Thiol-catalysed radical-chain redox rearrangement reactions of benzylidene acetals derived from terpenoid diols

Article abstract of DOI:10.1039/b309060b

The thiol-catalysed radical-chain redox rearrangement of cyclic benzylidene acetals derived from 1,2- and 1,3-diols of terpene origin has been investigated from both synthetic and mechanistic standpoints. The redox rearrangement was carried out either at ca. 70°C (using Bu^tON=NOBu^t as initiator) or at ca. 130°C (using Bu^tOOBu^t as initiator) in the presence of triisopropylsilanethiol or methyl thioglycolate as catalyst; the silanethiol was usually more effective. This general reaction affords the benzoate ester of the monodeoxygenated diol, unless rearrangement of intermediate carbon-centred radicals takes place prior to final trapping by the thiol to give the product, in which case structurally rearranged esters are obtained. For the benzylidene acetals of 1,2-diols prepared by vicinal ds-dihydroxylation of 2-carene, α-pinene or β-pinene, intermediate cyclopropylcarbinyl or cyclobutylcarbinyl radicals are involved and ring opening of these leads ultimately to unsaturated monocyclic benzoates. 1,2-Migration of the benzoate group in the intermediate p-benzoyloxyalkyl radical sometimes also competes with thiol trapping during the redox rearrangement of benzylidene acetals derived from 1,2-diols. Redox rearrangement of the benzylidene acetal from carane-3,4-diol, obtained by cis-dihydroxylation of 3-carene, does not involve intermediate cyclopropylcarbinyl radicals and leads to benzoate ester in which the bicyclic carane skeleton is retained. The inefficient redox rearrangement of the relatively rigid benzylidene acetal from exo,exo-norbornane-2,3-diol is attributed to comparatively slow chain-propagating β-scission of the intermediate 2-phenyl-1,3-dioxolan-2-yl radical, probably caused by the development of adverse angle strain in the transition state for this cleavage. Similar angle strain effects are thought to influence the regioselectivities of redox rearrangement of bicyclic [4.4.0]benzylidene acetals resulting from selected 1,3-diols, themselves prepared by reduction of aldol adducts derived from reactions of aldehydes with the kinetic lithium enolates obtained from menthone and from isomenthone.

Full text of DOI:10.1039/b309060b

View Article Online / Journal Homepage / Table of Contents for this issue
Thiol-catalysed radical-chain redox rearrangement reactions of
benzylidene acetals derived from terpenoid diols
Hai-Shan Dang, Brian P. Roberts* and Derek A. Tocher†
Christopher Ingold Laboratories, Department of Chemistry, University College London,
20 Gordon Street, London, UK WC1H 0AJ
Received 30th July 2003, Accepted 5th September 2003
First published as an Advance Article on the web 2nd October 2003
The thiol-catalysed radical-chain redox rearrangement of cyclic benzylidene acetals derived from 1,2- and 1,3-diols
of terpene origin has been investigated from both synthetic and mechanistic standpoints. The redox rearrangement
was carried out either at ca. 70 ЊC (using ButON᎐NOBut as initiator) or at ca. 130 ЊC (using ButOOBut as initiator) in
᎐
the presence of triisopropylsilanethiol or methyl thioglycolate as catalyst; the silanethiol was usually more effective.
This general reaction affords the benzoate ester of the monodeoxygenated diol, unless rearrangement of intermediate
carbon-centred radicals takes place prior to final trapping by the thiol to give the product, in which case structurally
rearranged esters are obtained. For the benzylidene acetals of 1,2-diols prepared by vicinal cis-dihydroxylation of
2-carene, α-pinene or β-pinene, intermediate cyclopropylcarbinyl or cyclobutylcarbinyl radicals are involved and ring
opening of these leads ultimately to unsaturated monocyclic benzoates. 1,2-Migration of the benzoate group in the
intermediate β-benzoyloxyalkyl radical sometimes also competes with thiol trapping during the redox rearrangement
of benzylidene acetals derived from 1,2-diols. Redox rearrangement of the benzylidene acetal from carane-3,4-diol,
obtained by cis-dihydroxylation of 3-carene, does not involve intermediate cyclopropylcarbinyl radicals and leads to
benzoate ester in which the bicyclic carane skeleton is retained. The inefficient redox rearrangement of the relatively
rigid benzylidene acetal from exo,exo-norbornane-2,3-diol is attributed to comparatively slow chain-propagating
β-scission of the intermediate 2-phenyl-1,3-dioxolan-2-yl radical, probably caused by the development of adverse
angle strain in the transition state for this cleavage. Similar angle strain effects are thought to influence the
regioselectivities of redox rearrangement of bicyclic [4.4.0]benzylidene acetals resulting from selected 1,3-diols,
themselves prepared by reduction of aldol adducts derived from reactions of aldehydes with the kinetic lithium
enolates obtained from menthone and from isomenthone.
tion of the thiol catalyst is to mediate the overall abstraction of
the benzylic hydrogen atom from the parent 1,3-dioxolane 4 by
the nucleophilic alkyl radical 6 to give the dioxolanyl radical 5,
Introduction
We have reported previously that monodeoxygenation of 1,2-
and 1,3-diols 1 (n = 0 or 1) can be readily achieved through a
which is also nucleophilic. The direct transfer of a hydrogen
thiol-catalysed radical-chain redox rearrangement of the
atom between these two nucleophilic carbon-centred radicals
derived benzylidene acetals 2, to give benzoate esters of the type
is relatively inefficient because it suffers from adverse polar
3 (Scheme 1).1,2 The propagation cycle for the rearrangement 2
effects. However, because the thiyl radical is electrophilic, both
steps A and B in the thiol-catalysed cycle that replaces the direct
abstraction benefit from favourable charge transfer in the
3 is illustrated in Scheme 2 for the simple case of the benzyl-
idene acetal derived from 2-methylpropane-1,2-diol.1 The func-
respective transition states. The thiol fulfils the role of a protic
polarity-reversal catalyst.3
In general, the chemoselectivity of the redox rearrangement
is determined by the regioselectivity of the ring-opening β-
scission step (e.g. 5
6), in conjunction with any rearrange-
ment reactions that the product alkyl radical may undergo prior
to trapping by the thiol. Thus, redox rearrangement of 4 gives
exclusively isobutyl benzoate 7 and none of the tert-butyl ester,
because β-scission of 5 occurs selectively by cleavage of the
C(4)–O bond to give the tertiary alkyl radical 6. We have
recently reported an experimental and computational study
of the factors that influence regioselectivity in the β-scission of
substituted 1,3-dioxolan-2-yl and 1,3-dioxan-2-yl radicals and
shown that the relative ease of C–O bond cleavage does not
always follow the simplistic ‘homolytic’ order 3Њ–C–O > 2Њ–C–
O > 1Њ–C–O.4 In particular, for bicyclic 2-phenyl-1,3-dioxan-2-
yl radicals such as those derived from carbohydrate benzylidene
acetals, examples are found where 1Њ–C–O bonds cleave in
marked preference to 2Њ–C–O bonds.2,4,5
Scheme 1
Naturally-occurring terpenes provide ready access to a
variety of cyclopropyl- and cyclobutyl-carbinyl radicals that
are prone to ring-opening rearrangements6 and we reasoned
that thiol-catalysed monodeoxygenation of 1,2- and 1,3-diols
of terpene origin could be of both mechanistic and preparative
Scheme 2
† Correspondence concerning the X-ray crystallography should be
directed to this author.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4073

View Article Online
Table 1 Redox rearrangement of the benzylidene acetal 10 as a func-
tion of thiol concentration in hexane at 70 ЊCa
interest. In the present paper we describe the thiol-catalysed
radical-chain redox rearrangements of benzylidene acetals
derived from selected terpenoid diols.
Conversion
b
Yield 11 ϩ 14
Entry TIPST/mol%
[TIPST]/M
(%)
Yield 12 ϩ 13
Results and discussion
1
2
3
4
5
6
7
8
1
2
3
0.0056
0.0111
0.0167
0.0276
0.0549
0.0829
0.1087
0.2116
70
78
85
≥98
96
97
97
95
40.0
32.3
13.3
9.0
4.0
3.2
Apart from pinane-2,3-diol which was obtained commercially,
the 1,2-diols studied in this work were prepared by vicinal cis-
dihydroxylation of terpenoid alkenes by treatment with tetra-
decylammonium permanganate in a two-phase solvent system
consisting of dichloromethane and aqueous sodium hydroxide,
as described by Hazra et al.,7 or by the classical reaction with
KMnO4 in alkaline aqueous tert-butyl alcohol. Selected 1,3-
diols derived from menthone and isomenthone, by reduction
of aldol adducts obtained from the kinetic lithium enolates of
these ketones,8 were also investigated. All diols were converted
into their benzylidene acetals by acid-catalysed condensation
with benzaldehyde in refluxing benzene or toluene, with azeo-
tropic removal of the water produced.
Redox rearrangements were carried out using two general
methods;2,4 in hexane solvent (bath temp. 70 ЊC) using di-tert-
butyl hyponitrite9 (TBHN) as initiator (method A) or in octane
solvent (bath temp. 140 ЊC) using di-tert-butyl peroxide (DTBP)
as initiator (method B). With both procedures the catalyst was
usually triisopropylsilanethiol (Pri3SiSH),2,4,10 although methyl
thioglycolate (MeO2CCH2SH) was used in some experiments
for comparison. It was usual to make multiple additions of
TBHN but, because of its relatively slow decomposition at
ca. 130 ЊC, a single initial addition of DTBP was sufficient. The
results are discussed below under separate headings for each of
the parent diols.
5
10
15
20
40
2.6
1.6
a Bath temperature. The concentration of the acetal 10 was ca. 0.55 M;
TBHN (5 mol%) was added initially and again after 20, 40 and 60 min;
the total reaction time was 2.5 h. A single addition of thiol was made at
the start of the reaction. b Estimated by 1H NMR analysis of the crude
reaction product.
esters could not be separated by column chromatography.
Examination of the reaction mixture after partial conversion
of 10 showed2 that the more abundant isomer 10a was slightly
(ca. 1.7 times) more reactive than the minor isomer 10b towards
abstraction of the benzylic hydrogen atom by Pr 3SiS at
ca. 70 ЊC. Of course, the same benzylic radical 15 is formed by
abstraction of hydrogen from either isomer.
i
ؒ
Carane-2,3-diol
cis-Dihydroxylation of (1S)-(ϩ)-2-carene 8 afforded the diol 9
and thence the benzylidene acetal 10 as a 57 : 43 mixture of the
diastereoisomers 10a and 10b, the H NMR spectra of which
The routes by which compounds 11–13 arise from the inter-
mediate benzylic radical are set out in Scheme 3. The product
1
1
distribution was determined by H NMR spectroscopy as a
were assigned on the basis of selective nuclear Overhauser
effect (NOE) experiments. Thus, irradiation of the benzylidene
proton at δ 5.81 resulted in strong enhancement of the peaks
arising from the C(3)-methyl protons at δ 1.33 and from H(2) at
δ 3.97 (carane numbering), confirming that these signals are
associated with the more abundant isomer 10a in which the
phenyl group is trans to the methyl group on the dioxolane
ring. No corresponding enhancements were observed when the
benzylidene proton at δ 6.00 was irradiated, confirming that this
peak is associated with the minor isomer 10b.
function of the initial thiol concentration and the results are
presented in Table 1. As expected, the ratio 12 : 13 (1 : 1) was
independent of the thiol concentration, but the relative total
yield of alkenes 11 and 14 decreased with increasing concen-
tration of TIPST, indicating that rearrangement of the radical
16 to 17 by the 1,2-shift of the benzoyloxy group2,11 is com-
peting with trapping of 16 by the thiol to give 12 and 13. Since
the product ratio 12 : 13 is 1 : 1, the rate constants for trapping
of radical 16 by TIPST from its exo or endo face are the same
within experimental accuracy. We propose that the benzoate 14
arises from thiol-catalysed radical-chain isomerisation of the
initial product 11, as shown in Scheme 4.
Assuming that the thiol concentration remains effectively
constant during the reaction, which would imply that chain
termination takes place mainly by coupling of two benzylic
radicals 15 (see later), and that k4 may be neglected in com-
parison with k6 (the rate constant for ring opening of the
cyclopropylcarbinyl radical 17) it can be shown that eqn. (1)
When the isomeric mixture of acetals was heated in hexane
(method A) for a total of 2.5 h with 5 mol% TBHN present
initially, followed by a further three additions of 5 mol% TBHN
after 20, 40 and 60 min, about 80% of the starting material
remained unchanged. However, in the presence of 5 mol% tri-
isopropylsilanethiol (TIPST) under otherwise identical con-
ditions, essentially none (≤ 2%) of the acetal 10 remained as
(1)
should hold.2 When the product distribution was monitored as
a function of time with 10 mol% TIPST as catalyst, there was
no significant change in the ratio 11 ϩ 14 : 12 ϩ 13 (determined
by removing small samples of the reaction mixture) as the con-
version of 10 increased from 30 to 97%. This supports the
adequacy of the assumption that the thiol concentration
remains essentially constant during the redox rearrangement in
this case. A plot of the combined yield of 11 and 14 relative to
that of 12 and 13 against 1/[TIPST], for runs using 10–40 mol%
thiol catalyst, gave a best straight line of slope 0.18 MϪ1 and an
1
judged by H NMR spectroscopy and the esters 11–13 were
detected as primary products of the redox rearrangement,
together with a trace of a fourth compound subsequently identi-
fied as the homoallylic benzoate 14. The ratio 11 : 12 : 13 : 14
in the crude product was 87 : 5 : 5 : 3 and an isomeric mixture of
similar composition was isolated in 90% yield; the individual
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4074

