2826
Letter
Vol. 40, No. 12 / June 15 2015 / Optics Letters
Twin-core fiber SPR sensor ZHIHAI LIU, YONG WEI, YU ZHANG,* YAXUN ZHANG, ENMING ZHAO, JUN YANG,
AND
LIBO YUAN
Key Lab of In-fiber Integrated Optics, Ministry Education of China, Harbin Engineering University, Harbin 150001, China *Corresponding author: [email protected] Received 15 April 2015; revised 8 May 2015; accepted 23 May 2015; posted 26 May 2015 (Doc. ID 238094); published 10 June 2015
We propose and demonstrate a novel surface plasmon resonance (SPR)-sensing approach by using the fundamental mode beam based on a twin-core fiber (TCF). Although normally in a fiber SPR sensor, a multimode fiber (MMF) has often been used to improve the coupling efficiency; for improving fiber SPR sensor sensitivity, single-mode beam is optimal. We provide a novel method to employ the single (fundamental)-mode beam to SPR sense based on the TCF. We grind the TCF tip to be a frustum wedge shape, and plate a 50-nm sensing gold film on the end face, and two 500-nm reflected gold films on the side faces of the wedge tip. By using this new configuration, we reduce the mode noise effectively and get a high testing sensitivity (the testing highest sensitivity reaches to 6463 nm/RIU). This SPR probe can be applied in a microfluidic chip and monitors the refractive index (RI) charges of the flow liquid in the microfluidic channel in real-time. The probe successfully monitors the refractive index of liquid ranged from 1.3333 to 1.3706, and the average sensitivity reaches to 5213 nm/RIU in the solution, which is much higher than most multimode SPR systems. © 2015 Optical Society of America OCIS codes: (060.2370) Fiber optics sensors; (060.4005) Microstructured fibers; (130.3990) Micro-optical devices. http://dx.doi.org/10.1364/OL.40.002826
Although the research of surface plasmon resonance (SPR) sensors has been developed for more than 20 years, nowadays it is still a research focus for the bio-sensing. Due to their high sensitivity without fluorescent labeling, SPR sensors have become a central tool for the study of bio-molecular interactions and have been widely applied in the detection of chemical and biological analytes [1–3]. As an alternative to the Kretschmann–Raether prism configurations, metal-coated optical fibers are proposed for the development of miniaturized SPR sensors [4–8]. The most common configuration previously studied is based on plastic cladding multimode fibers because it is easy to remove the cladding in the sensing region and reach the core. However, the high number of modes can excite several plasmon waves decreasing the accuracy of the sensor. For a typical multimode fiber, there exist hundreds of modes including all polarizations, 0146-9592/15/122826-04$15/0$15.00 © 2015 Optical Society of America
which implies that the light power coupled from the source is distributed over the total number of modes. For this reason, although coupling may occur between the surface plasmon (SP) mode and fiber modes at various wavelengths, over nearly the entire wavelength range when the phase matching condition is satisfied, and the overlap integral of the two modes does not vanish, only a fraction of the total launched power is coupled to the SP mode. The sensitivity of the MMF sensor based on the SPR is significantly reduced [7,8]. Therefore researchers develop few-mode [9] or single-mode fiber schemes (side polishing fiber [10,11], D-type fiber [12–15], chemically etched fiber [16–18] and tapered fiber [19–21]). However the fabrication of these single-mode optical fiber SPR sensor probes is rather complex and sophisticated. Besides, violation of the structural integrity of the fiber impairs the mechanical stability and reliability of the sensor. Consequently, a novel configuration is desired. A microstructure multi-core fiber is suitable to solve this problem. Here we employ a twin-core single-mode fiber. The twin cores in the fiber is geometry symmetric in the spatial structure, which is convenient for the sensing beam launching and receiving. The twin cores diameter is 3.8 μm, meeting the single-mode propagating condition during the light source waveband. Thus, only a fundamental mode beam can propagate in the fiber core, and the sensitivity of the sensor can be greatly improved. The twin-core fiber is grinded to be a frustum wedge tip configuration, and coated with a 50-nm gold film on the end face and two 500-nm reflected films on the side faces. This configuration has four advantages: First, we can employ the fundamental mode beam for the SPR sensing; Second, we can realize the light source incident angle adjustment by changing the fiber grinding angle; Third, we collect the incident light and the reflected light on the same end of the fiber leaving the sensor probe tip on the other end, which is convenient for sensing and testing in a narrow space; Fourth, the diameter of the twin-core fiber is 125 μm, which is easy for employment in a normal fiber system. The fiber grind and polish method is employed to fabricate the sensor probe. The schematic diagram of the twin-core optical fiber probe with a frustum wedge tip configuration is showed in the Fig. 1, where the grinding angle α is 37.5°, and the resonance angle θ 75°. The distance between the two cores is 53 μm, and the core diameter is 3.8 μm. The NA of the core is 0.12, and the single mode cut-off wavelength is 633 nm. The thickness of the reflected films on the side face (Surface A and B, seen Fig. 1) of the fiber tip is 500 nm, while
Letter
Vol. 40, No. 12 / June 15 2015 / Optics Letters
h1=450nm
(a)
Fig. 1. Sketch diagram of the SPR twin-core fiber frustum wedge tip.
