US3424578A - Method of producing permanent magnets of rare earth metals containing co,or mixtures of co,fe and mn - Google Patents

Description

Jan. 28, 1969 K. J. sTRNA-r ET A1. 3,424,578

METHOD QF' PRODUCING PERMANENT MAGNETS OF RARE EARTH METALS CONTAINING Co OR MIXTURES OF Co, Fe AND Mn Filed June 5, 1967 nited States Pate METHOD F PRDUCTNG PERMANENT MAG- NETS 0F RARE EARTH METALS CNTAHNING C0, OR MIXTURES 0F Co, Fe AND Mn Karl J. Strnat and Gary I. Hoter, Fairborn, and John C. Olson, Dayton, Ohio, and Werner Ostertag, Painted Post, NY., assignors to the United States of America as represented by the Secretary of the Air Force Filed June 5, 1967, Ser. No. 644,460 U.S. Cl. 75-213 Int. Cl. B22f 3/12; H01f 1/08 Claims ABSTRACT OF THE DHSCLOSURE BACKGROUND OF THE INVENTION Field of the invention The present invention is predicted on the discovery that certain known intermetallic compounds posses high crystal anisotropy and when subjected to certain processing steps will form permanent magnets.

Representative of such compounds is yttrium pentiacobaltide, YCo5. It is to be understood that neither the discovery tof YCo5 per se or its property of ferromagnetism is alleged but rather a specific process for treating YCo5 and related compounds to form a permanent magnet.

By way of further explanation and background, the terms ferromagnet and permanent magnet are by no means synonymous. The tirst term simply indicates the existence of an ordered `arrangement of the atomic magnetic spin moments, a basic physical phenomenon also found in iron and in many other substances. A permament magnre however, is a specific device for technological application. While a material has to be ferroor ferri-magnetic to qualify for use in permanent magnets, the properties demanded of the latter are not inherent in any ferromagnetic material, The material must have a certain combination of basic properties -which may be optimized by such measures as adding alloying elements or heat treating. Furthermore, rather complicated processing of the material is usually necessary to produce a permanent magnet with the best possible properties.

With the present invention, the basic property of YCo5, f-or example, which the present inventors believe they were the first to recognize and measure, was its extremely lange unitaxial crystal anisotropy. Only with the kno-wledge of this property and not on the basis of previously published values for Curie point, magnetization, etc., of YCO5 was it possible to predict and to develop process steps whereby permanent magnets could be made of YCO5.

The technology set forth herein with respect to making permanent magnets from known ferromagnetic YC05 represents a distinct and imarked advance in the art even as Lodex and Alnico magnets were of patentable merit. With regard to the Lodex magnets, pure iron, pure cobalt, and solid solution alloys of these elements have long been known to exist and to be ferromagnetic. However, the idea of using elongated single domain (ESD) particles for magnets, developed mainly in the last ten years, has resulted in a series of patents on Lodex magnets concerned with the basic principles and properties of such particles as well as details of particle and magnet preparation. In connection with Alnico magnets, ferromagnetic alloys of iron with cobalt and nickel have been known for many decades. Yet in the course of developing the family of alloys known .fas Alnicos, various patents were issued covering first the use of such alloys processed in a certain manner for permanent magnets and later minor alterations in composition, heat treatment, sintering, etc., which resulted in better permanent magnets. With such precedents, it is believed that the unusually strong permanent magnets made from YC05 by the process steps set forth herein are clearly of patentable merit.

Description of the prior art The most pertinent prior art is probably represented by U.S. Patent No. 3,102,002 to Wallace et al, who describe a group of ferromagnetic materials on ycompounds prepared froim certain of the lanthanide elements and the transition metals of the first long period which have formulae corresponding to AB5 where A is a lanthanide element Ior yttrium and B is manganese, cobalt or iron. However, Wallace et al. only describe some basic crystallographic and magnetic properties of these compounds, including YCo5. Furthermore, Wallace et al. make no speciiic statements about the merits of lany of their listed compounds for permanent magnet application. Especially, no mention is made of the key property, high crystal anisotropy, nor is there any reference to the necessity of making powders of the compounds.

