a novel spm system for determining quantum electronic
TRANSCRIPT
A Novel SPM System for Determining Quantum
Electronic Structure at the Nanometer-scale
Joseph A. Stroscio, NIST Fellow
Electron Physics Group
2011 Nanoelectronics Metrology, May 25, 2011
Presentation Outline
2
B
Microscopy Honeycomb Lattices Magnetic Fields
Graphene“Quartet” Graphene Devices
Some History of Microscopy
Occhiolino “Little Eyes” – 16th Century
First microscope was the optical microscope
Compound microscopes end of 16 century
Galileo Galilei's compound microscope in 1625
Occhiolino “Little Eyes”
3
18th century microscopes
Musée des Arts et Métiers, Paris http:/www.eatechnology.com
Wikipedia
Some History of Microscopy: Scanning Tunneling
Microscope a “Quantum” Microscope
Invented by Gerd Binnig and Heinrich Rohrer in 1981
Nobel Prize in Physics in 1986 with Ersnt Ruska (electron microscope)
4
Quantum Mechanical Tunneling
Distance
Current
2 dI e
from Wikipedia
Scanning Tunneling Microscopy
A “Quantum” Microscope
STM is an electron probe, sensitive to the energy resolved
local density of electron states (LDOS) – seeing in “color”
5
2
( , ) ( , )
( , ) ( ) ( )
( , )
F
F
E V
t
E
t t
t
I r E T E V dE
r E r E E
dIr E
dV
J. A. Stroscio, R. M. Feenstra, and A. P. Fein, PRL
57, 2579 (1986)
R. M. Feenstra, J. A. Stroscio, J. Tersoff, and A. P.
Fein, PRL 58, 1192 (1987)
GaAs(110)
,B ,Vg
Evolution of Cryogenic Scanning Tunneling
Microscopes
6
Exponential tunneling transmission selects out the last
atom on the probe tip
Allows to “see”, “feel”, and “hear” in the nanometer
scale world
Evolution of Cryogenic Scanning Tunneling
Microscopes
Desire stability and higher energy resolution
Resolution limited by the thermal Fermi-Dirac
distribution ~ 3kBT
Solution: Go to lower temperatures
Not so easy!
7
1990
1981
2004
2010
T = 4 K T = 295 K
T = 0.6 K
T = 10 mK
Competing Requirements to Achieve High
Resolution at Low Temperatures
Tunneling current changes by x10 with 1 Å change
< 1 picometer displacement fluctuation is required
Isolate from the environment to achieve small
fluctuations
Poor thermal transport
Bond strongly to environment to achieve good thermal
contact
Poor isolation
Solution is to do both!
8
Developing High-Energy Resolution SPM
Measurements
9
Processing Lab
ULTSPM Lab at NIST
Stage 1
Stage 2
Stage 3
Vibration Isolation
Developing High-Energy Resolution SPM
Measurements
Refrigeration to 10 mK using 3He-4He mixture
10
Magnet
IVC
UHV
≈700 mK
STILL
Shield ≈50 mK
ICP
Shield
STM
Silver rods
Mixing
Chamber
≈10 mK3He dilute
3He rich
STILL
3He Pumping
Station
3He
Pump
4LHe
1K pot
3He
Z3 Z1
Z2
E3 E1
E2
E4
E5
Comp
3He-4He Gas Handling System
(GHS)
1.5 K
4 K
700 m K
50 mK
10 m K
Vladimir Shvarts
Zuyu Zhao Y. J. Song et al. RSI (2010)
Developing High-Energy Resolution SPM
Measurements
11
Young Jae Song
Alexander F. Otte
Young Kuk
Joseph A. Stroscio
Developing High-Energy Resolution SPM
Measurements
Excellent performance down to lowest temperatures
JT is better than 1K pot
Z noise < 1 pm Hz1/2
I noise < 100 fA Hz1/2
12
Y. J. Song et al. RSI (2010)
Er atoms on CuN
8 nm
Graphene/SiC
5 nm
T=13 mK
Presentation Outline
13
B
Microscopy Honeycomb Lattices Magnetic Fields
Graphene“Quartet” Graphene Devices
From Honeycombs to the Dirac Hamiltonian
Graphene – Light-like Electrons
14
Savage, N., "Researchers pencil in graphene
transistors." IEEE Spec. 45, 13 (2008).
