Development of Enantioselective Lithium-Isothiourea-Boronate–Catalyzed Matteson Homologations

Abstract

Our group’s discovery of lithium-isothiourea-boronate–catalyzed Matteson homologations is chronicled. Chiral thiourea dual–hydrogen bond donors were initially found to promote enantioselective dichloromethyl boronate rearrangements, albeit with poor reproducibility. Systematic investigations of the fate of the thiourea led to the discovery that lithium-isothiourea-boronate derivatives were being generated in situ as highly enantioselective catalytically active species. The optimal lithium-isothiourea-boronate catalyst displays significant generality in the rearrangement of primary alkyl migrating groups, affording synthetically valuable α-chloro boronic ester products with consistently high enantioselectivities. The catalyst is proposed to act as a structurally rigid chiral framework that precisely positions two lithium cations to enable a dual-lithium–mediated chloride abstraction.

1 Introduction

2 Reaction Development

3 Discovery of Isothiourea-Boronate Catalysts

4 Synthetic Application

5 Mechanistic and Computational Studies

6 Conclusions and Outlook

Biographical Sketches

Jake Essman received his bachelor’s degree in chemistry from Princeton University in 2018, where he conducted research in the lab of Abigail Doyle. He is currently pursuing his Ph.D. at Harvard University working with Eric Jacobsen on the development of enantioselective alkali metal catalysis.

Hayden Sharma studied chemistry at Northwestern University, where he received his bachelor’s and master’s degrees in 2016 while conducting research in the lab of Karl Scheidt. He went on to earn his Ph.D. at Harvard University in 2021 working with Eric Jacobsen on the development of various catalytic enantioselective rearrangement reactions. In 2022, Hayden began as a postdoctoral researcher at The Scripps Research Institute, where he is working with Benjamin Cravatt III on developing new chemical approaches to explore the druggable proteome.

Eric Jacobsen is the Sheldon Emery Professor of Chemistry at Harvard. He is a native New Yorker of Cuban descent, and he carried out his undergraduate studies at NYU working in the lab of Yorke Rhodes. He earned his Ph.D. at U.C. Berkeley with Robert Bergman in the arena of organotransition metal chemistry and was an NIH postdoctoral fellow at MIT with Barry Sharpless. He began his independent career at the University of Illinois in 1988 and moved to Harvard as a full professor in 1993.

Introduction

The Matteson homologation reaction, in which a dichloromethyl-substituted tetracoordinate boronate intermediate undergoes stereospecific 1,2-rearrangement to afford an α-chloro boronic ester (Scheme [1]), has found extensive use in both academic and industrial synthesis.[1] [2] [3] [4] Highly diastereoselective variants of the Matteson homologation have been achieved using chiral diol ligands such as pinanediol or 1,2-dicyclohexylethanediol on boron. The power of this method is tied to the fact that the α-chloro boronic ester products can be further elaborated through the addition of an organometallic nucleophile, inducing a second stereospecific boronate rearrangement to afford chiral secondary boronic esters.

Our group became intrigued by the possibility of effective enantioselective catalytic Matteson homologations by drawing on our longstanding efforts in promoting asymmetric reactions through anion abstraction by chiral dual–hydrogen bond donor (HBD) catalysts.[5] [6] [7] [8] [9] The canonical diastereoselective Matteson homologation of boronic esters bearing chiral diol ligands is promoted by stoichiometric ZnCl₂, which is proposed to assist in chloride abstraction as a Lewis acid.[10] Jadhav and Man reported a chiral Lewis acid–mediated rearrangement of a dichloromethyl-substituted n-butyl pinacol boronate, wherein good levels of enantioselectivity were obtained using a Yb-based Lewis acid with superstoichiometric quantities of a chiral bis(oxazoline) ligand.[11] [12] We considered that a chiral HBD catalyst might similarly promote the desymmetrization of prochiral dichloromethyl boronates via selective abstraction of one enantiotopic chloride leaving group. In 2021 we reported the successful development of a protocol for catalytic asymmetric Matteson homologation reactions.[13] However, the path to that discovery took us far away from HBD catalysis and into an entirely unexpected and highly promising new area of main-group metal chemistry.

