A simple method to generate [18F]triflyl fluoride for 18F radiosynthesis

Article abstract of DOI:10.1016/j.tetlet.2021.153273

A simple, continuous-flow solid-phase radiosynthesis method has been developed for the generation of [¹⁸F]triflyl fluoride as a source of [¹⁸F]fluoride for the preparation of ¹⁸F labeled radiopharmaceuticals without the ne

Full text of DOI:10.1016/j.tetlet.2021.153273

Tetrahedron Letters 78 (2021) 153273
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
A simple method to generate [18F]triflyl fluoride for 18F radiosynthesis
b
Dong Zhou a, , John A. Katzenellenbogen
⇑
a Department of Radiology, School of Medicine, Washington University in Saint Louis, Saint Louis, MO 63110, United States
b Department of Chemistry and Cancer Center at Illinois, University of Illinois at Urbana-Champaign, IL 61801, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
A simple, continuous-flow solid-phase radiosynthesis method has been developed for the generation of
[
18F]triflyl fluoride as a source of [18F]fluoride for the preparation of 18F labeled radiopharmaceuticals
Received 19 April 2021
Revised 18 June 2021
Accepted 28 June 2021
Available online 10 July 2021
without the need for an azeotropic drying step. After [18F]fluoride was trapped and dried on an anion-
exchange resin cartridge, gaseous [18F]triflyl fluoride was generated directly by eluting the cartridge with
a solution of PhN(OTf)2, followed by isolation from the eluted reaction mixture via an empty cartridge
and conversion to [18F]fluoride in a basic solution for radiofluorination. A peristaltic pump was used to
maintain flow through the above stages, eventually providing reaction-ready [18F]fluoride in over 90%
radiochemical yield in less than 10 min. Up to 0.057 mg non-radioactive fluoride was introduced from
the reagents and components used during the above processes. Several radiolabeling reactions were car-
ried out using the [18F]fluoride generated in this manner, affording the labeled products in good to high
radiochemical yields. This method should have great potential for the safe and convenient preparation of
18F labeled radiopharmaceuticals.
Keywords:
[
18F]triflyl fluoride
Radiofluorination
Peristaltic pump
Generation of [18F]fluoride
Continuous flow solid-phase radiosynthesis
Ó 2021 Elsevier Ltd. All rights reserved.
Introduction
radiolabeling [8]. A rapid and reliable method for generating [18F]
triflyl fluoride ([18F]TfF) as a gaseous source of [18F]fluoride has been
Fluorine-18 labeled radiopharmaceuticals play an important
role in the field of nuclear imaging using positron emission tomog-
raphy (PET), and they are often prepared from [18F]fluoride pro-
duced by an 18O (p, n)18F reaction by proton irradiation of 18O
enriched water (95%) using a medical cyclotron. The hydrated
reported [2]. This method includes trapping [18F]fluoride on an
anion-exchange resin cartridge, eluting [18F]fluoride with an aque-
ous potassium salt solution, converting it to volatile [18F]triflyl flu-
oride using a bistriflate precursor PhN(OTf)2, distilling and drying
this gas through a column of P2O5, and trapping the activity with
base in a solvent suitable for radiolabeling. The suitability of using
[
18F]fluoride in the 18O water needs to be converted to more reac-
tive forms in an anhydrous solvent suitable for radiolabeling,
except for reactions that can be carried out in water [1]. Azeotropic
drying with acetonitrile is the most commonly used method for
this purpose but has many drawbacks [2]. Therefore, significant
efforts have been undertaken to develop methods for labeling
without this azeotropic drying step, but applications of these
methods have been limited [3].
Fluorine-18 labeled sulfonyl fluorides have shown promise as a
source of [18F]fluoride for the preparation of 18F labeled radiophar-
maceuticals [2,4–7]. This strategy takes advantage of the high affin-
ity of sulfonyl groups for fluoride, and the quantitative conversion of
sulfonyl fluorides to the fluoride ion under basic conditions, thereby
enabling the conversion of [18F]fluoride from an aqueous solution in
the 18O water to an anhydrous organic solvent that is suitable for
[
18F]TfF for radiolabeling has been demonstrated by model reac-
tions and the preparation of 18F labeled radiopharmaceuticals [2].
