Page 1
Self-consistent screening in graphene FOR1807 workshop: Advanced computational methods for strongly correlated quantum systems, Würzburg University February 26th 2015
Shaffique Adam Yale-NUS College and
Center for advanced 2D materials (formerly Graphene Research Center)
Department of Physics National University of Singapore
Page 2
[email protected] ! 2
Center for Advanced 2D materials (formerly Graphene Research Center)"
Theory colleagues"
Antonio Castro Neto
Showing NUS physicists and collaborators… (not shown: chemists, engineers, people at NTU etc.)!
Baowen Li Feng Yuan Ping
Hsin Lin Vitor Pereira
Zhang Chun
Experimental colleagues"
Barbaros Özyilmaz
Su Ying Quek
Giovanni Vignale Kian Ping Loh
Andrew Wee Goki Eda Jens Martin
Ji Wei
Slaven Garaj Utkur Misaidov
Sow Chorng Haur
Peter Ho
Jose Gomes
Lay-Lay Chua Wei Chen Christian Nijhuis
Page 3
[email protected] ! 3
Research Group"Postdoctoral Research Fellows!
Jeil Jung" Mirco Milletari" Derek Ho"
3
Graduate Student Researchers!
Indra Yudhistira " Navneeth Ramakrishnan" Tang Ho Kin"
Joao Rodrigues"
Jia Ning Leaw"
Page 4
[email protected] !
The perfect 2DEG"
4
1 nm
Graphene is all surface and no bulk!
“God made the bulk; surfaces were invented by the devil”! – Wolfgang Pauli!
Figure from G. Rutter
Page 5
[email protected] !
Delicate interplay between disorder, interactions and quantum effects"
5
Most experiments are in the regime, where all these effects are relevant, but not dominant!!
Disorder
Electron!Interactions
Quantum !Interference
Image credits: Fuhrer, Geim, Westervelt,
See e.g. !!
M. S. Fuhrer and S. Adam,!Nature, news and views, 458 38 (2009); and!!
S. Das Sarma, S. Adam, E. Hwang and E. Rossi, !Reviews of Modern Physics 83 407 (2011)
Page 6
[email protected] !
Semi-classical picture"
6
Disorder
Figure from M. Wayne
Disorder!Potential!
Electron gas!
Page 7
[email protected] !
Map to classical percolation"
7
M. Isichenko, Rev. Mod. Phys. (1992)
Potential !Maxima!
Potential !Minima!
Saddle !Point != bond!
Metal above percolation threshold, !!Insulator below percolation threshold!
Unlike conventional 2DEGs, graphene remains metallic even for strong disorder
Disorder
Page 8
[email protected] !
Dirac materials: interaction strength and density tuned independently"
8
Interactions rs =
e2
vFrs =
e2
vF κ1 +κ2( )
C. Juang, S. Adam, J-H. Chen, E. D. Williams, S. Das Sarma, and M. S. Fuhrer, Phys. Rev. Lett. 101, 146805 (2008).
Page 9
[email protected] !
Fermi liquid away from DP"
9
ε(q)=1+ 2πe2
κqΠ(q) § Screening is “metallic” on distances larger than the Fermi wavelength !
§ Screening is like a dielectric “insulator” on shorter distances!§ Long-range nature of coulomb tail can not be screened
Weak Interactions = Screening
Page 10
[email protected] !
What about the Dirac point?"
10
e.g. Sorella et al. Sci. Rep. (2012) F. Assaad (2012-2014) Semi-metal to AFM phase transition consistent with SU(2) SU(2) Heisenberg Gross-Neveu universality class
U = 0 U = oo
Uc
For strong interactions, !Monte Carlo reveals gapped phase (anti-ferromagnet)
Strong Interactions
Page 11
[email protected] !
Divergent Fermi velocity"
11
"
V. Kotov, B. Uchoa, V. Pereira, A. H. Castro Neto and F. Guinea, Rev. Mod. Phys. (2012)
G(k + q)
V (q)ε(q)
ve↵v0
= 1 +
rs4
log
⇠ 1
ka
�For weak interactions, perturbation theory reveals diverging Fermi velocity
Page 12
[email protected] !
