First observation and measurement of the ${}^{198}\text{Hg}$ bosonic transition in an optical lattice clock

Abstract

We report the first observation of the magnetic-field-induced (5d10 6s2)1S0-(5d10 6s6p)3P0 transition in a bosonic isotope of mercury, 198Hg, realized in an optical lattice clock. We characterize this new isotope, determining key features such as the quadratic Zeeman shift, the probe light shift, and the magic frequency. We also report a first comparison between the 198Hg optical lattice clock and 87Sr. In this comparison, the 198Hg clock has a relative frequency stability of 6x10-16/sqrt(tau/s) and a total relative systematic uncertainty of 6.9x10-16. This comparison yields the first direct determination of the 198Hg/87Sr optical frequency ratio: 198Hg/87Sr = 2.629 315 734 684 118 1, with the same relative uncertainty.

Figures (4)

• Figure 1: (Left) The ${}^{198}\text{Hg}$ bosonic clock transition magnetic-field induced mixing scheme: The doubly forbidden ${}^{1}\text{S}_{0}–{}^{3}\text{P}_{0}$ transition is enabled by a strong static magnetic field $\textbf{B}$, coupling the ${}^{3}\text{P}_{0}$ and ${}^{3}\text{P}_{1}$ states via the matrix element $\Omega_B$ through magnetic dipole interaction (dotted orange lines). The cooling transition ${}^{1}\text{S}_{0}–{}^{3}\text{P}_{1}$ with its electric dipole coupling $\Omega_L$ is indicated by the dotted pink line. (Right) Probing configuration for ${}^{198}\text{Hg}$: The probe beam polarization is aligned with the magnetic field from 3D MOT coils. The vertical lattice is polarized perpendicular to the probe beam.
• Figure 2: (Top) All spectra acquired for the search of the ${}^{198}\text{Hg}$ transition in the optical lattice. Each color corresponds to a scan covering a specific range of 100 kHz by steps of $120$ Hz. In total, more than 25 MHz were scanned before successfully identifying the transition (black arrow). (Bottom) Narrowest ${}^{198}\text{Hg}$ spectroscopy obtained, after optimization: for 200 ms probe time, the full width at half maximum is 4 Hz.
• Figure 3: Measurement of the quadratic Zeeman shift coefficient $\beta_{\text{A}}/2\pi$ with initial error bars (blue) and error bars incorporating a model with additional uncorrelated noise (orange). See text for more details. Blue and green dots represent differential measurements from Mar.-Apr. 2024 and Sep. 2025, respectively. The cross symbol is a value deduced from 3 direct measurements against a Sr clock at currents of 1.1 A, 3.6 A and 10.9 A note_1. The black line represents the estimated coefficient returned by the fit of the measurement values with the modeled error bars. The orange lines represents our final uncertainty range ($\pm1\sigma$), where $\sigma$ is defined here as the smallest modeled error bar.
• Figure 4: Magic frequency determination from two series of differential measurements: one with the Hg clock alone and one assisted by a Sr clock Lodewyck2016SrClock to improve stability. The reference used for the lattice frequency is $\nu_{\text{ref}}=826~855~240$ MHz. The linear fit (orange) returns the value of the magic frequency $\nu_{L}^{m}=826~855~228\pm 28~\text{MHz}$ for $^{198}$Hg. It also gives the experimental slope of $\eta=(-1.41\pm0.04)\times10^{-4}\text{~Hz}/E_r/\text{MHz}$. Both are used to estimate the lattice light shift correction and its uncertainty.