12 October 2003

 

New Features

 

·         The current version of the program FullProf is 2.50  Oct2003

 

·         Some actions have been included to control local divergence of some parameters. Now instead of fixing the diverging parameter to an arbitrary value, the parameter is fixed to the previous value but the multiplier associated with its refinement codeword is divided by two in order to make more successful the refinement in the next cycle. A warning is always output. This allows a better control by the user of the refinement process.

 

·         The refinement flags concerning different patterns are now output explicitly even if the user has not given zero values. Now the output of the NPATT line adopts the following aspect:

 

NPATT      3       1 1 1   <- Flags for patterns (1:refined, 0: excluded)

 

When using using zero flags for some patterns remember that the weighting factor of each pattern is changed by the programa, so that the sum of weights of all considered patterns is equal to 1. If one wants to re-start with more patterns, check the weighting factors before running the program.

 

·         The resolution function of the different banks corresponding to a TOF diffractometer can now be included in a single file. One has to repeat the name of the file for each pattern in the PCR file. The different bank information has to be separated by the keyword "END" (without quotes) withint the IRF file. Now the intrumental parameters + refined shifts are output as comments in the PCR file when an IRF file is used in TOF.

 

The keyword TOFRG is now used for limiting the TOF read in the input data file. Only the points satisfying: TOF_min <= TOF <= TOF_Max, are used in the program. Examples of IRF files will be provided in the subdirectory "IRF-files" of the FullProf suite directory.

 

·         New modes for treating preferred orientation have been included. Putting NOR=2 the user can give several (up to a maximum of 5 per phase) preferred orientation directions (POD) and refine the corresponding March-Dollase parameter and the fraction of crystallites having a particular preferred orientation direction. The fraction corresponding to the last POD cannot be refined: it is calculated from the other values in such a way as to get a sum equal to 1.

 

·         If NOR=3 the Multi-axial March-Dollase model for preferred orientation adapted for high pressure anvil cells in which the incident beam is in the same direction as the external force producing the high pressure and perpendicular to the image plate collecting the data. A numerical integration is performed to calculate the value of  corresponding to expressions (4),(5) and (6) of the paper:  "Observation and Modelling of Preferred Orientation in Two-Dimensional Powder Patterns" Journal of Synchrotron Radiation 3, 112-119 (1996).   by N. G. Wright, R. J. Nelmes, S. A. Belmonte and M. I. McMahon

 

The input in the PCR file corresponding to both models is similar. If the program detect that NOR=2 or NOR=3 a line containing the number of POD (NPOD) is read after the pair of lines following the text

"!  Pref1  Pref2   Asy1    Asy2    Asy3     Asy4"

 

or the text:

"!  Pref1  Pref2   alph0   alph1   beta0   beta1 "  for TOF,

 

or the text:

"!  alph0t    beta0t   alph1t   beta1t " in case of using the function   NPROF=10 for TOF.

 

If NOR=3 a second integer is read in the same line that corresponds to the number of steps for the numerical integration (is not given the program uses NOrstep=20)

 

It follows a set of NPOD lines containing the following information:   Indices of the POD direction (3 real numbers), March-Dollase parameter and refinement code (2 real numbers) and Fraction of crystallites and code for the current POD.

 

Example:

 

For NOR=2

....

!  Pref1    Pref2      Asy1     Asy2     Asy3     Asy4      S_L      D_L

  0.00000  0.00000  0.00000  0.00000  0.00000  0.00000  0.00000  0.00000

     0.00     0.00     0.00     0.00     0.00     0.00     0.00     0.00

!  March-Dollase multi-axial model for preferred orientation (naxes)

    2

!    Pref-hkl           MD-r   Code(MD-r)    Fraction  Code(Fract.)

  1.00  0.00  1.00      1.3807   91.000      0.6246  101.000

  2.00  1.00  1.00      0.6107  111.000      0.3754    0.000

.......

 

 

For NOR=3

....

!  Pref1    Pref2      Asy1     Asy2     Asy3     Asy4      S_L      D_L

  0.00000  0.00000  0.00000  0.00000  0.00000  0.00000  0.00000  0.00000

     0.00     0.00     0.00     0.00     0.00     0.00     0.00     0.00

!  March-Dollase-Wright multi-axial model for preferred orientation

!  (naxes, N_Int_steps)

    2    30

!    Pref-hkl           MD-r   Code(MD-r)    Fraction  Code(Fract.)

  0.00  0.00  1.00      1.1102   61.000      0.7200   71.000

  2.00  1.00  0.00      0.7107   81.000      0.2800    0.000

....

