Diffraction and Beyond: Thin Film Analysis by X-Ray Scattering... (Scott Speakman)
Sep 5, 2023
For more information, visit https://nanohub.org/resources/22632
Scott Speakman
6/3/15
"Diffraction and Beyond: Thin Film Analysis by X-Ray Scattering with a Multipurpose Diffractometer"
Thank you very much, it's great to be here this morning
and to interrupt your morning's lectures
on electron spectroscopy
with an all talk about X-ray scattering.
Except that it's not so odd
because we are going to be focusing
on the surface sensitive aspects of X-ray scattering
when you're looking at a thin film.
The types of aspects that are going to be very complimentary
to the electron spectroscopy techniques
we've been talking about this morning.
And so I'm going to be focusing
not so much on diffraction this morning
because I don't really wanna have to follow up
Morrow's talk yesterday.
And I'm going to be focusing on scattering.
And the fact that a modern diffractometer
does a lot more than just diffraction.
It's really a misnamed tool these days
because we have been able to expand
the capabilities so significantly.
Morrow touched on it yesterday.
What happens when an X-ray beam hits our sample?
Well, we have elastic scatter
which may produce the phenomenon of diffraction
which is going to give us atomic scale information.
We also have small angle scatter.
Another form of elastic scatter
which is giving us more information
on a mesostructural level, a larger scale
and in a different dimension.
We additionally have X-ray reflectivity,
X-rays that reflect from the surfaces of our sample.
And finally, we do have some X-ray fluorescence
which I'm not going to talk about
because I'm a crystallographer.
What can we probe?
What can we learn with a multipurpose diffractometer?
I actually ran out of room to put everything in this table
because the font was getting way too small.
But this is a fairly good overview
of the different types of information
we can learn from the thin film and a surface.
And what I want to point out here
is that we have to match what we want to learn
with some expectation of what our film is
so that we can identify the proper technique
that will give us the information that we are looking for.
There is no one size fits all,
one technique that will tell you everything
about any type of sample.
We need to know a little bit ahead of time,
at least have an educated guess
what it is we're looking at.
This morning, I'm going to focus
on our surface scattering techniques,
X-ray reflectivity, grazing incident sacs,
and what they can tell us about our thin films,
about our surfaces.
I want to reemphasize a point from yesterday's lecture.
When we are doing diffraction,
we do not just put a sample on the diffractometer,
hit go, get information, do an analysis.
We need to remember about the scattering vector.
We need to remember that we are only probing our sample
in the direction of the scattering vector.
In a coupled scan,
we are probing the vertical structure of our sample.
We are not getting significant information
about lateral structure.
If we want lateral structure information,
we need to change the scattering vector
to probe different directions in our sample.
So we always want to remember the scattering vector.
What are other important angles in the diffractometer?
We have the incident angle omega,
which sets the depth of interaction
of the X-rays with our sample.
A shallower omega gives us
a more surface-sensitive measurement.
We have our scattering angle two theta.
Then we have a chi tilt and a phi rotation
where we can move the sample to reposition the sample
with respect to the scattering vector
so that we are probing the direction
that we want information from in our measurement.
We take our modern multipurpose diffractometer.
We put on a fancy mirror.
We put on some anti-scatter optics and a good detector.
And now we can expand the capabilities
beyond traditional diffraction
to X-ray scattering techniques such as X-ray reflectometry.
So we're going to look at the reflection
of X-rays off of a surface.
We're all familiar with reflectometry.
We use it every morning when we check our hair.
What happens with X-ray reflectometry?
Well, we have some X-rays that reflect
from the top surface of our sample.
And some X-rays penetrate into our sample.
We get reflection at every interface
that has a different refractive index.
Basic fundamental physics.
What that looks like is we begin our scan
and we see a decay in the reflected signal
and we might begin to see some oscillations.
These oscillations are interference fringes
produced by interference of each separate wave
that is reflecting off of our sample.
We continue the scan out.
We continue to get more and more information.
Our X-rays are probing deeper as we increase omega.
So we are getting information from the topmost surface.
And now we are gaining information
from additional buried layers in our sample.
The ability to study a multi-layer thin film
and learn information about each individual layer
that is in our thin film.
What information are we learning?
We start with the density of the film.
The mass density correlates to the refractive index
and that is going to change our critical angle.
Below the critical angle, we have total external reflection.
Near 100% of the X-rays that hit our sample are reflected.
