12 December 2005

 

Tabulated variation of Alpha/Beta decay parameters in TOF

 

New Features

 

·         A table with the values of the decay parameters Alpha and Beta as a function of d-spacing can now be given in the Instrumental Resolution Function file for TOF when Res=5 using the profile function Npr=9. The program calculates the      value of Alpha and Beta for each reflection by linearly interpolating the list of given values. A maximum of 100 points is allowed. The keyword for introducing the list is: LISTALFBET followed by the number of points (NPo). After this line the program reads NPo lines containing three real values: the d-spacing Alpha and Beta. This procedure allows the treatment of moderators that cannot be modelled easily using a back-to-back exponential, convoluted with a pseudo-Voigt function, having a dependence of their decay parameters of the form:

 

   Alpha=Alpha0+Alpha1/dsp     and       Beta=Beta0+Beta1/dsp^4

 

An example is given below:

 

............. omitted lines

!       Gam-2     Gam-1     Gam-0

GAMMA   0.000     4.937     0.000

!          alpha      beta

LISTALFBET   24

 0.4277  0.580207  0.935069

 0.4508  0.550432  0.762171

 0.4709  0.526999  0.644441

 0.4878  0.508716  0.562862

 0.5023  0.494028  0.503393

 0.5206  0.476688  0.439694

 0.5437  0.456394  0.373471

 0.5671  0.437566  0.319440

 0.5903  0.420400  0.275896

 0.6206  0.399880  0.230396

 0.6609  0.375457  0.184648

 0.7043  0.352329  0.148817

 0.7809  0.317792  0.106979

 0.8554  0.290103  0.081954

 0.9017  0.275216  0.071148

 1.0412  0.238344  0.050993

 1.1044  0.224713  0.045555

 1.2098  0.205134  0.039299

 1.3526  0.183477  0.034180

 1.5618  0.158896  0.030199

 1.6313  0.152131  0.029381

 1.9128  0.129738  0.027355

 2.7052  0.091739  0.025649

 3.1237  0.079448  0.025400

END

 

Shift to these values can always be fitted provided the shifts verify the dependency versus d-spacing described above for the function Npr=9. The parameters that can be refined are D_Alpha0,D_Beta0,D_Alpha1,D_Beta1, where the prefix "D_" stands for the shifts.

 

 

6 December 2005

 

Change of the name of integrated intensity cluster files

 

·         The name of the file containing generated integrated intensity clusters after a run of FullProf in which More=1 and Jvi=11 (useful for simulated annealing) has been changed to avoid incompatibility with WinPLOTR. When WinPLOTR runs FullProf, after selecting a PCR (or NEW) file, looks for input data files (*.dat, *.hkl, *.int ...) if it finds a file with the extension "int" having the same code (with a numeric suffix) as the PCR file will send to FullProf a wrong input data file. The name is now of the form: "codefile"n_cltr.int. Where "codefile" stands for the name of the PCR file and "n" is the number of the phase. The suffix "_cltr" (cluster) indicates that the file has been generated by FullProf. It can be changed (if wished) for using the file as input for Simulated Annealing.

 

 

25 November 2005

 

Correction of bugs in magnetic phases in EdPCR Program

 

Bug

 

 

·         Long standing bugs in the edition of magnetic phases have been corrected in EdPCR. In particular the writing in bad fields (anisotropic thermal parameters) of magnetic parameters. The edition of symmetry operators and basis functions of irreducible representations of the propagation vector group has also been corrected.

The bug consisting in the modification of the number of propagation vectors when passing through the constraints windows has also been fixed. There are still other bugs that will be corrected progressively. Workarounds are always possible by editing the PCR file using the text editor of EdPCR.

