C−F bond activation enables synthesis of aryl difluoromethyl bicyclopentanes as benzophenone-type bioisosteres

Bioisosteric design has become an essential approach in the development of drug molecules. Recent advancements in synthetic methodologies have enabled the rapid adoption of this strategy into drug discovery programs. Consequently, conceptionally innovative practices would be appreciated by the medicinal chemistry community. Here we report an expeditous synthetic method for synthesizing aryl difluoromethyl bicyclopentane (ADB) as a bioisostere of the benzophenone core. This approach involves the merger of light-driven C−F bond activation and strain-release chemistry under the catalysis of a newly designed N-anionic-based organic photocatalyst. This defluorinative coupling methodology enables the direct conversion of a wide variety of commercially available trifluoromethylaromatic C−F bonds (more than 70 examples) into the corresponding difluoromethyl bicyclo[1.1.1]pentanes (BCP) arenes/difluoromethyl BCP boronates in a single step. The strategy can also be applied to [3.1.1]and [4.1.1]propellane systems, providing access to analogues with different geometries. Moreover, we have successfully used this protocol to rapidly prepare ADB-substituted analogues of the bioactive molecule Adiporon. Biological testing has shown that the ADB scaffold has the potential to enhance the pharmacological properties of benzophenone-type drug candidates.

][3][4][5][6][7][8][9][10][11] Consequently, the design of new bioisosteres has been adopted as a creative and effective manner to increase potential for developing lead compounds and creating new drugs.Over the last few decades, one tactic that has been explored to improve metabolic stability while maintaining bioactivity is the substitution of the ketone functional group with a difluoromethylene moiety. 6In addition, sp 3 -hybridized small-ring cage hydrocarbons, such as bicyclo-[1.1.1]pentane][9][10][11][12][13] While each of these bioisosteres has been extensively studied as a sole functional group isostere in drug development, their merger for the synthesis of difluoromethyl BCP arene as a new surrogate of the benzoyl group and the subsequent evaluation of their pharmacokinetic properties remain unexplored (Fig. 1b). 14,15][18][19] Incorporating the ADB moiety as a new bioisostere would create significant opportunities to access a unique chemical space for benzophenone-type drug design.As such, new reaction designs for the rapid installation of the ADB would be of considerable value in developing novel pharmaceuticals.
Several challenges must be addressed to achieve such a general defluorinative coupling transformation: (1) tuning the photocatalytic ability to enable the reductive activation of strong C−F bonds, regardless of their electronic substitution, despite typically high reductive potential; [36][37][38] (2) preventing overfunctionalization of the resulting difluoromethyl product as the strength of C−F bond decreases during defluorination; 37 (3) selectively trapping the electronphilic difluorobenzylic radicals with propellane before being quenched by the hydrogen atom donor; [37][38][39] (4) controlling deleterious propellane oligomerization. 51Herein, we report our successful efforts to develop an expedient route to ADB scaffold synthesis from readily accessible trifluoromethylarenes. 52A broad scope of trifluoromethylarenes coupling with [n.1.1]propellane systems has been demonstrated through two-component and threecomponent coupling (Fig. 1c).This method allowed for the rapid preparation of ADB analogues of known drugs, one of which is found to be more metabolically stable than its commercial progenitor.We anticipate this strategy would serve as a valuable tool for the synthesis and evaluation of ADB motifs as new bioisosteres of benzophenone-type drug derivatives, ultimately leading to the development of novel pharmaceuticals.

