Temperature dependence of laser-induced micro/nanostructures for femtosecond laser irradiation of silicon Guoliang Deng,1 Guoying Feng,1,2 Kui Liu,1 and Shouhuan Zhou1,3 1
School of Electronics and Information Engineering, Sichuan University, 610064, China 2
e-mail: [email protected] 3
e-mail: [email protected]
Received 6 December 2013; revised 1 April 2014; accepted 1 April 2014; posted 7 April 2014 (Doc. ID 202607); published 5 May 2014
The temperature dependence (from 25°C to 350°C) of laser-induced micro/nanostructures for multiple linearly polarized femtosecond laser pulse (pulse duration τ 35 fs, wavelength λ 800 nm) irradiation of silicon in air is studied experimentally. Distinct micro/nanostructures are fabricated at elevated temperature. Low spatial frequency, laser-induced periodic ripple structures (LSFL), which are perpendicular to the polarization of the laser beam, are formed at all temperatures. Micrometer-size grooves, which are oriented perpendicular to the LSFL ripples, have been observed in the central part of the irradiated area above 150°C. The threshold to fabricate the LSFL ripples goes from 1.65 to 2 kJ∕m2 while the temperature of the substrate increases from 25°C to 350°C. The possible mechanism of the temperature dependence of the micro/nanostructure generation is also discussed. These results demonstrate that temperature is an important parameter to be tuned to tailor the micro/nanostructure fabrication. © 2014 Optical Society of America OCIS codes: (050.6624) Subwavelength structures; (140.3390) Laser materials processing; (160.6000) Semiconductor materials. http://dx.doi.org/10.1364/AO.53.003004
1. Introduction
Micro/nanostructures on silicon, which are fabricated by femtosecond laser irradiation, have attracted much interest in the past decades because of their broad potential applications and rich physics [1–5]. Femtosecond lasers can create various micro/ nanoscale structures such as quasi-periodic sharp conical spikes [6] and high regular arrays of nanorods [7]. Another representative nanostructure which has been abundantly researched is the laser-induced periodic surface structure (LIPSS) [1,8–12]. Two kinds of ripples with different periods can be observed on the silicon surface after femtosecond laser irradiation with moderate fluence. 1559-128X/14/143004-06$15.00/0 © 2014 Optical Society of America 3004
APPLIED OPTICS / Vol. 53, No. 14 / 10 May 2014
The low-spatial-frequency LIPSS (LSFL) has a range of period between 0.6λ and λ, while the highspatial-frequency LIPSS has a period less than 0.6λ [8]. In addition, micrometer-size grooves, which are parallel to the polarization, can be found with either a large number of pulses or higher fluence irradiation [9,10]. The formation of these micro/ nanostructures depends on the strong reaction between the femtosecond laser pulses and the silicon surface. Much research has been done into changing the properties of these structures, such as the spatial period, by changing the parameters of the incident irradiation (wavelength, polarization, pulse number, incident angle, etc.) [9,13], but little research has been done into altering the property of the substrate [14]. Prior research has shown that the dielectric function and band gap of silicon can be tuned by changing the temperature [15–17]. This implies that
studying the properties of micro/nanostructures of silicon at elevated temperature can provide a new approach to understanding the rich physics of the process of micro/nanostructure generation and a new way to control the properties of these structures. In this work, the temperature dependence of micro/nanostructures formed on a silicon surface by femtosecond laser irradiation was studied experimentally. Micro/nanostructures are fabricated with an 800 nm femtosecond laser at elevated temperature of the substrate. The results show that the laser-induced micro/nanostructure has distinct temperature dependence. The threshold of the laser fluence for producing LSFL ripples increases as the temperature rises. In addition, parallel micrometer-size grooves are generated at higher temperatures. The results reveal that the temperature of the wafer is an important parameter for controlling micro/nanostructure fabrication. The possible mechanism of the temperature dependence of micro/nanostructure generation is also discussed. 2. Experiment Details
The silicon wafer used in this experiment was an n-type silicon wafer with resistivity less than 5 × 10−5 Ω · m. The silicon wafer was first cleaned ultrasonically with methanol and acetone in turn for 5 min, and then it was rinsed with deionized water and dried with N2 . The wafer was mounted onto an electric heater. The heater was mounted onto a three-axis translation stage. The temperature of the silicon wafer, which was measured real-time by a thermal imager (NEC H2640), can be controlled between 25°C and 350°C. The femtosecond pulsed laser system is composed of a Ti: sapphire oscillator (Coherent, Mantis) followed by a chirped-pulse-regenerative amplifier (Coherent, Legend) operating at 1 kHz with a central wavelength of 800 nm. The duration of the laser pulse is about 35 fs. The pulse energy can be up to 3.5 mJ. The linearly polarized laser pulses were focused by a 0.2 m focal length lens and incident normally to the silicon surface. The beam profile was analyzed with a CCD beam analyzer. In this work, x indicates the direction perpendicular to the polarization of the laser beam, while y indicates the direction perpendicular to x. The spot of the laser beam was an ellipse and had a spatially Gaussian intensity distribution with a radius of ωx0 90 μm, ωy0 60 μm (as measured at 1∕e2 of the maximum intensity) on the sample surface. The fluence of the incident laser can be adjusted continuously using a combination of a polarizer and a half-wave plate. The pulse energy used in the experiment was 21 μJ∕pulse, which gave a fluence of about 2.5 kJ∕m2 in the center of the beam. The wafer was moved 42 μm when it was heated up to 350°C, arising from the thermal expansion of the heater and the wafer itself. The fluence change caused by the thermal expansion was less than 0.03 kJ∕m2 .
