Use of 1,8-bis-(dimethylamino)-naphthalene/H18F complex as new radiofluorinating agent

Article abstract of DOI:10.1002/jlcr.823

Radiopharmaceuticals containing an ¹⁸F label are of increasing interest due to their utilization in PET imaging. However, the bottleneck in these applications is the limited methods available for introduction of this radionuclide into biologica

Full text of DOI:10.1002/jlcr.823

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS
J Label Compd Radiopharm 2004; 47: 373–383.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.823
Research Article
Use of 1,8-bis-(dimethylamino)-naphthalene/H18F
complex as new radiofluorinating agent
Giancarlo Pascali1,2,*, Dale O. Kiesewetter3, Piero A. Salvadori4 and William
C. Eckelman3
1 National Institutes of Health, Bdg. 10 Room 1C401, 10, Center Dr. MSC 1180,
Bethesda, MD 20892, USA
2 ISUFI-University of Lecce, via per Monteroni, 73100, Lecce, Italy
3 PET & Cyclotron Department, Clinical Center, NIH, USA
4 Radiofarmacia e Ciclotrone Department, IFC, CNR, Italy
Summary
Radiopharmaceuticals containing an 18F label are of increasing interest due to their
utilization in PET imaging. However, the bottleneck in these applications is the
limited methods available for introduction of this radionuclide into biologically
interesting molecules. In this work, we have evaluated a new radiofluorination method
based on the properties of the complex between 1,8-(dimethylamino)-naphthalene
(PS) and [18F]–HF. The results obtained on various model substrates suggest that, in
some limited cases, this new procedure can be regarded as a possible alternative to the
traditional nucleophilic route using K222/K2CO3 in CH3CN. Copyright # 2004 John
Wiley & Sons, Ltd.
Key Words: proton sponge; nucleophilic fluorination; [18F]fluoride; PET
Introduction
Positron emission tomography (PET) has earned increasing interest in the
biomedical field due to its ability to provide high-resolution molecular imaging of
various organs in vivo. The methodology introduces a b+-emitting radionuclide
into a molecule of known pharmacological and metabolic fates and quantita-
tively detects such radiolabelled molecule in vivo with temporal information.
A large number of b+-emitting nuclides are known, but only a few are used
for this purpose; 11C (t1=2 ¼ 20:4 min), 15O (t1=2 ¼ 2 min), 13N (t1=2 ¼ 10 min)
and 18F (t1=2 ¼ 109:7 min) are widely employed radionuclides because they offer
mimetic or quasi-mimetic labelling. If a longer synthetic pathway or longer PET
*Correspondence to: G. Pascali, National Institutes of Health, Bdg. 10 Room 1C401, 10, Center Dr. MSC
1180, Bethesda, MD 20892, USA; E-mail: gp114b@nih.gov
Copyright # 2004 John Wiley & Sons, Ltd.

374
G. PASCALI ET AL.
F
H
(79%)
NMe
NMe
Me
N
NMe
H
H
F
Cl
HF/Et2O, Et2O
r.t.
PS-HF, CH3CN
r.t. overnight
PS-HCl
PhCOCl
PhCOF
(100%)
PS
PS-HF
I
Cl
F
4 PS
3 [Et3N(HF)3]
N
100˚C, 144 hrs
N
Ph
N
Ph
N
II
Figure 1. Previous reactions involving the use of PS/HF (I: Chambers, II:
Darabantu)
acquisition time is required, 18F is the nuclide of choice because of its longer
half-life. A large and increasing variety of radiopharmaceuticals containing 18F
have been found to be useful in PET imaging; this leads to the increasing
importance of the development of new introductory routes for this nuclide in
interesting molecular contexts. 18F can be produced by a biomedical cyclotron as
carrier-added [18F]–F2 or as no-carrier-added [18F]–HFaq. [18F]–F2, or reagents
directly derived from it, can effect labelling reactions by electrophilic substitution
(SE) whereas, nucleophilic substitution (SN) employs the complex K222/K18F
(derived from [18F]–HFaq) in polar organic solvents (generally acetonitrile or
DMSO). This nucleophilic reaction is preferable for many reasons such as higher
specific activity and ease of handling. In this work, we have studied the use of the
complex between 1,8-bis(dimethylamino)-naphthalene (Proton Sponge, (PS))
and H18F as a new nucleophilic radiofluorinating agent.
The use of the non-radioactive PS/HF complex has been previously
reported by Chambers1 and Darabantu2 as a useful mild fluorinating agent
capable of aromatic nucleophilic substitution (SNAr) of chlorine atoms bound
to di- and tri-azine moieties. The PS/HF complex could be synthesized
separately or in situ using either ethereal solution of HF (Chambers) or the
complex Et3N ꢀ (3HF) (Darabantu) (Figure 1). The yields reported for the
fluorinated products were generally high for most of the substrates, but
required long (c2 h) reaction times, not suitable for a radiosynthesis.
We report the in situ synthesis of [18F]–PS/HF from an aqueous solutions of
H18F and the resulting reactions with a variety of substrates.
Results and discussion
Non-radioactive test reactions
In order to test the feasibility of these fluorinating methods, we have first studied
the reactivity of the PS/HF complex, produced in situ from a 48% HFaq
solution, towards various halogenated aromatics and heteroaromatic substrates.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

