Synthesis, Stereochemical Assignment, and Enantioselective Catalytic Activity of Late Transition Metal Planar Chiral Complexes

Abstract

Planar chirality is an important form of molecular chirality that can be utilized to induce enantioselectivity when incorporated into transition metal catalysts. However, due to synthetic constraints, the use of late transition metal planar chiral complexes to conduct enantioselective transformations has been limited. Additionally, the published methods surrounding the stereochemical assignment of planar chiral compounds are sometimes conflicting, making proper assignment difficult. This review aims to provide clarity on the methods available to assign planar chirality and provide an overview on the synthesis and use of late transition metal planar chiral complexes as enantioselective catalysts.

1. Introduction

Enantioselective catalytic transformations utilizing chiral late transition metal catalysts bearing C₂-symmetric, π-coordinated ligands have been extensively studied in recent years. Notable catalyst platforms from the Cramer, You, and Wang groups rely on steric blocking effects of the ligand to induce enantioselectivity.^1–3 The C₂-symmetric nature of these ligands allows for the generation of a single catalyst independent of which face of the π-complex the metal coordinates. While the use of C₂-symmetric ligands simplifies the synthesis of the chiral catalyst, designing ligands to generate planar chiral complexes provides a powerful alternative. Ligand scaffolds that are C₂-symmetric must independently contain all the chiral information necessary to impart selectivity; in contrast, the ligands used in planar chiral complexes can contain chiral information or be entirely pro-chiral prior to coordination of transition metals. Additionally, many planar chiral complexes demonstrate a unique ability to control chirality at a metal center, increasing their utility.^4–8 While there is a significant body of work detailing planar chiral complexes of early transition metals,^9,10 planar chiral late transition metal complexes (Group XIII and Group IX) are relatively underexplored. The focus of this review is to highlight planar chiral late transition metal complexes specifically used to catalyse enantioselective transformations and provide a thorough foundation for future research; therefore, planar chiral complexes whose catalytic activity has not been explored will not be included. Furthermore, this scope only includes planar chiral complexes where the metal center itself participates in catalysis; planar chiral complexes where the metal center is not involved in the catalytic cycle will not be included, and are detailed in other reviews.^11,12 This review is organized into three sections; first, we will review the definition and assignment of planar chirality; second, we will describe the syntheses of late transition metal planar chiral complexes; finally, we will review their actions as catalysts for enantioselective transformations.

1.1. Defining Planar Chirality

Planar chirality was first described by Cahn, Ingold, and Prelog in the context of cyclophanes.¹³ When viewing cyclophane 1, one can envision the molecule existing in two planes (I and II, Fig. 1A). In this representation, the substituted benzene ring exists in plane I and the ethereal chain exists in perpendicular plane II. Because the benzene ring bears non-equivalent substituents, plane II becomes a plane of asymmetry. This becomes clear in 2 when viewing plane I of the molecule along plane II from the direction bearing the most atoms outside of plane I, i.e. the face of the benzene ring over which the ethereal chain is linked. Thus, compound 1 is a planar chiral molecule where the chirality is, in this case, not reducible to central chirality. Subsequent work, in collaboration with Schlögl and Falk, defined planar chirality as it applied to transition metal π-complexes.^14,15 Similarly to cyclophanes, there exists a plane of asymmetry when a molecule such as 3 is viewed in two perpendicular planes (III and IV, Fig. 1B). When the molecule is viewed through the iron atom (4, unsubstituted cyclopentadienyl ring removed for clarity), plane IV is the plane of asymmetry due to the non-equivalent carbonyl group. Unlike the cyclophanes, however, the defined planar chirality of these types of molecules is reducible to the central chirality of the highest priority atom in the π-complex.¹⁴ Considering this review’s focus on planar chiral catalysts among late transition metal π-complexes, this mode of planar chirality will be the focus of the following section on planar chiral assignment.

Figure 1.

Approaches to representing and viewing planar chiral molecules

1.2. Assigning Planar Chirality

Assigning chirality to a stereogenic carbon following the standardized Cahn-Ingold-Prelog (CIP) rules is fundamental knowledge throughout chemistry.^13,14,16,17 Planar chirality of π-complexes, however, is less commonly studied and wasn’t fully established until the third addendum to the CIP rules in 1966.¹⁴ Conflicting reports regarding the rules for determining the absolute configuration of a planar-chiral complex have led to different planar R (pR) or planar S (pS) assignments.¹⁸ As such, we shall describe the most generally accepted rules for assigning the absolute configuration of planar chiral π-complexes as they are outlined by Cahn, Ingold, and Prelog^14,15 in conjunction with work conducted by Schlögl and Falk.

We will outline the methods for assigning planar chirality by demonstrating the different approaches on ferrocene derivative 3 (Fig. 2). The key to assigning planar chirality relies in establishing the centroid, an atom within the π-complex that will exhibit pseudo-central chirality that we can extend to planar chirality. To do this, we must first determine the plane to utilize in the assignment. In the case of metallocenes, the most-substituted planar-chiral ring is the focus (the ketone bearing ring in 3, Fig. 2).¹⁸ For centroid assignment, we will view the Cp ring from the face bearing the atom with the highest atomic number that is attached to, but not in, the plane of the π-complex. In late transition metal complexes, this refers to the metal coordinated to the π-complex, also known as the pilot atom (Fe atom in 3, Fig. 2).^19,20 Next, we must find the highest priority atom (following classical CIP rules) of the π-system that is bonded to the pilot atom. It is important to consider the iron atom as singly bound to each atom of the π-complex (5, Fig. 2),¹⁹ rather than coordinating to the entire π-system. This will make every atom of the π-system exhibit pseudo-central chirality. Following these rules, C1 (5, Fig. 2) is the highest priority atom of the π-complex as it is bound to the highest priority substituent as well as the iron atom and thus will be the centroid for assigning planar chirality.

Figure 2.

Methods for assignment of planar chirality

There are three methods to assign the absolute configuration from the chirality of the centroid that will, most often, provide the same conclusion on stereochemical assignment: the central chirality when considering the highest-priority atom in a pseudo-tetrahedral configuration, the direction of rotation of the trigonal substituents around the centroid (C1) with respect to the position of the pilot atom, and the application of the rules for assigning cyclophane chirality to π-complexes. However, due to the inconsistencies of applying cyclophane chirality rules to π-complexes, this review will focus only the first two methods.

Following the first method, the atoms about the centroid are assigned as a (iron), b (C2), c (C5), and d (C=O) where a > b > c > d according to CIP priority rules (3, Fig. 2A).^14,19 When making this assignment, recall that each atom of the π-complex must be considered to have a single bond to the iron atom 5,¹⁹ which explains why the carbonyl carbon ends up with the lowest priority. Next, view the centroid (C1) in a tetrahedral configuration (6, Fig. 2A). Finally, view the tetrahedron with the lowest priority atom (d) in the back to assign a clockwise (pR) or anticlockwise (pS) rotation following the direction of rotation of a to b to c, mirroring CIP rules for tetrahedral central chirality.^13,16

When utilizing the second method, begin the same as the first method by viewing the planar-chiral ring from the orientation of the pilot atom and assigning priority (a, b, c, and d, where a > b > c > d according to CIP priority rules) (5, Fig. 2B).¹⁴ For this method, we will focus on the trigonal-planar substituents around the centroid (C1). Assign rotation by following the direction of b to c to d, according to priority, as clockwise (pR) or anticlockwise (pS) when the pilot atom is positioned above the plane bearing the π-system (7, Fig. 2B).

These methods should provide the same result in most cases; note that both methods assign 3 as pS. Discrepancies mainly occur when the highest priority atom of the π-complex is substituted with an atom of higher priority than the pilot atom. This most often occurs in the case of halogen substitutions on complexes with first row transition metals. One case is seen in compound 8 (Fig. 2C). Flipping the molecule to view the pilot atom above the π-complex 9, method one (10) will assign pR while method two (11) will assign pS. In these cases, method one should be used as central chirality has higher priority than planar chirality according to CIP rules.^13,14

We note that method one provides the most accurate and consistent planar chiral assignments, while method two allows for rapid assignment with the caveat that errors can occur if one of the substituents is higher priority than the pilot atom.

