Labeling of benzodioxin piperazines with fluorine-18 as prospective radioligands for selective imaging of dopamine D4 receptors

Article abstract of DOI:10.1002/jlcr.3074

The D₄ receptor is of high interest for research and clinical application but puts high demands on appropriate radioligands to be useful tools for investigation. Search for adequate radioligands suitable for in vivo imaging is therefore still in progress. The potential neuroleptic drug 6-(4-[4-fluorobenzyl]piperazin-1-yl)benzodioxin shows high affinity and selectivity to the D₄ receptor. Derivatization of this lead structure by adding hydrophilic moieties was carried out in order to lower its lipophilicity what led to three new putative dopamine receptor D₄ ligands. A comprehensive description of the syntheses of standard compounds and corresponding labeling precursors is given which were obtained in satisfactory yields. Furthermore, the radiosyntheses by direct ¹⁸F-labeling and build-up synthesis were compared. All derivatives of 6-(4-[4-fluorobenzyl]- piperazin-1-yl)benzodioxin were successfully synthesized in ¹⁸F- labeled form with radiochemical yields of 9-35% and molar activities of 30-60 GBq/μmol using one-pot procedures. 2013 John Wiley & Sons, Ltd. Derivatives of the lead structure 6-(4-[4-fluorobenzyl]-piperazine-1-yl) benzodioxin were successfully synthesized, labeled via direct ¹⁸F-substitution or build-up synthesis leading to new putative radioligands for imaging of the dopamine D4 receptor. Reductive amination proved as the method of choice for preparation of substituted benzyl-derivatives of piperazine as precursors and standards, and also superior for build-up radiofluorination (RCY = 9 - 35;%) in comparison to direct ¹⁸F- labeling (RCY = 1 - 5;%).

Full text of DOI:10.1002/jlcr.3074

Research Article
Received 15 February 2013,
Revised 10 May 2013,
Accepted 19 May 2013
Published online 10 August 2013 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/jlcr.3074
Labeling of benzodioxin piperazines with
fluorine-18 as prospective radioligands for
selective imaging of dopamine D4 receptors
a
a
Fabian Kügler,a,b Johannes Ermert, and Heinz H. Coenen
*
The D4 receptor is of high interest for research and clinical application but puts high demands on appropriate radioligands
to be useful tools for investigation. Search for adequate radioligands suitable for in vivo imaging is therefore still in
progress. The potential neuroleptic drug 6-(4-[4-fluorobenzyl]piperazin-1-yl)benzodioxin shows high affinity and selectivity
to the D4 receptor. Derivatization of this lead structure by adding hydrophilic moieties was carried out in order to lower its
lipophilicity what led to three new putative dopamine receptor D4 ligands. A comprehensive description of the syntheses of
standard compounds and corresponding labeling precursors is given which were obtained in satisfactory yields. Furthermore,
the radiosyntheses by direct 18F-labeling and build-up synthesis were compared. All derivatives of 6-(4-[4-fluorobenzyl]-
piperazin-1-yl)benzodioxin were successfully synthesized in 18F-labeled form with radiochemical yields of 9–35% and molar
activities of 30–60GBq/mmol using one-pot procedures.
Keywords: dopamine D4 receptor; radioligand; n.c.a. [18F]fluoride; one-pot radiosynthesis; positron emission tomography
requirements of the aforementioned criteria is 1-phenyl-4-
benzylpiperazine. The fact that this compound is very
Introduction
The dopamine D4 receptor belongs to the type of receptors that
nonselective and has even higher affinity for D2 receptors
consistently cause problems in their determination when using
underlines the differences between known theory and reality.
labeling techniques (antibody, mRNA, or ligand labeling).
Therefore, hitherto, there exists neither an adequate radioligand
for autoradiographic studies nor for the molecular imaging using
positron emission tomography (PET) or single photon emission
computed tomography (SPECT), but it would be a powerful tool
allowing noninvasive D4 receptor imaging. The most probable
reason for the lack of a suitable radiotracer is the extremely
low concentration of this receptor subtype in the mammalian
However, a hydroxyl substitution in 4-position of the N-aryl ring
lowers D2 and increases D4 affinity. To this end, a methoxy
substituent in this position leads to a compound that shows
moderate D4 affinity, whereas it is inactive at the D2 and D3
receptor (cf. Scheme 1).11
Further systematic development by the Thurkauf group12 led to
the fluorinated benzodioxin D4 antagonist 6-(4-[4-fluorobenzyl]-
piperazin-1-yl)benzodioxin (1), which shows a good D4 subtype
selectivity with a D2;3/D4 ratio of >1250. As one of the rare selective
D4 antagonists available, it may be helpful to verify the D4
brain. In consequence, even the concentration of this receptor
is not very well known what is also true for its localization and
function as widely discussed in the literature.1,2 The dopamine
hypothesis of schizophrenia or, in case of verification, be used for
D4 receptor is part of the small group of dopaminergic receptors,
selective blocking of D4 receptors as an antipsychotic drug for
treatment of this disease.12
which show comparatively close similarities in their
pharmacology. There is evidence that this receptor system is
Although it is known that affinity and selectivity requirements
associated with functions and pathophysiological processes of
for an extremely low-expressed receptor are very demanding,
the human brain in schizophrenia, substance abuse,3 and
neurobehavioral dysfunctions such as attention deficit
problems developing suitable radioligands for PET or SPECT are
often not due to a lack of selective compounds. Rather, a high
hyperactivity disorder,4,5 libido dysfunctions,6,7 and novelty
lipophilicity of the radioligand causing adhesion to proteins
seeking.8,9 Nevertheless, a confirmation of these postulates and
their verification by in vivo molecular imaging methods in
humans is missing due to lack of suitable radioligands.
Chemical structures of known ligands selective for the D4
and lipids tends to increase nonspecific binding, an essentially
aInstitute of Neurosciences and Medicine, INM-5: Nuclear Chemistry,
Forschungszentrum Jülich GmbH, Jülich, Germany
receptor mostly consist of a central backbone of piperazine,
sometimes a piperidine, flanked with two aromatic moieties10
bPresent address: Nuklearmedizinische Klinik und Poliklinik, Technische Universität
München, Ismaninger Straße 22, D-81675 München, Germany
(cf. Scheme 1). One of these aromatic groups is always an
arylamine to which the other is linked by a methylene group.
Longer linkers cause an increased D2 and then D3 selectivity.
These findings are meanwhile also confirmed by CoMFA and
cloning studies.8 The simplest structure that fulfills the minimum
*Correspondence to: Johannes Ermert, Institute of Neurosciences and Medicine,
INM-5: Nuclear Chemistry, Forschungszentrum Jülich GmbH, Germany.
E-mail: j.ermert@fz-juelich.de
J. Label Compd. Radiopharm 2013, 56 609–618
Copyright © 2013 John Wiley & Sons, Ltd.

F. Kügler et al.
Scheme 1. Change of the dopamine receptor affinity in dependence of substitution in 4-position of the N-aryl of the piperazine D4 ligand backbone.
electron beam energy of 70 eV. High resolution electron spray mass
spectra were recorded on an LTQ FT Ultra (Thermo Fischer). Melting
points are uncorrected and were determined on a Mettler FP-61
apparatus in open capillaries.
nonsaturable component of the total tissue uptake. Here,
lipophilicity can become the more critical parameter regarding
the effectiveness of a radioligand.13 The calculated lipophilicity
of the previously described D4 antagonist 1 fits with a cLog P
value of 3.35 to a generally accepted range for suitable cerebral
neuroligands. However, with regard to the very small
concentrations of D4 receptors such a relatively high value is
particularly critical if the relative ratio of receptor-binding to
nonspecific binding becomes very low. Therefore, a series of
derivatives were developed starting from 1 as lead structure with
a lower lipophilicity in the range of logD of 1.8 to 2.3 at pH 7.4.
Generally, many criteria besides a suitable lipophilicity have to
be fulfilled in order to develop a successful imaging agent for
in vivo human studies using PET or SPECT. Thus, for the
Chemistry
The following compounds were synthesized according to reported
literature methods: 4-N,N-dimethylaminobenzaldehyde and 4-N,N-
dimethylamine-3-methoxybenzaldehyde were prepared by amination
of corresponding fluorine compounds as described by Wilson et al.;16
4-formylphenyltrimethylammonium triflate 12,17 3-benzyloxy-4-
formylphenyltrimethylammonium triflate 12b,18 3-benzyloxy-4-
fluorobenzaldehyde 6b,18 3-benzyloxy-4-nitrobenzaldehyde 12b0,18 and
1-(1,4-benzodioxin-6-yl)piperazine 212 were prepared as described in
the corresponding literature.
compounds of interest here,
a detailed pharmacological
4-Formyl-3-methoxyphenyltrimethylammonium triflate 12a
evaluation of their in vitro affinity and lipophilicity, as well as
their in vivo binding behavior, membrane permeability, and
biodistribution including metabolic stability was recently
published.14 Here, the syntheses of the non-radioactive standard
compounds and the 18F-radiosyntheses are described in detail
together with appropriate analytical conditions for the
identification of radiolabeled compounds.
