UW- Madison Geology 777
Electron Probe MicroAnalysis
Revised 3/30/2014
UW- Madison Geology 777
What’s the point?
If it don’t fit in the machine, you ain’t gonna get
any numbers.
If it ain’t polished right nice and purdy, you ain’t
going be able to trust the numbers like you want.
If your epoxy ain’t cured nice and solid, it’s
gonna bubble and degas inside the probe and
make John real unhappy.
If it ain’t conductive, it will charge and your
numbers will be in the toilet.
UW- Madison Geology 777
It does make a difference:
Proper sample preparation can be critical to the
success of your EPMA work.
A successful experiment (creating diffusion couple,
achieving multiphase equilibrium, locating critical rock
specimen, etc) can be result in unsuccessful EPMA
results if the proper sample mounting and polishing is
not done.
Don’t be hesitant about asking questions.
UW- Madison Geology 777
Goals
What is your goal? What are the “issues” with your samples? Make
sure your sample preparation is appropriate. Do some “dry run” preps
before you waste your precious experimental materials.
There are several different ways to prepare samples for the probe.
Generally people follow in the footsteps of those before them, which is
many times correct – but not always. You can learn new tricks from others
(I have).
Sometimes “normal” sample preparation (usually the polishing aspect)
yields a poor result for the type of EPMA the user desires, whereas for
another user, they could “make do” with it.
An example is a hard brittle material in epoxy. First, cutting rapidly
with high speed saw (in a hurry?) fractures the material. Then normal
polish (with loose abrasive) easily polishes the softer epoxy and makes
rounded edges of the material (the little that is left). This sample would
yield lousy diffusion profile data–but you could maybe get a few data
points out of the center where the curvature is slight.
UW- Madison Geology 777
Sizes
Specimen size - mount size - holder size
- probe shuttle size:
Shuttle: the base we insert into the probe. We
have 2 shuttles, both the same size. There
are 2 slots along the edges and a hole at the
‘south’ end, all for the insertion rod.
Holder: this is the thing we insert into the
shuttle; we have many different styles of
bases (thin sections, 1”, 1 1/4” and 2” rounds,
but they all have the same outside dimensions
to fit into the shuttle. It is held in the shuttle by
4 screws.
Mount: this is the thing you bring to the lab;
your specimen is attached somehow, either
embedded in plastic or epoxy, or glued to a
glass slide. Round mounts are held in place
with set screws; thin sections are held in place
with springs. They are top referenced (pushed
up from below against a stop) and therefore
the complete top must be flat and smooth.
Holder
Shuttle
This is the sample holder
for six 1” diameter mounts
UW- Madison Geology 777
Sample Mounts
90% of the samples used in
EPMA are 1” (outside diameter)
rounds, smaller plugs mounted in
brass 1” diameter holders, or
geological thin sections. The 1”
diameter rounds can have the
specimens mounted with epoxy
inside a 7/8” high phenolic ring
forms (e.g., Buehler 20-8151) or 1”
diameter compression mounts.
There are a few other sample sizes
that can be accepted by the probe;
ask if you have questions.
Do not use the (usually
red) soft plastic pipe
end covers for molds;
they deform (epoxy is
exothermic) and yield
improper angles.
X
UW- Madison Geology 777
Mounting grains
Many people need to mount crystals or grains. Typically like this:
1.Separate out the grains to be probed
2.Determine what size holder they will need to be mounted in
3.Place he grains on a sticky tape, within the required perimeter
4.Place the round holder over them, sticking firmly to the tape
5.Pour the epoxy over the grains, without creating air bubbles. This
can be difficult if using the very small (1/4”) mounts.
6.Wait for hardening
7.Remove and polish (itself a whole other procedure).
In 2014, we have come upon an important
improvement in step 3. Before, we used regular
double-sided sticky tape, which has a tendency to
pull up in the middle upon hardening, making it
hard to have all small grains polished to the same
extent. Now there is Kapton Tape. Much better.
UW- Madison Geology 777
Mounts cannot be more than 7/8” maximum height
If they are slightly greater, they can be
unknowingly forced into the chamber,
but they are impossible to remove, and
the chamber must be vented to remove
them manually.
See image to right, view of shuttle
upside down, showing mount bottom
extending above/ beyond the shuttle
bottom surface.
UW- Madison Geology 777
Mounting Media
Most samples need to be encapsulated for ease
of mounting in the probe or SEM and for
polishing; encapsulants are normally either
* compression mounts, curing at 3-4000 psi,
150°C, using Phenolics or Epoxies or
Acrylics
*or castable (cold), using
Epoxies or Acrylics
Do not use epoxies you purchase at a hardware
store. Only use ones created for
metallographic/microscopic applications (high
vacuum, low degassing). There are several
suppliers, with Buehler being one commonly
used (higher price but excellent product).
