4,6-Dimethyl-o-quinone methide and 4,6-dimethylbenzoxete

Article abstract of DOI:10.1021/jo981445x

4,6-Dimethyl-o-quinone methide (4) was produced by FVP of alcohol 3 or of the trimer 6 and matrix isolated in Ar at 7.6 K. Photolysis of 4 with long wavelength light (>345 nm) at this temperature afforded 4,6-dimethylbenzoxete (5), which was observable up to room temperature in the solid state in the absence of water. 5 can be converted back to 4 by UV irradiation at 254/190 nm. Quantum chemical calculations on the thermal interconversion of 4 and 5 indicate activation barriers of the order of 40 kcal/mol for 4 → 5, and 30 kcal/mol for 5 → 4. The dimer, trimer, and tetramer (8, 6, and 7) of 4 are characterized.

Full text of DOI:10.1021/jo981445x

9806
J . Org. Chem. 1998, 63, 9806-9811
4,6-Dim eth yl-o-qu in on e Meth id e a n d 4,6-Dim eth ylben zoxete
Greg GuangHua Qiao, Keith Lenghaus, and David H. Solomon*
Polymer Science Group, Department of Chemical Engineering, The University of Melbourne,
Parkville, Victoria 3052, Australia
Ales Reisinger, Ian Bytheway,1 and Curt Wentrup*
Department of Chemistry, The University of Queensland, Brisbane, Queensland 4067, Australia
Received J uly 23, 1998
4,6-Dimethyl-o-quinone methide (4) was produced by FVP of alcohol 3 or of the trimer 6 and matrix
isolated in Ar at 7.6 K. Photolysis of 4 with long wavelength light (>345 nm) at this temperature
afforded 4,6-dimethylbenzoxete (5), which was observable up to room temperature in the solid state
in the absence of water. 5 can be converted back to 4 by UV irradiation at 254/190 nm. Quantum
chemical calculations on the thermal interconversion of 4 and 5 indicate activation barriers of the
order of 40 kcal/mol for 4 f 5, and 30 kcal/mol for 5 f 4. The dimer, trimer, and tetramer (8, 6,
and 7) of 4 are characterized.
In tr od u ction
of o-quinone methide 4 and 4,6-dimethylbenzoxete (5),
the latter being observable till room temperature in the
solid state. Ab initio calculations of the spectra, energies,
and the barrier for interconversion of 4 and 5, are also
reported.
There has been a resurgence of interest in o-quinoid
chemistry, due to its important roles in organic and bio-
organic synthesis.2 Of particular interest is isomerization
to the four-membered ring compounds 2. Benzothietes
(2a )3 and benzazetines (2b)4,5 have been investigated in
considerable detail; until recently, little was known about
benzoxetes (2c). In 1993, Adam et al.6 reported the first
synthesis of 2H-benzoxetes carrying methyl and acyl
groups on position 2 of the four-membered ring, and
methoxy or allyl substitutes on the aromatic ring. Most
of these compounds are stable at -25 to -78 °C. More
recently, Tomioka et al.7 observed the parent benzoxete
(2c) and its interconversion with o-quinone methide 1c
in an Ar matrix at 10 K in the photolyses of benzofuran-
2-one and o-(diazomethyl)phenol. We have also generated
2c by flash vacuum pyrolysis (FVP) of benzofuran-2-one
as well as o-(hydroxymethyl)phenol, followed by photo-
chemical cyclization of the o-quinone methide 1.8 Our
warm-up experiments demonstrated that 2c is stable till
at least 155 K.8 We subsequently found that simple
methyl substitution on the benzene ring stabilizes the
benzoxete significantly. Here we report the generation
Resu lts a n d Discu ssion
1. Gen er a tion of Qu in on e Meth id e 4. Although
other compounds, such as (o-hydroxyphenyl)carbene,7
have been used as the precursors of benzoxetes, the
quinone methides proper are the most convenient, as they
are easily generated from a number of precursors and
can be isomerized to benzoxetes through photochemical
four-electron electrocyclization. Therefore, 2-(hydroxym-
ethyl)-4,6-xylenol (3) was subjected to an FVP experi-
ment9 to generate the quinone methide 4. Xylenol 3 was
sublimed at ca. 45 °C in high vacuum (∼4 × 10-6 mbar).
The vapor of sublimed 3 was mixed with argon as a
carrier gas and passed through a pyrolysis tube. The
pyrolysate was immediately condensed on a KBr, BaF2,
or CsI target as an argon matrix at 7-12 K for IR
spectroscopy. At a pyrolysis temperature of 500 °C, yellow
4,6-dimethyl-o-quinone methide (4) was observed in the
matrix together with unchanged xylenol 3 (Scheme 1).
Above 650 °C, no starting material 3 survived, and only
the yellow quinone methide 4 and the eliminated H2O
were trapped on the target. The observed IR spectrum
of 4 generated by pyrolysis of xylenol 3 at 600 °C and
isolated in an Ar matrix (Figure S1 in the Supporting
Information: identical with Figure 1b, where 4 is gener-
ated from precursor 6, as described below), was compared
with the theoretical one obtained by a B3LYP/6-31G*
calculation (Figure 1a). The good agreement between the
experimental and the theoretical spectra supports the
observation of quinone methide 4 in this matrix. Fur-
(1) Current address: Department of Chemistry, University of
Western Australia, Nedlands WA 6907 Australia.
