Improvement of the Rigid Body refinement
option
New Features
·
The
internal coordinates of rigid body objects can now be of four different types:
0: Spherical coordinates (previous
versions, default)
1: Cartesian coordinates (KIND=1)
2: Cylindrical coordinates (KIND=2)
3: Z-matrix form (P6=4, KIND=3)
·
The
angles can be given in radians (previous versions) or in degrees
·
The
efficiency and convergence radius has been greatly improved
·
A
small program for constructing rigid body objects adequate for pasting in a PCR
file is now provided within the FullProf suite. A more sophisticated program
for handling molecular objects is being prepared.
·
Correction
of the matrix given in note of 9 October 2004. It was just an error on writing
in the document.
In the following example the
internal coordinates of atoms of the molecular fragment are given in standard
spherical coordinates (KIND=0) and all angles are given in degrees (DEG=1)
!Atom Typ x
y z B
Occ P6 THETA
PHI Spc
!
r/xc/rho the/yc/phi phi/zc/z
X0 Y0 Z0
CHI P16:SAT DEG
KIND
Ur1
C 0.00000 0.50000
0.32811 0.50000 1.00000
1.00000 31.3200 42.000
0
0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00000 0.000 0.000
0.00000 0.50000 0.32811
-12.140 0.000 1
0
0.00
0.00 0.00 0.00
0.00 0.00 0.00
...................
Ur3
N 0.14466 0.64466
0.17816 0.50000 1.00000 0 0 0 0
0.00 0.00
0.00 0.00 0.00
1.34141 121.612 45.000
0.00
0.00 0.00
.............................................
In the following example the
internal coordinates of atoms of the molecular fragment are provided in form a
of a Z-matrix (P6=4, KIND=3) and all angles are given in degrees (DEG=1). At
present the internal coordinates cannot be refined using the Z-matrix option.
Only the orientation angles (THETA,PHI,CHI) and the position of the molecular
origin (X0,Y0,Z0) are allowed to be refined when using a least-squares method.
For using this formulation in Simulated Annealing jobs there is no restriction.
!Atom Typ x
y z
B Occ P6
THETA PHI Spc
!
dist Bond-ang Torsion-ang
X0 Y0 Z0
CHI Connectiv DEG KIND
Ur1
C 0.00000 0.50000
0.32811 0.50000 1.00000
4.00000 31.3200 42.000
0
0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
0.00000 0.000 0.000
0.00000 0.50000 0.32811
-12.140 0.000 1
3
0.00
0.00 0.00 0.00
0.00 0.00 0.00
Ur2
O 0.00000 0.50000
0.59574 0.50000 1.00000 1
0 0 0
0.00 0.00
0.00 0.00 0.00
1.25491 0.000 0.000
0.00
0.00 0.00
Ur3
N 0.14466 0.64466
0.17816 0.50000 1.00000 1
2 0 0
0.00
0.00 0.00 0.00
0.00
1.34141 121.612 0.000
0.00
0.00 0.00
Ur4
N -0.14466 0.35534
0.17816 0.50000 1.00000 1
2 3 0
0.00 0.00
0.00 0.00 0.00
1.34141 121.612 180.000
0.00
0.00 0.00
Ur5
H 0.14261 0.64261 -0.03550 0.50000
1.00000 3 1
2 0
0.00 0.00
0.00 0.00 0.00
1.00198 120.686 180.000
0.00
0.00 0.00
New version of FullProf and new version of
FP_Studio
New Features
·
Few
bugs in the output of the ATM and FST files have been corrected. The new way
describing helical magnetic structures have been tested positively in several
cases. See note of 9 October for details. The current version of FullProf has
been updated to:
**********************************************************
** PROGRAM FullProf.2k (Version 3.00 - Nov2004-LLB JRC) **
**********************************************************
·
Few
changes in WinPLOTR to handle the new version of DICVOL04 that is now
distributed within the FullProf Suite. See WinPLOTR.new or Winplotr_news.htm in
the DOCS directory of the FullProf Suite.
·
A
new version of Fp-Studio, allowing a better interactivity, is available. In the
forthcoming months new versions will be produced improving and adding new
functionalities. The input file concerning the magnetic part has been changed (it
is not compatible with the previous version). The current version of FullProf
produces always an output for Fp-Studio, even if it is not explicitly asked by
the user, according with the new input. In the current version only a single
magnetic block per file is allowed but several propagation vectors can be
grouped into a single magnetic block. The Fourier components are now given in a
separate line from MATOM. Depending of the way the user describes the magnetic
structure,in the case of several propagation vectors, the produced FST file may
have to be changed manually in order to represent the real structure.
The new interface has been written
using the Winteracter library. There are still some bugs (e.g. disappearance of
the image after the output of a BMP file) that will be corrected in future
releases. The functionalities that are not completely available in the
interface can be manually set by editing the FST file via a button in the
interface.
A manual of FullProf Studio is
available in the "docs" subdirectory of the FullProf Suite directory.
For details consult the manual. Here we give the most relevant changes with
respect to the alpha-version.
·
New
keyword
·
The
new keyword BKG followed by a legal color value and to be put in the line with
the name of the phase is now used
·
The
magnetic atoms are now given without the Fourier coefficients in the same line.
The new keyword GROUP is written in order to calculate the total magnetic
moment when the structure has several propagation vectors. If one prefers to
represent the arrows corresponding to each propagation vector, the keyword
GROUP should be removed.
·
The
Fourier coefficients of the magnetic structure are written just following the
MATOM lines in the following format:
SKP
n1 n2 Rx
Ry Rz Ix
Iy Iz MPhase
optional keywords ....
