Nanoimprint Lithography Based Approach for the Fabrication of Large‐Area, Uniformly‐Oriented Plasmonic Arrays

Localized surface plasmon resonance (LSPR), collective electron density oscillations found in noble metal nanostructures, has been studied extensively over the past decade due to its potential utility as the backbone for a number of photonic technologies capable of controlling light at nanoscopic dimensions well below the diffraction limit. Research in this field has been engendered by the tremendous growth in fabrication methods capable of producing an enormous variety of nanoparticle (NP) systems and nanostructured films. Numerous theoretical and experimental studies have established that LSPR is sensitive to the shape, size, interparticle distance, dielectric environment and material composition of the constituentNPs.Additionally, optical dichroismobserved from well-aligned nanoparticle arrays has demonstrated the polarization dependence of their LSPR response.One of themost promising applications of nanoparticle systems is their use as real-time chemical and biological sensors that originate from the aforementioned LSPR dependence on their dielectric environment. Such systems have been demonstrated using a variety of NP implementations including single-particle, onedimensional and two-dimensional array configurations on transparent substrates as well as solution-based methods. An abundance of nanofabrication techniques have been employed to produce the desired nanostructures utilized in LSPR studies with varied degrees of success as measured by parameters such as monodispersity. A few examples of these techniques include electronbeam lithography, templates, nanosphere lithography (NSL) and colloidal solution-based nanoparticle synthesis—ofwhich the latter two have been used quite extensively. Although NSL and solution-based methods have been effective for fundamental studies of the influence of NP characteristics on LSPR, there are still significant limitations of both techniques that limit their applicability to commercialized LSPR-based applications. Chemical synthesis techniques have the advantage of creating a wide array of exotic nanostructures based onmodification of the reaction parameters such as time, relative concentration of reactants and temperature. However, the monodispersity and reproducibility of desired structures can be difficult to achieve using this method. More importantly, this technique lacks the control of relative NP positioning and orientation in addition to the requirement of novel surface chemistries for the reduction of NP agglomeration and effective substrate attachment. The inability to precisely control the sample-to-sample LSPR response of immobilized NP systems is a severe limitation of this fabrication approach. Nanosphere lithography is an alternative fabrication method introduced byHulteen et al. to produce periodic particle arrays (PPAs) directly on a variety of substrates. This technique utilizes a closed-packed nanosphere mask that permits direct deposition of noble metal NPs onto a substrate through the interstitial regions of themask. NSL has been implemented in a single-layer and double-layer approach with extensive characterization and utilization of triangular nanoparticles resulting from the single-layer method. The precise control of PPA attributes afforded by this technique makes it a promising candidate as a fabrication method relevant to commercialized LSPR applications. Limitations of NSL include issues with surface coverage and the geometric constraints imposed by the nanospheremaskon thePPA lattice structure andNP shape characteristics which reduce the degrees-of-freedom available for the designed LSPR response of the PPA system. In order to address and supplement limitations encountered by a number of currentNP fabricationmethods, we propose the use of Nanoimprint Lithography (NIL) and two-dimensional nanostructure array (nanoblock) molds derived from onedimensional gratings to produce noblemetalNPAs.We believe this approach possesses a number of attributes that will not only enhance the fundamental study of NP systems, but may also play a key role in the production of marketable LSPR technologies. First, NIL is a mold-based, high-throughput and lowcost process capable of patterning large areas with sub-10 nm [*] Prof. L. J. Guo, Dr. J.-S. Kim Department of Electrical Engineering and Computer Science, University of Michigan Solid State Electronics Laboratory, 1301 Beal Ave., Ann Arbor, MI 48109 E-mail: guo@umich.edu