APPLIED PHYSICS LETTERS 93, 263102 共2008兲

Early stage formation of graphene on the C face of 6H-SiC N. Camara,1 G. Rius,1 J.-R. Huntzinger,2 A. Tiberj,2 L. Magaud,3 N. Mestres,4,a兲 P. Godignon,1 and J. Camassel2 1

CNM-IMB-CSIC—Campus UAB 08193 Bellaterra, Barcelona, Spain GES–UMR 5650, CNRS, and Université Montpellier 2, 34095 Montpellier cedex 5, France 3 Institut Néel, CNRS, and UJF, BP 166, 38042 Grenoble cedex 9, France 4 ICMAB-CSIC, Campus UAB 08193 Bellaterra, Barcelona, Spain 2

共Received 27 October 2008; accepted 4 December 2008; published online 29 December 2008兲 An investigation of the early stage formation of graphene on the C face of 6H-silicon carbide 共SiC兲 is presented. We show that the sublimation of few atomic layers of Si out of the SiC substrate is not homogeneous. In good agreement with the results of theoretical calculations it starts from defective sites, mainly dislocations that define nearly circular graphene layers, which have a pyramidal, volcanolike shape with a center chimney where the original defect was located. At higher temperatures, complete conversion occurs but, again, it is not homogeneous. Within the sample surface, the intensity of the Raman bands evidences inhomogeneous thickness. © 2008 American Institute of Physics. 关DOI: 10.1063/1.3056655兴 Graphene is a two-dimensional carbon system with relativisticlike electronic transport properties1 and the graphenebased devices should have the capability to breakthrough the silicon C-MOS leadership in the microelectronic industry. A widely used technique to fabricate mono or bilayers of graphene was developed 4 years ago by exfoliating highly oriented pyrolytic graphite 共HOPG兲 on an oxidized Si wafer. Unfortunately, the biggest flakes obtained in this way are only 10⫻ 10 ␮m2, which is not enough for industrial purposes. An alternative technique is the use of chemical vapor deposition deposition of carbon on a metal surface, such as 具0001典 Ru or 具111典 Ir.2,3 In this case the graphene transport properties are hidden by the metallic conductivity of the substrate and, for device application, complex transfer procedures are necessary. The last way of fabricating graphene is to sublimate few atomic layers of Si from a monocrystalline SiC substrate.4,5 This can be done from a complete SiC wafer or from a small prepatterned area,6 and constitutes by far the most promising technique to develop industrial applications.7 This is the one we present in this work. Whatever the technique, the main concerns for graphene electronics are the properties of the substrate to graphene interface, the crystalline structure with emphasis on the long range order along the c axis, the reproducibility, and finally, the homogeneity of layers. Two frequently used techniques to characterize the growth of few layers graphene 共FLG兲 by SiC sublimation are low electron energy diffraction and scanning tunneling microscopy. To gain more realistic information on a wider area, optical microscopy 共OM兲, atomic force microscopy 共AFM兲, scanning electron microscopy 共SEM兲, and Raman spectroscopy must be used. In this work we investigate the formation of FLG grown on the C face of a 6H-SiC substrate on which an atomically flat surface was prepared by Novasic.8 Sublimation was done in a radio frequency induction furnace at temperatures ranging from 1450 to 1550 ° C. The processing time was about 5 min. Then, the growth product was systematically investigated by OM, SEM, AFM, and Raman. a兲

Electronic mail: [email protected].