4-Formyl-3-methoxyphenyltrimethylammonium triflate 12a was obtained
from 3 g (19.46 mmol) 4-fluoro-3-methoxybenzaldehyde as a white solid
(5.9mg, 17.2 mmol, 88.4%) according to a method by Wilson et al..16 mp
155.5–156.4 ꢁC, 1H NMR (DMSO-d6, 200.13 MHz) d 3.69 (s, 9H), 3.73 (s, 3H),
7.77–7.82 (m, 3H), 10.09 (s, 1H) ppm. FT-MS (ESI): 195.33 m/z (100)
[M+ H]+, elemental analysis: found: %C, 41.98; %H, 4.70; %N, 4.05; expected:
%C, 41.98; %H, 4.70; %N, 4.08.
1-(1,4-Benzodioxin-6-yl)piperazine 2
Experimental
Under an atmosphere of dry argon, a mixture of 1.27 g (8.4 mmol) of 6-
amino-1,4-benzodioxine and 1.5 g (8.4 mmol) of bis(2-chloroethyl)amine
hydrochloride in 3 mL of dry ethylenglycol was heated at 150 ꢁC over
night. After cooling to room temperature, methanol was given to the
mixture in order to precipitate a white crystalline solid. The solid was
treated with a saturated sodium carbonate solution and extracted with
ethyl acetate. The combined organic layers were washed with brine,
dried over sodium sulfate, and concentrated in vacuo to give a pure
white solid. The mother liquor was further reduced in vacuo to get rid
of methanol and extracted with a saturated sodium carbonate solution
as described previously. The product was purified by column
General
All reagents and anhydrous solvents were purchased from Aldrich
(Steinheim, Germany) or Fluka (Buchs, Switzerland) and used without
further purification. 6-Fluoronicotinaldehyde 6d was obtained from Alfar
Aesar (Karlsruhe, Germany) and 4-fluoro-3-hydroxybenzaldehyde 6c from
ABCR (Karlsruhe, Germany).
Sep-Pak C-18 plus-cartridges were acquired from Waters (Eschborn,
Germany); EN cartridges and Li-Chrolut glass columns (65ꢀ 10mm) from
Merck (Darmstadt, Germany). Thin layer chromatography (TLC) was run
on precoated plates of silica gel 60 F254 or Alumina N from Merck
(Darmstadt, Germany). The compounds were detected by UV at 256nm.
HPLC was performed on the following system from Dionex (Idstein,
Germany): an Ultimate 3000 LPG-3400A HPLC pump, an Ultimate 3000
VWD-3100 UV/VIS-detector (272 nm), a UCI-50 chromatography interface,
and an injection valve P/N 8215. Reversed-phase HPLC was carried out
using a Gemini 5 mm C18 110A column, for analytical separations with a
dimension of 250ꢀ 4.6 mm (flow 1mL/min) and for semi-preparative
applications 250ꢀ 10 mm (flow 5 mL/min) from Phenomenex
chromatography (CHCl3/methanol 5:1) to obtain
a slight beige
solid (1.17 g, 5.3 mmol, 63%). mp 169–171 ꢁC. TLC (CHCl3/methanol,
5:1): Rf = 0.18. 1H NMR (200.13 MHz, DMSO-d6): d 2.78–2.92 (m, 8H), 4.17
(m, 4H), 6.43 (m, 2H), 6.70 (d, 1H) ppm. FT-MS (ESI): m/z 221.21 [M + H]+,
HRMS (221.1290) found: 221.1284.
1-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)-4-(4-fluorobenzyl)-
piperazine 1
(Aschaffenburg, Germany). Determination of molar activities was carried A mixture of 350 mg of 2 (1.6 mmol), 394 mg of 4-fluorobenzylbromid
out according to standard procedures by means of the previously described (2.08 mmol), 443 mg of K2CO3 (3.2 mmol), and 345 mg of potassium
semi-preparative HPLC, comparing UV-signals under radioactivity peaks iodide (2.08 mmol) was suspended in 20 mL of dry acetonitrile and
with a calibration UV/mass-curve, as described earlier.15 Radio TLC heated at 85 ꢁC for 18 h. After cooling to room temperature, the mixture
chromatograms were analyzed by UV detection and on a Packard Instant was treated with 50 mL of water and extracted with ethyl acetate. The
Imager. 1H, 13C, and 19F NMR spectra were recorded on a Bruker DPX combined organic layers were washed with brine, dried over sodium
Avance 200 spectrometer with samples dissolved in CDCl3 or DMSO-d6.
sulfate and concentrated in vacuo. The product was purified by column
All chemical shifts are given in the succeeding text in d ppm using the chromatography (n-hexane/ethyl acetate 1:2) to obtain a white solid
signals of the appropriate solvent as a reference. Mass spectra were (240 mg, 0.73 mmol, 46%). mp 120 ꢁC. TLC (n-hexane/ethyl acetate, 1:2):
obtained from a Finnigan Automass Multi mass spectrometer with an Rf = 0.7. 1H NMR (CDCl3, 400 MHz)
d 2.56 (t, J = 4.8 Hz, 4H), 3.06
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 609–618

F. Kügler et al.
(t, J = 4.8 Hz, 4H), 3.50 (s, 2H), 4.19 (m, J = 4.5 Hz, 2H), 4.21 (m, J = 4.6 Hz,
2H), 6.43 (m, 2H), 6.75 (m, 1H), 6.98 (m, 1H), 7.29 (m, 2H) ppm. 13C NMR
(CDCl3, 50.33 MHz) d 50.2, 53.0, 62.2, 64.2, 64.6, 105.6, 110.4, 115.0,
117.3, 130.6, 133.7, 137.3, 143.6, 146.5, 162.0 ppm. 19F NMR
General procedure for the preparation of 3-substituted 6-(4-
benzylpiperazine-1-yl)benzodioxin derivatives 7a–e19
In a 100-mL two-neck flask, 1-(1,4-benzodioxine-6-yl)piperazine (2)
(1.816 mmol,
1 equiv) and corresponding benzaldehyde (3a–e,
(188.28 MHz, DMSO-d6)
d
ꢂ115.6 ppm. FT-MS (ESI): m/z 329.166
[M + H]+. Elemental analysis: found: %C, 69.32; %H, 6.44; %N, 8.29;
expected: %C, 69.49; %H, 6.45; %N, 8.53.
2.724 mmol, 1.5 equiv) were dissolved in 20 mL of dry methanol.
After an addition of acetic acid (96%, 5.45 mmol, 3 equiv.), sodium
cyanoborohydride (2.724 mmol, 1.5 equiv.) was added in small
portions and rinsed with methanol. The reaction was heated for
12–24 h at 60 ꢁC. After cooling, the reaction was quenched by
addition of 50 mL of saturated sodium bicarbonate solution,
extracted three times with chloroform, and washed with brine.
The combined organic layer was dried over sodium sulfate and
evaporated to dryness after filtration. The residue was purified
General procedure for preparation of amides 9a, 9b, and 9c
The corresponding aromatic carboxylic acid (3.54 mmol) was dissolved in
20 mL of anhydrous dichloromethane (DCM) followed by a small amount
of DMF (20 mL) and oxalylchloride (5.3 mmol, 1.5 equiv). The mixture was
stirred at room temperature until evaporation of gas had stopped (2–3 h),
and the solvent was evaporated in vacuo to obtain the corresponding
acid chloride.
by chromatography over
acetate 1:2).
a silica gel column (n-hexane/ethyl
Under an argon atmosphere, 1-(1,4-benzodioxin-6-yl)piperazine (2)
was dissolved in DCM, and triethylamine (2.2 mmol, 0.6 equiv) was
added. The solution was cooled in an ice bath at 0 ꢁC and the respective
acid chloride redissolved in 5 mL of DCM, which was added dropwise.
The reaction mixture was warmed to room temperature and stirred
overnight for at least 10 h. To quench the reaction, 20 mL of saturated
sodium bicarbonate solution was added. The solution was extracted with
DCM, washed with water and saturated sodium chloride solution, and
dried over sodium sulfate. After evaporation of the solvent in vacuo,
the residue was purified by flash chromatography.
