Germanium-Based Quantum Computing

• Germanium-based quantum computing is a promising approach that employs hole-spin quantum dots with strong spin–orbit coupling and reduced hyperfine interaction for robust qubit realization.
• It utilizes advanced device architectures in platforms like Ge/SiGe and Ge₁₋ₓSnₓ/Ge, enabling precise qubit addressability and fast all-electrical manipulation.
• Key performance metrics, including millisecond T₁ relaxation times and Rabi frequencies up to 100 MHz, underscore its potential for scalable, error-corrected quantum processors.

Hole-spin quantum dots are nanoscale semiconductor devices in which the spin degree of freedom of an individual or few valence-band holes is isolated, controlled, and read out. Holes, the absence of electrons in the valence band, behave as positively charged, spinful quasiparticles with strong spin–orbit coupling and, in most materials/interfaces, significantly weaker hyperfine coupling to nuclear spins compared to electrons. These properties make holes in quantum dots (QDs) attractive quantum bit (qubit) candidates for scalable quantum information processing, offering prospects for rapid all-electrical manipulation, long-lived coherent spin states, and dense two-dimensional architectures. Realization platforms span silicon and planar Ge, GeSn, and III–V semiconductors, with key physical metrics—relaxation times, tuning rates, qubit addressability, and gate fidelities—now reaching thresholds suitable for error-corrected quantum computing.

1. Physical Realization: Device Architectures and Orbital Structure

Hole-spin quantum dots are fabricated using gate-defined, self-assembled, or nanowire-based approaches across group-IV (Si, Ge, GeSn) and III–V (InGaAs, GaAs) material systems. In planar architectures, electrostatic gates on semiconductor heterostructures (e.g., Ge/SiGe, Si/SiO₂, Ge₁₋ₓSnₓ/Ge) define confining potentials for holes, creating single or arrays of quantum dots with tunable tunnel coupling and local plunger control (Lawrie et al., 2020, Liles et al., 2018, Hendrickx et al., 2019, Liles et al., 2023, Rotaru et al., 24 Feb 2025). Self-assembled nanocrystal (NC) or nanowire (NW) quantum dots (e.g., InSb NWs, SiGe domes) also enable strong confinement and improved g-factor control (Pribiag et al., 2013, Ares et al., 2013).

The low-energy single-particle levels in a dot are accurately described by the two-dimensional (2D) Fock–Darwin spectrum: $$E_{n_r,m} = \hbar\Omega(2n_r + |m| + 1) - m\,\hbar\omega_c/2$$ where $\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$, $\omega_0$ is the confinement frequency set by gate potentials, and $\omega_c = eB/m^*$ is the cyclotron frequency. Absence of valley degeneracy in planar Ge and Si architectures ensures that, unlike electrons, each orbital level is only two-fold spin-degenerate, facilitating unambiguous shell filling sequences (Liles et al., 2018, Lawrie et al., 2020).

Strong Coulomb interactions between holes further suppress excited-state singlet-triplet splittings relative to single-particle gaps, with measured reductions to as low as 10% of orbital spacings in Si MOS devices (Liles et al., 2018).

2. Spin-Orbit Coupling and Electrical Manipulation

Hole states, formed from $p$-orbital valence bands with $j=3/2$, exhibit inherently strong spin–orbit coupling (SOC), originating both from atomic and structural inversion asymmetry (Rashba) in low-dimensional heterostructures. The dominant SOC terms include cubic-in-momentum Rashba ($\sim p^3$) and, depending on symmetry, Dresselhaus components, which mediate electric-dipole spin resonance (EDSR) transitions driven by microwave fields on gates (Fernandez-Fernandez et al., 2022, Terrazos et al., 2018, Rotaru et al., 24 Feb 2025).

The EDSR dipole moment and Rabi frequency are parametrically large for holes, with values approaching $d\sim 1$ $e$·pm and Rabi rates $f_{Rabi} \sim 10$–$\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$0 MHz (for $\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$1 kV/m), with the out-of-plane magnetic field maximizing dipole coupling (Rotaru et al., 24 Feb 2025, Terrazos et al., 2018). All-electrical $\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$2-tensor modulation, realized by gate tuning or displacement, further enables fast Rabi oscillations ($\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$3 MHz), sidestepping the need for local AC magnetic fields (Ares et al., 2013).

