The reaction of α-phenethyl radicals with 1,4-benzoquinone and 2,6-di-tert-butyl-1,4-benzoquinone

Article abstract of DOI:10.1016/j.tet.2010.09.034

The absolute rate constant for 1,4-benzoquinone (BQ) irreversibly trapping α-phenethyl radicals (3) has been determined as 4.4×10⁶ M^-1 s^-1 at 43 °C using acyclic cis azoalkane 9c as a radical precursor. These reactants afford the hydroquinone mono ether 4 at 30 °C but a mixture of products at elevated temperature. 2,6-Di-tert-butyl-1,4- benzoquinone (DTBQ) also reacts with 3 but the cyclohexadienone products are thermally labile.

Full text of DOI:10.1016/j.tet.2010.09.034

Tetrahedron 66 (2010) 8805e8814
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
The reaction of
a
-phenethyl radicals with 1,4-benzoquinone
and 2,6-di-tert-butyl-1,4-benzoquinone
*
Paul S. Engel , Hee Jung Park, Hua Mo, Shaoming Duan
Department of Chemistry, Rice University, PO Box 1892, Houston, TX 77251, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 June 2010
Received in revised form 8 September 2010
Accepted 9 September 2010
Available online 17 September 2010
The absolute rate constant for 1,4-benzoquinone (BQ) irreversibly trapping a-phenethyl radicals (3) has
been determined as 4.4ꢀ106 Mꢁ1 sꢁ1 at 43 ꢂC using acyclic cis azoalkane 9c as a radical precursor. These
reactants afford the hydroquinone mono ether 4 at 30 ꢂC but a mixture of products at elevated tem-
perature. 2,6-Di-tert-butyl-1,4-benzoquinone (DTBQ) also reacts with 3 but the cyclohexadienone
products are thermally labile.
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
also found some substitution on the quinone nucleus. Similarly,
Golubev et al. used ESR to determine by spin trapping that 2-cya-
Aside from their importance in biology,1,2 quinones have been
employed as free radical polymerization inhibitors for over 70
years. The reaction of benzoquinone (BQ) with free radicals was
first reported by Wieland, who used phenylazotriphenyl-methane
as the radical source.3 The products were hydroquinone bis-tri-
phenylmethyl ether, quinhydrone, and 2-phenyl benzoquinone, as
confirmed by Betterton and Waters.4 The first product represents
radical attack on BQ oxygen while the last one arises from attack on
a ring carbon, also called the nucleus. A few years later, Breitenbach
et al.5 showed that BQ inhibited the thermal polymerization of
styrene while Foord examined a large number of potential in-
hibitors, including BQ.6 Cohen studied the benzoyl peroxide initi-
ated polymerization of styrene using BQ and other quinones as
inhibitor.7 He suggested that quinones react with two radicals to
produce a phenol retarder arising from disproportionation of
phenoxy radicals with growing chains or by nuclear substitution on
BQ. Determining the nature of inhibition reactions is complicated
by the small percentage incorporation of the inhibitor end group
into the polymer structure. This point can be illustrated by the work
of Yassin and El-Reedy,8 who studied the quinone inhibited poly-
merization of styrene and reached the unlikely conclusion that the
growing chains attacked BQ on the ring (vide infra). To surmount
this problem, Waters’ group9,10 and later Gleixner et al.11 studied
the thermolysis of the common free radical initiator AIBN with BQ.
The 2-cyanopropyl radicals serve as a model for propagating
methacrylonitrile chains and produce low molecular weight com-
pounds that are easier to characterize. The main products were the
mono- and diethers of hydroquinone 1 and 2, though Gleixner et al.
nopropyl radicals react five times as fast on oxygen as on the
nucleus.12 The unreasonable products13 supposedly arising from
the ketenimine resonance structure of 2-cyanopropyl radicals were
thus shown to be in error.
CN
CN
O
O
O
O
O
OH
CN
BQ
2
1
Rembaum and Szwarc examined the reaction of methyl radicals
from acetyl peroxide with 18 quinones, leading to a scale of relative
reactivities termed methyl affinities.14 BQ was among the most
reactive compounds and it was suggested that methyl radicals
initially attack the quinone nucleus. The possibility was raised in
this work that the final isolated product might arise by intra-
molecular rearrangement of the initial radicaleBQ adduct. Most
relevant to the present report, Breitenbach et al.15 determined the
reaction products of BQ with 1-phenethyl radicals (3), which were
taken as a model for propagating styrene chains. Using azo-a-
phenylethane in benzene at 80 ꢂC as a radical source, they found
products analogous to those of AIBN, namely mono- and diethers 4
and 5. A third compound present in the GC trace was insoluble in aq
hydroxide, which was taken as evidence against a ring substituted
phenol, as exemplified by 6. Using the free radical clock technique,
Citterio et al. determined the rate constant for addition of 5-hexenyl
radicals to BQ as 2.0ꢀ107 Mꢁ1 sꢁ1 at 69 ꢂC.16
* Corresponding author. E-mail address: engel@rice.edu (P.S. Engel).
0040-4020/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2010.09.034

8806
P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
Since much of the early work on quinone radical chemistry
suffered from a lack of modern separation and identification
methods, we have delved into the reactions of two quinones with 3
in some detail. As mentioned above, 3 is structurally similar to
propagating styrene chains. Our goals were as follows: to confirm
and extend the BQ product studies of Breitenbach et al.,15 to analyze
the radical trapping products of DTBQ, to establish the extent of
reversibility, and to determine the rate of BQ radical trapping.
O
O
Ph
Ph
Ph
O
O
Ph
•
Ph
Ph
3
OH
OH
4
5
6
2. An azoalkane precursor of 3
Although radical 3 can be generated byavarietyof methods,28 we
preferred to avoid two component redox systems for our competi-
tion study and focused instead on unimolecular initiators. The fact
that azoalkanes are widely used to generate radicals of a specific
The much more hindered 2,6-di-tert-butylbenzoquinone
(DTBQ) has received less attention in the literature than BQ.
Roginskii and Belyakov17 reported that DTBQ is a useful trap for
alkyl radicals, and they recorded the ESR spectra of several of the
resulting hindered phenoxy radicals. Using kinetic ESR, Maeda and
Ingold18 determined the rate constant for DTBQ trapping of
6-hepten-2-yl radicals as 9.3ꢀ104 Mꢁ1 sꢁ1 at 40 ꢂC while Shlyap-
nikova et al. reported that the trapping rate constant for sec-decyl
radicals was 8ꢀ105 Mꢁ1 sꢁ1 at 50 ꢂC.19 Simandi and Tudos obtained
a value of 61 for the trapping to propagation rate ratio of styrene
polymerization at 50 ꢂC.20 Velikov et al. employed micro-calorim-
etry to study the effectiveness of various quinone inhibitors on the
autoxidation of cumene and polymerization of styrene. DTBQ
showed no induction period in the autoxidation, indicating poor
radical scavenging ability.21 Similarly, Fujisawa et al.22 found that
1 mol of DTBQ only scavenged 0.003 moles of radicals in the AIBN
initiated polymerization of methyl methacrylate. Finally, Hageman
reported the astonishing result that DTBQ trapped two 2-cyano-
propyl radicals, one on each oxygen, to form a hydroquinone di-
ether23 (see below).
structure29 makes Breitenbach’s azo- -phenylethane (9t)30 a logical
a
source of 3 (cf. Scheme 1). However, this compound undergoes
thermolysis only at rather high temperatures (>100 ꢂC)29,31 and we
feared that attack of 3 on our inhibitors would be reversible.
Although 9t is a useful initiator at higher temperatures, we also
sought low temperature sources of 3.
Δ
•
Ph
N
Ph
2
Ph
N
Ph
Ph
3
10
9t
Scheme 1. Thermolysis of 9t.
An appealing alternative is still based on 9t but would replace
thermolysis with ambient temperature photolysis in the presence
of quinones. Unfortunately, the nitrogen quantum yield of 9t at
366 nm in isooctane is only 0.035, the quinones are UV light ab-
sorbers, and excited quinones are photoreactive.32,33 We imagined
that it might be possible to first photoisomerize 9t to the cis isomer
9c, which should be considerably more labile than the trans iso-
mer34 (cf. Scheme 2). Although some acyclic cis azoalkanes revert
thermally to trans,35 our experience suggested that 9c would lose
nitrogen.34 After photoisomerization, the quinone would be added
and the mixture heated mildly to generate 3. Our concern that 9c
would be more prone to the tautomerization sometimes seen36,37
in 9t proved to be unfounded. In fact, we never observed tauto-
merization of either isomer of 9, even though the hydrazone is
more stable thermodynamically than the azo form.38,39
O
t-Bu
t-Bu
O
DTBQ
One factor governing the inhibition efficiency of quinones is
the degree to which the initial trapping step is reversible, il-
lustrated below for
a-phenethyl radicals 3. This topic has re-
ceived little attention in the literature, the most relevant
example being the work of Roginskii et al.24 on hindered phe-
noxy radical 8. It has also been shown that phenoxymethylation
of BQ on carbon25 and radical carboxylation of naphthoquinone
on carbon26 are reversible. The reversibility question was of
great concern to us in this investigation; consequently, we
sought to study BQ and DTBQ at relatively low temperatures.