2. Synthesis and Chiral Resolution of the Planar Chiral Ruthenium, Iron, Rhodium and Cobalt Complexes Used in Enantioselective Catalysis

A majority of the literature concerning late transition metal planar chiral catalysis for enantioselective transformations employs Ru, Fe, Rh and Co catalysts. As such, this section will outline the two main strategies used to synthesize these planar chiral complexes and the methods used to resolve them. The first strategy involves de novo synthesis of the ligand and subsequent complexation to form the metal complex. The second approach, more common with Shvo type catalysts,²¹ is to form the ligand and complex in one pot via direct annulation mediated by the precursor metal complex. Conversely, there is no general approach to resolving mixtures of planar chiral complexes. The most common method involves complexation of ligands with existing chirality followed by achiral chromatography to resolve the diastereomeric mixtures. Another common practice is to complex achiral ligands, but then utilize chiral HPLC to resolve the resultant enantiomers. More creative and tailored solutions have been developed to address the issue of delivering stereo-enriched planar chiral catalysts, such as steric blocking ligands and recrystallization. In this section, we have curated the syntheses and resolutions of planar chiral catalysts that are relevant to stereoselective transformations.

2.1. Synthesis and Resolution of Planar Chiral Ruthenium Catalysts

The current catalogue of planar chiral ruthenium catalysts (Fig. 3) for enantioselective transformations contains two main groups of catalyst scaffolds: cyclopentadienyl (Cp) derived complexes and η⁶-arene complexes, the former being more common. There are a variety of methods used to produce and resolve these complexes, which we will outline below.

Figure 3.

Planar chiral ruthenium catalysts that have been used for enantioselective transformations.

Ruthenium Cp complexes have been broadly applied toward catalysis and many planar chiral analogues have been developed. In the context of allylic substitution catalysis, Takahashi and co-workers have prepared a variety of tethered planar chiral Cp complexes such as 12 (Scheme 1) and an untethered planar chiral Cp complex 13 (Fig. 3).²² Takahashi and coworkers’ synthetic scheme is shown in Scheme 1 with representative complex pS-12. Enantiopure precursor pS-22 was accessed from a cyclopentadienyl thallate before being used as the starting material for their subsequent catalyst syntheses.²³ Diastereoselective crystallization of precursor complex pS-22 for the dimethyl Cp variant only yields 26% of pure diastereomer.²⁴ Ester pS-22 was hydrolysed, then converted to the acid chloride and re-esterified with the requisite aryl phosphine to form pS-23 in good yield over three steps. Irradiation of pS-23 in the presence of acetonitrile tethered the pendant phosphine and afforded pS-12 in four steps overall from pS-22. Untethered complex pS-13 was prepared more efficiently, in which pS-22 was irradiated directly.^23,24 The main weakness of this strategy was the subsequent manipulations carried out on precious transition metal complex to deliver the final active catalyst. Still, 12 and 13 were the earliest examples of planar chiral catalysts with planar chirality as their only source of chiral information that were successful in delivering enantioenriched products.

Scheme 1.

Synthesis of tethered Cp ruthenium catalysts with alkyl phosphine linker.

While many of the tethered planar chiral ruthenium complexes are Cp derivatives, Faller and co-workers have developed a series of tethered planar chiral benzene derived ruthenium complexes (Scheme 2).^26–28 Their syntheses began with complexing DavePhos with [Ru(benzene)Cl₂]₂ via irradiation to deliver racemic complex 14a in 94% yield. A single chloride was substituted with triphenylphosphine to deliver racemic 14b.²⁶ Crystallization of the phosphine complex in DCM/ether delivered enantiopure pS- or pR-14b crystals which were manually separated.²⁶ The Faller group introduced a more convenient resolution of (±) 14a that delivered pR-14a by complexing the racemate with (R)-BINAM to only precipitate crystals of pR-24 via fractional crystallization and reconstitution of the dichloro complex with HCl.²⁷ This resolution method was a significant improvement that precluded the manual separation of crystals. Enantiopure pR-14a was converted to the NHC ruthenium complex pR-15 in 65% yield over three steps.²⁸ Faller and co-workers delivered an efficient and concise synthesis of tethered planar chiral complex from relatively inexpensive phosphine.

Scheme 2.

Faller and co-workers provided several tethered η⁶ ruthenium catalysts that originate from DavePhos.

Cyclopentadienone complexes have become a topic of interest as alternatives to Shvo catalysts for hydrogenation reactions. A common strategy to deliver planar chiral cyclopentadienone catalysts is preparation of a diyne followed by one-pot cyclization and complexation with Ru₃(CO)₁₂. Yamamoto et al. followed this strategy to develop a planar chiral cyclopentadienone framework based on C-arylribosides to carry out asymmetric hydrogenations (Scheme 3).²⁹ The synthesis began with lactone 26, a previously synthesized intermediate. Addition of phenylacetylide, followed by substitution with propyn-1-ol, delivered diynes 27a-b. The diyne was then cyclized with Ru₃(CO)₁₂ to deliver a diastereomeric mixture of 28a-b. The TBS-protected diastereomers 28a-b proved inseparable; however, desilylation provided both diastereomers of 16a-b that were separated via silica chromatography (only the pR diastereomer is depicted in Scheme 3).

Scheme 3.

Shvo-type planar chiral Ru catalysts developed by Yamamoto, Wills, and Dou. *Complexes not shown.

Wills et al. added to the existing library of planar chiral ruthenium cyclopentadienones with a series of catalysts based on a framework they also developed for iron catalysis.³⁰ They accessed a variety of diynes (R)-33a-c in three steps from alkyne 29. The ketone was subjected to asymmetric hydrogenation catalysed by (R,R)-30, allowing for separation of the planar chiral complexes downstream. After converting enantiopure alkyne (R)-31 into respective diynes (R)-33a-c, direct complexation and cyclization with Ru₃(CO)₁₂ yielded cyclopentadienone catalysts 17a-c. Both diastereomers of 17b were accessible, while only the pR diastereomers of 17a and 17c were obtained. While the stereoenriched diynes required to synthesize the cyclopentadienone catalysts took several steps to produce, once complexed, the additional chiral information allowed for separation via achiral chromatography.

One of the more recent advancements in Ru cyclopentadieneone catalysts was made by Hayashi and Dou when they developed one of the first Shvo catalysts containing planar chirality as its only source of asymmetry.³¹ Hayashi and Dou forwent direct complexation and cyclization and instead prepared several cyclopentadienones that were then complexed onto Ru. Complex (pS,pS)-18a is shown as a representative example. Cyclopentadienone 34a was first complexed with Ru₃(CO)₁₂ to deliver the monomeric racemate 35a, which was then resolved via chiral HPLC and dimerized to provide typical Shvo-type catalyst (pS,pS)-18a.

Perekalin et al. have recently disclosed an untethered ruthenium complex featuring a camphor-derived arene ligand (Scheme 4).³² Synthesis of the complex began with (+)-camphor, which was converted to the hydrazone and subsequently iodinated to afford 4-iodocamphene 36 and 2-iodobornene 37. This mixture was carried into a Sonogashira coupling which selectively reacted with the vinyl iodide, providing tethered ene-yne fragment 39 that was then reduced and cyclized to furnish diene ligand 40. Complexation of ruthenium with 40 occurred diastereoselectively due to steric blocking from the camphor backbone, forming complex (pR,pR)-19. This is an example of a synthesis that utilizes a diastereoselective complexation rather than a chromatographic separation to afford enantiopure planar chiral catalysts.

Scheme 4.

Perekalin and co-workers’ camphor-derived ligand for diastereoselective synthesis of (pR,pR)-19.

The Wang group has further diversified the η⁶-arene class of planar chiral Ru catalysts with a unique paracyclophane-derived arene ligand (Scheme 5).³³ Synthesis began with enantiomerically-pure aldehyde pR-41, which was oxidized to the methyl ester and exposed to an excess of MeMgBr to afford the dialkylated alcohol pS-43. Subsequent deoxygenation provided the free C₂-symmetric ligand pS-44, which was complexed to [Ru(benzene)Cl₂]₂ to afford the sandwich complex (pS,pS)-45. Reduction with Red-Al exclusively occurs at the non-cyclophane arene ligand, and treatment with HCl furnishes the Ru chloride dimer (pS,pS)-20 in moderate yield. Similar to the previous ligand, this is an example of a Ru-arene complex that does not require a stereoselective purification due to the chiral nature of the ligand itself.

