Difference between revisions of "GSAS Profile Terms"

From Ug11bm
Jump to navigationJump to search
 
(50 intermediate revisions by the same user not shown)
Line 1: Line 1:
== Peaks Profile terms for Rietveld Analysis==
GSAS offers 5 different Constant Wavelength (CW) X-ray profile functions.  They are described in detail within the GSAS technical manual (see page 156).  11-BM users are encouraged to use either profile type 3 or type 4.
Most Rietveld refinement programs use a pseudo-Voigt term combining Gaussian and Lorentzian peak shapes (plus other correction terms).  In general, the X-ray source can be described by a Gaussian function, while sample effects are described by Lorentzian terms.


For the synchrotron powder XRD data from 11-BM, the instrumental resolution is well described by Gaussian terms. Sample effects, such as size and microstrain broadening (i.e. local variations in the lattice parameters) are usually fit best by Lorentzian terms.
A quick reference guide to terms in the CW profile type 3 and 4 functions is given below after a brief introduction. Users are ''strongly'' encouraged to read this section of the GSAS manual at least once before (or after) blindly using this guide!  The [http://www.aps.anl.gov/Xray_Science_Division/Powder_Diffraction_Crystallography/ Powder Diffraction Crystallography Video lectures] are also a great resource for both beginners and experts.


Please consult other references (such as the GSAS maunal) for details on Rietveld profile functions.
== Pseudo-Voigt Profile Functions ==


== Fitting 11-BM Data ==
These peak profile functions are a [http://en.wikipedia.org/wiki/Voigt_profile pseudo-Voigt] type, combining Gaussian (G) and Lorentzian (L) components.


Representative LaB6 data for 11-BM (high resolution powder XRD) can be [http://11bm.xor.aps.anl.gov/standards_data.html downloaded from the 11-BM webpage] here (pick your format):
The general Gaussian shape (as a function of angle θ) is described by the Cagliotti function


   
Gaussian Profile &asymp; U*tan<sup>2</sup>&theta; + V*tan&theta; + W + P/cos<sup>2</sup>&theta;


more details about this specific LaB6 dataset follow:
These U, V, W, and P variables match the GU, GV, GW, and GP profile terms you see below.


precise wavelength = 0.412235
The Lorentzian shape is more complex (check the manual), but includes size and strain broadening terms.  
data was collected on a spinning 0.8 mm diameter capillary of LaB6 660a  [NIST SRM certificate lattice value = 4.15691(1) A] .  Estimated muR is ~ 1.0
collection Temp: 295 K
2theta range: 0.5 deg - 50.0 deg
step size: 0.001 deg


My GSAS/EXPGUI refinement gives a good fit (image attached) with the following parameters:
== Asymmetry, Zero-Shift and Related Terms ==
=== Asymmetry ===
[http://youtu.be/SIz6Ng6UzAw Axial Divergence] (i.e. low angle peak asymmetry) is modeled in GSAS profile types 3 & 4 with the Finger-Cox-Jephcoat model (see GSAS manual). The profile terms ''S/L'' & ''H/L'' describe the intrinsic instrument asymmetry for low angle (< 5 deg theta for 30 Kev 11-BM data), and is related to the goniometer radius, or distance from the sample to the detector (about ~1000 mm for 11-BM).


SG: Pm-3m
Most 11-BM users will not need to refine these terms from the default (1.2E-03) values included in 11-BM GSAS instrument parameter files. If your fit shows an asymmetric misfit to strong low angle peaks, my might try to refine, but never refine both simultaneously.
a =  4.156917(1)
zero shift:  -0.00029 deg 2theta
La  @ 0, 0, 0 (Ui/Ue*100 =  0.62)
B  @  0.1984(1), 1/2, 1/2  (Ui/Ue*100 =  0.29)


GSAS Profile type 4:  (non-listed terms = 0.0)
=== Transparency ===
Coeff.  :  GU        GV        GW            LX        S/L          H/L
This is the '''trns''' profile term describing the x-ray penetration depth into a flat plate sample . It should *only* be used for fitting flat-plate reflection geometry (Bragg-Brentano) powder diffraction dataDo '''*not*''' use for 11-BM data which is collected in a transmission geometry (Debye-Scherrer)
Value    :  2.552E+00 -5.439E-01  5.990E-02  2.790E-01 1.2E-03  1.2E-03