From Pencil Drawings to High Speed Transistors to iPAD?
Or Galaxy Tab?
IBM and HRL
GHz Transistors
Nature Nanotech. (2010) SKKU, Korea
New Materials and State Variables
Graphene, TIs; Spin and Pseudo-Spin as State Variables
Electron spin
Graphene sub-lattice pseudo-spin
Graphene bilayer – layer pseudo-spin
Topological Insulator – spin locked to momentum
15
Graphene Dirac Fermions
Graphene Basics
16
Top View (real space)
Carbon with 4 valence electrons
Two atom basis in the unit cell
→ pseudo-spin
From Honeycombs to the Dirac Hamiltonian
17
Figure from droid-life.com
kx
ky
Ener
gy
Figure from P. Kim
Wallace (1947)
kx' ky'
K K’
Energy is linear with
momentum massless particles
From Honeycombs to the Dirac Hamiltonian
Low Energy Expansion: Dirac Hamiltonian
18
y
x
Real space: For behavior away from Dirac point, make an expansion:
Paul Dirac
4
0, ,3
y xK K K k ka
kx' ky'
E
/2
, /2
3
2
0
0
1 ( )
2
arctan
F nn
x y
F F
x y
i
F i
x
y
av
k ikH v v
k ik
eE v k
e
k
k
k
k
K
K
k
σ k
k
Reciprocal space:
The Independent Two Valleys
19
kx' ky'
E
kx' ky'
E
K
K’
0
0
x y
F
x y
k ikH v
k ik
K
0
0
x y
F
x y
k ikH v
k ik
K
Leads to additional degeneracy
Consequences of Dirac Hamiltonian
Pseudo-spin; reduced
backscattering
20
kx' ky'
E
Klein tunneling;
transmission through
potential barriers
Katsnelson et al. Nature Physics 2006
Presentation Outline
21
B
Microscopy Honeycomb Lattices Magnetic Fields
Graphene“Quartet” Graphene Devices
Geim & Novoselov Nature 2007
Landau Quantization in Graphene
Cyclotron motion in a magnetic field
Quantized orbits and energy levels
22
“Standard” Landau level spacing Graphene Landau level spacing
10 K@10 T
1000 K@10 T
Relativistic:
2sgn( ) 2nE n e c n B
Standard Model:
*( 1/ 2)n
eE E B n
m
Lev Landau
1908 - 1968
B
Scattering in the graphene
landscape
Effects of disorder and
interactions
Landau Quantization in Graphene
The Graphene Quartet
23
Four-fold degenerate
due to spin and valley
symmetries
STS provides direct
measure of energy gaps
and interaction effects
STS vs Transport Measurements
STS
24
http://en.wikipedia.org/wiki/Quantum_Hall
Wide energy spectrum
Localized states in the
mobility gaps
Spatial properties of
extended and localized
states
Energy gaps when
degeneracies are lifted
Correlation effects
Transport
Presentation Outline
25
B
Microscopy Honeycomb Lattices Magnetic Fields
Graphene“Quartet” Graphene Devices
Epitaxial Graphene on C-face SiC – Weak Disorder
26
C-Face termination
Si-Face termination
SiC SiC
4 - 100 ML
1 - 5 ML
(0001)
(0001)
Graphene layers
n~1012/cm2
n~1010/cm2
E
E
SiC
Induction Furnace Method
J. Hass et al., PRL 100, 1255504 (2008) Berger et al., J. Phys. Chem B 108, 19912 (2004) Berger et al., Science 312, 1191 (2006) de Heer et al., Sol. St. Commun., 143, 92 (2007)
Magnetic Quantization C-face Graphene at 4K
Direct measurement of graphene quantization
Weak disorder
27
Quantization obeys
graphene scaling
Full quantization of DOS
into Landau levels
Very sharp LLs
High mobility
-300 -200 -100 0 100 200 300
0
1
2
3
4
5
dI/
dV
(nS
)
Sample Bias (meV)