View Article Online
Scheme 3
Scheme 4
Scheme 5
intercept (= k2/k1) of 0.89. Thus, provided the mechanism is as
shown in the ion pair 21.16a Furthermore, some collapse of the
shown in Scheme 3, β-scission of the 2-phenyl-1,3-dioxolanyl
radical 15 to give the tertiary alkyl radical 16 is only marginally
favoured over the alternative cleavage to give the secondary
cyclopropylcarbinyl radical 17. We have reported similar
observations regarding the relative rates of β-scission of
ion pair 21 via bonding of the benzoyloxy group to the other
terminus of the allylic cation moiety to give the ester 22 might
be expected to occur, but this compound was not detected as a
product from the redox rearrangement under these conditions
at 70 ЊC. We conclude that if the 1,2-shift of benzoate takes
place via the ion pair 20, opening of the 3-membered ring in the
carene radical cation evidently does not compete with the very
rapid collapse of 20 to give 17.
ؒ
ؒ
radicals of the type ROC (Ph)CH2X, to give R and PhC(O)-
CH2X, when R is a tertiary alkyl or a secondary cyclo-
propylcarbinyl group.12 This is quite reasonable, because an
α-cyclopropyl group provides significantly greater stabilisation
to an attached carbon radical centre than does a simple α-alkyl
group.13
Combining the slope and intercept it can be estimated that
k5 is about 11 times greater than k3 at 70 ЊC and, since the
former rate constant should be in the region of 107 MϪ1 sϪ1 2,14,15
,
k3 must be about 106 Ϫ1. Once formed, the cyclopropyl-
s
carbinyl radical 17 should undergo very rapid ring opening6
(k6 ca. 108 sϪ1) to give the homoallylic radical 18 and thence the
product 11.
When the redox rearrangement was carried out at higher
temperature in refluxing octane (bath temp. 140 ЊC, internal
temp. ca. 130 ЊC) with 50 mol% DTBP as initiator and 5 mol%
TIPST as catalyst (method B), none of the benzylidene acetal
10 remained after 2.5 h. Under these conditions, the ester 22
was indeed detected in addition to 11–14 and the product
mixture (11 : 12 : 13 : 14 : 22 = 38 : 3 : 4 : 15 : 40) was isolated in
92% total yield by column chromatography. We propose that 22
is derived from its less stable isomer 11 by an allylic rearrange-
ment proceeding via the ion pair 23, formed by heterolysis of 11
at the relatively high temperature of these experiments.
It should be borne in mind that quantitative kinetic analysis
of the results using eqn. (1) could possibly be compromised
if the 1,2-benzoyloxy shift takes place via initial dissociation of
16 to give a contact ion pair 20.11a In this case, if the 2-carene
radical cation16 were to undergo ring opening before it collapses
with the benzoate anion to give 17, a pathway might exist
from 16 to 18 (and thence to 11) that does not involve the
intermediacy of 17, as shown in Scheme 5. If this were so, k2
could be smaller relative to k1 than indicated by the analysis
using eqn. (1). However, when the 2-carene radical cation was
generated by photo-mediated electron transfer from the parent
terpene, it was thought to rearrange to an allylic radical-tertiary
carbocation rather than to the isomeric distonic radical cation
Quantitatively different results were obtained when the redox
rearrangement of 10 was catalysed by methyl thioglycolate
(MTG). Conversion was low (ca. 40%) with 5 mol% MTG in
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4075

View Article Online
Table 2 Rearrangement of the benzylidene acetal 10 in the presence of methyl thioglycolatea
Final composition (%)
Entry
Solvent
Bath temp./ЊC
MTG/mol%
Collidine/mol%
Initiator/mol%
10
11
12
13
14
22
c
1
2
3
4
Hexane
Hexane
Octane
Octane
70
70
140
140
5
5
5
None
10
10
TBHN (4 × 5)b
TBHN (4 × 5)b
DTBP (1 × 50)
DTBP (1 × 50)
61
5
2
6
25
3
12
26
23
28
5
9
10
13
16
27
50
53
—
8d
12
2
10
10
2
2
a The concentration of 10 was ca. 0.6 M and the total reaction time was 2.5 h. b TBHN (5 mol%) was added initially and again after 20, 40 and 60
min. c Not detected. d In a similar experiment with TIPST as catalyst without collidine (see text), none of the ionic rearrangement product 22 was
detectable. The presence of collidine appears to facilitate the formation of the ionic rearrangement product 22.
hexane at 70 ЊC (Table 2, entry 1) under conditions for which
complete conversion was obtained using TIPST as catalyst.
However, in the additional presence of 10 mol% collidine
(2,4,6-trimethylpyridine),1,2 the conversion rose to 95% (entry
2). In both runs using MTG, much more of the benzoate 14 was
formed than when TIPST was used as catalyst: an independent
synthesis served to confirm the identity of this ester. At higher
temperature even more 14 was formed (entries 3 and 4). The less
bulky MTG evidently reacts with the intermediate radical 16 to
give relatively more of the trans-isomer 12 than does the more
sterically demanding TIPST, which gave similar amounts of the
isomeric benzoates 12 and 13. These results can be understood
if the radical centre in 16 is more accessible to a thiol from the
face cis to the benzoate group and trans to the cyclopropyl ring,
which leads to 13, than from the opposite face which leads to
12. An intrinsic preference for formation of the trans isomer 12,
which is presumably more stable than the cis isomer 13, could
then be offset in the case of reaction with the bulky TIPST
because of steric inhibition to the formation of 12. There was a
marked reduction in the yield of 22 relative to that of 14 when
the amount of MTG catalyst was increased to 10 mol% (entry
4), probably because the rate of radical rearrangement of 11 to
14 is increased relative that of ionic rearrangement of 11 to 22
when the concentration of thiol is greater.17
δ 6.26 in the minor isomer 25b was irradiated; the isomers were
not separable by column chromatography. When the redox
rearrangement of 25 was conducted in hexane at 70 ЊC for a
total of 2.5 h, with four additions of 5 mol% TBHN (made
initially and again after 20, 40 and 60 min), conversion was
only 65–70% with either TIPST or MTG as catalyst. The major
isomer 25a was ca. 4.3 times more reactive than 25b towards
i
ؒ
benzylic hydrogen abstraction by Pr 3SiS and ca. 3.5 times
ؒ
more reactive towards abstraction by MeO2CCH2S . β-Scission
of the benzylic radical 26 takes place essentially exclusively by
cleavage of the C(3)–O bond to give the tertiary radical 27, that
is subsequently trapped by thiol from the exo or endo face to
give the trans-ester 28 or the cis-ester 29 as the only rearrange-
ment products. The ratio 28 : 29 was 52 : 48 with TIPST and
58 : 42 with MTG. A third product, subsequently identified by
X-ray crystallography as the benzil bis(acetal) 30, was formed
under these conditions (method A) in a significant yield of
ca. 14% using either thiol as catalyst. Evidently, this compound
is produced by dimerisation of two radicals 26. Both of the
benzylic centres in 30 possess the R-configuration and the
saturated 6-membered rings adopt only very slightly twisted
boat conformations in the crystal. An interesting feature of the
1H NMR spectrum of 30, which was interpreted with the aid
1
of H-1H correlated spectroscopy (COSY) and selective NOE
When a partially purified sample of 11 (containing ca. 5%
each of 12 and 13, but only traces of 14 and 22) was subjected
to the conditions of entry 3, examination of the crude reaction
experiments, was an apparent 1 : 3 : 3 : 1 quartet centred at
δ Ϫ0.60 and assigned to H-6 and H-6Ј. The splitting pattern
arises because of very similar couplings to the three vicinal
protons and the origin of the negative chemical shift is evident
from the crystal structure, which shows that the two protons in
1
product by H NMR spectroscopy showed it to contain 11, 14
and 22 in the ratio 8 : 74 : 18 (along with unchanged 12 and 13).
However, when this experiment was repeated in the presence of
3 mol% 4,4Ј-methylenebis(2,6-di-tert-butylphenol) as a radical
scavenger, the ratio 11 : 14 : 22 in the crude product was very
different at 8 : 5 : 87. This result provides convincing support for
our proposal that while the ester 14 arises through a radical-
chain allylic rearrangement,17 as shown in Scheme 4, the
benzoate 22 is derived from 11 by a heterolytic rearrangement
proceeding via the ion pair 23 (and thus its formation is not
subject to inhibition by the phenol). The very different rates at
which MTG and TIPST induce the radical rearrangement of 11
to 14 is not surprising because, as we have pointed out pre-
viously,17 for this type of process to proceed efficiently both of
the steps A and B in Scheme 4 must be rapid. It appears that the
greater strength of the S–H bond in the silanethiol,12,18 com-
pared with that in MTG, causes step B to become relatively
slow when TIPST is present as the catalyst. With MTG, both
steps are evidently sufficiently favourable for significant radical-
chain isomerisation to occur even at 70 ЊC.
Carane-3,4-diol
cis-Dihydroxylation of (1S)-(ϩ)-3-carene afforded the diol 24
which was converted to an 87 : 17 mixture of the benzylidene
acetals 25a and 25b. In NOE experiments, the peaks from
H(4) and the C(3)-methyl protons showed strong enhancement
when the benzylidene proton at δ 5.74 in the major isomer was
irradiated, confirming that this is 25a. No corresponding
enhancements were observed when the benzylidene proton at
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4076

