crystallographica

The product. crystallographica is designed to run on IBM-compatible PCs (386 or higher, with at least 4 Mb RAM) under Windows (3.x, NT or 95). It has been developed to provide a flexible resource for those requiring access to crystallographic information and procedures. Whilst there are features that are useful to fully-fledged crystallographers the package is aimed principally at materials scientists and does not presume or require more than a modest amount of crystallographic knowledge. Essentially crystallographica is a Pascal interpreter, augmented by a large number of special functions and procedures, combined with two plotting packages--CrystalPlot to display the structure in hand and Graph to plot out data. The Multiple Document Interface allows a number of windows to be open simultaneously: besides CrystalPlot and Graph there can be Interpreter windows for interactive access, Console windows for data input, Editor windows for modifying Pascal scripts or datasets etc. The Graphical User Interface is well exploited, the functions of icons (not too many, not too small) are self-evident and the items in the pull-down menus are well supported by on-line help if one's personal RAM dips in performance.

Functions related to scattering handle X-rays or neutrons but not electrons. A very useful procedure has been included that produces the powder diffraction pattern of a specified material. A monochromatic version is produced directly whilst a Ka2 pattern may be added if appropriate. crystallographica can also be used to generate the .ins files that are required as input to the SHELXL structure refinement program. The SHELXL output can be reimported, the structure viewed and an updated .ins file produced.

Installation requires 5 Mb of free disk space. It is initiated in Windows Program Manager and proceeds effortlessly, involving little more than a few mouse clicks.

Use and testing. There are two ways in which the basic data for a crystal may be entered: either by importing a .CIF file (Crystallographic Information File), a generous number of which have been provided, or by entering atom types, positions, lattice parameters etc. through dialogue boxes. All this information is stored in a Pascal `data structure' and the program ensures that any changes made to these data do not introduce any inconsistency--thus if a lattice parameter change violates a space-group assignment the latter will be altered appropriately. The structure may be observed via CrystalPlot, atom radii and bonds tailored to taste, viewing angle selected and so on. One small criticism here is the lack of perspective-- a transformation that added depth would greatly help 3D visualisation.

To pass beyond the above minimum level of use brings the user up against Pascal. Be not afraid! The first 47 pages of the manual constitute a tutorial guide and provide plenty of simple examples of Pascal statements and syntax. The final 46 pages comprise a substantial Pascal primer. Those familiar with the language will be reassured to know that a ** operator has been implemented so that integral powers of numbers can be computed. The execution of an interactive session is facilitated by an `alias' capability, by name completion using the ESC key and by a UNIX-like history feature that allows short cuts to previous commands. Your reviewer, however, was frequently frustrated by this mode of use and much prefers to write simple Pascal scripts which may be added to the corpus of scripts already provided.

My own use of crystallographica has largely been to simulate the X-ray response of crystals to different perturbations: in-diffusion of ions from electrodes in LiNbO3 and ion-disordering in KTiOPO4 under electric field and inner displacement of atom positions in GaP under uniaxial stress. These all involve tracking the behaviour of the structure factor. Effects can be made vivid in CrystalPlot by linking colour with site-occupancy and in Graph by a plot-as-you-go build up. Incorrect hypotheses can be rapidly identified, the script edited and a new run started. This work revealed a serious shortcoming of the program---anomalous dispersion is not handled correctly in the vicinity of absorption edges so the structure factor may be wildly in error. It is possible to get round this by entering one's own data into the file USERPROF.i which is provided as a repository for any constants, data, aliases etc. that one wishes to set up. The widely-available Cromer-Liberman FPRIME program could be used for this purpose.

In another task a colleague provided lists of atomic positions representing exotic carbon structures, wishing to know the spacegroups. It was simple to scan his data, deduce the structures were orthorhombic or tetragonal and exhaust speedily the possibilities using crystallographica, obtaining informative pictures at the same time. All in all the product is very easy to use.

Conclusions. This is a very well-conceived package, loads of crystallographically-secure procedures that the user can weld together to solve problems which would otherwise require time-consuming programming. It is especially good for simulation where the gradual effects of some change may be clearly seen simultaneously in graphics and in plots. The powder diffraction pattern feature will appeal strongly to the increasing numbers of workers in high-pressure research, inter alia. I have some reservations about the quality of the hardcopy output from CrystalPlot but am informed that the version 1.1 release (this review relates to 1.03) due later this year will include the ability to export any crystal structure drawing as a VRML file. This can then be processed to give a Silicon Graphics-style photo-realistic plot which may include thermal ellipsoids, polyhedra and Miller planes in addition to standard features. (Editor's Note: The VRML ouput was on show at the IUCr in August 96, but the resolution of your hardcopy may still be limited to that of your screen, unless you have a smart Web Browser.)

Unlike software dedicated to limited tasks, an individual crystallographica implementation will grow by the accumulation of the user's own scripts, and thus become yet more powerful. An analogous process at the developers' end should lead rapidly to a solution of the anomalous dispersion problem mentioned above.

Value for money. I think it's over-priced though I know of no other similar product with which to compare it. However it's no more than the cost of an electronic module that might form a very small part of the experimental apparatus that produces the results that crystallographica is called upon to analyse. Look on it as an instrument!

Support is available by email or fax. Registered users are entitled to free upgrades until the next major release of the product. In addition a space on the [email protected] Click here to return to BCA homepage