JULY 7, 2009
Category Archives: Materials and Equipment
July 13, 2009 – After nearly a year in the red, semiconductor equipment manufacturers finally broke back above the “parity” mark for bookings vs. billings an indicator of future business stability and growth.
Worldwide bookings rose to $25B in June, up 40% vs. May, and billings rose 31% sequentially to $2.4B, for a book-to-bill ratio (B:B) of 1.01 — meaning $101 worth of product orders were received for every $100 recorded in sales. As other market watchers have noted, though, the levels though appearing stabilized are still well below where they were a year ago; bookings -41% from June 2008, and billings down -57%.
Nevertheless, “business activity is beginning to improve,” particularly for backend suppliers boosted by “soaring utilization rates at the subcons,” VLSI Research said in a statement. On the frontend of production lines, investments are technology-driven (vs. capacity additions), supporting a ramp of 3Xnm and 4Xnm node process technologies. The firm also cites a push in orders from the long-dormant memory sector, thanks to a transition to DDR3 memory.
by Howard Lovy, contributing editor
Like nanotechnology in general, carbon nanotubes (CNT) are seen — depending on the industry you’re in and your world outlook — as the enabling technology for cleantech or a possible environmental and health hazard. They are viewed as a creator of jobs and of new market opportunities. They are optimistically seen as an engine for an elevator to the stars and a lift out of economic recession.
They can, of course, already be seen in a few products, like tennis rackets and bicycles, but nothing on the scale that yet match the hype that has surrounded them in the decade and a half since they were first produced. However, if you take a look at some of the companies, new and old, that are developing new ways of making and applying CNTs, you might cautiously conclude that nanotube prosperity is right around the corner.
Here, then, are a few snapshots of companies that are betting the farm on nanotubes, trying their luck at the little miracle workers in a big way.
Bayer MaterialScience
In Chempark Leverkusen, in Germany’s Rhineland region, Bayer MaterialScience is attempting to conquer the market for nanotubes, which it estimates will grow at an annual rate of 25%. In 10 years, the German chemical and pharmaceutical giant believes annual carbon nanotube sales are expected to reach $2 billion. The company is preparing now for this expected windfall by building what it is calling the largest nanotube factory in the world, investing around $29 million on the project and creating about 20 jobs. The project adds to a pilot plant with an annual capacity of 60 tons that has been in operation in Laufenburg in southern Germany since 2007.
Bayer MaterialScience claims to be one of the few companies that can produce carbon nanotubes of consistently high quality on an industrial scale. “Bayer is investing in this, the world’s largest CNT production plant, because we are convinced of the technological and economic efficiency of the process,” said Wolfgang Plischke, a member of the Bayer AG Board of Management responsible for innovation, technology and the environment.
Bayer is not alone. Its efforts have the backing of more than 80 partners from industry and science, which have joined together to develop new technologies and applications for carbon nanotube-based materials. The effort is part of the “Innovation Alliance Carbon Nanotubes” (Inno.CNT), created with the support of Germany’s Federal Ministry for Education and Research.
Baytubes production process illustration. (Source: Bayer MaterialScience)
What Bayer is doing is important for a couple of other reasons beyond the mere scope of the project. While mainstream media coverage of nanotubes has focused primarily on what is not known about the environmental and health effects of nanotubes, Bayer is focusing its efforts on what is known about them — they help forge a path toward greener technology. As part of a companywide push to develop sustainable technologies, Bayer is working on membranes to produce fresh water through seawater desalination.
And in December, the US Environmental Protection Agency gave Bayer MaterialScience regulatory approval to sell its multiwall carbon nanotubes — what it calls Baytubes — in the United States.
BayTubes were developed through a collaboration between Bayer Technology Services and Bayer Material Science to develop a cost-effective production process for CNTs that paves the way for their industrial application. Baytubes make plastics not only electrically conductive, but also very stable and strong, while keeping the material extremely lightweight, the company says. These improved properties are already being put to use today in the production of various sports goods, such as ski poles and baseball bats.
Nanocomp
Nanocomp Inc., based in New Hampshire, is answering the call from the aerospace industry for materials that are lightweight and strong with high thermal and electrical conductivity. Nanotubes seem a natural for this purpose, and the US Air Force seems to agree, granting the company a number of contracts under the federal Small Business Innovation Research (SBIR) program.
With Nanocomp’s CNTs as a test case, the Air Force is hoping to finally replace copper wiring with nanotubes, which are much lighter and harder to break down. Nanocomp will use this funding to advance the suitability of nanotube-based material for a number of aerospace applications, ranging from thermal management and electromagnetic shielding to electrical and power generation system enhancements.
Nanocomp production furnaces. (Source: Nanocomp Inc.)
The first SBIR award builds upon Nanocomp’s successful demonstration, accomplished under a Phase I contract awarded in early 2008, of the use of lightweight conductive wires made from CNTs. During Phase II, Nanocomp will work toward optimizing processing and manufacturing methods to produce CNT wiring in the quantities and forms required for direct integration into aircraft electric power applications.
The Air Force awarded Nanocomp a second SBIR contract to develop carbon nanotubes as a viable substitute for nickel-based conductors in electrostatic discharge (ESD) and electromagnetic interference (EMI) shielding applications. This Phase I research has been designated as a “critical program,” indicating that the government places a high degree of importance on the research. The goals of this program are to optimize the properties of CNT sheet materials to meet shielding requirements, develop a process to integrate the mats into existing commercial EMI/ESD shielding systems, and develop on-line production quality-control methods.
Unidym
Unidym, through acquisition, is the company that ended up inheriting the groundbreaking work of the late Rick Smalley, one of the original discoverers of the carbon nanotube. Unidym is owned by Pasadena, CA-based Arrowhead Research Corp., which has its hands in many different types of nanotech companies. It appears to have pinned its highest hopes, however, on Unidym’s ability to turn its deep nanotube intellectual property into business success — Arrowhead recently upped its ownership in Unidym from 51.5% to 58.1%. Still, it was not able to stop burning cash and aborted a planned IPO, and laid off half its staff in Houston.
Still, Unidym plans its first product release in the second half of 2009: rolls of its carbon-nanotube-coated plastic films. The transparent, conductive films could make manufacturing LCD screens faster and cheaper, and enhance the life of touch panels used in ATM screens and supermarket kiosks. They might also pave the way for flexible thin-film solar cells and bright, roll-up color displays for cell phones, billboards, and electronic books and magazines. The company has said that Unidym is already working with leading touch-panel makers, and recently announced a year-long joint development partnership with LG Display.
Carbon nanotubes dispersed in polycarbonate, which is highly electrically conductive. Formed into an enclosure for sensitive electronics, this plastic protects the internal circuitry from external radio-frequency interference, and prevents escape of radio-frequency interference from within the enclosure. (Source: UniDym)
To further Unidym’s push into electronics and displays, it formed an important partnership with Continental Carbon Company (“CCC”), a global provider of carbon blacks, conductive blacks, and carbon nanotubes, to take over its bulk CNT business. The agreement is the latest in a series of steps taken by Unidym to accelerate the development of its CNT technology for application to the electronics industry and reduce its cost structure. It is intended to provide Unidym and its partners with a reliable and scalable source of high quality CNTs by leveraging CCC’s deep experience manufacturing large quantities of carbon materials. In addition, it will allow Unidym to capture additional value from its extensive CNT patent portfolio by enabling CCNI to focus on end-markets that are not core to Unidym’s business.