(2) Recent review: Wan, P.; Barker, B.; Diao, L.; Fischer, M.; Shi,
Y.; Yang, C. Can. J . Chem. 1996, 74, 465, and refs therein. Earlier
reviews: Densimoni, G.; Tacconi, G. Chem. Rev. 1975, 75, 651. Turner,
A. B. Q. Rev. 1964, 28, 347.
(3) Meier, H.; Gro¨schl, D. Tetrahedron Lett. 1995, 36, 6047. Schmidt,
M. Meier, H.; Niedermann, H.-P.; Mengl, R. Chem. Ber. 1990, 123,
1143. Meier, H.; Eckes, H.-L.; Niedermann, H.-P.; Kolshorn, H. Angew.
Chem., Int. Ed. Engl. 1987, 26, 1046. Kanakarajan, K.; Meier, H. J .
Org. Chem. 1983, 48, 881. Meier, H. J . Prakt. Chem. 1997, 338, 383.
Meier, H.; Mayer, A.; Groschl, D. Sulfur Rep. 1994, 16, 23. Wentrup,
C.; Bender, H.; Gross, G. J . Org. Chem. 1987, 52, 3838.
(4) Pfister-Guillouzo, G.; Gracion, F.; Senio, A.; Lettulle, M.; Ripoll,
J .-L. Tetrahedron Lett. 1992, 33, 5753. Smolinski, G. J . Org. Chem.
1961, 26, 4108. Burgess, E. M.; McCullagh, L. J . Am. Chem. Soc. 1966,
88, 1580. Murata, S.; Sugawara, T.; Iwamura, H. J . Am. Chem. Soc.
1985, 107, 6317.
(5) Letulle, M.; Guenot, P.; Ripoll, J . L. Tetrahedron Lett. 1991, 32,
2013.
(6) Adam, W.; Hadjiarapoglou, L.; Peter, K.; Sauter, M. J . Am. Chem.
Soc. 1993, 115, 8603. Adam, W.; Sauter, M.; Zu¨nkler, C. Chem. Ber.
1994, 127, 1115. Sauter, M.; Adam, W. Acc. Chemn. Res. 1995, 28,
289.
10.1021/jo981445x CCC: $15.00 © 1998 American Chemical Society
Published on Web 12/08/1998

4,6-Dimethyl-o-quinone Methide and 4,6-Dimethylbenzoxete
J . Org. Chem., Vol. 63, No. 26, 1998 9807
The same trimer and tetramer were also observed by
IR spectroscopy in a warm-up experiment of quinone
methide 4 isolated neat at 7.6 K on a KBr target. In this
experiment, the FVP was conducted at 700 °C without
Ar as a carrier gas. A very slow sublimation rate of the
precursor 3 was used, since this permits complete con-
version of 3 as demonstrated by the IR spectrum. After
deposition, the target was slowly warmed room temper-
ature. IR spectroscopy revealed that the quinone methide
4 had disappeared and precursor 3 had regenerated on
the target (reaction starting above -90 °C; see below).
Because the water eliminated from 3 was cocondensed
on the target, the regeneration of 3 was simply due to
reaction of 4 with H2O. Other major peaks appeared at
1732 and 1696 cm-1 and correspond to the main bands
of trimer 6 (1695 cm-1) and tetramer 7 (1695 and 1731
cm-1). The material on the target was dissolved in
chloroform and examined by GC-MS. Three peaks were
observed; peak 1 appeared at a retention time of 6.5 min,
contained 94.6% of the mixture according to integration
(uncorrected), and had a mass of 152 amu and a mass
spectrum identical with that of the precursor 3. Peak 2
appeared at 10.9 min, corresponded to ca. 3% of the
mixture, and had a mass of 402 amu. Based on a
comparison with the mass spectrum obtained in the
preparative work, peak 2 is due to trimer 6.
The third peak appeared at 12.4 min, represented ca.
2.5% of the mixture, and had a mass of 270 amu. This is
different from the tetramer 7 observed by IR spectroscopy
at room temperature in the warm-up experiment. As the
GC-MS sample was injected through a 200 °C injection
port, it seems likely that tetramer 7 decomposed to
another compound with a mass of 270 amu. This was
confirmed by injection of pure 7 into the GC-MS, when
the only resulting peak had a mass of 270 and the same
retention time as peak 3 above. In the mass spectrum of
this compound, the base peak is at m/z 135, which
suggests that the m/z 270 compound is a symmetrical
molecule. Thus, it is proposed that this compound is bis-
(2-hydroxy-3,5-dimethylphenyl)ethane (9) (Scheme 2).
Compound 9 was also synthesized by condensed phase
thermolysis of either trimer 6 or tetramer 7 at 180 °C
(see Experimental Section). Compound 9 is formally a
product of reductive dimerization of 4. We do not know
the mechanism of this reaction and can only speculate.
