Technical Physics, Vol. 50, No. 12, 2005, pp. 1605–1610. Translated from Zhurnal Tekhnichesko

œ

Fiziki, Vol. 75, No. 12, 2005, pp. 69–75.Original Russian Text Copyright © 2005 by Frolov, Yakovchuk, Seredkin, Iskhakov, Stolyar, Polyakov.

SOLIDS

Unidirectional Anisotropy in Ferromagnetic–Ferrimagnetic Film Structures

G. I. Frolov1, V. Yu. Yakovchuk1, V. A. Seredkin1, R. S. Iskhakov1, S. V. Stolyar1,2, and V. V. Polyakov1

1Kirenskiœ Institute of Physics, Siberian Division, Russian Academy of Sciences, Akademgorodok, Krasnoyarsk, 660036 Russia

e-mail: sva@iph.krasn.ru2Krasnoyarsk State University, Krasnoyarsk, 660041 Russia

Received April 5, 2005

Abstract—A mechanism of unidirectional anisotropy formation in an exchange-coupled ferromagnetic–ferri-magnetic film structure with orthogonal effective magnetizations in the layers is investigated. The reason forunidirectional anisotropy is the magnetic heterogeneity of the ferrimagnetic layer in the compensation range.Magnetization reversal in the magnetically soft layer of an (REE–transition metal)/NiFe film structure is dis-cussed based on a model of uniform rotation of magnetization. It is found that unidirectional anisotropy sharplydecreases the magnetic noise level in the magnetically soft layer. The field of application of these materials isoutlined. © 2005 Pleiades Publishing, Inc.

Exchange interaction between magnetically soft andmagnetically hard layers imparts intriguing propertiesto related structures, which are of great fundamentaland applied interest. One of these properties is unidi-rectional exchange anisotropy in the magnetically softlayer, which shifts hysteresis loop ∆H along the mag-netic field axis. Although this effect was discovered50 years ago [1], it has not been yet completely under-stood. Therefore, to gain a deeper insight into the natureof unidirectional anisotropy remains a topical problem.Hot interest in these materials as candidates, e.g., formagnetic memory devices [2], spintronics devices [3],and magnetic sensors [4] is giving an additional impe-tus to research in this field.

Unidirectional exchange anisotropy has been stud-ied mostly in ferromagnetic–antiferromagnetic(FM/AFM) film structures [5]. However, a number ofdisadvantages limit the application of these materials[6]. These are a poor temperature stability of ∆H, anincreased coercive force of the magnetically soft layercompared with that of a one-layer FM film, and the evo-lution of the hysteresis loop with the number of a mag-netization reversal cycle.

At the beginning of the 1980s, unidirectional anisot-ropy was discovered [7] in a TbFe/NiFe ferromagnetic–ferrimagnetic (FoM/FiM) film structure. These struc-tures attracted attention, since they, on the one hand,lack the disadvantages typical of their FM/AFM coun-terparts and, on the other hand, unidirectional anisot-ropy here appears in layers with orthogonal effectivemagnetizations. A large body of data concerning inves-tigation and application of FoM/FiM structures hasbeen gained in recent years [8–15]. In our opinion, it is

1063-7842/05/5012- $26.00 1605

an appropriate time to analyze the state of the art in thisfield. In this work, we consider the nature of unidirec-tional anisotropy, its effect on the magnetic perfor-mance of the FM layer, and applications of FoM/FiMstrictures.

MECHANISMS OF UNIDIRECTIONAL ANISOTROPY FORMATION

The phenomenological description of the effect ofunidirectional exchange anisotropy is straightforward.It is based on the assumption that the magneticmoments at the interface are collinear,

(1)

In [7], where the existence of unidirectional anisot-ropy in a TbxFe1 – x/NiFe exchange-coupled film struc-ture with orthogonal effective magnetizations of thelayers was reported for the first time, we applied someconsiderations to the origin of this effect. It was conjec-tured that an amorphous ferrimagnetic layer may havethe in-plane magnetization component resulting, forexample, from a chemical inhomogeneity across theamorphous alloy film. This inhomogeneity, in turn, pro-duces a compensation plane, in which the magnetiza-tion reverses.

