Remote substituent effects on the oxymercuration of 2-substituted norbornenes: An experimental and theoretical study

Article abstract of DOI:10.1021/jo010330h

The effect of a remote substituent on regioselectivity in the oxymercuration of 2-substituted norbornenes has been investigated experimentally and theoretically using density functional theory (DFT). Regioselectivities of 1:1 to 14:1 were observed with various 2-substituted norbornenes. Exo-2-substituted norbornenes always gave greater regioselectivities compared to the corresponding endo-2-substituted norbornenes. The effects of solvents on the regioselectivity have also been examined, and ethereal solvents were found to be the best choice giving the optimal yield and regioselectivity. The relative rate of oxymercuration was estimated by competition experiments. The least reactive substrate (X = OAc) gave the highest regioselectivity. According to DFT predictions, the increased difference between the reaction barriers that results in the greater regioselectivity is correlated directly with the larger polarity of the C=C double bond, which is attacked by the mercury and oxygen. A number of stable exo and endo conformers were predicted. All exo conformers show the same polarity of the double bond, while some endo conformers have a reversal of this polarity. All the conformers except those with the OAc substituent are very close in energy and thus should react. The existence of a mixture of endo conformers with the C=C double bond of opposite polarity clearly explains a decrease in regioselectivity for the endo species. The origin of the greatest regioselectivity for the OAc-2-norbornenes lies in the fact that the conformer with the largest polarity is notably lower in energy than others due to an internal C-H-O hydrogen bond.

Full text of DOI:10.1021/jo010330h

5182
J . Org. Chem. 2001, 66, 5182-5191
Rem ote Su bstitu en t Effects on th e Oxym er cu r a tion of
2-Su bstitu ted Nor bor n en es:
An Exp er im en ta l a n d Th eor etica l Stu d y
Peter Mayo, Galina Orlova, J ohn D. Goddard,*,† and William Tam*,‡
Guelph-Waterloo Centre for Graduate Work in Chemistry and Biochemistry,
Department of Chemistry and Biochemistry, University of Guelph, Guelph, Ontario, Canada N1G 2W1
wtam@uoguelph.ca
Received March 28, 2001
The effect of a remote substituent on regioselectivity in the oxymercuration of 2-substituted
norbornenes has been investigated experimentally and theoretically using density functional theory
(DFT). Regioselectivities of 1:1 to 14:1 were observed with various 2-substituted norbornenes. Exo-
2-substituted norbornenes always gave greater regioselectivities compared to the corresponding
endo-2-substituted norbornenes. The effects of solvents on the regioselectivity have also been
examined, and ethereal solvents were found to be the best choice giving the optimal yield and
regioselectivity. The relative rate of oxymercuration was estimated by competition experiments.
The least reactive substrate (X ) OAc) gave the highest regioselectivity. According to DFT
predictions, the increased difference between the reaction barriers that results in the greater
regioselectivity is correlated directly with the larger polarity of the CdC double bond, which is
attacked by the mercury and oxygen. A number of stable exo and endo conformers were predicted.
All exo conformers show the same polarity of the double bond, while some endo conformers have
a reversal of this polarity. All the conformers except those with the OAc substituent are very close
in energy and thus should react. The existence of a mixture of endo conformers with the CdC
double bond of opposite polarity clearly explains a decrease in regioselectivity for the endo species.
The origin of the greatest regioselectivity for the OAc-2-norbornenes lies in the fact that the
conformer with the largest polarity is notably lower in energy than others due to an internal C-H-O
hydrogen bond.
In tr od u ction
The study of the long-range stereoelectronic effects of
a remote substituent in controlling the regio- and ste-
reoselectivities in nucleophilic and electrophilic additions
to π-bonds has attracted considerable interest.1,2 To date,
studies of such an effect on the bicyclo[2.2.1]heptane and
bicyclo[2.2.2]octane systems (Figure 1) include the fol-
lowing: nucleophilic additions of hydride and alkyl-
lithium reagents to 7-norbornanones 1 and bicyclo[2.2.2]-
octan-2-ones 2;3 Baeyer-Villiger oxidation with peracids
F igu r e 1. Bicyclo[2.2.1]heptane and bicyclo[2.2.2]octane sys-
tems.
and ring expansion with diazomethane of 1;4,5 electro-
philic additions of dichlorocarbene to 7-methylenenor-
bornenes 3;6 hydroboration, epoxidation, and oxymercu-
ration of 3;7 electrophilic additions of PhSeX (X ) Cl, Br),
PhSCl, and Br2 to 7-oxabicyclo[2.2.1]hept-5-ene deriva-
tives 4;8,9 1,3-dipolar cycloadditions; Diels-Alder cycload-
ditions and Pauson-Khand reactions of 4;10-12 ring-
* To whom correspondence should be addressed. Tel: 519-824-4120.
† Fax: 519-766-1499. E-mail: (J .D.G.) jgoddard@uoguelph.ca.
‡ E-mail: wtam@uoguelph.ca.
(1) (a) Cieplak, A. S.; Tait, B. D.; J ohnson, C. R. J . Am. Chem. Soc.
1989, 111, 4635. (b) Adcock, W.; Cotton, J .; Trout, N. A. J . Org. Chem.
1994, 59, 1867. (c) Gung, B. W. Tetrahedron 1996, 52, 5263. (d) Fraser,
R. R.; Faibish, N. C.; Kong, F.; Bednarski, J . J . Org. Chem. 1997, 62,
6167.
(4) Mehta, G.; Mohal, N. J . Chem. Soc., Perkin Trans. 1 1998, 505.