Scheme 5.

Wang and co-workers’ paracyclophane-derived Ru-arene complex

2.2. Synthesis and Resolution of Planar Chiral Iron Complexes

Currently the only planar chiral iron catalysts that are involved in enantioselective transformations bear cyclopentadienone ligands. These Shvo-type complexes have been utilized as catalysts in enantioselective reductions of carbonyls and are an attractive alternative to ruthenium catalysts due to their comparatively low cost. The synthetic strategy to access cyclopentadienone iron complexes revolves around delivering a diyne intermediate that is cyclized and complexed in one step with Fe_x(CO)_x, mirroring the strategies used in Ru Shvo catalyst syntheses. Most of these cyclopentadienone catalysts possess additional chiral information alongside their planar chirality; however, some complexes possess solely planar chirality and require more involved resolution techniques.

Wills et al. synthesized and evaluated many of the early planar chiral iron complexes that contain cyclopentadienone ligands fused with furan rings (Scheme 6).^30,34 Diynes (R)-33a,b,d were synthesized from diyne 32 as they were for the ruthenium complexes from Wills 2010/2011. Subsequently, cyclopentadienone complexes 46a,b,d were accessed in one step via cyclization with pentacarbonyliron(0). The resultant diastereomeric mixtures were separated via silica chromatography.

Scheme 6.

Wills and co-workers’ direct synthesis of Shvo-type planar chiral iron complexes from diyne precursors

Wills et al. also synthesized a series of cyclopentadienone iron complexes without an ether linkage (Scheme 7).³⁵ Novel diynes 51 and 53 were synthesized in three steps from 5-(trimethylsilyl)pent-4-yn-1-al in 58% and 78% respective overall yields. Diastereomeric complexes pS-47a and pR-47b were manufactured over two steps and resolved via silica chromatography. However, complex 54 was delivered as a racemic mixture, so achiral chromatography could not resolve the catalyst. The Wills group addressed this challenge via kinetic resolution of the racemic mixture and utilized (S,S)-30 (catalyst structure shown in Scheme 3) to selectively convert the pS enantiomer to the alcohol pS-47b. The complexes were then separated on silica to deliver pS-47b and pR-48. As we have seen before, the chiral diyne substrates require a four-step synthesis, but once complexed, the resultant diastereomers are resolved via achiral chromatography.

Scheme 7.

Examples of iron cyclopentadienone catalysts with fused alkyl backbones

Gennari et al. expanded on the cyclopentadienone scaffold with complex 49, which contains a fused six-membered ring, and more typical Shvo type complex 50 (Scheme 7).³⁶ The synthesis of catalyst 49 followed a plan similar to previous cyclopentadienone complexes. Diyne 55 was accessed in three steps and 46% yield overall. Subsequent silylation provided functionalized diyne 56. The diyne was then cyclized with diiron nonacarbonyl to deliver racemic 49, which was resolved via chiral HPLC. Catalyst 50 developed by Gennari et al. diverged from the typical diyne strategy seen in previous Fe Shvo type catalysts by abandoning the fused ring scaffold.³⁶ Instead, an aldol condensation was used to form cyclopentadienone 58, which was then complexed onto iron with diiron nonacarbonyl. This afforded racemic complex 50 which was ultimately resolved via chiral HPLC. The catalysts provided by Wills and co-workers require enantioselective hydrogen transfer reactions, as opposed to the simpler syntheses of Gennari catalysts. It is notable, however, that the racemic catalysts produced by Gennari et al. require chiral HPLC, while Wills diastereomeric complexes were resolved via simple silica chromatography.

2.3. Rhodium Complexes

Planar chiral rhodium catalysts have been utilized for a variety of stereoselective transformations since 2002. The Rh scaffolds have all been built around an indenyl or Cp core (Fig. 5), with two main strategies to produce these planar chiral complexes. The most common strategy involves full synthesis of the ligand, then complexation to a Rh(I) center, followed by oxidation to a Rh(III) catalyst in some cases. Otherwise, a common strategy is production of a simpler substrate that is directly cyclized and complexed to the Rh(I) center.

Figure 5.

Planar chiral rhodium catalysts utilized in enantioselective transformations.

The first example of planar chiral rhodium complexes that would be used in enantioselective transformations was provided by Schumann et al. with menthyl substituted indenyl rhodium complexes (Scheme 8).⁷ Complexation of indenyl lithiate 70 with [Rh(COD)Cl]₂ provided a mixture of diastereomers that were separated via silica chromatography to give pR-59 in 18% yield and pS-59 in 29% yield.

Scheme 8.

Synthesis of indenyl Rh catalysts from a neomenthyl-substituted indenyl lithiate.

Waldmann et al. began to explore Rh Cp complexes with a library of JasCp ligands aimed at providing catalysts with broad applicability (Scheme 9).² Waldmann’s modular design afforded several planar chiral Rh complexes built around a bicyclic Cp core generated via a Cu(I) catalyzed [6+3] cyclization between fulvene 71 or 74 and imine 72 (Scheme 9). In the Cu(I) catalyzed annulation, ferrocene ligand L1 provided the endo-selective cyclized products 73a-h and 75o, whereas the phosphine ligand L2 resulted in the exo-selective cyclized ligands 75. In some cases, recrystallization was required to obtain pure enantiomers. Ligands 73i-k, 75f and 75q required an additional step to alkylate the piperidine nitrogen (Scheme 9, Ligand Alkylation). The enantiopure ligands were complexed with [Rh(C₂H₄)₂Cl₂]₂ to supply complexes 60a-k and 61a-r, whose resultant isomers were resolved via chromatography on neutral alumina under argon. Waldmann and coworkers demonstrated the potency of modular synthesis for delivering a large quantity of planar chiral catalysts by combining expedient synthesis with efficient resolution techniques.

Scheme 9.

Waldmann’s modular approach to a family of planar chiral Rh Cp complexes. Planar chirality not designated as individual assignment varies according to the substitution pattern.

Perekalin and coworkers have also synthesized several planar chiral Cp derived rhodium complexes (Scheme 10).^37–39 While Perekalin et al. detail two chiral complexes in their 2018 report, only complex 62 was resolved and utilized in stereoselective synthesis.³⁷ 62 (Scheme 10A) can be accessed in three steps from tert-butyl acetylene and [Rh(COD)Cl]₂, using a [2+2+1] cyclotrimerization to form the scaffold on Rh(I). Racemic 62 was resolved by forming the (S)-prolinate complex, which was selectively crystallized to deliver pR-77 in 98:2 dr. Ultimately, enantiopure (pR,pR)-62 was reconstituted with HI.

Scheme 10.

A) Synthesis of Perekalin and coworkers Cp catalysts via [2+2+1]cyclization with chiral resolution. B) An (R)-myrtenal derived ligand for diastereoselective complexation. C) Perekalin’s planar chiral Rh(III) complex derived from on-metal cyclization of ^tBu acetylene.

In 2020, Perekalin et al. added to the catalogue of Cp derived catalysts with a Cp ligand accessed from the terpene (R)-myrtenal.³⁸ Subjecting (R)-myrtenal to a Grignard addition and oxidation delivered ketone 78 (Scheme 10B). A Wittig olefination provided the diene 79, and subsequent cyclopropanation of the terminal olefin and Skattebøl rearrangement delivered cyclopentadiene 81 (Cp^myrH) and decomposed allene side-product. While 81 was not isolated, reduced temperature and diluted conditions improved the ratio of product to allene from 3:1 to 8:1. It is also noteworthy that the impurity was separated when 81 was complexed. Complexation of 81 proceeded in a straightforward manner with [Rh(COD)OAc]₂ in MeOH/Toluene. Due to the dimethyl bridge on the fused bicycle of the Cp^myr ligand, Rh only complexed to the unhindered face of the Cp ring, delivering a single diastereomer of the Rh(I) monomer complex pS-63 and eliminating the need for advanced resolution techniques. The Rh(I) complex was then exposed to Br₂ or I₂ to deliver the Rh(III) dimer. The ligand Perekalin showcases in this Cp complex requires extensive synthesis, but its bridged dimethyl precludes any need to separate diastereomers, demonstrating a useful strategy in planar chiral catalyst resolution.

