Femtosecond intrapulse evolution of the magneto-optic Kerr effect in magnetoplasmonic crystals

In magnetoplasmonics, it is possible to tailor the magneto-optical properties of nanostructures by exciting surface plasmon polaritons (SPPs). Thus far, magnetoplasmonic effects have been considered to be static. Here, we describe ultrafast manifestations of magnetoplasmonics by observing the nontrivial evolution of the transverse magneto-optic Kerr effect within 45-fs pulses reflected from an iron-based magnetoplasmonic crystal. The effect occurs for resonant SPP excitations, displays opposite time derivative signs for different slopes of the resonance, and is explained with the magnetization-dependent dispersion relation of SPPs.

Figure 1

Illustration of the ultrafast time-dependent TMOKE. The incident pulse is transformed into an SPP wave that has the dispersion law depending on the magnetization direction of the sample. The influence of SPPs is stronger at later time moments. Consequently, the reflected pulse profile depends on the sample magnetization, thereby yielding an intrapulse time-dependent TMOKE.

Figure 2

(a) Experimental setup for observation of the femtosecond evolution of TMOKE. (b) Reflection spectrum of the $p$-polarized light from the sample as a function of the angle of incidence. The rectangle indicates the angle of incidence and the spectral range chosen for the experiment. (c) Reflectance and static TMOKE spectra as measured with a Ti:sapphire laser in the cw regime. The dashed curves denote the spectra of the femtosecond laser pulses with carrier wavelengths ${\lambda }_c=795$, 808, and 822 nm in blue, green, and red, respectively.

Figure 3

(a) CFs measured for the $p$-polarized pulses at ${\lambda }_c=816$ nm for an in-plane external magnetic field in the forward (green curve) and backward (blue curve) directions. The time dependence of the TMOKE is represented by the measured $\Delta \left(\tau \right)=\left[I\right(+M)-I(-M\left)\right]/I\left(0\right)$ given by red dots and corresponding standard error bars. (b) Modeling results for the corresponding experimental data using Eqs. (10) and (14) (see the text for the parameter values). The time-dependent TMOKE, $\delta \left(t\right)$, is defined by Eq. (3); its CF counterpart $\Delta \left(\tau \right)$ is defined by Eq. (7).

Figure 4

(a) Intensity CF width measured with the correlation setup (solid dots) and with an autocorrelator placed in front of the sample (open dots) as a function of the carrier wavelength. The black dashed curve denotes the Fourier-transform limit. The gray curve shows the reflectance spectrum. (b)–(h) Correlation functions (curves) and the time-dependent TMOKE represented by the $\Delta \left(\tau \right)$ function, as denoted in Eq. (7) (dots with error bars), for the pulses with carrier wavelength values of 784, 795, 802, 808, 816, 822, and 830 nm, respectively.