Introduction
Step 1 - Download data
Step 2
Step 3
Conclusion
References
Results
Search-Match Round Robin - 2002

Organized by J.-M. Le Meins, L.M.D. Cranswick & A. Le Bail
May-June 2002


SMRR closed (June 30, 2002)
Results available

Introduction

Phase identification from a powder diffraction pattern is a widely used method. Modern computer search-match programs coupled with the ICDD Powder Diffraction File (PDF) database (or with custom databases, Pauling, etc) are now probably more than 20 [1-4], and use various algorithms. However, their respective efficiency is unknown : "To date, no search/match program has proved to be completely satisfactory in all cases..." [2]. Promises of better approaches by using the full trace of known powder patterns were also made : "Nevertheless, the use of PDF-3 for the identification of unknown samples may well become a standard technique at some time in the future "[2]. This future is still awaited. Wasting time because a compound was believed wrongly to be new since a search-match was negative is something that should not happen nowadays. Knowing more about this software category is imperative. This Round Robin is completely open to academic researchers, manufacturers, developers (given that developers know their program better than any occasional user, in principle), etc. Moreover, the SMRR appears timely, because the PDF has made a recent breakthrough by including calculated patterns from the Cambridge Structural Database atomic coordinates, increasing enormously the ability to identify organic and organometallic compounds by search-match procedures. This should attract more search-match users.

Large advertisement will be made (at the sci.techniques.xtallography Newsgroup, at the Rietveld and SDPD mailing lists, at ACA 2002, EPDIC 8, etc)  that the SMRR starts early May, 2002 (when this page is announced is when it starts), and that people will have to propose identifications for 4 powder patterns. 

These 4 patterns are not especially hard to identify. They correspond to typical real cases. And the PDF contains solutions or at least close solutions to these problems. Minimal information is given about the samples since many identifications by search-match methods are frequently made without prior knowledge or having only an idea of the chemical content. Data are available as .dat ASCII files as defined in PowBase, and in some other standard and proprietary formats. 

Collection of results will occur in two steps : 
Step 1, deadline June 15, 2002 - samples without chemistry 
Step 2, deadline June 30, 2002 - samples with chemistry information given June 16. 

We estimate that a simple search-match is routinely performed in 5-10 minutes, and a more complex one in 15-30 minutes. So that the total job would take 2 hours of your precious time. This is the price to pay for knowing really the value of your own software.

We are expecting to obtain one screen copy (in .gif or .jpg or .png format, with a 1024x768 or 1280x1024 or higher resolution) for each of the search-match results showing all details (it should be seen : the program name, the filename of the powder pattern, the conditions for the search, if possible - complete or reduced dataset, etc - the beginning of the list of solutions proposed by the software, and a fit of the selected proposition(s) superposed to the powder pattern). Examples of such screen copies are downloadable here (1, 2) using the easy to identify pattern N°66 in PowBase (KAlF4), or more complex multiphase cases (3, 4, 5). However, if you did not get that result by only applying your search-match software, then you may explain how you get it (using the Hanawalt search manual, indexing and then finding the compound into the ICSD or CSD databases, etc). Nevertheless, if  your search-match software cannot perform better than Hanawalt method, then why did you buy it, and why is it so expensive ?

Step 1 - Download data

Data in various formats are gathered in files compressed by Winzip :

Sample 1 - smrr-1.zip
About sample 1, the geologist said : please, can you identify the phases in that powder ? Probably this is not single phase.
Bragg-Brentano geometry - Cu-Kalpha. See it.

Sample 2 - smrr-2.zip
About sample 2, the chemist said : please, can you tell me if this powder from my last synthesis contains phases listed in the PDF ? In fact, it could be single phase.
Transmission geometry - Capillary - Cu-Kalpha-1. See it.

Sample 3 - smrr-3.zip
About sample 3, the pharmaceutical engineer said : please, would you be able to identify the phases in that powder from a search in the PDF ?
Bragg-Brentano geometry - Cu-Kalpha - Sample dusted through a sieve for avoiding preferred orientation. See it.

