The recent achievements of redox-neutral radicalC–C cross-coupling enabled by visible-light
Jin Xie, *a Hongming Jinb and A. Stephen K. Hashmi bc
Visible-light photoredox catalysis has been esteemed as one sustainable and attractive synthetic tool.
In the past four years, a new yet challenging trend, visible-light-driven redox-neutral radical C–C cross-
coupling involving putative radical intermediates, has been booming rapidly. Its advent brings a powerful
platform to achieve non-classical C–C connections, and should lead to fundamental changes in retro-
synthetic analysis. In this tutorial review, we highlight the recent achievements of visible-light-mediated
redox-neutral radical C(sp3)–C(sp2), C(sp3)–C(sp), and C(sp3)–C(sp3) bond formation, opening a new
window for C–C cross-coupling through the photoredox electron shuttling cycle between two coupling
partners. While radical–radical coupling steered by the persistent radical effect was proposed as a rational
explanation for the redox-neutral photoredox events, alternative kinetically driven chain propagation
and radical addition pathways cannot be ruled out. This tutorial review aims to highlight the recent
achievements of photoredox-neutral radical C–C coupling in synthetic chemistry.
Key learning points(1) The significant advancements of redox-neutral C–C coupling by means of photoredox catalysis.(2) The persistent radical effects for radical recombination.(3) Redox-neutral construction of C(sp3)–C(sp2), C(sp3)–C(sp) and C(sp3)–C(sp3) bonds.(4) Stereocontrolled C–C bond formation.(5) Radical C(sp3)–H bond functionalization.
1 Introduction
Solar energy is a unique and renewable natural source ofenergy.1 Chemists aim at using sunlight or visible-light to drivechemical reactions, with the potential to decrease pollution andenergy consumption to a greater extent and to develop newreaction modes. However, most organic compounds cannotdirectly absorb visible-light (l4 400 nm) efficiently. Consequently,visible-light photoredox catalysis with a catalytic amount of photo-sensitizers has become particularly popular.2 In the past decade,we have witnessed a blooming development of visible-light photo-redox catalysis for the concise construction of molecular architec-tures with different photosensitizers (Scheme 1).3–8 Inarguably,this emerging strategy is one of the most efficient and sustainable
methods to generate radical intermediates from a great number ofreadily available starting materials, and can complement tradi-tional radical chemistry which heavily depends on the use of toxic
Scheme 1 Representative photosensitizers.a Jiangsu Key Laboratory of Advanced Organic Materials, School of Chemistry and
Chemical Engineering, Nanjing University, Nanjing 210023, China.
organotin reagents, radical initiators and UV light (l o 300 nm).9
To date, several reviews have summarized the recent applicationsof visible-light photocatalysis in organic synthesis.10–14 In thistutorial review, we mainly discuss the achievements of visible-light-mediated redox-neutral radical C–C cross-coupling in thepast four years. To the community it offers a new strategy toconstruct targeted molecules beyond transition-metal-catalyzedclassical C–C coupling technologies.
A general mechanistic skeleton for light-induced redox-neutralC–C coupling is depicted in Scheme 2. In general, based on the
redox potential of the excited state photocatalyst, either reductivequenching with electron donors (amines, thiols, carboxylates etc.)or oxidative quenching with electron acceptors (R-X, 1,4-dicyano-benzene, high valent metal complexes etc.) can initiate thephotoredox cycle. In some cases, the combination of anothercatalytic cycle (organocatalysis or transition-metal catalysis)with photoredox catalysis is necessary for a successful coupling.In each case, double single electron transfer (SET) enableselectron shuttling to photocatalysts. It is noteworthy that thephotoredox electron shuttling can avoid using external sacrificialoxidants and reductants, offering a promising bond-formingstrategy. To account for the photoredox cycle, a plausibleradical–radical coupling based on the thermodynamic potentialsof the ground and the excited state photocatalyst and putativeradical intermediates was proposed for photoredox-neutral C–Ccoupling events. While these thermodynamic data sound reason-able for closed catalytic cycles, kinetically driven radical chainpropagation or radical addition/elimination pathways may alsoplay a determinative role.15 For example, the generation of
Scheme 2 The general mechanistic consideration.
