High Aspect Subdiffraction-Limit Photolithography via a Silver Superlens

Feb 29, 2012 - Institute of Materials Research and Engineering, Agency for Science, Technology and Research ... semiconductor and data storage industr...
0 downloads 0 Views 2MB Size
Download PDF
Letter pubs.acs.org/NanoLett

High Aspect Subdiffraction-Limit Photolithography via a Silver Superlens Hong Liu,† Bing Wang,† Lin Ke,† Jie Deng,† Chan Choy Chum,† Siew Lang Teo,† Lu Shen,† Stefan A. Maier,*,‡ and Jinghua Teng*,† †

Institute of Materials Research and Engineering, Agency for Science, Technology and Research (A*STAR), 3 Research Link, Singapore 117602 ‡ Department of Physics, Imperial College London, London SW7 2AZ, United Kingdom S Supporting Information *

ABSTRACT: Photolithography is the technology of choice for mass patterning in semiconductor and data storage industries. Superlenses have demonstrated the capability of subdiffraction-limit imaging and been envisioned as a promising technology for potential nanophotolithography. Unfortunately, subdiffraction-limit patterns generated by current superlenses exhibited poor profile depth far below the requirement for photolithography. Here, we report an experimental demonstration of sub-50 nm resolution nanophotolithography via a smooth silver superlens with a high aspect profile of ∼45 nm, as well as grayscale subdiffraction-limit three-dimensional nanopatterning. Theoretical analysis and simulation show that smooth interfaces play a critical role. Superlens-based lithography can be integrated with conventional UV photolithography systems to endow them with the capability of nanophotolithography, which could provide a cost-effective approach for large scale and rapid nanopatterning. KEYWORDS: Superlens, surface plasmons, nanophotolithography, subdiffraction-limit, surface roughness, aspect ratio The superlens concept, first proposed by Pendry in 2000,20 has drawn numerous interests. The first experimental demonstration of a thin slab silver superlens for near-field subdiffraction-limit imaging in the UV region was presented in 2005.21 This experiment became a milestone on the journey heading toward super resolution optical imaging.20−25 The concept relies on the generation of surface plasmon polaritons enhancing the evanescent fields to restore the near-field components of the Fourier decomposition of the source object, hence breaking the diffraction limit.26 Index-matching in nearfield superlens imaging systems is also widely applied to further improve the restauration of high-wave vector imaging components.21,27−29 Ag is the most commonly used superlens material due to its intrinsic low damping loss from the UV to the visible range as compared to other noble metals.30 A smooth Ag film is critical to reduce scattering loss to avoid the loss of the subwavelength information encoded in base wave vectors. Techniques to reduce surface roughness in Ag and Agdielectric thin film systems have been actively explored.31,32 Indeed, a smooth silver superlens with a thin germanium seed layer has succeeded in resolving half-pitch grating structures as small as 30 nm, which is so far the best image resolution achieved experimentally.29

N

anopatterning is becoming a critical step in many materials and device applications. A number of nanopatterning techniques have been developed, such as electron-1,2 and ion-3 beam lithography (EBL and IBL), focused ion beam (FIB) milling,4,5 dip pen lithography,6,7 and scanning probe lithography.8−10 However, these techniques and tools are costly and their throughputs are very low in terms of large-scale patterning. To break through the bottleneck, multiple11,12 and massively parallel13 electron beam lithography have been investigated to shorten the cycle time and increase the throughput, but are still facing the problem of beam positioning drift induced by thermal fluctuations, electrical charging, and magnetic field interference, etc.14 Nanoimprint lithography (NIL)15−18 offers a low cost and high throughput approach for large area patterning with nanometer scale resolution in comparison with those aforementioned nanolithography techniques. However, issues such as accuracy, defect control, alignment, uniformity, yield, and so forth still need further improvements before NIL can be used in mass fabrication. Optical lithography or photolithography has been the dominant patterning technology in the advancement of semiconductor industries for several decades due to its large scale and high throughput characteristics. With state-of-the-art deep UV (DUV) and extreme UV (EUV) photolithography tools, the pattern resolution has been reached sub-30 nm. However, the cost of the tools and associated processing are escalating, which has posed economic concerns.19 © 2012 American Chemical Society