 

Bugs

 

·         A new version of the program  EdPCR has been included in the FullProf Suite. The latest version had a bug: the change of profile function was disabled.

 

 

 

30 September 2003

 

Bugs

 

·         A bug concerning the scale factors in single twinned single crystal refinements was introduced in the May2003 version that was still in the latest version. Contrary to the precedent versions the refinement of several scale factors was diverging systematically (another editing error!). The bug has been corrected.

 

 

29 September 2003

 

Bugs

 

·         Sorry! a copy/paste error was introduced in the latest version (24 September) of FullProf interrupting the reading of reflections in the second run of a Le Bail fit before the end of the file. Now this bug is corrected.

 

 

24 September 2003

 

New Features

 

·         A new format for data files has been implemented INSTRM=14. It corresponds to the multibank format for TOF diffractometers existing at ISIS. The format is the following:

 

1.    The first line is considered as a title

2.    All following lines starting by the symbol # are considered as comments

3.    A line starting with the keyword BANK indicates that follows a series of lines containing three items: scattering variable (TOF/2theta/Energy), profile intensity and sigma. Internally the intensity and sigma are multiplied by a factor 1000. This is because the normalization procedures used at ISIS produce quite small numbers and FullProf expects "counts" as input for profile intensities.

The user can change this factor putting the keyword "fac_y" followed by the desired factor (a real number) in one of the comment lines before the keyword BANK. Remember that if the final intensities are too small numbers (order of unity) the calculation of RBragg factors may be in error. If within the line containing the keyword BANK there is the keyword RALF, the input intensity and sigma values are divided by the steps according to:

 

                  

                 

 

4.    If all banks are to be used, in the PCR file one has to repeat the name of the data file a number of times equal to the number of banks (number of patterns). If the user experiences that the generation of reflections takes long time, this is certainly due to too short times of flight, included in the data file, that do not contain useful information. This can be overcome just putting in the line containing the name of the phases the instructions 'DLIM n_pat d-spacing' See the end of the note of 20 January 2003 for details.

 

 

Bugs

 

·         Several small bugs of the latest version have been corrected:

 

1.    The new format for INSTRM=10 was incompatible to some older versions

 

2.    The CIF file generated for powder work was corresponding to single  crystal work and the profile file was not output. Now when RPA=-1 two CIF files are generated

 

3.    Some format errors have been corrected in the generated CIF files.

 

 

28 July 2003

 

New Features

 

·         The current version of the program FullProf is 2.45  Jul2003

 

 

·         Some changes have been introduced for treating the background:

 

1.    The polynomial background of 12 coefficients, for constant wavelength case, has been changed so that the last three coefficients correspond to inverse powers of 2theta. The ,  and  coefficients correspond to the powers ,  and  respectively. The background for the case Nba=-3 corresponds to the formula:

 

                

 

 

2.    Now there is the possibility to include several previously calculated profiles as contributing, through a linear combination, to the background of a powder diffraction pattern. The individual profiles are read in input files named "filedat_n.bac". Where "filedat" is the code of the data file corresponding to a diffraction pattern and the       index "n" is the number of the contributing profile. The format of these files should correspond to INSTRM=0 and must have the same number of points as the observed powder diffraction pattern. At present a maximum of 10 user-given profiles can be used as an additional contribution to the background calculated using whatever function available previously. The additional contribution to the background is calculated as:

 

                   

 

 

This is useful for complicated backgrounds or for adding calculated thermal diffuse scattering to a powder diffraction pattern. To use this new option the flag MORE in the same line as the variable "Zero"       must be 2 or 3. In this last case (MORE=3) Microabsortion coefficients are read first, then a line containing in the first position the number of profiles and the coefficients of the linear combination is read. An additional line is read for the refinement codewords of the coefficients.

 

An example of the relevant part of a PCR file is the following:

 

............................

!

!  Zero    Code    Sycos    Code   Sysin    Code  Lambda     Code MORE

   0.00000   0.00 -0.29309   0.00  0.00025   0.00 0.000000    0.00   3

!

! Microabsorption coefficients for Pattern#  1

!   P0    Cod_P0    Cp   Cod_Cp     Tau  Cod_Tau

 0.6950    0.00  0.6413    0.00  0.0452    0.00

!

! Num. and coeff. of profiles contributing to background of Pattern#  1

   1      1.2310

          51.0000

............................

 

For using only microabsoprtion coefficients MORE=1 and for using only     additional background contributions MORE=2. To use this option with     TOF neutron diffraction MORE=2.

