A germanium quantum well has produced a fully spin-polarized quantum Hall state at magnetic fields as low as 0.25 tesla — an order of magnitude below what semiconductor systems have conventionally required — demonstrating for the first time that a landmark topological quantum state is achievable in a CMOS-compatible chip without high-field superconducting magnets or millikelvin cryostats, according to a study by M. Myronov, W. Jiang, and S. Studenikin.

Published today as open access in Communications Materials, the study by M. Myronov of the University of Warwick and W. Jiang and S. Studenikin of the National Research Council of Canada does more than lower a threshold. It opens the physical possibility of building a single germanium chip that integrates individually addressable spin qubits with topological edge channels — without the multi-tesla magnets that have kept these two regimes of quantum physics in separate experimental worlds. That combination, if realized in hardware, would represent one of the cleaner paths yet identified toward fault-tolerant quantum computation in a manufacturable semiconductor platform.

Why Sub-Tesla Operation Changes What Quantum Engineers Have to Build

The integer quantum Hall effect is one of condensed matter physics' most precisely characterized phenomena, as covered in this topological quantum Hall transport review. When a two-dimensional conductor is cooled and placed in a perpendicular magnetic field, charge carriers organize into discrete energy levels called Landau levels, and their transport along the sample's physical edges becomes topologically protected — immune to scattering from impurities or geometric disorder. The Hall resistance locks to exact rational multiples of h/e², where h is Planck's constant and e is the electron charge. The precision is so absolute that it now anchors the international definition of electrical resistance.

Within this physics, the filling factor ν=1 state is the simplest fully topological state available: it requires that the lowest Landau level be occupied exclusively by carriers of one spin orientation. Reaching that state in conventional semiconductors such as GaAs has historically demanded several tesla, because two competing energy scales — the cyclotron energy separating Landau levels and the Zeeman energy splitting spins within each level — only resolve in the right ratio at very high fields. In most materials the Zeeman splitting is far smaller than the Landau-level spacing at modest fields, so spin polarization is incomplete until deep in the multi-tesla regime.

Germanium plays by different rules.

Holes — the positively charged quasi-particles that carry current in p-type germanium quantum wells — possess an unusually large effective Landé g-factor. This dimensionless quantity sets how strongly a particle's spin couples to an external magnetic field: the larger the g-factor, the greater the Zeeman splitting at any given field. In compressively strained germanium wells grown on silicon, the out-of-plane effective g-factor for holes ranges from roughly 13 to 24 depending on gate-tunable electric field conditions — roughly an order of magnitude larger than the g-factor of approximately 2 that applies to electrons in silicon.

This intrinsically large Zeeman energy is one of two features that make the result possible. The other is carrier density.

The team operated their device in what they describe as an ultra-dilute carrier regime: a two-dimensional hole gas with a density low enough that the Landau-fan structure places the ν=1 filling factor at anomalously low magnetic fields. At conventional carrier densities, Landau-level broadening — the smearing of individual energy levels by disorder in the material — exceeds the Zeeman splitting at modest fields, collapsing the spin gap. In the Warwick-NRC germanium device, magnetotransport and Landau-fan measurements confirm that broadening remains smaller than the Zeeman splitting all the way down to 0.25 tesla. The Hall resistance locks to h/e², deviating from its universal value by less than one part in many thousands — meeting the standard definition of a well-quantized integer quantum Hall plateau.

The engineering implication is direct. Dilution refrigerators — the ultralow-temperature cryostats most quantum hardware runs on — can generate modest magnetic fields with simple coil designs, but sustaining several tesla demands bulky, high-current superconducting solenoids adding cost, weight, and complexity. Operating at a quarter of a tesla relaxes all of those constraints. The result also holds to 1.5 kelvin, a temperature accessible with ordinary helium-3 cryostats rather than the dilution refrigerators required for millikelvin operation. Sub-tesla magnetic fields also reduce the risk of unwanted crosstalk in multi-qubit arrays — a growing concern as the field moves from handfuls of qubits toward processors with hundreds or thousands of them.

Germanium's Large g-Factor: Why the Number Matters

The physics connecting germanium's g-factor to practical quantum hardware is worth unpacking, because the number doing the work is not obviously significant without context.

The Zeeman energy — the energy difference between spin-up and spin-down states in a magnetic field — scales as E_Z = g·μ_B·B, where μ_B is the Bohr magneton and B is the field strength. For a material with g=2 (silicon electrons), achieving a Zeeman splitting of, say, 0.2 meV requires a field of roughly 1.7 tesla. For germanium holes with g=15, the same splitting is reached at 0.11 tesla. Combine that with the ultra-dilute carrier regime shifting the ν=1 Landau level to low fields, and the two conditions that make spin polarization difficult in conventional systems — insufficient Zeeman splitting and Landau-level broadening exceeding the spin gap — are both addressed simultaneously.