B= 5 T
-7-6-5
-4
-3
-2 -1
n=0
12
3 4 56
7
D. L. Miller, et al., Science 324, 924 (2009).
Resolving the Graphene Quartet
Tunneling Spectroscopy at ~10 mK
28
0 2 4 60
100
200
300
-200 -100 0 100 200
0
10
20
30
Linear Fit
B=2 T
B=3 T
B=4 T
EN-E
0 (
meV
)
(NB)1/2
N=1
B=0 T
B=2 T
B=3 T
B=4 T
dI/d
V (
nS
)
Sample Bias (mV)
N=0
Zero field Dirac point is at -125 meV
indicating a doping of ~1 x 1012 cm-2
Graphene on C-face SiC
Y. J. Song et al. Nature (2010)
Resolving the Graphene Quartet
Tunneling Spectroscopy at ~10 mK
29
Y. J. Song et al. Nature (2010)
-25 -20 -15 -10 -5 0 5-1
0
1
2
3
4
5
dI/d
V (
nS)
Sample Bias (mV)
B = 11.5 T
-200 -100 0 100 200
0
1
2
3
dI/d
V (
nS
)
Sample Bias (mV)
B= 2 TN=1
Weak disorder in graphene
on C-face SiC allows fine
features to be observed
ΔVrms = 1 mV
ΔVrms = 50 μV
Resolving the Graphene Quartet
Tunneling Spectroscopy at ~10 mK
30
Y. J. Song et al. Nature (2010)
-25 -20 -15 -10 -5 0 5-1
0
1
2
3
4
5
dI/d
V (
nS)
Sample Bias (mV)
B = 11.5 T
Weak disorder in graphene
on C-face SiC allows fine
features to be observed
ΔVrms = 50 μV
Smaller peak separation
– electron spin?
0 2 4 6 8 10 12 140.0
0.5
1.0
1.5
2.0
E
S(m
eV
)
Magnetic Field B (T)
gS = 2.26 0.05
Resolving the Graphene Quartet
Tunneling Spectroscopy at ~10 mK
31
Y. J. Song et al. Nature (2010)
-25 -20 -15 -10 -5 0 5-1
0
1
2
3
4
5
dI/d
V (
nS
)
Sample Bias (mV)
B = 11.5 T
Weak disorder in graphene
on C-face SiC allows fine
features to be observed
ΔVrms = 50 μV
Valley splitting is ten
times larger than
smaller energy splitting
~1 meV/T
Many Body Effects in Graphene
Polarizing Landau Levels
32
-15 -12 -9 -6 -3 0 3 5
0
3
6
9
0
3
6
90
5
10
0
5
10
0
5
10
15
0
5
10
15
B=11.75 T
B=11 T
B=11.5 T
B=11.375 T
B=11.25 T
ESL
ESR
Sample Bias (mV)
=6
EV
B=11.125 T
=11/2
1/2 filled LL
ESF
ESF
ESE
=5
dI/d
V (
nS)
=5
1/2 filled LL =9/2
=4
-6 -3 0 3 6 9 12
0
3
6
9
1/2 filled LL
B=14 T
=7/2
-15 -12 -9 -6 -3 0 3 5
0
3
6
9
dI/d
V (
nS)
B=11 T
ESL
ESR
Sample Bias (mV)
=6
Filling factor = 6
Fermi Level
Many Body Effects in Graphene
Polarizing Landau Levels
33
-15 -12 -9 -6 -3 0 3 5
0
3
6
9
0
3
6
90
5
10
0
5
10
0
5
10
15
0
5
10
15
B=11.75 T
B=11 T
B=11.5 T
B=11.375 T
B=11.25 T
ESL
ESR
Sample Bias (mV)
=6
EV
B=11.125 T
=11/2
1/2 filled LL
ESF
ESF
ESE
=5
dI/d
V (
nS)
=5
1/2 filled LL =9/2
=4
-6 -3 0 3 6 9 12
0
3
6
9
1/2 filled LL
B=14 T
=7/2
-15 -12 -9 -6 -3 0 3
0
5
10
Sample Bias (mV)
B=11.25 T
ESE
=5
dI/d
V (
nS)
Filling factor = 5
Enhanced spin splitting at
odd filling factors Enhanced valley splitting
at v=4
Many Body Effects in Graphene
Enhanced Exchange Interaction
