Highly enantioselective synthesis of no-carrier-added 6-[ 18F]fluoro-L-dopa by chiral phase-transfer alkylation

Article abstract of DOI:10.1002/ejoc.200400059

[¹⁸F]Fluoro-L-dopa, an important radiopharmaceutical for positron emission tomography (PET), has been synthesized using a phase-transfer alkylation reaction. A chiral quaternary ammonium salt derived from a Cinchona alkaloid served as phase-tra

Full text of DOI:10.1002/ejoc.200400059

FULL PAPER
Highly Enantioselective Synthesis of No-Carrier-Added 6-[18F]Fluoro-
by Chiral Phase-Transfer Alkylation
L
-dopa
[a]
[a]
[b]
[a]
[a]
´
¨
Christian Lemaire,* Steve Gillet, Stephane Guillouet, Alain Plenevaux, Joel Aerts,
[a]
´
and Andre Luxen
Keywords: Amino acids / Enantioselectivity / Fluorides / Radiochemistry / Radiopharmaceuticals / Phase-transfer catalysis
[
18F]Fluoro-
L
-dopa, an important radiopharmaceutical for po-
veloped. After labelling, the labeled [18F]fluoroveratralde-
hyde was trapped on a tC18 cartridge and then converted on
the cartridge into the corresponding benzyl halide derivati-
sitron emission tomography (PET), has been synthesized
using a phase-transfer alkylation reaction. A chiral quater-
nary ammonium salt derived from a Cinchona alkaloid ser- ves by addition of aqueous sodium borohydride and gaseous
ved as phase-transfer catalyst for the enantioselective alkyla- hydrobromic or -iodic acid. Hydrolysis and purification by
tion of a glycine derivative. The active methylene group of
this Schiff-base substrate was deprotonated with cesium hy-
droxide and rapidly alkylated by the 2-[18F]fluoro-4,5-di-
preparative HPLC made 6-[18F]fluoro-
L-dopa ready for hu-
man injection in a 25−30% decay-corrected radiochemical
yield in a synthesis time of 100 min. The product was found
methoxybenzyl halide (X = Br, I). The reaction proceeded to be chemically, radiochemically and enantiomerically pure
with high yield (Ͼ 90%) at 0 °C or room temperature in
various solvents such as toluene or dichloromethane. Prepa-
ration of the [18F]alkylating agent on a solid support was de- Germany, 2004)
(ee Ͼ 95%).
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Introduction
duction. This is undoubtedly the main limitation of this
technique, which nevertheless permits large-scale pro-
duction [Ͼ 100 mCi EOS (end of synthesis)] of 6-[18F]-
fluoro--dopa ready for PET injection with excellent chemi-
cal, radiochemical and enantiomeric purities (Ͼ 97%).
Numerous applications of phase-transfer catalysis (PTC)
for the enantioselective syntheses (92% Ͻ ee Ͻ 99.5%) of a
wide range of α-amino acids have been reported in the lit-
erature in the last few years.[7] These results prompted us to
evaluate the potential of this attractive synthetic approach
for the preparation of nca 6-[18F]fluoro--dopa. This paper
reports our first results with the PTC approach, which
should overcome the main drawbacks of chiral auxiliary
strategies.
Positron emission tomography (PET) is a molecular im-
aging technology that uses radiopharmaceuticals labeled
with positron-emitting radioisotopes.[1] Among these radio-
pharmaceuticals, 6-[18F]fluoro--dopa labeled with fluor-
ine-18 (half-life 109 min), has found wide application as a
tracer for in vivo cerebral studies of presynaptic dopami-
nergic functions in humans.[2] The need for reliable pro-
duction of this radiopharmaceutical has led, in the last two
decades, to reports on various electrophilic and nucleophilic
radiosyntheses of this compound.[3,4] Routine productions
of 6-[18F]fluoro--dopa have been conducted in our labora-
tory for some time, using a nucleophilic, multistep radio-
synthesis approach.[5] This method is based on the enanti-
oselective alkylation of a chiral auxiliary previously de-
scribed by Seebach et al. for the synthesis of amino acids.[6]
This no-carrier-added (nca) regioselective and enantioselec-
tive approach involves a sequence with sensitive reagents.
