Sir

In his Concepts essay “Crystal: In search of clarity” (Nature 423, 485; 2003) Gautam R. Desiraju asserts that “to a scientist, 'crystal' denotes solid matter in a more or less ordered form ... According to the International Union of Crystallography, a crystal is any solid that gives a discrete X-ray diffraction diagram.”

Yet there is more to crystals than solidity, for condensation into symmetrical arrays can be governed by repulsive as well as attractive forces. Some ordered arrays that fulfil the criterion of symmetrical diffraction consist not of atoms or molecules, but of discrete objects, or even organisms, such as the Iridoviridae. The rainbow diffraction of visible light by ordered arrays of hydrous silica microspheres is familiar as the play of colour in precious opal, but few realize that an equally arresting display can be seen when such microspheres are widely separated in aqueous suspension by electrostatic forces and Debye screening. The colourful diffraction testifies to long-range order, yet fish can swim through this crystalline glory and it reassembles in their wake.

There are also crystals of crystals, as monocrystalline nanoparticles are equally susceptible to falling into line in repulsive lattices, macrocrystallites of metals in planetary cores, and vast domains of crystalline order created within neutron stars by the strong nuclear force. Good luck in trying to detect their diffraction pattern.

At the limit of artifice are ordered arrays of atoms assembled by manipulation, which can defy the criterion of least energy that Desiraju's fascinating piece invokes. The many ramifications of this include the possibility of creating deliberate arrays of stable isotope atoms. This has important technical ramifications, for while the ballistic phonon free path, and the thermal conductivity of diamond (illustrated in the Concepts essay) can be enhanced by reducing the carbon 13 level in synthetic diamond feed stocks, that improvement can be transcended only when entropy as well as free energy is defied by ordering isotope defects instead of letting a random cloud of anharmonic oscillators destroy the thermal transport capacity of an otherwise perfect lattice (see R. Seitz, Science 250, 1194–1195; 1990).

Desiraju is certainly right in observing that “the meaning and scope of the term 'crystal' can only evolve and expand”.