Toshiba and Canon team up for developing 15nm nanoimprint lithography

Toshiba and Canon team up for developing 15nm nanoimprint technology

Toshiba and Canon, two of the major Japanese companies operating in the semiconductor business, have teamed up to develop nanoimprint technology for NAND applications. As we discussed already in this blog a few days ago Canon is seriously committed to nanoimprint lithography and they have recently bought Molecular Imprints, a company based in Texas, US, to gain traction with the new technology.

Nanoimprint lithography has the potential to revolutionize the lithography market but so far suffered from a series of problems that prevented full adoption for IC patterning: the main one being lack of alignment precision between two consecutive imprints.

If this and other problems (such as defectivity in the master mold and planarity control between the master mold and the substrate) will be overcome, nanoimprint lithography may become a serious alternative to conventional lithography due to the relatively inexpensive cost of imprinting a large number of wafers with a single master mold.

While logic patterning may still be challenging for nanoimprint lithography, NAND patterning has been seen for a while as the first step for full-scale adoption of the technology in mass production.

The strategy adopted by the two companies will require the installation of Canon machines at Toshiba` s plant in Mie prefecture. As for how the Research and development costs will be split between the two companies, a formal detailed announcement will be done later this year.

Currently, Toshiba is competing for the top spot in the NAND market with Samsung electronics of Korea with both contenders trying to reduce the cost of patterning per wafer and increasing the number of transistors packed per area.

The current patterning for NAND is based on 19nm lithography technology, with plans to move to a thinner and more precise 16nm~17nm process this summer.

Toshiba is already using ASML lithography machines and will add Canon imprint machines thanks to this collaboration.

Link (in Japanese): http://www.nikkan.co.jp/news/nkx0320140227aaaq.html?source=myce

Note:
If you are interested in nanoimprint lithography, please visit our nanoimprint lithography service page.

Subscribe to our newsletter to receive our new articles directly in your mail box.

If you liked this article, please give it a quick review in StumbleUpon, Facebook or Pinterest.

The two main categories of nanoimprinting

The two main categories of nanoimprinting

While there are many different flavours of nanoimprinting processes, two main ones are:

      • Thermal nanoimprinting (also called as hot embossing)
      • UV-nanoimprinting

Thermal nanoimprint technology

Thermal nanoimprint technology has been the first kind of nanoimprinting process ever used and the original one adopted by Prof. Chuo back in 1996.

The thermal nanoimprint process is very easy to understand and straightforward: a layer of thermoplastic polymer is deposited on the surface of the substrate by spin coating or alternative method; then the substrate is placed inside the imprint machine along with the mold and the mold is pressed against the substrate at a set pressure.

The substrate received a precise amount of heat from the machine and the temperature of the polymer rises over the glass-transition temperature and becoming soft.

The mold is kept pressed against the substrate for an amount of time, after that the substrate is cooled down and the mold is released

After the pattern is transferred to the mold polymer layer, it is then transferred to the beneath silicon substrate surface by etching process

UV-nanoimprint technology

UV-nanoimprint technology process is different from thermal nanoimprint in two aspects: it does not use heat to soften the polymer layer but UV light and requires a UV-transparent mold.

With UV-nanoimprint technology, the polymer used on the substrate surface is a UV-curable photopolymer and not a thermoplastic one as in the case of thermal nanoimprint; this allows the process to be effective at considerably lower pressures and helps to avoid the issue of thermal expansion

As discussed, the main advantage of thermal nanoimprint over UV-nanoimprint is that molds of any material can be used while, in case of UV-nanoimprint technology, only UV-transparent materials can be used

However, since UV-nanoimprint uses mainly UV light to deform the polymer, considerably lower pressure levels on the substrate are needed to be able to transfer the pattern to the polymer

Moreover, the advantage of not needing much heat to deform the polymer brings other advantages such as much lower special deformation of the pattern due to applied heat and therefore a better alignment accuracy and the ability to imprint larger areas.

As for UV-nanoimprint, there are two sub-categories: hard-mold UV-nanoimprint and soft-mold UV-nanoimprint: as the name implies, hard-mold UV-nanoimprint is a process that uses a quartz or other rigid mold, while soft-mold nanoimprinting uses a soft mold usually made in polymer.

Hard-mold UV-nanoimprint technology has two main shortcomings if compared to soft-mold UV-nanoimprint: the mold is rigid and therefore there are sometimes issues while releasing the mold from the substrate; moreover, there are other issues when trying to imprint large areas due to the inherent issues of flatness of the mold and the substrate.

Lack of perfect flatness in either the mold or the substrate can lead to defectivity in the pattern or, in worst cases, to a breakage or the substrate to the mold.

Using a soft mold, however, leads to some shortcomings too: while issues of flatness between the mold and the surface are greatly resolved while using a soft mold and releasing is far easier, problems like thermal expansion and non-uniformities are worsened due to the elasticity of the mold itself

Deformation of the pattern due to even small quantities of heat can become a significant problem at lower resolutions and this may lead to additional issues when the surface to be patterned is large and therefore the thermal expansion rate of the mold polymer varies across the substrate surface

A variant of the UV-nanoimprint process is the so-called Step and Flash ® process.
Using this process, initlally developed at the University of Texas, the polymer is deposited on the substrate not by spin coating or by any other machine, but by a pattern generator which is embedded in the imprint machine. After the polymer has been deposited, a soft mold is gently pressed over the substrate surface and the pattern exposed to a UV-lamp.
After the soft mold is detached, the patterned polymer is etched and the unnecessary features are removed.

Other further categorizations of the process are: single step or multiple step and single layer and multiple layer nanoimprint

Single step nanoimprinting involves the patterning as a one-time process while multiple step process (also called step-and repeat nanoimprint) involves multiple step patterning of usually large areas of substrate with smaller area molds.

Single layer nanoimprinting is the traditional process where one single layer of pattern is transferred into the substrate surface, multiple layer nanoimprinting instead requires thhe patterning of multiple layers of patterning over the same area.

Note:
If you are interested in nanoimprint lithography, please visit our nanoimprint lithography service page.

Subscribe to our newsletter to receive our new articles directly in your mail box.

If you liked this article, please give it a quick review in StumbleUpon, Facebook or Pinterest.

Virtual interview about nanoimprinting lithography and its current applications

Nanoimprinting lithography replica

Q: Let` s start to discuss about the current situation of nanoimprint technology and how do you see it as of now. There were a lot of expectations behind nanoimprint a few years ago, do you think nanoimprinting lithography has delivered as promised or not?

A: Until a few years ago, nanoimprinting lithography was one of the candidates of the so-called next-generation lithography that should have replaced conventional optical lithography for the production of ICs in large scale. In this sense, I believe that nanoimprint, along with other technologies as multi-beam and direct self-assembly, can not be seen today as a replacement of conventional stepper lithography. However, I do not think that nanoimprint technology has proven to be a failure, I see many areas of applications outside the conventional IC processing market and I believe that most of the potential of nanoimprinting lithography has yet to be tapped

Q: You have said that nanoimprint, along with other technologies, have not delivered as someone hoped as a potential replacement of optical lithography. Can you explain to us the reasons why this did not happen?

A: There are various reasons why nanoimprinting lithography and other technologies are having a hard time in replacing conventional (stepper) optical lithography, despite the enormous cost of the soon-to-come iteration of stepper machines (namely, EUV steppers). The main reason is that stepper lithography has so many advantages compared with other technologies that it has proven difficult for other candidate technologies to match them. For example: in the case of nanoimprint, alignment and defect control have been the two critical areas where nanoimprint could not match optical lithography. Simply put, a nanoimprinting lithography machine can do not deliver the same level of alignment accuracy that a stepper does. Another problem is defectivity: it is extremely difficult to keep the master mold clean after a large number of imprints and this inevitably translates into defects and impurities being transferred to the imprinted wafers. Therefore, stepper lithography is superior to nanomprinting lithography in such regards

Q: Since there has been a lot of discussions about the skyrocketing costs of stepper lithography for volume IC and DRAM production, do you think new technologies can provide different solutions?

A: Unfortunately, despite the high costs, I do not see viable alternatives other than stepper lithography for ICs and DRAM production,. There are a few companies out there that are promoting nanoimprinting lithography for IC production but I really do not see this change happening in the next few years, unless some really huge technological breakthrough comes forward. It is true that for stepper technology costs are going up with the future adoption of EUV machines but I do not see any alternative to both immersion lithography and EUV. Maybe direct self-assembly will be useful as a complementary technology but still the main lithography process will be stepper based

Q: Being this the case, what do you think about the future of nanoimprint. Will you see nanoimprinting lithography as a niche market application in the future or there is some hope that nanoimprint will become the technology of choice in some bigger markets?

A: As I was saying before, I believe the future for nanoimprinting lithography is bright. The reason being that many markets that today are regarded as niche or even are yet to be tapped and will grow extensively in the future. Nanoimprinting lithography has a lot of advantages in terms of cost and easiness of process when you do not need high alignment capabilities and you have some tolerance in terms of defects: LED patterning, patterning of substrates for cell culture, production of anti-reflecting films are just some of the markets where nanoimprint can be applied easily. I believe that the existing markets I have just listed and other markets will grow exponentially in the next few years

Q: LED patterning and other potential applications have been discussed for quite a while but still we do not see nanoimprinting lithography widely used. Is there a reason for this?

A: One of the reasons has historically been that most of nanoimprint research has focused on replacing stepper technology for volume IC and DRAM production, a technological application nanoimprint is not really fit for. This misconception has led the industry to ignore other potential applications where nanoimprint could have been the technology of choice as such markets were still too small. On the opposite, I believe that the role of nanoimprint will be exactly that of making such niche market grow and create new exciting applications. Let` s take for example roll-to-rollnanoimprint. It is now possible to pattern literally square meters of polymer with a nanopattern at a cost that is orders of magnitude lower than other technologies. Same goes for LED patterning, nanoimprint can be successfully used to pattern sapphire substrates at a cost unmatched by conventional lithography. I believe such are the markets where we should focus on.

Q: You are talking about roll-to-roll nanoimprint. What is the status of this particular technology?

A: Roll-to-roll nanoimprint is a technology that has been plagued by a number of issues in the past, such as keeping the film flow at a constant rate and at a constant thickness. Keeping thickness variation control has especially been a challenge for wide films. However, I see such hurdles being overcome so I think roll-to-roll will be definitely be used in many applications that require to have polymer surfaces patterned at a low cost and in high quantities, such as in the anti-reflection film market for example.

Q: Thank you for your time

A: Thank you.

Note:
If you are interested in nanoimprint lithography, please visit our nanoimprint lithography service page.

Subscribe to our newsletter to receive our new articles directly in your mail box.

If you liked this article, please give it a quick review in StumbleUpon, Facebook or Pinterest.