We present dynamic membrane projection lithography as a method to create three dimensional metallic traces in hemispherical cavities. The technique entails directional evaporation through perforations in a membrane covering a hemispherical unit-cell cavity. The sample is positioned on a rotating stage and tilted with respect to the incident evaporated beam, such that the traces are deposited on the interior face of the cavity. A simple self-aligned version and a more general two-step fabrication version are presented. Furthermore, by incorporating a fixed shutter, both closed-loop and split-loop structures are demonstrated.
©2011 Optical Society of America
The field of optics continues to evolve, aided in large part by continuing advances in our ability to fabricate structures with ever smaller dimensions. Emerging disciplines within optics such as photonic crystals [1,2], plasmonics , nano-antennas [4,5] and metamaterials  all leverage various subtleties of the interaction of incident electromagnetic fields with materials structured on ever-smaller size scales. At the same time, each of these sub-disciplines is motivated by unique applications and performance metrics, serves to refine our understanding of the mechanisms at play in the field-material interaction and helps identify new paradigms for manipulating light. The electromagnetic behavior of a nano-antenna can be modeled as the superposition of current loops which respond to incident magnetic fields and charge-separating dipole elements which respond to incident electric fields. The precise electromagnetic behavior depends on many different physical parameters including the symmetry, size scale and linewidths of the antenna. As the design wavelength shrinks to infrared and optical wavelengths, the ability to control and manipulate these physical features stresses even cutting edge lithography capabilities. Additionally, conventional semiconductor processing techniques are not capable of creating truly 3-D nano-antenna structures without resorting to stacking and layer-by-layer approaches. Chemical synthesis and colloidal self-assembly routes continue to progress, displaying a vast array of nano-scale shapes and morphologies , but typically yield structures with resonances in the visible and NIR wavelength range, and generate assemblies with geometries and physical arrangements subject to the statistical thermodynamics of the particular self-assembly route used to produce the structures.
Recently the role of symmetry and symmetry breaking has been identified as a key component in the scattering response of nano-antennas [8,9]. In addition, several routes to the production of nano-antenna structures with reduced or broken symmetries have been developed. Nano-caps, formed by casting PDMS over polystyrene spheres coated with evaporated metal, possess plasmon resonances which correlate with their shape and demonstrate hybridized electromagnetic modes [10,11]. Furthermore, the far field radiation patterns of these particles can be correlated to these specific resonances, and hence tuned to achieve angular control over emission, demonstrating the utility of creating 3-D nano-antennas with engineered symmetry . One limitation of this approach is that once a particular polystyrene sphere diameter and angle of evaporation is chosen, the resulting nano-antennas consist of blanket deposited metal over the surface of the sphere exposed to the incident evaporated metal flux.
We present a variant of membrane projection lithography (MPL) [12,13] as a fabrication approach capable of creating 3D metallic traces in hemispheroidal cavities containing multiple metallic loops on the interior surface of the cavity. The fabrication approach is general; capable of creating intricate antenna designs rather than simple hemispheroidal sections. Previous versions of MPL used a membrane patterned with the final desired resonator pattern and performed the angled evaporations on a fixed substrate. The present technique patterns the membrane with a finite number of holes which serve as deposition apertures, and then tilts and continuously rotates the substrate in-plane during the evaporation. Although the membrane is fixed to the cavity, the combination of tilt angle and rotation results in the deposition of a metallic trace on the interior face of the cavity where the line-of-sight through the aperture intersects the cavity face throughout the 2π rotation. The trace geometry can be engineered by selection of the number, shape, size and relative spacing of the deposition apertures with respect to the cavity perimeter. This technique is similar in some ways to nano-stencil lithography and dynamic stencil lithography, but here is applied to deposition in a three dimensional cavity with the membrane (stencil) fixed to the substrate rather than the separate stencil and planar substrates used in the nano-stencil work [14–16].
Formation of the suspended, patterned membrane can take one of two routes: 1) a general, two step fabrication process, or 2) a less general but simpler, single step self-aligned fabrication process. Figure 1 (Media 1) shows the two-step process for a single unit-cell with three deposition apertures. A layer of a barrier material is deposited on a substrate with a known isotropic dissolution etch (XeF2 for silicon, hydrofluoric acid for oxide, etc.) The barrier material must be resistant to the substrate etch chemistry, but have its own dissolution mechanism with excellent selectivity to the substrate. Fortunately a variety of materials can satisfy these requirements (photoresist, PMMA, oxide, nitride and metallic films.) The barrier layer is then patterned with an array of small holes, exposing the underlying substrate. The substrate is then etched through the patterned holes using the isotropic substrate etch chemistry so that hemispherical cavities are formed beneath the barrier layer, centered under the patterned holes. The barrier layer is removed, and the cavities are backfilled and planarized with a sacrificial material. A membrane material is deposited on the substrate and patterned with a single or multiple small holes over each underlying cavity. The perforations serve a dual-purpose, allowing removal of the sacrificial material and serving to define the deposited trace(s) in the interior of the cavity during the evaporation step (covered in detail below).
