On the Structure of Chiral Dirhodium ( II ) Carboxylate Catalysts : Stereoselectivity Relevance and Insights

Modern experiments have offered alternative interpretations on the symmetry of chiral dirhodium(II) carboxylate complexes and its relationship to their level of enantioselectivity. So, this contribution is to provide an insight on how the knowledge around the structure of these catalysts has evolved with a particular emphasis on the impact of this knowledge on enantioselectivity prediction and catalyst design.

Chiral dirhodium(II) carboxylate complexes constitute a very important class of dirhodium(II) catalysts.Their structure consists of two rhodium atoms connected together through a unique dirhodium bridge.This dirhodium axes is connected to four chiral electron withdrawing carboxylate ligands and they have two vacant axial positions available for carbene binding [50,51].Although several highly enantioselective dirhodium(II) carboxylate catalysts have been reported to date [2,52] and their application spectrum has expanded dramatically, limited evidence is known on how chirality is tailored at the active center of these complexes by its four chiral carboxylate ligands.As a consequence, this article is providing an insight on how the knowledge around the structure of dirhodium(II) carboxylate complexes has evolved.This is with a particular emphasis on the impact of this knowledge on enantioselectivity prediction and catalyst design.To the best of our knowledge, this topic was last reviewed by Hansen and Davies in 2008 [53] and since then subsequent investigations have offered alternative interpretations on the symmetry of chiral dirhodium(II) carboxylate complexes and its relationship to their level of enantioselectivity.Scheme 1. Dirhodium(II)-mediated chemistry of methyl styryldiazoacetate (1) with different substrates.
Figure 1.Main arrangements of ligand's blocking groups around the dirhodium(II) core and their point groups (the sterically blocking groups around the rhodium active sites are depicted as ovals) [2,4,36,53,54].
In the same context, Rh 2 (R-BTPCP) 4 was also reported as a catalyst that exhibited high enantioselectivity levels [73][74][75].For example, the Rh 2 (R-BTPCP) 4 -catalyzed cyclopropanation of styrene with methyl styryldiazoacetate generated the cyclopropane product in 92% ee [73].Based on its X-ray crystal structure, this enantioselectivity was again returned to the α,β,α,β-distribution of ligands around the dirhodium(II) core (Scheme 3) [73].To support the D2-symmetry hypothesis of Rh2(S-DOSP)4, the authors introduced the second generation prolinate-based complex, Rh2(S-biTISP)2 (Scheme 2) [13, 71,72].As the ligands themselves possess a C2-symmetry, Rh2(S-biTISP)2 was anticipated to adapt the higher D2-symmetrical structure.More importantly, the complex was expected to be more rigid at which, the ligands are conformationally locked in the α,β,α,β-arrangement around the catalyst active centers.For this catalyst, not only both faces are equivalent but also all staggered binding orientations of the axial carbene ligands are identical with respect to the approaching substrate [13,53,71].As expected by the authors, Rh2(S-biTISP)2 demonstrated excellent enantioselectivity in the enantioselective intermolecular cyclopropanation of styrene with methyl styryldiazoacetate leading to the generation of the cyclopropane product in 98% ee (Scheme 2b) [13,71].This result very well supported the proposed concept that the observed enantioselectivity of Rh2(S-DOSP)4 originates from an α,β,α,βconformation adapted during catalysis.However, the use of Rh2(S-biTISP)2 catalyst preferentially resulted in the generation of the (1S,2S)-cyclopropane product (Scheme 2b) which was the opposite to that obtained with the Rh2(S-DOSP)4-catalysed reaction (Scheme 2a).The authors returned this difference in asymmetric induction to the different preferred staggered orientation for the carbene binding which causes different channels of the metal-carbene intermediate to be exposed to the approaching alkene [13].
