Graphene growth by molecular beam epitaxy using a solid carbon source

Article abstract of DOI:10.1002/pssa.200982412

The direct elaboration of graphene by molecular beam epitaxy (MBE) has been studied, using a solid carbon source. Successful growth has been achieved on both the carbon and silicon terminated faces of silicon carbide substrates in the temperature range 1000-1100 °C, as confirmed by low energy electron diffraction (LEED) and X-ray photoemission spectroscopy (XPS) analysis. Atomic force microscopy (AFM) observations show that the initial substrate structure, i.e. flat atomic terraces and half-period high steps, remains almost unaffected during the growth, contrary to what is observed following the graphitization process.

Full text of DOI:10.1002/pssa.200982412

Phys. Status Solidi A 207, No. 2, 300–303 (2010) / DOI 10.1002/pssa.200982412
a
p s s
www.pss-a.com
applications and materials science
Graphene growth by molecular beam
epitaxy using a solid carbon source
*
E. Moreau , F. J. Ferrer, D. Vignaud, S. Godey, and X. Wallart
IEMN, UMR CNRS 8520, Avenue Poincar e´ PO Box 60069, 59652 Villeneuve d’Ascq Cedex, France
Received 26 June 2009, revised 16 September 2009, accepted 6 October 2009
Published online 15 December 2009
PACS 61.05.Jh, 68.37.Ps, 81.05.Uw, 81.15.Hi, 82.80.Pv
*
Corresponding author: e-mail eleonore.moreau@ed.univ-lille1.fr, Phone: þ33 320 19 78 65, Fax: þ33 320 19 78 92
The direct elaboration of graphene by molecular beam epitaxy scopy (XPS) analysis. Atomic force microscopy (AFM)
(
MBE) has been studied, using a solid carbon source. observations show that the initial substrate structure, i.e. flat
Successful growth has been achieved on both the carbon and atomic terraces and half-period high steps, remains almost
silicon terminated faces of silicon carbide substrates in the unaffected during the growth, contrary to what is observed
temperature range 1000–1100 8C, as confirmed by low energy following the graphitization process.
electron diffraction (LEED) and X-ray photoemission spectro-
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1
Introduction Since the pioneering work of diffraction (RHEED). The solid carbon cell was purchased
Novoselov et al. [1], the electronic properties of graphene from MBE-Komponenten. Carbon is sublimated from a
have encouraged a lot of fundamental and device-related heated graphite filament. The carbon flux has been evaluated
1
3
ꢁ2 ꢁ1
studies. Graphene has been mainly obtained by three ways. to be 2 ꢀ 10 atoms cm
s , by measuring the total
The first method consists in exfoliating graphene layers from thickness of the carbon layer deposited after a two hours run
bulk graphite and transferring it onto oxidized silicon [1]. on a Si substrate. The temperatures reported in this study
The second one, the so-called graphitization process, were measured using an optical pyrometer, its emissivity
involves heating bulk silicon carbide to induce Si prefer- being set to 0.5 (one face polished Si face substrates) or 0.35
ential sublimation, resulting in epitaxial graphene growth on (two face polished C face). The estimated error in the
the surface [2]. The last one is chemical vapour deposition temperature measurement is ꢂ10 8C. The initial surface
and relies on growth from hydrocarbon gases on metallic preparation included a first degassing step in the 800–900 8C
substrates (e.g. Ni) [3]. In the present work, our first results of temperature range, followed by a long smoothening under Si
direct growth ofgraphene by molecular beam epitaxy (MBE) flux exposure at 1050–1100 8C. This last process was shown
using a solid carbon source are reported. Silicon carbide to lead to a step-terrace structure with half-period step height
substrates have been used to reduce lattice mismatch [4], illustrated by the atomic force microscopy (AFM)
problems.
Figs. 1(d) and 2(a).
The sample temperature was then lowered to avoid any
2
Experimental conditions 6H-SiC n-type wafers contribution of the direct graphitization to the graphene
purchased from Cree or SiCrystal, with a nominal cut-off growth. Following this step, no particular surface recon-
angle <0.58 were used in this study. The as-received struction could be observed with the C face samples
substrates were polished by NovaSiC to remove scratches [Fig. 1(a)], whereas the Si face ones always exhibited the
and most other surface defects. A thin tungsten film was Si-rich (H3 ꢀ H3)R308 periodicity [Fig. 2(b)]. The gra-
deposited onto the rear face to allow UHV non-contact phene layers were finally grown on both the silicon (0001)
2
radiation heating. The wafers were then diced into 4 ꢀ 4 mm
and carbon (000–1) SiC faces.
The surface structure was investigated ex situ using low
substrates and cleaned with acetone and isopropyl alcohol.
The samples were then loaded into an ultra-high vacuum energy electron diffraction (LEED), X-ray photoemission
MBE chamber fitted with reflected high-energy electron spectroscopy (XPS) and AFM. The XPS experiments were
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Original
Paper
Phys. Status Solidi A 207, No. 2 (2010)
301
Figure 3 (online colour at: www.pss-a.com) LEED diagram at
2
160 eV (a), 2 ꢀ 2 mm wide AFM image (b), for 2.1 ML graphene
epitaxially grown on SiC (000–1).
Figure 1 (online colour at: www.pss-a.com) RHEED patterns
obtained on the SiC (000–1) surface along the (1–100) azimuth at
1
030 8C, initial (1 ꢀ 1) (a), (3 ꢀ 3) reconstruction after 20 s growth
2
(b) and graphene after 4 min (white arrow) (c), 10 ꢀ 10 mm AFM
picture showing flat atomic terraces with 0.75 nm high steps before reconstruction which is stable at lower temperature only
growth (d).
[6]. Although no atomic model for this C-rich (3 ꢀ 3) seems
yet accepted [6, 7], this result, associated with the measured
C flux, suggests that 3–4 carbon atoms are required for each
carried out using Al Ka (hy ¼ 1486.6 eV) radiation from a (3 ꢀ 3) surface unit cell. The (3 ꢀ 3) pattern soon disap-
monochromatic X-ray source and a hemispherical energy peared after a few more seconds of carbon exposure.
analyzer. Decomposition of the XPS spectra was performed
After 4 min of growth at 1030 8C, the RHEED graphene
using Gaussian–Lorentzian mixed components for sym- pattern is detected [Fig. 1(c)]. This is confirmed by the ex situ
metrical lines, adding some exponential tail parameter for LEED pattern [see Fig. 3(a)], which shows that the
asymmetrical lines. Angle resolved XPS (ARPXS) was used graphene lattice is rotated by 308 relative to SiC. Thicker
to calculate the graphene thickness. Surface topographic graphene layer present partial diffraction rings, indicating
characterization was achieved by AFM in tapping mode.
the presence of rotational disorder as it has already been
found after graphitization [5].
The corresponding C1s XPS spectrum is presented in
3
Results Graphene growth at temperature lower than Fig. 4. It is very similar to the spectra obtained after
000 8C soon ended with amorphous material, characterized graphitization [8], with a C–Si bulk component at the binding
1
by the vanishing of all diffraction spots. Thus, the growth energy of 282.8 eV, a graphene C–C component at 284.5 eV
temperature was restricted within the range 1000–1100 8C, and a weak peak at 291.1 eV which is identified as a shake-up
ꢄ
to avoid both amorphous deposit and purely thermal satellite of the graphene peak due to p–p transitions (see
graphitization.
Fig. 4, inset). In order to estimate the thickness of the
3
.1 Graphene growth on C face SiC After ꢃ20 s
of exposure to carbon flux at 1030 8C, the (3 ꢀ 3)
reconstruction appears [Fig. 1(b)]. The extra spots from the
1
/3 fractional order Laue zone are clearly observed, which
shows that most of the surface is reconstructed. This
reconstruction appears to be a precursor step towards
graphene growth, since it is also observed at the beginning
of the graphitization process [5, 6], at temperatures close to
1
100 8C in our experiments. This C-rich (3 ꢀ 3) reconstruc-
tion should not be mistaken for the Si-rich (3 ꢀ 3)
2
Figure 2 (online colour at: www.pss-a.com) 10 ꢀ 10 mm AFM
picture of a 6H-SiC (0001) surface after the Si preparation step (a).
Flat atomics steps are 0.75 nm high, i.e. half the 6H polytype period. Figure 4 (online colour at: www.pss-a.com) XPS C1s spectra of
In situ RHEED patterns along the [1–100] azimuth of the initial SiC C face at 458 (photoemission angle with respect to the sample
(
(
H3 ꢀ H3)R308reconstruction(b)andofthefinalSiCandgraphene surface) after 2.1 ML graphene growth at 1030 8C. Inset shows the
ꢄ
white arrow), after 3 min growth at 1050 8C (c).
shake-up satellite peak due to p–p transitions.
www.pss-a.com
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

