Large-Aperture and Grain-Boundary Engineering through Template

Subscriber access provided by READING UNIV

Article

Large-Aperture and Grain-Boundary Engineering through Template Assisted Metal Dewetting (TeAMeD) for Resonances in the Short Wave Infrared Jonathan Trisno, Liang Xing Lu, Jinfa Ho, Zhaogang Dong, Yong-Wei Zhang, and Joel K.W. K.W. Yang ACS Photonics, Just Accepted Manuscript • DOI: 10.1021/acsphotonics.7b01028 • Publication Date (Web): 31 Oct 2017 Downloaded from http://pubs.acs.org on October 31, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Photonics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

Large-Aperture and Grain-Boundary Engineering through Template Assisted Metal Dewetting (TeAMeD) for Resonances in the Short Wave Infrared Jonathan Trisno1,‡, Liangxing Lu2,‡, Zhaogang Dong3, Jin Fa Ho3, Yong Wei Zhang2,* and Joel K.W. Yang1,3,* 1

2

Singapore University of Technology and Design, 8 Somapah Road, Singapore 487372, Singapore Institute of High Performance Computing, A*STAR, 1 Fusionopolis Way, #16-16 Connexis North, Singapore 138632 3 Institute of Materials Research and Engineering, A*STAR, 2 Fusionopolis Way, Singapore 138634 ‡ These authors contributed equally * To whom correspondence should be addressed. Corresponding emails: [email protected] & [email protected]

Abstract We extend the fabrication method of Template Assisted Metal Dewetting (TeAMeD) to create near-infrared resonant nanostructures in an Au film without the need for etching or lift-off. TeAMeD has previously been used to generate high aspect-ratio sub-10 nm apertures, but struggles to generate larger apertures (>100 nm). In this work, we introduce a method to create larger apertures using templates consisting of fin-like patterns with radial symmetry. We also report evidence of grain boundary engineering, through the template pinning effect. Our threedimensional phase field model of TeAMeD predicts both the grain-boundary pinning and aperture opening effects that agree well with experiments. Combined with simulation design, TeAMeD can be established be as a grain engineering platform, allowing grain shape and boundary position to be controlled. Variations of template motif produces larger grains, and numerous possible outcomes, including suspended Au nanodisks and triangular apertures.

Keywords templated self-assembly, solid state-dewetting, phase-field simulation, grain boundary engineering, metallic hole arrays, IR plasmon resonances

ACS Paragon Plus Environment

ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Solid state dewetting of thin metal films deposited on patterned substrates is a thermodynamically driven process with the ability to produce controlled and ordered nanostructures.1,2 Based on this technique, Template Assisted Metal Dewetting (TeAMeD) has shown promise as an approach to generate ~10 nm apertures in thick (~60 nm) Au,3,4 with unique advantages over conventional methods such as etching, lift-off, and focused ion beam milling.5–7 Attractive features, such as potential reusability of templates, insensitivity of the process to the directionality of the metal deposition, formation of deep apertures with vertical sidewalls, reduced grain boundary and surface roughness, and the preservation of the purity of metals, could be advantageous to applications such as chemical sensing,8–10 extraordinary optical transmission (EOT),11,12 and spectroscopy.13,14 We previously reported a comprehensive study of aperture formation in an Au film deposited to a thickness matching the heights of cylindrical nanoposts of varying diameters. However, due to incomplete dewetting, TeAMeD has thus far not been able to generate apertures with dimensions larger than 100 nm.3 Here we introduce strategies to pattern apertures with larger dimensions (>100 nm) to achieve plasmon resonances. Several apertures with dimensions ranging from 20 nm to 1 µm (for slits) are also demonstrated. The new strategy allows TeAMeD to produce large apertures (>100 nm) while maintaining relatively short e-beam lithography writing time, providing flexibility of patterning arbitrary patterns over method such as interference lithography15,16 and nanosphere lithography17,18. Additionally, the expanded range of fabrication dimension (from micron down to ~10 nm) allows aperture sizes spanning more than an order of magnitude to be fabricated in a single lithographic step. The fabricated hole arrays were designed to exhibit resonances in the Short Wave Infrared (SWIR) wavelength region (0.9-1.7µm) with minimal response in the visible wavelengths. As the Au film is deposited on a Si substrate which has a high index, the metal-Si interface allows EOT to occur at longer wavelengths than with a low index substrate. Tunable resonance in the SWIR region is achieved through the variation of aperture periodicity, allowing for the realization of spectrally distinctive SWIR ‘color’. The resulting pattern will only be visible in SWIR, but remain completely invisible to the human eye. In addition to patterning structures that are visible in the SWIR spectrum, our experimental results show that grain shape and boundary positions can also be controlled by template structures. Grain-boundary growth is known to be affected by the presence of particles, which can induce a grain-pinning effect.19 Dispersion of these particles to control grain recrystallization has been studied.20 We observed that the templates used in TeAMeD exhibit a similar pinning effect. Here we designed several templates to manipulate the grain, with the aim of minimizing the number of grain-boundaries. Along with larger grain sizes, engineering the grain-boundary position could allow the minimization of loss from grain-boundary scattering, thus reducing damping of surface plasmon resonances,21 allowing for sharper resonance peaks for a plasmonic structure. This will be applicable to the field of color printing, spectroscopy techniques such as Surface Enhanced Raman Spectroscopy (SERS) or Surface Enhanced Infrared Absorbance (SEIRA), and light extraction from quantum emitters.22–25 Our results show the promise of finlike structures arranged into different template motifs in achieving a wide range of nanostructures via Au dewetting for applications where the purity of Au, or the minimization and positioning of grain-boundaries are crucial.