Not only was reversibility expected to be minimized, but low
temperatures should diminish the redox processes often seen
with quinones.27 Moreover, radical trapping rate constants are
not valid unless the trapping reaction is irreversible under the
measurement conditions.
hν
Ph
N N
•
Ph
Ph
N
2
N
Ph
Ph
X
3
9t
9c
Scheme 2. Formation and thermolysis of 9c.
Irradiation of 0.02 M 9t with a 500 W mercury lamp and
monochromator at 313 nm and ꢁ50 ꢂC for 2.5 h led to formation of
9c in 47.4% conversion (cf. Table 1), with the only by-products being
meso and dl-2,3-diphenylbutane 10, as seen by 1H NMR. Irradiation
at 366 nm was less effective and gave only 20% 9c at the photo-
stationary state, most likely because 9c absorbs light at longer
wavelength than does 9t.34,40 We also tried both irradiation
through a potassium chromate filter and 254 nm irradiation with
O•
O•
Table 1
Relative methine 1H NMR peak areas for 313 nm monochromator irradiation of
t-Bu
t-Bu
kf
•
0.02 M 9t at ꢁ50 ꢂC in toluene-d8
BQ
Ph
Time, h
9t
9c
Total 9
% 9c
% 9 remaining
kr
0
1.0
2.5
1.52
0.954
0.819
0
1.52
1.49
1.27
0
35.9
47.4
(100)
97.8
83.6
Ph
O
3
O
0.534
0.603
t-Bu
7
8

P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
8807
added naphthalene and pyrene as singlet sensitizers.41 Although
too much diphenylbutane (10) was formed, we learned in these
experiments that 0 ꢂC was cold enough to produce and store 9c. Use
of a fresh sample proved critical to achieve a photo-stationary state
•
O
O
Ph
Δ
Ph
O
+
2
CO2
+
Ph
O
3
11
rich in 9c, probably because
a-phenethyl radicals recombine to
Scheme 3. Formation of 3 by thermolysis of 11.
a small extent at the ortho position and produce a light-absorbing
triene.42 Indeed, the UV spectrum of 9c after warming in isooctane
showed a 306 nm band that would certainly interfere with 313 nm
irradiation. The UV spectrum of 9c exhibited a broad absorption
with lmax w376 nm, which is w10 nm longer than that of 9t.34,40
Thermolysis of 9c, which was monitored by 1H NMR (cf. Table 2),
proceeded smoothly at temperatures slightly above ambient, with
conducted in the presence of BQ and another inhibitor 2,6-di-tert-
butyl-4-benzylidenecyclohexa-2,5-dienone (QM)54 while the fourth
run included no added inhibitor (cf. Table 3). The observation that all
points fell on the same Eyring plot mitigates against self induced
decomposition of 11 and decomposition induced by the in-
hibitors.14,55,56 The activation parameters show that 11 is somewhat
more stable than 9c but far more labile than 9t. However, storage of
neat 11 at ꢁ25 ꢂC for several months resulted in much de-
composition. Extrapolation of the activation parameters for 11 leads
to a predicted half life of many years at this temperature, indicating
that self induced decomposition becomes important in neat
samples.
activation parameters
D
Hz¼28.3 kcal/mol,
D
Sz¼15.6 eu, and
D
Gz
(40 ꢂC)¼23.4 kcal/mol. No regeneration of 9t was observed, but
instead the products were the usual dimers 10. The thermolysis rate
of 9c was not accelerated by inclusion of BQ or DTBQ, thus showing
the expected43 absence of quinone-induced decomposition. Com-
pound 9c was also thermolyzed in acetonitrile, giving a roughly
3-fold slower decomposition rate, in accord with earlier reports
concerning cis azoalkanes in polar solvents.44
Table 3
a
Thermolysis rate constants of 0.02 M 11 in C6D6
Temp,ꢂC
105 k, sꢁ1
Table 2
a
2.81b
5.85c
12.1b
40.0b
Decomposition rate of 9c in toluene-d8
39.2
44.2
49.9
60.4
Temp, ꢂC
104 k, sꢁ1
Half life, h
30.2
37.8
46.2
0.708
2.12
7.81
2.72
0.908
0.246
Monitored by 1H NMR with TMSOTMS as internal standard.
With 0.02 M BQ and 0.02 M QM.
Compound 11 alone.
a
b
a
c
Concentration of 9c was 0.011e0.022 M.
One problem with using 9c as the source of 3 is that its prepa-
ration is always accompanied by unreacted 9t. Compound 9t will
then decompose on the GC injector to generate 3 at a far higher
temperature than the original experiment and introduce un-
certainty into quantification of quinones and their products.
However, this problem disappears if analysis is conducted by HPLC
or NMR, as was done in much of this work. With these analytical
methods, 9c proved to be a good low temperature source of 3. It
was later found possible to separate 9c and 9t by TLC, HPLC, and
column chromatography; hence, GC analysis could be used with
pure 9c as the radical precursor if no other thermally labile com-
pounds were involved.
The thermolysis products of 11 were analyzed by GC to ensure
that this perester was a viable source of 3. The dominant product
was dimer 10, accompanied by styrene, ethylbenzene, and styrene
oligomers. The other products contained oxygen or were of
unknown structure (10e15% of total GC peak area). No detailed
discussion of the products is presented here because they are
similar to those of bis-1-phenethylacyl peroxide.57 Although the
complexity of the product mixture from thermolyzed 11 alone was
a disadvantage in analysis of the BQ products, both 9c and 11
generate 3 in substantial yield, allowing us to study the reaction of
3 with quinones.
Not all a-phenethyl radicals from thermolysis of 9c are available
4. The trapping rate constant of 3 with BQ
for reaction on account of the cage effect. Quantitatively, the cage
effect is the percent of radicals formed that recombine before dif-
fusing into the bulk solution. Because this phenomenon has been
studied exhaustively,45,46 we did not measure the cage effect for 9c;
nevertheless, a good estimate can be derived from the 29e32%
value for 9t at 105 ꢂC.47,48 Since both cis and trans azoalkanes ex-
hibit the same cage effect at a given temperature,49,50 we extrap-
olated these figures to our typical temperature of 43 ꢂC using the
activation energy of 1.3 kcal/mol for the diffusion to recombination
ratio of cumyl radicals.51 The result is a 38% cage effect, corre-
sponding to a radical efficiency of 62% for 9c at 43 ꢂC. The un-
avoidable formation of 10 is a disadvantage of 9c but this product is
relatively inert and gives an easily recognized NMR and GC signa-
ture (1:1 mixture of diastereomers).
In the competition kinetics method, a trapping reaction of un-
known rate is allowed to compete against one whose rate is not
only known but, which falls in the same range as the one to be
determined.58 TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl) is
a common bimolecular trap59e61 but we soon found that it com-
bines with 3 much faster than does BQ. We therefore used the
stable nitroxyl radical SG1, whose reaction rate with 3 is a conve-
niently slow 4.0ꢀ106 Mꢁ1 sꢁ1 at 60 ꢂC.61 BQ and SG1 were simply
allowed to compete for 3 generated by thermolysis of 9c. Provided
that no unwanted reactions occur, the ratio of starting materials
reacted yields the rate ratio.
N
•O
P(OEt)2
3. A perester precursor of 3
O
SG1
Another mild source of 3 was found in the new perester 11, which
was used in a few of the studies reported here but ultimately proved
less desirable than 9c. The perester undergoes homolysis of the OeO
bond followed by loss of CO2 and acetone to afford two radicals 3 (cf.
Scheme 3).52,53 The activation parameters for thermolysis of 11 were
The advantage of this method is that the products can be
ignored; on the other hand, quantifying starting material disap-
pearance is subject to greater error than product appearance.
Initial control experiments showed that SG1 in benzene was
stable at 60 ꢂC. A solution of SG1 and BQ in benzene heated at 60 ꢂC
for 3 h showed no change in the reactant ratio and no new HPLC
determined as
D
Hz¼25.1 kcal molꢁ1
,
D
Sz¼0.90 eu, and
D
Gz (40 ꢂC)¼
24.8 kcal/mol based on four kinetic runs. Three of these runs were

8808
P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
peaks. For the competition experiment, a toluene solution of 0.37 M
SG1, 0.035 M BQ, 0.27 M 9c was divided into five ampoules, which
were degassed and sealed. Compound 9c had been prepared by
irradiating 9t for 5 h at ꢁ10 ꢂC. These ampoules were heated at
43 ꢂC for various times, removed from the bath, and analyzed
several times by HPLC without workup to determine the amount of
SG1 and BQ reacted (cf. Table 4). The wavelength was set to 450 nm
to avoid light absorption by any other compounds besides BQ and
SG1, which are colored.
of later retention time were also apparent. We assigned the first of
these to diether 5 because its mass spectrum (m/z¼318, 214, 110,
105) matched that of Breitenbach.15 The latest GC peak was
assigned as 6 in part because it was obviously an isomer of 5. Loss of
styrene from the molecular ion affords the base peak at m/z¼214,
which fragments further with loss of methyl. Both 4 and 5 exhibited
a hydroquinone ion at m/z¼110 but 6 showed no such peak. Al-
though the earlier workers had concluded on the basis of its in-
solubility in aq KOH that 6 was not a phenol, this test is unreliable
for a large phenol.
