---------------------- >> 3 October 2006: Magnetic domains in single X-tal refinements ---------------------- - The presence of magnetic domains of any type can now be handled easily with the current version of FullProf. Within the commands associated with a particular phase the new keyword MAGDOM has been created. To introduce a set of domains one has to write a line per domain with the keyword MAGDOM (including the domain 1 with identity matrix) followed by rotational operators, a real number (as for a magnetic symmetry operators used in MSYM) a colon symbol and the values and refinement codes for the populations. Two artificial examples are given below . . . . . . . . Relevant portion of the PCR file My Magnetic phase name ! COMMANDS magdom u, v, w, 0.0 : 0.650 121.00 magdom -v, u, w, 0.0 : 0.250 131.00 magdom u,-v, w, 0.0 : 0.100 0.00 END COMMANDS !Nat Dis Mom Pr1 Pr2 Pr3 Jbt Irf Isy Str Furth ATZ Nvk Npr More 2 0 0 0.0 0.0 1.0 1 4 -1 0 0 0.000 -1 7 0 . . . . . . . . In the example above there are: three domains in total. Notice that the first domain corresponds to that of the magnetic model described in the PCR file. The sum of the populations are always 1.0 The refinement code of the last domain should always be zero. It is calculated as a function of the previous values in order to ensure that the sum is always 1. . . . . . . . . Relevant portion of the PCR file My Magnetic phase name ! COMMANDS magdom u, v, w, 0.1 : 0.250 0.250 121.0 131.0 magdom u,-v, w, 0.1 : 0.200 0.300 141.0 0.0 END COMMANDS !Nat Dis Mom Pr1 Pr2 Pr3 Jbt Irf Isy Str Furth ATZ Nvk Npr More 3 0 0 0.0 0.0 1.0 1 4 -1 0 0 0.000 -1 7 0 . . . . . . . . In the example above there are: two domains + two chirality domains, four domains in total. Notice that the first domain corresponds to that of the magnetic model described in the PCR file. A value of the real number following the operator greater than 0.001 means that we have chirality domains (second population value and second refinement code). The sum of all populations should be equal to 1.0 and the last one should not have a refinement code. The rotational operators should be those of the paramagnetic group that are lost in the transition to the ordered state. Be careful with trying to refine all domain populations. There are situations where two domains give exactly the same module of the magnetic interaction vector so that the relative population cannot be determined using non-polarised neutron diffraction. - A bug (sign) in the derivative with respect to one imaginary component of magnetic Fourier coefficients in the case of using Jbt=1 (but not for Jbt=-1) has been corrected. The bug was producing poor convergence in case of refinement of "Iy". ---------------------- >> 18 September 2006: Lambda/2 in single X-tal refinements ---------------------- - Changes in the code concerned with the contribution of Lambda/2 harmonics in single crystal refinements have been introduced. Now the contribution of Lambda/2 is explicitly written in the output file and the forbidden reflections are not eliminated from the refinement when (2h, 2k, 2l) is allowed. The flag "Excl" is still written but it refers to (hkl) and not to (2h,2k,2l). - A bug introduced in the latest versions of Bond_Str, when using disordered compounds has been corrected. Now the bond valence of a chemical species, occupying partially a site is correct. - The current version of FullProf has been updated to: **********************************************************' ** PROGRAM FullProf.2k (Version 3.71 - Sep2006-ILL JRC) **' **********************************************************' --------------------- >> 20 August 2006: Simulated annealing using profile intensities --------------------- - A new cost function, using the profile intensities, for simulated annealing has been included in FullProf. The program generates a new output file after a Le Bail fit when global Ipr=-1 and Cry=0 with extension ".spr" together with the integrated intensity file (*.int) when More=1 & Jvi=11. In such a file the profile intensities of selected points together with the profile (and other constants) contributions of each reflection of the proper phase are written and will be used to calculate the new cost function for subsequent simulated annealing jobs To run a simulated annealing (Cry=3) job, the user should make the usual PCR file and he(she) has just to put Ipr=-1 if he(she) wants to use the new cost function. function. The new cost function corresponds to the profile Rp (unweighted) of selected points. When there is only one phase, then the quantity "yraw(i)-bac(i)" is used as observed profile intensity (yobs(i) in the SPR file). If the Le Bail fit is used for a second (or higher) phase (magnetic contribution, for instance) then "yobs(i)" is taken as the "calculated profile" of the corresponding phase. It is expected that this function improves the success ratio for structure solution because it contains more information than the clusters of integrated intensities. --------------------- >> 27 July 2006: Simulated