Page 2 of 18

Page 3 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

Results and Discussions Large Aperture Generation Figure 1(a) shows the schematic of the fabrication process. Arrays of dots/cross patterns with diameter/linewidth of ~20 nm were patterned in 60-nm thick hydrogen silsesquioxane (HSQ) resist on a Si substrate, cross-linked by Electron Beam Lithography (EBL). HSQ is a high contrast negative E-Beam resist, which will become an optically transparent material that resembles amorphous SiOx upon being cross-linked.26,27 Next, Au with the same thickness as the template was deposited using an electron beam evaporator. Finally, the sample was heated on a hot plate set at 400°C for ~2 hours, instead of a rapid thermal processor as reported previously3. During the heating process, the Au dewets and forms apertures with diameters corresponding to template design. To demonstrate the process, we compare the fabrication results of samples with a dot template and with a cross template. Figures 1(b-I, III) show the templates fabricated by EBL, where the bright structures are cross-linked HSQ. We fabricated dot array templates with diameter of~20 nm and periodicity of 200 nm, as shown in Fig. 1(b-I). Figure 1(b-II) shows high resolution apertures with diameters ~20 nm formed on an Au film by using TeAMeD with the dot template, similar to what we have demonstrated in our previous work3. It is possible to achieve higher resolution (smaller diameters), as the resolution achievable with TeAMeD is defined by the EBL resolution. We realized the patterning of larger apertures through the utilization of a fin-like cross shaped templates. Instead of patterning the whole aperture area on the template, we patterned only a cross template with linewidth of ~20 nm and diameter of ~140 nm, shown in Fig. 1(b-III). Interestingly, instead of forming a cross shaped aperture, the film further dewetted, and formed apertures that had a larger area than the template itself, shown in Fig. 1(b-IV). As one does not need to expose the entire area of the aperture, this strategy will reduce the overall e-beam patterning time.


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. (a) Schematic of the fabrication process. First, dot/cross shaped HSQ (thickness t= 60nm) are patterned using EBL on Si substrate. Next, Au with thickness t is deposited using electron beam evaporator. Finally, by heating at 400°‫ ܥ‬on a hot plate, metal sitting on top of HSQ will dewet away from the nanostructure, and form apertures as a result of capillaryassisted diffusion. (b) SEM images of (I) dot template, (II) ~20 nm diameter aperture arrays, (III) cross template, and (IV) ~130 nm diameter aperture arrays (scale bars=100 nm). We developed a 3D phase field model to simulate the aperture forming process. Based on the model, we observed that the metal-cap (metal that sat on the template) first dewetted from the template top to the pit through atomic diffusion, connecting to the metal that originally sat in the pit. As heating progressed, larger apertures started to form as the metal clustered together. This phenomenon is due to the tendency of thin metallic films deposited on non-wetting substrates to suffer from thermal instability when heated to high temperatures.28 When a metallic film is heated, holes in the metal film will merge with each other, resulting in the agglomeration (dewetting) of the metallic material to form clusters of random metallic islands.29 With TeAMeD, we are able to control the intermediate process as well as the final result through template geometry design. Instead of forming random metallic islands, the template allowed the formation of a more uniform pattern, which in our case was an array of apertures. The dewetting and forming of the larger apertures occurs as the Au film is driven to minimize its surface energy. In Fig. 1(b-IV), the dewetting tends to form aperture edges resembling straight lines instead of curves, as a straight edge has smaller surface energy compared to a curve. The shape/size non-uniformity observed in Fig. 1(b-IV) would most likely be linked to the initial conditions of the as-deposited metal film. Due to the random nature of the polycrystalline gold and its grain boundaries, dewetting is expected to proceed at different rates at each aperture site.


