TPA-induced long-period gratings in a photonic crystal fiber: inscription and temperature sensing properties

Transcription

1 Fotiadi et al. Vol. 24, No. 7/July 2007/J. Opt. Soc. Am. B 1475 TPA-induced long-period gratings in a photonic crystal fiber: inscription and temperature sensing properties Andrei A. Fotiadi, 1 Gilberto Brambilla, 2 Thomas Ernst, 3 Stephen A. Slattery, 3 and David N. Nikogosyan 3, * 1 Faculté Polytechnique de Mons, Service d Electromagnétisme et de Télécommunications, 31, Boulevard Dolez, B-7000 Mons, Belgium, and Ioffe Physico-Technical Institute of Russian Academy of Sciences, Politekhnicheskaya 26, St. Petersburg, Russia 2 Optoelectronics Research Centre, University of Southampton, Southampton SO17 1BJ, UK 3 Department of Physics, University College Cork, Cork, Ireland *Corresponding author: d.nikogosyan@aston.ac.uk Received February 15, 2007; accepted April 2, 2007; posted April 19, 2007 (Doc. ID 80104); published June 15, 2007 We report on the photochemical recording of long-period fiber gratings (LPFGs) in a photonic crystal fiber made of pure fused silica. Such inscription is based on two-photon absorption (TPA) of high-intensity 300 GW/cm nm 220 fs pulses and brings about LPFGs of high strength and narrow peak width. The characteristic fluence value for the inscription is 1 order of magnitude less than that for a standard telecom fiber irradiated under similar conditions. The temperature sensitivity of TPA-induced LPFGs is 300 pm/ C and overcomes that of LPFGs inscribed by other nonphotochemical methods by 2 orders of magnitude Optical Society of America OCIS codes: , , , , , INTRODUCTION Long-period fiber grating (LPFG) fabrication in a photonic crystal fiber (PCF) has been known since 2002 [1]. For this, nonphotochemical inscription techniques are commonly used that modify the refractive index in the fiber cladding either by heating (using CO 2 laser light [1 3] or an electric arc discharge [4 9]) or by applying mechanical pressure [10,11]. One of the serious disadvantages of these methods is the collapsing of fragile PCF holes, especially in the case of PCFs with relatively large holes d 20 m. Another disadvantage originates from the irregularity of period deformation for any of the abovementioned approaches. As a result, it is rather difficult to change the LPFG strength by repetition of the inscription process for the same number of periods and the only way to adjust the strength of the inscribed grating is to increase the number of periods. The increase in the number of periods with irregular refractive index changes leads obviously to the decrease of the main and suppression of small grating resonances. The first photochemical grating recording in PCF was achieved in 1999 [12]. However, in this work a special PCF with a photosensitive Ge-doped core was used and the induced refractive index changes originated from the linear absorption of Ge oxygen-deficient centers with the maximum at 242 nm (common single-quantum inscription mechanism). An ordinary PCF is made of pure fused silica, which is fully transparent in the UV spectral region [13]. Nevertheless, it is possible to induce refractive index changes in this material using two-photon absorption (TPA) of femtosecond UV radiation. Though the TPA-induced longperiod and Bragg grating fabrication by high-intensity GW/cm nm pulses in hydrogenated standard telecom fiber (possessing a very low linear absorption at the irradiation wavelength) has been known since 2002 [14 20], this approach was implemented to PCF just recently [21,22]. In this paper, we thoroughly investigate the process of photochemical LPFG fabrication in a hydrogenated standard PCF, following high-intensity 300 GW/cm nm fs irradiation, and discuss the gratings characteristics and temperature sensing properties in comparison with those of similar gratings recorded by nonphotochemical inscription techniques. 