GoKawiilChemicals and materials
Natural Kish graphite crystals (Grade 300, 99.99% purity) were procured from Graphene Supermarket. Hexagonal boron nitride crystals were provided by T. Taniguchi and K. Watanabe, and were used as received. Large, flat crystals of RuCl 3 were grown by chemical vapour transport following the procedure detailed in a previous study41. Briefly, commercial RuCl 3 powder (Alfa Aesar, anhydrous, Ru ≥ 47.7%) was loaded into a quartz ampoule in an argon glovebox, sealed under dynamic vacuum, and heated in a two-zone furnace with a temperature gradient and ramp rates of 1 K per min. The resulting crystals were collected from the cold end and stored in an argon-filled glovebox.
Si/SiO 2 wafers (0.5-mm-thick, 285 nm SiO 2 ) and polydimethylsiloxane stamps (PDMS) were obtained from NOVA Electronic Materials and MTI Corporation, respectively. Sn/In alloy (Custom Thermoelectric), poly(bisphenol-A carbonate), hexaammineruthenium(III) chloride (98%) and potassium chloride (>99%) were purchased from Sigma-Aldrich. Sulfuric acid (ACS grade, >95−98%, Thermo Fisher Scientific) was used as received. All aqueous electrolyte solutions were prepared using type I water (EMD Millipore, 18.2 MΩ cm). Solid KCl was added as a supporting electrolyte in Ru(NH 3 ) 6 3+ solution to a final concentration of 100 mM.
Sample fabrication
Graphene and hBN flakes were mechanically exfoliated onto SiO 2 (285 nm)/Si substrates from bulk crystals using Scotch tape42. Exfoliated flakes on SiO 2 /Si chips were identified by optical microscopy (Laxco LMC-5000). MLG flakes were distinguished by their approximately 7% optical contrast in the green channel14,43 and further verified by Raman spectroscopy (HORIBA LabRAM Evo). Extended Data Fig. 1 shows a representative optical contrast of around 7% in the green channel for MLG and about 14% for bilayer graphene. The thickness of hBN flakes was determined by atomic force microscopy (Park Systems NX10) (Extended Data Fig. 1c,d).
α-RuCl 3 crystals were exfoliated in an argon-filled glovebox onto SiO 2 (90 nm)/Si substrates to prevent degradation. Precise thickness control was not required, as even a single monolayer of α-RuCl 3 is sufficient to induce substantial hole doping in graphene30,41. Instead, emphasis was placed on selecting flakes smaller than the hBN to ensure complete encapsulation, and flatness was prioritized to minimize strain during stacking. Suitable flakes were identified with an optical microscope (Nikon) within the glovebox.
We selected the multilayer system comprising graphene, hBN, RuCl 3 and WSe 2 due to their complementary characteristics. Graphene offers a tunable and well-defined electronic platform, whereas hBN serves as an inert spacer that allows precise control of doping. The RuCl 3 and WSe 2 layers function as stable charge-transfer dopants, modulating graphene’s electronic properties without affecting its structural integrity. Together, these materials enable systematic tuning of interfacial doping while preserving the overall structural quality of the heterostructure. MLG–hBN–RuCl 3 heterostructures were assembled by a dry-transfer technique on a temperature controlled stage (Instec), equipped with an optical microscope (Mitutoyo FS70) and micromanipulator (MP-285, Sutter Instrument) in an argon glovebox. A poly(bisphenol-A carbonate) film on a PDMS stamp was used to pick up a RuCl 3 flake within 30 min of exfoliation to minimize moisture exposure, which could compromise its doping efficacy41. The picked RuCl 3 flake was then capped with an hBN flake (3–180 nm thick), followed by MLG, partially overlapping the RuCl 3 to leave a segment of graphene without RuCl 3 . A thick graphite flake (10–100 nm) was finally transferred to partially overlap the graphene, providing electrical contact with the heterostructure. The poly(bisphenol-A carbonate) film was delaminated from the PDMS stamp and placed onto a clean SiO 2 /Si chip. Electrical contacts with graphite were subsequently established using Sn/In microsoldering14.
SECCM measurements
Single-channel SECCM nanopipettes were fabricated from quartz capillaries (0.7 mm inner diameter, 1 mm outer diameter; Sutter Instrument) using a laser puller (P-2000, Sutter Instrument) with the following parameters: heat = 700, filament = 4, velocity = 20, delay = 127, and pull = 140. This procedure yielded pipettes with orifice diameters of 600–800 nm, as confirmed by bright-field transmission electron microscopy (TEM; Extended Data Fig. 3). Each nanopipette was filled with an electrolyte solution containing the redox species of interest and equipped with a silver wire coated with AgCl, serving as a quasi-reference or counter electrode.
Scanning electrochemical cell microscopy experiments were performed using a Park NX10 SICM module. The nanopipette was positioned above the sample using an optical microscope and approached the surface at 100 nm s–1 until meniscus contact was detected by a current increase above 3 pA. During approach, a −0.5 V bias was applied to facilitate diffusion-limited reactions. Cyclic voltammograms were recorded at multiple locations by sweeping the potential at 100 mV s−1 between –0.6 V and 0 V, with the half-wave potential, E 1/2 , defined as the potential at which i = i ∞ /2, where i ∞ represents the diffusion-limited current plateau. [Ru(NH 3 ) 6 3+/2+] was chosen as the redox couple because it has well-characterized, reversible, outer-sphere ET with no detectable adsorption on graphite electrodes, as confirmed by in situ Raman spectroscopy14,21. This ensures that the measured kinetics are sensitive to the electronic properties of the electrode while avoiding complications from surface-specific reactions.
... continue reading