The market
Other companies, such as Nantero, are continuing their quest to make nanotubes an integral part of future electronics — in Nantero’s case, nonvolatile memory. Others, like Cheap Tubes Inc., are going for bulk (as their name implies). And while some inroads have been made, much of the promise of nanotubes remain just that — a promise. Some analysts have said that nanotubes are a promising technology in search of a market that has largely been uninterested until now. Other analysts are more optimistic — the Freedonia Group projects the total demand for carbon nanotubes is expected to be about $1 billion by 2014, up from $6 million in 2004. Still others are cautious, as environmental and health regulatory agencies worldwide weigh in on the “unknowns” about environmental and health effects.
So, companies like Bayer, Unidym, and Nanocomp choose to remain optimistic that, during a time when nothing is certain, there is at least a certain market waiting for those who know how to find the right formula to produce and market this nanomaterial of the future.
Howard Lovy has been covering nanotechnology since 2001. E-mail: [email protected].
by Finlay Colville, Coherent, Inc.
July 9 2009 – Next-generation c-Si cell production tooling is finally poised to make an impact within the industry.
Production equipment for c-Si cell manufacturing is slowly adjusting to changes required for next-generation cell types. Three different c-Si cell concepts are forecast, each with unique tooling to enable efficiency- or cost-prioritized manufacturing. Championed during the 1980s and 1990s, high-efficiency concepts using non-contact laser-based tooling are seeing renewed focus, as market demand hastens their long-awaited entry.
Gloom and despair was clearly in evidence down through the supply-chain for c-Si cell manufacturing equipment earlier this year [1]. With the transition from healthy equipment order backlogs to one of near-empty production floors indeed being a wake-up call for most concerned. The impact however was not simply related to short-term bookings forecasts, but heralded an inflection point that looks set to redefine production equipment technologies once cell capex levels return to their halcyon days of the mid-2000s.
This article explains how changes in market supply-demand dynamics [2], coupled with less favorable financing conditions brought on by the current credit crunch climate, are together redefining the type of c-Si production equipment required by cell manufacturers over the next few years. The adoption of laser-based process tools — long considered a luxury by c-Si cell manufacturers — is used to illustrate the changes afoot, and the trends predicted for representative technologies now considered essential within specific tracks of proposed cell roadmaps.
Supplier’s market hinders progress
Historically, production technology utilized by cell manufacturers has been dictated by the supply-demand dynamics prevalent at the time. To understand this better, first let’s define relevant time periods. Throughout this article, comparison will be made directly to industry phases and assigned nomenclature, as borrowed from Stephen O’Rourke’s market-leading commentaries [2]: initial growth (2003-2008); shake-out (2008-2012); and build-out (2012-2020). The final category of maturity (beyond 2020) is omitted here, partly on account of the rate-of-change in production technologies anticipated during shake-out and build-out phases. Rather, an additional phase is introduced prior to initial growth, here labeled ‘R&D/Technology,’ to help understand the equipment utilized extensively prior to 2003 — and how this embryonic phase largely set the scene for production technology that would be applied widely during initial growth.
So, what about the implementation of laser-based tools within c-Si cell production lines during these time periods? It’s a mute point that, had search engines been around prior to the mid-80s, a likely candidate for most-hits when looking for matches between ‘lasers’ and ‘solar’ may well have been Ray Davies’ mid-60s chart-topping hit ‘Sunny Afternoon’ [3] when he calls out for “lazing on a sunny afternoon.” Certainly not as essential components for efficiency-enhancement above the 17% level. For the R&D/Technology phase was more about possibilities than a concerted effort to introduce non-contact production-line equipment offered by laser tools.
Prior to 2003, laser-assisted cell enhancement stages were confined mainly to the research activity at the University of New South Wales [4] (UNSW), Sandia Labs [5], European-funded projects such as LOWTHERMCELLS [6], and within the Fraunhofer-ISE [7]. With the focus then on R&D achievements (the strength of the academic community) and proof-of-concept as an indication of technical prowess, there was no fundamental driver to take these processes from lab to production. Nor was there an established equipment supply-chain willing to invest the time and resource to qualify laser-tools for production use. To compound matters, suitable industrial-grade laser-based tools of the late 1980s and 1990s were not aligned with cell manufacturing needs.
Ultimately, the lack of any appreciable market-pull (demand) stood as the underlying barrier to laser-tool adoption as standardized equipment. Interestingly, high-profile was afforded to the flagship c-Si cell of the 1990s, the laser grooved buried contact (LGBC) concept [8]. Originating at UNSW [4] and licensed by BP Solar, their Saturn lines were pioneering in every aspect [9]: efficiencies greater than 17% from standard c-Si cells with laser-prepared grooves, double-diffusion for selective emitter formation, and electroless plating. Just two things really were out of sync — the time period and the lack of a tangible market demand.
As initial growth kicked in, the order of the day was ramp-up of capacity to meet the demand originating from subsidies in Japan, Germany, Spain [10]. Equipment selection was dominated by tooling then qualified and readily available, and when configured as a complete production line, could output standard c-Si cells with efficiencies at the 13%-15% level, well below the benchmark set by research labs (UNSW PERL cell at 25% [11]) or the Saturn lines (over 20% demonstrated [12]). So, moderate-efficiency standard cell types became the mainstay of c-Si output during this time period, and as demand outstripped supply, the goal seemed simple enough — how quickly could additional lines be commissioned? For even modules based upon these cells at the $5/W mark were readily sold [13]. Thin-film competition from the likes of First Solar [14] had not fully matured to the dominant level it is today, and alternative thin-film technologies based upon a:Si seemed to be more biased towards announcements of fab orders than fully-commissioned capacity levels [15].
What did however emerge was the idea of a high-efficiency (HE) c-Si cell concept [16], but not based upon low-cost p-type silicon. Rather, the technologies of Sunpower [17] (interdigitated back-junction, or IBJ) and Sanyo [18] (heterojunction with intrinsic thin-film, or HIT) used high-grade n-type silicon and production-line equipment not ideally matched to the lowest cost means of cell production. Differentiation between these HE cells and standard c-Si cells is certainly influenced by the priority afforded to cell-making revenues, when this stage is isolated within the value-chain; the comparison most evident between pure-play cell makers and vertically-integrated companies more focused on selling energy further downstream. (Note: The back-contact cells of Advent Solar 19], based upon the Emitter Wrap-Through (EWT) cell concept, can also be included within this definition of high-efficiency.) However, a clear divide was established between cells manufactured either for the lowest cost (LC) or the highest efficiency (HE). Similarly, production line equipment availability and the technologies used therein also fell into these two categories.
Laser-based production line equipment certainly had a hard task getting into standard cell production lines during this time period. The HE cells — especially LGBC production and initial EWT lines [19] — were using laser-tooling as an enabling technology [20], but accounted for a very small percentage of production output worldwide. Standard c-Si cells did come to rely upon laser tools to perform the junction isolation step [21], but in the mid-2000s, plasma-etching very much led the way here on account of low capex and (low-tech) alignment with manual wafer handling (especially in Asia). Understanding however that junction isolation is really a loss-preventative step, and not an efficiency-enhancement enabler, puts some perspective on what many regarded then as the ‘big’ opportunity for laser-based tools within c-Si lines [21].
Shake-out to the rescue
Something had to change the landscape. Enter the events of the past 12 months which have been well documented within the press recently, with the label shake-out applied mainly to consolidation of cell or panel producers expected to emerge as winners over the next few years. Shake-out also relates to a lesser degree to the different cell technologies (c-Si, thin-film types, or wildcard entrants). But shake-out also applies to the production equipment used for cell manufacturing. And with process steps for c-Si cells far more standardized compared to many of the thin-film technologies promoted today, the impact of shake-out here will be more pronounced and immediate.
Figure 1. Bottom-up analysis of historical and projected c-Si capacity added per annum. Forecasting of c-Si cells with efficiencies exceeding 17% is borrowed from the categorization proposed in Ref. #16. Actual production outputs can be estimated by combining the data above with recent analyses showing fab-utilization rates [1] and capex [35]. (E) estimate; (F) forecast.