Catalysts 2017, 7, 347 7 of 19 in different asymmetric transformations [37,88,[91][92][93][94][95].For example, Rh2(S-PTAD)4-catalyzed cyclopropanation of styrene with α-diazobenzylphosphonate resulted into the generation of the corresponding cyclopropylphosphonate product in high levels of enantioselectivity (99% ee) compared to Rh2(S-DOSP)4 (34% ee), Rh2(S-biTISP)2 (88% ee) and Rh2(S-PTTL)4 (97% ee) (Scheme 6b) [37].Up to this point, Hansen and Davies theoretical assumption that only catalysts with α,α,β,β-and α,β,α,β-arrangements results in high enantioselectivity and not the others [53] was very well supported by scientific evidence.Nevertheless in 2009, Fox et al. were able to capture an X-ray crystal structure for Rh 2 (S-PTTL) 4 [78,87].Surprisingly, the obtained structure revealed that the catalyst is adapting the α,α,α,α-conformation in solid state and not the expected α,α,β,β-conformation (Figure 2).Furthermore, the structure showed that the catalyst is having a rectangular chiral cavity (C 2 -rather than C 4 -symmetry) with only one ethyl acetate axial ligand coordinated to the rhodium atom from the cavity face.[78,87].In this model, Rh2(S-PTTL)4 was proposed to keep its α,α,α,αconformation during catalysis and the carbene is anticipated to align itself within the wide dimension of the catalyst's rectangular cavity leaving its si-face accessible by the substrate.Also according to Fox model, the tert-butyl groups are crucial to restrict the reactivity to the chiral cavity face of the catalyst and limit the racemic background reaction.α-alkyldiazoesters (Scheme 7) [78,87].In this model, Rh 2 (S-PTTL) 4 was proposed to keep its α,α,α,α-conformation during catalysis and the carbene is anticipated to align itself within the wide dimension of the catalyst's rectangular cavity leaving its si-face accessible by the substrate.Also according to Fox model, the tert-butyl groups are crucial to restrict the reactivity to the chiral cavity face of the catalyst and limit the racemic background reaction.Based on the new information, Fox et al. introduced a new model to account for the enantioselectivity exhibited by Rh2(S-PTTL)4 in enantioselective cyclopropanation of styrene with αalkyldiazoesters (Scheme 7) [78,87].In this model, Rh2(S-PTTL)4 was proposed to keep its α,α,α,αconformation during catalysis and the carbene is anticipated to align itself within the wide dimension of the catalyst's rectangular cavity leaving its si-face accessible by the substrate.Also according to Fox model, the tert-butyl groups are crucial to restrict the reactivity to the chiral cavity face of the catalyst and limit the racemic background reaction.Fox model was very well supported by the results obtained earlier for Rh 2 (S-PTAD) 4 by being more enantioselective than Rh 2 (S-PTTL) 4 [37,[91][92][93]95,96].The results were very relevant as the much bulkier adamantyl groups would more efficiently block the carbene ligands from coordinating to the achiral rhodium face (shrouded by the adamantly groups) compared to the tert-butyl groups of Rh 2 (S-PTTL) 4 leading to the observed more enantioselective outcome.Moreover, Fürstner et al. recently succeeded in capturing an X-ray crystal structure of the Rh 2 (S-PTTL) 4 =C(MeOC 6 H 4 )CO 2 Me intermediate (Figure 3) [97].The obtained structure showed that the entire carbene resides within a long narrow pocket formed by the four N-phthalimido groups of Rh 2 (S-PTTL) 4 , that adapt an α,α,α,α-arrangement (C 2 -symmetric).The overall situation looked, therefore, in good agreement with the model proposed by Fox.But, low temperature NMR data for this intermediate returned a single set of signals for the four chiral carboxylate ligands [97].This observation was not in a good accord with a C 2 -symmetric binding site as proposed by the Fox model.Fox proposal was also challenged by data reported from different research groups.Adly, Ghanem and co-workers reported the X-ray crystal structure of Rh2(S-PTAD)4 [98].Similar to Rh2(S-PTTL)4, the obtained structure showed that the catalyst is adapting the α,α,α,α-conformation in solid state, however, the catalyst was observed to form a bis(EtOAc) adduct when crystalized from ethyl acetate/n-hexane (Figure 4).This observation contradicted with the assumption proposed by Fox et al. in their model and confirmed that the adamantyl groups are not completely blocking the "achiral" catalyst face.In other words, there is still enough room for a Lewis basic ligand to coordinate to the achiral axial rhodium site.This observation provided direct evidence that both Rh atoms are still accessible by the diazo substrates even after the introduction of the bulkier adamantyl groups [98].In addition, results have shown that complexes with tert-butyl groups at their α-carbon produced enantioselectivity levels similar and sometimes higher than Rh2(S-PTAD)4 which carries the bulkier adamantyl groups [98].