1-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)-4-(4-fluoro-3-methoxy-
benzyl)piperazine 7a
The methoxy derivative was obtained from the commercially available 4-
fluoro-3-methoxybenzaldehyde (6a) (420 mg, 2.724 mmol) as a pale
yellow crystalline solid (551 mg, 1.54 mmol, 85%). mp 100 ꢁC. TLC
(n-hexane/ethyl acetate, 1:2): Rf = 0.44. 1H NMR (CHCl3, 400 MHz)
d 2.56 (t, J = 4.8 Hz, 4H), 3.06 (t, J = 4.8 Hz, 4H), 3.48 (s, 2H), 3.88
(s, 3H), 4.18 (m, J = 1.6 Hz, 2H), 4.20 (m, J = 1.6 Hz, 2H), 6.44 (m, 1H),
6.45 (m, 1H), 6.75 (m, 1H), 6.82 (m, 1H), 6.98 (m, 2H) ppm. 13C NMR
(CDCl3, 50.33 MHz) d 50.2, 53.1, 56.2, 62.5, 64.3, 64.6, 105.6, 110.3,
113.9, 115.5, 117.3, 121.3, 134.5, 137.3, 143.6, 146.5, 147.4,
151.5 ppm. 19F NMR (188.28 MHz, DMSO-d6) d ꢂ137.8 ppm. FT-MS
(ESI): m/z 359.18 [M + H]+. HRMS (359.177) found: 359.17642.
(4-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)piperazine-1-yl)(4-fluoro-
phenyl)methanone 9a
Obtained from 4-fluorobenzoic acid (286 mg, 2.04 mmol) as white foamy 1-(3-(Benzyloxy)-4-fluorobenzyl)-4-(2,3-dihydrobenzo[b][1,4]dioxin-
solid (460 mg, 1.34 mmol, 66%). TLC (n-hexane/ethyl acetate 1:2): Rf =
0.56. 1H NMR (CHCl3, 400 MHz) d 3.04 (b, 4H, 4CH2); 3.7 (b, 4H, 4CH2);
4.19 (m, 2H, 2CH2); 4.22 (m, 2H, 2CH2); 6.44–6.45 (m, 2H, Ar-CH); 6.77
(m, 1H, Ar-CH); 7.08 (m, 2H, Ar-CH); 7.42 (m, 2H, Ar-CH) ppm. 13C NMR
(CDCl3, 50.33 MHz) d 42.3, 50.8, 64.2, 64.5, 106.5, 11.0, 115.6, 117.5,
129.4, 131.6, 138.1, 143.7, 145.9, 163.4, 169.4 ppm. 19F NMR
(188.28 MHz, DMSO-d6) d ꢂ110.1 ppm. FT-MS (ESI): m/z 343.19 (100)
[M + H]+. Elemental analysis: found: %C, 66.90; %H, 5.85; %N, 7.97;
expected: %C, 66.66; %H, 5.59; %N, 8.18.
6-yl)piperazine 7b
Obtained from 3-benzyloxy-4-fluorobenzaldehyde (6b) (160 mg,
0.7 mmol) as a white crystalline solid (175 mg, 0.41 mmol, 58%). mp
94 ꢁC. TLC (n-hexane/ethyl acetate, 1:2): Rf = 0.74. 1H NMR (CHCl3,
400 MHz) d 2.40 (m, 4H); 2.92 (m, 4H); 3.41 (s, 2H); 4.11 (m, 2H); 4.14
(m, 2H); 5.14 (s, 2H); 6.36 (d, J = 2.7 Hz, 1H); 6.38 (dd, J = 2.9 Hz/8.6 Hz;
1H); 6.66 (d, J = 8.7 Hz, d); 6.84 (m,1H); 7.14 (m, 2H); 7.31–7.42 (m, 5H)
ppm. 13C NMR (CDCl3, 50.33 MHz) d 49.7, 52.9, 61.8, 64.2, 64.7, 70.6,
105.2, 109.8, 116.1, 116.3, 117.4, 121.8, 128.3, 128.4, 128.9, 135.3, 137.0,
137.1, 143.8, 146.2, 146.5, 151.3 ppm. 19F NMR (188.28 MHz, DMSO-d6) d
ꢂ137.1 ppm. FT-MS (ESI): m/z 345.21 [M + H]+. Elemental analysis: found:
%C, 71.80; %H, 6.13; expected: %C, 71.87; %H, 6.26.
(4-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)piperazine-1-yl)(4-nitro-
phenyl)methanone 9b
5-((4-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)piperazine-1-yl)methyl)-
Obtained from 4-nitrobenzoic acid (267 mg, 1.6 mmol) as a yellow solid
1
(520 mg, 1.41 mmol, 88%). TLC (n-hexane/ethyl acetate 1:2): Rf = 0.49. H 2-fluorophenol 7c
NMR (CHCl3, 400 MHz) d 2.92 (b, 2H); 2.97 (b, 2H); 3.13 (b, 2H); 3.45
The hydroxyl derivative was obtained from 4-fluoro-3-hydroxybenzaldehyde
(b, 2H); 4.19 (m, 2H); 4.22 (m, 2H) 6.44–6.45 (m, 2H); 6.77 (d, 1H); 7.58
(dt, J = 8,7 Hz; 2H); 8.28 (dd, J = 8,3 Hz, 2H) ppm. 13C NMR (CDCl3,
50.33 MHz) d 42.3, 47.6, 50.6, 51.1, 64.2, 64.6, 106.6, 11.1, 117.6, 123.9,
128.1, 138.3, 141.7, 143.7, 145.6, 148.4, 167.9 ppm. FT-MS (ESI): m/z
370.14 (90) [M + H]+. Elemental analysis: found: %C, 61.49; %H, 5.50; %
N, 11.16; expected: %C, 61.78; %H, 5.18; %N, 11.38.
(6c) (70 mg, 0.5 mmol) as a white crystalline solid (112 mg, 0.35 mmol, 66%).
mp 186–187 ꢁC. TLC (n-hexane/ethyl acetate, 1:2): Rf = 0.35. 1H NMR (CHCl3,
400 MHz) d 2.57 (m, 4H); 3.06 (m, 4H); 3.45 (s, 2H); 4.18 (m, 2H); 4.20 (m,
2H); 6.44 (m, 2H); 6.74 (m, 1H); 6.79 (m, 1H); 6.98 (m, 2H); 9.84 (br, 1H)
ppm. 13C NMR (CDCl3, 50.33 MHz) d 50.1, 52.9, 62.2, 64.2, 64.6, 105.7, 110.4,
115.1, 117.3, 118.1, 121.3, 134.5, 137.4, 143.4, 143.6, 146.4, 150.3 ppm. 19F
NMR (188.28 MHz, DMSO-d6) d ꢂ142.6 ppm. FT-MS (ESI): m/z 345.19 (100)
[M + H]+. Elemental analysis: %C, 66.02; %H, 6.26; %N, 7.68; expected: %C,
66.26; %H, 6.15; %N, 8.13.
(4-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)piperazine-1-yl)(6-fluoro-
pyridine-3-yl)-methanone 9c
1-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)-4-((6-fluoropyridine-3-yl)-
methyl)piperazine 7d
Obtained from 2-fluoronicotinic acid (450 mg, 3.2 mmol) as a white solid
(680 mg, 2 mmol, 62%). TLC (n-hexane/ethyl acetate, 1:3): Rf = 0.48. 1H
NMR (CHCl3, 400 MHz) d 3.06 (br, 2H); 3.12 (br, 2H); 3.52 (br, 2H); 3.80
(br, 2H); 4.21 (m, 2H); 4.25 (m, 2H); 6.52 (s, 1H); 6.53 (dd, J = 2.8 Hz/
8.6 Hz, 1H); 6.79 (d, J = 9.2 Hz, 1H); 7.36 (dd, J = 2.5 Hz/8.4 Hz, 1H); 8.15
(ddd, J = 2.5 Hz/8.2 Hz/8.2 Hz, 1H); 8.42 (d, J = 2.0 Hz, 1H) ppm. 13C NMR
(CDCl3, 50.33 MHz) d 41.4, 46.9, 49.3, 49.6, 63.7, 64.1, 105.4, 109.5, 109.9,
116.9, 130.1, 137.0, 141.2, 143.3, 145.5, 146.2, 162.9, 165.6 ppm. 19F NMR
(188.28 MHz, DMSO-d6) d ꢂ66.7 ppm. FT-MS (ESI): m/z 344.14 (100)
[M + H]+, HRMS (344.14104) found: 344.1403.
