Regio- and stereoselectivity in the solvomercuriation and intramolecular alkoxymercuriation of cyclic unsaturated alcohols

Article abstract of DOI:10.1039/b006963i

The regio- and stereoselectivity in the solvomercuriation and intramolecular alkoxymercuriation of several cyclic olefinic alcohols were examined. The regioselecuvity is controlled mainly by electronic factors, while the stereoselectivity is controlled by

Full text of DOI:10.1039/b006963i

View Article Online / Journal Homepage / Table of Contents for this issue
Regio- and stereoselectivity in the solvomercuriation and
intramolecular alkoxymercuriation of cyclic unsaturated alcohols
Yasuhisa Senda,* Satomi Takayanagi, Tomomi Sudo and Hiroki Itoh
1
Department of Material and Biological Chemistry, Faculty of Science, Yamagata University,
Yamagata 990-8560, Japan
Received (in Cambridge, UK) 25th August 2000, Accepted 22nd November 2000
First published as an Advance Article on the web 10th January 2001
The regio- and stereoselectivity in the solvomercuriation and intramolecular alkoxymercuriation of several
cyclic olefinic alcohols were examined. The regioselectivity is controlled mainly by electronic factors, while the
stereoselectivity is controlled by steric factors as well as electronic ones. The optimised structure of the mercurinium
ion intermediate suggests that the attractive interaction between the hydroxy group in the molecule and the
mercurinium ion moiety affects the selectivity.
Introduction
The oxymercuriation–demercuriation procedure provides a
convenient synthetic method for effecting the Markovnikoff
hydration of the carbon–carbon double bond. The reaction of
(E)-pent-2-ene, which has an evenly substituted carbon–carbon
double bond with an equal number of substituents on each sp2
carbon, preferably yielded pentan-2-ol, but there is no clear
explanation for this regioselectivity.1 In cyclic systems, the regio-
and stereoselectivity of alkylcyclohexenes are controlled by
the torsional strain between the allylic alkyl group and the
incoming nucleophile to the mercurinium ion intermediate.2
Since 3-methylcyclohexene gave mainly trans-3-methylcyclo-
hexanol, Brown pointed out that the regio- and stereochemistry
are determined by the difference in the relative stabilities of the
two possible transition states caused by the gauche interactions
between the forming bond of the nucleophile with the mercur-
inium ion moiety and the bonds in the molecule.3 In this article,
we present the results of a study designed to evaluate the role of
the polar substituent on the regio- and stereochemistry in the
solvomercuriation of cyclic olefinic alcohols. The intramolecular
alkoxymercuriation of cyclohexenylalkanols was also examined.
Cyclohexenols such as cyclohex-2-enol 1, 2-methylcyclohex-2-
enol 2, 3-methylcyclohex-2-enol 3, 3,cis-5-dimethylcyclohex-2-
enol 4 and cyclopent-2-enol 5 were employed. 2-Isopropylidene-
cyclohexanol 6 and 2-isopropylidenecyclopentanol 7 were used
as substrates which have an equally substituted exocyclic
double bond. The regio- and stereoselectivity obtained from the
cyclic allylic alcohols were compared with those from 3-methyl-
cyclohexene 8 and 1-isopropylidene-2-methylcyclohexane 9.
The intra- and intermolecular alkoxymercuriations of cyclo-
hex-3-enylmethanol 10, 4-methylcyclohex-3-enylmethanol 11
and 2-(cyclohex-2-enyl)ethanol 12 were also examined (Fig. 1).
Methoxymercuriation was applied to the unsaturated alcohols,
and oxymercuriation occurred on the unsaturated hydro-
carbons. The analyses of the regio- and stereoselectivity of the
reaction were performed on the demercuriated products which
were obtained by NaBH4 reduction.
Fig. 1 Olefins.
activation between the routes to the diastereomeric products
is almost constant. The differences in the free energies of acti-
vation giving two diastereomeric products are adequately
approximated by the differences in the enthalpies of activation.
The regioselectivity in the nucleophilic attack on the mercurin-
ium ion intermediate is controlled generally by the electron
density of the attacking site. We previously showed that the
regioselectivity varied depending on the substituent(s) on the
phenyl group in the methoxymercuriation of substituted phenyl-
alkenols.4 On the other hand, the regioselectivity as well as the
stereoselectivity in the solvomercuriation of cyclic compounds
such as the cyclohexene derivatives is considered to be con-
trolled by not only electronic but also steric factors, that is, (i)
the electron density of the mercurinium ion moiety, (ii) the
energetics of the two different routes to the products via skew-
boat-like and chair-like transition states, when the ring opening
of the mercurinium ion takes place in the diaxial mode, and (iii)
the local steric environment of the approaching nucleophile
with respect to the substrate (steric and/or torsional strain). In
addition, both the regio- and stereoselectivity of the intra-
molecular alkoxymercuriation forming the new cyclic structure
are also controlled by (iv) the steric relationship of the ring
Results and discussion
In order to examine the consistency of the regio- and stereo-
selectivity, the reaction was performed at different temperatures
and for different times. No appreciable differences in results
were observed upon varying the reaction conditions (Table 1,