0003-6951/2008/93共26兲/263102/3/$23.00

We focus on three different types of samples obtained by different heating temperatures. They were 1 ⫻ 1 cm2 templates cut from an on-axis 6H-SiC wafer, n-type doped to ⬃5 ⫻ 1017cm−3. All chemical treatments used before sublimation were clean-room compatible and similar to the ones used before thermal oxidation. The vacuum limit was ⬃10−6 Torr. In order to remove any trace of native oxide, the temperature was raised to 1050 ° C for 10 min. Then, the three samples 共A, B, and C兲 were heated at 1450, 1500, and 1550 ° C, respectively, for 5 min. On the low temperature samples 共sample A兲 after heating at 1450 ° C for 5 min, we find that the only change is a large reconstruction of the initial surface. In the SiC literature, this surface reconstruction is best known as “step bunching” and originates from a small 共nonintentional兲 miscut of the nominally on-axis 6H-SiC wafer. It does not correspond to a single SiC bilayer 共BL兲 height and, in the case of sample A, resulted in large parallel terraces with ⬃1 nm height and 2 ␮m width. From OM, SEM, or AFM no evidence of graphene is found. For the Raman spectroscopy measurements, we used the 514 nm 共El = 2.14 eV兲 line of an Ar+ ion laser as exciting frequency and collected Raman spectra in the confocal backscattering configuration. The second order spectrum of the SiC substrate, with two main peaks at 1516 and 1714 cm−1 at room temperature,9 was used as internal reference. Despite intensive research, no graphene response from Raman investigations could be found. No evidence of Si aggregates before out diffusion was also found. This shows that working at 1450 ° C, under the pressure conditions used in this work, no efficient sublimation of Si atoms can occur. Upon heating, the surface only reorganizes to minimize energy. On sample B, the situation is entirely different. After heating at 1500 ° C for 5 min, the first graphene layers appeared. They were not distributed homogeneously but rather randomly, with nearly circular shape 关Fig. 1共a兲兴 and a common diameter 共⬃5 ␮m兲. The step-bunched SiC terraces already observed in sample A remained between the FLG areas. These layers were clearly evidenced by AFM 关Figs. 1共a兲 and 2共c兲–2共e兲兴, OM 关Fig. 2共a兲兴, and SEM 关Fig. 2共b兲兴. The density over a full sample was about 106 cm−2, which is the

93, 263102-1

© 2008 American Institute of Physics

Downloaded 06 Jan 2009 to 117.32.153.167. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

263102-2

Camara et al.

Appl. Phys. Lett. 93, 263102 共2008兲

FIG. 1. 共Color兲 Topographic AFM measurement in tapping mode of samples graphitized at 共a兲 1500 ° C, a typical 5 ␮m diameter flake of FLG 共confirmed by its Raman signature兲 is visible while outside the flake the SiC steps are visible and 共b兲 at 1550 ° C, the full wafer is covered by FLG 共as determined by Raman spectroscopy兲.

typical density of dislocations in a commercial, research grade, SiC wafer. This suggests that the dislocations act as catalyzing defects, of which sublimated Si escapes more easily. With longer annealing times, the in-plane shape of the flakes turns from 共nearly兲 circular to 共roughly兲 hexagonal, with some pyramidal profile and a depression at the periphery. If one considers that the growth starts from an extended defect, which acts like a “chimney” for a volcano, this is not so surprising. In this case, all constituting Si atoms in the topmost SiC BLs are progressively pumped by the central chimney, until some frontier is reached. This can be just the diffusion length 共but in this case the profile should remain flat at the periphery兲 or the boundary of the constitutive SiC grain. This is in better agreement with the final hexagonal shape and the depression at grain periphery, with no possible growth outside the periphery. This complex behavior was confirmed by Raman spectroscopy. Outside the flakes only the second-order Raman spectrum of the SiC substrate was detected while, inside the flakes we found the well-resolved graphite 共G and 2D兲 bands, which give absolute evidence of carbon sp2 reorganization.10–12

FIG. 2. 共Color兲 Summary of results collected on sample B. 共a兲 Wide range optical microscope view of the sample with the flakes clearly visible, 共b兲 SEM picture of the same sample, the darker flakes visible at SEM are the brighter seen at the optical microscope and correspond to a higher number of graphene layers. 共c兲–共e兲 are, respectively, the AFM measurement in tapping mode of the topography of the surface, the amplitude corresponding to contours, and the phase measurement.

FIG. 3. 共Color兲 共a兲 Wide range crossed polarized OM view of sample C, graphitized at 1550 ° C and 共b兲 the same area investigated in dark field mode. 共c兲 is the superposition of 共a兲 and 共b兲 highlighting the correspondence between the thicker areas and the dislocations acting as chimneys. 共d兲 Raman spectra collected on the same sample in three different parts: in the bright thin continuum 共bottom spectrum兲, in the dark sublimation chimney 共top spectrum兲, and in a gray intermediate part 共middle兲. From bottom to top, the number of FLG increases by, typically, a factor of 10.