Method A: Obtained from 6-fluoronicotinaldehyde (6d) (100 mg,
0.8 mmol) as a white crystalline solid (35 mg, 0.11 mmol, 14%).
Method B: A 25 mL flask with a silicone septum was twice
heated under vacuo while filled with argon. After cooling, 1 mL
of borane (1 M in THF) was filled into the flask and 150 mg of
7b (0.44 mmol) dissolved in 4 mL of dry THF were added
afterwards. The reaction was heated to 65 ꢁC for 7 h and a second
J. Label Compd. Radiopharm 2013, 56 609–618
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

F. Kügler et al.
at 160 ꢁC. Monitoring of the reaction progress was carried out by radio
HPLC from aliquots of about 30–50 mL (Gemini 5 m RP18 A110,
250ꢀ 4.6 mm, 1 mL/min, isocratic 60:40:0.1 v/v/v CH3CN/H2O/TEA) and
by radio TLC (1:2 n-hexane/ethyl acetate) determining RCYs and the
optimal reaction time. For reduction of the intermediate compound,
the reaction mixture was diluted with 20 mL of water, passed through a
Sep Pak C18 cartridge, washed with 5 mL of water, and dried with air.
The cartridge was eluted through a drying cartridge charged with 4 Å
molecular sieve and 270 mg dry sodium sulfate using 5 mL absolute
dichloromethane. After removal of the solvent in vacuo (800 up to
portion of borane (0.7 mL, 1 M in THF) was added. After another
22 h, the reaction was quenched by addition of small amounts
of ice, extracted three times with chloroform and washed with
brine. The combined organic layer was dried over sodium sulfate
and evaporated to dryness after filtration. The residue was
purified over a silica gel column (chloroform/methanol 8:1) to
obtain a white solid (43 mg, 0.132 mmol, 30%). mp 94 ꢁC. TLC
(chloroform/methanol 8:1): Rf = 0.63; (ethyl acetate/methanol/
1
diethyl amine 96:2:2): Rf = 0.78 . H NMR (CHCl3, 400 MHz) d 2.52
(br, 4H); 3.00 (br, 4H); 3.47 (s, 2H); 4.14 (m, 2H); 4.16 (m, 2H); 300 mbar) at room temperature, 0.2 mL of BH3THF (1 M, 0.2 mmol)
solution and 0.7 mL of dry THF were added, and the solution was stirred
for 10 min at 65 ꢁC. The reaction mixture was cooled in an ice bath,
carefully quenched with 20 mL of water, and passed through a second
Sep Pak C18 cartridge. After washing with water, drying in air and eluting
with 1 mL of acetonitrile, the solution was injected on a semi-preparative
HPLC system (Gemini 5 m RP18 A110, 250ꢀ 10 mm, 4 mL/min, isocratic
30:70:0.1 v/v/v CH3CN/H2O/TFA).
6.38 (m, 2H); 6.71 (m, 1H); 6.84 (dd, J = 2.8 Hz/8.4 Hz, 1H); 7.74
(ddd, J = 2.3 Hz/8.0 Hz/8.1 Hz, 1H); 8.09 (d, J = 1.9 Hz, 1H) ppm.
13C NMR (CDCl3, 50.33 MHz) d 49.7, 52.9, 58.5, 64.2, 64.7, 105.2,
109.8, 117.4, 124.4, 133.7, 137.1, 140.8, 143.8, 146.4, 149.4,
150.5 ppm. 19F NMR (188.28 MHz, DMSO-d6) d ꢂ71.2 ppm.
FT-MS (ESI): m/z 330.16 [M + H]+, HRMS (330.16172) found:
330.16124.
Method B: A solution of precursor 12 (8 mg, 25.5 mmol) in 0.5 mL of
anhydrous DMSO was added to a reaction vial at 110 ꢁC, and the reaction
was monitored via radio HPLC (Gemini 5 m RP18 A110, 250ꢀ 4.6 mm,
1-((6-Chloropyridine-3-yl)methyl)-4-(2,3-dihydrobenzo[b][1,4]-
dioxin-6-yl)piperazine 7e
1 mL/min, isocratic 60:40:0.1 v/v/v CH3CN/H2O/TEA).
A solution of
Also obtained by the general amination procedure from the
commercially available 6-chloronicotinaldehyde (6e) (380 mg, 2.68 mmol)
as a white crystalline solid (445 mg, 1.29 mmol, 71%). mp 94 ꢁC. TLC
(n-hexane/ethyl acetate, 1:2): Rf = 0.53. 1H NMR (CHCl3, 400 MHz)
d 2.45 (m, 4H); 2.94 (m, 4H); 3.51 (s, 2H); 4.10 (m,2H); 4.14 (m, 2H);
6.36 (d, J = 2.5 Hz, 1H); 6.38 (dd, J = 2.7 Hz/8.6 Hz, 1H); 6.66
(d, J = 8.6 Hz, 1H); 7.46 (d, J = 8.2 Hz, 1H); 7.77 (dd, J = 2.3 Hz/8.1 Hz,
1H); 8.31 (d, J = 2.2 Hz, 1H) ppm. 13C NMR (CDCl3, 50.33 MHz) d 50.2,
53.1, 61.9, 64.2, 64.6, 105.7, 110.4, 117.4, 124.0, 135.0, 137.5, 137.6,
143.6, 146.3, 148.2, 150.3 ppm. FT-MS (ESI): m/z 346.13 [M + H]+, HRMS
(346.1322) found: 346.1315.
intermediate 1 (8.7 mg, 40 mmol) in 40 mL acetic acid and 50 mL DMSO
and a solution of sodium cyanoborohydride (4 mg, 64 mmol) in 50 mL
DMSO were added successively. The reaction mixture was stirred for
15 min, diluted with water, and passed through a SepPak C18 cartridge,
washed with water (5 mL) and dried with air. The cartridge was eluted
with 1 mL of acetonitrile and injected on a semi-preparative HPLC system
(Gemini 5 m RP18 A110, 250ꢀ 10 mm, 4 mL/min, isocratic 30:70:0.1 v/v/v
CH3CN/H2O/TFA). The separated fraction was diluted with 15 mL of water
and passed through a second SepPak C18 cartridge. After washing with
water and drying in an argon stream, the cartridge was eluted with
5 mL of diethylether, which was evaporated in vacuo (800 up to
330 mbar). For in vitro studies, [18F]1 was formulated in 300 mL ethanol/
saline (5:1).
4-Formyl-2-benzyloxy-N,N-dimethylaniline 14 (alternative method
to Langer et al.18)
1-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)-4-(4-[18F]fluoro-3-methoxy-
benzyl)piperazine [18F]7a
Under an atmosphere of argon 4-bromo-2-benzyloxy-N,N-dimethylanilin
(13) (500 mg, 1.65 mmol) was dissolved in 6mL of anhydrous diethylether
and cooled in a carbon dioxide/acetone bath at ꢂ78 ꢁC. A 1.75mL of s-BuLi
(2.45mmol, 1.4M in cyclohexane) was added cautiously, and the reaction
was stirred for 45 min at ꢂ78 ꢁC. Subsequently, DMF (200 mL, 2.6mmol)
was added all at once, and the reaction mixture was stirred for 1h at room
temperature. For quenching, the solution was diluted with 15 mL of water
and 15mL of saturated aqueous ammonium chloride and extracted with
diethylether. After washing with saturated sodium chloride solution, the
organic phase was dried over sodium sulfate, filtered, and evaporated in
vacuo to dryness. The crude product was purified by flash chromatography
on silica gel (n-hexane/ethyl acetate: 4:1) and obtained as a clear, light
yellow liquid (240mg, 0.95 mmol, 57%). TLC (n-hexane/ethyl acetate: 4:1):
Rf = 0.44. 1H NMR (DMSO-d6, 200MHz) d 2.91 (s, 6H); 5.18 (s, 2H); 6.97 (d,
1H); 7.46 (m, 7H); 9.79 (s, 1H) ppm. MS (ESI): m/z 256.21 [M+ H]+.
