Takamizawa, Hideo (Tokyo, JA)
Yotsuyanagi, Keiichi (Tokyo, JA)

05/130343

10/02/1973

04/01/1971

Download PDF 3763045       PDF help

Click for automatic bibliography generation

Nippon Electric Company, Limited (Minato-ku, Tokyo, JA)

252/62.570

252/62.560, 252/62.630, 252/62.590

C04B35/26; C04B35/50

252/62.57,62.59,62.63,62.56

3447851

FERRIMAGNETIC LIGHT TRANSMISSION GARNETS

June 1969

Remeika et al.

Geller et al., "Journal of Applied Physics", Vol. 35, No. 3, pp. 570-572, March 1964. .
Yudin et al. "Physics Letters", Vol. 28A, No. 7, PP. 483-484, January, 1969..

Vertiz, Oscar R.

Cooper J.

what is claimed is

1. Ferrimagnetic garnets having compositions expressed by the formula:

2. 3 ≤ x ≤ 0.5,

3. 2 ≤ y ≤ 2.0, and

4. Ferrimagnetic garnets having compositions expressed by the formula:

5. 25 ≤ x ≤ 0.5,

6. 2 ≤ y ≤ 2.4,

7. 2 ≤ z ≤ 0.4,

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates to calcium-vanadium (Ca-V) ferrimagnetic garnets for use in microwave circuit elements operating in the VHF, UHF or SHF band range. Among the required characteristics of magnetic materials for use in such microwave circuit elements are low magnetic losses and small temperature variations of the saturation magnetization values (4 π Ms). Desired values of 4 π Ms will vary according to the application of the magnetic materials. The essential condition for reducing the magnetic loss is that the ferromagnetic resonance linewidth (ΔH) be as low as possible. It has been known that the higher the Curie temperature, the lower becomes temperature variation of the saturation magnetization (4 π Ms) and that in order to lower the value of the linewidth (ΔH), which is subject to change with the sintering density and the presence or absence of a second phase, the sintering densities must be made sufficiently large and there should be no second phase formation.

Description of the prior art

The yttrium-iron garnets (YIG) that have been most commonly used as magnetic materials in microwave applications offer the advantages of lower 4 π Ms values, higher Curie temperatures and lower magnetic losses than conventional spinel-type ferrites such as nickel series ferrites or magnesium-manganese series ferrites. These advantages of the yttrium-iron series garnets are considerably offset by detects such as the necessity for the use of yttrium oxide which is an expensive raw materials and the need for sintering at extremely high temperatures and for long time intervals which are not suited for large-scale industrial production.

It has been proposed to substitute part of yttrium-iron garnets with A1 ₂ 0 ₃ for lowering the saturation magnetization. See Physical Review, vol. 110, 1958, p. 73. Such garnets have another defect of a rapid decrease in the Curie temperature which inevitably causes a large variation of 4 π Ms with a temperature variation.

The unmodified Ca-V garnet is known for featuring high Curie temperatures in spite of low 4 π Ms, but its high Δ H rendered practical application extremely difficult.

It has been also proposed to substitute part of the Ca-V garnet with yttrium and indium oxides in order to decrease Δ H. See Material Research Bulletin, vol. 4, 1969, p. 825 -838. However, such substitution causes decrease in Curie temperature and increase in 4π Ms.

It is, therefore, the object of the invention to provide improved microwave circuit element materials for use in the VHF, UHF or SHF band range, having excellent characteristics such as low ferromagnetic resonance linewidth values and high Curie temperatures.

The garnet compositions are generally expressed by a normal formula unit (A ₃ ) (B ₂ ) (C ₃ )0 ₁₂ , where the first, second and third parentheses represent respectively the 24c, 16a and 24d sites and A, B and C denote atoms occupying the respective sites. Fe has a preference for the two different sub-lattice sites (the 16a and the 24d sites) and the Fe-Fe super-exchange-interactions in each of and between these sub-lattic sites cause the Fe magnetic moments at the 16a and 24d sites to be coupled anti-ferromagnetically. Under the situation of the relative site unbalance, wherein the magnetic moment at the 24d site is not equal to that at the 16a site, the garnet compositions manifest ferrimagnetism. It has been generally considered that the compositions manifest anti-ferromagnetism in case of the site balance and at which an abnormal phenomenon of Δ H occurs.

