Open Access
Add to BibSonomy
Abstract
We propose a perspective type of insulator-metal-insulator magnetoplasmonic crystal waveguide, composed of a gold grating placed between two garnet layers. Using an original non-perturbing method for the deposition of the upper magneto-dielectric layer, we fabricate the samples and provide experimental results evidencing the coupling of surface plasmon-polaritons propagating on the opposite Au/garnet interfaces. In contrast to traditional Au/garnet magnetoplasmonic crystals, spectra of the magneto-optical effect measured in transmission through this waveguide demonstrate rather specific features: a high-quality resonance for the long-range surface plasmon-polariton and a broad 60 nm wide resonance for the short-range surface plasmon-polariton. Our findings open new routes towards the development of high-sensitivity robust magnetoplasmonic sensors.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Corrections
4 February 2019: A typographical correction was made to the author listing.
1. Introduction
Research in the field of magnetoplasmonics is fueled with the possibility of tailoring light-matter interaction on the nanoscale [1] by localizing light in a gyrotropic medium via the surface plasmon resonance. This recipe was used to design various magnetoplasmonic systems demonstrating enhanced magneto-optical effects due to the excitation of localized [2–5] and propagating [3,6–8] surface plasmons. The dispersion relation of a surface plasmon polariton (SPP) can be shifted by magnetizing the medium in the direction parallel to the SPP magnetic field component. This results in intensity magneto-optical effects: odd in magnetization variation of the light reflection or transmission characterized by the magnetic contrast ρ = [T(+M) – T(−M)]T(0), where T (±M) is the transmission for the opposite magnetization directions, and (T(0) is that for zero magnetization. Plasmon-induced enhancement of this effect can lead to the development of high-sensitivity refractive index and magnetic field sensors [9, 10].
Among various designs for magnetoplasmonic structures that vary by the choice of the magnetic material (metal or dielectric) and the SPP excitation method (prism or grating coupling), one of the most promising is the combination of a noble metal grating (gold) and a magnetic dielectric (iron garnet) [8, 11, 12] (Fig. 1(a)). These structures known as magnetoplasmonic crystals (MPC) have much lower losses and thus higher quality factor of resonances as compared to the magnetoplasmonic structures based on ferromagnetic metals. Due to the grating coupling method, SPPs can be excited in a wide spectral range. At present, the ρ values in MPCs are quite high and exceed 1% [13], however, further increase of this parameter is desired for various applications in nanophotonics and the search for novel structures and approaches continues.
Fig. 1 (a) Magnetoplasmonic crystal (MPC). (b) Magnetoplasmonic crystal waveguide (MPCW). Blue and red colors show the Hy component distribution for symmetric and antisymmetric bound modes, correspondingly.
The magnitude of considered magneto-optical effect depends on two factors: the magneto-optical parameter Q and the SPP resonance quality. The first one is expressed as Q = εxz/ε, where ε is the diagonal element of the permittivity tensor in the magnetic material, and εxz is the transverse magnetization-induced non-diagonal element [15]. This parameter purely depends on the choice of the magnetic material. The second factor is defined by the materials used in the structure, the quality of the interfaces and the design of an MPC. High-quality resonances can be obtained in plasmonic waveguides, where a thin metal film is surrounded by identical dielectrics (insulator-metal-insulator, IMI). In such systems SPP excitations on both sides of a thin metal film couple and give rise to the symmetric and antisymmetric bound modes also known as long-range and short-range SPPs [16]. By using magnetic dielectric in such a structure, one can possibly achieve large magneto-optical effects in a high-quality structure. This idea was theoretically considered in a number of papers [17–20]. An important result of Ref. [17] is that in order to obtain magneto-induced shifts of SPP resonances in a magnetic IMI waveguide, the structure should not be purely symmetric. The authors suggest using cladding dielectric layers with either slightly different dielectric permittivities or with opposite magnetization directions. Recently a magnetoplasmonic structure with iron garnet layer bound by gold and titanium nitride layers was reported to demonstrate high-quality resonance in the transverse Kerr effect [10].
