betaFeedback

WorksPhysics and Astronomy

On-surface synthesis and characterization of Teranthene and Hexanthene: Ultrashort graphene nanoribbons with mixed armchair and zigzag edges

Abstract

Graphene nanoribbons (GNRs) exhibit a broad range of physicochemical properties that critically depend on their width and edge topology. While the chemically stable GNRs with armchair edges (AGNRs) are semiconductors with width-tunable band gap, GNRs with zigzag edges (ZGNRs) host spin-polarized edge states, which renders them interesting for applications in spintronic and quantum technologies. However, these states significantly increase their reactivity. For GNRs fabricated via on-surface synthesis under ultrahigh vacuum conditions on metal substrates, the expected reactivity of zigzag edges is a serious concern in view of substrate transfer and device integration under ambient conditions, but corresponding investigations are scarce. Using 10-bromo-9,9':10',9''-teranthracene as a precursor, we have thus synthesized hexanthene (HA) and teranthene (TA) as model compounds for ultrashort GNRs with mixed armchair and zigzag edges, characterized their chemical and electronic structure by means of scanning probe methods, and studied their chemical reactivity upon air exposure by Raman spectroscopy. We present a detailed identification of molecular orbitals and vibrational modes, assign their origin to armchair or zigzag edges, and discuss the chemical reactivity of these edges based on characteristic Raman spectral features.

Monobromo_MS_.pdf

keywords: on surface synthesis; nanographenes; scanning probe; length selectivity; Raman spectroscopy

Introduction

The interest in graphene nanoribbons (GNRs) has surged since graphene was first isolated, largely due to the potential for fine-tuning their physicochemical properties by manipulating their size and shape[1, 2]. The tailored fabrication of these one-dimensional systems, employing atomically precise control over their width and edge structure, enables the synthesis of GNRs with variable electronic band gaps[3, 4, 5, 6, 7], specific topological quantum phases[8, 9] or even spin-polarized edge states[10] in a single material class. The required atomic precision for such extended structures is achieved in a bottom-up approach based on the metal surface-catalyzed covalent assembly of carefully designed molecular precursors, which is nowadays called 'on-surface synthesis'. Not only does this approach enable the deterministic synthesis of GNRs (and other nanographenes) with the desired electronic and magnetic properties[8, 9, 11, 12, 13], it also opens up a great potential for their use in nanoelectronics[14, 15, 16] and spintronics[17, 18].

In order to fully leverage the GNRs properties into device platforms, a transfer step is necessary to bring them from their growth substrate to the technologically relevant ones[15, 19]. Next, a careful out-of-vacuum characterization is required to ensure that the GNRs' intrinsic properties remain preserved throughout the transfer and device integration process. Significant progress has been made in the ex situ characterization, mostly by spectroscopic methods[20, 21, 22, 23, 24, 25, 26], and in substrate transfer procedures of armchair-GNRs (AGNRs). Their robustness and stability under ambient conditions[19, 27] has allowed the realization of several types of AGNR field effect transistor (FET) devices[28, 29, 30, 31, 32, 33, 34].

Moreover, GNRs (and other nanographenes) containing zigzag edges host spin-polarized edge states[10, 35, 36, 37, 38] which render them interesting for applications in spintronics[39] and quantum computation[40, 41]. However, proper ex situ characterization and device integration are extremely challenging since zigzag edges are prone to oxidation[13, 42, 43, 44], which so far has hampered any experimental investigation of their transport characteristics.

Here we designed and synthesized two ultrashort GNR segments, namely teranthene (TA) and hexanthene (HA). With a comparable number of armchair and zigzag edge sites, TA and HA serve as model systems for the investigation of structure dependent electronic properties and chemical reactivity, in particular compared to long GNRs. TA and HA are synthesized from the precursor molecule 10-bromo-9,9′:10′,9″-teranthracene (1), by using dehalogenative aryl-aryl coupling and cyclodehydrogenation of 1 on a gold surface[3, 45]. We employ low-temperature scanning tunneling microscopy (STM) and non-contact atomic force microscopy (nc-AFM) under ultra-high vacuum (UHV) for the in situ investigation of structural and electronic properties of TA and HA at sub-molecular resolution. Moreover, we utilize vacuum Raman spectroscopy to probe the vibrational signatures of these pristine, atomically precise GNR segments. Characterizing the Raman modes of both TA and HA, before and after their ambient exposure, allows us to correlate changes in Raman spectroscopy signatures with the chemical reactivity associated with zigzag edges. Our results confirm that Raman spectroscopy can serve as a reliable method to detect the chemical degradation of highly reactive GNRs.

Results and Discussion

Ultrashort GNR segments, TA and HA, were achieved via the precursor 1 (10-bromo-9,9':10',9''-teranthracene) Fig. 1, (the details for the solution synthesis of 1 are in the supporting information, Fig. S1-S3). 1 was sublimated onto an Au(111) surface under UHV conditions and thermally activated to induce surface-assisted dehalogenation and subsequent C-C coupling. 80% of the dehalogenated precursor molecules underwent dimerization at 210° C, while 20% of the surface-stabilized radicals were passivated by hydrogen after dehalogenation of the precursor, thus preventing dimer formation (see Fig. S4 for STM images). Subsequently, intramolecular cyclodehydrogenation of the dimers and passivated monomers was induced by annealing the sample to 290 °C (Fig. 1b-d), which led to a majority of fully planar species coexisting with few partially cyclodehydrogenated molecules (featuring bright protrusions in Fig. 1c). To identify the resulting products, we conducted bond-resolved imaging using nc-AFM with a CO-functionalized tip[46] (Fig. 1d and Fig.S5), which conclusively proves the successful formation of TA and HA.

Synthesis and characterization of ultrashort graphene nanoribbon segments TA and HA

Figure 1.

