• 沒有找到結果。

From STEM‐EELS to multi‐dimensional and  multi‐signal electron microscopy

N/A
N/A
Protected

Academic year: 2022

Share "From STEM‐EELS to multi‐dimensional and  multi‐signal electron microscopy"

Copied!
59
0
0

加載中.... (立即查看全文)

全文

(1)

Taipei, 

November 22, 2011

F STEM EELS t lti di i l d

From STEM‐EELS to multi‐dimensional and  multi‐signal electron microscopy

Christian Colliex

L b i d h i d S lid ld 10

Laboratoire de Physique des Solides, Bldg 510 Université Paris Sud, 91405, Orsay, France

christian colliex@u‐psud fr christian.colliex@u‐psud.fr

(2)

Outline Outline

Signals, instrumentation and methods for STEM EELS Atomically resolved elemental and bonding maps

Mapping plasmons and EM fields When electrons and photons team up

(3)

Electron Matter interactions Electron – Matter interactions

Secondary Secondary 

Event  vs.

Primary Event

Transmission Electron Microscopy ‐ A Textbook for Material Science David Williams and Barry Carter, Fig. 1.3, page 7. 

(4)

Working modes for an transmission electron microscope

EDX spectroscopy

Electron source (0.5 à 2eV)

d ( )

Diffractions

object (gonio) condensers anode(s) 

a few 100kV

HREM  imaging object (gonio)

Objective  lens

Nanolaboratory : Specific specimen holders and stages

Intermediate lenses Projector lens

Hologram

holders and stages

Electron Energy Loss Spectroscopy

(EELS) Hologram

Energy filtered imaging Magnetic

prism Energy filtered imaging

p

(5)

1980 20

STEMs with EELS analysers at Orsay

1980 ‐ 20xx

2008 ‐ 2011

y

2011 ‐ 20xx

VG HB 501

NION UltraSTEM 100

NION UltraSTEM 200

(6)

20 nm

EELS spectrum

EELS spectrum‐image  at Orsay

A B

y

HADF image

450 400

350 300

Energy Loss (eV)

Magnetic spectrometer Spectrum 0.5 to 0.8 eV

1 ms to 5 s SPECTRE LIGNE

SPECTRUM LINE

Magnetic spectrometer E

E -E o

o Camera

CCD

Specimen

HADF detectors

A

Specimen

Probe

0-

(nm)

• 0.1 to 1nA

• in 0.5 to 1 nm

Scanning coils

I I I I

2 0 300 3 0 400

40- (nm)

Field emission gun 100 keV 250 300 350 400

Energy Loss (eV)

B

(7)

Multi‐dimensional microscopy in a composite space (x,y position, E energy and t time)

x

(x,y position, E energy and t time)

x

E y E

1D EELS t

x

0D : bit of information 1D : EELS spectrum

x

y 3D : spectrum image

data cube

y E (or x)

data cube

2D t li y

E 2D : spectrum line

or E‐ filtered image

(8)

Use of C correctors to reduce Use of Cs correctors to reduce

probe size or to increase probe current : i) <1 Å probe size at 100 kV, 

Å

<0.7 Å at 200 kV

ii) 200 pA of current in a 1.4 Å  probe

iii) 1 nA current in a 2 3 Å probe iii) 1 nA current in a 2.3 Å probe

Orsay Nion U‐STEM  100 acceptance tests

(9)

New UltraSTEM for aberration‐corrected nanoanalysis (delivered in Orsay in 2008)

(delivered in Orsay in 008)

The column is built from modules that all have the

a b modules that all have the

same mechanical

interface and are 100%

interchangeable.

interchangeable.

Each module has triple magnetic plus acoustic g p shielding.

Emphasis is on small probe formation and efficient coupling into detectors.

Everything including

sample exchange can be

t d t l

operated remotely.

