************************************************************************* ************************************************************************** *************** ******************* *************** Documentation for CRYSIZ ******************* *************** ******************* ************************************************************************** ************************************************************************** PROGRAM AUTHORS - C. R. Hubbard, B. Morosin & J. M. Stewart. ---------------------------------------------------------------------- CRYSIZ - A program for computing crystallite size and strain from the broadening of powder diffraction lines. This program uses the output generated by the program XRAYL. ----------------------------------------------------------------------- Please note that the executable of CRYSIZ can be found in the appropriate "/ccp14/operating-system/crysiz" directory on the CCP14 anonymous FTP site. Numerous example input files are included in the current directory. Please Note : this file is a condensed version of the full CRYSIZ documentation, which will become available, and which will replace this file, in April 1997. COMMAND LINES ------------- The CRYSIZ executables are invoked simply by typing "crysiz" on UNIX or PC machines, or by typing "run crysiz" on VMS. THEORETICAL BACKGROUND ---------------------- CRYSIZ uses the Warren-Averbach method to estimate domain size and root mean square strain. The value of rms strain and the average coherence length or crystallite size are obtained as a function of crystallographic direction. The effective crystallite size is derived from the slope of an analytical function, and the size coefficients are derived from the Stokes coefficients versus the column length. The results from this method are printed and plotted. USING THE PROGRAM CRYSIZ ------------------------ In order to use CRYSIZ, it is necessary to have two input files: one witht the reference data and one with the broadened data. In these files, each crystallographic line must be separated from all other nonsymmetrically overlapping lines. Exact superposition of lines must be treated as a single line as the lines cannot be separated. The program XRAYL is used to prepare these data files (see "xrayl.doc" in CCP14 directory "/ccp14/doc/xrayl"). The files will be "opened" if they have the designation .IDL, where is a name chosen by the user for the data sets; for example "NaClrf.IDL" for the reference pattern and "NaClb1.IDL" for broadened sample one. In the program, is restricted to no more than 11 characters (4 on a PC). Besides the data files, instructions for the run must be given interactively or, more conveniently, by preparation of an ".CIN" file. Whichever method is chosen, the options to be specified concern which methods are to be carried out, the plots generated, the extent of the printed details, and the reflections from the ".IDL" to be included in the analysis. In the interactive mode, the options are presented as a set of tedious questions to be answered in turn. The CRYSIZ program will open and process a number of prepared files and produce output files in the process. Each file type is defined by an "extension" mnemonic following a period. Before the peiod, an up-to-11-character code, here called the , is prefixed. The code should be mnemonic of the data set being treated. The extension codes are fixed by the program while the identities are chosen by the user. FILE TYPES ---------- ".CIN" ... Program input instructions file. Program may be run interactively, but the use of this file prepared by a local line editor is much more productive. ".IDL" ... Input files containing the separated, idealised input files ".IDM" ... Optional input files like ".IDL" but with least-squares parameters displaced by a specified number of standard deviations down from the best result. ".IDP" ... Optional input file like ".IDM" but with the parameters displaced up from the best result ".CPN" ... Output file containing summary of run suitable for printing ".CPL" ... Optional output file containing crude line printer plots for inspection ".CEX" ... Optional output file containing raw plot data for use with local plotting software .......................................................................... FORMAT OF ".CIN" FILE --------------------- This file, which specifies the calculations to be done and the results to be displayed, must consist of a series of "lines" with defining titles which are parsed in fields defined by blanks (spaces). These lines may be prepared by means of a line editor and then used repeatedly to drive the program. The first field of every line defines the function of the line in setting options. The six types of lines and the order in which they must be supplied in the file are : OPTIONS (required) QUANTITIES (optional) STOKES (optional, forces use of Fourier rather than numerical deconvolution) REFLEC (optional, if none supplied all reflections in file will be used) WARREN (optional) END (required) The "OPTIONS" line and the "END" line are mandatory. The other lines are optional, depending on the result desired. If they are not supplied, default actions will be taken as described below. The benefit of using a ".CIN" file is that it may be edited with the local line editor for use with other problems or samples. In CRYSIZ, these lines are parsed by explicit code so that the quotation marks (eg. in "REGIN_2_THETA" or "N") required in XRAYL for alphanumeric strings must not be used. FORMAT OF THE ".CIN" INPUT LINES -------------------------------- OPTIONS line - y/n implies do (y) or do not (n) FIELD CONTENTS PURPOSE OF FIELD USUAL CHOICE 1 OPTIONS 2 y/n Print out the Fourier coeffs. produced from n the input line profiles 3 y/n Print out the Stokes coeffs. of line profiles n 4 y/n Print out the A values of the line profiles n 5 y/n Calculate and print out the Warren-Averbach y size and strain values 6 y/n Plot the input line profiles n 7 y/n Print all intermediate results (dump!) n 8 y/n Plot the Stokes coeffs. of line profiles n 9 y/n Plot the unfolded line profile derived y from the Stokes coeffs. 