Effect of aldehyde and methoxy substituents on nucleophilic aromatic substitution by [18F]fluoride

Article abstract of DOI:10.1016/j.jfluchem.2007.08.003

For a series of benzaldehydes only with a leaving group or with both a leaving group and a single methoxy substituent ¹⁸F-fluorination via nucleophilic aromatic substitution (S_NAr) was studied in DMF and Me₂SO. In general, the radiochemical yields were clearly higher in DMF than in Me₂SO. In the fluorodehalogenation reaction (leaving group: halogen = Br, Cl), extremely low radiochemical yields were observed in Me₂SO (<1%). By monitoring labeling reactions using HPLC, oxidation of the aldehyde function of the precursor was detected. Especially, 2-bromobenzaldehyde was oxidized fastest in Me₂SO (within 3 min reaction time, 90% of the precursor was consumed; radiochemical yield = 1.0 ± 0.5%); however, in DMF oxidation was always kept at a low level during the entire reaction (<5% of the precursor was oxidized; radiochemical yield = 73.0 ± 0.2%). In DMF, nitrobenzaldehydes with a methoxy substituent (methoxy group in meta-position to the nitro group) were labeled with good radiochemical yields (4-methoxy-2-nitrobenzaldehyde: 87 ± 3%; 2-methoxy-4-nitrobenzaldehyde: 83 ± 3%; 2-methoxy-6-nitrobenzaldehyde: 79 ± 4%) comparable to the non-substituted nitrobenzaldehydes (2-nitrobenzaldehyde: 84 ± 3%; 4-nitrobenzaldehyde: 81 ± 5%). Moreover, for structurally similar compounds, radiochemical yields showed a good correlation with ¹³C-NMR ppm values of the aromatic carbon atom bearing the leaving group.

Full text of DOI:10.1016/j.jfluchem.2007.08.003

Journal of Fluorine Chemistry 128 (2007) 1461–1468
www.elsevier.com/locate/fluor
Effect of aldehyde and methoxy substituents on nucleophilic aromatic
substitution by [18F]fluoride
a
Bin Shen , Dirk Loffler , Klaus-Peter Zeller , Michael Ubele
a
b
a,ä
¨
¨
,
Gerald Reischl a, , Hans-Ju¨rgen Machulla
*
a
a
¨
Radiopharmacy, PET-Center, Eberhard Karls University Tubingen, Germany
¨
Institute of Organic Chemistry, Eberhard Karls University Tubingen, Germany
b
Received 4 April 2007; received in revised form 3 August 2007; accepted 6 August 2007
Available online 10 August 2007
Abstract
For a series of benzaldehydes only with a leaving group or with both a leaving group and a single methoxy substituent 18F-fluorination via
nucleophilic aromatic substitution (SNAr) was studied in DMF and Me2SO. In general, the radiochemical yields were clearly higher in DMF than in
Me2SO. In the fluorodehalogenation reaction (leaving group: halogen = Br, Cl), extremely low radiochemical yields were observed in Me2SO
(<1%). By monitoring labeling reactions using HPLC, oxidation of the aldehyde function of the precursor was detected. Especially, 2-
bromobenzaldehyde was oxidized fastest in Me2SO (within 3 min reaction time, 90% of the precursor was consumed; radiochemical
yield = 1.0 ꢀ 0.5%); however, in DMF oxidation was always kept at a low level during the entire reaction (<5% of the precursor was oxidized;
radiochemical yield = 73.0 ꢀ 0.2%). In DMF, nitrobenzaldehydes with a methoxy substituent (methoxy group in meta-position to the nitro group)
were labeled with good radiochemical yields (4-methoxy-2-nitrobenzaldehyde: 87 ꢀ 3%; 2-methoxy-4-nitrobenzaldehyde: 83 ꢀ 3%; 2-methoxy-
6-nitrobenzaldehyde: 79 ꢀ 4%) comparable to the non-substituted nitrobenzaldehydes (2-nitrobenzaldehyde: 84 ꢀ 3%; 4-nitrobenzaldehyde:
81 ꢀ 5%). Moreover, for structurally similar compounds, radiochemical yields showed a good correlation with 13C-NMR ppm values of the
aromatic carbon atom bearing the leaving group.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Nucleophilic aromatic substitution; Benzaldehydes; Aldehyde oxidation; 18F-labeled aromatic amino acids
1. Introduction
yield. By the strategy of SNAr, a direct reaction with
[18F]fluoride can be realized easily, if a substituent is present
that reduces the electron density of the benzene ring. The
aldehyde group appears to be particularly well suited as an
auxiliary substituent because it can either be removed
catalytically or transferred into a different functional group
for further synthetic steps. In the case of tyrosines, the hydroxy
group has to be protected when performing nucleophilic
substitutions with [18F]fluoride in order to avoid solvation
effects resulting in loss of the nucleoplilic properties of
[18F]fluoride, a common problem when performing reactions in
the typically low radiochemical concentrations. Therefore,
SNAr might be realized in presence of a methoxy group in
syntheses of tyrosines. However, it needs to be proven since the
methoxy group increases the electron density on the benzene
ring and generally decreases the radiochemical yield. Pursuing
earlier research on the 18F-labeling of benzaldehyde derivatives
[6–14], we evaluated the contrary influence of the aldehyde
In PET and recent PET/MRI studies, new interest in 18F-
labeled aromatic amino acids is apparent [1,2]. With regard to
the preparation of amino acids as no-carrier-added (nca) 18F-
tracers with high specific activities, nucleophilic aromatic
substitution (SNAr) is to be considered the most attractive type
of reaction since fluorine-18 can be applied as [18F]fluoride,
which is easily and efficiently produced at a medical cyclotron.
