AP621 Advanced Electron

Microscopy: Theory and

Practice
Dr. Zhao Jiong(趙炯)
Office: CD609; Tel: 27665692
Email: jiongzhao@polyu.edu.hk
 Today’s topics

 The class / your teacher

 The course
 Lec-01: Fundamentals of
electron optics
INTRO OF Dr. Zhao Jiong

3
 Learning Outcomes

Upon completion of the subject, students will be able to:

(a)Understand the fundamental theory of electron optics

(b)Understand the working principles of transmission electron microscopy(TEM) and
scanning electron microscopy(SEM)
(c) Identify the functions of analytical TEM
(d)Master the practical skills of TEM and SEM operations
(e)Select appropriate methods for particular purposes of real material characterizations
 Teaching & Learning
 Lectures/tutorials (~27h)//labs (~12h)
 References
1. Diffraction physics
John M. Cowley (ASU-Arizona State University)
Amsterdam: Elsevier Science B.V., 1995
2. Experimental high-resolution electron microscopy
John C. H. Spence. (ASU)
New York : Oxford University Press, 1980.
3. Transmission Electron Microscopy: A Textbook for Materials Science
Williams, David B., Carter, C. Barry, Springer, 2009 (Lehigh)
4. Electron microscopy analysis, lecture notes, 2005-2009, Yuan J., Yu R. (Tsinghua)

 Assessment
Continuous assessment: Assignments (20%)/
Laboratory Report (30%)/QUIZZES: (20%)
Final exam: 30%
Questions for this course
 How to describe the image formation
by wave theory
 How to account for the effect of lens
defects and operating conditions on
the image formed
 How to get a directional interpretable
high resolution images
 How to overcome the experimental
artifacts (lens aberration, etc.)

6
 Basic Optics and Coherent Waves (L1)

Fundamentals of electron optics

Basics and electron optics (L1, L2)

Electron diffraction (L2)

TEM (transmission electron microscopy) contrast (L2,L3)

HREM (high resolution electron microscopy) (L3)

Crystal defects and imaging

Defect analysis by diffraction and imaging (L4)

SEM (scanning electron microscopy), STEM (scanning transmission electron

microscopy), EDS (energy dispersive spectrum), EELS (electron energy loss
spectrum) (L5 and L6) Frontiers of Electron microscopy (research topics) (L7)

Operations of TEM, TEM sample preparations (L8)

TEM introduction

8
Rayleigh Criterion

λ is wavelength, µ is refraction index,

β is numerical aperture radius

For normal optical microscope, ~300 nm

9
 The wavelength of electrons
De Broglie wave
(matter wave)

10
gossips :
Brief History of electron microscopy
 1931 Ruska invented first TEM, and
awarded Nobel Physics Prize in 1986

11
TEM diffraction contrast
 P. Hirsh, A. Howie and M. J. Whelan
(1950s~1960s)
Cavendish Lab., U Cambridge

12
Micro/Nano beam diffraction and
atomic resolution TEM
 J. Cowley, ASU, 1960s~1980s

Some Chinese students and RAs in 1980s:

Zhu Jing, Zhang Ze, Wang Zhonglin, Peng Lianmao,
Zuo Jianmin…

student
(1970s)
S. Iijima

Nature 1991

13
STEM and atomic EELS
 S. Pennycook, Oakridge National Lab
(1980-1990s)
YBCO, Nature, 1988

14
Cs corrector commercialized

 M. Haider, H. Rose, K. Urban, CEOS

 (1980s~2000s)

15
New frontiers (atomic, in situ, low
voltage, chemical, electric, magnetic,
bio-…)
K. Suenaga et al (AIST) X. Zhong et al (Tsinghua)

16
The evolutions of TEM
1940s 1960s 1980s

1990s 2000s 2010s

17
Trend of the height of TEM
 Monochromator, Cs Corrector, Cc
corrector

18
Structure of A modern commercial TEM
JEOL 2010F

Electron SOURCE:
Field emission electron gun

ILLUMINATION SYSTEM
2 major condenser lens

Sample and stage

IMAGING SYSTEM
A strong objective lens

ENLARGING SYSTEM
Intermediate and projection lens

DETECTORS:
SCREEN+EYE, FILM and CCD camera
20
Conventional TEM

21
 CTEM Ray diagrams

Source

Illumination Imaging Enlarge

Specimen Camera
I III
II
 Lens modulation on incident wave(I)
 Lens modulation on exit wave(III)
Scanning TEM

