Iric.imet-db.ru

Journal of Research of the National Institute of Standards and Technology [J. Res. Natl. Inst. Stand. Technol. 101, 295 (1996)]
Siemens Software for Four-Circle and Susan K. Byram, Charles
Ideas for future access to this and other F. Campana, James Fait,
and Robert A. Sparks
Technology has been incorporated withSiemens single crystal software for data Key words: CCD diffractometer; crys-
four-circle diffractometers; NIST Crystal database is useful in the process of single the database is searched, are described.
Accepted: February 2, 1996
Introduction
As a commercial vendor of single crystal x-ray dif- chemicals, minerals and pharmaceuticals. In decreasing fractometers, Siemens1 both produces data which goes order of frequency, their fields are chemistry (inorganic, into crystallographic databases, and searches these data- organometallic, and organic), crystallography, mineral- bases during the process of single-crystal structure ogy, and materials science. In North America, their af- determination. This paper describes how NIST Crystal filiations are 75% academic, 15% industrial, and 10% Data is incorporated with the graphical user interfaces government. The primary application in searching NIST of the XSCANS software for single crystal diffractome- Crystal Data is compound identification. Specifically, it ters, and of the SMART software for two-dimensional is to compare the unit cell of a new single crystal spec- CCD diffractometers. Reasons for choosing this partic- imen being studied on the diffractometer to previously ular database and ideas for future access to this and studied unit cells in the database. It is important to note other crystallographic databases are described.
that this search can be done before spending the time tocollect the single-crystal data itself. Thus, the primary Primary Application of NIST
reason to search NIST Crystal Data is to maximize pro- Crystal Data Search
ductivity by not re-collecting data on a known com- Siemens small-molecule diffractometer users are also users of the NIST crystal data identification file. The How Often Known Structures are
diffractometer users are researchers studying new Redetermined
1 Certain commercial equipment, instruments or materials are identi- Even highly experienced researchers may find them- fied in this paper to foster understanding. Such identification does not selves redetermining a known compound as the number imply recommendation or endorsement by the National Institute ofStandards and Technology, nor does it imply that the materials or of new single crystal structures grows far beyond human equipment identified are necessarily the best available for the purpose.
ability to remember each one. There are 197 500 entries Journal of Research of the National Institute of Standards and Technology with lattice parameters in the 1994 Version of NIST the CD-ROM is accessed for detailed information on the crystal data, incorporating entries on all classes of crys- known unit-cell. This information includes the known talline materials such as inorganics, minerals, metals, unit cell parameters a , b , c , ␣ , ␤ , ␥ , and cell volume; intermetallics, organics, and organometallics. Cases of compound name and formula; literature reference; and crystal-structure redetermination of common starting some descriptive information. The information is dis- materials which were recrystallized, and redetermina- played and stored for future access.
tions of compounds previously reported in lesser known The original algorithms incorporate many ideas or foreign-language journals, have been reported to us derived from discussions with Alan Mighell, Vicky from time to time. To try to attach a quantitative figure Karen, and colleagues at NIST as far back as 1977 when for how often these redeterminations happen, the results NIST was the National Bureau of Standards [2,3,4]. The of the Pittsburgh Summer School in crystallography algorithms were updated after the 1992 American were surveyed from 1992 through 1995. This school Crystallographic Association meeting, where Rodgers provides a 10 day course of intensive lectures and hands- and LePage [5] discussed searching NIST Crystal Data on data collection for approximately 30 students each when large uncertainties are present. Cell-comparison summer. Each student is asked to bring a crystal of an techniques discussed by Andrews et al. [6] have also unknown compound on which to collect data and solve been considered. Algorithms are designed to ensure that the structure. Each year, one or two of these students has no known unit cells are missed in the search. The output unexpectedly found a match for the supposedly un- may sometimes present numerous candidates for a known specimen. Complete results are given in Table 1.
match, but this can be screened readily by the researcher If the NIST Crystal Data search is done prior to data and is not considered problematic since the search is collection, the student has time to study a different done only once per new crystal studied.
unknown compound. As more and more classes in-corporate complete structure determinations as projects, Four-Circle Serial Diffractometers
the NIST Crystal Data search will become more im- When the crystal data are to be collected on a serial four-circle diffractometer, the average time to proceedfrom mounting the unknown crystal on the diffractome- Table 1. Frequency of known compounds identified by a NIST Crys-
ter to determining the precise unit cell and Bravais lat- tal Data search at Pittsburgh Summer School in Crystallography tice is approximately 1 1/2 hours, as shown in Fig. 1.
This assumes no human intervention between initial screening for the suitability of the crystal and initiatingthe NIST Crystal Data search. This point, prior to data collection, is the best time to search the NIST Crystal Data to see if the compound on the diffractometer has been studied previously, since a further 6 h to 20 h typ-ically will be needed to collect the single-crystal dataand to solve the structure, as shown in Fig. 2. If the unit How the NIST Crystal Data
cell is very large or the crystal weakly diffracting, this Search is Implemented
time could extend to many days of data collection. Thusthe human and instrument time productivity is much The algorithms to search NIST Crystal Data have increased if this time is not repeated.
been written at Siemens, since it was necessary toembed them in the graphical user interfaces which Two-Dimensional CCD Diffractometer
control the serial four-circle or two-dimensional CCDdiffractometers. Both types of instruments are con- Siemens has recently introduced a new type of dif- trolled by a PC, with a CD-ROM attached which con- fractometer for collecting data on single crystals. This is tains the NIST Crystal Data Identification File obtained called SMART (Siemens Molecular Analysis Research through the International Center for Diffraction Data Tool) and incorporates a novel two-dimensional CCD [1]. An index file for very fast search is created locally detector capable of collecting many reflections simulta- on the PC hard disk by a Siemens utility program when neously [7]. With this new instrument, the average time the file is first installed and when it is updated annually.
to proceed from mounting the unknown crystal on the A simple point-and-click menu initiates the NIST Crys- CCD diffractometer to determining the precise unit cell tal Data search, utilizing the unit-cell information al- and Bravais lattice is only 12 min, as shown in Fig. 3.
ready determined by the diffractometer. When a match This short time leads us to believe that the SMART for the unit cell is found within user-chosen tolerances, system, coupled with a NIST Crystal Data search, could Journal of Research of the National Institute of Standards and Technology be used to identify unknown single crystals in much thesame way powder samples are identified using a powderdiffractometer today. The initial reason for searching theNIST Crystal Data, to avoid re-collecting data on aknown compound, is still important. From Fig. 4, we seethat a further 1 h to 8 h to collect the data and solve thestructure is required. Many researchers choose a CCDdiffractometer system for crystals which are very weaklydiffracting or very small, for example 10 ␮m on a side.
For such crystals data collection may take up to a day.
Examples of NIST Crystal Data Search
The search is initiated using precise unit-cell parame- ters, determined on the diffractometer immediatelyprior to the search, or typed into the search menu frompreviously determined data. Parameters used by the Fig. 2. Time to collect data on a four-circle diffractometer.
sarch are: unit-cell axial lengths a , b , c ; axial angles ␣ , ␤, ␥; tolerance for the match as a fraction; lattice center-ing type (such as P for primitive); whether to search theorganic or inorganic file or both; an output file name tostore results; and whether to display short or verboseoutput.
Typically no hits will be shown, meaning no match is found for the new unit cell. Sometimes a match indicatesan isomorphous compound with very similar unit cellbut different chemical elements. One example was amatch between the newly collected C44H54Br4Zn4N2Se6with a primitive cubic unit cell 17.79, 17.79,17.79, 90.00, 90.00, 90.00 and the data base compoundC44H54Br4Cd4N2S6 with unit cell parameters 17.869,17.869, 17.869, 90.00, 90.00, 90.00, published earlierby Dean et al. at the University of Western Ontario [8].
Another example was a badly split crystal of a vanadiumcompound. In spite of the poor crystal quality, sufficientreflections were found to determine the unit cell andsearch NIST Crystal Data to match vanadyl hydrogen Fig. 1. Four-circle diffractometer search time.
Journal of Research of the National Institute of Standards and Technology Fig. 3. CCD diffractometer search time.
Fig. 4. Time to collect data on a CCD diffractometer.
Students at the Pittsburgh Summer School in crystal- integrated with Siemens’ diffractometer control soft- lography are encouraged to search NIST Crystal Data ware. The volume of new small-molecule crystal struc- prior to collecting data. An example for adenine hy- tures created a real need to maximize productivity of drochloride hemihydrate, brought by N. Sparks [11] as both the researcher and the instrument. NIST Crystal a known compound, is shown in Fig. 5. Three hits in Data database contains the unit-cell information needed the organic file showed that the structure had been prior to data collection. However, it was not until NIST published previously three times, first in 1948.
made the database available on a simple medium, aCD-ROM in PC format, that it became feasible to access Data Base Accessibility to
it from vendor software. It was also important that reg- Vendors and Users
ular updates (annually) were easy to obtain throughICDD [1], and that the reasonable license fee to indus- The NIST Crystal Data is the only crystallographic trial as well as academic users made it commercially database for which the search algorithms have been Journal of Research of the National Institute of Standards and Technology Fig. 5. CDF search results on adenine hydrochloride hemihydrate.
