This work demonstrates a form of arrayed transmitting apertureless near-field photolithography, called apertureless beam pen lithography. An array of fully chromium-coated polyurethane acrylate (PUA) pyramidal microstructures was illuminated by a traditional Ultraviolet (UV) lamp to generate an array of massive UV beam pens for realizing apertureless beam pen lithography. Experimental results reveal that significant UV energy can pass through the apex of a fully metal-coated PUA pyramid even though the thickness of the metallic coating exceeded the penetration depth. The patterned photoresist profiles were 117 nm deep and the full-width-at-half-magnitude (FWHM) was 180 nm when the exposure dosage was 54 mJ/cm2 and the wavelength was 365 nm. Both depth and FWHM increased with exposure dosage, implying that the profiles depended on exposure dosage rather than on physical imprinting. With the adjustment of the thickness of the photoresist layer and the exposure parameters, the lift-off process yields arrayed metal dots with a diameter of 300 nm. Finite-element simulation of the intensity distribution near the apex of the pyramid and within the photoresist layer was carried out to reveal that the energy concentration within the pyramids is increased by approximately an order of magnitude, significantly enhancing the UV energy that passes through the fully metal-coated apex. The contrast curve model of the photoresist was used to calculate the patterned photoresist profiles for various energies. Experimental results, theoretical analysis and potential improvements of the method are presented.
© 2014 Optical Society of America
In recent decades, the application of near-field photolithography to nanofabrication has attracted increasing attention. Common near-field photolithography is based on scanning probe microscopy technology (SPM) using a tapered optical fiber probe with a sub-micrometer aperture on its apex. Ultraviolet (UV) light is guided within the core of the optical fiber and passed through the aperture to perform photolithography. When used with a nano-positioning system, so-called near-field scanning optical lithography (NSOL) yields arbitrary patterns. The NSOL provides sub-wavelength patterning without limitation of diffraction. However, the major drawback of NSOL is the low throughput to large-area patterning, and the fact that only one or a few patterns can be produced at once, even using arrayed fiber probes.
To improve the throughput of NSOL, Huo et al.  utilized a high-modulus polydimethylsiloxane (h-PDMS) mold on whose surface was an array of pyramid-shaped microstructures in place of the tapered fiber probes. The h-PDMS mold and the arrayed micro-pyramids on the surface were firstly negatively replicated from a master silicon mold using standard molding of h-PMDS materials. The h-PDMS mold was then coated with a thin metal layer to block the transmission of light. Several methods [1–3] have been developed to remove the metal film from the apexes of the pyramidal tips to form apertures through which light can pass. When a substrate on which photoresist (PR) has been deposited is placed either in contact or in close proximity with the pyramidal tips, UV light that is incident from the back of the h-PDMS mold can emerge from these aperture openings and expose the PR layer. Mechanically scanning the h-PDMS mold across over the PR-coated substrate enables multiple beams to produce UV patterns with arbitrary geometry in the PR layer simultaneously by in a direct-write and mask-less manner. This process is called beam pen lithography.
The advantage of h-PDMS-based beam pen lithography is its ease of implementation for large-area patterning with a high throughput. The patterning resolution of this process can be very high, and the line-width can be reduced to sub-micrometer or even nanometer scale [1–3] if the aperture opening is sufficiently small. However, one critical issue associated with h-PDMS-based beam pen lithography is the formation of the arrayed aperture openings by removing the metal film that originally covered the h-PDMS pyramidal tips. This process is quite difficult: forming the desired small aperture and maintaining the uniformity of size of the openings over the entire h-PDMS mold is particularly challenging.
With regard to recent developments in near-field scanning optical microscopy (NSOM), Vaccaro et al. [4–6] were the first to realize transmitting apertureless NSOM using a fully metal-coated SiO2 conical probe; they found that various metallic coatings and tapering angles of the SiO2 conical probe increased the transmittance of the probe and improved the spatial resolution of the apertureless NSOM. Kubicova et al. [7,8] used a completely aluminum-coated fiber probe to produce an apertureless NSOL; they patterned a 2D irregular structure the top surface of light emitting diodes (LED), using apertureless NSOL to increase their light extraction efficiency, and experimentally characterized the near-field electromagnetic field of this apertureless fiber probe to obtain evidence of the exposure phenomenon. This apertureless NSOL produced a smaller line-width as compared with common NSOL but was still limited in throughput when used for large-area patterning.