Table 4
Further support for structures 4, 5, and 6 was obtained by
treating a BQþ9t thermolysis mixture with diazomethane. The GC
peak due to 4 disappeared and was replaced by one of shorter re-
Relative trapping rate constant of 3 in PhMe by SG1 and BQ
Time, mina
SG1b
BQb
kSG1/kBQ
c
ꢃ
tention time whose MS showed a shift of Mþ and m/z¼110 to
0
1431
1366
1221
1125
1054
1330
1258
1050
980
15
30
45
60
0.85
0.67
0.79
0.90
higher mass by 14 units, indicative of a methyl ether. As expected,
the peak assigned to 5 remained unchanged while that due to 6
moved to shorter retention time. The MS of this new peak showed
the same mass shift as 4, consistent with the methyl ether of 6. A
number of other minor GC product peaks were noticed, some with
long retention times, but they were not investigated further. We
conclude that under mild conditions 4 is the major product of
BQþ3. NMR showed a minor, unidentified, labile product along
with some minor thermally stable compounds. Under no condi-
tions did we see 12 nor any cyclohexadienones.
946
a
Thermolysis time.
HPLC peak area at 450 nm, average of 5e6 injections.
Relative trapping rate by SG1 versus BQ calculated from parallel second order
b
c
reaction kinetics.58
The average value of kSG1/kBQ¼0.80 was converted to an abso-
lute value of kBQ¼4.4ꢀ106 Mꢁ1 sꢁ1 using the calculated absolute
rate kSG1¼3.5ꢀ106 Mꢁ1 sꢁ1 at 43 ꢂC taken from Sobek et al.61 We
Taking advantage of our three unimolecular sources of 3, we
determined the relative amount of the three GC products over
a wide temperature range, leading to the results in Table 5. These
data show an obvious increase in diadducts 5 and 6 at higher
temperatures. Although the GC trace of the 132 ꢂC reaction showed
some minor peaks besides ethylbenzene, styrene, and styrene
oligomers, TLC revealed no spot at the origin, suggesting a relatively
clean reaction. In contrast, Breitenbach15 obtained 42% of ‘non-
volatile, high molecular weight material’ in benzene at 80 ꢂC. We
observed no hydroquinone by GC at either low or high temperature,
unlike some earlier work with BQ and primary alkyl and phenyl
radicals.27,64e66
conclude that BQ is a relatively rapid scavenger of
radicals.
a-phenethyl
5. BQ product analysis
Because the literature on BQ contains a wide variation in
product structures depending upon the attacking radical and the
mode of radical generation, we carried out a careful product study
of BQ reacting with 3. First, a degassed solution of 9t in CD3CN was
thermolyzed at 120 ꢂC to obtain the 1H and 13C NMR spectra of the
two diastereomeric phenethyl dimers 10. Then a CD3CN solution of
9t was irradiated at 10 ꢂC until about half of the trans isomer had
been converted to cis. This solution exhibited small 1H NMR peaks
due to 10 plus smaller ones due to acetophenone, formed pre-
sumably by reaction of 3 with residual oxygen. Thermolysis of the
9c, 9t mixture at 30 ꢂC caused disappearance of 9c, a large increase
Table 5
Products of BQ with 3
Precursor
[Precursor], M
[BQ], M
Temp, ꢂC
4a
5a
6a
in 10, and a few tiny new peaks possibly due to a,o and a
,p dimers,42
whose position was noted to avoid confusing them with BQ prod-
ucts in the next experiment.
When BQ was reacted with pure 9c (isolated by preparative TLC)
in CD3CN at 30 ꢂC, the 1H and 13C NMR spectra clearly showed 4.62
Also apparent were 1-phenethyl alcohol, acetophenone, and a few
small, unidentified 13C NMR and 1H peaks. No tautomerization of 9c
to the corresponding hydrazone was detected nor was the product
distribution affected by changing the solvent to benzene-d6. No-
tably absent were any unexplained carbonyl signals and peaks due
to ring substitution product 12.62
9c
11
11
11
9t
9t
9t
9t
0.02
0.04
0.4
0.04
0.04
0.15
0.15
0.03
0.01
0.02
0.10
0.02
0.02
0.10
0.10
0.02
30
35
60
75
85
85
120
132
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
d
d
d
d
d
d
0.081
0.070
0.056
0.26
0.44
0.12
0.11
0.14
0.23
0.17
a
Relative GC peak areas.
6. DTBQ product analysis
OH
The products of 3 and DTBQ were investigated initially under
mild conditions using atom transfer radical addition (ATRA).67,68 A
mixture of DTBQ, 1-phenethyl bromide, CuBr, and PMDETA in tol-
uene was degassed and stirred at room temperature for 2 h. GC
analysis demonstrated consumption of phenethyl bromide and
formation of dimer 10. Interestingly, while the UV spectrum
revealed no absorption for DTBQ, the GC trace exhibited a peak for
this quinone. Flash chromatography allowed isolation of an oil
whose 1H NMR spectrum was consistent with a mixture of two
diastereomers of 13 (cf. Scheme 4). Diastereomer 13A exhibited
cyclohexadienone peaks at 5.99 and 6.56 ppm while 13B showed
these peaks at 6.16 and 6.42 ppm. The 13C NMR spectrum of 13B
showed a characteristic cyclohexa-2,5-dienone carbonyl peak69e71
Ph
OH
12
Further product analysis was carried by GC/MS. Heating BQ with
11 at 50 ꢂC yielded only one major product peak whose MS showed
it to be 4, in accord with Breitenbach et al.15 As expected,63
a characteristic fragmentation of the aryl ethers in this series was
loss of neutral styrene. When 3 was generated at 120 ꢂC using 9t as
the radical source, 4 was still the major product but two new peaks

P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
8809
O
O
However, 13C NMR showed a prominent quaternary carbon at
185.1 ppm, characteristic of the cyclohexa-2,5-dienone sys-
tem.69e71 It follows that moderately hindered radicals such as 3 and
2-cyanopropyl attack DTBQ at the less hindered oxygen and then at
the aromatic ring 4-position. The resulting cyclohexadienones are
thermally labile at elevated temperatures because a structure such
as 17 lacks the aromatic stabilization that its isomer 16 possesses.
This ground state destabilization is also found in 3-methylene-1,4-
cyclohexadienes, which dissociate much more easily than their
aromatic tautomers because the energy required to reach the rad-
icals is lower when starting from the non-aromatic compound.73
The disappearance of the diastereomeric mixture 13A,B fol-
lowed first order kinetics with kd¼1.39ꢀ10ꢁ4 sꢁ1 at 120 ꢂC
•
t-Bu
t-Bu
Ph
t-Bu
t-Bu
2
Ph
O
O
Ph
DTBQ
13
Scheme 4. Trapping of 3 by DTBQ.
at 185.6 ppm while close examination of the spectrum revealed
smaller peaks at 205.3 and 206.1 ppm, indicative of ortho re-
combination of 3 and 14.69,72
The diastereomers were separated by HPLC and a solution of
13B in toluene-d8 was degassed, sealed in an NMR tube, and heated
at 120 ꢂC. Compound 13B gradually disappeared while 13A built up
and then eventually disappeared also (cf. Scheme 5).
(t ¼1.39 h), which is the dissociation rate of 13A,B assuming that
½
the rates are the same. The product mixture was analyzed by NMR
and GC, revealing 15 and DTBQ in a 1:8 ratio, along with styrene.
The presence of DTBQ is interesting because it indicates that even
the first addition of 3 to DTBQ is reversible (cf. Scheme 5). Since 13
was formed in the original phenethyl bromide reaction, the dis-
appearance of DTBQ by UV but not by GC is readily explained as
thermolysis of 13 on the GC injector. In fact a GC trace of 13
exhibited not only sharp peaks for DTBQ, 10, and 15 but also a very
broad peak at w25 min indicative of a thermally labile compound.
An independent experiment with DTBQ and 9t gave similar
results. A solution of 0.01 M DTBQ and 0.03 M 9t in benzene was
degassed, sealed in a UV cell, and heated at 120 ꢂC for 3 h.
According to UV, DTBQ was gone but GC again showed the char-
acteristic peak for this quinone. We also saw phenol 15 and the
broad 25 min GC peak mentioned above. NMR showed a small
amount of 15 plus the diastereomers of 13, indicating that they had
partially survived the reaction.
O•
O
O
t-Bu
R
t-Bu
Ph
t-Bu
R
t-Bu
Ph
t-Bu
Ph
t-Bu
•
Ph
3
O
O
Ph
Ph
O
R
S
13B
14
13A
O
H
O
O
t-Bu
t-Bu
t-Bu
Ph
t-Bu
+
•
+
Ph
Ph
3
O
DTBQ
15
Scheme 5. Generation and fragmentation of phenoxyl radical 14.