For the self-aligned fabrication approach, the barrier material used to define the cavity is patterned with the final suite of perforations in each unit-cell, so that immediately after the isotropic substrate etch (Fig. 1 Step D) to create the cavities, the sample is ready for evaporation. Depending on the symmetry of the perforation pattern and isotropy of the substrate etch, an approximately hemispherical cavity forms under the membrane. At this point in either fabrication approach, the angled, rotating deposition can take place. The advantage of the self-aligned process is its simplicity, while the two-step process is capable of more general patterns and offers more control over the cavity and metallic trace shape.
Figure 2 contains a series of schematic images of the evaporation step for a single centered hole above a hemispherical cavity. For clarity, only a single unit-cell is shown, however the method generalizes large area arrays. In Fig. 2, the source is tilted with respect to the surface normal of the membrane, and then continuously rotated in-plane during evaporation. After deposition, the membrane is coated with a metal layer of thickness t. The metal incident on the perforation proceeds in a line-of-sight fashion until it impinges on the interior face of the cavity and is deposited on the wall, so that the thickness of the metal trace is given by t/L,where L is the total path length of the metallic trace. The linewidth of the deposited antenna trace is a function of the membrane perforation diameter, membrane thickness and evaporation tilt angle. By increasing the deposition thickness, the trace thickness can be increased, however clogging can occur for thick evaporations, and must be managed. The reentrant profile of the suspended membrane makes removal of the membrane straightforward.
While it is possible to create the tilted, rotating source of Fig. 2, a more practical arrangement is to place the sample on a rotation stage, tilted with respect to the evaporation source by an angle,ϕ, and then continuously rotate the stage during deposition (Fig. 3 ). Provided the deposition source is sufficiently far from the turntable, the incident flux is parallel throughout the entire rotation. Figure 4A shows a sequence of time-lapse schematic images demonstrating this deposition process to create complete closed loops (inset). In addition to simplifying the deposition apparatus, using the tilted rotation stage also allows for a very simple modification to enable fabrication of split rings. By positioning a single or multiple fixed shutter(s) in the deposition chamber, incident metallic flux is shadowed from the aperture during the portion of the revolution while the sample is under the shutter, opening a gap in the deposited trace (Fig. 4B). In this configuration, the size of the gap is controlled by the area of the rotation stage subtended by the shutter. The result is the ability to create antennas with both inductive loops and capacitive gaps.
The geometry in Fig. 5 is used to predict the resulting deposited trace pattern given a bowl radius, incident angle, and the number and position of membrane perforations. The hemispherical bowl is described by:Fig. 5, and θ represents the angle of in-plane rotation throughout the revolution of the rotation stage. The location of the deposition at a given ϕ, θ is found by substituting Eqs. (2) and (3) into Eq. (1), solving for parameter t, and substituting this result back into Eq. (2) to find the 3-dimensional coordinates of the trace at that point. Looping θ from 0 to 2π yields the entire trace perimeter. Multiple perforations are taken into account by including multiple traces and changing the x1, y1, and z1 to reflect their position above the cavity.
Figure 6 demonstrates the versatility of dynamic MPL. Figure 6A shows the top down view of a cavity of radius 1.8 μm with 3 perforations in the triangular configuration shown, with an evaporation tilt angle of 20 degrees. The three traces reside near the bottom of the cavity (Fig. 6B) and hence can be approximated by a planar model. Figure 6C shows the resulting traces for an identical cavity, but this time the perforations are moved near the rim of the cavity and the evaporation is at 45 degrees. The resulting traces occupy most of the bowl surface area, and hence are fully 3-dimensional antennas (Fig. 6D).
As a demonstration of the self-aligned version of dynamic membrane projection lithography fabrication, a 6 μm thick layer of base developable polyimide (ProLift 100-24) was spin-coated on a silicon wafer and softbaked at 200 °C for 5 minutes. A 225 nm thick PMMA layer was spin-coated on and baked at 175 °C for e-beam patterning. Samples with 5 x 5 mm rectangular arrays of perforation groupings were patterned on an 8 μm pitch. The individual perforations were square holes 300 nm per side and the groupings of 1-4 holes/unit-cell werearranged as shown in Fig. 7A . Timed submersion in standard TMAH based i-Line developer (AZ400K 1:4, RohmHaas) dissolved out the polyimide directly under the perforations, resulting in the formation of roughly hemispherical cavities beneath the patterns. A 20 nm titanium/200 nm gold evaporation was performed at 30 degrees in a Temescal FC2000 e-beam evaporator with ~65cm separating the crucible from the sample. The crucible is ~4.2 cm in diameter and holds 25 cc of material. Acetone liftoff was used to remove the gold coated PMMA membrane. Figure 7B shows a 3x3 unit-cell section of the sample with 4 perforations/unit-cell. Except for sporadic liftoff related defects, the sample was uniform over the entire 5mm X 5mm area. Figure 8 contains higher magnification SEM images of individual unit-cells for cases with one, two and three perforations created using either the standard evaporation (Figs. 8A–8C) or the shuttered evaporation (Figs. 8D–8F).
