Collective chirality flipping of dibenzopentacene molecules induced by an electric field

Abstract

We report a scanning tunnelling microscopy (STM) study on the collective chirality flipping of dibenzopentacene (DBPen) molecules on Cd(0001). It is observed that the DBPen monolayer formed on Cd(0001) is composed of parallel molecular rows, which exhibit two enantiomeric lattices (chiral domains). The flat-lying molecules in the parallel rows exhibit an S-like shape with clockwise or anticlockwise handedness (single-molecule chirality). In particular, the pulse voltages from the STM tip lead to the simultaneous rotations of molecular rows and long-molecular-axes. When the two rotations have the same angle, pure lattice rotation takes place in the DBPen monolayer. When the two rotational angles are different, collective chirality flipping takes place in the homochiral domains, accompanied by the reversal of lattice chirality. Statistical analysis demonstrates that the chirality flipping takes place only at large negative pulse voltage, and there is a weak dependence of the chiral reversal probability on the tunnelling current, suggesting that the electric field of the STM tip is the primary driving force for chirality reversal.

---

Introduction

Chirality plays an important role in chemistry, biology, and materials science,^1–7 due to the fundamental importance and wide applications in enantioselective heterogeneous catalysis and chemical sensors.^8,9 In recent years, with the aid of scanning tunnelling microscopy (STM), significant achievements have been made in the study of molecular chirality at surfaces.^10–19 Chirality switching is an efficient channel toward the formation of extended homochiral domains, representing the most dramatic effects of chirality steering. It was reported that a prochiral molecule on Au(111) exhibits thermally induced chirality flipping between the enantiomeric forms.²⁰ The Morgenstern group found that inelastic electron tunnelling (IET) induces the chirality flipping of the chloronitrobenzene molecules on Au(111).²¹ Parschau et al. also demonstrated that the two distinct enantiomeric states of propene molecules on Cu(211) can be interconverted by IET.²² It is noteworthy that the previous chiral switching events take place only at the single-molecule level, since chirality switching needs certain energy to overcome the barrier between the two enantiomers. To the best of our knowledge, switching the chirality of numerous molecules in a controlled and reversible manner has hitherto not been reported. It was recently reported that the electric field from the STM tip can modulate the interfacial adsorption and manipulate both the interfacial phase transition and chemical reactions.^22–26 In particular, the electric field-induced trans–cis isomerization of azobenzene molecules was observed on Au(111).²³ Thus, one may expect a dramatic effect of the tip electric field on the chirality controlling and chirality switching.

As a derivative of the organic semiconductor pentacene, the trans-dibenzopentacene (DBPen) molecule contains two additional benzene rings attached to the diagonal ends of the long-molecular axis of pentacene.²⁷ In the single crystals of DBPen, there is a dihedral angle of ∼4° between the two terminal benzene rings.²⁸ The distortion (twisting) from planarity reduces the molecular symmetry and brings chirality to the DBPen molecules, such as the clockwise (R) or anticlockwise (L) handedness illustrated in Fig. 1. Recently, the structure and electronic properties of DBPen molecules have received growing interest.^29–35 Thin films of DBPen have been grown on Au(111),²⁸ Ag(111),³⁰ Cu(111),^32,33 and glass.^29,34 In particular, the discovery of superconductivity in K-doped DBPen with a high T_c of 33 K has opened a new avenue for the organic superconductors.^36–40 However, there have been few studies on the chirality of DBPen molecules on a solid surface.

Fig. 1 Structure models of the planar and twisted DBPen molecules with L- or R-handedness.

In this study, a Cd(0001) thin film characterized by low electronegativity and negative electron affinity was selected as the substrate.^41–43 Compared to the noble metals Au, Ag, and Pt, the divalent hexagonal close-packed metal Cd has a smaller work function, which promotes the charge transfer from the Cd substrate to DBPen molecules. As a result, the adsorbed DBPen molecules on Cd(0001) may carry net negative charge or dipole moments, facilitating the STM manipulation of DBPen molecules via the tip electric field. On the other hand, it has potential applications to study the growth of DBPen thin films on Cd(0001), because metal Cd is usually used as an electrode material due to the high redox potential.

