Nonuniform STM Contrast of Self-Assembled Tri-n-octyl-triazatriangulenium Tetrafluoroborate on HOPG

We have assembled 4,8,12-tri-n-octyl-4,8,12-triazatrianguleniumtetrafluoroborate (TATA-BF4) on highly oriented pyrolytic graphite (HOPG) and have studied the structure and tunneling properties of this self-assembled monolayer (SAM) using scanning tunneling microscopy (STM) under ambient conditions. We show that the triazatriangulenium cations TATA+ form hexagonally packed structures driven by the interaction between the aromatic core and the HOPG lattice, as evidenced by density functional theory (DFT) modeling. According to the DFT results, the three alkyl chains of the platform tend to follow the main crystallographic directions of HOPG, leading to a different STM appearance. The STM contrast of the SAM shows that the monolayer is formed by two types of species, namely, TATA+ with BF4– counterions on top and without them. The cationic TATA+ platform gives rise to a seemingly higher appearance than neutral TATA-BF4, in contrast to observations made on metallic substrates. The variation of the STM tunneling parameters does not change the relative difference of contrast, revealing the stability of both species on HOPG. DFT calculations show that TATA-BF4 on HOPG has sufficient binding energy to resist dissociation into TATA+ and BF4–, which might occur under the action of the electric field in the tunneling gap during STM scanning.

model of the top layer of HOPG.As such, we initially have considered a configuration of the TATA + core being rotated with respect to the graphene hexagonal rings by 30° (Figure S1).Next, a configuration with the TATA + core being aligned with the hexagons of the graphene leading to a stacking in an AA fashion (Figure S2) was investigated.Finally, a configuration with the TATA + core being translated by half the unit cell to give an AB-like stacking with regard to the graphene (Figure S3) turned out to be the most favorable position.This configuration has been calculated after full DFT optimization and taking into account van der Waals interactions as described above.graphene.Chemical elements are color-coded (gray for carbon, blue for nitrogen, white for hydrogen, pink for boron, green for fluorine and red for carbon atoms in graphene).

2.
Filtering of the noise in some STM images The quality of all images is limited by the thermal drift and external acoustical noise, since all measurements have to be done in ambient atmospheric environment to gain a thermally induced selforganization of the molecules.This is the typical situation in ambient STM, and the figure quality is state-of-the-art for this method.As an example for the noise level in ambient STM measurements Fig. S5 shows the original data of Fig. 4 of the manuscript before (left) and after noise filtering (right) using the well-established software package SPIP (Digital Surf) 1 .

a) b)
Figure S5 Original (a) and filtered (b) image of the data presented in Fig. 4 in the main manuscript using the well-established software package SPIP 1

Calculations of the Projected Density of States (PDOS)
In order to fully discriminate the species observed in the STM images, we have performed electronic structure calculations on four different potential configurations on graphene and Au(111), respectively.Namely, we have considered TATA + /graphene and TATA + BF4 -/graphene and the same on Au (111).The atomic structure of these configurations is represented in Figure 5a) to d) in the main text.As for the other calculations, structural optimizations and electronic structure calculations have been performed using the Fireball code 1 .The whole methodology has been fully detailed previously. 2,3e corresponding PDOS for the four configurations are represented in Figure 5 e) in the main text.
As a result, we observe a pronounced maximum of the PDOS close to EF for TATA + /graphene, while no maximum is close to EF for the other three configurations.Comparing with TATA-BF4/graphene, we observe a similar width of the resonances while on Au(111) they are broadened.These trends can be explained with the much weaker van der Waals interaction between the TATA species and graphene compared to Au(111).The position of a molecular resonance close to EF, as observed for TATA + /graphene, signals the ionic nature of the TATA + platform and the weak charge transfer between TATA + and graphene.A high DOS at EF is a mandatory ingredient of a high electronic transmission. 2,3Although the absolute position of the resonance will depend on the positioning of the molecule with respect to the graphene lattice, we can safely expect a higher electronic transmission for the TATA + than for the neutral molecule TATA-BF4 provided a similar distance of the STM tip is chosen and similar molecular orbitals and electronic wavefunctions of the substrate are involved.We refrain from performing explicit transport calculations since the exact shape and position of the STM tip is not known in our experiment and will have a dominating influence on the absolute value of the transmission.Hence, these calculations give us the trend that when measuring the TATA with and without the BF4 -ion, there will be a clear difference in the obtained STM contrast.

Structural data of the adsorbed molecules
We provide here the .xyzfiles for the molecular configurations (TATA-BF4 on graphene and TATA+ on graphene, all coordinates are in Å).

Figure S1 :
Figure S1: Top view of DFT-optimized configuration of a TATA + cation randomly adsorbed on a 14 x 14 unit cell graphene sheet.

Figure S2 :
Figure S2: Top view of DFT-optimized configuration of a TATA + cation adsorbed on a 14 x 14 unit cell graphene sheet in an AA-like stacking with respect to the graphene sheet.

Figure S3 :
Figure S3: Top (Bottom): Top (side) view of DFT-optimized configuration of a TATA + cation adsorbed on a 14 x 14 unit cell graphene sheet in an AB-like stacking with respect to the graphene sheet.(original Figure 3a) of the manuscript).

Figure S4 .
Figure S4.Top and side view of DFT-optimized configuration of a TATA-BF4 molecule adsorbed on