Image not extracted

Figure 1 (PDF p. 4). a) Solution-synthetic route toward precursor 1, b) Schematic description of the on-surface synthesis of TA and HA via dehalogenation and aryl-aryl covalent coupling (radical passivation in the monomer case) followed by cyclodehydrogenation of 1. c) Large-scale STM image of the surface after deposition of 1 on Au(111) at RT and subsequent annealing to 290 °C. d) High-resolution STM (left panels) and Laplace-filtered nc-AFM frequency shift images (right panels) of TA (bottom) and HA (top) obtained using a CO-functionalized tip. Scanning parameters: c: V=−1.00V = -1.00 V, I=20I = 20 pA, d (STM, top): V=−20V = -20 mV, I=100I = 100 pA, d (STM, bottom): V=−20V = -20 mV, I=200I = 200 pA. See Figure S5 for the corresponding unfiltered nc-AFM images.

nc-AFM frequency-shift images of HA and TA and corresponding Laplace-filtered versions

Figure S5. Left panels: nc-AFM frequency-shift images of HA (upper panel) and TA (lower panel). Scanning parameters: (top) tip height offset Δz=195\Delta z=195 pm above the STM setpoint V=−5V=-5 mV, I=100I=100 pA, (bottom) Δz=190\Delta z=190 pm above the STM setpoint V=−5V=-5 mV, I=100I=100 pA. Right panels: Corresponding Laplace-filtered versions of the left panels (also shown in Fig. 1).

The electronic structure of TA and HA was probed using scanning tunneling spectroscopy (STS). Differential conductance (dI/dVdI/dV) spectra acquired on TA reveal four distinct peaks at -1.65 V, -0.10 V, 0.20 V and 1.00 V (Figure 2a–Figure 2d). In order to assign these states to specific molecular orbitals, we record spatially resolved dI/dVdI/dV maps over TA at the respective bias voltages and compared them to gas-phase tight-binding local density of state (TB-LDOS) maps of the molecular orbitals of TA (Figure S6 and Figure S7). As shown in Figure 2d, the dI/dVdI/dV maps at -0.10 V, 0.20 V and 1.00 V concur with the TB-LDOS maps of the HOMO, LUMO and LUMO+1 of TA, thus allowing for an unambiguous assignment of the frontier molecular orbitals. The dI/dVdI/dV map at -1.65 V, while containing features of both the HOMO–1 and HOMO–2 of TA (Figure S8), cannot be uniquely assigned to a single orbital resonance. Our TB simulations indicate that this resonance likely results from an intermixing of the HOMO–1 and HOMO–2, as has also been previously observed in related molecular systems[47] (Figure 2d and Figure S8). We also explored the electronic structure of HA via STS, where five resonances at -1.10 V, 0.10 V, 0.44 V, 1.25 V and 1.65 V are clearly resolved in dI/dVdI/dV spectroscopy (Figure 2e,Figure 2f). Based on the reasonable correspondence between the dI/dVdI/dV maps at the above biases and the TB-LDOS maps of HA, we assign the resonances at –1.10 V, 0.10 V, 0.44 V, 1.25 V and 1.65 V to the gas phase HOMO–1, HOMO, LUMO, LUMO+1 and LUMO+2 of HA, respectively (Figure 2g; also see Figure S6 and Figure S9). The experimentally measured resonance at 0.10 V, which matches the gas phase HOMO of HA, lies above the Fermi energy. This is in contrast with what was observed for TA and corresponds to the situation where electron transfer from HA to the Au(111) surface leads to the emptying of the HOMO, thus leading to its upshift above the Fermi energy. Our present results are in agreement with previous works, where hole-doping of 7-AGNR segments on Au(111) has been observed by both local scanning probe[48, 49] and ensemble electron spectroscopy and photoemission measurements[50]. However, while the reported STS data of extended 7-AGNRs on Au(111) show a band gap of 2.3 eV[51], we observe a significantly narrower gap in the case of TA and HA on Au(111) (0.30 and 0.34 eV, respectively). This gap reduction is likely due to the hybridization of the localized end states when the zigzag termini of the ribbon become close to one another. Conductance dI/dVdI/dV spectra of a suspended 7-AGNR revealed that

Electronic structure characterization of TA and HA

Figure 2. TA and HA. a) dI/dVdI/dV spectra acquired on TA and on the bare substrate (black curve). b) High-resolution dI/dVdI/dV spectroscopy acquired on TA in the vicinity of the Fermi level, showing the HOMO and LUMO resonances. c) STM image of TA with colored circles indicating the positions where the corresponding dI/dVdI/dV spectra in (a) and (b) were acquired. Scanning parameters: V=−0.10V = -0.10 V, I=300I = 300 pA. d) Upper panels: constant-current dI/dVdI/dV maps recorded at the indicated voltages, corresponding to the main features observed in (a) and (b). Bottom panels: Gas phase TB-LDOS maps of intermixed HOMO-2/HOMO-1, HOMO, LUMO and LUMO+1 of TA. e) dI/dVdI/dV spectra acquired on HA and on the bare substrate (black curve). f) STM image of HA with the colored circles indicating the positions where the corresponding dI/dVdI/dV spectra in (e) were acquired. Scanning parameters: V=0.05V = 0.05 V, I=100I = 100 pA. g) Upper panels: constant-current dI/dVdI/dV maps recorded at the indicated voltages, corresponding to the main spectral features observed in e). Bottom panels: Gas phase TB-LDOS maps of the HOMO-1, HOMO, LUMO, LUMO+1 and LUMO+2 of HA.

TB wave functions of HOMO–2 to LUMO+2 of TA and HA

Figure S6. TB wave functions of HOMO–2 to LUMO+2 of TA (top) and HA (bottom). Size of the filled circles denotes amplitude of the wave function, while the two colors denote opposite phases of the wave function.

TB calculations on TA

Figure S7. TB calculations on TA. (a) Gas phase TB energy spectrum of TA between ±3\pm3 eV. HOMO–2 to LUMO+2 states are labeled. The calculated frontier gap of TA is ∼0.5\sim0.5 eV at nearest-neighbor TB level of theory. (b, c) TB-LDOS maps of the labeled states in (a) for the unoccupied (b) and occupied (c) channels. The carbon backbone of TA (colored circles) is superimposed as a guide to the eye. The maps are calculated at a height of 7 Å above the molecular plane.

Intermixing of HOMO-1 and HOMO-2 states of TA

Figure S8. Intermixing of HOMO-1 and HOMO-2 states of TA. (a) dI/dVdI/dV map of the resonance at −1.65-1.65 V. (b) TB-LDOS maps of the HOMO-1 and HOMO-2 of TA. The dI/dVdI/dV map at −1.65-1.65 V contains features that are characteristic of both orbitals – that is, strong LDOS features at the termini (characteristic of HOMO-1, and to some extent, also HOMO-2) and slightly weaker LDOS features at the center of the armchair edges (characteristic of HOMO-2 exclusively). As shown in (c), mixing of the HOMO-2 LDOS (with a weight of 0.4) into the HOMO-1 LDOS clearly reproduces our experimental observation. The TB-LDOS maps are calculated at a height of 7 Å above the molecular plane.