In order to secure an authentic sample of 15, a solution of 1 M 13
and 300 mg thiophenol in 0.5 mL toluene-d8 was degassed, sealed
in an NMR tube, and heated at 120 ꢂC for 2 days. 1H NMR showed no
DTBQ but a large amount of 15, which could be isolated by pre-
parative reverse phase HPLC and identified by NMR. Obviously,
thiophenol effectively scavenged 14 before it could fragment.
This kinetic behavior suggested homolysis of 13B to 14, followed
by recombination of the formed radicals 14 and 3 to 13A in com-
petition with disproportionation of the radicals to 15þstyrene. The
designations R and S are arbitrary since we do not know the ab-
solute configuration of the diastereomers. In view of the relatively
high temperature needed for thermolysis of 13, we conclude that
these cyclohexadienone products are stable at 25 ꢂC where our
ATRA study was carried out. Formation of 13 is not surprising in
view of the dominant disproportionation of 3 and 7, coupled with
the large steric hindrance to this process in 14. In this regard, a re-
port23 on AIBN reacting with DTBQ stated that the exclusive
product formed in 90% yield was diether 16. While such a structure
(5) is observed with BQ and 3 at elevated temperature and with BQ
and 2-cyanopropyl radicals (2),9 recombination of 2-cyanopropyl at
the hindered phenoxy radical from DTBQ is most unlikely. The only
evidence for structure 16 was 1H NMR, which exhibited an ‘aro-
matic’ singlet at 6.80 ppm. Our experience with 13 and 15
7. Reversibility of BQ scavenging a-phenethyl radicals
Our planned approach to test for reversibility in the reaction of
BQ with 3 was to expose phenol 4 to a nitroxyl radical, hoping to
form 7, which would then cleave and lead to the trialkylhydroxyl-
amine from 3. However, the required phenol 4 proved to be a dif-
ficult synthetic target and it was unstable to normal phase HPLC on
silica gel. Although isolated 4 was unknown at the time of our work,
the recent literature contains one reference to this phenol, which
was made from an alkylcatecholborane and was isolated by reverse
phase HPLC.62 (See later discussion of 4.) When a small sample of 4
was exposed to the bicyclic nitroxyl ABNO at 25 ꢂC, the solution
rapidly formed a deep red solid. GC analysis showed many un-
known peaks but not the expected trapping product 18, even when
the reaction was repeated at 120 ꢂC.
suggested that this peak could just as well arise from the b-cyclo-
hexadienone protons of 17. To test this possibility, a sealed,
degassed NMR tube of DTBQ and AIBN in a 1.0:1.5 mole ratio in C6D6
was heated at 80 ꢂC for 8 h. 1H NMR revealed nearly complete
disappearance of AIBN and appearance of Hageman’s 6.80 singlet at
6.77 ppm.
Ph
O
N
CN
O
O
t-Bu
NC
t-Bu
t-Bu
t-Bu
CN
18
O
O
To circumvent the scarcity of 4, we carried out further experi-
ments with the commercially available 4-benzyloxyphenol 19,
reasoning that fragmentation of 20 would require the same process
in 7. Furthermore, we used TEMPO in place of the much less
CN
16
17

8810
P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
accessible ABNO, which behaved poorly with 4 (see above). The
high reactivity of ABNO relative to TEMPO is readily explained by its
6.6 kcal/mol higher OH bond strength.74 Another advantage of 19
over 4 is that radical trapping product 21 is more stable thermally
8. Discussion
We employed 11 and 9c as simple, unimolecular, low temper-
ature sources of free radicals in our competition kinetics experi-
ment to avoid possible complications from metal induced redox
reactions. ATRA chemistry was used in one product study and led to
than it’s
a-phenethyl analog and can therefore be analyzed by
GC.67,75e78
O
O
OH
O•
O•
O
O
N
O
Ph
O
Ph
O
O
O
O
Ph
Ph
Ph
20
19
21
22
23
We found that TEMPO did not react with 19 thermally nor upon
irradiation, which is known to enhance hydrogen atom transfer to
TEMPO.76,79 Even visible light and 313 nm irradiation of a mixture
of 19 and TEMPO at 100 ꢂC revealed no obvious disappearance of
starting materials and no formation of 21, as proven by GC com-
parison with an authentic sample. In a second approach, a mixture
of 19 and TEMPO was shaken at 100 ꢂC for 15 h with active lead
dioxide,80 which is expected to oxidize 19 to 20.24,81 Again there
was no sign of 21, though 19 was gone and TEMPO remained.
Control experiments proved that lead dioxide did not attack TEMPO
and although it destroyed 19, no products were recognizable by GC.
These results argue against fragmentation of 20. Finally, an attempt
was made to generate 20 at elevated temperatures by irradiation of
oxalate 22. It is known that diaryl oxalates decarbonylate upon
254 nm irradiation to generate phenoxy radicals.82 If 20 were to
fragment, we would expect to see BQ, bibenzyl, hydroquinone
dibenzyl ether, or toluene formed by hydrogen abstraction from
solvent. Samples of 22 were irradiated at 254 nm and 100 ꢂC in THF,
t-BuOH, CH3CN, and as a neat solid. None of these products were
formed; instead, we saw only 19. If 20 does not fragment at 100 ꢂC,
no unexpected behavior that could not be readily attributed to
reaction temperature differences. Other phenethyl radical sources
have been employed in the past16,27,28 but few are likely to compete
with 9c for chemical simplicity.
Our kinetic study of perester 11 (t1/2¼29 min at 60 ꢂC) showed it
to be somewhat more labile than the analogous 24 (t1/2¼2.16 h at
60 ꢂC)52 as expected from the greater stability of
a-phenethyl
radicals than tert-butyl radicals.86 The yield of free 3 is not quan-
titative, as evidenced by the formation of oxygenated products.
O
O
O
24
Azoalkanes 9t and 9c are much cleaner phenethyl radical pre-
cursors than 11. While 9t is only useful above 85 ꢂC, 9c decomposes
at temperatures slightly above ambient. Thus the same radicals can
be generated over a large temperature range. Not only does 9c
exhibit no induced decomposition nor tautomerization under our
conditions, but it can even be isolated pure. The behavior of 9c and
9t is in accord with the sizeable body of knowledge about ther-
molysis of cis and trans azoalkanes.29,34,40
Table 6 summarizes the present and previously published rate
constants for trapping of radicals by BQ. The figures span a range of
nearly 105 and depend on the structure of the attacking radical. Our
value of 4.4ꢀ106 Mꢁ1 sꢁ1 for 3 should be similar to the one for
propagating styrene chains, but it is actually about 50 times faster.
This difference could be due to slower diffusion in the polystyrene
solution or to the fact that the propagation rate of styrene is a poor
clock because the many literature reports of its rate are widely
is it possible that the a-methyl group of 7 allows fragmentation
during our competition kinetics measurement, which was run at
43 ꢂC (Table 4)?
The fragmentation rate of 7 at 43 ꢂC can be roughly evaluated
from a theoretical calculation on the 4-allyloxyphenoxy radical 23.
Homolysis of 23 to BQ plus allyl is calculated at the B3LYP level to be
endothermic by 12.9 kcal/mol.83 Since the benzyl radical is
w2 kcal/mol less stable than allyl,84 we estimate that cleavage of 20
is endothermic by 14.9 kcal/mol. Subtracting the 2.7 kcal/mol sta-
bilization energy of 3 versus benzyl, derived from azoalkane ther-
molysis rates,29 indicates that fragmentation of 7 is uphill by
12.2 kcal/mol. The activation energy for this process is certainly not
zero but should be at least 3 kcal/mol while the A factor will be
taken as 1013
, by comparison with similar reactions. Using
Ea¼15.2 kcal/mol and log A¼13 in the Arrhenius equation leads to
a fragmentation rate constant at 43 ꢂC of 310 sꢁ1, which is 500
times slower than the pseudo first order trapping rate constant
(0.035 Mꢀ4.4ꢀ106 Mꢁ1 sꢁ1) for 3þBQ. A far smaller fragmentation
rate for 7 has been calculated by Denizov,85 who obtained
Ea¼28.2 kcal/mol and used log A¼1014. These values lead to a frag-
mentation rate constant of only 3.2ꢀ10ꢁ6 sꢁ1, which is 5ꢀ1010
slower than trapping. We conclude that the reaction of BQ with 3 is
irreversible at 43 ꢂC, thus validating our measured trapping rate.
Furthermore, the concentration of SG1 in the competition kinetics
Table 6
Radical trapping rate constants of BQ
Entry
Radical
Temp, ꢂ
C
Rate constant, Mꢁ1 sꢁ1
A
Ref.