The gate-tunability of the g-factor (from roughly 13 to 24 depending on perpendicular electric field) is an additional tool: it means experimenters can fine-tune the Zeeman energy relative to the Landau-level spacing without changing the magnetic field, giving a new axis of control that is entirely absent in silicon. This tunability was documented in Myronov et al., Communications Materials 2023.

Topological Edge States and the Path to Exotic Quantum Physics

The ν=1 quantum Hall state achieved in the germanium device is not merely a convenient experimental marker. It is a genuinely topological state: the physical boundary of the device hosts a single, spin-polarized, chiral edge channel — a one-dimensional highway for holes that flows in one direction only and is immune to backscattering from defects or disorder. This edge channel is the foundational architecture for topological quantum transport.

At and near ν=1 in a fully spin-polarized system, theoretical predictions become rich. When the spin degree of freedom is fully aligned and Landau levels are sharp, exchange-driven instabilities can produce Skyrmionic spin textures: topological excitations in the spin density of the hole gas that behave as composite objects and carry unusual charge and statistics, as predicted by Sondhi et al., Phys. Rev. B. In certain correlated limits, filling factors near ν=1 also host fractional quantum Hall states — states whose quasi-particle excitations obey non-Abelian anyonic statistics. Non-Abelian anyons are the class of exotic excitations that topological quantum computers would exploit as an inherently fault-tolerant basis for encoding quantum information, as reviewed by Nayak et al., Rev. Mod. Phys..

What makes the Warwick-NRC result specifically significant for topological quantum computing research is that these exotic states were previously accessible in this class of experiment only in ultra-high-purity GaAs heterostructures at multi-tesla fields and millikelvin temperatures — experimental conditions that have essentially nothing in common with the CMOS manufacturing environment. The new germanium result places them, in principle, within reach of a standard helium-3 cryostat at sub-tesla fields, in a material that is already the substrate for a functioning spin-qubit quantum processor.

Equally significant is a companion paper from Myronov, Studenikin, and colleagues published in APL Quantum in June 2026, proposing quantum phononic links in compressively strained germanium-on-silicon for long-range qubit coupling, per Myronov et al., APL Quantum 2026. Together, the two papers indicate a coordinated research program using the same material platform to attack multiple quantum computing bottlenecks simultaneously.

Germanium's Moment in Semiconductor Quantum Computing

The new result arrives as germanium is consolidating its position as a leading candidate for scalable semiconductor quantum computing. The past two years have generated a sustained sequence of milestones in the material. A 10-spin qubit array in germanium from QuTech at Delft achieved single-qubit gate fidelities exceeding 99% across a two-dimensional 3-4-3 layout in late 2025, advancing the foundational geometry needed for surface-code error correction, per John et al., Nature Communications. Rabi frequencies exceeding 540 MHz have been demonstrated in germanium hole-spin qubits operating at just 100 millitesla — far above typical operating speeds in silicon spin qubits, as shown in Nature Communications, 2022.

In April 2026, Delft-based startup Groove Quantum announced the world's largest semiconductor spin-qubit processor at 18 qubits — built on CMOS-compatible germanium — alongside €16 million in combined equity and grant funding. CEO Anne-Marije Zwerver described the processor as marking a definitive transition from a promising concept to a scalable platform, noting that the company aims to achieve 100-qubit systems and transition production to established semiconductor foundries.

Germanium's advantages over silicon for spin qubits are several and specific. In silicon, the weak spin-orbit coupling makes it difficult to manipulate individual qubits without incorporating micromagnets or other magnetic structures — adding fabrication complexity and potential crosstalk. In germanium, the large effective g-factor means qubits respond strongly to modest electrical pulses and can be driven at high speed without those additional structures. Holes in germanium also benefit from having their wave function nodes at nuclear sites, suppressing unwanted hyperfine interactions with germanium's predominantly spin-0 nuclear isotopes.

Charge Noise Remains a Known Limitation for Scaling

The case for germanium is strong and growing, but the engineering challenges that remain are real and should not be understated. The same strong spin-orbit interaction that enables fast electrical qubit control also makes germanium hole-spin qubits sensitive to low-frequency charge noise — fluctuations in the electrostatic environment caused by defects in the gate dielectric or heterostructure interfaces. Martinez and Niquet, Phys. Rev. Applied published in January 2026 found that while charge properties of planar germanium qubits show moderate variability, spin properties — including g-factors and Rabi frequencies — show significant dispersion across a device, with implications for large-scale architectures. Managing this variability at the scale of hundreds or thousands of qubits remains an active and unsolved research problem.