Reaction optimization
Our investigation into this new defluoronative coupling reaction began with exposure of ArCF3 5 to various photocatalysts in the presence of [1.1.1]propellaneand γterpinene as hydro gen atom donor (Table 1).To our delight, the desired product, aryl difluoromethyl bicyclopetane 6, was observed in 36% yield when N-phenylphenothiazine (PTH) was used as the photocatalyst, along with a ArCF2-diBCP sideproduct 7, that arose from propellane dimerization.The less reducing photocatalyst fac-Ir(ppy)3 led to a decreased yield of ArCF2-BCP 6 (6% yield).At this stage, we surmised that selecting the appropriate photocatalyst would be key to improving reaction efficiency.Indeed, a highly reducing ophosphinophenolate (PO) photocatalyst was more efficient than PTH to increase the yield to 43%. 45 Based on previous studies on the effect of phosphine group facilitating photo-induced electron transfer processes, 45 we hypothesized that an ortho-PPh2 substituent on arylamine might have similar effect to enable a redox cycle of strong reductive ability. 53,54To this end, we designed and synthesized different types of ophosphinoarylamines (PNs) as strongly reductive photocatalysts.Next, we evaluated their effectiveness in catalyzing the defluoroalkylation of 5, a selection of which is shown in Table 1(see Supplementary Figure S13 for a list of all evaluated photocatalysts).The best results were obtained with carbazole-based catalyst PCN (1), a highly reducing photocatalyst ( E 1/2 red [ PN − * /PN • ] = -3.26V vs. the saturated calomel electrode (SCE) in DMSO) (see Supplementary Information for measurement), which provided the desired product 6 in 74% yield (entry 1).Structurally related PCN (2) also catalyzed the defluoroalkylation but with diminished yield (55% yield).We considered that the accelerated intersystem crossing process to access triplet state of long lifetime and steric protection by introducing the second adjacent -PPh2 account for the improved catalyst efficiency.Interestingly, 2-(diphenylphosphino)biphenyl amine PBN (3), a potential ligand for transition metal catalysis, was also effective to catalyze the reaction giving 48% yield.In contrast, the structurally similar phenyl-protected secondary amine PBN (4) was less effective (12% yield).Unfortunately, dimethylprotected tertiary amine diMe-PBN (I) and biphenyl amine BN (II) without ortho-PPh2 substitution, proved to be ineffective at all.These results suggest that both of the N-anion and ortho-PPh2 substituent are crucial for the photocatalytic reactivity.Increasing the amount of [1.1.1]propellaneto 3.0 or 5.0 equiv.resulted in a significant decrease in reaction efficiencies due to the unproductive propellane dimerization leading to dimer 7(entries 2−3).Finally, control experiments revealed that the photocatalyst, base and visible light were all essential for the success of this transformation (entries 5−7).The reaction proceeded poorly in the absence of γ-terpinene (entry 4).It is worth noting that catalyst PCN (1) is an air-stable white solid that can be prepared on gram-scale and stored under ambient condition for months without observing any decomposition.
A plausible mechanism for the proposed defluoroalkylation of ArCF3 with [1.1.1]propellaneis shown in Fig. 2. Visible-light photoexcitation of electron-rich N-anionic catalyst I would generate a highly reducing excited state II.Single-electron transfer to the trifluoromethylarene 5 ( E 1/2 red = -2.16V vs. SCE in DMSO) (see Supplementary Information for measurement) gives rise to the corresponding difluorobenzylic radical 8, which should be rapidly intercepted by [1.1.1]propellaneto form the resulting BCP radical 9 55 .The electrophilic radical 9 would then abstract a hydrogen atom from γ-terpinene, thus enabling the desired ADB product 6 and the cyclohexadienyl radical 10. 56 Finally, single electron transfer between 10 and oxidized photocatalyst III would regenerate photocatalyst I, thus completing the catalytic cycle.

