Technical Insight
SiC benefits from Title III effort
After working wonders for the US GaAs substrate industry in the 1990s, the Title III program has been attempting to do the same for SiC substrates, as Jon Newey describes.
For about 50 years now, the US Department of Defense (DoD) has had a very powerful legislative tool at its disposal that gives it license to nurture and develop domestic industries that it sees as crucial to its needs. This is known as the Title III program, which was borne out of the Defense Production Act passed in the 1950s. Title III of this act dealt with the ability to create, expand and maintain US domestic industrial capacity and supply of resources and technology for national defense. Responsibility for the overall management of the Title III program lies with the deputy under-secretary for defense for advanced systems and concepts. However, the US Air Force is responsible for the execution of individual projects within the program. Matthew Seaford is the manager of the Title III Program Office, which was established within the Air Force Research Laboratory Materials and Manufacturing Directorate, the director of which is Charles Browning.
At present, the Title III program consists of nine projects, worth about $75 million annually and covering an impressive range of technologies including structural materials, polymers, superconductors and semiconductors. The various programs have one common aim - to establish a production base that is self-sustaining after Title III involvement ends. This means that contractors are required to focus on business planning and marketing, as well as manufacturing capabilities. This combination brings long-term viability to industries seen as critical by the DoD.
GaAs: a big Title III successOver the years, the benefits to defense programs resulting from Title III contracts have been substantial. The most noteworthy success concerning compound semiconductors was the program covering semi-insulating GaAs substrates, which began in 1994. Back then, non-US firms dominated the global market for GaAs wafers and US producers were struggling financially, meaning that they lacked the financial muscle to invest in order to improve their products and market positions. The goal of the Title III program was to develop the domestic US capability for manufacturing high-quality, low-cost semi-insulating GaAs wafers in order to satisfy the growing military need for GaAs ICs. The program focused on improving yields, reducing production costs, and developing longer-term business and marketing plans.
The project was successful in turning around the fortunes of the US GaAs wafer industry. The dollar value of GaAs wafer sales of the three Title III contractors - AXT, Litton Airtron and M/A-COM - grew by approximately 300% during the project, despite a 40% reduction in the average price of GaAs wafers. However, Litton Airtron no longer exists, while AXT carries out most of its GaAs wafer manufacturing in China. The quality of the substrates being produced boosted the yields of MMIC device manufacturers, who in turn were able to lower their prices to the military, with GaAs ICs eventually finding their way into consumer electronic goods.
SiC comes to Title III s attentionIn 1999 the SiC wafer industry became the subject of a new Title III program, which aimed to emulate the runaway success of the GaAs wafer project. Three contracts were awarded to Cree, Sterling Semiconductor (now Dow Corning) and Litton Airtron, whose SiC activity is now part of II-VI Inc. In 1999 only 2 inch wafers of questionable surface quality were available, and the goal of the SiC program was to develop the US capability to produce high-quality 3 inch SiC substrates. With the project due to end in early 2004, 4 inch wafers are now being demonstrated, although they are not yet in widespread use.
When the project started in 1999, the situation with the US SiC suppliers was a little different from the one that the military faced with the US GaAs wafer industry in 1994. Cree was, and still is, the acknowledged world leader in SiC substrate production and also accounted for about 85% of the US market. The concern centered on the ability of Cree and its competitors to supply the large-area, high-quality substrates that would make the high-power devices required by the military available in volume and at a reasonable cost. The pull from the commercial semiconductor device industry that otherwise would have driven supply and quality was not great enough at the time for this to start happening.
"The semiconductor industry would not make the step of producing SiC devices unless it was viable for them to do so," said Seaford, explaining the origins of the Title III SiC project. "The substrate suppliers could not make the jump on their own of building capacity and quickly improving their products without the backing of customers." This is an example of the Catch-22 situations that Title III aims to address where it affects supply to the military.
While the Title III office recognizes the benefits of competition, it does not have a problem with a dominant domestic source (such as Cree) for any of its military needs, but the outcome of its programs must be a viable industry. "Having some competition drives improvements and increases viability," said Seaford.
Beyond substratesBy the time it closes, the SiC program will have cost about $10.7 million, with industry adding another $9.6 million from its own pocket and a further $2.9 million coming from other DoD organizations taking the opportunity to link into the Title III work for their own specific purposes. So, as the project nears its conclusion, can it be considered a success? Seaford certainly thinks so, citing the moves from the 2 to 3 to 4 inch substrates and the vast improvements in material quality over the last three years (figures 1 and 2; table 1). The fruits of the Title III SiC program are now getting early insertion into DARPA programs such as the Wide Bandgap Semiconductor Technology Initiative, which is making use of improved substrates to demonstrate devices that the DoD would like to see in its systems.
Title III is therefore building the industrial base in electronic materials (and the other technologies that it addresses) that the DoD needs to run its device-oriented programs. The Title III SiC program has successfully focused on substrates, but it is likely that its involvement will not end there. It is probable that SiC will be revisited by the Title III program in the future. Plans for a Title III wide-bandgap device initiative beginning in 2005 have been discussed, and Seaford is confident that these plans will be funded. "There are two or three companies that now know that they have a great technology and there are DoD program offices that want the technology," he said. "However, it will never become space qualified or reliability tested, and program offices don t have the funding to do it themselves. The companies are unlikely to go into mass production without guaranteed customers and so Title III may again step in to provide a way over that inertial barrier by purchasing the devices and field-testing them. Once two or three program offices accept the technology then you have the start of an industrial base."
