Collaboration key as LIGO is prepped for future breakthroughs

The Laser Interferometer Gravitational-Wave Observatory (LIGO) has to date made five confirmed observations of gravitational waves from its two sites in Washington and Louisiana, a spectacular success for the teams of engineers behind LIGO's optics. But developing those systems and making observations possible was a demanding operation for the researchers and industrial partners involved.

At the heart of gravitational wave detection inside the Advanced LIGO sites is a system of test masses - the 40 kg mirrors, each 34 cm across and 20 cm thick, against which the laser beam is reflected during the instrument's operation. These mirrors, along with an arrangement of compensation plates, recycling mirrors and beam splitter components, are collectively termed the "Core Optics" subsystem, and form the cavity optics where observation of a gravitational wave occurs.


Just a couple of weeks after Rainer Weiss, Kip Thorne and Barry Barish were awarded the 2017 Nobel Prize in Physics, the LIGO and Virgo teams revealed that they had detected the first direct gravitational-wave evidence of a neutron-star collision; the cataclysmic cosmic phenomenon that was then confirmed by observations of a resulting gamma-ray burst and spectacular signals across the electromagnetic spectrum. Credit: NSF/LIGO/Sonoma State University/A. Simonnet.

GariLynn Billingsley, LIGO senior engineer and manager of the Core Optics subsystem, confirms that development of an optics system able to meet the requirements of LIGO was a far from trivial matter, demanding the combined efforts of engineers from the LIGO project and several industrial partners from the optics sector.

"We were aware of how challenging the specifications were, but at the same time we knew that every gain we made in these optics would have a direct impact on the final noise floor of the instrument in use," commented Billingsley. "The effort needed to make the Core Optics as good as possible would ultimately be worth it when advanced LIGO was operational, so we worked ‘hand in glove' with the vendors involved."

The design challenges were ultimately met through the use of test masses made from ultra-high purity fused silica and coated with layers of doped tantala (tantalum oxide), with a micro-roughness of less than 0.16 nm RMS in order to meet stringent restrictions on acceptable light scatter.

Billingsley explained that the fused silica used was a Heraeus Suprasil material, with the test masses all super-polished at Coastline Optics in California, and then ion beam figured at what is now Zygo Extreme Precision Optics at Richmond, formerly the ASML Optics facility.

Other partners involved in the project included Laboratoire des Matériaux Avancés (LMA) at the University of Lyon in France, where the core optics were coated, while a now-dissolved group at CSIRO in Sydney, Australia, carried out coating operations on LIGO's Michelson interferometry optics and certain other components - efforts that Billingsley characterized as "spectacular" work.

The target was a final Core Optics design capable of delivering acceptably low losses for the optical signal when taken across the complete sub-system, requiring a delicate balancing act from the engineering team.

Acceptable losses
"Our budget for allowed loss was 75 ppm - that's 75 ppm total loss in a cavity four kilometers long, a very stringent requirement when allocated among the different sources of loss that we knew we were likely to be dealing with," observes Billingsley. "A standard approach would be to divide up the allowable defect losses per measurement instrument type and allow a certain number of ppm in each defect size range, but in reality it turned out that the optical defects were mostly point defects, with very few scratches. This allowed us to re-allocate the loss budget based on what we were actually finding."

In the end the team achieved losses of around 60 ppm, better than they were aiming for, with the quality of the final ion-beam polish playing a significant part in the achievement.

"We posed a significant challenge when we set our RMS limit for the polishing operation, knowing that it was a very tough specification," Billingsley recalled. "In the end, many pieces were delivered at 0.1 nm RMS over an area of 160 mm diameter, which is excellent work."

The coatings on the test masses were a further challenge, and LMA's engineers were tasked with producing a coating uniformity that was, in Billingsley's words, "at the far end of anyone's experience."

The recipe for the final coating arose from the need to tackle thermal noise, one of the most sensitive and important factors, while also retaining all desired optical properties from the test masses. A titania-doped tantala coating, layered with silica, was ultimately developed for the purpose. Tantala/silica is a known material for mirror coatings, but the exacting demands of LIGO turned out to require the addition of titania, in order to both reduce internal friction within the coating itself, and lower the thermal noise of the final coated component.

One of the 40-kilo test masses installed in the twin US LIGO facilities. The massive mirrors form part of the "Core Optics" subsystem that is critical to the detection of gravitational waves. Photo: LIGO/NSF.

Future generations of LIGO could employ radical new approaches to the thermal noise problem, perhaps by using silicon test masses cooled to cryogenic temperatures, while several different concept designs for the Core Optics sub-system as a whole are under consideration, to tackle the inherent trade-off between cost and sensitivity.

In the meantime, 2018 will see LIGO begin its third observation run, termed "O₃", and in preparation a program of modifications and enhancements to the existing instrument and its optics is currently under way.

"Development of the Core Optics sub-system has taught us a great deal about what's needed in the search for gravitational waves, and the existing system has produced some great results," said Billingsley. "But for O₃ we are putting in the latest-greatest test masses, combining the same optics design with the most uniform coatings of the bunch. We now have the opportunity to put in our best technology, and that's what we will be doing."

Laser stability
The specifications of the LIGO interferometer also posed significant challenges for the team tasked with developing a pre-stabilized laser subsystem suitable for use in the hunt for gravitational waves.

Design parameters for LIGO's multi-stage Nd:YAG laser, intended to be capable of supplying power in the 200 W range with exacting limits on the allowed frequency, intensity fluctuations and spatial impurities, meant that the simple purchase of a commercial laser system for use in the observatory was out of the question.

