Transmission Electron Microscopy

Dr Heath Bagshaw CMA

[email protected]

What is the difference between SEM and TEM?

Scanning Electron Microscope

Transmission Electron Microscope

In an SEM, electrons are scanned over the surface of the sample.

In a TEM, electrons are transmitted through the sample.

What is a Transmission Electron Microscope (TEM)?

Electrons pass through a (very thin) sample (i.e. are transmitted) to form an image.
Simplistically, In its operation a TEM can be thought of as analogous to a slide
projector:-

Slide projector
Projection screen
Objective lens

slide

Condenser lens

Light source

TEM
Fluorescent
screen
aperture

Specimen (thin)
Electron beam

Electron source

Objective lens

Condenser lens

What is a Transmission Electron Microscope (TEM) cont.

In a slide projector white light from the bulb is focussed onto the slide by a condenser
lens.

This produces an image which is focussed by the objective lens to a point

(crossover).
The image is then projected onto a screen and magnified by varying the
distance from the crossover to the screen.

In a TEM the electron beam is focussed on the sample using the condenser lens
system.
This produces an image which is focussed by the objective lens to a point
(crossover).
This image is then magnified by a series of projector lenses to vary the size
of the image on a fluorescent screen.
Changing the current of an electromagnetic lens alters its focal length

altering magnification.

Anatomy of a TEM
TEMs have basically the same design independent of manufacturer or age:-

a)

b)

Sample

Electron
Gun
c)

Condenser
lens system

Apertures

Sample

Apertures

Viewing
Screen

Examples

of

Transmission

Electron

Microscopes, a)Philips EM420 (1980s), b)

Cut through Schematic of Philips CM 200
(1990s) and c) JEOL 2100 (2000s)

Viewing
Screen

JEOL 2100 has digital imaging and

computer assisted operation.

main components of a TEM

The TEM can be broken down into a few main components, these are:The Gun which produces electrons.
The condenser system which forms the probe.
The sample sample preparation is important, and time consuming.
Image formation use of image plane or back focal plane.
Projection of the image (magnification).
Viewing and recording the image.

Lets start by looking at how electrons are produced.

Electron Guns
Two main types of gun Thermionic and Field Emission.

Thermionic sources produce electrons when heated.

Field emission sources produce electrons when exposed to an intense
electric field.
FEGs give much more brightness than thermionic systems.
FEGs give a more monochromatic

electron source and finer probe (i.e. better

resolution).

Comparison of the three types of source operating at 100kV

The Condenser System

The Wehnelt (or 2nd anode in a FEG) focuses the
beam to a crossover which is accelerated down the
column.
The first condenser de-magnifies the crossover to
give a smaller point source this is referred to as C1

or spot size.
The second condenser lens (C2) is used to either
converge or spread the beam of illumination on the
sample (intensity (Philips) or brightness (JEOL)).
A condenser aperture is placed in the beam path
to remove electrons far from the optic axis which
would reduce resolution.

The smaller the aperture the better the resolution,

but there is an associated decrease in brightness
need to compromise.

The Sample
Samples are typically 3mm in diameter and <80m thick.
Sample preparation is tricky and time consuming.
For a metallic sample:Cut\slice a section of material <1mm thick.
Produce 3mm diameter blanks by either Punching, or spark erosion.
Grind and polish blanks to <80m thick and 0.25m or better finish.
Either electropolish or, if a light metal Ion Beam mill to perforation region
around the hole should contain electron transparent material (thin area).

Thin area

For Al based alloys, due to build up of Al2O3 coatings in the presence of air, the
samples should be examined immediately after preparation (e.g. prepared in the
morning and examined in the afternoon).

The Sample (2)

For a non metallic sample:-

Cut\slice a section of material <1mm thick.

Mount on glass slide and Grind to <80m thick with a 0.25m or better finish.
Mount on support grids if necessary.
Dimple to leave ~10 m of material remaining.
Ion beam mill to perforation.
Some samples require coating to prevent charging effects.
For powders:-

Grind\mix the powder in chloroform or reagent grade methanol to create a

suspension.
Pipette a drop of suspension onto a support grid with a Carbon film coating.
Solvent will evaporate leaving powder dispersed on the grid.
Allow to dry and observe.
Biological samples require fixing and embedding before being stained with heavy metals
(e.g. OsO4) for contrast prior to ultra microtome sectioning. Very time consuming.

