On the addition of .OH radicals to the ipso positions of alkyl-substituted aromatics: Production of 4-hydroxy-4-methyl-2,5-cyclohexadien-1-one in the radiolytic oxidation of p-cresol

Article abstract of DOI:10.1021/jp021807b

4-Hydroxy-4-methyl-2,5-cyclohexadien-1-one has been conclusively identified by its ¹H and ¹³C NMR spectra as a significant initial product in the radiolysis of aqueous solutions of p-cresol. This product is formed as the result of ox

Full text of DOI:10.1021/jp021807b

12178
J. Phys. Chem. A 2002, 106, 12178-12183
•
On the Addition of OH Radicals to the Ipso Positions of Alkyl-Substituted Aromatics:
Production of 4-Hydroxy-4-methyl-2,5-cyclohexadien-1-one in the Radiolytic Oxidation of
p-Cresol
Robert H. Schuler,*,† Guadalupe Albarran,†,‡ Jaroslav Zajicek,† M. V. George,†,§
Richard W. Fessenden,† and Ian Carmichael†
Radiation Laboratory and Department of Chemistry and Biochemistry, UniVersity of Notre Dame,
Notre Dame, Indiana 46556, Instituto de Ciencias Nucleares-UNAM, Circuito Exterior, C.U., Mexico, D.F.,
04510, and Photochemistry Research Unit, Regional Research Laboratory, TriVandrum, India 695 019
ReceiVed: August 6, 2002; In Final Form: October 11, 2002
4-Hydroxy-4-methyl-2,5-cyclohexadien-1-one has been conclusively identified by its 1H and 13C NMR spectra
as a significant initial product in the radiolysis of aqueous solutions of p-cresol. This product is formed as the
result of oxidation of the hydroxycyclohexadienyl radicals produced by addition of ∼12% of the •OH radicals
to the aromatic ring at the ipso position adjacent to the methyl group. It has a strong absorption band at 228
nm that is similar to the 246 nm band of p-benzoquinone. Its proton NMR spectrum exhibits strong coupling
between two pairs of ring protons. A quartet is observed in the spectrum of the methyl-13C labeled product,
confirming that a methyl group is attached to the dienone ring.13C chemical shifts and 13C-H spin-spin
splittings of the dienone are reported. Although DFT calculations of the proton NMR parameters are in very
good agreement with the experimental values, the calculated chemical shifts of the ring 13C carbons are 4-8
ppm too high, indicating that the DFT calculations do not properly take into account the dienone π system.
This conclusion is confirmed by parallel experimental and theoretical studies of 4H-pyran-4-one that provides
a model system closely related to the dienone. The observation of 4-hydroxy-4-methyl-2,5-cyclohexadien-
•
1-one as a product in the radiolysis of aqueous solutions is important in demonstrating that addition of OH
radicals to aromatic rings at positions substituted with alkyl groups can be of appreciable importance in the
•OH oxidation of aromatic substrates.
Introduction
ferred to as dienone). The observation of this product is of
general importance in demonstrating the importance of addition
of OH at ipso positions of alkyl-substituted aromatics.
In the following, the radiolytic preparation, spectral properties,
and theoretical studies of 4-hydoxy-4-methyl-2,5-cyclohexadien-
1-one and its methyl-13C labeled analogue are described in some
detail. Parallel experimental and theoretical NMR studies on
4H-pyran-4-one (OdCC4H4O), also reported below, provide
comparative data on a ring system closely related to the dienone.
