Substituent dependence of the diastereofacial selectivity in iodination and bromination of glycals and related cyclic enol ethers

Article abstract of DOI:10.1021/jo000799x

The stereochemical course of the electrophilic iodination and bromination of tri-O-benzyl-D-glucal under various conditions has been compared to that of substituted dihydropyrans 2-5. IN₃ addition in acetonitrile affords trans-α-iodoazides (80-87%), besides small amounts of trans-β-adducts, in the presence or the absence of benzyloxy substituents at C-3 or C-4, and in agreement with bridged iodonium ion intermediates. In contrast, the diastereofacial selectivity of bromine addition in dichloroethane going through open bromo oxocarbenium ions depends strongly on the substituents. Whereas the trans-α-dibromides are the main (85-95%) adducts in the absence of C-4 and C-5 substituents, in their presence a moderate to exclusive selectivity for cis-α-addition (60-99%) is observed. The predominance of trans-α-addition is again observed whatever the substituents when the bromination is carried out in the same solvent but with a tribromide ion salt, supporting a concerted addition of the two bromine atoms under these conditions. Finally, bromine addition in methanol exhibits a completely different behavior with the nonselective formation of trans-α- and trans-β-methoxybromides and a small dependence on the substituents. In agreement with the absence of azide trapping of any cationic intermediate, it is concluded that these brominations which do not go through an ionic intermediate are concerted additions of bromine and methanol with very loose rate- and product-determining transition states. Finally, the substituent conformation at C-4 influences drastically the stereoselectivity in all these brominations. Evidence for α-anomeric control of the nucleophile approach at C-1 is given by the highly predominant formation of α-adducts, except in the methanolic bromination. The factors determining the versatile selectivity of the electrophile approach are discussed in terms of PPFMO theory and of the special mechanisms of glycal reactions.

Full text of DOI:10.1021/jo000799x

8470
J . Org. Chem. 2000, 65, 8470-8477
Su bstitu en t Dep en d en ce of th e Dia ster eofa cia l Selectivity in
Iod in a tion a n d Br om in a tion of Glyca ls a n d Rela ted Cyclic En ol
Eth er s
Antonella Boschi,† Cinzia Chiappe,*,† Antonietta De Rubertis,† and Marie Franc¸oise Ruasse*,‡
Dipartimento di Chimica Bioorganica e Biofarmacia, Universita` di Pisa, via Bonanno 33,
56126 Pisa, Italy, and Institut de Topologie et de Dynamique des Syste`mes,Universite´ Paris 7,
75005 Paris, France
ruasse@paris7.jussieu.fr
Received May 23, 2000
The stereochemical course of the electrophilic iodination and bromination of tri-O-benzyl-D-glucal
under various conditions has been compared to that of substituted dihydropyrans 2-5. IN3 addition
in acetonitrile affords trans-R-iodoazides (80-87%), besides small amounts of trans-â-adducts, in
the presence or the absence of benzyloxy substituents at C-3 or C-4, and in agreement with bridged
iodonium ion intermediates. In contrast, the diastereofacial selectivity of bromine addition in
dichloroethane going through open bromo oxocarbenium ions depends strongly on the substituents.
Whereas the trans-R-dibromides are the main (85-95%) adducts in the absence of C-4 and C-5
substituents, in their presence a moderate to exclusive selectivity for cis-R-addition (60-99%) is
observed. The predominance of trans-R-addition is again observed whatever the substituents when
the bromination is carried out in the same solvent but with a tribromide ion salt, supporting a
concerted addition of the two bromine atoms under these conditions. Finally, bromine addition in
methanol exhibits a completely different behavior with the nonselective formation of trans-R- and
trans-â-methoxybromides and a small dependence on the substituents. In agreement with the
absence of azide trapping of any cationic intermediate, it is concluded that these brominations
which do not go through an ionic intermediate are concerted additions of bromine and methanol
with very loose rate- and product-determining transition states. Finally, the substituent conforma-
tion at C-4 influences drastically the stereoselectivity in all these brominations. Evidence for
R-anomeric control of the nucleophile approach at C-1 is given by the highly predominant formation
of R-adducts, except in the methanolic bromination. The factors determining the versatile selectivity
of the electrophile approach are discussed in terms of PPFMO theory and of the special mechanisms
of glycal reactions.
In tr od u ction
facial selectivity.3 It has been reported4 that in the case
of glycals the below-plane approach of the electrophile
is generally preferred, with the exception of above-plane
attack of iodine2 and selenium.5 However, a thorough
search in the literature for glycal additions revealed that
for each electrophile the face selectivity depends highly
on glycal substitution and reaction conditions, and vari-
ous interpretations have been proposed to rationalize the
observed behavior.
For example, the electrophilic addition of halogens to
cyclic enol ethers has been investigated early by Lemieux
and Fraser-Reid6 who proposed a general mechanism
involving the initial formation of carbenium ions which,
upon nucleophilic attack by halide ion, give mainly the
products of thermodynamic control. Later on, it has been
shown that the product formation is under kinetic control
and that the stereoselectivity depends on the solvent
polarity,7-9 the structure of the enol ether, and the nature
For many natural products, a selective link-up between
an aglycon and a carbohydrate is the ultimate require-
ment for a total synthesis. Among the different methods
reported1 to obtain stereocontrolled constructions of
glycoside bonds one of the more fruitful approach consists
of the use of glycals.2 These compounds can indeed be
easily activated by addition of an electrophile which,
through the formation of an onium species (eq 1), makes
the addition of the suitable nucleophile at the anomeric
position possible.
In this synthesis strategy, in agreement with all the
other modern synthetic procedures, the central problem
is the possibility to predict and to control the diastereo-
(3) Diastereoselection. Chem. Rev. Special Thematic Issue 1999,
99(5).
(4) Grewal, G.; Kaila, N.; Franck, R. W. J . Org. Chem. 1992, 57,
2084.
(5) Perez, M.; Beau, J . M. Tetrahedron Lett. 1989, 30, 75.
