Open Access
Add to BibSonomy
Abstract
Surface Plasmon Resonance (SPR) based fiber optic curvature sensors have the advantage of being insensitive to temperature and axial strain. However, they have the disadvantage of low sensitivity and small curvature detection range. To improve the performance of SPR curvature sensors, we propose an intensity-modulated microstructured optical fiber (MOF) curvature sensor. In this sensor, two no-core fibers (NCFs) are used as input-output couplers, and MOF with silver film deposited is used as sensing arms. The light in the cladding is used to excite the SPR, and the exciting resonant valley is extremely sensitive to slight bending changes. The performance of this sensor is investigated theoretically and experimentally. Numerical results show that its cladding pattern is more favorable in the excitation of SPR effects. Experimental results show that the cladding mode of MOF is very sensitive to curvature changes, thus giving it a great advantage in bending measurements. Its sensitivity reaches 0.18 dB/m-1, and linearity reaches 0.995 in the curvature range of 0-30 m-1. The sensor has the advantages of high sensitivity, low temperature and axial strain crosstalk, compact structure, and easy fabrication, which make it attractive in the field of bending sensing.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Curvature is a fundamental and essential physical parameter in mechanical engineering, bridge construction, and high-precision instrumentation manufacturing [1–3]. In practical applications, many building structures are subject to bending deformation due to excessive loads, so it is vital to monitor the curvature health of the structure [4,5]. Fiber optic curvature sensors have been widely studied because of their small size, high sensitivity, easy integration, and resistance to electromagnetic interference [6,7]. Its curvature monitoring can be achieved by detecting the change of wavelength or intensity of the optical wave signal in the optical fiber. Over the past few years, sensors based on fiber Bragg gratings (FBGs) [8,9], long-period fiber gratings (LPFGs) [10,11], and fiber interferometers have been used effectively for curvature sensing [12,13]. However, this kind of curvature sensor is also sensitive to temperature and axial strain, which requires temperature and strain compensation, increasing the complexity of the sensing system. Surface plasmon resonance (SPR)-based fiber optic sensors are temperature and strain insensitive and offer unique advantages in curvature sensing [14].
For SPR fiber optic sensors, when the evanescent wave in the fiber resonates with the plasma wave on the metal surface, the energy in the fiber is coupled to the metal surface and transmitted, thus weakening the optical power received by the spectrometer. This phenomenon results in a distinct resonance valley in the spectrum [15]. When the fiber is bent, the depth of the resonance valley deepens, and the bending curvature can be determined by identifying the depth of the resonance valley (output optical power intensity). Therefore, the SPR fiber curvature sensor is an intensity-modulated type sensor. Fewer SPR-based fiber curvature sensors have been reported. In 2010, Takagi et al. proposed a fiber SPR curvature sensor with a hetero-core structure [16]. The light emitted in multimode fiber (MMF) was coupled to the cladding of a single-mode fiber (SMF), and the curvature sensing was achieved by using the SPR effect between the cladding mode and the gold film. The curvature range to be measured is 0-22.6 m-1. In 2020, Liu et al. improved the Takagi-designed sensor by misaligned fusion of MMF and SMF, which allowed more light to be coupled into the cladding, thus enhancing the leakage of the evanescent wave [17]. The experimental results show that the maximum power loss ratio is only 40%, and the average sensitivity is 0.00596 a.u./m-1 (0.0596 dB/m-1). The variation of the power loss determines the variation of the resonance valley depth. Too small power loss limits the depth of the resonant valley, resulting in low sensitivity. Although the misaligned fusion method allows more light to travel through the cladding, most of the light is still in the core, and the light in the core is difficult to leak when bending occurs. In addition, the misaligned fusion method reduces the mechanical strength of the fiber. To address the low sensitivity of SPR fiber curvature sensors, the main problem to be solved is how to effectively couple the light in the core to the cladding while maintaining the mechanical strength of the fiber.