It is observed that the DBPen monolayer formed on the Cd(0001) surface is composed of parallel molecular rows. The flat-lying molecules exhibit an S-like shape with R- or L-chirality. The pulse voltages from the STM tip lead to both rotations of DBPen molecules and molecular rows. When the two rotations have equal angles, the pure lattice rotation takes place in the monolayer. When the two rotations have different angles, the DBPen molecules in the homochiral domains undergo collective chiral flipping, accompanied by the reversal of the lattice chirality. The statistical analysis demonstrates that the chirality flipping takes place only at negative pulse voltage. No obvious dependence of flipping probability on the tunnelling current can be discerned.

Experimental details

The experiments were conducted in a Unisoku low-temperature STM with a base pressure of 2.0 × 10⁻¹⁰ mbar. A Si(111) wafer was degassed overnight at 880 K, and flashed to 1400 K to remove the surface oxides. Smooth Cd(0001) thin films were fabricated by depositing 15–20 monolayers of Cd atoms onto the Si(111)-7 × 7 surface with a rate of 0.35 ML min⁻¹, with the substrate held at room temperature. Fig. 2(a) shows a topographic STM image of the as-grown Cd(0001) crystalline thin film, which displays the hexagonal lattice (c₀ = 3.0 ± 0.2 Å). The trans-DBPen molecules were placed in a quartz crucible and heated to 370 K. During the deposition, the Cd(0001) substrate was kept at 100 K, and the deposition rate was approximately 0.18 ML min⁻¹. Here, we define one monolayer of DBPen as the coverage that covers the entire Cd(0001) surface before the appearance of a second layer of molecules. After deposition, the sample was quickly transferred to the STM chamber for structural characterization. All STM images were acquired at liquid nitrogen temperature (77.8 K) in constant-current mode.

Fig. 2 Parallel molecular rows of DBPen molecules on Cd(0001). (a) Topographic image of the Cd(0001) thin films (200 nm × 200 nm, 4.0 V, 50 pA). The inset shows the atomic-resolution image of Cd (0001) (3 nm × 3 nm, −1.2 V, 30 pA). (b) Two enantiomeric domains of the DBPen molecule rows (20 nm × 20 nm, 1.2 V, 20 pA). (c) High-resolution STM image of parallel rows of DBPen molecules (6.5 nm × 6.5 nm, 0.9 V, 20 pA). (d) Schematic model of the two enantiomeric domains composed of L-type and R-type molecules, respectively.

Results and discussion

Single-molecule chirality of the flat-lying molecules

First, small amounts (∼0.1 ML) of trans-DBPen molecules were deposited onto the Cd(0001) surface at 100 K. The image or trace of DBPen molecules cannot be found from the STM scanning. Instead, strong fuzzy noises appear in the STM images, suggesting that the DBPen molecules are highly mobile and collide frequently with the STM tip. This means that the deposited DBPen molecules formed a disordered two-dimensional (2D) molecular gas on the Cd(0001) surface. Previously, similar 2D gases were observed in other molecular systems, such as copper phthalocyanine on Cd(0001)⁴⁴ and pentacene on Ag(111), Ag(110), and Au(111) substrates.^45–47 When the molecular coverage exceeds a critical value, ordered condensed phases form on the substrates. Otto et al. found that the DBPen molecules form a static disordered structure, but not the mobile gas phase.³⁰ We speculate that the DBPen molecules adsorbed on Cd(0001) carry the dipole moment due to charge transfer effects. At low coverage, the intermolecular long-range electrostatic repulsion results in the formation of a 2D gas of DBPen molecules.

Increasing the DBPen coverage leads to the reduction of the intermolecular distance and the enhancement of the short-range attractive force, which would result in the condensation of a 2D molecular gas. When the coverage reaches 0.8 ML, the DBPen molecules condense into a monolayer phase composed of parallel molecular rows. As shown in Fig. 2(b), the parallel molecular rows form two enantiomeric domains, denoted as ρ-type and λ-type, respectively. Inside each domain, the molecular rows exhibit an oblique lattice with a = 19.4 ± 0.2 Å, b = 8.3 ± 0.2 Å, and θ = 74° ± 0.5°. The a-axes of the ρ- and λ-domain deviate +8° and −8° from the lattice direction of the Cd(0001) substrate, respectively. This kind of mirrored symmetry for the molecular arrangement is referred to as lattice chirality or organizational chirality. The transformation matrix between the molecular lattice of the ρ-domain and the substrate lattice is calculated as [−2.58, −0.33; −5.88, 6.92]. The transformation matrix for the λ-domain is changed to [0.33, 2.58; −6.92, 5.88]. No integer can be found from the columns of the matrices, indicating that the DBPen monolayer is incommensurate with respect to the Cd(0001) substrate.