 

 

3 July 2003

 

Bugs

 

·         A bug in the program BasIreps was introduced in the last version providing the wrong number of generators of the space group. This has been corrected and new output is available concerning the general expressions of the Fourier coefficients. The Fourier coefficients  are written in terms of symbolic free parameters.

 

·         The values of the weight fraction of the phases when using (JBT=-3, IRF=2) for quantitative analysis was not written in the output files. This has been corrected.  Now the written hkl-files (h,k,l,M, Struct-Fact ...) when Hkl=5 contain the value of ATZ. This ensures that the used ATZ-values agree with those used for calculating the structure factors.

 

 

10 June 2003

 

New Features

 

·         New documents have been written to explain the use of multipolar refinements in flipping ratio refinement of single crystals and to detail the peak shape functions used for TOF powder diffraction.

 

·         Reorganization of the TOF peak shapes and derivatives. The refinement of the instrumental parameters is now much more stable. Some internal changes make differences with respect to previous versions for the values of the profile parameters. The new peak shape (thanks to Laurent Chapon!) consisting in the convolution of a pseudo-Voigt function with the Ikeda-Carpenter function is now working. The TOF peak shapes used in FullProf and the meaning of each refinable parameter is now documented in the note: TOF_FullProf.PDF

 

 

Bugs

 

·         A bug in the program Fourier was introduced in the last version that was appearing with non-centrosymmetric structures. This has been corrected. The direct access to GFourier help-file is now working.

 

·         The interface program EdPCR is being modified and corrected extensively in order to include most of the capabilities of FullProf. There are still many things to be included and corrected but we intend to make so easy the manipulation of the PCR files that even experienced user will find advantageous the use of EdPCR for doing modifications in their PCR files. We have modified the creation of simple templates and simplified the creation of a new PCR file from the scratch.

 

 

10 May 2003

 

New Features

 

·         The current version of the program FullProf is 2.40  May2003

 

·         Different PANalytical formats and extensions are automatically recognized for INSTRM=13.

 

·         New keywords have been introduced for IRF files. The interface program EdPCR is now able to write an IRF using a template. A subdirectory called IRF-files will be created on installing the FullProf suite where several examples of IRF files can be found.

 

·         The program makes automatically constraints when two atoms occupy the same crystallographic position. This apply only for working in automatic mode.

 

·         Some modications in the profile matching modes have been performed. Now the case jbt=3 works as stated in the manual (jbt=-3 was already working well!). The case jbt=2 and several atoms were given, the program was starting "always" from the calculated structure factors of the known part of the structure. Now only for the first run (irf=0) the atoms are important. This option serves to bias the distribution of overlapped intensities according to the known atom positions.

 

·         Single crystal simulations (flipping ratios ...) are now possible by putting JOB=2,3 and CRY=4. The program generates reflections according to the provided ,  angles. The wavelength and other characteristics of the simulation are provided at end of the PCR file.

 

In the example below, the used wavelength is 0.711 angstroms, itypd=2 means that the program will calculte flipping ratios from the model given in the same file. Polarp and Polarm are the efficiency of the flipper for neutrons "up" and "down" respectively, UB_mat=1 means that the program will read the orientation matrix just below the current line. If itypd=0,1 the program will calculate structure factors and the output *.int file will contain squares of structure factors (itypd=0) or simply the modules of structure factors (itypd=1).

 

Example:

 

COMM Simulation of Flipping ratios

! Files => DAT-file: flipr-sim,  PCR-file: flipr-sim

!Job Npr Nph Nba Nex Nsc Nor Dum Iwg Ilo Ias Res Ste Nre Cry Uni Cor Opt Aut

   3   0   1   0   0   0   0   0   0   0   0   0   0   0   4   0   0   0   0

!

!Ipr Ppl Ioc Mat Pcr Ls1 Ls2 Ls3 NLI Prf Ins Rpa Sym Hkl Fou Sho Ana

   0   0   1   0   1   0   0   0   0   1   0   0   0   0   0   0   0

!

!NCY  Eps  R_at  R_an  R_pr  R_gl     Thmin       Step       Thmax    PSD    Sent0

  1  0.01  1.00  1.00  1.00  1.00     10.0000     0.1000    89.9000   0.000   0.000

!

  . . . . . . . . . . . . . . . . . . . . . . . . . . . .

  . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 

   Tail of the PCR file adapted for single crystal simulations.

  . . . . . . . . . . . . . . . . . . . . . . . . . . . .