In this work, the advantages of h-PDMS-based beam pen lithography and apertureless NSOL probes are exploited to realize so-called apertureless beam pen lithography. A large-area micro-pyramids array, fully coated with a chromium layer without an aperture opening, is utilized to perform apertureless beam pen lithography. The following sections present the experimental and theoretical feasibility of this new approach as well as to illustrate its underlying physical mechanism.
This section describes the details of an experiment on apertureless beam pen lithography. Figure 1 presents the complete flow diagram of the fabrication of a fully metal-coated polyurethane acrylate (PUA) mold. The process begins with the use of the conventional method of silicon bulk machining [9–11] for producing a silicon master mold with arrayed surface micro-cavities with an inverted pyramidal shape. First, a pre-cleaned <100> wafer is coated with a 100 nm-thick silicon nitride (Si3N4) film using low-pressure chemical vapor deposition (LPCVD). The silicon nitride film is then patterned with an array of holes using standard photolithography. The holes, each with a diameter of 2 μm, form a rectangular array with a center-to-center pitch of 10 μm. The overall patterned area of the silicon mold is 18x18 mm2. The nitride-patterned silicon wafer is then immersed in 45% potassium hydroxide (KOH) solution at 70 °C for 300 s to form an array of pyramidal micro-cavities. After the anisotropic wet etching of the silicon, the nitride film is removed by hydrofluoric (HF) acid etching . Figure 2 displays the SEM images of the KOH-etched pyramidal micro-cavities. The pyramidal cavities have a base length of 2 μm and a depth of 1.4 μm. To prevent the cured polymer mold from sticking to the silicon master mold during the molding process, the surface of the silicon master mold is modified from hydrophilic to hydrophobic by the vapor deposition of tridecafluoro-1,1,2,2-tetrahydrooctyl-1-trichlorosilane .
In this work, an UV-curable PUA (MINS-301RM, Minuta Technology Co., Ltd., Korea) is used to replicate negatively the array of pyramidal microstructures from the silicon master mold by UV-curing and molding processes. Cured PUA is transparent to UV and is used in many UV imprinting applications [14–16]. To produce sharp pyramidal tips with a small radius of curvature, a dual vacuum chamber mechanism was used herein to prevent the trapping of air in the molding process. Figure 3 presents the UV-curing molding process performed in a vacuum. A few drops of liquid PUA were poured onto a 2 inch-diameter amorphous polyethylene terephthalate (PET) substrate (ES303010, Goodfellow Cambridge Ltd., U.K.) and then spin-coated at 3000 rpm for 30 s. The backside of the PUA-coated PET substrate was then attached to an 8 inch-diameter flexible PET membrane using an adhesive layer. In Fig. 3(a), this membrane is tensioned and fixed in a fixture to keep the PUA-coated PET substrate aligned parallel to the silicon master mold at a fixed distance from it. The flexible PET membrane also divides the inner space of the fixture into an upper and a lower chamber. The critical step is to maintain the position of the membrane by controlling the pressure difference between the upper and lower chambers while vacuuming both chambers simultaneously. After the vacuuming process is stabilized, the upper chamber is vented to 1 atm and the lower chamber is still vacuumed so the pressure difference causes the flexible membrane to move downward, allowing the uncured PUA layer to fill the pyramidal cavities without trapping air. UV light with an energy density of 30 mW/cm2 and a wavelength of 405 nm is then incident from the backside of the PET substrate to cure PUA for 15 mins under a uniform pressure loading on the backside of the PET substrate, as presented in Fig. 3(c). After the PUA layer has been completely cured, both the upper and the lower chambers are vented to 1 atm, and the PET substrate with PUA micro-structures is then peeled off from the silicon master mold. Figure 4 displays an SEM image of the replica PUA mold. The PUA pyramids have a base length of 2 μm and a height of 1.4 μm. The radius of curvature of each pyramidal tip is estimated to be less than 100 nm. Finally, the replica PUA mold is fully coated with a 60 nm-thick chromium film using an electron-beam thermal evaporator (VT1-10CE, ULVAC, Japan). The average thickness of the chromium film is 66 nm, as determined by atomic force microscopy (SPA400, SII Technology, Japan).