footnote a). This indicates that the difference in the entropies of
270
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b006963i

View Article Online
Table 1 Selectivity in solvomercuriation and intramolecular alkoxy-
Table 2 Calculated heat of formation for reaction species
mercuriationa
Heat of
Heat of
Regioselectivityc
Stereoselectivityd
formation/
formation/
Yieldb
(%)
Compound
kcal molϪ1
Compound
kcal molϪ1
Compound
α
β
cis
trans
1a
Ϫ45.94
Ϫ46.25
27.88
16.45
27.56
7ag
Ϫ63.43
Ϫ63.13
9.60
0.48
17.01
Methoxymercuriation
1e
7eh
1ac1
1at1
1ec1
1et1
1et3
1ec4
1c
7ac1i
7at1j
7ec1k
7et1l
7ec3m
7et3n
7co
1
78
95
0
2
100
98
30
α 18
70
80
1
11
87
1e
β
1
89f
13f
0f
20.11
0.49
2
80
76
69
92
42
61
100
0
0
100
100
100
100
100
84
Ϫ174.57
Ϫ171.98
Ϫ174.30
Ϫ178.33
Ϫ39.03
36.16
Ϫ178.26
Ϫ183.36
Ϫ182.25
Ϫ185.84
Ϫ9.93
Ϫ10.30
58.78
3
4
0
0
0
0
100
5
19
62
59
α 12
β
81
38
41
4
1t
5
7tp
8a
8e
6
7
5c1a
5t1b
5c3c
5t3d
5ce
5tf
8g
16
25.92
8ac1
8at1
8ec1
8et1
8at3
8ec4
8c
6
79
Ϫ165.28
Ϫ170.73
Ϫ169.19
Ϫ171.25
Ϫ69.12
Ϫ65.46
Ϫ0.15
Ϫ5.01
1.08
3.05
Ϫ187.18
Ϫ186.43
Ϫ187.40
Ϫ186.95
51.59
58.95
54.14
9
10
5
43
52
0
48
α
1
51
2
3
β 46
97h
Ϫ138.98
Ϫ134.94
Ϫ135.69
Ϫ141.36
Ϫ32.68
Ϫ27.72
32.37
11
65
100
6a
6e
Intramolecular alkoxymercuriation
6ac1
6at1
6ec1
6et1
6ac3
6et3
6c
8t
10
11
12
30
48
20
100
0
100
0
100
0
100
100
100
0
0
0
9a
9e
9ac1
9at1
9ec1
9et1
a Reaction time 1 h (solvomercuriation) or 1.5 h (intramolecular
alkoxymercuriation). Temperature 25 ЊC. No apparent discrepancies of
regio- and stereoselectivity appeared depending on the reaction condi-
tions (time 24 h, temperature 0 and 40 ЊC). The discrepancies of selectiv-
ity were within 3% except in the case of 6 (6%, time 24 h, temp. 25 ЊC).
b Isolated yield. c The α is the proximal and the β is the distal carbon
atom of the mercurinium ion moiety to the allylic substituent. d The
relation between the two substituents on the ring unless otherwise indi-
cated. e Ref. 12. Oxymercuriation. Accompanied by 2% cyclohexane-
1,2-diols. Reaction time not mentioned. No appreciable discrepancy
was observed for the selectivity between the oxymercuriation and the
methoxymercuriation. f The relation between OH and OMe. g Ref. 3.
The reaction temperature, time and yield not reported. h The relation
between CH2OH and OMe.
33.91
36.20
40.63
6t
a Mercurinium ion, OH, Hg: opposite side. b Mercurinium ion, OH,
Hg: same side. c 5-Hydroxy-2-methoxycyclopentylmercuric acetate,
OMe, OH, HgOAc: equatorial. d 5-Hydroxy-2-methoxycyclopentyl-
mercuric acetate, OH: axial, OMe, HgOAc: equatorial. e 5-Hydroxy-2-
methoxycyclopentylmercuric acetate, OMe, OH, HgOAc: axial. f 5-
Hydroxy-2-methoxycyclopentylmercuric acetate, OH: axial, OMe,
HgOAc: equatorial. g 7, OH: axial. h 7, OH: equatorial. i Mercurinium
ion, OH: axial, OH, Hg: opposite side. j Mercurinium ion, OH: axial,
OH, Hg: same side. k Mercurinium ion, OH: equatorial, OH, Hg:
opposite side. l Mercurinium ion, OH: equatorial, OH, Hg: same side.
m 2-Hydroxy-1-(2-methoxypropan-2-yl)cyclopentylmercuric
acetate,
-CMe2OMe: axial, OH, HgOAc: equatorial. n 2-Hydroxy-1-(2-
methoxypropan-2-yl)cyclopentylmercuric acetate, HgOAc: axial,
-CMe2OMe, OH: equatorial. o 2-Hydroxy-1-(2-methoxypropan-2-
yl)cyclopentylmercuric acetate, OH, HgOAc: axial, -CMe2OMe:
equatorial. p 2-Hydroxy-1-(2-methoxypropan-2-yl)cyclopentylmercuric
acetate, OH, -CMe2OMe: axial, HgOAc: equatorial.
junction, if a condensed-ring structure is constructed, and (v)
the number of atoms in the newly formed ring.
Methoxymercuriation of cyclohexenols
We accepted the premise that the oxymercuriation of olefins
involves the rapid pre-equilibrium formation of a mercurinium
ion as an unstable intermediate followed by the rate- and prod-
uct-determining attack of a nucleophile.5 The stereochemistry
of the reaction in this study is discussed on the basis of the fact
that the attack of the nucleophile occurs trans to the mercury
and proceeds with preferential diaxial opening of the mercurin-
ium ion in the oxymercuriation of cyclohexenols, similar to the
case of simple cyclohexenes.6 The reaction of 1, which has
the evenly substituted carbon–carbon double bond, gave
the product in which the methoxy group was introduced at the
β-carbon (Table 1). The terms α and β designate a proximal and
a distal carbon of a mercurinium ion moiety from a substituent
on a six-membered cyclic system, respectively. A plausible reac-
tion pathway for 1 is illustrated in Scheme 1. There are two
conformers for 1, 1a and 1e, the hydroxy group of which is at
the pseudo-axial position in the former and the pseudo-
equatorial in the latter. The structure optimisation of 1 using
MO calculations indicated that 1a is more stable than 1e by 0.3
kcal molϪ1 † (Table 2).7 The mercuric (HgII) acetate approaches
the two diastereofaces of the carbon–carbon double bond for
each conformer to give four mercurinium ions, 1ac1, 1at1, 1ec1
and 1et1. We tried to calculate the optimised structures of these
diastereomeric mercurinium ions, the corresponding transition
Table 3 The distance between the mercury atom and the hydroxy
group oxygen in the mercurinium ion
Mercurinium ion
Hg ؒ ؒ ؒ OH Distance/Å
1ac1
1at1
1ec1
1et1
6ac1
6at1
6ec1
6et1
4.12
2.19
4.07
2.41
4.36
2.17
2.21
4.00
states and the products, the methoxymercurials. At present, the
transition-state structures of the reaction cannot be obtained.