(5) Mehta, G.; Khan, F. A. J . Chem. Soc., Perkin Trans. 1 1993, 1727.
(6) Mehta, G.; Gunasekaran, G. J . Org. Chem. 1994, 59, 1953.
(7) (a) Mehta, G.; Khan, F. A. J . Chem. Soc., Chem. Commun. 1991,
18. (b) Mehta, G.; Khan, F. A.; Gadre, S. R.; Shirsat, R. N.; Ganguly,
B.; Chandrasekhar, J . Angew. Chem., Int. Ed. Engl. 1994, 33, 1390.
(8) (a) Arjona, O.; de la Pradilla, R. F.; Plumet, J .; Viso, A.
Tetrahedron 1989, 45, 4565. (b) Arjona, O.; de la Pradilla, R. F.; Pita-
Romero, I.; Plumet, J .; Viso, A. Tetrahedron 1990, 46, 8199.
(9) Arjona, O.; de la Pradilla, R. F.; Garcia, L. Mallo, A.; Plumet, J .
J . Chem. Soc., Perkin Trans. 2 1989, 1315.
(10) Arjona, O.; Manzano, C.; Plumet, J . Heterocycles 1993, 35, 63
and refererences cited therein.
(11) Black, K. A.; Vogel, P. J . Org. Chem. 1986, 51, 5341.
(12) Arjona, O.; Csa´ky¨, A. G.; Murcia, M. C.; Plumet, J . J . Org.
Chem. 1999, 64, 7338.
(2) For recent reviews, see: (a) Cieplak, A. S. Chem. Rev. 1999, 99,
1265. (b) Ohwada, T. Chem. Rev. 1999, 99, 1337. (c) Mehta, G.;
Chandrasekhar, J . Chem. Rev. 1999, 99, 1437.
(3) (a) Mehta, G.; Khan, F. A. J . Am. Chem. Soc. 1990, 112, 6140.
(b) Mehta, G.; Khan, F. A.; Ganguly, B.; Chandrasekhar, J . J . Chem.
Soc., Chem. Commun. 1992, 1711. (c) Ganguly, B.; Chandrasekhar,
J .; Khan, F. A.; Mehta, G. J . Org. Chem. 1993, 58, 1734. (d) Mehta,
G.; Khan, F. A.; Ganguly, B.; Chandrasekhar, J . J . Chem. Soc., Perkin
Trans. 2 1994, 2275. (e) Mehta, G.; Khan, F. A.; Adcock, W. J . Chem.
Soc., Perkin Trans. 2 1995, 2189. (f) Mehta, G.; Khan, F. A.; Mohal,
N.; Narayan, I. N.; Kalyanaraman, P.; Chandrasekhar, J . J . Chem.
Soc., Perkin Trans. 1 1996, 2665. (g) Mehta, G.; Ravikrishna, C.;
Kalyanaraman, P.; Chandrasekhar, J . J . Chem. Soc., Perkin Trans. 1
1998, 1895.
10.1021/jo010330h CCC: $20.00 © 2001 American Chemical Society
Published on Web 07/04/2001

Oxymercuration of 2-Substituted Norbornenes
J . Org. Chem., Vol. 66, No. 15, 2001 5183
Sch em e 3
Sch em e 4
F igu r e 2. Exo- and endo-2-substituted norbornenes.
Sch em e 1
The endo-norbornene 7b was prepared from the Lewis
acid-catalyzed Diels-Alder reaction of cyclopentadiene
and methyl acrylate.17b,c Reduction of 6b and 7b by
LiAlH4 followed by protection with TBSCl provided the
exo-norbornene 6a and the endo-norbornene 7a . The exo-
norbornene 6c was prepared from norbornadiene 8
(Scheme 3). According to the literature, the oxymercurial
compound 9 could be made by oxymercuration using Hg-
(OAc)2 in AcOH.18 However, we found that under these
reaction conditions a significant amount of a rearrange-
ment product, the tricyclic compound 10, was formed
accompanied by the desired oxymercurial compound 9,
and separation of these compounds by chromatography
was very difficult. After several attempts to modify the
reaction conditions, we found that using THF as solvent
and slow addition of Hg(OAc)2 to an excess of norborna-
diene 8 led to the exclusive formation of the desired
oxymercurial compound 9. Demercuration and hydrolytic
cleavage of the acetate group was achieved in a single
step using sodium amalgam in NaOH to afford the
required exo-norbornene 6c in good yield. Oxidation of
6c followed by reduction with L-Selectride at -78 °C
provided the endo alcohol 7c with >99:1 endo/exo selec-
tivity (Scheme 4). Derivatization of 6c and 7c gave the
exo-norbornenes 6d -h and endo-norbornenes 7d -f.
Sch em e 2
openingcross-couplingmetathesis of4;13 and hydroboration
and electrophilic addition of PhSeX (X ) Cl, Br) and
PhSCl to 2-substituted norbornenes 5.14,15 In this paper,
we report our experimental and theoretical results on
remote substituent effects on regioselectivity in the
oxymercuration of 2-substituted norbornenes 5 (Scheme
1).16
Oxym er cu r a tion of 2-Su bstitu ted Nor bor n en es.
Unlike the oxymercuration of monocyclic olefin systems
that usually follows anti addition, oxymercuration of
bicyclic olefins often gives syn addition products. Traylor
and Baker have shown that oxymercuration of nor-
bornene (X ) H) gave entirely the syn-exo product.19 In
accord with this result, oxymercuration of all the 2-sub-
stituted norbornenes in our study was highly stereose-
lective, giving only the syn-exo products. Two different
regioisomers can be formed in the syn oxymercuration
of 2-exo-substituted norbornenes (Table 1). Hg could end
up on C6 (giving regioisomer 12) or on C5 (giving regio-
isomer 13). Addition of 1.2-3 equiv of Hg(OAc)2 to
2-substituted norbornenes 6a -h in THF afforded a
mixture of regioisomers in moderate to good yields.
