Substitution-reduction: An alternative process for the [ 18F]N-(2-fluoroethylation) of anilines

Article abstract of DOI:10.1002/jlcr.816

Substitution of a halo atom (chloro or bromo) in easily prepared N-haloacetyl-anilines with no-carrier added (NCA) cyclotron-produced [ ¹⁸F]fluoride ion (¹⁸F, t_1/2 = 109.8 min; β⁺ = 96.9%), followed by reduction with borane-tetrahydrofuran (BH₃-THF), provides an alternative route to NCA [ ¹⁸F]N-(2-fluoroethyl)-anilines. This two-step and one-pot process is rapid (～50 min) and moderately high yielding (～40% decay-corrected radiochemical yield (RCY) overall). In the nucleophilic substitution reaction, 18-crown-6 is preferred to Kryptofix 222 as complexing agent for the solubilization of the counter-ion (K⁺), derived from an added metal salt, in acetonitrile. Weakly basic potassium bicarbonate is preferred as the added metal salt. Inclusion of a small amount of water, equating to 4-5 molar equivalents relative to 18-crown-6, base or precursor (held in equimolar ratio), is beneficial in preventing the adsorption of radioactivity onto the wall of the glass reaction vessel and for achieving high RCY in the nucleophilic substitution reaction. BH₃-THF is effective for the rapid reduction of the generated [¹⁸F]N-fluoroacetyl-aniline to the [ ¹⁸F]N-(2-fluoroethyl)-aniline. Copyright

Full text of DOI:10.1002/jlcr.816

JOURNAL OF LABELLED COMPOUNDS AND RADIOPHARMACEUTICALS
J Label Compd Radiopharm 2004; 47: 217–232.
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jlcr.816
Research Article
Substitution-reduction: an alternative process for the
[18F]N-(2-fluoroethylation) of anilines
Emmanuelle Briard* and Victor W. Pike
PET Radiopharmaceutical Sciences, Molecular Imaging Branch, National
Institute of Mental Health, National Institutes of Health, Building 10,
Rm. B3 C346, 10 Center Drive, Bethesda, MD 20892-1003, USA
Summary
Substitution of a halo atom (chloro or bromo) in easily prepared N-haloacetyl-
anilines with no-carrier added (NCA) cyclotron-produced [18F]fluoride ion (18F, t1/2
=
109.8 min; b+=96.9%), followed by reduction with borane–tetrahydrofuran (BH3–
THF), provides an alternative route to NCA [18F]N-(2-fluoroethyl)-anilines. This two-
step and one-pot process is rapid (ꢀ50 min) and moderately high yielding (ꢀ40%
decay-corrected radiochemical yield (RCY) overall). In the nucleophilic substitu-
tion reaction, 18-crown-6 is preferred to Kryptofix1 222 as complexing
agent for the solubilization of the counter-ion (Kþ ), derived from an added metal
salt, in acetonitrile. Weakly basic potassium bicarbonate is preferred as the
added metal salt. Inclusion of a small amount of water, equating to 4–5 molar
equivalents relative to 18-crown-6, base or precursor (held in equimolar ratio),
is beneficial in preventing the adsorption of radioactivity onto the wall of the
glass reaction vessel and for achieving high RCY in the nucleophilic substitution
reaction. BH3–THF is effective for the rapid reduction of the generated [18F]N-
fluoroacetyl-aniline to the [18F]N-(2-fluoroethyl)-aniline. Copyright # 2004 John
Wiley & Sons, Ltd.
Key Words: fluorine-18; [18F]N-(2-fluoroethylation); anilines; N-haloacetyl-anilines;
nucleophilic substitution; reduction
Introduction
Fluorine-18 (t1/2=109.8 min; b+=96.9%) is a major positron-emitter for
labeling tracers for application in clinical research1–3 and drug development4,5
with positron emission tomography (PET).1–3 The usual source of no-carrier-
*Correspondence to: E. Briard, PET Radiopharmaceutical Sciences, Molecular Imaging Branch, National
Institute of Mental Health, National Institutes of Health, Building 10, Room B3 C346, 10 Center Drive,
Bethesda, MD 20892-1003, USA. E-mail: briarde@intra.nimh.nih.gov
Copyright # 2004 John Wiley & Sons, Ltd.
Received 25 November 2003
Revised 20 January 2004
Accepted 22 January 2004

218
E. BRIARD AND V. W. PIKE
added (NCA) fluorine-18 is the 18O(p,n)18F reaction on 18O-enriched water,
which provides aqueous NCA [18F]fluoride ion.6 Invariably, the use of
[18F]fluoride ion for labeling tracers must proceed with a nucleophilic
substitution reaction, either directly on a precursor or on another species
that can then be used as a labeling agent.7 Important amongst such labeling
agents are compounds of formula X(CH2)1n8F, where n may range from 1 to 4
and X may be halogen (Cl, Br or I),8–20 alkyl sulfonyloxy (OMs or
OTf)11,18,19,21 or aryl sulfonyloxy (OTs).10,11,16,20,22 These labeling agents are
usually prepared from the corresponding compounds with formula X(CH2)nX
and are often used to perform [18F]o-fluoroalkylations at nucleophilic centers
(e.g. OH, NH2, SH).7 This two-step process has given rise to several useful
radiotracers, including [18F]N-(2-fluoroethyl)-spiperone,23–25 [18F]FECNT26,27
and [18F]O-(2-fluoroethyl)-l-tyrosine.28
[18F](2-Fluoroethylation) (n=2) at a hetero atom Y (Y=N(H/R), O
or S) has been especially popular, because it provides a convenient means
to label organic molecules, often with little change to their lipophilicity,
pharmacology and pharmacokinetics. Acceptable radiochemical yields
can be obtained from the reactions of these labeling agents with good
nucleophiles (e.g. phenoxides,11,19,28 secondary aliphatic amines16,19,21,26 and
carboxylates20). However, there are few examples of this process being useful
for the [18F](2-fluoroethylation) of weak nucleophiles (e.g. aromatic
amines29,30).
An alternative means for creating an [18F]2-fluoroethyl group at a hetero
atom is to directly substitute X in RY(CH2)2X with [18F]fluoride ion.25,31,32
However, this single step process is sometimes low yielding. Also the precursor
RY(CH2)2X may be difficult to prepare or vulnerable, for example, to
elimination, hydrolysis or, in the case of Y= N(H/R), formation of an
aziridinum intermediate.21,33
In our recent experience the special case of direct [18F]N-2-fluoroethylation
of anilines is quite difficult to achieve.30 Here we report an alternative two-step
process for the synthesis of [18F]N-(2-fluoroethyl)-anilines based on direct
nucleophilic substitution with NCA [18F]fluoride ion in an N-haloacetyl-
aniline (halo=Cl or Br) followed by rapid reduction of the generated [18F]N-
fluoroacetyl-aniline in situ.
Results and discussion
The reference N-(2-fluoroethyl)-anilines 3a and 3b were prepared in three steps
from aniline and N-methyl-aniline, respectively (Figure 1). First the precursor
N-bromoacetyl- and N-chloroacetyl-anilines 1a–d were obtained in high yields
by treating the aniline with the corresponding haloacetic anhydride in glacial
acetic acid.34 1a–1d were converted into the corresponding N-fluoroacetyl-
anilines 2a and 2b in moderate yield by treatment with powdered potassium
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