To round out their planar chiral Cp catalogue, Perekalin et al. published catalyst 65 (Scheme 10C) in 2021.³⁹ Similar to their 2018 report, the new catalyst scaffold is also formed from cyclization of tert-butyl acetylene. In this case, cyclization of tert-butyl acetylene with [Rh(COD)Cl]₂ at room temperature with Rh(I) provided complex 83, which was oxidized to the Rh(III) chloride monomer 65 with Cl₂ gas. The racemic complex was then resolved by complexation with (R)-phenylglycinol and separation on preparatory TLC. Finally, enantiopure pR-65 was reconstituted with HCl.

Also in 2021, Perekalin demonstrated a similar diastereomeric complexation strategy to isolate the diastereomers of a planar chiral Rh(I) complex featuring a tetrafluorobenzobarrelene ligand (Scheme 11).⁴⁰ Racemic complex 66, which was accessed in two steps from 1,4-disubstituted benzenes and pentafluorobromobenzene (84 and 85), selectively produced one diastereomeric complex when reacted with Na[S-Salox], allowing for chromatographic separation of the diastereomers. Subsequent exposure to HCl restored both enantiopure Rh(I) complexes in up to >99% ee for the (pR,pR) enantiomer and up to 94% ee for the (pS,pS) enantiomer (>99% after recrystallization). In a later publication, the Perekalin group also found the racemic complex 66 could be selectively complexed to the Salox ligand in situ, bypassing the need for diastereomeric separation entirely.⁴¹ We note that this is the only example of a planar chiral diene ligand in this review; however, another recent review explores this class of ligands in further detail.⁴²

Scheme 11.

Synthesis of tetrafluorobenzobarrelene Rh catalyst by the Perekalin Group

Work from the Perekalin lab significantly advanced the overall strategy of directly cyclizing and complexing simple substrates onto metal centers. Coupling this strategy with the complexation of a simple, labile chiral ligand allowed their syntheses to remain concise while utilizing facile resolution techniques.

In 2020 Blakey et al. disclosed a planar chiral indenyl catalyst for enantioselective amidation (Scheme 12).⁴ Indene 92 was prepared from 2-methylindan-1-one in 81% yield. The prochiral indenyl ligand was complexed with [Rh(COD)Cl]₂ to provide racemic complex 93, which was then resolved via chiral HPLC. Each Rh(I) COD enantiomer was then oxidized to the diiodide dimer pre-catalyst 67 with molecular iodine.

Scheme 12.

Synthesis of Blakey planar chiral rhodium indenyl catalyst relying on chiral HPLC for resolution.

Wang and coworkers disclosed their own class of planar chiral Cp Rh(III) catalysts in 2022 featuring easily tunable, nonchiral ligand scaffolds (Scheme 13).⁴³ These ligands were based off an already existing scaffold first synthesized by Takahashi et al. in 1992;⁴⁴ however, their synthesis was extended to the Rh(III) catalyst. The synthesis began from methylacrylates 94, which were brominated and converted to the corresponding Wittig salt using PPh₃, followed by [3+2] annulations from haloacetone derivatives 97 to form the substituted cyclopentadienyl ligand cores 98. Subsequent transesterification with (L)-menthol followed by complexation with Rh(COD) provided Rh(I) diastereomers 100 that could be easily resolved via column chromatography. The enantiopure COD monomer was oxidized to deliver pre-catalyst 68. However, 100 could also be further diversified via exposure to organolithium reagents, which also cleaved the chiral ester during the subsequent addition reactions. The resultant alcohols 101 were either oxidized to pre-catalyst or subjected to etherification then oxidized, in both cases delivering a family of 69 analogues.

Scheme 13.

Wang and co-workers’ planar chiral Rh(III) Cp complex. Both nonchiral ligands and diastereomeric ligands were used to prepare the Cp complexes.

2.4. Cobalt Complexes

Only two planar chiral cobalt complexes have been synthesized and used in stereoselective reactions (Scheme 14); however, the few complexes synthesized have proven versatile across an array of [2+2+2] cyclization reactions. Schumann et al. synthesized the 1-menthyl and 1-neomenthyl ligands (103ab, Scheme 14) later used for synthesis of cobalt indenyl complexes.⁴⁵ Heller et al. complexed these ligands with [Co(PPh₃)₃Cl] to form diastereomers that were then separated via chromatography to provide enantioenriched complexes pR-104 and pS-104 respectively.⁴⁶ Jungk et al. further derivatized the complex developed by Heller and coworkers by exchanging the COD ligand with P(OEt)₃ which provides complex pS-105 in 96% yield from pS-104.⁴⁷ The menthyl substituted indenyl complexes have complex structures but can be easily prepared from commercially available menthol. In this case, complexity aids in resolution while not detracting from synthetic yield.

Scheme 14.

Synthesis of planar chiral menthyl indenyl cobalt catalysts bearing cyclooctadiene or phosphite ligands.

3. Enantioselective Catalysis with Late Transition Metal Planar Chiral Complexes

We have discussed the variety of methods used to deliver and resolve planar chiral catalysts containing late transition metals. We will now focus on the applications of these catalysts in enantioselective transformations. There are a range of stereoselective substitution, annulation, polymerization, and hydrogen transfer reactions that have been disclosed over the last twenty years.

3.1. Ruthenium Catalysed Enantioselective Transformations

The largest and most expansive body of work with planar chiral ruthenium catalysts began with Takahashi and coworkers.²⁵ Initially reported in 2001, the focus of the work in the context of enantioselective catalysis was the study of tethered catalysts pS-12a-e and untethered catalyst pS-13 (Fig. 6). The catalyst platform has been optimized for allylic substitution reactions; specifically, the enantioselective substitution of racemic allylic carbonates (106) with simple amine (107) and sodium malonate (108) nucleophiles (Table 1). In the case of both nucleophiles, the tethered catalysts outperformed the non-tethered catalyst in either yield, selectivity, or both. In the case of allylic substitution with amines, all catalysts provided 90–98% yield while catalyst pS-12e afforded the highest selectivity of 75% ee. The same catalyst, pS-12e, imparted the highest yield and selectivity in the allylic substitution with sodium malonate nucleophiles. Interestingly, in both C–N (107 to 109) and C–C (108 to 110) bond formation, a simple change in the -R group of the catalyst from a methyl (pS-12a) to a tert-butyl (pS-12b) group led to a reversal in the observed selectivity of the product, while maintaining the stereochemistry of the catalyst.

Figure 6.

Planar chiral ruthenium cyclopentadienyl catalysts used in enantioselective allylic substitutions and polymerization reactions.

Table 1.

Initial study of planar chiral tethered and untethered Ru catalysts in allylic substitution

Catalyst	Nuc.	Conv. (%)	Yield (%)	Selectivity (% ee)
pS-12a	107	100	97	35(−)
pS-12b	107	98	90	20(+)
pS-12c	107	100	99	64(+)
pS-12d	107	100	93	57(+)
pS-12e	107	100	98	74(+)
pS-13	107	93	86	65(+)
pS-12a	108	98	96	80(R)
pS-12b	108	97	91	91(S)
pS-12c	108	99	97	96(S)
pS-12d	108	100	98	97(S)
pS-12e	108	100	99	96(S)
pS-13	108	10	3	n.d.

Following the initial report of the catalyst platform, Onitsuka et al. next explored a kinetic resolution of allyl carbonates (111) through an asymmetric allylic alkylation akin to their previous report from 2001.⁴⁸ As was the case in the initial disclosure of catalytic results, catalyst pS-12a and catalyst pS-12e were utilized to achieve each enantiomer of the desired product 112 (Table 2). Optimization of the ratio between racemic pentenyl carbonate starting material 111 and sodium malonate nucleophile 108 was the primary requirement for modulating the observed enantioselectivity in the product and enantio-enrichment of the olefin. In general, the bulkier catalyst pS-12e outperformed pS-12a in the context of enantio-enrichment of both the carbonate starting material and malonate product. In contrast, catalyst pS-12a had the highest conversion of starting material (80%) with similar observed selectivities of product (up to 95% ee), but poor resolution of recovered carbonate.

Table 2.

