Enhanced magneto-optical effect in cobalt nanoparticle-doped optical fiber Helmut C. Y. Yu,1,2,* Martijn A. van Eijkelenborg,1 Sergio G. Leon-Saval,1 Alexander Argyros,1 and Geoff W. Barton2 1
Optical Fibre Technology Centre, University of Sydney, 206 National Innovation Centre, Australian Technology Park, Eveleigh, NSW 1430, Australia
2
School of Chemical and Biomolecular Engineering, University of Sydney, NSW 2006, Australia *Corresponding author:
[email protected] Received 15 August 2008; revised 21 October 2008; accepted 26 October 2008; posted 30 October 2008 (Doc. ID 100182); published 3 December 2008
An enhanced magnetic Faraday effect is demonstrated in cobalt nanoparticle-doped polymer optical fiber. Magneto-optically induced rotation of the plane of polarization proportional to both the dopant particle concentration and the magnetic field strength is demonstrated. Potential applications include magnetic field sensors, current sensors, and in-fiber optical isolators. © 2008 Optical Society of America OCIS codes: 060.2290, 060.2370, 060.4005, 160.3820, 230.2240, 350.4990.
1. Introduction
The magneto-optic Faraday effect induces a nonreciprocal circular birefringence proportional to a material’s Verdet constant and the applied longitudinal magnetic field strength. The magnetic field breaks the symmetry of the material as it induces a difference in refractive index for left- and right-handed circular polarizations. As a result, the angle of linear polarization of light passing through the material along the direction of the magnetic field will rotate. The Faraday rotation θ is given by θ ¼ VBL;
ð1Þ
where V is the material Verdet constant, B is the magnetic field strength, and L is the length of the sample. To achieve a given effect, it is desirable to utilize a material with a high V as this minimizes the requirement for both B and L. The magnetic Faraday effect is nonreciprocal; thus, propagating backward through the material does not return the polarization to the initial state. Therefore, it forms the basis of optical isolators that 0003-6935/08/356497-05$15.00/0 © 2008 Optical Society of America
are essential to protect light sources from unwanted reflections that can reduce the source stability, especially when fast switching and large bandwidth are important, such as in mode-locked lasers [1]. To date, only bulk-optic isolators have been developed, despite there being a strong need for compact, integrated isolators in many applications. The magnetic Faraday effect is also used in fiber-optic current sensing [2], magnetic field measurements [3], the development of optical modulators [4], and measurements of the toroidal plasma current in Tokomak fusion reactors [5]. The effect is very small in conventional silica and polymer fibers due to the low value of V for those materials [6]. Furthermore, any linear birefringence of the sensing fiber greatly influences the desired signal caused by the small magneto-optic effect [7–9]. For all these applications, the development of optical waveguides and fibers made from materials with greatly enhanced Verdet constants is clearly of importance. The Verdet constants of some optical fiber materials, such as chalcogenide glasses, have been investigated, with Ge33 As12 Se55 found to have the highest value of V ¼ 14 rad T−1 m−1 at 1550 nm, which is some 30 times that of silica [10]. One of the highest Verdet constants observed in bulk materials is for 10 December 2008 / Vol. 47, No. 35 / APPLIED OPTICS
6497
yttrium iron garnet (YIG), for which V ¼ 1309 rad T−1 m−1 when measured at >1000 nm wavelength [11]. This value of V means that the 45° of Faraday rotation necessary for isolator operation can be achieved in a path length of ∼2 mm using a magnetic field strength that can be readily obtained from a small permanent magnet (∼0:5 T). Unfortunately, optical fiber cannot be fabricated from this material. Composite materials can combine useful material properties. In particular, nanoparticles have been incorporated into transparent media in order to obtain enhanced optical functionalities [12,13]. Super paramagnetic nanoparticles have been incorporated into polymethyl methacrylate (PMMA) and the resulting V can be large. For example, the incorporation of γ-Fe2 O3 particles in PMMA was reported to give V ¼ 62:8 rad T−1 m−1 per weight percent at 632 nm [14,15]. Similarly, the incorporation of cobalt nanoparticles into PMMA thin films has led to a measured rotation of 1°=μm thickness at 50 wt: % cobalt concentration for a magnetizing field of 6:4 kA=cm [12]. This last value indicates that a concentration of ∼15 wt: % cobalt nanoparticles in PMMA should lead to a V equal to that of solid YIG. Light guidance in microstructured polymer optical fibers (mPOF) was demonstrated some years ago. By creating microscopic air channels inside the polymeric material, light can be guided along the fiber [16,17]. More recently, a “universal” doping technique was developed that allows a wide range of nanoparticles to be introduced within such an optical fiber [13]. This created the opportunity for developing diverse novel fibers. In this paper, we demonstrate the use of this method to achieve an enhanced magnetic Faraday effect in cobalt nanoparticle-doped mPOF.
