ANOMALOUS LINE SHAPES OF Fe3+ -MÖSSBAUER SPECTRA IN MAGNETICALLY ORDERED SYSTEMS — EFFECTS OF HEAT TREATMENT IMPURITY IONS AND Fe3+ ION-CONCENTRATION

HAL Id: jpa-00215761

https://hal.archives-ouvertes.fr/jpa-00215761

Submitted on 1 Jan 1974

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.

ANOMALOUS LINE SHAPES OF Fe3+

-MÖSSBAUER SPECTRA IN MAGNETICALLY ORDERED SYSTEMS - EFFECTS OF HEAT

TREATMENT IMPURITY IONS AND Fe3+

ION-CONCENTRATION

J. Srivastava, R. Sharma

To cite this version:

J. Srivastava, R. Sharma. ANOMALOUS LINE SHAPES OF Fe3+ -MÖSSBAUER SPECTRA IN

MAGNETICALLY ORDERED SYSTEMS - EFFECTS OF HEAT TREATMENT IMPURITY IONS

AND Fe3+ ION-CONCENTRATION. Journal de Physique Colloques, 1974, 35 (C6), pp.C6-663-C6-

668. �10.1051/jphyscol:19746146�. �jpa-00215761�

ANOMALOUS LINE SHAPES OF Fe3

-MOSSBAUER SPECTRA

IN MAGNETICALLY ORDERED SYSTEMS - EFFECTS OF HEAT TREATMENT IMPURITY IONS AND Fe3

ION-CONCENTRATION

J. K. SRIVASTAVA and R. P. SHARMA

Tata Institute of Fundamental Research, Bombay-400005, India

RBsumB. - Des mesures par effet Mossbauer ont montre qu'un recuit incomplet des solutions solides CrzO 3-Fez0 3 et A1 2 0 3-Fez0 ³ donne lieu a des spectres Mossbauer anormaux qui indiquent la presence soit de superparamagnetisme soit d'effets de relaxation. Des etudes detaillees des solu- tions solides Cr~03-Fez03 montrent que l'augmentation d'intensite des parties centrales des formes anormales de raie diminue si la temperature de recuit ou la durke du recuit augmente ou si l'kchan- tillon est trempe a temperature ambiante ou encore si la solution solide est mise sous pression avant recuit. D'autre part une augmentation de la concentration d'ions Fe3+ ou la presence d'impuretes ioniques, Al3+, Gd3+ et Ni2+, augmente l'excks central anormal. Des recuits sous vide de solutions solides de Cr203-Fe203 pures ou dopees par Fe2+ donnent lieu une espkce entierement nouvelle de spectres Mossbauer comportant un doublet quadrupolaire de Fez+ a 80 K et une large raie unique a 293 K.

Abstract. - Mossbauer measurements have shown that an incomplete annealing of the Cr203- Fez03 and A1203-Fez03 solid solutions gives rise to the anomalous Mossbauer spectra indicating the presence of either the superparamagnetic or the electronic relaxation effects in them. Detailed investigations made on the Crz03-Fez03 solid solutions show that the enhancement in the intensity of the central portions of the anomalous line shapes decreases when the annealing temperature or annealing time is increased or the sample is quenched to room temperature or the solid solution is subjected to pressure before annealing. On the other hand an increase in the Fe3+ ion-concentration or the presence of impurity ions, A13+, Gd3+ and Niz+, is seen to increase the anomalous central enhancement. Vacuum-annealing of the pure and Fez+-doped CrnO 3-Fez0 3 solid solutions gives rise to entirely new kind of Mossbauer spectra showing a Fez+-quadrupole doublet at 80 K and a broad single line at 293 K.