SUMMARY OF THE INVENTION The present invention consists essentially of a novel method of preparing a permanent magnetic material and the product resulting therefrom which is characterized by a high saturation magnetization, a reasonably high Curie temperature of several hundred degrees C., and high coercive force. The permanent magnets 'made by the present method are made from particles with magneto-crystalline anisotropy instead of shape anisotropy, and thereby avoid a number of disadvantages noted of the latter. The outstanding advantage of the permanent magnet material lmade by the present method is its potential energy product which is 29.2M g.0e. (1M g.oe.=l06 gaussxoersted) for percent packing of perfectly aligned, single-domain particles and 16AM goe. for 75 percent density packing. This compares with (a) 9.5M g.oe. for a platinum and cobalt alloy which is the maximum energy product known for a commercial magnet, (b) 12.5M g.oe. for the best laboratory magnet of Alnico, and (c) 6.5M goe. for the commercial Lodex (ESD) magnets.

The permanent magnets of the present invention nd application in communication equipment, control devices, navigational instruments, auxiliary power generators, etc. Specific examples include instruments which are based on the galvanometer principle, small electrical motors and generators, microwave tubes (in magnetrons and as focusing magnets in traveling wave tubes), biasing magnets for relays, microphones and telephones, and loudspeakers. The use of magnets in motors to replace the conventional stator windings is rapidly gaining acceptance and, while until recently only very small rotating electrical machines were built this way, permanent magnets are now invading tihe medium-power motor field. New concepts for equipment to be used on board of airor spacecraft are presently under study which require strong, large-volume, steady magnetic elds such as magnetohydrodyn-amic energy converters, devices which would direct the flow of hot plasma or of radiation particles around a space vehicle, and miagneto-plasmadynamic engines for space vehicles.

While present designs yare mostly based on the use of electromagnets, permanent magnets of appreciably larger energy density than presently available combined with good high-temperature performance would simplify designs considerably and reduce the equipment weight.

BRIEF DESCRIPTION OF THE DRAWING In the accompanying drawing:

FIGURE 1 is a graph showing the intrinsic coercive force of powders produced by ballmilling plotted as a function of grinding time and particle diameter; and

iFIGURE 2 is a graph showing the magnetization curves of a spherical single crystal of YCo5 measured in the easy and hard directions.

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention is predicated rupon the discovery that permanent magnet materials or alloys having a potential energy product which surpasses available permanent magnet material by a factor of over 2.5 can be prepared from (1) a rare earth metallic component selected from the group consisting -of yttrium, scandium, lauthanum, cerium, praeseodymium, neodymium, promethium, samarium, europium, gadolinium, terbium, dysprosium, lholminurn, erbiurn, thulium, ytterbium, and lutetivum including mixtures thereof and (2) a second metallic component selected from the group consisting of cobalt, manganese and iron including mixtures thereof. This discovery in turn is dependent upon the recognition and measurement for the first time of the extremely large uniaxial crystal anisotropy of the aforementioned materials as typified by YC05 coupled with certain process steps. The manner in 'which the magneto-crystalline anisotropy of YCo5 is measured is set forth in an article entitled Magnetocrystalline Anisotropy of Some Rare-Earth-Cobalt Compounds, published in the Journal of Applied Physics, March 1967. The ratio of rare earth metal employed to the second metallic component is 11 to 22 atomic percent rare earth metal to 78 to 89 atomic percent of Co, Fe of Mn.

In general, the permanent magnet materials of the present invention are prepared by melting together the desired amounts of the two general components, such as Y and Co, under a protective noble gas atmosphere or under a vacuum. rDhis step may be effected by arc melting on a cold copper hearth, induction melting in pure alumina crucibles, or containerless levitationmelting so as to avoid crncible contamination. The resulting alloy, such as YCo5, can be formed in -30 gram buttons and then crushed and thereafter ground in .a ball mill or a vibratory mill. By way of example, after 24 hours ball-milling time, all particles will be smaller than 53 microns, with further milling particles having a diameter as small as .3 to 3 microns can be obtained and effectively used in t-he present process. (See FIGURE l.) Vibration milling was found to` be considerably faster, yielding l to 5p size panticles in periods of 1/2 to 10 hours with charges of YCo5 varying from 1 to 25 grams. However, it should be noted that these particles are relatively large compared to tlhe 100 to 1,000 A. diameters of the particles required for shape anisotropy particles of comparable coercive force.