PS/H18 F IN RADIOFLUORINATIONS
375
Table 1. GC-MS yields of non-radioactive test reactions
Substrate
Thermal 2 h
Thermal 20 h
MW
1
2
3
4
2-chloropyrimidine
1.4
4
13
3.5
22.1
33.7
3.5
5.4
8.4
Trace
3.4
2-chloro-3-nitropyridine
5-chloro-1-phenyl-1H-tetrazole
4-chloro-3-nitrotoluene
3.5
In the first set of experiments, a 0.5 mmol scale was used and all the reagents
were added in equimolar amounts in a 1 ml V-vial. The excess water was
evaporated under Ar by azeotropic distillation. CH3CN was added and the
solutions stirred and heated at 1008C while samples, taken at 2 and 20 h, were
analyzed by GC–MS.
Ionic liquids, a class of chemical salts in which either the cation or anion is an
organic charged molecule, have interesting physical properties such as high
polarity, high conductivity and high boiling point.3 These compounds have been
used as co-solvents for fluoride substitution reactions on some aliphatic substrates.4
We have tested a 2:1 solution of 1-pentyl-3-methylimidazolinium bromide
(PMIMBr) and CH3CN as solvent in a microwave (MW) procedure employing
0.2 mmol of reagent, HFaq and PS in an open 7 ml glass tube. Pulsed MW
irradiation (10 cycles of {5 s ON, 15 s OFF}) was employed to avoid overheating.
Fluorinated products (calculated on the basis of the GC-MS peaks of the
starting substrates) were only detected for four compounds (Table 1) that are
activated to SNAr. No fluorination was observed using aromatic chlorides not
activated to SNAr, as well as various aromatic bromides and iodides.
Solubility of [18F]–PS/HF
In the previous literature, the PS/HF complex was reported as an acetonitrile
soluble fluoride source.1,2 We have used radioactive 18F to assess the claimed
solubility of this fluorinating system. For this purpose, irradiated water
(100 ml) was added to various quantities of PS. After azeotropic distillation,
the vial was washed with 0.6 ml of CH3CN. The recovered acetonitrile solution
was assayed for radioactivity content and compared to the total radioactivity
present in the vial. These data showed that the radioactivity recovered by this
procedure was related to the quantity of PS in a linear fashion, requiring
approximately 80 mg of PS in order to recover 50% of the 18F in solution.
These results show that the complex PS/HF is not highly soluble in
acetonitrile, at least at tracer levels.
Nucleophilic incorporation methods
After some initial experiments to settle issues such as the amount of substrate,
amount of PS and length of reaction, we selected three compounds (2, 3, and
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