Instead, a development team at the Max Planck Institute for Gravitational Physics (the Albert Einstein Institute, AEI) in collaboration with Laser Zentrum Hannover (LZH) selected a Mephisto laser source from Innolight as its starting point. After Innolight, now part of Coherent, had made some initial modifications to the source's diagnostic channels and driver electronics, the two German centers took the Mephisto laser as a seed source for their high-power laser development process, and set about creating a laser suitable for use in LIGO.

"It would be difficult for a commercial vendor to develop a laser with the specifications we needed, as it is unlikely that there would be any large market for such a source," commented Benno Willke from AEI, the leader of that development project. "We essentially took a commercial 2 W output seed laser with a very stable frequency and narrow linewidth, and then carried out the rest of the work ourselves."

The subsequent development program saw LZH design and build amplifiers able to boost the laser's output power in two stages, first from 2 W to 35 W, and subsequently from 35 W up to the 200 W capacity demanded by the full LIGO specifications. AEI then designed the systems needed to stabilize and characterize the laser output, which in practice included an effort to reduce the laser noise by orders of magnitude compared to what is considered acceptable for a commercial source.

"We were greatly helped by having experience of an existing gravitational wave interferometer, the GEO600, situated south of Hannover," explained Willke. "This 600-meter Anglo-German instrument, construction of which started in 1995, has helped us to understand what is actually required in terms of not just stability and reliability of the laser sources, but also the diagnostics. Adequate monitoring and diagnostics are essential, in order to rule out laser glitches in any meaningful observations of gravitational waves."

Even with this prior experience of gravitational wave observatories, the laser development program was a challenging one, with requirements approaching the limits of what was known to be possible in certain areas. Power stabilization, in particular, is one field where the expertise gained by AEI during its LIGO work has subsequently made it a world leader, according to Willke.

"We found that our LIGO targets could only be reached if we detected roughly 200 mW of light with a photodiode, which is typically not possible," he said. "So our solution was to split the light power onto several photodiodes and then combine the photocurrents. That was the only way to achieve sufficiently low noise levels for our control systems and allow the power stability we needed."

O₃ upgrades
For LIGO's upcoming O₃ operational run, some changes are being made to the pre-stabilized laser subsystem, building on lessons learned during the interferometer's previous operations. One significant alteration is a change to the laser power levels used. Although the source developed by LZH and AEI has a design capability of 200 W, significantly lower powers were actually used during the successful operational runs that saw LIGO detect the distant black-hole and neutron star collisions that prompted last year's physics Nobel.

"Due to thermal loading and certain noise coupling mechanisms, the LIGO interferometers have not been able to operate with the laser at its highest power levels, and instead used around 20 W during O₁ and O₂," said Willke. "The plan for O₃ is to increase this figure to something like 50 or 60 W."

That intermediate goal was itself challenging, as the interferometer has proved to be more sensitive to fluctuations in laser beam pointing than the designers had expected. An initial theoretical limit on the beam pointing had been calculated, based on the team's simulation of how it would influence the output of the detector, but studies of the completed LIGO runs have shown that the beam pointing was in fact limiting the overall sensitivity of the instrument.

For O₃, the solution will be a change to the second laser amplification stage. At present a master oscillator-power amplifier (MOPA) handles the first amplification stage from 2 W to 35 W, with an injection-locked high-power oscillator then employed to boost the 35 W output towards the full 200 W capability. For O₃ the plan is to replace this second amplifier stage with a modified version of the first amplifier, doubling the 35 W input into a 70 W beam.


The laser and vacuum equipment area at the Hanford, Washington, LIGO facility. For the forthcoming "O3" observation run, the laser power is being increased with a new amplifier stage that will reduce fluctuations in beam pointing that has limited overall sensitivity in previous runs. Photo: LIGO/NSF.

LZH's work on the initial 35 W amplifier has already led to the creation of a spin-out company, neoLASE, to exploit the commercial potential of this amplifier architecture for fields other than gravitational wave astronomy. Now LIGO intends to build further on neoLASE amplifier technology, and use it to generate a 70 W laser beam with the low beam pointing needed.

"Maik Frede, co-founder of neoLASE, was group leader of the solid-state photonics group at LZH, and neoLASE is now providing its Nd:vanadate amplifiers for current LIGO upgrades and to other potential end users," said Willke. "There are several possible applications in picosecond laser amplification and other areas that could be exploited commercially, using the technology neoLASE now has."

Virgo, the Italian gravitational wave observatory working alongside LIGO, is also set to benefit. Although its optics are generally similar to those at LIGO there are some differences, including a shorter interferometer arm and some consequent differences in Virgo's Core Optics sub-system, compared to those in the US facilities managed by GariLynn Billingsley.

The pre-stabilized laser is another point of difference, with Virgo requiring a source delivering 100 W, rather than the 70 W intended for use in LIGO's O₃. But the neoLASE laser amplification technology has been modified to suit this goal, and the company has shipped a unit to Virgo for installation. This kind of collaboration and shared expertise is likely to be a hallmark of future gravitational wave research, an evolution from earlier patterns of more isolated development work carried out by individual groups, according to Willke.

"The collaboration between the AEI, LZH, and neoLASE is unique, as it provides an environment with all the skills required for the development,fabrication and test of stabilized lasers for gravitational wave observatories," he said.

"Our research has developed what I think we can fairly call the most stable lasers in the world, and other facilities can start to benefit. Now that real gravitational wave astronomy is being carried out, there is an effort under way to bring the different groups currently working in the field together, and coordinate their laser research and development. That is the right approach, to ensure a bright future for lasers in gravitational wave detection."

Tim Hayes is a freelance writer based in the UK. He was previously industry editor of optics.org and Optics & Laser Europe magazine.

A version of this article appeared in the Photonics West Show Daily in February.

Related SPIE content:

LIGO gravitational wave researchers win Nobel Prize in Physics

Third gravitational wave detected