CHARACTERIZATION OF NANOSTRUCTURESCHARACTERIZATION OF NANOSTRUCTURES HRTEMHRTEM

ANANDH SUBRAMANIAMDepartment of Applied Mechanics

INDIAN INSTITUTE OF TECHNOLOGY DELHINew Delhi-

110016Ph: (+91) (11) 2659 1340, Fax: (+91) (11) 2658 1119

anandh333@rediffmail.com, anandh@am.iitd.ernet.inhttp://web.iitd.ac.in/~anandh

March 2007

The entire presentation is available at:

http://web.iitd.ac.in/~anandh/Nanotech2007.ppt

Transmission Electron Microscopy

David B. Williams and C. Barry Carter

Plenum Press, New York, 1996.

Monographs in Practical Electron Microscopy in Materials Science

J.W. Edington

Philips, Eindhoven, 1974.

High-Resolution Electron Microscopy for Materials Science

Daisuke Shindo

and Kenji. Hiraga

Springer-Verlag, Tokyo, 1998.

Experimental High-Resolution Electron Microscopy John C.H. Spence Oxford University Press, 1988.

High-Resolution Electron Microscopy John C.H. Spence Oxford University Press, 2003.

REFERENCESREFERENCES

When do we see anything ? –

Contrast

How do we see fine detail ? –

Resolution

How can we see better ? –

The TEM

What is meant by ‘High-Resolution’?

What things are there to be ‘seen’

? –

Imaging (as we go along)

HRTEM

Nanostructures by HRTEM: Image Gallery

OUTLINEOUTLINE

HRTEM

= High-Resolution Transmission Electron Microscopy

How can we define nanostructures in a general way ?

What do we want to see in nanostructured

materials ?

Why do we want to see these details ?

TEM of Nanomaterials

0.66 nm

1.4 nm

IGF

Grain-1

Grain-2

0.66 nm

1.4 nm

IGF

Grain-1

Grain-2

High-resolution micrograph from a Lu-Mg doped Si3

N4

sample showing the presence of an Intergranular Glassy Film (IGF).

TEM: High-resolution phase contrast image

Atomic-resolution scanning transmission electron microscope (STEM) images ofan intergranular

glassy film (IGF) in La-doped -Si3

N4

:a) High-angle annular dark-field (HAADF-STEM)b) Bright-field (BF-STEM)

N. Shibata, S.J. Pennycook, T.R. Gosnell, G.S. Painter, W.A. Shelton, P.F. Becher

“Observation of rare-earth segregation in silicon nitride ceramics at subnanometre

dimensions”

Nature 428 (2004) 730-33.

(a) (b)[0001]

STEM: High-Angle Annular Dark Field Image + Bright Field Image

HAADF-STEM images of the interface between the IGF

and the prismatic surface of an -

Si3

N4

grain. The -Si3

N4

lattice structure is superimposed onthe images. a)

La atoms are observed as the bright spots (denoted by red arrows) at theedge of the IGF. The positions of La atoms are shifted from that of Si atoms based on theextension of the -Si3

N4

lattice structure; these expected positions are shown by opengreen circles. b)

reconstructed image of a, showing the La segregation sites more

clearly. The predicted La segregation sites obtained by the first-principles calculations are shown by the open white circles.

N. Shibata, S.J. Pennycook, T.R. Gosnell, G.S. Painter, W.A. Shelton, P.F. Becher

“Observation of rare-earth segregation in silicon nitride ceramics at subnanometre

dimensions”

Nature 428 (2004) 730-33.

