A Practical Method for the Preparation of 18F-Labeled Aromatic Amino Acids from Nucleophilic [18F]Fluoride and Stannyl Precursors for Electrophilic Radiohalogenation

In a recent contribution of Scott et al., the substrate scope of Cu-mediated nucleophilic radiofluorination with [18F]KF for the preparation of 18F-labeled arenes was extended to aryl- and vinylstannanes. Based on these findings, the potential of this reaction for the production of clinically relevant positron emission tomography (PET) tracers was investigated. To this end, Cu-mediated radiofluorodestannylation using trimethyl(phenyl)tin as a model substrate was re-evaluated with respect to different reaction parameters. The resulting labeling protocol was applied for 18F-fluorination of different electron-rich, -neutral and -poor arylstannyl substrates in RCCs of 16–88%. Furthermore, this method was utilized for the synthesis of 18F-labeled aromatic amino acids from additionally N-Boc protected commercially available stannyl precursors routinely applied for electrophilic radiohalogenation. Finally, an automated synthesis of 6-[18F]fluoro-l-m-tyrosine (6-[18F]FMT), 2-[18F]fluoro-l-tyrosine (2-[18F]F-Tyr), 6-[18F]fluoro-l-3,4-dihydroxyphenylalanine (6-[18F]FDOPA) and 3-O-methyl-6-[18F]FDOPA ([18F]OMFD) was established furnishing these PET probes in isolated radiochemical yields (RCYs) of 32–54% on a preparative scale. Remarkably, the automated radiosynthesis of 6-[18F]FDOPA afforded an exceptionally high RCY of 54 ± 5% (n = 5).


Introduction
The enormous clinical potential of PET imaging still remains underexplored owing to the lack or poor accessibility of suitable molecular probes.Therefore, much effort has been spent in the last decades towards the development of novel labeling methods with PET nuclides for the preparation of structurally diverse imaging probes.Undeniably, 18 F-labeled ligands play an outstanding role in PET imaging.The popularity of 18 F is mainly due to its easy accessibility at a small cyclotron as well as its excellent nuclear decay properties like half-life and β + energy.Moreover, the half-life of 18 F is sufficiently long to allow shipping of 18 F-labeled probes to more distant PET facilities.Additionally, the relatively long half-life allows the accomplishment of demanding chemical conversions and long-time measurements (up to 6 h).Furthermore, the low β + energy (0.63 MeV) is ideally suited to acquiring PET images with high resolution.
In the last few years new emerging radiofluorination methods have facilitated access to probes which had been so far inaccessible or difficult to produce using conventional 18 F-labeling procedures.Particularly, methods for transition metal mediated radiofluorination pioneered by Ritter et al. and Coenen et al. have the potential to change the paradigm of radiochemistry [1][2][3].Obviously, procedures for Cu-mediated 18 F-fluorination discovered by Scott et al. and Gouverneur et al. [4][5][6][7] and further developed by inventors and others [8][9][10][11][12], enabling the preparation of 18 F-labeled aromatics and heteroaromatics regardless of their electronic properties using nucleophilic 18 F − have gained special interest.This is primarily due to the fact that these approaches do not require strictly controlled conditions (e.g., complete exclusion of oxygen and/or moisture), poorly accessible or extremely sensitive radiolabeling precursors.Moreover, these methods are easily amenable to automation [13,14].The latter is a main prerequisite for the implementation of radiolabeling procedures for cGMP production of clinically relevant PET probes.
In the seminal report on Cu-mediated radiofluorination (aryl)(mesityl)iodonium salts were used as labeling substrates [4].Nevertheless, these compounds are rather impractical for routine PET tracer production.Moreover, polyfunctionalized iodonium salts are relatively difficult to prepare.In many cases these compounds suffer from limited storage capability.Furthermore, Cu(MeCN) 4 OTf used in this procedure as a Cu source has only a short shelf life under ambient conditions.Accordingly, further efforts led to the development of procedures utilizing more stable and readily available radiolabeling precursors like aryl boronic acids and pinacol boronates [5,6].Recently, Scott et al. extended Cu-mediated 18 F-fluorination to arylstannanes.They produced a variety of 18 F-labeled arenes on a small scale and demonstrated amenability of the novel method to automation [7].Arylstannanes can be easily prepared and have a long shelf life.Stannylated compounds are well known substrates for electrophilic radiohalogenation [15,16].Fortunately, many of them are commercially available, including, precursors for 6-[ 18  Owing to the known drawbacks of electrophilic radiofluorination (i.e., max.50% RCY, significantly lower accessible amounts of [ 18 F]F 2 compared to that of 18 F − , impracticability of the preparation of n.c.a tracers, disadvantages of gas vs. liquid target, necessity to handle with F 2 /Ne gas, etc.) fluorodestannylation with nucleophilic 18 F − could substantially improve the availability of various PET probes.Unfortunately, the reported protocol is rather impractical for the production of labeled compounds on a preparative scale due to high losses of 18 F − (up to 70%) during 18 F-preprocessing before the radiolabeling step.Additionally, the applied Cu source Cu(OTf) 2 is extremely hygroscopic which may prevent its widespread application for routine radiosyntheses.
Recently, our group demonstrated that 18 F-labeled arylstannanes could be obtained by applying the protocol for alcohol-enhanced Cu-mediated radiofluorination.This approach utilizes not only bench stable Cu(py) 4 (OTf) 2 but also substantially simplifies the radiosynthesis by obviating time consuming azeotropic drying steps.However, RCYs obtained with stannyl substrates were found to be significantly lower than those with pinacol boronate or boronic acid precursors.
These preliminary findings prompted us to investigate Cu-mediated 18 F-fluorination of arylstannanes in more detail.The initial aim of this study was to devise a robust protocol for radiofluorination of commercially available stannyl precursors.First of all, the newly developed procedure should be applied for the production of 6-[ 18 F]FDOPA.This tracer is widely applied for the measurement of integrity and function of the nigrostriatal dopaminergic system, e.g., in Parkinson's disease [17][18][19][20][21] as well as for the detection and staging of neuroendocrine tumors [22][23][24][25].Numerous protocols for the production of 6-[ 18 F]FDOPA via nucleophilic radiofluorination have been published [9,13,[26][27][28].However, the majority of them are cumbersome, poorly reproducible and/or use insufficiently stable radiolabeling precursors and/or highly corrosive reagents [29].Moreover, the production of 6-[ 18 F]FDOPA is frequently used in the literature to demonstrate the potential of novel 18 F-labeling techniques.
Likewise, a broader clinical application of 6-[ 18 F]FMT, a structural analog of 6-[ 18 F]FDOPA with improved imaging properties, and other radiofluorinated aromatic amino acids is hampered by the lack of simple production routes [30,31].Therefore, the labeling method should also be applied to obtain these compounds in high yields.Finally, the method should be transferred to automated synthesis modules for cGMP production.

