Cross-coupling with organosilicon compounds
Cross-coupling with organosilicon compounds involves the use of a palladium or nickel catalyst to couple an aryl, alkenyl, or alkyl halide electrophile with a nucleophilic organosilane. Organosilanes typically require nucleophilic activation to transmetalate to palladium. Fluoride sources and Brønsted bases are used for this purpose; however, "self-activating" organosilanes that do not require an activator have also been employed in the reaction.
- 1 Introduction
- 2 Mechanism and Stereochemistry
- 3 Scope and Limitations
- 4 Synthetic Applications
- 5 Comparison to Other Methods
- 6 Experimental Conditions and Procedure
- 7 References
Classical cross-coupling reactions involve the combination of an organohalide electrophile and organometallic nucleophile under palladium or nickel catalysis. Historically, organoboranes, organostannanes, and organozinc reagents have been used as the nucleophilic partners in these reactions. However, with appropriate activation, organosilanes may also serve in this role. Organosilanes offer the advantages of reduced toxicity, ease of preparation and storage, and enormous structural variety compared to the classical organometallic reagents used for cross-coupling. Although organosilanes require activation by a nucleophile to transfer a group to palladium, a variety of nucleophiles may be used for this purpose, and recent work has led to the development of "self-activating" silanolate salts.(1)
Mechanism and Stereochemistry
The generally accepted mechanism of cross-coupling reactions involves oxidative addition of a carbon-halogen (or pseudohalogen) bond of the electrophile to a palladium(0) catalyst, transmetalation of a nucleophilic group to palladium, and reductive elimination to form a new carbon-carbon bond and regenerate palladium(0). Organosilanes serve as nucleophiles in cross-coupling reactions, and participate in transmetalation mechanisms that are signficantly different from transmetalation processes involving tin, boron, zinc, and other metals. Thus, this section will focus on the mechanistic details of transmetalation from silicon to palladium.
Silanes alone are not nucleophilic enough to transfer a group to palladium. However, in the presence of a Lewis base, silicon may expand its valency and form hypervalent 10-Si-5 compounds. These compounds, which may be generated in situ or (in some cases) prepared in advance of cross-coupling, are able to transfer a nucleophilic group to palladium. Nucleophilic fluoride sources or Brønsted bases (alkoxides) are most commonly used to effect group transfer (Eq. 2). A Hammett study has demonstrated that, at least for aryl transferrable groups and aryl electrophiles, the actual transmetalation event involves electrophilic attack of the arylpalladium(II) intermediate on the ipso carbon of the activated silane.(2)
Silanols exhibit an interesting kinetic profile that suggests unique dynamics under fluoride activation, prior to the transmetalation event. The silanol 1 rapidly forms siloxane 2 and hydrogen-bonded complex 3 in the presence of tetrabutylammonium fluoride (TBAF). Reaction of 2 with an additional molecule of TBAF leads to activated disiloxane 4, which (consistently with kinetic data) participates in rate-limiting, bimolecular transmetalation to palladium (Eq. 3).(3)
Brønsted bases may facilitate transmetalation through two distinct mechanisms. Under stoichiometric conditions, when palladium and silanolate K+5- are present in a 1:1 ratio, nucleophilic attack on palladium leads to a T-shaped palladium complex, which undergoes unimolecular transmetalation. Under catalytic conditions, a second molecule of silanolate likely becomes involved in the formation of a 10-Si-5 intermediate, which transfers a group to palladium at ten times the rate of the neutral T-shaped complex (Eq. 4).(4)
Stereochemistry and Site Selectivity
Cross-coupling of alkenylsilanes is, in the vast majority of cases, a stereospecific and retentive process. However, alkenylsilanes are subject to changes in site selectivity stemming from the possibility of reaction at the ipso or cine positions of the nucleophile. For cross-couplings with aryl electrophiles, more electron-rich electrophiles encourage formation of the cine product. In a study employing aryldiazonium salts, the cine product was interpreted as arising from a pathway involving multiple carbopalladation events. The ipso product arises from carbopalladation in the opposite sense and anti elimination of palladium and silicon (Eq. 5).(5)
Allylsilanes may be subject to both site selectivity and stereoselectivity issues in cross-coupling reactions. The site selectivity of reactions of allylsilanes depends strongly on the identity of the ligand used. π-Acidic ligands such as dibenzylidene acetone (dba) generally lead to selective reaction at the gamma position. Gamma-selective transmetalation from an allylsilane to palladium may be interpreted as an SE2' process. Under fluoride activation, the stereospecificity of this step depends on the fluoride source used. Sources containing non-coordinating counterions, such as TBAF, lead to open SE2' transition states and anti stereospecificity. Coordinating counterions (such as Cs+) lead to closed transition states and syn stereospecificity. For a similar reason, self-activating silanolates transmetalate in a syn SE2' fashion (Eq. 6).(6)
