Mendeleev Communications Electronic Version, Issue 4, 1999 (pp. 129–170) Enantiomeric NMR analysis of organic acids with the Corey chiral controller Margarita A. Lapitskaya, Georgy V. Zatonsky and Kasimir K. Pivnitsky* N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, 117913 Moscow, Russian Federation. Fax: +7 095 135 5328; e-mail: li@ioc.ac.ru (R,R)-1,2-Di-(2,4,6-trimethylbenzylamino)-1,2-diphenylethane (the Corey chiral controller) was found to be an effective chiral NMR shift reagent for the determination of the enantiomer ratio in chiral carboxylic and sulfonic acids and cyclic b-diketones. Determination of an enantiomer ratio (enantiomeric analysis) in the mixtures of enantiomers is a difficult but unavoidable procedure in any chiral total synthesis so widespread in the present days.There are several more or less general solutions of the problem. The most general solutions are NMR analysis with chiral lanthanide shift reagents and chromatography on chiral columns. The NMR analysis of diastereomeric mixtures of esters, amides and amine salts with chiral acids is by far the most popular substrate-specific method for enantiomeric analysis of alcohols and amines.1 The best method for the enantiomeric analysis of carboxylic acids is NMR analysis of diastereomeric salt mixtures formed in situ with (R,R)-1,2-diphenylethane-1,2-diamine (DPDAE).2 However, this diamine was found to be inapplicable to the analysis of ketoacid 1 because of rapid formation of Schiff bases in CDCl3 solution.As ketoacid 1 was important for us as a key intermediate in the total synthesis of eicosanoids3 so we were looking for a substitute of DPDAE devoid of its reactivity.Enantiomeric (R,R)- and (S,S)-1,2-di-(2,4,6-trimethylbenzylamino)- 1,2-diphenylethanes (DADPE) are excellent chiral controllers in an enantioselective olefin dihydroxylation reaction.4 The structure of DADPE has been tailored by E.J. Corey5 specially for this purpose. A complex of DADPE with osmium tetroxide binds olefins in a stereospecific manner due to the existence of an asymmetric groove in the predominant conformation of DADPE. It is believed that other types of complex formation with DADPE can be highly stereo- or enantiospecific as well. On the other hand, being hindered bis-secondary diamine, DADPE is much less reactive in comparison with DPDAE.We report here the salt formation of chiral acidic compounds with DADPE and its application to enantiomeric NMR analysis. The acidic compounds studied are presented in Scheme 1. These are chiral carboxylic acids 1–4, 9 and 10, sulfonic acid 8, and cyclic b-diketones 5–7. Acids 1–4 and 8 were used each as a pair of enantiomers or as a racemic mixture and a single enantiomer.Achiral acid 11 was used for comparison purposes. 1H NMR spectra were measured for mixtures of (R,R)-DADPE with the acids taken as racemic mixtures or, if available, as scalemic mixtures with known enantiomeric composition (approximately 4:6). The enantiomeric differences of signal shifts were taken from these spectra (Table 1).† The signal shifts induced by salt formation were calculated by comparison of the spectra of salts and corresponding individual compounds.The interaction of the diamine (R,R)-DADPE with an acid in a solution gives rise to a set of equilibria presented in Scheme 2, where NN and AH are a diamine and an acid, respectively. This interaction can lead to the formation of basic (1:1, acid:diamine) and neutral (2:1) salts depending on the composition of the mixture, existing as ion pairs (or triades), free ions or mixtures of them.In the case of an enantimer mixture, each enantiomer forms its own basic and (homo-)neutral salts, and the formation