Further investigations into the chemical composi-tion of REE–transition metal (REE–TM) amorphousalloy films obtained by thermal evaporation haveshown, however, that a concentration gradient normalto the plane is absent in such films. Therefore, othermechanisms come to the fore. For example, the in-plane magnetization component in amorphous REE–

jM1 M2⋅ JM1M2 M1^M2( ).cos=

© 2005 Pleiades Publishing, Inc.

1606

FROLOV

et al

.

TM films, which are characterized by perpendicularanisotropy, may be due to a local dispersion of the mag-netic anisotropy axis. As a result, the general directionof the easy magnetic axis deviated from the normal tothe film surface. If such a mechanism works and thecomposition of the REE–TM film is off-compensation,the amount of unidirectional anisotropy in exchange-coupled (REE–TM)/NiFe film structures is bound togrow, being minimal in those where the easy magneticaxis of the amorphous REE–TM layer lies in the planeof the film. However, our investigations [14], as well asthose carried out by other authors [9, 11], show that thereverse is true. In (REE–TM)/NiFe structures, unidirec-tional anisotropy exists only in that concentration rangeof the amorphous alloy where perpendicular anisotropyin the REE–TM layer is observed.

The existence of unidirectional anisotropy inTbxFe1 – x/NiFe and DyxCo1 – x/NiFe film structures,which was first discovered in [7], indicates a specificmagnetic microstructure of REE–TM layers. Namely,DyxCo1 – x and TbxFe1 – x alloys show nanoareas wherethe magnetization vector of the 3d metal sublattice hasthe in-plane component, and it is this component thattakes part in exchange interaction with the FM layer ofNiFe alloy.

To establish factors causing the in-plane componentof the 3d metal sublattice in REE–TM alloy with layers

–2015.0 17.5

∆H, Oe

Dy concentration, at.%

–10

20.0 22.5 25.0 27.5

0

10

20

∆H H

Hc

xcompM/M0

M/M0

H

0

0

NiFeDyCo

A

î1

î2 î1

î2I

II II

I

x < 22 at. % x > 22 at. %

(‡)

(b)

Fig. 1. (a) Hysteresis loop and concentration dependence ofbias field ∆H(x) and (b) the orientation scheme for the mag-netizations of 3d metals in exchange-coupled DyCo/NiFefilm structures. A, substrate; I, matrix; and II, impurity.

that have near-compensation compositions and featurea high degree of integral perpendicular anisotropy is achallenge. This is because the amorphous ferrimagneticREE–TM alloy has a low saturation magnetization(Ms 0) and a high magnetic hardness (for example,the coercive field in the (REE–TM)/DyCo alloy may beas high as > 10 kOe). In light of this, we suggested thatthe physical properties of such magnetically hard mate-rials be studied on NiFe/DyxCo1 – x/NiFe multilayerexchange-coupled film structures where the DyxCo1 – xlayer is much thinner than NiFe, dDyCo ! dNiFe.Exchange interaction between the Co sublattice of theferrimagnetic alloy and the NiFe layer substantiallymodifies the magnetic performance of such a well-stud-ied alloy as NiFe.

Based on the measurements of the dynamic andstatic magnetic characteristics of NiFe/DyxCo1 – x/NiFecomposites, we developed a microheterogeneousmodel of the DyCo amorphous layer [15, 16]. TheDyCo specific microstructure in the compensationrange that follows from this model has allowed us toexplain experimental data for FM resonance (FMR)and spin-wave resonance (SWR). To find specific fea-tures of DyCo, we (i) prepared NiFe/DyCo/NiFe three-layer structures with unidirectional exchange anisot-ropy and orthogonal effective magnetizations of thelayers, (ii) studied the FMR and SWR spectra of thesestructures, and (iii) found that the spin system of theamorphous DyCo alloy in the concentration range ofmagnetic compensation can be represented in the formof two subsystems with the TM magnetization prevail-ing in one of them (nanophase Φ1) and the REE magne-tization in the other (nanophase Φ2). This modelembodies the basic structural feature of amorphousalloys: natural fluctuation (topological and composi-tion) inhomogeneity. In the concentration ranges xi ±∆x ! xcomp (matrix Φ1) and xi ± ∆x @ xcomp (matrix Φ2),the magnetic microstructures of amorphous ferrimag-nets will differ substantially from the magnetic micro-structure in the range xi – ∆x < xcomp < x + ∆x. In thiscase, magnetic compensation point xcomp is defined by

the condition ⟨M⟩ = p + q = 0, where p andq are the volume fractions of nanophases Φ1 and Φ2 and

and are the effective magnetizations ofthese phases at xi – ∆x and xi + ∆x, respectively.