Page 4 of 18

Page 5 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

The choice of the metal film, template pattern, substrate, and annealing environment will affect the interfacial energies at the substrate-metal, template-metal and metal-environment boundaries. In our fabricated structure, the Au-SiOx interface energy was the most significant. The Au-SiOx interfaces occur between Au and the cross-linked HSQ template, and between Au and the thin native oxide that covers the Si substrate. The wetting angle between Au and SiOx is ~140 degrees.30 The large wetting angle means that there is a large interfacial energy between the two materials, which accounts for the tendency of the Au film to cluster together and form metallic islands instead of sticking to the substrate. This property enables the dewetting process to work better, while materials such as Al with a small wetting angle will be harder to dewet. The performance of TeAMeD, however, is also dependent on the height difference between template and metal. To achieve higher aperture yield, the thickness of the metal film has to be optimized. Our experimental results suggest that the metal film thickness needs to be in the range of -20 to +5 nm of the thickness of template. If the metal film is too thick (>5nm), the yield of apertures will become lower, partially caused by the fact that dewetting of thicker films will introduce more material that will cluster together. Thus, the thicker the film, the more space will be required to accommodate the final steady configuration after dewetting. If there is not enough space, the metal will cover the template, resulting in a lower aperture yield. On the other hand, if the metal film is too thin, incomplete dewetting will occur as some metal-caps will cluster on top of the template structures without slipping to the pit. While the process is sensitive to metal thickness, variables such as template linewidth, size, and periodicity will also influence the final state. For example, if the template linewidth is too large, it will be harder for the metal-cap to slip to the pit, as it has enough ground to stick to. Template size and periodicity, together with metal thickness, will influence the final state of dewetting too. Through appropriate template design and thickness control, different final states can be produced. Plasmonic Resonances in SWIR Wavelengths The hole arrays fabricated by TeAMeD exhibit resonances in the Short Wave Infrared (SWIR) wavelengths (0.9-1.7 µm), but with no observable resonances in the visible spectrum. Figure 2 shows the measured and FDTD-simulated reflectances. The reflectance is normalized to the reflectance of Au film on corresponding substrates (without nano-apertures). Figure 2(a) shows measured reflectance spectra in visible range (0.4-0.7 µm). Only minimal change was observed as the period of the apertures array was varied from 250-350 nm. Meanwhile, Fig. 2(b) shows its response in the SWIR, where dips in reflectance at different wavelengths depending on the aperture period was observed. These dips occur due to Extraordinary Optical Transmission (EOT). EOT is a result of Fano resonance,31 where light is able to pass through periodic subwavelength apertures in a metal film.32–34 The resonances of the aperture arrays are tunable in the SWIR by adjusting the periodicity of the apertures. As shown in Fig. 2(b), by increasing the periodicity from 250 nm to 300 nm, we observed that the main resonance dip red-shifted by ~165nm, from 980 nm to 1150 nm. The same result can be seen when the periodicity is increased from 300 nm to 350 nm, as the dip redshifted by ~165 nm, from 1150 nm to 1315 nm. Despite the non-uniform aperture size and shape,


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the experimental results show that resonance wavelengths are indeed tunable, as they are more dependent on periodicity of the array than the diameter of the apertures. The shape nonuniformity will cause weaker resonance dips, as shown in Fig. 2(c), where measured spectrum was compared to 3D FDTD simulation results. Here we see that the resonance dip occurs at similar wavelengths, but the measured curve is broader than the simulation results. As the plasmonic lifetime is dependent on aperture geometry,35 the irregular shapes in our structures will exhibit a spread in lifetimes, thus broader resonances. The spectral dip at 980 nm corresponds to light being extraordinarily transmitted through the aperture array. Figure 2(d) shows 3D FDTD simulation results to compare the effect of different substrates. Aperture arrays (periodicity= 300nm) on Si and glass substrates were compared. When a lowindex substrate such as glass was used, the dip occurred at a visible wavelength of ~600 nm. When the substrate was changed to a high-index Si substrate, the dip corresponding to the same mode was shifted to 1150 nm, in the SWIR region. As the Au film was fabricated on a high index Si substrate, the metal-Si interface enabled EOT to occur at longer wavelengths than with a low index substrate.

Figure 2. Measured and FDTD-Simulated normalized reflectance plots: (a) Measured reflectance in visible spectrum, showing minimal differences for the three periodicities. (b) Measured reflectance in SWIR, showing tunable resonance through periodicity variation. (c)


Page 6 of 18

Page 7 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

Comparison with FDTD simulations shows that the measured reflectance agrees with the simulated result (periodicity of 250 nm). (d) Simulated reflectance for aperture arrays on highindex (Si) and low-index (glass) substrates. Figure 3 shows the simulated normalized E-fields and Poynting vectors at the lateral crosssection (a, c) and Au-SiO2 interface (b, d) planes, at the resonance wavelength of 980 nm. It can be observed that field enhancement (about 4 times) is most prominent at the Au-Si interface, which includes a thin native oxide (2 nm). The strong field at the Au-Si interface instead of the Au-Air interface suggests that light was transmitted through the aperture, an EOT phenomenon due to Fano resonance. This can also be observed in the normalized Poynting vectors (S) in Fig. 3(c).

Figure 3. 3D FDTD-Simulated electric field enhancement plots (a, b) and time-averaged Poynting vector plots (c, d) for a structure with aperture diameter of 148 nm (periodicity =250 nm) at the resonance wavelength of 980 nm. Electric field enhancement calculated as electric field (E) divided by the incident field (Eincident), and the Poynting vector (S) was normalized to the incident Poynting vector (Sincident). Plane wave illumination is incident from above the aperture (in the z-direction) and is polarized along the x-axis. The independent control of the transmission in the visible and SWIR wavelength regions enables for the realization of a metal film with visually concealed SWIR plasmonic resonances. Figure


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

4(a) shows an electron micrograph of an aperture array fabricated using TeAMeD, with a periodicity of 250 nm. This pattern remains invisible under the optical microscope (Fig. 4(b)), but becomes detectable when viewed using an IR detector (Fig. 4(c)). This can be applied as a means to conceal information, where the image will only be visible in SWIR, but remain completely invisible to the human eye.