Window for intensityprofile

40 nm

16 nm

22 n

mA

A

A A

aFWHM ~ 3.2 nm

Window for intensityprofile

40 nm

16 nm

22 n

mA

A

A A

aFWHM ~ 3.2 nm

Diffuse Dark Field image from a grain boundary in SrTiO3

TEM: Diffuse Dark Field Image

We see something if-

Light enter our eyes

-

There is contrast in the imageWe have strong or weak contrast*

Contrast (difference in intensity between

two adjacent areas)

Amplitude contrast

Phase contrast ~ Fringes

1 2

2 2

IC I II I

When do we see anything ? –

Contrast

> 5-10 % we see

* But not bright or dark contrast (these terms refer to intensity and not to contrast)

Thickness Fringes

Bend Contours

Fresnel Fringes

Moiré

Patterns

Phase contrast

Lattice Fringes

In most of the situations both type of contrasts contribute to an image-

although one will tend to dominate

~ Fringes

Contrast

Amplitude Contrast

Diffraction contrast

Absorption contrastMass-thickness contrast small effect in a thin specimen

Bright Field & Dark Field Images

Resolution of human eyes ~ (0.1 –

0.2) mm Highest useful magnification is governed by the resolution Raleigh criterion

is used for the definition of resolution

How do we see fine details ? -

resolution

0.61sin

~2

1/ 2

1.22~E

100keV electrons λ = 4pm

Optical Microscope

TEM

→ Smallest distance that can be resolved

→ Wavelength of radiation

→ Semi-angle of collection

→ Refractive index of viewing mediumGreen light= 550 nm ~ 300 nm 1000 atomic diameters

2x

30x 10x

Magnification without detail in image does not help!

Rayleigh criterion

Not a fundamental rule but a practical definition. Figure of merit in terms of the eyes ability to distinguish separate

images of two self-luminous incoherent point sources. A single point source will not be imaged as a point even if aberrations

are absent. Any physical limit in the path of the rays (outer boundary of the lens /

aperture)

will lead to diffraction effects. Diffraction → point is imaged as a disc (Airy disc).

P2P1

Fully resolved

P1 P2

Unresolved

P2P1

0.61

Just resolved according

to Rayleigh criterion

20 %

What is meant by High-Resolution?

High Spatial Resolution

microscope’s ability to process high spatial frequencies

High Energy Resolution

spectrometer’s ability to resolve two closely positioned energy peaks

High Chemical Resolution

ability to detect small quantities of a given chemical species

In general

Additionally, one would like to differentiate two structures adopted by a material of a given composition

This is essentially related to the above-mentioned points

Phase contrast (lattice fringe) image but not ‘high’-resolution

High-resolution but not a lattice fringe image

Phase contrast image of as-cast Mg37

Zn38

Y25

alloy showing a 18 R modulated phase

Picture taken in a JEOL

2000FX microscope with W filament. Resolution ~ 2.8Å

2 nm2 nm2 nm

A

A

(b)(a)

1.2 nm

2 nm2 nm2 nm

A

A

(b)(a)

2 nm2 nm2 nm

A

A

(b)(a)

1.2 nm

Fourier Filtered Fresnel Contrast Image from Si3

N4

–

grain boundary having an IGFPicture taken in a JEOL

JEM

4000 EX, with LaB6

filament. Resolution ~ 1.8 Å

OBJECT PLANE

BACK FOCAL PLANE

IMAGE PLANE(Gaussian Image Plane)

Diffraction Pattern

f

u

v

Lens

How can we see better ? – The TEM

Absorbed Electrons Electron-Hole Pairs

Incident High-kV Beam

Direct Beam

SPECIMEN

Elastically Scattered Electrons

InelasticallyScattered Electrons

BremsstrahlungX-

rays

Auger ElectronsVisible Light

BackscatteredElectrons (BSE)

Secondary Electrons (SE)

Characteristic

X-rays

Incident electronbeam direction

Thin specimen

Diffractionpattern

Forward scatteredbeam directions

Incident electron beam With

Uniform intensity

Thin specimen

Image Scattered electrons with

varying intensity

Scattering by the specimen changes the spatial

and Angular

distribution of the electrons

Scattering

Elastic

Inelastic

Coherent

Incoherent

Elastic scattering is usually coherent if specimen is thin & crystalline In TEM elastic scattering is predominant in (1 -

10)

in the forward direction Elastically scattered electrons at large angles (> 10) can be incoherent Inelastic scattering almost always incoherentMost of scattered electrons < 5

