Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

www.advmat.

de
www.MaterialsViews.com
COMMUNICATION

Epitaxial Graphene Growth by Carbon Molecular Beam


Epitaxy (CMBE)
By Jeongho Park,* William C. Mitchel, Lawrence Grazulis, Howard E. Smith,
Kurt G. Eyink, John J. Boeckl, David H. Tomich, Shanee D. Pacley, and John. E. Hoelscher

Since graphene was first prepared by simple mechanical exfo- on the nucleation and growth of graphene on a metal surface.
liation of highly ordered pyrolytic graphite (HOPG),[1] many Al-Temimy et al.[11] demonstrated the possibility of the graphene
approaches have been studied for its growth or preparation growth by the evaporation of carbon. Hackley et al.[12] claimed
including annealing of silicon carbide (SiC),[2,3] exfoliation,[4] graphitic (amorphous) carbon formation, which is not graphene,
chemical vapor deposition (CVD) on transition metals,[5,6] by e-beam evaporation of carbon rod onto a Si (111) surface.
and the formation and subsequent decomposition of graphite None of these experiments, however, demonstrated the growth
oxide.[7] None of these approaches, however, involve direct of high quality graphene. Further these experiments focused on
deposition of carbon onto the substrate of interest. Among the evaporation of carbon from bulk graphite. We report here the
the above mentioned methods, the thermal decomposition of direct growth of epitaxial graphene on SiC substrates from C60
SiC[2,3] has attracted interest due to its potential for uniform, and bulk graphite filament carbon sources. We demonstrate that
wafer scale growth of graphene. An advantage of this method this direct growth approach provides superior quality and homo-
is that it eliminates the process of transferring the graphene geneous graphene layers. We also show that thickness, structure,
from the metal or bulk graphite to the substrate on which and electronic properties can be controlled by carbon flux, the
the graphene will ultimately be used. In this process the SiC types of carbon source, and substrate temperature.
decomposes at high temperatures during which Si evaporates Epitaxial graphene was grown on chemical-mechanical
and carbon remains to form graphene. It is generally accepted polished (CMP) on-axis 4H-SiC (0001) substrates by carbon
that a specific SiC surface reconstruction, 6√3 × 6√3 R30°, is MBE (CMBE) in an EPI Model 930 MBE system modified for
required for the carbon to form graphene. While great strides high temperature growth. Prior to growth a tantalum film was
have been made using this technique on both (0001) and deposited on the back of all samples for thermal management.
(0001) surfaces of SiC, it is limited by a lack of control over the The substrate temperature was monitored with a pyrometer.
graphene growth. The only major control parameters are the Prior to insertion in the chamber the SiC substrates were
substrate temperature, growth time, and chamber pressure. A cleaned with acetone and methanol, followed by a 10% HF
true epitaxial growth process would provide many benefits over dip to remove any native oxide. The manipulator of the MBE
simple decomposition. In addition to improved process control system was modified to obtain a maximum temperature of
through the potential to vary the carbon flux on the substrate, 1600 °C by use of a graphite filament heater powered by a DC
it would enable innovative graphene based heterostructures power supply. The substrate holder for all experiments was
such as the BisFET proposed by Banerjee et al.[8] This device further modified by opening the back side to expose the wafer
would consist of two graphene layers separated by a thin dielec- to the heater so as to increase the maximum temperature that
tric. Other novel graphene based heterostructures which could could be obtained. The carbon molecular beams were obtained
better exploit the bandgap tunabilty of bilayer graphene[9] for from two different sources. First, C60 was used with a conven-
opto-electronic applications might also be possible as well. tional effusion cell to provide a uniform carbon flux.[13] The
Molecular beam epitaxy (MBE) is a well known and widely used carbon flux was measured with an ion gauge. For graphene
technique for the production of high quality and homogeneous growth with C60 and for SiC thermal decomposition growth, in
wafer-scale epitaxial layers. It is highly reproducible with atomic situ SiC cleaning was performed by exposure to a Si flux from
layer control of thickness and composition. These properties a high temperature elemental source[13] at a substrate tempera-
make MBE an attractive growth method for graphene films. There ture of 850 °C.[14] After Si flux cleaning the shutter of the Si
have been only a few experimental attempts to grow the graphene effusion cell was closed and the temperature was ramped up
by the evaporation of carbon.[10–12] Logninova et al.[10] focused to desired growth temperature. The shutter of the C60 effusion
cell was then opened for the desired growth time after which
the temperature was ramped down to room temperature at a
[∗] Dr. J. H. Park, Dr. W. C. Mitchel, L. Grazulis, Dr. H. E. Smith, fixed rate. For comparison, graphene was also grown by SiC
Dr. K. G. Eyink, Dr. J. J. Boeckl, Dr. D. H. Tomich, thermal decomposition. As a second carbon source, the same
S. D. Pacley, J. E. Hoelscher pyrolytic graphite filament was used simultaneously as both
Air Force Research Laboratory the carbon source and the substrate heater by inverting the
Materials and Manufacturing Directorate sample in the open back substrate holder. Graphite filaments
AFRL/RXPS, Wright – Patterson Air Force Base, Ohio, 45433 (USA)
E-mail: Jeongho.Park@WPAFB.AF.MIL have been widely used as carbon sources for p-type doping
of GaAs.[15,16] Mass spectroscopy studies of the vapor from
DOI: 10.1002/adma.201000756 heated filaments confirmed that the vapor consists mainly of