annealing using rigid body and angles in degrees --------------------- - A bug in the output of the solution found by simulated annealing when using a rigid body description and angles in degrees has been corrected. The values of the angles were multiplied by the factor converting radians to degrees twice before the output to the new PCR and "*.so1" files. This has been corrected. There was no problem when using radians. Up to now the limits of the angles had to be given in radians even if the angles were provided in degrees, this has been changed to be consistent with the global description using degrees. The program uses internally radians as the unit for angles but this is transparent for the user. Comments on that are now output in the OUT file. --------------------- >> 24 July 2006: Additional background profile code-names provided in the PCR file --------------------- - The codes of the files containing background profiles can now be arbitrary. Remember that to access this option one has to put MORE=2,3 in the same line as the ZERO point appears. According to the note of 28 July 2003, the only accepted names were of the form: "filedat_n.bac". Where "filedat" was the code of the data file corresponding to the diffraction pattern to be refined and the index "n" was the number of the contributing profile. This is still true by default. Otherwise the user can provide the individual codes of the contributing profiles as given in the following example: ........ ! ! Number and coefficients of profiles contributing to background of Pattern# 1 3 1.2310 0.3310 2.3133 51.0000 61.0000 71.0000 ! Codes of files containing additional background profiles myfile1 qwerty21 file33389 ............................ The program read the line following the codewords of the refined weights and if it finds the word "Codes" then the codes of the files are read in the next line. In the above example three files containing background profiles are read. The complete names of the three files are: myfile1.bac, qwerty21.bac and file33389.bac The maximum length of the codes for this kind of files is of 9 characters. The restrictions described in the note of 28 July 2003 (format of files and number of points) are still valid. --------------------- >> 22 July 2006: Treatment of Single Crystal Electron Diffraction Data --------------------- - The current version of FullProf has been updated to: ********************************************************** ** PROGRAM FullProf.2k (Version 3.70 - Jul2006-LLB JRC) ** ********************************************************** - The new version of FullProf can handle automatically data taken form electron diffraction patterns provided they are close to kinematical scattering. The access to this option is exactly the same as for treating single crystal X-rays or neutron data except that the parameter Job should be equal to 4 (Job=4 or Jtyp=4). The program applies kinematical theory using the Mott-Bethe formula for calculating the scattering factor of the atoms or ions. If the parameter Cry is put equal to 4 (Cry=4) the program simulates a single crystal electron diffraction pattern using the provided wavelength and the values of 2theta-init (Thmin) and 2theta-end (Thmax) given in the PCR file (see note of 10 May 2003 for details). For refining of solving crystal structures by simulated annealing, experimental data should be given as structure factors squared. It is supposed that the program handling the diffraction images makes the proper corrections in order to obtain h,k,l,F2 and sigma(F2). --------------------- >> 20 July 2006: Refinement of magnetic structures with SXtal data and magnetic groups --------------------- - The refinement of magnetic structures with k=0 using time reversal operators for single crystal together with the crystal structure is now possible without skipping the systematic extinction due to the crystallographic space group. Up to now putting JBT=10 and using time reversal operators was impossible due to the fact that the program suppressed automatically the reflections of the list that were forbidden for the space group (but not for the magnetic group). Now the reflections are not suppressed and only the magnetic contribution is calculated. These reflections appear in the output list with Icod=-1. --------------------- >> 30 May 2006: The official Web site of the Fullprof Suite --------------------- - The new official Web site for the FullProf Suite has been installed in the web page of the Diffraction Group at the Institute Laue-Langevin. The page is: http://www.ill.fr/pages/science/IGroups/diff/Soft/fp/ All previous web sites (not mirroing this site) contain non updated versions of the programs. I cannot access anymore the CEA site and I know that some files have been corrupted in particular the GBasIreps program for windows is not working. Please inform me (fullprof@ill.fr) of possible errors in using the programs distributed from this page. --------------------- >> 25 May 2006: Another way of using quadratic form of anisotropic size --------------------- - It is now possible to use a quadratic form for lorentzian anisotropic size broadening (HL = FZ(h,k,l)/cosTheta), as defined by the expressions below cte_siz=0.001*360*lamda/pi^2 FZ(h,k,l,...)