analysis of sans and usans data...steve kline ncnr summer school neutron small angle scattering and...
TRANSCRIPT
![Page 1: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/1.jpg)
Steve Kline
NCNR Summer School
Neutron Small Angle Scatteringand Reflectometry
June 26-30, 2006
Analysis of SANS andUSANS Data
![Page 2: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/2.jpg)
Outline• Basic Equations
• Model-independent methods
Guinier, Porod, Invariant
• Non-Linear Model Fitting
Particles, Polymers, Materials
• Global Fitting
• Anisotropic Scattering
• Transforms
• Ab initio modeling
Easy
More involved
![Page 3: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/3.jpg)
The Basics
• Inhomogeneities in give rise to small angle scattering
Starting from:
• We can replace the sum over atoms with an integral overthe scattering length density
• Normalizing by sample volume and introducingscattering length density
is the “macroscopic cross section”
![Page 4: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/4.jpg)
• A statistical description may also be appropriate
Non-particulate correlation function
Scattering BasisDifferent systems each have a natural basis - and all are equivalent
• This is especially true if the scattering is from “countable” units
Polymers monomer unitParticulates per particleProteins polypeptide subunits
![Page 5: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/5.jpg)
Measured IntensityAfter reducing the raw data - you will typically have:
I(Q) vs. QUnits of intensity are cm-1 ster-1
Units of Q are length-1
![Page 6: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/6.jpg)
Model Independent AnalysisWhat information can you obtain?
• Invariant- volume fraction, data consistency
• Porod Limit- specific surface area, surfactant head group area
• Guinier Analysis- general or specific dimensions
• I(Q=0)- particle volume, molecular weight
![Page 7: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/7.jpg)
Scattering Invariant10 % black90 % whitein each square
• Scattered intensity for each would certainly be different
• For an incompressible, two-phase system:
• Domains can be in any arrangement *Guinier and Fournet, pp. 75-81.
**Need “wide” Q-range to do integration
![Page 8: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/8.jpg)
Porod Scattering
*Glatter and Kratky, pp. 30-31.
• At large q:
SV = specific surfacearea of sample
**Need “sharp” interface and “high” Q
![Page 9: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/9.jpg)
Guinier AnalysisGuinier Approximation:
Guinier Plot:
Guinier Radius = Rg= RMS distance from “centerof scattering density”
Sphere:
Cylinder:
Gaussian Coil:
**Need “dilute” particles and “low” Q
![Page 10: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/10.jpg)
Zero angle scatteringbecomes
In terms of concentration:
for N uniform particles involume V, each with SLDrp and volume Vp
**Need “dilute” particles and “low” Q
![Page 11: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/11.jpg)
Determining the Contrast Match Point• Make several measurements atdifferent solvent SLDs• Keep the same concentration• Extrapolate data to I(Q=0)• Plot sqrt(IQ=0) vs. SLD
• Don’t forget to correct for theincoherent background contribution
• For composite particles, I(Q=0)will never reach zero - but it will bea minimum at the average particlecontrast
**Need “dilute” particles and “low” Q
![Page 12: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/12.jpg)
Non-Linear Model Fitting
• propose a structural model• calculate I(Q)• adjust structural parameters• repeat until done
One of the most commonly used methods - a “forward” calculation - many structures and interactions to choose from
![Page 13: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/13.jpg)
ResolutionAll measured data is affected to some extent by the instrumentconfiguration = “Resolution Smearing”
“True” I(Q) -> Is(Q)
• SANS uses pinhole collimation• USANS uses slit collimation (more significant smearing)
Resolution effects should NEVER be ignored during analysis
- NCNR tools make it easy to include resolution
![Page 14: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/14.jpg)
Non-Linear Model Fitting• Non-linear least squaresfitting to experimental data
• Use all the information youcan to reduce the number offree model parameters-SLD’s-Concentrations-Lengths
• A “good” fit does notnecessarily guarantee aperfect representation of thestructure in the sample
**Need knowledge of sample and model
![Page 15: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/15.jpg)
Contrast VariationContrast Matching
reduce the number of phases “visible”
• The two distinct two - phase systems can be easily understood
(shell visible)
(core visible)
r solvent = r core
r solvent = r shell
or
![Page 16: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/16.jpg)
Global FittingContrast variation- same particles, different solvent- R, rp, f are the same- rsolv, background are different
SANS + USANS data- same sample (same cell)- all parameters are the same- smearing, scaling different
**Need “correct” model
![Page 17: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/17.jpg)
Anisotropic Scattering
Qx
Qy
H
Elongated particles aligned by shear
Magnetic domains under an applied field
DM = magnetic contrast
Analyze as a 2D pattern I(Q,f)
![Page 18: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/18.jpg)
Transforms - p(r)p(r) is the probability that 2randomly chosen points areat a distance r apart
If I(Q) is measured over a wide enoughQ-range, then one can compute p(r) as theinverse transform:
For a sphere:
*See: D. Svergun, O. Glatter**Need “dilute” particles and “wide” Q-range
![Page 19: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/19.jpg)
Ab initio methods• Calculate I(Q) for complex structures - Biological molecules made up of subunits - non-standard geometric shapes
• fill volume with spheres on a grid• need distance between every pair• can be computationally intensive• can optimize shape
- packages available (D. Svergun)
**Need “dilute” particles and “burly” computer
![Page 20: Analysis of SANS and USANS Data...Steve Kline NCNR Summer School Neutron Small Angle Scattering and Reflectometry June 26-30, 2006 Analysis of SANS and USANS Data Outline • Basic](https://reader034.vdocument.in/reader034/viewer/2022042201/5ea0cdac6edc5203cd3728ef/html5/thumbnails/20.jpg)
Summary• Start Simple
• Work up to more complex
• Must use all other information available• Must always make physical sense• Must always check that approximations are valid
-dilute, random, length scales, etc.
• Many tools available at the NCNR• Many tools available on the web