Europe sees 'biochips' as long-term endeavor
By Junko Yoshida, EE Times
September 23, 2003
URL: eetimes.com
Europe's leading semiconductor companies hope to grab the brass ring by applying fundamental semiconductor technologies to the field of medicine-in the form of microarrays, microfluidics and neuro chips.
Although biochip R&D activities are carried out on a global scale, Infineon, Philips and STMicroelectronics, no longer see the intersection of CMOS processing and biology as a 21st-century form of alchemy. Today, they identify biochip-tech as the one semiconductor growth market they can't afford to ignore.
Unlike Motorola, which plunged into the biochip market with a big splash in 1999 but quickly retreated, selling off its bioarray business in 2002 to a U.K.-based medical diagnostics company called Amersham, European chip vendors-operating much more frugally-say that they are committed for the long haul.
From lab to market
While tightly guarding the specific names of potential partners, they acknowledge they are forging strategic alliances with leading pharmaceutical, biotech and medical equipment firms, to move their technologies from lab to market.
Thomas Klaue, director of business development and manager of Infineon's biochip activities, believes biochips will change pharmaceuticals forever, just as PCs revolutionized the computer industry 20 years ago. Although still embryonic, biochip technology "strategically makes sense to us," said Klaue. It fits the bill for Infineon's latest corporate agenda to become "the number one semiconductor company in a pioneering field."
Among 200 companies already active in the biochip market, "Missing in small biotech chip firms is mass-production expertise," said Michaela Fritz, product manager, bioscience, at Infineon.
Indeed, large semiconductor manufacturers have much to contribute to biochip development and production. "This includes: patterning biological molecules on a surface of a substrate; sensing technologies to detect biological bindings and localize them; electro-fluidic packaging technology to apply pre-treatment to samples and add filtering to extract microelement from the sample," explained Menno Prins, principal scientist at Philips Research.
However, manufacturing priorities and semiconductor characteristics for biochips differ significantly from those of traditional ICs. Rather than faster clock rates, lower voltages and finer geometries, biochips integrated with sensors feature a lot of analog properties, thus requiring a process technology as large as 0.35 micron, according to Roland Thewes, senior director of corporate research at Infineon.
Biochips in every clinic
Driven by the Human Genome Project, many startups originally joined the biochip fray to speed up or automate the process of unlocking the keys to genes. Now that all genes in the human genome have been sequenced, the new race is to develop high-throughput biochips that exploit DNA information so that pharmaceutical companies can discover the next blockbuster drug faster. While the market for DNA or protein microarrays designed for the drug discovery process is expected to grow big, large semiconductor companies are looking further ahead, toward ultimately capturing a volume market.
The holy grail for biochip researchers is a low-cost chip, capable of sample preparation and pathogen detection, that doctors and nurses can use in intensive care units (ICU), emergency rooms or clinics, to quickly determine the right antibiotic for patients, explained John Hart, program manager, Molecular Imaging and Dx at Philips Medical Systems.
Calling infectious diseases "the killer app for biochips," Hart said that 50 percent of patients, even after successful surgeries, die in an ICU from infections. "The patient may be dead or develop a coagulated organ by the time the doctor finds out a broadband antibiotic originally prescribed didn't work, and then he switched to another antibiotic, based on guesswork," Hart said. "If a biochip can detect the DNA of a pathogen-say within 30 minutes-millions of lives can be saved," Hart said. Further, "we can save huge hospital costs, and physicians no longer have to cry in their sleep at home." Hart predicts that the so-called "point-of-care" biochip market, when materialized, will be 10 times bigger than that of DNA arrays used by drug companies.
"We are completely convinced that the molecular knowledge combined with technology that allows low-cost analysis leads to preventive health care," said Prins. "There, we see an unavoidable big market opportunity."
Today, "microarrays," designed to automate the process of identifying matching samples of DNA or protein with unknown samples, and "microfluidics" to prepare samples to extract the molecules or cells for examination, are separate steps performed by independent biochips or testing equipment. One of the challenges for the biochip industry is to integrate microarrays and microfluidics to lower the cost, said Hart. Currently, this integration remains in the prototype stage.
A recent market research report on Lab-On-A-Chip, which was prepared by Frost & Sullivan, indicated that biochip researchers are encountering the same stumbling block: "how to deal with 'real' fluids, and how to fit everything together on a single chip or on handheld equipment."
Infineon's Thewes agreed. "The proof-of-principle, to marry biology and microelectronics, is already here." The only technology hurdle left is "to fit all the components together with our partners."