SUBSTITUTION-REDUCTION: AN ALTERNATIVE
219
O
R
F
R
F
N
N
BH3-THF (1M)
65˚C, 1 h
KF
O
Di(ethylene
glycol)
150˚C, 2h
R
X
NHR
N
R = H, 2a, 53% from 1a
R = Me, 2b, 55% from 1c
R = H, 3a, 71%
R = Me, 3b, 68%
(XCH2CO)2O
AcOH,
O
18F]KF
RT, 2 h
[
18F
R
18F
R
18-Crown-6
N
N
R = H, Me
BH3-THF (1M)
65 oC,10 min
MeCN
130˚C, 10 min
R = H, X = Br, 1a, 72%
R = H, X = Cl, 1b, 74%
R = Me, X = Br, 1c, 78%
R = Me, X = Cl, 1d, 79%
R = H, 4a, 58% RCY from 1a R = H, 5a, ~ 70% RCY
48% RCY from 1b R = Me, 5b ~ 70% RCY
R = Me, 4b, 62% RCY from 1c
57% RCYfrom 1d
Figure 1. Synthesis of N-(2-fluoroethyl)-anilines and [18F]N-(2-fluoroethyl)-
anilines
fluoride in di(ethylene glycol).35,y Treatment of 2a and 2b with borane-
tetrahydrofuran complex36 (BH3–THF) gave 3a and 3b, respectively, in
moderately high yield.
In this study, we used N-bromoacetyl-aniline 1a as a model substrate for
optimizing the nucleophilic substitution of the bromo atom by NCA
[18F]fluoride ion. Complexing agent, amount of added water, metal salt,
amount of precursor, temperature, reaction time, leaving group and N-
substitution were varied. Favorable conditions were then applied to the chloro
analog 1b and the secondary amides 1c and 1d. Conditions were also
established for the reduction of the [18F]N-fluoroacetyl-anilines 4a and 4b to
the desired [18F]N-(2-fluoroethyl)-anilines 5a and 5b.
Optimization of nucleophilic substitution with NCA [18F]fluoride ion
Choice of complexing agent. Initially, radiolabeling was tried with 1a (5 mg)
under conditions widely applied to achieve successful aliphatic nucleophilic
substitution reactions with NCA [18F]fluoride ion, namely using acetonitrile as
solvent, Kryptofix1 222 as cryptand and potassium carbonate as added metal
salt.37 The decay-corrected radiochemical yield (RCY) of [18F]N-fluoroacetyl-
aniline 4a from a reaction performed in a glass vessel heated in an oil bath at
1308C for 10 min was 15%. In reactions where the amount of precursor was
increased, RCYs were even lower. The N-bromoacetyl compound 1a was
found to be capable of N-alkylating Kryptofix 222 under the reaction
y Fluoroacetyl chloride is no longer commercially available and hence it was not possible to prepare 2a and
2b by single step acylation of the anilines.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

220
E. BRIARD AND V. W. PIKE
N
O
O
N
Br
O
O
O
O
O
O
1a
K+
[
18F]F-
+
4a, 15% RCY
O
O
N
N
NH
K2CO3, MeCN
130˚C, 10 min
O
O
O
Kryptofix 222
6
O
O
O
O
O
1a
K+
O
[
18F]F-
4a, 37% RCY
K2CO3, MeCN
130˚C, 10 min
18-Crown-6
O
OH
18F]F-
O
O
O
1a
K+
O
[
4a, 41% RCY
O
K2CO3, MeCN
130˚C, 10 min
2-(Hydroxymethyl)-18-crown-6
Figure 2. Effect of complexing agent on the radiofluoridation of 1a
1
conditions, as evidenced by new H- and 13C-NMR signals after heating
Kryptofix 222 with one molar equivalent of 1a in CD3CN at 1308C for
10 min (see Experimental). The mono-alkylated product 6 (Figure 2), was
also detected by LC-MS. By switching the complexing agent to the
amine-devoid 18-crown-638, 4a was obtained in a much higher RCY (37%).
As expected, no reaction between the precursor and 18-crown-6 occurred
under these conditions. However, a high proportion (55%) of the
radioactivity, assumed to be [18F]fluoride ion, became adhered to the wall of
the glass vessel during the course of the reaction and was unavailable for
reaction.
The use of 2-(hydroxymethyl)-18-crown-6 as complexing agent was also
investigated. Under conditions where the molar ratio of complexing agent,
precursor 1a and potassium ion (as carbonate) was 4:1:1, respectively, [18F]N-
fluoroacetyl-aniline 4a was obtained in 41% RCY. Adsorption of radioactivity
onto the glass reaction vessel was 31% at the end of the reaction. Evidently, 2-
(hydroxymethyl)-18-crown-6 adequately solubilized the K+ ion in acetonitrile
and also promoted the nucleophilicity of the [18F]fluoride ion. This result
showed the tolerance of the reaction for the presence of free aliphatic hydroxyl
groups.
18-Crown-6 was used as the preferred complexing agent, since it gave higher
RCYs than Kryptofix 222 and was cheaper than its 2-hydroxymethyl
derivative (or indeed Kryptofix 222).
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

