High-Field, Thermal & Energy
Properties of 2D Devices & Layers
Eric Pop
Electrical Engineering (EE) and Precourt Institute for Energy (PIE)
Stanford University
http://poplab.stanford.edu
E. Pop
Acknowledgements
1
http://poplab.stanford.edu
• Our “flatland” (2D) subgroup and alumni:
– Dr. Vincent Dorgan, Dr. Myung-Ho Bae
– Dr. Enrique Carrion, Dr. Zuanyi Li, Dr. Sharnali Islam
– Chris English, Kirby Smithe, Michal Mleczko, Ning Wang
• Undergrads:
– Maryann, Yuan, Andrew, Tim, Justin
• Sponsors:
– Air Force Research (AFOSR)
– National Science Foundation (NSF)
• Collaborators:
– E. Reed, K. Goodson, Y. Nishi, I. Fisher, K. Saraswat, H.-S.P. Wong (Stanford)
– J. Lyding (UIUC), D. Akinwande (UTA), I. Knezevic (Wisc.), R. Sordan (Milano),
A. Roy, V. Varshney (AFRL)
E. Pop
2
2D Materials in Our Lab
CVD growth of graphene, BN and MoS2
MoTe2
HfSe2
Intensity (a.u.)
ZrSe2
BN
Si
A1g 194 cm-1
100
200
300
400
Raman Shift (cm-1)
20 µm
HfSe2
~
~
4
WTe2 (metallic)
x 10
WTe2
3.5
exfoliate
3-6 Layer
1.5
1
10 µm
Bulk
0.5
mono-MoS2
0
50
~
~
100 µm
2
~
~
Intensity (a.u.)
3
2.5
100
150
200
250
300
-1
Raman Shift (cm )
CVT growth of MoTe2, WTe2, ZrSe2, HfSe2
collab. H.S.-P. Wong, Y. Nishi, I. Fisher
E. Pop
3
Device (Transistor) Scaling
• Problem: 20th century transistors “carved” out of 3D materials (Si) 
surface roughness restricts mobility, band gap, scaling of dimensions
LG ~ (tchtox)1/2
tch
roughness
21st
• Solution:
century transistors with atomically thin 1D and 2D
materials (<1 nm) can re-enable LG scaling
104
1D carbon
nanotube (CNT)
1 nm
µ (cm2V-1s-1)
graphene
2D TMD (MoS2,
WTe2, ZrSe2)
WSe2
102
MoS2
<1 nm
10
SOI
0
sources: K. Uchida, A. Kis, IBM, our work
E. Pop
CNTs
103
1
2
3
4
tch or d (nm)
4
High-Field Transport in Suspended Graphene
V. Dorgan, A. Behnam, H. Conley, K. Bolotin, E. Pop, Nano Letters 13, 4581 (2013)
2000
1500
4000
suspended
graphene
VD
D
1000
500
T (K)
 (Wm -1 K-1 )
S
-
SiO2
Si
2000
steeper drop-off
than graphite
Suspended graphene allows us to study
intrinsic coupled electrical-thermal properties
experiments
1000
0
300
simulations
μ0 = 25,000
cm2/V/s
0.8
0.6
more
disordered
0.4
μ0 =
2,500
cm2/V/s
0.2
0
0
3 0
1
2
F (V/m)
1
2
F (V/m)
1,000
T (K)
2,000
5
cleaner
v (107 cm/s)
1
3000
1 μm
VG
I/W (mA/m)
This work (exfoliated)
This work (CVD)
Graphite (in-plane)
4
high-field
intrinsic vsat
3
2
1
0
0 2 4 6 8 10 12 14
Sample #
3
E. Pop
5
Simulation: Ambipolar + Poisson + Heating
models “GFETTool” and “S2DS” available on http://nanoHUB.org
2


C

1  Cox

V0  VGx    ox V0  VGx   4nix2 

2 q
 q



I  sgn( px  nx ) qW ( p x  nx )vx
Vx , j  Vx , j 1  sgn( p x  nx )
A
E. Pop
vx
 0 1  vx / vsat
  T 
k
  p ' g T  T0   0
x  x 
Charge
Continuity