Consequently, the alkylation reaction, which requires the
use of strong bases such as lithium diisopropylamide under
anhydrous conditions, can be difficult to handle under re-
mote control or to automate in the scope of routine pro-
Results and Discussion
In the early 1990s, the first amino acid synthesis involving
an asymmetric PTC appeared in the literature, although the
reported enantiomeric excesses were quite moderate.[8] In
1997, the Corey and Lygo groups independently described
very efficient PTC catalysts for amino acid synthesis.[9Ϫ11]
More recently, several new chiral phase-transfer catalysts
have also been presented in the literature.[7] Most of these,
affording high enantiomeric excesses (Ͼ 90%), are generally
derived from the quaternary ammonium salt of a cinchona
alkaloid derivative. Among the described catalysts, those
bearing an anthracenylmethyl substituent on the bridge-
[a]
B-30 Cyclotron Research Center,
Liege University, 4000 Belgium
Fax: (internat.) ϩ 32-43662946
E-mail: Christian.Lemaire@ulg.ac.be
[b]
CERMEP,
59 Bd Pinel, 69003 Lyon, France
Eur. J. Org. Chem. 2004, 2899Ϫ2904
DOI: 10.1002/ejoc.200400059
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2899

C. Lemaire, S. Gillet, S. Guillouet, A. Plenevaux, J. Aerts, A. Luxen
FULL PAPER
head nitrogen atom allow high catalytic enantioselective al-
kylations of N-(diphenylmethylene)glycine tert-butyl ester.
Another important feature of this methodology is that al-
most equal and opposite enantiomers can be induced by
switching from cinchonidinium to cinchoninium salts.[11] In
light of this data from the literature, Corey’s catalyst was
selected for the nca synthesis of 6-[18F]fluoro--dopa.[10]
The general synthetic pathway of this new radiochemical
PTC approach involves three major steps (Scheme 1).
Scheme 2. Radiochemical synthesis of the 2-[18F]fluoro-4,5-benzyl
t
halide alkylating agent on a C18 Sep Pak cartridge
In order to be able to perform the halogenation with high
yield (Ͼ 90%), the solid support has to be kept ‘‘wet’’ with
hexane. The aryl iodide was then synthesized by passing
gaseous HI, previously generated as described in the Exp.
Sect., over the support. These conditions afforded the iod-
ide derivative in nearly quantitative yield. This can be reco-
vered and purified by elution with dichloromethane or an-
other solvent through a small, commercially available Sep
Pak alumina column.
Scheme 1. Synthesis of nca 6-[18F]fluoro--dopa by phase-transfer
reaction; reagents and conditions: (a) [K/222] ϩ 18FϪ, DMSO, 110
°C, 45Ϫ50% EOB; (b) NaBH4 aq., room temp., 100%; (c) HX(g),
room temp., Ͼ 90%; (d) 4, (Ph)2CNCH2COOtBu (5), CsOH·H2O,
toluene, 0 °C, Ͼ 90%; (e) HI 57%, 200 °C, 20 min; (f) HPLC
The use of HBr instead of HI allows considerable re-
duction of the reaction time, as the bromination is complete
in less than 20 s. Nevertheless, the silica purification process
previously developed for the iodo derivative was inefficient
The first step implies the preparation of an [18F]fluoro- and the best results were obtained with a homemade
benzyl halide (2) derivative. The second step consists of the K2CO3 column.
alkylation reaction with a chiral catalyst (4) of a commer-
Since the total recovery volume does not exceed
cially available benzophenone imine, and the final step, the 1.8Ϫ2.5 mL, an additional evaporation step is not required
HPLC purification after hydrolysis.
before alkylation. The [18F]fluorobenzyl bromide derivative
Although numerous methods for the radiosynthesis of was obtained from no-carrier-added [18F]fluoride in
[18F]fluorobenzyl halide have been reported,[12,13] in our op- 35Ϫ40% radiochemical yields [EOB (end of bombardment)]
inion none of these methods is well suited for a simple auto- with a radiochemical purity of Ͼ 98% within 40 min.
mation and a reliable operation. Generally based on two or
The following alkylation reaction, using phase-transfer
three reaction steps, they imply time-consuming operations techniques, is especially noteworthy because of its simplicity
such as evaporation, extraction, drying, and heating, mak- in contrast to other procedures that require anhydrous con-
ing these syntheses inconvenient for routine use.
Reactions with reagents on a solid support have been re-
ditions for the deprotonation of a chiral auxiliary.
In practice, the [18F]fluoro-arylated product (3) of 6-
discovered with the development of combinatory chemis- [18F]fluoro--dopa was synthesized in a two-phase mixture
try.[14] This solid-support approach, which has previously containing N-(diphenylmethylene)glycine tert-butyl ester
been applied with success in our laboratory to the alkaline (5), O-allyl-N-(9-anthracenylmethyl)cinchonidinium (4),
synthesis of [18F]FDG (2-deoxy-2-[18F]fluoro--glucose), [18F]benzyl bromide 2, cesium hydroxide and dichlorometh-
should overcome these difficulties and avoid steps that are ane or toluene. Initial enantioselective catalytic phase-trans-
difficult to automate.[15] Thus, after labeling, the fluoro- fer alkylation was conducted by Corey’s group in dichloro-
t
veratraldehyde derivative was trapped on a C18 cartridge methane with 1 equiv. of the Schiff base, 0.1 equiv. of cata-
and quantitatively reduced on this same support with an lyst, 10 equiv. of cesium hydroxide and a large excess of
aqueous solution of NaBH4 (Scheme 2). The resulting alkylating agent (1.5Ϫ5 equiv.). Under these conditions, the
[18F]fluorobenzyl alcohol was then washed with water and reaction is described to proceed with high enantioselectivity
dried on the solid support using a non-polar solvent.
