Proximitized Topological Insulator Charge Island Fabricated via In Situ Multi-Angle Stencil Lithography

1. Problem Statement

The realization of topological superconductivity is a central goal in condensed matter physics, primarily due to its potential for fault-tolerant quantum computing via Majorana zero modes (MZMs). While semiconductor nanowires have been extensively studied, Topological Insulators (TIs) offer a promising alternative platform where topological phases can be tuned via magnetic flux or chemical potential.

However, creating hybrid superconductor-TI devices faces two critical material challenges:

1. Surface Degradation: TI surfaces (e.g., $(Bi,Sb)_2Te_3$) oxidize rapidly under ambient conditions, degrading interface transparency and preventing the formation of high-quality superconductor-TI interfaces.
2. Lack of Proximitized Charge Islands: While tunneling spectroscopy exists, proximitized TI charge islands (essential for studying charging effects and parity-dependent transport) have not been successfully demonstrated. Previous attempts were limited by the inability to create clean interfaces and well-defined tunnel barriers without post-growth processing, which introduces contamination.

2. Methodology: In Situ Multi-Angle Stencil Lithography

The authors developed a fully in situ fabrication technique combining Selective Area Growth (SAG) with multi-angle shadow evaporation using a double-mask stack. This approach eliminates post-growth processing, ensuring pristine interfaces.

• Mask Design: A double-mask stack is used:
  • A lower mask defines the selective growth trenches for the TI.
  • An upper, partially free-standing mask acts as a shadow mask for depositing subsequent layers.
• Fabrication Sequence:
  1. TI Growth: A $(Bi,Sb)_2Te_3$ nanoribbon (18 nm thick, 200 nm wide, 1 $\mu$m long) is grown via SAG under sample rotation.
  2. Diffusion Barrier & Superconductor: A 3 nm Pt diffusion barrier and a 20 nm Aluminum (Al) superconducting layer are deposited from the same angle without rotation, forming a proximitized island on the TI ribbon.
  3. Tunnel Barrier: A stoichiometric $Al_2O_3$ tunnel barrier is grown under rotation. Due to shadowing by the upper mask, the effective barrier thickness is reduced to $\approx$1 nm (despite a nominal 5 nm), allowing for tunable tunneling.
  4. Normal Contacts: 40 nm Platinum (Pt) contacts are evaporated from the opposite angle to define the measurement leads.
• Device Architecture: The resulting device features a TI nanoribbon with a central superconducting island (proximitized by Al) and normal metal contacts on either side, separated by ultrathin tunnel barriers.

3. Key Contributions

• Novel Fabrication Platform: The first demonstration of a fully in situ, multi-angle stencil lithography process for creating hybrid TI-superconductor nanostructures without post-growth etching or exposure to air.
• Pristine Interfaces: The technique prevents TI surface oxidation prior to superconductor deposition, addressing the primary bottleneck in TI-based hybrid devices.
• Scalability: The process allows for the fabrication of multiple devices on a single chip with reproducible tunnel barriers, enabling complex multi-barrier architectures.

4. Experimental Results

Transport measurements were performed in a dilution refrigerator (base temperature 10 mK, electron temperature ~50 mK).

• Coulomb Blockade: The device exhibits robust, periodic Coulomb blockade peaks as a function of gate voltage ($V_G$), confirming the formation of a well-defined charge island.
  • Charging Energy ($E_C$): Extracted as 95 $\mu$eV.
  • Gate Control: Stable Coulomb diamonds were observed over a wide gate voltage range, indicating excellent electrostatic control and high-quality tunnel barriers.
• Proximity-Induced Superconductivity:
  • Gap Suppression: At zero magnetic field, low-energy conductance outside the Coulomb diamonds is strongly suppressed, consistent with a proximity-induced superconducting gap ($\Delta^* \approx 150$ $\mu$eV).
  • Magnetic Field Response: As the in-plane magnetic field increases (up to 0.6 T), the superconductivity is suppressed, and the low-energy conductance recovers. The "gap" closes, confirming the superconducting nature of the island.
• Parity and Periodicity:
  • The study investigated the transition from $2e$ (Cooper pair) to $1e$ (single electron) charging periodicity, which is expected if MZMs are present.
  • Observation: No transition to $1e$ periodicity was observed; the peak spacing remained $2e$-periodic.
  • Analysis: The absence of parity preservation is attributed to subgap quasiparticle states (a "soft gap") within the island, likely caused by the diffusion barrier or normal metal leads acting as quasiparticle sources. This prevents the conservation of parity required to observe MZM signatures in this specific configuration.

5. Significance and Outlook

• Validation of Platform: This work establishes a versatile and scalable nanofabrication platform for TI-based hybrid quantum devices, overcoming the critical issue of surface degradation.
• Foundation for Future Research: While this specific device did not exhibit the parity protection associated with MZMs (due to quasiparticle poisoning), the successful creation of a proximitized TI charge island with a clean interface provides a necessary foundation.
• Future Directions: The authors suggest that optimizing material combinations (e.g., removing the diffusion barrier or using superconducting leads) and further interface engineering could suppress quasiparticle states, potentially enabling the observation of topological superconductivity and Majorana physics in future iterations of this device architecture.

In summary, the paper represents a significant technical breakthrough in the fabrication of TI-superconductor hybrids, shifting the focus from material synthesis challenges to the physics of topological superconductivity in controlled nanostructures.