In this report, we present a simple method (continuous-flow
solid-phase radiosynthesis) for the generation of gaseous [18F]TfF
for radiolabeling, using a peristaltic pump for loading, drying,
eluting, and distilling/converting the radioactivity in a safe and
controlled manner. [18F]TfF was generated directly on the anion-
exchange resin cartridge (generally termed solid-phase extraction
(SPE) cartridge) and separated from other components via an
empty cartridge. Some commonly used radiolabeling reactions
using [18F]TfF generated in this manner are reported, and the fac-
tors important for the preparation of 18F labeled radiopharmaceu-
ticals are also discussed.
Results and discussion
⇑
Corresponding author at: 510 S. Kingshighway Blvd, Saint Louis, MO 63110,
United States.
[
18F]Fluoride in 18O water from the cyclotron target was first
trapped in an anion-exchange resin cartridge to allow recovery of
E-mail address: zhoud@wustl.edu (D. Zhou).
https://doi.org/10.1016/j.tetlet.2021.153273
0040-4039/Ó 2021 Elsevier Ltd. All rights reserved.

D. Zhou and J.A. Katzenellenbogen
Tetrahedron Letters 78 (2021) 153273
the expensive isotopically enriched water, a procedure that is rou-
tinely used for the azeotropic drying protocol. We used an in-house
made anion-exchange resin cartridge (AG MP-1MÀHCO3), which is
equivalent to a 30-PS-HCO3 cartridge. Next, the [18F]fluoride and
the cartridge were dried by rinsing with acetonitrile. A manual elu-
tion with PhN(OTf)2 (5 mg) in acetonitrile resulted in release of
over 95% radioactivity from the cartridge within a few seconds,
with [18F]TfF as the only radioactive product (see Fig. S1). However,
while PhN(OTf)2 was still present in the eluted solution, a test
labeling directly with the eluted solution resulted in no radiofluo-
rination, presumably due to the interference with the large amount
of PhN(OTf)2. This initial result led us to further explore the elution
of radioactivity in a more controlled manner and to develop a
method to separate [18F]TfF from PhN(OTf)2 in the eluted solution.
Peristaltic pumps are typically used in medical devices to pump
clean/sterile fluids and in industry for containing highly reactive
chemicals during their transfer; for both of these purposes there
is a critical need that the flowing components be isolated from
the environment. As illustrated in Figs. 1 and 2, a peristaltic pump
can accomplish all of the tasks needed for the generation of gas-
eous [18F]TfF and can do so in a safe and controlled fashion. Differ-
ent solvents were first tested for the elution of [18F]TfF. Acetonitrile
provided the highest elution efficiency (96%) (Table 1), and it is
compatible with the pump material; therefore, it was selected as
the solvent for elution.
Fig. 2. Continuous-flow solid-phase apparatus for the synthesis of [18F] triflyl
fluoride for radiolabeling.
Table 1
Effect of eluting solvents on efficiency.
Entry
Solvent (mL)a
EE (%)b
1
3
4
5
MeCN (0.5 mL)
DMSO (0.5 mL)
Amyl alcohol (0.5 mL)
THF (0.5 mL)
96
68
94
95
Note: a. Eluted with Tf2NPh (5 mg) at 3 mL/min and distilled at 10 mL/min; b.
Eluting efficiency (EE) (%) = 1-radioactivity left in SPE/total radioactivity.
