Skip to main content
Nature Communications logoLink to Nature Communications
. 2026 Apr 23;17:3492. doi: 10.1038/s41467-026-71660-0

Ultra-narrow donor-acceptor nanoribbons

James Lawrence 1,2,, Luka Đorđević 3,9, Fabienne Bachtiger 1,10, Harry Pinfold 1, Marc Walker 4, Jiong Lu 2,5, Gabriele C Sosso 1, Davide Bonifazi 3,6, Giovanni Costantini 1,7,8,
PMCID: PMC13106672  PMID: 42026048

Abstract

Donor–acceptor (D–A) architectures underpin many high-performance conjugated polymers but remain largely unexplored in atomically precise nanoribbons. Here, we report the on-surface synthesis of ultra-narrow D–A nanoribbons using two complementary brominated precursors based on the electron donor peri-xanthenoxanthene and the acceptor anthanthrone. High-resolution scanning tunnelling microscopy, non-contact atomic force microscopy and scanning tunnelling spectroscopy reveal submolecular structural and electronic features of the resulting nanoribbons. Homopolymerisation of each precursor yields structurally well-defined donor-only and acceptor-only nanoribbons, whose electronic character strengthens with length. Co-deposition of both precursors produces mixed D–A nanoribbons with tuneable electronic structures governed by monomer sequence. The spatial character and energetic alignment of their frontier orbitals match gas-phase density functional theory calculations, while a simplified linear combination of molecular orbitals model captures dominant trends. This bottom-up synthetic strategy enables precise control over nanoribbon composition and functionality, offering a versatile platform for engineering π-conjugated nanostructures with tailored optoelectronic properties.

Subject terms: Scanning probe microscopy; Surface spectroscopy; Graphene; Surfaces, interfaces and thin films


Donor–acceptor architectures underpin many high-performance conjugated polymers but remain largely unexplored in atomically precise nanoribbons. Here, the authors report an on-surface strategy to synthesize ultra-narrow donor–acceptor nanoribbons with sequence-controlled electronic structure.

Introduction

Confining graphene into one dimensional nanoribbons via bottom-up on-surface synthesis has attracted considerable interest in the past fifteen years13. Unlike 2D graphene, graphene nanoribbons (GNRs) exhibit a band gap, making them promising semiconductor candidates47. This band gap can be tuned by varying the length, width and edge structure of graphene nanoribbons1,8. Recent advances in on-surface synthesis have enabled the fabrication of atomically precise nanoribbons through the controlled design of molecular precursors, allowing detailed studies of structure–electronic property relationships.

Beyond tuning basic structural parameters, heteroatom doping has also been shown to significantly modify the electronic properties of GNRs. Internal doping with boron911 and nitrogen12, along with edge functionalisation using groups containing nitrogen1318, sulfur19 and oxygen2022, enables GNRs to act as electron donor (p-type) or electron acceptor (n-type) materials. Moreover, increasing the length of a given GNR enhances its donor or acceptor character by reducing the band gap through quantum confinement effects.

Donor—acceptor (D-A) nanoribbons represent a natural extension of these strategies, combining distinct monomer units to enable further modulation of electronic properties. However, such architectures remain largely underexplored in the GNR field. Most reported examples involve junctions between all-carbon segments with different widths or edge structures2326, leading to variations in band gap and interesting topological effects2730, or the combination of sections incorporating various heteroatoms, which typically exhibit only limited electronic differences13,31.

In contrast, true D-A architectures—featuring alternating electron-rich (donor, D) and electron-poor (acceptor, A) units—have become a benchmark in high-performance conjugated polymers for organic electronics and photovoltaics32,33. Their success is largely due to this design’s ability to precisely tune energy levels and narrow the band gap. In particular, the hybridisation of frontier molecular orbitals positions the LUMO of D-A systems close to that of the acceptor and the HOMO near that of the donor, allowing near-independent control over both energy levels (see Fig. S1 and Section 21 in the attached Supplementary Information, SI)32,34. Moreover, selecting strong D and A moieties—i.e. with high-lying HOMOD and low-lying LUMOA—can yield narrow band gaps35. These features enhance device performance by reducing charge carrier injection barriers, improving ambient stability, enabling ambipolar transport and maximising light adsorption in the near-infrared region32,3638.

In polymer chemistry, D-A conjugated polymers are typically synthesised by preparing D and A monomers separately, followed by polymerisation through transition-metal-catalysed cross-coupling reactions39. Here, we adapt this strategy to an on-surface synthetic paradigm, employing two molecular precursors that are strong electron donor and acceptor analogues of anthanthrene (AA, Fig. S2a), to form ultra-narrow D-A nanoribbons. The resulting nanoribbons are effectively fully conjugated D-A ladder polymers40.

Peri-xanthenoxanthene (PXX) and anthanthrone (AO) were employed as D and A units, respectively (Fig. 1). Structurally, PXX is an O-doped, stable analogue of AA (Fig. S2a)41,42. The incorporated oxygen atoms are conjugated with the π-system, contributing electron density through resonance and thereby raising the HOMO energy level, which enhances its donor character. PXX has been previously used in flexible organic light-emitting diode (OLED) displays43,44, with substituted derivatives exhibiting good injection and transport properties as well as high chemical and thermal stability41,45. Structural modifications, such as extending the aromatic core or introducing alkyl-imide groups, have been shown to tune its optical and electronic properties, including emission yield and HOMO level4648. Short PXX-based nanoribbons have also been synthesised in solution through a lengthy multi-step procedure, showing progressively elevated HOMO levels with increasing length47,49,50. On the contrary, the quinoidal structure of AO withdraws electrons, stabilising the LUMO and conferring acceptor character to the molecule5154. AO can thus be viewed as the natural acceptor analogue of PXX (Fig. S2a). Its brominated derivative—commonly known as Vat Orange 3 (VO3) due to its application as a vat dye55,56—has also been explored as an organic semiconductor in OFETs, valued for its charge mobility, low toxicity and biodegradability5759. Additionally, AO derivatives have been employed as hole-transport materials in perovskite solar cells60.

