Elettra-Sincrotrone Trieste S.C.p.A. website uses session cookies which are required for users to navigate appropriately and safely. Session cookies created by the Elettra-Sincrotrone Trieste S.C.p.A. website navigation do not affect users' privacy during their browsing experience on our website, as they do not entail processing their personal identification data. Session cookies are not permanently stored and indeed are cancelled when the connection to the Elettra-Sincrotrone Trieste S.C.p.A. website is terminated.
More info

Graphene growth by molecular beam epitaxy: an interplay between desorption, diffusion and intercalation of elemental carbon species on islands

One of the most widely used techniques to synthesize large-area flakes of high-quality graphene is CVD (Chemical Vapour Deposition) growth, which is based on the decomposition of hydrocarbon molecules on a metal surface to provide the carbon feedstock. Its growth by MBE (Molecular Beam Epitaxy) from a solid carbon source, on the other hand, has shown great promise to overcome some of the major drawbacks of the CVD technique, for example the possibility to synthesize graphene directly on various surfaces including semiconductors and insulators. This technique is based on the sublimation of C species (mainly monomers and dimers) from a heated graphite rod.

While the individual steps of the CVD growth process have been already studied for several surfaces, such knowledge is still lacking for the case of MBE, even though it is a key ingredient to optimise its performance and effectiveness. In this work, we have performed a combined experimental and theoretical study comparing the growth mechanisms of the MBE and CVD processes on the Ir(111) single-crystal surface, a prototypical surface where the CVD growth of graphene has already been well characterised.
By employing high-resolution fast XPS at the SuperESCA beamline, we were able to follow the growth of both single- and multi-layer graphene in real time (Fig. 1a), and to identify the spectroscopic fingerprints allowing to distinguish graphene with a different number of layers (Fig. 1b). In addition, by comparing the rate of the graphene growth using CVD and MBE (Fig. 1c), we show that the latter technique can be employed to grow either a high-quality single- or multi-layers of graphene, in a controlled way.

Figure 1.  a: Real-time evolution of the C1s XPS spectrum during graphene growth, represented in colour scale (dark represents higher photoemission intensity, and therefore a higher C coverage). b: Selected C1s spectra corresponding to different stagesof single- and multi-layer graphenegrowth. The two components observed are originated by the first (A) and additional (B) C layers. c: Evolution of the growth rate of single- and multi-layer graphene as a function of C feedstock exposure in MBE, compared to the CVD growth.

Moreover, we demonstrate that the differences between these techniques can be ascribed to the much stronger interaction with graphene of the C species found in MBE with respect to the hydrocarbons used in CVD. Thanks to Density Functional Theory calculations -performed at King's College London, UK- of the desorption and diffusion barriers of C species on graphene (Fig. 2a), we demonstrate that in MBE there is a tight competition between processes occurring not only on the metal substrate, but also on the already nucleated, still growing graphene islands, as schematically shown in Fig. 2b. These concurrent processes are the diffusion and the desorption of the just-deposited carbon feedstock on the growing graphene flakes, and it is controlled both by the size of the flakes and by the composition of the feedstock. In fact, in MBE, the carbon species sticking both on the still bare metal surface and on the growing islands can participate to the growth, by diffusing towards the edges of such islands. However, the C feedstock deposited too far from these edges desorbs before having the opportunity to attach to them. The attachment of the C species to the edges of already nucleated islands allows their ordered growth with a very high crystallographic quality and is therefore a key component for the growth of graphene with a low density of defects. This is not the case instead for CVD, where only processes occurring on the uncovered regions of metal surface lead to the island’s accretion.

Figure 2 a: Adsorption configuration and energy of C monomers and dimers on graphene, as calculated by DFT. b: Schematic representation of the MBE growth process dynamics, as demonstrated by our results.

Our experiments and calculations highlight the role of the interaction between different C precursor species and the growing graphene flakes on the growth rate of graphene, providing an overview of the main differences between CVD and MBE growth and thus on the main parameters which can be tuned to optimise growth conditions.


This research was conducted by the following research team:

F. Presel1, Holly Tetlow2, Luca Bignardi3, Paolo Lacovig3, Cristian Alexandru Tache1, Silvano Lizzit3, Lev Kantorovich2, Alessandro Baraldi1,3,4

Physics Department, University of Trieste, Trieste, Italy
Physics Department, King's College London, London, UK 
Elettra - Sincrotrone Trieste S.C.p.A., Trieste, Italy
CNR-IOM, Laboratorio TASC, Trieste, Italy

Contact persons:

Alessandro Baraldi, email: baraldi@elettra.eu


F. Presel, Holly Tetlow, Luca Bignardi, Paolo Lacovig, Cristian Alexandru Tache, Silvano Lizzit, Lev Kantorovich, Alessandro Baraldi, "Graphene growth by molecular beam epitaxy: an interplay between desorption, diffusion and intercalation of elemental C species on islands", Nanoscale 10, 7396 (2018); DOI:10.1039/C8NR00615F

This work has been selected for the Cover of Nanoscale.

Last Updated on Monday, 30 July 2018 13:24