Enantioconvergent Copper-Catalysed Difluoromethylation of Alkyl Halides

Abstract Stereochemical-controlled hydrogen bond donors play essential roles in the pharmaceutical industry. Consequently, organic molecules that bear difluoromethyl (CF2H) groups at chiral centers are emerging as pivotal components in pharmaceuticals due to their distinct hydrogenbonding property. However, a general approach for introducing CF2H groups in an enantioselective manner remained elusive. Here, we show that enantioconvergent difluoromethylation of racemic alkyl electrophiles, through alkyl radical intermediates, represents a new strategy for constructing CF2H-containing stereocenters. This strategy is enabled by using copper catalysts bound with a chiral diamine ligand bearing electron-deficient phenyl groups, and a nucleophilic difluoromethyl-zinc reagent. This method allows for the high-yield conversion of a diverse range of alkyl halides into their alkyl-CF2H analogs with excellent enantioselectivity (up to 99% e.e.). Mechanistic studies, supported by DFT calculations, revealed a route involving asymmetric difluoromethylation of alkyl radicals and crucial non-covalent interactions in the enantio-determining steps.

recently showcased by the development of two pharmaceutical agents, Inavolisib 5 and LPC-233 6 (Fig. 1a).The efficacy of the former has been attributed to the hydrogen-bonding ability of its CF2H group with a serine residue of phosphoinositide 3-kinase, whereas the potency of the latter molecule has been attributed to the electrostatic interaction of the fluorine atoms with a lysine residue of the enzyme LpxC.
8][9] Conventional approaches for constructing CF2H-containing stereogenic centers relied on either the deoxyfluorination of enantioenriched aldehydes 10,11 or the asymmetric transformation of CF2Hcontaining prochiral molecules. 12,13 hese approaches generally suffered from either the narrow substrate scope of the fluorination reagents or the limited availability of the prochiral precursors.Recent advances in asymmetric difluoromethylation have focused on the manipulation of C(sp 2 ) electrophiles.For example, the Hu group has made progress for the asymmetric difluoromethylation of aldehydes 14 and imines 15 , albeit with moderate enantioselectivity.The Jacobsen group has reported an unusual migratory geminal difluorination of styrenes for the construction of CF2H-containing stereocenters at benzylic positions. 20Shen and Mikami have recently demonstrated the difluoromethylation of allyl substrates via either SN2' 16 or Michael addition 17 pathways, although these transformations yielded products with modest enantioselectivity.An alternative strategy in this theme involves the use of difluoroenoxysilanes as masked CF2H sources, as exemplified by Zhang's asymmetric difluoroalkylation of propargyl sulfonate 18 and Ma's synthesis of enantioenriched isoindolones. 19Enantioselective difluoromethylation via the insertion of difluorocarbene has recently emerged, enabling the synthesis of specific CF2H-containing moieties, including α-amino acids 20 and β-ketoesters. 21iven the unique properties of CF2H groups, a new and more general mode of asymmetric difluoromethylation is highly desired to expand the molecular architectures containing the CF2H bioisostere.Nucleophilic difluoromethylation of racemic C(sp 3 )-electrophiles in an enantioconvergent manner represents one of the most mechanistically straightforward approaches to the construction of chiral C(sp 3 )-CF2H moieties (Fig. 1b).Nonetheless, despite the considerable progress in the use of chiral transition-metal catalysts for the substitution of racemic alkyl electrophiles with carbon- [22][23][24] oxygen- 25 and nitrogen-centered nucleophiles, [26][27][28] enantioselective difluoromethylation via this pathway remained largely underdeveloped.0][31] This lag in the development of nucleophilic fluoroalkylation reactions is attributed to the sluggish oxidative addition of the metal-fluoroalkyl species with alkyl electrophiles and the slower reductive elimination of metal-fluoroalkyl species compared to their non-fluorinated counterparts.
Our group [32][33][34][35] and others 36 have recently demonstrated the potential of copper catalysts in facilitating the transfer of CF2H groups to aliphatic sites through alkyl radical intermediates.These catalytic methods enabled the synthesis of racemic CF2H-containing products from various alkyl electrophiles.We recently questioned the feasibility of enantioselective transfer of CF2H groups to alkyl radicals when the copper catalysts are in a chiral environment.This could enable a general strategy for the construction of difluoromethylated stereocenters.We realized that several challenges should be addressed to achieve such an enantioselective transformation.First, the intrinsic instability of the [Cu-CF2H] species compared to their CF3 counterparts 37,38 requires a swift single electron transfer (SET) event between this intermediate and the alkyl electrophile to prevent off-cycle decomposition.In addition, the formation of the anionic species [Cu I (CF2H)2] -, which is known to react with alkyl electrophiles in a racemic fashion, 39 needs to be minimized.The pinnacle challenge, however, is to enable the enantioselective transfer of CF2H groups to alkyl radicals, an endeavor that had yet to be achieved. 40,41 n this work, we solved all the above challenges by employing an electron-deficient chiral diamine ligand (Fig. 1c).We disclose herein a new mode for enantioselective difluoromethylation through alkyl radical intermediates, enabling the high-yield synthesis of a diverse range of CF2H-containing molecules with excellent enantioselectivity.DFT calculations demonstrate key hydrogen bonding and π-π interactions between the chiral diamine ligand and substrates.

