Optical activities of large-area SU8 microspirals fabricated by multibeam holographic lithography Xia Wang,1 Wensheng Gao,2 Jenny Hung,2 and Wing Yim Tam2,* 1
2
Physics Department, Institute of Photonic-Electronic Science and Technology, QingDao University of Science and Technology, China
Department of Physics and William Mong Institute of Nano Science and Technology, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China *Corresponding author: [email protected] Received 18 November 2013; revised 27 February 2014; accepted 1 March 2014; posted 5 March 2014 (Doc. ID 201321); published 7 April 2014
We report on the fabrication of large-area microspirals in SU8 photoresist using a 6 1 beam holographic lithography (HL) technique involving the interference of six linearly polarized side beams and one circularly polarized central beam. In contrast to common photoresist-substrate (glass) configuration, the spirals are fabricated on a substrate with a precured thin SU8 photoresist. This SU8-SU8-glass configuration strengthens the attachment of the spirals to the substrate, and hence enhances the quality of the fabricated spirals. The fabricated SU8 microspirals exhibit large optical activities with a polarization rotation close to 10 deg and a circular dichroism of about 0.5 in the visible range. Our precured substrate method could lift the limitations of the HL method in fabricating large and uniform microstructures or nanostructures. © 2014 Optical Society of America OCIS codes: (050.5298) Photonic crystals; (160.5298) Photonic crystals; (120.3180) Interferometry. http://dx.doi.org/10.1364/AO.53.002425
1. Introduction
Holographic lithography (HL) is a simple and effective technique in fabricating microstructures and nanostructures. HL involves the recording, using a photoresist, of an interference pattern obtained by the interference of multiple coherent beams and the removal of unexposed parts to obtain 2D or 3D microstructures. It has been demonstrated that periodic microstructures could be fabricated using the HL method [1,2]. In addition to the periodic structures, we have shown that quasi-periodic structures like the 2D Penrose structure and the 3D icosahedral quasi-crystals can also be fabricated using five and seven interfering beams, respectively 1559-128X/14/112425-06$15.00/0 © 2014 Optical Society of America
[3,4]. Furthermore, we have also shown that it is possible to fabricate chiral structures using circularly polarized beams, e.g., microspirals using a 6 1 beam HL configuration [5]. Recently instead of using separate beams for the HL method, complex photonic chiral structures are obtained by phase manipulation of a single beam using a spatial light modulator [6]. Chiral structures, having a sense of structural twist and no mirror symmetry, are fascinating. They exhibit very unique optical activities like the polarization rotation of linearly polarized light as observed in quartz and solutions of organic substances [7]. In addition to the polarization rotation, chiral materials could also exhibit polarization bandgaps in which the transmittance of left- and right-handed circularly polarized light is different, a phenomenon known as circular dichroism (CD) [8]. However, optical activities from natural materials 10 April 2014 / Vol. 53, No. 11 / APPLIED OPTICS
2425
are usually very weak, and thus hinder potential applications. Due to advances in micro- and nanotechnologies, artificial chiral microstructures are now readily fabricated. For example, the glancing angle deposition (GLAD) [9,10] and the two-photon direct laser writing (DLW) [11–13] techniques have been used in the fabrication of microspirals. However, these two methods have their own limitations. For example, it is not easy to produce large and uniform samples using the GLAD while it is very time consuming using the DLW because of the serial nature of DLW. Note that a more direct layer-by-layer nanolithorgraphy method has been used to fabricate chiral (nonspiral) structures exhibiting optical activities [14] and also achieving negative refractive index [15–17], which has amplified the interest in chiral metamaterials. The HL method is a good alternative because it is simple and effective as we have demonstrated that microspirals can be fabricated using circularly polarized beams [5]. In that work, we were able to fabricate only small areas of good spirals in SU8 photoresist because we had used seven individual beams with no control of the beam phases such that we could not achieve large lateral areas of good interference patterns [5]. Recently we have overcome this limitation by using an expanded beam approach such that all seven beams are extracted directly from a single large beam so that the interfering beams were more uniform and, more importantly, the phases and intensities of the beams were better controlled. We also used a high resolution dichromate gelatin (DCG) holographic emulsion to record the interference pattern [18]. Samples with areas ∼1 cm2 were obtained exhibiting optical rotation and CD in the visible range [18]. However, we were unable to obtain the 3D structural information of the fabricated spirals using SEM imaging because of the low refractive index contrast of the DCG (dense region/less-dense region ∼1.5∕1.4 1.07) [19]. The low-index contrast also limits the performance of the optical activities of the DCG spirals. Here we repeat the experiment using SU8 photoresist with a modification such that the spirals are fabricated on a precured SU8-glass substrate. The new SU8-SU8-glass approach enables the SU8 spirals to have a good attachment to the precured SU8 layer and hence enhances the quality of the fabricated spirals. This time we can image (using SEM imaging) the SU8 spirals clearly and, more importantly, the SU8 spirals exhibit larger optical rotation and CD because of the larger refractive index contrast of the SU8 (SU8/air ∼ 1.6) as compared to that of the DCG. Our precured buffer layer approach could be useful in the production of large and uniform samples for other similar methods in the fabrication of microstructures or nanostructures. A similar HL approach using a multiple (six) twobeam exposures has also been used in fabricating large area helical photonic crystals, demonstrating the flexibility of the HL method [20]. 2426
APPLIED OPTICS / Vol. 53, No. 11 / 10 April 2014
Fig. 1. Schematic setup for the 6 1 beam holographic interference. The inset shows the 6 1 beam configuration. The laser is argon-ion at 488 nm. The incidence angle ϕ is 42°.