As an example of [18F]amino acid, [18F]FDOPA can be
prepared in a multi-step synthesis with the radionuclide being
introduced in the first reaction step via SNAr and the amino acid
function is added afterwards [3–5]. Thus, high yields in the
labeling reaction are necessary for a good overall radiochemical
* Corresponding author. Tel.: +49 7071 2987443; fax: +49 7071 295264.
E-mail address: gerald.reischl@uni-tuebingen.de (G. Reischl).
ä
Deceased.
0022-1139/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2007.08.003

1462
B. Shen et al. / Journal of Fluorine Chemistry 128 (2007) 1461–1468
and the methoxy groups on the labeling yield in case of SNAr
reactions.
2.2. Benzaldehyde derivatives containing a leaving group
With respect to DMF as labeling solvent, in Me2SO the
extent of decrease of radiochemical yield was strongly
dependent on the leaving group. With the fluoro or nitro group
as substituent (compounds 1a, 2a, 1b, 2b), the radiochemical
yield was reduced from 75–82% in DMF to 40–68% in Me2SO.
In case of the chloro- or bromo-substituent (compounds 1c, 2c,
1d, 2d), in Me2SO the radiochemical yield decreased to less
than 2% whereas yields of 65–75% were obtained in DMF
(Table 1).
2. Results and discussion
2.1. General
In this study, all compounds were labeled in DMF and
Me2SO at 140 8C with 10 mg of precursor. In previously
reported results from our group [15,16], in general, the
radiochemical yield showed a strong dependence on the type of
solvent. This observation was consistent with findings of this
study (Tables 1 and 2). Radiochemical yields were higher in
DMF than in Me2SO which had usually been used by other
groups as solvent for SNAr reactions.
For a better understanding, we examined the labeling
reaction in DMF and Me2SO in the case of ortho-bromo-
benzaldehyde (1d) versus a possible effect of oxidation.
HPLC analyses showed the formation of the correspondingly
Table 1
Maximum radiochemical yields for 18F-labeling of benzaldehydes only with a leaving group
Leaving group
Radiochemical yielda (%)
Radiochemical yielda (%)
DMF
Me2SO
DMF
Me2SO
This study
Literature
This study
Literature
F (a)
75 ꢀ 1
64 ꢀ 4
68.0 ꢀ 0.4
0.6 ꢀ 0.3
1.0 ꢀ 0.5
–
78.0 ꢀ 0.5
81 ꢀ 5
56 ꢀ 4
–
NO2 (b)
Cl (c)
Br (d)
84 ꢀ 3
65 ꢀ 1
53–78 [6ꢁ12]
40 ꢀ 3
53–75 [6,8–13]
1–6 [12,13]
3–7 [12,13]
–
–
67 ꢀ 3
1.0 ꢀ 0.8
0.8 ꢀ 0.6
73.0 ꢀ 0.2
75.0 ꢀ 0.5
a
Maximum radiochemical yield at 20 min.
Table 2
Maximum radiochemical yields for 18F-labeling of single methoxy-substituted nitrobenzaldehydes
Radiochemical yield (%)
DMF
Me2SO
This study
Literature
3a (R1 = OMe)a
3b (R2 = OMe)a
3c (R3 = OMe)b
3d (R4 = OMe)b
4a (R1 = OMe)a
4b (R2 = OMe)a
67 ꢀ 3
87 ꢀ 3
55 ꢀ 5
79 ꢀ 4
75 ꢀ 1
83 ꢀ 3
6.0 ꢀ 0.6
74 ꢀ 4
17–23 [7]
75–87 [7,9]
5 [7]
4.0 ꢀ 0.7
41.0 ꢀ 0.6
23 ꢀ 1
–
29–55 [7,9,10]
74 [10]
75 ꢀ 8
a
Maximum radiochemical yield at 10 min.
b
Maximum radiochemical yield at 20 min.

B. Shen et al. / Journal of Fluorine Chemistry 128 (2007) 1461–1468
1463
Fig. 1. Oxidation of ortho-bromobenzaldehyde (1d) to ortho-bromobenzoic
acid during labeling reactions in DMF or Me2SO.
Fig. 3. Conversion of para-nitrobenzaldehyde (2b) during labeling reactions in
DMF or Me2SO.
Me2SO resulting in low labeling yields. Interestingly, no
labeled benzoic acid derivatives were observed.
The effect of the carbonyl group para or ortho to the leaving
group should be similar, when substituent effects are
considered on the basis of mesomeric interactions. In DMF
ortho- and para-isomers were similarly substituted by
[18F]fluoride independent of the leaving group (Table 1). In
Me2SO, no differences between ortho- and para-isomers were
noted except in the case of the nitro group as leaving group. The
ortho-isomer (1b) showed a radiochemical yield of 68 ꢀ 0.4%,
the para-isomer (2b) only 40 ꢀ 3%. For 1b and 2b, the labeling
reactions in both solvents were also analyzed by HPLC.