Received: December 14, 2011 Revised: February 21, 2012 Published: February 29, 2012 1549

dx.doi.org/10.1021/nl2044088 | Nano Lett. 2012, 12, 1549−1554

Nano Letters

Letter

deposition of a 35 nm thick Ag thin film. The rms surface roughnesses measured by AFM were 0.42 and 0.48 nm for the polymer spacer initial surface and after reflow, and 1.60 nm for the Ag surface, respectively. AFM images of the polymer spacer and Ag layer are shown in Figure S1 in the Supporting Information. Eventually, optical lithography was performed to transfer the mask pattern into photoresist through 365 nm UV light exposure and development. The resist patterns were characterized through AFM scanning. In Scheme 3D, depicted in Figure 1b, the Cr grating has a constant gap of 45 nm and pitch of 700 nm but its line width varies from 55 to 135 nm with a step of 20 nm in one grating period. During the ion milling process, the EBL resist on top of the small slits (55, 75, and 95 nm) was etched faster than that on the large ones (115 and 135 nm) due to the loading effect.34 Subject to a certain etching time, the small slits were hence over etched down to about 18 nm depth, while the large slits remained about 40 nm deep. Thus, Cr 3D nanostructures having L-shaped 2D in-plane gratings with varied grating heights were created to serve as a grayscale mask. The same planarization and Ag deposition processes as used in Scheme 2D were then applied. The surface profiles of the polymer spacer and the Ag superlens layer were the same as those in Scheme 2D, as shown in Figure S1 in the Supporting Information. Subject to UV exposure, the amount of light transmitted through this 3D grayscale mask was spatially modulated.35 After enhancement by surface plasmons across the Ag superlens, the distinct optical intensities over the corresponding locations were recorded in the photoresist. In such a way, 3D subdiffraction-limit patterns were produced through a combination of superlens and grayscale lithography. The fabrication and characterization procedures of 2D and 3D nanostructures are detailed in Section 2, Supporting Information. Figure 2 shows surface characterization results of the 2D Cr grating (mask) and the corresponding patterns recorded in a negative photoresist layer. Figure 2a is a scanning electron microscopy (SEM) image of the 2D Cr grating and the inset is

On top of subdiffraction-limit imaging, the superlens concept has also been envisioned for use in optical lithography.33 The aforementioned superlens experiments21,27−29 were indeed conducted via resist exposure through a thin metallic layer, that is, in a photolithographic manner. Unfortunately, all the attained subdiffraction-limit grating patterns in photoresist exhibited a poor profile depth of less than 10 nm, leading to poor image contrast evidenced by the presented atomic force microscopy (AFM) images and corresponding cross-sectional analysis. Table S1 in the Supporting Information summarizes the key features of mask and profile depth in resist pattern collected from those experiments. Apparently, the poor depth profile makes a superlens hardly a convincing tool for photolithography in which a sufficiently high aspect ratio is a fundamental prerequisite. Furthermore, up to this point all the resist patterns have been limited to two-dimensional (2D). Additional recording of information in the longitudinal direction is highly desirable for many applications, such as optical data storage and optical information processing. In this paper, we report experimental demonstrations of subdiffraction-limit 2D and grayscale 3D nanophotolithography with high aspect and high fidelity via a near-field silver superlens. Equipped with critically controlled smooth interfaces, our system has remarkably minimized information loss via enabling high efficiency surface plasmon excitation to enhance the evanescent fields. The thin flat smooth silver superlens is capable of not only transferring sub-50 nm 2D grating features from the mask with a close to five times improvement in aspect profile compared to previous reports but also 3D patterns in combination with grayscale lithography with high pattern fidelity ratio. The results explicitly unveil the applicability of a superlens for subdiffraction-limit optical lithography. Figure 1 summarizes the three main fabrication steps to achieve high aspect subdiffraction-limit patterns in 2D and 3D,

Figure 1. Schematics of three main fabrication steps of subdiffractionlimit nanolithography using Superlens. (a) The 2D scheme to transfer the L-shape, 100 nm pitch, and 40 nm thick Cr grating patterns into photoresist. (b) The 3D scheme to transfer the 700 nm pitch gratings with linewidths varied from 55 to 135 nm in a step of 20 nm separated by a constant gap of 45 nm into photoresist.

