More than a buzz word

Nanotechnology is currently an industry buzz word with a lot of hype and hope. A forum to discuss the potential of this emerging technology was held in the USA recently. Dick James of Chipworks was in Texas and provided this report for European Semiconductor.

Recently (14 – 17 Nov) I attended the SEMI Nanoforum in Austin, with the intent of seeing what impact evolving nanotechnology could have on the semiconductor business, and Chipworks business in particular. It was billed as “the first global event to bring together SEMI members and representatives from leading nanotechnology developers, in order to explore applications and strategies for commercialisation.”

“SEMI NanoForum will give semiconductor equipment and materials companies and nanotechnology professionals the opportunity to identify, through open discussions and informative presentations, areas where their industries have the greatest synergies.”

Topics covered over the event were Nanoelectronics/Photonics, Nanomaterials, Nanoprinting, Nanofabrication, and Nanocharacterisation.

The first session was a generic introduction to the field, to bring people up to speed with the state of the art. The best session of the tutorial was the one on lithography (Nanoprinting/Nanofabrication), by Grant Wilson, who has specialized in resist development for the last few years. He is a big proponent of Step and Flash Imprint Lithography (SFIL), and he makes a very persuasive argument for it. The nanomaterials session provided interesting feedback on the progress made in the production of nanowires from semiconductor materials. They are a long way from commercial usage, since there does not seem to be a way to orient the wires so that they are usable in an organized way. The last set of the day on characterisation was mostly on TEM imaging, with a few R&D notes on molecular probes.
Overall, the day was a good review, but from a university R&D perspective. The exception to that was the nano-imprint lithography, which seems to have progressed significantly in the last few years, to the point that it will soon be a manufacturable technique and could well challenge optical lithography in the next couple of process nodes.

Commercialising technology

The NanoForum itself was a two-day conference followed by a half-day workshop. The first day was focused on nanotechnology challenge and commercialisation, and the second on technology capabilities and synergies, basically a summary of the latest technologies, mostly in the semiconductor industry. The intent of the workshop was to synthesise and discuss the information from the conference and see what actions can be taken to focus activities in the field.

The first day went through the range of nanotechnologies and applications, from semiconductors and displays, through medical and defence applications, to automotive uses such as dirt-repellant paints and fuel cells. This was both depressing and exhilarating – depressing, since the range is so broad that its hard to know where to start, and exhilarating because the potential is so huge for both business and societal gains.

IBM had some interesting illustrations of self-assembling copolymer films that can be organized into patterns useful for nanocrystal flash memory, or embedded capacitors – no lithography needed! Motorola are looking at field emission displays (FEDs) using carbon nanotubes (CNTs), and conversely the Industrial Technology and Research Institute (ITRI) of Taiwan have demonstrated a large area CNT/FED backlight for a LCD panel (which to me, seems to be a more pragmatic and immediately commercial idea).

Lockheed Martin detailed the variety of nanotech applications being investigated by the defence industry, and Medtronic did the same for the medical business. In both areas there was a large emphasis on sensors, but also including how material modification on the nanoscale can meet many needs in a more passive way. Daimler Chrysler are doing the neat stuff in fuel cells and self-cleaning cars, and of course sensors, but the speaker did not mention those. Tom Cellucci, CEO of Zyvex gave a very high level talk about getting into the nanotech business - basically, dont do it, unless you are really sure of your market.

Perhaps the most interesting talk of the day was by Prof Rick Smalley, who won the Nobel Prize for the invention of Buckyballs, (AKA fullerenes, molecular carbon spheres, which have now evolved into nanotubes). He gave an impassioned speech about the impending doom of the energy crisis when oil runs out, as it will 30 – 80 years from now depending on who you believe and on the development of the Asian economies. His major point is that the energy crisis can be overcome, but it needs action now to enhance the technologies that will generate the energy to replace fossil fuels. In the interim, nuclear will help, but reactors cannot be built fast enough and there is not enough uranium for conventional reactors anyway.

His thesis is that there is enough energy reaching the planet for all conceivable needs, but we need to develop the technology to extract it, distribute it, and store it. That takes investment by government, and a concerted approach from industry. An additional problem is the lack of young people going into science and technology. Prof Smalleys message was that we should go out and proselytize, get politicians to realize the problem. What has that to do with nanotechnology? Undoubtedly nanotech will provide some of the solutions, whether it be batteries, fuel cells, or whatever.

Technological challenges

The second day was more detailed and featured a range of presentations by equipment, materials, and semiconductor manufacturers, starting with the materials potential from 3M and Dupont, then lithography from ASML, Hitachi (e-beam), and Molecular Imprint (SFIL), wet etch from FSI, deposition from Applied Materials and Genus, and metrology from FEI, KLA-Tencor, Veeco, and Philips.

On the materials side, it is obvious that a lot of work is being done (and any chemist these days would claim that theyve been doing nanotechnology from the start!), but there were a couple of interesting examples directly in the semiconductor field. 3M have developed nanoscale silica particles for filler in no-flow underfill for flip chip packages. Conventional filler in capillary underfill can have particles up to 50 microns in diameter; if these are used in a no-flow application, where the polymer is put down before the solder-bumped chip is flipped on to the substrate lands, they can prevent good contact. The filler can be left out, but then there is thermal mismatch, since the resins CTE is 60-70, solder and FR4 board is ~20, and silicon is ~3. Adding filler can bring the CTE down to ~35 and solve a lot of problems. In the area of organic semiconductors, 3M also found that putting down an organic monolayer on silica or alumina substrates before deposition of the semiconductor can improve the carrier mobility by 500%.

