My failure to respond on Yuli's breakthrough is a function of having to spend a little time working, etc., and less reading all the chat boards. But, now that I have a few minutes, based on my understanding of the historical known limitations with xray lithography, and the semi industry's past resistance to adopting it, I do believe that this is a significant development.
First of all, a bit of background for any newbies: in general, the current state of the art in lithography is using UV or Deep UV lasers (ala Cymer) to etch the fine circuitry on semi wafers. This is referred to as "optical" lithography. Through a lot of tricks, optical technology has been able to get to smaller features sizes than ever expected several years ago. But most people believe that optical litho will reach its limit soon. That is why DARPA (US Govt) and SEMATECH (consortium of semi manufacturers e.g., Intel)) have pumped multi-millions into "Next Generation Lithography". What are the "NGLs"? X-ray, EUV (extreme UV), SCALPEL and E-Beam. Each of these has their financial supporters. In X-ray it has been IBM, commercially, and DARPA, at the Govt.
It is important to note that ALL of the NGLs have some technical hurdles to overcome. With x-ray, the two main hurdles that most people have identified to its adoption are 1) that until JMAR came along, the main work in x-ray was using synchrotrons (huge facilities) that no semi manufacturer wants to build a fab around. Since JMAR has been developing a small footprint system (about the same size as today's optical systems), this should eliminate this concern (assuming JMAR can bring its system to market) and 2) the masks used in the lithography process had to be a 1:1 ratio (i.e., the same feature size on the mask as is projected on the wafer). In contrast, optical and perhaps other NGL technologies can use 5:1 masks, which allows a larger mask to be used to make features a fifth the size.
Until recently, it was accepted that x-ray litho had to use a 1:1 mask. These masks presently either extremely expensive and/or have other technical issues to overcome with them. But even with the mask issues x-ray litho was still of interest and probably on an even competitive basis with the other NGLs.
Why I and others believe this development is significant is because, IF IT WORKS, obviously it eliminates a major hurdle to the adoption of XRL. Some others have posted copies of the paper from the Singapore synchrotron scientists (recall that JMAR's Yuli was previously at Singapore). The following is a link to their earlier announcement on this:
nus.edu.sg
Just as none of us really knows for sure when JMAR will have a fully completed XRL system, we do not really know for sure that the x-ray demagnification which has been reported will ultimately do the trick. But, I am much more comfortable seeing announcements made by scientists reporting results, which are directed to the scientific community. Most of these scientist types are very concerned about their reputation in the scientific community, so I think there is some solid basis in this.
The bottom line: ANYTHING that removes barriers or makes x-ray lithography more acceptable and, in this case, gives x-ray litho a greater competitive advantage over other competing NGL technologies, moves JMAR's XRL system closer to adoption. So far, JMAR has benefited from the $25M+ of DARPA funds to support the development of its XRL system. This has allowed them to solve the many technological hurdles they have faced over the years. However, even if JMAR had already demonstrated the technical merits of its system, it isn't clear that the semi industry would have adopted it or supported it financially until now. If the x-ray de-magnification work pans out, this is one more step toward convincing the semi industry to adopt JMAR's point source X-ray system. |