Figure 4. Array of 250 nm pitched apertures under (a) SEM, (b) Optical Microscope, and (c) false color processed IR image.

Grain Shape and Boundary Engineering

Figure 5. Simulation results showing TeAMeD for cross template from initial stage, to early stage after heating and to final stage after longer heating, where grain control is clearly observed.

Fabrication results show that the grain shape and boundary can be controlled by the HSQ template design. Initial results of TeAMeD show that grain boundaries in Au terminate close to where the ends of the cross meet with Au. This phenomenon became more prominent when the sample has undergone a longer heating, as confirmed by simulation results. Figure 5 shows snapshots of the calculation results at different stages of dewetting. Calculations were done using a method reported previously.4 At first, the metal-caps dewet from the HSQ template crosses. Larger apertures are then formed. As dewetting progresses, the grains become larger, while the


Page 8 of 18

Page 9 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

aperture size remains relatively constant, with grain boundaries only present between edges of the template crosses. In Fig. 5, we can see that the simulation predicts a strong pinning of grain boundaries at the ends of the crosses.

Figure 6. Grain boundary pinning. (a) SEM image of ~140nm diameter aperture fabricated using the cross template and (b) its zoomed image (blue cross overlaid on SEM). Interestingly grain boundaries in Au follow the template shape, and end near the edges of the crosses. Red arrows show the distance between the grain-boundary and the ends of a representative HSQ cross. (c) Histogram shows the shift of grain-boundary from template edge. Forty percent of the boundaries are within 5 nm of the template, showing the potential of this method to be used as a grain-boundary engineering platform. Figure 6(a) shows grain control in the fabricated samples. A detailed image, Fig. 6(b), shows the grain boundaries in Au terminate close to where the ends of the crosses meet with Au, and appear to be pinned. To ascertain if the grain pinning effect was statistically significant, we randomly sample 400 grain boundaries, and measure their distances to the ends of the HSQ template crosses. The resulting measurement is presented as the histogram in Fig. 6(c), which shows that 40% of the boundaries terminate within 5 nm of the HSQ template. A significant correlation between the edge position of the template crosses and grain-boundary positions was observed.


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 18

Figure 7(a) Alternative templating strategy (hexagonal template) with periodicity of 300 nm and diameter of 225 nm fabricated through cross-linking of 60 nm thick HSQ by EBL. (b) A 50 nm Au film was deposited on the sample through E-beam evaporation deposition. (c) After thermally induced dewetting, hexagonal apertures were formed, showing grain-boundary pinning. (d) Interestingly, several grains merged together (Labelled A1, A2), reducing the overall number of grains. (e) Phase field simulation of solid state dewetting on hexagonal patterned substrate. Grain shape and boundary control can also be observed when other template designs are used. Figure 7 shows a hexagonal template formed by three lines with linewidth ~20 nm. Figure 7(c) shows that a hexagonal template will result in the formation of hexagonal apertures. The grain pinning effect can also be observed in Fig. 7(d). Here, the grain boundaries also terminate near the end of the template edges. Interestingly, the merging of grains was observed. The hexagonal shape of the template promotes the joining of two grains, as the hexagon is the native grain shape of Au on the plane. The merging of grains can be observed in regions labeled as A1 and A2. This design potentially allows for the minimization of grain boundaries, producing a film that is close to a monocrystalline film. Some hexagonal regions, however, were still occupied by more than one grain, for example the regions labeled as B1 and B2. These grains that remained were due to the presence of low-angled grain boundary. Figure 7(e) shows the simulation result of dewetting on a hexagon patterned substrate. Compared with Fig. 5, we can see that the pinning of grain boundaries using the hexagonal pattern is weaker than with the cross pattern. Regions that show merging of grains (A1 and A2) and regions that are occupied by more than one grain (B1 and B2) can also be observed in simulation results shown in Fig. 7(e).