(6) (a) Lemieux, R. U.; Fraser-Reid, B. Can. J . Chem. 1965, 43, 1460.
(b) Uriel, C.; Santoyo-Gonzalez, F. Synthesis 1999, 2049.
(7) Igarashi, K.; Honma, T.; Imagawa, T. J . Org. Chem. 1970, 35,
610.
† Dipartimento di Chimica Bioorganica e Biofarmacia, Universita`
di Pisa, via Bonanno 33, 56126 Pisa, Italy. E-mail: cinziac@farm.unipi.it.
‡ Institut de Topologie et de Dynamique des Syste`mes,Universite´
Paris 7, 75005 Paris, France.
(1) Toshima, K.; Tatsuta, K. Chem. Rev. 1993, 93, 1503.
(2) Danishefsky, S. J .; Bilodeau, M. T. Angew. Chem., Int. Ed. Engl.
1996, 35, 1380.
10.1021/jo000799x CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/14/2000

Diastereofacial Selectivity in Glycals
J . Org. Chem., Vol. 65, No. 25, 2000 8471
of the halogen.9 In particular, in the bromination reac-
tions a dependence of the product distribution on the
electron-donating or withdrawing effect of the substituent
at C-6 has been found and these results have been
interpreted9 in terms of the effect that the 6-substituent
may exert on the nonbonding electron pair of the ring
oxygen atom, affecting the charge distribution in the
intermediate whose predominant structure can be either
I, II, or III.
Sch em e 1
Sch em e 2
More recently, a stereoelectronic R-anomeric effect10
able to stabilize the transition states related either to
the electrophilic step or to the nucleophilic step has been
invoked11 to rationalize the product distribution obtained
in the additions of halogens to tri-O-benzyl-D-glucal.
Finally, polarized π-frontier molecular orbital theory
(PPFMO) in which the reagents and not the transition
states are taken into account, has been applied12 recently
for interpreting the stereochemical behavior of several
dihydrofurans and dihydropyrans observed both in pro-
tonation and sulfenylation, two reactions considered as
models of additions via open and bridged intermediates,
respectively. Also of interest in electrophilic additions to
glycals are the recent data13 and calculations14 on the
stereochemistry of the nucleophilic trapping of six-
membered-ring oxocarbenium ions showing the pseudo-
axial preference of these ions bearing benzyloxy substit-
uents.
and sulfenylation. Second, they are suitable substrates
to give indications about the contribution to the stereo-
selectivity of some specific factors, such as the stereo-
electronic R-anomeric effect, the polarization of the
π-orbitals, the effect due to the axial substituent at C-4
and their dependence on the bridging in the ionic
intermediates. While the nucleophile attack is mainly
determined by stereoelectronic R-anomeric effect what-
ever the substituents, it is shown that the kinetically
controlled stereochemistry of the electrophile approach
depends significantly on the substituents and on the
reaction mechanism.
In this paper, the stereochemical course of the elec-
trophilic additions of halogens, and in particular of
bromine, to tri-O-benzyl-D-glucal (1) is compared with
that of the dihydropyrans 2-5, to obtain further infor-
mation about the factors which contribute to the stereo-
facial selectivity in the electrophilic addition to cyclic enol
ethers.
Resu lts
1. Syn th esis of Dih yd r op yr a n s 4 a n d 5. 3,4-Dihy-
dro-2-benzyloxymethyl-2H-pyran (4) was obtained from
commercial 3,4-dihydro-2-hydroxymethyl-2H-pyran by
benzylation. 3,4-Dihydro-2-phenyl-2H-pyran (5) was pre-
pared according to the sequence of reactions reported in
Scheme 1.
Treatment of 4-benzoylbutyric acid (6) with NaBH4
afforded the corresponding alcohol 7, which has been
warmed, in the absence of solvent for 2 h, at 80-106 °C
under vacuum using a Dean-Stark trap to remove the
formed H2O. Reduction of 8 with DIBAL-H at -78 °C
afforded the diasteroisomeric mixture of lactols 9 that
may be converted by dehydration with P2O5 in toluene
to dihydropyran 5.
2. P r od u ct An a lysis. The halogen additions (IN3, Br2,
and Br3-) to compounds 2-5 were performed in the dark
at 0 and/or 25 °C. The products have been extracted from
the mixtures immediately after the end of the reactions.
The composition of the mixtures was determined by NMR
spectroscopy. The structures of the resulting products,
reported in Scheme 2, were established on the basis of
their 1H and 13C NMR spectra, by comparison with
previously reported data (gluco, manno, galacto and talo
derivatives) or from a combination of the following
factors: (i) the J 1,2 coupling constants, (ii) the chemical
shifts of H-1, (iii) the chemical shifts of C-1, and (iv) the
J 2,3 coupling constants.
These substrates were chosen for the following reasons.
First, compounds 1-4 were studied theoretically applying
the PPFMO method and experimentally in protonation
(8) Boullanger, P.; Descotes, G. Carbohydr. Res. 1976, 51, 55.
(9) Horton, D.; Priebe, W.; Varela, O. J . Org. Chem. 1986, 51, 3479.
(10) The stereoelectronic R-anomeric efect has been shown to occur
as a general effect in X-A-Y segments. It is described in terms of
endo- and exo-anomeric effects. The endo-effect is related to the
preference of an electronegative group at the anomeric carbon for the
axial orientation. The preference is due partly to inductive effects
(dipole-dipole interactions between the endo-cyclic oxygen lone pairs
and those of the electronegative substituent, X), partly to the stabilizing
overlap of the nx f σ/c-x orbitals. See: Kirby, A. J . Stereoelectronic
Effects; Oxford Science Publications: New York, 1996.
(11) Bellucci, G.; Chiappe, C.; D’Andrea, F.; Lo Moro, G. Tetrahedron
1997, 53, 3417.
(12) Franck, R. W.; Kaila, N.; Blumenstein, M.; Geer, A.; Ling
Huang, X.; Dannenberg, J . J . J . Org. Chem. 1993, 58, 5335.
