- Email: [email protected]
Photolithography using lateral surface of nanofibers
Optics Communications 343 (2015) 195–200
Contents lists available at ScienceDirect
Optics Communications journal homepage: www.elsevier.com/locate/optcom
Photolithography using lateral surface of nanofibers Zhen Wei, Jian Bai n, Chen Wang, Neibin Hu, Jianfeng Xu, Yuan Yao, Yiyong Liang, Kaiwei Wang, Changlun Hou, Guoguang Yang State Key Laboratory of Modern Optical Instrumentation, Department of Optical Engineering, Zhejiang University, Hangzhou 310027, China
art ic l e i nf o
a b s t r a c t
Article history: Received 10 July 2014 Received in revised form 14 November 2014 Accepted 23 November 2014 Available online 7 December 2014
To enhance the resolution of photolithography, we demonstrate a technique that confines the exposure area by using lateral surface of nanofibers. Due to evanescent wave and optical tunneling effect, the interaction area of optical energy and the photoresist layer is confined into sub-wavelength dimension. Illuminated by a He–Cd laser device with a 442 nm wavelength, exposed lines with sub-wavelength width were obtained by using a nanofiber with a 247 nm diameter. Furthermore, curve lines and annular lines were obtained by manipulating the shape of nanofibers on the photoresist layer. & 2015 Elsevier B.V. All rights reserved.
Keywords: Fiber optics Nanolithography Near-field optical recording Tunneling
1. Introduction Although electron-beam lithography and atomic force microscope (AFM) etching have obtained resolution beyond 50 nm [1–3], they could not satisfy mass production of nano-components. Enhancing the resolution of photolithography to subwavelength dimension is urgent for fabricating nano-sensors, nano-detectors and so on. Recently, exposed lines with 350-nmresolution were obtained by using the tip of a nanofiber as an exposure source [4]. Unfortunately, light divergence and diffraction limit impede further improvement of resolution. The purpose of this paper is to enhance the resolution of photolithography by narrowing down the interaction area of optical energy and the photoresist layer. Evanescent wave confined in sub-wavelength scale has brought about a lot of application, such as Scanning Near Fields Optical Microscope (SNOM), optical micromanipulation and so on [5–10]. Besides, based on the frustrated total internal reflection, a kind of optical tunneling effect, Photon Scanning Tunneling Microscope (PSTM) could form images with λ/10 lateral resolution [11–16]. Silica nanofibers with diameters ranging from 50 to several hundred nanometers were fabricated successfully. With the decrease of the nanofiber diameter, the proportion of optical fields propagating as evanescent wave increases dramatically [17–19]. Actually, nanofibers have brought about a lot of applications, such as nano-lasers, nano-interferometers and so on [20–23]. n
Corresponding author. E-mail address: [email protected] (J. Bai).
http://dx.doi.org/10.1016/j.optcom.2014.11.071 0030-4018/& 2015 Elsevier B.V. All rights reserved.
In this paper, the simulations demonstrated that nanofibers confine the interaction area of optical energy and the photoresist layer in sub-wavelength dimension so that the linewidth of the light spot in the photoresist layer decreases to 200 nm. Experimentally, using a 247-nm-diameter nanofiber, exposed lines with 75, 160 and 238 nm width were achieved under the illumination of different powers. This basic research provides a novel method for photolithography.
2. Principles and simulations 2.1. The principle of photolithography using lateral surface of nanofibers The model of photolithography using lateral surface of nanofibers is illustrated in Fig. 1(a): a nanofiber's lateral surface intimately contacts the photoresist layer (0 rxr 0.1, 0.4 ryr 0.4, 0 rzr 3.1) which is coated on a silicon wafer (0.1 rx r0.35, 0.4 ry r0.4, 0 rzr 3.1). The principle of the exposure could be explained by analyzing the XY plane of the model. As shown in Fig. 1 (b), because the refractive index of the photoresist layer is larger than that of the nanofiber, optical energy could be transmitted from the nanofiber into the photoresist layer at the connected point (x¼0, y¼0). However, at the disconnected region, optical energy could not be transmitted into the photoresist layer unless the optical tunneling effect occurs. Actually, the optical tunneling effect generates a transparent aperture in the air (blue region). As shown in Fig. 1(b), inside the critical plane which is defined as y¼ a(a≠0),
196
Z. Wei et al. / Optics Communications 343 (2015) 195–200
Fig. 1. (a) The model of photolithography using lateral surface of nanofibers and (b) the XY plane of the model. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
the evanescent wave is intense enough so that optical energy could penetrate the air and expose the photoresist layer, but optical energy could not penetrate the air if y4a. Consequently, the dimension of the transparent aperture is confined to 2a and lower power brings out narrower optical channel.