Thereafter the particles of the alloy are bonded together which may be effected by several methods. Use of an organic resin or plastic binder, such as an epoxy resin, is simple and will yield `magnets that are sturdy, easy to shape, and corrosion resistant. They will not be usable at temperatures much above room temperature. High temperature capability may be achieved by sintering or hotpressing the powder without a binder, or with an inorganic binder such `as boron nitride or a metal power. yIn either case, magnetic eld of at least several kilo-oersteds must be applied before or during the consolidation if oriented magnets with optimum magnetic properties are desired. By way of example, the nely divided particles of YCo5 have been formed into magnets by the following three different methods: (l) The powder was mixed with molten paraffin and the mixture permitted to solidify in a magnetic field of -l4K oe.; (2) The powder was compacted to ya density of -60 volume percent by means of a hydraulic press in a field of -6K oe. and the resulting magnet soaked in a polystyrene solution and subsequently dried; (3) The powder was stirred into a quick-setting epoxy resin which hardened while the magnet was in a homogeneous 20K oe. field.

The basic principles and considerations of the properties of permanent magnets of this kind lwere discussed prewiously in the publication of K. Strnat and G. Hoifer, YCo5 A Promising New Permanent Magnet Material, AFML-TR-65446, Air Force Materials Laboratory, Wright-Patterson Air Force Base, May 1966. These authors determined the magnetocrystalline anisotropy constants of YCo5 at room temperature from magnetization vs. field `curves measured on a spherical single crystal of -1 mm. diameter in the direction of the c-axis and normal to it. These curves are shown in FIGURE 2. YCo5 was found to have a single easy magnetic direction, the c-axis, and no detectable anisotropy in the basal plane in which the crystal is hardest to magnetize. The maximum applied external field of -45K oe. twas by far insucient to saturate the crystal in a direction in the basal plane. A straight line extrapolation of the M vs. H curve for the basal plane yields a saturation eld HAl32K oe. (also called anisotropy eld). Based on this extrapolation one can calculate the extremely high anisotropy constant The following examples are submitted to illustrate further the invention and not to limit the invention.

EXAMPLE I The metals yttrium (Y) and cobalt (Co), both commercial products of 99.9% nominal purity, were mixed in the weight ratio of l to 3.31 (atomic ratio 1:5). Peasize lumps or chips from machining on a lathe were used. In the latter case, it was found advantageous to precompact the charge to prevent loss of chips during melting. The charges of 5 to l0 lgrams were melted in a levitation furnace (USAF Technical Documentary Report No. ML-TDR-64-90, A Levitation Melting Apparatus for tihe Preparation of Ultrapure Samples of Reactive Materials, by John C. Olson, April 1964) and then cast and cooled rapidly by dropping the melt into a cold, thin- Iwalled porcelain crucible of 5 cm.3 capacity. Temperatures of over 1,600 C. were reached Iduring melting, a protective atmosphere of purified argon gas of typically -7 p.s.i.a. pressure was employed to prevent reaction of the yttrium with oxygen or nitrogen of the air. The resulting ngots were wrapped in tantalum foil, fused into evacuated quartz bulbs, and yannealed at a temperature of -1,000 C. for 100 hours. This treatment typically resulted in homogeneous alloys and metallographic sections appeared completely single-phase and coarsegrained. The ngots were then crushed in a hardened-steel mortar until the grains passed through a 60 mesh sieve.

IFifty (50) grams of this coarse powder together with cm.3 of hexane were placed in an alumina milling iarr (5" I.D. x 5" long) with l2 cylindrical alumina pieces O.D. x l" long) and tballmilled for 50 hours. The resulting slurry was removed from the jar, dried by letting the hexane evaporate at room temperature, and portions of the powder consolidated into magnets in the following different ways:

(a) 8 grams were placed in a l/z diameter cylindrical brass die to [be `compressed between two hardened' canbonsteel pistons inserted axially. An axial magnetic iield was applied with a solenoid surrounidng the die. The field was repeatedly turned on and off before pressure was applied, and was then maintained during compacting in an attempt to align the powder particles with their magnetic easy axes parallel to one another. Because the moving pistons also served as pole caps, the field acting on the sample during compression varied from an initial -6,000 oersted to -9,000 oe. A pressure of 51,000 p.s.i. was applied. A cylindrical magnet nesulted which bad a density of 4.5 g./c1m.3 (60% of massive YC05) and a powdery surface and the following magnetic properties: Br=3,6280 g., MHc=1,180 oe., BHC=930 oe., (BH)m-x= 1.1 l06 goe.

