Transcript Document 7223452

Applications of in-situ X-ray
Scattering Techniques
Sam Webb
SSRL Scattering Workshop
May 15, 2007
Overview


Why in situ?
Experimental Design





Beamlines
Sample prep
Analysis
Reactions with x-ray scattering
Example(s)
Why Should I Do Scattering When I
Have EXAFS Data?
30
Triclinic
Birnessite
3
k (k)
20
10
Hexagonal
Birnessite
0
d-MnO2
-10
2
4
6
8
10
12
14
Intensity (Arb. Units)
40
1.0
Triclinic Birnessite
0.5
Hexagonal Birnessite
0.0
10
d-MnO2
5
2
D (Å)
-1
k (Å )
EXAFS = Local Structure
WAXS = Long-Range Structure
1
Why In-situ


Traditional powder diffraction experiments require
dry, fine powders as samples
For many biological and environmental samples:

Drying = artifact



Dehydration, exposure to air
Powder = artifact
Other thoughts to consider…



Sample throughput
Sample textures
Timing/Reactions
Experimental Design

Does my sample need to be wet?



High resolution vs. low




Transmission vs. reflection
Tradeoffs due to backgrounds of sample holder
and water
Soller slits vs. analyzer vs. area detector
Data range
Exposure to beam?
Exposure to air?
Diffractometer (SSRL BL 2-1)
detector
collimating slits
scattered x-rays
analyzer
incident beam
sample
Powder Scattering Experiment



Monochromatic
Sample contains all crystal
orientations
Detector and sample
angles unchanged
X-ray source
(Synchrotron)
Mono
Detector
Slits
Sample
Beamstop
Diffractometer (SSRL 11-3)


Tight spaces in hutch
Samples:





Flat plate
transmission
Reflection (half of
area detector)
Capillary
BL software (Blu-Ice)
5-10 MB per picture
Diffractometer (SSRL 11-3)


Tight spaces in hutch
Samples:





Flat plate
transmission
Reflection (half of
area detector)
Capillary
BL software (Blu-Ice)
5-10 MB per picture
Sample Preparation (Flat plate)

Keep sample hydrated to avoid artifacts!



Change in oxidation state/mineralogy
Collapse of hydrated structures
Use transmission geometry

Why? - Better subtraction of background
scattering (water, windows)
 Window material important


top plate
lexan
windows

sample
shim spacer

bottom plate
Lexan is a good material for background
removal (WAXS)
Water peaks in similar places as silica
Optimize sample thickness
depending on l and sample
composition. Sample should
absorb ~ 20-50% of incident
beam. One “m” is about max.
Other sample holders –
goniometer head – sample
distance ~ 37 mm (11-3, 7-2)
What if I have a powder for
transmission?
Flat plate is poor for dry
samples


Particles are not generally
stable and settle – even out of
beam!
Need a better support – tape!
Kapton
not ideal
Scotch
Magic tape (translucent)
10000
8000
Kapton
Counts

6000
4000
Scotch
2000
0
0.5
1.0
1.5
2.0
-1
Q (A )
2.5
3.0
Data Analysis



CCD to diffractogram (2D to 1D)
Geometry corrections
Background subtraction



Windows, capillary, tape
Water
Other interferences (cotton, etc)
Integration of Powder Pattern

What Can it Tell?

Peak Positions:



Peak Shape & Width:



Crystal structure
FIT2D

2q
Crystallite size
Textures (preferential
orientation, multiple phases,
etc.)
Peak Intensity:


Phase identification
Lattice symmetry
http://www.esrf.fr/computing/
scientific/FIT2D/
Theta Dependent Effects

Absorption
t
q

Samples absorb the incident and transmitted beams
Abs = (t / cos q) exp(-mt/cos q)
Measure
sample absorption at
the beamline!
t cos q
Volume effect


1/cos q dependence
Compton Scattering

0.6
Highest at large q
In order to get proper
removal of background
(windows, water) these
corrections must be made.
Critical for thicker samples!
Raw
Corrected
0.5
Intensity

0.7
0.4
0.3
0.2
0.1
0.0
0
20
40
60
80
2q (degrees)
100
120
Background Subtraction
Background in experiments consists of lexan
windows and water

0.6
0.12
Raw Data
Lexan
Water
Intensity
0.4
0.3
0.2
0.08
0.06
0.04
0.1
0.0
0.10
Intensity
0.5
0
20
40
60
80
2q @ 10 keV
100
120
0.02
0
20
40
60
80
2q @ 10 keV
100
120
http://www-ssrl.stanford.edu/~swebb/xrdbs.htm
RDSUB


