Kinetic Study of the Phthalimide N-Oxyl Radical in Acetic Acid. Hydrogen Abstraction from Substituted Toluenes, Benzaldehydes, and Benzyl Alcohols

Article abstract of DOI:10.1021/jo0348017

The phthalimide N-oxyl (PINO) radical was generated by the oxidation of N-hydroxyphthalimide (NHPI) with Pb(OAc)₄ in acetic acid. The molar absorptivity of PINO^. is 1.36 × 10³ L mol ^-1 cm^-1 at λ_max 382 nm. The PINO radical decomposes slowly with a second-order rate constant of 0.6 ± 0.1 L mol^-1 s^-1 at 25°C. The reactions of PINO ^. with substituted toluenes, benzaldehydes, and benzyl alcohols were investigated under an argon atmosphere. The second-order rate constants were correlated by means of a Hammett analysis. The reactions with toluenes and benzyl alcohols have better correlations with σ⁺ (ρ = -1.3 and -0.41), and the reaction with benzaldehydes correlates better with σ (ρ = -0.91). The kinetic isotope effect was also studied and significantly large values of k_H/k_D were obtained: 25.0 (p-xylene), 27. 1 (toluene), 27.5 (benzaldehyde), and 16.9 (benzyl alcohol) at 25°C. From the Arrhenius plot for the reactions with p-xylene and p-xylene-d₁₀, the difference of the activation energies, E_a^D - E _a^H, was 12.6 ± 0.8 kJ mol^-1 and the ratio of preexponential factors, A^H/A^D, was 0.17 ± 0.05. These findings indicate that quantum mechanical tunneling plays an important role in these reactions.

Full text of DOI:10.1021/jo0348017

Kin etic Stu d y of th e P h th a lim id e N-Oxyl Ra d ica l in Acetic Acid .
Hyd r ogen Abstr a ction fr om Su bstitu ted Tolu en es, Ben za ld eh yd es,
a n d Ben zyl Alcoh ols
Nobuyoshi Koshino, Basudeb Saha, and J ames H. Espenson*
Department of Chemistry, Iowa State University of Science and Technology,
Ames, Iowa 50011
espenson@iastate.edu
Received J une 10, 2003
The phthalimide N-oxyl (PINO) radical was generated by the oxidation of N-hydroxyphthalimide
(NHPI) with Pb(OAc)4 in acetic acid. The molar absorptivity of PINO• is 1.36 × 103 L mol-1 cm-1
at λmax 382 nm. The PINO radical decomposes slowly with a second-order rate constant of 0.6 ( 0.1
L mol-1 s-1 at 25 °C. The reactions of PINO• with substituted toluenes, benzaldehydes, and benzyl
alcohols were investigated under an argon atmosphere. The second-order rate constants were
correlated by means of a Hammett analysis. The reactions with toluenes and benzyl alcohols have
better correlations with σ+ (F ) -1.3 and -0.41), and the reaction with benzaldehydes correlates
better with σ (F ) -0.91). The kinetic isotope effect was also studied and significantly large values
of kH/kD were obtained: 25.0 (p-xylene), 27.1 (toluene), 27.5 (benzaldehyde), and 16.9 (benzyl alcohol)
at 25 °C. From the Arrhenius plot for the reactions with p-xylene and p-xylene-d10, the difference
of the activation energies, Ea - EaH, was 12.6 ( 0.8 kJ mol-1 and the ratio of preexponential
D
factors, AH/AD, was 0.17 ( 0.05. These findings indicate that quantum mechanical tunneling plays
an important role in these reactions.
The autoxidation of hydrocarbons is a commercial
reaction that can be made to run economically.1 For
example, the aerobic oxidation of p-xylene produces
terephthalic acid which is in turn used to prepare
polyethylene terephthalate (PET). The industrial oxida-
tion processes mostly use a Co/Mn/Br combined catalyst
that is called the Amoco-MC catalyst.2 A major drawback
is the use of bromide ions that corrode expensive titanium
reactors and form a side product, CH3Br, that can
photocatalyze ozone loss. Ishii et al. have developed a new
catalyst for the oxidation of hydrocarbons using N-
hydroxyphthalimide (NHPI) and Co(OAc)2.3 The autoxi-
dation of p-xylene catalyzed by 10 mol % of NHPI with
Co(OAc)2 and Mn(OAc)2 gave 82% terephthalic acid in
HOAc at 100 °C.4
PINO radical in turn abstracts a hydrogen atom from
the hydrocarbon (RH) to produce a carbon-centered
radical (R•).5
The hydrogen abstraction mechanism is supported by
the kinetic isotope effect (KIE ) 3.8) for the catalytic
reaction of ethylbenzene, in which eq 1 is one component
step.6 The resulting carbon-centered radical (R•) reacts
with O2 to form a peroxyl radical ROO•, which is further
converted to an alcohol, aldehyde (or ketone), or carboxy-
lic acid. Despite its unique character, only one kinetic
study of PINO• has been reported by Masui et al., who
studied the kinetics in acetonitrile.7,8 The PINO radical
was generated electrolytically, and then its self-decom-
position and the reactivity of PINO• toward benzhydrol
and other substrates were investigated. A remarkably
large KIE (10.6) for the reaction of PINO• with benzhydrol
was found, which is larger than the theoretical maximum
of 7.8 based on the bond-stretching mode.9 Moreover, the
In the proposed scheme, Co(III) reacts with NHPI to
form the PINO radical, an oxidized form of NHPI. The
(5) Yoshino, Y.; Hayashi, Y.; Iwahama, T.; Sakaguchi, S.; Ishii, Y.
