Microstructural and magnetic studies of Mn–Al thin films

Microstructural and magnetic studies of Mn–Al thin films

P. C. Kuo

Institute of Materials Science and Engineering, National Taiwan University, Taipei 107, Taiwan Y. D. Yao

Institute of Physics, Academia Sinica, Taipei 115, Taiwan J. H. Huang and S. C. Shen

Institute of Materials Science and Engineering, National Taiwan University, Taipei 107, Taiwan J. H. Jou

Department of Materials Science and Engineering, Tsing Hua University, Hsinchu 300, Taiwan

Mn–Al thin films with high coercivity and high saturation magnetization were successfully fabricated by rf magnetron sputtering with properly controlled chemical composition, substrate temperature, and annealing temperature. A high coercivity of about 3000 Oe and a saturation magnetization of about 420 emu/cc have been achieved. We have observed that during annealing at 410 °C, the nonmagneticephase with a grain size of roughly 100 nm transforms into a metastable ferromagnetic t phase with a platelike grain size of roughly 300 nm. From the continuous measurement of the stress of the films in vacuum as a function of temperature, we observed a compression stress during heating below 220 °C, and a tension stress above 220 °C during cooling. The structure phase transformation from e to t phases was related to the stress variation from compression to tension. The high coercivity can be explained by the high magnetocrystalline anisotropy constant of thetphase and the magnetoelastic energy arises from the residual stress of Mn–Al films after the shear transformation. © 1997 American Institute of Physics.

@S0021-8979~97!21308-3#

INTRODUCTION

In the bulk Mn–Al alloy system, a wide range of com-positions has been studied extensively, and only the alloys containing about 50–60 at. % Mn exhibit a ferromagnetic phase. This ferromagnetic phase has been identified as a metastable t phase that is a tetragonal Ll₀-type superstruc-ture so that the magnetic moments of Mn atoms in alloys are parallel to each other.1–5 It is well known that the high-temperature nonmagnetic hexagonal-closed-packed e phase transforms into an orthorhombic e

8

In a filmtype MnAl alloy system, because the magnetic properties of Mn–Al alloys are processing sensitive, it is quite difficult to obtain pure MnAl t-phase alloy films with high saturation magnetization and high coercivity.6–10

EXPERIMENT

The MnAl alloy films containing Mn between 30 and 70 at. % were prepared by rf magnetron sputtering. The targets were made from a high purity Al~99.999%! disk and over-laid with small high purity Mn ~99.99%! pieces. This ar-rangement provides a wide range of effective target compo-sitions and, therefore, film compocompo-sitions. Films were deposited on glass substrates at a substrate temperature Ts

range between 30 and 200 °C. The substrate holder was ro-tated during deposition. The rf power was controlled with a deposition rate of 0.5 nm/s. The base pressure in the chamber was 531027Torr, and after the high purity Ar gas was in-troduced, the pressure was kept at 1 mTorr. A typical thick-ness of the films was 0.8mm. Thermal annealing was carried out at a temperature between 350 and 550 °C in vacuum. The

microstructure of the films was characterized by x-ray dif-fractometer and transmission electron microscopy diffraction technique and their compositions were determined by an electron probe microanalyzer ~EPMA! calibrated by a stan-dard bulk Mn₅₅Al₄₅alloy. A vibrating sample magnetometer was used for the magnetic studies. The stress variation of the film samples during the annealing process was studied by using a bending beam method. Under this method, the film sample was clamped in the vacuum oven, and from the re-flection of a He–Ne laser beam, we converted the signal of the deflection position of the laser beam into the stress value continuously.

