Single-Operation Rydberg Phase Gates via Dynamic Population Suppression

TL;DR

The paper tackles fast, high-fidelity two-qubit gates in neutral-atom quantum processors by leveraging a dynamic population suppression strategy that preserves Rydberg interactions as a phase resource. By using amplitude-modulated, zero-area pulses across two overlapping fields, Rydberg excitation is coherently canceled while an interaction-dependent phase accumulates, enabling single-operation gates with no finite-blockade error even when the Rabi frequency is comparable to the interaction energy. The authors derive and validate that singly excited manifolds contribute a phase α largely independent of the blockade, while the |11> manifold contributes a phase β that depends on V, with the entangling condition φ = 2α − β = π (mod 2π) yielding a perfectly entangling controlled-Z gate; analytical and numerical analyses reveal broad regions in parameter space supporting high-fidelity gates and robust performance against typical experimental fluctuations. The scheme achieves gate durations in the nanosecond regime and infidelities on the order of 10^{-3} under realistic noise, making it compatible with current neutral-atom platforms and extendable to metrology and scalable quantum information processing, including spin-squeezed states and enhanced sensing.

Abstract

We propose a versatile control protocol based on modulated zero-pulse-area fields that dynamically suppresses Rydberg excitation while retaining Rydberg-Rydberg interactions as an entangling phase resource. This mechanism enables single-step, perfectly entangling phase gates for arbitrary blockade strengths, eliminating finite-blockade errors even when the Rabi frequency approaches or exceeds the interaction energy. The approach defines a new operational regime for Rydberg-blockade quantum logic in which speed, fidelity, and robustness are achieved simultaneously within a simple dynamical framework. Owing to its simplicity and generality, the technique is compatible with a wide range of neutral-atom architectures and offers a promising route toward scalable, high-fidelity quantum computation and simulation.

Figures (6)

• Figure 1: (a) Energy level diagram for two identical qubits coupled via Rydberg-Rydberg interaction. (b) Level diagram of the relevant subspace for the dynamics of the state $\ket{1 1}$.
• Figure 2: Phase shifts produced by DE pulses and gate performance metrics. (a) Accumulated phase $\alpha$ for states $\ket{01}$ and $\ket{10}$ versus peak Rabi frequency $\Omega_0$ and modulation frequency $\omega_e$ for a blockade value $V = 50/t_p$. (b) Accumulated phase $\beta$ for state $\ket{11}$ versus peak Rabi frequency $\Omega_0$ and modulation frequency $\omega_e$ for a blockade value $V = 50/t_p$. (c) Fidelity of a controlled-Z gate and (d) entangling power of the gate normalized to the maximum possible value, both as functions of modulation frequency $\omega_e$ and the ratio of peak Rabi frequency $\Omega_0$ to blockade value $V$.
• Figure 3: (a) Population of state $\ket{10}$ (or $\ket{01}$) after applying the gate as a function of modulation frequency $\omega_e$ and Rabi frequency $\Omega_0$ for $V = 50/t_p$. (b) Population of state $\ket{11}$ after applying the gate.
• Figure 4: Fidelity landscape and optimal driving parameters for dynamically suppressed Rydberg phase gates. (a) Optimal peak Rabi frequency $\Omega_0$ (in units of $t_p^{-1}$) as a function of the modulation frequency $\omega_e$, obtained by imposing the condition $\alpha = -2\pi$ for the singly excited states $\ket{01}$ and $\ket{10}$. (b) Exact gate fidelity as a function of $\omega_e$ and the Rydberg–Rydberg interaction $V$, using the optimal $\Omega_0$ from (a). The dashed line indicates the locus predicted analytically from the condition $\beta = -3\pi$, derived in Appendix \ref{['Appendix:C']}, showing excellent agreement with the numerical results. (c) Final population of $\ket{11}$ after the gate operation for the same parameter set. The dashed line marks the region where population return is not guaranteed, consistent with the analytic analysis of Appendix \ref{['Appendix:B']}. (d–f) Same as (a–c), but for the higher-order condition $|\alpha| = 10\pi$. Increasing $|\alpha|$ yields additional high-fidelity branches, including solutions at negative $\omega_e$, and broadens the parameter regimes that support perfectly entangling gates. The final-state population in (f) again shows full population return except in narrow nonadiabatic regions near $V = 2\omega_e$ and at very small $|\omega_e|$.
• Figure 5: State dynamics under a controlled-Z gate. (a) Accumulated phase $\alpha$ for state $\ket{01}$. (b) Corresponding population dynamics. (c) Accumulated phase $\beta$ for state $\ket{11}$. (d) Corresponding population dynamics. When plotting populations, states that are not populated are omitted.