Bisphosphines: A Prominent Ancillary Ligand Class for Application in

Jun 20, 2018 - The Ni-catalyzed Csp2–N cross-coupling of NH substrates and (hetero)aryl (pseudo)halides for the synthesis of (hetero)anilines is in ...
0 downloads 0 Views 5MB Size
Download PDF
Perspective Cite This: ACS Catal. 2018, 8, 7228−7250

pubs.acs.org/acscatalysis

Bisphosphines: A Prominent Ancillary Ligand Class for Application in Nickel-Catalyzed C−N Cross-Coupling Christopher M. Lavoie and Mark Stradiotto*

Downloaded via STEPHEN F AUSTIN STATE UNIV on July 15, 2018 at 09:46:31 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Department of Chemistry, Dalhousie University, 6274 Coburg Road, P.O. Box 15000, Halifax, Nova Scotia B3H 4R2, Canada ABSTRACT: The Ni-catalyzed Csp2−N cross-coupling of NH substrates and (hetero)aryl (pseudo)halides for the synthesis of (hetero)anilines is in the midst of a resurgence. Reactivity breakthroughs that have been achieved in this field within the past five years have served to establish Ni catalysis as being competitive with, and in some cases superior to, more well-established Pd- or Cu-based protocols. Whereas the repurposing of useful ancillary ligands from the Pd domain has been the most frequently employed approach in the quest to develop effective Ni-based catalysts for such transformations, considerable progress has been made as of late in the design of ancillary ligands tailored specifically for use with Ni. Bisphosphine ancillary ligands have proven to be well-suited for such an approach, given their modular and facile syntheses; several variants have emerged recently that are particularly effective in enabling a range of otherwise challenging Nicatalyzed Csp2−N cross-couplings. This Perspective presents a comprehensive summary of the advancements within the field of Ni-catalyzed Csp2−N cross-coupling through the application of the bisphosphine ancillary ligand class. It is our intention that the discussion of key ancillary ligand design concepts and mechanistic considerations presented herein will provide a useful platform for researchers to initiate ancillary ligand design efforts for the development of high-performing Ni cross-coupling catalysts. KEYWORDS: nickel, C−N cross-coupling, ligand design, bisphophines, anilines

1. INTRODUCTION The palladium-catalyzed Csp2−N cross-coupling (herein C−N) of NH substrates and (hetero)aryl (pseudo)halides (i.e., Buchwald−Hartwig Amination, BHA) represents a widely employed method, along with Chan−Lam coupling, for the synthesis of (hetero)anilines, with applications in the preparation of natural products, pharmaceuticals, functional materials, and beyond.1−7 The rapid evolution of BHA chemistry over the past two decades is due in large part to advances in ancillary ligand and precatalyst design,8−11 which afforded generalized protocols devoid of salient limitations associated with early precursors to BHA chemistry (e.g., the use of toxic tin amido reagents, limited functional group compatibility, and elevated reaction temperatures). The development of broadly useful ligand classes for application in BHA was guided by extensive mechanistic analysis, wherein the preponderance of empirical data led to the identification of sterically demanding, relatively electron-rich ancillary ligands as being optimal (Figure 1).12−14 These ancillary ligand characteristics promote the formation of electron-rich, lowcoordinate LnPd0 species that are predisposed toward challenging oxidative additions (e.g., Csp2−Cl). Prominent ancillary ligand classes adhering to these basic design principles include trialkylphosphines (e.g., PtBu3 (L1),15 cataCXium A (L2),16), (hetero)biaryl monophosphines (e.g., RuPhos (L3),13,17 BrettPhos (L4),17 BippyPhos (L5),18,19), large bite angle bisphosphines (e.g., XantPhos (L6),20 CyPF-tBu JosiPhos (L7),21−23 dppf (L8),24,25), mixed P,N donors (e.g., Mor-DalPhos (L9),26−29 Me-DalPhos (L10),30), and sterically demanding N-heterocyclic carbenes (NHCs) (e.g., IPr (L11), © XXXX American Chemical Society

Figure 1. Prominent bulky, electron-rich ligands used in BHA chemistry.

IMes (L12),31,32),33 which collectively give rise to Pd catalysts capable of promoting the cross-coupling of a broad spectrum Received: May 15, 2018 Revised: June 13, 2018 Published: June 20, 2018 7228

