GSECARS X-ray Microprobe for Earth and Environmental Sciencesphome.postech.ac.kr/user/atl/note/xrsa/Microprobe-2.pdf · The GSECARS microprobe station is running well and productively,

ALS Users Meeting 19-Oct-2004

GSECARS X-ray Microprobe for Earth and Environmental Sciences

Matthew Newville, Peter Eng, Steve Sutton, Mark RiversConsortium for Advanced Radiation Sources (CARS)University of Chicago, Chicago, IL

X-ray Absorption (XANES / EXAFS)

Objectives for Earth and Environmental Sciences:

X-ray microprobe techniques:

oxidation state of selected elementnear-neighbor distances and coordination numbers

abundance and spatial correlations of heavy elementsX-ray Fluorescence (XRF), Fluorescence Mapping

Determine chemical associations, speciation, and structure of heavy elements on heterogeneous samples: soils, sediments, aggregates,plant material, isolated inclusions, or contaminants.

X-ray Micro-diffractionphase identification


Focusing: Kirkpatrick-Baez mirrors: Rh-coated Si, typically using 3x3μm spot sizes, at 50mm from end of mirrors.

Incident Beam:LN2 cooled Si (111) mono

Sample Stage: x-y-z stage, 1μm resolution

Fluorescence detector:16-element Ge detector

with DXP electronics Si-drift detector (shown) Lytle Detector Wavelength Dispersive

Spectrometer

Optical Microscope:5x to 50x objective with external video system.

GSECARS microprobe: APS 13-ID

Entrance Slits: typically 250μm, accepting ~30% of undulator beam

CCD Camera: Bruker 2k area detector


The table-top Kirkpatrick-Baez mirrors use four-point benders and flat, trapezoidal mirrors to dynamically form an ellipsis. They can focus a 300x300μm beam to 1x1μm.

With a typical working distance of 100mm, and a focal distance and spot-size independent of energy, they are ideal for micro-XRF and micro-EXAFS.

We use Rh-coated silicon for horizontal and vertical mirrors to routinely produce 2x3μm beams for XRF, XANES, and EXAFS.

Kirkpatrick-Baez Focusing Mirrors


Nicole Keon, Daniel Brabander, Harold Hemond (MIT), GSECARS

The Superfund site at the Wells G+H wetland, Woburn, MA (featured in A Civil Action) contains ~10 tons of arsenic within the upper 50 cm of the sediment. Most of the arsenic is held in the wetland sediments with relatively little As in the groundwater.

Usually an iron-reducing, anoxic environment such as a sediment would be expected to have high As mobility.

Can the metabolic activity of wetland plants, such as Typha latifolia (cattail) explain the sequestration of arsenic in the wetland?

Within ~100μm of the roots, Fe is oxidized to Fe(III) and forms a plaque on the root, even in these sediments. Could As be adsorbed to the ferric oxy-hydroxides formed at the root exteriors?

Where is As in the cattail roots? What elements (Fe) are associated with As? What is the As oxidation state in the roots?

Arsenic/Iron in cattail roots: XRF tomography

Physical slicing the root for 2D XRF mapping would damage the sample.

Fluorescence tomography can make a virtual slice of the root and show the elemental associations and concentrations in the slice.

As Fe

Zn Cu


X-ray computed microtomography (CMT) gives 3D images of the x-ray absorption coefficient.

An absorption image is collected as the angle ω is rotated through 180o, and the 3D image is reconstructed in software.

In some cases, element-specific images can be made by tuning the x-ray energy above and below an absorption edge.

Fluorescence x-ray tomography use a focused beam, scanned across the sample. The sample is rotated around ω and translated in x.

Fluorescence x-rays are collected as for XRF maps. Transmission x-rays are measured as well to give an overall density tomograph.

• data collection is relatively slow –one slice can be made at a time.

• can be complicated by self-absorption.

• can collect multiple fluorescence lines.

X-ray Fluorescence Tomography: Primer

broad x-ray beam

rotation stage

SamplePhosphor

CCD camera ω

fluorescence detector

Transmission detector

x

visible light

focused x-ray beam

Sample

ω

rotation and translation stages

fluoresced x-rays


Fluorescence detector:multi-element Ge detector

Sample stage: x-y-z-θ

Sample, mounted on silica fiber, or in ‘shrink-wrap’tube, on a goniometer head

KB mirrors,with Pb tape shield

Fluorescence Tomography: Experimental Setup

Optical microscope


x

ω

The raw fluorescence tomography data consists of elemental fluorescence (uncorrected for self-absorption) as a function of position and angle: a sinogram. This data is reconstructed as a virtual slice through the sample by a coordinate transformation of (x,ω) → (x, y). The process can be repeated at different z positions to give three-dimensional information.

Fluorescence sinograms collected simultaneously for Zn, Fe, and As for a cross-section of As-contaminated cattail root (photo, right): x: 1100μm in 10μm steps ω: 180° in 3° steps

Fluorescence Tomography: Sinograms

Zn Fe As


Wells G&H Typha latifolia root: reconstructed slices from fluorescence μ-tomography, showing As concentrated on the root exterior, associated with Fe.

Fluorescence Tomogram Slices of Cattail Roots

As Fe

Zn Cu

Quantitative XRF analysis of the As and Fe concentrations from these slices give an As/Fe molar ratio of ~10 ppm, consistent with the average from bulk, wet chemical techniques.

Though only a few virtual slices could be made, this gives us confidence that the slices made are representative of the average.

