Homolytic dissociation of 1-substituted cyclohexa-2,5-diene-1-carboxylic acids: An EPR spectroscopic study of chain propagation

Article abstract of DOI:10.1039/b104859g

Hydrogen abstraction from 1-substituted cyclohexa-2,5-diene-1-carboxylic acids containing linear, branched and cyclic alkyl substituents, as well as allyl, propargyl (prop-2-ynyl), cyanomethyl and benzyl substituents, has been studied by EPR spectroscopy. For each carboxylic acid, EPR spectra of the corresponding cyclohexadienyl radicals were observed at lower temperatures, followed by spectra due to ejected carbon-centred radicals at higher temperatures. Rate constants, for release of the carbon-centred radicals from the cyclohexadienyl radicals, were determined from radical concentration measurements for the above range of substituents. The rate of cyclohexadienyl radical dissociation increased with branching in the 1-alkyl substituent and with electron delocalisation in the ejected carbon-centred radical; 3,5-and 2,6-dimethyl-substitution of the cyclohexadienyl ring led to reductions in the dissociation rate constants. Rate data for abstraction of bisallylic hydrogens from the cyclohexadienyl acids were also obtained for ethyl, n-propyl and isopropyl radicals. These results indicated a sharp drop in the rate of hydrogen abstraction as the degree of branching in the attacking radical increased. Small decreases in the hydrogen abstraction rate constants were observed for cyclohexadienes containing CO₂R substituents.

Full text of DOI:10.1039/b104859g

View Article Online / Journal Homepage / Table of Contents for this issue
Homolytic dissociation of 1-substituted cyclohexa-2,5-diene-1-
carboxylic acids: an EPR spectroscopic study of chain
propagation†
2
Leon V. Jackson and John C. Walton
University of St. Andrews, School of Chemistry, St. Andrews, Fife, UK KY16 9ST.
E-mail: jcw@st-and.ac.uk; Fax: 44 (0)1334 463808; Tel: 44 (0)1334 463864
Received (in Cambridge, UK) 4th June 2001, Accepted 27th June 2001
First published as an Advance Article on the web 8th August 2001
Hydrogen abstraction from 1-substituted cyclohexa-2,5-diene-1-carboxylic acids containing linear, branched
and cyclic alkyl substituents, as well as allyl, propargyl (prop-2-ynyl), cyanomethyl and benzyl substituents, has
been studied by EPR spectroscopy. For each carboxylic acid, EPR spectra of the corresponding cyclohexadienyl
radicals were observed at lower temperatures, followed by spectra due to ejected carbon-centred radicals at
higher temperatures. Rate constants, for release of the carbon-centred radicals from the cyclohexadienyl radicals,
were determined from radical concentration measurements for the above range of substituents. The rate of
cyclohexadienyl radical dissociation increased with branching in the 1-alkyl substituent and with electron
delocalisation in the ejected carbon-centred radical; 3,5- and 2,6-dimethyl-substitution of the cyclohexadienyl
ring led to reductions in the dissociation rate constants. Rate data for abstraction of bisallylic hydrogens from the
cyclohexadienyl acids were also obtained for ethyl, n-propyl and isopropyl radicals. These results indicated a sharp
drop in the rate of hydrogen abstraction as the degree of branching in the attacking radical increased. Small decreases
in the hydrogen abstraction rate constants were observed for cyclohexadienes containing CO2R substituents.
Introduction
Results and discussion
1-Substituted cyclohexa-2,5-diene-1-carboxylic acids 1 are a
new “pro-aromatic” class of potentially ‘clean’ reagents for free
radical generation. Hydrogen abstraction takes place selectively
from the bisallylic sites of 1 to generate delocalised cyclo-
hexadienyl radicals 2 that dissociate by C–C bond scission at
moderate temperatures to produce benzoic acid and an alkyl
radical. The released alkyl radical may cyclise, if it is suitably
unsaturated, or react with an added alkene (Z), and hence be
Birch reduction–alkylation of benzoic acid affords 1-alkyl-
cyclohexa-2,5-dienecarboxylic acids of type 1 in moderate to
good yields.4–9 The 1-alkyl acids 1a–l were obtained in various
yields by means of a modification of the procedure due origin-
ally to Birch.4 The method was most efficient for primary and
secondary alkyl substituents and particularly with benzyl chlor-
ide, allyl and propargyl bromides. Alkylations with cyano-
methyl iodide, bromide and chloride were all examined in the
synthesis of the cyanomethyl acid 1k; yields, however, were
moderate in all cases. Attempts to prepare an alkoxycarbonyl-
methyl acid by similar methodology using esters of haloacetic
acids were not successful. Alkylation following Birch reduction
in liquid ammonia was much more difficult with sterically
hindered organohalides and, as described previously,2 the
pure tert-butyl acid 1h could only be obtained in low yield.
1-Isopropyl- and 1-isobutyl-3,5-dimethylcyclohexa-2,5-diene-1-
carboxylic acids 5a,b were prepared in a similar way starting
with 3,5-dimethylbenzoic acid. The 1-n-propyl-2,6-dimethyl-
cyclohexa-2,5-diene-1-carboxylic acid 6 was made from the cor-
responding substituted benzoic acid. 1-Benzyl-2,3,4,5,6-penta-
deuteriocyclohexa-2,5-diene-1-carboxylic acid 8 was obtained
by use of pentadeuteriobenzoic acid.
ؒ
transformed to a new C-centred radical RZ that can abstract
hydrogen from more 1. This constitutes a chain process that
has potential synthetic applications.1,2 The advantages of these
carboxylic acids over organotin reagents, apart from the safety
aspect, are that benzoic acid, which can easily be removed by an
alkaline extraction, is a comparatively benign co-product, and
that the H-transfer step is slower. Respectable yields of alkyl-
ated olefins were obtained for secondary, tertiary and delocal-
ised alkyl radicals;2 however, end product analyses showed that
competition from an alternative β-scission of radical 2 to afford
the hydroxyformyl radical 3 (formate radical) and an alkyl-
benzene (RPh) was significant for primary substituents R.
Kinetic information on the hydrogen transfer (kH) and dis-
sociation (kd) steps would be very desirable as a tool for effect-
ive synthetic planning. A preliminary communication outlined
the EPR spectroscopic method we developed for determining
the rate constants for the chain propagation steps.3 We now
report our study of the effect of differing substituents R, and of
alkyl substituents in the ring, on the kinetics of the dissociation
and on the hydrogen transfer steps.
