- The paper demonstrates that polymer-free wet transfer preserves the atomic structure of 9-AGNRs on epitaxial graphene, retaining their precise armchair-edge topology.
- The methodology employs low-temperature STM/STS to reveal transfer-induced fragmentation, showing reduced ribbon lengths and a widened band gap (1.7 eV) compared to Au substrates.
- The results highlight challenges with chemically sensitive GNRs and emphasize the need for alternative transfer protocols to maintain quantum properties in advanced devices.
Atomic-Scale Imaging of Graphene Nanoribbons on Graphene After Polymer-Free Substrate Transfer
Introduction and Motivation
The atomic-scale engineering of graphene nanoribbons (GNRs) via on-surface synthesis offers precise control over width, edge topology, and consequently the electronic band structure. Realizing scalable, functional integration of such GNRs into device architectures necessitates their transfer from noble metal growth substrates—typically Au(111)—to technologically relevant supports, including epitaxial graphene (EG), quasi-freestanding EG (QFEG), and other two-dimensional (2D) materials. Conventional transfer methods, especially those involving supporting polymers such as PMMA, introduce contamination and residuals which have deleterious effects on GNR integrity and device performance. Even optimized, polymer-free wet transfer protocols, although superior regarding cleanliness, frequently induce mechanical and chemical modifications, especially critical for GNRs where edge structure encodes quantum properties.
Atomic-resolution post-transfer characterization is required to address ensemble-averaging limitations inherent in Raman spectroscopy and to elucidate substrate-induced modifications. This work applies low-temperature STM/STS to systematically investigate the structural and electronic properties of 9-atom-wide armchair GNRs (9-AGNRs) and chemically sensitive GNR variants after polymer-free transfer onto EG and QFEG, establishing transfer-induced structure-property relationships essential for future GNR device engineering (2504.03815).
Methods and Experimental Design
9-AGNRs and topologically non-trivial variants (e.g., porphyrin-extended zigzag ribbons) were synthesized via UHV surface-catalyzed polymerization and cyclodehydrogenation of tailored monomers on gold single crystals or films. Two principal transfer protocols were deployed: (1) a polymer-free wet transfer leveraging the Au(111) film as sacrificial support, followed by delamination, substrate fishing, and KI-I gold etching; (2) an electrochemical (bubble) transfer employing a PMMA support layer for GNRs requiring preserved global alignment. Post-transfer decontamination required high-temperature (750 ℃) UHV annealing to render the surface amenable to STM analysis.
Structural and electronic properties were interrogated by room-temperature overview STM and high-resolution low-temperature STM/STS, with additional ensemble-averaged characterization via Raman spectroscopy to monitor post-transfer chemical modifications and vibrational mode retention. Control measurements on as-grown GNRs established the baseline for transfer-induced effects.
Structural Integrity of 9-AGNRs after Transfer
STM analysis revealed that polymer-free wet transfer preserves the atomic structure and characteristic armchair-edge topology of 9-AGNRs with high fidelity. Pre- and post-transfer images confirmed the persistence of atomically defined edges and, where present, ‘bite defects’ originating in the synthesis. A statistically significant reduction in average ribbon length (from 26 nm to 15 nm) was quantified, attributable to mechanical/chemical fragmentation during transfer and thermally induced breakage during UHV annealing—parameters that do not induce discernible edge oxidation or massive disorder.
Raman spectroscopy corroborated these findings, as the radial breathing-like mode (RBLM) at 312 cm⁻¹—indicative of precise ribbon width—remained prominent, though D and CH mode broadening was indicative of some edge disorder and occasional inter-ribbon fusion events, as detected in STM. Notably, transfer to EG (compared to metallic Au) does not catalyze detrimental chemical transformations, owing to its inertness.
Electronic Structure and Substrate Interactions
Low-temperature STS of 9-AGNRs on EG and QFEG revealed well-defined gaps, with the valence and conduction onsets located near –1 V and 0.7 V, respectively. Frontier orbital maps obtained by STM were in quantitative agreement with tight-binding simulations. Compared to Au(111), an increased GNR band gap of 1.7 eV on EG and QFEG (vs. 1.4 eV on Au) was measured—indicative of reduced electronic screening by graphene relative to metals.
Differences in work function between EG (4.2 eV), QFEG (4.7 eV), and Au (5.3 eV) translate into distinct Fermi level alignments, directly impacting carrier injection barriers. On EG, a nearly symmetric alignment of the valence and conduction bands with respect to the substrate Fermi level was observed, theoretically supporting efficient, ambipolar injection in device implementations. These experimental determinations of level alignment are crucial for rational contact and channel design in GNR-based FETs, quantum dot and spintronic architectures.
Stability of Chemically Sensitive GNRs
The transfer and annealing protocols that were successful for 9-AGNRs proved only partially compatible, or insufficient, for chemically sensitive GNR subtypes. For porphyrin-extended zigzag GNRs, STM revealed that while some porphyrin cores survived, substantial fragmentation and disorder were evident; Raman spectroscopy failed to provide an optical fingerprint due to either poor resonance or structural disintegration. For 7-AGNR-S(1,3) GNRs hosting topological quantum states, atomic-scale imaging post-transfer revealed complete structural loss, making them inappropriate for current wet-transfer/anneal protocols.
These observations unambiguously demonstrate that while armchair GNRs tolerate existing transfer processes—preserving both structure and key electronic features—open-shell or functionalized GNRs do not. The findings highlight an urgent need for alternative strategies, such as ultra-clean dry transfer, inert encapsulation, or edge-function-group stabilization during processing, to translate quantum-engineered GNRs into application-relevant contexts.
Implications and Future Directions
This work establishes, for the first time at atomic resolution, the transfer fidelity of atomically precise GNRs (specifically 9-AGNRs) to technologically relevant, 2D substrates. The quantitative retention of structure and the associated observation of a widened band gap with favorable band alignment on EG/QFEG directly informs the design of GNR-based electronic and quantum devices, where contact engineering and transport symmetry are paramount. The robust methodology presented serves as a template for scrutinizing GNR-substrate interactions beyond conventional metal surfaces.
The inability of current techniques to preserve chemically sensitive nanoribbon structures underlines a critical research objective: developing transfer and integration protocols that are compatible with exotic edge states and functional side-chain architectures. Success will open pathways to GNR-based quantum devices, low-dimensional spintronics, and optoelectronics exploiting topologically nontrivial states.
Conclusion
Atomic-scale STM/STS provides definitive evidence that polymer-free substrate transfer can maintain the structural integrity and enable intrinsic electronic characterization of 9-AGNRs on EG and QFEG. A significant widening of the electronic band gap and symmetric band-to-Fermi-level alignment—absent on metallic substrates—were demonstrated, directly benefitting device design. Chemically sensitive GNRs, especially those with open-shell character, do not survive current protocols, mandating the development of transfer strategies that minimize chemical and thermal stress. This work systematically bridges chemical synthesis, process integration, and atomic-resolution characterization, underpinning reliable GNR device development (2504.03815).