  4-term Shifted Chebyschev (type #1) background
=== Displacement ===
This is the '''shft''' profile term describing vertical displacement of the flat plate sample . It should *only* be used for fitting flat-plate reflection geometry (Bragg-Brentano) powder diffraction data. Do '''*not*''' use for 11-BM data which is collected in a transmission geometry (Debye-Scherrer)


wRp = 0.0645,    Rp = 0.0486
=== ZERO ===
  CHI**2 =  3.349    for  14 variables                   
Instead of '''trns''' or '''shft''', 11-BM users should refine the 2&theta zero shift term "ZERO". This term is found in EXPGUI under the Histogram tabIt describes any shift (error) in the absolute 2theta "0" position for the data and is measured in units of centi-degrees (100*2&theta;).


Let me know if you need any additional information.
GSAS users refining 11-BM data should refine this term near the end of a refinement to ensure the most accurate lattice parameters. Typical values for 11-BM might be in the range +/- 0.1 (i.e. 0.001 degrees).


== Background Function ==


The background function type 1 (Chebyshev polynomial) is recommended.  Use the minimum number of terms needed to fit the pattern background.  Start with 3 or 4 terms, and increase as needed.  Sometimes 10-20 tems are needed for complex backgrounds.


Just remember that the background also represents scattering from something (sample holders, air, sample?).  Try to think about what may contributes to the background for your sample.  This might contain valuable information.




==Conversion of pseudo-Voigt function terms==
==Suggested Profile Types & Terms for Fitting 11-BM Data==


11-BM users are encouraged to use the GSAS constant wavelength (CW) profile type 3 or type 4.  Profile #4 is best for cases in which anisotropic terms are required.


 
11-BM users will '''not''' (''usually!'') need to change or refine the default 'G' terms given in the instrumental parameter file.  For the high-resolution synchrotron powder data collected at 11-BM, the instrumental resolution is well described by Gaussian terms.


On the other hand, sample effects in 11-BM data, such as size and strain broadening are (''usually!'') best fit and refined using Lorentzian terms. Gaussian crystallite size broadening (GP) is *rarely* observed; this requires a very tight mono-disperse size distribution rarely encountered in powder samples (solid metal samples may be an exception).


GSAS <-> Fullprof
Many 11-BM diffraction patterns can be well fit by refining only the LX (size), LY (strain) and anisotropic Lorentzian size & microstrain terms. Profile type 4 is recommended if the later are required.
Gaussian Parameters
GSAS Term <=> Fullprof Term : (description)
GU = 1803.4 * U : (instrumental term, ~ tan^2 of theta)
GV = 1803.4 * V : (instrumental term, ~ tan of theta)
GW = 1803.4 * W : (instrumental term, ~ constant with theta)
GP = 1803.4 * IG : (size broadening)


note: 1083.4 => centidegrees squared divided by 8*ln(2)
For more info see a EXPGUI-GSAS Parameter [http://www.aps.anl.gov/Xray_Science_Division/Powder_Diffraction_Crystallography/GSASparameters.html tutorial video]
 
 
==Constant Wavelength X-ray GSAS Profile Type 3==
{| class="wikitable"
|-
| <font color="#C11B17">'''GU''' = Gaussian U term</font>
| <font color="#C11B17">'''GV''' = Gaussian V term</font>
| <font color="#C11B17">'''GW''' = Gaussian W term</font>
|-
| <font color="#C36900">'''GP''' = Gaussian crystallite size broadening</font>
| <font color="#2C820E">'''LX''' = Lorentzian isotropic crystallite size broadening </font>
| <font color="#2C820E">'''LY''' = Lorentzian isotropic strain broadening</font>
|-
| <font color="#C36900">'''S/L''' = Axial Divergence S term</font>
| <font color="#C36900">'''H/L''' = Axial Divergence H term
| <font color="#C11B17">'''trns''' = Sample 'Transparency' </font> ''(note 1)''
|-
| <font color="#C11B17">'''shft''' = Sample 'Displacement'</font> ''(note 1)''
| <font color="#C36900">'''stec''' = Lorentzian anisotropic strain broadening</font> ''(note 2)''
| <font color="#2C820E">'''ptec''' = Lorentzian anisotropic crystallite size broadening</font>
|-
| <font color="#C36900">'''sfec''' = Lorentzian sublattice anisotropic broadening</font> ''(note 3)''
| <font color="#C36900">'''LXX''' = Anisotropic Lorentzian microstrain</font> ''(note 2)''
| <font color="#C36900">'''LYY''' = Anisotropic Lorentzian microstrain</font> ''(note 2)''
|-
|}
Color denotes indicates terms 11-BM users should 
<font color="#2C820E">'''Refine''' (green) </font>,
<font color="#C36900">'''Sometimes''' Refine (yellow) </font>, and
<font color="#C11B17">''' Not''' Refine (red) </font>
 