! x-Lambda/2  +          Not yet used parameters

     0.03000     0.00000     0.00000     0.00000     0.00000

        0.00        0.00        0.00        0.00        0.00

! Parameters for Single X-tal calculations

! Lambda   itypd    ipow    Polarp    Polarm  UB_mat

  0.7110       2       0   0.89000   0.89000    1

! UB_matrix

   -0.017950    0.000610    0.188620

   -0.042370   -0.173200   -0.003470

    0.127620   -0.031460    0.012250

 

 

Bugs

 

·         Bug correction: the output SHELX *.ins files for profile matching mode   contains now the symmetry operators.

 

·         Some modications in the profile matching modes have been performed. Now the case jbt=3 works as stated in the manual (jbt=-3 was already working well!). The case jbt=2 and several atoms were given, the program was starting "always" from the calculated structure factors of the known part of the structure. Now only for the first run (irf=0) the atoms are important. This option serves to bias the distribution of overlapped intensities according to the known atom positions.

 

 

 

5 March 2003

 

New Features

 

·         A new option for treating the profile parameters of special reflections has been included in FullProf. This option works with Constant Wavelength and Time of Flight diffraction patterns. The user may select few reflections from the pattern to treat them in a special manner: an additional Gaussian and Lorentzian broadening with respect to the values calculated with the resolution parameters, as well as the shift with respect to the calculated position (from cell parameters) can bi fitted. This situation may be found in cases of defective materials for which the law governing the shifts and broadening is not known in advance, or in cases of instrumental defects(slight change of wavelength accross the pattern, etc).

 

At present 50 reflections per phase and per pattern is the maximum allowed.

 

For using this option for a given phase and a given pattern the user should provide the value of the variable "Nspec_ref(iphase,n_pat)" representing the number of reflections to be treated as "specials". This variable is given immediately after the variable Jtyp(iphase,n_pat)"

 

The program expect to read (in case of Nspec_ref /= 0) a list Nspec_ref lines containing: 

 

hkl, nvk, D-HG^2, code, D-HL, code, Shift, code

 

where nvk is the number of the propagation vector (if relevant), code   is the refinement code for the parameter. The list starts at the end of the profile parameters for a given pattern.

 

The meaning of the parameters is the following:

 

The Gaussian  for a special reflections is calculated as:

 

     

      

 is treated as a free parameter.

 

The Lorentzian  for a special reflections is calculated as:

       

      

 is treated as a free parameter.

 

    

The position of a special reflections is calculated as:

 

            

      

 

Shift is treated as a free parameter.

 

 

Example:

 

Case of PCR file corresponding to the format used with a single pattern

 

!-------------------------------------------------------------------------------

!  Data for PHASE number:   1  ==> Current R_Bragg for Pattern#  1:     1.06

!-------------------------------------------------------------------------------

 Myphase

!

!Nat Dis Ang Pr1 Pr2 Pr3 Jbt Irf Isy Str Furth     ATZ   Nvk Npr More

   6   0   0 0.0 0.0 1.0   0   0   0   0   0     5050.20   0   7   1

!

!Jvi Jdi Hel Sol Mom Ter  Brind   RMua    RMub    RMuc   Jtyp  Nsp_Ref

   0   0   0   0   0   0  1.0000  0.0000  0.0000  0.0000   1   3

!

P 3 1 c                  <--Space group symbol

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

!  Pref1    Pref2      Asy1     Asy2     Asy3     Asy4      S_L      D_L

  0.00000  0.00000  0.07373  0.01902  0.00000  0.00000  0.00000  0.00000

     0.00     0.00   251.00   241.00     0.00     0.00     0.00     0.00

! Special reflections:

!  h   k   l  nvk   D-HG^2  Cod_D-HG^2   D-HL    Cod_D-HL    Shift   Cod_Shift

   1   0   1    0  0.00000       0.000  0.04417   551.000  -0.01236    561.000

   2   0   0    0  0.00000       0.000  0.03056   571.000  -0.00274    581.000

   3   0   1    0  0.00000       0.000  0.00759   591.000  -0.00119    601.000

 

Bugs

 

·         A bug concerning the generation of reflections when particular propagation vectors of the form k=2H in centred cells has been corrected. In very special cases some reflections were not generated and it was necessary to put nvk=-1 to get all reflections generated. However this had, as a consequence, a bad calculation of the magnetic moment amplitude. The workaround was to create an hkl file to be read (irf=1) and putting back nvk=1. Now this is no more needed.

 

 

20 January 2003

 

New Features

 

·         A new automatic installation of the FullProf suite is available. The   programs FullProf (fp2k, wfp2k), GFourier, Fourier, WinPLOTR, EdPCR,   BasIreps, ITO, DICVOL91, TREOR, SuperCell, etc ... as well as the   corresponding documentation and examples are distributed in a single   executable installer.

