Comparative assessment of germanium-based spin-qubit modalities: donor, acceptor, gate-defined hole, and gate-defined electron platforms

Abstract: High-purity germanium (Ge) has re-emerged as a versatile semiconductor platform for spin-based quantum information processing because it combines mature materials processing, access to spin-free isotopes, high mobilities, small effective masses, and strong but engineerable spin--orbit coupling. However, ``Ge qubits'' are not a single technology. Donor spin qubits, acceptor spin qubits, gate-defined hole spin qubits, and gate-defined electron spin qubits exploit different parts of the Ge band structure and therefore make distinct trade-offs among coherence, controllability, fabrication complexity, and scalability. Here we compare these four Ge-based spin-qubit modalities on a common physical and architectural footing. We review the shared Ge materials physics, including isotopic purification, the multivalley (L)-point conduction band, the spin-(3/2) valence band, heavy-hole/light-hole mixing, strain, interfaces, disorder, and phonons. We also introduce a common framework for estimating phononic-crystal-modified (T_1) using a calibrated reference relaxation rate, a geometry-dependent strain-density-of-states suppression factor, and parasitic relaxation channels. The comparison shows that gate-defined Ge hole-spin qubits currently offer the strongest combination of all-electrical control, demonstrated multiqubit operation, and scalability. Donor, acceptor, and gate-defined electron qubits remain important complementary directions for memory, hybrid, and exploratory architectures. Overall, Ge supports a diverse qubit ecosystem, with gate-defined hole-spin qubits presently providing the clearest path toward scalable Ge-based quantum processors.

• The paper presents a comprehensive analysis comparing donor, acceptor, gate-defined hole, and gate-defined electron qubit platforms in germanium.
• It details experimental and numerical trade-offs including Stark tunability, strain sensitivity, and phononic crystal integration for enhanced coherence.
• It highlights gate-defined hole spin qubits as the most mature platform, offering high-fidelity control and promising scalability for quantum processors.

Comparative Assessment of Germanium-Based Spin-Qubit Modalities

Overview and Context

This paper presents a comprehensive technical comparison of four germanium (Ge)-based spin-qubit modalities: donor spin qubits, acceptor spin qubits, gate-defined hole spin qubits, and gate-defined electron spin qubits (2605.13680). Ge is emerging as a versatile semiconductor platform due to its mature processing, access to spin-free isotopes, small carrier masses, high mobility, and strong, tunable spin--orbit coupling. The paper systematically analyzes materials physics, qubit encoding mechanisms, dominant decoherence channels, control paradigms, device architectures, and scalability prospects.

Materials Physics Relevant Across Platforms

Ge’s suitability as a qubit platform arises from its access to both conduction-band electrons (L-valleys, with complex anisotropic effective masses and multivalley degeneracy) and valence-band holes (spin-$3/2$ heavy-hole/light-hole manifold at the $\Gamma$-point, valley-free). Isotopic purification ($^{73}$Ge depletion) and zone refinement yield spin-quiet environments and ultra-low impurity densities, which are critical for coherence and reproducibility. Strain and interfaces play central roles: they can either be engineered for favorable band structure or introduce uncontrolled disorder, with electron modalities particularly sensitive to valley-orbit effects and hole modalities to strain-induced spin-orbit coupling.

Figure 1: Estimated average linear strain in Ge as a function of impurity density for representative donor/acceptor species, highlighting the critical impact of purity on background strain and qubit uniformity.

Phonon coupling modulates both spin and charge states; phononic crystal engineering can suppress undesired relaxation channels and enable hybrid architectures with tailored acoustic modes.

Donor Spin Qubits in Ge

Donor spin qubits encode quantum information in the spin of an electron bound to a substitutional donor (typically P, As, Sb). Atom-like confinement and strong Stark tunability offer high reproducibility and large electrical addressability. The $L$-valley conduction band introduces multivalley complexity and extended wavefunctions, easing spacing constraints while complicating control/variability. The dominant decoherence channel is spin-lattice relaxation, enhanced by Ge’s strong spin--orbit interaction. Experimental studies report $T_1$ times on the order of $0.6$ ms at $B_0=0.44$ T and $T=0.35$ K; coherence becomes $T_1$-limited in isotopically enriched Ge [Sigillito2015].

Figure 2: Architecture schematic for a Ge donor-spin platform embedded in a phononic crystal, illustrating bandgap engineering for coherence protection and phonon-mediated coupling.

Strong numerical results include tunable Stark shifts exceeding ensemble linewidths and the possibility of phononic crystal integration to suppress one-phonon decay and permit long-range coupling [Smelyanskiy2014DonorGePhononic]. However, deterministic donor placement, valley-orbit sensitivity, and integration complexity remain challenging.

Acceptor Spin Qubits in Ge

Acceptor qubits employ a valence hole bound to an acceptor impurity; these qubits inherit spin-$3/2$ physics enabling quadrupolar interaction, electric tunability, and pronounced strain sensitivity. Interface effects split the fourfold ground state into electrically tunable Kramers doublets, facilitating EDSR and protected operation points. Acceptor qubits theoretically bridge atom-like reproducibility and electrical control, with strong coupling to strain and possibility for phononic integration.

Figure 3: Schematic of a Ge acceptor-spin platform integrated with a phononic crystal, enabling both electrical and strain-mediated control over the spin-$\Gamma$0 manifold.