View Article Online
question are positioned directly above the centres of the nearer
benzene rings.
Pinane-2,10-diol
Dihydroxylation of (1S)-(Ϫ)-β-pinene afforded pinane-2,10-
diol 36 and thence a 50 : 50 mixture of the benzylidene acetals
37a and 37b. There was an NOE correlation between the
benzylidene proton at δ 5.71 and the proton at δ 2.31 attached
to C(1), indicating that these signals are associated with the
isomer 37a. The benzylidene proton in 37b appears at δ 5.85.
Formation of 30 in this relatively high yield (maximum
possible yield 50%) indicates that the kinetic chain length
for the redox rearrangement is very short and that, at 70 ЊC,
termination by dimerisation of the 2-phenyl-1,3-dioxolan-2-yl
radical 26 is competing effectively with its β-scission to give 27.
In contrast, when the redox rearrangement of 25 was carried
out at higher temperature under the conditions of method B
(octane solvent, bath temp. 140 ЊC), in the presence of 5 mol%
TIPST, the benzylidene acetal was cleanly converted to an
inseparable 52 : 48 mixture of 28 and 29, which was isolated
in 89% yield. In similar experiments with MTG as catalyst,
conversion to a 57 : 43 mixture of 28 and 29 was also essentially
complete. The benzylic radical dimer 30 was not detected
amongst the products of these higher temperature reactions.
The activation energy for the β-scission of 26 will be appre-
ciably larger20 than that for its (presumably near-diffusion-
controlled) dimerisation and increasing the temperature will
lead to a large increase in the rate constant for β-scission
relative to that for dimerisation. In contrast to the results
obtained with the benzylidene acetal from carane-2,3-diol, no
products arising from opening of the 3-membered ring were
produced in this case, because no cyclopropylcarbinyl radical is
involved.
The redox rearrangement of 37 proceeded smoothly to
completion in the presence of 5 mol% TIPST, using either
method A or B. The allylic benzoate 38 was essentially the only
product and none of the bicyclic isomer 39 could be con-
clusively identified under either set of conditions; 38 was iso-
lated in 90–95% yield. These results show that the first-formed
benzylic radical undergoes regioselective β-scission to give the
tertiary cyclobutylcarbinyl radical 40 which undergoes rapid
opening of the 4-membered ring followed by thiol quenching to
give 38.
Pinane-2,3-diol
Commercially available (1R,2R,3S,5R)-(Ϫ)-pinane-2,3-diol 31,
derived from (1S)-(Ϫ)-α-pinene, was converted to a 74 : 26
mixture of the benzylidene acetals 32a and 32b. The benzyl-
idene proton singlet in 32a appeared at δ 5.81 and that in 32b
at δ 6.20; the assignments were made on the basis of NOE
experiments. β-Scission of the intermediate 2-phenyl-1,3-
dioxolan-2-yl radical evidently takes place with essentially
complete regioselectivity to give the tertiary radical 33,
although in refluxing octane with 5 mol% TIPST catalyst
(method B) the only redox rearrangement product detected was
the allylic benzoate 34, which was isolated in 95% yield and
arises from opening of the cyclobutylcarbinyl radical 33. The
isomer 32a of the benzylidene acetal, in which the benzylic
hydrogen atom is in the exo position on the bicyclic [4.3.0]
fragment of the molecule, is ca. 4.3 times more reactive towards
Norbornane-2,3-diol
Dihydroxylation of norbornene afforded the exo,exo-diol 41,
which was converted to the benzylidene acetal 42 containing
only a trace (ca. 1.7%) of the alternative diastereoisomer.19a
Even under the more forcing conditions of method B, with 5
mol% TIPST in refluxing octane, the redox rearrangement
of 42 was very sluggish and only ca. 30% conversion to exo-
norbornyl benzoate 43 was achieved after 3 h.
i
ؒ
Pr 3SiS than the isomer 32b in which this hydrogen atom is in
the corresponding endo position.
When the reaction was carried out at lower temperature
under the conditions of method A with 5 mol% TIPST as
catalyst, the redox rearrangement of 32 also proceeded to
completion. However, now a small amount (ca. 13%) of
another product, presumed to be the benzoate 35 that arises
from trapping of 33, was formed alongside the ring-opened
compound 34.
In contrast, the monocyclic analogue 44, prepared from
meso-butane-2,3-diol as a 57 : 43 mixture of the syn and anti
isomers, was completely converted to s-butyl benzoate 45 under
the conditions of method B. We attribute the markedly different
efficiencies with which 42 and 44 undergo redox rearrangement
to the relative rigidity of the polycyclic skeleton in the 2-phenyl-
1,3-dioxolan-2-yl radical 46, which leads to additional strain
in the transition state for its β-scission compared with the
analogous monocyclic radical 47. When chain termination
occurs by dimerisation of pairs of benzylic radicals, the relative
rates of redox rearrangement of 42 and 44 should be approxi-
mately equal to the ratio of the rate constants for β-scission of
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4077

View Article Online
46 and 47 (kβ46/kβ47), provided that the rates of initiation and the
rate constants for chain termination are similar for the two
systems.
larger (ϩ4.0Њ) for cleavage of the 2Њ–C–O bond, leading to a
preference for formation of the primary alkyl radical.
Although consideration of the computed UAS alone pro-
vides an internally consistent rationalisation of the regio-
selectivities observed experimentally for β-scission of a number
of 2-phenyl-1,3-dioxolan-2-yl and 2-phenyl-1,3-dioxan-2-yl
radicals, it is notoriously risky to quantitatively dissect the
total strain of a molecule into its various constituents, because
these are not independent of each other. In order to gain
further insight into this problem, we investigated the redox
rearrangement of some benzylidene acetals of 1,3-diols derived
from isomenthone 52 and menthone 53, on the basis that the
presence of the methyl and isopropyl ring substituents would
give information regarding the importance of torsional strain
interactions that might develop on moving to the transition
states for β-scission. Furthermore, such redox rearrangements
could provide useful synthetic routes to enantiomerically pure
substituted cyclohexyl benzoates.
We have reported previously that the relative rates of β-scis-
sion of 2-phenyl-1,3-dioxolan-2-yl radicals can be satisfactorily
predicted using density functional theory.4,20 Therefore, similar
calculations20 were carried out to model the β-scission of 46
and 47, using the Gaussian 98 package of programs.21 At the
UB3LYP/6-31G(d,p)//UB3LYP/6-31G(d,p) level of theory, the
computed activation energy for β-scission of 46 was 6.8 kJ
molϪ1 greater than that for 47 and this energy difference is very
similar (7.1 kJ molϪ1) at the UB3LYP/6-311ϩG(d,p)//UB3LYP/
6-31G(d,p) level. The β-scission of the more flexible 47 was also
favoured by entropic factors, such that the computed Arrhenius
pre-exponential factor for cleavage 47 was 3.2 times larger than
47
that for 46. Using these activation parameters, kβ46/kβ is pre-
dicted to be ca. 0.04 at 130 ЊC, in accord with the observed
much slower redox rearrangement of 42 compared with that
of 44. We have proposed4 that ‘umbrella angle strain’ (UAS)
at the emerging radical centre is a sensitive indicator of the
total strain in the transition state for β-scission of this class of
radical and, thus, of the relative rates of β-cleavage processes
that lead to formally similar types of carbon-centred radical
(in the case of both 46 and 47, to secondary alkyl radicals).
We defined UAS by eqn. (2),4 in which Σ is the sum of the
UAS = 352.4Њ – Σ
(2)
bond angles α, β and γ at the developing radical centre in the
generalised transition state structure 48 and reflects the degree
of pyramidalisation at this site. Consistent with our earlier
conclusions, the computed UAS in the transition state for β-
scission of 46 (ϩ2.9Њ) is appreciably larger than that (Ϫ0.6Њ) in
the transition state for cleavage of 47.
Treatment of the kinetic lithium enolate from (ϩ)-iso-
menthone 52 with acetaldehyde at Ϫ78 ЊC, as described by
Gardiner and co-workers,8c afforded the ketoalcohol 54 which
was reduced with LiAlH4 to give the 1,3-diol 55 as the major
diastereoisomer.8b The isomerically pure diol was converted to
the crystalline benzylidene acetal 56 and the structure of this
was confirmed by X-ray diffraction. The phenyl and two methyl
substituents occupy equatorial sites on the trans-fused bicyclo-
[4.4.0]dioxadecane core, while the isopropyl group is in an axial
site. A similar series of reactions starting from (Ϫ)-menthone
53 afforded the 1,3-diol 57 as a major product, which was
purified by recrystallisation and converted to the crystalline
benzylidene acetal 58. The structure of 58 was also confirmed
by X-ray diffraction, which showed that all four substituents
occupy equatorial sites on the trans-fused bicyclic skeleton.
1,3-Diols derived from menthone and isomenthone
We have reported that the stereoisomeric bicyclic [4.4.0] 1,3-
dioxan-2-yl radicals 49 and 50, in which the ring junction is
respectively trans or cis, undergo β-scission with markedly dif-
ferent regioselectivity.4 The selectivities for β-scission at 130 ЊC
are indicated on the structures of these radicals and the trans-
fused 49 shows a surprisingly strong preference for formation
of the primary alkyl radical. In contrast, without the con-
straints of a bicyclic skeleton, the monocyclic analogue 51
undergoes β-scission to give mainly the secondary alkyl radical
under similar conditions.1 We have suggested that an intrinsic
preference for formation of a secondary rather than a primary
radical is offset for 50, and completely outweighed for 49, by the
development of angle strain in the (chair–chair4) transition
states for cleavage of these bicyclic dioxanyl radicals and that
this strain becomes evident as UAS at the emerging radical
Redox rearrangement of the menthone-derived acetal 58 in
refluxing octane, with 5 mol% TIPST as catalyst and 50 mol%
DTBP as initiator (method B), resulted in its complete con-
version to the benzoate 59 which was isolated in 96% yield. The
same result was achieved at 70 ЊC using method A and 59 was
isolated in 94% yield; none of the isomeric benzoate 60 was
detected under either set of conditions. Without a thiol catalyst
under otherwise identical conditions, only 4% conversion of 58
to 59 took place using method A and this rose to just 10% using
method B. Redox rearrangement of the isomenthone-derived
benzylidene acetal 56, in which the methyl and isopropyl groups
centres.4 The computed UAS is negligible (0.0
0.2Њ) for
cleavage of either the 1Њ–C–O or the 2Њ–C–O bond in 51.
However, while the calculated UAS is still small (ϩ0.7Њ) for
cleavage of the 1Њ–C–O bond in the trans-fused 49, it is much
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4078