The sample was made by moving the silicon wafer when the laser was working. The speed was 1.5 mm∕s, which gave nearly 40 pulses per laser spot size by considering the beam radius mentioned above. After irradiation, the sample was analyzed using an optical microscope (Keyence) and scanning electron microscopy (JSM-7500F), respectively. 3. Results
The optical micrographs of the overall structures formed at elevated temperature are shown in Fig. 1. The arrows indicate the polarization direction of the laser. The structures generated at different temperatures have distinct features; even the laser parameters are kept the same. The groove-like structures can be observed when the temperature rises above 150°C. The fine representative SEM micrograph of Figs. 1(a) and 1(d), which are the structures formed at room temperature and 350°C, respectively, are shown in Figs. 2(a) and 2(c). The LSFL ripples were formed in both cases, which can be further demonstrated by Figs. 2(b) and 2(d); they are the Fourier transforms of Figs. 2(a) and 2(c), respectively. The two bright dots in the horizontal direction of Fig. 2(b) indicate that the vertical ripples, which are perpendicular to the polarization of the laser beam, are formed at room temperature [Fig. 2(a)]. The LSFL ripples are nearly uniform with a spatial period around 651 nm. Aside from vertical ripples, horizontal grooves are also generated at 350°C in the central part of the structured area, which is shown in Fig. 2(c). The spots in the vertical direction of Fig. 2(d) can further confirm that the horizontal grooves are formed. These grooves have a width of around 700–800 nm and a period around 1 μm. Interestingly, period structures are also observed within these grooves. The two faint, sickle-like spots in the horizontal direction in Fig. 2(d) indicate that ripples were formed as well in vertical direction with a period around 738 nm, which is bigger than those
Fig. 1. Optical micrograph of structures formed at different temperatures: (a) room temperature (25°C), (b) 150°C, (c) 250°C, and (d) 350°C. The arrow indicates the direction of the laser polarization. 10 May 2014 / Vol. 53, No. 14 / APPLIED OPTICS
3005
Fig. 2. Scanning electron micrograph of structures formed at different temperatures. (a) Structure formed at room temperature, (b) Fourier transform of Fig. 1(a), (c) structure formed at 350°C, and (d) Fourier transform of Fig. 1(c). The arrow indicates the direction of the laser polarization.
structured area to the edge while the distance increases because of the fact that the intensity of the femtosecond laser beam has a spatially Gaussian distribution. L is the half-width of the structured area, while ω0 is the radius of the laser spot (1∕e2 ). The arrow indicates the direction of the polarization of the laser beam. Figure 3(b) was the periphery of the structured area with a higher magnification. Horizontal straight ripples can be seen clearly with a direction perpendicular to the laser polarization. The location where the ripples can be barely formed can be found using the optical microscope image as shown in Fig. 3(b), and then L can be deduced. The widths of the structured areas formed at elevated temperature are shown in Fig. 4. They decrease from 80.8 to 58.9 μm monotonically when the temperature changes from 25°C to 350°C. The local fluence, which is the threshold when the LSFL ripples first appear, can be deduced according to [9] FL F max · exp
formed at room temperature. In addition, the LSFL ripples generated at 350°C are shallower than those fabricated at room temperature. However, as shown in Fig. 1, the structured area becomes smaller when the temperature goes higher. Since the LSFL can always be observed on the periphery of the structured area, the shrinkage of the structured area indicates that the threshold for generation of the LSFL ripples becomes higher as the temperature increases. The threshold of the LSFL ripples then was studied systematically at elevated temperature. The structured area in Fig. 3(a) was formed by scanning the silicon wafer with the laser parameters mentioned above at room temperature. The fluence decreases monotonically from the center of the
−2 · L2 ; ω2x0
(1)
where F max is the fluence in the center of the beam and ωx0 is the beam width (1∕e2 ) in the direction perpendicular to the polarization of the beam. The threshold at which the LSFL ripples can be observed in our experiment is around 1.65 kJ∕m2 at room temperature, which is in agreement with the previous results reported by Wang et al. [18]. The temperature dependence of the threshold is shown in Fig. 4. The threshold increases monotonically over the entire temperature range between 25°C and 350°C. The threshold at 350°C is 2 kJ∕m2 , corresponding to a 21.2% increase compared to that at room temperature. 4. Discussion
The temperature dependence of femtosecond laserinduced micro/nanostructures and the threshold at which these structures can be expected to be formed
Fig. 3. Optical micrograph of the structured area. (a) Structure generated by femtosecond laser scanning at room temperature; (b) periphery of the structured area. L is the half-width of the structured area and ω0 is the radius of the beam profile (1∕e2 ). The arrow indicates the direction of the laser polarization. 3006
APPLIED OPTICS / Vol. 53, No. 14 / 10 May 2014
Fig. 4. Threshold of the LSFL ripple generation and the width of the structured area at elevated temperatures.