Fig. 1. Chemical structures of the donor and acceptor units employed in this study.

Fig. 1

Chemical structure of peri-xanthenoxanthene (PXX), its brominated derivative (Br2PXX), anthanthrone (AO) and its brominated derivative (VO3).

In the following, brominated derivatives of PXX and AO—Br2PXX and VO3, respectively—are employed to synthesise polymeric nanoribbons via surface-catalysed Ullmann coupling followed by dehydrogenation1. Structural characterisation is carried out using scanning tunnelling microscopy (STM), bond-resolving STM (BR-STM) and non-contact atomic force microscopy (nc-AFM) under ultrahigh vacuum (UHV) conditions, while scanning tunnelling spectroscopy (STS) is used to probe their local electronic properties. We demonstrate the successful formation of all-donor and all-acceptor nanoribbons, in some cases exceeding 20 monomer units in length. Furthermore, we show that the on-surface copolymerisation of PXX and AO enables a systematic modulation of the electronic structure, yielding ultra-narrow donor–acceptor nanoribbons with composition-dependent frontier-state distributions. The electronic structure of these ribbons, as well as the spatial character of their frontier orbitals, is in excellent agreement with gas-phase DFT calculations of the corresponding molecular orbitals. Additionally, a simple linear combination of molecular orbitals (LCMO) model captures the main electronic trends observed across different donor–acceptor compositions. This synthetic strategy demonstrates a viable route for precisely tuning the properties of carbon-based nanoribbons and offers a foundation for achieving application-specific electronic behaviour through controlled subunit composition.

Results and discussion

Pure donor nanoribbons were synthesised by depositing Br2PXX via thermal sublimation onto a room temperature Au(111) surface, followed by post-annealing in 50 K increments. A representative STM image of the as-deposited Br2PXX molecules, forming a characteristic halogen-bonded kagome assembly61, is shown in Fig. 2a. Figure 2b displays an image acquired after annealing to 523 K, revealing the formation of nanoribbons. X-ray photoelectron spectroscopy (XPS) measurements performed during the nanoribbon growth process (Fig. S3d) indicate the onset of debromination and changes in the carbon bonding environment upon annealing above approximately 410 K, providing supporting evidence for the nanoribbon formation observed by STM. As shown in Fig. S3b, at intermediate annealing temperatures, nanoribbons assemble into compact islands alongside Br atoms released during the Ullmann coupling reaction. Upon further annealing to 523 K, most of the co-adsorbed Br atoms desorb from the surface (Figs. S3c and 2b), as confirmed by XPS (Fig. S3e). This also results in the dispersion of the PXX nanoribbons across the surface, as they do not self-assemble without co-adsorbed bromine atoms. Additional details on the intermediate steps of PXX nanoribbon synthesis are provided in Section 3 of the SI.

Fig. 2. Synthesis of pure PXX and AO nanoribbons on Au(111).

Fig. 2

a, b STM images of Br2PXX after deposition at room temperature and subsequent annealing to 523 K. c Constant-height nc-AFM frequency shift image (CO tip) of a PXX nanoribbon; the brighter upper and lower sections of the nanoribbon are related to the position of the ribbon relative to the underlying Au(111) herringbone reconstruction. d Corresponding chemical structure. e, f STM images of VO3 after deposition at room temperature and subsequent annealing to 573 K. g Constant-height nc-AFM frequency shift image (CO tip) of an AO nanoribbon. h Corresponding chemical structure. Imaging parameters: (a) IT = 2 nA, Vb = –0.50 V; b IT = 50 pA, Vb = –1.00 V; (c, g) oscillation amplitude = 50 pm; (e) IT = 130 pA, Vb = +1.09 V; (f) IT = 60 pA, Vb = –0.50 V. All images acquired at T = 4.3−7.0 K.

Owing to the pro-chirality of the Br2PXX precursor61, nanoribbons can form via two distinct coupling pathways: between PXX units of the same pro-chirality, yielding ‘straight’ junctions (Fig. S4a), or between units of opposite pro-chirality, resulting in ‘alternating’ junctions (Fig. S4b). The predominance of alternating junctions—causing the wavy morphology of most nanoribbons (Fig. 2c, g)—suggests that the initial relative positions of the reacting C-Br groups play a key role in determining the coupling efficiency. Specifically, the reaction appears to be favoured when the radicals generated by precursor debromination are aligned face-to-face. The same steric and configurational considerations apply to the VO3 precursors during the formation of the acceptor nanoribbons (see below). Further possible reaction mechanisms are discussed in Sections 4, 6 and 9 of the SI.

STM images reveal that many nanoribbons exhibit structural defects, likely caused by contaminants such as unintended regioisomers of Br₂PXX or trisubstituted derivatives that could not be removed61. For instance, molecules brominated at the 1-position instead of the 3-position (1 in Fig. S5) tend to form ‘straight’ rather than the regular ‘alternating’ junctions, while tribrominated species (2 in Fig. S5) may result in branched structures similar to those observed at the centre of Fig. S6b, highlighted by the red circle. Additional defects, such as kinks, are attributed to the formation of 5-membered rings arising from misoriented monomer coupling (Fig. S6), a phenomenon also observed in other brominated precursors used for nanoribbon synthesis62.

To determine the structure of the PXX nanoribbons with higher precision, we employed constant-height nc-AFM frequency shift imaging with a CO-functionalised tip (see Methods)63,64. A representative image is shown in Fig. 2c, confirming that the synthesised nanoribbons match the expected structure depicted in Fig. 2d. For PXX nanoribbons, nc-AFM proved more effective than BR-STM, as strong tunnelling current modulation in longer ribbons partially obscured their structure in BR-STM images. As discussed in Sections 7 and 16 of the SI, this effect is linked to the electronic structure of the PXX nanoribbons.