376
G. PASCALI ET AL.
2,3,4,5,6-pentamethylbenzyl 4-fluoro-3-nitrobenzoate [6]) to conduct compar-
ison radiofluoride incorporation reactions. The evaluated parameters were:
heating method (either MW or thermal), proportion of water, and comparison
of PS/HF versus K222/K2CO3. In all procedures using PS, irradiated water was
added to the melted base at 1008C. Substrates were added in 50 ml of solvent.
Procedures A1 and A2 employed PS (8 mg) with substrate (3 mg) in 50 ml of
solvent under thermal heating (1008C, 30 min). A1 used only 5 ml of irradiated
water and no drying procedure was done prior addition of the substrate in
CH3CN. A2 used 100 ml of irradiated water, which was azeotropically
evaporated under an Ar stream prior addition of the substrate dissolved in
10% water in CH3CN. These two methods would evaluate the effects of quantity
of irradiated water used and drying process on radiochemical incorporation.
Procedure B1 was designed to evaluate the effect of ionic liquids on the
radiochemical yield (RCY). Procedure A2 was followed, except that the
substrate was added in 50 ml of a 1:1 solution of CH3CN and 1-ethyl-3-
methylimidazolium triflate (EMIMTfO). Procedure B2, designed to evaluate
MW heating, followed the procedure B1 except that heating was conducted by
MW (200 W, 10 cycles of {5 s ON, 15 s OFF}).
Procedure C provided the comparison to the established K222/K2CO3
method. The vial was charged with K222 (6 mmol) and K2CO3 (3 mmol),
irradiated water was added, and CH3CN-assisted azeotropic drying was
conducted. Substrate (3 mg) was added in acetonitrile (50 ml) and the reaction
heated thermally (1008C, 30 min). Procedure D shows the effect of having no
nucleophilic assistance. Irradiated water was placed into a vial, evaporated
with the assistance of CH3CN, and substrate was added in CH3CN. After
thermal heating (1008C, 30 min), the solution was assayed for RCY.
The radiochemical yields obtained for all procedures are shown in Table 2.
For these three substrates, A1 provides the best decay corrected RCY and
does not require azeotropic distillation, but has the drawback of allowing
the use of only 5 ml of radioactive water and, hence, limited radioactivity.
Pre-concentration of aqueous 18F, as in A2, allowed use of more radioactivity
and provided similar RCY for 2 and 6. Interestingly, the addition of a 10%
of water in the reaction solvent (CH3CN) is optimal in these methods: if water
is not added to the vessel, no substitution reaction occurs on 2, 3 and 6 (data
not shown).
Recently, Kim and co-workers5 have shown the successful use of ionic
liquids in radiofluorination reactions. In procedures B1 and B2, we
investigated the effect of using an ionic liquid, ethyl-1-methyl-imidazolium
triflate (EMIMTfO), as co-solvent with CH3CN, using both a thermal and a
microwave heating method. As it can be seen from Table 2, the addition of
EMIMTfO lowered the yields with all the substrates studied, especially when
using a microwave irradiation.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

PS/H18 F IN RADIOFLUORINATIONS
377
Table 2. Substrates, yields and comparison of radiofluorination procedures
Substrates
Products
Procedures (RCY% ꢁ STD)a
A2 B1 B2
A1
C
D
NO
Cl
NO
2
2
30ꢁ8 30ꢁ7 18ꢁ5 10ꢁ4 8ꢁ2
0
2
18F-2
18
N
N
F
N
N
N
N
N
Ph
N
Ph
N
N
71ꢁ7 35ꢁ8 22ꢁ5
0
0
0
18F-3
18
Cl
F
3
O
O
88ꢁ10 70ꢁ7 65ꢁ9 28ꢁ7 80ꢁ9 5ꢁ4
F
NO
18F-6
6
a Yields are HPLC-recovered and based on the total 18F activity; all the numbers are averaged on a
minimum of three experiments.
In order to compare the data obtained from these procedures involving
the use of PS, we did some experiments using a traditional K222/K2CO3
in CH3CN6 approach (C) and one without adding any activating agent to
the irradiated water (D). The yields reported show that the PS method
can give comparable or better results than the Kryptofix method with the
three reported substrates: the difference is particularly large in the case of
2 and 3. Moreover, it seems that these new procedures are very mild since
the only radioactive peaks present in the radio-chromatograms (not repor-
ted) are the one of the product and the one for unreacted fluoride, while
with the traditional method, a number of other radioactive peaks can be
detected.
We have applied these methods to other model compounds (Table 3).
Aromatic NO2 and NMe+3 leaving groups are not displaceable with A1
(substrate 8–10). Aliphatic tosylates are also poor substrates but alkyl iodide is
somewhat better (cf. substrate 7–11).
Experimental
General
Chemistry: Ionic liquid EMIMTfO was purchased from Fluka Chemicals and
K222 from EM Science. Ionic liquid PMIMBr was synthetized by a literature
procedure.7 Compounds 6,10 7, F-7,8 89 and 1010 were synthetized according to
literature methods. All other chemicals (solvents, reagents unless indicated)
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