TB calculations on HA

Figure S9. TB calculations on HA. (a) Gas phase TB energy spectrum of HA between ±3\pm3 eV. HOMO-2 to LUMO+2 states are labeled. The calculated frontier gap of HA is ∼90\sim90 meV at nearest-neighbor TB level of theory, much less than that of TA at the same level of theory (Fig.

Bibliography

such end states extend up to ~ 1.1 nm from the ribbon terminus[52], which supports their overlap in both TA and HA, where the total molecular length amounts to 1.4 and 2.8 nm (STM-measured), respectively. This effect modifies the electronic properties of our model compounds with respect to those of longer 7-AGNRs, similarly to what has been reported for the distinct transport properties of finite vs long 5-AGNRs[15, 31].

A pertinent question that arises with respect to the electronic structure of TA and HA is their open- (magnetic) or closed-shell (non-magnetic) ground state on Au(111). Both TA and HA are members of the family of peri–condensed anthracenes denoted as anthenes. All anthenes may be chemically drawn as resonance hybrids of closed-shell Kekulé and open-shell biradical forms, with the later form being driven by the gain in the number of aromatic sextets. The smallest member of this family, bisanthene (C28H14\mathrm{C}_{28}\mathrm{H}_{14}), is considered a closed-shell system. However, X-ray crystallographic studies of the larger homologues, namely TA[35] and quarteranthene (C56H18\mathrm{C}_{56}\mathrm{H}_{18})[53], showed notable deviation of their geometrical structures from those of closed-shell forms. A more direct route to access the electronic structure of open-shell systems at the single-molecule scale is through STS, where open-shell molecules generally exhibit two distinguishing features. First, the open-shell molecules host singly occupied molecular orbitals (SOMOs), which appear similar in both the occupied and unoccupied regimes due to electron tunneling to/from the same orbitals[54, 55]. In contrast, the frontier states of closed-shell molecules consist of the hybridized HOMO and LUMO, which present different shapes and symmetries in dI/dVdI/dV mapping. Second, the presence of unpaired electrons in open-shell molecules additionally manifests as characteristic low-energy spectral features in STS such as Kondo resonances or spin excitations[55, 56]. Based on our STS measurements on both TA and HA, we find no indication of SOMOs and/or magnetic excitations in both the species, and we thus ascribe their electronic ground states to be closed-shell on Au(111).

Surface-averaging techniques like Raman spectroscopy offer an overview of the sample quality at large scales[19] while still being sensitive to the atomic structure of the investigated materials.

Raman is a particularly powerful technique, providing information on defects[57], chirality[58], width[21] and length[20] of sp2sp^{2} hybridized carbon nanostructures. Here, we use Raman spectroscopy to investigate the fingerprint vibrations of TA and HA. We develop a portable UHV chamber that can be mounted onto the UHV systems in which the samples were prepared and then transferred onto the scanning stage of our Raman microscope. Details of the construction and characterization of the UHV Raman suitcase will be presented elsewhere. This UHV Raman suitcase allows us to characterize the Raman spectroscopy of the pristine GNRs under ultra-high vacuum conditions, and thus allows us to characterize the influence of the environment (UHV vs ambient conditions) on their atomic structure.

The Raman spectrum of a sample with high surface coverage of TA and HA (Figure 3a) measured in ultra-high vacuum (∼5⋅10−8\sim5 \cdot10^{-8} mbar) is shown in Figure 3b. For both, TA and HA, the high-frequency range of the Raman spectrum is dominated by a peak at 1600 cm−1cm^{-1} that is related to the nanographenes' sp2sp^2 lattice (G mode). Additionally, we observe the D mode at 1346 cm−1cm^{-1} and C-H bending modes at 1220, 1252 and 1274 cm−1cm^{-1}, proving hydrogen passivation of TA and HA edges[59, 60, 61].

STM topography and Raman spectra of 1 on Au/mica

Figure 3. a) STM topography of a high coverage sample of 1 on Au/mica, annealed to 290 ∘C290\,^\circ\mathrm{C}. Scanning parameters: V=−1V=-1 V, I=20I=20 pA), b) Raman spectroscopy of the same sample as shown in a), measured in ultra-high-vacuum with a 532 nm laser. Inset shows a high resolution spectrum in the D region highlighting the C-H bending mode of zigzag edges. c) DFT simulated Raman spectra of TA and HA d) Comparison between Raman spectra acquired on the sample kept in UHV (black) and after exposure to air (red).

The inset in Figure 3b shows a zoom-in of the D and C-H region, where a peak at 1142 cm−1cm^{-1} and a peak at 1170 cm−1cm^{-1} are distinguished. These peaks corresponds to the C-H bending mode of the zigzag edges[62], of TA and HA, respectively, as indicated by our normal mode analysis (Figure S10). The C-H bending mode of the zigzag edges of the ribbon termini was so far not observed experimentally due to the low ratio between zigzag (ZZ) and armchair (AC) edge segments of the structures investigated to date. For TA and HA, we have a much higher proportion of zigzag edges due to their short length: 3 zigzag edge segments per 3 armchair edge segments (ZZ/AC=1) for TA and 3 zigzag edge segments per 6 armchair edge segments (ZZ/AC=½) for HA (see Figure S10). Raman simulations for both TA and HA show this mode at 1146 cm−1cm^{-1} and 1158 cm−1cm^{-1}, respectively, with higher normalized (to IGI_G) intensity for the teranthene structure (0.11 vs 0.05), in line with the higher ZZ/AC ratio for TA.

Raman normal mode analysis of teranthene and hexanthene

Figure S10. Raman normal mode analysis of teranthene (TA) (upper panel) and hexanthene (HA) (lower panel). Red arrows indicate amplitude and direction of atomic displacements.

