É. Lalande1, A. Lussier1, M. Ward1, R. Shink1, S. Roorda1, F. Schiettekatte1, E. Bousser2, B. Baloukas2, L. Martinu2, A. Ananyeva3, G Billingsley3, E. Gustafson3, G. Vajente3, R. Bassiri4, M. M. Fejer4, A. Markosyan4

1) Département de physique, Université de Montréal, Québec, Canada
2) Department of Engineering Physics, Polytechnique Montreal, Montreal, Quebec H3T 1J4, Canada
3) LIGO Project, California institute of technology, Pasadena, California, USA
4) Ginzton Laboratory, Stanford University, Stanford, USA

After more than 40 years of efforts to reduce the noise down to staggeringly low levels, the Laser Interferometer Gravitational-Wave Observatory (LIGO) detected, in 2015, gravitational waves emitted by a binary black hole merger. Since then, several other events have been detected, including a neutron star merger together with its electromagnetic counterparts. Still, the detection rate of events is about one per week due to noise sources limiting detector sensitivity. In particular, the thin films that form the LIGO mirrors play an important role for the following reason: they consist of Bragg reflectors made of Ti-doped tantala/silica quarter-wave stacks, and the internal mechanical dissipation (IMD) in these amorphous layers (especially in tantala) translates into fluctuations of the mirrors’ surface that form the dominant source of noise in LIGO’s most sensitive frequency band (~100Hz). The IMD is proportional to the loss angle by the fluctuation-dissipation theorem. We are investigating various means to minimize the IMD such as material deposition conditions and post-treatments. These include depositing by High-Power Impulse Magnetron Sputtering (HiPIMS) and tantala doping with Ti and Zr. Key parameters for data interpretation include the composition and density, both of which can be determined by ion beam analysis (with a separate thickness measurement regarding the density), and the Young’s modulus measured by depth-sensing indentation or deduced from frequency shifts in nodal suspension measurements during IMD evaluation. It is shown that while Ti doping decreases the IMD, Zr doping helps to frustrate crystallization, allowing to further decrease the IMD by post-annealing at more elevated temperatures. Indeed, 46% Zr-doped films remain amorphous til 800°C.