2. EXPERIMENTAL SETUP In the experiments, we used an endlessly single-mode PCF, ESM from Blaze Photonics (now Crystal Fibre A/S), and for comparison the standard Corning telecom fiber SMF-28 (supplied by Elliot Scientific). The ESM hasa12 m core diameter surrounded by four rings of holes (hole diameter is 3.7 m, hole pitch is 8 m, number of holes is 54), and its outside diameter is 125 m. The pieces of the PCF used in the experiments were m long. To decrease the rate of hydrogen out-diffusion from the PCF, the latter was pigtailed to two 0.5 m long pieces of SMF-28 fiber by splicing before hydrogenation. The telecom SMF-28 fiber has a core diameter of 8.2 m and a cladding diameter of 125 m. Both fibers were sensitized under similar conditions (in a hydrogen atmosphere at 150 bar, at 70 C, for 2 weeks). For LPFG inscription, we applied femtosecond UV laser pulses [ =264 nm, p 200 J, p =220 fs (FWHM), /07/ /$ Optical Society of America

2 1476 J. Opt. Soc. Am. B/ Vol. 24, No. 7/ July 2007 Fotiadi et al. w 0 =0.3 cm (FWHM), f=27 Hz] [23], which were directed by a 46.9 cm CaF 2 spherical lens through a 250 m rectangular slit onto the fiber (with the acrylate coating removed). The displacement of the lens with respect to the fiber allowed us to adjust the incident UV irradiation intensity in the GW/cm 2 range. The slit was centered relative to the beam axis and placed close to the fiber at 1 mm distance. The movement of the fiber and the control of the light fluence was accomplished according to the method described earlier [19,20,24]. The irradiated fiber was fixed on a 50 mm (1 m resolution) computer-controlled translation stage PI 405.DG (Physik Instrumente) and was exposed point by point with a period of 500 or 600 m. In each case the number of recorded periods was equal to 20, so the length of the fabricated LPFG was equal to 1 or 1.2 cm, respectively. The evolution of the LPFG transmission loss peaks was monitored in situ with a white-light source AQ4305 and an optical spectrum analyzer AQ6317C (both from Yokogawa Europe BV). The UV exposure and fiber translation processes were computer controlled using LABVIEW and a PCI-GPIB card (both from National Instruments). Annealing of the recorded LPFGs (5 h at 80 C) and temperature studies were performed using an oven Carbolite MTF 12/25/ TWO-PHOTON LPFG INSCRIPTION MECHANISM IN ALL-FUSED- SILICA PHOTONIC CRYSTAL FIBER Two-photon absorption (simultaneous absorption of two light quanta) was proposed in 1931 [25], but experimentally realized only 30 years later [26] after the appearance of powerful light sources (lasers). TPA in fused silica in the UV region was investigated from the end of the 1970s [27,28]. The most reliable data on TPA coefficients ( = cm/w at 264 nm [23,29]) were obtained in the past decade in the transmission experiments utilizing very short femtosecond UV pulses (to exclude the absorption of photoproducts formed during long pulse UV excitation; see, for example, [23]). To realize TPA in the presence of linear (single-photon) absorption, one should use rather high values of irradiation intensity, I, in order to make the product, I, overcome significantly the linear absorption coefficient,, at the irradiation wavelength (approximation of optically thin layer; see, for example, [30]). In the case of vitreous quartz = cm 1 for different commercial sorts of fused silica at 264 nm [13], this means that for the obvious superiority of TPA against the linear absorption we need to fulfil I, i.e., we need to use irradiation intensities at 264 nm much higher than 1 5 GW/cm 2. The hydrogenation of pure fused silica increases the linear absorption coefficient and decreases the bandgap energy, which could lead to the increase of the TPA coefficient. Our direct measurements show that the corresponding