Based on this, we propose the construction of the SPR bending sensor using grapefruit-type microstructured optical fiber (MOF). The sensor consists of MMF-no core fiber (NCF)-MOF-NCF-MMF. The grapefruit-type MOF fiber has a thick cladding and a small core diameter, which allows most of the light to be transmitted in the cladding without misalignment fusion. When the curvature changes, the intensity distribution of the cladding light will change, resulting in different SPR incidence angles and evanescent wave intensities. Then the curvature is measured by the intensity of the SPR resonance valley. The effects of different bending directions (0°, 90°, 180°, 270°) on the curvature sensing effect were experimentally investigated. The results show that the curvature measurement sensitivity is 0.18 dB/m-1 for all four bending directions in the curvature range of 0-30 m-1. In addition, we have confirmed that the sensor is insensitive to temperature and strain, which means that the proposed sensor has high detection accuracy and does not require a cascade parameter compensation system. Finally, the repeatability measurement experiments were carried out, and the results showed that the sensor has good repeatability and stability. The proposed intensity-modulated SPR sensor based on thick cladding MOF provides a new research direction for high-performance curvature sensing.
2. Sensor preparation and experimental principle
Figure 1(a) and Fig. 1(b) show the cross-section and three-dimensional (3D) schematic diagrams of the described SPR curvature sensor. The sensor consists of two MMFs, two NCFs, and one MOF. The main part of the sensor is a MOF in the middle, which a length is 1 cm, and its outer surface is coated with a silver film to excite plasma waves. The cross-section of the MOF is shown in Fig. 1(c), which has a small core area and six air holes surrounding the core. However, it has a large cladding area. The MOF can excite more light to the outer surface of the fiber due to the presence of internal air holes, thus improving the sensitivity of the cladding mode to curvature changes. The silver film was chosen to excite the SPR because it can obtain sharper resonance valleys than other metallic films [18]. The types and structural parameters of optical fibers are shown in Table 1. MMF is a commercial optical fiber (GIF-105) used to transmit optical signals. Since the core diameter of MMF is similar to the air hole diameter of MOF. If the two are directly fused, most of the optical energy will be input into the air holes. Therefore, we make a transition using NCF (special fiber without core), which allows more light energy to be input into the cladding, with an NCF length of 0.5 cm.
Fig. 1. (a) Schematic diagram of SPR curvature sensor structure, (b) 3D image of the sensor, (c) cross-section of MOF.
Table 1. Types and structure parameters of the optical fiber
Next, we analyzed the optical transmission characteristics of the MOF using the finite element method (FEM). The modeling procedure uses the parameters of the MOF in Table 1. The fiber material is silica, which is described by the Sellmeier equation [19]. In addition, a perfect matching layer (PML) and scattering boundary conditions are introduced outside the MOF to obtain more accurate results. The grid division is shown in Fig. 2(a). The simulation results show that the light is transmitted in the core and cladding of the MOF, as shown in Fig. 2(b). It can be seen from the figure that most of the light is distributed in the cladding region, which indicates that the cladding mode can effectively excite the SPR effect and is more sensitive to the change of bending.
Figure 3 shows the preparation process of the SPR curvature sensor. In this experiment, the silver mirror reaction was used to deposit the silver film because this method is low-cost and simple to operate [18]. As shown in Fig. 3(a), silver nitrate solution is added dropwise to a beaker on a magnetic stirrer in the first step. In the second step, add ammonia dropwise to the silver nitrate solution and wait for the mixed solution to change from brown to transparent. The third step is to add potassium hydroxide solution dropwise, at which point the mixed solution in the beaker is gray. In the fourth step, ammonia is added once more dropwise until the solution in the beaker becomes clear again. Finally, the glucose solution was added to the beaker, and the mixture was coated onto the MOF. After 5 minutes, the residual solution on the MOF is rinsed off with deionized water, and the silver coating work is completed. Here, the deposition time is an important parameter because the deposition time determines the thickness of the silver. In order to determine the optimal deposition time, several sets of controlled tests were done for different deposition times. Finally, a deposition time of 5 minutes was the best choice. The spectrum showed a smooth curve at this time and formed the ideal resonance valley. After depositing the silver film, the MOF needs to be fused into the optical path. In Fig. 3(b), a fiber cutter was used to cut off 1 cm of the deposited MOF with silver film. Then, a fusion splicer is used to fuse 0.5 cm of NCF at both ends of the MOF. Subsequently, multimode patch cords are fused at both ends of the NCF. The above steps can obtain the proposed SPR curvature sensor. The method is simple, feasible, and easy to implement.
