900 Materials Science and Engineering, A 134 ( 1991 ) 900-903

M6ssbauer measurements and susceptibility measurements on crystalline and icosahedral AICuFe alloys

F. Miiller and M. Rosenberg Lehrst. f Exp.-Phys. VI, Ruhr-Universitdt Bochum, D-4630 Bochum (F.R.G.)

W. Liu and U. K6ster FB Chemietechnik, Universitiit Dortmund, D-4600 Dortmund 50 (F.R.G.)

Abstract

In order to obtain more information on the atomic decoration and short range order of the icosahedral A1CuFe alloys, we compared the stable icosahedral phase with the crystalline phases of the parent com- position and four icosahedral alloys around the thermodynamical stable composition by means of mag- netization and M6ssbauer measurements. The measurements of the magnetic susceptibility show that in the AICuFe icosahedral alloys as well as in some of the parent crystalline compounds the magnetic moment of iron is suppressed, the magnetic behaviour being dominated by the electronic structure of aluminium and copper. It seems that the short range order in the studied icosahedral phases has some similarity with the one occurring around the 4(i) iron sites with 10 aluminium nearest neighbours in the crystalline structure of A113Fe4 •

1. Introduction

In the field of quasicrystalline materials there are a number of well elaborated theories today, describing quasilattices and explaining the par- ticular symmetry properties of quasicrystals [1]. However, up to now very little knowledge could be achieved about the atomic decoration of these quasilattices. It has been proposed to describe quasicrystals in terms of atomic units existing in crystalline phases of similar chemical composi- tion, and a comparison of structural data has been done, for example, for AI-Mn [2, 3]. For similar reasons the aim of the present study was to compare the M&ssbauer data and magnetic data of AICuFe quasicrystals and the crystalline phases of the parent composition [4].

2. Experimental details

For the present study we prepared the crystal- line samples AIsFe2, Al13Fe4, and A17fu2Fe , and the quasicrystalline samples A164Cu23Fe13 (stable phase [5]), A165Cu20Fe15, A170Cu17Fe13, A170fu20- Felo, A177Cu13Fe10, and Al70CuloFe20. All master

alloys were obtained by consecutive arc melting of high purity metals in argon atmosphere. The quasicrystalline samples were produced from the master alloys by conventional melt spinning in argon atmosphere at a roller speed of 25-35 m s- 1. The structures of both crystalline and quasi- crystalline samples were checked by X-ray dif- fraction analysis and proved to be single phase within the detection limit for crystalline A17Cu2Fe and i - A 1 6 4 C u 2 3 F e 1 3 annealed in two steps at (1058 + 5) K for 72 h and then at 1043 K for 3 h. The following secondary phases in an estimated amount below 5% occurred in the other samples: Al13Fe 4 in AIsFe2, as quenched A164Cu23Fe13 and A165Cu20Fe15 (both, as quenched and annealed), A12Cu in i-Al70CUlTFel3 and i-A170Cu20Fel0 , AI in i-A177Cu13Fe10 , and an unidentified phase in Al13Fe 4. A17oCul0Fe20 contained larger amounts of several phases. In addition, for the icosahedral samples Al70Cul7Fel3 , A170Cu20Fel0 and A177- Cu13Fe10 the structures were confirmed by elec- tron diffraction.

The magnetic measurements were performed with a vibrating sample magnetometer in the tem- perature range 4-300 K and fields in the range 0.8-1.75 T. All the M6ssbauer 57Fe spectra were

0921-5093/91/$3.50 © Elsevier Sequoia/Printed in The Netherlands

taken at room temperature using standard equipment, the iron content of the M6ssbauer adsorbers lying below 5 mg cm-2.

3. Results

3.1. Magnetization measurements The temperature dependent susceptibility

curves corrected for a contribution of core dia- magnetism are shown in Figs. 1 and 2. For quanti- tative evaluation we tried to fit the susceptibility data using a Curie-Weiss law. The samples Ai~sFe4, i-AlyoCUl7Fel3 , i-AlyoCu20Felo and i- A170Cu,)Fe:o required an additional T ~/2 term, according to the Bloch law for spin waves, prob- ably due to a small quantity of ferromagnetic phases. This leads to the following equation for X

901

C M0 - - + - - ( 1 - a T s/2) X= XP + T - ® H

Xp being the temperature independent Pauli sus- ceptibility, C the Curie constant, ~ the paramag- netic Curie temperature and M 0 the magnetization of the ferromagnetic phase. Using a fitting function of the form

al .3,,~ X = a() -~ I- a~ 7 -' ~

(T-a~)

one can see that % = Zp + Mo/H can no longer be identified with Zp if a ferromagnetic secondary phase is present.