•
Chromatographic and capillary electrophoretic studies of the
radiolytic oxidation of p-cresol at low doses show that four pro-
ducts are formed initially as the result of attack of •OH radicals
on the cresol.1 Of these, three are readily identified as 4-methyl
catechol, 4-methyl resorcinol, and p-hydroxybenzyl alcohol by
comparison of their absorption spectra and elution times with
reference samples. The first two are expected as the result of
oxidation of the hydroxycyclohexadienyl radicals initially pro-
•
Experimental Section
duced by addition of OH to the unsubstituted positions of the
cresol. The third indicates that abstraction of a methyl hydrogen
atom is also of some importance. However, in these studies the
identity of the fourth product, that has a strong absorption band
Radiolytic Preparation of the Dienone. Aqueous solutions
20 mM in p-cresol (Aldrich), also containing 20 mM potassium
ferricyanide and saturated with N2O (∼25 mM), were irradiated
to a dose of ∼1.5 megarads (15 kGy) in a 60Co source at a
dose rate of ∼1 megarad/h (10 kGy/h). The N2O served to
convert the hydrated electrons produced in water radiolysis to
•OH radicals and the ferricyanide to oxidize the hydroxycyclo-
hexadienyl radicals initially resulting from the addition of •OH
to the cresol. At this dose ∼50% of the cresol was consumed
and numerous secondary products were formed in addition to
the initial products. The irradiated samples were extracted four
times with an equal volume of diethyl ether and the extract
concentrated to a volume ∼1 mL. The dienone was then isolated
chromatographically (see below) by collecting samples from
individual injections of 100 µL. These samples were combined
and extracted with ether. After evaporation of the ether the
resultant dienone fraction was shown to be free of products other
1
at 228 nm, was not immediately evident. Its H and 13C NMR
spectra reported below conclusively show that this product is
4-hydroxy-4-methyl-2,5-cyclohexadien-1-one (subsequently re-
* Corresponding author. E-mail: schuler.1@nd.edu.
† University of Notre Dame.
‡ Instituto de Ciencias Nucleares-UNAM.
§ Photochemistry Research Unit, Trivandrum.
10.1021/jp021807b CCC: $22.00 © 2002 American Chemical Society
Published on Web 11/22/2002

Radiolytic Oxidation of p-Cresol
J. Phys. Chem. A, Vol. 106, No. 50, 2002 12179
than a small contribution of p-hydroxybenzyl alcohol. Irradiation
of 250 mL of the cresol solution yielded 2 mg of the purified
dienone. This sample, dissolved in 0.3 cm3 of Aldrich dichlo-
romethane-d2 (99.95% D), was used for the NMR analysis.
The dienone having the methyl group labeled with 13C was
prepared similarly from a 100 mg sample of 4-methylphenol-
13C (99% enrichment of CH3) obtained from Isotec Inc.
Although the amount of this labeled dienone was less than 1
mg, it was adequate for observation of the additional splittings
resulting from the methyl-13C substituent. It is pointed out that
the radiolytic approach used here made it possible to synthesize
the labeled dienone from the limited quantity of labeled cresol
available.
Chromatography. A Waters Millennium system with a 996
diode array detector was used to record liquid chromatographic
(LC) data in 3-D format. Product separation was with a 5µ
Phenomenex Luna C-8 column using gradient elution, initially
with 70:30 water-methyl alcohol, followed by a gradual
increase of the alcohol to 100%. The alcohol cleared the column
of the other products formed in the radiolysis. The dienone was
well separated from all radiolytic products except for p-
hydroxybenzyl alcohol that eluted immediately in front of the
dienone. Appropriate chromatograms and absorption spectra
were extracted from the recorded 3-D data. Chromatographic
sensitivities were determined from the areas of the recorded
peaks and the extinction coefficients determined spectrophoto-
metrically.
Electrophoretic Studies. Capillary electrophoretic (CE)
studies were with a Hewlett-Packard G1600 3-D CE system. A
micellar buffer (10 mM phosphate at pH 7 containing 50 mM
SDS) was used to optimize resolution. With this buffer the
dienone appears immediately after the water plug but is well
resolved from all other components. These electrophoretic
studies complemented the chromatography in tracking the
isolation of the dienone.
filling the F1 domain to 2048 and a Gaussian weighting function
were applied in both F1 and F2 domains prior to double Fourier
transformation.
The DEPT experiment allowed classification of the carbon
signals as representing methyl, methylene, methine, and qua-
ternary carbons. Specific assignments of the signals of carbons
with directly attached protons were made using HETCOR
spectra.3 The HETCOR spectrum was obtained using 4096
complex data points in the F2 domain and 128 t1 increments
(320 scans each), which were zero-filled to 1024 in the F1
domain. A sine-bell weighting function was applied in both
domains prior to double Fourier transformation.
Finally, the HMBC spectrum obtained using the pulse
sequence of Bax and Summers4 was used to assign quarternary
carbons and to check the correctness of the connectivities
established by the interpretation of the other spectra. Low-pass
J-filtering was used to suppress correlations due to one-bond
couplings. Delay ∆2 was set to 62.5 ms corresponding to the
average long range (through two or three bonds) carbon-proton
coupling constant of 8 Hz. In this experiment, 512 t1 increments
were sampled in 2048 complex data points using 196 scans for
each of the t1 increments. Data in the F1 domain were zero-
filled to 2048, and a Gaussian weighting function was applied
in both F1 and F2 domains prior to double Fourier transforma-
tion.