Oxymercuration of 6a with an essentially neutral sub-
stituent (X ) CH2OTBS) was not selective, giving a 1:1
Resu lts a n d Discu ssion
Syn th esis of 2-Su bstitu ted Nor bor n en es. To study
substituent effects on regioselectivity in the oxymercu-
ration of 2-substituted norbornenes, exo-2-substituted
norbornenes 6a -h and endo-2-substituted norbornenes
7a -f were prepared (Figure 2). The exo-and endo-
norbornenes 6b and 7b were prepared by Diels-Alder
reactions (Scheme 2). A thermal Diels-Alder reaction of
cyclopentadiene and methyl acrylate, followed by epimer-
ization of the endo/exo cycloadduct mixture with NaOMe
and separation of the exo and endo cycloadducts by
column chromatography, produced exo-norbornene 6b.17a
(13) Arjona, O.; Csa´ky¨, A. G.; Murcia, M. C.; Plumet, J . J . Org.
Chem. 1999, 64, 9739.
(14) Brands, K. M. J .; Kende, A. S. Tetrahedron Lett. 1992, 33, 5887.
(15) Arjona, O.; de la Pradilla, R. F.; Plumet, J .; Viso, A. J . Org.
Chem. 1991, 59, 6227.
(16) For preliminary results of this work, see: Mayo, P.; Poirier,
M.; Rainey, J .; Tam, W. Tetrahedron Lett. 1999, 40, 7727.
(17) (a) Cope, A. C.; Ciganek, E.; Lebel, N. A. J . Am. Chem. Soc.
1959, 81, 2799. (b) Inukai, T.; Kojima, T. J . Org. Chem. 1966, 31, 2032.
(c) Sauer, J .; Kredel, J . Tetrahedron Lett. 1966, 713.
(18) Bartsch, R. A.; Lee, J . G. J . Org. Chem. 1991, 56, 2579.
(19) (a) Traylor, T.; Baker, A. Tetrahedron Lett. 1959, 14. (b) Traylor,
T.; Baker, A. J . Am. Chem. Soc. 1963, 85, 2746. (c) Traylor, T. J . Am.
Chem. Soc. 1964, 86, 244. (d) Whitesides, G. M. Fillpo, J . S. J . Am.
Chem. Soc. 1970, 92, 6611. (e) J ensen, F. R.; Miller, J . J .; Cristol, S.
J .; Beckley, R. S. J . Org. Chem. 1972, 37, 4341.

5184 J . Org. Chem., Vol. 66, No. 15, 2001
Mayo et al.
Ta ble 1. Oxym er cu r a tion of Exo-2-su bstitu ted
Ta ble 3. Rela tive Ra te of Oxym er cu r a tion of
Exo-2-su bstitu ted Nor bor n en es in THF
Nor bor n en es
entry
norbornene
X
relative rate
12:13
entry norbornene
X
yielda (%) 12:13b
1
2
3
4
5
6
7
6i
H
8.6
7.8
4.3
2.3
2.3
1.6
1
1
2
3
4
5
6
7
8
6a
6b
6c
6d
6e
6f
CH2OTBS
COOMe
OH
91
72
49
58
83
83
74
75
1:1
5:1
6:1
6a
6c
6e
6d
6b
6h
CH2OTBS
OH
OTBS
OBn
COOMe
OAc
1:1
6:1
12:1
9:1
5:1
14:1
OBn
OTBS
9:1
12:1
12:1
14:1
14:1
OCOPh
OCO-C6H4-p-NO2
OAc
6g
6h
Ta ble 4. Effect of Solven t in Oxym er cu r a tion of
Nor bor n en e 6h
a
b
Isolated yields of pure products after chromatography. Mea-
sured by intergration of 400 MHz 1H NMR spectra before and after
column chromatography.
Ta ble 2. Oxym er cu r a tion of En d o-2-Su bstitu ted
Nor bor n en es
entry
solvent
toluene
DMF
CH3CN
DME
Et2O
THF
CH2ClCH2Cl
hexanes
yielda (%)
12h :13h b
1
2
3
4
5
6
7
8
91
68
75
91
68
99
92
84
6:1
7.5:1
9:1
14:1
14:1
14:1
14:1
15:1
entry
norbornene
X
yielda (%)
14:15b
1
2
3
4
5
6
7a
7b
7c
7d
7e
7f
CH2OTBS
COOMe
OH
OBn
OTBS
OAc
80
62
30
36
69
49
1:1
5:1
3:1
6:1
4:1
9:1
a
b
Isolated yields of pure products after chromatography. Mea-
sured by intergration of 400 MHz 1H NMR spectra before and after
column chromatography.
a
b
Isolated yields of pure products after chromatography. Mea-
sured by intergration of 400 MHz 1H NMR spectra before and after
column chromatography.
