2068
J. Cardinale, J. Ermert / Tetrahedron Letters 54 (2013) 2067–2069
31% yield proving the possibility of a one pot procedure. In an addi-
tional experiment starting from iodosobenzene and Meldrum’s
acid the desired product was isolated in 30% yield, showing no dif-
ference in the reactivity of diacetoxyiodobenzene and iodosoben-
zene which is consistent with the results of Goudreau et al.7
discussed above. In general, results of all procedures were compa-
rable under the chosen conditions with the yield of the desired
product 1 amounting to 33 3%.
Scheme 2. Resonance structures of phenyliodonium ylide 1.
As further examples of the synthesis of more complex iodonium
ylides, that is, isomeric iodophenoxyethers, 3-benzyloxy-phenyli-
odonium-(5-[2,2-dimethyl-1,3-dioxane-4,6-dione]) ylide and 4-
benzyloxy-phenyliodonium-(5-[2,2-dimethyl-1,3-dioxane-4,6-dione])
ylide were prepared by the one pot procedure (Table 1, entries 2
and 3). While the para-derivative was produced in 44% yield it
was not possible to isolate the desired product in the case of the
meta-derivative under the same reaction conditions. However, in
the case of the iodonium ylides derived from 2-, 3-, 4-iodoanisole
and 4-iodotoluene the one-pot procedure worked quite well
(Table 1, entries 4–7). The yield of 2-methoxyphenyl-iodonium-
(5-[2,2-dimethyl-1,3-dioxane-4,6-dione]) ylide amounting to 59 %
was surprisingly high. This could be due to an additional stabiliza-
tion of the iodine-III centre by the neighbouring oxygen. Addition-
ally 2-, 3- and 4-bromophenyliodonium-(5-[2,2-dimethyl-1,3-
dioxane-4,6-dione]) ylide, which are possible precursors for the
corresponding [18F]fluorobromobenzenes, were also synthesized
by the one-pot procedure (Table 1, entries 8–10). While the yields
of the ortho- and the para-derivatives were in the typical range of
about 30% (Table 1, entry 2), the yield of 13% of the meta-derivative
was noticeably low. This trend was also observed in the case of the
anisole derivatives and could explain the problems in the synthesis
of 3-benzyloxyphenyliodonium-(5-[2,2-dimethyl-1,3-dioxane-4,6-
dione]) ylide. It seems to be correlated with electronic factors influ-
encing either the oxidation or the stability of the product (or both).
Recently, we reported an improved synthesis of (4-iodo-
phenyl)(aryl)iodonium salts which are precursors, for example, for
the versatile intermediate 4-[18F]fluoro-1-iodobenzene.12 Since
4-iodophenyliodonium-(5-[2,2-dimethyl-1,3-dioxane-4,6-di-one])
ylide would be a further promising precursor for 4-[18F]fluoro-1-
iodo-benzene, an attempt was made to prepare this compound by
the one-pot procedure. The yield of 19% was relatively low (Table 1,
entry 11), which is presumably due to the previously reported diffi-
culties in the oxidation of 1,4-diiodobenzene.11 Thus, it was also
tried to couple 4-iodo-1-[hydroxyl(tosyloxy)iodo]benzene, a deriv-
ative of Koser’s reagent, to Meldrum’s acid under conditions similar
to those of Goudreau et al.7 This way, the desired product was
achieved in 42% yield, again demonstrating that coupling of various
iodine-III precursors is possible in order to obtain iodonium ylides.
In conclusion, the possibility of a one-pot procedure for the syn-
thesis of iodonium ylides was demonstrated. The application to two
iodobenzyloxybenzene isomers, however, was only successful in the
case of the 4-isomer while the synthesis of the corresponding ani-
sole derivatives was successful for all three isomers. Additionally,
the synthesis of the iodonium ylides derived from the o-, m- and
p-bromoiodobenzenes, possible precursors for the corresponding
Scheme 3. Synthesis of iodonium ylide 1.
Scheme 4. Equilibrium between the phenyliodosobase and its salts
ðX ¼ RCOꢀ2 ; ClꢀÞ.10
modifications (base, solvent) of the method according to Scheme 3
starting from compound 2 and Meldrum’s acid. Here, the best re-
sults were obtained with the procedure by Goudreau et al.7 deliv-
ering 1 in 31% yield. The yield obtained, however, was lower than
reported which was also observed in the following reactions (see
below). This is probably due to slight differences in the isolation
of the only moderately stable product. However, since the isolation
of the products here was always performed by the same protocol,
the results obtained are considered to be comparable.
Alternatively iodobenzene was first oxidized with meta-chloro-
peroxobenzoic acid (mCPBA) in DCM, an oxidation procedure
which has been used by Bielawski and Olofsson for the synthesis
of iodonium salts,11 and subsequently a suspension containing
Meldrum’s acid and KOH was added. Using this procedure product
1 was formed in 36% yield. Obviously meta-chlorobenzoate does
not have a negative impact on the reaction. Finally, the reaction
was also conducted as a two-step one-pot procedure adding KOH
and Meldrum’s acid directly as solids to previously mCPBA-oxi-
dized iodobenzene (Table 1). Here the product was formed in
Table 1
Iodonium ylides produced by the one-pot procedure
[
18F]fluorobromobenzenes, was accomplished. Furthermore, the
synthesis of 4-iodophenyl-iodonium-(5-[2,2-dimethyl-1,3-diox-
ane-4,6-dione]) ylide, a promising precursor of 4-[18F]fluoro-1-
iodobenzene, was accomplished by two different procedures.
Entry
R
Yield (%)
1
2
3
4
5
6
7
8
9
H
31
0
3-OBn
4-OBn
2-OMe
3-OMe
4-OMe
4-Me
2-Br
3-Br
4-Br
4-I
44
59
26
35
20
29
13
28
19
Acknowledgments
The Authors thank Prof. Dr. Heinz H. Coenen for his guidance
and constant support, Sven Humpert for laboratory assistance,
Dr. Markus Holschbach (all INM-5) and Dr. Sabine Willbold
(ZCH), all Forschungszentrum Jülich, for recording the spectro-
scopic data.
10
11