Selecting a single orientation for millimeter sized graphene sheets
R. van Gastel, A.T. N'Diaye, D. Wall, J. Coraux, C. Busse, F.-J. Meyer zu Heringdorf, N. Buckanie, M. Horn von Hoegen, T. Michely, B. Poelsema
aa r X i v : . [ c ond - m a t . m t r l - s c i ] J u l Selecting a single orientation for millimeter sized graphene sheets
R. van Gastel, A.T. N’Diaye, D. Wall, J. Coraux, C. Busse, F.-J. Meyer zuHeringdorf, N. Buckanie, M. Horn von Hoegen, T. Michely, and B. Poelsema University of Twente, MESA + Institute for Nanotechnology,P.O. Box 217, NL-7500AE Enschede, The Netherlands II. Physikalisches Institut, Universit¨at zu K¨oln, Z¨ulpicher Straße 77, 50937 K¨oln, Germany Department of Physics and Center for Nanointegration Duisburg-Essen (CeNIDE),Universit¨at Duisburg-Essen, D-47057 Duisburg, Germany Institut N´eel/CNRS, Bˆat. D, 25 Rue des Martyrs, F-38042 Grenoble Cedex 9, France (Dated: November 11, 2018)We have used Low Energy Electron Microscopy (LEEM) and Photo Emission Electron Microscopy(PEEM) to study and improve the quality of graphene films grown on Ir(111) using chemical vapordeposition (CVD). CVD at elevated temperature already yields graphene sheets that are uniformand of monatomic thickness. Besides domains that are aligned with respect to the substrate, otherrotational variants grow. Cyclic growth exploiting the faster growth and etch rates of the rotationalvariants, yields films that are 99 % composed of aligned domains. Precovering the substrate witha high density of graphene nuclei prior to CVD yields pure films of aligned domains extendingover millimeters. Such films can be used to prepare cluster-graphene hybrid materials for catalysisor nanomagnetism and can potentially be combined with lift-off techniques to yield high-quality,graphene based electronic devices.
Graphene potentially constitutes a new material forelectronic circuitry with vastly improved transport prop-erties over traditional silicon [1]. A large scale applicationof graphene crucially hinges on a fabrication method thatyields perfect graphene sheets, that is low-cost and reli-able. CVD growth of graphene on metals has recentlybeen demonstrated to yield large graphene sheets of uni-form monatomic thickness [3, 4, 5, 6]. This form of epi-taxial CVD, which occurs through ethylene decomposi-tion on the uncovered parts of the metal substrate, isself-limiting since the nature of the process limits thethickness of the graphene sheets to a single atomic layer,in contrast to e.g. graphene formed on heated SiC sub-strates [7, 8]. Metal CVD thus appears to be the routeof choice for fabrication of large graphene sheets. In situgrowth studies with LEEM and PEEM have however,highlighted a new problem. The orientation of the do-mains that make up the graphene sheet is not always inregistry with the substrate. For the case of e.g. Ir(111),four different orientations have been observed [9]. Theelectronic properties of graphene sheets depend sensi-tively on the relative orientation with respect to the sub-strate [10, 11, 12]. Also, applications of cluster super-lattices of magnetically or catalytically active materialsgrown on the graphene sheets [13] require full controlover the orientation of the domains. Here, we addressthis problem by tailoring the epitaxial process to grow amillimeter sized, monatomic thickness graphene sheet ofsingle, aligned orientation.An Ir(111) single crystal was heated to 1123 K and ex-posed to a 1 · − mbar partial pressure of O to removeresidual carbon contamination. CVD grwoth of graphenesheets was performed by exposing the surface to ethy-lene at elevated temperatures. Growth of the grapheneceases when the fractional surface coverage of grapheneapproaches 1 ML. Threshold PEEM images using a Hg discharge lamp yield a high intensity from the grapheneflakes and very low intensity from the bare Ir(111) sur-face. Contrast between the different rotational domainsis achieved in LEEM mode at various electron energies.First, growth of a graphene sheet was studied at a tem-perature of 1411 K by exposing to an ethylene partialpressure of 1 · − mbar. The formation of the grapheneis shown in Fig. 