(13) Romero, J . A. C.; Tabacco, S. A.; Woerpel, K. A. J . Am. Chem.
Soc. 2000, 122, 168.
(14) Mijkovic, M.; Yeagley, D.; Deslongchamps, P.; Dory, J . L. J . Org.
Chem. 1997, 62, 7597.

8472 J . Org. Chem., Vol. 65, No. 25, 2000
Boschi et al.
Ta ble 1. P r od u ct Distr ibu tion in Electr op h ilic Ad d ition s
Ta ble 2. P r od u ct Ra tios Obser ved for Br om in a tion of 2
to Com p ou n d s 1-5
in MeOH in th e P r esen ce of N3-, a t 25 °C
products
product ratio
a
kN3/kMeOH
(M-1
)
(kN3 + kB)/kN3
a
[NaN3]
Br, N3
Br, OCH3
a
b
c
0.15
0.2
0.27
5.6
10.6
13.8
94.4
89.4
86.2
reagents
N, E
N3, I
%
N, E
%
N, E
%
1
ICl/NaN3/-
CH3CNa
87 N3, I
13
0.595
1.006
4
5
80
84
20
16
a
Reference 19.
1
2
3
4
5
Br2/DCEb
Br, Br
40
20
<1
95
86
Br, Br
60
Sch em e 3
80
>99
5
14
1
2
Bu4NBr3/DCEb Br, Br
87 Br, Br
traces Br, Br
13
7
57
5
93
43
95
90
3
4c
5c
10
1
2
3
4
5
Br2/MeOHb
MeO, Br 55 MeO, Br 45
70
30
56
62
30
70
44
38
a
b
0 °C. 25 °C. c 20-25% unidentified products.
The product distributions observed for all compounds
2-5 are reported in Table 1, which also shows the
distribution data related to glycal 1. The NMR data of
all the addition products are given in Table S1.
tially to the R-trans adduct, a (N, E ) Br, Br). Neverthe-
less, in the case of enol ethers 4 and 5 the reaction was
characterized by the formation of a mixture (20-25%) of
unidentified products.
3. Ster eoch em istr y. In the IN3 addition (generated
in situ from NaN3 and ICl) to compounds 4 and 5, the
main products are the R-trans-azide-incorporated ad-
ducts,15 a (N, E ) N3, I). The ratio between the R-trans
adducts and the â-trans, b (N, E ) N3, I), is practically
the same as that observed in the reaction of 1, showing
that the presence of the substituents at C-3 and C-4 does
not affect significantly the product distribution for this
addition.
Bromine addition in 1,2-dichloroethane (DCE) gives
always two products: the R-trans, a (N, E ) Br, Br), and
the R-cis adduct, c (N, E ) Br, Br). However, the ratio is
markedly affected by the nature and position of the
substituents at C-3 and C-4. The cyclic enol ethers 4 and
5 lead to the preferential formation of the adduct a , while
all the other substrates give selectively or exclusively the
corresponding dibromides c. Moreover, this selectivity
depends on the nature of the substituents (compare 1 and
2) and on their orientation on the pyran ring (compare 1
and 3). When bromine addition is carried out in methanol
the trans-methoxy adducts, the R-trans, a (N, E ) OCH3,
Br) and the â-trans, b (N, E ) OCH3, Br), were formed
as the sole products. These compounds are obtained in a
ratio close to unity from 1, 4 and 5, indicating that in
this solvent the product stereochemistry does not depend
significantly on the substituents at C-4 and C-5. In
contrast, the stereochemistry at C-4 seems to affect
markedly the reaction.
4. Br om in e Ad d ition in Meth a n ol in th e P r esen ce
of Na N3. The addition of Br2 in methanol in the presence
of a large excess of NaN3, which generated in situ
BrN3,16,17 generally afforded the same product distribu-
tion as that observed in the absence of the added salt.
The bromomethoxy derivatives a and b were indeed the
sole products observed in the crude reaction mixtures
arising from the addition to 1, 3, 4 and 5. In other words,
at variance with previous results on bromination of
acyclic alkenes,17 there is no trapping of any cationic
intermediate by the strongly nucleophilic azide ion.18
Only in the case of 2, a detectable amount of bro-
moazide adducts, which increases with [N3-] is observed
(Table 2).
The product data of Table 2 fit fairly well the usual
equation17,18 (eq 2), based on Scheme 3 in which the
reaction products are assumed to arise from the nucleo-
phile trapping of an intermediate.
kB
kMeOH + kBr[Br-]
1
faz
1
) 1 +
+
(2)
-
(
)
k
k
N3
N3
[N3 ]
In this equation, faz is the fraction of bromoazide in
the total product mixture, kB, the rate constant of the
azide-catalyzed trapping of the intermediate by metha-
Finally, the electrophilic addition of bromine using a
tribromide salt proceeds, with the exception of the
galactal derivative 3, stereoselectively leading preferen-
nol, kN , the rate constant for the diffusion-controlled
3
azide trapping17,18 of this intermediate (in methanol, kN
3
(16) Chiappe, C.; Ruasse, M. F. unpublished results.
(17) Nagorski, R. W.; Brown, R. S. J . Am. Chem. Soc. 1992, 114,
7773.
(18) Richard, J . P.; Rothenberg, M. E.; J encks, W. P. J . Am. Chem.
Soc. 1984, 106, 1361. Fishbein, J . C.; McClelland, R. A. J . Am. Chem.
Soc. 1987, 109, 2824.
(15) R-trans (â-cis) adducts: R(or â) stands for the nucleophile
entering in the R- (or â-) anomeric position, i.e., for an axial (or
equatorial) or below-face (or top-face) nucleophilic attack. “trans” (or
“cis”) is related to the relative orientation of the electrophile, E, and
the nucleophile, N.

Diastereofacial Selectivity in Glycals
J . Org. Chem., Vol. 65, No. 25, 2000 8473
) 1010 M-1 s-1) and kMeOH and kBr , the rate constants
-
Sch em e 4
for the reaction of the cationic intermediate with metha-
nol and bromide ions. (In the present case, the term kBr
-
[Br-] can be neglected since no dibromide is obtained.)