SUBSTITUTION-REDUCTION: AN ALTERNATIVE
221
Effect of added water. The NCA [18F]fluoride ion available for our reactions at
the end of azeotropic drying is expected to be almost dehydrated or almost
‘naked’ i.e. to exist as Fꢁ(H2O)n where n is very small. The nucleophilicity of
fluoride ion (as R4N+ salts) in aliphatic substitution reactions in acetonitrile39
or aprotic solvents40 depends on its degree of solvation, such that when n=1.5
reaction rates may be about 100-fold higher than when n=8.5.40 Some
nucleophilicity may be retained when n is as high as 10.0.39 The N-haloacetyl-
anilines 1a–1d are expected to be very reactivez towards nucleophiles by virtue
of the expected activating effect of the a-carbonyl group.41,42 Tolerance for a
trace of water in some nucleophilic substitution reactions of [18F]fluoride ion is
known.43,44 Hence, the reactive N-haloacetyl anilines might be expected to
react with moderately hydrated fluoride ion (i.e. fluoride ion that is moderately
nucleophilic).
The adsorption of NCA [18F]fluoride ion from organic solvents to reaction
vessel walls is a well-known phenomenon.45 Adsorption of radioactivity onto
the glass reaction vessel wall was very evident in this study. ‘Naked’
[18F]fluoride ion most probably adsorbs onto the walls of glass vessels
through stabilizing interactions (hydrogen-bonding) with surface hydroxyl
groups. Strategies for reducing such adsorption have included silylation of
glass vessels and the use of vessels with non-hydroxylic surfaces (glassy carbon
or platinum).45 However, these strategies were not explored here.
Instead, we reasoned that addition of a low proportion of water to the NCA
[18F]fluoride ion might hydrate it sufficiently to hinder its adsorption onto the
glass vessel wall without destroying its nucleophilicity towards the reactive 1a.
Accordingly, we tried reactions of [18F]fluoride ion with 1a–1d to test this
hypothesis. In the presence of a small proportion of added water, the reactions
generally proceeded more reproducibly and with less adsorption of radio-
activity onto the glass vessel wall. These reactions occurred cleanly. There were
no radioactive by-products in the case of the secondary amides 1c and 1d (e.g.
Figure 3), while the primary amides 1a and 1b gave only one minor radioactive
by-product (510% of radioactivity in analyte; HPLC retention time
10.6 min). Unreacted precursor always remained in the reaction mixtures.
For three different amounts of precursor 1a (2.1, 3.0 and 4.5 mg; 9.8, 14 and
21 mmol), we varied the molar ratio of water while 1a, complexing agent and
metal salt were kept equimolar (Figure 4). In each series, the RCYs were
maximal when the molar ratio of water to each of the other components was
4:1. The maximal RCY was highest (58%) when using 3 mg (14 mmol) of 1a
(Figure 4).
The hydrations of 18-crown-6 and its potassium complex have been well
studied.46 A few water molecules of hydration greatly enhance the selectivity
z The ability of 1a to alkylate Kryptofix 222 is evidence of this reactivity.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

222
E. BRIARD AND V. W. PIKE
4b
1c
3
2
1
3
2
1
0
[18F]F-
0
0
2
4
6
8
10
12
14
16
Time (min)
Figure 3. Representative chromatogram from the HPLC analysis of the
products from a nucleophilic substitution reaction ([18F]fluoride ion with 1c to
give 4b)
60
40
20
0
0
2
4
6
8
Molar ratio of water to 18-crown-6
Figure 4. Effect of water on RCY of 4a from 1a treated with [18F]fluoride ion in
MeCN (0.65 ml) at 1308C for 10 min. (&) 1a (2.1 mg; 9.8 mmol): 18-crown-6
(9.8 mmol): KHCO3 (9.8 mmol) (n) 1a (3.0 mg; 14 mmol): 18-crown-6 (14 mmol):
KHCO3 (14 mmol). (}) 1a (4.5 mg; 21 mmol): 18-crown-6 (21 mmol): KHCO3
(21 mmol). ( ) 1a (3.0 mg; 14 mmol): 18-crown-6 (9.8 mmol): KHCO3 (9.8 mmol)
of 18-crown-6 for binding potassium ion versus other metal ions.47 Thus, trace
water added to the 18-crown-6-mediated reactions of 1a with [18F]fluoride ion
most probably serves to solubilize and stabilize the 18-crown-6-K+ complex
by hydration. The added water may also help to solubilize anions, principally
the bulk of added bicarbonate ion, but also trace fluoride ion, trace reaction
product and bromide ion.
Five molar equivalents of water (49 mmol) were necessary to reach a
maximal RCY of 4a of 60%, when using 21 mmol of amide precursor 1a and
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

SUBSTITUTION-REDUCTION: AN ALTERNATIVE
223
9.8 mmol each of 18-crown-6 and potassium bicarbonate (Figure 4). The
requirement for an extra molar equivalent of water under these conditions of
increased proportion of precursor 1a suggests that some of the added water is
consumed in hydration of the precursor. This might occur through hydrogen
bonding to the amido group.
The sharp decline in RCYs when the molar ratios of water to other
components of the reaction increased above 4 or 5 (Figure 4) suggests that
under these conditions the [18F]fluoride ion became increasingly hydrated and
less nucleophilic.
The inclusion of 4 molar equivalents of water in the reactions limited the
adsorption of radioactivity (assumed to be [18F]fluoride ion) onto the glass
wall of the reaction vessel to an acceptable level (average ꢀ20%; Table 1,
entries 8, 12–14); generally, the level of adsorption decreased with increased
proportion of water.
Having established the beneficial effects of an added trace of water on
the overall outcome of the nucleophilic substitution reaction, we proceeded
to investigate the effect of varying other parameters of the reaction, as
follows.
Table 1. Results from the radiofluoridations of N-haloacetyl-anilines under various
conditionsa
Entry
Salt
Precursor
(], mg; mmol)
Temperature
(8C)
Time
(min)
RCYb
(%)
RCYc
(%)
Solubilityd
(%)
1
2
3
Cs2CO3
K2CO3
KI
1a, 3, 14
1a, 3, 14
1a, 3, 14
130
130
130
10
10
10
7
35
37
33
57
59
23
57
64
4
5
6
KHCO3
KHCO3
KHCO3
1a, 3, 14
1a, 5, 23
1a, 7, 33
80
80
80
10
10
10
29
31
43
48
44
60
61
70
74
7
8
KHCO3
KHCO3
1a, 3, 14
1a, 3, 14
110
130
10
10
39
58
64
72
67
76
9
10
11
KHCO3
KHCO3
KHCO3
1a, 3, 14
1a, 3, 14
1a, 3, 14
130
130
130
5
20
30
44
48
43
74
67
58
70
72
74
12
13
14
KHCO3
KHCO3
KHCO3
1b, 3, 18
1c, 3, 13
1d, 3, 16
130
130
130
10
10
10
48
62
57
80
71
76
79
87
77
aInitials conditions: precursor (as listed), 18-crown-6 (14 mmol), salt (14 mmol in metal ion), H2O (56 mmol),
MeCN (0.65 ml).
bRCY from initial activity of NCA [18F]fluoride ion in reaction vessel.
cRCY from radioactivity in solution at the end of reaction.
dPercent of original radioactivity in solution at the end of reaction; remainder is adhered to the glass
reaction vessel wall.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

