X-ray Photoelectron Spectroscopy (XPS) and Auger Electron Spectroscopy (AES) Part 2 (Richard Haasch)
Sep 5, 2023
For more information, visit https://nanohub.org/resources/22629
Richard Haasch
6/3/15
"X-ray Photoelectron Spectroscopy (XPS) and Auger Electron Spectroscopy (AES)" Part 2
OK, so let me just pick up where I left off before.
And we're going to start with another mode
you can actually do XPS in, and that is imaging mode.
So if you have an instrument that is so equipped,
you can not only do high energy resolution spectroscopy
like you saw earlier this morning,
you can also, by having, in this case, a hybrid analyzer,
where it's basically two analyzers in one,
you can run the instrument in a different mode.
So in this case, we'll open up the aperture
at the entrance plane of the analyzer.
And so then in this mode, we'll transfer
a two-dimensional image of your sample to the detection plane.
And in this design, by putting the detection planes
both in the same plane, and they're both real imaging
you could jump back, actually back and forth
between spectra and imaging quite conveniently
without needing to move your sample.
So here's an example of some microcontact pricking
that was done here at the university, where
they wanted to confirm this step,
that the platinum catalyst that was
being used to etch the porosilicon pixels
was actually selectively depositing
inside these little circles.
And so by doing two X-pass images, carbon and blue
and platinum and orange, we're able to see
that the platinum was located inside the circles as desired.
Then we could jump out of the imaging mode
into the spectroscopy mode, collect some data within a dot.
We could see a fairly nice platinum signal within the dot,
and then move off the dot and actually see
quite a diminished signal.
We are still picking up a little bit of platinum,
so that could either be at the tail of the spot
that we're analyzing, or we are getting
small amounts of platinum deposited
around the surface.
So this is one example of XPS combined with ion sputtering
that I'd like to show.
This technique known as depth profiling.
Tim went through this in quite a bit of detail,
but you could actually do some depth profiling in XPS.
In most cases, though, with XPS, because
of the physical interactions of the ion beam
with your material, there is quite a bit
of chemical and electronic damage done to your material.
Tim showed some really beautiful models of how that looks.
And so in most cases, if you were
to sputter a metal oxide due to preferential removal
of the oxygen, the metal reduces.
But in this case, this group wanted
to study the interface of the amorphous silicon carbide
layer that's on top of the transparent tin oxide layer,
look for any evidence of reduction at this interface.
And such a reduction would affect the transparency
at the interface and affect the photovoltaics efficiency.
So as I mentioned, in most cases, the oxides are reduced.
But tin oxide is one lucky case where the oxide does reduce,
but it does not reduce fully to the metallic state.
So I thought, ah, I can probably do this experiment
and get a result.
So when we sputter from the surface
down to this interface right here,
you can actually see a little bump develop,
indicating the presence of tin metal.
So this does, in fact, show that there
was a little bit of reduction at the interface that
carried through.
And to demonstrate further that the sputtering process was not
inducing this reduction, we went deeper
within the tin oxide layer, and we
see that eventually that bump disappears as you
get into just tin oxide.
Here's another example in both XPS and UPS
of the effects of ion sputtering.
And so this is the first transition metal nitride,
titanium nitride, that I showed you earlier in the talk.
Remember, we had these satellites
that indicate unscreened final states.
And so after sputtering, and these
are conditions typical for sputter cleaning,
just a few minutes, we notice that these satellites
are diminished quite a bit.
And that's because we're preferentially
removing nitrogen. Nitrogen is reducing
to a lower oxidation state.
It's actually becoming more metallic,
and we're providing more screening electrons
to the titanium.
So you see, just by a few minutes of sputtering,
we alter the electronic structure
of the material quite a bit.
Here's another example of scandium nitride.
Scandium nitride happens to be a wide bandgap semiconductor,
which is evidenced by this diminishing density
states up to the Fermi level.
And so the black is before sputtering.
After a few minutes of argon ion sputtering,
you'll notice in this red curve, we actually
induced a little bit of a Fermi edge.
