Open Access
Add to BibSonomy
Abstract
A compact and robust Fabry-Perot interferometer (FPI) based on polymer core is proposed and experimentally demonstrated. The fabrication is low-cost and has simple processes, including fusion splicing and polymer injection. Its characteristic is that the polymer fills the entire capillary core, which is easy to demodulate, and provides a good platform for the refractive index measurement of the polymer after curing. The experimental result shows a linear temperature sensitivity of 1226.64 pm/°C between 39°C and 54°C. Furthermore, we also used the Vernier effect to improve the temperature sensitivity as high as −15.617 nm/°C. The proposed FPI structure provides potential application in the research of sensors and polymer optical fibers.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Temperature, one of the most common physical quantities, even subtle changes will have a great impact on our lives, such as medical and health, atmospheric climate, food engineering, etc [1]. Therefore, temperature monitoring has always been a hot topic for people. For decades, people have used a variety of electronic temperature sensors to monitor temperature, but these seem to have been unable to meet the needs of today's diversified development. Compared with traditional temperature sensors, optical fiber temperature sensors have received extensive attention due to their high sensitivities, immunity to electromagnetic interference, corrosion resistance, and long transmission distance. In recent years, many different types of fiber optic temperature sensors have been proposed and demonstrated, such as Fabry-Perot interferometers (FPI) [2–4], Mach-Zehnder interferometers [5], distributed sensors [6], fiber gratings [7], and so on. Among these sensors, the fiber-optic FPI has aroused people's interest due to its advantages of easy fabrication, low cost, and compact structure [7–9].
In the optical fiber, the FPI is generally composed of two reflective surfaces. Interference occurs when the light propagates in the fiber core and is reflected multiple times by the two reflective surfaces. Commonly, shift of ambient temperature will change the phase of the interfering light, which can be divided into two ways: (i) The refractive index of the medium changes with temperature, which can be expressed by the thermo-optical coefficient (TOC); (ii) The volume of the medium changes with temperature, which can be described by the thermal expansion coefficient (TEC). The TOC and TEC characterize the response ability of the materials to temperature. The sensitivity of the traditional all-silicon structure fiber sensor is still at a relatively low level, which is limited by the smaller TOC and TEC of silicon [10]. Therefore, to enhance the temperature sensitivity of the sensor, materials with high TOC and TEC could be applied to assist in the production of the sensor [11].
Recently, methods of combine the thermal sensitive materials with FPI based on optical fiber structures have been presented to improve temperature sensitivity. B. Sun et al. proposed a novel polymer-capped FPI and used it to measure the temperature and pressure, its temperature response is 249 pm/°C [12]. The fabrication process of this sensor is very simple, but its structural strength is insufficient because polymer cap is exposed to the air. To solve this problem, polymer was filled into the FPI inside the optical fiber [3,13,14], which made robustness greatly improved, but limited the role of the TEC of the polymer. In addition, the fabrication process of these sensors are relatively complicated or expensive. In 2018, M. Q. Chen et al. demonstrated the use of capillary action to fill Polydimethylsiloxane (PDMS) in a capillary tube, and its temperature sensitivity reached 2.7035 nm/°C [10]. D. Fu et al. [15]and C. Lang et al. [16] proposed to inject PDMS and dimethyl silicone oil into capillary respectively to form a sandwich air cavity structure. However, the above sandwich air cavity structure is composed of multiple FPIs, which is difficult to demodulate and the length is hard to control. In addition, there are still many materials used in the preparation of sensors, such as polymethyl methacrylate (PMMA) [17], Resin [18], Nafion [19] etc. Although the above materials have enhanced the sensitivity of the sensor in varying degrees, we still need to consider the difficulty of its preparation. Some materials such as PDMS needs to be heated for 8 hours at 80°C or placed in room temperature at least 24 hours to complete curing [4,20,21]. In comparison, UV adhesive only needs to be irradiated under UV light for less than 30 minutes [22].