The Title III SiC substrate program has attempted to follow the path that GaAs successfully took. Eventually the domestic commercial GaAs market took off and Title III was no longer needed. Now the DoD gets its GaAs devices and so does the commercial sector. "We consider the GaAs program a good blueprint that we intentionally tried to copy," said Seaford. "Nobody is in any doubt that GaAs changed society, and SiC has the same potential."
At present, the Title III program consists of nine projects, worth about $75 million annually and covering an impressive range of technologies including structural materials, polymers, superconductors and semiconductors. The various programs have one common aim - to establish a production base that is self-sustaining after Title III involvement ends. This means that contractors are required to focus on business planning and marketing, as well as manufacturing capabilities. This combination brings long-term viability to industries seen as critical by the DoD.
GaAs: a big Title III successOver the years, the benefits to defense programs resulting from Title III contracts have been substantial. The most noteworthy success concerning compound semiconductors was the program covering semi-insulating GaAs substrates, which began in 1994. Back then, non-US firms dominated the global market for GaAs wafers and US producers were struggling financially, meaning that they lacked the financial muscle to invest in order to improve their products and market positions. The goal of the Title III program was to develop the domestic US capability for manufacturing high-quality, low-cost semi-insulating GaAs wafers in order to satisfy the growing military need for GaAs ICs. The program focused on improving yields, reducing production costs, and developing longer-term business and marketing plans.
The project was successful in turning around the fortunes of the US GaAs wafer industry. The dollar value of GaAs wafer sales of the three Title III contractors - AXT, Litton Airtron and M/A-COM - grew by approximately 300% during the project, despite a 40% reduction in the average price of GaAs wafers. However, Litton Airtron no longer exists, while AXT carries out most of its GaAs wafer manufacturing in China. The quality of the substrates being produced boosted the yields of MMIC device manufacturers, who in turn were able to lower their prices to the military, with GaAs ICs eventually finding their way into consumer electronic goods.
SiC comes to Title III s attentionIn 1999 the SiC wafer industry became the subject of a new Title III program, which aimed to emulate the runaway success of the GaAs wafer project. Three contracts were awarded to Cree, Sterling Semiconductor (now Dow Corning) and Litton Airtron, whose SiC activity is now part of II-VI Inc. In 1999 only 2 inch wafers of questionable surface quality were available, and the goal of the SiC program was to develop the US capability to produce high-quality 3 inch SiC substrates. With the project due to end in early 2004, 4 inch wafers are now being demonstrated, although they are not yet in widespread use.
When the project started in 1999, the situation with the US SiC suppliers was a little different from the one that the military faced with the US GaAs wafer industry in 1994. Cree was, and still is, the acknowledged world leader in SiC substrate production and also accounted for about 85% of the US market. The concern centered on the ability of Cree and its competitors to supply the large-area, high-quality substrates that would make the high-power devices required by the military available in volume and at a reasonable cost. The pull from the commercial semiconductor device industry that otherwise would have driven supply and quality was not great enough at the time for this to start happening.
"The semiconductor industry would not make the step of producing SiC devices unless it was viable for them to do so," said Seaford, explaining the origins of the Title III SiC project. "The substrate suppliers could not make the jump on their own of building capacity and quickly improving their products without the backing of customers." This is an example of the Catch-22 situations that Title III aims to address where it affects supply to the military.
While the Title III office recognizes the benefits of competition, it does not have a problem with a dominant domestic source (such as Cree) for any of its military needs, but the outcome of its programs must be a viable industry. "Having some competition drives improvements and increases viability," said Seaford.
Beyond substratesBy the time it closes, the SiC program will have cost about $10.7 million, with industry adding another $9.6 million from its own pocket and a further $2.9 million coming from other DoD organizations taking the opportunity to link into the Title III work for their own specific purposes. So, as the project nears its conclusion, can it be considered a success? Seaford certainly thinks so, citing the moves from the 2 to 3 to 4 inch substrates and the vast improvements in material quality over the last three years (figures 1 and 2; table 1). The fruits of the Title III SiC program are now getting early insertion into DARPA programs such as the Wide Bandgap Semiconductor Technology Initiative, which is making use of improved substrates to demonstrate devices that the DoD would like to see in its systems.
Title III is therefore building the industrial base in electronic materials (and the other technologies that it addresses) that the DoD needs to run its device-oriented programs. The Title III SiC program has successfully focused on substrates, but it is likely that its involvement will not end there. It is probable that SiC will be revisited by the Title III program in the future. Plans for a Title III wide-bandgap device initiative beginning in 2005 have been discussed, and Seaford is confident that these plans will be funded. "There are two or three companies that now know that they have a great technology and there are DoD program offices that want the technology," he said. "However, it will never become space qualified or reliability tested, and program offices don t have the funding to do it themselves. The companies are unlikely to go into mass production without guaranteed customers and so Title III may again step in to provide a way over that inertial barrier by purchasing the devices and field-testing them. Once two or three program offices accept the technology then you have the start of an industrial base."
The Title III SiC substrate program has attempted to follow the path that GaAs successfully took. Eventually the domestic commercial GaAs market took off and Title III was no longer needed. Now the DoD gets its GaAs devices and so does the commercial sector. "We consider the GaAs program a good blueprint that we intentionally tried to copy," said Seaford. "Nobody is in any doubt that GaAs changed society, and SiC has the same potential."
