Compatibility of trapped ions and dielectrics at cryogenic temperatures

Abstract

We study the impact of an unshielded dielectric $\unicode{x2013}$ here, a bare optical fiber $\unicode{x2013}$ on a $^{40}$Ca${^+}$ ion held several hundred $μ$m away in a cryogenic surface electrode trap. We observe distance-dependent stray electric fields of up to a few kV/m due to the dielectric, which drift on average less than 10% per month and can be fully compensated with reasonable voltages on the trap electrodes. We observe ion motional heating rates attributable to the dielectric of $\approx$30 quanta per second at an ion-fiber distance of 215(4) $μ$m and $\approx$1.5 MHz motional frequency. These results demonstrate the viability of using unshielded, trap-integrated dielectric objects such as miniature optical cavities or other optical elements in cryogenic surface electrode ion traps.

Figures (7)

• Figure 1: Schematic of apparatus and relevant mechanisms. (a) Top view of the surface-electrode ion trap with attached bare optical fiber. An ion (orange circle) is loaded $\approx 325\,\mu$m from the optical fiber and transported to various ion-fiber distances $d$ to quantify stray fields and heating rates. Directions of laser beams, neutral Ca flux, and applied magnetic field $\vec{B}$ are indicated with arrows, and laser beam orientations at the transported ion position are indicated with shorter arrows. (b) Stray charges on the optical fiber surface produce a static electric field $\vec{E}$, while thermal fluctuators inside the bulk (cutaway view) radiate electric field noise with power spectral density $S_E(\omega)$ that heats ion motion.
• Figure 2: Distance dependence of the stray electric field $\vec{E}$. (a) Stray field data in the $x$ (blue circles), $y$ (red squares), and $z$ (purple diamonds) directions. The fitted $E_x$, $E_y$, and $E_z$ are shown as lines in solid blue, dashed red, and dot-dashed purple, respectively. (b) $E_x$ from five sequential data runs [labeled I-V, where II is the data run in (a)], with connecting lines to guide the eye, showing a slow relaxation over several months. Some error bars are smaller than the plot markers. (c) Temperature $T$ of the trap and optical fiber over the time period of the data runs shown in (b), including both full and partial cycles between $\approx$ 6 K and 300 K. All error bars represent 68 % confidence intervals.
• Figure 3: Heating rates versus ion-fiber distance in the (a) axial and (b) in-plane radial directions. The axial data are fit with a model based on thermal fluctuations inside the optical fiber bulk plus a constant heating rate from the trap (red line). The radial data are fitted only to determine the constant heating rate from the trap, with the fiber contribution calculated using the dielectric properties from the axial fit. The axial [radial] data are taken with $\omega/2\pi = 1.45(5)$ MHz [$4.5(6)$ MHz], where the error bars represent 68 % confidence intervals. The fitted heating rate from the optical fiber alone is shown in both panels (purple, dashed).
• Figure 4: Frequency dependence of axial heating rate at an ion position far from the fiber ($d=325(4)\,\mu$m, blue circles) and near the fiber ($d=225(4)\,\mu$m, red squares) with corresponding power law fits to the data (blue solid and red dashed lines) and 68% confidence intervals (shaded regions).
• Figure S1: Photograph of the surface electrode ion trap with an optical fiber attached approximately parallel to and centered on the trap axis. The segment is cleaved to have a clean facet on the side near the trapping region. It is kept in place with a sleeve of wire bonds tacked to the adjacent grounded electrodes and further constrained off-chip with a small amount of adhesive $\approx 2$ cm away from the trapping region (not shown).