Cairo University Graduate Level

Faculty of Engineering Modern Metallographic Techniques

Department of Metallurgy Mahmoud Talaat Abdu, PhD

Modern Metallographic Techniques

Part III
Scanning Electron Microscope

by
Mahmoud Talaat Abdu, PhD
Assistant Professor, Metallurgical Engineering
Faculty of Engineering
Cairo University

Textbook:
J. Goldstein, D. Newberry, D. Joy, C. Lyman, P. Echlin, E. Lifshin, L.
Sawyer, and J. Michael, Scanning Electron Microscope and X-Ray
Microanalysis, 3rd Ed., Springer, 2007.

Page 1
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

Chapter 2
The SEM and Its Modes of Operation

2.1 SEM Subsystems and Their Functions

a

Fig. 2.1 (a) Two major parts of the SEM: the electron column and the electronics
console, (b) Schematic drawing of the electron column showing the electron gun,
lenses, the deflection system, and the electron detectors.

The SEM (Fig. 2.1 a) consists of two main components: the electron column,
and the electronics console.

Page 2
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

A. the electron column:

The electron column (Fig. 2.1 b) consists of an electron gun, two or more
electron lenses which affect the path of the electrons travelling down the
electron column, the base of the column is evacuated to about 10-4 Pa (~ 10-6
torr).

A.1. Electron Gun and Lenses Produce a Small Electron Probe

The electron gun generates electrons and accelerates them to an energy from
0.1 to 30 keV. The spot size from the electron source is too large to produce
a sharp image unless electron lenses are used to demagnify it and place a
much small focused electron spot on the specimen, see Fig. 2.1 b. The spot
sizes of most SEMs are less than 10 nm such that the electron probe contains
sufficient probe current to form an acceptable image. The beam emerges
from the final lens into the specimen chamber until it interacts with the
specimen to a depth of 1 μm and generates the different signals used in SEM
characterization.

A.2 Deflection System Controls Magnification

Fig. 2.2 Deflection system inside the final

lens. W = working distance between the
specimen surface and the final pole piece.

Page 3
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

The scanned image is formed point by point. The deflection system (Fig.
2.2) causes the beam to move to a series of discrete locations along a line
and then along another line below the first, and so on, until a rectangular
‘raster’ is generated on the specimen. Simultaneously, the same scan
generator creates a similar raster on the viewing screen. The deflection
system is located inside the final lens (objective lens). It consists of two
pairs of electromagnetic deflection coils (scan coils) that are used to sweep
the beam across the specimen. The first pair of coils deflects the electron
beam off the optical axis whereas the second pair of scan coils deflects the
beam back to the optical axis at the pivot point of the scan (Fig. 2.2). The
magnification (M) is the ratio of raster length on the viewing screen to that
on the specimen. For example, a 100-μm-wide raster on the specimen
displayed on a 10-cm-wide viewing screen generates an image magnification
of 1000×. When the SEM operator requests an increase in magnification, the
scan coils are excited less strongly, so that the beam deflects across a smaller
distance on the specimen. Note that the magnification depends also on the
working distance (W); see Fig. 2.2. In modern SEMs, the magnification is
automatically compensated for each working distance to assure that the
indicated magnification is correct.

A.3 Electron Detectors Collect the Signal

Fig. 2.3 Schematic diagram showing the collection of

SE and BSE by the Everhart-Thornley (EHT) detector.

When the electron beam impinges on the specimen, many signals are
generated as will be discussed later and any of these signals can be displayed
as an image by using the signals detectors. The electronics of a detection

Page 4
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

system converts the signal to point-by-point intensity changes on the

viewing screen and produces an image. The two signals ever used to produce
SEM images are the SEs and BSEs. An example of the detectors used to
detect the latter two signals is the standard Everhart-Thornley (EHT)
detector shown in Fig. 2.3. EHT detector collects either both SEs and BSEs
or only BSEs depending on the bias of the collector grid. If the grid has a
positive bias (+300 V), it collects both SEs and BSEs for topographical
images. When the grid has a negative bias (−100 V), it repels the SEs
(having weak energy) and collects only BSEs (having strong energy) for
chemical or orientation contrast images.

