1 A. Optimizing the growth conditions of large-scale graphene films Figure S1. Optical microscope images of graphene films transferred on 300 nm SiO 2 /Si substrates. a, Images of the graphene films grown on different thicknesses of Ni layers and growth time. The sample grown on 100nm thick Ni for 30 s shows monolayer domains as large as 20x20 μm 2. b, An optical image showing linear crack patterns with an angle of 60 or 120, suggesting well developed crystallographic coordination of the graphene layers. c, A histogram of the number of graphene layers two representative growth with different Ni thickness and growth time. www.nature.com/nature 1
2 Figure S2. Optical microscope images of graphenes grown on different thickness Ni layers. As the thickness decreases, the Ni layers tend to form aggregates rather than film structures during the growth process (1000 C for 7 min). B. Properties of as-grown graphene films on Ni Figure S3. Raman spectra of the as-grown graphene film on Ni, corresponding to the colored spots in the inset optical microscope image. The darker colors correspond to thicker graphene layers. www.nature.com/nature 2
3 Fig. S4. AFM images of the ripple structures of graphene films grown on a thick Ni foil (a and b) and a 300 nm-thick Ni layer (c and d). The AFM profiles along the red and blue dots of a show that the height of the ripples increases with the periodicity (e). C. Transferring graphene films on SiO 2 substrates using polymers Table S1. Transfer yields for various substrates Substrate Ni to PDMS PDMS to SiO 2 Ni to PMMA PMMA to SiO 2 Yield 99 % 95 % 99 % 99 % * In the case of PDMS-to-substrate transfer, the roughness of the target substrate is important to increase the yield in addition to hydrophobicity or surface charge. Flexible graphene films tend to maximize contact area with the transferring substrates. Thus the rough surfaces with larger contact area interact stronger with graphene. Preferentially, the roughness of final substrates should be larger than that of stamping PDMS for efficient transfer process. Employing PMMA as a support, the transferring yield can be further enhanced (Ref. Reina, A. et al. "Transferring and identification of single- and few-layer graphene on arbitrary substrates" J. Phys. Chem. C 112, 17741-17744 (2008)). www.nature.com/nature 3
4 Fig. S5. Photograph images of large-scale graphene films floating on the solution. a, A photograph of the PMMA-supported graphene film floating on FeCl 3 etching solution. b, A photograph of Ni-supported graphene layers floating after etching SiO 2 layers in buffered oxide etchant (BOE) solution. D. Measuring electric transport properties in single layer graphene samples Fig. S6. Single layer graphene device images and their electric characteristics. a-c.optical microscope images of a typical CVD grown single layer/ bilayer graphene samples. The electrodes are arranged to form a quasi Hall-bar geometry. Multi-terminal geometry of the devices allow us to measure four-probe resistance and Hall resistance in order to obtain the conductivity and carrier density respectively. The scale bars represent 5 μm. d. Measured conductivity of device b as a function gate voltage. The inset shows a Hall resistance measurement as a function of magnetic field at a fixed gate voltage V g = -60 V. The carrier concentration can be obtained from the slope of Hall resistance. e. The resistance measured in sample in c as a function of V g at different temperatures. In order to measure the electric properties of single and bilayer samples, we first transferred the CVD grown graphene films using a PDMA stamp as described in the text, and then located the suitable isolated pieces of single or bilayer graphene using an optical microscope. Multiple electrodes were fabricated by the electron beam lithography followed by Cr/Au (5/30 nm) metal deposition. The electrodes were www.nature.com/nature 4
5 configured to form a quasi Hall bar contribution as shown in Fig. S6a-c. In this experiment, we did not perform further device fabrication process such as oxygen plasma etching in order to avoid the potential sample degradation. The transport properties, such as conductivity (σ) and Hall resistance, were measured by the four-terminal and the Hall probe configuration, utilizing the quasi Hall bar geometry of these devices. Fig. S6d shows the conductivity of the single layer graphene sample in S6b as a function of applied gate voltage V g. The inset shows the Hall resistance change as a function of magnetic field at a fixed gate voltage V g =-60 V. The slope of Hall resistance in magnetic field gives the density of the sample n = 6x10 12 cm -2. Once we obtain σ and n, the mobility μ was calculated from the Drude relation: μ=σ/(en). Due to the imperfect geometrical configuration of the quasi Hall bar devices, however, the mobility values obtained from these measurements suffers from a relatively large uncertainty, up to 10-40% in the absolute value scale. The measured mobility ranges between 1000-4000 cm 2 /Vsec for 5 different devices we measured. The temperature dependent of resistance (S6e) shows similar behavior to that of mechanically exfoliated graphene samples with similar mobility values. (Ref. Tan, Y. W.. et al. "Temperature Dependent Electron Transport in Graphene" Eur. Phys. J. 148, 15 (2007)). www.nature.com/nature 5