Positronium Laser Cooling via the $1^{3}S\text{−}{2}^{3}P$ Transition with a Broadband Laser Pulse

We report on laser cooling of a large fraction of positronium (Ps) in free flight by strongly saturating the $1^{3}S\text{−}{2}^{3}P$ transition with a broadband, long-pulsed 243 nm alexandrite laser. The ground state Ps cloud is produced in a magnetic and electric field-free environment. We observe two different laser-induced effects. The first effect is an increase in the number of atoms in the ground state after the time Ps has spent in the long-lived $2^{3}P$ states. The second effect is one-dimensional Doppler cooling of Ps, reducing the cloud’s temperature from 380(20) to 170(20) K. We demonstrate a 58(9)% increase in the fraction of Ps atoms with $v_{1\text{D}}<3.7\times {10}^4\mathrm{m}{\mathrm{s}}^{-1}$.

Figure 1

Experimental layout for Ps laser cooling. (a) Diagram of relevant Ps energy levels and transitions, highlighting transition wavelengths and annihilation or fluorescence lifetimes. (b) Front and top view of the vacuum setup, featuring the cooling (${\lambda }_{243}$) and the two probing lasers (${\lambda }_{205}$ and ${\lambda }_{1064}$), the ${\mathrm{PbWO}}_4$ detector, and the dichroic mirror used to selectively retroreflect the cooling beam.

Figure 2

SSPALS spectra of Ps in vacuum without lasers (black dotted curve), with the $205\mathrm{nm}+1064\mathrm{nm}$ lasers (red dashed curve), with the 243 nm laser only (green dash-dotted curve), and with all three lasers $243\mathrm{nm}+205\mathrm{nm}+1064\mathrm{nm}$ (blue solid curve). The 243 nm laser is firing during the time window from $-20$ to 50 ns (green band), while the $205\mathrm{nm}+1064\mathrm{nm}$ (red vertical line) are injected 75 ns after $e^{+}$ implantation time ($t=0\mathrm{ns}$). Each curve is an average of 90 individual spectra. The statistical error is smaller than the linewidths. For analysis, the spectra were integrated between 150 and 400 ns (light gray area).

Figure 3

Ps velocity distribution measured by SSPALS. (a) Transverse Doppler profile measured by two-photon resonant ionization. A Gaussian fit yields an rms width of 44(1) pm, which translates to a Ps rms velocity of $5.3\pm 0.2\times {10}^4\mathrm{m}{\mathrm{s}}^{-1}$ after deconvoluting the ${\sigma }_{205}$ laser bandwidth. (b) Velocity-resolved increase in the number of ground state Ps atoms, induced by the 243 nm transitory excitation to the $2^{3}P$ level. At resonance, the expected Lamb dip is observed. A 2-Gaussian fit yields an rms width of the enveloping Gaussian of 44(3) pm, which corresponds to a Ps rms velocity of $4.9\pm 0.4\times {10}^4{\mathrm{ms}}^{-1}$.

Figure 4

One-dimensional transverse Doppler profiles of the Ps cloud with (solid curve) and without (dotted curve) interaction with the 243 nm cooling laser beam at a fixed frequency detuning of $-200\mathrm{GHz}$. The semitransparent bands represent the statistical measurement error (one standard deviation of the average).

Figure 5

Number of Ps atoms with $v_{1\text{D}}<3.7\times {10}^4{\mathrm{ms}}^{-1}$, as a function of the cooling laser frequency detuning, normalized to the number of Ps atoms in the absence of all lasers. The dashed horizontal line represents the reference population of Ps in this velocity range with the cooling laser off. The highest observed relative increase is 58(9)% at a cooling laser frequency detuning of $-350\mathrm{GHz}$. The semitransparent bands represent the statistical uncertainties (one standard deviation of the average).