View Article Online
relative rates of cleavage of the 1Њ–C–O and 2Њ–C–O bonds in
the intermediate radical 69 are quite similar to those (91 : 9)
observed previously4 for the β-scission of the unsubstituted
parent 49, indicating that the regioselectivity is not significantly
influenced by the presence of the equatorial methyl and axial
isopropyl groups on the cyclohexane ring in 69. Taken together
with the results from the redox rearrangements of the benzyl-
idene acetals 56 and 58, we conclude that torsional strain inter-
actions are probably not of major importance in determining
the regioselectivity of β-scission for these bicyclic 2-phenyl-1,3-
dioxan-2-yl radicals.
on the cyclohexane ring are now cis, also proceeded efficiently
with 5 mol% TIPST catalyst using either method. The only
identifiable product was the benzoate 61, which was formed in
high yield (ca. 96%), although it was contaminated with a trace
of an inseparable unidentified compound with very similar
properties; this is believed not to be the regioisomeric benzoate
62.‡ For each of the intermediate benzylic radicals 63 and 64 the
two alternative modes of β-scission both lead to secondary
alkyl radicals. However, β-scission is completely regioselective
in each case and takes place to give exclusively the exocyclic
alkyl radical, in preference to the endocyclic alternative
that would result from cleavage of the bridgehead-C–O bond.
Evidently, any torsional interactions that develop between the
various ring substituents present in 63 and 64 on moving to
the product-like4 transition states do not offset the marked
resistance to cleavage of the bridgehead-C–O bond that was
observed previously for their unsubstituted trans-fused parent
49.
When the redox rearrangement of 66 was attempted in
refluxing hexane, under the conditions of method A, much
of the starting material remained unchanged and several
unidentified products were formed alongside relatively small
amounts of the benzoates 67 and 68. The difference in
behaviour of 66 at the lower temperature is presumably a con-
sequence of the comparatively slow β-scission of 69 leading to
the occurrence of side reactions. This result is reminiscent to
that obtained for the redox rearrangement of the benzylidene
acetal 25 at low temperature and, indeed, the head-to-head
dimer of 69 was probably present amongst the products from
the reaction of 66 at 70 ЊC.
Experimental
NMR spectra were recorded using a Bruker AVANCE 500
instrument (500 MHz for 1H, 125.7 MHz for 13C). Unless stated
otherwise, the solvent was CDCl3 and chemical shifts are
reported relative to residual CHCl3 (δH = 7.26) or to CDCl3
(δC = 77.0 ppm); J values are quoted in Hz and the use of
[multiplet] indicates an apparent multiplet associated with an
observed line spacing. Column chromatography and TLC were
carried out using Merck Kieselgel 60 (230–400 mesh) and
Kieselgel 60 F254 aluminium-backed pre-coated plates, respec-
tively. Optical rotations were measured using an AA Series
Polar 2000 polarimeter (Optical Activity Ltd.) in a 1 dm cell
and are given in units of 10Ϫ1 deg cm2 gϪ1. Infrared (IR) spectra
were obtained from liquid films or KBr pellets using a
Treatment of the kinetic enolate from (ϩ)-isomenthone with
gaseous formaldehyde at Ϫ78 ЊC,22a followed by reduction of
the ketoalcohol, afforded as the major product the diol 65
which has been prepared previously by a different route.22b The
structure of the derived benzylidene acetal 66 could be deduced
unambiguously from its 1H NMR spectrum, thus also confirm-
ing the identity of the diol 65.
Shimadzu FTIR-8700 spectrophotometer; wavenumbers (cmϪ1
)
are reported only for strong well-defined bands.
All redox rearrangements and reactions of air-sensitive com-
pounds were carried out under an atmosphere of dry argon or
nitrogen and all extracts were dried over anhydrous MgSO4.
Petroleum refers to the fraction of bp 40–60 ЊC.
Materials
Hexane (Aldrich) was dried by heating under reflux over
calcium hydride and then distilled from CaH2 under argon.
Anhydrous octane, di-tert-butyl peroxide (98%) and methyl
thioglycolate were obtained commercially (Aldrich) and were
used as received. Triisopropylsilanethiol was prepared accord-
ing to the method of Soderquist and co-workers10 or obtained
commercially (Aldrich). Di-tert-butyl hyponitrite was prepared
from sodium hyponitrite, tert-butyl bromide and zinc chloride
in diethyl ether, as described by Mendenhall.9b
Redox rearrangement of 66 in refluxing octane using 5 mol%
TIPST as catalyst (method B) proceeded to completion and the
benzoate esters 67 and 68 were formed in the ratio 87 : 13. The
‡ The benzoate 61 contained ca 4% of the unknown contaminant when
method B was used for the rearrangement, although this was reduced
to ca. 2% with method A or with method B when MTG was used
as catalyst. The only visible peak in the 1H NMR spectrum of
the unknown compound was what appeared to be a broad triplet
(J ca. 3 Hz) at δ 5.48 and this was assumed to arise from a single proton
in the molecule.
(ϩ)-Isomenthone 52 and (Ϫ)-menthone 53 were prepared by
oxidation with aqueous-acidic potassium dichromate of (ϩ)-
20
isomenthol {[α]D24 ϩ24.0 (c 4.2, EtOH)} or (Ϫ)-menthol {[α]D
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4079

View Article Online
Ϫ50.4 (c 5.6, EtOH, 99% ee}, respectively (both from Aldrich),
IR (KBr disc) 3456, 3332, 2959, 1456, 1367, 1136, 1006,
981. (Found: C, 72.2; H, 12.2. C12H24O2 requires C, 72.0; H,
12.1%).
as described by Brown and co-workers.23a
Diols
(1ЈR,1S,2R,3R,6S )-2-(1-Hydroxyethyl)-6-isopropyl-3-methyl-
cyclohexanol 57. This was prepared from (Ϫ)-menthone23a
using the same procedure as for 55 and on the same scale. The
final recrystallisation stage afforded the diol 57 as colourless
(1R,2R,3S,5R)-(Ϫ)-Pinane-2,3-diol 31 (97% ee) and meso-
butane-2,3-diol were obtained from Aldrich and used as
received. Where appropriate, because the compounds are
incompletely described in the literature, details of the prepar-
ation and/or characterisation of starting materials are given
below.
(2S,3R)-Carane-2,3-diol 9 was prepared by dihydroxylation
of (ϩ)-2-carene (Aldrich), using basic KMnO4 in aqueous tert-
butyl alcohol as described in the literature.24 The starting 2-
25
needles (3.57 g, 34%), mp 119–120 ЊC, [α]D Ϫ33.8 (c. 1.36,
CHCl3); δH 0.79 (3 H, d, J 7.0, CHMe2), 0.92 (3 H, d, J 7.0,
CHMe2), 0.95 (3 H, d, J 6.4, Me-3), 0.98 (2 H, m, ring-H), 1.05
(1 H, [td], J 10.0 and 2.8, H-2), 1.24 (1 H, m. ring-H), 1.31 (3 H,
d, J 6.6, Me-1Ј), 1.34 (1 H, m, H-3), 1.56 (1 H, m, HA-5), 1.66
(1 H, m, HB-5), 2.18 (1 H, septet of doublets, J 7.0 and 2.8,
CHMe2), 2.31 (1 H, brs, OH), 2.39 (1 H, brs, OH), 3.44 (1 H, [t],
J 10.0, H-1), 4.11 (1 H, m, H-1Ј); δC 15.9, 20.8, 21.1, 22.6, 25.0,
25.7, 34.4, 35.3, 49.8, 57.4, 69.5, 73.0; IR (KBr disc) 3345, 3333,
2950, 1448, 1371, 1124, 1019, 978. (Found: C, 72.3; H, 12.3.
C12H24O2 requires C, 72.0; H, 12.1%).
20
carene showed [α]D ϩ86.9 (c. 3.1, EtOH), which corresponds
to an ee of ca. 92% according to values in the literature.25
(3R,4R)-Carane-3,4-diol7a 24 was prepared similarly from
20
(1S)-(ϩ)-3-carene (Fluka), which showed [α]D ϩ17.2 (neat).
This diol showed mp 69 ЊC (lit.7a mp 68 ЊC); δH 0.63 (1 H, [t]d,
J 9.5 and 4.5, ring-H), 0.85 (1 H, [t], J 8.5, ring-H), 0.89 (3 H,
s, Me-7), 1.00 (3 H, s, Me-7), 1.20 (3 H, s, Me-1), 1.23 (1 H, dd,
J 15.6 and 4.5, ring-H), 1.67 (1 H, ddd, J 14.6, 8.5 and 4.5, ring-
H), 1.70 and 1.82 (2 H, brs, OH), 2.04 (1 H, dd, J 14.6 and 7.3,
ring-H), 2.10 (1 H, dd, J 15.6 and 9.5, ring-H), 3.18 (1 H, dd,
J 9.5 and 7.5, H-2); δC 15.4, 16.2, 17.5, 21.5, 25.4, 26.9, 28.5,
33.2, 70.2, 73.1.
Other isomeric diols were present in the residue from
the recrystallisation stage, but no pure compounds could be
isolated by column chromatography.
(1S,2R,3R,6R)-2-Hydroxymethyl-6-isopropyl-3-methylcyclo-
hexanol 65.22b The kinetic lithium enolate derived from (ϩ)-
isomenthone (10.00 g, 64.8 mmol) was prepared as described
before. Excess gaseous formaldehyde, from the pyrolysis of
paraformaldehyde (4.25 g, corresponding to 112 mmol of
the monomer), was conducted into the reaction flask during
20 min in a stream of dry argon. The tip of the transfer tube
terminated just above the surface of the solution, which was
stirred and maintained at ca. Ϫ78 ЊC during the addition.
The work-up was as described for the ketoalcohol 54 and the
crude product (11.5 g, ca. 62.5 mmol) was reduced with LiAlH4
(6.00 g, 158.0 mmol) in diethyl ether (100 mL), as described
before, to afford the diol 65 as a viscous oil (containing ca. 8%
of a presumed isomer); δH (C6D6) 0.92 (3 H, d, J 6.6, Me-3),
1.00 (3 H, d, J 6.5, CHMe2), 1.18 (3 H, d, J 6.5, CHMe2), 1.28
(5 H, m, ring-H), 1.60 (1 H, m, ring-H), 1.66 (1 H, m, ring-H),
1.83 (1 H, m, ring-H), 3.26 (2 H, brs, OH), 3.50 (1 H, dd, J 10.5
and 7.9, HA-1Ј), 3.81 (1 H, dd, J 10.5 and 4.0, HB-1Ј), 3.90 (1 H,
dd, J 7.4 and 3.6, H-1); δC 20.7, 22.0, 23.0, 24.6, 27.2, 29.7, 31.0,
45.6, 48.3, 64.8 and 75.7.
(2S)-Pinane-2,10-diol 36 was prepared by dihydroxylation
of (1S)-(Ϫ)-β-pinene (Aldrich), which showed [α]D Ϫ22.0
(neat), using tetradecyltrimethylammonium permanganate as
described by Hazra et al.;7 exo,exo-norbornane-2,3-diol19b 41
was prepared from norbornene in the same way.
20
(1ЈR,1S,2R,3R,6R)-2-(1-Hydroxyethyl)-6-isopropyl-3-methyl-
cyclohexanol 55.8a Butyllithium solution (1.6 M in hexanes,
44.0 mL, 70.4 mmol) was added dropwise under argon to a
stirred solution of dry diisopropylamine (7.58 g, 74.9 mmol) in
dry tetrahydrofuran (THF, 100 mL) cooled in an ice-water
bath. The resulting solution was stirred at this temperature for
a further 15 min, then cooled in a solid CO2-acetone bath. Iso-
menthone23b (10.00 g, 64.8 mmol) in THF (10 mL) was added
dropwise during 15 min with stirring and the resulting solution
was stirred for a further 1 h. Still with cooling in the solid CO2–
acetone bath, acetaldehyde (3.30 g, 74.8 mmol) was added
dropwise during 10 min and the resulting mixture was sub-
sequently stirred at this temperature for a further 1.5 h, before
being quenched with saturated aqueous NH4Cl (200 mL) at
Ϫ78 ЊC to room temperature. The organic layer was separated
and concentrated by evaporation. The aqueous phase was
extracted with diethyl ether (3 × 30 mL) and the combined
organic liquid was washed with saturated brine and dried. After
evaporation of the solvent, the residue (10.5 g, 53.0 mmol based
on the product ketoalcohol 54) was dissolved in dry diethyl
ether (50 mL) and added dropwise with stirring to LiAlH4
(5.00 g, 132 mmol) in diethyl ether (100 mL) cooled in an ice–
water bath. After the addition, the mixture was allowed to
warm to room temperature and stirred for 30 min, then heated
under gentle reflux for 1 h. The reaction mixture was cooled in
an ice–water bath and quenched by successive addition of water
(5 mL), aqueous NaOH (15% w/v, 5 mL) and water (10 mL).
The solid material was removed by filtration and washed on the
filter with diethyl ether. Evaporation of the solvent from
the filtrate afforded an oily residue that was crystallised from
hexane at Ϫ4 ЊC and then recrystallised from the same solvent
to afford the diol 55 (7.22 g, 68%), mp 98–100 ЊC, [α]D25 Ϫ14.5
(c 1.30, CHCl3); δH 0.93 (3 H, d, J 6.6, CHMe2), 1.00 (3 H, d,
J 6.6, CHMe2), 1.12 (3 H, d, J 7.1, Me-3), 1.15 (1 H, m, ring-H),
1.28 (3 H, d, J 6.3, Me-1Ј), 1.37 (3 H, m, ring-H and OH), 1.50
(3 H, m, ring-H), 1.56–1.72 (3 H, m, ring-H and OH), 3.87 (1 H,
[sextet], J 6.3, H-1Ј), 4.31 (1 H, [q], J 3.6, H-1); δC 20.5, 21.1,
21.3, 21.4, 23.4, 28.7, 28.9(9), 29.0(2), 44.4, 54.7, 68.2, 69.6;
This diol was used to prepare the benzylidene acetal 66 with-
out further purification.
Benzylidene acetals
A mixture of the diol (typically ca. 20 mmol), benzaldehyde
(ca. 23 mmol) and, unless stated otherwise, p-toluenesulfonic
acid (50 mg) in benzene (40 mL) was stirred and heated under
reflux for ca. 1 h, while water was removed azeotropically using
a Dean–Stark trap. The solution was allowed to cool, shaken
with calcium carbonate (ca. 0.5 g) to neutralise the acid and
the suspension was filtered through Celite. The filter cake was
washed with diethyl ether, the solvent was removed from the
filtrate by evaporation and the acetals were isolated, as oils
unless stated otherwise, by flash chromatography (light
petroleum–diethyl ether eluent 20 : 1), usually followed by dis-
tillation under reduced pressure. The characteristic properties
are given below. It should be noted that the optical rotations
reported for benzylidene acetals refer to mixtures of stated
diastereoisomeric composition.
Acetal 10 from carane-2,3-diol. Yield 88%, bp 115–118 ЊC/
0.05 mmHg as a mixture of 10a and 10b in a ratio of 57 : 43;
[α]D18 Ϫ29.8 (c 3.2, EtOH), [α]D18 Ϫ24.5 (c 1.5, CHCl3). (Found:
C, 78.8; H, 8.8. C17H22O2 requires C, 79.0; H, 8.6%).
Isomer 10a. δH 0.87 (1 H, m, ring-H), 0.97 (3 H, s, Me-7), 1.06
(3 H, s, Me-7), 1.25 (1 H, m, ring-H), 1.33 (3 H, s, Me-3), 1.55
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4080

