Gravitational-wave (GW) detectors are pushing the limits of precision interferometry. Fundamental quantum noise is known to compromise their sensitivity in both the high and low end of their frequency band. In the mid-range near 100 Hz, their sensitivity has progressed to the point that thermal noise in the mirrors - caused by the thermal excitation of the molecules that make up the reflective coatings - has become the limiting factor. A combination of cryogenics and novel coating materials is needed to provide coating thermal noise relief - but much research has yet to be done in this field to reach that point.

At ANU we are hosting optical coating deposition equipment that has produced some of the most critical optics of the two Laser Interferometric Gravitational-wave Observatory (LIGO) facilities in the past. The apparatus is a reactive ion beam deposition (IBD) system, in which a beam of high-velocity ionised atoms ablates a sputter target, creating a molecular plume of highly energetic particles. These impinge and settle on sample substrates and grow a thin coating on their surface. By alternating layers of different optical materials we can produce mirrors with extremely high reflection or filters with very high transmission. IBD is known to produce some of the most homogeneous an optically superior coatings of all popular deposition methods.

The amount of thermal noise that optical coatings generate in GW detectors depends on their detailed layer structure and material properties such as index of refraction (IOR), coefficient of thermal expansion (CTE), specific heat capacity, and thermal conductivity. These are fairly well known for mainstream coating materials and easy to measure with commercially available specialised equipment. Another, somewhat more elusive property of thin-film coatings that has a major influence on thermal noise is the mechanical loss angle. It characterises the lag in mechanical response of the coating material when subjected to external forces and represents internal friction in the material - a source of dissipation that causes thermal noise, a coupling mechanism described by the infamous fluctuation-dissipation theorem.

The coating materials of the future need to feature lower mechanical loss for lower thermal noise. Unfortunately, beyond intrinsic material properties, there is a large parameter space for deposition methods, process parameters, and thermal history (annealing) that affects the mechanical loss in the coating. With one of the world’s best deposition facilities for optical coatings and a full range of post-deposition treatment and optical inspection equipment at our disposal, the CGA is on the forefront of coating research for the next generation of precision interferometric devices, in particular GW detectors.

The IBD system is a recent acquisition for us at ANU, and we have a range of experiments planned for the characterisation of the mechanical loss in single- and multi-layer optical coatings that we produce in our coating facility.

For single layers, the default method is to observe the modal ringdowns of mechanical oscillators, such as cantilevers, membranes, double paddle oscillators, or entire disks/wafers. The amount of time it takes for the excited modes to lose energy to internal friction and decay in amplitude reveals the dissipation in the oscillator. With comparative measurements before and after coating deposition and annealing, one can isolate the coating effect on the quality factor of the system and estimate the coating loss angle. These measurements need to be performed down to cryogenic temperatures to determine the suitability of coating materials and their optimal working point.

For multilayers, we deposit full stacks of reflective mirror coatings such as those used in LIGO, and shrink the interferometric layout to the size of a table-top experiment. We will use a single optical cavity formed by two mirrors and lock separate lasers to different higher order modes of the cavity. These higher order modes are orthogonal and their intensity distributions on the mirrors probe different areas of the coated surface. Since the coating thermal noise even between different regions on a single mirror is uncorrelated, the two lasers sense different length fluctuations of the two-mirror cavity. This is a direct way of measuring the same effect that limits the full-scale GW interferometers and the ultimate test for novel coatings before integration into the main instruments can commence.