-Diphenyl-1-picrylhydrazyl Radical
(under nitrogen) in a mixture of 10 mL of MeOH and 6 mL of
10% aqueous KOH at room temperature for 1 h. The solution was
then acidified with HCl 1 N and extracted with ethyl acetate. The
organic layer was washed with water and saturated brine and then
dried over Na2SO4. After solvent removal, the residue (710 mg)
was purified over silica gel using cyclohexane/ethyl acetate (80:20
v/v) as eluent to give 456 mg (60%) of 4-methoxyresorcinol (12)
as a yellow oil. The EI-MS showed peaks at m/z 140 [M]+, 125
1
and 97. H NMR (MeOD): δ 3.81 (s, 3H), 5.5 (br, 2H), 6.31 (dd,
J ) 8.6, 2.6 Hz, 1H), 6.50 (d, J ) 2.6 Hz, 1H), 6.69 (d, J ) 8.6
Hz, 1H). 13C NMR (CDCl3): OCH3 at δ 56.5; 3 × CH at 102.8,
105.8, 111.7; 3 quaternary C’s at 140.8, 146.2 and 150.2.
Synthesis of 13. Reduction of 2,5-dimethoxy-1,4-benzoquinone
with sodium dithionite (Na2S2O4) in methanol/water at room
temperature yielded phenol 13 in high yield. Briefly, approximately
150 mg of this benzoquinone was treated with a mix of ca. 10 mL
of methanol and ca. 20 mL of water. The solution was degassed
with nitrogen, and then an excess of Na2S2O4 was added under
stirring. After 15 min, another portion of Na2S2O4 was added and
the solution was left under bubbling nitrogen at room temperature
until it gradually became colorless after approximately 1 h. The
volume of the solution was reduced on a rotovap and then extracted
with CH2Cl2 (3 × 10 mL). The organic phase was washed with
water, dried on Na2SO4, and filtered and the solvent removed to
afford ca. 130 mg of 13 (yield ca. 87%) as a white solid which
slowly turned a pale pinkish. The solid was stored in the refrigerator
The final crude residue of reaction containing 18 was purified
by column chromatography on silica gel using cyclohexane/ethyl
acetate (90:10 v/v) as eluent to give 2,4-dimethoxy-6-methylphenol,
18, as a white powder in 46% final yield. The EI-MS showed peaks
1
at m/z 168 [M]+, 153, 125. H NMR (CDCl3): δ ) 2.26 (s, 3H),
3.76 (s, 3H), 3.86 (s, 3H), 5.30 (s, 1H, OH), 6.30 (d, J ) 2.7 Hz,
1H), 6.36 (d, J ) 2.7 Hz, 1H).13C NMR (CDCl3): δ ) 15.4 (CH3),
55.5 (OCH3), 55.8 (OCH3), 96.7 (CH), 106.6 (CH) and 4 quaternary
C’s at 123.6, 137.7, 146.5, 152.5.
Kinetics. Solutions of dpph• were prepared in cyclohexane or
n-hexane at a concentration of ca. (2 - 10) × 10-5 M by sonicating
(at room temperature) the mixture until all dpph• crystals were
dissolved. Phenols were dissolved in cyclohexane or n-hexane at a
final concentration of 1-5 mM. To facilitate the dissolution of the
most polar phenols, small quantities of CH2Cl2 or ethyl acetate
(<2% vol) were added. The reservoir syringes of the stopped-flow
spectrophotometer were filled with the solutions which were then
purged at room temperature with dry nitrogen for 5 min at a low
flow rate (the concentrations were corrected for solvent evapora-
tion). During the purge phase, the solutions were transferred back
and forth into the drive syringes to remove the oxygen adsorbed
into the syringes. Then, the drive-syringes were filled and the
solutions were allowed to stand for 15 min for thermal equilibration.
Each single rate constant at a given temperature was obtained by
averaging the values obtained from three to six decay traces at
512nm. The observed deviations from the average were smaller
than ( 10%. The explored temperature range was 5 to 70 °C and
phenol concentrations were corrected for solvent expansion using
an averaged cubic expansion coefficient of 1.2 × 10-3 °C-1 for
cyclohexane and of 1.45 × 10-3 °C-1 for n-hexane. The program
SPECFIT (version 3.0.34) was used to make the simulations (see
text) of a few experimental traces.
1
at - 20 °C until use. H NMR in DMSO-d6: δ 3.64 (6H, s); 6.42
(2H, s) and 8.24 (2H, br s). 13C NMR in DMSO-d6: OCH3 at δ
56.7; CH at δ 103.2; 2 quaternary Cs at δ 139.0 and 141.5. The
UV-vis spectrum in EtOH shows a max at 259 nm.
Syntheses of 15-17. Ubiquinol-0, 15, was obtained by reduction
of ubiquinone-0 with ascorbic acid following the procedure reported
in ref 61a.