23
24
25
STEM-EELS

26
In situ TEM

27
Basics of Electron optics

28
 Fundamentals of Light
 The wave form of light/electromagnetic wave
 Electromagnetic wave E(r, t ) , satisfy
∂ 2E
∇ E ( r,t ) =
2
εµ 2
∂t
 If neglect the vector field, Ψ ( r,t )
∂ 2
Ψ
∇ Ψ ( r,t ) =εµ 2
2

∂t
 The solution is
Ψ ( r, t ) =
Ψ 0 eiωt +ik.r
ω =
is frequency, k is wave vector, k2 εµω
= 2
c 2ω 2
For electrons (wave-particle duality)

 The wave form of electrons

 Electrons have the wave-particle duality, electrons can
be described by wave function Ψ ( r,t )

 Electrons satisfy Schrodinger equation：

2me ∂ψ ( r, t )
∇ ψ ( r, t ) + 2 V =
2
−i
 ∂t
 In vacuum（V＝0），solution is plane wave

Ψ ( r, t ) =
Ψ 0 eiωt +ik.r
ω is the energy of electrons， k is wave vector；
p = k is the momentum of electrons
Wave optics and coherent
sources
 Mathematical Description of Wave
Motion
Sinusoidal Wave
  x 
Ψ ( x, t ) = A cos ω  t −  + ϕ  = A cos(ωt − k x x + ϕ )
  v 

A＝wave amplitude 波的振幅

w＝angular frequency 波动角动量
v＝wave speed 波速
kx= wave number 波数 ＝2p/l

31
Waves in three dimensions：
plane waves
Ψ (r , t ) = A cos(ωt − k x x − k y y − k z z + ϕ ) = cos(ωt − k ⋅ r + ϕ )

Spatial position

Optical path
O
k
θ
P
wavefront

Spatial position

32
 Phase and Phase shift
 The amplitude of a running wave at each point is a complex
number of a fixed amplitude and a time-varying phase.

Ψ 0 exp(ik.r ) exp(iφ ( t ))
Ψ 0 exp(iωt ) exp(ik.r ) =
Ψ (t ) =
 The relative motion of wave can be considered as phase shift

Phase shift
 Young’s double slit Experiment

 Wave front division

 To create two coherent sources from a single
source, we can place a double slit screen at some
distance from a single source

S1 θ

S2
Q
Thomas Young (1773-1829)
Mathematics of Young’s double slit experiment (far field
theory)
 At far field, the spherical wave front is
approximated as a plane wave front
 Linear Superposition of two plane wave
fronts originated from the two point
sources S1
θ
d
E1 = E0 cos ωt S2
 2πd  Q
E 2 = E 0 cos ωt + sin θ 
 λ 
 πd   πd  2πd
E = E1 + E2 = 2 E0 cos sin θ  cos ωt + sin θ  ∆= sin θ
 λ   λ  C
λ
 πd 
Data 1
5

I = E = 4 I 0 cos 2  sin θ 
2

 λ  4

 Classically only see light in the forward

position (θ=0)
2

 Interference pattern shows light at finite

1

angle from the original propagation 0

-40 -20 0 20 40

direction of the wave Theta(degree)

 Light bends (diffraction)! Interference pattern

 Spatial Coherence
 The maximum separation over which two
points on the same wavefront can add
coherently to produce interference pattern.
 coherent angle size of an extended source (θc)
 Four source at an angle, the interference pattern
From coherent source
will also shift accordingly. For an extended
source of size d, interference pattern originated
from different part of the source will be added
incoherently and they can cause a loss of
visibility of the overall interference pattern.

From partially
coherent source
S1 θ
b
d λ
S2
θc = <
Q R b

R
Temporal coherence

 Any physical source can only emit wavelet of finite length. As

different wavelets are emitted randomly at different time, their
initial phase are completely uncorrelated. Interference
between different wavelets emitted from the same point in the
source will not interfere.
 coherence length
 The length of the continues wave trains
 Ordinary light source: 1-100 cm
 Electron source in a TEM
 coherence time
 The time taken from a continuous wavelet to pass
 Tc=Lc/v
 Source coherence and monochromaticity

 Heisenberg Relationship for the energy

uncertainty ∆E of the wave with its spatial
extent ∆t: ∆E∆t ≤ 

c c∆λ
 For light: E = ω = ; ∆E = −
λ λ2
 λ2 λ2
∆t ≤ ≈ ; L = c∆t ≤
∆E c∆λ ∆λ

 For electrons
 v
∆t ≤ ≈; L = v∆t ≤
∆E ∆E
 Electron Guns (Thermionic and Field emission)

thermoionic gun field emission gun

Real source

virtual source
LaB6 Thermionic gun

40
FE Gun

41
 Characteristics of electron sources

Units W LaB6 Field Emission

Work function, φ eV 4.5 2.4 4.5

Richardson’s constant A/m2K2 6x105 4x105

Operation Temperature K 2700 1700 300

Current density A/m2 6x104 106 1010
Cross-over size µm 50 10 0.01

Energy spread eV 3 1.5 0.3

Temporal coherence length

Brightness A/m2sr 109 5x1010 1013

Brightest source in universe

 Magnetic Lens
 Principle
 Lorentz force

vx
F = −ev × H F
H

 Long Magnetic Lens

 F⊥v， v＝constant
 Electron trajectory bending
 focus
 Magnification=1
 Short Magnetic Lens
 Focus ability (convex les)
 Hx，y
 Magnification
∞
1 e
= ∫ z dz
2
H
f 8mV −∞