Results and Discussion
2.1.Effect of Different Salts on 18 F-Recovery from Anion Exchange Resin and 18 F-Incorporation Optimization of radiofluorodestannylation was carried out using trimethyl(phenyl)tin (PhSnMe 3 ) as a model substrate.First, the 18 F-elution capacity of different tetraethylammonium salts in MeOH was studied (Figure 1).Almost complete radioactivity recovery (95-97%) was achieved with 2.5 µmol of all four examined salts. 18F-Recovery decreased to 76-89% if only 0.5 µmol salt was used.Likewise, a broader clinical application of 6-[ 18 F]FMT, a structural analog of 6-[ 18 F]FDOPA with improved imaging properties, and other radiofluorinated aromatic amino acids is hampered by the lack of simple production routes [30,31].Therefore, the labeling method should also be applied to obtain these compounds in high yields.Finally, the method should be transferred to automated synthesis modules for cGMP production.

Effect of Different Salts on 18 F-Recovery from Anion Exchange Resin and 18 F-Incorporation
Optimization of radiofluorodestannylation was carried out using trimethyl(phenyl)tin (PhSnMe3) as a model substrate.First, the 18 F-elution capacity of different tetraethylammonium salts in MeOH was studied (Figure 1).Almost complete radioactivity recovery (95-97%) was achieved with 2.5 µmol of all four examined salts. 18F-Recovery decreased to 76-89% if only 0.5 µmol salt was used.Next, in trying to optimize the conditions for alcohol-enhanced Cu-mediated radiofluorination of aryl stannanes the dependency of radioactivity recovery and 18 F-incorporation on different salts was investigated using solutions of Et4NHCO3, Et4NOTf, KOTf/K2.2.2 and Bu4POMs in nBuOH.(Table 1).With KOTf/K2.2.2 > 80% of 18 F − was eluted from the resin.For other salts radioactivity recovery amounted to 71-76%.The resulting solutions were diluted with a solution of PhSnMe3 and Cu(py)4(OTf)2 in DMA and heated to give [ 18 F]FPh.Surprisingly, in the case of KOTf/K2.2.2 18 F-incorporation did not exceed a RCC of 10%.In contrast, if ammonium or phosphonium salts were applied, RCCs of 60-69% were achieved [32].

Dependency of 18 F-Recovery and 18 F-Incorporation Yields on the Type of Anion Exchange Cartridge
The type of anion exchange cartridge substantially influenced the efficacy of 18 F-elution and especially the subsequent radiolabeling step (Table 2).The highest radioactivity recovery was observed for Strata X-CO3 followed by QMA-CO3 cartridges (81% and 73%, respectively).However, while using Strata X-CO3 cartridges only fair RCCs of 24 ± 15% were obtained, 18 F-incorporation amounted to 73 ± 8% if QMA-CO3 cartridges were applied.Next, in trying to optimize the conditions for alcohol-enhanced Cu-mediated radiofluorination of aryl stannanes the dependency of radioactivity recovery and 18 F-incorporation on different salts was investigated using solutions of Et 4 NHCO 3 , Et 4 NOTf, KOTf/K 2.2.2 and Bu 4 POMs in nBuOH.(Table 1).With KOTf/K 2.2.2 > 80% of 18 F − was eluted from the resin.For other salts radioactivity recovery amounted to 71-76%.The resulting solutions were diluted with a solution of PhSnMe 3 and Cu(py) 4 (OTf) 2 in DMA and heated to give [ 18 F]FPh.Surprisingly, in the case of KOTf/K 2.2.2 18 F-incorporation did not exceed a RCC of 10%.In contrast, if ammonium or phosphonium salts were applied, RCCs of 60-69% were achieved [32].