The fit of the product data to eq 2 provides19 the kinetic
data shown in Table 2 with kMeOH ) 7 × 109 M-1 s-1. In
other terms, the lifetime in methanol of the intermediate
issued from the bromination of 2 is 6 × 10-9 s. Further-
more, this reaction is not azide-catalyzed since kB is
negligible (Table 2).
Discu ssion
The present interest3 in understanding and predicting
diastereofacial selectivity is not surprising considering
the ubiquity of glycal reactions in synthetic organic
chemistry. Despite much work, however, most of the
assumptions on the interactions contributing to the small
activation energy differences leading to the observed
selectivities are still controversial. In particular, as
regards electrophilic additions to cyclic enol ethers (gly-
cals) under kinetic control, apart from steric effects, a
number of factors, which can be classified as stereoelec-
tronic, conformational, inductive and electrostatic, have
been alternatively invoked7-11 to account for the observed
stereochemistry, while the PPFMO method has been
proposed12 to predict the selectivities. Upon accounting
for the expected differences in the mechanisms of the
investigated reactions, protonation implying electrophilic
attack at C-2 and sulfenylation both at C-1 and C-2, this
method has been suggested to be able to predict generally
the stereoselectivity of the electrophilic additions to
glycals since “it does not focus upon a single electronic
interaction of the dissymmetric σ framework as being
responsible for the selectivity. Rather, it accounts for all
electronic effects as part of a molecular orbital treat-
ment”.12 It is noteworthy that although this method
predicts an equatorial selectivity both for gluco and
galacto derivatives, as well as for 4, in reactions involving
electrophilic attack at both C-1 and C-2 to form a bridged
intermediate (sulfenylation), an axial selectivity in the
case of 4 and again equatorial for glucals and galactals
has been forecast when, as in deuteration, the reaction
implies the initial electrophilic attack at C-2.
It must be considered that, with the term “electrophilic
addition”, is regarded a large class of organic reactions
whose essential feature is that the alkene reacts with a
reagent in a such way that the transition state, or the
intermediate, has a cationic character. Electrophilic
addition can occur in a sole step, characterized by the
concerted attacks of the electrophile and nucleophile, at
both ethylenic carbon atoms without formation of an
intermediate with a significant lifetime in the presence
of nucleophiles. Alternatively, the electrophilic addition
can be a two step process involving the formation of an
ionic intermediate with a sufficient lifetime to diffuse
through the solvent before it reacts with the present
nucleophiles. Between these two limiting mechanisms
when an unstable intermediate can be formed with a very
short but significant lifetime, a preassociation stepwise
mechanism20 involving ionization within a ternary com-
plex between the electrophile (eq 3), the ethylenic sub-
strate and the nucleophile has been shown to occur in
electrophilic bromination.21 In this latter mechanism, the
stereochemical outcome of the addition is controlled by
the stereochemistry of the initially formed ternary com-
plex.
Therefore, taking into account the generally accepted
mechanisms for each of the investigated reaction and the
different effects which may affect the facial stereoselec-
tivity, the observed stereochemical behaviors can be
rationalized as reported below.
i. IN3 Ad d ition in Aceton itr ile. IN3 addition, as
sulfenylation and bromination in aprotic solvents, is a
stepwise reaction involving an ionic intermediate.22 In
most cases, the formation of an iodonium ion of type I by
iodination is reversible.22 However, when its trapping
agent is the strongly nucleophilic azide ion instead of less
nucleophilic species such as methanol or iodide ion, the
second step is likely very fast since N3- is known to react
with carbocations at diffusion-controlled rates.18 There-
fore, return (Scheme 4) is not very significant in the case
of IN3 addition, so that the product distribution shown
in Table 1 is close to be kinetically controlled. The
insignificant return is supported by the fact that
iodoalkoxylation24a of the same glucals by iodine addition
in protic solvents exhibits a stereoselectivity smaller than
that shown in Table 1 for azidoiodination. In agreement
with the complete iodine bridging of the intermediate,
the addition of the two electrophilic and nucleophilic
partners of IN3 is trans, whatever the substituents. The
product data of Table 1 establish clearly that the C-3 and
C-4 substituents do not influence markedly the stereo-
selectivity. Moreover, the stereochemistry of the sub-
(21) Ruasse, M.-F.; Lo Moro, G.; Galland, G.; Bianchini, R.; Chiappe,
C.; Bellucci, G. J . Am. Chem. Soc. 1997, 119, 12492.
(22) Schmid, G. H.; Garratt, D. G. In The Chemistry of Functional
Groups: The Chemistry of Double-bonded Functional Groups, Supple-
ment A; Patai, S., Ed.; Wiley: London, 1977; Part 2, p 828.
(23) Thiem, J .; Ossowski, P. J . Carbohydr. Chem. 1984, 3, 287.
(24) (a) Horton, D.; Priebe, W.; Sznaidman, M. Carbohydr. Res. 1990,
205, 71. (b) Lafont, D.; Descotes, G. Carbohydr. Res. 1987, 166, 1953.
Lafont, D.; Descotes, G. Carbohydr. Res. 1988, 175, 35. (c) Griffith, D.
A.; Danishefsky, D. A. J . Am. Chem. Soc. 1990, 112, 5811.
(19) According to eq 2, the plot of 1/faz versus 1/[N3-] gives a straight
line,17,18 the slope of which is kMeOH/kN since [Br-] is negligible as
3
compared to [N3-] and MeOH (no dibromide is observed). The intercept
affords kB.
(20) J encks, W. P. Chem. Soc. Rev. 1981, 10, 345. Amyes, T. L.;
Richard, J . P. J . Am. Chem. Soc. 1990, 112, 9507.

8474 J . Org. Chem., Vol. 65, No. 25, 2000
Boschi et al.
Sch em e 5
stituent at C-4 does not influence it either, since,
according to previous results,24b azidoiodination of 1 and
3 exhibits the same stereochemical behavior.
