Abstract:In order to progress from the lab to commercial applications it will be necessary to develop industrially scalable methods to produce large quantities of defect-free graphene.Here we show that high-shear mixing of graphite in suitable, stabilizing liquids results in large-scale exfoliation to give dispersions of graphene nanosheets. XPS and Raman spectroscopy show the exfoliated flakes to be unoxidised and free of basal plane defects. We have developed a simple model which shows exfoliation to occur once the local shear rate exceeds 10 4 s -1 . By fully characterizing the scaling behaviour of the graphene production rate, we show that exfoliation can be achieved in liquid volumes from 100s of ml up to 100s of litres and beyond. The graphene produced by this method performs well in applications from composites to conductive coatings. This method can be applied to exfoliate BN, MoS2 and a range of other layered crystals. Main Text:Due to its ultra-thin, 2-dimensional nature and its unprecedented combination of physical properties, graphene has become the most studied of all nano-materials. In the next decade graphene is likely to find commercial applications in many areas from high-frequency electronics to smart coatings.

In order to fulfil their potential for applications, it will be necessary to develop large-scale production methods for two-dimensional (2D) inorganic nanosheets. Here we demonstrate the large-scale shear-exfoliation of molybdenum disulphide nanosheets in aqueous surfactant solution using a kitchen blender. Using standard procedures, we measure how the MoS2 concentration and production rate scale with processing parameters. However, we also use recently developed methods based on optical spectroscopy to simultaneously measure both nanosheet lateral size and thickness, allowing us to also study the dependence of nanosheet dimensions on processing parameters. We found the nanosheet concentration and production rates to depend sensitively on the mixing parameters (the MoS2 concentration, Ci; the mixing time, t; the liquid volume, V; and the rotor speed, N). By optimising mixing parameters, we achieved concentrations and production rates as high as 0.4 mg/ml and 1.3 mg/min respectively. Conversely, the nanosheet size and thickness were largely invariant with these parameters. The nanosheet concentration is also extremely sensitive to the surfactant concentration. However, more interestingly the nanosheet lateral size and thickness also varied strongly with the surfactant concentration. This allows the mean nanosheet dimensions to be controlled during shear exfoliation at least in the range ~40-220 nm for length and ~2-12 layers for thickness. We demonstrate the importance of this by showing that the MoS2 nanosheets prepared using different surfactant concentrations, and so displaying different nanosheets sizes, perform differently when used as hydrogen evolution catalysts. We find the nanosheets produced using high surfactant concentrations, which gives smaller flake sizes, perform significantly better, consistent with catalysis occurring at nanosheet edges. Finally, we also demonstrate that shear exfoliation using a kitchen blender is not limited to MoS2 but can also be achieved for boron nitride and tungsten disulphide. ToC fig3

Liquid phase exfoliation is a powerful and scalable technique to produce defect-free mono- and few-layer graphene. However, samples are typically polydisperse and control over size and thickness is challenging. Notably, high throughput techniques to measure size and thickness are lacking. In this work, we have measured the extinction, absorption, scattering and Raman spectra for liquid phase exfoliated graphene nanosheets of various lateral sizes (90 ≤ 〈L〉 ≤ 810 nm) and thicknesses (2.7 ≤ 〈N〉 ≤ 10.4). We found all spectra to show well-defined dependences on nanosheet dimensions. Measurements of extinction and absorption spectra of nanosheet dispersions showed both peak position and spectral shape to vary with nanosheet thickness in a manner consistent with theoretical calculations. This allows the development of empirical metrics to extract the mean thickness of liquid dispersed nanosheets from an extinction (or absorption) spectrum. While the scattering spectra depended on nanosheet length, poor signal to noise ratios made the resultant length metric unreliable. By analyzing Raman spectra measured on graphene nanosheet networks, we found both the D/G intensity ratio and the width of the G-band to scale with mean nanosheet length allowing us to establish quantitative relationships. In addition, we elucidate the variation of 2D/G band intensities and 2D-band shape with the mean nanosheet thickness, allowing us to establish quantitative metrics for mean nanosheet thickness from Raman spectra.