Also, different reported observations confirmed the flexibility of the PTTL ligands around the rhodium active center, especially in solution and emphasized the effect of the shape and size of the axial ligand on the overall structure of the dirhodium(II)-carbene intermediate.For example, 2D heteronuclear NOESY data reported by the groups of Charette [54] and Duddeck [99], independently, confirmed that Rh2(S-PTTL)4 and analogues have a mobile conformation in solution.This confirms the existence of other conformers with at least one N-phthalimido group is flipped downward.Further and based on X-ray crystallographic data, Rh2(S-PTTL)4(NC-C6H4-CN) adduct was found to adapt a conformation in which one of the ligands is twisted by only about 80° in solid state [87].In other words, the ligand's blocking group lies in the periphery of the catalyst, the case that was originally supposed to be problematic due to steric constraints [36,53].Gardiner and Ghanem [2,100] also succeeded to grow crystals of Rh2(S-PTPA)4(EtOAc)2 and Rh2(S-PTPA)4(MeOH)(H2O) and they found that different conformations exist for each of these adducts at which, all of them feature inequivalent axial coordination sites unlike its original α,α,β,β-structure previously reported by Hashimoto [76,[85][86][87].The Rh2(S-PTPA)4(EtOAc)2 adduct exhibited an α,α,α,β-conformation, while Rh2(S-PTPA)4(MeOH)(H2O) adduct exhibited an α,α,α,α-conformation in solid state.Fox proposal was also challenged by data reported from different research groups.Adly, Ghanem and co-workers reported the X-ray crystal structure of Rh 2 (S-PTAD) 4 [98].Similar to Rh 2 (S-PTTL) 4 , the obtained structure showed that the catalyst is adapting the α,α,α,α-conformation in solid state, however, the catalyst was observed to form a bis(EtOAc) adduct when crystalized from ethyl acetate/n-hexane (Figure 4).This observation contradicted with the assumption proposed by Fox et al. in their model and confirmed that the adamantyl groups are not completely blocking the "achiral" catalyst face.In other words, there is still enough room for a Lewis basic ligand to coordinate to the achiral axial rhodium site.This observation provided direct evidence that both Rh atoms are still accessible by the diazo substrates even after the introduction of the bulkier adamantyl groups [98].In addition, results have shown that complexes with tert-butyl groups at their α-carbon produced enantioselectivity levels similar and sometimes higher than Rh 2 (S-PTAD) 4 which carries the bulkier adamantyl groups [98].
Also, different reported observations confirmed the flexibility of the PTTL ligands around the rhodium active center, especially in solution and emphasized the effect of the shape and size of the axial ligand on the overall structure of the dirhodium(II)-carbene intermediate.For example, 2D heteronuclear NOESY data reported by the groups of Charette [54] and Duddeck [99], independently, confirmed that Rh 2 (S-PTTL) 4 and analogues have a mobile conformation in solution.This confirms the existence of other conformers with at least one N-phthalimido group is flipped downward.Further and based on X-ray crystallographic data, Rh 2 (S-PTTL) 4 (NC-C 6 H 4 -CN) adduct was found to adapt a conformation in which one of the ligands is twisted by only about 80 • in solid state [87].In other words, the ligand's blocking group lies in the periphery of the catalyst, the case that was originally supposed to be problematic due to steric constraints [36,53].Gardiner and Ghanem [2,100] also succeeded to grow crystals of Rh 2 (S-PTPA) 4 (EtOAc) 2 and Rh 2 (S-PTPA) 4 (MeOH)(H 2 O) and they found that different conformations exist for each of these adducts at which, all of them feature inequivalent axial coordination sites unlike its original α,α,β,β-structure previously reported by Hashimoto [76,[85][86][87]

Local Ligand Symmetry
Following the classical way of catalyst design, all reported chiral dirhodium(II) carboxylate complexes derived from L-amino acid ligands were designed to have a symmetric N-protecting groups for the construction of the chiral ligands [4,53].However in 2004, Müller and Ghanem reported several Rh2(S-NTTL)4 analogues at which, one hydrogen on the N-protecting group is substituted to generate catalysts carrying asymmetric N-protecting groups for asymmetric cyclopropanation of styrene with dimethyl malonate [101] and Meldrum's acid [102] (Scheme 8).The authors were surprised when screening results revealed that Rh2(S-4-Br-NTTL)4-catalyzed cyclopropanation proceeded with far improved levels of enantioselectivity compared to Rh2(S-NTTL)4.

Local Ligand Symmetry
Following the classical way of catalyst design, all reported chiral dirhodium(II) carboxylate complexes derived from L-amino acid ligands were designed to have a symmetric N-protecting groups for the construction of the chiral ligands [4,53].However in 2004, Müller and Ghanem reported several Rh 2 (S-NTTL) 4 analogues at which, one hydrogen on the N-protecting group is substituted to generate catalysts carrying asymmetric N-protecting groups for asymmetric cyclopropanation of styrene with dimethyl malonate [101] and Meldrum's acid [102] (Scheme 8).The authors were surprised when screening results revealed that Rh 2 (S-4-Br-NTTL) 4 -catalyzed cyclopropanation proceeded with far improved levels of enantioselectivity compared to Rh 2 (S-NTTL) 4 .
