Phase Equilibria of the Pd-Rich Fe-Pd Eutectoid Region: Shockley's Controversial L1' Phase and Order to Order Transformations

The ferromagnetically desirable L10 phase is typically studied at or near equiatomic compositions and exhibits magnetocrystalline anisotropies comparable to rare-earth permanent magnets in Fe-Pt, Co-Pt, and Fe-Pd. The eutectoid decomposition A1→L10+ L12 may yield self-assembled microstructures with coherent nm-lengthscale domains, which can be tuned to elicit desirable bulk magnetic properties. Despite nearly a century’s worth of study on the Fe-Pd system, the Pd-rich eutectoid region has never been explicitly explored.

The dissertation provides novel and direct evidence of the L1’ phase first predicted by Shockley in 1938, but never identified in bulk. L1’ is a second ordered tetragonal phase on the FCC lattice which exhibits site-specific placement of excess (off-stoichiometric) atoms on (½, ½, 0). Identification of this phase is afforded by low-angle superlattice reflections in X-ray diffraction measurements in conjunction with electron microscopy. Magnetic measurements suggest that spin structures between L10 and L1’ are different, and may contribute to L1’ phase stability.

Characterization and analysis of the Pd-rich two-phase region is also provided, where coexistence of L10 + L12 (at 650℃) and L1’ + L12 (at 525℃) is found. With a narrow two-phase region (~1 at%), small variations in composition dramatically influence equilibrium phase fractions of L10 + L12. As such, two emerging microstructures are presented. For L12-lean regions, an L10 polytwinned microstructure is found with nm-lengthscale cubic wetting layers along domain boundaries. For L12-rich regions, a unique matrix + plate morphology of the phases is found. The coexistence L1’ + L12 dispels the notion that the L1’ phase is a metastable hybrid-ordering of L10 and L12. The L1’ + L12 microstructure is found to also be polytwinned, with (thicker) regions of L12 wetting anti-phase and domain boundaries.

Furthermore, an order→order+order transformation is found, where solid solution FCC orders to single-phase L12 prior to any L10 formation. The tweed pre-transition state was observed within single-phase L12, and a polytwinned plate growth mechanism is hypothesized. The observation of tweed in an ordered phase is relatively novel, as the tweed pretransition state is typically considered disorder-driven and identified in solid solution.