Full Year 2021
----------------------
>> 15 December 2021 : Changes in multi-pattern with symmetry modes.
----------------------
- The previous version of FullProf was unable to work properly when a phase treated using symmetry modes
was not contributing to the first pattern. Now this has been changed and corrected.
----------------------
>> 7 December 2021 : The program TOF_fit_LM can now be run using Orthogonal Distance Regression algorithm.
----------------------
- The program TOF_fit_LM has been improved by introducing the optimizacion algorithm ODR (Orthogonal Distance
Regression), which is based in a Levenberg-Marquardt method complemented a projection method for handling
box constraints. We use a wrapper to the package ODRPACK95 (Jason W. Zwolak et al., http://hdl.handle.net/10919/20185)
- Some minor changes in WinPLOTR for better handling .xrf files resulting from refinement programs XRFIT and TOF_fit_LM
----------------------
>> 2 December 2021 : New version of FullProf and TOF. Increasing the number of scale factors for single crystal data.
----------------------
- The current version of the FullProf program has been updated to:
**********************************************************
** PROGRAM FullProf.2k (Version 7.50 - Dec2021-ILL JRC) **
**********************************************************
- The effective number of refinable scale factors in single crystal patterns was in fact 21 and not 24 as expected.
This was due to a side effect that was putting equal to zero the last three parameters. Now this bug has been
corrected and the number of scale factors for the treatment of single crystal data has been increased up to 36.
- Some PCR files of the Examples subdirectory have been updated.
----------------------
>> 20 October 2021 : Correcting few bugs
----------------------
- The orthogonal system text written in the *.out files of FullProf was not correct.
It was marked as the old default frame:
"=> Cartesian frame: z // c; y is in the bc-plane; x is y ^ z = a*"
but the true orthogonal system used inside, changed the 18 September 2004, was:
"=> Cartesian frame: x//a; z is along c*; y is within the ab-plane"
- Adding a checking within the toolbar to avoid user miss-dimension of arrays in FullProf.
- Correcting a bug in DataRed, the averaged k-domain intensities was not correct when the
star of incommensurate k-vectors had more than 4 arms. A list of the changes in the indices
of measured reflections is now included in the output file.
- Correction of bugs in EdPCR when entering in the box-constraint and linear restraints dialogs.
- Other cosmetic changes have been performed in different programs.
----------------------
>> 28 July 2021 : Correcting few minor bugs
----------------------
- The calculation of R-Magnetic in the version of April 2021 was bad in some particular circumstances when
the maximal number of counts were very low. Now it should be OK.
- In some circumstances the calculation of TOF patterns was bad (NaN values) due to a problem of bad optimization
flags of the compiler. This has now been corrected.
----------------------
>> 17 May 2021 : New Energy x-space in WinPLOTR-2006. Better installation for Linux
----------------------
- Energy dispersive x-space has been introduced in WinPLOTR-2006 (It was already operational in WinPLOTR)
It is assumed that the energy is provided in KeV (Kilo-electron Volts).
- Change of the VESTA file generated by FullProf for magnetic structures: the binary .pgrid file has
been removed from the VESTA file when the purpose is to represent a magnetic structure.
- An example of Lebail fit for X-ray dispersive energy pattern has been included into the Examples directory.
- An item in the File Menu has been introduced in EdPCR for better control of the settings
- The setting file wplot050.set for WinPLOTR-2006 has been renamed as wplot-2006.set
- When installing the FullProf Suite for the first time in Linux the following command should be executed after
going to the directory in which it has been installed: "source Set_FULLPROF_Envi" (without quotes).
If for any reason the command does not work, you should follow the procedure in the Web page.
Remember that some programs of the FullProf Suite require a high stack memory, for that the best is
to execute, in the terminal from which the toolbar will be opened, the command: "ulimit -s unlimited"
----------------------
>> 30 April 2021 : Calculation of site energies in BondStr. New x-space in WinPLOTR-2006
----------------------
- When calculating BVEL maps whith Bond_Str, the calculation of the site-energy for all the atom sites of the
same species as the migrating species. This is useful for calculating the activation energies of the different
crystallographic sites by doing the difference with respect to the percolation energy. The proper subroutine
has been written by Nebil A. Katcho.
- A crash of the program GBondStr in Linux has been corrected.
- A new scattering variable called "Omega" has been introduced in WinPLOTR-2006 to put the correct label
when representing refined single crystal 1-D profiles by scanning the Omega angle. This is used in
*.xrf files generated by the integration program SXTAL_Fit. This program will be included within the
data reduction program Int3D to be released in the forthcoming months.
----------------------
>> 23 April 2021 : Changes in the VESTA management by the Toolbar and FullProf
----------------------
- The output files of FullProf concerning VESTA have been changed slightly. When the superspace method is used
the invocation of VESTA in the toolbar takes automatically the P1 mCIF file as default.
