2. Beyond the Basics
• Stereo SEM
• EDS of particles: beware!
• TTL Detectors
• X-ray mapping
• Feature Sizing, Chemical Typing
• EBSD
• FIB
• and on and on
3. Stereo SEM
Why? Graphical demonstration of 3D shapes of rough or
complex objects
How? Very generally …..
1. Acquire two SEM images of the same object at the same
magnification but by tilting the stage slightly.
2. Color each image red or green in Photoshop.
3. Then in Photoshop overlay the two images, while
wearing red-green stereo glasses, shift them slightly
apart until you have the correct 3D effect.
4. Show downstairs in 2nd floor visualization lab for
maximum effect…
Exact detailed instructions on wall in room 308
5. Stereo SEM Details
For rough objects, use 3-7° tilt
For smooth objects, use more, 7-15°
Use less tilt if higher magnification
If you need to refocus, do NOT change objective lense
setting, rather move sample stage Z up or down.
Cited reference in literature: Heuser, 1989, Protocol for 3-D
visualization of molecules on mica via the quick-freeze,
deep-etch technique, Journal of Electron Microscopy
Technique, vol 13, p. 244-263.
6. EDS of particles
… is easy…. Maybe too
easy?
--> And easy to make
mistakes!
7. Particles - 1
• Mass effect/error:
electrons escape from sides
of small particles if E0 is
large enough, so
quantitative analysis will
be in error because
different elements (with
different binding energies)
will be affected
differently
Goldstein et al, 1992, p. 479, 481
8. Particles - 2
Absorption effect of non-flat
upper surface: different path
length from normal flat
geometry and the “normal”
way we do quantitative
analysis is to use FLAT
polished standards for
calibration -- so we could
have “too much” x-ray
intensity for particles
Goldstein et al, 1992, p. 479, 481
9. Particles - 3
•Variable effect of geometry of
trajectory between beam impact
area on non-uniform surface and
the location of the detector -- so we
get different results from the
same material, depending upon
where we place the electron
beam.
Goldstein et al, 1992, p. 479, 481
10. Results from 2006 777 student project on
EDS of rough samples with VP-SEM
The first column shows the actual chemical composition,
followed by the average composition of 20 points on
different grains, followed by the variation (=standard
deviation) of those individual measurements
11. EDS X-ray Mapping-1
BSE imaging provides rapid distinction of some/many/most
chemically distinct phases in a multiphase sample.
However, in many cases, having explicit 2D chemical
information is value. This is where “x-ray mapping” comes
into play.
X-ray mapping has changed tremendously since it was first
introduced in 1956 as a combination of WDS (crystal
diffraction spectrometry) and SEM: in those days it was called
“dot mapping” as literally dots would be painted onto the CRT.
To capture it, a photograph would have to be snapped of the
screen. It generally was pretty grainy. And only 1 or 2 or 3
elements could be acquired simultaneously (depending on how
many spectrometers were on the electron probe).
12. EDS X-ray Mapping-2
EDS was developed in the late 1960s and soon began to
be used for X-ray mapping.
For the next 30 years or so, users would select “regions of
interest” (ROIs), essentially the peak areas of particular
elements -- sometimes limited to a finite value like 8 or 12
-- which would then be used to “color in” a 2D area over
which the beam would scan, either one very long and
slow scan, or many faster scans that would be averaged.
Note: To this point, all the above maps included both the
characteristic X-ray being chosen AND the background/
continuum under the characteristic peak.
13. EDS X-ray Mapping-3
Things have changed in the 40 years since EDS came on the
scene:
1. Digital pulse processing have taken over from the older
analog processing, making shorter time constants and thus
higher count rates possible (e.g. 30,000 cps vs 3,000 cps
before) which create many more opportunities -- larger areas,
shorter times.
2. And then the development of SDD (Si-Drift Detectors) have
boosted the count rates to the hundreds of thousands of cps.
14. X-ray Mapping and the clock
Due to the low count rate of detected
X-rays, dwell times in the past
generally needed to be hundreds of
milli-seconds. A 512x512 X-ray map at
100 msecs took ~8 hours to acquire.
The improvements to EDS systems
with improved digital processing
throughput allows 1-10 msec dwell
times, dropping the x-ray mapping
times down to the hour or less range, in
many cases
Reed, 1996, Fig 6.1, p. 102
15. Continuum Artifact
Goldstein et al, 1992, Fig 10.6, p. 535
Here is an example of false
compositional contrast, an artifact
of the background being a
function of Z (atomic number, or
mean at. nu., MAN, for
compound). Specimen is Al-Cu
eutectic; X-ray maps are (a) Al,
(b) Cu, (c) Sc. The contrast in (c)
suggests Sc is present in the Cu-
rich phase. However, there is no
Sc, only the background in the
Cu-rich phase is elevated
relative to the background in
the Al-rich phase. Thus one
needs to be aware whether the
background is or is not subtracted
from X-ray maps, esp. when
looking at minor elements where
the continuum is a major
component. Many published maps
do not state if bkg-subtracted, so
assume they aren’t.
False
impression
Correct
impression
Al Ka map Cu Ka map
Sc Ka map
Sc Ka
background
substracted
map
16. Continuum and Atomic Number
At a given energy (or l), the intensity of the continuum
increases directly with Z (atomic number) of the
material. This is of critical importance for minor or trace
element analysis, and also lends itself to a timesaving
technique (Mean Atomic Number,“MAN”).