Following the characterisation of donor-type nanoribbons, we turned to the synthesis of their acceptor counterparts using the VO3 precursor. When deposited at room temperature on Au(111), VO3 forms compact self-assembled islands spanning both fcc and hcp regions of the herringbone reconstruction (Fig. 2e). The resulting square lattice is primarily stabilised by type II halogen bonds, as revealed by BR-STM imaging and supported by a structural assignment (Fig. S10).

The polymerisation behaviour of VO3 closely resembles that of Br₂PXX. Annealing to 473 K results in the formation of short nanoribbons that are co-adsorbed alongside bromine atoms. Further annealing to 573 K leads to bromine desorption (Fig. S11 and 2f) and the formation of long nanoribbons. Unlike PXX nanoribbons, the AO nanoribbons tend to assemble into compact domains following bromine desorption. A tentative molecular arrangement for this assembly, based on weak hydrogen bonding between ketone groups and adjacent hydrogens, is presented in Fig. S12.

The structure of the AO nanoribbons is clearly resolved via CO-tip nc-AFM, with ketone groups appearing as sharp, distinct features (Fig. 2g, h)65. In many cases, bright junctions are observed connecting the termini of adjacent ribbons. These are likely the result of precursor misalignment during polymerisation, leading to the formation of a single C-C bond rather than a fully fused junction (Fig. S13). Although no distinct polymeric intermediate is observed at lower annealing temperatures, the reasons why such junctions appear more frequently in AO nanoribbons than in their PXX counterparts remain to be clarified.

STS measurements were performed to investigate the impact of length on the electronic structure of both PXX and AO nanoribbons, with a particular focus on their donor/acceptor strength. dI/dV spectra of short nanoribbons ranging from 1 to 4 units are shown in Fig. 3a, f. The most pronounced shifts in frontier energy levels occur in the short-length regime, consistent with gas-phase DFT calculations (Fig. S2b), while changes beyond four units are minimal. In both nanoribbon types, the HOMO shifts upwards and the LUMO downwards with increasing length, resulting in evident narrowing of the band gap. This trend reflects an increase in donor character for longer PXX nanoribbons and is qualitatively reproduced by theory. For PXX, it also aligns well with cyclic voltammetry and photophysical measurements on similar short PXX-derived nanoribbons synthesised in solution49.

Fig. 3. Electronic structure of pure PXX and AO nanoribbons on Au(111).

Fig. 3

a, f dI/dV spectra of PXX and AO nanoribbons of increasing lengths (monomers to tetramers), showing a progressive narrowing of the energy gap. Dotted lines mark the energetic positions of the frontier resonances. The faint lines for each nanoribbon are the Au(111) reference spectra. c, e Constant-height dI/dV images of a PXX trimer acquired at voltages corresponding to the lowest unoccupied and highest occupied resonances, respectively; h, j analogous images for an AO trimer (scale bars: 500 pm). b, d, g, i are simulated dI/dV images of the same resonances, using a probe with a 25% s- and 75% p-wave character to model the CO-functionalised tip. Measurement positions are indicated in Fig. S14.

Figure 3 also shows constant height dI/dV images of trimers for both nanoribbon types, acquired with a CO-functionalised tip at voltages corresponding to the frontier resonances. The spatial distributions of these states closely match the corresponding simulated dI/dV images. Additional comparisons—including higher and lower energy resonances and alternative tip terminations—are presented in Fig. S15 and show good agreement with theory too.

For longer PXX nanoribbons, the highest occupied state becomes increasingly difficult to resolve as it approaches the Fermi level, as seen for the tetramer in Fig. 3a. High-resolution dI/dV spectra around 0 V for a set of longer ribbons are shown in Fig. S16, with an example of a 21-unit nanoribbon in Fig. S17. The highest occupied state is found to cross the Fermi level at a length of 5 or 6 units, after which it empties via charge transfer to the surface and stabilises near +40 mV. This charging behaviour may contribute to the reduced tendency of longer PXX nanoribbons to self-assemble, possibly due to repulsive intermolecular interactions. The same state also accounts for the contrast observed in BR-STM at low bias (Fig. S7). Notably, the unoccupied resonances of PXX nanoribbons are significantly broader than the occupied states, a feature previously attributed to lifetime effects in surface-supported molecular systems66.

Longer AO nanoribbons exhibit a pronounced acceptor character, with the lowest unoccupied state shifting to lower energies as the ribbon length increases. This resonance appears at +500 mV for the AO tetramer, significantly below that of the PXX tetramer (+920 mV), while the highest occupied state is over 1 eV deeper. Interestingly, the energy gap of the AO tetramer (1.62 eV) is substantially larger than that of its PXX counterparts (0.93 eV), despite both being predicted to have identical gas-phase gaps of 1.63 eV. This discrepancy likely arises from differences in substrate interaction, charge transfer and electrostatic screening. Decoupling the nanoribbons from the metallic substrate would likely reduce this difference.

Having established that the on-surface homopolymerisation of Br2PXX and VO3 successfully yields donor and acceptor nanoribbons, respectively, we extended this strategy to mixed systems by co-depositing both precursors onto a Au(111) surface. Subsequent annealing to 473 K induced copolymerisation via the same surface-catalysed Ullmann coupling and dehydrogenation, resulting in large islands composed of various nanoribbons interspersed with dissociated Br atoms (Fig. 4a, b). High-resolution BR-STM imaging confirms that many nanoribbons incorporate both PXX and AO units, demonstrating the successful formation of D-A nanoribbons. The two components are readily distinguished in BR-STM and nc-AFM images due to their contrasting central ring features: the pyranopyranyl units of PXX appear darker, while the quinoidal moieties of AO show a brighter contrast with sharp features corresponding to their ketone groups67. Statistical analysis of the three possible junction types (D-D, D-A and A-A, see Fig. S18 and Table S1) reveals a clear preference for intermixing, with alternating D-A sequences occurring more frequently than D or A-block segregation. Interestingly, D-A junctions between PXX and AO units consistently result in straight ribbon segments, unlike D-D or A-A junctions, which often introduce bends (Fig. S19). This is likely due to the different positions of the bromine atoms in the two precursors. As a result, block-type ribbons tend to exhibit directional changes at the homojunctions, while alternating D-A sequences remain mostly straight.

