Fig. 2: Out-of-equilibrium magnetic phase diagrams.

a Temperature dependence of the change in second harmonic generation (SHG) intensity ΔI(2ω) at t = 0.3 ps, normalized to the un-pumped value at T = 80 K, for select pump excitation densities n_ex. b Out-of-equilibrium phase diagram mapped at t = 0.3 ps by either sweeping T or n_ex. Black squares mark the temperature onset of the surface magnetization-induced electric-dipole (SMIED) contribution at different fluences, T_N(n_ex). Dashed curve and blue-orange background are guides to the eye. The reported La-doping dependence of T_N is overlaid. c Comparison of the temperature dependence of ΔI(2ω) for an un-pumped sample with the fluence (F) dependence of ΔI(2ω) collected at t = 10 ps and T = 80 K, where F is converted into an effective temperature by assuming all of the pump energy goes into quasi-equilibrium heating. Dashed curve is a guide to the eye. d Out-of-equilibrium phase diagram mapped at t = 10 ps (black squares). Dashed black line and blue-orange background are guides to the eye. Overlaid are the phase boundary reproduced from panel b (white line) and that calculated assuming all of the pump energy goes into quasi-equilibrium heating (red line). e High fluence ΔI(2ω) and f magneto-optical Kerr effect transients acquired below and above T_N. Vertical error bars in panels a, c represent the range of intensity values obtained upon selecting different regions of the CCD camera for radial integration and background subtraction in order to convert raw data into SHG-RA intensity patterns. Horizontal error bars in panels c, d represent the uncertainty in fluence due to laser drift. Horizontal (vertical) error bars in panel b (panels b, d) represent the uncertainty in our estimate of the critical n_ex (T_N(n_ex)).