So in this case, we're doping this material
with charge carriers.
And so again, changing the electronic structure
of the material.
This is one quick mention of another application.
You can use UPS4, and that is work function measurements.
Like I mentioned earlier this morning,
there's a lot of attention being drawn
towards two-dimensional materials.
Graphene, hexagonal boron nitride, aluminum disulfide.
And there's a lot of also interest
in modifying the surface graphene to actually either
to induce a bandgap.
As you know, graphene is a zero bandgap semiconductor.
You could, by putting things selectively at the surface,
you can actually make a bandgap in various electronic
properties.
So this shows an example of how you
measure work function.
The work function, as you know, is
the difference between the Fermi level and the vacuum level.
And so basically, what we're looking for
is the onset of electron emission.
And then we can take the energy of the photon
and look at the difference between that energy
level and the Fermi level and deduce the work function.
This is just a quick summary.
XPS, sensitivity is in the tenths of an atomic percent.
Chemical information, yes, for most.
Depth resolutions in the nanometer range.
Sample types can go solid UHV compatible,
conducting, semiconducting, and insulating
with modern spectrometers.
UPS, not usually elemental analysis.
Chemical state information, yes, but it's complicated
because that usually requires modeling.
Depth resolution is a little bit less than XPS.
Lateral resolution, several microns.
That's limited by the lamp.
And sample type, solid UHV.
And mostly conducting or not very insulating
semiconductors.
OK, let's now move into Auger electron spectroscopy.
And then in this technique, traditionally, we
use an electron to excite the sample
and analyze with electron beam.
I should probably look to see how long I have.
OK, just want to make sure I don't go over.
So now Auger electron spectroscopy
is a widely used technique for elemental analysis.
It was first discovered in 1922 by Lisa Meitner,
and then later independently discovered once again
by Pierre Auger in 1925.
And so the name actually stuck with him,
Auger electron spectroscopy, which
was most likely the result of some scientific sexism
of the early 19th century.
Fortunately, Lisa was recognized in her own respect,
and the element 109 was actually named after her, Meitnerium.
OK, this is a cartoon demonstrating
the particle interaction with the solid
that we see with Auger.
So you'll see this diagram come up later in the afternoon.
So you have the focus electron beam
interacts with the sample.
Different than a photon, the electron
can transfer parts of its energy.
It doesn't have to transfer its entire energy
to an individual atom.
It can incrementally transfer energy,
in which case it can actually interact quite a bit
with the sample.
So depending on your energy, you can go down several microns.
Exciting particles in that whole volume.
And as I mentioned earlier in this morning's talk,
Auger electron spectroscopy, being a technique
that we're looking at electrons, it is very surface sensitive.
And so we'll be just concentrating
on these electrons here.
And because we're very close to the surface in this area,
we're close to the neck of the bottle.
So it turns out our analysis area
is pretty comparable to the probe dimensions.
So we'll get roughly maybe half micron lateral resolutions.
Here is, once again, a repeat of the physics.
In this case, the electron comes in,
excites a secondary electron.
And then through a relaxation process,
an upper shell electron falls to fill that hole.
And then the third electron, the Auger electron, is emitted.
And this is what we'll be looking at.
Just a reminder that the complementary process that
takes place is the emission of an X-ray photon.
And it so happens that the probability for, in this case,
a K-shell relaxation is much higher
for resulting in an Auger electron
for the lighter elements than the heavier elements.
So it's actually the complement of EDS.
But the nice thing is then, this gives you
access to the lighter elements.
EDS gives you access to the heavier elements.
And so it's nice to have both instruments in the same lab.
One is a little bit better at the surface, Auger, that is.
And one's a little bit better in depth, EDS.
OK, here's a diagram of the Auger spectrometer
that we have in the lab.
First thing you'll notice that it looks a lot like an SEM.
That's basically because that's what it is.
It's an SEM with an energy analyzer
surrounding the electron gun.
And so the electrons are projected in the sample here.
And they pass through this football-shaped pathway.
And that filters the energies.
And we'll collect them there.