For comparison purposes the NMR spectra of 4H-pyran-4-
one, obtained from Aldrich, were recorded under conditions
similar to the above.
Computational Details. Geometries were computed by
density functional theory (DFT) using the popular B3LYP
functional5 and a modest polarized split-valence basis set (6-
31G*).6 Chemical shifts were estimated relative to TMS from
calculations at the B3LYP/6-311+G(2d,p) level,7 using gauge-
including atomic (London) orbitals.8 Computed shifts at the ring
carbons were further refined by comparison to calculations on
benzene and toluene. Optical absorption spectra were evaluated
from time-dependent DFT adding a set of diffuse functions to
the heavy-atom basis (6-31+G*). All of these calculations were
performed with the Gaussian 98 series of quantum chemistry
codes.9 Indirect NMR spin-spin coupling constants were
calculated using finite (Fermi-contact) field double perturbation
theory10 and a specifically designed augmented double-ú basis
([5s2p1d,3s1p])11 with a modified version of the Gaussian
suite.12
Spectrophotometry. Absorption spectra were recorded with
a Hewlett-Packard 4865 spectrophotometer. The dienone spec-
trum recorded spectrophotometrically was essentially identical
to those extracted from the LC and CE data. For reference
purposes spectra of known concentrations of 4-methyl catechol
and p-benzoquinone were also recorded. In the radiolytic
experiments the signal of the catechol provided an internal
reference against which to compare the dienone. The extinction
coefficient of the dienone determined spectrophotometrically
is 11 600 M-1 cm-1 at 228 nm. This value compares with values
of 2400 M-1 cm-1 for 4-methyl catechol at 280 nm and 17 300
M-1 cm-1 for p-benzoquinone at 246 nm. The initial dienone
yield1 was determined at low doses, taking the relative sensitivi-
ties as proportional to these extinction coefficients. Because only
2 mg of the purified dienone was available, its LC and CE
sensitivities are estimated to be accurate to only ∼10%.
Results
As indicated in the Introduction, the LC and CE studies have
shown that a product having a strong absorption at 228 nm is
formed in the radiolysis of aqueous solutions of p-cresol.1 The
intensity of this absorption increases with radiation dose in
parallel with the intensities of the products formed by addition
1
NMR Spectra. H and 13C NMR spectra were recorded on
•
of OH to the ortho and meta positions of the cresol so it is
a three-channel Varian UnityPlus spectrometer using a 3 mm
clear that this product is also produced as an initial product in
the reaction of OH with the cresol.
1
Nalorac dual 1H/13C probe: H NMR spectra (1H, DQF-COSY,
•
HMBC) at 599.98 MHz and 13C spectra (13C, DEPT, HETCOR,
HMBC) at 150.86 MHz. All experiments were run at ambient
temperature (20 °C) in deuterated methylene chloride. Chemical
shifts (δ) are reported in ppm relative to TMS as an internal
standard. Values of scalar spin-spin couplings (J, i.e., NMR
splittings) are reported in hertz.
Proton connectivities were determined by DQF-COSY ex-
periments.2 These experiments were run in the phase sensitive
mode. The 512 t1 increments of 88 scans each were sampled in
2048 complex data points for each of the increments. Zero-
Comparison in Figure 1 of this product’s absorption spectrum
with that of p-benzoquinone indicates that it has structural
features closely resembling those of the quinone.
Although this product was initially thought to be the quinone
methide (OdC6H4dCH2), it proved to be quite stable, suggest-
ing otherwise. The NMR studies described below conclusively
identify this product as having the structure indicated in the
Introduction, i.e., that it is 4-methyl-4-hydroxy-2,5-cyclohexa-
•
dien-1-one. Presumably, it is produced by addition of OH at
the methyl position of the aromatic ring followed by loss of

12180 J. Phys. Chem. A, Vol. 106, No. 50, 2002
Schuler et al.
Figure 1. Absorption spectra of 4-methyl-4-hydroxy-2,5-cyclohexa-
dien-1-one (solid spectrum) and of p-benzoquinone (dashed spectrum).
the phenolic proton on oxidation of the resultant hydroxycy-
clohexadienyl radical.