When the tetramer 7 is used as the precursor, two
molecules of dimer 8 should form. Cleavage of the spiro-
C-O bond in 8 leads to a diradical analogue of 9, which
can furnish alcohol 9 by hydrogen abstraction. When the
trimer 6 is the precursor, a mixture of monomer 4 and
dimer 8 should form, the latter reacting as before. The
monomer 4 may dimerize to the same diradical in the
condensed phase thermolysis. Surprisingly, the yield of
9 increased to 100% when the glassware was first washed
with 1,1,1,3,3,3-hexamethyldisilazane. Photoacoustic IR
measurements demonstrated that silazane compounds
remain on the glass surface even after baking at 100° C.
A quantitative yield of compound 9 was obtained when
trimer 6 was thermolyzed in the presence of water.
The mass spectrum of trimer 6 indicates that its
molecular ion can easily fragment into dimer and mono-
mer. This suggests that trimer 6 might be a better
precursor of quinone methide 4, as no byproducts would
be formed, thus allowing an investigation of the inter-
conversions and stabilities of quinone methide and ben-
zoxete. In fact, this turned out to be the case.
F igu r e 1. IR difference spectrum of 4,6-dimethylbenzoxete
(5) (c, positive peaks) at 7.6 K in an Ar matrix, generated by
photolysis of quinone methide 4 (b, negative peaks) (obtained
by FVP of trimer 6) with light of λ > 345 nm for 30 min; a:
B3LYP/6-31G*-calculated IR spectrum of 4; d: B3LYP/6-31G*-
calculated IR spectrum of 5.
Sch em e 1
thermore, the main bands of 4 at 1668, 1642/1637, and
1569 cm-1, due to CdO and CdCH2 stretching, are
similar to those in the parent compound 1c in Ar matrix
(1668, 1616, and 1568 cm-1).7,8,10
Next, FVP was performed on a preparative scale,
whereby the thermolysate was isolated in a U-tube at
77 K. The use of a U-tube,11 rather than a coldfinger,
avoids regeneration of the starting material by reaction
of quinone methide 4 with eliminated H2O. At a pyrolysis
temperature of 800 °C, all starting material had reacted,
and a mixture of trimer 6 and tetramer 7 of the quinone
methide 4 was isolated from the U-tube in a nearly 1:1
molar ratio and in 98% absolute yield (Scheme 2). These
compounds were fully characterized spectroscopically (see
Experimental Section).

9808 J . Org. Chem., Vol. 63, No. 26, 1998
Qiao et al.
Sch em e 2
Pyrolysis of the trimer 6 was carried out at 850 °C.
The trimer was sublimed at ca. 105 °C with argon as a
carrier gas. Under these conditions, pure quinone me-
thide 4 was matrix isolated on a 7.6 K KBr target as
evidenced by the IR spectrum (see Supporting Informa-
tion). In a similar experiment, the pyrolysis of trimer 6
was carried out without argon. When the neat quinone
methide 4 was warmed above -92 °C, new IR bands
appeared, and the absorptions due to the quinone me-
thide decreased. These newly formed bands became much
stronger when the target was warmed further to -65 °C,
and they are attributed to the formation of dimer 8 and
trimer 6 by comparison with the IR spectrum of the
trimer obtained in the preparative FVP work. Moreover,
a low-temperature NMR experiment revealed the exist-
ence of dimer 8 (see below). After the target was warmed
to room temperature, additional bands due to tetramer
appeared. It seems obvious that dimerization would be
the first step of reaction of quinone methide 4. As this
dimer was never isolated from any pyrolysis products,
we concur with Ripoll5 that it is not stable at room
temperature. Therefore, a low-temperature NMR experi-
ment was carried out. Xylenol 3 was pyrolyzed at 650
°C, and quinone methide 4 was isolated on a coldfinger
at 77 K. After pyrolysis, the system was isolated from
the vacuum pumps, and CDCl3 was injected and con-
densed on the coldfinger. On warm up, the chloroform
melted and flowed with the pyrolysate into a round-
bottomed flask cooled to 77 K in liquid N2. The solution
was then transferred to a NMR tube and kept at 77 K
until the NMR measurement. The NMR spectrum of the
solution was first measured at -30 °C, revealing a
mixture of dimer 8 and trimer 6. When the probe was
warmed to room temperature (20 °C), all peaks due to
dimer 8 disappeared, and peaks due to tetramer 7
developed. This result demonstrates that tetramer 7 is
formed from the reaction of two dimer molecules, and this
reaction is far slower than the trimerization due to
addition of a molecule of monomer to the dimer.
F igu r e 2. IR difference spectrum obtained after irradiation
of 5 with UV light (254/190 nm) for 14 h in Ar matrix at 7.6
K. Positive peaks: regenerated quinone methide 4; negative
peaks: reacted benzoxete 5.
bands at 1609, 1482, 1463, 1402, 1175, 1120, 930, 859,
850, and 830 cm-1 (Figure 1c). After further photolysis
for 1 h, no quinone methide 4 remained. The new
compound is assigned as benzoxete 5 for the following
reasons: (1) As quinone methide 4 was matrix isolated
in Ar, the photolysis product should be an isomer of 4. A
four-electron electrocyclization of 4 to 5 (Scheme 1) is a
logical reaction. (2) The ab initio calculated IR spectrum
of 5 (Figure 1d) is in good agreement with the experi-
mental spectrum (Figure 1c). (3) The IR spectrum of 5 is
similar to that of the unsubstituted benzoxete7,8 (which
has the main bands at 1600, 1453, 1180, and 790 cm-1).
