The electronics industry is ruthless. miss a beat, get behind the competition in introducing new products, and customers will treat you like yesterday's newspaper.

	ROAD MAP Smaller and faster electronic devices mean new chemicals and materials. JSR CORP. PHOTO

Electronics is also now Exhibit A when economists describe the current downturn. The sector is experiencing an unprecedented slowdown that would be a catastrophe if the economy as a whole were going through the same trauma.

Dan Tracy, a senior market analyst at the trade association Semiconductor Equipment & Materials International, says global production of semiconductor silicon wafers in the third quarter was down 43% compared with the third quarter of 2000. He expects production to be down 25% for the year. The market for photoresists used in semiconductors will likely shrink from $810 million last year to $660 million in 2001.

Yet, in the face of declining sales and profit, companies supplying chemicals to semiconductor and flat-screen display manufacturers are not cutting their R&D budgets. Tracy comments, "I have a sense that a lot of chip manufacturers are pushing new technologies that will give them the latest and greatest functionality for their chip or will help them reduce their manufacturing costs."

Masami Nakamoto, president of Sumitomo Chemical's fine chemicals sector, is blunter: "In this field, we cannot reduce R&D. Otherwise, there is no future."

The determination of electronic chemicals companies to maintain their R&D budgets looks particularly brave considering the hardship experienced by their customers at semiconductor and flat-panel display plants. Tracy says plants in Taiwan, where much of the world's memory chips are made, are operating at a fraction of their capacity.

Overall, spending on capital equipment by semiconductor manufacturers worldwide will be down 31% this year. One source tells C&EN that Hynix, South Korea's largest and most indebted memory chip manufacturer, is putting its staff on unpaid holidays lasting weeks as equipment sits idle.

Over the past year, prices for thin-film-transistor liquid-crystal displays (TFT LCDs) have tanked. Yoshimitsu Tsutsumi, an electronic materials sales and marketing manager at Clariant in Osaka, says flat-screen prices are down by half since last year, with 15-inch units down by two-thirds.

Tsutsumi says Clariant is the leading supplier of photoresists used in the manufacturing of TFT LCDs. With the exception of one factory in China, production of TFT LCDs takes place exclusively in Japan, Taiwan, and South Korea; there are no plants in North America or Europe. The market for electronic materials for semiconductors is 10 times larger than the market for electronic materials used by flat-screen manufacturers, Tsutsumi estimates.

Manufacturers of semiconductors and flat screens are accustomed to a succession of boom-and-bust cycles lasting two to three years. The so-called silicon cycle exists because prices drop rapidly once there is more than one producer to supply the latest chip or enlarged flat screen. With semiconductors, such introductions take place every 18 months or so.

Tracy isn't predicting when a turnaround can be expected in the electronics industry. He says the industry consensus is that there is no clear indication of when a turnaround will occur.

Seiichi Hasegawa, a JSR Corp. board member acting as director of the electronic materials business, says the turnaround will occur in the middle of next year at the earliest and no later than the end of the year. "This is the general industry outlook," he says.

THE DOWNTURN is more severe this time because the trough in the silicon cycle coincides with one of the worst economic downturns in memory. Meanwhile, over the past two to three years, production capacity for both semiconductors and flat screens has greatly increased, leading to oversupply.

The TFT industry is 10 years old, Tsutsumi explains. At first, all flat screens were produced in Japan. A few years ago, Korean firms LG-Phillips and Samsung entered the business. The latter has since become the world's largest producer in terms of units. And since last year, Tsutsumi adds, there are four producers in Taiwan.

Philip Rose, Asia-Pacific region business director for Rohm and Haas's Shipley unit, says his company foresaw the glut in the market several years ago and decided not to develop a wide range of electronic materials used to manufacture TFTs.

Pressured by low sales and profit, TFT manufacturers are squeezing their suppliers for price concessions. For Clariant, this a depressing challenge. It cannot develop lower cost photoresist formulations because such a change would be unacceptable to customers who would need to thoroughly test the new material before using it.

Only two possibilities to achieve lower production costs remain, Tsutsumi says. One is for Clariant to squeeze its own suppliers for lower prices. The other is to produce photoresists in larger and more economical batch sizes at plants in Shizuoka, Japan, and in Seoul.