• As and Fe are both at root plaque, not in the root interior. As and Fe are ~98% correlated.

• Cu, Zn, and Pb (not shown) are less uniform on plaque, suggesting they are not co-precipitated with or sorbed onto the Fe phase.

• Bulk XAFS of Fe shows Fe(III).


As XANESXANES measurements on the Typha latifolia cattail roots show mixed As oxidation state. The As3+ fraction did vary between different root samples, and along a single root, ranging from ~10% to ~60% As3+.

The spectra here shows roughly equal portions As3+ and As5+, which is higher As3+ than average.

Porewaters in the wetland have ~50% As3+: the As is generally more oxidized in the root plaque, which is consistent with adsorption onto ferric oxyhydroxides.


As XANES, Oxidation State TomogramsFluorescence tomograms made at 2 different energies:

EAs total As concentration EAs3+ As3+ concentration

would show spatial dependence of the As oxidation state.

XANES measurements on the Typha latifolia cattail roots show mixed As oxidation state. The As3+ fraction did vary between different root samples, and along a single root, ranging from ~10% to ~60% As3+.

The spectra here shows roughly equal portions As3+ and As5+, which is higher As3+ than average.


The As3+ / As ratio is heterogeneous (boxed areas). As5+ appears correlated with metals (Fe, Cu, Zn).

There does not seem to be a systematic spatial As oxidation gradient.

As oxidation tomograms for Cattail Roots

As3+

As total

As3+

As3+ total

total

As5+ appears at location with high Cu and Zn.

As5+ appears at location with high Fe.

More detailed spatial and oxidation state information would need faster data collection rates.

N. Keon, et al., Environ Sci & Tech, Oct 2004 (on line already….).


Ni in hyperaccumulating Alyssum muraleDavid H. McNear Jr*, Edward Peltier, Univ of Delaware

Root cross-section

Stem cross-section

Leaf cross-section

Ni

Ni

Ni

Mn

Ca

FeZn

Zn

Zn Fe

Fe

high

low

Root cross-section

Stem cross-section

Leaf cross-section

Ni

NiNi

Ni

Mn

Ca

FeZn

Zn

ZnZn FeFe

FeFe

high

low

high

low


Ken Severin, Tom Trainor, Univ of Alaska, Fairbanks, Randy Brown, US Fish and Wildlife Service

Sr distribution in Artic fish ear bones

One promising approach is to study the composition of otoliths (ear bones) which are records of the water chemistry.

Sr is particularly interesting, as it seems to be strongly correlated with salinity, and can be incorporated directly into the carbonate phases.

Bulk XRD shows a mixture of vaterite and aragonite.

Thin section of sheefish otolith: the dark inclusion is presumed to be vaterite while the lighter area expected to be aragonite. The otoliths are ~96% CaCO3 (mostly aragonite) and ~4% protein.

Electron microprobe results: strong Sr heterogeneity, consistentwith aragonite / vaterite (aragonite can incorporate more Sr).

Artic migratory fish, such as sheefish (stenodus leucicthys), are difficult to track, which complicates fisheries management. For the sheefish, it is not even known how much time is spent in fresh and brackish waters. (Sheefish from the Selawik river in northern Alaska).

One promising approach is to study the composition of otoliths (ear bones) which are records of the water chemistry.

Sr is particularly interesting, as it seems to be strongly correlated with salinity, and can be incorporated directly into the carbonate phases.

Bulk XRD shows a mixture of vaterite and aragonite.


We start with XRF analysis:

A line scan (a 1-dimensional map) shows a strong variation in Sr concentration (from ~2000ppm down to ~200ppm), across the aragonite (high Sr) to vaterite (low Sr) regions.

There is a strong heterogeneity in Zn (at 10s of ppm level), while Cr, Mn, Fe, Co, and Ni are uniformly distributed across the section.

We then proceed to doing XAFS on selected spots (#1: high Sr and #2: low Sr / high Zn).

Sr distribution in Selawik Sheefish otolithsKen Severin, Tom Trainor, Univ of Alaska, Fairbanks, Randy Brown, US Fish and Wildlife Service


μXAFS (with ~10x10μm spot size) were collected on selected spots:

#1: high Sr

#2: low Sr / high Zn

Also shown is data from Sr in coral aragonite from Allison et al. (earlier GSECARS μXAFS measurements).

Sheefish Sr micro-XAFS

Spot #2 Spot #1

EXAFS analysis in progress….



Sheefish Sr micro-XRD

Spot #2 Spot #1Spot #7

μXRD (with ~10x10μm spot size) were collected on several spots. Spots #1 and #7 are confirmed to be aragonite and vaterite structures, but there are questions about spot #2 (high Zn).

More work needed, but a good start for 2 days of beam time.



Conclusions, Future Directions and ImprovementsThe GSECARS microprobe station is running well and productively,combining μXRF, mapping, tomography, μXRF, XANES, and EXAFS for a wide range of problems in geological, soil, and environmental sciences.

Areas for Improvement:Beam positional stability, especially during EXAFS. This has improved, but needs more work.

Ease of focus to below 2μm and of reliable defocus / refocus to a desired beam size.

Data collection speed and efficiency. We’re generally limited in time by the solid-state fluorescence detectors.

Ease of use for novice users: data collection and on-line analysis software for quantitative XRF, XANES, EXAFS, and XRD.

Documents

GSECARS X-ray Microprobe for Earth and Environmental Sciencesphome.postech.ac.kr/user/atl/note/xrsa/Microprobe-2.pdf · The GSECARS microprobe station is running well and productively,