Elettra | BaDElPh, SuperESCA Designer graphene: High and uniform atomic-level transition metal substitution achieved Doping graphene with foreign substitutional atoms such as transition metals can help tailor its properties for applications ranging from magnetism and spintronics to single-atom catalysis. However, substitutional incorporation of transition metal atoms in graphene is generally not energetically favorable. Ion implantation makes this possible by introducing single atoms in bulk materials as the implantation process drives the material into a non-energetically favorable system. The same approach can be applied to graphene and other two-dimensional (2D) materials, but the ion energy must be of the order of tens of eV compared to keV and MeV for bulk materials. To this end, ultralow-energy (ULE) ion implantation has been successfully used to substitutionally dope graphene with manganese (Mn). Nonetheless, a low substitution efficiency of ~1% was observed indicating that the used ion energy (40 eV) may not have been optimal. Although the substitutional concentration can in principle be increased by increasing ion fluence, it may be accompanied by the appearance of unwanted radiation defects. Achieving higher substitution efficiencies would allow for a better compromise between a high concentration of dopants and low irradiation-induced disorder. Graphene on Cu(111) was implanted with Mn+ ions at an energy of 60 eV using ULE ion implantation. Scanning tunneling microscopy (STM) was used to estimate the average surface density of substitutional Mn atoms. Compared to the surface density of C atoms in graphene, it corresponds to a substitutional Mn concentration of 0.50 ± 0.14%; and considering the implanted fluence, this amounts to a substitution efficiency (fraction of implanted ions that are substitutionally incorporated) of 13 ± 4%. Compared to Mn implantation performed at 40 eV to the same fluence as here, the implantation at 60 eV results in a 10-fold increase in substitutional concentration and substitution efficiency. This increase can be understood as due to the fact that an increase in kinetic energy of the Mn ions is associated with more efficient displacement of the C atoms, as required for substitutional incorporation to occur. We used synchrotron x-ray photoelectron spectroscopy (XPS) to further characterize the Mn incorporation as a function of annealing temperature. Fig. 1 (a) shows Mn 2p core-level data fitted with multiple Voigt components from where we determine the total Mn concentration at the surface. Although it is not possible to identify a Mn 2p component that is clearly associated with the substitutional Mn fraction identified using STM, these XPS data still provide useful insights. At the lowest annealing temperatures, approximately half of the implanted Mn ions are detected at the surface (2.8% with respect to C in graphene), indicating that the other half experienced backscattering during implantation. From the fraction that is detected at the surface, approximately 10% of the implanted Mn ions are incorporated substitutionally, based on the STM data. The remaining fraction consists of intercalated Mn atoms. The decrease in Mn detected at the surface from the 200 °C to the 700 °C annealing step can be attributed to the diffusion of intercalated Mn into the Cu substrate. After 700 °C, our XPS measurements still detect about 0.9% Mn at the surface, significantly higher than the 0.5% substitutional concentration determined using STM. Although the difference may be within the experimental uncertainty (especially considering that we are comparing XPS to STM quantification), it supports the notion that some of the non substitutional Mn remains sufficiently close to the surface to still contribute to the XPS signal. The Dirac character of graphene is preserved upon substitutional Mn doping, even in this relatively high concentration regime (0.5%), as observed using angle-resolved photoemission spectroscopy (ARPES) in Fig. 1 (b). The ARPES measurements show only a subtle effect of Mn substitution in the band structure, compared to pristine (non implanted) graphene. When considering the values for the Dirac point position, the apparent band gap, and the Fermi velocities, obtained from the analysis of the momentum distribution curves extracted from the ARPES maps, only the Fermi velocities appear to show a measurable and relatively mild reduction. All other key features of the band structure are essentially unaffected: the discontinuity in the graphene π-band in Fig. 1 (d), resulting from the hybridization with the Cu 3d-band; the Cu surface state is also clearly visible in Fig. 1 (e), indicating that no significant oxidation of the interface has occurred. ULE (60 eV) ion implantation of Mn ions into graphene on Cu(111) followed by thermal annealing in UHV at 700 °C, resulted in a relatively high substitutional efficiency (13%) and high substitutional concentration (0.50%), while keeping the graphene layer virtually as clean and defect-free as in the as-grown state, except for the incorporated substitutional Mn atoms. In particular, the Dirac character of graphene is preserved, even in this relatively high concentration regime. ULE ion implantation is shown to be well suited for the functionalization of graphene with substitutional transition-metal atoms for a wide range of applications, from magnetism and spintronics to single-atom catalysis. In particular, our data show no indication of a limit on the achievable substitutional concentration; higher doping levels can in principle be achieved simply by increasing the implanted fluence. Moreover, one can expect that other transition elements can be substitutionally incorporated following a similar approach, allowing exploration of the functionalization effects over several orders of magnitude of concentration of different elements. Elettra Highlights 2024–2025 16
RkJQdWJsaXNoZXIy MjQ4NzI=