h2=450nm
h1=450nm
(b)
2827
h1=500nm h2=50nm
(c)
Fig. 3. Sketch diagram of the gold film-coating fabrication (a) firsttime gold film coating, where the thickness of the gold film on the reflected surface A, B h1 450 nm, and the thickness of the gold film on the sensing surface h2 450 nm; (b) removing the gold film on the sensing surface while h1 450 nm and h2 0 nm; (c) secondtime gold film coating, while h1 500 nm and h2 50 nm.
the thickness of the SPR film on the end face (Sensing surface, seen Fig. 1) is 50 nm. According to the Fig. 1, the incident laser is launched into the core 1 and then is totally reflected on the Surface A, the reflected beam then propagates to the sensor region to generate a SPR wave when it is totally internal reflected on the Sensing surface, and finally the reflected beam, which brings with the sensor information, is totally reflected by the Surface B and then gets back through the core 2. Therefore, we realize the Kretchmann prism configuration by using this twin-core fiber with a 50-nm sensing gold film. A grinding and polishing technique makes use of the fabrication of the frustum wedge tip, as show in Fig. 2. Two dc motors are used to drive the abrasive disk and rotate the fiber simultaneously. The grind angle α of the twin-core fiber is defined as the angle between the rotational axis of the fiber and vertical direction. The twin-core fiber end in processing by using this grinding and polishing method is shown in the inset of Fig. 2. Here we employ the 8000-grit paper to grind the fiber and 12000-grit paper to polish the grinded surface. And then we place the grinded twin-core fiber tip in the plasma cleaner (PDC-MG) for 8 mins to clean the grinded surface for gold film coating. A vacuum plasma-sputtering method is employed to plate the gold film on the twin-core fiber tip. We develop a three-step method to plate: First, we plate the 450-nm gold film on the reflected surface A, B, and sensing surface by using the plasma sputtering apparatus(JS-1600, HTCY) [see Fig. 3(a)]; Second, we remove the gold film on the sensing surface by using the fiber end-grinding technology while keeping the gold films on the reflected surface A and B [see Fig. 3(b)]; Third, we
put the fiber into the vacuum chamber of plasma-sputtering apparatus again, and plate another 50-nm gold film on the reflected surface A, B, and the sensing surface [see Fig. 3(c)]. Therefore, we realize the 50-nm gold film coating on the end face (sensing surface) and two 500-nm reflected films on the side faces (reflected surface A and B). The gold film thickness is measured by the threedimensional morphology analyzer (NewView7200, Zygo). The gold film coated on the fiber surface is scratched as a crossed shape. The flatness of gold film surface and the depth of the groove are observed by the morphology analyzer (seen Fig. 4). The depth of the groove indicates the thickness of the coated gold film. Here the solution of the morphology analyzer is 0.1 nm. We simulate and calculate the SPR reflected light intensity of the fiber probe [seen Fig. 5(a)] with the condition of the resonance angle θ 75°. According to the Fig. 5(a), different gold film thinness causes different reflected spectrum valley. In order to get a deep valley, the film thinness should be 50– 60 nm. It is why we plate 50-nm gold film on the sensing surface of the fiber tip. Figure 5(b) shows the simulated results that different spectrum valley generates from different index, with the simulated condition of resonance angle θ 75°, the gold film thinness is 50 nm. The simulated results show that the SPR sensor with the sensing film of 50 nm and resonance angle of 75° can sense and test the refractive index changes of the solution. The simulated parameters are: the refractive index of the fiber core is 1.4634, the thickness of the gold film is 50 nm, and the dielectric constant of the gold film is obtained from the Ref. [22].
Fig. 2. Schematic diagram of the fiber grinding technology. (a) Image of the twin-core fiber being grinded.
Fig. 4. Film thickness measured results by using the three-dimensional morphology analyzer (a) the three-dimensional test result of the film surface with a groove and (b) the test result of the groove depth, which equals the film thickness.