J . Org. Chem. 1997, 62, 6810-6813.
(1) Cornils, B.; Herrmann, W. A. Applied Homogeneous Catalysis
with Organometallic Compounds; VCH: Weinheim, 1996.
(2) Partenheimer, W. Catal. Today 1995, 23, 69-158.
(3) Ishii, Y.; Sakaguchi, S.; Iwahama, T. Adv. Synth. Catal. 2001,
343, 393-427.
(4) Tashiro, Y.; Iwahama, T.; Sakaguchi, S.; Ishii, Y. Adv. Synth.
Catal. 2001, 343, 220-225.
(6) Ishii, Y.; Iwahama, T.; Sakaguchi, S.; Nakayama, K.; Nishiyama,
Y. J . Org. Chem. 1996, 61, 4520-4526.
(7) Ueda, C.; Noyama, M.; Ohmori, H.; Masui, M. Chem. Pharm.
Bull. 1987, 35, 1372-1377.
(8) After this manuscript was submitted, Pedulli et al. reported their
work in ref 12.
10.1021/jo0348017 CCC: $25.00 © 2003 American Chemical Society
Published on Web 11/04/2003
9364
J . Org. Chem. 2003, 68, 9364-9370

Phthalamide N-Oxyl Radical in Acetic Acid
difference of the activation energies for Ph2CH(OH) and
Ph2CD(OH) was 7.9 kJ mol-1, which is beyond the
difference in zero-point energies of C-H and C-D (4.6
kJ mol-1).
On the other hand, Ingold et al. obtained a large KIE
(15) for the reaction of (CF3)2NO• with toluene,10 which
implies quantum mechanical tunneling might play a role
there. The (CF3)2NO radical is one of the most reactive
nitroxyl radicals, which is reflected by a fairly large O-H
bond dissociation energy (BDE) of (CF3)2NO-H, 346 kJ
mol-1. Recently, Minisci et al. proposed that the O-H
BDE of NHPI might be >360 kJ mol-1 from a study of
an equilibrium with phenoxyl radical.11,12 On that basis,
one can imagine that the reactivity of PINO• must be
higher than that of (CF3)2NO•. Thus the remarkable
behavior of PINO• has attracted our interest in the fields
of catalytic and free radical chemistry.
In the present work, we generated PINO• in acetic acid
by oxidation of NHPI with Pb(OAc)4.13,14 The PINO
radical is reasonably persistent in HOAc, which enables
the direct study of its kinetics with other substrates. We
investigated the reactions of PINO• with substituted
toluenes, benzaldehydes, and benzyl alcohols under an
argon atmosphere. We will discuss its reactivity in terms
of polar effects (Hammett analysis) and kinetic isotope
effects.
F IGURE 1. UV-vis spectra of 20 mmol L-1 of Pb(IV) with
different concentrations of NHPI in acetic acid: 0 (1), 0.36 (2),
0.44 (3), and 0.53 (4) mmol L-1 of NHPI. Inset: plot of molar
absorptivity of the PINO radical against wavelength.
with a temperature-controlled cell holder. After 15-20 min,
the Pb(IV) solution was injected into the cell. The absorbance
at 382 nm increased until all Pb(IV) had reacted with NHPI.
Finally, p-xylene or toluene solution was injected into the cell.
The kinetic traces could be fitted by the first-order equation
even under the lowest concentration at each temperature,
which means that the decomposition of PINO• is negligible
under the experimental condition. For reactions in acetonitrile,
a solution of NHPI (1.5 mmol L-1) was prepared in acetonitrile,
and then a small amount of a Pb(OAc)4 solution in acetic acid
was added to the acetonitrile solution (finally, the mixture
contained ca. 1 vol % HOAc). After the addition, the solution
was bubbled by Ar for 15 min, and then an oxygen-free
acetonitrile solution of p-xylene or p-xylene-d10 was injected.
P r od u ct in th e Rea ction of P INO Ra d ica l w ith p-
Xylen e. An acetic acid solution of Pb(OAc)4 (0.2 mmol) was
added slowly to an acetic acid solution that contained NHPI
(0.4 mmol) and p-xylene (2 mmol) under Ar. After the comple-
tion of the reaction, all the solvent was removed under reduced
pressure and the residue was dissolved in CDCl3. The 1H NMR
spectrum indicated that the major product is PINO-CH2C6H4-
CH3, along with minor amounts of other compounds. The
mixture was purified by column chromatography; a silica gel
column was prepared in CHCl3, which was also used as the
Exp er im en ta l Section
Ma ter ia ls. NHPI and lead tetraacetate were purchased
from Aldrich and used as received. Glacial acetic acid and
anhydrous acetonitrile were used as solvents. All the substi-
tuted toluenes, benzaldehydes (except 4-carboxybenzaldehyde),
and benzyl alcohols were obtained commercially and used
without further purification. 4-Carboxybenzaldehyde was
provided from BP Chemicals. For the study of kinetic isotope
effects, p-xylene-d10 (99+% D), toluene-d8 (99+% D), benz-
aldehyde-d (98% D), and benzyl alcohol-d7 (98% D) were
purchased from Aldrich and used as received.
