Electrical Noise Produced by Micron-Sized Particles above a Surface Paul Trap

Technical Summary: Electrical Noise Produced by Micron-Sized Particles above a Surface Paul Trap

Problem Statement
Electric field noise generated by the surfaces of ion trap electrodes is a primary limitation on the fidelity of quantum computing operations. While this noise causes motional heating and subsequent decoherence, its microscopic origins remain unclear despite decades of investigation. A significant challenge in the field is the large variation in reported noise levels across different experiments, which hinders the reliable identification and mitigation of noise sources. This study addresses the need to determine the specific physical mechanisms responsible for anomalous noise levels in surface traps.

Methodology
The authors investigated electric field noise along a 1.2 mm segment of the symmetry axis of a linear surface Paul trap fabricated on a fused silica substrate. The trap electrodes were coated with 10 nm titanium, 10 nm platinum, and 1000 nm gold, with specific attention paid to preventing electrical shorts between electrodes. Following fabrication, the trap underwent a rigorous cleaning and baking process (heating to 450°C for one hour and 150°C for four days) within a controlled environment to minimize contamination.

To measure noise, the authors trapped a single $^{40}\text{Ca}^+$ ion and measured the heating rate ($\dot{n}$) of its axial ($z$-mode) motion. The heating rate was determined by monitoring the increase in the mean motional occupation number over time after Doppler cooling, using resonant Rabi oscillations on the $S_{1/2} \leftrightarrow D_{5/2}$ transition to infer the motional state. Measurements were taken at various trapping locations ($L1$ through $L7$) and across a range of motional frequencies (0.6–1.4 MHz) to characterize both the spatial and frequency dependence of the noise.

Following noise measurements, the trap surface was characterized using optical microscopy and scanning electron microscopy (SEM) to identify potential physical defects. The authors also developed two theoretical models to explain the observed noise: a numerical finite element analysis (FEA) treating particles as lossy dielectrics, and a simplified analytic model for thin dielectrics on a metal surface.

Key Results

1. Spatial Variation of Noise: The study revealed a dramatic spatial variation in noise levels. In the first 600 µm section of the trap (locations $L1$–$L3$), heating rates were constant at approximately 200 quanta/s, consistent with state-of-the-art room-temperature traps. However, in a subsequent 600 µm section ($L4$–$L7$), noise levels varied by three orders of magnitude, reaching a maximum of 77,000(7,000) quanta/s at location $L6$.
2. Identification of Particles: Optical and SEM imaging identified micron-sized particles located near the trapping sites with the highest noise levels. Two specific particles, $p1$ and $p2$, were identified near locations $L5$ and $L6$. These particles were several micrometers in diameter and situated along the edge of a trap electrode.
3. Frequency Dependence: In the low-noise region ($L1$–$L3$), the noise spectral density followed a power law $S_E(\omega) \propto 1/\omega^\alpha$ with $\alpha \approx 1.3$ and $0.9$. In the high-noise region dominated by the particles, the exponent was lower, with $\alpha \approx 0.9$ and $0.7$, suggesting a frequency dependence closer to $1/\omega$.
4. Modeling:
   • Lossy Dielectric Model: By modeling the particles as lossy dielectric chunks with an effective loss tangent $\tan\theta = 0.33(0.06)$ and relative permittivity $\epsilon_r = 5$, the authors successfully reproduced the magnitude, spatial dependence, and frequency dependence of the observed noise. The fit yielded a reduced chi-squared value of $\chi^2_\nu = 0.83$.
   • Analytic Model: A simpler analytic model for thin dielectrics qualitatively captured the spatial dependence but provided a poorer quantitative fit ($\chi^2_\nu = 3.4$).

Significance and Claims
The authors conclude that the strong position-dependent noise observed in their trap is caused by micron-sized particles on the electrode surface. They propose that modeling these particles as lossy dielectrics with a specific effective loss tangent accounts for the anomalous noise levels.

The paper posits that these findings may explain the wide range of heating rate data reported in existing literature. Specifically, the authors suggest that traps exhibiting significantly elevated noise levels may be limited by extrinsic factors—such as particles of various sizes, distributions, and materials—rather than the intrinsic surface electric field noise that affects all traps. This distinction implies that noise mitigation strategies may need to focus on particle removal or surface cleanliness in addition to addressing intrinsic material properties. The study does not claim to have identified the specific material composition of the particles, as they vanished from view during SEM imaging, but the dielectric model provides a robust phenomenological description of their effect.