(Ͼ 95%) with various benzyl halides at Ϫ78 °C for 24 h.[9]
2900
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org Eur. J. Org. Chem. 2004, 2899Ϫ2904

Highly Enantioselective Synthesis of No-Carrier-Added 6-[18F]Fluoro-L-dopa
FULL PAPER
Extrapolation of these reaction conditions to radiochem- also led to a light increase of the ee (Entry 13). However,
istry is not feasible. Indeed, a radiochemical synthesis start- at lower temperatures (Ϫ40 and Ϫ78 °C), the radiochemical
ing from 100 mCi of [18F]fluorobenzyl bromide of high yield decreased drastically, resulting in no alkylation reac-
specific activity (10 Ci/µmol) contains only a few microg- tion even after 1 h (Entries 11 and 12).
rams of reagent (2.48 µg). Therefore, in order to obtain the
A reaction carried out using 50% aqueous potassium hy-
best enantioselectivity and radiochemical yield, different re- droxide as base shows that the alkylation reaction can be
action parameters (base, PTC and imine amounts, solvent performed in the presence of large amounts of water (see
and reaction temperature), keeping constant solvent volume Exp. Sect.). However, this last approach was not optimized
(2 mL) and reaction time (10 min), were investigated. In all (ee 86%).
cases, the (S) absolute configuration and the enantiomeric
Taking into account all these results, a higher enantiose-
excess (ee) were established after conversion of the lectivity was obtained when the reaction proceeded at 0 °C
[18F]fluorinated alkylated product to the corresponding in toluene in the presence of 1 equiv. of Schiff base (5; 6.8
[18F]fluorodopa. After preparative HPLC purification, the µmol, 2 mg), 35 equiv. of cesium hydroxide (40 mg, 238
samples were analyzed on an analytical chiral column ac- µmol) and 4.2 equiv. of catalyst (4; 15 mg, 24.8 µmol). In
cording to the method exemplified in the Exp. Sect. Al- this case, we have to keep in mind that the above process
though levels of induction are not affected by stirring rate, cannot be called a catalytic process. The [18F]electrophilic
effective stirring is crucial in order to obtain rapid reac- agent (10 nmol, 2.48 µg) is present, but in a very small
tion.[16]
amount compared to both the Schiff base (1.5 ϫ 10Ϫ3
As can be seen from the results (Table 1), increasing the equiv.) and the quaternary ammonium salt (4 ϫ 10Ϫ4
concentration of catalyst (4) has a positive result on the ee equiv.). However, this is not a problem since the reaction is
(Entries 1 and 2). In both cases, a high conversion (Ͼ 90%) run on a small scale, and the chiral reagent, which is now
of the [18F]fluorobenzyl bromide into the corresponding al- commercially available, is easily prepared in two steps from
kylated product was also observed. From Entries 2Ϫ8, it inexpensive cinchonidine.[9] The other major improvement,
appears that the ee tends to be lower when the concen- compared to the initial procedure using chiral auxiliaries,[5]
tration of the Schiff base increases, although in this instance is the high enantioselectivity observed at 0 °C and even at
the enantioselectivity of Entry 8 is somewhat higher than room temperature. Under the best conditions, more than
for Entry 7. The use of imine in a range of 1Ϫ5 mg affords 97% of the  form of 6-[18F]fluoro--dopa is available at the
similar enantioselectivities (Entries 3Ϫ6).
end of the synthesis. Furthermore, this alkylation reaction
is fast and complete even after 5 min. This nca synthesis
of 6-[18F]fluoro--dopa is a good illustration of the main
differences that can exist between conventional and radio-
chemical chemistry.
Table 1. Enantioselectivity of the reaction with reaction conditions;
n ϭ number of runs
Entry CsOH
5
4
Solvent
T
ee
n
mg (µmol) mg (µmol) mg (µmol)
[°C] [%]
Conclusion
1
2
3
4
5
6
7
8
9
10
11
12
13
38 (226)
42 (250)
39 (232)
40 (238)
39 (232)
39 (232)
39 (232)
42 (250)
40 (238)
40 (238)
39 (232)
41 (244)
41 (245)
10 (32.8)
2 (3.1)
CH2Cl2 20 81
3
3
1
1
3
3
1
1
5
2
1
1
In summary, a systematic study of PTC reaction con-
ditions using various amounts of substrate, catalyst, solvent
at different temperatures has led to a simple, stereoselective
synthesis of 6-[18F]fluoro--dopa. After preparative HPLC
purification, the main radiochemical impurity was the 
isomer of 6-[18F]fluorodopa. When the alkylation reaction
was conducted in toluene at 0 °C, only 2.0Ϫ2.5% of this
isomer was detected. Although this simple method affords
a lower enantiomeric excess (96%) than other published nca
procedures, this general nca [18F]fluorination route is well
adapted for automation and future routine production of
nca 6-[18F]fluoro--dopa. Moreover, with further develop-
ment of the phase-transfer catalysts, it should be possible
in the future to optimize the efficiency of the alkylation
step. Research continues towards extending this process to
the preparation of other amino acids such as 2-fluoro--
tyrosine and to simplify the preparation of the alkylating
agent.