may be due to several mechanisms. Several researchers have reported the temperature dependence of the optical properties of silicon [15,17]. It was experimentally found that n and k (n and k are the real and imaginary part of the complex refraction index) are a linear and exponential function of temperature, respectively. Therefore, the temperature dependence of the reflectivity R of normal incidence can be deduced. R increases linearly with the temperature in the temperature range 25°C–350°C at the wavelength of 800 nm according to [15] RT R0 aR · T − T 0R ;
(2)
where T 0R is 25°C as the reference temperature, while aR is the temperature coefficient. aR is 3.06 × 10−5 ∕°C at 800 nm. Then we can obtain that the reflection relatively increases 3.24% at 350°C. This may contribute to part of the increase of the threshold of the ripple generation. However, the change is very small compared to the increase of the threshold. On the other hand, as silicon is an indirect band gap semiconductor, the absorption needs phonon assistance. At higher temperature, higher-momentum phonons can be generated to enhance the excitation of electrons. The linear absorption coefficient actually increases as temperature rises. The band gap of silicon also decreases as the temperature increases [16]. This mechanism can enhance the electron generation and increase the absorption coefficient further. Aside from the linear absorption coefficient, two-photon absorption also plays a very important role in carrier excitation [19]. There are some studies on the temperature dependence of the two-photon absorption coefficient on semiconductors [20] but few on silicon [21]. In general, the total absorption increases at elevated temperature. The increased absorption coefficient will then cause decrease of the ablation threshold of silicon. This coincides with our experiment results, in which the structures in the central part of the structured area changed from uniform ripples to ripples and grooves (which are usually obtained with high fluence or a large number of pulses) when the temperature increases, as mentioned earlier. This may result in an increase in absorption coefficient, which can deposit more energy with the same amount of incident light. Then the surface of silicon can have a thicker melt depth to form the grooves, but we should notice that the grooves generated here have a smaller size than usual, which is several micrometers. However, the threshold to generate LSFL ripples increases as the temperature increases, as shown in Fig. 4. This difference may be due to the mechanism of LSFL ripple generation. Bonse and several other groups proposed that the excitation of surface plasmon polaritons (SPPs) plays a crucial role in the formation of the femtosecond laser-induced LSFL ripples [1,8,9]. Bonse and co-workers successfully explained the role of SPPs in the formation of LSFL
by combining the generally accepted first-principles theory of Sipe with a Drude model. With this model, they can take into account the transient changes in the optical properties of the material due to the excitation of a dense electron-hole plasma [1]. This model shows that the energy deposited into the irradiated material is proportional to η × jbj, where η is a response function describing the efficacy, which has sharp peaks and leads to inhomogeneous absorption. The factor b represents the surface roughness, which is a slowly changing function for a surface with homogeneously distributed roughness. The increases of the threshold for LSFL ripple fabrication can be explained with this model qualitatively. Prior research has shown that the minimum carrier density required to generate SPPs is 4.61 × 1021 ∕cm3 in the case of a laser irradiation of silicon using a wavelength of 800 nm [22]. This high density of carriers can be attained with the assistance of
Fig. 5. (a) 2D gray maps of efficacy factor η for crystalline silicon as a function of the normalized LIPSS wave vector under excitation of 1.65 kJ∕m2 irradiation at 800 nm; (b) maximum value of the efficacy factor η for the LSFL feature as a function of the damping rate. 10 May 2014 / Vol. 53, No. 14 / APPLIED OPTICS
3007
defects created by multiple laser shots under the fluence we used. As shown in Figs. 5(a) and 5(b), the 2D gray maps of efficacy factor η for crystalline silicon as a function of the normalized LIPSS wave vector has been calculated under this critical carrier density by using a Drude–Sipe model. The effective mass of carriers mopt 0.18 and the Drude damping rate 1∕τD 1 fs−1 [19] have been chosen for room temperature. Following Refs. [1] and [8], the surface roughness was modeled with the values of s 0.4 and f 0.1, which represent the assumption of spherically shaped islands. The two sickle-like spots in the horizontal direction in Fig. 5(a) are associated with the LSFL, which are perpendicular to the polarization. A significant phonon–electron scattering will occur while the temperature of silicon substrate get higher, which will increase the damping rate and reduce the oscillation of the surface plasma [23,24]. We have evaluated the horizontal cross sections through the LSFL feature in the efficacy factor maps (K y 0). Figure 5(b) shows the maximum value of the efficacy factor η for the LSFL feature as a function of the damping rate. An efficacy factor η changes from 0.98 to 0.38 as the damping rate increases from 1 to 2 fs−1 . This means higher incident energy is required to maintain the SPP generation. Eventually, the threshold to fabricate ripples becomes higher. 5. Conclusion
