Buried waveguides in glass are manufactured by irradiation with femtosecond laser double pulses. The refractive index change Δn is determined by measuring the numerical aperture NA of the waveguides and by through light microscopy. The value of Δn shows a significant dependency on the time delay Δt of the fs-laser double pulses. A Δn of up to 2×10-3 in fused silica is reached at a Δt between 400 and 800 ps. Based on the results of the double pulse experiments the initial effects of the refractive index change are discussed, taking into account thermal effects and the formation of self trapped excitons (STE) and transient color centers and their interaction with the next laser pulse.
©2007 Optical Society of America
Refractive index modifications in the volume of glasses and crystals provide the possibility for integrated optical devices like phase plates, 3D-optics for beam forming of diode laser radiation, beam-splitter and combiner or active devices like amplifiers and waveguide lasers. Waveguiding structures can be manufactured by ion-exchange or diffusion into a transparent substrate, laser irradiation of photorefractive materials or lithographic methods. In recent years, the technique of direct femtosecond laser writing of optical waveguides has been demonstrated by several groups [1,2,3]. Nonelastic thermomechanical stress and/or the formation of color centers due to multiphoton absorption in the focal volume of tightly focused femtosecond laser radiation leads to a refractive index increase. Refractive index change by tightly focused fs-laser radiation is observed using oscillators with repetition rates of several ten MHz and pulse energies of some nJ as well as with amplified systems with kHz repetition rates and µJ pulse energies [4,5,6]. The possibility of waveguide writing in these two regimes can be explained by thermally induced processes due to heat accumulation . On the other hand, Guizard et al. showed the formation of a transient lattice distortion (self trapped exciton STE) within 250 fs, which relaxes to a transient color center. Lonzaga et al.[8,9] determined the lifetime of these color centers to be less than 10 ns. Pump probe absorption measurements in irradiated SiO2 demonstrate the existence of a competition between two different relaxation channels [10,11]. One of these two channels ends up in the formation of permanent color centers, the other one leads to a total relaxation of the lattice. Guizard et al. determined, that the permanent defect at the end of the process is a neutral oxygen vacancy, rather than an ionized one (E′1) . The formation of neutral oxygen vacancies needs another free electron, which can be delivered by a second fs-laser pulse during the lifetime of the transient color center. Nagata et al. show an effect on the damping and the resulting numerical aperture of buried waveguides written using fs-laser double pulses with a delay time between 0.5 and 200 ps . In this work, the time resolved free electron density is controlled using fs-laser double pulses in another time interval (Δt=0…10 ns). By varying the delay time and measuring the resulting refractive index modification, the lifetime of the STE and transient color centers is determined. Based on these results the contribution of the two observed processes, thermal melting and quenching in a denser state and on the other hand the formation of neutral oxygen vacancies by electronic processes to the refractive index change is discussed.
2. Experimental setup
2.1 Laser and sample processing setup
Irradiation is carried out by a Ti:Sapphire CPA laser system (Thales Concerto) with a central wavelength λ=800 nm and a pulse duration of τp=100 fs, running at a repetition rate of f=1 kHz with a maximum output power of Pav=1.5 W. The beam is transferred into a personal computer controlled positioning stage with three perpendicular translation axes and focused using different microscope objectives with numerical apertures NA of 0.4 and 0.7. The samples are processed by scanning the laser focus at a depth of 150µm through the volume of the glass.
Fs-double pulses have been formed using a Michelson interferometer consisting of a beam splitter, a fixed mirror and a moveable one on a delay line of 1.5 m length and a resolution of 1µm. Therefore, the delay time between the pulses is maximal Δt=10 ns with an accuracy of 33 fs. The splitting ratio for the double pulses is 50:50 in any case. The temporal adjustment of the two beamlines is made by overlaying both beams in a non-linear crystal and observing the signal intensity at λ=400 nm. The intensity increase at Δt=0 can be easily observed by the naked eye. The control of the spatial adjustment is made for every delay time by observing the beam positions on the beamsplitter as well as in the focus.
2.2 Waveguide characterization
The resulting waveguides have been lapped and polished from both sides using a Logitech PM-5 polishing machine. They are analyzed using Nomarsky microscopy to determine the size of the modified region. Using three through light images, focused in different sample depths a commercial software (QPM) calculates the absolute value of the refractive index n by solving the Zernicke phase contrast equations. The second method to determine the refractive index change in the glass is to derive the refractive index change from the measurement of the numerical aperture NA of the waveguides. Therefore, the Δn is calculated from the NA using equation 1.
For the determination of the NA HeNe-laser (λ=633 nm) radiation is coupled into the waveguide and the farfield of the guided light is projected on a screen. The spot size of the guided mode is determined from an image of the intensity distribution on a screen in the distance d by fitting a Gaussian to the intensity values of a selected part of a digital image. The full width of the half maximum of this gaussian fit was used as the beam radius r on the screen.
3. Results and Discussion
Waveguides written with fs-laser double pulses at a repetition rate f=1 kHz, pulse energies EP=0.65 µJ (2×0.325 µJ), a scanning velocity v=15 µm/s and a used writing NA of 0.4 show a strong dependency of the numerical aperture NA on the delay time Δt between the two pulses. For Δt>400 ps the NA increases by a factor of 1.5 comparing to the one written with single pulses (Δt=0).
For different samples, with low variations in writing speed or pulse energy this maximum in the NA between Δt=400 and 800 ps could be observed (Fig.1). As shown in Fig. 4, the measurement of the refractive index profile by microscopy and QPM software shows the same results: the refractive index change in fused silica is increased by a factor of more than 1.5 using fs-laser double pulses with delay times of Δt between 400 and 800 ps. A maximum Δn of 2,0×10-3 in fused silica was observed for EP=0.65 µJ, v=16 µm/s and Δt=500 ps (Fig. 2).
The maximum of the refractive index change for pulse delays of around Δt=600 ps cannot be explained by thermal effects. The temperature decreases after the first pulse exponentially. An increased absorption of the second pulse due to free electrons is probable. But the absorption of the second pulse must also decrease with the temperature and cannot show a maximum for a special delay time Δt. Therefore, the observed maximum is considered to be due to changes in the electronic structure of the material like the formation of transient color centers. As shown in , transient color centers are a relaxation product of self trapped excitons (STE) and have themselves two different relaxation channels: one to a total relaxation of the lattice and one to a non-bridging oxygen-vacancy (NBOHC), which means a permanent density and refractive index change. The effect of the peak in the refractive index change at Δt between 400 and 800 ps can be explained in two different ways: a higher absorption of the second pulse because of the color center existing at this time. The higher absorbed energy leads to a larger molten volume and increased permanent thermally induced stress in the material. The second possible explanation is the formation of a non-bridging oxygen-vacancy as a relaxation product of the color center itself with subsequent density change, induced by free electrons generated during the second pulse (Fig.3).
The experimental results indicate lifetimes of the STE of 400 ps and subsequent formation of transient color centers. After another 400 ps a total lattice relaxation or, if induced by free electrons, formation of a non-bridging oxygen-vacancy occurs. However, by using fs-laser double pulses with a Δt between 400 and 800 ps, a refractive index change in fused silica of Δn=2*10-3 is reached, increased by a factor of more than 1.5 compared to single pulse irradiation.
4. Summary and outlook
Refractive index modification in fused silica under irradiation with fs-laser double pulses was observed and shows a strong dependency on the delay time Δt. The observation of non-thermal effects, using fs-laser double pulses is explained by the existence of transient color centers during the time 400 and 800 ps after the first pulse. The NA of the produced waveguides and the Δn in in fused silica are increased by a factor of more than 1.5 by irradiation with fs-laser double pulses compared to single pulse irradiation. The effect of the peak in the refractive index change versus the delay time is explained by the formation of a non-bridging oxygen-vacancy with subsequent density change, induced by free electrons generated during the second pulse.
Further work will include variations of the energy ratio of the double pulses and the determination of non-bridging oxygen vacancies by spectroscopic measurements. Also a determination and quantification of this effect for various glasses, thermally sensitive and electronically sensitive ones, will be done.
We would like to thank M. Kachel, A. Horn, I. Mingareev and A. Glavic from the RWTH Aachen for a lot of work in the laboratory and helpful discussions.
References and links
1. K. Miura, J. Qiu, H. Inouye, T. Mitsuyu, and K. Hirao, “photowritten waveguides in various glasses with ultrashort pulse laser,” Appl. Phys. Lett. 71, 3329–3331 (1997). [CrossRef]
3. A. Saliminia, N.T. Nguyen, M.-C. Nadeau, S. Petits, L. Chin, and R. Vallee, “Writing optical waveguides in fused silica using 1 kHz fs infrared pulses,” J. Appl. Phys. 93, 3724–3728 (2003). [CrossRef]
4. C. Cerullo, R. Osellame, S. Tacche, M. Marangoni, D. Polli, R. Ramponi, P. Laporta, and S. DeSilvestri, “Femtosecond micromachining of symmetric waveguides at 1.5µm by astigmatic beam focusing,” Opt. Lett. 27, 1938–1940 (2002). [CrossRef]
5. A. Zoubir, M. Richardson, C. Rivero, C. Lopez, and K. Richardson “Direct fabrication of waveguiding structures with a fs-laser”
6. L. Shah, F. Yoshino, A. Arai, S. Eaton, H. Zhang, S. Ho, and P. Herman, “MHz-rate ultrafast fiber laser for optical waveguides in silica glasses,” Proceedings of SPIE, Applications of ultrafast and free electron lasers 5714, (2005).
7. S. Eaton, H. Zhang, P. Herman, F. Yoshino, L. Shah, A. Arai, and S. Ho, “Heat accumulation effects in femtosecond written waveguides with variable repetition rate,” Opt. Express 13, 4708 (2005). [CrossRef] [PubMed]
8. S. Guizard, P. Martin, G. Petite, and P. Meynadier, “Time resolved study of laser-induced colour centres in SiO2,” J. Phys.: Condens Matter 8, 1281–1290 (1996). [CrossRef]
9. J.B. Lonzaga, S.M. Avanesyan, S.C. Langford, and J.T. Dickinson, “Color center formation in soda-lime glass with femtosecond laser pulses,” J. Appl. Phys. 94, 7 (2003). [CrossRef]
10. R.A. Devine, C. Fiori, and J. Robertson, “Defects in glasses,” Materials research society , p.177 (1986).
11. M.G. Jani, R.B. Bossoli, and L.E. Halliburton, “Further characterization of the E′1 center in crystalline SiO2,” Phys. Rev. B 27, 2285 (1983). [CrossRef]
12. Nagata et al. “Optical waveguide fabrication with double pulse femtosecond lasers,” Appl. Phys. Lett. 86, 251103 (2005). [CrossRef]