Photoredox Imino Functionalizations of Olefins

Abstract Shown herein is that polyfunctionalized nitrogen heterocycles can be easily prepared by a visible‐light‐mediated radical cascade process. This divergent strategy features the oxidative generation of iminyl radicals and subsequent cyclization/radical trapping, which allows the effective construction of highly functionalized heterocycles. The reactions proceed efficiently at room temperature, utilize an organic photocatalyst, use simple and readily available materials, and generate, in a single step, valuable building blocks that would be difficult to prepare by other methods.

Small-molecule nitrogen heterocycles constitute the core of many drugs and agrochemicals and are one of the key epitopes evaluated in biological screenings. [1] As aresult, the invention of methods enabling their rapid construction is at opic of continuous scientific endeavour.A no verarching goal in the design of these methodologies is the identification of divergent approaches whereby simple starting materials are converted into ab road array of products which contain different functional groups and efficiently tap into chemical space.
Nitrogen radicals are ac lass of versatile synthetic intermediates,a nd their reactivity encompasses as eries of powerful bond-forming reactions,like intramolecular cyclization and 1,5-H abstraction. [2] In recent years,o wing to the power of visible-light photoredox catalysis [3] in realizing single-electron-transfer (SET) processes, [4] thec hemistry of nitrogen-radicals has witnessed ar emarkable resurgence in scientific interest. [5] We have developed aclass of electron-poor O-aryl oximes and aryloxy amides which allows,upon reductive SET,access to iminyl [6] and amidyl [7] radicals to use in radical hydroaminations (Scheme 1A). However,despite all our efforts,we have not been able to harness this reactivity mode for the development of much sought after, but also more challenging, amino-functionalization processes. [2i,8] Herein, we describe our work towards this goal, which has led to the development of ap owerful platform for the divergent assembly of polyfunctionalized nitrogen heterocycles (Scheme 1B). [9] While facing the difficulties of developing cascade photoredox imino functionalizations of electron-poor O-aryl oximes, [6] we realized that an inherent chemical reactivity issue was associated with the envisaged catalytic cycle.T he nitrogen radicals A are electrophilic species and undergo facile exo-trig cyclizations (Scheme 1C). [10] However,u pon ring closure,they generate the nucleophilic carbon radical B, which benefits from polar effects [11] in the reaction with SOMOphiles X À Y. This process leads to the formation of the imino-functionalized product C and the radical YC.T his radical, even though it is not involved in any bond-forming event, controls the fate of our photoredox manifold. In fact, owing to the required polarization of XÀYf or efficient reaction with B,YC is expected to be electrophilic.This species generates an electronic mismatch in our first-generation photoredox cycle,w hereby [PCC + + YC ! PC + Y + ]i s ae ndothermic process (Scheme 1D,l eft). Therefore,w e reasoned that should the iminyl A be generated by oxidative SET from the visible-light-excited photocatalyst (*PC), af acile SET between YC and PC ·À ought to take place,t hus ensuring catalytic activity (Scheme 1D,r ight). An additional challenge associated with the realization of this approach concerns the nature of the SOMOphile X À Y. [12] In fact, many of these species are excellent SET quenchers for *PC,a nd they have been exploited in many powerful atom/grouptransfer reactions and they can lead to the formation of onium ions. [12b] As ar esult, we would require the oxidative generation of the iminyl radical, its intramolecular cyclization, and the following intermolecular radical reaction to supersede any other possible photoredox and ionic pathway.
Inspired by the pioneering work of Forrester [13] and Zard, [2h] we speculated that the oxime I might serve as at raceless (n + s + s)e lectrophore and undergo as equence of two b-scissions upon in situ deprotonation (II)a nd SET oxidation (Scheme 2A). This step would deliver an iminyl radical in conjunction with the release of CO 2 and HCOH (III!IV). [14] As a-hydroxy acids (e.g., 1)have been employed in many photoredox decarboxylations, [12c, 15] we were confident that such as ubstrate design would be feasible.S urprisingly,e lectrochemical analysis of the model substrate 2a revealed its oxidation potential to be almost outside the range for photoredox oxidation (E 1/2 ox = 2.10 Vv s. SCE;S cheme 2B). In fact, previous activation methods required the use of strong oxidants either under forcing conditions or with the formation of the corresponding Barton ester. [2h,13] Hence,wespeculated that the addition of organyl groups at the methylenic position might inductively increase the electron density at the carboxylate,a nd lower its E 1/2 ox . Pleasingly,t he addition of am ethyl (2b)a nd ap henyl (2c) group lowered the E 1/2 ox by 0.25 and 0.3 V, respectively.More importantly,t he substrate 2d,w hich was prepared by condensation of acetophenone with acommercially available hydroxylamine,d isplayed E 1/2 ox = 1.65 V( vs.S CE), av alue which can be reached by some organic photoredox catalysts. [16] Having identified as uitable electrophore,w ep repared the substrate 3a and begun our investigations by looking at developing an oxidative hydroimination (Scheme 3). To accomplish this goal, we selected the methyl acridinium perchlorate 4 (E* 1/2 =+2.06 Vv s. SCE) [3c, 17] as the photoredox catalyst and evaluated ar ange of solvents and bases under blue LED irradiation. Pleasingly,inthe presence of an aryl-disulfide as H-atom relay catalyst, [18] and 2,6-lutidine