It is also observed that the molecules inside the ρ-domain exhibit a clockwise S-like shape, denoted as R-handedness, while those in the λ-domain exhibit an anticlockwise S-like shape, denoted as L-handedness. This indicates that chiral separation takes place spontaneously in the form of homochiral domains. Close inspection indicates that the S-like protrusions have a size of 18.5 Å × 8.1 Å, which are very close to the long axis (1.8 nm) and short axis (0.8 nm) of a DBPen molecule. This implies that the DBPen molecules adopt the flat-lying orientation when adsorbed on Cd(0001).

Fig. 2(c) shows a high-resolution STM image of the R-type molecules in a ρ-domain. Inside each DBPen molecule, the pentacene core manifests as five parallel lobes, corresponding to the five benzene rings. It is noticed that not all DBPen molecules exhibit exactly the same morphology. The molecule marked by a red arrow is different from the other molecules. There are no lobes appearing at the pentacene core, in spite of the two terminal benzene rings. This anomalous molecule is possibly related to the charge state due to the acceptance of some tunnelling electrons from the STM tip. Fig. 2(d) illustrates the enantiomeric relationship between the arrangements of R-type molecules inside the ρ-domain and the L-type molecules inside the λ-domain. The lattice directions of the two domains are oriented at the angles +8° and −8° with respect to the lattice direction of the Cd(0001) substrate, respectively. As a result, there is a 16° angle between the long-molecular-axes of R-type and L-type molecules.

Collective flipping of the single-molecule chirality induced by a pulse voltage

The DBPen molecules are easily rotated and moved under the influence of the electric field of the STM tip. Fig. 3(a) shows a ρ-domain composed of R-type molecular rows. The large bright protrusion located at the upper right corner serves as the reference point. In the process of a second round of scanning (from bottom to top), a voltage pulse of −4.0 V with 50 ms width was applied at the cross mark position. The scanning was continued after a pause of 5 seconds. As shown in the upper half panel of Fig. 3(b), the orientation of the molecular rows is rotated by an angle of −41° after the voltage pulse. Meanwhile, the long molecular axes of all molecules are rotated by −16°. Most importantly, it is found that the different rotational angles between the molecular rows and long-molecular axis lead to the chirality flipping. All the R-type molecules are changed into the L-type molecules, i.e. collective flipping of the single-molecule chirality occurs in the homochiral domains. Furthermore, the lattice chirality has also been found to switch from the ρ-domain to a λ-domain.

Fig. 3 Collective chirality flipping of the DBPen molecules induced by a voltage pulse. (a) A ρ-domain composed of R-type molecules before a voltage pulse is applied. (b) Chirality flipping occurs after applying a voltage pulse at the cross mark position. (c) The same sample area as (a) and (b) acquired in the third round of scanning. (d) Schematic models for the arrangements of DBPen molecules before and after chiral flipping. The vertical arrows show the slow-scanning direction. All image sizes are 20 nm × 20 nm, and scanning parameters are 1.8 V, 26 pA.

In fact, based on the result in Fig. 3(b), we can only conclude that the chirality flipping takes place in the upper half panel. It is not clear whether all the molecules in the whole panel undergo the chiral flipping. Fig. 3(c) shows the third round of scanning of the same sample area (from top to bottom). It is observed that, after the voltage pulse, not only the molecules in the upper half panel but also those in the lower half panel undergo the same chirality flipping. Thus, the voltage pulse from the STM tip leads to collective chirality switching involving hundreds of molecules.

Regarding the flipping mechanism of single-molecule chirality and lattice chirality, we propose that the pulse voltage of the STM tip generates a strong electrical field, which brings a perturbative potential to the DBPen molecules beneath the tip. The perturbative potential may overcome the energy barrier between two enantiomers and results in both the out-of-plane rotation of 180° (rolling) and in-plane rotation. It is the 180° rolling that leads to the chirality flipping from R- to L-handedness. Moreover, due to the intermolecular van der Waals attraction, the rolling of one molecule beneath the tip would induce similar motions of adjacent molecules. As a result, a cascade of reactions takes place in the homochiral domain and results in the collective chiral reversal of hundreds of molecules. Fig. 3(d) shows the handedness and lattice structures of the molecular rows before and after the chirality flipping, where the direction of the molecular rows is rotated by −41°, and the long-molecular axis is rotated by −16°.

