SPIN WAVE RESONANCE ON PdFe ALLOYS

HAL Id: jpa-00214017

https://hal.archives-ouvertes.fr/jpa-00214017

Submitted on 1 Jan 1971

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

SPIN WAVE RESONANCE ON PdFe ALLOYS

D. Hardison, E. Thompson

To cite this version:

D. Hardison, E. Thompson. SPIN WAVE RESONANCE ON PdFe ALLOYS. Journal de Physique Colloques, 1971, 32 (C1), pp.C1-565-C1-566. �10.1051/jphyscol:19711193�. �jpa-00214017�

JOURNAL DE PHYSIQUE Colloque C 1, supplkment au no 2-3, Tome 32, Fkvrier-Mars 1971, page C 1 - 565

SPIN WAVE

RESONANCE

Electrical Sciences and Applied Physics Division and Center for the Study of Materials, Case Western Reserve University, Cleveland, Ohio

R6sum6. - On a etudi6 la rksonance d'ondes de spin, sur des films de PdFe, kvapores sous vide, sur des supports de verre pour des concentrations en fer variant de 0,5 a 26 at. %. Les mesures ont Bte faites a 33 GHz pour des temp&

ratures comprises entre 4,2 et 8,l OK. Les modes d'ordres kleves sont quadratiquement espacks sur les films a hautes concentrations en fer ; mais l'espacement devient presque linkaire aux faibles concentrations. Les valeurs de g dkpendent de cette concentration et presentent un minimum pour 18 at. % de fer environ.

Abstract. - Spin wave resonance has been performed on PdFe films vacuum evaporated onto glass substrates with the iron concentration ranging from 0.5 to 26 atomic per cent. Measurements were made at 33 GHz at temperatures of 4.2 to 8.1 OK. The high order modes are quadratically spaced for the high iron concentration films, but the spacing becomes nearly linear for the lower iron concentrations. The g-values are found to exhibit a concentration dependence and display a minimum at approximately 18 atomic per cent iron.

The incipient ferromagnetism of palladium metal and the consequent existence of a ferromagnetic state in the palladium-iron alloys of low iron concentration has created considerable interest in the magnetic properties of these alloys. Crangle and Scott [I, 21 have measured the composition dependence of the saturation magnetization and the Curie temperature and also found a giant moment of approximately 11 Bohr magnetons per iron atom for a 1 at. % iron alloy, with the giant moment decreasing with increasing iron concentration. Magnetic diffuse neutron scattering experiments by Cable, Wollan, and Koehler [3] have shown that the moment residing on the iron atom itself is approximately 3 Bohr magnetons independent of alloying and that the remainder of the giant moment resides on the palladium atoms. Further neutron scattering work by Low and Holden [4] and Hicks, et. al. [5] has revealed the range of the palladium polarization to be about 10 A for very dilute alloys.

A theoretical calculation by Doniach and Wohlfarth [6] of the dynamical magnetic properties of the palla- dium-iron alloys based on an s-d interaction model has predicted a magnon dispersion having both acoustic and optical branches and an exchange stiffness which increases linearly with increasing iron concen- tration. Neutron small angle scattering experiments by Stringfellow [7] have verified this compositional dependence for low iron concentrations but reveal an exchange stiffness four times larger than the theore- tical figure. Exchange stiffness measurements have also been conducted by Oder [&I using temperature modu- lation techniques. Fischer, Herr, and Meyer 193 have performed ferromagnetic resonance on the related nickel-palladium alloys ranging in composition from 0 to 90 at. % palladium. They have concluded that their g-value results can be represented by a tightly- coupled two sub-lattice model using constant g-values for nickel and palladium (gNi = 2.18 and g,, = 2.58).

We report here the results of ferromagnetic and spin wave resonance studies on palladium-iron alloy

(*) Research sponsored by the Air Force Office of Scientific Research, Office of Aerospace Research, United States Air Force, under AFOSR number 68-1484.

films with compositions ranging from 0.5 to 26 at. %

iron. The films, 1 000 to 4 700 A in thickness, were vacuum deposited onto clean glass substrates heated to approximately 100 ^{O C .} Alloy sources, prepared from 99.99 % pure iron and 99.99 % pure palladium, were heated in an alumina coated tungsten filament crucible.

Pressures during evaporation were on the order of torr. Film compositions were determined from the saturation magnetization data of Crangle [l] using magnetization values obtained from the ferromagnetic resonance data.

Resonance measurements in both the parallel and perpendicular field orientations were performed using a reflection cavity microwave spectrometer of standard design and field modulation techniques. The klystron source was frequency stabilized on the cavity at 33.4 Ghz. A liquid helium cryostat and temperature controller were used to maintain isothermal conditions during resonance runs. Resonance measurements were conducted at constant temperatures ranging from 4.2 t o 8.1 OK.

From the parallel and perpendicular resonance fields the effective g-values of the alloy films may be calcu- lated. The results of such a calculation, depicted as curve A in figure 1, display two striking features.

