Experimental observation of dynamical blockade between transmon qubits via ZZ interaction engineering

TL;DR

This work demonstrates that strong, tunable longitudinal (ZZ) couplings between capacitively connected transmon qubits can be engineered with purely capacitive means, achieving $\zeta$ values from $\sim{10\ MHz}$ up to $\sim{350\ MHz}$ as qubits approach resonance. The authors validate the mechanism with two separate devices and connect the observed ZZ strength to a microscopic picture via perturbation theory and black-box quantization, then demonstrate a dynamical blockade where one excitation inhibits the neighbor’s excitation. They further develop a scalable design cycle using Foster synthesis and differential evolution to maximize ZZ interactions, enabling interaction-dominated dynamics and potential globally controlled quantum architectures in superconducting circuits. The results pave the way for blockade-enabled quantum simulators and cooperative many-body dynamics in solid-state qubit platforms, while addressing hardware scalability by reducing control wiring needs. The work integrates spectroscopic, time-domain, and advanced circuit quantization methods to provide a coherent, predictive framework for strong ZZ coupling and blockade phenomena in superconducting qubits.

Abstract

We report the experimental realization of strong longitudinal (ZZ) coupling between two superconducting transmon qubits achieved solely through capacitive engineering. By systematically varying the qubit frequency detuning, we measure cross-Kerr inter-qubit interaction strengths ranging from 10 MHz up to 350 MHz, more than an order of magnitude larger than previously observed in similar capacitively coupled systems. In this configuration, the qubits enter a strong-interaction regime in which the excitation of one qubit inhibits that of its neighbor, demonstrating a dynamical blockade mediated entirely by the engineered ZZ coupling. Circuit quantization simulations accurately reproduce the experimental results, while perturbative models confirm the theoretical origin of the energy shift as a hybridization between the computational states and higher-excitation manifolds. We establish a robust and scalable method to access interaction-dominated physics in superconducting circuits, providing a pathway towards solid-state implementations of globally controlled quantum architectures and cooperative many-body dynamics.

Figures (17)

• Figure 1: (a) Device under test, composed of two transmon-type qubits (Q1 and Q2) with mutual capacitive coupling. (b) Effect of the ZZ interaction. When Q1 and Q2 are in the ground state, they resonate at $\omega_{1,2}$ respectively (full lines). The excitation of one qubit (Q2) leads to a shift of the energy levels of the adjacent qubit (Q1) by $\zeta$ (dashed lines).
• Figure 2: (a) Q1 spectroscopy as a function of $\Delta / (2 \pi)$ when Q2 is either in the ground or excited state. The orange and red markers show the extracted Q1 frequency from the spectroscopy with a Lorentzian fit when Q2 is in $\ket{0}$ or $\ket{1}$ respectively. $\zeta$ is computed as the difference between these two frequencies. (b) Measured ZZ interaction as a function of $\Delta / (2 \pi)$. The black circles mark the ZZ extracted from the spectroscopy measurements shown in panel (a), while the red triangles are extracted from Ramsey measurements on Q1 conditional on the state of Q2. Additionally, we show in blue the ZZ extracted from the circuit Hamiltonian. The kink in the theoretical curve around $\Delta / (2 \pi) = 1.4$ GHz is caused by the readout resonator of Q2 located at that frequency.
• Figure 3: (a) Measured (points) and predicted (full lines) populations of Q1 versus relative delay for six different excitation pulse lengths ($16$ ns - $200$ ns). When long pulses are applied, exciting Q2 first (positive delay) inhibits the excitation of Q1, demonstrating blockade due to the qubit cross-Kerr coupling ($\zeta/(2\pi) = 19~\text{MHz}$). Top inset: pulse sequences for Q1 and Q2, illustrating positive and negative excitation delays. (b) The red points show the evolution of $P_1(e)$ as a function of the excitation pulse length for a delay of 100 ns. The blue points show, for each pulse length, the relative power of the drive pulse integrated over a spectral window centered around $(\omega_1 - \zeta) / (2 \pi)$ and with a width given by $1 / T_2^{Q1}$.
• Figure S1: The measurement setup. The sample is marked with green for clarity. The arrows near each FEM marks the direction of the signal flow.
• Figure S2: (a) Q1 population as a function of the delay between the excitation pulses of different lengths. The black dashed lines mark the fit to the yellow data ($\pi$-pulse length of 200 ns), from which the decay times are extracted. For the same pulse length, the other panels show the readout histograms of the ground (light gray) and excited (light red) state in the rotated IQ plane at selected delays: -75 $\mu$s (b), 100 ns (d) and 75 $\mu$s (c). The insets show the readout fidelity matrix.