2. Experiment A. Holographic Setup
Figure 1 shows the 6 1 beam holographic interference setup for our experiment. The inset shows the beam arrangement for generating spiral interference pattern [5]. The expanded beam approach involved the extraction of the seven (6 1) beams out of a collimated beam (∼5 cm in diameter), as shown in Fig. 1, by passing the beam through a mask (holder) with seven holes (size 7.5 mm with six holes equally distributed on a circle and one central hole). Since all seven beams were originated from a single beam, the phases of the beams were better controlled as the optical paths are all fixed and the side beams are symmetric with respective to the central beam such that opposite side beams can compensate each other on the effects due to phase difference as compared to that of the individual beam approach used in the first study [5]. Furthermore, the intensities and polarizations of the beams could be adjusted individually using half-wave plates and polarizers mounted on the top of the holes. The seven beams formed an umbrella configuration with the side beams incident at φ 42° with respect to the central beam by internal reflection at slanted surfaces of a truncated hexagonal pyramid. The interference pattern was recorded by a photoresist placed on the top of the pyramid. A cylindrical window placed on the top of the holographic plate provided the escape paths for the interfering beams. The 6 1 beam interference pattern can be calculated easily for different beam polarizations [5,18]. We used a left-handed circularly polarized (LCP) central beam and polarizations normal to the planes of incidence, i.e., 90°, for the side beams such that their polarizations do not change after reflected from the slanted surfaces of the hexagonal pyramid to produce LH spirals for this work. Details can be referred to [5] and [18].
B.
Fabrication Procedures
An argon ion laser (λ 488 nm) was used in our holographic interference experiment. The power was ∼1.2 mW and ∼2.5 mW for the side beams and the central beam, respectively. We used a highcontrast photoresist EPON SU8 (from Shell) as the raw polymer resin. SU8 resin was dissolved in γbutyrolactone (1: 0.9 wt.) with 2.1% wt. of photoinitiator Irgacure 261 (from Ciba Co.) to form a SU8 photoresist solution. The solution was first spin coated on a glass substrate to form a ∼1 μm thick SU8 buffer layer. It was cured by exposure to the argon laser to form a SU8-glass substrate. Then a second spin coating of ∼3–10 μm SU8 was applied to the SU8-glass substrate. (Note that the SU8 will shrink after curing and development.) The precured SU8 buffer layer strengthened the attachment of the second SU8 layer, and hence the fabricated spirals, to the substrate and thus prevented the deformation of the spirals during the development process, ensuring the quality of the samples. To fabricate the spirals, the photoresist was exposed to the interference pattern of the 6 1 beams as shown in Fig. 1 for 15–20 s, corresponding roughly to a total exposure dose of 240–320 mJ∕cm2 using seven 0.89 cm2 beams with mean power 2 mW. Matching fluid was applied to interfaces to reduce multireflections inside the photoresist during exposure. After exposure, the sample was baked at 90° C for about 45 min to complete the polymerization. Finally, the sample was developed in propyleneglycolmethyl-ether-acetate (developer) for more than 5 h to remove the underexposed regions and then cleaned with isopropanol for 5–10 min to create a SU8 copy of the 3D spiral interference pattern. 3. Results A.
SU8 Spirals
Figure 2 shows a composite transmission image taken with a 5× objective (together with a 10× eye-piece) of a SU8 spiral sample under crossed polarizers. The sample is ∼0.5 cm in diameter. The nonpatterned part (blue) corresponds to the
Fig. 2. Transmission image of SU8 spirals under crossed polarizers using a 5× objective. Holes (indicated by arrows) near locations p1–p4 are “drilled” by focused ion beam milling. The sample is broken along the white-dashed line for cross-section SEM imaging. The diameter (left to right) of the sample is ∼0.5 cm.
Fig. 3. (a) Normal and (b) 40° tilted SEM images at p2 in Fig. 2. (c)–(f) Cross sections at locations A–D in Fig. 2, respectively. Scales are 10 μm for black and 1 μm for white bars. Circles in (a) are regions where transmissions are taken with 100× objective.
extinction of the polarizers while the patterned part (SU8 spirals) shows uniformly nonzero transmission across the whole sample. This indicates that the polarization of incident light after passing through the spirals is altered, exhibiting some kind of optical activities. The sample was “drilled” out at several locations, indicated by the arrows, by focused ion beam milling to expose the 3D structure and also broken along the dashed line for cross-section imaging. Note that we can consistently obtain large areas of microspirals using this expanded beam method whereas we could only find small regions of microspirals in [5]. Figure 3 shows SEM images of the sample in Fig. 2 at various locations. Figures 3(a) and 3(b) show the normal and tilted SEM views at locations near p2 in Fig. 2, respectively. These two images show, in addition to the microspiral structure, secondary hexagonal structures of sizes ∼7–8 μm resulted from the slightly off-symmetry interfering side beams (i.e., they are not exactly 60° apart after internally reflected from the slanted surfaces of the truncated hexagonal pyramid) are also observed. The secondary structures, verified in simulations using slightly off-symmetry interfering side beams, are visualized as regions with high volume fraction [circled as (i) in Fig. 3(a)] and low volume fraction [circled as (ii) in Fig. 3(a)] microspirals, shown clearly in Fig. 3(b). In Fig. 3(b), due to damages from the focused ion beam milling, the spirals are not as obvious. The SU8 spirals, left-handed, are better visualized in the cross-section SEM images as shown Figs. 3(c)–3(f) taken at locations A–D along the dashed line in Fig. 2. Note that the ∼1 μm thick SU8 buffer layer beneath the spirals is also clearly shown. The spirals all have pitch ∼0.7 μm and lattice spacing ∼0.55 μm. The actual spiral size differs from the calculated value, 1.2 μm pitch and 0.53 μm lattice spacing, because of the shrinkage of the SU8 along the normal direction and slight expansion in the lateral direction after the curing and developing processes [5]. Before exposure, for a ∼5 μm thick 10 April 2014 / Vol. 53, No. 11 / APPLIED OPTICS
2427
SU8 film, there will be around four pitches of spirals, which is consistent with the observed numbers (more than three pitches) in Fig. 3. Thicker, ∼10 μm, samples show similar results. In general, samples with the SU8 buffer layer are more uniform and well attached to the substrate than those without the buffer layer, enabling more quantitative optical measurements. B.