In DMF, the formation of oxidation products was always
much slower and the concentration of benzoic acid remained
below 5% (Figs. 2 and 3) and good labeling yields i.e. 84 ꢀ 3%
and 81 ꢀ 5% for ortho- and para-nitrobenzaldehyde were
observed.
Fig. 2. Formation of ortho-nitrobenzoic acid by oxidation of ortho-nitroben-
zaldehyde (1b) in labeling reactions in DMF or Me2SO.
In Me2SO, almost 30% of 1b was oxidized to benzoic acid
within 30 min, the oxidation proceeded faster than in DMF,
without giving rise to any other byproducts (Fig. 2). Therefore,
radiochemical yield (68 ꢀ 0.4%) was lower than in DMF
(84 ꢀ 3%). Labeling of 2b in Me2SO was accompanied by a
high degree of oxidation. After 30 min there was almost no
precursor left due to the conversion to benzoic acid (ca. 50% of
substituted benzoic acid as result of the oxidation of the
carbonyl group (Fig. 1). In Me2SO more than 90% of 1d was
oxidized within 3 min. In DMF less than 5% of benzoic acid
was formed and after 1 min the amount of benzoic acid did not
change. These results show oxidation proceeds rapidly in
Fig. 4. Suggested mechanism for the oxidation of substituted benzaldehydes in Me2SO.

1464
B. Shen et al. / Journal of Fluorine Chemistry 128 (2007) 1461–1468
expected to decrease nucleophilic substitution at the benzene
ring. However, methoxy-substituted benzaldehydes are impor-
tant intermediates in the radiochemical synthesis of amino
acids. Therefore, different nitrobenzaldehydes with a single
methoxy group in different positions were studied (Table 2). In
spite of the strong electron donating effect of the methoxy
group, all radiochemical yields observed in DMF were higher
than 55 ꢀ 5%. With the methoxy group para to the nitro
substituent and meta to the aldehyde (3c), the lowest
substitution yield (55 ꢀ 5%) was obtained. When the methoxy
group was meta to the nitro group and ortho or para to the
carbonyl group (3b, 4b, 3d), even better yields (3b = 87 ꢀ 3%,
4b = 83 ꢀ 3% and 3d = 79 ꢀ 4%) were found compared to 2-
or 4-nitrobenzaldehyde (1b = 84 ꢀ 3% and 2b = 81 ꢀ 5%).
In the methoxy-substituted substrates 3 and 4 (Table 2) the
yield decrease in Me2SO was highest for compounds 3a, 3c and
4a (methoxy group meta to the aldehyde) and less expressed in
compounds 3b, 3d and 4b (methoxy ortho- or para-oriented to
the aldehyde). As discussed above for fluorobenzaldehyde, in
the latter compounds direct mesomeric resonance between the
methoxy and the carbonyl substituent reduces the electro-
philicity of the aldehyde carbon and decreases oxidation of the
benzaldehydes by Me2SO (Fig. 4).
Fig. 5. Correlation of radiochemical yields (RCY) in DMF and Me2SO with 13
C
NMR chemical shifts of the aromatic carbon atom bearing the leaving group.
the precursor) and an unidentified byproduct (Fig. 3), but
radiochemical yield (40 ꢀ 3%) did not decrease significantly.
Neither benzoic acid nor byproduct were detected as radio-
actively labeled products.
The small amount of oxidation of bromo- and nitrobenzal-
dehyde (Figs. 1–3) in DMF observed at the beginning of the
reaction may be due to dissolved oxygen. The high oxidation
rates observed in Me2SO, however, seem to be the result of an
oxidation process mediated by this solvent. Me2SO is known to
be a superior reagent for oxidation of alcohols to carbonyl
compounds when it is activated by an electrophile [17–19]. The
oxidation of benzaldehydes substituted with electron-with-
drawing groups by Me2SO in the presence of K2CO3 can be
rationalized as an electrophilic addition of the aldehyde to the
nucleophilic oxygen in Me2SO, followed by deprotonation of
the resulting sulfonium intermediate. The sulfur ylide, thus
formed, may finally decompose under internal hydrogen
abstraction into the anion of the benzoic acid and dimethyl-
sulfide, a mechanism as found similarly in the Swern oxidation
[18] (Fig. 4). In this study, benzoic acid was detected by HPLC
and dimethylsulfide by GC–MS analysis.
2.4. Correlation of radiochemical yields with 13C NMR
chemical shift values
Following earlier findings [7,10] establishing the correlation
between radiochemical yield and the 13C NMR chemical shift
of the aromatic carbon atom bearing the leaving group, we
studied this in both solvents. Good correlations were found for
all nitro compounds (Fig. 5, regression line y = 1.78x ꢁ 185;
r2 = 0.67 for DMF; regression line y = 5.45x ꢁ 763; r2 = 0.82
for Me2SO). The precursors with higher ppm values for the
aromatic carbon bearing the leaving group carry a lower
electron density on this reaction center and are more active for
nucleophilic substitution.