Initial study of planar chiral tethered and untethered Ru catalysts in allylic substitution

Catalyst	Ratio 111/108	Conv. 111 (%)	Recov.111 (% ee)	112 (% ee)
pS-12e	0.2	16	10(R)	80(S)
pS-12e	0.6	48	72(R)	91(S)
pS-12e	0.8	60	94(R)	92(S)
pS-12e	1.2	69	93(R)	89(S)
pS-12a	0.2	15	10(S)	72(R)
pS-12a	0.6	44	33(S)	87(R)
pS-12a	0.8	61	51(S)	87(R)
pS-12a	1.2	80	72(S)	87(R)
pS-12e	0.8a	58	90(R)	>99(S)
pS-12a	0.8a	57	30(S)	91(R)
pS-12e	0.8b	58	50(R)	92(S)
pS-12a	0.8b	54	18(S)	95(R)

^a.Et instead of Me in 111.

^b.Ph instead of Me in 111.

Early studies in diketone substitution reactions were only compatible with setting a single stereocenter, since the harsh reaction conditions required for these transformations would likely epimerize the potential stereocenter on the diketone. The Onitsuka group set out to combat this challenge in 2015.⁴⁹ Initially, diketones (114) were utilized in place of the pre-activated sodium malonate as the nucleophile. A screen of bases showed that sodium bicarbonate allowed for the highest level of diastereocontrol (6:1) combined with high yield (99%) and enantioselectivity (97% ee) (Table 3). In the scope of the reaction, the addition of functional groups at the para position of the cinnamyl chloride 113 were well tolerated, providing 115b-d in up to 6:1 dr, 99% yield, and 97% ee. For the diketone coupling partner, aryl-alkyl diketones (115a-e: up to 7:1 dr, 99% yield, and 92% ee) were much more diastereoselective than alkyl-alkyl diketones nucleophiles (115f: 1:1 dr, 83% yield, and 92% ee).

Table 3.

Allylic substitution of styrene derivatives with asymmetric diketones

In addition to diketone nucleophiles, indoles⁵⁰ and silyl enolates⁵¹ proved to be competent nucleophiles for C–C bond formation in the arylation of cinnamyl chloride. In the study of indoles, free indole provided the best combination of yield (79%) and enantioselectivity (117a, 85% ee, Table 4).⁵⁰ N-methylation of the indole (116b, Table 4) resulted in reduced enantioselectivity, while 2-methylation of the indole (116c, Table 4) resulted in reduced yield. The reaction between allylic chlorides 113 and silyl enolates 118 furnished the corresponding ketone products 119 in up to 99% yield and 93% ee (Table 4).⁵¹ Some of the most notable products were ethereal ketone 119a, α-β unsaturated ketone 119b, and ester functionality 119c (Table 4).

Table 4.

Allylic substitution with indole and enolate nucleophiles

Onitsuka and co-workers have continued to push the limits of their catalyst platform in the context of allylic substitution reactions. They applied their catalytic tool box toward C–O bond formation through allylic etherification of 113 with phenols and carboxylates (Table 5).^8,52 This work focused on allylic substitution of cinnamyl chloride derivatives with catalyst pS-12c. Enantioselective etherification of cinnamyl chlorides was successful for compounds of various substitution patterns in up to 99% yield and 95% ee (121a-d, Table 5).⁸ Replacement of the aryl component of both the alcohol and olefin substrates in the reaction with a methyl group was successful, albeit with reduced yields and enantioselectivity. Sodium carboxylates were also effective nucleophiles (122a-d) in up to 99% yield and 97% ee.⁵²

Table 5.

Allylic substitution with oxygen nucleophiles

The Onitsuka group also utilized pS-12 for the synthesis of chiral allylic alcohols, using water as the nucleophile (Table 6).⁵³ Phenyl substituted pS-12c was able to catalyse the reaction with 99% yield and 81% ee. This report examines several variations of pS-12c and matches the most effective catalyst with substrate for each reaction, highlighting the advantage catalyst libraries offer when exploring substrate scope.

Table 6.

Allylic substitution with water as nucleophile

^aUsed 12f

^bUsed 12g

^cUsed 12e

Perhaps the most explored use of catalyst scaffold pS-12 by Onitsuka and co-workers is in the realm of C–N bond formation. After the 2001 report of the amination of 1,3-diphenylpropenes in the context of catalyst development,²⁵ the next report of C–N bond formation is the 2013 study of the substitution of allylic chlorides 113 with amide nucleophiles (124, Table 7).⁵⁴ Primary amides and protected amides were both successful nucleophiles in the reaction, providing allylic amide products 125a-d in 80–99% yield and 90–95% ee. Further synthetic utility for these products was demonstrated by carrying the product of the allylic substitution with 126, bearing a terminal olefin and α-chloroalkyl functionality, forward through an atom-transfer radical cyclization to form the γ-lactam product 127 (Scheme 15). Experimental data indicated that catalyst pS-12c was necessary to catalyse the cyclization reaction, so the reaction was optimized to complete the allylic substitution and radical cyclization in one pot. As in the initial substitution reaction, the one-pot substitution-cyclization reaction proceeded with excellent yields (up to 93% isolated yield) and selectivity (up to >99% ee and 20:1 dr).

Table 7.

Allylic substitution with nitrogen nucleophiles

Scheme 15.

Asymmetric allylic amination and post-functionalization cyclization.

From 2014⁵⁵ until the most recent report in 2019,⁵⁶ Onitsuka and coworkers have utilized allylic substitution of cinnamoyl chlorides extensively in the context of polymer synthesis. Initial efforts focused on the application of C–N bond formation. Model studies showed that 1 mol % of pS-12c was able to catalyse the reaction in 99% yield and 97% ee with a >20:1 branched to linear ratio. With this high yielding result in hand, the chemistry was readily applied to the monomer unit 128 to provide polymer 129 in 65% yield with excellent branched-to-linear regioselectivity (>20:1) and, after purification on silica gel, molecular weight distribution of 1.5 (Table 8). Further optimization of reaction conditions improved the yield (73%) and molecular weight distribution (1.3) of the polymer. Several other monomers were also able to undergo polymerization with molecular weight distributions of 1.3–1.8. The monomers that have undergone the greatest investigation bear a 1,1-disubstutured olefin (130) which, after polymerization (131), is amenable to ring-closing metathesis with the terminal olefin remaining after the allylic substitution reaction (Scheme 16). This post-polymerization modification has enabled the group to synthesize various optically active polymers.

Table 8.

Allylic substitution for polymerization

Conc. 128 (M)	Temp. (°C)	Yield 129 (%)	M n	Mw/Mn
0.25	30	60	11,000	1.4
0.33	30	60	11,000	1.5
0.50	30	65	17,000	1.5
0.80	30	60	10,000	1.5
0.50	35	73	19,000	1.3
0.50	40	65	14,000	1.3
0.50	25	63	15,000	1.5

Scheme 16.

Polymerization of monomers optimized for 1.3 to 1.8 weight distributions.

Kanbayashi et al. also developed a polymerization system utilizing allylic substitution that incorporated a diketone nucleophile (Scheme 17).⁵⁷ In the case of the alkylation polymer, the chiral polymeric structure that results from the initial reaction features a terminal alkene which undergoes further functionalization prior to complete characterization. The completed polymers had much higher molecular weight distributions (2.51 and 2.73, respectively) than the polymers synthesized through allylic amination reactions.

Scheme 17.

Polymerization of diketones via allylic substitutions

The next example of planar chiral ruthenium catalysis came in the DavePhos derived complexes disclosed by Faller and D’Alliessi in 2003 (Table 9).²⁶ While the initial report deals primarily with the synthesis and characterization of the catalyst, an initial test of catalytic activity was explored with the Diels-Alder reaction between methacrolein 134 and cyclopentadiene 135. In the first report, the catalyst provided 136 in up to 23% ee with high diastereoselectivity for the exo products. Further optimization of the catalyst in Faller and co-workers’ 2005²⁷ and 2006²⁸ reports explored the non-tethered phosphine ligand on the catalyst as well as the counterion (Table 9). Replacement of the original chiral phosphine with BINOL derived phosphine (S)-MonoPhos and mixed counterions improved the enantioselectivity, demonstrating up to 70% ee. However, the selectivity imparted by (S)-MonoPhos conflicted with the innate selectivity of pR-14b, resulting in ee reduction to 19%. The use of an achiral NHC ligand, which also required a switch to a tetrafluoroborate counterion, proved less beneficial, with the selectivity decreasing to 23% ee.

Table 9.