Page 11 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

Figure 8. SEM images of apertures (a) metallic slit (LxW= 1 µm x 20 nm), (b) isolated nanostructure, (c) suspended Au nanodisk, (d) trefoil shaped aperture. Inset: corresponding templates used (scale bars=200 nm). Numerous aperture shapes can also be engineered through the variation of TeAMeD motifs. Figure 8 shows various apertures formed through TeAMeD. Figure 8(a) shows a metallic slit with length of 1 µm and width of 20 nm, showing potential of TeAMeD in fabricating large slits with dimension in micron range. Isolated nanostructures can potentially be formed using a mesh-like template, as shown in Fig. 8(b). Figure 8(c) shows an example of a Casie-Baxter state with isolated metal-caps that sit on top of the template crosses, creating a suspended nanodisk. Hexagonal templates can also be modified, e.g. by disconnecting the template, to produce two possible orientations of a set of three triangles as shown in Fig. 8(d). The flexibility of TeAMeD to pattern arbitrary motifs could also be applied to produce nanohole trimers or quadrumers36, as it will be fairly easy to pattern small clusters of apertures. More investigations will be required to improve uniformity of apertures and reduce the minimum separation between holes. Our results show the potential of fabricating structures >100 nm. The largest fabricated dimension is shown in the long axis of a slit, i.e. 1 μm (Fig. 8a). The largest radial dimension


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 18

obtained was 225 nm (Fig. 7c, d). Larger apertures could be fabricated with a suitable designed template that drives Au radially outwards. However, with the current design of radial lines, we expect that Au would still remain between the HSQ lines after dewetting, despite initially detaching from the SiOx interface. We believe that the dewetting of structures in the micron range will be inhibited by incomplete dewetting and a longer dewetting process. The tendency of incomplete dewetting is already apparent on smaller structures as seen on the outer part of the aperture (e.g. hexagonal aperture in Fig. 7d), which result in non-uniform aperture shapes. This incomplete dewetting might be caused by the presence of pinning sites that act as thermodynamic kinetic traps, preventing the system from reaching its final equilibrium state. Forming of larger aperture will also require more metal to be displaced, thus will increase the dewetting time significantly. Conclusions Template Assisted Metal Dewetting (TeAMeD) allows for the patterning of high aspect ratio, high resolution apertures in metal without the need of etching or lift-off. We introduce a new strategy to pattern larger apertures, using partially patterned guiding templates, reducing overall e-beam patterning time. The fin-like template design allows Au to dewet and cluster to form metallic islands in a controlled manner, thus forming arrays of apertures. These apertures exhibit resonances in the Short Wave Infrared (SWIR) wavelength region (0.9-1.7µm) with minimal response in the visible wavelengths. We also observed evidence that suggests the ability to control grain boundary positions. Grain boundaries terminate close to the edges of template crosses, suggesting pinning effect from the sharp corners of the template crosses. Our strategy allows TeAMeD to potentially be a platform for grain engineering, enabling the control of grain shape and boundary positions. Methods Electron Beam Lithography 2% Hydrogen Silsesquioxane (HSQ) was spin coated onto Silicon substrate at 1000rpm rampedup for 2s, to achieve a resist film thickness of 60nm. Resist thickness was measured through Filmetrics F20 thin-film measurements and SEM images of the side wall of collapsed HSQ templates. We used Raith eLine Plus to pattern the template with an accelerating voltage of 30kV in a field size of 100 µm x 100 µm, and step size of 2 nm. High resolution dot arrays were patterned through exposure of single pixel dots with dose of 0.08 pC. To achieve larger apertures, templates were designed through combination of single pixel lines, patterned with dose of 9000 pC/cm. We developed the structure in salty developer26 (1% NaOH+ 4% NaCl wt/wt in DI Water) in room temperature for 60 s with slight agitation. The sample was immediately immersed in DI water for 60s to stop the development. To allow cleaner drying process, the sample was rinsed with IPA before being gently blow dried with a steady stream of N2 gas. Metal Deposition


Page 13 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

We deposited Au with thicknesses of 30 nm to 60 nm using electron beam evaporator (Explorer Coating System, Denton Vacuum, New Jersey, USA) on the patterned sample, without the use of adhesion layers. The thickness of the Au will influence the final state of the apertures, but it generally needs to be close enough (+10 nm) to the thickness of the template for a successful process3. Deposition was performed at a rate of 1 Å/s, chamber process pressure of ~9 x 10-7 torr, and sample holder rotation speed of 12 rpm. Thermal Dewetting The sample was placed on a Corning #6798-400D hot-plate to induce heat and promote the dewetting process. The temperature was set at 400o C, with actual temperature of 350 o C measured by an IR thermometer. The sample was covered by a beaker glass to allow more uniform ambient temperatures, and heated for 2 hours. After heating, the sample was immediately put on a room temperature stainless-steel plate for 1 minute to cool down. Optical Measurement We performed bright field optical microscopy and reflection spectra measurements to investigate the optical responses of the fabricated structures. We used a UV−visible-NIR microspectrophotometer (CRAIC Technology Inc.) at normal incidence with a 75 W broadband xenon source to perform extinction spectrometry in reflection mode. Samples were illuminated with normal incident white light, using a 36x, 0.5 NA objective lens. The aperture was set to a detecting area of 7.1 µm × 7.1 µm. The reflected light was captured by spectrometer and CCD camera, recording reflectance spectrum at λ = 400−1600 nm and optical images of the sample respectively. The false color IR image was produced through post-processing steps, based on reflection spectrum collected from microspectrophotometer in SWIR region (0.7-1.7 µm). To show the ‘color’ in the false IR image, the recorded SWIR reflectance spectrums, both for patterned and unpatterned area, were mapped to visible range (0.4-0.75 µm), before calculating the corresponding xyz chromaticity coordinates. Dewetting and Grain-Boundary Simulation We developed a 3D phase field model to incorporate the dewetting process and grain-boundary evolution. In this model, two sets of phase field variables were employed to represent the phases and grains respectively. For distinguishing different phases, we used three conserved variables: [ϕ1,ϕ2 ,ϕ3 ] equal to [1,0,0] , [0,1,0] , [0,0,1] to represent the vapor phase, the metal film and the substrate, respectively. For distinguishing different grains, we used a set of non-conserved variables ηi with each of them represents for a random orientation. In our phase field code, the variables of ηi were stored as sparse matrix, so that the maximal number of grain orientations can be infinite with no limitations. In our phase field model, the total free energy of the coupled system is expressed as: m   F tot = ∫  gϕ (ϕi ) + p2 ⋅ gη (ηi ) + B(1 − p2 )∑ηi2 + f grad  dV (1) i =1  