In principle, if two consecutive pulse voltages are applied in the same homochiral domain, the chirality flipping event should take place twice, which would in turn make the chirality recover. Fig. 4(a) shows an STM image of the λ-domain composed of L-type molecules acquired before a voltage pulse. During the second round of scanning (from bottom to top) as shown in Fig. 4(b), a pulse voltage was applied at the dotted line. It is found that the molecular chirality is changed from L- to R-handedness, and the lattice chirality is also changed from the λ-domain to a ρ-domain. Meanwhile, the direction of molecular rows rotated by −81° and the long axis of DBPen molecules rotated by 76°. In the process of the third round of scanning (from top to bottom), once more we applied a pulse voltage at the dotted line in Fig. 4(c). It is found that the chirality flipping takes place in the low half panel. The direction of the molecular rows is rotated by an angle of −40°, and the long-molecular axes are rotated by an angle of −16°, which is the same angle as in Fig. 3(b). From the fourth round of scanning shown in Fig. 4(d), it is found that the second time of chirality flipping involves all the molecules in the whole panel. As a result, the single-molecule chirality has been recovered to the original L-handedness, and the lattice chirality is also restored to the λ-domain, in spite of the different directions of the molecular rows and different molecular orientations between Fig. 4(a) and (d).

Fig. 4 Chirality recovery after two times of chirality flipping. (a) A λ-domain composed of L-type DBPen molecules before a voltage pulse. (b) Chirality flipping occurs in the upper half panel after the first voltage pulse. (c) The second time of chirality flipping occurs in the lower half panel. (d) The fourth round of scanning showing the recovery of single-molecule chirality and lattice chirality. All image sizes are 20 nm × 20 nm. The scanning parameters are 2.25 V, 26 pA.

The dependence of chirality flipping probability on the voltage and tunnelling current

In principle, the phenomenon of chirality flipping on a solid surface may involve several mechanisms such as the IET induced excitation of molecular vibration,²¹ and the electric field induced deformation of the potential energy surface.²³ To identify the primary mechanism that is responsible for the collective chirality reversal, we performed statistical analysis on the dependence of the chiral reversal probabilities on pulse voltage and tunneling current. Fig. 5(a) shows the variation of the reversal probability versus the pulse voltage. The magnitude of pulse voltage varies from −5.0 V to +4.0 V with a step size of 0.5 V and a pulse width of 50 ms. Before applying the pulse voltage, the tip height was fixed by the scanning parameters of U = 1.5 V and I = 20 pA, and the feedback loop was switched off to avoid the influence from pulse voltages. It is found that chirality reversal takes place only at negative pulse voltages. At positive pulse voltages from 0–4.0 V, the reversal probability remains zero. For the small value of negative voltages, the chiral reversal probability is also zero. A clear threshold voltage of approximately −1.7 V has been observed. Beyond the threshold voltage, the flip probability starts to rise from zero. When the voltage is increased to −4.0 V, the flipping probability reaches a maximum value of 0.53. After this maximum, the reversal probability falls to 0.18 when the voltage is increased to −5.0 V. It was noticed that the voltage corresponding to the maximum probability appears at −4.0 eV, which is coincident with the work function (4.07 eV) of metal Cd. Thus, the reduction of flipping probability after the maximum value is associated with the enhanced charge transfer from the Cd substrate to the DBPen monolayer. The built-in electric field is opposite to the external electric field of the STM tip, and thus results in the reduction of the flipping probability.

Fig. 5 Statistics of chirality reversal probability. (a) The dependence of chiral reversal probability on the pulse voltage with tunnelling current kept at 20 pA. (b) Plot of chiral reversal probability vs. tunnelling current with the voltage kept at −4.0 V for 50 ms. Each data point is obtained by repeating the same measurement 20 times.

