High efficiency synthesis of F-18 fluoromethyl ethers: An attractive alternative for C-11 methyl groups in positron emission tomography radiopharmaceuticals

Article abstract of DOI:10.1021/ol401792n

A rapid and efficient method for the synthesis of O-fluoromethyl aliphatic and aromatic ethers is presented. This method is so mild that it can be used for the preparation of positron emission tomography (PET) radiopharmaceuticals bearing O-[¹⁸F]fluoromethyl groups.

Full text of DOI:10.1021/ol401792n

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
High Efficiency Synthesis of F‑18
Fluoromethyl Ethers: An Attractive
Alternative for C‑11 Methyl Groups
in Positron Emission Tomography
Radiopharmaceuticals
Chansoo Park,† Byoung Se Lee,‡ and Dae Yoon Chi*,†,‡
Department of Chemistry, Sogang University, 35 Baekbeomro Mapogu, Seoul 121-742,
Korea, and Research Institute of Labeling, FutureChem Co., Ltd.,
388-1 Pungnap-2-dong Songpagu, Seoul 138-736, Korea
dychi@sogang.ac.kr
Received June 28, 2013
ABSTRACT
A rapid and efficient method for the synthesis of O-fluoromethyl aliphatic and aromatic ethers is presented. This method is so mild that it can
be used for the preparation of positron emission tomography (PET) radiopharmaceuticals bearing O-[18F]fluoromethyl groups.
The carbon-11 radionuclide, widely used in positron-
emission tomography (PET), is generally produced locally
by cyclotron bombardment, according to the 14N(p,R)11C
nuclear reaction.1 The primary radiochemical prod-
ucts, [11C]CO2 or [11C]CH4, are then transformed into
other precursors ([11C]CO, [11C]CH3OH, [11C]CH3I,
[11C]CH3OTf, [11C]HCN, etc.) that are used for radiola-
beling more complex PET radiopharmaceuticals, with
heteroatom methylations (CH3ÀX; X = O, N, S) with
CH3I or CH3OTf being most commonly used for this
purpose. These C-11 labeled radiopharmaceuticals are
good for research; their commercialization, however, is
production and distribution.2À5 In the case of fluoro-
alkoxy groups (X = O), Tsukada et al. showed a nice
(2) (a) Coenen, H. H.; Elsinga, P. H.; Iwata, R.; Kilbourn, M. R.;
Pillai, M. R. A.; Rajane, M. G. R.; Wagner, H. N., Jr.; Zaknun, J. J.
Nucl. Med. Biol. 2010, 37, 727–740. (b) Zhang, M.-R.; Suzuki, K. Curr.
Top. Med. Chem. 2007, 7, 1817–1828.
(3) X = O: (a) Urakami, T.; Sakai, K.; Asai, T.; Fukumoto, D.;
Tsukada, H.; Oku, N. Nucl. Med. Biol. 2009, 36, 295–303.
(b) Kawamura, K.; Tsukada, H.; Shiba, K.; Tsuji, C.; Harada, N.; Kimura,
Y.; Ishlwata, K. Nucl. Med. Biol. 2007, 34, 571–577. (c) Kilbourn, M. R.;
Hockley, B.; Lee, L.; Hou, C.; Goswami, R.; Ponde, D. E.; Kung, M.-P.;
Kung, H. F. Nucl. Med. Biol. 2007, 34, 233–237. (d) Tsukada, H.; Sato, K.;
Fukumoto, D.; Kakiuchi, T. Eur. Nucl. Med. Mol. Imaging 2006, 33, 1017–
1024. (e) Tsukada, H.; Sato, K.; Fukumoto, D.; Nishiyama, S.; Harada, N.;
Kakiuchi, T. J. Nucl. Med. 2006, 47, 679–688. (f) Suzuki, M.; Yamaguchi,
K.; Honda, G.; Iwata, R.; Furumoto, S.; Jeong, M.-g.; Fukuda, H.; Itoh, M.
Annals Nucl. Med. 2005, 19, 589–595.
(4) X = N: (a) Shao, X.; Hockley, B. G.; Hoareau, R.; Schnau, P. L.;
Scott, P. J. H. Appl. Radiat. Isot. 2011, 69, 403–409. (b) Hodolic, M.