Gen er a l P r oced u r e. PINO• was prepared in glacial acetic
acid by the oxidation of NHPI with lead tetraacetate. An acetic
acid solution that contained 0.15 mmol L-1 of Pb(IV) and 1.5
mmol L-1 of NHPI was prepared, and the solution was bubbled
with Ar for 15 min. The absorbance of the resulting solution
was ca. 0.3 at 382 nm, which indicated that about 0.2 mmol
L-1 of PINO• had been generated in the solution (ꢀ382 ) 1.36
× 103 L mol-1 cm-1; see below).15 An oxygen-free acetic acid
solution that contained the substituted toluene, benzaldehyde,
or benzyl alcohol was injected into the PINO• solution, and
then the absorbance change was monitored at 382 nm. The
organic substrates were always in large excess in this study,
and each kinetic trace conformed excellently to first-order
kinetics. Most of the rate constants were measured at 25.0 (
0.1 °C spectrophotometrically. To study the temperature
dependence of the rate constants, oxygen-free solutions of
NHPI, Pb(IV), and p-xylene or toluene were prepared. The
solution of NHPI in a cell was placed in the spectrophotometer
1
eluting solution. The H NMR spectrum in CDCl3 obtained by
a Varian VXR-300 NMR spectrometer is shown in Figure S1.16
Resu lts a n d Discu ssion
Mola r Absor p tivity a n d Sta bility of th e P INO
Ra d ica l in Acetic Acid . Before we studied the kinetics
of PINO•, the molar absorptivity of PINO• was deter-
mined in HOAc. At first, 20 mmol L-1 of Pb(IV) solution
was prepared, and then three aliquots of NHPI were
injected into a stoichiometric excess of Pb(IV). It was
observed that the absorbance around 380 nm increased
strongly. The spectra of the generated transients are
shown in Figure 1. Because these spectra contained
contributions from the PINO radical and the excess Pb-
(IV),17 the absorbance from the excess of Pb(IV) was
subtracted from each spectrum. The spectra so obtained
were then divided by the concentration of NHPI to obtain
the molar absorptivity of PINO•. These spectra then
(9) Espenson, J . H. Chemical Kinetics and Reaction Mechanisms,
2nd ed.; McGraw-Hill: New York, 1995.
(10) Malatesta, V.; Ingold, K. U. J . Am. Chem. Soc. 1981, 103, 3094-
3098.
(11) Minisci, F.; Punta, C.; Recupero, F.; Fontana, F.; Pedulli, G. F.
Chem. Commun. 2002, 688-689.
(12) Amorati, R.; Lucarini, M.; Mugnaini, V.; Pedulli, G. F.; Minisci,
F.; Recupero, F.; Fontana, F.; Astolfi, P.; Greci, L. J . Org. Chem. 2003,
68, 1747-1754.
(16) 1H NMR (300 MHz, CDCl3, δ): 2.36 (s, 3H), 5.18 (s, 2H), 7.19
(d, 2H), 7.43 (d, 2H), 7.74 (m, 2H), 7.80 (m, 2H).
(17) The changes in the spectra of NHPI and Pb(IV) can be
neglected, because they are excess components and the reduced Pb-
(II) is colorless.
(13) Lemaire, H.; Rassat, A. Tetrahedron Lett. 1964, 2245-2248.
(14) Ishii, Y.; Sakaguchi, S. Catal. Surv. J pn. 1999, 3, 27-35.
(15) The stoichiometry of the reaction is [Pb(IV)]:[NHPI] ) 1:2.
J . Org. Chem, Vol. 68, No. 24, 2003 9365

Koshino et al.
coincided with each other, as shown in the inset in Figure
1. The PINO radical has a maximum absorption at 382
nm with a molar absorptivity of 1.36 × 103 L mol-1 cm-1
.
The reported value in acetonitrile is 1.46 × 103 L mol-1
cm-1 at 380 nm,7 a small difference probably due to
solvent polarity.
It was found that the formation of PINO• was followed
by a slower decomposition, which we investigated in
HOAc. Masui et al. reported that the self-decomposition
of PINO• in acetonitrile obeys second-order kinetics.7
F IGURE 2. Dependence of kobs on the concentrations of
p-xylene or p-xylene-d10 in HOAc at 25 °C.
The second-order rate constant can be obtained by
using eq 3, where Y ) absorbance, as presented in the
Supporting Information.
SCHEME 1. Rea ction Sch em e for P INO Ra d ica l
w ith Hyd r oca r bon
Y0 - Y∞
Yt ) Y∞ +
(3)
1 + [PINO•]0kdt
[PINO•]0 is the initial concentration of PINO•, and eq 3
can be represented by using its molar absorptivity:
[PINO•]0 ) (Y0 - Y∞)/ꢀ382
,
Y0 - Y∞
1 + (Y0 - Y∞)kdt/ꢀ382
Yt ) Y∞ +
(4)
Rea ction of P INO Ra d ica l w ith p-Xylen e. The
reaction of PINO• with p-xylene was investigated under
the pseudo-first-order conditions, [PINO•]0 , [p-xylene]0.
Figure 2 shows the dependence of kobs on [p-xylene].
The second-order rate constant for p-xylene is 11.9 L
mol-1 s-1 at 25 °C. The initial step of the reaction
represents hydrogen atom transfer from a methyl group
of p-xylene to PINO•, which produces NHPI and a
benzylic radical (R•), as shown in Scheme 1. The following
step is a radical recombination between the benzyl radical
and another PINO• to form an adduct PINO-R.