*''note 1: Do *not* use for 11-BM data, see above''
*''note 2: Better to use Profile Type #4 Anisotropic microstrain terms''
*''note 3: See GSAS Manual before using''
 
 
 
==Constant Wavelength X-ray GSAS Profile Type 4==
{| class="wikitable"
|-
| <font color="#C11B17">'''GU''' = Gaussian U term</font>
| <font color="#C11B17">'''GV''' = Gaussian V term</font>
| <font color="#C11B17">'''GW''' = Gaussian W term</font>
|-
| <font color="#C36900">'''GP''' = Gaussian crystallite size broadening</font>
| <font color="#2C820E">' '''LX''' = Lorentzian isotropic crystallite size broadening</font>
| <font color="#2C820E">''''ptec''' = Lorentzian anisotropic crystallite size broadening</font>
|-
| <font color="#C11B17">'''trns''' = Sample 'Transparency'</font> ''(note 1)''
| <font color="#C11B17">'''shft''' = Sample 'Displacement'</font> ''(note 1)''
| <font color="#C36900">'''sfec''' = Lorentzian sublattice anisotropic broadening</font> ''(note 2)''
|-
| <font color="#C36900">'''S/L''' = Axial Divergence S term</font>
| <font color="#C36900">'''H/L''' = Axial Divergence H term</font>
| <font color="#C36900">'''eta''' = Gaussian-Lorentzian mixing factor</font> ''(note 3)''
|-
| <font color="#2C820E">''''SXXX''' = Anisotropic microstrain broadening (Lorentzian)</font> ''(note 4)''
| <font color="#2C820E">''''SYYY''' = Anisotropic microstrain broadening (Lorentzian)</font>
| <font color="#2C820E">''''SZZZ''' = Anisotropic microstrain broadening (Lorentzian)</font>
|}
Color denotes indicates terms 11-BM users should 
<font color="#2C820E">'''Refine''' (green) </font>,
<font color="#C36900">'''Sometimes''' Refine (yellow) </font>, and
<font color="#C11B17">''' Not''' Refine (red) </font>
 
*''note 1: Do *not* use for 11-BM data, see above''
*''note 2: See GSAS Manual before using''
*''note 3: Changes pseudo-Voigt mix from pure Gaussian (eta=0) to pure Lorentzian (eta=1). Typical 11-BM data is fit well using (or at least starting with) eta = 1''
*''note 4: Number of Stephen's microstrain terms SXXX, SYYY etc changes with phase crystal symmetry (ie. a monoclinic phase will have more of these terms than a cubic symmetry phase)''
 
 
 
==Physical Meaning of Profile Terms==
 
Some of the CW profile terms can be interpreted to give physically meaningful strain and particle size information.  See the page 162 of GSAS technical manual for details
 
=== Particle size broadening ===
 
Using the isotropic '''LX''' profile term
 
then particle size = (18000*K*&lambda;)/(&pi;*LX)
 
where K = Scherrer constant (typically ~ 1), &pi; = 3.1416
 
and size units are in Angstroms (&#8491;), same units as wavelength (&lambda;), typically ~ 0.41 &#8491; for 11-BM data
 
 
=== Strain broadening ===
 
Strain is more difficult to quantify.
 