The value of 4 π Ms is determined by the relative site unbalance in the magnetic moment between the 16a and 24d sites for which Fe has a strong preference, while the temperature variation of 4 π Ms changes with the numbers of iron ions located on each sub-lattice site, kinds of non-magnetic ions replacing the iron ions, and kinds of ions located on the 24c site.

The unmodified calcium-vanadium garnet can be expressed by a normal formula unit (Ca ₃ ) (Fe ₂ ) (Fe ₂ ) (Fe ₁ .5 V ₁ .5)0 ₁₂ .

SUMMARY OF THE INVENTION

According to the invention, calcium-vanadium garnets are improved by substituting Ca ions on the 24c site and Fe ions on the 16a site with Y ions and Sn ions, respectively, or by further substituting ions on the 24d site in the thus-modified garnets with Ge ions. The improved calcium-vanadium garnets are featured by sufficiently low Δ H values, high Curie temperatures, 4πMs values within controllable suitable ranges, small temperature variations in the value of 4πMs and low manufacturing costs.

The calcium-vanadium garnet compositions substituted with Y and Sn of this invention are expressed by the chemical formula:

(Ca _{3 -y} Y _y ) (Fe _{2 -x} Sn _x ) (Fe ₁ .5 ₊₀ .5x ₊ 0.5y V ₁ .5 _-0 .5x _-0 .5y)0 ₁₂ ,

in which the values of x and y are required to satisfy the following relations:

x > 0

x ≤0.35y +0.3

1.5 - 0.5x -0.5y >0

0 <y ≤2.5

Where the difference between numbers of Fe ions located on the 16a site and the 24d site is 0.1 or less, an abnormal phenomenon of Δ H is observed and the values of Δ H cannot be improved. Of course, ferrimagnetism disappears if the difference is zero. Thus, the range of │(2 -x) - (1.5 + 0.5x +0.5y) │ ≤ 0.1 , i.e. 0.8 ≤3x +y ≤1.2, should be excluded from the above-mentioned Y-and Sn-substituted Ca-V garnet compositions of this invention.

It is preferable that the values of x and y are within the ranges of 0.3 ≤x ≤0.5 and 1.2 ≤y ≤2.0, respectively. With such compositions, the value of Δ H is less than 10 oersteds.

According to this invention, the Y-and Sn-substituted Ca-V garnets can be further substituted with Ge. The garnet compositions thus obtained can be expressed as

(Ca _{3 -y} Y _y )(Fe _{2 -x} Sn _x )(Fe ₁ .5 ₊₀ .5x ₊ 0.5y _- 0.5z Ge _z V ₁ .5 _-0 .5x _-0 .5y _-0 .5z)0 ₁₂ ,

in which x, y and z should be within the ranges of 0 < x ≤ 0.5, 1.0 ≤ y ≤ 2.4 and 0 < z ≤ 0.5, respectively and they should satisfy the relation of 1.5 - 0.5x - 0.5y - 0.5z > 0. Where the difference of the numbers of Fe ions between 16a and 24d sites │(2 - x) - (1.5 + 0.5x + 0.5y - 0.5z)│ is 0.1 or less, the abnormal phenomenon of Δ H is observed, as mentioned previously. Accordingly, the compositions satisfying the relation of 0.8 ≤ 3x + y - z ≤ 1.2 should be excluded. It is preferable that x, y and z lie within the ranges of 0.25 ≤ x ≤ 0.5, 1.2 ≤ y ≤ 2.4 and 0.2 ≤ z ≤ 0.4, respectively. The Y-, Sn- and Ge- substituted Ca-V garnets within such ranges have low Δ H of less than 20 oersteds.