  • Modes at 245 and 126 cm−1\text{cm}^{-1}: longitudinal compressive mode (LCM)

  • Mode at 403 cm−1\text{cm}^{-1}: radial breathing-like mode (RBLM)

  • Modes at 1146 and 1158 cm−1\text{cm}^{-1}: CH-bending of zigzag edges

  • Modes at 1225, 1261 and 1279 cm−1\text{cm}^{-1}: CH-bending armchair edges

  • Mode at 1342 cm−1\text{cm}^{-1}: D mode

  • Mode at 1599: G mode

The low frequency region shows the fingerprint modes of both width and length of the species on the surface. The fundamental transverse acoustic mode for GNRs, called radial breathing-like mode (RBLM), is specific for each GNR width[21, 63]. Both TA and HA may be structurally seen as short 7-AGNRs, and here we observe that their RBLM is in the same region than the one of long 7-AGNRs[64], around 394 cm−1cm^{-1}, but with a different peak shape. A similar behavior was observed for short 5-AGNRs (1-2.5 nm) where the RBLM showed side-peaks due to the splitting of the RBLM into normal modes with diagonal atomic displacement[20].

Recently we showed that in a sample containing AGNRs with broad length distribution, probing the longitudinal compressive mode (LCM) of AGNRs with Raman spectroscopy enables us to obtain surface-averaged length information on the growth substrate and upon transfer to technological relevant substrates[20]. Here, in virtue of the controlled selectivity of our precursor design, the short length of both TA and HA allows us to observe two characteristic length-dependent modes, one at 127 cm−1cm^{-1} for HA and another at 246 cm−1cm^{-1} for TA, in agreement with their DFT simulated spectra (Figure 3c). In line with the STM images, which shows 80% of HA on the surface, the LCM of this species has higher intensity relative to the TA (0.1 vs 0.02 – intensity normalized to IGI_G).

TA and HA have a larger ZZ to AC edge ratio in comparison to long 7-AGNRs, and thus have easily detectable zigzag edge vibrational modes at 1142 cm−1^{-1} (from TA) and 1170 cm−1^{-1} (from HA) in Raman spectroscopy (Figure 3b). Therefore, they serve as excellent model systems to investigate the chemical reactivity of nanographenes/GNRs upon exposure to air.

Figure 3d shows the high frequency Raman profiles of the edge-related modes (D and CH bending) and G mode measured in vacuum, with (in red) or without (in black) an air-exposure step in between. The Raman profiles reveal broadening and disappearance of the edge-related modes for the sample briefly (~5 minutes) exposed to air. Particularly we observe that the CH-bending modes assigned to the zigzag edges are not resolved anymore after air exposure along with broadening of all other CH-bending related modes as shown by their increased full-width at half maximum: mode at ~1220 cm−1cm^{-1} from FWHM = 5 cm−1cm^{-1} to 19 cm−1cm^{-1}, mode at ~1274 cm−1cm^{-1} from FWHM = 9 cm−1cm^{-1} to 43 cm−1cm^{-1}, both D modes from FWHM = 9 cm−1cm^{-1} to 27 cm−1cm^{-1}, and G mode from FWHM = 10 cm−1cm^{-1} to 23 cm−1cm^{-1}. This behavior is not observed for long 7-AGNRs where peak widths remains the same upon air exposure (Refs. 19,64 and Figure S11).

Raman spectra of 7-AGNR in air and in vacuum

Figure S11. Raman spectroscopy of long 7-AGNRs in ultra-high vacuum (in black) and in air (in red) measured with 532 nm laser energy.

The FWHM of Raman peaks is a useful tool to characterize sample crystallinity or purity. Recent work by Lawrence et al. demonstrated that exposing chiral (3,1)-GNRs to air results in their oxidation[43]. Moreover, this process generates several oxidized species, undermining the atomic precision achieved through on-surface synthesis. Considering that the armchair edges of the 7-AGNR are chemically unreactive, any oxidation is limited to their zigzag termini[13, 42]. The signal from these zigzag termini are negligible compared to the armchair edges because the 7-AGNRs have an average length exceeding 15-20 nm. However, by choosing to study the ultrashort 7-AGNR segments, we have amplified the signal of the zigzag edges with respect to the armchair edges. This allows us to detect their presence when the Raman is measured in UHV and their absence upon air exposure.

Furthermore, due to the small size of these segments, any chemical change such as ketone or carboxyl substitution, would have a significant effect on all vibrational modes in TA and HA. We expect a variety of such uncontrolled chemical changes to occur upon air exposure, yielding several different oxidized species, each with their own peak shifts (see Figure S12 for a simulated spectra of both TA and HA with OH-terminated zigzag edges, showing new peak shifts and broadening). Cumulatively, this broadens all the Raman peaks when compared with the pristine samples measured in UHV.

Simulated Raman spectra of HA and TA with OH-terminated zigzag edges

Figure S12. Simulated Raman spectra of HA and TA with OH-terminated zigzag edges.

Our findings reaffirm the chemical reactivity of the zigzag edges. We also establish that Raman spectroscopy is an extremely efficient technique for characterizing the chemical degradation occurring when these edges oxidize. Consequently, Raman spectroscopy serves as an effective tool for verifying the chemical integrity of systems at different stages of device integration. Our results underscore the inherent challenges in transferring GNRs with zigzag edges, emphasizing the need for novel fabrication and encapsulation methods to preserve these edges and the related spin-polarized edge states for their potential integration into spintronic platforms.

Conclusions

In conclusion, we successfully synthesized the ultrashort 7-AGNR model compounds teranthene (TA) and hexanthene (HA) from 10-bromo-9,9′:10′,9″-teranthracene precursor on Au(111), and unveiled their atomically-resolved chemical structures by STM and nc-AFM. Detailed electronic characterization and tight-binding calculations revealed electronic gaps of 0.34 eV and 0.30 eV between frontier orbitals of HA and TA on Au(111), respectively, with the former being positively charged on this substrate. Our precursor design allowed assessing Raman fingerprints of the length-dependent vibrational modes of TA and HA at 245 and 127 cm−1\mathrm{cm}^{-1}, respectively, which would be impossible with standard di-halogenated precursors that produce polymers and GNRs with broad length distributions. In virtue of the high ZZ to AC edge ratio of the nanostructures synthesized herein, we also identified for the first time CH-bending modes of the zigzag-edges at 1142 cm−1\mathrm{cm}^{-1} for TA and 1170 cm−1\mathrm{cm}^{-1} for HA. Finally, we probed the reactivity of TA and HA by measuring Raman spectra of samples kept in ultra-high vacuum and exposed to air. Our results showed broadening and suppression of the AC and ZZ edge-related modes, respectively, which we associate to oxidation of the highly reactive zigzag edges.