Focused Ion Beam (FIB)

Uses a beam of focused Ga ions to sputter material away.
Fairly new technique (~10-15 years old) which allows site specific micro machining to cut
out a specific area of a bulk material for TEM analysis.
Usual process is the lift out technique, a) to d).

e)
Another common method is the
trench milling technique, e).
More on this tomorrow...

Image Formation
sample

All rays from a point in the object are gathered

by the lens and converge to a point in the
image.
All parallel rays are focused in the focal plane.
The back focal plane of the objective lens

Objective
lens

contains groupings of rays that have left the

object at the same angle.
The back focal plane contains the diffraction
pattern of the sample.
Diffraction pattern and image are both formed
in the imaging process
The intermediate lens is then focused on
either the image plane (for the image), or the
back focal plane (for the diffraction pattern).

Imaging Modes

Two principle modes of TEM operation, A

Projecting the diffraction pattern,

B Projecting the image.

The intermediate lens selects either the Back Focal Plane or the image plane of the
objective lens.

Contrast
Conventionally, in order to get contrast in an image an objective aperture is placed in the
beam path at the back focal plane of the objective lens.

Any electrons which are scattered by interaction with the sample are prevented from
recombining in the image by the aperture.
Any regions where electrons are scattered will appear dark in the image.
The smaller the aperture used the greater the contrast, but at the expense of brightness.
Another method of achieving contrast is the Mass thickness
Contrast mechanism.
Thicker (or higher Z) areas of the sample scatter more electrons
than thinner (lower Z areas).
Therefore fewer electrons from the dark region fall onto the
screen, so the region appears darker in a Bright Field image.
Mechanism of mass
thickness contrast

Projection - Magnification
A series of projector lenses are then used to magnify the image formed by the
intermediate lens onto a viewing screen.
Electron Microscope lenses are Electro-magnetic in nature.

They consist of a cylindrical soft metal core (polepiece) with a hole drilled through it (bore)
wound with copper wire.
When a current is passed through the coils a magnetic field is created in the bore.
Changing the current in the windings changes the
magnetic field and effectively changes the focal
length of the lens.
Increase the current and the focal length (f) of the

lens decreases, so a weaker lens (f1) gives a higher

magnification then a stronger lens (f2) as image
distance v increases, but the object distance is
unchanged.

Viewing\Recording Images
The image, comprising of a beam of focused electrons falls onto a fluorescent screen
therefore producing a visible image.
This image can be recorded onto photographic film using the camera located underneath
the viewing screen.
Modern machines can capture images digitally using CCD cameras.

The signal from the beam can be output to external analysis equipment, e.g. STEM
detectors, EELS etc. for further analysis
Fluorescent viewing
screen

Digital CCD camera

Other analysis Techniques

One of the main strengths of TEM is the ability to obtain diffraction (structural information)
and images from the same area at the same time from the same sample from a small
volume.
There are, however many more analysis techniques available:Energy Dispersive Spectroscopy (EDS) for compositional analysis.
Electron Diffraction for crystallographic analysis.
Dark Field imaging enhanced contrast using crystallography.
High Resolution TEM (HRTEM) with resolution to atomic scale.
Scanning TEM (STEM) for increased resolution\contrast.
Electron Energy Loss Spectroscopy (EELS) to examine local electronic environment
allowing mapping of elemental distribution to atomic level.
Tomography to produce 3 dimensional models of samples.
Low dose and Cryo TEM to allow analysis of beam sensitive materials e.g. polymers and
biological samples.

Energy Dispersive Spectroscopy (EDS)

EDS analysis in the TEM can be performed on very small areas due to very small
interaction volume.
In SEM the interaction volume is ~ 1m3 from a 0.1m probe size.
In TEM the interaction volume is of the order of nm (~10-8 m3 for a FEG TEM) due to very
thin samples and very small probe sizes.

Comparison of interaction volumes in an SEM, Thermionic TEM and FEG TEM showing
much smaller interaction volume for TEM.

Electron Diffraction
Diffraction patterns give crystallographic information about a material from site specific
small volumes, unlike XRD which is a bulk analysis technique.
Can determine if a material is amorphous, crystalline or poly crystalline quickly and
effectively.
a)

b)

c)

a)Amorphous material, b) Crystalline material and c) polycrystalline material

Lattice parameter information, unit cell details and any periodicity present in the sample
arising from structural transformations or ordering processes can be seen.

Superlattice structure gives information about the long range make up of a material.