Kinetic EPR spectroscopic study of the dissociation of
1-substituted cyclohexadienyl radicals (2)
Photolyses of solutions of 1 (R = n-alkyl) and di-tert-butyl per-
oxide (DTBP) in tert-butylbenzene solvent, without any addi-
tives, gave rise to t-BuOH, benzoic acid and RH as the main
products, accompanied by minor amounts of alkylbenzenes
RPh from the undesired fragmentation. The mechanism and
propagation steps under these conditions were as shown in
Scheme 1, except that no alkene (Z) was present, and chain
† Electronic supplementary information (ESI) available: measured
radical concentrations and the corresponding rate constants at each
temperature for each cyclohexadienyl acid. Computed heats of form-
ation and selected geometric parameters are also available for each
cyclohexadienyl radical and the species they dissociate to. For direct
electronic access see http://www.rsc.org/suppdata/p2/b1/b104859g/
ؒ
termination was by bimolecular reactions of 2 and R (2kt).
Using the steady state approximation eqn. (1) may be derived,
1758
J. Chem. Soc., Perkin Trans. 2, 2001, 1758–1764
DOI: 10.1039/b104859g
This journal is © The Royal Society of Chemistry 2001

View Article Online
Table 1 9GHz EPR parameters for 1-substituted cyclohexadienyl-1-carboxylic acid radicals 2 in solutiona
Structure No.
R
T /K
g-factorb
a(H4)
a(H2,6
)
a(H3,5
)
a(Hother
)
2a
2b
2c
2d
2f
Me
Et
n-Pr
n-Bu
i-Pr
c-C5H11
t-Bu
Allyl
CH2CCH
CH2CN
CH2Ph
Ph
220c
295d
270d
245d
220c
215c
145e
160e
170f
245g
150f
220c
240c
290d
340d
—
2.0026
—
—
2.0027
—
2.0027
2.0026
2.0026
—
—
—
13.2
13.1
13.2
13.2
13.3
13.7
13.3
13.2
13.3
13.3
13.1
13.4
13.4
13.1
12.8
9.2
9.1
9.1
9.1
9.2
9.2
9.2
9.2
9.2
9.2
9.1
9.3
8.8
8.7
—
2.7
2.7
2.7
2.6
2.8
2.8
2.9
2.7
2.7
2.7
2.6
2.7
—
—
0.30(2H)
0.32(4H)
—
—
—
—
—
1.3(1H)
—
1.3(1H)
—
2g
2h
2i
2j
2k
2l
2m
ؒ
ؒ
5a(-H)
i-Pr
i-Bu
n-Pr
2.0027
—
—
2.5(6H)
2.4(6H)
8.7(6H)
5b(-H)
—
2.6
ؒ
6(-H)
a Hfs in G (10 G = 1.0 mT). b Non-quoted g-factors = 2.003 0.001. c Solvent PhBu-t. d Solvent neat DTBP. e Solvent cyclopropane ϩ ca. 20% t-
BuPh. f Solvent cyclopropane. g Solvent t-BuPh ϩ trace MeOH.
1-substituent. In a few cases small, long-range hfs from hydro-
gen atoms of the 1-substituents were resolved and, interestingly,
in the case of the low temperature spectra of the propargyl- 2j
and benzyl-substituted radicals 2l, only one of the two available
methylene hydrogens coupled with the unpaired electron. At
these low temperatures the radicals adopted a preferred
conformation about the ring–CH2 bond (9) in which only one
H-atom was favourably placed for interaction with the SOMO.
At higher temperatures the rate of internal rotation about this
bond became fast on the EPR timescale and the lines broad-
ened. Although the spectral lines were sharper at still higher
temperatures, the average hfs of the now equivalent methylene
hydrogens were too small to resolve.
For each carboxylic acid, EPR spectra were recorded over an
appropriate range of temperatures to test for alkyl radical
production. For acids 1a and 1m only the corresponding
cyclohexadienyl radical 2 could be observed up to 360 K and, as
ؒ
ؒ
expected, neither CH3 nor Ph radical production was
observed. For all the other acids, however, the spectrum due to
radical 2 weakened in some temperature range and was grad-
ؒ
ually replaced by that of the corresponding alkyl radical R . On
lowering the temperature, the inverse process was observed, and
the spectrum of 2 was restored. The EPR spectral parameters
of the released alkyl and substituted-alkyl radicals were all
essentially identical to those given in the literature. These obser-
vations indicated that the chain propagation rate constants
might be determined from EPR spectral measurements of the
steady state radical concentrations by use of eqns. (1) or (2).
Scheme 1
provided that the alternative β-scission to the hydroxyformyl
radical (3) [k2 ] is negligible. For low concentrations of acids 1,
d
or when kH is small, eqn. (1) simplifies to the usual expression
[eqn. (2)].10
The importance of the alternative, undesired β-scission (k2 )
d
1
2
ؒ
ؒ
ؒ
k /2kt ϩ kH/2kt{[1][R ]/[2]} = [R ] /[2] ϩ [R ]
(1)
(2)
producing the hydroxyformyl radical 3, was appraised from the
relative yields of alkylbenzene and benzoic acid determined by
GC and NMR. The measured [PhCO2H]/[RPh] ratios, from
photolyses in DTBP at 300 K, were found to be ca. 20 and ca.
13 for 1b and 1c respectively, and no alkylbenzenes could be
detected during photolyses of 1f–m. It follows that diversion of
cyclohexadienyl radicals 2 down this reaction channel can
safely be neglected.
d
1
2
ؒ
ؒ
k /2kt = [R ] /[2] ϩ [R ]
d
When solutions of individual acids 1a–m, 5–8 and DTBP in
an appropriate solvent were photolysed in the resonant cavity
of an EPR spectrometer the corresponding cyclohexadienyl
radicals 2 were generated by hydrogen abstraction by tert-
butoxyl radicals and observed over specific temperature ranges.
The hyperfine splittings (hfs) and g-factors for each radical are
recorded in Table 1.
