Radiosynthesis of 5-[18F]Fluoro-1,2,3-triazoles through Aqueous Iodine–[18F]Fluorine Exchange Reaction

In this report, a simple and efficient process to achieve fluorine-18-labeled 1,2,3-triazole is reported. The heteroaromatic radiofluorination was successfully achieved through an iodine–fluorine-18 exchange in an aqueous medium requiring only trace amounts of base and no azeotropic drying of fluorine-18. This methodology was optimized on a model reaction and further validated on multiple 1,2,3-triazole substrates with 18–60% radiochemical conversions. Using this strategy—the radiosynthesis of a triazole-based thiamin analogue—a potential positron emission tomography (PET) probe for imaging thiamin-dependent enzymes was synthesized with 10–16% isolated radiochemical yield (RCY) in 40 min (uncorrected, n > 5).

In the positron emission tomography (PET) imaging field, this framework has also been extensively utilized as a versatile linker for attaching short-lived radioisotopes. The heterocycle is normally stable under in vivo conditions [16] and also introduces some degree of polarity into the imaging tracer [17]. Additionally, it can be designed as a surrogate for the amide bond [18]. Among the methods of introducing fluorine-18 to a triazole, copper-catalyzed azide-alkyne 1,3-dipolar cycloaddition (CuAAC, click chemistry) is undoubtedly the most powerful approach [19][20][21][22]. The simplicity and highly efficient nature of CuAAC have enabled numerous applications of this process in the radiosyntheses of both small molecules [23][24][25] and macromolecules [26][27][28][29]. However, this method requires multi-step radiosynthesis: the azeotropic drying of fluorine-18, the incorporation of fluorine-18 on the alkyl-or azide-substrates, and the purification of the labeled substrate followed by click reaction with biomolecules. Late-stage, direct fluorination to the triazole ring remains challenging. To our knowledge, no methodology of appending fluorine-18 directly to 1,2,3-triazole heterocycles has been reported. In 2012, Fokin et al. first discovered an efficient and straightforward halogen exchange (Halex) reaction of 5-iodotriazoles to prepare 5-fluoro/chloro-triazoles [30]. Under Fokin et al.'s conditions, various fluoro-and chloro-substituted 1,2,3-triazoles were obtained in aqueous media containing a large excess of KF or KCl (Scheme 1A). A ring-opening mechanism was proposed to generate a reactive diazo/imidoyl iodide intermediate which further reacts with fluoride ions (Scheme 1B). Mild fluorinating agents such as KF or KHF 2 , fast reaction times (10 min), aqueous reaction conditions, and excellent functionalgroup tolerance further enhanced the feasibility of this methodology. Later on, Chu et al. reported the silver-mediated fluorination of 5-iodotriazoles with AgF to prepare 5fluorotriazoles [31]. Although reaction temperature was lowered (120 • C vs. 180 • C) when using AgF as a source of fluorine, the reaction time became much longer (20 h vs. 10 min). Nevertheless, these aqueous iodine-fluorine exchange (IFX) reactions-which require no time-consuming azeotropic drying of fluorine-18 or phase-transfer reagents-would be beneficial for the production of fluorine-18-labeled PET tracers. In this work, we report the application of this highly efficient heteroaromatic substitution in the radiosynthesis of fluorine-18-labeled 1,2,3-triazole analogs (Scheme 1C), as well as a triazole-based thiamin analogue, a potential PET probe for thiamin-dependent enzyme imaging. In 2012, Fokin et al. first discovered an efficient and straightforward halogen exchange (Halex) reaction of 5-iodotriazoles to prepare 5-fluoro/chloro-triazoles [30]. Under Fokin et al.'s conditions, various fluoro-and chloro-substituted 1,2,3-triazoles were obtained in aqueous media containing a large excess of KF or KCl (Scheme 1A). A ringopening mechanism was proposed to generate a reactive diazo/imidoyl iodide intermediate which further reacts with fluoride ions (Scheme 1B). Mild fluorinating agents such as KF or KHF2, fast reaction times (10 min), aqueous reaction conditions, and excellent functional-group tolerance further enhanced the feasibility of this methodology. Later on, Chu et al. reported the silver-mediated fluorination of 5-iodotriazoles with AgF to prepare 5fluorotriazoles [31]. Although reaction temperature was lowered (120 °C vs. 180 °C) when using AgF as a source of fluorine, the reaction time became much longer (20 h vs. 10 min). Nevertheless, these aqueous iodine-fluorine exchange (IFX) reactions-which require no time-consuming azeotropic drying of fluorine-18 or phase-transfer reagents-would be beneficial for the production of fluorine-18-labeled PET tracers. In this work, we report the application of this highly efficient heteroaromatic substitution in the radiosynthesis of fluorine-18-labeled 1,2,3-triazole analogs (Scheme 1C), as well as a triazole-based thiamin analogue, a potential PET probe for thiamin-dependent enzyme imaging. Scheme 1. Aqueous Halex reaction of 5-iodo-1,2,3-triazole to give 5-fluoro/chloro-1,2,3-triazole (A, literature) [30], the proposed mechanism (B, literature) [30] and aqueous [ 18 F]IFX approach to synthesize 5-[ 18 F]fluoro-1,2,3-triazole (C, this work).

Results and Discussion
A commercially available compound, 5-iodo-1,4-dimethyl-1H-1,2,3-triazole (1), was used to optimize the [ 18 F]IFX reaction. The [ 18 F]IFX radiofluorination was performed by adding co-solvent and [ 18 F]HF (in target water) to a microwave vessel containing the precursor 1. The effects of several factors on radiochemical conversions, such as base, co-solvent, temperature, and reaction time, were investigated (Table 1). First, the use of CH3CN was tested. However, due to the high temperature required vs. the low boiling point of CH3CN, the pressure in reaction vials can reach the limit of the microwave (MW) reactor and trigger safety cooling. On the other hand, the use of DMF or DMSO increases the radiochemical conversion (RCC, up to 50%, entry 3-10). As a result, we decided to use DMSO as the co-solvent of choice. Various reaction temperatures were also tested without Scheme 1. Aqueous Halex reaction of 5-iodo-1,2,3-triazole to give 5-fluoro/chloro-1,2,3-triazole (A, literature) [30], the proposed mechanism (B, literature) [30] and aqueous [ 18 F]IFX approach to synthesize 5-[ 18 F]fluoro-1,2,3-triazole (C, this work).

Results and Discussion
A commercially available compound, 5-iodo-1,4-dimethyl-1H-1,2,3-triazole (1), was used to optimize the [ 18 F]IFX reaction. The [ 18 F]IFX radiofluorination was performed by adding co-solvent and [ 18 F]HF (in target water) to a microwave vessel containing the precursor 1. The effects of several factors on radiochemical conversions, such as base, co-solvent, temperature, and reaction time, were investigated (Table 1). First, the use of CH 3 CN was tested. However, due to the high temperature required vs. the low boiling point of CH 3 CN, the pressure in reaction vials can reach the limit of the microwave (MW) reactor and trigger safety cooling. On the other hand, the use of DMF or DMSO increases the radiochemical conversion (RCC, up to 50%, entry 3-10). As a result, we decided to use DMSO as the co-solvent of choice. Various reaction temperatures were also tested without any significant differences in RCCs between 140-160 • C. When the temperature was 130 • C or below, little to no product was observed. This finding is in agreement with the literature suggesting that a minimum of 140 • C is required to initiate the reaction [30]. any significant differences in RCCs between 140-160 °C. When the temperature was 130 °C or below, little to no product was observed. This finding is in agreement with the literature suggesting that a minimum of 140 °C is required to initiate the reaction [30]. The effect of the amount of base was also investigated (entries 3,5,6,7,9). The reaction did not proceed in the absence of a base. This was expected since [ 18 F]HF is known to be non-nucleophilic without neutralization. Since the cyclotron-produced [ 18 F]HF had a very low mass concentration, the radiosyntheses were tested with trace levels of K2CO3. When the reaction was spiked with 15 µ g of K2CO3 (0.11 µ mol), 11% RCC was observed. The RCC (31%) was improved by increasing the amount of K2CO3 to 30 µ g (0.22 µ mol), and plateaued (50%) at 60 µ g/reaction (0.43 µ mol, Figure 1). Based on the model reaction, we settled on the following conditions: K2CO3 (60 µ g, 0.43 µ mol), DMSO/H2O, and 150 °C MW for 20 min.
With optimized conditions, a substrate scope study was performed to assess the feasibility of this methodology ( Figure 2). For the four substrates without C4 substitution (compound 3-6), no product was observed. This demonstrated that a C4 substitution is necessary, perhaps to stabilize the open diazo form of the intermediate (Scheme 1B) [30]. A variety of 1,4-disubstituted-5-iodo-1H-1,2,3-triazoles were designed to further evaluate the application scope (Compounds 7-13). These substrates were readily prepared by following a published procedure [32]. Under the standard condition, most of the radiofluorinated product was obtained in medium to good RCCs. For N1 and C4 substitution, both aromatic and aliphatic groups were well tolerated to produce the desired product. More specifically, for the N1 position, no significant differences between RCCs were noticed for substrates with electron-withdrawing and electron-donating groups. Similar or slightly The effect of the amount of base was also investigated (entries 3, 5, 6, 7, 9). The reaction did not proceed in the absence of a base. This was expected since [ 18 F]HF is known to be non-nucleophilic without neutralization. Since the cyclotron-produced [ 18 F]HF had a very low mass concentration, the radiosyntheses were tested with trace levels of K 2 CO 3 . When the reaction was spiked with 15 µg of K 2 CO 3 (0.11 µmol), 11% RCC was observed. The RCC (31%) was improved by increasing the amount of K 2 CO 3 to 30 µg (0.22 µmol), and plateaued (50%) at 60 µg/reaction (0.43 µmol, Figure 1). Based on the model reaction, we settled on the following conditions: K 2 CO 3 (60 µg, 0.43 µmol), DMSO/H 2 O, and 150 • C MW for 20 min. improved RCC was achieved when the benzyl group of 7 was replaced with a more electron-withdrawing 4-cyanobenzyl group (11), or a more electron-donating polyethylene glycol (PEG) group (13). However, for the C4 substitution, the replacement of the phenyl group (7) with methyl ester (8) completely inhibited radiofluorination. This finding is in agreement with the literature, indicating that no transformation occurs for substrates with an electron-withdrawing group on C4 [30]. In contrast to the previously reported examples, we found that both aromatic and aliphatic C4 substituents were able to successfully activate radiofluorination. All different C4 aliphatic substrates (9, 12) produced the desired product (18-40% RCC).  With optimized conditions, a substrate scope study was performed to assess the feasibility of this methodology ( Figure 2). For the four substrates without C4 substitution (compound 3-6), no product was observed. This demonstrated that a C4 substitution is necessary, perhaps to stabilize the open diazo form of the intermediate (Scheme 1B) [30]. A variety of 1,4-disubstituted-5-iodo-1H-1,2,3-triazoles were designed to further evaluate the application scope (Compounds 7-13). These substrates were readily prepared by following a published procedure [32]. Under the standard condition, most of the radiofluorinated product was obtained in medium to good RCCs. For N1 and C4 substitution, both aromatic and aliphatic groups were well tolerated to produce the desired product. More specifically, for the N1 position, no significant differences between RCCs were noticed for substrates with electron-withdrawing and electron-donating groups. Similar or slightly improved RCC was achieved when the benzyl group of 7 was replaced with a more electron-withdrawing 4-cyanobenzyl group (11), or a more electron-donating polyethylene glycol (PEG) group (13). However, for the C4 substitution, the replacement of the phenyl group (7) with methyl ester (8) completely inhibited radiofluorination. This finding is in agreement with the literature, indicating that no transformation occurs for substrates with an electron-withdrawing group on C4 [30]. In contrast to the previously reported examples, we found that both aromatic and aliphatic C4 substituents were able to successfully activate radiofluorination. All different C4 aliphatic substrates (9, 12) produced the desired product (18-40% RCC).    Figure 3), a potential imaging probe for thiamin-dependent enzymes. Thiamin plays a key role in numerous body functions such as energy metabolism, protein and nucleic acid biosynthesis [33]. It is particularly important in the function of the nervous system and protection against neurological disorders [34]. Recent studies have demonstrated the significance of thiamin-dependent enzymes in cancer cell metabolism [35]. This suggests that tumor cells will display an elevated uptake of probes targeting thiamin-dependent enzymes, a finding that may be of diagnostic value in the early detection of cancer through PET. So far, a triazole-based thiamin pyrophosphate, 14, is one of the most potent inhibitors for thiamin-dependent enzymes [36,37]. We reasoned, therefore, that the fluorine-substituted analog of 14, [ 18 F]15, would be an excellent candidate as a PET imaging agent (Figure 3).  The non-radioactive standard compound 15 and the radiolabeling precursor 16 were synthesized as depicted in Scheme 2. The azido intermediate 17 was synthesized from a commercially available thiamin chloride [36]. Iodo-precursor 16 was obtained by the Cu-AAC reaction of the azido intermediate 17 and the alkynyl compound 18 [38]. Compound 15 was then prepared from the iodo-counterpart by the Halex reaction [30,39]. The non-radioactive standard compound 15 and the radiolabeling precursor 16 were synthesized as depicted in Scheme 2. The azido intermediate 17 was synthesized from a commercially available thiamin chloride [36]. Iodo-precursor 16 was obtained by the CuAAC reaction of the azido intermediate 17 and the alkynyl compound 18 [38]. Compound 15 was then prepared from the iodo-counterpart by the Halex reaction [30,39].  (16) in DMSO (200 µ L) and an aqueous K2CO3 solution (0.3 mg/mL, 200 µ L) were added to a reaction vial containing aqueous [ 18 F]HF. The reaction vial was heated at 150 °C for 10 min via microwave irradiation. The crude reaction mixture was evaluated by analytical HPLC, which indicated 30-34% RCC (n = 4). The product was purified by semi-preparative HPLC to produce [ 18 F]15 in 10-16% RCY (uncorrected, n > 5) with radiochemical purity >98% ( Figure 4A). The synthesis was  F]HF. The reaction vial was heated at 150 • C for 10 min via microwave irradiation. The crude reaction mixture was evaluated by analytical HPLC, which indicated 30-34% RCC (n = 4). The product was purified by semi-preparative HPLC to produce [ 18 F]15 in 10-16% RCY (uncorrected, n > 5) with radiochemical purity >98% ( Figure 4A). The synthesis was completed in 40 min, including fluorination and HPLC purification to produce the inject-ready dose. In a typical production starting with 6.03 GBq (163 mCi) of [ 18 15 was confirmed by co-elution with its authentic nonradioactive standard on an analytical HPLC ( Figure 4B,C). It is worth noting that [ 18 F]IFX was successfully achieved with fully unprotected (amino and hydroxyl groups) iodo-precursor 16, which further confirmed the functional group compatibility of this radiofluorination method. With the trace amount of fluorine-18 and the strong rates of incorporation, this reaction displayed unusually high chemoselectivity. The functional groups (hydroxyl and amino) and water did not compete with the fluorine-18 as a nucleophile, perhaps suggesting a tight ion pair between [ 18 F]F − and the purported C5-N1 imine intermediate, which leads to fluorination.
iodotriazole precursor (16) in CH3CN or DMSO was reacted with azeotropically dried fluorine-18 ([ 18 F]KF/K2CO3/K222 or [ 18 F]TBAF/TBAB) at a temperature between 100 and 180 °C. However, no product formation was detected under these conditions. The decomposition of the precursor was observed for all conditions tested, although it is relatively more stable in TBAB vs. K2CO3/K222. Next, we tested the process of aqueous [ 18 F]IFX radiofluorination to prepare the compound [ 18 F]15 (Scheme 3). Cyclotron-produced [ 18 F]HF in target water was used directly without azeotropic drying. The iodo-precursor (16) in DMSO (200 µ L) and an aqueous K2CO3 solution (0.3 mg/mL, 200 µ L) were added to a reaction vial containing aqueous [ 18 F]HF. The reaction vial was heated at 150 °C for 10 min via microwave irradiation. The crude reaction mixture was evaluated by analytical HPLC, which indicated 30-34% RCC (n = 4). The product was purified by semi-preparative HPLC to produce [ 18 F]15 in 10-16% RCY (uncorrected, n > 5) with radiochemical purity >98% ( Figure 4A). The synthesis was completed in 40 min, including fluorination and HPLC purification to produce the injectready dose. In a typical production starting with 6.03 GBq (163 mCi) of [ 18 F]HF, 0.962 GBq (26 mCi) of [ 18 F]15 was received at the end of synthesis. The identity of [ 18 F]15 was confirmed by co-elution with its authentic nonradioactive standard on an analytical HPLC ( Figure 4B,C). It is worth noting that [ 18 F]IFX was successfully achieved with fully unprotected (amino and hydroxyl groups) iodo-precursor 16, which further confirmed the functional group compatibility of this radiofluorination method. With the trace amount of fluorine-18 and the strong rates of incorporation, this reaction displayed unusually high chemoselectivity. The functional groups (hydroxyl and amino) and water did not compete with the fluorine-18 as a nucleophile, perhaps suggesting a tight ion pair between [ 18 F]F − and the purported C5-N1 imine intermediate, which leads to fluorination.

Material and Methods
Compounds 3-5 were purchased from AstaTech Inc (Bristol, PA, USA). Compounds 7-11 were synthesized according to the literature method outlined in [31,32]. All other chemicals and solvents were received from Sigma Aldrich (St. Louis, MO, USA) and used without further purification. Fluorine-18 was received from the National Institutes of Health's cyclotron facility (Bethesda, MD, USA). Mass spectrometry (MS) was performed on a 6130 Quadrupole LC/MS (Agilent Technologies, Santa Clara, CA, USA), Agilent Technologies instrument equipped with a diode array detector. The 1 H, 13 C and 19 F NMR spectra were recorded on a Varian spectrometer (400 MHz) (Varian, Palo Alto, CA, USA). Chemical shifts (ppm) are reported relative to the solvent residual peaks. High-performance liquid chromatography (HPLC), for purification and analytical analysis, was performed on an Agilent 1200 Series instrument (Agilent Technologies, Santa Clara, CA, USA) equipped with multi-wavelength detectors along with an Eckert and Ziegler B-FC-3500 diode flow count radiodetector (Eckert and Ziegler, Berlin, Germany).

General Procedure to Synthesize Compounds 12-13
The procedure followed the literature, with minor modifications [32]. Generally, azides (1.65 mmol), terminal alkynes (1.5 mmol), Selectfluor (1.8 mmol), tetraethylammonium iodide (1.65 mmol), DIPEA (1.8 mmol), and CuI (0.15 mmol) were added to H 2 O (8 mL) and stirred for 5-12 h at 30°C (water bath). The reaction was monitored by TLC. After the reaction was completed, the mixture was partitioned between water and ethyl acetate (3 × 10 mL). The organic layer was then combined, dried over Na 2 SO 4 , and the filtrate was concentrated under reduced pressure. The crude products were further purified to produce pure compounds using flash chromatography on silica gel with hexanes/ethyl acetate as the eluent. The procedure followed the literature with minor modifications [30]. Briefly, the iodide starting material (1 equiv) and KF (5 equiv) were added to a 2-5 mL microwave vial equipped with a magnetic stir bar. To the solid mixture, CH 3 CN and water (1 mL each) were added. The vial was capped with a crimp cap with Teflon septum and this was placed into the microwave reactor at 180 • C for 30 min. After the vial had cooled to room temperature, the crude mixture was diluted with ethyl acetate and water (3 mL each) and extracted with a Pasture pipette. The resulting aqueous phase was extracted two additional times and the combined organic phase was pushed through a Pasture pipette filled with~3 g of Na 2 SO 4 . Volatiles were removed under reduced pressure and the residue  4-Iodo-but-3-yn-1-ol (18) To a solution of but-3-yn-1-ol (5.0 g, 71.0 mmol) and potassium iodide (13.0 g, 78.5 mmol) in methanol (30 mL), a 70% solution of aqueous tert-butyl hydroperoxide (100 mmol) was added, drop-wise, over 50 min and while stirring at room temperature. The reaction mixture was stirred for an additional one hour and quenched with saturated aqueous Na 2 S 2 O 3 . The product was extracted with ethyl acetate and dried over anhydrous MgSO 4 . The solvent was removed under vacuum and the crude product was purified by column chromatography using a hexane/ethyl acetate mixture to give an oily 4-iodo-but-3-yn-1-ol (9.2 g, 65%), 1 H NMR δ (400 MHz, CDCl 3 ): δ = 3.72 (t, J = 6.13 Hz, 2H), 2.63 (t, J = 6.01 Hz, 2H), 2.23-2.45 (br s, 1H). (17) Sodium sulfite (0.39 g, 3.0 mmol) was added to a solution of thiamin chloride (10.0 g, 30.2 mmol) and sodium azide (5.0 g, 76.0 mmol) in water (100 mL). The mixture was stirred for 5 h at 65 • C. Citric acid (22.0 g, 100 mmol) was added to adjust pH ≈ 4 and then the aqueous solution was washed with dichloromethane. Potassium carbonate was added to the aqueous phase to pH ≈ 8, upon which some precipitation of the product occurred. The suspension was filtered, and the filtrate was extracted with ethyl acetate and the combined organic layers were washed with brine, dried over MgSO 4 and evaporated under reduced pressure. The solid residue was pooled with the precipitate and recrystallized from ethyl acetate/hexane to give the azide derivative as fine needles  (18) Under an argon atmosphere, 4-iodo-but-3-yn-1-ol (1.2 g, 6.1 mmol), 5-Azidomethyl-2-methylpyrimidin-4-ylamine (1.0 g, 6.1 mmol), triethylamine (975 µL, 7 mmol), DMF (10 mL), and CuI (1.16g, 6.1 mmol) were successively added to a round conical flask. The mixture was stirred vigorously overnight; ethyl acetate (20 mL) was added to the reaction mixture and it was filtered through a pad of Celite. After the removal of the solvent, the crude product was purified by column chromatography using a hexane/ethyl acetate mixture to afford 1.17 g (69%) of pure product as an off-white powder; 1 H NMR (400 MHz, MeOD) δ 8.00 (s, 1H), 5.51 (s, 2H), 3.81 (t, J = 6.7 Hz, 2H), 2.89 (t, J = 6.8 Hz, 2H), 2.42 (s, 3H); 13  2-[1-(4-Amino-2-methyl-pyrimidin-5-ylmethyl)-5-Flouro-1H-[1,2,3]triazol-4-yl]-ethanol (15) A mixture of iodo-triazole derivative (50 mg, 0.14 mmol) and KF (40 mg, 0.7 mmol) was added to DMSO (250 uL). The reaction mixture was heated up to 180-190 • C for 5 min; DMSO was blown out by air flow. The crude product was purified using flash silica gel column chromatography (gradient, ethyl acetate/hexanes, 1:1) to afford fluoro-triazole derivative (