Results:
We first examined the difluoromethylation of α-haloamides given the prevalence of amide groups in bioactive molecules (see Table S1-3 for detailed optimization results).Our study revealed that the combination of copper salts and a commercially available chiral diamine ligand, L*, catalysed the difluoromethylation of racemic α-bromo-N-phenylbutanamide in 85% yield and with 95% e.e. (Fig. 2, Entry 1).A nucleophilic difluoromethyl zinc reagent, (DMPU)2Zn(CF2H)2, i.e.Vicic-Mikami reagent, 42,43 was essential to promote this reaction.The use of L * , which bears electrondeficient phenyl groups, was uniquely effective for this difluoromethylation reaction.Notably, the use of L * has not been reported in any reactions prior to this study.The sub-ambient temperature is necessary to achieve both high enantiocontrol and efficiency of the reactions, presumably by decelerating the decomposition pathways of the Cu I -CF2H intermediate.Additionally, using solvents with higher polarity significantly decreased the enantioselectivity of products presumably due to the partial formation of the anionic [Cu I (CF2H)2] -species.We next examined the generality of this asymmetric difluoromethylation reaction.α-bromoamides with simple alkyl substituents afforded the difluoromethylated products in good yield (79-85% yield), with enantioselectivity ranging from 90-95% e.e. (2-4).Alkyl bromides adjacent to secondary cyclic and acyclic groups were competent, affording the desired products in good yield and with high enantioselectivity (5-12, 65-79% yield, 90-99% e.e.).Notably, medicinally relevant heterocycles including piperidine (9), tetrahydropyran (10), and azetidine (13) were compatible with this protocol.In addition, a substrate that contained the bromide at a homobenzylic position, which was prone to β-hydrogen elimination, could afford its difluoromethylated product 14 in 66% yield and with 90% e.e.Various functional groups including ester (15), amide (16), carbamate (17-18), imide (19), and even unprotected alcohol (20) could be well tolerated (52-81% yield, 90-92% e.e.).