ؒ
The concentrations of 2 and R were determined by the EPR
method12 from photolyses of known concentrations of 1b,c and
1f with DTBP, either neat or in tert-butylbenzene, directly in the
ؒ
These hfs were similar to those of other cyclohexadienyl
EPR resonant cavity. From measurements of [R ] and [2] at two
radicals11 and were not very sensitive to the nature of the
(or more) different acid concentrations [1] k1 /2kt and kH/2kt
d
J. Chem. Soc., Perkin Trans. 2, 2001, 1758–1764
1759

View Article Online
Table 2 Kinetic data for release of C-centred radicals (k1 , E 1d) from cyclohexadienyl radicals 2
d
Acid
Delocalised radical
Released radical
k1 /103 sϪ1 (300 K)
log Ad/sϪ1
aE 1d/kcal molϪ1
bE 1d/kcal molϪ1
d
ؒ
1a
1b
1c
1f
1g
1h
1i
2a
2b
2c
2f
Me
Et
≤0.001
0.040
0.022
0.96
—
16.9
13.00
9.97
11.24
12.19
12.71
11.22
10.05
9.88
—
≥18
ؒ
21.3
16.01
10.14
11.21
8.85
8.45
10.14
9.92
6.90
13.61
16.2
15.0
ؒ
n-Pr
16.00
13.75
13.55
9.57
ؒ
i-Pr
ؒ
2g
2h
2i
c-C5H11
t-Bu
Allyl
HCCCH2
1.34
ؒ
1070
4450
22.6
1.54
1290
0.28
8.72
ؒ
1j
2j
11.87
13.47
9.46
14.48
14.11
ؒ
1k
8
5a
5b
2k
NCCH2
2lc
PhCH2
ؒ
ؒ
ؒ
ؒ
5a(-H)
5b(-H)
i-Pr
12.30
14.5
ؒ
Me2CHCH2
0.53
a 1 cal = 4.18 J. b Activation energies derived assuming all log (Ad/sϪ1) values = 13.0 sϪ1 c Pentadeuteriated isotopomer.
.
Fig. 1 Arrhenius plots of dissociation k1 and H-abstraction kH rate
d
constants from reactions of ethyl (1b), n-propyl (1c) and isopropyl (1f)
acids. Filled circles; kH for n-propyl radical abstracting from 1c. Open
circles; kH for ethyl radical abstracting from 1b. Crossed circles; kH for
isopropyl radical abstracting from 1f. Open squares; raw kd data for 4
mg acid sample. Crossed squares; raw kd data for 0.3 mg acid sample.
Filled squares; kd corrected according to eqn. (1).
Fig. 2 9 GHz EPR spectra of radicals obtained on H-abstraction from
3,5-dimethyl-1-isopropylcyclohexa-2,5-dienecarboxylic acid (5a) in
tert-butylbenzene solution. The upper spectrum at 250 K shows mainly
the 3,5-dimethylcyclohexadienyl radical and the lower spectrum at 290
K shows mainly the ejected isopropyl radical (central region).
were obtained. Satisfactory solutions were obtained in this way
ؒ
ؒ
for the two primary radicals Et and n-Pr (Table 2). Fig. 1
shows Arrhenius plots of the rate constants obtained for the
n-propyl-substituted acid 1c. The open squares and crossed
squares show the raw data for two concentrations of 1c plotted
according to eqn. (2) and the filled squares show the corrected
k1d data after use of eqn. (1) (see below).
was complex and overlapped too extensively with the
spectrum of the cyclohexadienyl radical 2l for concentration
measurements to be carried out successfully. To try and over-
come this problem 1-(4-tert-butylbenzyl)cyclohexa-2,5-diene-1-
carboxylic acid (7) was prepared. It was anticipated that the
simpler hyperfine splitting pattern of the 4-tert-butylbenzyl
radical would enable concentration measurements to be made.
However, it was found that this acid was too poorly soluble in
EPR compatible solvents for satisfactory spectra to be
obtained. Instead, therefore, 1-benzyl-2,3,4,5,6-pentadeuterio-
cyclohexa-2,5-diene-1-carboxylic acid (8) was prepared and
used. In this case the spectrum of the released benzyl radical
was clear but that of the deuteriated cyclohexadienyl radical
consisted of a single broad peak with no resolved fine structure.
This spectrum overlapped with the normal, broad, central
background signal present in control solutions, which had to be
subtracted from the composite spectrum at each temperature.
This correction procedure degraded the data to some extent
and hence the kinetic data for release of benzyl are less accur-
For the other acids 1g–l containing tertiary or stabilised rad-
ؒ
icals the [R ]/[2] ratios were effectively independent of the initial
acid concentration, to within the limits of the experimental
ؒ
measurements. The rate constants for H-abstraction by t-Bu ,
allyl, propargyl, cyanomethyl and benzyl radicals from 1 were
therefore too small for measurement by this technique. How-
ever, the simplified eqn. (2) could then be used, in conjunction
with the radical concentrations, to measure the dissociation rate
constants k1 . For the acids 1g–j and 5a,b, radical generation
d
was smooth and the absolute radical concentrations were read-
ily obtained by double integration of suitable peaks from the
two radicals. Fig. 2 shows an example of the quality of the data
obtained from the carboxylic acid 5a.
The rate constants obtained by use of eqn. (1), or eqn. (2) in
conjunction with the well established 2kt value of Fischer and
co-workers,13 corrected for changes in solvent viscosity as
described previously,14 are given in Table 2 for 300 K.
ate. A combined Arrhenius plot of the k1 values from all the
d
The cyanomethyl carboxylic acid 1k was poorly soluble in
tert-butylbenzene and therefore traces (10 µl) of methanol were
added to improve spectral quality. In the case of the 1-benzyl
carboxylic acid 1l the spectrum of the released benzyl radical
acids spans a 200 K temperature range and nearly 5 orders of
magnitude in kd (Fig. 3) and demonstrates the good calibre
of the data (see Supplementary Data for a complete list of rate
constants at each temperature studied).† The measured
1760
J. Chem. Soc., Perkin Trans. 2, 2001, 1758–1764

View Article Online
Fig. 3 Arrhenius plots of dissociation rate constants k1d for the range
of 1-substituted (1-carboxylato)cyclohexadienyl radicals. Lines have
been constrained to intersect the y-axis at log A = 13.0. Crossed circles;
allyl-substituted radical 2i. Open squares; tert-butyl-substituted radical
2h. Stars; benzyl-substituted radical from acid 8. Triangles apex-up;
propargyl-substituted radical 2j. Diamonds top and bottom filled;
cyanomethyl-substituted radical 2k. Triangles apex-down; cyclopentyl-
substituted radical 2g. Crossed squares; isopropyl-substituted radical
2f. Open diamonds; isobutyl radical from acid 5b. Open circles;
isopropyl radical from acid 5a. Squares filled left; ethyl-substituted
radical 2b. Squares filled right; n-propyl-substituted radical 2c.
Fig.