Where SKP is the keyword introducing
the Fourier coefficients. The integers n1 and n2 correspond to the number of
the propagation vector in the block and the number of magnetic matrices to be
applied (same meaning as in FullProf). The numerical values Rx,Ry,Rz,Ix,Iy,Iz
and MPhase correspond to the following expression of the Fourier coefficients
Sk = 0.5 { (Rx,Ry,Rz) + i (Ix,Iy,Iz)
} exp{-2 pi i Mphas}
when k is not equivalent to -k (so
both terms Sk and Sk* are included in the sum)If k is equivalent to -k (a
single term) then Sk=M= (Rx,Ry,Rz) and all I=0, Mphas=0
New multi-helical magnetic structure description
New Features
·
A
new way for describing helical magnetic structures in real space has been introduced
in FullProf. For accessing this option one has to put JBT=-1 (or JBT=-10) and
HEL=2 (option accessible when More=1). The list of refinable magnetic
parameters per atom is the following: Mr, Mi, Chi, Phi, Theta and Mphas. These
parameters are used to calculate the Fourier coefficients corresponding to a
helix as:
Sk = 0.5 ( Mr u + i Mi v)
exp{-2pi i Mphas}
where u and v are perpendicular unit
vectors. Together with the vector w = u x v they form an orthogonal system that
is associated with the magnetic atom. The orientation of this system with
respect to the Cartesian frame bound to the unit cell is determined by the
Euler angles (Chi, Phi, Theta) defined in the following way:
In the starting position the
cartesian frame (u,v,w) coincides with the crystallographic cartesian frame
(e1//a, e2 in the a-b plane and e3= e1 x e2). First an active rotation Chi
around the e3 axis is applied, then an active rotation Theta around the e2 axis
and finally an active rotation Phi around e3. The total rotation matrix is
R(Phi,Theta,Chi) = R(e3,Phi)
R(e2,Theta) R(e3,Chi) = [[ u, v, w]]
(Active rotations correspond to the
transpose of the rotation matrix relating coordinates of a same point in two
rotated reference frame)
The columns of the active rotation
matrix are the components of the unitary vectors u,v,w.
u = ( cosPhi cosTheta cosChi - sinPhi sinChi, sinPhi cosTheta cosChi+cosPhi sinChi, -sinTheta cosChi)
v = (-sinPhi cosChi - cosPhi cosTheta sinChi, cosPhi cosChi -sinPhi cosTheta sinChi, sinTheta sinChi)
w = ( cosPhi sinTheta, sinPhi sinTheta, cosTheta)
Notice that the refinable angles
(Phi,Theta) correspond to the spherical angles of the helix axis (w). The Chi
angle represent a rotation of (u,v) around the w-axis. The Chi angle is 100%
correlated to the phase MPhas, so they should not be refined simultaneously. If
one wants to described a circular envelope for the helix the two magnetic
moments Mr and Mi should be constrained to have the same value. The order in
which the refinable parameters are as given in the following ficticious
examples
Jbt=-1
------
!
!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 0 1
!
!Jvi Jdi Hel Sol Mom Ter Brind
RMua RMub RMuc
Jtyp Nsp_Ref Ph_Shift
3 0 2
0 0 0
1.0000 1.0000 0.0000
0.0000 1 0
0
!
P -1 <--Space group symbol
!Nsym Cen Laue MagMat
4 1 1 1
!
SYMM x, y, z
MSYM u, v, w, 0.00
.....
!Atom Typ Mag Vek
X Y Z
Biso Occ Mr
Mi Chi
!
Phi Theta unused
beta11 beta22 beta33
MagPh
Fe
MFE3 1 0
0.12340 0.02210 0.25000 0.00000 0.50000
3.450 3.450 0.000
0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
15.000 25.000 0.000
0.000 0.000 0.000 0.00000
0.00 .00 0.00
0.00 0.00
0.00 0.00
.....
Jbt=-10
-------
....
!Nat Dis Ang Pr1 Pr2 Pr3 Jbt Irf Isy
Str Furth ATZ Nvk Npr More
3 0 0 0.0 0.0 1.0 -10 4 -1 0
0 492.121 -1
0 1
!
!Jvi Jdi Hel Sol Mom Ter Brind
RMua RMub RMuc
Jtyp Nsp_Ref Ph_Shift
3 -1 2
0 0 0
1.0000 1.0000 0.0000
0.0000 1 0
0
!
P -1 <--Space group symbol
!Nsym Cen Laue MagMat
4 1 1 1
!
SYMM x, y, z
MSYM u, v, w, 0.00
...
!Atom Typ Mag Vek
X Y Z
Biso Occ N_type Spc/Fftype
!
Mr Mi Chi Phi
Theta unused MagPh
/ Line below:Codes
!
beta11 beta22 beta33
beta12 beta13 beta23
/ Line below:Codes
Fe
MFE3 1 0
0.12340 0.02210 0.25000
0.00000 0.50000 1
0
0.00 0.00
0.00 0.00 0.00
4.46000 4.46000 0.00000
10.00000 25.00169 0.00000
0.12110 <-MagPar
0.00 0.00 0.00
0.00 .00 0.00
0.00
....
New version of FullProf
New Features
·
The
current version of FullProf is now:
**********************************************************
** PROGRAM FullProf.2k (Version 2.90 - Sep2004-LLB JRC) **
**********************************************************
·
The
cartesian system used in different parts of FullProf has been unified. The
setting selected is: "x" axis is along "a" ([100]),
"y" is in the a-b plane and "z" is perpendicular to a-b.
This corresponds to the CarType="A" in CrysFML library. This change
may produce problems if one wants to run old PCR files for rigid body
refinements in triclinic, monoclinic and hexagonal/Rhombohedral bases. Only the
rotational angles of the rigid group are affected by this change.
·
The
constraint for using spherical components of magnetic moments (or their Fourier
coefficients) to systems where the c-axis had to be perpendicular to the a-b
plane has been removed. Spherical components referred to the above cartesian
system can be used within the triclinic system.
·
The
polar angles (Theta,Phi) of the helix axis in JBT=5 are now referred to the
unified cartesian system.
Bugs
·
A
bug has been corrected concerning TOF profile calculation. A temporary statement
during a testing for accelerating calculations was not removed. The calculation
was wrong for very broad peaks due to overflow in calculation of exponentials.
Sorry! (I have taken the opportunity for optimizing a little bit the
calculation of exponentials).
Corrections of bugs and improvements
New Features
·
The
cartesian system using simulated annealing and jbt=-1 has been changed in order
to be consistent with least squares part. The "x" axis is along
"a" ([100]), "y" is in the a-b plane and "z" is
perpendicular to a-b.
