Design and Characterization of a Cryogenic Vacuum Chamber for Ion Trapping Experiments

Abstract

We present the design and characterization of a cryogenic vacuum chamber incorporating mechanical isolation from vibrations, a high numerical-aperture in-vacuum imaging objective, in-vacuum magnetic shielding, and an antenna for global radio-frequency manipulation of trapped ions. The cold shield near 4 K is mechanically referenced to an underlying optical table via thermally insulating supports and exhibits root-mean-square vibrations less than 7.61(4) nm. Using the in-vacuum objective, we can detect 397 nm photons from a trapped $^{40}\mathrm{Ca}^{+}$ ion with 1.77% efficiency and achieve 99.9963(4)% single-shot state-detection fidelity in 50 $μ$s. To characterize the efficacy of the magnetic shields, we perform Ramsey experiments on the ground state qubit and obtain a coherence time of 24(2) ms, which extends to 0.25(1) s with a single spin-echo pulse. XY4 and XY32 dynamical decoupling sequences driven via the radio-frequency antenna extend the coherence to 0.72(2) s and 0.81(3) s, respectively.

Figures (4)

• Figure 1: (a) 3D-model cutaway illustrating the cryogenic vacuum chamber for ion trapping experiments. The cold (1) and intermediate (2) shields are thermally referenced to the cryocooler 4 K and 50 K stages via flexible copper straps (3). Mechanically, the chamber is referenced to the optical table via thermally insulating Vespel (4) and Macor (5) posts. Three layers of magnetic shielding material (6) are installed within the vacuum chamber. The imaging objective (7) is positioned on a piezo hexapod (8) at a working distance of 20.5 mm. (b) Enlarged view of the ion trap. Within the cold shield, two small magnets (9) establish the bias magnetic field, and a resonant coil (10) is placed above the trap to provide radio-frequency magnetic fields.
• Figure 2: (a) Schematic of vibration isolation components in the cryostat. Vespel posts (1) mechanically reference the cold shield to the intermediate shield, and Macor posts (2) reference the intermediate shield to the vacuum chamber (3) which is directly bolted to the optical table (4). (b) Time series vibration measurement off of the cold shield. (c) Frequency components of vibration measurement in (b) with resolution bandwidth of 0.05 Hz.
• Figure 3: (a) Schematic of imaging stack components shown with corresponding transmission at 397 nm. (b) Energy level diagram of $^{40}$Ca$^{+}$ with relevant transition wavelengths listed in nm. (c) Shelving transitions used to initialize an ion in the dark state for detection fidelity experiments. (d) Fluorescence histograms for $N = 1.02\times10^6$ trials each of bright and dark using a detection interval of 50 $\mu$s. Solid lines represent a detection model (discussed in the main text). For a bright/dark discrimination threshold of 5 counts, the detection fidelity is 99.9963(4)%, in good agreement with our modeled fidelity of 99.9969(4)%.
• Figure 4: Population contrast as a function of delay time for Ramsey experiments using various dynamical decoupling schemes with rf pulses. Error bars represent the contrast fit error. Solid lines represent a Gaussian fit to the data.