Simulated spin qubits in silicon quantum dots and enhancement of InGaAs photodetectors

Abstract

Semiconductor quantum dot spin qubits are a leading candidate for scalable, fault- tolerant quantum computing. Their advantages include nanoscopic device size, compat- ibility with foundry fabrication processes, and long coherence times relative to gate du- rations. The fabrication and control of a quantum processing unit composed of tens of thousands to millions of physical qubits pose many engineering challenges. These chal- lenges fall broadly into two categories: device design, such as optimizing the geometry for high-quality qubit formation, and qubit control, which involves the precise manipulation of spin or charge states in qubits that are capacitively coupled to numerous neighboring electrodes. In this thesis, we develop a simulation tool that accelerates device design iter- ation prior to fabrication by providing a priori knowledge of the quantum dot electrostatic potential landscape as a function of external electrode voltages. This enables effective spin and Hubbard Hamiltonian parameters to be computed before experimental charac- terization, facilitating early-stage control method development and device performance prediction. The tool, implemented as the Python-based QuDiPy package, integrates three- dimensional finite-element Poisson solutions with modules for electrostatic reconstruction, Hamiltonian parameter extraction, and control pulse optimization. Unlike previous dis- jointed toolchains, QuDiPy offers a unified workflow for full-stack qubit control simulations, including automated voltage-to-Hamiltonian mapping for exploring high-dimensional gate voltage spaces and mitigating crosstalk in dense qubit arrays. The simulator is designed to be memory- and CPU-efficient to enable computationally efficient simulation of linear quantum dot arrays consisting of several qubits. Simulation of small quantum dot arrays serves as a design tool for control protocols within multi-node quantum processors. Sim- ulation of spin qubit dynamics in many-qubit nodes connected in a network enables the study of required voltage ranges for maintaining stable charge configurations in the device. It also supports the design of experimental input pulses to generate maximally entangled Greenberger–Horne–Zeilinger (GHZ) states between nodes, a key step for implementing surface code error correction protocols.

Spin qubit control requires a precise understanding of the impact of experimental con- trols, such as electrode voltages or radio-frequency magnetic field amplitude and phase, on effective parameters such as electronic g-factor, exchange energy, chemical potential, etc. A mapping between experimental and effective parameters is created by performing effective parameter calculations on two-dimensional cross-sections of the electrostatic po- tential landscape obtained from a 3-dimensional Poisson solver nextnano++, a commercially available, 3D Poisson solver chosen for its robustness, flexibility in defining quantum device geometries, and proven accuracy in modeling semiconductor heterostructures at cryogenic temperatures. First, a 2D cross-section of the electrostatic potential landscape is taken along the growth direction of the quantum dot device, near the heterojunction where qubit formation occurs. This region is selected because it captures the horizontal confinement profile most relevant to charge localization and wavefunction shape. The cross-section is extracted for all simulated voltage configurations applied to the gate electrodes. Second, the single-particle ground state or first excited state wavefunctions are determined using a non-uniform grid Schrödinger solver for all voltage configurations and for each isolated quantum dot or nearest-neighbor quantum dot pair. The non-uniform grid provides higher spatial resolution near confinement potential minima, enabling more accurate modeling of localized wavefunctions where precision is most critical. The mapping between input voltage and single-particle wavefunctions is leveraged, along with numerical integration routines, to calculate the desired effective parameters as a function of voltage. The chemi- cal potential, tunnel coupling, and onsite and interdot Coulomb parameters are computed for each voltage configuration. This enables exact diagonalization of the Hubbard Hamil- tonian at every point in voltage space and identifies the regions of charge stability for a multiqubit quantum dot device. This step is essential for establishing control over the quantum processor.

The second part of this thesis investigates optoelectronic device enhancement using localized surface plasmons in nanocrystals. Fast and accurate detection of light in the near-infrared (NIR) spectral range plays a crucial role in alleviating speed and capacity bottlenecks in optical communications and in enhancing the control and safety of au- tonomous vehicles through NIR imaging systems. Several technological platforms are cur- rently under investigation to improve NIR photodetection, aiming to surpass the perfor- mance of established III–V semiconductor p-i-n (PIN) junction technology. These plat- forms include in situ-grown inorganic nanocrystals (NCs) and nanowire arrays, as well as hybrid organic–inorganic materials such as graphene-perovskite heterostructures. How- ever, challenges remain in NC and nanowire growth, large-area fabrication of high-quality 2D materials, and the fabrication of devices for practical applications. Here, we ex- plore the potential for tailored semiconductor NCs to enhance the responsivity of planar metal–semiconductor–metal (MSM) photodetectors. MSM technology offers ease of fabri- cation and fast response times compared to PIN detectors. We observe enhancement of the optical-to-electric conversion efficiency by up to a factor of ∼2.5 through the application of plasmonically-active semiconductor nanorods and NCs. We present a protocol for syn- thesizing and rapidly testing the performance of non-stoichiometric tungsten oxide (WO) nanorods and cesium-doped tungsten oxide (CsyWO) hexagonal nanoprisms prepared in colloidal suspensions and drop-cast onto photodetector surfaces. The results demonstrate the potential for a cost-effective and scalable method exploiting tailored NCs to improve the performance of NIR optoelectronic devices.