·
Improvements
in the calculation of simulated single crystal diffraction patterns. IRF can
now take whatever value. In particular IRF=-1 for generating only satellites
reflections is now allowed.
·
Changes
in WinPLOTR
o
A
Fourier based smoothing procedure has been added.
o
The
output for the forthcoming DICVOL 2004 program (Boultiff and Louer) has been
prepared.
o
A
buttom to access FP_Studio has been added. If the word fp_studio is added into
the WinPLOTR.set file the program is automatically lauched after running
FullProf to displaye the first phase.
Bugs
·
A
bug concerning the calculation of magnetic intensities (incommensurate
structures)using simulated annealing when jbt=-1 and isy=1,-1 has been
corrected. The error affected the phase of reflections with propagation vector
"-k" when the Fourier coefficients have imaginary components. The
calculation using least squares was correct.
FullProf Studio: Visualising crystal and
magnetic structures
New Features
·
An
alpha version of the FullProf Studio program will be distributed with the
forthcoming version of the FullProf Suite. The program has been written by
Laurent Chapon (ISIS, RAL) and it is based in the CrysFGL and CrysFML Fortran
95 crystallographic libraries. This is the result of an informal collaboration
between LLB(
RUNNING FullProf Studio
o
For
the moment the program runs only on Windows platforms. It will be ported to
Linux as soon as we have time.
o
The
program can be invoked from a DOS shell typing "fp_studio"
o
The
program asks for the name of the input file, normally "codfefiln.fst"
where "codefil" is the code of a PCR file and "n" is the
ordinal number of the phase within the PCR file. The program can also be
invoked with an argument as:
My_prompt> fp_studio codfefiln <cr>
o
If
there is no error the program opens a window with a plot of the structure that
can be rotated with the help of the mouse. At present, the only way to save an
image is by using the "prnt scrn" key and pasting it in a windows
application (Power Point, Word, MSpaint, etc...). We shall try to export
post-script and jpeg files as soon as wee can.
The input file (*.fst) for FullProf
Studio can be generated from FullProf using Jview=3 (for all phases to be
plotted) and the following directives:
o
If
a nuclear part is related to one or several magnetic phases the keywords
"magphn" must appear in the name of the phase. The final symbol
"n" should be substituted by the numeral (integer) representing a
magnetic phase related to the current crystallographic phase (eg. My_phase_name
magph2 magph3. This tells to the program to associate the magnetic phases 2 and
3 to the current crystallographic phase)
o
The
program generates automatically severall keywords (see below), but additional
plotting keywords can be added at the end of the atoms lines. To start the
plotting keywords the symbol "#" is used. For instance, the directive
"# RADIUS 0.8 COLOR 1 0.2 0.2
1 BOND Cu1 Cu 0.0 2.3" added in the
same line of an atom at the end of the normal PCR line will create the
appropriate keywords in the FST file. Remember that the BOND directive must
appear after other keywords affecting the current atom. The BOND directive can
make reference to different atoms. The keywords are case insensitive but not
the label used for atoms.
There is another version of the
program, called "fp_studio_dyn", that is useful for looking
dynamically the behaviour of the structure during a refinement or a simulated
annealing run. For that FullProf has to generate a *.fst file at each
refinement (or Montecarlo) cycle, this is obtained by putting the flag Ls2 = 5
(LSQ refinement or Simulated Annealing job) in addition to Jview=3. In the case
of a Simulated annealing job the name of the *.fst file is fixed to
"simann.fst". One can run FullProf in a shell or from WinPLOTR and
go, in a DOS shell, to the directory where the current files are read/written.
Then type "fp_studio_dyn simann" to see the behaviour of the atoms
during the structure solution process. FullProf and f_studion_dyn run
simultaneously and the whole process is slower.
CONTENT of the input file (FST
file)
o
All
lines starting with ! are considered as comments
o
The
file contents a list of keywords needed to plot the structure.
o
For
plotting a crystal structure the following keywords are needed:
SPACEG
is followed by the Hermann-Mauguin
symbol of the space group given in the same format as in FullProf (e.g. SPACEG
I 41/a m d) Instead of giving the space group a list of generators is also
admissible. The keyword is then GENER followed by the symmetry operator given
in symbolic form, e.g. GENER x,-y,z+1/2. Up to 15 generators are allowed.
CELL
is followed by six real numbers (a,b,c,alpha,beta,gamma)
defining the cell parameters (e.g. CELL
4.32 4.32 8.41 90.0 90.0 90.0)
BOX
is followed by six real numbers
representing the volume of the structure to be considered for plot (xmin,xmax,ymin,ymax,zmin,zmax)
(e.g. BOX -0.15 1.15
-0.15 1.15 -1.25 1.25)
ROTAX
This keyword and the forthcoming up
to ATOM refers to the orientation view of the unit cell the first time the
program is invoked. ROTAX is followed by four real numbers. The first (ang) is
an angle in degrees and the other three represent the components of a unit
vector in cartesian coordinates around which a rotation is performed. The
orientation of the system (if no orientation keyword is given) is a view along
the c-axis with the a-axis horizontal and directed to the right.
The values of ROTAX are output in
the DOS-shell each time one changes the orientation of the view using the
mouse. The user may copy and paste these values in the FST file for further
processing.
(e.g. ROTAX
88 1.0 0.0 0.0)
VIEW
is followed by three real values
representing the vector (in cartesian components) along which the structure
will be output on the screen
(e.g. "VIEW u v w",
with u,v,w, real numbers default VIEW 0 0 1)
SPHER
Followed by two real numbers
representing the spherical angles theta and phi of the orientation axis the
same as that given in VIEW.
(e.g. SPHER
87 10 )
ROTXYZ
Followed by three real numbers
representing the rotations (in degrees) along x,y and z to be applied to the
default orientation in order to obtain the desired view. The rotation are
applied in the following order first "rotx", then "roty"
and, finally, "rotz". A point P is trasformed to point P' as: P'=
rotz(roty(rotx(P))).