Reaction Development

The hypothesis that neutral HBD catalysts might promote enantioselective homologation reactions was tested using members of the emerging privileged catalyst family of arylpyrrolidine-tert-leucine–derived HBDs in the rearrangement of model boronate 1a generated by addition of n-butyllithium (nBuLi) to dichloromethyl boronic acid pinacol ester (DCM-BPin) (Scheme [2]A, ‘Approach A’).[11] The preliminary results were highly encouraging, with 20 mol% of the thiourea derivative 3a promoting formation of the α-chloro boronic ester product 2a with up to 81% ee. The corresponding urea derivative 3b was found to be far less enantioselective, and the squaramide analogue 3c generated only racemic product. We have often observed significant differences in enantioinduction between the corresponding urea, thiourea, and squaramide derivatives of a given chiral framework during optimization campaigns of HBD-catalyzed reactions,[14] so initially we did not attach any great significance to that finding in the present system. Only later would we come to recognize how the unique characteristics of the thiourea molecule were crucial to its successful application in this reaction.

Variation of the thiourea arylpyrrolidine moiety revealed a relatively flat selectivity landscape (Scheme [2]B), with the 4-chlorophenyl–substituted thiourea derivative 3f affording the α-chloro boronic ester product with the highest enantiomeric excess (88% ee) of the HBDs that were evaluated.

As we gained confidence that we might achieve a reasonably effective protocol for the enantioselective catalytic reaction, we began to contemplate scope studies. Given the great number of commercially available boronic esters and the corresponding dearth of commercial organolithium reagents, we knew that the variety of boronate substrates that could be prepared conveniently would be greatly expanded if we accessed the boronates via the addition of dichloromethyllithium (DCM-Li) to boronic esters (Scheme [3], ‘Approach B’). However, reactions run using boronates prepared in this manner were found to be highly irreproducible in both yield and enantioselectivity, only occasionally affording levels approaching those obtained in the original catalyst optimization efforts using Approach A. Clearly, both protocols should generate the same boronate intermediate. The different enantioselectivity outcomes with the same HBD indicated that any of the following three scenarios might be at play: (1) an uncatalyzed racemic background reaction might be contributing to different extents due to different rates of boronate generation, given that the chiral HBD is added last; (2) the methods of substrate generation might result in slightly different reaction conditions (e.g., the presence of iPr₂NH as a byproduct in Approach B); or (3) different catalytically active species might be generated under the different conditions. In retrospect, our allegiance to our initial hypothesis of HBD catalysis made us initially reluctant to give the third option the consideration it ultimately merited.

At this point it should be noted that a fundamental premise underlying our initial strategy for catalysis of the Matteson homologation was that the boronate intermediates would not be basic enough to deprotonate the weakly acidic HBD, and that the HBD must also elude decomposition by any reagents employed to generate the boronate intermediates. For the latter reason, and as outlined in the original screening protocol in Scheme [2]A, we were careful to avoid exposure of the HBD to the strongly basic nBuLi reagent used to generate the boronate adduct in Approach A. Thus, we combined an excess of DCM-BPin with nBuLi, and only afterwards added the chiral HBD. Within this protocol, we found that the ee of rearrangements was strongly dependent on the interval between addition of nBuLi and addition of thiourea 3f, with higher enantioselectivities achieved when 3f was added immediately following or even prior to nBuLi addition (Table [1]). We first dismissed that unexpected observation as an artifact of contributions from an uncatalyzed racemic pathway occurring prior to addition of the chiral HBD. Only with considerable skepticism about the likely outcome of the reaction did we perform what proved to be a crucial control experiment: incubation of thiourea 3f with a large excess of nBuLi prior to addition of DCM-BPin (Scheme [4]). Under those conditions, which presumably result in quantitative deprotonation of the HBD, product 2a was obtained in good yield with 73% ee. We were thus forced to confront the fact that the HBD was at very least not the only active catalyst, and that it was undergoing decomposition to one or more highly enantioselective catalytically active species.

