We report observation of optical bistability and enhanced thermal nonlinearity in a graphene-silicon waveguide resonator. Photo-induced Joule heating in the graphene layer gives rise to a temperature increase in the silicon waveguide core and a corresponding thermo-optic shift in the resonance of the Fabry-Perot resonator. Measurement of the nonlinear resonance spectra showed a 9-fold increase in the effective thermal nonlinear index due to the graphene layer compared with a bare silicon waveguide.

Many new materials emerging are strictly two dimensional (2D), often only one or two monolayers thick. They include transition metal dichalcogenides, such as MoS , and graphene. Graphene in particular appears to have many potential applications. Typically the crystalline film without contamination is of interest. Therefore, a reliable method is needed to routinely evaluate the quality of the synthesized samples. Here, we present one such candidate method that utilizes standard electron diffraction and low/medium magnification imaging in a rudimentary transmission electron microscope. The electron irradiation dose is very low thus reducing electron irradiation damage of the investigated samples. As an example, the method was applied to the evaluation of as-grown graphene sample quality and a study on heating-induced change in graphene. It can be used to evaluate the volume and areal ratio of crystalline to noncrystalline component. The method is amiable to automated film quality evaluation.

Graphene appears to be a material with many potential applications. A robust and convenient method is therefore needed to evaluate the quality of graphene films. Raman spectroscopy is often used and is fast and inexpensive; however, the information content in the Raman spectrum is limited. We demonstrate an electron diffraction based method that can be routinely used to evaluate the quality of graphene thin films.Graphene membranes were fabricated by transferring CVD grown graphene on to holey silicon nitride membranes [1]. TEM imaging and diffraction was performed using a Hitachi H-9500 environmental TEM operated at 100 kV. A typical low magnification image and select area electron diffraction pattern (DP) are shown in Fig. 1 a) and b). For analysis, the DPs were rotationally averaged using the DiffTools package [2] in Gatan Digital Micrograph. A sample rotationally averaged diffractogram is shown in Fig. 1 c) with peaks at 4.7 {100}, 8.1 {110}, and 9.4 nm -1 {200} [3]. After a background subtraction fit to regions beside the peak of interest, single or double Lorentzians were fit to the peak as shown in Fig. 1 d) and e). In order to perturb the quality of the graphene, the membrane was heated in the TEM from room temperature to 950 o C, with images and DPs collected in 50 o C intervals. The membranes seldom tear during the heat treatment. However, a shoulder begins to appear on the high spatial frequency (smaller lattice spacing) side of the {100} and {110} peaks in DPs at T C exceeding 150 to 350 o C, depending on the sample. The peaks are well fit with single Lorentzian below T C and with double Lorentzian above it. Inspection of the corresponding diffraction patterns in Fig. 2 a) reveals that the shoulder does not have the six-fold symmetry of the {100} peaks but is a ring close to the first order diffraction peak of graphene. The shoulder center at 5.1 nm -1 does not change with increasing temperature, as shown by the double Lorentzian fitting parameters in Fig. 2 c). Control experiments confirm that the growth of the shoulder is due to heating and that electron beam irradiation has no noticeable effect. Moreover, heating experiments in the presence of H2 at 10 -4 Pa confirm that the diffraction ring is not a result of graphene oxide formation [3]. Cooling the membrane to room temperature does not affect the observed DPs. The results are consistent with the formation of disordered centers with lattice spacing slightly smaller than graphene. One possible explanation is the rearrangement of six-membered graphene rings to five-membered rings [4,5]. The ratio of the crystalline and disordered components is evaluated by the intensity ratio of the six diffraction spots to the diffraction ring obtained using the arc function in DiffTools [2], as shown in Fig. 2 b). The ratio decreases when heating from room temperature to T C and becomes constant above T C , as shown in Fig. 2 d). Extensive hold time at 950°C does not further affect the ratio. Since both the spots and rings arise from carbon with known scattering...