2828
Vol. 40, No. 12 / June 15 2015 / Optics Letters
Letter The sensing experimental setup is shown in Fig. 6. A super continuum light source (SuperK compact, NKT Photonics) whose spectrum wide is 500–2400 nm, is launched into the core 1 of the twin-core fiber by using a lens transform system [seen Fig. 6(b), where we place the collimating lens in front of two single-mode fibers while placing the microscope objective lens (the magnification of 25) in front of the twin-core fiber. The light power is launching into core 1 by adjusting the distance and position of the lens in the transform system], and the reflected beam is received by an optical spectrum analyzer (AQ6370C, Yokokawa). We employ a programmable microinjection pumper (LSP01-1A, LongerPump) to inject the Glycerine-aqueous solution into the micro-fluidic channel. The Glycerine-aqueous solution index is calibrated by the Abbe refractometer (GDA-2S, Gold). We employ the smoothing and normalized method to processing data. For the distilled water (n 1.33) testing results, we employ the normal smooth and normalized algorithm in the Matlab Software to processing the light source spectrum data [see Fig. 7(a), magenta curve]; and similarly, we smoothing
Fig. 5. (a) SPR spectrum corresponding to different film thickness; (b) different index causes different spectrum valley with the simulated condition of resonance angle θ 75°; the gold film thinness is 50 nm.
We fabricate a micro fluidic device to demonstrate the SPR sensor advantage and application on the micro-fluidic chip. Our micro-fluidic device is fabricated by using the standard soft lithography and Polydimethylsiloxane (PDMS) molding techniques [23]. The inlet and outlet of the fluid is realized by using two stainless-steel joints. The sensor probe is fixed in the microfluid channel by using a fiber ceramic casing pipe, whose inter diameter is 125 μm [seen Fig. 6(c)].
Fig. 6. (a) The experiment setup sketch diagram of the two-core fiber SPR sensor; (b) the sketch diagram of the lens transform system; (c) the images the micro-fluidic device, where the inset image shows the details of the micro-fluidic channel and the fiber probe.
Fig. 7. Response of reflection spectrum of a two-core fiber SPR probe: (a) SPR spectrum in the aqueous solution; (b) SPR spectrum response to different solution refractive index; (c) relation between the resonance wavelength and the refractive index, where the fit goodness coefficient is R 99.63%.
Letter and normalize the testing SPR reflected spectrum, and then subtract the reflection spectrum with the light source spectrum [see Fig. 7(a), blue curve], from which we can find the resonance valley in the wavelength of 718 nm. By using this method, we obtain the SPR reflected spectrum of different index testing results [see Fig. 7(b), n 1.3333–1.3706]. According to the test results, the SPR spectrum produced by twin-core fiber SPR microsensor has a sensitive response to the changes of the solution refractive index. The relation between the resonance wavelength shift and the refractive index changing is linear [where the fit goodness coefficient R 99.63%, seen Fig. 7(c)]. The mean sensitivity can reach to 5213 nm/RIU during the solution refractive index range from 1.3333 to 1.3706, and the highest sensitivity reaches to 6463 nm/RIU. Compared with the theoretical and testing results in the Figs. 5(b) and 7(b), the discrepancy lies that the sensitivity is relatively lower, and the dips are getting shallow in the experimental results. In our opinion, the main reason is the gold film quality. The gold-film thickness difference and the roughness make the SPR resonance happen at different wavelengths, which around the center resonance wavelength, cause the testing spectrum sensitivity reducing and the dips getting shallow. Besides that, the errors generated from the fiber grind processing cause the sensing surface to be not smooth and flat, may be another reason. The unsmooth or inflating surface may make the SPR resonance wavelength shift, which also causes the testing spectrum sensitivity reducing and the dips getting shallow. In summary we develop a novel twin-core fiber SPR microsensor with a frustum wedge tip probe that can be applied in the micro-fluidic chip researching fields. We