BRIEF DESCRIPTION OF THE DRAWING

FIGS. 1, 2 and 3 show, respectively, ferromagnetic resonance linewidths (Δ H), Curie temperatures and 4 πMs as a function of x for the Y- and Sn-Substituted Ca-V ferrimagnetic garnets of this invention with the compositions expressed by the formula mentioned above;

FIG. 4 is a diagram illustrating the effective range of the values of x and y in the Y- and Sn-substituted Ca-V garnet compositions expressed as above;

FIG. 5 shows the effect of x on 4 πMs and Δ H for the Y-, Sn- and Ge-substituted Ca-V garnet composition expressed by the mentioned formula where y = 1.4 and z = 0.3;

FIGS. 6 and 7 show the effect of y on Δ H and Curie temperature, respectively for the Y- and Sn-substituted Ca-V garnets;

FIG. 8 shows influence of z(Ge) on Δ H and 4 πMs for the Y-, Sn- and Ge-substituted Ca-V garnet compositions expressed by the formula mentioned above; and

FIG. 9 is a graph of the 4πMs versus temperature characteristics of garnet compositions of the prior art and of this invention which illustrates the advantages of this invention.

DETAILED DESCRIPTION

Samples mentioned below were prepared by thee method well known in the art. In detail, starting materials CaCO ₃ , Fe ₂ 0 ₃ , V ₂ 0 ₅ , Sn0 ₂ , Ge0 ₂ , and Y ₂ 0 ₃ in such amounts were weighed, 350 grams in total in each case, so that each of the compositions shown in Tables may be finally obtained. These materials were admixed in a ball mill made of steel, presintered at 900° C for 4 hours, compressed into the desired shapes, and then sintered at a temperature of 1210° C to 1300° C for 10 hours in air. The sintered products were removed from the furnace when the furnace temperature cooled down to 300° C. Then the values of the saturation magnetization (4 πMs) at room temperature (23° - 25° C), the linewidth (Δ H) at 9.5 GHz, and the Curie temperature were measured.

Table 1 lists the results of measurements for the unsubstituted, Ge-substituted, Sn-substituted, Sn- and Ge-substituted, Y- and Sn- substituted and Y-, Sn- and Ge-substituted Ca-V garnet compositions to demonstrate the successively promoted substitution effects, as regards a decrease in the linewidth Δ H and an increase in the Curie temperature. Sample No. 1, or unmodified Ca-V garnet has low 4 πMs and high Curie temperature, but Δ H is as high as 370 (i.e., large magnetic loss). Therefore, practical use of this garnet is substantially impossible. It will be noted with No. 2 sample in which a fraction of Fe ion on the 24d site is replaced with Ge, that the Ge-substitution has contributed to a reduction in Δ H (110), or less than one-third of that for the unmodified Ca-V garnet. The effectiveness of Sn-substitution, or the substitution of Sn for a fraction of Fe ion on the 16a site of the unmodified Ca-V garnet compositions as indicated in sample Nos. 3 and 4, for reduction in both 4πMs and Δ H will also be appreciated. The Δ H value of bout 190 has become almost one-half of that for the unmodified Ca-V garnet composition. A comparison of No. 3 and No. 4 samples, however, indicates that the Curie point has decreased from 200° C to 135°C with an increase in the amount of Sn-substitution.

An inspection of samples No. 5 and No. 6 indicates that the simultaneous Ge- and Sn-substitution is effective for a further reduction in Δ H. It will be appreciated that the values of Δ H have become less than those for No. 3 and No. 4 samples, or less than one-fourth and one-tenth of that for the unsubstituted Ca-V garnet composition. The Curie temperature, however, has fallen to 150° C or to 120° C. To compensate for the lowering of Curie temperature with increasing Sn- or simultaneous Sn- and Ge-substitution, a fraction of Ca ion on the 24c site is further replaced with Y as seen in the No. 7 and No. 8 compositions. This has succeeded in a further reduction in Δ H, for example, 32 (No. 7 sample) and 26 (No. 8 sample) and at the same time, raising the Curie temperature, for example, 222° C (No. 7 sample) and 200° C (No. 8 sample). In other words, the Curie temperature which has been lowered by Sn- or simultaneous Sn- and Ge-substitution can be raised by the simultaneous Y-substitution. Thus, Y- and Sn- or Y-, Sn- and Ge-substitution in the Ca-V garnet composition has succeeded in realizing low 4πMs, low Δ H, high Curie temperature, and improved temperature stability of 4πMs.

