“This  document  is  the  Accepted  Manuscript  version  of  a  Published  Work  that  appeared  in  final  form  in   the  American  Society  copyright  ©  American  Chemical  Society  after  peer  review  and  technical  editing  by  the   publisher.  To  access  the  final  edited  and  published  work  see  [insert  ACS  Articles  on  Request  authordirectedlink  to  Pub-­‐ lished  Work,  see  http://pubs.acs.org/doi/abs/10.1021/jacs.6b02867     Pd-Catalyzed C(sp3)-H Functionalization/Carbenoid Insertion: All-Carbon Quaternary Centers via Multiple C–C Bond-Formation   Álvaro Gutiérrez-Bonet,† Francisco Juliá-Hernández,† Beatriz de Luis† and Ruben Martin†§* †Institute of Chemical Research of Catalonia (ICIQ), Av. Països Catalans 16, 43007 Tarragona, Spain § Catalan Institution for Research and Advanced Studies (ICREA), Passeig Lluïs Companys, 23, 08010, Barcelona, Spain Supporting Information Placeholder 3 ABSTRACT: A Pd-catalyzed C(sp )-H functionaliza- tion/carbenoid insertion is described. The method allows for the rapid synthesis of bicyclic frameworks, generating all-carbon quaternary centers via multiple C–C bond-formations in a straightforward manner. Over the last few years, there has been a growing consensus that C–H functionalization has profoundly changed the landscape of organic synthesis while establishing new paradigms in retrosynthetic analysis.1 While spectacular advances have been realized, this area of expertise primarily relies on the utilization of directing groups, particularly via C(sp2)–H functionalization. Indeed, a close inspection into the literature data reveals that the preparation of all-carbon quaternary centers2 via C(sp3)–H functionalization in the absence of directing groups still remains rather elusive.3,4 Scheme 1. C(sp3)–H Functionalization/Carbenoid Insertion. R H N2 + R1 H + X X =Br,I,Cl catalyst R2 R1 R2 1 Secondary (R=H) & tertiary centers (known) catalyst N2 R1 R H path a R2 R1,R2!H R1 path b R2 Single C–C bond-formation Multiple C–C bond-formation All-carbon quaternary centers (unknown) While originally designed for cyclopropanation events, carbenoid species have shown to be superb synthons in a myriad of relevant transformations.5 Indeed, these reagents have successfully been employed in C–H functionalization without the need for directing groups, allowing for installing secondary or tertiary carbon centers via single C–C bond-formation (Scheme 1, path a).6 To the best of our knowledge, all-carbon quaternary stereocenters derived from the corresponding carbenoid species are beyond reach in C–H functionalization.7,8 Undoubtedly, the ability to promote multiple C–C bondformations initiated by C(sp3)–H functionalization while installing all-carbon quaternary centers would be of particular interest (Scheme 1, path b).9 If successful, such a protocol would not only represent an unconventional, yet powerful, technique for our synthetic arsenal, but also a unique opportunity to improve our ever-growing knowledge in C–H functionalization. However, the difficulty for effecting C(sp3)–H functionalization in the absence of directing groups3 and the inherent propensity of carbenoids towards competitive dimerization5,6 constitute serious drawbacks to be overcome. To such end, we hypothesized that the intermediacy of in situ generated Pd-I10 via C(sp3)–H functionalization would be critical for success (Scheme 2). At the outset of our