In the absence of significant return, the marked
predominance of the R-trans-azido iodo adducts a is to
be related to a preferred top-side axial approach of the
electrophilic iodine atom to the double bond of 1, 4 and
5, i.e., on the face opposite to that observed and predicted
by PPFMO theory for sulfenylation. The large difference
between the steric requirements of the electrophilic sulfur
and iodine species are probably at the origin of this
disagreement. In contrast, our results are more consis-
tent with previous results showing the iodine preference
for an above-plane attack.24
ii. Br om in e Ad d it ion in t h e Ap r ot ic Dich lor o-
eth a n e. Bromination in DCE is also a two-step reaction
involving an ionic intermediate which, according to
previous results on acyclic enol ethers, is a bromo
oxocarbenium ion (III).25 Therefore, the formation of this
intermediate is irreversible and the stereochemistries of
the two consecutive steps are not necessarily related each
other. In agreement with the unbridged nature of the
intermediate, both cis and trans adducts a and c are
observed, in contrast with iodination. Analogous behavior
was previously found9 in the bromination of acetyl glycals
in various aprotic solvents.
it has been shown that the 5H4 conformation becomes
increasingly important for D-glucal derivatives when the
C(5) substituent becomes more electronegative. Although
glucal 2 exists preferentially in 4H5 conformation, it is
however possible that this compound reacts, at least
5
partly, in the H4 conformation, which is characterized
by a different face diastereoselectivity with respect to 4H5.
5
In particular, the two faces of the H4 conformation are
A striking result is that the capture of the bromo
oxocarbenium ions occurs always by nucleophilic R-at-
tack, whatever the number and the nature of the sub-
stituents of the six-membered ring. This is not readily
understood in terms of the recently established prefer-
ence of these cations to react via their pseudoaxial
conformation.13 In particular, benzyloxy substituents at
C-3 or C-4 were found to favor strongly this conformation
because of stabilizing electrostatic interactions between
the partially negatively charged hydroxyl oxygen and the
cationic oxygen atom.14 Nevertheless, the combination of
steric and polar effects of the three substituents in 1, 2,
and 3 and the differences in the requirements of the
involved nucleophiles make any stereochemical interpre-
tation hazardous.
differentiated to much extent by steric and electrostatic
factors. The upper face (â) is hindered by two axial
substituents, while the bottom (R) is shielded only by the
axial C(4)-OR group in glucals 1 and 2. Moreover, when
the latter substituent is an acyloxy group, as in 2, the
transition state leading to the ionic intermediate ii′
(Scheme 5) could be stabilized by a charge-dipole interac-
tion. The bromine addition to compounds 1-5 is almost
4
5
certainly slow compared to H5 S H4 interconversion.
Therefore, in agreement with the Curtin-Hammett prin-
ciple, the product distribution is determined exclusively
by the relative energies of the competing transition states
and is independent of the distribution of ground state
conformers. On this basis, assuming that the transition
state for Br2 addition is half-chairlike, glycals that
More interestingly, the diastereofacial selectivity of the
electrophilic attack of the carbon-carbon double bond is
strongly substituent-dependent. In the bromination of 4
and 5, i.e., in the absence of C-3 and C-4 substituents,
the top-side axial bromine approach is largely predomi-
nant, as it was found for iodination of the same sub-
strates and in agreement with PPFMO theory predictions
when an unbridged intermediate is involved.12 In con-
trast, the equatorial approach is moderately to exclu-
sively favored when benzyloxy or acetoxy substituents
are at C-3 or C-4 positions. A large preference for this
approach is observed for 3 in which steric 1,3-interac-
tions, opposing the effect of through-space electron-
donation of the axially oriented electronegative substit-
uent at C-4, play a role in determining the facial
selectivity of the electrophilic step. The increased dias-
tereoselectivity in favor of the gluco or galacto derivative
c, on going from 1 to 2 and 3 can be rationalized
considering that glycals are conformationally flexible23
5
preferentially adopt the H4 conformation in the transi-
tion state, as well as probably 2 and 3, should exhibit
higher diastereofacial selectivity in favor of the dibromo
adduct c. A similar interpretation has been recently
proposed to rationalize the facial stereoselectivity ob-
served in sulfenylation of glycals bearing electronegative
substituents at C-6 and polar substituents at C-4.26
iii. Br om in a t ion Usin g t h e Tr ib r om id e Sa lt ,
Bu 4NBr 3. In contrast with bromination by free bromine,
the very major product of the tribromide ion addition is
the trans adduct a arising from the diaxial attack of the
two bromine atoms, except in the case of 3. This result
agrees fairly well with the well-established mechanism
of bromination by tribromide salts.27 The trans-addition
of the two bromine atoms occurs via a concerted process,
i.e., without the involment of any ionic intermediate, by
rate- and product-determining nucleophilic attack of a
bromide ion on initially formed olefin-bromine π-com-
plexes (Scheme 6). The stereochemical results which,
again, are not predictable from the PPFMO theory, are
4
and therefore both the normal half-chair H5 conforma-
tion and conformationally inverted isomer, 5H4, may
contribute to the product distribution.23,24 Furthermore,
(26) Roush, W. R.; Sebesta, D. P.; Bennett, C. E. Tetrahedron 1997,
53, 8825.
(25) Ruasse, M. F.; Dubois, J . E. J . Am. Chem. Soc. 1984, 106, 3220.
Ruasse, M. F. Isr. J . Chem. 1985, 26, 414.
(27) Bellucci, G.; Chiappe, C.; Lo Moro, G. J . Org. Chem. 1997, 62,
3176.

