Patent Application
20140251500
This invention provides new magnetic materials containing cerium, iron, and small additions of a third element(s), and comprising compounds Ce(Fe12−xMx) having the ThMn12 tetragonal crystal structure (space group I4/mmm, #139). Compounds with M=B, Al, Si, P, S, Sc, Ti, V, Co, Ni, Zn, Ga, Ge, Zr, Nb, Mo, Hf, Ta, and W are identified theoretically, and one class of compounds based on M=Si has been synthesized. The Si cognates are characterized by large magnetic moments 4πMs (above 1.27 Tesla) and high Curie temperatures (264≦Tc≦305° C.). Further enhancement in Tc and magnetic moment is obtained by nitriding the cerium compounds through heat treatment in nitrogen gas while retaining the ThMn12 crystal structure; for example CeFe10Si2N1.29 has Tc==426° C.
There remains a need for permanent magnet materials in electric motors for many applications and in other magnet-containing articles of manufacture. Cerium-iron compounds are attractive candidates to explore as potential permanent magnet materials. However, they have a low Curie temperature which will impede their use in major automotive applications (e.g., traction motors) because they will not retain sufficient magnetic properties in a device at elevated operating temperatures. It appears that if cerium-iron materials are to be thus utilized their compositions will have to be modified.
This invention provides a new series of Ce—Fe-based permanent magnet materials based on the presence in the material of a major portion of one or more compounds of the form Ce(Fe12−xMx), where M is one or more elements selected from the group consisting of B, Al, Si, P, S, Sc, Co, Ni, Zn, Ga, Ge, Zr, Nb, Hf, Ta, and W. The material is prepared with the Ce(Fe12−xMx) compound(s) in the form of a stable ThMn12 tetragonal crystal structure (sometimes referred to as 1-12) to provide the permanent magnet properties. Preferably the value of x is in the range of 1-4. Compounds containing an M element from the above listing may additionally include one or more of Ti, V, Cr, and/or Mo along with one or more of the M-constituents. In general, it is preferred that the Ce—Fe-M magnetic materials be prepared by a suitable process, such as by rapid solidification from a melt of the constituent elements, to achieve the presence of a major phase of Ce(Fe12−xMx) in the ThMn12 tetragonal crystal structure and with x in the range of 1-4.
The above listed elements, M, forming a stable ThMn12-type crystal structure with cerium and iron are identified in this specification using first-principles theoretical calculations based on Density Functional Theory (DFT) using the representative compound, CeFe8M4. In addition to the DFT calculations, examples of stable ThMn12-type compounds have been synthesized with M=Si having stoichiometries CeFe12−xSix (x=1, 1.5, and 2).
Permanent magnet alloys containing CeFe11Si, CeFe10.5Si1.5, and CeFe10Si2 were prepared by combining stoichiometric quantities of elemental Ce, Fe, and Si in an ingot. Ingots of these materials were then melted under inert gas and subjected to a rapid solidification process to form ribbon particles. The ribbon particles were comminuted to a powder and magnetically characterized. The magnetic moment (saturation magnetization) 4πMs may be approximated by the value of the magnetization 4πM at the largest applied magnetic field (H) of 1.9 Tesla; given that the magnetization is still slowly increasing with H at 1.9 Tesla, the values of 4πMs presented in this application thus represent lower limits to the actual saturation magnetization. The three CeFe12−xSix alloys were found to have large magnetic moments 4πMs=1.04 to 1.27 Tesla and Curie temperatures, 264° C.<Tc<305° C., which are higher than the Curie temperatures of any previously known Ce—Fe-based compounds. Curie temperatures are further improved by heat treatment under nitrogen gas to form the corresponding CeFe12−xMxNy nitrides, while retaining the ThMn12 crystal structure. The nitride CeFe10Si2N1.29 boasts a Curie temperature of 426° C. and a higher magnetic moment than its precursor, CeFe10Si2.
Accordingly, we have prepared specific CeFe12−xSix compositions where x=1, 1.5, and 2, and demonstrated that they possess useful permanent magnetic properties. And we have determined that a family of compositionally related compounds is likely to be formable in a like manner into useful permanent magnet materials. These related compounds are Ce(Fe12−xMx), where M is one or more elements selected from the group consisting of B, Al, Si, P, S, Sc, Co, Ni, Zn, Ga, Ge, Zr, Nb, Hf, Ta, or W. In these compounds it is preferred that x have a value in the range of one to four. Proportions of one or more of Ti, V, Cr, and Mo may be combined with or substituted for up to about ninety percent of one of the M elements in our Ce—Fe-M magnetic material; for example, CeFe10.25Si1.5Ti0.25.
