Direct graphene growth on metal oxides by molecular epitaxy

Abstract

Direct growth of graphene on Co.sub.3O.sub.4(111) at 1000 K was achieved by molecular beam epitaxy from a graphite source. Auger spectroscopy shows a characteristic sp.sup.2 carbon lineshape, at average carbon coverages from 0.4-3 monolayers. Low energy electron diffraction (LEED) indicates (111) ordering of the sp.sup.2 carbon film with a lattice constant of 2.5 (0.1) characteristic of graphene. Six-fold symmetry of the graphene diffraction spots is observed at 0.4, 1 and 3 monolayers. The LEED data also indicate an average domain size of 1800 , and show an incommensurate interface with the Co.sub.3O.sub.4(111) substrate, where the latter exhibits a lattice constant of 2.8 (0.1) . Core level photoemission shows a characteristically asymmetric C(1s) feature, with the expected lr to lr* satellite feature, but with a binding energy for the three monolayer film of 284.9 (0.1) eV, indicative of substantial graphene-to-oxide charge transfer.

Claims

1. A method of forming a composition of matter comprising macroscopically continuous layers of graphene on a metal oxide surface, wherein said method comprises steps of: forming a layer of a metal consisting of cobalt and having a first major surface; annealing said first major surface in oxygen to form an oxide of said metal on said first major surface; and forming more than one layer of macroscopically continuous graphene on said metal oxide layer by carbon molecular beam epitaxy (MBE) thereby creating an oxide/graphene interface between said metal oxide layer and a first layer of graphene, wherein said oxide/graphene interface is incommensurate, without reconstruction of the metal oxide surface, thus yielding C.sub.6V symmetry in said first layer of graphene, wherein said step of annealing and said step of forming more than one layer of macroscopically continuous graphene are both conducted at temperatures of not more than one thousand degrees Kelvin.

2. A composition of matter comprising a macroscopically continuous multilayer film of graphene on a magnetically polarizable metal oxide creating a graphene/oxide interface between a surface of said metal oxide and a first layer of graphene, wherein said metal oxide consists of cobalt oxide and is in turn on a room temperature ferromagnet, wherein the graphene/oxide interface is incommensurate, without reconstruction of said metal oxide surface, thus yielding C.sub.6V symmetry in said first layer of graphene.

3. The composition of matter of claim 2, wherein said composition of matter comprises a graphene field effect transistor, a coherent spin field effect transistor, a magnetic tunnel junction or switch.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The accompanying drawings, which are incorporated herein and constitute part of this specification, illustrate exemplary embodiments of the invention, and, together with the general description given above and the detailed description given below, serve to explain the features of the invention.

(2) FIG. 1 is (a) Auger spectra acquired after deposition of 0.4 ML graphene (dashed line) and 3.0 ML graphene (solid line), insetdetail of C(KVV) spectral region showing characteristic sp.sup.2 lineshape; (b) Growth curve showing evolution of Auger-derived average carbon overlayer thickness (number above each data point, ML) as a function of deposition time at 1000 K. Lines are least squares fits to the data.

(3) FIG. 2 is LEED and corresponding line scan data for (a,b) 0.4 ML graphene on Co.sub.3O.sub.4(111), and (c,d) 3 ML graphene on Co.sub.3O.sub.4(111). Arrows (a,c) mark diffraction spots associated with Co.sub.3O.sub.4(111), as do inner spots in the outer ring of bifurcated features (e.g., O1,O2b,d). Outer spots in the outer ring of bifurcated features (e.g., G1, G2b,d) are graphene-related. LEED beam energy is 65 eV.

(4) FIG. 3 is (a) Variation in averaged LEED spot intensities for A site diffraction spots (dark trace) and B site spots (light trace) as a function of electron beam energy at 0.4 ML graphene coverage. Inset shows labeling of A site and B site spots in the diffraction pattern; (b) corresponding data at 3 ML coverage, with inset showing labeling of A site and B site spots in the diffraction pattern.

(5) FIG. 4 is (a) C(1s), (b) Co(2p) and (c) O(1s) XPS spectra for a 3 ML graphene film on Co.sub.3O.sub.4(111). The magnified portion of the C(1s) spectrum (a, inset) near 291 eV binding energy indicates the presence of a ir,ir* shakeup feature.

