We have experimentally developed a highly sensitive and a compact size current sensor by using the CdSe quantum dots-doped bend insensitive optical fiber, operating in the visible band of wavelength. The modified sensitivity of this sensor was about 675 μrad/(Turn.A.m) for the loop radius of just 10 mm, which is more than 16 times larger than that of the single mode optical fiber current sensor.
© 2010 OSA
Optical fiber current sensors are attracting worldwide researchers because of their advantages such as light weight, cost effectiveness and their use in developing all optical fiber devices (e.g. switches modulators, circulators, and isolators, etc.). Although less sensitive, optical fiber current sensors certainly have advantages over their counterpart bulk glass current sensors as they don’t need bulk optical devices such as polarizers, birefringent plates, special launching lenses; instead they use all optical fiber devices that can be simply spliced, thereby avoiding precision alignment and careful handling [1–9]. Reported optical fiber current sensors include: (a) Fiber Bragg gratings (FBG) based optical fiber arrays , which have disadvantages such as a high temperature sensitivity, handling difficulties, and a high cost of infrared pumping components needed to develop sensors, (b) Single mode optical fiber current sensors [10,11] that are very portable, light in weight, have small size, and are compatible with existing optical devices; however their low sensitivity to the magnetic field is a main concern, and (c) Specialty optical fiber current sensors e.g., Eu2+-ions doped optical fibers that show enhanced Faraday rotation  and CdSe quantum dots-doped optical fibers  with better current sensitivity than the single mode fiber current sensor. However, all these optical fiber current sensors need big windings of the size of 7.5 cm radius because sharp loops cause substantial radiation losses. In contrary, to make a handy, and a small size current sensor device, the optical fiber used has to be wound with a loop size as small as possible (e.g. 10 mm of radius). Therefore, existing silica glass optical fibers reported for current sensing are not clearly suitable to make small size devices. In this regards, flint glass fibers having smaller photo-elastic coefficient as compared to silica glass fibers have shown bend insensitivity (for birefringence) and enhanced current sensitivity, nearly twice than that of silica glass fibers [14,15]. A major drawback of such devices is their incompatibility with existing optical fibers and networks, which predominantly consist of silica glass fibers.
Recently, bending loss insensitivity in the optical fiber has been successfully addressed by various researchers including our group [16–22]. For the visible wavelength operating fibers, which are of interest to develop the current sensor, bending loss of just 0.23 dB/loop for the loop radius of 5 mm has been reported at 633 nm . It is noted that for a visible wavelength optimized single mode optical fiber, the bending loss exceeds well beyond 100 dB for one loop of 5 mm radius and explains well why these single mode optical fibers cannot support compact devices that need very sharp loops. Therefore, to build a small and a compact device, natural choice will be to use bending-loss insensitive optical fiber (BIF) operating in the visible region (so that low cost visible wavelength sources can be used). A limiation of BIFs for current sensing application is their low sensitivity to the magnetic field, which is the same as the single mode optical fiber. Bend insensitivity along with high magento-optic sensitivity can be achieved by unifying two technologies of BIF and CdSe quantum dots doped optical fiber to produce a highly sensitive magnetic field sensor. In a current communication, we address this issue by reporting the development of CdSe quantum dots (QDs) doped BIF, where quantum dots were incorporated in the core of the optical fiber. We measured the modified current sensitivity of the CdSe QDs-doped BIF to be more that 16 times than that of the single mode fiber at 633 nm. Describing these effects, we first describe the experimental part to discuss the fabrication of a specialty optical fiber doped with CdSe QDs and having a bend insensitivity function, then we determine the magnetic sensitivity of this fiber in terms of a Verdet constant and finally, we show current sensing results of the current sensor developed using this fiber. It is worth mentioning that the BIF that is discussed in this paper is bending-loss insensitive, and to make it bending-birefringence insensitive, one needs methods such as twisting, annealing, etc.