In the previous report, an inert gas was used to distill [18F]TfF
from the reaction mixture/vessel [2]. In our investigation, several
designs, an empty cartridge, a vial with needles, and cartridges
containing inert materials (illustrated in Table 2 and Fig. S2) were
tested in order to separate [18F]TfF from the eluted solution, with
the peristaltic pump providing the gas flow for distillation. Both
the empty cartridge and the vial methods worked equivalently
well, with less than 1% of radioactivity left in them. However, the
empty cartridge is a much simpler design. The inert material
design gave variable results, with an alumina N cartridge com-
pletely trapping the radioactivity. We took advantage of this by
using an alumina N cartridge as a vent at the end of the apparatus
to trap any escaped radioactivity. It is worthwhile mentioning that
as little as 15 mL air was needed for the efficient distillation of [18F]
TfF from the eluted solution. This low volume is ideal for the trap-
ping/converting step that follows.
Table 2
Summary of separators.
Entry
Type
Radioactivity left
1
2
3
4
5
Empty cartridge
Vial/Needle
Na2SO4
Silica gel
Alumina N
<1%
<1%
3.9 1.2% (n = 3)
27%
100%
cient conversion. When a 1:1 ratio is used, only 15.7% was trapped,
indicating that the conversion of TfF to fluoride is not ‘‘instant”.
Increasing the trapping temperature, however, did not help the
conversion.
The trapping of [18F]TfF with potassium bicarbonate/Kryptofix
222 (KHCO3/K222) was studied extensively in the previous report
[2]. In our investigation, the commonly used potassium carbon-
ate/Kryptofix 222 (K2CO3/K222) was studied. As shown in Table 3,
the trapping efficiency is dependent on the amount and type of
base, the type of solvent, the amount of fluoride, and the flow rate.
Contrary to the previous report, we found K2CO3/K222 to be better
at trapping than KHCO3/K222. At the flow rate of 10 mL/min, [18F]
TfF was trapped by 1 mg (1.1 mmol) K2CO3/K222 in 84.8% vs 32.3%
by 1.4 mg (2.9 mmol) KHCO3/K222. When amyl alcohol, a unique
solvent for radiofluorination, was used [9], the trapping efficiency
dropped significantly, from 97.8% in acetonitrile to 31.9% in amyl
alcohol. In some cases, non-radioactive fluoride was added to
The amount of non-radioactive fluoride released from the com-
ponents used in overall process was determined by derivatizing
fluoride as tosyl fluoride, which was then measured by HPLC
[10]. Excluding the fluoride from 18O water, it was found that a
total of 0.057 mg fluoride (including up to 0.008 mg from the
MP-1MÀHCO3 resin) was introduced from the overall process
using 10 mg PhN(OTf)2.
The reactivity of [18F]fluoride generated in this manner was
tested using several precursors, including those used to produce
FDG, 18F glutamine (FGln) [11] and [18F]FluorThanatrace (FTT)
[12], and the results are shown in Figure 3 and Table S1. The radio-
chemical conversions are comparable to those previously reported,
suggesting that [18F]fluoride generated by our method has high
reactivity. The radiosynthesis of [18F]FGln used only 2 mg precur-
sor and 1.5 mg K2CO3/K222 and afforded 59% radiochemical conver-
sion of the intermediate. Starting from 43 mCi of [18F]fluoride in
18O water, conversion via [18F]TfF to reactive [18F]fluoride in ace-
tonitrile was achieved in 89% radiochemical yield (RCY). [18F]FTT
was then synthesized from this converted [18F]fluoride in 57%
RCY (decay corrected), with molar activity of 43.7 GBq/mmol
(1182 mCi/mmol), and used for an animal study (conducted at
Washington University in St Louis (WUSTL) under WUSTL Animal
Studies Committee IACUC-approved protocols). This RCY is compa-
rable to those obtained using conventional methods [13], and the
molar activity of the starting [18F]fluoride, measured as [18F] tosyl
[
18F]fluoride to simulate the use of large amounts of radioactivity
(0.1 mmol and 1 mmol fluoride from 37 GBq (1000 mCi) with molar
activity of 370 GBq /mmol or 37 GBq /mmol, respectively). At least a
50-fold excess amount of K2CO3/K222 over TfF is needed for effi-
Fig. 1. Chemical transformation of [18F]fluoride from aqueous solution to organic
solution for labeling.