4140 wileyonlinelibrary.com © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2010, 22, 4140–4145
www.advmat.de
www.MaterialsViews.com

C3 molecules.[17,18] The experiments with the heated graphite shown here). But, as seen in Figure 1b, a slight improvement of

COMMUNICATION
filament used a different procedure from thermal decompo- the surface morphology was achieved after C60 CMBE growth
sition method and C60 CMBE growth in that the carbon flux (BEP: 5 × 10−8 Torr). Although the surface is still covered with
was controlled by changing the current through the filament pits, their depth is reduced to less than 5 Å. When the surface
and the Si flux cleaning was not employed since the orienta- was exposed to a higher carbon flux (BEP: 8.5 × 10−8 Torr) the
tion of the substrate was inverted with respect to the effusion surface morphology improved even more as seen in Figure 1c,
cells to allow exposure to the carbon flux from the heater. Also, where atomically flat, pit and island-free graphene terraces are
RHEED could not be used to determine the surface recon- visible. Wrinkles are found in this sample along and across
struction. This does limit the control parameters but further the step edges (inset Figure 1c). This indicates that the graphene
development could include a separate graphite filament in a film is continuous over the SiC steps. The step height at
normal configuration as a source. The formation of graphene the area having broken wrinkles is about 3.3 ± 0.1 Å, which is
was evidenced by a 2D peak in Raman spectra and the appro- the distance between graphene layers in few layer graphene.
priate C1s peak position in X-ray photoelectron emission The graphene grown by CMBE with the graphite filament
measurements. source at 1200 °C for 60 min. is shown in Figure 1d. A filament
The surface morphologies of the films grown by thermal current of ∼53 A was required to obtain this substrate tempera-
decomposition and C60 CMBE as measured by atomic force ture. The insert shows the surface morphology of graphene
microscopy (AFM) are shown in Figure 1a–c. The substrate grown by thermal annealing at 1200 °C for 30 min. Compared
temperature was kept at 1400 °C for 30 min.. The sample with the graphene grown by graphite filament CMBE, the sur-
grown by thermal decomposition shown in Figure 1a under- face of the graphene grown by the thermal decomposition has a
goes considerable modification of the surface morphology, such higher density of pits and is significantly rougher. Interestingly,
as significant roughening, the formation of many deep pits a higher density of wrinkles was formed on the surface. The
(≥ 1 nm), and the growth of isolated graphene islands (not wrinkles are uniformly distributed and are connected to each
other to form a hexagonally shaped “wrinkle net” mesh over the
surface. This indicates that the graphene layer
is continuous over the surface.
X-ray photoelectron spectroscopy (XPS)
was used to study the chemical composition
of the layers. The XPS C1s core level spec-
trum of the graphene grown by C60 deposi-
tion (8.5 × 10−8 Torr) is shown in Figure 2a.
This spectrum contains component peaks
from graphene, SiC, interface layers, and an
addition peak located at lower binding energy
region, which is possibly associated with two
coordinated carbon atoms in the SiC sub-
strate (see Supporting Information for more
details). The graphene C1s peak is shifted
∼0.9 eV to higher binding energy than the SiC
peak. These results are in good agreement
with graphene grown by vacuum annealing[19]
and in an Ar atmosphere.[3] The FWHM of
C1s component peak of SiC, which is ∼0.8 eV,
is comparable to that (∼0.75–0.8 eV) of other
groups’ reports.[3,19] The interfacial peak, how-
ever, experiences broadening compared to
other reports[3,19] (See detail in the Supporting
information). The broadening of interfacial
peak might suggest that it contains more
than one component. The carbon atoms in
the interface layer might experience various
local environments due to the disorder prop-
erties of this layer which could result in the
multiple components in the C1s spectra of
Figure 1. AFM images of surface morphology for various growth methods. a) Graphene this layer. Emtsev et al.[19] claimed that the
layer grown on SiC (0001) by conventional vacuum annealing in UHV at 1400 °C for 30 min. broadening of interfacial related peaks are “a
b) Graphene grown on SiC (0001) by carbon MBE (C60 BEP : 5 × 10−8 Torr) at 1400 °C for
consequences of strain in the carbon layers
30 min. c) Graphene with a higher C60 flux (BEP: 8.5 × 10−8 Torr). Insert: AFM image of 5 μm ×
5 μm scan area. d) Graphene grown by carbon MBE with the graphite filament source (Average that causes inhomogeneous broadening.” The
Input current: ∼53 A) at 1200 °C for 60 min. Insert: the AFM image of graphene by vacuum FWHM in their report ranges from ∼0.85 to
annealing at 1200 °C for 30 min. ∼0.9 eV. The FWHM of interfacial C1s peak of

Adv. Mater. 2010, 22, 4140–4145 © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 4141
www.advmat.de
www.MaterialsViews.com
COMMUNICATION

Figure 2. The properties of graphene grown with C60 (BEP : 8.5 × 10−8Torr) at 1400 °C during 30 min. a) XPS C1s core level spectrum of carbon MBE
grown graphene, showing contributions from the SiC substrate, the graphene layer (shifted by 1 eV from the SiC peak), and an interfacial layers.
b) Raman spectrum (λlaser = 514 nm) of epitaxial graphene grown with C60 (BEP : 8.5 × 10−8 Torr). c) 2D peak of graphene grown by thermal decomposi-
tion. d) 2D peak of graphene grown by carbon MBE with C60 (BEP : 8.5 × 10−8 Torr). e) Raman mapping of 2D peaks over a 1 mm × 1 mm area (λlaser =
514 nm, ∼1 μm spot size, 30 μm spatial resolution.

4142 wileyonlinelibrary.com © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2010, 22, 4140–4145
www.advmat.de
www.MaterialsViews.com

our graphene varies from ∼1 to ∼2 eV, which is much broader.

