Next-Generation Sustainability Gets More Challenging

The semiconductor industry has made major progress on reducing energy usage and water consumption, and effectively abating its emissions, as companies made sustainability a core requirement in their design of new processes and tools. But it’s about to get considerably harder. That means more opportunities to add value with innovative technologies, and also more need for collaboration.

Next-generation technologies will need to handle and treat both significantly more material, and more different and more challenging materials, noted the experts gathered at the four-day Sustainable Manufacturing Forum at SEMICON West 2014. EUV lithography will use 4-5X more energy for its high power sources, and larger amounts of hydrogen that can’t be recycled, noted ASML’s Burt Planting, product safety manager, in the “Next Generation EcoFab” session of the Forum. Even without 450mm, companies are also starting to look at megafabs, with 100,000s of wafer starts a month, which means a whole new scale of material volumes and chemical flows, which will be as much of a challenge as 450mm, pointed out Gene Karwacki, director for strategic collaborations and integration, Air Products Electronic Division.

More process steps, finer features, and tighter thermal budgets will all require higher flows of gases, as would eventual 450mm wafers. Increased use of III-Vs, nano materials and many new and less well studied materials present new challenges for emissions and waste management. Fabs are also using more rare materials like Xenon and Helium that will need to be re-used to be economic. And while improvements can most easily be made when designing new tools and new fabs, improved practices will also have to be ported back to the installed base, on which many of the devices for the IoT will be made.

Regulatory requirements are also getting more complex, with regions introducing more new and differing rules, sometimes based on broader categories like machine tools and furnaces that may not make sense for semiconductor equipment. Others noted that analysts now ask about sustainability, stakeholders expect companies to quantify and take accountability for their full impact, and customers can still ask for 30 percent less energy use in tools. “Just as it doesn’t make sense to make a 7nm device if it costs twice as much, it also doesn’t make sense to make it if it takes twice as much water and energy,” commented Steve Moffatt, CTO of front end products at Applied Materials.

Opportunities for solving high-value problems
The first key issue is figuring out how to measure, model and benchmark consumption and emissions. “We can only manage what we can measure,” noted Moffatt. “EHS has been a very qualitative area, but we will have to change that.”

Another speaker at the Sustainable Manufacturing Forum, Hans Lebon, VP fab and process step R&D&M at imec, ticked off a long list of further strategies for smart fabs and equipment suppliers to develop technologies to distinguish their processes, most of which will also need new monitoring and measuring technologies:

• Better air handling and re-use of heat could save energy, as could reducing the thermal load on the tools. The subfab presents another opportunity to reduce energy use and nitrogen consumption by introducing a sleep or idle mode for subfab tools when not in active use.
• More water can be reclaimed and re-used, by using water from de-humidifiers, and by treating waste water for other uses. That will require new sensors to monitor water quality. Adding separate drain channels to process tools for different fluids for separate handling would also help. For example, the dirty water from the first rinse and the much cleaner water from the last rinse could be treated differently. Better still would be equipment that did wet processing in a closed loop process.
• Process chambers that use less gas, or use gases more efficiently are needed to reduce chemical waste. Benign chemistries would be better still, but when that’s not possible, then we need better ways to reclaim, recondition and re-use them, or at least better destroy them so they are not released into the environment. Though most H2 used is almost free of contamination, we as yet have no proven technology for reclaiming and monitoring it at an acceptable cost of ownership.

It will take a village
No company can effectively do all this on its own any longer. “We have to take into account the needs of the different tools when we design a process, which needs coordination among the players,” noted Lebon. Any realistic evaluation of sustainability also has to look at the whole picture — one advanced process may use more energy, but that could enable other processes to use less, or reduce the total energy consumed by the finished electronic product over its lifetime, for a net reduction overall.

The problems often are at the interfaces. SEMATECH’s recent collaborative effort on how best to handle the energetic materials — such as organometallics now becoming more common in the fab — found most of the leaks, chemical reactions, fires, and explosions occurred between process steps. Users did not have enough information on the materials and how they should be handled, while materials suppliers often had no idea of the range of applications and interactions with these materials or how they would be handled. “We need to get everyone together to discuss the risks and controls needed for new materials early on for the process chamber and vacuum pump design,” reported Steve Trammell, SEMATECH EHS manager.

The cooperative project investigated the root causes of some 70 recent incidents across chip makers. Many of these incidents were quite serious not only for the immediate danger but also from the submicron soot generated that could take weeks or even months to remove from the cleanroom. The group then developed a best known practices document for handing these materials, based on the collaborative input from all parties. Now the group is working with SEMI International Standards to get wider industry consensus from stakeholders and then plans to formalize the best practices in a SEMI standard. “This is certainly a priority for IBM and the gas suppliers,” Trammell noted. “The best practices published on the SEMATECH website include a letter signed by most major IDMs that these requirements will be included in their purchase orders.”

While the adoption of 450mm is delayed, the Global 450 Consortium of the five major chip makers who might build 450mm fabs continues to work together on the technologies needed for this transition, including measuring and monitoring fluids coming in and out at all process steps on the demonstration tools at CNSE Albany Nanotech. “This is an unprecedented collaboration among the top 5 high volume fabs,” said Adrian Maynes, M+W Group, currently program manager for the Facilities 450mm Consortium for sustainability, noting that they have also been tracking benchmarks for past technologies as well. The goal is to track actual usage to be able to correctly size the utilities to deliver just enough water and energy, since each 10% reduction in power usage saves about 2% on the total capital expense for the fab, he said. But such data also helps in things like matching the abatement capacity to the actual effluent flows. “With more benchmarks from the past, we know more than ever to help decrease risk. This is an unprecedented collaboration for data for a sustainable future,” he adds.

Industry also needs to work together to educate government officials. “It really opened my eyes recently when I talked to a former senator who told me that congressmen really need to be taught about the implications of their regulations,” said Karwacki. “We can’t expect government to understand the complexities we’re working with, so it’s our job to explain it to them. We need to do a better job of educating our representatives. We must work collaboratively with all governments for the regulations that are needed. SEMI does a good job of this.”

“What has worked well is a collaborative approach, such as for the reduction of PFC emissions, where research consortia and suppliers got together and committed to reduce emissions,” noted Leo Kenny, senior materials development engineer, Intel, who is currently overseeing that company’s programs for green chemistry and alternative assessment development. “The take away is that collaborative approaches have worked well in the past, and we need to start thinking now about using them early on [when designing the next generation tools and fabs].”

“Continuing advancements in semiconductor manufacturing technology present a wide range of further challenges in achieving the industry’s sustainability goals. These Sustainable Manufacturing Forums at SEMICON West bring the players together to work out the paths towards solutions,” said Sanjay Baliga, senior manager of EHS at SEMI.

For more information on SEMI, please visit http://www.semi.org. Upcoming SEMICON exposition information is available here: http://www.semi.org/en/EventsTradeshows/ExpositionsTradeshows. For information on EHS, visit http://www.semi.org/ehs or contact Sanjay Baliga at [email protected].