To evaluate the effect of Sn-substitution, several samples were prepared by varying x, with y fixed at 0.5 and 0.8 for z = 0, and sintering at 1260° C for 10 hours and others by varying x, with y fixed at 1.2 and 1.5 for z = 0, and sintering at 1800° C for 10 in the Ca-V garnet compositions expressed as

(Ca _{3 -y} Y _y )(Fe _{2 -x} Sn _x )(Fe ₁ .5 ₊₀ .5x ₊₀ .5y _-0 .5z Ge _z V ₁ .5 _-0 .5x _-0 .5y _-0 .5z)0 ₁₂ .

FIG. 1 indicates the dependence of Δ H on x for these samples. The effectiveness of Sn-substitution for the improvement in Δ H will be readily appreciated from the fact that Δ H decreases rapidly with increasing x in curves y = 0.5, 0.8, 1.2, and 1.5.

For instance, Δ H exhibits an extraordinarily large value near x = 0.167 for y = 0.5 by the abnormal phenomenon mentioned previously. But, Δ H decreases with increasing x beyond this value to reach Δ H = 63 at x = 0.5. In other words, Δ H has been improved to a value less than 1/3 of that at x = 0. In similar manner, it takes an extraordinarily large value in the vicinity of x = 0.067 for y = 0.8, but decreases with increasing x beyond this point to reach Δ H = 60 at x = 0.5. Thus Δ H has been improved to a value of the order of 1/4 of that at x = 0. The value Δ H = 32 at x = 0.5 for y = 1.2 indicates an improvement of Δ H equivalent to 1/7 of that at x = 0, while Δ H = 14 at x = 0.7 indicates an improvement equivalent to 1/12 of that at x = 0. That Δ H can be improved by Sn-substitution under the Y-substitution will be readily evident from these curves in FIG. 1.

Referring to FIG. 3 which indicates dependence of 4πMs on x it will be seen, in each case, that 4 πMs increases with increasing x to reach a maximum at x = 0.5 and then, decreases with increasing x. It can also be noted that the 490 Ms maximums at 0.5 become higher with increasing Y-substitution, reaching the highest 4πMs = 1150 gauss for y = 1.5. These curves further demonstrate that 4πMs can be varied in a wide range, 200 to 1250 gauss, by the Sn-substitution under the presence of Y and Δ H is markedly improved in the range of large 4πMs values.

In order to evaluate the effect of Sn-substitution in case z ≠0, that is, under the coexistence of Y and Ge, the value of x was varied in the range 0 to 1.0 with z fixed at 0.3 and y at 1.4 in the formula mentioned above. Saturation magnetization 4πMs and linewidth Δ H as a function of x for these samples which were prepared by sintering at 1300° C for 10 hours are shown in FIG. 5. The 4πMs curve (solid line) indicates that 4πMs increases at first with increasing Sn-substitution to reach a maximum 680 at x = 0.5 and then, decreases with increasing x to become less than 100 at x = 1.0. The Δ H curve (dotted line) indicates that Δ H is 185 Oe at x = 0, decreases with increasing x to reach Δ H = 28 at x = 0.3 and Δ H = 20 at x = 0.5 and then, increases steadily with increasing x.