Table 1 Effect of delayed addition of thiourea 3f using boronate preparation Approach A.

time (min)	yield (%)a	ee (%)
15b	–	7
5	78	48
0.33	78	66
0	68	92

^a Yields refer to GC yields relative to an internal standard.

^b Reaction run using 0.5 equivalents of nBuLi and warmed from –78 °C to RT.

Discovery of Isothiourea-Boronate Catalysts

With the realization that decomposition of the HBD catalyst was a possible culprit underlying the disparate results obtained using boronate generation Approaches A and B, we directed our efforts towards characterizing the active species under the catalytic conditions. Attempted recovery of the chiral HBD catalyst from a 1.0 mmol scale reaction carried out under the conditions previously identified as optimal (Approach A, thiourea 3f added prior to nBuLi) resulted in isolation of a novel compound containing the arylpyrrolidine-tert-leucine motif but also exhibiting a broad ¹¹B NMR resonance around 12 ppm. High-resolution mass spectrometric analysis revealed a parent ion with m/z corresponding to [thiourea + DCM-BPin – chloride]⁺, suggesting that the thiourea had displaced one of the chlorides of DCM-BPin to form a covalent adduct.

The identity of this covalent adduct was established by X-ray crystallographic analysis as the novel heterocyclic isothiourea-boronate derivative 4a, formed by reaction of DCM-BPin with the thioureate produced by deprotonation of 3f (Figure [1]). As unexpected as that product was, in fact the first compound in this class that we succeeded in characterizing as a minor impurity was the even more exotic coupling product 4b, which derives from thioureate displacement of the remaining chloride of 4a. The spirocyclic structure of 4 finds precedent in reports from Matteson[15] and Biedrzycki,[16] who demonstrated that thiourea itself undergoes reaction with α-halo boronic esters to form analogous heterocycles. However, the properties of this unusual class of compounds had never been investigated to our knowledge, and certainly never in the context of asymmetric catalysis.

A systematic investigation of the fate of thiourea 3f under boronate preparation Approaches A and B revealed orthogonal catalyst speciation, consistent with the strongly differing outcomes of reactions carried out under the different conditions (Scheme [5]). Under Approach A, none of the thiourea 3f remained; the predominant HBD-derived species was the mono-adduct 4a, present in a roughly 3:1:1 ratio with its α-boryl epimer 4c and the dimeric adduct 4b. In contrast, under the Approach B conditions, roughly half of the thiourea 3f was recovered along with butyl-substituted derivative 4d, present in a roughly 4:1 ratio with its α-boryl epimer 4e.

The key concern after the discovery that isothiourea-boronates were being generated under the catalytic conditions was whether these compounds were catalytically relevant intermediates or simply inactive catalyst decomposition products. To address this question, we carried out a reaction under the Approach A conditions and then used the crude reaction mixture—containing product 2a, isothiourea-boronates 4a–c, and no detectable thiourea—as the catalyst for a second reaction with a different substrate (Scheme [6]). The rearrangement of boronate 1b prepared by Approach B was catalyzed by the mixture of isothiourea-boronates to afford product 2b reproducibly in 92% ee.

With the new insight that at least one of the isothiourea-boronates—and not the thiourea 3f—was the active catalyst responsible for highly enantioselective boronate rearrangements, we re-examined reactions of 1a prepared by Approach A in the presence of urea derivative 3b and squaramide derivative 3c (Scheme [2]A). In the reaction with 3b, an array of urea decomposition products was detected with similar spectral characteristics to the isothiourea-boronates, whereas no analogous adducts were observed using squaramide derivative 3c. We could thus conclude that the enantioselective rearrangement reaction is tied directly to the generation of the iso(thio)urea-boronate adducts.