In this Letter we suggest and implement a novel technological method for the fabrication of a garnet-based magnetoplasmonic crystal waveguide (MPCW), in which a gold grating is placed between two iron garnet (IG) layers: epitaxially grown bismuth-lutetium IG (BiLuIG) and sputtered bismuth IG (BiIG) cladding layer. Following the recipe described above, it combines a slight ”asymmetry” in the parameters of the garnet layers and a strong magneto-optical activity.
2. Experimental sample
The structure of the MPCW is schematically shown in Fig. 1(b). To fabricate the MPCW we used a (Bi,Lu)3(Fe,Ga)5O12 film grown by liquid phase epitaxy on Gd3Ga5O12 (GGG) substrate. Approximately 50 nm thick gold layer was ion-beam sputtered on the BiLuIG surface. Periodical slits with 600 nm period and 100 nm width were etched in gold with focused ion beam resulting in the fabrication of an Au/garnet MPC. Figure 2(a) shows a scanning electron microscopy (SEM) upper view of the MPC. Bi3Fe5O12 film was then ion-beam sputtered on the surface of the MPC. BiIG was crystallized as a high-density polycrystalline layer through vacuum fast annealing: gas composition 80%N2+20%O2, temperature T=580 C° (limited by the requirement for the gold grating to retain its geometry), duration 3÷5 min, pressure not above 20 Torr. Figures 2(b) and 2(c) show SEM upper view and cross-section of a resulting magnetoplasmonic crystal waveguide.
Fig. 2 SEM images of the gold grating on top of the BiLuIG layer before (a) and after (b) BiIG deposition. (c) SEM image of the MPCW cross-section. The platinum layer was deposited for the cross-section measurements. (d) Normal-incidence transmission spectra for the BiLuIG (black curve) and BiIG (green curve) films. Red curve is the analytical fit [14]. Transverse Kerr effect hysteresis loops for BiLuIG (e) and BiIG (f) films.
Figure 2(d) shows normal-incidence transmission spectra measured for the BiLuIG and BiIG films deposited on the GGG substrate under the same conditions as in the case of an MPCW fabrication. Using analytical expressions for transmission of a thin film with low losses [14] we have fitted the spectrum of a BiIG film in the transparency region (600 1000 nm, red line in Fig. 2(d)) and evaluated the dispersion of dielectric susceptibility (ε3 = 6÷.20 + 0.07i at λ = 800 nm) and the film thickness (hd = 130 nm). The spectrum for a BiLuIG film could not be fitted within this model due to a non-homogeneous thickness dependence of its refractive index.
Using a temperature-stabilized laser diode with 460 nm wavelength we have performed measurements of the transverse Kerr effect. An electromagnet was used to create the transverse magnetic field and p-polarized light was incident at 45 degrees on the sample. The magnetic field induced modulations of the intensity of the reflected light R(M) were registered using a photodiode. Hysteresis loops for a BiLuIG epitaxial film and BiIG film on a GGG substrate are shown in Figs. 2(d) and 2(e), correspondingly, with ΔR = R(M)/R(0) − 1. It is seen that the magneto-optical activity of the sputtered BiIG film is at least one order of magnitude lower than that of the epitaxially grown BiLuIG. The saturating magnetic field is 1 kG for BiLuIG and 250 G for BiIG. Lower values of the magneto-optical effect and the saturating magnetic field can be explained by the polycrystalline structure of the sputtered film. Optical and magneto-optical properties of the obtained thin BiIG film are similar to the previously reported [21, 22].
3. Experiment
The fabrication method allowed us to compare the properties of the structure prior to the formation of the upper garnet layer (Au/BiLuIG MPC) and after it (BiIG/Au/BiLuIG MPCW) (Fig. 1). The transmission and magnetic contrast spectra were measured using a Halogen lamp. The radiation was spatially filtered with an optical fiber and was polarized in the incidence plane. The light was focused in a 50 µm spot on the sample placed on a rotation stage between the poles of a permanent magnet. The transverse magnetic field of 3 kG could be reversed by rotating the magnet. Transmitted light spectrum was detected with a spectrometer and normalized to the transmission spectrum of the garnet sample without the gold grating.