Methods

STM/STS and nc-AFM

The on-surface synthesis experiments were performed under ultrahigh vacuum (UHV) conditions with base pressure below 2×10−102 \times10^{-10} mbar. The Au(111) single crystal (MaTeck GmbH) was cleaned by several cycles of Ar+\mathrm{Ar}^{+} sputtering (1 keV) and annealing (470 °C) until a clean surface with monoatomic terrace steps is achieved. Deposition of the precursor molecule was performed by thermal evaporation from a 6-fold organic evaporator (Mantis GmbH). A commercial low-temperature STM (Scienta Omicron) was used for sample characterization, operated at 5 K in constant-current mode using an etched tungsten tip. Bias voltages are given with respect to the sample. Constant-height dI/dVdI/dV spectra and maps were obtained with a lock-in amplifier (f=610f = 610 Hz). Nc-AFM measurements were performed at 5 K with a tungsten tip placed on a QPlus tuning fork sensor[65]. The tip was functionalized with a single CO molecule at the tip apex picked up from the previously CO-dosed surface[46, 66]. The sensor was driven at its resonance frequency (27000 Hz) with a constant amplitude of 70 pm. The frequency shift from resonance of the tuning fork was recorded in constant-height mode using Omicron Matrix electronics and HF2Li PLL by Zurich Instruments.

Raman Spectroscopy

After UHV preparation and characterization, the samples were transferred into a vacuum suitcase via the fast-entry lock of the UHV apparatus. The maximum pressure experienced by the samples during transfer and Raman characterization was in the order of 10−810^{-8} mbar.

Raman spectra were acquired with a WITec Alpha 300 R confocal Raman microscope in backscattering geometry. Measurements were performed with 532 nm excitation and 150g/mm grating. For maximum signal intensity spectra are recorded with a Zeiss 50x LD objective, NA=0.55, through an uncoated fused silica window of only 0.2mm thickness covering a hole of just 7mm diameter.

Raman simulations

The Raman spectra were calculated with the ORCA 4.2 DFT code[67] using the GGA PBE exchange-correlation functional, and the def2-SVP basis set in combination with the RI-J approximation for Coulomb integrals. The frequencies were calculated numerically. All atomic position are obtained from a geometry relaxation performed with the same settings as the Raman calculations.

Computational details

Tight-binding calculations: The tight-binding calculations of TA and HA were performed by numerically solving the nearest-neighbor hopping tight-binding Hamiltonian, considering only the carbon 2pz2p_z orbitals:

H^TB=−t∑⟨α,β⟩,σcα,σ†cβ,σ(1)\hat{H}_{\mathrm{TB}}=-t\sum_{\langle\alpha,\beta\rangle,\sigma}c_{\alpha,\sigma}^{\dagger}c_{\beta,\sigma} \tag*{(1)}

Here, cα,σ†c_{\alpha,\sigma}^{\dagger} and cβ,σc_{\beta,\sigma} denote the spin selective (σ∈{↑,↓}\sigma\in\{\uparrow,\downarrow\}) creation and annihilation operator at neighboring sites α\alpha and β\beta, and tt is the nearest-neighbor hopping parameter, with t=2.7t = 2.7 eV used throughout this work.

Orbital electron densities, ρ\rho, of the nthn^{\mathrm{th}}-eigenstate with energy EnE_n have been simulated from the corresponding state vector an,i,σa_{n,i,\sigma} by:

ρn,σ(r⃗)=∣∑ian,i,σϕ2pz(r⃗−r⃗i)∣2(2)\rho_{n,\sigma}(\vec{r})=\left|\sum_i a_{n,i,\sigma}\phi_{2p_z}(\vec{r}-\vec{r}_i)\right|^2 \tag*{(2)}

where ii denotes the atomic site index, and ϕ2pz\phi_{2p_z} denotes the Slater 2pz2p_z orbital for carbon.

Acknowledgments

This worked was supported by the Swiss National Science Foundation under grant no. 200020_182015, IZLCZ2_170184 and CRSII5_205987, the European Union Horizon 2020 research and innovation program under grant agreement no. 881603 (GrapheneFlagship Core 3), and the Office of Naval Research BRC Program under the grant N00014-18-1-2708. The authors also greatly appreciate the financial support from the Werner Siemens Foundation (CarboQuant). M.C. acknowledges support by the Swiss National Science Foundation under the Sinergia grant no. 189924 (Hydronics) and funding from the EC H2020 FET Open project no. 767187 (QuIET). M.L.P. acknowledges funding from the Swiss National Science Foundation under Spark grant no. 196795 and the Eccellenza Professorial Fellowship no. PCEFP2_203663, as well as support by the Swiss State Secretariat for Education, Research and Innovation (SERI) under contract number MB22.00076. Lukas Rotach is gratefully acknowledged for his technical support during the experiments.

For the purpose of Open Access (which is required by our funding agencies), the authors have applied a CC BY public copyright license to any Author Accepted Manuscript version arising from this submission.

Author Contribution

Conceptualization: GBB, MDG, TL, PR Data Curation: GBB, MDG, TL, JO, MP, SM ., Formal Analysis: GBB, MDG, TL, SM, PR. Funding Acquisition: RF, MP, PR, XF, MC, Investigation: GBB, MDG, TL. Supervision: RF, PR, MC, XF. Methodology: GBB, MDG, TL, JO, AK. Visualization: GBB, MDG, TL, Writing|original draft: GBB, MDG, TL, SM, PR Writing|review & editing: all

Conflicts of Interest

There are not conflicts to declare

References

Monobromo_SI_.pdf

Precursor synthesis and characterization

Unless otherwise stated, commercially available starting materials, catalysts, reagents, and dry solvents were used without further purification. Reactions were performed using standard vacuum-line and Schlenk techniques. All the starting materials were obtained from Sigma Aldrich, or TCI. Catalysts were purchased from Strem. Column chromatography was performed on silica (SiO2\mathrm{SiO}_{2}, particle size 0.063–0.200 mm, purchased from VWR). Silica-coated aluminum sheets with a fluorescence indicator (TLC silica gel 60 F254\mathrm{F}_{254}, purchased from Merck KGaA) were used for thin layer chromatography.