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

in SET processes has been exploited by pairing Ni catalysts with photosensitizers (e.g., [Ir(dtbpy)3]3+, dtbpy = 4,4′-di-tertbutyl-2,2′-dipyridyl),53 which are presumed to facilitate otherwise challenging cross-couplings (e.g., Csp2−N,54 Csp2− O,55 Csp3−Csp256) by modulating the oxidation state of Ni. Given the inherent advantages associated with using Ni in place of Pd, it is surprising that Ni has received comparatively little attention in C−N cross-coupling over the past two decades. The potential for more diverse reactivity and a broader range of accessible oxidation states undoubtedly discouraged Ni catalyst development, thereby contributing in part to the preferential study of Pd catalysts for C−N crosscoupling. Indeed, the complex reactivity characteristics of Ni complicate mechanistic analysis of such catalyst systems, rendering the identification of key ancillary ligand design criteria, and thus the development of high-performing Ni catalysts, rather problematic.57 In the face of these challenges, chemists have routinely defaulted to repurposing ligands that function well in BHA chemistry as a means of developing new Ni-based amination protocols−an approach that is predicated on the assumption that Ni catalysis is governed by Pd-like mechanisms (e.g., a Pd0/PdII cycle), with analogous ratelimiting steps (e.g., oxidative addition). This strategy has proven somewhat effective for the cross-coupling of a limited selection NH substrate/electrophile pairings, with Ni catalysts based on PPh3,58 IPr,59−64 JosiPhos,49,65−67 dppf,66,68−73 BINAP47,48 and DPEPhos,74 each exhibiting useful performance. Nonetheless, given the frequent failure of prominent ancillary ligands from the BHA domain in enabling more challenging Ni-catalyzed C−N cross-couplings involving sought after NH substrates, and in consideration of the divergent properties of Ni versus Pd, it is unlikely that repurposing BHA ligands will be a universal solution for developing superlative Ni catalysts for C−N cross-coupling. Consequently, the tremendous potential of Ni in C−N crosscoupling catalysis will remain untapped until our understanding of the mechanistic landscape of Ni-catalyzed C−N cross-coupling, and the influence of ancillary ligation on the elementary steps therein, rivals that of BHA chemistry. 1.1. Scope of Perspective. The aim of this Perspective is to highlight major recent advances made within the burgeoning field of Ni-catalyzed C−N cross-coupling, with an emphasis on the discussion of ancillary ligand classes that have proven particularly useful in enabling highly effective catalysts. Notwithstanding the utility of NHCs,59−63,75−84 monophosphines,85,86 bipyridines,87−90 and various other91,92 ancillary ligand classes, we focus our discussion herein on reports in the academic literature featuring the application of bisphosphine ancillary ligands, appearing up to the end of April 2018. Along with NHCs, bisphosphines have emerged as “bestin-class” ancillary ligands in a diversity of Ni-catalyzed C−N cross-couplings (Figure 2); however, whereas the application of NHCs in such transformations has been comprehensively documented elsewhere,64 a review devoted to documenting the major synthetic advances enabled by use of bisphosphineligated Ni catalysts, including recent mechanistic insights gained, has yet to appear. In highlighting key ligand design concepts, we hope to inform the further development of useful ancillary ligands with applications in the field of Ni-catalyzed C−N cross-coupling and beyond. An excellent overview of the development of Ni-catalyzed C−N cross-coupling chemistry has been published by Nicasio and co-workers.93 For guidelines pertaining to the design of Ni cross-coupling

of synthetically useful (hetero)aryl electrophiles and NH reagents. Notwithstanding the broad impact of BHA in synthetic chemistry, both cost and supply limitations of Pd have steered research efforts in the direction of more sustainable amination methodologies that utilize inexpensive, nonprecious metals. Among such metals, catalysts based on Ni, Cu, and Fe have each proven effective in enabling C−N cross-coupling (Figure 2). However, the scope of compatible reaction partners and

Figure 2. Approaches to metal-catalyzed C−N cross-coupling, and the focus of this Perspective.

mild reaction conditions enabled by use of Ni catalysts currently far exceeds that of Cu34−40 or Fe38,41−44 catalysts, making Ni the most promising substitute for Pd in such reactions. Beyond the significant cost benefits associated with using Ni in place of Pd, the reactivity profile of Ni can provide useful advantages in cross-coupling chemistry.45,46 For example, the relatively low electronegativity of Ni serves to enhance reactivity with sought-after (hetero)aryl chlorides and phenol derivatives (e.g., ethers, tosylates, mesylates, triflates, sulfamates, etc.),46 which are comparatively less reactive electrophiles under BHA conditions. Furthermore, the expanded range of readily accessible oxidation states (commonly 0 to III),45,46 and greater propensity for single-electron transfer (SET) for Ni leads to an inherently more complex, but potentially useful, mechanistic landscape versus Pd. Whereas Pd0/PdII cycles involving 2e− elementary steps are commonly observed in BHA,14 the situation is much less clear in the case of Ni. Whereas both Ni0/NiII47−50 and NiI/NiIII50−52 catalytic cycles have been invoked in Ni variants of BHA (Figure 3), the existence of elementary processes that link these manifolds (e.g., comproportionation, disproportionation, and SET) creates ambiguity regarding the dominant operative mechanism of catalysis. In fact, the tendency of Ni species to engage