Sample 4 - smrr-4.zip
About sample 4, the laboratory director said : I have received an industrial processing plant sample, please could you realize a powder pattern and then a search-match on it ? 
Bragg-Brentano geometry - Cu-Kalpha. See it.

All samples in one file : smrr.zip

Deadline for receiving results : June 15, 2002.

Results have to be sent (as soon as you wish) to J.-M. Le Meins : 
JM.Le_Meins@univ-mulhouse.fr preferably gathering the files in one 
having your name and compressed by WinZip (name.zip).  Summary results
of phases identified and software/method/database used should be
put in the ASCII text of the Email.

Contact L.M.D. Cranswick : l.m.d.cranswick@dl.ac.uk  if you have any
problems with the data and formats.

Step 2

Additional information about the sample chemistry, and more, will appear on this site on June 16, 2002. The search-match results will be collected in the same way as for step 1 (screen copies, etc). And the SMRR will be completely closed June 30, 2002.

Additional informations about the sample chemistry are given below :

Sample 1 - Contains : Ca, Fe, Al, P, Si, O, C and H, other traces possible, specimen from Rapid Creek, Yukon, Canada.

Sample 2 - Contains : Si, O, maybe F,  plus an organic molecule.

Sample 3 - Molecular formula : C13H10N2O4, possibly several polymorphs.

Sample 4 - Contains : Pb, S, O, trace C, and H. The customer says it is a cured plate from a lead acid battery plant - created by mixing lead oxide with sulphuric acid, pasted on a lead grid and cured at high humidity somewhere between 50 to 90 degrees Celcius.

Results have to be sent (as soon as you wish) to J.-M. Le Meins : 
JM.Le_Meins@univ-mulhouse.fr preferably gathering the files in one 
having your name and compressed by WinZip (name.zip).  Summary results
of phases identified and software/method/database used should be
put in the ASCII text of the Email. If it is your second participation to the SMRR,  please do clearly indicate in the subject field of your email the participation number which was given to you at step 1.

Contact L.M.D. Cranswick : l.m.d.cranswick@dl.ac.uk  if you have any
problems with the data and formats.

Step 3

Official solutions and participant solutions will then be made available at this web site in July 2002. Anonymity will be ensured, but it can be broken on demand from participants. Post-deadline comments (and complaints) from participants are welcome, and will be added to the SMRR stuff.

Conclusion

Should you continue to have confidence in your search-match software ? Individuals will conclude by themselves after looking at the SMRR results.

If developers have the good idea to adopt a common format for the entry file of their search-match software, that would already be a good result of this Round Robin (CIF or ASCII .dat...).

References

[1] D.K. Smith & S. Gorter, "Powder Diffraction Program Information 
     1990 Program List," J. Appl. Cryst. 24 (1992) 369-402.
     and, from the same authors, see the last release of the "World Directory
     of Powder Diffraction Programs."
[2] J.I. Langford & D. Louër, "Powder Diffraction," Rep. Prog. Phys. 59
     (1996) 131-234.
[3]  L.M.D. Cranswick, Available Search-Match Software
      L.M.D. Cranswick, Hints on Phase Identification Using Powder 
      X-ray Diffraction.
[4] A. Le Bail,  SDPD Tutorial.

Some publications about Search-Match :