COMMUNICATION
in graphene on SiC (0001) shifts to higher wave number with
This broadening might be due a combination of the strain increasing thickness whereas for graphene on SiC(0001) the
effect and the various local environments of carbon atoms in 2D peak shifts to lower wave number with thickness. There-
the interface layer. fore, the mapping of the 2D peak position is a convenient
As shown in Figure 2b, the Raman spectrum shows the typ- method to estimate the thickness uniformity of the film.
ical features of the graphene grown by thermal decomposition, Figure 2e shows the peak position versus position for a 1 mm ×
the 2D peak at 2750 cm−1 and the G peak at 1584 cm−1. The 1 mm area of CMBE graphene. As can be seen, most of the
quality of the film is indicated by a minimal defect related D material has a peak at 2750 cm−1, indicating that the uni-
peak at about 1325 cm−1. The other features are due to the SiC formity of the graphene layer over this area is good. The large
substrate. Figure 2c and 2d show Raman 2D peaks from UHV up-shift (50 cm−1) of the 2D peak compared to reported few
graphitized and C60 CMBE grown graphene, respectively. The layer graphene (FLG) grown by thermal decomposition might
number of components of the 2D spectrum is a function of the be explained by a larger compressive strain in the MBE grown
graphene thickness and the full width at half maximum (FWHM) material.[23]
of each component should be the same as that of monolayer The rather unorthodox use of the graphite heater as both
graphene in the case of AB or Bernal stacking.[20] Therefore, carbon source and substrate heater was done to help confirm
the FWHM of the individual components of the 2D peak is that the graphene grown by CMBE was due to carbon depo-
an indicator of the overall quality of the graphene film. Curve sition rather than SiC decomposition. For the experiment
fitting of the spectrum (solid color lines in Figure 2c) shows reported here the filament current was kept at about 53 A,
that the 2D peak of the graphene grown by thermal decompo- which resulted in a growth temperature of 1200 °C, below the
sition contains four Lorentzian components with FWHM of common growth temperatures for SiC decomposition growth.
36 cm−1, indicating bilayer graphene.[20] CMBE grown graphene The sample was exposed to the heater for 60 min. To verify
shows that the 2D spectrum is much broader than that of the that the graphite heater can be used as the carbon source for
graphene grown by thermal decomposition. The fitting (solid graphene formation, the graphene growth on metal substrate
color lines in Figure 2d) shows that it has six components, indi- (4000 Å Ta) was attempted. First, Ta side was placed opposite
cating the growth of trilayer graphene. The FWHM of the com- the filament. None of the graphene related Raman features
ponents are about 31 cm−1, which indicates much improved were present, indicating absence of carbon on the surface.
quality of the graphene compared to the thermally decomposed Graphene films, as demonstrated by the presence of a strong
material. Like the thermal decomposition, the 2D peak of the Raman 2D band, were found on the Ta surface facing to the
CMBE grown graphene does not show the shoulder and asym- graphite heater after 30 min at 1200 °C. Therefore, we can con-
metry observed for exfoliated graphene.[21] clude that the graphene results from deposition of carbon from