(10)

Nion UltraSTEM 200  performance at

Orsay

(11)

Imaging molecules containing heavy atoms

(a) (b) (c)

0.5 nm 0.5 nm

1 nm 1 nm

(d) (d)

2 nm2 nm 2 nm2 nm

polyoxometalate (POM; As2W20O70Co(H2O)) molecules grafted on C‐SWNT courtesy A. Gloter, Orsay (2011)

(12)

BN monolayer with impurities imaged by MAADF

Result of DFT calculation overlaid on Matt Chisholm’s experimental MAADF image C ring is

N

C ring is  deformed

Cx6

O

N

B C O

O

Longer

B C O

Longer  bonds

Na  C

adatom

C

(13)

Si substituting for C in monolayer graphene

Si 2 Å

Si

Si

Si N Si

Si Si

Si

M di l l d k fi ld (MAADF) i

Si at and near  topological defects Si in topologically 

correct graphene

Si at graphene’s edge

Medium angle annular dark field (MAADF) images.

Nion UltraSTEM100 at ORNL, 60 kV.  Image courtesy Matt Chisholm, ORNL,  sample courtesy Venna Krisnan and Gerd Duscher, U. of Tennessee.

(14)

Si substituting for C: 2 structures are possible

Si 2 Å

Si

Si Si

N Si

Si

Si Si

Si in defect‐free graphene strains (and  Si in defective, but less strained graphene is  buckles) the foil. 

(courtesy Matt  Chisholm)

more stable. (15 images added together, no  other processing, 

courtesy Juan‐Carlos Idrobo)

(15)

EELS spectroscopy : spectral domains

Phonons Plasmons Absorption edges IR

visible

UV X

2.5

) x 10 6 )

Low losses C l

1.5 2.0

ts number) Low losses Core lossesCK

1.0

sity (count

0.5

Energy loss (eV)

Intens

x50 x106

MnL2,3

0 100 200 300 400 500

0 600 700

(16)

EELS: Involved electron populations  and associated transitions

Energy (eV)

CB. O ‐K

0^3 40

50 60

0^3

EF

VB.

x 10

10 20 30 40

x 10

120 Plasmons

IT

O 1

L2,3 eV

10

520 530 540 550 560 570 580 590 600 610

K eV

60 80

100 IT

TM 2p O 1s

0^3 200

250 300 350

0^3

TM L2,3 TM L2,3

20 40

0 5 10 15 20 25

-530 eV

-710 eV

2 x 10

0 50 100

x 10 150

eV

eV -725 eV

eV

0 695 700 705 710 715 720 725 730 735

eV

EELS gives informations on the electronic structure

(17)

EELS spectroscopy : spectral domains

Core energy‐loss domain Low energy‐loss domain

CK

CK MnL

2,3

Core energy loss domain Low energy loss domain

Plasmon modes

250 300 350 400 250 300 350 400

Energy loss (eV)

690

630 650 670

Energy loss (eV)

i h hi h h

0 10 20 30 40 Energy loss (eV)

Map with high accuracy the nature, the position and bonding  of the atoms responsible for the

structural properties Map different physical 

parameters, electronic,  optical or magnetic, 

which are especially important structural properties 

of real materials

(defects, interfaces, nanomaterials)

R i i i h

which are especially important  for electronic industries Requires instruments adapted 

Requires instruments with best spatial and energy resolutions

(0.1 nm, 0.1 eV) to measure the properties of interest

at the relevant scale

Towards the nanolaboratoryIn all cases, develop the theory for interpreting spectroscopical data i e

Towards the nanolaboratoryspectroscopical data, i.e. 

a physics of excited states

(18)

Absorption

Absorption edgesedges domaindomain ::

Absorption

Absorption edgesedges domaindomain : : three

three types of informationtypes of information

Identification of elements

CK

Elementary quantification

250 300 350 400

Energy loss (eV)

Study of the unoccupied electron states distribution

(19)

Quantitative elemental analysis

Q y

Characteristic signal : proportional to the number of atoms per unit  area for the element detected in

BK S

area for  the element detected in  the analysed area

CK

S = ct. I N 

NK

sity

Atomic concentration ratios:

NA SA B

Intens

NB

=

SB

A

200 300 400

Energy loss (eV)

(20)

EELS core‐level spectroscopy:

EELS core level spectroscopy:

elemental and bonding maps with atomic resolution

resolution

1. Individual atoms 1. Individual atoms

2. Crystalline structures and interfaces 3. Application to Tunnel ElectroMagneto 3. Application to Tunnel ElectroMagneto

Resistance – TEMR

(21)

Single atom identification (signal/noise criteria)

Peapods :