Optical Characterization
We used a microscopic setup coupled to a spectrometer as reported earlier to perform the optical measurements [18]. Since the spiral samples are large and quite uniform, we could use a 5× objective covering an area of 150 μm × 150 μm to obtain more accurate results. Note that the detected area is large enough to cover many units of the secondary structures as mentioned above. To demonstrate the spatial dependence, we also performed measurements using a 100× objective to zoom into the high and low volume fraction regions for comparison. 1. Optical Reflection and Transmission Figure 4(a) shows the transmission under crossed polarizers of the SU8 spirals at various locations in Fig. 2. At the nonpatterned part (no spirals) the transmission is practically zero for wavelengths
Fig. 4. (a) Transmittance of SU8 spirals under crossed polarizers. (b) Reflectance and (c) transmittance of linearly polarizer incident light. 2428
APPLIED OPTICS / Vol. 53, No. 11 / 10 April 2014
Fig. 5. Stoke’s parameters of SU8 spirals: (a), (b) polarization rotation and (c), (d) ellipticity. (a) and (c) are forward incidence while (b) and (d) are backward incidence.
longer than 550 nm while there is nonperfect extinction at shorter wavelengths, which sets the limitation of the polarizers. As for the spirals, due to the optical activities of the spirals, there is ∼2% transmission over a wide range from ∼570 to ∼720 nm at various locations as indicated. The corresponding reflectance and transmittance of linearly polarized incident light are shown in Figs. 4(b) and 4(c), respectively. The results show that there is no obvious photonic bandgap in the measured range except a small peak at ∼690 nm for the reflectance. The performance is qualitatively the same at various locations, indicating the quality of the spirals fabricated. 2. Polarization Rotation To characterize the chirality of the spiral samples, we first measured the Stoke’s parameter using a setup as reported earlier [18]. Figures 5(a) and 5(b) show the polarization rotation at various locations in Fig. 2 for forward [substrate glass facing incident light, Fig. 5(a)] and backward [spiral facing incident light, Fig. 5(b)], respectively. While it is zero for the nonpatterned part as expected, polarization rotation is observed over a wide range from ∼550 to ∼750 nm and reaches almost 10° at 650 nm. More importantly, the rotation is independent of the incidence direction, forward or backward, in good accord with the expectation from the symmetry of the spiral. The ellipticity of the transmitted light shows corresponding trend as shown in Figs. 5(c) and 5(d). The nonzero ellipticity implies that circularly polarized incident light will be converted into elliptically polarized light after passing through the spiral sample due to the difference in the absorptions and/or conversions for left-handed (LCP) and right-handed (RCP) incident light. The optical activities of the SU8 spirals are much larger than those of the DCG spirals reported earlier [18], in accord with
the expectation from larger refractive index contrast of the SU8 over DCG. C.
Circular Dichroism
We then further characterized the spirals with CD defined as Δ 2 × T R − T L ∕T R T L , where T L and T R correspond to the transmittance of LCP and RCP incident light, respectively. We used the same setup as reported earlier to perform the CD measurements [18]. The left and middle columns of Fig. 6 show the transmittance of circularly polarized light of the SU8 spirals for forward and backward incidence, respectively. It is clear that the transmittances are different for LCP and RCP for the spirals in all locations except for the nonpatterned part where they are the same. Moreover, the transmittances are all semi-quantitatively the same at different locations and also incident (forward and backward) directions. Note that the transmittance
difference can be as high as ∼0.1 at some wavelengths. The corresponding CDs are shown in the right column of Fig. 6. The CD for the nonpatterned part is zero while it is nearly the same for all locations with positive values for wavelengths < ∼ 650 nm and negative values for longer wavelengths. Moreover, it reaches ∼j0.5j at ∼610 and ∼680 nm. Note also that the forward and backward CDs are the same within experimental errors supporting strongly the 3D nature of the spirals and also the good quality of the sample. To further study the spatial dependence of the SU8 spirals, Fig. 7 shows the results taken using a 100× objective covering an area of ∼7.5 μm × 7.5 μm for high [Figs. 7(a) and 7(b), at location p2-i in Fig. 2] and low volume fraction [Figs. 7(c) and 7(d), at location p2-ii in Fig. 2] regions for forward incidence. Despite the noise levels being much higher than those taken using the 5× objective shown in Figs. 4–6, the behavior is consistent with
Fig. 6. Forward transmittance (left column), backward transmittance (middle column), and CD (right column) of circularly polarized incident light of SU8 spirals in Fig. 2. 10 April 2014 / Vol. 53, No. 11 / APPLIED OPTICS
2429
References
Fig. 7. Forward transmittance and CD of circularly polarized incident light of SU8 spirals near p2 in Fig. 2 [circled in Fig. 3(a)] using 100× objective.