(< about 3

they are coherent)

Particle concept

Wave concept

~ (1-10)0

Forward

~ < 10

Forward

Z has important effect

SpecimenThin

Thick

Coherent forward scattering

Incoherent back scattering

Coherentincident beam

Incoherent elastic backscattered electrons Secondary electrons

from with in the specimen

Coherent elastic scattered electrons

Incoherent elastic forward scattered electrons

Thin specimen

Direct beamIncoherent inelastic scattered electrons

Bulk specimen

Incoherent elasticbackscattered electrons

Secondary electronsfrom with in the specimen

Incoherent elasticbackscattered electrons

Coherentincident beam

Interaction cross section (σ)

T elastic inelastic 2r

elastic

ZeVr

(approximate) neglects screening effects

The chance of a electron undergoing any interaction with an atom isdetermined by an interaction cross section.

σ (in area units)Probability that a electron scattering event will occur = Actual area of atom

r → effective radius of a scattering centre Z → atomic number

→ Scattering angle (greater than this) V → potential of incoming electron e → charge in esu

(4.803 x 1010

esu)

0 TT

N tprobability of scattering by the specimen P Q t

A

Transition metals 100 keV

electrons(Mean free path ( λMFP

))

22 210elastic m

10MFP s of nm

22 26 210 10inelastic m

0

1MFP

T

AN

Mass-thickness

contrast

Electron Cloud

Nucleus

Scattering by electron cloud

Scattering by nucleus (Rutherford scattering)

Elastic Scattering

At high angles Rutherford scattered electrons become incoherent

Forward scattered

Back scattered In TEM this signal is small

Image of beam entrance surface (Z + topography) used in SEM

Z contrast imaging atomic resolution microanalysis

A

B

• A, B processes may not be truly elastic (B may produce x-rays)•

Higher the angle that an electron emerges more the chance that it would have been inelastically

scattered

0

4

8

12

0.2 0.60.4(Sinθ)/λ (Å1)

f(θ)(Å)

Al Cu

Au

Incre

asing

Z

Change in atomic scattering factor with scattering angle

Variation of Rutherford cross section with scattering angle

60 120 180

-28

-26

24

-22

(degrees) Angle, Scattering

)(m

)(Log2

10

Beam Energy = 100 KeV

0

Au

Cu

C

Decreases orders of magnitude !

Basis for HAADF

High-angle scattering is incoherent

image contrast is due to Z and not

orientation of the specimen

Inelastic Scattering (Energy Loss)

Secondary electrons

Collective interactions / Oscillations

X-ray

Radiation damage

Characteristic x-rays

Bremsstrahlung

Electron-hole pairs and cathodoluminescence

Incoming electrons

(~100 keV)

Vacuum

Energy Loss electrons

Energy

levels

KE

1LE

2LE

3LE

Nucleus

K

1L

2L

3L

Characteristic x-rays(10−16s later)

Conduction Band

Valence Band

Characteristic x-rays

Valence Band

Conduction Band

Inner shells

Fast SE( 50 -200 keV)

Auger electrons

Up to 50%Energy loss

Slow SE

~ 50 eV

100s ev-few keV

FSE

are unavoidable and undesirable One does not form images with them as they can

come from deep with in the specimen

May degrade the quality of spectroscopic data

For surface chemistry

Topography in SEM

(resolution using FEG

~ 1nm at 30 keV)

+

STEM ( high-resolution)

Secondary electrons

SemiconductorsConduction band

Valence band

Band gap

e

e

hole

Hν

cathodoluminescence

(CL)

spectroscopy

Bias Electron beam induced current ( EBIC)

Charge collection microscopy

OR

Electron-hole pairs and cathodoluminescence

100 200 300 400

10-26

10-25

10-23

10-24

10-22

10-21

)(m 2

Aluminium

Plasmons

IonizationL

K

F S E

S E

Elastic scatter

Secondary Electrons

Incident beam energy (keV) →

Cross sections for the various scattering processes

Electron properties as a function of accelerating voltage

Accelerating voltage (kV)

