At long last, gravitational waves have been found. Now a different sort of hunt for those space-time ripples is picking up speed.
On a mountaintop in Chile and at the South Pole, a new generation of superconducting detectors is beginning the search for the imprint of gravitational waves on the cosmic microwave background (CMB), the universe’s oldest observable light. While gravitational waves passing through Earth can tell us about relatively recent events involving black holes or neutron stars, the discovery of an ancient signature could provide a window into the universe a minuscule fraction of a second after the big bang.
Cosmologists have been hunting for years for evidence of such primordial gravitational waves, which should show up as a “swirl” in the CMB’s polarization. And for a short time it seemed the cosmic quarry had been bagged: In 2014, the South Pole–based BICEP2 experiment declared that it had found the characteristic swirls—a particular polarization pattern called the B-mode. But in the end, the source was identified as something much closer to home—dust from our own galaxy.
A new generation of superconducting receivers, packed with more detectors, could help physicists isolate such confounding signals. While BICEP2 was sensitive to a single frequency—150 gigahertz—these new experiments can pick up multiple frequencies simultaneously. Because the CMB’s 13.8-billion-year-old light has a different spectrum than that of Milky Way dust emission, acquiring data at various frequencies can allow physicists to better identify and subtract the unwanted light.
“BICEP2 really illustrated [that] with one frequency you can only detect one type of signal—in other words, if you have one unknown and one measurement, you can only fit one thing to that one unknown,” says Brian Keating of the University of California, San Diego, who worked on the experiment. “Now the name of the game is multifrequency coverage.”
There are a number of ways to measure the polarization of the CMB at different frequencies. An experiment can use multiple single-frequency receivers or mount them on separate telescopes. Another option, recently employed by the Atacama Cosmology Telescope, in Chile, uses three-dimensional feed-horn antennas that pick up and funnel a wide band of microwaves to a multifrequency detector system.
The Polarbear team, of which Keating is a part, has settled on an approach that uses an array of silicon lenses, each over a single flat antenna. Developed by groups based at the University of California, Berkeley, and UC San Diego, the “sinuous antenna” contains four zigzagging niobium arms arranged more or less like a plus sign. Because of their fractal nature—having a structure that repeats at different scales—the antennas are capable of picking up a wide range of frequencies. The long parts of the arms pick out a particular polarization depending on their orientation. The rest of the system is similar to those used in other CMB receivers; the signals are filtered by frequency and sent to ultrasensitive superconducting devices called transition edge sensor bolometers, which register the microwave signals. The ability to pick up multiple frequencies simultaneously will not only help distinguish signals from unwanted sources of polarization swirls; it will also let physicists boost sensitivity without drastically increasing the size of the detector array.
High sensitivity per unit area is critical for CMB experiments, which demand temperatures less than a degree above absolute zero to operate effectively. That area “is more expensive than Manhattan real estate,” Keating says. “What you want to do is to make each square centimeter extract as much information from the photon field as you can.”
The basic antenna design was patented by engineer Raymond DuHamel in 1987. More recent work, by Gabriel Rebeiz of UC San Diego, and others, paired the structure with a silicon lens, with the aim of creating small antennas sensitive to terahertz and millimeter-wave light. The Berkeley team collaborated with Rebeiz on optimizing the design for CMB observations.
That work included shrinking the size of the antenna, says Berkeley’s Aritoki Suzuki, reducing the smallest feature size to about a micrometer and making the whole antenna and its silicon lens only about 5 millimeters across.
One challenge in this miniaturization, Suzuki says, was devising a method to connect signal-carrying feed lines to the antenna in a way that would be compatible with traditional chip-printing technology. In the end, the team found they could connect the lines by snaking them inward along the arms of the antenna.
In early March, workers were getting ready to construct two new telescopes that will house some of these detector arrays, more than 5,000 meters above sea level on Chile’s Cerro Toco. The first of the new receivers, dubbed Polarbear-2A, should arrive later this year. It will be sensitive to two frequencies and boast a detector count six times that of its single-frequency predecessor, which is mounted on the Polarbear team’s nearby Huan Tran telescope. Before the end of 2017, the group aims to have multifrequency receivers mounted on both of the new telescopes as well as on Huan Tran to create what will be called the Simons Array.
Each approach to multifrequency detection comes with benefits and trade-offs, says Clarence Chang of Argonne National Laboratory and the University of Chicago. Chang and his colleagues have opted to use the Berkeley team’s design to upgrade the sensor array on the South Pole Telescope. That telescope’s new receiver will be sensitive to three frequencies, which will be used to study the CMB as well as the evolution of galaxy clusters. The group plans to begin work installing the new array later this year, at the start of the austral summer.