Analogous to method B of the radiosynthesis of [18F]1, a solution of 12a
(7 mg, 20 mmol) in 0.5 mL of anhydrous DMSO was added to the reaction
vial containing the dried [18F]fluoride at 110 ꢁC, and the progress of
reaction was monitored as described previously by radio HPLC (Gemini
5 m RP18 A110, 250ꢀ 4.6 mm, 1 mL/min, isocratic 60:40:0.1 v/v/v CH3CN/
H2O/TEA). Without further separation, a solution of 2 (8.7 mg, 40 mmol)
in 40 mL of acetic acid and 50 mL of DMSO and a solution of sodium
cyanoborohydride (4 mg, 64 mmol) in 50 mL of DMSO were rapidly added
to the reaction vial at the same temperature. The mixture was stirred for
15 min, diluted with water, and passed through a SepPak C18 cartridge,
washed with water (5 mL) and dried with air. The cartridge was eluted
with 1 mL of acetonitrile and the eluate injected on a semi-preparative
HPLC system (Gemini 5 m RP18 A110, 250ꢀ 10 mm, 4 mL/min, isocratic
30:70:0.1 v/v/v CH3CN/H2O/TFA). The separated product fraction was
diluted with 15 mL of water, passed through a SepPak C18 cartridge
again, washed with water (5 mL), and dried in a stream of argon. The
cartridge was eluted with 4 mL of diethylether, and the solvent was
evaporated in vacuo (800 up to 330 mbar). For further experimental use
in in vitro autoradiography studies, [18F]7a was formulated in 300 mL
ethanol/saline (5:1).
Radiochemistry
N.c.a. [18F]fluoride was produced via the 18O(p,n)18F nuclear reaction by
bombardment of an isotopically enriched [18O]water target20 with
17 MeV protons at the JSW cyclotron BC 1710 (INM-5, Research Center
Juelich). The aqueous [18F]fluoride solution (10–50 mL, 75–375 MBq) was
filled into a 5 mL conical vial (Reactivial) containing 1 mL of acetonitrile,
10 mg of Kryptofix 2.2.2 and 13 mL of an aqueous 1 M potassium
carbonate solution.21 The solvent was evaporated under a stream of
argon at 80 ꢁC and 600 mbar. This azeotropic drying was repeated twice
using 1 mL of dry acetonitrile, followed by evaporation at 8–15 mbar for
5 min.
6-(4-[4-[18F]Fluoro-3-hydroxybenzyl]piperazine-1-yl)benzodioxin
[18F]7c
The benzyl-protected precursor 12b (3 mg, 7 mmol) dissolved in 0.5 mL of
anhydrous DMSO was added to the vial containing dried [18F]fluoride
and heated at 110 ꢁC, and the progress of the reaction was determined
by radio HPLC (Gemini 5 m RP18 A110, 250ꢀ 4.6 mm, 1 mL/min, isocratic
80:20 v/v CH3CN/H2O). A solution of 2 (6.5 mg, 30 mmol) in 40 mL acetic
6-(4-[4-[18F]Fluorobenzyl]piperazine-1-yl)benzodioxin [18F]1
Method A: A solution of precursor 9b (8 mg, 22 mmol) in 0.5 mL of
anhydrous DMSO was added to the vial containing the dried [18F]fluoride acid and 50 mL DMSO and a solution of sodium cyanoborohydride
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 609–618

F. Kügler et al.
(4 mg, 64 mmol) in 50 mL DMSO were immediately added to the reaction
vial and stirred for 7 min at 110 ꢁC. Subsequently, the solution was
diluted with 20 mL of water and passed through an EN cartridge. After
washing with 5 mL of water and drying in a stream of argon for 8 min,
the product was eluted from the cartridge and then passed through an
unconditioned Alumina N cartridge with 5 mL of anhydrous diethylether
in a second reaction vial. The ether was evaporated in vacuo, and the
residue was diluted with 1 mL of anhydrous methanol. Radio HPLC
analysis showed a nearly quantitative conversion of [18F]6b to [18F]7b.
Palladium (black) (20 mg) and 250 mg of ammonium formate were
synthesized as intermediates and coupled afterwards. Analogous
to the description of Liu et al.,22 the latter intermediate 1-(1,4-
benzodioxin-6-yl)piperazine (2) was synthesized via cyclization
of the commercially available aminodioxin 2,3-dihydrobenzo[b]-
[1,4]dioxin-6-amine with bis(2-chloroethyl)amine hydrochloride
in diethylene glycol or diethylene glycol monomethyl ether15
(cf. Scheme 2). Thereby, higher yields were found by saving
one reaction step in contrast to the palladium catalyzed coupling
with the corresponding dioxine-6-bromide as described earlier
added to the solution, and the resulting suspension was stirred for by Hodgetts et al.12 Double nucleophilic substitution on anilines
8 min at 60 ꢁC. The start of the reaction could be observed by a strong
evolution of gas. The mixture was cooled in an ice bath, filtered through
at high temperature (Prelog cyclization) constitutes the easiest
method to generate arylpiperazines when appropriate anilines
a small glass column with two frits (Merck: pore size 4), and washed with
are available and higher temperatures are tolerable. This
a small amount of methanol. After evaporation in vacuo, the residue was
concerns not only yield and reaction time but also the purchase
diluted with 1 mL of methanol/phosphate buffer and injected on a semi-
of suitable compounds as well as product purification.
preparative HPLC system (Gemini 5 m RP18 A110, 250ꢀ 10 mm, 4 mL/min,
The intermediate 2 is the common key structure for all final
radioactive and inactive compounds with the exception of those
isocratic 60:40 v/v MeOH/phosphate buffer pH 8.5). The separated
product fraction was diluted with 15 mL of water, passed through a
for direct labeling syntheses. Isolating the hydrochloride product
by precipitating the salt from methanol and liberation of the free
SepPak C18 cartridge, washed with water (5 mL), and dried in a stream
of argon. The cartridge was eluted with 4 mL of diethylether, and the
solvent was evaporated in vacuo (800 up to 330 mbar). For further base with Na2CO3 resulted in a pure white solid. Alternatively,
in vivo autoradiography studies, [18F]7c was formulated in 300 mL of
ethanol/saline (5:1).
the product was obtained from the supernatant by flash
chromatography, yielding a light yellowish solid that, however,
showed no differences, neither in analytical properties nor in
its reaction behavior.
1-(2,3-Dihydrobenzo[b][1,4]dioxin-6-yl)-4-((6-[18F]fluoropyridin-3-yl)-
methyl)-piperazine [18F]7d
Method A: The precursor for direct labeling 7e (13.8 mg, 40 mmol)
dissolved in 0.5 mL of anhydrous DMSO was added to the vial containing
the [18F]fluoride complex, and the solution was stirred at 160 ꢁC for
20 min. Progress of the reaction was monitored by radio HPLC (Gemini
5 m RP18 A110, 250ꢀ 4.6 mm, 1 mL/min, isocratic 50:50:0.1 v/v/v CH3CN/
H2O/TEA). Water was added to the solution, and it was passed through
a Sep Pak C18 cartridge, washed with water, and dried with air. The
cartridge was eluted with 2 mL of acetonitrile, and the solvent was
evaporated in vacuo. The residue was diluted with 1 mL acetonitrile/
water (50/50) and injected on a semi-preparative HPLC system (Gemini
5 m RP18 A110, 250ꢀ 10 mm, 4 mL/min, isocratic 40:60:0.1 v/v/v CH3CN/
H2O/TEA).
Method B: The radiochemical synthesis of [18F]7d by reductive
amination was conducted in an identical manner as described for
[
18F]1, [18F]7a, and [18F]7b, starting from precursor 6e (5 mg, 35.5 mmol)
dissolved in 0.5 mL of anhydrous DMSO and stirred for 2 min at 110 ꢁC.
The RCY of [18F]6d was determined by radio HPLC (Gemini 5 m RP18 A110,
250ꢀ 4.6mm, 1 mL/min, isocratic 60:40:0.1v/v/v CH3CN/H2O/TEA). After
reaching the optimal yield at 2min, a solution of 2 (6.5mg, 30mmol) in
40 mL of acetic acid and 50 mL of DMSO and a solution of sodium
cyanoborohydride (4mg, 64 mmol) in 50 mL of DMSO were added
immediately, and the reaction mixture was stirred for further 8 min. After
addition of 20 mL of water, the solution was passed through a Sep Pak
C18 cartridge followed by washing with 5 mL water and drying with
air. The cartridge was eluted with 1 mL of ethanol/water (80/20) and
injected on a semi-preparative HPLC system (Gemini 5 m RP18 A110,
250ꢀ 10mm, 4 mL/min, isocratic 40:60:0.1v/v/v CH3CN/H2O/TEA). The
separated fraction was diluted with 15 mL of water, passed through a
SepPak C18 cartridge, washed with 5 mL of water, and dried in a stream
of argon. The cartridge was eluted with 4 mL of diethylether, and the
solvent was evaporated in vacuo (800 up to 330 mbar). For further
experimental use, [18F]7d was formulated in 300 mL of saline (1% Tween 80).