Projected cost analyses of the various cell technologies at the levelized cost of energy (LCoE) point are differentiating again two discrete tracks specific to c-Si cell production [2], necessary to remain competitive with thin-film approaches in the medium-term. While other factors clearly come into play here (no more so than raw material costs [22]), the two routes for c-Si cell production (based upon HE and LC cell types) each have their own roadmap for competitiveness [2]. In fact, there are really three tracks opening up for c-Si cell production defined as:
- Ultra-HE cell types based upon high-grade silicon (typically n-type), characteristic of the IBJ and HIT cell types, with efficiencies exceeding 20%.
- A HE version of the standard c-Si cell, ideally based upon multicrystalline-Si (mc-Si), and providing efficiencies in the 17-20% level.
- A LC version of the standard c-Si cell, with comparable efficiencies to current cell outputs at the 14-16% level, but where tool capex and operating costs are significantly lower than used in production today.
It is therefore this differentiation, somewhat courtesy of the current shake-out phase of the industry, which forms the basis of the new production-line equipment and processes being forecast now for the next few years. Figure 1 (above) captures the projected trends out to 2013 for each of these c-Si cell types, with respective capacities added each year. The growth of high efficiency cells over this time period is illustrated in Figure 2 (below), which compares the cumulative installed capacity at year-end 2008 and 2013.
Supply chain falls into line
Fortunately, the equipment supply-chain has been primed for this change; a glance at the patent filings and license agreements over the past decade from some of the research labs flagged up earlier in this article backs this up. Further, recent EU-funded projects such as CrystalClear [23] and SOLASYS [24] are full of next-generation cell concepts awaiting industrial technology-transfer (much in the same way that the UK DTi-funded project Alpinism [25] between Exitech, Ltd. and BP Solar hastened industrial-grade laser tooling for the Saturn lines). It’s largely through these consortia that new laser-based processes have been illustrated, set to form part of new cell production stages to enhance the efficiency of c-Si cells.
Figure 2. With the introduction of high-efficiency c-Si cells from 2008, the proportion of installed c-Si capacity by year-end 2013 comprised of standard cells will decrease from ~90% to ~60%.
Opportunities certainly exist for new technology (including laser-based tools) to assist in next-generation cells based upon the ultra-HE cell types, but the fundamental driver there may be reduced capex and lower operating costs, rather than an increase of a few percent in cell efficiency. And scope for radically new approaches for LC cell production, possibly from reduced silicon consumption (e.g., sub-100μm thick cells [26]), may well see the introduction of other new tooling there. But the main driver for new equipment tooling and processes in the short-term is clearly track 2 above — the HE version of the standard c-Si cell concept [16]. Here, immediate incremental improvements can be made via upgrading existing lines or during new capacity expansions. The goal for now being efficiency-enhancement with minimal change to production capex. But at last, a clear route-in for laser-based process tools at key enabling production stages, with a well-defined long term market pull.
The various laser-based steps for high-efficiency cells have been summarized recently in a number of review articles [20, 27, 28]. Without getting too much into technicalities, laser-based tools can be categorized [29] as:
- Assisting in mask writing [30] (for subsequent metallization, secondary diffusion in selective emitters, or surface etching to texture);
- Localized secondary diffusion [31] (laser doping via phosphorous- or boron-contained precursor layers);
- Contact preparation (finger grooves, through-silicon-vias [32], interdigitated structuring [33]);
- Contact forming (sintering [34] or firing [7]).
When looking at the various schemes being considered today, it is indeed a testimony to the endeavors of the UNSW researchers and BP-Solar technologists that many of the efficiency-enhancement steps pursued within the current crop of HE cell candidates share so many of the concepts inherent to the original LGBC cell [8]. Only now, production technology has matured somewhat, there exists a buoyant equipment supply-chain eager to participate, and the timing for higher finesse variants being introduced to the market is just right!
Conclusion
Having waited in the wings for the better part of twenty years, and having patiently watched cell production line capex grow with CAGRs in excess of 40%, next-generation c-Si cell production tooling is finally poised to make an impact within the industry — thanks mainly to the market dynamics of the new shake-out phase. As the leading cell manufacturers start to announce average c-Si cell efficiencies on mc-Si in excess of 17%, others will be forced to follow with similar high efficiency claims or choose to pursue a lower cost approach to cell manufacturing. With the implementation of high-efficiency c-Si cells directly linked to advanced laser processing within new production lines, the next few years will see increasing reliance on laser tools as they account for a greater fraction of new cell capex releases. (A follow-on feature by the author will address additional factors required for successful laser tool adoption and the timing of the various new process steps into mainstream production.)
Finlay Colville received his BSc in physics at the U. of Glasgow in 1990 and PhD in laser physics at the U. of St. Andrews in 1995, and is director of marketing for solar at Coherent Inc., 5100 Patrick Henry Drive, Santa Clara, CA 95054 USA; ph +44-7802-238-775; email [email protected]; www.Coherent.com/Solar.
References
1. P. Mints, “The global economy fell down and went boom — will solar follow?,” Jan. 2009.
2. S. O’Rourke, “Solar PV Competitiveness and Market Dynamics,” Keynote session at 34th IEEE PVSC, Philadelphia, 2009.
3. R. Davies, “Sunny Afternoon.”
4. S. Wenham, M. Green, “Buried Contact Solar Cells,” Australian Patent 570309, 1985.
5. D. King et al., “Development of a Multi-purpose, Pulsed-laser System for Solar Cell Processing Applications,” 21st IEEE PVSC, Orlando, 1990.
6. L. Pirozzi et al., “Innovative Applications of Laser Technology in Photovoltaics,” Proc. SPIE Conf. on Laser Applications in Microelectronic & Optoelectronic manufacturing II, San Jose, 1997.
7. E. Schneiderlöchner et al., “Laser-fired Contacts,” 17th EUPVSEC, Munich, 2001.
8. F. Colville, “Into the Groove: How Buried Contacts Brought Lasers to Life in Solar,” InterPV, June 2009.
9. N. Mason et al., “The Technology and Performance of the Latest Generation Buried Contact Solar Cell Manufactured in BP Solar’s Tres Cantos Facility,” 19th EUPVSEC, Paris, 2004.
10. P. Mints, “Photovoltaic Industry 2009: a Journey into Uncertainty,” Photovoltaics International, 4, 2009.
11. J. Zhao et al., “24.5% Efficiency Silicon PERT Cells on MCZ Substrates and 24.7% Efficiency PERL Cells on FZ Substrates,” Prog. Photovolt: Res. Appl., 7, 1999.
12. N. Mason et al., “20.1% Efficient Large Area Cell on 140 micron Thin Silicon Wafer,” 21st EUPVSEC, Dresden, 2006.
13. P. Mints, “Solar tech sets sales record — just as demand slows,” May 2009.
14. M. Osborne, “Analyst Predicts First Solar to Become Largest Solar Module Manufacturer in ’09,” June 2009.
15. P. Mints, “As Demand for Solar Tech Deepens, Where Do Thin Films Stand,” June 2009.
16. N. Mason, “High-Efficiency Crystalline Silicon PV Cell Manufacture: Status and Prospects,” PVSAT-5, Glyndŵr, 2009.
17. “Sunpower Announces World-record Solar Cell Efficiency.“
18. “Sanyo Develops HIT Solar Cells with World’s Highest Energy Conversion Efficiency of 23.0%.“
19. “Ventura Solar Technology.”
20. F. Colville, et al., “Existing and Emerging Laser Applications within PV Manufacturing,” Photovoltaics International, 1, 2008.