OK, just a reminder that because the deep core
levels and the upper core levels both shift to higher binding
energy as you increase the atomic number,
as well as the separation between the orbitals
increase with the atomic number, that's
what allows us to identify the characteristic lines of all
the different elements.
And the same trend applies to Auger.
They can make energies for the transitions
increase with atomic number, as you would expect.
And here, I show two examples of Auger survey
spectra for the first-order transition metals,
similar to the x-based data we showed earlier.
The one thing you'll notice right away, though,
is that the Auger transition is much more complicated
in terms of number of lines and the energy space span
as compared to the photoelectron lines.
And that adds some complications to the spectroscopy
in not only being able to identify elements,
but also to be able to precisely do quantitative analysis.
That's why it's also good to have
an x-based system in the same lab as your Auger system.
Traditionally, Auger spectra can be
presented two different ways.
The first way is the NOV MOA, where
you'll see that the lines are fairly
small on a linear and intense background.
And so that case, we could also do
a first derivative of the spectrum.
That way, we enhance the signal quite a bit
and give us a little bit more sensitivity
to lower concentration elements.
OK, here's a couple of examples of Auger.
And these lines are now, I'm comparing Auger signal
that was produced by the electron beam.
And if you remember my survey spectra at the beginning,
we can also create Auger lines with an x-ray photon.
And as you can see, the lines look about the same,
which you'd expect, because the physics is the same.
Only the exciting particle is different.
Because of the design of our photoelectron spectrometer,
you'll notice that we get a little bit better energy
resolution with the x-ray-induced Auger
spectrum as compared to the Auger electron-induced
spectrum.
That's because in order to make the Auger
spectrometer of a coaxial design like we have,
they have to make some sacrifices
in terms of the energy resolution of the instrument.
And in most cases, we use Auger for elemental analysis.
But in many cases, we can get chemical selectivity.
Here's an example of silicon carbide
with highly ordered pyrolytic graphite.
And besides the peak energy shift,
I adjusted the kinetic energy scale of these lines
just for comparison.
I overlapped the two.
So besides the energy shift of this line,
there's a little bit of change in the shape of the pre-peak.
And you use that, then, to differentiate HOPG from SIC.
Again, this is a comparison of that same set of data
collected with the x-ray source instead.
A little bit extra detail you can
see on the x-ray-induced Auger spectrum.
OK, I mentioned earlier this morning
there's a trick I can play to extend
the range of what we can look at with a conventional x-ray
source.
OK, so this is how we produce the x-rays in our x-ray source.
We actually use the same physics that you
use to produce the x-ray photon that you use in EDS.
So we use that as our x-ray source.
So this is how our x-ray source emits characteristic x-rays.
But if you use an x-ray source that does not
have a monochromator, you can actually
watch how that electron interacts.
We'll understand that electron, as it is going into the anode,
will interact with your anode material.
And at any point in time, when the electron interacts
with a proton and it changes direction,
the momentum also changes.
And then, because momentum must be conserved,
an x-ray photon is emitted.
We call this x-ray Bragg-Strong, or Bragg-Beam Radiation.
And we can have these x-rays up to the energy
of the electron beam.
Now, this will be a broadband x-ray.
So we'd not be able to use this x-ray
to do photoelectron spectroscopy,
because we'd just get a random distribution
of photoelectron lines.
We can use it for oxyelectrons generation,
because those kinetic energies are only
dependent upon the differences in the orbitals.
So here's an example where I used an x-ray source
to excite the KLL transition of silicon, which
happens to be at 1,600 eV.
So that's already 230 eV higher than the source
energy of the monochromatic x-ray source that we have.
Again, we see that we do get some chemical selectivity
with the Auger lines, just like XPS.
I'm going to skip to here.
And so like XPS, you'd like to be
able to relate the signal to the quantity of species
that are within the material.
And you'll get an equation with Auger
that's similar to the form of the equation we've
seen for XPS, with the addition of two other parameters,
the ionization or scattered cross-section
for the incident electron, and the Auger transition
probability, and the secondary ionization coefficient.