These observations suggest that there is a large energy barrier that prevents the 共direct兲 diffusion of Si atoms through the topmost carbon layer. To cross check this statement, we performed ab initio calculations to evaluate the energy difference for a Si atom located 1.99 Å below a graphene layer and moving in the layer. We assumed two different geometries: 共i兲 a perfect 共infinite兲 honeycomb lattice and 共ii兲 a defective topmost layer with a Stone–Wales defect.13 Both calculations were done using the code VASP14 which is based on the density functional theory 共DFT兲. We used the generalized gradient approximation and ultrasoft pseudopotentials that have been extensively tested.15 Finally, a 3 ⫻ 3 ⫻ 3 k-point mesh was used so that, at convergence, the change of the total energy is below 0.001 eV. In the defect-free carbon layer, the lateral 共x , y兲 positions of the Si atom corresponded to the center of a honeycomb hexagon. With the Stone–Wales defect, it was fixed at the center of one of the defect heptagons. Both diffusion barriers calculated in this way were found very large. They are about 15.6 eV for the perfect layer and 9.7 eV in the presence of a Stone– Wales defect. The path through a defect is already lower but it is not enough to account for the strong difference shown in Figs. 1共a兲 and 1共b兲 for a 50 ° C temperature rise. This confirms that, in order to be evacuated, the direct jump of an

Downloaded 06 Jan 2009 to 117.32.153.167. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

263102-3

Appl. Phys. Lett. 93, 263102 共2008兲

Camara et al.

in-depth Si atom to the topmost graphene layer is hardly possible. In other words, the fast sublimation of Si atoms, as observed in Figs. 1共a兲 and 1共b兲, requires a completely different process with a large assisting defect. Finally, we investigated sample C graphitized at 1550 ° C. The temperature rise was low but, again, the situation appears completely different. The whole surface is now covered by FLGs with no bare SiC. This appears clearly visible from the AFM picture shown in Fig. 1共b兲 or the OM picture shown in Fig. 3共a兲. Simply, the graphene thickness is still not homogeneous. The dark areas in Fig. 3共a兲 in crossed polarization mode microscopy are the “volcanos” already identified in sample B. They are clearly associated with dislocations visible in dark field mode microscopy 关Fig. 3共b兲兴. The merged pictures of Figs. 3共a兲 and 3共b兲 revealed that the dislocations/chimneys are mainly responsible for the inhomogeneity in the graphene growth process 关Fig. 3共c兲兴. A clear continuum with less graphene and a rather smooth surface appears between the dark thicker FLG parts. This suggests a second 共different兲 mechanism which, according to the results of the DFT calculation, should not be intrinsic but associated with a second type of defects. From the AFM measurements we find that all domain sizes are similar 共in the range of few hundreds of nanometer兲 with, in some cases, incomplete coalescence 关dark parts in Fig. 1共b兲兴. To confirm these results we again used Raman spectroscopy. From the ratio of the D and G bands integrated intensities 共ID / IG兲 we deduced the domain size of FLG using the empirical relation of Cancado et al.16 as follows: La =

冉冊

560 nm eV4 ID IG E4l

−1

.

共1兲

Even in the worst case where the ratio ID / IG is ⬃1 / 50 关see Fig. 3共d兲兴 this gives crystallite size La in the range of 800 nm. This is in very reasonable agreement with the results of the AFM measurements. Apart from the intensity which can change by a factor of 10–20 when moving from the continuum to the top of a chimney, the main difference from spectrum to spectrum comes from the 2D band. Being stacking order sensitive,8,17 it should be used to follow the evolution from monolayer to bilayer or more complex turbostratic 共with only one broad peak兲 to, finally, three-dimensional HOPG 共with two main bands兲. We found mainly two types of spectra. The first one appears only in the continuum 共outside the flakes兲 giving evidence of thin and rather uniform FLG. The 2D band appears at 2710 cm−1 and the full width at half maximum is ⬃42 cm−1. The second series of spectra 共inside the darker areas兲 not only confirms the larger thickness, it also indicates

a different stacking. This may suggest that the speed of conversion plays a role in the three-dimensional organization of the layers. To summarize, in good agreement with the results of DFT calculations we have shown that the low pressure graphitization of 6H-SiC is not an intrinsic process. Within 50 ° C, on the C face of 6H-SiC, two different mechanisms manifest. The first one involves the dislocations which are inherent to the limited quality of actual SiC substrates. The second gives more homogeneous results and more work is in progress to identify the participating defects. To grow large, homogeneous FLG, the first process will have to be eliminated. We thank the French ANR and the European Community for partial support through the project. Financial support from the Spanish Government, Consolider NANOSELECT, CSD2007-00041 project, and 2006 Juan de la Cierva grant to N. C. is also acknowledged. 1

