A bright beam of BaF molecules
Bright beams of cold molecules provide valuable quantum sensors for measuring
permanent electric dipole moments that violate fundamental symmetries, which enables these tabletop experiments to probe energy scales exceeding those available with current or near-future particle colliders. Consequently, such beams can help shed light on major open questions of particle physics such as the observed matter-antimatter asymmetry. The prime example is a measurement of the electron electric dipole moment (eEDM), whose limit currently provides constraints on broad classes of new physics above 10 TeV (Science 381, Issue 6653).
The precision of such experiments generally scales with the number of molecules that are detected per unit time, typically a few meters downstream of the place where the molecular beam is created. So, improving the precision -and thereby probing new physics at higher and higher energy scales- requires increasing the downstream flux of molecules. This can be achieved by collimating the molecular beam to maintain a high number density. Luckily, these experiments using molecules can borrow techniques for control and manipulation from the extremely well-developed field of atomic physics. One such technique is laser cooling, which is a powerful method to achieve this boost in brightness, whose implementation however remains challenging due to the complex energy level structure of molecules. In this work, by our NL-eEDM collaboration, we showcase 2D transverse Doppler laser cooling of a focused cryogenic beam of BaF molecules. We conclude that with some follow-up steps this method will increase the beam brightness by a factor 100, a crucial milestone towards a high precision measurement of the eEDM.
Unforeseen challenges along the way
Laser cooling an atom has become almost trivial over the years; laser cooling a molecule is a wildly different story. Compared to an atom, a diatomic molecule has a broad landscape of energy levels associated with it's rotations and vibrations. These all have to be addressed before laser cooling can be successfully applied. A careful choice of the transition used for laser cooling can eliminate the problem of populating many different rotational levels. Eliminating the vibrational branching however can only be achieved by setting up a dedicated repump laser for each relevant vibrational level. Finally, each laser has to be stabilized to less than 1 ppb, and two to four separate radiofrequency sidebands are typically added per laser beam to cover hyperfine structure.
The main challenge that had to be overcome for the experiments presented here was the pointing stability of the cooling light, for two main reasons. First, the cooling cycle in the BaF molecule has many different levels, and consequently, it tends to scatter light at a rather modest rate. Secondly, Doppler cooling requires the detuning of the laser light to cancel out the Doppler shift due to the transverse velocity of each molecule, such that each molecule prefers to absorb photons from the laser opposing it's transverse velocity direction. Here, the detuning refers to the difference between the laser frequency and the resonance frequency of the cooling transition. However, the forward velocity of the molecular beam (200 m/s) is far, far larger than the transverse velocity that has to be dissipated (4 m/s). Therefore, any small deviations in the angle of the cooling light will lead to a significant Doppler shift from the forward velocity, and this additional detuning then 'pushes' molecules to a large, nonzero velocity which disturbs the cooling effect. In summary, the successful application of transverse laser cooling to BaF molecules required an interaction region of tens of centimeters where molecules can continually scatter light, from a laser whose detuning is constant to the 1 ppb level.
The easiest method to generate an extended volume of high intensity laser beams is to take one such beam, and let it reflect many times for both transverse directions. It turns out that BaF is exceptionally sensitive to slight misalignment angles of this cooling light due to the NIR wavelength of the transition used and the long excited state lifetime. It is crucial that every reflection of cooling light intersects the molecular beam at the same angle, as one degree change from the desired angle can make the difference between the obtaining the strongest possible cooling force, or the force entirely vanishing. In the end, facing these challenges, we however were happy to achieve sufficient stability and interaction length by splitting the cooling light for each direction into two parallel beams, and rotating the large cooling mirrors (pictured on the poster image) such that the first of these two beams crosses the second one exactly where the molecules fly through the light field.
Looking ahead
The successful implementation of 2D transverse laser cooling brings the NL-eEDM experiment a step closer to a high precision measurement of the electron EDM, which may help shed light on our universe's matter-antimatter asymmetry.