photo of technicians with squeezed-light thingy

Researchers install a new quantum squeezing device into one of LIGO’s gravitational wave detectors. [Image: Lisa Barsotti]

A bit more than two years ago, after their landmark detection of gravitational waves (GWs) from a pair of colliding neutron stars, the LIGO and Virgo laser interferometer observatories went offline for a significant overhaul that aimed to push their already legendary sensitivities to new heights. One of the more exotic-sounding new features of the upgrade would be equipment to inject a “squeezed” state of light at the photodetectors of the observatories. Doing so, it was said, would help the observatories overcome a persistent and elusive source of noise—photon-counting uncertainties attributable to fluctuations in the quantum vacuum itself.

Now—some eight months after the observatories re-started, kicking off their latest observing run—the LIGO and Virgo teams have detailed how the squeezed-light deed was done, and the results of the upgrade thus far (Phys. Rev. Lett., doi: 10.1103/PhysRevLett.123.231107, 10.1103/PhysRevLett.123.231108). The take-home message: Quantum squeezing could increase the facilities’ detection rates by as much as 50%, possibly allowing them to sniff out of a new GW event nearly every week.

Nagging background noise

aerial view of LIGO Livingston

The LIGO site near Livingston, LA, USA. [Image: Caltech/MIT/LIGO Lab]

LIGO and Virgo are essentially giant, L-shaped Fabry-Pérot laser interferometers, with arms 4 km long in the case of the LIGO facilities in Hanford, Washington, and Livingston, Louisiana, and 3 km long for the Virgo facility near Pisa, Italy. As a gravitational wave—an impossibly faint ripple in spacetime, triggered by a distant astronomical event such as a black-hole or neutron star collision—rumbles past Earth, it infinitesimally nudges the heavy mirrors at the end of each interferometer arm. That results in a tiny change in the relative lengths of the two interferometer arms, that's read as an interferometric signal by sensitive photodetectors at the end of the line.

The design of the advanced LIGO and Virgo systems has already made them the most sensitive instruments ever devised, capable of detecting displacements a thousand times smaller than the width of a proton. But at those sensitivity levels, quantum-mechanical noise starts to become an important source of error. And that, in turn, has put a cap on the observatories’ ability to reach deep into the cosmos.

Squeezing out uncertainty

squeezed state diagram

The quantum noise endemic to the big GW detectors comes in two flavors: low-frequency radiation-pressure noise, due to quantum uncertainties in the photon flux hitting the interferometer mirrors; and photon-counting or shot noise at the photodetectors, due to continuous background fluctuations in the quantum vacuum. The recent LIGO and Virgo upgrades aim to conquer the second source, quantum vacuum fluctuations at the interferometer’s so-called dark port, via the use of squeezed vacuum states.

Squeezing is basically a bit of nonlinear-optics sleight-of-hand, in which the Heisenberg uncertainty associated with one quadrature of the light field (say, phase) is reduced by transferring some of it to the other quadrature (amplitude). Using a squeezed state of light to reduce quantum uncertainty at GW detectors is not a new idea. Remarkably, it was first proposed 35 years before LIGO made the first, epoch-making direct detection of a gravitational wave, in a 1981 paper by the physicist Carlton Caves in Physical Review D.

Caves detailed mathematically how tweaking the uncertainties of the amplitude and phase quadratures could “squeeze” the error circle into an ellipse, with one error much smaller than the other. And he even noted that a specific nonlinear-optical tool—a degenerate parametric amplifier—could do the trick, if the amplifier’s pump frequency were twice that of the signal light.

Intricate system

schematic from LIGO paper

Schematic of squeezed-light setup at the LIGO facilities. [Image: M. Tse et al., Phys. Rev. Lett. 123, 231107 (2019); CC-BY 4.0] [Enlarge image]

The squeezed-light upgrades at LIGO and Virgo—which rest on pioneering development work at a smaller testbed facility in Germany, GEO 600—basically follow Caves’ parametric-amplification recipe. Broadly speaking, the systems begin with a 1064-nm “squeezing laser” that’s phase-locked to the high-power laser system driving the interferometer.  Part of the squeezing-laser output is split off to be up-converted to a 532-nm pump beam via second-harmonic generation.

The 532-nm pump field is then sent to an optical parametric amplifier (or oscillator), where it interacts with 1064-nm vacuum fluctuations. The result is a squeezed infrared vacuum field with reduced phase noise, which co-propagates with the light of the control field into the interferometer dark port.

At the dark port, the squeezed state is mode-matched to the output field of the main interferometer laser, allowing errors attributable to vacuum fluctuations in the GW readout to be reduced. The main laser also acts as a local oscillator for the co-propagating control field, and the beat note between the two can be fed back to the squeezing laser to control the amount of squeezing and prevent drift.

A brisker detection rate

graph of Virgo strain sensitivity

Measured strain sensitivities at Virgo without use of squeezed states (black), with squeezing (blue), and with “antisqueezing” (red). [Image: F. Acernese et al., Phys. Rev. Lett. 123, 231108 (2019); CC-BY 4.0] [Enlarge image]

The squeezed light sources at LIGO and Virgo have been operating since the observatories started up their most recent observation run (nicknamed O3) in April 2019. And so far, the teams at both projects seem pleased with the results.

According to the recent papers, adding squeezed light has upped the astronomical distance of binary neutron star (BNS) collisions at which the two LIGO facilities can pick up GWs by 12% to 14%, allowing them to listen for events up to 400 million light-years away. That, according to the LIGO team, implies a 40%–50% increase in the detection rate of such events. The squeezed-light dividend for Virgo is somewhat smaller; its range for BNS collision detection is calculated to have increased by 5% to 8%, which could drive a 16%–26% detection-rate increase.

Next step: Frequency-dependent squeezing

One downside of using squeezed vacuum states to reduce quantum uncertainty at the detector, though, is that—by definition—it tends to increase the other quantum noise component: low-frequency radiation-pressure noise at the mirror. The next upgrade at LIGO (nicknamed “A+”) will attempt to address this through “frequency-dependent squeezing.” That approach will use a narrow-linewidth filter to enable squeezing of the quadrature most relevant for the measurement being made—reducing phase uncertainty at higher frequencies to handle shot noise, and amplitude uncertainty at lower frequencies to take care of radiation-pressure noise.

Under that scheme, “we squeeze both quadratures at the same time, but not at the same frequency,” Michael Evans of the Massachusetts Institute of Technology, USA, told OPN in 2018. “So Heisenberg is always happy.”