Figure 1 shows the experimental dependences ofshift ∆H(x) of the hysteresis loop on the REE concen-tration in a planar DyCo/NiFe structure and the distri-butions of nanophases Φ1 and Φ2 in the DyCo layer. Itis seen that the curve ∆H(x) for the DyxCo1 – x/NiFestructure is described by the asymmetric function∆H(x – xcomp) and has singular points: the coordinate ofthe zero (minimal value of ∆H) and extreme points.Also, it is seen that ∆H(x) in this planar structurereaches a maximum at x ≈ 19 at.% for undercompensa-

MeffΦ1( )

MeffΦ2( )

MeffΦ1( )

MeffΦ2( )

TECHNICAL PHYSICS Vol. 50 No. 12 2005

UNIDIRECTIONAL ANISOTROPY 1607

tion compositions of DyCo and x ≈ 24 at.% for over-compensation compositions.

The above results are readily explained in terms ofthe suggested model of amorphous DyCo alloy struc-ture in the magnetic compensation range. Indeed, in theREE–TM concentration range x ≤ 16 at.% (x ≥27 at.%), the magnetic structure of amorphous DyCo iscompletely specified by magnetic nanophase Φ1 (Φ2).Hence, the magnetic moments of the Co and Dy sublat-tices are collinear with the perpendicular anisotropyaxis in the DyCo layer and, as a consequence, the effec-tive magnetizations of the DyCo and NiFe layers aremutually orthogonal (exchange coupling is absent).Such a situation takes place in the concentration rangexi – ∆x < xcomp < x + ∆x. Here, the magnetic structure ofDyCo is formed by randomly mixed nanophases Φ1 andΦ2. If phase Φ1 is a matrix, phase Φ2 is an impurity andvice versa. The only exception is compensation pointxcomp, where the volume fractions of the phases areroughly the same.

At any concentration xi from the above range, theeffective magnetization of matrix phase Φi in the DyColayer is aligned with the perpendicular anisotropy field(MCo and MDy are collinear with this field (as indicatedby the polar Kerr effect). In this case, the magnetizationof the Co sublattice, MCo, in impurity nanophase Φj

must have the in-plane component because of exchangeinteraction between the transition elements in the impu-rity and matrix and the effective magnetization ofnanophase Φj has a chance to align with an externalmagnetic field. We believe that exchange interactionbetween magnetization MCo of impurity phase Φj in theDyCo layer and the magnetization of the NiFe layerresults in exchange unidirectional anisotropy in NiFe.

The form of the experimental dependence in Fig. 1(asymmetry about xcomp) can also be treated in terms ofour model. In this planar system, a DyxCo1 – x layerforms where the magnetization of the matrix phase isaligned with the perpendicular anisotropy axis, whilethe effective magnetization of the impurity phase lies inthe plane of the payer. When a NiFe overlayer is grown,a permanent magnetic field is switched on to specify theeasy direction and the effective magnetization of theimpurity phase in the DyxCo1 – x layer is aligned withthis field and, hence, with the unidirectional anisotropyaxis in NiFe. In the range x < xcomp, the impurity phasemeets the inequality MCo < MDy, while at x > xcomp, theinverse inequality is valid. This means that, at x < xcomp,the magnetization vectors of the Co sublattice and NiFelayer are anticollinear, whereas, at x > xcomp, they arecodirected (see Fig. 1b). That is why the sign of ∆Hchanges in going through concentration xcomp (Fig. 1a).