10 y/n Determine the size and strain and produce the y plots for the Cauchy plotting method of Hall-Williamson, using the sum of coeffs. as a measure of broadening 11 y/n As in 10, but using integral breadth as the y measure of broadening 12 y/n As in 10, but using FWHM as the measure of y broadening 13 y/n Determine the size and strain and produce the y plots for the Gauss plotting method, using the sum of coeffs. as a measure of broadening 14 y/n As in 13, but using integral breadth y 15 y/n As in 13, but using FWHM y 16 y/n Determine the size and strain and produce the y plots for the Cauchy-Gauss plotting method, using the sum of coeffs. as a measure of broadening 17 y/n As in 16, but using integral breadth y 18 y/n As in 16, but using FWHM y 19 y/n Plot the effective crystallite size plot of y the Warren-Averbach method 20 y/n Plot the average strain as a function of column y length of the Warren-Averbach method 21 y/n Plot the Stokes Fourier coeffs. as a function y of the square of (1/d) for three l-values of the Warren-Averbach method QUANTITIES line - two successive blanks terminatescan of line --------------- FIELD CONTENTS PURPOSE OF FIELD USUAL CHOICE 1 QUINTITIES 2 Integer The number of points in the Stokes coeff. 100 tables. If zero, the dafault will be 120, the max. allowed. The max. column length treated will be the column length step size of field 7 times this integer 3 Integer The number of points at each end of the raw 10 profile data that are used to establish background intensity. If zero, the default is 20. No fewer than three points should be specified 4, Reals The three values of "l", in Angstroms, which 50., 5, are to be used in preparing the 100., 6 Warren-Averbach plots of Stokes Fourier 150. coeffs. vs. square of (1/d). If supplied as zero, the values 50.0, 100.0 and 150.0 will be used 7 Real The column length step size to be used in 10.0 the Warren-Avervbach method. The default is 10.0 Angstroms. STOKES line - ----------- Parameters for controlling the damping of Stokes coeffs. If used, there must be one STOKES line for each reflection in the ".IDL" files. The presence of a Stokes line switches the deconvolution procedure from direct to Fourier methods. FIELD CONTENTS PURPOSE OF FIELD USUAL CHOICE 1 STOKES 2 0/1/2 Zero directs a "log" decay, fit by least 0 squares to the Stokes coeffs. One directs the application of a linear decay to zero in the noise region. Two directs that no smoothing be done. 3 Real A weight control parameter for the Stokes 1.0 coeff. fitting process. If it is near 1.0, the fit will be to all regions of column length "l". If it is near 4.0, the fit will be weighted to short column lengths REFLEC line - ----------- Selects reflections from the ".IDL" data files. One reflection is specified per line. When these lines are omitted, all reflections will be included. FIELD CONTENTS PURPOSE OF FIELD 1 REFLEC 2 Integer The h index of the reflection 3 Integer The k index of the reflection 4 Integer The l index of the reflection WARREN line - ----------- Selects reflections to be associated in a Warren-Averbach analysis. One reflection is specified per line. Each associated group must be terminated with a "WARREN-99" line. When these lines are omitted, associations will be made of all reflections in the ".IDL" files which are orders of one another, eg. (1,1,1) with (2,2,2). The use of WARREN lines lines allows superseding of this choice. FIELD CONTENTS PURPOSE OF FIELD 1 WARREN 2 Integer The h index of the reflection or -99 to signal the end of the group 3 Integer The k index of the reflection 4 Integer The l index of the reflection END line - This line is not optional and MUST be the last line in the -------- ".CIN" file. ........................................................................ FORMAT OF THE ".DA" FILES -------------------------- The format of the input ".DA" files read by XRAYL is as shown below. These files contain, in FORTRAN formatted ASCII, the intensity measurements from regions of the 2theta "windows", in steps of 2theta taken from the diffractometer. RECORD FORMAT CONTENTS 1 A53 Title; up to the first 11 characters may be the used to identify the files of the compound 1 A19 Time/date as bhh:mm:ssbdd/MMM/yy where "b" represents a blank space 2 I4 h index of reflection 2 I4 k index of reflection 2 I4 l index of reflection 2 I4 Number of records of type 3, intensity data, NREC, which must be > 15 and < 1023 2 F10.2 Counting time per intensity point 2 F10.4 Beginning 2theta for intensity measurements 2 F10.4 Ending 2theta for intensity measurements 2 F10.4 2theta step size, lower limit 0.001 degrees 2 F10.6 Wavelength of K-alpha1 used for measurement 3 through F10.2 Intensity at measured point 2 + NREC ---------------------------------------------------------------------------