(6) FIG. 5 is the imaginary part of the dielectric optical response function for the 3 ML graphene/Co.sub.3O.sub.4(111)/Co(0001) system as interpreted directly from data (a). This is compared to the pure dielectric response of graphene sublimated from 4H-silicon carbide, on the carbon face (b), and on the silicon face (d); and graphene transferred to SiO.sub.2 (c), adapted from ref. 40. Negative numbers for (b-d) are the vertical offsets of the data.

(7) FIG. 6 illustrates the process of the invention in conventional flowchart format. Thus, the process begins with the formation of a metal layer, which in the second step is annealed in an oxygen atmosphere to form a metal oxide surface on the metal. Thereafter, in the third step of the process claimed a layer of macroscopically continuous graphene is formed on the surface of the metal oxide.

(8) As shown in FIG. 7, practice of this process results in a device that comprises ferromagnetic layer 100, on a major surface of which is formed a layer of metal oxide 200 by annealing the ferromagnetic layer 100 in an oxygen atmosphere. On the exposed surface of the metal oxide layer 200, a film of more than one layer of macroscopically continuous graphene 300 is formed by molecular beam epitaxy.

(9) The process shown in FIG. 7 may be schematically illustrated as set forth in FIGS. 8A-8C. Thus, in FIG. 8A a layer of metal 100 having a first major surface 102 is formed as shown. In the second step of the recited process, that major surface 102 of layer 100 is annealed to form a layer of the oxide of that metal, shown in FIG. 8B as oxide layer 200. In the third step of the process used to form the device of the claims, more than one layer 300 of macroscopically continuous graphene is formed on the metal oxide layer 200. Since these layers of macroscopically continuous graphene are formed by MBE, the interface 400 between the surface of metal oxide layer 200 and the first formed macroscopically continuous layer of graphene 300 is incommensurate with no reconstruction of the surface.

DETAILED DESCRIPTION OF THE DRAWINGS

(10) The examples of the invention can be best understood, initially, by reference to the Figures briefly described above. The descriptions set forth are applicable to the more detailed description immediately below. Wherever specific aspects of the drawings are referred to, such as 1b or 3a, the same numbers will be used throughout the drawings to refer to the same aspects.

(11) FIG. 1 displays the Auger electron spectra (FIG. 1a) acquired after depositions corresponding to 0.4 ML (first deposition, dashed line) and 3.0 ML (last deposition, solid line). The C(KVV) spectral region of the Auger electron spectra (FIG. 1a, inset) shows line/shape characteristic of sp.sup.2 carbon [28]. The growth curve showing the evolution of Auger spectra-determined average carbon thickness as a function of MBE deposition time, is plotted in FIG. 1b. The data are well fit by a series of straight lines, with changes in slope corresponding to completion of one monolayer and beginning of another. This is strongly suggestive evidence of layer-by-layer growth [29], consistent with the changing low energy electron diffraction intensity versus voltage curves, as discussed below. Thus, the Auger electron spectroscopy data demonstrate growth of multilayer sp.sup.2 carbon films by MBE on Co.sub.3O.sub.4(111).

(12) FIG. 2 shows LEED images and corresponding line scans for films of 0.4 ML (FIG. 2a,b) and 3 ML (FIG. 2c,d) average thickness. Both LEED images (FIG. 2a,c) consist of an outer six-fold array of bifurcated spots, and an inner array of spots (3 strong, 3 weak, marked by arrows, FIG. 2a) rotated 30 from the outer ring of spots. The outer spots in the bifurcated six-fold array (e.g., G1, G2FIG. 2b,d) increase in relative intensity with increasing carbon deposition while the other spots in the bifurcated outer ring (e.g., O1,O2FIG. 2b,d) as well as the inner ring of spots (arrows, FIG. 2a,c) decrease in intensity during this process. This identifies the outer spots in the bifurcated array as related to a six-fold sp.sup.2 C(111) overlayer-graphene. The other spots are characteristic of the Co.sub.3O.sub.4(111) substrate, and an estimate of the OO surface nearest neighbor distance can be obtained from the line scans, with the reciprocal length from G1 to G2 (FIG. 2b,d) corresponding to 2.5 (0.1) . The OO nearest neighbor distance is thus estimated at 2.8 (0.1) , in excellent agreement with that obtained for adsorbate-free Co.sub.3O.sub.4(111) films [30]. These data show that the graphene/oxide interface is incommensurate or encompasses a very large unit cell. The LEED data thus indicate that graphene grows on the Co.sub.3O.sub.4(111) substrate without significant in-plane reconstruction of the oxide surface.

