How salt sprinkles? It's as clear as crystal

This book stands very firmly in the tradition of seeking a molecular understanding of crystallisation. Nowadays such a book would be incomplete without a chapter on computer simulation issues. This is precisely the topic of the first chapter by Alexander Izmailov and Allan Myerson, where the reader can learn about techniques such as molecular dynamics and Monte Carlo firmly set in their background in statistical mechanics and thermodynamics, and put appropriately into the context of crystallisation.

This is followed by an excellent presentation by Myerson of the main concepts of crystallography and crystallisation mechanisms. With the background firmly established, Robert Docherty and Paul Meenan return to the main subject of the book: the computational aspects of crystallisation.

In the introduction we learn to appreciate how important crystallisation is to our lives, with issues ranging from the control of colour in pigments and dyes to the bio- availability of pharmaceuticals.

This is followed by an overview of the main issues and methods in the modelling of crystallisation. We encounter methods to address structure calculations, crystal shape prediction, property estimations and crystal-packing calculations, and we learn to appreciate crystallisation as a process of molecular recognition and self-assembly. While modelling turns out to be a useful tool when combined with experimental techniques, for example in the "screening" of potential substances, the ultimate goal is still some way off: the ab initio design of novel materials. This task requires the ability to predict the solid structure from the molecular formula and crystallisation environment alone.

The remaining chapters expand on some of the most important avenues that are being explored to achieve this goal. Isabelle Weissbuch, Meir Lahav and Leslie Leiserowitz believe that the structure formation during crystallisation can be understood as "surface-active" process. Understanding and modelling the molecular-recognition process at the surface enables us to control growth and dissolution of crystals by using tailor-made auxiliaries. These will affect the structure and morphology of the crystals, and the macroscopic properties. A simple material example is rock salt, which is crystallised together with an anti-caking agent to produce small grains rather than more elongated crystals, so the salt pours easily.

In chapter five, C. S. Strom et al . present an in-depth treatment of modelling of ionic crystals by means of the Hartman-Perdok theory, which has led to successful calculations of crystal habits in many systems. It impresses by its depth and clearly demonstrates the power of bringing together a matured theory and computer-simulation methods in generating quantitative results.

Chapter six, by Patrick Stahly and Stephen Byrn, returns to symmetry issues, in the context of the structure of chiral organic pharmaceuticals. We hear that the birth deformities that were caused by thalidomide are due to just one of its enantiomers. A better awareness of the different forms at the time could have prevented this. Similarly, different crystal forms of the same element or molecule, called polymorphs, may have very different properties, affecting, for example, the rate of in vivo delivery of drugs. Molecular modelling methods for the ab initio prediction of polymorphs have progressed greatly over the past decade. This includes, to my knowledge, even cases of chiral molecules, in contrast to what is stated by Stahly and Byrn.

While the editor has selected some topical issues in crystallisation, other important areas are hardly mentioned. The crystallisation of macromolecules, for example affects the properties of everyday materials as well as the structure determination of proteins. This deserves more than a paragraph on long-chain hydrocarbons in chapter four. The concluding statement that very little work has been done on crystals composed of macromolecules ignores years of research in this area. In this context a somewhat more extensive bibliography of related textbooks, review articles and websites would have been desirable.

The emphasis of the book is on the scientific rather than the computational interpretation of modelling. It counteracts the trend to regard modelling as a black box, which is to some extent brought about by the wide availability and ease of use of commercial modelling software. While it is commendable that the book aims to redress this balance, I would have liked to see a bit more enthusiasm for the impetus that computer simulation has given to the subject. Also, there is not much mention of visualisation, which in my view is an inspiration as well as a powerful tool for understanding and communication in molecular modelling. In summary, despite some minor shortcomings and small errors, the book can be warmly recommended to its intended readership of scientists and engineers from graduate level onwards.

Gerhard Goldbeck-Wood is assistant director of research, department of materials science and metallurgy, University of Cambridge.