Data recording, reduction and processing - EMBL Hamburg

Data recording, reduction and processing 

Manfred Roessle 

EMBL Hamburg

Outline 

Recording of Small Angle Scattering (SAS) data 

SAXS Variety: Laboratory Sources and Synchrotron based SAXS Beamlines 

What makes SAXS Beamlines special? 

X-rays in and X-rays out: From the source to the detector (via the sample?) 

Seeking and finding: What detects what? 

Primary Data Reduction 

1-dim is 1-dim 

Everything has to be normalized! 

Subtracting nothing from something a bit more than nothing 

The beauty of SAXS (raw) data 

What the hell does this all mean? 

Initial parameters to judge the success of the experiment 

Automated SAXS data recording and analysis 

Let’s have a beer, we are automated! 

2 26.10.2010 

EMBO Workshop on solution scattering EMBL Hamburg 25.10 to 01.11.2010

3 

SAXS Synchrotron based 

At all running synchrotron source are SAXS beamlines available! 

At all new constructed synchrotrons SAXS beamlines are planned! 

All SAXS beamlines are highly oversubscribed and “working horses” on their 

facilities! 

The scientific applications ranging from soft condensed matter, nano-science and 

fiber diffraction on ordered biological systems up to structural biology in solution. 

HASYLab, Hamburg ESRF, Grenoble 

Soleil, Orsay 

26.10.2010 

APS, Chicago 

Diamond, Didcot 

EMBO Workshop on solution scattering EMBL Hamburg 25.10 to 01.11.2010 

SPring 8, Hyogo 

Petra III, Hamburg

4 

SAXS Lab sources 

Bruker NanoStar 

Bruker NanoStar 

Anton Paar SAXSess 

Rigaku S-Max3000 

Hecus SAXSeye 

26.10.2010 


5 

The Electromagnetic Spectrum 

26.10.2010 


6 

Radiation from Synchrotron Storage Rings 

Production of X-rays 

Bending magnet 

Wiggler 

Undulator 

http://www.physics.uwa.edu.au 

26.10.2010 

Necessary part of a synchrotron. Many of 

these dipole magnets form the 

synchrotron storage ring. The electrons 

(positrons) are deflected and accelerated 

in the magnetic field. This acceleration 

generates the synchrotron light. The light 

is emitted tangential to the electron beam. 

Stack of magnetic dipoles. Put into the 

straight sections of the storage ring. 

Wigglers produces more light than 

bending magnets in a smaller source 

size. 

Most powerfull insertion device! A stack of 

magnetic dipoles generate a high flux of 

photons in a very small source size. The 

specific arrangement of the dipoles (d=n*λ) 

produces a discrete spectrum with 

coherent properties. 


7 

Radiation from Synchrotron Storage Rings 

Dipole bending magnet (APS) 

26.10.2010 


8 

Properties of X-rays from 

Synchrotron Sources 

1. High flux of photons 

2. Tangentially emitted with a 

central cone such like a 

spotlight 

3. Polychromatic light from 

infrared to X-rays 

26.10.2010 

X-ray optics for: 

1. Beam defining 

2. Focusing 

3. Monochromatization 


High heat load and radiation damage! 

Beamsize increases with distance from 

source point! 

Scattering/Diffraction experiments need 

monochromatic light!

9 

X-ray optics 

Beam defining 

Beam defining by slit pairs: 

From Glatter & Kratky 1982 

Small Angle Scattering 

Book online available! 

http://physchem.kfunigraz.ac.at/sm/Index.html 

Or google “Otto Glatter” 

26.10.2010 

Problem: 

Every edge cutting the beam produces 

parasitic scattering! Refractive streaks at low 

angles impedes SAXS experiments! 

Solution: 

Successive slit systems. First slits cutting the beam and 

second system cuts out the undesired parasitic scattering 

from source 

Slit system 1 

Beam defining 

Slit system 2 

Beam cleaning 

to detector 

For slits exposed to high head load, cooling is necessary 


10 

X-ray optics 

Focusing 

The glancing angle for X-rays on surfaces are very 

small. Only under grazing conditions X-ray mirrors 

can be used for focusing. 

