RESUMEN
The nature of the quantum-to-classical crossover remains one of the most challenging open question of Science to date. In this respect, moving objects play a specific role. Pioneering experiments over the last few years have begun exploring quantum behaviour of micron-sized mechanical systems, either by passively cooling single GHz modes, or by adapting laser cooling techniques developed in atomic physics to cool specific low-frequency modes far below the temperature of their surroundings. Here instead we describe a very different approach, passive cooling of a whole micromechanical system down to 500 µK, reducing the average number of quanta in the fundamental vibrational mode at 15 MHz to just 0.3 (with even lower values expected for higher harmonics); the challenge being to be still able to detect the motion without disturbing the system noticeably. With such an approach higher harmonics and the surrounding environment are also cooled, leading to potentially much longer mechanical coherence times, and enabling experiments questioning mechanical wave-function collapse, potentially from the gravitational background, and quantum thermodynamics. Beyond the average behaviour, here we also report on the fluctuations of the fundamental vibrational mode of the device in-equilibrium with the cryostat. These reveal a surprisingly complex interplay with the local environment and allow characteristics of two distinct thermodynamic baths to be probed.
RESUMEN
The giant plasticity of [Formula: see text]He crystals has been explained as a consequence of the large mobility of their dislocations. Thus, the mechanical properties of dislocation free crystals should be quite different from those of usual ones. In 1996-1998, Ruutu et al. published crystal growth studies showing that, in their helium 4 crystals, the density of screw dislocations along the c-axis was less than 100 per cm[Formula: see text], sometimes zero. We have grown helium 4 crystals using similar growth speeds and temperatures, and extracted their dislocation density from their mechanical properties. We found dislocation densities that are in the range of 10[Formula: see text]-10[Formula: see text] per cm[Formula: see text], that is several orders of magnitude larger than Ruutu et al. Our tentative interpretation of this apparent contradiction is that the two types of measurements are somewhat indirect and concern different types of dislocations. As for the dislocation nucleation mechanism, it remains to be understood.
RESUMEN
High-stress silicon nitride microresonators exhibit a remarkable room temperature Q factor that even exceeds that of single crystal silicon. A study of the temperature dependent variation of the Q of a 255 micromx255 micromx30 nm thick high-stress Si3N4 membrane reveals that the dissipation Q-1 decreases with lower temperatures and is approximately 3 orders of magnitude smaller than the universal behavior. Stress-relieved cantilevers fabricated from the same material show a Q that is more consistent with typical disordered materials. e-beam and x-ray studies of the nitride film's structure reveal characteristics consistent with a disordered state. Thus, it is shown that stress alters the Q-1, violating the universality of dissipation in disordered materials in a self-supporting structure.
RESUMEN
We have made reliable measurements of the sound velocity delta v/v(0) and internal friction Q(-1) in vitreous silica at 1.03, 3.74, and 14.0 kHz between 1 mK and 0.5 K. In contrast with earlier studies that did not span as wide a temperature and frequency range, our measurements of Q(-1) reveal a crossover (as T decreases) only near 10 mK from the T(3) dependence predicted by the standard tunneling model to a T dependence predicted if interactions are accounted for. We find good fits at all frequencies using a single interaction parameter, the prefactor of the interaction-driven relaxation rate, in contrast to earlier claims of a frequency dependent power law. We also show that the discrepancy in the slopes d(delta v/v(0))/d(log(10)T) below and above the sound velocity maximum (1: -1 observed, 1: -2 predicted) can be resolved by assuming a modified distribution of tunneling states.