These insights motivated us to perform structure–enantioselectivity relationship studies with different isothiourea-boronate analogues, where we focused on variations of the α-boryl substituent, the electronic properties of the aniline, the amino acid linker, and the arylpyrrolidine (Figure [2]). More than twenty structural analogs of 4a were successfully prepared, but all were found to catalyze the rearrangement of 1a with lower enantioselectivity; ultimately 4a, which was synthesized independently in pure form and on gram scale via reaction of excess racemic bromochloromethyl boronic acid pinacol ester with deprotonated thiourea 3f,[13] was found to be the most enantioselective catalyst. The N-methylated derivative 4f was unique among the catalysts examined in that it afforded only racemic product. This result led to the hypothesis that the N–H proton in 4 was instrumental in the formation of the active chiral catalyst. In particular, given the observation that thiourea 3f could be incubated with excess nBuLi and maintain catalytic activity (Scheme [4]), we hypothesized that the isothiourea N–H was undergoing lithiation under the reaction conditions and that this deprotonation step might be crucial for catalytic activity.

Analysis of isothiourea-boronate 4a by ReactIR spectroscopy in the presence and absence of a strong base provided definitive evidence for the lithiation of the isothiourea. The amide C–O and isothiourea N–C–N stretches of 4a were redshifted by roughly 40 cm^–1 upon addition of 1 equivalent of LiHMDS, consistent with the formation of a five-membered amide-lithium-isothiourea chelate. This structural hypothesis was supported by DFT calculations of 4a and Li-4a (Figure [3]A), and by characterization of a lithiated catalyst analogue Li-4l by X-ray crystallography (Figure [3]B).

Re-examination of the isothiourea-boronate derivatives 4a,c–e now as their lithiated derivatives under the previously irreproducible Approach B conditions revealed that chloro-substituted Li-4a was indeed optimal, affording the desired product in reproducibly high yield and 95% ee (Scheme [7]). Perhaps most significant, identical results were obtained with Li-4a in rearrangements of boronates generated by either Approach A or B. The α-boryl epimer Li-4c also promoted the model reaction with high enantioselectivity. In contrast, the butyl-substituted epimeric analogues Li-4d and Li-4e performed very differently, with Li-4d providing an ee value similar to Li-4c and Li-4e affording much lower enantioselectivity. The basis for these differences will be discussed later in this account.

Taken together, the observations described above allow us to account for the variable yet sometimes highly enantioselective reactions obtained under the original Approach B (thiourea 3f/nBu-BPin/DCM-Li) conditions. In those reactions, nBu-BPin and DCM were dissolved in Et₂O at –78 °C; 1.0 equiv. of LDA was added; and the mixture was stirred for 5 to 15 minutes with the intention of allowing complete conversion into boronate 1a, followed by addition of thiourea 3f. Since we now know that no active chiral catalyst is present initially as the mixture is warmed to room temperature, it is likely that an uncatalyzed reaction of boronate 1a takes place to produce a small amount of racemic 2a. Compound 2a can undergo reaction with thiourea 3f that has been deprotonated by any unreacted base (either LDA or DCM-Li) to afford active catalyst Li-4d together with smaller amounts of its epimer Li-4e.[17] Given that Li-4d is generated selectively (Scheme [5]), the reaction involves a kinetic resolution of 2a resulting in a slight upgrade of the ee of that compound. Catalyst Li-4d promotes the enantioselective rearrangement of boronate 1a to afford 2a in 88% ee, which may be further upgraded by kinetic resolution with deprotonated thiourea 3f to produce additional catalyst Li-4d in what can be seen overall as a two-step autocatalytic process. The observed product yield and ee are dependent on several factors but primarily on the amount of catalyst Li-4d generated, which itself is dependent on the quantity of unreacted LDA or DCM-Li available to deprotonate thiourea 3f upon its delayed addition.[18] The variability in the delay time before addition of 3f can thus explain the irreproducibility in the observed product yield and ee.