Scheme 2. Two considered synthesis routes for the preparation of the
intermediate 2. Reaction conditions: (a) Pd2(dba)3, P(o-tolyl)3, NaOtBu, 1-boc-
piperazine, toluene, 100 ꢁC; (b) TFA, CH2Cl2, rt; (c) C4H9NCl HCl, C2H6O2, 150 ꢁC,
12 h; (d) Na2CO3.
Results and discussion
Chemistry
The first objective was to synthesize three further derivatives of
lead structure 1, all substituted at the benzyl moiety. Therefore,
Scheme 3. General synthesis scheme of title structures and reference
compounds. Reaction conditions: (a) 2, K2CO3, KI, CH3CN, reflux, 18 h; (b) 2, CH3OH,
benzyl- and benzodioxinpiperazine moieties were separately CH3COOH, NaBH3CN, 60 ꢁC, 24 h.
J. Label Compd. Radiopharm 2013, 56 609–618
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

F. Kügler et al.
Scheme 6. Synthesis of (3-benzyloxy-4-formylphenyl)-trimethyl-ammonium
triflate (15b). Reaction conditions: (a) 1 sec-BuLi, C4H10O, ꢂ78 ꢁC, 45 min; 2. DMF,
rt, 1 h; (b) CF3SO3CH3, CH2Cl2, 40 ꢁC, 7 h.
Scheme 4. Synthesis of the precursor 9b for direct radiolabeling and its reference
compound 9a and the alternative synthesis of title compound 7d. Reaction
conditions: (a) C2O2Cl2, DMF, CH2Cl2, 50 ꢁC, 2 h; (b) 2, CH3Cl, (CH3)3 N o. C5H5N, rt,
3–24 h; (c) BH3THF, C4H8O, reflux, 24 h.
mentioned aldehydes (cf. Scheme 3). Despite an extended reaction
pathway and a low amination yield due to the high sensitivity of
6d, an overall yield of about 40% was obtained. Thus, this method
can be compared with the amide formation described earlier.
All aldehydes used as precursors for radiofluorination are listed
in Scheme 5. The aldehydes (4-formylphenyl)-trimethylammonium
triflate (12) and (4-formyl-3-methoxyphenyl)-trimethylammonium
triflate (12a) were synthesized by the well-known amination of
commercially available corresponding fluorine compounds
with dimethylamine hydrochloride and subsequent quaternization
with methyltrifluoromethane sulfonate.16,17 The benzyl protected
precursor (3-benzyloxy-4-formylphenyl)-trimethylammonium triflate
(12b) was synthesized from 2-(benzyloxy)-4-bromo-N,N-dimethylaniline
(10) (Scheme 6), which was obtained in four steps from 4-bromo-N,N-
dimethylaniline according to literature.23 The key step was the
formation of the aldehyde group from the bromo compound 11
with dimethylformamide (DMF) and sec-butyllithium. The nitro
precursors 120 and 12a0 were commercially available, whereas the
nitro-analog 12b0 was synthesized by a benzylation of 4-nitro-3-
hydroxybenzaldehyde at a temperature of 110 ꢁC following a
procedure described by Langer et al.18
A four-step route starting from 4-fluoro-3-methoxybenzaldehyde
was utilized to prepare the corresponding hydroxy and benzyloxy
reference compounds (7b, 7c) of the respective radiolabeled
compounds by a Lewis acid assisted cleavage of the methyl
group from 7a. The series of compounds obtained included
the reference compounds of the desired radiolabeled analogs
(1, 7a, 7c and 7d), the precursors for aromatic 18F-for-Cl and
18F-for-NO2 substitution (7e and 9b), and precursor compounds
12–12b and 6e suitable for a two-step radiosynthesis of [18F]1
and [18F]7a–d by reductive amination.
The coupling reaction of the starting intermediate 2 can be
performed with corresponding benzylhalides, which led in case
of 4-fluorobenzylbromide (6’) to good yields of up to 55% of 1
(cf. Scheme 3a). Reductive amination with the aldehydes 6–6e,
however, was selected as the preferred method because of the
easier synthesis and also better reproducibility. Thus, a series of
new 3-substituted 6-(4-[4-fluorobenzyl]piperazine-1-yl)benzodioxin
derivatives were obtained in 20–85% yield (cf. Scheme 3b). They
were all used as standard reference compounds with the exception
of the pyridine compound 7e, which makes the direct labeling
precursor to obtain [18F]7d.
As displayed in Scheme 4 the nitrobenzylamide 9b and its
corresponding fluorine analog 9a as standard were synthesized
from the corresponding carboxylic acids via formation of
acid chlorides. The fluoropyridine standard analog (4-(2,3-
dihydrobenzo[b][1,4]dioxin-6-yl)piperazine-1-yl)(6-fluoro-pyridine-
3-yl)methanone) (7d) was also synthesized that way with the
intention of subsequent reduction and thereby achieving a shorter
route of production. However, problems with the reduction
procedure occurred as described in the succeeding text.
Alternatively, reductive amination using 6-fluoro-pyridine-3-
carbaldehyde 6d yielded 7d as described for the other previously
Radiochemistry
Direct labeling
For direct labeling of arenes with n.c.a. [18F]fluoride, it is
principally necessary that they are activated for nucleophilic
substitution. Synthesis of n.c.a. 4-[18F]fluoro-phenylpiperazines
is therefore only practicable by labeling of the corresponding
amides followed by a radioactive reduction (cf. Scheme 7). The
nitro group was selected as the best adaptable leaving group
for an SNAr reaction when trimethylammonium salts are not
available. Trimethylammonium triflate as leaving group cannot
be generated with the compounds of interest due to the
competition of the two tertiary arylamines (dimethyl and
arylpiperazine group) with very similar nucleophilicities.
Kryptofix 2.2.2 was used as anion activator for nucleophilic 18F-
substitution in dimethyl sulfoxide (DMSO) or DMF. In DMSO,
substitution of the nitro group resulted in a radiochemical yield
(RCY) of up to 50% after 15 min. As a typical feature of direct
labeling of activated aromatic systems, adequate yields are only
Scheme 5. Radiosynthesis of [18F]1, [18F]7a, and [18F]7c–d by a two-step, one-pot
synthesis. Reaction conditions: (a) [18F]Fꢂ, K ⊂ 2.2.2., K2CO3, DMSO, 110 ꢁC; (b) 2,
NaBH3CN, DMSO, 110 ꢁC; (c) Pd(black), HCOONH4, MeOH, 60 ꢁC, 10 min.
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 609–618

F. Kügler et al.
It was expected that direct nucleophilic 18F-substitution on the
pyridine derivative 7d would proceed much easier. This is due to
the fact that pyridines themselves induce an aromatic activation
for nucleophilic substitution in ortho-position and para-position.
However, despite this and the high RCYs of 2-[18F]fluoropyridine
amounting to 94% described in the literature,24 only a poor RCY
of [18F]7d of less than 5% was obtained with the precursor 7e at
160 ꢁC in DMSO after 30 min.
Build-up synthesis
Since the first utilization of reductive amination of 18F-labeled
aldehydes by Wilson et al.,16 the method has proven convenient
to reliably produce corresponding n.c.a. benzylamines with
fluorine-18 in an aryl position. The labeling conditions were a
reaction temperature of about 130–150 ꢁC, DMSO as solvent,
and (K⊂2.2.2.)/carbonate as anion activator.25 As displayed in
Figure 2, RCYs obtained by precursors containing the
Scheme 7. Direct labeling of [18F]7d and [18F]9a and its reduction to [18F]1.
Reaction conditions: (a) [18F]Fꢂ, K ⊂ 2.2.2., K2CO3, DMSO, 160 ꢁC, 20 min; (b) [18F]Fꢂ,
K ⊂ 2.2.2., K2CO3, DMSO, 160 ꢁC, 15 min; (c) BH3THF, THF, 65 ꢁC, 15 min.
obtained with high reaction temperatures of at least 160 ꢁC. DMF trimethylammonium triflate leaving group (12–12b) reach up
cannot be used because of its lower boiling point, although it to 80% after 5–10 min and thus are about twice as high as RCYs
leads to considerably higher RCYs at equal temperatures obtained by the corresponding nitro precursors (120–12b0) at the
(cf. Figure 1). In this study, higher temperatures of DMF near its same reaction temperature (cf. Scheme 5 and Figure 2).
boiling point diminished the reproducibility of RCY.