In this paper, A high-sensitive and compact fiber-tip polymer filled probe for temperature sensing and polymer refractometer was proposed and experimentally demonstrated. The sensor was fabricated by splicing the single-mode fiber (SMF) and a short section of hollow-core fiber (HCF) which was filled with UV adhesive. Based on the principle of FPI, the sensor probe can be used to measure the refractive index of cured polymers. Experiments show the use of two different methods to improve the sensitivity of the sensor, with a maximum temperature sensitivity of −15.617 nm/°C and a linearity of 99.7%. The proposed device has the advantages of ultra-high temperature sensitivity, easy manufacturing, and good robustness, which can provide potential research insights for the research of polymer sensors and polymer optical fibers.
2. Structure fabrication and sensing principle
The preparation process of the proposed sensor can be divided into four steps. Firstly, a section of SMF (Corning SMF28e+) and HCF (TSP075150, with a core diameter of 75 µm and a cladding diameter of 150 µm) was spliced by commercial welding machine (Fujikura, Tokyo, Japan, 80C+), and then the HCF was cleaved with a certain length, i.e. SMF1, as shown in Fig. 1(a). Secondly, another cleaved SMF (i.e. SMF2) was dipped into the UV adhesive to obtain a polymer-cap. After that, SMF1 and SMF2 were both clamped by the fusion splicer and their positions were adjusted with a lateral offset, as shown in Fig. 1(b). Thirdly, SMF2 was moved toward SMF1 via the motor of the splicer until the liquid adhered on the end face of SMF2 was attached to the HCF end of SMF1. As a result, the liquid slowly flowed into the HCF along the inner wall due to its fluidity and gravity. It can be seen from Fig. 1(c) that the air inside and outside the HCF was always in circulation, so there was no bubble cavity during the filling process. Fourthly, after repeating the steps of Fig. 1(b)-(c) for many times, the liquid accumulated inside the HCF, and finally a fully filled UV adhesive core can be obtained, as represented in Fig. 1(d). Finally, the filled probe was irradiated by UV light source for 30 minutes to obtain a stable probe.
Fig. 1. Schematic diagram of the device fabrication. (a)The fabrication of hollow cavity. (b)-(d) The process of filling the capillary with UV polymer.
The schematic diagram of the device is shown in Fig. 2. The lead-in fiber-end surface (M1) and the end surface (M2) of the UV adhesive define the reflective surfaces of FPI. First, the light propagating in the SMF is partially reflected by the M1, and then the remaining light is transmitted in the UV cavity and then reflected by the M2. These reflected light beams will interfere in SMF and form reflection spectrum. The wavelength of the reflection spectrum dip can be defined as:
where λm is the dip wavelength of the m-order (m is an integer) interference fringes, n and L are the effective refractive index and cavity length of UV adhesive, respectively. The distance between two adjacent interference dip is called as free spectral range (FSR), which can be expressed as: where ƒ is signal frequency of reflection spectrum. It can be seen from Eq. (2) that the refractive index and length of the interferometer dominate its FSR and frequency. When the length is constant, the change of the refractive index of the material in the cavity will inevitably lead to the change of its frequency and FSR. Using this principle, the refractive index of the cured colloid in the filled cavity can be obtained without the need for expensive instruments such as a refractive index meter. In the fabrication of the sensor, we used the end face of another fiber SMF2 as the reflecting surface to measure the reflection spectrum of the air cavity and its corresponding fast Fourier transform (FFT), as shown in Fig. 3. After that, we used the same method to measure the reflectance spectrum and FFT of the filled polymer cavity after curing. Since the cavity length is fixed, the frequency of the reflectance spectrum increases after filling. Therefore, the refractive index of the UV adhesive after curing can be obtained by calculating the frequency ratio before and after filling to be about 1.4279 (assuming the refractive index of air is 1).Fig. 3. Reflection spectra and corresponding FFT results of unfilled cavity and filled cavity. Inset: the above is unfilled cavity and the below is filled cavity.
The micrographs of the three samples with lengths of 105.39 µm, 125.46 µm, and 135.90 µm and their corresponding reflection spectra are shown in Fig. 4. For the different lengths of the cured samples, increasing the cavity length results in a corresponding decrease in FSR, which is consistent with the description of Eq. (2). From the spectra we can see that longer cavity length correspond to larger losses, this may be caused by the light suffers more diffraction and scattering in the longer cavity. However, a high fringe visibility is favorable in the demodulation. For wavelength demodulation, the change of refractive index of medium or total length of FPI will lead to the shift of interference peak. By taking a derivative of Eq. (1), the temperature sensitivity can be expressed as [10]:
Fig. 4. Right: The microscope images of sensor samples of different length. Left: Reflectance spectra corresponding to different sensor samples.