investigations, it was unclear whether such scenario could ever be conducted given the known proclivity of Pd-I towards C–C reductive elimination (path b)11,12 or competitive [1,4]-shifts en route to 4 (path a).13 Herein, we report a mild catalytic C(sp3)–H functionalization/carbenoid insertion en route to indanes 3 bearing all-carbon quaternary centers (path c). This protocol is distinguished by a wide scope and excellent chemoselectivity profile, thus constituting a unique tool to rapidly build up molecular complexity. Scheme 2. Intermediacy of Pd-I in C-H Functionalization. R1 R2 Me Me H R1 R2 R H 4 R M path a Single C–C bond-formation X Pd(0)Ln R1 R2 Pd Pd-I L L R1 R2 5 1 (X = Br, Cl) path b R1 R2 Table 1. Optimization of the Reaction Conditions.a Me + 1a 3aa Entry Deviation from the standard conditions Ph CO 2Me Me Me 3aa (%) b 5a 5a (%) b 1 none 93 (80)c 0 2 using L2 as the ligand 36 0 3 using L3 as the ligand 0 0 4 Using L4 as the ligand 83 0 5 path c 4 this work 3 R3 R C(sp3)-H functionalization Multiple C-C bond-formation All-carbon quaternary centers We initiated our study by investigating the reaction of 1a with 2a (Table 1). After considerable experimentation,14 a protocol based on PdCl2(SMe2)2, L1, PivOH and Cs2CO3 in DMF at 80 ºC provided the best results (entry 1). Although the structure of 3aa was evident by NMR spectroscopy, we univocally assigned its structure by comparison with 3aa’ derived from the hydrolysis of 3aa by X-ray crystallography.14 Not surprisingly, the ligand backbone had a critical impact on both reactivity and selectivity (entries 2-6). While the significant lower reactivity of L2 and L3 might suggest an intimate interplay of steric and electronic effects, care must be taken when generalizing this since we found that L4 was equally effective. The use of monodentate phosphines (entries 5 and 6) had a deleterious effect; strikingly, the utilization of PtBu3 resulted in a selectivity switch, obtaining exclusively 5a.11c Similarly, the base and the solvent exerted a profound influence on reactivity (entries 7-10), with toluene favoring the formation of 5a (entry 9). Interestingly, inferior results were found for protocols based on Pd(OAc)2 (entry 11). The higher reactivity of PdCl2(SMe)2 is tentatively attributed to its high solubility; at present, we cannot rule out that Me2S facilitates the reduction to Pd(0) while forming DMSO. Additionally, otherwise related aryl chlorides, iodides and triflate congeners failed to deliver 3aa. As anticipated, control experiments univocally revealed that all parameters were essential for the reaction to occur.14,15 Me PivOH (50 mol%) Cs 2CO3 (1.30 equiv) DMF, 80 ˚C Br N2 R3 R4 2 Pd catalyst PhC(N 2)CO2Me (2a) PdCl 2(SMe 2) 2 (5 mol%) H L1 (7.5 mol%) Using PtBu 3·HBF 4 (15 mol%) as the ligand 0 58 6 Using PCy 3 (15 mol%) as the ligand 0 0 7 Using CsOPiv (1.30 equiv) as the base d 43 0 8 Using CsOAc as the base 38 0 9 Using PhMe instead of DMF 15 73 10 Using DMA instead of DMF 49 0 11 Using 5 mol% Pd(OAc) 2 73 0 PR 2 PR 2 Me Me PPh 2 O R = Cy, L1 R = Ph, L2 PPh 2 O PPh 2 L3 PPh 2 3aa' L4 a 1a (0.10 mmol), 2a (0.18 mmol), PdCl2(SMe2)2 (5 mol%), L1 (7.50 mol%), PivOH (50 mol%), Cs2CO3 (0.13 mmol), DMF (0.25 M) at 80 ºC. b GC yields using o-xylene as standard. c Isolated yield. d No PivOH was added. Prompted by these results, we sought to examine the influence of the carbenoid species (Table 2). As