= cte_siz*1000/D, D in same units as Lambda FZ(h,k,l,...)= cte_siz*d(hkl)^2*(c1 h^2+ c2 k^2+ c3 l^2+ c4 2hk+ c5 2hl+ c6 2kl), for whatever case of strain model. Up to now the use of this expression in microstructure refinement was limited by putting Str=2 or 3, now the only condition is to put Isizemodel=30. In particular it is possible to use spherical harmonics (Str= 4 + StrainModel) or the quartic form for strains (Str= 1 + StrainModel) together with the quadratic form for size broadening (Isizemodel=30). Up to six coefficients can be refined. An example of PCR file in which appear the size model 30 is given below: ................................................................................ ! Scale Shape1 Bov Str1 Str2 Str3 Strain-Model 10.000 0.00000 0.00000 0.00000 0.00000 0.00000 2 11.00000 0.000 0.000 0.000 0.000 0.000 ! U V W X Y GauSiz LorSiz Size-Model 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 30 0.000 0.000 0.000 0.000 0.000 0.000 0.000 ! a b c alpha beta gamma #Cell Info 4.871308 2.979645 9.287657 90.000000 95.000000 90.000000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 ! Pref1 Pref2 Asy1 Asy2 Asy3 Asy4 S_L D_L 1.00000 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 ! Generalised size parameters (quadratic form): ! h2 k2 l2 2hk 2hl 2kl 0.52036 2.4900 0.069229 0.0000 0.0000 0.0000 231.00 241.00 121.00 0.00 0.00 0.00 ! Y00 Y20 Y22+ Y22- Y40 (Strain) 88.000000 338.000000 117.000000 0.000000 -10.000000 61.00 51.00 71.00 0.00 41.00 ! Y42+ Y42- Y44+ Y44- (Strain) 11.000000 117.000000 0.000000 0.000000 31.00 21.00 0.00 0.00 ! Lorentzian strain coeff.+ code 0.20000 0.00000 ................................................................................. --------------------- >> 24 May 2006: A new option of anisotropic strains using spherical harmonics --------------------- - It is possible now to use spherical harmonics (SH) to handle anisotropic strains. Up to now the use of SH in microstructure refinement was limited to the use as modelling for anisotropic lorentzian size effects. We have included an option putting Str=4 and StrainModel=n, where n=1,2,-2,3,4,5,6,7,8,9,10,11,12,13,14. The value of n corresponds to the Laue classes: 1 2 -2 3 4 5 6,8 7,9,10 11 12 13 14 -1 12/m1 112/m mmm 4/m 4/mmm -3 -3m 6/m 6/mmm m3 m3m The spherical harmonics are referred to a cartesian frame where the z-axis is always along the principal symmetry axis. In the case of trigonal system the hexagonal setting is used and "c" is along the z-axis. The order of the harmonics has been arbitrarily limited as it is the case for size broadening. This may be changed in the future. The input is similar to that used for S_HKL coefficients except that for some cases the number of parameters is different. There are n1 and n2 values to be read in different lines. The codewords are always below the corresponding parameter. After the lines containing the coefficients of the spherical harmonics an additional line containing the degree of Lorentzian character of the strain, together with its codeword, is given. (n1,n2) ( 5, 1) (-1) : coeff. Y00,Y20,Y21+,Y21-,Y22+,Y22- (6 parameters) ( 5, 4) (2/m) : coeff. Y00,Y20,Y22+,Y22-,Y40,Y42+,Y42-,Y44+,Y44- (9 parameters) ( 6, 0) (mmm) : coeff. Y00,Y20,Y22+,Y40,Y42+,Y44+ (6 parameters) ( 4, 4) (4/m) : coeff. Y00,Y20,Y40,Y44+,Y44-,Y60,Y64+,Y64- (8 parameters) ( 4, 2) (4/mmm) : coeff. Y00,Y20,Y40,Y44+,Y60, Y64+ (6 parameters) ( 4, 1) (-3) : coeff. Y00,Y20,Y40,Y43+,Y43- (5 parameters) ( 4, 3) (-3m) : coeff. Y00,Y20,Y40,Y43-,Y60,Y63-,Y66+ (7 parameters) ( 4, 2) (6/m) : coeff. Y00,Y20,Y40,Y60,Y66+,Y66- (6 parameters) ( 4, 1) (6/mmm) : coeff. Y00,Y20,Y40,Y60,Y66+ (5 parameters) ( 5, 0) (m3) : coeff. K00,K41,K61,K62,K81 (5 parameters) ( 4, 0) (m3m) : coeff. K00,K41,K61,K81 (4 parameters) For instance a case corresponding to the Laue class 2/m is introduced in the PCR file (remember that Str=4) as follows: ................................................................................ ! Scale Shape1 Bov Str1 Str2 Str3 Strain-Model 10.000 0.00000 0.00000 0.00000 0.00000 0.00000 2 11.00000 0.000 0.000 0.000 0.000 0.000 ! U V W X Y GauSiz LorSiz Size-Model 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0 0.000 0.000 0.000 0.000 0.000 0.000 0.000 ! a b c alpha beta gamma #Cell Info 4.871308 2.979645 9.287657 90.000000 95.000000 90.000000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 ! Pref1 Pref2 Asy1 Asy2 Asy3 Asy4 S_L D_L 1.00000 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 ! Y00 Y20 Y22+ Y22- Y40 (Strain) 88.000000 338.000000 117.000000 0.000000 -10.000000 61.00 51.00 71.00 0.00 41.00 ! Y42+ Y42- Y44+ Y44- (Strain) 11.000000 117.000000 0.000000 0.000000 31.00 21.00 0.00 0.00 ! Lorentzian strain coeff.