2.2. The simulated results of photolithography model To investigate the transmission of optical energy, the model is simulated with the Lumerical FDTD Solutions. The simulations are set with the following parameters: the refractive indices of
Fig. 2. (a) Poynting vector in the X-direction. of an isolated 250-nm-diameter. nanofiber (blue line) and the exposure model (red line). (b) Poynting vector in the XZ(y¼0) plane of the model. (c) Poynting vector in the YZ (x ¼0.05) plane of the photoresist layer. (d) Poynting vector in the XY (z¼ 0.555224) plane of the photoresist layer. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Z. Wei et al. / Optics Communications 343 (2015) 195–200
nanofibers, the photoresist layer and the silicon wafer are 1.47, 1.69 and 4.76 respectively, the diameter and the length of the nanofiber are 250 nm and 3.6 μm ( 0.5 r zr3.1) respectively, the thicknesses of the photoresist layer and the silicon wafer are 100 nm and 350 nm respectively, and the amplitude of Gauss source is 1 V/ m. In the simulations, broadband light source is used, and the wavelength is fixed at 442 nm. In order to analyze the interaction of optical energy and the photoresist layer, the real part of the total poynting vector was studied. As shown in Fig. 2(a), Poynting vector (blue line) in the Xdirection of an isolated 250-nm-diameter nanofiber demonstrates that a part of optical energy propagates outside the nanofiber as evanescent wave. Besides, Poynting vector (red line) in the X-direction of the model implies that a majority of optical energy is transmitted into the photoresist layer. Fig. 2(b) and (c) illustrates that the transmission mainly occurs in the range of 0.1 rz r1.0, and optical energy is maximal at z¼ 0.55522 in the photoresist layer. There is little optical energy in the nanofiber of the range of z 41. Fig. 2(d) illustrates that the light spot in the photoresist layer is approximately restricted in a rectangular area whose width is about 200 nm. Given the exposure threshold, the interaction area of optical energy and the photoresist layer could be further reduced by adopting proper light dose. The simulations demonstrated that the diameter of nanofibers determine the full width at half maximum (FWHM) value of the light spot in the photoresist layer. Fig. 3(a) shows the distribution of Y-direction Poynting vector of light guided by silica nanofibers with diameters ranging from 50 nm to 400 nm. They present Gauss distribution and their FWHM values were calculated and drawn in Fig. 4(b). Fig. 4 illustrates that the peak value and the FWHM values of Y-direction Poynting vector change with the variation of the nanofiber diameter. Attractively, not only the peak value of Poynting vector achieves maximum, but also the FWHM value reaches minimum (202 nm) when the nanofiber diameter is 250 nm. Importantly, compared with Laser direct writing using the tip of a nanofiber as a point source, the smallest FWHM value of the light spot in the photoresist layer is smaller, indicating that the exposure resolution is increased. Furthermore, given the exposure threshold and Gauss distribution of Poynting vector, the power of the laser also exerts significant influence on the width of exposed lines.
197
3. Experiment scheme Photolithography using lateral surface of nanofibers is schematically shown in Fig. 4. The short and inflexible nanofibers used in Fig. 4(a) could expose lines. As shown in Fig. 4(a), a He–Cd laser device with a 442 nm wavelength is used in the experiment. The laser is coupled into a single mode (SM) fiber with a coupler fixed on the laser device and its output power can be tuned by the SM variable attenuator. The tail end of the fiber is fabricated into a section of nanofiber. A silicon wafer covered with photoresist (AZ MIR701) is placed on a XYZ motorized stage with a resolution of 50 nm/pulse, and a fiber clamp installed on a manual XYZ stage fixes the fiber. The surface of two stages keep parallel with the horizontal plane. The whole process of the experiment is monitored by a microscope camera. The movie of the nanofiber and its shadow is displayed on the screen. Firstly, the substrate moves in the negative X-direction to approach the nanofiber and then stops when light vanishes, because optical energy is transmitted from the nanofiber into the photoresist layer. Secondly, the substrate moves in the Y-direction with designed distance. Repeating the above steps could obtain another line. Besides, as shown in Fig. 4(b), using the tapered fiber fixed on another manual XYZ stage, the silica nanofiber could be manipulated to expose some patterns such as curves and rings. Then turn on the laser and light is injected into the silica nanofiber to expose the photoresist layer. After that, the substrate moves in the X-direction so that the nanofiber leaves the photoresist layer and the whole pattern could be exposed on the photoresist layer. Compared with laser direct writing using the tip of a nanofiber as a point source, the photolithography using lateral surface of nanofibers has several advantages. First, optical energy avoids divergence so that narrower linewidth could be obtained. Second, due to confined evanescent wave and optical tunneling effect, the resolution of photolithography increases obviously.
4. Experimental process and results 4.1. The process of fabricating nanofibers Using a spin coater with the speed of 6000 revolutions per minute, dilute photoresist (photoresist: thinner ¼1:3) was spun onto a silicon wafer to form the photoresist layer whose thickness was about 100 nm. Besides, the exposure threshold of photoresist
Fig. 3. (a) The distribution of Y-direction Poynting vector of light guided by silica nanofibers with diameters varying from 50 nm to 400 nm in the photoresist layer. (b) The FWHM values of Y-direction Poynting vector of light guided by silica nanofibers with different diameters in the photoresist layer.