(4) Further photolysis of 5 at 254 nm regenerates the
quinone methide 4. In this second photolysis, benzoxete
5 was irradiated with a low-pressure UV lamp (254/190
nm). After 14 h photolysis, about 20% of 5 had reverted
to quinone methide 4 (Figure 2). (5) The same benzoxete
5 was also observed after the alcohol 3 was pyrolyzed at
650 °C, and the matrix-isolated product was photolyzed
as above.
2. P h otolysis of Qu in on e Meth id e 4. Pyrolysis of
trimer 6 at 800 °C yielded pure quinone methide 4 in an
Ar matrix at 7.6 K (Figure 1b). This matrix was irradi-
ated at 340 nm using a high-pressure mercury lamp with
a cutoff filter (λ >340-245 nm). The reaction was
monitored by IR spectroscopy. After 30 min photolysis,
most of the yellow quinone methide 4 had disappeared,
and the new compound (colorless) formed had major IR

4,6-Dimethyl-o-quinone Methide and 4,6-Dimethylbenzoxete
J . Org. Chem., Vol. 63, No. 26, 1998 9809
F igu r e 3. IR spectra taken during warm up experiment of benzoxete 5 in an Ar matrix, produced by photolysis of quinone
methide 4 for 27 h, from 7.6 K to room temperature in Ar atmosphere, and at rt in air. IR spectra of quinone methide 4 and
alcohol 3 are also shown for comparison. Peaks marked with * are the IR absorption bands due only to benzoxete 5.
The stability of benzoxete 5 was tested in a warm up
experiment. For this purpose, the quinone methide 4 was
isolated on the 7.6 K target by pyrolysis of trimer 6 at
800 °C but using a lower Ar flow rate. The resulting
higher substance/Ar ratio facilitates the subsequent
evaporation of Ar. The sublimation of trimer 6 into the
pyrolysis tube was controlled at a slow rate, and no
unchanged 6 condensed on the target. Long wavelength
irradiation (>340 nm) was performed until all of the
quinone methide 4 had been converted into benzoxete 5.
Subsequently, benzoxete 5 was slowly warmed while
monitoring the IR spectrum. The initial warming rate
was very low, so the Ar evaporated very slowly without
carrying the material away from the target. After com-
plete evaporation of Ar, the IR spectrum of neat 5 showed
that, as expected, all bands had broadened compared
with the Ar matrix, but were still clearly observable
(Figure 3). At 200 K, the system was isolated from the
vacuum line, and the cryostat was turned off. To prevent
any evaporation of the compounds on the target on
further warming, the system vacuum was reduced by the
addition of dry Ar. It needs to be mentioned here that
the IR spectrum of alcohol 3 is very similar to that of 5.
Due to small amounts of water that might leak into the
system, it is important to ascertain the difference be-
tween the two compounds. By comparison with the IR
spectrum of 3 as shown in Figure 3, it is seen that bands
at 705, 824, and 1114 cm-1 (marked with *) belong to 5
only. When the target was further warmed to -30 °C,
benzoxete 5 remained unchanged. At room temperature,
these bands were still clearly visible in the spectrum,
although new bands probably due to 3 increased, indicat-
ing some reaction of 5 with adventitious water to form
alcohol 3. Air was then introduced into the system,
resulting in an immediate reduction in intensities of the
bands due to 5. After 3 to 4 min, only alcohol 3 was
observed on the target.
Since benzoxete 5 was found to be stable in the solid
form, an attempt was made to isolate it in solution for
NMR spectroscopy. Unfortunately, preparative photolysis
could not completely convert neat quinone methide 4 on
a 77 K coldfinger into benzoxete 5, and a low temperature
(-30 °C) NMR measurement of the products isolated
from the coldfinger after photolysis showed the major
species to be dimer and trimer (8 and 6), together with a
small amount of photolysis product. It was not possible
to identify any peaks in this photolysis product as
compound 5. Most likely, benzoxete 5 is only stable as a
pure solid, i.e., without any other reactive species present.
3. Qu a n tu m Ch em ica l Ca lcu la tion s. It was initially
surprising that solid benzoxete 5 was stable at room
temperature, as such molecules are usually regarded as
reactive intermediates.7 The stability of this molecule
with respect to formation of 4 was therefore investigated
using quantum chemical methods.