The absorbance change at 382 nm was monitored, and
eq 4 was used for fitting (Figure S2). From several
experimental runs with different concentrations of NHPI
and Pb(IV), the second-order rate constant for the self-
decomposition of PINO•, kd, was obtained as 0.6 ( 0.1 L
mol-1 s-1 at 25 °C in HOAc. One can calculate the half-
life (t1/2) of PINO• by eq 5.
1
t1/2
)
(5)
There are several reports for radical recombination
reactions of TEMPO• with carbon-centered radicals (mostly
benzyl radical), and the rate constants of such reactions
[PINO•]0kd
For example, when [PINO•]0 ) 0.2 mmol L-1, the half-
life is 2.2 h, which is long enough to study the reactions
of PINO• with p-xylene and other hydrocarbons. There-
fore, 0.2 mmol L-1 of PINO• was used for the subsequent
kinetic studies. Because self-decomposition of PINO• is
unimportant with respect to its reactions with RH and
the experiments used, [RH]0 g 10[PINO•]0, the kinetic
data can be analyzed by pseudo-first-order kinetics: Yt
) Y∞ + (Y0 - Y∞) exp(-kobst), where kobs represents the
pseudo-first-order rate constant referred to subsequently.
Masui reported kd ) 24.1 L mol-1 s-1 at 25 °C for the
decomposition of PINO• in acetonitrile solution containing
0.1 mol L-1 of NaClO4 (as supporting electrolyte) and 5
mmol L-1 of pyridine (for electrochemical reversibility).7
We also monitored the decay of PINO in acetonitrile
containing 1% of acetic acid, and the decay was almost
the same as that in acetic acid, even with 0.1 mol L-1 of
NaClO4. On the other hand, addition of pyridine acceler-
ated the decay of PINO• in acetonitrile (Figure S3). The
effect of pyridine on the decomposition of PINO• lies
outside the scope of this subject, however.
are normally ca. 107 L mol-1 s-1 18-20
Moreover, a radical
.
recombination of TEMPO• with benzoyl radical takes
place very rapidly (1 × 109 L mol-1 s-1).21 As we know,
TEMPO• is a fairly stable radical, and its O-H BDE is
293 kJ mol-1, which is much lower than that of NHPI.11,22
These data indicate that radical recombination reactions
of PINO• with a carbon-centered radical must be faster
than those of TEMPO•. Therefore, it can be concluded
that the reactions of PINO• with hydrocarbons take place
through Scheme 1. We also considered the self-reaction
between two benzyl radicals. Although this reaction has
a nearly diffusion-controlled rate constant (ca. 109 L mol-1
(18) Bowry, V. W.; Ingold, K. U. J . Am. Chem. Soc. 1992, 114, 4992-
4996.
(19) Arends, I. W. C. E.; Mulder, P.; Clark, K. B.; Wayner, D. D. M.
J . Phys. Chem. 1995, 99, 8182-8189.
(20) Sobek, J .; Martschke, R.; Fischer, H. J . Am. Chem. Soc. 2001,
123, 2849-2857.
(21) Brown, C. E.; Neville, A. G.; Rayner, D. M.; Ingold, K. U.;
Lusztyk, J . Aust. J . Chem. 1995, 48, 363-379.
(22) Minisci, F.; Recupero, F.; Punta, C.; Gambarotti, C.; Antonietti,
F.; Fontana, F.; Pedulli, G. F. Chem. Commun. 2002, 2496-2497.
9366 J . Org. Chem., Vol. 68, No. 24, 2003

Phthalamide N-Oxyl Radical in Acetic Acid
TABLE 1. Secon d -Or d er Ra te Con sta n ts (L m ol-1 s-1
for th e Rea ction of th e P INO Ra d ica l w ith Su bstitu ted
Tolu en es, Ben za ld eh yd es, a n d Ben zyl Alcoh ols in HOAc
a t 25 °C
)
s-1),19 the concentration of benzyl radical at the steady
state is quite low in comparison with that of PINO•. In
that case, the reaction scheme can safely ignore radical
dimerization (eq 7). A kinetic simulation was performed
with the KinSim program to confirm this reaction pat-
tern.23 The kinetic simulation result is shown in Figure
S4.
no. of H
kPR
kPR (per H)
X-C6H4CH3
p-MeO
p-Me
(p-xylene-d10
m-Me
o-Me
p-H
(toluene-d8)
p-Br
p-COOH
m-COOH
o-COOH
p-NO2
3
6
6
6
6
3
3
3
3
3
3
3
10.3
5.95
0.240
3.06
2.99
0.620
0.0229
0.595
0.272
0.200
0.113
0.119
3.43
0.992
0.0400
0.509
0.498
0.207
0.00763
0.198
0.0907
0.0667
0.0375
0.0395
According to Scheme 1, the observed rate constant, kobs
,
)
corresponds to 2kPR[RH]. Therefore, the net second-order
rate constant, kPR, for the reaction of PINO• with p-xylene
can be obtained as 5.95 L mol-1 s-1 at 25.0 °C in HOAc.
kobs ) 2kPR[RH]
(8)
Following Scheme 1, an adduct of PINO• with a carbon-
centered radical is produced. This kind of adduct was
reported by Ishii; however, its NMR spectrum was not
reported.3 We confirmed the product in the reaction of
PINO• with p-xylene by 1H NMR spectroscopy. The major
p-X-C6H4CHO
MeO
Me
H
(benzaldehyde-d)
COOH
1
1
1
1
1
1
13.9
12.4
10.6
0.386
13.9
12.4
10.6
0.386
3.16
1.73
1
product was identified as PINO-CH2C6H4CH3. The H
NMR spectrum obtained is shown in Figure S1, which is
similar to that of N-benzyloxyphthalimide (PINO-
CH2C6H5).24
3.16
1.73
NO2
p-X-C6H4CH2OH
Figure 2 also shows a large KIE of 25.0 for the reaction
of PINO• with p-xylene, which supports the rate-
determining step being hydrogen abstraction. However,
the KIE is much larger than the maximum value for a
normal hydrogen abstraction on the basis of bond-
stretching modes and zero-point energies. The large KIE
implies quantum mechanical tunneling participates in
this reaction. We will discuss the tunnel effect later.