(13) In FIG. 4 the C(1s) spectrum (FIG. 4a) shows an asymmetric main feature characteristic of graphene [13]. Referencing the spectrometer energy scale to the Al(2p) binding energy of 74.4 eV, characteristic of bulk sapphire [34], yields a Co.sub.(2p3/2) peak maximum binding energy of 778.3 (0.1) eV (FIG. 4b), in excellent agreement with literature values for both bulk and thin film Co [35]. A shoulder at higher binding energies (FIG. 4b) is also consistent with the formation of CO.sub.3O.sub.4 [36], and the intensity ratio of the CO.sub.(2p3/2) oxide and metal features indicates an oxide thickness of 3 ML. The O(1s) peak maximum is at 530.9 (0.1) eV, consistent with O(1s) spectra from both bulk sapphire [35] and Co.sub.3O.sub.4(111) thin films [36], with a small shoulder at higher binding energy suggesting some hydroxylation. The C(1s) main peak binding energy (FIG. 4a) is at 284.9 (0.1) eV, substantially higher than the 284.5 eV binding energy commonly reported for bulk graphite, but similar to the 284.7 eV binding energy reported for monolayer grapheme grown by graphitization of SiC(0001) [37], and indicating charge transfer from graphene to the oxide substrate.

(14) FIG. 5 displays the experimental imaginary part of the dielectric response function for the 3 ML graphene/Co.sub.3O.sub.4(111)/Co(0001) sample. The feature with a maximum near 5.5 eV (FIG. 4, Co.sub.3O.sub.4) attributed to both a it to ir* transition and it plasmon excitation [38], is comparable to but blue-shifted with respect to features in the dielectric function reported for graphene physically transferred to SiO.sub.2 [24, 39, 40] and for graphene grown on Si substrates [24, 38]. These energies have been shown to be somewhat dependent on both the graphene substrate [24] and number of graphene layers [38]. We suggest, however, that the observed blue shift (FIG. 5) is consistent with the significant graphene-to-oxide charge transfer deduced from the XPS C(1s) binding energy (FIG. 4). The spectroscopic ellipsometry data confirm that the dielectric response of the graphene film and CO.sub.3O.sub.4 system are similar to few-layer graphene films grown on SiC(0001) or transferred to SiO.sub.2.

DETAILED DESCRIPTION OF THE INVENTION

(15) Direct graphene growth by chemical or physical vapor deposition has been reported on monolayer h-BN(0001)/Ru(0001) [17] and bulk MgO(111) single crystals [18] at or below 1000 K. In both cases, however, strong substrate/graphene interactions, consistent with commensurate graphene/substrate interfaces [17, 19], result in pronounced substrate-induced modification of the graphene electronic structure [18, 20]. Recently, the growth of continuous graphene layers by chemical vapor deposition on Al.sub.2O.sub.3(0001) has been reported [21], but with growth temperatures >1800 K required for films with few defects. The growth temperature reported here, 1000 K, is fully compatible with Si CMOS front end processing, and is significantly lower than that previously reported (1300 K) [12, 13] for MBE growth on SiC(0001). Other suitable device substrates can be employed at the temperatures employed herein, such as Germania, gallium arsenide and indium.