Nonrelativistic wavelength (nm)

Relativistic wavelength (nm)

Mass (x mo

)Velocity

(x 108 m/s)100 0.00386 0.00370 1.196 1.644

120 0.00352 0.00335 1.235 1.759

200 0.00273 0.00251 1.391 2.086

300 0.00223 0.00197 1.587 2.330

400 0.00193 0.00164 1.783 2.484

1000 0.00122 0.00087 2.957 2.823

Resolution in the TEMResolution in the TEM

Electron Source

Imaging System

Resolution

= f(Voltage), High coherency FEG

↑

voltage

Can’t increase as lens aberrations ↑

with β61.0

rth

Note : Electron sources are coherent can’t use Rayleigh criterion.

P2P1

Fully resolved

P1 P2

Unresolved

P2P1

0.61

Just resolved according

to Rayleigh criterion

Can be increased by a larger lens aperture but electron lenses are not good

Electron Source

Lenses

Imaging System

Resolution

= f(Voltage), High coherency FEG

Apertures

Control the quality of• Images• Diffraction patterns• Analytical signals

•Lens defects affects the resolution but give us increased ‘Depth of Focus’

& ‘Depth of Field’• Can suffer from 10 kinds of aberration

Spherical Aberration

Chromatic Aberration

Control • Beam current and Convergence• Cutoff paraxial rays which suffer aberration• Image contrast• Type of image (BF / DF)• Select areas from which we want diffraction patterns (SAD)

Astigmatism

“The current best electromagnetic lens is like using a coke bottle for a magnifying lens” “If the lens of our eyes was as good as the best electromagnetic lens available then we

would be legally blind”

Lenses aberrations

Spherical Aberration

Chromatic Aberration

3sph sr = C β

chr cEr = C β

E

Most important in objective lens in TEM & condenser lens in STEM

Cs

~ Focal lengh

~ (1-3 mm)

Current high tension supplies are stable to 1 in 106

(100 keV

beam 0.1 eV) For electrons passed through a sample (50-100 nm thick) E = (15-25) eV

Astigmatism astr = β f Maximum difference in focus introduced by astigmatism

Due to non-uniform magnetic field in the path of electrons Can be corrected with a stigmator

Gaussian image planeDisk diameter = 2CS

β3

Plane of least confusionDisk diameter = CS

β3

P

P’

CS = 0

β0

dmin

CS

≠

0 The spherical aberration of the lens

causes wavefronts

from a point object P

to be spherical distorted

The point is imaged as a disc with a

minimum radius in the plane of least confusion

The disc is larger in the Gaussian image plane

Lens

This figure has been exaggerated in the angular dimension

Effect of Spherical Aberration

Spherical aberration

AssumeSpecimen is thin → low chromatic aberration

Astigmatism is corrected

14

min30.91

opt scr 143

sc

1 14 4

0.77opt

s sc c

rsph

limits the resolution Further resolution f(Rayleigh

criterion, aberration error)

0.0037nm 3

smmc

15opt

mrad min 0.3nmr

HREM min 0.15nmr

100 kev electrons

Useful magnification

Take resolution on specimen ~ 2 Å

→ Magnify the resolution of eye 0.2 mm3

69

0.2 10 100.2 10usefulM

Inelastic scattering causes loss of the energy of electrons Electron-electron interaction Loss in energy + Change in momentum Thin specimen required EELS spectrometer has a very high energy resolution

[(FEG

~ 0.3 eV), (XEDS

→ resolution ~ 100eV)]

Note: beam energy can be 400 kV Can be used in forming energy filtered images + diffraction patterns

Electron Energy Loss Spectrometry (EELS)Electron Energy Loss Spectrometry (EELS)

EELS

Inter of intra band transitions

Plasmon excitation

Phonon excitation

Inner shell ionization

Part of the zero loss peak. Not resolved. Causes specimen to heat up.(~ 0.2 eV, 5-15 mrad)

(5-25 eV, 5-10 mrad)

(~5-25 eV, < 0.1 mrad)