4
Plot of the experimental activation energies for the
1
cyclohexadienyl dissociations bEd against the dissociation enthalpies
computed using the AM1 method.
optimised with respect to all geometric variables. The RHF and
UHF options were used for closed shell stable molecules and
open shell radicals respectively (heats of formation of indi-
vidual species, together with selected structural parameters, are
given in the Supplementary Data).† The overall dissociation
enthalpies are listed in Table 3. The rings of radicals 2 were all
computed (AM1) to be planar with ЄC2–C1–C6 in the range
112.7 0.7Њ, r(C1–C(O)OH) in the range 1.522 0.002 Å and
ЄR–C1–C(O)OH in the range 108 1.5Њ, i.e. structures were
insensitive to the nature of the substituent R. The exception to
this was the R–C1 distance, which was longest at 1.565 for
R = t-Bu, dropping to 1.556 for R = i-Pr, to 1.544 for R = Et
and to 1.535 Å for R = Me. For all other substituents (R) this
distance was computed to be 1.546 0.002 Å. The PM3 com-
putations showed a similar trend. The lengthening of this bond
with increasing size of the attached substituent gave a clear
signal that frontal strain in the radicals was important.
Arrhenius pre-exponential factors covered a considerable range
(Table 2). The temperature ranges of individual experiments
were quite short (40 10 K) and hence accurate A-factors
could not be obtained with this technique because of the long
extrapolations involved. Deviations from the ‘normal’ value for
unimolecular processes [log (A/sϪ1) = 13] were not gross, hence
it is probable that the true A-factors are close to this for all the
dissociations. The final column of Table 2 shows the activation
energies that correspond to assumed log (A/sϪ1) factors of 13.
The experimental k1 values confirmed that the ease of
d
fragmentation of the cyclohexadienyl radicals 2 increased
dramatically with the degree of branching of the released alkyl
ؒ
radical. In fact, release of t-Bu was nearly 5 orders of magni-
tude faster than primary alkyl radical production at 300 K.
Similarly, comparison of the data for dissociation of 2 to allyl
or benzyl radicals, with that for dissociation to primary
radicals, indicated that electron delocalisation in the released
radical also led to sharply increased dissociation rates. Both the
thermodynamic stabilisation of the released alkyl radical and
steric strain in the initial cyclohexadienyl radical appeared to be
important in controlling the fragmentation rates. 3,5-Dimethyl
substitution in the cyclohexadienyl ring caused a reduction in
The AM1 computed reaction enthalpies ∆Ho (Table 3) for
Ϫ1
ؒ
dissociation to R and PhCO2H reduced from 33.4 kcal mol
endothermic for R = Me, with increasing branching of the sub-
stituent, to 7.2 kcal molϪ1 exothermic for R = t-Bu. The com-
ؒ
puted ∆Ho values for the alternative dissociation to CO2H and
RPh were all endothermic and, apart from the larger value for
the R = Me case, were comparatively insensitive to the nature of
R. In agreement with experiment, the AM1 computed ∆Ho
values showed that, on thermodynamic grounds, dissociation
of 2 to R and PhCO2H should be favoured for all substituents
(except Me). The results of the PM3 computations corre-
sponded less well with experiment, i.e. they implied that
dissociation to CO2H and PhR should be favoured for
R = primary alkyl, as well as Me, contrary to observation.
ؒ
kd of about a factor of 3 for dissociation to i-Pr . The effect of
ؒ
2,6-dimethyl substitution was also investigated by means of the
n-propyl-2,6-dimethyl acid 6. For the unsubstituted acid (1c)
dissociation of radical 2c took place, under EPR conditions, in
the temperature range 340–375 K. Surprisingly, however, for 6
only the 2,6-dimethylcyclohexadienyl radical could be observed
over this whole temperature range, i.e. dissociation required
higher temperatures, outside the accessible range. It follows that
kd for this acid radical is significantly smaller. In the case of
2,6-dimethyl substitution some steric assistance to dissociation
might have been expected. However, the experimental result
shows that this must have been minor and was outweighed by
the additional stabilisation of the cyclohexadienyl radical
due to methyl substitution. This latter effect would be expected
to increase the activation energy of the dissociation step, as
observed.
ؒ
A plot of the AM1 computed ∆Ho values, against the
experimental activation energies (bEd ), is shown in Fig. 4. The
1
data divide into two groups with points for primary and
delocalised radicals, all of type RCH2 falling on one line,
ؒ
and points for branched radicals on another line. Steric effects
are expected to be small for all the radicals initially attached to
the cyclohexadienyl ring by CH2 groups and hence the lower
line expresses the reduction in activation energy associated with
increasing resonance stabilisation of the ejected radical. For
this group of radicals, unit reduction in computed dissociation
enthalpy corresponded to about 0.3 unit reduction in activation
energy. It is noteworthy that the largest deviation from this
straight line is for the benzyl case (radical 2l) for which the
measured activation energy had the largest error limits (see
above).
To probe the controlling influences underlying these dissoci-
ations, the structures and energies of a range of substituted
cyclohexadienyl radicals 2, and the products of both their
fragmentation modes, were computed using the AM1 and PM3
semi-empirical SCF MO methods.15,16 All structures were fully
J. Chem. Soc., Perkin Trans. 2, 2001, 1758–1764
1761

View Article Online
Table 3 Computed dissociation enthalpies (∆Ho) for substituted cyclohexadienyl radicals 2a
∆Ho (AM1)→
∆Ho (AM1)→
∆Ho (PM3)→
∆Ho (PM3)→
ؒ
ؒ
ؒ
ؒ
Cyclohexadienyl
R
R ϩ PhCO2H
CO2H ϩ PhR
R ϩ PhCO2H
CO2H ϩ PhR
2a
2b
2c
2e
2f
2g
2h
2i
Me
Et
n-Pr
i-Bu
i-Pr
c-C5H9
t-Bu
33.4
11.1
12.0
9.0
1.8
3.5
Ϫ7.2
1.1
4.9
8.8
Ϫ0.6
Ϫ4.5
29.0
15.8
15.8
14.4
14.6
15.5
13.7
17.1
15.1
14.5
22.8
17.3
19.0
9.2
9.1
8.9
0.4
Ϫ0.1
Ϫ8.9
0.4
5.6
7.8
7.6
7.6
7.7
4.3
5.5
5.2
4.3
7.0
7.2
6.5
Allyl
2j
CH2CCH
CH2CN
CH2Ph
Pentadienyl
2k
2l
2
Ϫ1.7
Ϫ4.9
10.4
7.1
a ∆Ho values in kcal molϪ1
.