Figure 3. Comparison of palladium- and nickel-catalyzed C−N crosscoupling. 7229

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

2.3. N-Heterocyclic Carbenes (NHCs). In contrast to monophosphine ligands, NHCs have proven highly effective in enabling both Pd- and Ni-catalyzed C−N cross-coupling chemistry. The seminal report of such Ni-catalyzed aminations utilizing NHC ancillary ligation was disclosed by Fort and coworkers in 2001,104 wherein the coupling of (hetero)aryl chlorides with anilines and secondary cyclic and acyclic amines was enabled by employing mixtures (1:4) of Ni(acac)2 and SIPr·HCl (SIPr = 1,3-bis(2,6-diisopropylphenyl)dihydroimidazol-2-ylidene), in the presence of NaH and tBuOH. Tremendous progress has been made since, and Ni catalysts based on NHCs have been shown to promote crosscouplings of a broad spectrum of both (hetero)aryl electrophiles (chlorides, 59,75,77,78,80,84,105 bromides, 60 tosylates,61,103,106 phosphates,62 pivalates,63 methyl ethers,79 carbamates,82 and sulfamates83), and NH reagents (primary62,82,106 and secondary amines,61−63,75,77−79,82 anilines,61,75,77,82,106 hydrazones,60 indole,59 and carbazole59). A common feature of highly effective amination protocols of this type is the application of well-defined Ni precatalysts (e.g., [Ni(NHC)CpCl], 75 [Ni(NHC)(allyl)Cl],77 [Ni(NHC)(styrene)2],59,61 etc.). This has proven to be a superior approach to catalysis, relative to in situ catalyst generation via ligand exchange between a Ni source (e.g., Ni(acac)2, Ni(cod) 2 , etc.) and the NHC. Though only limited mechanistic data pertaining to Ni-catalyzed C−N crosscouplings employing NHC ancillary ligation are available, current literature findings implicate a Ni0/NiII cycle analogous to that proposed for BHA, involving oxidative addition, amine binding and deprotonation, and product forming C−N reductive elimination;78,80,103,107 the last of these steps has been predicted to be rate-limiting for the cross-coupling of aryl carbamates with secondary amines on the basis of DFT analysis.108 Complementary mechanisms involving a NiI/NiIII cycle have also been postulated51 when employing mononuclear, Y-shaped NiI(NHC)X precatalysts. In these systems it is proposed that initial amine binding and HX loss, followed by aryl halide oxidative addition to the resultant [NiI(NHC)(amido)] intermediate, and subsequent product-forming C−N reductive elimination from a [NiIII(NHC)(Ar)(amido)X] species occurs.

catalysts by minimizing off-cycle species, please see a very recent Perspective by Balcells and Nova.57 Finally, given that the benefits of employing appropriately designed Ni precatalysts is a recurring theme in this Perspective, readers are directed to a review in this area by Hazari and coworkers.10

2. ANCILLARY LIGANDS Ni-CATALYZED C−N CROSS-COUPLING: A BRIEF OVERVIEW To provide context for the main discussion of this Perspective regarding the application of bisphosphine ancillary ligands in Ni-catalyzed C−N cross-coupling (Section 3), a brief overview of other prominent ligand classes employed in such transformations is given below. 2.1. Chelating N,N-Donor Ligands. The field of Nicatalyzed C−N cross-coupling employing aryl chlorides was initiated in 1997, when Wolfe and Buchwald68 reported the cross-coupling of electronically varied electrophiles with selected primary and secondary amines, enabled by use of in situ generated Ni catalysts based on dppf (L8) or 1,10phenanthroline (phen), and employing sodium tert-butoxide as the base. While L8 has since been exploited in a plethora of Ni cross-couplings (e.g., selected examples: Csp2−N,66,69,70 Csp2− Csp2,94,95 Csp2−SCF3,96,97 Csp2−CF2H,98 Csp2−CN;99 the utility of L8 in C−N cross-coupling is described more fully in Section 3.1 of this Perspective), chelating N,N-donor ligands (e.g., phen,68 3,5,6,8-tetrabromo-1,10-phenanthroline,87 and bipyridine88,89 (bpy)) have seen limited use in Ni-catalyzed C−N cross-coupling since the mid 2000s. This is despite the fact that Wolfe and Buchwald68 observed that the use of phen in place of L8 in some cross-couplings led to improved reactivity and selectivity. Recently there has been considerable interest in the use of Ni catalysts supported by such N,N-donor ligands in the context of metallophotoredox-enabled crosscoupling applications.53 2.2. Monodentate Phosphines. Despite figuring prominently in BHA chemistry, monophosphine ancillary ligands including some of those featured in Figure 1 have generally proven inferior to bisphosphines in challenging Ni-catalyzed C−N cross-couplings (e.g., ammonia monoarylation100). Among the useful Ni amination catalysts based on monophosphines that have been reported, most feature simple ligands such as PPh3101 or PMe3,86 and in no cases do such catalysts represent the state-of-the-art. It is unclear as to why monophosphines that perform well in BHA are generally ineffective in Ni-catalyzed C−N cross-couplings. However, the enhanced stability of Ni complexes supported by chelating bisphosphines presumably plays an important role in suppressing catalyst deactivation pathways, which are likely more prevalent in Ni chemistry versus Pd (vide supra).57 For example, recent computational analyses conducted by Maseras and co-workers102 predicts that halide abstraction is competitive with aryl halide oxidative addition for complexes of the type Ni(PMenPh(3−n))4, thus providing a direct route to paramagnetic intermediates (e.g., NiI, NiIII), which may diminish catalytic activity under circumstances where a product-forming Ni0/NiII cycle is operative. It is noteworthy, however, that well-defined Ni coordination complexes supported by monophosphines, including [Ni(PPh3)2Cl2]60,101 and [Ni(PPh3)2(1-naphthyl)Cl],62,103 can serve as useful precatalyst precursors, wherein the target precatalyst may be obtained via PPh3 substitution with a desired ancillary ligand.

