Using CIF for Powder Diffraction
Most crystallographers are aware of the IUCr's efforts to
define a standardized format for crystallographic information
known as the Crystallographic Information File (CIF). Those
crystallographers who are not may wish to review the following
section, A Very Brief Introduction to CIF or consult the
enclosed references. The initial definitions developed for CIF
were designed to suit the needs of the small-molecule single-
crystal crystallographic community. However, the importance of
efforts to broaden the applicability of CIF to suit the needs of
the macromolecular and the powder diffraction community were
recognized early.
Over the last few years, Syd Hall, Ian Langford, and I,
with the assistance of many people, have been working on
expanding the definitions in CIF for powder diffraction
applications. These efforts have been assisted by the members of
the Commission on Powder Diffraction (CPD) and the Commission on
CIF (COMCIFS). A draft of these additions has been completed in
the form of a supplementary dictionary for powder diffraction.
The draft dictionary is available to the crystallographic
community for review prior to its adoption as a standard (see
On-line Access to Information About CIF, below) Those
crystallographers who do not have access to electronic mail may
contact the Research and Development Officer of the IUCr, Brian
McMahon, for printed information.(see sidebar for address)
The assistance of the powder diffraction community is
needed to ensure that definitions used in the dictionary conform
to proper usage and are sufficiently comprehensive to include
the vast majority of powder diffraction applications. It should
be noted that powder diffraction CIF files will very likely use
definitions in both the original (core) CIF dictionary and the
powder CIF dictionary, so both dictionaries should be reviewed
together. Comments about the draft dictionary may be addressed
to the chair of COMCIFS, I. David Brown (IDBrown@mcmaster.ca) or
Brian McMahon. Questions may be addressed to the principal
author of the powder dictionary, Brian Toby
(tobybh@ttown.apci.com).
While the goal of the initial CIF effort was to define a
file format to contain information typically found in a single-
crystal structure manuscript (and supplementary materials), the
powder CIF effort had the additional goals of documenting and
archiving experimental data. It is intended that powder CIF be
used for communication of completed studies, but also for data
exchange between laboratories. This is frequently done at shared
diffraction facilities such as neutron and synchrotron sources.
The powder dictionary was written anticipating needs from
conventional x-ray diffractometers, as well as from synchrotron,
CW neutron, TOF neutron, and energy dispersive x-ray
instruments. Due to my lack of familiarity, a more cursory
attempt was made for film instruments. We have delayed
addressing requirements for image plates and area detector
instruments until issues relating to storing 2-D arrays in CIF
are resolved.
A powder CIF may contain descriptions of:
(1) the experimental configuration
(2) sample preparation, characterization and mounting details
(3) references to calibration information
(4) the experimental (raw) diffraction dataset
(5) the processed diffraction dataset
(6) a calculated diffractogram
(7) a peak table
(8) reflection assignments and intensities
(9) derived structural results
A powder CIF can also document diffraction data where multiple
phases are indexed or refined. It is also possible to document a
Rietveld refinement that incorporates multiple datasets.There is
no requirement that powder CIFs include all of the above
information. Indeed no powder CIF will ever utilize all of the
data items defined in the powder dictionary.
A Very Brief Introduction to CIF.
The Crystallographic Information File (CIF) is a mechanism
for storing crystallographic information in a computer-readable
ASCII file. The format is also intended to be human-readable and
editable. The syntax of CIF is a subset of the Self-defining
Text Archive and Retrieval (STAR) standard.
A CIF file consists of a series of data items (entries)
and corresponding values, e.g:
_cell_length_a 5.959(1)
A data item may also have a series of values associated with it
by preceding the data item with "loop_". The following example
shows a list of four symmetry operators associated with a single
data item:
loop_ _symmetry_equiv_pos_as_xyz
+x,+y,+z
1/2-x,-y,1/2+z
1/2+x,1/2-y,-z
-x,1/2+y,1/2-z
It is also possible to group a series of values together, in
this case the coordinates and atom label for a set of atoms:
loop_
_atom_site_label
_atom_site_fract_x
_atom_site_fract_y
_atom_site_fract_z
O1 .4154(4) .5699(1) .3026(3)
C2 .5630(5) .5087(2) .3246(1)
C3 .5350(5) .4920(2) .3997(1)
Related data items are grouped together in a block. The
beginning of a block is designated by the string "data_"
prefixing the name of the block. Data items can be recognized
because they always begin with an underscore (_) and values are
delimited by spaces, quotes or pairs of lines beginning with a
semicolon. Quotation marks are used to delimit a value that
contains spaces, but do not span lines:
_chemical_formula_sum 'C18 H25 N O3'
and semicolons are used for values that span more than one line:
_publ_author_address
; Research School of Chemistry
Australian National University
GPO Box 4, Canberra, A.C.T.
Australia 2601
;
Thus, CIF is largely free format. Two restrictions were
implemented to allow facile transmission of CIFs by e-mail:
lines may not exceed 80 characters and only printable ASCII
characters may be used.
Data items used in CIF are described in a dictionary,
which defines meaning and usage. For example, the dictionary
entry defining _cell_length_a specifies that the value will be a
number in angstroms and that an ESD is allowed. The dictionary
is itself a STAR file, where the dictionary syntax is defined in
a separate Dictionary Definition Language (DDL) file.
References:
CIF: S. Hall, F. Allen & I. D. Brown, Acta Cryst. A 47 (1991),
655-685.
STAR: S. Hall & N. Spadaccini, J. Chem. Inf. Comput. Sci. 34
(1994), 505-508.
On-line Access to Information About CIF
There are a number of methods for accessing CIF-related
information via Internet. Much of the information is independent
of the method used for access, but World Wide Web (WWW) and
gopher offer a more friendly user interface and more
introductory material.
WWW (using Mosaic, Lynx,...)
* Connect to URL http://www.iucr.ac.uk/welcome.html and
follow references to bring you to the CIF home page.
Gopher:
* Point your favorite rodent to gopher.iucr.ac.uk (port 70).
Look for the section titled "CIF Dictionary files." Of
greatest interest: the "draft powder" and "draft revised
core" dictionaries.
* Also see "General CIF-related" files for some software
(ciftbx, ciftex, quasar and cyclops).
Anonymous ftp:
* Connect to ftp.iucr.ac.uk. Look in directory cifdics.
Files of interest are: cifdic.c94 (core dictionary) and
cifdic.p94 (powder dictionary). Also note the PostScript
versions of these files (with added suffix .ps)
* Also, see directory pub for some software (ciftbx, ciftex,
quasar and cyclops).
E-mail:
* Send a message to SENDCIF@IUCR.AC.UK containing the single
word "help." Most of the files listed above are available
through SENDCIF.
If all else fails, contact:
Brian McMahon, Research and Development Officer, IUCr, 5
Abbey Square, Chester CH1 2HU, England. (tel: +44 244
342878, fax: +44 244 314888, or e-mail bm@iucr.ac.uk)
! Brian H. Toby, Air Products & Chemicals, Allentown, PA 18195-1501 (USA) !
! net: tobybh@ttown.apci.com voice: (610) 481-4198 FAX: (610) 481-6517 !
! Opinions expressed herein are mine and do not represent my employer. !
! Any resemblance between these opinions and fact is purely accidental. !