core, and the sleeve. This procedure creates very thin bridges that are ideal for the confinement of light within the core [18]. Fiber drawn with a core rod doped with 0:05 wt: % cobalt nanoparticles is shown in Fig. 1. 3.
Faraday Rotation Experiments
Experiments were conducted on both the doped rods and the drawn fiber. A balanced bridge detection arrangement, as shown in Fig. 2, is used to accurately measure the polarization rotation [3]. By using a broadband supercontinuum source and two optical spectrum analyzers (Ocean Optics USB4000), the Faraday rotation can be determined over a 500 nm wide wavelength range. A 1 cm length of cobaltdoped rod is placed inside a 6:3 cm long solenoid (BLP Components) orientated to produce a B field along the propagation direction. A polarizer oriented at 45° to the horizontal was placed between the (unpolarized) source and the solenoid. A polarizing beam splitter (Glan laser calcite polarizer GL10) was placed after the solenoid to split the transmitted light into the x and y polarized components, labeled as I x ðλÞ and I y ðλÞ. Magnification lenses (4×) coupled the beams into the two spectrum analyzers. The intensity spectra I x ðλÞ and I y ðλÞ are recorded for solenoid currents of 9:63 A. The background spectra, recorded when blocking the beam, were subsequently subtracted. The balanced bridge arrangement allows Eq. (2) to give a rotation curve for the two magnetic field currents. A Faraday rotation curve θðλÞ is then determined by taking half the
2. Fabrication
PMMA particles of 0:5 mm diameter (Sigma-Aldrich) were mixed with either (i) 0:0055 wt: % cobalt nanoparticle powder wetted with toluene or (ii) 0.050, 0.125, and 0:25 wt: % functionalized cobalt nanoparticles having a L-cysteine ethyl ester coating and suspended in ethanol. The doped powder was ground using a small ball mill (Retsch MM 301) for 10 min. This achieved a homogeneous mixture while reducing the size of the PMMA particles to around 50 μm. The mixed powder was placed in a mold and slowly lowered under vacuum into a furnace at 225 °C, fusing the powder and creating a doped rod (approximately 60 mm long and 5 mm in diameter) as reported previously [13]. Each doped rod was subsequently annealed at 90 °C for three days. Microstructured polymer optical fibers with a composite core are fabricated by the “stack and draw” technique, whereby the doped rod is placed in the center of six circular 5 mm diameter capillaries stacked inside a 17 mm outer diameter PMMA sleeve. The capillaries are sealed at the end where vacuum is applied during the drawing process to remove the air present between the capillaries, the 6498
APPLIED OPTICS / Vol. 47, No. 35 / 10 December 2008
Fig. 1. (Color online) Microstructured polymer optical fiber (mPOF) with a PMMA/cobalt nanoparticle composite material core. The fiber has an external diameter of 400 μm and a core of ∼50 μm.