Figure 2. Proton NMR spectra of the dienone for the groups centered
at chemical shifts of 6.870 and 6.086 ppm: the experimental spectrum
is given in black, and the red spectrum is that simulated (see text) using
the parameters of Table 1.
HOC6H4CH3 + •OH -8 [HOC6H4 (OH)CH3]•
(I)
-e
1
TABLE 1: Dienone H NMR Parametersa,b
-H
[HOC6H4 (OH)CH3]•
+8 [HOC6H4 (OH) CH3]+ (II)
experimentala
calculateda,b
δ2,6
δ3,5
δCH
J2,6
J3,5
6.086
6.870
1.456
2.00
6.3
7.0
1.4
1.9
3.0
10.4
132
2.6
[HOC6H4 (OH) CH3]+ f OdC6H4(OH)CH3 (III)
3
The radiation chemical yield of this product, determined from
the initial slopes of its production in LC and CE experiments,1
is 0.7 molecules per 100 eV of absorbed energy, i.e., ∼12% of
3.09
J2,3 ) J5,6
1J(CmeHme
3J(CmeH3)
10.12
129.24c
2.17d
•
the OH radicals formed in the radiolysis of N2O saturated
) 3J(CmeH5)
solutions add at the methyl position of p-cresol. Clearly its
4J(CmeH2)
) 4J(CmeH6)
(-)0.76e
-0.9
production in significant yield demonstrates that, in general,
•
addition of OH to aromatic rings at positions substituted with
a Chemical shifts (δ) of the unlabeled dienone in ppm relative to
TMS; spin-spin coupling constants (J) are in hertz. b In the 13C-methyl
labeled dienone the spin-spin coupling constants are not observably
affected by the isotopic substitution, but the chemical shifts are
marginally affected (see text). c The one-bond spin-spin coupling
between 13C and the methyl protons in the 13C-methyl labeled dienone.
d The three-bond spin-spin coupling between 13C and the ring protons
at C3 and C5 in the 13C-methyl labeled dienone. e The four-bond spin-
spin coupling between 13C and the ring protons at C2 and C6 in the
13C-methyl labeled dienone.
aliphatic groups can be of importance.
NMR Spectra of the Dienone. The structure of the isolated
dienone was determined by analysis of one- and two-
dimensional 1H and 13C NMR spectra. The 1H NMR spectrum
showed a singlet at δ 1.456 that, from its intensity, is assignable
to the methyl protons (see also below) and to two multiplets at
δ 6.086 and 6.870. These multiplets are assignable to the two
pairs of ring protons. Although the latter patterns, illustrated in
Figure 2, are quite complex, they are virtually superimposeable,
so they clearly represent strong coupling among the ring protons.
The complexity noted in Figure 2 is similar to that in the case
of 1,1-difluorethylene where a less than obvious spectrum was
observed for both hydrogen and fluorine as the result of a
strongly coupled AA′XX′ spin system.13 The values of δ and J
for the ring protons, given in Table 1, were obtained by spin
simulation using the program PERCH.14 Figure 2 shows the
quality of agreement between the experimental spectra (black)
and the simulated one (red). Coupling constants are fitted to
better than 0.02 Hz and chemical shifts are accurate to 0.001
ppm.
TABLE 2: Dienone 13C NMR Parametersa
experimentala
calculated
δ1
185.65
127.57
152.54
67.62
191
135
160
73.2
27.4
δ2 ) δ6
δ3 ) δ5
δ4
δCH
27.16
3
a Chemical shifts (δ) relative to TMS in the unlabeled dienone in
ppm.
methine pairs at C2 and C6 and at C3 and C5. The HETCOR
spectrum also showed that the proton signal at 1.456 correlates
with the methyl carbon signal at δ 27.16, confirming its
assignment to the methyl protons.
