Effect of elevated substrate temperature deposition on the mechanical losses in tantala thin film coatings

Abstract

Brownian thermal noise in dielectric multilayer coatings limits the sensitivity of current and future interferometric gravitational wave detectors. In this work we explore the possibility of improving the mechanical losses of tantala, often used as the high refractive index material, by depositing it on a substrate held at elevated temperature. Promising results have been previously obtained with this technique when applied to amorphous silicon. We show that depositing tantala on a hot substrate reduced the mechanical losses of the as-deposited coating, but subsequent thermal treatments had a larger impact, as it reduced the losses to levels previously reported in literature. We also show that the reduction in mechanical loss correlates with increased medium range order in the atomic structure of the coatings using x-ray diffraction and Raman spectroscopy. Finally, a discussion is included on our results, which shows that the elevated temperature deposition of pure tantala coatings does not appear to reduce mechanical loss in a similar way to that reported in the literature for amorphous silicon; and we suggest possible future research directions.

In silicobroadband mechanical spectroscopy of amorphous tantala

“…Experimental data from Refs. [48,49] (d1), [50] (d2), [51] (d3), [48,49] (d4), [52] (d5), [53] (d6), and [54] (d7). The dashed lines in panels (a) and (c) bound the uncertainty of the NEMD data from the present paper (PW).…”

“…deposition [48][49][50][51][52][53][54]. This seemingly hazardous step was, in part, driven by the missing, to the best of our knowledge, mechanical data concerning quench-cooled glassy tantala but, it is also fully justified in view of recent experiments concluding that the amorphous tantala film shares a common progenitor state with a hypothetical glass quenched from the melt [47].…”

“…6(b) shows that, although the losses evaluated by the MD-TARP model for the glass cooled with rate q = 3.5×10 4 K/ns [36] are inconsistent with the experiment, our NEMD data for the glass with q = 10 4 K/ns deviate from the experiment not more than ∼50%, capturing the weak role of the temperature. In a further effort, we profited from the fact that suitable annealing yields mechanical losses much less dependent on the deposition procedure [48,49,[52][53][54]. Figures 6(c) and 6(d) compare the measurements concerning annealed amorphous tantala films [48,49,[52][53][54] with the NEMD results concerning a glass prepared by slower melt cooling (q = 10 2 K/ns).…”

“…More specifically, amorphous tantala is an ideal candidate as a benchmark system, given the wide set of experimental, e.g., Refs. [47][48][49][50][51][52][53][54][55][56], and numerical [36,57] studies.…”

Amorphous Silicon with Extremely Low Absorption: Beating Thermal Noise in Gravitational Astronomy

“…Thermal noise performance.-To estimate the thermal noise performance of these coatings, fused silica cantilevers were coated at the same temperatures as the disc samples, to facilitate studies of the mechanical loss. Coating mechanical loss may be calculated from the difference between the free amplitude decay of the cantilevers' resonant modes before and after coating [39]. Figure 6 shows the coating mechanical loss as a function of deposition temperature.…”