Enantiomeric Perturbation of Equilibria. Differential Solvation of a Chiral Lithium Amide by the Enantiomers of 2-Methyltetrahydrofuran Measured by NMR Spectroscopy Go ¨ran Hilmersson, Per Ahlberg,* and O ¨ jvind Davidsson* Department of Organic Chemistry, Go ¨ teborg UniVersity S-41296 Go ¨ teborg, Sweden ReceiVed December 6, 1995 We wish to report the first direct observation of diastereo- selective solvation of a chiral organolithium compound by the enantiomers of a chiral ether, i.e. solvation by (R)- and (S)- 2-methyltetrahydrofuran (2-MTHF) of the dimer of lithium (2- methoxy-(R)-1-phenylethyl)((S)-1-phenylethyl)amine ((LiA) 2 ) (Scheme 1). A novel method for determination of accurate equilibrium constants (K) and their temperature dependence (ΔH and ΔS) using NMR chemical shifts is also presented. It has been used to measure the differential solvation of (LiA) 2 by the R- and S-isomers of 2-MTHF. The 13 C NMR spectrum of (LiA) 2 in toluene with 2-MTHF added ([(LiA) 2 ] tot ) 0.25 M and [2-MTHF] tot ) 0.1 M) shows separate signals of equal intensity for (R)-2-MTHF and (S)-2- MTHF in the diastereomeric complexes (LiA) 2 ‚(R)-2-MTHF and (LiA) 2 ‚(S)-2MTHF at all temperatures (-80 to +25 °C). All 2-MTHF molecules are complexed, and (LiA) 2 has been shown by integration to form 1:1 complexes with 2-MTHF. With excess of 2-MTHF at -80 °C separate signals are observed for the complexes as well as for uncomplexed 2-MTHF (Figure 1). Some signals (C 2 and C 5 ) from noncoordinated 2-MTHF and 2-MTHF in the diastereomeric complexes are shown. The signals from 5-carbons (C 5 ) in the (R)- and (S)-isomers, respectively, of 2-MTHF appear with different intensities, indicating that the solvation of (LiA) 2 shows stereoselection and that diastereomeric complexes are formed, i.e., K in Scheme 1 differs from unity. The assignment of each set of peaks to the stereoisomers of 2-MTHF has not yet been made. The signals from the 2-carbons (C 2 ) appear at about the same chemical shift (δ 76.24 and 76.31 at -80 °C). In Figure 1 these signals are not resolved. Obviously, the exchange between coordinated and noncoordinated 2-MTHF is slow at -80 °C, since separate sharp signals are observed for the coordinated and noncoordinated (R)- and (S)-2-MTHF molecules. Since no 13 C signals from uncomplexed (LiA) 2 are observed, it is concluded that all (LiA) 2 molecules are complexed by 2-MTHF. When the temperature is raised, the rate of exchange between coordinated and noncoordinated ether molecules increases and coalescence is observed. At even higher temperature (-40 °C) pairs of sharp signals are observed for each of the two types of carbon (C 2 and C 5 ). At 25 °C it is obvious that the intensities of the signals in each pair is equal. Thus, it is concluded that one of the signals in a pair originates from (R)-2-MTHF molecules in (LiA) 2 ‚(R)-2-MTHF and non- coordinated (R)-2-MTHF molecules, which are rapidly exchang- ing. Similarly, the other one originates from exchanging (S)- 2-MTHF molecules. The chemical shift difference between the signals in the pair originating from C 2 is mainly due to the fact that the (R)- and (S)-2-MTHF molecules have different complexation constants since the shift difference between C 2 of 2-MTHF in (LiA) 2 ‚- (R)-2-MTHF and (LiA) 2 ‚(S)-2-MTHF is only ca. 0.07 ppm. To our knowledge this is the first example of chiral “discrimination” by a chiral organolithium compound between the enantiomers of an ethereal solvent. Previously slow ethereal ligand exchange on the NMR time scale between ethers coordinated to achiral lithium amides and noncoordinated ethers has been reported. 1,2 Slow exchange between (LiA) 2 coordi- nated to achiral ethers and noncoordinated achiral ethers has also been recently reported. 3,4 The 13 C NMR and the 6 Li NMR spectra of the solutions used in the above experiments also showed the exclusive presence of the lithium amide dimer (LiA) 2 . However, addition of large amounts of 2-MTHF to the toluene solution of (LiA) 2 at -80 °C results in the appearance of separate signals of the solvated monomer (LiA) in addition to the signals from the dimer. Increase of temperature strongly disfavors the monomers. Figure 1 shows that the shift difference between the signals from the diastereomeric solvates shows temperature dependence. The results presented above indicate that accurate equilibrium constants (K) and their temperature dependence (ΔH and ΔS) may be determined as shown below for the equilibrium in Scheme 1. Thus the differential solvation of (LiA) 2 by the enantiomers of 2-MTHF could be accurately measured. The equilibrium constant is defined as in eq 1. K ) [(LiA) 2 ‚(R)-2-MTHF][(S)-2-MTHF] [(LiA) 2 ‚(S)-2-MTHF][(R)-2-MTHF] (1) At fast ligand exchange the average chemical shift (δ R ) for (R)-2-MTHF is given by eq 2. (1) (a) Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1994, 116, 6009. (b) Lucht, B. L.; Collum, D. B. J. Am. Chem. Soc. 1995, 117, 9863. (2) Slow ligand exchange rates have also been reported with HMPA and diamines, see: (a) Reich, H. J.; Green, D. P. J. Am. Chem. Soc. 1989, 111, 8729. (b) Barr, D.; Doyle, M. J.; Mulvey, R. E.; Raithby, P. R.; Reed, D.; Snaith, R.; Wright, D. S. J. Chem. Soc., Chem. Commun. 1989, 318. (c) Fraenkel, G.; Chow, A.; Winchester, W. R. J. Am. Chem. Soc. 1990, 112, 1282. (d) Boche, G.; Fraenkel, G.; Cabral, J.; Harms, K.; van Eikema Hommes, N. J. R.; Lohrenz, J.; Marsch, M.; Schleyer, P.v.R. J. Am. Chem. Soc. 1992, 114, 1562. (e) Hoffman, R. W.; Klute, W.; Dress, R. K.; Wenzel, A. J. Chem. Soc., Perkin Trans. 2 1995, 1721. (3) Hilmersson, G.; Davidsson, O ¨ . J. Organomet. Chem. 1995, 489, 175.; J. Org. Chem. 1995, 60, 7660. (4) The structure of THF monosolvated (LiA) 2 has been determined by X-ray crystallography. Hilmersson, G.; Davidsson, O ¨ ; Håkansson, M. Manuscript in preparation. Figure 1. Partial 13 C NMR spectra at 125 MHz of 0.25 M [ 6 Li](LiA)2- 0.5 M 2-MTHF in toluene-d 8 at different temperatures. Scheme 1 3539 J. Am. Chem. Soc. 1996, 118, 3539-3540 0002-7863/96/1518-3539$12.00/0 © 1996 American Chemical Society