Diels Alder catalysis with tethered planar chiral ruthenium complexes

Catalyst/Ligand	Anion 1	Anion 2	exo/endo	ee (%)
pS-14b/PPh3	SbF6−	SbF6−	91/9	23
pS-14a/(S)-MonoPhos	BF4−	BF4−	93/7	65
pS-14a/(S)-MonoPhos	BF4−	SbF6−	93/7	70
pR-14b/(S)-MonoPhos	SbF6−	SbF6−	92/8	19
pR-15/NHC	SbF6−	SbF6−	95/5	23

^*Only pS stereoisomers shown for 14 and 15

We next highlight Shvo type Cp complexes that have primarily been used in asymmetric transfer hydrogenations (Fig. 7). Yamamoto et al. reported the synthesis and use of a spirocyclic C-riboside ruthenium complexes 16a-b used in the hydrogenation of acetophenone 137 (Scheme 18).²⁹ Despite the presence of a bulky spirocycle, the planar chiral environment around the active portion of the catalyst seems to play a larger role in enantioinduction, as more symmetric versions of the catalyst failed to induce enantioselectivity. The use of pre-catalysts 16a-b provided up to 100% conversion of acetophenone 137 to the phenylethyl alcohol 138, but with modest selectivity (up to 21% ee, Scheme 18).

Figure 7.

Ruthenium catalysts designed for stereoselective hydrogen transfer reactions.

Scheme 18.

The use of spirocyclic C-riboside derived Shvo catalysts in asymmetric hydrogen transfer reactions.

The ruthenium catalysts from the Wills group were, ultimately, less effective at inducing enantioselectivity (up to 17% ee) than either the iron counterparts (up to 25% ee) or the spirocyclic Yamamoto²⁹ catalysts (Table 10).³⁰

Table 10.

Hydrogen Transfer Hydrogenation with bicyclic Ru Shvo Type Catalysts

Catalyst	Temp (°C)	Time (h)	Conv. (%)	Selectivity (% ee)
pR-17b	60	160	50	12
pS-17b	60	160	61	3
pR-17a	60	18	31	17
pR-17c	60	150	48	5

A suite of more effective catalysts was delivered by Dou and Hayashi in the form of several planar chiral ruthenium Shvo catalysts (18a-c) for the asymmetric transfer hydrogenation of ketimines 139 and various ketones 141 (Table 11).³¹ The hydrogenation of ketimines 139a-c provided secondary amine products 140a-c in up to 99% yield and 64% ee. In these results, the assertion by Yamamoto²⁹ and Hopewell³⁰ that the steric environment around the catalytically active C–O bond is most important for asymmetric induction is further supported by the use of catalyst 18c. The larger 9-anthracenyl substituent affords amine product 139a in the highest selectivity (64% ee). The two best catalysts, 18a and 18c, were then used in the hydrogenation of ketones 141a-d (Table 12). In addition to examining acetophenone 141d, the catalysts were competent with trifluoromethyl acetophenone 141a, an α-β diketo system 141b, and an α-β unsaturated α-β diketo compound 141c in up to 99% yield and 56% ee.

Table 11.

Stereoselective hydrogen transfer reactions of imines with traditional planar chiral Shvo catalysts

Entry	Catalyst	Product	Yield (%)	Selectivity (% ee)
1	(pR, pR)-18a	140a	99	54 (R)
2	(pR, pR)-18a	140b	99	45 (S)
3	(pR, pR)-18a	140c	36	62 (S)
4	(pR, pR)-18b	140a	99	47 (R)
5	(pR, pR)-18c	140a	79	64 (S)

139a/140a: Ar1 = Ar2 = Ph, 139b/140b: Ar1 = Ph, Ar2 = 4-MeO-Ph, 139c/140c: Ar1 = 2-Me-Ph, Ar2 = Ph

Table 12.

Stereoselective transfer hydrogenation of ketones with traditional planar chiral Shvo catalysts

Entry	Catalyst	Product	Yield (%)	Selectivity (% ee)
1	(pR, pR)-18a	142a	99	56 (S)
2	(pR, pR)-18a	142b	91	53 (S)
3	(pR, pR)-18a	142c	99	46 (S)
4	(pR, pR)-18a	142d	27	4
5	(pR, pR)-18c	142a	99	40 (S)
6	(pR, pR)-18c	142b	99	17 (S)
7	(pR, pR)-18c	142c	99	26 (S)
8	(pR, pR)-18c	142f	80	3

141a/142a: R₁ = Ph, R₂ = CF₃, 141b/142b: R₁ = Ph, R₂ = CO₂Et, 141c/142c: R₁ = (E)-PhCH=CH, R₂ = CO₂Me, 141d/142d: R₁ = Ph, R₂ = Me

We will close this section with discussion of the non-tethered Ru-arene complexes 19 and 20. The Perekalin group was able to demonstrate the catalytic ability of their camphor-derived ruthenium complex in the asymmetric hydrogenation of acetophenone (Scheme 19).³² By combining chiral catalyst (pR,pR)-19 with achiral N-tosyl-1,2-ethylenediamine, they prepared (S)-1-phenylethanol in 86% yield and 64% ee; however, reducing the catalyst loading from 5 mol% to 0.5 mol% provided a boost in enantioselectivity up to 85% ee, though with a loss in yield, down to only 38%. To date, these are the most effective planar chiral catalysts for enantioselective transfer hydrogenation reactions.

Scheme 19.

Asymmetric hydrogenation of acetophenone with camphor derived Ru catalyst.

The Wang group applied their paracyclophane-derived Ru catalyst very effectively to the synthesis of axially chiral isoquinolones 145, featuring yields up to 99% and ee’s up to 96% (Table 13).³³ The reaction was effective for a variety of N-methoxybenzamides 143, though steric hindrance did diminish ee despite good yields (145d). Several aryl alkynes were also tolerated in the reaction, though the introduction of a heterocycle did lead to diminished reactivity and enantioselectivity (145e).

Table 13.

Annulation reaction using paracyclophane Ru catalyst

3.2. Iron Complexes

Currently the only planar chiral iron catalysts utilized in asymmetric transformations are adaptations of pre-existing Shvo catalysts.³⁰ The Wills group explored the majority of these complexes in the early 2010s. Complexes 46a,b,d were successful in the asymmetric transfer hydrogenation of acetophenone 137, providing alcohol product 138 in up to 91% yield and 15% ee (Table 14). Complexes pS-46b and pR-46d were superior for enantioinduction (up to 25% ee) albeit with moderately reduced yields (up to 80% yield).

Table 14.

Hydrogen transfer hydrogenation with Wills’ catalysts in 2012

Catalyst	Temp (°C)	Time (h)	Conv. (%)	Selectivity (% ee)
pS-46a	28	48	25	15
pS-46a	40	48	36	10
pR-46a	40	96	10	10
pS-46b	28	48	40	25
pS-46b	28	96	69	23
pR-46b	28	48	46	11
pS-46b	40	96	80	23
pR-46b	40	48	91	11
pR-46d	40	96	66	25
pS-46d	40	96	17	5

Continuing with the asymmetric hydrogenation of acetophenone, Wills and co. later explored several different ligands on their Shvo-type iron catalyst (47–48).³⁵ Solvent variations applied based on whether the transformation was an asymmetric hydrogenation or an asymmetric transfer hydrogenation (Table 15). Optimization of the activator influenced the yield in most cases, often with minimal changes in selectivity. The notable exception was asymmetric hydrogenation versus transfer hydrogenation using catalyst pS-47b, which showed the highest selectivity of any transformation in this report (24% ee) albeit with one of the lowest yields (4% yield). The carbocyclic variants of the first-generation catalysts tended to be more active than the original catalysts; however, lower enantioselectivities were observed in almost all cases.

Table 15.