TECHNICAL PHYSICS Vol. 50 No. 12 2005

QUASI-STATIC MAGNETIZATION REVERSAL IN FERROMAGNETIC–FERRIMAGNETIC FILM

STRUCTURES

In exchange-coupled structures, the state of onelayer may substantially affect the state of another. Forexample, direct exchange interaction between ferro-magnetic layers (positive coupling) with differentdegrees of anisotropy may lead to the situation whereboth layers experiencing magnetization reversal willbehave as a whole with a coercive force intermediatebetween the coercive forces of the layers [6]. InFM/AFM film structures, unidirectional anisotropyshifts the hysteresis loop, results in only one easy axis,stabilizes the domain structure of the ferromagneticlayer, etc. [17, 18].

In FoM/FiM structures, the magnetization reversalprocess somewhat differs from that in FM/AFM films.Figure 2a shows the angular dependence of the hyster-esis loop for a DyCo/NiFe film (the layer thicknessesare, respectively, 70 and 210 nm; the easy axis coin-cides with the unidirectional anisotropy axis). In theeasy axis direction (α = 0), the hysteresis loop is thewidest (Hc = 2 Oe) and is shifted along the field axis by∆H = 10.5 Oe. As angle α increases, the loop gets thin-ner and collapses at α = 25° (in the range 25°–90°, theprocess of magnetization reversal is hysteresis-free).The shift also decreases, and, at α = 90°, the “loop”becomes symmetric (the anisotropy field determinedfrom this curve is Ha = 15 Oe).

Figure 2b demonstrates the variation of the hystere-sis loop for the same film when permanent magneticfield H⊥ is applied along the hard axis. At H⊥ = 4 Oe,

α = 0°

α = 15°

α = 25°

α = 45°

α = 90°

α = 0°

H~ = 23 Oe H~ = 23 Oe

H⊥ = 20 Oe

H⊥ = 4 Oe

H⊥ = 15 Oe

H⊥ = 0

(‡) (b)

Fig. 2. (a) Angular dependence of ∆H and (b) dependenceof the hysteresis loop on H⊥ in the exchange-coupledDyCo/NiFe film structure.

1608 FROLOV et al.

the loop collapses and takes the form of the magnetiza-tion reversal curve for α = 25°. At H⊥ = Ha = 15 Oe, theloop is still asymmetric.

Such an observation (collapsed hysteresis loop) dif-fers from experimental data obtained on NiFe/FeMnfilms [17]. Magnetization reversal curves taken of two-layer exchange-coupled structures are usually treated interms of the model assuming that the magnetizationvector of one layer is fixed, while the magnetizationvector of the other spirals under the action of an exter-nal magnetic field [19, 20]. The analytical solution pre-sented in [20] gives the following results. If magnetiza-tion reversal takes place along the easy axis (α = 0) infilms with dFM < dcr, the hysteresis loop collapses andshifts along the field axis. In films with dFM > dcr, theloop “opens up” and expands with dFM. (Here, dcr =

2A/MsHa > 12/π2, where A is the exchange interac-tion constant and Ms, Ha, and dFM are, respectively, thesaturation magnetization, anisotropy field, and thick-ness of the magnetically soft layer.)

However, in none of the numerous experimentswhere magnetization reversal took place along the easyaxis in FM/AFM and FoM/FiM structures did the hys-teresis loop collapse even for ∆H @ HC. The reason forthe discrepancy between the analytical and experimen-tal data seems to be an inadequate estimate of dcr. Foran exchange-coupled structure with parameters Ms =800 G and Ha = 3 Oe for the magnetically soft materialand exchange constant A = 10–6 erg/cm3, the criticalthickness was estimated as dcr = 260 nm [20]. At thesame time, experiments with NiFe/FeMn films showedthat the exchange constant is two orders of magnitudesmaller [21]; that is, dcr = 3 nm. However, the hysteresisloop did not collapse even when magnetization reversal

dFM2

4

0

BN, nT/Hz1/2

∆H; H0, Oe2 4 6

8

20

60

100

1

2

Fig. 3. Magnetic noise intensity (1) in the exchange-cou-pled DyCo/NiFe film structure vs. the amount of unidirec-tional anisotropy ∆H and (2) in the reference NiFe film vs.applied permanent magnetic field H0.

along the easy axis took place in structures with dFM(dNiFe) < 3 nm [21].