For profile type 3, using the isotropic '''LY''' term, then
 
isotropic strain (%) = 100% * LY * (&pi;/18000)
 
For profile type 4, consult the GSAS manual.  The anisotropic strain are best visualized using gnuplot.  See the Robert Von Dreele's Rietveld Mailing List post from July 28th, 2011 below:
 
Q: What is the "best" order for incorporating GSAS profile 4 anisotropy Sxxx terms into a refinement?
A: Robert Von Dreele writes: "Usually, I do the S4xx ones first & then the others.
The thing is fairly stable after that. Do all of them at once in any case in the end.
The S4xx must be > 0 but the rest can be any sign. If you want a quick way of seeing the shape,
run mustrplot in GSAS; you'll need gnuplot to see the plot.
For monoclinics, it might look like two beans back-to-back."
-Bob




Lorentzian Parameters
GSAS Term <=> Fullprof Term : (description)
LX = 100 * Y : (size broadening)
LY = 100 * X : (microstrain)


note: 100 => degrees to centidegrees
== Example 11-BM Profile Fit and Terms ==
[http://en.wikipedia.org/wiki/Lanthanum_hexaboride Lanthanum Hexaboride (LaB6)] is used an X-ray powder diffraction profile standard because of its sharp peak shape. 


Representative LaB6 data for 11-BM (high resolution powder XRD) can be downloaded from the 11-BM wiki:


Finger-Cox-Jephcoat asymmetry parameters
https://wiki-ext.aps.anl.gov/ug11bm/index.php/Standards_Data
GSAS S/L = Fullprof S_L
GSAS H/L = Fullprof D_L


Note: terms are equivalent
On the same Wiki page find more details about Instrument Parameter and Input Files, plus example fits to LaB6 data measured at 11-BM


S / L = the sample “half height”/diffractometer radius
H / L = the slit half height/diffractometer radius


==Convert GSAS profile terms to Fullprof terms==


"Typical values of Rietveld instrument profile coefficients" Kaduk J, Reid J. Powder Diffraction (2011) vol. 26 pp. 88
11-BM provides users with instrument profiles terms for GSAS. Convert these values to Fullprof profile terms using the formulas below.


For more info see: [http://dx.doi.org/10.1154/1.3548128 "Typical values of Rietveld instrument profile coefficients" Kaduk J, Reid J. Powder Diffraction (2011) vol. 26 pp. 88] 


Gaussian Parameters
GSAS Term <=> Fullprof Term : (description)
GU = 1803.4 * U : (instrumental term, ~ tan^2 of theta)
GV = 1803.4 * V : (instrumental term, ~ tan of theta)
GW = 1803.4 * W : (instrumental term, ~ constant with theta)
GP = 1803.4 * IG : (size broadening)
note: 1083.4 => centidegrees squared divided by 8*ln(2)


size and microstrain broadening (i.e. local variations in the lattice parameters) are Lorentzian for 11-BM data
Lorentzian Parameters
GSAS Term <=> Fullprof Term : (description)
LX = 100 * Y : (size broadening)
LY = 100 * X : (microstrain)
note: 100 => degrees to centidegrees


Only rarely is Gaussian size broadening observed; this re-  
Finger-Cox-Jephcoat asymmetry parameters are equivalent in GSAS & Fullprof
quires a very tight monodisperse distribution, which is rarely
GSAS S/L = Fullprof S_L
encountered in powder specimens but is sometimes seen in
GSAS H/L = Fullprof D_L
polycrystalline solid specimens such as pieces of metal.

Latest revision as of 13:31, 20 March 2014

GSAS offers 5 different Constant Wavelength (CW) X-ray profile functions. They are described in detail within the GSAS technical manual (see page 156). 11-BM users are encouraged to use either profile type 3 or type 4.

A quick reference guide to terms in the CW profile type 3 and 4 functions is given below after a brief introduction. Users are strongly encouraged to read this section of the GSAS manual at least once before (or after) blindly using this guide! The Powder Diffraction Crystallography Video lectures are also a great resource for both beginners and experts.

Pseudo-Voigt Profile Functions

These peak profile functions are a pseudo-Voigt type, combining Gaussian (G) and Lorentzian (L) components.

The general Gaussian shape (as a function of angle θ) is described by the Cagliotti function

Gaussian Profile ≈ U*tan2θ + V*tanθ + W + P/cos2θ

These U, V, W, and P variables match the GU, GV, GW, and GP profile terms you see below.

The Lorentzian shape is more complex (check the manual), but includes size and strain broadening terms.