(b) 10 g. of the powder were intimately mixed with 2.5 cmI of a clear lacquer (Plastiklear No. 225, fan acrylic ester resin in colloidal solution containing -12% by rweight solids, manufactured by the Illinois Bronze Powder Company, Chicago, 111.), using a porcelain mortar and pestal. The mixture was dried completely in 1a stream of warm air of -50 C. and repowdered in the mortar. The powder was then compacted as described before, except that 1" diameter dies and pistons were used and the field varied from 11,000 to 15,000 oe.

The product was a disc magnet, -1/s" thick and l62% dense (massive YCo5=100%) which is mechanically mudh stronger than the magnet made without a binder. The properties measured in the alignment (=pressin1g) direction normal to the disc face are Br=3,500 g., MHC: 960 oe., BHc=750 e., (BH)m,x=o-.7M gee.

(c) In an attempt to use an inorganic binder which would not interfere with use of the magnet at elevated temperatures, l0 g. of YCo5 powder were mixed with 2.0 g. of 325 mesh Iboron nitride powder in the porcelain mortar. Ilhe mixture was compressed as in (a). The compact had mechanical strength and .cohesion superior to those of the binder-free magnet, but not las good as those of the acrylic-plastic bonded one, with magnetic data inferior to both.

EXAMPLE II In contrast to the procedure of Example I, the YCo5 alloy was prepared by fusing the alloying constituents Y and Co in an arc melting furnace having a water-cooled copper hearth and a nonconsumable tungsten electrode. Melting was done under a protective atmosphere of either pure argon gas or an argon-helium mixture, the charges weighing between 30 and `60 grams. Each charge was melted and resolidied three to four times to assure good mixing, the buttons were turned over between meltings. The ingot usually broke into several pieces under the thermal stress. They were vacuum-annealed for five days at 1,100 C. The material prepared in this manner was again crushed to a coarse powder in the steel mortar. 100 g. of powder were ballmilled as described in Example I, but more pieces of grinding medium were used (20 alumina cylinders) `and 100 cm.3 of hexane were initially added. Samples of powder rwere taken hom the jar at regular time intervals for coercive force measurements, and hexane 'was replenished as needed to maintain the same consistency of the thin slurry.

Specimens for coercive force measurements were made by mixing a small amount (200300 mg.) of the dried powder with ca. times its weight of an expoxy resin (Allaco Twenty/Twenty), then putting the thick liquid in a I0.4" I.D. x 0.6" cylindrical mold of Teflon, and letting the epoxy harden at a temperature of -70 C. while a magnet field of -l 5,000 oe. was applied to orient the particles in axial direction. Measurements of the intrinsic coercive force in this direction as a function of milling time and estimated average particle size are summarized in FIGURE 2. A maximum value of MH=1,850 oe. for -5,u. particles is followed by a drop-off on prolonged milling which is attributed to plastic deformation of the particles which destroys the favorable magnetic symmetry. It is expected that this undesirable overmilling effect can be overcome when the powders are vacuumannealed at temperatures between 300 and 600 C., or if they are prepared by a technique which avoids plastic deformation such as grinding below room temperature,

but for powders prepared by ballmilling at room temperature coercive forces appear to be limited by it.

IEXAMPLE III An alloy was made of 28.3 weight percent yttriumrich mischmetal (Y-MM) and 71.7 weight percent Co by arc melting as in Example Il and vacuum annealing for 160 hours at l,000 C. The resultant material was -95% single-phase and brittle. It was mortar-crushed to -60 mesh isize, a small amount of the powder was imbedded in epoxy resin and this binder was allowed to harden in a magnetic field as described in Example II. Magnetization curves were measured on this aligned powder sam-ple for the alignment direction and normal to it, using a maximum field of 45K oe. The magnetization curves resemble those of the YCo5 single crystal (see FIGURE 1). The following room temperatu-re data were determined from the measurements on this alloy:

Saturation induction, BS-9,500 g. Anisotropy field strength, H ,tpe-142 K oe.

