The critical phase transition in ferromagnetic ultrathin Fe/W(110) films has been studied using the magnetic ac susceptibility. A statistically objective, unconstrained fitting of the susceptibility is used to extract values for the critical exponent γ, the critical temperature Tc, the critical amplitude χo and the range of temperature that exhibits power-law behaviour. A fitting algorithm was used to simultaneously minimize the statistical variance of a power law fit to individual experimental measurements of χ(T). This avoids systematic errors and generates objective fitting results. An ensemble of 25 measurements on many different films are analyzed. Those which permit an extended fitting range in reduced temperature lower than approximately 4.75 × 10 −3 give an average value γ=1.76 ±0.01. Bilayer films give a weighted average value of γ = 1.75 ± 0.02. These results are in agreement with the 2-dimensional Ising exponent γ= 7 4 . Measurements that do not exhibit powerlaw scaling as close to Tc (especially films of thickness 1.75ML) show a value of γ higher than the Ising value. Several possibilities are considered to account for this behaviour.

Power-law scaling of the relaxation time associated with critical slowing down has been experimentally measured in the dynamics of the magnetization of a bilayer of iron grown on top of a W͑110͒ substrate using the complex magnetic ac susceptibility ͑T͒. The observed value of the critical exponent for the slowing down above the Curie transition of this two-dimensional Ising ferromagnetic system is z = 2.09± 0.06 ͑95% confidence͒, in agreement with most contemporary theories and simulations. Further analysis reveals that dynamical effects cause ͑T͒ to deviate from power-law scaling as the temperature is decreased towards T c , whereas the saturation of the correlation length due to finite-size effects ͑on the order of 500 lattice spaces͒ limits the divergence of .

No abstract

The magnetic susceptibility, χ=∂M/∂H, can be determined for ultrathin films using the surface magneto-optic Kerr effect and an ac technique (ac-MOKE) where the response to a small, modulated H field is measured. Optimization of the signal-to-noise ratio is imperative if the measurements are to be used to extract critical scaling exponents from the small tail of χ in the paramagnetic region. These optimal conditions are different than those appropriate for more conventional static H measurements (dc-MOKE) used to generate magnetic hysteresis loops, because the relative noise contributions from the laser stability, sample stability, and detector are changed. A simple method for determining optimum conditions is demonstrated, and the technique is used to show that a 1.8 ML Fe/W(110) film has the critical exponent γ of the 2D Ising model.