Another model of quasi-static magnetization rever-sal in exchange-coupled structures that assumes coher-ent rotation of the magnetization in a magnetically softlayer was suggested in [17]. Here, the interactionbetween the layers is described in terms of specific sur-face energy Es. In this case, we deal with thickness-averaged unidirectional anisotropy, the amount ofwhich is inversely proportional to the thickness of themagnetically soft layer. Such a dependence of ∆H onthe thickness of the magnetically soft layer is confirmedexperimentally [9, 11]. When external magnetic field His applied to the plane of the film at angle α to the easyaxis, magnetization M of the magnetically soft layerrotates through angle β relative to the easy axis. Then,the energy of the FM layer can be written in the form

(2)

where ϕ = α – β and Ea is the unidirectional anisotropyconstant.

Using the standard computational method, wederived an expression for the hysteresis loop for any α(except α = 0 and π). It was found that the loop col-lapses at certain α > αcr, αcr depending on ratio ∆H/Ha,where ∆H = Es/MdFM. This model leads us to concludethat, first, magnetization reversal at an angle to the easyaxis of an FM layer may proceed by rotation of vectorM and, second, unidirectional anisotropy can be simu-lated by permanent magnetic field H = ∆H. Both con-clusions were corroborated in experiments, whichmeans that the model adequately describes quasi-staticmagnetization reversal in exchange-coupled structures.

MAGNETIC NOISE IN FERROMAGNETIC–FERRIMAGNETIC FILM STRUCTURES

It is known that the magnetic noise in thin-film mag-netic devices can be depressed, e.g., by applying a mag-netic field, which makes the process of magnetizationreversal more uniform. The same results can beobtained using structures with unidirectional anisot-ropy.

We performed special experiments to see how theexternal magnetic field and unidirectional anisotropyinfluence the magnetic noise in Permalloy films[22, 23]. The object of interest was fluctuations of thetransverse emf in the presence of a constant bias (H0,∆H) applied along the easy magnetic axis or a high-fre-quency field (Hhf < Ha, H0) applied along the hard mag-netic axis.

Figure 3 shows the variation of magnetic noise BN inthe NiFe film incorporated into an exchange-coupledDyCo/NiFe structure 500 nm thick with bias field ∆H(curve 1) and in a reference NiFe film magnetized bymagnetic field H0 (curve 2).

E HM ϕcos– Ek βsin2 Es

dFM-------- β,cos–+=

TECHNICAL PHYSICS Vol. 50 No. 12 2005

UNIDIRECTIONAL ANISOTROPY 1609

As follows from Fig. 3, the magnetic noise in theexchange-coupled film structure is much lower than inthe reference film. Presumably, unidirectional anisot-ropy is not totally equivalent (is superior) to the appliedfield in efficiency and exchange interaction in the struc-ture makes the process of magnetization reversal in themagnetic layer more uniform.

APPLICATION OF EXCHANGE-COUPLED STRUCTURES

Film magnetic materials have found wide applica-tion in various fields of technology [2–4]. Here, we willconcentrate on using (REE–TM)/NiFe FoM/FiM struc-tures with unidirectional anisotropy as magnetoopticmemory devices and weak-field magnetic sensors.

Memory devices. It is believed that, in this field, thefilm structures under consideration may decrease thepower consumption and raise the speed of data writingand erasing.

Magnetooptic data writing on films with a magneti-zation normal to the film surface is in common use indisk storages of PCs. Data writing and erasing isaccomplished by the thermomagnetic method in writ-ing magnetic field H3 = 400–500 Oe applied normallyto the film (Fig. 4).

As materials for magnetooptic data carriers, REE–TM alloys are usually employed. Their basic disadvan-tages are (i) the need to apply high writing/erasing mag-netic fields, which influence the electrodynamic sus-pension of the focusing lens holder and (ii) a large timedelay between writing and erasing, which depends onthe inductance of the magnetic field source winding[24].