GUI for removal
of background
and thickness
corrections
Designed for
use with Fit-2D
output (chi files)
Reactions

Mineral-solution reactions





Time scale of minutes to hours
Redox reactions
Cation exchange
Colloid transport
Sample prep = miniaturized “columns” (i.e., particles
packed in a capillary)

Lexan capillary



Particle size and porosity


Better background (no overlap with water like silica)
Doesn’t break!
Clogging
Flow rate

Stalling of pump
Reaction Flow Setup (not to scale)
Tubing
Scattered Beam
Cotton
Sample
Gasketed capillary holder
Incident beam
120 mL Syringe pump
Flow collection system
Beamline Setup (BL 11-3)
Future Improvements…

Peristaltic pump vs. syringe pump



Development of better column packing materials
Gas impermeable tubing


Easy loading of capillary
Fraction collector


Improve anaerobic conditions
Injection loop


Better flow and ability to change reactant solutions
Analysis of post-reaction fluids
Fluorescence detector

Monitor elemental changes in sample if reactions lead to
deposition / removal of compounds
Examples

Mn biomineral structures


Real time biogenic Mn oxidation


Area detectors in reactions
MnOxide reactions with metals


Compare 2-1 and 11-3 data quality
Area detectors in reactions
Sulfide mineral oxidation


Wet-dry artifacts for air sensitive minerals
Air exposure
Mn Oxide Biomineral Structure



BL 11-3
2 minute exposure


360 degrees are better
than 1!
BL 2-1
Sum of 4 to 5 scans, ~8
hours total
Mg
Mg
Ca
Intensity
Intensity
Ca
Sr
Ba
Sr
Ba
Na
Na
K
Rb
Cs
K
Rb
Cs
14
10
2
1
14
10
D(A)
Tradeoff between noise-resolution-time
2
D (A)
1
Biogenic Mn Oxidation
Triclinic
80 hr
0.05
50 hr
24 hr
Triclinic
0.25 Birnessite
Hexagonal
(310,020)
(311,021)
(200)
(110)
(201,111)
*
Hexagonal
Birnessite
12 hr
0.00
a*
b
(002)
Intensity (Arb. Units)
(002)
0.10
a
a
0.50
(003)
Intensity (Arb. Units)
0.15
(310,020)
(311,021)

Mn oxidation in seawater
progresses through
symmetry changes in
oxide structure
Due to the effect of Ca
present in interlayers
(200,110)
(201,111)

0.00
5
3
2
D (Å)
1
5
3
2
D (Å)
1
Manganese in-situ Oxidation
Mn(II)
Spores
1d
2d
Mn(IV)
50
Triclinic peaks
Mn(IV)
Oxide
Q
Q (nm-1)
40
30
20
Spores
10
Scan No. (~20 min between scan)
Manganese oxide reaction with metals
Co(II)
Mn(IV)
Decrease in (001) amplitude


Time (h)
Q (A-1)
Co(II) reacts with
pre-formed
biogenic oxides to
oxidize to Co(III).
Mn-oxides are
reduced
No evidence of new
Co(III) minerals
Biogenic MnOxides + Co(II)


001 peak broadens with reaction and shifts to larger
d-spacings
Changes follow pseudo-first order reaction kinetics

Slow and fast steps of Co(III) incorporation
8.61
0.85
th-1 = 1.27
th-2 = 15.16
8.60
0.80
8.58
(001) FWHM
(001) Position
8.59
8.57
8.56
8.55
0.75
0
0.70
10
20
30
40
50
0.65
0.60
8.54
0.55
8.53
0
10
20
30
Time (h)
40
50
60
0
10
20
30
Time (h)
40
50
60
Wet-Dry Artifacts
Measurements of
anaerobic, dried sample
lead to formation of
peaks with different
texture
3000
Counts

Paste
Dry
2000
1
2
3
4
-1
Q (A )
FeS dried
FeS paste, anaerobic
5
Fe-Sulfide oxidation reactions
~1.2 mm from end
2-line Ferrihydrite
10
O2
Counts
8
FeS
O2
6
4
2
0
1
2
3
4
5
Time (h)
-1
Q (A )
t=0
Mackinawite
t=7
t=8
t=9
t=10
Acknowledgements




Anna Obraztsova and Greg Dick (SIO)
Apurva Mehta (SSRL)
Tanya Gallegos (U of M)
Funding:



NSF-CRAEMS
DOE-BER
DOE-BES
R/V Knorr, Black Sea, 2003