View Article Online
(1 H, m, ring-H), 1.67 (1 H, d, J 5.6, H-3), 1.78 (1 H, m, ring-
H), 2.07 (1 H, m, ring-H), 3.79 (1 H, s, H-2), 5.81 (1 H, s,
PhCH ), 7.37 (3 H, m, Ph), 7.53 (2 H, m, Ph); δC 15.7, 16.5, 17.0,
21.8, 24.4, 25.9, 29.0, 34.3, 77.8, 79.7, 100.3, 127.0, 128.3, 129.2,
137.8.
Isomer 10b. δH 0.87 (1 H, m, ring-H), 0.99 (3 H, s, Me-7), 1.09
(3 H, s, Me-7), 1.28 (1 H, m, ring-H), 1.34 (3 H, s, Me-3), 1.56
(1 H, m, ring-H), 1.67 (1 H, dd, J 5.6 and 1.6, H-3), 1.77 (1 H,
m, ring-H), 2.07 (1 H, m, ring-H), 3.97 (1 H, s, H-2), 6.00 (1 H,
s, PhCH ), 7.37 (3 H, m, Ph), 7.53 (2 H, m, Ph); δC 15.6, 16.4,
17.0, 21.6, 23.8, 27.1, 28.7, 32.2, 77.6, 79.4, 101.4, 126.7, 128.3,
129.0, 138.3.
(2 H, d, J 1.3, H-3,7), 5.56 (1 H, s, PhCH ), 7.38 (2 H, m, Ph),
7.54 (2 H, m, Ph); δC 23.0, 31.8, 39.8, 83.0, 102.6, 126.8, 128.4,
129.4, 136.3.
Acetal 44 from meso-butane-2,3-diol.26 Yield 95%, bp 63–68
ЊC/0.05 mmHg, as a mixture of isomers 44a and 44b in the ratio
57 : 43. Determination of the stereochemistry at the benzyl-
idene centres in 44a and 44b was based on NOE experiments
1
and our assignments of the two H NMR spectra are reversed
from those reported in the literature.26 When the benzylidene
proton appearing at δ 5.77 in the major isomer was irradiated,
the signal at δ 4.34 from H-3,4 on the dioxolane ring in the
same isomer showed a strong NOE enhancement, identifying
this compound as 44a. No corresponding enhancement was
observed when the benzylidene proton at δ 6.11 in the minor
isomer 44b was irradiated.
Acetal 25 from carane-3,4-diol. Yield 83%, bp 113–115 ЊC/
0.05 mmHg (solidified at room temperature, mp 65–72 ЊC) as a
18
mixture of the isomers 25a and 25b in a ratio of 87 : 13; [α]D
ϩ3.9 (c 2.44, CHCl3). (Found: C, 79.2; H, 8.4. C17H22O2
requires C, 79.0; H, 8.6%).
(2S,4R,5S)-4,5-Dimethyl-2-phenyl-1,3-dioxolane 44a. δH 1.27
(6 H, m, Me), 4.34 (2 H, m, MeCH ), 5.77 (1 H, s, PhCH ), 7.38
(3 H, m, Ph), 7.50 (2 H, m, Ph); δC 15.5, 75.1, 102.8, 126.8,
128.4, 129.3, 137.8.
(2R,4R,5S)-4,5-Dimethyl-2-phenyl-1,3-dioxolane 44b. δH 1.24
(6 H, m, Me), 4.35 (2 H, m, MeCH ), 6.11 (1 H, s, PhCH ), 7.35
(3 H, m, Ph), 7.47 (2 H, m, Ph); δC 14.5, 74.6, 101.5, 126.0,
128.3, 120.2, 139.9.
Isomer 25a. δH 0.70–0.98 (4 H, complex, ring-H), 1.00 (3 H,
s, Me-7), 1.13, (3 H, s, Me-7), 1.31 (3 H, s, Me-3), 2.16 (1 H, dd,
J 10.2 and 7.2, ring-H), 2.28 (1 H, ddd, J 15.4, 7.2 and 2.5,
ring-H), 3.99 (1 H, [t], J 2.5, H-4), 5.74 (1 H, s, PhCH ), 7.38
(3 H, m, Ph), 7.55 (2 H, m, Ph); δC 14.9, 16.4, 17.6, 18.2, 23.1,
25.4, 28.5, 29.6, 79.5, 80.9, 100.6, 127.2, 128.3, 129.5, 137.3.
Isomer 25b. δH 0.70–0.98 (4 H, complex, ring-H), 0.98 (3 H,
s, Me-7), 1.12, (3 H, s, Me-7), 1.32 (3 H, s, Me-3), 2.17 (1 H,
dd, J 10.1 and 7.2, ring-H), 2.29 (1 H, ddd, J 15.4, 7.2 and 2.6,
ring-H), 4.14 (1 H, [t], J 2.6, H-2), 6.26 (1 H, s, PhCH ), 7.38
(3 H, m, Ph), 7.50 (2 H, m, Ph); δC 14.3, 17.6, 19.0, 19.2, 24.4,
27.8, 28.4, 30.5, 80.5, 81.0, 104.4, 126.4, 128.3, 128.9, 139.6.
Acetal 56 from the 1,3-diol 55. Prepared using PPTS as cata-
lyst. Colourless crystals from MeOH (yield 94%), mp 103–104
ЊC, [α]D25 ϩ3.1 (c 1.80, CHCl3); δH (in C6D6, which gives better
signal dispersion than CDCl3) 0.80 (3 H, d, J 6.5, CHMe2), 0.90
(3 H, d, J 6.5, CHMe2), 0.92 (1 H, m, ring-H) 1.00–1.18 (2 H,
complex, ring-H), 1.25 (1 H, tt, J 13.5 and 4.1, ring-H), 1.31
(3 H, d, J 6.6, Me-3), 1.36 (3 H, d, J 6.1, Me-1Ј), 1.37 (1 H, m,
ring-H), 1.63–1.75 (2 H, m, CHMe2 and ring-H), 1.94 (1 H, m,
ring-H), 3.36 (1 H, dq, J 9.0 and 6.1, H-1Ј), 3.49 (1 H, dd, J 10.5
and 4.2, H-1), 5.40, (1 H, s, PhCH ), 7.12 (1 H, m, Ph), 7.20
(2 H, m, Ph), 7.70 (2 H, m, Ph); δC 22.1, 22.4, 23.4, 23.9, 26.2,
27.5, 31.7, 33.7, 44.4, 47.0, 78.2, 84.7, 101.3, 126.8, 128.2,
128.6, 140.3. (Found: C, 79.2; H, 9.9. C19H28O2 requires C, 79.1;
H, 9.8%).
Acetal 32 from pinane-2,3-diol. Prepared using pyridinium
p-toluenesulfonate (PPTS ca. 70 mg) as catalyst in place of
p-toluenesulfonic acid. Yield 83%, as a mixture of isomers 32a
18
and 32b in a ratio of 74 : 26; [α]D Ϫ19.3 (c 1.71, CHCl3).
(Found: C, 78.9; H, 8.8. C17H22O2 requires C, 79.0; H, 8.6%).
Isomer 32a. δH 0.92 (3 H, s, Me-6), 1.35 (3 H, s, Me-6), 1.51
(3 H, s, Me-2), 1.96 (1 H, m, ring-H), 2.10 (1 H, ddd, J 14.6,
3.6 and 1.1, ring-H), 2.14–2.25 (3 H, m, ring-H), 2.30 (1 H, m,
ring-H), 4.15 (1 H, d, J 7.4, H-3), 5.81 (1 H, s, PhCH ), 7.38
(3 H, m, Ph), 7.57 (2 H, m, Ph); δC 24.1, 25.2, 25.6, 27.2, 32.8,
37.9, 40.0, 50.8, 77.5, 83.8, 100.6, 127.0, 128.4, 129.4, 136.7.
Isomer 32b. δH 0.93 (3 H, s, Me-6), 1.34 (3 H, s, Me-6), 1.37
(3 H, s, Me-2), 2.03 (1 H, m, ring-H), 2.15–2.35 (4 H, m, ring-
H), 2.43 (1 H, m, ring-H), 4.43 (1 H, dd, J 9.0 and 3.0, H-3),
6.20 (1 H, s, PhCH ), 7.37 (3 H, m, Ph), 7.50 (2 H, m, Ph);
δC 24.2, 26.8, 27.2, 28.4, 34.1, 38.8, 39.9, 52.3, 77.4, 83.6, 103.6,
126.4, 128.3, 128.8, 139.7.
Acetal 58 from the 1,3-diol 57. Prepared using PPTS as cata-
lyst. Colourless crystals from MeOH (yield 92%), mp 110 ЊC,
25
[α]D Ϫ48.7 (c 1.84, CHCl3); δH 0.83 (3 H, d, J 7.0, CHMe2),
0.90(3 H, d, J 7.0, CHMe2), 1.04 (3 H, d, J 6.3, Me-3), 1.18
(3 H, m, ring-H), 1.28 (1 H, m, H-3), 1.46 (3 H, d, J 6.0, Me-1Ј),
1.52–1.72 (4 H, m, ring-H), 2.26 (1 H, m, CHMe2), 3.32 (1 H,
dd, J 10.1 and 9.4, H-1), 3.76 (1 H, dq, J 9.2 and 6.0, H-1Ј),
5.53, (1 H, s, PhCH ), 7.31 (1 H, m, Ph), 7.36 (2 H, m, Ph), 7.51
(2 H, m, Ph); δC 16.0, 20.4, 22.4, 23.1, 23.2, 25.5, 33.2, 36.4,
46.4, 52.8, 77.5, 82.2, 100.0, 125.9, 128.1, 128.3, 139.2. (Found:
C, 79.3; H, 9.7. C19H28O2 requires C, 79.1; H, 9.8%).
Acetal 37 from pinane-2,10-diol. Prepared using PPTS as
catalyst. Toluene was used in place of benzene as the solvent for
azeotropic removal of water and the reflux period was extended
to 4 h. Yield 72%, as a mixture of isomers 37a and 37b in a ratio
Acetal 66 from the 1,3-diol 65. Prepared using PPTS as cata-
lyst. Colourless crystals from MeOH (yield 84%), mp 35 ЊC,
[α]D25 Ϫ13.4 (c 1.04, CHCl3); δH 0.90 (3 H, d, J 6.0, Me-3), 0.96
(3 H, d, J 6.7, CHMe2), 1.19 (3 H, d, J 6.7, CHMe2), 1.20 (2 H,
m, ring-H), 1.38–1.48 (2 H, complex, ring-H), 1.70 (1 H, m,
ring-H), 1.80 (1 H, m, ring-H), 1.92 (1 H, m, ring-H), 2.20 (1 H,
m, CHMe2), 3.53 (1 H, [t], J 11.0, Hax-1Ј), 3.76 (1 H, dd, J 10.5
and 4.4, H-1), 4.37 (1 H, dd, J 11.0 and 4.7, Heq-1Ј), 5.45 (1 H, s,
PhCH ), 7.35 (3 H, m, Ph), 7.51 (2 H, m, Ph); δC 18.5, 22.0, 23.8,
25.7, 27.5, 29.8, 33.4, 39.9, 43.8, 71.5, 85.2, 102.0, 126.2, 128.2,
128.6, 139.0. (Found: C, 78.6; H, 9.8. C18H26O2 requires C, 78.8;
H, 9.6%).
18
50 : 50; [α]D Ϫ2.80 (c 1.82, CHCl3). (Found: C, 78.8; H, 8.7.
C17H22O2 requires C, 79.0; H, 8.6%).
Isomer 37a. δH 0.88 (3 H, s, Me-6), 1.26 (3 H, s, Me-6), 1.70–
2.35(8 H, complex, ring-H), 3.87 (1 H, d, J 8.1, OCHAHB), 3.92
(1 H, d, J 8.1, OCHAHB), 5.58 (1 H, s, PhCH ), 7.36 (3 H, m,
Ph), 7.50 (2 H, m, Ph); δC 22.7, 24.1, 26.6, 26.9, 29.5, 38.2, 40.2,
51.2, 77.6, 86.1, 101.9, 126.8, 128.3, 129.2, 138.3.
Isomer 37b. δH 0.88 (3 H, s, Me-6), 1.29 (3 H, s, Me-6), 1.70–
2.35 (8 H, complex, ring-H), 3.83 (1 H, d, J 8.1, OCHAHB), 4.01
(1 H, d, J 8.1, OCHAHB), 5.71 (1 H, s, PhCH ), 7.36 (3 H, m,
Ph), 7.53 (2 H, m, Ph); δC 22.8, 23.8, 26.1, 26.8, 29.2, 38.1, 40.1,
49.4, 77.3, 86.4, 101.9, 126.7, 128.3, 129.1, 138.2.
1
The H NMR spectrum was assigned with the aid of 2D-
COSY experiments and the structure of 66 was confirmed on
the basis of selective NOE experiments. The chemical shift of
the benzylidene proton (δ 5.45) confirms2 that, as expected, the
phenyl group occupies an equatorial site. When the benzylidene
proton was irradiated, H-1 (δ 3.76) and one of the protons
Acetal 42 from norbornane-2,3-diol.19a Yield 89%, bp 100–
102 ЊC/0.05 mmHg; δH 1.00 (2 H, m, ring-H), 1.12 (1 H, d[quin-
tet], J 10.3 and 1.4, ring-H), 1.53 (2 H, m, ring-H), 1.90 (1 H,
d[quintet], J 10.3 and 1.8, ring-H), 2.44 (2 H, m, ring-H), 4.03
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4081