Figure 3(a) shows transmission of an MPC versus the wavelength and the incidence angle. Solid and dashed curves correspond to Au/BiLuIG and Au/air SPP excitations calculated using the SPP dispersion relation kSPP = k0[εmεd/(εm + εd)]1/2 and the phase matching condition kSPP k0 sin θ +2πm/p, where k0 is the light wave vector in vacuum, εm and εd are the dielectric constants of gold [24] and the dielectric, θ is the angle of incidence and m is an integer number corresponding to the excitation order. The dielectric constant of BiLuIG was varied in the reasonable range [23] to match the SPP resonances (εd = 5.21 + 0.03i at λ = 800 nm). The magnetic contrast ρ (Fig. 3(b)) is enhanced in the regions of Au/BiLuIG SPP excitation and exceeds 1%. The magnitude of the considered effect for a bare BiLuIG is of the order of 10−4. The corresponding spectra for the MPCW are shown in Figs. 3(c) and 3(d). The MPCW spectrum contains less number of spectral resonances: the SPP at the Au/air interface almost disappears and the waveguide modes in the BiLuIG layer are suppressed. The reason for a residual trace of the Au/air SPP excitation may be in the existence of air bubbles (∼50 nm diameter) in BiIG and Au films (dark spots in Figs. 2(b), 2(c)). The magnetic contrast demonstrates a complicated form, which will be further explained by the interaction of SPPs on the opposite sides of the gold layer.
Fig. 3 Transmission (a) and magnetic contrast (b) of an MPC versus wavelength and incidence angle. Solid curves correspond to calculated regions for Au/BIG (solid curves) and Au/air (dashed curves) SPP excitation. Transmission (c) and magnetic contrast (d) of a MPCW versus wavelength and incidence angle. Solid curves indicate calculated regions for long-range (blue curves) and short-range (red curves) SPP excitation.
Figure 4 shows a cross section of the measured 2D spectra for a 16° incidence angle in a wavelength range from 780 nm to 920 nm. In the case of an MPC the transmission spectrum demonstrates a characteristic asymmetric Fano-shape resonance with maximum and minimum. Corresponding magnetic contrast enhancement also has an asymmetric shape with a maximal absolute value of 1.2% and the full width at half minimum (FWHM) of 30 nm. After application of the BiIG layer both the transmission and magnetic contrast spectra change. Transmission in the vicinity of the SPP clearly increases, which can be explained by a better coupling of light at the opposite surfaces of the gold grating. The magnetic contrast for MPCW demonstrates one sharp blue-shifted minimum reaching −1% with 10 nm FWHM, and a broad red-shifted maximum reaching 0.6 % with 60 nm FWHM. Narrow features at shorter wavelengths correspond to the waveguide modes in the BiLuIG layer, which are also sensitive to the magnetization and thus make a weak contribution to the magneto-optical effect.
Fig. 4 Transmission (a) and magnetic contrast (b) spectra for MPC (black curves) and MPCW (green curves) for 16° angle of incidence. Vertical dashed lines are guide to the eye.
4. Discussion
For analytical calculations we used an MPCW model shown in Fig. 5(a). Gold and BiIG layers were bounded by semi-infinite BiLuIG and air regions. Based on Maxwell equations and boundary conditions we obtain the following equation for the dispersion of the bound modes:
where , βi = γi/εi and the subscript i corresponds to layers as noted in Fig. 5(a). Using experimentally obtained parameters of the MPCW (see Fig. 2) we numerically solve the Eq. (1) and obtain the dispersion for symmetric and antisymmetric bound modes shown in Fig. 5(b). High-quality symmetric mode has longer propagation length than the low-quality antisymmetric mode as seen from the mode propagation length dependence on the gold layer thickness hm for 1.55 eV photon energy (Fig. 5(c)).Fig. 5 (a) The MPCW model with Hy distributions of symmetric (blue) and antisymmetric (red) modes calculated for 1.55 eV photon energy (800 nm wavelength). (b) Dispersion curves for a single-interface Au/BiLuIG SPP (black curve), for symmetric (blue curve) and antisymmetric (red curve) modes supported by a 50 nm Au film. Dashed line shows the light dispersion in BiLuIG. (c) Propagation length for symmetric (blue curve) and antisymmetric (red curve) bound modes versus the gold film thickness for 1.55 eV photon energy.
