Miniature and robust optical fiber in-line Mach ... - OSA Publishing

Letter

Vol. 40, No. 15 / August 1 2015 / Optics Letters

3516

Miniature and robust optical fiber in-line Mach–Zehnder interferometer based on a hollow ellipsoid H. GONG,1,2 D. N. WANG,1,2,3,4,* B. XU,1,2 K. NI,1,2 H. LIU,1,2

AND

C. L. ZHAO1

1

College of Optical and Electronic Technology, China Jiliang University, Xiasha Higher Education Park, Hangzhou, China Department of Electrical Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China 3 The Hong Kong Polytechnic University Shenzhen Research Institute, Shenzhen, China 4 School of Electrical, Electronic and Information Engineering, Hubei Polytechnic University, Huangshi, China *Corresponding author: [email protected] 2

Received 4 May 2015; revised 8 July 2015; accepted 8 July 2015; posted 9 July 2015 (Doc. ID 240205); published 23 July 2015

An optical fiber in-line Mach–Zehnder interferometer based on a hollow ellipsoid fabricated by femtosecond laser micromachining and fusion-splicing technique is demonstrated. The surface of the hollow ellipsoid acts as an internal mirror that can be utilized for the construction of an interferometer. Such an interferometer device is miniature and robust and can perform external refractive index, curvature, and hightemperature sensing in a mutually independent way, and hence a simultaneous multiple parameter measurement capability can be readily achieved. © 2015 Optical Society of America OCIS

codes:

(230.3990)

Micro-optical

devices;

(120.3180)

Interferometry; (230.1150) All-optical devices. http://dx.doi.org/10.1364/OL.40.003516

The optical fiber in-line Mach–Zehnder interferometer (MZI) is an attractive photonic-sensing device because of its merits of all-fiber structure, convenient operation, and good sensitivity for many physical parameters, such as temperature, curvature, and refractive index (RI). Various types of fiber in-line MZI structures have been proposed, including those based on fiber tapering [1,2], lateral offset splicing [3], peanut-shape pair [4,5], fiber bitaper splicing [6,7], long-period fiber grating (LPG) pair [8,9], and mismatched fiber core [10,11], etc. The fibers employed include single-mode fiber (SMF) and other special fibers, such as photonic crystal fiber (PCF), which has the merit of low sensitivity to temperature, but is much more expensive than conventional SMF. However, the sizes of above-mentioned structures are usually large, close to 10 mm. Recently, by use of femtosecond (fs) laser-micromachining technique, fiber in-line MZIs with small size have been developed [12,13]. For instance, a MZI based on an inner aircavity adjacent to the fiber core having a sensor head size of 35 μm is used for high-temperature sensing [14]. Another 0146-9592/15/153516-04$15/0$15.00 © 2015 Optical Society of America

MZI formed by a pair of hollow spheres and used for stain and refractive index sensing exhibits a size of less than 200 μm [15]. A microfiber in-line MZI formed by initially creating a hollow sphere and followed by tapering has been reported, which has a diameter of down to 22 μm and size of up to 1250 μm [16,17]. The fabrication of a hollow sphere adjacent to the fiber core or a pair of nearly identical hollow spheres is difficult and time consuming and the device based on hollow sphere within microfiber is rather fragile. In this Letter, we propose and demonstrate a miniature and robust fiber in-line MZI based on a hollow ellipsoid with major axis length of ∼50 μm. The interesting feature of the device is that the surface of the hollow ellipsoid and the fiber claddingair interface act as the internal mirrors that reflect the light beam traveling in the fiber core to the air-cladding interface, and then back into the fiber core, thus forming one arm of the interferometer. The device can be used for simultaneous measurement of external RI and temperature, or curvature and temperature, in the temperature range from 24°C to 60°C. Moreover, the device exhibits excellent thermal stability at high temperature up to 1100°C. The operating principle of the proposed MZI is illustrated in Fig. 1. Part of the incident light traveling in the fiber core is reflected at the left surface of the hollow ellipsoid, and propagates along path 1. When arriving at the interface between the

Fig. 1. Schematic of the proposed fiber in-line MZI.