RESULTS AND DISCUSSION

The Mn–Al alloy films containing Mn between 30 and 70 at. % were prepared by the rf magnetron sputtering tech-nique. The compositions were determined by the EPMA technique. The relation between the substrate temperature and the magnetic properties of the films has been measured systematically. At first, we noticed that the concentration of the e phase in the as-deposited samples varied as the sub-strate temperature was varied. For example, Fig. 1~a! shows the x-ray diffraction patterns for the as-deposited Mn50Al50

films with substrate temperatures~a! 200 °C, ~b! 100 °C, and

~c! 30 °C. It is clear that the intensity of the ephase peaks increased with decreasing substrate temperature and the half-width of the peaks also decreased with decreasing the sub-strate temperature. This indicates that the films deposited at a lower Ts have a larger and more perfect crystallinee-phase

structure than the films deposited at higher Ts. From the

magnetization measurement of all the as-deposited films, the magnetizations of all the as-deposited samples are very low. 5621 J. Appl. Phys. 81 (8), 15 April 1997 0021-8979/97/81(8)/5621/3/$10.00 © 1997 American Institute of Physics

After annealing at temperatures between 350 and 550 °C in vacuum for 30 min, only the sample with Mn₅₀Al₅₀ showed the highest saturation magnetization, the highest coercivity, and almost pure t phase. For explanation, Fig. 1~b! shows the x-ray diffraction patterns of three MnAl film samples with Mn concentrations of ~a! 44, ~b! 50, and ~c! 56 at % after annealing at 410 °C for 30 min. Only the Mn50Al50

sample shows an almost puret-phase diffraction peak. There are always coexisting other phases for all samples besides Mn50Al50. For example,g phase in Mn44Al56, andb-Mn in

Mn56Al44as shown in Fig. 1~b!.

From the magnetic measurements of the samples after annealing between 350 and 550 °C, we found that the best condition was annealing at 410 °C for 30 min and the ferro-magnetic phase appeared at a composition range between 40 and 60 at. % Mn. Table I lists the saturation magnetization

Msand coercivity Hcfor various MnAl films after annealing

at 410 °C for 30 min and the subtrate temperature T_s530 and 100 °C. The maximum M_s of about 420 emu/cc and a maximum H_c of about 3000 Oe was obtained for the Mn₅₀Al₅₀sample with T_s530 °C. In general, thee→tphase transformation occurred during the annealing treatment. Since the M_sof the annealed films decreased with increasing T_s, the transformation fraction should decrease with increas-ing T_s, i.e., the films deposited at lower T_shave a large and more perfect crystalline structure than the films deposited at higher T_s. In other words, at a low substrate temperature the sputtered atoms arriving on the substrate should not have enough energy to form the metastabletphase or the equilib-riumbandgphases, and theephase was formed due to the superquenching effect.

Figure 2 shows the stress–temperature curve of the Mn50Al50 film during the whole annealing process, which

were recorded at a heating rate of 5 °C/min from room tem-perature to 450 °C and then furnace cooled to room tempera-ture. We observed a compression stress for the heating run roughly below 220 °C, and a tension stress roughly above 220 °C during the cooling run. This suggests that the struc-ture phase transformation from e to e

8

FIG. 1. ~a! X-ray diffraction patterns of the as-deposited Mn50Al50 films

with substrate temperature~a! 200 °C, ~b! 100 °C, and ~c! 30 °C. ~b! X-ray diffraction patterns of the MnAl films with Mn concentration of~a! 44 at. %, ~b! 50 at. %, and ~c! 56 at. %, after annealing at 410 °C for 30 min.

FIG. 2. The stress of the Mn50Al50film as a function of temperature between

room temperature and 450 °C ~closed square: heating run; open square: cooling run!.

TABLE I. Magnetic properties of various MnAl film samples after anneal-ing at 410 °C for 30 min.

Film composition Ms~emu/cc! Hc~Oe! TS530 °C TS5100 °C TS530 °C TS5100 °C Mn30Al70 0 0 0 0 Mn35Al65 10 5 20 15 Mn40Al60 30 20 150 130 Mn43Al57 140 40 1580 700 Mn48Al52 340 90 2400 1300 Mn50Al50 420 220 3000 1750 Mn54Al46 270 80 1650 1100 Mn60Al40 25 30 200 170 Mn64Al36 15 10 50 35 Mn70Al30 0 0 0 0

5622 J. Appl. Phys., Vol. 81, No. 8, 15 April 1997 Kuoet al.

roughly between 220 and 450 °C. After cooling to room tem-perature, the residual stress s of this t-phase film is about 133108 N/m2. Therefore, the high coercivity is explained due to the high magnetocrystalline anisotropy constant K1of

the t phase ~K₁>107 erg/cm3!1 and the magnetoelastic en-ergy E_me ~Ref. 11! arises from the residual stress

~Eme}s513310

8_N/m2_{! of MnAl films after the shear}

trans-formation.