 

·         Patterson calculation when using profile matching modes (JBT=2,3) can   be performed automatically using Fou=4 and the program GFourier.

 

·         A new value for the IRESO variable for the instrumental resolution function has been implemented. When IRESO=-1 the program expects to read the following variables:

 

     U-inst   V-inst  W-inst    ETA-inst   X-inst   Y-inst

 

The program takes the values of FWHM and ETA a pseudo-Voigt function (NPRO=5) deduced from the above values as:

 

               

 

                  

 

and calculates for each reflection the Gaussian (HG-inst) and Lorentzian (HL-inst) components of the corresponding Voigt function using the De Keijser formula (J.Appl.Cryst. 1982, 15, 308-31). These are added quadratically for the Gaussian and linearly for Lorentzian components of the T-C-H (NPRO=7) pseudo-Voigt of the sample.

 

·         Instrumental resolution function (IRF) files have been implemented for TOF patterns. At present it is available for profile functions NPROF=9 (Res=5) and NPROF=10 (Res=6). When an IRF file is read, all the corresponding refinable parameters read in the PCR file are additive, so that the refined parameters are shifts with respect to the instrumental values.

 

The format of the IRF files is the following:

First line: considered as a TITLE

All empty lines and lines starting with "!" or "#" are considered as comments. The rest of lines contain the following keywords and real values. The order of keywords is irrelevant but the keywords should be written in upper case characters.

 

      D2TOF   Dtt1  Dtt2

      TWOTH   2ThetaBank

      SIGMA   Sig-2  Sig-1  Sig-0

      GAMMA   Gam-2  Gam-1  Gam-0

      ALFBE   alph0  beta0  alph1   beta1

      ALFBT   alph0  beta0  alph1   beta1

 

Below there are two examples of IRF files

 

Example of IRF file for RES=5, NPROF=9

----------------------------------- file starts in the next line

Instrumental resolution function of TOF of GPPD at IPNS (Argonne)(from NBS-Si)

 

! Test of the IRF for TOF in FullProf. RES=5

! This is a comment

# This is also a comment

# IRF valid for profile function number 9 in Fullprof

!        Dtt1      Dtt2

D2TOF  7476.910   -1.540

!        2ThetaBank

TWOTH    144.845

!       Sig-2     Sig-1     Sig-0

SIGMA   0.000    49.020     5.987

!       Gam-2     Gam-1     Gam-0

GAMMA   0.000     7.642     0.000

!          alph0       beta0       alph1       beta1

ALFBE   0.000000    0.042210    0.597100    0.009460

 

 

Example of IRF file for RES=6, NPROF=10

---------------------------------- file starts in the next line

Instrumental resolution function of TOF of GPPD at IPNS (Argonne)

 

! Test of the IRF for TOF in FullProf. IRESO=6 Jason Hodges' TOF vs d

# IRF valid for profile function number 10 in Fullprof

#        Dtt1

D2TOF  7478.332

#        Dtt1t       Dtt2t    x-cross    Width

D2TOT  7459.7754    19.9522    1.2640    2.3630

#     TOF-TWOTH of the bank in degrees

TWOTH    144.845

#       Sig-2     Sig-1     Sig-0

SIGMA   0.000    33.000     0.000

#       Gam-2     Gam-1     Gam-0

GAMMA   0.000     0.000     0.000

#          alph0       beta0       alph1       beta1

ALFBE   1.470500    2.483000   -0.853300    9.298000

#         alph0t      beta0t      alph1t      beta1t

ALFBT   7.387000   27.180000    4.012400    4.733000

 

·         A new keyword has been introduced to modify the behaviour of the program in relation with the generation of reflections. In one writes in the line containing the name of the phase the word DLIM followed by the number of the pattern and the value of a d-spacing, only the reflections having a d-spacing greater than the provided value are used in the refinement of the data.

 

Example:

 

!---------------------------------------------

My phase name         DLIM 2 2.14    DLIM 3 1.5

!Nat Dis ....

 

The reflections of pattern number 2 are limited to d-spacing greater than d=2.14 angstroms and those of pattern number 3 to d-spcing greater than 1.5 angstroms. This may be useful for magnetic structure refinement when using a diffraction pattern going far in Q (for instance in TOF case).

 

Bugs

 

·         Some minor bugs in the generation of CIF files have been corrected.   In particular, related coordinates of type (x,2x,z) or (x,-x+1/4,z) were output with numerical values corresponding to (x,x,z).

 

·         Correction of some bugs in calculating the background for polynomial and Debye-like refinements before starting the refinement cycles. This was without consequences in the refinement. This bug was introduced when the refinement of the linearly interpolated background was implemented.