No consistent picture at DP"
12
1. Renormalization Group [e.g. Sachdev (1998) / Guinea (1997) ]
U = 0 U = oo
Fixed point is free Dirac theory where interactions are “dangerously irrelevant”
2. Diagrammatic perturbation approaches [e.g. Das Sarma et al. (2007)] “marginal Fermi liquid”
3. Hubbard model on a honeycomb lattice (semi-metal to AFM Mott transition occurs at interaction strengths outside the experimental window) e.g. Sorella, Assaad, Katsnelson (2012-2014). “qualitatively similar to a Fermi liquid”
4. Lattice Monte Carlo applied to Dirac fermions with momentum cut-off e.g. Drut and Lahde (2009-2014). “chiral symmetry breaking insulating state for suspended graphene”
U = 0 U = oo
Uc
Page 13
[email protected] !
Perfect transmission"
13
Classical world
Quantum world
Klein world
P-N junctions are completely transparent for direct incidence. Results in beam collimation.
Quantum Interference
Page 14
[email protected] !
Universal ballistic σmin"
14
N"
N"
N"
N"
P"
N"
€
→4e2
πh
M. Katsnelson (2006);!C. Beenakker et al. (2006)
Without disorder and without interactions, both Kubo and Landauer formalisms give a universal ballistic minimum conductivity at the Dirac point.
Quantum Interference
Page 15
[email protected] !
Delicate interplay between disorder, interactions and quantum effects"
15
Most experiments are in the regime, where all these effects are relevant, but not dominant!!
Disorder
Electron!Interactions
Quantum !Interference
Page 16
[email protected] !
Scaling theory of localization"
16
Schrödinger 2DEG
2DEG with SOC
Graphene Disorder + Quantum Interference
Page 17
[email protected] !
No Anderson localization"
17
0 # per puddle2K electronsπ≈
Semi-classical regime Quantum regime
S. Adam, P. W. Brouwer and S. Das Sarma, Phys. Rev. B R 79, 201404 (2009)
Self-consistent diffusive Boltzmann transport theory
Ballistic universal quantum theory
Dirac point conductivity (numerically exact)
Page 18
[email protected] !
How does an inhomogeneous Fermi liquid screen external potentials? "
18
Vscreened =Vbareε[q]
Example of Thomas-Fermi screening: Interactions + Disorder
Page 19
[email protected] !
How does an inhomogeneous Fermi liquid screen external potentials? "
19
Vscreened =Vbareε[q]
Regular 2DEG: ✏(q) = 1 +2e2
~2me
q
Example of Thomas-Fermi screening: Interactions + Disorder
Page 20
[email protected] !
How does an inhomogeneous Fermi liquid screen external potentials? "
20
Vscreened =Vbareε[q]
Regular 2DEG: ✏(q) = 1 +2e2
~2me
q
Dirac electrons: ✏(q) = 1 +
e2
~vF
p⇡n
q
Example of Thomas-Fermi screening: Interactions + Disorder
Page 21
[email protected] ! 21
Vscreened =Vbareε[q]
Regular 2DEG: ✏(q) = 1 +2e2
~2me
q
Dirac electrons: ✏(q) = 1 +
e2
~vF
p⇡n
q
Example of Thomas-Fermi screening:
vF2DEG�! ~
p⇡n
2m
How does an inhomogeneous Fermi liquid screen external potentials? "
Interactions + Disorder
Page 22
[email protected] !
How does an inhomogeneous Fermi liquid screen external potentials? "
22
Vscreened =Vbareε[q]
Regular 2DEG: ✏(q) = 1 +2e2
~2me
q
Dirac electrons: ✏(q) = 1 +
e2
~vF
p⇡n
q
Example of Thomas-Fermi screening:
Inhomogeneous Dirac: ✏(q) = 1 +
e2
~vF
p⇡nrmsp3q
Interactions + Disorder
Page 23
[email protected] !
What is nrms(nimp)?"