Gd@C82@SWCNT

Element selective single‐atom  imaging

A HREM i

A : HREM image

B : Schematic presentation

C : Superposed maps of the Gd N45  and C K signals extracted from a  32x128 pixels spectrum image (C in 32x128 pixels spectrum‐image (C in  blue, Gd in red)

K. Suenaga et al., Science (2000)

(22)

STEM imaging of peapods at 30 and 60 kV with Delta STEM imaging of peapods at 30 and 60 kV with Delta 

corrector

30kV

60kV

Damage drastically reduced at 30kV

Courtesy Suenaga, Sawada & Sasaki (2010)

(23)

Single atom imaging by STEM‐EELS at low voltage with the  g g g y g delta corrector

Endohedral fullerenes  M@C82 (M= La, Ce, Er) Iizumi and Okazaki

(24)

Atom by atom labeling at 60kV y g

Courtesy K. Suenaga (AIST, Tsukuba, 2010)

(25)

Valence state identification of  individual atoms

La3+ Ce3+

La3+ in LaCl3 Ce3+ in CeCl3 Ce4+ in CeO2

Courtesy K. Suenaga (AIST, Tsukuba, 2010)

(26)

EELS spectrum‐imaging across interfaces

S C C lli N t N&V (2007) See C. Colliex, Nature N&V (2007)

HAADF micrograph

(27)

Elemental maps recorded with NION UltraSTEM at Orsay

( t L B h 2011)

(courtesy Laura Bocher, 2011)

(28)

Spectroscopic imaging of LMO down the pseudocubic <110> axis. The sketch shows the  projected structure of LMO down this direction. In green, the O K edge image; in blue the  simultaneously acquired Mn L2,3 image and in red the La M4,5 image. The RGB overlay  of the three elemental maps is also shown.

From M. Varela et al. to be published in MRS bulletin 01/2012

(29)

D. Muller et al. Science 319 (2008) 1073)

(30)

High resolution Z contrast image of a LCMO/YBCO/LCMO heterostructure The inset High resolution Z‐contrast image of a LCMO/YBCO/LCMO heterostructure. The inset  marks the region where an EELS spectrum image was acquired, along with the 

simultaneous ADF signal. (b) O K, Mn L2,3, Ba M4,5 and La M4,5 atomic resolution  images (c) RGB overlay of the Mn (red) La (green) and Ba (blue) images in (b) The images. (c) RGB overlay of the Mn (red), La (green) and Ba (blue) images in (b). The  sketch shows the interface structure. 

From M. Varela et al. to be published in MRS bulletin 01/2012

(31)

Ferroelectric control of spin polarization (Tunnel ElectroMagneto Resistance – TEMR) (Tunnel ElectroMagneto Resistance – TEMR)

 tunnel junctions with ferromagnetic electrodes for large nonvolatile control of carrier  spin polarization by switching ferroelectric polarization

ultrathin BTO ferroelectric

half‐metallic LSMO as spin detector

Fe electrode

p

NGO: substrate

Information provided from STEM/EELS analyses

Garcia V. et al. Science DOI: 

10.1126/science.1184028

* Structural quality of the film growth and the interfacial area

* Termination planes at the interfaces

* Oxidation states of the TM at the atomic scale

d d d h l l h / f

In order to understand the electromagnetic coupling at the FE/FM interface

Garcia V. et al. Science (2010)

(32)

Ferroelectric control of spin polarization

V Garcia et al (Thales/CNRS Palaiseau LPS Orsay U Cambridge) V. Garcia et al (Thales/CNRS Palaiseau, LPS Orsay, U. Cambridge)

BTO

Fe LSMO NGO

10 nm

Fe BTO

[001]

[110]

1 nm

[001]

USTEM data courtesy A. Gloter & L. Bocher

TMR (H) measured for reversed bias polarities on the ferroelectric junction

y

L. Bocher et al. submitted (2011)

(33)

BaTiO3/Fe interface

i) one possible structure model model ii) image simulation

iii) and iv) HAADF experimental images) ) p g iv) elemental profiles 

Courtesy L. Bocher & A. Gloter JOM 62 (12/2010) 53‐57

(34)

STEM imaging the interface BTO‐Fe

HAADF BF

(35)