Radiol. Oncol. 2011, 45, 17–21. (c) Gao, M. Z.; Miller, M. A.; DeGrado,
T. R.; Mock, B. H.; Lopshire, J. C.; Rosenberger, J. G.; Dusa, C.; Das,
M. K.; Groh, W. J.; Zipes, D. P.; Hutchins, G. D.; Zheng, Q.-H. Bioorg.
Med. Chem. 2007, 15, 1289–1297. (d) Yamaura, G.; Yoshioka, T.;
Fukuda, H.; Yamaguchi, K.; Suzuki, M.; Furumoto, S.; Iwata, R.;
Ishioka, C. Eur. Nucl. Med. Mol. Imaging 2006, 33, 1134–1139. (e) Chi,
D. Y.; Kilbourn, M. R.; Katzenellenbogen, J. A.; Welch, M. J. J. Org.
Chem. 1987, 52, 658–664.
very challenging because the short half-life of C-11 (t1/2
=
20 min) makes mass production and FDA-mandated
cGMP qualification analysis very impractical. Thus, many
researchers have sought F-18 labeled alternatives for the
RX-11CH3 group, namely, RX-CH218F, RX-CH2CH218F,
and RX-CH2CH2CH218F groups, which, because of the
longer half-life of F-18 (t1/2 = 110 min), enable regional
† Sogang University.
(5) X = S: (a) Robins, E. G.; Zhao, Y. J.; Khan, I.; Wilson, A.;
˚
‡ FutureChem Co., Ltd.
Luthra, S. K.; Arstad, E. Bioorg. Med. Chem. Lett. 2010, 20, 1749–1751.
(b) Zessin, J.; Eskola, O.; Brust, P.; Bergman, J.; Steinbach, J.; Lehikoinen,
(1) (a) Pimlott, S. L.; Sutherland, A. Chem. Soc. Rev. 2011, 40, 149–
162. (b) Li, Z.; Conti, P. S. Adv. Drug Delivery Rev. 2010, 62, 1031–1051.
P.; Solin, O.; Johannsen, B. Nucl. Med. Biol. 2001, 28, 857–863.
r
10.1021/ol401792n
XXXX American Chemical Society

comparison of the D- and L-isomers of O-18F-fluoro-
methyl, O-18F-fluoroethyl, and O-18F-fluoropropyl tyro-
sine as tumor imaging agents in mice.3d Inthis comparison,
D-isomers were more metabolically stable in vivo than
L-isomers, but the stabilities of different chain lengths,
from fluoromethyl to fluoropropyl, were similar. It is
notable, however, that due to synthetic difficulties, there
are very few reports of the synthesis of RX-CH218F
systems; typical radiochemical yields are less than 5%,
which is insufficient for commercial production.6 For
RNHCH2CH218F and RNHCH2CH2CH218F syntheses,
massproduction wasstill difficult until ourdevelopment of
tertiary alcohol labeling conditions.7 Fluoroalkyl groups
have also been added to aryl or alkenyl units,8 and a study
of the in vivo stability of fluoromethyl, fluoroethyl, and
fluoropropyl arenes showed the fluoromethyl to be only
somewhat more unstable than the others.8c There are,
of course, many RX-11CH3 compounds that behave well
as PET radiopharmaceuticals,9 but if one replaces the
11C-methyl group with a fluoroethyl or fluoropropyl
group, the analog becomes not only larger but also more
lipophilic by the addition of a CH218F or CH2CH218F unit
onto the methyl group. Given the small size of a fluorine
atom, replacement of a 11C-methyl group with a 18FCH2
group is a more isostructural change.
Table 1. Optimization of Fluorination of 1,2,3-Triazolium
Triflate 1a
yieldb (%)
temp
time
(h)
entry
MF
KF
solvent
(°C)
1
1p
1c
2c
3
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
t-BuOH
acetone
THF
80
80
rt
24
100
57
0
CsF
24
38
19
64
89
81
45
85
79
37
19
62
84
d
TBAF
TBAF
TBAF
TBAF
TBAF
TBAF
TBAF
KF/K222
KF/K222
KF/K222
KF/K222
24
À
d
4
80
80
80
80
80
80
rt
15 min
À
d
5
1
À
d
6
1
À
d
7
1
À
d
8
40 min
20 min
24
À
d
9
DMF
À
d
10
11
12
13
CH3CN
CH3CN
CH3CN
CH3CN
À
d
40
60
80
1
À
d
1
À
d
15 min
À
During a study of alcohol protecting groups, we exam-
ined a triazolium methyl group (compound 1) as a replace-
ment for a benzyl group, and we found that this group
functioned as a very good nucleofuge for displacement
by the fluoride ion. Herein, we describe a convenient
method for the fluoromethylation of alcohols that can be
used for the synthesis of O-[18F]fluoromethylated PET
radiopharmaceuticals.
