Volume Editor Dr. Robert D. Larsen Department of Process Research Merck and Co., Inc. P.O. Box 2000 Rahway, NJ 07065 USA E-mail: [email protected] Editorial Board Dr. John M. Brown Prof. Pierre Dixneuf Dyson Perrins Laboratory Campus de Beaulieu South Parks Road Université de Rennes 1 Oxford OX1 3QY, Av. du Gl Leclerc E-mail: [email protected] 35042 Rennes Cedex, France E-mail: [email protected] Prof. Alois Fürstner Prof. Louis S. Hegedus Max-Planck-Institut für Kohlenforschung Department of Chemistry Keiser-Wilhelm-Platz 1 Colorado State University 45470 Mühlheim an der Ruhr, Germany Fort Collins, Colorado 80523-1872, USA E-mail: [email protected] E-mail: [email protected] Prof. Peter Hofmann Prof. Paul Knochel Organisch-Chemisches Institut Fachbereich Chemie Universität Heidelberg Ludwig-Maximilians-Universität Im Neuenheimer Feld 270 Butenandstr. 5–13 69120 Heidelberg, Germany Gebäude F E-mail: [email protected] 81377 München, Germany E-mail: [email protected] Prof. Gerard van Koten Prof. Shinji Murai Department of Metal-Mediated Synthesis Faculty of Engineering Debye Research Institute Department of Applied Chemistry Utrecht University Osaka University Padualaan 8 Yamadaoka 2-1, Suita-shi 3584 CA Utrecht, The Netherlands Osaka 565, Japan E-mail: [email protected] E-mail: [email protected] Prof. Manfred Reetz Max-Planck-Institut für Kohlenforschung Kaiser-Wilhelm-Platz 1 45470 Mülheim an der Ruhr, Germany E-mail: [email protected] Preface The impact of organometallics on the synthesis of medicinal agents during the last decade cannot be overstated. Although metals or organometallic species have been used for sometime in the pharmaceutical industry, the advances in the variety of reactions and the tremendous selectivity that is often achieved has expanded the toolbox of reagents available to the medicinal and process chem- ist. These reagents have helped to streamline the synthetic approaches devel- oped for medicinal agents as can be seen by the increasingly sophisticated strat- egies used to synthesize complex drug products. In the areas of asymmetric syn- thesis, pharmaceuticals can be prepared using methodology only dreamed of a few decades ago. This volume will highlight some of the more active areas where organometal- lics are playing an important role in process chemistry. With the chelation effects of many chiral ligands, organolithium reagents have become key intermediates in asymmetric synthesis. Similarly, because of the great propensity of titanium to chelate and bind to heteroatoms, organotitanium reagents are extremely useful and versatile in carrying out a number of selective organic transformations. Al- though the asymmetric hydrogenation of a-amidoacrylates to prepare amino ac- ids has been known for sometime, the discovery and application of new ligands and reaction classes have expanded the number of substrates that can be convert- ed to chiral products with Rh and Ru-mediated reactions. The cyclopropyl group is a common moiety found in pharmaceutical agents. Metal-mediated methods for preparing this ring are reviewed herein. Non-reductive means to prepare the enantiomers of oxy-systems remained elusive until the reports of chiral salen lig- ands and osmium-mediated reactions over the past two decades. Organopalladi- um reactions are now so commonly used in the synthesis of complex molecules that one can forget that palladium was once used only in hydrogenation reac- tions. The use of metals does come with a price when preparing pharmaceuticals for human consumption. Because of the potential toxicities of the metals, only low ppm levels can remain in the active pharmaceutical ingredient. Some of the methods that have been utilized in process chemistry to remove residual metals are presented. Organometallics will certainly continue to be used extensively in the pharmaceutical industry as newer methods and applications are discovered. Rahway, USA, January 2004 Robert D. Larsen Contents Organolithium in Asymmetric Processes George G. Wu and Mingsheng Huang . . . . . . . . . . . . . . . . . . . . . . 1 Applications of Organotitanium Reagents David L. Hughes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Rhodium/Ruthenium Applications Kenzo Sumi and Hidenori Kumobayashi . . . . . . . . . . . . . . . . . . . . 63 Development of Transition Metal-Mediated Cyclopropanation Reactions Albert J. Delmonte, Eric D. Dowdy, Daniel J. Watson . . . . . . . . . . . . . . 