The effects of different solvents used in the oxymer-
curation of norbornene 6h (X ) OAc) were studied, and
the results are shown in Table 4. Although the use of an
aromatic solvent such as toluene still provided a good
yield in the oxymercuration, the regioselectivity was
much lower than that obtained using THF as solvent
(Table 4, entry 1). Oxymercuration in DMF and CH3CN
also gave lower regioselectivities as well as lower yields
(Table 4, entries 2 and 3). The use of ethereal solvents
such as DME, Et2O, and THF and a chlorinated solvent
such as 1,2-dichloroethane all provided an excellent
regioselectivity of 14:1 (Table 4, entries 4-7). Using
hexanes as solvent gave the highest regioselectivity of
15:1 (Table 4, entry 8), although the yield was lowered
to 84%.
mixture of regioisomers 12a and 13a . With an ester
(COOMe) functionality, 6b gave a moderate regioselec-
tivity of 5:1. With oxy substituents, the regioselectivities
increased further. When X was changed from OH to OBn,
to OTBS, and to OCOR, the regioselectivities increased
from 6:1 to 14:1. The highest regioselectivity was ob-
tained with X) OAc. The oxymercuration reaction of
2-endo-substituted norbornenes 7a -f is shown in Table
2. Although a similar trend was observed, the regiose-
lectivities of the endo substituents ranged from 3:1 to 9:1,
and these were consistently lower than the corresponding
ratios for the exo substituents.
Several competition experiments were carried out in
order to estimate the relative rates of oxymercuration of
2-substituted norbornenes. For example, a mixture of
equimolar amounts of 6c and 6h was allowed to react
with 0.25 equiv of Hg(OAc)2 (in order to approach pseudo-
first-order conditions), and the relative rate of oxymer-
curation of 6c and 6h was determined by the ratio of the
oxymercuration products, which was measured from the
integration of the 400 MHz 1H NMR spectrum of the
crude reaction mixture. Similar competition experiments
were carried out between various norbornenes and 6c,
and the results are shown in Table 3. The rates of
oxymercuration of all the norbornenes examined (with
X * H) are slower than norbornene (X ) H). Although
there is no good correlation between the relative rate of
oxymercuration and the regioselectivity observed, the
most reactive substituted norbornene (6a ) gave the
lowest regioselectivity and the least reactive substituted
norbornene (6h ) gave the highest regioselectivity.
Except for the neutral group CH2OTBS, which showed
no selectivity, the major regioisomers formed in all cases
were the ones with the Hg attached to C6 and the OAc
attached to C5 (Tables 1 and 2). The regiochemistry of
all of the isomers was identified by spectroscopic tech-
niques and/or by chemical means. NOESY experiments
showed that in the major regioisomers 12 (e.g., 12e, when
X ) OTBS), positive NOE were observed between Ha and
Hc but no NOE were observed between Ha and Hb (Figure
3). In contrast, no NOE was observed between Ha and
Hc in the minor regioisomer 13e, but a positive NOE was
observed between Ha and Hb. We have also confirmed this
identification by chemical means. For example, for the
regioisomers 12e and 13e with X ) OTBS, the regioiso-
mers were converted to 16e and 17e by demercuration
with Na/Hg in NaOH or with LiAlH4, followed by protec-
tion with TBSCl (Scheme 5). Compound 16e is C2-
symmetric, and therefore, only four carbon signals from
the bicyclic framework were observed in the 13C NMR

Oxymercuration of 2-Substituted Norbornenes
J . Org. Chem., Vol. 66, No. 15, 2001 5185
positive charge at C6 in 19b would be further disfavored.
Thus, the formation of 18b and the regioisomer 12 would
be more probable energetically.
Th eor etica l An a lysis. To gain further understanding
of remote substituent effects on the regioselectivities in
the oxymercuration of 2-substituted norbornenes as well
as the relative rate of oxymercuration of these 2-substi-
tuted norbornenes, we performed a density functional
study of the oxymercuration of 2-substituted norbornenes.
The mechanism of the oxymercuration is of particular
interest since it previously has not been examined
theoretically. Our aim was to consider the charge distri-
bution in the reactants, as well as the energetics of initial
complexation of the Hg to the norbornenes and the
oxymercuration reaction path itself.
F igu r e 3. Determination of regiochemistry by NOE.
Sch em e 5
Since different solvents do not change qualitatively the
substituent effects and the highest regioselectivity was
obtained with nonpolar solvents, the effect of media was
not included in the computational research. We have
examined the relative energies of and the charge distri-
butions in conformers of exo- and endo-2-norbornenes,
with OH, OAc, OSiH3, OCH3, and CH2OH substituents.
The last three groups model the experimentally used
OTBS, OBn, and CH2OTBS substituents, respectively.
The potential energy profiles for the reactions of unsub-
stituted norbornene, of OH-exo-2-norbornene, of OH-
endo-2-norbornene, and of OAc-exo-2-norbornene with
mercury diformate, Hg(O2CH)2, are predicted. Hg(O2CH)2
was studied theoretically instead of the experimental
reactant, Hg(OAc)2, to lessen the computational complex-
ity.
Sch em e 6
All theoretical predictions were performed with the
GAUSSIAN 98 suite of programs.20 The local spin density
(LSD) SVWN functional with Slater exchange (S)21 and
Vosko, Wilk, and Nusair correlation (VWN)22 and the
D95V basis set23,24 were used to predict a variety of
conformers of 2-norbornenes. The charges were obtained
from a natural population analysis.25 The hybrid B3PW91
functional with Becke’s B3 exchange26 and Perdew and
Wang correlation (PW91)27 and the LanL2DZ/D95 core
potential and basis set24,28 were used for prediction of the
reaction pathways. This exchange-correlation functional
and these basis sets previously have predicted very
spectrum. In the case of compound 17e, a plane of
symmetry is present in the norbornane, and therefore,
five carbon signals from the bicyclic framework were
observed in the 13C NMR spectrum.