21. F. Colville, “Lasers Scribing Tools Edge in Front,” Global Solar Technology, Vol. 2, No. 2, March/April 2009.
22. M. Thirsk, “Opportunities and Challenges for Materials Makers in the Global PV Market”, 1st Annual SEMI Silicon Valley Photovoltaic Forecast Luncheon, Santa Clara, September 2008.
23. “CrystalClear: the Next Generation in Crystalline Silicon Technology.”
24. “SOLASYS: Next Generation Solar Cell and Module Laser Processing System.”
25. N. Mason, J. Fieret, “Advanced Laser Processing for Industrial Solar Cell Manufacturing (Alpinism),” 2006.
26. A. Skumanich, “Advanced Ultra-thin Wafer Processing: Challenges and Recent Developments,” Challenges in crystalline silicon manufacturing, Intersolar, San Francisco, July 2009.
27. F. Colville, “Laser Scribing Exposed: the Role of Laser-based Tools in the Solar Industry,” Photovoltaics International, 3, 2009.
28. F. Colville, “Laser Processing Enables High-efficiency Silicon Cell Concepts,” Photovoltaics World, 1, March/April 2009.
29. F. Colville, “High Efficiency Roadmap Demands Clear Dialogue for Laser Adoption,” In print, InterPV, September 2009.
30. F. Colville, “Selective Criteria: Lasers Go Short for Dielectric Ablation of Silicon Solar Cells,” Solar: A PV Management Magazine, Issue 2, 2009.
31. F. Colville, “Laser-assisted Selective Emitters and the Role of Laser Doping,” In print, Photovoltaics International, 5, 2009.
32. F. Colville, “The Hole Story: Lasers take the Wrap,” InterPV, April 2009.
33. A. Schoonerdbeek et al., “Laser Technology for Cost Reduction in Silicon Solar Cell Production,” Proc. of 69th Laser Materials Processing Conference, ISBN 4-947684-70-4, 2007.
34. T. Rublack et al., “Evaluation of the Laser Melting Process of Different Materials for the Front-side Metallization of Silicon Solar Cells,” 23rd EUPVSEC, Valencia, 2009.
35. “Solar PV CapEx Will Ease Capacity Growth in Time for Recovery,” IC Insights Research Bulletin; June 2009.
This article was originally published by Photovoltaics World.
by Dr. Paula Doe, contributing editor
July 8, 2009: The much-hyped properties of materials at the nanoscale are finally starting to be applied to some real electronics applications, ranging from near-ideal thermoelectric material based on spray-on semiconductor nanocrystals, to transparent conductive films made from carbon nanotubes and self assembled silver nanoparticles, to ultrasensitive nanoscale MEMS gas sensors.
Nanoscale materials properties are enabling efficient, low-cost thermoelectric materials. Though long studied, thermoelectric conversion of heat to electricity have never been efficient enough to be practical for most applications. “They were stuck with natural materials,” says Evident Technologies CEO Clinton Ballinger. “But with nano-structured materials you have a lot more materials to work with. You can change the thermal properties.” That means it is possible to design something that approximates the ideal thermoelectric material, conducting electricity well but not heat. Modeling suggests that the most efficient structure would be point sources of excited electrons distributed evenly in a matrix — and that ideal efficient material can be approximated quite well by a low-cost solution, using colloidal ink containing semiconductor nanocrystals to create a bulk material while retaining the nano properties.
Ballinger suggests that some of the first markets for this technology will be in the semiconductor industry, where it could enable efficient, flexible, solid-state cooling for integrated circuits and LEDs. This could greatly reduce the size or need for a heat sink, he argues, and potentially improve performance. “Right now we’re just spraying it on with an airbrush,” he notes. “So it could likely be coated right on the chip for thermoelectric cooling.”
First application is likely to be for less sophisticated solid-state cooling, though, such as spot cooling for things like wine coolers. But eventual markets for low-cost roll-to-roll coated thermoelectric films could also include waste heat recovery in automobiles and central power stations, general heating and cooling, and even power generation.
Likely closer to market are transparent conductive films using innovative nanomaterials to potentially challenge ITO. Unidym is sampling a transparent conductive film based on carbon nanotubes for touch panel displays and readying production capacity. CimaNanotech is similarly sampling its flexible film product based on self assemble of silver nano particles, for which Toray Advanced Films is the production coating partner.
Pushing MEMS to the nanoscale opens up new potential as well. “The advantages go beyond scaling,” says Caltech professor Michael Roukes, whose lab has been driving developments for the last 15 years. “The physics scales in a profound way.” This means MEMS-based detectors in an electronic nose can be made significantly more sensitive, as well as scaled down in size by about a million fold, compared to the existing state-of-the-art — and made with efficient wafer-scale processes.
Roukes’ lab and CEA-Leti are now routinely mass producing arrays of these nano MEMS sensors on 8-in. wafers, and recruiting corporate partners to their Alliance for Nanosystems VLSI for the final stage of developing the MEMS and CMOS processes to integrate them into practical low-cost gas-phase chemical sensors, to monitor toxic industrial gases and gas phase processes, or to analyze human breath to detect diseases.
The detectors are essentially arrays of nanoscale MEMS resonators — fancy versions of guitar strings — set within MEMS flow channels. The resonators are coated with a kind of chemical sponge that absorbs the target material, which changes the mass of the resonator. The gas is first sent through a chipscale version of a gas chromotograph process, to simplify the identification problem.
Though first markets will likely be military and industrial, the most interesting potential may be in medical diagnostics. “There are a few validated tests for detecting lung cancer and other diseases from the gases in the breath, enough to suggest this is a fertile area,” says Roukes, even though studies so far require large-scale lab instrumentation, so are hard to do. “The more easily and routinely this could be deployed, the more deeply it could be studied,” he notes.
All these structures can be made at 90nm, though 45nm would be preferable, says Roukes. He notes that with the device arrays successfully being produced at wafer-scale, current efforts are directed towards precursor systems including surface chemical functionalization and integration en masse with both MEMS flow channels and CMOS circuits for data post processing.
These companies will be among those discussing their latest developments in the program on Emerging Commercial Applications of Nanoelectronics at SEMICON West, July 14-16 in San Francisco. SRC Nano Electronics Initiative director Jeff Welser will also give a mini keynote on the interesting properties of graphene and spin wave transistors with potential to impact the semiconductor industry further out. The program is part of the Extreme Electronics series on emerging technology opportunities for the semiconductor manufacturing supply chain. For details, see www.semiconwest.org.
This article appears in the Summer 2009 issue of Small Times.
July 7, 2009: SouthWest Nanotechnologies, a developer of single-wall and specialty multiwall carbon nanotubes, is receiving $3M equity from Troy, MI-based Insight Technology Capital Partners LP to push product development, manufacturing, and marketing for both existing and new nanotube products.
Last year the company opened a new manufacturing facility in Norman, OK, increasing SWNT capacity 100× and lowering unit costs by 90%. The company also has an applications development center near Boston. Taking on Insight as an investor brings “extensive financial resources, broad industry knowledge and a successful track record of investing in advanced materials companies,” said company CEO Dave Arthur, in a statement.
“We welcome the opportunity to support SWeNT’s continued development as a world leader in carbon nanotechnology. Their leadership in product quality and performance, coupled with their proprietary scalable synthesis process, convinced us that SWeNT is the right vehicle for investment in this burgeoning area of materials technology,” stated Insight Principal Joe Nathan.
July 2, 2009: Researchers at Georgia Tech have developed a new statistical analysis technique that could lead to more precise and reliable measurements of nanomaterials and nanostructures.
The technique, “sequential profile adjustment by regression” (SPAR), identifies and removes system bias, noise, and equipment-based artifacts, which at the nanoscale may be only slightly weaker than true signals of interest. It also can help reduce the amount of experimental data required to make conclusions, and help distinguish true nanoscale phenomena from experimental error.