378
G. PASCALI ET AL.
Table 3. Substrates and yields in procedures utilizing PS compared to K222/K2CO3/
CH3CN method
PSa
K222/K2CO3
a
Substrates
Products
O
O
A1, 2
72b
N
N
18
F
OTs
O
O
18F-7
7
O
O
O
O
A1, 0
C, 62
18
N
F
TfO
18F-8
8
O
N
NO
18F
NO
2
2
2
A1, 0
A1, 0
C, 19
C, 0
18F-9
9
18
F
Br
Me
N
Br
3
18F-10
OTf
10
18
O
F
O
I
8
8
A2, 10
C, 27
11
18F-11
a The yields are HPLC-recovered and are based on the total 18F activity.
b Optimized literature data8.
were purchased from Aldrich (analytical grade when available) and were used
without purification unless otherwise stated. PS was generally used without
purification; in case of visible oxidation (brown color), it can be recrystallized
from cold pentane. Analytical TLC was run on pre-coated Silica Gel 60 F245
plates (Merck) and the results read under UV light (l ¼ 254 nm). 1H and 13C-
NMR were run on a Varian Gemini 2000 using TMS as internal standard and
200 MHz for 1H and 50 MHz for 13C. GC-MS analysis were conducted using a
Thermo Finnegan Trace-GC/Trace DSQ system equipped with a DB5 Restek
column (15 m ꢂ 0.25 mm; coating 0.25 mm thin) and a EI ion source (2008C).
Radiochemistry: 18F aqueous solutions were prepared by a 18O(p,n)18F
reaction in a GE PETrace cyclotron using a 1.8 ml target of 95% enriched
water irradiated by a 14.1 MeV beam at 20–25 mA for 60–90 min.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

PS/H18 F IN RADIOFLUORINATIONS
379
The HPLC system consisted of 2 Gilson pumps (Model 305 and 306), a
Gilson high-pressure mixer, a Rheodyne injector and an analytical C-18
column (Phenomenex Luna, 250 ꢂ 4.6 mm, 5 mm); chromatograms were
obtained using a Gilson Model 151 UV–VIS detector and a Bioscan
radioactivity detector. All the analysis were done using varying composition
H2O/CH3CN eluent systems and the radioactive fractions were collected and
assayed for radioactivity in a Capintec Dose calibrator.
Chemistry
PS-HF fluorination test reactions, thermal: 0.5 mmol of PS (107 mg), 48%
HFaq (18.5 ml) and 0.5 mmol of substrates were added in a 1 ml V-vial and
azeotropically dried with 200 ml of CH3CN (1008C, Ar stream). The vial was
capped after addition of 50 ml of CH3CN and heated (1008C, heating block)
without stirring for 20 h. Samples were taken at 2 and 20 h and analyzed by
GC-MS.
PS-HF fluorination test reactions, microwave: 0.2 mmol of PS (42.8 mg),
48% HFaq (7.4 ml) and 0.2 mmol of substrates were added in a 7 ml test tube
together with 100 ml of PMIMBr and 50 ml of CH3CN. Microwave irradiation
was conducted using a pulsed heating at 200 W: 10 cycles of {5 s ON, 15 s
OFF}. The final mixture was extracted with 1 ml of Et2O and analyzed by
GC-MS.
Synthesis of 2-fluoro-3-nitropyridine F-2: 2-chloro-3-nitropyridine (0.635 g,
4 mmol), PS (1.822 g, 8 mmol) and Et3N ꢀ 3HF (3.9 ml, 24 mmol) were
placed in a 5 ml V-vial and heated at 908C under vigorous stirring for
4 days. After cooling, the residue was extracted with 10 ml of CH2Cl2 and
the organic phase washed with 3 portions of water. The crude product (65%
of fluoro-product and 35% of unreacted chloro-compound by GC-MS)
was purified by flash chromatography (AcOEt/Hexane=2/1+5% MeOH;
Rf ¼ 0:24) affording 0.340 g of yellow oil (60% yield). GC-MS: >99% purity;
m/z: 142, 112, 96, 76, 69. 1H-NMR (CDCl3): (d ppm, J Hz) 8.65–8.60 (2H, m),
7.57 (1H, dd, 8.0, 4.8); 13C-NMR (CDCl3): 155.6 (d, J ¼ 248 Hz), 153.3, 153.0,
137.3, 122.8.
Synthesis of 5-fluoro-1-phenyl-1H-tetrazole F-3: 5-chloro-1-phenyl-1H-
tetrazole (0.541 g, 3 mmol), PS (0.642 g, 3 mmol) and Et3N ꢀ 3HF (1.5 ml,
9.3 mmol) were placed in a 5 ml V-vial and heated at 908C under vigorous
stirring for 5 days. After cooling, the reaction vessel was washed with 10 ml
of CH3CN; the solid remained was discarded and the solution (>98% of
fluoro-product by GC-MS) evaporated and purified by flash chromato-
graphy (CHCl3, Rf ¼ 0:6) affording 0.364 g of white crystals (74% yield).
1
GC-MS: >99% purity; m=z: 164, 136, 116, 109, 85, 64. H-NMR (CDCl3):
7.69–7.58 (5H, m); 13C-NMR (CDCl3): 137.5 (d, J ¼ 290 Hz), 131.9, 130.4,
124.7, 121.9.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