1.90

0 20 40 60 8 0 1 0 0

ATOMIC PERCENT IRON IN P A U A O I U M

FIG. 1. - Compositional dependence of g-values for palladium (B) and palladium iron alloys (A).

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19711193

C 1 - 566 D. L. HARDISON AND E. D. THOMPSON First, the effective g-value displays a minimum at

about 18 at. % iron. Second, the effective g-value varies rapidly with composition in the range less than 10 at. % iron and precludes an extrapolation to pure palladium. Curve A in figure 1 has been extrapolated over the entire composition range taking 2.1 as the g-value of pure iron.

We have analyzed the g-value data using as a model a two sub-lattice system having strong exchange coupling between the two sub-lattices. The assumption of strong coupling appears valid since the exchange field for a 0.1 at. % Co palladium cobalt alloy is estimated to be greater than lo5 Oe [lo]. Under this assumption a molecular field equation of motion calculation [ l l ] leads to an effective g-value, geff, for the alloy

In equation (1) palladium and iron form the two sub- lattices with their corresponding g-values and magne- tizations. Equation (1) corresponds to a weighting of each sub-lattice g-value according to the angular momentum density of that sub-lattice.

The fractional sub-lattice magnetization can be calculated using the data of Crangle [I] and Cable, et. al. [3]. Assuming the g-value for iron to be 2.1 and alloy independent, an assumption which appears consistent with the iron moment determinations of Cable, et. al. [3], we may use equation (1) in conjunc- tion with the experimental data to calculate the g- value for palladium. The results, shown as curve B in figure 1, indicate the palladium g-value to be alloy independent for iron concentrations greater than 25 at. %. Below 25 at. % iron the palladium g-value increases rapidly with decreasing iron content, This behavior correlates with the decreasing palladium atom moment for alloys with less than 25 at. % iron and the palladium moment saturation at 0.34 Bohr magne- tons for disordered alloys having greater than 25 at. %

iron [3]. If this moment arises from the splitting of the d-conduction bands of palladium, then we are led to the conclusion that the palladium g-value depends strongly on this band splitting, reaching the almost free electron value for widely split bands. This result contrasts sharply with the palladium-nickel alloy results of Fischer, et. al. [9].

Resonance measurements performed in the perpen- dicular field orientation display evidence of spin Refer

waves at all iron concentrations. The spin wave modes appear discrete, however, only for iron concentrations greater than about 12 at. %. Figure 2 depicts typical

13 15 17 19 2L

If, kilo-oersteds

FIG. 2. - Typical perpendicular absorption traces for palladium iron alloy films. Vertical scale is in arbitrary units.

perpendicular field absorption traces for films of 24.1 and 12.0 at. % iron. Below 6 at. % iron an asym- metric bump which we attribute to spin waves appears on the low field side of the uniform resonance. This bump becomes less apparent with decreasing iron concentration. Calculation of spin wave mode sepa- rations based on the film thickness used and the extrapolated exchange stiffness data of Stringfellow [7]

reveals that discrete modes should have been resolved below 6 at. % iron. This suggests the possibility of spin pinning difficulties in the low iron concentration alloys. The spacing of the modes in the higher iron concentration films yields a quadratic magnon disper- sion, but the mode spacing bscomes almost linear as the iron concentration is decreased. This has precluded an evaluation of the exchange stiffness and a compa- rison with Stringfellow's small angle neutron scattering data [7].

We have also found that the uniforms resonance linewidth for both parallel and perpendicular field orientations increases with decreasing iron concentra- tion, the rate of increase being larger at the lower concentrations. For instance, the perpendicular uni- form resonance linewidth is 440 Oe for a 12 at. %

iron film and 1 500 Oe for a 1.0 at. % iron film. The lineshapes are non-Lorentzian for the lower iron concentrations.

[I] CRANGLE (J.), Phil. Mag,, 1960, 5 , 335. [6] DOMIA'CH (S.) and WOHLFARTH (E.), Proc. Roy. Soc.

[2] CRANGLE (J.) and SCOTT (W.), J. Appl. Phys., 1965, 4, 1967, 296, 442.

36, 921. 173 STRINGFELLOW (M.), J. Phys. C (Puoc. Phys. Soc.), 1968, 1, 1699.

P I ^{CABLE (J.), WOLLAN} (E.) and KOEHLER (W.), Plays,

181 (R.), j. Appl. phys., 1969, 40, 1204.

Rev., 1965, 138, A 755. [9] FISCHER (G.), HERR (A.) and MEYER (A.), J. Appl.

[4] LOW (G.) and HOLDEN (T.), Proc. Phys. SOC., 1966, _Phys., 1968, 39, 545.

89, 119. [lo] HORNFELDT (S.), KETTERSON (J.) and WINDMILLER (L.), [5] HICKS (T.), HOLDEN (T.) and Low (G.), J. Phys. C Phys. Rev. Lett., 1969, 23, 1292.

(Proc. Phys. Soc.), 1968, 1 , 528. [ll] WANGSNESS (R.), Phys. Rev., 1953, 9 1 , 1085.