- The CIF and VESTA files generated by FullProf when Shubnikov groups are used have been simplified. Up to now
two files (with extensions .cif and .mcif) were generated and the VESTA file imported the CIF file. Now,
the generated VESTA file imports directly the mCIF file.
- The scale factor used by VESTA for arrows is now calculated by FullProf before writing the VESTA file.
----------------------
>> 6 April 2021 : New version of WinPLOTR for Windows
----------------------
- The new version of the program WinPLOTR (32 bits), from Thierry Roisnel, has been included in the last version
of the FullProf Suite distribution.
- The mCIF file generated by FullProf in P1 for VESTA includes now all atoms described in the superspace phase:
magnetic and non-magnetic.
----------------------
>> 30 March 2021 : The program MHall. Corrections of few bugs and improvements
----------------------
- The program MHall (console only) has been added to the FullProf Suite distribution. This program allows to generate
the magnetic Hall symbol as described in the paper:
"Extension of Hall symbols of crystallographic space groups to magnetic space groups",
Javier Gonzalez-Platas, Nebil A. Katcho and Juan Rodriguez-Carvajal
Journal of Applied Crystallography 54(1), 338-342 (2021)
https://doi.org/10.1107/S1600576720015897
The manual and instructions to how use the program MHall can be obtained from the supporting information corresponding
to the above paper. The magnetic Hall symbols allows to describe unambiguously the generators of the magnetic space groups
for reasonable arbitrary settings. It is suggested that articles concerned with conmmensurate magnetic structures should
provide the magnetic Hall symbol of the magnetic group actually used in the refinement.
- A bug in the MCIF file in P1 for VESTA when using superspace formulation of magnetic structures has been corrected.
The error affected mostly the superspace groups containing non-zero rational components (thanks to Ovidiu Garlea for
communicating the error), however in some other cases some atoms had a wrong orientation.
The generated files have been now tested extensively with MVISUALIZE at the Bilbao Crystallographic Server
that reads the files *_ssg*.mcif (incommensurate mCIF) that VESTA is unable to read properly.
Remember that the supercell approach used in the mCIF file generated by FullProf, of name CodePCRn.mcif (n=1, 2...),
when read by VESTA an unavoidable error occurs because for VESTA the system is periodic. Atoms in edges and faces of the
supercell may have a wrong orientation (in one of the opposite edges or faces).
----------------------
>> 10 January 2021 : New version FullProf. TOF improvements.
----------------------
- The current version of the FullProf program has been updated to:
**********************************************************
** PROGRAM FullProf.2k (Version 7.40 - Jan2021-ILL JRC) **
**********************************************************
- Improvement of the TOF peak shapes. As suggested by the POWGEN instrument team we have modified the
dependency of peak shape parameters with d-spacing (D) for the TOF-function number 9 (convolution of
back-to-back exponentials with pseudo-Voigt). Old PCR files can be read automatically and converted
to the new format in the output PCR(NEW) file.
We have added a new term (Dtt_1overD) in the TOF versus d-spacing. Other terms have also been included
in the dependency of the Gaussian part of the Voigt function versus d-spacing and in the rise and decay
terms of the exponentials.
The formulae for profile number 9 we use now in FullProf are the following:
TOF(microseconds) = Zero + Dtt1 * D + Dtt2 * D^2 + Dtt_1overD)/D (with D in Angstroms)
Gaussian variance and Lorentzian FWHM of the Voigt function:
Sigma^2 = sig_2 * D^4 + sig_1 * D^2 + sig_0 + sig_q / D^2
Gamma = gam_2 * D^2 + gam_1 * D + gam_0
Rise "constant" of the exponential:
Alpha = alpha_0 + alpha_1 / D + alpha_q / Srqt(D)
Decay "constant"of the exponential
Beta = beta_0 + beta_1 / D^4 + beta_q / D^2
The free parameters of the profile function number 9 are now
Peak positions:(Zero, Dtt1, Dtt2, Dtt_1overD)
Peak shape: (sig_2,sig_1,sig_0,sig_q, gam_2, gam_1, gam_0, alpha_0, alpha_1, alpha_q, beta_0, beta_1, beta_q)
(one can select the desired expressions by nullifying the appropriate terms)
Of course these parameters cannot be refined simultaneously. One has to study a standard sample to fit the instrumental
parameters, for a particular machine and detector bank, and fix them for real samples, for which only size and strain
coefficients can be refined.
These new parameters are now being tested to see what is the best way of working. In fact it is quite difficult to
obtain the parametres just starting at random and the correlation between them is quite strong.
- Remember that another option existing in FullProf is the possibility of providing an instrumental resolution file
in which Sigma, Gamma, Alpha, Beta and Shift are numerically provided as a function of D-spacing. The particular
instrumental parameters Sigma, Gamma, Alpha, Beta and Shift for a particular peak is interpolated linearly.
