Semiconductor Qubits In Practice

Anasua Chatterjee; Paul Stevenson; Silvano De Franceschi; Andrea Morello; Nathalie de Leon; Ferdinand Kuemmeth

Source-linked AI summary

Semiconductor Qubits In Practice

Anasua Chatterjee, Paul Stevenson, Silvano De Franceschi, Andrea Morello, Nathalie de Leon, Ferdinand Kuemmeth

arXiv:2005.06564v1May 13, 2020cond-mat.mes-hallquant-ph

TL;DR

Semiconductor qubit implementations must be assessed in relation to the applications and hardware properties that shape their usefulness. This Review surveys charge and spin qubits, relates their strengths and prospects to quantum applications, and identifies fabrication and scaling challenges.

Problem
Different quantum applications depend strongly on the specific properties of their underlying hardware, motivating a broader assessment than suitability for universal quantum computing alone.
Method
The Review summarizes the state of the art in semiconductor charge and spin qubits, emphasizing gate-controlled quantum dots, dopants, and color centers and relating them to practical applications.
Results
The Review maps semiconductor qubit categories to their respective strengths and prospects for quantum simulation, computation, sensing, and communication.
Takeaways & Limitations
Semiconductor qubits offer application-relevant platforms spanning charge and spin implementations, including quantum dots, shallow dopants, and color centers.
Takeaways & Limitations
Scalable and reproducible fabrication, large-array calibration and operation, and cross-talk and placement remain significant challenges for semiconductor quantum circuits.

Abstract

from arXiv · show

In recent years semiconducting qubits have undergone a remarkable evolution, making great strides in overcoming decoherence as well as in prospects for scalability, and have become one of the leading contenders for the development of large-scale quantum circuits. In this Review we describe the current state of the art in semiconductor charge and spin qubits based on gate-controlled semiconductor quantum dots, shallow dopants, and color centers in wide band gap materials. We frame the relative strengths of the different semiconductor qubit implementations in the context of quantum simulations, computing, sensing and networks. By highlighting the status and future perspectives of the basic types of semiconductor qubits, this Review aims to serve as a technical introduction for non-specialists as well as a forward-looking reference for scientists intending to work in this field.

Introduction

Semiconductor qubits encompass diverse materials, physical systems, and operating conditions, motivating a review organized around their application-specific strengths. The Review surveys charge and spin implementations and benchmarks them across sensing, simulation, computation, and communication.

Introduction: Quantum applications depend strongly on hardware properties, so the Review considers sensing, simulation, computation, and communication beyond universal-computing potential.
Introduction: Semiconductor qubits range from engineered potential wells to single-atom impurities and can operate from millikelvin to room-temperature conditions.
Introduction: The Review focuses on charge and spin qubits in gate-controlled quantum dots, dopants, and color centers, while excluding several other implementations.
Introduction: The Review summarizes the field’s current state and relates implementation categories to their strengths and prospects for practical applications.
Introduction: Each qubit category is benchmarked against four applications using criteria specific to sensing, simulation, computing, and communication.

Charge in gate-controlled structures

Gate-controlled quantum dots offer versatile control of charge states and spatially engineered Hamiltonians, but charge coherence is vulnerable to electrical and environmental noise. Consequently, charge qubits are especially valuable for sensing, initialization, readout, and resonator-mediated coupling.

Charge encoding and control: Charge qubits encode information in an electron’s location between adjacent quantum dots, with |L⟩ and |R⟩ as the basis states.
Charge qubit operation and coherence: Charge oscillations and two-axis control have been demonstrated in double quantum dots, including silicon devices exploiting valley-state superpositions.
Charge qubit operation and coherence: 0.4 µs coherence time was reported for a silicon strained Si/SiGe double quantum dot, with similarly long coherence obtained in a GaAs/AlGaAs device.
Limitations and applications: Charge degrees of freedom enable sensing, spin-to-charge readout, and coherent coupling through superconducting resonators over millimeter-long on-chip distances.
Device architecture: Gate-patterned quantum dots support spatial engineering of specific Hamiltonians, and increasing gate complexity enables tunnel-coupled multiple-dot systems.
Limitations and applications: Electrical and environmental charge noise limit charge coherence, favoring charge qubits as readout or initialization tools rather than long-coherence computing qubits.

