Research
My PhD research was in particle physics. More specifically, I built and operated a detector to search for a hypothetical particle called the axion whose existence could explain dark matter.
For my postdoc I moved into quantum information science. Specifically, I designed and characterized devices that use tiny vibrating membranes to enable disparate quantum systems to communicate with each other.
An overarching theme of my research was expanding the reach of quantum technology to physical systems that usually behave classically. I’ll try to explain what all this means in plain(ish) English below.
On a day-to-day level, my work entailed measuring very cold things using lasers or microwave-frequency electrical signals. I spent a lot of time battling what physicists call “noise” — random fluctuations that can mask or degrade a signal of interest. And occasionally I found myself awkwardly crouched over a piece of uncooperative equipment.
You can find a complete list of my scientific publications on Google Scholar. If you don’t have access to academic journals through a university or other institution, all these papers are also freely available on the arXiv.
The Haloscope at Yale Sensitive to Axion CDM (HAYSTAC)
HAYSTAC is the detector I helped design, build, and operate for my PhD research, and the best acronym I’ve ever come up with. It contains another acronym and a few terms that need defining. Let’s walk through it backwards.
CDM stands for cold dark matter, the invisible stuff believed to constitute 85% of the matter in the universe. We don’t know anything about its microscopic nature, except that it is not made of atoms or any other known particles. Whatever it is, its only known interactions with normal matter are through gravity, which only plays a significant role on galactic scales. Any other interactions must be very weak — indeed, dark matter is constantly passing through us as our solar system orbits the galactic center.
The axion is a hypothetical fundamental particle — it might not exist at all, but if it does, it would have all the right properties to explain dark matter. This is an attractive hypothesis because the existence of the axion would also resolve a totally unrelated problem in particle physics that I won’t go into here.
A haloscope is a type of axion detector. At the heart of the detector is the shiny copper cylinder shown in the picture to the right, which is about 10″ tall and is called a microwave cavity. The cavity is inserted into the bore of a powerful electromagnet and then cooled to a fraction of a degree above absolute zero. I won’t go into how it works here, but if you’re interested, I’ve written a blog post on the subject. Suffice it to say that an axion dark matter signal would manifest as a very tiny oscillating electric field that must be detected against a much larger noise background.
The laws of quantum mechanics impose a fundamental limit on how small this noise is allowed to be. Historically, haloscope detectors have been limited by more mundane sources of noise. The main result of my PhD research was to reduce other sources of noise to the level where quantum fluctuations were the dominant contribution.
The key to reducing noise was integrating technology developed for superconducting quantum computing research into a haloscope detector. One of the main challenges here was the very large magnetic field required for axion detection. Generally, superconductors and magnetic fields don’t play nice with each other, and the energy stored in our magnetic field was about equal to that of a small car traveling at 45 miles per hour. On one unfortunate occasion, a power outage dumped all this energy into the experimental apparatus, warping the normally parallel plates of our cryostat into shapes resembling clocks painted by Salvador Dali.
I didn’t discover dark matter during my PhD, and if I were you I wouldn’t hold my breath for a discovery any time soon. Theorists have proposed many hypothetical particles that might constitute dark matter. Even if we restrict our focus to axions, there is a vast range of possible masses that the axion particle could have, and a given experiment will only be sensitive to a tiny sliver of this mass range. The way I think about this research is that we need to solve all of the hard problems facing haloscopes today in order to make future detectors sufficiently sensitive.
During my postdoc, I remained plugged into this field in a small way, contributing to the conceptual development of novel analysis techniques and schemes that exploit quantum technology to circumvent the noise limit imposed by quantum fluctuations. One of these methods has now been implemented in the HAYSTAC detector.
For another introduction to my PhD research that places more emphasis on the theoretical motivation, you can check out the first chapter of my thesis, which was written with non-physicist readers in mind.
Membrane optomechanics for quantum electro-optic transduction
Present-day computer networks use electrical circuits to process information and laser pulses in undersea optical fibers to distribute that information over long distances. This is a pleasing complementarity: lasers are no good for building computers, and electrical cables have much larger transmission losses than optical fibers.
These basic principles also apply to quantum information science. While a true quantum computer is still a long way off, many of the most successful quantum processors are based on superconducting electrical circuits, and long-distance quantum communication experiments invariably use lasers at near-infrared wavelengths.
The problem is that superconducting circuits and lasers don’t naturally interface with each other. We would like a device that can faithfully map the quantum state of a superconducting circuit onto laser light; this is much harder than the analogous task in classical communication because of the extreme fragility of quantum states. The physical implementation of this key component of a future “quantum internet” remains an open problem. I’ve written more about the motivation for quantum transduction here.
My postdoctoral research, supervised by Konrad Lehnert and Cindy Regal, was focused on developing a quantum electro-optic transducer using a tiny tightly stretched membrane (the yellowish square in the picture on the upper left; about half a millimeter on each side) as an intermediate element. The pattern of pads and tethers surrounding the membrane is for acoustic isolation from the outside world.
The basic idea here is that motion can couple to pretty much any physical system you can think of. Specifically, the membrane interacts with both a superconducting circuit and an “optical cavity” formed by two highly reflective mirrors that trap laser light, thus enabling these two systems to talk to each other. The fully assembled transducer (about 1″ tall) is shown in the lower left picture. I plan to eventually write a blog post describing how it works in more detail.
The main impediment to quantum transduction is (you guessed it) noise. In our design, the primary source of noise is random motion of the membrane, which is substantial even in a cryostat at 40 thousandths of a degree above absolute zero. The main result of my postdoctoral research was cooling the vibrations of the membrane to the quantum ground state — the lowest-energy state permitted by the laws of quantum mechanics. We accomplished this using laser cooling techniques borrowed from atomic physics.
This was just the latest result in a long-running and ambitious project that predates my involvement: earlier results were published in 2014 and 2018. We didn’t quite reach the level where we can faithfully transduce quantum states, because there are are other noise sources we haven’t managed to totally eliminate. But ground-state cooling is an important milestone that puts us very close to the threshold for quantum operation. We also demonstrated a first application for our transducer, using it to measure the quantum state of a superconducting qubit.
Though this was an applied physics experiment, it touched on one of the most profound fundamental questions in modern physics: the quantum behavior of macroscopic objects. If you are inclined to scoff at my use of the word macroscopic here, consider that the membrane contains more than a quadrillion atoms!
The efforts of various research groups in the field of quantum optomechanics may be regarded as baby steps towards coaxing more dramatic quantum behavior out of macroscopic objects. For example, we could imagine putting something like this membrane into a quantum superposition of two observably distinct vibrational states — a cruelty-free variation on the famous Schrödinger’s cat thought experiment, in which a cat in a box is often said to be simultaneously dead and alive until it is measured.1This is not an entirely satisfactory framing of the Schrodinger’s cat thought experiment for a number of reasons. I’ve written a bit more about how to think about quantum superposition in this post. For those who are curious, I’ve written a short blog post about quantum optomechanics experiments here.