Atomic structure at the interface : 

comparison experiment, models and simulations

L. Bocher et al. submitted (2011)

(36)

Elemental composition and Fe EELS L23 fine  structures across the interface

Presence of oxidized Fe   (in the Fe++ as well as in the Fe+++

state) over one atomic layer at the interface

L. Bocher et al. submitted (2011)

(37)

Modelling the interface and electronic structure calculations

DFT calculations

of the spin polarisation L. Bocher et al. submitted (2011)

(38)

Orsay STEM

Trends of the accessible 

1

0.2 2

performance in terms of spatial  90s

and spectral resolution (updated in 2010)

1 1

EFTEM 70s Cs correctors

(updated in 2010)

0.3 0.3

V)

IBM STEM

Must be accompanied with a parallel 90s

development in data processing and  modelization tools (propagation of a

(eV

modelization tools (propagation of a  sub‐angström electron probe across a  thin specimen, physics of the inelastic

scattering, calculation of electron

0.1 0.1

Monochromators 80 00 U‐STEM

2010

g,

density of states…)

0.2 1 2

80s‐00s

0.1 2010

Where are we now ?

∆x (nm)

(39)

Mapping plasmons and EM fields

Mapping plasmons and EM fields

(40)

4D (

4D (x,y,E,tx,y,E,t) Spectrum) Spectrum‐‐imagingimaging modemode

e e e e e e e e e e e e e e

Improved energy resolution

(0 2 eV) (0.2 eV)

Sample

0 / i l

at each pixel

• 50 spectra/pixel

• 3 ms/spectrum

d l ti

+ HAADF signal

• deconvolution

(41)

New possibilities for studying the low energy‐loss domain

Deconvolution techniques open the way to investigating nanophotonics  with electrons

2.25 eV I.R. U.V.

1.8 eV

1 15 eV

Increase in ΔE 

(f 0 35 V t 0 25 V) 1.15 eV

0 2 4 6

Energy loss (eV)

(from 0.35 eV to 0.25 eV) Cut‐off of zero loss signal

at 0.9 eV (IR)

New technical possibilities (A. Gloter, A Douiri, M. Tencé)

Higher

signal‐to‐background ratio

(42)

Mapping surface plasmon resonances of triangular silver nanoprisms

B of triangular silver nanoprisms

78 nm edge long 78 nm edge long

nanoprism nanoprism

A C B

D

Energy map of the “tip” mode

J. Nelayah et al. Nature Physics, 3, 348‐353 (2007)

(43)

EELS simulations of triangular Ag nanoprisms

(courtesy J. Garcia de Abajo, Madrid)

1.9 eV 2.9 eV

0.8

3.4 eV 1.0

0.2 0.4 0.6

0.0

100 Kv electrons

78 nm long and 10 nm  thick Ag prismg p

(44)

Modes in an Ag nanoantenna (aspect ratio L/r increases)

(coll Cambridge Stuttgart courtesy P Midgley)

1.2eV 1.4eV 1.6eV

(coll. Cambridge‐Stuttgart, courtesy P. Midgley)

1.2eV 1.4eV 1.6eV

2 3

1 8 V 2 0 V 2 2 V 2 4 V

1.8eV 2.0eV 2.2eV 2.4eV

4 5 6

2 6 V

EFTEM i 2.6eV 2.8eV 3.0eV 3.2eV

EFTEM series on  SESAM machine 660 nm Ag nanorod

7

0.2eV slit width

(45)

Silver nanoantennas EELS (2)

Nano Lett 11, 1499 (2011)

Experiments versus simulations 

(DDA of |Ez|2 at 60 nm above antenna)

(46)

When electrons and photons  team up

« Multi‐signal microscopy »

(47)

« When electrons and photons and photons 

team up »

by F.J. Garcia de Abajo (Nature 462 (2009) 861)

(48)

Light detection (inserting a parabolic mirror within the  VG pole piece)

VG pole piece) 

home-made cold stage + light

d t t (L Z l S

detector (L. Zagonel+ S.

