Catalytic Isohypsic‐Redox Sequences for the Rapid Generation of Csp3‐Containing Heterocycles

Abstract Cross‐coupling reactions catalyzed by transition metals are among the most influential in modern synthetic chemistry. The vast majority of transition‐metal‐catalyzed cross‐couplings rely on a catalytic cycle involving alternating oxidation and reduction of the metal center and are generally limited to forging just one type of new bond per reaction (e.g., the biaryl linkage formed during a Suzuki cross‐coupling). This work presents an Isohypsic‐Redox Sequence (IRS) that uses one metal to effect two catalytic cycles, thereby generating multiple new types of bonds from a single catalyst source. We show that the IRS strategy is amenable to several widely used transformations including the Suzuki–Miyaura coupling, Buchwald–Hartwig amination, and Wacker oxidation. Furthermore, each of these reactions generates value‐added heterocycles with significant sp3‐C (3‐dimensional) content. Our results provide a general framework for generating complex products by using a single metal to fulfill multiple roles. By uniting different combinations of reactions in the isohypsic and redox phases of the process, this type of catalytic multiple bond‐forming platform has the potential for wide applicability in the efficient synthesis of functional organic molecules.

Transitionm etal catalysis is among the most commons trategies in organic synthesis for the formation of CÀCa nd CÀX bonds. [1] The vast majority of organometallic reactions rely on as ingle catalytic cycle to generate one new bond. [2] Although the power of transition metal catalysis to effect previously unknown reactions has provedt ob et remendously enabling, this "one reaction-oneb ond" limitation fails to maximize the complexity of the products generatedu sing these methods. Catalytic multiple-bond-forming strategies carry vast potential to impact the "economies of synthesis" through the rapid evolution of molecular complexity. [3] We set out to develop am ulti-ple-bond-forming reaction sequence that would use as ingle metalt oe ffect multiple catalytic cycles by uniting an isohypsic reactionm anifold with more common redox catalytic cycles. [4] Most transition-metal-catalyzed reactions form as ingle new bond via amechanistic cycle that involves alternating oxidative and reductives teps with respect to the metal catalyst.Asmaller,b ut still widely used, set of metal-catalyzed processes occurs without changes in the metalo xidation state,t hat is, in an isohypsic manifold. Common examples of such catalytic cyclesi nclude the conjugate additiono fo rganoboronates to a,b-unsaturated carbonyl compounds, [5] Au I or Co III -catalyzed alkynea ctivations, [6] metallocarbenoid reactions (e.g.,R h II -catalyzed reactions of a-diazocarbonyls), [7] and the chain propagation phase of metal-catalyzed alkene polymerization. [8,9] The preference for redox-active catalytic cycles is maintained in the wide field of Pd-catalyzedf ine chemical synthesis. Most processes occur by some variant of the well-knowni terative sequence of oxidative addition, transmetallation, and reductive elimination. Nevertheless, many isohypsic Pd-catalyzed processesa re known, such as the additiono fo rganoboronates to activated p-bonds, [10] cycloisomerization processes that terminate by protonation or b-halide elimination, [11] allylic rearrangementso fe sters or imidates, [12] and the halo-allylation of alkynes, [13] among others. [14] The mechanistic distinctionb etween redox-activea nd isohypsicc atalysis carriesa ni mportant consequence from as ynthetic perspective, namely that functionality that is inert to the metalo xidation state presenti nt he isohypsic process (such as the aryl halides typicallyi nvolved in oxidative addition to Pd 0 ) should be tolerated during an isohypsic reaction at ad ifferent oxidation state (e.g.,P d II ). Subsequent alteration of the metal oxidation state (for example by the addition of an ew reagent) allowsf or as econd catalyticb ond formation to occur using the same metal ( Figure 1). This type of transition from isohypsic to redoxm anifolds is an example of assistedt andem catalysis where one precatalyst effects two distinct catalyticp rocesses using sequential reagentc ombinationst oc ontrol the change in mechanism. [15] Despite its potentialf or broad utility (based on the number of well elucidated catalytic cycles), this isohypsic-redoxs trategy has seldom been used in the field of Pd-catalysis,a nd never in the context of alkene difunctionalization. [16] We have previouslyd eveloped aP d-catalyzed alkene difunctionalizationr eaction that forms ah eterocycle with concomitant creation of an sp 3 -sp 3 CÀCb ond ( Figure 2b). [17] This methodology was specificallyd esigned to generate heterocycles with significant sp 3 -C content,a ss tudies of clinical success rates indicate ac orrelationb etween the progress of drug candidates through clinical trials and enhanced three-dimensionality. [18] An isotopic labeling study suggested the alkene heteroallylation process proceeds via an isohypsic mechanism involving as omewhat unusual b-halide elimination step. [17,19] Here, we describe the development of an isohypsic-redox sequence (IRS) based on the unification of alkene heteroallylation with transformativeP d-catalyzedr edox-activep rocesses such as the Suzuki-Miyaura coupling, Buchwald-Hartwiga mination,a nd both the Wacker and Feringa-Grubbs aldehyde-selective Wacker oxidation protocols (Figure 2c). [16d, 20] This IRS approach enhances molecular complexity by generating three new bonds in as ingle process while also forming ah eterocycle and an ew sp 3 -sp 3 CÀCbond.