Letter fiber cladding and air, it is reflected again, and incidents on the right surface of the hollow ellipsoid before being reflected back into the fiber core. The rest of the incident light travels across the hollow ellipsoid and recombines with the light of path 1 in the fiber core. The hollow ellipsoid is fabricated by the fs laser micromachining and fusion-splicing technique. The fs laser pulses with the wavelength of 800 nm, pulse width of 120 fs, and repetition rate of 1 kHz were focused onto a cleaved fiber end by a 20× objective lens with NA value of 0.5 and working distance of 2.1 mm. A section of SMF was placed on a computercontrolled translation stage with a resolution of 40 nm. A CCD camera was used to monitor the micromachining process. A micro-square centered at 20 μm away from the center of the cleaved fiber end, with side length of ∼30 μm and depth of ∼40 μm, was inscribed with fs laser power of ∼1 mW. The fiber end with micro-square was cleaned by use of ethanol to remove the dust before being fusion spliced together with another cleaved SMF end without micro-square to form a hollow ellipsoid. The fusion splicer used was FSM-45P (Fujikura), and the parameters set for fusion splicing were the same as that of splicing two SMFs, i.e., discharge time of 2000 ms, intensity of 16.2 mA. After fusion splicing, the hollow ellipsoid with major axis length of ∼50 μm was fabricated, as shown in Fig. 2(a). To demonstrate the propagating path 1 shown in Fig. 1, a red light beam was launched into the optical fiber with a hollow ellipsoid. When the red light illuminated the surface of the hollow sphere, strong reflection could be clearly observed as shown in Figs. 2(b) and 2(c), respectively, depending on the incident light direction. The transmission intensity of the MZI can be expressed by two beam interference theory as pffiffiffiffiffiffiffiffiffi (1) I I 1 I 2 2 I 1 I 2 cos2πΔnL∕λ; where I 1 and I 2 represent the light intensity of the two interferometer arms, respectively, λ is the wavelength of light beam, and ΔnL is the optical path difference between the two interferometer arms. The minimum intensity appears at the wave-

Fig. 2. Microscope image of the hollow ellipsoid (a) without red light illumination; (b) when red light incidents from left to right; (c) when red light incidents from right to left.


3517

lengths λm 2ΔnL∕2m 1, where m is an integer. Thus, the free spectral range (FSR) of the interference fringe can be described as [15] FSR λ2 ∕ΔnL:

(2)

As one of the arms of the MZI is an air-cavity, the optical path difference of the two arms becomes large due to the large RI difference between silica glass and air, thus the FSR is relatively small. The RI response of the device is investigated in the experiment. A broadband source was launched into the device, and the transmission spectrum was measured by an optical spectrum analyzer (OSA, AQ6317) with a resolution of 0.1 nm. The fiber device was fixed on a translation stage, and immersed by different liquids including pure water and RI oil (Cargille Laboratories). The transmission spectra with different RI values are shown in Fig. 3(a). The insertion loss of the device is ∼28 dB. Such a relatively high insertion loss is due to the double reflections at the two sides of the hollow ellipsoid, and the curvature of the ellipsoid surface that diverge the reflection beam. The loss is expected to be reduced if a large major axis length of the ellipsoid could be achieved. It can be seen from the figure that the visibility of the interference fringe decreases with the increase of RI. The visibility obtained at 1552 nm is ∼7.2 dB when the device is in air. When the device is in the liquid with RI of larger than 1.333, the visibility decreases rapidly and becomes less than 2.4 dB. The Fourier transform of the transmission spectra under air and RI of 1.37 are shown in Fig. 3(b), where it can be observed that the cladding mode is very weak when the device is immersed into the RI liquid. It is due to the fact that the RI liquid introduces a weak reflection at the air-cladding interface, and thus the intensity of the light from the path 1 becomes very weak, which reduces the visibility of the interference fringe pattern. The fringe visibility corresponding to different RI values ranging from 1.333 to 1.40 are displayed in Fig. 3(c), where the sensitivity obtained is ∼ − 14.3 dB∕RIURI unit. The curvature response of the MZI is also investigated. The MZI is placed on two translation stages with an initial separation of 140 mm, and one stage is moved inwardp with a step ffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi of 50 μm, and the curvature is calculated by C ≈ 24x∕L30 [5], where x is the travel distance of the moving stage, and L0 is the initial separation. Figure 4(a) shows the transmission spectra corresponding to different curvatures. When the curvature is varied from 0 to 2.88 m−1 , the intensity of the transmission spectrum decreases gradually, and a maximum variation of ∼1.8 dB is obtained at the dip wavelength of 1566.5 nm. Figure 4(b) shows the dip intensity versus curvature, and the sensitivity of −0.61 dB∕m−1 is obtained. It can be noticed from Fig. 4(b) that the transmission spectra only experience intensity variation while the dip wavelength is nearly a constant, which is due to the fact that when the MZI is bent, the reflection on the cladding-air interface is varied. The high temperature sensing capability and thermal stability of the device are tested. The fiber MZI is placed inside a tube furnace (CARROLITE MTF12/38/250). First, the temperature is increased from 24°C to 1100°C at an average rate of 15°C/min, and stays at 1100°C for one hour to eliminate the

3518


Letter

Fig. 4. (a) Transmission spectra of the device corresponding to different curvatures, and the inset shows the enlargement of the dip wavelength; (b) dip wavelength intensity variation with the curvature, and the inset shows the dip wavelength versus curvature.