The sub-tesla quantum Hall result itself is a proof-of-concept at ultra-dilute carrier densities. Practical integration of topological edge channels with a multi-qubit spin-qubit array — the hybrid chip that would represent the full payoff of this research line — has not been demonstrated. The authors identify that as a primary direction for future work, alongside probing exchange-driven instabilities near ν=1 and searching for fractional quantum Hall states at sub-tesla fields.

What Was Funded and What Comes Next

The paper was funded by the Quantum Sensing Program of the National Research Council of Canada and by EPSRC in the United Kingdom. It was received in January 2026, accepted in June, and published today as open access.

Immediate experimental directions include characterizing the topological edge states directly using local probe techniques, searching for fractional quantum Hall states in the sub-tesla regime, and probing Skyrmionic exchange instabilities near ν=1 — a class of correlated physics that has never been studied in a CMOS-compatible semiconductor at these accessible conditions. Longer-term, the ability to produce a well-defined fully spin-polarized topological state in the same material used for spin qubits creates an opening for hybrid architectures in which topological edge channels and individually addressable spin qubits coexist on the same germanium-on-silicon chip.

For the quantum computing field, the message is substantive: germanium is not merely a convenient material for spin qubits. It is rich enough in accessible quantum physics to support the class of topological behavior that, until today's result, required either completely different material systems or experimental conditions far beyond practical reach.

A germanium chip. A magnet weaker than a typical MRI scanner's fringe field. A temperature reachable with liquid helium-3. One of quantum physics' most pristine topological states.

Frequently Asked Questions

Why does germanium produce a topological quantum Hall state at such low magnetic fields when other semiconductors cannot?

Germanium holes — the positive charge carriers in p-type germanium quantum wells — have an effective Landé g-factor of 13 to 24, roughly an order of magnitude larger than the g-factor of electrons in silicon or gallium arsenide. The g-factor sets how strongly a carrier's spin responds to a magnetic field; a larger g-factor produces a larger Zeeman spin-splitting energy at any given field. Combined with an ultra-dilute carrier density that places the ν=1 Landau level at anomalously low field values, the Zeeman splitting in germanium exceeds the disorder-induced energy smearing that normally washes out spin polarization — and does so at 0.25 tesla rather than the several tesla required in conventional materials. The g-factor is also gate-tunable by an applied electric field, giving experimenters an additional control axis that silicon lacks entirely.

What is the connection between this result and topological quantum computing?

The ν=1 quantum Hall state is the simplest fully topological state in a two-dimensional semiconductor: its physical edge supports a single, spin-polarized, chiral channel immune to backscattering. Near ν=1 in a spin-polarized system, theory predicts Skyrmionic spin textures and, in certain correlated regimes, fractional quantum Hall states whose quasi-particle excitations are non-Abelian anyons — the exotic particles that topological quantum computers would exploit as inherently fault-tolerant qubits. Until now, studying those states in a semiconductor required ultra-high-purity GaAs at multi-tesla fields. The Warwick-NRC result brings them within experimental reach, in principle, in a CMOS-compatible germanium chip at sub-tesla fields and above-1-kelvin temperatures.

Could a future germanium chip integrate both spin qubits and topological edge channels on the same device?

That is precisely the architecture the paper points toward, and it represents the result's largest long-term implication. Because the quantum Hall experiment and existing germanium spin-qubit processors use the same compressively strained germanium-on-silicon heterostructure, the physical platform already exists in which both phenomena could in principle coexist. No one has yet demonstrated a working hybrid device, and the engineering challenges — controlling charge noise, achieving compatible operating conditions, and wiring both subsystems — are substantial. But lowering the required magnetic field from several tesla to sub-tesla removes what had been one of the most fundamental obstacles: the impossibility of operating a spin-qubit array and a topological edge channel under the same modest magnetic field on the same chip.

Why does operating at 1.5 kelvin instead of millikelvin temperatures matter practically?

Dilution refrigerators — required to reach millikelvin temperatures — are large, expensive, slow to cool, and require specialist operation. A helium-3 cryostat reaches 1.5 kelvin with a fraction of the engineering overhead: it is faster, smaller, cheaper, and far more accessible to research groups and eventually to industrial facilities. More significantly, most near-term quantum computing demonstrations at scale will need to run inside refrigerators that can accommodate hundreds or thousands of qubits. A physical regime that works at 1.5 K rather than at 10 mK dramatically broadens the thermal budget available for those systems — in terms of heat dissipation, wiring density, and classical control electronics co-located near the qubits.