7 Problem a

Problem a Electron optics.

i) What is the task of the lenses in the SEM?
Reduce the beam diameter and thus improve the resolution (secondary electron imaging)
ii) Name the 3 most important electron-optical lens defects.
Spherical aberration, chromatic aberration and astigmatism
iii) These lens errors can be reduced by reducing a common parameter, which one? How
can operator reduce this parameter?
The lens errors are reduced by reducing the beam divergence. This is done by reducing
the objective lens aperture and/or increase the working distance.

8 Problem b
Assume a (cheap) tabletop model SEM with fixed objective lens aperture and fixed working
distance. This configuration results in a fixed beam divergence of 2·10 rad.The operator can
choose between two acceleration voltages, 10 kV and 15 kV, and two beam currents,10nA and
30nA.
i. Which values of acceleration voltage and beam current give the best resolution?

The resolution is given by the beam diameter: d=k

√ i 1
Eα
where k is a constant, i is the beam current, E is the acceleration voltage and 𝜶 is the
beam divergence. Best resolution (thinnest beam) is achieved by using the lowest beam
current and highest acceleration voltage, i.e. 10nA and 15k
ii. Calculate the ratio between worst and best resolution for this SEM.
Since k and 𝜶 are both constants, the relationship becomes:

√
30
10
=2.12

iii.
√
10
15
Calculate an approximate value for the depth of field at the lowest magnification of 10X.
The depth of field is given by:

D= [δ
M
−dp
]1
α
3 δ
δ=0.1 mm , M =10 , α =2∗10 , ≫ dp
M

D= [ ]
δ 1
M α
=5 mm

9 Problem c
Assume a SEM equipped with an EDS (energy dispersive spectrometer).
I. You experience that the EDS analysis takes a long time due to low counting speed. How
do you want to as an operator could increase the counting speed so that the analysis time
is shorter?
In order to increase the counting speed, more electrons must be sent down onto the
sample per unit of time, that is increase the beam current. This is done by e.g. to insert a
larger aperture and/or change the condenser lens setting (increase the focal length). An
increase in the acceleration voltage will also increase the beam current.

II. Calculate the characteristic energy of K radiation from molybdenum (Z=42). A steel
sample which contains molybdenum is analyzed with EDS, but no K radiation from it is
detected molybdenum. What could be the cause?

The acceleration voltage may be too low (should be more than 20kV) or it may be
something (topography) which prevents the radiation from reaching the EDS detector.
III. An EDS analysis of a non-electrically conductive sample must be carried out in a low
vacuum. What problem occurs with EDS analyzes in low vacuum and how can it be
solved minimize this problem?
In low vacuum, the electrons are scattered by the air molecules in the sample chamber,
the so-called "skirt effect". The problem can be reduced by reducing the pressure,
reducing the working distance or increasing it the acceleration voltage.