Fig. 5(b) shows the dependence of chirality flip probability on the tunneling current. It is found that the probability is weakly dependent on the tunneling current. When the tunneling current increases from 50 pA to 250 pA, the flip probability fluctuates between 0.35 and 0.55. The nonmonotonic variation of probability implies that the non-elastic electron tunneling (IET) mechanism does not play a dominant role in the chirality flip process. Otherwise, the flip probability should be proportional to the n-th power of tunneling current, where n is an integer. Since the reversal probability depends on the direction of the electric field, i.e., nonzero probability only appears in the negative voltage, the primary mechanism of chiral flipping can be attributed to the electric field from the STM tip, which deforms the potential energy surface and lowers the barrier between the two opposite enantiomers.

Lattice rotation without chirality flipping

Based on the statistical analysis, the maximum chirality reversal probability is 0.53, which means that not all pulse voltages can lead to the reversal of single-molecule chirality and lattice chirality. In other words, the minimum probability of chirality conversion is 0.47 after the voltage pulse. In some cases, we can only observe the lattice rotation without the flipping of single-molecule chirality and lattice chirality. When the lattice rotation takes place, both the direction of the molecular rows and the molecule orientation have been changed. Fig. 6(a) displays a ρ-domain composed of R-type molecules. During the second round of scanning, a voltage pulse of −4.0 V was applied at the blue cross position, as shown in Fig. 6(b). After the pulse voltage, the molecular chirality still retains R-handedness and the lattice chirality keeps the ρ-domain. However, close-inspection indicates that the direction of the molecular rows in the lower half panel is rotated by −60° relative to that in the upper half panel. Meanwhile, the long-molecular axes are also rotated by an angle of −60°. Based on these observations, we conclude that, in the case of lattice rotation, both the directions of the molecular rows and the molecular orientations are rotated for the same angle. When the two rotational angles mentioned above have different values, collective chirality flipping takes place in the homochiral domain.

Fig. 6 Lattice rotation induced by a voltage pulse. (a) A ρ-domain composed of R-type molecules before a voltage pulse. (b) The same sample area as (a) after applying a voltage pulse at the cross mark position. The image sizes are 20 nm × 20 nm. The scanning parameters are 1.8 V, 26 pA.

Electronic structure of the DBPen monolayer

Through the measurement of differential conductance, the scanning tunnelling spectra (STS) of the DBPen monolayer have been obtained. As shown in Fig. 7, the highest occupied molecular orbital (HOMO) of the DBPen molecules is located at −1.77 eV, and the lowest unoccupied molecular orbital (LUMO) is at +1.72 eV. Thus, the HOMO–LUMO gap is 3.49 eV for the DBPen monolayer. Additionally, we observed three weak peaks located at −0.56 eV, +0.53 eV, and +1.26 eV, respectively, which can be assigned to the quantum well states of Cd(0001) thin films based on the previous STS measurement.⁴⁸

Fig. 7 STS of the flat-lying DBPen molecules in the molecular rows. The tip height was fixed by U = 1.0 V, I = 50 pA.

Based on the experiment of valence-band photoemission, Mahns et al. found that the ionization energy of the DBPen films grown on a KBr substrate is 5.25 eV, and the calculated electron affinity is 1.67 eV.²⁷ According to the definitions of ionization energy and electron affinity, the difference between the ionization energy and electron affinity is the HOMO–LUMO gap. Thus, the HOMO–LUMO gap is 3.58 eV for the DBPen thin films grown on KBr, consistent with the 3.49 eV gap of the DBPen monolayer on Cd(0001). Otto et al. studied the electronic states of the DBPen thin film on an Ag(111) substrate by ultraviolet photoelectron spectroscopy (UPS).³⁰ They found that the HOMO level of the DBPen molecules is at −2.0 eV, and the calculated LUMO level is at +0.2 eV. Thus, the DBPen monolayer on Ag(111) exhibits a HOMO–LUMO gap of 2.2 eV, which is much smaller than our measured gap of 3.49 eV. The smaller gap of the DBPen monolayer on Ag(111) can be attributed to the stronger electronic screening from the noble metal substrate relative to the Cd(0001) thin films.

Conclusions

In this work, we demonstrated that electric field-induced collective chirality switching occurred in the DBPen monolayer on a Cd(0001) surface. Such kind of chirality reversals take place at the scale of tens of nanometres, involving hundreds of DBPen molecules. Not only the single-molecule chirality but also the lattice chirality can be switched by applying negative voltage pulses. These results provide an efficient way to realize large-scale chirality manipulations, which are beneficial to the applications in molecular electronic devices (chiral switches and chiral sensors) and selective catalysis. This work represents an important step toward the chirality steering and provides key insights into the interplay between molecular chirality and external electric fields.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data that support the findings of this study are available within the article.

