Letter Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Ni-catalyzed Reductive Deaminative Arylation at sp3 Carbon Centers Raul Martin-Montero,†,‡ Veera Reddy Yatham,†,∥ Hongfei Yin,†,∥ Jacob Davies,† and Ruben Martin*,†,§ †
Downloaded via IDAHO STATE UNIV on March 29, 2019 at 16:08:29 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Institute of Chemical Research of Catalonia (ICIQ), The Barcelona Institute of Science and Technology, Av. Països Catalans 16, 43007 Tarragona, Spain ‡ Universitat Rovira i Virgili, Departament de Química Analítica i Química Orgànica, c/Marcel·lí Domingo, 1, 43007 Tarragona, Spain § ICREA, Passeig Lluís Companys, 23, 08010 Barcelona, Spain S Supporting Information *
ABSTRACT: A Ni-catalyzed reductive deaminative arylation at unactivated sp3 carbon centers is described. This operationally simple and user-friendly protocol exhibits excellent chemoselectivity profile and broad substrate scope, thus complementing existing metal-catalyzed cross-coupling reactions to forge sp3 C−C linkages. These virtues have been assessed in the context of late-stage functionalization, hence providing a strategic advantage to reliably generate structure diversity with amine-containing drugs.
T
Scheme 2. Forging sp3 C−C Bonds via sp3 C−N Cleavage
he ubiquity of sp3-hybridized carbon atoms in biologically active molecules has prompted chemists to develop catalytic late-stage sp3 C−C bond formations aimed at streamlining the preparation of new chemical entities in the early drug discovery stage.1,2 In this context, the prevalence of alkyl amines in pharmaceuticals and preclinical candidates makes them ideal counterparts to generate structural diversity via sp3 C−N cleavage (Scheme 1),3,4 holding promise to provide a strategic advantage when compared to conventional platforms based on alkyl (pseudo)halides in lead generation approaches.2,5 At present, catalytic methods for forming sp3 C−aryl architectures from alkyl amines have been developed with Watson providing an elegant approach using organometallic reagents (Scheme 2, path a)6 and Glorius contributing Minisci reactions with electron-poor heteroarenes.7 Recent reports have expanded the scope of deaminative functionalizations using Scheme 1. Prevalence of Alkyl Amines in Pharmaceuticals
pyridinium salts such as alkyl-Heck-type reactions,8 alkynylation/allylation,9 and C-heteroatom bond-forming reactions.10,11 Despite these contributions, the pursuit of synthetically enabling and user-friendly catalytic technologies that provide an orthogonal gateway to forge sp3 C−aryl bonds without recourse to stoichiometric organometallic reagents or biased heteroarenes represents a worthwhile endeavor for chemical invention. Under this premise, we anticipated that this approach might not Received: March 22, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.9b01016 Org. Lett. XXXX, XXX, XXX−XXX
Letter
Organic Letters
Scheme 3. Scope of Reductive Deaminative Arylationa,b
only excel in terms of chemoselectivity but also provide a new handle for late-stage applications with improved practicality and generality. As part of our studies in catalytic reductive couplings, 12 we report herein a Ni-catalyzed reductive deaminative arylation of pyridinium salts (I) derived from alkyl amines, with aryl halides (Scheme 2, path b).13,14 In preparative terms, this transformation is distinguished by its wide scope and excellent chemoselectivity profile. At the strategic level, these virtues have been assessed in late-stage functionalization,15 hence providing new tactics for lead generation in drug discovery.5 Preliminary experiments suggest a mechanistically distinct pathway from conventional Nicatalyzed cross-electrophile couplings16 or related redox-neutral C−N scissions (path a).6 This observation advocates the notion that, in our case, Ni catalysts are not responsible for generating open-shell species II via single-electron transfer (SET) processes, thus paving the way for designing novel bondforming technologies (path b). Our study began by evaluating the reaction of 2a with pyridinium salt 1a, which is readily accessible in bulk quantities from the corresponding alkyl amine in a single step (Table 1). Table 1. Optimization of the Reaction Conditionsa
As Table 1 (entry 1), 0.20 mmol scale, rt to 60 °C. bIsolated yields, average of two runs. cAt rt. dUsing ArI. eDMA at 15 °C. fAt 45 °C. g 5.44 mmol scale. hNaI (2.0 equiv). a
corresponding aryl bromide was largely inconsequential for the reaction outcome (3−16). While ortho-substituted (12, 21) or electron-rich (4, 5) aryl halides could be employed as substrates, low yields were found in the former for particularly hindered combinations, whereas the latter was better accomplished with aryl iodides.17 Although extensions to secondary acyclic alkyl counterparts are often plagued by parasitic β-hydride elimination or homodimerization,2 this was not the case, and 17−18 were obtained in good yields, not even traces of alkene sideproducts were identified in the crude mixtures. Similarly, primary alkylpyridiniums were expected to be problematic due to the inherent reluctance to generate primary alkyl radicals via SET; however, a subtle electronic modification with methoxy groups at the para position19 enabled the coupling of primary alkyl congeners with similar ease (19−33), including nitrogencontaining heterocycles (24, 25, 31−33). The chemoselectivity of this reaction was illustrated by the fact that esters (3, 16, 26, 27), ketones (7, 18), nitriles (12, 22), carbamates (14), alkenes (23, 28), alkynes (5), acetals (19−25, 30), sulfonamides (29), or unprotected phenols (11) could be well-accommodated. Vinyl bromides could also be utilized, albeit in lower yields (30). Most noteworthy was the compatibility with aldehydes (6, 13, 14, 15, 17) that would otherwise not be tolerated with classical organometallic approaches,6 even on a gram scale (13), constituting an additional bonus from a both synthetic and
a 1a (0.14 mmol), 2a (0.10 mmol), NiBr2·glyme (10 mol %), L3 (14 mol %), Mn (0.15 mmol), NMP (0.1 M) at rt. bGC yields using decane as internal standard. cIsolated yield.