3. Conclusion
In the case of the isomers 1 and 2 the substitution by
[18F]fluoride is reduced because the Cl- and Br-group are poorer
leaving groups (Table 1). In these cases, the competing oxidation
prevails. As already pointed out, the radiolabeled benzaldehydes
formed by nucleophilic displacement by [18F]fluoride do not
suffer from subsequent oxidation by Me2SO. In all cases the
fluorine is introduced either ortho or para to the aldehyde group.
Despite itsstrongelectron-withdrawingcharacter, fluorineisalso
a strong +M-donor. This mesomeric effect lowers the electro-
philic character of the aldehydic carbon and may therefore
protect the aldehyde from oxidation.
DMF is superior to Me2SO as solvent in the labeling reaction
for benzaldehydes by SNAr. Benzaldehydes with bromo- or
chloro-substituents as leaving group give very low radio-
chemical yields in Me2SO due to fast oxidation. With one
methoxy substituent at the nitrobenzaldehyde highest radio-
chemical yields are obtained in DMF when the methoxy group
is meta to the nitro group and in ortho-, para-position to the
aldehyde group.
4. Experimental
4.1. General
2.3. Benzaldehyde derivatives with a leaving group and an
additional methoxy substituent
For performing the labeling reactions, Me2SO and DMF as
solvents (stored over molecular sieve) were purchased from
Fluka, Germany. Acetonitrile (for DNA synthesis) and
Kryptofix 222 were obtained from Merck (Darmstadt,
The methoxy group is an electron donating substituent
(strong +M effect) and thus will not favor SNAr. It is to be

B. Shen et al. / Journal of Fluorine Chemistry 128 (2007) 1461–1468
1465
Germany). For the MPLC system (Bu¨chi, Switzerland) silica
gel 60 (0.040–0.063 mm; Merck) was used, eluents were
mixtures of petrolether (60/90) and ethylacetate. Precursors and
reference standards were characterized by their melting point
(Gallenkamp MPG 350, Germany), IR (Spectrum One FT-
ATR-IR, Perkin–Elmer, Boston, USA), MS (Finnigan MAT
TSQ 70, Bremen, Germany) and NMR (Bruker Avance 400,
Rheinstetten, Germany). For NMR measurements, as internal
standards the deuterated solvents were used (DMSO-d6:
d = 2.49 in H and d = 39.5 in 13C; CDCl3: d = 7.25 in H
and d = 77.0 in 13C). The chemical shift d in ppm was referred to
the internal standard. All radiochemical yields given in this
work represent an average of three to five experiments.
IR (n¯, cmꢁ1): 3104 (85%), 3066 (82%), 2945 (84%), 2841
(85%), 1679 (56%), 1600 (59%), 1564 (77%), 1524 (41%),
1500 (56%), 1458 (74%), 1435 (70%), 1403 (79%), 1353
(52%), 1320 (60%), 1282 (54%), 1241 (38%), 1200 (65%),
1184 (52%), 1140 (59%), 1067 (48%), 1025 (44%), 963 (78%),
910 (56%), 897 (47%), 825 (29%), 775 (57%), 758 (54%), 698
(76%), 675 (73%), 655 (62%).
MS-EI (70 eV): m/z: 181 ([M]ꢂ+, C8H7NO4, 4%); 151
([M ꢁ NO]+, C8H7O3, 100%); 134 ([M ꢁ NO ꢁ OH]+,
C8H6O2, 21%); 119 (4%); 108 (18%); 106 (26%); 95 (22%);
92 (12%); 77 ([C6H5]+, 8%); 63 ([C5H3]+, 23%); 50 (3%).
HRMS (EI): m/z measured: 181.03411 ([M]ꢂ+, C8H7NO4);
m/z calculated: 181.037486.
1
1
4.2. Precursors and reference standards
4.2.2. 5-Methoxy-2-nitrobenzaldehyde (3c)
To a stirred suspension of 5-hydroxy-2-nitrobenzaldehyde
(3.99 g, 24 mmol) and K2CO3 (3.18 g, 24 mmol) in DMF
(40 mL), methyliodide (1.49 mL, 24 mmol) was added drop-
wise at 0 8C. After the addition, the suspension was stirred at
room temperature. The reaction was monitored by TLC. After
3 h, the starting material was completely reacted. The reaction
mixture was poured into water and extracted with 3ꢃ 50 mL of
diethylether. The combined organic layers were washed with
100 mL of water and dried over Na2SO4. After evaporation of
the solvent, the residue was purified by MPLC using
petrolether:ethylacetate (1:1) to yield 3.77 g 3c as yellow
crystals (87%).