Figure 2. Surface characterizations of 2D mask and its corresponding resist pattern. (a) SEM picture of 100 nm pitch Cr grating mask (inset: the cross sectional image was taken at 60° tilt angle). (b) AFM image (1 × 0.5 μm2) of the Cr grating mask, the inset shows its crosssectional profile. (c) The 3D plot of surface contour of Cr grating mask. (d) AFM image (3 × 3 μm2) of corresponding resist pattern; inset plots its cross-sectional profile. (e) The 3D plot of the surface profile of resist pattern. (The color scales for all AFM images are from 0 to 100 nm.)

respectively. In scheme 2D, depicted in Figure 1a, 40 nm thick chrome was initially deposited onto a quartz substrate through electron beam evaporation. An L-shape 100 nm-pitch grating separated by a constant gap of 45 nm was patterned on resist by EBL, which was subsequently transferred into the Cr film by ion milling to form the mask. After that, the mask was planarized with a 20 nm thick polymer spacer followed by 1550

dx.doi.org/10.1021/nl2044088 | Nano Lett. 2012, 12, 1549−1554

Nano Letters

Letter

its cross sectional image. Figure 2b shows a close-up AFM image (1 × 0.5 μm2) confirming a pitch of 100 nm. The crosssectional contour is plotted in the inset and the average height of the gratings has been measured to about 40 nm, close to the targeted thickness in Cr film deposition. A close-up 3D AFM image in Figure 2c clearly depicts the grating profile. Figure 2d shows a plan view of the acquired pattern (3 × 3 μm2) in resist and its cross-sectional contour in the inset. The average depth or altitude of the resist is measured to around 45 nm, which can be seen clearly in the 3D surface contour in Figure 2e, a big leap from the 6.5) and sub-50 nm resolution only if the surface roughness is well controlled to be less than 2 nm. Without a superlens, the control sample of 138.6 nm fwhm is not able to resolve the mask. One should note that the photolithography process has been performed based on the technical data provided by the resist manufacturer, which offers a relatively wide processing window. Hence, achieving such high aspect subdiffraction-limit patterns is due to the well controlled critical interfacial roughness rather than process optimization. The approach of PSF computation and the influence of surface roughness on superlens are detailed in Supporting Information. Using the finite-difference timedomain (FDTD) method, Figure 3c compares the electric field intensities for 2D nanostructures at various distances from the superlens surface as a function of lateral position. As expected, the electric field intensity at the surface (d = 0) exhibits the highest amplitude under the resonated evanescent field. It decays with the distance from the superlens surface but keeps its profile without divergence. Ultimately, the contrast of electric field intensity becomes negligible at about 100 nm away in the current superlens configuration. It theoretically implies 1551

dx.doi.org/10.1021/nl2044088 | Nano Lett. 2012, 12, 1549−1554

Nano Letters

Letter

peaks in the valley, following the topography of mask. The peaks on the plateau exhibit an average height of about 25 nm, and those in the valley are about 18 nm in depth. Within each period, the absolute altitude is averaged at about 64 nm. Figure 5a plots the experimental cross-sectional contours of grayscale mask and the 3D resist pattern in one period for a

that such a near-field superlens lithography approach is able to achieve a subdiffraction-limit pattern with a depth profile approaching ∼100 nm. The surface characterizations of the grayscale Cr mask and the corresponding 3D resist patterns are depicted in Figure 4.