Bending of the highly 

polished mirror surface 

permits focusing on the 

parabolic mirror profile 

26.10.2010 

reflectivity 

1 

0.9 

0.8 

0.7 

0.6 

0.5 

0.4 

0.3 

0.2 

0.1 

0 

Reflectivity for Rh@8keV 

0 0.2 0.4 0.6 0.8 1 

angles [deg] 

In order to use the full beam size X-ray mirrors on 

synchrotrons are typically in the range of 30cm to 

1m. The grazing angle can be increased by using 

high Z-elements (e.g. Rhodium, Platin, Nickel) as 

reflecting surface. These mirrors are acting as well 

as high energy filter; important for 

monochromatization 


11 

X-ray Optics 

Monochromator 

Monochromatization of the X-ray is achieved by using single crystals in 

Bragg diffraction geometry 

26.10.2010 

nλ= 

2dsin 

( Θ) 

For instance a Si 111 crystal 

with d-spacing (=distance 

between the crystal lattice) of 

d=3.14 Å deflects the beam for 

a wavelength of λ=1 Å to Θ ~ 

10°(9.16°). 


Θ 

Θ 

2Θ 

Problem: 

The Bragg condition is also fulfilled for 

multiple orders of the wavelength n. 

These higher order have to be filtered 

by the X-ray mirror

12 

X-ray optics 

X33 mirror 

1m Rh coated on Zerodur substrate 

26.10.2010 


X33 monochromator 

Si 111 single crystal in asymmetric 

cut 

This device is also used for horizontal 

focusing.

13 

Detectors for SAXS 

Seeking and finding 

26.10.2010 

Fiber optic tapered CCD cameras for X-rays. 

+ fast readout (10Hz framing possible) 

+ large area possible 

- low dynamic range 

- sensitive to overexposing 

- spatial correction due to the fiber optic taper 

- intrinsic background of CCD chip 


14 

Detectors for SAXS 

Seeking and finding 

26.10.2010 

Novel pixel detectors 

PILATUS system Swiss light source 

+ high dynamic range (10 20 phot/s!!!) 

+ no intrinsic background 

+ fast framing 

Ideal detector for solution scattering! 

500k model was installed successfully at 

X33 and upgraded to a 1M prototype 


SAXS 

Detector 

15 

BM4 

Camera Tube 

26.10.2010 

Experimental Hutch 

WAXS 

Detector 

Sample 

Area 

BM3 

Shutter 

Hutch Wall 

1.5m 1.2m 0.1m 2.2m 2.5m 0.7m 

2.7m 

X33-Beamline Schematic 

Not to scale 

Attenuator 

Optics Hutch 

SS4 SS3 SS2 Mirror SS1 Mono 

Aperture 

PS – Primary Slit SS – Secondary Slits BM – Beam Monitor 

31.6m 


0.9m 

BM2 

1.1m 

21.4m 

BM1 

+ 

PS 

+ 

Source

16 

Schematic X33 SAXS setup 

Beamstop with diode for measurement 

of transmitted beam 

PILATUS 1M 

26.10.2010 

1400 mm 

s: 0.1 nm -1 to 4.3nm -1 

d : 65 nm to 15 nm 

Pilatus WAXS detector up to 13° 

q~10 nm -1 d~ 6 Å 

1000 mm 


Sample cell 

Beamshutter 

with diode for 

measurement of 

incident beam 

(prior to 

exposure)

17 

EMBL’s X33 Beamline 

26.10.2010 


18 

Data Reduction 

From 2-dim to 1-dim 

Intensity 

1e+5 

1e+4 

1e+3 

1e+2 

26.10.2010 

d 

0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 

Channels 

Radial (azimuthal) averaging: 

The 2-dim detector intensities are stored as 

a images with three values: 

Intensity counts; pixel X and pixel Y 

I(x,y) 

The x=0 and y=0 position 

can be determined by 

the concentric diffraction 

cycles of Silver Behenate 

powder. 


Radial Averaging 

Log I(s) 

Normalization against: 

• data collection time, 

• concentration, 

• transmitted sample intensity. 

s, nm -1

20 

Data Reduction 

Assigning the s-axis 

Peak order d-spacing s-value d-Channel 

1 st order 5.834 nm 1.076 nm -1 495 

2 nd order 2.917 nm 2.153 nm-1 995 

3 rd order 1.944 nm 3.231 nm -1 1400 

4 th order 1.458 nm 4.304 nm -1 1990 

26.10.2010 

2π 

s= 

sin 

λ 

1 

( 2Θ) 

= 

d 

Intensity 

s-value [nm -1 ] 

5 

4 

3 

2 

1 

1e+12 

1e+11 

1e+10 


1e+9 

1e+8 

1e+7 

0 500 1000 1500 2000 2500 

Channels 

0 

0 500 1000 1500 2000 

Peak position [Channels]

21 

Background subtraction 

26.10.2010 

Standard protocol: 

1. Measurement: Buffer 

2. Measurement: Protein 

3. Measurement: Buffer 


We are looking to protein 

signals of less than 0.5% 

above the background level!