(~30-1000 eV, 1-5 mrad)

Bulk plasmon

Surface plasmon

λp

~ 100nm

• Transverse waves•

Half the energy of bulk plasmons

Signature of the structure

• Longitudinal waves

COLLECTIVE OSCILLATIONS

PLASMONS PHONONS

Collective oscillations of free electronsMost common inelastic interaction Damped out in < 1015

s Localized to < 10 nm

Predominant in metals (high free electron density)

Collective oscillations of atoms Can be generated by other inelastic

processes. (Auger / X-ray energy)Will heat up the specimen Small energy loss < 0.1 eV

Phonon scattered electrons

to large angle (5 –

15 mrads) Diffuse background

Plasmon excitation

Longitudinal wave like oscillations of weakly bound electrons Rapidly damped (lifetime ~10-15

sec, localized to < 10 nm) Dominate in materials with free electrons (n) (Li, Na, Mg, Al )

But occur in all materials EP

= f(n) microanalytical

information Carry contrast formation, limit image resolution through chromatic aberration Can be removed by energy filtering

Inter of intra band transitions

Change in orbital state of the core electron Interactions with molecular orbital → can be used for finger printing Secondary electron emission (< 20eV)

(hence in same energy-loss regime as band transitions) Weakly bound outer-shell electrons control the reaction of a material to an

external field → controls the dielectric response (can measure with the signal

from the < ~ 10eV region)

Inner shell ionization

The other side of the coin of XEDS. Small cross section (ΔE ↑

cross section ↓) K , L (L1

, L2

, L3

) , M (M1

, M2

, M3

, M4;5

). Combination of ionization loss with plasmon loss can occur.

There are intensity variations superimposed on the ionization edge

ELNES

(starting from about 30 eV

of the edge and extending ~100s of eV)

EXELFS

(~50 eV

after ELNES) ELNES

and EXELFS

arise due to the ionization process imparting more than critical energy (Ec

) for ionization

0 40 80v

Io

Plasmon peak

250 300 350 400 450Energy-loss (eV)

ELNES

EXELFS

Zero loss peak

Forward scattered (cone of few mrad) 000 spot of DP Bragg diffracted peak (~20mrad) rarely enters the spectrometer) Includes energy loss of ~0.3 eV Includes phonon loses EELS does not resolve phonon loses

FWHM

defines the energy resolution

E resolution

kV resolution

Low loss region ~ 50 eV

Ionization Edge

Bonding effects

~50 eV

after ELNES

Diffraction effects from

the atoms surrounding

the ionized atom

The EELS spectrum

ELNES-

Energy Loss Near Edge StructureEXELFS-

EXtended

Energy Loss Fine StructureCoordination, Bonding effects,

Density of states, Radial distribution function

Thin specimen is better (plasmon peak intensity < 1/10 th

zero loss peak) Use high E0

(scattering cross-section , but benefit is in

of plural scattering + edge signal to noise ratio )

Energy resolution limited by electron source

Chemical analysis (structural and elemental) using EELS

Spatial ResolutionTEM

STEM Limited by size of probe (~1nm)

Limited by selecting aperture at spectrometer entrance (its effective size at the plane of the specimen)

Energy Resolution 1 eV

(incident energy 200-400 eV!)

• Somewhat better than XEDS• FEG

< 1nm resolution

• Using FEG

DSTEM

Krivanek

et al. [Microsc. Microanal. Microst. 2,

257 (1991).] detected single atom of Th

on C film

No visible Grain Boundary

2.761 Å

Fourier filtered image

Dislocation structures at the Grain boundary

counts

eV

10000

11000

12000

13000

14000

15000

16000

17000

18000

19000

counts

1850 1900 1950 2000 2050 2100 2150 2200 2250eV

counts

eV

11000

12000

13000

14000

15000

16000

17000

18000

19000

20000

counts

1850 1900 1950 2000 2050 2100 2150 2200 2250eV

Si peak at 1839 eV Sr

L2,3

peaks

Grain Boundary

Grain

eV1900 2000

EELS

2100 2200

~8º

TILT BOUNDARY IN THE SrTiO3

POLYCRYSTALGB23x4

MX23x4

Dist 5 nm

VG microscope

[1] Using FEG

DSTEM

Krivanek

et al. [Microsc. Microanal. Microst. 2,

257 (1991).] detected single atom of Th

on C film[2] Chapter 40 in Transmission Electron Microscopy by David B. Williams and C. Barry Carter, Plenum Press, New York, 1996.