The magnetic material may be prepared in powder form for compacting, molding, resin bonding, or other shaping practice into a useful permanent magnet body for use in an electric motor or other magnet application. Other objects and advantages of our invention will be apparent from the following sections of this specification.
First principles Density Functional Theory (DFT) was applied in order to computationally identify elements M for which CeFe12−xMx compounds having the prototypical tetragonal ThMn12-type crystal structure may form. In that structure the Th ions occupy 2a crystallographic sites; the Mn ions reside on 8i, 8j, and 8f sites. Neutron diffraction studies of known RFe12−xMx materials (R=rare earth) demonstrate that the M ions show distinct site preferences among the 8i, 8j, and 8f sites. Within the preferred crystallographic site, however, the Fe and M ions are disordered. Treating the intra-site disorder on such high occupancy sites is a daunting computational challenge. Instead, elements M that might stabilize the ThMn12 structure are qualitatively identified via a much more tractable approach: element M is assumed to fully occupy the 8i, 8j, or 8f sites in the ThMn12 structure, corresponding to the stoichiometry, CeFe8M4, and the enthalpy of formation, ΔH, is computed for each of the three cases. A negative AH suggests the formation of CeFe12−xMx.
All calculations reported here rely on DFT as implemented in the Vienna ab initio simulation package (VASP) within a plane wave basis set. Potentials constructed by the projector-augmented wave (PAW) method were employed for the elements; the generalized gradient approximation was used for the exchange-correlation energy functional. As a consequence of 4f shell instability, the cerium ion in intermetallic compounds is often in a mixed-valent, α-like state that is incompatible with a local 4f magnetic moment. In view of the fact that only 3+ (one 4f electron in a frozen core) and 4+ (one 4f electron treated variationally with two 5s, six 5p, and three 5d-6s electrons) PAW potentials are available in VASP, the latter was chosen as the preferable approximation for the materials studied. Lattice constants and atomic positions were optimized by simultaneously minimizing all atomic forces and stress tensor components via a conjugate gradient method. Dense reciprocal space meshes having spacings <0.10 Å−1 were used throughout. In all computations the plane wave cutoff energy was 900 eV, the total energy was converged to 10−6 eV per cell, and the force components relaxed to at least 10−4 eV/Å. No fewer than three successive full-cell optimizations were conducted to ensure that the structural parameters and cell energies were fully converged. Total energies were derived by integration over the irreducible Brillouin zone with the linear tetrahedron method.
The electronic total energies Eel obtained with VASP enable calculation of ΔHel(CeFe8M4), the standard enthalpy of CeFe8M4 formation at zero temperature in the absence of zero point energy contributions:
ΔHel(CeFe8M4)≡Eel(CeFe8M4)−Eel(Ce)−8Eel(Fe)−4Eel(M) (1).
In the case of the progenitor compound CeFe12 this yields
ΔHel(CeFe12)=Eel(CeFe12)−Eel(Ce)−12Eel(Fe)=11 kJ/mole CeFe12 (2);
the positive value is consistent with the experimental observation that CeFe12 does not form under normal conditions.
Table I presents ΔHel, the magnetic moment μ, and cell volume V calculated for CeFe8M4 with M one of 26 elements other than Fe populating the 8i, 8j, or 8f sites in the ThMn12 structure. The bold-data cells highlight the cases for which ΔHel is the most negative, indicating the greatest stability with respect to the elemental constituents, for a given M and lattice position.
The results suggest that CeFe12−xMx may be stabilized by M=B, Al, Si, P, S, Sc, Ti, V, Co, Ni, Zn, Ga, Ge, Zr, Nb, Mo, Hf, Ta, and W with Sc, Ti, V, Zr, Nb, Mo, Hf, Ta, W preferring the 8i site and B, Al, Si, P, S, Co, Ni, Zn, Ga, Ge preferring the 8j site. C, Na, Mg, Mn, Cu, and Sn are definitely not favorable in view of the large, positive ΔHel values. The small but positive ΔHel for CeFe8Cr4 (Cr filling the 8i site) is consistent with the fact that RFe12−xCrx compounds are known only for x≦2.