Scope and Limitations
Trimethyl- and tert-butyldimethylsilanes are deactivated with respect to group transfer relative to other classes of silicon reagents. The incorporation of a Lewis basic group onto an aryl or alkenyl transferable group leads to intramolecular activation after coordination to copper(I). It was recognized that for groups that transfer more readily than phenyl rings, 2-hydroxymethylphenyl can serve as a "universal intramolecular activating group." However, a Brønsted base and copper(I) salt are still required to accomplish group transfer when the 2-hydroxymethylphenyl group is employed. Heteroaryl groups, on the other hand, are more commonly transferred from trialkylsilanes with the help of a fluoride source.(7)
Among monohalosilanes, fluorides are most commonly used. Fluorides are most commonly used to effect the transfer of alkenyl groups. In general, monohalosilanes are less reactive than dihalosilanes.(8)
The attachment of a second halogen to the silicon center helps suppress undesired products resulting from alkyl group transfer, particularly when an aryl transferable group is employed. Conditions favorable for reactions of difluorosilanes do not work for dichlorosilanes; however, it has been observed that pretreatment of dichlorosilanes with potassium fluoride before the introduction of palladium is an effective method for cross-coupling of this class of nucleophiles.(9)
Although the reactivity trend of mono- and dihalosilanes may suggest that trihalosilanes are even more effective cross-coupling partners, this is not generally true. Only trifluorosilanes have been used to date in several specific applications. For instance, trifluorosilanes with aryl transferable groups undergo cross-coupling with alkyl bromides and iodides under nickel catalysis in the presence of cesium fluoride.(10)
Silanols, Silanolates, and Alkoxysilanes
Silanols may be used as nucleophiles in cross-coupling reactions under either fluoride or Brønsted base activation. The observation that Brønsted base activation leads to the formation of stable, storable silanolate salts led to the development of these salts as "self-activating" coupling partners. These reagents obviate the need for a separate activation reagent and avoid issues associated with the use of neutral silanols, such as dimerization to disiloxanes, limited amounts of active nucleophile present in the reaction mixture, and competitive coordination of Brønsted base to palladium.(11)
Silyl ethers may also be used as coupling partners, but are limited by available methods of preparation to primarily alkenyl transferable groups. The bis(silyl) linchpin in Eq. 12 is prepared by a combination of platinum-catalyzed hydrosilylation and ruthenium-catalyzed silylative coupling reaction. The methyl-substituted ether reacts more rapidly than the isopropyl-substituted ether under activation by TMSOK.(12)
Disiloxanes are synthetic equivalents of silanols whereas polysiloxanes are synthetic equivalents of silanediols (acid or base catalysis produces the monomeric alcohols). All four silyl groups in the reagent D4V may be transferred to alkenyl and aryl electrophiles.(14)
Bis(catecholato)siliconates may be used as nucleophiles without fluoride activation. Although alkenylsiliconates in this class of reagents give significant amounts of cine substitution, aryl groups may be cleanly transferred to aryl triflates.
Fluoride and Brønsted base activation modes may be employed independently of one another within a single substrate in order to synthesize complex conjugated carbon frameworks from a single linchpin nucleophile. The synthesis of RK-397 nicely illustrates this concept. Linchpin reagent 6 possesses a silanol group that reacted selectively under conditions of Brønsted base activation to furnish triene 7. The use of TBAF in a subsequent cross-coupling led to reaction of the remaining trialkylsilyl group. Further elaboration of tetraene 8 led to RK-397.(15)
Comparison to Other Methods
Compared to tin-, boron-, and zinc-based nucleophiles, tetracoordinate silicon-based reagents vary enormously in structure, and a wide variety of methods for activating silicon-based nucleophiles are available. This fact may seem daunting to the synthetic chemist interested in a simple cross-coupling method that "just works." On the other hand, the variety of silicon-based cross-coupling methods makes the reaction highly versatile, and the substrate scope, at least for sp2-sp2 couplings, is as wide as cross-couplings with tin, boron, and zinc reagents. Additionally, organosilanes are straightforward to prepare, exhibit low toxicity, and often cost less relative to other nucleophiles for cross-coupling.
Experimental Conditions and Procedure
Organosilanes for cross-coupling may be prepared using one of four methods: (1) addition of organometallic nucleophiles to silicon electrophiles, (2) hydrosilylation, (3) silylative coupling and metathesis, and (4) insertion with silanes or disilanes. Because of the added variable in these reactions of the structure of the silane, determining the optimal conditions for coupling can be a challenging problem. Nonetheless, cross-couplings employing organosilanes bear a number of similarities to other cross-coupling reactions. All of the most common palladium(0) and palladium(II) catalysis used with other nucleophiles have been successfully employed with organosilanes. The ligand required to accomplish coupling depends largely on the nature of the electrophile. Iodides react under "ligandless" conditions, but bromides and chlorides require phosphine ligands. More weakly donating phosphine ligands may be employed with an increase in reaction temperature. Tetrahydrofuran and dimethylformamide are the most common solvents used. Additives may be needed to facilitate transmetalation (copper(I) salts) or extend catalyst lifetime (triphenylphosphine oxide, surfactants). Both hydrated and anhydrous fluoride sources have been used, and the most common Brønsted bases employed are silver oxide, silver carbonate, and the less expensive alkali metal trimethylsilanolates.
To an oven dried, 5-mL, single-neck, round-bottomed flask, containing a magnetic stir bar, equipped with a reflux condenser and an argon inlet capped with a septum was added Pd(dba)2 (28.8 mg, 0.05 mmol, 0.05 equiv). The flask was then sequentially evacuated and filled with argon three times. The aryl bromide (1.0 mmol) was then added by syringe. Sodium 2-butenyldimethylsilanolate (308 mg, 2.0 mmol, 2.0 equiv), pre-weighed into a 10-mL, two-necked, round-bottomed flask in a dry-box, was then dissolved in toluene (2.0 mL) and then norbornadiene (nbd, 5.2 µL, 0.05 mmol, 0.050 equiv) was added by syringe. The solution of silanolate and nbd in toluene was then added to the aryl bromide by syringe. The reaction mixture was heated under argon to 70 ºC in a preheated oil bath. After complete consumption of the aryl bromide was observed by GC analysis, the mixture was cooled to rt, filtered through silica gel (2 cm x 2 cm) in a glass-fritted filter (C, 2 cm x 5 cm) and the filter cake washed with ether (3 x 10 mL). The filtrate was concentrated (ambient temp., 20 mm Hg) and the residue was purified by silica gel chromatography (20 cm x 20 mm, hexane/EtOAc, gradient 100:0 to 20:1) followed by Kugelrohr distillation, to afford the product (155 mg, 67%) as a clear, colorless oil: bp 155º (5 mmHg, ABT); TLC Rf 0.33 (hexane/EtOAc, 20:1) [silica gel, UV]; GC γ-tR : 6.70 min, α-tR : 6.88 and 6.95 min (25:1 [3.2:1], γ:α [E or Z]); IR (film) 2975 (m), 2931 (m), 1718 (s), 1700 (m), 1636 (w), 1610 (m), 1578 (w), 1507 (w), 1477 (m), 1457 (m), 1413 (m), 1392 (m), 1368 (m), 1291 (s), 1256 (m), 1167 (m), 1117 (m), 1017 (m), 915 (m), 849 (m), 771 (m), 708 (m) cm-1; 1H NMR (500 MHz, CDCl3) δ 7.93 (m, 2 H, HC(3)), 7.26 (m, 2 H, HC(2)), 5.98 (ddd, J = 17.7, 9.8, 6.4, 1 H, HC(7)), 5.05 (m, 2 H, HC(8)), 3.51 (ap, J = 7.0 1 H, HC(6)), 1.59 (s, 1 H, HC(11)), 1.37 (d, J = 7.0, 3 H, HC(5)); 13C NMR (126 MHz, CDCl3) δ 166.0 C(9), 150.6 C(1), 142.7 C(7), 130.2 C(4), 129.9 C(3), 127.4 C(2), 114.0 C(8), 81.0 C(10), 43.3 C(6), 28.5 C(11), 20.8 C(5); EIMS (70 eV) m/z (%): 232 (M+ , 12), 176 (48), 159 (53), 131 (100), 117 (37), 84 (18). Anal. calcd. for C15H20O2: C, 77.55; H, 8.68 Found: C, 77.28; H, 8.72.