B. Electronics Console ‫وحدة التحكم‬

In modern SEMs, the electronics console (Fig. 2.1) consists of a computer
connected with a key board containing the operators’ controls, a viewing
monitor, and a mouse and printer. The electronics console allows the
operator, via specialized software, to adjust the operations parameters and
digitally collect and store images and EDS/WDS analyses and maps, in
addition to collecting EBSD patterns and processing them, using special
software, orientation maps.

B.1 Camera or Computer Records the Image

Older SEMs have a separate CRT (Cathode Ray Tube) for slow-scan
recording of images onto photographic films or, alternatively, a video printer
to produce a hard copy of the CRT images. Modern SEMs store images
digitally in a computer for later processing and printing.

B.2 Operator Controls

First, the SEM operator has to learn how to align the electron beam. Second,
he/she has to master the accelerating voltage, emission current, and working
distance controls. Third, the lens controls come: the condenser lens control
determines both the amount of beam current available and the minimum
beam size, and the objective lens control allows the beam to be focused so
that the smallest diameter is located exactly at the specimen surface. Finally,
the contrast control has to be adjusted for the too light and too dark areas on
the screen to be darkened or lightened, respectively, until a sound image
appears on the monitor.

Page 5
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

2.2 Electron Guns

The electron guns provide a stable beam of electrons of adjustable energy.

There are two main types of electron guns: thermionic emission electron
guns and field emission electron guns. Older SEMs use tungsten or LaB 6
thermionic emitters, but, increasingly, new microscopes are equipped with
cold, thermal, or Shottky field emitters because they provide enhanced
performance, reliability, and lifetime. Thermionic emission depends on the
resistive heating of the filament to high temperatures to enable the electrons
in the cathode to overcome the work function and escape into vacuum.
Conversely, in field emission guns, a negative field is applied to a sharp tip
of the emitter. When the field reaches 10 V/nm, the electron can ‘tunnel’
through the tip material and escape into vacuum.

2.2.1 Tungsten Hairpin Electron Guns

The tungsten electron gun is the most common of electron guns and consists
of three components: tungsten wire filament (hairpin) serving as the cathode
(negative electrode), grid cap (Whnelt cylinder) serving as a control
electrode, and the anode plate (positive electrode) as shown schematically in
Fig. 2.4. These components are maintained at different electrical voltages by
appropriate connections to the high voltage supply, which is variable in the
range 0.1–30 kV. For example, if the accelerating voltage is set to 20 kV, the
filament will be set at −20,000 V with respect to the anode, which is at
ground potential. The grid cap is set to a voltage slightly more negative than
the filament to provide a focusing effect of the beam.

2.2.1.1 Filament

Although tungsten hairpin is now over 70 years old, it remains the best
choice for low-magnification imaging and x-ray microanalysis because it is
reliable, easy-to-understand, and inexpensive. The filament is a tungsten
wire, about 100 μm in diameter, bent into a V-shaped hairpin with a tip
radius of about 100 μm; see Fig. 2.5. Thermionic emission produces
significant beam current only when the filament is at white heat. Thus, the
filament is resistively heated by a current if to a temperature of 2000−2700
K. Thermionic electrons are then emitted from the filament tip over an area
of 100 μm × 150 μm.

Page 6
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

Fig. 2.4 Schematic diagram of the conventional, self-biased, thermionic

tungsten hairpin electron gun.

Fig. 2.5 Conventional tungsten hairpin filament electron gun. (a) Filament wire
spot-welded to support posts, (b) tip of wire showing wire drawing marks, and (c)
exploded view of electron gun showing filament assembly, grid cap, and anode
displaced from one another.