Synthetic Application

Boronate rearrangements catalyzed by 20 mol% of Li-4a proceeded with excellent levels of enantioselectivity using a wide variety of primary alkyl and small secondary alkyl migrating groups (Scheme [8]). While boronates bearing larger secondary alkyl groups (2n), aryl groups (2o,p), and alkenyl groups (2q,r) underwent migration with lower levels of enantioselectivity, the scope of highly enantioselective reactions with primary migrating groups was excellent.[19] Treatment of the enantioenriched α-chloro boronic ester products with Grignard reagents, organolithium reagents, or various heteroatomic nucleophiles led to fully stereospecific displacement of the chloride, affording highly enantioenriched secondary boronic esters with a wide variety of unusual substitution patterns.[13] The migrating group limitations in the enantioselective rearrangement are absent from the subsequent stereospecific reaction, permitting access to a wide assortment of secondary boronic ester products. Further elaborations of the pinacol boronic ester moiety using known stereospecific methods generated products bearing a wide variety of trisubstituted stereocenters.[13]

Mechanistic and Computational Studies

By analogy to the role proposed for ZnCl₂ in the diastereoselective Matteson homologation, we initially hypothesized that Li-4a might be acting as a simple chiral lithium-centered Lewis acid that promotes boronate rearrangement via enantioselective chloride abstraction. However, a Gutmann–Beckett[20] [21] [22] study of the Lewis acidities of representative lithiated compounds (Table [2]) revealed that Li-4a is likely no more Lewis acidic than the lithium boronate substrate itself, as assessed using the unreactive lithium boronate substrate mimic Li-5.[13] Thus, Lewis acidity alone does not account for the efficient enantioselective catalysis of boronate rearrangement reactions by Li-4a. Additional transition state stabilizing effects must be at play in the enantiodetermining step.

Table 2 Gutmann–Beckett analysis of Li-4a in comparison to other Li-based Lewis acids.

Lewis acid	acceptor number (Et2O)
LiHMDS	33
LiNTf2	36
Li-4a	37
Li-5	38
LiBArF	40

Analysis of a calculated electrostatic potential map of Li-4a (Figure [4]A) reveals the site of highest electron density to be the boronate oxygens, consistent with significant delocalization of negative charge from the deprotonated isothiourea nitrogen to the boronate moiety. This delocalization may serve both to enhance the Lewis acidity of the catalyst Li cation and to provide a distant Lewis basic anchoring site for the Li cation of the boronate substrate. Kinetic characterization of Li-4a–catalyzed rearrangement reactions was hampered by several technical challenges, but for the purposes of developing a reasonable computational model for the mode of catalysis we assumed a 1:1 substrate:catalyst stoichiometry in the enantiodetermining transition states (Figure [4]B). In the lowest-energy computed transition state structure leading to the major enantiomer of product, the substrate Li cation is chelated by the anterior boronate oxygen and α-boryl chloride of the catalyst. The catalyst and substrate Li cations are thus positioned to induce cooperative abstraction of the chloride leaving group. The location of the chloride leaving group in the corresponding minor transition state geometry permits only the substrate Li to promote chloride abstraction. The moderate increase in ee with chloro-substituted catalyst Li-4a over its epimer Li-4c, along with the severe decrease in ee with butyl-substituted catalyst Li-4e over its epimer Li-4d, are both consistent with the proposed substrate Li binding mode (Figure [4]C). The generality of this reaction with primary alkyl migrating groups may be ascribed to the solvent-facing position of the migrating group, with the catalyst-facing remainder of the dichloromethyl boronate substrate being identical across the entire reaction scope.

Conclusions and Outlook

The unexpected discovery of an entirely new class of chiral catalysts offers tantalizing possibilities for future development and application. Clearly, development of catalysts with expanded scope as well as improved catalytic efficiency and synthetic accessibility will be crucial to the widespread adoption of this methodology. Specific efforts to exploit and expand the synthetic utility of the lithium-isothiourea-boronate–catalyzed Matteson homologation include application to diastereoselective catalyst-controlled iterative homologations.[23] Success in that direction will rely on the development of catalysts that mediate highly stereoselective rearrangements of boronates bearing bulky secondary alkyl migrating groups. More broadly, we are hopeful that the discovery of these sterically and electronically tunable chiral frameworks might facilitate the development of new alkali-metal–based catalysts and their application in new asymmetric catalytic reactions.