We suggest another approach [16] to writing magne-tooptic information on REE–TM/NiFe FoM/FiM struc-tures with exchange anisotropy (see Fig. 5).Writingmagnetic field Hwr > HC + ∆H (10–15 Oe) is applied toa FoM/FiM structure (Fig. 5a) antiparallel to the initialmagnetization in layer 3 (Fig. 5b). The magnetizationof layer 3 reverses, while that of layer 2 does not. Afterlayer 2 has been locally heated by a thermal pulse to atemperature close to Curie temperature TC, the heatedarea turns into the paramagnetic state (Fig. 5c) and themagnetic state of layer 3 remains unchanged, since TCof NiFe is much higher than TC of DyCo. Due toexchange interaction between layers 2 and 3 (duringcooling), the magnetization of the heated area of layer2 reverses in accordance with the magnetization direc-tion in the NiFe layer (Fig. 5d). After the thermal pulseis terminated and the writing field is switched off, theinverted state of the magnetic moment of the heatedarea in layer 2 persists (which corresponds to writing abit of information), while the magnetization of the localarea in layer 3 takes on a nonequilibrium (helical) struc-ture (Fig. 5d), representing a compressed “spin spring.”Data readout is accomplished using the polar Kerreffect.

TECHNICAL PHYSICS Vol. 50 No. 12 2005

To erase information, it suffices to heat the samelocal area of layer 2 to the Curie temperature. Exchangeinteraction between this area and layer 3 then disap-pears, and the magnetization of the local area in layer 3becomes aligned with the magnetization of the entirelayer (“the helix untwists”). After the pulse is termi-nated and the local area of layer 2 cools, the magnetiza-tion of this area changes sense because of exchangeinteraction with layer 3 and the information is erased.

Thus, in our approach, writing magnetic field Hwr ≥HC + ∆H is much lower than the writing field in the caseof standard REE–TM films. Moreover, data erasingdoes not require a magnetic field at all. Accordingly, themagnetic field energy spent on data writing/erasingsubstantially decreases. At the same time, a low writingmagnetic field, which makes it possible to generate H3

pulses as short as several nanoseconds, and no need forswitching the magnetic field at writing and/or erasingprovide a higher speed of the processes. It should beemphasized that all the related advantages of REE–TMlayers are retained.

Hwr < Hc(T < TC) Hwr < Hc(T = TC) Hwr = 0(T < TC)Hwr Hwr

B M CREE–TM

A

Fig. 4. Thermomagnetic data writing on REE–TM filmswith perpendicular magnetic anisotropy. A, substrate; B,radiation; and C, domain.

Hwr = 0

M

Hwr >_ Hc + ∆H

Hwr >_ Hc + ∆H Hwr = 0

B C

DyCo

NiFe

A

(‡) (b)

(d)(c)

Fig. 5. Thermomagnetic data writing on exchange-coupled(REE–TM)/NiFe film structures. A–C, the same as in Fig. 4.

1

1

2

2

3

3

1610 FROLOV et al.

Film magnetometers. FoM/FiM structures withunidirectional anisotropy were used to design a weak-field magnetic detector and study the dependence of themagnetic noise of this device on the static characteris-tics of the film magnetic structure and on excitationmodes [22, 23]. The operating principle of this magne-tometer is akin to that of a bubble magnetometer [25].However, an exchange-coupled film structure used as asensitive element made it possible to considerably sup-press the low-frequency noise, which was found to be≈2 × 10–11 T/Hz1/2 at a frequency of 1 Hz. The devicewas put to field tests, which shows that it can be appliedin geophysics, specifically, in geoelectrical prospectingof shallow ore-bearing rocks and in shallow oil-and-gasprospecting [12].

CONCLUSIONSWe touched upon three issues concerning exchange-

coupled FoM/FiM film structures consisting of a mag-netically hard material (amorphous films of REE–TMalloys) and a magnetically soft Permalloy film. Theseare the mechanism of formation of exchange couplingbetween the layers, the magnetic properties of thestructures, and application of the structures. In spite ofextensive investigation, the problems considered herestill need greater insight. However, wide application ofthese composites stimulates further effort in this field.

ACKNOWLEDGMENTSThis work was supported by the Russian Foundation

for Basic Research, grant no. 04-02-16099a.