x
iterate
iterate
p x , nx 
Poisson
Thermal
6
Transport at Graphene Grain Boundaries
K. Grosse, V. Dorgan, D. Estrada, J. Wood, I. Vlassiouk, G. Eres, J. Lyding, W. King, E. Pop, Appl. Phys. Lett. 105, 143109 (2014)
Goal: Understanding nanometer scale graphene transport, heating and reliability
– Measured heating at graphene grain boundaries (GB)
– Deduced grain boundary resistance; Collaboration with ORNL
technique:
scanning Joule expansion microscopy (SJEM)
E. Pop
7
High-Field Transport in MoS2
A. Serov, V. Dorgan, C. English, E. Pop, Device Research Conf. (2014)
V. Dorgan, A. Serov, C. English, E. Pop, submitted (2015)
VD
S
300
Au
I D ( A/ m)
tch
VG
120 K
250
MoS2
• Negative differential conductance
(NDC) at high-field in MoS2
80 K
160 K
VGT ≈ 30 V
200
200 K
150
300 K
100
400 K
• Simulations suggest combination of:
– Self-heating and µ(T) ~ T-2
– Intervalley scattering (K to Q valley)
50
500 K
1 µm
0
Au
0
2
4
V (V)
6
• Results shed light on band structure
8
D
4
8
7
2
300 K
1.5
500 K
1
0
F (V/ m )
E. Pop
2
0
4
ΔE = 130 meV
3
1
2
Q
4
0.5
0
Γ
5
6
1
0
experiment (T0 = 80 K)
0
2
4
6
-1
-1
0
1
k (nm -1)
x
100
1
10-5
0
10-10
y
200 K
y
2.5
Upper
Valley
(Q)
Lower
Valley
(K)
K
6
vd (10 6 cm/s)
vd (10 6 cm/s)
3
k (nm-1)
80 K
k (nm-1)
3.5
-1
-1
0
1
k (nm -1)
10-15
x
F (V/ m )
8
Some Recent Results on Monolayer CVD MoS2
K. Smithe, C. English, S. Suryavanshi, E. Pop, Device Research Conf. (2015)
• Growth of monolayer MoS2 by CVD on PTAS/SiO2/Si
• Effective mobility (µeff) comparable to exfoliated material
Good current achieved (~275
µA/µm) in 80 nm device, but contact
resistance remains dominant
E. Pop
9
Summary of Challenges in 2D Devices
Contact Resistance (RC):
Metals
Interfaces:
VTG
VD
Graphene
VS
Metals
High-κ
Graphene
Si (p++)
Substrate
LT
o Pre vac. anneal
o Post vac. anneal
Si (p++)
VBG
W
L
SiO2 (90 or 300 nm)
VD
Scope
VTG
GFET
50 Ω
VBG
Si (p++)
RL
VP
tON,D
tON,TG
VD
tOFF
1.5
Material Quality:
SiC
Growth
+Variability!
[2]
R [K]
[1]
1.6
1.2
0
(b)
tON,TG
1
tON,D
0.5
VTG
CVD
Rprobe Cprobe
Cpad
VS
VBG
V [V]
Oxide
d
0
100 200
Time [ns]
DC
tON,TG = 100 µs
= 10 µs
0.8
0.4
-8
= 400 ns
(air)
-4
0
VTG [V]
4
8
E. Carrion, E. Pop et al, IEEE TED, 61, 1583 (2014)
E. Pop
10
A Few Words on Thermal 2D Work
• Large in-plane thermal conductivity of graphene, BN (>500 W/m/K)
• Ultra-low cross-plane thermal conductivity of layered WSe2 (<0.1 W/m/K)
– Lower than plastics and comparable to air
• Huge thermal anisotropy in all layered 2D materials (>10-100x)
• MRS Bulletin review with AFRL:
E. Pop, V. Varshney, A.K. Roy, "Thermal
Properties of Graphene: Fundamentals and
Applications," MRS Bulletin 37, 1273 (2012)
• Large thermopower in TMDs (S ~ 0.5 mV/K)  Thermoelectrics?
E. Pop
11
Ballistic Phonons in Graphene Ribbons
M.-H. Bae, Z. Li, Z. Aksamija, P. Martin, F. Xiong, Z.-Y. Ong, I. Knezevic, E. Pop, Nature Comm. 4, 1734 (2013)
First measurement of quasi-ballistic
heat flow in graphene near room
temperature
graphene
nanoribbons
First measurement of phonon edge
scattering in graphene nanoribbons
9
W~130 nm
W~85 nm
W~65 nm
W~45 nm
-2
-1
G/A (Wm K )
10
10
8
~T1.5
10
E. Pop
long graphene13
(L ~ 10 μm)
7
50
100
T (K)
200
300 400
“long”
Maximum heat flow also
limited by edge scattering.
But, “short” graphene devices
reach up to ~35% of theoretical
ballistic thermal limit at room T.
(diffusive)
“short”
quasi-ballistic
12
Looking Ahead: Future Opportunities
Could we:
collab: E. Reed, K. Goodson, K. Saraswat, H.-S.P. Wong, Y. Cui
– Exploit anisotropy for routing heat? (thermal diode)
– Separate thermal and electrical flow? (thermal transistor)
– Design electronics with built-in thermoelectric cooling?
– Achieve transparent heat spreaders and flexible thermoelectrics?
E. Pop
13
Looking Ahead: Future Opportunities
Could we:
collab: E. Reed, K. Goodson, K. Saraswat, H.-S.P. Wong, Y. Cui
– Exploit anisotropy for routing heat? (thermal diode)
– Separate thermal and electrical flow? (thermal transistor)
– Design electronics with built-in thermoelectric cooling?
– Achieve transparent heat spreaders and flexible thermoelectrics?
electrolyte LiPF6
Li
Al / MoS2
G (MW/m2/K)
Thermal Switching
(
)
Pristine MoS2
9
3
1
0
2
4
Time (min)
6
8
A. Sood, F. Xiong, […], E. Pop, Spring MRS (2015)
E. Pop
14
Summary
• Goals: understand 2D materials and
devices for analog (e.g. graphene) and
digital (e.g. MoS2) electronics
• High-field transport experiment & models
• Insights into band structure, grains, power
dissipation
• Unique 2D thermal properties
• Next:
Contact:
W: http://poplab.stanford.edu
E: [email protected]
– Quasi-ballistic transport evaluation
– Role of contact geometry
– Exploit heterostructures, thermal
anisotropy, thermoelectric properties
E. Pop
15
E. Pop
16