But like I said earlier, because we calculate the percentages
as a concentration, we can remove
the instrument-dependent terms, log all the rest
of these terms into a relative sensitivity factor,
and then use that for quantification.
Now Auger, because of its difficulties of overlap
with other peaks, it's not quite as quantitative as XPS.
It's going to be more on the semi-quantitative side.
And its sensitivities span the same range,
three orders of magnitude as XPS.
So the one thing you'll notice on this log scale
is that they'll hit a peak.
And then the sensitivity will sharply drop.
That's because the Auger transition
is a many-body process.
You have to get three things to work out right
for the transition to work.
And so usually, that only works for a small range of elements.
And then you'd lose signal.
But we can play with the electron beam energy.
And so we go from 3 to 10 kV to enhance
the sensitivity of some of these heavier elements.
But not lost, even though a lot of the elements
have a low sensitivity, we can't see a lot of them.
We do drop out when we get to the higher Z elements.
And so if you get to something like gold,
we may only be able to see maybe our limited detection
will be in the several atomic percent range.
Let's jump up to some other capabilities.
So because the Auger behaves like an SEM,
you can use it as an SEM.
And so here, we have an SEM image of a sample.
And they're interested in studying
the etching of iridium from this patterned sample.
Indium, I'm sorry, indium.
And so they're concerned at what the edges of the pads
look like.
So we located a little pad right here,
increased the magnification, then switched over
to ocean molar, collected an indium map.
And we see that the etch process did not
fully remove indium from the edges of the pad.
And so that would be an example where
these people would have to modify their etch process
to correct that.
Here's another example of a combination of imaging
with spectroscopy.
And this is a pretty simple one.
We see a lot of contrast here in this SEM in which we
have a light area and a darker area.
And because we can move the electron probe around,
we can look at a section we'll call A, which is above,
very nickel-rich, B, the darker area, very chromium-rich.
And so by now, we can collect a chromium and carbon
map while using this example.
We see a lot of areas that are chromium-rich here,
and then areas where carbon-rich are down here.
By plotting the intensity of each of the pixels
for carbon and chromium, we can look at where they correlate.
And so we can now select an area here, and then an area here
for the carbon, and then identify the various phases
present in this material.
And here's an example where we're
going to combine spectroscopy, imaging, and depth profiling.
So here, we're studying a boron-carbon fiber
and a boron-nitride matrix.
And a lot of times, in these composite materials,
the properties of the composite relate
to the interaction of the fiber with the matrix.
And so sometimes, the producers of these fibers
will want to modify this surface to vary the properties,
so the interaction between the two.
And so we can, first of all, collect an image,
select the fiber we want to zoom in on,
collect the spectrum, identify all the elements
that we have present, then look at these elements
as a function of depth as we sputter into them.
And so here, we can see we're combining
a bunch of different methods with the same technique.
So it's the same reason you would actually
use a bunch of different techniques for your analysis
here is an example where you'd want
to use a bunch of different methods with the same technique
to enhance the amount of information
you can extract from your material.
And so that leaves me to just a quick summary.
OJ, we're going to be able to see lithium in a ball.
The one drawback of OJ is, to make the OJ process work,
you have to start with at least three electrons.
And lithium happens to be the first element
that we have three electrons.
And so I want all you guys that can remember my OJ transition
process, the physics behind the process,
to try to figure out in your minds
how we get an OJ electron out of lithium.
It's not as easy as you think.
Non-destructive, we will suffer some beam damage
if you have electronegatives.
Mostly elemental analysis, semi-quantitator.
Chemical state of information from several elements,
not for everything.
Not always as easy to extract as with XPS.
Depth resolution comparable to XPS.
Lateral resolution, with our instance,
about a half a micron.
You can now go sub-10 nanometer with the field emission OJ
systems.
Sample type, once again, just like XPS, UHV compatible.
We can work with conducting, semiconducting
for very insulating samples.
You may have some limitations.
You may have to coat your sample.
And with that, I'd like to conclude.
And once again, thank our sponsors.
Thank you.

