top of page
Locked
applscilettersa>vol-01>issue>04>Stacking-Mediated Diffusion of Ruthenium Nanoclusters in Graphite
Research
Stacking-Mediated Diffusion of Ruthenium Nanoclusters in Graphite
James G. McHugh
Pavlos Mouratidisy
Abstract
The diffusion, penetration, and intercalation of metallic atomic dopants is an important question for various graphite applications in engineering and nanotechnology. We have performed systematic first-principles calculations of the behaviour of ruthenium nanoclusters on a graphene monolayer and intercalated them into a bilayer. Our computational results show that at a scientifically high density of single Ru atom interstitials, intercalated atoms can shear the surrounding lattice to an AA stacking confiduration, an effect which weakens with increasing cluster size. Moreover, the interlayer stacking configuration, in turn, has a signcant affect cluster diffusion. We therefore nd different trends in diffusivity as a function of cluster size and interlayer stacking. For monolayer graphene and an AA graphene bilayer, the formation of small clusters generally lowers diffusion barriers, while the opposite behaviour is found for the preferred AB stacking conduration. These results demonstrate that conditions of local impurity concentration and interlayer disregistry are able to regulate the diffusivity of metallic impurities in graphite.
Introduction
The energetic and dynamical properties of transition metal impurities adsorbed on top of graphene and intercalated between the layers of graphite is a recurring topic of considerable interest in materials science. Foreign elemental impurities are one of the most promising ways to modify the physical properties of pristine graphene [1], and they hold considerable promise in engineering desirable electronic phases [2]. For example, transition-metal doping of monolayer graphene (MLG) and bilayer graphene (BLG) can be exploited to increase the weak intrinsic spin-orbit effects of the native carbon atoms [3{5], allowing the engineering of novel quantum states with advantageous transport properties and prospective applications as topological insulators or in quantum computing [6{8]. Intercalated metallic species are also of interest in their own right, and the layered structure of van der Waals materials, such as graphite, provides an excellent platform to grow quasi-two-dimensional sheets of selected transition metals [9{12]. These two-dimensional transition metal sheets have many desirable properties, which are greatly enhanced by their lower-dimensional topology. Due to the higher (2D) bulk to surface ratio and the associated change in coordination, layered metals can completely change electronic properties such as band gaps and transport properties [13, 14]. Two-dimensional metallic layers have been grown underneath the top monolayer of highly-oriented pyrolytic graphite (HOPG) [11, 15], a process which is known to proceed via the diffusion of impurities through lattice defect "entry portals" (monovacancy and multi-vacancy complexes) [12, 16], and vacancy sites are also known to promote the intercalation of other elemental species such as Cs and Dy [17, 18]. Transition metals are also important from the perspective of graphite applications. Some of the most important nuclear fusion products are transition metals [19], and the penetration of these impurities into the bulk and subsequent diffusion through the graphite lattice is a pressing problem in the design of new reactors and in the assessment of the safety and decommissioning of retired reactors [20{22].
Conclusion
In this work, we have performed extensive simulations of the adsorption, intercalation, clustering and diffusion of ruthenium nanoparticles in graphite, which we nd has a pronounced impact on structural, energetic and dynamical properties. Most importantly, we observe that interstitial Ru has the capability of shifting the surrounding graphene layers to an AA stacking configuration. This is a particularly interesting effect as it is accompanied by a substantial increase in the transition barrier, which severely impedes diffusion relative to other cluster sizes and stacking configurations. In addition, we nd that penetration into the bulk of graphite is energetically preferably for all Ru nanoclusters sizes, however, this tendency becomes much weaker with increasing cluster size, as shown in Fig. 10 (b), which shows the segregation energy (difference in adsorption and intercalation energies) vs cluster size extracted from our DFT calculations. We conclude that large clusters are more likely to become trapped at the surface and in porous regions of more disordered graphite grades. Only smaller clusters, which have a larger thermodynamic incentive to intercalate, can proceed via defective entry portals into the bulk. Thus, the decreasing thermodynamic preference for intercalation with increasing cluster size necessitates cluster break-up as an active part of the intercalation process. Notably, it has been found recently that the dynamical process underlying the penetration of metal intercalants into the bulk is mediated by a low energy barrier bond-breaking process of the metal dimer at defect sites [39]. This can eectively lter clustered surface fragments into isolated single atoms, which then proceed to intercalate into graphite. In combination with our observation of higher barriers for single interstitial Ru, and the general preference for these isolated ruthenium atoms for defected regions such as vacancies and basal dislocations, this suggests that the high temperature breakup of clusters into single Ru atoms at entry sites will impede overall diffusion into the bulk and throughout the microstructure. In general, the modelling of the penetration, diffusion and interaction of ruthenium atoms and clusters with graphite is a challenging one which is vital to many applications and prospective uses of graphite. A fuller understanding of these issues requires larger cells and longer timescales than those which are currently feasibly using.
bottom of page