Individual two-axis control of three singlet-triplet qubits in a micromagnet integrated quantum dot array

Abstract

We report individual confinement and two-axis qubit operations of two electron spin qubits in GaAs gate-defined sextuple quantum dot array with integrated micro-magnet. As a first step toward multiple qubit operations, we demonstrate coherent manipulations of three singlet-triplet qubits showing underdamped Larmor and Ramsey oscillations in all double dot sites. We provide an accurate measure of site-dependent field gradients and rms electric and magnetic noise, and we discuss the adequacy of simple rectangular micro-magnet for practical use in multiple quantum dot arrays. We also discuss current limitations and possible strategies for realizing simultaneous multi-qubit operations in extended linear arrays.Fabrication of large array of qubits and demonstration of coherent multi-qubit operations are necessary steps for realizing scalable quantum processing unit 1-4 . Harnessing spin degree of freedom in gate defined quantum dots (QDs) has attracted significant interests [5][6][7][8][9][10][11][12][13][14] owing to spin's long coherence time in solid state system and potential scalability including well-established fabrication technology 2,15,16 . Depending on the degree of spin-charge mixing, elementary quantum operations of Loss-DiVincenzo (single electron) 12,14,17,18 , singlet-triplet (ST0, two electron) 5,10,11,[19][20][21] , and various three electron spin qubits [22][23][24] have been demonstrated. Electron loading and charge state manipulation are also performed in ~ 10 coupled linear dots 3,25 and ~ 4 site two dimensional arrays 1,4 . Moreover, QD Hamiltonian parameters have shown to be widely tunable 19,[26][27][28] . With possibility of efficient electrical control, QD systems are also emerging as promising quantum simulators 26,29,30 .Toward realizing multi-qubit operations in semiconductor quantum chips, coherent qubit addressing beyond ground state property tuning is important. In this work, we demonstrate individual operations of three ST0 qubits formed in different double dot sites in a sextuple linear quantum dot array with rf-single-electron transistor (rf-set) sensors. Following pioneering works such as recent demonstration of four individual Loss-DiVincenzo qubit operations in GaAs 31 , we employed proximal rectangular micro-magnet enabling sizable magnetic field gradient 32 and two-axis control for all ST0 qubits in our device. This work addresses spatial range of experimentally usable field gradient by simple micro-magnet using qubit oscillations as sensitive magnetic probes, and we show that magnitude of electric and magnetic noise of all qubits are comparable to the previous works. Moreover, we discuss examples of adverse effects of surface micro-magnet structure fabricated on the device without interposing dielectrics, which calls for further optimization of sensor and magnet position to enable simultaneous qubit and sensor operations in similar device geometries.

Deep reinforcement learning for universal quantum state preparation via dynamic pulse control

“…Future quantum computers promise exponential speed-ups over their classical counterparts in solving certain problems like search and simulation [1]. A wide variety of promising modalities emerges in the race to realize the quantum computer, such as trapped ions [2,3], photonic system [4][5][6][7], nitrogen-vacancy centers [8], nuclear magnetic resonance [9], superconducting circuits [10,11] and semiconductor quantum dots [12][13][14][15][16][17][18]. Among these the semiconductor quantum dots is a powerful competitor for potential scalability, integrability with existing classical electronics and well-established fabrication technology.…”

Modularized and Scalable Compilation for Double Quantum Dot Quatum Computing

“…In the present work, the model of interest is the semiconductor DQDs system, where the S-T 0 qubit is encoded in the collective spin states of two electrons confined in a double-well potential [14,42]. The Hamiltonian of a single S-T 0 qubit, governed by external electric pulses is [8,9,43,44]…”

“…σ z and σ x are Pauli matrices and represent rotations of the quantum state around the z-and x-axes with rates J(t) and h, respectively. The coefficient h describes the Zeeman energy splitting between |S and |T 0 , commonly raised by nearby deposited permanent micromagnet [43] and its value is difficult to vary during the quantum gate time in experiment [42,43] (although tunable splitting by Overhauser field has been observed [14], its time consumption is much longer than a typical gate time). We assume h = 1 here to facilitate our theoretical treatment, and take this as the energy unit.…”

Modularized and scalable compilation for double quantum dot quantum computing

“…In the present work, the model of interest is the semiconductor DQDs system, where the S-T 0 qubit is encoded in the collective spin states of two electrons confined in a double-well potential [14,42]. The Hamiltonian of a single S-T 0 qubit, governed by external electric pulses is [8,9,43,44]…”

“…σ z and σ x are Pauli matrices and represent rotations of the quantum state around the z-and x-axes with rates J(t) and h, respectively. The coefficient h describes the Zeeman energy splitting between |S⟩ and |T 0 ⟩, commonly raised by nearby deposited permanent micromagnet [43] and its value is difficult to vary during the quantum gate time in experiment [42,43] (although tunable splitting by Overhauser field has been observed [14], its time consumption is much longer than a typical gate time). We assume h = 1 here to facilitate our theoretical treatment, and take this as the energy unit.…”