None penetrate into our sample.
When we go above the critical angle,
now the X-rays begin to penetrate
and to probe the interior structure of our sample.
So we can determine some density information.
We can very easily extract the thickness of our film,
the thickness of every film in a multi-layer
by looking at the spacing of the periodic fringes
that are produced.
We can quantify this in many different ways.
We could take a Fourier transform
and look for peaks in the Fourier transform
since this is a periodic repeating oscillation.
We could do a direct measurement,
just measure the distance there to there,
plug it into an equation, get a value.
Or we can do simulation and modeling and refinement
because we do understand fairly well the physics
that are behind this phenomenon.
And that's going to give us the most complete,
most detailed understanding of our sample.
The last thing that we can learn is roughness.
We might want to assume that we have
an absolute sharp boundary at every interface in our sample.
But the reality is that our surfaces
are not perfectly smooth.
We can use an AFM to probe the topmost surface
of our sample.
But to learn about the roughness of the interface
within our sample, we need an additional technique
like X-ray reflectometry.
We see here as the roughness increases,
and this is for the interface between germanium and silicon,
the buried interface,
the strength of the oscillations decays.
Reflectometry is a technique
that has been around for quite a while.
It is a well-established, well-understood technique.
The solution is rarely as straightforward
as we would hope it would be.
But we know what we're doing.
And we've been doing on diffractometers for a while.
What we've been ignoring until recently,
at least some of us,
is why is that signal decaying?
Where are the X-rays going?
Well, the X-rays, instead of being reflected,
are being diffusely scattered.
If we have a perfect, sharp, smooth interface,
we get near-perfect reflection.
When we have an imperfect interface,
then some of the X-rays are specularly reflected,
and some of the X-rays are scattered off-specular,
what we will call diffuse scatter.
Notice that now the diffuse scatter
is probing a different direction in our sample,
because the scattering vector for the diffuse scatter
is in a different direction
than the reflected scattering vector.
So if we look at and analyze that diffuse scatter,
we are going to begin adding information to our analysis,
because we are going to begin probing
the lateral direction in our sample,
and we're going to begin to understand,
perhaps even model,
the interface between our substrate and our film.
How are we going to do that?
Again, we need to change the direction
of the scattering vector.
It's quite easy.
We take our instrument, and we put in a little,
you know, here's our specular reflection.
Omega, theta, these two angles are equal.
The scattering vector is normal.
And we're just going to put in a little bit of an offset.
On older instruments, we would actually tilt the sample.
So we would call this, perhaps, a rocking curve,
because we're rocking the sample back and forth.
Modern instruments, the sample's usually stationary,
because that's just much easier to mount the sample.
And we tilt the diffractometer around the sample.
A little bit more of an engineering challenge,
but it makes our data collection much easier.
We're not fighting gravity.
So we are now able to probe in different directions,
and collect our off-specular scan.
What's the first thing an off-specular scan can tell us?
Well, we look at this reflection curve.
This is data from a thin film.
No, really.
The user insisted, that's a thin film.
There should be reflection fringes.
Why aren't there reflection fringes?
Well, from a student's point of view,
it's because the instrument's bad.
From the lab manager's point of view,
my instruments are impeccable.
It's the sample.
How can we figure out who is right, who is wrong?
We're gonna do an off-specular scan.
And now, we have our omega scan.
And now we know what we knew all along.
My instrument is perfectly aligned, and in great shape,
because this intensity here,
between the UNIDA wing and this reflected peak,
this is our diffuse scatter.
If I had a perfectly smooth film,
there would be a lot less intensity here,
and a much more pronounced UNIDA wing.
Every roughness to the interface
takes some intensity away from the specular reflection,
and scatters it into this region.
And in this case, we've lost a lot of intensity
from our specular reflection,
and a lot of it has gone into the diffuse scatter.
This film is not good.
In fact, when you actually go
and look at it under a microscope,
you see it's not even a continuous film.
It's little splotches around on the substrate.
We need to go back to our reactor and try again.
If we do have a good film,
we can actually begin to model the interface
using a couple of different models.
And we may be able, with a good, careful measurement,
to quantify, to model the diffuse scatter in our sample.
In this case, we took an iridium thin film,
and we are able to model it with a fractal interface,
and find that there was some lateral coherence.
There was order to the roughness of the surface.
It wasn't random.