Dependency of 18 F-Recovery and 18 F-Incorporation Yields on the Type of Anion Exchange Cartridge
The type of anion exchange cartridge substantially influenced the efficacy of 18 F-elution and especially the subsequent radiolabeling step (Table 2).The highest radioactivity recovery was observed for Strata X-CO 3 followed by QMA-CO 3 cartridges (81% and 73%, respectively).However, while using Strata X-CO 3 cartridges only fair RCCs of 24 ± 15% were obtained, 18 F-incorporation amounted to 73 ± 8% if QMA-CO 3 cartridges were applied.]Fluoride (~50 MBq) was loaded onto a QMA cartridge from the male side.The cartridge was washed with nBuOH (1 mL) in the same direction and flushed with air (5 mL).Afterwards 18 F -was eluted from the female side with a solution of the respective salt (11 µmol) in nBuOH (300 µL).A solution of PhSnMe 3 (14.5 mg, 60 µmol) and Cu(py) 4 (OTf) 2 (20.3 mg, 30 µmol) in DMA (700 µL) was added, the reaction mixture was heated at 100   F-Recovery as well as RCCs were strongly dependent on the nature of the respective alcohol (Figure 2).Whereas 18 F-recovery was highest for short-chained alcohols (for MeOH, EtOH and TFE > 85%), RCCs increased if higher alcohols were used (cf.RCCs for nBuOH, tBuOH and nAmOH > 70%).While TFE allowed efficient 18 F-elution from the anion exchange resin, no 18 F-incorporation was observed if this alcohol was used as reaction co-solvent.This should be attributed to the acidic nature of TFE (pK a = 12.4) [33] and even more to a very strong hydrogen bond donor power of trifluoroethanol [α (TFE) = 1.51] [34].Consequently, TFE solvates halogenide ions much stronger than MeOH [35] and, therefore, should strongly decrease nucleophilicity of 18 F − .nBuOH and nAmOH represented a reasonable compromise, between, on the one hand, sufficient 18 F-recovery and, on the other hand, high RCCs.Notably, the reaction was very sensitive to water: 18 F-incorporation halved at the water content of 0.5% (Figure 3).After addition of 15 µL H 2 O (1.5% final concentration) RCC fell below 10%.Next, we evaluated the influence of nBuOH content on RCCs (Figure 4).Addition of up to 20-30% nBuOH was well tolerated and did not cause a significant decrease of RCCs.Further increase of nBuOH concentration resulted in lower RCCs of 58 ± 9 and 42 ± 2% in 40% and 50% n-butanolic solutions, respectively.In contrast to Cu-mediated radiofluorination of arylboronic acids and pinacol arylboronates where a pronounced increase of RCCs in the presence of nBuOH took place (nBuOH content of 1-30%), no increase of 18 F-incorporation yield was observed for the stannylated precursor.
The remarkable tolerance of Cu-mediated radiofluorination towards alcohols could be presumably attributed to solvation of 18 F − with alcohols, which obviously decreases its basicity.This interaction is, however, not strong enough to significantly affect the nucleophilicity of [ 18 F]fluoride [36,37].
We proposed that the first steps of Cu-mediated radiofluorination of boronate and stannyl precursors consist of anion metathesis followed by air oxidation of the Cu(II) to the Cu(III) complex stabilized by py, and in alcohol-containing media by alcoholate ligands (Figure 5) [38].Thereafter, transmetalation should afford 18 F-fluorinated (probably, polynuclear) [39] aryl(III)cuprates with RO and py ligands.Finally, reductive elimination liberates either the desired [ 18 F]ArF or the ArOR side product from the Chan-Lam coupling [40].
In our opinion, the beneficial effect of alcohols could be mainly attributed to their capability to stabilize the transition state of the rate limiting B/Cu(III) transmetalation step by hydrogen bonding interactions between the hydroxyl hydrogens of alcohol molecules and oxygens of B(OH) 2 or BPin groups.This beneficial effect should be more pronounced for B(OH) 2 than for BPin and cannot occur in the case of aryltrialkylstannanes where no hydrogen bond formation can occur.Indeed, a more distinct beneficial effect was observed for aryl boronic acids than for aryl pinacol boronates [9].This effect was absent for arylstannanes since no hydrogen bond formation can take place.Additionally, in tBuOH medium where the stabilization via hydrogen bond formation, especially in the case of pinacol boronate substrates, is limited by sterical hindrance of tBu group, a deleterious and much less pronounced beneficial effect was observed for ArBpin and ArB(OH) 2 substrates, respectively.In contrast, in the case of arylstannane precursors the highest RCCs were observed in the presence tBuOH in comparison to the other alcohols [36,37].
Molecules 2017, 22, 2231 5 of 25 molecules and oxygens of B(OH)2 or BPin groups.This beneficial effect should be more pronounced for B(OH)2 than for BPin and cannot occur in the case of aryltrialkylstannanes where no hydrogen bond formation can occur.Indeed, a more distinct beneficial effect was observed for aryl boronic acids than for aryl pinacol boronates [9].This effect was absent for arylstannanes since no hydrogen bond formation can take place.Additionally, in tBuOH medium where the stabilization via hydrogen bond formation, especially in the case of pinacol boronate substrates, is limited by sterical hindrance of tBu group, a deleterious and much less pronounced beneficial effect was observed for ArBpin and ArB(OH)2 substrates, respectively.In contrast, in the case of arylstannane precursors the highest RCCs were observed in the presence tBuOH in comparison to the other alcohols [36,37].      in the corresponding anhydrous alcohol (300 µL) (see legend of Table 1); to this solution a solution of trimethyl(phenyl)tin (14.5 mg, 60 µmol) and Cu(py) 4 (OTf) 2 (20.3 mg, 30 µmol) in DMA (700 µL) was added, the reaction mixture was heated at 100 molecules and oxygens of B(OH)2 or BPin groups.This beneficial effect should be more pronounced for B(OH)2 than for BPin and cannot occur in the case of aryltrialkylstannanes where no hydrogen bond formation can occur.Indeed, a more distinct beneficial effect was observed for aryl boronic acids than for aryl pinacol boronates [9].This effect was absent for arylstannanes since no hydrogen bond formation can take place.Additionally, in tBuOH medium where the stabilization via hydrogen bond formation, especially in the case of pinacol boronate substrates, is limited by sterical hindrance of tBu group, a deleterious and much less pronounced beneficial effect was observed for ArBpin and ArB(OH)2 substrates, respectively.In contrast, in the case of arylstannane precursors the highest RCCs were observed in the presence tBuOH in comparison to the other alcohols [36,37].

