Magnetoelastic evidence for a local structural transformation in Co-rich glasses

Abstract

The magnetostriction of Fe-rich glasses is large and positive ( λ = +32×10 −6) whereas that of polycrystalline α-Fe is small and negative ( λ = -7×10 −6). This is understood to be a consequence of the sharp difference in local atomic order between these two materials: a 12-fold coordinated short-range order dominated by trigonal prismatic arrangements of Fe about B atoms in the glass, versus an eight-fold coordinated bcc structure in the crystal. Co-rich glases on the other hand show temperature and compositional dependences which closely parallel those of polycrystalline hcp (ϵ) cobalt and Co -Fe alloys. Specifically: (1) the addition of iron to ϵ-Co (polycrystalline λ = -16×10 −6) drives a martensitic transformation to the γ (fcc) phase in Co 1- x Fe x at approximately x = 0.07. Near this composition λ = 0 and for x 62 0.09, λ is positive. Similar behavior is observed with the addition of iron to amorphous Co 80B 20 ( λ = -4×10 −6) in the series (Co 1- x Fe x ) 80B 20 where λ passes through zero at x = 0.06. (2) The martensitic transformation from the hcp to the fcc phase can also be thermally driven and is observed in Co at 420°C, near which temperature λ again changes sign and is positive in the high-temperature phase. Similarly, amorphous Co 80B 20 shows by extrapolation to above crystallization [1] (and Co 78Fe 2B 20, Co 72Mn 8B 20 and Co 70Cr 10B 20 show explicitly [2]) a thermally induced transition to a positive magnetostriction phase at 400°C. To the extent that the magnetostriction of polycrystalline cobalt signals the occurrence of the long-range ε-γ transformation, the magnetostriction of amorphous cobalt-rich alloys can be taken to indicate a similar, but local, transformation in the glasses. These transformations are reversible in both the crystals and glasses. In the amorphous materials most other macroscopic manifestations of the cooperative, local atomic rearrangements involved in the transformation are masked by the random orientation of the local structural units. The electronic origin of the transformations in both crystalline and noncrystalline materials is suggested to be a Jahn-Teller mechanism involving two states close to E F which have orbital character similar to d x 2- y 2 (near E F- k B T/2) and d z 2- r 2 (near E F + k B T/2) levels. The ground state of the system is then characterized by an oblate spheroidal d-charge distribution and the ϵ phase (basal plane expansion, λ<0) is stable in the crystal and a similar local order is assumed to obtain in the glass. Increasing temperature (or shifting E F relative to these nearly degenerate levels by alloying) will tend to equalize their populations and drive the system to the higher-symmetry local order seen in the crystal as the γ phase.