3. Experiment and results discussion
To verify the temperature response of the proposed sensor, three samples with lengths of 110.5 µm, 127.9 µm, 141.1 µm were made and tested. The temperature sensing system is shown in Fig. 5. An amplified spontaneous emission (ASE, with a wavelength ranging from 1530 nm to 1610 nm) broadband light source supplies light to the proposed interferometer via a circulator, and the reflected light is then collected by an optical spectrum analyzer (OSA, MS9740A with a resolution of 0.1 nm). The FPI structure was fixed on the heating table (with an accuracy of 0.1°C), and the plastic petri dish was covered on the sensor probe to ensure the uniformity of the space temperature. In the experiment, the heating platform was adjusted from 34 to 51°C, collecting data at 3°C intervals.
Figure 6 displays the relationship between the wavelength and temperature response of three samples. As the temperature increases, the spectrum of each sensor shifts toward long wave. At the same time, as the length of the sensor cavity increases, the temperature sensitivity gradually increases. Another sample S4 with a length of 145.09 µm was made and tested in the same way, and its temperature response performance is shown in Fig. 7. In the illustration, the peak near 1545 nm is selected for tracking. As the temperature increases, the peak redshifts, which is the same as the other samples, and its temperature sensitivity reached 1226.64 pm/°C. This is still consistent with theoretical analysis. However, the limitation of sensor production also needs to be considered, that is, the longer the probe, the more difficult the filling process is.
Fig. 6. The linear fitting curves of the temperature-wavelength relationship corresponding to the three samples. S1: 105.39 µm, S2: 125.46 µm, S3: 135.90 µm.
Fig. 7. The relationship between the wavelength and temperature of the reflection spectrum of sample S4. Interpolation: with temperature changes, part of the reflection spectrum of S4.
For the sensor, its stability is a very important indicator. In order to test its stability, we tested the S4 at 25°C and recorded the reflection spectrum every 10 minutes for 1 hour. Four peaks near different wavelengths (1550 nm, 1570 nm, 1590 nm, 1610 nm) were selected and their wavelength shifts were monitored, as shown in Fig. 8. The maximum wavelength shifts of the different peaks were 0.94 nm, 1.17 nm, 1.09 nm and 0.55 nm, corresponding to temperature shifts of 0.77°C, 0.95°C, 0.89°C and 0.45°C, respectively. The slight wavelength shift may be caused by the disturbance of the environment and system noise.
The repeatability of the proposed sensor also tested using S4 under different temperature range, as shown in Fig. 9. The temperature was increased in steps of 3°C from 39 to 54°C and then cooling down to 39°C for two cycles. As a result, a significant hysteresis can be observed, with the peak permanently shifting to shorter wavelengths after successive temperature cycles are completed, approximately 3 nm per cycle. The average deviations between the heating process and the cooling process are 15.64 pm/°C and 88.62 pm/°C, respectively.
We also studied the response of the proposed sensor to gas pressure. In this experiment, the sample (S4) was placed in a sealed air chamber connected with an adjustable pressure pump and a pressure gauge. The Fig. 10 exhibits the gas pressure response of the sensing probe ranging from 100 to 800 kPa with a gradient of 100 kPa. As the air pressure increases, the peak wavelength red shifts, but the wavelength shift is not linear, which is caused by the structure of our sensor. The deformation of the polymer is suppressed due to the reverse stress of the capillary wall and the end face of the optical fiber. We employed the cubic polynomial fitting method to estimate the relationship between resonance wavelength of spectral and gas pressure. The slope of the fitted curve gradually decreases and tends to be parallel to the air pressure axis, which means that the greater the air pressure applied, the wavelength shift gradually decreases.
Fig. 10. Gas pressure characteristics of sensor sample S4. Interpolation: the reflection spectrum with gas pressure ranging from 100 to 500 kPa.