. Anisotropy constant, K1-l-K2-5.4 x 107 erg/cm3.

Density, d=8.06 g./cm.3,

From these results the conclusion is drawn that (Y MM)Co5, as a fine particle permanent magnet material, behaves basically like YCo5. The upper limit for the enengy product is (BH)mm=(Bs/ l2)2=22.5 10i g.oe. The advantage of using the mischmetal instead of pure yttrium is a substantial reduction in the raw material cost with only a small sacrifice in magnet performance.

A typical analysis of Y-rich mischmetal (Y-MM) supplied by the Research Chemical Corporation, Phoenix, Ariz., was as follows.

In addition, there were also traces of other rare earths and other elements.

It is to be understood that the above mixtures are simply illustrative of the application of the basic principles of the present invention. By Way of example other plastic or resinous materials can be employed as |binders in addition to the aforementioned Plastiklear No. 225 and Allaco Twenty/Twenty. In general the requirements for a binder to be used in dense magnets in the manner outlined above (Example Ib) are as follows:

The application form must be a true solution or a colloidal suspension of particles having an average size below -0.l,u. The application viscosity must be centipoise (cps.), preferably on the order of 1 cp. The boiling point of the solvent should tbe under C., preferably near 50 C; the coating left on the particles after evaporation of the solvent must bond them into a solid body under pressures of less than 50,000 p.s.i. at temperatures below 70 C.; the resulting plastic must have very low absorption for water from the atmosphere.

Another satisfactory commercial product of the same general type as Plastikl'ear No. 225, is GC Koloid-Clear Acrylic, which like Plastiklear is available in the form of an aerosol container made by the `GC Electronics Company, Rockford, Ill. rl'lhis is a colloidal solution of methyl methacrylate in butyl acetate, 40% solids.

On the other hand, a -resin to be effective as a binder or matrix for coercive force specimens (see Example Il above) should have the following general properties:

A chemically hardening two-liquid system without a ller material having a low application viscosity 5,000 cp. at 25 C.) and a pot life longer than 15 minutes. It must cure into a hard solid rather than a rubber-like substance) in no more than 2 -hours at room temperature, no more than 30 minutes from the atmosphere.

In addition to Allaco Twenty/Twenty, another commercial product which gives satisfactory results as a binder or matrix for coercive force specimens is Allaco Crystal-Clear also made by Allaco Products, Incorporated, 238 Main St., Cambridge, Mass.

What we claim is:

1. A method of making a permanent magnet comprising the steps of:

pulverizing an alloy consisting essentially of a rst component A selected from the group consisting of Y, Sc, La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu an-d mixtures thereof and a second component B selected from the group consisting of (a) Co alone, (b) Co and at least one metal selected from the group consisting of Mn and Fe in the ratio of 11 to 22 atomic percent for component A and 78 to 89 atomic percent for component B,

mixing the pulverized alloy with a binding agent,

subjecting said alloy to a magnetic eld in the easy direction of the alloy so that the powder particles of said alloy lbecome aligned with the applied magnetic ield, and

thereafter permanently bonding together the aligned powdered alloy by sintering.

2. The method of claim 1 in which component A is yttrium and component B is cobalt.

3. The method of claim 1 in which component A is yttrium-rich tmischmetal and component B is cobalt.

4. The method of claim 1 in which the magnitude of the applied magnetic field is at least several kilooersteds.

5. The method of claim 1 in which the permanent bonding is effected by hot pressing.

References Cited UNITED STATES PATENTS 1,893,330 1/1933 Jones 75-222 X 2,384,215 9/ 1945 Toulmin 75-226 X 2,661,387 12/1953 Ackermann 75-222 X 2,825,670 3/1958 Adams 75-200 X 2,849,312 8/1958 Peterman 75-201 3,180,012 4/1965 Smith 75-213 X 3,287,112 11/1966 Blaha 75-222 3,322,536 5/1967 Stoddard 75-211 X FOREIGN PATENTS 590,392 7/ 1947 Great Britain.

CARL D. QUARFORTH, Primary Examiner. A. I. STEINER, Assistant Examiner.

U.S. C1. XJR.

Images