Figure 5. The analysis and simulation of 3D subdiffraction-limit patterning. (a) The measured cross-sectional profiles of mask and resist pattern, which clearly shows the individual subdiffraction-limit grating profile with a high pattern fidelity ratio. (b) The 3D contour of computational distribution of electric field intensity (|E|2) at a distance of 40 nm from an Ag superlens.

Figure 4. Surface characterization of grayscale mask and its corresponding resist pattern. (a) SEM picture of grayscale mask (inset: the cross sectional image was taken at 60° tilt angle). (b) AFM image (6 × 6 μm2) of mask, the inset shows its cross-sectional profile. (c) The 3D plot of surface contour of the mask (d) AFM image (6 × 6 μm2) of the 710 nm-period L-shaped resist pattern; inset plots its cross sectional profile. (e) The 3D plot of surface contour of resist pattern, which verifies a 3D subdiffraction-limit pattern has been achieved. (The color scales for all AFM images are from 0 to 100 nm.)

clear comparison. The threshold points over one period of mask (1′−14′) and resist pattern (1−14) have been individually labeled for identification. The three peaks in the valley (2′, 4′, and 6′) of grayscale mask correspond to the small slits (55, 75, and 95 nm) of about 18 nm thick while the four peaks on the plateau (8′, 10′ and 12′, 14′) correspond to the large slits (115 and 135 nm) of about 40 nm thickness. This shows that the profile of resist pattern is the reverse of that of mask. After examining all the five slit profiles in resist pattern, it shows that each of them exhibits a subdiffraction-limit resolution profile with a small discrepancy in terms of fwhm. This analysis hence confirms the capability for subdiffraction-limit photolithography in three dimensions with high aspect resist profile and high pattern fidelity. Using the FDTD method, Figure 5b plots a 3D contour of the electric field intensity distribution (|E|2) at a distance of 40 nm from the superlens on the resist side when the 3D mask is subjected to UV illumination. Since the UV light is unpolarized, computations have been implemented for light incidence with different linear polarizations. The field intensity profile clearly shows that the periodicity of the field intensity is almost equal to that of the mask while the intensity after penetrating the thinner Cr slits is stronger than that of the thicker slits. By analyzing the field intensity value at a distance of 40 nm after the Ag superlens, it is found that the electric field intensity above the thin Cr slits and their adjacent gaps are about 5.5 and 3.7 times higher than those above thick Cr slits and adjacent gaps, respectively. Therefore, it shows that the varied width and thickness of the slits in the 3D mask and their different distances from the superlens make the electric field intensity spatially modulated at corresponding locations. The 3D electric field intensity profile due to the different trasmissivity of the grayscale mask is subsequently transferred to the negative photoresist after development, producing a 3D resist profile with the thicker resist on top of the thinner Cr slits. The result validates that 3D high aspect subdiffraction-limit pattern is

Figure 4a shows the SEM image of the Cr mask and its cross sectional image in the inset, which shows the Cr at the opening area has been completely etched without any residual layer left in the valley. Figure 4b shows the corresponding AFM image of the Cr mask containing gratings of different linewidths. The inset shows each periodic contour contains four peaks sitting on the plateau, and three peaks in the valley. A 3D AFM contour plot (Figure 4c) facilitates such an intuitional visualization of the periodic slits with alternating altitudes. It confirms the formation of grayscale Cr masks with alternating grating depths generated by the aforementioned loading effect. Referring to the SEM image in Figure 4a and its inset, the line edges of large slits (115 and 135 nm) actually correspond to these four peaks on the plateau, due to the much more severe edge effect during etching, while the three peaks in the valley correspond to small slits (55, 75, and 95 nm) in each period (Figure 4c). It should be noted that ∼5 nm depth modulation has been observed between the large slits and ∼2 nm modulation among the small ones. To simplify our simulation work later, an average Cr thickness of 40 nm is used for the large slits, and 18 nm for the small slits. The AFM images of resist pattern is depicted over an area of 6 × 6 μm2 in Figure 4d, which exhibits an L-shaped slit pattern with alternating altitudes, similar to the mask profile in Figure 4b. A Fourier spectrum analysis of the cross sectional profile of the image, plotted in the inset, confirms an average periodicity of about 710 nm, slightly larger than the 700 nm periodicity of the mask. The small discrepancy of about 10 nm can be attributed to material loss. A more elaborative 3D topographic contour of resist pattern is illustrated in Figure 4e. Each periodic contour of resist pattern contains four peaks on the plateau and three 1552