Rea ction s of P INO Ra d ica l w ith Su bstitu ted
Tolu en es, Ben za ld eh yd es, a n d Ben zyl Alcoh ols. The
rate constants for the reactions of PINO• with substituted
toluenes, benzaldehydes, and benzyl alcohols were mea-
sured in HOAc at 25 °C. The second-order rate constants
are listed in Table 1, and the rate constants for the
reactions with para- and meta-substituted substrates
were plotted against Hammett substituent parameters
(σ or σ+) as shown in Figure 3.25,26
MeO
Me
H
(benzyl alcohol-d7)
Br
Cl
CF3
NO2
2
2
2
2
2
2
2
2
44.9
22.4
13.7
5.65
0.335
8.03
6.75
4.93
5.53
27.4
11.3
0.670
16.1
13.5
9.85
11.1
resonance effect by the electron-donating group (correla-
tion coefficient, r ) 0.976 for σ+ vs 0.927 for σ). The
obtained F value for σ+ is -1.3, which is fairly close to
previously reported values for hydrogen abstraction from
substituted toluenes.29-31 One might speculate that elec-
tron-donating groups weaken the C-H bond of the
methyl group and thus cause the rate to increase.32-34
However, Ingold et al. calculated the bond dissociation
energies of substituted toluenes and concluded that any
substituents, even electron-withdrawing groups, in the
para position of toluenes make the C-H bonds weaker.35
Therefore, it can be concluded that the reactions of PINO•
with substituted toluenes have a contribution from the
polar effect, which supports the previously proposed
transition state.
These plots show that each reaction is accelerated by
electron-donating substituents, which suggests that the
reaction center is more positive at the transition state.
This result corresponds with the proposed transition
state for PINO• reactions.3,27
On the other hand, for the reaction of benzaldehydes,
the correlation with σ is better than that with σ+ (r )
0.992 for σ vs 0.932 for σ+). There are a few studies of
The F value of the reaction with substituted toluenes
is larger than those of the reactions with benzaldehydes
and benzyl alcohols.28 For the reaction with toluenes, the
correlation with σ+ values is more suitable than that with
σ values, which indicates a significant contribution of a
(28) In the reactions of p-tolualdehyde and p-methylbenzyl alcohol,
there are two reactive groups, -CH3 and -CHO or -CH2OH. The
obtained rate constants are fairly large for hydrogen abstraction from
the CH3 group. Therefore, the obtained rate constants must be for
hydrogen abstraction from the -CHO and -CH2OH groups.
(29) Pearson, R. E.; Martin, J . C. J . Am. Chem. Soc. 1963, 85, 354-
355.
(30) Walling, C.; Rieger, A. L.; Tanner, D. J . Am. Chem. Soc. 1963,
85, 3129-3134.
(23) Barshop, B. A.; Wrenn, R. F.; Frieden, C. Chemical Kinetic
Simulation Software Ver. 4.0; Washington University, 1997.
(24) Zhou, W.-J .; Hsieh, Y.-L.; Pratt, L. M.; Patten, T. E. Polym. Bull.
2001, 46, 43-50.
(31) Metelski, P. D.; Espenson, J . H. J . Phys. Chem. A 2001, 105,
5881-5884.
(32) Zavitsas, A. A.; Pinto, J . A. J . Am. Chem. Soc. 1972, 94, 7390-
7396.
(25) McDaniel, D. H.; Brown, H. C. J . Org. Chem. 1958, 23, 420-
427.
(33) Zavitsas, A. A.; Fogel, G.; Halwagi, K. E.; Legotte, P. A. D. J .
Am. Chem. Soc. 1983, 105, 6960-6962.
(34) Zavitsas, A. A. J . Phys. Chem. A. 2002, 106, 5041-5042.
(35) Pratt, D. A.; Wright, J . S.; Ingold, K. U. J . Am. Chem. Soc. 1999,
121, 4877-4882.
(26) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165-
195.
(27) Wentzel, B. B.; Donners, M. P. J .; Alsters, P. L.; Feiters, M. C.;
Nolte, R. J . M. Tetrahedron 2000, 56, 7797-7803.
J . Org. Chem, Vol. 68, No. 24, 2003 9367

Koshino et al.