198
Z. Wei et al. / Optics Communications 343 (2015) 195–200
Fig. 4. (a) The experiment scheme of writing lines. (b) The experiment scheme of writing patterns.
Fig. 5. (a) The process of heating and drawing single mode fiber to generate silica microfibers or nanofibers. (b) The process of corroding silica microfiber by a drop of hydrofluoric acid to form a section of nanofiber. (c) SEM image of a 22-mm-length nanofiber. (d) The SEM image of a 212-nm-diameter nanofiber. (e) The SEM image of a 247nm-diameter nanofiber. (f) The SEM image of a length of nanofiber.
Fig. 6. (a) The SEM image of an exposed line with160 nm width. (b) SEM image of an exposed line with 238 nm width. (c) The SEM image of an exposed line with 75 nm width.
(AZ MIR 701) is 125 J/m2 per 100 nm thickness, therefore the exposure threshold is 0.25 μJ for a 200-nm-width line with a 1 cm length. Fig. 2(b) shows that optical energy is mainly guided into the photoresist layer in the range of 0.1 rz r1.0, thus the maximal length of the exposed line is about 900 nm. Fortunately, flexible
nanofibers could contact the photoresist layer section by section, so that longer lines, curve lines and rings could be exposed. Fig. 3 illustrates that the nanofiber with a 250 nm diameter gives rise to the smallest FWHM value of Poynting vector of the photoresist layer. The process of fabricating nanofibers used in the
Z. Wei et al. / Optics Communications 343 (2015) 195–200
199
Fig. 7. The relationship between the amplitude of the source and the width of exposed lines.
experiment is shown in Fig. 5(a) and (b). Firstly, the tail end of the single mode fiber was heated by butane flame (1300 °C flame temperature), meanwhile, it was drawn rapidly (about 0.4 m/s) to form a length of microfiber with a 2 μm diameter. Secondly, the microfiber was corroded about 20 min by dipping into a drop of dilute hydrofluoric acid to form a length of nanofiber. Fig. 5 (c) shows a scanning electron microscope (SEM) image of a 22mm-length nanofiber and Fig. 5(d) shows that its diameter is 212 nm. Another SEM image of a 247-nm-diameter nanofiber is shown in Fig. 5(e). They were so short and inflexible that straight lines could be exposed. Besides, nanofibers used in the experiment scheme of writing patterns were directly fabricated by heating and drawing the fiber slowly (about 0.01 m/s). Fig. 5(f) shows a length of nanofiber which is so long and flexile that some patterns, such as curves and rings could be exposed.
Fig. 9. (a) The optical microscope image of a straight exposed line. (b) The SEM image of an exposed line with a 294 nm width. (c) The optical microscope image of a curve exposed line. (d) The optical microscope image of an exposed ring.
4.2. The experimental results of photolithography Photolithography using lateral surface of nanofibers was carried out with a 247-nm-diameter nanofiber, and the SEM images of the exposed lines are shown in Fig. 6. When the output power was 0.26 μW, a 160-nm-width line shown in Fig. 6(a) was obtained. And when we increased the output power to 0.52 μW, a 238-nm-width exposed line was obtained, as shown in Fig. 6(b). The duration of exposure was about 2000 milliseconds. Especially, because the scattering of evanescent wave led to smaller optical energy in the tail end of the nanofiber, some exposed lines with sub-100 nm width, existed at the tail end of an exposed line, were
Fig. 8. Poynting vector in the XY plane of the model with different nanofibers. (a) 50-nm-diameter nanofiber. (b) 100-nm-diameter. nanofiber. (c) 150-nm-diameter. nanofiber. (d) 200-nm-diameter. nanofiber. (e) 250-nm-diameter. nanofiber. (f) 300-nm-diameter. nanofiber. (g) 350-nm-diameter. nanofiber. (h) 400-nm-diameter. nanofiber.
200
Z. Wei et al. / Optics Communications 343 (2015) 195–200
observed in the experiment. For example, an exposed line with a 75 nm width was shown in Fig. 6(c). Experiments demonstrated that photolithography using lateral surface of nanofibers could obtain exposed lines with widths ranging from 75 nm to 294 nm. However, the width of exposed lines could not decrease infinitely because optical energy below exposure threshold leads to incomplete exposure. The relationship between the amplitude of the source and the width of exposed lines could be calculated with the help of FDTD simulations. The diameter of the nanofiber in the simulations is set as 250 nm. As shown in Fig. 7, with the enhancement of the amplitude of the source, the calculated value of the width of exposed lines increases from 156 nm to 616 nm, and the slope decreases increasingly. The relationship between the diameter of nanofibers and the FWHM value of Poynting vector in the Y-direction of the photoresist layer is shown in Fig. 3. The reason why the FWHM value of Poynting vector achieves minimum when the diameter of a nanofiber is 250 nm could be explained by the distribution of Poynting vector in the XY plane. As shown in Fig. 8, with the diameter of nanofibers decreasing from 250 nm to 50 nm, more and more optical energy propagates in the form of evanescent wave. And with the diameter of nanofibers increasing from 250 nm to 400 nm, the area of the nanofiber's lateral surface facing the photoresist layer increases correspondingly. Consequently, using the nanofiber with a 250 nm diameter, the interaction area between the optical energy and the photoresist layer reaches minimum value. With the experience and knowledge mentioned above, long exposed lines and patterns could be exposed. Fig. 9(a) shows an optical microscope image of a long exposed line and Fig. 9 (b) illustrates that its width is 294 nm. Fig. 9(c) shows an image of a curve line and Fig. 9(d) shows an image of an annular line. The magnification of the microscope is 1000 and the length of these lines is several hundred micrometers. Obviously, the exposed patterns are decided by the form of nanofibers, and their length depends on the nanofiber length. With this technology, it is easy to fabricate some patterns, such as curves and rings.
