Technology seems to improve year after year, like magic. But behind every incremental improvement and breakthrough revolution is a group of hard-working scientists and engineers.
UC Santa Barbara professor Ben Mazin is developing sophisticated optical sensors for telescopes and observatories.published in Physical Review Lettershe and his team improved the spectral resolution of superconducting sensors, an important step toward their ultimate goal: analyzing the composition of exoplanets.
“We were able to roughly double the spectral resolution of the detector,” said first author Nicholas Zobrist, a doctoral student in Mazin’s lab.
“This is the largest increase in energy resolution we’ve ever seen,” Mazin added. “It opens up a whole new avenue to scientific goals that we couldn’t achieve before.”
Mazin Labs uses a sensor called MKID. Most light detectors—such as the CMOS sensors in cell phone cameras—are silicon-based semiconductors. They operate via the photoelectric effect: a photon strikes a sensor, knocking out an electron, which can then be detected as a signal suitable for processing by a microprocessor.
MKIDs use superconductors, where electricity can flow without resistance. In addition to zero resistance, these materials have other useful properties. For example, semiconductors have gap energies that need to be overcome to knock electrons out. The associated gap energy in a superconductor is about 10,000 times less, so it can detect even weak signals.
What’s more, a single photon can knock many electrons out of a superconductor, compared to just one in a semiconductor. By measuring the number of moving electrons, MKID can actually determine the energy (or wavelength) of incoming light. “The energy of a photon or its spectrum tells us a lot about the physics of emitting photons,” Mazin said.
Researchers have reached the limit of how sensitive it is to make these MKIDs. After multiple reviews, they found that energy leaked from the superconductor into the sapphire crystal wafer that made the device. As a result, the signal appears weaker than it actually is.
In a typical electronic product, the current is carried by moving electrons. But they tend to interact with their surroundings, dispersing and losing energy in so-called drag. In a superconductor, two electrons would pair up — one spinning up, one spinning down — and this Cooper pair, as it’s called, is able to move without resistance.
“It’s like a couple in a club,” Mazin explained. “You have two people paired up, and then they can walk through the crowd together without resistance. And one person stops and talks to everyone along the way, slowing them down.”
In a superconductor, all electrons are paired in pairs. “They all danced together, walked around, but rarely interacted with other couples because they all looked deeply into each other’s eyes.
“A photon hitting the sensor is like someone came in and spilled a drink on one of the buddies,” he continued. “This breaks the couple up, causing one of the partners to stumble upon the other couple and cause interference.” This is the mobile electronic cascade measured by MKID.
But sometimes it happens on the edge of the dance floor. The offended side stumbled out of the club without hitting anyone else. Good for other dancers, but not for scientists. If this happens in MKID, the light signal will appear weaker than it actually is.
Mazin, Zobrist and their co-authors found that a thin layer of metallic indium between the superconducting sensor and the substrate greatly reduces the amount of energy leaking from the sensor. Indium is basically like a fence around the dance floor, keeping the jostled dancers in the room and interacting with the rest of the crowd.
They chose indium because at the temperatures at which MKID operates, it is also a superconductor, and neighboring superconductors tend to cooperate with each other if they are thin. The metal did present a challenge for the team, though. Indium is softer than lead, so it has a tendency to cake. That’s not great for making the thin, uniform layers that researchers need.
But their time and effort paid off. The technology reduces wavelength measurement uncertainty from 10 percent to 5 percent, the study reports. For example, photons with a wavelength of 1,000 nanometers can now be measured with 50 nm accuracy using the system. “This has real implications for the science we can do,” Mazin said, “because we can better resolve the spectrum of the object we’re looking at.”
Different phenomena emit photons with a specific spectrum (or wavelength), and different molecules absorb photons of different wavelengths. Using this light, scientists can use spectroscopy to identify the composition of objects nearby and throughout the visible universe.
Mazin is particularly interested in applying these probes to exoplanet science. Currently, scientists can only perform spectroscopic analysis of a small number of exoplanets. The planet needs to pass between its star and Earth, and it must have a thick atmosphere to allow enough light to pass through it for researchers to use. Still, the signal-to-noise ratio is pretty bad, especially for rocky planets, Mazin said.
With a better MKID, scientists can take advantage of light reflected from the planet’s surface, rather than just being transmitted through its narrow atmosphere. With the capabilities of the next-generation 30-meter telescope, this will soon be possible.
Mazin’s group is also experimenting with a completely different approach to the energy loss problem. While the paper’s results are impressive, Mazin said he believes the indium technology could become obsolete if his team succeeds in this new endeavor. Either way, scientists are rapidly getting closer to their goal, he added.