Electron probe microanalysis EPMA

1
UW- Madison Geology 777
Electron Probe MicroAnalysis
Sample Preparation for EPMA
2
Whats the point?
UW- Madison Geology 777
If it dont fit in the machine, you aint gonna
get any numbers. If it aint polished right nice
and purdy, you aint going be able to trust the
numbers like you want. If your epoxy aint cured
nice and solid, its gonna bubble and degas
inside the probe and make John real unhappy. If
it aint conductive, it will charge and your
numbers will be in the toilet.
3
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. Dont be hesitant about asking
questions.
4
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 databut you could maybe get a
few data points out of the center where the
curvature is slight.
5
Sizes
UW- Madison Geology 777
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.
Shuttle
Holder
This is the sample holder for six 1 diameter
mounts
6
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.
7
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.
8
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, 150C, 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/microscop
ic applications (high vacuum, low degassing).
There are several suppliers, with Buehler being
one commonly used (higher price but excellent
product).
9
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.
10
UW- Madison Geology 777
11
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.

12
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).

13
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.
14
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
15
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
16
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 X-Ray 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/cnt10
7-6.htm
17
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 (lt1 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).
18
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.
19
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 Labs 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 lt1 ?m alumina or silica) and then recoated.
Watch your eyes!
20
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 enterring
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
21
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
Windows, 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
22
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.)
23
UW- Madison Geology 777
Documentation
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.