dx.doi.org/10.1021/nl2044088 | Nano Lett. 2012, 12, 1549−1554

Nano Letters

Letter

(5) Si, G.; Zhao, Y.; Liu, H.; Teo, S.; Zhang, M.; Huang, T. J.; Danner, A. J.; Teng, J. Annular aperture array based color filter. Appl. Phys. Lett. 2011, 99, 033105. (6) Piner, R. D.; Zhu, J.; Xu, F.; Hong, S.; Mirkin, C. A. Dip pen nanolithography. Science 1999, 283, 661−663. (7) Zhang, H.; Chung, S. W.; Mirkin, C. A. Fabrication of sub-50 nm solid-state nanostructures on the basis of dip-pen nanolithography. Nano Lett. 2003, 3, 43−45. (8) Yin, X.; Fang, N.; Zhang, X.; Martini, I. B.; Schwartz, B. J. Near field two-photon nanolithography using an apertureless optical probe. Appl. Phys. Lett. 2002, 81, 3663−3665. (9) Suez, I.; Backer, Scott A.; Frechet, M. J. Generating an etch resistant resist layer from common solvents using scanning probe lithography in a fluid cell. Nano Lett. 2005, 5, 321−324. (10) Shin, M.; Kwon, C.; Kim, S. K.; Kim, H. J.; Roh, Y.; Hong, B.; Park, J. B.; Lee, H. Formation of λ-DNA’s in parallel- and crossed-line arrays by molecular combing and scanning-probe lithography. Nano Lett. 2006, 6, 1334−1338. (11) Muraki., M.; Gotoh., S. New concept for high-throughput multielectron beam direct write system. J. Vac. Sci. Technol. B. 2000, 18, 3061−3066. (12) Chang, T. H. P.; Mankos, M.; Lee, K. Y.; Muray, L. P. Multiple electron-beam lithography. Microelectron. Eng. 2001, 57−58, 117−135. (13) Su, M. S.; Tsai, K. Y.; Lu, Y. C.; kuo, Y. H.; Pei, T. H.; Yen, J. Y. Architecture for next-generation massively parallel maskless lithography system (MPML2). Proc. SPIE 2010, 7637, 76371Q. (14) Chen, S. Y.; Tsai, K. Y.; Ng, P. C. W.; Ng, H. T.; Liu, C. H.; Shen, Y. T.; Kuan, C. H.; Chen, Y. Y.; Kuo, Y. H.; Wu, C. j.; Yen, J. Y. In situ beam drift detection using a two-dimensional electron-beam position monitoring system for multi-electron-beam-direct-write lithography. J. Vac. Sci. Technol. B 2011, 29, 041607. (15) Chou, S. Y.; Krauss, P. R.; Renstrom, P. J. Imprint lithography with 25-nanometer resolution. Science 1996, 272, 85−87. (16) Chou, S. Y.; Keimel, C.; Gu, J. Ultrafast and direct imprint of nanostructures in silicon. Nature 2002, 417, 835−837. (17) Wu, W.; Tong, W. M.; Bartman, J.; Chen, Y. F.; Walmsley, R.; Yu, Z. N.; Xia, Q. F.; Park, I.; Picciotto, C.; Gao, J.; Wang, S. Y.; Morecroft, D.; Yang, J.; Berggren, K. K.; Williams, R. S. Sub-10 nm nanoimprint lithography by wafer bowing. Nano Lett. 2008, 8, 3865− 3869. (18) Kreindl, G.; Glinsner, T.; Miller, R. Next-generation lithography: making a good impression. Nat. Photonics 2010, 4, 27−28. (19) International Technology Roadmap for Semiconductors (ITRS) 2009, http://www.itrs.net/links/2009itrs/home2009.htm (accessed September 8, 2011). (20) Pendry, J. B. Negative refraction makes a perfect lens. Phys. Rev. Lett. 2000, 85, 3966−3969. (21) Fang, N.; Lee, H.; Sun, C.; Zhang, X. Sub-diffraction-limited optical imaging with a silver superlens. Science 2005, 308, 534−537. (22) Taubner, T.; Korobkin, D.; Urzhumov, Y.; Shvets, G.; Hillenbrand, R. Near-field microscopy through a SiC superlens. Science 2006, 313, 1595. (23) Liu, Z.; Lee, H.; Xiong, Y.; Sun, C.; Zhang, X. Far-field optical hyperlens magnifying sub-diffraction-limited objects. Science 2007, 315, 1686. (24) Smolyaninov, I. I.; Huang, Y. J.; Davis, C. C. Magnifying superlens in the visible frequency range. Science 2007, 315, 1699− 1701. (25) Kehr, S. C.; Liu, Y. M.; Martin, L. W.; Yu, P.; Gajek, M.; Yang, S. Y.; Yang, C. H.; Wenzel, M. T.; Jacob, R.; Ribbeck, H. G. V.; Helm, M.; Zhang, X.; Eng, L. M.; Ramesh, R. Near-field examination of perovskite-based superlenses and superlens-enhanced probe-object coupling. Nat. Commun. 2011, 2, DOI: 10.1038/ncomms1249. (26) Kik, P. G.; Maier, S. A.; Atwater, H. A. Image resolution of surface-plasmon-mediated near-field focusing with planar metal films in three dimensions using finite-linewidth dipole sources. Phys. Rev. B 2004, 69, 045418. (27) Melville, D. O. S.; Blaikie, R. J. Super-resolution imaging through a planar silver layer. Opt. Express 2005, 13, 2127−2134.