The problem of anomalous Mossbauer line shapes have been studied earlier both experimentally [I-31 and theoretically [4-61. In the present work study has been made regarding the line shapes of Fe3+-Moss- bauer spectra in magnetically ordered Cr203-Fe203 and A1203-Fe,03 systems. I t is seen that the appea- rance of anomalous line shapes in these solid solutions is a consequence of their incomplete annealing. This incomplete annealing arises when either the annealing temperature, Tan,, is low or the annealing time, tan,, is short. Figure 1 shows a few representative Mossbauer spectra of the (1-x)Cr203 - x Fe203 system (x = 3.85. wt %) annealed at 1250OC in air for a period of 2 hours before being slow-cooled to room temperature. These spectra clearly show the presence of anomalous line shapes which are characterised by the enhancement in the intensity of their central portions. Annealing the Cr20T3.85 wt. % Fe203 sys- tem at 1400 OC in air for a period of 10 hours has been found to make this anomalous central enhancement disappear. This has been seen to be true whether the sample is quenched or slow-cooled from 14000C.

Figure 2 shows Mossbauer spectra of the sample quenched from 1400 OC. We see only normal Moss- bauer spectra with no indication of any central enhan- cement. Similar results have been obtained in the Al2O3-29 mol. % Fe203 system also which when annealed at 1350 OC for 10 hours in air, before being slow-cooled, has been found to give rise to the ano- malous Mossbauer line shapes. Increasing the annea- ling temperature to 1500 OC has been found to make the anomalous central enhancement disappear. This shows that this phenomenon of heat treatment-depen- dence of Mossbauer line shapes is general in nature and is not specific to any particular solid solution. In order to have a better understanding of this pheno- menon, extensive Mossbauer measurements have been made on the (1 - x)Cr203 - x Fe203 system

(x < 7.41 wt. %) at various temperatures between

80 and 300 K. About two hundred and fifty Mossbauer spectra of the Cr203-Fe203 solid solutions have been recorded after subjecting them to various heat treat- ments, after varying a-Fe203 concentration in them, after subjecting them to pressure before annealing and

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19746146

C6-664 J. K. SRIVASTAVA AND R. P. SHARMA

I I I I I I I ~ I , ~ , ~ , l .

60 80 100 120 140 160 180 200

C H A N N E L NUMBER-+

FIG. 1. - Mossbauer spectra of theCr~0~-3.85 wt. % Fez03 system annealed at 1 250 OC for 2 hours in air before being slow-cooled to room temperature (category C). The arrow indicates the zero velocity position and the velocity calibration is : 1 channel = 0.11 2 18 mmls. The gamma ray source was 5 7Co diffused in a copper matrix and was kept at room temperature.

after doping some of the solid solutions with non- magnetic and magnetic impurity ions. For making these extensive studies, following samples were prepared :

CHANNEL NUMBER -

FIG. 2. - Mossbauer spectra of the CrzO3-3.85 wt. % Fez03 system annealed at 1 400 OC for 10 hours in air and then quenched to room temperature (category B). The arrow indicates the zero velocity position and the velocity calibration is : 1 chan- nel = 0.129 96 mmls. The source used was 57C0 diffused in a

copper matrix and was kept at room temperature.

2 t psi (tan, = 60 hours) and 3.75 t psi (tan, = 10 hours) respectively before being annealed in air at 1 250 OC (t psi = ton per square inch). These samples were slow- cooled to room temperature after annealing.

(5) Samples of category E contained various concen- trations of Fe3' ions (x = 1.77, 3.85, 5.66 and (1) Samples of category A, with x = 3.85 wt. %, ^{7.41 wt.} 2) and were annealed in air at 1 250 OC for were annealed in air at 1 400 OC fol a period of 2, 4, 10 hours before being slow-cooled to room tempe- 6 and 10 hours before cooling them slowly to room rature.

temperature. (6) Samples of category F contained nonmagnetic

(2) The sample of category B (x = 3.85 wt. %) was and magnetic impurity ions, A13+, Gd3', Fe2' annealed in air at 1400OC for 10 hours and then and Ni2', doped in various concentrations in quenched in water to room temperature. Cr203 - 3.85 wt. % Fe203 system in the form of A120,, Gd203, FeTiO, and NiTiO, respectively.