2

D. Zhou and J.A. Katzenellenbogen
Tetrahedron Letters 78 (2021) 153273
Table 3
Summary of the trapping efficiency of [18F]TfF.
Entry
Basea
(mg)
FR (mL/min)b
TE (%)c
Note
1
3
4
5
6
7
8
9
10
11
12
13
K2CO3/K222
K2CO3/K222
K2CO3/K222
K2CO3/K222
K2CO3/K222
K2CO3/K222
K2CO3/K222
K2CO3/K222
K2CO3/K222
K2CO3/K222
KHCO3/K222
KHCO3/K222
5
5
5
5
2.5
2.5
1
1
1
3
96.4 2.6 (n = 5)
96.5 2.1 (n = 6)
98.3 0.1 (n = 2)
91.4
97.8 1.2 (n = 3)
31.9
92
84.8
–
–
10
10
3
3
3
0.1 mmol fluoride (KF) added
Solvent: 1:4 MeCN/amyl alcohol
–
Solvent: amyl alcohol (300 mL)
–
–
Trapping at 105 °C
1 mmol fluoride (KF) added
–
–
3
10
10
3
3
10
76.3
15.7
1
4.76
1.4
98.2 0.5 (n = 2)
32.3
Note a. K2CO3/K222 1:2 (MW 891.19, 5 mg/5.6 mmol), KHCO3/K222 1:1 (MW 476.61, 4.76 mg/10 mmol) in acetonitrile (0.5 mL) except for as noted; b. FR: Flow rate of
distillation; c. TE: Trapping efficiency of [18F]TfF (%) = trapped radioactivity in the reactor/trapped radioactivity in the reactor and in the alumina N trap.
Fig. 3. Labeling precursors and their radiochemical conversions using [18F]TfF as the source of [18F]fluoride. Radiochemical conversion was determined by radio-TLC and
confirmed by radio-HPLC. Note: a. 0.1 mmol KF was added as carrier to simulate a large amount of radioactivity; b. HPLC isolated radiochemical yield (decay corrected).
fluoride, was 118.4 GBq/ mmol (3200 mCi/mmol) at the same time
point. This [18F] tosyl fluoride method [10] can be used to deter-
mine the molar activity of fluoride at any stage of the conversion
process.
with KHCO3/K222 prepared by dissolving the complex in anhydrous
acetonitrile, is expected to generate dry and highly reactive [18F]
fluoride. However, very dry [18F]fluoride is not always preferred
for radiolabeling because it may be absorbed non-specifically onto
the surface of reaction vessels and might result in more side reac-
tions (see Table S1). In this regard, we found that 20–30% of con-
verted [18F]fluoride in the acetonitrile/trapping container could
become insoluble if not used right away. This may have con-
tributed to the irreproducible results in the production of [18F]
FES and [18F]FET noted in the previous report [2]. Furthermore, it
has also been reported that long term storage of K2CO3/K222 in
stock solutions (water/acetonitrile) resulted in lower labeling
yields [14] and it is well known that overheating [18F]fluoride/
K2CO3/K222 during azeotropic drying can also reduce the yield of
labeling reactions. Hence, we believe that dry, water-free [18F]flu-
oride obtained by the conversion of 18F in the 18O water via [18F]TfF
under these mild and controlled protocols should be well suited for
18F fluorination in general, similar to that of other reagents ([18F]
AcF [15] and [18F]Me3SiF [16]) that have been used as the source
of [18F]fluoride for labeling. The trapping of [18F]TfF in the base
solution is critical for the application of this method in radiolabel-
ing. Compared with the conventional method which requires about
1 mg K2CO3 to efficiently elute radioactivity from a cartridge, much
less K2CO3 (0.15 mg) in K2CO3/K222 (1 mg) is needed to trap over
90% radioactivity transferred via [18F]TfF, but this depends to some
degree on the amount of radioactivity and its molar activity. The
small amount of base allows for the use of small amounts of a
base-liable precursor for labeling (e.g., [18F]FGln) and should
facilitate the purification step. We have used this method to
