Self-disproportionation of enantiomers of (R)-ethyl 3-(3,5-dinitrobenzamido)-4,4,4-trifluorobutanoate on achiral silica gel stationary phase

https://doi.org/10.1016/j.jfluchem.2005.11.004Get rights and content

Abstract

The very unusual phenomenon of separation of enantiomers by self-disproportionation on achiral silica gel stationary phase has been studied in detail using (R)-ethyl 3-(3,5-dinitrobenzamido)-4,4,4-trifluorobutanoate as an example. The appearance and magnitude of self-disproportionation were shown by this study to depend on the optical purity of the starting compounds and the nature (polarity) of an eluent used. Based on the results described by us and on previous literature data we conclude that it is not safe to assume that any type of chromatography is a reliable method for purification of non-racemic mixtures for determination of their enantiomeric purities. We recommend that chemists dealing with non-racemic compounds should first run an ESD-test (enantiomer self-disproportionation test) proposed herein before using chromatography for any purification. Moreover, one of the serious consequences of our finding is that all fluorine-containing chiral reagents and drugs, which are currently on the market, as well as all previously reported literature data on the stereochemical outcome of asymmetric transformations involving fluorine-containing compounds, should be carefully re-evaluated if column chromatography on silica gel was involved as a purification step.

Introduction

Until asymmetric synthesis reaches the level of ultimate perfection that allows preparation of chiral compounds in enantiomerically pure form, the separation of enantiomers will remain of paramount importance in production of enantiomerically pure compounds. Conceptually, there are two available approaches to separation of enantiomers: using either external or internal chirality. The former, the most explored strategy that clearly dominates the separation methodology, features various well-known and practically useful methods based on the formation of strong [1], medium [2] and weak [3] diastereomeric relationships. The second approach, making use of internal chirality, is known to every chemist as a re-crystallization technique where the separation of enantiomers takes place due to the strong crystal lattice forces that bring about diastereomeric interactions in the aggregate [4]. However, the re-crystallization approach can be applied only to already highly enantiomerically enriched (>90% ee) samples. Furthermore, the search for efficient solvents can be very time consuming, with success being uncertain. In contrast, separation methods based on the preferential homochiral/heterochiral interactions leading to formation of diastereomeric relationships in solutions of optically enriched solutes are virtually unknown and are largely unexplored as a separation technique. The basic principle for such separations is illustrated in Fig. 1.

Considering the intermolecular interactions between enantiomers of a chiral compound in solution, one can envision two modes of association based on preferential formation of heterochiral or homochiral species (Fig. 1). In both cases formation of dimeric or more complex oligo/polymeric associates can be expected. Thus, when heterochiral associations are preferred one can expect formation of racemic species 2nSR or (SR)2n plus excess of enantiomer (R), which are different chemical entities and therefore may be separated without application of any external element of chirality. In the case of preferential homochiral associations, the situation is a bit subtle as the formation of dimers will result in different number of enantiomeric (S)(S) and (R)(R) pairs with identical scalar properties. These dimers therefore cannot be separated. However, if oligo/polymeric associations are favored, formation of aggregates of different molecular weight is possible and separation of these may be achieved.

Such interactions, in general, may be expected to be very weak and therefore evaluation of the enantiomeric association mode in solution has not been an easy task. The first physical observation of the associations of enantiomers discussed above was reported by Williams et al. [5], who studied 1H NMR spectra of non-racemic dihydroquinines and detected two sets of peaks for some protons with peak areas being proportional to the relative ratio of the enantiomers. Further evidences of enantiomeric discrimination in solution were reported to be observed in the 1H NMR spectra of the nonracemic samples of certain chiral compounds such as carboxamides [6], dicarboxamides, [7] phosphinamides [8], and phosphinothioic acids [9], all of which can easily form relatively strong intermolecular hydrogen bonds.

These associations of enantiomers leading to mixtures of compounds with different scalar properties are believed to be responsible for several remarkable phenomena observed for nonracemic compounds, such as nonlinear behavior of optical rotation [10] and UV absorbance [11], as well as nonlinear effects in asymmetric catalysis [12]. Furthermore, separation of enantiomers based on solely internal chirality of nonracemic mixtures could lead to models for amplification of optical activity of chiral compounds under prebiotic conditions [13]. However, the most fascinating opportunity offered by the enantiomer associations is the possibility of using internal chirality of nonracemic mixtures for their optical purification/separation on achiral phase chromatography [14].