224
E. BRIARD AND V. W. PIKE
Effect of added metal salt. Reactions of precursor 1a with [18F]fluoride ion, in
which the added metal salt was cesium carbonate, potassium carbonate,
potassium iodide or potassium bicarbonate gave 4a in 7, 35, 37 and 58%
RCY, respectively (Table 1, entries 1–3, 8). The low RCY from the use of
cesium carbonate is due to the poor solubilization of ‘cesium [18F]fluoride’ by
18-crown-6; indeed 77% of the radioactivity was adsorbed onto the vessel wall
at the end of the reaction (Table 1, entry 1). The higher RCY with potassium
bicarbonate is attributed to less decomposition of the reactive precursor under
the more weakly basic conditions. Potassium bicarbonate was therefore used
as the preferred added metal salt.
Effect of precursor quantity. Increasing the amount of precursor 1a from 3 mg
(14 mmol) to 7 mg (33 mmol) increased the RCY from 29 to 43% for reactions
carried out for 10 min in an oil-bath at 808C (Table 1, entries 4–6). At higher
temperature (1308C), 3 mg (14 mmol) of precursor gave the highest RCY (61%)
in the precursor range 2.1–4.5 mg (Figure 4). Thus the reaction works with a
moderately low amount of precursor, as might be demanded by a need to
conserve expensive or rare precursor in radiopharmaceutical production.
Effect of temperature. By increasing the temperature from 80 to 1308C, the
RCY of 2a increased from 29 to 58% (Table 1, entries, 4, 7, 8).
Effect of reaction time. For reactions carried out at 1308C with 3 mg of 1a,
RCYs of 2a increased to 58% for reaction times up to 10 min and then,
because of the formation of radioactive by-products, declined to 43% for
reactions up to 30 min (Table 1, entries 8–11).
Effect of leaving group. A bromo leaving group gives the desired [18F]N-
fluoroacetyl-aniline 4a or 4b in appreciably greater RCYs than a chloro
leaving group (Table 1, cf. entries 8 with 12 and 13 with 14). The
N-bromoacetyl-anilines are just as easily prepared as the N-chloroacetyl-
anilines and are stable to storage. A bromo precursor is therefore preferred.
Effect of N-substitution. In a previous study, where [18F]fluoride ion was used
to displace tosyl from N-a-tosylmethyl-anilines, it was noted that primary
amides gave low RCYs (ꢀ6%), whereas their secondary N-ethyl amido
analogs, gave moderately high RCYs (ꢀ60%).48 This marked difference was
attributed to deprotonation of the primary amido group under the basic
conditions (Kryptofix 222-K2CO3) and its adverse effect on the nucleophilicity
of the [18F]fluoride ion. In this study, for the same leaving group, the primary
amides 1a and 1b gave just marginally lower RCYs of [18F]N-fluoroacetyl-
anilines than the secondary amides 1c and 1d (Table 1, cf. entries 13 with 8,
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

SUBSTITUTION-REDUCTION: AN ALTERNATIVE
225
and 14 with 12). The conditions used here are probably insufficiently basic to
depress the RCYs of the primary amides greatly.
Reduction conditions
The strong carbon–fluorine bond is generally quite resistant to chemical
reducing agents, even strong reducing agents that readily reduce other carbon–
halogen bonds.49–51 Borane-THF was selected as a convenient reagent for the
reduction of 4a to [18F]N-(2-fluoroethyl)-aniline 5a in situ. Reduction was
achieved by treating the crude reaction mixture from the nucleophilic
substitution reaction with borane-THF under mild conditions (10 min,
658C). The conversion of 4a into 5a was estimated to be 65–70% from HPLC
analysis. NCA [18F]N-(2-fluoroethyl)-aniline 5a was produced in 40% RCY
overall within 50 min of radionuclide production. The reaction proceeded
cleanly. The only other radioactive species detected at the end of reaction was
unreacted [18F]fluoride ion. Similarly 4b was reduced to 5b in about 70% RCY
(representing an overall RCY from [18F]fluoride ion of about 40%).
By-products identified by MS were the corresponding 2-hydroxy-N-methyl-
N-phenyl-acetamide 7 and N-(2-bromoethyl)-N-methyl-aniline 8. These were
easily separable from the corresponding radioactive product (Figure 5).
A requirement of this method of [18F]N-(2-fluoroethylation) is that the
compound to be labeled should not contain other functional groups that might
be easily and undesirably reduced by the selected reducing agent. For wider
application of this method, other reducing agents might be considered for the
reduction of the intermediate 18F-labeled amide to the required [18F]N-(2-
fluoroethyl)-aniline, depending on the structure of the compound to be
reduced.
5b
7
8
2
1
1c
0.20
0.10
0.0
[18F]F-
0
2
4
0
10
6
8
12
14
16
Time (min)
Figure 5. Representative chromatogram from the HPLC analysis of the reaction
mixture from a reduction reaction (of 4b to 5b). 7: 2-hydroxy-N-methyl-N-
phenyl-acetamide, 8: N-(2-bromoethyl)-N-methyl-aniline
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