10.2 (34.5) 14.3 (22.4) CH2Cl2 20 88
0.7 (2.4) 16 (25.1) CH2Cl2 20 92
1.2 (4.1) 17.5 (27.4) CH2Cl2 20 92
1.9 (6.4) 15.5 (24.3) CH2Cl2 20 92
5 (16.9) 15.4 (25.4) CH2Cl2 20 92
26.4 (89.4) 14.6 (22.9) CH2Cl2 20 53
34 (115)
1.5 (5.1) 15.3 (24.0) CH2Cl2
1.6 (5.4) 14.7 (23.0) CH2Cl2 Ϫ20 95
1.8 (6.1) 15.5 (24.3) CH2Cl2 Ϫ40 Ϫ
1.9 (6.4) 16.1 (26.6) CH2Cl2 Ϫ78 Ϫ
16.1 (25.2) CH2Cl2 20 57
0 95
1.5 (5.1) 16.3 (25.5) toluene
0 96 Ͼ 10
In a last set of experiments, the temperature effect on
the alkylation enantioselectivity was probed (Entries 9Ϫ12).
Alkylations with the nca electrophilic agent were conducted
at five different temperatures: Ϫ78 °C, Ϫ40 °C, Ϫ20 °C, 0
°C and ambient temperature. From these data, it appears
that reaction temperature has only a slight effect on the
asymmetric induction. A decrease of the temperature from
room temperature to 0 °C and Ϫ20 °C increased the ee
Experimental Section
Starting Materials: The aminopolyether ‘‘Kryptofix 222’’
from 92 to about 95%. Substitution of CH2Cl2 by toluene (4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]hexacosan,
K/
Eur. J. Org. Chem. 2004, 2899Ϫ2904
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2901

C. Lemaire, S. Gillet, S. Guillouet, A. Plenevaux, J. Aerts, A. Luxen
FULL PAPER
222), potassium carbonate, cesium hydroxide monohydrate, and
hydroiodic acid were obtained from Merck. Sodium borohydride
was purchased from Acros. C18, Al2O3, SiO2, QMA Sep Pak car-
0.8 mL/min. The retention time of the  and  enantiomers was 8
and 11.2 min, respectively.
t
Nca [18F]F؊ Production: Nca [18F]fluoride was produced by proton
irradiation (18 MeV, 30 µA, 20 min) of 95% enriched 18O water
in a silver target (2.2 mL internal volume) using an IBA compact
cyclotron (Louvain La Neuve, Belgium) The target contents were
delivered through 75 m of small-bore Teflon tubing (1.6 mm o.d.
ϫ 0.8 mm id) to the chemistry laboratory. The separation of nca
[18F]fluoride from 18O-enriched water was realized on a small
QMA Sep Pak light cartridge previously conditioned with 10 mL
of 0.5  K2CO3 and water (20 mL). Enriched water was recovered
and after distillation used again for further production. The
[18F]FϪ trapped on the column was eluted into a round-bottomed
open vial (2.5 mL, Wheaton) with 400 µL of a solution previously
prepared by mixing of an equal volume of potassium carbonate in
water (35 mg/mL) and kryptand in acetonitrile (110 mg/mL).
tridges were obtained from Waters (Milford, MA, USA). The pro-
chiral reagent tert-butyl glycinate, benzophenone imine and gase-
ous hydrobromic acid were purchased from SigmaϪAldrich. 6-
[19F]Fluoro-,-dopa was purchased from RBI (USA). Solvents
were of HPLC grade from Baker and were filtered before use and,
unless described, used without further purification. The trifluoro-
methanesulfonate salt of 2-dimethylaminoveratraldehyde (1) and
the O-allyl-N-(9-anthracenylmethyl)cinchonidinium bromide cata-
lyst (4) were synthesized according to literature procedures[5,9] and
characterized by standard methods. 18O-enriched water (95%) was
obtained from Rotem Industry (Israel).
General: IR spectra were obtained with a Mattson Genesis series
FTIR spectrometer. Melting points were recorded with a Büchi
melting point apparatus and are corrected. NMR spectroscopy was
carried out with a Bruker 400 NMR spectrometer (400 MHz) using
CDCl3 as a solvent unless noted and tetramethylsilane (TMS) as
internal standard. The chemical shifts are reported in parts per
million (ppm) from TMS. MS were recorded with an Automass
solo from ThermoFinnigan (EI: 70 eV) or with LC MSMS using a
‘‘Triple-Stage quadrupole TSQ 7000’’ from Finnigan with electros-
pray source (ESI, HV 4.5 KV). Column chromatography was per-
formed on 230 mesh silica gel from Merck. Thin layer chromatog-
raphy (TLC) was carried out on Merck silica gel 60 F254 silica
plates and analyzed on a Berthold TLC scanner. TLC elution con-
ditions varied and are specified individually for each compound.