effectively solve the problems that the normal multimode fiber SPR sensors may meet, such as, it is hard to match the fiber SPR resonance angle, and it is difficult to improve the sensitivity. Besides that, we can change the resonance angle θ by changing the grinding angle α of the fiber tip. This twin-core fiber SPR micro sensor collects the incident light and reflected light on the same end of the fiber, and the other end of the fiber is fabricated to be a probe, which is used to test and sense. The advantage of this sensor configuration is that it is easy to get into a narrow space to sense. Besides that, the sensing test region is on the fiber tip, which is convenient for testing the rare and precious sample; therefore, it improves the utilization ratio of samples, reduced the experiment cost. When the grinding angle α is 37.5° and SPR resonance θ is 75°, the mean sensitivity can reach to 5213 nm/RIU with the solution refractive index range is from 1.3333 to 1.3706. If we employ more cores in a multi-core optical fiber, such as fourcore, six-core even annular-core to the SPR sensing, we may realize distributed SPR micro-sensor on the fiber end face. Although the testing sensitivity is not as high as the professional refractive index sensors based on the photonic crystal fibers [24–28], we aim to provide new ideas for the development of the optical fiber SPR sensor based on the microstructure
Vol. 40, No. 12 / June 15 2015 / Optics Letters
2829
multi-core fiber, which may be meaningful for the SPR sensor research. 111 project (B13015); Fundamental Research Funds for Harbin Engineering University of China; National Natural Science Foundation of China (NSFC) (11204047, 61227013, 61275087 and 61377085); Postdoctororal Science Foundation Fund of China (2014M550181); the natural science foundation of Heilongjiang Province of China (F201338). REFERENCES 1. X. Guo, J. Biophotonics 5, 483 (2012). 2. P. Zijlstra, P. M. R. Paulo, and M. Orrit, Nat. Nanotechnol. 7, 379 (2012). 3. G. Xiao and W. J. Bock, eds. Photonic Sensing: Principles and Applications for Safety and Security Monitoring (Wiley, 2012). 4. R. C. Jorguenson and S. S. Yee, Sens. Actuators B 12, 213 (1993). 5. M. Mitsushio, S. Higashi, and M. Higo, Sens. Actuators A 111, 252 (2004). 6. W. B. Lin, N. Jaffrezic-Renault, A. Gagnaire, and H. Gagnaire, Sens. Actuators B 84, 198 (2000). 7. M. Skorobogatiy and A. Kabashin, Opt. Express 14, 8419 (2006). 8. A. Hassani and M. Skorobogatiy, Opt. Express 14, 11616 (2006). 9. H. S. Jang, K. N. Park, C. D. Kang, J. P. Kim, S. J. Sim, and K. S. Lee, Opt. Commun. 282, 2827 (2009). 10. M. Piliarik, J. Homola, Z. Manikova, and J. Ctyroky, Sens. Actuators B 90, 236 (2003). 11. W. J. H. Bender, R. E. Dessy, M. S. Miller, and R. O. Claus, Anal. Chem. 66, 963 (1994). 12. M. H. Chiu and C. H. Shih, Sens. Actuators B 131, 596 (2008). 13. M. H. Chiu, C. H. Shih, and M. H. Chi, Sens. Actuators B 123, 1120 (2007). 14. M. H. Chiu, S. F. Wang, and R. S. Chang, Opt. Lett. 30, 233 (2005). 15. S. F. Wang, M. H. Chiu, and R. S. Chang, Opt. Eng. 44, 030502 (2005). 16. L. Coelho, J. M. M. M. de Almeida, J. L. Santos, R. A. S. Ferreira, P. S. Andre, and D. Viegas, Plasmonics 10, 319 (2015). 17. L. C. C. Coelho, J. M. M. M. de Almeida, H. Moayyed, J. L. Santos, and D. Viegas, J. Lightwave Technol. 33, 432 (2015). 18. K. Kurihara, H. Ohkawa, Y. Iwasaki, O. Niwa, T. Tobita, and K. Suzuki, Anal. Chim. Acta 523, 165 (2004). 19. A. Diez, M. V. Andres, and J. L. Cruz, Sens. Actuators B 73, 95 (2001). 20. A. J. C. Tubb, F. P. Payneay, R. B. Millington, and C. R. Lowe, Sens. Actuators B 41, 71 (1997). 21. J. Villatoro, D. Monzon-Hernandez, and E. Mejia, Appl. Opt. 42, 2278 (2003). 22. P. B. Johnson and R. W. Christy, Phys. Rev. B 6, 4370 (1972). 23. S. M. Kim, M. A. Burns, and E. F. Hasselbrink, Chip Anal. Chem. 78, 4779 (2006). 24. D. K. C. Wu, B. T. Kuhlmey, and B. J. Eggleton, Opt. Lett. 34, 322 (2009). 25. W. Yuan, G. E. Town, and O. Bang, IEEE Sens. J. 10, 1192 (2010). 26. P. Torres, E. Reyes-Vera, A. Díez, and M. V. Andrés, Opt. Lett. 39, 1593 (2014). 27. C. Markos, W. Yuan, K. Vlachos, G. E. Town, and O. Bang, Opt. Express 19, 7790 (2011). 28. B. Sun, M. Y. Chen, Y. K. Zhang, J. Yang, J. Yao, and H. X. Cui, Opt. Express 19, 4091 (2011).