3. BISPHOSPHINE LIGANDS FOR ENABLING Ni-CATALYZED C−N CROSS-COUPLING Whereas electron-rich, sterically demanding monodentate phosphines (e.g., L1−L5; Figure 1) currently represent some of the most effective ancillary ligands known for application in BHA, these ligands often perform poorly when applied in Nicatalyzed C−N cross-coupling. Conversely, building on the pioneering 1997 report by Wolfe and Buchwald68 regarding the use of dppf (L8), several other bisphosphines repurposed from BHA chemistry have proven effective for such Nicatalyzed transformations, including BINAP (L13), DPEPhos (L14), and variants of JosiPhos (L15−L17), with most reports of this type appearing within the past five years (Figure 4). A less well-developed, yet highly effective approach in the pursuit of useful Ni C−N cross-coupling catalysts has involved the rational development of ancillary ligands specifically tailored for use with Ni, including ortho-phenylene bisphosphines (L18−L22; Figure 4) such as PAd-DalPhos (L18). Collectively, such studies involving both repurposed and new bisphosphines have enabled the identification of high-performing Ni catalysts for C−N cross-coupling, while also laying the 7230

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

enabled the cross-coupling of electronically diverse aryl chlorides with primary or secondary amines in synthetically useful yields, when conducting reactions at elevated temperatures (70−100 °C) in the presence of sodium tert-butoxide (Scheme 1). Notably, the chelating N,N-donor ligand 1,10phenanthroline (phen), which is not generally effective under BHA conditions, outperformed the Ni(cod)2/L8 catalyst system in select substrate pairings, affording higher product yields and lower quantities of reduced arene byproduct. Nonetheless, reactions involving phen required higher catalyst loadings than those featuring L8 (5% vs 2%). In an effort to circumvent the use of air-sensitive Ni(cod)2 as a catalyst precursor, Wolfe and Buchwald employed air-stable precatalysts of the type [NiCl2(L8)] (C1), in conjunction with catalytic quantities of MeMgBr as a reducing agent; such mixtures exhibited comparable performance to catalysts generated in situ from Ni(cod)2. While the use of aryl chlorides under the reported conditions generally gave excellent product yields, analogous aryl bromides gave rise to substantial quantities of reduced arene byproducts. Such observations may be rationalized in terms of a mechanism involving electron transfer from Ni to the aryl bromide, and/or deleterious side-reactions involving [Ni(L8)(aryl)Br] intermediates (e.g., comproportionation). Several years later in 2000, Lipshutz and Ueda71 utilized dppf (L8) ligation for developing the first heterogeneous Nibased catalyst for the amination of aryl chlorides (Scheme 2). To achieve an active Ni0 catalyst, a NiII salt supported on charcoal is reduced by n-BuLi in the presence of L8. Notably, the monodentate ligands examined in this study (PCy3, PPh3, and P(o-tolyl)3) were considerably less effective than L8. Under refluxing conditions in either toluene or 1,4-dioxane, this catalyst system enabled the efficient coupling of aryl chlorides and primary alkylamines, anilines, or secondary cyclic amines, using lithium tert-butoxide as base. While examples of electron-rich (e.g., p-chloroanisole), electron-poor (e.g., pchlorobenzonitrile) and hindered ortho-substituted (e.g., 2chloro-4-fluorotoluene) substrates are included in this report, the scope of electrophiles is limited to only five examples total, and the capacity for the catalyst system to tolerate broad functionality with respect to both coupling partners is not wellestablished. Furthermore, the need for high catalyst loadings in select pairings (10/20% Ni:L8), and two equivalents of the amine coupling partner, are limitations of the protocol. In 2003, the Lipshutz research group reported a simplified protocol109 that does not require prereduction of the NiII/C precursor by n-BuLi, and instead achieves an active Ni0 species via an in situ reduction of NiII/C by action of an amine/LiO-tBu mixture. In examining the influence of ancillary ligand electronic factors on reaction kinetics, they found that more

Figure 4. Bisphosphine ancillary ligands featured in this Perspective (PS = polystyrene, R = para-tert-butyl).

groundwork for establishing an understanding of the mechanistic landscape associated with such transformations. 3.1. 1,1′-Bis(diphenylphosphino)ferrocene (dppf). As alluded to earlier in this document, the breakthrough 1997 report by Wolfe and Buchwald68 is commonly viewed as having established Ni-catalyzed C−N cross-coupling as a bona f ide synthetic approach, especially given that aryl chloride electrophiles, which are considerably less reactive under BHA conditions, were employed successfully. While many of the ligands from the field of BHA performed poorly in test Ni cross-coupling reactions (e.g., BINAP, dppp, dppe, PPh3, P(otolyl)3), the chelating ferrocenyl bisphosphine dppf (L8) was found to be highly effective when used in conjunction with Ni(cod)2. Application of the Ni(cod)2/L8 catalyst system

