6 7 I H I G H L I G H T S 2 0 2 1
to optimise the CVD process on the liquid phase, but the lack of precise, multi-scale in-situ monitoring techniques enabling direct feedback on the growth parameters (including temperature, gas composition and pressures) has led primarily to empirical recipes. Such recipes intrinsically suffer from a limited understanding of the graphene formation process and low reproducibility of the product due to the complex and stochastic nature of the growth phenomena.
Recent publications report the development  and successful implementation  of four in-situ techniques for multi-scale monitoring of graphene growth on liquid copper at 1370 K and under atmospheric-pressure CVD conditions. The LMCat reactor at beamline ID10 enables the tailoring of the crystal size, shape, and quality of graphene while optimising growth speeds (Figure 52a). Radiation-mode optical microscopy provides essential information on growth morphology and dynamics in real- time at macroscopic scales. In-situ Raman spectroscopy confirms the presence of monolayer graphene and yields information about its crystallinity and defects from mesoscopic to nano scales. At the atomic scale, the lattice constant and corrugation of graphene sheets floating on liquid copper are derived from the positions and angular spread of the Bragg rods measured by grazing-incidence X-ray diffraction. The number of graphene layers, roughness, and the separation between graphene and
liquid copper are provided by synchrotron X-ray reflectivity (Figure 52d). Multi-scale simulations were used to analyse the experimental results [4,5].
To demonstrate the wealth of information and control capability that can be achieved by multi-scale in- situ monitoring, CVD growth processes for which the nucleation of graphene seeds is induced by an injection of a short pulse of methane at high partial pressure were investigated. This procedure produces many flakes that form a super-ordered assembly due to short-range electrostatic and long-range capillary interactions (Figure 52b). Simulations reproduced the observed assembly of flakes into a 2D hexagonal network on liquid copper. Such spontaneous ordering is not possible on a solid catalyst due to the immobility of the flakes on a solid surface. Ultimately, the flakes merged into a continuous film; however, slight misorientations of neighbouring flakes remained upon their coalescence, leaving domain boundaries where they merged. Next, monitoring and feedback-control was used to improve the ordering of the flakes and reduce the remnant defects upon merging. Finally, the growth parameters were tailored to nucleate only a single flake and grow it to millimetre size (Figure 52c). The spectra obtained using X-ray scattering and Raman spectroscopy compare well to those of single- layer exfoliated graphene .
Fig. 52: a) Configuration of in-situ monitoring methods applied to a graphene layer grown on liquid copper. b) An example of in-situ radiation- mode optical microscopy of self-organised hexagonal graphene flakes on liquid copper. c) One single-crystal flake growing to millimetre size . d) In-situ XRR data for liquid copper with and without graphene overlay.