Abstract
Co-Pt alloys with compositions in the narrow eutectoid range (~ Co40Pt60) can undergo eutectoid transformation via the pseudospinodal decomposition mechanism to produce a highly regular nanochessboard structure. This microstructure comprises the hard magnetic L10 phase (CoPt) and the soft magnetic L12 phase (CoPt3) interleaved on the nanoscale. The resulting periodic nanocomposite exhibits exchange coupling between its hard and soft ferromagnetic constituents, thus suggesting a potential for high magnetic energy storage. The magnetic response of Co-Pt nanochessboards has, surprisingly, not been characterized so far. This investigation focuses on establishing and understanding interrelationships between processing variables and the resulting microstructure, magnetic properties in near-eutectoid Co-Pt alloys. The effects of thermal processing were characterized using X-Ray Diffraction (XRD), Vibrating Sample Magnetometry (VSM) and Transmission Electron Microscopy (TEM).
Since the phases obtained from eutectoid decomposition are non-stoichiometric, we first studied the individual Co41.7Pt58.3 and Co37.6Pt62.4 compositions that lie slightly outside the two-phase coexistence region. The magnetic properties of Co41.7Pt58.3 were observed to vary intricately as a function of annealing time and temperature. These changes corresponded to changes in microstructural length-scales, phases present, magnetocrystalline anisotropy of the ordered phase and dominant pinning mechanisms. This alloy exhibited a maximum coercivity of 4 kOe along with a maximum remanence to saturation ratio of 0.8 upon annealing at 600 ◦C for 4 days, and a microstructure showing tweed contrast on a ~ 15 nm lengthscale. The peak coercivity alloy was established to be in the initial stages of phase transformation, containing fine L10 particles with low tetragonality embedded in the disordered FCC matrix. Despite the fine particles sizes, we argue for a domain wall pinning model based on gradients in the wall energy. This arises from exchange coupling to the soft FCC matrix, and demonstrates good agreement with the experimental coercivity, highlighting the contribution of nanoscale modulation to magnetic hardness. As the phase transformation continued towards completion, the microstructure evolved into the polytwinned structure associated with lower coercivity compared to the peak coercivity sample. Longer annealing at 700 °C led to the formation of the L12 ordered phase in addition to L10. The Co37.6Pt62.4 alloy ordered exclusively to the soft L12 phase.
Chessboard microstructure was observed in Co40.2Pt59.8 and Co38.8Pt61.2 alloys. Slow cooling a Co40.2Pt59.8 sample from 750 ºC to 600 ºC at 40 ºC/day followed by aging for 1 week produced a chessboard structure associated with maximum coercivity, 2300 Oe, along with maximum remanence to saturation ratio, 0.66. An analysis of the corresponding magnetic susceptibility curve and XRD indicated the co-existence of both ordered phases as well as the disordered phase at this stage, wherein the hard L10 and soft L12 were exchange-coupled, but the untransformed FCC phase was uncoupled. This would require the residual FCC phase to be spatially segregated, either in macroscopic regions or in regions near the peripheries of the chessboard colonies. Overaging led to an apparent broadening/breakdown of the chessboard structure, but clear trends with process conditions are yet to be isolated.
Most of our two-phase samples (L10 + FCC or L10 + L12) demonstrated strong evidence of exchange coupling. However, recoil analysis showed unexpectedly limited reversibility in all cases. This may be due to the dominant effect of domain wall motion in controlling the magnetization reversal.