Scheme 1. First Ni-Catalyzed C−N Cross-Coupling of (Hetero)aryl chlorides

7231

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis Scheme 2. Heterogeneous Ni-Catalyzed Amination of Aryl Chlorides

Scheme 3. Ni-Catalyzed C−N Cross-Coupling/Hydroamination of ortho-Alkynyl-Haloarenes for Indole Synthesis

Whereas monodentate phosphine ancillary ligands (PPh3, PCy3, (1-Ad)2P(O)H, XPhos) provided unsatisfactory product yields ( P(p-tolyl)3 > PPh3 ≥ P((p-trifluoromethyl)phenyl)3). This result was not anticipated, given that product-forming C−N reductive elimination is often considered to be rate-limiting in Ni-based C−N cross-coupling, and in this scenario electron-deficient ancillary ligands would be expected to give rise to faster reaction rates. By comparison, L8 afforded superior reaction rates and higher product yields relative to each of the triarylphosphines examined, indicating that chelating bisphosphines are superior to monophosphine ligands when utilized in this particular reaction setting. Given that aryl halide oxidative addition is typically facile with Ni0 species,46 the authors propose that complexation/deprotonation of the amine is likely rate limiting for heterogeneous Nicatalyzed aminations of this type. In considering mechanistic scenarios leading to NiII amido species, a pathway involving initial complexation of tert-butoxide to form a NiII(L)(O-t-Bu) intermediate, followed by subsequent complexation/deprotonation of the amine, is invoked. This proposal is based on the observed reactivity trend for electronically varied triarylphosphines (vide supra), in which electron-rich ligands afforded the fastest rates. Under this proposed mechanism, such ancillary ligands should accelerate complexation/deprotonation of the amine by increasing the basicicity of tert-butoxide in a NiII(L)(O-t-Bu) type intermediate. An important addition to the substrate scope of Ni-catalyzed aminations came in 2011, when Ackermann and co-workers reported72 an efficient synthesis of substituted indoles via sequential Ni-catalyzed C−N cross-coupling/hydroamination of ortho-alkynyl haloarenes (X = Br, I) employing alkyl and aryl amines (Scheme 3). The reaction is promoted by an in situ generated catalyst based on a 1:1 mixture of L8 and Ni(cod)2 (5 mol %), in the presence of NaO-t-Bu (for the C−N crosscoupling step), and KO-t-Bu (for the hydroamination step). 7232

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

Scheme 4. Ni-Catalyzed Amination of (Hetero)aryl (Pseudo)halides Enabled by Use of the Air-Stable [Ni(L8)(o-tolyl)Cl] Precatalyst C2

Scheme 5. Direct Ni-Catalyzed Amination of Phenols via C−O Bond Activation by Using 2,4,6-Trichloro-1,3,5-triazine (TCT)

weaker base K3PO4 was necessary. Notably, the protocol requires an equivalent of MeCN and additional L8 to achieve optimal performance, which represent limitations of this chemistry. The Stewart group110 subsequently reported the synthesis and application of an air-stable [Ni(L8)[P(OPh)3]2] precatalyst (C3), in an effort to establish methodologies that do not rely on the costly and air-sensitive Ni(cod)2 precursor. Precatalyst C3 activates via dissociation of phosphite ancillary ligands, and is complementary to the [Ni(L13)(η2-NCPh)] precatalyst (C4) developed by Hartwig and co-workers (see Section 3.2.1)47 in that it is already in the requisite Ni0 oxidation state for oxidative addition. This study was initiated by screening the performance of various ligands used in conjunction with the Ni[P(OPh)3]4 precatalyst for the C−N

The scope of effective phenol-derived (pseudo)halide coupling partners was expanded in 2014 by Buchwald and co-workers69 to include synthetically attractive aryl mesylate electrophiles, transformations that were enabled by use of the air-stable [Ni(L8)(o-tolyl)Cl] (C2) precatalyst (Scheme 4). Such NiII precatalysts are experiencing increased usage, given that they are typically air-stable unlike most Ni0 species, and activate rapidly under catalytic conditions without the need for additives or exogenous reductants.10 Employing C2 allowed for the cross-coupling of a range of structurally diverse (hetero)aryl (pseudo)halides (X = Cl, OTf, OSO2NMe2, OMs) with anilines and secondary alkylamines in synthetically useful yields. Although (hetero)aryl chlorides were aminated in the presence of LiO-t-Bu, in moving toward electrophiles with base-sensitive functionality, it was found that switching to the 7233

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

Figure 5. Summary of dppf (L8) utility in Ni-catalyzed C−N cross-coupling.

