Letter Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Ylide-Functionalized Phosphine (YPhos)−Palladium Catalysts: Selective Monoarylation of Alkyl Ketones with Aryl Chlorides Xiao-Qiang Hu,§,† Dominik Lichte,† Ilja Rodstein,‡ Philip Weber,† Ann-Katrin Seitz,† Thorsten Scherpf,‡ Viktoria H. Gessner,*,‡ and Lukas J. Gooßen*,† †
Downloaded via NOTTINGHAM TRENT UNIV on August 30, 2019 at 23:23:25 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
Faculty of Chemistry and Biochemistry, Evonik Chair of Organic Chemistry, Ruhr-Universität Bochum, Universitätsstr. 150, 44801 Bochum, Germany ‡ Faculty of Chemistry and Biochemistry, Chair of Inorganic Chemistry II, Ruhr-Universität Bochum, Universitätsstr. 150, 44801 Bochum, Germany S Supporting Information *
ABSTRACT: Ylide-functionalized phosphine (YPhos) ligands allow the palladium-catalyzed α-arylation of alkyl ketones with aryl chlorides with record setting activity. Using a cyclohexylsubstituted YPhos ligand, a wide range of challenging ketone substrates was efficiently and selectively monoarylated under mild conditions. A newly designed YPhos ligand bearing tert-butyl groups on the coordinating phosphorus atom is already active at room temperature. The synthetic potential was demonstrated by gram-scale reactions and the succinct synthesis of εcaprolactone derivatives. he synthesis of α-aryl ketones is of substantial interest in pharmaceutical research and natural product synthesis.1 Semmelhack et al. pioneered the transition-metal-catalyzed αarylations of ketones in their synthesis of cephalotaxinone.2 Efficient catalysts were later developed featuring electron-rich, sterically hindered phosphines or N-heterocyclic carbenes (NHCs) in combination with palladium or nickel catalysts.3 Owing to contributions, for example, by Hartwig,4 Buchwald,5 Miura,6 Beller,7 and Nolan,8 the reaction concept was advanced to a stage that is close to synthetic maturity. The key advantage of Pd/Ni systems is that they permit the use of inexpensive aryl chlorides. Aryl bromides or iodides can also be converted with copper catalysts,9 and highly reactive aryl substrates, such as hypervalent iodine reagents,10 benzynes,11 diazonium salts,12 nitroarenes,13 are amenable to metal-free arylation strategies. Despite the many advances made in this field,14 the coupling of small, unbranched alkyl ketones to aryl chlorides still remains a challenge, and even the most sophisticated catalyst systems reach their performance limit (Scheme 1a). The main obstacle is the particularly slow reductive elimination of aryl enolates from Pd for sterically unhindered ketones.15 The monoarylation products are inherently more reactive because the additional aryl group facilitates the reductive elimination step. Moreover, the additional aryl group increases the acidity of any remaining α-C−H bonds, so that predominantly the product molecules are deprotonated when using equimolar amounts of base. The resulting formation of mixtures of polyarylated products is a particular problem when elevated temperatures are required to convert unreactive aryl chloride substrates.5b,16
T
© XXXX American Chemical Society
In this context, the coupling of cyclohexanone to aryl chlorides is the ultimate performance test for any α-arylation catalyst. The efficient monoarylation of cyclohexanones under mild conditions has so far been reported only with aryl bromides or imidazolyl sulfonates.4c,5a,17 In the case of aryl chlorides, Buchwald reported a 4.9:1 mono-to-diarylation selectivity and 70% yield for cyclohexanone with 2dicyclohexylphosphino-2′-methylbiphenyl at 80 °C,5b Beller reported a high selectivity but only 38% yield using nbutylbis(1-adamantyl)phosphine (nBuPAd2) at 100 °C,7 and Nolan reported a high yield but moderate selectivity using a bulky NHC ligand in refluxing THF.18a However, the catalysts were tested only for cyclohexanone itself, not for substituted derivatives or other cyclic ketones, and mostly only in combination with simple chlorobenzene (Scheme 1b). After many years of optimization, the mature ligand structures described above set a standard that is hard to match with any conceptually distinct ligand type. Nevertheless, NHCs and phosphines seem to reach their inherent performance limits, and innovative concepts are vital for further performance leaps. We recently disclosed a new class of ylide-functionalized monophosphine ligands (YPhos).19 The ylide group induces donor qualities that match or even exceed those of NHCs. Metal−YPhos complexes are easily generated from benchstable phosphonium salts, which, in turn, are accessible in great structural diversity from simple precursors. The first catalytic applications of YPhos in Au-catalyzed hydroamination and hydration reactions19b as well as Pd-catalyzed Buchwald− Received: August 9, 2019
A
DOI: 10.1021/acs.orglett.9b02830 Org. Lett. XXXX, XXX, XXX−XXX
Letter
Organic Letters Table 1. Optimization of the Arylation Reactiona
Scheme 1. Transition-Metal-Catalyzed α-Arylation of Alkyl Ketones with Aryl Chlorides
entry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15b 16b 17b 18b,c 19d 20b,d 21e 22b,e
Hartwig aminations20 showcase the potential of this ligand class. The unusual activity of the catalysts can be explained by the high donor strength of the YPhos ligands. With YPhos−Pd complexes, oxidative additions of aryl chlorides have been found to be facile even at low temperatures.20 In addition, for YPhos derivatives with bulky substituents, highly reactive monoligated Pd complexes are preferentially formed. These properties gave us confidence that YPhos catalysts would enable the α-arylation of carbonyl compounds with aryl chlorides, whereas the modular tunability of the 3D ligand structure would facilitate the design of selective monoarylation catalysts (Scheme 1c). To benchmark our YPhos ligands against the best known catalysts, we chose the challenging arylation of cyclohexanone 1a with 4-chlorotoluene 2a as a test reaction. The temperature was set to only 60 °C at a catalyst loading of 1 mol % (Table 1). At most, trace amounts of product were observed when employing Pd(COD)Cl2 together with the benchmark ligands n BuPAd2, BINAP, dppf, and Xantphos (entries 1−4). Solely 2di-tert-butylphosphino-2′-(N,N-dimethylamino)biphenyl (tBuDavePhos) gave the monoarylation product 3aa in 37% yield (entry 5). To our delight, our YPhos ligand CyYMePCy2 (L1) was more active than these state-of-the-art systems under the same conditions. With L1, 3aa was obtained in >70% yield along with only traces (
