An atom is stereogenic if switching any two atoms or groups of atoms that are bound to it results in a pair of stereoisomers. So far, molecules with no or only one stereogenic atom have been discussed. Very often the situation is more complex; indeed, there can be several stereogenic atoms in a molecule. A molecule with only one stereogenic atom has only two stereoisomers—the R and S enantiomers. If there are two stereogenic atoms in a molecule, both can be either R or S. Thus, there are four possibilities: RR, SS, RS, and SR. Three stereogenic atoms would lead to eight possibilities: RRR, RRS, RSR, SRR, SSR, SRS, RSS, and SSS. The formula for finding the maximum number of stereoisomers X is X = 2ⁿ, where n is the number of stereogenic atoms in the molecule.

The formula X = 2ⁿ reliably gives the maximum number of stereoisomers, but in situations of high symmetry it fails to give the real number. For example, it fails for 2,3-dichlorobutane [H₂Cl₂(CH₃)₂]. One pair of enantiomers, SS and RR, does appear. But the other combination gives an identical “pair” of SR compounds. This happens because 2,3-dichlorobutane contains an internal plane of symmetry. The result is fewer than the maximum number of stereoisomers predicted by the formula. Three stereoisomers are possible: one pair of enantiomers (A and B) and an achiral molecule C, called a “meso compound.” A meso compound is an achiral molecule that nonetheless contains a stereogenic atom.

In order to find molecules that are enantiomers, one must draw the mirror image of the original and see if they are superimposable. That is the only absolutely safe way to do it. It might be suggested that there is something special about a molecule containing four different groups attached to one carbon. The question now is whether the presence of such an atom (usually carbon) is either sufficient or necessary for the molecule to be chiral. The answer is no in each case. Although looking for such carbons is a good way to start a search for enantiomers, there is no way to avoid the ultimate necessity of writing out the mirror image and checking for superimposability. To test the question of sufficiency, for example, look at the meso compound C of 2,3-dimethylbutane. It certainly does contain a carbon attached to four different groups. The indicated carbon C₂ is attached to hydrogen, a methyl group, a chlorine, and the rest of the molecule. Yet C is achiral.

There are many compounds whose molecular architecture makes them chiral but that do not contain an atom attached to four different groups. One classic example is hexahelicene, a molecule composed of six benzene rings connected to each other. The molecule coils in the form of a spiral so that the atoms of the last ring do not impinge on the atoms of the first ring. The result is a left- or right-handed screw form, and the molecule is chiral.

Diastereomers

Cyclohexane is achiral, as are both axial and equatorial methylcyclohexane. The two methylcyclohexanes (axial and equatorial methyl group) are stereoisomers, but they are not enantiomers. Such isomers—stereoisomers that are not mirror images—are called diastereomers. The molecules cis- and trans-2-butene are diastereomers, as are cis- and trans-1,2-dimethylcyclopropane. However, in dimethylcyclopropane, the cis compound is achiral, but the trans compound exists as a pair of enantiomers. Therefore, there are three stereoisomers of 1,2-dimethylcyclopropane.