1
2
ؒ
Table 4 Kinetic data for H-abstraction from 3- R,3- R-substituted cyclohexa-2,5-dienes by radicals R in solution
1R
2R
kH/105 sϪ1 (T /K)
log (AH/sϪ1
)
EH/kcal molϪ1
Ref. No.
ؒ
R
ؒ
Me
H
H
H
H
Et
1.3 (300)
0.58 (300)
0.14 (340)
0.4 (340)
2.3 (323)
0.8 (413)
0.2 (421)
5 (323)
0.003 (272)
0.094 (300)
79 (298)
0.3 (300)
540 (295)
9.1
—
8.4
8.7
—
—
—
—
7.9
—
—
—
—
5.5
—
6.6
6.6
—
—
—
—
6.6
—
—
—
—
17
17
twc
twc
18
21
1
18
twc
17
20
17
19
ؒ
Et
ؒ
Et
HO2C
HO2C
H
Me
HO2C
H
HO2C
H
H
ؒ
n-Pr
n-Pr
H
Hex-5-enyl
Hex-5-enyl
Hex-5-enyl
O2CRa
Ra
H
b
ؒ
C6H11
ؒ
i-Pr
i-Pr
H
H
H
H
ؒ
t-Bu
Cyclopropyl
CCl3
t-BuO
ؒ
H
H
ؒ
a R = hex-5-enyl. b 2,2-Dimethylbut-3-en-1-yl. c tw = this work.
Increased branching in R will cause increased frontier strain
in the initial cyclohexadienyl radical 2. In addition, however,
α-methyl substitution in R will also contribute to stabilisation
of the released radicals due to hyperconjugation. The upper line
expresses a combination of the two effects and shows that unit
reduction in computed dissociation enthalpy corresponds to
ca. 0.2 unit reduction in activation energy. These straight lines
define Evans–Polanyi relationships (E = α∆Ho ϩ C) having
α values of 0.57 (lower line) and 0.32 (upper line) (Fig. 4). Many
types of radical reactions obey such relationships. The observed
Experiments with the ethyl acid (1b) showed that a 10-fold vari-
ation in the concentration of the acid (0.2 to 0.02 mol dmϪ3
)
ؒ
caused about a 3-fold variation in {[1b][Et ]/[2b]}. Plots of
2
ؒ
ؒ
ؒ
{[Et ] /[2b] ϩ [Et ]} against {[1b][Et ]/[2b]} for four different
acid concentrations, in the above range, were naturally rather
scattered, but satisfactory gradients and intercepts were
obtained at a series of temperatures (see Supplementary Data,†
and Fig. 1 for an Arrhenius plot of kH) and the kinetic param-
eters derived in this way are given in Tables 2 and 4. For the
n-propyl acid 1c, measurements were made for two acid concen-
trations and the resulting Arrhenius plot for kH is also shown in
ؒ
α value, midway between 0 and 1 for the RCH2 radicals, is
ؒ
indicative of a transition state roughly halfway between react-
ants and products. The lower α value for branched radical
expulsion implies an earlier, more reactant-like, transition state.
This is entirely consistent with the longer R–C1 bonds com-
puted for these radicals and with the participation of frontier
strain. The AM1 computed ∆Ho value for the dissociation of
the 1-(penta-2,4-dienyl)cyclohexa-2,5-diene-1-carboxylic acid
radical was Ϫ4.5 kcal molϪ1 (Table 3), hence, by use of the
above empirical Evans–Polanyi relationship, a very low activ-
ation energy of 6.4 kcal molϪ1 is predicted. The large extent of
Fig. 1. For dissociation of radical 2f to the i-Pr radical the
effect of changing the concentration of 1f was small because of
the smaller rate of H-abstraction. From experiments with 1f
concentrations of 0.013 and 0.13 mol dmϪ3, meaningful results
could only be obtained for three temperatures in the middle of
the dissociation range (Fig. 1) and the error limits on these
kinetic parameters are consequently high. No effect on the
radical concentrations was observed for different initial concen-
trations of acid 1e, indicating that, as expected, kH is even lower
for the tert-butyl radical.
resonance stabilisation present in the ejected pentadienyl rad-
Rate constants (kH) were derived from eqn. (1), in conjunc-
tion with the 2kt values of Fischer and co-workers,13 and are
compared in Table 4 with literature data for a range of radicals
abstracting from cyclohexa-1,4-diene17–20 and various deriv-
atives.1,21 The estimated log(AH) values were normal for bi-
molecular reactions of this type but, as indicated above, the
Arrhenius parameters were not very reliable because of the
limited temperature ranges, and therefore the rate constants are
of most significance. In comparing the different kH values it
should be noted that cyclohexa-1,4-diene has 4 bisallylic hydro-
gens in comparison with the 2 bisallylic hydrogens of the 1,1-
disubstituted analogues and hence, for comparison purposes,
the kH values for the former should be reduced by a statistical
Ϫ1
1
ical leads to the prediction of very rapid dissociation [kd /s
(calc.) = 2 × 108 at 300 K].
Hydrogen abstraction from substituted cyclohexa-2,5-dienes
The concentrations of the initial radicals 2b and 2c and the
ejected n-alkyl radicals were significantly affected by major
changes in the initial concentrations of 1b and 1c (see above
and Fig. 1) and hence kH values could be determined for H-
ؒ
ؒ
abstraction by Et and n-Pr radicals. Reference to eqn. (1) indi-
2
ؒ
ؒ
ؒ
cates that plots of {[R ] /[2] ϩ [R ]} against {[1][R ]/[2]} should
be linear having kH/2kt as gradient and k1 /2kt as intercept.
d
1762
J. Chem. Soc., Perkin Trans. 2, 2001, 1758–1764

View Article Online
ؒ
factor of 2. In accord with expectation, the t-BuO radical
eum ether (PE) refers to the fraction boiling between 40 and
abstracts most rapidly, followed by the σ-type cyclopropyl
60 ЊC. Other organic compounds were used as received.