cross-coupling between para-chloroanisole and para-toluidine, to furnish a secondary N,N-diarylamine. While monodentate (Ad2PBu, XPhos, DavePhos) and bidentate (XantPhos, dppe, 1,10-phenanthroline, TMEDA) ligands performed poorly, the bisphosphines BINAP (L13) and dppf (L8) were highly competent, with the Ni[P(OPh)3]4/L8 system providing the best catalytic performance in the survey. Under optimized conditions, the cross-coupling of various aryl (pseudo)halides (X = Cl, Br, I, OTf, OTs) with anilines, primary alkylamines, and secondary cyclic/acyclic amines employing C3 (5 mol %) and additional L8 (5 mol %) was achieved. Good functional group tolerance regarding the (hetero)aryl electrophile was demonstrated, whereby substrates featuring alkyl, nitrile, trifluoromethyl, ketone, ether, and halide functionality were transformed effectively. The scope of nucleophiles was much less expansive, with the majority of the examples being derived from para-toluidine; however, the use of 5 mol % L13 in place of L8 enabled the pairing of other amine classes such as primary alkylamines (e.g., benzylamine and n-hexylamine), secondary acyclic (N-methylbenzylamine) and secondary cyclic amines (e.g., morpholine) with chlorobenzene. In contrast to a related report by Hartwig and co-workers47 where rate-limiting C−Cl oxidative addition is invoked, preliminary mechanistic analysis of cross-couplings involving C3 indicates a slow product-forming C−N reductive elimination, which the authors propose is accelerated by the addition of L8 in the reforming of the initial C3 species.110 Iranpoor and co-workers92,111 reported a direct Ni-catalyzed C−N cross-coupling of phenols via C−O bond activation with a 2,4,6-trichloro-1,3,5-triazine (TCT) reagent (Scheme 5). Whereas the use of phenols in such amination chemistry typically requires prior conversion into an appropriate aryl C− O electrophile so as to facilitate cross-coupling (e.g., aryl ethers, 112 carbamates, 67,82,113,114 pivalates, 63,67 sulfamates67,83,114), the Iranpoor methodology allows for the direct transformation of phenols by generating activated Ar-O-TCT intermediates in situ, which are then C−N cross-coupled in the presence of an appropriate Ni catalyst. In optimizing the reaction conditions, a series of [Ni(L)Cl2] (L = PPh3, PCy3, dppe, dppp, and dppf) complexes were screened in the crosscoupling of para-cresol and morpholine. Whereas [Ni(PPh3)2Cl2] (C5) and [Ni(PCy3)2Cl2] (C6) precatalysts gave moderate product yields (40−55%), the chelating

bisphosphine-based precatalysts [Ni(dppp)Cl2] (C7), [Ni(dppe)Cl2] (C8), and C1 (i.e., [Ni(dppf/L8)Cl2]) each gave high product yields, with the C1 being superior for this reaction. Employing C1 (5 mol %) in conjunction with K3PO4 in refluxing 1,4-dioxane permitted the amination of phenols with an impressive scope of amine nucleophiles, including secondary cyclic and acyclic amines, anilines, primary alkylamines, pyrazole, imidazole, indazole, indole, carbazole and ammonium salts, all in synthetically useful yields. In an effort to evaluate the effect of varying the dppf (L8) PR2-donor groups on catalytic performance, our group70 tested a series of ten structurally varied 1,1′-bis(di(alkyl/aryl)phosphine)ferrocene ancillary ligands in competitive crosscouplings of (hetero)aryl chlorides with furfurylamine, morpholine, and indole, employing L/Ni(cod)2 catalyst mixtures. It was found that in addition to the generally excellent performance of parent L8 in the test transformations, selected dialkylphosphino (e.g., DiPPF), and meta-disubstitutied diarylphosphino variants exhibited comparable, and in some cases superior, reactivity to parent dppf (L8). In particular, an electron-deficient variant featuring 3,5,-bis(trifluoromethyl)phenyl groups on phosphorus generally demonstrated enhanced catalytic performance relative to parent L8 in the N-arylation of indole, ultimately enabling the first example of a room-temperature indole N-arylation by using Ni catalysis. Notably, L8 variants featuring orthosubstituted diarylphosphino groups (e.g., R = ortho-tolyl, 1naphthyl), or sterically demanding dialkylphosphino analogues (e.g., R = tBu) generally proved ineffective. The comparable performance in some instances of the meta-disubstituted diarylphosphino variants examined (e.g., R = 3,5-bis(trifluoromethyl)phenyl, 3,5-dimethoxyphenyl, or 3,5-xylyl) indicates that electronic perturbations of the diarylphosphino moiety do not exert a strong influence on the behavior of Ni in these transformations. A summary of the application of L8 in Ni-catalyzed C−N cross-coupling is provided in Figure 5. Note that a survey of L8-derived Ni precatalysts in varying oxidation states is discussed as part of Section 3.4.1. 3.2. 1,1′-Binaphthalene-2,2′-diyl)bis(diphenylphosphine (BINAP) and (Oxydi-2,1phenylene)bis(diphenylphosphine) (DPEPhos). Despite the tremendous utility of dppf (L8) in enabling Ni-catalyzed C−N cross-couplings (vide supra), the successful application of 7234

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

Scheme 6. Ni-Catalyzed Amination of Unactivated (Hetero)aryl Halides Using Primary Aliphatic Amines, Catalyzed by C4a

a

dme = dimethoxyethane.