Hydrogen transfer hydrogenation by Wills group in 2018

Catalyst	Activator	Solvent	Conv. (%)	Selectlvlty (% ee)
pR-47a	K2CO3	IPA/H2O	68	7.4
pR-47a	TMAO	IPA/H2O	>99	6.2
pR-47a	TMAO	FA/TEA	91	8.1
pS-47a	K2CO3	IPA/H2O	82	11.6
pS-47a	TMAO	IPA/H2O	>99	12.2
pS-47a	TMAO	FA/TEA	88	9.9
pS-48	K2CO3	IPA/H2O	>99	9.2
pS-48	TMAO	IPA/H2O	>99	4.2
pS-47b	TMAO	IPA/H2O	>99	0.4
pS-47b	TMAO	FA/TEA	4	24.2

The latest addition to planar chiral Fe catalysis was delivered by the Gennari group, in the form of fused Cp catalyst 49 and triphenyl cyclopentadienone 50.³⁶ These catalysts were applied towards traditional ketone substrates for asymmetric hydrogenation (Table 16). While pS-49 maintained the better activity in terms of yield and to a lesser degree selectivity, unfortunately, substrate selection heavily influenced enantioinduction and most substrates translated into poor ee for the chiral alcohol product. The small scope of ketimine substrates were more productive targets for enantioselective hydrogen transfer catalysis (Table 17). Conversion remained good among all products with catalyst pS-50 providing consistently higher enantioselectivities (18 – 54% ee). While these catalysts were not applied towards a wide range of substrates, Gennari et al. displayed the possibility of Shvo type catalysts to expand beyond the realm of poor enantioinduction and maintain efficient catalyst development.³⁶

Table 16.

Hydrogenation of ketones by the Gennari group

Table 17.

Hydrogenation of ketimines by the Gennari group

3.3. Rhodium Complexes

There is a wide breadth of rhodium catalysts that have been studied in enantioselective transformations. To our knowledge, the earliest example is from 2002, where Shumann et al. utilized a 1-menthyl-4,7-dimethylindenyl rhodium(I) COD pre-catalyst 59 for the enantioselective hydrogenation of itaconic acid (150, Scheme 20).⁵⁸ This paper examined several 2-menthyl variations for their catalytic activity, but they found that only 3-substituted pR-59 and a variant with the 2-menthyl group substituent at the 2-position of the indene were both catalytically active with some enantioselectivity. Planar chiral pre-catalyst pR-59 afforded the hydrogenation product 151 in a modest 16% ee.

Scheme 20.

Rh-catalyzed hydrogenation of itaconic acid.

The next series of planar chiral complexes comes from the work of Antonchick, Waldmann, and coworkers as they developed a tunable catalyst platform to improve upon the established C₂ symmetric chiral Cp ligands of the Cramer¹, and You³ groups as well as the biotin/streptavidin construct from the Rovis lab.⁵⁹ To this end, they developed a series of catalysts (60–61, Fig. 9) and explored the generality of this platform to catalyse many different reactions.²

Figure 9.

The first planar chiral Rh indenyl catalyst used in asymmetric hydrogenation and two Jas Cp pre-catalysts used in a variety of transformations.

Antonchick and Waldmann first explored ortho C–H functionalizations of aryl hydroxamates with a variety of nucleophiles.² Annulation of 152 with styrenes, cyclic alkenes, and heteroaromatics such as benzofuran and benzothiophene were shown to deliver the cyclized products 154a-e in up to 90% yield, 93% ee, and most often with excellent regioselectivity (Table 18). Allylation of methyl benzohydroxamate 143 was also effective in generating ortho functionalized 156 in up to 91% yield and 94% ee (Table 19). The group also explored the ability to generate axially chiral biaryl 159 by utilizing diazonaphthoquinones 158. Various functionality around the benzamide was tolerated providing biaryl atropisomers in up to 93% yield and 91% ee (Table 20).

Table 18.

Annulation reactions with Rh JasCp catalyst

Table 19.

Allylation reaction with chiral Rh JasCp catalyst

Table 20.

Enantioselective carbene insertion reaction with chiral Rh JasCp catalyst

In another recent disclosure of the Antonchick and Waldmann planar chiral pre-catalyst, yet another variation on the tunable platform (pR-61b) was found to be the optimal catalyst for the spiroannulation of α-arylidine pyrazolones with alkynes to enantioselectively generate spiropyrazolones.⁶⁰ Initial optimization and scope of the α-arylidine pyrazolones 160 and alkynes 161 afforded spirocyclic products 162a-h in up to 94% yield and 97% ee (Table 21). One primary goal of this report was to apply this reaction to the late-stage modification of drug and natural products. To this end, alkynes featuring steroids, terpenes, and other chiral functionalities were examined and successfully afforded their respective spirocyclic products. These reactions all proceeded in similar yield and enantiomeric excess regardless of which enantiomer of the catalyst was utilized.

Table 21.

Enantioselective carbene insertion reaction with chiral Rh JasCp catalyst

Over the next several years, the focus on planar chiral rhodium complexes as catalysts for enantioselective catalysis was mostly developed in the context of pre-catalysts bearing a planar chiral Cp ring. Perekalin and co-workers have contributed significantly in developing planar chiral Rh catalysts as well as evaluating their activities (Fig. 10). The first publication from the Perekalin group on this front came in 2018, with Rh(III) diiodide complex (pR,pR)-62.³⁷ This pre-catalyst was utilized in an annulation reaction between aryl hydroxamic acids (163) and alkenes (164) to generate dihydroisoquinolones (165a-m, Table 22). This catalyst system relied on steric blocking interactions for the induction of enantioselectivity. The reaction worked extremely well in most cases with up to 97% yield and up to 95% ee observed for internal alkenes (Table 22). Terminal alkenes worked as well, but with reduced ee (up to 47%).

Figure 10.

Rhodium catalysts developed by the Perekalin Group.

Table 22.

Annulation reaction with di-tert-butyl Cp Rh^III catalyst

Recent examples of planar chiral rhodium catalysts for enantioselective catalysis are complexes (pS,pS)-64 and pR-65 from Perekalin and coworkers.^38,39 Catalyst (pS,pS)-64 features a myrtenal derived ligand that was utilized for the annulation of benzamide 166 with simple alkenes (164a-d). Under the conditions examined, the catalyst can successfully form products 165a-c in high yield (77–93% yield) with a wide range of enantioselectivities (16–94% ee, Table 23). Catalyst pR-65 was utilized under similar conditions to generate compounds 165b and 165d in low to moderate yields and selectivities.³⁹ The two catalysts can be compared head-to-head with compound 165b where pR-65 afforded a much lower yield than (pS,pS)-64a (35% versus 93%) but in a much higher, albeit still only moderate, selectivity (58% ee versus 30% ee).

Table 23.

Enantioselective annulation with a myrtenal-derived Rh catalyst

Perekalin and coworkers also utilized their Rh(I) tetrafluorobenzobarrelene catalyst (66, Fig. 10) in the addition of donor-acceptor diazo compounds 167 into B–H and Si–H bonds (Scheme 21).⁴⁰ The demonstrated reactivity is excellent for carbene additions into a borane N-methylpyrrolidine adduct (168), offering yields of 83–99% and ee’s ranging from 89–98% for a range of diazo compounds. Reactivity was also good for carbene additions into triethylsilane (170), with yields of 78–89% and ee’s of 87–97%. However, the catalyst demonstrated substantially diminished enantiocontrol when reacting with N–H bonds, giving no enantioselectivity when reacting with benzamide (174a) and only providing 33% ee with phthalimide (174b) and 42% ee with carbazole (174c).

Scheme 21.

Rh carbene insertion into B–H, Si–H, and N–H bonds.

The next planar chiral pre-catalyst system comes from the Blakey group. In this report, (pS,pS)- and (pR,pR)-67 were utilized in the enantioselective allylic C–H amidation reaction of several activated and unactivated olefins (175, Table 24).⁴ This catalyst variation is the only planar chiral indenyl catalyst where the source of chirality arises exclusively from planar chirality. Data obtained from a crystal structure and the computational support shows that the indene begins to “slip” from an η⁵ coordination to an η³-like coordination where the phenyl moiety of the indene lifts off from the rhodium. This elicits an electronic asymmetry in the π-allyl intermediate via the trans effect which allows for reductive elimination of a Rh(V) imido intermediate to take place at the longer Rh–C bond. This chemistry was amenable to a variety of terminal (177a,h,j) and internal (177e,f), symmetrical (177e,f) and asymmetrical olefins (Table 24). Separate from the dioxazolone scope, the use of phthalimidoyl glycine dioxazolone alongside tert-butyl dioxazolone in the olefin scope demonstrates that the yield and ee are the result of a subtle balance of steric and electronic effects, ultimately providing products in up to 90% yield and 98% ee (177f-k).

Table 24.