The HMBC spectrum showed that the carbon signal at δ
152.54 correlates with the methyl proton signal, allowing its
assignment to the carbon pair at C3 and C5. The HMBC
spectrum also allows the carbon signal at δ 67.62 to be assigned
to the C4 carbon as it exhibits cross-peaks with the methyl
proton signal at δ 1.45 due to two-bond linkage and with the
proton signal at δ 6.086 due to three-bond linkage. Finally, the
The 13C DEPT experiment at a natural abundance level
revealed a total of five carbon resonance signals having the
chemical shifts reported in Table 2. From its intensity, the signal
at δ 27.16 can be assigned to the methyl carbon and is observed
as a quartet in the sample having the methyl carbon enriched
in 13C (see below). The unique signal at δ 185.65 is readily
assignable to the carbonyl carbon. The HETCOR spectrum
showed that the carbon signal at δ 127.57 correlates with the
proton multiplet at δ 6.087 and that at δ 152.54 with the
multiplet at δ 6.870. They are assignable, respectively, to the

Radiolytic Oxidation of p-Cresol
J. Phys. Chem. A, Vol. 106, No. 50, 2002 12181
TABLE 3: Dienone 13C-1H Coupling Constantsa
experimental
calculated
1J(C2H2)
1J(C3H3)
2J(C2H3)
4J(C2H5)
3J(C2H6)
2J(C3H2)
4J(C3H6)
3J(C3H5)
2J(C3Hme)
2J(C1H2)
3J(C1H3)
2J(C4H3)
3J(C4H2)
3J(C4Hme)
166.15
162.64
<1
171
167
0.2
-1.0
3.9
0.4
-0.9
5.9
4.0
1.4
10.1
2.0
9.1
<1
3.44
<1
<1
5.5
3.9
n.o.b
n.o.
n.o.
n.o.
n.o.
-2.9
a Spin-spin coupling constants observed in the spin uncoupled mode.
Ring positions of the C and H are given parenthetically. b no: not
observed. Signals of C1HX and C4HX were too weak to be observed
in the uncoupled mode.
Figure 3. Proton NMR spectra of the methyl-13C labeled dienone
illustrating the additional 13C splittings of 2.17 Hz for the group at δ
6.870 and 0.76 Hz for the group at δ 6.086. The red spectrum gives
the fit to the experimental spectrum in black.
HMBC spectrum shows that the carbon signal at δ 185.65
correlates with the proton multiplet at δ 6.870, confirming its
assignment to the carbonyl carbon at C1.
negligible effect on the splittings by the ring protons. Taking
into account the additional splittings indicated above, the spectral
simulations in Figure 3 gave values for the proton splittings
identical to those of the unlabeled compound. The optimized
spectra obtained with PERCH are superimposed on the experi-
mental spectra in Figure 3. In the optimization, changing the
sign of either or both of these spin-spin splittings had no effect.
Slight differences between the chemical shifts of the labeled
and unlabeled dienones are, however, noted. The chemical shift
of the protons at C3 and C5 is 0.0013 ppm (0.77 Hz) lower for
the labeled than for the unlabeled compound. This difference
is evident by comparing the upper spectra in Figures 2 and 3.
It may be caused by differing vibrational amplitudes in the 12C-
and 13C-containing compounds. An even smaller difference is
observed the patterns of the C2 and C6 protons, where the shift
of the labeled dienone is marginally (0.12 Hz) higher than that
of the unlabeled dienone.
Theoretical Studies. Initial theoretical attempts to identify
the carrier of the 228 nm absorption band were guided by its
similarity to the well-known benzoquinone band at 246 nm.
Time dependent DFT calculations on quinone methide, with a
structure optimized at the B3LYP/6-31G* level of theory, gave
an absorption maximum at 276 nm as compared to a calculated
maximum of 253 nm for p-benzoquinone. This comparison
essentially ruled out the methide as a candidate. Similar
calculations for the dienone, for which some significant
structural parameters are displayed in Figure 4, predict an
absorption maximum at 231 nm, in excellent agreement with
the experimental value (228 nm). Also the calculated oscillator
strength, 70% of that of the corresponding band in p-benzo-
quinone, is comparable to that determined from the ratio of the
extinction coefficients.
These various correlations permit unequivocal assignment of
the chemical shifts observed in the 13C spectrum to the five
different types of carbon atoms in the dienone. These assign-
ments are further substantiated by the DFT calculations reported
below.
In addition to the above, in uncoupled spectra, the 13C signals
of the two pairs of methine carbons are split by the protons
adjacent to the carbons. Additionally, the signals of the C2
carbons are observed as doublets because of couplings to the
protons at C6 and the signals of C3 appear as a doublet of
quartets corresponding to coupling to the proton at C5 and the
methyl protons. The other protons broaden these observed lines
somewhat but are not resolved. At natural abundance levels the
13C signals of the methyl carbon and those at C1 and C4 were
too weak in the uncoupled spectra to manifest additional
structure. The observed 13C-1H coupling constants are sum-
marized in Table 3.