For polarized LL, symmetric spin and antisymmetric
space wavefunction leads to enhanced exchange
interaction
34
Pauli Exclusion Principle
Wolfgang Pauli
Presentation Outline
35
B
Microscopy Honeycomb Lattices Magnetic Fields
Graphene“Quartet” Graphene Devices
Developing SPM Measurements for Devices
Graphene is Not Ideal in Real Devices
36
Y. Zhang et al. Nature (2010)
K. Novoselov et al. Nature (2005)
SPM Measurements in Graphene Devices
Potential Disorder in Graphene/SiO2
37
SiO2
Graphene
Impurities
EF @ Vg2
EF @ Vg1
Disorder potential variation
Vg2 > Vg1
Puddle size @ Vg2
Gate Electrode
N. M. R. Peres et al. PRB (2006)
E. H. Hwang et al. PRL (2007)
J. Martin et al. Nature Phys. (2008), (2009)
E. Rossi and S. Das Sarma PRL (2009)
Y. Zhang et al. Nature Phys. (2009)
Etc……
• Mobility
• Minimum conductivity
• Localization…
How does disorder affect:
SPM Measurements in Graphene Devices
Device Fabrication / Experimental Set-up
38
- Mechanically exfoliated graphene on SiO2/ Si substrate - Single / bilayer confirmed by Raman spectroscopy - Stencil mask evaporation
200 μm
Optical viewing and probe alignment in
CNST STM
S. Jung et al. Nature Physics (2010)
LDOS vs Transport Measurements
Gate Mapping Tunneling Spectroscopy
39
Vary density with
applied back gate
Spatially map density
fluctuations
Examine interaction
effects at EF
Vg = V1
Gate insulator Gate electrode
Vg = V2
Map dI/dV(E,Vg)
-300
-200
-100
0
100
200
300S
ampl
e B
ias
(mV
)
dI/dV (nS)-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30
Gate Voltage (V)
SPM Measurements of Graphene Devices
Gate Mapping Tunneling Spectroscopy (simulation)
40
LL0
LL1
LL0
LL-1
0
D F
G
E v n
n V V
LL-1
5 10 15 20 25 30 35-300
-200
-100
0
100
200
300
Gate Voltage (V)
Sam
ple
Bia
s (
mV
)
-2 0 2 4 6Filling Factor
SPM Measurements of Graphene Devices
Gate Mapping Tunneling Spectroscopy in An Electron
Puddle
41
LL0
LL1
Potential Disorder Map
electron-rich hole-rich
0
2 nS
dI/d
V
B=8 T STS Measurement of Dirac
Point Fluctuations
S. Jung et al. Nature
Physics (2010)
0.4 dI/dV (nS) 1
Sa
mp
le B
ias (
mV
)
dI/dV (nS)
N=0
N=-2
N=-1
-10 0 10 20 30 40 50
-250
-125
0
125
250
Sam
ple
Bia
s (
mV
)
Gate Voltage (V)
SPM Measurements of Graphene Devices
Evolution of Localization in Graphene Devices
42
B = 0 T B = 2 T B = 4 T B = 5 T B = 6 T B = 7 T B = 8 T LL0
LL-1
LL-2
LL0
LL1
LL2
SPM Measurements of Graphene Devices
Graphene Quantum Dot Formation in High Field
Coulomb blockade – Groups of four diamonds due to spin and valley
degeneracy
43
E
Vb
STM tip Sample
QD
Vacuum
barrier
I.S.