In addition, when the performance of a single sensor probe is known, we can still use the Vernier effect to greatly increase its sensitivity. To generate the Vernier effect, the optical lengths of the sensing probe and reference cavity need to be precisely matched by carefully controlling the two cavity lengths. Equation (4) shows that as long as the optical path length of the reference interferometer is properly configured, the desired magnification can be obtained. However, for the actual situation, we have to consider the sensitivity of a single sensor, the bandwidth of the device, and other constraints [24]. Based on the above considerations, since the sensitivity of the sensor probe S4 is 1226.64 pm/°C and the bandwidth of the laser source is 80 nm, we have made two air cavities with lengths of 184.05 µm and 222.56 µm as reference interferometers. In the following discussion, they will be named after the short reference cavity and the long reference cavity. According to Eq. (4), the theoretical magnifications of the short reference cavity and the long reference cavity can be calculated as 8.99 and 13.33 respectively.
In the system configuration of the Vernier effect, the parallel configuration scheme was applied. The schematic diagram of the experimental system is shown in Fig. 11. Since the reference interferometer is physically separated from the sensing probe, we only need to connect the reference interferometer to the system for testing.
Fig. 11. Experiment setup of the temperature sensing system for Vernier effect of parallel structure.
The temperature test results of the sensing probe connected to the reference interferometer are shown in Fig. 12. It can be found that due to the amplification of the Vernier effect, the shift of the spectrum occupies almost the entire bandwidth of the light source, and the temperature measurement range is also reduced accordingly. We can find that the temperature rises from 41°C to 45°C in both Fig. 12(a)-(b), and their reflectance spectra shift in opposite directions. The reason for this phenomenon is mainly due to the change of the optical path length of the reference cavity. If the optical path length of the reference interferometer is greater than that of the sensing probe, the shift direction of the total spectrum is opposite. Conversely, the total reflection spectrum shifts in the same direction as a single sensing probe. To investigate the sensing characteristics of total sensor, the linear fitting results are demonstrated in Fig. 13. The linear fitting sensitivity of the short reference cavity is 13.154 nm/°C, and the linearity is 0.999. In the long reference cavity, the sensitivity is −15.617 nm/°C and the linearity is 0.997. Therefore, the calculation shows that compared with a single sensor, the sensitivity is increased by 10.72 and 12.73 times respectively, which are consistent with the theoretical calculation. During the test, we only connected the reference cavity to the system without making any changes to the sensor probe. In fact, the function of the reference interferometer can be regarded as a modulation effect on the sensing signal. Therefore, all performances of the sensing probe remain unchanged, only the signal was modulated by the reference cavity.
Fig. 12. Variation of the Vernier effect spectrum generated in parallel with the reference cavity in the temperature range of 41°C∼45°C. (a) Short reference cavity; (b) Long reference cavity.
Fig. 13. The linear fitting of the relationship between the shift of the reflection spectrum and the temperature under the Vernier effect. Black: parallel with long reference cavity; Red: parallel with short reference cavity.
The sensing performance of the proposed FPI temperature probe is compared with the recently proposed FPI fiber sensors, as shown in Table 1. Compared with [12], the robust of the sensor’s structure we proposed has been increased due to the surrounding of the capillary tube, and the sensitivity has also been greatly improved. In [15], on account of the multi-segment PDMS, multi-beam interference appeared in the spectrum, which led to the difficulty of demodulation. The fully filled polymer core in this work solves this problem. The completely hermetic filling method suppresses the thermal expansion properties of the material and causes its sensitivity to be limited, but its robustness is better [13]. Although a relatively high temperature sensitivity was obtained in [16], but the stability of the liquid and the filling length were difficult to control, which led to the optical path matching problem of the reference interferometer that realized the Vernier effect. In our sensor, the length of the polymer can be determined by the length of the hollow cavity, and the parallel reference interferometer can also solve the problem of optical path matching. In addition, we can still further improve the sensitivity by reducing the optical path difference between the reference interferometer and the sensing interferometer, but this requires a larger light source bandwidth. As a comparison, the sensitivities of polymer sensors are 2∼3 orders of magnitude higher than the temperature performance of all-silicon-structured sensors in [25].