Radioactivity was measured in a dose calibrator and all radio-
chemical yields are decay-corrected unless noted. The synthesis was
performed with a Zymate Laboratory Robot (Zymark Corp),
which was described previously elsewhere.[17]
The [18F]Fluorinating Agent: The water was evaporated under a
stream of nitrogen for 5 min on an aluminum heating block at 120
°C. Acetonitrile was then added (100 µL) and the solution concen-
trated to dryness. The azeotropic evaporation step was repeated
twice (2 ϫ 100 µL) and finally a dry fluorinating agent in the form
of a [18F]potassium Kryptofix complex ([K/222]/18FϪ) was ob-
tained.
Preparation of the Nca 2-[18F]Fluoro-4,5-dimethoxybenzaldehyde: A
solution of 15 mg of the trifluoromethanesulfonate salt of dimeth-
ylaminoveratraldehyde in 0.9 mL of DMSO was added to the pre-
vious residue of [K/222]/18FϪ. The vial was then closed with a screw
cap and heated at 90 °C for 5 min. After labeling, the DMSO solu-
tion was diluted with water (30 mL) and the whole solution passed
t
through a previously activated environmental C18 Sep Pak car-
tridge. The radiochemical yield of the nca [18F]fuoroveratraldehyde
intermediate depends mainly of the quality of the starting triflate
precursor and averaged between 40 and 50% EOB [Rf (CH2Cl2) ϭ
0.46; tR (HPLC) ϭ 2.9 min (System A, CH3CN/water, 80:20)].
High-Performance Liquid Chromatography (HPLC) Separations.
System A: Analytical HPLC was performed with a Waters 600E
pump, a manual Rheodyne injector (200 µL loop), a Waters PDA
and the Millennium Chromatography Manager software from
Waters. The analytical reverse phase Waters Symmetry C18 col-
umn (150 ϫ 3.9 mm, 5 µm particle size), protected with a Waters
SentryԹ guard column, was connected in series to an NaI detector
for radioactivity detection. The elution was performed at a con-
stant flow rate of 1 mL/min with a mixture consisting of CH3CN/
18 Megohms water. System B: The preparative HPLC of 6-
[18F]fluoro--dopa was conducted on an Econosphere C8 column
(250.0 ϫ 9 mm, 10 µm particle size Altech) with a Waters 600
pump, a manual Rheodyne injector (5000 µL loop), a Waters
Lambda Max Model 481 UV spectrophotometer (281 nm). The
mobile phase was aqueous acetic acid (0.1%) with ascorbic acid
(0.01%) and EDTA (1 m) and the flow rate was 5 mL/min. The
radioactive elution profile was monitored with a homemade GM
radioactivity detector.
Preparation of the Nca 2-[18F]Fluoro-4,5-dimethoxybenzyl Alcohol:
For the on-column reduction of the aldehyde to the alcohol, a solu-
tion of NaBH4 in water (0.5 g/10 mL) was slowly passed through
t
the C18 cartridge (1 min). For subsequent quality controls, the 2-
[18F]fluoro-4,5-dimethoxybenzyl alcohol was eluted from the C18
cartridge with 2 mL of CH3CN. TLC and HPLC analyses indicated
that this reduction step was quantitative (95Ϫ100%). As an alterna-
tive to this procedure, reduction was conducted directly in the
DMSO phase. In this case, after labeling, the DMSO solution was
diluted with 20 mL of sodium borohydride solution (1 g/10 mL)
and the whole solution passed through a previously activated en-
t
vironmental C18 Sep Pak cartridge (Waters). This process pro-
ceeded also with quantitative yield [Rf (CH2Cl2) ϭ 0.10; tR
(HPLC) ϭ 2.5 min (System A, CH3CN/water, 80:20)].
Preparation of the Nca 2-[18F]Fluoro-4,5-dimethoxybenzyl Bromide
Alkylating Agent (2, X ؍ Br): After reduction, the Sep Pak column
was washed with 10 mL of highly pure water. The residual water
on the cartridge was then displaced with hexane (3Ϫ5 mL). Hexane
Enantiomeric Excess Determination: The enantiomeric purity of the
radiopharmaceutical was determined after HPLC purification un-
der two different conditions. In both cases, similar enantiomeric
excesses were obtained. System C: A chiral ProϭSi 100 column
t
(500 µL) was kept in the buffer volume above the C18 cartridge.