Fig. 4. Mixed PXX-AO nanoribbons on Au(111).

Fig. 4

a Large-scale STM image showing mixed PXX-AO nanoribbons and interspersed Br atoms. b Smaller scale constant-height BR-STM image of the same sample. c dI/dV spectra of the PXX-AO dimer and of all possible mixed PXX-AO trimers. Frontier orbital resonances are marked by dashed vertical lines; energy gaps are indicated. The faint lines for each nanoribbon are the Au(111) reference spectra. dh Chemical structures, nc-AFM images, constant height dI/dV maps (CO tip) and simulated dI/dV images of the PXX-AO dimer and each trimer. For (d, e, h), a probe with 25% s- and 75% p-wave character was used to model the CO-functionalised tip. For (f, g), a probe with 75% s- and 25% p-wave character was used instead. Scale bars: 500 pm. dI/dV measurement locations are shown in Fig. S20. Imaging parameters: a IT = 170 pA, Vb = –0.89 V; b Vb = +0.03 V; (dh) nc-AFM oscillation amplitude = 50 pm.

While a complete analysis of the electronic structure of mixed D-A nanoribbons across different lengths and sequences lies beyond the scope of this study, we focus here on the characterisation of all four distinct PXX-AO trimers and the corresponding mixed dimer, using STS (Fig. 4c) and dI/dV imaging (Fig. 4d–h). This choice also reflects the expectation that these short oligomers capture the main changes in electronic structure, based on the trends previously observed for homopolymer nanoribbons. While the full set of STS data may appear complex at first glance, a clear initial trend can be identified: trimers containing two AO units (blue and purple spectra in Fig. 4c) exhibit lower HOMO and LUMO levels than those with a higher PXX content (red and green), as also shown in Fig. S23.

This trend can be effectively interpreted using a simplified linear combination of molecular orbitals (LCMO) approach, described in Section 21 of the SI. Within this framework, the HOMO and LUMO energies of a trimer are estimated from those of its constituent monomers and dimers, with variational corrections that scale inversely with the energy spacing between their corresponding orbitals. As a result, trimers with two A (or two D) units show stronger orbital mixing and larger energy shifts—downward for the LUMO and upward for the HOMO—reflecting enhanced acceptor (or donor) character. This simple method (Fig. S23c) qualitatively reproduces the experimental trends (Fig. S23a) and generally aligns well with more accurate DFT calculations (Fig. S23b), while offering a convenient tool for interpreting and predicting the electronic structures of mixed D-A oligomers.

The STS data further reveal that the two block-type trimers (PXX-PXX-AO and AO-AO-PXX) exhibit energy gaps 0.15–0.2 eV smaller than their alternating polymers, despite having the same monomer composition. This narrowing likely arises from enhanced orbital overlap between adjacent identical units, which promotes greater intrablock delocalisation and stabilises the frontier orbitals. While this effect is not captured by the simplified LCMO model (see Section 21), it is correctly reproduced by DFT (Table S2). However, some features are not fully accounted for by the DFT calculations either, most likely due to differences between the idealised vacuum environment assumed in the simulations and the real experimental conditions. In particular, the measured trimers are adsorbed on Au(111) and subject to local variations in their surroundings, including possible charge transfer effects and the presence of co-adsorbed Br atoms (as shown by the ‘pure’ donor or acceptor nanoribbons, further annealing to fully remove Br leads to longer ribbons, complicating the preservation of short mixed oligomers). To assess the magnitude of these effects, we compared dI/dV spectra from several PXX-AO dimers with different surroundings and terminations. The resulting variations in the positions of the frontier resonances—up to approximately 0.1 eV (Fig. S24)—are in line with those measured for different Br2PXX precursors (Fig. S25).

Constant height dI/dV images recorded with a CO-functionalised tip (Fig. 4d–h) reveal the spatial distribution of the frontier states. The highest occupied states are primarily localised on the donor PXX units, while the lowest unoccupied states show greater intensity over the acceptor AO units. A degree of orbital extension across the PXX-AO heterojunctions is expected due to the formation of a shared perylene moiety. This is indeed observed, with the HOMO states partially delocalised into adjacent AO units, for example, in the AO-AO-PXX trimer shown in Fig. 4h. Comparison with the corresponding simulated dI/dV images (displayed below the experimental dI/dV images in Fig. 4) shows excellent agreement in both shape and spatial distribution, considering the p-wave character of the CO tip. Variations due to tip effects are discussed in Sections 24 and 25 of the SI.

These results demonstrate the feasibility of synthesising D-A nanoribbons via on-surface (co)polymerisation by tuning the donor and acceptor character of the molecular building blocks used as (co)monomers, closely mirroring the design strategies employed in solution-phase conjugated polymer chemistry. Crucially, the atomic precision afforded by on-surface synthesis, combined with the advanced characterisation techniques of UHV surface science, enables structural and electronic analysis of the resulting nanoribbons with unmatched spatial resolution.

