1882
C. Hultsch et al. / Tetrahedron Letters 49 (2008) 1881–1883
MeO2C
mized protocol are summarized in Column 3 of Table
N
MeO2C
MeO2C
1.14 Although a variety of variations regarding reaction
time, temperature and excess of substrates were investi-
gated, suppression of hydrogen abstraction from arecoline
2 and ecgonidine methylate 3 could not be achieved. Con-
sequently, [18F]fluorobenzene was formed in up to 60%
yield (see Supplementary data).
N
N
2
TiCl3
F
F
5
6
MeO2C
N
N2+ X-
3
In contrast to the fluoroarylation of arecoline 2 and
ecgonidine methylate 3, where significantly lower yields
were obtained for radiosyntheses compared to ‘cold’ syn-
theses, the small reaction scale and the increase in substrate
equivalents turned out to be beneficial for the formation of
stilbene 7. Radiochemical yields of up to 70% were
observed after reaction times of only 5 min.15 The opposed
behaviour of arecoline 2 and ecgonidine methylate 3 com-
pared to bromostyrene 4b in the radiochemical experiments
is most probably due to undesired hydrogen abstraction by
the aryl radical. In the non-radioactive optimization exper-
iments with arecoline, only a minor increase in yield (from
50% to around 60%) was found when the amount of arec-
oline was doubled from 2.5 to 5 equiv, since the side reac-
tion also benefits from higher substrate concentrations.
Experiments with bromostyrene 4b gave a comparatively
larger relative increase in yield from 21% to 32% when con-
ducted with four instead of two equivalents. In conclusion,
substrates which undergo no specific side-reactions (e.g.,
hydrogen abstraction from allylic positions, radical addi-
tion to an alternate functional groups) are much more
likely to benefit from the manifold excess in radiochemical
syntheses. In the case of stilbene 7, the dramatic increase in
equivalents made a synthetically low-yielding process
become an efficient method in radiochemistry.
TiCl3
F
1
N
N
X
4
TiCl3
4a: X = Cl
4b: X = Br
7
F
Scheme 1. Fluoroarylation of arecoline 2, ecgonidine methylate 3 and
styrenes 4a and 4b.
Table 1
Chemical and radiochemical fluoroarylation of arecoline 2, ecgonidine
methylate 3 and styrenes 4a and 4b
Substrate
Product,
yielda (%)
Product, radiochemical
yield (%)
Arecoline 2
5, 54b
6, 51c
7, 28
7, 32
[
[
—
18F]-5, 613d,b
Ecgonidine methylate 3
Chlorostyrene 4a
Bromostyrene 4b
a
18F]-6, 616d
[
18F]-7, 60–70e
Reactions according to general procedure. Isolated yields.
Diastereoselectivity: trans:cis = 58:42 (non-radioactive), trans:cis =
b
50:50 (radioactive).
c
Diastereoselectivity >80% (43% of 2b,3a-isomer, 8% for remaining
three isomers).
d
Reaction time:10 min.
Reaction time: 5 min.
e
In combination with the preparation of the intermediate
4-[18F]fluorobenzenediazonium salt, the Alzheimer plaque
imaging reagent 7 is now accessible in 80 min total reaction
time with overall radiochemical yields of 30–45% (decay
corrected) starting from [18F]fluoride, which is the superior
to the previously reported procedures.16
In summary, the first successful application of aryl rad-
icals for the synthesis of radiopharmaceuticals is reported
and exemplified for the production of 18F-labelled stilb-
enes, one class of Alzheimer imaging agents. Exploiting
the characteristics of earlier aryl radical reactions under
comparable conditions, this methodology should be insen-
sitive towards many functional groups (e.g., amino groups,
hydroxyl groups, carboxylic acids, nitriles and ketones),5,6
and opens a new route to radiopharmaceuticals not avail-
able with commonly used approaches. Furthermore, based
on the experiences with 18F-substituted, deactivated aro-
matic systems it can be assumed that radiopharmaceuticals
synthesized by this methodology will show excellent stabil-
ity towards defluorination in vivo.
the anti-depressive drug paroxetine,9 as well as the (2b, 3a)-
aryltropane 6, which has been found to have high affinity
(21 nM) for the dopamine transporter (DAT),10 were
accessible in synthetically useful yields, the latter even with
remarkable diastereoselectivity. Hydrogen abstraction
from arecoline and ecgonidine methylate, leading to the
formation of fluorobenzene, was determined to be the main
competing process.11 We therefore found it surprising that
the experiments with both styrenes 4a and 4b, which do not
bear easily abstractable hydrogen atoms, gave significantly
lower yields. Styrenes 4a and 4b were chosen since stilbene
7 is an important lead structure for Alzheimer plaque
imaging.12
In contrast to the preliminary experiments described
above, syntheses of no-carrier-added radiopharmaceuticals
are usually performed on a much smaller scale (usually lM
for the labelling precursor and pM for fluorine-18). The
yields obtained for the ‘cold’ syntheses therefore do not
necessarily allow a prediction for yields obtained for radio-
active syntheses. Based on the literature-known prepara-
tion of 4-[18F]fluorobenzenediazonium salts4,13 and the
procedure developed for the non-radioactive synthesis, we
then investigated the [18F]fluoroarylation under radio-
chemical conditions. The results obtained with the opti-
Acknowledgements
This project was financed by the Fonds der Chemischen
Industrie (Liebig fellowship). We are grateful to the Leon-