 

Note: In some cases EdPCR quits suddenly (giving or not an error message) after clicking in a button. This normally occurs on XP when EdPCR has been invoked from WinPLOTR. We do not know what is the cause of this extrange behaviour. If the user experiments this problem, a workaroung is to invoke EdPCR by clicking on it, from a link on the desktop or from the command line. Please inform me in what circumstances this behaviour occurs

 

 

20 November 2005

 

Improvements in running FP_CYC from WinPLOTR

 

New Features

 

·         Several changes and additions have been performed for running FullProf in sequential mode. Additional information, in particular for plotting the behaviour of individual reflections coming from Le Bail fits of many files treated in sequential mode, can be found in the news of WinPLOTR.

 

1.    Concerning "fp2k", the argument giving the code for the data (datcode) files (can be now extended with an extension (ext) that may be different of ".dat". For instance the third argument can be of the form "datcode.xy", where ext=".xy" is the extension of the data files. The extension can also be given in the dialog box of WinPLOTR.

 

2.    Two kind of formats for the names of the data files are now accepted. In the first one the names of the data files are of the form: "datcode<n>ext", where "<n>" is an integer number and "ext" is the extension of the data files (by default ".dat"). The length of the names is variable. The second format for the names has a fixed length and it is of the form: "datcode<000xxx>ext", where <000xxx> is an integer with zeroes on the left         to complete a fixed format. This format can have a maximum of 9 digits, the program determines the format automatically from an analysis of the existing first file.

 

For instance, invoking FullProf from the command line (or from WinPLOTR) as:

 

       > fp2k  cyc  my_pcr  datcode.uxd  23  268  y n y

 

The program will select ".uxd" as extension of the data files. Moreover the program will determine if the file "datcode23.uxd" exist or not. In the case the file is found, the program assumes that the length of the names is not       fixed (first format) and will treat the files from "datcode23.uxd" up to "datcode268.uxd" even if there is some lacking files in between. In the case the file is not found the program will try to find files with names of the form "datcode00...23.uxd", once a file, for instance "datcode0023.uxd", is found the program assumes that all the other files have the same length, so the final file must be "datcode0268.uxd". The from the last sequence "y n y", the program will save the ".prf" and ".mic" and will discard ".hkl" files.

 

8 November 2005

 

Basis vectors of IrReps and Fp_Studio

 

New Features

 

·         The current version of FullProf has been updated to:

 

      **********************************************************

      ** PROGRAM FullProf.2k (Version 3.40 - Nov2005-LLB JRC) **

      **********************************************************

 

·         In the current version of FullProf the FST file generated when the magnetic structure is described using basis vectors of the irreducible representations of the propagation vector group is now adequate for visualising the magnetic structure. FullProf converts the basis functions provided by BasIreps to Fourier coefficients readable by Fp_Studio. This is done even in the case where the MSYM-like representation of the symmetry is not appropriate.

 

·         A convergence condition has been implemented for Le Bail fits. Until now no convergence condition was applied as soon as a phase was treated using abs(Jbt)=2, so that the program runs for the prescribed number of cycles even if no improvement was produced. At present,  the convergence condition is that the programs stops doing cycles if the difference between successive Chi2 values is less than 1.0E-3.

 

21 October 2005

 

Bug in magnetic structure factor when using Isy=-2

 

Bug

 

·         A bug has been corrected when using magnetic incommensurate structures calculations and basis functions coming from symmetry analysis with BasIreps. The bug was only for the case of using Isy=-2: numeric form of complex basis vectors. Internally the calculation of the Fourier coefficients was done using the complex conjugate of the basis vectors instead of the basis vectors themselves. This has been corrected in the current version of FullProf. This error does not affect the case of Isy=1,-1: direct use of Fourier coefficients and MSYM operators deduced from the same basis functions (alphanumeric symbols describing Fourier coefficients in BasIreps). Of course for Isy=-2 but with no complex components the calculation was correct in the previous version.

 

·         The FST file for FullProf Studio when JBT=5 was wrong. A phase factor was with the wrong sign in generating the Fourier coefficients. This has been corrected in the current version of FullProf.

