Louis Pasteur was the first documented person to separate stereoisomers. He did so by noticing that crystals of tartaric acid had either a left-handed crystal or a right handed crystal, and then he used a microscope and tweezers to separate the crystals from each other. The discovery of stereoisomerism in tartaric acid crystals, like Fleming's discovery of penicillin, was serendipitous, since very few stereoisomers form separate crystals, and tartaric acid only forms such separate crystals at cold temperature (it is speculated that it was a cold day when Pasteur synthesized the tartaric acid, and that he had recrystallized it on his cold windowsill).

So how are enantiomers currently separated? Even if one could use Pasteur's method of separation, one would be hard-pressed to find workers as patient as Pasteur, willing to spend their days under a microscope separating the different crystal forms with a pair of tweezers.

Compounds that are enantiomers of each other have the same boiling points, refractive indices, reactivities, melting points, and solubilities. So if a mixture of enantiomers is obtained, how can they be separated? They can't be recrystallized, since they have the same solubility properties; they can't be distilled, since they have the same boiling points; they can't be run on regular achiral silica chromatography columns since they have the same attractions to the stationary phase. Because of their difficulty in separation, a great deal of current study today is done to generate catalysts that form only a single stereoisomer of a compound. There are, however, techniques that can be employed to separate enantiomers.

Conversion to diastereomers
One way to separate enantiomers is to chemically convert them into species that can be separated: diastereomers. Diastereomers, unlike enantiomers, have entirely different physical properties--boiling points, melting points, NMR shifts, solubilities--and they can be separated by conventional means such as chromatography or recrystallization.

If it was desired to separate a mixture of an R and S carboxylic acid, for example, this mixture could be reacted with a single enantiomer of a chiral amine to make the diastereomic ammonium salts that could then be separated. Once the diastereomic salts have been separated, mineral acid can reprotonate the carboxylic acid to reform the original enantiomers. This is a general, three step, technique for separating enantiomers:

1. React the enantiomers with a single enantiomer of another compound to form diastereomers
2. Separate the diastereomers by conventional means (chromatography, recrystallization)
3. Regenerate the original enantiomers, now separated

S-Brucine

A common amine used in these reactions with carboxylic acids is S-Brucine, an alkaloid found in only its S enantiomer. S-Brucine is used because it is commercially available, although in theory any amine that is purely one enantiomer should work just as well.

                                               Easily separated

Other reactions that form diastereomers from various functional groups have been described in the literature, although this one with carboxylic acids is particularly effective because it involves a noncovalent modification, and the reactions are quick and performed in high yields.

Chiral chromatography
Another technique for separating enantiomers is chiral chromatography. While enantiomers cannot be distinguished in achiral environments, such as a solvent system or by normal silica gel chromatography, they can be distinguished in chiral environments, such as in the active site of an enzyme, or in a chiral stationary phase of a column. In a chiral column, achiral silica gel (SiO₂)is converted into a chiral stationary phase by a reaction with a chiral molecule. Once the enantiomers that need to be separated are run down the column, one enantiomer will "stick" to the stationary phase better than the other, and there will be separation (of course, a disadvantage is that chiral silica gel is much more expensive than standard silica gel).

In this hypothetical example of an interaction between a chiral stationary phase (left) with an enantiomer of a biphenyl derivative (right), there is a three-point interaction, with the carboxy groups aligning with the amino groups and the aromatics lining up with each other to form pi stacking interactions. The enantiomer of this biphenyl would not be able to have all three of these interactions because its groups would not be aligned correctly, and, consequently, it would stick less to the chiral stationary phase and filter off the column first.

Click here for the animation of chiral flash chromatography (40kb, opens in a popup window).

A diagram of chiral column chromatography: the enantiomer of the biphenyl that can form the three-point interaction with the stationary phase (red band) sticks better and filters off the column after its enantiomer (green band).