(3) of ^{' 9} = 3.85 wt. %, were In the samples containing ~ 1 3 + and ^{~ ~} impurity ^{2 '} annealed in air at 1 250 OC for a period of ²⁷ 99 '07 ions, the ratio of the concentrations of impurity ions 26 and 6o hours and then and Fe3' ions, Xi,,,,/XFe3+, was 0.5. The samples

temperature. containing Gd3+ had

(4) Samples of category D, x ⁼ 3.85 wt. %, were

subjected to a pressure of 1 t psi (tar,, = 3,10,17 hours), X,,,/X,,,+ = 0.25 and 0.5

whereas for the Ni2' doped samples Ximp/XFe3+ was 0.5, 2.0 and 7.0. All these samples were fired in air, except the Fez'-doped one which was annealed in vacuum, at 1 250 OC for 10 hours and then were slow- cooled to room temperature.

(7) Similar to the case of AI2O3-Fe203 system [7], vacuum-annealing is expected to create Fez+ ions in the Crz03-Fe203 system also. In order to see the effect of these Fe3+ ions, if produced, on the Fe3+ Moss- bauer line shapes, sample of category G was prepared by annealing the (1 - x) Cr203-xFez03 mixture (x = 3.85 wt. %) in vacuum at 1 250 OC for 10 hours and then cooling it slowly to room temperature. For a better understanding of the results obtained in this sample, more samples were prepared. Sample of cate- gory H, x = 3.85 wt. %, was first annealed in vacuum for 10 hours and then was refired in air for the next 10 hours. The sample of category I (x = 3.85 wt. %) was annealed in vacuum for 20 hours whereas the sample of category J was just cr-FezO, annealed in vacuum for 10 hours. All these samples (belonging to categories H, I, J) were annealed at 1 250 OC and were slow-cooled to room temperature.

The results obtained are summarised in figure 3 where the temperature interval TN - T*, during which the anomalous line shapes are observed, is plotted as a function of annealing time, concentration of impurity ions, Xi,,, and the Fe3' ion-concentration,

x . We see an exponential decrease in TN - T* with

increasing tan, (Fig. 3a, TN = NCeI temperature).

Increasing the annealing temperature or quenching the sample after annealing or subjecting it to high pressure before annealing has also been found to decrease TN - T*. However, the effect of Tan, has been found to be more pronounced than the effects of tan,, rate of cooling and applied pressure. An annealing for even 2 hours at 1 400 OC has been seen to be enough to make TN - T* - 0. As seen in figure 3b, ^c, the effect of both the increasing xi,, and the increasing x is to increase TN - T*. For all these samples, the Nee1 temperature, TN, have been determined by the sta- tionary absorber technique [8]. The definition of the temperature T* is a bit arbitrary. We have taken it as the temperature at which the enhancement in the intensity of the central portion of Mossbauer spectrum starts becoming somewhat evident when the tempe- rature is increased towards TN. More specifically, we have taken T* as the temperature where 13/11 = 1.76 ; here I, is the average amplitude of the central peaks and I, is the average amplitude of the outermost peaks in the Mossbauer spectrum.

The vacuum-annealed CrZO3-3.85 wt. % Fez03 solid solution (category G) and the Fe2+-doped sample (category F), which was also annealed in vacuum, have been found to show identical Mossbauer spectra. They do not show any six-line or anomalous Mossbauer pattern, instead they show a quadrupole doublet, with an isomer shift and splitting characte-

X ( w t . % ) -

FIG. 3. - Variation of the temperature interval T N - T*, during which the anomalous Mossbauer line shapes are observed, in the (1 - x) Cr203-x Fez03 system with (a) the annealing time tann(x ⁼ 3.85 wt. %, category C), (b) the ratio of the concentrations of impurity ions and Fe3+ ions, ximP/xres +, in the impurity-doped samples (x ⁼ 3.85 wt. %, tann = 10 hours, category F) and (c) the concentration, x , of Fe3+ ions

(tann = 10 hours, category E). All the samples have been anneal- ed in air at 1 250 OC and then cooled slowly to room temperature