Hongming Jin
Hongming Jin was born in Anhui(China) in 1989 and received hisbachelor’s degree from AnhuiNormal University in 2011 anda master’s degree from NanjingUniversity in 2014. He iscurrently a third-year PhD candi-date in the group of ProfessorA. S. K. Hashmi at HeidelbergUniversity. His recent contribu-tions are focused on gold-catalyzedtandem reactions for heterocyclesynthesis. A. Stephen K. Hashmi
Prof. Dr A. Stephen K. Hashmistudied chemistry at LMU Munich,where he obtained his diploma andPhD with Prof. G. Szeimies in thefield of nickel- and iron-catalysedcross coupling of strained organiccompounds. His postdoctoralresearch with Prof. B. M. Trostat Stanford University coveredtransition metal-catalysed enzymemetathesis. After his habilitationon enantiomerically pure organo-palladium compounds andpalladium-catalysed conversions
of allenes with Prof. J. Mulzer at the FU Berlin, the JWG-UniversityFrankfurt and the University of Vienna, in 1998 he was awarded aHeisenberg fellowship of the DFG for a proposal on gold-catalysedreactions for organic synthesis – still a major focus of the group. Hisnext stations were at the University of Tasmania in 1999, andMarburg University during 1999–2000; in 2001 he was appointedProfessor for Organic Chemistry at Stuttgart University, and since2007, he has occupied a chair for Organic Chemistry at HeidelbergUniversity. Since 2013 he has been Vice President for Research andStructure, since 2016 Vice President for Research and Transfer atHeidelberg University.
Jin Xie
Dr Jin Xie was born in Chongqing,China, in 1985. He received hisbachelor’s degree from NortheastForestry University in 2008, anda PhD degree in 2013 fromNanjing University, working underthe direction of Prof. ChengjianZhu. From 2014 to 2017, he wasa postdoctoral research associatein the group of Prof. A. S. K.Hashmi at Heidelberg University.In 2017, he came back to NanjingUniversity to start his inde-pendent career. His current
research interests lie in transition-metal catalysis, biomimeticradical transformations and homogeneous gold catalysis.
nucleophilic radical B holds potential to add to electron accep-tor AX (Path B), or the resulting intermediate can undergoradical propagation with the starting material BH to affordproducts and initiate another new catalytic cycle (Path C) orseveral competing pathways work concurrently. Herein, weask the readers to be aware of the latter overlooked pathways.Of note, the visible-light-mediated redox-neutral C–C couplinginarguably represents a mild and efficient bond-formationkit-tool for synthetic manipulations. The future focus on thedetailed mechanistic study of photoredox-neutral C–C couplingwill further help the chemist to entirely predict a possiblereaction.
Notoriously, owing to the high reactivity of open-shellradicals, controllable radical–radical heterocoupling is a highlychallenging task in organic synthesis. With profuse efforts, thecross-coupling selectivity of two radicals can be steered by thepersistent radical effect (PRE). The PRE is based on a kineticeffect; if one ‘‘persistent’’ (more long-lived, radical A) and one‘‘transient’’ (more short-lived, radical B) species are generatedat equal rates, the persistent radical A is less prone to undergoself-termination to form A–A but the transient radical B canundergo homo-coupling to form the dimer B–B. During theinitial step, the concentrations of both radicals are very low(usually o10�7 M), which enables the homo-coupling of tran-sient radical B and the cross-coupling of radicals A and Bunfavorable. After this period, the concentrations of persistentradical A and transient radical B increase gradually and thenthe high concentration of transient radical B makes its self-termination preferred. Along with the consumption of transientradical B, the concentration of persistent radical A continues toslowly increase. When the concentration of persistent radical Aturns out to be much higher than that of transient radical B, thenewly generated radical B is subjected to an increased con-centration of persistent radical A and then rapid formation ofA–B becomes much more favourable than the homo-couplingof radical B (Scheme 3). We would like to provide two com-prehensive reviews for further reading about the PRE.16,17
Although the PRE principle was initially discovered with per-sistent radicals, this kinetic phenomenon may also be suitablefor the selective heterocoupling between two transient radicalsif they have significantly different lifetimes. Indeed, manyvisible-light-mediated redox-neutral C–C coupling reactionsmay proceed via other pathways (e.g., radical chain propagation)rather than the proposed radical–radical coupling. We would
like to witness future work on possible mechanistic studies bycombination of experiments and theoreticians.