radiosynthesize [18F]FTT with trapping and the reaction conducted
The use of 18F labeled sulfonyl fluorides is a promising strategy
for the preparation of 18F labeled radiopharmaceuticals without
the need for an azeotropic drying step. [18F]TfF was generated
directly on the anion-exchange resin using the high affinity of
the sulfonyl group for fluoride, and was isolated from the reaction
mixture in gaseous form. While the high volatility of [18F]TfF
makes it extremely dangerous with respect to radiation safety,
the use of a peristaltic pump provides an ideal solution for effect-
ing the elution and distillation in a well-controlled manner that
isolates this radioactive gas from the environment. As shown in
Figure 2, fewer components are needed for the generation of
[
18F]TfF by our method than by the previously reported one, which
should result in more reliable production of [18F]TfF. The use of an
empty cartridge greatly simplifies the procedure for the distillation
of [18F]TfF. The small amount of non-radioactive [19F]fluoride,
which is found after the processing step (0.057 mg), is presumed
to originate from the resin (up to 0.008 mg from the MP-1MÀHCO3;
for comparison, 0.011 mg would come from 30-PS-HCO3) and PhN
(OTf)2 (10 mg). Consequently, lower amounts of PhN(OTf)2 are pre-
ferred to ensure that the final products are obtained in high molar
activity. This is a particular concern when a small amount of
radioactivity is to be used (e.g., 10 mCi), because even small
amounts of non-radioactive fluorine could lower the molar activity
of the labeled radiopharmaceuticals to levels that would be insuf-
ficient for imaging molecular targets of low concentration. The
water-free conversion of dry [18F]TfF to [18F]fluoride, especially
3

D. Zhou and J.A. Katzenellenbogen
Tetrahedron Letters 78 (2021) 153273
in one pot. The molar activity of [18F]FTT is somewhat less than
that of [18F]fluoride itself in the 18O water measured as [18F]TsF,
but the non-radioactive fluoride from the processing only con-
tributes 23.8 % (0.057 mg out of 0.24 mg) of total fluoride obtained
when 10 mg of PhN(OTf)2 is used for elution. Finally, it is notable
that by using the [18F] tosyl fluoride method [10] it is possible to
determine the molar activity of fluoride at any stage of the conver-
sion and radiolabeling processes, and thereby identify steps that
might be compromising the utility of the final tracer material
due to reduced molar activity and implement changes to minimize
this dilution.
References
[1] L. Cai, S. Lu, V.W. Pike, Chemistry with [18F]Fluoride Ion, Eur. J. Org. Chem. 2008
(17) (2008) 2853–2873.
[2] A. Pees, C. Sewing, M.J.W.D. Vosjan, V. Tadino, J.D.M. Herscheid, A.D.
Windhorst, D.J. Vugts, Fast and reliable generation of [18F]triflyl fluoride, a
gaseous [18F]fluoride source, Chem. Commun. 54 (72) (2018) 10179–10182.
[3] P. He, S.J. Haswell, N. Pamme, S.J. Archibald, Advances in processes for PET
radiotracer synthesis: Separation of [18F]fluoride from enriched [18O]water,
Appl. Radiat. Isot. 91 (2014) 64–70.
[4] D. Zhou, W. Chu, J. Katzenellenbogen, Sulfonyl
[
18F]fluorides open new
landscape for radiofluorination, J. Nucl. Med. 58 (supplement 1) (2017) 403.
[5] B.o. Zhang, B.H. Fraser, M.A. Klenner, Z. Chen, S.H. Liang, M. Massi, A.J.
Robinson, G. Pascali, [18F]Ethenesulfonyl fluoride as a practical radiofluoride
relay reagent, Chemistry 25 (32) (2019) 7613–7617.