In the lithography area the imprint technique again impressed; there is obviously a lot of investment inertia in deep-UV lithography, but I think most attendees were convinced that with more development to take the technology to a volume production scale, it could be the lithography method of the next decade. This conclusion was supported in a survey by Lux Research, which pegged 193 nm immersion lithography as the short-term winner to succeed the current dry 193 nm litho, but put nano-imprint ahead of EUV and X-ray projection as the long-term solution. Ive been in the business long enough to remember the problems caused by nitride stress in the early days of LOCOS isolation, and then nitride passivation, and yet now its being used for mobility enhancement – will we see a similar re-run for the imprint version of contact printing?

On the metrology front, Philips has developed laser-based techniques to look at metallization and dielectrics. This seems to be an enhancement of the thermo-acoustic technique that has been around for years for implant metrology, but it has evolved to measure copper thickness and resistivity, and the elastic and thermal properties of dielectrics, and on patterned wafers too.

Novellus, AMAT, and Genus all affirmed that their technologies were well capable of meeting the nanoscale challenge, so for the foreseeable future it seems that the dielectric/metal stack can be built as the nodes shrink. However, nobody mentioned fundamental problems such as electron surface scattering as metal linewidths diminish.


On Tuesday evening there was a banquet at which the guest speaker was George Atkinson, Science and Technology Advisor to the Secretary of State. The theme was “Science and Technology in Foreign Policy”, which in the USA is a function of the State Department. His address discussed USA issues and went on to consider the number of foreign students enrolling in US institutions, which has dropped dramatically since September 2001, and the efforts being made to counter that trend. Not only are universities and colleges hurting from the reduced income, but also the long-term future of US S&T is affected, since immigrant talent has been a huge part of both academic and industrial success in the last fifty years. Dr. Atkinson illustrated the need for S&T within the government by describing efforts the State Department is making to reduce the barriers being raised by the Immigration Department on security grounds, but which is having a drastic effect on conference attendance and staff exchanges. In the Q&A session afterwards there was a bit of a division in the listeners attitudes to internationalizing US S&T, between those who thought it should be kept in the US, and those with a less nationalistic approach. Dr. Atkinsons conclusion was that science and technology was a global commodity, and the interests of the USA are better served by encouraging that outlook. All in all, it was an illuminating discourse on the internal workings of the US government.

What is nanotechnology?

SEMI as an organization is obviously taking a proactive stance to keep ahead of the nanotech wave, even though we are already pretty well into it in simple dimensional terms. Part of the discussion over the three days was about a definition of nanotechnology; and while no formal definition was posted, there seemed to be a consensus that the revolutionary changes in technology will be in the development of “bottom-up” ways of making devices, rather than the current “top-down” methods. For example, using self-assembly techniques to create conductive elements instead of defining them by lithography. This actually coincides fairly well with another tentative definition, whereby nanotechnology is the use of techniques at a scale where surface material properties dominate the bulk properties.

In the context of the semiconductor business, the industry will continue into nano-dimensions by evolving the current techniques – immersion lithography will move us into the next couple of process nodes, and it seems that imprint lithography could be a significant disruptive technology. The materials and processing techniques also seem capable of coping with this evolution. The interesting phase will be as we get beyond the ITRS “red brick wall”, and we reach the physical limits of the materials, whether it is a single dopant atom under a gate, or the resistance of a too-thin copper line.

And that brings me to where I thought the NanoForum was lacking. The current players showed us that we can continue with the “top-down” techniques, and there was a good consideration of “true” nanotechnology, still well in the R&D mode, but there was hardly any discussion of the transition that will have to occur between the two. Will we see carbon nanotubes as interconnect? Since they have ballistic electron transport and therefore higher conductivity, unlimited by surface scattering, nanotubes could be a viable candidate. Nanotube transistors are out there, but how do you orient them in a controllable manner?

There were two examples of nanotubes in field emission displays, so in the display world it seems we will see product soon, and there is R&D on nanotubes as contact/via structures, so maybe that is where the transition will start. We will see – I look forward to IEDM!

I was also disappointed not to have more coverage of analytical techniques. Imaging was reasonably well dealt with – we know we can see at the atomic scale with TEM and scanning probe techniques, but how do we measure the doping in a nanotube? Or in a 20 nm gate, for that matter? There are single atom probes that are still experimental tools, but it seems to me that they will have to be commercialized if the nanotech business is going to be able to know what it does as it builds its products. On the positive side, we actually have a convergence of trends here, since atom probes look at very small amounts of material, and devices are shrinking into their range.

Aside from those two gripes, the conference was well put together, with a good roster of speakers and attendees. Overall, my guess is that nanotube or nanowire transistors are at least 10 years away in commercial product. If we really blue-sky, we could eventually end up with structures with nanotube conductors treated to give a semiconducting portion, with a gate layer wrapped around it – freedom from the tyranny of a substrate! Interconnect them in a low-k matrix, and then we would really have 3-D nanoelectonics! With any luck, I might be retired by then..

Dick James is a Senior Technology Analyst with Chipworks. Originally from the UK, Dick has over 30 years experience working in the process development, design, manufacturing, packaging and reverse engineering of semiconductor devices.He joined the Chipworks team in March 1995 as Manager, Microstructural Characterization, and is currently Senior Technology Analyst. He can be contacted on www.chipworks.com