(Serva, Polylab, Antwerp, Belgium; 4.6 ϫ 250 mm) was eluted at a Halogenation was then performed directly on the cartridge with
flow rate of 1.5 mL/min with a mixture consisting of 50 m gaseous hydrogen bromide during 20 s. An exothermic reaction
KH2PO4 and 1 m  CuSO4 at pH ϭ 4. Under these conditions, took place on the support and hexane was then removed. After
the retention time of the  and  enantiomers was 6 and 9 min,
respectively. System D: A CrownPak CR (ϩ) column (4.0 mm I.D
ϫ 150 mm, 5 µm) from Daicel Chemical Industries (Tokyo, Japan)
was eluted with aqueous HClO4 (pH ϭ 2.0) at a flow rate of
reaction, the Sep Pak cartridge was flushed with nitrogen for 2 min
and the corresponding [18F]fluorobenzyl bromide derivative eluted
with dichloromethane or toluene (3.5 mL). The residual unchanged
HBr was removed by passing the solution through a small potas-
2902
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2899Ϫ2904

Highly Enantioselective Synthesis of No-Carrier-Added 6-[18F]Fluoro-L-dopa
sium carbonate column (5 mm i.d. ϫ 30 mm) [Rf (CH2Cl2) ϭ 0.6; ether (dry, 70 mL) saturated with HBr(g) was stirred at room tem-
FULL PAPER
tR (HPLC) ϭ 3.2 min (System A, CH3CN/water, 80:20)].
perature for 30 min. At this time, as indicated by HPLC, the reac-
tion was quantitative. The reaction mixture was then treated with
a saturated NaHCO3 solution. The ether phase was recovered and
dried with MgSO4. The solvent was then removed under reduced
pressure to give 2 (2.4 g, 90%) as a white solid which decomposes
slowly. M.p. 59Ϫ61 °C. IR (KBr): ν˜ ϭ 2925, 2854, 1515, 1131, 1108
Preparation of the Nca 2-[18F]Fluoro-4,5-dimethoxybenzyl Iodide
Alkylating Agent (2, X ؍ I): Halogenation was also performed on
the cartridge with gaseous hydrogen iodide generated by heating
liquid HI at 150 °C under a flow of nitrogen. In this case the reac-
tion was slower (10 min). After iodination, the support was flushed
with nitrogen for 2 min and the corresponding [18F]fluorobenzyl
iodide derivative eluted with dichloromethane or toluene (4 mL).
The residual unchanged HI and other impurities were removed by
passing the mixture through an SiO2 Sep Pak column. The yield
and radiochemical purity of the product were determined by radio
TLC and HPLC analyses [Rf (CH2Cl2) ϭ 0.6; tR (HPLC) ϭ 3.2 min
(System A, CH3CN/water, 80:20)].
1
cmϪ1. H NMR(400 MHz, CDCl3): δ ϭ 3.85 (s, 3 H, OCH3), 3.86
3
(s, 3 H, OCH3), 4.51 (s, 2 H, CH2OH), 6.61 (d, JH,F ϭ 10.92 Hz,
4
1 H, ArH), 6.81 (d, JH,F ϭ 7.0 Hz, 1 H, ArH) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 26.5 (CH2Br), 56.2 (OCH3), 56.4 (OCH3),
100.1 (d, 2JC,F ϭ 27.5 Hz, Ar-C), 112.7 (d, 3JC,F ϭ 4.85 Hz, Ar-C),
115.4 (d, 2JC,F ϭ 16.2 Hz, Ar-C), 145.4 (Ar-C), 145.4 (Ar-C), 150.4
(Ar-C), 154.9 (d, JC,F ϭ 241 Hz, ArϪF) ppm. EI-MS: m/z (%) ϭ
248 (18) [Mϩ, 79Br], 169 (100) [M Ϫ 79Br]ϩ.
General Reaction Procedure for the [18F]Alkylation of the Schiff
Base 5. Method I: Alkylation was performed under agitation in a
conical flask (10 mL, Wheaton) containing 2 mg (6.8 µmol) of the
Schiff base 5, 15 mg (24.8 µmol) of the O-allyl-N-(9-anthracenylme-
thyl)cinchonidinium bromide catalyst (4) and 40 mg (238 µmol) of
cesium hydroxide monohydrate as base. To this vial, cooled to 0
°C, was added the nca fluorobenzyl halide 2 in toluene (2 mL). The
alkylation reaction was performed at this temperature for 10 min
and the organic mixture quenched at 0 °C with 250 µL of HI (57%
w/v) previously distilled from red phosphorus. In order to evaluate
the yield of the alkylation step, an aliquot was assayed before HI
addition and directly analyzed both by TLC and HPLC [Rf
(CH2Cl2) ϭ 0.17; tR (HPLC) ϭ 6.5 min (System A, CH3CN/water,
80:20)]. The experiments were also conducted at different tempera-
tures with various amounts of base, imine and catalyst (see Results
and Discussion section). Method II: A liquid/liquid PTC was also
realized with 50% aqueous potassium hydroxide as base. In this
case, 1 mL of this solution was mixed with 1 mL of toluene con-
taining 2 mg (6.8 µmol) of the prochiral Schiff base, 10 mg (16.5
µmol) of the O-allyl-N-(9-anthracenylmethyl)cinchonidinium bro-
mide catalyst and the nca 2. The reaction was conducted at 0 °C
or room temperature for 10 min [RY (radiochemical yield): 85%].