To provide experimental evidence of apertureless beam pen lithography, the variation of the patterned photoresist structure with exposure dosage is studied. First, a 1.8 μm-thick PR (AZ1500 20 cPs, AZ Electronic Materials Taiwan Co., Ltd., Taiwan) layer is spin-coated on a silicon wafer at 3000 rpm for 40 s. After soft-baking at 100 °C for 90 s, the chromium-coated PUA mold is brought into contact with the PR layer using the dual vacuum chamber mechanism but the silicon master mold is replaced with the PR/silicon specimen. A collimated UV source (ELS-201SA, ELS System Technology Co., Ltd., Taiwan) with an energy intensity of 18 mW/cm2 and a wavelength of 365 nm is incident from the backside of the PUA mold. The UV energy that emerges from the pyramidal tips will expose the PR layer. After the photoresist develops (AZ300MIF, AZ Electronic Materials Taiwan Co., Ltd., Taiwan), a UV-patterned PR layer with surface micro-structures of a particular kind is formed.
Figure 5 shows the topographic cross-section profiles of each patterned structure, determined by atomic force microscopy, with exposure times of 3 s, 4 s and 5 s. Gaussian curve fitting is used to estimate the depth and full-width-at-half-magnitude (FWHM) of the measured profile. Figure 6 plots the variation of the cross-section profiles with exposure time. When the exposure time is 3 s, the depth of the patterned structure is 117 nm and the FWHM of the measured surface profile is 180 nm. When the exposure time is increased to 4 s and 5 s, the depths of the patterned structures are 140 nm and 154 nm, respectively and the FWHM are 231 nm and 269 nm, respectively.
To determine whether the physical contact between the PUA micro-pyramid and the PR layer causes any imprinting on the PR layer, the experiment is repeated but without UV exposure. Figure 7 shows the SEM images of (a) an indented pattern without UV exposure and PR developing, and (b) an etched profile after UV exposure and developing process. From Fig. 7(a), there was no significant pattern revealed by SEM inspection, verifying that all of the obtained PR surface structures arose from UV photolithography and not from physical imprinting. Furthermore, Fig. 7(a) also implies that no chromium was transferred to the PR layer during contact-mode lithography owing to the good adhesion between chromium and PUA. These experimental results indicate that a particular UV energy can pass through the apex of the fully metal-coated PUA micro-pyramids, even when the metallic coating is thicker than the penetration depth of UV light. Both the depth and the FWHM of the patterned PR structure are proportional to the exposure time.
Based on the above experimental results, the thickness of the PR layer and the exposure parameters are adjusted to pattern arrayed metal dots on the silicon substrate through the lift-off process. AZ1500, with a viscosity of 20 cPs, is pre-diluted with propylene glycol monomethyl ether acetate (PGMEA, AZ Electronic Materials Taiwan Co., Ltd., Taiwan) 1:5 w/w and then spin-coated on a silicon wafer at 3000 rpm for 40 s to produce a ~100 nm-thick PR layer. Exposure by the beam pens for 5 s and a subsequent metal lift-off process yield a pattern of arrayed metal dots, each with a diameter of 300 nm, as displayed in Fig. 8.
3. Numerical simulations
In this section, the finite-element method (FEM) is utilized to simulate the experimentally observed phenomena and thereby elucidate the mechanism of UV photolithographic patterning using the PUA micro-pyramids that are fully coated with a metal layer without apertures. The purpose is to determine the electromagnetic wave intensity distribution around the tip of a fully chromium-coated PUA micro-pyramid and within the photoresist layer to provide evidence of the effect of exposure by UV that was observed in the preceding experiments.
