Exploring Pathways Towards Heterostacks
of Graphene and Hexagonal Boron Nitride
Jürg Osterwalder,
Physics Department,
University of Zürich, Switzerland
US-EU Workshop on 2D Layered Materials and Devices
Arlington, VA, April 22-24, 2015
Collaborators
External Collaborations
Current:
Carlo Bernard
Thomas Kälin
Huanyao Cun
Elisa Miniussi
Single-Crystal Metal Films
Michael Weinl
Stefan Gsell
Matthias Schreck
/ University of Augsburg
Thomas Greber
Former:
Adrian Hemmi
Silvan Roth
2D Materials
Preparation of artificial 2D
materials and devices
r
olaye
n
o
M
fer
trans
h-BN
Exfoliation
metal
metal
metal
MoS2
K. S. Novoselov and
A. H. Castro Neto
Phys. Scr. T146, 014006 (2012).
NbSe2
metal
3
Our approach:
single-crystalline metal substrates
at the four-inch wafer scale
Goals:
<110
>
• single domain growth of
h-BN and graphene
using UHV-CVD
• explore growth modes of
heterostacks
4” Si(111) wafer with 150 nm Rh(111) film
(from Matthias Schreck, University of Augsburg)
… same for Ir(111), Ni(111), Ru(0001), Pt(111)…
4
• transfer and stack single layers
with defined lattice orientation
Contents
• single-layer CVD growth on metal surfaces
• single crystal substrates for wafer scale 2D films
• CVD growth of g/h-BN heterostack on (111) surfaces
• transfer of single crystal 2D films
5
Growth of a h-BN monolayer on Ni(111)
leak valve
CVD under UHV conditions
pump
base pressure: 2x10-10 mbar
borazine exposure: 5x10-6 mbar
glas tube cont.
liquid borazine
borazine
(BN)3H6
XPS
B 1s
1050 K
Borazine+Ni(111)
h-BN on Ni(111)+3H2
6
Nagashima, Oshima et al., PRB 51, 4606 (1995)
h-BN on Ni(111)
- perfect monolayer growth on Ni(111)
upon exposure to ~100 L of borazine
-  good match of lattice constant (mismatch=+0.8%)
-  atomic and electronic structure well known
recipe by Nagashima et al., PRB 51, 4606 (1995)
I=1nA, V=0.1V
7
200 nm
26 Å
I=9nA, V=-23mV
W. Auwärter et al., Surf. Sci. 429, 229 (1999)
Chemically sensitive
structure probe
(probing depth ~ 1nm)
Can identify different
phases at the interface
countintensity
rate (arb. units)
X-ray photoelectron
Diffraction (XPD)
experiments:
N1s
20
B1s
1340
1346
1544
1550
10
B
N
0
800
1000
1200
1400
1600
electron kinetic energy (eV)
N 1s
__
(1342 eV) [112]
B 1s
(1550 eV)
__
[112]
XPD signals:
forward scattering:
- B to N n.n.
- N to B n.n.
A=48%
A=36%
-> interference fringes
absence of peaks near
center => single layer
W. Auwärter et al., Surf. Sci. 429, 229 (1999).
A (12x12) h-BN/Rh(111) superstructure
Single h-BN layer
- 13x13 h-BN on 12x12 Rh units
- periodicity: 3.2 nm
- not a simple moiré pattern
- pores of ca. 20 Å diameter of
strongly bonded h-BN
- wires of weakly bonded h-BN
A boron nitride nanomesh ! 9
LEED:
superstructure
M. Corso et al., Science 303, 217 (2004).
Single-layer model for h-BN nanomesh
13 BN
12 Rh
pore
wire
(NB)=(top,fcc)
10
R. Laskowski, P. Blaha et al., PRL 98, 106802 (2007)
Pentanone as a precursor
for graphene growth
g/Rh(111)
11
S. Roth et al., Surf. Sci., 605, L17 (2011).
Contents
• single-layer CVD growth on metal surfaces
• single crystal substrates for wafer scale 2D films
• CVD growth of g/h-BN heterostack on (111) surfaces
• transfer of single crystal 2D films
12
Single-crystalline metal films on 4 Si wafers
Group of Matthias Schreck
Institut für Physik, Universität Augsburg
Single-crystalline Ir films for
epitaxial diamond growth
PVD, 800°C,
micrometers
e-beam evaporation,
650°C, 100 nm
PLD, 750°C,
20-40 nm
Works for (001) growth …
13
S. Gsell, M. Schreck et al.,
Appl. Phys. Lett. 84, 4541 (2004)
Single-crystalline metal films on 4 Si wafers
… but also for (111) growth:
XPD -> twin-free diamond nucleation
14
XRD pole figures
-> twin free, low mosaicity
M. Fischer, M. Schreck et al.,
J. Appl. Phys. 104, 123531 (2008)
Single-crystalline metal films
… and also for Rh(111) growth:
15
S. Gsell, M. Schreck, unpublished.
The four-inch wafer growth chamber at UZH
Cleanroom
ISO 8
+
flow box inside
16 A. Hemmi et al., Rev. Sci. Instrum. 85, 035101 (2014)
Quality control of h-BN growth across wafer surface
60 eV
LEED patterns at 65 eV
Use LEED superstructure as a quality measure for h-BN film
(after transfer through air + annealing)
17 A. Hemmi et al., Rev. Sci. Instrum. 85, 035101 (2014)
Contents
• single-layer CVD growth on metal surfaces
• single crystal substrates for wafer scale 2D films
• CVD growth of g/h-BN heterostack on (111) surfaces
• transfer of single crystal 2D films
18
CVD growth beyond the monolayer: g/h-BN/Cu(111)
LEED: lattice constants
h-BN / Cu(111)
5x10-6 mbar borazine
Careful analysis of
LEED spots:
• each layer has a
different lattice constant
graphene / h-BN / Cu(111)
2.2 mbar 3-pentanone
• the peak positions are
consistent with the
intrinsic lattice constants
of h-BN and graphene
(2.50 Å and 2.46 Å)
=> Only small angle domains in both layers (~ 2°).
19
S. Roth et al., Nano Lett. 13, 2668 (2013)
Characterisation by quantitative XPS
Composition and coverage:
graphene / h-BN
h-BN
After borazine exposure:
fairly stoichiometric, single layer of h-BN
After pentanone exposure:
h-BN still stoichiometric, single layer,
intensities slighly attenuated by
single graphene layer on top
20
carbon – h-BN intensity ratios
increase towards grazing emission
=> graphene layer is on top of h-BN layer
S. Roth et al., Nano Lett. 13, 2668 (2013)
XPD from the heterostack – a puzzle !
N 1s
C 1s
B 1s
SSC theory
(free-standing)
⇒  Single layer of h-BN
⇒  Single graphene layer
Both well ordered.
But:
Why no forward scattering
of the B and N signal
through the graphene
layer?
Answer:
Because the layers are incommensurate !
21
S. Roth et al., Nano Lett. 13, 2668 (2013)
XPD: commensurate vs. incommensurate overlayer
commensurate
incommensurate
N 1s
N 1s
strong forward
scattering signal
22
forward scattering
signals smear out
completely
S. Roth et al., Nano Lett. 13, 2668 (2013)
Confirmation: STM sees a moiré pattern
Moiré pattern on most of the terraces
Periodicity ~ 9 nm
100 nm
8 Å
Regular ribbons (>100 nm x 9 nm, ca. 0.8 nm high):
=> formation of grafolds ?
23
K. Kim et al.,
Phys. Rev. B 83, 245433 (2011).
S. Roth et al., Nano Lett. 13, 2668 (2013)
ARPES data: the π bands of h-BN and graphene
g/h-BN/Cu(111)
h-BN/Cu(111)
graphene
Cu sp
Cu 3d
h-BN
h-BN
π band
h-BN
σ band
Γ
In the heterostack,
we observe the π bands
of both individual layers.
• h-BN maintains a large
band gap
• graphene shows linear
dispersion up to EF
• the two layers are
electronically decoupled
(exhibit slightly different BZ)
K
24
S. Roth et al., Nano Lett. 13, 2668 (2013)
Contents
• single-layer CVD growth on metal surfaces
• single crystal substrates for wafer scale 2D films
• CVD growth of g/h-BN heterostack on (111) surfaces
• transfer of single crystal 2D films
25
A refined ‘bubbling’ method using hydrogen intercalation
Approach similar to that used
for graphene in
E. Koren, P. Sutter at al,
Appl. Phys. Lett. 103, 121602 (2013)
Schematic potentiostat set-up
for the bubbling transfer process.
Solution: 0.1 M HClO4
Hydrogen intercalation
prior to bubbling weakens
the bonding to the metal substrate
26
B 1s
Si 2s
Si 2p
N 1s
O 1s
Rh 3d
XPS analysis of transferred h-BN monolayer
27
Conclusions
Single-layer CVD growth on metal surfaces:
• We can grow large domains of high quality h-BN and graphene
Single crystal substrates for wafer scale 2D films:
• Single layer quality approaching that of single crystal substrates
• Scalable approach
CVD growth of g/h-BN heterostack on (111) surfaces:
• Growth beyond single layer difficult to control
• Quality inferior to single layer growth
Transfer of single crystal 2D films:
• Transfer of h-BN single layers appears to be more difficult than
graphene – optimization in progress (collaboration within Flagship)
• Can get aligned flakes
28
Collaborators
External Collaborations
Current:
Carlo Bernard
Thomas Kälin
Huanyao Cun
Elisa Miniussi
Single-Crystal Metal Films
Michael Weinl
Stefan Gsell
Matthias Schreck
/ University of Augsburg
Thomas Greber
Former:
Adrian Hemmi
Silvan Roth