Synthesis of the Alkylated Schiff Base 3: A solution of 2-fluoro-4,5-
dimethoxybenzyl bromide (2; 624 mg, 3.4 mmol) in CH2Cl2 (2 mL)
was added to a mixture of tert-butyl N-(diphenylmethylene)glycin-
ate (5; 1.0 g, 3.4 mmol), CsOH·H2O (2.8 g, 17 mmol), and the cata-
lyst Bu4NϩBrϪ (11 mg, 0.034 mmol) in CH2Cl2 (5 mL). The reac-
tion mixture was then stirred vigorously at room temperature. After
30 min, the alkylation was quantitative as demonstrated by analyti-
cal HPLC (System A). Water was added (20 mL), the two phases
were separated and the water solution extracted twice with diethyl
ether (2 ϫ 10 mL). The combined ether phases were dried with
magnesium sulfate, then concentrated under reduced pressure to
give the crude product 3 as a yellow oil (1.34 g, 2.89 mmol, 85%).
Purification of the crude product by column chromatography on
silica gel afforded the alkylation product 3 (hexanes/CH2Cl2/ET3N,
70:30:3) as a yellow oil (1.1 g, 70%). TLC: Rf (silica gel) ϭ 0.38
(hexanes/CH2Cl2/Et3N, 70:30:3). IR (neat, NaCl): ν˜ ϭ 2963, 2932,
2870, 1732, 1514, 1447, 1286, 1227, 1191, 1152, 700 cmϪ1 1H
.
NMR(400 MHz, CDCl3): δ ϭ 1.43 [s, 9 H, C(CH3)3], 3.09 (dd, J ϭ
9.6, 13.2 Hz, 1 H, H-3Ј), 3.22 (dd, J ϭ 3.6, 13.6 Hz, 1 H, H-3), 3.6
(s, 3 H, OCH3), 3.78 (s, 3 H, OCH3), 4.14 (dd, J ϭ 4.0, 9.3 Hz, 1
H, H-2), 6.49 (d, J ϭ 7 Hz, 1 H, ArH), 6.58 (br. d, J ϭ 10.9 Hz, 1
H, ArH),), 6.67 (d, J ϭ 6.6 Hz, 2 H, ArH), 7.37Ϫ7.24 (m, 6 H,
ArH), 7.57 (d, J ϭ 7.3 Hz, 2 H, ArH) ppm. 13C NMR (CDCl3/
100 MHz): δ ϭ 27.9 [C(CH3)3], 31.8 (CH2), 55.50 (OCH3), 55.53
(OCH3), 66.0 (CH), 80.5 [C(CH3)3], 99.2 (d,2JC,F ϭ 27.5 Hz, Ar-
C), 113.8 (d,3JC,F ϭ 4.85 Hz, Ar-C), 115.1(d,2JC,F ϭ 17.8 Hz, Ar-
C), 126.6, 127.4, 127.6, 128.2, 129.7, 135.7, 139.0, 144.2 (10 Ar-C),
6-[18F]Fluoro-
-dopa: Toluene was evaporated at 130 °C and ad-
l
ditional HI added (750 µL). The vial was closed and the solution
refluxed (200 °C) for 15 min. Following hydrolysis, the solution was
cooled and diluted with 1.5 mL of HPLC eluent. The 6-[18F]fluoro-
-dopa was finally purified by preparative HPLC (Condition B).
Under these conditions 6-[18F]FDOPA was eluted between 8 and
9 min.
1
147.8 (3JC,F ϭ 9.7 Hz, Ar-C), 154.8 (d, JC,F ϭ 237.8 Hz, Ar-C),
169.8 (CϭN), 170.0 (CO) ppm. ES-MS (ESI): m/z ϭ 464.3 [M ϩ
2-[19F]Fluoro-4,5-dimethoxybenzyl Alcohol: A solution of NaBH4
(2 g, 52.9 mmol) in water (20 mL) was slowly added to a suspension
of 6-fluoroveratraldehyde (5 g, 27.2 mmol) in ethanol (200 mL).
The reaction mixture was stirred at room temperature for 15 min
and then diluted with water (500 mL). The water/ethanol mixture
was extracted twice with EtOAc. The extracts were dried with
MgSO4, filtered and the solvents evaporated to dryness. Crystalli-
zation from diethyl ether afforded the benzylic alcohol as white
crystals (4.8 g, 95%). M.p. 55Ϫ56 °C. IR (KBr): ν˜ ϭ 3292, 1625,
H]ϩ.