In this work, we have demonstrated the on-surface synthesis of ultra-narrow donor-acceptor nanoribbons via the surface-assisted polymerisation of PXX and AO molecular precursors and characterised their structure and electronic properties at submolecular resolution using BR-STM, nc-AFM and STS. Homopolymerisation of each precursor yields structurally well-defined nanoribbons with distinct electronic character: PXX-based ribbons exhibit a progressive increase in donor strength with length, while AO-based ribbons display pronounced electron-accepting behaviour. Copolymerisation of the two precursors produces mixed donor–acceptor nanoribbons, where the local electronic structure is strongly modulated by monomer sequence. The experimentally observed orbital distributions and energy levels are well described by gas-phase DFT calculations, while the overall trends in level shifts are captured by a simple linear combination of molecular orbitals model, offering an intuitive framework for predicting the electronic structure of D–A oligomers. This strategy parallels solution-phase D–A copolymer design yet harnesses the atomic precision and high-resolution characterisation uniquely enabled by on-surface synthesis in ultrahigh vacuum. Together, these results open new avenues for the atomically precise engineering of functional π-conjugated nanostructures with tuneable optoelectronic characteristics.

Methods

Materials

The synthesis of Br2PXX is described in our previous work61 and reported in Section 27 of the SI. VO3 was purchased commercially from Carbosynth.

STM and nc-AFM

STM, nc-AFM and STS experiments were performed on two systems: a Createc LT-STM at 7 K and a Scienta-Omicron LT-SPM at 4.3 K. The Au(111) crystal was prepared via standard cycles of sputtering and annealing. Br2PXX and VO3 were both deposited onto the room temperature surface via sublimation using an organic molecular beam deposition system (Dodecon Nanotechnology). PXX was sublimed at 165 °C, Br2PXX at 210 °C, and VO3 at 270 °C. Typical bias voltages (Vb) in the range of −2 V to +2 V and tunnelling current set points (IT) of 50–300 pA were used for standard STM imaging.

To obtain bond-resolving nc-AFM and STM images, CO was deposited onto the cold (7–10 K) Au(111) surface by leaking CO gas into the chamber at a pressure of 10−7 mbar for 20–30 s. It was then picked up from the surface either by scanning with typical tunnelling parameters (negative bias voltage) or by approaching the tip over CO clusters. nc-AFM images were recorded with a QPlus sensor (Q ≈ 120k) at a constant height with a typical oscillation amplitude of 50 pm. BR-STM images were recorded with a low bias voltage (5–40 mV) at a constant height with the tip in close proximity to the molecules.

STS

dI/dV measurements were acquired using an internal lock-in with oscillation amplitudes of 5–40 mV and frequencies at 780–1080 Hz. Constant height dI/dV images were recorded using similar oscillation parameters. For some of the nanoribbons, the constant height dI/dV spectra for positive and negative bias voltages were recorded separately. This is due to spatial variations in the intensity of the dI/dV signal coming from the different occupied and unoccupied states. In some cases, dI/dV images were FFT-filtered to remove high-frequency electronic noise.

XPS

XPS experiments were performed on an Omicron Multiprobe system with a monochromated Al Kα source and a photon energy of 1486.7 eV in an analysis chamber with a base pressure lower than 1 × 10−10 mbar. The analyser work function and binding energy scale were calibrated according to photoemission peaks and the Fermi level position of a polycrystalline silver foil with a known work function. A pass energy of 10 eV (resolution approximately 0.47 eV) was used for recording core-level spectra from a 1.1 mm diameter area on the surface. Samples were prepared in an adjoining preparation chamber, with molecules deposited via sublimation using an organic molecular beam deposition system (Dodecon Nanotechnology), before being transferred to the analysis chamber for XPS measurements. Annealing was performed in the main analysis chamber using a filament built into the manipulator and temperature monitoring via a chromel/alumel thermocouple in close proximity to the sample.

DFT

All DFT calculations were carried out using the mixed Gaussian and Plane-Waves (GPW) method implemented in the CP2K package68. Given the sensitivity of intermolecular interactions and electronic level alignment to the choice of exchange-correlation (XC) functional, we explored a range of functionals with differing levels of sophistication. Specifically, we employed the fully self-consistent non-local vdW-DF XC functional69 as well as two different hybrid XC functionals: B3LYP70 and HSE0671.

All calculations were performed in the gas phase, with no periodic boundary conditions, starting from the experimental molecular geometries. The Kohn–Sham orbitals were expanded in a triple-zeta valence plus double polarization (TZV2P) Gaussian basis set. Core electrons were treated using Goedecker–Teter–Hutter (GTH) pseudopotentials72, with four, one, six, five, and three valence electrons for carbon, hydrogen, oxygen, nitrogen, and boron atoms, respectively. The plane-wave cutoff for the finest level of the multigrid used to solve the Poisson equation68 was set to 400 Ry. These settings were sufficient to converge the total energy to 4 meV/atom.

Supplementary information

Acknowledgements

We thank Richard Walton for kindly providing the VO3 precursor used in this study. J. Lawrence acknowledges the support from the Agency for Science, Technology & Research (A*STAR), Singapore, through its MTC YIRG Grant (Project ID: M23M7c0121). G.C. acknowledges funding from the EU through the European Research Council (ERC) Grant “VISUAL-MS” (Project ID: 308115). G.C.S. gratefully acknowledges the use of SULIS, which was funded by the EPSRC (EP/T022108/1), via the HPC Midlands+ Consortium. G.C.S. and F.B. would also like to acknowledge the high-performance computing facilities provided by the Scientific Computing Research Technology Platform (SCRTP) at the University of Warwick. D.B. gratefully acknowledges the EU through the MSCA-RISE (project: VIT, No. 101008237) and the University of Vienna for generous financial support. J. Lu acknowledges the support of the Ministry of Education (MOE) Singapore (MOE-T2EP10124-0004).