1
2
3
4
5
6
Me2CCN
Me3C
35
26
50
69
43
65
6ꢀ102
4.1ꢀ107
8.8ꢀ104b
2.0ꢀ107
4.4ꢀ106
5.2ꢀ106
12
12
20
16
This work
90
RCHPha
RCHꢃ2
PhCHMe
CH3
a
Polystyryl radical.
experiment was 10.6 times higher than that of BQ. Any a-phenethyl
radicals arising from fragmentation of 7 are therefore unlikely to
recombine with BQ.
b
Calculated from the known propagation rate of styrene (170 Mꢁ1 sꢁ1 at 50 ꢂC)91
and the reported ratio of the inhibition rate constant to the propagation rate
constant.

P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
8811
scattered.87 The 9-fold faster addition rate of tert-butyl (entry 2)
than methyl radicals (entry 6) is surprising but the much faster rate
of 3 compared to 2-cyanopropyl (entry 1) is probably due to the
polar effect in radical addition reactions.88,89
t-Bu
O
t-Bu
t-Bu
O
O
t-Bu
t-Bu
t-Bu
OR
t-Bu
8.1. BQ product studies
R = Me, Et, n-Pr, PhCH2-, t-Bu
The reaction of BQ with free radicals often leads to ring sub-
stitution products analogous to 12.12,13,16,27,92e98 Many of these
studies were carried out using metal ions or strong oxidizers, which
could affect the nature of the products, and others did not employ
modern methods of structure determination. In the case of phe-
nethyl radicals, the products we found (4e6) arise by attack at
quinone oxygen. Breitenbach et al. reported 4 and 5 in roughly a 6:1
ratio starting with BQ and 9t in chlorobenzene while Kumli et al.62
found 4 and 12 in a 7:1 ratio using a borane precursor. The effect of
temperature on the BQ product distribution has not been observed
previously but was easily studied here because three sources of 3
were available. When comparing products of quinones with various
radicals, it is now necessary to consider the possible effect of re-
action temperature. In view of our findings, the claim8 that prop-
agating polystyrene chains substitute BQ on the ring becomes
doubtful.
A possible explanation for the formation of 12 in the work of
Kumli et al.62 versus its absence in our study is that 12 might be
a hydrolysis product of 6. Although 4 is unstable to silica gel, we
found by GC that subjecting a mixture of 4, 5, and 6 to the published
workup (washing with saturated NH4Cl and drying with MgSO4)62
led to no perceptible loss of these products relative to dimers 10.
Even if 12 were a hydrolysis product of 6, our analysis did not reveal
6 at ambient temperature by 1H NMR. We suggest that our product
distribution differs from that of Kumli et al.62 because a non-radical
pathway intervenes in the borane chemistry.99
27
28
Dimer 26 resembles 28, which can be isolated but dissociates
rapidly in solution.101e103 Although 26 does not suffer the front
strain present in 28, radical 7 is substantially stabilized by the ad-
ditional oxygen at the para position.104,105 Thus 26 may form but
dissociate rapidly even at 30 ꢂC in our NMR experiments. Simple
phenoxyl radicals also dimerize carbon to carbon106,107 but again,
we did not see any unexplained cyclohexadienone protons nor
carbonyls in our 30 ꢂC NMR experiment.
Our surprisingly high yield of 4 may arise because electron-rich
radicals such as 3 reduce BQ to the radical anion 29 (cf. Scheme 7).
Recombination then takes place on oxygen because it has more
negative charge than the ring carbons.16,20,93,108 After 3 attacks BQ
to form phenoxy radical 7, the subsequent reaction at lower tem-
peratures is disproportionation to 4 rather than recombination to 5.
Recombination of 3 and 7 must have a higher activation energy
than disproportionation since 5 and 6 appear only at elevated
temperatures. The dominance of disproportionation (3þ7/4, not
5) is noteworthy because the delocalized radicals 3 exhibit roughly
nine times more self recombination than disproportionation.109 On
the other hand, alkoxy radicals exhibit far more disproportionation
than recombination, both in self reaction and in cross reaction with
alkyl radicals.109 Apparently, the 3þ7 pair takes on much of the
character of the alkoxy pairs, possibly due to involvement of an-
ionic phenoxide structures.110
Some of the plausible reactions (dashed arrows) and the main
observed reactions (solid arrows) of 3þBQ are shown in Scheme 6.
The absence of possible product 25 is not due to its homolytic
dissociation under these mild conditions because analogs 13 and
2724,100 are readily isolable. Instead, the failure of 7 to produce 25
O•
O
Ph
O
+
•
3
+
+
Ph
7
Ph
3
-
O
O
O
O•
•
•
OH
BQ
Ph
OH
Ph
29
4
3
3
+
Ph
Scheme 7. Electron transfer from 3 to BQ.
O
Ph
O
Ph
O
The exclusive formation of 4 rather than 5 at low temperatures
7
•
4
dimerize
stands in contrast to the behavior of 2-cyanopropyl radicals from
AIBN, which lead to both mono- and diether (1 and 2)9,11 plus some
attack on the quinone ring.11,12 In view of the lack of modern ana-
lytical techniques employed by some of the earlier workers, it
would be interesting to re-examine the AIBNþBQ products and to
determine whether the temperature effect on product distribution
found here extends to other radicals.
Ph
1
O
O
Ph
O
O
Ph
Ph
O
Ph
O
25
26
8.2. Reversibility
Scheme 6. Reaction of BQ with 3.
We have demonstrated that
a-phenethyl radical trapping by
leads to the interesting conclusion that disproportionation to 4 and
styrene is a faster reaction than para recombination and suggests
that the bulky tert-butyl groups slow down disproportionation of
the phenoxy and phenethyl radicals that actually lead, especially at
lower temperatures, to 13 (i.e., tert-butyl groups slow down the
reaction 7þ3/4). Based on our observation of 6, it is likely that 25
forms at high temperatures, but this compound would not survive
the GC injector. Because these product studies were carried out
with a clean, unimolecular source of 3 without workup, we believe
that they reflect the inherent reactivity of BQ with 3.
DTBQ on oxygen is reversible. In a rarelycited paper, Roginskii et al.24
showed in 1981 that phenoxy radical 8 undergoes fragmentation
with a half life of 14 min at 94 ꢂC, which is directly analogous to the
fragmentation of 14. In contrast to 14 at 120 ꢂC, the same process in 7
is slow at 43 ꢂC. Besides the expected large effect of temperature, 14
is a persistent phenoxy radical;81 hence, it survives long enough to
undergo a slow reaction like fragmentation while 7 quickly follows
other reaction paths like disproportionation.
We find that phenoxy radical 14 recombines in part with a sec-
ond a-phenethyl radical to afford cyclohexa-2,5-dienone 13. This

8812
P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
type of reaction was suspected earlier in the reaction of DTBQ with
sec-decyl radicals, but the only evidence was an UV absorption
band at 240 nm.19 Since we now know that 2-cyanopropyl radicals
behave the same way, it is likely that addition of two radicals is
common during hindered quinone inhibited polymerization of
monomers. The cyclohexadienones are thermally labile, compli-
cating the radical scavenging process at elevated temperatures.
10.3. trans Azo-
a-phenylethane 9t
This trans azoalkane was prepared from acetophenone in 40%
yield according to a published procedure.30 1H NMR (CD3CN)
d
7.35
(m, 10H), 4.63 (q, 2H), 1.48 (d, 6H). 13C NMR (CD3CN)
128.5, 128.4, 77.3, 20.6.
d 142.5, 129.7
10.4. Photoisomerization of 9t
9. Summary
A 40 mg sample of 9t was dissolved in 0.3 mL CD3CN or C6D5CD3
in a medium wall NMR tube. To avoid thermolysis of 9c, the tube
was placed in a copper cooling coil that kept the temperature
at ꢁ20 to ꢁ10 ꢂC. The cold solution was irradiated for 5e6 h at
313 nm and the conversion to 9c was monitored by the methine
proton of 9c at 4.83 ppm and 9t at 4.63 ppm (C7D8). Compound 9c:
We have developed methods for generating a-phenethyl radi-
cals 3 by mild thermolysis of precursors 9c and 11. cis Azoalkane 9c
was employed in a competition study to determine the rate con-
stant of trapping 3 with BQ. While previous reports claimed oxygen
and ring attack on BQ, we established that the initial attack of 3 is
on oxygen but that the product distribution is temperature-de-
1H NMR (CD3CN)
d
7.20ꢁ7.29 (m, 10H), 5.32 (q, J¼6.4 Hz, 2H), 1.67
(d, J¼6.4 Hz, 6H). 13C NMR (CD3CN)
d 143.2, 130.0, 130.0, 128.7, 68.4,
pendent. Surprisingly,
3 and 7 disproportionate rather than
22.4. TLC on silica gel eluting with 20% EtOAc in hexane: Rf: 9t 0.67,
9c 0.39.
recombine at lower temperatures. 2,6-Di-tert-butylbenzoquinone
(DTBQ) scavenges two radicals 3 to afford thermally labile cyclo-
hexa-2,5-dienone 13. The trapping reaction of DTBQ is found to be
reversible, most likely diminishing its inhibition efficiency at ele-
vated temperatures.
