Supplementary Information
Controlled assembly of graphene-capped nickel, cobalt and iron silicides
O. Vilkov1,2, A. Fedorov1,3, D. Usachov1, L. V. Yashina4, A. Generalov1,2,
K. Borygina1, N. I. Verbitskiy4,5, A. Grüneis3,5, and D. V. Vyalikh1,2,∗
1 St. Petersburg State University, Ulyanovskaya str. 1, St. Petersburg 198504, Russia
2 Institute of Solid State Physics, Dresden University of Technology, Zellescher Weg 16, D-01062 Dresden, Germany.
3 Leibniz-Institut für Festkörper- und Werkstoffforschung Dresden, Helmholtzstr. 20, P.O. Box 270116, D-01171 Dresden, Germany
4 M.V. Lomonosov Moscow State University, Leninskie Gory 1/3 119991 Moscow, Russia
5 Faculty of Physics, University of Vienna, Boltzmanngasse 5 A-1090 Vienna, Austria
*Correspondence to
Supplementary Figures
Supplementary Figure S1: Scheme of the LEED pattern formation for the Ni3Si phase under graphene. The LEED image is recorded at the beam energy of 93 eV. Brighter spots in the image correspond to the (1x1) structure of graphene with respect to the initial Ni(111) surface. With increasing of the electron beam energy the difference between the intensities of integer and fractional reflexes almost vanishes indicating the bulk structure of Ni3Si silicide.
Supplementary Note 1:
The samples were prepared in ultra-high vacuum conditions with a base pressure in the vacuum chamber of 1x10-10 mbar. A W(110) single crystal in form of a disk with a radius of 5 mm was used as a substrate. To prepare a single layer of graphene on metal surfaces like a nickel, cobalt or iron, we have used a well-elaborated procedure which is discussed in ref. 14, 15, 38, 44. In the beginning, the substrate was carefully cleaned by annealing it at 1300 °C in an oxygen atmosphere at a partial pressure of 5 x 10-8 mbar, and subsequent flashing to 2200 °C. The surface cleanness was conclusively confirmed by observation of a sharp (1x1) LEED pattern, indicating an absence of carbon- or oxygen-related superstructures. A thin metal (Ni, Co or Fe) film of usually 10 nm thickness was produced on top of the W(110) by means of molecular beam epitaxy, using a deposition rate of 1-2 Å/min. It is worth noting that in the case of Ni and Co films the growth occurs in Nishiyama-Wassermann regime, leading to well-oriented crystalline structure of the Ni(111) and Co(0001) surfacesS1, S2. The growth of Fe film results in a well-oriented Fe(110) surfaceS3. The high quality of the produced monocrystalline metal films was confirmed by LEED and photoemission. In order to be sure of the homogeneity and consistency of the produced metal film on W(110), we checked for W 4f photoemission lines. In case of high quality film these lines have never been observed. However, when we deviate from the synthesis protocol, W 4f emission lines immediately appear, clearly showing that the film is broken.
Graphene synthesis on Ni(111) and Co(0001) surfaces was performed by heating the sample up to the synthesis temperature of 550°C followed by exposing the surface to propylene gas at a pressure of 10−6 mbar for 15 min. In the case of the iron-based system the exposure was increased to avoid carbidization and to produce a complete graphene layer. We used propylene pressure of 5 x 10−6 mbar at a temperature of 700°C for 25 min.
It is well known that the graphene growth at these conditions is self-limited to one monolayer. In case of nickel substrates graphene forms in a high-quality crystalline structure as inferred from the sharp (1x1) LEED pattern and explained by the perfect match of the lattice constants of the Ni(111) surface and graphene. The Co(0001) surface has slightly larger lattice constant with respect to graphene, and therefore, besides the (1x1) structure, graphene domains with different orientations are present14. The situation for the Fe(110) surface is even more complicated. It is commensurate to graphene only in one direction. Therefore, graphene grown on top becomes corrugated due to the lattice mismatch in other direction15.
Silicon atoms were deposited onto the graphene surface from a pure Si crystal, heated directly by electric current up to the temperature of sublimation. The pressure in the vacuum chamber was kept below 5 x 10-10 mbar during deposition, which was enough to avoid silicon oxidation before subsequent intercalation. The deposition rate was 1-2 Å/min and the thickness of deposited Si was controlled by a calibrated quartz microbalance.
Supplementary References:
S1Kämper, K.-P., Schmitt, W., Güntherodt, G., and Kuhlenbeck, H. Thickness dependence of the electronic structure of ultrathin, epitaxial Ni(111)/W(110) layers. Phys. Rev. B38, 9451 (1988).
S2 Pratzer, M., Elmers, H.J., and Getzlaff, M. Heteroepitaxial growth of Co on W(110) investigated by scanning tunneling microscopy.Phys. Rev. B67, 153405 (2003).
S3 Stankov, S. et al. Phonons in iron: From the bulk to an epitaxial monolayer. Phys. Rev. Lett.99, 185501 (2007).