Table 1. Sensing performance comparison of the proposed sensor with reported sensor.
Nevertheless, the proposed sensor still has some shortcomings, such as the polymer end face exposed to the air is still easily affected, and the lower reflectivity leads to lower spectral contrast of the sensor. Therefore, plating a protective film with high reflectivity on the end surface is a good improvement plan.
4. Conclusion
In conclusion, we proposed a compact, high-sensitivity fiber tip polymer core sensor. The sensing probe is filled with UV glue, which is simple to prepare and easy to demodulate. Based on the principle of FPI, we demonstrated its function as a cured polymer refractometer. Experiments prove that the sensor has high temperature sensitivity and good stability. In addition, we use the Vernier effect to further improve the temperature sensitivity of the sensor as high as −15.617 nm/°C. This work will provide potential insights for the research of polymer-based sensors and polymer optical fibers.
Funding
Hainan Provincial Natural Science Foundation (2019RC054); Finance Science and Technology Project of Hainan Province (ZDKJ2020009).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. P. Roriz, S. Silva, O. Frazao, and S. Novais, “Optical Fiber Temperature Sensors and Their Biomedical Applications,” Sensors 20(7), 2113 (2020). [CrossRef]
2. E. Vargas-Rodriguez, A. D. Guzman-Chavez, R. Baeza-Serrato, and M. A. Garcia-Ramirez, “Optical fiber FP sensor for simultaneous measurement of refractive index and temperature based on the empirical mode decomposition algorithm,” Sensors 20(3), 664 (2020). [CrossRef]
3. M. Li, Y. Liu, R. Gao, Y. Li, X. Zhao, and S. Qu, “Ultracompact fiber sensor tip based on liquid polymer-filled Fabry-Perot cavity with high temperature sensitivity,” Sensors and Actuators B: Chemical 233, 496–501 (2016). [CrossRef]
4. L. X. Kong, Y. X. Zhang, W. G. Zhang, Y. S. Zhang, L. Yu, S. Wang, P. C. Geng, and T. Y. Yan, “High-sensitivity and fast-response fiber-optic micro-thermometer based on a plano-concave Fabry-Pérot cavity filled with PDMS,” Sensors and Actuators A: Physical 281, 236–242 (2018). [CrossRef]
5. Z. Liu, H. Xiao, M. Liao, X. Han, W. Chen, T. Zhao, H. Jia, J. Yang, and Y. Tian, “PDMS-assisted microfiber MZ interferometer with a knot resonator for temperature sensing,” IEEE Photonics Technol. Lett. 31(5), 337–340 (2019). [CrossRef]
6. Y. Mizuno, A. Theodosiou, K. Kalli, S. Liehr, H. Lee, and K. Nakamura, “Distributed polymer optical fiber sensors: a review and outlook,” Photonics Res. 9(9), 1719–1733 (2021). [CrossRef]
7. Y. Xiong and F. Xu, “Multifunctional integration on optical fiber tips: challenges and opportunities,” Adv Photonics 2(06), 64001 (2020). [CrossRef]
8. A. Miliou, “In-Fiber Interferometric-Based Sensors : Overview and Recent Advances,” Photonics 8(7), 265 (2021). [CrossRef]
9. C. Li, W. Yang, M. Wang, X. Yu, J. Fan, Y. Xiong, Y. Yang, and L. Li, “A review of coating materials used to improve the performance of optical fiber sensors,” Sensors 20(15), 4215 (2020). [CrossRef]
10. M. Chen, Y. Zhao, F. Xia, Y. Peng, and R. Tong, “High sensitivity temperature sensor based on fiber air-microbubble Fabry-Perot interferometer with PDMS-filled hollow-core fiber,” Sensors and Actuators A: Physical 275, 60–66 (2018). [CrossRef]
