workshop jana2006 powder option in jana2006 lecturer jan rohlíček authors václav pet ř í č ek,...
TRANSCRIPT
Workshop Jana2006
Powder option in Jana2006
LecturerJan Rohlíček
AuthorsVáclav Petříček, Michal Dušek andLukáš PalatinusInstitute of Physics ASCRPraha, Czech Republic
Workshop Jana2006
Single crystal diffraction
Powder diffraction pattern is a combination of a large number single crystal diffraction patterns
Diffracted pattern
Workshop Jana2006
Powder diffraction pattern of alaptide powder on the area detector
Diffracted pattern
Workshop Jana2006
Problems and limitations
• Peak overlap – problem of powder data.
What is the intensity of overlapped reflections?
2 Theta
Workshop Jana2006
Overview of the functionality for powdersJana2006 is a refinement program originally developed for single crystal data. Except for the refinement of basic structural parameters it allows also:
ADP parameters till 6th order – ionic conductors, positional disorder
e3=0.000
-3.0 -1.0 1.0 3.0e1-2.5
-1.5
-0.5
0.5
1.5
2.5
e2
Charge density studies - multipole refinement
Refinement of modulated and composite structures
E.Gaudin, F.Boucher, V.Petříček, F.Taulelle and M.Evain, Acta Cryst. (2000), B56, 402-408.
Melilite – Bindi L., Bonazzi P., Dušek M., V.Petříček and Chapuis G., (2001), Acta
Cryst., B57, 739-746.
Workshop Jana2006
Overview of the functionality for powders
magnetic structures
Jana2006 is a refinement program originally developed for single crystal data. Except for the refinement of basic structural parameters it allows also:
Workshop Jana2006
Powder data of one or more phases
unit cell parameters, dimension
Le Bail refinement
Space group test
Structure determination (Superflip, Expo)
Rietveld refinement
JANA2006
Overview of the functionality for powders
Workshop Jana2006
Le Bail refinementSeparate intensities from powder pattern
in Jana2006 called as „Make only profile matching“
Structure solutionUse separated intensities to solve the
crystal structure
Rietveld refinementRefine the model from the structure
solution process by using powder data
Overview of the functionality for powders
Workshop Jana2006
The powder diffraction option (Rietveld refinement + leBail technique) was implemented in 2001: Dušek,M., Petříček,V., Wunschel,M., Dinnebier,R.,E. & Van Smaalen,S. (2001). J. Appl. Cryst. 34, 398-404.
„Edit Profile“:
Profile functions: Gauss, Lorentz and pseudo-Voigt
Background corrections: Legendre, Chebyshev polynomials, “cos” functions
Peak asymmetry: Simpson, Berar-Baldinozzi, Finger-Cox-Jephcoat
fundamental approach (Cheary-Coelho)
Preferred orientation: March-Dollase, Sasa-Uda
Roughness: Pitchke-Hermann-Matter, Suorti
Anisotropic line-broadening according to Stephens modified by Leineweber and Petricek to include modulated phases
„Profile Viewer“:
Edit manual background
checking calculated and measured data
Powder options in Jana2006
Workshop Jana2006
Gaussian:
where is Full-Width at Half-Maximum
22 2/exp2
1, G
GG bx
bxbG
GH
2ln8
22 GG
Hb
Lorentzian: 221
12,
LLL
bxbxbL
LL Hb
where is Full-Width at Half-MaximumLH
Powder profile parameters in Jana2006
Workshop Jana2006
Voigt function:
is a convolution of Gaussian and Lorentzian function. For powder profile we use a simpler analytical approximation of the Voigt function called pseudo-Voigt function:
xHGxHLxHpV ,1,,
543223455 07842.047163.442843.269269.2 LLGLGLGLGG HHHHHHHHHHH
the parameters and are functions of and H GH LH
32
11116.047719.036603.1
H
H
H
H
H
H LLL
xdxxbLxbGxbbV LGLG ,,,,
Powder profile parameters in Jana2006
Workshop Jana2006
For the Gaussian term the formula Cagliotti, Pauletti & Ricci, 1958 (Nucl.Instrum., 3, 223) is used:
2
22
costantan
PWVUbG
In the original formula only three first terms were used. The last one was introduced later as a Scherrer term and it is connected with crystalline size. But from the fact that:
2
2
cos
11tan
follows that only two of three coefficients can be refined simultaneously.
For the Lorentzian part we use the same terms as for pure crystal broadening but, as mentioned above, they cannot be used directly to find sample characteristics.
PWU ,,
tancos LLL YXb
Powder profile parameters in Jana2006
Workshop Jana2006
For modulated structures this method has been generalized by A. Leineweber and V. Petříček, (2007). J. Appl. Cryst., 40, 1027-1034.
Stephens model for anisotropic broadening
Workshop Jana2006
Simpson’s method – Peak is combined with several shifted peaks having the identical shape
Berar-Baldinozzi method - based on Hermite polynomials
By axial divergence – according to Finger, Cox and Jephcoat, (1994) J.Appl.Cryst. 27, 892-900.
Two parameters are used: “height” and “sample”
These parameters are strongly correlated. Our recommendation is to estimate their ratio and keep it as a restriction during the refinement.
LH LS
Fundamental approach – it follows the method introduced by Cheary and Coelho, (1998), J. Appl. Cryst. 31, 851-861. This method can estimate the profile asymmetry just on the base of experimental parameters. For Bragg-Brentano geometry it works very nicely.
Asymmetry options in Jana2006
Workshop Jana2006
Background correction
Several possibilities of description: Legendre polynomials, Chebyshev polynomials, Cos-ortho, cos-GSAS
Manual background – the background is expressed as a set of background intensities over the diffraction interval. The actual value is calculated by a linear interpolation. This method is can very effectively describe even very complicated background profiles. But it need some user assistance to select it properly. Moreover this first background estimation can be combined with some of previous continuous functions.
Shift parametersShift – it defines the zero shift (again in units of 0.01 deg). This value is to be added to the
theoretical peak position to get a position in experimental profile
sycos – in analogy with the Fullprof:is connected with a specimen displacement
sysin – in analogy with the Fullprof:is connected with a transparency correction
Background, shift, displacement
Workshop Jana2006
In Jana2006 is also used to predict symmetry by comparing of profile fits for different space groups:
Centering and space group test
Workshop Jana2006
Practical notes•Refining of many profile parameters together causes endless process, which is not
converging. Note that only two of U, W and P can be refined simultaneously, also S/L and H/L cannot be refined together.
•It is true that using huge number of background points slightly improve agreement profile factors. The reasonable number of background points is up to 10 for flat sample (without amorphous content) and up to 30 for capillary samples (with amorphous content).
•Manual background makes Le Bail refinement more stable.
Workshop Jana2006
•It is good idea to start Rietveld refinement with disabled unit cell and profile parameters which you got by previous Le Bail refinement. Enable these parameters after several cycles, when you are sure that your model is close enough to the correct position.
•Bond and angle restraints are very powerful but also very dangerous tools.
•Using small damping factor e.g. 0.1 at the beginning of the Rietveld refinement can protect you from repeating the last step again.
•Make backup copies
Practical notes