10.5. Thermolysis of 9c
Toluene-d8 solutions of 9c were degassed, sealed into NMR
tubes, and thermolyzed at 30.2 ꢂC (0.022 M 9c), 37.8 ꢂC (0.02 M 2c),
and 46.2 ꢂC (0.0114 M 9c). The disappearance of 9c was monitored
by 1H NMR of its methine protons using TMSTMS as internal
standard. The rate constants are shown in Table 2.
10. Experimental section
10.1. Materials
All reagents and NMR solvents (Cambridge Isotope Laboratories)
were used as supplied unless otherwise noted. HPLC solvents were
HPLC grade and were sonicated before use. Aqueous solutions were
made with laboratory deionized water.
10.6. Purification and analysis of 9c
The cis isomer was purified by HPLC using a C18 column with
30% H2Oþ70% CH3CN and a flow rate of 0.7 mL/min. Retention
times: 9c 5.15 min, 9t 8.84 min.
10.2. General methods
10.7. N,N-(1,1-Dimethylethyl-1)-(1-diethylphosphono-2,2-
Melting points (uncorrected) were obtained on a Mel-Temp
apparatus. NMR spectra were recorded on a Bruker 250, 400 MHz
dimethylpropyl-1)-nitroxyl (SG1)
or 500 MHz spectrometer. Chemical shifts (
d
, ppm) are based on
SG1 was prepared as a red oil in 35% yield according to the lit-
erature.111 The 1H NMR spectrum of SG1 was obtained on SG1-H by
d (CDCl3) 4.02 (m,
4H), 3.16 (d, J¼20.0 Hz, 1H), 1.23 (m, 6H), 1.09 (s, 9H), 1.03 (s, 9H).
TMS, hexamethyldisilane (TMSTMS) or solvent signal (C6D5CD3,
CD3CN) as reference. UV/visible spectra were run on a Hewlett
Packard 8452A diode array spectrometer. HPLC analyses and pu-
rifications were conducted with a Thermo Separation Products
Spectra System P2000 HPLC system equipped with a Hewlett
Packard 1050A UV detector interfaced to a PC. The normal phase
adding a few drops of phenylhydrazine. 1H NMR
10.8. 2-Methyl-3-phenyl-but-2-yl hydroperoxide112
CN and reversed phase C18 columns (5
m
m, 250 mmꢀ4.6 mm)
To a solution of 100 mL 50% hydrogen peroxide and 10 mL 85%
H3PO4 was carefully added 2.5 g 2-methyl-3-phenyl-butan-2-ol.113
The mixture was stirred at 55 ꢂC for 24 h. When the starting ma-
terial disappeared according to NMR, hexane was added. The or-
ganic layer was separated and washed with 10% NaHCO3, then with
water. NaOH (40%) was added to the solution to precipitate the
hydroperoxide as its sodium salt. The precipitate was filtered off
and acidified with dilute HCl. After extraction into CH2Cl2, the hy-
droperoxide was isolated by evaporating the solvent under vac-
were purchased from Alltech. The solvents were either H2O/
CH3CN or 2-propanol/hexane. The flow rate was 0.7 mL/min and
the UV/visible absorbance was monitored simultaneously at 266,
313, 330, 360 and 450 nm. GC analyses were carried on a Hewlett
Packard 5890 instrument with FID detectors and a J & W scientific
DB-5 capillary column (0.25
m
m, 30 mꢀ0.25 mm). This GC was
interfaced to a PC running HP Chemstation software. The usual GC
conditions were: injector 230 ꢂC, detector 250 ꢂC, initial column
temperature 35 ꢂC for 5 min, program at 10 ꢂC/min to 250 ꢂC and
hold for 10 min.
uum. 1H NMR
d
7.8 (s, 1H), 7.05 (s, 5H), 3.13 (q, J¼7.0 Hz, 1H), 1.28 (d,
J¼7.0 Hz, 3H), 1.13 (s, 3H), 1.12 (s, 3H).
Samples for thermolysis or photolysis were prepared in
7 mmꢀ300 Pyrex tubes or 5 mm medium wall NMR tubes and were
freeze/thaw degassed three times at ꢁ78 ꢂC and sealed on a high
vacuum line. The tubes were immersed completely in a DC-200
silicone oil bath contained in a 1.5 gallon Dewar flask with me-
chanical stirrer. The temperature was regulated precisely by a Bay-
ley Model 123 temperature controller and was measured with
a Hewlett Packard Model 3456A digital voltmeter and a platinum
thermometer. An Oriel 500 W high pressure mercury lamp fitted
with a Bausch and Lomb high intensity grating monochromator
was employed in photolysis work.
10.9. 2-Methyl-3-phenyl-2-butyl per(2-phenylpropionate) 11
2-Methyl-3-phenyl-but-2-yl hydroperoxide (0.18 g, 1 mmol)
was dissolved in 1 mL pentane. KOH (40%, 1 equiv) was added
dropwise and the mixture was stirred for 30 min. Most of the
pentane was carefully aspirated away from the precipitate and the
residual solid was evacuated to dryness. The solid was dissolved in
1 mL CH2Cl2 and the solution was cooled to ꢁ2 ꢂC. After dropwise
addition of 0.15 mL 2-phenyl-propionyl chloride, the solution was
stirred at ꢁ2 ꢂC for 3 h. The CH2Cl2 solution was separated from the

P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
8813
white solid, washed three times with cold 10% aq NaOH, and dried
10.14. N-(1-Phenethyloxy)-9-azabicyclo[3.3.1]nonane 18
over MgSO4. Evaporation of the solvent gave an oil whose NMR
spectrum showed that it contained unreacted hydroperoxide,
2-phenyl-propionyl chloride, and perester 11. The starting mate-
rials were removed by dissolving the oil in 2 mL cold pentane and
cooling in an ice bath. An excess of freshly pentane-washed NaH
was added and the mixture was shaken for 5 min. The solid was
removed by filtration and the solvent was evaporated under vac-
uum to leave the diastereomers of 11 (125.3 mg, 40%). Obtaining
a completely pure sample of 11 was impossible because of its
thermal lability, the fact that CDCl3 induces its decomposition, and
An authentic sample of nitroxyl trapping product 18 was pre-
pared by heating a benzene solution of 0.122 M 9t and 0.0714 M 9-
azabicyclo[3.3.1]nonane-N-oxyl (ABNO) at 120 ꢂC for 3 h in a sealed
tube. GC analysis of the resulting mixture showed dimers 10 and
a peak at 23.6 min, whose GC/MS/CI spectrum exhibited m/z¼246
(base) and fragments at 142, 124, and 105. Preparative TLC on silica
gel with 5% Et2O/hexane afforded 3.5 mg (20%) of 18 (Rf 0.33). 1H
NMR (250 MHz, CDCl3)
d
7.23e7.39 (m, 5H), 4.71 (q, J¼6.5 Hz, 1H),
3.21 (br s, 1H), 2.97 (br s, 1H), 2.33 (m, 2H), 1.80 (m, 8H), 1.44 (d,
because it is a mixture of diastereomers. 1H NMR
d
7.1e7.4 (m, 10H),
J¼6.5 Hz, 3H),1.26 (2H m). 13C NMR (62.5 MHz CDCl3)
d 144.8,127.9,
3.82 (q, J¼7.2 Hz, 2H), 2.94e3.07 (two q, J¼7.2 Hz, 1H), 1.62 (t,
J¼7.0 Hz, 6H), 1.36 (d, J¼7.0 Hz, 3H), 1.29 (d, J¼7.0 Hz, 3H), 1.17 (s,
6H), 1.10 (s, 3H), 1.05 (s, 3H).
126.9, 126.5, 78.7, 54.8, 54.2, 23.2, 22.0, 20.0, 19.9. HRMS calcd for
(C16H24NOþH): 246.1858, found 246.1868.
10.15. Bis(4-benzyloxyphenyl)oxalate 22
10.10. Competition kinetics experiments with BQ, SG1, and
a
-phenethyl radicals
The oxalate ester was made by treating 4-benzyloxyphenol with
oxalyl chloride according to the procedure of Lahti et al.82 The
white, solid product was recrystallized from CH3CN. Mp
A toluene solution containing 0.035 M BQ, 0.37 M SG1, and
159e160 ꢂC. 1H NMR (CDCl3)
d
7.33 (m, 10H), 7.11 (d, J¼9.2 Hz, 4H),
0.27 M 9c was degassed three times, sealed in four 7 mm Pyrex
tubes, and heated at 43 ꢂC. The disappearance of BQ and SG1 was
monitored at 450 nm by HPLC using a silica gel column (2-prop-
anol/hexane¼5:95) after every 15 min of thermolysis.
6.95 (d, J¼9.2 Hz, 4H), 5.01 (s, 4H). 13C NMR
d 157.4, 156.3, 143.8,
136.8, 128.9,128.4, 127.7,122.0, 115.9, 70.7. The exact mass of 22 was
determined on a Thermo Electron Corp. LTQ Orbitrap mass spec-
trometer. NaHCO3 used in the workup left Na in the sample. Calcd
for C28H22O6Na: 477.13141, found 477.13092.