The Sn-substitution under the coexistance of Ge and Y readily reveals a marked contribution to reduction in Δ H in the range in which 4πMs values are not so large, for instance, 680. This advantage is considerably offset, however, by the decreasing tendency of Curie temperature with increasing x as will be seen in FIG. 2 for z = 0 and in Table 2 for z ≠0. It will be noted in FIG. 2 that whereas the Curie temperature decreases with increasing x, it becomes higher with increasing Y-substitution. For instance, the Curie temperature becomes less than 100° C in the vicinity of x = 0.6, 0.7, and 0.9 respectively for y = 0.5, 0.7, and 1.2, while it remains well over 100° C even at x = 1.0 for y = 1.5. Furthermore, the slopes of these curves become less steep with increasing y. For instance, the Curie temperature decreases 125, 110, 97, and 87 degrees Centigrade respectively with an increase in x from 0 to 0.5 for y = 0.5, 0.8, 1.2, and 1.5.

Although the Curie temperature rises with increasing y, but the results of extensive experimentation conducted by us has proven that the Ca-V garnet compositions, which exhibit Curie temperatures most suitable for practical application, must have x and y values meeting the relation of x ≤ 0.35y + 0.3. The compositions which do not meet this relationship have been found to be difficult for reduction to practice in having low Curie temperature and large variations of 4πMs with temperature. Referring to FIG. 4, the hatched area indicates the effective range of the values of x and y, where z = 0, in the garnet compositions expressed by the formula mentioned above. The area between the lines of 3x + y = 0.8 and 3x + y = 1.2 is excluded, because the abnormal phenomenon of Δ H occurs there.

An inspection of Table 2 will reveal that the Curie temperature decreases with increasing x, or Sn-substitution, when z ≠0, falling below 150° C for x in excess of 0.5. Accordingly, the range of x suitable for practical application should be x ≤ 0.5 when temperature variations of 4πMs are taken into consideration.

To conclude, materials suitable for practical use with loww Δ H, high 4πMs (1250 gauss at maximum), satisfactory Curie temperatures can be manufactured within the range of x meeting the relationship x ≤0.35y + 0.3 for z = 0 or within the range of x meeting the relationship 0 < x ≤ 0.5 for z ≠0.

The evaluate the effectiveness of Y-substitution, several samples with x = 0.3, 0.5, and 0.7 for z = 0 were prepared by sintering at 1260° C for 10 hours for y ≤0.8 and at 1300° C for 10 hours for y > 0.8. FIG. 6 shows Δ H as a function of y and FIG. 7 shows Curie temperature as a function of y for these samples.

Referring to each curve in FIG. 6, it will be noted that Δ H decreases with increasing y until it reaches a minimum and then, increases with increasing y. For instance, Δ H reaches a minimum 30 at y = 1.5 for x = 0.3, a minimum 20 at y = 1.5 for 0.5, a minimum 14 at y = 1.5 for x = 0.7. The values of these minimums decrease with increasing Sn-substitution and beyond these points, all increase in Δ H steadily to reach values approximately equal to or larger than those at y = 0 such as Δ H = 340 at y = 2.7 for x = 0.3, Δ H = 318 at y = 2.5 for x = 0.5, Δ H = 280 at = 2.3 for x = 0.7, provided Δ H corresponds to y which meets the equation 1.5 - 0.5x - 0.5y = 0 (i.e. no vanadium in the composition).

A rapid increase in Δ H for y >2.4 is attributed to the insufficiency of the prescribed sintering temperature of 1300° C for the maturity of sintering with increasing Y. In view of this fact, x and y values are defined as 0 < y ≤2.5; 1.5 - 0.5x - 0.5y > 0 in order to lower Δ H. The degradation of the sintering property with increasing Y will also be evident from the experimental data (z ≠ 0) set forth in Table 3, in which Δ H becomes a minimum Δ H = 20 at y = 1.4 (sample No. 5) and then increases the Y-substitution.

Referring to FIG. 7 which shows the effect of Y-substitution on Curie temperature, there is a tendency toward a gradual increase in the Curie temperature with increasing Y, demonstrating the possibility of compensating for the lowering of Curie temperature with increasing Sn content by Y-substitution. FIG. 7 also indicates a tendency toward the lowering of Curie temperature for 1.5 - 0.5x - 0.5y = 0 -- that is, when vanadium content is nil.