NMR Spectra of the 13C-Methyl Labeled Dienone. The
product produced from a sample of p-cresol enriched with 13
C
at the methyl position exhibits a 13C 129.24 Hz quartet centered
at a chemical shift of δ 27.14. The observation of this quartet
additionally confirms the identity of the 13C line with a similar
chemical shift in the unlabeled dienone as representing the
methyl carbon. Each of the lines in the quartet also shows
unresolved structure resulting from coupling to the ring protons.
The observation of this quartet, which clearly demonstrates that
a methyl group is attached to the ring system, was crucial in
the initial identification of the product as the dienone.
In the case of the methyl-13C labeled dienone, the proton line
at δ 1.456 in the spectrum of the unlabeled dienone is observed
as a doublet split by 129.24 Hz. The agreement of this splitting
with the proton splitting of the 13C quartet, noted above, allows
unequivocal assignment of the line at δ 1.456 in the proton
spectrum to the methyl protons. In the labeled dienone its
chemical shift is slightly lower at δ 1.454.
The structure for the dienone given in Figure 4 shows a Cs
equilibrium structure with the OH bond located under the ring
(anti- to the exocyclic CC bond). A barrier of 14.5 kJ mol-1 is
predicted to inhibit free rotation of the OH. Local minima at
+11.1 kJ mol-1 exist when the OH bond is approximately anti-
to either CC ring bond. The calculated harmonic frequency for
In Figure 3 the proton spectrum of the 13C labeled sample
shows line groups corresponding to those in Figure 2 in which
each pattern has been additionally split, as indicated in Table
1, by 2.17 Hz for the group at δ 6.870 and by 0.76 Hz for the
group at δ 6.086. The experimental spectra accurately depict
the superposition of two patterns of Figure 2 separated by the
above splittings, indicating that the methyl-13C group has a
OH torsion in this potential well is 396 cm-1
.
The DFT calculations of the proton NMR parameters given
in Table 1 confirm the experimental assignments of the chemical
shifts and spin-spin coupling constants. Whereas the computed
shifts are slightly larger (∼0.2 ppm) than the experimental ones,

12182 J. Phys. Chem. A, Vol. 106, No. 50, 2002
Schuler et al.
TABLE 4: Comparative NMR Parameters for Dienone and
Pyranonea
dienone
pyranone
Proton Parameters
δ2
6.086
6.870
2.00
δ3
6.283
7.727
2.9
1.0
5.9
δ3
δ2
J(H2H6)
J(H3H5)
J(H2H5)
J(H3H5)
J(H2H6)
J(H2H5)
3.09
10.12
13C Parameters
δ1
185.65
127.57
152.54
166.15
162.64
<1
δ4
178.04
118.15
155.44
168.6
198.6
7.46
3.55
7.7
7.4
δ2
δ3
δ3
δ2
J(C2H2)
J(C3H3)
J(C2H3)
J(C2H6)
J(C3H2)
J(C3H5)
J(C3H3)
J(C2H2)
J(C3H2)
J(C3H5)
J(C2H3)
J(C2H6)
Figure 4. Computed (B3LYP/6-31G*) structural parameters.
3.44
<1
5.5
the indirect nuclear spin-spin coupling constants are in close
accord with those deduced from the analysis presented above.
Also, the DFT calculations for the methyl-13C labeled dienone
gave values for the couplings between the methyl carbon and
the methyl and ring protons, in good agreement with the
experimental values. We note here that the theoretical work
indicates that the sign of the spin-spin coupling between the
methyl carbon and the protons at C2 and C6 is negative, as
expected for four bond coupling.