The mercurinium ion 1at1 whose hydroxy group is placed at the
pseudo-axial position and on the same side as the mercurinium
ion moiety is the most stable among the four. This infers that the
close contact between the hydroxy oxygen and the mercurinium
ion moiety results in the generation of the attractive inter-
action. This is supported by the interatomic distance between
the oxygen atom and the mercury atom (Table 3). In this reac-
tion a ligand exchange is also expected between the acetate ion
on the mercury and the alkoxide in the solvent or the substrate
† 1 cal = 4.184 J.
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278
271

View Article Online
Scheme 1 Reaction pathway of 1 and 8.
Table 4 The calculated formal charge and the coefficient of the
LUMO of carbons in the mercurinium ion moiety
before addition. In order to compare these two intermediates,
viz. the mercurinium acetate with which the hydroxy group is
undergoing intramolecular interaction (1at1) and the ligand-
exchanged species, the optimisation was performed on these
two different structures. For simplification, the calculation was
performed on allyl alcohol as a model compound. The result
indicated that the former is far more stable than the latter (the
difference in energy is 14.26 kcal molϪ1) probably because of
the difference in the distances between the mercury atom and
the alcoholic oxygen (the former: 2.21 Å and the latter: 1.95
Å). It is likely that the contribution of the species in which
the ligand is exchanged with the intramolecular hydroxy
group adds little to the reaction. When the hydroxy group is
placed pseudo-equatorially, then 1et2, whose hydroxy group
is placed on the same side as the mercurinium ion, is ener-
getically more stable compared with 1ec1 whose hydroxy
group is placed on the opposite side to that of the mercurin-
ium ion moiety. This also indicates the presence of the inter-
action between the hydroxy group and the mercurinium ion
moiety.
Formal charge
|Coefficient of LUMO|a
Mercurinium
ion
α
β
α
β
1ac1
1at1
1ec1
1et1
6ac1
6at1
6ec1
6et1
8ac1
8at1
8et1
8ec1
8et1
10a1
12a1
Ϫ0.1714
Ϫ0.1452
Ϫ0.1734
Ϫ0.1401
Ϫ0.2278
Ϫ0.1798
Ϫ0.1539
Ϫ0.2203
Ϫ0.1098
Ϫ0.1153
Ϫ0.1140
Ϫ0.1292
Ϫ0.0961
Ϫ0.1076
Ϫ0.0762
Ϫ0.0755
Ϫ0.0802
Ϫ0.0791
0.0000
Ϫ0.0543
Ϫ0.0941
Ϫ0.0124
Ϫ0.1069
Ϫ0.0990
Ϫ0.1040
Ϫ0.0964
Ϫ0.1158
Ϫ0.1075
0.0867
0.0321
0.1587
0.0150
0.0322
0.0300
0.1146
0.0193
0.1205
0.0832
0.1780
0.0512
0.1297
0.1210
0.2263
0.3336
0.1643
0.2892
0.3339
0.2563
0.1454
0.2715
0.1886
0.2175
0.1385
0.2381
0.1848
0.1902
a 2pπ component.
Since the rate-determining step of the oxymercuriation is the
introduction of a nucleophile to the mercurinium ion, then
species 1ac1, 1at1, 1ec1 and 1et1 are in pre-equilibrium. Since
the opening of the three-membered cyclic mercurinium ion by
attack of the nucleophile occurs in the diaxial mode, the
transition-state structures from 1at1 and 1ec1 take skew-boat-
like conformations, while those from 1ac1 and 1et1 take chair-
like conformations. These transition-state species will then
change to the corresponding, more stable conformers of the
two diastereomeric products. The trans product will be formed
from 1at1 and 1et1; the former is more stable than the latter by
3.66 kcal molϪ1. According to the Curtin–Hammett principle
the product ratios depend only on the energy difference in the
diastereomeric transition states.8 Since the energy difference
between the chair and skew-boat conformations has been esti-
mated to be about 5.5 kcal molϪ1,9 the energy of 1et2 will be
smaller than that of 1at2 at the transition state and so trans-
isomer formation mainly proceeds from the mercurinium ion
1et1 via 1et2. Likewise, cis-product formation proceeds from
the mercurinium ion, 1ac1 via 1ac2.
Since it is considered that the regio- and stereoselectivity in
the solvomercuriation of six-membered cyclic olefins are pri-
marily controlled by the electronic situation of the mercurinium
ion moiety and the deformation of the ring during the reaction,
the formal charges and coefficients of the LUMO of the
carbons in the mercurinium ion moiety were calculated using
the MO method (Table 4). The formal charges of the β-carbons
for both 1ac1 and 1et1 are apparently higher than those of the
α-carbons, and the coefficients of the LUMO of the carbons in
the mercurinium ion moiety for 1et1 to the trans product are
higher than in that of 1ac1 to the cis product (Table 4). These
are in accord with the experimental evidence that the nucleo-
phile attacked at the β-position and that the trans product was
preferably formed.
The stabilisation energies obtained using the modified
Klopman’s equation, which consists of the Coulombic term
and the frontier orbital term, give an additional clue to the
272
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278

View Article Online
Table 5 Stabilisation energy calculated by the modified Klopman’s
equation (kcal molϪ1
β-isomer is larger than that of the cis β-isomer is one of the
)
reasons why the formation of the trans product is preferred.