Fig. 3. Fabrication process of (a) reagent required for silver mirror reaction, (b) the proposed SPR curvature sensor.
When the evanescent wave at the cladding interface and the plasma wave at the metal surface propagate in the same direction with the same wave vector and frequency, they will undergo SPR. As shown in Fig. 4(a), light is incident from the NCF into the cladding of the MOF, and the incident wave reaches the interface between the metal and the medium called the evanescent wave, whose intensity decays rapidly with the increase of the penetration depth. The wave vector component of the evanescent wave at the interface is [20]:
In addition, we used Rsoft software to simulate the light field transmission of the sensing system in both straight and bent states. The MOF and NCF were modeled using the parameters in Table 1. As shown in Fig. 5, light is injected from the NCF into the MOF. Figure 5(a) shows the MOF in the straight state, and Fig. 5(b) shows the MOF in the bent state. The horizontal coordinates in the figure indicate the diameter of the fiber and the degree of bending, the vertical coordinates indicate the length of the fiber, and the color axis indicates the color scale, with different colors indicating different light intensities. From Fig. 5, we can see that when the MOF is bent, the light leakage in the cladding occurs significantly. Therefore, when the outer cladding of the MOF is coated with a silver film, the evanescent field of the cladding mode in contact with the silver film excites the SPR. The leakage of the evanescent field is enhanced with the increase of the fiber bending curvature. Therefore, the depth of the SPR resonance valley deepens.
Fig. 5. (a) Transmission of light field in MOF in straight state, (b)transmission of light field in MOF in bent state.
3. Results and discussion
The prepared sensor is placed on the curvature template, as shown in Fig. 6(a). The curvature template is shown in Fig. 6(b), and curves with different curvatures are etched on the template. We can determine the curvature of the fiber bend according to the curvature template. The MMF at one end of the sensor is connected to a halogen light source (AvaLight-Hal-(S)-mini, 360-2500 nm), and the MMF at the other end is connected to a spectrometer (Ocean Optics Inc. USB 4000, 350-1100 nm). The output light signal is obtained by computer demodulation. It is important to note that the outer surface of the fiber is covered with a layer of deionized water with a refractive index of 1.333 to make a reference valley in the spectrum corresponding to a curvature of 0 m-1. In addition, we verified that the proposed sensor is temperature-insensitive and strain-insensitive, using devices such as the strain controller in Fig. 6(c) and the temperature controller in Fig. 6(d), which vary the axial strain of the fiber by stretching the fiber.
Fig. 6. (a) Experimental apparatus for testing curvature sensor, (b) curvature of the template, (c) strain controller, (d) temperature controller.
3.1 Influence of MOF length on shape of resonance valley
The length of the MOF is a crucial factor affecting the shape of the resonance valley. We performed three controlled experiments to determine the optimal MOF length. Figure 7 shows the three curves of the loss spectra at MOF lengths of 0.5 cm, 1 cm, and 1.5 cm, respectively. The refractive index of water outside the fiber at this time is 1.3330. As can be seen from Fig. 7, the resonance valley becomes deeper with the increase of MOF length. This is because as the MOF length increases, more light in the cladding occurs SPR with the silver film, which enhances the resonance intensity and reduces the power of the output light. It is well known that narrow, sharp resonance valleys have high detection accuracy and are ideal for sensing discrimination valleys. However, if the resonance valley is too deep, the amplitude of the resonance valley is limited, and the range of curvature to be measured is reduced. In summary, the MOF length of 1cm is the best choice.
3.2 Curvature sensing performance under different bending directions
Next, we explored the curvature sensing effect of the proposed sensor in four directions. The purpose of doing so is to verify the stability of the sensor for sensing applications in different bending directions. In some specific cases, instead of requiring the sensor to recognize the bending direction, we require the sensor to maintain the same sensing effect in different directions. As shown in Fig. 8, these four directions are forward bending (0° direction), backward bending (180° direction), downward bending (90° direction), and upward bending (270° direction), using the optical fiber as the coordinate axis. The whole process is done on the curvature template. In forward and backward bending, the curvature template is placed horizontally, and the curvature is calibrated according to the circle on the template. For upward and downward bending, the curvature template is placed vertically. During this process, we fixed the multimode jumpers at both ends to ensure that the fiber probe did not twist during the bending process and to ensure the accuracy of the measurement results. We bent the fabricated sensor in each of the four directions and recorded the change of its resonance valley with curvature.