shown, the scope was insensitive to electronic changes at the para and meta positions on the aromatic ring (2f-2l). Likewise, the substitution pattern on the ester motif was inconsequential to the reactivity profile (2a-2c), invariably leading to the targeted products in high yields. The chemoselectivity profile of our protocol is nicely illustrated by the fact that a wide variety of diazoester derivatives bearing aryl halides (2f, 2j and 2m), esters (2e and 2h), ketones (2l) or acetals (2o) were all well accommodated. Notably, nitrogen-containing heterocycles posed no problems (2p). Particularly interesting was the observation that the presence of alkene on the side chain did not interfere, affording 3ad in high yields without traces of intramolecular cyclopropanation being observed in the crude mixtures. Gratifyingly, the diazo compound derived from Isoxepac (2l),16 a nonstereoidal anti-inflammatory drug (NSID), could be employed with equal ease. Notably, this transformation was not limited to diazoester derivatives, as diaryldiazomethanes could also be coupled, albeit in lower yields (2q-r). Unfortunately, donor/donor diazocompounds and monosubstituted carbene precursors could not participate in the targeted reaction, recovering starting material unaltered. Table 2. Scope of Diazo Compounds.a,b Me Me H N2 + Br 1a N2 Ph R R PdCl2(SMe2)2 (5 mol%) L1 (7.5 mol%) PivOH (50 mol%) Me R1 R2 2a-o Cs2CO3 (1.30 equiv) R2 3aa-3ao R1 DMF, 80 ˚C Me R=CO2Me; 80% (2a) N2 N2 R=CO2Bn; 80% (2b) O EtO2C CO2Et R=CO2tBu; 70% (2c) Ph 68% (2e)c O 87% (2d) N2 N2 CO2Et R O CO2Me CO2Me N2 R=F; 75% (2f) R=CF3; 71% (2g) R=CO2Et; 59% (2h) N2 Cl R=OMe; 69% (2i) R=F; 53% (2j) R=OCF3; 63% (2k) N2 O 61% (2l) H R1 63% (2n) N2 N 83% (2p)d 53% (2q)d CF3 tBu Ph CO2Me R=CHO; 79% (3ea) R=CO2Me; 77% (3fa) R OTIPS CO2Et R2 R3 R1 CO2Me 3ba-3ma Ph Me Me F Me Me R 42% (2r)d Ph CO2Me 48% (3ca) Me Me R2N Ph CO2Me R=H; 40% (3ga)d R=Me; 71% (3ha) R R Ph CO2Me 81% (3da) R Me R Ph CO2Me R=H, 40% (3ia)c,d R=Me, 84% (3ja)c,d R F F3C Me Me Ph CO2Me 73% (3ba) 47% (2o) N2 CO2Et Cs2CO3 (1.30 equiv) DMF, 80 ˚C Br Me Me Me O 43% (2m)c N2 + 2a 1b-1l N2 O CO2Et CO2Me PdCl2(SMe2)2 (5 mol%) L1 (7.5 mol%) PivOH (50 mol%) R2 R3 Me Ph CO2Me R = H, 73% (3ka)c,d R = Me, 0% (3ka')c,d CF3 Ph CO2Me R=Me; 70% (3la) R=OMe; 51% (3ma)d a As Table 1 (entry 1), 0.50 mmol scale. b Isolated yields, average of at least two independent runs. c PdCl2(SMe)2 (10 mol%) at 100 ºC. d PdCl2(SMe)2 (10 mol%). a Next, we turned our attention to study the substitution pattern on the aryl halide backbone (Table 3). As shown, the preparative scope was rather general regardless of whether electron-donating or electron-withdrawing groups were present or not. Notably, a variety of aryl fluorides (3da), aldehydes (3ea), esters (3fa), amines (3ga and 3ha) or silyl ethers (3ka) could perfectly be tolerated. Importantly, even free amines could be employed as substrates, albeit in lower yields (3ga). Although the presence of an ortho t-butyl group statistically accelerates the key C(sp3)–H functionalization,17 we found that a variety of ortho substituents other than tbutyl groups could be equally accommodated (3ia-3na). In all cases analyzed, the targeted C(sp3)–H functionalization occurred exclusively at the primary C(sp3)–H bonds of methyl groups, leaving the corresponding methylene positions intact. In line with this notion, no reaction occurred when employing 