If several rotation instructions are
given in the file, only the last one is applied in practice.
(e.g. ROTXYZ 88 10 0)
ATOM
This keyword is followed by the
label of the atom, the chemical symbol the fractional coordinates and,
optionally, other keywords. The additional keywords are given for plotting
purposes. At present they are: DISPLAY (default), NODISPLAY, RADIUS and COLOR
The number of ATOM keywords is not
limited.
(e.g. ATOM Cu1 CU 0.0 0.0
0.5 RADIUS 0.8 COLOR 0.8 0.8 0.1
1)
BOND
Followed by two atom labels and two
real numbers representing the distance range between the two given atoms for
making a bond between them. Additional plotting keywords may be added in the
same line. There is no limit for the number of BOND keywords
(e.g. BOND Cu1 Cu1 0 3.3 RADIUS 1.0 COLOR 1 1 1 1,
BOND Cu1 O1 0 2.4 RADIUS 0.2 COLOR 0.2 0.2 0.5 1 NODISPLAY )
o
For
plotting magnetic structures, we need in addition the definition of the
propagation vector, magnetic symmetry and Fourier coefficients of the magnetic
moment. For starting the magnetic part description a brace "{" must
appears in the first column. The magnetic description bloc finish with a line
containing "}" in the first column. Several magnetic blocks may be
defined. The content of the magnetic part is the following:
K
Followed by three real numbers
representing the components of the propagation vector with respect to the
reciprocal basis of the conventional unit cell.
(e.g. K 0.5
0.0 0.123)
LATTICE
This keyword is, normally, the
lattice symbol of the Space group.
(e.g. LATTICE I)
A block of symmetry operators
similar to that appearing in the PCR file for a magnetic phase when Isym=-1. An
example is given below.
MATOM
Similar to ATOM keyword to which the
Fourier coefficients of the magnetic moments are added just after the
fractional coordinates, real and imaginary parts along a,b,c are given between
parenthesis. The plotting keyword SCALE followed by a real value, can be added
in order to re-scale the magnetic moments. There's no limit for the number of
MATOM lines.
Here is a complete example of the
magnetic part.
{
K
0.5 0.5 0.5
LATTICE I
SYMM x,y,z
MSYM u,v,w,0.0
SYMM -x+1/2,-y,z-1/2
MSYM u,v,w,-0.25
SYMM y+3/4,-x+1/4,-z+3/4
MSYM u,v,w,0
SYMM -y+3/4,x+3/4,-z+5/4
MSYM u,v,w,0.25
MATOM Cu1 Cu 0.00 0.00 0.50 (0.4786,0) (0,0.4786) (0,0)
0.0 SCALE 3.0 COLOR 0.1 0.6 0.0 1.0
#color 1 0 1 1
}
MagDraw v1.6 (plotting and editing magnetic
structures with LAMP)
Info
·
A
new tool to visualise magnetic structures has been incorporated within the
program LAMP (Large Array Manipulation Program) developed at
The whole LAMP can be obtained from
the ILL ftp server:
ftp://ftp.ill.fr/pub/cs/lamp_runtimes/
A short account of the capabilities
of the program is given in:
ILL News 41, p 10-11, June 2004
The tool is able to represent
incommensurate magnetic structures and the propagation vector formalism is
fully implemented. MAGDRAW contains an editor in which the user can implement
by hand his(her) own model of a magnetic structure. In MagDraw v1.6 some bugs
have been corrected. Please send an e-mail didier.richard@ill.fr,
ouladdiaf@ill.fr,
Energy dispersive input
New Features
·
For
Energy Dispersive data treatment, it is not needed to give "STE1" if
the value of 2SinTheta (or 2Theta) of the detector is provided. The program
considers that 2Theta is provided, instead of 2SinTheta, if 2SinTh > 2.0.
The parameters relating the energy in KeV with the inverse of the d-spacing
(s=1/d) are as given in the expression: E(KeV) = zero + (STE1 + STE2 * s) * s.
For the first run, if STE1=0.0 and 2sinTh is given, the program calculates STE1=
hc/2sinTh in KeV*Angstrom. In the new PCR file this value is written and the
value of 2SinTh is replaced by the 2theta angle of the detector. Of course the
parameters zero,STE1 and STE2 may be refined using a standard sample with fixed
cell parameters.
Correction of integrated intensities
(external texture, ...)
New Features
·
A
multiplicative correction of integrated intensities for each phase and pattern
can be performed in Rietveld refinements, just putting IRF=2 (provided that
abs(JBT)=0,1,4,5,... but not for abs(JBT)=2,3). The program will read
reflections as for profile matching modes except that the variable representing
the intensity for those modes is now a multiplier included in the Lp-factor.
The program expects to read a file “codfiln_pat.hkl", n: number of the
phase, pat: number of the pattern, in which the first two lines are considered
as comments.
In the case of a single pattern the
name of the file does not contain "_pat". If there is no propagation
vectors the program reads (up to the end of the file) a list of reflections
containing the items: h,k,l,m,corr. Corresponding to the reflection indices,
multiplicity and correction factor for the integrated intensity. If the phase contains
propagation vectors, a line with the number of propagation vectors (nv) is
first read in the third line. Then "nv" lines containing the number
of the propagation vector and the components (nvk,vk1,vk2,vk3) is read. After
that the program reads a list of reflections containing the items: h,k,l,ivk,m,corr.
The items (h,k,l,etc) are read in
free format, so all of them have to be provided. This correction may be useful
for different purposes. In particular when a polycristalline plate of known
texture is measured.
05 May 2004
Note
about generation of reflections in magnetic refinements
New Features
·
The
versions of FullProf starting with number 2.6 behave differently than previous
ones when using the space group "L 1" (L: P,A,B,C,I,F,R) to generate
magnetic reflections. The whole set of reflections including Friedel pairs
(even with powder data) was generated with multiplicity 2. This error can be
overcome avoiding the use of "L 1" as symbol for generating
reflections and putting instead
"L -1". A protection against the use of the "L 1" symbol in
magnetic refinement has now been included.
