OPTICS LETTERS / Vol. 38, No. 23 / December 1, 2013
Transformations induced in bulk amorphous silica by ultrafast laser direct writing Vitor Oliveira,1,2,* Sahendra P. Sharma,2,3 Pilar Herrero,4 and Rui Vilar2,3 1
Instituto Superior de Engenharia de Lisboa, Avenida Conselheiro Emídio Navarro no 1, 1959-007 Lisbon, Portugal 2 ICEMS—Instituto de Ciência e Engenharia de Materiais e Superfícies, Avenida Rovisco Pais no 1, 1049-001 Lisbon, Portugal 3 Instituto Superior Técnico, Avenida Rovisco Pais no 1, 1049-001 Lisbon, Portugal 4
Instituto de Ciencia de Materiales de Madrid (ICMM-CSIC), Cantoblanco, Madrid 28049, Spain *Corresponding author: [email protected]
Received September 10, 2013; revised October 24, 2013; accepted October 25, 2013; posted October 25, 2013 (Doc. ID 197489); published November 20, 2013
A transmission electron microscopy study of nanogratings formed in bulk amorphous silica by direct writing with an ultrafast pulsed laser with a radiation wavelength of 1030 nm and pulse duration of 560 fs is presented. The results achieved show that the nanogratings are composed of planar nanostructures with an average periodicity of 250 nm and typical thickness of about 30 nm, consisting of alternating layers of heavily damaged material and layers of material where a dense precipitation of nanocrystals occurred. The crystallization of silica to form these nanocrystals can be explained by the large pressures and temperatures reached in these regions as a result of nanoplasma formation and recombination. © 2013 Optical Society of America OCIS codes: (320.7110) Ultrafast nonlinear optics; (350.3390) Laser materials processing; (160.6030) Silica. http://dx.doi.org/10.1364/OL.38.004950
In recent years, femtosecond lasers have found application in the direct writing of internal structures in a wide variety of transparent materials [1–9]. This technique involves focusing an intense ultrafast laser pulse in a very small focal volume within the transparent material. Due to their extremely high peak intensity and short pulse duration, femtosecond lasers induce nonlinear electron excitation processes such as multiphoton and tunnel ionization processes, which promote electrons to the conduction band and eventually lead to the formation of a plasma . Since the interaction is highly nonlinear, the excitation is localized to a very small volume. This localization is further enhanced because heat transport to the material during the duration of the laser pulse is negligible. As the plasma recombines and its energy is dissipated, permanent structural changes can be induced in the material. Depending on the numerical aperture of the focusing lens and laser parameters, such as pulse duration, wavelength, energy, and repetition rate, three types of modifications can be achieved: an isotropic refractive index change [11,12], a birefringent refractive index change [8,13,14], and void formation [15–17]. Among these, the formation of birefringence domains embedded within the bulk of fused silica has attracted considerable attention, due to its potential applications in photonic devices [18,19]. The origin of this anisotropy lies in the formation of self-ordered structures with subwavelength periodicity oriented perpendicularly to the polarization of the incident laser beam radiation. These so-called nanogratings or nanoplanes have a period proportional to the incident laser radiation wavelength . In this Letter, we present a transmission electron microscopy (TEM) study of nanogratings formed in the bulk of amorphous silica by ultrafast laser direct 0146-9592/13/234950-04$15.00/0
writing. On the basis of these results, it is shown that nanogratings consist of a parallel arrangement of damaged material layers with a thickness between 20 and 30 nm, alternating with layers of material with embedded nanocrystals. The nanocrystals are likely formed due to the pressure wave and temperature increase arising from plasma formation and recombination within the focal volume. We inscribed nanogratings in fused silica samples using a mode-locked Yb:KYW chirped-pulse-regenerative amplification laser system (s-Pulse HP, Amplitude Systèmes) with a nominal pulse duration of 560 fs, a repetition rate of 1–100 kHz, and a radiation wavelength of 1030 nm. The experiments were carried out by translating the sample relative to the stationary laser beam using a computer-controlled XYZ specimen stage. The laser beam was focused at a certain depth within the material via a high-NA (0.66) 40× microscope objective [Fig. 1(a)]. The laser pulse energy was E 5 μJ, the writing speed s 0.1 mm∕s, the focal spot diameter D ≅ 10 μm, and the pulse repetition rate f 10 kHz. After the laser treatment, a thin disk 3 mm in diameter and about 70 μm thick containing buried laser tracks was cut with an ultrasonic cutter and polished on both sides by mechanical polishing with a diamond abrasive. After mechanical polishing the disk was further thinned from both sides by ion milling carried out with a Fischione 1010 system with a double Ar beam regulated at 5 keV until transparency to the electron beam was achieved [Fig. 1(b)]. The TEM observations were performed with a JEOL JEM-3000F fieldemission gun microscope using an acceleration voltage of 300 kV. A low magnification TEM micrograph of the laserprocessed region is depicted in Fig. 1(c). The micrograph reveals the presence of a self-organized parallel © 2013 Optical Society of America
December 1, 2013 / Vol. 38, No. 23 / OPTICS LETTERS
Fig. 1. (a) Schematic of the nanograting inscription process. The sample is translated in the y axis with a scanning speed of s 0.1 mm∕s. Nanogratings are engraved using a laser pulse energy of 5 μJ and a pulse repetition rate of 10 kHz. (b) Schematic of the preparation of TEM samples. A 70 μm thick disk containing nanogratings is made by mechanical polishing and further thinned by ion milling until TEM visualization is possible. (c) TEM micrograph of self-organized nanogratings in fused silica. The nanogratings are composed of parallel dark and light nanoplanes oriented perpendicularly to the laser radiation polarization (E).