Secondary electron images of an object imaged at 10kV (left) and 1kV (right), respectively. How
would you explain the differences in contrast/information in these images? ii) Secondary electron
image (left) and backscattered electron image (right) of an object imaged with 10kV. How would
you explain the differences in contrast/information in these images?
The image on the left shows the same image and signal as the figure on the left in i). The picture
on the right shows same sample and same voltage, but now depicted with backscattered electrons
instead secondary electrons. Backscattered electrons generally have higher energy than
secondary electrons and thus the detected signal comes from a deeper area i the sample. Also,
backscattered electrons will show atomic number contrast, something like in this the case shows
spherical particles inside the sample. Since the particles are bright, it means that the particles has
a higher average atomic number than its surroundings
star_border
iv) How large is the spatial resolution of X-ray radiation as a function of acceleration
voltage and the material's average atomic number? (qualitative description only).
In general, the emission volume of X-rays increases with increasing acceleration voltage
and decreasing atomic number. Since the resolution is generally given by the dimension
of the signal emission volume, this means that the spatial resolution of the X-ray signal
becomes worse with increasing acceleration voltage and decreasing atomic number.
11 Problem e
The most common detector in SEM is the secondary electron detector (SED).
i. What energy (range) do secondary electrons have?
Secondary electrons are emitted electrons with energy from 0ev to 50ev.
ii. How can the SED be operated so that it does not detect secondary electrons, but
backscattered electrons?
By switching off the collector voltage (200-300V) and/or the scintillator voltage (10-12kV) the
secondary electrons do not gain enough (motional) energy to activate the scintillator. It must
moreover, be "free sight" for backscattered electrons so that they can hit the scintillator and thus
being detected.
ii. Why can't SED be used in low vacuum?
Both the collector voltage and the scintillator voltage are automatically switched off
when the pressure increases i the sample chamber. This is to avoid sparking. The air
molecules that are admitted into the sample chamber at low vacuum conducts current.
i. What are the bandwidths in an EBSP expression for?
The bandwidth corresponds to 2θB where θB is the Bragg angle in Bragg's law: 2dsin θB
= λ = constant. Where d is the lattice plane distance to h,k,l planes which for cubic
a
crystals is given by d= a is the lattice parameter d= 2 2 2
√ h +k +l
ii. What do intersections between bands represent?
Intersections between bands correspond to intersections between lattice planes which are
crystal directions (zone axes).
iii. Assume the EBSP above is from aluminum. What would the corresponding pattern look
like if the material had been nickel? Both aluminum and nickel have a cubic crystal
structure (FCC). The lattice parameters of aluminum and nickel are 0.405nm and
0.353nm respectively. Only qualitative description is required.respectively. Only
qualitative description is required. Replace with question text
If one replaces an FCC structure with another FCC structure, it is only the lattice parameter that
changes. The zone axes (crystal directions) do not change, mao will the intersections between the
bands remain unchanged. But since the lattice parameter a changes, so will the plane distance d
and thus θB changes, cf. Bragg's law. As θB changes, the bandwidth changes. The lattice
parameter of nickel (0.353) is smaller than the lattice parameter of aluminum (0.405nm), thus the
planar distances are also smaller for nickel. As d becomes smaller, sin θB must become larger
since 2ꞏdꞏ sin θB = constant. An increase of sin θB and thus increase of θB means that that the
bandwidths in EBSPs from nickel are larger. (θB is typically a few degrees, see calculation
example below:
Intro XRD
You have recently been employed as a material characterization specialist at SuSol Tech - a
Norwegian company research center for sustainable solar cell technology. You are part of a
project that looks at the use of TiO-based solar cells. You use an X-ray powder diffractometer
with Cu K ≈ 1.54 Å radiation and standard Bragg-Bretano setup.
13 XRD-1a (1p)
Your first task is to train the lab engineer (Engineer Stråle) to do powder diffraction in your new
lab. Engineer Stråle has taken a few measurements of the samples, but is struggling with
understanding what is what. In a few sentences, explain Bragg's law to Ingeniør Stråle.
In a few sentences explain Braggs law to the engineers. Braggs law sets the condition for
constructive interference of x-rays being scattered by atomic planes. If Braggs law is
fulfilled, we get a diffraction peak. The law is 2dsin = n, where d is the distance between
crystallographic planes,  is the incoming angle (2 is the outgoing angle) of the x-rays,  is the
wavelength and n is an integer for the order of diffraction
Above is a typical diffractogram Engineer Ståle has measured. Engineer Stråle has discovered
that every peak has a secondary peak of lower intensity at slightly higher angles. Explain in a
few sentences what could be the reason for this secondary peak.

The second peak is a product of K2 radiation (characteristics: slightly higher angles, half
intensity). We generate X-rays by bombarding with electrons a metal, in this case Cu, which
remove electrons from inner K shells, and other electrons are then free to drop from outer
shells. That produces radiation with the exact energy of the quantized shell. The K1 radiation
comes from electrons dropping from 2s to 1s, while K2 radiation comes from 2p to 1s. Unless
we have a monochromator we will get peaks from both these radiations. 1p (0.5 for Kbeta
instead

Mr. Ray, the engineer is also having problems with reproducibility. Above are some examples of
scans from the same batch of samples. What can be reasons the intensity and peak shapes are
changing between each measurement when the material is exactly the same?
Sample preparation is the issue here. If there is a powder valley you can get the intensities
wrong, if there is a powder mountain you could get wrong shapes and peak splitting.
XRD 2a (3p)

