Toriumi-san is on a roll. He has for two years now reported on theoretical calculations designed to discover the dominant reaction channels opened up during exposure of resist by high energy electrons. [1, 2]
In simplest terms, electron beam exposure is highly inefficient compared with optical exposure of resists. An optical resist is designed so that a photon has a high probability of exciting a resist component molecule into a reactive state, from which the desired reaction (chain) occurs. In scattering theory terminology, only one “channel” is opened by the incoming photon.
Such simplicity is not possible in EUV or in electron exposure. These energy sources open a huge number of channels, with little regard to the precise chemical make-up of the resist material. Specifically, EUV and electron beam dump most of their reactive energy into the polymer matrix rather than into photo-active additives. The process creates all manner of excited states.
To get an idea of how difficult it is to sort out the resulting mess, it suffices to go to http://scitation.aip.org/ and enter the search string ((acid generation)\<and\>(electron beam)). Among other things, it was found long ago that electron beam exposure leads to cross-linking of the polymer matrix. This runs contrary to the chain scission that one wants in a positive tone resist. Interestingly, electron beam exposure also leads to chain scission. To some extent, the polymer can be designed for greater resistance to one or the other reaction, but this may not be compatible with chemical amplification.
In addition, only a small fraction of the electron energy is transferred to the resist on first pass. Many things have been tried, in order to increase the cross-section of the initial energy transfer. Among the simpler, adding halogens to the resist introduces another whole set of competing reactions, usually poisoning the desired reactions. Adding heavy metals results in X-ray generation, with attendant blurring.
So what to do?
Perhaps we have to leave behind chemical amplification for the next generations of lithography. [3-6]
References:
[1] Theoretical analysis of energy degradation of electrons in the resists
Minoru Toriumi
Proc. SPIE 7273, 72732X (2009), DOI:10.1117/12.813940
http://spiedl.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=PSISDG00727300000172732X000001&idtype=cvips&gifs=yes&ref=no
[2] Theoretical analysis of energy dissipation of electron in the resists II
Minoru Toriumi
Proc. SPIE 7639, 76392N (2010), DOI:10.1117/12.846503
http://spiedl.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=PSISDG00763900000176392N000001&idtype=cvips&gifs=yes&ref=no
[3] Negative-type extreme ultraviolet Resist Materials based on Water-Wheel-like Cyclic Oligomer (Noria)
Hiroyuki Seki, Yuki Kato, Hiroto Kudo, Hiroaki Oizumi, Toshiro Itani and Tadatomi Nishikubo
Jpn. J. Appl. Phys. 49 (2010) 06GF06, DOI: 10.1143/JJAP.49.06GF06
http://jjap.ipap.jp/link?JJAP/49/06GF06/
[4] Development of New Positive-Tone Molecular Resists Based on Fullerene Derivatives for Extreme Ultraviolet Lithography
Hiroaki Oizumi, Katsutomo Tanaka, Kiminori Kawakami and Toshiro Itani
Jpn. J. Appl. Phys. 49 (2010) 06GF04, DOI: 10.1143/JJAP.49.06GF04
http://jjap.ipap.jp/link?JJAP/49/06GF04/
[5] Alternatives to chemical amplification for 193nm lithography
Burak Baylav, Meng Zhao, Ran Yin, Peng Xie, Chris Scholz, Bruce Smith, Thomas Smith, and Paul Zimmerman,
Proc. SPIE 7639, 763915 (2010), DOI:10.1117/12.846924
http://spiedl.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=PSISDG007639000001763915000001&idtype=cvips&gifs=Yes&ref=no
[6] Characteristics of main chain decomposable star shaped polymer on EUV lithography
Taku Hirayama, Jun Iwashita, Sachiko Yoshizawa, Kenri Konno, and Takeshi Iwai
Proc. SPIE 7639, 76390Q (2010), DOI:10.1117/12.848439
http://spiedl.aip.org/getabs/servlet/GetabsServlet?prog=normal&id=PSISDG00763900000176390Q000001&idtype=cvips&gifs=Yes&&ref=no