Scalable bottom-up synthesis of Co-Ni-doped graphene

Abstract

Introducing heteroatoms into graphene is a powerful strategy to modulate its catalytic, electronic, and magnetic properties. At variance with the cases of nitrogen (N)- and boron (B)-doped graphene, a scalable method for incorporating transition metal atoms in the carbon (C) mesh is currently lacking, limiting the applicative interest of model system studies. This work presents a during-growth synthesis enabling the incorporation of cobalt (Co) alongside nickel (Ni) atoms in graphene on a Ni(111) substrate. Single atoms are covalently stabilized within graphene double vacancies, with a Co load ranging from 0.07 to 0.22% relative to C atoms, controllable by synthesis parameters. Structural characterization involves variable-temperature scanning tunneling microscopy and ab initio calculations. The Co- and Ni-codoped layer is transferred onto a transmission electron microscopy grid, confirming stability through scanning transmission electron microscopy and electron energy loss spectroscopy. This method holds promise for applications in spintronics, gas sensing, electrochemistry and catalysis, and potential extension to graphene incorporation of similar metals.

Fig. 1.. During-growth incorporation of Co and Ni into Gr on Ni(111) substrate.

Details of the used recipe are sketched in the diagram of the substrate temperature versus time (A). The growth mechanism, schematically represented in (B), involves Co and Ni adatoms diffusing across the substrate surface before being incorporated in the growing edge of the Gr layer.

Fig. 2.. STM, DFT, and XPS characterization of the Co-Ni–doped Gr layer.

The large-scale room temperature STM image in (A) shows the abundance of Ni versus Co dopants; the latter marked by green arrows. Three possible orientations of the Co dopants exist [see models in (B), where the C atoms coordinated to Co are colored in red]; two of them are visible in the image (A)—indicated by the arrow directions—and can be distinguished by their different appearance. Panel (C) shows a comparison between experimental and simulated STM images of Co and Ni atoms incorporated in the Gr mesh (V_b = −0.1 V, size 2 × 1.8 nm²). In (D), the related models in the top and side views are displayed [corresponding for Co to the configuration shown in (B) in the middle]. Panel (E) reports XPS spectra of the Co 2p core level at room temperature (R.T.) and after annealing at 400°C, respectively. Tunneling parameters: (A) V_b = −0.15 V, I_t = 1.0 nA, [(C), Co] V_b = −0.1 V, I_t = 1.7 nA, [(C), Ni] V_b = −0.2 V, I_t = 1.7 nA.

Fig. 3.. Substitutional Ni and Co dopants in Gr characterized by STEM and EELS after transferring the layer onto a TEM grid.

The transfer process is schematically displayed in (A), illustrating the initial stage on the Ni substrate, the delamination through bubbling, and the transfer onto the Quantifoil Au grid stack. STEM–medium angle annular dark-field (MAADF) images of two different regions of ML Gr, containing respectively one substitutional Ni atom [(B), blue arrow] and one substitutional Co atom [(C), green arrow]. As seen in the images, the Gr lattice also contains other defects (white arrows), such as vacancies and lighter impurities, mainly Si. (D) EELS point spectra for single Ni and Co atoms are shown in blue and green, respectively. The spectra were scaled to have a similar magnitude for the peaks. The black spectrum was recorded as a reference from a metal cluster containing both Co and Ni atoms (see Supplementary Materials for the corresponding STEM-MAADF image).