MALDI-TOF spectra were recorded on a Bruker Autoflex Speed MALDI-TOF MS (Bruker Daltonics, Bremen, Germany). All of the samples were prepared by mixing the analyte and the matrix, 1,8-dihydroxyanthracen-9(10H)-one (dithranol, purchased from Fluka Analytical, purity >98%) in the solid state.

NMR data were recorded on a Bruker AV-II 300 spectrometer operating at 300 MHz for 1H^{1}\mathrm{H} and 75 MHz for 13C^{13}\mathrm{C} with standard Bruker pulse programs 333 K. Chemical shifts (δ\delta) are reported in ppm. Coupling constants (JJ) are reported in Hz. Tetrachloroethane-d2_2 (δ(1H)=5.91\delta(^{1}\mathrm{H}) = 5.91 ppm, δ(13C)=74.2\delta(^{13}\mathrm{C}) = 74.2 ppm) was used as solvents. The following abbreviations are used to describe peak patterns as appropriate: s = singlet, d = doublet, t = triplet, q = quartet, and m = multiplet. Tetrachloroethane-d2_2 (99.9 atom% D) was purchased from Euriso-top.

Melting points were determined on a Büchi Melting Point M-560 in a range of 50–400 °C with a temperature rate of 10 °Cmin−1^{-1}.

10-bromo-9,9′:10′,9″-teranthracene (1)

Figure S1. Figure S1. 10,10′-dibromo-9,9′-bianthracene (100 mg, 0,195 mmol), 9-anthraceneboronic acid (56.4 mg, 0.254 mmol, 1.3 eq.), tripotassium phosphate (207.2 mg, 0.976 mmol, 5.0 eq.), DPEphos (42 mg, 0.078 mmol, 0.4 eq.) and Pd2_2(dba)3_3 (42.1 mg, 0.078 mmol, 0.4 eq.) were added to a 50 ml schlenk flask and evacuated and refilled with argon three times. After adding the degassed solvents (5 ml toluene, 1 ml water and 0.6 ml ethanol), the reaction mixture was heated to 90 °C for 2 days. After quenching with water the product was extracted with DCM, washed with brine and dried over MgSO4_4. The solvents were removed under vacuum. The crude material was purified by column chromatography (chloroform:isohexane = 1:5 to 1:1). After recrystallization (chloroform/MeOH) the title compound 1 was afforded in a yield of 14% (16 mg).

mp: > 400 °C; Rf_f: 0.30 (DCM:isohexane = 1:3); 1H^{1}\mathrm{H} NMR (300 MHz, C2_2D2_2Cl4_4, 333 K) δ\delta (ppm) = 8.71 (d, JJ = 6.9 Hz, 3H), 8.16 (d, JJ = 8.5 Hz, 2H), 7.65 – 7.56 (m, 2H), 7.52 – 7.43 (m, 2H), 7.34 – 7.23 (m, 8H), 7.21 – 7.13 (m, 4H), 7.11 – 7.01 (m, 4H); 13C^{13}\mathrm{C} NMR (76 MHz, C2_2D2_2Cl4_4, 333K) δ\delta (ppm) = 132.74, 131.91, 131.66, 128.97, 128.40, 127.79, 127.61, 127.30, 127.06, 126.68, 126.42, 126.00, 125.74; HR-MS (MALDI-TOF) m/z: [M+^{+}] calcd for C42_{42}H25_{25}Br: 608.1130; found: 608.1128; error: -0.33 ppm.

MALDI-TOF spectra

Figure S2. a) HR-MALDI-TOF spectrum of 1; b) HR-MALDI-TOF magnified spectrum of 1 (black line) is in agreement to the expected isotopic distribution pattern (red line).

NMR spectra

${}^{1}\mathrm{H}$-NMR spectrum of 1 with expanded aromatic region and molecular structure

Figure S3a. 1H{}^{1}\mathrm{H}-NMR spectrum of 1 dissolved in tetrachloroethane-d2d_2, 300 MHz, 333 K.

Figure S3b. 13C{}^{13}\mathrm{C}-NMR spectrum of 1 dissolved in tetrachloroethane-d2_2, 75 MHz, 333 K.

Figure S3c. 1H/{}^{1}\mathrm{H}/{}^{1}\mathrm{H}$-COSY-NMR spectrum of 1 dissolved in tetrachloroethane-d2_2, 300 MHz, 333 K.

Figure S3d. HSQC-NMR spectrum of 1 dissolved in tetrachloroethane-d₂, 75 MHz, 333 K.

Figure S3e. HMBC-NMR spectrum of 1 dissolved in tetrachloroethane-d₂, 75 MHz, 333 K.

STM images of the Au(111) surface after deposition of precursor 1 at room temperature (RT), followed by successive annealing steps

Figure S4. STM images of the Au(111) surface after deposition of precursor 1 at room temperature (RT), followed by successive annealing steps at the given temperatures (a-f). Inset (e): magnified STM image with superimposed chemical structures of the dimers after dehalogenative aryl-aryl coupling of 1 on the surface. Scanning parameters: a) V=−1.0V=-1.0 V, I=30I=30 pA, b) V=−1.0V=-1.0 V, I=30I=30 pA, c) V=−1.0V=-1.0 V, I=20I=20 pA, d) V=−0.5V=-0.5 V, I=30I=30 pA, e) V=−0.3V=-0.3 V, I=30I=30 pA, f) V=−1V=-1 V, I=20I=20 pA.

S7). (b, c) TB-LDOS maps of the labeled states in (a) for the unoccupied (b) and occupied (c) channels. The carbon backbone of HA (colored circles) is superimposed as a guide to the eye. The maps are calculated at a height of 7 A˚\text{\AA} above the molecular plane.