Fig. 8. (a) Schematic diagram of bending direction, (b) bending forward, (c) bending downward, (d) bending upward, (e) bending backward.
The test results are shown in Fig. 9. These four transmission spectra demonstrate the variation regular of resonance valleys with curvature when the external refractive index is 1.333 for water. It can be seen from the figure that the changes in the resonance valley keep the same regularity under the four curvature directions. As the curvature of the sensor increases, the SPR is enhanced, and the resonance valley is gradually deepened. In addition, it is noteworthy that the resonance wavelength is stable at 580 nm. This means that the position of the resonance valley does not shift due to the curvature change. This is because the sensor utilizes the cladding mode as the excitation source. Since the cladding mode is divergent, the incident angle does not affect its resonant wavelength. However, the variation of the incident angle affects the leakage intensity of the evanescent wave and thus changes the resonant intensity of the SPR. Therefore, the proposed sensor has low requirements for light sources and spectrometers [23,24]. When the optical fiber probe is bent, and the curvature is 30m-1, the power loss ratio is 92%, so the upper limit of the curvature is selected as 30m-1. The power loss has converged to the maximum value.
Sensitivity and fitting coefficient are essential parameters to evaluate sensor performance. Therefore, we calculated the sensitivity of the proposed sensor under four bending directions and made a fit, as shown in Fig. 10. The following equation can calculate the sensitivity [8]:
3.3 Validation of temperature insensitivity
Subsequently, we conducted temperature sensing experiments on the proposed sensor. After connecting the optical path, the prepared sensor was placed on the temperature controller. The temperature was controlled to rise from 10°C to 50°C, and the data was recorded every 10°C. The test results are shown in Fig. 11, it can be seen that with the increase of temperature, a slight red-shift of the resonance valley occurs, but the depth of the resonance valley does not change. This phenomenon illustrates that temperature does not affect the curvature sensitivity of the sensor and exhibits temperature insensitivity. This is because the temperature does not change the leakage strength of the evanescent wave, so the depth of the resonant valley does not change. The red-shift occurs because the dielectric constant of silica (6 × 106 RIU/°C) is changed by temperature, resulting in a slight change in the resonance valley's location [25].
3.4 Validation of axial strain insensitivity
Then, we did axial strain experiments on the proposed sensor. Because a significant drawback of the curvature sensor based on fiber grating and fiber interferometer is that it is sensitive not only to radial strain (bending) but also to axial strain. So it is necessary to verify that the proposed sensor is insensitive to axial strain. We fixed the prepared sensor between two 2D adjusting bases of the strain controller, which has an accuracy of 0.01 mm. When the left 2D base is fixed, the strain can be applied to the sensor by moving the right 2D base longitudinally. The amount of strain change is determined as follows [26]:
3.5 Verification of sensor stability
Finally, another test was performed to verify the stability and repeatability of the proposed curvature sensor. We bent the sensor in the 0° direction with curvature ranging from 0 m-1 to 30 m-1 and recorded the depth of the resonance valley at different curvatures. We repeated the same procedure every half hour, five times in a row. As shown in Fig. 13, the experimental results show that the maximum resonance valley depth error is 0.0213 dB. The performance of our sensor is relatively stable and repeatable, which confirms its application value.
In the above analysis, our proposed sensor's sensitivity and detection range are better than those of the reported SPR curvature sensors, as shown in Table 2. This is thanks to the MOF used in this work because the light is transmitted mainly in its cladding region, and a slight curvature change can cause a change in the cladding mode. Therefore, the performance of the proposed sensor is superior. In addition, this sensor maintains the same sensing effect in any bending direction, which can play an essential role in some areas of curvature measurement. Moreover, the proposed sensor is insensitive to temperature, axial strain, and bending direction, as the SPR valley is sensitive only to changes in the nearby refractive index. However, it has an obvious drawback: the sensor must be covered with deionized water during the application, which limits its application to some extent. Therefore, we need to improve this problem in a subsequent study.