a All reactions were carried out on a 1.0 mmol reaction scale of
triazolium salt 1 using 1.5 mmol of fluoride in 4.0 mL of organic solvent
in a sealed reaction vial. b Isolated yield. c 3 mmol of fluoride were used. d The
triflate anion of starting material was immediately changed to fluoride.
in Supporting Information (SI)): Upon treatment with
sodium azide, O-chloromethyl ethers were converted to
O-azidomethyl ethers; the triazole was formed by a Click
reaction with an alkyne in EtOAc, and treatment with
methyl triflate in CH3CN gave the 1,2,3-triazolium tri-
flates. Most of the starting 1,2,3-triazolium triflates and
imidazolium triflate are stable after freezer storage for
two months at À15 °C (see details in SI). As shown in
Table 1, triazolium salt 1, prepared as a model substrate,
was subjected to nucleophilic fluorination under various
reaction conditions. When 1 was treated with 3 equiv of
KF at 80 °C for 24 h in CH3CN, starting material (1) was
recovered quantitatively (entry 1). However, treatment
with CsF under the same conditions gave the desired
O-fluoromethylated product 1p in 38% yield, with 57%
recovered starting material 1 (entry 2). When TBAF was
used, the reaction was complete within 1 h in good yield
(64% at 15 min, entry 4; 89% at 1 h, entry 5). The reaction
proceeded slowly at rt (entry 3), but curiously, starting
material 1 could not be recovered. Careful NMR analysis
showed that under these conditions the triflate anion of 1
was immediately replaced by the fluoride ion; this anion
exchange did not occur under the conditions of entries 1
and 2, presumably because the fluoride ion has a very
tight ionic interaction with potassium or cesium (see
NMR in SI). In searching for the optimal reaction
solvent, we explored fluorination with TBAF at 80 °C
in other solvents (t-BuOH, acetone, THF, and DMF)
The starting 1,2,3-triazolium triflates were prepared
from the corresponding chloromethyl ethers (see procedure
(6) (a) Neal, T. R.; Apana, S.; Berridge, M. S. J. Label. Compd.
Radiopharm. 2005, 48, 557–568. (b) Zhang, M.-R.; Ogawa, M.;
Furutsuka, K.; Yoshida, Y.; Suzuki, K. J. Fluorine Chem. 2004, 125,
1879–1886. (c) Zheng, L.; Berridge, M. S. Appl. Radiat. Isot. 2000, 52,
55–61. (d) Iwata, R.; Pascali, C.; Bogni, A.; Furumoto, S.; Terasaki, K.;
Yanai, K. Appl. Radiat. Isot. 2002, 57, 347–352.
(7) (a) Kim, D. W.; Ahn, D.-S.; Oh, Y.-H.; Lee, S.; Kil, H. S.; Oh,
S. J.; Lee, S. J.; Kim, J. S.; Ryu, J. S.; Moon, D. H.; Chi, D. Y. J. Am.
Chem. Soc. 2006, 128, 16393–16397. (b) Lee, S. J.; Oh, S. J.; Chi, D. Y.;
Kang, S. H.; Kil, H. S.; Kim, J. S.; Moon, D. H. Nucl. Med. Biol. 2007,
34, 345–351. (c) Lee, S. J.; Oh, S. J.; Chi, D. Y.; Kil, H. S.; Kim, E. N.;
Ryu, J. S.; Moon, D. H. Eur. J. Nucl. Med. Mol. Imaging 2007, 34, 1406–
1409. (d) Lee, S.-S.; Kim, H.-S.; Hwang, T.-K.; Oh, Y.-H.; Park, S.-W;
Lee, S.; Lee, B. S.; Chi, D. Y. Org. Lett. 2008, 10, 61–64. (e) Lee, S. J.; Oh,
S. J.; Chi, D. Y.; Lee, B. S.; Ryu, J. S.; Moon, D. H. J. Label. Compd.
Radiopharm. 2008, 51, 80–82. (f) Kim, D. W.; Jeong, H.-J.; Lim, S. T.;
