 |
March 31, 2003 — Experts generally agree that if another way isn’t found to make computer chips, the industry will soon show Moore the door — or at least the “law” named after Intel co-founder Gordon Moore, who predicted a doubling of computer chip transistors every 18 months.
 |
While clever engineers have found ways to push the limits of what is physically possible with optical lithography, companies like Mapper Lithography in the Netherlands are experimenting with new methods. The spinout from Delft University is looking at small tech to enable 40 nanometer resolution and beyond.
The most commonly used method to obtain nanometer feature size is to use electron beams to write a pattern on a wafer. This technique has been around for decades and is commonly used for fabrication of photomasks. The problem is you cannot really use it for mass production because it’s a slow process that gets even slower as the features get smaller.
 |
“Mapper’s technology is based on a combination of e-beam writing and conventional optical imaging that enables high-resolution and high-speed lithography,” said Boudewijn Baud, mapper’s chief executive. “Furthermore, it does not need a conventional photomask, which is often very expensive and therefore not feasible for medium volume production typically seen with ASICS (application-specific integrated circuits).”
The designed pattern is not imaged on the wafer directly, but on a converter plate, which consists of an array of tiny micromachined sharp tips with a definition of approximately 100 nanometers. These function as electron emitters and convert the beams of deep ultraviolet light into parallel electron beams.
Competing technologies that use extreme ultraviolet (EUV) light require quite expensive optical lenses. “Generally, the optics accounts for 50 percent of the costs for a wafer stepper,” Baud said. “Since we use less critical wavelengths, our optics are approximately 10 times cheaper.”
John Cossins, product manager for ASML Holding NV, a Dutch lithography thermal systems company, said that “for next generation lithography, ASML has narrowed the focus down to extreme ultraviolet solutions, while Canon and Nikon are looking more at electron beam-related solutions.”
As for Mapper’s technology, he said, “the alternative they’re working on, though promising, is nowhere near a real product yet, and therefore not serious competition in the short term.”
Mapper is indeed still working on the basic technology. “Our first goal is to demonstrate that the concept works,” Baud said. “It will take us another half year to reach the definite answer. The next step is the fabrication of a prototype, which, at the moment, we plan to finish in 2005.”
According to Robert Mariner, managing director of VLSI Research Europe Ltd., today’s integrated circuit production is approaching 0.1 micrometer, but only less than 20 percent of the total volume (which consists mostly of microprocessor and memory applications) is driving the demand for these high resolutions. For 2003, the company expects a total market for wafer steppers of just under $3 billion, which will grow to $6.2 billion in 2005.
He called EUV and Scalpel — another high-throughput e-beam related technology — the two “mainstream leading options.”
“These technologies have gained a lot of momentum and are backed quite strongly by the large semiconductor companies such as Intel and IBM,” Mariner said.
According to Mariner, Mapper technology, fluid immersion and nano-imprint technology are among the “low-probability, high-risk” alternatives. “We have no reason to believe that these technologies are not working — they might work very well and they might find good niche markets — it’s just that the semiconductor industry, though leading edge in a lot of fields, is still quite conservative. They go for the lowest-risk option, even if they have to pay more.”
As for Mapper, he said, “they have a difficult task ahead proving the concept, and then convincing the market.”
null
Company file: Mapper Lithography B.V.
(last updated March 31, 2003)
Company
Mapper Lithography B.V.
Headquarters
Lorentzweg 1
2628 CJ Delft
The Netherlands
History
The company was incorporated in July 2000 as a spinoff from Delft University, based on research by Professor Pieter Kruit. It is located on the university campus.
Industry
Integrated circuit production
Employees
20
Small tech-related products and services
Mapper is developing a wafer stepper for use in 22-45 nanometer semiconductor lithography. The stepper uses a high-resolution lithography Mapper (Multi-Aperture Pixel-by-Pixel Enhancement of Resolution) technique that combines components of electron beam and optical lithography to achieve the goals of high resolution, high throughput and low cost. There are approximately 30-50 potential large-volume corporate customers for this technology.
Selected strategic partners and customers
The company actively seeks partnerships, including co-marketing/co-branding, joint R&D, distribution and funding.
Investment history
In July 2001, Mapper completed an initial round of financing, garnering nearly $2.2 million from Residex Venture Capital Network (round leader), Delft University of Technology and individual investors. The company underwent a corporate round in August 2002 (amount withheld) with participation from KT Venture Group.
Barriers to market
Mapper’s technology has been dubbed by some analysts “high-risk, low-likelihood” and is only one among several alternative methodologies that will compete in a conservative marketplace with the more mainstream EUV and Scalpel techniques. The company is still in a proof-of-concept stage and estimates that it will not have a prototype until 2005.
Competitors
Goals
Short term: “To demonstrate the proof of concept,” said Boudewijn Baud, chief executive. Long term: “To become a world player in high-throughput advanced lithography.”
Why they’re in small tech
Micromachining is a key technology for Mapper to achieve light-to-electron-beam conversion.
What keeps them up at night
“The excitement to bring a disruptive technology to the market that enables the extension of Moore’s Law in an economically viable way,” Baud said.
Contact
— Research by Gretchen McNeely