Spin in gate-controlled structures

Gate-controlled spin qubits use diverse encodings and electrical or magnetic control, with spin-to-charge conversion providing the primary readout route. Their coherence, gate performance, and material-dependent trade-offs support quantum-computing applications while leaving fabrication and noise challenges.

Performance: Tens of microseconds, occasionally up to 120 µs, are routinely exceeded by spin-qubit coherence times, longer than manipulation times.This coherence advantage supports coherent control of gate-defined spin qubits.
Qubit encodings: Gate-controlled spin qubits use single-spin, singlet-triplet, exchange-only, hybrid, and multi-dot encodings, each balancing control, coherence, complexity, and noise sensitivity.Singlet-triplet qubits can use decoherence-free operation, while exchange-only and hybrid designs trade additional complexity or charge sensitivity for control advantages.
Qubit encodings: 100 ps π-rotations have been demonstrated for an all-electrical charge-spin hybrid qubit.The hybrid encoding combines charge-like speed with spin-like increased coherence.
Initialization and readout: Spin-to-charge conversion is the primary readout method because charge displacement is easier to detect than an individual spin’s magnetic moment.For singlet-triplet qubits, conditional tunneling distinguishes singlets from triplets through Pauli spin blockade.
Control: Electrical control uses exchange pulses or electric-dipole spin resonance, while magnetic control uses resonant microwave pulses matched to the qubit Larmor frequency.EDSR obtains an effective time-varying magnetic field from wavefunction motion through natural or synthetic spin-orbit fields.
Materials and applications: Silicon spin qubits combine coherence times up to 28 ms with foundry compatibility, but valley-related leakage and fabrication variability constrain scalability.Valley splitting is affected by fabrication defects, inhomogeneities, and nanowire, interface, or heterostructure step edges.

Dopants in silicon

Shallow dopants provide atom-like semiconductor qubits with long-lived spin states, while silicon isotopic purification substantially extends electron coherence. Their readout can use single-electron-transistor charge sensing.

Dopant physics: Shallow group-V donors in group-IV materials are solid-state analogues of hydrogen atoms, with donor electrons coupled to spin-1/2 nuclei.For phosphorus in silicon, the electron occupies a weakly bound 1s-like orbital with an envelope Bohr radius of about 2 nm.
Dopant physics: Donor-spin properties reveal effects of conduction-band valley degeneracy, spin-orbit coupling, and valley-orbit coupling in silicon.These properties made dopant quantum states important benchmarks for semiconductor band-structure theories.
Readout: Single-electron-transistor charge sensors read out shallow-dopant qubits, alongside optical and photocurrent methods for color centers.Figure 3 distinguishes proximal and gate-integrated sensors from the readout mechanisms used by different qubit platforms.
Coherence: Electron coherence reached approximately 60 ms with 800 ppm residual 29Si and was extended to 10 seconds using dipole-dipole suppression and 50 ppm 29Si.These results demonstrate the impact of isotopic enrichment on donor-spin coherence.

Single-donor 31P spin qubits

Single-donor silicon qubits combine exceptionally long coherence with high-fidelity control, while their scalability depends on engineering practical inter-qubit couplings and readout. Donor platforms also support sensitive magnetic-field detection, simulation, and emerging optical interfaces.