Acknowledgements

Financial support from the National Natural Science Foundation of China (Grant No. 11874304 and 11574253) is gratefully acknowledged.

Notes and references

1. S. M. Barlow and R. Raval, Surf. Sci. Rep., 2003, 50, 201–341.
2. K. H. Ernst, Supramolecular surface chirality, in Supramolecular Chirality, ed. M. CregoCalama and D. N. Reinboudt, Springer-Verlag Berlin, Berlin, 2006, vol. 265, pp. 209–252.
3. J. Elemans, I. De Cat and H. X. S. De Feyter, Chem. Soc. Rev., 2009, 38, 722–736.
4. M. O. Lorenzo, C. J. Baddeley, C. Muryn and R. Raval, Nature, 2000, 404, 376–379.
5. A. Kühnle, T. R. Linderoth, B. Hammer and F. Besenbacher, Nature, 2002, 415, 891–893.
6. S. Haq, N. Liu, V. Humblot, A. P. J. Jansen and R. Raval, Nat. Chem., 2009, 1, 409–414.
7. O. Stetsovych, M. Svec, J. Vacek, J. V. Chocholousová, A. Jancarík, J. Rybácek, K. Kosmider, I. G. Stará, P. Jelínek and I. Stary, Nat. Chem., 2017, 9, 213–218.
8. M. O. Lorenzo, S. Haq, T. Bertrams, P. Murray, R. Raval and C. J. Baddeley, J. Phys. Chem. B, 1999, 103, 10661–10669.
9. G. Fukuhara and Y. Inoue, Chem. – Eur. J., 2010, 16, 7859–7864.
10. G. P. Lopinski, D. J. Moffatt, D. D. Wayner and R. A. Wolkow, Nature, 1998, 392, 909–911.
11. R. Fasel, M. Parschau and K. H. Ernst, Nature, 2006, 439, 449–452.
12. T. Huang, Z. P. Hu, A. D. Zhao, H. Q. Wang, B. Wang, J. L. Yang and J. G. Hou, J. Am. Chem. Soc., 2007, 129, 3857–3862.
13. H. Cao and S. De Feyter, Nat. Commun., 2018, 9, 9.
14. S. Y. Li, T. Chen, J. Y. Yue, D. Wang and L. J. Wan, Chem. Commun., 2017, 53, 11095–11098.
15. T. Chen, W. H. Yang, D. Wang and L. J. Wan, Nat. Commun., 2013, 4, 8.
16. X. Liao, E. Minamitani, T. Xie, L. Z. Yang, W. H. Zhang, S. Klyatskaya, M. Ruben and Y. S. Fu, J. Am. Chem. Soc., 2024, 146, 5901–5907.
17. Z. W. Liang, S. R. Wu, J. M. Wang, Y. F. Qin, F. Cheng, L. Cao and H. Xu, ACS Nano, 2023, 17, 10938–10946.
18. H. R. Liu, H. P. Li, Y. He, P. Cheng, Y. Q. Zhang, B. J. Feng, H. Li, K. H. Wu and L. Chen, Nat. Commun., 2023, 14, 10.
19. M. Schied, D. Prezzi, D. D. Liu, S. Kowarik, P. A. Jacobson, S. Corni, J. M. Tour and L. Grill, ACS Nano, 2023, 17, 3958–3965.
20. S. Weigelt, C. Busse, L. Petersen, E. Rauls, B. Hammer, K. V. Gothelf, F. Besenbacher and T. R. Linderoth, Nat. Mater., 2006, 5, 112–117.
21. V. Simic-Milosevic, J. Meyer and K. Morgenstern, Angew. Chem., Int. Ed., 2009, 48, 4061–4064.
22. M. Parschau, D. Passerone, K. H. Rieder, H. J. Hug and K. H. Ernst, Angew. Chem., Int. Ed., 2009, 48, 4065–4068.
23. M. Alemani, M. V. Peters, S. Hecht, K. H. Rieder, F. Moresco and L. Grill, J. Am. Chem. Soc., 2006, 128, 14446–14447.
24. B. Borca, T. Michnowicz, R. Pétuya, M. Pristl, V. Schendel, I. Pentegov, U. Kraft, H. Klauk, P. Wahl, R. Gutzler, A. Arnau, U. Schlickum and K. Kern., ACS Nano, 2017, 11, 4703–4709.