attainable through this superlens structure. In contrast, the morphological profile of a control sample plotted in Figure S4 (Supporting Information) shows a very poor resolution and contrast, by which the subdiffraction-limit objects cannot be clearly resolved without a superlens. In conclusion, we have experimentally demonstrated sub-50 nm resolution nanophotolithography by using a smooth Ag superlens under 365 nm UV light in a conventional photolithography setup. The pattern exhibits a high aspect profile of ∼45 nm and a high fidelity ratio of 0.6 with a superlens surface roughness controlled at 1.6 nm. Numerical calculations of the roughness effect reveal that a smooth Ag surface is critical for achieving high performance superlens lithography. We have also demonstrated that the Ag superlens was capable of performing 3D subdiffraction-limit grayscale lithography. The results show that a superlens could be a novel and powerful tool for subdiffraction-limit, high throughput, and cost-effective nanophotolithography for applications in semiconductor and data storage industries.

■

ASSOCIATED CONTENT

S Supporting Information *

Lengthy experimental details of the sample preparation, fabrication, and characterization together with theoretical simulation and analysis. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*(S.A.M.) E-mail: [email protected]. Tel: +44-20-75946063. Fax: +44-20-7594-2077. (J.T.) E-mail: [email protected]. Tel: +65-6874-8590. Fax: +65-6872-0785. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge Dr. M. S. Zhang from Data Storage Institute, A*STAR and Mr. G. Y. Si from National University of Singapore for their help on ion-milling process as well as Dr. E. S. P. Leong and Ms. A. B. Chew from IMRE for their help on superlens process. The work is financially supported by the Agency for Science, Technology and Research (A*STAR) under Grants 0921540099 and 0921450030, the U.K. Physical Sciences and Engineering Research Council (EPSRC) and Leverhulme Trust.