Unlike the oxymercuration of monocyclic olefin systems
which usually follows anti addition, oxymercuration of
bicyclic olefins often gives syn addition products.19 Cyclic
transition states have been proposed to account for these
results. Thus, for the oxymercuration of unsymmetrically
substituted bicyclic alkenes, the regioselectivity of the syn
addition of Hg-OAc across the double bonds will be
dependent on the following: (i) the charge distribution
of the olefinic carbons in the Hg-olefin complex and (ii)
the relative stability of the partial cations once the Hg-
carbon bond has formed. In our study, the major regio-
isomers (12) in all cases were formed with the OAc
attached to C5 and the HgOAc attached to C6 (Scheme
6). Initial complexation of Hg(OAc)2 with the double bond
in the norbornenes will lead to the formation of the Hg-
olefin complexes 18a and 19a . The charge distribution
in these complexes of Hg to the double bond will be one
factor that influences the regioselectivity of the oxymer-
curation. Attack of the mercury on C5 (19b) will lead to
a partial positive charge on C6, while attack on C6 (18b)
will lead to a partial positive charge on C5. The relative
stability of the two transition states will be another factor
that controls the regioselectivity. When X is an electron-
withdrawing group, the partial positive charge on C6
expected in the transition state 19b would be less
favorable, and therefore, a transition state like 18b would
be preferred in the oxymercuration and lead to the
observed major regioisomer 12. As the electron-with-
drawing power of the substituent X increases, the partial
(20) Gaussian 98: Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.;
Scuseria, G. E.; Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.;
Montgomery, J . A.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.;
Millam, J . M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.;
Tomasi, J .; Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli,
C.; Adamo, C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P.
Y.; Cui, Q.; Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu,
G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R.
L.; Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J . L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J . A. Gaussian, Inc., Pittsburgh, PA, 1998.
(21) Slater, J . C. In Quantum Theory of Molecular and Solids. Vol.
4: The Self-Consistent Field for Molecular and Solid; McGraw-Hill:
New York, 1974.
(22) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J . Phys. 1980, 58, 1200.
(23) Leininger, T.; Nicklass, A.; Stoll, H.; Dolg, M.; Schwerdtfeger,
P. J . Chem. Phys. 1996, 105, 1052.
(24) Dunning T. H., J r.; Hay, P. J . In Modern Theoretical Chemistry;
Schaefer, H. F., III; Plenum: New York, 1976; Vol. 3, p 1.
(25) (a) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J . Chem. Phys.
1985, 83, 735. (b) Reed, A. E.; Weinhold, F.; Curtiss, L. A.; Pochatko,
D., J . Chem. Phys. 1986, 84, 5687.
(26) Becke, A. D. J . Chem. Phys. 1993, 98, 5648.
(27) Burke, K.; Perdew, J . P.; Wang, Y. In Electronic Density
Functional Theory: Recent Progress and New Directions; Dobson, J .
F., Vignale, G., Das, M. P., Eds.; Plenum: New York, 1998.
(28) Hay, P. J .; Wadt, W. R. J . Chem. Phys. 1985, 82, 299.

5186 J . Org. Chem., Vol. 66, No. 15, 2001
Mayo et al.
F igu r e 4. Conformers of exo- and endo-2-norbornenes along with the relative energies ∆E in kcal/mol predicted with the SVWN
exchange-correlation functional and the D95V basis set. The positions of the excess negative charges on C5 or C6 determined by
a natural population analysis are shown. Values in parentheses for selected 2-norbornenes were predicted with B3PW91/D95. All
relative energies include zero-point vibrational energy corrections.
reasonable results for organometallic compounds.29 The
nature of each stationary point (minimum, transition
state) on the adiabatic potential energy surface was
verified by computation of the Hessian matrix. To confirm
the connections of the saddle points to minima along the
reaction pathways, Minyaev’s gradient reaction line
approach was used.30 All relative energies include cor-
rections for the zero-point vibrational energies.
Figure 4 shows the predicted conformers of the exo-
and endo-2-norbornenes along with their relative ener-
gies. For OH, OSiH3, OCH3, and CH2OH substituents,
the conformers are all very close in energy (within 0.1 to
4 kcal/mol) and apparently correspond to shallow minima
since the barrier to rotation about a single bond does not
exceed a few kcal/mol. For the OAc-2-norbornenes, the
energy differences between the conformers are somewhat
greater. The lowest energy conformers, OAc-exo-20 and
OAc-endo-20 are stabilized by weak hydrogen bonding
between the hydrogen atom attached to C2 of the nor-
bornene and the carbonyl oxygen of the OAc substituent.
Obviously, for all substituents, the predicted conformers
are of relatively low energy and thus may exist at
ambient temperature either in nonpolar or in weakly
polar solvents.
Natural population analysis shows that the charges on
the C5 and C6 atoms are slightly negative and slightly
different. Figure 4 shows the “extra” negative charges,
δ-, located on either C5 or C6 of the conformers. Recall
that δ- measures the excess negative charges on one
carbon center relative to the other. Naturally, the mer-
cury atom in Hg(O2CH)2 possesses a positive charge and
attack is favored on the carbon atom with the largest
negative charge. We propose that the ratio “product-Hg-
C6/product-Hg-C5” could be estimated on the basis of the
relative energies of the conformers and of the position of
the “extra” negative charge. The position of δ- is pre-
dicted to depend on the remote exo and endo substituents.
For the exo form, δ- is always on C6 for any conformation
studied theoretically and with any substituent. For the
endo forms, the position of δ- depends on the conforma-
tion of the substituent and, in particular, on the orienta-
tion of the lone electron pair of the oxygen atom in the
substituent. For any substituent, there exists at least one
(29) (a) Orlova G.; Scheiner, S. J . Phys. Chem. A. 1998, 102, 4813.