03 May 2004
New option for anomalous scattering plots
New Features
·
If
Anm is given the value 2, two kind of anomalous scattering files are generated
to be plot with WinPLOTR. See note of
a)
Files
with extension ".pgf" containing observed and calculated resonant
patterns as a function of 2theta. The calculated non-resonant pattern transformed
to the conditions of the resonant
pattern and the difference ycal(res)-ycal(non-res) ( e.g. ycal(NE)-ycal(FFE) ).
b)
Files
with extension ".prf" containing as observed pattern the difference
of the observed resonant (NE) pattern minus a simulated non-resonant (FFE) pattern
transformed to the (profile) conditions of the resonant pattern. As calculated
pattern the files contain the corresponding calculated differences. As for
other kind of PRF-files Bragg positions, etc ... are automatically accessed
from WinPLOTR.
Correction of small bugs in FullProf and
WinPLOTR
Bugs
·
The
hkl file automatically generated in magnetic refinements at the end of a run
can now be re-read with IRF=1 in subsequent refinements, without he need of
removing some header lines. Thanks to Francois Fauth for calling my attention
to this small bug.
·
Corrected
a bug in the writing of some standard deviations of correlated parameters.
·
The
automatic treatment of sequential refinements from WinPLOTR has been improved.
In particular the restriction in the sense of counting the dat files
(ascending) has been removed. This was not a restriction of FullProf.
Correction of small bugs in BasIreps,
FullProf and WinPLOTR
Bugs
·
In
BasIreps the precision for giving propagation vectors has been relaxed in order
to take the highest symmetry point in the Brillouin Zone. For instance 0.333 is
automatically changed to 1/3, etc.
·
Protection
against bad input data in the PCR file has been reinforced.
Multipattern with anomalous dispersion
difference plots
Bugs
·
Small
bugs in the output file concerned with the information in the IRF files for
Res=8 have been corrected
New Features
·
The
current version of FullProf is now:
**********************************************************
** PROGRAM FullProf.2k (Version 2.70 - Apr2004-LLB JRC) **
**********************************************************
·
New
output files, useful when working with several x-ray diffraction patterns of
different wavelengths (anomalous scattering synchrotron data collection), have
been implemented. The files represent differences of near-edge (NE) patterns
minus the far-from-edge (FFE) pattern for both observed and calculated
patterns. Before performing the differences the profile intensities are
corrected from background and Lorentz-polarization, and the different
scattering angles are transformed (by linear interpolation) to common s=1/d
(modulus of scattering vector). The intensities of the NE patterns are
normalized to the intensities of the FFE pattern, e.g. multiplied by
norm=sum_FFE/sum_NE.
To access this
option the new flag "Anm" appears after the flag "Cor" in
the PCR file. The FFE pattern should be attributed the value Anm=-1, and
those from
which the differences have to be output should be attributed the value Anm=1.
The name of the output files has the same code of the PCR file and appended
with "_" and the numbers of the difference patterns. For example if
in the PCR file called "my_pcrfile.pcr" the flag Anm take the
values as in:
. . . . . . .
!Job Npr Nba Nex Nsc Nor Iwg Ilo Res Ste
Uni Cor Anm
0
7 39 1
4 0 0
0 0 0
0 0 -1
0
7 36 1
4 0 0
0 0 0
0 0 1
0
7 36 1
4 0 0
0 0 0
0 0 0
!File names of data(patterns) files
. . . . . . . .
the patern #1
is the FFE pattern and the pattern #2 is a NE pattern from which the user wants
to get the difference pattern. FullProf will create an anomalous scattering difference
pattern in a single file called: my_pcrfile_21.pgf. This can be visualized with
WinPLOTR using going to the
"File"
menu and selecting the item "Open .Pgf file".
Phase-dependent shifts
New ILO for multi-films
New Features
·
A
new option concerning global shifts of reflections has been included for
Bragg-Brentano and Debye-Scherrer geometry, for constant wavelength case, that may be useful in some
circumstances. Typical examples are phases contributing to the diffraction
pattern that are not in the optical center of the diffractometer, samples
formed by several polycrystalline thin films, etc.
A zero shift,
systematic cosine (SyCos, e.g. displacement) and systematic sine (SySin, e.g.
transparency) shifts depending on each individual phase
and pattern can
be now refined.
To access this
option the new flag "Ph_Shift" must be equal to "1". This
is accessible using More=1 in the traditional single pattern format for the
PCR file. For
all cases this flag appears in the same line and after the flag
"Nsp_Ref".
If Ph_Shift=1,
the program reads the parameters Zero_ph, SyCos_ph, SySin_ph and their
corresponding refinement codes appearing just after the items
concerned with
the asymmetry parameters and before the items concerned with multi-axial
preferred orientation parameters (if used).
A example is given below:
!-----------------------------------------------------------------------------
! Data for PHASE number: 3
==> Current R_Bragg for Pattern#
1: 4.01
!-----------------------------------------------------------------------------
Name: Rutile
!
!Nat Dis Ang
Pr1 Pr2 Pr3 Jbt Irf Isy Str Furth
ATZ Nvk Npr More
2
0 0 0.0 0.0 1.0 0
0 0 0
0 159.798 0 7 1
!
!Jvi Jdi Hel
Sol Mom Ter Brind RMua
RMub RMuc Jtyp
Nsp_Ref Ph_Shift
0
0 0 0
0 0 1.0000
0.0000 0.0000 0.0000
0 0 1
!
P 42/m n m <--Space group symbol
.................................................................
! a
b c alpha beta
gamma
4.594301
4.594301 2.960153 90.000000
90.000000 90.000000
111.00000
111.00000 121.00000 0.00000
0.00000 0.00000
! Pref1
Pref2 Asy1 Asy2
Asy3 Asy4 S_L
D_L
0.00000
0.00000 0.00000 0.00000
0.00000 0.00000 0.03300
0.03100
0.00 0.00
0.00 0.00 0.00
0.00 0.00 0.00
! Zero_ph SyCos_ph SySin_ph
0.027186 -0.000917 0.000000
31.000000 251.000000 0.000000
!-------------------------------------------------------------------------------
.........................
