TECHNOLOGY FOCUS EUV LiTHOGrapHY
Winner takes all T
he recent announcement that Komatsu will buy out Gigaphoton from its partner Ushio is a major shakeup in the extreme ultraviolet (EUV) lithography
market, a technology deemed next-generation lithography because of its potential to shrink the features printed on an integrated circuit further than current photolithography techniques allow. Gigaphoton was founded in 2000 as a 50/50 joint venture between Komatsu and Ushio to develop excimer laser light sources for lithography tools. Its recent EUV development programme has centred on laser-produced plasma (LPP) technology as a means of generating EUV energy, as opposed to the competing discharge-produced plasma (DPP) method (see Nadya Anscombe’s article in the December 2010/January 2011 issue of Electro Optics for more on LPP and DPP). In addition to its stake in Gigaphoton, Ushio also purchased Philips’ Extreme UV business a year ago through its subsidiary Xtreme Technologies (Aachen, Germany), which focuses on developing DPP sources. Therefore, up until the sale of Gigaphoton, Ushio had an interest in both of the two competing methods for producing EUV light: its subsidiary Xtreme Technologies is working on DPP, while Gigaphoton (Oyama, Japan) is concentrating on LPP. The sale effectively means Ushio has chosen to continue its EUV development along the DPP path, leaving Komatsu to pursue LPP as a means to produce EUV light through its subsidiary Gigaphoton. ‘The ironic situation prior to the sale was that Gigaphoton and Xtreme, two half-brothers, would be in competing segments of the same future market,’ comments Phil Alibrandi, director of sales and account management at Gigaphoton. ‘The sale certainly simplifies things – following a successful 11 year DUV partnership, Ushio and Gigaphoton amicably went their separate ways. We are now competitors without any confusing allegiances.’ Extreme UV light is produced by a high-
temperature plasma, typically derived from tin. The method of doing this is either via DPP using an electrical discharge or LPP using a laser. According to Alibrandi, there might only be room for one of the EUV technologies: ‘The industry has yet to make a decision on which technology, LPP or DPP, will win out. Both demonstrated vendors’ technologies are relatively immature
34 ELECTrO OpTiCS l june 2011
Greg Blackman speaks to Phil Alibrandi of Gigaphoton, about the recent buyout of the company and what that means for the euV lithography community
Gigaphoton’s Lpp system uses an Nd:YaG laser to produce a fine mist of tin particles and then a CO2
laser to produce the plasma. Image courtesy of Gigaphoton
and neither has been proven to have enough power to hit the economic threshold of 60 to 100 wafers an hour production rate. All three competitors [Xtreme, Gigaphoton, and Cymer, the latter developing LPP technology] are working closely with ASML and the industry has to make a decision – is it a LPP world or a DPP world? The data isn’t sufficient, yet, to know either way.’ The technologies are yet to be commercialised and scanner manufacturer ASML is evaluating both LPP and DPP tools. Gigaphoton plans to deliver its first HVM-level LPP tool to ASML later this year. The threshold at which EUV becomes economically viable is generally considered as a production rate of 60 to 100 wafers per hour and one of the points discussed in Nadya Anscombe’s earlier article was the ability to scale the power to reach this level, with spokespeople from both DPP and LPP camps arguing their technology could scale. ‘Gigaphoton LPP is more scalable simply by adding power to the drive laser,’ explains Alibrandi, adding that, from his perspective, it’s unclear how DPP scales with regards to power. ‘Just by the way DPP produces EUV, it would seem impractical for the technology to scale to hundreds of watts of power. It’s unclear
how they’ll do it, but proponents of DPP might have methods that we’re unaware of. Ushio has announced that DPP does scale and we must leave it to them to demonstrate it.’ However, Alibrandi concedes that neither technology has demonstrated enough power yet. ‘Gigaphoton believes that its LPP is scalable with further refinement to the process,’ he comments. The throughput of 60 to 100 wafers an
hour, at which EUV becomes economically viable, is compared to existing deep UV (DUV) photolithography at 193nm. In order to shrink the features on a chip to the 22nm and even the 16nm nodes, clever processing techniques like double or triple patterning have to be employed. Multiple patterning is a lithography technique where scanners make up to five passes per layer compared to one or two with conventional lithography for critical layers. These additional steps make the process expensive, not just in terms of cycle time, but also capital equipment costs. EUV lithography, with a wavelength of 13.5nm, would reduce the processing down to one or two passes – an EUV step and another DUV lithography step – thereby lowering the cost. Gigaphoton has a roadmap agreed with ASML to deliver its LPP technology in Q4 2011. At that point ASML will have three different takes on EUV sources. ‘Everyone is working frantically to prove to ASML and the litho community that their technology is the best – DPP or LPP. And, within the two LPP vendors, Gigaphoton is working to prove it can demonstrate cleaner power and higher conversion efficiency with a better cost of operation,’ Alibrandi states. EUV systems contain a large collector mirror inside the vessel focusing the EUV energy. If that collector mirror becomes soiled with debris, it has to be removed and changed, resulting in costly downtime of the system. Gigaphoton’s goal is that its debris mitigation technology ensures the collector mirror lasts a targeted minimum of one year between replacements. This is vital to maximise machine uptime. ‘When it’s a $100m scanner, it needs to have the maximum uptime,’ comments Alibrandi.
The race between LPP and DPP is ongoing. Ushio and Komatsu have put their cards on the table and plumped for a technology, but we will have to wait and see which one will ultimately be successful. l
www.electrooptics.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36