Acknowledgments
This work, the preliminary results of which were previously pre-
sented as an oral communication at the Thirteenth International
Symposium on Radiopharmaceutical Chemistry, St Louis, USA,
June 27ϪJuly 1, 1999, was supported by the ‘‘Fonds National de
la Recherche Scientifique’’ (grants 3.4526.96 and 3.4513.01), the
1519, 1234, 1196, 1107 cmϪ1
.
1H NMR (400 MHz, CDCl3): δ ϭ
´
Fonds speciaux ULG and the INSERM/CFB exchange. A. P. is a
3.84 (s, 3 H, OCH3), 3.85 (s, 3 H, OCH3), 4.49 (s, 2 H, CH2OH),
research associate of FNRS Belgium.
3
4
6.59 (d, JH,F ϭ 10.96 Hz, 1 H, ArH), 6.79 (d, JH,F ϭ 7.3 Hz, 1
H, ArH) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 57.1 (OCH3),
[1]
2
E. S. Garnett, G. Firnau, C. Nahmias, Nature 1983, 305,
57.3 (OCH3), 59.8 (CH2), 100.9 (d, JC,F ϭ 27.5 Hz ArϪC), 112.6
3
2
137Ϫ138.
(d, JC,F ϭ 4.9 Hz, Ar-C), 119.2 (d, JC,F ϭ 16.2 Hz, Ar-C), 146.3
(Ar-C), 150.3 (d, 3JC,F ϭ 9.7 Hz, Ar-C), 156.7 (d, JC,F ϭ 237.8 Hz,
ArϪF) ppm.
[2]
M. E. Phelps, Annu. Rev. Nucl. Part. Sci. 2002, 52, 303Ϫ338.
A. Luxen, M. Guillaume, W. P. Melega, V. W. Pike, O. Solin,
[3]
R. Wagner, Nucl. Med. Biol. 1992, 19, 149Ϫ158.
E. F. de Vries, G. Luurtsema, M. Brüssermann, P. H. Elsinga,
2-[19F]Fluoro-4,5-dimethoxybenzyl Bromide (2): A solution of 2-
[4]
[19F]fluoro-4,5-dimethoxybenzyl alcohol (2 g, 10.7 mmol) in diethyl
W. Vaalburg, Appl. Radiat. Isot. 1999, 51, 389Ϫ394.
Eur. J. Org. Chem. 2004, 2899Ϫ2904
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2903

C. Lemaire, S. Gillet, S. Guillouet, A. Plenevaux, J. Aerts, A. Luxen
FULL PAPER
[5]
[13]
C. Lemaire, P. Damhaut, A. Plenevaux, D. Comar, J. Nucl.
R. Iwata, C. Pascali, A. Bogni, G. Horvath, Z. Kovaacs, K.
Med. 1994, 35, 1996Ϫ2002.
D. Seebach, E. Dziadulewicz, L. Behrendt, S. Cantoreggi, R.
Yanai, T. Ido, Appl. Radiat. Isot. 2000, 52, 87Ϫ92.
S. V. Ley, I. R. Baxendale, R. N. Bream, P. S. Jackson, A. G.
[6]
[14]
Fitzi, Liebigs Ann. Chem. 1989, 12, 1215Ϫ1232.
Leach, D. A. Longbottom, M. Nesi, J. S. Scott, R. I. Storer, S.
J. Taylor, J. Chem. Soc., Perkin Trans. 1 2000, 3815Ϫ4195.
C. Lemaire, P. Damhaut, B. Lauricella, C. Mosdzianowski, J.-
[7]
K. Maruoka, T. Ooi, Chem. Rev. 2003, 103, 3013Ϫ3028.
[8]
[15]
M. J. OЈDonnell, W. D. Bennett, S. Wu, J. Am. Chem. Soc.
1989, 111, 2353Ϫ2355.
L. Morelle, M. Monclus, J. Van Naemen, E. Mulleneers, J.
Aerts, A. Plenevaux, C. Brihaye, A. Luxen, J. Labelled Compd.
Rad. 2002, 45, 435Ϫ447.
[9]
E. J. Corey, F. Xu, M. C. Noe, J. Am. Chem. Soc. 1997, 119,
12414Ϫ12415.
[10]
[16]
E. J. Corey, M. C. Noe, F. Xu, Tetrahedron Lett. 1998, 39,
5347Ϫ5350.
B. Lygo, P. G. Wainnwright, Tetrahedron Lett. 1997, 38,
8595Ϫ8598.
M. J. O’Donnell, F. Delgado, C. Hostettler, R. Schwesinger,
Tetrahedron Lett. 1998, 39, 8775Ϫ8778.
C. Brihaye, C. Lemaire, Ph. Damhaut, A. Plenevaux, D.
[11]
[17]
Comar, J. Labelled Compd. Rad. 1994, 35, 160Ϫ162.
[12]
C. Lemaire, P. Damhaut, A. Plenevaux, R. Cantineau, L. Chri-
Received January 30, 2004
stiaens, M. Guillaume, Appl. Radiat. Isot. 1992, 43, 485Ϫ494.
2904
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2899Ϫ2904