As precursors and reference standards, the following
compounds of the highest purity available were from either
Sigma–Aldrich, Fluka, ABCR, Alfa Aesar, AVOCADO or
Fluorochem and were used as received (authenticity of the
fluoro compounds used as standards was confirmed by NMR
and MS): 2-fluorobenzaldehyde (1a; purity 97%), 4-fluoro-
benzaldehyde (2a; >98%), 2-nitrobenzaldehyde (1b; 98%), 4-
nitrobenzaldehyde (2b; 98%), 2-chlorobenzaldehyde (1c;
99%), 4-chlorobenzaldehyde (2c; 97%), 2-bromobenzaldehyde
(1d; >98%), 4-bromobenzaldehyde (2d; >97%), 3-methoxy-2-
nitrobenzaldehyde (3a; >97%), 4-nitro-2-methoxybenzalde-
hyde (4b; 97%), 2-fluoro-3-methoxybenzaldehyde (97%), 2-
fluoro-4-methoxybenzaldehyde (98%), 2-fluoro-5-methoxybe-
nzaldehyde (98%), 2-fluoro-6-methoxybenzaldehyde (98%), 4-
fluoro-3-methoxybenzaldehyde (98%), 4-fluoro-2-methoxy-
benzaldehyde (98%), 2-bromobenzoic acid (98%), 2-nitroben-
zoic acid (98%) and 4-nitrobenzoic acid (>98%).
mp: 81–82 8C (83 8C [21]).
1H NMR (CDCl3, 250 MHz): d = 10.46 (s, 1H, –CHO), 8.13
(d, 1H, J = 9.1 Hz, Harom), 7.30 (s, 1H, Harom), 7.13 (d, 1H,
J = 9.1 Hz, Harom), 3.94 (s, 3H, –OCH3). 13C NMR (CDCl3,
60 MHz): d = 55.7, 113.5, 118.9, 127.6, 134.6, 142.6, 164.3,
188.8.
All other precursors in Table 2 were synthesized as follows.
IR (n¯, cmꢁ1): 3102 (81%), 2982 (82%), 2946 (79%), 2845
(82%), 2666 (83%), 2166 (87%), 1922 (87%), 1689 (37%),
1609 (77%), 1580 (29%), 1499 (31%), 1479 (47%), 1459
(61%), 1442 (47%), 1419 (51%), 1390 (45%), 1326 (24%),
1286 (25%), 1233 (29%), 1187 (41%), 1159 (33%), 1079
(39%), 1024 (38%), 1001 (48%), 963 (66%), 929 (27%), 892
(28%), 847 (31%), 834 (22%), 771 (52%), 741 (23%), 688
(42%), 667 (45%).
4.2.1. 4-Methoxy-2-nitrobenzaldehyde (3b)
4-Bromo-3-nitroanisole (1.75 g, 7.5 mmol) was dissolved
in anhydrous THF (80 mL) under argon. The solution was
cooled to ꢁ78 8C, and phenyl lithium (4 mL, 7.5 mmol) was
added dropwise. The temperature was kept between ꢁ70 and
ꢁ90 8C for 2 h. Then, DMF (1.5 mL, 20 mmol) was added
dropwise and the reaction solution was stirred between ꢁ70
and ꢁ90 8C for an additional hour. After warming to 0 8C
within 1 h the solution was quenched by adding acetic acid
(100%; 5 mL). The resulting mixture was partitioned between
100 mL of diethylether and 50 mL of water. The organic layer
was washed with 50 mL of saturated NaHCO3 solution, 50 mL
of water and 50 mL of brine, dried over Na2SO4 and
concentrated to afford a crude product, which was purified
by MPLC (petrolether:ethylacetate, 9:1) to yield 0.9 g 3b as
yellow crystals (65%).
MS-EI (70 eV): m/z: 181 ([M]ꢂ+, C8H7NO4, 10%); 180
([M ꢁ H]+, C8H6NO4, 29%); 163 ([M ꢁ H2O]ꢂ+, C8H5NO3,
5%); 151 ([M ꢁ NO]+, C8H7O3, 41%); 133 ([M ꢁ H2O
ꢁ NO]ꢂ+, C8H5O2, 21%); 123 (32%); 108 (65%); 95 (54%);
77 ([C6H5]+, 24%); 64 (37%); 63 ([C5H3]+, 100%); 52
([C4H4]ꢂ+, 21%); 49 (14%).
HRMS (EI): m/z measured: 181.03488 ([M]ꢂ+, C8H7NO4);
m/z calculated: 181.037486.
4.2.3. 2-Methoxy-6-nitrobenzaldehyde (3d)
mp: 95–96 8C (96–97 8C [20]).
To a stirred solution of hexamethylenetetramine (3.63 g,
25.92 mmol) in trifluoro acetic acid (35 mL) 3-nitrophenol
(3 g, 21.6 mmol) was added. The solution was refluxed at
130 8C under Ar. After 27 h, TLC showed complete conversion
of the starting material. The reaction solution was cooled down
and quenched by adding 6N HCl (120 mL). After 20 min of
1H NMR (CDCl3, 400 MHz): d = 10.28 (s, 1H, –CHO), 7.96
(d, 1H, J = 8.5 Hz, Harom), 7.50 (d, 1H, J = 2.6 Hz, Harom), 7.22
(d, 1H, J = 8.8 Hz, Harom), 3.95 (s, 3H, –OCH3). 13C NMR
(CDCl3, 100 MHz): d = 56.3, 109.6, 119.1, 123.4, 131.5, 151.6,
163.7, 187.0.

1466
B. Shen et al. / Journal of Fluorine Chemistry 128 (2007) 1461–1468
stirring, the resulting solution was extracted with 2ꢃ 100 mL of
dichloromethane. The combined organic layers were washed
with 2ꢃ 50 mL of saturated NaHCO3 solution, 50 mL of
saturated NaCl and 50 mL of water and dried over Na2SO4.
