Intertwined magneto-optical and plasmonic effects in Ag/Co/Ag layered structures

Surface plasmon polariton (SPP) excitation effects on the magneto-optical (MO) activity of Au capped Ag/Co/Ag trilayers are studied as a function of Co thickness. An enhancement of the transverse MO Kerr signal under SPP excitation as compared with that obtained without SPP excitation is measured with a maximum value of 150 times obtained for the trilayer with 8 nm Co. Such enhancement on the magneto-optical activity due to SPP excitation is also five times higher than that obtained in Au/Co/Au trilayers in similar conditions. The lower optical absorption in the studied range and the sharper plasmon resonance of Ag vs Au are responsible for these values. On the other hand, magnetic field-induced SPP wavevector modulation ${\left(\Delta k/k\right)}_{\text{SPP}}$ is studied for these trilayers and compared both with previous results in the Au/Co/Au system as well as with the theory. In the wavelength considered here, the obtained values are similar for both Ag- and Au-based structures and on the order of ${10}^{-4}$, pinpointing the role of the magnetic layer on the SPP wave-vector modulation.

Figure 1

(Color online) (a) AFM image of the 5 nm Co thick sample. (b) AFM profiles of two Au/Ag/Co/Ag multilayers with 2 and 11 nm Co thickness, respectively. The topographic profile of a glass substrate is also included as a reference. (c) Evolution of the mounts height with the Co thickness.

Figure 2

(Color online) (a) Kerr loops of some representative samples (4–6 and 9–11 nm Co thicknesses) measured in polar (left side of the figure) and transversal (right side of the figure) configuration. (b) Evolution of the polar saturation field and the transverse coercivity with Co thickness.

Figure 3

(Color online) Optical constants and magneto-optical constants (real part ${\epsilon }_1$ and imaginary part ${\epsilon }_2$) of the metals composing the multilayers analyzed in this work, obtained by ellipsometric (optical constants) and polar Kerr (MO constants) spectra: (a) Au; (b) Ag; (c) Co; and (d) MO constants of Co. All the data have been extracted from 20-nm-thick films. In the case of Ag and Co, they were covered by a 10 nm Au layer to prevent oxidation.

Figure 4

(Color online) (a) Rotation and (b) ellipticity spectra of the Au/Ag/Co/Ag layers for four representative values of Co thickness: 2, 5, 8, and 11 nm.

Figure 5

(Color online) Theoretical Kerr rotation and ellipticity spectra for continuous Au/Ag/Co/Ag films for the same Co thicknesses as in Fig. 4 (symbols) and a nanostructured (Co islands on an Ag matrix) thin film (crosses) to estimate the effect of the layers continuity loss. The inset focuses on the high-energy region for the continuous and nanostructured film with the same Co amount (2nm).

Figure 6

(Color online) (a) Schematic view of the measurement configuration, allowing SPP excitation if light is incident from the prism side at the appropriate angle (Kretschmann configuration). (b) Measurements of the R and transverse MO Kerr effect signal $\left(\Delta \text{R}/\text{R}\right)$ for the 8 nm Co thickness sample.

Figure 7

(Color online) Co thickness dependence of the transversal MO Kerr signal $\left(\Delta \text{R}/\text{R}\right)$. (a) Experimental $\Delta \text{R}/\text{R}$ versus incidence angle curves for different Co thickness upon SPP excitation. (b) Co thickness dependence of the ${\left(\Delta \text{R}/\text{R}\right)}_{\text{Max}}$, obtained both experimentally (lines with open symbols) and from simulations (lines). The continuous blue line and the dashed red line with squares correspond to the $\Delta \text{R}/{\text{R}}_{\text{Max}}$ values upon SPP excitation, whereas the dashed-pointed green line and the dashed black line with circles correspond to ${\left(\Delta \text{R}/\text{R}\right)}_{\text{Max}}$ values multiplied by ten times without SPP excitation. (c) Comparison between Au/Co/Au (squares) and Au/Ag/Co/Ag (circles) structures. The Au/Co/Au curve comes from Ref. 16. In both multilayers, the nominal Co thickness corresponds to the case of maximum SPP excitation and thus maximum $\Delta \text{R}/{\text{R}}_{\text{Max}}$ (5.5 nm Co for Au/Co/Au and 8 nm for Au/Ag/Co/Ag).

Figure 8

(Color online) (a) Simulated reflectivity versus angle of incidence curves obtained for the sample 5nmAu/5nmAg/7nmCo/7nmAg/Glass with the magnetization saturated in one direction ${\theta }_{\text{min}}\left(H_{+}\right)$ (continuous line) and in the opposite direction ${\theta }_{\text{min}}\left(H_{-}\right)$ (dashed line). The angular shift due to the magnetic field modulation, $\Delta \theta ={\theta }_{\text{min}}\left(H_{+}\right)-{\theta }_{\text{min}}\left(H_{-}\right)$, is shown. (b) Angular derivative of the experimental reflectivity curve divided by the reflectivity and times $\Delta \theta$ [i.e., $\left(\delta \text{R}/\delta \theta \right)\Delta \theta /\text{R}$] compared with the experimental transverse MO Kerr signal $\left(\Delta \text{R}/\text{R}\right)$. $\Delta \theta$ is the only fitting parameter. (c) Co thickness dependence of the ${\left(\Delta k/k\right)}_{\text{SPP}}$ experimental (points) and simulated (line) values. The dashed line corresponds to the simulation with the Au/Co/Au structure.