the 5× objective results. However, the results seem to show a slightly larger transmittance difference and CD for the low volume fraction spirals. Unfortunately our samples are still not good enough to determine a more detailed dependency. Overall, the results of the SU8 spirals are robust and provide a clear verification for the optical activities of the spirals. 4. Conclusion
We have fabricated large-area and good quality SU8 microspirals using a 6 1 beam holographic interference. The spirals exhibit large optical activities— polarization rotation and circular dichroism—in the visible range. The good quality of the spirals is resulted from a better control of the quality of the interfering beams by extracting the beams from a single expanded beam and also by using a precured SU8 buffer layer for better attachment of the fabricated spirals to the substrate. Our precured buffer layer approach could be applied to other techniques for fabrication of microstructures or nanostructures. Support from Hong Kong RGC grants RPC10SC06 and HKUST2/CRF/11G is gratefully acknowledged. The technical support of the Raith-HKUST Nanotechnology Laboratory for the focused ion beam milling and SEM image facility at MCPF (SEG_ HKUST08) is hereby acknowledged. X. W. acknowledges support from the National NSFC (11274189) and the Excellent Youth Foundation of Shandong Scientific Committee (JQ201018).
2430
APPLIED OPTICS / Vol. 53, No. 11 / 10 April 2014
1. M. Campbell, D. N. Sharp, M. T. Harrison, R. G. Denning, and A. J. Turberfield, “Fabrication of photonic crystals for the visible spectrum by holographic lithography,” Nature 404, 53–56 (2000). 2. X. Wang, J. F. Xu, H. M. Su, Z. H. Zeng, Y. L. Chen, H. Z. Wang, Y. K. Pang, and W. Y. Tam, “Three-dimensional photonic crystals fabricated by visible light holographic lithography,” Appl. Phys. Lett. 82, 2212–2214 (2003). 3. X. Wang, C. Y. Ng, W. Y. Tam, C. T. Chan, and P. Sheng, “Large area two-dimensional mesoscale quasicrystals,” Adv. Mater. 15, 1526–1528 (2003). 4. J. Xu, R. Ma, X. Wang, and W. Y. Tam, “Icosahedral quasicrystals for visible wavelength by optical interference holography,” Opt. Express 15, 4287–4295 (2007). 5. Y. K. Pang, J. C. W. Lee, H. F. Lee, W. Y. Tam, C. T. Chan, and P. Sheng, “Chiral microstructures (spirals) fabrication by holographic lithography,” Opt. Express 13, 7615–7620 (2005). 6. J. Xavier and J. Joseph, “Tunable complex photonic chiral lattices by reconfigurable optical phase engineering,” Opt. Lett. 36, 403–405 (2011). 7. A. Lakhtakia, ed., Selected Papers on Natural Optical Activity (SPIE, 1990). 8. J. C. W. Lee and C. T. Chan, “Polarization gaps in spiral photonic crystals,” Opt. Express 13, 8083–8088 (2005). 9. K. Robble and M. J. Brett, “Chiral sculptured thin films,” Nature 384, 616 (1996). 10. M. O. Jensen and M. J. Brett, “Square spiral 3D photonic bandgap crystals at telecommunications frequencies,” Opt. Express 13, 3348–3354 (2005). 11. S. Maruo, O. Nakamura, and S. Kawata, “Three-dimensional microfabrication with two-photon-absorbed photopolymerization,” Opt. Lett. 22, 132–134 (1997). 12. M. Deuble, G. von Freymann, M. Wegener, S. Pereira, K. Busch, and C. M. Soukoulis, “Direct laser writing of threedimensional photonic-crystal templates for telecommunications,” Nat. Mater. 3, 444–447 (2004). 13. K. K. Seet, V. Mizeikis, S. Matsuo, S. Juodkazis, and H. Misawa, “Three-dimensional spiral-architecture photonic crystals obtained by direct laser writing,” Adv. Mater. 17, 541–545 (2005). 14. See, e.g., A. V. Rogacheva, V. A. Fedotov, A. S. Schwanecke, and N. I. Zheludev, “Giant gyrotropy due to electromagnetic-field coupling in a bilayered chiral structure,” Phys. Rev. Lett. 97, 177401 (2006). 15. E. Plum, J. Zhou, J. Dong, V. A. Fedotov, T. Koschny, C. M. Soukoulis, and N. I. Zheludev, “Metamaterial with negative index due to chirality,” Phys. Rev. B 79, 035407 (2009). 16. J. B. Pendry, “A chiral route to negative refraction,” Science 306, 1353–1355 (2004). 17. C. Monzon and D. W. Forester, “Negative refraction and focusing of circularly polarized waves in optically active media,” Phys. Rev. Lett. 95, 123904 (2005). 18. J. Hung, W. Gao, and W. Y. Tam, “Optical activities of microspiral photonic crystals fabricated by multi-beam holographic lithography,” J. Opt. 13, 119501 (2011). 19. R. Ma, J. Xu, and W. Y. Tam, “Wide bandgap photonic structures in dichromate gelatin emulsions,” Appl. Phys. Lett. 89, 081116 (2006). 20. K. Raub and S. R. J. Brueck, “Large area 3D helical photonic crystals,” J. Vac. Sci. Technol. B29, 06FF02 (2011).