■

REFERENCES

(1) 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-5-nm-half-pitch lithography. J. Vac. Sci. Technol. B 2009, 27, 2622−2627. (2) Duan, H.; Winston, D.; Yang, J. K. W.; Cord, B. M.; Manfrinato, V. R.; Berggren, K. K. Sub-10-nm half-pitch electron-beam lithography by using poly(methyl methacrylate) as a negative resist. J. Vac. Sci. Technol. B 2010, 28, C6C58−C6C62. (3) Winston, D.; Manfrinato, V. R.; Nicaise, S. M.; Cheong, L. L.; Duan, H.; Ferranti, D.; Marshman, J.; McVey, S.; Stern, L.; Notte, J.; Berggren, K. K. Neon ion beam lithography. Nano Lett. 2011, 11 (10), 4343−4347. (4) Wu, M. C.; Aziz, A.; Witt, J. D. S.; Hickey, M. C.; Ali, M.; Marrows, C. H.; Hickey, B. J.; Blamire, M. G. Structural and functional analysis of nanopillar spin electronic devices fabricated by 3D focused ion beam lithography. Nanotechnology 2008, 19, 485305. 1553

dx.doi.org/10.1021/nl2044088 | Nano Lett. 2012, 12, 1549−1554

Nano Letters

Letter

(28) Jeppesen, C.; Nielsen, R. B.; Boltasseva, A.; Xiao, S.; Mortensen, N. A.; Kristensen, A. Thin film Ag superlens towards lab-on-a-chip integration. Opt. Express 2009, 17, 22543−22552. (29) Chaturvedi, P.; Wu, W.; Logeeswaran, V. J.; Yu, Z.; Islam, M. S.; Wang, S. Y.; Williams, R. S.; Fang, N. A smooth optical superlens. Appl. Phys. Lett. 2010, 96, 043102. (30) Palik, E. D., Ed.; Handbook of Optical Constants of Solids; Academic Press: New York, 1991. (31) Liu, H.; Wang, B.; Leong, E. S. P.; Yang, P.; Zong, Y.; Si., G.; Teng, J.; Maier, S. A. Enhanced surface plasmon resonance on a smooth silver film with a seed growth layer. ACS Nano 2010, 4, 3139− 3146. (32) Leong, E. S. P.; Liu, Y.; Wang, B.; Teng, J. Effect of surface morphology on the optical properties in metal-dielectric-metal thin film systems. ACS Appl. Mater. Interfaces 2011, 3, 1148−1153. (33) Zhang, X.; Liu, Z. W. Superlens to overcome the diffraction limit. Nat. Mater. 2008, 7, 435−441. (34) Bourouina, M.; Masuzawa, T.; Fujita, H. The MEMSNAS process: microloading effect for micromachining 3-D structures of nearly all shapes. J. Microelectromech. Syst. 2004, 13, 190−199. (35) Kim, J.; Joy, D. C.; Lee, S. Y. Controlling resist thickness and etch depth for fabrication of 3D structures in electron-beam grayscale lithography. Microelectron. Eng. 2007, 84, 2859−2864. (36) Lee, E.; Hahn, J. W. The effects of photoresist contrast on the exposure profiles obtained with evanescent fields of nanoapertures. J. Appl. Phys. 2008, 103, 083550. (37) Hakkarainen, T.; Setala, T.; Friberg, A. T. Near-field imaging of point dipole with silver superlens. Appl. Phys. B: Laser Opt. 2010, 101, 731−734. (38) Hakkarainen, T.; Setala, T.; Friberg, A. T. Subwavelength electromagnetic near-field imaging of point dipole with metamaterial nanoslab. J. Opt. Soc. Am. A 2009, 26, 2226−2234. (39) Micro Resist Technology GmbH, mr-I T85-Thermoplastic Polymer for Nanoimprinting Lithography. www.microresist.com, 2011. (40) Madou, M. J. Fundamentals of Microfabrication, 2nd ed.; CRC: New York; 2002; Chapter 1, p 9. (41) Yang, X.; Zeng, B.; Wang, C.; Luo, X. Breaking the feature sizes down to sub-22 nm by plasmonic interference lithography using dielectric-metal multilayer. Opt. Express 2009, 17, 21560−21565.

1554

dx.doi.org/10.1021/nl2044088 | Nano Lett. 2012, 12, 1549−1554