The calculated values of the chemical shifts of the methyl
and ring 13C carbons are given in Table 2. Whereas the value
for the methyl carbon is very close to that observed, the
computed values for the ring carbons are in the range 5.5-7.5
ppm too high. This difference is discussed below. The calculated
13C-1H splittings given in Table 3 permit the specific assign-
ments indicated in the table. We note that these calculations
show three-bond C3Hme, C3H5 and C2H6 couplings to be
responsible for the additional structure observed experimentally,
that the two-bond C2H3 and C3H2 couplings are very small,
and that the four-bond C3H6 and C3H5 couplings are somewhat
larger than the two-bond couplings but negative.
a Because the ring carbons of dienone are referenced to the carbonyl
position and those of pyranone to the ring oxygen C2, C3, and C4 of
the dienone correspond to C3, C2, and C1 of the pyranone.
broad, as is frequently the case in solvents such as methylene
chloride. Wipf and Kim also examined the 13C spectrum at
natural abundance and reported five carbons with 13C chemical
shifts of δ 186.0. 153.1, 126.6, 66.9, and 26.7. Although these
shifts were unassigned, the values are very similar to those in
Table 2.
As mentioned above, attempts to directly compute the
chemical shifts produced values for the carbonyl and C2 and
C6 carbons 5.5 ppm and those for C3 and C5 carbon 7.5 ppm
higher than the experimental. To better calibrate this method
for chemical shifts, we have compared the DFT predictions with
experiment in a number of related systems. For example, the
calculated values are 5-9 ppm too high for the ring carbons of
benzoquinone, 2-cyclopenten-1-one (a model for the enone
motif), cyclopenta-1,3-diene, benzene, toluene, phenol, and
p-cresol. Similar discrepancies can therefore be expected for
other aromatic like systems.
We have also examined 4H-pyran-4-one (subsequently called
pyranone) as a model system more closely related to the present
case. In pyranone, oxygen replaces the dienone C4 carbon and
the conjugated structure of the dienone ring is preserved. Its
optical spectrum is similar to that of the dienone with, however,
a shift of the absorption maximum to 250 nm. Because pyranone
has only four protons and five ring carbons, its NMR spectra
are considerably simpler than those of the dienone. The various
experimental and theoretical parameters of its proton spectrum
are given in Table S1 of the Supporting Information and the
13C chemical shifts and 13C-1H splittings are given in Table
S2. Assignments, made here on the basis of comparison with
values from the DFT calculations, are discussed in the Sup-
porting Information.
Discussion
The NMR results reported here conclusively identify the
dienone as the product of the OH oxidation of p-cresol that
absorbs at 228 nm and allow specific assignments of the
observed chemical shifts and coupling constants. As indicated
above 12% of the OH radicals add to the ring at the methyl
•
•
•
position of the cresol. This yield indicates that attack of OH
on the ring at the methyl position is approximately statistical.
Wipf and Kim have previously prepared the dienone by
chemical oxidation of p-cresol.15 They reported that the dienone
was produced in a 61% yield. Such a high yield indicates that
chemical oxidation must involve a mechanism other than OH
radical attack on the cresol.
•
In their study Wipf and Kim examined the proton NMR
spectrum of this dienone in deuterated chloroform and reported
a singlet at δ 1.39 that they assigned to the methyl protons, in
agreement with the assignment given above for the line observed
in deuterated methylene chloride at δ 1.456 in the present study.
They also found lines at δ 5.98 and 6.83 that they interpreted
as doublets of doublets. Their analysis of these structures differs
from that given in Table 1 in indicating splittings of 8.4 and
1.8 Hz for the signals at δ 5.98 and 8.4 and 1.6 Hz for those at
δ 6.83. Wipf and Kim also observed a singlet at δ 3.95 that is
likely assignable to the hydroxyl proton. In our experiments
this latter signal was not observed, probably because it is very
Selected NMR values are compared with those at the
corresponding positions of the dienone in Table 4. Though the
overall patterns are similar, these comparisons show that the
ring oxygen has an appreciable effect on the proton parameters
related to its adjacent carbons. In pyranone, for example, there
is a large increase in δ2 and an appreciable decrease in the
coupling between adjacent ring protons (J2,3). These increases
manifest a significant effect of the ring oxygen on the electronic
distribution. This effect is also manifest in the 13C chemical
shifts and 13C-H splittings associated with the nuclei adjacent
to the oxygen. There are relatively small differences in the
chemical shifts of the carbonyl carbon and the adjacent carbons

Radiolytic Oxidation of p-Cresol
J. Phys. Chem. A, Vol. 106, No. 50, 2002 12183
that indicate that the effects of the ring oxygen are to a large
extent localized in the region of the oxygen.
Bangalore, India. This is contribution NDRL4396 from the
Notre Dame Radiation Laboratory.