In the case of
a multipattern-like PCR file the part that is different from the above example
is:
!-------------------------------------------------------------------------------
! Data for PHASE number: 3 ==> Current R_Bragg for Pattern# 1:
4.04
!-------------------------------------------------------------------------------
Name: Rutile
!
!Nat Dis Ang
Jbt Isy Str Furth ATZ Nvk More
2
0 0 0
0 0 0
159.7980 0 0
!Contributions
(0/1) of this phase to the 1 patterns
1
!Irf Npr
Jtyp Nsp_Ref Ph_Shift for Pattern# 1
0
7 0 0
1
! Pr1 Pr2
Pr3 Brind. Rmua
Rmub Rmuc for Pattern# 1
0.000
0.000 1.000 1.000
0.000 0.000 0.000
!
P 42/m n m <--Space group symbol
......................................................
·
A
new ILO value has been included and may be used together with the new shift
options of the above paragraph. If ILO=5 it is supposed that the symmetric Bragg-Brentano
geometry is used together with a sample formed by layers of different phases. The
product of the effective linear absorption coefficient "mu" by the
thickness "t" for each phase is given in the variable RMUA (that is
also used for other purposes, see manual). A fixed absorption correction is
applied to each phase. Notice that the order of phases in the PCR file must be
the same as that of the sample starting with the phase in the surface exposed
to the X-ray beam. The multiplicative correction applied to the integrated
intensities of the phase "p" is given by the expression:
Corr(p)=
Product(i=1,p-1){exp(-2*Rmua(i)/sinTh)}*(1-exp(-2*Rmua(p)/sinTh)/(2*Rmua(p))
·
A
modification in the generation of reflections with propagation vectors has been
performed in order to simplify the cases with centred lattices and integer propagation vectors. If the
space group for generation was "L -1" (L: lattice symbol) everything
was O.K., however some repetitions
were observed using higher symmetry space groups when L /= P and
k=integer in the list. Using a higher symmetry group (the propagation vector
group) all reflections are now generated with the proper multiplicity. Remember
that the propagation vector group is the higher configurational symmetry of a
magnetic structure. The real magnetic structure may have lower symmetry and
then the space group for generating the reflections should have the same
symmetry. Using the triclinic "L -1" for generating the whole set of
reflections is safer in the first stages of solving the structure.
Correction of bugs in WinPLOTR, GFourier
Bugs
·
New
versions of WinPLOTR, GFourier and Fourier have been included in the FullProf
Suite. Several small bugs have been corrected. In particular the editor of
GFourier (version 4.00) was not working properly: the user had to modify by
hand the *.inp file for changing the type of Fourier map. The extra peaks found
in the peak search procedure are now visualized in maps when this is asked
clicking on the "Active atom representation" button.
·
Some
changes have also been performed in Fourier concerned with the merging of peaks
and atom/peaks distances.
·
Other
small bugs in latests versions of WinPLOTR have been corrected. In particular
the manual changes of plot styles, pen width, sized of symbols, etc ... are now
working properly.
Output for Fourier using twinned single
crystals
New Features
·
The
current version of FullProf is now:
**********************************************************
** PROGRAM FullProf.2k
(Version 2.60 - Mar2004-LLB JRC) **
**********************************************************
·
Fourier
files for the case of twinned single crystals and mutiple data collection data
(up to 24 scale factors per phase) are now generated using Fou=5. The Fourier
file can be read by Fourier and Gfourier programs.
An additional
file of extension *.mrg is generated to check the details of the reflection
merging. In order to unscramble the original data, containing observations with
different domain contributions, I have used a method similar to that used in
powder diffraction to get the "observed" integrated intensity. In
order to obtain an observed 'Fo(h)' of a particular reflection contributing to
F2obs of a cluster the following formula is used:
'Fo(h)' = Fc(h) *
sqrt(F2obs/F2cal)
where F2obs if
the experimentally determined F^2 and
F2cal = Sumk { Fc(k)^2 *
Scale(k) * Corr(k) }
The sum is
extended to all reflections contributing to the current observation. Corr(k)
includes absorption and extinction corrections. When there is only a reflection
contributing to an observation one obtains the obvious result:
'Fo(h)' = Fc(h) * sqrt(
F2obs/{Fc(h)^2* Scale(h) * Corr(h)} )
'Fo(h)' = Fo(h) = sqrt(
F2obs/{Scale(h) * Corr(h)} )
The use of
Fou=5 is only valid for CRY /= 0
(integrated intensity data alone). If one uses several artificial phases
(for instance to separate
different class
of reflections, different crystals, or for whatever other reason), it is
supposed that all phases refer to the same crystal structure.
The reflections
read for each particular phase are merged into a single Fourier file.
Minor bug in the SUM files
Bugs
·
An
error in the *.sum files, in the case of using the option Ana=1 (final analysis
of the quality of powder diffraction patterns) when multiple patterns are
present, has been corrected. The error was in the writing of the area fraction
of the different phases in different patterns.
Several cosmetic changes have
also been performed.
IMPORTANT:
NON-BACKWARD COMPATIBLE MODIFICATION
Bugs
·
In
the last version on EdPCR the values of wavelengths defined on Profile Data
Information were wrong. Now it was resolved.
New Features
·
IMPORTANT: NON-BACKWARD
COMPATIBLE MODIFICATION
An internal change to unify the description of
magnetic refinements for powder and single crystal cases has been performed.
The
correction concerns mostly the people working in single crystal diffraction,
however there is an additional modification that changes the way of treating
the occupation numbers of a magnetic phase in the case of centred lattices.
Now in all cases the nuclear and magnetic
structure factors are calculated for the content of the conventional unit cell,
that is the old (internal) structure factor was calculated only for the
asymmetric part of the unit cell ( ,
no lattice centring, no centre of symmetry) and only after that calculation the
square of the structure factor was multiplied by to obtain the conventional structure factors .
Where
NLAT: Number of lattice points per cell (1 for
P; 2 for A,B,C,I; 3 for R and 4 for F lattices),
ICEN=1 for non-centro and ICEN=2 for
centrosymmetric case.