23
250 nm x 250 nm
2. Screened potential correlation function:
( ) (0)V r V
ξ
1. Histogram of the carrier density (distribution function):
[ ]P nrmsn∝
r
2 (0)V
Characterizes the puddle depth and length n
Any physical observable could then be calculated
Statistical properties of density fluctuations and the Dirac point
Page 24
[email protected] !
What is nrms(nimp)?"
24
250 nm x 250 nm
2. Screened potential correlation function:
( ) (0)V r V
ξ
1. Histogram of the carrier density (distribution function):
[ ]P nrmsn∝
r
2 (0)V
Characterizes the puddle depth and length n
Any physical observable could then be calculated
Statistical properties of density fluctuations and the Dirac point
Calculate dimensionless quantity: ?=imp
rms
nn
Page 25
[email protected] !
Ansatz for Inhomogeneous screening"
25
),,()()(nrqqqs
barescr ε
φφ =
+ + -
n(r)
c)(dielectriconst )( =qε
(metal) q
q 1 )( TF+=qε(Dirac)
k 2q ;41
2 ;2
1 ),,(
F⎪⎪⎩
⎪⎪⎨
⎧
>>+
<<+=
qrk
kqr
rkqsF
Fs
sF
π
ε
Page 26
[email protected] !
Ansatz for Inhomogeneous screening"
26
( )0,,)()(=
=nrqqq
s
barescr ε
φφ
( ),...,,,])[,,( 32 nnrqnPrq ss εε →
Page 27
[email protected] !
Ansatz for Inhomogeneous screening"
27
( )0,,)()(=
=nrqqq
s
barescr ε
φφ
( ),...,,,])[,,( 32 nnrqnPrq ss εε →
Page 28
[email protected] !
Ansatz for Inhomogeneous screening"
28
),,()()(rmss
barescr nrq
qqεφ
φ ≈
( )2,,])[,,( nrqnPrq ss εε ≈
Page 29
[email protected] !
Ansatz for Inhomogeneous screening"
29
),,()()(rmss
barescr nrq
qqεφ
φ ≈
[ ]rmsssimp
rms ndrCrnn , 2 0
2=
S. Adam, E. H. Hwang, V. M. Galitski and S. Das Sarma Proc. Nat. Acad. Sci. USA 104, 18392 (2007).
Page 30
[email protected] !
Local screening vs. global screening"
30
S. Adam, E. H. Hwang, V. Galitski, S. Das Sarma, Proc. Nat. Acad. Sci. USA 104, 18392 (2007). !
E. Rossi and S. Das Sarma !PRL 101, 166803 (2008)!!E. Rossi, S. Adam and S. Das Sarma, PRB 79, 245423 (2009) !
Vscreened =Vbare
ε[q,nrms ]
ε[q,nrms ]
Page 31
[email protected] !
Numerical verification"
31
nrms
nimpS. Adam, E. H. Hwang, V. M. Galitski and S. Das Sarma Proc. Nat. Acad. Sci. USA 104, 18392 (2007).
E. Rossi, S. Adam and S. Das Sarma Phys. Rev. B 79, 245423 (2009)
Page 32
[email protected] !
Puddle Formation"
32
20 nm !
STM topography image!
Electron"puddle"
Hole "puddle!
20 nm !20 nm !
Single Layer Graphene" Bilayer Graphene"
Electron puddle!
Hole puddle!
With J. A. Stroscio (NIST) Phys. Rev. B 84, 235421 (2011)
Page 33
[email protected] ! 33
Agreement with experiments [1]"
33
20 nm !
STM topography image!
Electron"puddle"
Hole "puddle!
20 nm !20 nm !
Single Layer Graphene" Bilayer Graphene"
Electron puddle!
Hole puddle!
Collaboration with J. A. Stroscio (NIST) Phys. Rev. B 84, 235421 (2011)
No adjustable parameter!
Page 34
[email protected] !
Monolayer graphene from a Brian LeRoy (Arizona)"
34
Agreement with experiments [1]"
34
20 nm !
Electron"puddle"
Hole "puddle!
20 nm !
Bilayer Graphene"
Electron puddle!
Hole puddle!
Collaboration with J. A. Stroscio (NIST) Phys. Rev. B 84, 235421 (2011)
No adjustable parameter!
No adjustable parameter!