2 Redox-neutral C(sp3)–C(sp2)coupling
In 2011, MacMillan’s group developed an early and rare example ofvisible-light-mediated radical–radical heterocoupling of a-aminoalkylradicals and arene radical anions.18 It represents the first redox-neutral a-amino C(sp3)–H arylation protocol with broad applications(Scheme 4). Both aromatic and aliphatic tertiary amines areuniformly a-arylated. In particular, several different types ofelectron-deficient arenes and heteroaromatics can be applied.As illustrated in Scheme 4, oxidative quenching of *Ir(ppy)3 pro-duces a long-lived arene radical anion 7 and a strong oxidantIr(IV) [E1/2(IrIV/IrIII) = +0.77 V versus SCE]. The formed Ir(IV) speciesthen undergoes an SET from tertiary amines and subsequentlygenerates transient a-aminoalkyl radical 6 by deprotonation ofradical cation 5. The cross-coupling of a-aminoalkyl radical 6with arene radical anion 7 leads to the desired product 4 byexpulsion of a cyanide group. One major drawback of thisprotocol is the generation of toxic cyanide salts. However, thislimitation can be addressed by dual photoredox and nickelcatalysis,19 allowing for a-arylation of amines with aromatichalides.
The benzylic amine is a privileged structural motif in bio-logically important reagents. To further enrich the access tothis class of compounds, MacMillan and co-workers expandedthis radical heterocoupling scope, by switching tertiary aminesto commercially available a-amino acids 9 (Scheme 5).20 Thesuccessful application of abundant and inexpensive a-aminoacids represents a practical route to many structurally diversebenzylic amines. Its excellent chemoselectivity was explainedaccording to the PRE by the authors, in which a-amino radical
Scheme 3 The general synopsis of radical–radical cross-coupling. Scheme 4 a-Amino arylation with electron-deficient arenes.
12 is a highly reactive radical and arene radical anion 13 is amore stable radical.
The selective functionalization of C(sp3)–H bonds consti-tutes the state-of-the-art in synthetic chemistry.21 However,visible light-mediated C(sp3)–H bond functionalizations areusually limited to the a-C(sp3)–H bond of amines. Therefore,the predictable C(sp3)–H functionalization of other classes ofsubstrates beyond amines is a worthwhile objective. On thebasis of the bond dissociation energies (BDEs) of differentkinds of C–H bonds (Scheme 6), selective functionalization ofallylic and benzylic C–H bonds (relatively low C–H bond BDE)has been realized.
In 2013, a dual photoredox and organocatalytic arylation ofbenzylic C(sp3)–H bonds was reported (Scheme 7).23 The mainadvance of this protocol is proton-coupled electron transfer(PCET) oxidation of a thiol organocatalyst with an excited-statephotocatalyst to generate transient thiyl radical 19, which ishighly reactive towards hydrogen atom abstraction from weakbenzylic C–H bonds. The low C–H bond BDE and high hydro-gen atom transfer (HAT) rate constant are two crucial factors forsuccess. Under the standard conditions, both benzylic ethersand benzylic alcohols are efficient coupling partners, deliveringthe desired radical heterocoupling products 17 in good yields.Intriguingly, for benzylic alcohol substrates, the additive octanalhas two distinct functions: formation of a hemiacetal inter-mediate and sequestering of the CN� anion. In the absence ofoctanal, benzylic alcohol was oxidized to benzaldehyde underthe same conditions. This redox-neutral radical C–C couplingprovides a new route to a series of important diarylmethylalkylethers.
