The chemistry of materials for artificial Darwinian systems

The best materials for artificial systems able to evolve under natural selection are unlikely to be the same as for natural organisms. This is so particularly for genetic materials, where the overriding need for high fidelity of replication points to crystal growth processes as providing the key part of the mechanism rather than organic polymerizations. The genetic information would then have to be in the form of an aperiodic structure based on some relatively simple crystal defect capable of an immense number of permutations. Here I discuss particularly polytypism and related intergrowth phenomena, using a number of examples to illustrate suitable and unsuitable features.