Letter
Vol. 40, No. 12 / June 15 2015 / Optics Letters
Twin-core fiber SPR sensor ZHIHAI LIU, YONG WEI, YU ZHANG,* YAXUN ZHANG, ENMING ZHAO, JUN YANG,
AND
LIBO YUAN
Key Lab of In-fiber Integrated Optics, Ministry Education of China, Harbin Engineering University, Harbin 150001, China *Corresponding author: [email protected] Received 15 April 2015; revised 8 May 2015; accepted 23 May 2015; posted 26 May 2015 (Doc. ID 238094); published 10 June 2015
We propose and demonstrate a novel surface plasmon resonance (SPR)-sensing approach by using the fundamental mode beam based on a twin-core fiber (TCF). Although normally in a fiber SPR sensor, a multimode fiber (MMF) has often been used to improve the coupling efficiency; for improving fiber SPR sensor sensitivity, single-mode beam is optimal. We provide a novel method to employ the single (fundamental)-mode beam to SPR sense based on the TCF. We grind the TCF tip to be a frustum wedge shape, and plate a 50-nm sensing gold film on the end face, and two 500-nm reflected gold films on the side faces of the wedge tip. By using this new configuration, we reduce the mode noise effectively and get a high testing sensitivity (the testing highest sensitivity reaches to 6463 nm/RIU). This SPR probe can be applied in a microfluidic chip and monitors the refractive index (RI) charges of the flow liquid in the microfluidic channel in real-time. The probe successfully monitors the refractive index of liquid ranged from 1.3333 to 1.3706, and the average sensitivity reaches to 5213 nm/RIU in the solution, which is much higher than most multimode SPR systems. © 2015 Optical Society of America OCIS codes: (060.2370) Fiber optics sensors; (060.4005) Microstructured fibers; (130.3990) Micro-optical devices. http://dx.doi.org/10.1364/OL.40.002826
Although the research of surface plasmon resonance (SPR) sensors has been developed for more than 20 years, nowadays it is still a research focus for the bio-sensing. Due to their high sensitivity without fluorescent labeling, SPR sensors have become a central tool for the study of bio-molecular interactions and have been widely applied in the detection of chemical and biological analytes [1–3]. As an alternative to the Kretschmann–Raether prism configurations, metal-coated optical fibers are proposed for the development of miniaturized SPR sensors [4–8]. The most common configuration previously studied is based on plastic cladding multimode fibers because it is easy to remove the cladding in the sensing region and reach the core. However, the high number of modes can excite several plasmon waves decreasing the accuracy of the sensor. For a typical multimode fiber, there exist hundreds of modes including all polarizations, 0146-9592/15/122826-04$15/0$15.00 © 2015 Optical Society of America
which implies that the light power coupled from the source is distributed over the total number of modes. For this reason, although coupling may occur between the surface plasmon (SP) mode and fiber modes at various wavelengths, over nearly the entire wavelength range when the phase matching condition is satisfied, and the overlap integral of the two modes does not vanish, only a fraction of the total launched power is coupled to the SP mode. The sensitivity of the MMF sensor based on the SPR is significantly reduced [7,8]. Therefore researchers develop few-mode [9] or single-mode fiber schemes (side polishing fiber [10,11], D-type fiber [12–15], chemically etched fiber [16–18] and tapered fiber [19–21]). However the fabrication of these single-mode optical fiber SPR sensor probes is rather complex and sophisticated. Besides, violation of the structural integrity of the fiber impairs the mechanical stability and reliability of the sensor. Consequently, a novel configuration is desired. A microstructure multi-core fiber is suitable to solve this problem. Here we employ a twin-core single-mode fiber. The twin cores in the fiber is geometry symmetric in the spatial structure, which is convenient for the sensing beam launching and receiving. The twin cores diameter is 3.8 μm, meeting the single-mode propagating condition during the light source waveband. Thus, only a fundamental mode beam can propagate in the fiber core, and the sensitivity of the sensor can be greatly improved. The twin-core fiber is grinded to be a frustum wedge tip configuration, and coated with a 50-nm gold film on the end face and two 500-nm reflected films on the side faces. This configuration has four advantages: First, we can employ the fundamental mode beam for the SPR sensing; Second, we can realize the light source incident angle adjustment by changing the fiber grinding angle; Third, we collect the incident light and the reflected light on the same end of the fiber leaving the sensor probe tip on the other end, which is convenient for sensing and testing in a narrow space; Fourth, the diameter of the twin-core fiber is 125 μm, which is easy for employment in a normal fiber system. The fiber grind and polish method is employed to fabricate the sensor probe. The schematic diagram of the twin-core optical fiber probe with a frustum wedge tip configuration is showed in the Fig. 1, where the grinding angle α is 37.5°, and the resonance angle θ 75°. The distance between the two cores is 53 μm, and the core diameter is 3.8 μm. The NA of the core is 0.12, and the single mode cut-off wavelength is 633 nm. The thickness of the reflected films on the side face (Surface A and B, seen Fig. 1) of the fiber tip is 500 nm, while
Letter
Vol. 40, No. 12 / June 15 2015 / Optics Letters
h1=450nm
(a)
Fig. 1. Sketch diagram of the SPR twin-core fiber frustum wedge tip.