Fig. 3. (a) Transmission spectra of the device corresponding to different RI values; (b) spatial frequency spectra of the device; (c) fringe visibility versus RI value.

influence of the fiber coating. Then, the temperature is cooled down from 1100°C to 24°C. Finally, the temperature is increased to 100°C and further to 1100°C with a step of 100°C and stay for 20 minutes at each step. The device is kept at 1100°C for 2 hours before cooling down, following the same process as in heating. The dip wavelength shift versus temperature variation is shown in Fig. 5, where a red wavelength shift

with the increase of temperature can be found. The temperature sensitivity obtained is 19.4pm/°C, and the repeatability is good. Such a sensitivity is similar to that reported in [18,19], and lower than that in [20,21]. The fibers used in our device are just conventional SMF, and when compared with PCF, our device has the advantage of low cost. Moreover, we have applied a maximum strain of ∼3000 με to the sensor, and found that the dip wavelength had no obvious shift, thus proving the robustness of the device. As our fiber in-line MZI is miniature and robust, that makes it a good candidate for high-temperature sensing. The spectra of different temperatures in the range from 24°C to 60°C have no obvious wavelength shift, as shown in the inset of Fig. 5. From the experimental results obtained for external RI, curvature, and temperature sensing, it can be noticed that the RI merely changes the fringe visibility, and the curvature only varies the fringe dip wavelength intensity, while the temperature just shifts the dip wavelength position in the range from 24°C to 60°C. Thus a simultaneous measurement of RI and temperature, or curvature and temperature can be achieved within the temperature range above mentioned.

Letter


3519

Polytechnic University Research (4-BCBF); National Natural Science Foundation of China (NSFC) (61377094). REFERENCES

Fig. 5. Dip wavelength shift versus temperature, and the inset shows the dip wavelength shift with different temperatures from 24°C to 60°C.

In conclusion, a miniature and robust fiber in-line MZI based on a hollow ellipsoid is demonstrated. The surface of the hollow ellipsoid essentially acts as an internal mirror that reflects the light in fiber core to the fiber cladding-air interface, and also the light from the cladding-air interface back into the fiber core. Such a device is sensitive to the external RI, curvature, and temperature in an independent manner, which allows a simultaneous measurement of external RI and temperature or curvature and temperature in the temperature range from 24°C to 60°C. Moreover, the hollow ellipsoid-based device has good high temperature sustainability. The excellent features of the device demonstrated make it a promising candidate for versatile photonic application. Funding. Hong Kong Government GRF (General Research Fund) (PolyU 152163/14E); Hong Kong

1. P. Lu, J. Harris, Y. Xu, Y. Lu, L. Chen, and X. Bao, Opt. Lett. 37, 4567 (2012). 2. Z. Tian, S. S. Yam, and H. P. Loock, Opt. Lett. 33, 1105 (2008). 3. J. Zhou, C. Liao, Y. P. Wang, G. Yin, X. Zhong, K. Yang, B. Sun, G. Wang, and Z. Li, Opt. Express 22, 1680 (2014). 4. D. Wu, T. Zhu, K. S. Chiang, and M. Deng, J. Lightwave Technol. 30, 805 (2012). 5. H. Gong, X. Yang, K. Ni, C. L. Zhao, and X. Dong, IEEE Photon. Technol. Lett. 26, 22 (2014). 6. Y. F. Geng, X. J. Li, X. L. Tan, Y. L. Deng, and Y. Q. Yu, IEEE Sens. J. 11, 2891 (2011). 7. H. Song, H. Gong, K. Ni, and X. Dong, Sens. Actuators A 203, 103 (2013). 8. O. Frazão, J. Viegas, P. Caldas, J. L. Santos, F. M. Araújo, L. A. Ferreira, and F. Farahi, Opt. Lett. 32, 3074 (2007). 9. J. H. Lim, H. S. Jang, K. S. Lee, J. C. Kim, and B. H. Lee, Opt. Lett. 29, 346 (2004). 10. H. T. Wang, S. L. Pu, N. Wang, S. H. Dong, and J. Huang, Opt. Lett. 38, 3765 (2013). 11. H. Fu, N. Zhao, M. Shao, H. Li, H. Gao, Q. Liu, Z. Yong, Y. Liu, and X. Qiao, IEEE Sens. J. 15, 520 (2015). 12. Y. Wang, M. W. Yang, D. N. Wang, S. J. Liu, and P. X. Lu, J. Opt. Soc. Am. B 27, 370 (2010). 13. P. Lu and Q. Chen, Opt. Lett. 36, 268 (2011). 14. T. Y. Hu, Y. Wang, C. R. Liao, and D. N. Wang, Opt. Lett. 37, 5082 (2012). 15. T. Y. Hu and D. N. Wang, Opt. Lett. 38, 3036 (2013). 16. C. R. Liao, H. F. Chen, and D. N. Wang, J. Lightwave Technol. 32, 2531 (2014). 17. C. R. Liao, D. N. Wang, and Y. Wang, Opt. Lett. 38, 757 (2013). 18. D. Monzon-Hernandez, V. P. Minkovich, and J. Villatoro, IEEE Photon. Technol. Lett. 18, 511 (2006). 19. G. Coviello, V. Finazzi, J. Villatoro, and V. Pruneri, Opt. Express 17, 21551 (2009). 20. Y. Wang, Y. Li, C. Liao, D. N. Wang, M. Yang, and P. Lu, IEEE Photon. Technol. Lett. 22, 39 (2010). 21. R. M. Gerosa, D. H. Spadoti, L. S. Menezes, and C. J. S. Matos, Opt. Express 19, 3124 (2011).