Dependency of RCC on Reaction Solvent
The type of reaction solvent had a significant influence on RCCs (Figure 6).Thus, DMA and N-methyl-2-pyrrolidone (NMP) afforded the highest RCCs of [ 18 F]FPh of 72% and 73%, respectively.

Dependency of RCC on Reaction Solvent
The type of reaction solvent had a significant influence on RCCs (Figure 6).Thus, DMA and N-methyl-2-pyrrolidone (NMP) afforded the highest RCCs of [ 18 F]FPh of 72% and 73%, respectively.

Dependency of RCC on Reaction Solvent
The type of reaction solvent had a significant influence on RCCs (Figure 6).Thus, DMA and N-methyl-2-pyrrolidone (NMP) afforded the highest RCCs of [ 18 F]FPh of 72% and 73%, respectively.

Dependency of RCC on Temperature and Time
The dependency of temperature (Figure 7) and time (Figure 8) on RCC revealed rapid reaction kinetics.Maximal RCCs were achieved already after 5 min incubation at 100 • C. The optimal reaction temperature amounted to 100 In N,N,N',N'-tetramethylurea (TMU), DMF and DMSO RCCs amounted to only 28%, 9% and 7%, respectively.In pyridine, N-methylformamide (NMF) and tBuOH no 18 F-incorporation took place.

Dependency of RCC on Temperature and Time
The dependency of temperature (Figure 7) and time (Figure 8) on RCC revealed rapid reaction kinetics.Maximal RCCs were achieved already after 5 min incubation at 100 °C.The optimal reaction temperature amounted to 100 °C.1).A solution of PhSnMe3 (14.5 mg, 60 µmol) and Cu(py)4(OTf)2 (20.3 mg, 30 µmol) in DMA (700 µL) was added.The reaction mixture was heated at 100 °C for different times, cooled down, diluted with H2O (1 mL) and analyzed by HPLC.All experiments were carried out in triplicate.

Dependency of RCC on Temperature and Time
The dependency of temperature (Figure 7) and time (Figure 8) on RCC revealed rapid reaction kinetics.Maximal RCCs were achieved already after 5 min incubation at 100 °C.The optimal reaction temperature amounted to 100 °C.1).A solution of PhSnMe3 (14.5 mg, 60 µmol) and Cu(py)4(OTf)2 (20.3 mg, 30 µmol) in DMA (700 µL) was added.The reaction mixture was heated at 100 °C for different times, cooled down, diluted with H2O (1 mL) and analyzed by HPLC.All experiments were carried out in triplicate.