The geometries of quinone methide 4 and benzoxete
5, and the transition state connecting them, were initially
calculated at the HF/STO-3G level of theory. Subsequent
calculations were then performed at the HF, B3LYP, and
MP2 levels in conjunction with the standard 6-31G* basis
set. At each level of theory, analysis of the normal modes
of vibration confirmed that the optimized geometries of
4 and 5 were true local minima. The transition states
were similarly analyzed, and at all levels of theory the
imaginary frequency corresponding to formation of the
O-C bond in 5 was located. The energies of 4 and 5 were
also investigated by performing additional single point
calculations using the expanded 6-311+G(3df,2p) basis
in conjunction with the B3LYP/6-31G* and MP2/6-31G*
geometries (optimized and transition state) and zero-

9810 J . Org. Chem., Vol. 63, No. 26, 1998
Qiao et al.
documents. At all levels of calculation, the results are
similar for the parent system and for 4/5. For example,
at the highest level (MP2/6-31G*//MP2/6-311+G(3df,2p)
the parent benzoxete 2c is ca. 6 kcal/mol above 1c, and
the barrier is ca. 32.4 kcal/mol for the reversion of 2c to
1c. Thus, there is probably little intrinsic difference
between the parent system (1c/2c) and the methyl-
substituted one (4/5). As mentioned in the Introduction,
2c is observable till at least 155 K.8 It appears that 2c is
just more reactive than 5 (i.e., a kinetic effect).
F igu r e 4. Ab initio calculated structures of quinone methide
4, benzoxetene 5, and the transition structure TS4,5 for their
interconversion (MP2/6-31G*).
Exp er im en ta l Section
Ta ble 1. Ca lcu la ted Rela tive En er gies (k ca l/m ol; zer o
p oin t en er gies a d d ed ) of Qu in on e Meth id es 1c a n d 4,
Ben zoxetes 2c a n d 5, a n d th e Tr a n sition Sta tes TS1C,2c
a n d TS4,5 In ter con ver tin g Th em
Ap p a r a tu s. Preparative FVP was carried out in a heated
quartz tube, 25 cm in length and 2 cm in diameter. Samples
were sublimed into the horizontal pyrolysis tube using a
sublimation oven. The system was generally evacuated to ca.
10-4 mbar and remained attached to the vacuum line during
the pyrolysis period. The pyrolysis products were trapped on
a coldfinger or in a U-tube at 77 K. Further details of the FVP
apparatus and the purpose of the U-tube have been pub-
lished.9,11
Matrix isolation was carried out using a 10 cm long, 0.8 cm
diameter quartz tube in an oven directly attached to the
vacuum shroud of a closed cycle liq He APD Cryogenics
cryostat, CSW-204SL-6.5K, equipped with a Lakeshore Model
330-44 temperature controller. Argon was used as the matrix
medium, which was passed over the sample while it was
sublimed and cocondensed as a matrix at ca. 6.5-12 K on
BaF2, KBr, or CSI2 targets for IR spectroscopy. A turbomo-
lecular pump was used to maintain a vacuum of 10-6-10-7
mbar.
level of theory
HF/6-31G*
1c TS1c,2c 2c
4
TS4,5
5
0
0
0
55.9
39.3
39.2
15.9
11.0
12.5
0
0
0
58.6 18.3
41.4 13.6
41.1 14.5
B3LYP/6-31G*
B3LYP/6-31G*//
B3LYP/6-311+G(3df,2p)
MP2/6-31G*
0
0
40.2
38.4
7.7
6.0
0
0
41.2 10.5
40.0
MP2/6-31G*//
9.0
MP2/6-311+G(3df,2p)
point energies. The resulting relative energies of 4, 5 and
the transition state linking them (TS4,5), calculated for
thermal reactions in gas phase, are given in Table 1.
Imaginary frequencies for the transition state geometries
are contained in Table S3, and full geometrical data is
given in Table S4. All calculations were performed using
the Gaussian 94 software package.12
As shown in Table 1, at the higher computational levels
and with large basis sets, benzoxete 5 is found to be only
9-10 kcal/mol above the quinone methide 4. While the
aromaticity of 5 would stabilize this molecule, strain
increases the energy. The optimized structures of 4, 5,
and TS4,5 are shown in Figure 4. The activation barrier
is ca. 40 kcal/mol for the conversion of 4 to 5, and about
31 kcal/mol for the reversion of 5 to 4. This is sufficient
to explain the persistence of 5 at room temperature in
the absence of water and other reactive molecules. It is
a thermodynamically stable molecule.
Photolysis was carried out with a low-pressure Hg lamp (75
W, 254 and 190 nm output; Gra¨ntzel, Karlsuhe, Germany) or
a high-pressure Xe-Hg Lamp (1000 W; Hanovia). In the
preparative photolysis, a 2000 W white light theater Lamp
(SEIECON, Luminaire Manufacturer, CP91, Auckland, New
Zealand) was used. Melting points are uncorrected.