To develop the bend insensitive optical fiber, an optical fiber preform with germano-silicate glass composition was fabricated by using the MCVD technique where low index trenches were formed by boron doping during the fabrication of preform. CdSe quantum dots were incorporated in the core of preform by using the solution doping technique where the toluene solution containing CdSe quantum dots (Sigma Aldrich, peak absorption < = 600 nm, 7.5 mg in 1.5 ml solution) was used. After the solution doping process, subsequent drying of the soaked preform was carried out and an additional glass layer was deposited to reduce possible evaporation of dopants. The optical fiber was drawn with an outer diameter of 125 µm at 2000 °C using a drawing tower. The optical fiber had the refractive index profile as shown in Fig. 1 . A cutoff wavelength was measured to be about 600 nm by using the bend reference technique.
Bending loss characteristics of the CdSe QDs-doped optical fiber were measured by using 1 to 18 loops of various diameters. Input power was applied to the fiber by using a wideband white light source and the output power was detected by the optical spectrum analyzer. The bending loss characteristics of the fiber at 633 nm are shown in Fig. 2 for various numbers of loops at 5 mm of radius. The mean bending loss of the CdSe QDs-doped BIF was about 0.47 dB/loop at 633 nm for 5 mm of a loop radius. Variations of the bending loss of this fiber at various loops of different radii are show in Fig. 3 where a negligible bending loss can be observed for 20 mm bending radius while just 0.18 dB/loop bending loss can be observed for 10 mm of bending radius at 633 nm. These results prove that the bend insensitivity of the CdSe QDs-doped BIF was quite good.
We measured the absorption spectrum of the fiber by using the cutback technique and this spectrum was used to determine the existence of quantum dots. As shown in Fig. 4 , two distinct absorption peaks were observed at 608 nm and 585 nm that are related to the CdSe quantum dots in the optical fiber . To further ascertain the presence of CdSe quantum dots, we carried out the transmission electron microscopy (TEM) measurement of the preform sample. As shown in Fig. 5 , CdSe quantum dots with radius less than about 2 nm can be observed. The concentration of CdSe quantum dots can be determined by using the absorption coefficient data of Fig. 4. An approximate equation representing the relationship between the absorption cross section and the size of CdSe QDs is expressed as :Fig. 5). The concentration of CdSe QDs (per m3) can also be approximately evaluated using the relationship :Equation (2) gives CdSe QDs concentration of about 2 × 1025 quantum dots per m3. With this high concentration of CdSe QDs, it is expected that the magneto-optic sensitivity of the CdSe QDs-doped BIF should be high .
Next, to determine the magnetic sensitivity of the CdSe QDs-doped BIF in terms of the Faraday rotation, we launched the linearly polarized light from the HeNe gas laser (633 nm) into the CdSe QDs-doped BIF (71 cm) that was kept under the influence of magnetic field generated by the dc solenoid and the output power was detected and applied to the polarimeter, which determined the Faraday rotation angle by using a Poincare sphere. Care was taken to minimize the linear birefringence of the fiber and measurements were performed at room temperature (25 °C), which was maintained constant throughout the experiment. Various measurements of the Faraday rotation angle with the applied magnetic field at 633 nm are shown in Fig. 6 . Considering the beat length of BIF to be about 10 m (as discussed in the next section), the fiber sample with 71 cm can be approximately considered to be free from random polarization effect, although a small increasing spread in the readings with increased magnetic field does indicate the linear polarization effect. It can be noticed that the mean Faraday rotation angle of the polarization plane was about 40.8 degrees for 0.14 T (as shown in Fig. 6), which is about 2.6 times greater than the Faraday rotation angle of the single mode optical fiber at 633 nm . This enhanced magnetic field sensitivity can be contributed to the high CdSe quantum dots concentration. The Verdet constant of the CdSe QDs-doped BIF can be determined from measurements of Fig. 6 as follows. When the plane of polarization of the light beam passes through the medium that is under uniform magnetic field parallel to the light propagation direction, it rotates by an angle given by,13].