Dependency of RCC on Precursor Amount and Precursor to Cu(py)4(OTf)2 Ratio
The amount of the stannyl substrate (Figure 9) and Cu(py)4(OTf)2 (Figure 10) was adjusted to reduce costs and simplify the purification step.If 30-60 µmol PhSnMe3 were applied, [ 18 F]FPh was

Dependency of RCC on Precursor Amount and Precursor to Cu(py) 4 (OTf) 2 Ratio
The amount of the stannyl substrate (Figure 9) and Cu(py) 4 (OTf) 2 (Figure 10) was adjusted to reduce costs and simplify the purification step.If 30-60 µmol PhSnMe 3 were applied, [ 18 F]FPh was obtained in RCCs of ≥70%.At 20 and 10 µmol precursor, a decline of 18 F-incorporation to 63% and 44%, respectively, was observed.Consequently, all further experiments were performed with 30 µmol of the corresponding stannyl precursor.This precursor amount is higher in comparison to that used by Makaravage et al. which amounted to 10 µmol [7].However, owing to the reasonable accessibility of arylstannanes this quantity may be considered as acceptable for the majority of applications.Occasionally, it may be difficult to separate larger amounts of radiolabeling precursor and/or product of its protodestannylation from a radiolabeled compound even when using preparative HPLC.Yet, for all PET samples described herein this problem has not been encountered.The novel protocol for 18 F-fluorodestannylation was rather insensitive to the stannane/Cu salt ratio.Comparable RCCs were achieved at PhSnMe3/Cu(py)4(OTf)2 ratios of 3:4 to 2:3.A marked decrease of conversion was first observed at a substrate/Cu salt ratio of 3:1.1).A solution of PhSnMe3 (7.2 mg, 30 µmol) and a given amount of Cu(py)4(OTf)2 in DMA (700 µL) was added, the mixture was heated under air at 100 °C for 10 min, cooled down, diluted with H2O (1 mL) and analyzed by HPLC.All experiments were carried out in triplicate.

Optimized Protocol of 18 F-Fluorodestannylation
Based on the optimization study, a novel protocol of radiofluorodestannylation was developed.In order to obviate the notable loss of radioactivity during 18 F-recovery using nBuOH, we modified obtained in RCCs of ≥70%.At 20 and 10 µmol precursor, a decline of 18 F-incorporation to 63% and 44%, respectively, was observed.Consequently, all further experiments were performed with 30 µmol of the corresponding stannyl precursor.This precursor amount is higher in comparison to that used by Makaravage et al. which amounted to 10 µmol [7].However, owing to the reasonable accessibility of arylstannanes this quantity may be considered as acceptable for the majority of applications.Occasionally, it may be difficult to separate larger amounts of radiolabeling precursor and/or product of its protodestannylation from a radiolabeled compound even when using preparative HPLC.Yet, for all PET samples described herein this problem has not been encountered.The novel protocol for 18 F-fluorodestannylation was rather insensitive to the stannane/Cu salt ratio.Comparable RCCs were achieved at PhSnMe3/Cu(py)4(OTf)2 ratios of 3:4 to 2:3.A marked decrease of conversion was first observed at a substrate/Cu salt ratio of 3:1.1).A solution of PhSnMe3 (7.2 mg, 30 µmol) and a given amount of Cu(py)4(OTf)2 in DMA (700 µL) was added, the mixture was heated under air at 100 °C for 10 min, cooled down, diluted with H2O (1 mL) and analyzed by HPLC.All experiments were carried out in triplicate.

Optimized Protocol of 18 F-Fluorodestannylation
Based on the optimization study, a novel protocol of radiofluorodestannylation was developed.In order to obviate the notable loss of radioactivity during 18 F-recovery using nBuOH, we modified This precursor amount is higher in comparison to that used by Makaravage et al. which amounted to 10 µmol [7].However, owing to the reasonable accessibility of arylstannanes this quantity may be considered as acceptable for the majority of applications.Occasionally, it may be difficult to separate larger amounts of radiolabeling precursor and/or product of its protodestannylation from a radiolabeled compound even when using preparative HPLC.Yet, for all PET samples described herein this problem has not been encountered.The novel protocol for 18 F-fluorodestannylation was rather insensitive to the stannane/Cu salt ratio.Comparable RCCs were achieved at PhSnMe 3 /Cu(py) 4 (OTf) 2 ratios of 3:4 to 2:3.A marked decrease of conversion was first observed at a substrate/Cu salt ratio of 3:1.

