Ultrafast selective mid-infrared sublattice manipulation in the ferrimagnet $FeCr_2S_4$

TL;DR

The study demonstrates ultrafast, sublattice-selective manipulation in the ferrimagnet FeCr2S4 by resonantly exciting Fe sublattices with mid-infrared pulses targeting Fe $d$-$d$ transitions, monitored via time-resolved MOKE. By comparing resonant $0.30$ eV pumping with non-resonant higher-energy pumping and tuning the probe energy relative to Fe $d$-$d$ transitions, the authors isolate Fe-sublattice contributions and reveal slower, sublattice-specific dynamics and fluence-dependent spin precession. The work also highlights coherent artifacts in degenerate mid-IR configurations and shows how excitation density and temperature-induced MO coefficient changes shape the pump-probe responses, including large relative changes near zero-crossings. Overall, the results establish ultrafast, sublattice-selective control in a ferrimagnetic spinel using mid-infrared resonant excitation, with implications for targeted ultrafast spin manipulation and a reminder of penetration-depth and MO-coefficient temperature effects in interpreting pump-probe signals.

Abstract

$FeCr_2S_4$ is a ferrimagnet with two oppositely ordered spin sublattices (Fe and Cr), connected via superexchange interaction, giving a non-zero net magnetic moment. We show, using time-resolved measurements of the magneto-optic Kerr effect, how the magnetic dynamics of the sublattices can be selectively manipulated by resonantly perturbing the Fe sublattice with ultrashort laser pulses. The mid-infrared excitation through intra-atomic Fe $d$-$d$ transitions triggers markedly slower dynamics in comparison to an off-resonant pumping affecting both of the two sublattices simultaneously. By changing probe wavelength to move in and out of resonance with the Fe $d$-$d$ transitions, we also show the specific contributions of the Fe sublattice to these dynamics.

Figures (14)

• Figure S1: Time-resolved reflectivity of the horizontal component in the (a) 0.30 eV pump - 0.30 eV probe (b) 3.10 eV pump - 3.10 eV probe wavelength combinations. The scans were recorded at $T$=75 K.
• Figure S2: (a) Power reflectivity and (b) penetration depth in FeCr2S4 as a function of the photon energy. The curves were obtained using the dielectric function data reported by Ogasawara et al. Ogasawara2006.
• Figure S3: Coherent artifacts in the dynamics for the pump 0.30 eV - probe 0.30 eV configuration. (a) 7 ps (b) 150 ps windows. The different colors represent subsequent scans. The scans were recorded at $T$=75 K under a 1.1 mJ/cm2 incident fluence.
• Figure S4: Fourier transform of the oscillatory response in Fig. 3 (0.30 eV probe photon energy) of the Main Text after subtracting a quadratic background in the time ranges (a) 1.2-6.0 ps, (b) 1.2-6.0 ps, (c) 10-130 ps and (d) 10-130 ps.
• Figure S5: Time-resolved MOKE signal for FeCr2S4 using a 0.30 eV pump and 0.30 probe photon energy before time zero. (a) Time domain (b) Frequency domain obtained through a Fourier transform of the data between -3.3 and -0.9 ps after subtracting a quadratic background. The scan was acquired at $T$=75 K.