Bicyclopentylation of Alcohols with Thianthrenium Reagents

Herein we present the first method for the synthesis of bicyclo[1.1.1]pentyl (BCP) alkyl ethers from alcohols. The reaction uses BCP–thianthrenium reagents and is catalyzed by a dual copper/photoredox catalyst system. Unlike known alkylations of tertiary alcohols via carbocation intermediates, our Cu-mediated radical process circumvents the labile BCP carbocations. The approach demonstrates a broad tolerance for functional groups when applied to primary, secondary, and even tertiary alcohols. In addition, we highlight the utility of this method in late-stage functionalizations of both natural products and pharmaceuticals as well as in the rapid construction of BCP analogs of known pharmaceuticals that would otherwise be difficult to access.

A pproximately 45% of marketed small-molecule pharma- ceuticals contain phenyl rings. 1 In modern medicinal chemistry, replacement of planar phenyl rings with sp 3 -rich bioisosteres can lead to increased metabolic stability, membrane permeability, and increased solubility. 2 Since the first study of the BCP analogue of (S)-(4-carboxyphenyl)glycine was described by Pellicciari and co-workers in 1996, 3 the rigid three-dimensional 1,3-disubstituted BCPs have emerged as promising bioisosteres for para-substituted benzenes in drug development that maintain the exit vectors in a 180°dihedral angle. 4Over the last 15 years, numerous efforts have been devoted to the development of substituted BCPs, such as BCP amines, 5 BCP arenes, 5h,6 BCP alkanes, 6f,7 and BCP aryl ethers 5h via the functionalization of [1.1.1]propellane 8or cross-coupling reactions of BCP-based reagents.5h,6c,d,7b,9 However, alkylation of alcohols with BCP scaf folds has remained elusive.Such a new method would enable access to BCP analogs of alkyl aryl ethers that are prevalent in pharmaceuticals, such as in the selective serotonin reuptake inhibitor fluoxetine and the Parkinson's medication safinamide.
Here we report the first bicyclopentylation of alcohols using BCP−thianthrenium (TT + ) reagents.Distinct from wellestablished alkylation of alcohols with tertiary electrophiles such as tert-butyl bromide via carbocation intermediates, 10 our method proceeds through a metal-mediated radical process that can bypass the unstable BCP cations.The favorable high reduction potential and rapid mesolytic cleavage rate of BCP− TT + allow the reaction to proceed under mild conditions with a wide variety of functional groups present and alcohols used as limiting reagents.Mechanistic studies imply that TT may potentially act as an electron-transfer mediator between the photoredox catalyst and copper catalyst.The substituted BCP−TT + reagents can be used as structural linkers or end groups, which allows for rapid preparation of promising BCP analogs of pharmaceutically relevant ethers.To the best of our knowledge, this study represents the first example of directinggroup-free transition-metal (TM)-catalyzed alkylation of tertiary alcohols that does not involve classic carbocation intermediates.
Alkyl aryl ethers are important structural motifs in pharmaceutically active compounds, 11 such as in fluoxetine, 12 butoxycaine, 13 safinamide, 14 and pranlukast. 15The current approach to BCP alkyl ethers involves the alkylation of a hydroxyl-substituted BCP with reactive alkylating reagents.Therefore, the approach is limited to reactive primary electrophiles, such as benzyl bromide, in the presence of a strong base for a conventional Williamson ether synthesis (Figure 1A, left).6c Alternatively, a conceptually distinct approach is the bicyclopentylation of alcohols with BCPbased reagents, which remains challenging due to the sluggish development of syntheses of tertiary alkyl ethers (Figure 1A, right).The Baran 16a and Ohmiya 16b groups reported elegant syntheses of hindered dialkyl ethers from tertiary alkyl carboxylic acids and alcohols via electrochemistry and organophotoredox catalysis, respectively.Key to the success lies in the generation of tertiary carbocations or carbocationlike intermediates under nonacidic conditions, which can be subsequently trapped by alcohol nucleophiles.However, the BCP carboxylic acid, as a bridgehead 3°carbocation precursor, does not afford the corresponding product under these conditions (Figure 1B), 16a possibly because of the low stability of BCP cation intermediates (energy barrier to ring opening: ∼19 kcal/mol). 17The use of transition metals to catalyze the bicyclopentylation of alcohols involving kinetically stable BCP radical intermediates (energy barrier to ring opening: ∼26 kcal/mol), 18 may avoid the skeletal rearrangement of the BCP intermediates.To date, BCP metals, 6b BCP halides, 6d BCP boronates, 6c,7b,9 BCP amines, 5a−d and BCP carboxylate derivatives 6a,e are the most used reagents for the construction of BCPs.However, none of them have been reported for the functionalization of alcohols to construct ether bonds.Considering the potential of BCP alkyl ethers in drug discovery, we sought to develop the first TM-catalyzed bicyclopentylation of alcohols using BCP−TT + reagents (Figure 1C).
