Novel routes to new intermetallic compounds

S.H. Kilcoyne, P. Manuel (Univ. St. Andrews), C. Ritter (ILL).

Rare-earth transition-metal compounds play a central role in both the science and technology of magnetism. On the one hand, they offer the opportunity of varying magnetic exchange, anisotropy, conduction electron density and local environment as a means of probing the most fundamental aspects of spin fluctuations, moment formation and magnetic order, whilst on the other hand their properties can be modified and developed for applications such as high-density permanent magnets, magnetic storage media or giant magnetoresistive alloys for technological applications. Consequently, the structural phase diagrams of most binary rare-earth transition-metal systems have been studied in great detail and it is generally considered that most parent phases have been identified and classified and only minor modifications associated with perhaps interstitial substitution or site selectivity remain to be characterised. Indeed, no entirely new binary rare-earth transition-metal structural phases have been identified for some years. However, it should be noted that the majority of equilibrium phase diagrams have been determined under conditions that are far from equilibrium. Generally, binary mixes are annealed at pre-determined temperatures and subsequently quenched in the hope of retaining phases and phase proportions which are characteristic of those at the annealing temperature. High-intensity kinetic neutron diffraction, such as that afforded by the D20 diffractometer, can be used to circumvent this limitation by allowing an in situ exploration of the phase diagrams of binary rare-earth transition-metal alloys. The true equilibrium phase diagrams can be mapped in detail by studying phase formations and transformations at the temperatures at which they occur.

Figure 1: Neutron thermogram of crystallisation of Y7Fe3 on ramping the temperature from 250°C to 550°C. Note the presence of the intermediate Y-Fe phase between 400°C and 450°C. The final phase is Y plus YFe2.

A survey of binary rare-earth transition-metal phase diagrams using kinetic diffraction techniques is now underway. We have adopted an annealing procedure, involving amorphous rare&endash;earth transition-metal precursors, which has previously proved effective in the preparation of high quality superconducting cuprates. In the present studies high purity amorphous rare-earth transition-metal melt spun ribbons, in which an intimate mixing of the elemental components is assured, are crystallised in situ and the resulting phase formations monitored via neutron thermograms. An unexpected and exciting outcome of these studies has been the discovery of several hitherto unreported binary structural and magnetic phases.

Figure 2: Temperature dependence of the normalised intensities at the positions of the Y (100), YFe (100) and YFe2 (111) Bragg peaks.

As an example consider the kinetic neutron diffraction of the crystallisation of amorphous Y₇Fe₃ metallic ribbons. The ribbons were prepared at the University of St. Andrews by a radio-frequency melt spinning technique in which molten Y₇Fe₃ is ejected at high velocity onto a rapidly rotating copper wheel, leading to quench rates of the order of 10⁶ Ks^-1. Diffraction patterns were collected on D20 from only 3 g of ribbon in a vanadium can loaded into a furnace. The sample temperature was ramped smoothly from 250°C to 550°C at a rate of 60°C per hour with diffraction patterns collected every four minutes. The resulting neutron thermogram is shown in Fig. 1. It can be seen in the thermogram, or perhaps more clearly in Fig. 2 where the temperature dependences of the normalised intensities at the positions of characteristic Bragg peaks are plotted, that partial crystallisation of Y₇Fe₃ first occurs at approximately 300°C.

At this temperature a pure hexagonal-close-packed elemental Y phase is formed, with the Fe atoms segregating to form an Fe-rich amorphous phase. At 390°C, there is a dramatic crystallisation of the entire sample, witnessed by an abrupt loss of the amorphous contribution to the diffraction pattern. At this same temperature new Bragg peaks appear signifying the formation of an intermediate phase which coexists with the Y matrix to 450°C. At this temperature, the Bragg peaks associated with the intermediate phase rapidly decrease in intensity, there is a concomitant further increase in the intensity of the Y peaks, and the well known cubic C15 Laves phase YFe₂ emerges. The final crystallisation product is YFe₂ coexisting in equilibrium with crystalline elemental Y.

It is the intermediate phase that is of particular interest here. Intensity calculations identify its composition as stoichiometric YFe, a phase that is entirely absent from all published phase diagrams of the Y-Fe system. This absence is perhaps not surprising, as under equilibrium conditions the phase is truly stable only over a temperature range of 40°C, and requires the intimate atomic mixing of constituents afforded by the amorphous precursor in order to form. Using an ab initio two-phase Rietveld refinement we have established that the YFe phase crystallises with the hexagonal space group P63mc with unusually large cell dimensions of a = 12.90 Å and c = 11.71 Å and with = 120°.

We have since found that it is possible to stabilise this phase at room temperature, where it coexists in equilibrium with pure Y and we are currently attempting to identify the atomic positions associated with Y and Fe sites within this unusual structure. Our Mössbauer spectroscopy and magnetisation measurements have also revealed that the new YFe compound is a weak itinerant electron ferromagnet with a Curie temperature of 60 K and a magnetic moment of less than 1 µ_B.

In addition to the new YFe phase obtained from the crystallisation of amorphous Y₇Fe₃ ribbons, our kinetic recrystallisation measurements on D20 have provided evidence of many similar new and exotic binary magnetic and crystallographic phases in other systems such as Y-Co and Er-Fe. Such studies may pave the way to many new scientifically interesting and technologically important compounds in which the magnetic properties and associated functionality can be tailored for specific applications.

Finally, the results of these studies will also form a valuable input into theoretical ab initio caculations for the understanding of the interplay of structural, electronic and magnetic properties of precisely the families of intermetallic compounds upon which our investigations are based. In many cases the agreement between the predictions of augmented spherical wave and linear muffin-tin orbital calculations and experimental data are extremely good. However, the parameter space available to the theoreticians is often restricted by the paucity of appropriate stable intermetallic compounds against which the calculations can be tested. Our kinetic neutron diffraction studies are already providing novel but related crystallographic and magnetic phases which will undoubtedly stimulate further calculations and provide a stringent test of the predictive capabilities of these methods.