EELS microanalysis detection of a single atom of Th

on C film

EELS microanalysis

Si L edge ELNES

changes across a Si- SiO2

interface due to change in Si bonding atomic level images with spectra localized to individual atomic columns

[1]

[2]

Energy-loss (eV) →

v

280 290 300 310 320

Graphite

Diamond

Differences between C edge between graphite and diamond

Change in Cu L edge as Cu metal is oxidized

Energy-loss (eV) →

v

920 940 960 980

CuO

Cu

L3

L2

[1] [1]

[1] Chapter 40 in Transmission Electron Microscopy by David B. Williams and C. Barry Carter, Plenum Press, New York, 1996.

ELNES

2D image of 3D specimen

Diffraction contrast

Contrast depends on lenses + apparatus

Absorption contrast is usually small

Phase contrast related to specimen and not to topography (or “depth”)

“Seeing is believing “

? ! ?

High Resolution Microscopy

With Reference to Lattice Fringe Imaging in a TEM

)( ),( rfyxf

)( ),( rgyxg

)(rh

Point spread function )(uHContrast transfer function

Specimen

Image

Diffraction patternBack focal plane

( )F u

( )G u

Rea

l spa

ceR

eciprocal space

Fourier transform

2 ) ()(

432 uCufu s

)()()( rhrfrg

)()()( uFuHuG

Point spread function

Convolution in real space gives multiplication in reciprocal space

Contrast transfer function

Aperture function

Envelope function

Aberration function

)()()()( uBuEuAuH

Property of lens

) , , ,()( sCuffu

)](exp[)( uiuB

Specimen transmission function

Phase distortion function

High resolution implies ability to see two closely spaced features in the sample (real ‘r’

space) as distinct

This corresponds to high spatial frequencies large distances from the optic axis in the diffraction pattern (reciprocal (u) space)

Rays which pass through the lens at such large distances are bent through a larger angle by the objective lens

Due to an imperfect lens (spherical aberration) these rays are not focused at the same point by the lens

Point in the sample disc in the image (spreading of a point in the image)

Objective lens magnifies the image but confuses the detail (hence the resolution is limited)

Each point in the final image has contributions from many points

in the specimen (no linear relation between the specimen and the image)

Is a linear relationship possible?

yxieyxAyxf , ),(),(

f(x,y) → specimen function

→ phase (function of V(x,y,z))

V(x,y,z) → specimen potential Vt

(x,y) → projected potential

→ Interaction constant (=/E)

Phase Object approximation

yxieyxf ,),(

Weak Phase Object approx.

),( 1),( yxVyxf t

d

V(x,y,z) Specimen

Phase change = f(Vt

)

yxVi teyxf , ),(

WPOA For a very thin specimen the amplitude of the transmitted wave function will be linearly related to the projected potential

Specimen and approximations

)],(, [),( yxyxVi teyxf

dzzyxVyxVt ),,(),(

),( ),( yxVyxVE

d tt

Including

absorption

meEh

2In vacuum

),,((2'

zyxVEmeh

In the specimen

dzdzd

22'

Phase change dzzyxVE

d ),,( dzzyxVd ),,(

dzzyxVyxVt ),,(),(),( ),( yxVyxVE

d tt

→ Interaction constant (=/E) tends to a constant value as V increases(energy of electron proportional to E & 1, variables tend to compensate)

2 3 4

11! 2! 3! 4!

x x x x xe

POA

holds good only for thin specimens

If specimen is very thin {Vt

(x,y) << 1} WPOA

In WPOA

{f(x,y) = 1

i

Vt

(x,y)} the amplitude of the transmitted wave will be linearly related to the projected potential of the specimen