a
p s s
3
02
E. Moreau et al.: Graphene growth by molecular beam epitaxy
graphitic layer from XPS data, we used the approach
proposed by Cumpson [9]. It is based on the attenuation of
the bulk component through the assumed uniform graphene
layer. One finds a coverage of 0.69 ꢂ 0.07 nm or 2.1
monolayers (ML) of graphene.
The topographic structure of this 2.1 ML sample is
shown on the AFM picture [Fig. 3(b)]. The original substrate
structure is almost preserved, with large atomically flat
terraces approximately 900 nm wide. The mean height
between these main terraces is 0.75 nm, that is half the SiC
substrate unit cell, which directly reflects the initial substrate
surface. These terraces are separated by narrower ones, the
average step height between them being ꢃ0.25 nm. This
value corresponds to the characteristic SiC bilayer stacking
along the <0001> direction, but it is not very far either from
the height of an extra partial graphene layer (ꢃ0.33 nm). It is
thus not yet clear if this particular topographic structure
comes from the epitaxial growth process or from an
incomplete surface preparation prior to the growth itself.
Further work is required to clarify this point.
Figure 6 (online colour at: www.pss-a.com) XPS C1s spectra at
4
1
58 (photoemission angle with respect to the sample surface) of
.5 ML graphene grown on Si face SiC.
3
.2 Graphene growth on Si face SiC Graphene
growth was studied starting from the (H3 ꢀ H3)R308
surface reconstruction on Si face SiC [Fig. 2(b)]. After C–C bonds at the interface between the SiC substrate and
min growth at 1050 8C, the graphene RHEED pattern the graphene layer. Using one component (see Fig. 6) or two
appeared [Fig. 2(c)] and its presence was confirmed by ex situ components [8] for the interface contribution does not affect
LEED and XPS measurements.
the calculated graphene thickness as the difference is lower
3
Once again, the graphene LEED spots presented in than 1%. The energy shift between Si–C bulk bonds and the
Fig. 5(a) show that its lattice is rotated by 308 with respect to graphene or interface components are 1.0 and 1.6 eV,
the SiC one, as was already observed after graphitization respectively. In the case of graphitization, this interface
[
10]. In the same time, flat atomic steps of half-period height layer is usually associated with the (6H3 ꢀ 6H3)R308
remain unchanged after the growth [Fig. 5(b)]. Thanks to the surface reconstruction which precedes graphene growth
temperature lower than the one required for graphitization [8]. Although the (6H3 ꢀ 6H3)R308 pattern could not be
(
ꢃ1100 8C for the Si face), the strong surface reorganization clearly observed during MBE growth, it is important to point
usually associated with this process is not efficient during the out that graphene on SiC requires an interface layer when
MBE growth of graphene.
grown on the Si face but not on the C face, whatever the
The ex situ C1s XPS spectra presented in Fig. 6 shows elaboration method is. From the attenuation of the bulk C1s
three components, a main peak at 284.7 eV indicating the component through the interface and graphene layers, a total
presence of the graphene C–C bonds, a second peak at thickness of 0.49 ꢂ 0.05 nm or 1.5 ML is found after 3 min
2
83.7 eV linked to the substrate C–Si bonds and a third peak growth.
at 285.3 eV. By analogy with the corresponding and almost
identical C1s XPS spectra measured after graphitization [8],
it is suggested that this component is associated with
4
Conclusions Graphene layers have been success-
fully grown on both faces of SiC substrates by MBE using a
solid carbon source, as evidenced by electron diffraction and
XPS analysis. Noticeably, it has been shown that the C-rich
(
3 ꢀ 3) surface reconstruction could be generated after a
short C exposure. Furthermore, AFM topographical studies
have shown that the strong surface reorganization associated
with the graphitization could be avoided by MBE growth of
graphene.
Figure 5 (online colour at: www.pss-a.com) LEED diagram at
98 eV (a), 2 ꢀ 2 mm wide AFM image (b), for 1.5 ML graphene (project Xp-Graphene) and the Region Nord-Pas-de-Calais are
epitaxially grown on SiC (0001).
Acknowledgements Financial support from the ANR
2
greatly acknowledged.
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.pss-a.com