Ma ter ia ls. 4,6-Dimethyl-2-(hydroxymethyl)phenol (3) was
synthesized according to a known method.13
P r ep a r a tive F VP of 3. A 703.7 mg sample of 3 was
sublimed at 45 °C (10-4 mbar) and pyrolyzed at 750 °C in the
preparative apparatus. The pyrolysis product was trapped on
a coldfinger at 77 K. After warming to room temperature, it
was dissolved in chloroform, and a column chromatographic
separation (silica gel, 250-400 mesh) afforded as the first
fraction 357.2 mg (56%) of trimer 6. A second fraction (206.6
mg) was recrystallized from ethyl acetate to yield 102.8 mg
(25%) of tetramer 7. Their spectroscopic data are as follows:
Tr im er 6: mp 201-202 °C (lit.14 199-201 °C); 1H NMR (400
MHz, CDCl3): δ 1.63 (s, 3H), 1.78 (d, J ) 1.2 Hz, 3H), 1.99-
2.27 (m, 3H), 2.12 (s, 3H), 2.15 (s, 3H), 2.15 (s, 3H), 2.17 (s,
3H), 2.33-2.73 (m, 3H), 2.89-3.01 (m, 1H), 6.43 (s, 1H), 6.59
(s, 1H), 6.65 (s, 1H), 6.74 (s, 2H); 13C NMR (400 MHz, CDCl3):
δ 15.45 (CH3), 15.89 (CH3), 15.94 (CH3), 20.43 (CH3), 20.53
(CH3), 21.90 (CH2), 25.62 (CH2), 29.19 (CH3), 30.23 (CH2), 49.41
(CH), 77.20 (Cq), 80.59 (Cq), 119.53 (Cq), 124.78 (Cq), 125.53
(Cq), 125.84 (Cq), 125.95 (CH), 126.70 (CH), 128.92 (Cq), 129.35
(CH), 129.63 (CH), 130.01 (Cq), 133.15 (Cq), 144.86 (CH), 148.97
(Cq), 149.22 (Cq), 197.39 (CdO). The definitive NMR signal
assignments were based on DEPT and two-dimensional 13C-
1H correlations with both short (HMQC) and long (HMBC)
range couplings. IR (KBr): 669 w, 723 w, 852 m, 857 m, 910
m, 957 w, 1034 m, 1042 m, 1078 m, 1159 m, 1179 m, 1202 w,
1228 s, 1247 m, 1435 m, 1482 s, 1669 s, 2916 m, 2944 w, 3005
w cm-1; MS m/z 402 (M+, 5%), 268 (45), 134 (100), 112 (30),
91 (60), 77 (20), 44 (44). Anal. Calcd for C27H30O3: C, 80.56;
H, 7.51; O, 11.92. Found: C, 80.10; H, 7.72.
Tomioka and Matsushita7a have reported calculations
for the parent system 1c/2c at the MP2/6-31G*//HF/6-
31G* level of theory. To be able to compare directly with
the results for 4/5, we have carried out the corresponding
calculations on the parent system 1c/2c. The calculated
energies and full geometries are listed in the supporting
(7) (a) Tomioka, H.; Matsushita, T. Chem Lett. 1997, 5, 399. Main
IR bands of 1c in Ar matrix: 1668, 1616, 1568, 1540, 852, 788 cm-1
.
2c: 1600, 1453, 1180, 896, 874, 796, 740 cm-1. (b) Tomioka, H. Pure
Appl. Chem. 1997, 4, 837.
(8) Morawietz, J .; Wentrup C., Unpublished results.
(9) Wentrup, C.; Blanch, R.; Briehl, H.; Gross, G. J . Am. Chem. Soc.
1988, 110, 1874.
(10) The main IR bands of the parent quinone methide 1c (neat,
-196 °C) have been reported at 1656, 1565, and 1539 cm-1: McIntosh,
C. L.; Chapman, O. L. J . Chem. Soc., Chem Commun. 1971, 771.
(11) . Qiao, G. G.; Wong, M. W.; Wentrup, C. J . Org. Chem. 1996,
61, 8125.
(12) . Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Robb, M. A.; Cheeseman, J . R.; Keith, T. A.; Peterson,
G. A.; Montgomery, J . A.; Raghavachari, K.; Al-Laham, M. A.;
Zakrzewski, V. G.; Ortiz, J . V.; Foresman, J . B.; Cioslowski, J .;
Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,
P. Y.; Chen, W.; Wong, M. W.; Andres, J . L.; Replogle, E. S.; Gomperts,
R.; Martin, R. L.; Fox, D. J .; Binkley, J . S.; DeFrees, D. J .; Baker, J .;
Stewart, J . J . P.; Head-Gordon, M.; Gonzalez, C.; Pople, J . A. Gaussian
94; Gaussian Inc., Pittsburgh, PA, 1995.
(13) Ninagawa, A.; Kawazoe, M.; Matsuda, H. Makromol. Chem.
1979, 180, 2123.
(14) Merijan, A.; Shoulders, B. A.; Gardner, P. D. J . Org. Chem.
1963, 28, 2148.