Diastereofacial Selectivity in Glycals
J . Org. Chem., Vol. 65, No. 25, 2000 8475
Sch em e 6
readily rationalized considering that the formation of
R-trans adducts a is markedly favored by stereoelectronic
R-anomeric effects. The much smaller selectivity observed
for this addition to 3 may be attributed to unfavorable
steric and electrostatic interactions between the axial C-4
benzyloxy group and the electrophile, which slow mark-
found previously for bromination of acyclic enol ethers
in protic solvents.25 The most important result for the
interpretation of these data is the absence of any azide
incorporation in the trapping experiments in the presence
of azide ions, which suggests strongly that there is no
intermediate in the methanolic bromination of these
substrates, i.e., that these reactions are concerted and
not stepwise (Scheme 7). This result agrees fairly well
with previous findings that nucleophilic substitutions, a
class of carbocation-forming reactions exhibiting some
common features with bromination,21 on glycosyl deriva-
tives cannot involve carbocationic intermediates in the
presence of strong nucleophiles.29 For example, rate
constants of 1012 s-1and 1019 s-1, much larger than that
expected for diffusion control, has been extrapolated for
the reaction of the glycosyl cation with water and azide
ion, respectively, showing that this cation cannot exist
when strong nucleophiles are present. Because of the
destabilizing effect of the electron-withdrawing hydroxyl
substituents, probably reinforced by the â-bromo sub-
stituent in bromination, these ionic species are too
unstable, too short-lived to exist in the presence of strong
nucleophiles and in particular in methanol. In contrast,
they can exist in DCE in which the only nucleophile is
the bromide ion liberated by the formation of the inter-
mediate.
edly the C(2)-Br bond formation in the axial position.
-
Unexpectedly, the minor products in these Br3
-
additions are not the trans adducts b arising from
â-anomeric attack of Br- but those resulting from a cis-
addition of the two bromine atoms from the bottom face.
Clearly, the unfavorable â-anomeric effect inhibits totally
the trans-addition leading to b. The formation of the cis-
addition products c could be attributed to the addition
of some free bromine since adducts c are also obtained
in the reaction of this electrophile (vide supra). Some free
bromine arising from a preliminary dissociation of the
tribromide ion according to eq 4 can, actually, be present
since tribromide addition to glycals can be slow as
compared to its dissociation28 (k+ ) 5 × 102 s-1 in DCE),
which is not generally observed in the
k+
\
Br3- y
z Br2 + Br-
(4)
k-
Br3--bromination of acyclic alkenes.
iv. Br om in a tion in Meth a n ol. The stereochemical
outcome of bromine addition in methanol is markedly
different from that of the reaction of the same electrophile
in the nonprotic dichloroethane, although identical bromo
oxocarbenium ion intermediates could be postulated a
priori. Still more surprising is the exclusively trans-
addition of the two electrophilic and nucleophilic part-
ners, as it was observed for iodination via a bridged
iodonium ion but not for the concerted Br3--bromination.
Moreover, in contrast with iodination, the methanolic
bromination of 1, 4 and 5 are almost nonselective, the
two adducts a and b being obtained in closely similar
amounts. Only in the case of 2 or 3, some diastereofacial
selectivity is observed. This poor selectivity and its small
sensitivity to substituent effects is consistent with very
early and weakly charged transition states, as it was
The marked differences in the diastereofacial selectivi-
ties of the tribromide ion addition and the methanolic
bromination which both are concerted reactions, can be
rationalized in terms of tight and loose transition states,
respectively (Schemes 6 and 7). Very early transition
states for bromination in methanol are supported by
unusually small solvent and substituent kinetic effects
on the very fast reaction of enol ethers25 whereas tribro-
mide ion additions involve likely late transition states
with significant charge development. Therefore, the steric
and electronic interactions30 contributing to the stereo-
selectivity of the two reactions are markedly stronger
when the nucleophile is a bromide ion.
(29) (a) Young, P. R.; J encks, W. P. J . Am. Chem. Soc. 1977, 97,
8238. (b) Amyes, T. L.; J encks, W. P. J . Am. Chem. Soc. 1989, 111,
7888. (c) Banait, N. S.; J encks, W. P. J . Am. Chem. Soc. 1991, 113,
7951.
(30) Chre´tien, J . R.; Coudert, D.; Ruasse, M. F. J . Org. Chem. 1993,
58, 1917.
(28) Bellucci, G.; Bianchini, R.; Ambrosetti, R.; Ingrosso, G. J . Org.
Chem. 1985, 50, 3313. Ruasse, M. F.; Aubard, J .; Galland, B.; Adenier,
A. J . Phys. Chem. 1986, 90, 4382.

8476 J . Org. Chem., Vol. 65, No. 25, 2000
Boschi et al.
Sch em e 7
In this context, the significant lifetime of the oxocar-
benium ion derived from 2 (kMeOH ) 3 × 107 s-1) in which
the substituents are acetoxy, instead of benzyloxy groups,
deserves a peculiar attention. First, although the electron
withdrawing effects of these two substituents are very
close,31 the trapping of the corresponding intermediate
is drastically slown down. Second, a significant diaste-
reofacial selectivity, analogous to that observed for iodi-
nation or tribromide ion addition, is restored, with a
predominance of top-side axial approach of the bromine
to the ethylenic bond. These two results suggest a
preassociation mechanism (eq 2) for the reaction of 2 in
which methanol on one side and bromine on the other
side preassociate with the enolic double bond before
ionization within the ternary complex in a cationic
intermediate formed in sandwich between the leaving
bromide and the entering methanol.21
axially top-side for the reactions of 4 and 5, except again
for bromination in methanol. The axial or equatorial
conformation of the substituent at C-4 plays also an
important role. On going from 1 to 3 with an equatorial
or axial benzyloxy substituent, respectively, the prefer-
ence for the axial approach of the electrophile is strongly
attenuated (Br3--addition, methanolic bromination) or
disappears (bromination in DCE) at the benefit of an
equatorial approach, in agreement with strong steric and
electrostatic 1,3-interactions. Finally, the substitution of
a benzyloxy by an acetoxy group has also a noticeable
effect on the stereochemistry of the bromine addition in
the polar and apolar solvents.