)()()( rhrfrg

),()],( 1[),( yxhyxVyx t

WPOA

)],( ),([)],( 1[),( yxSiniyxCosyxVyx t

[1 2 ( , )] ( , )tI V x y Sin x y

I = * Neglecting 2

Only imaginary part of B(u) contributes

to intensity

)](exp[)( uiuB

)]([2)( uSinuB

)]([2 )()()( uSinuEuAuT

Objective lens transfer function Incoherent illumination T(u) = H(u)

T(u) ve

+ve

phase contrast (atoms would appear dark)

In WPOA

T(u) is sometimes called CTFApproximately

)]([)(2)( uSinuAuT

Or

coherencespatialchromaticeffective EEuTuT )()(

Envelope is like a virtual aperture at the back focal plane

u (nm1)

→

T(u)

= 2

Sin

()→

+2

2

v

2 4 6

Cs

= 1 mm, E0

= 200 keV, f = 58 nm)

2.5Å

)]([)(2)( uSinuAuT

u (nm1)

→

T(u)

= 2

Sin

()→

Cs

= 2.2 mm, E0

= 200 keV, f = -100 nm)

)]([)(2)( uSinuAuT

T eff

ectiv

e(u)→

u (nm1)

→

coherencespatialchromaticeffective EEuTuT )()(

Scherzer

defocus and Information Limits 124

3Scherzer sf C

1 34 4

Scherzer su =1.51 C

1 34 4

Scherzer s= 0.66 C r

Instrumental resolution limit

can use nearly intuitive arguments to interpret the contrast

Information limit

Damping envelope

IMAGE SIMULATIONIMAGE SIMULATION

Computer simulation (a) and electron micrograph (b) of a Cu Cubeoctahedron

consisting of seven shells (1415 atoms) in [001] orientation. Structures of nanoclusters

by high-resolution transmission electron microscopy by J. Urban, Volume 10, p.161 in Encyclopedia of Nanoscience

and Nanotechnology (Ed.: Hari

Singh Nalva), American Scientific Publishers, Stevenson Ranch, 2004.

(a) (b)

Thickness dependency of lattice images of a Si crystal (200 kV, f = 65 nm). Thickness changes from 1nm (a) to 86 nm (r) in steps of 5 nm.

High-Resolution Electron Microscopy for Materials Science by Daisuke Shindo

and Kenji Hiraga, Springer-Verlag, Tokyo, 1998.

Defocus dependency of lattice images of a Si crystal (200 kV, t = 6 nm). Thickness changes from 20 nm (a) to 90 nm (l) in steps of 10 nm.

High-Resolution Electron Microscopy for Materials Science by Daisuke Shindo

and Kenji Hiraga, Springer-Verlag, Tokyo, 1998.

In lattice fringe imaging mode:

Things which look like ‘atoms’

are not atoms!

Use image simulation to confirm what you see

Do not interpret details below the resolution limit of the microscope

HRTEM-

some tips

Spherical aberration correctors for illumination as well as imaging lens systems,

Electron beam monochromators

reducing the energy spread of the electron beam

to < 0.1 eV

High-energy resolution spectrometers (Sub-eV

mandoline

filter for EELS)

Imaging energy filters

Advanced specimen holders for in-situ experiments

High-sensitivity and high-dynamic-range detectors for images, diffraction patterns,

electron energy loss spectra (EELS) and energy dispersive X-ray spectra (EDXS)

Advanced methods Focal series reconstruction

Electron Holography

In-focus ‘Fresnel’

reconstruction

It is possible to record phase contrast images with point resolutions of well below

0.1 nm, or alternatively, scan electron probes of size < 0.1 nm

across the

specimen, recording high-angle scattering or high-energy resolution spectra from

areas as small as a single atomic column.