Synthetic application
We further showcased that this asymmetric difluoromethylation protocol could facilitate the efficient synthesis of CF2H analogs of molecules with pharmaceutical significance (Fig. 3).First, the diverse reactivity of amide groups allowed for the synthesis of different CF2H-containing molecules without affecting the chiral centers (Fig. 3A).For example, the reduction of compound 10 (98% e.e.) with borane provided the β-difluoromethylamine 56 in 90% yield with 97% e.e.In addition, the N-para-methoxy phenyl amide group in compound 57 (97% e.e.) could be removed via the cerium ammonium nitrate (CAN) oxidation to afford a primary amide 58 in 60% yield with 96% e.e.The scalability of this difluoromethylation protocol has also been demonstrated by the synthesis of compound 57 at a half-gram scale (60% yield, 97% e.e.).Moreover, recognizing the benefits of fluorinated analogs of nonsteroidal anti-inflammatory drugs (NSAIDs), we synthesized the difluoromethylated counterpart of (S)-ibuprofen (Fig. 3A).The CF2H amide 59 was formed in 87 % yield and with 87% e.e. from the corresponding benzylic chloride.A two-step approach allows for the conversion of 59 to the CF2H analog of (S)-ibuprofen (61, 87% e.e.) without compromising the enantioselectivity.
Given the widespread presence of amide functionalities in pharmaceuticals and agrochemicals, our approach offered an invaluable way for the rapid synthesis of their enantioenriched fluorinated bioisosteres (Fig. 3B).BMS-270394, a selective agonist for the human retinoic acid receptor (hRAR), possesses a hydroxyl moiety that binds to the methionine sulfur atom within the protein's active site. 44,45 his hydroxyl group plays a pivotal role in the activity of BMS-270394, while its enantiomer remains inactive.Recognizing the potential of CF2H groups as bioisosteres of OH groups, we successfully transformed the benzyl chloride precursor (62) into the difluoromethylated analog of BMS-270394 (63) in 72% yield and with 92% e.e.Moreover, given the escalating relevance of fluorinated herbicides and the increasing frequency of chiral centers in these molecules, 46 we applied this asymmetric difluoromethylation protocol to the synthesis of a chiral CF2H analog of pentanochlor, a pre-and post-emergence herbicide.Thus, the treatment of the alkyl bromide 64 under the standard difluoromethylation conditions allowed for the synthesis of 65 in 83% yield and with 94% e.e.Furthermore, this protocol enabled the synthesis of a precursor to the CF2H analog of acebutolol, 47 a beta-blocker for the treatment of high blood pressure (67, 60% yield, d.r.= 93.5 : 6.5).Notably, the tolerance of a reactive epoxide group further demonstrated the mild conditions of this Cu-catalysed protocol.
We have extended this protocol to the late-stage difluoromethylation of natural products (Fig. 3C).A facile α-bromination followed by the amide bond formation of oleic acid affords the corresponding alkyl bromide (68), which could be converted to its difluoromethylated product (69) with high enantioselectivity (73% yield, 95% e.e.).Similarly, two steroid derivatives, cholic acid and lithocholic acid (in their acetate forms), were converted to their difluoromethylated analogs with excellent control of stereochemistry (70-73, 63-69% yield, d.r.> 20:1) The high diastereoselectivity with either enantiomer of the diamine catalyst L* further highlights the catalyst-controlled selectivity of this difluoromethylation protocol.
Finally, this late-stage asymmetric difluoromethylation reaction protocol could be applied to the functionalization of pharmaceutical agents (Fig. 3D).Thus, a chemotherapy medication chlorambucil 48 and a γ-secretase inhibitor MK-0752 49 could be readily converted to their corresponding alkyl bromides (74 and 76).Difluoromethylation of these two alkyl bromides under the standard conditions afforded their CF2H products (75 and 77) in good yield and with high enantioselectivity (48 and 66% yield, respectively, 95% e.e.).

Mechanistic studies
These results shown herein represent a rare example of highly enantioconvergent fluoroalkylation of alkyl electrophiles.Detailed mechanistic insights into these catalytic reactions should inspire the future development of asymmetric fluoroalkylation reactions.The radical nature of this protocol was confirmed by the difluoromethylation of a cyclopropyl-containing substrate (78), which afforded the only ring-opening products (79 and 80), whereas the unrearranged product (81) was not observed (Fig. 4a).In addition, we employed a radical trap DMPO (5,5-dimethyl-1pyrroline N-oxide) to trap the transient alkyl radicals formed in the reaction mixture.The addition of DMPO to the reaction mixtures containing [Cu(MeCN)4]PF6, L*, and an alkyl bromide 82, with or without (DMPU)2Zn(CF2H)2, consistently revealed the formation of the trapping product 83, which was detected by electron paramagnetic resonance (EPR) spectroscopy.This observation further suggests that the reaction mechanism involves a SET step from the copper(I) species, leading to the generation of an alkyl radical (Fig. 4a).EPR simulations indicated hyperfine coupling to both nitrogen and hydrogen nuclei, with hyperfine constants AN = 6.5 G, 6.5 G, 34 G and AH = 23 G, 21 G, 23 G. Notably, the average hyperfine splitting values at low temperature align closely with those previously reported for DMPO radicals at room temperature. 24 further applied EPR spectroscopy to investigate the potential formation of Cu II intermediates through SET (Fig. S2).A solution of diamagnetic [Cu(CH3CN)4]PF6 with the chiral diamine ligand L* exhibited a negligible EPR signal from the cavity.The addition of an alkyl bromide to this solution led to the emergence of an EPR signal with hyperfine splitting characteristics of a Cu II species, implying the formation of a Cu II intermediate in the reaction pathway.In analogous experiments, we introduced (DMPU)2Zn(CF2H)2 to the reaction mixture.Interestingly, the resulting EPR spectra were silent, suggesting the rapid reaction between the transient alkyl radicals with the paramagnetic [Cu II -CF2H] species.
To shed light on the enantioconvergent process of this reaction, we conducted difluoromethylation of an enantiomerically pure alkyl bromide 84 (Fig. 4b).Under standard reaction conditions, both enantiomers were converted to the same difluoromethylated product 2 with an identical e.e. value and in similar yield.Moreover, analysis of the unreacted alkyl bromides throughout the reaction revealed that no racemization occurred during the difluoromethylation reactions.These results are consistent with a stereoablative enantioconvergent process 50 rather than a simple kinetic resolution or a dynamic kinetic resolution.
Additionally, a linear correlation was observed between the catalyst and product e.e. value, suggesting a 1:1 copper-to-ligand ratio in the enantio-discrimination complex (Fig. S3).X-ray crystallographic studies of L*Cu II (OAc)2, synthesized by mixing L* with Cu(OAc)2, validated the bidentate binding of the diamine catalyst with copper center (CCDC-2303333, Fig. S4).These results supported the involvement of a mononuclear copper species coordinated with a single chiral ligand as the active intermediate in the reaction.Based upon these experimental observations, our current hypothesis for the mechanism of this Cucatalysed enantioselective difluoromethylation reaction is shown in Fig. 4c.Diamine-bound copper(I) complex I undergoes transmetallation with the zinc-difluoromethyl reagent to provide copper(I)-CF2H complex II.The reaction between complex II and the alkyl electrophile generates copper(II) complex III and an organic radical IV.The radical then recombines with the complex III to form an alkyl-copper(III)-CF2H species V, 51,52 which reductively eliminates to form the difluoromethylated product and regenerate the copper(I) catalyst. 53,54