K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva, and A. A. Firsov, Science 306, 666 共2004兲. 2 P. W. Sutter, J. I. Flege, and E. A. Sutter, Nature Mater. 7, 406 共2008兲. 3 J. Coraux, A. T. N’Diaye, C. Busse, and T. Michely, Nano Lett. 8, 565 共2008兲. 4 I. Forbeaux, J. M. Themlin, and J. M. Debever, Phys. Rev. B 58, 16396 共1998兲. 5 C. Berger, Z. M. Song, X. B. Li, X. S. Wu, N. Brown, C. Naud, D. Mayo, T. B. Li, J. Hass, A. N. Marchenkov, E. H. Conrad, P. N. First, and W. A. de Heer, Science 312, 1191 共2006兲. 6 N. Camara, G. Rius, J.-R. Huntzinger, A. Tiberj, N. Mestres, P. Godignon, and J. Camassel, Appl. Phys. Lett. 93, 123503 共2008兲. 7 J. Kedzierski, P. L. Hsu, P. Healey, P. W. Wyatt, C. L. Keast, M. Sprinkle, C. Berger, and W. A. de Heer, IEEE Trans. Electron Devices 55, 2078 共2008兲. 8 www.novasic.com. 9 J. C. Burton, L. Sun, F. H. Long, Z. C. Feng, and I. T. Ferguson, Phys. Rev. B 59, 7282 共1999兲. 10 A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov, S. Roth, and A. K. Geim, Phys. Rev. Lett. 97, 187401 共2006兲. 11 C. Faugeras, A. Nerriere, M. Potemski, A. Mahmood, E. Dujardin, C. Berger, and W. A. de Heer, Appl. Phys. Lett. 92, 011914 共2008兲. 12 J. Rohrl, M. Hundhausen, K. V. Emtsev, R. Graupner, and L. Ley, Appl. Phys. Lett. 92, 201918 共2008兲. 13 A. J. Stone and D. J. Wales, Chem. Phys. Lett. 128, 501 共1986兲. 14 G. Kresse and J. Hafner, Phys. Rev. B 47, 558 共1993兲. 15 J. Hass, F. Varchon, J. E. Millan-Otoya, M. Sprinkle, N. Sharma, W. A. de Heer, C. Berger, P. N. First, L. Magaud, and E. H. Conrad, Phys. Rev. Lett. 100, 125504 共2008兲. 16 L. G. Cancado, K. Takai, T. Enoki, M. Endo, Y. A. Kim, H. Mizusaki, A. Jorio, L. N. Coelho, R. Magalhaes-Paniago, and M. A. Pimenta, Appl. Phys. Lett. 88, 163106 共2006兲. 17 P. Poncharal, A. Ayari, T. Michel, and J. L. Sauvajol, Phys. Rev. B 78, 113407 共2008兲.

Downloaded 06 Jan 2009 to 117.32.153.167. Redistribution subject to AIP license or copyright; see http://apl.aip.org/apl/copyright.jsp

Early stage formation of graphene on the C face of 6H ...

SiC sublimation are low electron energy diffraction and scanning tunneling microscopy. ... plates cut from an on-axis 6H-SiC wafer, n-type doped to. 51017cm−3.

509KB Sizes 3 Downloads 147 Views

Recommend Documents

The Rise of Graphene
phenomena, some of which are unobservable in high-energy physics, can now be mimicked and tested in table-top ..... An alternative explanation for the half-integer ..... focused in this direction, and such companies as Intel and IBM fund this ...

Chemical Synthesis of Graphene - Arkivoc
progress that has been reported towards producing GNRs with predefined dimensions, by using ..... appended around the core (Scheme 9), exhibit a low-energy band centered at 917 .... reported an alternative method for the preparation of a.