Enhanced Accessibility of Crystallographic
Longer term recommendations can be more sweep- Databases
ing. Even now, the majority of single-crystal structuresare not accessible in the databases because they have not For crystallographic databases in general, some yet been published. The barriers to publication include short-term recommendations can be made for today’s lack of time to prepare the publication; lack of desire to computing environment. At present Siemens small- publish because the structure is not ‘‘good enough’’ but molecule users prefer access on PC computers, followed the chemistry is adequately determined; or the com- by Silicon Graphics work stations. Siemens protein- pound is proprietary. To address this, worldwide elec- crystallography users prefer Silicon Graphics work- tronic deposition of unpublished data, with validation stations. CD-ROM as a media for local database storage safeguards, must be strongly encouraged. Easy ways to is universal, simple, and inexpensive. While access over create and search local databases of proprietary com- the Internet is attractive and growing rapidly, not every pounds could be provided. With the advent of two-di- laboratory is attached. The cost of license fees, espe- mensional detectors, the number of new single-crystal cially if access to several databases is needed, can be a structures will grow exponentially. In the field of protein barrier, particularly for industrial users. Database ven- crystallography, this has already happened with the in- dors might project maximum revenue using different troduction of area detectors over a decade ago. In small scenarios ranging from low market price for a large molecule crystallography, we are just beginning the tran- number of users, to high market price for a smaller sition. From Fig. 2 we can see that if a serial diffrac- number of users. Market awareness of some of the data- tometer takes from a few hours to a few days to collect bases should be enhanced, as the biggest barrier to using data, then a sustained productivity of up to 150 new them is potential users lack of knowledge of what they structures a year is possible for one such instrument.
can be used for, or indeed of their very existence.
With the new CCD diffractometer, from Fig. 4 we can Journal of Research of the National Institute of Standards and Technology see that this number could increase to as many as two tothree structures per day if suitable crystals are available.
Database providers will need to discover innovativeways to find and validate this data, including that inforeign-language and lesser known publications, and toupdate the databases ever more rapidly.
Acknowledgments
Alan Mighell, and colleagues at NIST for provocativediscussions over many years, and John Chambers ofSiemens for implementation of the CDF search withinthe SMART CCD diffractometer software.
References
[1] NIST Crystal Data is a product of the NIST Crystal and Electron Diffraction Data Center, National Institute for Standards andTechnology, Gaithersburg, MD 20899.
[2] A. D. Mighell, The Reduced Cell: Its Use in The Identification of Crystalline Materials, J. Appl. Cryst. 9, 491-498 (1976).
[3] A. D. Mighell and J. R. Rodgers, Lattice Symmetry Determina- tion, Acta Cryst. A36, 321-326 (1980).
[4] A. D. Mighell and V. L. Karen, Compound Identification and Characterization using Lattice-Formula Matching Techniques,
Acta Cryst. A42, 101-105 (1986).
[5] J. R. Rodgers and Y. LePage, Robust and Discriminating Search Parameters for Lattice Retrieval from Crystallographic Data-bases in the Presence of Large Experimental Uncertainties,American Crystallographic Association, PA 106 (1992).
[6] L. C. Andrews, H. J. Bernstein, and G. A. Pelletier, A Perturba- tion Stable Cell Comparison Technique, Acta Cryst. A36, 248-
252 (1980).
[7] H.D. He, J. L. Chambers, J. Fait, C. F. Campana, and R. A.
Sparks, A New Small Molecule Single Crystal DiffractometerUsing [8] P. A . W. Dean, J. J. Vittal, and N. C. Payne, Inorg. Chem. 26,
[9] C. Torardi and J. Calabrese, E. I. DuPont de Nemours and Com- pany, Wilmington, DE, private communication (1987).
[10] J. Johnson, D. Johnston, A. Jacobson, and J. Brody, J. Am.
Ceram. Soc. 106, 8123 (1984).
[11] N. Sparks, student, Pittsburgh Summer School in Crystallogra- About the authors: Susan K. Byram is product
manager for crystallographic systems, Charles F.
Campana is a senior applications scientist, James Fait
is a senior software engineer, and Robert A. Sparks is a
consultant and founder of the original analytical x-ray
group, all of the Analytical Instrumentation group of
Siemens Energy and Automation.

Source: http://iric.imet-db.ru/PDF/51.pdf