[6] A. Pees, A.D. Windhorst, M.J.W.D. Vosjan, V. Tadino, D.J. Vugts, Synthesis of
Conclusion
[
18F]Fluoroform with High Molar Activity, Eur. J. Org. Chem. 2020 (9) (2020)
1177–1185.
[7] K. Dahl, A. Garcia, N.A. Stephenson, N. Vasdev, ‘‘In-loop” 18F-fluorination: A
proof-of-concept study, J Labelled Comp Radiopharm 62 (7) (2019) 292–297.
[8] J. Dong, L. Krasnova, M.G. Finn, K.B. Sharpless, Sulfur(VI) Fluoride Exchange
(SuFEx): Another Good Reaction for Click Chemistry, Angewandte Chemie Int.
Edition 53 (36) (2014) 9430–9448.
We have developed a simple method for the generation of [18F]
triflyl fluoride as the source of [18F]fluoride for radiolabeling. This
method is an example of continuous-flow solid-phase radiosynthe-
sis, capable of generating labeling-ready [18F]fluoride in high radio-
chemical yield and high molar activity within a short time. This
method has advantages for radiolabeling over the conventional
azeotropic drying method, although with some limitations, as dis-
cussed above. Nevertheless, it should have great potential for expe-
diting the preparation of 18F labeled radiopharmaceuticals.
[9] D.W. Kim, D.-S. Ahn, Y.-H. Oh, S. Lee, H.S. Kil, S.J. Oh, S.J. Lee, J.S. Kim, J.S. Ryu, D.
H. Moon, D.Y. Chi, A New Class of SN2 Reactions Catalyzed by Protic Solvents:
Facile Fluorination for Isotopic Labeling of Diagnostic Molecules, J. Am. Chem.
Soc. 128 (50) (2006) 16394–16397.
[10] D. Zhou, J.A. Katzenellenbogen, Determination of molar activity of [18F]
fluoride by HPLC via sulfonyl derivatizatio, in preparation for Applied
Radiation and Isotope.
[11] W. Qu, Z. Zha, K. Ploessl, B.P. Lieberman, L. Zhu, D.R. Wise, C. B. Thompson, H.F.
Kung, Synthesis of optically pure 4-fluoro-glutamines as potential metabolic
imaging agents for tumors, J Am Chem Soc 133 (4) (2011) 1122–1133.
[12] D. Zhou, W. Chu, J. Xu, L.A. Jones, X. Peng, S. Li, D.L. Chen, R.H. Mach, Synthesis,
Declaration of Competing Interest
[
18F] radiolabeling, and evaluation of poly (ADP-ribose) polymerase-1 (PARP-
The authors declare that they have no known competing finan-
cial interests or personal relationships that could have appeared
to influence the work reported in this paper.
1) inhibitors for in vivo imaging of PARP-1 using positron emission
tomography, Bioorg. Med. Chem. 22 (5) (2014) 1700–1707.
[13] S. Li, A. Schmitz, H. Lee, R.H. Mach, Automation of the Radiosynthesis of Six
Different 18F-labeled radiotracers on the AllinOne, EJNMMI Radiopharmacy
and Chemistry 1 (1) (2016) 15.
Acknowledgments
[14] A. Svadberg, T. Wickstrøm, O.K. Hjelstuen, Degradation of acetonitrile in eluent
solutions for [18F]fluoride PET chemistry: impact on radiosynthesis of [18F]
FACBC and [18F]FDG, J. Labelled Compd. Radiopharm. 55 (3) (2012) 97–102.
[15] H. Jiang, S.G. DiMagno, T.R. DeGrado, Production and transport of gaseous
18F-Synthons: 18F-Acyl Fluorides, J. Fluor. Chem. 180 (2015) 181–185.
[16] S.J. Gatley, Rapid production and trapping of [18F]fluorotrimethylsilane, and its
use in nucleophilic fluorine-18 labeling without an aqueous evaporation step,
Int. J. Rad. Appl. Instrum. A 40 (6) (1989) 541–544.
This work was supported by the National Institutes of Health
[R01CA025836 (JAK)].
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tetlet.2021.153273.
4