edfas.org 23 ELECTRONIC DEV ICE FA I LURE ANALYSIS | VOLUME 25 NO . 1 discussed it and agreed this approach was too risky. It takes a long time to identify a failure site and prepare it to thepoint of lift-out. Youneed tooptimize for 100% success rates, which would not be possible using the “Florida method,” as we called it. We believed the in situ lift-out path was the right approach. Tom approached Rocky and me to form a company, saying, “I can’t do this by myself, and I can’t pay you. But if you agree to help, you will get partial ownership in the company.” I knew liftout would change the industry and I wanted to be a part of that. Rocky and I both enthusiastically agreed. Tom funded the entire development himself, and venture capital was never used at any time. Tom had one rule we all had to follow: the lift-out work had to be done off site, outside normal TI hours and no TI resources could be used. We brainstormed company names and Tom came up with Omniprobe. His vision was a manipulator with a central core that could enable all kinds of tools—not just a probe tip for lift-out preparation, but fiber optics for light delivery or wires for electrical testing as well. I learned Adobe Illustrator and designed the company logo. Tom requested the “O” in Omniprobe to appear as a wafer with a notch (at the time, 8-in. wafers were just emerging, and they had a notch as an orientation mark). Envisioning extension of lift-out to life sciences, I created an “O” that not only was a wafer with the requested notch, but also could resemble a TEMcross-hatch grid like those commonly used by biologists (Fig. 5). It took nearly 15 years, but the adoption of FIB as amainstreammethod for biological TEM preparation is now realized, and cryo lift-out is currently undergoing adoption.[14-18] From 1998 until we each joined as paid employees, Tom, Rocky, and I worked nights and weekends in addition to our day jobs at TI. Weekly meetings were held at the Omniprobe “office,” otherwise known as Tom’s condominium. In 2001 J. “Mark” Anthony, the former SIMS labmanager in CRL, also joined the Omniprobe teamand providedextensive technical andbusiness advice.Weused an SEM at the University of Texas at Dallas to validate performance against the design, as well as for quality control of first production units. When a FIBwas needed for more involved testing, Tom and Rocky piled into Tom’s car for the three-hour drive to Sematech in Austin, Texas. Fig. 4 Steps of the FIB in situ lift-out method developed by TI’s Central Research Labs in 1996, performed with a prototype nanomanipulator. A through E are secondary electron images, while F is a TEM image. Images B and F were presented at the Spring ECS meeting in 1997.[10] Alasdair McDowall’s holey lacey film is visible at top of image C. Fig. 6 Omniprobe founders Tom Moore, Rocky Kruger, and Cheryl Hartfield celebrate the first product installation in 1999. Fig. 5 The “O” of the Omniprobe logo (a) is designed as a 200mmwafer with a notch (b) and also a cross hatch grid (c) used by life scientists. (a) (b) (c)
RkJQdWJsaXNoZXIy MTMyMzg5NA==