10.11. The reaction of DTBQ with 9t
A solution of 0.01 M DTBQ and 0.03 M 9t in benzene was
degassed, sealed in an UV cell and heated at 120 ꢂC for 3 h. UV
showed no DTBQ absorbance but GC revealed DTBQ, attributed to
thermolysis of an intermediate on the GC injector. After separation
by column chromatography (silica gel, 1% EtOAc in hexane), a mix-
ture of two diastereomers 13A,B was isolated. 1H NMR (C7D8): di-
Acknowledgements
We thank the Robert A. Welch Foundation for continued support
(Grant No. C-0499) and GE Betz, The Woodlands, TX for financing
the early stages of this work. Helpful comments from Dr. Sherif
Eldin of GE Betz are especially appreciated.
astereomer 1:
d
6.90e7.50 (m, 5H), 6.56 (d, J¼3.0 Hz, 1H), 5.99 (d,
J¼3.0 Hz, 1H), 4.35 (q, J¼6.6 Hz, 1H), 3.08 (q, J¼7.2 Hz, 1H), 1.39 (s,
References and notes
9H), 1.38 (d, J¼6.6 Hz, 3H), 1.30 (d, J¼7.2 Hz, 3H), 0.969 (s, 9H);
diastereomer 2:
d
7.02e7.60 (m, 5H), 6.42 (d, J¼3.0 Hz, 1H), 6.16 (d,
1. Inouye, H.; Leistner, E. In Chemistry of Quinonoid Compounds; Patai, S., Rappo-
port, Z., Eds.; Interscience: London, New York, NY, 1988; Vol. 2, pp 1293e1349.
2. Asche, C. Mini-Rev. Med. Chem. 2005, 5, 449e467.
3. Wieland, H. Justus Liebigs Ann. Chem. 1935, 514, 145e181.
4. Betterton, A. J.; Waters, W. A. J. Chem. Soc. 1953, 329e331.
5. Breitenbach, J. W.; Springer, A.; Horeischy, K. Chem. Ber. 1938, 71, 1438e1441.
6. Foord, S. G. J. Chem. Soc. 1940, 48e56.
J¼3.0 Hz, 1H), 4.35 (q, J¼6.6 Hz, 1H), 3.05 (q, J¼7.0 Hz, 1H), 1.39 (d,
J¼6.6 Hz, 3H), 1.34 (d, J¼7.0 Hz, 3H), 1.27 (s, 9H), 1.06 (s, 9H). 13C
NMR (C7D8) diastereomer 2: d 185.6, 149.9, 147.9, 147.3, 141.8, 141.8,
129.7, 128.8, 128.3, 128.1, 127.9, 127.5, 127.3, 126.4, 78.5, 74.2, 50.4,
35.3, 35.1, 29.8, 29.2, 29.2, 26.5, 15.9. HRMS calcd for (C30H39O2þH):
431.29505, found 431.29377.
7. Cohen, S. G. J. Am. Chem. Soc. 1947, 69, 1057e1064.
8. Yassin, A. A.; El-Reedy, A. M. Eur. Polym. J. 1973, 9, 657e667.
9. Bickel, A. F.; Waters, W. A. J. Chem. Soc. 1950, 1764e1769.
10. Aparacio, F. J. L.; Waters, W. A. J. Chem. Soc. 1952, 4666e4674.
11. Gleixner, G.; Breitenbach, J. W.; Olaj, O. F. Makromol. Chem. 1978, 179, 73e77.
12. Golubev, V. B.; Mun, G. A.; Zubov, V. P. Russ. J. Phys. Chem. 1986, 60, 347e350.
13. Yassin, A. A.; Rizk, N. A. Eur. Polym. J. 1976, 12, 393e399.
14. Rembaum, A.; Szwarc, M. J. Am. Chem. Soc. 1955, 77, 4468e4472.
15. Breitenbach, J. W.; Gleixner, G.; Olaj, O. F. Makromol. Chem., Suppl. 1975, 1,
177e190.
10.12. Thermolysis of 13
A 0.023 M solution of diastereomer 13B (d 6.15) in toluene-d8
containing TMSTMS as internal standard was degassed, sealed in an
NMR tube, and heated at 120 ꢂC. NMR showed the disappearance of
this diastereomer and formation of the other diastereomer 13A (
6.00). The final product also contained DTBQ and 2,6-di-tert-butyl-
4-[1-phenylethoxy] phenol 15.
d
16. Citterio, A.; Arnoldi, A.; Minisci, F. J. Org. Chem. 1979, 44, 2674e2682.
17. Roginskii, V. A.; Belyakov, V. A. Dokl. Akad. Nauk SSSR 1977, 237, 1404e1407.
18. Maeda, Y.; Ingold, K. U. J. Am. Chem. Soc. 1979, 101, 4975e4981.
19. Shlyapnikova, I. A.; Roginskii, V. A.; Miller, V. B. Russ. Chem. Bull. 1978,
2215e2219.
20. Simandi, T. L.; Tudos, F. Eur. Polym. J. 1985, 21, 865e869.
21. Velikov, A.; Matisova-Rychla, L.; Broska, R.; Rychly, J. J. Therm. Anal. Calorim.
1999, 57, 473e486.
22. Fujisawa, S.; Kadoma, Y.; Yokoe, I. Chem. Phys. Lipids 2004, 130, 189e195.
23. Hageman, H. J. Eur. Polym. J. 2000, 36, 345e350.
10.13. 2,6-Di-tert-butyl-4-[1-phenethoxy] phenol 15
A solution of 0.5 mmol 13 and 300 mg thiophenol in 0.5 mL
toluene-d8 was degassed, sealed in an NMR tube, and heated at
120 ꢂC for 2 days. NMR was used to monitor the reaction products.
The solution was washed with saturated aq NaHCO3 and water. The
solvent was evaporated and the residue was separated by reverse
phase HPLC (C18, CH3CN/water¼90:10) to isolate pure 15. This
phenol was very easily oxidized by air and should be kept in
24. Roginskii, V. A.; Dubinskii, V. Z.; Miller, V. B. Izv. Akad. Nauk SSSR, Ser. Khim.
1982, 2808e2812.
25. Jacobsen, N.; Sharma, S. C.; Torssell, K. Acta Chem. Scand. B 1979, 33, 499e502.
26. Sharma, S. C.; Torssell, K. Acta Chem. Scand. B 1978, 32, 347e353.
27. Barton, D. H. R.; Bridon, D.; Zard, S. Z. Tetrahedron 1987, 43, 5307e5314.
28. Braslau, R.; Burrill, L. C.; Siano, M.; Naik, N.; Howden, R. K.; Mahal, L. K.
Macromolecules 1997, 30, 6445e6450.
a sealed tube. 1H NMR (C6D6)
d 7.2 (m, 5H), 7.12 (s, 2H), 5.18 (q,
29. Engel, P. S. Chem. Rev. 1980, 80, 99e150.
J¼7.0 Hz, 1H), 4.59 (s, 1H), 1.56 (d, J¼7.0 Hz, 3H), 1.42 (s, 18H). 13C
30. Daub, G. H.; Cannizzo, L. F. J. Org. Chem. 1982, 47, 5034e5035.
31. Cohen, S. G.; Groszos, S. J.; Sparrow, D. B. J. Am. Chem. Soc. 1950, 72, 3947e3951.
32. Gilbert, A. In CRC Handbook of Organic Photochemistry and Photobiology, 2nd
ed.; Horspool, W., Lenci, F., Eds.; CRC: Boca Raton, FL, 2004.
NMR (C6D6)
d 152.1, 148.2, 144.8, 137.3, 129.0, 127.6, 126.2, 113.4,
76.9, 34.8, 30.5, 25.1.

8814
P.S. Engel et al. / Tetrahedron 66 (2010) 8805e8814
33. Horspool, W. M. Photochemistry (2004); Royal Society of Chemistry: London,
2006; Vol. 35, pp 17e46.
72. Urano, S.; Matsuo, M. Lipids 1976, 11, 380e383.
73. Hart, H.; DeVrieze, J. D. Tetrahedron Lett. 1968, 4257e4260.
74. Mahoney, L. R.; Mendenhall, G. D.; Ingold, K. U. J. Am. Chem. Soc. 1973, 95,
8610e8614.
75. Ciriano, M. V.; Korth, H.-G.; van Scheppingen, W. B.; Mulder, P. J. Am. Chem.
Soc. 1999, 121, 6375e6381.
76. Keana, J. F. W.; Dinerstein, R. J.; Baitis, F. J. Org. Chem. 1971, 36, 209e211.
77. Marque, S.; Le Mercier, C.; Tordo, P.; Fischer, H. Macromolecules 2000, 33,
4403e4410.
34. Engel, P. S.; Bishop, D. J. J. Am. Chem. Soc. 1975, 97, 6754e6762.
35. Chae, W.-K.; Baughman, S. A.; Engel, P. S.; Bruch, M.; Ozmeral, C.; Szilagyi, S.;
Timberlake, J. W. J. Am. Chem. Soc. 1981, 103, 4824e4833.