Formation of contacts on semiconductor substrates
Dec 23, 2010 - The surface resistivity of cadmium-based substrates in par ticular, for example a CdZnTe semiconductor substrate, is degraded when the ...

Emission of spherical cesium-bearing particles from an early stage ...
Emission of spherical cesium-bearing particles from an early stage of the Fukushima nuclear accident.pdf. Emission of spherical cesium-bearing particles from ...

The Influence of Prior Expectations on Emotional Face ...
Jun 1, 2012 - divided into 4 experimental blocks of ∼9 min each. The order of se- ... sented on a laptop using the Cogent Software Package (http://www.

E-Books Early-Stage Technologies
Download Book Early-Stage Technologies: Valuation and Pricing (Intellectual Property-General, Law, Accounting Finance, Management, Licensing, Special Topics), Read Book Early-Stage Technologies: Valuation and Pricing (Intellectual Property-General, L

Schonenberger, Bandstructure of Graphene and Carbon Nanotubes ...
Schonenberger, Bandstructure of Graphene and Carbon ... otubes, An Exercise in Condensed Matter Physics.pdf. Schonenberger, Bandstructure of Graphene ...

Early Salesian Regulations: Formation in the preventive system
RASOR John M., “Early Salesian Regulations: Formation in the preventive system”,. Journal of Salesian Studies, vol 8, no 2, 1997, 206-268.

on the role of disks in the formation of stellar systems: a ...
dimensionless parameters in Section 2. We describe the ini- tial conditions and the numerical code used in Section 3. In Section 4, we derive analytic predictions for the behav- ior of disks as a function of our parameters. We describe the main resul

Review on Different Cluster Head Formation Algorithm of ... - IJRIT
IJRIT International Journal of Research in Information Technology, Volume 3, .... On the other hand, intra cluster communication cost reflects the node degree or ...

Review on Different Cluster Head Formation Algorithm of ... - IJRIT
as locations under heavy traffic load [1]. Nodes in such areas quickly exhaust energy resources, leading to disconnection in network services. Cluster based ...

Some Observations on the Early History of Equilibrium ...
ing contributions of Michael Fisher to statistical mechanics and critical phe- nomena. ... tributed an account(1) of Michael's activities at King's College London,.

Authorization of Face Recognition Technique Based On Eigen ... - IJRIT
IJRIT International Journal of Research in Information Technology, Volume 2, ..... computationally expensive but require a high degree of correlation between the ...

The Face of Computer Science
research, we need to advocate for ourselves, and we need to convince those in the public eye that ... We fund alternative energy, we save the the planet.

KNLF121212 Declaration of The Formation of the Khmer National ...
KNLF121212 Declaration of The Formation of the Khmer National Liberation Front.pdf. KNLF121212 Declaration of The Formation of the Khmer National ...

Read Magic on the Early English Stage Online Ebook
Magic on the Early English Stage investigates the performance of magical tricks, illusions, effects and their staged appearance in the medieval and early English theatre. Performers who created such magic were not known as conjurors, as we might refe

VARIATIONS IN THE FORMATION OF THORACIC splanchnic ...
VARIATIONS IN THE FORMATION OF THORACIC splanchnic nerves.pdf. VARIATIONS IN THE FORMATION OF THORACIC splanchnic nerves.pdf. Open.

RESERVOIR CHARACTERIZATION OF THE JERIBE FORMATION ...
RESERVOIR CHARACTERIZATION OF THE JERIBE F ... LLS IN HAMRIN OIL FIELD, NORTHERN IRAQ.pdf. RESERVOIR CHARACTERIZATION OF THE ...

The Face of Computer Science
The NSF provides 86% of the federal funds for computer science research in ... companies, including Amazon.com, investing in cloud computing, the NSF's ...

The unique electronic properties of graphene – a ... - Semantic Scholar
In a time when cutting-edge scientific research is ex- pensive and complex, it ... stack of graphene layers; carbon nanotubes are made of rolled-up sheets of ...

The Formation of Financial Networks
settings although feasible, would merely add complexity without adding much ...... between effi cient outcomes and individual incentives is a classical theme.