Arene radiofluorination enabled by photoredox-mediated halide interconversion

Positron emission tomography (PET) is a powerful imaging technology that can visualize and measure metabolic processes in vivo and/or obtain unique information about drug candidates. The identification of new and improved molecular probes plays a critical role in PET, but its progress is somewhat limited due to the lack of efficient and simple labelling methods to modify biologically active small molecules and/or drugs. Current methods to radiofluorinate unactivated arenes are still relatively limited, especially in a simple and site-selective way. Here we disclose a method for constructing C–18F bonds through direct halide/18F conversion in electron-rich halo(hetero)arenes. [18F]F− is introduced into a broad spectrum of readily available aryl halide precursors in a site-selective manner under mild photoredox conditions. Notably, our direct 19F/18F exchange method enables rapid PET probe diversification through the preparation and evaluation of an [18F]-labelled O-methyl tyrosine library. This strategy also results in the high-yielding synthesis of the widely used PET agent l-[18F]FDOPA from a readily available l-FDOPA analogue. A photoredox-mediated SNAr reaction has now been developed for the direct radiofluorination of unactivated aryl halides. A series of arenes can be radiofluorinated in a site-selective manner from readily available aryl halide precursors under mild conditions. This strategy allows efficient 19F/18F isotopic exchange, enabling rapid PET probe diversification and clinical tracer preparation.

P ositron emission tomography (PET) is a routinely used non-invasive imaging technique for the real-time diagnosis and monitoring of human diseases, most notably oncological and neurological disorders 1,2 . Advances in radiotracer development have accelerated the clinical adoption of PET and fostered a growing interest in robust methodologies for synthesizing PET agents via the late-stage installation of short-lived radionuclides 3 . Because many small-molecule pharmaceuticals and therapeutics contain aromatic or heteroaromatic systems within their framework [4][5][6][7] , it is highly desirable to introduce radionuclides on these moieties in a synthetically facile and efficient manner. Late-stage radiolabelling is especially preferred considering the decay of short-lived PET isotopes [8][9][10][11] . Fluorine-18 ( 18 F) is arguably the most widely used short-lived PET isotope (t 1/2 ≈ 110 min) because of its excellent imaging properties, wide availability and ideal half-life. A common strategy used for constructing aryl C(sp 2 )-18 F bonds is nucleophilic aromatic substitution (S N Ar), which substitutes a (pseudo) halide with 18 F fluoride ( 18 F − ) (Fig. 1a). This strategy is routinely used for the synthesis of PET agents with high molar activity [12][13][14] , but is limited to electron-deficient (hetero)aromatic systems 15 . Consequently, systems for the nucleophilic radiofluorination of electron-neutral and -rich aromatics have been investigated over the past decade 3,16 . Progress towards this goal has focused on either transition-metal-mediated methods [17][18][19][20][21][22][23] or the development of specialized nucleofuges [24][25][26][27][28][29][30][31][32] , with several of the aforementioned strategies adapted for automation 17,[19][20][21][22][23][24][25][26][27][28]31,33,34 , which streamlines their translation to clinical use. As a major substrate class for arene functionalization, aryl (pseudo)halides are commonly used intermediates en route to synthesizing organometallic or prefunctionalized arene precursors for radiofluorination (such as aryl-palladium 17 / nickel complexes 18 , aryl boronic acids 24 /esters [19][20][21] , aryl stannanes 25 and aryl iodonium salts/ylides 22,23,[26][27][28] ; Fig. 1b). Clearly, methods that could directly radiofluorinate electron-rich aryl halides are highly desired. However, there are few examples of radio-fluorination for unactivated aryl (pseudo)halides except a recent method disclosed by the Scott and Sanford groups, which allows the ligand-directed conversion of aryl halides to radiofluorinated arenes using an N-heterocyclic carbene (NHC)-copper complex 34 . Although the installation of 18 F in electron-neutral and -rich arenes is impressive, the requisite substrate-directing group could limit the generality of this approach (Fig. 1c). Additionally, this method is limited to aryl iodides and bromides, with chloro-arenes being minimally reactive and no 19 F/ 18 F exchange product observed. Given the dearth of methods for the direct radiofluorination of electron-rich aryl chlorides and fluorides, developing a simple strategy to directly radiofluorinate these motifs is highly desired, especially considering their stability and abundance in therapeutics 7,35 . Importantly, the success of this approach would also enable the direct conversion of fluorinated aromatics into 18 F-labelled radiopharmaceuticals through simple late-stage 19 F/ 18 F exchange 36,37 .
We recently disclosed two acridinium photoredox-mediated methods for arene radiofluorination in which the photoactivated acridinium oxidizes arenes to arene cation radicals, thus enabling site-selective radiofluorination of C(sp 2 )-H (ref. 38 ) and C(sp 2 )-O (ref. 32 ) bonds. Inspired by our previous success, we explored the feasibility of directly converting an aryl halide into its 18 F-labelled congeners to obtain radiolabelled electron-rich arenes. Upon single-electron oxidation, we envisioned that the resulting electron-deficient cation radical would capture the 18 F − at the halide-bearing carbon. Subsequent reduction and expulsion of the halide nucleofuge would furnish the desired radiofluorinated arene (Fig. 1d). This radiofluorination strategy would obviate the need for lengthy, multistep precursor synthesis, greatly simplifying product isolation, and open up new ways to prepare clinically significant PET agents. Positron emission tomography (PET) is a powerful imaging technology that can visualize and measure metabolic processes in vivo and/or obtain unique information about drug candidates. The identification of new and improved molecular probes plays a critical role in PET, but its progress is somewhat limited due to the lack of efficient and simple labelling methods to modify biologically active small molecules and/or drugs. Current methods to radiofluorinate unactivated arenes are still relatively limited, especially in a simple and site-selective way. Here we disclose a method for constructing C-18 F bonds through direct halide/ 18 F conversion in electron-rich halo(hetero)arenes. [ 18 F]F − is introduced into a broad spectrum of readily available aryl halide precursors in a site-selective manner under mild photoredox conditions. Notably, our direct 19 F/ 18 F exchange method enables rapid PET probe diversification through the preparation and evaluation of an [ 18 F]-labelled O-methyl tyrosine library. This strategy also results in the high-yielding synthesis of the widely used PET agent l-[ 18 F]FDOPA from a readily available l-FDOPA analogue.