(iii) The usual interpretation in terms of the stereo-
electronic preference for R-anomeric attack10,11 still holds
in most of the investigated reactions. However, a more
detailed interpretation of the product-determining nu-
cleophilic steps in terms of the preferred pseudoaxial
conformation of the C-3 and C-4 benzyloxy substituents
of oxocarbenium ions13 does not fit the results quite well,
probably because of the presence at C-2 of a large,
polarizable halogen atom.
Con clu d in g Rem a r k s
Several conclusions emerge from our stereochemical
results which correspond to a kinetically controlled
product formation, either since the ionic intermediate is
obtained irreversibly (bromination in DCE) or quasi-
irreversibly (IN3 iodination) or since the electrophilic
addition is concerted (tribromide ion reaction, bromina-
tion in methanol).
(i) Whatever the six-membered ring substituents, the
favored mode of addition of the nucleophilic partner N
of an electrophilic reagent EN is a below-plane approach,
i.e., the stereoelectronic R-anomeric effect controls pre-
dominantly the widely preferred formation of adducts a
and c. The only exception to this rule is for the reaction
of bromine in methanol which does not exhibit any
marked selectivity for R- or â-attack.
(iv) Finally, in regard to the electrophile approach, the
PPFMO theory, which takes into account the orbital
interactions in the glycal itself,12 does not predict con-
sistently the top-face or below-face attacks of the iodine
and bromine observed in this work. This theory has been
developed for models of additions via open or bridged
intermediates, electrophilic protonation and sulfenyla-
tion, respectively, and could be a priori extrapolated to
bromination via oxocarbenium ions and iodination via
iodonium ions. Nevertheless, it is well-known that iodi-
nation and sulfenylation exhibit very different stereo-
chemical behaviors. It is also quite obvious that bromine
and proton are very different electrophiles not only in
their electronic and steric requirements but also in the
mechanism of their reactions.
(ii) In contrast, the diastereofacial selectivity of the
electrophile approach is more versatile and depends
significantly on the substituents. The substituents at C-3
and C-4 influence markedly the selectivity since in their
absence the electrophile approach is almost exclusively
(v) In this respect, the most important result is
probably to point out that the selectivities of these
reactions are closely related to their specific mechanisms.
The three bromination methods used in this work involve
(31) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119.

Diastereofacial Selectivity in Glycals
J . Org. Chem., Vol. 65, No. 25, 2000 8477
) 9.8 Hz), 1.36-2.06 (m, 6H). 13C NMR (CDCl3, δ ppm):
126.5-128.9, 97.4, 92.8, 79.1, 71.5, 18.4-34.2.
all usual brominating agents but the reaction mecha-
nisms and, therefore, the stereochemistries are very
different. The addition of free bromine in an aprotic
solvent is a two-step reaction involving bromo oxocarbe-
nium ions and not bromonium ions so that the stere-
ochemistries of the two consecutive electrophilic and
nucleophilic steps can be determined by different factors.
The reaction of tribromide ions is a concerted addition
of the electrophile and the nucleophile, implying an
enforced cooperation between the stereochemistries of the
two partners. Finally, bromination of the same substrates
in nucleophilic methanol appears to be also a concerted
addition, even though the same reaction in the same
solvent of other ethylenic compounds is well-established
to be stepwise.32 This is fairly well consistent with the
small kinetic and stereochemical sensitivity to substitu-
ent effects25 and also with previous findings on hydrolysis
of glycosyl derivatives.29 More work is in progress to
understand in more details the reaction mechanisms of
glycals and related cyclic enol ethers and their differences
with those established from more usual ethylenic com-
pounds.
2-P h en yl-3,4-d ih yd r o-2H-p yr a n 35 (5). To a solution of 9
(1.2 g) in dry toluene (45 mL) was added P2O5 (8 mg), and the
solution was allowed to reflux for 2 h. After being cooled at
room temperature, the reaction mixture was washed with an
aqueous solution of NaHCO3 and the organic phase was dried
over anhydrous MgSO4. Filtration and concentration of the
filtrate in vacuo provided 1.08 g of a crude residue wich was
chromatographed on a silica gel column. Elution with hex-
anes-ethyl acetate (8:2) gave 500 mg of pure 5 (46%). 1H NMR
(CDCl3, δ ppm): 7.23-7.37 (5H), 6.53 (d, 1H, J ) 6.09 Hz),
6.74-4.85 (m, 2H), 1.9-2.05 (m, 4H). 13C NMR (CDCl3, δ
ppm): 144.1, 141.9, 128.3, 125.8, 100.6, 77.0, 30.2, 20.2.
2-Ben zyloxym eth yl-3,4-d ih yd r o-2H-p yr a n 36 (4). To a
solution of 2-hydroxymethyl-3,4-dihydro-2H-pyran (1.37 g, 12
mmol) in 26 mL of THF containing 0.5% of H2O were added
138 mg (0.52 mmol) of 18-crown-6 and 2.6 g of powdered KOH.
The mixture was stirred at room temperature for 20 min, and
1.5 mL (12 mmol) of benzyl bromide was added. After 24 h at
the same temperature, 6 mL of MeOH was added. The reaction
mixture was stirred for 10 min, and the solvent was evaporated
under reduce pressure to give a residue which was diluted with
100 mL of CH2Cl2. The organic phase was washed with water,
dried (MgSO4), filtered, and concentrated in vacuo. The residue
was chromatographed on a silica gel column. Elution with
hexane-ethyl acetate (7:3) gave 1.78 g of pure 2-benzyloxy-
methyl-3,4-dihydro-2H-pyran (4). 1H NMR (CDCl3, δ ppm):
7.34-7.4 (m, 5H), 6.42 (d, J ) 6.2 Hz), 4.8-4.5 (m, 3H) 4.04
(m, 1H), 3.57 (ddd, 2H), 1.4-1.2 (m, 4H). 13C NMR (CDCl3, δ
ppm): 143.5, 138.1-127.6, 100.4, 74.0, 73.3, 72.4, 24.5-19.3.