(b) Orlova G.; Scheiner, S. J . Phys. Chem. A. 1998, 102, 260.
(30) Minyaev, R. M. Int. J . Quantum Chem. 1994, 49, 105.

Oxymercuration of 2-Substituted Norbornenes
J . Org. Chem., Vol. 66, No. 15, 2001 5187
F igu r e 5. Potential energy profile for the reaction between Hg(O2CH)2 and ethylene predicted by B3PW91 with the LanL2DZ
core potential and basis set. The relative energies (in kcal/mol) include zero-point vibrational energy corrections. The geometries
are reported in angstroms and degrees. The transition-state vector is sketched. The total natural charge, q, on the ethylene
fragment of the pre-reaction complex is reported.
conformer (with the lone pair of the oxygen oriented
toward norbornene) that possesses a δ- on C5. Thus, for
the endo forms, there is a competition between the
conformers with δ- at C6 and the conformer or conformers
with δ- at C5. The latter conformer is always slightly
higher in energy than those with δ- at C6. Therefore, the
endo form slightly favors the C6 product. Such a competi-
tion is absent for the exo form, where all conformers have
δ- on C6. Thus, the exo form shows greater regioselec-
tivity than does the endo form.
For the exo form, the ranking of substituents (from
greatest regioselectivity to least) may be established on
the basis of the relative energies and the values of the
“extra” negative charge at C6: OAc > OSiH3 > OCH3 >
OH ∼ CH2OH.
smaller δ- of -0.008, is only 0.3 kcal/mol higher in
energy. Thus, OSiH3-exo-21 also reacts leading to less
regioselectivity.
Obviously, the concept of interacting charges of the
reagents is too rough an approximation to determine the
whole reaction process, although it appears valid for the
initial step. To examine whether the statement “the
mercury prefers the carbon with the largest negative
charge” remains true along the entire reaction pathway,
we examined the potential energy profiles for the reaction
of mercury diformate with a series of reagents. Ethylene
(a simpler model reagent), unsubstituted norbornene (the
parent reaction), both conformers of OH-exo- and of OH-
endo-2-norbornenes, and the most stable conformer of the
OAc-exo-2-norbornene were considered. It should be
noted that the experimentally observed product ratios of
6:1 for OH-exo- and of 3:1 for OH-endo-2-norbornenes are
relatively small. Such a small product ratio corresponds
to a reaction involving an energy difference between the
transition states of only ca. 1.0 kcal/mol. Thus, rather
delicate distinctions between the reaction profiles for the
various conformers should be expected.
This order is in full accord with the experimental
results. From a theoretical analysis, the OAc substituent
shows the greatest regioselectivity since the OAc-exo-20
and OAc-exo-21 conformers possess the largest δ- of
-0.015 and -0.013, respectively. Moreover, the OAc-exo-
20 is 5.8 kcal/mol lower in energy than the OAc-exo-21
and thus must be the predominant reactant. The OSiH3
substituent is less regioselective. Although OSiH3-exo-
20 possesses a δ- of -0.015 similar to the OAc-exo-20
case, the second conformer, OSiH3-exo-21, with much
The potential energy profile of the reaction between
Hg(O2CH)2 and ethylene is shown in Figure 5. This
reaction models electrophilic attack on a double bond

5188 J . Org. Chem., Vol. 66, No. 15, 2001
Mayo et al.
F igu r e 6. Potential energy profiles for the reactions between Hg(O2CH)2 and 2-substituted norbornenes (X ) Y ) H; X ) OH,
Y ) H; X ) H, Y ) OH; X ) OAc, Y ) H) predicted by B3PW91 with the LanL2DZ core potential and basis set. The computed
relative energies are listed in Table 5. Transition-state vectors are sketched.
without significant steric effects. Hg(O2CH)2 possesses a
tetrahedral D2d structure, with equal Hg-O bond lengths.
The O2CH fragments may rotate easily about the sym-
metry axis via a D2h transition structure, with a small
barrier of 1.3 kcal/mol. In the first step, a pre-reaction
complex (PRC) forms, with a binding energy of 5.1 kcal/
mol. The ethylene fragment possesses a slightly positive
charge. A weak withdrawal of electron density from
ethylene to Hg(O2CH)2 is indicative of the electrophilic
character of the mercury attack on the double bond.
These results are in accord with earlier theoretical
predictions31 on the reaction intermediate formed from
Hg(OAc)2 attack on ethylene. The PRC undergoes a
rearrangement to the product via a six-membered transi-
tion structure, with a barrier of 21.0 kcal/mol. The
product is essentially isoenergetic with the reactants. The
transition structure possesses Cs symmetry with the
bonds to be formed in the plane of the six-membered ring.
The carbon which interacts with mercury is almost
tetrahedral, and the C-Hg bond is only slightly longer
than that in product. In contrast, the O-C interaction
is at a much earlier stage, and thus, the formation of the
O-C bond is the limiting step on the reaction pathway.
The potential energy profile for the reaction between
Hg(O2CH)2 and the selected norbornenes is shown in
Figure 6. In the first step, a PRC forms much like that
in the model reaction with ethylene. Then the reaction
proceeds somewhat differently. A six-membered transi-
tion structure was not located despite repeated attempts.
