edfas.org ELECTRONIC DEV ICE FA I LURE ANALYSIS | VOLUME 25 NO . 1 22 FIB industry, like Tom and Rocky had done years before with acousticmicroscopy.[9] Ultimately it would take seven design generations over the course of five years to lock down the “volume ready” final design for the open-loop Omniprobe 100.7 manipulator, the world’s first commercialized lift-out product. In 1996, an internal patent disclosure on the FIB in situ lift-out method was submitted to TI’s patent committee. The method is summarized in the six steps of Fig. 4. TI’s patent committee declined to file any patent, as “TI is not an instrument company.” This opened the opportunity Fig. 2 A lift-out sample approaches a cross hatch Cu TEM grid thatwas cut in half with a razor blade. Remnants of holey lacey film are visible at edges of the Cu, especially in the lower left corner of the image. The lamella is roughly 10 µmwide. Fig. 3 a) 1996 photo of the first Omniprobe working prototype (arrow) on a high-angle port of the FEI 820 FIB-SEM inTexas Instruments’ Central ResearchLabs. b) The AutoProbe 200, developed by the early 2000s, was an eighth-generationmanipulator. In 2007 itwas used to demonstrate the world’s first automated lift-out to several microscope manufacturers at Omniprobe’s facility in Dallas (unpublished data). to publish and talk openly about in situ lift-out. In May 1997, TomShaffner, then branchmanager of TI’sMaterials Characterization Labs in CRL, gave an invited talk at the spring Electrochemical Society (ECS) meeting.[10] The talk included the two lift-out images shown in steps B and F of Fig. 4. The region of interest for TEM analysis between the T-gate and substrate is visible in the zoomed call-out images in steps D-F. The lift-out images were published in the ECS abstract booklet that was available online by March 1997 and in the ECS proceedings that published that same November. This is of historical significance, as the date of March 1997 qualifies it as prior art for a broad patent that issued years later having a July 1997 priority date. This andmany other patentsmight have been restricted tomore narrowclaims if the patent examiners had been aware of the ECS prior art. Ultimately a “patent thicket” developed in the U.S. with hundreds of patents on FIB in situ lift-out—a testament to its importance. Eventually two large microscope manufacturers battled each other in a three-year infringement case that was settled in 2012.[11,12] OMNIPROBE: FOUNDING A START-UP COMPANY In 1997 the in situ lift-out progress in CRL was going well, yet little did anyone know that an axe was about to fall. The first inklingof troublewas the reassignment of CRL to a newcost center, the first time this had happened in its history. And then only a fewmonths after Tom Shaffner’s ECS presentation, TI’s Central Research Labs were shut down, along with TI’s three libraries on campus. With the closure of CRL, all lift-out work halted. Since TI had abandoned all interest in FIB lift-out and its development, Rocky, Tom, and I switched into survival mode, adapting into new roles in new departments at TI and having to shift positions twice within a year to stay employed as the company continued to evolve after closing CRL. Knowing the promise offered by the invention, Tom Moore wanted to continue commercialization of in situ lift-out. After lengthy discussions, in 1998 Tom obtained from TI a “statement of noninterest” granting the right to proceed with patent applications, to sell, or to use the invention in any manner desired. Meanwhile, the International Symposium for Testing and Failure Analysis (ISTFA) was in Dallas that year and Tom and I attended the FIB User Group meeting where someone mentioned a group of people in Florida were developing a method using a glass rod and “static attraction.” As we know to- day, that technique is now called ex situ lift-out and is effective through Van der Waal’s forces.[13] After ISTFA, we (a) (b)
RkJQdWJsaXNoZXIy MTMyMzg5NA==