After evaporation of the solvent, 1.5 g of crude 2-hydroxy-6-
nitrobenzaldehyde were obtained. This product was used
directly in the following synthetic step.
IR (n¯, cmꢁ1): 3108 (87%), 3047 (87%), 2868 (85%), 2756
(87%), 1699 (65%), 1607 (59%), 1520 (48%), 1488 (63%),
1467 (67%), 1422 (79%), 1389 (60%), 1361 (58%), 1316
(61%), 1272 (51%), 1189 (78%), 1162 (61%), 1145 (62%),
1096 (71%), 1028 (56%), 1006 (72%), 962 (65%), 869 (64%),
841 (49%), 826 (45%), 760 (60%), 737 (47%), 703 (70%), 679
(74%).
MS-EI (70 eV): m/z: 181 ([M]ꢂ+, C8H7NO4, 100%); 163
([M ꢁ H2O]ꢂ+, C8H5NO3, 2%); 151 ([M ꢁ NO]+, C8H7O3,
68%); 134 ([M ꢁ NO ꢁ OH]+, C8H6O2, 64%); 119 (54%); 104
(10%); 95 (23%); 91 (12%); 77 ([C6H5]+, 34%); 63 ([C5H3]+,
29%); 51 ([C4H3]+, 22%); 50 (15%); 39 ([C3H3]+, 11%).
HRMS (EI): m/z measured: 181.03632 ([M]ꢂ+, C8H7NO4);
m/z calculated: 181.037486.
2-hydroxy-6-nitrobenzaldehyde was methylated to 3d by the
same method as used for the synthesis of 3c: 2-hydroxy-6-
nitrobenzaldehyde (1.45 g, 9.0 mmol), K2CO3 (1.38 g,
10 mmol) and methyliodide (0.62 mL, 10 mmol) in DMF
(17 mL) were stirred at room temperature for 3 h, two products
were obtained after purification on MPLC (petrolether:ethy-
lacetate, 1:1): 0.33 g of 2-methoxy-4-nitrobenzaldehyde (8.6%,
based on 3-nitrophenol) and 0.59 g of 3d (15.8%, based on 3-
nitrophenol).
4.3. Production of [18F]fluoride
3d:
mp: 108–109 8C (110–111 8C [22]).
No-carrier-added (nca) [18F]fluoride was produced at the
PETtrace cyclotron (General Electric Healthcare, Uppsala,
Sweden) via the 18O(p,n)18F nuclear reaction by irradiating
1.5 mL of >95% enriched [18O]water (Rotem, Israel) with 16.5
MeV protons.
1H NMR (CDCl3, 400 MHz): d = 10.36 (s, 1H, –CHO), 7.59
(t, 1H, J = 8.3 Hz, Harom), 7.42 (d, 1H, J = 8.1 Hz, Harom), 7.22
(d, 1H, J = 8.3 Hz, Harom), 3.94 (s, 3H, –OCH3). 13C NMR
(CDCl3, 100 MHz): d = 55.7, 115.5, 116.0 120.5, 133.6, 148.7,
159.7, 187.6.
IR (n¯, cmꢁ1): 3092 (86%), 3034 (87%), 2954 (86%), 2879
(82%), 1959 (88%), 1706 (42%), 1609 (67%), 1527 (34%),
1468 (50%), 1437 (59%), 1388 (73%), 1353 (42%), 1309
(62%), 1269 (37%), 1208 (57%), 1190 (58%), 1162 (51%),
1048 (38%), 971 (62%), 900 (58%), 833 (50%), 804 (41%), 792
(33%), 733 (30%).
4.4. Labeling
The radioactivity was introduced into a 5.0 mL sealed vial
containing 100 mL of 3.5% aqueous K2CO3 and 15.0 mg
Kryptofix 222. The [18F]fluoride solution was dried for 20 min
under a mild stream of argon (ca. 2 mL minꢁ1) at 140 8C by
azeotropic distillation with acetonitrile (2ꢃ 1 mL). Then,
10 mg of precursor in 1.0 mL of solvent were added into the
vial containing the [Kryptofix 222] K18F complex. The sealed
vial was heated to 140 8C. TLC samples were withdrawn (1–
5 mL) for determination of the radiochemical yield at 1, 3, 7, 10,
20 and 30 min.
MS-EI (70 eV): m/z: 181 ([M]ꢂ+, C8H7NO4, 10%); 163
([M ꢁ H2O]ꢂ+, C8H5NO3, 10%); 151 ([M ꢁ NO]+, C8H7O3,
100%); 136 ([M ꢁ COOH]+, C7H6NO2, 66%); 133
([M ꢁ H2O ꢁ NO]+, C8H5O2, 17%); 119 (11%); 108 (41%);
92 (18%); 77 ([C6H5]+, 27%); 76 (38%); 63 ([C5H3]+, 19%); 50
(6%); 43 (5%).
HRMS (EI): m/z measured: 181.03436 ([M]ꢂ+, C8H7NO4);
m/z calculated: 181.037486.