Probably due to steric effects in the PRC, the formation
of a six-membered transition structure appears to be
energetically unfavorable. Instead, one of the Hg-O
bonds breaks followed by a rotation of the CHO group
about the remaining Hg-O bond and the reaction
pathway reaches a broad plateau on the potential energy
surface. This flat region is ca. 12 kcal/mol higher in
energy than the PRC and corresponds to a very loose
complex, with a five-coordinate mercury. Due to the
extremely flat potential energy surface, we were not able
to locate rigorously any critical point in this region. Since
the formation of the 5-coordinated-Hg complex is a
relatively low energy process, it does not appear to be
the limiting step on the reaction pathway. The 5-coordi-
nated-Hg complex transforms directly to the reaction
product, product-Hg-C5, via a four-membered transition
structure, TS-C5, or to the product-Hg-C6 via a four-
membered transition structure, TS-C6. These transition
structures, TS-C5 and TS-C6, are ca. 19-21 kcal/mol
higher in energy than the corresponding PRC. The
product-Hg-C5 and product-Hg-C6 are very close to or
very slightly higher in energy than the PRC. Similar
reaction profiles, with small deviations in the relative
energies were predicted for all the norbornenes consid-
ered. Note that despite the differences in the mecha-
nisms, the reactions of mercury diformate with ethylene,
via a six-membered transition structure, and with nor-
bornenes, via a four-membered transition structure, have
similar energetics. The relative energies for the reaction
profiles in Figure 6 are listed in Table 5. Selected
geometrical parameters for the pre-reaction complexes,
PRC, are shown in Figure 7.
For the parent reaction with norbornene, a PRC of Cs
symmetry might be expected. In fact, the Cs structure is
a transition state. It leads to two degenerate minima for
the PRC-norbornene, upon “swinging” of the Hg(O2CH)2
(31) Yamabe, S.; Minato, T. Bull. Chem. Soc. J pn. 1993, 66, 3339.

Oxymercuration of 2-Substituted Norbornenes
J . Org. Chem., Vol. 66, No. 15, 2001 5189
Ta ble 5. Bin d in g En er gies of th e P RC (∆EBin d ), th e
Rea ction Ba r r ier s for TS-C6 (∆EqC6) a n d TS-C5 (∆EqC5),
the carbon with the “extra” negative charge. A second
possible PRC, with the Hg shifted to the carbon with the
smaller negative charge does not appear to exist on the
potential energy surface. For all the PRC, the total
natural charges on the Hg(O2CH)2 fragments reported
in Figure 7 are slightly negative. Thus, similar to the
reaction with ethylene, there is a typical withdrawal of
the electron density from the olefin for these electrophilic
reactions. Naturally, all predicted PRC are very loose
structures. Distortions of the (O2CH) fragments involving
twisting have the lowest vibrational frequencies (ca. 4-50
cm-1) while those involving the stretching of the Hg-C5
and Hg-C6 bonds are at ca. 100 cm-1. As a result, the
rearrangement of the 6-coordinated PRC to the 5-coor-
dinated Hg complex is a low energy process.
th e Differ en ces betw een th e Ba r r ier s, ∆∆Eq (∆∆Eq
)
∆Eq - ∆EqC6), a n d th e En er gies of th e P r od u cts
C5
Refer r ed to th e P RC, ∆EC6 a n d ∆EC5, for th e P oten tia l
En er gy P r ofiles in F igu r e 6
system
X ) Y ) H
∆Ebind ∆Eq ∆Eq ∆∆Eq ∆EC6 ∆EC5
C6 C5
-8.6 20.0 20.0
-7.2 19.6 20.9
-8.0 19.4 20.2
0.0
3.2 3.2
X ) OH, Y ) H, exo-20
X ) OH, Y ) H, exo-21
1.3 -0.3 0.9
0.8
1.6
0.0 0.6
3.1 4.8
X ) H, Y ) OH, endo-20 -6.3 19.9 21.5
X ) H, Y ) OH, endo-21 -9.0 20.0 19.2 -0.8 -0.2 1.1
X ) OAc, Y ) H, exo-20 -7.3 19.4 20.8
1.4 -0.4 1.1
a
The relative energies in kcal/mol include zero-point vibrational
energy corrections. All results were predicted with B3PW91/
LanL2DZ.
To compare the four-membered transition structure
with a six-membered one, consider the geometries for the
TS-C5, TS-C6, and corresponding products for the OH-
endo-20 conformer, which possesses the largest excess
negative charge on C6 among all the conformers whose
reaction profiles were studied. Selected geometrical
parameters are shown in Figure 8. Similar to a six-
membered transition structure, the formation of the
fragment about the imaginary axis (indicated by the
arrow in Figure 7), which lies in the Cs plane of the
norbornene and crosses the Hg atom. The Hg-C5 and
Hg-C6 bond lengths in PRC-norbornene are virtually
equal. In contrast, for the PRC formed with all the exo
forms and with OH-endo-20, the Hg-C6 distances are
notably shorter compared to Hg-C5 while for the OH-
endo-21, the Hg-C5 bond is shorter, i.e., Hg is shifted to
F igu r e 7. Selected geometrical features for the pre-reaction complexes (PRCs) of the potential energy profiles from Figure 6 as
predicted using B3PW91/LanL2DZ. The geometries are reported in angstroms and degrees. The total natural charges, q, on the
Hg(O2CH)2 fragments are reported.

5190 J . Org. Chem., Vol. 66, No. 15, 2001
Mayo et al.
kcal/mol, and the resulting effect on regioselectivity
becomes small. Given that the OAc-exo-20-conformer is
notably more stable (5.8 kcal/mol) than the second
possible conformer, OAc-exo-21, the effective ∆∆Eq should
be close to 1.4 kcal/mol.