Column chromatography was performed using BDH silica gel
(40– 63 µm).
radical. The C-centred alkyl radicals show a large reduction in
ؒ
ؒ
ؒ
ؒ
ؒ
kH along the series Me > Et ∼ n-Pr > i-Pr > t-Bu . This is
readily explicable because the enthalpies of the hydrogen
abstraction reactions also reduce in parallel. No major effect on
kH due to 1,1-disubstitution would be expected because the
substituents are comparatively remote from the site of
H-abstraction and attached to the ring at sp3 hybridised
C-atoms. However, the kH values for the substituted acids 1 are
somewhat smaller, even after taking account of the statistical
1-Methylcyclohexa-2,5-diene-1-carboxylic acid 1a22 and 1-
cyclopentylcyclohexa-2,5-diene-1-carboxylic acid 1g2 were pre-
pared as described in the literature. 1-Isopropylcyclohexa-2,5-
diene-1-carboxylic acid 1f was available from previous work.1
General procedure for preparation of 1-alkylcyclohexa-2,5-diene-
1-carboxylic acids
ؒ
Ammonia (300 cm3) was added to benzoic acid (5 g, 41 mmol)
with careful stirring. To this, Li (0.8 g, 0.115 mol) was added
portionwise until a permanent blue colour persisted, followed
by dropwise addition of the alkyl halide (0.117 mol). The reac-
tion mixture was left for 1 h whilst the NH3 evaporated and ice
was added to the remaining solid followed by dilute H2SO4. The
product was extracted with ether (3 × 150 cm3) and the com-
bined ethereal extracts were dried (MgSO4) and the solvent was
evaporated leaving a solid which was recrystallised in light
petroleum, or purified by flash chromatography.
factor. Compare, for example, the two kH values for Et or the
ؒ
ؒ
kH values for i-Pr and t-Bu (Table 4). This is probably not an
artifact of the different experimental techniques because the
product analysis “clock” method, for the primary hex-5-enyl
radical abstracting from 1-alkyl-1-CO2R-substituted cyclo-
hexadienyl rings,1,21 showed a similar reduction in kH compared
to unsubstituted analogues. We conclude, therefore, that this
type of substitution leads to a small reduction in the rate of
H-abstraction by C-centred radicals.
Conclusions
1-Ethylcyclohexa-2,5-diene-1-carboxylic acid (1b)7,23
The rate of hydrogen donation by 1 to branched radicals is
ؒ
comparatively slow and hence radicals RZ formed from, for
Prepared using ethyl iodide (20.8 g, 0.133 mol). Purification by
distillation on a Kugelröhr apparatus gave the title compound
as a colourless oil (4.32 g, 69%); δH 0.85 (3H, t, CH3), 1.74 (2H,
example, addition to 1,1-disubstituted alkenes, will not be able
to sustain chain reactions effectively. However, transformations
ؒ
of R which produce primary radicals [or O-centred, or vinyl],
q, CH ), 2.65 (2H, s, bisallylic-H), 5.67–5.98 (4H, m, C᎐CH),
᎐
2
for example 5-exo-cyclisations, should be well suited to this
8.80 (1H, br s, CO2H); δC 10.2 (CH3), 25.8 (CH2), 27.4 (CH2),
46.4 (C), 125.8–126.9 (4 × CH), 180.7 (CO).
methodology. The most useful cyclohexadienyl acids will con-
tain branched R (high k1 ) which are transformed to primary
d
1-n-Propylcyclohexa-2,5-diene-1-carboxylic acid (1c)23
radicals, vinyl radicals or O-centred radicals on cyclisation
ؒ
(larger kH). In no case will premature reduction of R to RH
Prepared from 1-iodopropane (20.9 g, 0.123 mol) in dry ether
(5 cm3). The product was purified by Kugelröhr distillation to
yield the title compound (4.47 g, 66%) as a white powder; mp
47 ЊC (lit.23 46–48.5 ЊC); δH 0.89 (3H, t), 1.28 (2H, m), 1.67 (2H,
m), 2.65 (2H, s), 5.69–5.98 (4H, m), 11.10 (1H, br s); δC 14.1
(CH3), 17.4 (CH2), 26.0 (CH2), 41.7 (CH2), 47.7 (C), 125.8,
be a problem because the kH values are nearly 2 orders of
magnitude less than for H-donation by organotin hydrides.10
Experimental
1H NMR spectra were recorded at 200 or 300 MHz and 13C
NMR spectra at 75 MHz, in CDCl3 solutions with tetramethyl-
silane (δH = δC = 0) as reference. Coupling constants are
expressed in Hz. EI mass spectra were obtained with 70 eV
ionisation and CI spectra were obtained with isobutane as
target gas on a VG Autospec spectrometer. GC-MS analyses
were run on a Finnigan Incos 50 quadrupole instrument
coupled to a Hewlett Packard HP 5890 chromatograph fitted
with a 25 m HP 17 capillary column (50% phenyl methyl sili-
cone). For the calculation of yields from GC data, the detector
response was calibrated with known amounts of authentic
materials (or close analogues). EPR spectra were obtained with
a Bruker EMX 10/12 spectrometer operating at 9.5 GHz with
100 kHz modulation. Samples of the substrate (0.3 to 40 mg)
and di-tert-butyl peroxide (0.01 to 0.5 cm3), or tert-butyl-
benzene (up to 0.5 cm3), in 4 mm od quartz tubes, were
de-aerated by bubbling nitrogen for 20 min, and photolysed in
the resonant cavity by light from a 500 W super pressure
mercury arc lamp. For reactions performed in cyclopropane,
the solution was degassed on a vacuum line using the freeze–
pump–thaw technique, and the tube was flame sealed. In all
cases where spectra were obtained, hfs were assigned with the
aid of computer simulations using the Bruker Simfonia soft-
ware package. For kinetic measurements, carboxylic acid
samples were used in ‘single shot’ experiments, i.e. new samples
were prepared for each temperature and each acid concen-
tration, to minimise sample depletion effects. Signals were
double integrated using the Bruker WinEPR software and
radical concentrations were calculated by reference to a known
concentration of DPPH, as described previously.14
126.7 (4 × CH), 181.6 (C᎐O).
᎐
1-n-Butylcyclohexa-2,5-diene-1-carboxylic acid (1d)
Prepared using 1-iodobutane (22.04 g, 0.122 mol). The product
was successfully purified by chromatography [SiO2, light
petroleum–EtOAc (4 : 1)] to give the carboxylic acid 1d (1.74 g,
24%) as oily, colourless crystals; mp 59 ЊC; δH 0.94 (3H, t, J 6,
CH3), 1.26 (4H, m, 2 × CH2), 1.75 (2H, t, J 4, CH2C), 2.68 (2H,
s, allylic-H), 5.75–5.98 (4H, m, olefinic-H); δC 14.0 (CH3), 19.6–
26.4 (CH2), 26.1 (CH2), 39.3 (C), 126.0–126.8 (4 × CH), 180.5
(C᎐O).
