Cross-sectional study of high spatial frequency ripples performed on silicon using nanojoule femtosecond laser pulses at high repetition rate

Cross-sectional study of high spatial frequency ripples performed on silicon using nanojoule femtosecond laser pulses at high repetition rate

Applied Surface Science 305 (2014) 670–673 Contents lists available at ScienceDirect Applied Surface Science journal homepage: www.elsevier.com/loca...

1MB Sizes 0 Downloads 39 Views

Applied Surface Science 305 (2014) 670–673

Contents lists available at ScienceDirect

Applied Surface Science journal homepage: www.elsevier.com/locate/apsusc

Cross-sectional study of high spatial frequency ripples performed on silicon using nanojoule femtosecond laser pulses at high repetition rate R. Le Harzic a,∗ , M. Menzel b , S. Henning b , A. Heilmann b , F. Stracke a , H. Zimmermann a,c a

Fraunhofer Institute for Biomedical Engineering (IBMT), Ensheimer Str. 48, D-66386 St. Ingbert, Germany Fraunhofer Institute for Mechanics of Materials (IWM), Halle, Walter-Hülse-Straße 1, 06120 Halle, Germany c Saarland University, Chair Molecular & Cellular Biotechnology/Nanotechnology, D-66123 Saarbrücken, Germany b

a r t i c l e

i n f o

Article history: Received 10 February 2014 Received in revised form 21 March 2014 Accepted 22 March 2014 Available online 31 March 2014 Keywords: Nano-ripples HSFL Femtosecond laser Patterning Silicon Nanograting

a b s t r a c t We report on investigations of fine cross-sectional cuts performed by focused ion beam on periodic high spatial frequency ripples generated on large areas of silicon in air under ultrashort low energy pulses irradiation at high repetition rate. The morphology, depth profile, and aspect ratio which was found to be of 1:1 are quite independent from the energy and number of pulses applied even if a slight decrease of the aspect ratio was found at high energies and high number of applied pulses. Furthermore, even if the orientation of the HSFL is perpendicular to the polarization the profile and aspect ratio of the cross sections are not influenced by the rotation of the polarization. © 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

Introduction High spatial frequency laser-induced periodic surface structures (HSF-LIPSS or HSFL) have been observed on a variety of sample surfaces following irradiation of ultrashort laser pulses [1–13]. HSFL have typical spatial periods in a range of /4–/6 and are solely generated for multiple pulse irradiation with low energy. The origin of these nano-ripples is still quite controversially discussed in the current literature. Proposed mechanisms for HSFL generation include surface instability and self-organization [2], second harmonic generation (SHG) [1,3,13,14], excitation of surface plasmon polaritons [10,15] or Coulomb explosion [16]. If several studies have been performed on the influence of parameters such as the energy per pulse, the number of applied pulses, the polarization, the wavelength, etc. on the morphology and the periodicity of the ripples on the surface of Silicon (Si), much less information is available about the cross-sectional structure of the nano-patterns. Some studies on cross-sectional cuts of ripples with periods closed to the wavelength, known as low spatial frequency LIPSS (LSFL), have been performed in Si [17–20]. Cross-sectional investigations on

∗ Corresponding author. Tel.: +49 6894 980 167. E-mail address: [email protected] (R. Le Harzic).

HSFL generated in Si under water confinement have been presented [21]. These cross-sectional studies give essentially information on the nature and structural phase of the ripples. In this work we focus on the influence of illumination parameters on the cross section profile. We study the influence of the pulse energy and the number of applied pulses as well as the polarization on the morphology, depth profile, and aspect ratio of the HSFL generated on Si in air. For this purpose, cross-sectional cuts of structured fields entirely filled with HSFL were performed by Focused Ion Beam (FIB) under low fluence pulses (nJ) at high repetition rate (MHz). Scanning electron microscopy (SEM) observations on cross-sectional cuts reveal a quite homogeneous sinusoidal-like depth profile of the HSFL with a relative constant aspect ratio until a certain level of energy and pulse number. The orientation of the HSFL is dependent on the direction of the polarization and perpendicular to this latter but no influence of the direction of the polarization has been observed on the profile and aspect ratio of the cross sections of HSFL. Material and methods HSFL have been generated in air on standard one-side-polished, boron-doped p-type Si wafers with a thickness of 525 ␮m and a crystallographic orientation 100 using a compact ultrafast

