Magnetic field detectors, or magnetometers, are used for metal detection, medical imaging and geological observation, but the devices as currently designed have limitations. Many require “buffer gas” chambers filled with caesium vapor, which are hefty and expensive. Some work within only a narrow frequency range.

The new device, detailed in the latest issue of Nature Physics, attempts to improve upon a magnetometer technique using synthetic diamonds.

A tiny synthetic diamond — smaller than one-twelfth of a thumbnail — contains trillions of minuscule defects called nitrogen vacancies (NVs). When NVs are hit with lasers, the light is absorbed and re-emitted by the nano-sized deficiencies. The light, bounced back by NVs, carried information about magnetic fields nearby.

Despite the promise of synthetic diamond chips, organizing this complex process in a compact and efficient device has proven difficult. In previous models, the process of tagging the diamonds with laser photons had proven largely inefficient. Much of the light is launched straight through the diamond, failing to become captured by one of the NVs.

Researchers at MIT found that they if they could shape and angle both the laser and diamond chip just right, they could force the laser photons to bounce around the prism until all of the light was absorbed by NVs. Their technique sees a angled laser beam shot into a sawed-off corner to square-like diamond chip.

“We gain an enormous advantage by adding this prism facet to the corner of the diamond and coupling the laser into the side,” Hannah Clevenson, a graduate student in electrical engineering, explained in a recent press release. “All of the light that we put into the diamond can be absorbed and is useful.”

The researchers say their newly designed synthetic diamond chip is 1,000 times more efficient than older models.

“What’s cool about this is that they’re using the sample itself kind of like a waveguide, to bounce the light around,” Frank Narducci, a physicist at the U.S. Naval Air Systems Command, told MIT. “Their sample is quite small. Because the laser doesn’t have to be anything particularly special, that could be small, too. So you could envision very small magnetometers. And correspondingly, you could make them very cheap.”