Growth and electronic structure of nitrogen-doped graphene on Ni(111)

We report on experimental and theoretical investigations of nitrogen-doped graphene. The incorporation of nitrogen was achieved during chemical-vapor deposition on Ni(111) using pyridine as a precursor. The obtained graphene layers were investigated using photoelectron spectroscopy. By studying C $1s$ and N $1s$ core levels, we show that the nitrogen content is influenced by the growth temperature and determine the atomic arrangement of the nitrogen atoms. Valence-band photoelectron spectra show that the incorporation of nitrogen leads to a broadening of the photoemission lines and a shift of the $\pi$ band. Density functional calculations for two possible geometric arrangements, the substitution of carbon atoms by nitrogen and vacancies in the graphene sheet with pyridinic nitrogen at the edges, reveal that the two arrangements have opposite effects on the band structure. For the present experimental approach, vacancies with pyridinic nitrogen are dominant. In the latter case the vacancies generated by the nitrogen doping, not the nitrogen itself, have the main effect on the band structure. By intercalating gold between the doped graphene layer and the Ni(111) substrate electronic decoupling is achieved. After intercalation the doping remains.

Figure 1

The two different structures for nitrogen (blue) incorporated in a graphene sheet considered in the present work: (a) substitutional sites and (b) pyridinic sites. The latter consist of a nitrogen atom accompanied by an adjacent vacancy.

Figure 2

Differentiated photoelectron intensity maps vs. binding energy and parallel momentum of graphene grown by pyridine CVD at (a) $800{}^{\circ }$C and (b) $580{}^{\circ }$C, resulting in nitrogen contents of 0.00 ML and 0.02 ML, respectively. (c) and (d) display data taken from the samples shown in (a) and (b) after gold intercalation. The photon energy was $h\nu =40$ eV. (e) and (f) band structures from DF calculations (black lines) superimposed on the photoemission data of (a) and (c), nitrogen-free graphene on Ni(111) and Au/Ni(111), respectively. Note that in (f), for a better comparison with the experimental results, the band structure calculated in a $p(2\times 2)$ supercell is unfolded to be comparable to band structures referring to the $p(1\times 1)$ unit cell. Yellow and orange dots denote the carbon $p_z$ and $p_{x,y}$ contributions (corresponding to the dot size) for the given band, respectively.

Figure 3

High-resolution x-ray photoelectron spectra of the intercalation of nitrogen-doped graphene on Ni(111). Shown are the (a) C $1s$, (b) N $1s$, and (c) Ni $2p$ regions.

Figure 4

Energy distribution curves (EDCs) of different ARPES spectra at the $\Gamma$ point. (a)–(c) Spectra from preparations at different temperatures. (d)–(f) Gold-intercalated graphene layers.

Figure 5

Binding energy of the bottom of the $\pi$ band of N-doped graphene on Ni(111) and Au/Ni(111) as a function of the concentration of pyridinic nitrogen (N${}_{\mathrm{py}}$, left side) and substitutional nitrogen (N${}_s$, right side). The position of the Dirac point is included for the Au/Ni(111) case as well. Note that the Dirac cone is opened in our calculations of N-doped graphene independently of the substrate or specific doping geometry due to the reduction of symmetry upon doping. Therefore, in these cases $E_D$ is set equal to the midpoint of the arising band gap at $K$.