The allylic C–H bond has a C–H BDE comparable to or evenlower than that of the benzylic C–H bond (allylic C–H bond ofcyclohexene at 83.2 kcal mol�1 versus benzylic C–H of ethyl-benzene at 85.4 kcal mol�1). As a consequence, the photoredox
C(sp3)–H bond arylation was expanded to various allylic C–Hbonds.24 The synergistic organocatalysis and photocatalysiscompleted a broad allylic C–H bond scope. As illustrated inScheme 8, a wide array of useful functional groups was toleratedand moderate to good yields were achieved. In this redox-neutralcross-coupling reaction, the PRE was proposed to control itsselectivity between the ‘‘persistent’’ arene radical anion and the‘‘transient’’ allyl radical. Although ethylbenzene was a goodcoupling partner (69%), the competition experiment with equiva-lent ethylbenzene and cyclohexene gave a surprising result—onlyallylic arylation product. The authors envisioned that the weakerBDE and more hydridic allylic C(sp3)–H bonds would benefit theH-atom abstraction to thiyl radicals.
In the past decade, organocatalysis with a catalytic amountof primary or secondary amines was established as a newopportunity to enantioselectively functionalize the a-C(sp3)–Hbonds of aldehydes and ketones.25 However, organocatalyticb-C(sp3)–H bond functionalization of carbonyl compoundsremains elusive. Based on their success with the a-arylationof tertiary amines, the MacMillan group introduced a novel5pe� activation mode into carbonyl compounds, and b-C(sp3)–Hbond arylation of aliphatic aldehydes and cycloketones wasaccomplished with electron-deficient arenes (Scheme 9).26 Itoffers an intriguing b-monoarylation of aliphatic aldehydes andcycloketones under mild conditions via synergistic photocata-lysis and aminocatalysis.
When irradiated with visible light, an oxidative quenchingof excited state *Ir(ppy)3 (a strong reductant, E1/2(IrIV/*IrIII) =�1.73 V versus SCE) with electron-deficient arenes can produceIr(IV) species [a strong oxidant, E1/2(IrIV/IrIII) = +0.77 V versusSCE] and an arene radical anion 32 (Scheme 10). The enamineintermediate 29, generated in situ from aldehydes and ketones,undergoes an SET process with the Ir(IV) species to produceradical cation 30, which can be deprotonated to give the 5pe�
intermediate 31. The radical–radical heterocoupling of 31 and32 finally leads to the desired b-arylation products 27 or 28.Interestingly, only b-monoarylation products were formed.
The authors speculated that the produced adducts 27 and 28failed to undergo the second b-arylation reaction mainly becausethe corresponding b-enaminyl radicals had lower nucleophilicity.With a chiral aminocatalyst (cinchona-derived organocatalyst),moderate enantioselectivity can be achieved (50% ee). Thecombination of organocatalysis and photoredox catalysisinarguably brings us a new radical–radical coupling blossom.It can be expected that the electron-withdrawing arene radicalanions should selectively couple with a lot of transient carbon-centered radicals in line with the PRE principle.
The monofluoroalkenyl group is a very valuable structuralmotif for potential application in synthetic chemistry, medicinalchemistry and high-performance materials. Very recently, Hashmiand co-workers reported a photoredox-neutral strategy accessinga-monofluoroalkenylated tertiary amines (Scheme 11).27 It repre-sents the first dual C(sp3)–H bond and C(sp2)–F bond function-alization by visible-light photoredox catalysis. Aromatic andaliphatic tertiary amines and different gem-difluoroalkenes wereapplied to construct valuable tetra-substituted monofluoroalkenes
Scheme 9 b-C(sp3)–H bond monoarylation of aliphatic aldehydes andcycloketones.
Scheme 10 Proposed mechanism for b-C(sp3)–H bond monoarylation. Scheme 11 Late-stage a-amino monofluoroalkenylation.
with satisfactory results. Its mild reaction conditions, broadsubstrate scope, excellent functional group compatibility andgood selectivity make this protocol very promising. Its syntheticrobustness was convincingly established by late-stage mono-fluoroalkenylation of some top-selling drugs and complex mole-cules such as (+)-diltiazem, citalopram, rosiglitazone, venlafaxineand dihydroartemisinin (Scheme 11). Although a radical–radicalcoupling model was proposed for this monofluoroalkenylationaccording to initiatory mechanistic studies, alternative pathwayscannot be ruled out.