We have previously demonstrated that the cationic BCP− TT + reagents can serve as BCP radical sources, which can engage in TM-mediated functionalization of phenols, various N-nucleophiles, and (het)aryl bromides.5h However, these conditions failed to deliver the desired product when alcohols were used as substrates due to the inherent reactivity differences between alcohols and the other, softer nucleophiles, like phenols and amines.For instance, metal alkoxides, crucial intermediates in many ether syntheses, exhibit both higher basicity and stronger reducing power compared to metal phenoxides, 19,20 and alcohols are less nucleophilic than many N-nucleophiles.Consequently, in many of the TM-catalyzed Based on our working hypothesis, we found that the reaction between 4-methoxyphenethyl alcohol (1a) and trifluoromethyl bicyclopentyl thianthrenium salt (CF 3 -BCP−TT + BF 4 − , 1b) occurred with Cu(acac) 2 as the catalyst, Na 2 CO 3 as the base, and Ir[dF(CF 3 )ppy] 2 (dtbbpy)PF 6 as the photocatalyst under 460 nm irradiation to give 1 in 85% yield (Figure 2A).The use of Cu(acac) 2 was crucial for the high yield, possibly because the β-diketonate ligand may facilitate the oxidative ligation of BCP radical A or stabilize the high-valent metal alkoxides C (entries 1−4).5d,23a,24,25 While Cu(TMHD) 2 (TMHD = 2,2,6,6-tetramethyl-3,5-heptanedionate) can afford higher yields, we only opted for this more expensive copper source when Cu(acac) 2 afforded less than 60% yield (entry 5).Control experiments showed that the copper catalyst, iridium photocatalyst, 3 Å molecular sieves, and blue LED light irradiation are necessary for efficient reaction (entries 9−15). 26he addition of 2.0 equiv of the radical trapping reagent 2,2,6,6,-tetramethylpiperidin-1-oxyl (TEMPO) suppressed the formation of 1, and a TEMPO−BCP radical adduct was observed (entry 16), consistent with the formation of BCP radicals.
While the scope for the alcohol is large, the scope of the BCP−TT + reagents is limited to a rather small subset.Two aspects are conceptually challenging to expand the diversity of these reagents, at least in the chemistry presented here.Based on the data, it appears most likely that the reaction proceeds through a radical chain transfer, in which after photochemical initiation a TT radical cation is transferred from the starting material to the product (Figure 4A).For chain initiation and propagation, homolytic cleavage of the C−S bond in CF 3 -TT + (31) is facile, which is the case for a rather stabilized CF 3 radical.However, other substituents such as the methyl substituent in Me-TT + (33) are bound more strongly, as supported by DFT calculations (Figure 4B).Second, other compounds that also feature a small bond dissociation energy (BDE) and would possibly engage in productive chain transfer could not be synthesized based on their lack of stability.For example, the difluoromethyl analogue of 31 (32; Figure 4 C) is unstable, presumably due to fast elimination from the cationic sulfonium salt.The cyano-substituted BCP compound 1c was synthesized through a different procedure 5h and displays a unique synthesis that could not be extended to other nucleophiles, presumably due to competitive redox chemistry with oxidation of nucleophiles other than cyanide by the thianthrenium radical cation. 30,33hile the scope of the BCP−TT + reagents is limited to only a few cases, we have highlighted the synthetic utility of our methodology to drug discovery with the few but relevant substituents in four BCP analogs of known pharmaceuticals (Figure 5A−D).The BCP analogue of fluoxetine hydrochloride, 36, was synthesized in only two steps and 85% overall yield from 1b and readily available starting material 35a.The cyano group of BCP reagent 1c can serve as a linchpin for the synthesis of other 1,3-disubstituted BCPs.For instance, cyano-BCP butyl ether (37) was obtained in 71% yield.Subsequently, hydrolysis of the cyano group to the corresponding carboxylic acid followed by alkylation afforded BCP-butoxycaine (38).Similarly, 41 underwent hydrolysis and amide condensation to give BCP-pranlukast (42).The cyano-BCP can also be selectively reduced to a BCP aldehyde, enabling reductive amination for the synthesis of BCP-safinamide (40).To the best of our knowledge, none of these analogs have been reported before.In addition, the cyano group in the BCP scaffold gives access to methylcarboxylate, methylamino, amide, and amino substituents (Figure 5E).
In conclusion, we have described an efficient bicyclopentylation of alcohols with BCP−thianthrenium reagents, providing a variety of BCP alkyl ether products, even at a late stage.Importantly, several bioisosteric replacements of aryl rings in small-molecule drugs were realized, which would otherwise be difficult to access currently by other methods.We anticipate that our approach can facilitate the development of saturated analogs of alkyl aryl ether drugs in the pharmaceutical industry.
Experimental procedures, spectroscopic data, and NMR spectra of all products (PDF) Accession Codes CCDC 2286412 contains the supplementary crystallographic data for this paper.These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing data_request@ccdc.cam.ac.uk, or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, U.K.; fax: +44 1223 336033.

Figure 2 .
Figure 2. Reaction development and a mechanistic investigation.a Yields were determined by 19 F NMR. b Isolated yield.c The TEMPO−BCP adduct was observed by HRMS.d The Stern−Volmer plot of Cu(acac) 2 was corrected due to the inner filter effect.

Figure 3 .
Figure 3. Substrate scope.a Yields were determined by 19 F NMR or 1 H NMR.
a plausible catalytic cycle is shown in Figure 2C.After quenching of the excited Ir(III) photoreodoxcatalyst, the ensuing Ir(II) species can reduce the BCP−TT + salt (Ir III / Ir II = −1.4V, E 1/2 = −1.4V for 1b, versus SCE in MeCN), 5h generating BCP radical III.The TT radical cation can oxidize the Cu(I) catalyst to generate Cu(II) (supported by EPR analysis; Figure S8), which then undergoes ligand exchange with the O-nucleophile.Oxidative ligation with intermediate III affords the Cu(III) complex IV.Finally, the desired product is formed via reductive elimination from IV.All experimental observables are consistent with the proposed mechanism. 32