|
SAFC Hitech puts Sheboygan--and itself--on the semiconductor materials map |
|
|
|
May 01, 2008 at 06:04 PM |
Nestled along the western shore of Lake Michigan an hour north of Milwaukee, the mythic burg of Sheboygan is not the first location that comes to mind when you think of semiconductor materials--or anything in the chipmaking value chain for that matter.
Kohler makes its bathroom fixtures there, and beer icon Pabst used to run a plant in the Wisconsin city too, but microelectronics were not in the mix.
Until a few years ago, the name Sigma-Aldrich wasn't synonymous with the semiconductor world either, although the chemical company's SAFC business unit was certainly well known in the bio/biopharma sectors. Things started to change when the company saw an opportunity to become a force in the electronics and material sciences market and eventually bought UK-based Epichem in February 2007, laying the foundation for its revamped SAFC Hitech operation. The company's Sheboygan campus plays a key role in its semiconductor materials strategy, and in a nod to the area's traditions, the facilities include a production building that once housed a Pabst sterilization plant.
The company invited a select group of trade journalists to Sheboygan last week for a briefing and tour of its brand-new manufacturing cleanroom and distillation areas. The facilities expand on the already-prodigious production capabilities at the site, which Geoff Irvine, director of marketing and commercial development for SAFC Hitech (and one of the key strategists behind the unit's creation and growth), described as "multipurpose, highly configurable plants. We can leverage off a common asset base but can also specialize."
The new 5000-square-foot building features an ISO Class 4 cleanroom (that's Class 10 or so for you old-school , FS-209-type microcontamination geeks), distillation suite, and offices. Its capabilities include the cleaning and filling of containers, ranging in size from point-of-use, on-tool quantities of 500-cc to fab-level amounts of 200-L. A high-end Thermo Electron ICP-MS tool offers some of the best trace metals analysis money can buy--or at least a good chunk of the $9 million spent on the facility.
With the need for ultra-high-purity (with metallics at the top of the list), there are few metal surfaces in the cleanroom, something evident in the custom-welded PVC walls and ceilings and other customized touches. Since the parent company has as good an accident record as any in the chemical industry, safety was uppermost in the minds of the designers, so elements like an explosion-proof room and state-of-the-art chemical exhaust and containment features have been integrated along with ultra-high-purity handling and workstation capabilities.
Inside SAFC Hitech's shiny new Sheboygan cleanroom.
(Photo: Mark Button)
A new distillation building contains six purpose-built suites, where ALD precursors and other materials are brought for processing and packaging. The lines are capable of manufacturing metric tons annually of low-k and high-k dielectrics, and the 200-L electropolishing stations add the finishing touches to the recipe. The cycle time from start to finish is about 10 weeks, with final purification accounting for about two weeks of that span, according to the company's Reaves Prater. Some gear is off-the-shelf, such as the Pope Scientific vessels, while other tools have been designed and built by SAFC Hitech gadgetheads, such as a patented magnetic, air-driven stirring mechanism.
Of course, without a proper design and development cycle in place, all of this shiny new capability would be a waste of time and money. "There's one helluva gap between something that seems to work in R&D to an actual commercially and technically acceptable high-volume molecule," said Peter Heys, group R&D director and public house pontificator. "A massive amount of work goes into it."
He described the rigorous, multistep process, starting with literature review and molecule design, through the small quantity synthesis and chemical analysis phases, on through the primary physical property studies, film deposition tests, layer characterization, to eventual proof of concept and precursor selection. Of course, as Heys pointed out, if those molecules are not stable enough, the cycle starts over, all the way back to the design phase. Even if a precursor is selected, it still has many iterative hoops of process development, scale-up, further characterization, environmental analyses, and packaging to jump through--with the chipmakers and OEMs playing critical roles.
At the end of the proverbial day, Heys said that only a very small percentage of the precursor candidates examined actually make the cut. In the case of hafnium compounds run through their paces, two out of about 500 were eventually chosen. "A molecule might work for CVD but not for ALD, for example," he explained.
Successfully developing and delivering a new materials may not be enough if the market isn't there. Just because a company may have "the chemical talent to get to a solution," Irvine noted, "will it be in time to meet the window [of opportunity] of the OEMs for the IDMs?" He said that the integrated device folks are starting to tell the equipment suppliers which materials suppliers to use--a list that SAFC Hitech hopes to be on in many a memory, logic, and foundry fab operation.
With the opening of the new cleanroom and 'stills facility, the Sheboygan campus represents the entire electronics chemical supply chain (or at least SAFC Hitech's links in said chain), with production, packaging, and analysis all synergistically located in one site. Given the rampant plundering of the periodic table of elements by the semiconductor industry that's been going on since the turn of the century, this "one-stop shop," as SAFC head honcho Frank Wicks put it, offers the kind of robust investment and capability that chipmakers insist on--especially as other materials suppliers have jumped ship in the electronics space. The company's pervasive global sales and service support infrastructure doesn't hurt either.
Business is good and shows no sign of slowing down, with a 10% growth target this year. "We can sell everything we can make. The challenge is keeping up with demand," noted Irvine. SAFC Hitech (which itself is made up of three groups: silicon semi, compound semi, and performance materials) accounted for $70 million in sales in 2007, which represented a significant chunk of SAFC's overall revenues of just south of $600 million last year. Close to 200 dedicated employees work at the group's seven global manufacturing sites.
Another sign of SAFC Hitech's serious intent is its five-year silicon precursor roadmap rolled out to the industry last year. Divided into six categories--dielectric layers, additional functional memory layers, electrodes, barrier layers, wiring (interconnect), and low-k dielectrics-- the document shows the status of the company's work in those areas--from R&D to pilot to volume production--and where it expects to be by 2012. Enabled in part by a strong collaborative model with customers and partners, the core strength area is thin-film deposition, according to Irvine, concentrated in the areas of dielectrics, barrier seeds, endcaps, and gate materials.
The gameplan for 2008 includes focused acquisitions and alliances, investment in manfacturing and services (with Asia being a major beneficiary--SAFC Hitech would like to get 25% of its revenues from the region by 2010), the obligatory next-gen materials development, a buttressing of sales and support for the compound semi materials biz, and the debut of expanded offerings in the PV, organic LED, sensors, and related optoelectronics sectors.
SAFC Hitech has set the bar high for itself, but the company believes it has a more than legit chance to greatly increase its electronics-focused business, raise its industry profile, and further deploy what Wicks calls its core competency: "the handling of difficult-to-handle materials."
(For more info and other views on SAFC Hitech's new facilities, check out the reports from my esteemed (but not erstwhile) colleagues, Solid State Technology's Ed Korczynski and Semiconductor International's Laura Peters.)
|
|