The
multiplication of by N was performed in different
places outside the structure factor procedures. See for instance note on
extinction correction of 8 April
2002.
The factor N was not applied to the single
crystal refinements, so that the scale factor was higher because was used instead of .
This obliged to use JBT=10 for doing magnetic refinements or to apply a rule different
from that of powder case to relate the scale factors of a nuclear and a
magnetic phase in order to obtain the correct values of magnetic moments.
Consequences
using old PCR files for single crystals
This change has as a consequence that the old
scale factor of a crystal or magnetic structure refinement using single
crystals (including JBT=10) should be reduced by a factor 1.0/N in order to be
consistent with the current formulation.
In most cases it suffices to perform a
refinement with a sufficient number of cycles and the program will find the new
scale factor straightforwardly. In more critical cases it is convenient to fix
the structural parameters and refine only the scale and extinction parameters.
Consequences
using old PCR files for powders
For POWDER DIFFRACTION there is no consequence
EXCEPT IN THE CASE OF CENTRED CELLS when a magnetic phase has been defined
using a centred symbol for generating the magnetic reflections. Now the number
of lattice points per cell is taken into account for calculating the final
magnetic interaction vector squared .
Up to now if one wanted to calculate the magnetic contribution using the same
scale factor than that of the corresponding nuclear counterpart, it was
necessary to take into account (if one was using the content of a primitive
cell) a multiplication of the occupation factors by NLAT (ICEN was already
included in the magnetic structure factor calculation) or give all the atoms in
the conventional cell. This is no more necessary.
The examples given in the "examples"
subdirectory of the FullProf Suite have been slightly modified to take into
account the new way of handling centred cells.
Unified rules
(powders + sxtals) for doing magnetic refinements
The rule to get the correct values of magnetic
moments when using a magnetic phase separated of the nuclear one (JBT /= 10) is
the following:
o
Use
of the same cell [propagation vector formalism, or k=(000)] and the same scale
factor for both the nuclear and the magnetic phases.
Nuclear phase:
M : general multiplicity of the space group
m : multiplicity of the site of a
particular atom
f : factor given by the user to get
"convenient" occupation factors
The structure factor is calculated using Oc, so
only in the case of f=1 the structure factor coincides with that of the
conventional cell in absolute units (electrons for X-rays and for neutrons).
Magnetic phase:
Mm: number of symmetry operators given to
describe the magnetic phase.
mm: number of different atoms
generated by applying the Mm symmetry operators (+ Center of symmetry if
Icen=2) to a particular atom.
Icen: Value given by the user to
tell the program if the magnetic structure is centrosymmetric (Icen=2), that
is, the centre of symmetry does not change the orientation of the magnetic
moment or non-centrosymmetric (Icen=1).
Nlat: Number of lattice points per
conventional unit cell
If the lattice is centred, the centring
translations operators should not be given among the Mm symmetry operators. In
the propagation vector formalism the calculation does not need this redundant
information. The total number of magnetic atoms of a particular site in the conventional
cell is "m", so the number to be given in the description is
Nmag=m/Nlat. Nmag is the number of magnetic atoms in a primitive cell. If the
Mm symmetry operators (+ Center of symmetry if Icen=2) are able to generate at
least all the Nmag atoms we obtain the following prescriptions.
To get proper values of magnetic moments the
occupation number, Ocm,to be used is that giving the same number of magnetic
atoms per unit volume in both the nuclear and the magnetic phase. This gives
the rule:
Ocm = Oc * M / (Mm*Icen*Nlat) (1)
If we call Oca = mm/(Mm*Icen), it is easy to
verify that
Ocm= f * Oca * m /(mm*Nlat) = f * Oca * Nmag /mm (2)
The formula (1) is the most general and easy to apply.
It assumes implicitly that, for each magnetic atom site in the asymmetric
crystallographic unit, there is only one magnetic atom in the asymmetric
magnetic unit.
If the user prefers to describe individually all the
"Nmag" magnetic atoms (sublattices) corresponding to a single atom
site (crystallographic
Ocm = Oc * M / (Mm*Icen*Nlat) = Oc *
M / (Nmag*Icen*Nlat)
Ocm = Oc * M / m = f *
m/M * M/ m = f * 1 = f
·
If
one uses the magnetic unit cell (not recommended, the propagation vector
formalism is of general applicability!), then use a P symbol for generating the
magnetic reflections (in order to fix Nlat=1). In this case the same
prescriptions as above apply except that the scale factor of the magnetic phase
is related to that of the crystallographic part by
Sc: crystallographic scale factor, Vc: Unit
cell volume
Sm: magnetic scale factor, Vm: Magnetic unit
cell volume
In the case of large magnetic cells it can be
more convenient to modify the occupation numbers of magnetic atoms in such a
way that the two scale factors coincide. The occupation factors of the magnetic
part should then be multiplied by the factor: Vc/Vm.
New files for
Single Xtal Laue/TOF refinements
Bug in the Pearson
VII function
Bugs
·
An
error in the Pearson VII function when the exponent was close to 1 has been
corrected. This caused instabilities in refinement. The lowest m-value accepted
by the program is m=0.6
New Features
·
The
number of scale factors to be used in the refinement of single crystal data has
been augmented up to 24. A new format adapted to data taken with different
wavelengths (Laue or TOF) has been implemented. The JOB variable should be
given the value JOB=-1. The format of the intensity file is the following:
o
Firt
line is considered as a title
o
Whatever
number of comment lines (starting with ! or #) can follow the title
o
A
line with the format item is searched by the program by looking for balanced
parentheses (starting with "(" and finishing with ")" ...
without quotes). An example is: (3i4,2f14.4,i3,4f10.4)
o
After
the format item the program tries to read the number (nvk) of propagation
vectors that must coincide with what is given in the PCR file. And immediately
the program reads the propagation vectors themselves in the following nvk
lines.
o
Finally
the program read lines using the format item from which it extracts the
following items: h,k,l, F2, sigma(F2), cod, lambda, 2theta, absorption, Tbar.
In the case of propagation vectors an additional integer following hkl is read.