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 18

where, gϕ (ϕi ) and gη (ηi ) are bulk free energy of pure phase system and pure grain system, respectively, f grad is the gradient energy, p2 is a interpolation function representing the volume fraction of the metal film in a given position. The third term in the integration is a coupling term which constrains the metal grains in the metal film. The detailed expressions of gϕ (ϕi ) , gη (ηi ) ,

f grad and p2 are: 3

gϕ (ϕi ) = ∑ (α ijϕi2ϕ 2j )

(2)

i< j

gη (ηi ) =

1 m 2 3 m−1 m −1 (ηi − 1)2 + ∑ηi2η 2j − ∑ 4 i =1 4 i< j 4 3

m −1

i< j

i< j

f grad = −∑ ( kijϕ ∇ϕi ∇ϕ j ) − ∑ ( k η ∇ηi∇η j )

(3) (4)

p2 =ϕ22 (5) Based on the total free energy in Eq. (1), the kinetic evolutions of phases and grains are described by: ∂ϕ2 δ F tot = ∇M ∇ (6) δϕ2 ∂τ ∂ηi δ F tot = −L ∂τ δηi where, M and L are the related mobilities.

(7)

3D FDTD Simulation Commercial finite-difference time domain (FDTD) solver (Lumerical FDTD Solutions) was used to model the aperture arrays. The simulated structure was an aperture in Au film, on Silicon substrate with 2 nm native oxide layer in between. In the directions of periodicity (x and y directions), the simulation span was set to the size of 1 unit cell (equal to periodicity). To better match the experimental structure, the aperture was modeled as a diamond shape, with cross shaped SiO2 template in the middle. Mesh size of 2.5 nm was used for the simulation, with the mesh in native oxide region set to 0.5 nm in the z direction. The refractive index for Au was selected from Werner37, while Si and SiO2 were selected from Palik38 from the Lumerical FDTD Material database. The source was set as a plane wave (polarized in x axis) normal incidence to substrate surface. Field monitors were placed above the source, under the aperture, and at the cross sections of aperture to record the reflectance, transmittance, electric field, and power distributions. Perfectly matched layer (PML) boundaries were used for top and bottom sides (xy planes), while anti-symmetric and symmetric boundaries were used for xz and yz planes respectively to model the periodic array of apertures, while shortening the simulation time. Acknowledgement We acknowledge financial funding from Singapore’s National Research Foundation (NRF) (Grant No. NRF2015NRF-CRP001-021) and Digital Manufacturing and Design (DManD)


