Computational alloy design of (Co1-xNix)88Zr7B4Cu1 nanocomposite soft magnets

Abstract

The dependence of coercivity on composition is an important factor for establishing optimized soft magnetic properties. In this study, we have used the random anisotropy and coherent rotation models to estimate the variation of coercivity with composition in (Co1-xNix)88Zr7B4Cu1 nanocomposite alloys. Our calculations that the magnetoelastic anisotropy contribution to coercivity dominates for Ni rich compositions (x > 0.5). A small range of compositions (0.65 < x < 0.75) is predicted to result in low values of coercivity (<10 A/m). To validate this prediction, (Co1-xNix)88Zr7B4Cu1 nanocomposites in this range were prepared by melt spinning followed by 3600 s isothermal annealing at the primary crystallization peak temperature (∼673 K). Hysteresis loops were measured using vibrating sample magnetometry at room temperature and saturation magnetostriction was measured using a strain gage based magnetostrictometer. Moderately small coercivities (30-40 A/m) and magnetostrictions (3-4 ppm) were measured at for samples with 0.685 < x < 0.725. Our measured coercivity had a minimum value of 32 A/m at x = 0.725, a shift in composition of about 5 at% in the direction of higher Ni content and without the anticipated low value of coercivity. Several reasons for the inaccuracy of this approach are described, including: ignored contributions from amorphous phase (especially in magnetoealstic anisotropy), composition segregation during crystallization leading to unpredictable compositional shifts in prediction, and the general observation that the predictability of minimum coercivity from minimal combined anisotropies has unexplained deviation even in far less complicated materials.