REFERENCES1. W. H. Meiklejohn and C. P. Bean, Phys. Rev. 102, 1413

(1956).2. C. Tsang and R. E. Fantana, IEEE Trans. Magn. 18, 1149

(1982).3. A. V. Vedyaev, Usp. Fiz. Nauk 72, 1458 (1960) [Sov.

Phys. Usp. 3 (1960)].4. K. M. H. Lenssen, A. F. M. Vierman, and J. J. T. Douk-

ers, J. Appl. Phys. 81, 4915 (1997).5. J. Nogues and I. K. Shuller, J. Magn. Magn. Mater. 192,

203 (1999).6. A. Yelon, in Physics of Thin Films, Ed. by H. Francombe

and R. W. Hoffman (Academic, New York, 1971; Mir,Moscow, 1973), pp. 228–333.

7. V. A. Seredkin, G. I. Frolov, and V. Yu. Yakovchuk,Pis’ma Zh. Tekh. Fiz. 9, 1446 (1983) [Sov. Tech. Phys.Lett. 9, 621 (1983)].

8. W. C. Cain, W. H. Meiklejohn, and M. H. Kryder,J. Appl. Phys. 61, 4170 (1987).

9. V. A. Ceredkin, G. I. Frolov, and V. Yu. Yakovchuk, Fiz.Met. Metalloved. 63, 457 (1987).

10. F. Hellman, R. B. Dover, and E. M. Gyorgy, Appl. Phys.Lett. 50, 296 (1987).

11. W. C. Cain and M. H. Kryder, J. Appl. Phys. 67, 5722(1990).

12. A. N. Babitskiœ, E. P. Blinnikov, A. G. Vladimirov, et al.,Geofiz. Apparatura, No. 94, 21 (1991).

13. V. A. Ceredkin, R. S. Iskhakov, V. Yu. Yakovchuk, et al.,Fiz. Tverd. Tela (St. Petersburg) 45, 883 (2003) [Phys.Solid State 45, 927 (2003)].

14. R. S. Iskhakov, V. A. Seredkin, S. V. Stolyar, et al.,Pis’ma Zh. Éksp. Teor. Fiz. 80, 743 (2004) [JETP Lett.80, 638 (2004)].

15. R. S. Iskhakov, V. A. Seredkin, S. V. Stolyar, et al.,Pis’ma Zh. Éksp. Teor. Fiz. 76, 779 (2002) [JETP Lett.76, 656 (2002)].

16. V. A. Seredkin, S. V. Stolyar, G. I. Frolov, and V. Yu. Yak-ovchuk, Pis’ma Zh. Tekh. Fiz. 30 (19), 46 (2004) [Tech.Phys. Lett. 30, 820 (2004)].

17. A. A. Glazer, A. P. Potapov, R. I. Tagirov, and Ya. S. Shur,Fiz. Tverd. Tela (Leningrad) 8, 3022 (1966) [Sov. Phys.Solid State 8, 2413 (1966)].

18. M. Sh. Erukhimov and V. A. Seredkin, Fiz. Met. Metall-oved. 44, 757 (1977).

19. A. Ahoroni, E. H. Frei, and H. Strikman, J. Appl. Phys.30, 1956 (1959).

20. Yu. G. Canoyan and K. A. Egiyan, Fiz. Met. Metalloved.38, 231 (1974).

21. R. Jungblut, R. Coehoorn, M. T. Johnson, et al., J. Appl.Phys. 75, 6659 (1994).

22. V. V. Polyakov and G. I. Frolov, Magnetic Materials forRadio Electronics (Krasnoyarsk, 1982), pp. 179–185 [inRussian].

23. V. V. Polyakov, A. G. Vladimirov, and V. A. Seredkin,Amorphous Film Alloys of Transition and Rare-EarthMetals (Krasnoyarsk, 1988), pp. 219–223 [in Russian].

24. US Patent No. 4610009 (February 4, 1986).25. G. Chiron and G. Delapierre, IEEE Trans. Magn. 15,

1815 (1979).

Translated by V. Isaakyan

TECHNICAL PHYSICS Vol. 50 No. 12 2005