A colleague in your team, Dr Ana Tile, is testing different synthesis parameters for TiO powder.
You uses powder XRD to check the crystallinity of the samples. Above is three diffraction
patterns for wood different samples. The peaks are marked with a,b,c,d,e,f,g and h. 2 Print / PDF
18/34 By visually looking at the measurements, what are the differences between the three
samples and what does it indicate about the crystallinity of the samples?
Sample one has both amorphous and crystalline phases (big background and sharp peaks a,d, h),
Sample two seem single phase crystalline of high symmetry (a few sharp peaks b, c, e, g,),
Sample three seem a mix of crystalline phases, as it has all the peaks of both sample 1 and 2.
XRD 2b (2p)
TiO has several polymorphs, of which the two most common are Rutile (space group no 136
P4 /mnm, tetragonal, a=b= 4.653 Å, c= 2.969 Å) and Anatase (space group no 141 I4 /amd,
tetragonal, lattice parameters a=b= 3.802 Å, c= 9.748 Å).From the lattice system (first letter in
the space group) what Miller index do you expect that correspond to the first reflections
(diffraction peaks) of the two phases, anatase and rutile?
P indicates a primitive unit cell and so no extinction rules, I indicates bodycentered and so
h+k+l must be even. Hence Rutile would have the first reflection to be (100). For Anatase the
first reflection would be (110). 2p (must use selection rules to explain what we expect – not just
state the first reflections from the two)
XRD 2c (2p)
You create your structures in VESTA and find out that due to special extinction rules have not
rutile a (100) peak, but a (200) peak and that the first peak of Anatase is (101). Fromthe rest of
the lattice information given, calculate 2 the angle of rutile (200) and antase (101).
Info from task 2b): TiO has several polymorphs, of which the two most common are Rutile
(space group no 136 P4 /mnm, tetragonal, a=b= 4.653 Å, c= 2.969 Å) and Anatase (space group
no 141 I4 /amd, tetragonal, lattice parameters a=b= 3.802 Å, c= 9.748 Å).
Because both have a tetragonal unit cell we can use this formula to find d:

1 /𝑑2 =( (ℎ 2+𝑘 2 )/𝑎2 )+ (𝑙 2 /𝑐 2 )and then Braggs law 2dsin = . Rutile {200}: a= d = 4.653 Å,
2 = 38.66 1p Anatase {101}: d = 3.542 Å, and a 2 = 25.11 1p
Use what you found in 2c) (or from the internet if you failed c)). Which sample has anatase and
rutile phases?
Sample 1 and 3 has a peak at position a, which is just over 25 degrees and fit with our
calculations of Anatase (101). Sample 2 has a peak at 38,66 which fit with Rutile, this peak is
also possibly in sample 3. Suggesting that sample 1 is anatase and amorphous, sample 2 is
Rutile, and sample 3 is a mix og anatase and rutile.
XRD 3a (3p)
When you test the samples in a solar cell, it is clear that the antase phase is the best. Dr. Ana Tile
argue that the samples must have the most Anatase since the first peak (anatase (101)) has
highest intensity. Briefly explain whether this is correct or incorrect, by explaining what
determines the intensity of peaks in a diffractogram.
In general, a higher amount of a phase give rise to a higher intensity, however other factors has
to be taken into account. The intensity of a diffractogram peak is the absolute of the structure
factor squared. The structure factor consists of an atomic factor (intensity depends on the atomic
number of the atom) and the symmetry of the crystal. In this case the atoms are the same, but the
anatase phase has a higher symmetry giving rise to a higher intensity for the same planes. 3p
(points for the quality of reasoning, 3p for arguing right and wrong, 2p for choosing one and
giving good arguments, 1p for explaining factors that contribute to intensity)

You are trying to adapt the diffractograms in TOPAS with the Rietveld method. First you try
with just the Anatase phase and you get the result given which is the screenshot above. How will
you evaluate this adaptation? What will be your next step to improve your fit?
The fit is quite good, all the big peaks are accounted for and the intensities look right. However,
there are some small peaks not accounted for (for example at 28 degrees). You can try with a
small part Rutile. 4p (2p points for evaluation quality, 2p for next step)
XRD 3c (3p)
Another colleague, Miss Williamson-Hall, makes TiO2 nanoparticles - but does not understand
why The XRD peaks look worse (less intensity and wider peaks) when the particles get smaller?
Explain the different mechanisms for diffraction peaks to become wider, and how you can
distinguish between them?
Broadening of peaks come from 1) the instrument, 2) the size of the crystallites – in this case it
could be the size of the nanoparticles 3) strain in the sample; imperfection of the crystallinity,
defects, dislocations etc. The different cases have different dependence on angle.