UW- Madison Geology 777
Epoxy Evaluations
The SIMS lab at Edinburgh has a web page where they
give the results on a variety of tests of various epoxies
used possibly for SIMS work. This is exactly relevant to
the needs of an electron microprobe or SEM (vacuum
outgassing, beam damage, etc).
http://www.geos.ed.ac.uk/facilities/ionprobe/EpoxyResins/
Also valuable sample preparation information from the
Stanford SHRIMP lab:
http://shrimprg.stanford.edu/Sample%20Preparation.htm
UW- Madison Geology 777
Mounting Media - comparison
On this and the next slide are 2 tables from the Buehler (online)
catalog, reprinted here to give you a feel of the different
important parameters to consider in choosing mounting media.
UW- Madison Geology 777
UW- Madison Geology 777
Mounting Media - devil in the details
• Epoxy mixing and curing: it is very
critical to get some things right:
*fresh epoxy —old is no good; if not
sure, mix up a small batch and make
a dummy mount to test it (write date
on new epoxy), and
*correct proportions (always use a
scale to measure the ingredients, and
mix well).
*Wear gloves, use hood if possible.
• Conductive filler: in some cases, it is
useful to have a filler (e.g., C, Cu) that
is electrically conductive, especially if
you do not need/want to carbon coat.
UW- Madison Geology 777
Edge Retention
Edge retention: how critical
is it for your research?
If you are looking at
diffusion from the very
edge inward, it is very
important.
If so, you need to pay
specific attention to media
used and to minimizing
shrinkage (a filler helps).
Cross-sections
UW- Madison Geology 777
Cutting, Grinding, Polishing
Your goal is to prepare the specimen
so that the surface is “mirror
smooth” to minimize errors in the
matrix correction (path length).
Cutting, coarse grinding and then
polishing with finer abrasives can
introduce various artifacts, which
may or may not be relevant to
(hinder) your investigation, e.g.,
plucking of grains; preferential
removal of ‘softer’ phases;
imparting chemical change and/or
deformation to the surface of the
sample.
Fig 27 from Remond et al (NIST J of Research,
2002, v. 107, p. 639-662): Schematic
illustration of physical & chemical
modification as a function of depth of a
polished material.
Clearly the surface becomes
more important when operating
with low E0 and looking at light
elements.
UW- Madison Geology 777
Cutting, Grinding, Polishing
Polishing is an art (a science) in itself. Originally it fell under
the heading of metallography, the science of polishing,
etching and microscopic examination of surfaces of materials
(using reflected light). One excellent reference that describes
“how to” methods is Samuels*.
Only a few comments will be made here, mainly as warnings
if you intend to do this yourself:
* Loose coarse abrasive works OK, but I
prefer embedded papers and mylar, “lapping
film” (cleaner and less erosion of material).
* Different hardness materials in one mount
are difficult to polish; best to mount and polish
separately if possible.
L. E. Samuels, Metallographic Polishing by Mechanical Methods, 1982, 3rd edition,
American Society for Metals, 388 pp
UW- Madison Geology 777
Cutting, Grinding, Polishing
* Length of time spent polishing:
Depends on exact situation; sometimes no
inherent relationship between time spent
and goodness of polish of sample (vs polish
of epoxy); you may get very good polish
with diamond-embedded mylar in 10
seconds; hours of automatic polishing may
grossly erode material and ruin geometry of
mount.
* Smearing of soft material (phases) is
possible*; also tiny fragments of holder
material (e.g., brass) is known to get
trapped in fractures and in mica cleavage
* Broster, B.E. and Hornibrook, E.R.C. (1994) Tin-lead contamination in polished epoxy
grain mounts of heavy minerals. Sedimentary Geology, 88, 185-191
UW- Madison Geology 777
Cutting, Grinding, Polishing
Two recent accessible* references are papers from the 2002
NIST-MAS workshop on accuracy in EPMA:
Implications of Polishing Techniques in Quantitative XRay Microanalysis by Remond, Nockolds, Phillips and
Roques-Carmes, NIST J of Research, 107, p. 639-662.
Sample Preparation for Electron Probe Microanalysis —
Pushing the Limits by Geller and Engle, NIST Journal of
Research, 107, p. 627-638.
Either from the class webpage www.geology.wisc.edu/~johnf/g777/777NISTarticles.html
or the NIST webpage nvl.nist.gov/pub/nistpubs/jres/107/6/cnt107-6.htm
UW- Madison Geology 777
Cleaning
After polishing, samples must be well cleaned and degreased, to
minimize outgassing and contamination of the probe vacuum.