Table 2. Comparison of sensing performance of intensity-modulated curvature sensors
4. Conclusion
In this work, we experimentally demonstrate a highly sensitive SPR curvature sensor. A silver film is deposited on a MOF with thick cladding so that the transmission spectrum exhibits resonance valleys sensitive to bending changes. The bending curvature can be detected by monitoring the resonance valley's depth. Compared with the reported curvature sensors of the same type, the sensor has three advantages: (1) High sensitivity and a wide range of curvature to be measured. (2) The sensor is compact and does not require a cascade structure to solve the cross-sensitivity problem, which is in line with the trend of device miniaturization. (3) The sensor utilizes the cladding mode as the light source for the sensor and is simple to prepare without misalignment fusion. In addition, the experimental results show that the sensitivity reaches 0.18 dB/m-1 in the curvature range of 0-30 m-1, and the linearity is up to 0.995. This sensor is an important reference in curvature measurement without the demand for bending direction.
Funding
National Natural Science Foundation of China (42174162, 12074331); National Key Research and Development Program of China (2019YFB2204001).
Disclosures
The authors declare no conflict 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. A. A. Sipos and P. L. Várkonyi, “The longest soft robotic arm,” Int.J. Non-Linear Mech. 119, 103354 (2020). [CrossRef]
2. C. Leung, K. Wan, D. Inaudi, X. Bao, W. Habel, Z. Zhou, J. Ou, M. Ghandehari, H. Wu, and M. Imai, “Review: Optical fiber sensors for civil engineering applications,” Mater. Struct. 48(4), 871–906 (2015). [CrossRef]
3. Y. Zheng, Z. W. Zhu, W. Xiao, and Q. X. Deng, “Review of fiber optic sensors in geotechnical health monitoring,” Opt. Fiber Technol. 54, 102127 (2020). [CrossRef]
4. J. P. Moore and M. D. Rogge, “Shape sensing using multi-core fiber optic cable and parametric curve solutions,” Opt. Express 20(3), 2967–2973 (2012). [CrossRef]
5. W. Wang, “Magnetic Resonance–guided Active Catheter Tracking,” Magn. Reson. Imaging Clin. N. Am. 23(4), 579–589 (2015). [CrossRef]
6. X. Lian, G. Farrell, Q. Wu, W. Han, C. Shen, Y. Ma, and Y. Semenova, “Spectral Dependence of Transmission Losses in High-Index Polymer Coated No-Core Fibers,” J. Lightwave Technol. 38(22), 6352–6361 (2020). [CrossRef]
7. K. Tian, G. Farrell, X. Wang, Y. Xin, Y. Du, W. Yang, H. Liang, E. Lewis, and P. Wang, “High sensitivity temperature sensor based on singlemode-no-core-singlemode fibre structure and alcohol,” Sens. Actuators, A 284, 28–34 (2018). [CrossRef]
8. H. Wang, J. Luo, W. Dai, S. Chen, and H. Fu, “A Differential Intensity-Modulated Refractive Index Sensor Using a Droplet-Like Fiber Cascaded With FBGs,” J. Lightwave Technol. 40(9), 3098–3103 (2022). [CrossRef]
9. M. Jang, J. S. Kim, S. H. Um, S. Yang, and J. Kim, “Ultra-high curvature sensors for multi-bend structures using fiber Bragg gratings,” Opt. Express 27(3), 2074–2084 (2019). [CrossRef]
10. P. Geng, W. Zhang, S. Gao, H. Zhang, J. Li, S. Zhang, Z. Bai, and L. Wang, “Two-dimensional bending vector sensing based on spatial cascaded orthogonal long period fiber,” Opt. Express 20(27), 28557–28562 (2012). [CrossRef]