After evaluation of all reaction parameters,17 we found that a combination of NiBr2·glyme, L3, and Mn in NMP at rt provided the best results, giving rise to 3 in 90% isolated yield. A comparison of entries 1−6 reveals that the nature of the ligand played a crucial role, with bipyridine ligands possessing donor substituents at 4,4′-position (L3) providing the best interplay between electronic and steric effects. The use of Ni(cod)2 as precatalyst was neither necessary nor beneficial (entry 7). The choice of the solvent, concentration, stoichiometry, and reducing agent markedly influenced the reaction outcome, with NMP (0.1 M) and Mn affording the best results (entries 8− 10).17 While similar results were found for 2a-I, the use of 2a-Cl resulted in traces of 3 (entry 11). Control experiments indicated that all the reaction parameters were essential for the reaction to occur (entry 12).18 Encouraged by these results, we turned our attention to examining the generality of our reductive deamination event (Scheme 3). As shown, the electronic and steric nature of the B
DOI: 10.1021/acs.orglett.9b01016 Org. Lett. XXXX, XXX, XXX−XXX
Letter
Organic Letters practical standpoint. Furthermore, the ability to tolerate organoboron groups (20, 28, 29) or aryl halides (8, 11, 19) provided a handle for further functionalization via crosscoupling reactions.20,21 The successful preparation of 34−39 further illustrates the virtues of our protocol for generating structural diversity in latestage applications (Scheme 4). Indeed, a variety of biologically
Scheme 5. Mechanistic Experiments
Scheme 4. Late-Stage Functionalizationa
AgCl) rather than to Ni-1 (E1/2 [NiII/NiI] = −1.37 V vs Ag/ AgCl).17 An otherwise similar scenario can be rationalized for both 1m (E1/2 = −0.90 vs Ag/AgCl) or in our photochemical regime based on Ir-1 (E1/2 [IrIII/IrII] = −1.32 V vs Ag/AgCl).30 These interpretations gain credence by applying a constant potential of −0.8 V in the reaction of Ni-1 with 1a, observing clean formation of 9 via selective SET to 1a.17 Although a comprehensive mechanistic coverage might await further investigations, one may conclude that sp3 C−C bond-formation occurs by intercepting alkyl radicals II with Ni-1 followed by final reductive elimination.31 In conclusion, we have developed a mild, modular, and exceedingly tolerant platform for forging sp3 C−C bonds via catalytic reductive deamination of alkyl amines. These virtues bear strategic relevance in late-stage functionalization, thus opening a gateway to structural diversity in lead generation approaches. Preliminary mechanistic studies reveal a mode of action that might expand the toolbox of available reductive events via C−N scission. Further applications along these lines are underway.
As in Table 1 (entry 1), 60 °C for 48 h. bUsing ArI.
a
active compounds containing amine groups such as leelamine, histamine, primaquine, mexiletine, or mosapride could all partake in reductive deaminative arylations in good yields (34−38). Aimed at extending the versatility of our reaction, we surmised that the propensity of Ir-photoredox catalysts for noninvasive outer-sphere SET might pave the way for tackling challenging combinations under blue-LED irradiation while avoiding stoichiometric Mn.22,23 This assumption proved correct in that no conversion to 39 was found with a Ni/L3 protocol based on Mn,24 whereas a dual photoredox/Ni catalytic regime with Hantzsch ester as terminal reductant cleanly furnished the targeted product in 58% yield. At present, we have no explanation for such striking difference in reactivity. While the generally accepted rationale of Ni-catalyzed crosscoupling reactions suggests a pathway consisting of oxidative addition of sp3 C−N bonds to Ni(0)Ln or alkyl radical species II (Scheme 2) generated via SET from either Ni(0)Ln or Ni(I)Ln,6,16,20 we have gathered evidence that supports an alternative scenario (Scheme 5).25 First, negligible conversion of 1a (