22 

Primary Data Analysis 

Judging the data quality 

26.10.2010 


First experiment 

BSA standard solution 

This standard is used 

for calibration and has 

to be freshly prepared 

and measured.

23 

First test 

Quality check on intensity data 

Ideal solution of particles Repulsive particle interactions Attractive particle interactions 

26.10.2010 


24 


The Guinier Fit 

26.10.2010 


Comparison of two a 

protein sample to the BSA 

standard. 

The difference at low 

angles < 1 nm -1 because 

of the different molecular 

weights (MW). 

The scattering at low 

angles is proportional the 

product of c protein · MW. 

At known protein 

concentration and a 

proper BSA standard the 

molecular weight of the 

sample can be estimated 

using the Guinier-Fit.

25 


The Guinier Fit 

26.10.2010 

Radius of Gyration Forward scattering I 0 

BSA standard 3.07 nm 185 units 

Sample protein 5.58 nm 867 units 


I( 

s) 

≅ I 

lnI 

MW 

MW 

( s) 

protein 

protein 

0 

e 

Rg 

− 

3 

≅ln 

I 

0 

2 

s 

Rg 

− 

3 

66kDa 

= ⋅867 

185 

2 

= 307kDa 

2 

s 

2

Protein folding state 

Kratky plot 

The Kratky plot is typically used to analyze the conformation of proteins, but can be used to 

analyze the random walk model of polymers. 

I(s)*s2 A Kratky plot can be made by plotting: 

versus s. 

Folded globular protein Completely unfolded protein Partially folded protein 

26 26.10.2010 


Porod Volume 

Rule of thumb: Porod volume is approximately two times molecular weight 

Here BSA: MW 66 kDa; Porod volume 115 

27 26.10.2010 


Porod approximation: 

Q= 

I(s) = K*s -4 

K 

∝ 

Q 

V 

∞ 

∫ 

s= 

0 

s 

S 

V 

2 

I( 

s) 

ds 

Q = Porod invariant 

S shape to 

volume ratio

28 

Automated SAXS data treatment “Pipeline” 

water.dat 

s-buffer1.dat 

sample-hi-conc.dat 

sample-mid-conc.dat 

sample-check.dat 

sample-lo-conc.dat 

s-buffer2.dat 

bsa-buffer1.dat 

bsa.dat 

bsa-buffer2.dat 

26.10.2010 

AUTORG 

AUTOSUB 

AUTOSUB 

AUTORG 

AUTOSUB 

AUTORG 

qual. 

qual 

. 

qual 

. 

.dat 

.dat 

I 0 

I 0 

.dat .out 

Merge AUTOGNOM DAMMIF 

MolMass 

R g ,s min 

MM 


… 

…

29 

BioSAXS sample changer evaluation setup 

Integrated in experiment control and analysis pipeline 

In user operation since 2010 

26.10.2010 


Joint project with EMBL 

Grenoble (project leader and 

construction of the device) 

and ESRF 

In Hamburg: 

Fully automated operation 

implemented and integrated 

in the analysis pipeline 

Filling faster than with the 

Fraunhofer changer 

35 µl sample volume (before 

80 µl)

Automation Results: what the users see 

30 26.10.2010 


31 

Automated Data Analysis: AutoRg 

Estimation of radius of gyration Rg and molecular weight 

Solution 

scattering data 

1. Automated (no help from a user) 

2. Implemented into processing pipline 

3. Input used for Autognom 

26.10.2010 

file.dat 


Program autorg.exe 

Rg – radius of gyration 

σ – accuracy 

Io – zero angle intensity 

Used range 

Quality 

Petoukhov, M.V., Konarev, P.V., Kikhney, A.G. & 

Svergun D.I. (2007) ATSAS 2.1 - Towards Automated 

and Web-supported Small-angle Scattering Data Analysis. 

J.Appl.Cryst. 40, 223-228.

32 

Estimation of D max with AUTOGNOM 

D max underestimated D max overestimated D max ok. 

26.10.2010 


33 

Automation 

Easy going and money is for nothing? 

26.10.2010 

Ab initio model with closed shutter!! 


34 

Conclusions 

26.10.2010 

1. SAXS beamlines are special equipped for 

making excellent solution scattering 

experiments 

2. Trust the beamline automation, it is working! 

3. Don’t trust the beamline automation – if you 

find something strange going on 

4. Validate your results! 

5. Ask the beamline staff in the case of troubles – 

they will help you

Data recording, reduction and processing - EMBL Hamburg

Create successful ePaper yourself

Delete template?

Save as template?