View Article Online
(δ 3.53) attached to C(1Ј) showed strong NOE enhancement,
establishing that all three of these protons occupy axial sites.
An NOE enhancement was also observed for Me-3 when the
equatorial proton (δ 4.37) attached to C(1Ј) was irradiated,
indicating that this methyl group is also equatorial. The large
coupling (10.5 Hz) between H-1 and H-2 confirms the trans ring
junction stereochemistry and the value of JH(1)–H(6) (4.4 Hz)
confirms that the isopropyl group is axial.
was added and the aqueous phase was extracted with diethyl
ether (3 × 30 mL). The ethereal solution was washed
successively with dilute HCl, saturated aqueous NaHCO3 and
saturated brine, then dried. The solvent was removed by
evaporation and the residue was purified by flash chromato-
graphy, using petroleum ether (20 : 1) as eluent A final dis-
tillation gave the benzoate 14 (2.23 g; bp 114 ЊC/0.05 mmHg)
as an oil, containing ca. 5% of the corresponding saturated
benzoate (that arose from saturated ketone present in the
original 4-isopropylcyclohex-3-enone); δH 0.99 (6 H, d, J 6.9,
CHMe2), 1.66 (3 H, s, ring-Me), 1.84 (1 H, dddd, J 14.0, 8.4, 5.8
and 0.8, ring-H), 2.04 (1 H, m, ring-H), 2.13 (1 H, m, ring-H),
2.21 (1 H, [septet], J 6.9, Me2CH ), 2.34 (1 H, m, ring-H), 2.38
(1 H, ddd, J 11.5, 5.7 and 1.5, ring-H), 2.67 (1 H, m, ring-H),
5.32 (1 H, m, H-3), 7.40 (2 H, m, Ph), 7.51 (1 H, m, Ph), 7.97
(2 H, m, Ph); δC 21.4, 21.5, 23.6, 24.3, 33.0, 34.7, 37.3, 81.1,
115.3, 128.2, 129.3, 132.0, 132.4, 142.6, 165.9. (Found: C, 78.7;
H, 8.8. C17H22O2 requires C, 79.0; H, 8.6%).
General procedures for redox rearrangement
Method A. The acetal (1.0 mmol), dry hexane (1.5 mL),
TBHN (0.05 mol) and the thiol catalyst (0.05 mol) were succes-
sively introduced into a dry, argon-filled, two necked 10 mL
round-bottomed flask, containing a dry magnetic stirrer bar
and fitted with a condenser through which a slow downward
flow of argon was maintained. The side neck was closed with a
stopper and the flask was immersed in an oil bath that had been
pre-heated to 70 ЊC. Further portions of TBHN (0.05 mmol)
were added after 20, 40 and 60 min. After the final addition,
the mixture was stirred for a further 1.5 h, allowed to cool
and the volatile material was removed by evaporation. The
crude product was examined by 1H NMR spectroscopy to
determine its composition and estimate the extent of conver-
sion to benzoate esters, before the latter were isolated by flash
chromatography using petroleum-diethyl ether eluent (20 : 1).
The ester 22,29a containing only traces of the isomers 11–14,
was isolated as the major product from the rearrangement of
partially purified 11 in the presence of 3 mol% 4,4Ј-methyl-
enebis(2,6-di-tert-butylphenol) (see text): Oil, δH 0.88 (3 H, d,
J 6.9, Me), 0.98 (3 H, d, J 6.9, Me), 1.50 (1 H, m, ring-H), 1.67
(1 H, m, ring-H), 1.70 (3 H, s, Me-1), 1.83 (1 H, m, ring-H), 2.04
(2 H, m, CHMe2 and ring-H), 2.05 (1 H, m, ring-H), 5.45 (1 H,
m, H-2), 5.52 (1 H, m, H-3), 7.45 (2 H, m, Ph), 7.51 (1 H, m,
Ph), 8.06 (2 H, m, Ph); δC 18.1, 20.9, 21.3, 23.2, 26.9, 29.6, 44.1,
72.7, 120.9, 128.3, 129.6, 130.6, 132.7, 139.8, 166.6. These
NMR data are consistent with those reported for the corre-
sponding acetate.29b (Found: C, 78.8; H, 8.8. C17H22O2 requires
C, 79.0; H, 8.6%).
Method B. The procedure was similar to that for Method A,
except that the solvent was octane and a single initial addition
of DTBP (0.50 mmol) served as initiator in place of TBHN.
The reaction mixture was stirred and heated for 2.5 h in an oil
bath maintained at 140 ЊC.
The characteristics of the product benzoates are described
below.
Redox rearrangement of 25
The benzoates 28 and 29 were obtained as an inseparable
approximately equimolar mixture; their structures were con-
firmed by NMR spectroscopy in comparison with the data
reported for the known acetates.27a Because of the equimolar
composition of the mixture it was not possible to assign all
the peaks to individual isomers. The mixture (an oil) showed
δH 0.90 (3 H, d, J 6.5, Me-7), 0.95 (3 H, d, J 6.5, Me-7), 1.02
(6 H, s, Me-3), 1.40–2.40 (14 H, complex, ring-H), 4.61 (1 H,
Redox rearrangement of 10
The distribution of products 11–14 varied depending on the
reaction conditions (see text). From the product mixture
obtained at 70 ЊC, the ester 11 containing only small amounts
(ca. 1–2%) of the isomers 12–14 was isolated as an oil by
column chromatography; δH 0.90 (3 H, d, J 6.8, CHMe2), 0.92
(3 H, d, J 6.8, CHMe2), 1.48 (1 H, m, ring-H), 1.62 (1 H, m,
ring-H), 1.65 (3 H, s, Me-1), 1.85 (1 H, m, ring-H), 2.05 (1 H, m,
CHMe2), 2.13 (1 H, m, ring-H), 2.22 (1 H, m, ring-H), 5.75
(1 H, ddd, J 10.3, 2.6 and 0.9, H-2), 6.06 (1 H, ddd, J 10.3, 2.6
and 1.2, H-3), 7.41 (2 H, m, Ph), 7.52 (1 H, m, Ph), 8.00 (2 H, m,
Ph); δC 19.5, 19.7, 22.9, 26.0, 31.7, 34.2, 41.1, 81.0, 128.2, 129.4
131.0, 131.8, 132.4, 132.6, 165.6. (Found: C, 79.2; H, 8.5.
C17H22O2 requires C, 79.0; H, 8.6%).
ddd, J 15.8, 8.8 and 6.7 for 28), 4.96 (1 H, m, W 9.0 Hz for
½
29), 7.42 (4 H, m, Ph), 7.54 (2 H, m, Ph), 8.05 (4 H, m, Ph);
δC 15.1, 16.0, 17.2, 17.4, 17.8, 17.9, 18.2, 18.9, 20.0, 21.3, 24.1,
24.7, 26.8, 28.2, 28.7, 29.1, 31.0, 34.0, 73.3, 78.1, 128.2(6) and
128.2(9), 129.4(7) and 129.4(9), 130.8 and 131.0, 132.6 (2C),
166.3(5) and 166.3(7). (Found: C, 78.8; H, 8.8. C17H22O2
requires C, 79.0; H, 8.6%).
The esters 12 and 13 could not be obtained free from 11 and
14; they were identified from key features of their NMR spectra
by comparison with those of the corresponding acetates.27
In particular, the absorption at δ 4.47 (dd, J 10.8 and 2.8) in
The benzil bis(acetal) 30
Colourless crystals from methanol, mp 197–200 ЊC; [α]D18 Ϫ5.7
(c 1.92, CHCl3); δH (carane skeleton numbering) Ϫ0.60 (2 H,
[q], J 8.3, H-6,6Ј), 0.02 (2 H, [q], J 8.3, H-1,1Ј), 0.48 (2 H, dd,
J 15.8 and 8.4, HA-2,2Ј), 0.51 (2 H, ddd, J 15.8, 8.4 and 1.6,
HA-5,5Ј), 0.62 (6 H, s, 2 Me), 0.68 (6 H, s, 2 Me), 0.70 (6 H, brs,
2 Me), 1.86 (2 H, dd, J 15.8 and 8.0, HB-2,2Ј), 2.13 (2 H, ddd,
J 15.8, 7.5 and 3.3, HB-5,5Ј), 3.46 (2 H, brs, H-4,4Ј), 7.22 (6 H,
m, Ph), 7.60 (4 H, m, Ph); δC 14.6, 18.0, 18.4, 18.7, 23.1, 26.2,
28.2, 30.7, 81.7, 82.2, 111.4, 126.1, 127.0, 128.8, 141.5. (Found:
C, 79.2; H, 8.1. C34H42O4 requires C, 79.3; H, 8.2%).
1
the H NMR spectrum of the mixture was assigned to H-2 in
the ester 12 in which H-2 and H-3 are trans. The equivalent
absorption for the corresponding acetate appears at δ 4.20
(dd, J 10.6 and 2.0).27a In the isomeric ester 13 H-2 appears at
δ 5.12 (dd, J 4.2 and 2.4).
The ester 14 was identified by comparison with the authentic
compound prepared by benzoylation of 1-methyl-4-isopropyl-
cyclohex-3-en-1-ol,20a itself prepared by treatment of 4-iso-
propylcyclohex-3-enone28b with MeMgI. The alcohol showed
δH 0.99 (6 H, d, J 7.2, CHMe2), 1.22 (3 H, s, ring-Me), 1.55 (1 H,
m, ring-H), 1.67 (2 H, ring-H), 1.90–2.40 (4 H, m, ring-H), 5.29
(1 H, m, H-3); δC 21.4, 21.5, 23.6, 28.1, 34.6, 35.6, 39.8, 68.6,
115.9, 142.8.
Redox rearrangement of 32
(1S,5R)-2-Methyl-5-isopropylcyclohex-2-enyl benzoate 34.
This compound was isolated from the redox rearrangement
Benzoyl chloride (1.78 g, 12.7 mmol) was added to a stirred
solution of the alcohol (1.62 g, 10.5 mmol) in pyridine (20 mL),
with cooling in an ice–water bath, and the mixture was then
stirred overnight at room temperature. Ice-cold water (30 mL)
of 32 in refluxing octane. Oil, [α]18 Ϫ167.8 (c 1.5, CHCl3);
D
δH 0.88 (3 H, d, J 6.7, CHMe2), 0.89 (3 H, d, J 6.7, CHMe2),
1.50 (2 H, m, CHMe2 and ring-H), 1.59 (1 H, m, ring-H), 1.73
(1 H, s, Me-2), 1.75 (1 H, m, ring-H), 2.02 (1 H, d[q], J 14.0 and
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4082