Several aspects of fabrication are apparent from Fig. 8. In the self-aligned fabrication approach used to make these structures, the cavity shape is obviously affected by the arrangement of the perforations as evidenced by the oblong cavity in Fig. 8B for the two loop antenna, and the rounded triangular cavity in Fig. 8C surrounding the 3-Loop antenna. Second, the traces are fairly thin. The ~1μm diameter rings have a perimeter of ~3 μm, so that the trace thickness should be ~10 nm thick. Finally, being a self-aligned process, where the same perforations that create the cavity are used to create the traces, the perforations cannot be placed arbitrarily inside the unit-cell to create extremely non-planar antennas as in Fig. 6B. This shortcoming is remedied by the two-step fabrication process outlined earlier. Attempts to measure the optical properties of these structures were complicated by the large pitch relative to the anticipated operational wavelength, small trace thickness, and polyimide material absorption bands. We are currently investigating two-step fabricated structures in hemispherical silicon cavities on a smaller pitch with thicker traces to address these issues, and design functional nano-antennas.
We have presented a new variant of membrane projection lithography, dynamic membrane projection lithography, capable of creating three-dimensional metallic inclusions. The method can produce complicated arrangements of loops as well as structures with gaps, so that both electric field and magnetic field excitation are possible. The technique provides several mechanisms to exercise control over the symmetry for creation of antennas. Asymmetrically arranged shutters can be included to create traces with asymmetric gaps. In the two-step fabrication process, the source apertures need not be symmetrically disposed about the cavity, creating further asymmetry if desired. Creation of these fully three-dimensional structures offers many variables to the optical antenna designer. For instance, the highly non-planar 3-Loop structure of Figs. 6C–6D would experience significant self inductance for a normally incident TEM plane wave as well as mutual inductance from the neighboring loops. Both of these effects are drastically increased by the significant out-of-plane portion of the current loops, absent in planar structures. Such flexibility is certain to be important given the challenges which must be addressed in creating nano-antennas capable of interfacing to nano-scale sources .
This work was performed, in part, at the Center for Integrated Nanotechnologies, a U. S. Department of Energy, Office of Basic Energy Sciences user facility. Supported by the Laboratory Directed Research and Development program at Sandia National Laboratories. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy’s National Nuclear Security Administration under Contract DE-AC04-94AL85000.
References and links
9. M. Liu, T.-W. Lee, S. K. Gray, P. Guyot-Sionnest, and M. Pelton, “Excitation of dark plasmons in metal nanoparticles by a localized emitter,” Phys. Rev. Lett. 102(10), 107401 (2009). [CrossRef] [PubMed]
10. J. Liu, A. I. Maaroof, L. Wieczorek, and M. B. Cortie, “Fabrication of hollow metal nanocaps and their red-shifted optical absorption spectra,” Adv. Mater. (Deerfield Beach Fla.) 17(10), 1276–1281 (2005). [CrossRef]
12. D. B. Burckel, J. R. Wendt, G. A. Ten Eyck, J. C. Ginn, A. R. Ellis, I. Brener, and M. B. Sinclair, “Micrometer-scale cubic unit cell 3D metamaterial layers,” Adv. Mater. (Deerfield Beach Fla.) 22(44), 5053–5057 (2010). [CrossRef] [PubMed]
13. D. B. Burckel, J. R. Wendt, G. A. Ten Eyck, A. R. Ellis, I. Brener, and M. B. Sinclair, “Fabrication of 3D metamaterial resonators using self-aligned membrane projection lithography,” Adv. Mater. (Deerfield Beach Fla.) 22(29), 3171–3175 (2010). [CrossRef] [PubMed]
14. M. Graff, S. K. Mohanty, E. Moss, and A. B. Frazier, “Microstenciling: a generic technology for microscale patterning of vapor deposited materials,” J. Microelectromech. Syst. 13(6), 956–962 (2004). [CrossRef]
15. N. Takano, L. M. Doeswijk, M. A. F. Boogaart, J. Auerswald, H. F. Knapp, O. Dubochet, T. Hessler, and J. Brugger, “Fabrication of metallic patterns by microstencil lithography on polymer surfaces suitable as microelectrodes in integrated microfluidic systems,” J. Micromech. Microeng. 16(8), 1606–1613 (2006). [CrossRef]
16. S. Aksu, A. A. Yanik, R. Adato, A. Artar, M. Huang, and H. Altug, “High-throughput nanofabrication of infrared plasmonic nanoantenna arrays for vibrational nanospectroscopy,” Nano Lett. 10(7), 2511–2518 (2010). [CrossRef] [PubMed]
17. L. Novotny and N. van Hulst, “Antennas for light,” Nat. Photonics 5(2), 83–90 (2011). [CrossRef]