Unfortunately, the authors had no success in growing X-ray quality crystals for Rh 2 (S-4-Br-NTTL) 4 .Instead, the X-ray crystal structure of Rh 2 (S-NTTL) 4 (Figure 5) was used as a model to account for the enhanced enantioselectivity that accompanied Rh 2 (S-4-Br-NTTL) 4 [102].The authors proposed that, if Rh 2 (S-4-Br-NTTL) 4 is retaining a similar structure to Rh 2 (S-NTTL) 4 during catalysis (following Fox et al. proposal), the bromo-substituents would be situated at the cavity rim and are likely to exert a strong influence on the enantiofacial discrimination of the approaching substrate leading to the observed enhanced enantioselectivity.The authors also returned the improved performance of the 4-Br substituted catalyst over the 4-Cl analogue to the size of the substituents.The larger bromosubstituents are expected to exert more influence at the cavity rim leading to higher enantioselectivity in the case of Rh 2 (S-4-Br-NTTL) 4 .
Later, Adly and Ghanem investigated this further and they introduced a new approach for the design of dirhodium(II) carboxylate complexes.This approach is through the reduction of the N-protecting group symmetry by the inclusion of a bulky substituent at the catalyst's cavity rim.By the application of their new approach, the authors were able to introduce their catalyst, Rh 2 (S-tert PTTL) 4 (Scheme 9) [89,98,103].For this catalyst, the planarity of the N-phthalimido groups is maintained but the local symmetry of the N-protecting group was reduced from C 2v to C s by virtue of the substituents.Screening of Rh 2 (S-tert PTTL) 4 in different enantioselective cyclopropanation reactions demonstrated that, generally, Rh 2 (S-tert PTTL) 4 is more enantioselective than Rh 2 (S-PTTL) 4 , while returning comparable enantioselectivity to Rh 2 (S-PTAD) 4 (Scheme 9) [98].
Following the classical way of catalyst design, all reported chiral dirhodium(II) carboxylate complexes derived from L-amino acid ligands were designed to have a symmetric N-protecting groups for the construction of the chiral ligands [4,53].However in 2004, Müller and Ghanem reported several Rh2(S-NTTL)4 analogues at which, one hydrogen on the N-protecting group is substituted to generate catalysts carrying asymmetric N-protecting groups for asymmetric cyclopropanation of styrene with dimethyl malonate [101] and Meldrum's acid [102] (Scheme 8).The authors were surprised when screening results revealed that Rh2(S-4-Br-NTTL)4-catalyzed cyclopropanation proceeded with far improved levels of enantioselectivity compared to Rh2(S-NTTL)4.Scheme 8. Dirhodium(II)-catalyzed enantioselective cyclopropanation of styrene with dimethyl malonate [101] and Meldrum's acid [102].
Unfortunately, the authors had no success in growing X-ray quality crystals for Rh2(S-4-Br-NTTL)4.Instead, the X-ray crystal structure of Rh2(S-NTTL)4 (Figure 5) was used as a model to account for the enhanced enantioselectivity that accompanied Rh2(S-4-Br-NTTL)4 [102].The authors proposed that, if Rh2(S-4-Br-NTTL)4 is retaining a similar structure to Rh2(S-NTTL)4 during catalysis (following Fox et al. proposal), the bromo-substituents would be situated at the cavity rim and are likely to exert a strong influence on the enantiofacial discrimination of the approaching substrate leading to the observed enhanced enantioselectivity.The authors also returned the improved performance of the 4-Br substituted catalyst over the 4-Cl analogue to the size of the substituents.The larger bromo-substituents are expected to exert more influence at the cavity rim leading to higher enantioselectivity in the case of Rh2(S-4-Br-NTTL)4.Later, Adly and Ghanem investigated this further and they introduced a new approach for the design of dirhodium(II) carboxylate complexes.This approach is through the reduction of the Nprotecting group symmetry by the inclusion of a bulky substituent at the catalyst's cavity rim.By the application of their new approach, the authors were able to introduce their catalyst, Rh2(Stert PTTL)4 (Scheme 9) [89,98,103].For this catalyst, the planarity of the N-phthalimido groups is maintained but the local symmetry of the N-protecting group was reduced from C2v to Cs by virtue of the substituents.Screening of Rh2(Stert PTTL)4 in different enantioselective cyclopropanation reactions demonstrated that, generally, Rh2(Stert PTTL)4 is more enantioselective than Rh2(S-PTTL)4, while returning comparable enantioselectivity to Rh2(S-PTAD)4 (Scheme 9) [98].