Page 7
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

2.2.1.2 Grid Cap

Emitted electrons spread out into a broad cone from the filament emission
area; see Fig. 2.4. The grid cap (Whnelt Cylinder) acts to focus electrons
inside the gun to a crossover of diameter do and a divergence angle αo
between the grid cap and the anode, and to control the amount of current
emission (Fig. 2.4). Later, we shall see that condenser and objective lenses
produce demagnified image of this crossover to a final probe diameter dp on
the specimen surface.
The grid cap is connected to the filament by a variable resistor. The
emission current leaving the filament is replaced by an equal current to the
filament through this resistor. This generates a negative bias between the
grid cap and the filament. Fig. 2.4 shows lines of constant electrostatic field
potential varying from negative to zero to positive. Electrons move toward
positive potential, so they leave the filament where positive electrostatic
field lines meet the filament surface.

2.2.1.3 Anode

Emitted electrons in the gun are accelerated from the high negative potential
of the filament (e.g. −20,000 V) to ground potential (0 V) at the anode. A
hole in the anode allows a fraction of these electrons to proceed down the
column toward the lenses. Electrons collected on the anode and elsewhere
down the electron column return via the electrical ground to the high-voltage
power supply.

2.2.1.4 Emission Current, Beam Current and Probe Current

The important parameters for any electron gun are the amount of current it
produces and the stability of that current. The emission current (ie) is the
total current emitted by the filament. The beam current (ib) is the amount of
current that leaves the gun through the hole in the anode. At each lens the
beam current becomes smaller and it becomes several orders of magnitude
smaller when it is measured at the specimen as the probe current (ip).

2.2.1.5 Operator Control of the Electron Gun

Because the information from the SEM is recorded serially as a function of

time, the probe current must be the same for each point in the image. To
Page 8
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

ensure a stable beam current, the filament heating current if is adjusted for
‘saturation’ condition where a small increase or decrease in the filament
current if will not change the beam current ib because a plateau has been
reached, see Fig. 2.6. At saturation electrons are emitted from the tip of the
filament and are focused into a tight bundle under the influence of a negative
bias voltage.

Fig. 2.6 Saturation of a tungsten hairpin electron gun. At low filament heating some
electrons are emitted into various directions from the unsaturated filament (left
filament image). At higher filament heating, near saturation, electrons are
concentrated into a tight bundle (right filament image). Further filament heating
provides a slight increase in beam current. The beam should be operated just at the
saturation point. The images of the filament are taken via the filament imaging mode
available in most SEMs.

Note that the gun should operate just at the saturation point; further
increase in the filament heating by further increasing the filament current
(called oversaturation) will not affect the beam current but it will reduce the
filament lifetime.
Saturation is a self-regulating negative feedback process that ensures a
stable beam current. If the filament heating current if increases, then the
filament emission current ie will increase. The increase in ie causes a further
increase in the negative bias voltage of the grid cap, which opposes the
increase in the emission. This self-regulation of the electron gun warrants a
stable beam current. Each accelerating voltage requires a different setting to

Page 9
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

achieve filament saturation. Fortunately, Modern SEMs compensate for this

saturation automatically once the operator selects the accelerating voltage.

2.2.2 Electron Gun Characteristics

The electron gun characteristics measure the electron gun performance and
involve electron emission current, brightness, lifetime, source size, energy
spread, and stability. Brightness is the most important parameter because the
image quality at high magnifications depends on this parameter.

2.2.2.1 Emission Current

ie ~ 10 μA for tungsten hairpin, and ie ~ 10−30 μA for field emission sources.

The portion of emission current leaving the anode is the beam current ib; ib is
smaller than ie, and decrease further with moving down the electron column.
Therefore, the emission current cannot be considered as a sufficient judging
parameter for the electron gun performance.

2.2.2.2 Brightness

Brightness is defined as the beam current per unit area per solid angle. It is
important because it is “conserved”, i.e. the brightness at any point down the
column is the same as the brightness at the electron source itself even as the
individual values of ip, d, and α change. The brightness β can be calculated
from the following equation:

( )

The subscript p denotes the probe where ip in A is the probe current, dp in cm

is the probe diameter, and αp is the probe convergence angle, see Fig. 2.7,
the solid angle is measured in steradians (sr). Brightness equation is also
important because it helps the microscopist develop a practical
understanding of electron-optical limitations on SEM images. The
theoretical brightness of a tungsten hairpin ~ 105 A/cm2.sr at 20 kV.
Brightness increases linearly with the accelerating voltage, so at 10 kV the
electron source is 10 times brighter than at 1 kV. For thermionic guns, the

Page 10
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

brightness is optimized at the saturation point and varies with bias voltage
and filament-to-grid cap spacing, which is why SEM manufacturers provide
detailed instructions on for optimizing these adjustments.