Electrical tunability of the hole $\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$4-factors is exceptionally strong, with tuning slopes $\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$5 up to $\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$6 MHz/mV on target gates, and cross-talk rates below $\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$7 MHz/mV on neighbors in array geometries (Lawrie et al., 2020).

3. Spin Relaxation and Coherence: Hyperfine and Phonon Limits

The relaxation ($\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$8) and dephasing ($\Omega = \sqrt{\omega_0^2 + (\omega_c/2)^2}$9) times of hole spins are determined by a competition between spin–orbit-assisted phonon coupling and hyperfine interactions. Millisecond $\omega_0$0 have been measured for single-hole Ge dots at $\omega_0$1 T, with lower values for multi-hole states (e.g., $\omega_0$2 ms) (Lawrie et al., 2020, Hendrickx et al., 2019).

The leading spin relaxation mechanism is spin–orbit–assisted phonon emission, with the rate scaling as

$\omega_0$3

at low fields, and potentially as $\omega_0$4 out-of-plane in GeSn due to higher-order admixture terms (Rotaru et al., 24 Feb 2025). The underlying physical origin is the necessity of both spin–orbit mixing (linear in $\omega_0$5) and phonon density of states ($\omega_0$6), yielding $\omega_0$7 scaling at low $\omega_0$8. For $\omega_0$9, this mechanism dominates over hyperfine-induced relaxation in group-IV systems where nuclear spins are sparse or can be isotopically purified.

Dephasing $\omega_c = eB/m^*$0 is governed by hyperfine noise (significantly reduced for $\omega_c = eB/m^*$1-like holes; the Ising form $\omega_c = eB/m^*$2 dominates), but nonzero heavy-hole/light-hole (HH-LH) mixing and nuclear dipolar fields yield $\omega_c = eB/m^*$3–$\omega_c = eB/m^*$4 ns in typical In(Ga)As QDs (Dahbashi et al., 2011), extendable to $\omega_c = eB/m^*$5 $\omega_c = eB/m^*$6s–ms regime by narrowing the nuclear bath and/or isotopic purification (Fischer et al., 2010, Chekhovich et al., 2010).

4. Qubit Operation: Initialization, Manipulation, and Readout

Purely electrical manipulation of hole spins is implemented via gate-driven EDSR or $\omega_c = eB/m^*$7-tensor modulation. Single-shot spin blockade and readout exploit Pauli spin blockade (PSB): triplet-triplet transitions are forbidden, enabling high-contrast charge-state conversion for spin-to-charge measurement (Hendrickx et al., 2019, Lawrie et al., 2020, Liles et al., 2023).

Optical initialization and readout (in III–V dots) leverage ultrafast resonant or two-photon Raman transitions in $\omega_c = eB/m^*$8-systems defined by the heavy-hole ground states and optically excited trions (Godden et al., 2011, Barcan et al., 7 Mar 2025). Optical spin pumping yields initialization fidelities exceeding 99% in double-$\omega_c = eB/m^*$9 schemes, with site-controlled InGaAs pyramidal QDs demonstrating robust, high-fidelity preparation and readout across a broad range of conditions (Barcan et al., 7 Mar 2025).

Two-qubit and multi-qubit operations are enabled through capacitively tuned exchange coupling (in double or coupled dots), optically gated exchange, or cavity-mediated sideband gates. Experimental gate fidelities above 99% for universal operations have been demonstrated with fast-adiabatic protocols ($p$0–200 ns gate times) (Fernandez-Fernandez et al., 2022, Greilich et al., 2011).

5. Materials Platforms and Tunability

Hole-spin QDs have been realized and analyzed in:

• Si/SiO₂ (MOS): Planar dots with full CMOS compatibility; no valley degeneracy; $p$1–1.4, strong shell structure, and high interaction energies (90% singlet-triplet suppression) (Liles et al., 2018, Liles et al., 2023).
• Ge/SiGe, Ge₁₋ₓSnₓ/Ge: Highly tunable HH-LH splitting (from direct strain and Sn composition), light effective mass ($p$2), Rabi rates up to 100 MHz, no valley physics, and strong spin–photon interfaces in Sn-alloyed direct-gap material (Terrazos et al., 2018, Rotaru et al., 24 Feb 2025).
• SiGe domes and InSb nanowires: Large, electrically tunable $p$3-factors, fast Rabi control via g-tensor modulation, and strongly anisotropic spin-orbit coupling; weak hyperfine interaction ($p$4) (Pribiag et al., 2013, Ares et al., 2013).
• III–V self-assembled QDs (InAs, InGaAs; [001] and [111] growth): Strongly suppressed hyperfine coupling (Ising form), fast optical gates (ps–ns), and optically controllable arrays (Godden et al., 2011, Dahbashi et al., 2011, Barcan et al., 7 Mar 2025). Symmetry engineering in quantum dot molecules (QDMs) enables tunable spin–orbit-enhanced mixing for electrically switchable spin gates (Planelles et al., 2015, Economou et al., 2012).
• Curved and nanomembrane architectures: Ge/Si curved quantum wells and suspended structures provide highly optimized SOC, single- or multimodal microwave photon coupling (spin–photon $p$5–100 MHz), and sweet spots for charge noise insensitivity (Bosco et al., 2022, Sagaseta et al., 6 Oct 2025).

6. Figures of Merit and Scalability Prospects

Key operational metrics as realized in advanced devices include:

Metric	State-of-the-Art Benchmark	Reference
$p$6 (single hole, Ge QD)	$p$7 ms	(Lawrie et al., 2020)
$p$8 (5 holes, Ge QD)	$p$9 ms	(Lawrie et al., 2020)
$j=3/2$0 (Ge QD, Ramsey)	$j=3/2$1–$j=3/2$2 μs	(Hendrickx et al., 2019, Liles et al., 2023)
$j=3/2$3 (Si QD, echo)	$j=3/2$4 μs	(Liles et al., 2023)
$j=3/2$5 (EDSR/$j=3/2$6-modulation)	$j=3/2$7–$j=3/2$8 MHz	(Ares et al., 2013, Rotaru et al., 24 Feb 2025)
$j=3/2$9 (gate tun.)	$\sim p^3$0 MHz/mV	(Lawrie et al., 2020, Hendrickx et al., 2019)
Neighbor cross-talk ratio	$\sim p^3$10.03	(Lawrie et al., 2020)
Initialization/readout fidelity	$\sim p^3$2 (optical, PSB)	(Barcan et al., 7 Mar 2025, Hendrickx et al., 2019)
Two-qubit gate fidelities	$\sim p^3$3 (sim./exp.)	(Fernandez-Fernandez et al., 2022, Greilich et al., 2011)

These figures, together with the suppression of valley degeneracy (group IV), high addressability, and compact 2D gate layouts, establish hole-spin quantum dots as primary contenders for integration into scalable, error-corrected quantum processors.

7. Outlook and Open Questions

Ongoing directions include deeper engineering of the spin–orbit field via composition, strain, and multi-gate designs to optimize $\sim p^3$4 and Rabi rates, further suppression and control of charge noise (including operation at "sweet spots"), and integration with high-Q microwave photonic resonators for spin–photon interfaces and long-range two-qubit gates (Sagaseta et al., 6 Oct 2025). The interplay of symmetry, strain, and dot geometry enables novel, electrically switchable spin states and potentially ultrafast, spatially non-local entangling operations (Planelles et al., 2015, Bosco et al., 2022). Direct-gap GeSn holes provide a platform both for coherent spin–photon coupling and hybrid quantum memories (Rotaru et al., 24 Feb 2025). The approach to the fault-tolerance threshold for fidelity, coupled with site-controlled arrays and robust optical/electrical initialization, now positions hole-spin quantum dots for systematic scaling in large quantum architectures.

References: (Lawrie et al., 2020, Rotaru et al., 24 Feb 2025, Liles et al., 2018, Hendrickx et al., 2019, Dahbashi et al., 2011, Ares et al., 2013, Fernandez-Fernandez et al., 2022, Chekhovich et al., 2010, Terrazos et al., 2018, Fischer et al., 2010, Godden et al., 2011, Pribiag et al., 2013, Bosco et al., 2022, Barcan et al., 7 Mar 2025, Greilich et al., 2011, Planelles et al., 2015, Liles et al., 2023, Economou et al., 2012, Hung et al., 2016, Sagaseta et al., 6 Oct 2025).