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgment

We thank Shao-Liang Zheng (Harvard University) for determination of X-ray crystal structures.

• References

• 1 Matteson DS, Mah RW. H. J. Am. Chem. Soc. 1963; 85: 2599
  CrossrefSearch in Google Scholar
• 2 Matteson DS. Tetrahedron 1998; 54: 10555
  CrossrefSearch in Google Scholar
• 3 Matteson DS. J. Org. Chem. 2013; 78: 10009
  CrossrefPubMedSearch in Google Scholar
• 4 Matteson DS, Collins BS. L, Aggarwal VK, Ciganek E. Org. React. 2021; 105: 427
  Search in Google Scholar
• 5 Reisman SE, Doyle AG, Jacobsen EN. J. Am. Chem. Soc. 2008; 130: 7198
  CrossrefPubMedSearch in Google Scholar
• 6 Park Y, Harper KC, Kuhl N, Kwan EE, Liu RY, Jacobsen EN. Science 2017; 355: 162
  CrossrefPubMedSearch in Google Scholar
• 7 Bendelsmith AJ, Kim SC, Wasa M, Roche SP, Jacobsen EN. J. Am. Chem. Soc. 2019; 141: 11414
  CrossrefPubMedSearch in Google Scholar
• 8 Kutateladze DA, Strassfeld DA, Jacobsen EN. J. Am. Chem. Soc. 2020; 142: 6951
  CrossrefPubMedSearch in Google Scholar
• 9 Forbes KC, Jacobsen EN. Science 2022; 376: 1230
  CrossrefPubMedSearch in Google Scholar
• 10 Fasano V, Aggarwal VK. Tetrahedron 2021; 78: 131810
  CrossrefSearch in Google Scholar
• 11 Jadhav PK, Man H.-W. J. Am. Chem. Soc. 1997; 119: 846
  CrossrefSearch in Google Scholar
• 12 Smith K, Saleh BA, Alshammari MB, El-Hiti GA, Elliott MC. Org. Biomol. Chem. 2021; 19: 4279
  CrossrefPubMedSearch in Google Scholar
• 13 Sharma HA, Essman JZ, Jacobsen EN. Science 2021; 374: 752
  CrossrefPubMedSearch in Google Scholar
• 14 Wendlandt AE, Vangal P, Jacobsen EN. Nature 2018; 556: 447
  CrossrefPubMedSearch in Google Scholar
• 15 Matteson DS, Schaumberg GD. J. Org. Chem. 1966; 31: 726
  CrossrefSearch in Google Scholar
• 16 Biedrzycki M, Scouten WH, Biedrzycka Z. J. Organomet. Chem. 1992; 431: 255
  CrossrefSearch in Google Scholar
• 17 Regardless of its intrinsic enantioselectivity, minor catalyst Li-4e is present in small quantities and is speculated to be poorly efficient, thus exerting little effect on the reaction outcome.
• 18 Using 1 equivalent of LDA, complete conversion of 20 mol% of thiourea 3f into 20 mol% of Li-4d would consume 0.4 equivalents of LDA and 0.2 equivalents of product 2a, leading to a maximum theoretical yield of 2a of 40%.
• 19 While addition of excess DCM assisted in solubilizing certain boronate substrates, addition of THF led invariably to racemic product, possibly due to catalyst inhibition.
• 20 Mayer U, Gutmann V, Gerger W. Monatsh. Chem. 1975; 106: 1235
  CrossrefSearch in Google Scholar
• 21 Beckett MA, Strickland GC, Holland JR, Sukumar Varma K. Polymer 1996; 37: 4629
  CrossrefSearch in Google Scholar
• 22 Beckett MA, Brassington DS, Coles SJ, Hursthouse MB. Inorg. Chem. Commun. 2000; 3: 530
  CrossrefSearch in Google Scholar
• 23 Stymiest JL, Dutheuil G, Mahmood A, Aggarwal VK. Angew. Chem. Int. Ed. 2007; 46: 7491
  CrossrefPubMedSearch in Google Scholar