And that lateral correlation
was kinda on the length scale of 20 nanometers.
And this is very interesting.
If we have a non-smooth interface that has order,
that means there's a mechanism behind that.
This is something we want to study, we want to understand.
We can go even further and do a diffuse scatter map
to understand the two-dimensional properties.
We're going to skip over this a little bit,
because I do want to get to our next mapping capability,
grazing incident sacs.
So, grazing incident sacs is a closely related technique,
a hybrid of X-ray reflectometry,
diffuse scatter mapping, and sacs.
With X-ray reflection, we just come in with our beam,
and we're looking at the reflected beam,
and we're probing in one direction.
With grazing incident sacs,
we come in with a more narrowly defined beam,
and now we look at all the X-rays that scatter,
not just in the equatorial direction,
but also in the axial direction,
and every direction in between.
Now, if we're going to probe this,
we need a detector that can see all of this information.
So, we're going to need a two-dimensional detector,
a hybrid pixel detector, an image plate,
something that can capture all of this information.
How does this relate to our reflectivity curve?
How does this relate to our diffuse scatter mapping?
Because they're not independent.
They're all related to the same fundamental physics.
So, we have our reflectivity scan,
a linear scan in the specular direction,
our diffuse scatter map,
which is probing lateral structure of the roughness,
still in one direction, and then we have this,
which is probing the lateral direction,
is probing the vertical direction, the equatorial direction,
the axial direction, and in between.
Every spot on this map is information
from a different direction in our sample,
from a different scattering vector.
Now, we can begin to analyze and quantify
more information about the structure of our sample.
Skip over the math, get to the pictures,
because that's what's impressive.
We can most effectively distinguish
between order or disorder in our samples.
We take these examples.
These are two types of ordered pores in our sample.
These are actually pores that were filled
with a cobalt nanoparticles.
If the pores are stacked vertically,
we get two lines here, which gives us information
about the lateral correlation,
the lateral spacing between those pores.
And we find out from center to center
is a distance of about 37 nanometers.
If the pores are perfectly ordered, perfectly structured,
we might even observe diffraction from the structure,
from the ordering of the pores, constructive interference.
In this case, we have horizontally stacked pores.
Our structure is now much more interesting,
a way cooler picture, and again,
contains a wealth of information
where we are able to correlate lateral and vertical
structure in the pore assembly.
We're also able to tell that this is
a well-ordered pore system.
This is a very useful tool,
because sometimes our processing isn't so good,
and we don't get a well-ordered pore system.
So we've seen this before.
This is the same type of structure,
but now the pores, instead of being ordered
and nicely stacked, have some random,
some disorder to them.
And instead of stating these sharp features,
we're getting a more broad, less dimensional ring.
So we are now able to extract information
about the structure of our surface
using our grazing incident SACS experiment.
We can go one step further.
A nice, simple case of just some iron nanoparticles
on the surface, and we are now going to be able
to quantify the size of those nanoparticles.
Where we use this a lot right now
is with quantum dot research,
where we can determine the size of the quantum dots,
and are they self-assembled on the surface
of the substrate, or are they more randomly assembled?
Are they disordered?
Depending on the properties we want,
we might want one or the other.
We now have the tool to probe and understand this
using this big, expensive instrument
we probably already have sitting in the basement,
as long as it can support a nice two-dimensional detector
that can resolve this structural information.
So, in summary, we are taking our traditional instrument,
our traditional capabilities, and as new optics,
as new detectors, as new electronics become available,
we're able to expand the capability.
And this is a very, a technique that is,
it's in its infancy.
You know, our ability to model it
and quantify it right now is a little bit limited,
but we are developing it very, very quickly,
and so the capabilities are going to quickly expand,
and that's when things are exciting.
That's when you're looking at something
that nobody has ever looked at before.
So, I think with that, I'd like to conclude.
If I can just do one quick more plug
that I want everybody to be aware of,
totally unrelated, the Panalytical Award.
Any of you who use X-ray scattering or X-ray fluorescence
to publish a novel, significant paper,
and you're the first author on it, a student or postdoc,
you can submit to the Panalytical Award.
It's a nice 5,000-euro award,
and A, I wanna make you aware of it,
and B, I wanna make you aware
that this is to promote science, not a brand.
So, we do not care what instrument
you use to do the research.
We have an independent review board.
We would love to see some papers
coming from this area, from UIUC.
So, please consider submitting to this.