Optimized Protocol of 18 F-Fluorodestannylation
Based on the optimization study, a novel protocol of radiofluorodestannylation was developed.In order to obviate the notable loss of radioactivity during 18 F-recovery using nBuOH, we modified the elution procedure.We used Et 4 NOTf in MeOH for 18 F-elution according to Richarz et al. [41,42].After elution, low boiling methanol was removed within 2-5 min at 100 • C, and a solution of arylstannane precursor and Cu(py) 4 (OTf) 2 (30 µmol of each) in pure DMA (1 mL) was added to the residue.Thus, owing to the absence of the beneficial effect we did not use nBuOH.After that, the reaction mixture was heated under air at 100 • C for 10 min.
The scope of this protocol was evaluated using several model arylstannanes (Figure 11).The method worked equally well if either SnMe 3 or SnBu 3 precursors were applied.Substrates with electron-donating and electron-neutral substituents in mand p-positions (Figure 11, entries 2, 5 and 6) were radiolabeled in moderate to high RCCs.The introduction of a methoxy group into o-position (entry 4) resulted in lower RCCs, presumably due to unfavorable interactions of the substituent with the leaving group, thereby impeding transmetalation.Notably, Scott et al. prepared o-[ 18 F]fluoroanisole in a RCC of 48 ± 4% using Cu(py) 4 (OTf) 2 formed in situ from Cu(OTf) 2 in the presence of an excess of pyridine [7].The excess of pyridine, presumably, can additionally stabilize the Cu-complex and thus can overcome the deleterious effect of the o-MeO group.Fair to moderate RCCs were obtained with precursors with electron-withdrawing substituents (entry 3).
Molecules 2017, 22, 2231 9 of 25 the elution procedure.We used Et4NOTf in MeOH for 18 F-elution according to Richarz et al. [41,42].After elution, low boiling methanol was removed within 2-5 min at 100 °C, and a solution of arylstannane precursor and Cu(py)4(OTf)2 (30 µmol of each) in pure DMA (1 mL) was added to the residue.Thus, owing to the absence of the beneficial effect we did not use nBuOH.After that, the reaction mixture was heated under air at 100 °C for 10 min.The scope of this protocol was evaluated using several model arylstannanes (Figure 11).The method worked equally well if either SnMe3 or SnBu3 precursors were applied.Substrates with electron-donating and electron-neutral substituents in m-and p-positions (Figure 11, entries 2, 5 and 6) were radiolabeled in moderate to high RCCs.The introduction of a methoxy group into o-position (entry 4) resulted in lower RCCs, presumably due to unfavorable interactions of the substituent with the leaving group, thereby impeding transmetalation.Notably, Scott et al. prepared o-[ 18 F]fluoroanisole in a RCC of 48 ± 4% using Cu(py)4(OTf)2 formed in situ from Cu(OTf)2 in the presence of an excess of pyridine [7].The excess of pyridine, presumably, can additionally stabilize the Cu-complex and thus can overcome the deleterious effect of the o-MeO group.Fair to moderate RCCs were obtained with precursors with electron-withdrawing substituents (entry 3).Finally, 18 F-labeled anle186b was successfully prepared for the first time in RCC of 62% and in 48% isolated RCY.This 3,5-diaryl substituted pyrazole is able to bind to pathological protein aggregates in α-synucleinopathies and prion disease [43,44].Consequently, [ 18 F]anle186b could be potentially suitable for imaging of such pathologies.

Preparation of 18 F-Labeled Aromatic Amino Acids
Once the optimized protocol for radiofluorination of arylstannanes had been established, we turned to the production of clinically relevant 18 F-labeled aromatic amino acids.Unfortunately, direct radiolabeling of commercially available N-monoBoc 6-SnMe3 substituted phenylalanine derivatives Finally, 18 F-labeled anle186b was successfully prepared for the first time in RCC of 62% and in 48% isolated RCY.This 3,5-diaryl substituted pyrazole is able to bind to pathological protein aggregates in α-synucleinopathies and prion disease [43,44].Consequently, [ 18 F]anle186b could be potentially suitable for imaging of such pathologies.

Preparation of 18 F-Labeled Aromatic Amino Acids
Once the optimized protocol for radiofluorination of arylstannanes had been established, we turned to the production of clinically relevant 18 F-labeled aromatic amino acids.Unfortunately, direct radiolabeling of commercially available N-monoBoc 6-SnMe 3 substituted phenylalanine derivatives afforded radiolabeled intermediates in poor RCCs of 5-6%, presumably, due to concurrent intramolecular Chan-Lam coupling.This will furnish the respective indolines instead of the desired radiolabeled products via attack of the intermediate arylcuprate on the amide anion formed by the proton abstraction with sufficiently basic "naked" fluoride [5].