Sohn, M.-H.; Katzenellenbogen, J. A.; Chi, D. Y. J. Org. Chem. 2008,
73, 957–962. (g) Shinde, S. S.; Lee, B. S.; Chi, D. Y. Org. Lett. 2008, 10,
733–735. (h) Kim, D. W.; Jeong, H.-J.; Lim, S. T.; Sohn, M.-H.; Chi,
D. Y. Tetrahedron 2008, 64, 4209–4214.
(8) (a) Doi, H.; Ban, I.; Nonoyama, A.; Sumi, K.; Kuang, C. X.;
Hosoya, T.; Tsukada, H.; Suzuki, M. Chem.;Eur. J. 2009, 15, 4165–
4171. (b) Sachin, K.; Jeong, H.-J.; Lim, S. T.; Sohn, M.-H.; Chi, D. Y.;
Kim, D. W. Tetrahedron 2011, 67, 1763–1767. (c) Lee, K. C.; Lee, S.-Y.;
Choe, Y. S.; Chi, D. Y. Bull. Korean Chem. Soc. 2004, 25, 1225–1230. (d)
Lee, B. S.; Chu, S.; Lee, K. C.; Lee, B.-S.; Chi, D. Y.; Kim, S. E.; Choe,
Y. S.; Song, Y. S.; Jin, C. Bioorg. Med. Chem. 2003, 11, 4949–4958. (e)
Lee, K. C.; Chi, D. Y. J. Org. Chem. 1999, 64, 8576–8581. (f) Choe, Y. S.;
Song, D. H.; Lee, K.-J.; Kim, S. E.; Choi, Y.; Lee, K. H.; Kim, B.-T.; Oh,
S. J.; Chi, D. Y. Appl. Radiat. Isot. 1998, 49, 73–77.
(9) Tu, Z.; Mach, R. H. Curr. Top. Med. Chem. 2010, 10, 1060–1095.
B
Org. Lett., Vol. XX, No. XX, XXXX

(entries 6À10). All solvents gave comparable yields, ex-
cept acetone (entry 7). Because a short reaction period is
essential for F-18 labeling, the more rapid reaction in
DMF, which was complete in 20 min, would be advanta-
geous (entry 9). Another powerful fluorination reagent
KF/K222 was also tested in CH3CN and proved to be
superior to the others (entries 10À13), giving some pro-
duct even at rt (entry 10) and complete reaction within
15 min at 80 °C (entry 13). As the reaction even proceeds
at 60 °C (entry 12), this fluoromethylation protocol
appears promising for labeling thermally sensitive mate-
rials, such as amino acids, oligopeptides, and potentially
even proteins, without causing epimerization.
CH3CN) was applied to more complicated triazolium salts
(Scheme 1), and in all cases the desired products were
obtained in comparable yields. The precursor 7, having
both a primary mesylate and a triazolium salt, gave
principally the desired fluoromethyl compound 7p in
67% yield, along with some difluorinated product (7%).
Even though a mesylate is a better leaving group than a
triazolium salt, these oxymethylene triazolium salts have
very high reactivity toward fluoride ion substitution because
the ether oxygen labilizes the CÀN bond and would also
stabilize any incipient carbocation character in the transition
state. Thus, fluoride attack on the methylene group is doubly
activated, by oxygen as well as by the triazolium group.
Table 2. Optimization of Substituents in Fluorination of
1,2,3-Triazolium Triflates 1À5 and Imidazolium Triflate 6a
Scheme 1. Fluorinations of Various Compounds (Isolated
Yield)
entry
X
R, compound
yieldb
1
2
3
4
5
6
N
-t-Bu, 1
66
42
87
72
89
57
N
-COOCH3, 2
-Ph, 3
N
N
-4-CH3O-Ph, 4
-3,5-(CF3)2-Ph, 5
-Ph, 6
N
CH
a All reactions were carried out on a 1.0 mmol reaction scale of
triazolium salt using 1.5 mmol of fluoride in 4.0 mL of CH3CN.
b Isolated yield.
To find the optimal heterocyclic leaving group for this
fluorination reaction, four triazolium salts and an imida-
zolium salt were prepared and examined under the same
fluorination conditions (1.5 equiv of TBAF, 80 °C, 10 min,
CH3CN, Table 2). We expected that electron-withdrawing
groups (EWGs) on the triazolium salt might accelerate the
reaction rate by labilizing the methyleneÀnitrogen bond of
the precursor and electron-donating groups (EDG) would
decrease the rate by stabilizing this bond. Although
measuring yields at 10 min does not provide accurate
reaction kinetics, the triazolium salt with an EWG, (3,5-
ditrifluoromethyl)phenyl, gave the best yield (89%, entry
5). The triazolium salt with a methyl ester (EWG), how-
ever, gave a lower yield (42%, entry 2), presumably
because of poor precursor stability. Other triazolium salt
precursors afforded comparable yields (entries 1, 3, 4),
implying that the reaction proceeded rapidly regardless of
the electronic nature of the salt. Comparison of phenyl-
substituted triazolium and imidazolium salts (entries 3, 6)
showed that the triazolium salt (87%) was superior to the
imidazolium salt (57%).
