ABSTRACT
The term crystal engineering was introduced by Schmidt [1] in the context of the organic solid state, specifically the control of photochemical solid-state reactions. However, the field remained dormant until the groups of Desiraju [2,3] and the late Etter [4] broadened the scope of crystal engineering to include organic solids in general. Nevertheless, despite the seminal work of these groups, the state of the art in 1988 was described by Maddox [5] as follows: "it remains one of the continuing scandals in the physical sciences that it remains impossible to predict crystalline architecture from a knowledge of chemical composition." This statement remains largely true today; i.e., it is unrealistic to attempt to predict crystal structure from a molecular formula. Indeed, Maddox's comments were recently echoed by his successor as editor of Nature [6]. Fortunately, our ever more advanced understanding of supramolecular chemistry and the strength and directionality of noncovalent bonding has afforded some notable successes in recent years, primarily when molecules or mixtures of molecules are carefully selected for their ability to engage in intermolecular interactions such as hydrogen bonding (including weaker (C-H . . . 0 and X-H . . . pi) and pi-stacking. There has been particular success at using hydrogen bonds to contrast diamondoid solids from single [7,8] or multiple component systems [9,10]. Indeed, other noncovalent interactions such as N . . . Br interactions have also been exploited to generate diamondoid solids [11]. The result of such studies, especially those in recent years, has demonstrated that the relationship between supramolecular chemistry and crystal engineering is indeed close. Desiraju has even gone as far as to say that crystals are supramolecules par excellence; i.e., crystal structures are de facto examples of supramolecular chemistry [3]. However, it should be realized that, although a primary goal of crystal
engineering is to develop de novo methods for crystal structure prediction, pattern recognition and exploitation thereof are presently the modus operandi.