97 Asymmetric Processes Catalyzed by Chiral (Salen)Metal Complexes Jay F. Larrow and Eric N. Jacobsen . . . . . . . . . . . . . . . . . . . . . . . . 123 Non-Salen Metal-Catalyzed Asymmetric Dihydroxylation and Asymmetric Aminohydroxylation of Alkenes. Practical Applications and Recent Advances Steven J. Mehrman, Ahmed F. Abdel-Magid, Cynthia A. Maryanoff, and Bart. P. Medaer. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Palladium-Catalyzed Heck Arylations in the Synthesis of Active Pharmaceutical Ingredients Mahavir Prashad . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Palladium-Catalyzed Cross-Coupling Reactions in the Synthesis of Pharmaceuticals Anthony O. King and Nobuyoshi Yasuda. . . . . . . . . . . . . . . . . . . . . 205 Stereospecific Introduction of Cephalosporin Side Chains Employing Transition Metal Complexes Joydeep Kant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 Removal of Metals from Process Streams: Methodologies and Applications Jeffrey T. Bien, Gregory C. Lane, Matthew R. Oberholzer . . . . . . . . . . . 263 Author Index 1–6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 Topics Organomet Chem (2004) 6: 1–35 DOI 10.1007/b11769 Organolithium in Asymmetric Processes George G. Wu · Mingsheng Huang Chemical Process Research and Development, Schering-Plough Research Institute, Union, New Jersey 07083, USA E-mail: [email protected] Abstract The development of asymmetric processes has been the focus of industrial research as most of the molecules of pharmaceutical interest contain chiral center(s). Many of the re- ported processes employ organometallic reagents in their key transformations. This review surveys chemical processes involving organolithium species in their enantioselective steps published in the past decade. Keywords Organolithium · Asymmetric process · Chiral alkylation · Chiral imine addition · Chiral aldol/Michael addition 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 Organolithium in Enantioselective Alkylations . . . . . . . . . . . . 2 2.1 Chiral Auxiliary-Mediated Alkylation. . . . . . . . . . . . . . . . . . 3 2.2 Chiral Pool-Based Chemistry. . . . . . . . . . . . . . . . . . . . . . . 11 2.3 Intramolecular Alkylation . . . . . . . . . . . . . . . . . . . . . . . . 14 2.4 Chiral Additive-Mediated Alkylation . . . . . . . . . . . . . . . . . . 14 3 Chiral Nucleophilic Addition of Organolithium Reagents . . . . . . 17 3.1 Addition to C=N Bond . . . . . . . . . . . . . . . . . . . . . . . . . . 17 3.2 Alkynylation of Imines and Ketones. . . . . . . . . . . . . . . . . . . 21 3.3 Addition to Esters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 4 Organolithium in Chiral Aldol Condensations . . . . . . . . . . . . 25 4.1 Auxiliary-Induced Aldol Reaction . . . . . . . . . . . . . . . . . . . . 26 4.2 Lewis Acid-Mediated Aldol Reaction . . . . . . . . . . . . . . . . . . 27 4.3 Catalytic Asymmetric Aldol Condensation. . . . . . . . . . . . . . . 28 5 Organolithium in Chiral Conjugated Additions. . . . . . . . . . . . 29 5.1 Conjugated Addition to a Chiral Acceptor . . . . . . . . . . . . . . . 29 5.2 Auxiliary-Induced Asymmetric Conjugate Additions. . . . . . . . . 30 © Springer-Verlag Berlin Heidelberg 2004 2 G. G. Wu · M. Huang 5.3 Chiral Additive-Mediated Conjugate Addition . . . . . . . . . . . . 32 5.4 Catalytic Asymmetric Conjugate Addition . . . . . . . . . . . . . . 32 6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 1 Introduction The development of asymmetric processes for drug candidates has become in- creasingly important as most of the structures of interest contain one or more chiral centers. There is a wealth of literature on asymmetric processes published in the past decade. Many of the enantioselective processes reported involve or- ganolithium species in the asymmetric step. This review will focus on industrial applications involving the use of organolithium as a nucleophile. The most commonly used lithium reagents are lithium diisopropyl amide (LDA), butyllithium, phenylithium, and lithium hexamethyldisilylamide (LiH- MDS). These lithium reagents are commercially available in bulk quantities and are easy to handle in the plant. When appropriate, reactions with organolithium will be contrasted with related cations, such as organosodium, organopotassi- um, and organozinc reagents. Certain additives, such as LiCl, CuX, and water are often introduced to organolithium reactions to enhance either the reactivity or the enantioselectivity. While this review focuses on the industrial applications of organolithium in asymmetric syntheses, there are some good reviews of general applications of organolithium reagents. Advances in chiral lithium amides and enantioselective protonation as well as asymmetric synthesis via lithium intermediates have been reviewed previously [1, 2]. 