of a hetero-neutral salt becomes also possible. As the equilibria are rapid in the NMR time scale, the observed chemical shifts, for erxample, of R-acid signals are a weighted average of the chemical shifts in all species containing this enantiomer.This is also true for the S-enantiomer of the acid. Chemical shifts of acid enantiomer residues in the corresponding basic, homo- and hetero-neutral salts, and cations thereof, all diastereomeric by pairs, should be different from each other; therefore, enantiomeric differences for acid signals in the spectrum are possible.‡ The other reason for enantiomeric differences can be the inequality of association constants for the formation or dissociation of diastereomeric salts. This reason is believed to be at least as significant for an enantiomer discrimination in salts of monobasic amines and DPDAE due to anisochronicity.1,2 For DADPE we assume that the association constants for salt formation are sufficient to provide the nearly quantitative salt † Spectra of salts were recorded on a Bruker DRX500 spectrometer (500 MHz) at 30 °C in CDCl3 for the solutions 0.05 M in (R,R)-DADPE and 0.1 M in acidic compounds.Compounds with low solubility in CDCl3 (5–7, 9, 10) are much more soluble in the presence of DADPE and were used as saturated solutions with the acid:diamine ratios shown in Table 1.‡ The situation is quite different for the signals of the diamine (R,R)- DADPE. Although the positions of the signals in the mixtures with acids depend on the acid structure and an acid:diamine ratio, the salt formation with the mixtures of acid enantiomers never induced any enantiomeric splitting of enantiomerically pure diamine signals due to the rapid equilibria of a single diamine residue between all diamine-containing species.Scheme 1 For chiral compounds, only enantiomers with the R-configuration of a chiral centre proximal to an acidic functional group are presented. HN Ph NH Ph Me Me Me Me Me Me (R,R)-DADPE Me O OH COOH OH (R)-1 Ph COOH H OMe (R)-2 Ph COOH MeO CF3 (R)-3 O COOH O 1 2 3 4 5 6 7 8 (R)-4 O (R)-5 X = S (R)-6 X = SO2 O X Ph O (R)-7 O Br Br SO3H Me Me (R)-8 S N Ph COOH (R)-9 Me Me COOH NHCHO (R)-10 COOH 11 OMendeleev Communications Electronic Version, Issue 4, 1999 (pp. 129–170) formation in deuterochloroform solution, and dissociation of the formed salts into ions is insignificant. This assumption is based on the following observations.The induced chemical shifts (Dd = dsalt – dfree acid) of acid (S)-1 in mixtures with the diamine (R,R)-DADPE vary insignificantly with changing acid– base ratios from 0.6:1 to 2:1.§ An analogous result was obtained previously6 with Mosher acid 3. In addition, the substitution of deuterochloroform as a solvent for deuterobenzene does not increase significantly enantiomeric differences in the spectrum of acid 2 (cf.ref. 7). An acid:diamine ratio of 2:1 is most convenient from the practical point of view, and it was used throughout if possible.† At this ratio, acid (S)-1 with (R,R)-DADPE in CDCl3 produces a salt solution stable during several days (NMR data). The NMR spectrum of the salt exhibits significant induced Dd values for both acid and diamine signals.§ These Dd values reach –0.234 ppm for one of the acid signals and +0.91 ppm for one of the diamine signals.The reason of these large Dd values is not only ionization but, in a greater extent, complexation of the molecules due to an additional functionality in the acid residue. This is evident from the spectrum of the salt of model acid 11 where Dd are no higher than +0.04 and +0.23 ppm for the acid and diamine signals, respectively.¶ The spectra of 2:1 salts of enantiomeric acids (R)-1 and (S)-1 have different Dd.