Finally, an experimental arrangement for sensing the current by using the CdSe QDs-doped BIF is shown in Fig. 7 . The He-Ne gas laser emitting at 633 nm (10 mW) was used as the input source. The linearly polarized light at 633 nm was launched into the optical fiber using a collimator. The optical fiber with just 4 m of length was twisted manually with 15 twists per meter to minimize the linear birefringence effect in the optical fiber and then it was wound on a plastic drum (10 mm radius) by using 64 loops. Its output was directly launched into a photo detector attached to the polarimeter connected to a personal computer (PC), which displayed the Faraday rotation angle by using the built-in software. A conductor carrying a high current of 40 A was inserted in the hollow portion of the drum. The magnetic field generated due to flow of the current was detected in terms of Faraday rotation angle by the Pioncare sphere on PC. All measurements were carried out at room temperature.
3. Results and discussion
As explained earlier, for CdSe quantum dots concentration of 2 × 1025 QDs/m3, we obtained the Verdet constant to be about 7.2 rad/(T.m). This high value of the Verdet constant is contributed to the existence of CdSe quantum dots and it is quite higher than the reported mean value of 4.6 rad/(T.m) in the other CdSe QDs-doped optical fiber  due to high concentration of quantum dots in our fiber. The current sensor shown in Fig. 7 was realized by using the CdSe QDs-doped BIF with 10 mm of loop radius, which is the smallest size optical fiber current sensor reported so far. Without twists, Pioncare measurements were random due to the presence of significant linear birefringence in the BIF. As per earlier studies, by twisting the fiber by 10 to 15 twists/m, the effect of linear birefringence can be minimized [11,25]. With the twisting rate of 15 twists/m for the CdSe QDs-doped BIF, the Pioncare measurements showed observable results.
Variations of the Faraday rotation angle at 633 nm with the current flowing in the conductor are shown in Fig. 8 for the CdSe QDs-doped BIF. Performance of the optical fiber current sensor is usually determined in terms of the current sensitivity defined as:Table 1 . It can be observed that CdSe QDs-doped BIF current sensor showed the modified current sensitivity of 675 μrad/(A.Turn.m) that is over 8 times and 16 times larger than the modified current sensitivities of the CdSe QDs-doped optical fiber (non-BIF) current sensor and the single mode optical fiber current sensor, respectively.
With regards to the birefringence in the BIF, which is defined as , where neff is the effective index and suffixes x and y represent x-polarization and y-polarization direction, respectively. The birefringence in units of radian/m can be expressed by:
In earlier studies regarding birefringence of optical fibers, the single mode optical fiber had shown the beat length of about 30 m to 100 m and the dispersion shifted fiber with boron doping had shown higher birefringence as compared to the single mode optical fiber with the beat length value of about 10 m and the polarization mode dispersion of about 0.061 ps/(km)1/2 [26,27]. Typical commercial bend insensitive optical fiber of  has the polarization mode dispersion of about 0.06 ps/(km)1/2, and by considering the boron doping and similarity between our and reported fibers, for the qualitative discussion, the beat length of our BIF can be assumed to be about 10 m.