380
G. PASCALI ET AL.
Synthesis of 1,8-ditosyloctane 12: In a 150 ml round bottom flask, 1,8-
octanediol (6.66 g, 45.6 mmol) was dissolved in 50 ml of pyridine; the solution
was cooled to 08C and then tosyl chloride (26.8 g, 140.6 mmol) was added. The
mixture was stirred at 08C for 30 min and then stored in a refrigerator
overnight. The solution was then poured into ice water and the white solid
precipitate filtered and dried affording 13.6 g of pure product (66% yield based
1
on the diol). MS (FAB+) m=z: 455. H-NMR (CDCl3): (d ppm, J Hz) 7.79
(4H, d, 8.2), 7.35 (4H, d, 8.2), 4.01 (4H, t, 6.8), 2.46 (6H, s), 1.61 (6H, b), 1.21
(6H, bm); 13C-NMR (CDCl3): 144.8, 133.2, 130.0, 128.0, 70.7, 28.8, 28.7,
25.2, 21.7.
Synthesis of 2-(8-tosyloctyloxy)-naphthalene 13: In a 10 ml round flask,
b-naphthol (0.130 g, 0.9 mmol) was dissolved in 4 ml of CH3CN and 1 ml of
1.52 M Bu4NOH was added as base; compound 12 (0.520 g, 1.52 mmol),
dissolved in 4 ml of CH3CN, was added to the solution in five portions. The
solution was stirred overnight at room temperature. A white precipitate
formed that was identified as the undesired product resulting from
dimerization of two naphthol unit through the octyl bridge. Hence, the
solution was filtered, the filtrated concentrated and purified using flash
chromatography (75/20/5=Hexane/CHCl3/EtOAc, Rf ¼ 0:43) obtaining
0.163 g (0.38 mmol) of a yellow oil (42% yield). This compound was
crystallized from cold CH3CN containing a few drops of water yielding pale
yellow crystals. MS (FAB+) m=z: 427. 1H-NMR (CDCl3): (d ppm, J Hz) 7.76
(5H, m), 7.33 (4H, m), 7.13 (2H, m), 4.04 (4H, dd, 6.4, 12.8), 2.44 (3H, s), 1.82
(2H, m), 1.64 (2H, m), 1.30 (8H, b); 13C-NMR (CDCl3): 157.2, 144.8, 134.8,
133.5, 130.0, 129.5, 129.1, 128.1, 127.8, 126.9, 126.5, 123.7, 119.2, 106.8, 70.8,
68.1, 29.4, 29.2, 26.2, 25.5, 21.8.
Synthesis of 2-(8-iodooctyloxy)-naphthalene 11: Compound 13 (0.148 g,
0.35 mmol) was placed in a 25 ml round bottom reaction flask and dissolved in
10 ml of acetone; NaI (0.065 g, 0.43 mmol) was added. The solution was heated
under gentle reflux with magnetic stirring overnight. The solvent was
evaporated and the residue partitioned between CH2Cl2/H2O; the evaporation
of the organic solvent gives 0.041 g (31% yield) of 11. GC-MS: 93% purity;
1
m=z: 382, 157, 144, 127, 115, 69. H-NMR (CDCl3): (d ppm, J Hz) 7.73 (3H,
m), 7.35 (2H, m), 7.16 (2H, m), 4.07 (2H, t, 6.4), 3.19 (2H, t, 6.8), 1.84 (4H, m),
1.38 (8H, b); 13C-NMR (CDCl3): 157.3, 134.8, 129.5, 129.1, 127.8, 126.9,
126.5, 123.7, 119.2, 106.8, 68.1, 33.7, 30.6, 29.4, 28.7, 26.24.
Synthesis of 2-(8-fluorooctyloxy)-naphthalene F-11: Compound 13 (0.163 g,
0.38 mmol) was placed in a 25 ml round bottom reaction flask and dissolved in
5 ml of acetonitrile; Me4NHF2 (0.170 g, 1.5 mmol) was added. The solution
was heated to reflux under magnetic stirring for 2 h. The solvent was
evaporated and the residue partitioned between CH2Cl2/H2O. Evaporation of
the organic solvent gives 0.057 g (55% yield) of F-11. GC-MS: >99% purity;
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

