TOPOS3.2: a new version of the program package for multipurpose crystal- chemical analysis

The principal features of the package are as follows. (i) Management of crystal structure information with DBMS (database management system) which has gateways to the CSD and ICSD databases. (ii) Comprehensive analysis of geometrical and topological properties of crystal structures (employing the programs Dirichlet, AutoCN, ADS, DiAn, IsoCryst and IsoTest). (iii) Special facilities for statistical analysis of large sets of crystal structures (the program StatPack). In addition to the programs contained in the previous version (Blatov et al., 1999), version 3.2 includes the following two programs. (a) IsoTest, which provides an automatic search for the topological similarity (isotypism) in large groups of stoichiometrically and structurally different compounds, on three levels: the whole topological and geometrical similarity (crystal-chemical isotypism; Lima-deFaria et al., 1990); only the whole or partial topological similarity of crystal structures (topological isotypism; Blatov, 2000); topological similarity of separate atomic subnets and packings. (b) HSite, which searches for hydrogen positions in crystal structures of organic, organometallic and inorganic compounds. Besides traditional methods of geometrical and statistical analysis, and graphical representation of crystal structures (the programs DiAn, IsoCryst and StatPack), there are two novel concepts used in TOPOS algorithms: the concept of an atomic domain represented as a Voronoi±Dirichlet polyhedron (the program Dirichlet; Blatov et al., 1995) and the concept of a periodic net described as a contracted graph (the programs ADS, AutoCN and IsoTest; Blatov, 2000). The program IsoTest automatically enumerates all variants of topological representation of crystal structures and ®nds similar ones through a given list of compounds by comparing coordination sequences (Brunner & Laves, 1971) of corresponding atomic subnets (Blatov, 2000). The program HSite uses characteristics of Voronoi±Dirichlet polyhedra to predict the optimal positions of hydrogen atoms and orientation of atomic groups.