This tendency is notably conspicuous for the data for z≠ 0 contained in Table 3. For instance, the Curie temperature for sample No. 1 (Y = 0) is less than 50°C, but it increases with increasing Y-substitution to reach 200°C at y = 2.0 (sample No. 6). However, it decreases thereafter with increasing Y to reach a low value for sample No. 7 for which 1.5 - 0.5x - 0.5y - 0.5z = 0.

The Y-substitution is effective for lowering ΔH and at the same time, elevating the Curie temperature. But, with an increase in Y-concentration, there arises the need for elevating the sintering tmperature to bring sintering to maturity. For these reasons, the effective x and y ranges are defined as follows: 0 < y ≤2.5 and 1.5 - 0.5x - 0.5y >0 for z = 0; or 1.0 ≤y≤2.4 and 1.5 - 0.5x - 0.5y - 0.5z > 0 for z ≠ 0.

To evaluate the effect of Ge under the coexistence of Sn and Y, several samples of different compositions with z from 0 to 0.8, x = 0.3 and y = 1.4 in the composition (Ca _{3 -y} Y _y )(Fe _{2 -x} Sn _x ) (Fe ₁ .5 ₊₀ .5x ₊₀ .5y _-0 .5z Ge _z V ₁ .5 _-0 .5x _-0 .5y _-0 .5z)O ₁₂ . were prepared by sintering at 1300°C for 10 hours.

FIG. 8 shows the results of measurement for these samples, indicating both 4πMs and Δ H as a function of z. Data on Curie temperatures for these samples are shown in Table 4.

Referring to FIG. 8, it is noted that 4πMs decreases with increasing Ge substitution. For instance, values of 4πMs are 880, 530, and 200 respectively at z = 0, 0.3, and 0.8. The value of Δ H, however, decreases from 50 Oe at z = 0 with increasing z to reach a minimum 24 at z = 0.5. As this point is passed, Δ H gradually increases with increasing Ge-substitution.

The tendency toward a decrease in both 4πMs and Δ H by Ge-substitution under the coexistence of Sn and Y is clearly observed from these curves. Data of Table 4 indicate a steady decrease in Curie temperature with increasing Ge-substitution (z).

As z exceeds 0.5, the Curie temperature falls and the temperature variation of 4πMs becomes large. Therefore, an optimum z range for low Δ H and small temperature variation of 4πMs can be defined as 0 < z ≤ 0.5. Therefore, materials suitable for practical use with low Δ H, such as less than 50 within 0 < z ≤ 0.5 for x = 0.3 and y = 1.4, and optimum 4πMs values ranging between 880 and 380 gauss can be manufactured.

To evaluate how the proposed Ca-V garnet compositions contribute to the improvement in Δ H, several samples with x = 0.3 and 0.5, y from 1.2 to 1.8, and z = 0 were prepared by sintering between 1300°C and 1350°C for 10 hours. Table 5 lists values of 4πMs, Δ H, and Curie temperature for these samples.

Table 5 indicates that Δ H has been reduced to 10 Oe or less, demonstrating that these materials are eminently suitable for use in low-loss devices for microwave applications. The value Δ H = 2.5 for sample No. 6 was attained by use of No. 4 sample after careful polishing for very gentle sphere shaping. Low values of Δ H of this order were also achieved with other samples, Nos. 1, 2, 3, and 5, by observing the same polishing technique. The known Y- and In-substituted Ca-V garnets had Δ H of the order of 2.0 Oe, Curie temperatures as low as 140°C, and high 4πMs in excess of 1400 gauss. Generally speaking, the lower the value of 4πMs, the more difficult it becomes to reduce Δ H. The substituted Ca-V garnet compositions contemplated by this invention can claim to be superior, in practical application, to the known Y- and In-substituted Ca-V garnets in that Curie temperatures are as high as 160°C and 4πMs values are less than about 1200 gauss.