+ code 0.20000 0.00000 ................................................................................. --------------------- >> 21 April 2006: Some new options for single crystal --------------------- - A version of FullProf without the correction of the bug described in the note of 21 October 2005 has been distributed inadvertently. - A new option treating twinned crystals has been introduced. For using it the variable "N_Domain" appearing only when More=1 (in the line after the name of the phase) should be put equal to the number of domains (number of twins). In such a case instead of using a scale factor for each domain (as it is the normal option) a single scale factor is used and the fraction of N_Domains-1 can be refined. The fraction of the last domain is calculated with the restriction that Sum(fract)=1.0. This option is to be used compulsory when a heterogeneous (multidetectors) data collection has been performed in a twinned crystal and one wants to use all data simultaneously. Example: Portion of a PCR file using this new option !------------------------------------------------------------------------------- ! Data for PHASE number: 1 ==> Current R_Bragg for Pattern# 1: 9.02 !------------------------------------------------------------------------------- BiSrMnO3 a ! !Nat Dis Ang Pr1 Pr2 Pr3 Jbt Irf Isy Str Furth ATZ Nvk Npr More 14 0 0 0.0 0.0 1.0 0 4 0 0 0 3085.695 0 7 1 ! !Jvi Jdi Hel Sol Mom Ter Brind RMua RMub RMuc Jtyp Nsp_Ref Ph_Shift N_Domains 0 0 0 0 0 0 1.0000 1.0000 0.0000 0.0000 1 0 0 6 ! I b m m <--Space group symbol !Atom Typ X Y Z Biso Occ In Fin N_t Spc /Codes Bi BI 0.00031 0.00000 0.25000 2.50231 0.35000 0 0 0 1 71.00 0.00 0.00 81.00 0.00 .......................................................................... O1 O 0.04823 0.50000 0.25000 3.61239 0.50000 0 0 0 5 161.00 0.00 0.00 181.00 0.00 O2 O 0.75000 0.25000 0.04333 3.09902 1.00000 0 0 0 5 0.00 0.00 171.00 191.00 0.00 .................................................................... ! Scale Factor + Domain Fractions ! Scale Dom1 Dom2 Dom3 Dom4 Dom5 Dom6 64.26 0.136345 0.196609 0.181173 0.171022 0.178615 0.136237 11.00 21.00 31.00 41.00 51.00 61.00 .......................................................................... --------------------- >> 6 February 2006: A new version of WinPLOTR for Windows, Linux and MacOS X --------------------- - A complete re-writing of WinPLOTR has been undertaken in order to make it portable to other platforms. We have included a beta version of the program WinPLOTR-2006 in the present distribution of the FullProf Suite. The windows version of WinPLOTR (written using RealWin) will continue to be distributed waiting for the availability of all the functions in the new WinPLOTR-2006 (written using Winteracter). Our purpose is to maintain a unique version of the program that can be run in all common platforms. The new WinPLOTR-2006 has been developed thanks to the stay of Oscar Baltuano (from the Peruvian Nuclear Energy Institute: IPEN) at the LLB. The new WinPLOTR-2006 has new capabilities, absent in the windows version of WinPLOTR, like the possibility of visualising 3D surfaces using OpenGL. The preferences can be totally accessed interactively and saved. For the moment no manual is available. Users of WinPLOTR will find quite easy to use this new version of the program. Some comments and hints for using the program are given below: 1: The first time the program is run, a message telling the user that no "preferences" file is available appears. The user can save the preferences (Menu: Options) concerning all the aspects of WinPLOTR-2006 either globally (in the FullProf Suite directory) or locally (in the current directory). 2: The automatic peak and background options (Menu: Calculations) must be first made "enabled" before using them. After finishing the session of peak/background search the user should made the option "disabled" in order to do other things. 3: Open data files can be done by selecting several files simultaneously provided that all of them have the same format. They can be opened in an order that is not under the program control. The user can re-arrange the files in the appropriate way by selecting "Sorting" in the menu "Profiles". Once the user is satisfied with the ordering, the list of opened files can be saved as a buffer file that can be directly read by WinPLOTR-2006 in subsequent sessions. 4: When using the OpenGL visualisation of a surface representing powder diffraction patterns as a function of temperature or whatever, the user may see some gliches or holes in the surface if the data are noisy. This can be repared by changing the scale factor along X and Z and/or using a low pass filter (both in the menu "Processing" of the OpenGL window). Changes in colour of the surface or in the orientation of the illuminating light can also be performed by selecting the appropriate menus. In some PCs with old graphic cards this option may not work properly. 5: The most common programs of the FullProf Suite can be launched from the appropriate button or menu from WinPLOTR-2006. 6: The fitting procedure is not yet implemented The MacOS X version of the whole FullProf Suite will be available soon.