PS/H18 F IN RADIOFLUORINATIONS
1
381
m=z: 274, 254, 157, 144, 127, 115, 69 H-NMR (CDCl3): (d ppm, J Hz) 7.81
(3H, m), 7.36 (3H, m), 7.15 (1H, m), 4.42 (2H, dt, 6.4, 47.8), 4.07 (2H, t, 6.4),
1.75 (3H, m), 1.35 (9H, b); 13C-NMR (CDCl3): 156.5, 134.3, 129.2, 128.3,
127.4, 126.6, 126.3, 123.4, 118.7, 106.6, 83.7 (d, J ¼ 160 Hz), 67.4, 29.9, 29.6,
28.6, 25.4, 24.7, 24.6
Radiochemistry
Solubility test (general procedure): Various quantities of PS (from 0.5 to
80 mg) were melted at 1008C in a V-vial; 100 ml of [18F]–Faꢃq were added and
azeotropic distillation was conducted with five portions of 100 ml of CH3CN
(1008C, Ar stream). 50 ml of CH3CN were added and the vessel cooled down;
the content of the vial and a 600 ml CH3CN rinse were transferred, assayed for
radioactivity and compared to the total radioactivity.
Radiolabelling procedures
Procedure A1: PS (8 ꢁ 1 mg) was melted at 1008C in a 1 ml V-vial; to that,
[18F]–Fꢃaq (5 ml), substrate (3 ꢁ 0.5 mg) and 50 ml of CH3CN were added. The
reaction vessel was then closed and heated (1008C, heating block) without
stirring for 30 min. After cooling, the content of the vial and a 600 ml CH3CN
rinse were transferred, assayed for radioactivity and analyzed by Radio-
HPLC.
Procedure A2: [18F]–Faꢃq (100 ml) was added to 8 ꢁ 1 mg of PS melted at
1008C in a 1 ml V-vial. Azeotropic distillation conducted with five portions of
100 ml of CH3CN (1008C, Ar stream) was followed by the addition of
3 ꢁ 0.5 mg of substrate and of 50 ml of 10% H2O in CH3CN as solvent. The
reaction was heated in a closed vessel for 30 min without stirring (1008C,
heating block); after cooling, the content of the vial and a 600 ml CH3CN rinse
were transferred, assayed for radioactivity and analyzed by Radio-HPLC.
Procedure B1: [18F]–Faꢃq (100 ml) was added to 8 ꢁ 1 mg of PS melted at
1008C in a 1 ml V-vial. Azeotropic distillation conducted with five portions of
100 ml of CH3CN (1008C, Ar stream) was followed by the addition of
3 ꢁ 0.5 mg of substrate and of 50 ml of a 1:1 mixture of CH3CN and 1-ethyl-3-
methylimidazolinium triflate (EMIMTfO) as solvent. The reaction was heated
in a closed vessel for 30 min without stirring (1008C, heating block). After
cooling, the content of the vial and a 600 ml CH3CN rinse were transferred,
assayed for radioactivity and analyzed by Radio-HPLC.
Procedure B2: [18F]–Faꢃq (100 ml) was added to 8 ꢁ 1 mg of PS melted at
1008C in a 1 ml V-vial. Azeotropic distillation conducted with five portions of
100 ml of CH3CN (1008C, Ar stream) was followed by the addition of
3 ꢁ 0.5 mg of substrate and of 50 ml of a 1:1 solution of CH3CN and 1-ethyl-3-
methylimidazolinium triflate (EMIMTfO) as solvent. The reaction was heated
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