4,6-Dimethyl-o-quinone Methide and 4,6-Dimethylbenzoxete
J . Org. Chem., Vol. 63, No. 26, 1998 9811
Tetr a m er 7:15 mp 206-209 °C dec; 1H NMR (400 MHz,
CDCl3): δ 1.23 (s, 3H), 1.41 (s, 3H), 1.81 (d, J ) 1.3 Hz, 3H),
1.88 (d, J ) 1.4 Hz 3H), 2.01 (s, 3H), 2.06 (s, 3H), 2.16 (s, 3H),
2.18 (s, 3H), 2.20-2.43 (m, 4H), 2.57-2.83 (m, 4H), 2.92 (d, J
) 2.0 Hz 1H), 3.42 (t, J ) 2.2 Hz 1H), 5.18 (s, 1H), 6.02 (s,
1H), 6.57 (s, 1H), 6.64-6.65 (m, 3H); 13C NMR (400 MHz,
CDCl3): δ 12.70 (CH3), 15.83 (CH3), 16.05 (CH3), 16.59 (CH3),
20.39 (CH3), 20.42 (CH3), 21.59 (CH3), 22.08 (CH2), 22.11 (CH2),
23.56 (CH3), 27.45 (CH2), 33.98 (CH2), 44.45 (CH), 45.65 (Cq),
47.79 (CH), 58.33 (Cq), 75.98 (Cq), 79.73 (Cq), 119.06 (Cq),
120.05 (Cq), 124.94 (Cq), 125.02 (Cq), 126.78 (CH), 126.85 (CH),
128.56 (CH), 128.66 (Cq), 129.02 (Cq), 129.56 (CH), 129.58 (CH),
134.35 (Cq), 142.45 (CH), 144.40 (Cq), 148.54 (Cq), 149.44 (Cq),
197.42 (CdO), 208.94 (CdO). The definitive NMR signal
assignments were based on DEPT and two-dimensional 13C-
1H correlation’s with both short (HMQC) and long (HMBC)
range couplings. IR (KBr): 710 w, 854 m, 983 m, 1149 s, 1250
s, 1378 m, 1480 s, 1695 s, 1730 vs cm-1; MS m/z 536 (M+, 5%),
tive yield). Thus, all trimer 6 had reacted, and any alcohol 3
that had formed as an intermediate had thermolyzed again
to the monomer (dimer, trimer), which produced 9 as the final
product.
Dim er (8). Xylenol 3 (30 mg) was sublimed at 45 °C (10-4
mbar) and pyrolyzed at 800 °C in the preparative apparatus.
After the pyrolysis products were trapped on a coldfinger at
77 K, the system was isolated from the vacuum line, and 2
mL of CDCl3 (or acetone-d6) was injected and cocondensed on
the coldfinger. The coldfinger was allowed to warm until the
CDCl3 (or acetone-d6) melted and flowed, together with the
pyrolysate, into a round-bottom flask which was cooled at 77
K in liquid nitrogen. The solution was transferred at low
temperature into a NMR tube, and the NMR measurements
were conducted at a -30 °C probe temperature. The NMR
spectra showed a mixture of dimer 8 and trimer 6 in a ca. 2:1
ratio. NMR data for dimer 8 are as follows: 1H NMR (400
MHz, CDCl3) δ 6.82 (s, 1H), 6.72 (s, 1H), 6.62 (s, H), 6.05 (s,
1H), 2.4-2.8 (m, 2H), 2.23 (s, 3H), 2.15 (s, 3H), 1.96 (t, J )
6.7 Hz, 2H), 1.92 (s, 3H), 1.88 (s, 3H); 13C NMR (100 MHz,
acetone-d6): δ 201.83, 150.16, 141.05, 134.13, 132.53, 130.96,
129.88, 128.71, 127.79, 126.47, 120.45, 80.21, 30.24, 21.52,
21.07, 20.33, 16.21, 15.16.
402 (3.6), 268 (100), 253 (22), 251 (20), 225 (18), 135 (25), 106
12
(10). HRMS calcd for
C H40O4 536.29266; found 536.29289.
36
Anal. Calcd for C36H40O4: C, 80.56; H, 7.51; O, 11.92. Found:
C, 80.65; H, 7.63.
4,6-Dim eth yl-o-qu in on e Meth id e (4). (a ) F VP of 3.
Alcohol 3 was sublimed (10-6 mbar) at 45 °C with argon as
the carrier gas and pyrolyzed at 650 °C. The compound isolated
in the Ar matrix at 6.5 K was assigned as 4,6-dimethyl-o-
quinone methide (4): IR (Ar KBr, 6.5 K) 488 w, 492 w, 501 w,
505 w, 603 w, 613 w, 624 w, 798 m, 849 m, 943 m, 1052 m,
1059 m, 1389 m, 1397 m, 1432 w, 1455 s, 1569 s, 1637 vs,
1642 s, 1668 s, 3521 w, 3533 w, 3704 m cm-1. (b) F VP of 6.
Trimer 6 was sublimed (10-6 mbar) at 110 °C with Ar as a
carrier gas and pyrolyzed at 900 °C. Pure quinone methide 4
was isolated in argon matrix. IR (Ar KBr, 6.5 K) as above.
Bis(2-h yd r oxy-3,5-d im eth ylp h en yl)eth a n e (9). (a) An
ampule was washed with 1,1,1,3,3,3-hexamethyldisilazane and
dried in an oven at 100 °C. A 14.2 mg (0.05 mmol) amount of
trimer 6 was sealed in this tube under a nitrogen atmosphere
before immersion in an oil both at 180 °C. After 4 h, the tube
was cooled to room temperature and the product analyzed by
NMR. Bis(2-hydroxy-3,5-dimethylphenyl)ethane (9) was iso-
lated as a white solid in quantitative yield and further purified
by sublimation: mp 164-166 °C; 1H NMR (300 MHz, CDCl3):
δ 2.19-2.25 (m, 12H), 2.77 (s, 4H), 6.84 (s, 4H); 13C NMR (75.4
MHz, CDCl3): δ 16.22, 20.30 (CH3), 32.12 (CH2), 123.95,
127.17, 128.00, 129.39, 129.55, 150.34; IR (KBr) 459 w, 546
w, 741 w, 788 w, 854 m, 1033 w, 1159 m, 1210 vs, 1256 w,
1310 w, 1376 w, 1389 m, 1459 m, 1484 s, 2860 w, 2917 m,
3010 w, 3387 s/br cm-1; GC-MS m/z 270 (M+, 33%), 135 (100),
91 (25), 79 (10), 41 (8). Anal. Calcd for C18H22O2: C, 79.96; H,
8.20; O, 11.84. Found: C, 79.80; H, 8.22.