1. Initially, only a single phase formsthat has its lattice vectors parallel to the substrate lat-tice. Later, graphene domains that are rotated with re-spect to the substrate lattice are observed to form at theedges of the original nuclei and grow at a rate that issubstantially faster. In what follows, we shall refer tothese as aligned and rotated domains, respectively. Thestructure of the graphene sheet after it has completed isshown in Fig. 1(b). From the simple observation thatcontrast between different domains is observed in thesethreshold PEEM images, we conclude that there is a sig-nificant variation of the electronic structure of the filmbetween different domains. Even though the sheet thick-ness is very uniform and could already be characterizedas a high quality graphene film, further control over therotational orientation of the domains is desired.One way to produce a high quality graphene film of asingle rotational phase is done by exploiting the higher re-activity of the edges of the rotated domains. Not only dothe three types of rotated domains grow at a rate that ishigher than that of aligned domains, they are also etchedaway by oxygen at an increased rate [14]. Fig. 2 high-lights this experimental approach. The Ir(111) surfacewas alternatingly exposed to ethylene and O at partialpressures of 5 · − mbar. Exposure to ethylene leadsto the formation of new nuclei and continued growth ofaligned domains. It also gives rapid growth of any ro-tated domains that have formed. Exposure to O , shownin Fig. 2(b), then preferentially etches away the rotated FIG. 1: 50 µ m Field Of View (FOV) PEEM images ofgraphene growth on Ir(111) at T = 1411 K. (a, t = 265s) Two different types of domains are observed to form. Thebrighter of the two is aligned with the substrate, whereas theother, darker, type of domain is rotated by approximately30 ◦ with respect to the substrate. (b, t = 2120 s) The filmhas fully closed to form a graphene sheet consisting of variousrotational domains. (c)
Relative area fractions of the two dif-ferent types of domains that are visible in panels (a)-(b). Themajority of the graphene sheet consists of rotated domains. domains until only parallel domains remain. Cyclic rep-etition of this procedure yields a monolayer thick, nearuniform graphene sheet. A relative fraction of 99 % ofthe sheet consists of aligned domains. The drawback ofthis method is that closure of the film to produce a “per-fect” graphene sheet is not possible. The last step in thegrowth process always has to be a growth step, implyingthat the nucleation of rotated domains can not be fullyprevented.The nucleation of rotated domains occurs at the edgesof parallel domains [9]. In our measurements, we alsoobserve that growth of aligned domains and the nucle-ation of rotated domains occurs predominantly at thoseedges that do not run parallel to the substrate lattice vec-tors. This observation was exploited to further improvethe quality of the films beyond what was demonstratedin Fig. 2 with the cyclic recipe. A monolayer of ethy-lene was preadsorbed on the surface at room tempera-ture. Upon heating the substrate to the growth tem-perature this leads to the formation of a high densityof small aligned graphene domains that have edges par-allel to the substrate lattice [4]. This effectively forces
FIG. 2: 102 µ m FOV PEEM images of graphene growth onIr(111) obtained at a temperature of 1126 K. (a, t = 344 s) Several parallel and rotated domains have nucleated after thefirst growth cycle. (b, t = 265 s) The first O etching stephas completely removed all rotated domains. (c, t = 795 s) After two more etch and growth steps, the graphene film nowconsists of a majority of parallel domains. Rotated domainscontinue to nucleate at every growth step, as is visible in theimage. (d, t = 2120 s) The film has fully closed to forma graphene sheet. A relative fraction of 99 % of the sheetconsists of parallel domains. (e)