The results of the fits are listed in Table 1. The effective magnetic moment of the iron atoms was

T A B L E 1

Magnetic properties"

Sample C P~ ~ 0 X,, ( 1 0 - " c m 3 g I g ) (~13) (K) (10 ScmSg ')

Crystalline AI5 Fe~ 541 0.73 - 4.5 ( 105 ) AI].~Fej 167 0.44 - 5.4 (626) AIvCuzFe ~< 10 -<-<0.18 I0

lcosahedral AI~4Cu2~Fel ~ 1 1 0.16 - 6.1 6

a. AI64Cu23FeI3 ~< 10 ~<0.16 33 A I ~ C u ~F'e~ s ~< 10 ~<0.15 AIT~Cu].~Fe,0 < 5 < 0.12 20 AIT(~Cu2(IFe m < 10 < 0 . 1 7 Al7.Cu]oFe2. 300 0.66 - 7.4 ( 1321 )

~C = Curie constant; Pelf = effective moment of i ron-atoms; O = paramagnet ic Curie temperature; XP = Pauli paramagnet ic suscep- tibility; a. annealed sample; values in brackets due to ferromagnetic secondary phases.

40

35

30

25 7

o 15

".~o

5

- 5

{O) AI~Fe~ (b) AI~3FE 4 (C} A17Cu~Fe

k)

50 :tO0 150 200 250

T ( K ) 300

(B) A 17oCLI ~.TF e t3 k {b] AI64Cu23Fe ~3

I "~f~.!a) (c) a. AIB4Cu23Fe~3

u "~ , . (b} : ~~,,',;-.~:,.,::,.,.; ..

~< (c)

_ ~. 510 I . . I I I _ i O 0 150 200 250 300

T ( K )

Fig. 1. Tempera ture dependent susceptibility of crystalline Fig. 2. Typical susceptibility curves of icosahedral A l - (Cu) -Fe samples. A1-Cu-Fe samples (a. annealed sample).

902

TABLE 2

(a) M6ssbauer parameters for crystalline samples a

Sample Doublet 1 Doublet 2

Int. IS QS LW Int. IS QS LW (%) (mms- ' ) (mms -I) (mms -I) (%) (mms 1) (mms- l ) (mms 1)

AlsFe 2 65.6 0.137 0.561 0.282 34.4 0.118 0.337 0.264 A17Cu2Fe 100 0.045 0.142 0.321 All3Fe 4 61 0.093 0.403 0.293 39 0.095 0.121 0.274

alnt. = rel. intensity; IS = isomershift; QS = quadrupol splitting; LW = line width at half maximum.

-7 #

[a) (b)

-t o o.o t.'o -t o -o.o 1.0

relotive velocity / mm/s

•"•'•S,,I \ //~,,~

t!1

to) V

-I.0 -0.0 1.0

Fig. 3. Mrssbauer spectra of (a) crystalline AI7Cu2Fe, (b) crystalline AI t 3Fe4, and (c) crystalline AIsFe 2.

7

0.0 1.0

7

(b) ,~ -I ~. 0 O. 0

(a)

-i ~. 0 l.~O

r Q l a t i v e v Q l o c i t y / m m / s

Fig. 4. Typical Mrssbauer spectra of icosahedral AI-Cu-Fe alloys: (a) i -ml70fuzoFelo; (b) i-A164Cu23Fe13 .

calculated from the Curie constant C assuming that all iron atoms possess the same magnetic moment, the copper and aluminum atoms being nonmagnetic. Among the crystalline samples, A15Fe 2 and Al13Fe4 show a pronounced

Curie-Weiss paramagnetism which cannot be due to secondary phases of concentration below the sensitivity of the X-ray diffraction. Contrary to this, in the case of A17Cu2Fe the paramagnetic behaviour is at the detection limit of the mag- netometer ( C = 2 x 10-6-10 -5 cm 3 g- 1 K).

Like AI7Cu2Fe , all icosahedral samples, except i -Al70Cul0Fe20, s h o w n o o r v e r y w e a k Curie-Weiss paramagnetism as reported earlier in the case of i-A165Cu20Fe15 [6, 7]. The magneti- zation data of i-A170Cu17Fe13 and i-A170Cu20Fel0 could not be well fitted by the function intro- duced above. Therefore the corresponding values are less certain.

If the paramagnetic Curie temperature E) could be determined by the fit it was in the range of - 7.5 to - 4.5 K for both crystalline and quasi- crystalline samples, indicating a tendency to anti- ferromagnetic ordering. Provided no ferro- magnetic phase was present, Xp values were rather low, the higher value of 105 x 10 -8 cm 3 g-1 for A15Fe 2 possibly resulting from a small quantity of a ferromagnetic phase as in A113Fe 4 .

TABLE 2

(b) M6ssbauer parameters for icosahedral samples ~

Sample ISA QSA (rams ') (mms ')

AI~4Cu2~Fe L~ 0.122 0.374 a. AI~>aCu2 ~Fel3 0.113 0.383

Al~,~Cu2.Fe~ 0.127 0.350 a. Al~,sCuz~Fe 15 0.112 0.325

AlToCul :Fel 3 0.102 0.374 Al711CuzltFel0 0.080 0.385 AlToCu j t~Fe21 ~ 0.102 0.258 A177Cu13Fell , 0.(191 0.383

alSA, QS A = averaged values of both doublets; a. annealed subsequently.

3.2. M6ssbauer spectra Magnetic ordering phases could not be de-

tected in any of the M6ssbauer spectra. The fit parameters for the crystalline samples are listed in Table 21a) and the corresponding spectra are shown in Fig. 3. Tetragonal AI7CuzFe has a single iron site with 9 aluminium nearest neighbours [4, 8] resulting in low values of IS and QS. A third of the iron atoms in monoclinic Al l3Fe 4 have a similar environment [9]. The rest of the iron atoms in AII3Fe4 have 10 nearest neighbours followed by an iron atom at a distance of about 0.3 nm, yielding higher QS and IS values. In orthorhombic A15Fe z [10], in addition to a doublet with high QS and IS values another one with low values is found for which a structural reason cannot be given.

At first glance it can be seen that the spectra of the quasicrystalline samples (Fig. 4) are not simi- lar to any of the crystalline phases. A fit by two doublets is possible: however, this need not mean the actual existence of two distinct atomic envi- ronments, as a fit with correlated distributions of both QS and IS might be better adopted to the problem [4, 7, 1 1]. Therefore we decided to use

903

the average values ISA, QS A (Table 21b)), which are less dependent on the kind of fit [7]. IS A and QS a do not show any definite correlation to the composition of the sample. However, there is an increase in symmetry of the spectra with decreas- ing iron content. The values of IS A and Q S A for the quasicrystalline samples are next to the narrow doublet of A15Fe 2 and the wide one of Al13Fe 4. This may indicate an average iron envi- ronment in AICuFe quasicrystals similar to some extent to the one stated above for Alt ,~Fe4.

Acknowledgment

The financial support of the Deutsche For- schungsgemeinschaft is gratefully acknowledged.

References

1 See, for example, T. Janssen, Phys. Rep., 168 (1988) 53-113.

2 E. A. Stern, Y. Ma and C. E. Bouldin, Phys. Rev. Lett., 55 11985) 2172-2175.

3 A. Yamamoto and K. Hiraga, Phys. Rev. B, 37 (1988) 6207-6214.

4 For a similar approach, see N. Kataoka, A. R Tsai, A. lnoue, T. Masumoto and Y. Nakamura, Jpn. J. Appl. Phys., 2711988) L1125-1127.

5 C. Dong, M. DeBoissieu, J. M. Dubois, J. Pannetier and C. Janot, J. Mater. Sci. Lett., 8 (1989) 827-830.

6 S. Matsuo, T. Ishimasa, H. Nakano and Y. Fukano, J. Phys. F, 1811988)L175-L180.

7 Z. M. Stadnik, G. Stroink, H. Ma and G. Williams, Phys. Rev. B, 3911989) 9797-9805.

8 W. B. Pearson (ed.), Structure Reports" of the International Union of Crystallography, Utrecht, 26 ( 1961 ) 7.

9 W. B. Pearson (ed.), Structure Reports" of the International Union of Crystallography, Utrecht, 19 (1955) 22.

10 W. B. Pearson (ed.), Structure Reports of the International Union of Crystallography, Utrecht, 17(1953) 23. S. Nasu, U. Gonser and R. S. Preston, J. de Physique ( , 41 11980) C1-385f.

11 Z. M. Stadnik and G. Stroink, Phys. Rev. B, 38 (1988) 10447-10453.