Transcript Fluorescence Tomography Poster

GeoSoilEnviroCARS
The University of Chicago
Element-specific Microtomographic Imaging
of Metals in Environmental Systems
M. Newville, S. R. Sutton, M. L. Rivers, and P. J. Eng, Consortium
for Advanced Radiation Sources, University of Chicago
C. Hansel, S. Fendorf, Department of Geological and
Environmental Sciences, Stanford University
N. Keon, D. Brabander, Department of Civil and Environmental
Engineering, MIT
The Distribution of Toxic Metals in Roots
The role of root-borne carbonate nodules in the attenuation of
contaminant metals in aquatic plants has been investigated with
EXAFS, SEM and x-ray fluorescence imaging.
Getting the elemental distribution and correlations across the
root cross-section is very important for determining the phases
of metals in and on the root. Physically slicing the root would
do enough damage that elemental maps would be compromised.
The raw fluorescence tomograms 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,)  (x, y). The process could be repeated for different slices to give threedimensional information.
Most of the arsenic is sequestered in the
wetland peat sediments with relatively
little getting in the groundwater.
In contrast, riverbed sediments near the
wetland (2m away) have higher
concentrations of aqueous (mobile)
arsenic despite lower solid phase
concentrations.
x
Wells G&H Typha root 2
Fluorescence Sinograms for Zn, Fe, and As collected simultaneously for a section
of contaminated root (see below): x: 300mm in 5mm steps : 180 in 3 steps
Fe
180 2
20 5
Pb
Ca
Zn
2 0.5
Zn
Selected Further Reading:
G.F. Rust, and J. Weigelt IEEE TRANSACTIONS ON NUCLEAR
SCIENCE, 75, pp 14 (1998)
can collect multiple fluorescence lines
A. Simionovici, et al. in Developments in X-Ray Tomography II, SPIE
Proceedings 3772, 304-310 (1999)
can be complicated by self-absorption
A. Simionovici, et al, Nuclear Instruments and Methods in Physics
Research A, 467-468, pp 889-892 (2001)
data collection is relatively slow
C. G. Schroer, Applied Physics Letters, 79 (12), 1912-1914 (2001)
Experimental Setup:
x-ray microprobe
Horizontal and
Vertical
KirkpatrickBaez focusing
mirrors
Typical spot
size: 3 – 5mm
Fluorescence
detector:
multi-element
Ge detector
Optical microscope (10x to 50x) with video system
The GSECARS x-ray microprobe station: APS sector 13-ID.
3+
As
/
5+
As
Nicole Keon, Daniel Brabander (MIT)
For assessing As contamination in roots, knowing the total elemental concentration is not enough: the oxidation
state is also desired. By selecting the incident x-ray energy, we can preferentially detect As3+ or As5+.
As3+
E3
Fluorescence Tomography Features:
Sample on a
silica fiber
on standard
goniometer
head
20
Oxidation State Tomograms:
x
Pb
Photograph of root and reconstructed
slices of fluorescent x-ray CT for
2 selected elements (color bars in fg/mm3).
Mn
rotation stage
300 mm
Cu
0.4

Sample x-y-z- stage: 1mm step sizes
As
Fe
Sample
Transmission
fluoresced x-rays
detector
thin x-ray beam
transmitted x-rays
A standard x-ray microprobe station with an undulator, fluorescence
detector, and a rotation stage is all that is needed for fluorescence
tomography measurements!
These sinograms are reconstructed to cross-sectional images of the
300 mm root (Phalaris arundinacea), showing Fe and Pb uniformly
distributed in the root epidermis while Zn and Mn are correlated with
nodules. Arsenic is poorly correlated with the epidermis, suggesting a
non-precipitation incorporation.
20
Fluorescence x-ray tomography is done by scanning a pencilbeam across the sample. The sample is rotated around  and
translated in x and the x-ray fluorescence intensity from
selected elements are measured, making an x- intensity map
or sinogram. This can be reconstructed by software into a
virtual slice through the sample. Transmission x-rays are
measured as well to give an overall density tomogram.
translation stage
As
Fe
A tomogram gives cross-sectional information about a sample
without having to slice it physically.
In a similar project, this group is studying
a Superfund site (Wells G+H wetland)
that gained notoriety in A Civil Action, with
a reservoir of approximately 10 tons of
arsenic within the upper 50 cm of the
sediment profile.
The metabolic activity of the wetland
plants may help to explain the
sequestration of arsenic in the wetland.
Zn
Fluorescence Tomography
Nicole Keon,
Daniel Brabander (MIT)
Colleen Hansel, Scott Fendorf (Stanford University)
GeoSoilEnviroCARS is supported by:
NSF EAR-9906456
DOE DE-FG02-94ER14466.
“As5+
The Problem with
Self-Absorption
Formally, the fluorescence intensity for a tomogram is complex:
”
E5
Where
is the atomic density for element i (which is what we really want
to know). Here
ET
detector
As total
is the absorption of the incident beam along the incident
beam-path, and
incident
beam
Fluorescence tomograms were made with incident
energy tuned to the As3+ white-line, the As5+ whiteline, and well above the As K-edge.
Weighted redox: As3+=43%; As5+=57%.
As3+/
The differences in the resulting tomograms tells the
differences in the distribution of As3+ and As5+.
As5+ is
generally heterogeneous (boxed
areas) and there is a tendency for As5+ to be
on the exterior (circled area).
is the absorption of the fluoresced beam along
the path
to the detector.
The self-absorption problem is difficult to solve in general, requiring a selfconsistent solution for
.
The root samples have low self-absorption, so we can set g = 1, and apply
standard tomographic methods to convert the measured sinograms to
.