Current and Future Developments |
Current and Future Developments
The lay-out, structure and nature of the database as described here is largely historical so as to give the student a clear understanding of the concepts of search/matching and the PDF. In fact the database and search/matching is now almost entirely computerised and has developed over the last 15 years to the point where not only the card index system is redundant, but so too are the accompanying books and tables. This is a logical way to proceed as computing is relatively inexpensive and the database is now so large that it is not economical or practical for most end-users to posses a complete set of printed books.
Currently the PDF database is incorporated into the data acquisition and analysis software supplied by the manufacturers of laboratory X-ray powder diffractometers. Such software includes the manufacturers own data interface and search/match capabilities. The database is also available directly from the ICDD on CD-ROM with associated retrieval software and is marketed as PDF-2. It is essentially the "traditional" database in CD-ROM form. Sub-divided into several material type categories, all entries are in the one database and may be searched using Boolian logic. Results can be read out in many formats including the original card index style. More recently, the database includes many calculated entries derived from other crystallographic databases.
Already available, but still being heavily developed, is the ICDD's PDF-4. This is a "Relational Data Base" (RDB) on DVD which essentially is a library of databases (31 in the 2002 release) categorised under different headings and giving much more flexibility to Boolian searches. Subfiles for specialist use include minerals, metals & alloys, pharmaceuticals, organics, polymers, ceramics, corrosion products, cement materials, forensics, and explosives.
Also set to change is the traditional numbering system. This has been necessary because of the inclusion and proliferation of the many calculated patterns made available by incorporating other databases such as the Inorganic Crystal Structure Database (ICSD) and the Cambridge Structural Database (CSD). The change is outlined in the table below and has already been included in the 2003 PDF-4/Organics release. Where cells in the table have double numbers, the top one refers to the first ever entry for that particular source and the bottom one to the last entry of the 2002 release.
Database Source | Database Code | Set | Entry Number | Old 6-digit PDF # |
New 9-digit PDF # |
---|---|---|---|---|---|
PDF-2 Experimental Entries |
00 | 01 52 |
0001 2500 | 01-0001 52-2500 |
00-001-0001 00-052-2500 |
ICSD | 01 | 70 89 | 0001 6878 |
70-0001 89-6878 |
01-070-0001 01-089-6878 |
CSD | 02 | 60 72 | 0001 2828 |
60-0001 72-2028 |
02-060-0001 02-072-2828 |
NIST Database | 03 | 65 65 | 1000 3990 |
65-1000 65-3990 |
03-065-1000 03-065-3990 |
Finally, included in the 2002 release are approximately 1000 fully-digitised experimental patterns, i.e. full experimental data sets rather than just extracted d and I values. The concept of storing full profiles as a PDF-3 database was originally proposed many years ago, but the incorporation of other crystallographic databases together with the development of relational databases that now form PDF-4 has taken precedence. This, along with the software to deal with these files, is seen by the ICDD as the trend for the future. The implication for end-users is the ability to use raw profile data for whole-pattern matching. On the negative side, at least for some users, is the current trend towards the licensing of the PDF database on an annual subscription basis, as opposed to the outright purchase of a copy with an indefinite lifetime. How successful this turns out in practice remains to be seen: watch this space!
© Copyright 1997-2006. Birkbeck College, University of London. |
Author(s):
Martin Vickers |