increase of at 264 nm is by approximately four times (Fig. 1). Therefore, taking into account the possible 1 Fig. 1. (Color online) UV absorption spectrum of pure fused silica before and after hydrogenation. rise of a TPA coefficient, one can conclude that the hydrogenation of fused silica does not drastically change the condition (1). In the experiments described below, the irradiation intensity was GW/cm 2, so condition (1) was fulfilled. Similar estimates made for fused silica while irradiating at 193 nm show that for the realization of TPA we need to use intensities much higher than 0.1 GW/cm 2 (using for estimates the value given in [31] and the value from [13]). In [32], the nanosecond fiber Bragg grating (FBG) inscription in PCF at 193 nm was reported. Though the authors claim the two-photon inscription process, no proof of TPA occurrence is given. Besides, the used value of irradiation intensity was 0.01 GW/cm 2, which is obviously too small for the realization of TPA in this particular case. 4. RESULTS AND DISCUSSION A. Inscription Results Figures 2(a) 2(c) present the spectrum of transmission loss, the peak wavelength, and amplitude dependencies versus the incident fluence, respectively, for an LPFG with 500 m period and 1 cm length inscribed in a hydrogenated ESM fiber. Figures 3(a) 3(c) depict the corresponding dependencies for an LPFG with 600 m period and 1.2 cm length fabricated in the same fiber. Finally, Figs. 4(a) 4(c) represent the similar dependencies for an LPFG with 500 m period and 1 cm length prepared in a standard H 2 -loaded SMF-28 fiber. The first statement to be made is that the LPFG spectra [Figs. 2(a) and 3(a)] created in the PCF by highintensity 264 nm fs light pulses show very strong resonance peaks with up to 20 db grating strength [Figs. 2(c) and 3(c)]. As the bandgap energy value of pure fused silica is 9.3 ev [33], this bandgap could only be bridged by two 264 nm light quanta with an energy of 4.7 ev, which is in line with the recent experiment on FBG inscription in a PCF using high-intensity 267 nm pulses [21]. The loss spectra, shown in Figs. 2(a) and 3(a), demonstrate an excellent LPFG quality (regular form of the peaks and absence of out-band losses) in spite of the two splices decreasing the initial signal level by 15 db. All transmission loss peaks are shifting monotonically toward the longer wavelengths with the increase in incident

3 Fotiadi et al. Vol. 24, No. 7/July 2007/J. Opt. Soc. Am. B 1477 fluence [Figs. 2(b) and 3(b)], though the rate of this shift is different for different peaks [note the two-fold difference for peaks G and H in Fig. 3(b)]. From Fig. 2(c) it follows that at the incidence fluence of 10 J/cm 2 the peak A reaches its maximum (coupling factor is equal to /2 [34]). Such an effect is common for LPFG inscription with a high value of excitation energy (i.e., using 157 nm light [35]) and has never before been demonstrated in a PCF. But the most striking feature of the presented data is the extremely low value of fluence necessary for the recording of an LPFG in a hydrogenated PCF [Figs. 2(c) and 3(c)]. Fig. 3. (Color online) Inscription of an LPFG with 600 m period in hydrogenated ESM fiber with high-intensity 264 nm fs pulses: (a) transmission loss spectrum recorded with the irradiation intensity of 316 GW/cm 2 and total incident fluence of 8.4 J/cm 2 ; (b) shift of wavelengths corresponding to the transmission loss peaks F, G, H, and I (designated in the spectrum above) versus the total incident fluence; (c) transmission loss amplitudes for the peaks F, G, H, and I versus the total incident fluence. Fig. 2. (Color online) Inscription of an LPFG with 500 m period in hydrogenated ESM fiber with high-intensity 264 nm fs pulses: (a) transmission loss spectrum recorded with the irradiation intensity of 298 GW/cm 2 and total incident fluence