The regiochemistry in the methoxymercuriation of 2, 3, and
4 followed the Markovnikoff rule. The methoxy group was
introduced at the carbon with a methyl group. When a
nucleophile is introduced at the 2-position, the cis-isomers were
preferentially obtained, and at the 3-position the trans ones
were preferred.
a
b
Compound
Stereoselectivity
∆EC
∆EF
∆E
∆∆E
1
5
6
7
8
cis-β
trans-β
cis-β
trans-β
cis-β
trans-β
cis-β
trans-β
cis-α
cis-β
trans-α
trans-β
19
81
30
70
59
41
62
38
12
4
Ϫ0.09
Ϫ0.09
Ϫ0.07
Ϫ0.10
0.00
Ϫ0.06
0.06
Ϫ0.06
Ϫ0.13
Ϫ0.12
Ϫ0.15
Ϫ0.11
0.15
0.54
0.32
0.50
0.55
0.32
0.58
0.35
0.17
0.10
0.01
0.29
0.06
0.45
0.25
0.40
0.55
0.26
0.64
0.29
0.04
0.40
0.15
0.29
Methoxymercuriation of 2-isopropylidenecyclohexanol 6
0.35
0.05
The nucleophile was introduced at the β carbon of the olefinic
double bond in the methoxymercuriation of 6 which has an
exocyclic double bond, and the stereochemistry of the preferred
product was cis. The presumed reaction pathways are illus-
trated in Scheme 2. The geometry optimisation of the two pos-
sible conformations for 6 by theoretical calculations indicated
that the one (6a) in which the hydroxy group takes the axial
position is more stable than the alternative (6e) by 4.5 kcal
molϪ1. This may be due to the torsional strain between the
hydroxy group and the neighboring exocyclic carbon–carbon
double bond. The mercuric acetate attacks these two con-
formers to give four different mercurinium ion intermediates,
6ac1, 6at1, 6ec1 and 6et1. The conformer whose hydroxy group
is placed on the same side as the mercurinium ion is more stable
than the alternative by 4.8 kcal molϪ1 when the hydroxy group
is in the axial direction. This suggested the presence of an inter-
action between the hydroxy group and the mercurinium cation.
In contrast to the conformer whose hydroxy group takes the
axial position, the conformer (6ec1) whose hydroxy group is
placed on the opposite side to that of the mercurinium ion
moiety is more stable than the alternative (6et1). The distance
between the hydroxy group and mercury atom is shorter in 6ec1
(2.17 Å) than in 6et1 (4.00 Å).
Ϫ0.02
Ϫ0.14
0.18
6
79
0.32
a Coulombic term. b Frontier orbital term.
stereoselectivity of the reaction,10 as shown in eqn. (1) in
∆E = ϪQrQs/Rrsؒε ϩ
2(crcsؒ∆β)2/|ELUMO (mercurinium ion) Ϫ EHOMO (MeOH)
|
(1)
which Qr is the charge of the methanol oxygen (Ϫ0.309), Qs is
the charge of the β-carbon atom of the mercurinium inter-
mediate, Rrs is the distance between the methanol oxygen and
the β-carbon of the mercurinium intermediate at the transition
state of 2.70 Å (although it was impossible to find the transition
state of the reaction for 1 by MO calculations, we succeeded in
determining the transition state of the more simple substrate,
cyclohexene, in which the distance between the methanol
oxygen and the β-carbon atom is 2.70 Å), ε is the relative
permittivity of methanol (32.6), cr is the AO coefficient of the
methanol oxygen (0.805), cs is the AO coefficient of the
β-carbon of the mercurinium ion intermediate, ∆β is the C–O
resonance integral at 2.70 Å (Ϫ1.02), ELUMO is the energy level
of the LUMO for the mercurinium ion intermediate, and
EHOMO is the energy level of the HOMO for methanol (Ϫ11.1).
The calculated stabilising energies are summarised in Table 5.
Since the product distribution is controlled by the energy differ-
ence in the transition states, the fact that the ∆E for the trans
Since the introduction of the nucleophile to these four
diastereomeric mercurinium ion intermediates of 6 is con-
sidered not to cause an appreciable deformation in the ring
structure, the reaction will mainly produce the cis isomer from
6ac1 and the trans from 6at1, which are energetically more
stable than the corresponding alternatives, 6ec1 and 6et1,
respectively. The heats of formation for the two isomeric
Scheme 2 Reaction pathway of 6.
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278
273

View Article Online
product oxymercurials are reversed in magnitude compared
with the corresponding substrates, the mercurinium ion inter-
mediates, in this reaction step. The preferred formation of the
cis product indicates that the energy of 6ac2 is smaller than that
of 6at2 in the transition state; even the heat of formation for
6at1 is smaller than that of 6ac1. Again, the formal charges of
the β-carbons of both 6ac1 and 6at1 are larger than those of
the corresponding α-carbons and the coefficients of the LUMO
of 6ac1 are larger than those of 6at1 (Table 4). The stabilisation
energy for the cis β-product is higher than that for the trans
β-product 6t.
Fig.
2
Calculated parameters of cations from 9. The LUMO
coefficient indicates the 2pπ component.
Oxymercuriation of 3-methylcyclohexene 8 and 1-isopropyl-
idene-2-methylcyclohexane 9
the coefficients of the LUMO of the mercurinium ion moiety in
the optimised structure of 10a1 are tabulated in Table 4. The
substituents, which are separated from the double bond, do not
appear to contribute to the regioselectivity of the nucleophilic
attack. It has been reported that the oxymercuriation of
homoallylic cyclohexenol, cyclohex-3-enol, predominantly gave
the product in which the nucleophile is introduced at the
4-position and the stereochemical relationship between the
hydroxy group and the incoming nucleophile is trans.6 When
the substituent at the homoallylic position is the tert-butyl
or hydroxymethyl group, these substituents are fixed in the
equatorial position on the mercurinium ion intermediate, while
the hydroxy group is placed in axial as well as equatorial
positions. The interaction between the mercurinium ion and
the hydroxy group at the axial position may affect the stereo-
chemistry of the reaction.