11. U. O. Uyor, A. P. I. Popoola, O. M. Popoola, and V. S. Aigbodion, “Polymeric cladding materials under high temperature from optical fibre perspective: a review,” Polym. Bull. 77(4), 2155–2177 (2020). [CrossRef]
12. B. Sun, Y. Wang, J. Qu, C. Liao, G. Yin, J. He, J. Zhou, J. Tang, S. Liu, Z. Li, and others, “Simultaneous measurement of pressure and temperature by employing Fabry-Perot interferometer based on pendant polymer droplet,” Opt. Express 23(3), 1906–1911 (2015). [CrossRef]
13. K. Cao, Y. Liu, and S. Qu, “Compact fiber biocompatible temperature sensor based on a hermetically-sealed liquid-filling structure,” Opt. Express 25(24), 29597–29604 (2017). [CrossRef]
14. C. Wang, B. Zhou, H. Jiang, and S. He, “Agarose filled Fabry–Perot cavity for temperature self-calibration humidity sensing,” IEEE Photonics Technol. Lett. 28(19), 2027–2030 (2016). [CrossRef]
15. D. Fu, X. Liu, J. Shang, W. Sun, and Y. Liu, “A Simple, Highly Sensitive Fiber Sensor for Simultaneous Measurement of Pressure and Temperature,” IEEE Photonics Technol. Lett. 32(13), 747–750 (2020). [CrossRef]
16. C. Lang, Y. Liu, Y. Liao, J. Li, and S. Qu, “Ultra-sensitive fiber-optic temperature sensor consisting of cascaded liquid-air cavities based on Vernier effect,” IEEE Sens. J. 20(10), 5286–5291 (2020). [CrossRef]
17. T. T. Salunkhe, H. W. Choi, S. J. Park, J. H. Kim, I. T. Kim, and others, “High sensitivity temperature sensor based on Fresnel reflection with thermosensitive polymer: control of morphology and coating thickness,” Jpn. J. Appl. Phys. 59(SG), SGGG06 (2020). [CrossRef]
18. C. E. Dominguez-Flores, D. Monzon-Hernandez, J. I. Moreno-Basulto, O. Rodriguez-Quiroz, V. P. Minkovich, D. Lopez-Cortes, and I. Hernandez-Romano, “Real-Time Temperature Sensor Based on In-Fiber Fabry-Perot Interferometer Embedded in a Resin,” J. Lightwave Technol. 37(4), 1084–1090 (2019). [CrossRef]
19. S. Liu, Y. Ji, J. Yang, W. Sun, and H. Li, “Nafion film temperature/humidity sensing based on optical fiber Fabry-Perot interference,” Sensors and Actuators A: Physical 269, 313–321 (2018). [CrossRef]
20. X. Wei, X. Song, C. Li, L. Hou, Z. Li, Y. Li, and L. Ran, “Optical Fiber Gas Pressure Sensor Based on Polydimethylsiloxane Microcavity,” J. Lightwave Technol. 39(9), 2988–2993 (2021). [CrossRef]
21. L. Hou, C. Zhao, B. Xu, B. Mao, C. Shen, and D. N. Wang, “Highly sensitive PDMS-filled Fabry–Perot interferometer temperature sensor based on the Vernier effect,” Appl. Opt. 58(18), 4858–4865 (2019). [CrossRef]
22. X. L. Tan, Y. F. Geng, X. J. Li, Y. L. Deng, Z. Yin, and R. Gao, “UV-curable polymer microhemisphere-based fiber-optic Fabry–Perot interferometer for simultaneous measurement of refractive index and temperature,” IEEE Photonics J. 6(4), 1–8 (2014). [CrossRef]
23. J. Tian, Z. Li, Y. Sun, and Y. Yao, “High-sensitivity fiber-optic strain sensor based on the Vernier effect and separated Fabry–Perot interferometers,” J. Lightwave Technol. 37(21), 5609–5618 (2019). [CrossRef]
24. H. Gao, J. Wang, J. Shen, S. Zhang, D. Xu, Y. Zhang, and C. Li, “Study of the vernier effect based on the fabry–perot interferometer: Methodology and application,” Photonics 8(8), 304 (2021). [CrossRef]
25. B. Xu, Y. M. Liu, D. N. Wang, and J. Q. Li, “Fiber Fabry–Pérot interferometer for measurement of gas pressure and temperature,” J. Lightwave Technol. 34(21), 4920–4925 (2016). [CrossRef]
