Molecular graphs as topological objects in space

Stereochemistry deals primarily with distinctions based on rigid geometry, e.g., bond angles and lengths. But some chemical species have molecular graphs (such as knots, catenanes, and nonplanar graphs K5 and K3.3) that reside in space in a topologically nontrivial way. For such molecules there is hope of using topological methods to gain chemical information. Viewing a molecular graph as a topological object in space makes it unrealistically flexible; but if one proves that a certain graph is “topologically chiral” or that two graphs are “topological diastereomers,” then one has ruled out interconversion under any physical conditions for which the molecular graph still makes sense. In this paper, we consider several kinds of topological questions one might ask about graphs in space, methology and results available, and specific topological properties of various molecules.