Relative area fractions ofthe two different types of domains that are visible in panels(a)-(d). any graphene domains that impinge on existing nuclei tomaintain their aligned orientation. Fig. 3 highlights thesubsequent growth when the substrate is exposed to anethylene partial pressure of 1 · − mbar. Figs. 3(b) and(c) illustrate that growth of aligned domains is observedonly in those locations where domain edges are rough,having an orientation deviating from the dense packedsubstrate directions. Small domains with edges oriented FIG. 3: 4 µ m FOV LEEM images of graphene growth onIr(111) at a temperature of 1113 K and recorded with an elec-tron energy of 18.6 V. The dark spot in the top of the imagesis an MCP defect. (a, t = 0 s) Start of ethylene exposureof the surface. The Ir(111) surface has been precovered withmany small graphene nuclei, appearing dark in the LEEMimages. (b, t = 139 s) Several of the predeposited domainshave started to grow. Those domains that have edges alongthe substrate crystallographic directions are not observed togrow. (c, t = 631 s) The graphene film has nearly evolvedinto a sheet. No rotated domains that would yield a higherintensity than the aligned domains, are observed. Severalprecovered domains still persist and do not grow. (d, t =1279 s) The film has fully closed to form a perfectly alignedgraphene sheet. (e)
A 0.25 µ m FOV STM image of the growthof the parallel phase taken after the graphene film was onlypartially completed. The smaller nuclei with straight edgesrunning along substrate crystallographic directions have notgrown significantly, whereas domains with edges of differentorientation have (I t = 0.5 nA, V t = 0.5 V). (f) µ LEED pat-tern obtained of the closed graphene film. The orientation ofthe graphene is unaltered when the beam is scanned over anarea of several millimeters. parallel to substrate lattice vectors are not observed togrow, illustrated by the STM image shown in Fig. 3(e).Rotated domains do not form. The µ LEED pattern thatis shown in Fig. 3(f) is measured over several millime-ters of our 6 mm wide sample. Defects are sporadicallyfound, but always in locations where we have to presumethat they were induced by features present on the Ir(111)substrate. The graphene sheet that is formed throughthis recipe has the added advantage that its orientationis uniquely determined by the orientation of the Ir(111)substrate.In conclusion, we have grown millimeter sized,graphene films of a single orientation. Cyclic growth ofthe graphene film exploiting the different growth and O etching speeds of the domain variants yields films that arealigned to the substrate dense packed orientation up to afraction of 99 %. The final approach, using preadsorptionof ethylene on the Ir(111) surface at room temperature,followed by CVD growth at elevated temperatures yieldsperfectly aligned sheets that are ready for application. Acknowledgments
Financial support through Deutsche Forschungsge-meinschaft is gratefully 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 , 666 (2004). [2] X.K. Lu, M.F. Yu, H. Huang and R.S. Ruoff, Nanotech-nology , 269 (1999).[3] A.T. N’Diaye, J. Coraux, T. Plasa, C. Busse and T.Michely, New J. Phys. , 043033 (2008).[4] J. Coraux, A.T. NDiaye, M. Engler, C. Busse, D. Wall,N. Buckanie, F.-J. Meyer zu Heringdorf, R. van Gastel,B. Poelsema and T. Michely, New J. Phys. , 023006(2009).[5] P.W. Sutter, J.-I. Flege and E.A. Sutter, Nat. Mater. ,406 (2008).[6] E. Loginova, N.C. Bartelt, P.J. Feibelman and K.F. Mc-Carty, New J. Phys. , 093026 (2008).[7] R.M. Tromp and J.B. Hannon, Phys. Rev. Lett. ,106104 (2009).[8] K.V. Emtsev, A. Bostwick, K. Horn, J. Jobst, G.L. Kel-logg, L. Ley, J.L. McChesney, T. Ohta, S.A. Reshanov, J.R¨ohrl, E. Rotenberg, A.K. Schmid, D. Waldmann, H.B. Weber and T. Seyller, Nat. Mat. , 203 (2009).[9] E. Loginova, S. Nie, K. Th¨urmer, N.C. Bartelt and K.F.McCarty, submitted to Phys. Rev. B (2009).[10] L. Pletikosi´c, M. Kralj, P. Pervan, R. Brako, J. Coraux,A.T. N’Diaye, C. Busse and T. Michely, Phys. Rev. Lett. , 056808 (2009).[11] G. Giovannetti, P.A. Khomyakov, G. Brocks, V.M.Karpan, J. van den Brink and P.J. Kelly, Phys. Rev.Lett. , 026803 (2008).[12] P.A. Khomyakov, G. Giovannetti, P.C. Rusu, G. Brocks,J. van den Brink and P.J. Kelly, Phys. Rev. B , 195425(2009).[13] A.T. N’Diaye, S. Bleikamp, P.J. Feibelman and T.Michely, Phys. Rev. Lett.97