(16) Growth studies were carried out in a chamber equipped for Auger electron spectroscopy and reverse-view low energy electron diffraction (LEED) I(V) measurements, as described previously [17], but with the addition of a commercial multi-target electron beam evaporator for MBE. Co and graphite rod sources were used, and chamber pressure during deposition was <110.sup.8 Torr. The intensities of LEED diffraction spots were determined using commercial imaging software, as described previously [18], and by plotting I(V) intensity curves. X-ray photoemission (XPS) spectra were acquired in a separate system [22] with a hemispherical analyzer in constant pass energy mode (23.5 eV), using an unmonochromatized Al K.sub.ct x-ray source operated at 15 kV, 300 W. Spectroscopic ellipsometry data were acquired and analyzed by methods described previously [23, 24] using recent optical data of .sub.Co3O4 [25]. Micro-Raman spectra were acquired at two separate spots 200 nm in diameter, physically separated by a distance of 3 mm, using a 514 nm excitation source. Identical spectra (see supplemental information) were observed at each location, indicating that the film is uniform over macroscopic distances, with G/2D intensities consistent with multilayer graphene.

(17) Co films were deposited onto commercially available 1 cm.sup.2 Al.sub.2O.sub.3(0001) substrates at 750 K, with estimated Co thicknesses >40 , as determined by Auger spectroscopy. These films yielded well-defined LEED images characteristic of Co(111) [26]. Subsequent annealing resulted in the surface segregation of dissolved oxygen and the evolution of a 3 monolayer (ML) Co oxide film with LEED pattern identical to that of few-layer Co.sub.3O.sub.4(111) [27]. Graphene films were subsequently grown at 1000 K by carbon MBE.

(18) An estimate of the graphene domain size (R) can be obtained from the FWHM of the graphene diffraction peaks (FIG. 2), according to [31]

(19) $\begin{array}{cc}R=\frac{b}{2\left(1+E/E\right)s}& \left(1\right)\end{array}$

(20) In (1), b is the fidelity factor, generally estimated at 0.35, is the electron wavelength (1.52 ), and (E/E) is the energy spread for the electron gun, estimated at 0.015. .sub.s is the angular divergence estimated from LEED diffraction beams (e.g., G1, FIG. 2b), normalized to the sample-screen distance (150 mm). These data yield an estimate of a typical domain size of 1800 , roughly comparable to HOPG [32].

(21) The LEED data in FIG. 2d might indicate unequal intensities for G1 and G2, which is consistent with a transition to 3-fold symmetry, but which can also be induced by spurious experimental effects. In order to determine whether the graphene overlayers exhibited true six-fold symmetry and graphene lattice A site/B site chemical equivalence, the average integrated intensities of A site and B site diffraction spots are plotted as a function of beam voltage (I(V) curves) in FIG. 3, at both 0.4 ML (FIG. 3a) and 3.0 ML (FIG. 3b) average graphene coverage. The data show very similar I(V) curves for averaged A site and B site intensities at both low and high coverages, indicating the persistence of C.sub.3v LEED symmetry from 0.4 ML to 3 ML graphene coverage. .sub.C6v symmetry was also observed for LEED features at 1 ML (not shown). The sharpness of the features, and the existence of different maxima and minima at 0.4 ML (FIG. 3a) than at 3.0 ML (FIG. 3b), especially in the surface sensitive region <150 eV electron beam energy, also support the possibility of layer-by-layer growth, since island formation would result in the simultaneous existence of regions of low and high coverage, with the resulting I(V) curve being the average of I(V) curves characteristic of each coverage. The Auger electron spectroscopy and LEED data (FIGS. 1-3) indicate the controlled growth of multilayer grapheneto an average thickness of at least 3 MLwith the exhibition of .sub.C6v LEED symmetry at both low and high graphene coverages.

(22) The same conclusion was reached after the evaluation of average A site and B site intensities using commercial imaging software. Six-fold symmetry is expected for graphene in contact with an incommensurate substrate, as both graphene A sites and B sites will experience an ensemble of substrate environments. In contrast, C.sub.3v symmetry may be expected for multilayers, if the layers exhibit Bernal stacking (A-B-A) and the graphene layer is single domain. A more detailed analysis of the LEED data, including dynamical scattering analysis, is required for further elucidation of the structural details.

(23) Following the acquisition of Auger electron spectra and LEED data, the sample was exposed to ambient, and then re-inserted into UHV. Subsequently acquired Auger electron spectra and LEED data showed no change from those in FIG. 1a (solid line) and FIG. 2c, respectively, indicating that the sample was inert to ambient exposure. This effect has been observed for other reactive surfaces covered with graphene layers [17, 18, 20, 26, 33] and indicates that the graphene multilayers form a macroscopically continuous film shielding the 3 ML Co.sub.3O.sub.4(111)/Co(0001) film from reaction with ambient. The XPS spectra of the 3 ML graphene film on Co.sub.3O.sub.4(111)/Co(0001) substrate are shown in FIG. 4.