When the optical fiber is straight, by using Eq. (6) to Eq. (8), its linear birefringence can be calculated to be 6.33 × 10−8 or 0.628 radians/m at 633 nm with σ = 2.73 × 10−7. However, when this optical fiber is bent with the bending radius of Rb, the linear birefringence (rad/m) due to bending is expressed as 30], which is quite less to make the circular birefringence a dominating factor. As we have already discussed, after twisting, the Faraday rotation was about 1° for 40 A of current, which was although enough to detect the current, was quite small and indicates the presence of stray linear birefringence in the bent optical fiber. To elaborate this point, let’s take a specific example. In the presence of the circular birefringence and the linear birefringence, the signal R (in arbitrary units) at the photo-detector is given by :Eq. (9), i.e., to be = 52.6 rad/m, gives R = 4.3 × 10−7. Now, consider a case where the linear birefringence is negligible in the bent BIF, i.e., = 0, and we get R = 0.035, a huge improvement in the value of R. Thus, in the presence of linear birefringence, one needs sensitive-photo detector at the receiving end so that even a low quality signal can be accepted. In the earlier work, the problem of linear birefringence was addressed by annealing the optical fiber at 850 °C followed by slow cooling . In another approach to reduce the linear birefringence effect, elliptically birefringent fibers were produced by spinning preforms during the drawing process . Although for our CdSe QDs-doped optical fiber current sensor we did not carry out annealing and spinning (rather we used twisting), it detected current from 1A to 40A reasonably well in the temperature controlled laboratory environment. It is noted that, to make it more stable during the harsh environment, one would need annealing the BIF before applying it for the temperature sensing application .
We have developed an optical fiber current sensor by using the CdSe QDs-doped bend insensitive optical fiber. The bending loss of the fiber was mere 0.47 dB for one loop of 5 mm radius, and 0.18 dB for one loop of 10 mm radius at 633 nm. This allowed us to build a very tiny size optical fiber current sensor (10 mm of radius), which showed the modified current sensitivity to be 16 times larger than that of the single mode optical fiber operating at 633 nm.
This work was supported by the GIST Top Brand Project (Photonics 2020), Ministry of Science and Technology, South Korea, by the National Core Research Center (NCRC) for Hybrid Materials Solution of Pusan National University, by the Brain Korea-21 Information Technology Project, Ministry of Education and Human Resources Development, South Korea, and by the GTI, Gwangju Institute of Science and Technology, Gwangju, South Korea.
References and links
1. K. Tanaka, K. Fujita, N. Matsuoka, K. Hirao, and N. Soga, “Large Faraday effect and local structure of alkali silicate glasses containing divalent europium ions,” J. Mater. Res. 13(7), 1989–1995 (1998). [CrossRef]
2. M. W. Shafer and J. C. Suits, “Preparation and Faraday rotation of divalent europium glasses,” J. Am. Ceram. Soc. 49(5), 261–264 (1966). [CrossRef]
3. J. T. Kohli and J. E. Shelby, “Magneto-optical properties of rare earth aluminosilicate glasses,” Phys. Chem. Glasses 32, 109–114 (1991).
4. J. F. Owen, P. B. Dorain, and T. Kobayasi, “Excited-state absorption in Eu+2:CaF2 and Ce+3: YAG single crystals at 298 and 77K,” J. Appl. Phys. 52(3), 1216–1223 (1981). [CrossRef]
5. T. Sato and I. Sone, “Development of bulk-optic current sensor using glass ring type Faraday cells,” Opt. Rev. 4(1), 35–37 (1997). [CrossRef]
6. G. Li, M. G. Kong, G. R. Jones, and J. W. Spencer, “Sensitivity improvement of an optical current sensor with enhanced Faraday rotation,” IEEE J. Lightwave Technol. 15(12), 2246–2252 (1997). [CrossRef]
7. T. D. Maffetone and T. M. McClelland, “345 kV substation optical current measurement system for revenue metering and protective relaying,” IEEE Trans. Power Deliv. 6(4), 1430–1437 (1991). [CrossRef]
9. J. C. Yong, S. H. Yun, M. L. Lee, and B. Y. Kim, “Frequency-division-multiplexed polarimetric fiber laser current-sensor array,” Opt. Lett. 24(16), 1097–1099 (1999). [CrossRef]
10. A. H. Rose, S. M. Etzel, and K. B. Rochford, “Optical fiber current sensors in high electric field environments,” IEEE J. Lightwave Technol. 17(6), 1042–1048 (1999). [CrossRef]
11. M. Grexa, G. Hermann, G. Lasnitschka, and A. Scharmann, “Faraday rotation in a single-mode fiber with controlled birefringence,” Appl. Phys. B 35(3), 145–148 (1984). [CrossRef]
12. D. H. Kim, H. Y. Yang, B. H. Kim, U. C. Paek, and W.-T. Han, in Proceedings of Opto-Electronics and Communications Conference (OECC/COIN 2004), Yokohama Japan (2004), p.512.