as the base, 5a was obtained in 81 %yield, which, to the best of our knowledge,r epresents the first fully catalytic radical hydroimination of olefins.C ontrol experiments established the requirement for 4 and continuous irradiation and the calculated quantum yield [19] F = 0.09 supports the photoredox nature of the process. [ We next turned our attention to defining the capacity for imino functionalization in the presence of external coupling partners.B yemploying NCS,d iethyl-bromo-malonate,N IS, and NFSI, we developed olefin imino halogenations in excellent to good yield (5b-e). Interestingly,t he use of Selectfluor as ar adical fluorinating agent [20] provided 5e in considerably lower yield. Imino iodinations and imino fluorinations have not been reported previously.T he use of arylsulfonyl azide and DEAD as coupling partners enabled the preparation of substrates with two nitrogen-based functional groups in different oxidation states [imine and azide (5f)o rh ydrazine (5g)] and with orthogonal reactivity.T his aspect opens the potential for further functionalization and application in click chemistry ligation. As thioether motifs are frequently found in pharmaceutical and agrochemical compounds, [21] we evaluated their introduction as part of our strategy.P leasingly,w ew ere able to form the imines 5h,i which incorporated SÀCF 3 and SÀPh groups.I nt his case, CsOBz [12h] proved to be uniquely effective in promoting good conversions.T he formation of 5h is particularly relevant because the introduction of the SÀCF 3 group in organic molecules is as ignificant but challenging task in medicinal chemistry. [22] This reactivity was also successfully extended to the introduction of aS ef unctionality (5j). Finally,w e evaluated this approach for the installation of organic groups based on the cascade formation of C À Na nd C À C bonds.Byusing aMichael acceptor,weobtained the product 5k,w hich represents ar adical imino Michael cascade. Furthermore,b yu sing hypervalent IBX reagents, [23] we have developed unprecedented radical imino cyanation (5l), olefination (5m), and alkynylation (5n)p rocesses.T he successful formation of 5l is interesting, as recent literature reports described the cyanating reagent to be able to react with only C(sp 3 )-radicals a to ah eteroatom. [12i] Ther eagent used for the formation of 5m [24] has not been used before in radical manifolds,and makes this example the first use of IBX reagents as vinylation partners in free-radical processes.
Regarding the optimization of these 14 different imino functionalizations,wewere able to use 4 as the photocatalyst across all the systems,b ut the base,s olvent, and reaction concentration had to be adjusted to maximize the reaction efficiency.Ingeneral, we have found that the inorganic bases Cs 2 CO 3 and K 2 CO 3 ,a nd the solvents CH 2 Cl 2 and toluene, provided the best results. [16] We have also conducted additional electrochemical Stern-Volmer and quantum yield (F) studies,w hich corroborate our working hypothesis for the reaction mechanism. [16] According to these studies we believe that, in general, the iminofunctionalizations proceed by Scheme 4. Reaction scope. See Scheme 3f or solvent and base used in each imino functionalization reaction. [17] as ingle photoredox catalytic cycles tarting with the SET oxidation of 3a.H owever,i nt he case of the IBX reagents (imino-cyanation 5l and alkynylation 5m), the quantum yields 2 < F < 5s uggest that very short-lived productive radical-chain processes might be operating together with the main photoredox manifold. It is worth mentioning, that none of these processes provided the desired products using our first generation iminyl radical precursors.T his outcome clearly highlights the importance of polar effects and redox balance in the development of cascade reactions of nitrogen radicals.
Having developed ap latform for the divergent assembly of poly-functionalized nitrogen heterocycles,w ee xpanded the reaction scope.A ss hown in Scheme 4, we were able to implement substrates containing terminal olefins (3b), which lead, via the intermediacyofaprimary radical, to the imines 6a-h.Pyridine substituents were tolerated, thus providing the nicotine analogues 6i-k.S ubstrates containing disubstituted olefins (3d,e)w ere also successfully engaged, thus proving that the strategy is amenable to the formation and functionalization of secondary benzylic (7a-e)a nd a-ester (7f-j) radicals.B ye mbedding the olefin into ac ycle( 3f), we generated polyfunctionalized bicyclic molecules in good yield and good to moderate diastereoselectivity,t hus favoring the all-syn isomer (7j-q). We also expanded the scope of coupling partners in terms of Michael acceptors (8a,b)a nd IBX reagents for imino olefination (8c)a nd alkynylation (8d,e). Substrates prepared from a-ketoacids could also be employed leading to the formation of proline-like products (8f-i). Lastly,this radical platform was evaluated as an enabling tool for the late-stage imino functionalization of biologically active molecules.T herefore,w et ested the structurally complex and densely functionalized morphane derivative thevinone (9), [25] which was successfully engaged in the formation of products derived from imino azidation (10 a), imino amination (10 b), and imino selenation (10 c)o ft he olefin moiety.
In conclusion, we have developed ag eneral method for the fast and divergent assembly of polyfunctionalized nitrogen heterocycles.T he reaction involves the organo-photoredox generation of iminyl radicals by oxidative SET of at raceless electrophore and as ubsequent cyclization/functionalization cascade with abroad range of SOMOphiles.This strategy generates useful building blocks in asingle step,and its application to an umber of more complex examples highlights its broad applicability.