General
Chemicals and solvents were purchased from Sigma-Aldrich GmbH (Steinheim, Germany), Fluka AG (Buchs, Switzerland), TCI EUROPE N.V. (Zwijndrecht, Belgium), ChemPUR GmbH (Karlsruhe, Germany), Merck KGaA (Darmstadt, Germany) and ABCR GmbH (Karlsruhe, Germany) and used as delivered.Anhydrous solvents were purchased from Sigma-Aldrich GmbH (Steinheim, Remarkably, 6-[ 18 F]FDOPA was obtained in a high RCY of 54 ± 5% (n = 5) and in excellent enantiomeric, chemical and radiochemical purity.To the best of our knowledge, this is the highest value reported for the synthesis of this important PET tracer.The highest RCYs of n.c.a.6-[ 18 F]FDOPA achieved to date according to the protocols for the automated preparation of this tracer reported by Lemaire et al. [29,53,54] and by Hoepping et al. [29,55] To a solution of 1-(benzo[d] [1,3]dioxol-5-yl)ethan-1-one (1.5 g, 9.1 mmol), in anhydrous THF (20 mL) was added 1 m LiHMDS in THF (27.3 mL) and the resulting solution was stirred for 1 h at −80 • C. The solution was warmed to room temperature and stirred for 2 h.Thereafter, it was cooled to −80 • C and 3-bromobenzoyl chloride (1.2 mL, 2.0 g, 9.1 mmol) was added dropwise.The solution was allowed to warm to room temperature and stirred for additional 18 h.Afterwards, a saturated solution of NH 4 Cl (50 mL) was added, the pH was adjusted to 7.0 and the mixture was extracted with EtOAc (3 × 50 mL).The combined organic layers were washed with brine (100 mL), dried and concentrated under reduced pressure.The residue was purified by column chromatography (Et 2 O/petroleum ether, 1:4) affording the title compound [57] 6.91 mmol) and hydrazine monohydrate (1 mL, 98%, 13.83 mmol, 2 eq.) in ethanol (30 mL) was refluxed for 3 h.Water was added to the clear yellow solution and resulting precipitate was collected by filtration, washed with water and dried under vacuum to provide the title compound [57].1H-pyrazole (550 mg, 1.6 mmol) and Pd(PPh 3 ) 4 (185 mg, 0.16 mmol, 0.1 eq.) were evacuated and purged with argon (three times).Anhydrous 1,4-dioxane (2 mL) followed by hexamethylditin (830 µL, 1.31 g, 4 mmol, 2.5 eq.) was added, and the reaction mixture was heated to 100 • C for 18 h.The black suspension was filtered through a plug of Celite. 1 m TBAF in THF (2 mL) was added to the filtrate; the mixture was stirred for 30 min and diluted with EtOAc (50 mL).The resulting solution was washed with water (50 mL), brine (50 mL), dried and concentrated under reduced pressure.The crude product was purified by column chromatography and by recrystallization from hexane contained a small amount of CH 2 Cl 2 .

General Procedures
All radiosyntheses were carried out using anhydrous DMA and nBuOH stored over molecular sieves (available from "Acros", Geel, Belgium, or "Aldrich").Cu(OTf) 2 (py) 4 was stored under ambient conditions without any precautions.
[ 18 F]Fluoride was produced by the 18 O(p,n) 18 F reaction by bombardment of enriched [ 18 O]water with 16.5 MeV protons using a BC1710 cyclotron (The Japan Steel Works Ltd., Shinagawa, Japan) at the INM-5 (Forschungszentrum Jülich).
All radiolabeling experiments were carried out under ambient or synthetic air.Each radiochemical experiment was carried out at least in triplicates if not otherwise mentioned.Standard deviations (SD) were calculated by the least-square method.All experiments were carried out by using one-pot procedure.Before the determination of radiochemical conversions (RCCs), reaction mixtures were always diluted with H 2 O (1-4 mL) to dissolve any 18 F-fluoride adsorbed onto the reaction vessel walls.The loss of radioactivity on the vessel walls did not exceed 13 ± 2% from the starting activity (n > 100).All radiochemical yields (RCYs) are decay corrected and radiochemical purities (RCPs) were determined after purification.

Processing [ 18 F]Fluoride
Aqueous [ 18 F]fluoride was loaded onto an anion-exchange resin (e.g., QMA cartridge).It should be noted that aqueous [ 18 F]fluoride was loaded onto the cartridge from the male side, whereas flushing, washing and 18 F − elution were carried out from the female side.If the QMA cartridge had been loaded, flushed and eluted from the female side only, sometimes a significant amount of [ 18 F]fluoride remained on the resin (this is probably because QMA-light (46 mg) cartridges have a single frit on the male side but four frits on the female side).
For automated syntheses the following system was used: WellChrom Spectro-photometer K-2501 UV/Vis detector, BlueShadow Pump 80P from KNAUER Wissenschaftliche Geräte GmbH, Berlin, Germany and AD 1422 PIN-photodiode and scintillator detector from Eckert & Ziegler Strahlen-und Medizintechnik AG, Berlin, Germany was connected directly to the automated module.UV and radioactivity detectors were connected in series, giving a time delay of 0.1-0.9min depending on the flow rate. 18F-Labeled compounds were identified by co-injection of the unlabeled reference compounds.The completeness of the radioactivity elution was controlled by analyzing of the same sample amount choosing a column bypass.