“At the nanoscale, small errors are amplified,” explained Zhong Lin Wang from Georgia Tech’s School of Materials Science and Engineering. “This new technique applies statistical theory to identify and analyze the data received from nanomechanics so we can be more confident of how reliable it is.”
Specifically, the research focused on a data set measuring the deformation of zinc oxide nanobelts, to determine the material’s elastic modulus. Theoretically, applying force to a nanobelt with the tip of an atomic force microscope should produce consistent linear deformation — but their experimental data showed that sometimes less force appeared to create more deformation and the deformation curve was not symmetrical, and simple data-correction techniques didn’t solve the mystery. “The measurements they had done simply didn’t match what was expected with the theoretical model,” noted Georgia Tech prof. C.F. Jeff Wu. The new modeling technique “uses the data itself to filter out the mismatch step-by-step using the regression technique,” he said.
Georgia Tech researchers illustrate how their new technique improves measurement of nanostructure properties, in this case a graph of elastic modulus of nanobelts. (Georgia Tech Photo: Gary Meek)
In addition to correcting the errors, the technique’s precision could make it easier to produce reliable experimental data on nanostructure properties. “With half of the experimental efforts, you can get about the same standard deviation as following the earlier method without the corrections,” Wu stated. The technique also targets industrial manufacturing environments — i.e., commercialization — “because industrial users cannot afford to make a detailed study of every production run […] the significant experimental errors can be filtered out automatically,” Wu noted.
Future work will target the statistical technique to analysis of the properties of nanowires, whose structure will require a separate model using the same SPAR techniques to correct data errors, Wu noted. The technique also will be applied to past research to possibly generate new findings. “What may have seemed like noise could actually be an important signal,” Wang said. “This technique provides a truly new tool for data mining and analysis in nanotechnology.”
The research, sponsored by the NSF, was published June 25 by the journal Proceedings of the National Academy of Sciences.
SEM images showing a zinc oxide nanobelt on a trenched substrate. An atomic force microscope tip was scanned along the length of the nanobelt with a constant force applied. A series of such scans with the application of different forces produced a bending profile of the nanobelt. These bending measurements were then evaluated using the new SPAR technique to provide information on the nanostructure’s elastic modulus. (Images courtesy of Zhong Lin Wang/Georgia Tech)
Vesselin Shanov and Mark Schulz, U. of Cincinnati
Most natural fibers and nanofibers are produced only in relatively short lengths, and most applications require a bulk or continuous material–but there is no effective method for using short fibers or carbon nanotube (CNT) powders to achieve breakthrough properties in bulk materials. The most promising approach to use nano-fibers in bulk material is to form an intermediate material by spinning CNT fibers into yarn, which can displace conventional fiber in composite materials and other applications. This article examines recent advances that are allowing spinning smaller CNT fibers, and their use in new applications.
Prior to discussing the recent advances, it is important to give due credit to antecedent work. The length and diameter of the fiber play critical roles in the success of spinning; diameters of fibers such as cotton used for spinning since the 16th century are in the micron range, whereas carbon nanotube diameters are much smaller, in the 10nm range. Spinning small diameter fibers also increases twist by about the same factor. Short and long CNTs can be compared to cotton fibers [1]. For traditional spinning technologies such as rotor and ring spinning, there is a separate process called combing where short-length cotton fibers (less than ½ inch long) are removed from the raw material mass before spinning is performed. Typically, as much as 16% of the cotton raw material mass is short fiber, and removed; premium cottons such as Pima or Egyptian have less short fiber but still require combing where typically 8%-10% of the raw material is removed.
These factors and subsequent analyses indicate long CNTs will improve strength and electrical properties, and we have focused research at the University of Cincinnati [2-15] to produce long and strong CNT, spinning carbon nanotube yarns that have superior properties, are economically and commercially viable, and will meet the long-range needs of defense and commercial sectors.
Approaches to spinning CNTs
Spinning CNTs into thread is a relatively new topic of research. There are two main approaches: spinning thread from substrate grown forests of CNT, and direct spinning of CNTs into thread from a vertical reactor that uses the floating catalyst method of synthesis. Spinning from the array is done by a handful of research groups around the world [2-4, 5-9], and direct spinning from a vertical reactor using the floating catalyst method is done by just a few research groups around the world [10, 15]. Mechanical measurements indicate that yarn produced using both methods has a uniform strength of about 0.5N/Tex, which is equivalent to about 1.0GPa; short sections of yarn and special cases have shown higher strength. The electrical resistivity of thread is about 1×10-4 ohm cm and the current density is about 1×109 amp/m2. If the properties are divided by the density of the yarn, the specific properties are competitive with existing materials and the combination of properties can exceed those of existing materials. The mechanical and electrical properties are improving as the number of defects in the thread are reduced through improving the synthesis and spinning processes. The goal is to produce yarns that are strong, creep resistant, highly conducting, and reversibly deformable over relatively large strains to absorb energy.
Fiber properties for spinning. Understanding fiber spinning is important to move CNT out of the lab. Fibers must have certain particular properties to be able to be spun into thread, including strength, stiffness, and pliability–in other words, an openness and ease of fiber separation and toughness, and appropriate bending and radial stiffness. These aspects, along with quality and reproducibility, are of extreme importance in producing yarn. Spinning can be done using different approaches, the details of which are partly confidential; dry spinning from an array is discussed here. The relative size of the yarn being made commercially and the twist uniformity of a strand are important. Our initial target CNT yarn is a 10Tex size yarn (10g/1000m of yarn). CNT length is important in spinning yarn. The most important property of a CNT forest that is required for solid-state processing is that whenever the CNTs at the edge of the array are pulled away from the forest, the CNTs cling together (due to van der Waals forces) to form a continuous strand [5].
Properties of CNT yarn. The mechanical and electrical properties of CNT yarn depend on the number of defects in the CNT and in the yarn. Each gap or junction at the end of the nanotube can be considered to be a defect in the yarn. It can be considered that each CNT has one defect, which is the gap or junction between the next nanotube. Thus there are N nanotubes and N gap defects in the thread, besides defects in the walls that can cause the CNT to become wavy and weak. The gap interrupts the load transfer from CNT to CNT and requires that friction between CNTs carry the load. The effect of length of the CNT on strength of the thread can predicted based on a conventional thread-spinning model which was discussed by R. Baughman [5, 9]. For the CNT thread, a shorter migration length gives better strength–i.e. more fibers must run from the surface to the inside of the yarn in a short interval of length to make the fiber strong. A higher friction coefficient gives better fiber strength. However, the number of turns per unit length–the helix angle–plays an important role in fiber strength; when the number of turns increases the fiber strength decreases drastically. Similar to mechanical load transfer, electrical conduction is interrupted by the gaps between nanotubes, and thus electrical conduction must occur laterally from nanotube to nanotube probably by electron hopping. The resistance of the thread is equal to the longitudinal resistance of the nanotube plus the lateral resistance of the nanotube. The CNT yarn also has resistance, super-inductance, and super-capacitance properties, which are being studied to develop carbon electronics or “carbotronics” that have superior properties in certain applications compared to conventional copper components.
Direct spinning from the array. Centimeter-long “Black Cotton” [a type of CNT trademarked by UC spinoff General Nano] can be spun into thread for electrical wire and as fiber to reinforce composite materials supplementing or replacing carbon fiber. The long CNTs allow dry spinning, which is an advantage in terms of strength, cost, electrical conductivity, and scale-up to manufacturing commodity levels. The U. of Cincinnati’s spinning machine, specially designed to spin Black Cotton into thread, has two DC motors to independently control the twisting and winding while drawing thread directly from the CNT array. The spinner has independent orthogonal control of winding and twisting using a yoke assembly. The thread is twisted and wound onto a spindle. A post-treatment stage allows further processing of the thread, such as thermal annealing or coating with an insulating material.