5. Conclusions In conclusion, avoiding tip divergence, lateral surface of nanofibers for confined exposure was used to enhance the resolution of photolithography. Experimentally, using a 247-nm-diameter
nanofiber, exposed lines with width below half-wavelength, even below 100 nm were obtained. Furthermore, exposed curve lines and annular lines implied large potential of this technique for high-efficiency photolithography.
Acknowledgments This research was supported by the National Natural Science Foundations of China (No. 61076109).
References [1] C. Vieu, F. Carcenac, A. Pépin, Y. Chen, M. Mejias, A. Lebib, L. Manin-Ferlazzo, L. Couraud, H. Launois, Appl. Surf. Sci. 164 (2009) 111–117. [2] Georg H. Enevoldsen, Henry P. Pinto, Adam S. Foster, Mona C.R. Jensen, Angelika Kühnle, Michael Reichling, Werner A. Hofer, Jeppe V. Lauritsen, Flemming Besenbacher, Phys. Rev. B 78 (2008) 045416. [3] E. Koetter, D. Drakova, G. Doyen, Phys. Rev. B 53 (1996) 16595–16608. [4] Guoguang Feng Tian, Jian Yang, Jianfeng Bai, Changlun Xu, Yiyong Hou, Liang, Kaiwei Wang, Opt. Express 17 (2009) 19960–19968. [5] Emil Wolf, John T. Foley, Opt. Lett. 23 (1998) 16–18. [6] Martin Oheim, Dinah Loerke, Robert H. Chow, Walter Stühmer, Phil. Trans. R. Soc. Lond. B 354 (1999) 307–318. [7] Peter J. Reece, Veneranda Garcés-Chávez, Kishan Dholakia, Appl. Phys. Lett. 88 (2006) 221116. [8] S. Kawata, T. Tani, Opt. Lett. 21 (1996) 1768–1770. [9] C. Zettner, M. Yoda, “Particle velocity field measurements in a near-wall flow using evanescent wave illumination,”, Exp. Fluids 34 (2003) 115–121. [10] Athanasios N. Chryssis, Simarjeet S. Saini, Sang M. Lee, Hyunmin Yi, William E. Bentley, Mario Dagenais, IEEE J. Sel. Top. Quantum Electron. 11 (2005) 864–872. [11] Ph. Balcou, L. Dutriaux, Phys. Rev. Lett. 78 (1997) 851–854. [12] N.J. Harrick, Phys. Rev. Lett. 4 (1960) 224–226. [13] N.J. Harrick, Phys. Rev. Lett. 125 (1962) 1165–1170. [14] S. Zhu, A.W. Yu, D. Hawley, R. Roy, Am. J. Phys 54 (1986) 601–606. [15] JunHo Kim, Ki-Bong Song, Micron 38 (2007) 409–426. [16] R.C. Reddick, R.J. Warmack, D.W. Chilcott, S.L. Sharp, T.L. Ferrell, Rev. Sci. Instrum. 61 (1990) 3669–3677. [17] Jacques Bures, R.éne Ghosh, J. Opt. Soc. Am. A 16 (1999) 1992–1996. [18] L. Tong, L. Hu, J. Zhang, J. Qiu, Q. Yang, J. Lou, Y. Shen, J. He, Z. Ye, Opt. Express 14 (2006) 82–87. [19] L.M. Tong, R.R. Gattass, J.B. Ashcom, S.L. He, J.Y. Lou, M.Y. Shen, I. Maxwell, E. Mazur, Nature 426 (2003) 816–819. [20] G. Sagué, A. Baade, A. Rauschenbeutel, New J. Phys. 10 (2008) 113008. [21] X. Jiang, L. Tong, G. Vienne, X. Guo, A. Tsao, Q. Yang, D. Yang, Appl. Phys. Lett. 88 (2006) 223501. [22] Pavel Polynkin, Alexander Polynkin, N. Peyghambarian, Masud Mansuripur, Opt. Lett. 30 (2005) 1273–1275. [23] M. Sumetsky, Y. Dulashko, A. Hale, Opt. Express 12 (2004) 3521–3531.
