- Email: [email protected]
Random period arc-induced long-period fiber gratings
Optics & Laser Technology 44 (2012) 1176–1179
Contents lists available at SciVerse ScienceDirect
Optics & Laser Technology journal homepage: www.elsevier.com/locate/optlastec
Research Note
Random period arc-induced long-period fiber gratings A. Martinez-Rios a,n, I. Torres-Gomez a, D. Monzon-Hernandez a, G. Salceda-Delgado a, V.M. Duran-Ramirez b, G. Anzueto-Sanchez c ´ ptica, Lomas del Bosque 115, Colonia Lomas del Campestre, 37150 Leo ´n, Guanajuato, Me´xico Centro de Investigaciones en O Centro Universitario de los Lagos, Universidad de Guadalajara, Lagos de Moreno Jal., Me´xico c ´n en Ingenierı´a y Ciencias Aplicadas CIICAp, Universidad Auto ´noma del Estado de Morelos, Cuernavaca, Morelos 62210, Me´xico Centro de Investigacio a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 September 2011 Received in revised form 22 November 2011 Accepted 23 November 2011 Available online 7 December 2011
We report the fabrication of arc-induced long-period fiber gratings with strong random variations in the period. Long-period fiber gratings with standard deviations in the period from 8.50 to 36.98 mm were fabricated. The spectral position of the resonant bands is determined by the average period value, being similar to that observed in a long-period fiber grating with a fixed period equal to the average period of the random grating. Moreover the notch bands keep the shape characteristics like wideband and depth compared with a long-period grating with a constant period. In addition, their sensitivity to external parameters such as ambient refractive index is not too different with that of fixed period longperiod gratings. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Long-period fiber gratings Random period Refractive index
1. Introduction Long-period fiber gratings (LPFGs) are devices where the coupling between co-propagating core and cladding modes results in the formation of well defined loss bands at wavelengths determined by the period of the grating and the fiber characteristics [1]. LPFGs can be fabricated in optical fibers by the application of a periodic perturbation of the refractive index, which can be induced by UV [2] or CO2 [3] laser irradiation, by the application of mechanical stress [4], by the application of an electric arc [5], or even using an UV lamp [6]. An important factor to note is that in the case of LPFGs fabricated by laser irradiation the affected zone where the refractive index is perturbed is of a few microns in the case of UV irradiated fibers and maybe less than 50 mm for CO2 laser irradiation, while for arc-induced LPFGs the affected zone is higher than 300 mm. Thus, in the first two cases even small variations in the period may significantly affect their transmission characteristics. In Ref. [7], the effect of slight random variations in amplitude and period during the grating fabrication were simulated, finding that random period variations degrade the spectral characteristics of the gratings. On the other hand, in Ref. [8] a nonuniform LPFG was numerically designed to provide a linear sensitivity curve to external refractive index. In this case, the grating period has a cosine-like period variation, with 3 mm variation range (i.e. the difference between the maximum and minimum period). In contrast, the perturbation
n
Corresponding author. Tel.: þ52 477 441 42 00x197; fax: þ 52 477 441 42 00. E-mail address: [email protected] (A. Martinez-Rios).
0030-3992/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.optlastec.2011.11.037
zone in arc-induced LPFGs has a gradient along the longitudinal and transverse directions that may vary from point to point and affects their reproducibility under the same assumed fabrication conditions, particularly in poorly aligned setups [9]. In the size of the affected zone and its gradient along the longitudinal and transverse directions we may see arc-induced LPFGs having a random chirp, which make its reproducibility difficult unless special measures are taken during the fabrication [10]. In this work, we present the formation of LPFGs with random variation ranges up to 125 mm (standard deviation of 36.98 mm) in dispersion-shifted fiber (DSF) fabricated by the electric arc method. We found that the position of the resonance peaks is approximately determined by the mean period, i.e., the number and position of the resonances correspond to that obtained with a fixed period grating with period close to the mean value. Moreover, the spectral characteristics does not degrade, for example, the FWHM spectral width is not different from that obtained with fixed period LPFGs and in some cases is slightly smaller. Furthermore, we have measured the response to changes in external refractive index of the random period LPFGs without finding any significant difference in the response curve with the dispersion in the random period.
2. Experimental setup Fig. 1 shows a sketch of the experimental setup used for the fabrication of the arc-induced LPFGs. The fiber holders, V-groves, and electrodes are contained within a commercial fusion splicer S-176 from Fitel. One of the fiber holders (the right holder in
A. Martinez-Rios et al. / Optics & Laser Technology 44 (2012) 1176–1179
1177
Fig. 1. Experimental setup used for the fabrication of the random period gratings.
Fig. 1) was removed from the fusion splicer and the V-grooves were displaced from its aligned position in manual mode to enhance the asymmetry of the electric arc gradient in all directions. On the side where the fiber holder was removed, there is a pulley through which the fiber slides while it is kept under tension by a dead weight. At each discharge the left fiber holder is closed keeping the fiber in position, while the V-groves determine the position of the fiber with respect to the electrodes, and the pulley and weight allow the application of a controlled tension during the LPFG inscription. When the translation stage controlled by a step motor is moved to the next position the fiber holder is released and the fiber advances by itself to the next point due to the applied tension. The arc power for each discharge was kept at 10 mW, with an arc duration of 200 ms. It is worth to note that the lowest allowed values of arc power and arc duration for this particular fusion splicer is 5 mW and 100 ms, while a typical power for a fusion splice is 460 mW and 4800 ms. Here, we only concentrate in fabricating LPFGs in dispersion-shifted fiber (DSF). To choose the random period, in other words the random chirp of the LPFG, we used random number generator (60 numbers or 60 periods in all cases) corresponding to the ‘‘steps’’ of the step-motor. For our particular setup each step of the motor corresponds to an advance of approximately 1.25 mm of the translation stage, i.e., 400 steps correspond to 500 mm of linear advance. The LPFG thus fabricated were monitored continuously during the inscription by coupling the light of a white light source to one end of the LPFG, while the output light from the other end was monitored by an optical spectrum analyzer with a scanning range from 600 to 1650 nm and minimum resolution of 2 nm.
Fig. 2. LPFG with a random period variation in a range of 37.5 mm around 500 mm.
3. Spectral characteristics of random-period arc-induced LPFGs Fig. 2 shows the transmission spectrum of a random-period fiber grating fabricated in DSF. The inset graph shows the random variation of the period around the central period value of 500 mm. The number of periods was 60 with a total length of 30.023 mm, and the range of the random variation was 37.5 mm, with a mean value of 500.39 mm, as shown in the inset table of Fig. 1. The standard deviation of the period of the grating of Fig. 2 was 10.4 mm, showing that even for a relatively high dispersion in the grating period the LPFG is still formed in the fiber. In order to evaluate the maximum dispersion in the random period that still allows the formation of the LPFG, several LPFGs with different variation ranges around the central value of 500 mm were fabricated. Fig. 3 shows the transmission spectra of five random period LPFGs (labeled as LPG1, LPG2, LPG3, LPG4, and LPG5) and a LPFG fabricated with a fixed period of 500 mm (upper graph). In all cases, the number of periods was 60, although in some cases deeper peaks were observed for a number of period lower than 60. Table 1 shows the descriptive statistics of the five random period LPFGs of Fig. 3. It is clearly seen that the position of the resonance peaks of the random period LPFGs are very close to the position of the resonance peaks of the fixed
Fig. 3. Transmission spectra of LPFGs with different random period range around the central period of 500 mm.
Table 1 Descriptive statistics of the random period variation for the LPFGs of Fig. 3. Range (mm) LPG1 25 LPG2 50 LPG3 75 LPG4 95 LPG5 125
Mean (mm)
Std. dev. (mm)
Min. (mm)
Median (mm)
Max. (mm)
500 499.6 500.875 500.6 501.6
8.50 15.22 22.47 25.80 36.98
487.5 475 462.5 450 437.5
502.5 493.125 502.5 507.5 497.5
512.5 525 537.5 545 562.5
1178
A. Martinez-Rios et al. / Optics & Laser Technology 44 (2012) 1176–1179
period gratings, even for a random period with a standard deviation of 36.98 mm (LPG5), showing that the mean period value determines their transmission spectrum. In addition, the FWHM of the outermost resonance peak does not modify appreciably when compared to that of the fixed period LPFG. The FWHM spectral width of the outermost resonance peak at 1598.5 nm of the fixed period LPFG was 7.5 nm, while the corresponding widths for the same resonance peaks of the LPFG1–LPFG5 were 6.1 nm, 4.95 nm, 5.25 nm, 4.55 nm, and 7.4 nm, respectively. This tells us that the randomization of the period does not affect the spectral width, as in the case of linearly chirped LPFGs where the increase in the spectral width of the resonance peaks is expected. It is worth to note that the position of the peaks of Figs. 2 and 3 differ significantly despite having the same central period value of 500 mm. The explanation for this difference is due to the fact that the gratings were fabricated under different conditions of humidity and ambient temperature, which is known to affect the fabrication process [9]. Another random period LPFG (not shown) was fabricated, with a variation range of 146.25 mm, and a standard deviation of 44.06 mm (mean value of 492.3 mm) whose transmission spectrum showed small modulations ( 6 dB) at positions that do not correspond to the mean period value. This should be close to the limit of dispersion in the random period to observe a well formed LPFG. To further prove the assertion that the mean period value of the random period LPFGs determines their transmission spectrum, we fabricated three samples with a central period value of 431 mm (Fig. 4). As can be observed the position of the outermost resonance peak is close in the three cases, as indicated in Fig. 4. In this particular case, the resonance is at longer wavelengths for a higher mean period value, as in the case of fixed period gratings.
4. Sensitivity to ambient refractive index
Fig. 4. Transmission spectra of LPFGs with different random period ranges around the central period of 431.5 mm.
Fig. 5. Sensitivity of the outermost resonance peaks (fourth peak to the right in Fig. 2) to changes in the external refractive index.
In order to evaluate the possible different response to external factors in random period LPFGs with different random period variations, we evaluate the sensitivity to changes in the external refractive index. Fig. 5 shows the results from the measurement of the wavelength shift of the outermost peak with changes in the refractive index for the gratings LPG1 and LPG4. As can be observed there are no noticeable differences in the sensitivity even though the dispersion in the random period is significantly different for both fibers (8.5 mm standard deviation for LPG1, and 25.8 mm standard deviation for LPG4). This is a confirmation that the average characteristics, in particular, the mean period value, determines the performance of the random LPFGs. Our results contrast with the theoretical findings in Ref. [7], where it was found that a small random variation in the period degrade the LPFG characteristics. On the other hand, in Ref. [8] it was theoretically found that small variations in period do not affect substantially the position and shape of the resonance peaks. However, it was found that the form of the refractive index sensitivity curve can be engineered by properly designing the period variation. In our case, we did not observe this behavior since the randomization of the period averages out any accumulated effect. In addition the fibers considered in these works are induced with UV irradiation, where the affected zone is by far smaller than that in arc-induced LPFG. In fact, our results are more correlated with the theoretical findings of Ref. [11], where the effects of high-frequency and low-frequency random fluctuations in the period of fiber Bragg gratings were analyzed. Our case corresponds to the case of high-frequency random fluctuations, where the transmission spectrum is almost the same as that of a uniform grating with a period equal to the average period. As stated in [11], the coupling is the effect over a number of periods and the coupling wavelength is proportional to the average period and the average index change. Also, as was mentioned earlier, for an LPG with a standard deviation in the period of 44.06 mm the observed peaks do not correspond to mean period value, which also can be explained in Ref. [11], where for a large highfrequency random fluctuation destroy the period structure dividing the gratings in pieces and observed peaks are a result of the interference between them. It is worth to note, that the random period LPFGs presented here have more resemblance with the LPFGs inscribed by CO2 laser in random hole fibers [12], where although the period is fixed, the index profile changes at each
A. Martinez-Rios et al. / Optics & Laser Technology 44 (2012) 1176–1179
point. Then, as is claimed for random hole fibers, the random period LPFGs can alleviate the penalties arising from the need of strict ordering of the fiber perturbations during the fabrication of LPFGs. For example, if during the fabrication of a fixed period LPFG a period is lost, or its separation from the next perturbation is lower than the nominal period, the LPFG can become a phaseshifted LPFG with a totally different transmission spectrum. We did not give any value of refractive index sensitivity in RIU (refractive index units) since the reported sensitivities depend strongly on the measurement apparatus, i.e. the available power at the spectral region being monitored and the spectral resolution of the measurement apparatus. For the refractive index measurements of Fig. 5, we used a resolution of 2 nm since our interest was only in strong deviations from the usual behavior of fixed period LPFGs. As in the case of the refractive index sensitivity of random period LPFGs, we do not expect any significant difference in their response to temperature and strain changes. However, it may be some significant impact in the polarization dependent loss and waveguide dispersion that we will explore in our future work.
5. Conclusions In conclusion, we have demonstrated experimentally for the first time to our knowledge, the formation of LPFGs with strong random variations in the period. We have found that the position of the resonance peaks corresponds to the position of peaks in a fixed period LPFGs, which have a period close to the mean value of the random period. The random period variation around the central period value can be as high as 125 mm, with a standard deviation of 36.98 mm with practically no degradation of the spectral transmission characteristics. In particular, there are no noticeable changes in the FWHM spectral width of the resonance peaks, and in some cases is slightly lower for random period LPFGs. The sensitivity to changes in the external refractive index was measured, showing that the response from random period LPFGs is similar, independently of the dispersion in the random
1179
period. The randomization of the period can alleviate the strict tolerances needed during the fabrication of the LPFGs with fixed period, since in random period LPFGs the mean period value determines the transmission characteristics.
Acknowledgments The authors acknowledge CIO for their support. References [1] Vengsarkar Ashish M, Renee Pedrazzani J, Judkins Justin B, Lemaire Paul J, Bergano Neal S, Davidson Carl R. Long-period fiber-grating-based gain equalizers. Optics Letters 1996;21:336–8. [2] Kalachev AI, Pureur V, Nikogosyan DN. €lnvestigation of long-period fiber gratings induced by high-intensity femtosecond UV laser pulses. Optics Communications 2005;246:107–15. [3] Rao Y-J, Wang Y-P, Ran Z-L, Zhu T. Novel fiber-optic sensors based on longperiod fiber gratings written by high-frequency CO2 laser pulses. IEEE Journal of Lightwave Technology 2003;21:1320–7. [4] Savin S, Digonnet MJF, Kino GS, Shaw HJ. Tunable mechanically induced longperiod fiber gratings. Optics Letters 2000;25(10):710–2. [5] Rego G, Falate R, Santos JL, Salgado HM, Fabris JL, Semjonov SL, Dianov EM. Arc-induced long-period gratings in aluminosilicate glass fibers. Optics Letters 2005;30:2065–7. [6] Mizunami T, Sho Y, Yamamoto K, Ishida Y. Long-period fiber-gratings produced by exposure with a low-pressure mercury lamp and their sensing characteristics. Optics Communications 2009;282:4699–705. [7] Chung K-W, Yin S. Analysis of random grating period and amplitude errors in ultra-thin long-period grating. Microwave and Optical Technology Letters 2001;30:178–81. [8] Flores-Llamas I, Svyryd V, Khotiaintsev SN. Design of long-period fiber grating refractometric sensors with linear response by a genetic algorithm. IEEE Sensors Journal 2008;8:1130. [9] Ivanov OV, Rego G. Origin of coupling to antisymmetric modes in arc-induced long-period fiber gratings. Optics Express 2007;15:13936–41. [10] Garcia-de-la-Rosa LA, Torres-Gomez I, Martinez-Rios A, Reyes-Gomez J. Background loss minimization in arc-induced long-period fiber gratings. Optical Engineering 2010;49:065001. [11] Lu C, Cui J, Cui Y. Reflection spectra of fiber Bragg gratings with random fluctuations. Optical Fiber Technology 2008;14:97–101. [12] Wang K, Pickrell G. Long period gratings in random hole optical fibers for refractive index sensing. Sensors 2011;11:1558.
Contents lists available at SciVerse ScienceDirect
Optics & Laser Technology journal homepage: www.elsevier.com/locate/optlastec
Research Note
Random period arc-induced long-period fiber gratings A. Martinez-Rios a,n, I. Torres-Gomez a, D. Monzon-Hernandez a, G. Salceda-Delgado a, V.M. Duran-Ramirez b, G. Anzueto-Sanchez c ´ ptica, Lomas del Bosque 115, Colonia Lomas del Campestre, 37150 Leo ´n, Guanajuato, Me´xico Centro de Investigaciones en O Centro Universitario de los Lagos, Universidad de Guadalajara, Lagos de Moreno Jal., Me´xico c ´n en Ingenierı´a y Ciencias Aplicadas CIICAp, Universidad Auto ´noma del Estado de Morelos, Cuernavaca, Morelos 62210, Me´xico Centro de Investigacio a
b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 6 September 2011 Received in revised form 22 November 2011 Accepted 23 November 2011 Available online 7 December 2011
We report the fabrication of arc-induced long-period fiber gratings with strong random variations in the period. Long-period fiber gratings with standard deviations in the period from 8.50 to 36.98 mm were fabricated. The spectral position of the resonant bands is determined by the average period value, being similar to that observed in a long-period fiber grating with a fixed period equal to the average period of the random grating. Moreover the notch bands keep the shape characteristics like wideband and depth compared with a long-period grating with a constant period. In addition, their sensitivity to external parameters such as ambient refractive index is not too different with that of fixed period longperiod gratings. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Long-period fiber gratings Random period Refractive index
1. Introduction Long-period fiber gratings (LPFGs) are devices where the coupling between co-propagating core and cladding modes results in the formation of well defined loss bands at wavelengths determined by the period of the grating and the fiber characteristics [1]. LPFGs can be fabricated in optical fibers by the application of a periodic perturbation of the refractive index, which can be induced by UV [2] or CO2 [3] laser irradiation, by the application of mechanical stress [4], by the application of an electric arc [5], or even using an UV lamp [6]. An important factor to note is that in the case of LPFGs fabricated by laser irradiation the affected zone where the refractive index is perturbed is of a few microns in the case of UV irradiated fibers and maybe less than 50 mm for CO2 laser irradiation, while for arc-induced LPFGs the affected zone is higher than 300 mm. Thus, in the first two cases even small variations in the period may significantly affect their transmission characteristics. In Ref. [7], the effect of slight random variations in amplitude and period during the grating fabrication were simulated, finding that random period variations degrade the spectral characteristics of the gratings. On the other hand, in Ref. [8] a nonuniform LPFG was numerically designed to provide a linear sensitivity curve to external refractive index. In this case, the grating period has a cosine-like period variation, with 3 mm variation range (i.e. the difference between the maximum and minimum period). In contrast, the perturbation
n
Corresponding author. Tel.: þ52 477 441 42 00x197; fax: þ 52 477 441 42 00. E-mail address: [email protected] (A. Martinez-Rios).
0030-3992/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.optlastec.2011.11.037
zone in arc-induced LPFGs has a gradient along the longitudinal and transverse directions that may vary from point to point and affects their reproducibility under the same assumed fabrication conditions, particularly in poorly aligned setups [9]. In the size of the affected zone and its gradient along the longitudinal and transverse directions we may see arc-induced LPFGs having a random chirp, which make its reproducibility difficult unless special measures are taken during the fabrication [10]. In this work, we present the formation of LPFGs with random variation ranges up to 125 mm (standard deviation of 36.98 mm) in dispersion-shifted fiber (DSF) fabricated by the electric arc method. We found that the position of the resonance peaks is approximately determined by the mean period, i.e., the number and position of the resonances correspond to that obtained with a fixed period grating with period close to the mean value. Moreover, the spectral characteristics does not degrade, for example, the FWHM spectral width is not different from that obtained with fixed period LPFGs and in some cases is slightly smaller. Furthermore, we have measured the response to changes in external refractive index of the random period LPFGs without finding any significant difference in the response curve with the dispersion in the random period.
2. Experimental setup Fig. 1 shows a sketch of the experimental setup used for the fabrication of the arc-induced LPFGs. The fiber holders, V-groves, and electrodes are contained within a commercial fusion splicer S-176 from Fitel. One of the fiber holders (the right holder in
A. Martinez-Rios et al. / Optics & Laser Technology 44 (2012) 1176–1179
1177
Fig. 1. Experimental setup used for the fabrication of the random period gratings.
Fig. 1) was removed from the fusion splicer and the V-grooves were displaced from its aligned position in manual mode to enhance the asymmetry of the electric arc gradient in all directions. On the side where the fiber holder was removed, there is a pulley through which the fiber slides while it is kept under tension by a dead weight. At each discharge the left fiber holder is closed keeping the fiber in position, while the V-groves determine the position of the fiber with respect to the electrodes, and the pulley and weight allow the application of a controlled tension during the LPFG inscription. When the translation stage controlled by a step motor is moved to the next position the fiber holder is released and the fiber advances by itself to the next point due to the applied tension. The arc power for each discharge was kept at 10 mW, with an arc duration of 200 ms. It is worth to note that the lowest allowed values of arc power and arc duration for this particular fusion splicer is 5 mW and 100 ms, while a typical power for a fusion splice is 460 mW and 4800 ms. Here, we only concentrate in fabricating LPFGs in dispersion-shifted fiber (DSF). To choose the random period, in other words the random chirp of the LPFG, we used random number generator (60 numbers or 60 periods in all cases) corresponding to the ‘‘steps’’ of the step-motor. For our particular setup each step of the motor corresponds to an advance of approximately 1.25 mm of the translation stage, i.e., 400 steps correspond to 500 mm of linear advance. The LPFG thus fabricated were monitored continuously during the inscription by coupling the light of a white light source to one end of the LPFG, while the output light from the other end was monitored by an optical spectrum analyzer with a scanning range from 600 to 1650 nm and minimum resolution of 2 nm.
Fig. 2. LPFG with a random period variation in a range of 37.5 mm around 500 mm.
3. Spectral characteristics of random-period arc-induced LPFGs Fig. 2 shows the transmission spectrum of a random-period fiber grating fabricated in DSF. The inset graph shows the random variation of the period around the central period value of 500 mm. The number of periods was 60 with a total length of 30.023 mm, and the range of the random variation was 37.5 mm, with a mean value of 500.39 mm, as shown in the inset table of Fig. 1. The standard deviation of the period of the grating of Fig. 2 was 10.4 mm, showing that even for a relatively high dispersion in the grating period the LPFG is still formed in the fiber. In order to evaluate the maximum dispersion in the random period that still allows the formation of the LPFG, several LPFGs with different variation ranges around the central value of 500 mm were fabricated. Fig. 3 shows the transmission spectra of five random period LPFGs (labeled as LPG1, LPG2, LPG3, LPG4, and LPG5) and a LPFG fabricated with a fixed period of 500 mm (upper graph). In all cases, the number of periods was 60, although in some cases deeper peaks were observed for a number of period lower than 60. Table 1 shows the descriptive statistics of the five random period LPFGs of Fig. 3. It is clearly seen that the position of the resonance peaks of the random period LPFGs are very close to the position of the resonance peaks of the fixed
Fig. 3. Transmission spectra of LPFGs with different random period range around the central period of 500 mm.
Table 1 Descriptive statistics of the random period variation for the LPFGs of Fig. 3. Range (mm) LPG1 25 LPG2 50 LPG3 75 LPG4 95 LPG5 125
Mean (mm)
Std. dev. (mm)
Min. (mm)
Median (mm)
Max. (mm)
500 499.6 500.875 500.6 501.6
8.50 15.22 22.47 25.80 36.98
487.5 475 462.5 450 437.5
502.5 493.125 502.5 507.5 497.5
512.5 525 537.5 545 562.5
1178
A. Martinez-Rios et al. / Optics & Laser Technology 44 (2012) 1176–1179
period gratings, even for a random period with a standard deviation of 36.98 mm (LPG5), showing that the mean period value determines their transmission spectrum. In addition, the FWHM of the outermost resonance peak does not modify appreciably when compared to that of the fixed period LPFG. The FWHM spectral width of the outermost resonance peak at 1598.5 nm of the fixed period LPFG was 7.5 nm, while the corresponding widths for the same resonance peaks of the LPFG1–LPFG5 were 6.1 nm, 4.95 nm, 5.25 nm, 4.55 nm, and 7.4 nm, respectively. This tells us that the randomization of the period does not affect the spectral width, as in the case of linearly chirped LPFGs where the increase in the spectral width of the resonance peaks is expected. It is worth to note that the position of the peaks of Figs. 2 and 3 differ significantly despite having the same central period value of 500 mm. The explanation for this difference is due to the fact that the gratings were fabricated under different conditions of humidity and ambient temperature, which is known to affect the fabrication process [9]. Another random period LPFG (not shown) was fabricated, with a variation range of 146.25 mm, and a standard deviation of 44.06 mm (mean value of 492.3 mm) whose transmission spectrum showed small modulations ( 6 dB) at positions that do not correspond to the mean period value. This should be close to the limit of dispersion in the random period to observe a well formed LPFG. To further prove the assertion that the mean period value of the random period LPFGs determines their transmission spectrum, we fabricated three samples with a central period value of 431 mm (Fig. 4). As can be observed the position of the outermost resonance peak is close in the three cases, as indicated in Fig. 4. In this particular case, the resonance is at longer wavelengths for a higher mean period value, as in the case of fixed period gratings.
4. Sensitivity to ambient refractive index
Fig. 4. Transmission spectra of LPFGs with different random period ranges around the central period of 431.5 mm.
Fig. 5. Sensitivity of the outermost resonance peaks (fourth peak to the right in Fig. 2) to changes in the external refractive index.
In order to evaluate the possible different response to external factors in random period LPFGs with different random period variations, we evaluate the sensitivity to changes in the external refractive index. Fig. 5 shows the results from the measurement of the wavelength shift of the outermost peak with changes in the refractive index for the gratings LPG1 and LPG4. As can be observed there are no noticeable differences in the sensitivity even though the dispersion in the random period is significantly different for both fibers (8.5 mm standard deviation for LPG1, and 25.8 mm standard deviation for LPG4). This is a confirmation that the average characteristics, in particular, the mean period value, determines the performance of the random LPFGs. Our results contrast with the theoretical findings in Ref. [7], where it was found that a small random variation in the period degrade the LPFG characteristics. On the other hand, in Ref. [8] it was theoretically found that small variations in period do not affect substantially the position and shape of the resonance peaks. However, it was found that the form of the refractive index sensitivity curve can be engineered by properly designing the period variation. In our case, we did not observe this behavior since the randomization of the period averages out any accumulated effect. In addition the fibers considered in these works are induced with UV irradiation, where the affected zone is by far smaller than that in arc-induced LPFG. In fact, our results are more correlated with the theoretical findings of Ref. [11], where the effects of high-frequency and low-frequency random fluctuations in the period of fiber Bragg gratings were analyzed. Our case corresponds to the case of high-frequency random fluctuations, where the transmission spectrum is almost the same as that of a uniform grating with a period equal to the average period. As stated in [11], the coupling is the effect over a number of periods and the coupling wavelength is proportional to the average period and the average index change. Also, as was mentioned earlier, for an LPG with a standard deviation in the period of 44.06 mm the observed peaks do not correspond to mean period value, which also can be explained in Ref. [11], where for a large highfrequency random fluctuation destroy the period structure dividing the gratings in pieces and observed peaks are a result of the interference between them. It is worth to note, that the random period LPFGs presented here have more resemblance with the LPFGs inscribed by CO2 laser in random hole fibers [12], where although the period is fixed, the index profile changes at each
A. Martinez-Rios et al. / Optics & Laser Technology 44 (2012) 1176–1179
point. Then, as is claimed for random hole fibers, the random period LPFGs can alleviate the penalties arising from the need of strict ordering of the fiber perturbations during the fabrication of LPFGs. For example, if during the fabrication of a fixed period LPFG a period is lost, or its separation from the next perturbation is lower than the nominal period, the LPFG can become a phaseshifted LPFG with a totally different transmission spectrum. We did not give any value of refractive index sensitivity in RIU (refractive index units) since the reported sensitivities depend strongly on the measurement apparatus, i.e. the available power at the spectral region being monitored and the spectral resolution of the measurement apparatus. For the refractive index measurements of Fig. 5, we used a resolution of 2 nm since our interest was only in strong deviations from the usual behavior of fixed period LPFGs. As in the case of the refractive index sensitivity of random period LPFGs, we do not expect any significant difference in their response to temperature and strain changes. However, it may be some significant impact in the polarization dependent loss and waveguide dispersion that we will explore in our future work.
5. Conclusions In conclusion, we have demonstrated experimentally for the first time to our knowledge, the formation of LPFGs with strong random variations in the period. We have found that the position of the resonance peaks corresponds to the position of peaks in a fixed period LPFGs, which have a period close to the mean value of the random period. The random period variation around the central period value can be as high as 125 mm, with a standard deviation of 36.98 mm with practically no degradation of the spectral transmission characteristics. In particular, there are no noticeable changes in the FWHM spectral width of the resonance peaks, and in some cases is slightly lower for random period LPFGs. The sensitivity to changes in the external refractive index was measured, showing that the response from random period LPFGs is similar, independently of the dispersion in the random
1179
period. The randomization of the period can alleviate the strict tolerances needed during the fabrication of the LPFGs with fixed period, since in random period LPFGs the mean period value determines the transmission characteristics.
Acknowledgments The authors acknowledge CIO for their support. References [1] Vengsarkar Ashish M, Renee Pedrazzani J, Judkins Justin B, Lemaire Paul J, Bergano Neal S, Davidson Carl R. Long-period fiber-grating-based gain equalizers. Optics Letters 1996;21:336–8. [2] Kalachev AI, Pureur V, Nikogosyan DN. €lnvestigation of long-period fiber gratings induced by high-intensity femtosecond UV laser pulses. Optics Communications 2005;246:107–15. [3] Rao Y-J, Wang Y-P, Ran Z-L, Zhu T. Novel fiber-optic sensors based on longperiod fiber gratings written by high-frequency CO2 laser pulses. IEEE Journal of Lightwave Technology 2003;21:1320–7. [4] Savin S, Digonnet MJF, Kino GS, Shaw HJ. Tunable mechanically induced longperiod fiber gratings. Optics Letters 2000;25(10):710–2. [5] Rego G, Falate R, Santos JL, Salgado HM, Fabris JL, Semjonov SL, Dianov EM. Arc-induced long-period gratings in aluminosilicate glass fibers. Optics Letters 2005;30:2065–7. [6] Mizunami T, Sho Y, Yamamoto K, Ishida Y. Long-period fiber-gratings produced by exposure with a low-pressure mercury lamp and their sensing characteristics. Optics Communications 2009;282:4699–705. [7] Chung K-W, Yin S. Analysis of random grating period and amplitude errors in ultra-thin long-period grating. Microwave and Optical Technology Letters 2001;30:178–81. [8] Flores-Llamas I, Svyryd V, Khotiaintsev SN. Design of long-period fiber grating refractometric sensors with linear response by a genetic algorithm. IEEE Sensors Journal 2008;8:1130. [9] Ivanov OV, Rego G. Origin of coupling to antisymmetric modes in arc-induced long-period fiber gratings. Optics Express 2007;15:13936–41. [10] Garcia-de-la-Rosa LA, Torres-Gomez I, Martinez-Rios A, Reyes-Gomez J. Background loss minimization in arc-induced long-period fiber gratings. Optical Engineering 2010;49:065001. [11] Lu C, Cui J, Cui Y. Reflection spectra of fiber Bragg gratings with random fluctuations. Optical Fiber Technology 2008;14:97–101. [12] Wang K, Pickrell G. Long period gratings in random hole optical fibers for refractive index sensing. Sensors 2011;11:1558.