First, a 3D model that comprises a PUA pyramid, a 60 nm-thick chromium coating layer, and a 1 μm-thick AZ1500 photoresist layer, deposited on a silicon substrate, is generated in commercial FEM software (COMSOL Multiphysics®, COMSOL, Inc., USA). To reduce the computer memory required, the original model is reduced to a one-quarter model using two planes of symmetry, as presented in Fig. 9(a). The dimensions of the PUA pyramid are a base width of 2 μm and a height of 1.41 μm. Table 1 presents the optical refractive index (n) and extinction coefficient (k) of each material at a wavelength of 365 nm [17–20]. In the simulation, the scattering boundary condition is applied to both the top and the bottom surfaces of the model while perfect electric conductor (PEC) and perfect magnetic conductor (PMC) boundary conditions are applied at the symmetric surfaces perpendicular to the incident E-field and H-field, respectively.
To simulate the distribution of the UV energy, a plane harmonic wave of UV light with a wavelength of 365 nm and an average power density (P0) of 18 mW/cm2 is incident on the top surface of the model. The finite element method is used to calculate the electromagnetic wave field within the model and the average power density distribution (P) of the TE-plane and the TM-plane is considered to be that of randomly polarized UV light, as presented in Fig. 9(b). As can be seen in Fig. 9(b), most of the UV energy is blocked by the chromium layer, whose thickness is six times the penetration depth (d). The penetration depth of chromium is approximately 9 nm at a wavelength of 365 nm, as calculated from ,
However, around the tip of the pyramid, significant amount of UV energy can pass through the chromium coating and reach the AZ1500 photoresist layer, where it reacts with the photoactive compound (PAC).
A patterned photoresist profile for a particular energy distribution can be calculated theoretically from the contrast curve model of a photoresist [22–24]. Figure 9(c) presents the calculated patterned photoresist profile. The intensity distribution at each depth (z) within the photoresist layer, E(r, z), can be extracted by multiplying the power density distribution that is presented in Fig. 9(b) by the exposure time. The exposure dosage at each depth, E(z), is calculated according to Eq. (2).Fig. 9(c). Figure 9(d) plots the variation of the calculated profile with the exposure time. The calculated profiles are qualitatively similar to those obtained from the experimental results that are shown in Fig. 6, and both the depth and the FWHM of the profile of a patterned structure increase with exposure time. The calculated profiles do not exactly match the experimental results perhaps because 1) the geometric deviation between experiment and modeling, 2) the divergence and broadband spectrum of the UV source that was used in experiment, and 3) the details of the properties of the materials, and particularly those of the chromium thin film that was deposited using an electron-beam evaporator. However, this analysis still theoretically demonstrates the feasibility of using the fully chromium-coated PUA mold in apertureless beam pen lithography and the mechanism on which it depends.
Figure 10 compares the normalized power density distributions of variously surface-structured PUAs with complete chromium coatings. In Fig. 10(a), a plane harmonic wave of UV light is incident from the top surface of the PUA layer; passes through the flat bottom surface that is coated with a 60 nm-thick chromium layer, and finally reaches the photoresist layer. The transmitted power is less than 1% because the thickness of the chromium coating is six times the penetration depth. In Fig. 10(b), arrayed pillars with a diameter of 1 μm and a height of 1 μm are modeled. A comparison of Fig. 10(b) with Fig. 10(a) reveals an enhancement of the power density by a factor of 5.1 in the PUA pillar structure by the diffraction and reflection from the structured surface. Figure 10(c) shows the result for a pyramidal structure with a base length of 2 μm, a height of 1 μm, and, therefore, a base angle of 45°. The increase of the power density by the PUA pyramidal structure is approximately 7.5. Figure 10(d) shows the result for a pyramidal structure with a base length of 2 μm and a base angle of 54.74°: the power density increased by a factor of 9.3. These results prove that the energy concentration is increased within the pyramidal structures by approximately an order of magnitude, so the UV energy that passes through the fully metal-coated apex is considerably enhanced, and so-called transmitting apertureless near-field lithography is thereby realized.