References

  1. [1]Narita, A., Wang, X.-Y., Feng, X. & Müllen, K. New advances in nanographene chemistry. Chem. Soc. Rev. 44, 6616–6643 (2015).DOI
  2. [2]Corso, M., Carbonell-Sanromà, E. & de Oteyza, D. G. Bottom-Up Fabrication of Atomically Precise Graphene Nanoribbons. in On-Surface Synthesis II (eds. de Oteyza, D. G. & Rogero, C.) 113–152 (Springer International Publishing, 2018).DOI
  3. [3]Cai, J. et al. Atomically precise bottom-up fabrication of graphene nanoribbons. Nature 466, 470–473 (2010).DOI
  4. [4]Talirz, L. et al. On-Surface Synthesis and Characterization of 9-Atom Wide Armchair Graphene Nanoribbons. ACS Nano 11, 1380–1388 (2017).
  5. [5]Kimouche, A. et al. Ultra-narrow metallic armchair graphene nanoribbons. Nat. Commun. 6, 10177 (2015).DOI
  6. [6]Yamaguchi, J. et al. Small bandgap in atomically precise 17-atom-wide armchair-edged graphene nanoribbons. Commun. Mater. 1, 1–9 (2020).DOI
  7. [7]Chen, Y.-C. et al. Tuning the Band Gap of Graphene Nanoribbons Synthesized from Molecular Precursors. ACS Nano 7, 6123–6128 (2013).
  8. [8]Rizzo, D. J. et al. Topological band engineering of graphene nanoribbons. Nature 560, 204–208 (2018).DOI
  9. [9]Gröning, O. et al. Engineering of robust topological quantum phases in graphene nanoribbons. Nature 560, 209–213 (2018).DOI
  10. [10]Ruffieux, P. et al. On-surface synthesis of graphene nanoribbons with zigzag edge topology. Nature 531, 489–492 (2016).
  11. [11]Cai, J. et al. Graphene nanoribbon heterojunctions. Nat. Nanotechnol. 9, 896–900 (2014).DOI
  12. [12]Slota, M. et al. Magnetic edge states and coherent manipulation of graphene nanoribbons. Nature 557, 691–695 (2018).DOI
  13. [13]Lawrence, J. et al. Probing the Magnetism of Topological End States in 5-Armchair Graphene Nanoribbons. ACS Nano acsano.9b10191 (2020) doi:10.1021/acsano.9b10191.DOI
  14. [14]Chen, Z., Narita, A. & Müllen, K. Graphene Nanoribbons: On-Surface Synthesis and Integration into Electronic Devices. Adv. Mater. 32, 2001893 (2020).DOI
  15. [15]Borin Barin, G. et al. Growth Optimization and Device Integration of Narrow-Bandgap Graphene Nanoribbons. Small 2202301 (2022) doi:10.1002/smll.202202301.
  16. [16]Zhang, J. et al. Tunable Quantum Dots from Atomically Precise Graphene Nanoribbons Using a Multi-Gate Architecture. Adv. Electron. Mater. 9, 2201204 (2023).
  17. [17]Manipatruni, S., Nikonov, D. E. & Young, I. A. Material Targets for Scaling All-Spin Logic. Phys. Rev. Appl. 5, 014002 (2016).DOI
  18. [18]Behin-Aein, B., Datta, D., Salahuddin, S. & Datta, S. Proposal for an all-spin logic device with built-in memory. Nat. Nanotechnol. 5, 266–270 (2010).
  19. [19]Borin Barin, G. et al. Surface-Synthesized Graphene Nanoribbons for Room Temperature Switching Devices: Substrate Transfer and ex Situ Characterization. ACS Appl. Nano Mater. 2, 2184–2192 (2019).
  20. [20]Overbeck, J. et al. A Universal Length-Dependent Vibrational Mode in Graphene Nanoribbons. ACS Nano 13, 13083–13091 (2019).
  21. [21]Overbeck, J. et al. Optimized Substrates and Measurement Approaches for Raman Spectroscopy of Graphene Nanoribbons. Phys. Status Solidi B 0,.
  22. [22]Zhao, S. et al. Optical Investigation of On-Surface Synthesized Armchair Graphene Nanoribbons. Phys. Status Solidi B 254, 1700223 (2017).
  23. [23]Zhao, S. et al. Optical Imaging and Spectroscopy of Atomically Precise Armchair Graphene Nanoribbons. Nano Lett. 20, 1124–1130 (2020).DOI
  24. [24]Senkovsky, B. V. et al. Making Graphene Nanoribbons Photoluminescent. Nano Lett. 17, 4029–4037 (2017).
  25. [25]Mutlu, Z. et al. Transfer-Free Synthesis of Atomically Precise Graphene Nanoribbons on Insulating Substrates. ACS Nano 15, 2635–2642 (2021).DOI
  26. [26]Darawish, R. et al. Quantifying Alignment and Quality of Graphene Nanoribbons: A Polarized Raman Spectroscopy Approach. Preprint at https://doi.org/10.48550/arXiv.2307.09490 (2023).
  27. [27]Backes, C. et al. Production and processing of graphene and related materials. 2D Mater. 7, 022001 (2020).
  28. [28]Bennett, P. B. et al. Bottom-up graphene nanoribbon field-effect transistors. Appl. Phys. Lett. 103, 253114 (2013).DOI
  29. [29]Llinas, J. P. et al. Short-channel field-effect transistors with 9-atom and 13-atom wide graphene nanoribbons. Nat. Commun. 8:633, 1–6 (2017).
  30. [30]Martini, L. et al. Structure-dependent electrical properties of graphene nanoribbon devices with graphene electrodes. Carbon 146, 36–43 (2019).DOI
  31. [31]El Abbassi, M. et al. Controlled Quantum Dot Formation in Atomically Engineered Graphene Nanoribbon Field-Effect Transistors. ACS Nano 14, 5754–5762 (2020).DOI
  32. [32]Braun, O. et al. Optimized graphene electrodes for contacting graphene nanoribbons. Carbon 184, 331–339 (2021).arxiv.org/abs/2102.13033
  33. [33]Passi, V. et al. Field-Effect Transistors Based on Networks of Highly Aligned, Chemically Synthesized N = 7 Armchair Graphene Nanoribbons. ACS Appl. Mater. Interfaces 10, 9900–9903 (2018).DOI
  34. [34]Richter, N. et al. Charge transport mechanism in networks of armchair graphene nanoribbons. Sci. Rep. 10, 1988 (2020).DOI
  35. [35]Konishi, A. et al. Synthesis and Characterization of Teranthene: A Singlet Biradical Polycyclic Aromatic Hydrocarbon Having Kekulé Structures. J. Am. Chem. Soc. 132, 11021–11023 (2010).DOI