To demonstrate the effect of the proposed Ca-V garnets for improving the sintering properties, several samples according to the invention and a sample of the conventional yttrium iron garnet composition (sample No. 1) were prepared by sintering at various temperatures for 10 hours and their theoretical densities (sintering density/X-ray density) × 100 and Δ H values were measured, as listed in Table 6.

Table 6 indicates that the Ca-V garnets according to this invention maintain excellent theoretical densities in excess of 97 percent, whereas the theoretical density of the yttrium iron garnet is only 91.4 percent. In other words, to raise this value to the order of 97 percent, sintering at higher temperatures, 1450°C or higher for instance, would be necessary, as has been known among those skilled in the art. Data contained in this table demonstrate that the Ca-V garnet materials according to this invention can be manufactured at sintering temperatures lower than 1450°C by more than 150° Centigrade.

To demonstrate the feasibility of manufacture of Ca-V garnets with low 4πMs and small temperature variation of 4πMs, two samples of the composition according to this invention, (Ca ₁ .6 Y ₁ .4)(Fe ₁ .7 Sn ₀ .3) (Fe ₂ .2 Ge ₀ .3 V ₀ .5)O ₁₂ (curve a) and (Ca ₂ .0 Y ₁ .0)(Fe ₁ .7 Sn ₀ .3)(Fe ₂ .15 V ₀ .85)O ₁₂ (curve b), and an Al-substituted yttrium iron garnet Y ₃ Fe ₄ Al ₁ O ₁₂ (curve c) were prepared. All of these samples had room-temperature 4πMs values of the order of 500 gauss. FIG. 9 shows variation of 4πMs with temperature for these three samples.

A comparison of these curves a, b, and c readily reveals that the variation of 4πMs with temperature for the two samples according to the proposed composition is less, at or near room temperature, than that for the sample of the Al-substituted yttrium iron garnet. The values of Δ H for the sample corresponding to curve a and the Al-substituted yttrium iron garnet corresponding to curve c were in the range 40 to 60 Oe, while the value of Δ H for the sample corresponding to curve b was 26 Oe.

Table 1

Sinter- Sample Composition ing 4πMs ΔH Curie No. Temp. Temp. (°C) (gauss) (Oe) (°C) 1. (Ca ₃ )(Fe ₂ )(Fe ₁ .5 V ₁ .5)O ₁₂ 1210 520 370 210 2. (Ca ₃ )(Fe ₂ )(Fe ₁ .35 Ge ₀ .3 V ₁ .35)O ₁₂ 1210 785 110 197 3. (Ca ₃ )(Fe ₁ .9 Sn ₀ .1)(Fe ₁ .55 V ₁ .45)O ₁₂ 1210 400 198 200 4. (Ca ₃ )(Fe ₁ .7 Sn ₀ .3)(Fe ₁ .65 V ₁ .35)O ₁₂ 1210 276 190 135 5. (Ca ₃ )(Fe ₁ .9 Sn ₀ .1)(Fe ₁ .4 Ge ₀ .3 V ₁ .3)O ₁₂ 1210 550 90 150 6. (Ca ₃ )(Fe ₁ .7 Sn ₀ .3)(Fe ₁ .5 Ge ₀ .3 V ₁ .2)O ₁₂ 1210 390 35 120 7. (Ca ₁ .6 Y ₁ .4)(Fe ₁ .7 Sn ₀ .3)(Fe ₂ .35 V ₀ .65)O ₁₂ 1300 960 32 222 8. (Ca ₁ .6 Y ₁ .4)(Fe ₁ .7 Sn ₀ .3)(Fe ₂ .2 Ge ₀ .3 V ₀ .5)O ₁₂ 1300 530 26 200

TABLE 2

Curie Temperature x y z (°C) 0 1.4 0.3 240 0.1 1.4 0.3 220 0.3 1.4 0.3 200 0.5 1.4 0.3 152 0.8 1.4 0.3 110 1.0 1.4 0.3 70

TABLE 3

Sinter- Sample Composition ing 4πMs ΔH Curie No. Temp. Temp. (°C) (Gauss) (Oe) (°C) 1. (Ca ₃ )(Fe ₁ .5 Sn ₀ .5)(Fe ₁ .6 Ge ₀ .3 V ₁ .1)O ₁₂ 1210 <60 120 <50 2. (Ca ₂ .8 Y ₀ .2)(Fe ₁ .5 Sn ₀ .5)(Fe ₁ .7 Ge ₀ .3 V ₁ .0)O ₁₂ 1210 100 120 70 3.(Ca ₂ .2 Y ₀ .8)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .0 Ge ₀ .3 V ₀ .7)O ₁₂ 1250 320 35 100 4. (Ca ₂ .0 Y ₁ .0)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .1 Ge ₀ .3 V ₀ .6)O ₁₂ 1250 510 30 125 5. (Ca ₁ .6 Y ₁ .4)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .3 Ge ₀ .3 V ₀ .4)O ₁₂ 1300 730 20 160 6. (Ca ₁ .0 Y ₂ .0)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .6 Ge ₀ .3 V ₀ .1)O ₁₂ 1300 990 48 200 7. (Ca ₀ .8 Y ₂ .2)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .7 Ge ₀ .3)O ₁₂ 1300 1150 100 110 8. (Ca ₀ .6 Y ₂ .4)(Fe ₁ .7 Sn ₀ .3)(Fe ₂ .75 Ge ₀ .2 V ₀ .05)O ₁₂ 1300 1280 125 229

TABLE 4

x y z Curie Temperature (°C) 0.3 1.4 0 230 0.3 1.4 0.1 220 0.3 1.4 0.3 200 0.3 1.4 0.5 168 0.3 1.4 0.8 125

TABLE 5

Sinter- Sample ing 4πMs ΔH Curie No. x y z Temp. Temp. (°C) (Gauss) (Oe) (°C) 1. 0.3 1.2 0 1300 700 10.5 195 2. 0.3 1.6 0 1330 1030 10.0 206 3. 0.3 1.8 0 1330 1140 10.7 210 4. 0.5 1.6 0 1330 1120 9 160 5. 0.5 1.8 0 1350 1230 8 167 6. 0.5 1.6 0 1350 1120 2.5 160

TABLE 6

Sinter- Theore- Sample Composition ing tical ΔH No. Temp. Density (°C) (%) (Oe) 1. (Y ₃ )(Fe ₂ )(Fe ₃ )O ₁₂ 1300 91.4 170 2. (Ca ₁ .6 Y ₁ .4)(Fe ₁ .7 Sn ₀ .3)(Fe ₂ .2 Ge ₀ .3 V ₀ .5)O ₁₂ 1300 99.0 28 3. (Ca ₁ .6 Y ₁ .4)(Fe ₁ .7 Sn ₀ .3)(Fe ₂ .1 Ge ₀ .5 V ₀ .4)O ₁₂ 1300 99.1 24 4. (Ca ₁ .6 Y ₁ .4)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .3 Ge ₀ .3 V ₀ .4)O ₁₂ 1300 99.3 20 5. (Ca ₁ .0 Y ₂ .0)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .6 Ge ₀ .3 V ₀ .1)O ₁₂ 1300 97.9 48 6. (Ca ₂ .0 Y ₁ .0)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .1 Ge ₀ .3 V ₀ .6)O ₁₂ 1250 97.0 125 7. (Ca ₀ .8 Y ₂ .2)(Fe ₁ .7 Sn ₀ .3)(Fe ₂ .75 V ₀ .25)O ₁₂ 1300 98.9 100 8. (Ca ₁ .5 Y ₁ .5)(Fe ₁ .3 Sn ₀ .7)(Fe ₂ .6 V ₀ .4)O ₁₂ 1300 99.5 14 9. (Ca ₁ .8 Y ₁ .2)(Fe ₁ .5 Sn ₀ .5)(Fe ₂ .35 V ₀ .65)O ₁₂ 1300 99.0 30