 

Warning

Remember that, when a magnetic structure is described in two different phases (for instance if there are two propagation vectors), the user must edit the FST file and merge the two magnetic phases in a single one. Otherwise the propagation vector of the first phase is attributed wrongly to the second phase.

 

 

14 June 2005

 

Manual for flipping ratio refinements

 

New Features

 

·         The current version of FullProf has been updated to:

 

      **********************************************************

      ** PROGRAM FullProf.2k (Version 3.30 - Jun2005-LLB JRC) **

      **********************************************************

 

·         Some improvements for the output of *.fst files have been introduced. The output was not complete when using numeric form for symmetry operators and pure magnetig phase in single crystal refinements.

·         A manual describing the use of FullProf for flipping ratio refinements has been included in the FullProf Suite. The document is called Flipping_Ratios_in_FullProf.pdf and it is included in the "docs" subdirectory of the FullProf_Suite directory.

 

 

31 May 2005

 

Correction of a recent bug in generated sub-profiles

 

Bug

 

·         In the current version of FullProf there was an error in generating sub-profiles. This does not affected any important calculation, only the output of the profiles corresponding to the different phases (for plotting purposes) was wrong. This was due to an internal re-arrangement. Now the *.sub files are correct. Sorry!

·         The output CIF file, containing the profile and structure factors for powder patterns, has also been improved and corrected.

 

 

9 May 2005

 

Correction of a bug in generation of atom multiplicities

 

Bug

 

·         In the current version of FullProf there was an error in calculating multiplicities of atom sites in some particular circumstances. This affected only quantitative analyses if ATZ was generated automatically. This error was introduced in the latest version due to an internal change in the libraries: the setting up of an "epsilon" value was not properly done due to the move of a particular subroutine to another module. The bug has been fixed. Sorry!

 

25 April 2005

 

Correction of a bug in satellite option of rigid bodies

 

Bug

 

·         In the current version of FullProf there was an error, appearing only in some ircumstances, for the satellite groups in rigid bodies. ometimes the second reference atom was not properly detected. This was due to a change from the function nearest integer (NINT) to integer (INT).This has been corrected now coming back to the correct expression.

 

 

10 March 2005

 

New version of Bond_Str

 

New Features

 

·         The current version of FullProf has been updated with a new internal version of the program Bond_Str. The stand-alone program Bond_Str has also been updated and a new GUI is also available. In the previous version the calculation of bond-valence sums were performed using the worse bond-valence parameter of the file "bvpar.cif". This was an error of mine! When there were several (d0,B0) sets corresponding to the same cation-anion species I selected the last set ... just the worse! as written in the header of the "bvpar.cif" file. This has been corrected in the current new version.

 

·         A small program to prepare Rigid body objects for FullProf is now provided within the FullProf suite. See note of 12 December 2004. The program, called Mol_tPCR, can be run in a console as follows:

 

                     d:\my_prompt> mol_tpcr  myfile

 

where "myfile.cfl" is a previously existing file in wich the molecule(s) to be read is(are) written. The main use of this program is to prepare internal coordinates of a rigid group when one knows a crystal structure where this group is present. The internal coordinates are independent of the crystal structure and can be ported to a new unit cell for purposes of solving a new crystal structure by positioning known groups.

 

An example of such a file, it can be used for testing purposes, is given below

 

------------------  start File: urea.cfl -------------------------------------------

!

! Example of molecule given in a crystal

!

!

CELL     5.584001   5.584001   4.688990  90.0  90.0  90.0

!