UW- Madison Geology 777
17. However…
“X-ray mapping” of predefined elements is now
pretty much a thing of the past (or of EDS
systems purchased before 2000 or so).
18. Data Cube
With the large increases in
computing power, memory, and
storage, it is now possible to
acquire a complete spectrum at
each pixel -- within the time frame
of the acquisition time.
Therefore it is possible to
• subtract the background
• find elements that were not
known to be present beforehand
Kotula et al, 2003, p. 2
19. Spectral (“hyperspectral”) Imaging Here any and
all x-rays
detected are
mapped to
each pixel
over which
the beam is
scanning. This
is both very
powerful (see
elements not
known to be
present before
starting), but
also loses
something
relative to
other ‘slower
old-
fashioned’
maps--lower
counts in peak
channels.
20. One of the issues with SEMs…
Notice the
smearing of
the spectal
image… for
LONG maps,
the SEM is
not meant as a
high stability
platform
(images take
seconds to
acquire).
Some
software (cost
extra $$) use a
reference
point for each
frame and if
there is drift,
correct for it.
21. Multivariate statistical analysis
Paul Kotula et al
demonstrated the
usefulness of
applying “principal
component analysis”
to spectra images.
They developed a
procedure for
converting from
abstract principal
components to
physically
meaningful pure
components.
Kotula et al, 2003, p. 3
22. Multivariate statistical analysis
Here is an example of the
MSA to a braze between
copper and alumina, where
7 distinct chemical
components were detected
with the automated
spectral image analysis
described by Kotula and
coworkers.
Kotula et al, 2003, p. 6
23. Needle in a haystack
Because the entire
spectrum is acquired at
each spectrum, the
software can sum up
the entire “extracted
spectrum” for the
whole image. If there is
just one pixel with a
“rare” element (and
enough passes have
been acquired to get
enough counts) it is
possible to locate the
“needle in the
haystack” using these
spectrum images.
24. X-ray Mapping
3 X-ray maps combined; each element set to a
color, and then all merged together in
Photoshop. The maps took ~8 hours to collect.
This is an
old-fashioned
map using
older EDS
software +
WDS
channels with
the SX51
electron
probe, slowly
moving
across the
sample, then
applying
colors to
elements.
This
instrument is
built to be
more stable than the SEM so there is no smearing as in the previous image.
25. Feature Sizing, Chemical Typing
This is a valuable feature of the SEM for locating and identifying a
particular mineral in a mixed population (e.g. K-rich minerals mixed
with K-poor minerals.
1. Acquire BSE image with good contrast so the K-rich mineral is distinct
2. Set a desired brightness level as the test for the mineral in question (if
brighter than, then acquire short EDS spectrum of center of grain)
3. Set elements to be evaluated; set short (~2 second) EDS acquisition
time
4. Set the stage coordinates of the corners of the area; Run the program
5. Return the next day and look at the list of grains it acquired EDS
spectra; order from high to low K; then drive to each grain to verify it is
the grain you want.
The next 3 slides go through the process…
26. 1. Set the BSE intensity and then set the range
to select one phase
28. 3. After SEM done, check the grains the
software suggests are the ones you want
29. “What will they think
of next?”
The veritable scanning electron microscope
keeps providing more and more
opportunities for improvement and
innovation…
30. TTL
“Through-
the-lens
Detectors”
Zeiss (Leo)’s “Gemini” field emission SEM has a neat detector:
TTL. The objective lens projects a magnetic field downward
which traps SE1,2 and draws them up thru the lens where they
are detected by a scintillator-light guide-photomultiplier. This
enhances SE images, in eliminating both SE3s and most BSEs.
…Sharper images…
31. Energy filtering of SEs
At the August 2010 Microscopy and Microanalysis meeting in
Portland OR, I found one talk to be very interesting:
Someone working with FEI, and operating a low E0s, has found
a way to “energy filter” the very low energy secondary
electrons, producing a technique to “see through” the common
surface contamination on materials. The slightly higher energy
SEs (say 10 eV) result from the impact of the E0 electrons with
the surface contamination layer, whereas those SEs coming
from the material under the contamination have lower energy
(say 2 eV).
32. Electron Backscatter Diffraction
Over the past decade, EBSD has rapidly become a
desirable attachment to the SEM. The SEM permits easy
imaging of micron-sized domains, and the EBSD detector
permits grabbing -- and analyzing -- the diffraction
patterns.
Thus in addition to images, EDS spectra, now there is
crystallographic and orientation information available.
We will spend another lecture of this class going into
detail on EBSD.
33. Focused Ion Beam Instruments
The past decade has seen the rapid acceptance of SEMs
built to hold FIB sources. A “gas injection source” of
liquid gallium is aimed precisely at a region imaged by
the SEM, and can precisely sputter away sample material
• to expose material below the surface that was otherwise
unaccessible for SEM imaging or EDS characterization,
or
• to “dissect” the sample into very small (sub-nanometer)
slices for TEM study
34. Focused Ion Beam Instruments
From JAMES D. SCHIFFBAUER and SHUHAI XIAO, 2009
37. “Personal
SEM”
ASPEX has developed this
small portable SEM for use
“in the field” mainly for
quality control/maintanence
work
• runs on 110 or 220 volts
• puts out 2-20 kV
• BSE images and EDS
• portable, easily fits in the
back of a van or SUV
From ASPEX literature