Ad d ition s of ICl a n d Na N3. Iodine monochloride (0.55
mmol) was added to a suspension of 1.5 mmol of NaN3 in 2.5
mL of CH3CN cooled at 0 °C. After 10 min, the cyclic enol ether
(4 or 5) (0.5 mmol) dissolved in 2.5 mL of the same solvent
and precooled at 0 °C was added. The mixture was stirred at
0 °C for 15 min, diluted with a 10% of aqueous NaHSO3,
extracted with CH2Cl2, dried, and evaporated. The crude
residue was analyzed by 1H and 13C NMR. All experiments
were carried out at least in triplicate.
Exp er im en ta l Section
1H and 13C NMR spectra were registred in CDCl3 containing
TMS as the internal reference. All solvents were reagent grade
and were used without further purification. Commercial tri-
O-benzyl-D-glucal (97%), tri-O-benzyl-D-galactal (98%), tri-O-
acetyl-D-glucal, tetrabutylammonium tribromide (98%), bro-
mine (1 mL sealed ampules), iodine monochloride, sodium
azide, 2-benzoylbutyric acid and 2-hydroxymethyl-3,4-dihydro-
2H-pyran were used as supplied.
2-P h en yl-δ-va ler ola cton e33 (8). To a water (23 mL) solu-
tion of benzoylbutyric acid (6) (2.1 g, 10.92 mmol) containing
NaOH (5%) was added 0.5 g of solid NaBH4 (13.21 mmol), and
the reaction mixture was stirred at room temperature for 6 h.
After 20 h at room temperature, CHCl3 (25 mL) was added,
and the aqueous phase was acidified by addition of HCl. The
aqueous phase was then extracted with CHCl3 (3 × 5 mL),
and the combined organic extracts were dried (MgSO4) and
evaporated in vacuo. The crude product34 (7) (2.07 g) was
pyrolized at 80-106 °C in vacuo, collecting the water in a
Dean-Stark trap. Recrystallization from diethyl ether gave
1.46 g of compound 8 as a crystalline solid (78% yield). 1H NMR
(CDCl3, δ ppm): 7.32 (5H), 5.32 (dd, J ) 10.2, 3.2 Hz, 1H),
2.51-2.65 (m, 2H), 1.79-1.98 (m, 4H). 13C NMR (CDCl3, δ
ppm): 171.2, 139.4, 125.4-128.2, 81.3, 30.1, 29.1, 18.2.
2-P h en yl-δ-va ler ola ctol33 (9). To a solution (95 mL) of
2-phenyl-δ-valerolactone (8) (1.45 g, 8.3 mmol) in dry CH2Cl2
was added DIBAL-H (10.5 mL, 1.0 M in hexane) at -78 °C.
The reaction mixture was stirred for 2 h, quenched with a
saturated aqueous NaHCO3 solution (30 mL), diluted with 20
mL of CH2Cl2, and warmed at room temperature under
magnetic stirring. After washing with an aqueous solution of
HCl 10% and NH4Cl (1:1.5), the organic phase was separated,
dried (MgSO4), filtered, and concentrated under reduced
pressure to afford 1.68 g of lactol 9 (94%) as a white crystalline
solid. 1H NMR (CDCl3, δ ppm): 7.1-7.5 (5H), 5.39 (s, 1H), 4.99
(dd, 1H, J ) 2.3 Hz), 4.78 (d, 1H, J ) 9.6 Hz), 4.43 (d, 1H, J
Ad d ition s of Br om in e a n d Bu 4NBr 3. (a ) In 1,2-Dich lo-
r oeth a n e. 1,2-Dichloroethane solutions of Br2 or Bu4NBr3 (5
mL, 10-2 M) were mixed with 5 mL of a 10-2 M solution of 2,
or 3, or 4 or 5 in the same solvent. After the rapid disappear-
ance of color the mixture was evaporated (washed with water
for the Bu4NBr3 reactions) and the residue was analyzed by
1H and 13C NMR.37 All experiments were carried out at least
in triplicate.
(b) In Meth a n ol. The bromination in MeOH was carried
out in the same way as bromination in 1,2-dichloroethane. At
the end of reaction the mixture was diluted with water,
extracted with CH2Cl2, and analyzed by 1H and 13C NMR.37
All experiments were carried out at least in triplicate.
(c) In Meth a n ol, in th e P r esen ce of Na N3. To solutions
of NaN3 (0.1-0.285 M) in methanol (20 mL), containing the
glycal (1 × 10-2 M), an equivalent of bromine in same solvent
(1 mL) was added under stirring and the reaction mixtures
were stored in the dark at 25 °C. After the rapid disappearance
of color the reaction mixtures were diluted with water and the
products were extracted with CH2Cl2. The combined organic
layers were dried (MgSO4), evaporated in vacuo, and analyzed
1
by H and 13C NMR. Dibromo adducts were always insignifi-
cant. All experiments were carried out at least in triplicate.
Ack n ow led gm en t. The authors acknowledge the
financial contributions of the CNR and the University
of Pisa (Italy).
(32) Ruasse, M. F. Adv. Phys. Org. Chem. 1993, 28, 207.
(33) Downham, R.; Edwards, P. J .; Entwistle, D. A.; Hugues, A. B.;
Kim, K. S.; Ley, S. V. Tetrahedron: Asymmetry 1995, 6, 2403.
(34) Colonge, J .; Costantini, M.; Ducloux, M. Bull. Soc. Chim. Fr.
1966, 2005.
(35) Arai, I.; Daves, G. D. J . Org. Chem. 1979, 44, 21.
(36) Saroli, A.; Descours, D.; Carret, G.; Anker, D.; Pacheco, H.
Carbohydr. Res. 1980, 84, 71.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
data of adducts a , b, and c (Table S1). 1H and 13C NMR spectra
(Figures S1-S9) of adducts a , b, and c from 3, 4, and 5. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(37) Bromination products of 2 in DCE and in MeOH: (a) Lemieux,
R. U.; Fraser-Reid, B. Can. J . Chem. 1964, 42, 532. (b) Teichmann,
M.; Descotes, G.; Lafont, D. Synthesis 1993, 889.
J O000799X