 F⊥v F = −ev × H
 Rotation of beam
∞

∫
0.51 vx Fx
θ= H z dz Fy Hy
V
−∞ Hz vz
 Strong short lens
 Pole pieces

Pole-piece

Traditional Round Objective lens

46
 Huygens-Fresnel Principal
(Christian Huygens 1629-1695) Augustin-Jean Fresnel, 1788-1827

 Huygens suggests that the propagation of wave front can be

taken to be equivalent to superposition of secondary wavelets
originated from the existing wave front.
 Fresnel improved the Huygens intuitive description by allowing
the resulting intensity to be linear superposition of the
intensity of the all secondary wavelets.
 Huygens-Fresnel Principal
 Mathematical description
 The sample exit wave is represented
by the amplitude and the initial Ψexit (u , v )e iϕ (u ,v )
phase angle of the secondary waves
source

 The secondary wave propagates in

vacuum as a spherical wave 1 ik⋅r
e
r
 Fresnel diffraction pattern (near
field theory)

Ψexit (k ) = ∫∫ e Ψexit ( x, y )e iϕ ( x , y )dxdy

1 ik ⋅r
u ,v
r
 Fresnel Propagating Factor (near field theory)

 The intensity after a plane wave propagates

over distance R：
u 2 +v2
i 2πi
Ψexit ikR
e e Rλ
Rλ
Fresnel Propagation factor
 Aperture diffraction
 Airy Pattern and Airy
Disk
 Aperture limits high
frequency
information to
contribute to pass
through
Fraunhofer Diffraction of the Exit Waves (far field theory)

样品

J. V. Fraunhofer,1787—1826

Ψexit ( x, y ) φ (k x , k y )
FT (Fourier Transform)

φ (k x , k y ) ∝ ∫∫ Ψ (x, y )e
ixk x + iyk y
dxdy
 Diffraction

Near field vs Far field

Function of lens: shorten the distances

52
Fourier Transformation

53
Addition of Waves

54
Fourier Transform

55
Fourier Transform

56
Fourier Transform(FT)

57
2D FT

58
Properties of FT

59
Properties of FT

60
FT examples

61
Phase and Amplitude Reconstruction
by FT and FT-1

62
Convolution

63
Properties of convolution

64
Delta function

65
Delta function and convolution

66
 Abbe theory of image formation
 Image quality of a finite lens is non-perfect
because of the loss of details due to high
frequency spatial information
I3
G
O1
O
O2 I2
-G
O3
I1

Specimen plane Back focal plane Image plane

 The mathematical description of image formation

 Specimen exit wave: Ψexit ( x, y )

 First Fourier Transform（exit wave
Fraunhofer diffraction）

∫∫φ (k , k , z )e
1 ixk x + iyk y
Ψexit ( x, y, z ) = dk x dk y
2π
x y

 Lens：Focus diffraction on back focal plane φ k x , k y ( )

 Second Fourier transform （Back focal plane
Fraunhofer diffraction）

Ψimage (ξ ,η ) = C '
1
∫∫ φ ( p , q )e ipξ + iqη
dpdq ∝ Ψexit (− Mx,− My )
2π

Linear imaging system

The information flow in an imaging system

T(kxy): Transfer function

Channels
E I
X High f(requency) m
Specimen

FT
I Inverse FT a
T g
Low f e
W
A w
V High f a
E v
Input Output e

S (k ) S ' (k ) = T (k )S (k )
Ideal system：
T（k）＝1
Information transfer function and point spread
function

ℑ(T (u , v )) = h(ξ ,η )

T (u , v) h(ξ ,η )
FT

T=Aperture function
h＝Airy function
Image formation of an non-ideal lens

S BFP image

Ψexit ( x, y ) S (k x , k y )T (k x , k y ) Ψimage (ξ ,η )

S (k x , k y ) ∝ ψ (k x , k y ) = ∫∫ Ψexit ( x, y, z = t )e
− ik x x −ik y y
dxdy