14. K. Kurosawa, S. Yoshida, and K. Sakamoto, “Polarization properties of the flint glass fiber,” IEEE J. Lightwave Technol. 13(7), 1378–1384 (1995). [CrossRef]
15. K. Kurosawa, I. Masuda, and T. Yamashita, “Faraday effect current sensor using flint glass fiber for the sensing element,” in Proceedings of Optical Fiber Sensor Conference, Florence, USA (1993), pp. 415–418.
16. S. Matsuo, M. Ikeda, and K. Himeno, “Bend insensitive and low splice loss optical fiber for indoor wiring in FTTH,” in Proceedings of Optical Fiber Communication Conference (OFC), Anaheim, USA, Feb.23–27,2004 (Optical Society of America, Technical Digest, 2004), Paper number: ThI3.
17. M.-J. Li, P. Tandon, D. C. Bookbinder, S. R. Bickham, M. A. McDermott, R. B. Desorcie, D. A. Nolan, J. J. Johnson, K. A. Lewis, and J. J. Englebert, “Ultra-low bending loss single-mode fiber for FTTH,” in Proceedings of OFC/NFOEC-2008, San Diego, USA, Feb.24–28,2008 (Optical Society of America, Technical Digest, 2008), Paper number: PDP10.
18. I. Sakabe, H. Ishikawa, H. Tanji, Y. Terasawa, M. Ito, and T. Ueda, “Enhanced bending loss insensitive fiber and new cables for CWDM access networks,” in Proceeding of 53rd International Wire and Cable Symposium, Philadelphia, USA, November 14–17 (2004), pp. 112–118.
19. K. Himeno, S. Matsuo, N. Guan, and A. Wada, “Low bending loss single mode fibers for Fiber-to-the-Home,” IEEE J. Lightwave Technol. 23(11), 3494–3499 (2005). [CrossRef]
23. C. A. Leatherdale, W.-K. Woo, F. V. Mikulec, and M. G. Bawendi, “On the absorption cross section of CdSe nanocrystal quantum dots,” J. Phys. Chem. B 106(31), 7619–7622 (2002). [CrossRef]
24. J. H. Kratzer and J. Schroeder, “Magnetooptic properties of semiconductor quantum dots in glass composition,” J. Non-Cryst. Solids 349, 299–308 (2004). [CrossRef]
26. M. Legre, M. Wegmuller, and N. Gisin, “Investigation of the ratio between phase and group birefringence in optical single mode fibers,” IEEE J. Lightwave Technol. 21(12), 3374–3378 (2003). [CrossRef]
27. A. Galtarossa, L. Palmieri, M. Schiano, and T. Tambosso, “Measurements of beat length and perturbation length in long single-mode fibers,” Opt. Lett. 25(6), 384–386 (2000). [CrossRef]
28. Samsung bend insensitive optical fiber data-sheets (2010).
31. A. Ghatak, and K. Thyagarajan, Introduction to Fiber Optics (Cambridge University Press, USA, 1998).
32. D. Tang, A. H. Rose, G. W. Day, and S. M. Etzel, “Annealing of linear birefringence in single mode fiber coils: application to optical fiber current sensors,” IEEE J. Lightwave Technol. 9(8), 1031–1037 (1991). [CrossRef]
33. R. I. Laming and D. N. Payne, “Electric current sensors employing spun highly birefringent optical fiber,” IEEE J. Lightwave Technol. 7(12), 2084–2094 (1989). [CrossRef]