nanograting with a period of about 250 nm, composed of alternate dark and bright regions. This period is roughly half of the laser radiation wavelength in the material, as predicted by the nanoplasmonics model proposed by Taylor et al. [18,20]. The nature of these nanogratings has been the subject of intense discussion in recent years. Taylor and coworkers  suggested that nanogratings are composed of parallel planar nanocracks with thicknesses between 5 and 30 nm, depending on the laser pulse duration. Subsequently, Canning et al.  reported that nanogratings consist of an arrangement of nanoplanes containing pores with typical diameters between 10 and 30 nm. Raman measurements suggested that these pores are filled with oxygen and produced as a result of the decomposition of SiO2 into SiO21−x xO2 . More recently, it was reported that the nanoplanes are isolated thin, disklike cavities with a thickness around 30 nm and transverse diameter of 200 to 300 nm . Our results do not confirm any of these suggestions. Figure 2 depicts a magnified bright field TEM image of the alternate dark/ bright planar regions forming the nanograting. It is clear that there are no nanocracks or nanopores as previously reported [18,21,22], but there are alternate regions of different thickness, generating electron absorption contrast, the main source of contrast taking into consideration that the solid is amorphous. This discrepancy may be due to the fact that the laser radiation intensity used in
Fig. 2. TEM image showing bright and dark nanoplanes composing a nanograting.
the present work (∼1012 W∕cm2 ) is lower than the threshold value for void formation in silica , but one must also consider that nanocracks and nanopores are usually visualized by scanning electron microscopy after selective etching of the damaged material with diluted hydrofluoric acid [18,23], or after mechanical cleaving [18,21], which may irrevocably alter the structure of the nanogratings and contribute to the formation of the nanocracks and nanopores. In contrast, we propose that the differences in the planar region thickness observed in Fig. 2 arise from a larger sputter yield in the laserdamaged regions as compared to the regions in pristine condition in the ion milling process used to prepare the TEM thin foils. Ion bombardment of solids with energetic particles is characterized by a sputtering yield Y , which, in the linear cascade regime and according to Sigmund’s sputtering theory , is given by Y 3F D ∕4NC 0 U, where F D is the energy deposited per unit depth at the sputtered surface, U is the surface binding energy, N is the target atom density, and C 0 is a constant. According to Kazansky and co-workers [4,8], nanogratings consist of nanoplanes of weakly modified material with a refraction index and density slightly higher than the unmodified material, alternating with regions of lower refraction index and density, depleted in oxygen and rich in dangling-bond-type defects such as E-centers and nonbridging oxygen hole centers . Since the sputtering yield depends on the target density and surface binding energy, different etching rates and thicknesses between alternating nanoplanes are to be expected during the ion beam milling process. Note that from this point of view, the dark regions correspond to weakly modified material and the bright regions to highly damaged material. Close examination of the electron micrograph of Fig. 2 reveals the presence of black dots, more conspicuous in the weakly damaged regions. The high-resolution TEM images of Fig. 3(a) show that these dark spots are actually nanoparticles with a mean diameter of about 4 nm embedded in the amorphous silica matrix. High-resolution TEM images of the same specimen away from the laserinscribed track [Fig. 3(b)] show that these nanoparticles did not exist prior to the laser treatment and are not a sideeffect of the ion milling process. The inset of Fig. 3(a) depicts a high-resolution TEM image of a single particle and
Fig. 3. (a) High-magnification TEM image of nanograting nanoplanes showing the presence of nanoparticles. The inset depicts a high-resolution TEM image of a single nanoparticle and shows a nanocrystal with a lattice interplanar spacing of about 0.21 nm. (b) High-magnification TEM image of pristine silica showing that nanoparticles do not exist prior to the laser treatment.
OPTICS LETTERS / Vol. 38, No. 23 / December 1, 2013
shows that nanoparticles are randomly oriented nanocrystals displaying a lattice interplanar spacing of about 0.21 nm. This interplanar spacing is similar within the limits of experimental error to interplanar distances in any of the silica polymorphs known to form in low-pressure conditions (quartz, tridymite, and cristobalite), so the possibility of its being one of these phases cannot be discarded. However, due to the large amount of energy deposited in the focal volume by the femtosecond laser pulses, very high temperatures and pressures are reached [1,15], so it is plausible that the nanocrystals may consist of a high-pressure silica polymorph such as stishovite or coesite. In fact, the laser-induced formation of stishovite has been observed in amorphous silica treated with nanosecond-pulse-duration lasers . The formation of these nanocrystals can be explained by the nanoplasmonics model [18,20]. According to this model, ultrashort laser pulses focused into silica produce ionization hot spots due to localized inhomogeneous nonlinear multiphoton absorption at defects or color centers. These hot spots produced in the first laser pulses evolve into planar nanoplasmas over several pulses. As the nanoplasmas recombine and transfer their energy to the surrounding material, regular planar regions of highly modified material are formed. Simultaneously, extremely high pressures and temperatures are reached locally, leading to energy transfer to the surrounding amorphous silica matrix and favoring its crystallization. A similar behavior has been observed after low-energy neutral beam irradiation of amorphous silica : α-cristobalite and α-quartz nanocrystals were observed in amorphous SiO2 irradiated with more than 1017 neutrals∕cm2 . This transformation of silica into a denser crystalline phase was explained by the very high local pressures and temperatures induced by the neutral atom bombardment. Interestingly, in the same study , the authors reported that crystallization does not occur when an ion beam, instead of neutral atoms, is used for sputtering, and they explain this difference by the fact that, contrarily to neutral beam bombardment, ion-beam bombardment causes a preferential sputtering of oxygen from amorphous SiO2 , leaving a Si-rich material that cannot transform into a crystalline phase of SiO2 . This argument can explain why a lower number of nanocrystals should form in the damaged material: since these nanoplanes are composed of oxygen-depleted material [4,8], crystallization of silica is more difficult. Finally, it is interesting to compare the present results to those of a recent study of Richter and coworkers , who employed small-angle x-ray scattering (SAXS) to characterize the structure of planar nanogratings produced in silica by femtosecond laser pulses. Independently of the energy and number of pulses used, the authors observed features with two different sizes in the laser-treated material, with characteristic dimensions of about 6 and 28 nm, respectively. By using a combination of focused ion beam and scanning electron microscope observations, they could identify the larger features as self-organized nanoplanes but were unable to identify the smaller features. Taking into consideration their size, we suggest that the small features observed by Richter and co-workers  using SAXS are the nanocrystals observed in the present work by TEM.
In conclusion, TEM analyses revealed that the nanogratings formed in bulk amorphous silica when this material is irradiated by high-intensity femtosecond laser pulses consist of alternating layers of heavily damaged material with a typical thickness between 20 and 30 nm and layers where nanocrystals about 4 nm in diameter precipitate. The crystallization of silica to form these nanocrystals can be explained by the large pressures and temperatures reached in these regions as a result of nanoplasma formation and recombination in adjacent areas. This work was carried out with financial support from Fundação para a Ciência e a Tecnologia, through project PTDC/FIS/102127/2008, and from Spanish Ministerio de Ciencia e Innovación, through Project FUNCOATCSD2008-00023-CONSOLIDER INGENIO. One of the authors, S. P. Sharma, gratefully acknowledges Instituto Superior Técnico and Fundação para a Ciência e a Tecnologia (FCT) for the postdoctoral fellowship provided through projects PTDC/FIS/102127/2008 and SFRH/ BPD/78871/2011. Technical support from M. J. Castelo from Servicio de Microscopia TEM (ICMM-CSIC) and the use of equipment from ICTS Centro Nacional de Microscopıa Electrónica (UCM) is gratefully acknowledged. References 1. E. N. Glezer and E. Mazur, Appl. Phys. Lett. 71, 882 (1997). 2. V. R. Bhardwaj, E. Simova, P. B. Corkum, D. M. Rayner, C. Hnatovsky, R. S. Taylor, B. Schreder, M. Kluge, and J. Zimmer, J. Appl. Phys. 97, 083102 (2005). 3. E. Bricchi and P. G. Kazansky, Appl. Phys. Lett. 88, 111119 (2006). 4. B. Bricchi, B. G. Klappauf, and P. G. Kazansky, Opt. Lett. 29, 119 (2004). 5. M. Budiman, E. M. Hsu, H. K. Haugen, and G. A. Botton, Appl. Phys. A 98, 849 (2010). 6. C. Hnatovsky, E. Simova, P. P. Rajeev, D. M. Rayner, P. B. Corkum, and R. S. Taylor, Opt. Lett. 32, 1459 (2007). 7. K. Itoh, W. Watanabe, S. Nolte, and C. B. Schaffer, MRS Bull. 31, 620 (2006). 8. Y. Shimotsuma, P. G. Kazansky, J. R. Qiu, and K. Hirao, Phys. Rev. Lett. 91, 247405 (2003). 9. K. M. Davis, K. Miura, N. Sugimoto, and K. Hirao, Opt. Lett. 21, 1729 (1996). 10. P. Balling and J. Schou, Rep. Prog. Phys. 76, 036502 (2013). 11. L. Sudrie, M. Franco, B. Prade, and A. Mysyrewicz, Opt. Commun. 171, 279 (1999). 12. B.Poumellec,M.Lancry,A.Chahid-Erraji,andP.G.Kazansky, Opt.Mater.Express1,766(2011). 13. V. R. Bhardwaj, E. Simova, P. P. Rajeev, C. Hnatovsky, R. S. Taylor, D. M. Rayner, and P. B. Corkum, Phys. Rev. Lett. 96, 057404 (2006). 14. S. Richter, M. Heinrich, S. Doring, A. Tunnermann, S. Nolte, and U. Peschel, J. Laser Appl. 24, 042008 (2012). 15. S. Juodkazis, H. Misawa, T. Hashimoto, E. G. Gamaly, and B. Luther-Davies, Appl. Phys. Lett. 88, 201909 (2006). 16. E. N. Glezer, M. Milosavljevic, L. Huang, R. J. Finlay, T. H. Her, J. P. Callan, and E. Mazur, Opt. Lett. 21, 2023 (1996). 17. S. Juodkazis, K. Nishimura, S. Tanaka, H. Misawa, E. G. Gamaly, B. Luther-Davies, L. Hallo, P. Nicolai, and V. T. Tikhonchuk, Phys. Rev. Lett. 96, 166101 (2006). 18. R. Taylor, C. Hnatovsky, and E. Simova, Laser Photon. Rev. 2, 26 (2008). 19. J. Canning, Laser Photon. Rev. 2, 275 (2008).
December 1, 2013 / Vol. 38, No. 23 / OPTICS LETTERS 20. P. P. Rajeev, M. Gertsvolf, C. Hnatovsky, E. Simova, R. S. Taylor, P. B. Corkum, D. M. Rayner, and V. R. Bhardwaj, J. Phys. B 40, S273 (2007). 21. J. Canning, M. Lancry, K. Cook, A. Weickman, F. Brisset, and B. Poumellec, Opt. Mater. Express 1, 998 (2011). 22. S. Richter, A. Plech, M. Steinert, M. Heinrich, S. Doring, F. Zimmermann, U. Peschel, E. B. Kley, A. Tunnermann, and S. Nolte, Laser Photon. Rev. 6, 787 (2012).
23. S. Richter, M. Heinrich, S. Doring, A. Tunnermann, and S. Nolte, Appl. Phys. A 104, 503 (2011). 24. P. Sigmund, Phys. Rev. 184, 383 (1969). 25. A. Salleo, S. T. Taylor, M. C. Martin, W. R. Panero, R. Jeanloz, T. Sands, and F. Y. Genin, Nat. Mater. 2, 796 (2003). 26. T. Mizutani, J. Non-Cryst. Solids 181, 123 (1995).