The findings are in qualitative overall agreement with experiment inasmuch as (i) CeFe12−xMx (M=Ti, V, Cr, Mo) compounds have been reported previously and (ii) CeFe12−xSix (x=1.0, 1.5, 2.0) has been synthesized as part of this work. Table I indicates that CeFe12−xMx (M=B, Al, P, S, Sc, Co, Ni, Zn, Ga, Ge, Zr, Nb, Hf, Ta, and W) merit attempts to synthesize as well. The Sc material, even if it were to form, is not interesting from a technological perspective in view of the scarcity and associated enormous cost of Sc. The M=Co, Ni, Zn, Ga, and Ge compounds, on the other hand, may be particularly interesting since their magnetic moments per formula unit in Table I are about twice those of the M=Ti, V, Cr, and Mo compounds, which would afford magnets with substantially greater energy products and likely larger Curie temperatures. The relatively large cell volume of CeFe8Zr4 may foreshadow the formation of trivalent Ce, which would have a 4f magnetic moment that would contribute to the overall magnetization and provide magnetocrystalline anisotropy.
Alloys of CeFe11Si, CeFe10.5Si1.5, and CeFe10Si2 were prepared by combining stoichiometric quantities of elemental Ce, Fe, and Si. Ingots were prepared by induction melting the elements under argon inert gas at 1420-1450° C., holding the molten alloy at that temperature for 3-5 minutes to ensure complete homogenization by induction stirring. Pieces of the resulting homogenized ingot were placed in a quartz ampule having a 0.65±0.01 mm diameter orifice in the bottom, remelted by induction heating to 1420-1450° C., and melt-spun by applying a 2.5-3.5 psi overpressure to eject the molten alloy onto the circumference of a rapidly rotating chromium-plated copper wheel (diameter D=25.4 cm). The surface speed, vs, of the wheel was varied between 5 and 40 m/s to alter the quench conditions. The resulting ribbon materials were collected, ball milled into powder, and their properties examined by X-ray diffraction (XRD) to determine crystal structure and phase composition. Table II summarizes the compositions, wheel speeds, and selected results.
Rietveld analysis was applied to the XRD patterns from CeFe11Si, CeFe10.5Si1.5, and CeFe10Si2 ribbons melt-spun at various wheel speeds. An example is shown in
Curie temperatures Tc were measured for each CeFe12−xSix alloy melt spun at 15 m/s, and the results are given in Table II. Values of I′, were obtained by monitoring the temperature dependence of the magnetic force in a small applied magnetic field using a Perkin-Elmer System 7 thermogravimetric analyzer (TGA). The Curie temperature is taken as the point where the contribution to the magnetic force (i.e., the magnetization) due to CeFe12−xSix vanishes. The Curie temperatures are the highest observed in Ce—Fe-based compounds to date. Notably, Tc increases with Si content even though the Fe content of the Ce(Fe12−xMx) compound is reduced.
Nitriding of selected Ce(Fe12−xMx) ribbons with pure nitrogen gas was performed in a Hiden Isochema Intelligent Gravimetric Analyzer (IGA). The typical nitriding profile is set as the following: temperature (T) 450-500° C., time (t) 1-48 hours, and pressure (P) 20 bar of nitrogen gas. The powders were sieved to 25-45 μm sized particles prior to nitriding. The nitrogen uptake was calculated from the change in sample weight at approximately 1 bar and room temperature (20° C.), before and after nitrogenation, in order to eliminate the confounding effect of buoyant forces at elevated pressure and temperature. Typically Ce(Fe12−xMx) compounds can absorb one to three nitrogen atoms per formula unit after being fully saturated by the nitrogenation process.
XRD examinations of the nitrides show that the ThMn12 tetragonal crystal structure is retained, and that insertion of N atoms into the lattice results in an overall increase in the unit cell volume. Accompanying the lattice and volume expansions (shown in
Previous literature reports on relevant RFe12−xMx suggest that due to the atomic size difference, Ti and Si preferentially occupy different sites in the lattice. The DFT calculations performed on CeFe8Ti4 and CeFe8Si4 indicate that Ti preferentially occupies the 8i site in the 1-12 lattice, while the Si preferentially occupies the 8j site. The preferential substitution of Ti and Si at different sites suggests that a series of hypothetical quaternary compounds of the form CeFe10+xSi2−2xTix could result in lattice distortion different from a single element substitution scheme, which offers a new variable to tune the magnetic properties. The quaternary CeFe10+xSi2−2xTix could be perceived as a solid solution of ternary CeFe10Si2 and CeFe11Ti.