- Oxford Cryosystems, (1999). Crystallographica Search-Match. J. Appl. Cryst. 32, 379-380. 
- Al-Farhan K.A. (1999). FARHAN - a qualitative and quantitative PC program for X-ray powder diffraction. Powder Diffraction 14, 16-21.
- Jenkins R. (1994). Adv. X-ray Anal. 37, 117-121.
- Nusinovici J. and Winter M.J. (1994). Adv. X-ray Anal. 37, 59-66. 
- Smith D.K., Hoyle S.Q., Johnson Jr. G.G. (1994). Adv. X-ray Anal. 37,   67-77.
- Smith D.K., Hoyle S.Q., Johnson Jr. G.G. (1993). Adv. X-ray Anal. 36, 287-299.
- Nusinovici J. and Bertelmann D. (1993). Adv. X-ray Anal. 36, 327-332.
- Lin S.F., Lu, X.J., Zheng W.F., Zhang J.B. (1993). A novel search match system for X-ray powder diffraction data. Chemometrics and Intelligent Laboratory Systems 20, 85-91.
- Wright D., Liu C.L., Stanley D., Chen H.C. Fang J.H. (1993). XRAYS - A fuzzy expert-system for qualitative XRD analysis. Computers & Geosciences 19, 1429-1443.
- Toby B.H., harlow R.L. and Holomany M.A. (1990). Powder Diffr. 5, 2-7.
- Caussin P., Nusinovici J. and Beard D.W. (1989). Adv. X-ray Anal. 32, 531-538.
- Caussin P., Nusinovici J. and Beard D.W. (1988). Adv. X-ray Anal. 31, 423-430.
- Marquart R.G. (1988). µPDSM (micro powder diffraction search match). Abstracts of papers of the American Chemical Society 195, 30.
- O'Connor, B. H. & Chang, W.-J. (1984). A procedure for search/match analysis of X-ray powder diffraction patterns using compact data bases. J. Appl. Cryst. 17, 212-214. 
- Goehner R.P. and Garbauskas M.F. (1984). X-ray Spectr. 13, 172-179. 
- Goehner R.P. and Garbauskas M.F. (1983). Adv. X-ray Anal. 26, 82-86.
- Cherukuri C. and Snyder R.L. (1983). Adv. X-ray Anal. 26, 99-104.
- Lin T.H., Zhang S.Z., Chen L.J, Cai, X.X. (1983). An improved program for searching and matching of X-ray-powder diffraction patterns. J. Appl. Cryst. 16, 150-154.
- Toyohisa S., Fujiwara I., UI T., Asada E. (1983). Search march implications of the frequency-distribution of d values in the JCPDS powder data file. Advances in X-ray Analysis 26, 89-92.
- Hummel W.; Gottschalk M. (1983). Fast powder diffraction search program by microcomputer. Fortschritte der Mineralogie 61, 99.
- Schreiner, W. N., Surdukowski, C. & Jenkins, R. (1982). A new minicomputer search/match/identify program for qualitative phase analysis with the powder diffractometer. J. Appl. Cryst. 15, 513-523. 
- Frevel L.K. (1982). Structureèsensitive search match procedure for powder diffraction. Analytical Chem. 54, 691-697.
- Snyder R.L. (1981). Adv. X-ray Anal. 24, 83-90.
- Edmonds, J. W. (1980). Generalization of the ZRD-SEARCH-MATCH program for powder diffraction analysis. J. Appl. Cryst. 13, 191-192. 
- Marquart, R. G., Katsnelson, I., Milne, G. W. A., Heller, S. R., Johnson Jnr, G. G. & Jenkins, R. (1979). A search-match system for X-ray powder diffraction data. J. Appl. Cryst. 12, 629-634. 
- Jenkins R. (1979). Qualitative-analysis with the JCPDS powder data file. Abstracts of papers of the American Chemical Society, Iss Sep, 34-.
- Frevel, L. K., Adams, C. E. & Ruhberg, L. R. (1976). A fast search-match program for powder diffraction analysis. J. Appl. Cryst. 9, 199-204.
- Johnson Jr. G.G. and Vand V. (1967). Indust. Eng. Chem. 59, 19-31. 
- Nichols M.C. (1966). A Fortran II program for the identification of X-ray powder diffraction patterns UCRL-70078, Lawrence Livermore Laboratory.
- Frevel L.K. (1965). Anal. chem. 37, 471-482.
- Hanawalt J.D., Rinn H.W., Frevel L.K. (1938). Indust. Chem. Anal. Ed. 10, 457-512.
 




J.-M. Le Meins : JM.Le_Meins@univ-mulhouse.fr
L.M.D. Cranswick : l.m.d.cranswick@dl.ac.uk
A. Le Bail : alb@cristal.org

May-June 2002 - Sponsored by the SDPD Internet Course