Very recently, Jamison and co-workers developed an efficientamino acid synthesis protocol (Scheme 12).28 Owing to thehigher reduction potential of CO2 (E0 = �2.21 V versus SCE)than typical photocatalysts, they finally used p-terphenyl 38 as ametal-free photosensitizer (E0 = �2.63 V vs. SCE), and designed aredox-neutral carboxylation of tertiary amines using a photoredox-catalysed continuous flow technique. A variety of amines can bedirectly converted into valuable a-amino acids. The authorsproposed a radical–radical coupling of an a-aminoalkyl radicaland a CO2 radical anion which is formed via single electronreduction of CO2 under photoredox conditions.
3 Redox-neutral C(sp3)–C(sp)coupling
In 2015, Hashmi and co-workers developed the first sunlight-mediated, gold-catalyzed a-amino alkynylation of tertiary aliphaticamines with 1-iodoalkynes.29 As shown in Scheme 13, it has abroad substrate scope with regard to both coupling partners.Although the known A3 coupling and oxidative a-amino alkylationcan get access to propargylic amine products, this developedredox-neutral C(sp3)–H alkynylation protocol showcases its syn-thetic advantages. Some products bearing versatile functionalgroups such as –CHO, –PhCH2O and –OH, cannot be obtainedwith the previously known alkynylation methods. Moreover, usingthis strategy, the direct alkynylation of important drug citalopramcan be easily achieved.
4 Redox-neutral C(sp3)–C(sp3)coupling
The previous routes to b-amino ethers usually need multiplesteps with the use of sensitive organometallic reagents. As a
follow-up work in dual photocatalysis and organocatalysis,visible-light-induced redox-neutral C–C coupling of benzylicether and imines affords a compelling access to b-amino etherswhich represent an important class of structural units in bio-logically important compounds (Scheme 14).30 Under the opti-mized reaction conditions, a diverse range of benzylic ethers 45and Schiff bases 46 were examined. Its good functional grouptolerance and simple operation delivers a mild and efficientprotocol for the synthesis of b-amino ethers. It is no doubt thatthe development of enantioselective version will be of greatimportance in future.
Similarly, an unprecedented b-(sp3)-H aminoalkylation ofcycloketones with Schiff base was developed (Scheme 15).31
Although the substrate scope of the b-Mannich reaction is limitedto only cyclic ketones, it opens a new approach to construct1,4-aminoketones under mild conditions. To further increasethe viability of this new protocol, a three component couplingof cyclohexanone, benzaldehyde and anisidine was explored(Scheme 16). Interestingly, in the presence of DABCO, the
b-Mannich product 55 was exclusively formed in 70% yield,while the classical a-Mannich product 54 was obtained as themajor product (54%) along with minor b-Mannich product 55(19%) when DABCO was omitted from the standard reactionconditions. The corresponding asymmetric version allowingaccess to nitrogen-bearing tetrasubstituted carbon centerswould be very attractive for medicinal applications.
In early 2015 the a-allylation of tertiary and secondaryamines was reported by Xiao and Lu through dual photoredoxand palladium catalysis (Scheme 17).32 Both N-Ar tetrahydro-isoquinolines and N-Ar a-amino ketones were good couplingpartners, while the other N-Ar tertiary amines exhibited lessreactivity. The authors performed an electron paramagneticresonance (EPR) spin trapping experiment with DMPO to deter-mine a-aminoalkyl radicals. Although a-aminoalkyl radicals areproposed as important reactive intermediates in photoredoxcatalysis, to the best of our knowledge, this is the first case todirectly verify it by EPR. Its practice was also underlined by theeasy scale up to the gram scale.
As illustrated in Scheme 18, Xiao and co-workers appliedthis allylation strategy as a key step (61 to 62) to the formalsynthesis of potential anticancer candidate 8-oxoprotoberberinederivative 65. Although the radical–radical coupling was proposedas a likely pathway for photoredox-neutral allylation, radical addi-tion of a-aminoalkyl radicals to organopalladium was also possible.