Original
Paper
Phys. Status Solidi A 207, No. 2 (2010)
303
References
[5] I. Forbeaux, J.-M. Themlin, and J.-M. Debever, Surf. Sci.
42, 9 (1999).
4
[
[
1] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang,
Y. Zhang, S. V. Dubonos, I. V. Griogrieva, and A. A. Firsov,
Science 306, 666 (2004).
[
[
[
[
6] K. Heinz, J. Bernhardt, J. Schardt, and U. Starke, J. Phys.:
Condens. Matter 16, S1705 (2004).
7] F. Takeuchi, R. Fukuyama, Y. Hoshino, T. Nishimura, and
Y. Kido, Surf. Sci. 601, 2203 (2007).
8] K. V. Emtsev, F. Speck, T. Seyller, L. Ley, and J. D. Riley,
Phys. Rev. B 77, 155303 (2008).
9] P. J. Cumpson, in: Surface Analysis by Auger and X-Ray
Photoelectron Spectroscopy, edited by D. Briggs and
J. T. Grant (IM Publications, and SurfaceSpectra Limited,
Chichester, UK, 2003), chap. 23.
2] C. Berger, Z. Song, T. Li, X. Li, A. Y. Ogbazghi, R. Feng,
Z. Dai, A. N. Marchenkov, E. H. Conrad, P. N. First,
and W. A. de Heer, J. Phys. Chem. B 108, 19912
(
3] K. S. Kim, Y. Zhao, H. Jang, S. Y. Lee, J. M. Kim, K. S. Kim,
2004).
[
[
J.-H. Ahn, P. Kim, J.-Y. Choi, and B. H. Hong, Nature 457,
706 (2009).
4] F. J. Ferrer, E. Moreau, D. Vignaud, S. Godey, and [10] I. Forbeaux, J.-M. Themlin, and J.-M. Debever, Phys. Rev. B
X. Wallart, Semicond. Sci. Technol. 24, 125014 (2009).
58, 16396 (1998).
www.pss-a.com
ß 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim