It is anticipated that predicted/enumerated crystal structures
may soon (during the next decades) exceed in number the
really determined ones.
It is believed that the "one stucture - one publication" which
is almost the rule for the real crystal data will never be applied
for the virtual ones.
Therefore, it is offered room for storage of these predicted or
enumerated crystal structures at a new database, the PCOD
(Predicted Crystallography Open Database), a subset of the
COD containing real structures : [Only registered users see links. ]
As a beginning, PCOD contains an enumeration of 150 SiO2
polymorphs built up by following geometrical rules. Predicting
software should propose in first place in their lists the really
observed compounds. This is the case of the GRINSP beta
version which always places quartz, tridymite, cristobalite, EDI,
ABW, BIK, JBW, SOD, YUG, etc, at the first places of the smallest
predicted cells, with parameters close to the real ones.
The next step for PCOD, if high standards are attained in the
quality of the predicted models, is to produce a set of predicted
powder patterns usable by search-match software. Thanks to
the PCOD, phase identification would be extended to the hitherto
unknown compounds, saving time in difficult structure determinations.
But that dream is far from being realized yet, in spite of thousands
of predicted crystal structures listed in the past 20 years (including
thousands of zeolites).
So, do not only mention and show nice pictures of your huge quantities
of predicted crystal structures in your manuscripts, also deposit their
atomic coordinates at PCOD, as CIF files ! Thanks.