Common ultrasound transducers utilize epoxy for bonding the active piezoelectric to the rest of the acoustic stack. This paper explores some of the potential benefits of replacing this soft epoxy layer with a harder metallurgical bond. The influence of varying the bond layer thickness was evaluated for epoxy-and intermetallic bonding layers in an ultrasound transducer, as was the effect of voids and delamination that may occur during fabrication of intermetallic bonds. It was found that the intermetallic bond could be made much thicker than an epoxy bond without degrading the transducer performance, and the influence by thickness variations was less. Small voids in the intermetallic where shown to cause a downshift in resonance frequency, whereas regions with delamination will cause reflections that degrade the performance of the transducer.

This paper investigates the influence of high pressure on Au-Sn solid-liquid interdiffusion (SLID) bonds formed by bonding Si substrates to dies of either lead-zirconate titanate (PZT) with high surface roughness or Si with low surface roughness. Bonded samples were exposed to 1000 bar pressure in a silicone oil filled pressure vessel. Samples were characterized before and after exposure by means of scanning acoustic microscopy, optical microscopy and scanning electron microscopy with energy dispersive x-ray spectroscopy. All but one sample successfully passed the pressure exposure. This failed sample had a delamination in the proximity of a large void in the intermetallic layer.

This study presents a finite element approach to estimate acoustic parameters of layers of arbitrary compositions using 2D and 3D models. In this approach the resonance frequency of a layer to be investigated is found by exciting the layer with plane waves and studying the reflected and transmitted sound pressure from the layer. Compressional and shear modes can be excited separately by varying the angle of incidence. A script for generating inhomogeneous layers with randomly distributed inclusions of arbitrary shape and size was developed for this study. A Matlab application was built for processing the result and comparison with analytical calculations. The 2D and 3D models were validated by comparing derived acoustic parameters of known materials with no more than 0.06% deviation from expected values. Estimated parameters for a layer of gold with 10.6% volume fraction of spherical inclusions of voids of 3 µm and 5 µm diameter was found to range from 2540 m/s to 2652 m/s for compressional sound speed and from 1039 m/s to 1067 for shear speed of sound.

No abstract

No abstract