3ka’. Unfortunately, no diastereoselection was observed in the presence of gem-dimethyl groups (3ia-3ka), even in the presence of bulky silyl or aromatic motifs (3ja-3ka).18 Likewise, tertiary benzylic carbons (R2=H) resulted in β-hydride elimination, even with bulkier mesityl groups. Taken together, the results in Tables 2 and 3 show the prospective impact of our protocol for rapidly preparing indane skeletons bearing all-carbon quaternary centers. Ph CO2Me 75% (3na) Scheme 3. Mechanistic Experiments. Table 3. Scope of Aryl Bromides.a,b As for Table 1 (entry 1), but at 0.50 mmol scale.b Isolated yields, average of at least two independent runs. c 1:1 diastereomeric ratio. d PdCl2(SMe)2 (10 mol%) at 100 ºC. Me Me Cl Pd H 6 79% yield Reductive Me Me elimination 5a DMF, 80 ˚C 99% yield Me 2a/L1 COD PivOH / Cs 2CO3 DMF, 80 ˚C 7 (1) NaOH 99% yield (2) L4 / PhH Me 2a (1.75 equiv) DMF, 80 ˚C Pd 6 L4 99% yield as for Table 1 (entry 1) No reaction Me Me CO 2Me 3aa Ph Next, we decided to gather indirect evidence on the mechanism by examining the reactivity of 1a with PivOD. Interestingly, a non-negligible deuteration at ortho position of 3aa was observed, suggesting that Pd-I (Scheme 2) might coexist in equilibrium with homobenzylic Pd(II) intermediates generated upon protonolysis with PivOD via [1,4]-shift.10c,11c,13,14 Next, we studied the reactivity of the putative metallacycle Pd-I. Following the methodology described by Cámpora,10b we prepared 6 from 7 in high yield (Scheme 3, bottom), which was fully characterized by X-ray structure analysis.14 Interestingly, while 6 rapidly underwent reductive elimination en route to 5a in the absence of 2a,11 3aa was exclusively obtained with 2a (Scheme 3, bottom).19,20 Notably, 3aa was not obtained from 5a, thus ruling out the possibility of a C–C cleavage event. We believe these results reinforce a scenario consisting of Pd-I via concerted metallation-deprotonation from II (Scheme 4).11,21 While Pd-I might coexist in equilibrium with III upon protonolysis with PivOH, a 1,2-insertion of a diazo compound10a,22,23 might generate IV that ultimately delivers the targeted product via reductive elimination. At present, we cannot rule out the intermediacy of V via rapid equilibration with III and Pd-I,24 as traces of cyclopropane derivatives via reductive elimination from V were detected in reactions of aryl bromides possessing bulky groups at the geminal position.25 (3) (4) Scheme 4. Mechanistic Hypothesis. R1 R 2 R1 H R2 H Br Pd L1 II Cs2CO 3/PivOH Br R1 Pd(0)L n R1 R 2 R1 R 2 R1 R 2 R2 +PivOH Pd -PivOH Pd-I L1 (5) III L1 Pd OPiv R 2=Me R1 4 R3 R IV Pd L1 R3 R4 Pd L1 N2 R3 R4 V In conclusion, we have developed a mild and robust Pdcatalyzed C(sp3)-H functionalization/carbenoid insertion event. This technique represents a unique synthetic tool in the C(sp3)–H functionalization arena for building up bicyclic frameworks in which the all-carbon quaternary center is derived from carbenoid species. (6) (7) ASSOCIATED CONTENT Supporting Information. Experimental procedures and spectral data. This material is available free of charge via the Internet at http://pubs.acs.org. (8) AUTHOR INFORMATION Corresponding Author (9) * rmartinromo@iciq.es Funding Sources (10) No competing financial interests have been declared. ACKNOWLEDGMENT We thank ICIQ, the European Research Council (ERC277883), MINECO (CTQ2012-34054 & Severo Ochoa Excellence Accreditation 