Imprint as the next NGL

By Mike Watts

For imprint watchers, SPIE’s Advanced Microlithography in February was very interesting, in that there were several presentations that appeared to be linked, but the presenters denied that there were links ….which makes me think !

Toshiba (Tokyo, Japan) started the ball rolling with a detailed status report on their evaluation of the Molecular Imprints (MII – Austin TX)  Step and Repeat imprint tool. They showed lots of data and a frank assessment that the system  was not ready for production yet, but  that all the key performance characteristics were close. Throughput and overlay were within 2 x of target. Defects were 100x from target and were the critical barrier. Finally work was needed on a number of infrastructure issues.  When questions from the audience asked “did that mean that Toshiba believed that the problems were fixable or that they were so daunting that they were giving up ? “ , the author hedged !

Sematech (Albany NY) then  reported  on their MII tool that is a similar but a slightly older  rev tool. Their data was very similar. They had values for random defectivity, after removal of defects from known sources, that met the magic 0.1 cm^-2 number. Canon (Tokyo, Japan) stood up and showed more data of their evaluations and some improvements they had made to  their MII system. Finally, MII presented on a number of technical improvements focused on throughput and defectivity, and further data that was consistent with the 3 prior presenters.  They also announced that they had shipped 4 tools this year and one of those was an imprint module that was integrated in to a stepper by their “equipment partner”.  Another system was a mask replication tool that went to Dai Nippon Printing (DNP – Tokyo, Japan).

So there is  a major Japanese end user, and  a major Japanese stepper company both evaluating MII equipment , and MII working with a “equipment partner” who is integrating the MII module into a stepper. All 3 are working with DNP to obtain templates, who bought a mold replication system. The obvious conclusion is that there is a 4 way program here, and one that looks like it makes sense if there is a serious implementation program going on.

The second  observation  is that the data from these multiple sources all looks pretty consistent, although it was not clear which incremental improvements were behind which data. Finally, imprint is  really close to being viable. Questions from the audience suggested that the defect densities were comparable to EUV. Finally the reports from multiple multibeam e-beam  companies showed that they are still factors of 1000 away from throughput targets. All in all, it was a pretty positive story from the imprint stand point.

Full disclosure – as an ex- founding team member of MII,  I still have some shares but no special insights.

About the Author

Mike Watts has been patterning since 1 um was the critical barrier, in other words, for a longtime. I am a tall limey who is failing to develop a Texas accent here in Austin. I have a consulting shingle at www.impattern.com.

My blog “ImPatterning” will focus on the latest developments in the business and technology of patterning. I am particularly interested in trying to identify how the latest commercial applications will evolve.