(24) While the present invention has been disclosed with references to certain embodiments, numerous modification, alterations, and changes to the described embodiments are possible without departing from the sphere and scope of the present invention, as defined in the appended claims. Accordingly, it is intended that the present invention not be limited to the described embodiments, but that it has the full scope defined by the language of the following claims, and equivalents thereof.

REFERENCES

(25) [1] Novoselov K S, Geim A K, Morozov S V, Jiang D, Zhang Y, Dubonos S V, Grigorieva I V, Firsov A A. 2004 Science. 306 666. [2] Novoselov K S, Geim A K, Morozov S V, Jian D, Katsnelson M I, Grigorieva I V, Dubonos S V. Firsov A A. 2005 Nature. 438 197. [3] Neto A H C., Guineau F, Peres N M R, Noveoselov K S. Geim A K. 2009 Rev. Mod, Phys. 81 109. [4] Haugen H, Huertas-Hernando D, Brataas A. 2008 Phys. Rev, B. 77 115406. [5] Tombros N, Jozsa C, Popinciuc M, Jonkman H T, van Wees B J. 2007 Nature. 448 571. [6] Li X. Cai W, An J, Kim S, Nah J, Yang D, Piner R. Velamakanni A, Jung I, Tutuc E, Banerjee S K, Colobmo L, Ruoff R S. 2009 Science. 324 1312. [7] Reina A, Thiele 5, Jia X, Bhaviripudi S, Dresselhaus M S, Schaefer J A, Kong J. 2009 Nano. Res. 2 509. [8] Berger C, Song Z, Li T, Li Z. Ogbazghi A Y, Feng R. Dai Z, Marchenkov A N, Conrad, E H, First, P N, de Heer W A. 2004 J. Phys. Chem. B. 108 19912. [9] Berger C, Spong Z, Li X, Wu X, Brown N, Naud C, Mayou D, Li T, Hass J, Marchenkov A N., Conrad E H, First P N, de Heer W A. 2006 Science. 312 1191. [10] Forbeaux I, Themlin J-M, Debever J-M. 1998 Phys. Rev. B 58 16396. [11] Zhang Y, Brar V W, Girit C, Zettl A, Crommie M F. 2009 Nature Phys. 5 722. [12] Park J, Mitchel W C, Grazulis L, Smith H E, Eyink K G, Boeckl J J, Tomich D H, Pacley S D, Hoelscher J E. 2010 Adv Mat. 22 4140. [13] Moreau E, Ferrer F J, Vignaud D, Godey S, Wallart X. 2010 Phys. Stat. Solidi A. 207 300. [14] Zheng Y, Ni G, Toh C, Zeng M, Chen S, Yao K, Ozyilmaz B. 2009 Appl. Phys. Lett. 94 163505. [15] Imam S A, Deshpande T, Guermune A, Siaj M, Szkopek T. 2011 Appl. Phys Lett. 99 082109. [16] Michetti P, Recher P, Iannaccone G. 2010 Nano Lett. 10 4463. [17] Bjelkevig C, Mi Z, Xiao J, Dowben P A, Wang L, Mei W, Kelber J A. 2010 J. Phys: Cond. Matt. 22 302002. [18] Gaddam S, Bjelkevig C, Ge S, Fukutani K, Dowben P A, Kelber J A. 2011 J. Phy.: Cond Matt. 23 072204. [19] Kelber J A, Gaddam S, Vamala C, Eswaran S, Dowben P A. 2011 Proc. SPIE 8100 81000Y-1. [20] Kong L, Bjelkevig C, Gaddam S. Zhou M, Lee Y H, Han G H, Jeong H K, Wu N, Zhang Z, Xiao J, Dowben P A, Kelber J A. 2010 J. Phys. Chem. C. 114 21618. [21] Fanton, M A. Robinson, J A, Weiland, B E, Labella, M. Trumbel K, Kasarda, R., Howsare, C, Hollander M, Snyder D