ADVANCESADVANCES

ELECTRON HOLOGRAPHYELECTRON HOLOGRAPHY

High-resolution (a) amplitude and (b) phase images of the aberration-corrected object wave reconstructed from an electron hologram of [110] Si, obtained at

300 kV on a CM 30 FEG

TEM. The characteristic Si dumbbell structure is visible only after aberration correction. A. Orchowski

et al., Phys. Rev. Lett. 74, 399, 1995.

(a) (b)

-40 0 40-20 20Distance (Å)

RTHTCooled to RT

20 nm20 nm20 nm

A

A2.3 nm

20 nm20 nm20 nm

B

B

2.5 nm

20 nm20 nm20 nm

C

C 3.2 nm

RT

HT

Cooled to RT

FRESNEL RECONSTRUCTIONFRESNEL RECONSTRUCTION

Exit face wave reconstructed profiles, of the imaginary part of the inner potential, using

zero-defocus Fresnel fringe images

NANOSTRUCTURESNANOSTRUCTURES

IMAGE GALLERYIMAGE GALLERY

HRTEM

micrographs of MWNT: (a) 5-walled, diameter = 6.7 nm; (b) 2-walled, diamter

= 5.5 nm; (c) seven-walled, diameter = 6.5 nm (hollow

diameter = 2.2 nm).

S. Iijima, Nature, 354, p.56, 1991.

HRTEM

image of a BN

multi-walled nanotube

(inner shell distance of ~ 0.33 nm).D. Golberg

et al., J. of Appl. Phys. 86, 2364, 1999.

NANOTUBESNANOTUBES

(a) (b) (c)

Cross section of TiN/NbN

nanolayered

coatings: (a) Conventional TEM micrograph with SAD patterns, (b) HRTEM

of {200} lattice fringes.M. Shinn et al., J. Mater. Res. 7, 901, 1992.

(a) (b)

NANOLAYERSNANOLAYERS

High-resolution micrograph of Si3

N4

–

TiN

composite prepared by CVD.High-Resolution Electron Microscopy for Materials Science by Daisuke Shindo

and Kenji Hiraga, Springer-Verlag, Tokyo, 1998.

Gas atomized Al-Si powders:

(a) HREM, (b) SEM, (c) BFI.High-Resolution Electron Microscopy for Materials Science by Daisuke Shindo

and Kenji Hiraga, Springer-Verlag, Tokyo, 1998.

[001][110]

(a)

(b)

(c)

[110]Al

NANOPARTICLESNANOPARTICLES

NANOTWINSNANOTWINS

[011]

Twin Plane

HREM

of ZrO2

showing formation of nanotwins.High-Resolution Electron Microscopy for Materials Science by Daisuke Shindo

and Kenji Hiraga, Springer-Verlag, Tokyo, 1998.

NANO(poly)CRYSTALLINENANO(poly)CRYSTALLINE

MATERIALMATERIAL

HREM

image of nanocrystalline

PdG.J. Thomas et al., Scripta

Metall. Mater. 24, 201, 1990.

MOIRMOIRÉÉ

FRINGESFRINGES

HRTEM

image of a Kr nanocluster

on a Mg substrate showing Moiré

fringes. The lattice parameter of Kr can be calculated from the Moiré

fringe spacing: [1/dfringes

= |(1/dMgO

)

(1/dKr

)|] Encyclopedia of Nanoscience

and Nanotechnology (Ed.: Hari

Singh Nalva), American Scientific Publishers, Stevenson Ranch, 2004.

Incident convergent beam

Specimen

BF detector

HAADF detector

θ3

θ2

θ1 > 50 mrad off axis

θ3 < 10 mrad

θ2 >10-50 mrad

θ1

ADF detector

High-Angle Annular Dark Field Imaging

HAADF-STEM images of the interface between the IGF

and the prismatic surface of an -Si3

N4

grain.N. Shibata, S.J. Pennycook, T.R. Gosnell, G.S. Painter, W.A. Shelton, P.F. Becher

“Observation of rare-earth segregation in silicon nitride ceramics at subnanometre

dimensions”, Nature 428 (2004) 730-33.

Intergranular Glassy FilmIntergranular Glassy Film