MOLEX   8   Urea      F

!    Xc         Yc       Xc         Phi     theta    Chi     TypeAngles   TypeThermal

   0.00000  0.50000  0.32811         0.0     0.0     0.0         P          ISO

     0          0         0          0       0       0           !Refinement codes

!Atm Spe        x        y        z      N1  N2  N3    Biso      Occ

Ur1  C       0.00000  0.50000  0.32811    0   0   0   0.50000   1.00000

Ur2  O       0.00000  0.50000  0.59574    0   0   0   0.50000   1.00000

Ur3  N       0.14466  0.64466  0.17816    0   0   0   0.50000   1.00000

Ur4  N      -0.14466  0.35534  0.17816    0   0   0   0.50000   1.00000

Ur5  H       0.14261  0.64261 -0.03550    0   0   0   0.50000   1.00000

Ur6  H       0.25589  0.75589  0.28398    0   0   0   0.50000   1.00000

Ur7  H      -0.14261  0.35739 -0.03550    0   0   0   0.50000   1.00000

Ur8  H      -0.25589  0.24411  0.28398    0   0   0   0.50000   1.00000

!             x(P1)    y(P1)    z(P1)     x(P2)    y(P2)    z(P2)    x(P3)     y(P3)    z(P3)

XYZ_Frame   0.00000  0.50000  0.59574   1.00000  0.50000  0.32811   0.00000  0.50000  0.32811

!

------------------  End File: urea.mol -------------------------------------------

 

The meaning of the keyword CELL and other parameters are obvious from the given names.

 

The keyword MOLEX must be followed by the number of atoms in the molecule, the name of the molecule and the kind of coordinates provided below (F: for fractional coordinates with respect the the given cell, C: cartesian, Z: for Z-matrix or S: for spherical coordinates )

 

The parameters Xc, Yc, Zc are the fractional coordinates of the origin of the internal Cartesian system with respect to the given unit cell.

 

The angles Phi,Theta and Chi are Euler angles (two types, P: Phi and Theta are the polar spherical angles of the Z-axis of the Cartesian internal frame with respect to the Cartesian frame attached to the unit Cell, E: conventional Euler angles).

 

The keyword "XYZ_Frame" is followed by the fractional coordinates of three points in the unit cell. These three points will serve to position the molecule in the crystal and to calculate the centre and Euler angles. The origin of the Cartesian internal system will be put in the third given point. The z-axis is parallel to the vector P1-P3, the x-axis lies within the plane (P1-P3,P2-P3), and the y-axis completes the direct frame.

 

The numbers N1,N2,N3 are integers defining the connectivity of the molecule as it is conventional in the Z-matrix formulation. If all of them are put to zero, the program will generate automatically a connectivity table taking the first atom as the origin, the second along "x" and the third in the x-y plane. This internal Cartesian frame will be put

 

 

27 February 2005

 

New version of FullProf. User given Form-factors and ADPs

 

New Features

 

·         The current version of FullProf has been updated to:

 

      **********************************************************

      ** PROGRAM FullProf.2k (Version 3.20 - Feb2005-LLB JRC) **

      **********************************************************

 

·         The note of 12 December 2004 concerning rigid bodies has been updated.

 

·         A new option for atomic types has been added. Putting n_t=5 for an atom or object (SASH,...), it is supposed that the form factor is like that of n_t=4 but additional ADP are read after the normal reading for n_t=4. This allows the use of anisotropic temperature factors in calculating flipping ratios or structure factors of complex object described by different kinds of form factors. Two additional lines (one with beta11, beta22, ... and the other with the corresponding codes) are read for an atom after the reading of form factor coefficients.

 

·         The calculation of errors in microstructural parameters are now propagated in order to give the estimated errors of size and strain values along different crystallographic directions. The sigmas appear in files *.siz and *.str that can be obtained by putting JVI=5.

 

Bug

 

·         A bug concerning the absorption correction of powder patterns treated in integrated intensity mode has been corrected. The absorption correction when IRF=4 and IPOW=2 was not applied. This has as a consequence that, in cases of non-negligible absorption, the magnetic moments obtained when using simulated annealing were lower than those obtained when using the Rietveld method. Now the INT file generated after a Le Bail fit (Jbt=2 and Jvi=11) includes the muR coefficient for absorption correction.