Theoretical calculations
We next performed Density Functional Theory (DFT) calculations to understand the origin of enantioselectivity in this difluoromethylation process (Fig. 5).Previous work on copper-catalysed asymmetric functionalization of alkyl radicals indicated that the group transfer from Cu II intermediates to alkyl radicals is the enantio-determining step. 55Therefore, we focused our DFT studies on the reaction between the Cu II intermediate [L * Cu II (CF2H)Br] and the alkyl radical IV (Fig. 5a).The addition to the Re face of the alkyl radical IV, which proceeded with a 1.9 kcal/mol free energy barrier at transition state TSR, led to the formation of a Cu III intermediate Int-R.In contrast, the addition from the Si face of IV occurred with a higher free energy barrier of 4.4 kcal/mol at transition state TSs, resulting in the formation of the Cu III intermediate Int-S.Both steps were exergonic, with the Int-R being 3.4 kcal/mol more stable (∆G = -10.7 kcal/mol) than its diastereomer Int-S (∆G = -7.3kcal/mol).The C-CF2H bond formation was found to proceed via concerted reductive elimination from Cu III intermediates to afford either enantiomer of the difluoromethylated product.The transition state that formed the (R) product, TS-RER, was associated with a lower free energy (∆G ‡ = 3.2 kcal/mol) than its (S)-counterpart, TS-REs (∆G ‡ = 6.0 kcal/mol), with a difference in free energy of ∆∆G ‡ = 2.8 kcal/mol.These computation results demonstrate that both the radical substitution and the reductive elimination steps favored the formation of the (R)-product, aligning well with the experimental results.
We also considered the possibility of the involvement of a bisdifluoromethyl Cu II intermediate, L * Cu II (CF2H)2 (Fig. S5).Although low barriers were found to form [L * Cu III (CF2H)2(alkyl)] intermediates (∆G ‡ < 4 kcal/mol), the reductive elimination of resultant Cu III intermediates was accompanied by high free energy barriers (∆G ‡ > 29 kcal/mol).Such high energies are unreasonable for low-temperature (-40 °C) reactions.Additionally, an alternative radical substitution pathway where the alkyl radical directly attacks the CF2H group on the Cu II center, without the involvement of a Cu III intermediate, was ruled out due to its high activation energy (∆G ‡ = 23.2kcal/mol, Fig. S6).bis-trifluoromethyl phenyl groups of the ligand occupied the pseudo equatorial positions, consistent with the crystal structure of L * Cu II (OAc)2 (Fig. S7).This conformation of the phenyl groups is unlikely to induce significant steric repulsion with the alkyl radical intermediate.On the other hand, the transition state structures for radical combination and reductive elimination all exhibited favorable interactions between the ligand and the substrate-derived radical (Fig. 5a).Interaction Region Indicator analysis 56 revealed the hydrogen bond interactions between the carbonyl group of the substrate and the NH bond of the diamine ligand (Fig. 5b).Furthermore, an edge-to-face π-π interaction was observed between the electron-rich phenyl groups in substrates and the electron-deficient phenyl groups in diamine catalysts.These non-covalent interactions dictated the conformation of the substrate radical during its approach to the Cu II intermediate and the structure of the resultant Cu III intermediates.Notably, the transition state that led to the S product (TS-REs) was destabilized due to the steric repulsion between the methyl group on the substrate and the N-methyl group on the ligand, along with the methyl group and N-H group on the substrate.Consequently, (R)-2 was formed as the major product, consistent with the experimental results.This stereochemical model aligned well with the experimental observations.For example, using diamine ligands with either mono-trifluoromethyl or non-substituted phenyl groups resulted in significantly reduced enantioselectivity (Table S4), likely due to weaker hydrogen-bonding and π-π interactions.Additionally, the limitation with a primary amide (Extended Data Fig. 1, 85) and N-alkyl secondary amides (86 and 87) could be attributed to the lack of π-π interactions.Nonetheless, these limitations with N-alkyl amides could be overcome through a two-step synthesis strategy, involving the deprotection of a difluoromethylated N-aryl-N-alkyl amide 88.