h2=450nm
h1=450nm
(b)
2827
h1=500nm h2=50nm
(c)
Fig. 3. Sketch diagram of the gold film-coating fabrication (a) firsttime gold film coating, where the thickness of the gold film on the reflected surface A, B h1 450 nm, and the thickness of the gold film on the sensing surface h2 450 nm; (b) removing the gold film on the sensing surface while h1 450 nm and h2 0 nm; (c) secondtime gold film coating, while h1 500 nm and h2 50 nm.
the thickness of the SPR film on the end face (Sensing surface, seen Fig. 1) is 50 nm. According to the Fig. 1, the incident laser is launched into the core 1 and then is totally reflected on the Surface A, the reflected beam then propagates to the sensor region to generate a SPR wave when it is totally internal reflected on the Sensing surface, and finally the reflected beam, which brings with the sensor information, is totally reflected by the Surface B and then gets back through the core 2. Therefore, we realize the Kretchmann prism configuration by using this twin-core fiber with a 50-nm sensing gold film. A grinding and polishing technique makes use of the fabrication of the frustum wedge tip, as show in Fig. 2. Two dc motors are used to drive the abrasive disk and rotate the fiber simultaneously. The grind angle α of the twin-core fiber is defined as the angle between the rotational axis of the fiber and vertical direction. The twin-core fiber end in processing by using this grinding and polishing method is shown in the inset of Fig. 2. Here we employ the 8000-grit paper to grind the fiber and 12000-grit paper to polish the grinded surface. And then we place the grinded twin-core fiber tip in the plasma cleaner (PDC-MG) for 8 mins to clean the grinded surface for gold film coating. A vacuum plasma-sputtering method is employed to plate the gold film on the twin-core fiber tip. We develop a three-step method to plate: First, we plate the 450-nm gold film on the reflected surface A, B, and sensing surface by using the plasma sputtering apparatus(JS-1600, HTCY) [see Fig. 3(a)]; Second, we remove the gold film on the sensing surface by using the fiber end-grinding technology while keeping the gold films on the reflected surface A and B [see Fig. 3(b)]; Third, we
put the fiber into the vacuum chamber of plasma-sputtering apparatus again, and plate another 50-nm gold film on the reflected surface A, B, and the sensing surface [see Fig. 3(c)]. Therefore, we realize the 50-nm gold film coating on the end face (sensing surface) and two 500-nm reflected films on the side faces (reflected surface A and B). The gold film thickness is measured by the threedimensional morphology analyzer (NewView7200, Zygo). The gold film coated on the fiber surface is scratched as a crossed shape. The flatness of gold film surface and the depth of the groove are observed by the morphology analyzer (seen Fig. 4). The depth of the groove indicates the thickness of the coated gold film. Here the solution of the morphology analyzer is 0.1 nm. We simulate and calculate the SPR reflected light intensity of the fiber probe [seen Fig. 5(a)] with the condition of the resonance angle θ 75°. According to the Fig. 5(a), different gold film thinness causes different reflected spectrum valley. In order to get a deep valley, the film thinness should be 50– 60 nm. It is why we plate 50-nm gold film on the sensing surface of the fiber tip. Figure 5(b) shows the simulated results that different spectrum valley generates from different index, with the simulated condition of resonance angle θ 75°, the gold film thinness is 50 nm. The simulated results show that the SPR sensor with the sensing film of 50 nm and resonance angle of 75° can sense and test the refractive index changes of the solution. The simulated parameters are: the refractive index of the fiber core is 1.4634, the thickness of the gold film is 50 nm, and the dielectric constant of the gold film is obtained from the Ref. [22].
Fig. 2. Schematic diagram of the fiber grinding technology. (a) Image of the twin-core fiber being grinded.
Fig. 4. Film thickness measured results by using the three-dimensional morphology analyzer (a) the three-dimensional test result of the film surface with a groove and (b) the test result of the groove depth, which equals the film thickness.