᎐
1-tert-Butylcyclohexa-2,5-diene-1-carboxylic acid (1h)2
Prepared using tert-butyl iodide (30.01 g, 0.164 mol). The
product was chromatographed [SiO2, light petroleum–EtOAc
(4 : 1)] to give 1h (3.2 g, 22%) as colourless needles; mp 100–
102 ЊC (lit.2 101 ЊC); δH 1.00 (9H, s, C(CH3)3), 2.57–2.63 (2H, s,
allylic-H), 5.88–6.10 (4H, m, olefinic-H); δC 26.0 (3 × CH3),
26.2 (CH ), 38.6 (C), 52.9 (C), 125.5, 126.2 (CH), 180.4 (C᎐O).
᎐
2
1-Allylcyclohexa-2,5-diene-1-carboxylic acid (1i)26
Prepared from allyl bromide (14.9 g, 0.123 mol) in dry ether
(5 cm3). The product was purified by Kugelröhr distillation to
yield the title compound (4.28 g, 64%) as a colourless oil;
δH 2.46 (2H, d), 2.65 (2H, s), 5.09 (2H, d), 5.66 (1H, m), 5.69–
5.98 (4H, m), 11.19 (1H, br s); δC 26.0 (CH2), 44.0 (CH2), 47.5
(C), 118.3 (CH2), 126.1, 126.3 (4 × CH), 132.7 (CH), 180.8
(C᎐O).
᎐
Ether refers to diethyl ether. THF and ether were distilled
under nitrogen from sodium benzophenone ketyl prior to use.
Where dry DCM was used, it was distilled over CaH2. Petrol-
1-(Prop-2-ynyl)cyclohexa-2,5-diene-1-carboxylic acid (1j)
Prepared using propargyl bromide (26.45 g, 0.124 mol). The
J. Chem. Soc., Perkin Trans. 2, 2001, 1758–1764
1763

View Article Online
product was distilled under reduced pressure to yield the title
compound as a white solid; mp 74 ЊC; δH 2.0–2.04 (1H, s,
acetylenic-H), 2.57–2.63 (2H, s, allylic-H), 2.69–2.72 (2H, s,
CH2CCH), 5.81–6.02 (4H, m, olefinic-H); δC 26.3 (CH2), 30.0
(CH2), 47.3 (C, ring), 70.9 (CH), 79.7 (C, chain), 125.5, 127.2,
dry ether (5 cm3). The product was purified by recrystallisation
from pentane to yield the title compound (3.65 g, 33%); mp
271 ЊC; δH 1.25 (9H, s, tert-butyl-H), 2.14–2.40 (2H, m, allylic-
H), 2.88 (2H, s, benzylic-H), 5.46–5.83 (4H, m, olefinic-H),
6.90–7.12 (4H, m, aromatic-H).
128.5, 129.2 (CH), 178.9 (C᎐O) (Found: MHϩ 163.0766,
᎐
1-Benzyl-2,3,4,5,6-pentadeuteriocyclohexa-2,5-diene-1-carb-
oxylic acid (8)
C10H11O2 requires MH 163.0759).
1-(Cyanomethyl)cyclohexa-2,5-diene-1-carboxylic acid (1k)
Prepared from perdeuteriobenzoic acid (5 g, 39 mmol) and
benzyl chloride (15.62 g, 0.123 mol) in dry ether (5 cm3). The
product was purified by recrystallisation from pentane to yield
the title compound (4.81 g, 56%) as a white solid; mp 76–77 ЊC;
δH 2.32, 2.51 (2H, 2 × s, allylic-H), 3.06 (2H, br s, CH2), 6.8–7.8
(5H, m, Ar-H), 10.34 (1H, br s, CO2H) (Found: MHϩ 220.1390,
C14H10D5O2 requires MH 220.1386).
Prepared from iodoacetonitrile (10.2 g, 0.061 mol) in dry ether
(20 cm3). The product was purified with activated charcoal and
recrystallised from pentane to yield the title compound (2.93 g,
58%) as white plates; mp 110 ЊC; δH 2.76 (2H, s), 2.81 (2H, s),
5.72–6.17 (4H, m); δC 26.1 (CH2), 28.3 (CH2), 45.8 (C), 116.5
(CN), 123.4, 129.4 (CH), 177.3 (C᎐O); m/z (%) 163 (Mϩ, 6), 123
᎐
(100), 122 (29), 117 (35), 115 (47), 105 (27), 91 (35), 79 (66), 77
(55), 51 (19) (Found: Mϩ 163.0638, C9H9O2N requires M
163.0633). This compound was also prepared from bromo- and
chloro-acetonitrile, in similar yields.
Acknowledgements
We thank the EPSRC (grant GR/L49185) and NATO for finan-
cial support of this research.
1-Benzylcyclohexa-2,5-diene-1-carboxylic acid (1l)24,25
Prepared from benzoic acid (8 g, 82 mmol) and benzyl chloride
(29.2 g, 0.233 mol). The product was purified by recrystallis-
ation from pentane to yield 1l (7.24 g, 41%) as white needles;
mp 76–77 ЊC (lit.24 76–77 ЊC); δH 2.27–2.63 (2H, m, allylic-H),
3.03 (2H, s, benzylic-H), 5.80–5.90 (4H, m, olefinic-H), 7.11–
7.29 (5H, m, Ar-H); δC 25.0 (allylic-C), 46.1 (CH2), 48.8 (C),
References
1 P. A. Baguley, G. Binmore, A. Milne and J. C. Walton, Chem.
Commun., 1996, 2199.
2 P. A. Baguley and J. C. Walton, J. Chem. Soc., Perkin Trans. 1, 1998,
2073.
3 L. V. Jackson and J. C. Walton, Tetrahedron Lett., 1999, 40, 7019.
4 A. J. Birch, J. Chem. Soc., 1950, 2, 1551; A. J. Birch and J. Slobbe,
Tetrahedron Lett., 1975, 9, 627.
126.5, 126.7, 127.9, 130.7, 136.1 (CH), 179.7 (C᎐O).
᎐
5 R. B. Woodward, Pure Appl. Chem., 1968, 17, 519.
6 E. W. Kaiser, Synthesis, 1972, 391.
7 I. K. Zhurkovich and D. V. Ioffe, J. Org. Chem. USSR, 1974, 10, 216.
8 D. J. Hart and C.-P. Chuang, J. Org. Chem., 1983, 48, 1782.
9 R. J. Hamilton, L. N. Mander and S. P. Sethi, Tetrahedron, 1986, 42,
2881.
10 J. C. Walton, J. Chem. Soc., Perkin Trans. 2, 1998, 603, and
references cited therein.