226
E. BRIARD AND V. W. PIKE
Conclusions
The described two-step one-pot process provides an alternative route for
the NCA [18F]N-2-fluoroethylation at a weakly nucleophilic center (aromatic
amine). The overall RCY is about 40% and acceptable for application
to the radiosynthesis of PET radiopharmaceuticals. The preferred
bromo precursors are easily prepared and stored. Inclusion of a fixed
proportion of water in the first stage, the nucleophilic substitution reaction
with NCA [18F]fluoride ion in a glass vessel, affords high RCY and
reproducibility.
Experimental
Materials
Ammonium formate (96%, J.T. Baker) and acetonitrile (high purity solvent,
Burdick & Jackson) were used as supplied. Kryptofix 2221 (4,7,13,16,
21,24-hexaoxa-1,10-diazabicyclo[8,8,8]hexacosane) (98+%), 18-crown-6 (1,4,
7,10,13,16-hexaoxacyclooctadecane) (99.5+%), 2-hydroxymethyl-18-crown-6
(95+%), borane-THF complex (1 M in THF in Sure/SealTM bottle), aniline
(99.5+%), N-methyl-aniline (99+%), bromoacetic anhydride (99+%),
chloroacetic anhydride (97+%), potassium carbonate (99.995%), potassium
iodide (99+%), potassium bicarbonate (ReagentPlusTM), potassium fluoride
(99.99+%, anhydrous), THF (99.9+%, inhibitor-free, anhydrous) and other
common chemicals were purchased from Aldrich and used as received. Glass
vessels (5 ml; borosilicate glass; tapered cone; fitted with phenolic caps and
TFE/silicone liners; Mini-vial 20/400) for radiochemistry were purchased from
Alltech (Deerfield, USA) and were cleaned in water and acetone, and then
dried at 1008C before use.
General methods
Radioactivity was measured using a calibrated dose calibrator (AtomlabTM
300; Biodex Medical Systems) with correction for background and physical
decay.
Each prepared compound was analyzed by high performance liquid
chromatography (HPLC) performed on apparatus (System Gold; Beckman
Coulter) equipped with a reverse phase column (C18; 10 m particle size;
250 ꢂ 4.60 mm o.d.; Phenomenex) eluted with MeCN-0.01 M ammonium
formate at 2 ml/min [gradient (v/v) from 20:80 to 70:30 over 12 min then
isocratic (70:30) for 8 min]. The eluate was monitored for absorbance (254 nm;
166 detector, Beckman Coulter) and where appropriate simultaneously for
radioactivity (diode or PM Flow-count detector; Bioscan). Each 18F-labeled
compound was identified by retention time (Rt) and co-injection of the
corresponding reference compound.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

SUBSTITUTION-REDUCTION: AN ALTERNATIVE
227
RCYs of labeling reactions were estimated by analytical HPLC, with
integration of radioactive peaks; they represent conversions, not isolated
yields. The area of the product peak (P, in units of V min) was compared to the
integral of radioactivity from all eluted peaks (T, in units of V min) and the
radiochemical yield calculated as (P/T ꢂ 100%). No corrections were made for
radioactive decay in the short time between the first and last eluted peak
(ꢀ12 min). Regular checks were performed, confirming that all radioactivity
injected onto HPLC was eluted during the course of analysis.
1H- 13C- and 19F-NMR spectra were recorded on an Avance-400 spectro-
1
meter (Bruker). Chemical shifts in H- and 13C-NMR are reported in d units
(ppm) downfield relative to tetramethylsilane. Chemical shifts in 19F-NMR are
reported in d units (ppm) downfield from fluorotrichloromethane as reference.
Abbreviations s, d, dt, t and m denote singlet, doublet, double triplet, triplet
and multiplet, respectively.
IR analysis was performed on a Spectrum One FT-IR spectrometer (Perkin
Elmer).
Melting points are uncorrected and were determined with an Electrothermal
apparatus (MEL-TEMP1).
LC-MS analysis was performed on a LC apparatus (Thermo Finnigan
Surveyor), equipped with a reverse phase column (C18; 5 mm, 150 ꢂ 2 mm i.d.;
Phenomenex), eluted with a gradient of MeOH-H2O-0.5% AcOH at 0.15 ml/
min, linked to electrospray MS (ThermoFinnigan LCQDECA). GC-MS
analysis was performed on a Thermo Finnigan Polaris Q instrument equipped
with a capillary column (Rtx-5MS 30 m ꢂ 0.25 mm i.d.; Restek). The carrier
gas was helium (1 ml/min) and the oven temperature heated from 60 to 1808C
at 208C/min. The source temperature of MS was 2508C and the analysis
performed in the electron ionization (+ve) mode.
Synthesis of N-haloacetyl-anilines (1a–1d)
General procedure: aniline or N-methyl-Aniline (20 mmol) was dissolved in
glacial acetic acid (10 ml). As appropriate, bromoacetic anhydride or
chloroacetic anhydride (20 mmol) was added dropwise to this solution at
RT. The reaction mixture was stirred for 2 h at RT. Acetic acid was removed
under reduced pressure. Aqueous NaOH (1M) was added to the residue and
the product extracted with CH2Cl2 (3 ꢂ 10 ml). The organic layer was washed
with cold water, dried over MgSO4, filtered and concentrated. Thus prepared
were:
N-bromoacetyl-aniline 1a as slightly beige crystals (72%) after recrystallization
(EtOH-H2O). M.p.: 1328C. HPLC: Rt=9.1 min.1H-NMR (CDCl3) d: 4.02 (s,
2 H), 7.17 (t, 1 H, J=7.4 Hz), 7.36 (t, 2 H, J=7.5 Hz), 7.53 (d, 2 H, J=7.5 Hz),
8.20 (s, 1 H, NH); cf. Lit.52 13C-NMR (CDCl3) d: 29.49, 120.04 (2C), 125.21,
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