Ψimage (ξ ,η ) = ∫∫ T (k x , k y )S (k x , k y )e
ik xξ + ik yη
dxdy ≠ Ψexit (− Mx,− My )
Physical origin of the components of transfer
function

 finite aperture
 Aperture function: 1 θ ≤ θ c
A(q) A(θ ) =
 Virtual apertures： 0 θ > θ c
 Electon source and camera
have finite sizes

 2θ 
4

 Cc： ECc ECc (θ ) = exp − (πλCc ) 

Chromatic Aberration coefficent （  2 
C c）

 θ2 
 Defocus: T∆f T∆f (θ ) = exp − iπ∆f 
defocus ∆f  λ

 Cs:TCs  θ4 
Spherical Aberration coefficient （Cs） TCs (θ ) = exp − iπCs 
 2 
 wave aberration
 A real imaging device produces a non-spherical wavefront
 The difference between the real wavefront and the ideal
spherical wavefront is called wave aberration
 Aberrations
 Calculate the ray diagram for
real lens with its actual shape
 Ideal focus requires a glass
lens to have parabolic
surface
 Most glass lenses have
spherical surfaces, leading
to imperfect focusing
 Alternatively a convergent
wavefront have to be spherical t
form a point image.
 Real lens has defects, so the
actual convergent wavefront
non-spherical or aberrated.
Chromatic Aberration
 Chromatic Aberration
Gauss focus
Objective plane High energy electron

Sustained Circle of
Angle θ least confusion

∆rc

∆r'c

Optical axis ∆zc

Low energy electron Transverse
Longitudinal aberration
aberration
Simple thin lens
77
 (Third order)monochromatic aberrations

 Third order approximation

ρ α3 α5  α 3 
= sin (α ) ≈ α − + − .... s ≈ r 1 − 1 − α + + ... 
r 3! 4!  3 ! 
 
 Mainly five kind of aberrations：
 Spherical aberration
 Coma
 Astigmatism
 Curvature of field)
 distortion
 Coma Aberration
Objective plane

Optical axis

Gauss focus
Simple thin lens
 Stigmatism
Gauss focus
Objective plane Ray in y-direction
y

∆rx
x

∆ry
Optical axis ∆f
Ray in x-direction Difference in
focal lens
Simple thin lens
Stigmation correction

81
 Spherical Aberration
Gauss focus
Objective plane Peripherial ray

Sustained Circle of
Angle θ least confusion

∆rs

∆rt
Optical axis
Paraxial ray ∆z
Transverse
Longitudinal aberration
aberration
Simple thin lens
Spherical Aberration

83
Cs correction

84
85
Early Cs Corrector (Tetragonal-
lens)

86
Cs aberration corrector
(dedicated STEM)

87
88
89
90
 Radial Order Azimutha
Typke and
Krivanek l
Dierksen Wave Nomenclature
notation Ray Symmetr
notation (k) y

C0,1 A0 0 1 1 Image Shift

C1,2 A1 1 2 2 Two-fold axial astigmatism (or axial astigmatism of the 1st order)

Defocus (overfocus positive, or spherical aberration of the 1st order;

C1,0 C1 1 2 0, ∞ Real numbers and describing rotationally symmetric contributions to
the wave aberration) (alt: Δf)
Three-fold axial astigmatism (or axial astigmatism of the 2nd order)
C2,3 A2 2 3 3

C2,1 B2 2 3 1 Axial coma

C3,4 A3 4 4 Four-fold axial astigmatism or axial astigmatism of the 3rd order Cs
Twofold astigmatism of Cs (or Third order twofold astigmatism, or
C3,2 B3 4 2
Axial star aberration of the 3rd order)
Third-order spherical aberration (always positive for round lenses
C3,0 C3 4 0, ∞ [3]; Real numbers and describing rotationally symmetric
contributions to the wave aberration) (alt: Cs )

C4,5 A4 5 5 Five-fold axial astigmatism or axial astigmatism of the 4th order

C4,1 B4 5 1 Fourth-order axial coma
C4,3 D4 4 5 3 Fourth order threefold astigmatism (or Three lobe aberration)
Six-fold axial astigmatism or sixfold axial astigmatism of the 5th
C5,6 A5 6 6
order
C5,4 R5 5 6 4 Fourfold astigmatism of C5 (or Fifth order rosette aberration)
C5,2 S5 5 6 2 Twofold astigmatism of C5 (or Fifth-order axial star aberration)
C5,0 C5 6 0, ∞ Fifth-order spherical aberration
D5 Four lobe aberration of the 5th order

91
Question:
 The wavelength of a slow neutron,
and that of X-ray are also atomic
dimensions. Why only electron
microscopy is commonly found ?
 Answer: only electrons can be easily
focused using electron optics