Contents lists available at ScienceDirect
Optics Communications journal homepage: www.elsevier.com/locate/optcom
Photolithography using lateral surface of nanofibers Zhen Wei, Jian Bai n, Chen Wang, Neibin Hu, Jianfeng Xu, Yuan Yao, Yiyong Liang, Kaiwei Wang, Changlun Hou, Guoguang Yang State Key Laboratory of Modern Optical Instrumentation, Department of Optical Engineering, Zhejiang University, Hangzhou 310027, China
art ic l e i nf o
a b s t r a c t
Article history: Received 10 July 2014 Received in revised form 14 November 2014 Accepted 23 November 2014 Available online 7 December 2014
To enhance the resolution of photolithography, we demonstrate a technique that confines the exposure area by using lateral surface of nanofibers. Due to evanescent wave and optical tunneling effect, the interaction area of optical energy and the photoresist layer is confined into sub-wavelength dimension. Illuminated by a He–Cd laser device with a 442 nm wavelength, exposed lines with sub-wavelength width were obtained by using a nanofiber with a 247 nm diameter. Furthermore, curve lines and annular lines were obtained by manipulating the shape of nanofibers on the photoresist layer. & 2015 Elsevier B.V. All rights reserved.
Keywords: Fiber optics Nanolithography Near-field optical recording Tunneling
1. Introduction Although electron-beam lithography and atomic force microscope (AFM) etching have obtained resolution beyond 50 nm [1–3], they could not satisfy mass production of nano-components. Enhancing the resolution of photolithography to subwavelength dimension is urgent for fabricating nano-sensors, nano-detectors and so on. Recently, exposed lines with 350-nmresolution were obtained by using the tip of a nanofiber as an exposure source [4]. Unfortunately, light divergence and diffraction limit impede further improvement of resolution. The purpose of this paper is to enhance the resolution of photolithography by narrowing down the interaction area of optical energy and the photoresist layer. Evanescent wave confined in sub-wavelength scale has brought about a lot of application, such as Scanning Near Fields Optical Microscope (SNOM), optical micromanipulation and so on [5–10]. Besides, based on the frustrated total internal reflection, a kind of optical tunneling effect, Photon Scanning Tunneling Microscope (PSTM) could form images with λ/10 lateral resolution [11–16]. Silica nanofibers with diameters ranging from 50 to several hundred nanometers were fabricated successfully. With the decrease of the nanofiber diameter, the proportion of optical fields propagating as evanescent wave increases dramatically [17–19]. Actually, nanofibers have brought about a lot of applications, such as nano-lasers, nano-interferometers and so on [20–23]. n
Corresponding author. E-mail address: [email protected] (J. Bai).
http://dx.doi.org/10.1016/j.optcom.2014.11.071 0030-4018/& 2015 Elsevier B.V. All rights reserved.
In this paper, the simulations demonstrated that nanofibers confine the interaction area of optical energy and the photoresist layer in sub-wavelength dimension so that the linewidth of the light spot in the photoresist layer decreases to 200 nm. Experimentally, using a 247-nm-diameter nanofiber, exposed lines with 75, 160 and 238 nm width were achieved under the illumination of different powers. This basic research provides a novel method for photolithography.
2. Principles and simulations 2.1. The principle of photolithography using lateral surface of nanofibers The model of photolithography using lateral surface of nanofibers is illustrated in Fig. 1(a): a nanofiber's lateral surface intimately contacts the photoresist layer (0 rxr 0.1, 0.4 ryr 0.4, 0 rzr 3.1) which is coated on a silicon wafer (0.1 rx r0.35, 0.4 ry r0.4, 0 rzr 3.1). The principle of the exposure could be explained by analyzing the XY plane of the model. As shown in Fig. 1 (b), because the refractive index of the photoresist layer is larger than that of the nanofiber, optical energy could be transmitted from the nanofiber into the photoresist layer at the connected point (x¼0, y¼0). However, at the disconnected region, optical energy could not be transmitted into the photoresist layer unless the optical tunneling effect occurs. Actually, the optical tunneling effect generates a transparent aperture in the air (blue region). As shown in Fig. 1(b), inside the critical plane which is defined as y¼ a(a≠0),
196
Z. Wei et al. / Optics Communications 343 (2015) 195–200
Fig. 1. (a) The model of photolithography using lateral surface of nanofibers and (b) the XY plane of the model. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
the evanescent wave is intense enough so that optical energy could penetrate the air and expose the photoresist layer, but optical energy could not penetrate the air if y4a. Consequently, the dimension of the transparent aperture is confined to 2a and lower power brings out narrower optical channel.