The general procedure is to separate the labeled aldehyde by
Because of the absence of further reduction-sensitive moieties, solid phase extraction or HPLC purification and then to perform
at first, the strong reduction system LiAlH4/AlCl3 in diethylether the amination reaction in a mixture of MeOH and CH3CO2H at
or THF was applied on the amide [18F]9a. However, the RCY varied temperatures of about 60 ꢁC. As a variation of the Leuckart–
from zero to 100% after 5 min reaction time, and no reproducibility Wallach reaction, sodium cyanoborohydride is commonly
could be achieved. Furthermore, rapid hydrogen evolution chosen as a reducing agent instead of an excess of formic acid.
hampered the working performance. Using instead the BH3/THF Tian et al.26 showed that reductive amination reactions can also
complex as reducing agent led to a conversion of 42 ꢃ 4% but be performed in the presence of Kryptofix 2.2.2. and DMSO.
required 65 ꢁC and 15min reaction time. Despite an expected Labeling and amination could therefore be performed in a
overall RCY of about 21% (calculated from the 50% and 42% yields one-pot reaction in DMSO at the same temperature of 110 ꢁC,
of radiofluorination and reduction steps, respectively), the RCYs of thus facilitating remote controlled syntheses.
[18F]1 were only just above 1% after separation with a semi-
In the amination reactions performed here, it was found
preparative HPLC column. This was especially caused by the unnecessary to add methanol to the reaction mixture in contrast
threefold solid phase extractions, where each time a loss could to acetic acid that is obviously an essential reagent for the rapid
be observed as well as losses during the HPLC separation. The poor intermediate formation of the immonium ion.
RCYs consequently led to the decision to turn away from direct
labeling to applying build-up radiosynthesis.
Figure 2. Time dependence of radiochemical yield of 18F-labeling of
benzaldehyde derivatives. For comparison, RCYs obtained from nitro precursors
(120–12b0) are displayed in dashed blue lines, whereas RCYs obtained from
benzaldehyde trimethylammonium triflates (12–12b) are shown in solid red lines.
Figure 1. Time dependence of radiochemical yield of [18F]9a at different reaction
temperatures in DMSO (red) and DMF (black) by direct 18F-substitution on 9b.
6-[18F]Fluoropyridinealdehyde ([18F]6d), obtained from the chloro analog, is also
displayed in red. Reaction conditions: n.c.a. [18F]Fꢂ, [12–12b, 6e] = 20–60 mmol/L,
[K ⊂ 2.2.2.] = 40 mmol/L, [K2CO3] = 20 mmol/L, DMSO = 0.5 mL, 110 ꢁC. Data are
expressed as the mean ꢃ SD (n = 2–4). This figure is available in colour online at
wileyonlinelibrary.com/journal/jlcr
Reaction conditions: n.c.a. [ ⊂ 2.2.2.] = 40 mmol/L,
18F]Fꢂ, [9b] = 40 mmol/L, [K
[K2CO3]= 20mmol/L, solvent= 0.5mL. Data are expressed as the meanꢃ SD (n=2–3).
This figure is available in colour online at wileyonlinelibrary.com/journal/jlcr
J. Label Compd. Radiopharm 2013, 56 609–618
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

F. Kügler et al.
activating components (aldehyde and pyridine) seems to
polarize the C–F binding so strongly that even weaker
nucleophiles (every halide, hydride, or hydrogen) in low
concentrations may displace [18F]fluoride. Lower precursor
concentrations than 40 mM are therefore problematic because
deviation from the optimal time of quenching the reaction and
starting the next step has a serious influence on RCY. Despite
these findings, no comparable observation is described in
literature for similar nicotinic compounds (ortho pyridine and
para carbonyl) used for 18F-fluorination. A back reaction with
chloride as nucleophile, which is theoretically also conceivable,
would lead to an equilibrium that, however, was not observed.
This defluorination procedure was also the critical side reaction
when the amination reaction with [18F]6d was performed. Any
change of reaction parameters, however, such as reaction time,
lowering of the amine concentration or reaction temperature or
using weaker reduction systems, resulted in very poor RCYs.
For the production of the 18F-labeled benzyl protected
aldehyde compound [18F]6b, 18F-fluorination and amination
were performed as described previously. To separate the
intermediate product [18F]6b from the reaction solution,
adsorption on a polymer adsorbent LiChrolut EN cartridge
offered an excellent way because the benzyl group led to a
considerable higher lipophilicity. Thus, the aforementioned
described problems with final HPLC purification are not valid for
the hydroxyl derivative [18F]7c upon reductive debenzylation of
[18F]7b in methanol with ammonium formate and palladium black
at 60 ꢁC under inert conditions. Before this cleavage, however,
complete drying of the benzyl-intermediate is necessary by
applying a stream of argon and elution through an AluminaNW
cartridge. A water tolerating cleavage with strong acids such as
trifluoroacetic acid that additionally would possibly save one
separation step is not suitable because of a rapid degradation of
the dioxin component and of enhancing even more side reactions.
A similar degradation can be observed when cleavage of the
methoxy substituent of [18F]7a was attempted with BBr3.
Figure 3. Time dependence of the total RCY of the amination products ([18F]1 and
([18F]7a–d) related to [18F]Fꢂ (blue) as well as that related to the intermediate aldehyde
compounds ([18F]6–6d) (red). Reaction conditions: [2]=46mmol/L, NaBH3CN = 64 mmol,
DMSO = 0.6 mL, AcOH = 40 mL, 110 ꢁC. Data are expressed as the mean ꢃ SD (n= 2–3).
This figure is available in colour online at wileyonlinelibrary.com/journal/jlcr
Furthermore, fewer radioactive side products were found by
using DMSO instead of methanol. As graphically depicted in
Figure 3, a nearly quantitative conversion of the 18F-labeled
aldehyde to the tertiary amine in DMSO could be observed upon
radio-HPLC analysis. The total RCYs obtained over both reaction
steps, that is, related to initial [18F]fluoride, were about 30–50%.
Performing the one-pot reaction with DMF, which is also a
suitable solvent for radiofluorination of aromatic aldehydes, or
using lower temperatures, however, led to an extreme loss of
RCY. A general disadvantage of the one-pot synthesis, however,
is the accumulation of inactive reactants and unidentified side
products in the final reaction solution. Depending on the
respective product, HPLC purification can therefore be
obstructed. Nevertheless, after semi-preparative HPLC isolation,
radiochemical purity of all products was >98% and their molar
activities reached 60 GBq/mmol for [18F]1 and [18F]7a, and about
30–40 GBq/mmol for [18F]7c and [18F]7d.
Despite the high debenzylation yield and the good pre-
purification, the RCY of [18F]7c obtained after semi-preparative
radio HPLC separation was lower than with all other products
[18F]1 and [18F]7. Losses can be due to the additional reduction
and extraction steps and due to adsorption processes during
the cleaving step.
A higher amount of the precursor probably acted as a pseudo
carrier and diminished the rate of decomposition as well as the
rate of product formation (cf. Figure 4). The influence of both
Comparison of methods
Although direct labeling is the method of choice with short-lived
radionuclides and build-up syntheses are normally reserved for
cases when the former failed,27,28 in direct comparison of both
methods here, the latter led to better results. In both cases of
direct labeling of 9b and 7e, only RCYs of about 1–5% were
achieved at the end of synthesis (cf. Table 1), whereas build-up
syntheses led to RCYs of 9–35% as summarized in Table 2.
Table 1. Radiochemical yields of [18F]1 and [18F]7d by direct
labeling in DMSO at 160 ꢁC according to Scheme 7
Final
product
Total synthesis
time (min)
Precursor
RCY (%)
[
18F]6d with different
9b
7e
[18F]1
5 ꢃ 2
120
50
Figure 4. Time dependences of the total RCY of
[18F]7d
1 ꢃ 0.5
concentrations of precursor 6e. Data are expressed as the mean ꢃ SD (n = 2–3). This
figure is available in colour online at wileyonlinelibrary.com/journal/jlcr
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 609–618

F. Kügler et al.
Table 2. Total radiochemical yields of consecutive radiofluorination and amination to obtain [18F]1 and [18F]4a–d in a one-pot
reaction according to Scheme 5
Precursor
[18F]Fluoro-aldehyde
RCY (%)
Time* (min)
Final product
RCY (%)
Total synthesis time (min)
~ 90
12
[18F]6
75 ꢃ 5
38 ꢃ 1.5
72 ꢃ 5
30 ꢃ 5
80 ꢃ 2
48 ꢃ 7
80 ꢃ 6
10
12
10
15
5
[18F]1
35 ꢃ 5
120
12a
12a0
12b
12b0
6e
[18F]6a
[18F]6b
[18F]6d
[18F]7a
[18F]7c
[18F]7d
20 ꢃ 5
9 ꢃ 4
~ 90
~ 120
~ 80
10
2
15 ꢃ 5
*Time of radiofluorination.