In summary, the temperature dependence of the micro/nanostructures for femtosecond laser irradiation of silicon is studied experimentally. The result shows that the laser-induced micro/nanostructure has distinct temperature dependence. LSFL ripples can be generated at almost all of the structured area at room temperature, while at 350°C it can only be found at the periphery of the area. Aside from LSFL ripples, groove-like structures are generated in the central part of the structured area when the temperature is higher than 150°C. This probably comes from the enhancement of the absorption of light at elevated temperature. The threshold to fabricate the LSFL ripples goes from 1.65 to 2 kJ∕m2 , while the temperature of the substrate increases from 25°C to 350°C. The significant damping of the carrier oscillation by the phonon–electron scattering under high temperature is the possible mechanism of the increase of the threshold. These results demonstrate that temperature can be used to tailor the nanostructure fabrication. This work was supported by Major Program of National Natural Science Foundation of China (60890200). The authors thank Professor Jianguo Chen and Dr. Yang Li for helpful discussions. References 1. J. Bonse, A. Rosenfeld, and J. Kruger, “On the role of surface plasmon polaritons in the formation of laser-induced periodic surface structures upon irradiation of silicon by femtosecondlaser pulses,” J. Appl. Phys. 106, 104910 (2009). 3008
APPLIED OPTICS / Vol. 53, No. 14 / 10 May 2014
2. M. Halbwax, T. Sarnet, P. Delaporte, M. Sentis, H. Etienne, F. Torregrosa, V. Vervisch, I. Perichaud, and S. Martinuzzi, “Micro and nano-structuration of silicon by femtosecond laser: application to silicon photovoltaic cells fabrication,” Thin Solid Films 516, 6791–6795 (2008). 3. T. Sarnet, G. Kerrien, N. Yaakoubi, A. Bosseboeuf, E. Dufour-Gergam, D. Débarre, J. Boulmer, K. Kakushima, C. Laviron, M. Hernandez, J. Venturini, and T. Bourouina, “Laser doping for microelectronics and microtechnology,” Appl. Surf. Sci. 247, 537–544 (2005). 4. N. C. Kerr, S. E. Clark, and D. C. Emmony, “High resolution imaging studies into the formation of laser-induced periodic surface structures,” Appl. Opt. 28, 3718–3724 (1989). 5. Y. Peng, D. Zhang, H. Chen, Y. Wen, S. Luo, L. Chen, K. Chen, and Y. Zhu, “Differences in the evolution of surfacemicrostructured silicon fabricated by femtosecond laser pulses with different wavelength,” Appl. Opt. 51, 635–639 (2012). 6. T.-H. Her, R. J. Finlay, C. Wu, S. Deliwala, and E. Mazur, “Microstructuring of silicon with femtosecond laser pulses,” Appl. Phys. Lett. 73, 1673–1675 (1998). 7. M. Shen, J. E. Carey, C. H. Crouch, M. Kandyla, H. A. Stone, and E. Mazur, “High-density regular arrays of nanometerscale rods formed on silicon surfaces via femtosecond laser irradiation in water,” Nano Lett. 8, 2087–2091 (2008). 8. M. Huang, F. Zhao, Y. Cheng, N. Xu, and Z. Xu, “Origin of laser-induced near-subwavelength ripples: interference between surface plasmons and incident laser,” ACS Nano 3, 4062–4070 (2009). 9. J. Bonse and J. Kruger, “Pulse number dependence of laser-induced periodic surface structures for femtosecond laser irradiation of silicon,” J. Appl. Phys. 108, 034903 (2010). 10. J. Bonse, M. Munz, and H. Sturm, “Structure formation on the surface of indium phosphide irradiated by femtosecond laser pulses,” J. Appl. Phys. 97, 013538 (2005). 11. J. Z. P. Skolski, G. R. B. E. Römer, J. V. Obona, V. Ocelik, A. J. Huis in ’t Veld, and J. T. M. De Hosson, “Laser-induced periodic surface structures: fingerprints of light localization,” Phys. Rev. B 85, 075320 (2012). 12. T. J. Y. Derrien, R. Torres, T. Sarnet, M. Sentis, and T. E. Itina, “Formation of femtosecond laser induced surface structures on silicon: insights from numerical modeling and single pulse experiments,” Appl. Surf. Sci. 258, 9487–9490 (2012). 13. O. Varlamova, F. Costache, M. Ratzke, and J. Reif, “Control parameters in pattern formation upon femtosecond laser ablation,” Appl. Surf. Sci. 253, 7932–7936 (2007). 14. K. Liu, G. Feng, G. Deng, and W. Li, “Difference in microstructures induced by femtosecond laser scanning on silicon surface at different temperature,” Chin. J. Lasers 39, 56–60 (2012). 15. B. K. Sun, X. Zhang, and C. P. Grigoropoulos, “Spectral optical functions of silicon in the range of 1.13–4.96 eV at elevated temperatures,” Int. J. Heat Mass Trans. 40, 1591–1600 (1997). 16. K. P. O’Donnell and X. Chen, “Temperature dependence of semiconductor band gaps,” Appl. Phys. Lett. 58, 2924–2926 (1991). 17. G. E. Jellison, Jr. and F. A. Modine, “Optical functions of silicon between 1.7 and 4.7 eV at elevated temperatures,” Phys. Rev. B 27, 7466–7472 (1983). 18. C. Wang, H. Huo, M. Johnson, M. Shen, and E. Mazur, “The thresholds of surface nano-/micro-morphology modifications with femtosecond laser pulse irradiations,” Nanotechnology 21, 075304 (2010). 19. K. Sokolowski-Tinten and D. von der Linde, “Generation of dense electron-hole plasmas in silicon,” Phys. Rev. B 61, 2643–2650 (2000). 20. S. Krishnamurthy, Z. G. Yu, L. P. Gonzalez, and S. Guha, “Temperature- and wavelength-dependent two-photon and free-carrier absorption in GaAs, InP, GaInAs, and InAsP,” J. Appl. Phys. 109, 033102 (2011). 21. J. Thorstensen and S. E. Foss, “Temperature dependent ablation threshold in silicon using ultrashort laser pulses,” J. Appl. Phys. 112, 103514 (2012).
22. T. J.-Y. Derrien, T. E. Itina, R. Torres, T. Sarnet, and M. Sentis, “Possible surface plasmon polariton excitation under femtosecond laser irradiation of silicon,” J. Appl. Phys. 114, 083104 (2013). 23. H. P. Chiang, C. W. Chen, J. J. Wu, H. L. Li, T. Y. Lin, E. J. Sánchez, and P. T. Leung, “Effects of temperature on the
surface plasmon resonance at a metal–semiconductor interface,” Thin Solid Films 515, 6953–6961 (2007). 24. H. P. Chiang, Y. C. Wang, P. T. Leung, and W. S. Tse, “A theoretical model for the temperature-dependent sensitivity of the optical sensor based on surface plasmon resonance,” Opt. Commun. 188, 283–289 (2001).
10 May 2014 / Vol. 53, No. 14 / APPLIED OPTICS
3009
School of Electronics and Information Engineering, Sichuan University, 610064, China 2
e-mail: [email protected] 3
e-mail: [email protected]
Received 6 December 2013; revised 1 April 2014; accepted 1 April 2014; posted 7 April 2014 (Doc. ID 202607); published 5 May 2014
The temperature dependence (from 25°C to 350°C) of laser-induced micro/nanostructures for multiple linearly polarized femtosecond laser pulse (pulse duration τ 35 fs, wavelength λ 800 nm) irradiation of silicon in air is studied experimentally. Distinct micro/nanostructures are fabricated at elevated temperature. Low spatial frequency, laser-induced periodic ripple structures (LSFL), which are perpendicular to the polarization of the laser beam, are formed at all temperatures. Micrometer-size grooves, which are oriented perpendicular to the LSFL ripples, have been observed in the central part of the irradiated area above 150°C. The threshold to fabricate the LSFL ripples goes from 1.65 to 2 kJ∕m2 while the temperature of the substrate increases from 25°C to 350°C. The possible mechanism of the temperature dependence of the micro/nanostructure generation is also discussed. These results demonstrate that temperature is an important parameter to be tuned to tailor the micro/nanostructure fabrication. © 2014 Optical Society of America OCIS codes: (050.6624) Subwavelength structures; (140.3390) Laser materials processing; (160.6000) Semiconductor materials. http://dx.doi.org/10.1364/AO.53.003004
1. Introduction
Micro/nanostructures on silicon, which are fabricated by femtosecond laser irradiation, have attracted much interest in the past decades because of their broad potential applications and rich physics [1–5]. Femtosecond lasers can create various micro/ nanoscale structures such as quasi-periodic sharp conical spikes [6] and high regular arrays of nanorods [7]. Another representative nanostructure which has been abundantly researched is the laser-induced periodic surface structure (LIPSS) [1,8–12]. Two kinds of ripples with different periods can be observed on the silicon surface after femtosecond laser irradiation with moderate fluence. 1559-128X/14/143004-06$15.00/0 © 2014 Optical Society of America 3004
APPLIED OPTICS / Vol. 53, No. 14 / 10 May 2014
The low-spatial-frequency LIPSS (LSFL) has a range of period between 0.6λ and λ, while the highspatial-frequency LIPSS has a period less than 0.6λ [8]. In addition, micrometer-size grooves, which are parallel to the polarization, can be found with either a large number of pulses or higher fluence irradiation [9,10]. The formation of these micro/ nanostructures depends on the strong reaction between the femtosecond laser pulses and the silicon surface. Much research has been done into changing the properties of these structures, such as the spatial period, by changing the parameters of the incident irradiation (wavelength, polarization, pulse number, incident angle, etc.) [9,13], but little research has been done into altering the property of the substrate [14]. Prior research has shown that the dielectric function and band gap of silicon can be tuned by changing the temperature [15–17]. This implies that
studying the properties of micro/nanostructures of silicon at elevated temperature can provide a new approach to understanding the rich physics of the process of micro/nanostructure generation and a new way to control the properties of these structures. In this work, the temperature dependence of micro/nanostructures formed on a silicon surface by femtosecond laser irradiation was studied experimentally. Micro/nanostructures are fabricated with an 800 nm femtosecond laser at elevated temperature of the substrate. The results show that the laser-induced micro/nanostructure has distinct temperature dependence. The threshold of the laser fluence for producing LSFL ripples increases as the temperature rises. In addition, parallel micrometer-size grooves are generated at higher temperatures. The results reveal that the temperature of the wafer is an important parameter for controlling micro/nanostructure fabrication. The possible mechanism of the temperature dependence of micro/nanostructure generation is also discussed. 2. Experiment Details
The silicon wafer used in this experiment was an n-type silicon wafer with resistivity less than 5 × 10−5 Ω · m. The silicon wafer was first cleaned ultrasonically with methanol and acetone in turn for 5 min, and then it was rinsed with deionized water and dried with N2 . The wafer was mounted onto an electric heater. The heater was mounted onto a three-axis translation stage. The temperature of the silicon wafer, which was measured real-time by a thermal imager (NEC H2640), can be controlled between 25°C and 350°C. The femtosecond pulsed laser system is composed of a Ti: sapphire oscillator (Coherent, Mantis) followed by a chirped-pulse-regenerative amplifier (Coherent, Legend) operating at 1 kHz with a central wavelength of 800 nm. The duration of the laser pulse is about 35 fs. The pulse energy can be up to 3.5 mJ. The linearly polarized laser pulses were focused by a 0.2 m focal length lens and incident normally to the silicon surface. The beam profile was analyzed with a CCD beam analyzer. In this work, x indicates the direction perpendicular to the polarization of the laser beam, while y indicates the direction perpendicular to x. The spot of the laser beam was an ellipse and had a spatially Gaussian intensity distribution with a radius of ωx0 90 μm, ωy0 60 μm (as measured at 1∕e2 of the maximum intensity) on the sample surface. The fluence of the incident laser can be adjusted continuously using a combination of a polarizer and a half-wave plate. The pulse energy used in the experiment was 21 μJ∕pulse, which gave a fluence of about 2.5 kJ∕m2 in the center of the beam. The wafer was moved 42 μm when it was heated up to 350°C, arising from the thermal expansion of the heater and the wafer itself. The fluence change caused by the thermal expansion was less than 0.03 kJ∕m2 .
The sample was made by moving the silicon wafer when the laser was working. The speed was 1.5 mm∕s, which gave nearly 40 pulses per laser spot size by considering the beam radius mentioned above. After irradiation, the sample was analyzed using an optical microscope (Keyence) and scanning electron microscopy (JSM-7500F), respectively. 3. Results
The optical micrographs of the overall structures formed at elevated temperature are shown in Fig. 1. The arrows indicate the polarization direction of the laser. The structures generated at different temperatures have distinct features; even the laser parameters are kept the same. The groove-like structures can be observed when the temperature rises above 150°C. The fine representative SEM micrograph of Figs. 1(a) and 1(d), which are the structures formed at room temperature and 350°C, respectively, are shown in Figs. 2(a) and 2(c). The LSFL ripples were formed in both cases, which can be further demonstrated by Figs. 2(b) and 2(d); they are the Fourier transforms of Figs. 2(a) and 2(c), respectively. The two bright dots in the horizontal direction of Fig. 2(b) indicate that the vertical ripples, which are perpendicular to the polarization of the laser beam, are formed at room temperature [Fig. 2(a)]. The LSFL ripples are nearly uniform with a spatial period around 651 nm. Aside from vertical ripples, horizontal grooves are also generated at 350°C in the central part of the structured area, which is shown in Fig. 2(c). The spots in the vertical direction of Fig. 2(d) can further confirm that the horizontal grooves are formed. These grooves have a width of around 700–800 nm and a period around 1 μm. Interestingly, period structures are also observed within these grooves. The two faint, sickle-like spots in the horizontal direction in Fig. 2(d) indicate that ripples were formed as well in vertical direction with a period around 738 nm, which is bigger than those
Fig. 1. Optical micrograph of structures formed at different temperatures: (a) room temperature (25°C), (b) 150°C, (c) 250°C, and (d) 350°C. The arrow indicates the direction of the laser polarization. 10 May 2014 / Vol. 53, No. 14 / APPLIED OPTICS
3005
Fig. 2. Scanning electron micrograph of structures formed at different temperatures. (a) Structure formed at room temperature, (b) Fourier transform of Fig. 1(a), (c) structure formed at 350°C, and (d) Fourier transform of Fig. 1(c). The arrow indicates the direction of the laser polarization.
structured area to the edge while the distance increases because of the fact that the intensity of the femtosecond laser beam has a spatially Gaussian distribution. L is the half-width of the structured area, while ω0 is the radius of the laser spot (1∕e2 ). The arrow indicates the direction of the polarization of the laser beam. Figure 3(b) was the periphery of the structured area with a higher magnification. Horizontal straight ripples can be seen clearly with a direction perpendicular to the laser polarization. The location where the ripples can be barely formed can be found using the optical microscope image as shown in Fig. 3(b), and then L can be deduced. The widths of the structured areas formed at elevated temperature are shown in Fig. 4. They decrease from 80.8 to 58.9 μm monotonically when the temperature changes from 25°C to 350°C. The local fluence, which is the threshold when the LSFL ripples first appear, can be deduced according to [9] FL F max · exp
formed at room temperature. In addition, the LSFL ripples generated at 350°C are shallower than those fabricated at room temperature. However, as shown in Fig. 1, the structured area becomes smaller when the temperature goes higher. Since the LSFL can always be observed on the periphery of the structured area, the shrinkage of the structured area indicates that the threshold for generation of the LSFL ripples becomes higher as the temperature increases. The threshold of the LSFL ripples then was studied systematically at elevated temperature. The structured area in Fig. 3(a) was formed by scanning the silicon wafer with the laser parameters mentioned above at room temperature. The fluence decreases monotonically from the center of the
−2 · L2 ; ω2x0
(1)
where F max is the fluence in the center of the beam and ωx0 is the beam width (1∕e2 ) in the direction perpendicular to the polarization of the beam. The threshold at which the LSFL ripples can be observed in our experiment is around 1.65 kJ∕m2 at room temperature, which is in agreement with the previous results reported by Wang et al. [18]. The temperature dependence of the threshold is shown in Fig. 4. The threshold increases monotonically over the entire temperature range between 25°C and 350°C. The threshold at 350°C is 2 kJ∕m2 , corresponding to a 21.2% increase compared to that at room temperature. 4. Discussion
The temperature dependence of femtosecond laserinduced micro/nanostructures and the threshold at which these structures can be expected to be formed
Fig. 3. Optical micrograph of the structured area. (a) Structure generated by femtosecond laser scanning at room temperature; (b) periphery of the structured area. L is the half-width of the structured area and ω0 is the radius of the beam profile (1∕e2 ). The arrow indicates the direction of the laser polarization. 3006
APPLIED OPTICS / Vol. 53, No. 14 / 10 May 2014
Fig. 4. Threshold of the LSFL ripple generation and the width of the structured area at elevated temperatures.
may be due to several mechanisms. Several researchers have reported the temperature dependence of the optical properties of silicon [15,17]. It was experimentally found that n and k (n and k are the real and imaginary part of the complex refraction index) are a linear and exponential function of temperature, respectively. Therefore, the temperature dependence of the reflectivity R of normal incidence can be deduced. R increases linearly with the temperature in the temperature range 25°C–350°C at the wavelength of 800 nm according to [15] RT R0 aR · T − T 0R ;
(2)
where T 0R is 25°C as the reference temperature, while aR is the temperature coefficient. aR is 3.06 × 10−5 ∕°C at 800 nm. Then we can obtain that the reflection relatively increases 3.24% at 350°C. This may contribute to part of the increase of the threshold of the ripple generation. However, the change is very small compared to the increase of the threshold. On the other hand, as silicon is an indirect band gap semiconductor, the absorption needs phonon assistance. At higher temperature, higher-momentum phonons can be generated to enhance the excitation of electrons. The linear absorption coefficient actually increases as temperature rises. The band gap of silicon also decreases as the temperature increases [16]. This mechanism can enhance the electron generation and increase the absorption coefficient further. Aside from the linear absorption coefficient, two-photon absorption also plays a very important role in carrier excitation [19]. There are some studies on the temperature dependence of the two-photon absorption coefficient on semiconductors [20] but few on silicon [21]. In general, the total absorption increases at elevated temperature. The increased absorption coefficient will then cause decrease of the ablation threshold of silicon. This coincides with our experiment results, in which the structures in the central part of the structured area changed from uniform ripples to ripples and grooves (which are usually obtained with high fluence or a large number of pulses) when the temperature increases, as mentioned earlier. This may result in an increase in absorption coefficient, which can deposit more energy with the same amount of incident light. Then the surface of silicon can have a thicker melt depth to form the grooves, but we should notice that the grooves generated here have a smaller size than usual, which is several micrometers. However, the threshold to generate LSFL ripples increases as the temperature increases, as shown in Fig. 4. This difference may be due to the mechanism of LSFL ripple generation. Bonse and several other groups proposed that the excitation of surface plasmon polaritons (SPPs) plays a crucial role in the formation of the femtosecond laser-induced LSFL ripples [1,8,9]. Bonse and co-workers successfully explained the role of SPPs in the formation of LSFL
by combining the generally accepted first-principles theory of Sipe with a Drude model. With this model, they can take into account the transient changes in the optical properties of the material due to the excitation of a dense electron-hole plasma [1]. This model shows that the energy deposited into the irradiated material is proportional to η × jbj, where η is a response function describing the efficacy, which has sharp peaks and leads to inhomogeneous absorption. The factor b represents the surface roughness, which is a slowly changing function for a surface with homogeneously distributed roughness. The increases of the threshold for LSFL ripple fabrication can be explained with this model qualitatively. Prior research has shown that the minimum carrier density required to generate SPPs is 4.61 × 1021 ∕cm3 in the case of a laser irradiation of silicon using a wavelength of 800 nm [22]. This high density of carriers can be attained with the assistance of
Fig. 5. (a) 2D gray maps of efficacy factor η for crystalline silicon as a function of the normalized LIPSS wave vector under excitation of 1.65 kJ∕m2 irradiation at 800 nm; (b) maximum value of the efficacy factor η for the LSFL feature as a function of the damping rate. 10 May 2014 / Vol. 53, No. 14 / APPLIED OPTICS
3007
defects created by multiple laser shots under the fluence we used. As shown in Figs. 5(a) and 5(b), the 2D gray maps of efficacy factor η for crystalline silicon as a function of the normalized LIPSS wave vector has been calculated under this critical carrier density by using a Drude–Sipe model. The effective mass of carriers mopt 0.18 and the Drude damping rate 1∕τD 1 fs−1 [19] have been chosen for room temperature. Following Refs. [1] and [8], the surface roughness was modeled with the values of s 0.4 and f 0.1, which represent the assumption of spherically shaped islands. The two sickle-like spots in the horizontal direction in Fig. 5(a) are associated with the LSFL, which are perpendicular to the polarization. A significant phonon–electron scattering will occur while the temperature of silicon substrate get higher, which will increase the damping rate and reduce the oscillation of the surface plasma [23,24]. We have evaluated the horizontal cross sections through the LSFL feature in the efficacy factor maps (K y 0). Figure 5(b) shows the maximum value of the efficacy factor η for the LSFL feature as a function of the damping rate. An efficacy factor η changes from 0.98 to 0.38 as the damping rate increases from 1 to 2 fs−1 . This means higher incident energy is required to maintain the SPP generation. Eventually, the threshold to fabricate ripples becomes higher. 5. Conclusion
In summary, the temperature dependence of the micro/nanostructures for femtosecond laser irradiation of silicon is studied experimentally. The result shows that the laser-induced micro/nanostructure has distinct temperature dependence. LSFL ripples can be generated at almost all of the structured area at room temperature, while at 350°C it can only be found at the periphery of the area. Aside from LSFL ripples, groove-like structures are generated in the central part of the structured area when the temperature is higher than 150°C. This probably comes from the enhancement of the absorption of light at elevated temperature. The threshold to fabricate the LSFL ripples goes from 1.65 to 2 kJ∕m2 , while the temperature of the substrate increases from 25°C to 350°C. The significant damping of the carrier oscillation by the phonon–electron scattering under high temperature is the possible mechanism of the increase of the threshold. These results demonstrate that temperature can be used to tailor the nanostructure fabrication. This work was supported by Major Program of National Natural Science Foundation of China (60890200). The authors thank Professor Jianguo Chen and Dr. Yang Li for helpful discussions. References 1. J. Bonse, A. Rosenfeld, and J. Kruger, “On the role of surface plasmon polaritons in the formation of laser-induced periodic surface structures upon irradiation of silicon by femtosecondlaser pulses,” J. Appl. Phys. 106, 104910 (2009). 3008
APPLIED OPTICS / Vol. 53, No. 14 / 10 May 2014
2. M. Halbwax, T. Sarnet, P. Delaporte, M. Sentis, H. Etienne, F. Torregrosa, V. Vervisch, I. Perichaud, and S. Martinuzzi, “Micro and nano-structuration of silicon by femtosecond laser: application to silicon photovoltaic cells fabrication,” Thin Solid Films 516, 6791–6795 (2008). 3. T. Sarnet, G. Kerrien, N. Yaakoubi, A. Bosseboeuf, E. Dufour-Gergam, D. Débarre, J. Boulmer, K. Kakushima, C. Laviron, M. Hernandez, J. Venturini, and T. Bourouina, “Laser doping for microelectronics and microtechnology,” Appl. Surf. Sci. 247, 537–544 (2005). 4. N. C. Kerr, S. E. Clark, and D. C. Emmony, “High resolution imaging studies into the formation of laser-induced periodic surface structures,” Appl. Opt. 28, 3718–3724 (1989). 5. Y. Peng, D. Zhang, H. Chen, Y. Wen, S. Luo, L. Chen, K. Chen, and Y. Zhu, “Differences in the evolution of surfacemicrostructured silicon fabricated by femtosecond laser pulses with different wavelength,” Appl. Opt. 51, 635–639 (2012). 6. T.-H. Her, R. J. Finlay, C. Wu, S. Deliwala, and E. Mazur, “Microstructuring of silicon with femtosecond laser pulses,” Appl. Phys. Lett. 73, 1673–1675 (1998). 7. M. Shen, J. E. Carey, C. H. Crouch, M. Kandyla, H. A. Stone, and E. Mazur, “High-density regular arrays of nanometerscale rods formed on silicon surfaces via femtosecond laser irradiation in water,” Nano Lett. 8, 2087–2091 (2008). 8. M. Huang, F. Zhao, Y. Cheng, N. Xu, and Z. Xu, “Origin of laser-induced near-subwavelength ripples: interference between surface plasmons and incident laser,” ACS Nano 3, 4062–4070 (2009). 9. J. Bonse and J. Kruger, “Pulse number dependence of laser-induced periodic surface structures for femtosecond laser irradiation of silicon,” J. Appl. Phys. 108, 034903 (2010). 10. J. Bonse, M. Munz, and H. Sturm, “Structure formation on the surface of indium phosphide irradiated by femtosecond laser pulses,” J. Appl. Phys. 97, 013538 (2005). 11. J. Z. P. Skolski, G. R. B. E. Römer, J. V. Obona, V. Ocelik, A. J. Huis in ’t Veld, and J. T. M. De Hosson, “Laser-induced periodic surface structures: fingerprints of light localization,” Phys. Rev. B 85, 075320 (2012). 12. T. J. Y. Derrien, R. Torres, T. Sarnet, M. Sentis, and T. E. Itina, “Formation of femtosecond laser induced surface structures on silicon: insights from numerical modeling and single pulse experiments,” Appl. Surf. Sci. 258, 9487–9490 (2012). 13. O. Varlamova, F. Costache, M. Ratzke, and J. Reif, “Control parameters in pattern formation upon femtosecond laser ablation,” Appl. Surf. Sci. 253, 7932–7936 (2007). 14. K. Liu, G. Feng, G. Deng, and W. Li, “Difference in microstructures induced by femtosecond laser scanning on silicon surface at different temperature,” Chin. J. Lasers 39, 56–60 (2012). 15. B. K. Sun, X. Zhang, and C. P. Grigoropoulos, “Spectral optical functions of silicon in the range of 1.13–4.96 eV at elevated temperatures,” Int. J. Heat Mass Trans. 40, 1591–1600 (1997). 16. K. P. O’Donnell and X. Chen, “Temperature dependence of semiconductor band gaps,” Appl. Phys. Lett. 58, 2924–2926 (1991). 17. G. E. Jellison, Jr. and F. A. Modine, “Optical functions of silicon between 1.7 and 4.7 eV at elevated temperatures,” Phys. Rev. B 27, 7466–7472 (1983). 18. C. Wang, H. Huo, M. Johnson, M. Shen, and E. Mazur, “The thresholds of surface nano-/micro-morphology modifications with femtosecond laser pulse irradiations,” Nanotechnology 21, 075304 (2010). 19. K. Sokolowski-Tinten and D. von der Linde, “Generation of dense electron-hole plasmas in silicon,” Phys. Rev. B 61, 2643–2650 (2000). 20. S. Krishnamurthy, Z. G. Yu, L. P. Gonzalez, and S. Guha, “Temperature- and wavelength-dependent two-photon and free-carrier absorption in GaAs, InP, GaInAs, and InAsP,” J. Appl. Phys. 109, 033102 (2011). 21. J. Thorstensen and S. E. Foss, “Temperature dependent ablation threshold in silicon using ultrashort laser pulses,” J. Appl. Phys. 112, 103514 (2012).
22. T. J.-Y. Derrien, T. E. Itina, R. Torres, T. Sarnet, and M. Sentis, “Possible surface plasmon polariton excitation under femtosecond laser irradiation of silicon,” J. Appl. Phys. 114, 083104 (2013). 23. H. P. Chiang, C. W. Chen, J. J. Wu, H. L. Li, T. Y. Lin, E. J. Sánchez, and P. T. Leung, “Effects of temperature on the
surface plasmon resonance at a metal–semiconductor interface,” Thin Solid Films 515, 6953–6961 (2007). 24. H. P. Chiang, Y. C. Wang, P. T. Leung, and W. S. Tse, “A theoretical model for the temperature-dependent sensitivity of the optical sensor based on surface plasmon resonance,” Opt. Commun. 188, 283–289 (2001).
10 May 2014 / Vol. 53, No. 14 / APPLIED OPTICS
3009