·
The
meaning of Cod (integer) is the number of the scale factor to be applied for
treating the current observation. It may be used for treating twinning or to
combine reflections obtained in different circumstances (e.g. different banks with
non-normalized data)
·
The
program uses the additional information to calculate wavelength-dependent extinction
corrections (four different models are available).
·
For
using more than 6 scale factors, the user must add the number of scale factors
at the end of the same line where they are provided. Add lines (up to 6 scale
factors/line plus the line of refinement codes) with the additional scale
factors.
Bug (2 days in the
web)
Bugs
·
Sorry,
during few hours the version in the CEA site was wrong for profile matching
modes using two wavelengths!
IRF files. DLIM in
TOF. Changes in profile matching modes
Generation of *.int
files for TOF
New files of
extension *.seq putting Rpa=-2 in sequential refinements
New Features
Several
changes have been performed in the current version of FullProf.
·
The
keyword "WAVE" followed by three real numbers can now be included in
IRF files corresponding to Res=4, before giving the number of points to be
read.
Example (header of the file):
My resolution file: diffractometer with Cu ka1-ka2
!This is a comment: File prepared using LaB6
and WinPLOTR
WAVE 1.5406 1.5444 0.497
25
...........
·
Another
IRF format is available. Putting Res=8 a similar file to that described for
Res=4 is read. The only difference is that in each line the asymmetry
parameters S_L,D_L due to axial divergence (L. Finger, et al.) are also read.
In principle all the asymmetry parameters should have the same values but this
has been included in each line to take into account other possible causes of
asymmetry. Then the lines read with Res=4 have as items "2theta HG
HL" and those read with Res=8 are of the form "2theta
HG HL S_L
D_L".
·
In
the IRF files for Res=5 (T.O.F.) the keyword "NPROF" has been
included. It tells the program what kind of profile function has been used to
create the IRF file itself. NPROF is followed by an integer that can be 9 for
the back-to-back exponentials or 13 for
the Ikeda-Carpenter function convoluted with a pseudo-Voigt. In the lines
starting with the keyword D2TOF a third item (in addition to Dtt1_i and Dtt2_i)
can be read representing the zero point of the bank in micro-seconds.
·
Some
tuning of special variables controlling the range in different banks of TOF
multi-pattern refinement has been performed. This concerns mostly the cutting
at low d-spacing and the way of generating satellite reflections at low Q (high
d-spacing) for propagation vectors near the zone boundary. The range of TOF in
the instrumental resolution files has a higher priority than the effective
range read from the input data file.
·
The
generation of *.int files for simulated annealing in TOF has been totally
revised. Instead of storing sums of integrated intensities in a cluster
(j.Lp.C.F2), we use just sums of F2 (without multiplicities and Lorentz
factors) in order to get the same units for data coming from different banks.
The reflections in *.int files, for each bank, are re-ordered in order to put
on top the most important ones (those at low Q).
·
For
sequential refinements it is now possible to generate a file of extension
".seq", putting Rpa=-2. This file is better than the old
".rpa" file because it contains the summary of effectively refined
parameters together with their symbolic
names. In the console version of FullProf ("fp2k.exe" or
"fp2k") a sequential refinement can be performed by using the
necessary information in the command line. For instance a command like:
>fp2k
cyc my_pcr datcode 2341
2680 y n
means
the following:
The program is invoked for sequential
refinements by using as the first command line argument the keyword
"cyc", the code of the PCR file that will be used for treating all
the data is given as the second argument "my_pcr". The code of the data files is the third argument
"datcode". The arguments 4 and 5 correspond to the numbering of data
files to be treated. Eg. in our example all files between
"datcode2341.dat" and "datcode2680.dat" will be treated.
The last two arguments are given to ask the program to save (y) or not (n) the
files with extension ".prf" and ".hkl" respectively.
Bugs
·
An
error was introduced in profile matching modes few month ago. This cause
re-starting the refinement with a bad fit in some circumstances and a slowing down
of the convergence. I have taken the oportunity for changing the output values.
Now, in all cases, what is updated in each cycle of a profile matching refinent
is the square of the structure factor not the observed integrated intensity.
Depending of what kind of refinement is being done the output could be the
structure factor itself (Jbt=-2) or the integrated intensity (j.Lp.C.F2 for
Jbt=2). If one wants to use correction factors stored in C (preferred
orientation, absorption, etc ... but not extinction!) it must be taken into
account that the output j.Lp.C.F2 (Jbt=2) contains everyting. The integrated
intensity will be independent of what corrections we are using, for different
corrections the extracted structure factors will be different in the files
obtained with Jbt=-2. The format of some output files controled by the flag
"Hkl" has also been changed in order to cope with very different
magnitudes.
New versions of EdPCR,Fourier,
GFourier and WinPLOTR.
Fitting procedure
in WinPLOTR with axial divergence asymmetry
New Features
·
The
current version of the program FullProf is 2.55 Jan2004
·
New
versions of EdPCR, Fourier, GFourier and WinPLOTR have been included in the new
release of the FullProf Suite. The suite is also distributed for Linux. A
program GProfile/WinPLOTR for Linux has been written using the Winteracter library
for Linux. This program has not all the capabilities that has WinPLOTR for
Windows, however it may satisfy most of the requests of Linux users.
·
An
additional subfolder/subdirectory containing instrumental resolution files of
several instruments (TOF machines at ISIS, powder diffractometers at
·
Fourier/GFourier
provides now, within the peak search procedure distances between new peaks and
the atoms included in the model, as well as the inter-peaks distances and
angles.
·
The
fitting procedure within WinPLOTR includes now the axial divergence treatment
of asymmetry for constant wavelength. Now it is possible to refine individual
peaks in TOF powder diffraction patterns (See WinPLOTR.new for details).
Bugs
·
Several
corrections of small bugs in output files have been introduced. In particular
for the anomalous scattering in X-rays the output file for GFourier was
neglecting Df". The default output file for GFourier includes now more instructions.
·
EdPCR
has been improved and many small bugs corrected.
·
Now
Gfourier create a plot file into Isosurface mode.