( T N = NCel temperature).

ristic to the Fe2+ compounds, at 80 K which is seen

to merge into a broad single line at room temperature

(Fig. 4a). An interesting observation is that when

the vacuum-annealed Cr,O,-3.85 wt. % Fez03 system

(category G) is reannealed, for the next 10 hours,

in air at 1 250 OC (category H), anomalous Mossbauer

spectra are observed again (Fig. 4b). This reappea-

rance of the anomalous line shapes is not due to any

prolonged annealing since the Mossbauer spectra of the

sample of category I, where tan, = 20 hours, have

C6-666 J. K. SRIVASTAVA AND R. P. SHARMA

CHANNEL NUMBER --+

FIG. 4. - Mossbauer spectra of the Cr203-3.85 wt. % Fez03 system when (a) annealed in vacuum for 10 hours (category G ) and (6) annealed first in vacuum for 10 hours and then in air for the next 10 hours (category H). In both the cases, the annealing temperature is 1 250 OC and the samples have been slow-cooled to room temperature after annealing. The arrows indicate the zero velocity position and the velocity calibration is : 1 chan- nel ⁼ 0.125 38 mm/s. The source used was 57Co diffused in a

chromium matrix and was kept at room temperature.

been seen to be exactly same as those of the sample of category G (Fig. 4a). It may be mentioned here that the studies made with the sample of category J show that unlike the Mossbauer spectra of the Cr203-3.85 wt. % Fe203 solid solution, the Mossbauer spectra of the undoped a-Fe,03 system are unaffected by the vacuum-annealing.

We will discuss the results of the air-fired and the vacuum-annealed solid solutions separately. The appearance of the anomalous line shapes in the incompletely annealed air-fired samples indicates the presence of either the superparamagnetic effects or the electronic relaxation effects in these solid solutions [9].

We will analyse our results on the basis of these two effects and try to investigate that which of these effects are likely to be present in the solid solutions studied here.

It has been shown in some cases [lo-151 that an incomplete annealing of the host-impurity system produces a nonuniform distribution of the ions of the diffusing species in the host lattice. Such an inhomo- geneous diffusion of the impurity ions causes clustering of these ions in the matrix of the host. These impurity ion-clusters are superparamagnetic [lo] and therefore their direction of magnetisation fluctuates ran- domly with a flip-time given by [6]

where n - ^{1 (k =} Boltzmann's constant, X ⁼ aniso- tropy energy constant and V = volume of the cluster).

These superparamagnetic clusters give rise to a six-line or an anomalous or a single line Mossbauer pattern depending upon whether z, > z, or z, - ^{z, or}

z, < z, which, at any temperature, is essentially

decided by the volume of the cluster ; evidently, bigger the cluster-size, larger is the value of z, (7, ⁼ Nuclear Larmor precession time).

It is quite likely that in the presently studied Cr203-Fe,O, solid solutions also, superparamagnetic Fe3' ion-clustering is taking place when the solid solutions are annealed at 1 250 OC for short annealing times. Assuming this to be the case, z, has been deduced by comparing the observed line shapes with the theore- tically computed ones [9, 16-17]. This gives us z, as a function of tan,, x and xi,,. Now from eq. (1) and assuming that z, and X do not depend on the cluster- volume, we get the ratio V2/V1 as :

where C = kT/KV, - ^{1 for z,} - ^{7 , .} The dependence of V2/Vl on tan,, x and xi,, are shown in figure 5 ; here V , is the normalising average Fe3+ cluster- volume defined suitably in figure caption for each case.