F IGURE 3. Plots of the rate constants for the reactions of the PINO radical with substituted toluenes, benzaldehydes, and
benzyl alcohols against (a) σ constants and (b) σ+ constants in HOAc at 25 °C
hydrogen abstraction from benzaldehydes.36,37 We pro-
pose that σ values should be applied to the hydrogen
abstraction from benzaldehydes. This is because benzoyl
radicals are stabilized by a conjugation between an
unpaired electron on a carbon atom and a lone pair on
an oxygen atom, however, there is no contribution by
resonance with benzene ring.38
catalyzed oxidation of benzyl alcohols in acetonitrile.11
Even the oxidation of unsubstituted benzyl alcohol has
a remarkable selectivity. However, the rate constants for
benzyl alcohol and benzaldehyde in Table 1 show no
difference between the two reactions in HOAc. We
presume that the selectivity in oxidation of benzyl alcohol
can be attributed to a solvent effect and thus the
measured rate constants for hydrogen abstraction by
PINO• in acetonitrile (Table S1). The rate constants for
p-xylene and benzaldehyde in HOAc are larger than
those in acetonitrile probably due to thermodynamic
favorability in HOAc; the O-H BDE of NHPI in HOAc
could be higher than that in CH3CN because hydrogen
bonding between NHPI and HOAc increases the O-H
BDE.12,43 On the other hand, the reaction of PINO• with
benzyl alcohol in HOAc is slower than that in acetonitrile.
In this case, the O-H group in benzyl alcohol forms a
hydrogen bond with HOAc, which increases the C-H
BDE in benzyl alcohol. Presumably, selectivities for the
oxidation of benzyl alcohol by NHPI catalysis would vary
in different solvents.
Kin etic Isotop e a n d Tu n n el Effects. As shown in
Figure 2, a large kinetic isotope effect (KIE ) 25.0) was
observed in the reaction of PINO• with p-xylene in acetic
acid at 25 °C. This KIE value is much larger than that
reported (3.8) for ethylbenzene in the overall catalytic
system.6 We also studied the NHPI/Co(OAc)2-catalyzed
oxidation of p-xylene by monitoring O2 consumption to
determine the initial rate and found KIE ) 3.4 in HOAc
at 70 °C.44 Because of the multistep reaction for the
catalytic system, the observed KIE must be a mixed value
of several reaction steps.5,45 Therefore, we emphasize that
the KIE for the specific reaction of PINO• with p-xylene
is 25.0, which is far from the usual ones. Even if a
secondary isotope effect is considered, the KIE value is
still beyond the theoretical maximum. Moreover, the KIE
for the reaction of benzaldehyde, which does not have a
secondary isotope effect, is 27.5 (see below). Therefore,
the large KIE implies tunneling takes part in this
hydrogen abstraction reaction. We also investigated the
For the reaction with substituted benzyl alcohols, it is
hard to say which of the σ or the σ+ should be used
because of large experimental errors. However, it seems
that the correlation with σ+ is slightly better than that
with σ (r ) 0.897 for σ+ vs 0.790 for σ). If we select the
correlation with σ+ for the reaction, the F value is -0.41.
The kinetic studies of hydrogen abstraction from substi-
tuted benzyl alcohols are quite limited; however, there
are several studies of catalyzed oxidations of substituted
benzyl alcohols.39-41 The reported F values against σ or
σ+ lie between -0.32 and -0.58, even though the reaction
mechanisms are different. Moreover, in the reaction of
PINO• with substituted benzyl alcohols, probably the
obtained F has contributions not only from polar effect
but also from the differences in the C-H BDEs of benzyl
alcohols.35,42 One can notice from Figure 3 that the
reaction of PINO• with substituted benzyl alcohols is less
sensitive to its substituents than that of PINO• with
substituted benzaldehydes. Generally speaking, there-
fore, electron-withdrawing groups on benzyl alcohol lead
to higher selectivity in its oxidation to benzaldehyde.
Indeed, high selectivities were reported in the Co/NHPI-
(36) Lee, K. H. Tetrahedron 1968, 24, 4793-4803.
(37) Yoshida, T.; Namba, K. Kogyo Kayaku Kyokai-shi 1968, 29,
353-360.
(38) Solly, R. K.; Benson, S. W. J . Am. Chem. Soc. 1971, 93, 1592-
1595.
(39) Stavber, S.; Kosir, I.; Zupan, M. J . Org. Chem. 1997, 62, 4916-
4920.
(40) Fung, W.-H.; Yu, W.-Y.; Che, C.-M. J . Org. Chem. 1998, 63,
2873-2877.
(43) Koshino, N.; Cai, Y.; Espenson, J . H. J . Phys. Chem. A 2003,
107, 4262-4267.
(44) Saha, B.; Koshino, N.; Espenson, J . H. Manuscript to be
submitted.
(41) Dijksman, A.; Marino-Gonzalez, A.; Mairata i Payeras, A.;
Arends, I. W. C. E.; Sheldon, R. A. J . Am. Chem. Soc. 2001, 123, 6826-
6833.
(42) Wen, Z.; Li, Z.; Shang, Z.; Cheng, J .-P. J . Org. Chem. 2001, 66,
1466-1472.
(45) Zakharov, I. V.; Geletii, Y. V.; Adamyan, V. A. Kinet. Catal.
1991, 32, 31-36.