This option is probably the most precise given the complexity of the dependency with d-spacing of TOF peak
shapes.
As above only the coefficients related to strain and size are to be refined.
- For facilitating the creation of an IRF file for Res=5 using the interpolation we have created a new program, called
TOF_fit_LM (only console, see the text documents in the Toolbar), allowing to refine limited ranges of the patterns,
wich produces *.irf files that can be merged into a single IRF file. Otherwise, FullProf is also producing this kind
of files after a refinement and its generation can be forced even if we are already reading a raw IRF file. For doing
that one can include in the COMMANDS block the keyword: Generate_new_IRF, as in the following example
. . . . . . . . . . .
!-------------------------------------------------------------------------------
! Data for PHASE number: 1 ==> Current R_Bragg for Pattern# 1: 2.7425
!-------------------------------------------------------------------------------
NAC FIX xyz b
!
COMMANDS
Generate_new_IRF
END COMMANDS
!Nat Dis Ang Jbt Isy Str Furth ATZ Nvk More
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
- We have included now the T-C-H formulation for the Voigt function (profile number 7) in the TOF patterns. This
(symmetric) Voigt function in TOF is only adequate for TOF machines with relatively long pulses in which peak
shaping by using choppers is operative. In some cases when the broadening of the sample is clearly bigger than
the instrumental contribution this function may be used. Examples with well crystallized samples (Si and NAC)
are provided to see the mismatch with respect the to true back-to-back exponential convoluted with pV.
The dependency on D-spacing of TOF, Sigma and Gamma are the same as for profile function number 9.
Rise and decay constant are absent in this formulation.
----------------------
>> 1 January 2021 : Simulated annealing and preferred orientation.
----------------------
Information for working with simulated annealing in presence of preferred orientation
=====================================================================================
It works only with Nor=1 and full profile SAnn: Ipr=-2 and Accept negative.
The preferred orientation parameter Pref1 should be known in advance or refined in the same conditions for the nuclear phase.
Once the parameter Pref1 is known, it is necessary to prepare a LeBail fit of the additional peaks maintaining fixed the
nuclear phase with the corresponding Pref1. The parameter Pref1 should be provided also in the phase with the LeBail fit putting
also: More = 1 and Jvi=11. The preferred orientation is incorporated in the information about the peak shapes within the
file *.spr, so the use of SAnn is done as usual.
Crystal structure case:
======================
An example of simulated data with a resolution of crystal structure is provided in the subdirectory CrystalStructure-SAnnPrefOr within
the Examples directory of the distribution.
Simulated patterns of LaMnO3 without preferred orientation ("lamn_calc_nor.dat") and preferred orientation ("lamn_calc_pref.dat")
with Pref1=0.66 and preferred orientation axis [0,0,1]. A profile matching (LeBail) fit for each pattern generates profile
files *.spr and intensity files *.int (PCR files: "lamn_pm_nor.pcr" and "lamn_pm_pref.pcr" for free and oriented pattern respectively).
PCR files for SAnn are also provided for solving the crystal structure. The first file (lamn_san_nor.pcr) corresponds to the
solution of the crystal structure using the pattern file "lamn_nor.spr" and the *.int file "lamn_pm_nor1_cltr.int". The second
file ("lamn_san_pref.pcr") corresponds to the solution of the crystal structure using the pattern file "lamn_pref.spr" and the
*.int file "lamn_pm_pref1_cltr.int"
One can verify that the results provide the same crystal structure (or equivalent) as seen using FullProf Studio for visualizing
both structures. This demonstrates that the method works correctly.
Magnetic structure case:
=======================
An example of simulated data with a resolution of a magnetic structure is provided in the subdirectory MagneticStructure-SAnnPrefOr within
the Examples directory of the distribution.
Simulated patterns of Ho2BaNiO5 without preferred orientation ("hobk_cal_nor.dat") and preferred orientation ("hobk_cal_pref.dat")
with Pref1=1.48 and preferred orientation axis [1,0,1]. A profile matching (LeBail) fit for each pattern generates profile
files *.spr and intensity files *.int (PCR files: "hobk_pm_nor.pcr" and "hobk_pm_pref.pcr" for free and oriented pattern respectively).
PCR files for SAnn are also provided for solving the magnetic structure. The first file (hobk_san_nor.pcr) corresponds to the
solution of the magnetic structure using the pattern file "hobk_nor.spr" and the *.int file "hobk_pm_nor1_cltr.int". The second
file ("hobk_san_nor.pcr") corresponds to the solution of the magnetic structure using the pattern file "hobk_pref.spr" and the
*.int file "hobk_pm_pref1_cltr.int"
One can verify that the results provide the same magnetic structure (or an equivalent domain) as seen using FullProf Studio for
visualizing both structures. This demonstrates that the method works correctly.