Catalyst and substrates for growing of spinnable CNT arrays
It is observed that dense and aligned arrays are more spinnable, and the thread obtained from such an array is stronger. Continuous thread can be drawn from dense aligned arrays of nanotubes. In order to achieve this goal, increased catalyst particle density on the substrate is required. There is consensus that double-wall carbon nanotubes (DWCNT) are very appropriate for spinning into threads. A procedure developed for CVD of CNT was modified for synthesis of well-aligned and high-purity DWCNT arrays. A new catalyst based on an iron alloy with increased catalyst particle density was introduced, deposited on a Si/SiO2/Al2O3 substrate by e-beam deposition. After thermal annealing a uniform distribution of high-density catalyst particles was achieved, which was proven by AFM. The growth was performed at 750°C in a First Nano EasyTube 3000 reactor using a gas system consisting of ethylene (C2H4), water vapor, hydrogen, and argon with optimized concentrations, deposition temperature, and flow rates. Critical for this study was to maintain low-carbon partial pressure in the reaction zone. Two-hour growth with this catalyst produced a 1.1mm long array with excellent properties for spinning. Extremely long (up to 18mm) CNT arrays have been made; in one example an 11mm long CNT array was peeled completely off the substrate, which with no additional processing was reused to grow and yield an 8mm long CNT array.
 By using multiple spools, University of Cincinnati doctoral student Chaminda Jayasinghe is able to spin bigger-diameter threads from long (4-5.6mm) CNT arrays. |
Magnetron sputtering is a further improvement in substrate preparation being evaluated by North Carolina A&T and produces very uniform catalyst deposition and highly spinnable arrays.
Device-quality CNT thread, yarn, and ribbon
Our team has developed techniques for spinning long CNT directly from the array into thread, yarn, and ribbons (figure 1). This technique has produced CNT thread with strength of 1 GPa and electrical conductivity of 0.8×104 (ohm-cm)-1. This strength and electrical conductivity are at least 10X lower than the corresponding properties of perfect individual nanotubes. We assume that the lower properties of thread are due to: (i) defects in the walls of the nanotubes; (ii) large number of gaps in the thread at the ends of the CNT; and (iii) considerable open volume between the nanotubes in the threads. We are developing techniques that we believe will reduce these problems. One technique is to use electric current to fuse the ends of nanotubes together while the nanotubes are being spun into thread. This technique has been shown to work under a microscope and the challenge is to scale it up for mass production. Another technique is to perform secondary annealing/welding of the nanotube thread to reduce the number of defects. Other post-spinning treatments of the nanotube thread, such as ion irradiation and ozone or UV exposure [16-19], also have potential to improve the thread’s electrical and mechanical properties. The strategic importance of the research is to produce CNT thread that surpasses the properties of any existing material in terms of strength, weight, and electrical current carrying capability.
 Figure 1. Four types of CNT materials fabricated by the UC team: a) as grown CNT bundles, b) single thread, (c) two strand yarn, and (d) ribbon. |
null
Preparing thread and ribbon from CNT arrays
At the present time CNTs cannot be grown beyond about 2cm in length [20-37]. At the U. of Cincinnati, DWCNTs several millimeters in length are being spun into thread (with micron-range diameter) to produce a strong and tough bulk material with novel properties. Multiple threads have been woven together to form a yarn, which can be used to form tows and unidirectional plies; they also can be woven into a “smart fabric” with two-directional properties, which can then be used to fabricate strong, electrically conductive composite materials, or used as a wearable sensor embedded in clothing. Nanotube thread can also be used to form carbon electronic components (“carbotronics”), or electromagnetic devices such as an antenna to communicate with sensors inside the body. Thin narrow sheets of nanotubes called ribbon have also been drawn from the array [8, 11].
Spinning thread from DWCNT arrays
Closely aligned CNTs (spaced <100nm apart) are weakly held together by van der Waals intermolecular forces. In this forest configuration, they can be harvested by pulling a small bundle of CNTs away from an edge of the array in the direction that keeps the “centerlines” parallel and maintains the close spacing. The CNTs next to the first bundle will be pulled along also by van der Waals forces. As these CNT bundles are pulled away from the “forest” they form a long line in which all of the CNT centerlines are aligned in parallel. In the spinning process (pulling and twisting), CNTs are pulled from the array and held together by twisting around neighboring nanotubes [38-42], which prevents the CNTs from slipping along the lengths of their neighbors when axial force is applied. As the CNTs are twisted and pulled, more nanotubes are added to form a long, strong thread. CNT arrays that we have used for spinning range in length from 1mm to 0.5cm–and threads with lengths of 100m have been spun in our facility. The diameter of the thread can be controlled by the length of the CNT array and by the spinning parameters [43-53]. Figure 2a shows CNT thread being wound onto a spool.
 Figure 2. Manufacturing of CNT thread and ribbon at UC: (a) thread being wound onto a spool, (b) pulling and winding ribbon. |
As the quality of the CNTs improves, progressively longer CNTs will be used to spin thread. Long CNTs can improve thread properties by reducing the number of gaps in the thread at the nanotube ends, and by providing a longer length for mechanical interlocking each nanotube with its neighbors. If the quality and spin-ability of the 1.5-2cm long CNT that we currently produce are improved, thread properties could improve by an order of magnitude which would open up many applications. The excitement of this research is that the properties of thread are being continuously improved and are getting closer to the properties of individual nanotubes. If CNT thread reaches a strength of 10GPa (20% of the strength of individual nanotubes), this would be a large breakthrough in nanotechnology. Electrical conductivity would also be expected increase roughly in proportion to the strength increase.
Pulling ribbon from CNT arrays
A material called carbon nanotube ribbon (~200nm thick, 5mm wide) was also produced from our arrays, possessing a different morphology from threads. Winding CNT ribbon is shown in Figure 2b. The width of the ribbon is limited only by the lateral size of the array [11].
Specific properties of CNT and thread will be important for weight-critical applications and are calculated via dividing the property by the density of the material. However, since a single wall of each nanotube is one atom thick, defects can greatly reduce the strength and electrical conductivity of CNTs. Reducing intrinsic defects will greatly improve the properties of future CNT threads, ribbons, and yarns. Re-spinning, treating with solvent, and other post-processing is being done to improve the properties of thread. Several techniques to post-treat CNT yarn are being evaluated, mostly consisting of applying energy to the spun yarn in the form of electron flow or heat. This will be in a controlled atmosphere to prevent oxidation. The intent is to “fuse” the twisted CNT together so they will resist slipping when a lateral force is applied to the yarn.
Applications of CNT thread
CNTs’ high specific strength and stiffness, electrical and thermal conductivity, and compatibility with electronics and sensing applications are a key enabling medium for the convergence of textiles with fully integrated functions. Smart fabrics or interactive textiles have many potential applications such as physiological monitoring, power bus systems and communications, medical care, multifunctional exteriors, harvesting of energy and water, and passive and active thermal management. CNT yarns may eventually find applications in composite materials, in electrically conductive wire, bulletproof vests, light emitters by incandescence, antennas, and materials that block electromagnetic waves. The macroscopic CNT yarns may find application as mechanical actuators for artificial muscles, flexible conductors for textile sensors, power buses for communication, flexible batteries, and solar cells.
One example of new application is in signal communications. Researchers at the U. of Cincinnati have applied a 25µm spun carbon nanotube thread to create a dipole cell-phone antenna [54], with transmission close to that of copper but at a fraction of the weight [44]. CNT yarns also could be woven into cloths for use as a very lightweight reflector or dish antenna–one that could be deformed to change the focus of the antenna, e.g. to be molded directly into electronic device casings or aircraft structures. Nanocomp Inc. [15] has demonstrated fabrication of coaxial cables using CNT threads as center conductors and CNT sheets for outer conductors or shields, which would be substantially lighter than similarly sized copper coaxial cables; USB cables using only CNT threads and yarns also have been made.
 David Mast, U. of Cincinnati associate professor of physics, demonstrates new wireless applications of the spun carbon nanotubes. |
The surface of CNT materials also can impart different functionalities–e.g., absorption of gas molecules to make ultrasensitive sensors for toxic gases or biological agents. Fabrication of such functionalized CNT sensors with integral CNT antenna for wireless sensor applications is currently being investigated. These threads and yarns can also be wound into small loops or spirals for micro-miniature antenna for possible bio-medical applications.
The CNT yarns are good electrical conductors and can carry enough current to act as an incandescent filament or to emit electrons to produce light from phosphorescent screens. Electrons field emitted from the side of a cold, negative nanotube yarn electrode hit a transparent fluorescence screen to provide light emission. Low voltages are possible for field emission because of both the field enhancing effect of the yarn shape and the high aspect ratio of nanotubes that protrude from the sides of the yarn. The CNT yarns can be used as electron field emitters for light sources (lighting and displays) and X-ray sources that could be in a micro-catheter used for medical applications. The individual nanotubes are anchored into the yarn by twist, which should enhance electron emission stability and device lifetime.
Conclusions
Steady progress in CNT synthesis and spinning by a small number of groups around the world is moving nanotube yarn technology ever closer to the point where it can be come a disruptive material that can offer multi-functional properties that cannot be achieved by any other materials on earth. CNT yarn may supplement or displace carbon, glass, and aramid fibers and copper wire in high-performance critical applications.
Dr. Vesselin Shanov is associate professor of chemical and materials engineering at the University of Cincinnati.
Dr. Mark Schulz is an associate professor of mechanical engineering at the University of Cincinnati, and deputy director of the National Science Foundation’s Engineering Research Center for Revolutionizing Metallic Biomaterials located at North Carolina A&T.
Shanov and Schulz are co-directors of the UC Nanoworld and Smart Materials and Devices Laboratories at the University of Cincinnati, and are affiliated with the start-up company General Nano LLC in Cincinnati that is commercializing the Black Cotton material.
References for this story are available online, at www.smalltimes.com.
by Li-Shun Wang, Ana Hunter, Samsung Semiconductor, Inc.; Seung-Mahn Lee, Samsung Electronics Co.
Executive overview
June 26, 2009 – As IC design, manufacturing costs and complexities continue to increase, collaboration with customers from the earliest stages of technology development through high volume production is vital for the next generation foundries. In the earliest stage, collaboration with customers establishes specification, feature definition and process technology requirement. By co-optimizing physical design with manufacturing process technology, foundries can preserve design intent in actual silicon and ensure rapid time-to-market and right-the-first time working designs. During the product and technology development stage, strong partnership in data sharing and FA collaboration enable accelerated yield learning cycles.
To sustain requisite yield, manufacturing process control and defect performance has to be significantly improved. Manufacturing innovation becomes essential for the next generation foundries to keep pace with the scaling advantages of high-k metal gate (HK+MG) in 32nm node and beyond. Process variation in smaller geometry is becoming a pronounced engineering and manufacturing challenge. In this article, we will discuss the advancements in advanced manufacturing using Samsung’s S1 fab, a state-of-the-art 300mm foundry line, as an example in the areas of patterning and closed loop variation control systems.
Introduction
Leading foundries are moving towards 32nm high-k metal gate (HK+MG) and beyond today. The introduction of HK+MG in the 32nm node though the Common Platform technology alliance, a collaboration among IBM, Chartered and Samsung, offers major benefits of high performance, reduced static and dynamic power [1]. The advanced technology node delivers the expected Moore’s law benefits of increased density and performance and also requires significant investment from both chip design and manufacturing. For HW/SW design in an advanced technology design project, the non-recurring engineering (NRE) cost could go up from $120M in 45nm node to $180M in 32/28nm node [2]. With these development costs, the need for close collaboration and a fast ramp to volume production is imperative for return-on-investment (ROI) for both the customer and the foundry.
Today’s complicated designs and advanced technologies dictate the need for close collaboration between the customer and the foundry. Manufacturing variations can affect device performance and conversely, design layout can greatly impact process variations. Manufacturability of advanced technology is determined by both the design and the manufacturing process. Controlling these factors requires close customer collaboration, new manufacturing methods, and possibly new product design architectures.
Foundry and customer collaboration from the earliest stages of technology development through high-volume production is crucial to success. From the outset, collaboration with customers between design and process technology establishes specification and feature definition. As design rules for each new node change, design-for-manufacturing (DFM) methodology has to improve for continued scaling. Understanding the delta between design intent and silicon results leads to co-optimization of physical design and process technology and thus drives DFM improvement to ensure right-the-first time working design. During the product and technology development stage, a strong partnership in data sharing and failure analysis (FA) collaboration accelerates yield learning cycles.
Yield improvement requires the control of both design marginality and process defects including random, systematic, and parametric defects. Each category contains various issues that the foundries and chip designers must work together to resolve. Fast yield ramp and world class quality can only be achieved with a strong technical relationship between customers and foundries to take advantage of the combined expertise in manufacturing, design and test.
To sustain requisite yield from generation-to-generation, manufacturing process control and defect performance has to be significantly improved. Manufacturing innovation becomes essential for next generation foundries to keep pace with the scaling advantages of HK+MG at the 32nm node and beyond. Lithography, for example, is a significant challenge in semiconductor manufacturing as the device features are well below the most advanced manufacturing worthy lithography tool optical wavelength (Table 1).
Semiconductor Manufacturing Roadmap by ITRS 2007.
CLICK HERE to view larger image
Physical gate length in logic circuits today is getting down to <30nm level requiring 3σ variation control of <2nm. At these geometries, variability in the manufacturing process has a much greater impact on device operation than in previous technology nodes. Variation control and containing systematic and parametric defects is a pronounced engineering and manufacturing challenge. This article looks at the advancements in next generation manufacturing using Samsung's S1 fab (Fig. 1), a state-of-the-art 300mm foundry line, as an example in the areas of patterning and closed-loop variation control systems implementation.
Figure 1. Inside Samsung’s 300mm logic process facility, S1 Line, located on a 350-acre site in Giheung, South Korea.
Early customer collaboration brings significant advantages
The early engagement between device and process engineers and chip designers offers significant advantages to both the foundry and the customer. Customers gain access to the latest technology enabling higher performance, lower power and increased SOC integration with time-to-market advantage. The foundry gets real product design feedback to validate and fine-tune the technology and faster time-to-volume, accelerating ROI. [3]. Without early customer collaboration, the foundry works serially on process unit development, process flow, and specifications. Once these are established, the foundry delivers its design rules and models to the customer. The customer designs a new product and then delivers the design database to the foundry. In this scenario, problems are addressed at a late stage in the flow, whereas it would be much more effective to identify the issues upstream to avoid design delay or respin.
Under close customer collaboration, using a concurrent and interactive work flow, the foundry and customer co-optimize the design and process early in the development cycle through detailed discussions on interaction between process, layout structures and circuit design. The customer provides direct feedback and has significant influence on the manufacturing design rules and technology offerings.
The early collaboration across process and design ensures a better understanding and balance of all requirements to establish the technology specification and feature definition. With this feedback mechanism in place, the foundry provides multiple Vt core transistors, I/O transistors, and SRAM options for size, speed and power trade-offs best suited to designs requiring the latest advances in process technology. The customer can analyze design trade-offs and select the best strategy for the silicon process and design methodology to achieve the product goal within the required power consumption budget and with the best possible performance.