In 2014, Tunge and co-workers disclosed an interestingintramolecular redox-neutral decarboxylative allylation.33 Asshown in Scheme 19, the dual photoredox and palladium catalysiscombination make the decarboxylative allylation of intractablea-amino and phenylacetic allyl ethers feasible without the use ofany additives. To avoid the preparation of allylic esters, Tunge andco-workers also applied this dual catalysis tactic into the inter-molecular radical decarboxylative allylation of phenylacetic acidswith allyl methyl carbonates as allyl sources. From the perspectiveof one organic chemist, the moderate yields don’t compromise itssynthetic value, because it offers a rare example of a site-specificdecarboxylative radical allylation protocol, complementing Xiao’sa-allylation of amines.32
In the next year, Tunge and co-workers thoroughly expandedthe substrate scope of this dual photoredox and palladiumcatalysis system.34 Photoredox-neutral decarboxylative allyl-ation of N-protected a-amino allyl esters 70 were accomplished(Scheme 20).
Gratifyingly, the mechanistic study indicates that the reactionpathway is closely related to the stability of a-aminoalkyl radicals.
Scheme 15 b-C(sp3)–H aminoalkylation of cycloketones with Schiff bases.
Scheme 16 The selectivity of three component coupling of cyclohexa-none, benzaldehyde and anisidine.
Scheme 17 The cross-coupling of an a-aminoalkyl radical and an allyradical.
Scheme 18 Formal synthesis of an 8-oxoprotoberberine derivative.
Scheme 19 Decarboxylative allylation of a-amino and phenylacetic allylethers.
It was found that the stable a-aminoalkyl radical favours theradical–radical heterocoupling steered by the PRE principle,while the less stable a-aminoalkyl radical would be prone toundergo a Pd(III)-mediated reductive elimination pathway. Forexample, as depicted in Scheme 21, among starting materials70A–C, both 70A and 70B can form relatively more stablea-amino radicals than 70C. Therefore, a-amino radicals derivedfrom 70A and 70B are preferable for the radical heterocouplingwith allyl radicals, and the less stable radical derived from 70Cprefers to add to the Pd(II)-p-allyl species to form a Pd(III)-complex 73. Furthermore, with a chiral palladium catalyst itwas found that for the less stable radical, moderate enantio-selectivity (50% ee) was achieved. Conversely, the more stableradicals led to only racemic products or very low enantioselec-tivity. Much better enantioselectivity for starting material 70C isevidence for the formation of Pd(III)-bound intermediate 73which can be tuned by the chiral ligand.
At the end of 2015, Ooi and co-workers reported an unpre-cedented, visible-light-driven redox-neutral enantioselectiveC–C coupling of N-arylaminomethanes and N-sulfonyl aldimines,allowing for the synthesis of enantioenriched 1,2-diamines(Scheme 22).35 Under photoredox conditions, a single electronoxidation of N-arylaminomethanes 75 and a single electronreduction of aldimines 74 are known to generate a-aminoalkyl
radicals and aldimine radical anions, respectively. Based ontheir rich experience in chiral ionic Brønsted acid catalysis,36
they hypothesized that the chiral tetraaminophosphoniumion-type catalyst 77 would form a chiral ion-pair 78 with thephotoredox-generated aldimine radical anion. Finally, the radical–radical heterocoupling of chiral ion pair 78 with a-aminomethylradical 79 delivers the coupling products in good yields withexcellent enantioselectivities.
To demonstrate the synthetic value of this new strategy, arapid synthesis of chiral benzopiperazine derivatives was illus-trated (Scheme 23). First, a radical–radical heterocoupling anda subsequent intramolecular amination furnished 83 in 63%total yield (two steps) and with 97% ee. Although currently thesubstrate scope is limited to N-arylaminomethanes only, webelieve that this piece of work represents a significant advancein the photoredox-neutral enantioselective C–C coupling andhighlights the potential of visible-light photoredox catalysis inthis area.
