Bottom-up nanopatterning and nanofabrication by soft replication of spontaneously faceted crystal surfaces
Gabai-malka, Racheli; Ismach, Ariel; Joselevich, Ernesto
Israel

The large-scale patterning of materials and molecules down to nanometer feature sizes is a critical issue of nanoscience and nanotechnology. Photolithography is limited by the wavelength of light, which is 193 nm for the Deep-UV technology used today in industry, while Extreme-UV technology under development can reach 50 nm feature sizes, aiming for 32 nm in 2009. These limitations have motivated an extensive exploration of alternative lithographies based on self-assembly. Soft lithography is a versatile approach for the replication of patterns into a broad variety of media, but the original pattern, as in photolithography, must usually be generated by other methods. Nanopattern generation methods include electron-beam lithography and scanning probe nanolithographies, such as dip-pen and constructive nanolithography, which can reach feature sizes down to 5 nm, 30 nm and 9 nm, respectively. However, these nanopattern generation processes are serial, and hence extremely time-consuming for large areas. We present a new type of nanopatterning and nanofabrication based on the spontaneous faceting of unstable crystal surfaces. In the present study, periodic arrays of highly straight and parallel nanofacets are generated by annealing different unstable faces of sapphire (α-Al2O3). These self-generated nanopatterns are replicated by a standard soft lithography procedure to produce patterned self-assembled monolayers (SAMs), that we called "nanofacet lithography" (NFL).
NFL allowed us to form periodic linear and crossbar nanostructures on Au and Si/SiO2 substrates on a large scale, and to pattern SAMs into straight and parallel lines as thin as 10 nm with a pitch as small as 40 nm over 1 cm2 in a few seconds, as well as Au nanowires and nanogrids, and Si V-nanogroves and nanowaffles. Just as a reference, producing a pattern with similar dimensions by standard electron-beam or scanning probe nanolithography would take, respectively, 9 months or 80 years.
Reference: R. Gabai, A. Ismach, and E. Joselevich. Adv. Mat. 2007, in press.
back