4.5. Analytical assay
4.2.4. 3-Methoxy-4-nitrobenzaldehyde (4a)
4.5.1. TLC analysis
To the stirred suspension of 3-methoxy-4-nitrobenzylalcohol
(1.13 g, 6.2 mmol) and sodium acetate (1.1 g) in dichloro-
methane (30 mL) at 0 8C, pyridinium dichromate (2.32 g,
17.1 mmol) was introduced slowly over 15 min. After 2 h at
room temperature the mixture was filtered and dichloromethane
evaporated. To the residue 100 mL of diethylether were added
and the mixture was treated in an ultrasonic bath. The
suspension was filtered and the filtrate washed with 50 mL of
1N HCl to remove the pyridinium salt. The diethylether
solution was dried over Na2SO4, evaporated, and the residue
was purified by MPLC (petrolether:ethylacetate, 2:1), yielding
0.55 g of 4a (49%) as yellow crystals.
An aliquot of the reaction solution on a silica gel
plate (Polygram1 Silica G/UV254, 8 cm ꢃ 4 cm, Macherey&-
Nagel, Germany) was developed with petrolether/ethylacetate
(3/1, v/v). The radioactive spots were quantitatively assessed by
means of electronic autoradiography (InstantImager, Canberra
Packard, USA). The size of the TLC plate and the location of
the reference standard (from UV detection; Rf values are
presented in Table 3) were marked by radioactive spots on the
plate, thus a correlation between radioactive labeled product
and non-radioactive standards was assured.
4.5.2. HPLC analysis
mp: 105–107 8C (104–105 8C [23]).
HPLC was applied for identification of labeled products and
for monitoring of side reactions during the labeling reaction.
HPLC was carried out by means of a Hewlett-Packard Model
1050 equipped with a NaI(Tl)-scintillation detector and an UV
detector (280 or 254 nm) in series. Two columns with different
selectivities were used to assure identity of products. Analysis
1H NMR (Me2SO-d6, 400 MHz): d = 10.07 (s, 1H, –CHO),
8.05 (d, 1H, J = 8.1 Hz, Harom), 7.81 (s, 1H, Harom), 7.64 (d, 1H,
J = 8.3 Hz, Harom), 3.59 (s, 3H, –OCH3). 13C NMR (Me2SO-d6,
100 MHz): d = 57.0, 114.2, 121.8, 125.4, 139.6, 142.7, 151.9,
192.1.

B. Shen et al. / Journal of Fluorine Chemistry 128 (2007) 1461–1468
1467
Table 3
Rf, Rt and k0 for precursors and standards, determined by two different HPLC methods
Compound
Rf
HPLC method A
HPLC method B
Rt (min)
k0
Rt (min)
k0
1a
2a
1b
2b
1c
2c
1d
2d
3a
3b
3c
3d
4a
4b
0.69
0.63
0.51
0.51
0.71
0.66
0.72
0.65
0.24
0.42
0.51
0.21
0.55
0.39
0.61
0.56
0.44
0.68
0.61
0.59
–
7.49a
7.21a
6.68a
6.65a
10.88a
10.38a
11.43a
11.58a
7.25b
8.53b
8.71b
7.42b
8.77b
7.56b
8.22b
7.69b
6.44b
8.82b
7.94b
7.75b
4.24a
4.85a
5.13a
2.94
2.79
2.52
2.5
7.62b
4.04
4.00
–
7.56b
–
–
–
4.73
4.47
5.02
5.1
–
–
–
–
–
–
–
–
2.82
3.49
3.58
2.91
3.62
2.98
3.33
3.05
2.39
3.64
3.18
3.08
1.23
1.55
1.7
–
–
–
–
–
–
–
–
–
–
–
–
4-Fluoro-3-methoxybenzaldehyde
4-Fluoro-2-methoxybenzaldehyde
2-Fluoro-6-methoxybenzaldehyde
2-Fluoro-5-methoxybenzaldehyde
2-Fluoro-4-methoxybenzaldehyde
2-Fluoro-3-methoxybenzaldehyde
2-Nitrobenzoic acid
4-Nitrobenzoic acid
2-Bromobenzoic acid
9.97b
10.22b
7.15b
11.15b
9.62b
9.69b
–
5.60
5.76
3.73
6.38
5.37
5.41
–
–
–
–
–
–
-
Method A: C18 column (Luna, 5 mm, 250 mm ꢃ 4.6 mm, Phenomenex); flow: 1 mL minꢁ1; eluent: acetonitrile/water (50/50, v/v). Method B: Phenyl-hexyl column
(Luna, 5 mm, 250 mm ꢃ 4.6 mm, Phenomenex); flow: 2 mL minꢁ1; eluent: acetonitrile/water (30/70, v/v).
a
UV at 280 nm.
b
UV at 254 nm.
was performed in two runs. First, the reaction mixture was
analyzed and in a second run the identity of the product was
verified by co-injection of the standard with the reaction
mixture. Retention times of product and byproduct (i.e. benzoic
acid) were in agreement with the UV peaks of the reference
compounds, Rt and k0 values are presented in Table 3.
precursors were used to calibrate the integration area of the UV
peak for normalization.
4.5.3. GC–MS analysis
The dimethylsulfide formed by oxidation was analyzed by
means of GC–MS (Hewlett-Packard 5890 Series II Plus) by
taking samples from the gas phase above the reaction solution
at 30 min reaction time. Column: Chrompack Poraplot Q
(27.5 m ꢃ 0.32 mm); He-flow: 9.6 mL minꢁ1; temperature
program: 10 8C minꢁ1 from 85 to 200 8C. Rt (dimethylsulfide):
8.29 min.
