G/Rh(111)

G/Rh(111) can be grown by CVD with high quality. A precise selection of preparation parameters is necessary, since carbidic phases can be formed and dissolution of C in the Rh surface and eventually segregation were found to play a crucial role for the quality of G/Rh(111) (see also [190]). The samples presented here are prepared according to the recipe described in section 4.4.2.

The quality of the graphene is judged based on LEED [Figure 5.3]. For high quality graphene preparations diffraction spots in LEED suggest a parallel alignment of graphene and substrate lattice (~k_Gand~k_Rh). Moiré superstructure spots are sharp and several or-ders are observed surrounding the substrate and graphene spots indicative of large single crystal domains with identical orientation. The moire periodicity measured from the

dis-Figure 5.3 | Graphene on Rh(111). LEED image of graphene moiré on Rh(111). The graphene and substrate lattices are parallel giv-ing rise to aR0^◦moiré. LEED experiment by O.

Zander. ^{58 eV}

(a)

km

k_G kRh

5.1 Graphene on iridium and rhodium substrates 63

(a)

0 2 4 6 8

0 0.2 0.4 0.6 0.8

Lateral position (nm)

Height (Å)

A (a) (b)

A

1 nm

(a)

(c)

ring-atop ring-hcp ring-fcc ring-bridge

Rh 1st Rh 3rd Rh 2nd C Rh [110]

C [1120]

Rh [112]

Figure 5.4| Structure ofR0^◦G/Rh(111). (a)Atomically resolved STM topography of the moiré. The white rhombus outlines the moiré supercell.(b)Apparent height profile along the Rh[112]direction crossing the diagonal of the moiré supercell. (c) Ball model of a 12C/11Rh moiré superstructure.

Scanning parameters: 9_×9 nm²,V₌0.47 V,I₌4.2 nA,T₌300 K.

tances between superstructure spots (~km) amounts toam =(2.90±0.05) nm compared to an expected value ofa_m=2.88 nm derived according to equation 2.21 witha_C=2.46 Å [35] anda_Rh=2.69 Å [109].

Atomic structure, topographic corrugation and bonding

The atomically resolved STM topography of G/Rh(111) is shown in Figure 5.4 (a). The im-age shows the moiré supercell delimited by bright protrusions at the corners marked by dashed circles. Very pronounced features within the cell are five depressions, one in the middle of the cell and four at the cell borders marked by a small full circle. Three adjacent depressions constitute a triangle which gives the characteristic appearance of the moiré on larger scales. The two high-symmetry positions (marked with a star and a square) within the moiré cell enclosed by a triangular arrangement of the dark depressions each, are inequivalent in height as visible in the line profile in Figure 5.4 (b). Hence, the bright protrusions can be assigned to the ring-atoppositions, whereas the two high-symmetry points enclosed by a triangular arrangement of dark depressions are the ring-fcc and ring-hcppositions. According to the 12C/11Rh commensurate ball model in Figure 5.4

(c) the depressions at the moiré cell borders correspond to ring-bridgepositions, where the carbon ring is placed betweenfccandhcphollow sites. The brighter one of the two sites enclosed by the ring-bridgepositions within the moiré cell was assigned to the ring-fccarrangement according to DFT calculations [112].

The spatial modulation of the apparent STM height across the unit cell depends on the scanning parameters and was found to vary between 0.5 Å and 1.5 Å. For the origin of the local variation in apparent height both real topographic corrugation and LDOS variations have to be considered, both not unambiguously separable in STM. The line profile in Figure 5.4 (b) of G/Rh(111) is found to follow the topographic buckling of the graphene layer known from DFT [112, 88] largely and locally varying DOS contributions appear to be less important for the STM appearance compared to the real topographic buckling. The appearance in STM predominantly shows bright ring-atopcontrast, sug-gesting a considerable variation in the local bonding of the graphene layer, since the lat-ter leads to a large topographic buckling of the layer. This is in contrast to the findings for G/Ir(111), where the topographic buckling of the layer is small and the apparent height in STM images frequently shows inverted contrast.

Compared to G/Ir(111), the structure of the moiré on Rh(111) shows a more intri-cate substructure and suggests the bridging positions to be of special interest, see Figure 5.4 (b). Indeed, ring-bridgesites were shown to be energetically preferred on G/Rh(111) [112] and G/Ru(0001) [119] and yield relatively strong graphene-substrate bonding. They represent the closest adsorption site in G/Rh(111). The moiré corrugation found in DFT is large ranging from 1.1 Å and 1.6 Å with differences arising from the type of imple-mented DFT functional [112, 88].

Comparison and Discussion

The presented graphene on metal systems represent candidates for weakly interacting and strongly interacting systems. Both systems show a comparable lattice mismatch of roughly 10%. Since the degree of interaction drives the amount of real topographic buck-ling of the graphene layer, LDOS modulations are more important in flatter G/Ir(111), ev-ident in inverted imaging contrasts compared to real topography. The different degrees of interaction correlate with PES experiments reported in literature, which testify a Dirac cone which is largely preserved in G/Ir [125, 28] compared to a destroyed Dirac cone and severely shiftedπ-band in G/Rh [191]. A special importance of the ring-bridgeregions in G/Rh is found evident in a decreased graphene-substrate distance compared to sur-rounding atoms in difference to systems such as G/Ru(0001)[192] and G/Ir(111) [110].

5.2 Intercalation 65

5.2 Intercalation

The following section will show that pre-deposited metal atoms can be inserted into the interface between graphene and substrate upon annealing, referred to as intercalation [127, 32]. The term originates from graphite intercalation compounds, where a certain material is inserted between the weakly bonded graphene planes of graphite eventually forming a regular sequence of intercalant-graphite layers [127]. Later, the term interca-lation was extended to graphite monolayers on metals, where intercainterca-lation refers to the insertion of material into the graphene-substrate interface [32], which corresponds to the intercalation term used in this work.

Intercalation experiments are performed as sketched in Figure 5.5. Full graphene monolayers on either Rh(111) or Ir(111) metal surfaces are prepared using CVD growth resulting in mainlyR0° oriented graphene on Rh(111) or Ir(111) as discussed in section 5.1. After deposition of the metal, the sample is annealed between 400–600^◦C as speci-fied in the text and investigated by STM after cool down.

Ir(111)

Ni deposition

G/Rh(111) Annealing G/Ni/Rh(111)

Figure 5.5 |Schematic representation of the intercalation experiment. (a)A full monolayer of graphene on metal surfaces is prepared by CVD. (b) The metal is deposited commonly at room temperature.(c)The sample is annealed at temperatures between 400–600 °C.(d)The intercalation sample is investigated in STM at room temperature.