Short answer: Spin squeezing uses entanglement between the atoms of an optical lattice clock to push stability below the standard quantum limit. Published experiments with strontium atoms in a cavity have reported gains of roughly 2 dB. Theory work cautions that with typical laser noise the benefit holds only for fairly small ensembles. All of this is laboratory research.
The limit being attacked
Optical atomic clocks are limited by quantum projection noise, which comes from fluctuations in the population of the many atoms interrogated. Each atom's projection noise is uncorrelated, which sets the standard quantum limit (SQL). Lattice clocks interrogate large ensembles to reduce that noise, but density-dependent frequency shifts make it hard to keep adding atoms. Entanglement offers another route, because it can surpass the SQL, though an open question has been whether it can give an advantage at the stability levels of state-of-the-art clocks.
Experiments reported
A Nature Physics paper (2023) describes an optical clock platform combined with a cavity for quantum non-demolition measurements. A movable lattice lets the cavity address two independent subensembles, so two squeezed clock ensembles can be squeezed in turn and compared without the influence of clock laser noise. The authors state that the comparison remains above the effective standard quantum limit, while directly verifying a 1.9(2) dB clock stability enhancement at the 10^-17 level without subtracting technical noise contributions.
A later Physical Review paper reports clock performance beyond the SQL, with a fractional frequency precision of 1.1 x 10^-18 for a single spin-squeezed clock. The authors prepared two spin-squeezed ensembles of about 30,000 strontium atoms in a two-dimensional optical lattice using cavity-based measurements. A synchronous comparison with a 61 ms interrogation time gave a metrological improvement of 2.0(2) dB beyond the SQL after correcting for state preparation and measurement errors. They describe it as the most precise entanglement-enhanced clock to date.
Where theory sets limits
A 2020 Nature Communications analysis models a closed servo loop of an optical clock with dead time and laser noise. It finds that for usual cyclic Ramsey interrogation of single ensembles, even with the most stable lasers of the time, spin squeezing can only improve stability for ensembles below a critical atom number of about one thousand in an optical strontium lattice clock. A tenfold better laser would still leave the critical number below 100,000. The authors note that clocks with smaller, non-scalable ensembles, such as ion clocks, can already benefit from squeezed states with current lasers.
How to read these results
The experiments and the analysis are not in conflict as stated: the analysis covers a specific interrogation scheme, and the experiments use cavity measurements and a different comparison. Numbers are quoted from the papers and depend on their methods. Nothing here describes a clock in routine operation with squeezing.
FAQ
What is the standard quantum limit for a clock?
The limit set by the uncorrelated quantum projection noise of the atoms.
How large was the reported improvement?
About 1.9 to 2.0 dB in the two experiments cited, depending on how the comparison is corrected.
Why not just use more atoms?
Density-dependent frequency shifts make scaling the atom number difficult.
Can ion clocks use squeezing?
A 2020 analysis says clocks with smaller, non-scalable ensembles, such as ion clocks, can already benefit with current clock lasers.
How is the squeezing made?
Through cavity-based quantum non-demolition measurements of the atoms.
All field notes