†† The enantiomeric differences of Dd (DDd = = DdR-enantiomer – DdS-enantiomer) reach +0.056 ppm for one of the acid signal (Table 1, entry 2).However, the spectra of mixtures of acid enantiomers could differ from the superposition of the spectra of individual enantiomers due to formation of a heteroneutral RS-salt mentioned above (Scheme 2).Indeed, we found that the mixing of two solutions of diastereomerically individual salts produces a new spectrum with single diamine signals, but with enantiomerically splitted signals of the acid with new DDd values. Luckily enough these new DDd values are even larger for some signals (Table 1, entry 1) thus providing an excellent measurement method for acid enantiomer ratios.In a particular case of acid 1, the detection of 1% of minor enantiomer is easy § Dd/ppm for 0.6:1 and 2:1 salts, respectively. (S)-1 residue: –0.23, –0.26 and –0.208 (d, 1H, H2, J 3.6–4.0 Hz, shift from 4.37 ppm), –0.09, –0.14 and –0.138 (br. t, 1H, H3, J 5.8–6.0 Hz, shift from 4.15 ppm), +0.02, 0.00 and +0.03 (q, 2H, C8H2, J 6.4–7.3 Hz, shift from 1.75 ppm), –0.02, –0.06 and –0.04 (s, 3H, C11H3, shift from 2.20 ppm); (R,R)-DADPE residue: +0.26, +0.36 and +0.38 (d, 1H, CHAHB, J 11–13 Hz, shift from 3.47 ppm), +0.27, +0.36 and +0.36 (d, 1H, CHAHB, J 11–13 Hz, shift from 3.53 ppm), +0.19, +0.38 and +0.75 (br.s, 1H, CHN, shift from 3.70 ppm), £ 0.06 (all other).¶ Dd/ppm for 2:1 salt. Acid 11 residue: –0.02 (t, 2H, C2H2, J 7.5 Hz, shift from 2.50 ppm), +0.02 (quint., 2H, C3H2, J 7.5 Hz, shift from 1.84 ppm), +0.04 (td, 2H, C4H2, J 7.5 and 2.5 Hz, shift from 2.26 ppm), +0.03 (t, 1H, H6, shift from 1.95 ppm); (R,R)-DADPE residue: +0.14 (d, 1H, CHAHB, J 12.9 Hz), +0.23 (d, 1H, CHAHB, J 12.9 Hz), +0.12 (br. s, 1H, CHN), £ 0.05 (all other).††Dd/ppm for 2:1 salt. (R)-1 residue: –0.234 (m, 1H, H2), –0.082 (br. t, 1H, H3, J 6.0 Hz), +0.04 (q, 2H, C8H2, J 7.4 Hz), –0.05 (s, 3H, C11H3); (R,R)-DADPE residue: +0.36 (d, 1H, CHAHB, J 11.2 Hz), +0.38 (d, 1H, CHAHB, J 11.2 Hz), +0.91 (br. s, 1H, CHN), £ 0.10 (all other). by measurement of signal intensities of the proton H3 in spite of a multiplet character (W1/2 19 Hz) of this signal reporter.Analogously, with the 2:1 (if possible) salts of the enantiomer mixtures, DDd were measured for all other chiral acids shown in Scheme 1 (entries 3–11 in Table 1). In all cases, with only two exceptions, the enantiomeric differences were observed with DDd values sufficient to measure the ratio of enantiomers. One of the obvious exceptions is acid 9.Amino acid 9 exists as an internal salt and is not transformed into a salt with the (less basic) external base (R,R)-DADPE thus lacking considerable Dd. The second exception, Mosher acid 3, is most probably the result of an accident. The spectrum of the corresponding salt shows substantial Dd for acid signals‡‡ and large Dd values as well as DDd for diamine signals. The latter were used for the analysis of the enantiomer composition of DADPE.6 A specific and sometimes unique property of DADPE as a chiral shift reagent is its long-range action in enantiomeric proton discrimination.Significant DDd are generated for the signals of protons separated from the location of a charge in an anion of the acid by five and even larger number of bonds. In acid 8 (Table 1, entry 9) DADPE produces DDd 0.059 ppm for the signal of one of the methyl groups remote from the charge by more than 5 Å.For comparison, with DPDAE DDd 0.013 ppm was observed in a similar case (methyl group of camphanic acid).2 Thus, acyclic (1, 2 and 10), cyclic (5–7), and threedimensional bicyclic acids (4 and 8) can be analysed with DADPE for the enantiomer ratio. The shifting properties