Enantioselective allylic amidation with planar chiral indenyl catalyst

^a: Uses (pS,pS) version of catalyst

Finally, Wang et al. evaluated several catalysts in a pair of enantioselective C–H activation reactions (Table 25).⁴³ While many variations of 68 and 69 were tested, 68a proved most effective in the [2+2+2] annulation of 2-indolinones (178) with diarylacetylenes (179) to form axially chiral N-aryloxindoles (180). This reaction proceeded in 45–95% yields with ee up to 78–94% (absolute stereochemistry not defined, Table 25). In the exploration of annulations with O-Boc hydroxamates and norbornenes, 68b proved to be the more competent catalyst. In this case, the reaction proceeded in 62–92% yield with 65–93% ee’s (Table 26). In both cases, the authors note the importance of steric bulk around the pendant methylene unit of the catalyst, as well as modified reactivity between the free hydroxyl and the ether linkage.⁴³

Table 25.

Enantioselective [2+2+2] cyclization with Wang’s Rh^III catalyst

Table 26.

Enantioselective annulation with Wang’s Rh^III catalyst

3.4. Cobalt Complexes

To date, only two planar chiral cobalt complexes have been engaged in enantioselective catalysis (Fig. 12). While there is a limited number of Co catalysts, they have proven quite broad in their reactivity, mainly focused on annulation reactions. To begin examining cobalt complexes, Gutnov et al. first explored planar-chiral pR-104 in the context of [2+2+2] cycloadditions for the synthesis of atropisomerically enriched 2-arylpyridines (185a-c, Table 27).⁶¹ A total of six cobalt catalysts were explored but only, pR-104 and pS-104 featured planar chirality. In all cases of initial catalyst investigation where there is a head-to-head comparison, planar-chiral catalysts pR-104 and pS-104 were more selective than the chiral catalysts that didn’t feature planar chirality, albeit with reduced yields. Fortunately, the Heller group had more success when applying pS-104 towards intramolecular cyclization of diyne 186 under 420nm light (Table 27). They observed substantially higher yields and great selectivity (88% – 93% ee). Later the Hapke group investigated whether modifying the ligands of 104 would enable thermal activation.⁴⁷ Initial tests with pS-105 showed that irradiative activation provided a much higher yield than thermal activation (86% and 20% respectively) and better selectivity (93% ee and 66% ee respectively, Table 27). With this information in hand, several nitriles were examined under the best thermal conditions, and successfully generated product, albeit in low to trace yields, with moderate to good selectivity.

Figure 12.

The only planar chiral Co catalysts used in asymmetric catalysis to date.

Table 27.

Co-catalyzed atroposelective [2+2+2] cycloaddition

^[a]50 °C

^[b]toluene, 80 °C

Having established a system whereby a cobalt catalysed [2+2+2] cyclization could generate axially chiral 2-aryl pyridines in good yield and enantiomeric excess, Heller et al. sought to apply this system to phosphorus bearing axially chiral biaryls (189, Table 28).⁶² The phosphoryl moiety subsequently would offer a handle for reduction to the phosphine, providing an efficient synthesis of axially chiral phosphine ligands for use in asymmetric catalysis. The menthyl functionalized cobalt catalysts proved invaluable in developing this technology. Optimization showed that the planar chiral variants pR-104 and pS-104 were, once again, the optimal catalysts in the generation of stable atropisomers. Each substrate required some optimization of temperature and time to find the best conditions and products were generated in 56–83% ee in the reaction mixture, with most products subsequently crystallized to >99% ee (Table 28). Mechanistic work was conducted to demonstrate that the cobalt catalyst must react with two alkynes to generate an intermediate similar to 191 (Fig. 13), which will have oxidized the cobalt from Co(I) to Co(III), allowing for coordination to the nitrile followed by migratory insertion and metallocycle ring expansion.

Table 28.

Co-catalyzed atroposelective phosphoryl [2+2+2] cycloaddition

Product	R	Catalyst	Yield (%)	Selectivity (%) ee
(R)-190a	Ph	pS-104	45	63
(S)-190a	Ph	pR-104 [b] [e]	49	79
(S)-190b	p-OMePh	pR-104 [b]	51	72
(S)-190c	3,5-(CF3)Ph	PR-104 [b]	24	56
(S)-190d	tBu	PR-104 [e]	61	82
(S)-190e	1-adamantyl	PR-104 [f]	44	83
(S)-190f	NMe2	PR-104 [a] [c] [d] [f]	80	68

^[a]1 mol % cat.

^[b]45 °C

^[c]55 °C

^[d]12 h

^[e]48 h

^[f]toluene

Figure 13.

Co(III) intermediate of [2+2+2] cycloaddition

To this point, 104 had only been utilized in the introduction of axial chirality through the [2+2+2] cyclotrimerization reaction. While Heller et al. were utilizing axially chiral ligands generated from their earlier reports, they examined the nickel-catalyzed [2+2+2] cyclotrimerization of triyne 192 to form helicene 193 (Table 29).⁶³ Nickel has been demonstrated as a privileged metal for this transformation; however, the group also explored the use of their Co(I) catalyst pS-104 in the transformation. Optimization of time and temperature ultimately provided >99% yield for the transformation with minimal enantioinduction of 25% ee. Comparatively, the use of axially chiral (−)-(aS)-NAPHEP, the reduced form of 190a, as the ligand in the Ni(0) catalysed transformation afforded >99% yield and 64% ee, the highest observed for this reaction.

Table 29.

Co-catalyzed cyclotrimerization to form [7]-helicene

Run	Catalyst	Time/Temp	Yield (%)	Selectivity (%) ee
1	pS-104	18/−10	>99	25
2	pS-104	20/25	>99	20
3	pS-104	20/10	>99	20
4	pS-104	20/0	>99	9
5a	Ni(COD)2	20/−20	>99	64

^[a]20 mol % (-)-(aS)-NAPHEP

The formation of axially chiral biaryls has ultimately been the primary focus for the development of Co(I) complex 104.⁴⁶ While previous results had demonstrated the privileged nature of the methoxy substituent on the naphthyl ring for generating high yield and ee,^61,62 the dramatically increased steric bulk of the ring formed in the [2+2+2] cyclotrimerization of 194 leads to reduced yield and selectivity. During this investigation of rotational stability, pR-104 was utilized in the formation of a 2-naphthylpyridine 195 where the pyridine is a part of a 6-5-6 fused ring system.⁶⁴ While the catalyst did provide stereoinduction, the reaction proceeded in low yield (17%) and with minimal selectivity (44% ee, Scheme 22).

Scheme 22.

Co-catalyzed intermolecular [2+2+2] cycloaddition

The Hapke group also investigate the ability of 104 to install two stereocenters in one annulation.⁴⁷ The intramolecular [2+2+2] cyclotrimerization of 196 to set two atropisomeric stereocenters (197) was also carried out under thermal conditions providing a reasonable yield of 77% but only 7% ee (Scheme 23). All these results, taken together, provided the first example of this sort of complex undergoing thermal activation instead of irradiative, while also demonstrating that the original COD complex 104 is a more efficient catalyst under irradiative conditions than the phosphite bearing complex 105.

Scheme 23.

Co-catalyzed di-[2+2+2] cycloaddition to set two atropisomeric stereocenters

4. Conclusions

We have presented a thorough overview of the use of late transition metal planar chiral complexes as catalysts for enantioselective transformations. Planar chirality remains a largely underutilized source of chiral induction, as the use of planar chiral complexes as catalysts provides unique opportunities for electronic and steric control in reaction design. However, synthetic limitations have restricted systematic exploration and the range of reactions examined remains relatively narrow. Currently, balancing ligand synthesis with the ease of resolution of the resultant planar chiral complexes remains a challenge in the field. Generally, more complex ligands bearing point chirality offer easier purification of the diastereomeric planar chiral catalysts, while simpler ligand designs require more challenging resolution of the enantiomeric complexes. Therefore, the development of robust stereoselective methods for the synthesis of planar chiral complexes would represent a significant advance. We expect that a more thorough understanding of planar chirality will allow for the expanded use and development of late transition metal planar chiral complexes to induce enantioselectivity in new and exciting ways.

Figure 4.

Planar chiral iron catalysts that have been used for enantioselective transformations.

Figure 8.

Planar chiral iron catalysts used in asymmetric hydrogen transfer reactions.

Figure 11.

Rh indenyl catalyst from the Blakey group and two Rh Cp catalysts from the Wang Group