W. 2011 Catalytic Free Synthesis of High Mobility Graphene on Sapphire Abstract for Graphene 2011 (Bilbao, Spain, April 11-14) (obtained from www.phantomsnet.net). [22] Bjelkevig C, Kelber J. 2009 Electrochim Acta. 54 3892. [23] Hofmann T. Boosalis A, Kuhne P. Herzinger C M, Woollam J A. Gaskill D K, Tedesco J L, Schubert M. 2011 Appl. Phys. Lett. 98 041906. [24] Boosalis, A., Hofmann, T., Schoche, S., Dowben, P., Gaddam, S., Vamala, C., Kelber, J., Darakchieva, V., Yakimova, R., Nyakiti, L. O., Wheeler, V. D., Myers-Ward, R. L., Eddy Jr., C. R., Gaskil, D. K., Schubert, M. Optical properties of graphene on SiC polytypes determined by spectroscopic ellipsometry from terahertz to the VUV. 2011 Conference Abstract, Nano DDS (Brooklyn, N.Y. August 29). [25] Donders M E, Knoops H C M, van de Sanden, M C M, Kessels W M M, Notten P H L. 2011 J Electrochem Soc. 158 G92. [26] Ago H, Ito Y, Mizuta N, Yoshida K, Hu B, Orofeo C M, Tsuji M, Ikeda K, Mizuno S. 2010 ACS Nano. 4 7407. [27] Vaz C A F, Prabhakaran D, Altman E I, Henrich V E. 2009 Phys. Rev. B. 80 155457. [28] Viljoen P E, Roos W D, Swart H C, Holloway P H. 1996 Appl. Surf. Sci. 100/101 612. [29] Argile C, Rhead G E. 1989 Surf Sci Rep. 10 277. [30] Meyer W, Biedermann K, Gubo M, Hammer L, Heinz K. 2008 J. Phys.: Cond. Matt. 20 265011. [31] Ertl G, Kuppers J. 1974 Low energy electrons and surface chemistry (Verlag-Chemie, Weinheirn, FRG) 159. [32] Lu X, Yu M, Huang H, Ruoff R S. 1999 Nanotechnol. 10 269. [33] Dedkov Y S., Fonin M, Rudiger U, Laubschat C. 2008 Phys. Re.v Lett. 100 107602. [34] Moulder J F, Stickle W F, Sobol P E, Bomben K D. 1995 Handbook of X-ray photoelectron spectroscopy (Physical Electronics, Eden Prairie, Minn., USA). [35] Chambers S A., Droubay T, Jennision D R, Mattson T R. 2002 Science. 297 837. [36] Petito S C., Langell M A. 2004 J. Vac. Sci. and Technol. A. 22 1690. [37] Emtsev K V, Speck F, Seyller T, Ley L. 2008 Phys. Rev. B. 77 155303. [38] Santoso I, Gogoi P K, Su H B, Huang H, Lu Y, Qi D, Chen W. Majidi M A, Feng Y P, Wee A T S, Loh K P, Venkatesan T, Saichu R P, Goos A, Kotlov A, Rubhausen M, Rusydi A. 2011 Phys. Rev. B. 081403. [39] Kravets V G., Grigorenko A N, Nair R R, Blake P, Anissimova S, Novoselov K S, Geim A K. 2010 Phys Rev B. 81 155413. [40] Nelson F J, Kaminenei V K, Zhang T, Comfort E S, Lee J U, Diebold A C. 2010 Appl. Phys. Lett. 97 253110. [41] Lazarov V K, Plass R, Poon H-C, Saldin D K, Weinert M, Chambers S A, Gajdardziska-Josifovska M. 2005 Phys. Rev. B. 71 115434. [42] Rhr F, Wirth K, Libuda J, Cappus D, Baumer M, Freund H-J. 1994 Surf Sci. 315 L977. [43] Cappus D, Xu C, Ehrlich D, Dillmann B, Ventrice Jr. C A, Al Shamery K, Kuhlenbeck H, Freund H-J. 1993 Chem. Phys. 177 533. [44] Goniakowski J, Noguera C. 2002 Phys. Rev. B. 66 085417.