Fabrication and Characterization of Van der Waals Heterostructures

This study details the fabrication of van der Waals (vdW) heterostructures and their characterization using scattering-type scanning near-field optical microscopy (s-SNOM) and magnetic force microscopy (MFM).

These heterostructures were formed by positioning vdW materials, ranging from 10-120 nm thick, onto flat surfaces of κ-ET single crystals. A total of six devices were fabricated, employing various crystal surface preparations including as-grown, cleaved, or exfoliated.

Device Fabrication Process

κ-ET Crystal Surface Preparation

κ-ET crystal surfaces were prepared through several methods: as-grown, by cleaving bulk crystals, or by exfoliating single crystals. For bulk samples, κ-ET crystals were mounted on silicon substrates using epoxy or silver paint.

Large, atomically flat surfaces were subsequently identified using optical and atomic force microscopy. While as-grown surfaces were consistently flat, cleaved surfaces generally exhibited greater roughness. Significantly, all fabricated devices maintained their integrity over periods of weeks to months, enduring multiple thermal cycles without any observed degradation.

Microcrystal Exfoliation and Stacking

hBN and RuCl3 microcrystals were exfoliated using a standard scotch-tape technique onto a polydimethylsiloxane (PDMS) stamp, then identified via optical and atomic force microscopy. Stacking of these materials was performed using a commercial transfer stage under precise temperature control.

hBN microcrystals were released at approximately 60 °C, while RuCl3 microcrystals in devices 1 and 5 were released at 110 °C and 130 °C, respectively. Isotopically pure hBN (10BN) was specifically utilized to ensure a narrow linewidth for the TO phonon, a critical factor for making HPhPs visible in s-SNOM measurements.

hBN/BSCCO Sample Preparation

hBN/BSCCO samples were prepared within an inert gas glovebox environment. BSCCO was exfoliated on SiO2 substrates, and hBN microcrystals were subsequently transferred onto the BSCCO at room temperature.

s-SNOM Measurement Methodology

Nano-infrared scattering experiments were conducted using a home-built cryogenic s-SNOM system, housed within an ultrahigh-vacuum chamber.

This advanced s-SNOM system is built upon a tapping-mode atomic force microscope (AFM), operating with a tapping frequency of approximately 285 kHz and an amplitude of 70 nm. The technique works by scattering focused laser light from a sharp AFM tip, thereby enabling nanometer-scale resolution for probing near-field interactions.

A tunable quantum cascade laser served as the light source, which was precisely focused onto the metallized AFM tip using a parabolic mirror. Backscattered light was then detected by a mercury cadmium telluride detector. This signal was demodulated using a pseudoheterodyne scheme at the third or fourth harmonic of the tip-tapping frequency to effectively eliminate far-field background interference.

Magnetic Force Microscopy (MFM) Procedures

MFM measurements were performed utilizing an Attocube cantilever-based cryogenic atomic force microscope, operating within a helium exchange gas environment. For these measurements, Nanosensors PPP-MFMR probes with hard magnetic coatings and force constants around 2.8 N m−1 were employed.

The resonant frequency shift (∆f) of the cantilever was recorded during measurements, which was then converted to ∂zFz using a specific formula. The tip oscillation amplitude typically ranged from 50–70 nm. It is important to note that hBN, being non-magnetic, does not interfere with the detection of the Meissner effect, and RuCl3 exhibits low susceptibility.

To minimize noise and ensure data quality, ∆f(z) curves were averaged over multiple measurements with high z-resolution.

Additionally, the tip was electrically biased to compensate for any local contact potential difference. Averaged curves underwent smoothing, and linear backgrounds were meticulously subtracted. For comparative analyses, measurements for hBN/κ-ET, bare κ-ET, and RuCl3/κ-ET were thickness-compensated to align the κ-ET surface at zero.

Constant-height MFM images were collected with the tip at a fixed height above the κ-ET surface, presenting raw data after conversion from ∆f to ∂zFz. Data points for superfluid density analysis were derived either by fitting ∂zFz(z) curves or by converting constant-height MFM images into maps of local λin (decay length for parallel currents), which was then further converted to superfluid density.

HPhP Dispersion Simulation

No details were provided for HPhP dispersion simulation methodology in the provided text.