Asymmetry, Zero-Shift and Related Terms

Asymmetry

Axial Divergence (i.e. low angle peak asymmetry) is modeled in GSAS profile types 3 & 4 with the Finger-Cox-Jephcoat model (see GSAS manual). The profile terms S/L & H/L describe the intrinsic instrument asymmetry for low angle (< 5 deg theta for 30 Kev 11-BM data), and is related to the goniometer radius, or distance from the sample to the detector (about ~1000 mm for 11-BM).

Most 11-BM users will not need to refine these terms from the default (1.2E-03) values included in 11-BM GSAS instrument parameter files. If your fit shows an asymmetric misfit to strong low angle peaks, my might try to refine, but never refine both simultaneously.

Transparency

This is the trns profile term describing the x-ray penetration depth into a flat plate sample . It should *only* be used for fitting flat-plate reflection geometry (Bragg-Brentano) powder diffraction data. Do *not* use for 11-BM data which is collected in a transmission geometry (Debye-Scherrer)

Displacement

This is the shft profile term describing vertical displacement of the flat plate sample . It should *only* be used for fitting flat-plate reflection geometry (Bragg-Brentano) powder diffraction data. Do *not* use for 11-BM data which is collected in a transmission geometry (Debye-Scherrer)

ZERO

Instead of trns or shft, 11-BM users should refine the 2&theta zero shift term "ZERO". This term is found in EXPGUI under the Histogram tab. It describes any shift (error) in the absolute 2theta "0" position for the data and is measured in units of centi-degrees (100*2θ).

GSAS users refining 11-BM data should refine this term near the end of a refinement to ensure the most accurate lattice parameters. Typical values for 11-BM might be in the range +/- 0.1 (i.e. 0.001 degrees).

Background Function

The background function type 1 (Chebyshev polynomial) is recommended. Use the minimum number of terms needed to fit the pattern background. Start with 3 or 4 terms, and increase as needed. Sometimes 10-20 tems are needed for complex backgrounds.

Just remember that the background also represents scattering from something (sample holders, air, sample?). Try to think about what may contributes to the background for your sample. This might contain valuable information.


Suggested Profile Types & Terms for Fitting 11-BM Data

11-BM users are encouraged to use the GSAS constant wavelength (CW) profile type 3 or type 4. Profile #4 is best for cases in which anisotropic terms are required.

11-BM users will not (usually!) need to change or refine the default 'G' terms given in the instrumental parameter file. For the high-resolution synchrotron powder data collected at 11-BM, the instrumental resolution is well described by Gaussian terms.

On the other hand, sample effects in 11-BM data, such as size and strain broadening are (usually!) best fit and refined using Lorentzian terms. Gaussian crystallite size broadening (GP) is *rarely* observed; this requires a very tight mono-disperse size distribution rarely encountered in powder samples (solid metal samples may be an exception).

Many 11-BM diffraction patterns can be well fit by refining only the LX (size), LY (strain) and anisotropic Lorentzian size & microstrain terms. Profile type 4 is recommended if the later are required.

For more info see a EXPGUI-GSAS Parameter tutorial video


Constant Wavelength X-ray GSAS Profile Type 3

GU = Gaussian U term GV = Gaussian V term GW = Gaussian W term
GP = Gaussian crystallite size broadening LX = Lorentzian isotropic crystallite size broadening LY = Lorentzian isotropic strain broadening
S/L = Axial Divergence S term H/L = Axial Divergence H term trns = Sample 'Transparency' (note 1)
shft = Sample 'Displacement' (note 1) stec = Lorentzian anisotropic strain broadening (note 2) ptec = Lorentzian anisotropic crystallite size broadening
sfec = Lorentzian sublattice anisotropic broadening (note 3) LXX = Anisotropic Lorentzian microstrain (note 2) LYY = Anisotropic Lorentzian microstrain (note 2)

Color denotes indicates terms 11-BM users should Refine (green) , Sometimes Refine (yellow) , and Not Refine (red)

  • note 1: Do *not* use for 11-BM data, see above
  • note 2: Better to use Profile Type #4 Anisotropic microstrain terms
  • note 3: See GSAS Manual before using