2.2. The simulated results of photolithography model To investigate the transmission of optical energy, the model is simulated with the Lumerical FDTD Solutions. The simulations are set with the following parameters: the refractive indices of
Fig. 2. (a) Poynting vector in the X-direction. of an isolated 250-nm-diameter. nanofiber (blue line) and the exposure model (red line). (b) Poynting vector in the XZ(y¼0) plane of the model. (c) Poynting vector in the YZ (x ¼0.05) plane of the photoresist layer. (d) Poynting vector in the XY (z¼ 0.555224) plane of the photoresist layer. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Z. Wei et al. / Optics Communications 343 (2015) 195–200
nanofibers, the photoresist layer and the silicon wafer are 1.47, 1.69 and 4.76 respectively, the diameter and the length of the nanofiber are 250 nm and 3.6 μm ( 0.5 r zr3.1) respectively, the thicknesses of the photoresist layer and the silicon wafer are 100 nm and 350 nm respectively, and the amplitude of Gauss source is 1 V/ m. In the simulations, broadband light source is used, and the wavelength is fixed at 442 nm. In order to analyze the interaction of optical energy and the photoresist layer, the real part of the total poynting vector was studied. As shown in Fig. 2(a), Poynting vector (blue line) in the Xdirection of an isolated 250-nm-diameter nanofiber demonstrates that a part of optical energy propagates outside the nanofiber as evanescent wave. Besides, Poynting vector (red line) in the X-direction of the model implies that a majority of optical energy is transmitted into the photoresist layer. Fig. 2(b) and (c) illustrates that the transmission mainly occurs in the range of 0.1 rz r1.0, and optical energy is maximal at z¼ 0.55522 in the photoresist layer. There is little optical energy in the nanofiber of the range of z 41. Fig. 2(d) illustrates that the light spot in the photoresist layer is approximately restricted in a rectangular area whose width is about 200 nm. Given the exposure threshold, the interaction area of optical energy and the photoresist layer could be further reduced by adopting proper light dose. The simulations demonstrated that the diameter of nanofibers determine the full width at half maximum (FWHM) value of the light spot in the photoresist layer. Fig. 3(a) shows the distribution of Y-direction Poynting vector of light guided by silica nanofibers with diameters ranging from 50 nm to 400 nm. They present Gauss distribution and their FWHM values were calculated and drawn in Fig. 4(b). Fig. 4 illustrates that the peak value and the FWHM values of Y-direction Poynting vector change with the variation of the nanofiber diameter. Attractively, not only the peak value of Poynting vector achieves maximum, but also the FWHM value reaches minimum (202 nm) when the nanofiber diameter is 250 nm. Importantly, compared with Laser direct writing using the tip of a nanofiber as a point source, the smallest FWHM value of the light spot in the photoresist layer is smaller, indicating that the exposure resolution is increased. Furthermore, given the exposure threshold and Gauss distribution of Poynting vector, the power of the laser also exerts significant influence on the width of exposed lines.
197
3. Experiment scheme Photolithography using lateral surface of nanofibers is schematically shown in Fig. 4. The short and inflexible nanofibers used in Fig. 4(a) could expose lines. As shown in Fig. 4(a), a He–Cd laser device with a 442 nm wavelength is used in the experiment. The laser is coupled into a single mode (SM) fiber with a coupler fixed on the laser device and its output power can be tuned by the SM variable attenuator. The tail end of the fiber is fabricated into a section of nanofiber. A silicon wafer covered with photoresist (AZ MIR701) is placed on a XYZ motorized stage with a resolution of 50 nm/pulse, and a fiber clamp installed on a manual XYZ stage fixes the fiber. The surface of two stages keep parallel with the horizontal plane. The whole process of the experiment is monitored by a microscope camera. The movie of the nanofiber and its shadow is displayed on the screen. Firstly, the substrate moves in the negative X-direction to approach the nanofiber and then stops when light vanishes, because optical energy is transmitted from the nanofiber into the photoresist layer. Secondly, the substrate moves in the Y-direction with designed distance. Repeating the above steps could obtain another line. Besides, as shown in Fig. 4(b), using the tapered fiber fixed on another manual XYZ stage, the silica nanofiber could be manipulated to expose some patterns such as curves and rings. Then turn on the laser and light is injected into the silica nanofiber to expose the photoresist layer. After that, the substrate moves in the X-direction so that the nanofiber leaves the photoresist layer and the whole pattern could be exposed on the photoresist layer. Compared with laser direct writing using the tip of a nanofiber as a point source, the photolithography using lateral surface of nanofibers has several advantages. First, optical energy avoids divergence so that narrower linewidth could be obtained. Second, due to confined evanescent wave and optical tunneling effect, the resolution of photolithography increases obviously.
4. Experimental process and results 4.1. The process of fabricating nanofibers Using a spin coater with the speed of 6000 revolutions per minute, dilute photoresist (photoresist: thinner ¼1:3) was spun onto a silicon wafer to form the photoresist layer whose thickness was about 100 nm. Besides, the exposure threshold of photoresist
Fig. 3. (a) The distribution of Y-direction Poynting vector of light guided by silica nanofibers with diameters varying from 50 nm to 400 nm in the photoresist layer. (b) The FWHM values of Y-direction Poynting vector of light guided by silica nanofibers with different diameters in the photoresist layer.
198
Z. Wei et al. / Optics Communications 343 (2015) 195–200
Fig. 4. (a) The experiment scheme of writing lines. (b) The experiment scheme of writing patterns.
Fig. 5. (a) The process of heating and drawing single mode fiber to generate silica microfibers or nanofibers. (b) The process of corroding silica microfiber by a drop of hydrofluoric acid to form a section of nanofiber. (c) SEM image of a 22-mm-length nanofiber. (d) The SEM image of a 212-nm-diameter nanofiber. (e) The SEM image of a 247nm-diameter nanofiber. (f) The SEM image of a length of nanofiber.
Fig. 6. (a) The SEM image of an exposed line with160 nm width. (b) SEM image of an exposed line with 238 nm width. (c) The SEM image of an exposed line with 75 nm width.
(AZ MIR 701) is 125 J/m2 per 100 nm thickness, therefore the exposure threshold is 0.25 μJ for a 200-nm-width line with a 1 cm length. Fig. 2(b) shows that optical energy is mainly guided into the photoresist layer in the range of 0.1 rz r1.0, thus the maximal length of the exposed line is about 900 nm. Fortunately, flexible
nanofibers could contact the photoresist layer section by section, so that longer lines, curve lines and rings could be exposed. Fig. 3 illustrates that the nanofiber with a 250 nm diameter gives rise to the smallest FWHM value of Poynting vector of the photoresist layer. The process of fabricating nanofibers used in the
Z. Wei et al. / Optics Communications 343 (2015) 195–200
199
Fig. 7. The relationship between the amplitude of the source and the width of exposed lines.
experiment is shown in Fig. 5(a) and (b). Firstly, the tail end of the single mode fiber was heated by butane flame (1300 °C flame temperature), meanwhile, it was drawn rapidly (about 0.4 m/s) to form a length of microfiber with a 2 μm diameter. Secondly, the microfiber was corroded about 20 min by dipping into a drop of dilute hydrofluoric acid to form a length of nanofiber. Fig. 5 (c) shows a scanning electron microscope (SEM) image of a 22mm-length nanofiber and Fig. 5(d) shows that its diameter is 212 nm. Another SEM image of a 247-nm-diameter nanofiber is shown in Fig. 5(e). They were so short and inflexible that straight lines could be exposed. Besides, nanofibers used in the experiment scheme of writing patterns were directly fabricated by heating and drawing the fiber slowly (about 0.01 m/s). Fig. 5(f) shows a length of nanofiber which is so long and flexile that some patterns, such as curves and rings could be exposed.
Fig. 9. (a) The optical microscope image of a straight exposed line. (b) The SEM image of an exposed line with a 294 nm width. (c) The optical microscope image of a curve exposed line. (d) The optical microscope image of an exposed ring.
4.2. The experimental results of photolithography Photolithography using lateral surface of nanofibers was carried out with a 247-nm-diameter nanofiber, and the SEM images of the exposed lines are shown in Fig. 6. When the output power was 0.26 μW, a 160-nm-width line shown in Fig. 6(a) was obtained. And when we increased the output power to 0.52 μW, a 238-nm-width exposed line was obtained, as shown in Fig. 6(b). The duration of exposure was about 2000 milliseconds. Especially, because the scattering of evanescent wave led to smaller optical energy in the tail end of the nanofiber, some exposed lines with sub-100 nm width, existed at the tail end of an exposed line, were
Fig. 8. Poynting vector in the XY plane of the model with different nanofibers. (a) 50-nm-diameter nanofiber. (b) 100-nm-diameter. nanofiber. (c) 150-nm-diameter. nanofiber. (d) 200-nm-diameter. nanofiber. (e) 250-nm-diameter. nanofiber. (f) 300-nm-diameter. nanofiber. (g) 350-nm-diameter. nanofiber. (h) 400-nm-diameter. nanofiber.
200
Z. Wei et al. / Optics Communications 343 (2015) 195–200
observed in the experiment. For example, an exposed line with a 75 nm width was shown in Fig. 6(c). Experiments demonstrated that photolithography using lateral surface of nanofibers could obtain exposed lines with widths ranging from 75 nm to 294 nm. However, the width of exposed lines could not decrease infinitely because optical energy below exposure threshold leads to incomplete exposure. The relationship between the amplitude of the source and the width of exposed lines could be calculated with the help of FDTD simulations. The diameter of the nanofiber in the simulations is set as 250 nm. As shown in Fig. 7, with the enhancement of the amplitude of the source, the calculated value of the width of exposed lines increases from 156 nm to 616 nm, and the slope decreases increasingly. The relationship between the diameter of nanofibers and the FWHM value of Poynting vector in the Y-direction of the photoresist layer is shown in Fig. 3. The reason why the FWHM value of Poynting vector achieves minimum when the diameter of a nanofiber is 250 nm could be explained by the distribution of Poynting vector in the XY plane. As shown in Fig. 8, with the diameter of nanofibers decreasing from 250 nm to 50 nm, more and more optical energy propagates in the form of evanescent wave. And with the diameter of nanofibers increasing from 250 nm to 400 nm, the area of the nanofiber's lateral surface facing the photoresist layer increases correspondingly. Consequently, using the nanofiber with a 250 nm diameter, the interaction area between the optical energy and the photoresist layer reaches minimum value. With the experience and knowledge mentioned above, long exposed lines and patterns could be exposed. Fig. 9(a) shows an optical microscope image of a long exposed line and Fig. 9 (b) illustrates that its width is 294 nm. Fig. 9(c) shows an image of a curve line and Fig. 9(d) shows an image of an annular line. The magnification of the microscope is 1000 and the length of these lines is several hundred micrometers. Obviously, the exposed patterns are decided by the form of nanofibers, and their length depends on the nanofiber length. With this technology, it is easy to fabricate some patterns, such as curves and rings.
