UC Berkeley

One of the more attractive sources of green energy has roots in the popular recycling theme of other green technologies, now known by the term "energy scavenging." In its most promising conformation, energy scavenging converts cyclic mechanical vibrations in the environment or random mechanical pressure pulses, caused by sources ranging from operating machinery to human footfalls, into electrical energy via piezoelectric transducers. While commercial piezoelectrics have evolved to favor lead zirconate titanate (PZT) for its combination of superior properties, the presence of lead in these ceramic compounds raises resistance to their application in anything "green" due to potential health implications during their manufacturing, recycling, or in-service application, if leaching occurs. Therefore in this study we have pursued the application of aluminum nitride (AlN) as a non-toxic alternative to PZT, seeking processing pathways to augment the modest piezoelectric performance of AlN and exploit its compatibility with complementary-metal-oxide semiconductor (CMOS) manufacturing.

Such piezoelectric transducers have been categorized as microelectromechanical systems (MEMS), which despite more than a decade of research in this field, is plagued by delamination at the electrode/piezoelectric interface. Consequently the electric field essential to generate and sustain the piezoelectric response of these devices is lost, resulting in device failure. Working on the hypothesis that buried conducting layers can both mitigate the delamination problem and generate sufficient electric field to engage the operation of resonator devices, we have undertaken a study of silver ion implantation to experimentally assess its feasibility. As with most ion implantation procedures employed in semiconductor fabrication, the implanted sample is subjected to a thermal treatment, encouraging diffusion-assisted precipitation of the implanted species at high enough concentrations. The objective of this study is to understand the resulting phase transformation behavior during Ag precipitation with the intent to ultimately control the electrical operation of AlN piezoelectric resonators in energy scavenging applications.

In this work, multiple source reactive ion sputtering was employed to deposit a thin film of AlN on a 525 microns thick Si substrate, followed by ion implantation (Ag cathode) into the aluminum nitride, and subsequent thermal annealing. Computer simulations were conducted to elucidate the projected range of the silver in the AlN epilayer as a result of the ion implantation process. A myriad of characterization methods including Rutherford Backscattering Spectrometry (RBS), x-ray diffraction (XRD), rocking curve, electron microscopy was employed to quantify the concentration of silver, morphology of silver precipitates, as well as the composition, crystallinity and degree of damage in the ion-implanted AlN samples with respect to thermal annealing conditions. The presence, or lack of precipitates in the samples was utilized to draw conclusions about the feasibility of developing a buried conductive layer in a ceramic matrix via ion implantation.

Computer simulations results obtained via TRIM and TRIDYN confirmed that the maximum concentration of silver lied within 30 - 47 nm from the surface. The RBS data verified the presence of Si, Al, N, Ag, and O2, whose concentration varied with temperature. X-ray diffraction and electron microscopy corroborated the crystallinity of the AlN epilayer. Electron diffraction confirmed both the epitaxy of the AlN film on the (001) Si substrate and the crystalline quality of the epilayer prior to and after the thermal annealing treatment. Electron microscopy revealed that the sputtered AlN film grew epitaxially in a columnar morphology and silver precipitates did form in some of the aluminum nitride samples implanted but only in those implanted with a higher concentration of Ag under high-energy implantation conditions.

It is concluded that the Ag implanted region does indeed have potential as a buried contact layer for piezoelectric activation and sensing if the critical concentration and appropriate thermal conditions can be attained.