edfas.org 47 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 25 NO. 4 ensure that our students are prepared to tackle its ongoing challenges and contribute to advancements in the field.” Stephen Phillips, electrical engineering professor and director of the School of Electrical, Computer and Energy Engineering, believes these new efforts will further ASU’s leadership in microelectronics. “By establishing itself as a center of excellence in metrology and failure analysis, Arizona State University becomes an attractive destination for students and researchers seeking to make significant contributions to the semiconductor sector,” Phillips says. “In turn, this bolsters the local talent pool and enhances the region’s competitiveness in the semiconductor market.” The type of heterogeneous integration Dr. Celano refers to is illustrated by the work of Trevor Thornton and his research team. Dr. Thornton’s and his team are investigating the realization and characterization of wide bandgap material heterostructures for high power, high temperature applications. His recent work involves the study of cubic boron nitride (c-BN) and p-type polycrystalline diamond MOSCAPs. Individually, these materials possess remarkable thermal conductivity and large bandgaps, allowing them to efficiently dissipate heat even under extreme voltage conditions, but characterization of their heterostructure has not yet received adequate attention from researchers. Electron cyclotron resonance plasma-enhanced chemical vapor deposition (ECR-PECVD) is utilized to deposit c-BN films on polycrystalline diamond. The advantage of this method over other methods such as physical vapor deposition or molecular beam epitaxy is that high-energy ion bombardment of diamond substrate as well as high voltage bias to the substrate is not present. In addition, purity of grown c-BN films are higher thanks to the low pressure, about 10-5 torr. Polycrystalline diamond substrates are first wet-cleaned, then inserted into the ECRPECVD chamber. Microwave power of 1.4 KW is utilized to ionize the gas species inside the growth chamber. Growth was based on fluorine chemistry in addition to He and Ar to help build the plasma inside the chamber. A low voltage bias of -60V is applied to the substrate to facilitate growth of c-BN and to prevent incorporation of other crystallographic phases of BN. Depending on the desired thickness, growth duration varies between two to six hours. Grown films are then characterized using x-ray photoelectron spectroscopy (XPS) and transmission electron microscopy (TEM) to verify the presence of c-BN (Fig. 1). To create the MOSCAPs for characterization, three polycrystalline diamond samples, measuring approximately 1cm x 1cm, were prepared for the growth of boron nitride layer with estimated thicknesses of 100 nm, 200 nm, and 300 nm. Using contact-based lithography, circular pads with diameters of 400 µm, 200 µm, and 100 µm were created. The front and back surfaces of the samples were subsequently coated with a metal stack of Ti/Pt/Au, using an e-beam deposition system. The final structures obtained after the lift off process are illustrated in Fig. 2, which shows an optical view as well as a scanning electron microscope image. Fig. 1 A TEM image showing an example of c-BN film grown on diamond. The capacitor structures have been characterized for DC isolation, breakdown, and capacitance vs. bias. These characterizations showed excellent isolation, high dielectric strength, stable capacitance versus bias, and linear capacitance scaling versus pad area. The author would like to acknowledge Vishal Jha and Ali Ebadi Yekta of Arizona State University for their generous contributions to this article. Fig. 2 Top view of the sample showing many capacitor pads under optical microscope. The magnified inset view to the right shows a scanning electron microscope image of a small area on the sample with the diameters of the capacitor pads indicated.
RkJQdWJsaXNoZXIy MTMyMzg5NA==