Portable laser-cooled ytterbium beam clock based on an ultra-narrow optical transition

Abstract

The highest performance atomic clocks are based on interrogation of ultra-narrow optical transitions. There is now significant interest in developing these systems as a source of GNSS-independent time in deployed, dynamic environments. We report on the development and field trial of a portable optical atomic clock interrogating the 10mHz wide $^1$S$_0\rightarrow ^3$P$_0$ transition in ytterbium-171. To enable measurement of this ultra-narrow transition in a deployed setting we combine an atom-vapor based pre-stabilization reference with all-digital control and continuous clock spectroscopy of a transversely-cooled thermal atomic beam. Characterization of the short-term frequency stability within the lab demonstrates a modified Allan deviation of $2\times 10^{-14}/\sqrtτ$ for integration times up to 100s, reaching a best performance of $1.9\times 10^{-15}$ at 200s. The clock demonstrated the same performance after transport and install aboard a ship for field trial, and operated uninterrupted for multiple days whilst at sea. These results show a pathway towards truly portable optical frequency references based on the interrogation of ultra-narrow transitions.

Figures (7)

• Figure 1: (a) Schematic of the clock physics package. Atomic beam passes from left to right through cooling, interrogation, and detection lasers, which address the 399 nm, 578 nm, and 556 nm transitions in neutral $^{171}$Yb, respectively. Experimental measurement of the atomic flux distribution as a function of (b) transverse and (c) longitudinal velocity as measured in the detection stage, without (red) and with (blue) cooling. The dotted lines show the longitudinal velocity distribution which is excited during RB interrogation (see Supplementary Material Sec. 1 for details), and the yellow shading indicates the range of velocities that contribute to the clock measurement. The black curve in (c) shows a fit to a Maxwell--Boltzmann flux distribution with a temperature of $454^{\circ}$C.
• Figure 2: Portable vacuum system for the Yb beam clock (a) showing the in-vacuum optical components of the cooling, interrogation and detection stages, and (b) showing the externally mounted PMT and cooling optics. An off-the-shelf vacuum pump (not shown) maintains a base pressure of $1.5\times 10^{-9}$ Torr.
• Figure 3: (a) RB spectroscopy signal for a longitudinal atomic velocity of 95 m$\text{s}^{-1}$ and (b) RB spectroscopy over a reduced scan range for different longitudinal velocities.
• Figure 4: (a) Schematic of the photonics subsystem of the clock and (b) frequency noise power spectral density of the 1112 nm laser within the vapor reference and the 1156 nm clock laser, scaled to their frequency doubled outputs. Shaded gray region indicates noise floor of the interferometer. Sharp tones from 100 Hz to 1 kHz in the stabilized data are produced by acoustic noise within the interferometer.
• Figure 5: (a) The portable Yb beam clock, comprising 1. Uninterruptible power supply, 2. Power and oversight system, 3. Commercial cooling laser, 4. Commercial fiber amplifier for interrogation laser, 5. Physics package, 6. vapor reference. (b) Frequency instability ModADEV of the Yb beam clock (orange) and vapor reference (green) in lab-based tests, with specified frequency stability of the Microchip MHM-2020 hydrogen maser for comparison. Black line shows $2\times 10^{-14}/\sqrt{\tau}$. Shaded regions indicate stability limits due to predicted atom and photon shot noise from measured atomic fluxes, fluorescence and background scattering.