Effect of Interlayer Coupling on Ultrafast Charge Transfer from Semiconducting Molecules to Mono- and Bilayer Graphene

Graphene is used as flexible electrodes in various optoelectronic devices. In these applications, ultrafast charge transfer from semiconducting light absorbers to graphene can impact the overall device performance. Here, we propose a mechanism in which the charge-transfer rate can be controlled by varying the number of graphene layers and their stacking. Using an organic semiconducting molecule as a light absorber, the charge-transfer rate to graphene is measured by using time-resolved photoemission spectroscopy. Compared to graphite, the charge transfer to monolayer graphene is about 2 times slower. Surprisingly, the charge transfer to $A\text{−}B$–stacked bilayer graphene is slower than that to both monolayer graphene and graphite. This anomalous behavior disappears when the two graphene layers are randomly stacked. The observation is explained by a charge-transfer model that accounts for the band-structure difference in mono- and bilayer graphene, which predicts that the charge-transfer rate depends nonintuitively on both the layer number and stacking of graphene.

Figure 1

(a) Raman spectra of SG and $AB$-stacked BG directly grown on ${\mathrm{SiO}}_2/\mathrm{Si}$ substrates. (b) Raman spectra of the transferred SG and BG samples. For the BG sample, the two layers are transferred successively and have a random-stacking relationship.

Figure 2

Energy-level alignment at the interface. (a) UPS (left) and 2PPE (right) spectra for 0.5 nm ZnPc films deposited on various surfaces, which show the occupied and unoccupied band structures, respectively. The energy is referenced with respect to the Fermi level $E_F$. (SG, single-layer graphene; BG, bilayer graphene; HOPG, graphite; and Au, gold.) (b) The alignment of the HOMO (bottom) and $S_1$ state (top) of ZnPc with respect to $E_F$. The unit for the numbers is eV.

Figure 3

Time-resolved photoemission spectrum for ZnPc on single-layer graphene showing ultrafast electron transfer. (a) 0.5 nm ZnPc on SG and (b) 10 nm ZnPc on SG. The color scale represents the photoemission intensity which is proportional to the excited state population. The rapid intensity decay for the $S_1$ peak for the 0.5-nm sample is due to electron transfer from ZnPc to graphene.

Figure 4

The intensity of the $S_1$ state as a function of time for the (a) 0.5-nm samples and (b) 10-nm samples. For the 0.5-nm samples, the intensity decay is attributed to CT to the substrates. The initial CT rate is fitted with an exponential function convoluted with the instrumental response function (dashed line). The fits are shown as solid lines. The results for 10-nm samples represent the intrinsic $S_1$ dynamics of the ZnPc film. (c) The CT dynamics for 0.5 nm ZnPc deposited on transferred graphene. “$AB$” in the legend represents $AB$-stacked BG, while “Random” represents randomly stacked BG.

Figure 5

(a) Electronic band structure of BG calculated by using a tight-binding model. The schematic on the right shows the energy alignment at the interface. (b) The electron probability amplitude at $A$ and $B$ sites for the CB1 band of BG. Near the $K$ point, only $B$ sites are occupied. The figure on the right shows the crystal structure of BG. The top and bottom layers are shown by black and orange lines, respectively. On the top layers, only $B$ sites are occupied (blue shaded regions).