We see an increase in the Fe3+ cluster-volume with increasing tan, (Fig. 5a). It is possible to understand this result on the basis of the above mentioned super- paramagnetic Fe3+ ion-clustering hypothesis. Anneal- ing the solid solutions for a longer time is expected to diffuse the Fe3+ ions more uniformly in the host matrix. This homogeneous diffusion increases the distance between the Fe3 ' ions resulting in an increase in the volume of the Fe3' ion-clusters. As the effect of increasing the Tan, or quenching the sample or sub- jecting it to pressure before annealing is to decrease T, - T*, it follows that these treatments also give rise to a more homogeneous diffusion of the Fe3+ ions in the host lattice. Similarly the decrease observed in the average Fe3+ cluster-volume with increasing x and xi,, (Fig. 5b, c) indicates that the presence of Fe3+

ions in large concentration or those of the impurity

ions causes a more nonuniform distribution of the

Fe3' ions in the matrix of the host. This is physically

understandable since the solubility of the ions of the

diffusing species in the host matrix, at any temperature,

FIG. 5. - Dependence of the normalised average Fe3+ cluster- volume, in the (1 - x) Cr203-x Fez03 system, on (a) the annealing time tann (X = 3.85 wt. %, category C), (b) the Fe3+

ion-concentration x (tann ⁼⁼ 10 hours, category E) and (c) the ratio of the concentrations of impurity ions and Fe3+ ions, xi,,/xp,s+, in the impurity-doped samples (x = 3.85 wt. %,

tann = 10 hours, category F). All the samples have been annealed in air at 1 250 OC and then slow-cooled to room temperature.

These samples have a distribution in the size of Fe3+ ion-clusters and V here represents the average Fe3+ cluster-volume (ACV) ;

Vt = ACV corresponding to the annealing time t, V,_,_ = ave- rage volume of the Fe3+-clusters formed in the solid solution annealed for 2 hours, Vz = ACV corresponding to the Fe3+- concentration x, Vxmax = average volume of the Fe3+-clusters formed in the Cr203-7.41 wt. % Fez03 system, Vg = ACVexis- ting in the Cr203-3.85 wt. % Fez03 solid solutions doped with an impurity ion of concentration ^Ximp and ^{v ~ i 2 +} is the average

Xmax

0

volume of the Fe3+-clusters present in the Ni2+-doped Cr203- 3.85 wt. % Fez03 system when ximp/xFe3 + = 2.0.

I I I I ^I I ^I

2 4 6 8

. u X ( w t % ) -

is inversely proportional to the concentration of these ions [18]. Noticing that an increase in the Fe3' cluster- volume means a decrease in the concentration of Fe3+

ions at any point q and using the simple diffusion equation [18, 191,

( c )

:I< ^{A No ~ impurity d ~ +}

0 ~ i * +

1.0 El A L ~ +

where D is the diffusion coefficient, xo is the initial concentration of the Fe3' ions at q ⁼ 0, t ⁼ 0 and t = tan,, we have obtained q2/4 D as a function of tan, [17]. We see an increase in q2/4 D with increasing annealing time. This dependence of D on t,,,, at any point q, is probably due to the fact that D is expected to depend on the concentration of Fe3' ions [18, 191 which changes at the point q when tan, changes owing to a change in the volume over which the Fe3' ions are distributed. For evaluating D, it is therefore necessary to take this effect into account. Such a calculation is being done [17].

We will now examine the possibility of the presence of electronic relaxation effects in the 1 250 OC-annealed samples in which the anomalous line shapes have been observed. The existence of nonuniformly distri- buted Fe3' ions and those of the various types of lattice defects, both of which are almost absent in the 1 400 OC-annealed samples as they have been annealed out in them, creates a kind of magnetic disorder in the lattice of these 1 250 OC-annealed solid solutions. This magnetic disorder is expected to highly damp the long-range collective motions of the ionic spins existing in a normal magnetically ordered system [20].

Such a damping of collective modes greatly affects the electronic relaxation mechanisms operative in the system and the system effectively starts behaving like a paramagnetic one as far as the relaxation processes are concerned. This phenomenon can make the elec- tronic relaxation time, z,, ,,, -- z, giving rise to the anomalous Mossbauer line shapes. It may be men- tioned here that the NCel temperature TN, the T/TN dependence of the hyperfine field and the temperature dependence of the magnetic structure of the samples of categories A and B, which show normal Moss- bauer spectra, have been found to be exactly identical with the corresponding quantities of the samples of category C where anomalous line shapes have been seen. This results puts some doubt on the presence of magnetic disorder and hence on the existence of electronic relaxation effects in the 1 250 OC- annealed samples. However, t o understand this result on the basis of superparamagnetic effects is also somewhat difficult and thus it is not possible to predict conclusively whether the observed anomalous spectra are due to superparamagnetic effects or the electronic relaxation effects [17].