To compare the role of the methyl and the hydroxy groups,
the regio- and stereoselectivity in the oxymercuriation of 8
and 9 were examined. The oxymercuriation of 8 has already
been reported, by Brown et al. in 1979.3 The reaction path-
ways for 8 are illustrated in Scheme 1. The heats of form-
ation for the two conformers of 8 and four diastereomeric
mercurinium ion intermediates were calculated using the MO
method (Table 2). Similar to the case of 1, the reaction of 8
will also proceed mainly through 8et1 to the trans and
through 8ac1 to the cis oxymercurial. Again, the calculated
formal charges, the coefficients of the LUMO of the possible
mercurinium ion intermediates, and the stabilisation energies
were calculated and are consistent with the experimental
results.
The reaction of 9 occurred with much difficulty. The conver-
sion of the substrate to the product was less than 10% and most
of the starting olefin was recovered unchanged. The product
was not the expected oxymercurial but was instead 2-(6-methyl-
cyclohex-1-enyl)propan-2-ol 9a4 which indicates the double-
bond migration and the introduction of the hydroxy group
during the reaction. In general, the allylic position is oxidised
by mercuric acetate to give unsaturated acetates in acetic acid
at high temperature.11 We confirmed that 9 was oxidised at
the allylic position under the oxymercuriation conditions. We
would like to propose the following reaction pathways shown in
Scheme 3.5 Compound 9 exists in two conformations (9a and
9e), on which the attack by mercuric acetate produces four dias-
tereomeric mercurinium ion intermediates (9at1, 9ac1, 9et1 and
9ec1). The oxymercuriation should have occurred by the intro-
duction of a water molecule, whereas the elimination of the
allylic hydrogen and the cleavage of the C–Hg bond occurs to
give the allylic mercurials. If this elimination is supposed to be
of the E2 type, the axial hydrogen at the C-6 position of 9ac1
participates to give allylic mercurial (9a2), followed by the
formation of the allylic cation (9a3) by the elimination of Hg0.
The nucleophile then attacks the more stable tertiary cation to
form 9a4. Since there are two axial hydrogens at C-2 and C-6 in
9et1, the formed allylic cations should be 9a3 and 9e3, which
produce two allylic alcohols, respectively. Consequently, four
kinds of allylic alcohols, one of which consists of two stereo-
isomers, would be expected. Only the formation of 9a4 is con-
sistent with the theoretical parameters of the two allylic cations,
that is, the nucleophilicity is the highest and the coefficient of
the LUMO is the largest at the β position of 9a3 among the
allylic cations (Fig. 2).
The methoxymercuriation of 11 followed the Markovnikoff
rule to give only 4-methoxy products in which the cis (CH3O ؒ ؒ ؒ
CH2OH) isomer was 97% of the total and the trans isomer was
3%. The reaction of 12 in methanol did not proceed to give the
methoxylated product.
Intramolecular alkoxymercuriation
The intramolecular alkoxymercuriations of 10, 11 and 12 were
examined in 2-methylpropan-2-ol. Although the reaction of
10 was expected to produce two types of compound, only
6-oxabicyclo[3.2.1]octane was obtained after demercuriation.
The methoxymercuriation of 10 did not show any regioselectiv-
ity, while the intramolecular reaction proceeded with the attack
of the nucleophile in the molecule to the α position to form the
tetrahydrofuran ring. On the other hand, the reaction of 11
followed the Markovnikoff rule to give only 1-methyl-2-
oxabicyclo[2.2.2]octane. The different selectivity depending on
the intra- or intermolecular reaction of 10 indicates that the
nucleophilic reaction is controlled not only by the electronic
effect but also by other factors. The possible reaction pathways
for 10 and 11 are shown in Scheme 4. In the intramolecular
alkoxymercuriation, the addition of mercuric acetate to the two
diastereofaces of the olefinic bond of each of the two cyclo-
hexene conformers results in the formation of four kinds of
mercurinium ion intermediates. In order to form the new ring
system, the hydroxymethyl groups of 10 and 11 have to be
placed in the axial position and the attack should be in the cis
direction to it because the intramolecular nucleophile at the
equatorial position would have difficulty in attaining a trajec-
tory to attack the mercurinium ion moiety. The nucleophilic
attack on 11 follows the Markovnikoff rule and the reaction is
forced to proceed via the energetically unfavorable skew-boat-
like transition state. The nucleophilic attack on 10 occurred
only at the α carbon of the mercurinium ion moiety, and the
reaction proceeds via the energetically favorable chair-like tran-
sition state (10a2) to form the five-membered tetrahydrofuran
ring. The energy-minimised structure of 10a1 in which the
hydroxymethyl group is in the axial position and on the oppos-
ite side of the mercurinium ion moiety indicated that the formal
charge is higher at the α carbon than at the β, and that the
Methoxymercuriation of other cyclohexenyl compounds
The methoxy group was introduced at the cis 4-position in 46%
yield and the trans 3-position in 51% yield accompanied by
trans-4-methoxycyclohexanemethanol (2%) and cis-3-methoxy-
cyclohexanemethanol (1%) in the methoxymercuriation of
10, in which the hydroxymethyl group is at the homoallylic
position. This result was analogous to the oxymercuriation
of 4-tert-butylcyclohexene whose tert-butyl group is at the
homoallylic position.7 The calculated formal charges and
274
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278

View Article Online
Scheme 3 Reaction pathway of 9.
Scheme 4 Reaction pathway of 10 and 11.
coefficient of the LUMO of the β carbon is higher than that of
the α.
Intramolecular nucleophilic reaction occurred to give cis-7-
definitive factor determining the selectivity because the formal
charges of both carbons of the mercurinium ion moiety for
12a1 showed almost the same values, and the coefficient of the
LUMO was larger at the β carbon than at the α one (Table 4).
