8558
A. P. Lightfoot et al. / Tetrahedron Letters 45 (2004) 8557–8561
1
0
The iodo-deboronation processes outlined in Scheme 1
are applicable to other hindered esters besides pinacol.
Hence, hexylene glycol ester 1b also undergoes smooth
conversion to either iodoalkenes 2 or 3, with both high
stereoselectivities and yields.
tion product 5. Treatment of this 1:1 mixture of dia-
stereoisomers with NaOMe gave both the E-and Z-
iodoalkenes 2 and 3, respectively, as a 1:1 mixture of
iodoalkene stereoisomers (Scheme 2).
5
Although there is precedent for the syn-addition of
1
1
The mechanism of iodo-deboronation of alkenyl boro-
nates 1, with either retention or inversion of the alkenyl
dihalogen compounds to carbon multiple bonds, in
the present case, we were doubtful that such a process
was operating on these alkenyl boronate substrates 1.
However, it was noteworthy that 2Mequiv of ICl–pyr-
idine were required for the complete conversion of bor-
onate 1a to the 1:1 diastereoisomeric mixture of 4 and 5
(Scheme 2), perhaps suggesting that either chloride ion,
or pyridine were essential to assist the reaction. The
effect of chloride ion was directly tested by addition of
tetra-n-butylammonium chloride to the ICl addition
product 4, which completely reversed the ICl addition
reaction, resulting in a clean conversion back to the bor-
onate 1a. This shows that it must be the iodonium ion,
which is coordinated and controlled by chloride ion con-
centration since the ICl adduct 4 undergoes complete
elimination of ICl to regenerate the starting boronate
1a. Chloride must cause an anti-periplanar elimination
of ICl from adduct 4 by direct attack on iodine, rather
2
boronate geometry has been previously described by us
and was based upon the following: (1) addition of ICl to
1
boronate 1a (followed by H NMR) resulted in the for-
mation of a single diastereoisomer, assumed to be 4 on
the basis of an anti-addition of ICl via an iodonium
6
ion intermediate, hence, elimination with methoxide re-
sults in inversion of alkene geometry to give 3; (2) con-
version of boronate 1a to alkenyl iodide 2 was presumed
to occur as described by Brown et al. when using iodine-
1
hydroxide conditions. However, following problems
associated with clean iodo-deboronation of a substrate,
7
which contained an N-Boc protecting group, and
production of Z-alkenyl chloride by-products from E-
8
alkylalkenyl boronates (vide infra), we examined
alternative electropositive sources of iodine for achiev-
ing iodo-deboronation of alkenyl boronates, including
4
the readily prepared ICl–pyridine complex and
than causing S 2-mediated epimerisation of diastereo-
N
obtained some unexpected results that have led us to
re-examine the mechanisms operating in Scheme 1 and
in related iodo-deboronation processes.
isomer 4 to 5. Indeed, the addition of pyridine to adduct
4, results in attack at both iodine, and boron, which not
only regenerates ICl–pyridine, causing both diastereo-
isomer epimerisation and elimination, but also results
in the formation the starting material 1a and both alk-
enyl iodides 2 and 3. Hence, the alternative suggestion
that the mechanism of the ICl addition reaction initially
involves formation of anti-addition product 4, followed
by an equilibrium-controlled conversion to the syn-addi-
The reaction of E-boronate 1a with ICl can be followed
by H NMR and results in the formation of an interme-
1
diate initially presumed to be ICl adduct 4, which we
9
have found can be isolated as a single diastereoisomer
2
and confirms our previous findings. Indeed, subsequent
treatment of ICl adduct 4 with sodium methoxide results
in the rapid and efficient formation of the Z-iodoalkene
tion product 5 by S 2 inversion via chloride ion does
N
not explain why 2Mequiv of ICl–pyridine are required
for complete conversion of alkenyl boronate 1a, since
3 (Scheme 2). However, the formation of the corre-
sponding E-iodoalkene 2 from 1a could not be studied
the S 2 equilibration would merely require catalytic
N
1
by H NMR due to the high rate of reaction, that is, ini-
chloride ion, which is not the case. The alternative expla-
nation for the effect of chloride ion is to intercept the
expected intermediate iodonium ion formed by the addi-
tial treatment of boronate 1a with sodium methoxide,
followed by ICl addition provides alkenyl iodide 2 too
quickly to allow the observation of any intermediates.
+
tion of I to alkenyl boronate 1a, that is, producing an
4
In contrast, replacing ICl with ICl–pyridine complex
iodonium ion, which exists as a stable ion pair with chlo-
ride ion and probably further stabilised by pyridine,
since ICl alone rapidly adds across the alkenyl boronate
(to give 4). Such ion-paired intermediates, especially
involving iodide-iodonium ions have been inferred pre-
resulted in a slower reaction, which could be followed
by NMR. Two addition products (1:1 ratio) were ob-
served; one identical to that ascribed to the anti-addition
product 4, and the other being isomeric and presumed to
be the diastereoisomer of 4, that is, the formal syn-addi-
1
1
viously in related syn-addition reactions. In the present
Cl
H
I
I
Ph
NaOMe
H
B O
2 ICl.Py, CDCl3
ICl, CDCl3
H
H
B O
Ph
3
+
4
1a
Cl
O
O
5
4
2
n-Bu NCl
4
NaOMe
+ 3
2 Pyridine
2
1
1
2 + 3 + 1a
1 : 1 :
100 % conversion
:
1
2
00 % conversion
Scheme 2.