Summary
Researchers compared two room-temperature 176Lu+ optical frequency references and measured a fractional frequency difference consistent with zero. Each reference had an evaluated systematic uncertainty near 1 × 10−19.
Researchers compared two independent optical frequency references, each based on a single trapped 176Lu+ ion, and found their frequencies agreed within a statistical uncertainty of 5.7 × 10−19. The measured fractional difference was [−0.1 ± (5.7)stat ± (1.0)sys] × 10−19. The references’ evaluated systematic uncertainties were 1.2 × 10−19 and 1.3 × 10−19; the comparison tested their agreement at a somewhat larger, precision-limited level.
What the researchers compared
An optical clock uses the frequency of light associated with an atomic transition as a reference. Here, the team used the 848-nanometre transition in 176Lu+ ions, held in two separate traps called Lu-1 and Lu-2. The comparison combined 11 measurements totalling 200 hours.
To compare the ions, the researchers used correlation spectroscopy: both were interrogated together using a shared clock laser. Noise in the laser’s phase that affects both measurements in common can be rejected, allowing the comparison to track the relative phase evolution of the two atomic references. The team reported interrogation times up to 10 seconds and a comparison instability of 4.8 × 10−16(τ/s)−1/2.
The comparison is a direct test of whether two independent references of the same atomic species agree after known frequency shifts are assessed. That matters because an uncertainty budget is a prediction about how closely corrected references should match. The researchers also deliberately varied the magnetic field to stress-test the quadratic Zeeman shift, the largest systematic shift in their assessment.
How the lutetium reference limits shifts
An atomic frequency can be shifted by its surroundings, including magnetic fields, ambient thermal radiation and the ion’s motion. The team’s method averages three hyperfine components of the clock transition, using states labelled F = 6, 7 and 8. This averaging largely cancels sensitivity to quadratic Zeeman and quadrupole shifts. A hyper-Ramsey pulse sequence is used to suppress shifts caused by the clock laser itself.
The researchers report that this 176Lu+ transition is particularly insensitive to blackbody radiation and magnetic fields, and that the ion’s relatively large mass reduces motional shifts compared with lighter species. They assessed the known systematic effects for both references. The individual evaluated systematic uncertainties were near 1 × 10−19, while the estimated systematic uncertainty in their difference was 1.0 × 10−19.
The measured difference was consistent with zero, but its statistical uncertainty of 5.7 × 10−19 sets the precision of the experimental agreement test. In other words, the work pairs very small evaluated uncertainty budgets with a direct same-species comparison that checks them at the level the measurement reached.
Why the result matters for timekeeping
Optical frequency references are being assessed as candidates for a future redefinition of the SI second. This experiment adds a same-species comparison at a level below 10−18, alongside a detailed evaluation of the shifts affecting each reference. It is a contribution to that wider effort, not a redefinition of the second.
The authors also point to possible uses beyond time standards. Comparing precise clocks at different locations can reveal gravitational redshift—the change in clock rate associated with gravitational potential—and could support chronometric levelling at the millimetre scale. The paper presents this as a potential application of more accurate, well-validated clocks; the reported experiment itself compared two references within one laboratory.