The results obtained from 10 and 12 indicate that the regio-
selectivity is not mainly determined by electronic factors, but
by steric ones, that is, the formation of the five-membered ring
is kinetically preferable to that of the six-membered ring.12
Baldwin proposed that the opening of three-membered rings to
form cyclic structures generally follows the exo-mode.13
oxabicyclo[4.3.0]nonane in the reaction of 12. The formation of
a new oxygen-containing ring was examined using a Dreiding
model. When the hydroxyethyl group is placed at the pseudo-
axial position, the intramolecular nucleophile can possibly
attack both the α and β carbon atoms. On the other hand, when
the hydroxy group is at the pseudo-equatorial position, the
hydroxy oxygen can be introduced only at the α carbon because
the stereochemistry of the ring junction of the forming
oxabicyclo[3.3.1]nonane-type compound should be diaxial
(Scheme 5). Similarly to the case of the intramolecular
alkoxymercuriation of 10, the electronic situation cannot be the
Methoxymercuriation of five-membered cyclic compounds
The regio- and stereochemistry of the oxymercuriation of 53
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278
275

View Article Online
Scheme 5 Reaction pathway of 12.
and 7 showed similar trends to those seen in the corresponding
Experimental
six-membered-ring compounds. The methoxymercuriation of 5
exclusively gave the product in which the methoxy group was
introduced at the β-carbon, and the stereochemistry was 33%
cis and 67% trans. The reaction of 7 also gave only the product
in which the methoxy group was introduced at the β-carbon,
and the stereochemistry was 65% cis and 35% trans. The
energy-minimised structure of 5 exists as only one conformer
whose ring is nearly flat. The two diastereomeric addition
products (the one: the hydroxy group and mercurinium ion
moiety are placed on the same side, and the other: the hydroxy
group and mercurinium ion moiety are placed on opposite
sides) of mercuric acetate from 5 had a difference in their heats
of formation of 10.24 kcal molϪ1. The former is more stable
than the latter. On the other hand, there are two conformers
for 7, one of which has a pseudo-axial hydroxy group and the
other has a pseudo-equatorial one. The difference in the heats
of formation for these two diastereomers is only 0.30 kcal
molϪ1, which is smaller than that of 9. Among the four mer-
curinium ions obtained by the addition of mercuric acetate,
the two which have the hydroxy group and the mercurinium
ion moiety on the same side are apparently more stable than
the others.
General
13C NMR and H NMR spectra were obtained with a JEOL
JNM α-400 instrument in the pulse Fourier mode. J-Values are
in Hz. Gas chromatographic analyses were performed on a
Shimadzu model GC-8AIF with a Carbowax 20M chemical
bonded silica capillary column (0.25 mm × 25 m) at a column
temperature of 120 ЊC. The isomeric ratios were determined by
the comparison of NMR and GLC analyses with authentic
samples. Mps are uncorrected. Methanol for methoxymercuri-
ation was distilled from Mg foil. THF for oxymercuriation was
simply distilled. Solvents for column chromatography ethyl
acetate was dried over K2CO3 and hexane was dried over CaCl2,
and both were distilled before use. Column chromatography
was achieved using Nacalai Tesque silica gel 60 (70–230 mesh).
SiO2–AgNO3 was prepared by mixing SiO2 and aq. AgNO3
followed by drying. This was activated at 120 ЊC for 2 h before
use.
1
Substrates
Cyclohex-2-enol14 1, 2-methylcyclohex-2-enol15 2, 3-methyl-
cyclohex-2-enol16 3, 3,cis-5-dimethylcyclohex-2-enol17 4,
cyclopent-2-enol18 5, 2-isopropylidenecyclopentanol19 7, cyclo-
hex-3-enylmethanol20 10, 4-methylcyclohex-3-enylmethanol21
11, and 2-(cyclohex-2-enyl)ethanol22 12 have been previously
reported.
Conclusions
(1) The regioselectivity in the solvomercuriation of an evenly
substituted olefin is mainly controlled by electronic conditions
(shown by the formal charge and the coefficient of the LUMO)
of the mercurinium ion intermediate. In particular, MO con-
siderations using the modified Klopman’s equation shows that
the contribution of the frontier orbital term is larger than that
of the Coulombic term in determining the stabilisation energy.
In the intramolecular alkoxymercuriation, the regioselectivity
is controlled not only by electronic factors but also by steric
factors.
(2) The stereoselectivity of the solvomercuriation of evenly
substituted olefins is controlled by both steric and electronic
factors. The degree of stabilisation by the interaction between
the substrate and the nucleophile in the transition states also
contributes to the stereochemistry.
2-Isopropylidenecyclohexanol 6
2-Isopropylidenecyclohexanol was prepared by the LiAlH4
reduction of 2-isopropylidenecyclohexanone,23 mp 56 ЊC
(Found: C, 77.2; H, 11.3. C9H16O requires C, 77.1; H, 11.5%);
δH (400 MHz; CDCl3) 1.15–1.26 (1H, m), 1.40 (1H, s), 1.42–
1.46 (2H, m), 1.67 (3H, s, CH3), 1.73 (3H, s, CH3), 1.75–1.80
(2H, m), 1.92 (1H, d, J 13.4), 2.10 (1H, t, J 9.1), 2.78 (1H, t,
J 13.9), 4.83 (1H, br t, J 1.79, 1-H); δC (100 MHz; CDCl3) 19.6,
20.1, 20.2, 24.9, 27.0 (C-3), 33.7 (C-6), 66.5 (C-1), 125.1 (C᎐),
132.8 (C-2).
᎐
2-Isopropylidene-1-methylcyclohexane 9
(3) The intramolecular electrostatic interaction between any
polar substituent and the mercurinium ion is responsible for the
selectivity of the reaction.
Methyl 2-methylcyclohexanecarboxylate (10.0 g, 0.064 mol, cis/
trans 93/7) as a solution in diethyl ether (30 cm3) was added to a
276
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278

View Article Online
solution of CH3MgI in the same solvent prepared from methyl
iodide (36.4 g, 0.256 mol) and Mg (6.14 g, 0.256 mol) at 0 ЊC.
After the addition was complete, the mixture was heated under
reflux for 3 h. The product was hydrolysed with dilute HCl
to give 2-(2-methylcyclohexyl)propan-2-ol (8.60 g, 86.0%), bp
84–102 ЊC/50–53 mmHg.