Results and discussion
electron-rich arenes, we adapted our previously reported radiodeoxyfluorination conditions to 1-chloro-4-methoxybenzene (1-Cl), with the intent of promoting halide extrusion. Irradiating a reaction solution containing the substrate, [ 18 F]n-tetrabutylammonium fluoride ([ 18 F]TBAF) and acridinium S1 in a multicomponent dichloroethane(DCE)/t-BuOH/MeCN solvent system with a 450-nm laser under consistent air bubbling for 30 min leads to the formation of 1-18 F, albeit in 1.7% radiochemical conversion (RCC) as calculated by HPLC isolation. In our previous (radio)deoxyfluorination study 32 , we found that methoxy groups were ineffective due to the competitive deprotonation of acidic O-methyl C-H bonds upon formation of the arene cation radical-this deleterious pathway competes with extrusion of halide under aerobic conditions 39 . To circumvent this potential side reaction, the radiofluorination of 1-Cl was conducted under an N 2 atmosphere, which successfully increased the isolation yield of 1-18 F to 12.8 ± 0.3% (number of experiments, n = 3) with 71.25 GBq µmol −1 molar activity (A m ).
With these preliminary results, we then screened other nucleofuges commonly used in S N Ar reactions ( Table 1). The radiofluorination of haloarene 1-bromo-4-methoxybenzene (1-Br) yields an RCC comparable with that of 1-Cl, whereas 1-iodo-4-methoxybenzene (1-I) was surprisingly less reactive, with a tenfold lower RCC. Interestingly, 1-fluoro-4-methoxybenzene (1-F) was found to provide the highest yield of radiofluorinated product through direct 19 F/ 18 F exchange (79.7%). This observation represents a important breakthrough in the field because it allows simple and efficient direct conversion of electron-rich fluorinated bioactive compounds or pharmaceuticals to 18 F-labelled PET radiotracers. Previous reports of fluoroarene 19 F/ 18 F exchange have been restricted to electron-deficient compounds and generally require relatively high temperatures 36,37,[40][41][42] . A potential limitation of this strategy is the decreased molar activity (A m ) as a consequence of using the HPLC-inseparable 19 F-fluoroarene. However, A m values can be greatly improved by limiting the amount of aryl fluoride, as has been demonstrated previously 36 . Next, we found that aryl triflate 1-OTf was successfully radiofluorinated, albeit with lower yields. Interestingly, 1-methoxy-4-nitrobenzene (1-NO 2 ) was unreactive, despite the well-documented substitution of nitro nucleofuges in traditional S N Ar.
To better understand the effect of substitution pattern on the halogen/ 18 F interconversion, we evaluated a range of aromatic and heteroaromatic substrates (Table 1). Increasing alkylation at the α-carbon relative to oxygen in O-alkylated 4-chlorophenol derivatives resulted in a two-to four-fold increase in RCC, suggesting that labile C-H bonds adjacent to the O atom are inhibitory to the reaction (2, 3, 4-Cl, 5-Cl). This effect is less pronounced for 19 F to 18 F conversion, where minimal RCC differences were observed. For example, both 4-F and 5-F were obtained in excellent RCC based on HPLC isolation. We also performed the halide/ 18 F interconversion reactions with a blue light-emitting diode (LED) instead of a laser on substrates 5-Cl and 5-F. We were pleased to find that the Cl/ 18 F conversion proceeded on 5-Cl, albeit with lower RCC. Meanwhile, the 19 F/ 18 F isotopic exchange on 5-F afforded comparable RCC with LED irradiation. A meta-methyl substituent (6) resulted in moderate RCC improvement, while substituents ortho to the chlorine nucleofuge resulted in more efficient halide/ 18 F conversion (7)(8)(9). This observation is tentatively attributed to the enhanced stability of a putative captodative cation radical intermediate invoked in our previous findings 32, [43][44][45] . 2,4-Dimethoxy-substituted aryl halides (10-Cl, 10-Br, 10-I) were labelled with 18 F − , leading to 10-18 F in good, moderate and low RCCs respectively. Interestingly, 10-NO 2 , which is more electron-rich than mono-methoxy 1-NO 2 , was successfully radiofluorinated with 48.3% RCC. This observation suggests that the lack of reactivity with 1-NO 2 is probably due to a mismatch in redox potentials between 1-NO 2 and the acridinium catalyst. The radiofluorination of 2,4,6-and 2,3,4-substituted chlorobenzenes (11)(12)(13)(14) was accomplished with moderate to excellent RCCs. Desymmetrization of dihalogenated aromatics (15 and 16) was also demonstrated, although the reduced solubility of dibrominated 16 resulted in a lower RCC than for the more soluble dichlorinated analogue (15). More pronounced changes in radiofluorination efficiency were observed with O-alkylated haloarenes bearing ortho-nucleofuges (17)(18)(19)(20). Chloro-and fluoro-naphthalenes and their alkoxy-substituted derivatives were also successfully radiofluorinated (21)(22)(23). Aryl fluorides containing protected amines (24, 25-Cl, 26, 27) undergo 19 F/ 18 F conversion with moderate to good RCCs under standard labelling conditions, with aryl chloride 25-Cl demonstrating lower radiofluorination efficiency. Chloro-and fluoro-substituted heterocycles were also successfully radiofluorinated via Cl/ 18 F or 19    quinazoline fragment 37 and its analogue 38, common pharmacophores in kinase inhibitors 46 , were found to be competent substrates for direct 19 F/ 18 F exchange. Several substrate limitations exist for this methodology ( Supplementary Fig. 138). Generally, the presence of an activated benzylic group or α-heteroatom methylene unit leads to decreased yield or results in no reaction, probably because the deprotonation by fluoride is kinetically favoured. Additionally, substrates bearing more oxidizable functional groups
Method applications. We next applied our halide/ 18 F interconversion strategy to the radiolabelling of known pharmaceuticals and bioactive molecules (Fig. 3a). Clofibrate (43) and Boc-protected atomoxetine (44) were directly converted to 18 (Fig. 3b). Additionally, [ 18 F]fluorouracil (53-18 F), an important PET agent in oncology, was readily obtained through aryl-Cl/ 18 F conversion followed by a simple deprotection (Fig. 3c). The key pyrimidine intermediate 52-18 F obtained from easily available chloroarene 51 showed much higher labelling efficiency than previously reported aryl nickel (II) complex 33 and aryl iodonium ylide 27 . Moreover, the precursor can be obtained in fewer synthetic steps than the diarylether analogue that we previously used for radiodeoxyfluorination 32 . Taken together, our Cl/ 18 F exchange strategy offers a synthetically accessible alternative to [ 18 F]fluorouracil with high A m .
PET tracer exploration. We further evaluated the application of direct 19 F/ 18 F conversion within biologically relevant electron-rich arenes because of its exceptional efficiency and simplicity. Although the resulting PET agents may have reduced A m , it still represents a broadly useful technology for studying the pharmacokinetics/ pharmacodynamics of fluorine-containing drugs 36,37 or imaging transporter-mediated processes, such as synthesizing fluorinated amino-acid agents for large neutral amino-acid transporter (LAT1) imaging 48,49 . Amino-acid metabolism represents another important    class of pathways in cancer progression in addition to glucose metabolism. We were particularly interested in synthesizing 18 F-labelled tyrosine analogues 48,49 . In a proof-of-principle study, our direct 19 F/ 18 F exchange method allowed us to develop new 18 F-labelled tyrosine probes 50 via facile installation of 18 F within the aromatic core of a small library of O-methyl fluorotyrosine derivatives (54- 63; Fig. 4a). Good to excellent RCCs were observed when the fluorine nucleofuge was located at the ortho or para position relative to the methoxy group (54, 56-60, 62, 63). Lower but notable labelling was observed when the fluorine was positioned meta to the methoxy group (55, 61). After a simple deprotection (see Supplementary Section 3.6 for experimental details), ten 18 F-labelled O-methyl tyrosines were obtained and evaluated as potential PET agents in MCF-7 breast cancer models (Fig. 4b). Although radiotracers 54-58 and 60-62 are constitutional isomers, the relative position of 18 F and methoxy substitution greatly impacts their tumour uptake and clearance profiles. Most of the PET tracers demonstrated initial prominent tumour uptake at 1 h post injection (p.i) followed by washout at 3 h (54-18 F-COOH, 55-18 F-COOH, 56-18 F-COOH, 58-18 F-COOH, 59-18 F-COOH, 62-18 F-COOH, 63-18 F-COOH). By contrast, PET agents 57-18 F-COOH, 60-18 F-COOH and 61-18 F-COOH showed high and persistent retention in the MCF-7 tumour within the same timeframe (Fig. 4c). Although amino-acid analogues with initial high uptake and clearance (leading to high contrast) are great candidates for imaging applications, other analogues with prolonged tumour retention provide amino-acid backbones for potential therapy applications in which radioiodinated ( 131 I) or boronated ( 10 B) analogues can be used as cancer treatments via radioactive iodine therapy 51 or boron neutron capture therapy 52 , respectively. Although most of the 18 F-labelled tyrosine analogues demonstrated apparent pancreatic uptake, introducing one extra fluorine to the arene ring (difluorinated 63-18 F-COOH) greatly reduced uptake in the pancreas while still maintaining prominent tumour uptake. The effect of amino-acid stereodefinition was also studied using 61-18 F-COOH, which demonstrated high and persistent tumour retention. Although the contrast between the two enantiomers remained comparable at 1 h post-injection (p.i.), using L-61-18 F-COOH doubled the tumour uptake with increased tumour retention at 3 h p.i. compared with D-61-18 F-COOH. Taken together, these results suggest that our 19 F/ 18 F conversion strategy can be used to systematically investigate new PET agents with different pharmacokinetic properties for cancer imaging. Further evaluation of other O-methyl tyrosine enantiomers and other amino-acid scaffolds is currently under way.
FDOPA synthesis. In addition to developing new PET agents, we are interested in applying our halide/ 18 F interconversion method to streamline the preparation of existing PET agents (Fig. 5). For example, 6-[ 18 F]fluoro-l-DOPA ([ 18 F]FDOPA) has been used as an important radiopharmaceutical for Parkinson's disease (PD), brain cancer and other diseases since the 1980s 53 . Owing to its medical significance, multiple syntheses have been developed to produce [ 18 F]FDOPA for clinical use 19,[54][55][56][57][58][59] . Given the readily available precursor and simple reaction procedure of our halide/ 18 F exchange approach, we evaluated it for [ 18 F]FDOPA synthesis using O-methylated DOPA precursors L-64-Cl and L-64-F. We were pleased to find that radiofluorination proceeds in 4.2% and 73.4% RCC for L-64-Cl and L-64-F, respectively. The resulting product L-64-18 F was easily deprotected to l-[ 18 F]FDOPA with 97.1% RCC and >99% enantiomeric excess (e.e.). No racemization was observed under our photoredox labelling conditions. We also found that the 19 F/ 18 F conversion in L-64-F remains highly efficient after replacing the laser with more readily available LEDs, performing the reaction without ice cooling, reducing precursor concentration or shortening the reaction to 5 min at room temperature. Although these conditions would lead to decreased yield, they may be more suitable for automated synthesis (Fig. 5a). The methoxymethyl (MOM)-protected analogue (L-65) and three constitutional isomers of DOPA (66-68) were efficiently labelled via Cl/ 18 F and/or direct 19 F/ 18 F conversion as well (Fig. 5b). Encouraged by the initial successes of our method, we explored the feasibility of synthesizing l-[ 18 F]FDOPA on clinically relevant scales. Starting from 0.93-1.11 GBq of [ 18 F]TBAF, [ 18 F]FDOPA was isolated in 42% and 37.5% non-decay corrected radiochemical yield (n.d.c. RCY) when 0.01 and 0.005 mmol of L-64-F were used, respectively (Fig. 5c). Further increasing the scale to ~37 GBq [ 18 F]F − led to >11 GBq of [ 18 F]FDOPA with >30% n.d.c. RCY (>99% e.e., 1.51 GBq μmol −1 ) in 100 min (Fig. 5d). Compared with the well-established methods reported by Luxen and Lemaire 55 , our radiosynthesis has decreased A m but features the use of a stable and readily available precursor that greatly shortens the post-radiofluorination reaction sequence. Although high A m is not mandatory for investigation of the neuronal dopaminergic metabolism 54,60 , we anticipate that it could be further improved by reducing precursor loading and starting from higher amounts of activity. Overall, our 19 F/ 18 F exchange strategy for electron-rich arenes enables a new preparation of [ 18 F]FDOPA with high enantiomeric purity from the readily available isotopic precursor. Future work will focus on improving and automating this radiolabelling process to meet cGMP (current good manufacturing practice) compliance standards of clinical applications.

Conclusions
The developed photoredox-mediated halide/ 18 F conversion offers a strategy for the radiofluorination of electron-rich haloarenes -a limitation of current S N Ar methodology. The success of this method is most pronounced for aryl chlorides and fluorides, with the latter transformation enabling direct 19 F/ 18 F exchange. Its application is demonstrated by the successful radiofluorination of electron-rich halo(hetero)arenes, commercial pharmaceuticals and physiologically active compounds. Its utility for the synthesis of new PET agents is demonstrated by the rapid preparation of 18 F-labelled O-methyl tyrosine analogues and the characterization of their efficacy towards cancer imaging in a breast cancer model. Finally, direct 19 F/ 18 F exchange enables a scalable, late-stage radiosynthesis of [ 18 F]F-l-DOPA, with up to 11 GBq of the radiotracer obtained. Further improvements to this methodology, especially future work on automating this radiosynthesis protocol, should enable access to new and/or clinically important PET agents. Overall, the photoredox-mediated 18 F-labelling of unactivated aryl halides opens up new strategies for radiotracer synthesis via S N Ar disconnections.

Online content
Any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests; and statements of data and code availability are available at https://doi.org/10.1038/ s41557-021-00835-7.
Corresponding author(s): Li, Zibo; Nicewicz, David Last updated by author(s): Oct 6, 2021 Reporting Summary Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency in reporting. For further information on Nature Research policies, see our Editorial Policies and the Editorial Policy Checklist.

Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.

n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly The statistical test(s) used AND whether they are one-or two-sided Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient) AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals) For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted

Software and code
Policy information about availability of computer code Data collection PET acquisition was conducted using SuperArgus 4R (SEDECAL, Madrid, Spain). The PET images were reconstructed into a single frame by OSEM 3D option in SedecalRecon-200.123 provided by the vendor of PET scanner.

Data analysis
PET images were analyzed using AMIDE version 1.0.4 a free access tool for viewing, analyzing, and registering volumetric medical imaging data sets (http://amide.sourceforge.net). GraphPad Prism 8.0.1 was used to generate bar graph. and calculate means. HPLC data were analyzed on LabSolutions software (Version 5.6) For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.

Data
Policy information about availability of data All manuscripts must include a data availability statement. This statement should provide the following information, where applicable: -Accession codes, unique identifiers, or web links for publicly available datasets -A list of figures that have associated raw data -A description of any restrictions on data availability All the data generated or analyzed during this study are included in this article (and its Supplementary Information files). The PET imaging raw data for the animal study were deposited to a public repository (zenodo, DOI: 10.5281/zenodo.5220725 (2021)" ).

Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.

Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf

Life sciences study design
All studies must disclose on these points even when the disclosure is negative.

Sample size
The sample size was determined empirically. We used total of 15 female nude (3 batches for MCF7 xenografts) for the evaluation of different 18F-labeled tracers described in this manuscript. We took advantages of noninvasive nature of PET imaging, short half-life of F18, and the relatively slow growth of MCF-7 xenografts, and reused the animals for different tracer testing after 18F was fully decayed. Each tracer was tested in at least 3 mice bearing unilateral or bilateral subcutaneous tumors.
Data exclusions No data was excluded.

Replication
The tracer 57-18F-COOH was tested on two dates because of PET scanner had some issues and we could not collect the 3h post-injection imaging data. From these two sessions, the quantitative analysis showed that the 1 hour post injection results were reproducible. In addition, the tracer 61-18F-COOH was synthesized and tested on 10/16/2019 and 01/08/2020 using two different batches of animals; the tracer L-61-18F-COOH was synthesized and tested on 02/12/2020 and 07/24/2020 using two different batches of animals. The quantitative analysis showed that the results were reproducible for both tracers. No additional reproducibility test was performed for other tracers. No additional reproducibility test was performed for other tracers.
Randomization Each PET tracer was tested randomly in at least 3 mice bearing unilateral or bilateral subcutaneous tumors to obtain an average uptake value.

Blinding
The investigator (H.W.) who conducted PET image acquisition and data analysis was blinded to the tested compounds.
Reporting for specific materials, systems and methods We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material, system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.

Animals and other organisms
Policy information about studies involving animals; ARRIVE guidelines recommended for reporting animal research

Laboratory animals
Female BALB/c Nude Mouse (4-6 weeks old) were used for this study. Mice were group housed (no more than five mice per cage) with air temperature 24 ± 3°C, humidity 60 ± 4%, and a 12-hours light/12-hours dark cycle. Food and water were provided ad libitum.