Improving Detection of Gravitational Waves using Condensed Matter Physics: Understanding Entropy and Defects in Amorphous Materials

Gravitational waves are ripples in space-time that produce a time-dependent strain, in which distances are compressed in one direction and stretched in the perpendicular direction. Their detection relies on extraordinary precision measurements of distances, specifically displacements on the order of 10-19m, 10000 times smaller than the diameter of a proton, accomplished through laser interferometry between large mirrors 4 km apart from each other. The absorption of a metallic mirror is far too large, necessitating the use of Bragg reflecting mirrors which are ¼ λ high index/low index insulating amorphous (non-crystalline) materials. The present limit on detection is due to thermal noise associated with mirror surface fluctuations, due to poorly understood motions of atoms in the amorphous structure which dissipate energy, causing mechanical loss and thermal noise. At cryogenic temperature, these losses are associated with tunneling of many atoms between states of similar energy, known as tunneling level states (TLS), which are ubiquitous in amorphous materials. Amorphous materials lack structural order, making them difficult to describe and making it difficult to calculate and predict their properties compared to crystalline materials which consist of spatially repeated atoms. This difficulty, however, does not preclude their importance to both applications and scientific impact. The properties of an amorphous material depend strongly on how it was produced, and there are some well-defined known defects, but it is not clear how to describe the different amorphous structures produced by different methods. Intriguingly, there exists evidence for an "ideal glass", which while remaining disordered, lacks imperfections in that disorder and thus approaches the uniqueness and low entropy and energy of a crystal. Amorphous silicon (a-Si) is to date the single material where the mechanical loss and thermal noise can be nearly eliminated; TLS can be tuned via preparation conditions over several decades, from below detectable limits to high in the range commonly seen in amorphous systems. A strong correlation with atomic density is seen, as well as with other structural parameters, but TLS vary by orders of magnitude while other measures of disorder vary by less than a factor of two. The lowest loss a-Si is grown in thin film form at temperatures slightly below the theoretical glass transition temperature Tg of Si, similar to results on polymer films and suggestive that high surface mobility during growth produces materials close to an ideal glass, with low entropy and energy, high density, and low losses due to few nearby configurations with similarly low energy.

Speaker: Frances Hellman, UC Berkeley