🔬 materials science

Solid-state reaction synthesis of superconducting refractory-metal germanides for Ge/SiGe quantum devices

This study demonstrates the low-temperature solid-state synthesis of superconducting refractory-metal germanides (V-Ge, Ta-Ge, and Nb-Ge) for Ge/SiGe quantum devices, revealing that while high-quality superconducting layers can be formed at 390°C, limited interfacial reaction depth prevents effective integration with buried quantum wells.

Original authors: Sebastiaan R. Roelofs, Jan Cornelis Wolff, Nick van Loo, Karina L. Hudson, Giordano Scappucci, Greg P. Mazur

Published 2026-10-06

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Sebastiaan R. Roelofs, Jan Cornelis Wolff, Nick van Loo, Karina L. Hudson, Giordano Scappucci, Greg P. Mazur

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the quest to build the next generation of quantum computers, scientists are turning to a material called germanium. Imagine a crystal lattice, a perfectly ordered grid of atoms, where tiny particles called holes—places where an electron is missing—can move with incredible speed and spin like tops. These holes are promising candidates for storing quantum information, the fundamental data of a future computer. However, for these quantum bits to work together in a complex machine, they need to talk to one another over distances. To facilitate this conversation, researchers want to place superconductors, materials that conduct electricity with zero resistance, right next to the germanium. The challenge is that these two materials do not naturally mix well. If you simply place a superconductor on top of the germanium, the connection is often weak or damaged. Scientists have been searching for a way to fuse them together at low temperatures, because heating the germanium too much would destroy the delicate quantum properties they are trying to preserve.

A team of researchers set out to solve this puzzle by testing three specific metals: vanadium, tantalum, and niobium. These are known as refractory metals, meaning they are incredibly tough and have high melting points. The goal was to see if these metals could react with germanium to form a new, superconducting layer without needing the extreme heat usually required for such chemical changes. The researchers worked with two types of samples. The first was a simple block of pure germanium. The second was a more complex structure, a sandwich of silicon and germanium layers, which is the actual platform used for building quantum devices. In this sandwich, the useful germanium is buried deep inside, protected by a top layer of silicon-germanium alloy. The team deposited thin films of the three metals onto both types of samples and heated them to a modest 365 to 390 degrees Celsius, a temperature low enough to keep the quantum device safe.

The results on the pure germanium were a clear success. When the metals were applied in small steps, with pauses to let the heat do its work, they reacted completely through the entire film. The vanadium, tantalum, and niobium mixed thoroughly with the germanium atoms to create new compounds that became superconducting. The tantalum mixture began conducting electricity without resistance at a temperature of 2.1 degrees above absolute zero, while the niobium mixture did so at 2.8 degrees. The niobium layer was particularly robust, remaining superconducting even when exposed to a strong magnetic field. This proved that it is physically possible to create these useful superconducting layers using only low heat.

However, the story changed when the researchers applied the same process to the complex silicon-germanium sandwich. Instead of reacting all the way through the metal film, the chemical reaction stopped almost immediately. The metals only mixed with the top layer of the sandwich for a distance of about 5 to 12 nanometers, which is roughly the width of a few dozen atoms. The rest of the metal film sitting on top remained unreacted and metallic. When the team measured the electrical properties of these samples, the superconducting signal they detected came almost entirely from this leftover, unreacted metal layer, not from the thin layer of new material that had formed at the interface. The reaction simply did not have the energy or time to penetrate deep enough to reach the buried germanium quantum well where the action needs to happen.

This finding highlights a specific bottleneck in the path toward better quantum computers. The researchers demonstrated that the low-temperature chemistry works perfectly on pure germanium, but the silicon-germanium barrier acts as a gatekeeper, slowing down the reaction so much that it cannot reach the target. The team suggests that the difference lies in how the atoms move. In the pure germanium, the atoms can shuffle around and mix relatively easily. In the silicon-germanium alloy, the presence of silicon atoms creates a more crowded and complex environment that hinders this movement. The researchers propose that to make this work for real devices, they might need to heat the samples closer to the limit of what the device can survive, or perhaps find a way to make the top barrier thinner so the reaction has less distance to travel. For now, the study confirms that the materials can be made, but the journey to get them to the right spot inside the device remains a significant hurdle.

Technical Summary: Solid-state reaction synthesis of superconducting refractory-metal germanides for Ge/SiGe quantum devices

Problem Statement
Planar Ge/SiGe heterostructures are a leading platform for hole-spin qubits, offering high mobility, strong spin-orbit coupling, and isotopic purity. Integrating superconductors into these heterostructures is essential for creating hybrid devices such as Josephson junctions, proximitized quantum dots, and topological superconductors. However, current methods typically rely on etching the SiGe barrier to deposit metals, which can damage the semiconductor surface and reduce interface transparency. An alternative strategy involves forming the superconductor directly within the heterostructure via solid-phase reaction. While previous work demonstrated this with Pt-based alloys (PtSiGe), these materials suffer from low critical temperatures ( $T_c \approx 0.5$ K) and low critical fields. Refractory-metal germanides (V, Ta, Nb) offer the potential for higher energy and magnetic-field scales, including A15 phases with $T_c$ up to ~23 K. The central challenge is whether these refractory metals can form useful superconducting phases within the strict thermal budget of Ge/SiGe heterostructures (typically capped at 500°C to preserve the quantum well) and whether the reaction can penetrate the SiGe barrier sufficiently to reach the buried quantum well.