2828
Vol. 40, No. 12 / June 15 2015 / Optics Letters
Letter The sensing experimental setup is shown in Fig. 6. A super continuum light source (SuperK compact, NKT Photonics) whose spectrum wide is 500–2400 nm, is launched into the core 1 of the twin-core fiber by using a lens transform system [seen Fig. 6(b), where we place the collimating lens in front of two single-mode fibers while placing the microscope objective lens (the magnification of 25) in front of the twin-core fiber. The light power is launching into core 1 by adjusting the distance and position of the lens in the transform system], and the reflected beam is received by an optical spectrum analyzer (AQ6370C, Yokokawa). We employ a programmable microinjection pumper (LSP01-1A, LongerPump) to inject the Glycerine-aqueous solution into the micro-fluidic channel. The Glycerine-aqueous solution index is calibrated by the Abbe refractometer (GDA-2S, Gold). We employ the smoothing and normalized method to processing data. For the distilled water (n 1.33) testing results, we employ the normal smooth and normalized algorithm in the Matlab Software to processing the light source spectrum data [see Fig. 7(a), magenta curve]; and similarly, we smoothing
Fig. 5. (a) SPR spectrum corresponding to different film thickness; (b) different index causes different spectrum valley with the simulated condition of resonance angle θ 75°; the gold film thinness is 50 nm.
We fabricate a micro fluidic device to demonstrate the SPR sensor advantage and application on the micro-fluidic chip. Our micro-fluidic device is fabricated by using the standard soft lithography and Polydimethylsiloxane (PDMS) molding techniques [23]. The inlet and outlet of the fluid is realized by using two stainless-steel joints. The sensor probe is fixed in the microfluid channel by using a fiber ceramic casing pipe, whose inter diameter is 125 μm [seen Fig. 6(c)].
Fig. 6. (a) The experiment setup sketch diagram of the two-core fiber SPR sensor; (b) the sketch diagram of the lens transform system; (c) the images the micro-fluidic device, where the inset image shows the details of the micro-fluidic channel and the fiber probe.
Fig. 7. Response of reflection spectrum of a two-core fiber SPR probe: (a) SPR spectrum in the aqueous solution; (b) SPR spectrum response to different solution refractive index; (c) relation between the resonance wavelength and the refractive index, where the fit goodness coefficient is R 99.63%.
Letter and normalize the testing SPR reflected spectrum, and then subtract the reflection spectrum with the light source spectrum [see Fig. 7(a), blue curve], from which we can find the resonance valley in the wavelength of 718 nm. By using this method, we obtain the SPR reflected spectrum of different index testing results [see Fig. 7(b), n 1.3333–1.3706]. According to the test results, the SPR spectrum produced by twin-core fiber SPR microsensor has a sensitive response to the changes of the solution refractive index. The relation between the resonance wavelength shift and the refractive index changing is linear [where the fit goodness coefficient R 99.63%, seen Fig. 7(c)]. The mean sensitivity can reach to 5213 nm/RIU during the solution refractive index range from 1.3333 to 1.3706, and the highest sensitivity reaches to 6463 nm/RIU. Compared with the theoretical and testing results in the Figs. 5(b) and 7(b), the discrepancy lies that the sensitivity is relatively lower, and the dips are getting shallow in the experimental results. In our opinion, the main reason is the gold film quality. The gold-film thickness difference and the roughness make the SPR resonance happen at different wavelengths, which around the center resonance wavelength, cause the testing spectrum sensitivity reducing and the dips getting shallow. Besides that, the errors generated from the fiber grind processing cause the sensing surface to be not smooth and flat, may be another reason. The unsmooth or inflating surface may make the SPR resonance wavelength shift, which also causes the testing spectrum sensitivity reducing and the dips getting shallow. In summary we develop a novel twin-core fiber SPR microsensor with a frustum wedge tip probe that can be applied in the micro-fluidic chip researching fields. We