Fig. 2. Setup to measure the magnetic Faraday rotations in nanoparticle-doped PMMA samples. SCG, super-continuum generation source; P, polarizer; S, solenoid; PBS, polarizing beam splitter; L, 4× objective lens; OSA, optical spectrum analyzer.
difference between these two rotations: Ix − Iy 1 θðλÞ ¼ arcsin Ix þ Iy 2
ð2Þ
It was found that the solenoid did not produce a uniform magnetic field along the longitudinal axis, which meant that the position of the sample within the solenoid is a critical parameter. Consequently, the magnetic field profile inside the solenoid was measured and found to have a maximum field strength of 0:29 T, with a drop of some 30% per centimeter away from the maximum. By always positioning the end of any sample at z ¼ 0, the average magnetic field strength Bav experienced by the sample is equal to
Bav
1 ¼ L
ZL BðzÞdz:
ð3Þ
0
The Verdet constant, VðλÞ, is then calculated using Eq. (1) by dividing θðλÞ by Bav and L. The experimental arrangement for measurements on doped fiber required several modifications. First, a 10× lens is placed after the first polarizer to couple the supercontinuum source into an mPOF having a suspended core of 20 μm and an outer diameter of 430 μm [19]. This mPOF was not doped and was simply used to launch light into a 3:8 cm length of doped fiber by butt coupling. The doped fiber was located inside the solenoid. A 25× magnification lens was placed at the output of the doped fiber to image the end face onto a pinhole that selectively passes light from the core of the fiber only onto the beam splitter and the OSAs. 4. Results
The Verdet constant measurements as a function of wavelength are shown in Fig. 3. A clear λ−2 dependence is observed in all measurements, as expected [10,20]. The dashed curve is the theoretical value of VðλÞ for PMMA calculated using Becquerel’s equation [10] along with the Cauchy equation for the refractive index of PMMA [21]. The other curves correspond to measurements in bulk form (rods) with cobalt dopant concentrations of: a, 0%; b, 0.0055%; c, 0.05%; d, 0.125%; and e, 0.25%. Curve f corresponds to the fiber result for a concentration of 0.05%. Measurements for undoped PMMA (curve a) agree with
Fig. 3. (Color online) Verdet constant versus wavelength for different concentrations of cobalt nanoparticles in PMMA. The dashed curve is the calculated undoped-PMMA result. The cobalt concentrations for the other curves are: a, 0%; b, 0.0055%; c, 0.05%; d, 0.125%; e, 0.25%; and f, 0.05%. The Verdet constant was measured using bulk samples for curves a–e, and using a fiber sample for curve f.
the calculated values to within 5% while, at the lowest dopant concentration (curve b), there is little change. Starting from 0:05 wt: % (curve c), however, a clear enhancement in V is observed, which increases proportionally with the dopant concentration (see curves d and e). Note that as the dopant concentration increases, the scattering becomes stronger (especially at shorter wavelengths) and the measurement range over which sufficient signal could be obtained shifts to longer wavelengths. The fiber result at 0:05 wt: % (curve f) is in good agreement with the bulk sample measurement at the same concentration (curve c). Note that the measurements for the doped fiber extend to shorter wavelengths as a more powerful supercontinuum source was used for fiber characterization. Fiber measurements could not be obtained at higher concentrations as transmission losses were too high to provide a meaningful signal. To verify the linear dependence of the Verdet constant on dopant concentration, the measurements shown in Fig. 3 were integrated over their 150 nm wide common wavelength range (centered on 825 nm) and plotted against cobalt concentration, as shown in Fig. 4. A reasonable linear dependence of V on concentration is observed, with a best-fit curve a þ bx being added. The fitted parameter a ¼ 1:42 rad T−1 m−1 is in fair agreement with the calculated pure PMMA result of 1:58 rad T−1 m−1. The measured dependence on the cobalt concentration is b ¼ 23:3 rad T−1 m−1 per weight percent. The only comparable literature value available is the reported rotation of 1°=μm at 60 wt: % cobalt, a magnetizing field of 6:4 kA=cm, and a wavelength of 500 nm as measured on a PMMA thin film [12]. Using an estimated relative magnetic permeability μr ¼ 5:5 [22] and scaling to 825 nm using the established λ−2 dependence, it is possible to convert this result to 10 December 2008 / Vol. 47, No. 35 / APPLIED OPTICS
6499
ments should enhance the achievable Verdet constant to a point where magnetic field and current sensors, as well as an all-fiber optical isolator, are realistic possibilities. 6.