Reaction substrate scope
We next evaluated the generality of this defluorinative coupling reaction.As shown in Table 2, a myriad of unactivated trifluoromethylaromatics could be engaged under these conditions.A diverse range trifluoromethylarenes bearing electron-donating and electron-withdrawing substituents were found to be competent substrates (6, 12-29, 30-74% yield).Notably, functional groups such as amides, ethers, anilines, unprotected amines and alcohols, were all well compatible, allowing for practical handles for further derivatization.Given the importance of nitrogen-containing heterocycles in the production of bioactive molecules, we were delighted to find that a diverse array of pyridine substrates with trifluoromethyl groups at 2-, 3-, and 4-positions were readily accommodated (30−35, 50−85% yield).Moreover, this reactivity could be extended to bicyclic heteroaromatics affording good yields (36−39, 51−72% yield).Particularly, perfluoroalkylarene was readily employed for selectively defluorinative coupling with [1.1.1]propellane(40, 61% yield).Additionally, bis(trifluoromethyl) benzenes yielded selective activation of only one of the C−F bonds giving synthetically useful yields (41−48, 32−71% yield).Finally, the compatibility of the defluoroalkylation protocol with biorelevant molecules was also examined.To our delight, a large variety of biologically relevant systems were successfully employed to install the ADB moiety (49−54, 32−73% yield).It is worth noting that pharmaceutical analogs, including fluoxetine, cinacalcet and triflupromazine, were effective substrate for the late-stage C−F functionalization, delivering the corresponding ADB adducts smoothly (52−54, 32−73% yield).These results further highlight the real-world utility of this new defluorinative coupling technology.
The installation of bicyclo[3.1.1]heptane(BCH) scaffold as a novel bioisostere of meta-substituted benzenes in bioactive molecules has become an increasingly popular research area. 57  Pioneer works by Anderson 58 and Uchiyama 59 group have established a practical protocol for accessing functionalized BCHs employing a radical-based approach. 60,61Gratifyingly, our protocol for this defluorinative coupling can be extended to [3.1.1]heptanesystem (Table 3).An array of trifluoromethylaromatics, including pharmaceutical agent, can be utilized to afford ArCF2−BCH products in moderate to good yields (55−59, 35−80% yield).Furthermore, the less-strained [4.1.1]octanesystem can also be employed in the same way to generate ArCF2−BCO products (60−64, 21−62% yield). 62To our knowledge, this is the first instance of the [4.1.1]octanesystem engaging in a photocatalytic radical addition process to date.

Preparation of Pharmaceutical Analogues
To demonstrate the potential of these ArCF2−BCP cores as bioisosteres for benzophenone-type drugs and the applicability of our procedures in drug discovery settings, we conducted defluoronative coupling reactions to the rapid synthesis of ADB-substituted analogues of Adiporon, an established adiponectin receptor agonist. 68,69As show in Fig. 3, a two-step sequence employing condensation and photocatalytic defluorinative coupling is applied to produce three structurally distinct bioisosteres of Adiporon (79 [1.1.1]BCP−Adiporon,80 [3.1.1]BCH−Adiporonand 81 [4.1.1]BCO−Adiporon).We then tested these ADB pharmaceutical analogues in comparison to their benzophenone-containing counterparts.Interestingly, [1.1.1]BCPsubstituted analog 79 was found to be metabolically stable, with reduced clearance rates in human liver microsomes, although its membrane permeability (Caco-2) was slightly decreased compared to its parent drug.These findings underline the potential of the ADB scaffold as a beneficial motif for enhancing the pharmacological properties of drug candidates containing a benzophenone core.

Conclusion
In summary, we describe herein that C−F bond functionalization can be merged for the first time with strainrelease coupling for the expeditous synthesis of difluoromethyl BCP arenes/difluoromethyl BCP boronates.A diverse array of trifluoromethylaromatics can be employed to couple with [1.1.1]propellaneas well as [3.1.1]heptaneand [4.1.1]octanestrain ring systems delivering a diversifiable ArCF2−BCP synthetic linchpin.We demonstrate that this transformation can be applied to the synthesis of complex drug substrates, highlighting its potential application in drug discovery.Biological testing of ADB analogues of known bioactive compound showed their value to the pharmaceutical industry as a replacement for benzophenone core.

Fig 1 .
Fig 1. Development of a C-F bond activation strategy for the synthesis of benzophenone-type bioisosteres.a, Representative benzophenone core in drug design.b, Replacement of benzophenone core with aryl difluoromethyl bicyclopentane scaffold.c, C-F bond activation enables synthesis of aryl difluoromethyl bicycloalkanes (this work).BCP, bicyclopentane; PN, N-anionic-based photocatalyst; H, hydrogen atom donor; B, borylation reagent.