Methodology
The authors investigated the solid-state reactions of Vanadium (V), Tantalum (Ta), and Niobium (Nb) with elemental Germanium (Ge) and with the Si $_{0.2}$ Ge $_{0.8}$ top barrier of Ge/SiGe heterostructures.

Substrates: Experiments were conducted on both bulk undoped Ge(001) wafers and Ge/SiGe heterostructures (16 nm strained Ge quantum well buried under 55 nm Si $_{0.2}$ Ge $_{0.8}$ ).
Processing: Films (30–40 nm total thickness) were deposited via electron-beam evaporation in ultra-high vacuum at temperatures between 365°C and 390°C. Two deposition strategies were compared:
1. Step-wise: Deposition of ~2 nm metal increments followed by 5–15 minute in-situ anneals.
2. Continuous: Full thickness deposition followed by a single anneal.
Characterization:
- Transport: Four-terminal resistance measurements in a dilution refrigerator (20 mK to several Kelvin) with vector magnet fields up to 1 T (out-of-plane) and 5.5 T (in-plane). $T_c$ and critical fields ( $B_c$ ) were defined at 10% of the normal-state resistance ( $R_N$ ).
- Microscopy: Cross-sectional High-Angle Annular Dark-Field Scanning Transmission Electron Microscopy (HAADF-STEM) and Energy-Dispersive X-ray Spectroscopy (EDX) were used to analyze layer thickness, intermixing depth, and atomic composition.

Key Results
The study reveals a stark contrast between reactions on elemental Ge and on the SiGe barrier:

Reactions on Elemental Ge:
- All three metals (V, Ta, Nb) reacted completely through the deposited films when processed step-wise at 390°C.
- V–Ge: Formed a disordered/multiphase layer (V:Ge $\approx$ 1:1) with $T_c \approx 0.14$ K.
- Ta–Ge: Formed a reacted layer (Ta:Ge $\approx$ 2:3) with $T_c \approx 2.1$ K and an out-of-plane $B_c \approx 0.19$ T at 1.8 K.
- Nb–Ge: Formed a reacted layer (Nb:Ge $\approx$ 5:4) with $T_c \approx 2.8$ K and an out-of-plane $B_c \approx 0.77$ T at 2.1 K.
- Mechanism: Microscopy of the V/Ge sample indicated a Kirkendall-type reaction where Ge diffuses into the metal faster than the metal diffuses into the substrate, creating voids on the Ge side.
Reactions on Ge/SiGe Heterostructures:
- Under identical thermal budgets, the reaction was confined to a thin interfacial region (5–12 nm) near the metal–SiGe interface.
- A substantial fraction of the deposited metal remained unreacted.
- Transport Interpretation: The measured superconducting transitions in heterostructure samples (e.g., $T_c \approx 4.15$ K for V, $1.04$ K for Ta, $4.7$ K for Nb) are attributed primarily to the residual unreacted metal films rather than the thin reacted interfacial layer. For instance, the Ta sample's properties resemble disordered $\beta$ -Ta rather than Ta-germanides.
- Nb/SiGe: Showed the deepest intermixing (~12 nm), consisting of a ~7 nm Nb-Ge-rich region and a ~5 nm Nb-Si-Ge region, yet the superconducting response was still dominated by the remaining Nb film.

Significance and Claims
The paper establishes that low-temperature solid-state reactions (365–390°C) are sufficient to form superconducting refractory-metal germanides on elemental Ge, with Nb–Ge showing the most favorable properties ( $T_c = 2.8$ K, $B_c \approx 0.8$ T).

However, the authors conclude that the principal obstacle to applying this approach to buried quantum wells is the limited interfacial reaction kinetics within the SiGe barrier. The reaction stalls after penetrating only a few nanometers, leaving the quantum well isolated from the superconductor by a thick, unreacted metal layer. The authors do not claim to have solved the problem of reaching the buried well but rather identify the kinetic limitation as the specific barrier to be overcome.

The discussion suggests three potential paths forward without asserting their success:

Raising the processing temperature to the heterostructure growth limit (500°C), as demonstrated with PtIrSiGe.
Utilizing substantially longer post-deposition anneals to increase reaction depth.
Reducing the top barrier thickness to minimize the distance between the reaction zone and the quantum well.

The work successfully decouples the ability to form superconducting germanides at low temperatures from the kinetic challenge of extending those reactions through a SiGe alloy barrier.

More like this