228
129.12 (2C), 136.91, 163.32. GC-MS, m/z=214.06 and 216.06 [M+H]+ IR
(KBr): 1660 cmꢁ1
E. BRIARD AND V. W. PIKE
.
N-chloroacetyl-aniline 1b as white crystals (74%) after recrystallization
1
(EtOH-H2O). M.p.: 1308C. HPLC: Rt=8.7 min. H-NMR (CDCl3) d: 4.20
(s, 2 H), 7.17 (t, 1 H, J=7.4 Hz), 7.35 (t, 2 H, J=7.6 Hz), 7.53 (d, 2 H,
J=7.6 Hz), 8.20 (s, 1 H, NH); cf. Lit.34 13C-NMR (CDCl3) d: 42.89, 120.13
(2C), 125.26, 129.15 (2C), 136.68, 163.77. GC-MS, m/z=170.09 and 172.09
[M+H]+. IR (KBr): 1632 cmꢁ1
.
N-bromoacetyl-N-methyl-aniline 1c as slightly yellow crystals (78%) after
1
recrystallization (EtOH-H2O). M.p.: 478C. HPLC: Rt=9.4 min. H-NMR
(CDCl3) d: 3.30 (s, 3 H), 3.70 (s, 2 H), 7.26–7.32 (m, 2 H), 7.36–7.49 (m, 3 H);
cf. Lit.53 13C-NMR (CDCl3) d: 26.76, 38.10, 127.03 (2C), 128.55, 130.03 (2C),
143.09, 166.58. GC-MS, m/z=227.00 and 229.02 [M]+. IR (KBr): 1643 cmꢁ1
.
N-chloroacetyl-N-methyl-aniline 1d as slightly purple crystals (79%) after
1
recrystallization (EtOH-H2O). M.p.: 698C. HPLC: Rt=9.1 min. H-NMR
(CDCl3) d: 3.30 (s, 3 H), 3.85 (s, 2 H), 7.24–7.27 (m, 2 H), 7.36–7.50 (m, 3 H);
cf. Lit.53 13C-NMR (CDCl3) d: 38.01, 41.51, 127.09 (2C), 128.59, 130.09 (2C),
142.75, 166.30. GC-MS, m/z=183.05 and 185.07 [M]+. IR (KBr): 1687 cmꢁ1
.
Synthesis of N-fluoroacetyl-anilines (2a and 2b)
General procedure: A mixture of compound 1a or 1c (3.5 mmol), dry
potassium fluoride (8.9 mmol), and di(ethylene glycol) (5 ml) was stirred and
heated rapidly to 1508C and kept at this temperature for 2 h. The reaction
mixture was cooled, diluted with water (10 ml), and extracted with toluene
(3 ꢂ 10 ml). The organic layer was washed with sodium bicarbonate solution
(5% w/v; 10 ml), dried over MgSO4, filtered and concentrated. Thus, obtained
were:
N-fluoroacetyl-aniline 2a as white crystals (53%) after column chromatogra-
phy (silica gel; hexane-AcOEt, 80:20 v/v). M.p.: 1318C; cf. Lit.35 HPLC:
1
Rt=7.4 min. H-NMR (CDCl3) d: 4.91 (d, 2 H, J=47.4 Hz), 7.18 (t, 1 H,
J=7.4 Hz), 7.35 (t, 2 H, J=7.5 Hz), 7.57 (d, 2 H, J=7.6 Hz), 8.00 (s, 1 H, NH).
13C-NMR (CDCl3) d: (80.41, 82.27, JCꢁF=188 Hz), 121.26 (2C), 126.28,
130.30 (2C), 137.58, 166.69. 19F-NMR (CFCl3) d: ꢁ221.44. GC-MS, m/z =
154.12 [M+H]+. IR (KBr): 1673 cmꢁ1
.
N-fluoroacetyl-N-methyl-aniline 2b as slightly beige crystals (55%) after
chromatography (silica gel; hexane-AcOEt, 80:20 v/v): M.p.: 898C; cf. Lit.54
HPLC: Rt=7.5 min. 1H-NMR (CDCl3) d: 3.31 (s, 3 H), 4.65 (d, 2 H,
J=47.2 Hz), 7.18–7.26 (m, 2 H), 7.38–7.48 (m, 3 H). 13C-NMR (CDCl3) d:
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

SUBSTITUTION-REDUCTION: AN ALTERNATIVE
229
37.30, (77.67, 79.42, JCꢁF=176.1 Hz), 126.98 (2C), 128.73, 130.16 (2C),
141.37, 166.76. 19F-NMR (CFCl3) d: ꢁ225.01. GC-MS, m/z=168.10
[M+H]+. IR (KBr): 1668 cmꢁ1
.
Synthesis of N-(2-fluoroethyl)-anilines (3a and 3b)
General procedure: A solution of BH3 (1 M; 30 mmol) in THF was added
dropwise for 1 h to a solution of the N-fluoroacetyl-aniline 2a or 2b (3 mmol)
in anhydrous THF (3 ml) under N2 at 08C. The mixture was stirred for 30 min
at 08C and refluxed for 1 h. Excess BH3 was quenched with methanol (3 ml) at
08C and the solution evaporated to dryness. The oily residue was treated with
a mixture of CH2Cl2 and brine. The organic layer was separated, dried over
MgSO4, filtered and concentrated. Thus prepared were:
N-(2-fluoroethyl)-aniline 3a as an oil (71%) after column chromatography
1
(silica gel; hexane, 0.5% NEt3-AcOEt; 80:20 v/v). HPLC: Rt=11.1 min. H-
NMR (CDCl3) d: 3.40 (dt, 2 H, J=26.5, 4.9 Hz), 3.90 (s, 1 H, NH), 4.60 (dt,
2 H, J=47.4, 4.8 Hz), 6.63 (d, 2 H, J=8.6 Hz), 6.73 (t, 1 H, J=7.3 Hz), 7.18 (t,
2 H, J=7.4 Hz); cf. Lit.55 13C-NMR (CDCl3) d: (44.13, 44.33, JCꢁF=20.5 Hz),
(81.64, 83.30, JCꢁF=167 Hz), 113.16 (2C), 118.09, 129.38 (2C), 147.61. 19F-
NMR (CFCl3) d: ꢁ223.79. LC-MS, m/z=140.10 [M+H]+.
N-(2-fluoroethyl)-N-methyl-aniline 3b as an oil (68%) after column chroma-
tography (silica gel; hexane, 0.5% NEt3-AcOEt, 80:20 v/v). HPLC:
Rt=12.7 min.1H-NMR (CDCl3) d: 3.00 (s, 3 H), 3.64 (dt, 2 H, J=24.0,
5.2 Hz), 4.59 (dt, 2 H, J=47.2, 5.3 Hz), 6.67–6.76 (m, 3 H), 7.20–7.27 (m, 2 H);
cf. Lit.56 13C-NMR (CDCl3) d: 39.05, (52.68, 52.87, JCꢁF=21.4 Hz), (81.0,
82.7, JCꢁF=169.8 Hz), 112.37 (2C), 116.87, 129.29 (2C), 148.99. 19F-NMR
(CFCl3) d: ꢁ221.33. LC-MS, m/z=154.10 [M+H]+.
NMR and LC-MS study of alkylation of Kryptofix 222 by N-2-bromoacetyl-N-
phenyl-aniline
A solution of 1a (2.3 mg; 10 mmol), Kryptofix 222 (3.9 mg; 10 mmol) and
KHCO3 (1.0 mg; 10 mmol) in CD3CN (0.5 ml) was heated at 1308C in a sealed
glass vial for 10 min. The following NMR signals were recorded from this
1
mixture at the end of reaction: H-NMR (CD3CN) d: 2.35 (m, 12 H, CH2N),
3.36 (m, 12 H, CH2O), 3.45 (s, 12 H, OCH2CH2O), 4.30 (s, 2 H), 7.25 (m, 4 H),
7.65 (m, 1 H). 13C-NMR (CD3CN) d: 53.41, 67.15, 69.93, 125.33, 126.77,
128.77. LC-MS of the reaction mixture: m/z=511.4 ([M+1]+; calculated for
C26H45N3O7, 510.3). Reference Kryptofix 222 gave the following NMR
spectra: 1H-NMR (CD3CN) d: 2.5 (t, 12 H, CH2N), 3.46 (t, 12 H, CH2O), 3.55
(s, 12 H, OCH2CH2O); 13C-NMR (CD3CN) d: 55.70, 69.33, 70.11.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

230
E. BRIARD AND V. W. PIKE
Radiochemistry
Radionuclide production. [18F]Fluoride ion was produced by the 18O(p,n)18F
reaction, by irradiation of 18O-enriched water (95 atom %; 1.8 ml) with a beam
of protons (14.1 MeV; 20–25 mA). Generally, this method produces [18F]fluor-
ide ion with a specific radioactivity exceeding 400 GBq/mmol. Portions (20–
200 ml) of the whole irradiated water were used for individual experiments in
this study.
Nucleophilic substitution reaction–general procedure. A glass reaction vial
(5 ml; V-vial) was loaded with an aqueous solution (10 ml) of KHCO3 (2.3 mg;
23 mmol) plus a solution of 18-crown-6 (6.2 mg; 23 mmol) in MeCN (0.15 ml).
Cyclotron-produced NCA [18F]fluoride ion (574 GBq) in [18O]water (20–
200 ml) was added and the solvent evaporated under reduced pressure
(controlled with a bleed of nitrogen) while heating at 1108C (oil bath). Water
was removed azeotropically during three cycles of addition-evaporation of
MeCN (0.3 ml each cycle). A solution of H2O (1 ml; 56 mmol) in MeCN (0.5 ml)
was added to the dry residue, followed by addition of N-haloacetyl-aniline
(3 mg; ꢀ14 mmol) in MeCN (0.15 ml). The reaction vial was sealed, heated at
1308C for 10 min and then cooled to RT. The solution was then withdrawn.
Radioactivities in solution and adhered to the reaction vessel were measured.
The reaction solution was then analyzed by radio-HPLC. Reactions were
repeated with altered parameters as indicated in Table 1.
Reduction–general procedure. The reaction mixture from the nucleophilic
substitution reaction was cooled in a dry ice–acetone bath for 1 min, and then
concentrated to 0.1 ml under reduced pressure at 208C. The mixture was
dissolved in THF (0.5 ml) containing water (1 ml). A solution of BH3 in THF
(1M; 0.1 ml) was then added. The reaction vial was heated at 658C for 10 min
and then cooled in a dry ice–acetone bath for 1 min. Excess BH3 was then
hydrolyzed with H2O (1 ml). The reaction mixture was cooled and then
analyzed by radio-HPLC.
Acknowledgements
The authors are grateful to Drs Frederick T. Chin and John L. Musachio
(MIB, NIMH) for technical advice during this work, Dr Umesha Shetty (MIB,
NIMH) for NMR and MS measurements, and Dr Shui-Yu Lu (MIB, NIMH)
for critical discussion.
References
1. Phelps ME, Mazziotta JC. Science 1985; 228: 799–809.
2. Phelps ME. Neurochem Res 1991; 16: 929–940.
3. Phelps ME. Proc Natl Acad Sci USA 2000; 97: 9226–9233.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