The electronic sector's dire straits are occurring at a time of watershed innovations for the semiconductor industry. The shift to low-k dielectric materials and copper in the manufacturing of denser chips is prompting electronic chemicals suppliers to sustain, accelerate, or restructure their R&D efforts (see page 43).

The buyers of electronic materials--be they operators of flat-panel display plants or "foundries" where chips are made--are peculiar customers. Their cost structure is dominated by fixed expenses, mostly in the form of debt incurred in the purchasing of manufacturing equipment. In percentage terms, the sourcing of chemicals and other electronic materials from the likes of Shipley, JSR, or Cabot Microelectronics is comparatively small.

ANOTHER PECULIARITY of these customers is that they select their suppliers only when they are developing their production processes. Once the process is decided, the operators do not change it or the materials used until the plant is overhauled.

Electronic materials suppliers that fail to develop new generations of materials when foundries are qualifying suppliers will not be able to enter the market later. Tsutsumi notes that Clariant was a latecomer in the market for semiconductor photoresists, and that's why it chose to focus initially on the emerging market for TFT resists.

Sumitomo Chemical, Nakamoto says, was slow to develop deep-UV photoresists--used to create semiconductor circuit lines that are 0.13 µm wide--when foundries were qualifying suppliers. It had to cede the market, he says, to a newcomer, Shin-Etsu Chemicals. Shin-Etsu was not a player in the earlier generation g-line and i-line photoresists--used to make lines 0.25 µm and wider--where Shipley, Tokyo Ohka Kogyo, Sumitomo, and JSR are the dominant forces.

However, Shin-Etsu made a concerted effort to leapfrog its way into deep-UV photoresists, taking advantage of Sumitomo's overconfidence, Nakamoto acknowledges. He adds that, because Sumitomo does not have a chance in the deep-UV arena, it is developing next-generation "deeper UV" photoresists used to make lines as small as 0.10 Êm.

Another challenge that electronic chemicals suppliers face when introducing new products is the ballooning cost of R&D. Rose cites the example of deep-UV photoresists. He estimates the cost of developing a product in this category at $20 million and rising as the equipment needed gets more expensive. Total market demand, however, will be only $150 million. It takes a while to recoup R&D costs, so profitability depends to a large extent on the number of competitors, he says.

Echoing Rose, Hasegawa maintains that there are too many photoresist suppliers. "I think this recession is an opportunity for JSR," he says. "There are seven to eight players in the photoresist market now. Later, I think it will come down to three or four."

In a presentation at the SEMI Microelectronic Materials Strategy Symposium in San Francisco in September, J. J. Johnson, manager of Micron Technology's photolithography section, said photoresist manufacturers will have to form alliances with their customers or other suppliers in order to afford the rising cost of developing new products.

Looking past the current downturn, Shipley foresees a return to high growth for electronic materials. Shipley now supplies g-line, i-line, and deep-UV photoresists, along with antireflectants, developers, and strippers. The total value of these markets in 1998 was $820 million, the company estimates.

In 2008, the same markets will be worth $2 billion. But by 2008, Shipley expects it will also have developed new products for complementary markets that will be worth $2.2 billion by then. The emerging products consist of various types of low-k dielectrics, chemical mechanical planarization (CMP) slurries and cleaners, and copper.

Only emerging, the market for low-k material is now worth $6 million per year. By 2008, market consulting firm Kline & Co. estimates it will be worth $1.2 billion. Similarly, Merrill Lynch reports that Cabot Microelectronics, the top CMP slurry producer, expects its sales to return to a 50% annual growth rate within a year. Sales in the fourth quarter were flat.

Some suppliers of electronic materials do not foresee similar growth. T. V. Srinivasan, Tokyo-based executive vice president of HD Microsystems, says that, although the amount of material used per device is shrinking, the total market is still growing. He explains that HD's products--polyimide engineering polymers used in chip manufacturing--are applied on the silicon wafer through a spin-on process. HD is a joint venture of Hitachi Chemical and DuPont.