You try to fit your data set with both the Rutile and Anatase phases and you decide to adjust all
the size and voltage parameters to obtain accurate data for the particle size. A screenshot of the
customization is shown above. What is the nano-particle size from this the adaptation? What
would you do differently to reduce uncertainty?
Anatase you get 68 nm (FWHM) and 0 strain this seems correct. For rutile you get 4.8 nm and
(cry size G is 9694 – which is physically not correct). This is due to the smacontribution from
Rutile. Would be better to not fit the size/strain for rutile. In general it’s a better idea to do one
parameter at the time and not all at once
4. A third colleague (Dr. Grace Ingalls) has made 1µm thick TiO2 thin films on LaAlO3
substrates. The films are expected to be polycrystalline, but textured. You measure the films both
in the conventional powder XRD setup and the Gracing Incidence setup. a) Explain briefly how
Gracing Incidence (GI) setup is different from the conventional powder setup (Bragg Bretano)
and why you would use it in this case?
GI the incoming angle (theta) is very shallow and fixed. Hence, we measure only the
surface/film and not the substrate.

Above is the measurement from the Gracing Incidence (GI) setup. It has been matched with the
Rutile database entry. Briefly explain what you can learn about the TiO2 thin film from the data
given. What can be the origin of unidentified peaks? How can you see if the film is textured?
You can use the indexing of peaks found in XRD 1 if you want.
The GI diffractogram has only the Rutile phase peaks + one peak that are is very sharp and not
found by the PDF for Rutile. That peak is the substrate which is a single crystal. No sign of
Antase peaks. The 110 peak (first peak ~27 degrees) and the 220 peak (at 57 degrees) is much
higher intensity than the PDF indicating a preferred orientation in that direction
Underneath is the measurement from the powder diffraction setup. The peaks marked with an
asterix belong to the substrate, while A and R denote Antase and Rutile respectively. Briefly
explain what you can learn about the TiO2 thin film when considering both the information in b
and c.
In the powder diffraction we see a peak from both antase and rutile, where the Anatase 004 is
quite pronounced, while the rutile 200 is weak but there. No sign of Rutile 110. This indicates
that the film is both phases in the bulk, while the surfaces is mostly rutile and with a different
orientation than in the bulk of the film.
XRD 5a (4p)
Colleague number 4, Dr. Rex Tinction, has found that synthesizing TiO2 under nitrogen is good
for absorption of solar rays. He wants to find out which reactions actually take place underneath
the synthesis. He has three different scenarios;
1) TiO2 reacts with nitrogen to form TiN
appear as a second phase with different peaks
2). The nitrogen atoms dope the TiO2 structure as interstitial atoms.
nitrogen interstitials would appear as a slight shift in lattice parameters due to strain in the
structure and possibly a broadening of the peaks as the crystallinity is less perfect.
2) TiO2 gets oxygen vacancies due to the reducing atmosphere. Describe in 3-5 sentences how
you can tell the difference between these three scenarios with XRD.
oxygen vacancies could lead to a different polymorph of TiO2, or if a lot of oxygen
disappears it would lead to TiO structure which would have different peaks
XRD 5b (2p)
You will investigate the reaction as it happens by doing an in-situ experiment. Explain someone
advantages of doing the experiment at the synchrotron?
Advantages include: higher flux which enables faster measurements with higher data quality,
and thus more measurements in less time (2D plate detector enables very rapid data collection);
higher brilliance and the possibility of doing spatially resolved diffraction if the samples can be
moved with a suitable stage; smaller or thicker samples can be measured
XRD 5c (3p)
The positions of the O and N atoms in the crystal structure are very important for understanding
the structure. But, when you do Rietveld fitting of the X-ray diffraction data from the in-situ
experiment you get large uncertainties in these positions. Explain why the uncertainty of the
O/N positions is much greater than for Ti?
The X-ray scattering cross section, or atomic scattering strength fi in the structure factor Fhkl, is
approximately proportional to the square of the atomic number, ~Z 2 . The atomic numbers of Ti,
N and O are 81, 7 and 8, respectively. This means that the O/N contribution to the intensity of
the Bragg reflection is very small compared to those of Ti.

Above are some of the results from the in-situ experiment, with diffraction measurements of
TiNOx as a function of temperature. The peaks annotated with Ana are from the anatase phase of
TiO2. Describe in 3-5 sentences what happens to the peaks as the temperature increases.
At first we have both TiO2 and TiNOx, as temp is increased the peaks shift to lower angles and
TiNOx peaks decrease in intensity as well as broadens. When we get to ‘350 degrees the
antatase peak quickly increases intensity, while the TiNOx decreases. Hence there is a phase
transition where the TiNOx decomposes to TiO2 anatase phase. The peak shift to lower angles
for anatase comes from the oxygen vacancies in the structure