2
Physics Department, Institute of Photonic-Electronic Science and Technology, QingDao University of Science and Technology, China
Department of Physics and William Mong Institute of Nano Science and Technology, Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China *Corresponding author: [email protected] Received 18 November 2013; revised 27 February 2014; accepted 1 March 2014; posted 5 March 2014 (Doc. ID 201321); published 7 April 2014
We report on the fabrication of large-area microspirals in SU8 photoresist using a 6 1 beam holographic lithography (HL) technique involving the interference of six linearly polarized side beams and one circularly polarized central beam. In contrast to common photoresist-substrate (glass) configuration, the spirals are fabricated on a substrate with a precured thin SU8 photoresist. This SU8-SU8-glass configuration strengthens the attachment of the spirals to the substrate, and hence enhances the quality of the fabricated spirals. The fabricated SU8 microspirals exhibit large optical activities with a polarization rotation close to 10 deg and a circular dichroism of about 0.5 in the visible range. Our precured substrate method could lift the limitations of the HL method in fabricating large and uniform microstructures or nanostructures. © 2014 Optical Society of America OCIS codes: (050.5298) Photonic crystals; (160.5298) Photonic crystals; (120.3180) Interferometry. http://dx.doi.org/10.1364/AO.53.002425
1. Introduction
Holographic lithography (HL) is a simple and effective technique in fabricating microstructures and nanostructures. HL involves the recording, using a photoresist, of an interference pattern obtained by the interference of multiple coherent beams and the removal of unexposed parts to obtain 2D or 3D microstructures. It has been demonstrated that periodic microstructures could be fabricated using the HL method [1,2]. In addition to the periodic structures, we have shown that quasi-periodic structures like the 2D Penrose structure and the 3D icosahedral quasi-crystals can also be fabricated using five and seven interfering beams, respectively 1559-128X/14/112425-06$15.00/0 © 2014 Optical Society of America
[3,4]. Furthermore, we have also shown that it is possible to fabricate chiral structures using circularly polarized beams, e.g., microspirals using a 6 1 beam HL configuration [5]. Recently instead of using separate beams for the HL method, complex photonic chiral structures are obtained by phase manipulation of a single beam using a spatial light modulator [6]. Chiral structures, having a sense of structural twist and no mirror symmetry, are fascinating. They exhibit very unique optical activities like the polarization rotation of linearly polarized light as observed in quartz and solutions of organic substances [7]. In addition to the polarization rotation, chiral materials could also exhibit polarization bandgaps in which the transmittance of left- and right-handed circularly polarized light is different, a phenomenon known as circular dichroism (CD) [8]. However, optical activities from natural materials 10 April 2014 / Vol. 53, No. 11 / APPLIED OPTICS
2425
are usually very weak, and thus hinder potential applications. Due to advances in micro- and nanotechnologies, artificial chiral microstructures are now readily fabricated. For example, the glancing angle deposition (GLAD) [9,10] and the two-photon direct laser writing (DLW) [11–13] techniques have been used in the fabrication of microspirals. However, these two methods have their own limitations. For example, it is not easy to produce large and uniform samples using the GLAD while it is very time consuming using the DLW because of the serial nature of DLW. Note that a more direct layer-by-layer nanolithorgraphy method has been used to fabricate chiral (nonspiral) structures exhibiting optical activities [14] and also achieving negative refractive index [15–17], which has amplified the interest in chiral metamaterials. The HL method is a good alternative because it is simple and effective as we have demonstrated that microspirals can be fabricated using circularly polarized beams [5]. In that work, we were able to fabricate only small areas of good spirals in SU8 photoresist because we had used seven individual beams with no control of the beam phases such that we could not achieve large lateral areas of good interference patterns [5]. Recently we have overcome this limitation by using an expanded beam approach such that all seven beams are extracted directly from a single large beam so that the interfering beams were more uniform and, more importantly, the phases and intensities of the beams were better controlled. We also used a high resolution dichromate gelatin (DCG) holographic emulsion to record the interference pattern [18]. Samples with areas ∼1 cm2 were obtained exhibiting optical rotation and CD in the visible range [18]. However, we were unable to obtain the 3D structural information of the fabricated spirals using SEM imaging because of the low refractive index contrast of the DCG (dense region/less-dense region ∼1.5∕1.4 1.07) [19]. The low-index contrast also limits the performance of the optical activities of the DCG spirals. Here we repeat the experiment using SU8 photoresist with a modification such that the spirals are fabricated on a precured SU8-glass substrate. The new SU8-SU8-glass approach enables the SU8 spirals to have a good attachment to the precured SU8 layer and hence enhances the quality of the fabricated spirals. This time we can image (using SEM imaging) the SU8 spirals clearly and, more importantly, the SU8 spirals exhibit larger optical rotation and CD because of the larger refractive index contrast of the SU8 (SU8/air ∼ 1.6) as compared to that of the DCG. Our precured buffer layer approach could be useful in the production of large and uniform samples for other similar methods in the fabrication of microstructures or nanostructures. A similar HL approach using a multiple (six) twobeam exposures has also been used in fabricating large area helical photonic crystals, demonstrating the flexibility of the HL method [20]. 2426
APPLIED OPTICS / Vol. 53, No. 11 / 10 April 2014
Fig. 1. Schematic setup for the 6 1 beam holographic interference. The inset shows the 6 1 beam configuration. The laser is argon-ion at 488 nm. The incidence angle ϕ is 42°.