Page 15 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

(Grant No. RGDM1530302). Authors would like to thank Singapore University of Technology and Design (SUTD), Institute of High Performance Computing (IHPC), Institute of Material Research and Engineering (IMRE), and A*STAR Computational Resource Centre (A*CRC) for tools and facilities provided. Supporting Information Phase-field simulation showing evolution of grain coarsening, dewetting, and aperture opening for gold film on cross-template. References (1) Kim, D.; Giermann, A. L.; Thompson, C. V. Solid-State Dewetting of Patterned Thin Films. Appl. Phys. Lett. 2009, 95 (25). (2) Wang, D.; Ji, R.; Schaaf, P. Formation of Precise 2D Au Particle Arrays via Thermally Induced Dewetting on Pre-Patterned Substrates. Beilstein J. Nanotechnol. 2011, 2 (1), 318–326. (3) Wang, Y. M.; Lu, L.; Srinivasan, B. M.; Asbahi, M.; Zhang, Y. W.; Yang, J. K. W. High Aspect Ratio 10-Nm-Scale Nanoaperture Arrays with Template-Guided Metal Dewetting. Sci. Rep. 2015, 5, 9654. (4) Lu, L.; Wang, Y.; Srinivasan, B. M.; Asbahi, M.; Yang, J. K. W.; Zhang, Y.-W. Nanostructure Formation by Controlled Dewetting on Patterned Substrates: A Combined Theoretical, Modeling and Experimental Study. Sci. Rep. 2016, 6 (1), 32398. (5) Melli, M.; Polyakov, A.; Gargas, D.; Huynh, C.; Scipioni, L.; Bao, W.; Ogletree, D. F.; Schuck, P. J.; Cabrini, S.; Weber-Bargioni, A. Reaching the Theoretical Resonance Quality Factor Limit in Coaxial Plasmonic Nanoresonators Fabricated by Helium Ion Lithography. Nano Lett. 2013, 13 (6), 2687–2691. (6) Duan, H.; Hu, H.; Hui, H. K.; Shen, Z.; Yang, J. K. W. Free-Standing Sub-10 Nm Nanostencils for the Definition of Gaps in Plasmonic Antennas. Nanotechnology 2013, 24 (18), 185301. (7) Chen, X.; Park, H.-R.; Pelton, M.; Piao, X.; Lindquist, N. C.; Im, H.; Kim, Y. J.; Ahn, J. S.; Ahn, K. J.; Park, N.; Kim, D.-S.; Oh, S.-H. Atomic Layer Lithography of Wafer-Scale Nanogap Arrays for Extreme Confinement of Electromagnetic Waves. Nat. Commun. 2013, 4, 2361. (8) Stark, P. R. H.; Halleck, A. E.; Larson, D. N. Short Order Nanohole Arrays in Metals for Highly Sensitive Probing of Local Indices of Refraction as the Basis for a Highly Multiplexed Biosensor Technology. Methods 2005, 37 (1), 37–47. (9) Neubrech, F.; Huck, C.; Weber, K.; Pucci, A.; Giessen, H. Surface-Enhanced Infrared Spectroscopy Using Resonant Nanoantennas. Chemical Reviews. 2017, pp 5110–5145. (10) Gordon, R.; Sinton, D.; Kavanagh, K. L.; Brolo, A. G. A New Generation of Sensors Based on Extraordinary Optical Transmission. Acc. Chem. Res. 2008, 41 (8), 1049–1057. (11) Degiron, A.; Lezec, H. J.; Yamamoto, N.; Ebbesen, T. W. Optical Transmission Properties of a Single Subwavelength Aperture in a Real Metal. Opt. Commun. 2004, 239 (1), 61–66. (12) Lesuffleur, A.; Im, H.; Lindquist, N. C.; Lim, K. S.; Oh, S.-H. Laser-Illuminated Nanohole Arrays for Multiplex Plasmonic Microarray Sensing. Opt. Express 2008, 16 (1), 219–224. (13) Brolo, A. G.; Kwok, S. C.; Moffitt, M. G.; Gordon, R.; Riordon, J.; Kavanagh, K. L.


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(14)

(15) (16)

(17)

(18) (19) (20) (21)

(22) (23) (24) (25)

(26)

(27)

(28) (29)

(30) (31)

Enhanced Fluorescence from Arrays of Nanoholes in a Gold Film. J. Am. Chem. Soc. 2005, 127 (42), 14936–14941. Coe, J. V.; Rodriguez, K. R.; Teeters-Kennedy, S.; Cilwa, K.; Heer, J.; Tian, H.; Williams, S. M. Metal Films with Arrays of Tiny Holes: Spectroscopy with Infrared Plasmonic Scaffolding. J. Phys. Chem. C 2007, 111 (47), 17459–17472. Luo, X.; Ishihara, T. Surface Plasmon Resonant Interference Nanolithography Technique. Appl. Phys. Lett. 2004, 84 (23), 4780–4782. Li, J.; Iu, H.; Lei, D. Y.; Wan, J. T. K.; Xu, J. B.; Ho, H. P.; Waye, M. Y.; Ong, H. C. Dependence of Surface Plasmon Lifetimes on the Hole Size in Two-Dimensional Metallic Arrays. Appl. Phys. Lett. 2009, 94 (18). Haynes, C. L.; Van Duyne, R. P. Nanosphere Lithography: A Versatile Nanofabrication Tool for Studies of Size-Dependent Nanoparticle Optics. J. Phys. Chem. B 2001, 105 (24), 5599–5611. Cheung, C. L.; Nikolić, R. J.; Reinhardt, C. E.; Wang, T. F. Fabrication of Nanopillars by Nanosphere Lithography. Nanotechnology 2006, 17 (5), 1339–1343. Smith, C. S. Zener Pinning. Trans. Met. Soc. AIME 1948, 15–51. Longworth, H. P.; Thompson, C. V. Abnormal Grain Growth in Aluminum Alloy Thin Films. J. Appl. Phys. 1991, 69 (7), 3929–3940. Bosman, M.; Zhang, L.; Duan, H.; Tan, S. F.; Nijhuis, C. A.; Qiu, C.-W.; Yang, J. K. W. Encapsulated Annealing: Enhancing the Plasmon Quality Factor in LithographicallyDefined Nanostructures. Sci. Rep. 2014, 4, 5537. Kumar, K.; Duan, H.; Hegde, R. S.; Koh, S. C. W.; Wei, J. N.; Yang, J. K. W. Printing Colour at the Optical Diffraction Limit. Nat. Nanotechnol. 2012, 7 (9), 557–561. Nie, S.; Emory, S. R. Probing Single Molecules and Single Nanoparticles by SurfaceEnhanced Raman Scattering. Science (80-. ). 1997, 275 (5303), 1102 LP-1106. Nishijima, Y.; Nigorinuma, H.; Rosa, L.; Juodkazis, S. Selective Enhancement of Infrared Absorption with Metal Hole Arrays. Opt. Mater. Express 2012, 2 (10), 1367. Curto, A. G.; Volpe, G.; Taminiau, T. H.; Kreuzer, M. P.; Quidant, R.; van Hulst, N. F. Unidirectional Emission of a Quantum Dot Coupled to a Nanoantenna. Science (80-. ). 2010, 329 (5994), 930 LP-933. Yang, J. K. W.; Berggren, K. K. Using High-Contrast Salty Development of Hydrogen Silsesquioxane for Sub-10-Nm Half-Pitch Lithography. J. Vac. Sci. Technol. B Microelectron. Nanom. Struct. Process. Meas. Phenom. 2007, 25 (6), 2025. Yang, J. K. W.; Cord, B.; Duan, H.; Berggren, K. K.; Klingfus, J.; Nam, S.-W.; Kim, K.B.; Rooks, M. J. Understanding of Hydrogen Silsesquioxane Electron Resist for Sub-5Nm-Half-Pitch Lithography. J. Vac. Sci. Technol. B Microelectron. Nanom. Struct. Process. Meas. Phenom. 2009, 27 (6), 2622–2627. Thompson, C. V. Solid-State Dewetting of Thin Films. Annu. Rev. Mater. Res. 2012, 42 (1), 399–434. Gadkari, P. R.; Warren, A. P.; Todi, R. M.; Petrova, R. V.; Coffey, K. R. Comparison of the Agglomeration Behavior of Thin Metallic Films on SiO2. J. Vac. Sci. Technol. A Vacuum, Surfaces, Film. 2005, 23 (4), 1152–1161. Ressel, B.; Prince, K. C.; Heun, S.; Homma, Y. Wetting of Si Surfaces by Au-Si Liquid Alloys. J. Appl. Phys. 2003, 93 (7), 3886–3892. Liu, H.; Lalanne, P. Microscopic Theory of the Extraordinary Optical Transmission. Nature 2008, 452 (7188), 728–731.