Our magnetoplasmonic crystal waveguide (see Fig. 2) is asymmetric with respect to both the refractive index of two dielectrics and to their magnetization in the external saturating field. This should result in sensitivity of the bound modes to the external transverse magnetic field [17], which is confirmed by the experimental results shown in Figs. 3,4. Solid curves on Figs. 3(c), 3(d) correspond to the calculated dispersion of the MPCW (Fig. 5(b)) and are in a good agreement with the positions of the observed high-quality and low-quality resonances of the MPCW.
Narrow feature in the magnetic contrast associated with the symmetric bound mode can be used for sensing applications. By optimizing the MPCW parameters one can achieve a more narrow resonance and larger penetration depth of the mode into the adjacent layers and, particularly, in air, so that the resonance would become sensitive to the refractive index on the upper (BiIG/air) boundary of the MPCW. Further, the cladding BiIG layer protects the gold grating, which results in higher durability of the magnetoplasmonic structure. At the same time the antisymmetric mode provides large values of the magnetic contrast in a broad spectral range, which can be useful for applications with femtosecond lasers that have large spectral width of the pulse.
Summing up, we have presented a new type of magnetoplasmonic structure: magnetoplasmonic crystal waveguide composed of a gold grating bounded by magnetic dielectrics. Magneto-optical response of the structure is strongly modulated by both symmetric and antisymmetric bound modes, the former reveals a narrow resonant peak in the magnetic contrast. The results are promising for the development of high-sensitivity refractive-index and magnetic field sensors.
Funding
Russian Foundation for Basic Research (RFBR) (16-08-00933, 16-29-05204, 16-02-01060).
References and links
1. G. Armelles, A. Cebollada, A. García-Martín, and M.U. González, “Magnetoplasmonics: Combining Magnetic and Plasmonic Functionalities,” Adv. Opt. Mater. 1, 10–35 (2013). [CrossRef]
2. S. Tomita, T. Kato, S. Tsunashima, S. Iwata, M. Fujii, and S. Hayashi, “Magneto-optical Kerr effects of yttrium-iron garnet thin films incorporating gold nanoparticles,” Phys. Rev. Lett. 96, 167402 (2006). [CrossRef] [PubMed]
3. G. Armelles, A. Cebollada, A. García-Martín, J.M. García-Martín, M.U. González, J.B. González-Díaz, E. Ferreiro-Vila, and J.F. Torrado, “Magnetoplasmonic nanostructures: systems supporting both plasmonic and magnetic properties,” J. Opt. A: Pure Appl. Opt. 11, 114023 (2009). [CrossRef]
4. A.V. Baryshev, H. Uchida, and M. Inoue, “Peculiarities of plasmon-modified magneto-optical response of gold-garnet structures,” J. Opt. Soc. Am. B 30, 2371–2376 (2013). [CrossRef]
5. I.A. Kolmychek, A.N. Shaimanov, A.V. Baryshev, and T.V. Murzina, “Magneto-optical response of two-dimensional magnetic plasmon structures based on gold nanodisks embedded in a ferrite garnet layer,” JETP Lett. 102, 46–50 (2015). [CrossRef]
6. G.A. Wurtz, W. Hendren, R. Pollard, R. Atkinson, L.L. Guyader, A. Kirilyuk, T. Rasing, I.I. Smolyaninov, and A.V. Zayats, “Controlling optical transmission through magneto-plasmonic crystals with an external magnetic field,” New J. Phys. 10, 105012 (2008). [CrossRef]
7. V.V. Temnov, G. Armelles, U. Woggon, D. Guzatov, A. Cebollada, A. Garcia-Martin, J.M. Garcia-Martin, T. Thomay, and A. Leitenstorfer, “Active magneto-plasmonics in hybrid metal–ferromagnet structures,” Nat. Photonics 4, 107–111 (2010). [CrossRef]