Almost simultaneously, Meggers et al. reported a similarprocess; a visible-light-induced, bifunctional chiral Ir(III)-typephotocatalyst enabled redox-neutral stereocontrolled C–C cross-coupling of trifluoromethylated ketones and tertiary amines.37
This protocol allows for the direct construction of chiral 1,2-amino alcohols from easily available trifluoromethyl ketones andtertiary amines (Scheme 24). It is noteworthy that the other
Scheme 20 Radical decarboxylative allylation of N-protected a-aminoallyl esters.
Scheme 21 Mechanistic study of radical decarboxylative allylation.
Scheme 22 Redox-neutral asymmetric a-coupling of N-arylamino-methanes with aldimines.
Scheme 23 Asymmetric synthesis of benzopiperazine derivatives.
commercially available photocatalysts (1a and 1d) are ineffective,thus stressing the unique property of this kind of bifunctionalIr-based photocatalyst. A variety of enantioenriched 1,2-aminoalcohols were constructed under mild reaction conditions. TheIr-bound radical pair was regarded as one key combination forits successful asymmetric induction. Later, the employment ofa-silylamines instead of tertiary amines allows for a broaderamine scope.38 Interestingly, almost at the same time as thosereported by Ooi and Megger, racemic but more general exampleswere documented by Rueping39 and Xiao40 respectively.
The robustness of visible-light-mediated redox-neutral intra-molecular C–C coupling of ketones and a-C(sp3)–H of tertiaryamines was achieved by Zhu and co-workers, providing a conciseroute to a wide range of 4-6-membered N-heterocycles 88(Scheme 25).41 The reaction is brilliant with excellent selectivityand good synthetic yields. It enriches our toolbox for the mildsynthesis of N-heterocycles bearing one quaternary carbon centre.
Much more recently, Cheng and co-workers developed anelegant photoredox-neutral C(sp3)–C(sp3) coupling accessing
congested ketones (Scheme 26).42 A wide range of 4-alkylHantzsch esters 90 were used as efficient tertiary alkyl radicalsources by means of visible-light photoredox catalysis. Its promis-ing synthetic values were showcased by the concise connectionof two contiguous quaternary carbon centres and the formalsynthesis of hydroxysteroid dehydrogenase inhibitors. In thepast four years, we have witnessed the rapid growth of redox-neutral C–C coupling in organic synthesis and some examplesare not discussed in detail.43–46
5 Conclusions
In conclusion, the recent achievements of visible-light-inducedredox-neutral radical C–C cross-coupling during the past fouryears have been summarized. Its advent offers a powerful C–Ccoupling strategy under mild reaction conditions, which is beyondwhat traditional coupling chemistry would bring to syntheticchemistry. Inarguably, it will lead to fundamental changesin non-classical C–C disconnection in retrosynthesis analysis.Perhaps more importantly, it should stimulate forward develop-ment of visible-light-driven, redox-neutral enantioselectiveC–C coupling, which is almost underdeveloped yet deceptivelychallenging. At the moment, reaction discovery is outpacing theexacting analysis of individual reaction mechanisms,15,47–50
and most literature reviewed in this tutorial review can onlypropose a tentative radical–radical coupling pathway based onthe thermodynamic potentials of the ground and the excited statephotocatalyst and putative radical intermediates. Accordingly, it isa good chance for the readers to reconsider the possible mecha-nism. Nonetheless, given the significant advances of photoredox-neutral radical C–C coupling in synthetic chemistry, we believethat this timely tutorial review will draw great enthusiasm ofchemists in the near future.
Abbreviations
PRE Persistent radical effectSET Single electron transferSSCE Sodium saturated calomel electrodeSCE Saturated calomel electrodeBDE Bond dissociation energyPCET Proton-coupled electron transferHAT Hydrogen atom transferDMA N,N-DimethylacetamideDABCO 1,4-Diazabicyclo[2.2.2]octaneEPR Electron paramagnetic resonanceDMSO Dimethyl sulfoxide
Scheme 24 Stereocontrolled C–C coupling for chiral 1,2-aminoalcohols.
Scheme 25 Intramolecular redox-neutral C–C coupling for the synthesisof 4–6-membered N-heterocycles.
Scheme 26 Intermolecular radical–radical C(sp3)–C(sp3) coupling for thesynthesis of congested ketones.