Fig. 4. Average Verdet constant at 825 nm versus cobalt concentration.
20:3 rad T−1 m−1 per weight percent, which is in good agreement with our result. 5. Discussion
Demonstrating the magnetic Faraday effect in a nanoparticle-doped mPOF is a promising first step toward an all-fiber optical isolator and other magneto-optical fiber devices. One issue that remains, however, is the transmission loss with such composite materials. Measured at 808 nm, we find that the losses of the bulk samples doped with cobalt nanoparticles are 3:2 dB=cm for 0.0055%, 7:8 dB=cm for 0.05%, and 25:6 dB=cm for 0.125%. These losses are much larger than expected. They are attributed predominantly to the presence of the nanoparticles, since the intrinsic loss of PMMA (∼0:019 dB=cm) and the loss arising from the powder fusing process (∼0:12 dB=cm) are much smaller. Transmission electron microscope images of the cobalt nanoparticles showed that their average size was 212 90 nm, somewhat larger than the supplier’s quoted value of 10 nm (Strem Chemicals, Inc.). Particles of this size will lead to large Rayleigh scattering at visible wavelengths. In addition, agglomerated cobalt “lumps” were found with a size of 1 to 2 μm, which can lead to wavelength-independent scattering losses. Both loss mechanisms can be reduced by using smaller particles and functionalizing the particle surface so as to prevent clustering. This was verified by making a doped PMMA rod using smaller, functionalized iron oxide particles, which could be readily obtained as small as 6:5 3 nm (SigmaAldrich). A rod doped with 0.124% these particles had a measured loss of only 3:9 dB=cm at 808 nm. Unfortunately, such particles do not lead to an enhanced Verdet constant. Work is continuing to reduce the cobalt particle size and to functionalize the surface. In addition, we aim to eliminate the formation of microscopic voids in the material that also contribute to optical loss. Furthermore, by embedding the dopant particles in fluorinated polymer materials that operate at near infrared wavelengths, the scattering losses can be reduced with a λ−4 dependence, noting that the Verdet constant decreases more slowly as λ−2. It is anticipated that a combination of such improve6500
APPLIED OPTICS / Vol. 47, No. 35 / 10 December 2008
Conclusion
A doped polymer optical fiber, fabricated by fusing a mixture of cobalt nanoparticles and PMMA powder, was shown to exhibit rotation of the polarization of light under the influence of a longitudinal magnetic field provided by a small solenoid. Agreement was found between the calculated and measured Verdet constants for undoped PMMA (1.58 and 1:42 rad T−1 m−1 , respectively). An enhanced magnetic Faraday effect proportional to the cobalt nanoparticle concentration was demonstrated, with the Verdet constant increasing linearly at 23:3 rad T−1 m−1 per weight percent. This demonstration is an encouraging first step toward the development of a range of practical fiber-based magneto-optical devices. We thank the Australian Research Council (ARC) for funding of this work and the University of Sydney’s Electron Microscope Unit (EMU) for the transmission electron microscope images. References 1. A. Hideur, T. Chartier, M. Brunel, M. Salhi, C. Özkul, and F. Sanchez, “Mode-lock, Q-switch and CW operation of an Ybdoped double-clad fiber ring laser,” Opt. Commun. 198, 141–146 (2001). 