2 Organolithium in Enantioselective Alkylations A majority of the reported enantioselective alkylations can be divided into two distinct groups – those using a covalently bonded chiral auxiliary and those us- ing a chiral additive. The auxiliary is generally removed at the end of induction. However, in some examples, the auxiliary is incorporated as part of the mole- cule, the so-called chiral pool-based approach. Alkylation using a chiral additive is more straightforward as it does not require the attachment and removal of the auxiliary. While primary electrophiles are used in the majority of chiral alkyla- tions, secondary electrophiles also work well under certain conditions. Chiral auxiliary-based chemistry has been reviewed previously [3]. New approaches such as 1,3-asymmetric induction, the use of sterically rigid templates, and in- tramolecular trans-alkylation have also been developed. Organolithium in Asymmetric Processes 3 2.1 Chiral Auxiliary-Mediated Alkylation Diastereoselective processes using covalently bonded chiral auxiliaries have been widely used in enantioselective alkylation of lithium enolates for asymmet- ric C-C bond formation. Reactions involving organolithium species often afford chelation-controlled products. Lithium can be replaced with potassium or sodi- um as the countercation when chelation is not preferred. Oxazolidinones are frequently used as chiral auxiliaries for enantioselective alkylation. For example, Holla et al. of Hoechst have described a short and ster- eoselective synthesis of (2S)-3-(1¢,1¢-dimethylethylsulfonyl)-2-(1-naphthylme- thyl)-propionic acid 1, a very potent N-terminal component in aspartyl protease inhibitors, using an oxazolidinone 2 as an auxiliary [4]. The enantioselectivity was achieved by a stereoselective alkylation of a lithium enolate of the thioether oxazolidinone carboximide 3. Enolization with LDA followed by treatment with 1-(bromomethyl)-naphthalene and removal of the auxiliary with concurrent ox- idation of the thio group produced the alkylated product 1. An overall yield of 37% and 99% optical purity was obtained on a 100-g scale (Scheme 1). Scheme 1 Stereoselective alkylation using chiral oxazolidinones Hilpert of Hoffmann-La Roche also used the oxazolidinone auxiliary to estab- lish two consecutive chiral centers for the synthesis of Trocade, a matrix metal- loproteinase inhibitor [5]. A first enolization of the cyclopentyl propionic amide 4 with LDA followed by alkylation of the lithium enolate with tert-butyl bromoa- cetate gave the chelation controlled product 5 in 99.6% ee (Scheme 2). After re- moval of the auxiliary, the free acid 6 was converted to its corresponding piperi- dine amide 7. A chemoselective enolization of the amide carbonyl in 7 with a base followed by alkylation with bromomethyl hydantoin 8 gave either syn- or anti-succinate 9, depending on the cation of the base used. A lithium base such as LDA furnished preferentially syn-9 (anti-/syn-=15:85); whereas, potassium bases like KHMDS afforded predominantly anti-9 (anti-/syn- ratio up to 99:1). In the second alkylation, it was assumed that the chelation of the lithium enolate with the adjacent carbonyl group induces the alkylation from the sterically less hindered face, leading to the syn-9 isomer. In contrast, the potassium enolate 4 G. G. Wu · M. Huang Scheme 2 Diastereoselective alkylation process for trocade was assumed to favor the non-chelated, thermodynamically more stable confor- mation, consequently affording anti- alkylation. In a related investigation, Hilpert has shown that alkylation of the lithium enolate of succinic acid 10 with cinnamyl bromide gave a 93:7 mixture favoring the syn-isomer 11 [5]. As shown in Scheme 3, the syn-isomer 11 was converted to its corresponding anti-isomer via a second enolization with LDA. These selec- tivities were rationalized by a chelation effect of the lithium enolate which is alkylated on the sterically less hindered side leading to syn-isomer 