Constant Wavelength X-ray GSAS Profile Type 4

GU = Gaussian U term GV = Gaussian V term GW = Gaussian W term
GP = Gaussian crystallite size broadening ' LX = Lorentzian isotropic crystallite size broadening 'ptec = Lorentzian anisotropic crystallite size broadening
trns = Sample 'Transparency' (note 1) shft = Sample 'Displacement' (note 1) sfec = Lorentzian sublattice anisotropic broadening (note 2)
S/L = Axial Divergence S term H/L = Axial Divergence H term eta = Gaussian-Lorentzian mixing factor (note 3)
'SXXX = Anisotropic microstrain broadening (Lorentzian) (note 4) 'SYYY = Anisotropic microstrain broadening (Lorentzian) 'SZZZ = Anisotropic microstrain broadening (Lorentzian)

Color denotes indicates terms 11-BM users should Refine (green) , Sometimes Refine (yellow) , and Not Refine (red)

  • note 1: Do *not* use for 11-BM data, see above
  • note 2: See GSAS Manual before using
  • note 3: Changes pseudo-Voigt mix from pure Gaussian (eta=0) to pure Lorentzian (eta=1). Typical 11-BM data is fit well using (or at least starting with) eta = 1
  • note 4: Number of Stephen's microstrain terms SXXX, SYYY etc changes with phase crystal symmetry (ie. a monoclinic phase will have more of these terms than a cubic symmetry phase)


Physical Meaning of Profile Terms

Some of the CW profile terms can be interpreted to give physically meaningful strain and particle size information. See the page 162 of GSAS technical manual for details

Particle size broadening

Using the isotropic LX profile term

then particle size = (18000*K*λ)/(π*LX)

where K = Scherrer constant (typically ~ 1), π = 3.1416

and size units are in Angstroms (Å), same units as wavelength (λ), typically ~ 0.41 Å for 11-BM data


Strain broadening

Strain is more difficult to quantify.

For profile type 3, using the isotropic LY term, then

isotropic strain (%) = 100% * LY * (π/18000)

For profile type 4, consult the GSAS manual. The anisotropic strain are best visualized using gnuplot. See the Robert Von Dreele's Rietveld Mailing List post from July 28th, 2011 below:

Q: What is the "best" order for incorporating GSAS profile 4 anisotropy Sxxx terms into a refinement?
A: Robert Von Dreele writes: "Usually, I do the S4xx ones first & then the others.
The thing is fairly stable after that. Do all of them at once in any case in the end.
The S4xx must be > 0 but the rest can be any sign. If you want a quick way of seeing the shape,
run mustrplot in GSAS; you'll need gnuplot to see the plot.
For monoclinics, it might look like two beans back-to-back."
-Bob


Example 11-BM Profile Fit and Terms

Lanthanum Hexaboride (LaB6) is used an X-ray powder diffraction profile standard because of its sharp peak shape.

Representative LaB6 data for 11-BM (high resolution powder XRD) can be downloaded from the 11-BM wiki:

https://wiki-ext.aps.anl.gov/ug11bm/index.php/Standards_Data

On the same Wiki page find more details about Instrument Parameter and Input Files, plus example fits to LaB6 data measured at 11-BM


Convert GSAS profile terms to Fullprof terms

11-BM provides users with instrument profiles terms for GSAS. Convert these values to Fullprof profile terms using the formulas below.

For more info see: "Typical values of Rietveld instrument profile coefficients" Kaduk J, Reid J. Powder Diffraction (2011) vol. 26 pp. 88

Gaussian Parameters

GSAS Term <=> Fullprof Term : (description)
GU = 1803.4 * U : (instrumental term, ~ tan^2 of theta)
GV = 1803.4 * V : (instrumental term, ~ tan of theta)
GW = 1803.4 * W : (instrumental term, ~ constant with theta)
GP = 1803.4 * IG : (size broadening)

note: 1083.4 => centidegrees squared divided by 8*ln(2)

Lorentzian Parameters

GSAS Term <=> Fullprof Term : (description)
LX = 100 * Y : (size broadening)
LY = 100 * X : (microstrain)

note: 100 => degrees to centidegrees

Finger-Cox-Jephcoat asymmetry parameters are equivalent in GSAS & Fullprof

GSAS S/L = Fullprof S_L
GSAS H/L = Fullprof D_L