25. P. Matvija, F. Rozboril, P. Sobotík, I. Ost'ádal, B. Pieczyrak, L. Jurczyszyn and P. Kocán, Sci. Rep., 2017, 7, 8.
26. G. Velpula, J. Teyssandier, S. De Feyter and K. S. Mali, ACS Nano, 2017, 11, 10903–10913.
27. B. Mahns, F. Roth, A. König, M. Grobosch, M. Knupfer and T. Hahn, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 86, 5.
28. L. Graf, F. P. Liu, M. Naumann, F. Roth, B. Debnath, B. Büchner, Y. Krupskaya, A. A. Popov and M. Knupfer, ACS Omega, 2022, 7, 36902.
29. A. Tovstopyat, E. Zojer and G. Leising, J. Appl. Spectrosc., 2016, 83, 20–26.
30. F. Otto, T. Huempfner, M. Schaal, C. Udhardt, L. Vorbrink, B. Schroeter, R. Forker and T. Fritz., J. Phys. Chem. C, 2018, 122, 8348–8355.
31. A. Jancarik, G. Levet and A. Gourdon, Chem. – Eur. J., 2019, 25, 2366–2374.
32. J. Simbrunner, B. Schrode, S. Hofer, J. Domke, T. Fritz, R. Forker and R. Resel, J. Phys. Chem. C, 2021, 125, 618–626.
33. J. Simbrunner, J. Domke, F. Sojka, A. Jeindl, F. Otto, M. Gruenewald, O. T. Hofmann, T. Fritz, R. Resel and R. Forker, Acta Crystallogr., Sect. A: Found. Adv., 2022, 78, 272–282.
34. B. Manna, A. Nandi and R. Ghosh, J. Phys. Chem. C, 2022, 9.
35. M. Rosenkranz, L. Graf, B. Büchner, M. Knupfer and A. A. Popov, J. Mater. Chem. C, 2023, 11, 12714–12729.
36. M. Q. Xue, T. B. Cao, D. M. Wang, Y. Wu, H. X. Yang, X. L. Dong, J. B. He, F. W. Li and G. F. Chen, Sci. Rep., 2012, 2, 4.
37. F. Roth, A. König, B. Mahns, B. Büchner and M. Knupfer., Eur. Phys. J. B, 2012, 85, 5.
38. T. Kato, J. Phys. Chem. C, 2013, 117, 17211–17224.
39. G. H. Zhong, C. Zhang, X. W. Yan, X. G. Li, Z. Du, G. X. Jing and C. C. Ma, Mol. Phys., 2017, 115, 472–483.
40. M. Aoki and S. Masuda, J. Electron Spectrosc. Relat. Phenom., 2015, 204, 68–74.
41. R. W. Stark and L. M. Falicov, Phys. Rev. Lett., 1967, 19, 795–798.
42. S. Daniuk, T. Jarlborg, G. Kontrymsznajd, J. Majsnerowski and H. Stachowiak, J. Phys.: Condens. Matter, 1989, 1, 8397–8406.
43. H. S. Chauhan, L. Ilver, P. O. Nilsson, J. Kanski and K. Karlsson, Phys. Rev. B: Condens. Matter Mater. Phys., 1993, 48, 4729–4734.
44. Y. L. Wang, K. Sun, Y. B. Tu, M. L. Tao, Z. B. Xie, H. K. Yuan, Z. H. Xiong and J. Z. Wangd, Phys. Chem. Chem. Phys., 2018, 20, 7125–7131.
45. M. Eremtchenko, R. Temirov, D. Bauer, J. A. Schaefer and F. S. Tautz, Phys. Rev. B: Condens. Matter Mater. Phys., 2005, 72, 9.
46. Y. L. Wang, W. Ji, D. X. Shi, S. X. Du, C. Seidel, Y. G. Ma, H. J. Gao, L. F. Chi and H. Fuchs, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 69, 5.
47. N. Koch, A. Vollmer, S. Duhm, Y. Sakamoto and T. Suzuki, Adv. Mater., 2007, 19, 112–116.
48. M. L. Tao, H. F. Xiao, K. Sun, Y. B. Tu, H. K. Yuan, Z. H. Xiong, J. Z. Wang and Q. K. Xued, Phys. Rev. B, 2017, 96, 6.

---