Automated Radiosyntheses
All automated radiosyntheses were carried out in a home-made synthesis module.FFKM valves (Christian Bürkert GmbH&Co.KG, Ingelfingen, Germany) were applied.All connections between the valves were made using PTFE tubes and PEEK fittings.The flow scheme for the preparation of radiolabeled amino acids is depicted in Figures 13 and S1.Synthetic air and He (Westfalen AG, Muenster, Germany) were used as operating gases.(700 µL) was added, the mixture was heated under air at 100 • C for 10 min, cooled down, diluted with H 2 O (1 mL) and analyzed by HPLC.
3.5.17.Optimized Procedure for 18 F-Fluorodestannylation-General Procedure (GP6) [ 18 F]Fluoride (50-100 MBq) was loaded on an anion exchange cartridge (QMA-CO 3 , preconditioned with 1 mL water and dried with air) from the male side.The cartridge was rinsed with MeOH (1 mL) and dried with air, then [ 18 F]fluoride was eluted with a methanolic solution (500 µL) of Et 4 NOTf (2.79 mg, 10 µmol).Methanol was removed under reduced pressure (600 mBar) in a stream of argon at 100 • C within 3 min.Afterwards, the pressure was reduced to 50 mBar and the reaction vial was purged with air.A solution of the corresponding precursor (30 µmol) and Cu(OTf) 2 (py) 4 (20.3 mg, 30 µmol) in DMA (1 mL) was added, the reaction mixture was stirred at 100 • C for 10 min and cooled down to room temperature in an ice bath.The reaction mixture was quenched with water (4 mL) and analyzed by HPLC.
3.5.18.Manual Synthesis of Radiolabeled Amino Acids-General Procedure 7 (GP7) [ 18 F]Fluoride (200-300 MBq) was loaded onto an anion exchange cartridge (QMA-CO 3 preconditioned with 1 mL water and dried with air) from the male side.The cartridge was washed with MeOH (1 mL) and dried with air.Thereafter, [ 18 F]fluoride was eluted into the reaction vial using a solution of Et 4 NOTf (2.79 mg, 10 µmol) in MeOH (500 µL).MeOH was removed under reduced pressure (600 mBar) using a stream of air at 100 • C within 5 min.A solution of Cu(OTf) 2 (py) 4 (40.7 mg, 60 µmol) and the corresponding precursor (30 µmol) in DMA (1 mL) was added.The reaction mixture was stirred at 100 • C for 10 min, and cooled down to room temperature in an ice bath.The reaction mixture was quenched with water (2 mL) and loaded in Sep-Pak C18 Plus light Cartridge.The cartridge was washed with 5 mL water and the product was eluted with 1 mL EtOH.EtOH was removed under reduced pressure (600 mBar) using a stream of air at 120 • C within 5 min.48% HBr (1 mL) was added and the reaction mixture was stirred at 130 • C for 10 min.Hydrolysis of the protected [ 18 F]OMFD was carried out using 38% HCl at 100 • C for 10 min.The reaction mixture was cooled down, diluted with H 2 O (3 mL) and analyzed by HPLC.RCC was calculated from amount of 18 F − loaded onto QMA cartridge, radioactivity amount in the reaction vial after hydrolysis step and HPLC chromatogram.

Figure 1 .
Figure 1. 18F-Recovery from anion exchange resin with MeOH solutions of different tetramethylammonium salts.Conditions: [ 18 F]Fluoride (~50 MBq) was fixed on a QMA-CO3 cartridge from the male side and the cartridge was rinsed with MeOH (1 mL) in the same direction.Finally, [ 18 F]fluoride was eluted with a solution of Et4NX in MeOH (500 µL) from the female side.

Figure 1 .
Figure 1. 18F-Recovery from anion exchange resin with MeOH solutions of different tetramethylammonium salts.Conditions: [ 18 F]Fluoride (~50 MBq) was fixed on a QMA-CO 3 cartridge from the male side and the cartridge was rinsed with MeOH (1 mL) in the same direction.Finally, [ 18 F]fluoride was eluted with a solution of Et 4 NX in MeOH (500 µL) from the female side.

Figure 11 .
Figure 11.Substrate scope of the optimized protocol for 18 F-fluorodestannylation. a RCC ± SD. b RCY, single experiment was carried out.

Figure 11 .
Figure 11.Substrate scope of the optimized protocol for 18 F-fluorodestannylation. a RCC ± SD. b RCY, single experiment was carried out.

Molecules 2017 ,
22, 2231 10 of 25 radiolabeled products via attack of the intermediate arylcuprate on the amide anion formed by the proton abstraction with sufficiently basic "naked" fluoride [5].

3. 5 . 7 .
Recovery of18 F − from Anion Exchange Resin with MeOH Solutions of Different Tetramethylammonium Salts [ 18 F]Fluoride (~50 MBq) was fixed on QMA-CO 3 cartridge from the male side, the cartridge was washed with MeOH (1 mL) in the same direction.Finally, [ 18 F]fluoride was eluted with a solution of Et 4 NX in MeOH (500 µL) from the female side.

Table 2 .
• C for 10 min (300 µL) under air, diluted with H 2 O (1 mL) and analyzed by HPLC.All experiments were carried out in triplicate.Dependency of18F-recovery and18F-incorporation yields on the type of anion exchange cartridge.Conditions: [ 18 F]Fluoride (~50 MBq) was eluted from the respective anion exchange cartridge with a solution of Et 4 NOTf (3.1 mg, 11 µmol) in nBuOH (300 µL) (see legend of Table1