Figure 6. Proposed mechanism for the C4-catalyzed C−N cross-coupling of primary alkylamines and (hetero)aryl halides. Ar = aryl.

other known bisphosphine ancillary ligand classes in addressing challenging transformations of this type did not appear until 2014, when the Hartwig group47 disclosed an effective primary alkylamine cross-coupling protocol that makes use of the BINAP-based precatalyst [Ni(L13)(η2NCPh)] (C4). In the ensuing years, other bisphosphine ancillary ligands have been successfully repurposed in Nicatalyzed C−N cross-coupling chemistry, including DPEPhos

(L14, Section 3.2.2) and JosiPhos variants (L15−L17, Section 3.3). Application of these known ancillary ligands, along with new ortho-phenylene bisphosphines (Section 3.4), has collectively led to a rapid expansion of the reaction scope in such cross-couplings. Within the following sections, reports documenting the use of BINAP (L13) and DPEPhos (L14) in this context are discussed. Precatalysts featuring these wide bite-angle ancillary ligands are particularly useful in enabling 7235

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis Scheme 7. C−N Cross-Coupling of (Hetero)aryl (Pseudo)halides with Primary Alkylamines Catalyzed by C9

for an operative Ni0/NiII catalytic cycle in this particular system. On the basis of stoichiometric reactions and kinetic analyses, Hartwig and co-workers47 propose a mechanism depicted in Figure 6. Depending on if aryl chlorides or bromides are employed, the initial C4 resting state species either undergoes reversible dissociation of PhCN followed by oxidative addition in the case of aryl chlorides, or reacts directly with aryl bromides followed by PhCN dissociation. In either case, oxidative addition is presumed to be rate-limiting on the basis of the observed first-order rate-dependence on the concentrations of both the catalyst and aryl halide. The [Ni(L13)Ar(halide)] intermediate subsequently reacts with primary amine and base to form an [Ni(L13)Ar(amido)] species, followed by C−N reductive elimination of the N-alkyl aniline product in the presence of PhCN to regenerate C4.47 In 2015, Stewart and co-workers48 developed an air-stable phosphite-ligated [Ni(L13)(P(OPh)3)2] precatalyst (C9) that is complementary to the [Ni(L13)(η2-NCPh)] precatalyst (C4) developed by Hartwig and co-workers.47 Employing C9 enabled the efficient coupling of primary alkylamines with structurally diverse aryl chlorides and bromides (Scheme 7). In comparison to Hartwig’s methodology for primary alkylamine cross-coupling employing C4 (vide supra),47 the demonstrated substrate scope achieved by using C9 is much less expansive;48 for example, the use of heterocyclic electrophiles is limited to two examples involving pyridine and quinaldine in the Narylation of n-hexylamine. Furthermore, slightly higher catalyst loadings, the addition of free L13 (5 mol %), and higher reaction temperatures relative to the C4-based protocol are apparently needed to obtain suitably high product yields.48 Nonetheless, Stewart and co-workers were able to extend the scope of reactivity to include transformations of p-chlorotoluene or p-bromotoluene with benzophenone imine in the presence of C9. This nucleophile had not been utilized previously in a Ni-catalyzed C−N cross-coupling, and hydrolysis of the N-aryl imine cross-coupled product furnishes p-toluidine. While in this context benzophenone imine is serving as an ammonia surrogate, it is worth mentioning that the groups of Hartwig,49 Stradiotto,65,67,100 and Schranck, Tlili and co-workers113 have developed Ni-catalyzed methodologies for the monoarylation of ammonia (Sections 3.3 and 3.4), providing a direct route to aniline derivatives.115

the Ni-catalyzed N-arylation of primary and secondary alkylamines. 3.2.1. BINAP (L13). Prior to 2014, the nucleophile scope of Ni-catalyzed C−N cross-couplings was limited primarily to secondary amines and anilines. Work by the Hartwig group47 served to expand the scope to include a variety of primary alkylamines, which were arylated using unactivated (hetero)aryl chlorides and bromides (Scheme 6). The application of a new L13-ligated Ni0 precatalyst C4 (1−4 mol %) stabilized by a side-bound benzonitrile coligand was key to enabling this transformation under relatively mild reaction temperatures (50−80 °C). Hartwig and co-workers47 initiated their study by screening of a variety of phosphine ancillary ligands in combination with Ni(cod)2, for the coupling of 3-chloroanisole and n-octylamine. Whereas the application of monodentate phosphines (PPh3, PCy3) and alkyl bisphosphines (dppe, dppp, dppb) resulted in minimal substrate turnover (90%) to the target 4-aminobiphenyl was noted. The divergent performance of L18 and L23 highlights the importance of ancillary ligand sterics in enabling such transformations. In keeping with the concept that the design of L18 is particularly well-matched to Ni, no conversion was achieved when this ancillary ligand was used with [Pd(cinnamyl)Cl]2 in place of Ni(cod)2. In an effort to differentiate the contribution of ancillary ligand steric and electronic properties in Ni-catalyzed ammonia monoarylation chemistry, the bulky, relatively electron-poor CgP fragment in L18 was replaced with a sterically similar, yet comparatively electron-rich di-tert-butylphosphine group (L25). While inferior to L18, the competent performance of L25 indicates that ancillary ligand sterics are particularly important in engendering desired reactivity for this class of Ni-catalyzed C−N cross-coupling reactions.100 In an effort to circumvent the use of Ni(cod)2, the air-stable precatalyst [Ni(L18)(o-tolyl)Cl] (C16) was prepared in a straightforward manner as outlined in Scheme 11. In addition to routinely out-performing Ni(cod)2/L18 catalyst mixtures, precatalyst C16 has proven effective in enabling a range of otherwise challenging C−N cross-couplings at relatively low