The phenol (169 mg, 0.92 mmol) was then dissolved in
anhydrous acetone (10 mL), and the solution was treated with
Cs2CO3 (359 mg, 1.10 mmol), degassed with argon, and heated to
ca. 60 °C. After a few minutes, ca. 200 µL of CH3I was added
under stirring, and the solution was left to reflux for about 3 h
under argon. The color of the solution slowly changed from dark
green (after the addition of Cs2CO3) to yellow at the end of the
reaction. The solution was then acidified with 2 N HCl and extracted
(3 × 10 mL) with CH2Cl2. The organic phase was in turn extracted
(3 × 10 mL) with an aqueous solution of 2 N NaOH to separate
the dimethylated from the monomethylated ethers of ubiquinol-0.
The aqueous phase (containing the monomethylated ethers) was
acidified with 2 N HCl and re-extracted (3 × 10 mL) with CH2Cl2
to give ca. 70 mg of a crude residue which was purified on a
preparative TLC plate. This was initially eluted with cyclohexane/
acetone 90:10 v/v (to remove the last traces of the dimethylated
ether) and then with CH2Cl2/cyclohexane/acetone 70:27:3 v/v to
yield 39.4 mg (23%) of 16 and 2.4 mg (1.4%) of 17 as yellow oils.
The structure of the isomer 16 was confirmed by NOESY
experiments which showed a correlation between the ring proton
(6.44 ppm; see below) and one OCH3 (3.80 ppm; see below) and
the absence of any correlation between the ring methyl (2.21 ppm;
see below) and one OCH3. 1H NMR of 15 (CDCl3): δ 2.18 (3H, d,
J ) 0.6 Hz); 3.89 (3H, s) and 3.92 (3H, s); ca. 5.3 (2H, br); 6.49
The kinetics of reaction of the hydrocarbons 28 and 29 (TH2)
with dpph• were complicated by autocatalytic phenomena. Indeed,
the rate of reaction, -d[dpph•]/dt, with [TH2] . [dpph•] accelerated
as the dpph• radical was consumed.62 This autocatalysis cannot be
attributed to a double bond shift to form a more reactive 1,3-
cyclohexadienes since the rate constant for reaction of dpph• with
1,3-cyclohexadiene was found to be smaller than that for reaction
with 1,4-cyclohexadiene. We consider it most likely that the initial
TH• radical may self-quench with formation of a more reactive
TH-TH dimer and (TH)n oligomers (reaction of 1,3-cyclohexadiene
with dpph• gave a white polymer). Rate constants for the reaction
of 28 and 29 with dpph• were therefore determined from initial
rates.
1
(1H, br q). H NMR of 16 (CDCl3): δ 2.21 (3H, s); 3.80 (3H, s);
3.87 (3H, s) and 3.95 (3H, s); 6.44 (1H, s); ca. 5.4 (1H, br). 13C
NMR of 16 (CDCl3): CH3 at δ 15.3; 3 × OCH3 at 56.5, 60.72 and
61.0; CH at 109.5; 4 quaternary C’s at 117.8, 139.8, 141.0, 145.8.
1H NMR of 17 (CDCl3): δ 2.19 (3H, s); 3.78 (3H, s); 3.92 (3H, s);
3.93 (3H, s); 6.50 (1H, s). 13C NMR of 17 (CDCl3): CH3 at δ 15.4;
3 × OCH3 at 60.4, 60.5 and 61.0; CH at 110.5; quaternary C’s at
126.7, 131.1, 137.9, 144.7. The EI-MS spectra of 16 and 17 showed
the following main peaks at m/z 198 [M]+, 183, and 140.
Synthesis of 18. This phenol was prepared (46% yield) according
to the procedure reported in ref 61b and summarized in the
following reaction scheme.
Acknowledgment. G.A.D. thanks the Program for Energy
Research (PERD) for funding, the Centre for Excellence in
Integrated Nanotools, the Academic Information and Com-
munication Technologies (AICT), and Professor Pierre Bou-
langer (all of the University of Alberta) for access to computing
resources. M.C.F. thanks Dr. Paolo Bovicelli (CNR, Rome) for
J. Org. Chem. Vol. 73, No. 23, 2008 9281

Foti et al.
kindly providing samples of phenol 5. We are grateful to Annick
Le Quellec, Christian DiLabio, and Justin DiLabio for contribut-
ing to the animations provided in the Supporting Information.
Dr. Jon Johansson’s help in preparing the animations is also
appreciated.
hydrogen bonded complexes of 4-methoxyphenol with dpph•,
and additional experimental details. Animations showing more
detailed views of the orbitals illustrated in Figures 3 and 4 are
also provided. This material is available free of charge via the
Internet at http://pubs.acs.org.
JO8016555
Supporting Information Available: Optimized Cartesian
coordinates for dpph•, the hydrogen bonded structure involving
the N• center of dpph• and 4-methoxyphenol, and the structures
of hydrogen transfer transition states involving phenol, 3-meth-
oxyphenol and 4-methoxyphenol with dpph•, with the corre-
sponding locally dense basis set assignments, details of some
(61) (a) Foti, M.; Ingold, K. U; Lusztyk, J. J. Am. Chem. Soc. 1994, 116,
9440–9447. (b) Godfrey, I. M.; Sargent, M. V.; Elix, J. A. J. Chem. Soc., Perkin
Trans. 1 1974, 1353–1354.
(62) The decay traces at 512 nm were oriented with the convexity toward
the abs axis.
9282 J. Org. Chem. Vol. 73, No. 23, 2008