Lattice Defects in Rydberg Atom Arrays

TL;DR

This work addresses how geometric defects in a 1D Rydberg-atom array influence universal Ising criticality under nearest-neighbor blockade. The authors combine a defect-augmented continuum Ising CFT with DMRG validation to show that site-centered kinks act as cutting defects (Delta_phi = 1/8 governing one-point observables), while bond-centered kinks drive an exotic defect RG flow toward an intermediate fixed point; away from criticality, bond-centered kinks can also cause a localization-delocalization (quantum wetting) transition of a domain wall. The results reveal defect-type dependent destinies and demonstrate that one-point correlators can capture essential two-point physics in the presence of defects. Overall, defects emerge as powerful probes of critical behavior in Rydberg arrays and underscore the importance of defect control for faithful realization of pristine Ising physics in experiments.

Abstract

Rydberg atom arrays have become a key platform for studying quantum many-body systems. In these setups, defects arise naturally due to various imperfections and can significantly modify the theoretical predictions compared to an ideal model. Here, we investigate the impact of geometric defects in the simplest situation -- a one-dimensional Rydberg atom array, both at and away from its emergent Ising criticality. In the presence of defects, we demonstrate that relevant physical quantities can be extracted from one-point correlation functions. At the critical point, we show that different types of kinks yield distinct outcomes corresponding to their respective spatial-internal symmetries: site-centered kinks can effectively break the array at the kink position regardless of the kink angle, while bond-centered kinks lead to interesting intermediate-coupling fixed points. In the latter case, due to a special renormalization group flow trajectory, the whole system can appear ordered if the system is not large enough. Additionally, away from criticality, the bond-centered kink induces a localization-delocalization transition of the domain wall, characteristic of quantum wetting. These findings highlight the utility of kinks as experimental probes and stress the importance of controlling defects so that experimental observations remain faithful to the pristine model.

Figures (10)

• Figure 1: (a) Site-centered kink deformation (Eq. \ref{['eq:site_kink']}) increases the interaction strength between its two neighbors, effectively breaking the chain at the $\mathbb{Z}_2$ critical point. (b) Bond-centered kink deformations (Eq. \ref{['eq:bond-center_kink']}) preserves an extra symmetry compared to (a), driving the system to intermediate-coupling fixed points at $\mathbb{Z}_2$ criticality. Additionally, a localization-delocalization transition emerges in the ordered phase.
• Figure 2: Plot of $(-1)^i\delta Z_i = (-1)^i\langle \hat{Z}_i - \hat{Z}_{i+1} \rangle/2$ vs sites $i$ with site-centered kink at angle $\theta=60^\circ$, for (a) even $N$ and (b) odd $N$ cases (insets). The main figures depict data collapsed onto a universal curve by setting $N^{\Delta_\phi} (-1)^i\delta Z_i$ vs $x/L$.
• Figure 3: (a) Log-log plot of $|\delta Z_i|$ vs $N\sin(\pi x)$, for $\theta=60^\circ$ and $\theta=120^\circ$ (fader colors). (b) Fitted scaling dimension $\Delta_{\phi}^{\text{fit}}$ with increasing system size $N$ for site-centered kink angles $\theta=60^\circ, 90^\circ, 100^\circ, 120^\circ$. Crosses on each curve label the estimated limiting system size that begins to obey CFT prediction, set by $g = 60$.
• Figure 4: Plot of $\delta Z_i$ vs even sites $i$ with kink angle $\theta=140^\circ$ positioned at $k_\text{b}=0$, for (a) even $N$ and (b) odd $N$ cases. The main figures depict data collapsed onto a curve by setting $N^{2\Delta_\phi} \cdot (-1)^i \delta Z$ vs $x$, where $\Delta_\phi=1/8$.
• Figure 5: (a) For $\theta=60^\circ$, with increasing system size up to $N\simeq400$, $(-1)^i \delta Z_i$ is still increasing. (b) 1-loop perturbtive RG. The color bar refers to the flow speed.