Figure 8. Summary of JosiPhos (L15−L17) utilization in Nicatalyzed C−N cross-coupling.

ancillary ligands specifically tailored for use in enabling Nicatalyzed C−N cross-coupling was not reported until 2016, nearly 20 years after the seminal report on Ni-catalyzed amination by Wolfe and Buchwald.68 3.4.1. PAd-DalPhos (L18). In the context of our research program directed toward the rational development of ancillary ligands for use in addressing challenging Ni-catalyzed C−N cross-couplings and beyond, we reported in 2016 on the synthesis and application of the air-stable ortho-phenylene bisphosphine ancillary ligand PAd-DalPhos (L18), featuring 1,3,5,7-tetramethyl-2,4,8-trioxa-6-phosphaadamantanyl (CgP) and di-ortho-tolylphosphino donor fragments.100 In the pursuit of highly effective ancillary ligands for such C−N crosscoupling applications, we targeted bisphosphines that were both sterically demanding and relatively poor electron donors, given that such ancillary ligands might promote rate-limiting 7239

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

Scheme 11. Ligand Screen in the Ni-Catalyzed Monoarylation of Ammonia, and the Synthesis of L18-Derived Precatalyst C16a

a

Conversions reported as % 4-aminobiphenyl (% unreacted 4-chlorobiphenyl).

Scheme 12. Ni-Catalyzed C−N Cross-Coupling of Ammonia or Primary Alkylamines with (Hetero)aryl (Pseudo)halides Enabled by C16a

a

M = Na or Li.

Later that year, we reported120 on the application of C16 in enabling the first Ni-catalyzed N-arylation of primary amides and lactams with (hetero)aryl (pseudo)halides (Scheme 13). This study was initiated by screening ligands that function well in analogous Pd-catalyzed C−N cross-coupling of amides, including JackiePhos,121 BippyPhos (L5 122), XantPhos (L6123), and dppf (L8124); none proved to be particularly effective in the test cross-coupling reaction between 4chlorobenzonitrile and benzamide, in combination with Ni(cod)2. In targeting ancillary ligands that had proven effective in enabling Ni-catalyzed C−N cross-couplings involving alternative NH nucleophiles, BINAP (L13) and SIPr (a saturated variant of L11) performed poorly, whereas both CyPF-Cy (L15) and PAd-DalPhos (L18) afforded >90%

catalyst loadings (1−5 mol %), and without recourse to precious metal cocatalysts54 or other additives (e.g., nitriles; vide supra). In our initial report on the synthesis and application of C16,100 the utility of this precatalyst in promoting the cross-coupling of a broad scope of (hetero)aryl (pseudo)halides (X = Cl, Br, I, OTf, OTs, OMs, imidazolylSO3) and NH coupling partners (ammonia, primary alkylamines, and their derived ammonium salts) was disclosed, including room-temperature transformations, as well as the first pairing of ammonia with (hetero)aryl mesylate electrophiles by action of any transition metal catalyst (Scheme 12). Proof-of-principle cross-couplings of indole and related nucleophiles were also reported.100 7240

DOI: 10.1021/acscatal.8b01879 ACS Catal. 2018, 8, 7228−7250

Perspective

ACS Catalysis

Scheme 13. Ni-Catalyzed C−N Cross-Coupling of Primary Amides and Lactams with (Hetero)aryl (Pseudo)halides, Employing Precatalyst C16

Scheme 14. Ni-Catalyzed Cyclization of in Situ Generated ortho-Chlorobenzophenone Hydrazones, Enabled by Use of C16

chemistry include the poor reactivity of C16 with unactivated and deactivated (hetero)aryl (pseudo)halides, as well as with secondary amides. In a subsequent collaboration with the Bonacorso group, we reported125 on the application of C16 in promoting the first Ni-catalyzed cyclization of in situ generated ortho-chlorobenzophenone hydrazone derivatives to afford 1H-indazole products, which can be transformed subsequently in a onepot procedure via Cu-catalyzed N-arylation with (hetero)aryl bromides for the assembly of 1,3-di(hetero)aryl-1H-indazoles (Scheme 14). Notably, neither the CyPF-Cy (L15) nor dppf (L8)-derived [Ni(L)(o-tolyl)Cl] precatalysts were effective in enabling the formation of 1H-indazole products (