The X-ray crystal structure of Rh2(Stert PTTL)4 showed that the catalyst is adapting the α,α,α,αconformation in solid state (Scheme 9) featuring the tert-butyl substituents similarly disposed towards the "corner" of the nearly square-shaped cavity.The cavity is approximately square if compared to the reported structure of Rh2(S-PTTL)4.The authors claimed that the added tert-butyl substituent on the N-phthalimido group in Rh2(Stert PTTL)4 reinforced the overall chiral twist of the cavity, while, at the same time, not placing added steric hindrance to the binding of axial carbene The X-ray crystal structure of Rh 2 (S-tert PTTL) 4 showed that the catalyst is adapting the α,α,α,α-conformation in solid state (Scheme 9) featuring the tert-butyl substituents similarly disposed towards the "corner" of the nearly square-shaped cavity.The cavity is approximately square if compared to the reported structure of Rh 2 (S-PTTL) 4 .The authors claimed that the added tert-butyl substituent on the N-phthalimido group in Rh 2 (S-tert PTTL) 4 reinforced the overall chiral twist of the cavity, while, at the same time, not placing added steric hindrance to the binding of axial carbene ligand [98].The authors proposed that, if this geometry is relevant during catalysis, it is likely to be the reason for the observed enhanced enantioselectivity relative to its parent, Rh 2 (S-PTTL) 4 .This led to the generation of the lower symmetry heteroleptic complex, Rh 2 (S-TCPTTL) 3 (PTAiB) [104].The X-ray crystal structure of this complex revealed that it is adapting the α,α,α,β-conformation in solid state (Scheme 10).Surprisingly, screening results demonstrated that lowering the global symmetry of the catalyst had a beneficial impact on its enantioselectivity.In the cyclopropanation reaction of alkenes with α-nitro diazoacetophenones, Rh 2 (S-TCPTTL) 3 (PTAiB) resulted into the formation of the cyclopropane product in 96.4% ee which was higher compared to its parent Rh 2 (S-TCPTTL) 4 (92.4% ee) (Scheme 10).

Conclusions
From the above demonstration, all models proposed to account for the enantioselectivity observed with different dirhodium(II) carboxylate catalysts relied mainly on the obtained X-ray crystal structures of these catalysts (except for Rh2(S-DOSP)4 and related prolinate-based complexes).However, due to crystal packing, π-stacking and/or existence of axial and lattice solvent molecules within the crystal unit cell, the static pose of a catalyst and its chiral ligands portrayed in its X-ray crystal structure does not validate the adaption of the catalyst to the same conformation in solution and during catalysis.In other words, based on the size and the shape of the axial carbene ligand, as Scheme 11.Enantioselective cyclopropanation of styrene with ethyl α-diazobutanoate and structure of Rh 2 (S-PTTL) 3 (TPA) according to its X-ray crystal structure [105].

Conclusions
From the above demonstration, all models proposed to account for the enantioselectivity observed with different dirhodium(II) carboxylate catalysts relied mainly on the obtained X-ray crystal structures of these catalysts (except for Rh 2 (S-DOSP) 4 and related prolinate-based complexes).However, due to crystal packing, π-stacking and/or existence of axial and lattice solvent molecules within the crystal unit cell, the static pose of a catalyst and its chiral ligands portrayed in its X-ray crystal structure does not validate the adaption of the catalyst to the same conformation in solution and during catalysis.In other words, based on the size and the shape of the axial carbene ligand, as well as the reaction solvent used, the metal-carbene intermediate may adapt a different conformation during catalysis to the one observed in its X-ray crystal structure.
In fact, carboxylate ligands contain conformationally mobile units [13,54,87,98,99] which might give the ability to the dirhodium(II) catalyst to adapt its conformation to the particular substrate undergoing the reaction.Then enantioselection mechanisms can mainly originate from the different preferred reaction channels as a function of the different steric interactions taking place between the approaching substrate and the formed dirhodium(II)-carbene complex.This can explain the compatibility of a particular dirhodium(II) carboxylate catalyst with a particular substrate returning high levels of enantioselectivity, while the same catalyst might not retain the same level of enantioselectivity after switching to a different substrate.As a result, catalyst optimization is always necessary in method development to find the optimum catalyst that can return the highest possible level of enantioselectivity with the particular substrate undergoing the reaction.Based on that, researchers must not only limit their focus on the four main conformations of dirhodium(II) carboxylate complexes and an expansion in the considered conformational space is essential.