36. Corley, R. C.; Gibian, M. J. J. Org. Chem. 1972, 37, 2910e2911.
37. Lee, W. K.; Chung, C. T. Bull. Korean Chem. Soc. 1992, 13, 694e695.
38. Engel, P. S.; Wang, C.; Chen, Y.; Rüchardt, C.; Beckhaus, H.-D. J. Am. Chem. Soc.
1993, 115, 65e74.
39. Ioffe, B. V.; Stopskii, V. S. Zh. Org. Khim. 1968, 4, 1501e1510.
40. Schmittel, M.; Rüchardt, C. J. Am. Chem. Soc. 1987, 109, 2750e2759.
41. Engel, P. S.; Fogel, L. D.; Steel, C. J. Am. Chem. Soc. 1974, 96, 327e332.
42. Skinner, K. J.; Hochster, H. S.; McBride, J. M. J. Am. Chem. Soc. 1974, 96,
4301e4306.
78. Skene, W. G.; Belt, S. T.; Connolly, T. J.; Hahn, P.; Scaiano, J. C. Macromolecules
1998, 31, 9103e9105.
79. Johnston, L. J.; Tencer, M.; Scaiano, J. C. J. Org. Chem. 1986, 51, 2806e2808.
80. Kuhn, R.; Hammer, I. Chem. Ber. 1950, 83, 413e414.
81. Altwicker, A. R. Chem. Rev. 1967, 67, 475e531.
43. Engel, P. S.; Wu, H.; Smith, W. B. Org. Lett. 2001, 3, 3145e3148.
44. Schulz, A.; Rüchardt, C. Tetrahedron Lett. 1976, 3883e3886.
45. Frank, J.; Rabinowitch, E. Trans. Faraday Soc. 1934, 30, 120e130.
46. Koenig, T.; Fischer, H. In Free Radicals; Kochi, J. K., Ed.; Wiley: New York, NY,
1973; Vol. 1, p 157.
82. Lahti, P. M.; Modarelli, D. A.; Rossitto, F. C.; Inceli, A. L.; Ichimura, A. S.; Ivatury,
S. J. Org. Chem. 1996, 61, 1730e1738.
83. Smith, W. B. Texas Christian University.
84. Hrovat, D. A.; Borden, W. T. J. Phys. Chem. 1994, 98, 10460e10464.
85. Denisov, E. T. Kinet. Catal. 2006, 47, 662e671.
47. Greene, F. D.; Berwick, M. A.; Stowell, J. C. J. Am. Chem. Soc. 1970, 92, 867e873.
48. Seltzer, S.; Hamiliton, E. J. J. Am. Chem. Soc. 1966, 88, 3775e3781.
49. Hammond, G. S.; Fox, J. R. J. Am. Chem. Soc. 1964, 86, 1918e1921.
50. Engel, P. S.; Bishop, D. J.; Page, M. A. J. Am. Chem. Soc. 1978, 100, 7009e7017.
51. Nelsen, S. F.; Bartlett, P. D. J. Am. Chem. Soc. 1966, 88, 143e149.
52. Nakamura, T.; Busfield, W. K.; Jenkins, I. D.; Rizzardo, E.; Thang, S. H.; Suyama,
S. J. Org. Chem. 2000, 65, 16e23.
53. Rüchardt, C.; Grundmeier, M. Chem. Ber. 1975, 108, 2448e2464.
54. Volod’kin, A. A.; Ershov, V. V. Russ. Chem. Rev. 1988, 57, 336e349.
55. Koenig, T. In Free Radicals; Kochi, J. K., Ed.; Wiley: New York, NY, 1973; Vol. 1,
p 113.
56. Sawaki, Y. In Organic Peroxides; Ando, W., Ed.; Wiley: New York, NY, 1993;
pp 427e477.
57. Greene, F. D. J. Am. Chem. Soc. 1955, 77, 4869e4873.
58. Newcomb, M. Tetrahedron 1993, 49, 1151e1176.
59. Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc. 1992, 114, 4992e4996.
60. Engel, P. S.; He, S.-L.; Banks, J. T.; Ingold, K. U.; Lusztyk, J. J. Org. Chem. 1997, 62,
1210e1214.
61. Sobek, J.; Martschke, R.; Fischer, H. J. Am. Chem. Soc. 2001, 123, 2849e2857.
62. Kumli, E.; Montermini, F.; Renaud, P. Org. Lett. 2006, 8, 5861e5864.
63. Silverstein, R. M.; Webster, F. X.; Kiemle, D. J. Spectrometric Identification of
Organic Compounds; Wiley: Hoboken, NJ, 2005.
64. Wieland, H.; Heymann, K.; Tsatsas, T.; Juchum, D.; Varvoglis, G.; Labriola, G.;
Dobbelstein, O.; Boyd-Barrett, H. S. Justus Liebigs Ann. Chem. 1934, 514,
145e181.
65. Yassin, A. A. Makromol. Chem. 1975, 176, 2571e2578.
66. Yassin, A. A.; Rizk, N. A.; Ghanem, N. A. Makromol. Chem. 1973, 164, 105e113.
67. Matyjaszewski, K.; Woodworth, B. E.; Zhang, X.; Gaynor, S. G.; Metzner, Z.
Macromolecules 1998, 31, 5955e5957.
68. Sciannamea, V.; Jerome, R.; Detrembleur, C. Chem. Rev. 2008, 108, 1104e1126.
69. McCarroll, A. J.; Crayston, J. A.; Walton, J. C. Tetrahedron 2003, 59, 4275e4280.
70. Miller, B. J. Am. Chem. Soc. 1965, 87, 5115e5120.
86. Brocks, J. J.; Beckhaus, H.-D.; Beckwith, A. L. J.; Rüchardt, C. J. Org. Chem. 1998,
1935e1943.
87. Brandrup, J.; Immegut, E. H. Polymer Handbook, 3rd ed.; Wiley Interscience:
New York, NY, 1989.
88. Fischer, H.; Radom, L. Angew. Chem., Int. Ed. 2001, 40, 1340e1371.
89. Lalevee, J.; Allonas, X.; Fouassier, J.-P. J. Org. Chem. 2005, 70, 814e819.
90. Zytowski, T.; Fischer, H. J. Am. Chem. Soc. 1997, 119, 12869e12878.
91. Engel, P. S.; Duan, S.; Arhancet, G. B. J. Org. Chem. 1997, 62, 3537e3541.
92. Citterio, A. Tetrahedron Lett. 1978, 2701e2704.
93. Citterio, A. Gazz. Chim. Ital. 1980, 110, 253e258.
94. Jacobsen, N.; Torssell, K. Acta Chem. Scand. 1973, 27, 3211e3216.
95. Giordano, C.; Belli, A.; Citterio, A.; Minisci, F. J. Org. Chem. 1979, 44, 2314e2315.
96. Barton, D. H. R.; Sas, W. Tetrahedron 1990, 46, 3419e3430.
97. Yassin, A. A.; Rizk, N. A.; Ghanem, N. A. Makromol. Chem. 1973, 168, 59e76.
98. Matsui, M.; Kondoh, S.; Shibata, K.; Muramatsu, H. Bull. Chem. Soc. Jpn. 1995,
68, 1042e1051.
99. Hawthorne, M. F.; Reintjes, M. J. Am. Chem. Soc. 1965, 87, 4585e4587.
100. Tashiro, M.; Fukata, G.; Yoshiya, H. Synthesis 1979, 988e989.
101. Becker, H.-D. J. Org. Chem. 1965, 30, 982e989.
102. Williams, D. J.; Kreilick, R. J. Am. Chem. Soc. 1968, 90, 2775e2779.
103. Omura, K. J. Org. Chem. 1991, 56, 921e927.
104. Lucarini, M.; Mugnaini, V.; Pedulli, G. F.; Guerra, M. J. Am. Chem. Soc. 2003, 125,
8318e8329.
105. Lucarini, M.; Pedulli, G. F. J. Org. Chem. 1994, 59, 5063e5070.
106. Mahoney, L. R.; Weiner, S. A. J. Am. Chem. Soc. 1972, 94, 585e590.
107. Joschek, J.-I.; Miller, S. I. J. Am. Chem. Soc. 1966, 88, 3273e3281.
108. Yassin, A. A.; Rizk, N. A. Eur. Polym. J. 1977, 13, 441e446.
109. Gibian, M. J.; Corley, R. C. Chem. Rev. 1973, 73, 441e464.
110. Benson, S. W.; DeMore, W. B. Annual Review of Physical Chemistry; Annual
Reviews: Palo Alto, CA, 1965; Vol. 16, pp 397e450.
111. Benoit, D.; Chaplinski, V.; Braslau, R.; Hawker, C. J. J. Am. Chem. Soc. 1999, 121,
3904e3920.
71. Wenkert, E.; Wovkulich, P. M.; Pellicciari, R.; Ceccherelli, P. J. Org. Chem. 1977,
42, 1105e1107.
112. Kropf, H.; Wallis, H. V. Synthesis 1981, 633e635.
113. Greene, F. D. J. Am. Chem. Soc. 1959, 81, 2688e2691.