SuperLattice Reflections
Patterns a) to c) are electron diffraction patterns taken from CaTiO3 super lattice reflections
are seen showing the presence of Cation Shifts (cs), In-phase tilting (it) and antiphase tilting(at).
a)

000

001

b)

010

c)

0-11

at

001

1-10

000

cs

at

-110
It(ii)
It(i)

cs

001

e)

d)

f)

010
cs
cs

000

10-1

000

1-10

o2 000

001
a
t

o1

o1

a
t

it

0-11
o1

000
o2

-110

10-1
it

o1

it

Patterns d) to f) are electron diffraction patterns from Sr(Zn1/3)Nb2/3)O3 the same superlattice
reflections are seen as in CaTiO3, but with additional ordering reflections (o1 and o2).
Both CT and SZN have the same tilt system (a-a-c+) so are related, but SZN has long range
ordering as well.
CT space group is Pnma (orthorhombic), SZN is P21/c (Monoclinic - = 125.501).

SuperLattice in XRD and ED

In XRD traces the appearance of superlattice reflections is
influenced by the size difference of the cations.

Ordering Peak
Intensity

Intensity
2Theta

SMN ordering
peak in XRD

The ordering peak (~20 2 Theta) in SMN is much larger

(Approx. 12X) than the same peak in SZN.
2Theta

SZN ordering
peak in XRD

Consequently, the ordering peak in SZN is easy to overlook.

This is not a problem in Electron Diffraction.
Major perovskite
reflection

Major perovskite
reflection
(hk0)In phase
tilting reflection
<111> electron diffraction
pattern of CT

(hkl)anti phase
tilting reflection
1/3

<110> electron diffraction

pattern of SMN

(hkl)ordering
reflection

Superlattice reflections are clearly visible in electron diffraction patterns.

Electron diffraction therefore provides a powerful tool for the investigation of
superlattice reflections in materials (e.g. Tilting, ordering and layered structures).

Symmetry Analysis
Convergent Beam Electron Diffraction (CBED) is a powerful tool for analysing and
identifying symmetry in materials.
a)

b)

a)Convergent Beam Electron Diffraction (CBED) pattern showing Six fold rotational
symmetry and b)Selected area Diffraction Pattern (SADP) of same area.
Unknown materials can be analysed and their space groups be found to aid
identification\characterisation.

Symmetry Analysis 2
Overlaying of CBED patterns gives a map of the crystal which you can navigate around to
find the region\ zone axis (direction) of interest.
<111>

<110>
Overlayed CBED patterns for a FCC system.

<100>

Dark Field Imaging

Dark Field images are formed from specific diffracted beams, giving orientation contrast.
Bright Field images are formed using spot A.
A typical Dark Field images would be formed using spot B.
Bright field images contain information from all electrons that have passed through the
sample therefore subtle features are hidden.
With Dark Field images, a specific plane can be viewed in order to see specific
crystallographic features such as dislocations, regions of ordering or strain fields etc.
Dark field images typically have considerably higher contrast than bright field images
although the intensity is greatly reduced.

Dark Field Image formation

Can select the spot by moving the aperture or tilting the beam to bring the selected
spot to the optical axis of the microscope.

Image formation for a Bright Field, Displaced Aperture Dark Field (DADF) and Centred
Aperture Dark Field (CADF) image.

Dark Field Imaging Example

c)

a)

b)

001
at

000

110

d)

010
000
1.0 m

10-1

cs

Image a) illuminated with Cation Shift (cs) spot in c).

Image b) illuminated with Antiphase tilt (at) spot in d).
c) and d) both <110> zone axis diffraction patterns (ZADP).
Clear crystallographic difference between the two area is seen visually.

1.0 m

Dark Field Imaging Example 2

b)
a)

300nm

Image a) Dark Field image of unshadowed and unstained DNA, Image b) illuminated
with ordering superlattice reflection regions of ordering clearly visible.

Dark Field Advantages and Disadvantages

Advantages and disadvantages of Dark Field Imaging :Advantages

Provides high contrast for examining molecules with very low contrast such as
DNA.
For crystalline objects, specific diffraction spots can be selected in the back
focal plane of the objective lens in order to form a dark field image only from
the electrons scattered by a chosen set of crystal planes.
Disadvantages
More difficult to focus and correct for astigmatism since phase contrast is not
present.
Image brightness is low, since the objective aperture transmits only a small
fraction of the scattered beam,,
Longer exposure times needed to get good photographic images.

Consequently, specimens are subjected to greater levels of radiation damage.

