Conformal Embedding of Cluster Superlattices with Carbon

Abstract

Iridium cluster superlattices on the graphene moiré with Ir(111) are embedded with elemental carbon through vapor-phase deposition. Using scanning tunneling microscopy and spectroscopy, we find that carbon embedding is conformal and does not deteriorate the excellent order of the iridium clusters. The thermal and mechanical stability of the embedded clusters is greatly enhanced. Smoluchowski ripening as well as cluster pick-up by the scanning tunneling microscopy tip are both suppressed. The only cluster decay path left takes place at an elevated temperature of around 1050 K. The cluster material penetrates through the graphene sheet, whereby it becomes bound to the underlying metal. It is argued that conformal carbon embedding is an important step towards the formation of a new type of sintering-resistant cluster lattice material for nanocatalysis and nanomagnetism.

Carbon Embedding of Pt Cluster Superlattices Templated by Hexagonal Boron Nitride on Ir(111)

“…This is supported by the SXRD data presented in Figure S1 of the Supporting Information, where an Ir cluster superlattice embedding on h-BN/Ir(111) was investigated . In conclusion, the carbon embedding of the Pt cluster lattice on h-BN/Ir(111) proceeds similar as for an Ir cluster lattice on Gr/Ir(111) investigated in ref .…”

“…Because all components of the system (substrate, 2D-layer, clusters, and embedding matrix) are composed of different elements, the chemical processes during carbon embedding can be tracked without ambiguity through XPS. The present study extends our previous combined XPS/STM investigation of binding and thermal stability for Pt/h-BN/Ir(111) and compares to our carbon embedding investigation for Ir/Gr/Ir(111) . The latter had the disadvantage of having the embedding matrix and the 2D-layer being composed of the same chemical element, C, making it difficult to disentangle the processes with XPS.…”

“…Further annealing to 1050 and 1250 K as shown in Figures c and d hardly changes the morphology, and even at the highest annealing temperature the carbon cover of the clusters stays intact. The stability of the embedding matrix is quite similar to the case of carbon embedding of Ir cluster superlattices on Gr/Ir(111) . However, STM imaging is limited to the embedding matrix, and no information about the encased Pt clusters can be inferred from these topographs.…”

“…Beyond a C coverage of 0.2 ML the peak position and width remain constant. We interpret these changes as follows: Initially, the deposited C preferentially covers the Pt clusters, similar to the C embedding of iridium clusters . These Pt–C bonds cause a rapid change of the Pt 4f level up to 0.2 ML C deposited.…”

“…Indeed, the absence of all indications for C–N and C–B bonds in the C 1s spectra after annealing is consistent with this interpretation. Also, scratching experiments with the STM tip show that the embedding matrix together with the clusters can be removed from the h-BN after heating (compare Figure S3), while this is impossible when Gr is used as a 2D layer . Thus, it appears open whether a lift-off procedure will delaminate the membrane between Ir(111) and the 2D layeras for Gror between the 2D layer and the a-C matrix with the enclosed clusters.…”