Con clu sion s
We have demonstrated a remarkable remote substitu-
ent effect on the regioselectivity of the oxymercuration
of a 2-substituted norbornene system. Regioselectivities
of 1:1-14:1 were observed with various 2-substituted
norbornenes. Exo-2-substituted norbornenes always gave
better regioselectivities than the corresponding endo-2-
norbornenes. The use of ethereal solvents was found to
give the highest regioselectivities in the oxymercuration
reactions. Study of the relative rate of oxymercuration
with various substituted norbornenes showed that the
least reactive substrate (X ) OAc) gave the highest
regioselectivity.
Our theoretical investigations allow us to rationalize
the effects of the remote substituents on regioselectivity
in terms of the position of the “extra” negative charge at
the carbon atoms, C5 or C6, which will be attacked by
Hg. The positive Hg atom prefers interaction with that
carbon that possesses the “extra” negative charge. The
charge redistribution between C5 and C6 upon substitu-
tion at the 2-position of norbornene, depends on the
position of the substituent (exo or endo) and on its
conformation. For the exo forms, the “extra” negative
charge is at C6 for all conformers while for the endo forms
there exists at least one conformer with “extra” negative
charge at C5. The competition between endo conformers
with different positions of the “extra” negative charge
decreases the regioselectivity compared to the exo form.
The oxymercuration of norbornenes proceeds via for-
mation of a pre-reaction complex with a six-coordinate
Hg in the initial step. This pre-reaction complex rear-
ranges easily to a very loose complex with a five-
coordinate Hg that transforms to the product via a four-
membered transition structure. The preference for Hg to
attack the more negative carbon remains along the whole
reaction profile and results in a lowering of the activation
barrier to the product of 0.8-1.6 kcal/mol compared to
the “unfavorable” attack at the carbon with the smaller
negative charge.
F igu r e 8. Selected geometrical features for the transition
structures TS-C6 and TS-C5 and for the product-Hg-C6 and
product-Hg-C5 of the OH-endo-20 conformer as predicted using
B3PW91/LanL2DZ. The geometries are reported in angstroms
and degrees. Natural charges (in italics, underlined) on C5 and
C6 are reported.
C-Hg bond is almost complete in the four-membered
transition structure while the O-C interaction is at an
initial stage. Thus, the formation of the O-C bond is a
rate-limiting step in the oxymercuration of norbornenes
similar to the reaction with ethylene. Natural charges
on C5 and C6 in TS-C5 and TS-C6 in Figure 8 indicate
significant withdrawal of electron density from the
carbon interacting with oxygen to the carbon interacting
with mercury. The TS-C5 corresponds to a less favorable
attack of mercury but, more importantly, also of oxygen
since the oxygen must interact with the more negative
carbon. As a result, the TS-C5 is a later transition state
than the TS-C6.
For the exo forms, the OAc-exo-20 conformer with the
largest “extra” negative charge is stabilized by an inter-
nal C-H‚‚‚O hydrogen bond and thus is notably lower
in energy than OAc-exo-21, with a smaller “extra” nega-
tive charge at C6. Therefore, OAc-exo-20 exhibits the
greatest difference between the activation barriers and
provides the highest regioselectivity.
Further investigations of remote substituents effects
in the 2-substituted norbornene systems for other reac-
tions are ongoing in our laboratory.
The reaction barriers corresponding to TS-C6 and TS-
C5 reveal the same preferences for the Hg-C and O-C
interactions as do the PRC and transition structures.
Table 5 shows that for OH-exo-20, OH-exo-21, OAc-exo-
20, and OH-endo-20, with “extra” negative charge at C6,
the ∆Eq reaction barriers are lower than ∆Eq by 1.3,
C6
C5
0.8, 1.4, and 1.6 kcal/mol, respectively. In contrast, for
OH-endo-21, which possesses the “extra” negative charge
at C5, the ∆Eq is 0.8 kcal/mol lower in energy than
C5
Ack n ow led gm en t. We thank the Natural Sciences
and Engineering Research Council (NSERC) of Canada
and the University of Guelph for generous financial
support of our program. W.T. thanks Boehringer Ingel-
heim (Canada) Ltd. for a Young Investigator Award, and
P.M. thanks NSERC for postgraduate scholarships
(PGS A and PGS B). Mr. Marc Poirier and Mr. J an
Rainey are thanked for preliminary experiments, and
Ms. S. N. Sullivan is thanked for preliminary analysis
∆EqC6. Thus, the presence of OH-endo-2 cancels the large
regioselective effect of the OH-endo-20, which would be
expected given the largest difference between the activa-
tion barriers. The energy difference between OH-exo-20
and OH-exo-21 conformers of 0.1 (1.1) kcal/mol (Figure
4) is insignificant, and both conformers are involved in
the reaction. Thus, the effective ∆∆Eq (∆Eq - ∆Eq
)
C5
C6
lies approximately in the middle of the range of 1.3-0.8

Oxymercuration of 2-Substituted Norbornenes
J . Org. Chem., Vol. 66, No. 15, 2001 5191
oxymercuration reactions; Cartesian coordinates and total
energies predicted theoretically for the structures in Figures
4-8; and 1H and 13C NMR spectra of 6a ,e-g, 7a ,e, 12f/13f,
and 12g/13g. This material is available free of charge via the
Internet at http://pubs.acs.org.
of the performance of theoretical methods for the
description of oxymercuration reactions. Ms. Valerie
Robertson is thanked for NMR experiments and discus-
sion of NMR data.
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures for preparation of substituted norbornenes and their
J O010330H