Alloys of CeFe10Si2, CeFe10.25Si1.5Ti0.25, CeFe10.5Si1Ti0.5, CeFe10.75Si0.5Ti0.75, and Ce1.1Fe11Ti were prepared by combining stoichiometric quantities of elemental Ce, Fe, Si, and Ti. Ingots were prepared by induction melting the elements under argon inert gas at 1375-1450° C., holding the molten alloy at that temperature for 3-5 minutes to insure complete homogenization by induction stirring. Pieces of the resulting homogenized ingot were placed in a quartz ampoule having a 0.65±0.01 mm diameter orifice in the bottom, re-melted by induction heating to 1380-1450° C., and melt spun by applying a 2.5-3.5 psi overpressure to eject the molten alloy onto the circumference of a rapidly rotating chromium-plated copper wheel (D=25.4 cm). The surface speed, vs, of the wheel was varied between and 10 and 45 m/s to alter the quench conditions. The resulting ribbon materials were collected, ball milled into powder, and their properties examined by X-ray diffraction (XRD) to determine crystal structure and phase composition.
Nitriding of selected CeFe10+xSi2−2xTix ribbons was performed in a Hiden Isochema Intelligent Gravimetric Analyzer (IGA). The typical nitriding profile is set as the following: temperature (T) 450° C., time (t) 1-16 hours, and pressure (P) 20 bar. The powders were sieved to smaller than 45 μm sized particles prior to nitriding. The nitrogen uptake was calculated from the change in sample weight at approximately 1 bar and room temperature (20° C.) before and after nitrogenation, in order to eliminate the confounding effect of buoyant forces at elevated pressure and temperature. CeFe10Si2 exhibits the highest Tc=305° C. and CeFe11Ti has the lowest Tc=215° C.; the latter is in good agreement with the value of Tc=233° C. previously reported in the literature for CeFe11Ti. The Tc for the quaternary nitrides decreases monotonically with x. Curie temperatures are greatly increased after nitrogenation, with the smallest ΔTc=121° C. from CeFe10Si2 and the largest ΔTc=215° C. from CeFe11Ti. Quaternary compounds of the form CeFe10+xSi2−2xTix with x=0.25, 0.5, and 0.75 exhibit a Curie temperature enhancement exceeding 150° C., a larger enhancement compared to ternary CeFe10Si2. Magnetic moment has also been increased in the nitrides with the smallest increase of 12.8% in CeFe10.25Si1.5Ti0.25 and the largest increase of 20.6% in CeFe10.75Si0.5Ti0.75.
Table III summarizes the lattice constants, magnetic moment 4πMs, and Curie temperature for quaternary CeFe10+xSi2−2xTix and their nitrides. For the nitrides, the rightmost column also gives the number y of N atoms per CeFe10+xSi2−2xTixNy formula unit as determined from measured nitrogen weight gain during nitriding. CeFe10.25Si1.5Ti0.25 and CeFe10.5SiTi0.5 were melt spun at wheel speed vs=15 m/s while CeFe10.75Si0.5Ti0.75 was melt spun at vs=10 m/s. Except for CeFe11Ti, the nitrides listed in the table have been nitrided at nitrogen pressure of 20 bar at 450° C. for 16 hours. As stated above, the CeFe11Ti starting material listed in Table III was initially formed using 10 at % excess Ce in the starting composition in order to promote formation of the ThMn12 phase in both the as-formed melt-spun products and the nitrided products. For CeFe11Ti the nitriding was completed at a reduced pressure and temperature of 10 bar at 410° C. for 18 hours.
Thus, we have described a new family of permanent magnet materials that contain a major weight proportion of one or more compounds of CeFe12−xM, having the ThMn12 crystal structure (space group I4/mmm, #139) and with M being one or more of the elements B, Al, Si, P, S, Sc, Co, Ni, Zn, Ga, Ge, Zr, Nb, Hf, Ta, and W. Preferably, x is in the range of one to four. In addition, one or more of Ti, V, Cr, and Mo may be combined with, or substituted for, up to about ninety atomic percent of an M element in the CeFe12−xMx compound.
The material may be prepared from a melt of the constituent elements by rapid solidification to form with a major portion of the CeFe12−xMx compound. The material may be prepared in the form of a powder or other form for shaping and consolidating into a permanent magnet body for an electric motor or other desired product application. And the permanent magnet material may be nitrided to increase its Curie temperature and its permanent magnet properties.
Practices of the invention have been illustrated by specific examples which are not intended to limit the scope of the invention.
This invention was made with U.S. Government support under Agreement No. DE-AR0000195 awarded by the Department of Energy. The U.S. Government may have certain rights in this invention.