- ↑ Chang, W.-T. T.; Smith, R. C.; Regens, C. S.; Bailey, A. D.; Werner, N. S.; Denmark, S. E. Org. React. 2011, 75, 213. doi: (10.1002/0471264180.or075.03)
- ↑ a b Echavarren, A. M.; Cardenas, D. J. In Metal-Catalyzed Cross-Coupling Reactions, 2nd ed.; de Meijere, A., Diederich, F., Eds.; Wiley-VCH: Weinheim, 2004; Vol. 1, pp 1-40.
- ↑ Chuit, C.; Corriu, R.; Reye, C.; Young, J. Chem. Rev. 1993, 93, 1371.
- ↑ Hosomi, A.; Kohra, S.; Tominaga, Y. Chem. Pharm. Bull. 1988, 36, 4622.
- ↑ Denmark, S. E.; Smith, R. C.; Chang, W.-T.T. Tetrahedron 2011, 67, 4391-4396.
- ↑ Denmark, S. E.; Sweis, R. F.; Wehrli, D. J. Am. Chem. Soc. 2004, 126, 4865.
- ↑ Denmark, S. E.; Smith, R. S. J. Am. Chem. Soc. 2010, 132, 1243.
- ↑ Denmark, S. E.; Kallemeyn, J. M. J. Am. Chem. Soc. 2006, 128, 15958.
- ↑ Matsuhashi, H.; Hatanaka, Y.; Kuroboshi, M.; Hiyama, T. Heterocycles 1996, 42, 375.
- ↑ a b Denmark, S. E.; Werner, N. S. J. Am. Chem. Soc. 2008, 130, 16382.
- ↑ Denmark, S. E.; Werner, N. S. J. Am. Chem. Soc. 2010, 132, 3612.
- ↑ S. Akai; T. Ikawa; S. Takayanagi; Y. Morikawa; S. Mohri; M. Tsubakiyama; M. Egi; Y. Wada; Y. Kita. Angew. Chem. Int. Ed. 2008, 47, 7673.
- ↑ Pierrat, P.; Gros, P.; Fort, Y. Org. Lett. 2005, 7, 697.
- ↑ Hatanaka, Y.; Goda, K.; Hiyama, T. J. Organomet. Chem. 1994, 465, 97.
- ↑ a b Hatanaka, Y.; Goda, K.; Okahara, Y.; Hiyama, T. Tetrahedron 1994, 50, 8301.
- ↑ Powell, D. A.; Fu, G. C. J. Am. Chem. Soc. 2004, 126, 7788.
- ↑ Denmark, S. E.; Neuville, L.; Christy, M. E. L.; Tymonko, S. A. J. Org. Chem. 2006, 71, 8500.
- ↑ Denmark, S. E.; Ober, M. H. Org. Lett. 2003, 5, 1357.
- ↑ Denmark, S. E.; Smith, R. C.; Chang, W.-T. T.; Muhuhi, J. M. J. Am. Chem. Soc. 2009, 131, 3104.
- ↑ Marciniec, B.; Pietraszuk, C. In Topics in Organometallic Chemistry; Bruneau, C., Dixneuf, P. H., Eds.; Springer: Berlin, 2004; Vol. 11, pp 197-248.
- ↑ Murakami, M.; Matsuda, T.; Itami, K. Ashida, S.; Terayama, M. Synthesis 2004, 1522.
- ↑ Mowery, M. E.; DeShong, P. J. Org. Chem. 1999, 64, 1684.
- ↑ a b Ranu, B. C.; Dey, R.; Chattopadhyay, K. Tetrahedron Lett. 2008, 49, 3430.
- ↑ a b Denmark, S. E.; Butler, C. R. Chem. Commun. 2009, 20.
- ↑ Seganish, W. M.; DeShong, P. J. Org. Chem. 2004, 69, 1137.
- ↑ Denmark, S. E.; Liu, J. H.-C. Angew. Chem. Int. Ed. 2010, 49, 2978.
- ↑ Denmark, S. E.; Fujimori, S. J. Am. Chem. Soc. 2005, 127, 8971.
- ↑ Brook, M. A. Silicon in Organic, Organometallic, and Polymer Chemistry; Wiley: West Sussex, 1999.
- ↑ Denmark, S. E.; Smith, R. S.; Tymonko, S. Tetrahedron 2007, 63, 5730.