This paper acts both as documentation and as a citable reference for the automatic tests, implemented in the program package PLATON, that are carried out to validate the result of a small-molecule single-crystal structure determination. Common problems such as wrongly assigned space group symmetry, missed solvent molecules in solvent-accessible voids in a crystal structure, and wrong implied chemistry are addressed.

  What influenced you to design this program?

“...the chemical community that depends on reliable crystallographic results in terms of three-dimensional molecular structures [benefits from using PLATON].”

Validation of small-molecule crystal structure determinations was pioneered by the International Union of Crystallography (IUCr). The number of papers, reporting the results of a structure determination, that were submitted to the IUCr and other journals started to explode. The number of experts available to review such papers was decreasing with an increasing number of non-expert authors. The number of avoidable errors in the literature, sometimes with severe implications, was increasing correspondingly.

The first step taken to address this issue was to create a standard computer-readable format to report the result of a crystal structure determination. This resulted in the so-called CIF standard (largely through the efforts of Syd Hall, editor of the IUCr journal Acta Cryst. C at that time).This opened the way to implement automatic checking of the reported numerical data. I got involved when I was asked to join in this effort with the additional checking features that were already present in my PLATON package.

  What are some of the features/advantages of PLATON?

PLATON is compatible with the widely used SHELX software for structure determination by G.M. Sheldrick. It can be implemented locally and used in any stage of a crystal structure determination. In this way, errors can be detected early rather than during the publication stage where it is often laborious to correct. The program will produce a validation report of issues to be addressed (the so called ALERTS).

  Who benefits from using PLATON?

In general: the chemical community that depends on reliable crystallographic results in terms of three-dimensional molecular structures. Authors, referees, and editors can now concentrate on sorting out the reported issues rather than spending time on trivial checking of the data. Validation also points at unusual features in a crystal structure that may be of interest for further investigation.

PLATON is a collection of software tools. Structure validation is just one of the tools made available in this package that is freely available for academics. This paper is also cited when other related PLATON tools have been used. The validation tool in PLATON is currently part of a WEB service offered by the IUCr in Chester (UK). Most major journals currently require a validation report to accompany a submitted paper that reports a crystal structure. The electronic submission process to the two IUCr journals Acta Cryst. C & E is now fully automated. The paper is automatically checked. In case of serious ALERTS the paper will be returned automatically with a request to the authors to address the problem.

  Has PLATON been further developed since the publication of the 2003 Journal of Applied crystallography paper?

Yes. With validation procedures in place and used, we have detected even more issues to be checked or reported on. As an example, there is now also checking of the reported absolute configuration.

Prof. Dr. A.L. Spek
Crystal and Structural Chemistry
Utrecht University
Utrecht, the Netherlands