Figure 11 compares the normalized power density distributions of variously thickness of chromium coating on PUA micro-pyramids. In Fig. 11(a), a bare PUA micro-pyramid without metallic coating is molded. An enhancement of the power density by a factor of 43.6 is revealed. The enhancement within PUA pyramid is caused by the Fresnel’s reflection of light at the PUA/air interface. Figure 11(b) shows the result of a PUA micro-pyramid with a 60 nm-thick chromium coating, and the enhancement of power density is 9.3. Figure 11(c) and (d) show the results of a PUA micro-pyramid coated with a chromium layer of 54 nm and 66 nm in thickness, respectively, and the power density is enhanced by a factor of 9.2 and 10.5, respectively. The results demonstrate that the variation of thickness of metallic coating within ± 10% does not affect the enhancement factor significantly. In practical applications, the variation of transmittance with respect to the deviation of thickness of metal layer could be compensated by adjusting the exposure time to achieve desired exposure dosage.
Figure 12 compares the enhancement of power density distributions inside a PUA micro-pyramid with various sizes. The base angle of all PUA micro-pyramids is kept at 54.74°. In Fig. 12(a), a PUA micro-pyramid with a base length of 400 nm is molded and the enhancement of power density is by a factor of 4.8. Figure 12(b) shows the result of a PUA micro-pyramid with a base length of 2 μm, and the power density enhancement factor is 9.3. Figure 11(c) shows the result of a PUA micro-pyramid with a base length of 1.8 μm: the power density increased by a factor of 9.8. Figure 11(d) shows the result of a PUA micro-pyramid with a base length of 2.2 μm: the power density increased by a factor of 10.0. These results demonstrate that the enhancement factor is reduced significantly while the size of pyramids is close to the wavelength of incident light. However, the deviation of pyramid size with respect to desired one within ± 10% does not influence the whole mechanism of apertureless beam pen lithography.
This work provides both experimental evidence and numerical analysis of apertureless beam pen lithography. An array of 60 nm-thick fully chromium-coated PUA pyramidal microstructures is illuminated by a traditional UV lamp to form an array of massive UV beam pens for realizing apertureless beam pen lithography.
Experimental results reveal that significant UV energy can pass through the apex of a fully metal-coated PUA pyramid even when the thickness of the metallic coating exceeds the optical penetration depth of UV light. The patterned profiles of PR depend on exposure dosage rather than physical imprinting. The depth of the patterned PR structures is 117 nm, 140 nm, and 154 nm when the exposure time is 3 s, 4 s, and 5 s, respectively. The respective FWHM of the patterned PR structures are 180 nm, 231 nm, and 269 nm. An arrayed metal dot with a diameter of 300 nm was produced by apertureless beam pen lithography followed by a lift-off process.
A finite-element simulation of the intensity distribution near the apex of a pyramid and within the PR layer is performed to provide evidence of the exposure phenomenon of apertureless pyramidal tips. The analysis shows that UV energy can pass through a fully metal-coated pyramid without any opening at the apex. The contrast curve model of a photoresist is used to determine the variation of the theoretical patterned PR profile with different exposure dosage, which reveals the mechanism of observed exposure phenomenon.
Topographic measurements of patterned PR structures indicate that the proposed apertureless beam pen lithography cannot precisely produce high aspect-ratio PR structures owing to the poor contrast of the UV intensity field distribution. However, the simulation indicates that the contrast of the UV intensity field distribution can be improved by adjusting the tapering angle of the pyramidal tips. Further experimental works and theoretical analysis of possible improvement of the proposed apertureless beam pen lithography are still under investigation.
The authors would like to thank the National Science Council of Taiwan for financially supporting this research under Contract No. NSC 102-2120-M-006-001-CC1.
