edfas.org ELECTRONIC DEV ICE FA I LURE ANALYSIS | VOLUME 25 NO . 1 38 who is retired from IBM. Albert started by describing the structure of an MOS device, how to connect a parametric analyzer to a single transistor, and how to collect a family of curves and transfer curves. From these curves, he explained how to derive the important characteristics of a transistor such as Vt, Ion, Ioff, Idlin, and gm. The discussion was kicked off by going around the room and having each attendee introduce themselves and their analytical or research focus. The audience represented a broad cross section of the FA community, and we had a Q&A session that was engaging and informative. ISTFA 2022 FOCUSED ION BEAM USER GROUP Chair/Co-Chairs: Valerie Brogden, Steven Herschbein, Michael Wong, and Edward Principe vbrogden@uoregon.edu, steven.herschbein@gmail.com, mike.wong@thermofisher.com, eprincipe@synchres.com The FIB User Group meeting drew in 105 “Fibbers” this year. Though the User Groups were shortened to an hour, the format was essentially a carryover from prior years, with three short “lightning talks” to stimulate discussion and lots of audience participation between and during each talk. FIB is a widely diverse topic, with several primary applications and tool configurations. We provided a list of 40 FIB topics as discussion suggestions at the end of the presentation, assembled from past inputs, and a QR code for attendees to scan to give real-time feedback on which topics they would like to discuss. The QR code had 12 participants, with five indicating that they would be interested in speaking at the UG next year. However, due to robust audience participation during each talk, there wasn’t time for these suggested topics. To open themeeting, Cecile Bonifacio, Fischione, presented a solution to amorphization and Ga implantation issues fromGa LMIS FIBwith concentrated ion beam (CIB) milling. This techniquemitigates artifacts in FIB-prepared TEM samples using an inert gas at ultra-low Ar milling energies of 2 kV and below. The spot size for CIB tools is anywhere from 1 um to 600 nm. She showed some TEM images of silicon before and after CIBmilling, with a clear improvement in crystallinity after CIB milling. A question from the audience asked which beam currents are used, to which Bonifacio answered 100 pA for all energies. There was also a discussion of whether amorphization is a critical concern. Bonifacio emphasized that this is not necessarily so important for metals, but it is essential for silicon, a common substrate for advanced devices. Much of today’s FA work involves closely packed repetitive structures, requiring a TEM sample that is so thin it could all be damaged by Ga implantation. The ability to stop FIB thinning while still somewhat thick and finish off with a nonmetallic ion species at ultra-low kV could enable much sharper imaging on silicon fins where the structure is critical. Adam Stokes, Thermo Fisher, gave a talk to stimulate audience participation surrounding future needs for PFIB and laser tools. Stokes put forward several questions to the audience on future needs for PFIB applications, including delayering, large volume removal with both PFIB and laser, and other applications, including TEM prep, ECCI, and STEM-in-SEM. The audience was given an opportunity to respond to questions such as, in two to five years, what kind of planarity, precision, throughput, and length scales will be needed for PFIB delayering? What are the primary application spaces we foresee for laser PFIB? Packaging? Display? Therewas a lively discussionon these topics with several varying opinions. In a discussion regarding the different species of the Helios 5 Hydra Dual system, Stokes explained that the interactive volume of each ion species could be controlled by its acceleration voltage, which mitigated a concern that lighter species could introduce a more extensive amorphous damage layer. Another interesting question from the audience was regarding the milling uniformity while performing delayering. Stokes explained that uniformity could be controlled using Thermo Fisher Scientific’s proprietary Dx chemistry and adjusting the raster parameters; the thickness and density of the metal layers also play a role “MUCH OF TODAY'S FA WORK INVOLVES CLOSELY PACKED REPETITIVE STRUCTURES, REQUIRING A TEM SAMPLE THAT IS SO THIN IT COULD ALL BE DAMAGED BY GA IMPLANTATION.”
RkJQdWJsaXNoZXIy MTMyMzg5NA==