With higher carbon flux, we are able to obtain thicker the graphite filament
graphene. The graphene layer thickness in monolayers (ML) Figure 3a shows the XPS C1s spectrum for graphene grown
was estimated by using the attenuation model for carbon by heated graphite filament. The curve fitting results are sim-
photoemission lines from SiC and graphene.[22] The nominal ilar to those for C60 CMBE shown above and for UHV- and Ar
thickness varies from 1.8 ML to 3.7 ML for C60 fluxes in the atmosphere grown graphene from the literature.[3,19] The C1s
range between 5 × 10−8 and 3.9 × 10−7 Torr. Interestingly, the peak consists of a SiC peak from the substrate, a peak due to
graphene to SiC XPS C1s peak ratio (C1sG/C1sSiC) saturated at sp2 bonded graphene, two interface peaks, and an additional
a flux of 1.2 × 10−7 Torr which corresponds to about 3.5 ML. peak at lower binding energy region (See detail in the Sup-
The Raman 2D peaks of the graphene grown with larger carbon porting Information). The graphene peak is shifted by 1 eV
flux (1.2 × 10−7 Torr) consist of only three components, indi- from the SiC peak. Interestingly, the C1s position (∼283.8 eV) of
cating four layer graphene[20] as well. The observed multi-peak SiC from graphene grown by graphite filament has a peak shift
structure of 2D peak with increasing layer number in C60 ∼0.4eV lower than that from graphene grown by C60. In addi-
CMBE graphene is similar to that of Bernal stacked exfoliated tion, Si 2p3 measurement of SiC substrate showed that there
graphene.[20] Therefore, it is likely that the stacking of our C60 is peak shift between the two different carbon sources. Si 2p3
CMBE graphene is also Bernal. The FWHMs of the individual from graphene grown by the graphite filament has a peak shift
components of the 2D peak as a function of carbon flux show ∼0.5 eV lower than that from graphene grown by C60. Seyller
that lower FWHMs (31 cm−1) were obtained at a flux of 8.5 × et al.[22] reported that the change of the surface reconstruction
10−8 Torr, which gives flat, pit free graphene over the surface from 3 × 3 to 6√3 × 6√3 on the SiC surface causes the peak shift
(Figure 1c). As the carbon flux is increased, the FWHM experi- of the SiC bulk components. They claimed the peak shift of the
ences significant broadening to 46 cm−1 and then saturates at SiC bulk components results from the change in surface band
higher flux. The analysis of the XPS C1s and the Raman spectra bending. Our graphene shows the different doping proper-
suggests that C60 CMBE graphene on SiC (0001) has a similar ties. The C60 and graphite filament growth result in n-type and
electronic structure and phonon properties to graphene grown p-type, respectively. Therefore, the change of the surface state
by thermal decomposition of SiC (0001). from n-type to p-type might lead to the change of Fermi level
To evaluate the thickness uniformity of the films, we per- pinning on these surfaces. These might cause the observed
formed Raman mapping as shown in Figure 2e. The meas- peak shift of SiC bulk components on our graphene samples.
ured 2D peak position depends strongly on the graphene The Raman spectrum in Figure 3b shows the presence
thickness and can indicate variations in thickness under of the 2D peak at 2676 cm−1 and the G peak at 1584 cm−1,
proper conditions. It has been shown that the 2D peak energy which is also in good agreement with results for the graphene