9368 J . Org. Chem., Vol. 68, No. 24, 2003

Phthalamide N-Oxyl Radical in Acetic Acid
Such low preexponential factors were also obtained for
the reaction with benzhydrol and benzhydrol-d in aceto-
nitrile by Masui (Figure S10).7 The low preexponential
factors might be attributed to the large tunnel effect,
since if tunneling participates in the reaction, the Ar-
rhenius plot becomes curved, and a limited temperature
range often leads low preexponential factors.47,48
Ingold et al. studied the reaction of (CF3)2NO• with
toluene and other hydrocarbons.10,51 They found a large
KIE in the reaction with toluene (kH/kD ) 15 at 25 °C),
and Ea - Ea was 6.7 kJ mol-1. These results indicate
that the tunnel effect takes part in the reactions of (CF3)2-
NO•. They also obtained low preexponential factors in the
reaction (log A ) 3.6-6.5).51 However, they claimed that
the low preexponential factors cannot be attributed to
the tunnel effect, and they mentioned an intermediate
formation of a hydrogen-bonded complex or the triplet
repulsion in the transition state.51 The existence of such
a hydrogen-bonded complex is supported by several
studies of (tert-Bu)2NO• or TEMPO• by spectroscopic data
and theoretical calculations.52-54
D
H
F IGURE 4. Arrhenius plot for the reaction of the PINO
radical with p-xylene and p-xylene-d10 in HOAc. T ) 17-65
°C.
same reaction in acetonitrile containing 1 vol % acetic
acid to evaluate the solvent effect on the KIE. The KIE
in acetonitrile is 28.4, as shown in Figure S5. Therefore,
it is noteworthy that the large KIE value of the PINO•
reaction does not originate from a solvent effect such as
hydrogen bonding. Sometimes, a large KIE value is
explained by a loss of one or more C-H bending modes
at the transition state.46 However, according to Bell’s
calculation, the maximum semiclassical isotope effect
after allowance for bending modes is 10 at 25 °C.47,48
Therefore, the large KIE value for the PINO• reactions
cannot be explained without allowance for the tunnel
effect. We also investigated the reaction of PINO• with
p-xylene and p-xylene-d10 at various temperatures. Fig-
ure 4 shows the Arrhenius plot for p-xylene and p-xylene-
It is often proposed that unusually low preexponential
factors might be attributed to the formation of a hydrogen-
bonded intermediate in hydrogen abstraction from
phenols.55-57
kb
9
X• + HR yka z X‚‚‚HR
8 XH + •R
(9)
k-a
kakb
rate )
[X•][RH]
(10)
k-a + kb
When k-a , kb, the rate-determining step is ka and the
Arrhenius preexponential factor will be high (≈1011
L
mol-1 s-1), which is a normal value for a diffusion-
controlled reaction in normal viscosity solvents. On the
other hand, if k-a . kb, the rate equation is obtained as
eq 11.
d
10. The difference of activation energies, EaD - EaH, was
12.6 ( 0.8 kJ mol-1, which is almost twice as large as
the difference in zero-point energies of C-H and C-D
bonds.47 Moreover, the ratio of preexponential factors, AH/
AD, was 0.17 ( 0.05 which is remarkably larger than
those of normal hydrogen abstraction reactions. Without
a tunnel effect, AH/AD is usually close to unity.47,49 The
kinetic isotope effects for the reactions of PINO• with
other substrates were also investigated, and the values
of KIE were 27.1 (toluene), 27.5 (benzaldehyde), and 16.9
(benzyl alcohol) in HOAc at 25 °C (Figures S7-S9). These
results further indicate that tunneling plays an impor-
tant role in the PINO• reactions.50
rate ) Kakb[X•][RH]
(11)
In such a case, the preexponential factor is expressed by
eq 12.
ln(Aobs/L mol-1 s-1) ) ln(Ab/s-1) + ∆S°a/R (12)
Thus, if the PINO• reaction involves the formation of a
hydrogen-bonded intermediate, the observed preexpo-
nential factor will be smaller by the entropy loss, ∆S°a/
R. If hydrogen and deuterium have different equilibrium
constants for the formation of a hydrogen-bonded inter-
mediate, one cannot say whether this reaction is really
The preexponential factors for the reactions of PINO•
with p-xylene and p-xylene-d10 are quite low (log A ≈ 6)
compared with those for other hydrogen abstraction
reactions.47 The temperature dependence for the reac-
tions of PINO• with toluene and toluene-d8 was also
investigated, and low preexponential factors were ob-
tained: AH ) (8.8 ( 0.9) × 105 L mol-1 s-1, AD ) (2.2 (
0.9) × 106 L mol-1 s-1 (Figure S8). We must ask why the
preexponential factors are so low in the PINO• reactions.
(51) Doba, T.; Ingold, K. U. J . Am. Chem. Soc. 1984, 106, 3958-
3963.
(52) Otsuka, T.; Motozaki, W.; Nishikawa, K.; Endo, K. J . Mol.
Struct. 2002, 615, 147-151.
(53) Franchi, P.; Lucarini, M.; Pedrielli, P.; Pedulli, G. F. Chem-
PhysChem. 2002, 3, 789-793.
(54) Shenderovich, I. G.; Kecki, Z.; Wawer, I.; Denisov, G. S.
Spectrosc. Lett. 1997, 30, 1515-1523.
(46) Howard, J . A. In Peroxyl Radicals; Alfassi, Z. B., Ed.; J ohn
Wiley & Sons: Chichester, 1997; p 283.
(47) Bell, R. P. The Tunnel Effect in Chemistry; Chapman and
Hall: London, 1980.
(48) Bell, R. P. Chem. Soc. Rev. 1974, 3, 513-544.
(49) Kwart, H. Acc. Chem. Res. 1982, 15, 401-408.
(50) The difference in KIE was discussed in ref 43.
(55) Mahoney, L. R.; DaRooge, M. A. J . Am. Chem. Soc. 1972, 94,
7002-7009.
