Qubit syndrome measurements with a high fidelity Rb-Cs Rydberg gate

Abstract

We demonstrate an inter-species entangling Rydberg gate between rubidium (Rb) and cesium (Cs) atoms with fidelity $\mathcal F = 0.975\pm 0.002$. The two-species atom array enables in-place quantum non-demolition (QND) qubit measurements which are a key capability for quantum error correction. We demonstrate this functionality with multi-atom error syndrome measurements achieving QND measurement fidelities of ${\mathcal F}_{\rm QND} = 0.933(12)$ and 0.865(17) for two- and three-qubit plaquettes, respectively.

Figures (14)

• Figure 1: a) Dual species atom array. $^{87}$Rb and $^{133}$Cs atoms are prepared in a checkerboard pattern. Individual Rydberg state addressing beams with wavelengths 421, 1005, 459, and 1040 nm are used to excite neighboring Rb-Cs pairs to Rydberg states (top inset). Gate model circuits for performing two- and three- qubit QND measurements (middle and bottom insets). b) Illustration of independent species readout via two-sided imaging. Scattered photons for Rb and Cs are captured via independent optical paths and are imaged on a shared EMCCD camera. A set of dichroic mirrors and narrow-line filters are used to perform wavelength discrimination. c) Readout level diagrams for each species.
• Figure 2: Characterization of single qubit operations. a) Randomized benchmarking for one Rb qubit, yielding $\mathcal{F}=0.99963(5)$, measured simultaneously with Cs. b) Randomized benchmarking for one Cs qubit, yielding $\mathcal{F}=0.99962(5)$, measured simultaneously with Rb. c) Ramsey-Stark oscillations for Rb demonstrating ${\sf R_z}(\phi)$ oscillations at $f= 1.175(2)$ MHz and $f\tau = 18(3)$. See Miles2026SM for details. d) Same as c), but for Cs, at $f= 0.966(2)$ MHz and $f\tau = 34(15)$.
• Figure 3: a) Rubidium ground-Rydberg Rabi oscillations with rate $\Omega=2\pi\times1.2085(17)$ MHz. b) Cesium ground-Rydberg Rabi oscillations at $\Omega=2\pi\times1.2219(15)$ MHz. c) Rydberg spectroscopy of cesium with a neighboring rubidium atom in the ground state. d) Rydberg spectroscopy of Cs with a neighboring Rb atom in the Rydberg state $\ket{63s_{1/2}, m_j=-1/2}$. The shift in the Cs Rydberg resonance is due to the Rb-Cs interaction.
• Figure 4: Randomized benchmarking of Rb-Cs $\sf CZ$ gates. Both fits are to the functional form $P_{\rm measure}=AP^n$, where $P_{\rm measure}$ is the measured probability of both atoms being present giving $P_{\rm ret}=P$ or both atoms being in the bright state after blowaway of the upper hyperfine level giving $P_{\rm bb}=P.$ The probability without any $\sf CZ$ gates due to state preparation and measurement errors is $A$, and $n$ is the number of $\sf CZ$ gates.
• Figure 5: QND measurement results. a) Two-atom QND with target qubit as Rb. A state of $\ket{01}$ denotes Cs in $\ket{0}$ and Rb in $\ket{1}$. b) Target qubit as Cs. c) Three-atom QND with target qubit as Rb. The kets are labeled $\ket{{\rm RbCsCs}}$.