Author contributions

J. Lawrence performed the STM, BR-STM, STS, and nc-AFM measurements, carried out the XPS measurements, analysed the data, proposed the use of complementary donor and acceptor comonomers for the on-surface synthetic strategy, and co-wrote the first version of the manuscript. L.Đ. synthesised Br₂PXX. F.B. contributed to perform, analyse and interpret the DFT calculations. H.P. performed an initial set of DFT calculations. M.W. contributed to the XPS measurements and data analysis. J. Lu supervised the nc-AFM measurements. G.C.S. performed, analysed and interpreted the DFT calculations, generated the simulated dI/dV images, and supervised the computational aspects of the work. D.B. proposed the on-surface synthesis of PXX nanoribbons and supervised the synthetic work. G.C. supervised the STM, BR-STM, and STS measurements, developed the simplified LCMO model, co-wrote the first version of the manuscript, and supervised and coordinated the entire project. All authors contributed to the editing and revision of the manuscript.

Peer review

Peer review information

Nature Communications thanks Oscar Custance, Sabine Wenzel and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The data supporting this paper have been deposited in a publicly available Zenodo repository with DOI:10.5281/zenodo.18851738. The data in question includes: (a) the STM, BR-STM, STS, and nc-AFM measurements presented in this paper and; (b) the input files, geometries, and post-processing scripts utilised to perform and analyse the DFT simulations (and simulated STM images) featured in this work. All data are available from the corresponding author upon request.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

James Lawrence, Email: jmlaw91@nus.edu.sg.

Giovanni Costantini, Email: g.costantini@bham.ac.uk.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-026-71660-0.