Coherence and control: Isotopically enriched 28Si extends single-donor coherence times to T2e = 0.56 s and T2n = 35.6 s using dynamical decoupling.These record coherence times were reported for electron and nuclear spins, respectively.
Coherence and control: Randomized benchmarking measured 99.94% electron and 99.98% nuclear 1-qubit Clifford gate fidelities in enriched silicon donor qubits.Electron-nuclear entanglement was also demonstrated through a Bell-inequality violation with Bell signal S = 2.7097.
Sensing: A single 31P donor achieved a noise floor equivalent to 18 pT/√Hz, demonstrating strong magnetic-field sensitivity enabled by high spin coherence.Donor sensing remains constrained by low-temperature operation and integration with charge-readout devices.
Simulation and interfaces: Atomically precise dopant placement supports Hubbard-model simulations and single-site measurement by scanning tunneling spectroscopy, while deeper or rare-earth dopants offer optical-interface routes.Group-VI donors such as 77Se and erbium are discussed as more promising systems for spin-optical interconnects than shallow donors.
Scalability: The short range of exchange interaction complicates large-scale donor layouts, motivating interposer quantum dots, dopant spin chains, ferromagnetic couplers, and electric-dipole encodings.Alternative coupling proposals include a 4 ns SWAP between acceptor spins at 20 nm and a 40 ns SWAP between flip-flop qubits at 200 nm.

Optically-Addressable Quantum Defects

Optically addressable quantum defects combine spin control with optical initialization and readout, supporting sensing, quantum simulation, computation, and communication while retaining important material and scalability challenges.

Initialization, Manipulation and Readout: Optically addressable defects couple a spin degree of freedom to optical transitions, combining spin-resonance control with individual optical addressing.This spin-photon interface supports manipulation of the spin and interactions with its environment.
Initialization, Manipulation and Readout: Microwave pulses commonly provide coherent spin control, while optical pumping initializes defects into a defined ground state.Room-temperature initialization can use off-resonant excitation and spin-dependent intersystem crossings, although fidelity is limited.
Initialization, Manipulation and Readout: Optical readout enables spin-photon interfaces and, at low temperatures, single-shot quantum nondemolition measurements with fidelities reaching 99.7%.Photon extraction from high-refractive-index materials remains a challenge addressed through micromachining and photonic structures.
Sensing: Color centers provide magnetic and thermal sensing capabilities, including 500 nT/√Hz DC magnetometry, 4.3 nT/√Hz AC magnetometry, and thermometry sensitivities down to 100 mK/√Hz.These systems also support nanoscale imaging when combined with scanning probe microscopes.
Challenges and Applications: Near-surface sensing requires close defect-target proximity, but surface charge traps and magnetic noise can degrade stability and coherence; controlled diamond surfaces have achieved coherence times exceeding 100 µs.Defect formation randomness and limited spatial control also complicate arbitrary Hamiltonian construction for quantum simulation.
Simulation, Computation and Communication: Quantum defects support disordered-system experiments and small quantum registers, but their probabilistic formation and limited suitability for large-scale computation constrain broader architectures.NV centers have enabled registers with tens of qubits through hyperfine-mediated nuclear-spin manipulation.
Quantum Communication: Defect-based quantum communication seeks indistinguishable photons with low propagation loss, motivating nanophotonic structures, frequency conversion, and alternative defects with reduced spectral diffusion.SiV− offers demonstrated cavity-mediated interactions and spin-photon interfaces, but limited coherence and millikelvin operation remain challenges.

Outlook

Semiconductor qubits offer a broad ecosystem of applications and device architectures, while their future depends on scalable fabrication, reliable operation, and integration with industry manufacturing.

Outlook: Semiconductor qubit platforms support diverse applications, including computation, sensing, simulation, communication, light-matter networks, quantum memories, and designer simulation arrays.The review emphasizes complementary advantages and trade-offs across qubit encodings.
Outlook: Strong charge-photon and spin-photon coupling underpins proposed light-matter networks, while optically active qubits could support distant entanglement and secure quantum communication.Quantum-dot spin arrays have demonstrated proof-of-principle condensed-matter simulations, and silicon dopants are candidates for quantum memories and dense arrays.
Challenges: Scalable and reproducible fabrication, calibration, stabilization, cross-talk control, and classical readout placement remain major challenges for semiconductor quantum circuits.The small physical dimensions create unique engineering problems despite the prospect of integrating hundreds of millions of qubits on a chip.
Conclusion: Greater semiconductor-industry engagement is expected to improve qubit density and device reliability while expanding engineered quantum systems for useful devices and fundamental science.The conclusion presents diversity and flexibility as reasons for pursuing complementary research directions.