SEMICON WEST PREVIEW
Expanded innovation showcase spans gamut from desktop e-beam generation to measuring low-k stacks
By Dr. Paula Doe, Contributing Editor
It may be getting harder, but plenty of folks are still coming up with bold new ideas to break into the semiconductor equipment business.
Once again, visitors to SEMICON West will have the opportunity to conveniently meet with a range of entrepreneurs with new ideas. This year SEMI is expanding its Technology Innovation Showcase to three separate and more focused areas — one devoted to innovations in DFM and frontend processing, one to MEMS and nanotechnology, and one specifically to final manufacturing technologies.
“We were looking for people who were thinking outside the box,” said Ron Leckie, president of Infrastructure Advisors, one of the judges among a volunteer committee of industry executives. The committee demanded data that showed the technologies worked, but many of the companies are still looking for experienced partners for help developing applications. Others are more established companies with significant new technologies ready for market.
What could a fab do with a desktop e-beam generator? Advanced Electron Beams Inc. has a compact field-emission unit already used by other industries, and is now gearing up to develop applications for the semiconductors, such as curing thick photoresist without heat damage, or point-of-generation destruction of volatile organic compounds, or maybe even trying projection e-beam lithography again. “We think it will allow the semiconductor industry to think in new ways to do some new things,” said director of marketing Anne Testoni. “We’re looking for partners to help develop them.”
Key to the system is a very efficient small solid-state power supply, and some classic electrostatics. A keg-sized hermetically sealed vacuum can forms an electrostatic lens that directs electrons from a heated tungsten filament out of windows in a flood. It generates a shower — not a point source — of beams up to 150kV.
The seven-year old company, based in Wilmington, MA, has sold 90 of the units so far, to 40 different customers, mostly for curing polymers in the printing and packaging industries. In the last year it has closed on $12.5 million in venture capital, jumped from 10 to 35 employees, and brought in a bunch of equipment industry veterans to lead the company to new markets.
Actinix, Scotts Valley, CA, has a solid-state alternative for a excimer 193nm laser. Initially developed under an NSF grant as a stable source for the company’s planned phase-measuring microscope for photomask metrology, the simple system for making a high-quality beam with usable speed and power (5kHz, 10mW), without the gas maintenance issues of an excimer laser, has generated interest on its own. Now a major chipmaker is working on joint development of a version for evaluating immersion photoresist. First beta versions are slated to ship in the fall.
“The advantage is its simplicity,” says CEO Jim Jacob, defining the system as “only crystals and mirrors.” It utilizes an improved optical parametric oscillator, which is naturally narrow band, for efficient frequency mixing, allowing a simple architecture with few optical elements. A standard Nd:YAG pump laser beam is split and converted to UV and IR frequencies, through a cesium lithium borate crystal and the optical parametric oscillator respectively, which are then combined through a sum-frequency mixer crystal of beta-barium borate (BBO) to create the 193nm beam.
Another small company, better known for depositing superconducting films, says it has a way to do non-contact electrical test of low-k dielectrics through the deposition process — before there’s any metal. Neocera, Beltsville, MD, makes what’s essentially a parallel plate capacitor to measures the capacitance between the probe and a bottom layer of the structure, using a high-frequency (4GHz) signal that can measure across the air gap. At high frequencies the impedance of the measured film is comparable to that of the air gap, so a calibration can extract out the electrical properties of the film.
Innovations also are coming from more established players. Accretech USA, a subsidiary of Tokyo Seimitsu, aims to replace wet cleaning with a flame-assisted chemical etch that removes film residue from the wafer edge, with an initially developed application for removing polymer. “Wet chemicals can’t dissolve polymer, they have to undercut it to flake it off,” according to SVP Dave Duncavage. “Ours is the only technique that vaporizes the polymer.”
The new approach uses a needle-sized torch to direct reactive gas to a precisely controlled region on the wafer, without the need for masking, and limits heat damage by controlling the wafer’s rate of rotation through the flame. The beta version has been tested for BEOL ILD bevel polymer removal, and the company is now getting ready to ship a commercial model with throughput of 50 wafers/hour. The process also could remove BARC, resist, tantalum and dielectric films, according to the company, which is showing AFM pictures of resist with an atomically smooth edge after cleaning.
Startup RASIRC aims to replace the torches now used to combust pure hydrogen and oxygen to make ultrapure water vapor for wet oxide growth in diffusion and RTP by filtering steam instead. The system boils de-ionized water, then sends the steam through a hydrophilic membrane filter, which effectively removes metals, organics, urea, and dissolved gases. Eliminating the need for pure hydrogen and oxygen could bring significant cost savings, to say nothing of eliminating the occasional explosion. Company president Jeffrey Spiegelman sees eventual potential for purified steam for single-wafer cleaning as well, as steam could quickly penetrate deep features. The San Diego, CA-based company has one system in the field so far, and plans to ship several commercial systems in the next few months to both fabs and OEMs. — Dr. Paula Doe, Contributing Editor