2. H. Lassing, W. J. Mastop, A. F. G. van der Meer, and A. A. M. Oomens, “Plasma current measurements by Faraday rotation in a single-mode fiber,” Appl. Opt. 26, 2456–2460 (1987). 3. E. H. Hwang and B. Y. Kim, “Pulsed high magnetic field sensor using polymethyl methacrylate,” Meas. Sci. Technol. 17, 2015–2021 (2006). 4. S. D. Jacobs, “Faraday rotation, optical isolation and modulation at 10:6 μm using hot-pressed CdCr2 S4 and CoCr2 S4 ,” J. Electron. Mater. 4, 223–241 (1975). 5. G. I. Chandler and F. C. Jahoda, “Current measurements by Faraday rotation in single-mode optical fibers,” Rev. Sci. Instrum. 56, 852–862 (1985). 6. C. Z. Tan and J. Arndt, “Faraday effect in silica glass,” Physica B 233, 1–7 (1997). 7. H. Harms, A. Papp, and K. Kempter, “Magneto-optical properties of index-gradient optical fibers,” Appl. Opt. 15, 799–801 (1976). 8. A. M. Smith, “Polarization and magneto-optic properties of single-mode optical fiber,” Appl. Opt. 17, 52–56 (1978). 9. S. C. Rashleigh and R. Ulrich, “Magneto-optic current sensing with birefringent fibers,” Appl. Phys. Lett. 34 (11), 768–770 (1979). 10. Y. Ruan, R. A. Jarvis, A. V. Rode, S. Madden, and B. LutherDavies, “Wavelength dispersion of Verdet constants in chalcogenide glasses for magneto-optical waveguide devices,” Opt. Commun. 252, 39–45 (2005). 11. A. E. Turner, R. L. Gunshor, and S. Datta, “New class of materials for optical isolators,” Appl. Opt. 22, 3152–3154 (1983). 12. A. Horikawa, K. Yamaguchi, M. Inoue, T. Fujii, and K. I. Arai, “Magneto-optical effect of films with nano-clustered cobalt particles dispersed in PMMA plastics,” Mater. Sci. Eng. A 217, 348–352 (1996), see Fig. 6a.
13. H. C. Y. Yu, A. Argyros, G. Barton, M. A. van Eijkelenborg, C. Barbe, K. Finnie, L. Kong, F. Ladouceur, and S. McNiven, “Quantum dot and silica nanoparticle doped polymer optical fibers,” Opt. Express 15, 9989–9994 (2007). 14. J. Wouters, “Superparamagnetic nanoparticles for Faraday rotation,” http://www.kuleuven.be/inpac/presentations/ Woueters_WP6_INPAC.pdf. 15. V. K. Valev, J. Wouters, and T. Verbiest, “Precise measurements of Faraday rotation using ac magnetic fields,” Am. J. Phys. 76, 626–629 (2008). 16. M. A. van Eijkelenborg, M. C. J. Large, A. Argyros, J. Zagari, S. Manos, N. A. Issa, I. Bassett, S. Fleming, R. C. McPhedran, C. M. De Sterke, and N. A. P. Nicorovici, “Microstructured polymer optical fiber,” Opt. Express 9 (7), 319–327 (2001). 17. M. C. J. Large, L. Poladian, G. W. Barton, and M. A. van Eijkelenborg, Microstructured Polymer Optical Fibers (Springer Verlag, 2008).
18. N. A. Issa, “High numerical aperture in multimode microstructured optical fibers,” Appl. Opt. 43, 6191–6197 (2004). 19. R. Lwin, G. Barton, L. Harvey, J. Harvey, D. Hirst, S. Manos, M. C. J. Large, L. Poladian, A. Bachmann, H. Poisel, and K.-F. Klein, “Beyond the bandwidth-length product: graded index microstructured polymer optical fiber,” Appl. Phys. Lett. 91, 191119 (2007). 20. A. Jain, J. Kumar, F. Zhou, and L. Li, “A simple experiment for determining Verdet constants using alternating current magnetic fields,” Am. J. Phys. 67, 714–717 (1999). 21. J. D. Swalen, R. Santo, M. Tacke, and J. Fischer, “Properties of polymeric thin films by integrated optical techniques,” IBM J. Res. Dev. 21 (2), 168–175 (1977). 22. S. Kimura, T. Kato, T. Hyodo, Y. Shimizu, and M. Egashira, “Electromagnetic wave absorption properties of carbonyl iron-ferrite/PMMA composites fabricated by hybridization method,” J. Magn. Magn. Mater. 312, 181–186 (2007).
10 December 2008 / Vol. 47, No. 35 / APPLIED OPTICS
6501
