Looking back: Chiral organometallic reagents from Knochel

 Citation:

Skotnitzki, J; Kremsmair, A.; Keefer, D.; Gong, Y.; de Vivie-Riedle, R.; Knochel, P. Stereoselective Csp3-Csp2 Cross-Couplings of Chiral Secondary Alkylzinc Reagents with Alkenyl and Aryl Halides. Angew. Chem. Int. Ed. 2019 59, 1, 320-324. 

Summary Figure:

Background: 

In general, for an asymmetric Csp2-Csp3 cross-coupling reaction, there are 2 ways to set the stereochemistry. 

The first method is to ablate the existing stereocenter on the sp3 coupling partner, usually by forming an alkyl radical (although not always), and then use a chiral ligand to form a chiral alkyl-metal species. 

The second method is to use existing stereochemistry of the sp3 coupling partner and activate it with stereospecific elementary steps, usually oxidative addition. Famously, the Tsuji Trost reaction proceeds through an SN2-like oxidative addition, which inverts the stereochemistry at the electrophile:

(note how the X was on the dash, and the Pd is now on the wedge)

Usually, we only consider oxidative addition for this process, because we have known mechanisms for stereospecific oxidative addition that are fairly reliable (given the right circumstances). This means that usually only electrophiles are engaged in this fashion, which makes sense- usually you instill chirality as an alkyl alcohol, then derivatize. 

For example, you can access enantioenriched alkyl alcohols via chiral reductions of ketones (CBS reductions or the carrot reduction), asymmetric 1,2 additions from aldehydes, or kinetic resolutions from the racemic alcohol (lipase). Once that's made, you can do a number of protection/functionalization steps like esterification or etherification that proceed with retention of configuration, or Appel or Mitsunobu reactions to install many different functional groups with inversion of configuration. 

In contrast, making chiral organometallic reagents is usually very difficult* because (a) often the transmetalation epimerizes the chiral center (see: https://pubs.acs.org/doi/10.1021/ja025638k) and (b) even if you get the transmetalation to be stereospecific, the reagent may be configurationally labile and epimerize anyways. For example, see this paper from the Dieter lab, which generates chiral alkyl cuprate reagents using sparteine as a ligand, and still sees them degrade very quickly: https://pubs.acs.org/doi/10.1021/jo035845i.

*With the exception of boron reagents, which do not epimerize as readily as most organometallic reagents. There are a number of good ways to make chiral boron reagents for Suzuki reactions, although then the difficulty is in getting the Suzuki reaction to work with an sp3 boron reagent. For example: https://pubs.rsc.org/en/content/articlelanding/2022/cc/d1cc06186k

The Knochel lab found a way past this issue, through the hazardous sequence of adding chiral alkyl iodides dropwise into the solution of tert-butyllithium (!!) at -100 C. As the Knochel lab reports, "The configurational stability of these secondary alkyllithiums is rather moderate (ca. 1 h at -100 C in a hexane ether mixture)" I would consider only surviving 1 hour at -100 C as very unstable, but maybe I'm just slow. If you transmetalate with copper the stability marginally improves to "several hours at -50 C in THF), which the authors admit is still not good enough and restricts the reactivity and utility of these reagents. Eventually, they find that alkylzinc reagents are able to survive at room temperature for several hours without epimerizing. 

How it works:

Studying transmetalation is quite difficult, as most of the intermediates can't be studied by the same techniques used for studying other elementary steps like oxidative addition. The authors did do some DFT calculations to explain why the alkylzincs don't epimerize, which mostly boil down to (a) the zinc-carbon bond is too strong to homolytically break (~35 kCal/mol) and (b) the barrier to form a planar transition state to invert the stereochemistry is far too high (~95 kCal/mol), which is fairly expected. Notably, coordination of solvent molecules (such as ether) to the zinc does not significantly affect either pathway, which is somewhat surprising. Often these organometallic reagent are strongly solvent dependent, but this DFT calculation explains why these organozinc reagents are competent in both ether solvents and toluene. The authors also showed that epimerization is also not feasible in the transmetalation between zinc and palladium (which is commonly known to proceed with retention) and that the palladium-carbon bond is unlikely to epimerize via homolysis (also expected, otherwise any stereospecific palladium coupling would have this problem). 

Initial Questions and Key Findings:

1. Can you generate chiral organometallic reagents?

A: Yes, lithium halogen exchange with tBuLi and chiral alkyl iodides proceeds stereospecifically with retention of configuration of the iodide, forming a short-lived chiral alkyllithium. 

2: Can you generate chiral organometallic reagents that survive more than 5 minutes on a sunny day?

A: Yes, if you transmetalate these chiral alkyllithium reagents with zinc salts you can generate the chiral alkylzinc reagent, which is stable at room temperature for several hours.

3. Can you use these chiral alkylzinc reagents to make chiral compounds?

A: Yes, the authors proceed to do a number of Negishi-type cross-coupling reactions with the alkylzinc reagents to make enantioenriched products. 

Key takeaways:

The operational difficulty with these procedure probably precludes it from being useful on large scale, or in industry, or by people with shaky hands. However, it represents a huge chemical space that would lead to what I think is one of the holy grails of cross-coupling chemistry: an sp3-sp3 cross-coupling reaction that sets stereocenters on both partners. To my knowledge, this reaction does not currently exist, and I don't think we're close to this being a major method. However, we have a lot of methods to set stereocenters with oxidative addition. Adding a new method to the sparse other half makes this goal a lot more feasible.