In Table 2S we particularly note that the 13C chemical shifts
of the ring carbons are, as for the dienone, somewhat higher
than those predicted by the DFT calculations. Clearly, the DFT
calculations, which are optimized for alkyl-like carbons, do not
produce reliable values for the ring carbons. As mentioned
above, similar discrepancies are noted for other aromatic
systems. We stress here that TMS is an inappropriate compu-
tational reference for the DFT calculations of the chemical shifts
of ring carbons in aromatic like systems. Apparently, at the level
used, these calculations do not adequately take into account the
π electrons.
Supporting Information Available: NMR properties of 4H-
pyran-4-one as a model for the dienone system are described
in the Supporting Information. Included are Tables S1, listing
the proton chemical shifts and coupling constants, and Table
S2, the 13C chemical shifts and 13C-1H coupling constants. This
material is available free of charge via the Internet at http://
pubs.acs.org.
References and Notes
(1) Albarran, G.; Schuler, R. H. To be submitted for publication.
(2) Piatini, U.; Soerenson, O. W.; Ernst, R. R. J. Am. Chem. Soc. 1982,
104, 6800-6801.
(3) Bax, A.; Morris, G. A. J. Magn. Reson. 1981, 42, 501-905.
(4) Bax, A.; Summers, M. F. J. Am. Chem. Soc. 1986, 108, 2093-
2094.
(5) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(6) Hariharan, P. C.; Pople, J. A. Chem. Phys. Lett. 1972, 16, 217-
219.
(7) Cheeseman, J. R.; Trucks, G. W.; Keith, T. A.; Frisch, M. J. J.
Chem. Phys. 1996, 104, 5497-5509.
(8) Wolinski, K.; Hinton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112,
8251-8260.
(9) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. A.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A.
D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi,
M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;
Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,
D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.;
Ortiz, J. V.; Baboul, A. G.; Stefanov; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-
Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian98, revision A.7 ed.;
Gaussian, Inc: Pittsburgh, PA, 1998.
Summary
The present example represents a special case in that addition
•
of OH radical at the methyl position of the cresol results in a
•
readily observable product. Addition of OH radicals at ipso
positions of most other alkyl-substituted aromatics produces
intermediates that are not readily converted to products that
retain information pertaining to the site of addition. For example,
no product other than the expected cresols was found in a
detailed study of the radiolytic oxidation of toluene.17 There
currently is no quantitative information on the importance of
addition of •OH to alkyl-substituted positions of aromatics from
either time-resolved or product analysis studies. Attack of •OH
at positions of aromatics substituted with groups other than
alkyls clearly must also be of some importance, but such
processes are frequently masked by complicating secondary
reactions. For example, in the case of phenol the intermediate
•
radical produced by OH addition at the hydroxyl position
rapidly loses H2O to produce a phenoxyl radical. The observa-
tion that reaction I occurs to a significant extent is of particular
importance in radiolytic studies of aqueous solutions because
•OH radicals frequently dominate the oxidation processes. In a
more general sense similar processes must also be considered
in chemical studies where oxidation occurs via the reactions of
hydroxyl radicals.
(10) Kowalewski, J.; Laaksonen, A.; Roos, B.; Siegbahn, P. J. Chem.
Phys. 1979, 71, 2896-2902.
(11) Cloran, F.; Zhu, Y.; Osborn, J.; Carmichael, I.; Serianni, A. S. J.
Am. Chem. Soc. 2000, 122, 6435-6448.
(12) Carmichael, I. J. Phys. Chem. 1993, 97, 1789-1792.
(13) McConnell, H. M.; McLean, A. D.; Reilly, C. A. J. Chem. Phys.
1955, 23, 1152-1159.
(14) PERCH (v 1/2000), University of Kuopio, Finland.
(15) Wipf, P.; Kim, Y. J. Am. Chem. Soc. 1994, 116, 11678-11688.
(16) Previous reports indicate doublets-of-doublets in these regions;
cf.: Loots, M. J.; Weingarten, L. R.; Levin, R. H. J. Am. Chem. Soc. 1976,
98, 4571-4577.
Acknowledgment. Research described herein was supported
by the Office of Basic Energy Sciences of the U.S. Department
of Energy, by CONACyT-Mexico, grant No. 33752-E, and by
the Jawaharlal Nehru Centre for Advanced Scientific Research,
(17) Albarran, G.; Schuler, R. H. Radiat. Phys. Chem. 2002, 63, 661-
663.