Numerical and theoretical analyses indicate that strain gradients as low as $\Gamma$1 nm$\Gamma$2 can enhance hole-spin Rabi frequencies by an order of magnitude [AbadilloUriel2023StrainSOI]. Nonetheless, central-cell, interface, and strain-induced spectral sensitivity—combined with residual hyperfine interaction from the dopant nucleus—pose significant reproducibility issues. Experimentally, Ge acceptor qubits lag behind their Si counterparts and Ge hole qubits.

Gate-Defined Hole Spin Qubits in Strained Ge/SiGe

Gate-defined hole spin qubits in strained Ge/SiGe heterostructures constitute the most mature Ge qubit modality. Electrostatic gates confine hole spins in a two-dimensional channel, enabling fast all-electrical control via strong and tunable spin--orbit coupling. Pauli blockade, EDSR, and exchange gates allow multiqubit operation; sweet-spot bias tuning suppresses charge-noise sensitivity while maintaining rapid manipulation.

Strong experimental milestones have been achieved: single-shot readout [Vukusic2018], two-qubit logic with fidelities $\Gamma$3 [Hendrickx2020Fast], four-qubit processors [Hendrickx2021FourQubit], and sweet-spot operation above 1 K with $\Gamma$4 gate fidelity [Hendrickx2024Sweet]. Quantum shuttling and scaling to 10-spin arrays with robust local control have been demonstrated [John2025, vanRiggelen2024].

Figure 4: Device schematic for a gate-defined Ge hole-spin qubit platform integrated with a phononic crystal cavity, showing EDSR control and engineered acoustic environment.

Integration of phononic crystals is now a credible pathway for dissipation engineering and local acoustic mode coupling, potentially enabling hybrid architectures with tailored $\Gamma$5 via bandgap suppression of phonon-mediated relaxation [Mei2025QST].

Gate-Defined Electron Spin Qubits in Ge

Gate-defined electron spin qubits encode quantum information in spin-$\Gamma$6 electrons confined in Ge/SiGe quantum dots. Conceptually simpler than hole qubits, electron qubits are compatible with well-established Si quantum-dot architectures. However, they inherit $\Gamma$7-valley degeneracy, anisotropic masses, and intervalley mixing, which severely complicate device physics, spectral uniformity, and reproducibility. Electrically tunable $\Gamma$8 tensors and valley engineering have been explored, but no strong numerical or circuit-level results matching hole-qubit achievements have been reported.

Figure 5: Schematic for a gate-defined Ge electron-spin qubit platform with phononic crystal integration, highlighting acoustic bandgap engineering for suppression of relaxation and possible hybrid coupling.

Currently, gate-defined electron qubits in Ge are scientifically attractive but experimentally immature, with uncertain paths to scalability or reproducibility.

Cross-Modality Trade-Offs and Implications

A structured technical comparison reveals key outcomes:

• Donor qubits: Atom-like, strong Stark tunability, but limited by spin-lattice relaxation and multivalley band structure.
• Acceptor qubits: Spin-$\Gamma$9 hole encoding, quadrupolar and strain tunability, but highly sensitive to central-cell and interface environment and not yet mature in Ge.
• Gate-defined hole qubits: All-electrical control, valley-free band structure, strong system-level scalability and demonstrated multi-qubit hardware; sensitivity to charge noise and $^{73}$0-tensor anisotropy that is mitigated by special bias points.
• Gate-defined electron qubits: Simple encoding, but experimental challenges due to $^{73}$1-valley complexity render them less viable than hole qubits at present.

Ge supports a layered qubit ecosystem. For processor-oriented scaling, gate-defined hole-spin qubits are most advanced, while donor and acceptor modalities are compelling for hybrid memory, sensor, and transducer architectures. Phononic crystal engineering is a crucial cross-platform axis, governing relaxation, coherence protection, and coupling design.

Future Directions in Ge-Based Quantum Technology

Priorities for future work include:

• Advancing deterministic donor placement and hybrid phononic integration.
• Establishing reproducible, interface-engineered acceptor qubits and benchmarking their quadrupolar physics.
• Scaling gate-defined hole qubits in clean, low-noise strained Ge/SiGe with robust $^{73}$2-tensor control, shuttling, and high-fidelity multi-qubit gates.
• Exploring gate-defined electron qubits to identify regimes where $^{73}$3-valley physics can be an asset, not a liability.

A realistic roadmap is layered: gate-defined hole qubits as the scalable processor core, donor and acceptor qubits for memory or hybrid connections, and phononic engineering as a systems-level enabler for dissipation control and coupling.

Conclusion

Germanium offers a genuinely diverse qubit ecosystem, with distinct advantages and liabilities across donor, acceptor, gate-defined hole, and gate-defined electron modalities. Among these, gate-defined hole qubits presently represent the clearest route toward scalable, high-fidelity, and hardware-efficient quantum processors, supported by strong numerical results and robust experimental demonstrations. Donor, acceptor, and electron platforms provide valuable complementary and exploratory pathways—especially for hybrid architectures, quantum memory, and sensor applications. The theoretical and practical implications are profound: Ge’s band structure, materials purity, and compatibility with phononic engineering will continue to shape the future of scalable quantum information processing architectures.