Fig. 2.7 Four different electron beam parameters defined when the electron beam
impinges on the specimen: electron probe current ip, electron probe diameter dp,
electron probe convergence angle αp, and electron beam accelerating voltage Vo.

2.2.2.3 Lifetime

Because the filament operates at white-hot temperatures, it gradually

evaporates with time. Eventually the tungsten wire becomes thin and fails.
Oversaturating the filament increases the evaporation rate causing premature
failure of the filament.

2.2.2.4 Source Size, Energy Spread, and Beam Stability

do ~ 50 μm for tungsten hairpins at the crossover, and depends on the gun

configuration and operating conditions. Larger do means higher
demagnification is required to achieve the required SEM resolution. More
advanced electron emitters have smaller do, and thus considerably higher
brightness, e.g. LaB6 sources have do ~ 5 μm and field emitters have do ~
5−25 nm.
The electron beam energy spread ΔE is the spread of electron energies
leaving the filament. ΔE ~ 3.0 eV for tungsten hairpin, ~ 1.5 eV for LaB 6
emitters, and ~ 0.3−0.7 eV for field emitters. This parameter is the major
limiting factor for low-voltage operations.
The stability of an electron gun is a measure of how constant the
electron emission is over periods of minutes or hours. The most stable

Page 11
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

source (1%/h) is Shottky field emitters (SFEs), but all thermionic emitters
possess good performance. Cold field emitters (CFEs) have poor stability.
2.2.2.5 Improved Elecron Gun Characteristics

Better electron guns should have higher brightness to reduce the electron
probe size without losing the electron beam current. Higher brightness can
be achieved by changing the emitter material (to LaB6 instead of the
tungsten hairpin) or the mechanism of emission (field emission instead of
thermionic emission). These advanced electron guns are replacing the
conventional tungsten filament where the highest performance is required.
Table 2.1 compares the major electron gun characteristics for the various
electron sources.

2.2.3 Lanthanum Hexaboride (LaB6) Electron Guns

LaB6 emitter is a tiny block of LaB6 single crystal ~ 100 μm in diameter and
~ 0.5 mm long (Fig. 2.8). LaB6 has a lower work function than tungsten and
do not need to reach saturation as tungsten filaments. Thus, LaB 6 filaments
have 5−10 times greater brightness and a longer lifetime than tungsten
filaments. As the case for all thermionic emitters, the filament is supported
and resistively heated by either graphitic carbon or rhenium, because they
don’t react with LaB6. LaB6 filaments are usually polished to a 1-μm tip
radius, to ease the escape of electrons from the crystal. The vacuum for these
filaments has to be better than 10-4 Pa (10-6 torr), so differential pumping of
the gun with an additional ion pump is usually required.

Page 12
Cairo University Graduate Level
Faculty of Engineering Modern Metallographic Techniques
Department of Metallurgy Mahmoud Talaat Abdu, PhD

The brightness of LaB6 sources varies from 5×105 A/cm2.sr to 5×106

A/cm2.sr at 20 kV depending on the sharpness of the tip. Sharp tips provide
higher brightness but have shorter lifetimes; blunt or truncated emitters have
lower brightness but longer lifetimes. LaB6 emitters cost 10 times more than
tungsten hairpins, but this higher cost is justified in the higher brightness
(~10 times higher brightness at 20 kV) and longer lifetimes (> 1000 hours
compared to 40−100 hours for tungsten filaments) of LaB6 filaments.

Fig. 2.8 LaB6 source. (a) LaB6 source mounted on support, (b) higher
magnification showing the finely ground tip with 10-μm radius, (c) Failed
LaB6 filament with evidence of evaporation and oxide formation.

2.2.3.2 Operation of LaB6 Source

There is no saturation condition for LaB6 filaments because the high electric
field at the tip eases the escape of the electron into vacuum.

Page 13