European semiconductor companies today are dabbling in the emerging biochip market, using analysis technologies that range from optical and electronic to magnetic.
Infineon is working on three types of biochips. They include: a silicon-based optical biochip, called a "Flow-Thru Chip," designed to accelerate drug discovery; a DNA biochip with integrated electronics for evaluation and analysis of molecular-scale samples; and a Neuro-chip. The optical Flow-Thru Chip, developed with MetriGenix and already on the market, employs an Infineon-developed etching process that builds tiny pores into the silicon substrate to speed test times and increase sensitivity. By applying gene-carrying DNA sections to precisely defined locations on the pore walls, and repeatedly pumping molecules through the chip for analysis, the optical Flow-Thru Chip can make the assay more sensitive and accelerate reaction speed during hybridization. The hybridized samples go through an optical analysis by a separate test device fitted with a CCD camera.
Infineon's electronic DNA chip, designed for clinical diagnostics and patient-specific medication, differs from the optical chip. It uses the flow of electric current as a basis of measurement and analysis. By marking the substances to be examined with an enzyme, sensors equipped with gold electrodes integrated on the biochip "measure the current increases in proportion to the measured time," according to Thewes. The chip is still two to three years away from commercialization.
Infineon's third biochip, called the Neurochip, is developed for neurobiologists to analyze interactions among different areas of the brain, by applying single nerve cells that grow into neural networks on the chip surface.
Infineon etches about a million pores within a surface of a 1-cm2 silicon chip, which can simultaneously analyze the reaction of up to 400 known genes to a specific substance.
STMicroelectroncis is also actively exploring the biochip area. Last fall, the company showed off a Lab-on-Chip prototype, based on Micro-Electro-Mechanical-Systems (MEMS) technology, that integrates both DNA amplification and detection.
Meanwhile, Philips Research is developing biochips based on magnetics, in contrast to more conventional biochips using optical or electronic analysis. Under the Dutch giant's scenario, target molecules that bind themselves to detector molecules on the chip are labeled by a nanometer-sized magnetic particle, according to Philips Research's Prins. The presence of the magnetic particle, then, can be measured with a special sensor based on the Giant Magneto-Resistive effect integrated at each detector location, he explained. Noting that the two most important qualities of biochips are "sensitivity and specificity," Prins said that the magnetic method can sense biological bindings even in their lowest concentration, while it can separate desired signals from unwanted background noises very well.
Despite a broad range of biochip activities-along with some notable breakthroughs among European chip vendors-it remains unclear whether the semiconductor industry will ultimately become the lead pioneer in the biochip market. Chip vendors' future role might still be limited to playing second fiddle to the pharmaceutical industry. Although chip companies are responsible for developing a platform for biochips, pharmaceutical companies own the actual bio-content. Clearly, successful entry to the biochip market for semiconductor companies hinges on access to bio-content through partnerships.
European chip vendors appear to share a very realistic view on this issue. Andrea Cuomo, ST's corporate vice president and general manager, Advanced System Technology, said that the biggest challenge in the biochip market might not be the technology itself, but rather, to "really understand the business." Noting that the two industries-semiconductor and pharmaceutical-don't even share a common terminology, Cuomo said, "Personally, I have big doubts if we will ever have a pure understanding of the medical world." Nonetheless, projecting a junction where genetics, biology, chemistry, physics, mathematics and mechanical science will meet electronics and semiconductors to create a brand new life science industry, Cuomo said, "We will do everything to be a part of it, because we want to make it happen."
Strong partnering
The key, however, is not in necessarily buying up or investing in what they see as critical for entry into the nascent biochip market-which was Motorola's approach during the prosperous 90s. Instead, "You start with what you know," said Cuomo. And "you have to be able to work extremely well with your partners."
Infineon's Klaue agreed. He said the pharmaceutical industry operates on a completely different set of margins, marketing approaches, and time-to-market principles from those of semiconductors. While it often takes 10 years to develop a new drug and another 10 years to test it, the semiconductor industry demands a less than 10-month turnaround for every new chip. At Infineon, rather than mergers and acquisitions, "We've been taking a step-by-step approach to determine where the safe place is for us to invest, to expand and to forge alliances," he said.
ST is spending only 5 percent of its research and development money in biochips, which translates into "a couple of million dollars," said Cuomo. ST, starting small by using "technologies we already know"-such as combining inkjet printer chip with electronic and fluidic elements-thinks the return on investment in the company's biochip development program is "doable in a few year years," he said.
Infineon, which already succeeded in delivering its first commercial biochip after just two years, is operating on an R&D budget "below $10 million," according to Klaue. |
