Ti-Catalyzed Reformatsky-Type Coupling between r-Halo Ketones and Aldehydes Rosa E. Este ´vez, Miguel Paradas, Alba Milla ´n, Tania Jime ´nez, Rafael Robles,* Juan M. Cuerva,* and J. Enrique Oltra* Department of Organic Chemistry, UniVersity of Granada, Faculty of Sciences, Campus FuentenueVa s/n, E-18071 Granada, Spain joltra@ugr.es; jmcuerVa@ugr.es; rrobles@ugr.es ReceiVed October 9, 2007 We describe the first Ti-catalyzed Reformatsky-type coupling between R-halo ketones and aldehydes. The reaction affords -hydroxy ketones under mild, neutral conditions compatible with ketones and other electrophiles. The catalytic cycle possibly proceeds via bis(cyclopentadienyl)titanium enolates. As long ago as 1887, Reformatsky reported the coupling between ethyl R-haloacetates and aldehydes or ketones promoted by zinc dust, thus establishing the basis of the Reformatsky reaction. 1 Currently, the Reformatsky reaction is considered in a broad sense as being the process that results from the insertion of a metal into a carbon-halogen bond activated by a carbonyl-, carbonyl-derived, or carbonyl-related group in a vicinal (or vinylogous) position, followed by coupling of the enolate thus formed with aldehydes, ketones, or other kinds of electrophile. 2a In recent years, the Reformatsky reaction has been the subject of renewed interest, due largely to the replacement of hetero- geneous zinc dust by homogeneous metals and metal derivatives, which have helped to improve the poor stereochemical control of the classic Reformatsky reaction and facilitated the develop- ment of metal-catalyzed versions of the process, among other advantages. 2,3 In this context, Little and co-workers introduced the use of Cp 2 TiCl, a mild, homogeneous, single-electron- transfer reagent 4 to promote the Reformatsky-type reaction between R-halo esters and aldehydes. 5 This method proceeds at room temperature under mild conditions and affords good yields (78-95%) of -hydroxy esters but requires stoichiometric proportions of Cp 2 TiCl. Subsequently, Cozzi and co-workers developed a titanium-catalyzed version of the process, using (CF 3 CO) 2 O as scavenger, which gives moderate yields of -hydroxy esters. 6 Here we extend the Ti-based procedure to the Reformatsky-type coupling between R-halo ketones and aldehydes to obtain -hydroxy ketones (aldol-like products) under neutral conditions. The reaction is catalyzed by substo- ichiometric proportions of the titanium complex, using Mn dust as stoichiometric reductant and a combination of Me 3 SiCl and 2,4,6-collidine, developed in our laboratory, as titanocene- regenerating agent. 7 Because Mn dust (unlike Zn) does not promote Reformatsky reactions, 5,6,8 we chose this metal to generate Cp 2 TiCl for our experiments to avoid Zn-derived competing processes, which might generate misleading observations. 5,6,9 Thus, on the basis of our own experience with Ti-catalyzed reactions, 10 we anticipated the catalytic cycle depicted in Scheme 1. According to our hypothesis, an R-halo ketone such as 1 would react with 2 equiv of Cp 2 Ti III Cl to give a titanium(IV) enolate such as 2, releasing 1 equiv of Cp 2 Ti IV Cl 2 (Scheme 1). Enolate 2 could subsequently react with an aldehyde (3) to give adduct 4. Titanocene-regenerating agent 5, presumably derived from the Me 3 SiCl/collidine mixture used, 7a would generate Cp 2 - TiCl 2 from 4, releasing 6, which after the final acidic quenching would give the desired -hydroxy ketone 7. Eventually the Mn present in the medium would reduce Cp 2 Ti IV Cl 2 to Cp 2 Ti III Cl, thus closing the catalytic cycle. (1) Reformatsky, S. Ber. Dtsch. Chem. Ges. 1887, 20, 1210-1212. (2) For a recent review, see: (a) Ocampo, R.; Dolbier, W. R., Jr. Tetrahedron 2004, 60, 9325-9374. For recent reports on Reformatsky- type reactions catalyzed by SmI2, [ClMn(salen)], low-valent Fe, and Co I , see: (b) Lannou, M. I.; He ´lion, F.; Namy, J. L. Tetrahedron 2003, 59, 10551-10565. (c) Orsini, F.; Lucci, E. M. Tetrahedron Lett. 2005, 46, 1909-1911. (d) Cozzi, P. G. Angew. Chem., Int. Ed. 2006, 45, 2951- 2954. (e) Durandetti, M.; Pe ´richon, J. Synthesis 2006, 1542-1548. (f) Lombardo, M.; Gualandi, A.; Pasi, F.; Trombini, C. AdV. Synth. Catal. 