Method A
A C18 column (Luna, 5 mm, 250 mm ꢃ 4.6 mm, Phenom-
enex, USA) was used with a flow rate of 1 mL minꢁ1. Eluent
was acetonitrile/water (50/50, v/v).
Method B
References
A phenyl-hexyl column (Luna, 5 mm, 250 mm ꢃ 4.6 mm,
Phenomenex, USA) was used with a flow rate of 2 mL minꢁ1
Eluent was acetonitrile/water (30/70, v/v).
.
[1] F.W. Floeth, D. Pauleit, H.J. Wittsack, K.J. Langen, G. Reifenberger, K.
Hamacher, M. Messing-Junger, K. Zilles, F. Weber, W. Stummer, H.J.
Steiger, G. Woebker, H.W. Muller, H. Coenen, M. Sabel, J. Neurosurg. 102
(2005) 318–327.
For product purity control, an aliquot of the reaction mixture
(at 10 or 20 min) was injected onto HPLC. After separation, the
product fraction was collected and measured by means of a
gamma-counter (1480 Wallac WIZARD 300, Perkin–Elmer,
USA). The radiochemical product yield was calculated by
relating the radioactivity of the product peak with the amount of
radioactivity injected onto the column.
[2] D. Pauleit, F. Floeth, K. Hamacher, M.J. Riemenschneider, G. Reifen-
berger, H.W. Muller, K. Zilles, H. Coenen, K.J. Langen, Brain 128 (2005)
678–687.
[3] C. Lemaire, S. Gillet, S. Guillouet, A. Plenevaux, J. Aerts, A. Luxen, Eur.
J. Org. Chem. (2004) 2899–2904.
[4] P.N. Krasikova, V.V. Zaitsev, S.M. Ametamey, O.F. Kuznetsova,
ˇ
ˇ
O.S. Fedorova, I.K. Mosevich, Y.N. Belokon, S. Vyskocil, S.V. Shatik,
M. Nader, P.A. Schubiger, Nucl. Med. Biol. 31 (2004) 597–603.
[5] G.N. Reddy, M. Haeberli, H.F. Beer, A.P. Schubiger, Appl. Radiat. Isot. 44
(1993) 645–649.
For determination of the benzoic acid arising from
oxidation, an aliquot of the reaction mixture (at 1, 3, 10 and
30 min) was injected onto HPLC. The eluent included 0.1%
trifluoro acetic acid. UV peaks of benzoic acids and precursors
were detected. Concentration curves of benzoic acids and the
[6] A. Plenevaux, C. Lemaire, A.J. Palmer, P. Damhaut, D. Comar, Appl.
Radiat. Isot. 43 (1992) 1035–1040.

1468
B. Shen et al. / Journal of Fluorine Chemistry 128 (2007) 1461–1468
[7] Y.-S. Ding, C.-Y. Shiue, J.S. Fowler, A.P. Wolf, A. Plenevaux, J. Fluorine
[14] T. Tierling, Berichte des Forschungzentrums Ju¨lich (2002) 3952.
[15] A. Al-Labadi, K.P. Zeller, H.-J. Machulla, Radiochim. Acta 94 (2006)
143–146.
Chem. 48 (1990) 189–205.
[8] I. Ekaeva, L. Barre, M.C. Lasne, F. Gourand, Appl. Radiat. Isot. 46 (1995)
777–782.
[16] A. Al-Labadi, K.P. Zeller, H.-J. Machulla, J. Radioanal. Nucl. Chem. 270
(2006) 313–318.
[9] C. Lemaire, P. Damhaut, A. Plenevaux, R. Cantineau, L. Christiaens, M.
Guillaume, Appl. Radiat. Isot. 43 (1992) 485–494.
[10] R. Rengan, P.K. Chakraborty, M.R. Kilbourn, J. Label. Compd. Radio-
pharm. 33 (1993) 563–572.
[17] F.W. Sweat, W.W. Epstein, J. Org. Chem. 32 (1967) 835–838.
[18] A.J. Mancuso, D. Swern, Synthesis (1981) 165–185.
[19] T.T. Tidwell, Synthesis (1990) 857–870.
[11] D.-R. Hwang, C.S. Dence, J. Gong, M.J. Welch, Appl. Radiat. Isot. 42
(1991) 1043–1047.
[20] M.A. Gal’bershtam, Z.N. Budarina, Zhurnal Organicheskoi Khimii 5 (5)
(1969) 953–956.
[12] S. Stone-Elander, N. Elander, Appl. Radiat. Isot. 44 (1993) 889–893.
[13] C. Lemaire, M. Guillaume, L. Christiaens, R. Cantineau, Appl. Radiat.
Isot. 38 (1987) 1033–1038.
[21] M. Smith, E. Elisberg, M.L. Sherrill, J. Am. Chem. Soc. 68 (1946) 1301.
[22] S.M. Kupchan, H.C. Wormser, J. Org. Chem. 30 (1965) 3792–3800.
[23] F. Rieche, Chem. Ber. 22 (1889) 2347–2364.