Conclusion:
Overall, we report herein a Cu-catalysed approach for highly enantioselective difluoromethylation, enabling the synthesis of a broad spectrum of enantioenriched molecules containing the CF2H group.This reaction highlights the benefit of using nucleophilic fluoroalkyl reagents for the enantioselective functionalization of alkyl radical intermediates.More importantly, the successful realization of asymmetric difluoromethylation of alkyl radicals opens a new avenue for constructing stereochemical centers that contain CF2H and other fluoroalkyl groups.Finally, the manipulation of the hydrogen bonding and π-π interactions of this chiral amine ligand holds promise for inspiring the development of new asymmetric transformations.

Methods:
General procedure for copper-catalysed enantioconvergent difluoromethylation of alkyl halides: In a glovebox filled with argon, an oven-dried 4 mL vial equipped with a stir bar was charged with [Cu(CH3CN)4]PF6 (7.5 mg, 0.02 mmol, 10 mol%), ligand L* (20.0 mg, 0.04 mmol, 20 mol%) and anhydrous isopropyl ether (1.2 mL).The mixture was stirred at room temperature for 15 minutes, then taken out of the glovebox and cooled to -40 °C.Subsequently, a solution of the alkyl halide (0.2 mmol, 1.0 equiv.)and (DMPU)2Zn(CF2H)2 (88 mg, 0.2 mmol, 1.0 equiv.) in DMPU (0.3 mL) was added dropwise to the reaction mixture using a syringe.After stirring for 36 h at this temperature, the reaction mixture was allowed to warm to room temperature.The difluoromethylated product was purified through column chromatography on silica gel using hexanes/EtOAc as the mobile phase.The enantiomeric excesses of the products were determined by HPLC, using AD-H, OD-H, OJ-H, IA, or IB columns.

Fig. 1
Fig. 1 Development of Cu-catalysed enantioconvergent difluoromethylation of alkyl halides.a. Importance of CF2H-bearing chiral centers in the pharmaceutical industry.b. nucleophilic enantioconvergent fluoroalkylation as a

Fig. 3
Fig. 3 Synthetic applications of enantioconvergent difluoromethylation reaction.A. synthetic transformation of enantiopure CF2H products; B. synthesis of analogs of pharmaceuticals and agrochemicals; C. late-stage asymmetric difluoromethylation of natural products; D. late-stage difluoromethylation of medicinal agents

Fig. 4
Fig. 4 Mechanistic studies and proposed catalytic cycle.a. Radical clock and DMPO trapping experiments support the involvement of alkyl radical intermediates.b.Difluoromethylation of enantiomerically pure alkyl bromides supports a stereoablative enantioconvergent process.c.Proposed catalytic cycle for the enantioconvergent difluoromethylation reaction.

Fig. 5
Fig. 5 DFT calculations on the Cu-catalysed enantioconvergent difluoromethylation reaction at the B3LYP-D3BJ/def2-TZVP//B3LYP-D3BJ/def2-SVP level of theory.a. Free energy profile of the reactions between the Cu II intermediate with the alkyl radical.b. origin of the enantioselectivity.To further elucidate the origin of the enantioselectivity, we examined the structures of the transition states in the C-CF2H bond-forming step.Interestingly, DFT calculations revealed that