Thulium as a core of optical clock?


Optics department at Lebedev Physical Institute started to work with thulium atoms in 2008.

Narrow inner-shell transition

In 1999, it was proposed at the Laboratory of Optics of Active Media (LPI RAS) to use the magnetic dipole 4f136s2(J=7/2) → 4f136s2(J=5/2) transition in a thulium atom at a wavelength of 1.14 µm with a spectral width of ~1 Hz as a promising candidate for optical clocks. This idea was supported by I.I. Sobelman and the study of rare-earth atoms was later initiated at the Laboratory.

Low sensitivity to BBR

Magnetic-dipole transitions between the ground state fine structure components in hollow shell lanthanides are strongly shielded from external electric fields by the closed outer 5s2 and 6s2 shells. Such shielding can facilitate the use of inner-shell transitions in optical frequency metrology due to their low sensitivity to external electric fields and collisions. For the clock transition in thulium atom, this results in extremely low polarisabilities and shift of only 0.6 mHz due to blackbody radiation at room temperature.
Learn more

Magic wavelength stabilisation

In 2018 we discovered magic wavelength 813.3 nm for the clock transition in thulium atoms. It appeared that the polarisabilities of the ground and excited states cross at such low angle that for optical clock application the required lattice frequency stabilisation is about 3 MHz, which can be achieved with a wavelength meter.
In 2019 we discovered that there is yet another magic wavelength near 1064 nm. While using optical lattice at 1064 nm, we see that the related frequency shift is less than 3 Hz. Active laser frequency stabilisation to the clock transition is required for more precision.

World-class uncertainty

The list of dominant frequency shifts and corresponding uncertainties is presented in this paper. All of these can be well characterized and corrected to a high degree using moderate assumptions and established experimental techniques. As a result, the systematic frequency uncertainty of the proposed Tm optical clock at 1.14 μm can be reduced to 5x10−18 in fractional units.
The uncertainty mainly comes from the quadratic Zeeman effect. Using clock transition between the other hyperfine sublevels, it is possible to compensate this effect.