11. A second deprotonation of 11 to the chelated enolate 12 and protonation again from the sterically less hindered side affords the anti-13. Hilper’s methods have provided efficient and practical approaches to the ma- trix metalloproteinase inhibitor Trocade and to TNF-a converting enzyme in- hibitor TACE. Both of the processes have been operated in the plant on multi-ton and multi-kilogram scale [5]. cis-Aminoindanols are another important class of chiral auxiliaries that have been extensively investigated by various Merck groups. They were originally in- vestigated because the aminoindanol structure was part of the target molecule, but have since become important auxiliaries in their own right. The first example was reported by Askin et al. for the synthesis of hydroxyethylene dipeptide isos- tere (HDI) inhibitors of HIV-1 protease 14 [6]. As shown in Scheme 4, enolization Organolithium in Asymmetric Processes 5 Scheme 3 Asymmetric synthesis of TACE Scheme 4 Diastereoselective synthesis of HIV-1 protease inhibitor 6 G. G. Wu · M. Huang of 3-phenylpropionic amide of aminoindanol 15 with n-BuLi followed by alkyla- tion with H C=C(CH I) gave a C symmetry dimer 16. The lithium enolate could 2 2 2 2 also be trapped with an epoxide to give g-hydroxyamide 17 in 90% yield and 98% ee. The outcome of the R-stereochemistry at the C-2 positions of products 16 and 17 suggested that both the alkyliodide and the epoxide electrophiles approached from the least hindered face of the lithium enolate. The facial selectivity observed here is contrasted by those reactions with prolinol amide enolates reported earli- er [6e]. In the epoxide coupling reactions, Grignard reagents such as isopropyl magnesium chloride, gave lower yields (~60%) of the desired product. Another successful application of the rigid tricyclic aminoindanol lithium ac- etonide was the asymmetric synthesis of the orally active HIV protease inhibitor Crixivan, one of the leading drugs for the treatment of AIDS [6]. Two alternative approaches were developed. Both approaches started with enolization of inda- nol amide 15 with LiHMDS. In the first approach, the lithium enolate was react- ed with epoxy tosylate 18 to give the epoxy derivative 20. Chemoselectivity was obtained between the displacement of the tosylate and the opening of the epox- ide. In the second approach, the lithium enolate was alkylated with an allyl bro- mide followed by epoxidation to afford 20. In these syntheses, the cis-aminoin- danol unit remains as a part of the molecule (Scheme 5). Scheme 5 Diastereoselective alkylation process for crixivan Alkylation of the cis-aminoindanol-modified glycine enolate 21 with a number of alkyl halides in the presence of LiCl gave the corresponding alkylated product 22 in 90~99% diastereoselectivity [7]. Benzylic or naphthyl bromide gave 99% de and 85% yield. Allylic or primary alkyl halides typically gave 95 to 98% de. The diastereoselectivity was slightly lower (91%) with a secondary acy- clic iodide. Both the ee and the yield suffered in the absence of LiCl. The auxil- iary could be effectively removed under epimerization-free conditions. This provided a practical synthesis of a-amino acids (Scheme 6). Organolithium in Asymmetric Processes 7 Scheme 6 Diastereoselective alkylation for chiral a-amino acids In the synthesis of the sidechain of an endothelin receptor antagonist, Song et al. used (1R,2S)-cis-aminoindanol for a chiral alkylation [8]. As shown in Scheme 7, enolization of a propionyl amide 23 with LiHMDS followed by alkyla- tion with benzylchloride gave 2-methyl-3-phenylpropionic amide 24 in 96% de. Removal of the auxiliary by hydrolysis gave the free acid in 60% overall yield. Scheme 7 Asymmetric synthesis of endothelin receptor antagonist (1S,2R)-1-Aminoindanol was also used by Kress et al. as a chiral auxiliary for an efficient, diastereoselective [2, 3]-Wittig rearrangement of the a-allyloxya- mide lithium enolate 25 [9]. LiHMDS was found to be a much better base than NaHMDS, KHMDS, and n-BuLi. Addition of additives, such as HMPA and DM- PU, was also necessary. After removal of the auxiliary, the resulting optically ac- tive a-hydroxyl acids were transformed to their corresponding functionalized amino acid derivatives (Scheme 8). Pseudoephedrine is another class of chiral auxiliary used in alkylations. For example, Dragovich et al. of Agouron have employed pseudoephedrine as a chi-