(56) Chenier, J . H. B.; Furimsky, E.; Howard, J . A. Can. J . Chem.
1974, 52, 3682-3688.
(57) Foti, M.; Ingold, K. U.; Lusztyk, J . J . Am. Chem. Soc. 1994,
116, 9440.
J . Org. Chem, Vol. 68, No. 24, 2003 9369

Koshino et al.
tunneling or not. There is a report of a study of the
hydrogen-bonding interaction of C6H5OH and C6H5OD
with acetone and quinuclidine, where no detectable
difference is to be expected in the enthalpy of adduct
formation.58 If the hydrogen abstraction by PINO• really
involves such a hydrogen-bonded intermediate, the reac-
tion should be classified as a case of proton-coupled
electron transfer.59 Recently, quite large KIEs were
reported in proton-coupled electron transfers from nitro-
gen, sulfur, and phosphorus (KIE ) 31-178).60-62
Con clu sion s
Kinetic studies on the reactions of the PINO radical
with substituted toluenes, benzaldehydes, and benzyl
alcohols were performed under an argon atmosphere. The
reactions take place through hydrogen atom abstraction
from C-H bonds by the PINO radical. This step is rate
controlling and is followed by radical recombination of
the carbon-center radical so generated with another
PINO radical. From the analysis of the Hammett plot, it
was found that the reaction with substituted toluenes has
a large polar effect compared with the other reactions.
The detailed investigations of kinetic isotope effect
indicate quantum mechanical tunneling participates in
the PINO radical reactions. We expect that our findings
will lead to further understanding of the tunnel effect in
hydrogen atom transfer reaction.66 However, the reason
the preexponential factors in Arrhenius plots are so small
is less clear. Hydrogen abstraction by the PINO radical
might be an instance in which not only quantum me-
chanical tunneling but also the formation of a hydrogen-
bonded intermediate contribute.
Further kinetic studies in the presence of air or O2 are
necessary to understand the NHPI-catalyzed aerobic
catalytic system. To do so, precise information such as
the bond dissociation energy of NHPI will be helpful to
interpret the catalytic mechanism. From this work, it can
be said that the PINO radical is more reactive than other
nitroxyl radicals. The reactivities of nitroxyl radicals
must be related to bond dissociation energies of the
corresponding hydroxylamines. Direct evaluation of the
O-H bond dissociation energy of NHPI will help to define
the reactivity of PINO radical.43
On the other hand, the triplet repulsion is also a large
factor in hydrogen abstraction.51,63-65
X• + HR f [Xv‚‚‚HV‚‚‚Rv]q f XH + R•
In hydrogen abstraction, the transition state has parallel
spins, which leads to triplet repulsion. It is reported that
reaction of a nitroxyl radical with a hydrocarbon has a
large triplet repulsion energy (ET ) 28 kJ mol-1) but also
a large negative electron affinity energy (EEA ) -23 kJ
mol-1).65 It is not clear how much the triplet repulsion
energy or the electron affinity energy affect the pre-
exponential factor. However, it might be necessary to
take account of such energies because the electron affinity
of PINO• must be larger than those of usual nitroxyl
radicals such as TEMPO•.
In this article, we have discussed some possible reasons
why the preexponential factor of the PINO• reaction is
so small compared with other hydrogen abstraction
reactions. As Ingold proposed, the transition state in the
reaction of (CF3)2NO• as well as PINO• might have a
highly restrictive geometry that demands a very specific
orientation.51 More kinetic studies of PINO• with several
substrates such as phenols and hydroperoxides could be
helpful to understand the origin of the low preexponential
factor.
Ack n ow led gm en t. The authors are grateful for
support from the National Science Foundation (Grant
CHE-020409) and from BP Chemicals. We thank Dr.
Yu. V. Geletii, Prof. E. T. Denisov, and Dr. P. D.
Metelski for their helpful comments. This research was
in part carried out in the facilities of the Ames Labora-
tory of the U.S. Department of Energy, operated under
Contract W-7405-Eng-82 by Iowa State University of
Science and Technology.
(58) Boettcher, J . W.; Drago, R. S. J . Phys. Chem. 1974, 78, 429-
431.
(59) Mayer, J . M.; Hrovat, D. A.; Thomas, J . L.; Borden, W. T. J .
Am. Chem. Soc. 2002, 124, 11142-11147.
(60) Huynh, M. H. V.; Meyer, T. J .; White, P. S. J . Am. Chem. Soc.
1999, 121, 4530-4531.
Su p p or tin g In for m a tion Ava ila ble: The NMR spectrum
of PINO-CH2C6H4CH3, additional plots of kinetic data in
acetic acid and acetonitrile. This material is available free of
charge via the Internet at http://pubs.acs.org.
(61) Huynh, M.-H. V.; White, P. S.; Meyer, T. J . Angew. Chem., Int.
Ed. 2000, 39, 4101-4104.
(62) Huynh, M. H. V.; Meyer, T. J . Angew. Chem., Int. Ed. 2002,
41, 1395-1398.
J O0348017
(63) Mahoney, L. R.; DaRooge, M. A. J . Am. Chem. Soc. 1975, 97,
4722-4731.
(64) Denisov, E. T. Kinet. Catal. 1994, 35, 671-691.
(65) Denisov, E. T. Russ. Chem. Rev. 1997, 66, 859-876.
(66) McMahon, R. J . Science 2003, 299, 833-834 and references
therein.
9370 J . Org. Chem., Vol. 68, No. 24, 2003