The microstructure grain sizes and the phases of the samples were studied by transmission electron microscopy. Figure 3 shows the transmission electron bright field images and diffraction patterns of the as-deposited Mn₅₀Al₅₀ samples with T_s530 °C @Figs. 3~a! and 3~c!#, and the Mn₅₀Al₅₀ samples with T_s530 °C and annealing at 410 °C for 30 min@Figs. 3~b! and 3~d!#. It is clear that the nonmag-netic e phase as shown in Fig. 3~c! with a grain size of roughly 100 nm as shown in Fig. 3~a! for the as-deposited samples transforms into a metastable ferromagnetic tphase as shown in Fig. 3~d! with a platelike grain size of roughly 300 nm as shown in Fig. 3~b!.

In conclusion, we observed that during annealing at 410 °C, the nonmagneticephase with a grain size of roughly 100 nm transforms into a metastable ferromagnetictphase with a platelike grain size of roughly 300 nm. The structure phase transformation frometotphases were related to the variation of the stress from compression to tension. The high coercivity can be explained by the high magnetocrystalline

anisotropy constant of the t phase, and the magnetoelastic energy arises from the residual stress of Mn–Al films after the shear transformation.

ACKNOWLEDGMENT

The authors are grateful for financial support by the Na-tional Science Council of the ROC under Grant Nos. NSC86-2112-M-001-014 and NSC85-2216-E-002-017.

1_{H. Kono, J. Phys. Soc. Jpn. 13, 1444~1958!.} 2

Y. Z. Vintaykin, V. A. Udovenko, I. S. Belyatskaya, N. N. Luarsabishvili, and S. Y. Makushev, Fiz. Met. Metalloved. 38, 398~1974!.

3_{J. Van Landuyt, G. Van Tendeloo, J. J. Van Den Broek, and H.}

Donker-sloot, J. Magn. Magn. Mater. 15, 1451~1980!.

4

P. C. Kuo, Y. D. Yao, J. H. Huang, and C. H. Chen, J. Magn. Magn. Mater. 115, 183~1992!.

5_{J. H. Huang and P. C. Kuo, Mater. Sci. Eng. B 22, 256~1994!.} 6_{A. Morisako, M. Matsumoto, and M. Naoe, IEEE Trans. Magn. 23, 2470}

~1987!.

7

A. Morisako, M. Matsumoto, and M. Naoe, J. Appl. Phys. 61, 4281 ~1987!.

8_{J. X. Shen, R. D. Kirby, and D. J. Sellmyer, J. Appl. Phys. 67, 4929}

~1990!.

9

A. Morisako, N. Kohshiro, M. Matsumoto, and M. Naoe, J. Appl. Phys. 67, 5655~1990!.

10_{M. Matsumoto, A. Morisako, and J. Ohshima, J. Appl. Phys. 69, 5172}

~1991!.

11_{B. D. Cullity, Introduction to Magnetic Materials~Addison-Wesley,}

Lon-don, 1972!, p. 270.

FIG. 3. ~a! Transmission electron bright field images of the as-deposited Mn50Al50samples with Ts530 °C. ~b! Transmission electron bright field images of the Mn50Al50samples with Ts530 °C and annealing at 410 °C for 30 min. ~c! Transmission electron diffraction patterns of the as-deposited Mn50Al50samples

with Ts530 °C. ~d! Transmission electron diffraction patterns of the Mn50Al50samples with Ts530 °C and annealing at 410 °C for 30 min.

5623

J. Appl. Phys., Vol. 81, No. 8, 15 April 1997 Kuoet al.