2. Experiment A. Holographic Setup
Figure 1 shows the 6 1 beam holographic interference setup for our experiment. The inset shows the beam arrangement for generating spiral interference pattern [5]. The expanded beam approach involved the extraction of the seven (6 1) beams out of a collimated beam (∼5 cm in diameter), as shown in Fig. 1, by passing the beam through a mask (holder) with seven holes (size 7.5 mm with six holes equally distributed on a circle and one central hole). Since all seven beams were originated from a single beam, the phases of the beams were better controlled as the optical paths are all fixed and the side beams are symmetric with respective to the central beam such that opposite side beams can compensate each other on the effects due to phase difference as compared to that of the individual beam approach used in the first study [5]. Furthermore, the intensities and polarizations of the beams could be adjusted individually using half-wave plates and polarizers mounted on the top of the holes. The seven beams formed an umbrella configuration with the side beams incident at φ 42° with respect to the central beam by internal reflection at slanted surfaces of a truncated hexagonal pyramid. The interference pattern was recorded by a photoresist placed on the top of the pyramid. A cylindrical window placed on the top of the holographic plate provided the escape paths for the interfering beams. The 6 1 beam interference pattern can be calculated easily for different beam polarizations [5,18]. We used a left-handed circularly polarized (LCP) central beam and polarizations normal to the planes of incidence, i.e., 90°, for the side beams such that their polarizations do not change after reflected from the slanted surfaces of the hexagonal pyramid to produce LH spirals for this work. Details can be referred to [5] and [18].
B.
Fabrication Procedures
An argon ion laser (λ 488 nm) was used in our holographic interference experiment. The power was ∼1.2 mW and ∼2.5 mW for the side beams and the central beam, respectively. We used a highcontrast photoresist EPON SU8 (from Shell) as the raw polymer resin. SU8 resin was dissolved in γbutyrolactone (1: 0.9 wt.) with 2.1% wt. of photoinitiator Irgacure 261 (from Ciba Co.) to form a SU8 photoresist solution. The solution was first spin coated on a glass substrate to form a ∼1 μm thick SU8 buffer layer. It was cured by exposure to the argon laser to form a SU8-glass substrate. Then a second spin coating of ∼3–10 μm SU8 was applied to the SU8-glass substrate. (Note that the SU8 will shrink after curing and development.) The precured SU8 buffer layer strengthened the attachment of the second SU8 layer, and hence the fabricated spirals, to the substrate and thus prevented the deformation of the spirals during the development process, ensuring the quality of the samples. To fabricate the spirals, the photoresist was exposed to the interference pattern of the 6 1 beams as shown in Fig. 1 for 15–20 s, corresponding roughly to a total exposure dose of 240–320 mJ∕cm2 using seven 0.89 cm2 beams with mean power 2 mW. Matching fluid was applied to interfaces to reduce multireflections inside the photoresist during exposure. After exposure, the sample was baked at 90° C for about 45 min to complete the polymerization. Finally, the sample was developed in propyleneglycolmethyl-ether-acetate (developer) for more than 5 h to remove the underexposed regions and then cleaned with isopropanol for 5–10 min to create a SU8 copy of the 3D spiral interference pattern. 3. Results A.
SU8 Spirals
Figure 2 shows a composite transmission image taken with a 5× objective (together with a 10× eye-piece) of a SU8 spiral sample under crossed polarizers. The sample is ∼0.5 cm in diameter. The nonpatterned part (blue) corresponds to the
Fig. 2. Transmission image of SU8 spirals under crossed polarizers using a 5× objective. Holes (indicated by arrows) near locations p1–p4 are “drilled” by focused ion beam milling. The sample is broken along the white-dashed line for cross-section SEM imaging. The diameter (left to right) of the sample is ∼0.5 cm.
Fig. 3. (a) Normal and (b) 40° tilted SEM images at p2 in Fig. 2. (c)–(f) Cross sections at locations A–D in Fig. 2, respectively. Scales are 10 μm for black and 1 μm for white bars. Circles in (a) are regions where transmissions are taken with 100× objective.
extinction of the polarizers while the patterned part (SU8 spirals) shows uniformly nonzero transmission across the whole sample. This indicates that the polarization of incident light after passing through the spirals is altered, exhibiting some kind of optical activities. The sample was “drilled” out at several locations, indicated by the arrows, by focused ion beam milling to expose the 3D structure and also broken along the dashed line for cross-section imaging. Note that we can consistently obtain large areas of microspirals using this expanded beam method whereas we could only find small regions of microspirals in [5]. Figure 3 shows SEM images of the sample in Fig. 2 at various locations. Figures 3(a) and 3(b) show the normal and tilted SEM views at locations near p2 in Fig. 2, respectively. These two images show, in addition to the microspiral structure, secondary hexagonal structures of sizes ∼7–8 μm resulted from the slightly off-symmetry interfering side beams (i.e., they are not exactly 60° apart after internally reflected from the slanted surfaces of the truncated hexagonal pyramid) are also observed. The secondary structures, verified in simulations using slightly off-symmetry interfering side beams, are visualized as regions with high volume fraction [circled as (i) in Fig. 3(a)] and low volume fraction [circled as (ii) in Fig. 3(a)] microspirals, shown clearly in Fig. 3(b). In Fig. 3(b), due to damages from the focused ion beam milling, the spirals are not as obvious. The SU8 spirals, left-handed, are better visualized in the cross-section SEM images as shown Figs. 3(c)–3(f) taken at locations A–D along the dashed line in Fig. 2. Note that the ∼1 μm thick SU8 buffer layer beneath the spirals is also clearly shown. The spirals all have pitch ∼0.7 μm and lattice spacing ∼0.55 μm. The actual spiral size differs from the calculated value, 1.2 μm pitch and 0.53 μm lattice spacing, because of the shrinkage of the SU8 along the normal direction and slight expansion in the lateral direction after the curing and developing processes [5]. Before exposure, for a ∼5 μm thick 10 April 2014 / Vol. 53, No. 11 / APPLIED OPTICS
2427
SU8 film, there will be around four pitches of spirals, which is consistent with the observed numbers (more than three pitches) in Fig. 3. Thicker, ∼10 μm, samples show similar results. In general, samples with the SU8 buffer layer are more uniform and well attached to the substrate than those without the buffer layer, enabling more quantitative optical measurements. B.