Over time, as customers refine their production processes, they reduce waste and use less polyimide per square inch of wafer surface area. Additionally, the size of uncut wafers is increasing, reducing the total need for polyimide because larger wafers are a more efficient way to produce the same number of chips. Offsetting this shrinkage, the market is growing from year to year because more chips are produced.

Although the industry is moving at a brisk pace overall, electronic material suppliers have to limit themselves to incremental improvements in their product lines, Srinivasan says, hampering "blue sky" research. Because of the high cost of setting up facilities, foundry operators are averse to risk when qualifying new materials. Srinivasan says: "The industry is very conservative in terms of changing materials. There must be reliable data. For new products, there is no customer base; it is a huge risk to pioneer a new market."

DuPont, the inventor of nylon, Kevlar, and Teflon, envisions the introduction of new materials and technologies that completely change the rules of the game. Not a player in the TFT industry, DuPont is instead pioneering electroluminescent displays, a technology that could deliver a severe blow to LCD manufacturers (see page 49). But for Shipley, JSR, and others, it remains business as usual. Despite the economic downturn, materials must be continually improved, whether doing so is expensive or not.

Top

NICHE OR NOT?

Focus Or Broad Range--Suppliers Differ In Product Strategy

Pick one: To be the best at one thing, or to offer buyers a broad range of products. Manufacturers of electronic materials differ in their business strategies and advance convincing arguments on both sides.

At one end of the spectrum is Cabot Microelectronics. The firm has an 80% share of the market for polishing slurries used to manufacture semiconductors and hard disks.

Dow Corning's electronic materials product offering consists of dielectric materials, and the company is perfectly satisfied with its narrow focus, says Chris Konitzer, marketing manager in the firm's semiconductor fabrication materials business.

Standing in sharp contrast are firms that offer a much broader product range. Shipley, part of Rohm and Haas, is representative of this approach. Under the banner of a concept it calls Mosaic, its product offering includes a range of photoresists, low-k dielectrics, electroplating materials, and ancillaries; its sister company Rodel provides CMP slurries.

JSR Corp. claims to offer a similarly broad line of materials under a concept it calls Total Solutions.

Other companies offer a broad but discontinuous range of electronic materials. Hitachi Chemical official Nori Imai says the company strives to "pinpoint niche products within the electronic chemicals market." Dow Chemical can provide customers with low-k dielectric materials and epoxy resins, but its product offerings are not integrated.

Sumitomo Chemical is developing very deep UV photoresists and is a leading producer of i-line photoresists, which are the most widely used type of semiconductor resists. The company is also a supplier of filters for thin-film-transistor liquid-crystal displays as well as high-purity gallium. But it does not offer related materials such as CMP slurries or first-generation deep-UV photoresists.

There are strong arguments in favor of the focused product range. Konitzer says Dow Corning's narrow range of products allows it to "stay ahead of the curve in terms of innovation and providing unique solutions to our customers." Similarly, Cabot believes its focus on CMP slurries enables it to maintain its position as the market leader.

But firms offering a continuous range of products are equally certain of their strategy. "There is a high synergy in having a broad range of materials," says Seiichi Hasegawa, a JSR Corp. board member acting as director of the firm's electronic materials business. He adds that because JSR offers both low-k dielectric materials and CMP slurries, it has an intimate knowledge of the interplay between the two.

Shipley's Mosaic is a "concept that is used to answer customers that are experimenting with different materials and processes, that are wondering how a change in one material affects other components of the process," says Philip Rose, the firm's Asia-Pacific region business director.

The Mosaic concept partly explains why Rohm and Haas made several acquisitions in the past two years. Rose also believes that only large players will be able to afford ballooning R&D costs as electronic materials become increasingly sophisticated.

Air Products & Chemicals is another company that believes in the broad product offering. Joseph Stockunas, worldwide marketing manager for electronics, claims that the company--with its gases, chemicals, and gas-delivery systems--has a broader offering than almost any other.

Stockunas acknowledges that some customers fear becoming "hostage" to a dominant supplier. However, he contends that most believe strong partnerships are worth it. "The companies that make the most money are the ones that get the product to market first and fastest," he says. "As a result, there is an acceptance of working with suppliers who can really bring solutions to market faster."