High Resolution TEM (HRTEM)

High resolution TEM images are formed by a process known as phase contrast.
Contrast is formed by constructive or destructive interference between electron wave
passing through a material.
No objective aperture is used.
The appearance of the image can change by altering the focus.
Perfect focus is known as Scherzer defocus all electron waves are in phase no
contrast.
As focus is altered (either de-focused or over focused) the waves move out of phase
giving rise to interference patterns.
These patterns form the actual image.
Image simulations are then needed to verify interpretation on the image recorded
very involved.

High Resolution TEM Examples

HR 500kV lattice fringe image of as

grown silicon on sapphire showing
complex twinning and stacking
defects

Structure image of <100> tilt grain

boundary in nickel oxide.

Images from HRTEM and associated techniques

Buseck, Cowley and Eyring Oxford science pub

High Resolution TEM Image Simulation

Simulated HRTEM images showing how features change as the focus changes (f).

Scanning TEM (STEM)

In conventional TEM imaging an objective aperture is used to select the transmitted or
scattered beam in order to select the source of contrast.
In STEM the transmitted or scattered beam is selected in an equivalent way by using
detectors rather than apertures.

Comparison of the use of an objective aperture in TEM to select the transmitted or

scattered beam forming BF or DF image respectively
In STEM an on-axis, or annular detector is used to perform equivalent operations.

Why use STEM?

For DF imaging the annular detector collects more electrons than an aperture.
STEM ADF images are less noisy then TEM DF images as no lenses are used to form them.
Contrast in STEM images is greater than standard DF images.
a)

b)

c)

Comparison of TEM DF and STEM ADF images of the same sample shows clear contrast
difference
The low contrast in the TEM image (a) is enhanced in the STEM image (b) by signal
processing.
Image c) is a digitised TEM DF image treated with contrast enhancement software.

EELS

Electron Energy Loss Spectroscopy.

As electrons interact with matter they
lose energy to it.
A magnetic prism disperses these

electrons according to energy, then a

spectrometer located at the correct
position can collect electrons of a specific
energy.
Can not only detect positions of
individual atoms, but also the oxidation
state.

EELS - Examples
Typical EELS spectrum showing absorption edges
for three different elements, different oxidation

states

give

different

shaped

edges,

takes

experience to analyse properly.

Elemental distribution maps for C, Ti, O and Cu , maps taken by filtering particular edges.
Can clearly see atomic distribution at nm scale.

Tomography
Sample rotated through large angles in the electron beam and photos taken at each angle.
Powerful computers then used to render 3D model.
Very time consuming process, automation software such as Leginon* is often used for
image capture process.

*(The Scripps Research Institute Journal of Structural Biology, Vol 151, Issue 1, P 41-60)

Cryo TEM
To reduce damage to sensitive organic samples (biological\polymeric) Cryo TEM technique
is used.
Samples are prepared and analysed at low temperatures (usually liquid Nitrogen -196C) to
minimise heating effect of electron beam.

Longitudinal section through rat ventricular myocytes. A- Nucleus, B- Z-Line, C- A- Band

and D Mitochondria*
*Dept Cellular Sciences emlab.ubc.ca www.emlab.ubc.ca/sub/gallery.html

Conclusions
TEM is a very versatile analysis technique.
Many different types of analysis can be performed (hardware dependent).

Complimentary information can be obtained from distinct small (nm) regions allowing full
nano-scale characterisation.

Sample preparation is very time consuming requiring expertise to get the right results.
Sampling ratio is poor the sample is 3mm in diameter and ~ 80-100m thick is this
really representative of the sample ? have to assume it is.

Future improvements, such as aberration correctors and monochromators will give finer
control of probe sizes leading to higher resolution.

ED example - Perovskites
The prototype perovskite structure with general formula ABO3 can be described as a cubic
arrangement of corner-shared BO6 anion octahedra enclosing a single A-site
cubooctahedral cation arranged within a three dimensional O2- network.
The perovskite structure is able to sustain a wide range of cation sizes for both the A and B
sites.
The most well known perovskite is BaTiO3 although the

name derives from CaTiO3.

The most common phase transitions in perovskites involve
rotation, or tilting, of the corner shared octahedra making

The cubic unit cell of BaTiO3

up the structure.
This results in a doubling of the unit cell and gives rise to a reflection in reciprocal space
at a integer position.

Schematic diagram illustrating octahedral tilting around an axis perpendicular to the paper.