Our first task in achieving the planned IRS was to identify an appropriate substrate for the alkene heteroallylation process that contained af unctional handle for use in ad iverse array of subsequentr edox reactions. As aryl halides are the most commonly used coupling partners in standard Pd-catalyzed processes, bromophenol 1 was selected as our initial test case ( Figure 3). Gratifyingly,t his alkenyl phenol underwent the desired heteroallylationr eaction to generate benzofuran 2 in good yield under our previously optimized conditions without engaging the aryl bromide, as expected by the all Pd II catalytic cycle. [17] Once the heteroallylation in the presence of an aryl bromide had been demonstrated, we set out to establish our first IRS using the Suzuki-Miyaura cross-coupling, the most common CÀCb ond-forming reactionu sed by medicinal chemists. [21] In this process, we were relying on the well-precedented reduction of Pd II to Pd 0 by boronic acids to initiate the redox catalytic cycle. [22] After optimization, [23] including use of Buchwald dialkylbiaryl phosphine ligands, [24] we were able to generate the desired biaryl coupling products in good yield through the two catalytic cycles (Figure 4). Substrate scoping studies demonstrated that both electron-withdrawing and electron-donating substituents were tolerated.B ym odifying the phosphine ligand to XPhos in the case of thiophene (3e), [25] andP Phos in     Chem. Eur.J.2018, 24,17201 -17204 www.chemeurj.org 2018 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim the case of pyridine( 3f), [26] we were able to effectively couple these heterocycles.
Having demonstrated the capacity to form CÀCb onds in the redox phase of the IRS process, we next chose to studyC À Nb ond formation using the Buchwald-Hartwig amination. [27] In this instance,r eduction of theP d II wase nvisaged to occurv ia a b-hydride eliminationf romaPd II -amine complex. [28] Useo fa dialkylbiarylp hosphine wasa gain foundt ob ea dvantageous in coupling with hexyla mine (Figure5). In addition to primary amines,t he coupling proceeded well with secondary amines to generate morpholine 4c,piperazine 4d,and aniline 4e.
In order to extend the scope of the IRS platform beyond functionalized benzofurans, as well as to make use of the double bond that is installed by the isohypsic heteroallylation, we set out to combine the synthesis of N-containing heterocycles with oxidationo ft he doubleb ond as ar edox step ( Figure 6). [29] The Pd 0 generateda tt he end of the Wacker process would be re-oxidized by an external oxidant to complete ar edox cycle. After screening ar ange of conditions fort he standard Wacker oxidation, such as varying the re-oxidant system, [23] we found that benzoquinonew as the most effective (5 to 6). Am ethyl ketone was successfully installed in compounds containing both the isoquinolone andp yrrolopyrazinone ring systems. We then turnedo ur attention to the possible aldehyde-selective Wacker-type alkene oxidation developedb yF eringa and Grubbs et al. [16d, 20] Using silver nitrite and copper(II) chloride as co-catalysts resultedi nf ormation of the expected aldehyde as the major product in am odest overall yield consistent with the yields reported for theset wo processes in isolation (5 to 7). [20e] Interestingly,t he presence of a nitrile ligand (as used in earlier work by Feringa and Grubbs et al.) was found to be essential for the reaction to proceed.
In summary,w ehave developedasuite of tandemc atalytic processes based around the concept of linking the isohypsic (redox neutral) alkene heteroallylation reaction with well-known redox catalytic cycles including the Suzuki-Miyaura, Buchwald-Hartwig, and Wacker transformations. In all cases, one metal is used to effect two different catalytic cycles, thereby providing as trategy for the rapid evolution of molecular complexity in the context of forming 3D heterocycles. Given the number of well-elucidated catalytic cycles,e xpansiono f the IRS concept has vast potential both within the field of Pdcatalysis and beyond.