The results of the vacuum-annealed samples (Fig. 4)

can be understood either by assuming the formation

of FeCr204, embedded in the Cr20, host, upon

C6-668 J. K. SRIVASTAVA AND R. P. SHARMA

vacuum-annealing [21] or by visualising the presence vacuum-annealed samples [22]. Further measurements of randomly fluctuating electric field gradient owing are being carried out for understanding these to the random jump of vacancies (lattice defects) i n the results [17].

References

[I] SRIVASTAVA, J. K. and SHARMA, R. P., Phys. Stat. Sol. 35 (1969) 491.

[2] MORUP, S. and THRANE, N., Phys. Rev. B 4 (1971) 2087.

[3] SRIVASTAVA, J. K. and PRASAD, K. G., Phys. Stat. Sol. (b) 54 (1972) 755.

[4] CLAUSER, M. J. and BLUME, M., Phys. Rev. B 3 (1971) 583.

[5] GONZALEZ-JIMENEZ, F., IMBERT, P. and HARTMANN- BOUTRON, F., Phys. Rev. B 9 (1974) 95.

[6] COLLINS, D. W., DEHN, J. T. and MULAY, L. N., Mossbauer Effect Methodology, Vol. 3, ed. Gruverman, I. J.

(Plenum Press, New York) 1967 p. 103.

[7] BHIDE, V. G. and DATE, S. K., Phys. Rev. 172 (1968) 345.

[8] SRIVASTAVA, J. K. and SHARMA, R. P., Phys. Stat. Sol. (b) 49 (1972) 135.

[9] SRIVASTAVA, J. K., Phys. Stat. Sol. (b) 55 (1973) K 119.

[lo] RUBINSTEIN, M., Solid State Commun. 8 (1970) 919.

1111 FERRANDO, W. A., SEGNAN, R. and SCHINDLER, A. I , J. Appl. Phys. 41 (1970) 1236.

[12] FERRANDO, W. A., SEGNAN, R. and SCHINDLER, A. I., Phys. Rev. B 5 (1972) 4657.

[13] GONSER, U. and WIEDERSICH, H., J. Phys. Soc. Japan 18 (1963) SuppI. I1,47.

[14] BERKOWITZ, A. E. and FLANDERS, P. J., J. Appl. Phys. 30 (1959) 111s.

1151 TROUSDALE, W. L., LONGWORTH, G. and KITCHENS, T. A., J. Appl. Phys. 38 (1967) 922.

1161 MCNAB, T. K., Fox, R. A. and BOYLE, A. J. F., J. Appl.

Phys. 39 (1968) 5703.

[17] SRIVASTAVA, J. K. and SHARMA, R. P., Details will be published elsewhere.

1181 GRAY, T. J., DETWILER, D. P., RASE, D. E., LAW- RENCE, W. G., WEST, R. R. and JENNINGS, T. J., The Defect Solid State (Interscience Publishers, Inc., New York and Interscience Publishers Ltd., London) 1957.

[19] SUPTITZ, P. and TELTOW, J., Phys. Stat. Sol. 23 (1967) 9.

[20] BHARGAVA, S. C. and IYENGAR, P. K., Phys. Stat. Sol. (b) 53 (1972) 359.

[21] GREENWOOD, N. N. and GIBB, T. C., MiiSSbauer Spec- troscopy (Chapman and Hall Ltd., London) 1971, p. 263.

[22] DANON, J., Chemical Applications of Mossbauer Spee- troscopy, ed. Goldanskii, V . I., and Herber, R. H.

(Academic Press, New York and London) 1968, p. 249.