Researchers from ICFO (The Institute of Photonic Sciences) and Hangzhou Dianzi University have revealed that the effect known as “squeezed light” — the organization of photons in motion that ensures that they arrive on a target in a stream-like formation, more like water from a tap than raindrops falling on a roof, and thereby making the light stream quieter and measurement of it more precise — to improve the accuracy of magnetometric measurement. Currently, the most sensitive magnetometer instruments are optically pumped, using laser light to probe magnetically sensitive atoms. However, the sensitivity of these instruments is limited by random variations, or noise, in the light source. Reducing noise can increase the sensitivity of the magnetometer and enable the detection of smaller changes in the magnetic field. Charikleia Troullinou and Dr. Vito Giovanni Lucivero working in the experimental setup in the lab at ICFO. Courtesy of ICFO. Although lasers are designed to be as noise-free as possible, there are limits; light arrives as packets of energy, that is, photons, and the random arrival produces a noise known as “shot noise.” Even the quietest laser still has shot noise, and this often sets a limit on how precise a measurement can be. Still, the shot noise limit is not absolute. Scientists have shown that squeezed light can improve gravitational wave detectors. Applying the technology to magnetometry, however, has so far shown mixed results. ICFO researchers showed that the critical factor is the evasion of measurement back-action. That is, the light that probes the atoms must only disturb the atoms in ways that do not change their response to the magnetic field. The researchers constructed a back-action-evading magnetometer, applied squeezed light, and saw that this improved the sensitivity. The team built a Bell-Bloom (BB) optically pumped magnetometer (OPM) and used polarization squeezed light to observe the response of a dense hot cloud of rubidium atoms (87Rb) to a magnetic field. “We use linearly polarized light to probe the magnetic properties of the hot dense atomic ensemble and implemented a very sensitive magnetometer limited mainly by quantum noise. On top of that, the generation of squeezed light and its use for probing instead allowed us to suppress the photon shot noise in the signal,” researcher Charikleia Troullinou said. “We showed that this directly improves the magnetometer’s performance, making it more sensitive and better in its response to fast signals.” In measuring a microscopic system like an electron or an atom, the microscopic system influences the measuring instrument — it causes a detectable change. This influence is the “action” of the microscopic system on the instrument. According to the Heisenberg uncertainty principle, the instrument must also cause a “quantum measurement back-action” on the microscopic system. For example, when measuring the position of an electron, the back-action disturbs its momentum. More complex measurements can be spoiled by this back-action; for example, when attempting to measure the electron’s velocity by measuring its position, waiting, and then measuring the position again, the result will be inaccurate — the back-action of the first position measurement disturbs its momentum and thus the velocity, before the measurement is finished. A “back-action-evading measurement” doesn’t have this problem, the microscopic system is disturbed by the measurement, though not in a way that spoils the measurement procedure. “We figured out that, in the context of atomic sensors, the Bell-Bloom measurement scheme is naturally back-action evading, since the back-action noise affects the spin component that is not measured,” researcher Vito Giovanni Lucivero said. “Then the effect of squeezed light is beneficial over the entire frequency spectrum.” According to ICREA Professor at ICFO Morgan Mitchell, who led the work, the technique could be applied on magnetometers used in geotechnical applications. The research was published in Physical Review Letters (www.doi.org/10.1103/PhysRevLett.127.193601).