View Article Online
2.0, ring-H), 2.15 (1 H, m, ring-H), 5.49 (1 H, m, H-1), 5.78
(1 H, m, H-3), 7.44 (2 H, m, Ph), 7.55 (1 H, m, Ph), 8.06 (2 H,
m, Ph); δC 18.2, 19.4, 19.9, 20.7, 28.9, 31.7, 32.7, 34.9, 71.7,
120.9, 128.5, 129.6 130.8, 131.0, 132.7, 166.5. (Found: C, 78.7;
H, 8.9. C17H22O2 requires C, 79.0; H, 8.6%).
Redox rearrangement of 66
The benzoates 68 and 69 were obtained as an inseparable 87 : 13
mixture; their structures were verified by NMR spectroscopy.
(Found: C, 78.5; H, 9.7. C18H26O2 requires C, 78.8; H, 9.6%).
The benzoate 35, in which the pinane skeleton is retained,
was tentatively identified from its partial H NMR spectrum,
in particular from the absorption at δ 5.40 (dd, J 9.6 and 5.4)
which was assigned to the CHOBz group in the ester 35, by
analogy with NMR data reported for the corresponding
acetate.30
(1S,2S,3R,6S )-2,3-Dimethyl-6-isopropylcyclohexyl benzoate
67. δH 0.91 (3 H, d, J 6.7, CHMe2), 0.92 (3 H, d, J 6.7, CHMe2),
1.04 (6 H, d, J 7.1, Me-2 and 3), 1.40–2.00 (8 H, complex,
CHMe2 and ring-H), 5.08 (1 H, dd, J 5.2 and 3.1, H-1), 7.44
(2 H, m, Ph), 7.55 (1 H, m, Ph), 8.06 (2 H, m, Ph); δC 17.4, 20.7,
21.3, 21.5, 22.8, 27.8, 28.1, 34.3, 38.7, 42.2, 78.5, 128.3, 129.5,
130.9, 132.6, 165.9.
1
Redox rearrangement of 37
(1R,2R,5R)-1-Benzoyloxymethyl-2-methyl-5-isopropylcyclo-
hexane 68. δH 0.88 (3 H, d, J 6.7, CHMe2), 0.89 (3 H, d, J 6.7,
CHMe2), 1.03 (3 H, d, J 7.1, Me-2), 1.42–1.98 (10 H, complex,
CHMe2 and ring-H), 4.25 (1 H, dd, J 10.8 and 1.0, HA-1Ј), 4.37
(1 H, dd, J 10.8 and 5.2, HB-1Ј), 7.43 (2 H, m, Ph), 7.56 (1 H, m,
Ph), 8.05 (2 H, m, Ph); δC 19.7, 20.5, 20.8, 24.9, 26.3, 28.8, 29.1,
31.9, 38.6, 39.5, 74.2, 126.4, 129.4, 130.6, 132.8, 166.7.
(4S )-1-Benzoyloxmethyl-4-isopropylcyclohex-1-ene 38. Oil,
[α]18 Ϫ46.5 (c 1.95, CHCl3); δH 0.90 (3 H, d, J 6.7, CHMe2),
D
0.91 (3 H, d, J 6.7, CHMe2), 1.05–1.40 (3 H, m, ring-H), 1.50
(1 H, [octet], J 6.7 CHMe2), 1.81 (2 H, m, ring-H), 2.13 (2 H, m.
ring-H), 4.71(2 H, m, CH2OBz), 5.82 (1 H, m, H-2), 7.44 (2 H,
m, Ph), 7.55 (1 H, m, Ph), 8.06 (2 H, m, Ph); δC 18.2, 19.6, 19.9,
25.9, 26.7, 28.7, 32.2, 39.8, 69.0, 126.2, 128.3, 129.6, 130.4,
132.7(9), 132.8(1), 166.5. (Found: C, 79.2; H, 8.4. C17H22O2
requires C, 79.0; H, 8.6%).
X-Ray crystallography§
Single crystals were mounted on glass fibres and all geometric
and intensity data were taken from these samples on a Bruker
SMART APEX CCD diffractometer, using graphite-mono-
chromated Mo-Kα radiation (λ = 0.71073 Å). Data reduction
and integration were carried out with Bruker SAINTϩ soft-
ware33 and absorption corrections were applied using the pro-
gram SADABS.34 Structures were solved by direct methods and
developed using alternating cycles of least-squares refinement
and difference-Fourier synthesis. All non-hydrogen atoms
were refined anisotropically. Hydrogen atoms were placed in
calculated positions and their thermal parameters linked to
those of the atoms to which they were attached (riding model).
Absolute stereochemistry was not determined in the crystallo-
graphic experiments. Structure solution and refinement used
the SHELXTL PLUS V6.12 program package.35
Redox rearrangement of 42
2-exo-Benzoyloxynorbornane 4331. This ester was not isolated
in a pure state; it was identified by spectroscopic comparison
with the authentic compound prepared in 97% yield from exo-
norbornan-2-ol and benzoyl chloride in pyridine, as described
for 14. Oil; δH 1.20–1.27 (3 H, m, ring-H), 1.50 (1 H, m, ring-H),
1.54–1.70 (3 H, m, ring-H), 1.84 (1 H, ddd J 9.6, 7.1 and 2.4,
ring-H), 2.34 (1 H, m, ring-H), 2.45 (1 H, d, J 4.9, ring-H), 4.86
(1 H, brd, J 6.7, H-1), 7.42 (2 H, m, Ph), 7.54 (1 H, m, Ph), 8.02
(2 H, m, Ph); δC 24.3, 28.2, 35.4, 40.0, 41.6, 78.1, 128.2, 129.4,
130.8, 132.7, 166.2.
Redox rearrangement of 44
s-Butyl benzoate32 45 was isolated as an oil; δH 0.98 (3 H, t,
J 7.6, Me), 1.34 (3 H, d, J 6.3, Me), 1.69 (1 H, ddq, J 14.5, 7.6
and 6.3, CHAHBMe), 1.75 (1 H, ddq, J 14.5, 6.3 and 7.6,
CHAHBMe), 5.10 (1 H, [sextet], J 6.3, H-2), 7.43 (2 H, m, Ph),
7.55 (1 H, m, Ph), 8.05 (2 H, m, Ph); δC 9.7, 19.6, 28.9, 72.9,
128.3, 129.5, 130.9, 132.7, 166.3.
Crystal data for the benzil bis(acetal) 30. Data collected at
293 K. C34H42O4, M = 514.68, orthorhombic, space group
P212121, a = 10.6743(7), b = 16.4346(10), c = 16.6813(10) Å,
U = 2926.4(3) Å3, Z = 4, F(000) = 1112, Dc = 1.168 g cmϪ3
,
µ(Mo–Kα) = 0.075 mmϪ1, colourless crystal 0.48 × 0.48 ×
0.08 mm3. Full matrix least-squares refinement on 349 param-
eters gave R = 0.0426 (Rw = 0.1077) for 5908 independent
reflections [I > 2σ(I )] and R = 0.0507 (Rw = 0.1136) for all 6887
independent reflections for θ in the range 1.74 to 28.33Њ. The
final electron density map was featureless with the largest peak
Redox rearrangement of 56
(1S,2S,3R,6R)-2-Ethyl-3-methyl-6-isopropylcyclohexyl benz-
oate 56. Viscous oil, [α]D25 ϩ29.5 (c 1.65, CHCl3); δH 0.88 (3 H,
d, J 6.7, CHMe2), 0.93 (3 H, d, J 6.7, CHMe2), 0.99 (3 H, t,
J 7.5, CH2Me), 1.06 (3 H, d, J 7.1, Me-3), 1.30 (1 H, m,
CHMe2), 1.35–1.75 (9 H, complex, ring-H), 5.28 (1 H, [t], J 2.8,
H-1), 7.45 (2 H, m, Ph), 7.55 (1 H, m, Ph), 8.06 (2 H, m, Ph);
δC 12.3, 20.9(6), 21.0(3), 21.1, 21.2, 24.7, 27.6, 28.9, 30.3, 42.4,
45.6, 75.7, 128.4, 129.6, 131.0, 132.7, 165.9. (Found: C, 78.9; H,
10.1. C19H28O2 requires C, 79.1; H, 9.8%).
0.134 e ÅϪ3
.
Crystal data for the benzylidene acetal 56. Data collected at
150 K. C19H28O2, M = 288.41, orthorhombic, space group
P212121, a = 8.804(3), b = 10.274(3), c = 18.428(5) Å, U =
1666.9(8) Å3, Z = 4, F(000) = 632, Dc = 1.149 g cmϪ3, µ(Mo–Kα)
= 0.072 mmϪ1, colourless crystal 0.50 × 0.04 × 0.02 mm3.
Full matrix least-squares refinement on 194 parameters gave
R = 0.0579 (Rw = 0.1220) for 2927 independent reflections [I >
2σ(I )] and R = 0.0817 (Rw = 0.1329) for all 3899 independent
reflections for θ in the range 2.21 to 28.26Њ. The final electron
Redox rearrangement of 58
density map was featureless with the largest peak 0.250 e ÅϪ3
.
(1S,2S,3R,6S )-2-Ethyl-3-methyl-6-isopropylcyclohexyl benz-
25
oate 59. Viscous oil, [α]D Ϫ19.7 (c 1.59, CHCl3); δH 0.82(9)
(3 H, d, J 6.9, CHMe2), 0.83(2) (3 H, t, J 7.5, CH2Me), 0.88
(3 H, d, J 6.9, CHMe2), 0.95 (3 H, d, J 6.3, Me-3), 1.00–1.20
(3 H, m, ring-H and CH2Me), 1.30–1.52 (4 H, m, ring-H), 1.65–
1.80 (3 H, ring-H and CHMe2), 5.03 (1 H, [t], J 10.3, H-1), 7.45
(2 H, m, Ph), 7.55 (1 H, m, Ph), 8.07 (2 H, m, Ph); δC 8.4, 16.2,
19.5, 19.8, 21.2, 22.9, 26.5, 33.1, 34.7, 48.5, 49.3, 75.5, 128.3,
129.7, 130.6, 132.7, 166.2. (Found: C, 79.4; H, 10.1. C19H28O2
requires C, 79.1; H, 9.8%).
Crystal data for the benzylidene acetal 58. Data collected at
293 K. C19H28O2, M = 288.41, orthorhombic, space group
P212121, a = 8.5195(11), b = 10.0386(13), c = 19.370(2) Å, U =
1656.6(4) Å3, Z = 4, F(000) = 632, Dc = 1.156 g cmϪ3, µ(Mo–Kα)
§ CCDC reference numbers 216435–216437. See http://www.rsc.org/
suppdata/ob/b3/b309060b/ for crystallographic data in .cif or other
electronic format.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4083