References

  • 1.Corso, M., Carbonell-Sanromà, E. & de Oteyza, D. G. Bottom-up fabrication of atomically precise graphene nanoribbons. In On-SurfaceSynthesisII (eds de Oteyza, D. G. & Rogero, C.) 113–152 (Springer, 2018).
  • 2.Talirz, L., Ruffieux, P. & Fasel, R. On-surface synthesis of atomically precise graphene nanoribbons. Adv. Mater.28, 6222–6231 (2016). [DOI] [PubMed] [Google Scholar]
  • 3.Cai, J. et al. Atomically precise bottom-up fabrication of graphene nanoribbons. Nature466, 470–473 (2010). [DOI] [PubMed] [Google Scholar]
  • 4.Marmolejo-Tejada, J. M. & Velasco-Medina, J. Review on graphene nanoribbon devices for logic applications. Microelectron. J.48, 18–38 (2016). [Google Scholar]
  • 5.Rajaji, U. et al. Graphene nanoribbons in electrochemical sensors and biosensors: a review. Int. J. Electrochem. Sci.13, 6643–6654 (2018). [Google Scholar]
  • 6.Shekhirev, M. et al. Highly selective gas sensors based on graphene nanoribbons grown by chemical vapour deposition. ACS Appl. Mater. Interfaces12, 7392–7402 (2020). [DOI] [PubMed] [Google Scholar]
  • 7.Han, W., Kawakami, R. K., Gmitra, M. & Fabian, J. Graphene spintronics. Nat. Nanotechnol.9, 794–807 (2014). [DOI] [PubMed] [Google Scholar]
  • 8.Houtsma, R. S. K., de la Rie, J. & Stöhr, M. Atomically precise graphene nanoribbons: interplay of structural and electronic properties. Chem. Soc. Rev.50, 6541–6568 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Kawai, S. et al. Atomically controlled substitutional boron-doping of graphene nanoribbons. Nat. Commun.6, 8098 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Cloke, R. R. et al. Site-specific substitutional boron doping of semiconducting armchair graphene nanoribbons. J. Am. Chem. Soc.137, 8872–8875 (2015). [DOI] [PubMed] [Google Scholar]
  • 11.Carbonell-Sanromà, E. et al. Electronic properties of substitutionally boron-doped graphene nanoribbons on a Au(111) surface. J. Phys. Chem. C122, 16092–16099 (2018). [Google Scholar]
  • 12.Kawai, S. et al. Multiple heteroatom substitution to graphene nanoribbon. Sci. Adv.4, 1–8 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Cai, J. et al. Graphene nanoribbon heterojunctions. Nat. Nanotechnol.9, 896–900 (2014). [DOI] [PubMed] [Google Scholar]
  • 14.Carbonell-Sanromà, E. et al. Doping of graphene nanoribbons via functional group edge modification. ACS Nano11, 7355–7361 (2017). [DOI] [PubMed] [Google Scholar]
  • 15.Bronner, C. et al. Aligning the band gap of graphene nanoribbons by monomer doping. Angew. Chem. Int. Ed.52, 4422–4425 (2013). [DOI] [PubMed] [Google Scholar]
  • 16.Marangoni, T., Haberer, D., Rizzo, D. J., Cloke, R. R. & Fischer, F. R. Heterostructures through divergent edge reconstruction in nitrogen-doped segmented graphene nanoribbons. Chem. Eur. J.22, 13037–13040 (2016). [DOI] [PubMed] [Google Scholar]
  • 17.Li, J. et al. Band depopulation of graphene nanoribbons induced by chemical gating with amino groups. ACS Nano14, 1895–1901 (2020). [DOI] [PubMed] [Google Scholar]
  • 18.Zhang, Y. et al. On-surface synthesis of a nitrogen-doped graphene nanoribbon with multiple substitutional sites. Angew. Chem. Int. Ed.61, e202204736 (2022). [DOI] [PubMed] [Google Scholar]
  • 19.Nguyen, G. D. et al. Bottom-up synthesis of N = 13 sulfur-doped graphene nanoribbons. J. Phys. Chem. C120, 2684–2687 (2016). [Google Scholar]
  • 20.Nguyen, G. D. et al. Atomically precise graphene nanoribbon heterojunctions from a single molecular precursor. Nat. Nanotechnol.12, 1077–1082 (2017). [DOI] [PubMed] [Google Scholar]
  • 21.Liu, M. et al. Graphene-like nanoribbons periodically embedded with four- and eight-membered rings. Nat. Commun.8, 14924 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Zhao, Y. et al. Engineering graphene nanoribbons via periodically embedding oxygen atoms. Angew. Chem. Int. Ed.64, e202500490 (2025). [DOI] [PubMed] [Google Scholar]
  • 23.Jacobse, P. H. et al. Electronic components embedded in a single graphene nanoribbon. Nat. Commun.8, 119 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Chen, Y.-C. et al. Molecular bandgap engineering of bottom-up synthesised graphene nanoribbon heterojunctions. Nat. Nanotechnol.10, 156–160 (2015). [DOI] [PubMed] [Google Scholar]
  • 25.Bronner, C. et al. Hierarchical on-surface synthesis of graphene nanoribbon heterojunctions. ACS Nano12, 2193–2200 (2018). [DOI] [PubMed] [Google Scholar]
  • 26.Blankenburg, S. et al. Intraribbon heterojunction formation in ultranarrow graphene nanoribbons. ACS Nano6, 2020–2025 (2012). [DOI] [PubMed] [Google Scholar]
  • 27.Rizzo, D. J. et al. Topological band engineering of graphene nanoribbons. Nature560, 204–208 (2018). [DOI] [PubMed] [Google Scholar]
  • 28.Gröning, O. et al. Engineering of robust topological quantum phases in graphene nanoribbons. Nature560, 209–213 (2018). [DOI] [PubMed] [Google Scholar]
  • 29.Cao, T., Zhao, F. & Louie, S. G. Topological phases in graphene nanoribbons: junction states, spin centres, and quantum spin chains. Phys. Rev. Lett.119, 076401 (2017). [DOI] [PubMed] [Google Scholar]
  • 30.Sun, K. et al. Synthesis of regioisomeric graphene nanoribbon junctions via hetero precursors. J. Phys. Chem. C123, 17632–17638 (2019). [Google Scholar]
  • 31.Rizzo, D. J. et al. Length-dependent evolution of type II heterojunctions in bottom-up-synthesised graphene nanoribbons. Nano Lett.19, 3221–3228 (2019). [DOI] [PubMed] [Google Scholar]
  • 32.Brus, V. V. et al. Solution-processed semitransparent organic photovoltaics: from molecular design to device performance. Adv. Mater.31, 1900904 (2019). [DOI] [PubMed] [Google Scholar]
  • 33.Kim, M. et al. Donor–acceptor-conjugated polymer for high-performance organic field-effect transistors: a progress report. Adv. Funct. Mater.30, 1904545 (2020). [Google Scholar]
  • 34.Dou, L., Liu, Y., Hong, Z., Li, G. & Yang, Y. Low-bandgap near-IR conjugated polymers/molecules for organic electronics. Chem. Rev.115, 12633–12665 (2015). [DOI] [PubMed] [Google Scholar]
  • 35.van Mullekom, H. A. M., Vekemans, J. A. J. M., Havinga, E. E. & Meijer, E. W. Developments in the chemistry and band gap engineering of donor–acceptor substituted conjugated polymers. Mater. Sci. Eng. R Rep32, 1–40 (2001). [Google Scholar]
  • 36.Yuen, J. D. & Wudl, F. Strong acceptors in donor–acceptor polymers for high performance thin film transistors. Energy Environ. Sci.6, 392–406 (2013). [Google Scholar]
  • 37.Steckler, T. T. et al. Very low band gap thiadiazoloquinoxaline donor–acceptor polymers as multi-tool conjugated polymers. J. Am. Chem. Soc.136, 1190–1193 (2014). [DOI] [PubMed] [Google Scholar]
  • 38.Beaujuge, P. M., Amb, C. M. & Reynolds, J. R. Spectral engineering in π-conjugated polymers with intramolecular donor–acceptor interactions. Acc. Chem. Res.43, 1396–1407 (2010). [DOI] [PubMed] [Google Scholar]