11. K. Ren, L. Ren, J. Liang, X. Kong, H. Ju, and Z. Wu, “Highly strain and bending sensitive microtapered long-period fiber gratings,” IEEE Photonics Technol. Lett. 29(13), 1085–1088 (2017). [CrossRef]
12. X. P. Zhang and W. Peng, “Bent fiber interferometer,” J. Lightwave Technol. 33(15), 3351–3356 (2015). [CrossRef]
13. R. Zhao, X. Shu, and P. Wang, “High-Performance Bending Sensor Based on Femtosecond Laser-Inscribed in-Fiber Mach–Zehnder Interferometer,” J. Lightwave Technol. 38(22), 6371–6378 (2020). [CrossRef]
14. Y. Wei, C. Liu, C. Liu, L. Li, L. Hu, Y. Zhang, X. Zhao, T. Jiang, R. Wang, and C. Shi, “Single-mode fiber curvature sensor based on SPR,” Appl. Opt. 61(15), 4620–4626 (2022). [CrossRef]
15. Z. Zhu, L. Liu, Z. Liu, and Y. Zhang, “Surface-plasmon-resonance-based optical-fiber temperature sensor with high sensitivity and high figure of merit,” Opt. Lett. 42(15), 2948–2951 (2017). [CrossRef]
16. K. Takagi, H. Sasaki, A. Seki, and K. Watanabe, “Surface plasmon resonances of a curved hetero-core optical fiber sensor,” Sens. Actuators, A 161(1-2), 1–5 (2010). [CrossRef]
17. C. Liu, J. Hu, Y. Wei, Y. Su, and X. Zhao, “Fiber cladding SPR bending sensor characterized by two parameters,” Chin. Phys. B 29(12), 120701 (2020). [CrossRef]
18. C. Caucheteur, T. Guo, and J. Albert, “Review of plasmonic fiber optic biochemical sensors: improving the limit of detection,” Anal. Bioanal. Chem. 407(14), 3883–3897 (2015). [CrossRef]
19. Z. Yin, X. Jing, and Heng Zhang, “Refractive Index and Temperature Sensor of Micro-groove Photonic Crystal Fiber Based on Surface Plasmon Resonance,” Plasmonics 17(4), 1731–1741 (2022). [CrossRef]
20. Y. N. Zhang, E. Siyu, R. Tao, Q. L. Wu, and B. Han, “Reflective SPR sensor for simultaneous measurement of nitrate concentration and temperature,” IEEE Trans. Instrum Meas. 68(11), 4566–4574 (2019). [CrossRef]
21. Y. Zhao, Q. L. Wu, and Y. N. Zhang, “Theoretical analysis of high-sensitive seawater temperature and salinity measurement based on C-type micro-structured fiber,” Sens. Actuators, B 258, 822–828 (2018). [CrossRef]
22. Y. D. Liu, H. L. Chen, H. Y. Li, S. Zhang, Z. G. Gao, Y. H. Feng, Y. Y. Zhang, and S. G. Li, “High-performance surface plasmon resonance refractometer based on a no-core fiber coated with a silver film,” J. Opt. Soc. Am. B 38(9), 2536–2542 (2021). [CrossRef]
23. C. Zhang, S. Xu, J. Zhao, H. Li, H. Bai, and C. Miao, “Intensity-modulated refractive index sensor with anti-light source fluctuation based on no-core fiber filter,” Opt. Laser Technol. 97, 358–363 (2017). [CrossRef]
24. Z. Jia, S. Pu, J. Rao, Y. Zhao, Y. Li, and T. Yao, “Temperature self-compensative all-fiber magnetic field sensing structure based on no-core fiber cascaded with fiber Bragg gratings,” Opt. Lasers Eng. 119, 26–29 (2019). [CrossRef]
25. Z.Y. Yin, X.L. Jing, Y.H. Feng, Z.G. Gao, B. Wu, and C.J. Wang, “Refractive index and temperature dual parameter sensor based on a twin-core photonic crystal fiber,” J. Phys. D: Appl. Phys. 55(15), 155108 (2022). [CrossRef]
26. Y. Liu, H. Chen, Q. Chen, B. Li, and S. Li, “Experimental Study on Dual-Parameter Sensing Based on Cascaded Sagnac Interferometers With Two PANDA Fibers,” J. Lightwave Technol. 40(9), 3090–3097 (2022). [CrossRef]