A variety of structurally diverse phenol derivatives 8À13
were smoothly converted to the corresponding fluoro-
methylated phenyl products 8pÀ13p in moderate to ex-
cellent yield. Electronic environment variations on the
aromatic substituent had little influence on the reaction
although EWGs seemed to increase the reaction rate. The
precursor bearing a 3-acetamidophenol derivative 12 also
gave a good yield of 12p (76%), indicating tolerance of the
amide functionality. Similarly, the steroid having a ketone
also gave a good yield (13p), implying similar tolerance of
carbonyl groups. Fluorination on the benzylic (14), pri-
mary alkyl (15), or secondary alkyl (16) positions also
proceeded smoothly to afford the fluoromethoxy products
in very good yields (77À79%).
After establishing the optimal reaction conditions and
best salt form, the reaction (1.5 equiv of TBAF, 80 °C,
With completion of the F-19 fluorination study with a
variety of precursors, this protocol was extended to F-18
Org. Lett., Vol. XX, No. XX, XXXX
C

Scheme 2. Fluorinations of Various Compounds with
[18F]Fluoride (Nondecay Corrected Isolated Radiochemical Yield)
Figure 1. Proposed mechanism of intramolecular fluorination
for the synthesis of fluoromethyl ether (numbers are chemical
shifts of protons).
later at 0 °C after adding TBAF, due to the acidity of that
proton.10 At 60 min later at rt, the tiny peak at 9.63 ppm
which is presumably the C5 proton peak of triazolium
fluoride was shown. At 10 min later at 80 °C, the tiny peak
at 9.63 ppm disappeared with completion of fluorination.
Based on this experiment, the proposed mechanism of
intramolecular fluorination for the synthesis of fluoro-
methyl ether was shown in Figure 1. N-Heterocyclic zwit-
ter ionic carbene is a favored form in equilibrium at rt. The
displacement of fluoride proceeds slowly at 80 °C from
triazolium fluoride.
In conclusion, the method we have described for the
preparation of O-fluoromethylated aliphatic and aromatic
ethers should enable the facile preparation of a number
of compounds bearing O-fluoromethyl groups of varying
structure. In addition, because this method is efficient
and rapid, it will be of use for the preparation of
O-fluoromethylated compounds labeled with the PET radio-
nuclide fluorine-18 (t1/2 = 110 min) and could expand the
availability of commercially viable F-18 labeled radio-
pharmaceuticalsconsiderably. Weare currently expanding
this chemistry to S- and N-fluoromethylation.
radiolabeling. Several selected precursors were sub-
jectedto a [18F]fluorination condition ([18F]TBAF, 110 °C,
tert-amyl alcohol) to check the utility of this method in
F-18 labeling. The results are given in Scheme 2, and
overall, reasonable yields were obtained under favorable
reaction conditions. [18F]Fluorination of 1 afforded
[18F]fluoromethoxynaphthalene ([18F]1p) in 58% isolated,
nondecay corrected radiochemical yield after 10 min. The
overall preparing time including HPLC was 45 min with
230 GBq/μmol specific activity. In other cases, the corre-
sponding [18F]fluorine-substituted compounds were pro-
duced in great yields, except for precursor 10 including
EDGssuchasdimethoxygroups. A noteworthy advantage
of using this triazolium salt as a labeling precursor is that
it is so simple to separate the small amount of labeled
product from excess starting material by simple chromato-
graphic retention of the precursor salt and elution of the
neutral product.
To find the effect of the triazolium ring in a fluorination
reaction, we performed the fluorination reaction in an
NMR tube. All protons’ peaks of starting compound 1
changed their chemical shifts when tetrabutylammonium
fluoride (TBAF) was added (see SI, Figure S5). NMR
spectra were taken before adding TBAF, 10 min later at
0 °C after adding TBAF, at 60 min later at rt after removal
of the ice bath, and 10 min later at 80 °C. The C5 proton
peak shown at 8.60 ppm completely disappeared at 10 min
Acknowledgment. This research was supported by the
Converging Research Program (2012K001486) through
the Ministry of Science, ICT & Future Planning. We thank
Prof. John A. Katzenellenbogen, Dept. of Chemistry,
University of Illinois at UrbanaÀChampaign for helpful
discussions.
Supporting Information Available. Typical procedure
for the preparation of starting triazolium salts and fluor-
ination, F-18 fluorination, NMR experiment for finding
mechanism of fluorination, and NMR spectra. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(10) Massey, R. S.; Collett, C. J.; Lindsay, A. G.; Smith, A. D.;
O’Donoghue, A. C. J. Am. Chem. Soc. 2012, 134, 20421–20432.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX