Reflective long-period fiber grating-based sensor with Sagnac fiber loop mirror for simultaneous measurement of refractive index and temperature Jianying Yuan, Chun-Liu Zhao,* Yumeng Zhou, Xiangdong Yu, Juan Kang, Jianfeng Wang, and Shangzhong Jin College of Optical and Electronic Technology, China Jiliang University, Hangzhou 310018, China *Corresponding author: [email protected] Received 29 April 2014; revised 27 June 2014; accepted 17 July 2014; posted 18 July 2014 (Doc. ID 211133); published 15 August 2014
In this paper, we propose a reflective long-period grating-based sensor with a Sagnac fiber loop mirror (SFLM) for simultaneous measurement of refractive index (RI) and temperature. By cascading the SFLM to the end of a long-period fiber grating (LPFG), the LPFG works as a reflection operation, which is convenient in some applications. Further, the SFLM and the LPFG have different sensitivities to RI and temperature. As a result, RI and temperature measurement can be simultaneously achieved by monitoring the wavelength shifts of the LPFG and the SFLM’s dips in the reflection spectrum. Experimental results show that the temperature sensitivity can reach 1.533 nm/°C, and the RI sensitivity is from 16.864 nm/RIU (refractive index unit) to 113.142 nm/RIU when the RI range is from 1.333 to 1.430. The application for 40 km long-distance RI and temperature measurement shows that the sensor has potential application in long-distance sensing. © 2014 Optical Society of America OCIS codes: (060.2370) Fiber optics sensors; (060.2300) Fiber measurements; (120.6780) Temperature; (120.0280) Remote sensing and sensors. http://dx.doi.org/10.1364/AO.53.000H85
1. Introduction
In the monitoring and controlling of chemical, biological, and environmental industries, the accurate measurement of refractive index (RI) is an important issue because RI is regarded as one of the most important parameters of physical and chemical properties [1–3]. Since the long-period fiber grating (LPFG) is well known to be sensitive to ambient RI, various kinds of RI sensors are fabricated based on LPFGs [4–6]. However, the LPFG is normally used as a sensor head with transmission operation because it couples light from the fundamental core mode to 1559-128X/14/290H85-06$15.00/0 © 2014 Optical Society of America
some cladding modes and leads to dips in the transmission spectrum. This transmission operation is inconvenient and impractical in many sensing applications. More recently, several forms of reflective LPFG-based sensors have been developed, and they are convenient and compact. In 2009, Jiang et al. proposed a reflective LPFG sensor based on a cladding-mode-selective fiber end-face mirror for external RI sensing [7]. In 2013, Alwis et al. coated a sliver mirror on the distal end of a LPFG to form a relative humidity sensing probe [8]. In the same year, Qi et al. made a highly reflective LPFG sensor that worked well in long-distance RI measurement [9]. These reflective LPFG-based sensors have overcome the limitation of the LPFG’s transmission operation. However, the temperature and RI cross 10 October 2014 / Vol. 53, No. 29 / APPLIED OPTICS
H85
sensitivities in such reflective LPFG-based sensors are not considered enough. Furthermore, the sliver mirror coating process is complicated, and the fiber tip is fragile. In this paper, we propose a reflective LPFG sensor assisted by a cascaded Sagnac fiber loop mirror (SFLM) for RI and temperature measurement at the same time. The SFLM is formed by a 3 dB optical coupler and a piece of polarization maintaining fiber (PMF). When a conventional LPFG is cascaded with a PMF-SFLM, the transmission spectrum of the LPFG will be reflected, and a reflective LPFG is formed. Simultaneous RI and temperature measurement is achieved by monitoring the reflected dip wavelength of the LPFG and one dip of the SFLM at the same time since their sensitivities to RI and temperature are different. This dual-parameter sensor with the reflection operation is easy to fabricate. The good performance indicates that it provides an attractive platform for biomedical, chemical, and environmental sensing, especially for long-distance sensing applications.
The LFPG was point-by-point inscribed into a piece of standard SMF by a CO2 laser. The grating period of the LPFG was 420 μm. The LPFG can couple light from the fundamental core mode to some cladding mode, resulting in a series of attenuations in the transmission spectrum. Owing to the thermooptic effect and the thermal expansion coefficient differences between the fiber core and the cladding material, the phase-matching condition of the LPFG changes with the surrounding temperature. So the temperature sensitivity of the LPFG is mainly determined by the thermal-optic and the thermal expansion coefficients of the fiber materials. On the other hand, the variation of external RI can also cause a change of the phase-matching condition, so LPFGbased sensors exhibit an inherent crosstalk problem between temperature and RI in actual measurements. According to the literature [10,11], the response of the LPFG to RI and temperature can be expressed as
2. Setup and Sensing Principle
where ΔλLPFG is the dip wavelength shift of the LPFG, K L;T is the sensitivity coefficient of temperature, and K L;n n is the response function of the RI. ΔT and Δn are the changes of the surrounding temperature and the external RI, respectively. The SFLM was fabricated with a 3 dB optical coupler and a section of PMF. The length, birefringence index, and beat length of the used PMF were 4.3 cm, 5.74 × 10−4 , and 2.7 mm, respectively. Input light is split by the 3 dB coupler equally into two counterpropagating waves, which subsequently recombine. Interference will occur due to the optical path difference of the two counterpropagating waves caused by the PMF [12]. It is known that the birefringence and the length of the PMF are both affected by the surrounding temperature, so that the interference spectrum will shift with temperature. On the other hand, the section of PMF is insensitive to external RI when the PMF is covered with a polymer coating. Assuming that the birefringence and length of the PMF vary directly with the surrounding temperature, there is a linear correlation between the interference wavelength and temperature, which can be simply expressed as [13]
Figure 1 shows the schematic of the proposed sensor. The setup contains an optical broadband source (BBS) with 200 nm wavelength range, an optical spectrum analyzer (OSA), an optical circulator, a LPFG, and a PMF-SFLM. The single-mode fiber (SMF) link that is used for transmitting optical light bidirectionally is 1 km in length. As shown in Fig. 1, the output light from the light source transmits through the circulator, propagates along the SMF link, and then reaches the LPFG. The transmission of the LPFG inputs to the PMF-SFLM, which is formed by a 3 dB coupler and a piece of PMF. In the PMF-SFLM, the input light is split into two counterpropagating beams equally by the 3 dB coupler. Then, these two beams combine at the 3 dB coupler after propagating around the loop and interfere. Thus the transmission of the LPFG returns to the LPFG again. The LPFG assisted by the PMF-SFLM works like a reflection device. The reflection light goes back to the SMF link and is detected by the OSA.
ΔλLPFG K L;T ΔT K L;n nΔn;
ΔλSFLM K S;T ΔT;
(1)
(2)
where ΔλSFLM and ΔT are the variations of the interference dip (or peak) wavelength and temperature, respectively. K S;T is the temperature sensitivity coefficient of the SFLM. From Eqs. (1) and (2), we can get a matrix that can be written as follows:
Fig. 1. Schematic of the dual-parameter measurement sensor. H86
APPLIED OPTICS / Vol. 53, No. 29 / 10 October 2014
Δn ΔT
K L;n n 0
K L;T K S;T
−1
×
ΔλLPFG : ΔλSFLM
(3)
In this matrix, n can be expressed as n0 Δn, where n0 is the initial external RI value. These
Fig. 2. Reflection spectra of (a) LPFG, (b) SFLM, and (c) proposed cascaded device.
coefficients K L;T ; K S;T and the response function K L;n n can be obtained from the fitting curves of the measured data, respectively. So we can measure the RI and temperature simultaneously. Furthermore, the matrix is not affected by the distance between the tested medium and the measurement equipment. The proposed reflective LPFG sensor is suitable for remote sensing within 50 km, a length that is limited by the loss of the SMF link. 3. Experiment and Results
The reflective spectra of the LPFG and the SFLM are individually measured with the OSA, and are displayed in Figs. 2(a) and 2(b), respectively. The reflective spectrum of the LPFG with the cascaded SFLM is shown in Fig. 2(c). It can be seen from Fig. 2(a) that the loss of the reflection spectrum of the LPFG is high, approximately 20 dB, and the lowest intensity of the spectrum is about −48.14 dBm. It can be explained by the low reflectivity of the interface between the fiber end and air. The attenuation dip of the LPFG is located at 1498.5 nm, and its extinction ratio is about 30 dB. In Fig. 2(b), the reflection spectrum of the SFLM has an obvious interference fringe, and the insertion loss is approximately 2.8 dB, which is caused by the mismatching between the SMF and the PMF in two fusing splices (the insertion loss is near to zero if the SMF matches perfectly with the PMF). It shows that using the SFLM as a reflection mirror is an excellent choice for a remote sensing. The free spectral range between two resonant dips is 76.0 nm. In Fig. 2(c), the attenuation dips of the LPFG and the SFLM on the spectrum can be distinguished clearly. Due to a higher reflectivity of the SFLM compared with the Fresnel reflection between the fiber end and air, the lowest intensity of the reflection spectrum of the used LPFG is increased from −48.14 to −37.53 dBm. In our measurements, the loss of the SMF link is 0.192 dB/km, the intensity of the light source is fixed at −21 dBm, and the sensitivity of the OSA is −80 dBm, so that the measurement distance can reach 55.91 km in theory. It
Fig. 3. Measured reflection spectra of the cascaded device with different external RIs.
indicates that the proposed sensor can be used in some applications that require long-distance 55 km sensing. To confirm its feasibility and to characterize the sensitivity of the proposed sensor to the RI and temperature, the sensor area was immersed in liquid with different RI and temperature to observe the variations of the reflection spectra. The external RI response was obtained by placing the sensor area that contains the LPFG and the PMF into glycerol solutions with various concentrations while the environment temperature was kept stable at 20°C. The corresponding RI values of these glycerol solutions were accurately measured by an Abbe refractometer. In our measurement, the LPFG and the PMF were packaged together in a glass slide. During each test, the glass slide was injected with the testing solution, and kept the LPFG and the PMF close within the glycerol solution. Figure 3 shows the reflection spectra of the proposed sensor when the RI of the solution is different. In Fig. 3, we choose three attenuation dips (dip1, dip2, and dip3) in the reflection spectrum for the sensitivity analysis. As shown in Fig. 3, dip1 shifts toward to shorter wavelengths continuously as the external
Fig. 4. Relationships of the wavelength of dips and external RIs. 10 October 2014 / Vol. 53, No. 29 / APPLIED OPTICS
H87
RI increases, meanwhile the variations of dip2 and dip3 are slight. Figure 4 shows the relationships of the wavelengths of three tips as a function of the external RI. It can be seen from Fig. 4 that the attenuation dip of LPFG exhibits a nonlinear response to RI. We can use an inverse function to fitting the measured data [11], which is λdip1 1503.39 0.421∕ n − 1.491 at the RI range from 1.333 to 1.430 with a fitting degree of 0.9962. Thus the RI response function K L;n n inside Eq. (3) can be obtained from the fitting line. As already analyzed above, the PMF is coated with coating material so that the dips of the SFLM are insensitive to external RI. The temperature response of this sensor was measured by placing the sensor area in a temperaturecontrolled container. Figure 5 shows the reflection spectra of the proposed sensor when temperature increases from 20°C to 50°C. As shown in Fig. 5, different dips move independently as temperature increases. We can also observe that the extinction ratio of dip1 has a large variation, which is a result of the superimposed effect of two spectra. In order to avoid this superposition and make the measurements clear, a SFLM with a shorter length or a higher birefringence of the PMF is suggested to produce a larger separation between these two dips (dip1 and dip2). In our experiment, due to the larger extinction ratio of the LPFG, the attenuation dip of the LPFG affects dip1 dominantly, and the dip of the PMF-SFLM (dip2) will not affect the results clearly in the measurement. Figure 6 shows the relationships of the wavelengths of dip1 and dip3 with temperature. Two linear-fitting curves are λdip1 0.069T 1497.30 and λdip3 −1.533T 1619.18, with fitting degrees of 0.9810 and 0.9998, respectively. So the temperature sensitivities of dip1 and dip3 are 69 pm∕°C and −1.533 nm∕°C, respectively. This means better temperature measurement will be obtained by monitoring the dip3 wavelength shift, since dip3 has higher temperature sensitivity. In our experiment, the
With the sensor matrix, we can simultaneously measure the RI and temperature of an unknown sample, or determine the concentration of one solution. When the temperature and RI change at the same time, the reflection spectra are affected by both. As shown in Fig. 7, we measured the response of the
Fig. 5. Measured reflection spectra of the cascaded device with different surrounding temperatures.
Fig. 7. Measured reflection spectra of the proposed sensor with both temperature and RI changing.
H88
APPLIED OPTICS / Vol. 53, No. 29 / 10 October 2014
Fig. 6. Relationships of the wavelength of dips and surrounding temperature.
measurement precision of temperature is mainly determined by the PMF-SFLM. According to the above RI and temperature measurements, these coefficients K L;T , K S;T and the response function K L;n n can be obtained. The sensor matrix can be written as
Δn
ΔT −0.421∕n−1.4932 nm∕RIU 69.0×10−3 nm∕°C −1 ×
ΔλLPFG ΔλSFLM
0
:
−1.533nm∕°C (4)
proposed sensor with both temperature and RI changing. In the initial condition, when the RI was 1.333 and the temperature was 20°C, the wavelengths of dip1 and dip3 were located at 1501.1 and 1588.6 nm, respectively. When the temperature increased to 25°C and the RI changed to 1.340, the wavelength shifts of dip1 and dip3 were 0.219 and −7.688 nm, respectively. According to the matrix (4), the temperature and RI could be calculated and were 25.015°C and 1.33967, respectively. Similarly, the wavelength shifts of dip1 and dip3 were −4.197 and −53.632 nm when the temperature and RI were changed to 55°C and 1.419, as shown in Fig. 7. Therefore, the calculated temperature and RI were 54.985°C and 1.41902, respectively. The measurement values were accordant with the true values. This indicates that the proposed sensor works well when temperature and RI change at the same time. In an actual measurement, we can directly obtain the temperature value by monitoring the wavelength shift of dip3 since dip3 is only affected by temperature and their relationship is linear with a high sensitivity. Then the RI value at that temperature can be easily calculated by eliminating the temperature effect. The OSA with a 0.02 nm measurement precision is used in our experiments, so considering the worst condition when the RI is 1.333, the measurement precisions of the RI and temperature can reach about 1.2 × 10−3 refractive index units (RIU) and 1.3 × 10−2 °C, respectively. To verify the long-distance application of the proposed sensor, we replaced the 1 km SMF link with a 40 km one, and measured the RI and temperature response. Figure 8(a) shows the spectra with different external RIs when the environment temperature remains at 20°C. Figure 8(b) shows the reflective spectra when the temperature increases from 20°C to 50°C when the sensor remains in air. It can be seen from Fig. 8 that the spectra are decreased by 17.16 dB relative to that of the 1 km measurement. However, the wavelength variations of these dips are almost consistent with those of 1 km measurement, which certifies that the sensor works well in the 40 km measurement. The experimental results show that such a sensor has potential applications in remote sensing areas, such as some dangerous chemical industries, in which remote monitoring of the concentrations and temperature of strong acid (or/and strong base) is highly necessary for safety production. Finally, the measurement errors and repeatability with different distances were investigated. Through substituting the measured λdip1 and λdip3 in their fitting functions, the calculated RI and temperature could be obtained. The true values of RI and temperature were measured by an Abbe refractometer and electronic thermometer, respectively. The results are shown in Table 1, which reveals that the maximum absolute RI and temperature error of this sensor are about 1.38e-3 RIU and 2.1e-2°C, respectively.
Fig. 8. Measured reflection spectra of the proposed sensor with (a) different external RI and (b) different surrounding temperature in 40 km measurement.
It is worth noting that the performance of the proposed sensor is highly dependent on the splicing procedure, the measurement range of the OSA, and the bandwidth of the BBS. There was no doubt that with a more proper splicing procedure, the optical loss of the proposed sensor could be reduced. And if a high-power light source and a wide measurement range OSA were available, our cascaded sensor system would have a longer measurement distance.
Table 1.
Comparison of the Measured Values and True Values
RI Value at 21.4°C True Value 1.36233 1.37312 1.38743 1.39356 1.40434 1.43428
Measured Value
Temperature Value in Air
1 km
40 km
True Value/°C
1.36358 1.37403 1.38723 1.39356 1.40431 1.43425
1.36371 1.37392 1.38725 1.39351 1.40432 1.43421
25.00 30.00 35.00 40.00 45.00 50.00
Measured Value/°C 1 km
40 km
25.015 30.010 35.012 39.991 44.983 50.021
25.008 30.013 34.983 40.018 45.006 50.017
10 October 2014 / Vol. 53, No. 29 / APPLIED OPTICS
H89
4. Conclusion
In summary, a dual-parameter measurement scheme based on a reflective LPFG assisted by a SFLM was proposed and experimentally demonstrated. Instead of using the traditional transmission operation of the LPFG, the LPFG works in a reflection operation when a SFLM is cascaded to the end of the LPFG. Experimental results show the proposed reflective LPFG sensor assisted by the cascaded SFLM works well for different sensing distances. Such a reflection operation is more proper in remote application fields. The proposed sensor has a compact design, and dualparameter measurement performance with high sensitivities. The results show that the proposed sensor has good potential in some applications that require long-distance monitoring for dual parameters, such as marine environmental monitoring, chemical procedure control, and some chemical industrial fields. This work was supported by the National Natural Science Foundation of China under Grant No. 61108058, the International Technological Cooperation Projects of Zhejiang under Grant No. 2013C24018, the National Basic Research Program of China (973 program) under Grant No. 2010CB327804, the National Key Technology R&D Program 2011BAF06B02, the Science and Technology Commission of Shanghai Municipality of China under Grant No. 10595812300, and The Program Funded by Scientific Research and Innovation Team of College Students in Zhejiang Province under Grant No. 2013R409051.
3. 4.
5.
6.
7.
8. 9.
10.
11.
12.
References 1. K. Milenko, ´ D. J. J. Hu, P. P. Shum, T. Zhang, J. L. Lim, Y. Wang, and W. Tong, “Photonic crystal fiber tip interferometer for refractive index sensing,” Opt. Lett. 37, 1373–1375 (2012). 2. C. Zhong, C. Shen, Y. You, J. Chu, X. Zou, X. Dong, and J. Wang, “A polarization-maintaining fiber loop mirror based
H90
APPLIED OPTICS / Vol. 53, No. 29 / 10 October 2014
13.
sensor for liquid refractive index absolute measurement,” Sens. Actuators B 168, 360–364 (2012). L. Chen, X. Guo, and J. Hao, “Refractive index measurement by fiber point diffraction longitudinal shearing interferometry,” Appl. Opt. 52, 3655–3661 (2013). R. Garg, S. M. Tripathi, K. Thyagarajan, and W. J. Bock, “Long period fiber grating based temperature-compensated high performance sensor for bio-chemical sensing applications,” Sens. Actuators B 176, 1121–1127 (2013). A. Martinez-Rios, D. Monzon-Hernandez, and I. TorresGomez, “Highly sensitive cladding-etched arc-induced long-period fiber gratings for refractive index sensing,” Opt. Commun. 283, 958–962 (2010). C. Guan, X. Tian, S. Li, X. Zhong, J. Shi, and L. Yuan, “Long period fiber grating and high sensitivity refractive index sensor based on hollow eccentric optical fiber,” Sens. Actuators B 188, 768–771 (2013). M. Jiang, A. P. Zhang, Y. C. Wang, H. Y. Tam, and S. He, “Fabrication of a compact reflective long-period grating sensor with a cladding-mode-selective fiber end-face mirror,” Opt. Express 17, 17976–17982 (2009). L. Alwis, T. Sun, and K. V. Grattan, “Analysis of polyimidecoated optical fiber long period grating-based relative humidity sensor,” IEEE Sens. J. 13, 767–771 (2013). L. Qi, C. L. Zhao, J. Yuan, M. Ye, J. Wang, Z. Zhang, and S. Jin, “Highly reflective long period fiber grating sensor and its application in refractive index sensing,” Sens. Actuators B 193, 185–189 (2014). H. Juan, J. Dora, J. L. Lim, M. Jiang, Y. Wang, F. Luan, P. Ping Shum, and W. Tong, “Long period grating cascaded to photonic crystal fiber modal interferometer for simultaneous measurement of temperature and refractive index,” Opt. Lett. 37, 2283–2285 (2012). D. W. Kim, F. Shen, X. Chen, and A. Wang, “Simultaneous measurement of refractive index and temperature based on a reflection-mode long-period grating and an intrinsic Fabry-Perot interferometer sensor,” Opt. Lett. 30, 3000– 3002 (2005). C. L. Zhao, X. Yang, C. Lu, W. Jin, and M. S. Demokan, “Temperature-insensitive interferometer using a highly birefringent photonic crystal fiber loop mirror,” IEEE Photon. Technol. Lett. 16, 2535–2537 (2004). J. Wang, C. Shen, C. Zhong, J. Chu, and X. Dong, “Liquid refractive index sensor based on a polarizationmaintaining fiber loop mirror,” IEEE Sens. J. 13, 1721–1724 (2013).
In this paper, we propose a reflective long-period grating-based sensor with a Sagnac fiber loop mirror (SFLM) for simultaneous measurement of refractive index (RI) and temperature. By cascading the SFLM to the end of a long-period fiber grating (LPFG), the LPFG works as a reflection operation, which is convenient in some applications. Further, the SFLM and the LPFG have different sensitivities to RI and temperature. As a result, RI and temperature measurement can be simultaneously achieved by monitoring the wavelength shifts of the LPFG and the SFLM’s dips in the reflection spectrum. Experimental results show that the temperature sensitivity can reach 1.533 nm/°C, and the RI sensitivity is from 16.864 nm/RIU (refractive index unit) to 113.142 nm/RIU when the RI range is from 1.333 to 1.430. The application for 40 km long-distance RI and temperature measurement shows that the sensor has potential application in long-distance sensing. © 2014 Optical Society of America OCIS codes: (060.2370) Fiber optics sensors; (060.2300) Fiber measurements; (120.6780) Temperature; (120.0280) Remote sensing and sensors. http://dx.doi.org/10.1364/AO.53.000H85
1. Introduction
In the monitoring and controlling of chemical, biological, and environmental industries, the accurate measurement of refractive index (RI) is an important issue because RI is regarded as one of the most important parameters of physical and chemical properties [1–3]. Since the long-period fiber grating (LPFG) is well known to be sensitive to ambient RI, various kinds of RI sensors are fabricated based on LPFGs [4–6]. However, the LPFG is normally used as a sensor head with transmission operation because it couples light from the fundamental core mode to 1559-128X/14/290H85-06$15.00/0 © 2014 Optical Society of America
some cladding modes and leads to dips in the transmission spectrum. This transmission operation is inconvenient and impractical in many sensing applications. More recently, several forms of reflective LPFG-based sensors have been developed, and they are convenient and compact. In 2009, Jiang et al. proposed a reflective LPFG sensor based on a cladding-mode-selective fiber end-face mirror for external RI sensing [7]. In 2013, Alwis et al. coated a sliver mirror on the distal end of a LPFG to form a relative humidity sensing probe [8]. In the same year, Qi et al. made a highly reflective LPFG sensor that worked well in long-distance RI measurement [9]. These reflective LPFG-based sensors have overcome the limitation of the LPFG’s transmission operation. However, the temperature and RI cross 10 October 2014 / Vol. 53, No. 29 / APPLIED OPTICS
H85
sensitivities in such reflective LPFG-based sensors are not considered enough. Furthermore, the sliver mirror coating process is complicated, and the fiber tip is fragile. In this paper, we propose a reflective LPFG sensor assisted by a cascaded Sagnac fiber loop mirror (SFLM) for RI and temperature measurement at the same time. The SFLM is formed by a 3 dB optical coupler and a piece of polarization maintaining fiber (PMF). When a conventional LPFG is cascaded with a PMF-SFLM, the transmission spectrum of the LPFG will be reflected, and a reflective LPFG is formed. Simultaneous RI and temperature measurement is achieved by monitoring the reflected dip wavelength of the LPFG and one dip of the SFLM at the same time since their sensitivities to RI and temperature are different. This dual-parameter sensor with the reflection operation is easy to fabricate. The good performance indicates that it provides an attractive platform for biomedical, chemical, and environmental sensing, especially for long-distance sensing applications.
The LFPG was point-by-point inscribed into a piece of standard SMF by a CO2 laser. The grating period of the LPFG was 420 μm. The LPFG can couple light from the fundamental core mode to some cladding mode, resulting in a series of attenuations in the transmission spectrum. Owing to the thermooptic effect and the thermal expansion coefficient differences between the fiber core and the cladding material, the phase-matching condition of the LPFG changes with the surrounding temperature. So the temperature sensitivity of the LPFG is mainly determined by the thermal-optic and the thermal expansion coefficients of the fiber materials. On the other hand, the variation of external RI can also cause a change of the phase-matching condition, so LPFGbased sensors exhibit an inherent crosstalk problem between temperature and RI in actual measurements. According to the literature [10,11], the response of the LPFG to RI and temperature can be expressed as
2. Setup and Sensing Principle
where ΔλLPFG is the dip wavelength shift of the LPFG, K L;T is the sensitivity coefficient of temperature, and K L;n n is the response function of the RI. ΔT and Δn are the changes of the surrounding temperature and the external RI, respectively. The SFLM was fabricated with a 3 dB optical coupler and a section of PMF. The length, birefringence index, and beat length of the used PMF were 4.3 cm, 5.74 × 10−4 , and 2.7 mm, respectively. Input light is split by the 3 dB coupler equally into two counterpropagating waves, which subsequently recombine. Interference will occur due to the optical path difference of the two counterpropagating waves caused by the PMF [12]. It is known that the birefringence and the length of the PMF are both affected by the surrounding temperature, so that the interference spectrum will shift with temperature. On the other hand, the section of PMF is insensitive to external RI when the PMF is covered with a polymer coating. Assuming that the birefringence and length of the PMF vary directly with the surrounding temperature, there is a linear correlation between the interference wavelength and temperature, which can be simply expressed as [13]
Figure 1 shows the schematic of the proposed sensor. The setup contains an optical broadband source (BBS) with 200 nm wavelength range, an optical spectrum analyzer (OSA), an optical circulator, a LPFG, and a PMF-SFLM. The single-mode fiber (SMF) link that is used for transmitting optical light bidirectionally is 1 km in length. As shown in Fig. 1, the output light from the light source transmits through the circulator, propagates along the SMF link, and then reaches the LPFG. The transmission of the LPFG inputs to the PMF-SFLM, which is formed by a 3 dB coupler and a piece of PMF. In the PMF-SFLM, the input light is split into two counterpropagating beams equally by the 3 dB coupler. Then, these two beams combine at the 3 dB coupler after propagating around the loop and interfere. Thus the transmission of the LPFG returns to the LPFG again. The LPFG assisted by the PMF-SFLM works like a reflection device. The reflection light goes back to the SMF link and is detected by the OSA.
ΔλLPFG K L;T ΔT K L;n nΔn;
ΔλSFLM K S;T ΔT;
(1)
(2)
where ΔλSFLM and ΔT are the variations of the interference dip (or peak) wavelength and temperature, respectively. K S;T is the temperature sensitivity coefficient of the SFLM. From Eqs. (1) and (2), we can get a matrix that can be written as follows:
Fig. 1. Schematic of the dual-parameter measurement sensor. H86
APPLIED OPTICS / Vol. 53, No. 29 / 10 October 2014
Δn ΔT
K L;n n 0
K L;T K S;T
−1
×
ΔλLPFG : ΔλSFLM
(3)
In this matrix, n can be expressed as n0 Δn, where n0 is the initial external RI value. These
Fig. 2. Reflection spectra of (a) LPFG, (b) SFLM, and (c) proposed cascaded device.
coefficients K L;T ; K S;T and the response function K L;n n can be obtained from the fitting curves of the measured data, respectively. So we can measure the RI and temperature simultaneously. Furthermore, the matrix is not affected by the distance between the tested medium and the measurement equipment. The proposed reflective LPFG sensor is suitable for remote sensing within 50 km, a length that is limited by the loss of the SMF link. 3. Experiment and Results
The reflective spectra of the LPFG and the SFLM are individually measured with the OSA, and are displayed in Figs. 2(a) and 2(b), respectively. The reflective spectrum of the LPFG with the cascaded SFLM is shown in Fig. 2(c). It can be seen from Fig. 2(a) that the loss of the reflection spectrum of the LPFG is high, approximately 20 dB, and the lowest intensity of the spectrum is about −48.14 dBm. It can be explained by the low reflectivity of the interface between the fiber end and air. The attenuation dip of the LPFG is located at 1498.5 nm, and its extinction ratio is about 30 dB. In Fig. 2(b), the reflection spectrum of the SFLM has an obvious interference fringe, and the insertion loss is approximately 2.8 dB, which is caused by the mismatching between the SMF and the PMF in two fusing splices (the insertion loss is near to zero if the SMF matches perfectly with the PMF). It shows that using the SFLM as a reflection mirror is an excellent choice for a remote sensing. The free spectral range between two resonant dips is 76.0 nm. In Fig. 2(c), the attenuation dips of the LPFG and the SFLM on the spectrum can be distinguished clearly. Due to a higher reflectivity of the SFLM compared with the Fresnel reflection between the fiber end and air, the lowest intensity of the reflection spectrum of the used LPFG is increased from −48.14 to −37.53 dBm. In our measurements, the loss of the SMF link is 0.192 dB/km, the intensity of the light source is fixed at −21 dBm, and the sensitivity of the OSA is −80 dBm, so that the measurement distance can reach 55.91 km in theory. It
Fig. 3. Measured reflection spectra of the cascaded device with different external RIs.
indicates that the proposed sensor can be used in some applications that require long-distance 55 km sensing. To confirm its feasibility and to characterize the sensitivity of the proposed sensor to the RI and temperature, the sensor area was immersed in liquid with different RI and temperature to observe the variations of the reflection spectra. The external RI response was obtained by placing the sensor area that contains the LPFG and the PMF into glycerol solutions with various concentrations while the environment temperature was kept stable at 20°C. The corresponding RI values of these glycerol solutions were accurately measured by an Abbe refractometer. In our measurement, the LPFG and the PMF were packaged together in a glass slide. During each test, the glass slide was injected with the testing solution, and kept the LPFG and the PMF close within the glycerol solution. Figure 3 shows the reflection spectra of the proposed sensor when the RI of the solution is different. In Fig. 3, we choose three attenuation dips (dip1, dip2, and dip3) in the reflection spectrum for the sensitivity analysis. As shown in Fig. 3, dip1 shifts toward to shorter wavelengths continuously as the external
Fig. 4. Relationships of the wavelength of dips and external RIs. 10 October 2014 / Vol. 53, No. 29 / APPLIED OPTICS
H87
RI increases, meanwhile the variations of dip2 and dip3 are slight. Figure 4 shows the relationships of the wavelengths of three tips as a function of the external RI. It can be seen from Fig. 4 that the attenuation dip of LPFG exhibits a nonlinear response to RI. We can use an inverse function to fitting the measured data [11], which is λdip1 1503.39 0.421∕ n − 1.491 at the RI range from 1.333 to 1.430 with a fitting degree of 0.9962. Thus the RI response function K L;n n inside Eq. (3) can be obtained from the fitting line. As already analyzed above, the PMF is coated with coating material so that the dips of the SFLM are insensitive to external RI. The temperature response of this sensor was measured by placing the sensor area in a temperaturecontrolled container. Figure 5 shows the reflection spectra of the proposed sensor when temperature increases from 20°C to 50°C. As shown in Fig. 5, different dips move independently as temperature increases. We can also observe that the extinction ratio of dip1 has a large variation, which is a result of the superimposed effect of two spectra. In order to avoid this superposition and make the measurements clear, a SFLM with a shorter length or a higher birefringence of the PMF is suggested to produce a larger separation between these two dips (dip1 and dip2). In our experiment, due to the larger extinction ratio of the LPFG, the attenuation dip of the LPFG affects dip1 dominantly, and the dip of the PMF-SFLM (dip2) will not affect the results clearly in the measurement. Figure 6 shows the relationships of the wavelengths of dip1 and dip3 with temperature. Two linear-fitting curves are λdip1 0.069T 1497.30 and λdip3 −1.533T 1619.18, with fitting degrees of 0.9810 and 0.9998, respectively. So the temperature sensitivities of dip1 and dip3 are 69 pm∕°C and −1.533 nm∕°C, respectively. This means better temperature measurement will be obtained by monitoring the dip3 wavelength shift, since dip3 has higher temperature sensitivity. In our experiment, the
With the sensor matrix, we can simultaneously measure the RI and temperature of an unknown sample, or determine the concentration of one solution. When the temperature and RI change at the same time, the reflection spectra are affected by both. As shown in Fig. 7, we measured the response of the
Fig. 5. Measured reflection spectra of the cascaded device with different surrounding temperatures.
Fig. 7. Measured reflection spectra of the proposed sensor with both temperature and RI changing.
H88
APPLIED OPTICS / Vol. 53, No. 29 / 10 October 2014
Fig. 6. Relationships of the wavelength of dips and surrounding temperature.
measurement precision of temperature is mainly determined by the PMF-SFLM. According to the above RI and temperature measurements, these coefficients K L;T , K S;T and the response function K L;n n can be obtained. The sensor matrix can be written as
Δn
ΔT −0.421∕n−1.4932 nm∕RIU 69.0×10−3 nm∕°C −1 ×
ΔλLPFG ΔλSFLM
0
:
−1.533nm∕°C (4)
proposed sensor with both temperature and RI changing. In the initial condition, when the RI was 1.333 and the temperature was 20°C, the wavelengths of dip1 and dip3 were located at 1501.1 and 1588.6 nm, respectively. When the temperature increased to 25°C and the RI changed to 1.340, the wavelength shifts of dip1 and dip3 were 0.219 and −7.688 nm, respectively. According to the matrix (4), the temperature and RI could be calculated and were 25.015°C and 1.33967, respectively. Similarly, the wavelength shifts of dip1 and dip3 were −4.197 and −53.632 nm when the temperature and RI were changed to 55°C and 1.419, as shown in Fig. 7. Therefore, the calculated temperature and RI were 54.985°C and 1.41902, respectively. The measurement values were accordant with the true values. This indicates that the proposed sensor works well when temperature and RI change at the same time. In an actual measurement, we can directly obtain the temperature value by monitoring the wavelength shift of dip3 since dip3 is only affected by temperature and their relationship is linear with a high sensitivity. Then the RI value at that temperature can be easily calculated by eliminating the temperature effect. The OSA with a 0.02 nm measurement precision is used in our experiments, so considering the worst condition when the RI is 1.333, the measurement precisions of the RI and temperature can reach about 1.2 × 10−3 refractive index units (RIU) and 1.3 × 10−2 °C, respectively. To verify the long-distance application of the proposed sensor, we replaced the 1 km SMF link with a 40 km one, and measured the RI and temperature response. Figure 8(a) shows the spectra with different external RIs when the environment temperature remains at 20°C. Figure 8(b) shows the reflective spectra when the temperature increases from 20°C to 50°C when the sensor remains in air. It can be seen from Fig. 8 that the spectra are decreased by 17.16 dB relative to that of the 1 km measurement. However, the wavelength variations of these dips are almost consistent with those of 1 km measurement, which certifies that the sensor works well in the 40 km measurement. The experimental results show that such a sensor has potential applications in remote sensing areas, such as some dangerous chemical industries, in which remote monitoring of the concentrations and temperature of strong acid (or/and strong base) is highly necessary for safety production. Finally, the measurement errors and repeatability with different distances were investigated. Through substituting the measured λdip1 and λdip3 in their fitting functions, the calculated RI and temperature could be obtained. The true values of RI and temperature were measured by an Abbe refractometer and electronic thermometer, respectively. The results are shown in Table 1, which reveals that the maximum absolute RI and temperature error of this sensor are about 1.38e-3 RIU and 2.1e-2°C, respectively.
Fig. 8. Measured reflection spectra of the proposed sensor with (a) different external RI and (b) different surrounding temperature in 40 km measurement.
It is worth noting that the performance of the proposed sensor is highly dependent on the splicing procedure, the measurement range of the OSA, and the bandwidth of the BBS. There was no doubt that with a more proper splicing procedure, the optical loss of the proposed sensor could be reduced. And if a high-power light source and a wide measurement range OSA were available, our cascaded sensor system would have a longer measurement distance.
Table 1.
Comparison of the Measured Values and True Values
RI Value at 21.4°C True Value 1.36233 1.37312 1.38743 1.39356 1.40434 1.43428
Measured Value
Temperature Value in Air
1 km
40 km
True Value/°C
1.36358 1.37403 1.38723 1.39356 1.40431 1.43425
1.36371 1.37392 1.38725 1.39351 1.40432 1.43421
25.00 30.00 35.00 40.00 45.00 50.00
Measured Value/°C 1 km
40 km
25.015 30.010 35.012 39.991 44.983 50.021
25.008 30.013 34.983 40.018 45.006 50.017
10 October 2014 / Vol. 53, No. 29 / APPLIED OPTICS
H89
4. Conclusion
In summary, a dual-parameter measurement scheme based on a reflective LPFG assisted by a SFLM was proposed and experimentally demonstrated. Instead of using the traditional transmission operation of the LPFG, the LPFG works in a reflection operation when a SFLM is cascaded to the end of the LPFG. Experimental results show the proposed reflective LPFG sensor assisted by the cascaded SFLM works well for different sensing distances. Such a reflection operation is more proper in remote application fields. The proposed sensor has a compact design, and dualparameter measurement performance with high sensitivities. The results show that the proposed sensor has good potential in some applications that require long-distance monitoring for dual parameters, such as marine environmental monitoring, chemical procedure control, and some chemical industrial fields. This work was supported by the National Natural Science Foundation of China under Grant No. 61108058, the International Technological Cooperation Projects of Zhejiang under Grant No. 2013C24018, the National Basic Research Program of China (973 program) under Grant No. 2010CB327804, the National Key Technology R&D Program 2011BAF06B02, the Science and Technology Commission of Shanghai Municipality of China under Grant No. 10595812300, and The Program Funded by Scientific Research and Innovation Team of College Students in Zhejiang Province under Grant No. 2013R409051.
3. 4.
5.
6.
7.
8. 9.
10.
11.
12.
References 1. K. Milenko, ´ D. J. J. Hu, P. P. Shum, T. Zhang, J. L. Lim, Y. Wang, and W. Tong, “Photonic crystal fiber tip interferometer for refractive index sensing,” Opt. Lett. 37, 1373–1375 (2012). 2. C. Zhong, C. Shen, Y. You, J. Chu, X. Zou, X. Dong, and J. Wang, “A polarization-maintaining fiber loop mirror based
H90
APPLIED OPTICS / Vol. 53, No. 29 / 10 October 2014
13.
sensor for liquid refractive index absolute measurement,” Sens. Actuators B 168, 360–364 (2012). L. Chen, X. Guo, and J. Hao, “Refractive index measurement by fiber point diffraction longitudinal shearing interferometry,” Appl. Opt. 52, 3655–3661 (2013). R. Garg, S. M. Tripathi, K. Thyagarajan, and W. J. Bock, “Long period fiber grating based temperature-compensated high performance sensor for bio-chemical sensing applications,” Sens. Actuators B 176, 1121–1127 (2013). A. Martinez-Rios, D. Monzon-Hernandez, and I. TorresGomez, “Highly sensitive cladding-etched arc-induced long-period fiber gratings for refractive index sensing,” Opt. Commun. 283, 958–962 (2010). C. Guan, X. Tian, S. Li, X. Zhong, J. Shi, and L. Yuan, “Long period fiber grating and high sensitivity refractive index sensor based on hollow eccentric optical fiber,” Sens. Actuators B 188, 768–771 (2013). M. Jiang, A. P. Zhang, Y. C. Wang, H. Y. Tam, and S. He, “Fabrication of a compact reflective long-period grating sensor with a cladding-mode-selective fiber end-face mirror,” Opt. Express 17, 17976–17982 (2009). L. Alwis, T. Sun, and K. V. Grattan, “Analysis of polyimidecoated optical fiber long period grating-based relative humidity sensor,” IEEE Sens. J. 13, 767–771 (2013). L. Qi, C. L. Zhao, J. Yuan, M. Ye, J. Wang, Z. Zhang, and S. Jin, “Highly reflective long period fiber grating sensor and its application in refractive index sensing,” Sens. Actuators B 193, 185–189 (2014). H. Juan, J. Dora, J. L. Lim, M. Jiang, Y. Wang, F. Luan, P. Ping Shum, and W. Tong, “Long period grating cascaded to photonic crystal fiber modal interferometer for simultaneous measurement of temperature and refractive index,” Opt. Lett. 37, 2283–2285 (2012). D. W. Kim, F. Shen, X. Chen, and A. Wang, “Simultaneous measurement of refractive index and temperature based on a reflection-mode long-period grating and an intrinsic Fabry-Perot interferometer sensor,” Opt. Lett. 30, 3000– 3002 (2005). C. L. Zhao, X. Yang, C. Lu, W. Jin, and M. S. Demokan, “Temperature-insensitive interferometer using a highly birefringent photonic crystal fiber loop mirror,” IEEE Photon. Technol. Lett. 16, 2535–2537 (2004). J. Wang, C. Shen, C. Zhong, J. Chu, and X. Dong, “Liquid refractive index sensor based on a polarizationmaintaining fiber loop mirror,” IEEE Sens. J. 13, 1721–1724 (2013).