Optical Characterization
We used a microscopic setup coupled to a spectrometer as reported earlier to perform the optical measurements [18]. Since the spiral samples are large and quite uniform, we could use a 5× objective covering an area of 150 μm × 150 μm to obtain more accurate results. Note that the detected area is large enough to cover many units of the secondary structures as mentioned above. To demonstrate the spatial dependence, we also performed measurements using a 100× objective to zoom into the high and low volume fraction regions for comparison. 1. Optical Reflection and Transmission Figure 4(a) shows the transmission under crossed polarizers of the SU8 spirals at various locations in Fig. 2. At the nonpatterned part (no spirals) the transmission is practically zero for wavelengths
Fig. 4. (a) Transmittance of SU8 spirals under crossed polarizers. (b) Reflectance and (c) transmittance of linearly polarizer incident light. 2428
APPLIED OPTICS / Vol. 53, No. 11 / 10 April 2014
Fig. 5. Stoke’s parameters of SU8 spirals: (a), (b) polarization rotation and (c), (d) ellipticity. (a) and (c) are forward incidence while (b) and (d) are backward incidence.
longer than 550 nm while there is nonperfect extinction at shorter wavelengths, which sets the limitation of the polarizers. As for the spirals, due to the optical activities of the spirals, there is ∼2% transmission over a wide range from ∼570 to ∼720 nm at various locations as indicated. The corresponding reflectance and transmittance of linearly polarized incident light are shown in Figs. 4(b) and 4(c), respectively. The results show that there is no obvious photonic bandgap in the measured range except a small peak at ∼690 nm for the reflectance. The performance is qualitatively the same at various locations, indicating the quality of the spirals fabricated. 2. Polarization Rotation To characterize the chirality of the spiral samples, we first measured the Stoke’s parameter using a setup as reported earlier [18]. Figures 5(a) and 5(b) show the polarization rotation at various locations in Fig. 2 for forward [substrate glass facing incident light, Fig. 5(a)] and backward [spiral facing incident light, Fig. 5(b)], respectively. While it is zero for the nonpatterned part as expected, polarization rotation is observed over a wide range from ∼550 to ∼750 nm and reaches almost 10° at 650 nm. More importantly, the rotation is independent of the incidence direction, forward or backward, in good accord with the expectation from the symmetry of the spiral. The ellipticity of the transmitted light shows corresponding trend as shown in Figs. 5(c) and 5(d). The nonzero ellipticity implies that circularly polarized incident light will be converted into elliptically polarized light after passing through the spiral sample due to the difference in the absorptions and/or conversions for left-handed (LCP) and right-handed (RCP) incident light. The optical activities of the SU8 spirals are much larger than those of the DCG spirals reported earlier [18], in accord with
the expectation from larger refractive index contrast of the SU8 over DCG. C.
Circular Dichroism
We then further characterized the spirals with CD defined as Δ 2 × T R − T L ∕T R T L , where T L and T R correspond to the transmittance of LCP and RCP incident light, respectively. We used the same setup as reported earlier to perform the CD measurements [18]. The left and middle columns of Fig. 6 show the transmittance of circularly polarized light of the SU8 spirals for forward and backward incidence, respectively. It is clear that the transmittances are different for LCP and RCP for the spirals in all locations except for the nonpatterned part where they are the same. Moreover, the transmittances are all semi-quantitatively the same at different locations and also incident (forward and backward) directions. Note that the transmittance
difference can be as high as ∼0.1 at some wavelengths. The corresponding CDs are shown in the right column of Fig. 6. The CD for the nonpatterned part is zero while it is nearly the same for all locations with positive values for wavelengths < ∼ 650 nm and negative values for longer wavelengths. Moreover, it reaches ∼j0.5j at ∼610 and ∼680 nm. Note also that the forward and backward CDs are the same within experimental errors supporting strongly the 3D nature of the spirals and also the good quality of the sample. To further study the spatial dependence of the SU8 spirals, Fig. 7 shows the results taken using a 100× objective covering an area of ∼7.5 μm × 7.5 μm for high [Figs. 7(a) and 7(b), at location p2-i in Fig. 2] and low volume fraction [Figs. 7(c) and 7(d), at location p2-ii in Fig. 2] regions for forward incidence. Despite the noise levels being much higher than those taken using the 5× objective shown in Figs. 4–6, the behavior is consistent with
Fig. 6. Forward transmittance (left column), backward transmittance (middle column), and CD (right column) of circularly polarized incident light of SU8 spirals in Fig. 2. 10 April 2014 / Vol. 53, No. 11 / APPLIED OPTICS
2429
References
Fig. 7. Forward transmittance and CD of circularly polarized incident light of SU8 spirals near p2 in Fig. 2 [circled in Fig. 3(a)] using 100× objective.