http://dx.doi.org/10.1016/j.apsusc.2014.03.159 0169-4332/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-SA license (http://creativecommons.org/licenses/by-nc-sa/3.0/).

R. Le Harzic et al. / Applied Surface Science 305 (2014) 670–673

671

Fig. 1. (a) SEM image of HSFL generated in Si at 800 nm, 2.5 nJ/pulse and 160 000 applied pulses after etching. The double arrow represents the polarization of the laser irradiation. (b) SEM image of the corresponding FIB section.

Ti:Sapphire tuneable laser system (80 MHz, Emax ∼ 40 nJ/pulse,  = 140 fs (FWHM),  = 800 nm) coupled to a laser workstation specially designed and developed for accurate micro- and nanoprocessing. Details on the experimental setup can be found in a previous work [22]. Large areas were processed by scanning the sample

in order to yield representative numbers of ripples per profile. The low energy femtosecond laser pulses were focused by the use of a 20× focusing objective with a relatively high numerical aperture (NA) of 0.75 to reach the fluence threshold of HSFL generation. The focus diameter is about 1.3 ␮m at the surface of the sample.

Fig. 2. SEM images of cross sectional profile of HSFL in Si. (a) FIB section overview (refer to Fig. 1(b)). (b) Cross sectional profile overview of the HSFL. (c) Details of the cross sectional morphology and profile of the HSFL. (d) and (e) High magnification of HSFL profiles.

672

R. Le Harzic et al. / Applied Surface Science 305 (2014) 670–673

The SEM images of the surfaces and the depth profiles as well as the FIB preparation were performed in a combined ESEM/FIB instrument FEI Quanta 3D FEG after etching with ammonium fluoride to remove recast matter at the borders of the grooves and in the structures. The FIB cross-sections were performed after the deposition of a protective platinum (Pt) layer on the surface of the sample in the investigation region. A damage sample surface can then be avoided in the preparation and study area. Si material was extracted stepwise using gallium ions with a continuous decreasing energy down to 30 keV for an ultimate fine removal and polishing. The whole FIB procedure i.e., platinum layer deposition, successive removal of Si layer in the depth as well as fine polishing takes about 50 min for a cross section of 15 ␮m × 5 ␮m.