382
G. PASCALI ET AL.
by microwave irradiation: 10 cycles of {5 s ON, 15 s OFF} at 200 W. After
cooling, the content of the vial and a 600 ml CH3CN rinse were transferred,
assayed for radioactivity and analyzed by Radio-HPLC.
Procedure using C: K222 (2.26 mg, 6 mmol) in acetonitrile (30 ml) and of
K2CO3 (0.415 mg, 3 mmol) in water (15 ml) were added to a 1 ml V-vial. [18F]–
Fꢃaq (100 ml) was added and azeotropically dried with five portions of 100 ml of
CH3CN (1008C, Ar stream); the addition of 3 ꢁ 0.5 mg of substrate and of
50 ml of CH3CN followed. The reaction was heated in a closed vessel for
30 min without stirring (1008C, heating block); after cooling, the content of the
vial and a 600 ml CH3CN rinse were transferred, assayed for radioactivity and
analyzed by Radio-HPLC.
Procedure using D: [18F]–Faꢃq (100 ml) was added and azeotropically dried
with five portions of 100 ml of CH3CN (1008C, Ar stream). 3 ꢁ 0.5 mg of
substrate and 50 ml of CH3CN were added and the vessel closed and heated
(1008C, heating block) for 30 min without stirring. After cooling, the content
of the vial and a 600 ml CH3CN rinse were transferred, assayed for
radioactivity and analyzed by Radio-HPLC.
Conclusion
A new method for [18F]radiofluorination utilizing PS has been studied and
developed. This procedure allows 18F incorporation through halogen
exchange (F, Cl, I are reported) that is comparable or even superior to the
traditional K222/K2CO3 in CH3CN procedure for some of the model
compounds reported. The best reaction conditions require as solvent 10%
H2O in CH3CN, since radiofluorination does not occur from an azeotropically
dried medium. It is clear from solubility studies that the reactivity of this
system is not due to the better availability of the fluoride ion in pure
acetonitrile, as proposed in the previous literature. The procedure works best
on strongly deactivated aromatic rings while substrates not activated to SNAr
do not seem to be amenable for fluorine exchange. For aliphatic substitutions,
displacement proceeds better with iodide than tosylate as leaving group.
Although PS/HF may be rather limited in scope, it may be an alternative to
more established methods in some cases.
Acknowledgements
FAB mass spectra were performed by the Mass Spectrometry Facility of the
NIDDK, Bethesda, MD. We thank the NIH PET cyclotron engineers and
operators for production of F-18. We acknowledge the helpful comments and
suggestions from Dr. Y. Ma, Dr. L. Lang and Dr. M. B. Sassaman.
References
1. Chambers RD, Korn SR, Sandford G. J Fluorine Chem 1994; 69: 103–108.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383

PS/H18 F IN RADIOFLUORINATIONS
383
!
2. (a) Darabantu M, Lequeux T, Pommelet J-C, Ple N, Turk A. Tetrahedron Lett
!
2000; 41: 6763–6767; (b) Darabantu M, Lequeux T, Pommelet J-C, Ple N, Turk
A. Tetrahedron 2001; 57: 739–750.
3. Rogers RD, Seddon KR. Science 2003; 302 (5646): 792–793.
4. Kim DW, Song CE, Chi DY. J Am Chem Soc 2002; 124: 10278–10279.
5. Kim DW, Choe YS, Chi DY. Nucl Med Biol 2003; 30: 345–350.
6. Snyder SE, Kilbourn MR. Chemistry of Fluorine-18 Radiopharmaceuticals. In
Handbook of Radiopharmaceuticals: Radiochemistry and Applications, Michael
JW, Carol R (eds). Wiley: West Sussex, England, 2003; 195–228.
7. Namboodiri VV, Varma RS. Org Lett 2002; 4 (18): 3161–3163.
8. Tewson TJ. Nucl Med Biol 1997; 24: 755.
9. Lang L, Jagoda E, Schmall B, Vuong B-K, Adams HR, Nelson DL, Carson RE,
Eckelman WC. J Med Chem 1999; 42: 1576–1586.
10. Gail R, Coenen HH. Appl Radiat Isotopes 1994; 45 (1): 105–111.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 373–383