of DADPE are in some contrast with the observation that its close analogue, N,N'-dibenzyl-DPDAE, has greatly reduced enantiodifferentiating properties.2 However, this distinction is in agreement with the remark5 that, in contrast to DADPE, N,N'-dibenzyl-DPDAE was not an effective controller for enantioselective dihydroxylation reaction due to greater rotational flexibility.Another possible application of (R,R)-DADPE (or its equally available enantiomer6) as a chiral shift reagent is the detection of compound chirality. An example is dibromide 7 with an unknown relative configuration of bromine atoms.§§ The observation of large DDd values in the spectrum of its salt (entry 8, ‡‡Dd/ppm for 2:1 salts. (R)- or (S)-3 residues: –0.09 (br. s, 3H, OMe, shift from 3.59 ppm), +0.07 (d, 2H, 2o-HPh, J 7.6 Hz, shift from 7.49 ppm); Dd for (R,R)-DADPE residue see in ref. 6.§§ Dibromide 7 was prepared by partial debromination of the corresponding 2,4,5-tribromide.8 ARH + NN NNH+·AR – NNH2 ++·AR –AR – R-salt RR-salt NNH+ + AR – (or AS –) NNH2 ++·AR –AS – R- or S-anion RS-salt R-cation S-cation NNH2 ++·AR – + AS – (or AR –) NNH2 ++·AS – + AR – (or AS –) ASH + NN NNH+·AS – NNH2 ++·AS –AS – S-salt SS-salt Scheme 2 Salt formation equilibria with the participation of (R,R)-DADPE.ASH ARH ASH ARH aMolar ratios. The sum of enantiomers is taken as an acid concentration. bDDd = d (signal of R-acid) – d (signal of S-acid). Therefore ‘+’ sign means a more pronounced induced shift in low field or a less pronounced shift in high field.cDifferences between two spectra of individual enantiomers. dSome broadening of the OMe signal is observed. eNo sign for DDd is available due to an arbitrary assignment of the enantiomers. Table 1 Enantiomeric differences of chemical shifts induced in the 1H NMR spectra of racemic or scalemic mixtures of acid enantiomers by (R,R)- DADPE. Entry Acid Acid:DADPE ratioa Chemical shift differences of splitted signals (DDd/ppmb) 1 1 2:1 –0.018 (H2), +0.069 (H3) 2 1c 2:1 –0.026 (H2), +0.056 (H3) 3 2 2:1 +0.008 (H2), +0.010 (OMe), +0.032 (o-HPh) 4 3 2:1 No differencesd 5 4 2:1 –0.010 (HA 4 ), –0.027 (HB 4 ), –0.010 (H5), –0.005 (H6) 6 5e 2:1 0.093 (=CH), 0.000 (CHAHB), 0.008 (CHAHB), 0.084 (CHS), 0.040 (o-HPh) 7 6e 0.6:1 0.106 (=CH), 0.000 (CHAHB), 0.042 (CHAHB), 0.360 (CHSO2) 8 7e 1.5:1 0.129 (=CH), 0.064 (CHBr) 9 8 2:1 –0.059 (C9H3), –0.011 (C10H3) 10 9 2:1 No differences 11 10e 0.009 (CHN), 0.016 (CHO), 0.015 (CAH3), 0.000 (CBH3)Mendeleev Communications Electronic Version, Issue 4, 1999 (pp. 129–170) Table 1) proves unambigously the chirality of the molecule and hence the trans-configuration of the bromine atoms. In summary, DADPE enantiomers appear to be a useful addition to a set of chiral shift reagents for the NMR analysis of enantiomer ratio of various acidic substances. This work was supported by the Russian Foundation for Basic Research (grant no. 98-03-32996) and by ASGL Research Laboratories. References 1 D. Parker, Chem. Rev., 1991, 91, 1441. 2 R. Fulwood and D. Parker, J. Chem. Soc., Perkin Trans. 2, 1994, 57. 3 P. M. Demin, T. A. Manukina, C. R. Pace-Asciak and K. K. Pivnitsky, Mendeleev Commun., 1996, 130. 4 H. C. Kolb, M. S. van Nieuwenhze and K. B. Sharpless, Chem. Rev., 1994, 94, 2483. 5 E. J. Corey, P. D. Jardine, S. Virgil, P.-W. Yuen and R. D. Connell, J. Am. Chem. Soc., 1989, 111, 9243. 6 M. A. Lapitskaya and K. K. Pivnitsky, Izv. Akad. Nauk, Ser. Khim., 1997, 101 (Russ. Chem. Bull., 1997, 46, 96). 7 F. J. Villani, M. J. Costanzo, R. R. Inners, M. S. Mutter and D. E. McClure, J. Org. Chem., 1986, 51, 3715. 8 C. H. DePuy and R. D. Thurn, J. Org. Chem., 1962, 27, 744. Received: 16th December 1998; Com. 98/1413