of 10.1 J/cm 2 ; (b) shift of wavelengths corresponding to the transmission loss peaks A, B, C, D, and E (designated in the spectrum above) versus the total incident fluence; (c) transmission loss amplitudes for the peaks A, B, C, D, and E versus the total incident fluence. Indeed, to fabricate a FBG in a hydrogenated PCF with a reflection peak of 7 db [21], one needs a huge fluence value of 70 kj/cm 2. In our experiments with FBG inscription in H 2 -loaded SMF-28, a fluence value of 1.2 kj/cm 2 was sufficient to record a reflection peak of more than 45 db [17,18], meaning that Bragg grating fabrication in a PCF, in comparison with standard telecom fiber, requires 2 orders more fluence. On the contrary, for LPFG fabrication in PCF, only 10 J/cm 2 fluence is enough to record a 20 db peak [Figs. 2(a) and 2(c)] in comparison with J/cm 2 fluence value necessary for the in-

4 1478 J. Opt. Soc. Am. B/ Vol. 24, No. 7/ July 2007 Fotiadi et al. the presence of the holes decreases the contrast of the interference picture created by the phase mask, but this circumstance is not important in the case of LPFG fabrication where the grating period is much larger than the wavelength of the inscribing UV radiation. Concerning LPFG recording, the much smaller fluence value necessary for inscription in a PCF in comparison with a standard telecom fiber points to a much stronger coupling resonance between the cladding and core modes in the LPFG-PCF case. Such enhancement of the mode coupling could be explained by the peculiarities of the PCF drawing process. Indeed, the PCF is pulled out from a stack of Fig. 4. (Color online) Inscription of an LPFG with 500 m period in H 2 -loaded SMF-28 fiber with high-intensity 264 nm fs pulses: (a) transmission loss spectrum recorded with the irradiation intensity of 311 GW/cm 2 and total incident fluence of 157 J/cm 2, the peak K at this fluence value moved beyond the 1700 nm; (b) shift of wavelengths corresponding to the transmission loss peaks K, L, M, and N versus the total incident fluence; (c) transmission loss amplitudes for the peaks K, L, M, and N versus the total incident fluence. scription of the 24 db K (or L) peak in SMF-28 [Figs. 4(a) and 4(c)]. Although both these effects, the difficulty in writing a FBG in a PCF and the easiness of recording an LPFG, could be related to a quick redistribution of hydrogen from the core to the cladding before the inscription, which reduces the photosensitivity of core, there are some other points that are specific for either FBG or LPFG fabrication in a PCF. For example, during the FBG inscription Fig. 5. (Color online) UV absorption spectrum of an LPFG with 500 m period fabricated in hydrogenated ESM fiber by 264 nm fs pulses with the irradiation intensity of 293 GW/cm 2 and total incident fluence of 13.7 J/cm 2 : (a) before and (b) after annealing; (c) shift of the resonance wavelength for different transmission loss peaks versus temperature in annealed LPFG, for each dependence the temperature sensitivity values (in pm/ C) are indicated.

5 Fotiadi et al. Vol. 24, No. 7/July 2007/J. Opt. Soc. Am. B 1479 Fig. 6. (Color online) UV absorption spectrum of an LPFG with 600 m period fabricated in hydrogenated ESM fiber by 264 nm fs pulses with the irradiation intensity of 316 GW/cm 2 and total incident fluence of 10.3 J/cm 2 : (a) before and (b) after annealing; (c) shift of the resonance wavelength for different transmission loss peaks versus temperature in annealed LPFG, for each dependence the temperature sensitivity values (in pm/ C) are indicated. well-ordered fused-silica tubes. During the drawing the PCF acquires additional mechanical stress associated with the mutual interrubbing of the tubes. It is well known that the presence of mechanical stress could enhance fiber photosensitivity [36]. However, in this particular case, the enhancement could be even more pronounced as the geometry of the stress distribution reproduces the symmetry of the inner cladding region of the PCF (containing holes). Besides, the stress distribution is not uniform but concentrated in the tube interrubbing region. Hence the photosensitivity profile follows the hexagonal cladding mode symmetry, which could drastically enhance the overlap integral in comparison with uniform stress distribution. The LPFG spectra of losses recorded in PCF are more complicated than that in SMF-28 [cf. Figs. 2(a), 3(a), and 4(a)]. The comparison between the peaks for the same grating period in PCF and SMF-28 shows that the peaks A, B, C, and D in a PCF correspond well to peaks K, L, M, and N in SMF-28, though the latter cluster occupies a wider wave region. The peak E is specific to our PCF, which is in line with the highest value of the slope of peak wavelength dependence for peak E in comparison with peaks A, B, C, and D [Fig. 2(b)]. B. Annealing and Temperature Sensitivity Results To remove the remaining hydrogen, our LPFGs in PCF were annealed. The annealing was accompanied by the wavelength shift of all LPFG peaks toward blue spectral range [Figs. 5(a), 5(b), 6(a), and 6(b)]. After that the temperature sensitivity of LPFG peaks was investigated. From Figs. 5(c) and 6(c) it is clear that all photochemically inscribed peaks in the PCF LPFG spectrum are very sensitive to temperature. While for SMF-28, the observed temperature sensitivity value was 38 pm/ C (for peak M, Fig. 7), which is consistent with the data of other photochemical experiments on LPFG inscription, 54 pm/ C for 500 m period and 1520 nm resonance peak [37], the data for 500 m period LPFG in ESM reveals a much stronger and negative temperature sensitivity [ 290 pm/ C for peak A, 250 pm/ C for peaks B and C, 330 pm/ C for peak D and 490 pm/ C for peak E; see Fig. 5(c)]. Similar data were obtained for 600 m period LPFG in PCF [Fig. 6(c)]. C. Comparison between the Characteristics of LPFGs Prepared by Different Methods Table 1 provides the comparison between LPFGs inscribed in the same ESM fiber by electric arc, mechanical pressure, and femtosecond laser. From Table 1, it follows that the use of high-intensity UV irradiation allows the fabrication of long-period gratings with the highest strength and narrowest bandwidth (per unit of grat- Fig. 7. Shift of the resonance wavelength versus temperature for an LPFG with 500 m period in H 2 -loaded SMF-28 fiber fabricated in hydrogenated SMF-28 fiber by 264 nm fs pulses with the irradiation intensity of 311 GW/cm 2 and total incident fluence of 255 J/cm 2. The initial transmission loss peak at nm was moved to nm during annealing. The temperature sensitivity value (in pm/ C) is indicated.

6 1480 J. Opt. Soc. Am. B/ Vol. 24, No. 7/ July 2007 Fotiadi et al. Table 1. Characteristics of LPFGs Inscribed in Endlessly Single-Mode ESM Fiber by Different Methods Method Period m Length L (cm) Wavelength (nm) Strength (db) FWHM (nm) d /dt pm/ C Reference Electric arc [7] Electric arc [8] Electric arc [8] Mechanical [11] pressure Mechanical [11] pressure UV femtosecond laser This work irradiation UV femtosecond laser irradiation This work ing length). The temperature sensitivity of LPFGs prepared by the two-photon approach is 300 pm/ C and overcomes that of LPFGs inscribed by other nonphotochemical methods by 2 orders of magnitude. 5. CONCLUSION Concluding, the two-photon high-intensity UV femtosecond approach allows us to produce efficiently LPFGs in PCF with narrow bandwidth using very low fluences in comparison with standard telecom fiber. Such an approach possesses also the significant thermal sensing properties in comparison with LPFGs inscribed in PCF by other nonphotochemical methods. ACKNOWLEDGMENTS We are indebted to Philip St. J. Russell for providing us with the PCF sample, his continuous support, and valuable comments; and Jonathan Knight and Jovana Petrovic for useful discussions. A. Fotiadi is supported by the Interuniversity Attraction Pole program (IAP VI 10) of the Belgian Science Policy. T. Ernst, S. A. Slattery, and D. N. Nikogosyan are grateful to Science Foundation Ireland for financial support (grant 04/IN3/I608). REFERENCES 1. G. Kakarantzas, T. A. Birks, and P. St. J. Russell, Structural long-period gratings in photonic crystal fibers, Opt. Lett. 27, (2002). 2. Y. Zhu, P. Shum, J.-H. Chong, M. K. Rao, and C. Lu, Deep-notch, ultracompact long-period grating in a largemode-area photonic crystal fiber, Opt. Lett. 28, (2003). 3. Y. Zhu, P. Shum, H.-W. Bay, M. Yan, X. Yu, J. Hu, J. Hao, and C. Lu, Strain-insensitive and high-temperature longperiod gratings inscribed in photonic crystal fiber, Opt. Lett. 30, (2005). 4. G. Humbert, A. Malki, S. Février, P. Roy, and D. Pagnoux, Electric arc-induced long-period gratings in Ge-free airsilica microstructure fibres, Electron. Lett. 39, (2003). 5. G. Humbert, A. Malki, S. Février, P. Roy, and D. Pagnoux, Characterizations at high temperatures of long-period gratings written in germanium-free air-silica microstructure fiber, Opt. Lett. 29, (2004). 6. K. Morishita and Y. Miyake, Fabrication and resonance wavelengths of long-period gratings written in a pure-silica photonic crystal fiber by the glass structure change, J. Lightwave Technol. 22, (2004). 7. H. Dobb, K. Kalli, and D. Webb, Temperature-insensitive long period grating sensors in photonic crystal fibre, Electron. Lett. 40, (2004). 8. J. S. Petrovic, V. Mezentsev, H. Dobb, D. J. Webb, K. Kalli, and I. Bennion, Multiple period resonances in long period gratings in photonic crystal fibres, Opt. Quantum Electron. 38, (2006). 9. H. Dobb, K. Kalli, and D. J. Webb, Measured sensitivity of arc-induced long-period gratings sensors in photonic crystal fibre, Opt. Commun. 260, (2006). 10. J. H. Lim, K. S. Lee, J. C. Kim, and B. H. Lee, Tunable fiber gratings fabricated in photonic crystal fiber by use of mechanical pressure, Opt. Lett. 29, (2004). 11. K. N. Park, T. Erdogan, and K. S. Lee, Cladding mode coupling in long-period gratings formed in photonic crystal fibers, Opt. Commun. 26, (2006). 12. B. J. Eggleton, P. S. Westbrook, R. S. Windeler, S. Spälter, and T. A. Strasser, Grating resonances in air-silica microstructured optical fibers, Opt. Lett. 24, (1999). 13. D. N. Nikogosyan, Properties of Optical and Laser-Related Materials. A Handbook (Wiley, 1997). 14. A. Dragomir, D. N. Nikogosyan, A. A. Ruth, K. A. Zagorulko, and P. G. Kryukov, Long-period fibre grating formation with 264 nm femtosecond radiation, Electron. Lett. 38, (2002). 15. A. Dragomir, D. N. Nikogosyan, K. Zagorulko, and P. G. Kryukov, Inscription of long-period fibre gratings by femtosecond UV radiation, Proc. SPIE 4876, (2003). 16. A. Dragomir, D. N. Nikogosyan, K. A. Zagorulko, P. G. Kryukov, and E. M. Dianov, Inscription of fiber Bragg gratings by ultraviolet femtosecond radiation, Opt. Lett. 28, (2003). 17. S. A. Slattery, D. N. Nikogosyan, and G. Brambilla, Fiber Bragg grating inscription by high intensity femtosecond UV laserlight: comparison with other existing methods of fabrication, J. Opt. Soc. Am. B 22, (2005). 18. S. A. Slattery, D. N. Nikogosyan, and G. Brambilla, Fiber Bragg grating inscription by high intensity femtosecond UV laserlight: comparison with other existing methods of fabrication: erratum, J. Opt. Soc. Am. B 22, (2005). 19. A. I. Kalachev, V. Pureur, and D. N. Nikogosyan, Investigation of long-period fiber gratings induced by highintensity femtosecond UV laser pulses, Opt. Commun. 246, (2005). 20. A. I. Kalachev, V. Pureur, and D. N. Nikogosyan, Investigation of long-period fiber gratings induced by highintensity femtosecond UV laser pulses: erratum, Opt. Commun. 251, (2005).

7 Fotiadi et al. Vol. 24, No. 7/July 2007/J. Opt. Soc. Am. B L. B. Fu, G. D. Marshall, G. A. Bolger, P. Steinvurzel, E. C. Mägi, M. J. Withford, and B. J. Eggleton, Femtosecond laser writing Bragg gratings in pure silica photonic crystal fibres, Electron. Lett. 41, (2005). 22. G. Brambilla, A. A. Fotiadi, S. A. Slattery, and D. N. Nikogosyan, Two-photon photochemical long-period grating fabrication in pure-fused-silica photonic crystal fiber, Opt. Lett. 31, (2006). 23. A. Dragomir, J. G. McInerney, and D. N. Nikogosyan, Femtosecond measurements of two-photon absorption coefficients at =264 nm in glasses crystals and liquids, Appl. Opt. 41, (2002). 24. A. I. Kalachev, D. N. Nikogosyan, and G. Brambilla, Longperiod fiber grating fabrication by high-intensity femtosecond pulses at 211 nm, J. Lightwave Technol. 23, (2005). 25. M. Göppert-Mayer, Über Elementarakte mit zwei Quantensprüngen, Ann. Phys. 401, (1931). 26. W. Kaiser and C. G. B. Garrett, Two-photon excitation in CaF 2 :Eu 2+, Phys. Rev. Lett. 7, (1961). 27. P. Liu, W. L. Smith, H. Lotem, J. H. Bechtel, N. Bloembergen, and R. S. Adhav, Absolute two-photon absorption coefficients at 355 and 266 nm, Phys. Rev. B 17, (1978). 28. P. Liu, R. Yen, and N. Bloembergen, Two-photon absorption coefficients in UV window and coating materials, Appl. Opt. 18, (1979). 29. A. Dragomir, J. G. McInerney, D. N. Nikogosyan, and P. G. Kazansky, Two-photon absorption properties of commercial fusedsilica and germanosilicate glass at 264 nm, Appl. Phys. Lett. 80, (2002). 30. D. N. Nikogosyan, A. A. Oraevsky, and V. I. Rupasov, Two-photon ionization and dissociation of liquid water by powerful laser UV irradiation, Chem. Phys. 77, (1983). 31. O. Kittelmann and J. Ringling, Intensity-dependent transmission properties of window materials at 193-nm irradiation, Opt. Lett. 19, (1994). 32. N. Groothoff, J. Canning, E. Buckley, K. Lyttikainen, and J. Zagari, Bragg gratings in air-silica structured fibers, Opt. Lett. 28, (2003). 33. Z. A. Weinberg, G. W. Rubloff, and E. Bassous, Transmission photoconductivity and the experimental band gap of thermally grown SiO 2 films, Phys. Rev. B 19, (1979). 34. M. N. Ng and K. S. Chiang, Thermal effects on the transmission spectra of long-period fiber gratings, Opt. Commun. 208, (2002). 35. K. P. Chen, P. R. Herman, R. Tam, and J. Zhang, Rapid long-period grating formation in hydrogen-loaded fibre with 157 nm, Electron. Lett. 36, (2000). 36. E. Salik, D. S. Starodubov, and J. Feinberg, Increase of photosensitivity in Ge-doped fibers under strain, Opt. Lett. 25, (2000). 37. L. Qin, Z. X. Wei, Q. Y. Wang, H. P. Li, W. Zheng, Y. S. Zhang, and D. S. Gao, Compact temperaturecompensating package for long-period fiber gratings, Opt. Mater. 14, (2000).