  36. [36]Konishi, A., Hirao, Y., Kurata, H. & Kubo, T. Investigating the edge state of graphene nanoribbons by a chemical approach: Synthesis and magnetic properties of zigzag-edged nanographene molecules. Solid State Commun. 175–176, 62–70 (2013).DOI
  37. [37]Wang, S. et al. Giant edge state splitting at atomically precise graphene zigzag edges. Nat. Commun. 7, 11507 (2016).DOI
  38. [38]Kinikar, A. et al. Quantized edge modes in atomic-scale point contacts in graphene. Nat. Nanotechnol. 12, 564–568 (2017).DOI
  39. [39]Pesin, D. & MacDonald, A. H. Spintronics and pseudospintronics in graphene and topological insulators. Nat. Mater. 11, 409–416 (2012).DOI
  40. [40]Trauzettel, B., Bulaev, D. V., Loss, D. & Burkard, G. Spin qubits in graphene quantum dots. Nat. Phys. 3, 192–196 (2007).DOI
  41. [41]Recher, P. & Trauzettel, B. Quantum dots and spin qubits in graphene. Nanotechnology 21, 302001 (2010).DOI
  42. [42]Ma, C. et al. Oxidization stability of atomically precise graphene nanoribbons. Phys. Rev. Mater. 2, (2018).DOI
  43. [43]Lawrence, J. et al. Circumventing the stability problems of graphene nanoribbon zigzag edges. Nat. Chem. 14, 1451–1458 (2022).DOI
  44. [44]Berdonces-Layunta, A. et al. Chemical Stability of (3,1)-Chiral Graphene Nanoribbons. ACS Nano 15, 5610–5617 (2021).
  45. [45]Talirz, L., Ruffieux, P. & Fasel, R. On-Surface Synthesis of Atomically Precise Graphene Nanoribbons. Adv. Mater. 28, 6222–6231 (2016).
  46. [46]Gross, L., Mohn, F., Moll, N., Liljeroth, P. & Meyer, G. The Chemical Structure of a Molecule Resolved by Atomic Force Microscopy. Science 325, 1110 (2009).DOI
  47. [47]Mishra, S. et al. Topological Defect-Induced Magnetism in a Nanographene. J. Am. Chem. Soc. 142, 1147–1152 (2020).DOI
  48. [48]Koch, M., Ample, F., Joachim, C. & Grill, L. Voltage-dependent conductance of a single graphene nanoribbon. Nat. Nanotechnol. 7, 713–717 (2012).DOI
  49. [49]van der Lit, J. et al. Suppression of electron–vibron coupling in graphene nanoribbons contacted via a single atom. Nat. Commun. 4, 2023 (2013).
  50. [50]Bronner, C. et al. Electronic structure of a subnanometer wide bottom-up fabricated graphene nanoribbon: End states, band gap, and dispersion. Phys. Rev. B 86, (2012).DOI
  51. [51]Ruffieux, P. et al. Electronic Structure of Atomically Precise Graphene Nanoribbons. ACS Nano 6, 6930–6935 (2012).DOI
  52. [52]Chong, M. C. et al. Bright Electroluminescence from Single Graphene Nanoribbon Junctions. Nano Lett. 18, 175–181 (2018).DOI
  53. [53]Konishi, A. et al. Synthesis and Characterization of Quarteranthene: Elucidating the Characteristics of the Edge State of Graphene Nanoribbons at the Molecular Level. J. Am. Chem. Soc. 135, 1430–1437 (2013).DOI
  54. [54]Repp, J., Meyer, G., Paavilainen, S., Olsson, F. E. & Persson, M. Imaging Bond Formation Between a Gold Atom and Pentacene on an Insulating Surface. Science 312, 1196–1199 (2006).pubmed.ncbi.nlm.nih.gov/16728636
  55. [55]Mishra, S. et al. Topological frustration induces unconventional magnetism in a nanographene. Nat. Nanotechnol. 15, 22–28 (2020).DOI
  56. [56]Li, J. et al. Single spin localization and manipulation in graphene open-shell nanostructures. Nat. Commun. 10, 1–7 (2019).DOI
  57. [57]Verzhbitskiy, I. A. et al. Raman Fingerprints of Atomically Precise Graphene Nanoribbons. Nano Lett. 16, 3442–3447 (2016).DOI
  58. [58]Jorio, A. et al. G-band resonant Raman study of 62 isolated single-wall carbon nanotubes. Phys. Rev. B 65, (2002).DOI
  59. [59]Saito, R., Furukawa, M., Dresselhaus, G. & Dresselhaus, M. S. Raman spectra of graphene ribbons. J. Phys. Condens. Matter 22, 334203 (2010).DOI
  60. [60]Gillen, R., Mohr, M., Thomsen, C. & Maultzsch, J. Vibrational properties of graphene nanoribbons by first-principles calculations. Phys. Rev. B 80, (2009).DOI
  61. [61]Casiraghi, C. & Prezzi, D. Raman Spectroscopy of Graphene Nanoribbons: A Review. in GraphITA : Selected papers from the Workshop on Synthesis, Characterization and Technological Exploitation of Graphene and 2D Materials Beyond Graphene (eds. Morandi, V. & Ottaviano, L.) 19–30 (Springer International Publishing, 2017). doi:10.1007/978-3-319-58134-7_2.DOI
  62. [62]Kim, J. et al. Distinguishing Zigzag and Armchair Edges on Graphene Nanoribbons by X-ray Photoelectron and Raman Spectroscopies. ACS Omega 3, 17789–17796 (2018).DOI
  63. [63]Liu, D., Daniels, C., Meunier, V., Every, A. G. & Tománek, D. In-plane breathing and shear modes in low-dimensional nanostructures. Carbon 157, 364–370 (2020).arxiv.org/abs/1910.08134
  64. [64]Fairbrother, A. et al. High vacuum synthesis and ambient stability of bottom-up graphene nanoribbons. Nanoscale 9, 2785–2792 (2017).
  65. [65]Giessibl, F. J. Atomic resolution on Si(111)-(7×7) by noncontact atomic force microscopy with a force sensor based on a quartz tuning fork. Appl. Phys. Lett. 76, 1470–1472 (2000).
  66. [66]Bartels, L. et al. Dynamics of Electron-Induced Manipulation of Individual CO Molecules on Cu(111). Phys. Rev. Lett. 80, 2004–2007 (1998).DOI
  67. [67]Neese, F. The ORCA program system. WIREs Comput. Mol. Sci. 2, 73–78 (2012).

Paper details

vlnc:W_s63yx7xj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents