JPS62260039A - Permanent magnet alloy having improved residual magnetization and its bulk magnetic substance - Google Patents

Abstract

PURPOSE:To obtain a permanent magnet alloy showing isotropically superior magnetic properties and having improved residual magnetization, by applying quantum-mechanical exchange bonds among crystallites. CONSTITUTION:As a magnetic material, the magnetically isotropic one which has >1 coercivity coefficient Q, >15MGOe maximum magnetic energy product, and >9KG residual magnetization and is composed of a solid lump of crystallites and in which respective crystallites are in contact with adjacent crystallites via grain boundaries and are not oriented in the crystallologically same directions as those of crystallites adjacent via grain boundaries is used. Moreover, in the above material, respective crystallites have easy magnetization axes and the dimensions and geometries of crystallites and grain boundaries are regulated so that quantum-mechanical exchange bonds are produced across the grain boundaries among surface atoms of adjacent crystallites.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a permanent magnetic alloy material, its bulk magnetic body, and its manufacturing method.

BACKGROUND OF THE INVENTION There is a hyperbolic demand for relatively inexpensive, strong, and high performance permanent magnets. Such high-performance permanent magnets are characterized by relatively large magnetic parameters such as coercive force (Hc), coercive force, remanent magnetization, and maximum energy product.

Furthermore, an ideal high-performance permanent magnet exhibits a rectangular magnetic hysteresis loop. That is, coercive force 1-
When an external magnetic field H larger than 1c is applied, all the microscopic magnetic moments must be aligned parallel to the direction of force application, and a saturation magnetization Ms must be obtained. Shih\ also applies this alignment state not only when H=O (residual magnetization Mr) but also when 1-
It must also be maintained when a magnetic force smaller than 1c is applied in the opposite direction. This corresponds to the fact that the maximum magnetic energy product (maximum negative value of BH) is (M r /4)=(M's2/4). Unfortunately, such ideal conditions are at best metastable against domain formation in the other direction, which drives Mr and B Hnax to decrease.

Conventional high performance permanent magnets pursuing square loop behavior have the following four general requirements.

1. The material is mainly selected from the Curie temperature Tc and the applicable temperature T.
A ferromagnetic element or compound must be formed which is considerably higher than a and has a large f-magnetic Ms at Ta. In reality, it means that Fe or Co must be the main component.

2. In order to obtain a high coercive force, the material must be composed of an aggregate of fine particles or crystallites.

3. These particles or crystallites must exhibit microscopic magnetic anisotropy. In other words, it must have a desirable quantum ``easy axis'' of magnetism. This is due to shape anisotropy and magnetocrystalline anisotropy 4 interaction (magn
eto-crystalline 1ntcrac-t
ion).

4. Such microscopically anisotropic particles must be arranged substantially parallel in the macroscopic assembly to give a 1vlr value close to MS, ie, a square loop behavior.

According to the teachings of the prior art, a good permanent magnet, for example one with a maximum magnetic energy product of about 15 megagauss Oersted, has a non-interacting, crystallographically substantially oriented 7-axis Composed of agglomerates of particles. When a sufficiently large magnetic field is applied in a certain direction, the individual magnetization vectors of these particles are oriented along the direction of the applied magnetic field. This state corresponds to the maximum value or saturation value MS of net magnetization. When the applied magnetic field is reduced to zero, the magnetization vector of each particle returns to the easy axis direction of the particle, and the net residual magnetization MrG, becomes less than or equal to tMS.

This is explained more clearly by the following geometrical model in which the magnetization "easy axis" lies along some preferred axis C. When one uniformly magnetized particle is separated,
The magnetization vector M is along the C axis at zero applied magnetic field. When a magnetic field is applied in any direction Z, the magnetization rotates from the C-axis, and finally, if (component) is positive, when the magnetic field is removed, the magnetization returns to be parallel to the C-axis.

E, C, 5tun (3r and E, V, WohHarth
Paper phi1. ■rans.

Royal Soc, (London), A.
240, 599 (1948) Te calculated the hysteresis loop of such particles for cases where the orientation of the C axis with respect to the Z direction is different. In the case of a sample consisting of many particles that are oriented in a certain direction and do not interact,
The magnetic properties of the material or sample are the sum or average of the properties of the individual particles. Such a sample or material will hereinafter be referred to as an anisotropic material. Anisotropic materials have at least one magnetic property that is strongly dependent on the direction of measurement. Such a material is characterized by having one axis of easy magnetization whose characteristic value greatly exceeds the characteristic values in other magnetization directions. If there is no interaction between particles, the maximum energy product is 0.25 (MS) when Z is parallel to the C axis.
It varies from 2 to O when Z is perpendicular to the C axis. M
A theoretically anisotropic material selected with S of 16 and HC greater than 1yls has a theoretical maximum value of the energy product of the hysteresis loop of 64 megagauss Oersteds.

5tOner and Wohlfarth used a similar method to analyze an ideal array of randomly oriented, non-interacting, uniformly magnetized particles. Because the array is isotropic, the hysteresis loop is independent of the orientation of the applied magnetic field. The theoretical maximum value of the energy product of such a loop depends on Ms and HC. If Ms is 16 kilogauss and )-1c is chosen to be much larger than Ms, the maximum energy product will be 16 megagauss oersted.

Thus, the prior art teachings are that the maximum energy product of a perfectly oriented, non-interactive material (anisotropic) is the same as that of a randomly oriented material ′j! ? is at least four times the maximum energy product for the case (isotropic).

If the orientations of non-interacting particles have a general distribution, then the simple vector geometry conclusion, where 0 is the angle between the applied electric field and the easy axis of a given particle, (M r /Ms)=[Co5(θ)1, and the results shown in double parentheses represent the average of all particles weighted according to particle size. As is well understood in the anti-collision field, for a perfectly oriented, non-interactive permanent magnetic sample (anisotropic), M along the orientation direction
r/M S-1, and in the case of a completely unoriented and non-interactive sample (isotropic), M in all directions
r/M S = 0.5. For example, R9A,
Hc Currie's paper “Determination of the alignment of the easy axis of magnetization in uniaxial permanent magnets for residual magnetization measurements J, J.
Appl, Phy.

Vol, 52. (No. 12), p. 7344-73
46 (December 1981) See wo. 1 of this paper! The measurement result (0bSerVat 1on) is consistent with the above two hypotheses. For example, in the paper by J. F. tlerbst and J. C. Tracy [X-ray pole figure data (X-r
Regarding estimation of residual magnetization from aypole figure data) T J J, 81) Dl, Phys.
VOl, 50 (No, 6), D, 4283~
See No. 4284 (June 1979).

The aqueous number Q, which we refer to as the coercive coefficient or magnetic retention parameter, is: Q = S um (M r /M 5)2x,
It is represented by y and z. Here, Ms and Mr are measured while applying magnetic fields along three orthogonal directions. Theoretically speaking, for prior art magnetic materials, Q approaches 1 for particles or crystallites that are completely oriented and have no interactions (anisotropic); For crystallites without (isotropic), it approaches 0.75. Regarding the values reported for permanent magnet materials in the prior art, the values of Q tend to be significantly lower than the theoretical values. For example, the aforementioned HcCurrie paper Herbst and T.
racy paper, 5tOne and Wohlf-arth
See paper.

5toner and WohHarth papers, and which have no interaction, prior art systems that are compatible with the hypotheses and models of the It does not deviate in any way from the theory and model described in (which is one of the basis of the claim).

If the particles are allowed to interact, (Mr /Ms) = [Co5(
θ)] can be expected. Some of the magnetic recording literature makes this kind of suggestion, and A, K,
Bhatia's paper [particle aggregates]
Interaction effect on the saturation remanent magnetization of
Hagnetics, HAG-9, P, 127-133
(1983), and the paper by R, F, 5OOhOO [
Effect of particle interaction on coercive force and squareness (or squareness) of thin film recording media WJ, J, ApDl, P
hys,, Vol.

52(3), I), 2459-2461 (1981). However, questions have been raised regarding such hypotheses regarding interactions. For example, P, H, D
avis paper, "Interactional effects on the hysteresis properties of collections of randomly oriented magnetic or electric moments", J. Appl, phys.
,, Vol., 52(2).

p, 594-600. (1980).

E., Ca1len, Y., L., Liu, J., R., Cu1l.
en's paper "Initial magnetization, residual magnetization and coercive force of random anisotropic amorphous ferromagnets J, PhyS, ReV, B,
VOl, 16. p., 263-270 (1977),
Regarding amorphous iron-rare earth alloys at cryogenic temperatures, short-range interactions based on exchange interactions have been suggested.

Regarding isotropic permanent magnetic materials, this author does not prove that the Mr value is larger than the Mr value predicted by 5toner and Farth.

In contrast to the limited and negative teachings of the prior art, the present inventors have determined that the mother-child dynamics between crystallites can be measured in any spatial direction, i.e., equal A type of permanent magnetic alloy is provided that exhibits excellent directional magnetic properties.

The magnitude of the magnetic parameters reaches levels that could only be achieved in the prior art in one direction, ie, anisotropically, and only with aligned materials.

The ratio of net remanent magnetization (Mr) to net saturation magnetization (Ms) of the magnetic material of the present invention exceeds 0.5 in all directions with almost no crystallite orientation in a preferred direction;
Habo reaches 10. This means that to bring Mr closer to MS, i.e. to achieve square hysteresis loop behavior, the crystallites are microscopically anisotropic crystallites oriented substantially parallel in the macroscopic bulk magnetic material. This clearly overturns the conclusions of the 5tOner and WOhlfarth models and the assumptions of the prior art that it should be true.

The coercive coefficient Q of the permanent magnetic material of the present invention is greater than 1 as described above. The theoretical limits of coercivity for the materials of the present invention are 1.0 for the prior art aligned (anisotropic) non-interactive materials and non-aligned (isotropic) non-interactive materials, respectively. and 0.75, it is thought that it will be 3.

In the case of a ribbon-shaped sample that has been rapidly cooled and has not been further processed, the residual magnetization Mr is 9 kilogauss or more, and the coercive force)
(C is 8 kg or more, preferably 11 km or more, maximum energy product < BH) wax is 1
5 megagauss or more, and the same value is obtained when measured in all directions, that is, within the ribbon plane and perpendicular to the ribbon plane. Standard corrections for shape anisotropy of the ribbon when measured perpendicular to the ribbon plane (e.g. R, H, Bozorth 7y "Ferromagnetic J D, Va
n No5trandCo, New York, (1
951), pages 845-847).

The saturation magnetization MS of the ribbon, that is, the magnetization at the extreme when the applied magnetic field is large, for example, about 50 kilogauss or more, is also 15 to 16 kilogauss in all directions. To directly measure the saturation magnetization Ms, the applied magnetic field must be at least three times the coercive force Hc. Alternatively, the value of yls can be predicted from coercive force values for compositionally similar materials. The above values correspond to values of Mr/Ms of 0.6 or more and coercivity Q of 1 or more, which the prior art clearly teaches for microscopically isotropic, non-interactive materials. contradicts what is there.

Typical magnetic parameters for the magnetic alloy of the present invention are shown below (Japanese Patent Application No. 61-40160 (Japanese Unexamined Patent Publication No. 61-2431).
54→ to (corresponding to Table V of the aforementioned U.S. Patent Application No. 816,778) (using a 16 kilogauss MS)
.

Table ■ 650" 68r) **Electric ECD: Energy Conversion Device Incorporated [1st axis *l [2nd axis 1 12.2 >22 31.6 10 >2
2 21.51 0.75 1.41
2.4 8.5 24.2 11.3 B, Go
22.1 0.75 1,710.4 1
2.222.7 9.912.2 20.8
0.75 1.210.2 11.6
20.8 10.1 116 20.2 0
．． 75 1.29.6 9.817.7
9.3 9.5 15.9 0.75
1.059.6 13.618.6 8.81
4.2 15.8 0.75 1.03
In one embodiment of the present invention, the as-spun ribbon material can be further processed to create shaped solidified magnets. As an effect of the processing steps used, it is also possible to create a degree of magnetic anisotropy in the material, making it suitable for applications requiring such magnetic materials.

As can be seen from Table 3, the material of the present invention produced by the method disclosed herein has the following properties throughout the bulk solid in each sample:
It has excellent magnetic parameters that support the existence of quantum mechanical exchange coupling between crystallites. In particular, the patent application 1986-40
The magnetic properties are superior when compared with the magnetic properties of prior art isotropic materials listed in Table 1 of No. 160 (Japanese Unexamined Patent Publication No. 61-243154) to the aforementioned U.S. Patent Application No. 816.778. Table ■ (Patent Application No. 1983-40160 (Japanese Patent Application No. 1983-
No. 243154) to the aforementioned U.S. Patent Application No. 816,77
When compared with the prior art anisotropic materials listed in Table 1 of No. 8, the samples of the present invention are comparable in magnetic properties and do not require the costly and complex alignment or orientation process required by the prior art. does not require.

The invention may be better understood by reference to the accompanying drawings.

According to the present invention, there are provided a magnetic alloy material having excellent magnetic properties, a method for synthesizing the magnetic alloy material, a shaped solidified body of the magnetic alloy material, and a method for forming the shaped solidified body.

The alloy material of the present invention has 5 tons of non-interactive particles.
This interaction, in which individual particles or crystallites interact across grain boundaries and is inconsistent with the er and -ohlfarth hypothesis, is probably a ferromagnetic exchange type interaction via conduction electrons.

This alloy is a substantially non-crystallographically oriented, substantially magnetically isotropic alloy with interactions between adjacent crystallites. Substantially isotropic means that the material has similar properties in all directions. In constant terms, a substantially isotropic material is a material in which the value of C05(θ)1 defined above is approximately 0.75 or less in all directions, where C05
(θ) is averaged over all crystallites.

The material of the present invention has an isotropic maximum energy product of 15 megagauss or more, a coercive coefficient Q of 1.0 or more, a coercive force of about 8 kilogauss or more, and a residual magnetization of about 9 kilogauss or more, preferably about 11 kilogauss. There are 11 permanent (hard) stones.

For ribbon-like and flake-like samples of materials that have been quenched and have not been further processed, the residual magnetization Mr is 9 kgauss or more, the coercive force HC is 8 kgauss or more,
The maximum energy product (B l-1) max is greater than or equal to 15 megagauss, and these values are the same whether measured in any direction, i.e. in the ribbon plane or perpendicular to the ribbon plane. Become. The values measured perpendicular to the ribbon plane were obtained after standard corrections for ribbon shape anisotropy (shape-dependent demagnetization correction).

The saturation magnetization MS of the ribbon, ie, the magnetization at the extreme of increasing the applied magnetic field, is also 15-16 kilogauss in all directions. These l pieces correspond to an M r / M s value of 0.6 or more and a coercive coefficient Q of 1 or more, which is what the prior art clearly teaches regarding microscopically isotropic materials. contradicts.

The magnetic material of the present invention is composed of small crystalline ferromagnetic grains. Each crystal grain is a single magnetic domain. That is, all spins in a given grain are oriented in the same direction. Therefore, we can reduce the magnetic moment of each grain to a single giant spin (
or "superspin"). This means that each grain has a preferred axis of easy magnetization.

This means that the directions of the easy axes of magnetization are random and there is no correlation order between crystal grains.

The existence of a crystallographic easy axis is associated with magnetic anisotropy energy or anisotropy. The magnetic anisotropy energy, or anisotropy field, tends to keep the spins aligned with the easy axis. In the case of NdFe14B, this anisotropic magnetic field is H(
It is reported that the anisotropy) = 70 kOersted. For example, J. B. Livington gave a presentation at the 8th International Study Group on Rare Earth Magnets and Their Applications held in Deaton, Ohio, May 6-8, 1985.
Paper “Iron/Rare Earth Permanent Magnets J Paper No. Vl
-1 (obtained from: Univ, ofDayton, H
agnetics, KL-365, Dayton Oi
l, 45469. (USA). Since the anisotropic magnetic field acts on every spin in the crystal grain, it is essentially independent of grain size. Therefore, the theoretical hysteresis loop predicted for a random isotropic collection of non-interacting anisotropic grains can be calculated using the simple domain reversal model of 5tOner and Wohlfarth described above. 5tOner and Wohlfar mentioned above
The th model ignores the interaction between grains, but
This can become so large that it cannot be ignored.

)-18=70 pkOe, and saturation magnetization MS=1
In the case of 6kG, the hysteresis loop calculated using the above-mentioned 5tOner and farth model is as shown in FIG. 1, and the large energy product a is about 3 Hmax is about 14.

The crystal grains are structurally and metallically abutted along their surfaces, that is, along their grain boundaries.

Thus, the surface spins of a given grain interact ferromagnetically with the surface spins of adjacent grains via monic electron exchange. This is the same quantum mechanical interaction that produces the fundamental ferromagnetic alignment within the grain. As a result of this surface exchange interaction, effective or effective coupling occurs between the superspins.

Within the scope of a simplified model (mean field theory), exchange interactions can also be treated as effective magnetic fields acting on individual grains. Regarding crystalline iron, see Kittle
The results of standard calculations regarding the exchange magnetic field at the atomic level reported in "Introduction to Solid State Physics" (3rd edition) by the author, 10
Obtain the value of Hoe. This value of 10 Hoe assumes that all spins around a given iron atom are aligned in the same direction. This exchange magnetic field has a temperature of TC (1100 degrees for Fe, Nd Fe1
For 4B, it is proportional to 600℃), so Nd2Fe14B
Calculating the exchange magnetic field at the atomic level for approximately H(
Spin・Spin)=6HOO.

To calculate the effective interaction field [H(grain/grain)] acting between grains on a larger scale, it is necessary to use a simple scaling method. Interaction between grains is a surface phenomenon. Therefore, H (crystal grain, crystal grain) is proportional to the ratio of surface area to volume. Therefore, for the scale of grain size R, the effective angle spin interaction field is expressed as H(grain, crystal grain>(R)=H(spin, spin) x (a/R
). Here, a represents a typical Fe-Fe atomic distance m<about 2.5 angstroms). H(
In the case of Nd2Fe14B, the crystal grains, crystal grains) are Ro −[
H (spin, spin)/H (anisotropy)] X a = 2
H (anisotropy) on a scale of 00 angstroms
= 70 kOe. This scale is believed to be close to the particle size that provides optimal magnetic performance as described below.

The degree to which magnetism is enhanced is determined by the grain size distribution and grain size relative to this characteristic scale.

At the limit of large grains, the grain size of most grains is RO.
This corresponds to the case where the above is the case. That is, in this case, 1'' (anisotropy) is much larger than H (crystal grain, crystal grain),
The interaction becomes weaker and has almost no effect. Therefore, the magnetic parameters are essentially the same as in the non-interactive case. Additionally, if some grains are large enough to contain multiple magnetic domains, the grains will separate into multiple domains, resulting in a substantial reduction in coercive force and magnetic energy product.

At the smaller limit of grains, the grain size of most grains is R
It is below O. In this case, 1" (crystal grain, crystal grain) is H
(anisotropy), the interaction field becomes dominant and the magnetic moments of adjacent grains are essentially fixed. This increases Mr to approximately 1yls. The size of the effective magnetic domain falls within a range that includes the majority of highly correlated grain moments. That is, a single magnetic domain actually contains a plurality of individual crystal grains. Since such a supermagnetic domain contains many crystal grains with randomly distributed easy axes of magnetization, the anisotropy of the magnetic domain becomes almost zero when averaged.

As a result, the coercive force 1-1c also becomes smaller.

Therefore, in the state of small particle size, B Hll1ax becomes very small regardless of Mr.

For the intermediate state, grain size RO=200 angstroms, the interaction serves to bring the superspin moments together without substantially reducing the coercivity. Therefore, both Mr and (BH)max become substantially large. An estimate of the magnitude of this effect can be determined using a simple model as shown below.

In this case, grains due to surface exchange between adjacent crystallites -
The crystal grain interaction ring H (crystal grain, crystal grain) becomes substantially equal to the magnetic anisotropy energy of each crystallite.

The increase in interactions within the range of the average approximation can be treated similarly to the Weiss molecular field model for typhoidity (
For example, by 1ttel” Introduction
o 5olid 5tate Physics”,
3rd, Ed, pp 455-458)
. Since not all grains are aligned, the average total interaction ring acting on a given grain is: H(interaction) = H(grain, grain) x (M/Ms)
It is. The effect of interactions on the hysteresis loop can be calculated by considering that a given particle "sees" the entire effective field: H(effective) - H(interaction) = H(applied).

Then M vs. H(effective) should be the same as in the non-interactive case, by distorting the curve (instead of M vs. H(effective))
can be converted to . Such "distortion J" is carried out in the same manner as the standard treatment of the anti-11 field, which here corresponds to a negative demagnetizing coefficient of -H(grain, grain)/Ms (H(grain, grain) (See the dotted line in Figure 1 for the case where the grain size is 48 koe).

This average ring model gives the (erroneous) suggestion that both the residual magnetization and the energy product increase infinitely as the H (crystal grain, crystal grain) increases. This is a typical error seen when using the mean ring model outside its valid range.

The mean ring model rests on the assumption that the magnetic moments of the grains around a given central grain are independent of the central grain magnetic moment 1~. This hypothesis states that when the interaction is weak (that is, when H (crystal grain, crystal grain) is larger than H (anisotropy)), there is inevitably a correlation between magnetic moments between adjacent particles. Become.

This correlation cannot be handled by the mean ring model.

To the best of the inventors' knowledge, no model has been published that can solve problems down to the extremely small grain size.

However, from the curve in Figure 1, it is clear that for interaction,
It is therefore possible to estimate the energy product as a function of the scale length Ro (FIG. 2).

The dotted line showing a clear increase around 200 angstroms is an estimated value that is considered to be qualitatively correct.

Within the mean ring model with such qualitative corrections, energy products of up to 40 HGOe or more may be possible for the RE2TM, 4B, system. This corresponds to a coercive coefficient Q of approximately 3. Please refer to Figure 3.

The scaling methods described above assume that all grains are essentially the same size and have the same interactions among themselves. Improving magnetic properties is related to behavior that is correlated with grain scale, so the distribution of grain and intergrain geometry, including grain size and grain size distribution, can improve magnetic properties. This is an absolutely indispensable condition if you do it properly. When a grain of optimal size is in contact with a smaller grain, the coercive force of the smaller grain tends to weaken, making it difficult or even impossible to achieve optimal performance.

In any real system, there will naturally be a distribution of grain size and shape as well as a distribution of intergrain exchange coupling due to the presence of grain boundary inclusions and minor phases.

Any such defects tend to reduce optimal performance, and the actual material's optimal average grain size may deviate from the optimal values listed above.

In the above, the explanation regarding muntin mechanical exchange coupling beyond grain boundaries is applicable to rare earth/transition metal/boron materials with a crystal structure of P42/mn1ll symmetry in a tetragonal system, especially "d2
Although this has been carried out with respect to Fe14Bi-based materials, this phenomenon can be said to be a general phenomenon that can be similarly applied to other systems.

However, the optimum characteristic particle size RO may be different in a relaxed system.

According to the prediction, in the case of Pr Nd Fe14B1, X2-x
For any value of x, R8 is expected to be approximately 200 angstroms. For SmCo5, for example, Curie temperature Tc = 900K, saturation magnetization Ms = 12kGSH (
Anisotropy) = 300 kOe, H (spin, spin) = 9 MOe, then RO = (9 Hoe) / (3
00koe) x 2.5 angstroms, which is approximately 8
It becomes 0 angstrom. Similarly for Sm C017, R8= (12MOe) / <80 koc
) x 2.5 angstroms = 400 angstroms.

For optimally sized crystallites oriented randomly, the expected magnetic enhancement due to the magnetic magnetic coupling is similar to that predicted above for N d Fe,4B-based materials. It is about the same level, and B Hmax is the same as the above 5to
ner and costal farth model predicted values.
~3 times as much.

In one example, the magnetic alloy material is an alloy of iron (optionally also including transition metals such as cobalt), one or more rare earth metals, boron, and a modifying element. In another example, the magnetic alloy is an alloy of iron or cobalt, a lanthanide, such as samarium, and a modifying element.

A modifying element is a type of element that, when added to a magnetic material, works to improve the isotropic magnetic properties of the resulting material through appropriate processing techniques compared to unmodified materials. or more alloying elements. Examples include silicon, aluminum, and mixtures thereof.

Alternative or auxiliary modifying elements also include lithium, hydrogen, fluorine, phosphorus, sulfur, germanium, and carbon. The modifying element is a grain refining agent.
agent) and provides a crystallite morphology and crystallite size distribution that enhances interaction.

The m of the modified 71 element is set to such a value that, in combination with the quenching parameter, the above-mentioned isotropic magnetic parameter and the x8 pattern of E) are obtained.

Although the alloy described here has modifying elements that are thought to control grain nucleation and growth, the crystallite size and size distribution can be controlled by appropriately selecting and controlling the solidification technique used. It's okay.

Instead of rapid solidification of the melt, solidification methods such as gas atomization, metallization, chemical vapor deposition, etc. can be used. During solidification from a liquid phase, or grain nucleation or growth from an amorphous state, modifying elements act, for example, as grain viscosity agents or nucleating agents, to provide the necessary property enhancements. Give the distribution of crystallite size and morphology -.

The magnetic alloy is a [rare earth metal]-[transition metal]-[modifying element] system, for example, [Sml-[Fe, Col[Si,/
M].

Another interactive alloy is [Rare Earth Gold9.1-[Transition Metal]
-Boron- [modifying element] system, for example [rare earth metal]
- [Fe, Co] - boron - [modifying element] and [rare earth metal] - [Fe, Co, Mn1 - boron - [modifying element].

An example of a magnetic alloy material is represented by a chemical formula (Fe, Go,
N i >8 (Nd, Pr)bB.

(Ag, Srid, etc. For example, Fe (Nd, pr) bso (AR, Si).

It becomes a. At this time, a, b, c, and d are determined by energy dispersive spectroscopy (EDS) and wavelength dispersive spectroscopy (WDC) using a scanning electron microscope.

It represents the atomic percent of each component in the alloy, namely iron, one or more rare earth metals, boron, and silicon, and a + b + c + d = 100a is 75 to 8
5°b is 10~20. In particular, it is 11-13.5, and C is 5-13.5.
10, and d is a value effective to provide a morphology and crystallite distribution that, when combined with a particular solidification technique or solidification and heat treatment reaction, can enhance interactions with respect to magnetic parameters, e.g. to 5.0.

The rare earth metals are lanthanides selected from neodymium and praseodymium, and optionally other lanthanides (one or more la.

Ce, Sm, Eu, Gd, Tb, Dy, Ho.

Er, Tm, Yb, Lu), Sc, Y, and mixtures thereof can be added. Although various combinations of rare earth metals may be used without departing from the concept of the present invention, rare earth metals having one or more of the following properties are particularly preferred. (1) Those in which the number of electrons in the f-shell is not O (like La), 7 (like Gd), or 14 (like LLJ),
(2) La.

Low molecular weight lanthanides such as Ce, Pr, Nd, and Sm; (31 Lanthanides with large magnetic moments that ferromagnetically bind to iron such as Nd and Pr; (4) La, Ce,
Relatively inexpensive lanthanides such as Pr, Nd, especially Nd
and Pr are preferred. Various commercial Mitshu metals and/or Mitsush metals as by-products can also be used.

Particularly suitable Mitsushi metals are those rich in Nd and/or pr.

The crystallographic groove 1Δ of the interactive RE Fe14B-Si alloy of the present invention is consistent with the presence of a predominant portion consisting of rare earth-iron-boron-based phases, compared to the rare-earth-iron-boron-based phase of the prior art. Rietveld (Rietve
ld) Has a precisely analyzed X-ray powder diffraction pattern.

Although the structural differences between the magnetic materials of the present invention and conventional materials are minor, when combined with the appropriate microscopic structure, they are mutually effective as bulk permanent materials at standard temperatures and pressures. (cr
itical) is considered to be a difference.

An example of a hard magnetic alloy material has a crystallographic structure of P42.
/mnm symmetry, tetragonal RE-Fe-B, RE
-The main part consists of a tetragonal phase composed of Fe-B-3i, RE-Fe-B-(Si, AR) or RE-Fe-8-11, and the lattice constant matches the α-iron phase. It is composed of a crystal structure exhibiting a Rietveld precision X-ray powder diffraction pattern that is consistent with the presence of a phase considered to be a body-centered cubic phase, as well as other eluted phases below the detection level of X-ray diffraction. Furthermore, the Rietveld precision analysis X-ray powder diffraction pattern is consistent with a shift in boron position compared to that of the unmodified material. Such a shift in the boron position is due to the Si-containing M in one Si.
increases as the value increases.

The above-mentioned findings do not contradict the case where, for example, in the tetragonal RE-Fe-B Rakuko, Sl, which replaces Fe, is used as a modifying element, and the modifying element acts as a grain refining agent and competes. The rates of nucleation and grain growth are adjusted to result in crystal dimensions necessary for magnetic interaction between adjacent crystallites.

It, A. for the Rietveld analysis method.

Youna and D. B. Wiles, “Application of the Rietveld method for structural analysis using powder diffraction data” J. Adv., X-ray A.
nal, Vol. 24. p, 1-24 (1981) and “Contour shape function J in Rietveld precision analysis.
, J,APPL,Cryst, ,Vol.,15. p,
430-438 (1982). Rietveld analysis is a method of generating precise values for structural parameters starting from powder diffraction data, such as X-ray diffraction data or neutron diffraction data. In the Rietveld analysis method, a least squares method is used to obtain the best overall fit among all (plural) powder diffraction patterns obtained by calculation. This allows the coordinates of atoms, thermal motion,
fS structural parameters such as site occupancy parameters can be obtained.

As described in the above-mentioned Young et al. report, no attempt is made from the outset to assign the observed intensities in the powder diffraction pattern to individual Bragg reflections or to separate overlapping reflections. Instead, the expected powder diffraction pattern for angle 2θ is calculated from a model of the crystal structure. Next, the difference between the observed powder diffraction and the calculated powder diffraction is used to calculate changes in parameters to improve the degree of coincidence. Continue this operation using all observed intensity information.

A typical example of the magnetic alloy material of the present invention is P42/mnm and has a tetragonal crystallographic structure, and the occupancy, crystal lattice constant, and scale factor according to Rietveld precision analysis of X-ray diffraction data are shown in Table 3. That's right.

Table ■ High energy product Nd-F-B-8i alloy as rapidly cooled (Nd: 12.7 at%, Fe: 79.6 at%, B: 6
atomic%, 3i: 1.7 atomic%) 376^VO8 glue 1tbelt refined Nd (110,
2736(8) 0.2736(8) 0.000
0 0.24(1)Nd(2) 0.1417(8
) 0. M17(8) 0.0000 0.23
(1) Fe(1) 0.2220(3) 0.56
53(13) 0.5653(13) 1.05(
4) Fc(2) 0.0380(11) 0.35
77(12) 0617f32(9) 1.11(
4) Fe(3) 0.1063(14) 0.10
63(14) 0.273(12) 0.50(2
)Fe(4) 0.3141(10) 0.314
1 (10) 012603 (13) 0.55 (2
)Fe(5) 0.5000 0.5000 0
．． 1150(24) 0.23(1)Fe(6)
o, oooo O, 50000, 00000, 25
00B (1) 0.339 (13) -0,33
9 (13) 0.0000 0.28 (8) C axis (
12.203-12.221a axis (Angstrom) 8.79G-8,801 Scale factor (rare earths, iron, boron, relative alpha iron) 74, 29
/1 Rietveld precision analysis X-ray diffraction of the ribbon material revealed that the material was substantially isotropic.

Patent Application No. 61-40160 to U.S. Patent No. 816,77
Figure 2 of No. 8 graphically shows the relationship between boron lattice position versus silicon-containing book. The boron lattice trust in the rare earth/iron/boron lattice of square products is J and F.

1-1erbst, J., Croat,
W. B. Yellon's report “Nd Fe14B
"Structural and magnetic properties of", J. Appl, Phy.
s, Vol, 57. No, 1. P.

4086-4090 (April 1985) for the Rietwerd precision analysis neutron diffraction data on rare earth/iron/boron tetragonal rice grains that do not contain silicon.
It can be seen that there is a correlation between the gradual deviation and silicon addition. This is beyond experimental error and corresponds to a non-interstitial position of silicon in the lattice.

One of the methods for manufacturing magnetic alloys having the above-mentioned magnetic anisotropy and the above-mentioned crystal properties is the melt spinning method, which is disclosed in Japanese Patent Application No. 61-40160 to U.S. Special Application No. 316,778.
There is a method of rapidly solidifying and cooling a molten alloy on a moving cooling surface, such as a rotating cooling surface, as disclosed in the US Pat.

By adjusting the quenching parameters, it is possible to direct the solidification front, control its speed, and tune the fineness of the grains.

Quenching the alloy at a suitable rate results in shapes and morphological, crystallographic, atomic and electronic structures that provide novel interactive magnetic properties. By carefully controlling the quenching parameters, a suitably fine-grained structure is produced which, in combination with the above-mentioned modifying elements, results in the desired permanent magnetic material.

Alternatively, the alloy may be quenched at a rate that produces a precursor microstructure, in which case appropriate heat treatment results in a structure with the excellent magnetic parameters described above. The flakes thus obtained are much larger than the characteristic particle size RO. A typical flake may contain at least 10'' ([!a) grains of characteristic size RO.

The individual milled-spun fragments are recovered as a product from the melt-spinning process. Individual particles can also be obtained by milling seven relatively brittle ribbon pieces. When the ribbon is crushed, particles such as flake-like particles and individual plate-like particles are produced.

Particulate VvA materials have the practical advantage of being easier to subsequently process into final products. Because the magnetic particles of the present invention are isotropic in their magnetic behavior, they can be pressed, shaped, and solidified without being arranged in a particular crystal orientation and/or magnetic alignment.

Thus, in accordance with one embodiment of the present invention, the ribbon material may be roughly reduced in size, eg, to a maximum dimension of 0.5 millimeters, and then cold pressed and bonded. In this way, the magnetic powder density can be reduced by about 70%.
Thus, the (B H) max value can be greater than 50% of the original fully dense or 100% dense starting material.

The magnetic material (powder) of the present invention has a theoretical density of approximately 90% without going through a magnetic alignment process while maintaining high magnetic properties of the material.
It can be compressed up to %.

Compression molded solidified bulk hard material [
The isotropic magnetic energy of the 4i material is about 15 megagauss or more, and the coercive coefficient Q is 1 or more.

The invention may be better understood by reference to the following examples.

Example A, Test Summary The results for each of the following items are summarized by first melting a suitable mixture of iron, neodymium, praseodymium, boron, silicon, and aluminum into a macroscopically homogeneous ingot (master alloy). ) was created.

Each portion of the ingot is then melted and rapidly quenched using melt spinning to form ribbon pieces. These as-quenched ribbon samples were individually weighed and magnetically measured. Note that the sample was normally magnetized in advance using a large pulsed magnetic field. In some cases, the ribbon samples were subjected to further heat treatment before being magnetically measured again. Using scanning electron microscopy for both ingot and final ribbon samples,
Its microstructure and elemental composition were investigated. In addition, some samples showed X
Its crystallographic structure was analyzed using a combination of line diffraction and Rietveld precision analysis technology.

Several batches of ribbon samples were roughly ground, molded and solidified to form a bulk magnetic material, and then the magnetism was measured again.

B. Preparation of ingot (master alloy) Precursors or master alloys are generally iron (99.99% pure electrolytic iron flakes), boron (99.7% crystalline boron), pure Nd and Pr. Made from the following elemental components: rod (99.9% rare earth metal), silicon (99.99% Si crystal), in some cases using materials with higher purity, and in other cases using industrial grade materials. (commerci
A rare earth product (al-grade) containing up to 15% by weight of iron and up to t% of powdered rare earth metals other than Nd and Pr was used. Each component was weighed in appropriate proportions and melted by arc melting on a cooled A4 hearth or by high frequency induction heating in a crucible made of fused quartz or sintered magnesium oxide ceramic. The arc melting sample was melted and turned back six times, while the induction melting sample was kept at a temperature of about 1400°C or higher for 30 minutes to 2 hours while stirring the melt thoroughly.
A macroscopically endogenous alloy was obtained. After solidification and cooling, the ingot was recovered from the crucible, the shell of reaction products was removed, and the ingot was crushed into particles with a characteristic dimension of 1 cm. The composition of each sample of the ingot material was examined to determine its homogeneity.

To prepare C0 quenched ribbon, use any of the following three melt spinning systems.
A quenched ribbon was prepared from the ingot. Two of the three systems are simple box spinners using a 10" diameter, 1" thick copper wheel (10" spinner) and a 12" diameter, 2" thick copper wheel (12" spinner), respectively. be.

The chamber shall be open for refilling with an inert processing atmosphere after being depressurized. These spinners do not mask the crucible. In the third system (20" spinner), the copper wheel is shelled 20" outside diameter, 4" wide, and 3" thick in a chamber that is continuously flushed with an inert process gas. This wheel is housed. The crucible is surrounded by a shroud formed by a flow of inert gas. The flow of inert gas flowing from the crucible in a direction counter to rotation causes the wheel surface to pull the gas along its surface. In all three systems, the rotational speed of the spinner wheel is normally 15 to 30 in terms of surface speed.
m/sec.

For 12" and 20" spinners, the crucible should be
A transparent fused silica cylinder with a diameter of 5 mm and a length of 40 cy+ is used, and in the case of a 10" spinner, the crucible is a similar cylinder with an inner diameter of 1
It shall be 7m long and 25cm long. The orifice of the crucible was usually a circular hole with a bottom diameter of 0.5 to 1.5 m, and the crucible was positioned so that the orifice was 5 to 10 m from the wheel surface.

Put several pieces of ingot alloy into a crucible, 450k
Hz induction furnace (10kHz2 induction furnace for 12″ spinner)
Melting was performed until the temperature measured with an optical pyrometer reached a desired temperature (usually about 1200 to 1300°C). While high-frequency heating was continued, the inside of the crucible was pressurized with an inert gas, and molten metal was jetted from the orifice toward the rotating wheel. The injection continues until the crucible is empty, or until the molten metal remaining in the crucible is too large for efficient radio frequency heating and the orifice is clogged. In some cases, the molten metal may splash onto the crucible, causing the orifice to become partially or completely clogged at an early stage. Such factors also contribute to similar non-reproducibility of nominal process parameters in the properties of the final material.

Both the ingot material and the quenched ribbon samples were examined for microstructure and composition using a JEOL scanning electron microscope. In most cases, samples were placed using standard metallographic procedures and inspected after polishing. In some cases, the ribbon sample was etched with an ethanol solution containing 2% M of nitrate to further reveal the structure of crystal grains V4. The composition was measured using energy dispersive X-ray spectroscopy (EDS), and the concentrations of Fe, Nd, pr, Si, and Afl were measured, and the boron In both cases, the composition was investigated to a depth of less than 1 micron.In the lateral dimension, areas as small as 1 square micron, as well as larger areas, were investigated using the scanning method. This is especially important for ingots that exhibit substantial phase separation on the order of 10 to 100 microns. For the optimal magnetic ribbon materials, almost no phase separation is observed. figure,
The main phase was apparently homogeneous with no separable grain structure above at least 0.2 microns. Preliminary transmission electron microscopy analysis is described in Example I and Example ■.

E. Crystal structure and Rietveld precision analysis The crystallographic structure of the ingot material and the subsequently obtained ribbons was determined using X-ray diffraction techniques.

Measurements were made using CuKα radiation (wavelength 1.54).
The measurement was carried out using an Otelco (manufactured by Phillips) powder diffractometer. A graphite reflected beam monochromator was used to eliminate the Fe fluorescence background. 40 for Xal
An excitation voltage of kilovolts and a current of 20 milliamps were used. Prior art θ-20 X-ray powder diffraction scans were analyzed for peak positions and intensities to confirm the crystal structure of the material. The numerical output of the 7X, T, degree step scanning method with 200 seconds per step and 0.05 degrees per step was further computer analyzed using Rietveld Powder Pattern Precision Analysis & Techniques to obtain accurate lattice constants and principal N for the selected materials.
The positions and occupancies of atoms in the d2Fe and 4B structures were measured.

Using this precision analysis technique, a small amount of bccα iron was also quantitatively identified.

F, magnetic measurement maximum applied magnetic field of 22 kOe L D J , I nc
, 19500 type computer system i! The magnetic properties were measured using a vibrating sample magnetometer (VSM). Measurements of the magnetic field value T1 were carried out under feedback control using a calibrated Hall effect detector. The measurement software was changed in-house.
It is now possible to measure both the main hysteresis loop and the small hysteresis loop of permanent magnet materials with strong coercive force. Before performing a series of measurements, the magnetization M was always calibrated using a standard (soft magnetic) nickel ball (American National Standards Bureau) that had been repeatedly measured. To calculate the magnetization of a magnetic material, Cahn
-21 Automatic electronic balance (accuracy up to 1 microgram) was used to measure the sample size (length 5 m + nFi! degrees, width 2 m).
(about 1 milligram or less) for a typical ribbon piece of about 30 to 50 microns thick, giving an estimated density of 19. For each of the materials in the following examples, the density was all matched to 7.4(]/CC, which is slightly less than 76 g/cc, which is the value appropriate for pure chemical proportion NdFe14B. Proportionally, the calculated value of magnetization would have been 6 more than the value reported in the example below.

Each electrical ribbon sample was attached to a sample holding rod using adhesive tape. In most cases, the samples were premagnetized in a given direction using a pulsed magnetic field (peak magnitude up to 120 kOe) generated by an LDJ, Jnc, capacitor discharge magnetizer. Such premagnetization is necessary for proper magnetic measurements of the high performance permanent magnet materials of the present invention, since the maximum magnetic field of a VSM magnet is usually insufficient to completely saturate the magnetic moment. There were many things that happened. After this, the sample is placed in the gap of the VSM magnet,
It was positioned at the saddle point of the detection coil. After standard procedures and before magnetization measurements, the magnetic field applied to the premagnetized sample was zero, while the field applied to the unmagnetized (virgin) sample was 5.
A magnetic field of kOe was applied. The measurement is performed by increasing the magnetic field from zero to the maximum value (usually 22 kOe), passing through zero and decreasing to the negative maximum value, and then increasing it again through zero to the positive maximum value. The entire <m M vs. applied magnetic field H) was recorded. The program then determines the main magnetic parameters, namely the residual magnetization M (positive y-intercept of the hysteresis curve) measured in kilogauss.
, the intrinsic coercive force Hc in kilo Oersteds (the negative
(maximum negative value of the product) was determined.

In most of the examples below, the ribbon sample is
Along the direction! ! Magnetic measurements were taken. First, in the ribbon plane, in the "PX direction" parallel to the spin (melt spinning) direction (parallel to the longitudinal direction), then in the ribbon plane, in the Y direction (parallel to the width, not perpendicular to the spin direction), and finally in the ribbon Measurements were made in the Z direction perpendicular to the plane. In each case, the sample was premagnetized in the appropriate direction using a pulsed magnetic field. In the perpendicular Z direction. Generally, a demagnetizing field coefficient of N = 0.75 is reasonable. I found that I could get the fix: make it bigger,
Typical N=1 for measurements perpendicular to the thin plate. If a value of f71 is used, the magnetic measurements would be even higher than the values given in the example below (although the hysteresis loop would appear in a physically impossible shape). The coercive coefficient Q8 in the Z direction is calculated using N=0.75.

Value of coercive coefficient Q'S u mx, y, 'z (Mr/Ms)?ゲ Calculate using the Z direction km relationship T 〇 = 0.75. In all examples we use the theoretical maximum value of G of 16 for Ms.

Do not underestimate the applied magnetic field required to fully saturate these materials. In particular, once a virgin (unmagnetized) sample is magnetized in one arbitrary direction, it is often extremely difficult to completely remagnetize it in another direction. An example of such a magnetic training effect will be described next.

Example I - Samples 471 and 477 Series Alloy grains were arc melted in high purity argon (99.98%) to solidify ingots. Starting materials are Fe99.99%, Nd99
．． The quality was 9%, 8997%, and 3i 99.99%. Remelt and restore alloy grains to 1nvert
ed) was performed five times. A Zr getter (for oxygen) was used, but there was no noticeable contamination.

A pair of ingot pieces and magnetically measured melt-spun ribbons (
(twin) was examined using energy dispersive scanning electron microscopy. [Twin J refers to a section taken next to the ribbon piece that was magnetically measured; therefore, the composition and microstructure are generally considered to be the same. The results are shown below.

Example FeNd5iB ingot: 471AC80,513,506 ingot: 471A 79.6 14.4 0 6
Ribbon: 471ACO1■ 79.6 13.2 1
．． 2 6 Ribbon: 471ACO1 (4) 79.
5 13.1 1.4 6 [The 471AC alloy incorporated about 1.3 at.% of Sl after melt spinning] Example pe ■ Dan-L Dan ingot: 477AB 78.2 14.2 1
．． 6 6 Ribbon: 477AAO1■ 77.8 1
3.9 2.3 6 ribbons = 477＾AO1(4)
80.6 10.7 2.7 6 ribbon: 477
AAO1 (1) 76.4 14.1 3.5 6
[477AA has approximately 1.2 at.% S after melt spinning
Each ingot was spun under identical conditions except for wheel speed (melt spinning, i.e. ribbon formation in a single roll process). The wheel speed is the surface speed, 20.25.3
It was set to 0 m/sec. The conditions for the count were as follows.

Crucible type 2 Quartz crucible diameter: 17mm (inner diameter) 19+++
m (outer diameter) Crucible orifice diameter: 1 mm Crucible pressurizing gas: Argon crucible discharge pressure: 2 psi Crucible wheel opening I11:8111 Crucible axis:
Vertical (90') Wheel type: Cu Wheel diameter: 10 inches Wheel smoothness: #600 grid Indoor gas:
Argon air pressure: 1 atmospheric pressure filling maximum = 18-21Q Melting means: High-frequency induction heating at 400 kHz Time: 30-50 seconds Extrusion temperature: 1350°C [With an optical pyrometer with emissivity set to 0.55] Spin time : It should be noted that the melt spun ribbon is always higher than the ingot in the 3I content of 2 to 4 seconds. The molten alloy takes in Si from the quartz crucible. This shows that it is necessary to measure the composition at each stage of manufacturing.

The average value of the magnetic properties was taken from the results of three separate tests. Tests were conducted using stacks of 10 individual ribbon pieces each. 120kOe to the ribbon for each measurement
A magnetic pulse was applied three times for 1 millisecond. Magnetization measurement is
The test was conducted while applying a maximum magnetic field of 220 kOe in VSM.

A magnetic field was applied parallel to the spin direction within the ribbon plane.

Sample number Wheel BHm Hc
Mr Ribbon Ribbon width Ta (HGQe)
(koe) (kG) (7) Thickness II
/sec) (micron) (number) 4118801 20
14.3±0.6 16.1±0.3 9.
7±0.1 41±113 1.5±0.6
471AC012518,6±1.2 12.4±0.
2 10.5±0.2 33±8 1.4±0.44
718 DOI 30 14.7±1.
2 16.4±0.5 10.4±0.1
30±8 1.3±0.34778801
20 13.0±1.6 17.
4±0.5 9.2±0.3 42±7
1,8±0.34778BO1ning 25
12.3±1.3 16.9±0.6
9.2±0.1417^001 25 15.2
±2.7 17.9±0.8 9.2±0437±1
1 1.5±0.44778 COI 3
0 11.6±1.6 16.3±0.6
10. i±0.2 35±11 1
．． 7±0.8 Note *11 orifices were chained and clogged during this melt spinning.

The data above shows the peak energy product versus wheel surface speed. The optimum wheel speed is 25 m/sec for a 10" diameter wheel.

Correlation between microstructure and magnetic properties Sample number B11m Hc Mr Microstructure (HGOe) (kOe) (kG) 471A
CO1 (3) 25.3 12.7 11.0
Homogeneous fine crystal grains 4778^01 (5) 21,
8 17.7 10.3 Form only 471AD
O1 (4) 11.5 15.2 9.0
Fine grain area and thick 477 product 01 (3) 6,5
11.1 8.5 Even if the fine grain region (located on the wheel side of the melt spin ribbon) consisting of a thin grain region is polished and etched with an ethanol solution containing 1 to 2% nitric acid, it does not decompose (observed). Although it was not done, 2,0
It was determined to be less than 0.00 angstroms. This fine-grained region is identified as having good quality magnetic material, ie, high energy product and high remanent magnetization.

The other two samples exhibit a superposition of good magnetic material and bad material (the latter being fixed as bad by the presence of large grain areas on the free side of the ribbon) as shown in the magnetization curves.

Take・K Regarding the magnetic ribbon section that is the “pair” of sample #477ADO1 (1) with the quenching parameters and magnetic flux reported above, we performed a high-resolution scanning transmission electron microscopy f:ll mirror (ST
Preliminary analysis was performed using EM).

The ribbon sample is placed using standard micrometallurgical microscopy techniques and the thin section of the sample is thinned to about 1,000 angstroms using ion beam etching means.

Two ribbon surfaces (argon ions allow the upper half to
After removing microns by etching), the wheel side and free side were observed.

The microstructure on the wheel side is composed of very fine grains of predominantly tetragonal phase (rare earth to iron ratio consistent with Nd2Fe14B), with grain sizes ranging from 100 to 1000 angstroms, with electron Linear diffraction revealed that these grains were randomly oriented. There were also a small number of alpha iron grains with typical grain sizes greater than 500 Angstroms.

One or more other phases enriched in neodymium and sulfur compared to the main phase are present in the form of spherical inclusions at the boundaries between the large grains of the main tetragonal phase. Evidence found. Initially, sulfur was not considered a constituent of the material.

The position of Si remained difficult to measure.

On the free surface side of the ribbon, the grains of the main phase were much larger, reaching approximately 1 micron in diameter. There was slightly more silicon on the free surface side than on the wheel side. However, in this case as well, the exact position occupied by silicon in the microstructure was not clarified.

Example ■-Sample 376^QO9 (2) An ingot consisting of iron, neodymium, boron, and silicon was prepared according to the method described above. Average elemental analysis of the ingot was performed by EDS and W[)S, and the following results were obtained (
The ingot pieces were then placed in individual quartz crucibles, melted, and rapidly cooled to form ribbons as described above. The quenching parameters were as follows.

Crucible: 3/4″ Quartz crucible orifice diameter: 0.92mm Crucible pressurized gas:
Argon crucible discharge pressure: 5 psi Distance between crucible wheels: 1/8″ Fill height: 40 g Crucible axis: Vertical shroud gas: He 10 psi Scraper gas: Ar 15 psi Room gas: N
2 Indoor pressure (gauge): 0 psi Wheel diameter: 20 inches Wheel speed: 1200 rpm Wheel smoothness: 240 Grid Ribbon thickness = 15-20 microns Ribbon width: 1.8 mm + production rate): 32 g As-quenched ribbon As a result of magnetic measurements, 3
It was found that the daily max was low at 1 to 2 megagauss Oersted, and the holding power was also low at 1 to 3 koe.

The as-quenched magnetic alloy material is considered to be substantially isotropic.

The ribbons obtained from the products were wrapped in tantalum foil and encapsulated in quartz in the presence of argon. After annealing at 650° C. for 6 hours, the following magnetic parameters were obtained using the method described above.

Q (measured value) 0.9IQ
1.03 Other ribbon parameters taken from this material were:

Energy product Coercive force (HGOc)
(KOe) 1B, 89.8 16.2 13.7 As is clear from the above results, another way to obtain a material with Q of 1 or more is to rapidly cool the melt and change the precursor structure, e.g. This is a method of forming an amorphous or microcrystalline structure smaller than the characteristic grain size RO. This non-interactive structure is converted into a suitable structure with enhanced magnetic properties by appropriate heat treatment.

In the example given here, the maximum magnetic properties of the final product are a
Not optimized.

Example ■-Sample 4698A12 (1) An ingot consisting of iron, praseodymium, neodymium, boron, and silicon was prepared according to the method described above. The average elemental analysis of the ingot was carried out by EDS and WDS, and the following results were obtained (atomic%).

Iron 15.4
Neodymium 15.1 Boron 6.3 Silicon 0.5 Aluminum 2.7 Next, the ingot pieces were placed in individual quartz J crucibles.
It was melted and rapidly cooled to form a ribbon as described above. The quenching parameters were as follows.

Crucible = 2″ quartz (Tw) Crucible orifice diameter: O, [i5m Crucible pressurization gas: Argon crucible discharge pressure; 2pSi crucible distance between wheels: 5rIP Filling: 500’j Crucible @:
Vertical room gas: Argon room pressure (gauge): -6 (l-H) Wheel diameter: 12 inches Wheel speed: 900 rpm Wheel smoothness: 600 Grid wheel thickness: 35-62 ± 6.35 microns Ribbon width' 1.49±0.349
mm Yield: 375 g The magnetic parameters obtained for the as-quenched ribbon using the method described above were as follows.

Q (measured value) 1.33Q
Ribbon fragments were analyzed by shape-dependent demagnetically corrected 1.45 EDS and WDS of the magnetic field perpendicular to the ribbon plane. The elemental analysis results of the ribbon fragments were as follows (unit: atomic %): Iron 75.
99 neodymium 14.78 boron
6,64 Silicon 2.58 Other ribbon parameters spun and quenched from this material were: energy product;
Coercive force ()IGOe)
(to Qe)24.4
Greater than 22.0 16.6 1
8.8 Such materials illustrate that measurements of high coercivity materials are difficult to accurately measure. When a ribbon that has been checked for isotropy is measured at 180 degrees from the original magnetization direction, the value of B Hmax decreases from 30.5 to 26.8, and when remeasured in the original direction it decreases to 31.6. Got the value. Since these two measurement directions are crystallographically equivalent, it can only be assumed that the reason for such a difference is that the material cannot be completely saturated in the opposite direction. At this time, the material is normally magnetized with the same 12T pulse in each direction, so once the material has been magnetized, the magnetic field required to reverse the magnetization direction must be larger than the magnetic field required to initially magnetize it. It shows that there is a need. In reality, this is
In what is called a "training" phenomenon, the initial magnetization direction becomes the material's preferred magnetization direction.

Example ■ (Transmission electron micrograph of a cross-section of a magnet flake that has been rapidly cooled) A set of Nd-Pr-Fe-B-3ila alloy flakes that have been cooled was prepared and examined using high-resolution transmission electron microscopy. Inspected. For high-resolution transmission electron microscopy, thin flakes were cut in the thickness direction so that the uniformity of the microstructure across the flake thickness (30 to 40 microns) could be examined.

One of the same flakes was crushed and magnetically measured. 2
The material obtained from the WA milling process was used. Ingots and flakes were prepared using the method used in Examples 1 to 1 above.

The composition of manufactured portion 502ABO1 (according to plasma and chemical analysis measurements) is N d 12.0 atomic %, Pr 0.05 atomic %, 85.6 atomic %, 3i 1.7 atomic %, balance FeT.
``There was. 4009 loading material from 0.8 training hole to speed 2
It was discharged at a pressure of 3 psi onto a copper wheel rotating at 2 h/sec. During part of this operation, the temperature of the melt was raised to over 1400°C to eliminate clogging of the orifice.

”The composition of N400AA10 (according to plasma and chemical analysis measurements) is 69.8 atomic% of N, pr2.5
atomic%, B5.9 atomic%, 3i0.5 atomic%, balance 1"
It was e. w7A rotating a 329 charge material from a 0.8++o+ diameter orifice at a speed of 24m/sec! II
It was discharged onto the wheel at a pressure of 3 psi.

Three flakes with significantly different magnetic properties were selected from manufactured valve 502ABO1 and subjected to detailed analysis.

Sample 502^BO1 (4) was a large grained ("incompletely quenched") sample. The fine structure is shown in a micrograph in Fig. 4Δ. Figure 4B shows the corresponding hysteresis loop. The characteristic grain size of this crystal grain is 1000 Å
Above, the magnetic performance is B Hmax is only 28GO
It is inferior to c.

Sample 502ABO1 (8) was a small grained ("over-quenched") sample. The fine structure is shown in Fig. 5Δ. Figure 5B is the corresponding hysteresis loop. Although the structure appears to be amorphous, electron diffraction shows the presence of some crystallites with a particle size of about 50 angstroms or less. The magnetic energy product is the coercive force)
-1c was less than 4000e, so it was very low (I
HGOe). Since the interaction effect was strong, the residual magnetization Mr was large.

Sample 5502ABO1 (35) was selected on the basis of having grains close to the optimum grain size to enhance interaction. The microstructure is shown in FIG. 6A, and the corresponding hysteresis loop is shown in FIG. 6B. The grains are approximately 200-300 angstroms in diameter and randomly oriented with respect to each other. The particle size distribution around the mean value can be determined graphically and quantitatively by well-known methods. When further enlarged (not shown), it was found that the grain boundaries were clear with no noticeable second phase. Magnetic energy product is 18HGOe, residual magnetization is 9
KG or more, which corresponds to a Q value of 1.0 or more.

Sample 400AA10(6) is a sample taken from a different production valve and also shows the optimum particle size. That fine a
The 4M structure is shown in FIG. 7A and the corresponding hysteresis loop is shown in FIG. 7B. In this case as well, the diameter of the crystal grains is 200~
300 angstroms, and the microstructure looks very similar to that in Figure 6Δ, but the composition is slightly different. The magnetic parameters are also very similar to sample 502ABO1 (35), with (BH)max of 158 GOc or more and Q of 1.0 or more.

Judging from the results of electron diffraction, all samples showed RE2Fe14B as the main phase (related to the material), although the samples that had been cooled too quickly had a considerably broader peak than the others.
It is considered that it has a tetragonal system and a Pr2/mnm structure.

Although the present invention has been explained with reference to the preferred examples and preferred embodiments, this does not limit the scope of the invention, and the scope of the present invention is limited only by the claims. .

[Brief explanation of drawings]

Figure 1 shows the theoretical M vs. H hysteresis loop of an isotropic material when H (grain, grain) = 48 KOe for non-interacting grains (solid line) and interacting grains (dotted line). FIG. Figure 2 shows (BH) of an isotropic material with uniaxial grains.
It is a graph theoretically representing the relationship between max and the characteristic grain size R of exchange interaction crystal grains. Figure 3 shows the coercive coefficient Q of an isotropic material with uniaxial grains.
3 is a graph theoretically representing the relationship between R and the characteristic grain size R of the material of FIG. 2 having exchange-interacting crystal grains. Figure 4A shows an 8 inch
This is a transmission electron micrograph of a cross section of an inch copy enlarged at a ratio of 500 angstroms to 1,1α. Figure 4B is a hysteresis loop for the material shown in Figure 4A. FIG. 5A is a transmission electron micrograph of a cross section of an 8 inch by 10 inch copy of the small grain, low energy product alloy of Example IV, magnified at a ratio of 500 Angstroms = 1.3 cm. Figure 5B shows the hysteresis loop for the material shown in Figure 5A. Figure 6A shows the high energy product material of Example 5 (Sample 502ABO) having a particle size and particle size distribution suitable for quantum mechanical exchange coupling.
1(35)) is a transmission electron micrograph of a cross section of an 8 inch x 10 inch copy enlarged at a ratio of 500 angstroms to 1.3 cm. FIG. 6B is a hysteresis loop for the material shown in FIG. 6A. Example IV High Energy Product Material (Sample 400^A10(6
)) is a cross-sectional transmission electron micrograph of an 8 inch x 10 inch copy enlarged at a ratio of 500 angstroms = 1.65o++. FIG. 7B is a hysteresis loop for the material shown in FIG. 7A. Engraving of drawings (no change in content) Engraving of drawings (no change in content) Fj&,'LA Engraving of drawings (no change in content) Engraving of drawings (no change in content) Procedural amendment March 1988 12th 1, Incident Display Patent Application No. 3108 of 1988 2,
Title of the invention Permanent magnetic alloy with increased residual magnetization and its bulk magnetic material 3, Relationship with the Xi case for amendment Name of patent applicant (665) Nippon Steel Corporation stock * Hoe 5, Date of amendment order Spontaneous addition do. (No change in content) Procedural amendment (method) % formula % 1. Indication of case Patent Application No. 3108 of 1988 2.
Title of the invention Permanent magnetic alloy with increased residual magnetization and its bulk magnetic material 3, Relationship with the amended person's case Name of patent applicant (665) Nippon Steel Corporation 4, Agent Shinjuku, Shinjuku-ku, Tokyo 1-1-14 Yamada Doru 8, Contents of the amendment (1) In the specification, on page 82, line 3, the phrase “cross-sectional transmission type...” was replaced with “cross-sectional crystal structure transmission type. ...”
and correct it. ■ In the layer, page 82, line 10 says, “Transmission type of cross section...
'' should be corrected to read ``transmission type of crystal structure in cross section...''. G) In the layer, on page 82, line 17, it is written that ``Transmission type of cross section.
``...'' is corrected to ``Transmission type of crystal structure in cross section...''. (4) In the layer, page 83, line 7 says, “Transmission type of cross section.
``...'' is corrected to ``Transmission type of crystal structure in cross section... Regarding Figure 7A (photo), we submitted it in the written amendment to the application dated March 12, 1988 (voluntarily).

Claims (16)

[Claims] (1) The coercive coefficient Q is greater than 1.0, the maximum magnetic energy product is greater than 15HGOe, the residual magnetization is greater than 9KG, and the crystallites are substantially composed of a solid mass of crystallites, each of which is adjacent to each other. A magnetically substantially isotropic magnetic material that is in contact with crystallites through grain boundaries and whose orientation is not crystallographically aligned with adjacent crystallites at the grain boundaries, (a) each crystallite has an axis of easy magnetization; (b) the crystallites and grain boundaries are dimensioned such that quantum mechanical exchange coupling across the grain boundaries occurs between surface atoms of adjacent crystallites; A hard magnetic material having a geometry. (2) The crystallites have an average radius such that the magnetic field caused by grain-to-crystalline interactions due to surface exchange interactions between adjacent crystallites is substantially equal to the anisotropic magnetic field of each individual crystallite. The hard magnetic material according to claim 1, which has R_0. (3) The hard magnetic material according to claim 2, in which the magnetic surface exchange energy is strong enough to orient crystal grains in a direction deviated from their axis of easy magnetization. (4) The hard magnetic material according to claim 2, wherein the anisotropy energy of each crystallite becomes stronger as the coercive force becomes larger than about 8 KOe. (5) The hard magnetic material according to claim 1, which comprises an alloy of a rare earth metal and a transition metal. (6) The hard magnetic material according to claim 5, wherein the alloy further contains boron. (7) The hard magnetic material according to claim 6, wherein the nominal composition of the alloy is RE_2TM_1_4B_1, where RE is one or more rare earth metals and TM is one or more transition metals. . (8) The hard magnetic material according to claim 7, wherein the rare earth metal is selected from the group consisting of praseodymium and neodymium. (9) The hard magnetic material according to claim 7, wherein the transition metal is selected from the group consisting of iron, cobalt, and nickel. (10) The hard magnetic material according to claim 7, which comprises a modifying element. (11) The hard magnetic material according to claim 10, wherein the modifying element is selected from the group consisting of Al, Si, and mixtures thereof. (12) The modifying element is a grain refining agent (grain refining agent).
The hard magnetic material according to claim 10, which is a fining agent. (13) The crystal grain refining agent adjusts the competitive nucleation rate and crystal grain growth rate, and when the average crystallite radius R_0 is about 200 Å, the reduction in coercive force and the formation of multiple domains in the crystallites are substantially suppressed. 13. The hard magnetic material according to claim 12, which is made into a solid hard magnetic material having a crystallite radius distribution that can be avoided. (14) 75<a<85, 10<b<20, 5<c<1
When 0 and d are up to 5, the formula Fe_a(Nd,Pr)_bB_c(Si,Al)_d
The hard magnetic material according to claim 1, having a tetragonal main phase with a nominal composition of P4_2/mnm. (15) RE_2Fe_1 where the boron does not interact
_4B is displaced relative to the boron atom site,
A hard magnetic material according to claim 14. (16) The hard magnetic material according to claim 14, wherein the average radius R_0 of each crystallite is about 200 Å.