SUBSTITUTION-REDUCTION: AN ALTERNATIVE
231
4. Wagner HN. J Nucl Med 1991; 32: 561–564.
5. PET for drug development and evaluation, Comar D (ed.). Kluwer Academic
Publishers, Dordrecht, The Netherlands, Dev Nucl Med 1995; 26: 1–384.
6. Guillaume M, Luxen A, Nebeling B, Argentini M, Clark JC, Pike VW. Appl
Radiat Isot 1991; 42: 749–762.
7. Ferrieri RA. In Handbook of Radiopharmaceuticals, Welch MJ, Redvanly CS
(eds.), Ch. 7. John Wiley & Sons Ltd: Chichester, UK, 2003; 229–282.
.
8. Coenen HH, Colosimo M, Schuller M, Stocklin G. J Label Compd Radiopharm
1986; 23: 587–595.
9. Shiue CY, Bai L-Q, Teng R-R, Wolf AP. J Label Compd Radiopharm 1987; 24:
55–65.
.
10. Block D, Coenen HH, Stocklin G. J Label Compd Radiopharm 1987; 24:
1029–1042.
.
11. Block D, Coenen HH, Stocklin G. J Label Compd Radiopharm 1988; 25: 201–215.
12. Chesis PL, Welch MJ. Appl Radiat Isot 1990; 41: 259–265.
13. Oh S-J, Choe YS, Chi DY, Kim SE, Choi Y, Lee KH, Ha H-J, Kim B-T. Appl
Radiat Isot 1999; 51: 293–297.
14. Bergman J, Eskola O, Lehikoinen P, Solin O. Appl Radiat Isot 2001; 54: 927–933.
15. Zheng L, Berridge MS. Appl Radiat Isot 2000; 52: 55–61.
.
.
16. Comagic S, Piel M, Schirrmacher R, Hohnemann S, Rosch F. Appl Radiat Isot
2002; 56: 847–851.
17. Zhang M-R, Tsuchiyama A, Haradahira T, Yoshida Y, Furutsuka K, Suzuki K.
Appl Radiat Isot 2002; 57: 335–342.
18. Iwata R, Pascali C, Bogni A, Furumoto S, Terasaki K, Yanai K. Appl Radiat Isot
2002; 57: 347–352.
19. Zhang M-R, Furutsuka K, Yoshida Y, Suzuki K. J Label Compd Radiopharm
2003; 46: 587–598.
20. Wilson AA, Da Silva JN, Houle S. Appl Radiat Isot 1995; 51: 765–770.
21. Skaddam MB, Kilbourn MR, Snyder SE, Sherman PS, Desmond TJ, Frey KA. J
Med Chem 2000; 43: 4552–4562.
22. Chaly Jr T, Matacchieri R, Dahl R, Dhawan V, Eidelberg D. Appl Radiat Isot
1999; 51: 299–305.
.
23. Block D, Coenen HH, Laufer P, Stocklin G. J Label Compd Radiopharm 1986; 23:
1042.
.
24. Coenen HH, Laufer P, Stocklin G, Wienhard K, Pawlik G, Bocher-Schwarz,
.
Heiss W-D. Life Sci 1987; 40: 81–88.
25. Satyamurthy N, Barrio JR, Bida GT, Huang S-C, Mazziotta JC, Phelps ME. Appl
Radiat Isot 1990; 41: 113–129.
26. Goodman MM, Kilts CD, Keil R, Shi B, Martarello l, Xing D, Votaw J, Ely TD,
Lambert P, Owens MJ, Camp VM, Malveaux E, Hoffman JM. Nucl Med Biol
2000; 27: 1–12.
27. Davis MR, Votaw JR, Bremner JD, Byas-Smith MG, Faber TL, Voll RJ,
Hoffman JM, Grafton ST, Kilts CD, Goodman MM. J Nucl Med 2003; 44:
855–861.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232

232
E. BRIARD AND V. W. PIKE
28. Wester HJ, Herz M, Weber W, Heiss P, Senekowitsch-Schmidke R, Schwaiger M,
.
Stocklin G. J Nucl Med 1999; 40: 205–212.
29. Piel M, Bauman R, Schirrmacher R, Rosch F. J Label Compd Radiopharm 2003;
46: S207.
30. Chin FT, Musachio JL, Cai L, Pike VW. J Label Compd Radiopharm 2003; 46:
S172.
31. Loustau-Then I, Ponchant M, Fuseau C, Karmenka JM, Vignon J, Crouzel C.
Nucl Med Biol 1997; 24: 513–518.
32. Mulholland GK, Jung YW, Wieland DM, Kilbourn MR, Kuhl DE. J Label
Compd Radiopharm 1993; 33: 583–591.
33. Chi DY, Kilbourn MR, Katzenellenbogen JA, Welch MJ. J Org Chem 1987, 52:
658–664.
34. Akturk Z, Filic F, Erol K, Pabuccuoglu V. Il Farmaco 2002; 57: 201–206.
35. Shahak I, Bergmann ED. J Chem Soc (C) 1967: 319–320.
36. Baillet S, Buisine E, Horvath D, Maes L, Bonnet B, Sergheraert C. Bioorg Med
Chem 1996; 4: 891–899.
.
37. Hamacher K, Coenen HH, Stocklin G. J Nucl Med 1986; 27: 235–238.
38. Liotta CL, Harris HP. J Am Chem Soc 1974; 96: 2250–2252.
39. Albanese D, Landini D, Penso M. J Org Chem 1998; 63: 9587–9589.
40. Landini D, Maia A, Rampoldi A. J Org Chem 1989; 54: 328–332.
41. Conant JB, Kirner WR, Hussey RE. J Am Chem Soc 1925; 47: 488–501.
42. Bordwell FG, Brannen Jr WT. J Am Chem Soc 1964; 86: 4645–4650.
43. Kilbourn MR, Brodack JW, Chi D, Dence CS, Jerabek PA, Katzenellenbogen
JA, Patrick TB, Welch MJ. J Label Compd Radiopharm 1986; 23: 1174–1176.
44. Korguth ML, DeGrado TR, Holden JE. J Label Compd. Radiopharm 1988; 25:
369–380.
45. Brodack JW, Kilbourn MR, Welch MJ, Katzenellenbogen JA. Appl Radiat Isot
1986; 37: 217–221.
46. Feller D. J Phys Chem 1997; 101: 2723–2731.
47. Krongsuk S, Kerdcharoen T, Hannongbua S. J Phys Chem B 2003; 107: 4175–
4181, and references therein.
"
48. Rogers GA, Stone-Elander S, Ingvar M, Eriksson L, Parsons SM, Widen L. Nucl
Med Biol 1994; 21: 219–230.
49. Gambaretto G, Conte L, Fornasieri G, Zarantonello C, Tonei D, Sassi A, Bertani
R. J Fluor Chem 2003; 121: 57–63.
50. Choe YS, Lidstrom PJ, Chi DY, Bonasera TA, Welch MJ, Katzenellenbogen JA.
J Med Chem 1995; 38: 816–825.
51. Hendrix JA, Stefany DW. Tetrahedron Lett 1999; 40: 6749–6752.
52. Zucchi G, Scopelliti R, Buenzli JC. J Chem Soc Dalton Trans 2001; 13: 1975–1985.
53. Diehl K, Himbert G, Henn L. Chem Ber 1986; 119: 2430–2443.
54. Kano S, Taniguchi K. Japan Pat 1966; 14,987. Chem Abstr 1966; 65: 20055e.
55. Wimalesema K, May SW. J Am Chem Soc 1987, 109: 4036–4046.
56. Fuchigami T, Ichikawa S. J Org Chem 1994; 59: 607–615.
Copyright # 2004 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2004; 47: 217–232