In 1983 Cundy and Crooks reported the separation of an excess 14C-labelled (S)-(−)-nicotine enantiomer (second fraction) from the racemate (first fraction) on an achiral HPLC system [15]. Since that time a handful of sporadic publications on such separations of various organic compounds under HPLC [16], MPLC [17], flash [18] and regular [19] chromatography have appeared in the literature. However, this phenomenon still remains virtually unknown [20] to most practitioners in academic and industrial laboratories. As a result of this lack of awareness of potential enantiomeric enrichment, column chromatography on silica gel is generally accepted and routinely used as a safe method for purification of optically active compounds that will not alter the original enantiomeric composition.

Where previously reported in the literature, different terminology has been used to describe this phenomenon [21]. Most frequently it is called “enantiomeric enrichment on achiral phase chromatography”. This definition is obviously incorrect and rather misleading, since part of a non-racemic sample indeed undergoes enantiomeric enrichment while the rest of the sample becomes more racemic. In a recent communication [22] we have introduced a term “self-disproportionation of enantiomers” or “enantiomer self-disproportionation” to describe this phenomenon. We have also reported on the truly remarkable amplification of enantiomer self-disproportionation induced by a trifluoromethyl group. There are many chiral fluorine-containing drugs on the market and chromatographic procedures are frequently used to assess enantiomeric purity. Thus, one of the serious consequences of our finding is that it may be appropriate to re-evaluate data on optical purity of many of these. In addition, previously reported literature data on the stereochemical outcome of asymmetric transformations involving fluorine-containing compounds should be carefully re-evaluated.

To provide the first systematic study of the phenomenon of self-disproportionation of enantiomers, we report in this paper full experimental data on self-disproportionation of the enantiomers of (R)-ethyl 3-(3,5-dinitrobenzamido)-4,4,4-trifluorobutanoate under the conditions of achiral silica gel stationary phase chromatography.

Section snippets

Results and discussion

In the course of our studies on the 1,3-proton shift reaction that we had discovered [23], we were developing the asymmetric version of this reaction using (S)-1-phenylethanamine (PEA) as a source of chirality. In particular, keto-ester 1 (Scheme 1) chemoselectively [24] was converted to enamine 2 and isomerized, through formation of the intermediate imine 3, to Schiff base 4. For determination of the stereochemical outcome in this reaction, Schiff base 4 was hydrolyzed to amino-ester 6 and

Experimental

All experiments were conducted under the same conditions. Silica gel (mesh 200–300) was purchased from Natland international corporation. The temperature was T = 23 °C. The amount of sample was almost the same in all of the experiments m = 0.0300 ± 0.0001 g. The columns that were used had the same features: diameter d = 16 mm and longitude of silica gel along the column l = 230 mm. The amount of each fraction was V = 7 mL. Analysis of enantiomeric excess was determined by HPLC analysis. (Sumichiral QA = 4500, l = 25

References (28)

  • P. Diter et al.

    J. Org. Chem.

    (1994)
  • H. Takahata

    Yuki Gosei Kagaku Kyokaishi

    (1996)
    K. Tanaka et al.

    Chem. Soc., Perkin Trans.

    (1998)
    B.V. Ernholt et al.

    Chem. Eur. J.

    (2000)
  • For instance, in the most recent book (a) and review article (b) on non-conventional methods for resolution of...F. Toda

    Enantiomer Separation

    (2004)
    E. Fogassy et al.

    Synthesis

    (2005)
  • T. Katagiri et al.

    Chem. Lett.

    (1996)
  • An example: formation of diastereomers using a covalent bond. In this case almost any available techniques can be used,...P. Newman
    (1978/1981/1984)
    T. Vries et al.

    Angew. Chem., Int. Ed. Engl.

    (1998)
    J. Jacques et al.

    Enantiomers, Racemates and Resolutions

    (1994)
  • R. Tamura et al.

    J. Am. Chem. Soc.

    (2002)
  • T. Williams et al.

    J. Am. Chem. Soc.

    (1969)
  • A. Dobashi et al.

    J. Am. Chem. Soc.

    (1986)
    B.S. Jursic et al.

    J. Org. Chem.

    (1992)
  • M.T. Cung et al.

    Biopolymers

    (1978)
  • M.J.P. Harger

    J. Chem. Soc., Perkin Trans.

    (1977)
  • M.J.P. Harger

    J. Chem. Soc., Perkin Trans.

    (1978)
  • A. Horeau

    Tetrahedron Lett.

    (1969)
    A. Horeau et al.

    Tetrahedron

    (1974)
  • Cited by (0)

    View full text