5. Conclusions In conclusion, avoiding tip divergence, lateral surface of nanofibers for confined exposure was used to enhance the resolution of photolithography. Experimentally, using a 247-nm-diameter
nanofiber, exposed lines with width below half-wavelength, even below 100 nm were obtained. Furthermore, exposed curve lines and annular lines implied large potential of this technique for high-efficiency photolithography.
Acknowledgments This research was supported by the National Natural Science Foundations of China (No. 61076109).
References [1] C. Vieu, F. Carcenac, A. Pépin, Y. Chen, M. Mejias, A. Lebib, L. Manin-Ferlazzo, L. Couraud, H. Launois, Appl. Surf. Sci. 164 (2009) 111–117. [2] Georg H. Enevoldsen, Henry P. Pinto, Adam S. Foster, Mona C.R. Jensen, Angelika Kühnle, Michael Reichling, Werner A. Hofer, Jeppe V. Lauritsen, Flemming Besenbacher, Phys. Rev. B 78 (2008) 045416. [3] E. Koetter, D. Drakova, G. Doyen, Phys. Rev. B 53 (1996) 16595–16608. [4] Guoguang Feng Tian, Jian Yang, Jianfeng Bai, Changlun Xu, Yiyong Hou, Liang, Kaiwei Wang, Opt. Express 17 (2009) 19960–19968. [5] Emil Wolf, John T. Foley, Opt. Lett. 23 (1998) 16–18. [6] Martin Oheim, Dinah Loerke, Robert H. Chow, Walter Stühmer, Phil. Trans. R. Soc. Lond. B 354 (1999) 307–318. [7] Peter J. Reece, Veneranda Garcés-Chávez, Kishan Dholakia, Appl. Phys. Lett. 88 (2006) 221116. [8] S. Kawata, T. Tani, Opt. Lett. 21 (1996) 1768–1770. [9] C. Zettner, M. Yoda, “Particle velocity field measurements in a near-wall flow using evanescent wave illumination,”, Exp. Fluids 34 (2003) 115–121. [10] Athanasios N. Chryssis, Simarjeet S. Saini, Sang M. Lee, Hyunmin Yi, William E. Bentley, Mario Dagenais, IEEE J. Sel. Top. Quantum Electron. 11 (2005) 864–872. [11] Ph. Balcou, L. Dutriaux, Phys. Rev. Lett. 78 (1997) 851–854. [12] N.J. Harrick, Phys. Rev. Lett. 4 (1960) 224–226. [13] N.J. Harrick, Phys. Rev. Lett. 125 (1962) 1165–1170. [14] S. Zhu, A.W. Yu, D. Hawley, R. Roy, Am. J. Phys 54 (1986) 601–606. [15] JunHo Kim, Ki-Bong Song, Micron 38 (2007) 409–426. [16] R.C. Reddick, R.J. Warmack, D.W. Chilcott, S.L. Sharp, T.L. Ferrell, Rev. Sci. Instrum. 61 (1990) 3669–3677. [17] Jacques Bures, R.éne Ghosh, J. Opt. Soc. Am. A 16 (1999) 1992–1996. [18] L. Tong, L. Hu, J. Zhang, J. Qiu, Q. Yang, J. Lou, Y. Shen, J. He, Z. Ye, Opt. Express 14 (2006) 82–87. [19] L.M. Tong, R.R. Gattass, J.B. Ashcom, S.L. He, J.Y. Lou, M.Y. Shen, I. Maxwell, E. Mazur, Nature 426 (2003) 816–819. [20] G. Sagué, A. Baade, A. Rauschenbeutel, New J. Phys. 10 (2008) 113008. [21] X. Jiang, L. Tong, G. Vienne, X. Guo, A. Tsao, Q. Yang, D. Yang, Appl. Phys. Lett. 88 (2006) 223501. [22] Pavel Polynkin, Alexander Polynkin, N. Peyghambarian, Masud Mansuripur, Opt. Lett. 30 (2005) 1273–1275. [23] M. Sumetsky, Y. Dulashko, A. Hale, Opt. Express 12 (2004) 3521–3531.