View Article Online
= 0.073 mmϪ1, colourless crystal 0.50 × 0.10 × 0.08 mm3. Full
matrix least-squares refinement on 195 parameters gave R =
0.0535 (Rw = 0.1178) for 3275 independent reflections [I >
2σ(I )] and R = 0.0682 (Rw = 0.1251) for all 3974 independent
reflections for θ in the range 2.10 to 28.26Њ. The final electron
16 (a) D. R. Arnold, X. Y. Du and H. J. P. Delijser, Can. J. Chem., 1995,
73, 522; (b) H. D. Roth and T. Herbertz, J. Am. Chem. Soc., 1993,
115, 9804; (c) J. Oxgaard and O. Wiest, Eur. J. Org. Chem., 2003,
1454.
17 A. J. Fielding and B. P. Roberts, Tetrahedron Lett., 2001, 42, 4061.
18 S. J. Cole, J. N. Kirwan, B. P. Roberts and C. R. Willis, J. Chem. Soc.,
Perkin Trans. 1, 1991, 103.
density map was featureless with the largest peak 0.171 e ÅϪ3
.
19 (a) A. Bazbouz, H. Christol, J. Coste and F. Plénat, Bull. Soc. Chim.
Fr., 1978, II-305; (b) J. E. Taylor, Synthesis, 1985, 1142.
20 A. J. Fielding, P. Franchi, B. P. Roberts and T. M. Smits, J. Chem.
Soc., Perkin Trans. 2, 2002, 155.
Acknowledgements
We would like to express our thanks to Dr Abil E. Aliev for his
invaluable assistance with several of the NMR spectroscopic
assignments and we also wish to thank Professor D. Crich for
helpful comments. Support from the EPSRC for this work is
acknowledged.
21 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery,
Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam,
A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi,
V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala,
Q. Cui, K. Morokuma, N. Rega, P. Salvador, J. J. Dannenberg,
D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman,
J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong,
J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle and
J. A. Pople, GAUSSIAN 98, revision A.11.3, Gaussian, Inc.,
Pittsburgh PA, 2002.
Notes and references
1 B. P. Roberts and T. M. Smits, Tetrahedron Lett., 2001, 42, 137.
2 H.-S. Dang, B. P. Roberts, J. Sekhon and T. M. Smits, Org. Biomol.
Chem., 2003, 1, 1330.
3 B. P. Roberts, Chem. Soc. Rev., 1999, 28, 25.
4 Y. Cai, H.-S. Dang and B. P. Roberts, J. Chem. Soc., Perkin Trans. 1,
2002, 2449.
5 B. P. Roberts and T. M. Smits, Tetrahedron Lett., 2001, 42, 3663.
6 For example (a) S. Navarre, M. Vialemaringe, M.-J. Bourgeois,
M. Campagnole, E. Montaudon and J. S. Lomas, J. Chem.
Soc., Perkin Trans. 2, 2000, 2129; (b) M. Newcomb, R. Shen, Y. Lu,
M. J. Coon, P. F. Hollenberg, D. A. Kopp and S. J. Lippard, J. Am.
Chem. Soc., 2002, 124, 6879.
7 (a) B. G. Hazra, V. S. Pore, M. D. Chordia, B. B. Bahule and S. Basu,
J. Chem. Res. (S), 2001, 500; (b) B. G. Hazra, M. D. Chordia,
B. B. Bahule, V. S. Pore and S. Basu, J. Chem. Soc., Perkin Trans. 1,
1994, 1667.
8 (a) J. M. Gardiner, P. D. Crewe, G. E. Smith and K. T. Veal,
Chem. Commun., 2003, 618; (b) J. M. Gardiner, J. H. Chughtai
and I. H. Sadler, Tetrahedron: Asymmetry, 1998, 9, 599;
(c) J. H. Chughtai, J. M. Gardiner, S. G. Harris, S. Parsons,
D. W. H. Rankin and C. H. Schwalbe, Tetrahedron Lett., 1997, 38,
9043.
9 (a) H. Kiefer and T. G. Traylor, Tetrahedron Lett., 1966, 6163;
(b) G. D. Mendenhall, Tetrahedron Lett., 1983, 24, 451.
10 E. I. Miranda, M. J. Diaz, I. Rosado and J. A. Soderquist,
Tetrahedron Lett., 1994, 35, 3221.
11 (a) D. Crich, in Radicals in Organic Synthesis, ed. P. Renaud and
M. P. Sibi, Wiley-VCH, Weinheim, 2001, ch. 3.2, vol.2, p. 188;
(b) H. Zipse, J. Am. Chem. Soc., 1997, 119, 1087; (c) H. Zipse and
M. Bootz, J. Chem. Soc., Perkin Trans. 2, 2001, 1566.
12 Y. Cai and B. P. Roberts, Tetrahedron Lett., 2003, 44, 4645.
13 C. Roberts and J. C. Walton, Chem. Commun., 1984, 1109.
14 Y. Cai and B. P. Roberts, J. Chem. Soc., Perkin Trans. 2, 2002, 1858.
15 M. Newcomb, A. G. Glenn and M. B. Manek, J. Org. Chem., 1989,
54, 4603.
22 (a) G. Stork and J. d’Angelo, J. Am. Chem. Soc., 1974, 96, 7114;
(b) R. Murphy and R. H. Prager, Aust. J. Chem., 1976, 29, 617.
23 (a) H. C. Brown and C. P. Garg, J. Am. Chem. Soc., 1961, 83, 2952;
(b) A. L. J. Beckwith and G. Phillipou, Aust. J. Chem., 1976, 29, 877.
24 N. Irako, Y. Hamada and T. Shioiri, Tetrahedron, 1995, 51, 12731.
25 H. C. Brown, J. V. N. V. Prasad and M. Zaidlewicz, J. Org. Chem.,
1988, 53, 2911.
26 M. Kinugasa, T. Harada and A. Oku, J. Am. Chem. Soc., 1997, 119,
9067.
27 (a) F. Fringuelli, O. Piermatti, F. Pizzo and A. M. Scappini, Gazz.
Chim. Ital., 1995, 125, 195; (b) K. Gollnick and G. Schade, Justus
Liebigs Ann. Chem., 1969, 133.
28 (a) G. Accrombessi, P. Geneste, J.-L. Olive and A. A. Pavia,
Tetrahedron, 1981, 37, 3135; (b) N. A. Nelson and G. A. Mortimer,
J. Org. Chem., 1957, 22, 1146.
29 (a) A. L. J. Beckwith and G. Phillipou, Aust. J. Chem., 1976, 29,
1277; (b) Á. Cantín, C. Lull, J. Primo, M. A. Miranda and E. Primo-
Yúfera, Tetrahedron: Asymmetry, 2001, 12, 677.
30 T. Hirata, Bull. Chem. Soc. Jpn., 1972, 45, 3169, 3458.
31 D. Despeyroux, R. B. Cole and J.-C. Tabet, Org. Mass. Spectrom.,
1992, 27, 300.
32 (a) K. Nakamura and K. Takenaka, Tetrahedron: Asymmetry, 2002,
13, 415; (b) J. E. Kaminska, Z. J. Kaninski and J. Góra, Synthesis,
1999, 593.
33 Area detector control and data integration and reduction software,
Brüker AXS, Madison, MI, USA, 2001.
34 G. M. Sheldrick, SADABS, University of Göttingen, Göttingen,
Germany, 1997.
35 SHELXTL PLUS V6.12, Brüker AXS, Madison, MI, USA, 2001.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 4 0 7 3 – 4 0 8 4
4084