Regardless of how or who makes your sample mounts, you are
responsible for cleaning and drying.
* Put your mount in a small beaker with isopropyl or ethyl alcohol,
not acetone, and ultrasonically clean for a few (2-5) minutes
(acetone is dirty and can leave a residue upon drying.) However,
some compression mounts will dissolve, so ultrasonicate them in
solvent for very short period (<1 minute) only if oil or diamond
paste (~oil) used.
* Rinse 2-5 minutes with
distilled water, shake off excess
water and blot dry with clean
kimwipe (to minimize water
marks on surface to be probed).
UW- Madison Geology 777
Cleaning
* Do not blow water off with
building compressed air (oil is in
the lines). Do not use canned
air (it leaves hydrocarbons on
the surface).
* Then dry on a hot plate or oven
at low heat to remove all
adsorbed water.
UW- Madison Geology 777
Carbon Coating
Non-conductive samples (e.g,. rock thin sections)
will need to be carbon coated. They should be
coated with the Probe Lab’s evaporator in Room
307; this is normally done by lab personnel, though
students who have lots of thin sections to coat can
be trained.
Experience has shown that coating elsewhere with carbon
sputtering does not yield favorable quantitative results on the
probe here.
Old samples that were coated many
years ago should probably have their
coats cleaned (i.e., with <1 m
alumina or silica) and then recoated.
Watch your eyes!
UW- Madison Geology 777
Carbon Coating —Thickness
We monitor the thickness of carbon by
observing the interference colors on a polished
brass mount (Kerrick et al, 1973)*
Variation in the carbon thickness between
standard and unknowns can create additional
uncertainty and error: 1) the electrons entering
the specimen lose energy passing through the
coating, reducing the x-ray intensity produced
within the specimen, and 2) emergent x-rays are
absorbed by the carbon as they leave the sample
on the way to the detector. These effects are
largest for the “light” elements, but also higher Z
elements also.**
150Å
Orange
200Å
Indigo red
250Å
Blue
300Å
Bluish Green
350Å
Green Blue
400Å
Pale Green
450Å
Silver gold
Color representatives not very
accurate! Need to see Indigo Red.
* Kerrick, D. M., Eminhizer,
L.B. and Villaume, J.F. (1973)
The role of carbon film
thickness in electron
microprobe analysis, Am.
Min., 58, 920-925.
** Armstrong, J.T. (1993) Effects of carbon coat thickness and contamination on quantitative analysis:
a new look at an old problem, in Proceedings of the 27th Annual MAS Meeting, S13-14
UW- Madison Geology 777
Carbon Coating —Light Elements
If you are analyzing “light elements” (F, O, C, N, B) in
insulators that require conductive coating, it is critical that your
samples and standards be coated simultaneously, preferably in a
rotating apparatus, for even coating.
Donovan will probably be implementing a carbon-coat
thickness option in Probe for EPMA, so for high accuracy, C Ka
can be measured on both standards and unknowns and the various
values computed within the matrix correction.
Armstrong, J.T. (1993) Effects of carbon coat thickness and contamination on quantitative analysis: a
new look at an old problem, in Proceedings of the 27th Annual MAS Meeting, S13-14
UW- Madison Geology 777
Stained Mounts —
NOT IN THE PROBE!
There are occasions where thin
sections are stained to distinguish various
phases (carbonates, feldspars).
Samples that have been stained ARE
NOT ALLOWED TO BE PUT INTO THE
ELECTRON MICROPROBE.
One probe laboratory where high
current hit the stain/dye had a catastrophe
where the optical components (mirrors and
lenses) were coated with the stain and it was
VERY EXPENSIVE to repair.
UW- Madison Geology 777
Storage
Protect your specimen mounts!
* Avoid touching surface with
fingers, use gloves if possible.
* Do not wrap in kleenex (paper
fibers); if necessary, use kimwipe
instead
* Keep in containers: plastic boxes
of various sizes are available in the
lab; also cardboard boxes for 25
rectangular thin sections
* Label the boxes with your name,
so if you leave them in the lab they
can find their way back to you.
(Durphy Packaging in PA
sells many sizes of small
plastic boxes; Palouse Petro
Products sells inexpensive
cardboard boxes for thin
sections.)
UW- Madison Geology 777
Documentation - Navigation
If your samples have any ambiguities or complexities, it
pays well to make maps or drawings beforehand,
showing orientations (e.g., North arrow) and regions of
interest.
It is difficult to find a 100 micron grain in a 25000
micron forest without landmarks.
We now can scan mounts, import the images to the
Probe for EPMA software, “register” the image with
actual stage coordinates, and then click on the image to
move to that area. Slick!