the 5× objective results. However, the results seem to show a slightly larger transmittance difference and CD for the low volume fraction spirals. Unfortunately our samples are still not good enough to determine a more detailed dependency. Overall, the results of the SU8 spirals are robust and provide a clear verification for the optical activities of the spirals. 4. Conclusion
We have fabricated large-area and good quality SU8 microspirals using a 6 1 beam holographic interference. The spirals exhibit large optical activities— polarization rotation and circular dichroism—in the visible range. The good quality of the spirals is resulted from a better control of the quality of the interfering beams by extracting the beams from a single expanded beam and also by using a precured SU8 buffer layer for better attachment of the fabricated spirals to the substrate. Our precured buffer layer approach could be applied to other techniques for fabrication of microstructures or nanostructures. Support from Hong Kong RGC grants RPC10SC06 and HKUST2/CRF/11G is gratefully acknowledged. The technical support of the Raith-HKUST Nanotechnology Laboratory for the focused ion beam milling and SEM image facility at MCPF (SEG_ HKUST08) is hereby acknowledged. X. W. acknowledges support from the National NSFC (11274189) and the Excellent Youth Foundation of Shandong Scientific Committee (JQ201018).
2430
APPLIED OPTICS / Vol. 53, No. 11 / 10 April 2014
1. M. Campbell, D. N. Sharp, M. T. Harrison, R. G. Denning, and A. J. Turberfield, “Fabrication of photonic crystals for the visible spectrum by holographic lithography,” Nature 404, 53–56 (2000). 2. X. Wang, J. F. Xu, H. M. Su, Z. H. Zeng, Y. L. Chen, H. Z. Wang, Y. K. Pang, and W. Y. Tam, “Three-dimensional photonic crystals fabricated by visible light holographic lithography,” Appl. Phys. Lett. 82, 2212–2214 (2003). 3. X. Wang, C. Y. Ng, W. Y. Tam, C. T. Chan, and P. Sheng, “Large area two-dimensional mesoscale quasicrystals,” Adv. Mater. 15, 1526–1528 (2003). 4. J. Xu, R. Ma, X. Wang, and W. Y. Tam, “Icosahedral quasicrystals for visible wavelength by optical interference holography,” Opt. Express 15, 4287–4295 (2007). 5. Y. K. Pang, J. C. W. Lee, H. F. Lee, W. Y. Tam, C. T. Chan, and P. Sheng, “Chiral microstructures (spirals) fabrication by holographic lithography,” Opt. Express 13, 7615–7620 (2005). 6. J. Xavier and J. Joseph, “Tunable complex photonic chiral lattices by reconfigurable optical phase engineering,” Opt. Lett. 36, 403–405 (2011). 7. A. Lakhtakia, ed., Selected Papers on Natural Optical Activity (SPIE, 1990). 8. J. C. W. Lee and C. T. Chan, “Polarization gaps in spiral photonic crystals,” Opt. Express 13, 8083–8088 (2005). 9. K. Robble and M. J. Brett, “Chiral sculptured thin films,” Nature 384, 616 (1996). 10. M. O. Jensen and M. J. Brett, “Square spiral 3D photonic bandgap crystals at telecommunications frequencies,” Opt. Express 13, 3348–3354 (2005). 11. S. Maruo, O. Nakamura, and S. Kawata, “Three-dimensional microfabrication with two-photon-absorbed photopolymerization,” Opt. Lett. 22, 132–134 (1997). 12. M. Deuble, G. von Freymann, M. Wegener, S. Pereira, K. Busch, and C. M. Soukoulis, “Direct laser writing of threedimensional photonic-crystal templates for telecommunications,” Nat. Mater. 3, 444–447 (2004). 13. K. K. Seet, V. Mizeikis, S. Matsuo, S. Juodkazis, and H. Misawa, “Three-dimensional spiral-architecture photonic crystals obtained by direct laser writing,” Adv. Mater. 17, 541–545 (2005). 14. See, e.g., A. V. Rogacheva, V. A. Fedotov, A. S. Schwanecke, and N. I. Zheludev, “Giant gyrotropy due to electromagnetic-field coupling in a bilayered chiral structure,” Phys. Rev. Lett. 97, 177401 (2006). 15. E. Plum, J. Zhou, J. Dong, V. A. Fedotov, T. Koschny, C. M. Soukoulis, and N. I. Zheludev, “Metamaterial with negative index due to chirality,” Phys. Rev. B 79, 035407 (2009). 16. J. B. Pendry, “A chiral route to negative refraction,” Science 306, 1353–1355 (2004). 17. C. Monzon and D. W. Forester, “Negative refraction and focusing of circularly polarized waves in optically active media,” Phys. Rev. Lett. 95, 123904 (2005). 18. J. Hung, W. Gao, and W. Y. Tam, “Optical activities of microspiral photonic crystals fabricated by multi-beam holographic lithography,” J. Opt. 13, 119501 (2011). 19. R. Ma, J. Xu, and W. Y. Tam, “Wide bandgap photonic structures in dichromate gelatin emulsions,” Appl. Phys. Lett. 89, 081116 (2006). 20. K. Raub and S. R. J. Brueck, “Large area 3D helical photonic crystals,” J. Vac. Sci. Technol. B29, 06FF02 (2011).