2014-2018, SEV-2013-0319) and Cellex Foundation for support. Johnson Matthey, Umicore and Nippon Chemical Industrial are acknowledged for a gift of metal & ligand sources. We sincerely thank E. Escudero, E. Martin for X-Ray crystallographic data and Prof. E. Gómez-Bengoa for invaluable theoretical calculations. (11) (12) REFERENCES (1) (2) (a) Gutekunst, W. R.; Baran, P. L. Chem. Soc. Rev. 2011, 40, 1976. (b) Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147. (c) Chen, X.; Engle, K. M.; Wang, D. H.; Yu, J. –Q. Angew. Chem. Int. Ed. 2009, 48, 5094. All-carbon quaternary centers are defined as carbon atoms to which four distinct carbon substituents are attached. (a) Quasdorf, K. W.; Overman, L. E. Nature, 2014, 516, 181. (b) Das, J. P.; Marek, I. Chem. Commun, 2011, 47, 4593. (13) (14) (15) (a) Li, H.; Li, B.-J.; Shi, Z.-J. Catal. Sci. Technol, 2011, 1, 191. (b) Jazzar, R.; Hitce, J.; Renaudat, A.; SofackKreutzer, J.; Baudoin, O. Chem. Eur. J. 2010, 16, 2654. (c) Herrmann, P.; Bach, T. Chem. Soc. Rev. 2011, 40, 2022. (a) Pedroni, J.; Saget, T.; Donets, P. A.; Cramer, N. Chem. Sci. 2015, 6, 5164. (b) Davies, H. M. L.; Hansen, T.; Churchill, M. R. J. Am. Chem. Soc. 2000, 122, 3063. (c) Janody, S.; Jazzar, R.; Comte, A.; Holstein, P. M.; Vors, J. – P.; Ford, M. J.; Baudoin, O. Chem. Eur. J. 2014, 20, 11084. (d) Saget, T.; Cramer, N. Angew. Chem. Int. Ed. 2012, 51, 12842. (e) Martin, N.; Pierre, C.; Davi, M.; Jazzar, R.; Baudoin, O. Chem. Eur. J. 2012, 18, 4480. (f) Holstein, P. M.; Vogler, M.; Karini, P.; Pilet, G.; Clot, E.; Baudoin, O. ACS Catal. 2015, 5, 4300. (g) Yu, S.; Liu, S.; Lan, Y.; Wan, B.; Li, X. J. Am. Chem. Soc. 2015, 137, 1623. (a) Ford, A.; Miel, H.; Ring, A.; Slattery, C. N.; Maguire, A: R.; McKervey, A. Chem. Rev. 2015, 115, 9981. For selected carbenoid insertions into organopalladium intermediates: (b) Kudirka, R.; Devine, S. K. J.; Adams, C. S.; VanVranken, D. L. Angew. Chem. Int. Ed. 2009, 48, 3677. (c) Premachandra, I. D.; Nguyen, T. A.: Shen, C.; Gutman, E. S.; VanVranken, D. L. Org. Lett. 2015, 17, 5464. (d) Zhou, L.; Ye, F.; Zhang, Y.; Wang, J. Org. Lett. 2012, 14, 922. (e) Barluenga, J.; Escribano, M.; Aznar, F.; Valdes, C. Angew. Chem. Int. Ed. 2010, 49, 6856. For selected reviews of C–H functionalization dealing with diazoderivatives: (a) Caballero, A.; Pérez, P. J. Chem. Soc. Rev., 2013, 42, 8809. (b) Davies, H. M. L.; Morton, D. Chem. Soc. Rev. 2011, 40, 1857. For selected syntheses of quaternary center not derived from carbenoids via C–H activation: (a) Fuentes, M. A.; Muñoz, B. K.; Jacob, K.; Vendier, L.; Caballero, A.; Etienne, M.; Pérez, P. J. Chem. Eur. J. 2013, 19, 1327. (b) ref. 4c. While this paper was under preparation, a functionalization of activated C(sp3)–H bonds using expensive Rh catalysts followed by carbenoid insertion was reported: Zhou, B.; Chen, Z.; Yang, Y.; Ai, W.; Tang, H.; Wu, Y.; Zhu, W.; Li, Y. Angew. Chem. Int. Ed. 2015, 54, 12121. For syntheses of not all-carbon quaternary centers via C(sp2)–H/carbenoid insertion: (a) Ye, B.; Cramer, N. Angew. Chem. Int. Ed. 2014, 53, 7896. (b) Hyster, T. K.; Ruhl, K. E.; Rovis, T. J. Am. Chem. Soc. 2013, 135, 5364. For selected stoichiometric transformations of these Pd(II) metallacycles: (a) Cámpora, J.; Palma, P.; del Río, D.; López, J. A.; Valerga, P. Chem. Commun. 