Size can also be an advantage in providing technical support to large semiconductor manufacturers. Electronic-component manufacturers are increasingly welcoming assistance from their materials suppliers in recovering or recycling toxic chemicals. Taiwanese firms are particularly keen on this assistance, given the critical shortage of government-approved toxic waste processors on the island.

Wide but discontinuous product lines are a little more difficult to justify. Sumitomo Chemical's product range derives from its R&D capabilities, says Masami Nakamoto, president of the company's fine chemicals sector. Thus, the know-how to manufacture optically functional films for flat screens derives from the company's long involvement in polyolefins. Similarly, Sumitomo's photoresist business derives from the company's broad expertise in synthetic chemistry, particularly for agrochemicals.

"The 'best' product mix strategy does not actually exist," comments Jeff Kinnicutt, director of Honeywell Electronic Materials' dielectric materials business. Customers are most concerned about the total cost of use and ownership of their production process. Companies that provide multiple materials can help customers optimize materials for use together and also assist customers in minimizing their R&D expenses. On the other hand, materials such as encapsulation resins and dielectrics perform independently. A materials supplier that can provide both cannot tell a customer that the two can be optimized for best performance.

The jury is still out on whether focus or broad product range is the way to go when supplying electronic materials. In a recent report, Merrill Lynch analysts called Cabot a long-term buy and predicted that the company's sales will leap forward as use of CMP slurries surges. In contrast, firms like JSR, Air Products, and Shipley believe in size. As R&D costs continue to rise, Shipley's Rose asserts, "you need to be big."

Top

MARKET SHARE

Polyimide Versus Polybenzimidazole

Executives at Clariant Japan firmly believe that their polybenzimidazole is superior to DuPont's polyimide. However, they face formidable hurdles in convincing customers.

Among the many products that the semiconductor and flat-display industries purchase from chemical companies are engineering polymers. DuPont's Vespel-brand polyimide is used in lieu of metal to hold silicon wafers during semiconductor production. Polyimide does not conduct electricity and has high thermal resistance, two essential qualities for containers used in semiconductor manufacture. The main competitor of polyimide has traditionally been ceramic.

Like polyimide, polybenzimidazole was invented by the National Aeronautics & Space Administration in the 1960s, says Noriyuki Niwa, manager of Clariant's life sciences and electronic chemicals division in Japan. Polyimide was licensed to DuPont, which commercialized it in the form of films, fibers, and engineering plastics. Hoechst Celanese licensed polybenzimidazole, a material made from aromatic bis-o-diamines and dicarboxylates. Niwa says Celanese added lithium chloride to impart stability and marketed the product mainly as a fireproof fiber, under the trademark PBI.

Ten years ago, Niwa says, what was then known as Hoechst Japan agreed to purchase polybenzimidazole from what was then Hoechst Celanese. Hoechst companies traditionally did not share patents. Hoechst Japan proceeded to develop and mold stock shape parts. Clariant Japan is still the only patent holder for this process, Niwa says.

Only recently, Clariant decided to seriously market these parts to makers of semiconductor fabrication equipment. The opportunity is far from insignificant: Niwa says the market for high-heat-resistant molded plastic amounts to $100 million in Japan, up to $300 million in the U.S., and $100 million in Europe.

Niwa believes that polybenzimidazole has significant advantages over polyimide. For one, it is a thermoplastic, whereas polyimide is a thermoset. When heated, polyimide has the undesirable tendency to emit polyamic acid, Niwa claims, and as a result doesn't maintain its purity when molded into a stock part.

Vespel product specialist Shin-ichi Nakagawa counters that polyimide, which is made from polyamic acids, does not revert back into or emit polyamic acid when heated. "We see only an increase in part integrity when molded or heated," he says.

Clariant Japan will have difficulty gaining customer acceptance. Niwa admits that DuPont's share is over 90% in the markets Clariant will target. Moreover, Clariant is not allowed to market the product outside Japan; in both Europe and North America, Celanese and Clariant have agreed that other companies have exclusive marketing rights for polybenzimidazole stock shape parts.

[Previous Story] [Next Story]

Chemical & Engineering News
Copyright © 2001 American Chemical Society