Mazzucco + M. Kociak) Monochromatic electron beam

Electron Induced Radiation Emission (Cathodoluminescence)

Electron Energy-Loss

Absorption by EELS

gy Spectrum

Emission by CLy Absorption by EELS

2 patents (licensing opportunities)

(49)
(50)

2D photon  emission spectral‐

emission spectral‐

imaging

Spatial sampling: 0.7 nm

S t l li 2 ( 8 V)

VG HB501

L. Zagonel, M. Kociak et al., NanoLetters 2011

Spectral sampling: 2 nm (ca 8 meV)

Individual QD optical properties revealed!

(51)

Spectral Imaging with electrons

EELS CCD  camera EELS 

spectrometer

EELS scintillator

HAADF

EELS aperture EELS scintillator

c

Sample

Secondary  Electrons

CL Spectrometer

Obj ti l CL 

Mirror

CL Spectrometer

CL PM

Electron gun tip Scan coils

Objective lens

M. Kociak et al., patents pending L. Zagonel, A. Losquin et al.,  unpublished,

Absorption and emission on the same object 

@ nm resolutions..

p ,

(52)

Absorption and Emission

multi-detection: HADF+

EELS + EIRE (CL) first proof of principle for simultaneous EELS/EIRE

symmetry of the modes, modal decomposition?

CL

Energy 2.2 eV

EELS

(radiative and non 

CL 

(only radiative modes)

radiative modes) No energy resolution limitation by the PSF detection system:

optical spectroscopy at a true nanometer scale

(53)

Nature, 17 december 2009

(54)

and now?

With synchronised light injection (cf Zewail’s group at UCLA)

New spectroscopies synchronizing electrons and photons (injecting light)

•(i) electron energy‐GAIN spectroscopy

•(ii) dynamics of excited states

•(ii) dynamics of excited states 

(55)

The most The most

recent

textbook on  the market…

(56)

MRS Bulletin, january 2012 on

“Spectroscopic Imaging in Electron Microscopy”

Ed S P k & C C lli Eds. S. Pennycook & C. Colliex

Invited contributions : Invited contributions : G. Botton, McMaster, Canada

M Varela et al ORNL USA and Madrid Spain M. Varela et al. ORNL, USA and Madrid, Spain

M. Kociak, Orsay, France & J. Garcia de Abajo, Madrid, Spain K. Suenaga et al. AIST, Japan

L J All t l M lb A t li L.J. Allen et al.  Melbourne, Australia M. Aronova & R.D. Leapman, NIH, USA

(57)

The Orsay team enabling the future (june 2011)

http://www.lps.u‐psud.fr/stemlps

(58)

Thanks to all my colleagues at LPS Orsay Thanks to all my colleagues at LPS Orsay

Guillaume Boudarham, Laura Bocher N th li B Al d Gl t

Nathalie Brun, Alexandre Gloter, Mathieu Kociak, Katia March, Stefano Mazzucco, Claudie Mory,

Odile Stéphan, Marcel Tencé, Almudena Torres-Pardo, Mike Walls,

Luis Zagonel and Alberto Zobelli from LPS Orsay, France

Thanks to CNRS CNRS and to EC funded programs EU Thanks to CNRS CNRS and to EC funded programs EU

SPANS et ESTEEM and to all our partners who have submitted scientific issues to solve and suited specimens

Thank you very much for your attention

(59)

參考文獻

相關文件

A trait implementation class which contains the definitions for the provided methods of the trait, proxy fields for the user of the trait and all used traits, as well as

the prediction of protein secondary structure, multi-class protein fold recognition, and the prediction of human signal peptide cleavage sites.. By using similar data, we

We do it by reducing the first order system to a vectorial Schr¨ odinger type equation containing conductivity coefficient in matrix potential coefficient as in [3], [13] and use

Receiver operating characteristic (ROC) curves are a popular measure to assess performance of binary classification procedure and have extended to ROC surfaces for ternary or

 Promote project learning, mathematical modeling, and problem-based learning to strengthen the ability to integrate and apply knowledge and skills, and make. calculated

Wang, Solving pseudomonotone variational inequalities and pseudocon- vex optimization problems using the projection neural network, IEEE Transactions on Neural Networks 17

Define instead the imaginary.. potential, magnetic field, lattice…) Dirac-BdG Hamiltonian:. with small, and matrix

Microphone and 600 ohm line conduits shall be mechanically and electrically connected to receptacle boxes and electrically grounded to the audio system ground point.. Lines in