2004, 1490. (b) Cámpora, J.; López, J. A.; Palma, P.; del Río, D.; Carmona, E. Inorg. Chem. 2001, 40, 4116. (c) Cámpora, J.; López, J. A.; Palma, P.; Valerga, P.; Spillner, E.; Carmona, E. Angew. Chem. Int. Ed. 1999, 38, 147. (a) Kefalidis, C. E.; Davi, M.; Holstein, P. M.; Clot, E.; Baudoin, O. J. Org. Chem. 2014, 79, 11903. (b) Rousseaux, S.; Davi, M.; Sofack-Kreutzer, J.; Pierre, C.; Kefalidis, C. E.; Clot, E.; Fagnou, K.; Baudoin, O. J. Am. Chem. Soc. 2010, 132, 10706. (c) Chaumontet, M.; Piccardi, R.; Audic, N.; Hitce, J.; Peglion, J. –L.; Clot, E.; Baudoin, O. J. Am. Chem. Soc. 2008, 130, 15157. For preparing otherwise related benzocyclobutenones: (a) Flores-Gaspar, A.; Gutiérrez-Bonet, A.; Martin, R. Org. Lett. 2012, 14, 5234. (b) Álvarez-Bercedo, P.; Flores-Gaspar, A.; Correa, A.; Martin, R. J. Am. Chem. Soc. 2010, 132, 466. (a) Barder, T. E.; Walker, S. D.; Martinelli, J. R.; Buchwald, S. J. Am. Chem. Soc., 2005, 127, 4685. (b) Hoshi, T.; Honma, T.; Mori, A.; Konishi, M.; Sato, T.; Hagiwara, H.; Suzuki, T. J. Org. Chem., 2013, 78, 11513. See Supporting information for details. The use of tosyl hydrazones as carbenoid species resulted in dehalogenation while not observing even traces of 3aa. (16) Scott, J.; Huskinsson, E. C. Rheumatology, 1982, 21, 48 (17) For selected C(sp3)-H functionalization of ortho t-butyl groups: (a) Yan, J.-X.; Li, H.; Liu, X.-W.; Shi, J.-L.; Wang, X.; Shi, Z.-J. Angew. Chem. Int. Ed. 2014, 53, 4945. (b) Lafrance, M.; Gorelsky, S. I.; Fagnou, K. J. Am. Chem. Soc. 2007, 129, 14570. (d) Desai, L. V.; Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc. 2004, 126, 9542. (18) DFT calculations (B3LYP & M06) showed that the energy difference of the transition states leading to the two possible diastereoisomers (IV, Scheme 4) is negligible (0.3-2.3 kcal·mol-1). Additionally, the overall energy barrier for [1,2]-insertion was found to be 1.7-3.3 kcal·mol-1 (ref.14). (19) Although 6-L1 could be isolated and characterized by X-ray crystallography (see ref. 14), its insolubility prevented its characterization by NMR spectroscopy. Still, 6-L1 could be converted into either 5a or 3aa in quantitative yields. (20) 6 and 7 were found to be catalytically competent on the conversion of 1a into 3aa. See ref. 14. R (21) (a) Lapointe, D.; Fagnou, K. Chem. Lett. 2010, 39, 1118. (b) Gorelsky, S. I.; Lapointe, D.; Fagnou, K. J. Am. Chem. Soc. 2008, 130, 10848. (c) ref. 20c. (d) Garcia-Cuadrado, D.; de Mendoza, P.; Braga, A. A. C.; Maseras, F.; Echavarren, A. M. J. Am. Chem. Soc. 2007, 129, 6880. (22) We propose the intermediacy of Pd carbenoid species from Pd-I prior 1,2-insertion at the C(sp3)–C bond en route to IV. Preliminar DFT calculations (B3LYP & M06) revealed that 1,2-insertion at C(sp2)–C was less favorable by 12-15 kcal·mol-1 (ref.14). (23) (a) Hu, F.; Xia, Y.; Ma, C.; Zhang, Y.; Wang, J. Chem. Commun. 2015, 51, 7986. (b) Xia, Y.; Zhang, Y.; Wang, J. ACS Catal. 2013, 3, 2586. (24) (a) Miyashita, A.; Ohyoshi, M.; Shitara, H.; Nohira, H. J. Organomet. Chem. 1988, 338, 103. (b) Jennings P. W.; Johnson L. L. Chem. Rev. 1994, 94, 2241. (25) 0% ee was observed by reacting diethyl 2-diazomalonate with aryl halides containing gem-dimethyl groups. See Supporting information for a mechanistic rationale. Catalytic C(sp3)–H Functionalization / Carbenoid Insertion R3 N2 R2 R1 R1 2 R4 R H Pd catalyst / Ligand R All-carbon quaternary centers Br R4 Multiple C–C Bond-Formations R3 Wide substrate scope 31 examples up to 87% yield R1 via R2 R Pd L Isolated & characterized