Probing the Bonding and Electronic Structure of Single Atom Dopants in Graphene with Electron Energy Loss Spectroscopy

Abstract

A combination of scanning transmission electron microscopy, electron energy loss spectroscopy, and ab initio calculations reveal striking electronic structure differences between two distinct single substitutional Si defect geometries in graphene. Optimised acquisition conditions allow for exceptional signal-to-noise levels in the spectroscopic data. The near-edge fine structure can be compared with great accuracy to simulations and reveal either an sp(3)-like configuration for a trivalent Si or a more complicated hybridized structure for a tetravalent Si impurity.

Correlative electron and X-ray microscopy: probing chemistry and bonding with high spatial resolution

“…Using a model system of Gd-metallofullerene molecules trapped within a single walled nanotube (SWNT), individual Gd atoms were mapped using the Gd N edge, though counting 40 statistics were understandably low. 22 In a more recent study, Ramasse et al 23 were able to obtain EEL spectra over individual substitutional Si atoms in graphene, and observed changes in fine structure for different defect geometries (figure 3 d,e). Such single-atom EELS studies are often limited by sample instabilities 45 and electron beam damage, even in these relatively ideal systems.…”

“…In the metallofullerene experiment, Gd atoms were observed to move within their fullerene cages within the short (35 ms) acquisition times, 22 and the silicon dopants in the graphene experiment were occasionally observed to jump to neighbouring 50 sites. 23 As single-atom spectroscopy is limited by these sample instabilities, an increase in probe brightness or acquisition time alone may not be suitable for achieving better signal-to-noise ratios (SNR) or for investigating other elements with lower scattering cross-sections. Instead, improvements in detector 55 sensitivity are required for single-atom measurements to become more routinely achievable in EELS.…”

Electronic Structure Modification of Ion Implanted Graphene: The Spectroscopic Signatures of p- and n-Type Doping

“…1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 B 4 C, 49 or single6layer boron nitride samples, 45,[49][50][51] only a suppressed feature is discernible at 189 eV where one might normally expect a π* peak, complemented on the low6energy side by a weak shoulder at 188 eV. To confirm this highly unusual profile, additional high signal6 to6noise6ratio B K spectra were acquired, by accumulating the EELS signal while using a small subscan window around the substitutional B atom, similar to the method used by Ramasse et al 33 The corresponding spectra are again entirely similar to those recorded via spectrum imaging: see Figure …”

“…31 This ion6implantation technique, commonly used by the modern semiconductor industry for doping Si wafers, for instance, has the advantage of allowing the uniform incorporation over a large area of single dopants on a pre6screened, single6layer, suspended graphene sample, and of producing comparatively few defects or ad6atom configurations. 29 Recent progress in the application of Scanning Transmission Electron Microscopy (STEM) based spectroscopy to the study of 26dimensional materials has demonstrated the technique's ability to fingerprint single dopant atoms in graphene [32][33][34][35] and to differentiate between different electronic structure configurations, such as trivalent and tetravalent single atom Si impurities using subtle changes in the near edge fine structure of the Si L 2,3 ionisation edge in electron energy loss spectroscopy (EELS). 33,35 In this type of study ab initio calculations are essential tools in rationalising the experimental observations and in providing further insight into the nature of bonding around the foreign species; such a combined STEM6EELS and ab initio calculations approach was also used recently to probe the bonding of single nitrogen atoms in graphene 36,37 and N6doped single6walled carbon nanotubes.…”

“…While the discrepancy with N K edges observed in SWCNTs could be attributed to curvature 44 and geometry effects 45 (it should be noted that in the modelling studies in this work, non6planar dopant geometries were rejected as less thermodynamically favourable structures), it is more difficult to pinpoint the exact reason for the discrepancy with the N6doped graphene system. An obvious difference may lie in the fact that Warner et al 33 carried out their observations at elevated temperature. Nevertheless, here, the experiments were repeated multiple times, on several distinct N atoms, systematically yielding a clearly6discernible π* peak.…”