Results and discussion Series of square areas of 500 ␮m × 500 ␮m entirely covered with HSFL have been nanostructured at the surface of Si for FIB crosssectioning. For each area, an experimental parameter has been modified, the energy par pulse at a constant number of applied pulses, the number of applied pulses at a constant energy and the polarization orientation of the laser beam at a constant energy and number of applied pulses. Fig. 1(a) shows details of HSFL generated on Si at an energy of 2.5 nJ/pulse (0.4 J/cm2 ) and 160 000 applied pulses. The number of applied pulses is defined as the effective number of laser pulses per laser spot diameter Npulses = D × f/V with D the spot diameter, f the repetition rate and V the scan velocity. The corresponding FIB section is displayed in Fig. 1(b) where both the HSFL and the Pt cover layer are visible. Successive magnified SEM images of the cross section and profile of HSFL are given in Fig. 2 for more details. As a first remark, one can observe the slight undulation of the whole profile with a period of 1 ␮m which corresponds to the shifting of 1 ␮m between laser line scans during the fabrication of the nanostructured scan field of 500 ␮m × 500 ␮m. The parabolic shape of individual line scan has to be directly related to the Gaussian shape of the laser beam with a higher ablation rate toward the spot center, i.e., higher laser intensity than at the edges. Looking yet more in details on the HSFL themselves, one can observe relative homogeneous ripples in morphology, spacing and depth with a repeatable sinusoidal-like profile. Furthermore the images did not reveal regions of different brightness indicating that the silicon material is quite homogeneous. High magnified images of representative HSFL structures are depicted in Fig. 2(d) and (e). The spacing between the rims which was measured at different points of the profile is quite constant at a value of 125 ± 5 nm. The periodicity follows a /2n law which is in agreement with previous results showing that second harmonic generation (SHG) is involved in the formation of HSFL in silicon [23,24]. The depth is also found to be around 130 ± 5 nm giving an aspect ratio of about 1:1. Experiments have been repeated at different energies up to 7 nJ (1 J/cm2 ) per pulse at the same number of applied pulses. Results are resumed in Fig. 3. The aspect ratio has been calculated as a function of the energy per pulse. In insert a corresponding representative magnified SEM image of the ripple is depicted with its width and depth. As a first remark, there is no influence on the periodicity of the HSFL when the energy increases, its value is constant at 125 ± 5 nm. This is once again in accordance with previous work [22]. The depth, i.e., the aspect ratio is relatively constant toward 1:1 up to 4.5 nJ (0.7 J/cm2 ) per pulse. It decreases then down to about 3:4 at higher energies. It has to be noticed that the profile of the HSFL is more chaotic at higher energies due to a stronger ablation rate and more material removal. Fig. 4 gives measured values of the aspect ratio of the HSFL as a function of the number of applied pulses at a constant energy of

Fig. 3. Aspect ratio of the HSFL as a function of the energy per pulses at a constant number of 160 000 applied pulses. A typical high magnified SEM image of the HSFL is depicted in insert for all the different energy per pulses.

2.5 nJ per pulse. The aspect ratio is relatively constant toward 1:1 once again from 16 000 to 250 000 applied pulses. At upper number of applied pulses the aspect ratio tends also to decrease as for higher energies. The slight reduction in depth at higher energies and/or at a higher number of applied pulses could be simply explained by the redeposition of a significant amount of ablated matter or particles in the HSFL grooves. It could also result from a physical transformation of the solid material itself such as amorphization. Schade et al. [20] showed that ripples consisted of distorted crystalline silicon rich of defects with amorphous silicon in the hollows between the ripples. We have shown that HSFL initiation, formation and arrangement at low fluence and high repetition rate combine structural modification of the surface initiated by a thermal process with second harmonic generation (SHG) [13]. Heat accumulation of successive pulses occurs up to the melting temperature after multi pulses (MHz) laser irradiation at low fluence below or close to the melting threshold and the temperature increases with the energy and/or number of pulses. Generally, a transition between the crystalline and the amorphous phase following laser irradiation is related to a threshold behavior in the solidification velocity [25]. If solidification proceeds with liquid–solid interface velocities higher than a critical value, irradiation leads to the amorphization of the processed material. The solidification speed is strongly affected by the local temperature gradient and the transient cooling conditions after melting. In our case then, an increase of the local temperature with the energy and/or the number of applied pulses as well as a

Fig. 4. Aspect ratio of the HSFL as a function of the number of applied pulses.

R. Le Harzic et al. / Applied Surface Science 305 (2014) 670–673

low dissipation of the heat from the focal between successive pulses could affect the solidification velocity up to a critical value and consequently induce amorphization. However, Straub et al. have not observed evident amorphization of HSFL generated on Si immersed in water [21]. The origin of these observations is still subject to experimental studies. In another experiment, the polarization orientation of the laser beam was modified by the introduction of a half wave plate (/2) in the laser beam path. The polarization was rotated at three angles, 0◦ , 45◦ and 90◦ at the same energy and applied pulses of 2.5 nJ/pulse and 160 000, respectively. Apart the fact that the HSFL direction is perpendicular to the polarization orientation as often observed in the literature and rotate consequently with the polarization angle, the profile of the cross sections of HSFL were quit similar. The aspect ratio was also found to be constant around 1:1 once again. Conclusion SEM investigations of fine cross-sectional cuts performed by FIB on HSFL generated on Si in air under ultrashort low energy pulses irradiation at high repetition rate, reveal a quite homogeneous and repeatable sinusoidal-like depth profile on long distances with constant periodicities and depths of, respectively, 125 ± 5 nm and 130 ± 5 nm giving a corresponding aspect ratio of 1:1. The aspect ratio of 1:1 is relatively constant when the energy per pulse increases at a constant number of applied pulses, same when the number of applied pulses increases at a constant energy per pulse. The morphology and depth of the HSFL are quite independent from the energy and number of pulses applied even if a slight decrease of the aspect ratio was found at high energies and high number of applied pulses. It is not clear up to now if this slight reduction in depth is simply due to the redeposition of ablated particles in the HSFL grooves or due to physical transformations of the superficial material itself. Cross sectional investigations have to be performed on the structural phase or nature of the silicon present in the bulk of the HSFL but also at the rims, and in the bottom of the grooves by TEM analysis for example. Furthermore, no influence of the polarization has been observed on the morphology, depth profile and aspect ratio of the cross sections of HSFL. Acknowledgment This work was supported by the German Research Foundation (DFG, priority programme SPP 1327). References [1] A. Borowiec, H.K. Haugen, Subwavelength ripple formation on the surfaces of compound semiconductors irradiated with femtosecond laser pulses, Appl. Phys. Lett. 82 (2003) 4462–4464. [2] F. Costache, M. Henyk, J. Reif, Surface patterning on insulators upon femtosecond laser ablation, Appl. Surf. Sci. 208/209 (2003) 486–491.

673

[3] T.Q. Jia, H.X. Chen, M. Huang, F.L. Zhao, J.R. Qiu, R.X. Li, Z.Z. Xu, X.K. He, J. Zhang, H. Kuroda, Formation of nanogratings on the surface of a ZnSe crystal irradiated by femtosecond laser pulses, Phys. Rev. B: Condens. Matter 72 (2005) 125429. [4] R. Le Harzic, H. Schuck, D. Sauer, T. Anhut, I. Riemann, T. Velten, K. König, Sub100 nm nanostructuring of silicon by ultrashort laser pulses, Opt. Express 13 (2005) 6651–6656. [5] T.H.R. Crawford, H.K. Haugen, Sub-wavelength surface structures on silicon irradiated by femtosecond laser pulses at 1300 and 2100 nm wavelengths, Appl. Surf. Sci. 253 (2007) 4970–4977. [6] D. Dufft, A. Rosenfeld, S.K. Das, R. Grunwald, J. Bonse, Femtosecond laserinduced periodic surface structures revisited: a comparative study on ZnO, J. Appl. Phys. 105 (2009) 034908. [7] C. Wang, H. Huo, M. Johnson, M. Shen, E. Mazur, The thresholds of surface nano/micro-morphology modifications with femtosecond laser pulse irradiations, Nanotechnology 21 (2010) 075304. [8] U. Chakravarty, R.A. Ganeev, P.A. Naik, J.A. Chakera, M. Babu, P.D. Gupta, Nano-ripple formation on different band-gap semiconductor surfaces using femtosecond pulses, J. Appl. Phys. 109 (2011) 084347. [9] J. Bonse, J. Krüger, S. Höhm, A. Rosenfeld, Femtosecond laser-induced periodic surface structures, J. Laser Appl. 24 (2012) 042006. [10] M. Straub, M. Afshar, D. Feili, H. Seidel, K. Konig, Surface plasmon polariton model of high-spatial frequency laser-induced periodic surface structure generation in silicon, J. Appl. Phys. 111 (2012) 124315. [11] G. Miyaji, K. Miyazaki, K. Zhang, T. Yoshifuji, J. Fujita, Mechanism of femtosecond-laser-induced periodic nanostructure formation on crystalline silicon surface immersed in water, Opt. Express 20 (2012) 14848–14856. [12] M. Barberoglou, D. Gray, E. Magoulakis, C. Fotakis, P.A. Loukakos, E. Stratakis, Controlling ripples periodicity using temporally delayed femtosecond laser double pulses, Opt. Express 21 (2013) 18501–18508. [13] R. Le Harzic, F. Stracke, H. Zimmermann, Formation mechanism of femtosecond laser-induced high spatial frequency ripples on semiconductors at low fluence and high repetition rate, J. Appl. Phys. 113 (2013) 183503. ˇ ˇ [14] R. Buividas, L. Rosa, R. Sliupas, T. Kudrius, G. Slekys, V. Datsyuk, S. Juodkazis, Mechanism of fine ripple formation on surfaces of (semi)transparent materials via a half-wavelength cavity feedback, Nanotechnology 22 (2011) 055304. [15] G. Miyaji, K. Miyazaki, Origin of periodicity in nanostructuring on thin film surfaces ablated with femtosecond laser pulses, Opt. Express 16 (2008) 16265–16271. [16] Y. Dong, P. Molian, Coulomb explosion-induced formation of highly oriented nanoparticles on thin films of 3C–SiC by the femtosecond pulsed laser, Appl. Phys. Lett. 84 (2004) 10–12. [17] E. Coyne, J.P. Magee, P. Mannion, G.M. O’Connor, T.J. Glynn, STEM (scanning transmission electron microscopy) analysis of femtosecond laser pulse induced damage to bulk silicon, Appl. Phys. A 81 (2005) 371–378. [18] X. Fu, L. Li, Y. Wu, N. Geng, Measuring, Machining of ripples on silicon surface with femtosecond pulse laser, Key Eng. Mater. 381–382 (2008) 101–104. [19] T. Tomita, R. Kumai, S. Matsuo, S. Hashimoto, M. Yamaguchi, Cross-sectional morphological profiles of ripples on Si, SiC, and HOPG, Appl. Phys. A 97 (2009) 271–276. [20] M. Schade, O. Varlamova, J. Reif, H. Blumtritt, W. Erfurth, H.S. Leipner, Highresolution investigations of ripple structures formed by femtosecond laser irradiation of silicon, Anal. Bioanal. Chem. 396 (2010) 1905–1911. [21] M. Straub, M. Schüle, M. Afshar, D. Feili, H. Seidel, K. König, Sub-15 femtosecond laser-induced nanostructures emerging on Si(1 0 0) surfaces immersed in water: analysis of structural phases, Appl. Phys. A 115 (2014) 221–228. [22] R. Le Harzic, D. Dörr, D. Sauer, M. Neumeier, M. Epple, H. Zimmermann, F. Stracke, Formation of periodic nanoripples on silicon and germanium induced by femtosecond laser pulses physics, Phys. Procedia 12 (2011) 29–36. [23] R. Le Harzic, D. Dörr, D. Sauer, M. Neumeier, M. Epple, H. Zimmermann, F. Stracke, Large-area, uniform, high-spatial-frequency ripples generated on silicon using a nanojoule-femtosecond laser at high repetition rate, Opt. Lett. 36 (2011) 229–231. [24] R. Le Harzic, D. Dörr, D. Sauer, F. Stracke, H. Zimmermann, Generation of high spatial frequency ripples on silicon under ultrashort laser pulses irradiation, Appl. Phys. Lett. 98 (2011) 211905. [25] J. Bonse, S.M. Wiggins, J. Solis, Dynamics of femtosecond laser-induced melting and amorphization of indium phosphide, J. Appl. Phys. 96 (2004) 2352.