Some theoretical models predict that WIMPs are small in mass: between 1 and 30 GeV. In order to test for light WIMPs, it is advantageous to use a target nucleus that itself has a low target mass, in order for the WIMPs to transfer their energy effectively through elastic scattering. At Yale we testing a new approach using helium as such a detection material. The scheme I have in mind is a significant departure from previous techniques; in addition to detecting scintillation light and charge, we propose to detect metastable diatomic helium molecules in liquid helium (LHe) using laser-induced fluorescence. The technique of using a cycling transition (where each atom scatters many photons) to detect individual atoms is common in atomic physics, but one must observe each atom for a long time in order to maximize the total signal. Fortunately, metastable helium molecules have an extremely long lifetime in LHe - 13 seconds. The molecules only exist in the helium as a result of ionizing radiation, and the ability to detect individual molecules would allow a detector with an energy threshold of order 100 eV. By comparison, more typical WIMP detectors have thresholds of about 15,000 eV. By driving a two-photon transition, we estimate that of order 10,000 red fluorescence photons could be extracted from each molecule, enabling the optical imaging of very low energy events.

To study laser-induced fluorescence in liquid helium, we built a LHe cryostat and superfluid-filled optical cell, and purchased a laser system to test the idea. We performed two-photon spectroscopy of the metastable helium molecules to determine the optimal excitation wavelength, showed that the molecules may be repeatedly optically cycled, and demonstrated the first imaging of helium molecule tracks produced by individual gamma ray-scattered electrons. We are now investigating the drift of electrons through liquid helium, to pair this signal with proportional light due to ionization charge from individual events.

Along with its applications in particle astrophysics, this method can also be used to image fluid flow and turbulence in LHe, as demonstrated by several new papers from my group. Fluid velocity fluctuations in highly turbulent systems, where disturbances cascade down from large length scales and produce turbulence at smaller and smaller length scales until finally the energy is dissipated as heat, are not well understood, and any universal statistical properties of turbulence are likely to be found in this regime (i.e. at high Reynolds number). Because of its extremely low viscosity, LHe has long been recognized as an attractive medium for such studies. However, the use of LHe in this capacity has been hampered by the lack of a suitable tracer particle. We have now demonstrated that helium molecules may be used for this purpose. Helium molecules will also bind to quantized vortices in superfluid helium, and can be used to image their dynamic behavior at low temperature.