8. V.I. Belotelov, I.A. Akimov, M. Pohl, V.A. Kotov, S. Kasture, A.S. Vengurlekar, A.V. Gopal, D.R. Yakovlev, and A.K. Zvezdin, “Enhanced magneto-optical effects in magnetoplasmonic crystals,” Nat. Nanotechnol. 6, 370–376 (2011). [CrossRef] [PubMed]
9. D.O. Ignatyeva, G.A. Knyazev, P.O. Kapralov, G. Dietler, S.K. Sekatskii, and V.I. Belotelov, “Magneto-optical plasmonic heterostructure with ultranarrow resonance for sensing applications,” Sci. Reports 6, 28077 (2016). [CrossRef]
10. J. Qin, Y. Zhang, X. Liang, C. Liu, C. Wang, T. Kang, H. Lu, L. Zhang, P. Zhou, X. Wang, B. Peng, J. Hu, L. Deng, and L. Bi, “Ultrahigh figure-of-merit in metal–insulator–metal magnetoplasmonic sensors using low loss magneto-optical oxide thin films,” ACS Photonics 4, 1403–1412 (2017). [CrossRef]
11. A.L. Chekhov, V.L. Krutyanskiy, A.N. Shaimanov, A.I. Stognij, and T.V. Murzina, “Wide tunability of magneto-plasmonic crystals due to excitation of multiple waveguide and plasmon modes,” Opt. Express 22, 17762–17768 (2014). [CrossRef] [PubMed]
12. I. Razdolski, S. Parchenko, A. Stupakiewicz, S. Semin, A. Stognij, A. Maziewski, A. Kirilyuk, and T. Rasing, “Second-harmonic generation from a magnetic buried interface enhanced by an interplay of surface plasma resonances,” ACS Photonics 2, 20–26 (2015). [CrossRef]
13. O.V. Borovkova, H. Hashim, M.A. Kozhaev, S.A. Dagesyan, A. Chakravarty, M. Levy, and V.I. Belotelov, “TMOKE as efficient tool for the magneto-optic analysis of ultra-thin magnetic films,” Appl. Phys. Lett. 112, 063101 (2018). [CrossRef]
14. R. Swanepoel, “Determination of the thickness and optical constants of amorphous silicon,” J. Phys. E: Sci. Instruments 16, 1214–1222 (1983). [CrossRef]
15. A. Zvezdin and V.A. Kotov, Modern magnetooptics and magnetooptical materials (CRC, 1997). [CrossRef]
16. A.E. Craig, G.A. Olson, and D. Sarid, “Experimental observation of the long-range surface-plasmon polariton,” Opt. Lett. 8, 380–382 (1983). [CrossRef] [PubMed]
17. B. Sepulveda, L. Lechuga, and G. Armelles, “Magnetooptic effects in surface-plasmon-polaritons slab waveguides,” J. Light. Technol. 24, 945–955 (2006). [CrossRef]
18. Y.-C. Lan and C.-M. Chen, “Long-range surface magnetoplasmon on thin plasmon films in the Voigt configuration,” Opt. Express 18, 12470–12481 (2010). [CrossRef] [PubMed]
19. G. Mathew, C. Bhagyaraj, A. Babu, and V. Mathew, “Effect of gyrotropic substrates on the surface plasmon polaritons guided by metal films of finite width,” J. Light. Technol. 30, 273–278 (2012). [CrossRef]
20. M. Khatir and N. Granpayeh, “An exact analysis method of SPP propagation in the anisotropic magneto-optic slab waveguides: I. Transversal configuration,” Optik - Int. J. for Light. Electron Opt. 124, 276–281 (2013). [CrossRef]
21. N. Adachi, V.P. Denysenkov, S.I. Khartsev, A.M. Grishin, and T. Okuda, “Epitaxial Bi3Fe5O12(001) films grown by pulsed laser deposition and reactive ion beam sputtering techniques,” J. Appl. Phys. 88, 2734–2739 (2000). [CrossRef]
22. T. Tepper and C. Ross, “Pulsed laser deposition and refractive index measurement of fully substituted bismuth iron garnet films,” J. Cryst. Growth 255, 324–331 (2003). [CrossRef]
23. V. Doormann, J.P. Krumme, C.P. Klages, and M. Erman, “Measurement of the refractive index and optical absorption spectra of epitaxial bismuth substituted yttrium iron garnet films at UV to near-IR wavelengths,” Appl. Phys. A 34, 223–230 (1984). [CrossRef]
24. P.B. Johnson and R.W. Christy, “Optical constants of the noble metals,” Phys. Rev. B 6, 4370–4379 (1972). [CrossRef]