Adv. Mater. 2010, 22, 4140–4145 © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 4143
www.advmat.de
www.MaterialsViews.com
COMMUNICATION

Figure 3. XSP and Raman spectroscopy of graphene grown by carbon MBE from a heated graphite filament. a) C1s core level spectrum of graphene
grown at 1200 °C for 60 min. with the graphite heater source. The spectrum consists of four components due to the SiC substrate, the graphene, and
two interfacial layers. b) Raman spectrum (λlaser = 632.8 nm, ∼1 μm spot size) for this sample. c) Raman 2D peak shift as a function of graphene thick-
ness. d) Raman map (λlaser = 514 nm, ∼1 μm spot size, 30 μm spatial resolution on χ and y axes) of graphene grown at 1500 °C for 3 min (∼3.1 ML
nominal thickness). The bulk of the wafer has a 2D peak position at 2710 cm−1 with small islands at either 2705 or 2715 cm−1, indicating the epitaxial
growth of uniform thickness graphene over the large-area surface. Note that the apparent peak shift between Figure 3C and 3D is due to the different
excitation wavelength.

by thermal decomposition and C60 grown graphene. Even 3.1 ML. Interestingly, the 2D peaks for all graphene samples
though this growth is on SiC (0001), the 2D peak position grown by the graphite filament technique are composed of a
is very close to that of thermally decomposed material on single Lorentzian component. The single Lorentzian fits sug-
SiC (0001.[24] Interestingly, the defect-related D peak was gest that these samples have non-Bernal stacking, indicating
strongly suppressed in this sample, implying a high-quality that the graphene layers are electronically decoupled from
film. As seen in the inset of Figure 3b, the 2D peak consists each other. The variation of peak position with thickness and
of a single Lorentzian curve with a FWHM of about 39 cm−1, a single Lorentzian 2D line are very similar to the behavior of
characteristic of monolayer graphene. The graphene thick- the graphene from thermal decomposition on SiC (0001).[24]
ness in these experiments was controlled by the heater and This clearly indicates that the graphene grown here with a
exposure time up to a nominal thickness of 3.1 ML. Figure 3c graphite filament source has similar electronic structure to
shows the variation of the 2D peak position with different the graphene on carbon face SiC. The single Lorentzian fit
graphene thicknesses. The peak position undergoes a shift to also indicates that our graphene has Dirac-like electronic
lower wave number (about 11 cm−1) as the graphene thick- states.[24] To estimate the thickness uniformity of these sam-
ness increases from about 0.8 ML to a nominal thickness of ples we mapped the 2D peak position of graphene grown for

