Columbia University

I was admitted to Columbia to do high-energy physics, and switched to AMO before I started. Quantum mechanics had been winning me over through my undergraduate coursework, and from what I could tell, graduate work in high-energy nuclear physics meant spending most of my time writing code. I wanted the diverse daily experience of an experimentalist with hardware in front of him. Tanya Zelevinsky gave me a shot in her lab, on the molecular cooling experiment.

There was an enormous learning curve — almost nothing I had learned as an undergraduate was directly useful in AMO experiment — but I live for steep learning curves, and within a year I was comfortable. I owe that to Tanya, and to Ivan Kozyrev, our postdoc, and Reese McNally, the senior graduate student, both excellent experimentalists and generous mentors.

We were laser-cooling barium hydride, with the eventual aim of using it as a source of micro-kelvin hydrogen. Tanya is a specialist in molecular photo-dissociation — using a laser to break a molecule non-adiabatically into its constituents without adding kinetic energy to the system. Hydrogen itself cannot be laser-cooled to ultra-low temperatures because its transitions are in the ultraviolet, which fundamentally limits how cold light can take it. Cool a hydride instead, on much lower-energy infrared transitions, and the mass of the alkaline metal bound to the hydrogen carries away more energy still.

That is elegant in principle and difficult in practice, because of the internal quantum structure of molecules. Laser cooling relies on scattering hundreds of thousands of photons. In atoms this is straightforward: parity and angular-momentum selection rules let you pick out a two-state system ideal for the scattering process. Molecules have no such rules — a vibrational transition is possible from any excited-state decay — so as you scatter photons and slow the molecule, you are constantly leaking population into vibrational states that are not resonant with your laser. The fix is to repump those molecules back into the band where scattering is happening, which means many more lasers and a great deal more optics.

We demonstrated laser cooling of BaH. In the same work we showed it was a dead end for trapping: slowing BaH far enough to load a magneto-optical trap would take more scattering cycles than the vibrational leakage allows, and we measured the magnetic moment of the BaH excited state as too weak to trap molecules in any meaningful quantity.

That result is where the group emptied out. Ivan left for finance and Reese graduated into industry, and at the same time the pandemic began. New York in those early months was full of anxiety and uncertainty — hospital tents in Central Park, military hospital ships in the Hudson, tens of thousands dead in the first few months, most of them elderly people in care homes. It felt like being inside an apocalypse film. I decided I was going to finish my thesis anyway, even if the world was ending around me.

So I built the new molecular cooling experiment alone. We chose calcium monohydride: lighter, with cooling transitions in the 600 nm range that were possible to address. I rebuilt the cryostat and the laser system and did all the spectroscopy needed to find the transitions, and six months later I had my graduating result — magnetically assisted Sisyphus cooling, taking CaH down to millikelvin temperatures. Then I moved back to Denver, out of the asphyxiating atmosphere of pandemic New York, wrote the thesis in two months, and graduated.