1.19–1.32 (3H, m), 1.55 (1H, m), 1.64–1.74 (3H, m), 1.99 (1H,
m), 3.26 (3H, s, OCH3), 3.60 (1H, dt, J 10.0, 4.42, 1-H), 5.31
(1H, s); δC (100 MHz; CDCl3) 19.9 (CH3), 23.5 (CH3), 24.6,
26.1, 27.1, 35.3, 48.5 (C-2), 50.7 (OCH3), 72.5 (C-1), 80.8
[C(CH3)2OCH3].
(4-Methoxy-4-methylcyclohexyl)methanol,
bp
109.0–
Dehydration of 2-(2-methylcyclohexyl)propan-2-ol (23.6 g,
0.151 mol) was catalysed by toluene-p-sulfonic acid (0.1 g) in
boiling benzene (150 cm3). The reaction mixture was treated in
the usual manner. The dehydrated product [16.9 g, 81.1%; bp
74–80 ЊC/40 mmHg; two other isomers as well as 1-isopropyl-
idene-2-methylcyclohexane 9 were detected by GLC] was par-
tially hydrogenated over Rh–C catalyst. The product com-
position was monitored by GLC. After diminution of the less
substituted olefinic isomers by the hydrogenation, the reaction
mixture was separated by column chromatograpy [SiO2–
AgNO3 (25%), hexane–ethyl acetate 9:1]. 1-Isopropylidene-2-
methylcyclohexane 9 (2.1 g, 12.3%) had bp 71 ЊC/19 mmHg
(Found: C, 86.5; H, 13.3. C10H18 requires C, 86.9; H, 13.1%);
δH (400 MHz; CDCl3) 1.00 (3H, d, J 7.32, CH3), 1.05–2.00
(7H, m), 1.63 (3H, d, J 1.47, CH3), 1.65 (3H, d, J 2.2, CH3), 2.48
(1H, ddq, J 14.1, 2.86, 1.47), 2.92 (1H, m); δC (100 MHz;
CDCl3), 18.0 (CHCH3), 20.4 (CH3), 20.5 (CH3), 20.8, 24.9,
115.3 ЊC/20 mmHg (Found: C, 68.6; H, 11.5. C9H18O2 requires
C, 68.3; H, 11.5%); δH (400 MHz; CDCl3) 1.10 (3H, s, CH3),
1.17–1.23 (4H, m), 1.44 (1H, m), 1.53 (2H, q, J 4.14, 7.81), 1.86
(2H, dd, J 2.86, 8.56), 1.99 (1H, br s, OH), 3.15 (3H, s, OCH3),
3.46 (2H, d, J 6.34, CH2O); δC (100 MHz; CDCl3) 24.4 (C-2, -6),
24.8 (CH3), 34.9 (C-3, -5), 39.6 (C-1), 48.3 (OCH3), 68.3
(CH2O), 72.8 (C-4).
2-(2-Methoxypropan-2-yl)cyclopentanol, bp 120–121 ЊC/30
mmHg (Found: C, 68.1; H, 11.3. C9H18O2 requires C, 68.3; H,
11.5%). The cis-isomer was separated by column chromato-
graphy (SiO2; hexane–ethyl acetate 9:1), cis-isomer; δH (400
MHz; CDCl3) 1.19 (3H, s, CH3), 1.39 (3H, s, CH3), 1.55–1.69
(4H, m), 1.86–1.96 (3H, m), 3.24 (3H, s, OCH3), 4.37 (1H, br s,
C-1), 4.58 (1H, br s, OH); δC (100 MHz; CDCl3) 21.4 (CH3),
22.2 (CH3), 24.0, 24.1, 34.9, 48.9 (C-2), 55.0 (OCH3), 73.9 (C-1),
77.0 [C(CH3)2OCH3]. It was impossible to isolate the trans-
isomer.
27.7, 31.1 (C-2) , 33.2 (C-6), 124.1 [᎐C(CH ) ], 134.9 (C-1).
᎐
3
2
2-(6-Methylcyclohex-1-enyl)propan-2-ol 9a4. The crude
oxymercuriation-demercuriation product of 9 was purified by
column chromatography (SiO2; hexane–ethyl acetate 9:1). 9a4,
(Found: C, 77.8; H, 11.7. C10H18O requires C, 77.9; H, 11.8%);
δH (400 MHz; CDCl3) 1.14 (3H, d, J 6.83, axial-CH3), 1.35 (3H,
s, CH3), 1.36 (3H, s, CH3), 1.52–1.78 (5H, m), 2.05 (2H, m), 2.46
Methoxymercuriation
A solution of a substrate (0.600 mmol) in methanol (1 cm3) was
added to a solution of Hg(OAc)2 (0.230 g, 0.723 mmol) in
methanol (4 cm3) and the mixture was stirred at room temper-
ature. Sodium hydroxide (3.0 M; 1.08 cm3) and NaBH4 (20.4
mg, 0.537 mmol) in aq. NaOH (3.0 M; 1 cm3) were added at
0 ЊC. The precipitated Hg was removed by filtration. The prod-
uct was isolated by diethyl ether extraction. After drying over
Na2SO4, the solvent was removed and the residue was analysed
by GLC.
(1H, m, 2H), 5.72 (1H, t, J 3.91, ᎐CH); δC (100 MHz; CDCl3)
᎐
16.9 (axial-CH3), 21.6 (CH3), 25.0 (CH3), 28.0, 30.3, 30.8, 31.4
(CHCH ), 73.5 (COH), 120.0 (᎐CH), 148.7 (᎐C).
᎐
᎐
3
References
1 H. C. Brown and P. J. Geoghegen, Jr., J. Org. Chem., 1972, 37,
1937.
Oxymercuriation
2 D. J. Pasto and J. A. Gontarz, J. Am. Chem. Soc., 1970, 92,
A solution of a substrate (0.320 mmol) in THF (5 cm3) was
added to a solution of Hg(OAc)2 (0.123 g, 0.386 mmol) in water
(5 cm3) and the mixture was stirred for 60 min at room temper-
ature. Sodium hydroxide (3.0 M; 0.576 cm3) and NaBH4 (10.9
mg, 0.287 mmol) in aq. NaOH (3.0 M; 0.576 cm3) were added at
0 ЊC. The precipitated Hg was removed by filtration. The prod-
uct was isolated by diethyl ether extraction. After drying over
Na2SO4, the solvent was removed and the residue was analysed
by GLC. In the case of 9 the precipitation of Hg0 was observed
during the reaction. At the end of the reaction the mixture was
treated with NaBH4 in the usual manner to remove toxic
Hg(OAc)2 from the solution.