4,6-Dim eth ylben zoxete (5). Ar matrix-isolated quinone
methide 4 at 6.5 K, generated from either alcohol 3 or trimer
6, was irradiated with a mercury lamp with a cutoff filter (λ
g 340-345 nm). After 2 h 15 min, 4 was completely converted
to 4,6-dimethylbenzoxete (5): IR (Ar, KBr, 6.5 K) 628 w, 686
w, 689 w, 708 w, 830 vs, 850 s, 859 s, 930 w, 1113 m, 1120 m,
1175 vs, 1253 w, 1342 w, 1402 w, 1463 s, 1482 s, 1609 m, 1653
w, 2874 w, 2935 w, 2961 w cm-1
.
Wa r m u p E xp er im en t on 4,6-Dim et h ylb en zoxet e (5).
Trimer 3 was pyrolyzed at 800 °C, and the resultant quinone
methide 4 was matrix isolated in Ar at 7.6 K. The sublimation
rate of 3 was controlled at a slow rate, so that only the quinone
methide was isolated on the target. The subsequent photolysis
with the high-pressure mercury lamp (>340 nm) was per-
formed for 27 h to convert 4 completely to benzoxete 5. The
target was slowly warmed, so that the Ar evaporated slowly.
The warm-up was constantly monitored by IR spectroscopy.
IR bands at 705, 824, and 1114 cm-1 belong to 5 only and were
used to monitor the presence of 5. At 129 K, 5 remained
unchanged. At -30 °C, the intensities of IR bands due to 5
decreased slightly, and bands due to alcohol 3 started to
appear. 5 remained observable till room temperature as
described in the text. Details are shown in Figure 3.
(b) Trimer 6 was also heated at 180 °C in the same way but
without coating with hexamethyldisilazane. After 4 h, 16.7%
of 9 was isolated.
(c) An ampule was washed with hexamethyldisilazane and
dried in an oven at 100 °C. A 13.8 mg amount of tetramer 7
was sealed in this tube under a nitrogen atmosphere before
immersion in an oil both at 180 °C. After 3 h, the tube was
cooled to room temperature and opened, and the content was
extracted with CHCl3. A 9.1 mg amount of 9 was isolated from
the solution (65.5%).
Su p p or tin g In for m a tion Ava ila ble: Figures S1 and S2
showing observed matrix IR spectra of quinone methide 4
generated by pyrolysis of alcohol 3 at 650 °C and trimer 6 at
850 °C; Figure S3 showing IR spectra during warm up of 4
from 77 K to room temperature, indicating the disappearance
of 4 and formation of dimer between -65 and -33 °C; Table
S1 showing Cartesian coordinates at the B3LYP/6-31G* and
MP2/6-31G* levels of theory for 1c, 2c and the transition
structure TS1c,2c for their interconversion, and 4, 5 and the
transition structure TS4,5 for their interconversion; Table S2
showing calculated electronic and zero point corrected energies
at all levels for 1c, 2c, TS1c,2c, 4, 5, and TS4,5; Table S3
showing harmonic vibrational frequencies at HF, B3LYP, and
MP2 levels with 6-31G* basis set for TS1c,2c and TS4,5;
Tables S4A and Tables S4B showing the geometrical param-
eters at the B3LYP/6-31G* and MP2/6-31G* levels of theory
for 1c, 2c, TS1c,2c, 4, 5, and TS4,5 (17 pages). This material
is contained in libraries on microfiche, immediately follows this
article in the microfilm version of the journal, and can be
ordered from the ACS; see any current masthead page for
ordering information.
(d) A 14.7 mg amount of tetramer 7 was heated at 180 °C
in the same way but without coating with hexamethyldisila-
zane. After 3 h, 2.3 mg of 9 (15.5%) was isolated.
(e) Thermolyses of trimer 6 in the presence of water were
carried out in sealed tubes at 180 °C for 2 and 4 h. The NMR
spectrum of the product from the 2 h reaction indicated the
presence of three compounds, alcohol 3, trimer 6, and com-
pound 9, in a ratio of ca. 1:14: 2.5. As the amount of the 3
was very small, it was confirmed by adding authentic 3 to the
mixture, which resulted in a growth of the CH2 peak. After
the 4 h thermolysis, only compound 9 was obtained (quantita-
(15) A similar structure of the tetramer of 1c has been reported:
Faure, R.; Thomas-David, G.; Bartnik, R.; Cebulska, Z.; Graca, E.;
Lesniak, S. Bull. Soc. Chim. Fr. 1991, 128, 378.
J O981445X