11 A. Berndt, in Landölt-Bornstein; Magnetic Properties of Free
Radicals, eds. H. Fischer and K. Hellwege, Springer, Berlin, 1977,
vol. 9b, p. 452; A. Berndt, in Landölt-Bornstein; Magnetic Properties
of Free Radicals, eds. H. Fischer and K. Hellwege, Springer, Berlin,
1987, vol. 17c, p. 88.
1-Isopropyl-3,5-dimethylcyclohexa-2,5-diene-1-carboxylic acid
(5a)
Prepared from 3,5-dimethylbenzoic acid (5.0 g, 33 mmol) and
2-iodopropane (18.38 g, 0.10 mol). The product was purified by
recrystallisation from pentane, yielding 5a (4.59 g, 71%) as
white needles; mp 94–95 ЊC; δH 0.79–0.82 (6H, d, J 6, 2 × CH3),
1.78 (6H, s, 2 × ring CH3), 2.0–2.23 (1H, septet, J 6, CHMe2),
2.48 (2H, s, allylic-H), 5.45 (2H, s, olefinic-H); δC 17.3 (CH3),
22.9 (ring CH3), 36.1 (CHMe2), 36.3 (allylic-C), 54.2 (C), 119.4
(CH), 134.2 (C), 181.6 (C᎐O) (Found: MHϩ 195.1389, C H O
᎐
12 19
2
requires MH 195.1385).
12 D. Griller and K. U. Ingold, Acc. Chem. Res., 1980, 13, 193;
D. Griller and K. U. Ingold, Acc. Chem. Res., 1980, 13, 317.
13 H. Schuh and H. Fischer, Helv. Chim. Acta, 1978, 61, 2130.
14 B. Maillard and J. C. Walton, J. Chem. Soc., Perkin Trans. 2, 1985,
443.
1-Isobutyl-3,5-dimethylcyclohexa-2,5-diene-1-carboxylic acid
(5b)
Prepared with 1-bromo-2-methylpropane to give a white
solid which was purified by recrystallisation from pentane to
yield 5b (4.20 g, 61%) as white plates; mp 71 ЊC; δH 0.75–0.89
(6H, d, J 6, 2 × CH3), 1.5–1.63 (2H, d, J 6, CH2), 1.8 (6H, s, ring
CH3), 2.2 (1H, m, CHMe2), 2.48 (2H, s, allylic-H), 5.45 (2H, s,
olefinic-H); δC 17.3 (2 × CH3), 22.9 (ring CH3), 36.1 (CHMe2),
36.3 (allylic-C), 45.2 (CH2), 54.2 (C), 119.4 (CH), 134.2 (C),
15 M. J. S. Dewar, E. G. Zoebisch, E. F. Healy and J. J. P. Stewart,
J. Am. Chem. Soc., 1985, 107, 392; J. J. P. Stewart, QCPE No. 455,
University of Indiana, 1987.
16 J. J. P. Stewart, J. Comput. Chem., 1989, 10, 209; J. J. P. Stewart,
J. Comput. Chem., 1989, 10, 221; J. J. P. Stewart, J. Computer-Aided
Mol. Design, 1990, 4, 1; J. J. P. Stewart, J. Comput. Chem., 1990,
11, 543; J. J. P. Stewart, in Reviews of Computational Chemistry,
eds. K. Lipkovitz and D. B. Boyd, VCM, New York, 1990.
17 J. A. Hawari, P. S. Engel and D. Griller, Int. J. Chem. Kinet., 1985,
17, 1215.
18 M. Newcomb and S. U. Park, J. Am. Chem. Soc., 1986, 108, 4132.
19 H. Paul, R. D. Small and J. C. Scaiano, J. Am. Chem. Soc., 1978,
100, 4520; A. Effio, D. Griller, K. U. Ingold, J. C. Scaiano and
S. J. Sheng, J. Am. Chem. Soc., 1980, 102, 6063.
181.6 (C᎐O) (Found: MHϩ 209.1549, C13H21O2 requires MH
᎐
209.1542).
1-n-Propyl-2,6-dimethylcyclohexa-2,5-diene-1-carboxylic acid
(6)
20 L. J. Johnston, J. C. Scaiano and K. U. Ingold, J. Am. Chem. Soc.,
1984, 106, 4877.
21 G. Binmore, J. C. Walton and L. Cardellini, J. Chem. Soc., Chem.
Commun., 1995, 27.
22 G. Binmore, L. Cardellini and J. C. Walton, J. Chem. Soc., Perkin
Trans. 2, 1997, 757.
23 H. Van Bekkum, C. B. Van den Bosch, G. Van Minnen-Pathius,
J. C. Mos and A. M. Van Wijk, Recl. Trav. Chim. Pays-Bas, 1971, 90,
137; H. Van Bekkum, Recl. Trav. Chim. Pays-Bas, 1973, 92, 379.
24 H. Plieninger and G. Ege, Chem. Ber., 1961, 94, 2095.
25 B. M. R. Bandara, A. J. Birch and W. D. Raverty, J. Chem. Soc.,
Perkin Trans. 1, 1982, 1755; B. M. R. Bandara, A. J. Birch and
W. D. Raverty, J. Chem. Soc., Perkin Trans. 1, 1982, 1763.
26 D. Bland, G. Chambournier, V. Dragan and D. J. Hart, Tetrahedron,
1999, 55, 8953.
Prepared from 1-iodopropane (17.0 g, 0.1 mol) and 2,6-
dimethylbenzoic acid (5 g, 33 mmol). The product was success-
fully purified by recrystallisation from pentane, to yield the title
compound as white plates (3.42 g, 53%); mp 96–97 ЊC; δH 0.91
(3H, t, J 6, CH3), 0.97–1.13 (2H, m, CH2), 1.70 (6H, s,
2 × CH3), 1.79–1.88 (2H, m, CH2), 2.72 (2H, s, allylic-H), 5.72
(2H, s, ᎐CH); δ 14.3 (CH3), 16.6 (CH2), 19.3 (2 × CH3), 27.1
᎐
C
(CH ), 32.5 (CH ), 55.2 (C), 123.5 (CH), 130.3 (C), 179.9 (C᎐O)
᎐
2
2
(Found: MHϩ 195.1382, C12H19O2 requires MH 195.1385).
1-(4-tert-Butylbenzyl)cyclohexa-2,5-diene-1-carboxylic acid (7)
Prepared from 4-tert-butylbenzyl bromide (24.9 g, 0.123 mol) in
1764
J. Chem. Soc., Perkin Trans. 2, 2001, 1758–1764