2007, 349, 465-468. (3) Cozzi, P. G. Angew. Chem., Int. Ed. 2007, 46, 2568-2571. (4) Bis(cyclopentadienyl)titanium(III) chloride (Nugent’s reagent) can be generated in situ by stirring commercial Cp2TiCl2 with Zn or Mn dust in THF, where it exists as an equilibrium mixture of the monomer Cp2TiCl and the dinuclear species (Cp2TiCl)2; see: (a) RajanBabu, T. V.; Nugent, W. A. J. Am. Chem. Soc. 1994, 116, 986-997. (b) Enemærke, R. J.; Larsen, J.; Skrydstrup, T.; Daasbjerg, K. J. Am. Chem. Soc. 2004, 126, 7853- 7864. (c) Daasbjerg, K.; Svith, H.; Grimme, S.; Gerenkamp, M.; Mu ¨ck- Lichtenfeld, C.; Gansa ¨uer, A.; Barchuck, A.; Keller, F. Angew. Chem., Int. Ed. 2006, 45, 2041-2044. (d) Gansa ¨uer, A.; Barchuk, A.; Keller, F.; Schmitt, M.; Grimme, S.; Gerenkamp, M.; Mu ¨ck-Lichtenfeld, C.; Daasbjerg, K.; Svith, H. J. Am. Chem. Soc. 2007, 129, 1359-1371. For the sake of clarity, we represent this complex as Cp 2TiCl. (5) Parrish, J. D.; Shelton, D. R.; Little, R. D. Org. Lett. 2003, 5, 3615- 3617. (6) Sgreccia, L.; Bandini, M.; Morganti, S.; Quintavalla, A.; Umani- Ronchi, A.; Cozzi, P. G. J. Organomet. Chem. 2007, 692, 3191-3197. (7) For the Me 3SiCl/2,4,6-collidine combination as titanocene-regenerat- ing agent, see: (a) Barrero, A. F.; Rosales, A.; Cuerva, J. M.; Oltra, J. E. Org. Lett. 2003, 5, 1935-1938. For pioneering work on collidine hydrochloride and other pyridine hydrochloride derivatives, see: (b) Gansa ¨uer, A.; Bluhm, H.; Pierobon, M. J. Am. Chem. Soc. 1998, 120, 12849-12859. (8) In fact, after 6 h of stirring decanal (8) with 9 (2 equiv), Mn dust (8 equiv), 2,4,6-collidine (8 equiv), and Me 3SiCl (4 equiv) in the absence of Ti, an 81% yield of 8 was recovered unchanged and only a trace of coupling product 10 was detected. Comparison of this result with that presented in entry 1 of Table 1 (80% yield of 10 after the same reaction time) indicated that under our conditions the potential Reformatsky reaction promoted by Mn/Me 3SiCl would be substantially slower than the Ti-catalyzed process. (9) It may be presumed that not only Cp2TiCl but also Zn plays an important role in Zn/Cp2TiCl2-promoted Reformatsky reactions; see: (a) Ding, Y.; Zhao, Z.; Zhou, C. Tetrahedron 1997, 53, 2899-2906. (b) Chen, L.; Zhao, G.; Ding, Y. Tetrahedron Lett. 2003, 44, 2611-2614. (10) (a) Rosales, A.; Oller-Lo ´ pez, J. L.; Justicia, J.; Gansa ¨uer, A.; Oltra, J. E.; Cuerva, J. M. Chem. Commun. 2004, 2628-2629. (b) Justicia, J.; Rosales, A.; Bun ˜ uel, E. Oller-Lo ´ pez, J. L.; Valdivia, M.; Haı ¨dour, A.; Oltra, J. E.; Barrero, A. F.; Ca ´rdenas, D. J.; Cuerva, J. M. Chem. Eur. J. 2004, 10, 1778-1788. (c) Justicia, J.; Oltra, J. E.; Cuerva, J. M. J. Org. Chem. 2004, 69, 5803-5806. (d) Justicia, J.; Oller-Lo ´pez, J. L.; Campan ˜a, A. G.; Oltra, J. E.; Cuerva, J. M.; Bun ˜uel, E.; Ca ´rdenas, D. J. J. Am. Chem. Soc. 2005, 127, 14911-14921. (e) Este ´vez, R. E.; Oller-Lo ´pez, J. L.; Robles, R.; Melgarejo, C. R.; Gansa ¨uer, A.; Cuerva, J. M.; Oltra, J. E. Org. Lett. 2006, 8, 5433-5436. 1616 J. Org. Chem. 2008, 73, 1616-1619 10.1021/jo702189k CCC: $40.75 © 2008 American Chemical Society Published on Web 01/12/2008