4144 wileyonlinelibrary.com © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Mater. 2010, 22, 4140–4145
www.advmat.de
www.MaterialsViews.com

3 min at 1500 °C with a 514 nm laser for a 1 mm × 1 mm Acknowledgements

COMMUNICATION
area as for the C60 grown material. As seen in Figure 3d, most
The authors wish to thank S. N. Kim for Raman measurement support.
of the area has a peak position of about 2710 cm−1 indicating This work was supported by the Air Force Office of Scientific Research
homogeneous growth over a large area. (Dr. Donald Silversmith and Dr. Harold Weinstock).
To investigate if the structure and graphene properties are
related to the type of carbon sources, the effect of the initial Received: March 1, 2010
substrate condition was explored. First, the same initial SiC Published online: August 20, 2010
substrate was prepared with Si flux cleaning and took out from
the chamber. The orientation of the substrate was inverted to
allow exposure to the carbon flux from the heater. After that, [1] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang,
the sample was immediately reloaded into growth chamber. S. V. Dubonos, I. V. Grigorieva, A. A. Firsov, Science 2004, 306, 666.
The graphene growth was followed at 1200C for 70 min. [2] C. Berger, Z. Song, X. Li, X. Wu, N. Brown, C. Naud, D. Mayou, Ti,
XPS data showed that the graphene peak was shifted to Li, J. Hass, A. N. Marchenkov, E. H. Conrad, P. N. First, W. A. De
∼1 e V higher energy compare to SiC bulk component. The peak Heer, Science 2006, 312, 1191.
[3] K. V. Emtsev, A. Bostwick, K. Horn, J. Jobst, G. L. Kellogg, L. Ley,
position of graphene was also the same as that of graphene
J. L. McChesney, T. Ohta, S. A. Reshanov, J. Röhrl, E. Rotenberg,
on untreated SiC. Thickness of graphene was ∼0.85 ML. A. K. Schmid, D. Waldmann, H. B. Weber, Th. Seyller, Nat. Mater.
The thickness slightly increases due to the increase the depo- 2009, 8, 203.
sition time. The 2D peak was confirmed by the Raman meas- [4] X. Li, G. Zhang, X. Bai, X. Sun, X. Wang, E. Wang, H. DaiHighly,
urement. The position of 2D peak was ∼2686 cm−1, which Nat. Nanotechnol 2008, 3, 538.
is almost identical position to that of graphene grown on [5] K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim, J. H. Ahn,
untreated SiC by graphite filament. The surface morphology P. Kim, Y. Y. Choi, B. H. Hong, Nature 2009, 457, 706.
of graphene grown on between Si flux treated and untreated [6] P. W. Sutter, J. I. Flege, E. I. Sutter, Nat. Mater. 2008, 7, 406.
SiC has almost identical feature including hexagonal shaped [7] G. G. Eda, G. Fanchini, M. Chhowalla, Nat. Nanotechnol. 2008, 3, 270.
“winkle net”. These presumably suggest that the properties [8] S. K. Banerjee, L. F. Register, E. Tutuc, D. Reddy, A. H. MacDonald,
IEEE Electron Device Lett. 2009, 30, 158.
difference is not related to initial surface condition but the
[9] Y. Zhang, T. T. Tang, C. Girit, Z. Hao, M. C. Martin, A. Zettl,
type of carbon sources. M. F. Crommie, Y. R. Shen, F. Wang, Nature 2009, 459, 820.