7480.
3 H. C. Brown, G. J. Lynch, W. J. Hammar and L. C. Liu, J. Org.
Chem., 1979, 44, 1910.
4 Y. Senda, H. Kanto and H. Itoh, J. Chem. Soc., Perkin Trans. 2,
1997, 1143.
5 W. Kitching, Organomet. Chem. Rev., 1968, 1, 61; W. L. Waters
and E. F. Kiefer, J. Am. Chem. Soc., 1967, 89, 6261; W. L. Waters,
W. S. Linn and M. C. Caserio, J. Am. Chem. Soc., 1968, 90,
6741.
6 H. C. Brown and P. J. Geoghegen, Jr., J. T. Kurek and G. J. Lynch,
Organomet. Chem. Synth., 1970/1971, 1, 7; D. J. Pasto and
J. A. Gontarz, J. Am. Chem. Soc., 1971, 93, 6902.
7 MOPAC93, J. J. P. Stewart, QCPE Bull., 1989, 9, 10; revised by
Fujitsu Co. Ltd. for SUM Machine, JCPE Newsletter, 1994, 5,
84.
8 E. L. Eliel, Stereochemistry of Carbon Compounds, McGraw-Hill,
New York, 1962, p. 206.
9 E. L. Eliel, Stereochemistry of Carbon Compounds, McGraw-Hill,
New York, 1962, p. 211.
10 G. Klopman, J. Am. Chem. Soc., 1968, 90, 223: I. Fleming, Frontier
Orbitals and Organic Chemical Reactions, Wiley, New York,
1976.
11 W. Treibs and M. Weissenfels, Chem. Ber., 1960, 93, 1374.
12 E. L. Eliel, Stereochemistry of Carbon Compounds, McGraw-Hill,
New York, 1962, p. 198.
Products
Among the reaction products, 3-methoxycyclohexanol,24
2-methoxy-2-methylcyclohexanol,25
3-methoxy-3-methyl-
cyclohexanol,25 3-methoxy-3,5–dimethylcyclohexanol,25 3-
methoxycyclopentanol,26 (3-methoxycyclohexane)methanol,27
6-oxabicyclo[3.2.1]octane,28
1-methyl-2-oxabicyclo[2.2.2]-
octane,29 and 7-oxabicyclo[4.3.0]nonane30 have been previously
reported.
2-(2-Methoxypropan-2-yl)cyclohexanol, bp 121–126 ЊC/27
mmHg (Found: C, 69.5; H, 11.5. C10H20O2 requires C, 69.7; H,
11.7%); the isomers were separated by column chromato-
graphy (SiO2; hexane–ethyl acetate 9:1). cis-isomer; δH (400
MHz; CDCl3) 1.19 (3H, s, CH3), 1.32 (3H, s, CH3), 1.16–1.29
(4H, m), 1.56–1.88 (5H, m), 3.20 (3H, s, OCH3), 4.11 (1H, d,
J 1.95, 1-H), 4.33 (1H, br s, OH); δC (100 MHz; CDCl3) 19.6
(CH3), 20.1 (CH3), 23.0, 23.4, 26.4, 33.2, 48.9 (C-2), 50.5
(OCH3), 66.6 (C-1), 78.2 [C(CH3)2OCH3]; trans-isomer; δH (400
MHz; CDCl3) 0.91 (1H, m), 1.17 (3H, s, CH3), 1.22 (3H, s, CH3),
13 J. E. Baldwin, J. Chem. Soc., Chem. Commun., 1976, 734.
14 M. R. Johnson and B. Rickborn, J. Org. Chem., 1969, 34, 2781.
15 W. G. Dauben and G. H. Brezin, J. Am. Chem. Soc., 1953, 75, 468.
16 M. R. Johnson and B. Rickborn, J. Org. Chem., 1970, 35, 1041.
17 Y.-D. Wu, K. N. Houk and B. M. Trost, J. Am. Chem. Soc., 1987,
109, 5560.
18 H. C. Brown and E. F. Knight, J. Am. Chem. Soc., 1968, 90, 4439.
19 T. J. Howard and B. Morley, Chem. Ind. (London), 1967, 73.
20 J. H. Chan and B. Rickborn, J. Am. Chem. Soc., 1968, 90, 6406.
21 M. Mousseron, M. Mousseron-Chnet and M. Granier, C. R. Hebd.
Seances Acad. Sci., 1958, 247, 382.
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278
277

View Article Online
22 M. Dzieduszycka, M. Smulkowski and E. Borowski, Pol. J. Chem.,
1982, 56, 1569.
23 J. Mukherji and J. Vig, J. Indian Chem. Soc., 1956, 33, 853.
24 E. L. Eliel and T. J. Brett, J. Org. Chem., 1963, 28, 1923.
25 Y. Senda, H. Sakurai, S. Nakano and H. Itoh, Bull. Chem. Soc. Jpn.,
1996, 69, 3297.
27 M. K. Karin, W. H. Linnell and L. K. Sharp, J. Pharm. Pharmacol.,
1960, 12, 74.
28 S. W. Baldwin, R. J. Doll and S. A. Haut, J. Org. Chem., 1974, 39,
2470.
29 N. L. Weinberg and G. F. Wright, Can. J. Chem., 1965, 43,
24.
26 R. A. B. Bannard, A. A. Casselman, E. J. Langstaff and R. Y. Mori,
Can. J. Chem., 1967, 45, 2605.
30 Y. Yamamoto, R. J. Gargiulo and D. S. Tarbell, J. Org. Chem., 1971,
36, 846.
278
J. Chem. Soc., Perkin Trans. 1, 2001, 270–278