 

 

 

14 February 2005

 

A new section in the PCR file for putting commands

 

New Features

 

·         A bug concerning restraints has been corrected. If there was no distance restraints, angle restraints were not applied. Now angle restraints are applied even if no distance restraint is given.

 

·         New commands to facilitate the refinement strategy are available. Using the automatic mode (Aut=1) and the COMMANDS instruction explained in the note of 15 January, the new extensions of VARY and FIX are as follows. Within an instruction VARY of an instruction FIX individual parameters of selected atoms are allowed.

The available structural parameters are "x_", "y_", "z_", "xyz_" and "b_". The names of the selected atoms must be identical to those given in the list of the asymmetric unit (The instructions are case sensitive). As usual the instructions are applied as they appear. For instance a FIX instruction after a VARY may suppress a refinement code.      For instance the list of commands, appearing just after the phase name, given below

 

         COMMANDS

           VARY  xyz_Fe  xyz_O1

           VARY  x_Mn1  z_O27  y_Ho1 b_Al

           FIX   b_Mn1 x_C23

         END COMMANDS

 

tells to the program that it should refine all positional parameters of atoms Fe and O1, the "x" coordinate of Mn1, the "z" coordinate of O27, the "y" coordinate of Ho1 and the temperature factor(s) of atom Al. If Al is anisotropic all beta-parameters compatible with the site symmetry are refined. The instruction FIX avoid the refinement of the thermal parameters of Mn1 and the "x" coordinate of atom C23.

 

 

15 January 2005

 

Introduction of some facilities to treat layered structures

 

New Features

 

·         A section for commands has been introduced in the PCR file. A part of using the string containing the name of the phase for passing keywords and command, it is now possible to include up to 20 commands per phase. These commands are given starting with keywords. The command section for each phase starts after the name of the phase with the word COMMANDS and ends few lines below with the words END COMMANDS. In the middle one can give commands/keywords instructing the program for different purposes (VARY,DLIM, etc...). This works only in the automatic mode.

 

·         (Updated, February 15, 2005)

Among the available commands, the keywords LAYER_A, LAYER_B and LAYER_C are used to tell the program that the current phase will be treated as a single layer, so the reflections (H00),(0K0) or (00L), respectively, will be suppressed and the atom coordinates will be given in angstroms along the corresponding direction.

Considering an artificial (quasi-empty) supercell cell with a very long axis in the perpendicular direction to the layer, the 3D diffraction pattern approach to that of the single layer diffraction pattern without need of integrating the rods in reciprocal space. To eliminate the ripples a long axis together with a special broadening is enough. The method is based in the idea developed by K. Ufer et al.(Z. Kristallogr. 219, 519 (2004)) making the method perfectly compatible with the Rietveld method.

 

The keywords LAYER_A, LAYER_B and LAYER_C must be followed by the order of the supercell used, as in:

 

        COMMANDS

          LAYER_C 15

          .....

        END COMMANDS

 

This means that corresponding supercell parameter (as given in the unit cell line) is c=15*cs, being cs the c-parameter of the subcell. This is needed to conserve the (0,0,l=L/15) reflections that are treated separately. This option is still at a testing stage. A complete account will be given as soon as possible.

 

 

12 January 2005

 

The program Bond_Str is now available as stand alone application

 

New Features

 

·         A program for calulating distance and angles of a crystal structure as well as the bond valence sums for all ions is now available within the FullProf Suite. The program can also be obtained independently of the whole package just from the following ftp area:

                         ftp://ftp.cea.fr/pub/divers/Bond_Str

 

                    http://valmap.dfis.ull.es/fullprof

 

 

There are two programs: Bond_Str.exe and GBond_Str.exe. The first one performs the calculations and the second one is just a GUI for running the first or to import files. The calculation is the same as that already performed within FullProf,      however the advantage of the program is that it can be used to read CIF files and make the calculations without running FullProf. The input file can be a CIF file or a CFL file. For details read the manual Bond_Str.pdf or Bond_Str.htm.