Page 16 of 18

Page 17 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Photonics

(32) (33)

(34)

(35)

(36)

(37)

(38)

Ebbesen, T. W.; Lezec, H. J.; Ghaemi, H. F.; Thio, T.; Wolff, P. A. Extraordinary Optical Transmission through Sub-Wavelength Hole Arrays. Nature 1998, 391 (6668), 667–669. Li, W.; Hu, J.; Chou, S. Y. Extraordinary Light Transmission through Opaque Thin Metal Film with Subwavelength Holes Blocked by Metal Disks. Opt. Express 2011, 19 (21), 21098–21108. Li, W.-D.; Ding, F.; Hu, J.; Chou, S. Y. Three-Dimensional Cavity Nanoantenna Coupled Plasmonic Nanodots for Ultrahigh and Uniform Surface-Enhanced Raman Scattering over Large Area. Opt. Express 2011, 19 (5), 3925–3936. Lei, D. Y.; Li, J.; Fernández-Domínguez, A. I.; Ong, H. C.; Maier, S. A. Geometry Dependence of Surface Plasmon Polariton Lifetimes in Nanohole Arrays. ACS Nano 2010, 4 (1), 432–438. Zhan, Y.; Lei, D. Y.; Li, X.; Maier, S. A. Plasmonic Fano Resonances in Nanohole Quadrumers for Ultra-Sensitive Refractive Index Sensing. Nanoscale 2014, 6 (9), 4705– 4715. Werner, W. S. M.; Glantschnig, K.; Ambrosch-Draxl, C. Optical Constants and Inelastic Electron-Scattering Data for 17 Elemental Metals. J. Phys. Chem. Ref. Data 2009, 38 (4), 1013–1092. Palik, E. D. Handbook of Optical Constants of Solids; Academic Press: San Diego, 1998.


ACS Photonics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 18

For Table of Contents (TOC) Use Only

Title: Large-Aperture and Grain-Boundary Engineering through Template Assisted Metal Dewetting (TeAMeD) for Resonances in the Short Wave Infrared Authors: Jonathan Trisno, Liangxing Lu, Zhaogang Dong, Jin Fa Ho, Yong Wei Zhang, and Joel K.W. Yang The table of contents (TOC) graphic shows modeling results of samples before and after annealing through Template Assisted Metal Dewetting (TeAMeD) using fin-like cross shaped templates. The annealed sample produces large apertures (~100 nm) and exhibits grain-boundary pinning by the template. Optical micrograph of the sample which consists of a square array of apertures (~140 nm) with a periodicity of 250 nm on a Si substrate, shows low visibility between the patterned and surrounding region. However, the calculated color-mapped image of the same sample shows that pattern is detectable when imaged in the SWIR spectrum.


Large-Aperture and Grain-Boundary Engineering through Template

Recommend Documents