Figure 2 shows the design-process collaborative flow between the customer and the foundry. As result of close collaboration, the customer and the foundry arrive at a set of design rules and silicon options with built in optimization that meet both advanced process and design requirements.
Figure 2. The design-process collaborative flow between customer and foundry.
Accelerating yield learning cycles with data sharing, failure analysis collaboration
Historically, foundries have driven yield improvement by controlling particles from the process tools and environment, manufacturing process flow management and use of statistical process control (SPC). Today, more and more yield-loss issues are related to systematic design and process interaction marginalities. Improving product yield requires both product design tuning and process technology improvement. The design tuning is conducted via the early collaboration described above. Once the design is completed, to achieve stable volume yields, further process tuning requires design topology information input.
An understanding of the device’s physical and electrical characteristics, such as areas of high density features, critical speed paths and the location of precision analog blocks are used to create measurement recipes and fed into inspection systems. The customer and foundry work together to characterize nonfunctioning units and perform physical failure analysis (FA) to determine causes of yield loss, which is then fed to foundry for yield improvement activity. By partnering with the customer to share wafer-level bit map from device functional testing, the foundry can compare wafer-level bit maps to inline inspection results driving a better understanding of failure modes and root cause fixes. Failure bins classification from the functional testing analysis delivered in a real-time feedback path from the customer to the foundry is a key success element for a close customer collaboration strategy. Data sharing and FA collaboration offer high-level diagnostics, which enables accelerated yield learning cycles throughout the product development stage.
To improve yield from the development stage to a mature production level, the foundry must control random and systematic defects and parametric issues quickly. The speed of yield learning goes straight to the business’ bottom line. Random defects are mostly controlled by managing process tools and production environment cleanliness. Systematic and parametric defects control requires both design optimization and manufacturing innovation. The next section discusses the manufacturing advancements in gate patterning and closed loop variation control in Samsung’s S1 foundry line in addressing systematic and parametric defects.
Manufacturing innovation keeps pace with 32nm HK+MG and beyond
With the aggressive scaling of devices made possible with 32nm HK+MG, the required gate CD control comes down to the level of 1-2nm in 3σ. Even with the latest 193nm and 1.35 numerical-aperture (NA) immersion lithography tools, the mask and lithography processes are pushed to the extreme to meet this tight variation control. One option is double patterning whereby very dense tight pitch structures are divided onto two separate masks and “double” exposed. This is not desirable due to the negative impact on cost and throughput. In order to maintain single exposure, novel methods for lithography and etch variation control are required. As an example, Samsung uses a ASML DoseMapper to improve gate patterning quality in its S1 logic foundry fab.
Equipped with DoseMapper capability, ASML lithography tools modulate the exposure dose across the exposure field for process quality enhancement. DoseMapper, if applied correctly, can improve lithographic quality by compensating for external sources of CD variation such as reticle non-uniformity. The manufacturing process for reticles often creates a systematic feature size variation across the reticle field. Using DoseMapper, the dose profile is optimized to compensate for the reticle induced feature size variation in both the slit and scan axes. This technique has been studied in Samsung’s S1 manufacturing process to improve lithography process and systematically improves production quality. For gate patterning, poly interfield variation (AWLV) and process Cpk have been improved by >25% by using DoseMapper as shown in Fig. 3.
Figure 3. Inline parametric Cpk improved by >25% after using DoseMapper in gate patterning.
Closed-loop variation control: design, manufacturing, final test
Process variation is a pronounced engineering and manufacturing challenge today. The application of advanced process control (APC) has been dramatically increased in recent years as the most effective methodology to reduce variation in next generation foundries. Closed-loop variation control systems incorporating APC have been applied from design through manufacturing to final test in semiconductor manufacturing. In Samsung’s S1 fab, more than 100 unit processes are now fully covered and controlled in real time by framework based APC systems.
Although the implementation of APC algorithms requires in depth process and equipment understanding and characterization, the concept is simply how to feed back and/or feed forward the previous lot/wafer’s data to current and/or future lot/wafer to reduce the cumulative variation effects of each process step while maintaining the process feature at the specification target. To design an APC model, the APC team has to determine that it is controlled by each process, each chamber, each recipe, each product, etc, which is decided based on the communication with process and device teams. In Samsung’s S1 fab, APC systems are now in effect for almost all film processes and have demonstrated Cp and Cpk improvement by more than 25% in actual product data.
The lithography processes, key etch and implant processes are controlled by an APC system and the control performance shows similar improvements in ADICD, ACICD and related electrical testing (ET) data. Figure 4 shows the control performance of gate loop APC system on related key ET data. The Cp was improved by 40% by use of APC in the gate module. Advanced tool features such as the ASML DoseMapper that can adjust exposure dose in response to APC information from the reticle and etching process are making it possible to attain high yields in spite of the technical challenges at 32nm and below.
Figure 4. The implementation of closed loop control at gate loop showed >40% Cp improvement in key electrical testing data.
Conclusion
With today’s escalated development costs and complicated chip manufacturing, the need for close collaboration and a fast ramp to volume production is imperative for ROI for both the customer and the foundry. By partnering with the foundry, the customer achieves early understanding of the new process technology, which translates to fast time to market and superior product design with built-in process optimization and proven foundry technology. In addition, manufacturing innovation in 32nm HK+MG technology and beyond will continue to bridge the gap down to the process technology roadmap by containing both process variations and manufacturing defects to sustain requisite yield for the next generation foundries.
Acknowledgments
The authors would like to thank Dr. Steve Hah and Moon Won Lee of Samsung’s S1 fab, Dr. CS Choi of Samsung Foundry, Dr. Insik Chin of System Engineering Group in SEC, and Lisa Warren-Plungy of Samsung Semiconductor, Inc. for their support and contribution to this article.
References
- K. M. Choi, “Advantage of 32nm High-K & Metal Gate: Design & Product Perspective,” Common Platform Technology Forum, San Jose, CA 2008.
- M. Keating, “Reshaping Chip Design — From Architecture to Tapeout,” SNUG San Jose, CA 2009.
- Y.H. Su et al., “Inter-firm Collaboration Mechanism in Process Development and Product Design between Foundry and Fabless Design House,” Semiconductor Manufacturing Technology Workshop Proceedings, pp. 47-50, 2004.
Biographies
Li-Shun Wang received a PhD in materials science and engineering from Northwestern U., and a BS in physics from the U. of Science and Technology of China. He is a foundry engineering manager at Samsung Semiconductor, Inc (SSI) in San Jose, CA. E-mail: [email protected].
Seung-Mahn Lee has a PhD in materials science and engineering from the U. of Florida and is a principal engineer at Samsung, S1 fab, Samsung Electronics, Giheung, Korea.
Ana Hunter serves on the GSA board of directors and is VP of foundry in Samsung Semiconductor, Inc (SSI) in San Jose, CA.
June 26, 2009: KLD Energy Technologies is launching US sales of an electric scooter based on an transmissionless motor system that uses a nanocomposite material to boost energy efficiency.
The firm’s “Neue” motor, which it says can reach speeds of up to 65mph (Twice that of other electric motors) and a range of 100mi on a standard lithium battery, is built on a nanocrystalline composite material that the company says conducts energy up to 10× more efficiently than iron-core motors (2500Hz vs. 250Hz), with no need for additional cooling systems.
With a high frequency and low RP the motor does not require a transmission, and enables the scooter to achieve speeds and performance levels comparable to gas-powered vehicles, the firm says.
Base price for the scooter is $3288, plus a $500 reservation; deliveries are slated to start in 3Q10.