In conclusion, we successfully demonstrated that the direct [10] L. Loginova, N. C. Bartelt, P. J. Feibelman, K. F. MaCarty, New J. Phys.
deposition of carbon from solid sources, C60 and a graphite 2008, 10, 093026.
filament, is a promising approach for graphene growth and [11] A. Al-Temimy, C. Riedl, U. Strake, Appl. Phys. Lett. 2009, 95, 231907.
produces superior quality, homogeneous graphene layers in [12] J. Hackley, D. Ali, J. DiPasquale, J. D. Demaree, C. J. K. Richardson,
comparison with the thermal annealing method. The direct Appl. Phys. Lett. 2009, 95, 133114.
deposition of carbon yields atomically flat and pit-free graphene [13] W. V. Lampert, C. J. Eiting, S. A. Smith, K. Mahalingham,
terraces. Wrinkles along and across the step edges and map- L. Grazulis, T. W. Haas, J. Crystal Growth 2002, 234, 369.
ping of the 2D peak position indicated homogeneous graphene [14] L. Magaud, F. Hiebel, F. Varchon, P. Mallet, H. -Y. Veuillen,
Phys. Status Solidi RRL 2009,3, 172.
growth over the substrate. In addition, the direct deposition
[15] R. J. Malik, J. Nagle, M. Micovic, R. W. Ryan, T. Harris, M. Geva,
of graphene gave significantly improved controllability of L. M. Hopkins, J. Vandenberg, R. Hull, R. F. Kopf, Y. Anand,
the graphene growth. The thickness and electronic structure W. D. Braddock, J. Crystal Growth 1993, 127, 686.
related to the stacking order could be controlled by C60 flux and [16] W. E. Hoke, P. J. Lemonias, P. S. Lyman, H. T. Hendriks, D. Weir,
the type of carbon source, respectively. C60 CMBE graphene has P. Colombo, J. Crystal Growth 1993, 111, 269.
electronic and structural characteristics similar to graphene [17] M. Hoch, P. E. Blackburn, D. P. Dingledy, H. L. Johnston, J. Chem.
(Bernal Stacking) grown on SiC (0001) by decomposition but Phys. 1954, 22, 126.
the graphite filament CMBE graphene resembled decomposi- [18] M. Joseph, N. Sivakumar, P. Manoravi, Carbon 2002, 40, 2031.
tion grown graphene (Non-Bernal Stacking) on SiC (0001). This [19] K. V. Emtsev, F. Speck, Th. Seyller, L. Ley, Phys. Rev. B 2008, 77,
indicates that these differences of electronic structure associ- 155303.
[20] L. M. Malard, M. A. Pimenta, G. Dresselhaus, M. S. Dresselhaus,
ated with stacking order and phonon properties of graphene
Phys. Reports 2009, 473, 51.
might be due to different graphene growth mechanisms asso- [21] D. S. Lee, C. Riedl, B. Krauss, K. V. Klitzing, U. Starke, J. H. Smet,
ciated with different the solid carbon sources we used in the Nano lett. 2008, 8, 4320.
experiment. [22] Th. Seyller, K. V. Emtsev, K. Gao, F. Speck, L. Ley, A. Tadich,
L. Broekman, J. D. Riley, R. C. C. Leckey, O. Rader, A. Varykhalov,
A. M. Shikin, Surf. Sci. 2006, 600, 3906.
[23] J. Röhrl, J. M. Hundhausen, K. V. Emtsev, Th. Seyller, R. Graupner, L.
Supporting Information Ley, Appl. Phys. Lett. 2008, 92, 201918.
Supporting Information is available from the Wiley Online Library or [24] C. Faugeras, A. Nerrière, M. Potemski, A. Mahmood, E. Dujardin,
from the author. C. Berger, W. A. De Heer, Appl. Phys. Lett. 2008, 92, 011914.

Adv. Mater. 2010, 22, 4140–4145 © 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 4145

You might also like