  • 39.Wu, J.-S., Cheng, S.-W., Cheng, Y.-J. & Hsu, C.-S. Donor–acceptor conjugated polymers based on multifused ladder-type arenes for organic solar cells. Chem. Soc. Rev.44, 1113–1154 (2015). [DOI] [PubMed] [Google Scholar]
  • 40.Lee, J., Kalin, A. J., Yuan, T., Al-Hashimi, M. & Fang, L. Fully conjugated ladder polymers. Chem. Sci.8, 2503–2521 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Wang, L., Duan, G., Ji, Y. & Zhang, H. Electronic and charge transport properties of peri-xanthenoxanthene: the effects of heteroatoms and phenyl substitutions. J. Phys. Chem. C116, 22679–22686 (2012). [Google Scholar]
  • 42.Zanetti, D., De Luca, C. & Bonifazi, D. A bright comeback for peri-xanthenoxanthene. Nat. Chem.17, 1138 (2025). [DOI] [PubMed] [Google Scholar]
  • 43.Kobayashi, N., Sasaki, M. & Nomoto, K. Stable peri-xanthenoxanthene thin-film transistors with efficient carrier injection. Chem. Mater.21, 552–556 (2009). [Google Scholar]
  • 44.Kobayashi, N., Sasaki, M. & Ohe, T. Semiconductor device, method of manufacturing the same, and method of forming multilayer semiconductor thin film. US Patent # 8,546,796 B2 (2013).
  • 45.Lv, N. et al. Synthesis, properties, and structures of functionalised peri-xanthenoxanthene. Org. Lett.15, 2382–2385 (2013). [DOI] [PubMed] [Google Scholar]
  • 46.Miletić, T. et al. Tailoring colours by O annulation of polycyclic aromatic hydrocarbons. Chem. Eur. J.23, 2363–2378 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Stassen, D., Demitri, N. & Bonifazi, D. Extended O-doped polycyclic aromatic hydrocarbons. Angew. Chem. Int. Ed.55, 5947–5951 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Sciutto, A. et al. Customising photoredox properties of PXX-based dyes through energy level rigid shifts of frontier molecular orbitals. Chem. Eur. J.24, 4382–4389 (2018). [DOI] [PubMed] [Google Scholar]
  • 49.Berezin, A., Biot, N., Battisti, T. & Bonifazi, D. Oxygen-doped zig-zag molecular ribbons. Angew. Chem. Int. Ed.57, 8942–8946 (2018). [DOI] [PubMed] [Google Scholar]
  • 50.Di Santo, G. et al. Orbital mapping of semiconducting perylenes on Cu(111). J. Phys. Chem. C125, 24477–24486 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Della Pia, A. et al. Two-dimensional ketone-driven metal–organic coordination on Cu(111). Chem. Eur. J.22, 8105–8112 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Kawano, S. I., Baumgarten, M., Chercka, D., Enkelmann, V. & Müllen, K. Electron donors and acceptors based on 2,7-functionalised pyrene-4,5,9,10-tetraone. Chem. Commun.49, 5058–5060 (2013). [DOI] [PubMed] [Google Scholar]
  • 53.Lin, Y. et al. An electron acceptor challenging fullerenes for efficient polymer solar cells. Adv. Mater.27, 1170–1174 (2015). [DOI] [PubMed] [Google Scholar]
  • 54.Zhan, C., Zhang, X. & Yao, J. New advances in non-fullerene acceptor based organic solar cells. RSC Adv.5, 93002–93026 (2015). [Google Scholar]
  • 55.Maki, T. & Hashimoto, H. Vat dyes of acenaphthene series. V. constitution of dibromoanthanthrone (indanthrene brilliant orange RK). Bull. Chem. Soc. Jpn26, 348–351 (1953). [Google Scholar]
  • 56.Schmidt, M. U. et al. Experimental and predicted crystal structures of pigment red 168 and other dihalogenated anthanthrones. Acta Crystallogr. B66, 515–526 (2010). [DOI] [PubMed] [Google Scholar]
  • 57.Irimia-Vladu, M. et al. Biocompatible and biodegradable materials for organic field-effect transistors. Adv. Funct. Mater.20, 4069–4076 (2010). [Google Scholar]
  • 58.Irimia-Vladu, M., Sariciftci, N. S. & Bauer, S. Exotic materials for bio-organic electronics. J. Mater. Chem.21, 1350–1361 (2011). [DOI] [PubMed] [Google Scholar]
  • 59.Irimia-Vladu, M. et al. Environmentally sustainable organic field effect transistors. Org. Electron.11, 1974–1990 (2010). [Google Scholar]
  • 60.Pham, H. D. et al. One step facile synthesis of a novel anthanthrone dye-based, dopant-free hole transporting material for efficient and stable perovskite solar cells. J. Mater. Chem. C6, 3699–3708 (2018). [Google Scholar]
  • 61.Lawrence, J. et al. Combining high-resolution scanning tunnelling microscopy and first-principles simulations to identify halogen bonding. Nat. Commun.11, 2103 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Berdonces-Layunta, A. et al. Order from a mess: the growth of 5-armchair graphene nanoribbons. ACS Nano15, 16552–16561 (2021). [DOI] [PubMed] [Google Scholar]
  • 63.Gross, L. et al. Atomic force microscopy for molecular structure elucidation. Angew. Chem. Int. Ed.57, 3888–3908 (2018). [DOI] [PubMed] [Google Scholar]
  • 64.Gross, L. et al. The chemical structure of a molecule resolved by atomic force microscopy. Science325, 1110–1114 (2009). [DOI] [PubMed] [Google Scholar]
  • 65.Lawrence, J. et al. Circumventing the stability problems of graphene nanoribbon zigzag edges. Nat. Chem.14, 1451–1458 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Kimouche, A. et al. Ultra-narrow metallic armchair graphene nanoribbons. Nat. Commun.6, 10177 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Berdonces-Layunta, A. et al. Chemical stability of (3,1)-chiral graphene nanoribbons. ACS Nano15, 5610–5617 (2021). [DOI] [PubMed] [Google Scholar]
  • 68.VandeVondele, J. et al. Quickstep: fast and accurate density functional calculations using a mixed Gaussian and plane waves approach. Comput. Phys. Commun.167, 103–128 (2005). [Google Scholar]
  • 69.Dion, M., Rydberg, H., Schröder, E., Langreth, D. C. & Lundqvist, B. I. Van der Waals density functional for general geometries. Phys. Rev. Lett.92, 246401 (2004). [DOI] [PubMed] [Google Scholar]
  • 70.Becke, A. D. Density-functional thermochemistry. III. the role of exact exchange. J. Chem. Phys.98, 5648–5652 (1993). [Google Scholar]
  • 71.Krukau, A. V., Vydrov, O. A., Izmaylov, A. F. & Scuseria, G. E. Influence of the exchange screening parameter on the performance of screened hybrid functionals. J. Chem. Phys125, 224106 (2006). [DOI] [PubMed] [Google Scholar]
  • 72.Goedecker, S., Teter, M. & Hutter, J. Separable dual-space Gaussian pseudopotentials. Phys. Rev. B54, 1703–1710 (1996). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data Availability Statement

The data supporting this paper have been deposited in a publicly available Zenodo repository with DOI:10.5281/zenodo.18851738. The data in question includes: (a) the STM, BR-STM, STS, and nc-AFM measurements presented in this paper and; (b) the input files, geometries, and post-processing scripts utilised to perform and analyse the DFT simulations (and simulated STM images) featured in this work. All data are available from the corresponding author upon request.


Articles from Nature Communications are provided here courtesy of Nature Publishing Group

RESOURCES