Late‐Stage 18F‐Difluoromethyl Labeling of N‐Heteroaromatics with High Molar Activity for PET Imaging

Abstract Despite a growing interest in CHF2 in medicinal chemistry, there is a lack of efficient methods for the insertion of CHF18F into druglike compounds. Herein described is a photoredox flow reaction for 18F‐difluoromethylation of N‐heteroaromatics that are widely used in medicinal chemistry. Following the two‐step synthesis for a new 18F‐difluoromethylation reagent, the photoredox reaction is completed within two minutes and proceeds by C−H activation, circumventing the need for pre‐functionalization of the substrate. The method is operationally simple and affords straightforward access to radiolabeled N‐heteroaromatics with high molar activity suitable for biological in vivo studies and clinical application.


Experimental procedures and analytical data a. General Information
All reagents purchased from commercial sources were used as received. Technical solvents were bought from VWR international and used as received.
Flow reactions were performed on a R-series Vapourtec system, using a LED photoreactor (450 nm, 24 W, 10 mL coil reactor with FEP tubing).
Semi preparative (Semi-PREP) purifications were performed using SQD Waters single quadrupole mass spectrometer. This spectrometer is equipped with an ESI source, Waters 2535 quaternary pump coupled with 2767 sample Manager and with diode array detector (210 to 400 nm.) The column used is a Waters Sunfire ODB MS C18 column (5µm, 30 x 50 mm) for acidic purification and a Waters XBridge OBD MS C18 column (5µm, 30 x 50 mm) for basic purification.
NMR spectra were recorded on a BRUKER AVANCE III Ultrashield Nanobay 400 MHz NMR Spectrometer and on a BRUKER AVANCE III HD Ascend 500 MHz NMR Spectrometer fitted with a 5 mm Prodigy BBO 500 S1 cryoprobe. The compounds were analyzed in d6-DMSO solution at a probe temperature of 300 K. Chemical shifts are given in ppm downfield from TMS (tetramethylsilane) as internal standard. For 19 F NMR, chemical shifts are given in ppm downfield from TFA (trifluoroacetic acid,  -76.50) as internal standard. The NMR multiplicity signals are reported as s = singlet, d = doublet, t = triplet, m = multiplet, br = broad, or combinations of thereof. Coupling constants J are quoted in Hz and reported to the nearest 1 Hz.
HRMS were obtained using a SYNAPT G2-SI Waters Q-TOF mass spectrometer. This spectrometer is equipped with an ESI source and a Waters Acquity H-class UPLC with diode array detector (210 to 400 nm.) An Acquity UPLC HSS T3 C18 column (1.8µm, 2.1 x 50 mm) was used.
As less than 1 mg of product was obtained, carbons assignments were determined by HSQC and HMBC analysis.

S-27
No-carrier-added [ 18 F]fluoride was produced via the 18 O(p,n) 18 F nuclear reaction by bombardment of 18 O-enriched water (>95%) with 18 MeV protons using a cyclone 18/18 (IBA) . 18 O-enriched water was purchased from Rotem or ABX. At the end of bombardment (EOB), the activity was transferred to the hot lab cell with helium pressure through Teflon tubing (~50 m). Radioactivity was measured in a dose calibrator (Veenstrat). All the radiochemical yields are decay corrected.
First experiments were realized with a low level of radioactivity (37-185 MBq (1-5 mCi)). For synthesis at higher level (111 GBq (3 Ci)), an automated FASTlab TM module from GE Healthcare was used.
Thin layer chromatography (TLC) were realized silica gel Polygram ® SIL G/UV254 pre-coated TLC sheets eluted with MeOH (100%). The same eluent was used for all the radioactive analyses. The radioactive spots were quantitatively detected on a Berthold TLC scanner (model AR200). The TLC identity of all the labelled compounds was confirmed by UPLC after injection and co-injections on the same analytical system (see above) of the corresponding 19 F-fluorinated references.
UPLC analyses were carried out on an ACQUITY UPLC ® system (Waters) equipped with a PDA UV (200-400 nm) and a gamma-ray NaI detectors. The system was controlled by the Empower software. The ACQUITY UPLC ® CSH TM C18 column (2.1 x 100 mm, 1.7 μm; Waters), heated at 45°C, was eluted in gradient mode with a mixture consisting of MeCN/H2O (0,05 % HCOOH) (See Table S1).  The semi-preparative HPLC purification was conducted on a X-Terra® RP18 HPLC column (10 x 250 mm, 10 μm, Waters), connected to a stand alone HPLC. The loop of a motorized rheodyne valve was lined to the outlet of the synthesizer module. The Waters system (600 pump, 996 PDA UV detector (190-400 nm) was controlled by the Empower software. The radioactive elution profile was monitored with a custom-made Geiger-Müller (GM) radioactivity detector. The column was eluted with an isocratic mixture of MeCN/H2O (40/60 (v/v)) at a flow rate of 5 mL min -1 .
Flow reactions for difluoromethylation were performed using the Futurechem FlowStart EVO system, with a microchip of 100 µL and a 2 W LED of 470 nm.

b. Automated radiosynthesis of [ 18 F]3
The whole radiosynthesis of [ 18 F]3 was performed on a FASTlab TM synthesizer from GE Healthcare. The reagents and solvents used for the radiosynthesis of [ 18 F]3 were placed in small sealed vials according to a process previously reported in our laboratory [ 5] . Reagents were prepared and positioned on the FASTlab TM manifold as described and illustrated in Figure S2.

c. Optimization of the labeling and oxidative steps
The conditions screened for the labeling and oxidative steps are summarized in the following table (Table S2) and each radiochemical experiment was conducted three times (n=3). The Radio-LC and a LC of the references are also shown after the first table to confirm the formation of the desired product ( Figures S3-6).     For each experiment, the RCY (decay-corrected) of [ 18 F]2 and [ 18 F]3 was determined after SPE purification on an aliquot of the final solution according to the following formula taking into account the TLC and HPLC radiochemical purity of the isolated product.

d. Isolation and determination of molar activity of the sulfone [ 18 F]3
The fully automated synthesis of the 18 F-labeled sulfone was realized with the FastLab module as described in section 2b using the best conditions reported in section 2c ( Table 2, entry 7). The molar activity of [ 18 F]3 was determined on a aliquot of the DMSO solution (5 µL). After UPLC elution, the radioactive peak associated to the non-radioactive sulfone was collected and counted and the UV area of the peak determined ( = 239 nm). The decay-corrected activity was calculated and the corresponding amount of 3 was determined using the calibration curve previously obtained with the non-radioactive reference ( Figure S7). A molar activity between 74 and 185 GBq/µmol (2 and 5 Ci/µmol) was calculated.
These analyses were performed with the UPLC system and the conditions reported above.

SUPPORTING INFORMATION
S-33  Figure S8). The exited solution was analysed by Radio-TLC and Radio-UPLC for radiochemical yield determination.   NB : In some cases, some peaks at 0.6 and 0.9 min can be observed on the radio-UPLC chromatograms. Those two peaks were collected and analysed by radio-TLC. Their Rf being of 0, they were not taking into account for the determination of the LC purity (as they were already taken into account in the determination of the TLC purity).
The results of the different optimization tests are summarized in the following  Only a few solvents allows a complete solubilization of the substrate, DMSO gives better results than DMF. The temperature is also an important parameter: too high temperature leads to more degradation. The best conditions for the 18 F-difluoromethylation are 35°C, [Ir(ppy)3] (0.01 µmol), 2 minutes of residence time and DMSO as solvent (see entry 1, Table S4). These conditions were applied to the scope.

f. Isolation and Molar activity of CHF 18 FAcyclovir [ 18 F]5
After completion of the photoredox reaction that was realized as described in the previous section, H2O (5 mL) was added. The purification of [ 18 F]5 was then conducted at 5 mL/min on the semi preparative HPLC column described above. The eluent was a mixture of H2O (95%) and MeCN (5%). Based on the amount of sulfone used for the photochemistry reaction, the RCY of [ 18 F]5 purified by HPLC was of 42 ± 4 % (dc).
Then following the same process as for compound 3, a calibration curve ( Figure S12) was realized with 5 to determine the molar activity of the product. The molar activity (dc) was of 44.4 ± 11.1 GBq/µmol (1.2 ± 0.3 Ci/µmol). No product was obtained in the absence of light or photocatalyst (entries 1 and 2, Table S5) and in presence of TEMPO (entry 3), suggesting that radical species are involved, as proposed in the mechanism. The exited solution was analyzed by Radio-TLC and Radio-UPLC for radiochemical yield determination.
i. Scope

4-(difluoromethyl)-5-methyl-pyrimidin-2-amine [ 18 F]15 [ 18 F]15
General procedure using 5-methylpyrimidin-2-amine (2.2 mg, 20 µmol) yielded to 43 ± 3% as RCY (ndc, on crude product) of the title compound (see Table S15 and Figures S36 and S37 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using 2-methyl-6,8-dihydro-5H-pyrido[2,3-d]pyrimidin-7-one (3.3 mg, 20 µmol), with a residence time of 4 min and [Ir(ppy)3] (0.05 µmol) yielded to 32 ± 4% as RCY (ndc, on crude product) of the title compound (see Table S16 and Figures S38 and S39 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using caffeine (3.9 mg, 20 µmol) yielded to 51 +/-1% as RCY (ndc, on crude product) of the title compound (see Table S20 and Figures S46 and S47 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using theophylline (3.6 mg, 20 µmol) yielded to 42 ± 6% as RCY (ndc, on crude product) of the title compound (see Table S21 and Figures S48 and S49 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using pentoxyfilline (5.6 mg, 20 µmol) yielded to 30 ± 5 % as RCY (ndc, on crude product) of the title compound (see Table S22 and Figures S50 and S51 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using dimethyl-uracil (2.8 mg, 20 µmol) yielded to 54 ± 1% as RCY (ndc, on crude product) of the title compound (see Table S23 and Figures S52 and S53 for radio LC of the labelled compound and LC analysis of the reference).   General procedure using cytosine (3.2 mg, 20 µmol) yielded to 60 ± 2 % as RCY (ndc, on crude product) of the title compounds. The two isomers were not distincted in UPLC analyses, the ratio was then not determined. (see Table S24 and Figures S54 and S55 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using uridine (4.8 mg, 20 µmol) yielded to 71 ± 4 % as RCY (ndc, on crude product) of the title compound (see Table S26 and Figures S58 and S59 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using cytidine (4.8 mg, 20 µmol) yielded to 65 ± 2 % as RCY (ndc, on crude product) of the title compound (see Table S27 and Figures S60 and S61 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using adenosine (5.3 mg, 20 µmol) yielded to 59 ± 5% as RCY (ndc, on crude product) of the title compound (see Table S28 and Figures S62 and S63 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using guanosine (5.7 mg, 20 µmol) yielded to 62 ± 6% as RCY (ndc, on crude product) of the title compound (see Table S29 and Figures S64 and S65 for radio LC of the labelled compound and LC analysis of the reference).  General procedure using moxonidine analogue (4.5 mg, 20 µmol) yielded to 65 ± % 4 as RCY (ndc, on crude product) of the title compound (see Table S30 and Figures S66 and S67 for radio LC of the labelled compound and LC analysis of the reference).    Table S31 and Figures S68 to S71 for radio LC of the labelled compound and LC analysis of the reference).

Radio-TLC purity (%)
Different analyses conditions were used, to ensure a better separation of the different isomers.
The new analytic gradient is disclosed in the table below : (Table S32)

Comparison batch fluorine-19/flow fluorine-18 conditions
Some of the substrates were also tested in non-radioactive chemistry to able a comparison between non-radioactive and radioactive chemistry and put in evidence the difference of reactivity between both conditions. The general procedure and the results are reported just below.

. Results
The general procedure was applied to 10 compounds of the scope and the results are presented in the following table (Table S33) : It is interesting to note that in most cases, yields are way lower in non-radioactive chemistry (molecules 5, 8, 9, 12, 14, 23, 28) while similar results are obtained with both fluorine-19 conditions. Noteworthy, adding new portion of the difluoromethylating agent (3), and catalyst doesn't lead to much yield improvement suggesting the formation of a poisoning reagent preventing the reaction to continue.
Finally, the isomers ratio can be slightly different between all conditions. All these results clearly showed an improved reactivity of the reaction using fluorine-18.