References and links
2. X. Liao, K. A. Brown, A. L. Schmucker, G. Liu, S. He, W. Shim, and C. A. Mirkin, “Desktop nanofabrication with massively multiplexed beam pen lithography,” Nat Commun 4, 2103 (2013). [CrossRef] [PubMed]
4. L. Vaccaro, L. Aeschimann, U. Staufer, H. P. Herzig, and R. Dandliker, “Propagation of electromagnetic field in fully coated near-field optical probes,” Appl. Phys. Lett. 83(3), 584–586 (2003). [CrossRef]
5. L. Aeschimann, T. Akiyama, U. Staufer, N. F. De Rooij, L. Thiery, R. Eckert, and H. Heinzelmann, “Characterization and fabrication of fully metal-coated scanning near-field optical microscopy SiO2 tips,” J. Microsc. 209(3), 182–187 (2003). [CrossRef] [PubMed]
6. E. Descrovi, L. Vaccaro, W. Nakagawa, L. Aeschimann, U. Staufer, and H. P. Herzig, “Collection of transverse and longitudinal fields by means of apertureless nanoprobes with different metal coating characteristics,” Appl. Phys. Lett. 85(22), 5340–5342 (2004). [CrossRef]
7. I. Kubicova, D. Pudis, J. Skriniarova, J. Kovac, J. Kovac Jr, J. Jakabovic, L. Suslik, J. Novak, and A. Kuzma, “2D irregular structure in the LED surface patterned by NSOM lithography,” Appl. Surf. Sci. 269, 116–119 (2013). [CrossRef]
8. I. Kubicova, D. Pudis, L. Suslik, and J. Skriniarova, “Spatial resolution of apertureless metal-coated fiber tip for NSOM lithography determined by tip-to-tip scan,” Optik (Stuttg.) 124(14), 1971–1973 (2013). [CrossRef]
10. G. T. A. Kovacs, N. I. Maluf, and K. E. Petersen, “Bulk micromachining of silicon,” Proc. IEEE 86(8), 1536–1551 (1998). [CrossRef]
11. I. Barycka and I. Zubel, “Silicon anisotropic etching in KOH-isopropanol etchant,” Sens. Actuator A-Phys. 48(3), 229–238 (1995). [CrossRef]
12. K. R. Williams and R. S. Muller, “Etch Rates for Micromachining Processing,” J. Microelectromech. Syst. 5(4), 256–269 (1996). [CrossRef]
14. J. Park, J. H. Park, E. Kim, C. W. Ahn, H. I. Jang, J. A. Rogers, and S. Jeon, “Conformable solid-index phase masks composed of high-aspect-ratio micropillar arrays and their application to 3D nanopatterning,” Adv. Mater. 23(7), 860–864 (2011). [CrossRef] [PubMed]
15. J. Y. Kim, K. S. Park, Z. S. Kim, K. H. Baek, and L. M. Do, “Fabrication of low-cost submicron patterned polymeric replica mold with high elastic modulus over a large area,” Soft Matter 8(4), 1184–1189 (2012). [CrossRef]
16. P. J. Yoo, S. J. Choi, J. H. Kim, D. Suh, S. J. Baek, T. W. Kim, and H. H. Lee, “Unconventional patterning with a modulus-tunable mold: from imprinting to microcontact printing,” Chem. Mater. 16(24), 5000–5005 (2004). [CrossRef]
17. http://www.minuta.co.kr/products/products_mold_template.html (Accessed February 28, 2014)
18. D. W. Lynch and W. R. Hunter, “Chromium (Cr),” in Handbook of Optical Constants of Solids II, E.D. Palik, ed. (Academic, 1991).
19. R. A. Norwood and L. A. Whitney, “Rapid and accurate measurements of photoresist refractive index dispersion using the prism coupling method,” Proc. SPIE 2725, 273–280 (1996). [CrossRef]
20. D. F. Edwards, “Silicon (Si),” in Handbook of Optical Constants of Solids, E.D. Palik, ed. (Academic, 1985).
21. Z. M. Zhang, Nano/Microscale Heat Transfer, (McGraw-Hill, 2007), Chap. 8.
22. E. Lee and J. W. Hahn, “The effect of photoresist contrast on the exposure profiles obtained with evanescent fields of nanoapertures,” J. Appl. Phys. 103(8), 083550 (2008). [CrossRef]
23. E. Lee and J. W. Hahn, “Modeling of three-dimensional photoresist profiles exposed by localized fields of high-transmission nano-apertures,” Nanotechnology 19(27), 275303 (2008). [CrossRef] [PubMed]
24. Y. Kim, H. Jung, S. Kim, J. Jang, J. Y. Lee, and J. W. Hahn, “Accurate near-field lithography modeling and quantitative mapping of the near-field distribution of a plasmonic nanoaperture in a metal,” Opt. Express 19(20), 19296–19309 (2011). [CrossRef] [PubMed]