Dr. D. Bier, Dr. J. Castillo Meleán, and S. Humpert, all INM-5,
FZJ, for recording the spectral data and laboratory support,
respectively.
Furthermore, the required reaction time of 120 min in case of
[18F]1 could also not match up with the 80 to 90 min of the
two-step radiofluorination/amination reaction.
This is due to the high 18F-labeling yields of the
trimethylammonium triflate benzaldehydes within short reaction
times of 2 to 10 min. Further reasons include the insensitivity
and, therefore, reproducibility of the reductive amination
reaction and its high conversion rates that make this reaction
type remarkably appropriate for n.c.a. 18F-chemistry. Because of a
possible one-pot operation, the two-step reaction is characterized
by a short total reaction time and low product losses.
Conflict of interest
The authors did not report any conflict of interest.
References
[1] D. Noaín, M. E. Avale, C. Wedemeyer, D. Calvo, M. Peper,
M. Rubinstein, Eur. J. Neurosci. 2006, 24, 2429–38.
An exception is the pyridine chloride derivative 7e because
generation of the 18F-labeled product [18F]7d could be
performed in a single step. However, 18F-fluorination of the 6-
chloronicotinic aldehyde 6e leads to many inactive side products
which complicate product isolation procedures and decrease the
yields. On the other hand, the extremely poor yield obtained
from direct labeling of the chloropyridine–methylpiperazine
compound 7e is surprising, as normally good results with
nucleophilic substitution at the ortho-position or para-position
of pyridines are achieved. Therefore, it would be worthwhile to
examine other leaving groups such as the nitro substituent in
further experiments. Using microwave irradiation, as described
beneficial in many earlier 18F-labeling procedures of pyridines,
might be a second reasonable alternative.
[2] J. N. Oak, J. Oldenhof, H. H. Van Tol, Eur. J. Pharmacol. 2000, 405, 303–27.
[3] R. P. Ebstein, Mol. Psychiatry 1997, 2, 381–384.
[4] J. Swanson, J. Oosterlaan, M. Murias, S. Schuck, P. Flodman, M. A.
Spence, M. Wasdell, Y. Ding, H. C. Chi, M. Smith, M. Mann, C. Carlson,
J. L. Kennedy, J. A. Sergeant, P. Leung, Y. P. Zhang, A. Sadeh, C. Chen,
C. K. Whalen, K. A. Babb, R. Moyzis, M. I. Posner, Proc. Natl. Acad. Sci.
2000, 97, 4754–4759.
[5] S. V. Faraone, A. E. Doyle, E. Mick, J. Biederman, Am. J. Psychiatry
2001, 158, 1052–1057.
[6] I. Z. Ben Zion, R. Tessler, L. Cohen, E. Lerer, Y. Raz, R. Bachner-
Melman, I. Gritsenko, L. Nemanov, A. H. Zohar, R. H. Belmaker,
J. Benjamin, R. P. Ebstein, Mol. Psychiatry 2006, 11, 782–786.
[7] M. R. Melis, S. Succu, M. S. Mascia, A. Argiolas, Neurosci. Lett. 2005,
379, 59–62.
[8] R. P. Ebstein, O. Novick, R. Umansky, B. Priel, Y. Osher, D. Blaine, E. R.
Bennett, L. Nemanov, M. Katz, R. H. Belmaker, Nat. Genet. 1996, 12,
78–80.
[9] J. Gelernter, H. Kranzler, E. Coccaro, L. Siever, A. New, C. L. Mulgrew,
Conclusion
Am. J. Hum. Genet. 1997, 61, 1144–1152.
[10] S. Löber, H. Hübner, W. Utz, P. J. Gmeiner, Med. Chem. 2001, 44,
2691–2694.
[11] A. Thurkauf, X. Chen, Preparation of 1-phenyl-4-benzylpiperazines as
dopamine receptor subtype specific ligands (D4). PCT Int. Appl. WO
9833784, 1998.
[12] K. J. Hodgetts, A. Kieltyka, R. Brodbeck, J. N. Tran, J. W. Wasley, A.
Thurkauf, Bioorg. Med. Chem. 2001, 9, 3207–3213.
[13] S. Sebai, M. Baciu, O. Ces, J. Clarke, V. Cunningham, R. Gunn, R. Law, X.
Mulet, C. Parker, C. Plisson, R. Templer, A. Gee, Neuroimage 2006, 31, T56.
[14] F. Kügler, W. Sihver, J. Ermert, H. Hübner, P. Gmeiner, O. Prante, H. H.
Coenen, J. Med. Chem. 2011, 54, 8343–8352.
[15] J. Ermert, S. Stüsgen, M. Lang, W. Roden, H. H. Coenen, Appl. Rad.
Isot., 2008, 66, 619–624.
[16] A. A. Wilson, R. F. Dannals, H. T. Ravert, H. N. Wagner Jr., J. Label.
Compd. Rad. 1990, 8, 1189.
[17] M. S. Haka, M. R. Kilbourn, L. Watkins, S. A. Toorongian, J. Label.
Compd. Radiopharm. 1989, 27, 823–833.
Four derivatives of 6-(4-[4-fluorobenzyl]piperazin-1-yl)benzodioxin
(1) were successfully 18F-labeled in sufficient yields by using a
reductive amination reaction of aldehydes after their previous n.
c.a. nucleophilic radiofluorination. Because of the viability of both
reaction steps without interim separation, radiosynthesis of the
typical 1-phenyl-4-benzylpiperazine backbone of D4 ligands, 18F-
labeled at the benzyl moiety, can be performed quickly and easily.
These reaction conditions that enable the possibility of convenient
one-pot radiosyntheses can be transferred to many different
derivatives of [18F]fluorobenzylpiperazine beyond those anticipated
here for imaging dopaminergic receptors. This facilitates altering
physicochemical parameters such as lipophilicity in order to achieve
variations in binding behavior of radioligands.
[18] O. Langer, F. Dollé, H. Valette, C. Halldin, F. Vaufrey, C. Fuseau, C.
Coulon, M. Ottaviani, K. Någren, M. Bottlaender, B. Maziére, C.
Crouzel, Bioorg. Med. Chem. 2001, 9, 677.
[19] S. J. Oh, K. C. Lee, S. Y. Lee, E. K. Ryu, H. Saji, Y. S. Choe, D. Y. Chi, S. E.
Kim, J. Lee, B. T. Kim, Bioorg. Med. Chem. 2004, 12, 5505.
Acknowledgements
The authors thank Dr. S. Willbold, Dr. D. Hofmann, and
J. Bachhausen, all ZCH, FZJ, as well as Dr. M. Holschbach and
J. Label Compd. Radiopharm 2013, 56 609–618
Copyright © 2013 John Wiley & Sons, Ltd.
www.jlcr.org

F. Kügler et al.
[20] S. M. Qaim, J. C. Clark, C. Crouzel, M. Guillaume, H. J. Helmecke, B. Nebeling, [24] F. Dollé, Curr. Pharm. Des. 2005, 11, 3221–3235.
V. W. Pike, G. Stöcklin, in Radiopharmaceuticals for PET (Eds.: G. Stöcklin, V. [25] C. Lemaire, P. H. Damhaut, A. Plenevaux, R. Cantineau, L. Christiaens,
W. Pike), Kluwer Academic Publishers, Dordrecht, 1993, pp. 15–26. M. Guillaume, Appl. Radiat. Isot. 1992, 43, 485.
[21] H. H. Coenen, M. Schüller, G. Stöcklin, B. Klatte, A. Knöchel, J. Label. [26] H. B. Tian, D. Z. Yin, J. L. Li, L. Zhang, C. F. Zhang, Y. X. Wang, W. Zhou,
Compd. Radiopharm. 1986, 23, 455–466.
[22] K. G. Liu, A. J. Robichaud, Tetrahedron Lett. 2005, 46, 7921.
Radiochim. Acta 2003, 91, 241–245.
[27] J. Ermert, H. H. Coenen, Curr. Rad. 2010, 3, 109–126.
[23] R. Huisgen, F. Bayerlein, W. Heydkamp, Chem. Ber. 1959, 92, 3223–3241 [28] J. Ermert, H. H. Coenen, Curr. Rad. 2010, 3, 127–160.
www.jlcr.org
Copyright © 2013 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2013, 56 609–618