Layer-resolved photoemission tomography: The $p$-sexiphenyl bilayer upon Cs doping

The buried interface between a molecular thin film and the metal substrate is generally not accessible to the photoemission experiment. With the example of a sexiphenyl ($6\mathrm{P}$) bilayer on Cu we show that photoemission tomography can be used to study the electronic level alignment and geometric structure, where it was possible to assign the observed orbital emissions to the individual layers. We further study the Cs doping of this bilayer. Initial Cs exposure leads to a doping of only the first interface layer, leaving the second layer unaffected except for a large energy shift. This result shows that it is in principle possible to chemically modify just the interface, which is important to issues like tuning of the energy level alignment and charge transfer to the interface layer. Upon saturating the film with Cs, photoemission tomography shows a complete doping ($6p^{4-}$) of the bilayer, with the molecular geometry changing such that the spectra become dominated by $\sigma$-orbital emissions.

Figure 1

He I ARUPS spectra of the $6\mathrm{P}$ bilayer on Cu(110) at increasing Cs exposure from (a) normal emission ($0^{\circ }$ photoelectron takeoff angle) and (b) ${50}^{\circ }$ photoelectron takeoff angle, referenced to the Fermi-level energy $E_F$. For each Cs dosing step the measured work function $\varphi$ and the Cs dosing unit (DU) is indicated.

Figure 2

Schematic of a $6\mathrm{P}$ bilayer on Cu(110) in side (a) and top (b) view. Molecules in the first layer are in white, second-layer molecules are yellow. Azimuthal directions of the Cu(110) substrate are given. In (a), tilt and twist around the molecular axis are indicated.

Figure 3

ARUPS data of the $6\mathrm{P}$ bilayer on Cu(110) ($h\nu =35$ eV, $\alpha ={40}^{\circ }$). In the band map (a) in the $\left[1\overline{1}0\right]$ direction the energies of the hybridized LUMO (L) and HOMO (H) of the first layer, and the HOMO and nonbonding orbitals [20] (HOMO-3 to HOMO-8) of the second layer are indicated. The energy region between Fermi edge and 2 eV binding energy has been reproduced with increased intensity and inverted grayscale. The inset spectrum (white line) has been extracted from $k_{\left[1\overline{1}0\right]}=1.4{\AA }^{-1}$. The $k$ maps (b)–(d) of the LUMO and HOMOs are compared to simulations (e)–(g) of these orbitals for the corresponding molecular conformation (first-layer molecules planar and flat, second-layer molecules twisted and tilted). The direction of the band map is indicated for each $k$ map.

Figure 4

ARUPS data of the $6\mathrm{P}$ bilayer at initial doping ($h\nu =35\mathrm{eV},\alpha ={40}^{\circ }$). In the band map (a) the energy positions of the first-layer LUMO, and the second-layer HOMO and nonbonding orbitals (H-3 to H-8) are labeled. The assumed energy of the first-layer HOMO is marked with a white dotted line. The energy region between Fermi edge and 2 eV binding energy has been reproduced with increased intensity. The inset spectrum has been extracted from $k_{\left[1\overline{1}0\right]}=1.4{\AA }^{-1}$. The experimental momentum maps of the first-layer LUMO (b) and the second-layer HOMO (c) have been reproduced by simulated momentum maps (d), (e) with the respective geometric conformations. The direction of the band map is indicated for each $k$ map.

Figure 5

ARUPS data of the quadruply doped $6p^{4-}$ bilayer ($h\nu =35\mathrm{eV},\alpha ={40}^{\circ }$). In the band map (a) the $\mathrm{LUMO}+1$, LUMO, and HOMO energies as well as the top of an emerging $\sigma$ band are labeled. The inset spectrum has been extracted from $k_{\left[1\overline{1}0\right]}=1.4{\AA }^{-1}$. The measured momentum maps of the $\mathrm{LUMO}+1$ (b), LUMO (c), and HOMO (d) are reproduced by the corresponding simulated momentum maps (e) to (g) for molecules with $\pm {22}^{\circ }$ and $\pm {75}^{\circ }$ tilt angle. The direction of the band map is indicated for each $k$ map.

Figure 6

Simulated GGA band maps along the long molecular axes for isolated, planar $6\mathrm{P}$ molecules forming a charge transfer complex with four Cs atoms for $0^{\circ }, \pm {25}^{\circ }, \pm {50}^{\circ }$, and $\pm {75}^{\circ }$ tilt angle around the long molecular axis with respect to the substrate. The most significant $\pi$-orbital emissions $\mathrm{LUMO}+1$ (L+1), LUMO (L), HOMO (H), nonbonding orbitals (H-3 to H-8), and the top of the $\sigma$ band are indicated. The energies were aligned to the experimental $\mathrm{LUMO}+1$ binding energy.

Figure 7

Top: energy diagrams of the characteristic parameters of the $6\mathrm{P}$ bilayer in the undoped, doubly doped ($6p^{2-}$), and quadruply doped ($6p^{4-}$) bilayer. Indicated are orbital binding energies for both first and second layer (with respect to the Fermi edge), work function $\mathrm{\Phi }$, HOMO-LUMO (transport) gap (full and dashed gray lines). The energies have been yielded by home-laboratory ARUPS measurements at room temperature. Bottom: suggested schematic of the geometric molecular configuration of the bilayer for the three investigated doping cases. The structure for the doped layers has been derived from ARUPS data, the model for the undoped case supported by LT-STM [30].