effectively solve the problems that the normal multimode fiber SPR sensors may meet, such as, it is hard to match the fiber SPR resonance angle, and it is difficult to improve the sensitivity. Besides that, we can change the resonance angle θ by changing the grinding angle α of the fiber tip. This twin-core fiber SPR micro sensor collects the incident light and reflected light on the same end of the fiber, and the other end of the fiber is fabricated to be a probe, which is used to test and sense. The advantage of this sensor configuration is that it is easy to get into a narrow space to sense. Besides that, the sensing test region is on the fiber tip, which is convenient for testing the rare and precious sample; therefore, it improves the utilization ratio of samples, reduced the experiment cost. When the grinding angle α is 37.5° and SPR resonance θ is 75°, the mean sensitivity can reach to 5213 nm/RIU with the solution refractive index range is from 1.3333 to 1.3706. If we employ more cores in a multi-core optical fiber, such as fourcore, six-core even annular-core to the SPR sensing, we may realize distributed SPR micro-sensor on the fiber end face. Although the testing sensitivity is not as high as the professional refractive index sensors based on the photonic crystal fibers [24–28], we aim to provide new ideas for the development of the optical fiber SPR sensor based on the microstructure
Vol. 40, No. 12 / June 15 2015 / Optics Letters
2829
multi-core fiber, which may be meaningful for the SPR sensor research. 111 project (B13015); Fundamental Research Funds for Harbin Engineering University of China; National Natural Science Foundation of China (NSFC) (11204047, 61227013, 61275087 and 61377085); Postdoctororal Science Foundation Fund of China (2014M550181); the natural science foundation of Heilongjiang Province of China (F201338). REFERENCES 1. X. Guo, J. Biophotonics 5, 483 (2012). 2. P. Zijlstra, P. M. R. Paulo, and M. Orrit, Nat. Nanotechnol. 7, 379 (2012). 3. G. Xiao and W. J. Bock, eds. Photonic Sensing: Principles and Applications for Safety and Security Monitoring (Wiley, 2012). 4. R. C. Jorguenson and S. S. Yee, Sens. Actuators B 12, 213 (1993). 5. M. Mitsushio, S. Higashi, and M. Higo, Sens. Actuators A 111, 252 (2004). 6. W. B. Lin, N. Jaffrezic-Renault, A. Gagnaire, and H. Gagnaire, Sens. Actuators B 84, 198 (2000). 7. M. Skorobogatiy and A. Kabashin, Opt. Express 14, 8419 (2006). 8. A. Hassani and M. Skorobogatiy, Opt. Express 14, 11616 (2006). 9. H. S. Jang, K. N. Park, C. D. Kang, J. P. Kim, S. J. Sim, and K. S. Lee, Opt. Commun. 282, 2827 (2009). 10. M. Piliarik, J. Homola, Z. Manikova, and J. Ctyroky, Sens. Actuators B 90, 236 (2003). 11. W. J. H. Bender, R. E. Dessy, M. S. Miller, and R. O. Claus, Anal. Chem. 66, 963 (1994). 12. M. H. Chiu and C. H. Shih, Sens. Actuators B 131, 596 (2008). 13. M. H. Chiu, C. H. Shih, and M. H. Chi, Sens. Actuators B 123, 1120 (2007). 14. M. H. Chiu, S. F. Wang, and R. S. Chang, Opt. Lett. 30, 233 (2005). 15. S. F. Wang, M. H. Chiu, and R. S. Chang, Opt. Eng. 44, 030502 (2005). 16. L. Coelho, J. M. M. M. de Almeida, J. L. Santos, R. A. S. Ferreira, P. S. Andre, and D. Viegas, Plasmonics 10, 319 (2015). 17. L. C. C. Coelho, J. M. M. M. de Almeida, H. Moayyed, J. L. Santos, and D. Viegas, J. Lightwave Technol. 33, 432 (2015). 18. K. Kurihara, H. Ohkawa, Y. Iwasaki, O. Niwa, T. Tobita, and K. Suzuki, Anal. Chim. Acta 523, 165 (2004). 19. A. Diez, M. V. Andres, and J. L. Cruz, Sens. Actuators B 73, 95 (2001). 20. A. J. C. Tubb, F. P. Payneay, R. B. Millington, and C. R. Lowe, Sens. Actuators B 41, 71 (1997). 21. J. Villatoro, D. Monzon-Hernandez, and E. Mejia, Appl. Opt. 42, 2278 (2003). 22. P. B. Johnson and R. W. Christy, Phys. Rev. B 6, 4370 (1972). 23. S. M. Kim, M. A. Burns, and E. F. Hasselbrink, Chip Anal. Chem. 78, 4779 (2006). 24. D. K. C. Wu, B. T. Kuhlmey, and B. J. Eggleton, Opt. Lett. 34, 322 (2009). 25. W. Yuan, G. E. Town, and O. Bang, IEEE Sens. J. 10, 1192 (2010). 26. P. Torres, E. Reyes-Vera, A. Díez, and M. V. Andrés, Opt. Lett. 39, 1593 (2014). 27. C. Markos, W. Yuan, K. Vlachos, G. E. Town, and O. Bang, Opt. Express 19, 7790 (2011). 28. B. Sun, M. Y. Chen, Y. K. Zhang, J. Yang, J. Yao, and H. X. Cui, Opt. Express 19, 4091 (2011).