Even by relying on the catalyst X-ray crystal structures alone to justify the observed enantioselectivity, it is obvious from the demonstration above that dirhodium(II) carboxylate catalysts with α,α,α,α-, α,α,α,β-, α,α,β,β-and α,β,α,β-ligand arrangements were all reported to lead to high levels of enantioselectivity.As a consequence, the original proposal that catalysts with α,α,β,β-and α,β,α,β-geometries will only result in high enantioselectivity [53] can no longer be considered as valid.In other words, symmetry of the catalyst is not a mandatory prerequisite for achieving high levels of enantioselectivity.Subsequently, the use of high symmetry ligands as the traditional way to produce high symmetry catalysts [53] may not be as essential as previously thought.
All in all, although the knowledge around the structure of dirhodium(II) carboxylate complexes has been expanded dramatically in the last decade, the topic is still in its infancy.With the fast development in X-ray crystallography, computational chemistry and NMR spectroscopy, researchers may prosper in solving the mystery behind this topic.This, in turn, will aid to a great extent in the design of highly stereoselective chiral dirhodium(II) catalysts in the future.

Figure 1 .
Figure 1.Main arrangements of ligand's blocking groups around the dirhodium(II) core and their point groups (the sterically blocking groups around the rhodium active sites are depicted as ovals) [2,4,36,53,54].

Figure 2 .
Figure 2. (a) Space filling structure of mono(EtOAc) adduct of Rh2(S-PTTL)4 (top view); Schematic illustration of the rectangular binding pocket of Rh2(S-PTTL)4; (b) Space filling structure of mono(EtOAc) adduct of Rh2(S-PTTL)4 (side view).(Adapted from Adly, F.G.; Gardiner, M.G.; Ghanem, A. Chem.Eur.J. 2016, 22, 3447, Copyright 2017, with permission from John Wiley and Sons).Based on the new information, Fox et al. introduced a new model to account for the enantioselectivity exhibited by Rh2(S-PTTL)4 in enantioselective cyclopropanation of styrene with αalkyldiazoesters (Scheme 7)[78,87].In this model, Rh2(S-PTTL)4 was proposed to keep its α,α,α,αconformation during catalysis and the carbene is anticipated to align itself within the wide dimension of the catalyst's rectangular cavity leaving its si-face accessible by the substrate.Also according to Fox model, the tert-butyl groups are crucial to restrict the reactivity to the chiral cavity face of the catalyst and limit the racemic background reaction.

Scheme 7 .
Scheme 7. Facial selectivity of the Rh2(S-PTTL)4-stabilized carbene according to the model proposed by Fox et al.[78,87].Fox model was very well supported by the results obtained earlier for Rh2(S-PTAD)4 by being more enantioselective than Rh2(S-PTTL)4[37,[91][92][93]95,96].The results were very relevant as the much bulkier adamantyl groups would more efficiently block the carbene ligands from coordinating to the achiral rhodium face (shrouded by the adamantly groups) compared to the tert-butyl groups of Rh2(S-PTTL)4 leading to the observed more enantioselective outcome.Moreover, Fürstner et al. recently

Scheme 7 . 7 .
Scheme 7. Facial selectivity of the Rh2(S-PTTL)4-stabilized carbene according to the model proposed by Fox et al.[78,87].Fox model was very well supported by the results obtained earlier for Rh2(S-PTAD)4 by being more enantioselective than Rh2(S-PTTL)4[37,[91][92][93]95,96].The results were very relevant as the much bulkier adamantyl groups would more efficiently block the carbene ligands from coordinating to the achiral rhodium face (shrouded by the adamantly groups) compared to the tert-butyl groups of Rh2(S-PTTL)4 leading to the observed more enantioselective outcome.Moreover, Fürstner et al. recently formed by the four N-phthalimido groups of Rh2(S-PTTL)4, that adapt an α,α,α,αarrangement (C2-symmetric).The overall situation looked, therefore, in good agreement with the model proposed by Fox.But, low temperature NMR data for this intermediate returned a single set of signals for the four chiral carboxylate ligands[97].This observation was not in a good accord with a C2-symmetric binding site as proposed by the Fox model.