26 27 28 29
Gate Voltage (V)
1 2 3 4
Double barrier tunneling due to vacuum
barrier and incompressible regions
SPM Measurements of Graphene Devices
Graphene QDs Formed in Disorder Potential
44
Si (back gate)
SiO2 Cg Vg
A
I STM tip
Vb
Landau level formed
(Compressible region)
Resistive region
(Incompressible strip) B > 0 T
SPM Measurements of Graphene Devices
Graphene QDs Formed in Disorder Potential
45
Si (back gate)
SiO2 Cg Vg
A
I STM tip
Vb
Landau level formed
(Compressible region)
Resistive region
(Incompressible strip) B > 0 T
EF
EF
Metallic
Compressible
Insulating
Incompressible
SPM Measurements of Bilayer Graphene Devices
STS Allows Direct Measurement of Bilayer Potentials
46
E
k
ΔU = 0
±LL0, 1
LL2
LL3
LL4
LL2
LL3
LL4
ΔU > 0
LL2
LL3
LL4
LL2
LL3
LL4
+LL0, 1 (top)
-LL0, 1 (bottom)
ΔU < 0
LL2
LL3
LL4
LL2
LL3
LL4
-LL0, 1 (bottom)
+LL0, 1 (top)
!STS Selects Layer
Polarized States
20 nm
STM topography image Disorder potential
Electron
puddle
Hole
puddle
20 nm
Probing Spatial Distribution of Disorder Potential
47
20 nm
Single Layer Bilayer
Electron puddle
Hole puddle
SPM Measurements of Bilayer Graphene Devices
Gate Mapping Allows Direct Measurement of Bilayer Gap
48
Quantitative determination of bilayer gap
Variation on a microscopic scale in both
magnitude and sign
G. Rutter et al. Nature Physics (2011)
0 10 20 30 40 50 60
-60
-40
-20
0
20
40
,
,
,
Gate Voltage (V)
Ene
rgy
Gap
, U
(m
eV)
Hole puddle
Electron puddle
8 T 6 T 0 T
What’s the Next in Atomic Scale Measurement
Development
Coordinated approach to combine new atomic-scale
measurement methods, synthesis, and device fabrication
Atomic scale and macroscale measurements on the same test devices
How does microscale properties from substrates/gate insulators, contacts
etc… determine macroscale performance
Develop measurements for new qraphene device concepts, i.e. Veslago lens
BiSFET device
Fabrication and measurement of topological insulators – more Dirac
MBE and bulk crystal growth, atomic characterization studies
Combined STM, AFM and spin-polarized STM on device geometries
New high-throughput STM/AFM/SGM system
Multi-terminal STM/STS measurements on devices that combine
simultaneous transport and atomic characterization measurements to
optimize device performance
Continue to seek collaborations that leverage our capabilities
49
Collaborators
50
Jeonghoon Ha Jungseok Chae Young Kuk
Sander Otte Young Jae Song
Graphene/TI mK Crew
Niv Levy Tong Zhang
Collaborators
51
Graphene Device Crew
Greg Rutter Suyong Jung Nikolai Klimov
Nikolai Zhitenev Dave Newell Angie Hight-
Walker
Collaborators
52
GT Epitaxial Graphene Crew Graphene Theory Crew
Yike Hu
Phil First
Walt de Heer Hongki Min Shaffique Adam
Mark Stiles Allan MacDonald Britt Torrance Eric Cockayne