Corresponding Author

Publication History

Received: 29 April 2023

Accepted after revision: 24 May 2023

Accepted Manuscript online:
24 May 2023

Article published online:
02 August 2023

© 2023. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany

• References

• 1 Matteson DS, Mah RW. H. J. Am. Chem. Soc. 1963; 85: 2599
  CrossrefSearch in Google Scholar
• 2 Matteson DS. Tetrahedron 1998; 54: 10555
  CrossrefSearch in Google Scholar
• 3 Matteson DS. J. Org. Chem. 2013; 78: 10009
  CrossrefPubMedSearch in Google Scholar
• 4 Matteson DS, Collins BS. L, Aggarwal VK, Ciganek E. Org. React. 2021; 105: 427
  Search in Google Scholar
• 5 Reisman SE, Doyle AG, Jacobsen EN. J. Am. Chem. Soc. 2008; 130: 7198
  CrossrefPubMedSearch in Google Scholar
• 6 Park Y, Harper KC, Kuhl N, Kwan EE, Liu RY, Jacobsen EN. Science 2017; 355: 162
  CrossrefPubMedSearch in Google Scholar
• 7 Bendelsmith AJ, Kim SC, Wasa M, Roche SP, Jacobsen EN. J. Am. Chem. Soc. 2019; 141: 11414
  CrossrefPubMedSearch in Google Scholar
• 8 Kutateladze DA, Strassfeld DA, Jacobsen EN. J. Am. Chem. Soc. 2020; 142: 6951
  CrossrefPubMedSearch in Google Scholar
• 9 Forbes KC, Jacobsen EN. Science 2022; 376: 1230
  CrossrefPubMedSearch in Google Scholar
• 10 Fasano V, Aggarwal VK. Tetrahedron 2021; 78: 131810
  CrossrefSearch in Google Scholar
• 11 Jadhav PK, Man H.-W. J. Am. Chem. Soc. 1997; 119: 846
  CrossrefSearch in Google Scholar
• 12 Smith K, Saleh BA, Alshammari MB, El-Hiti GA, Elliott MC. Org. Biomol. Chem. 2021; 19: 4279
  CrossrefPubMedSearch in Google Scholar
• 13 Sharma HA, Essman JZ, Jacobsen EN. Science 2021; 374: 752
  CrossrefPubMedSearch in Google Scholar
• 14 Wendlandt AE, Vangal P, Jacobsen EN. Nature 2018; 556: 447
  CrossrefPubMedSearch in Google Scholar
• 15 Matteson DS, Schaumberg GD. J. Org. Chem. 1966; 31: 726
  CrossrefSearch in Google Scholar
• 16 Biedrzycki M, Scouten WH, Biedrzycka Z. J. Organomet. Chem. 1992; 431: 255
  CrossrefSearch in Google Scholar
• 17 Regardless of its intrinsic enantioselectivity, minor catalyst Li-4e is present in small quantities and is speculated to be poorly efficient, thus exerting little effect on the reaction outcome.
• 18 Using 1 equivalent of LDA, complete conversion of 20 mol% of thiourea 3f into 20 mol% of Li-4d would consume 0.4 equivalents of LDA and 0.2 equivalents of product 2a, leading to a maximum theoretical yield of 2a of 40%.
• 19 While addition of excess DCM assisted in solubilizing certain boronate substrates, addition of THF led invariably to racemic product, possibly due to catalyst inhibition.
• 20 Mayer U, Gutmann V, Gerger W. Monatsh. Chem. 1975; 106: 1235
  CrossrefSearch in Google Scholar
• 21 Beckett MA, Strickland GC, Holland JR, Sukumar Varma K. Polymer 1996; 37: 4629
  CrossrefSearch in Google Scholar
• 22 Beckett MA, Brassington DS, Coles SJ, Hursthouse MB. Inorg. Chem. Commun. 2000; 3: 530
  CrossrefSearch in Google Scholar
• 23 Stymiest JL, Dutheuil G, Mahmood A, Aggarwal VK. Angew. Chem. Int. Ed. 2007; 46: 7491
  CrossrefPubMedSearch in Google Scholar