Hydrogen annealing effects on polycrystalline La0.67Ba0.33MnO3 compound

Hydrogen annealing effects on polycrystalline La0.67Ba0.33MnO3 compound

ARTICLE IN PRESS Journal of Magnetism and Magnetic Materials 297 (2006) 21–25 www.elsevier.com/locate/jmmm Hydrogen annealing effects on polycrystal...

680KB Sizes 2 Downloads 66 Views

ARTICLE IN PRESS

Journal of Magnetism and Magnetic Materials 297 (2006) 21–25 www.elsevier.com/locate/jmmm

Hydrogen annealing effects on polycrystalline La0.67Ba0.33MnO3 compound Yimin Cuia, Liuwan Zhanga,b,c,, Chunchang Wanga, Kai shia, Bisong Caoa,b,c a Physics Department, Tsinghua University, Beijing 100084, People’s Republic of China Laboratory of Advanced Materials, Tsinghua University, Beijing 100084, People’s Republic of China c Materials Research Center, Tsinghua University, Beijing 100084, People’s Republic of China

b

Received 14 December 2004; received in revised form 12 January 2005 Available online 16 March 2005

Abstract Upon annealing polycrystalline La0.67Ba0.33MnO3 bulk samples in flowing 95%Ar:5%H2 mixed gas at 700 1C for different time, the insulator–metal transition temperature, T P and the amplitude of AC magnetic susceptibility were decreased first, then increased, finally decreased again. While the resistivity was increased monotonically. This anomalous behavior was explained by the combinational effects of oxygen loss and Ba ion vacancies caused by the segregation of Ba ion related impurity phase. r 2005 Elsevier B.V. All rights reserved. PACS: 81.40.Rs; 71.30.+h; 75.30.Kz Keywords: A. Manganites; B. AC Magnetic susceptibility; C. Hydrogen annealing

1. Introduction The properties of the perovskite oxides, such as high-T c superconductors, ferroelectric materials, and colossal magnetoresistive (CMR) manganites, are very sensitive to the oxygen content which can usually be varied by annealing the samples at Corresponding author. Physics Department, Tsinghua University, Beijing 100084, People’s Republic of China. Tel.: +86 10 62772762; fax: +86 10 62781604. E-mail address: [email protected] (L. Zhang).

different temperatures in different atmospheres. Extracting the intrinsic relations between the oxygen content and the properties is helpful to improve the understanding of superconductivity, ferroelectrics and CMR effects. For stoichiometric CMR materials La1xAxMnO3 (where A is divalent ions, such as Ca, Sr, and Ba), the strong ability to take up more oxygen makes them very hard to lose oxygen. In order to reduce their oxygen content, gas getter [1,2] or hydrogen [3] is often used during annealing at high temperatures. In the later case some other effects such as

0304-8853/$ - see front matter r 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jmmm.2005.02.008

ARTICLE IN PRESS Y. Cui et al. / Journal of Magnetism and Magnetic Materials 297 (2006) 21–25

hydrogen incorporation into the lattice [4,5] may appear in addition to the oxygen loss, which might lead us to the incorrect conclusions about the dependence of the properties of the materials on the oxygen content. Many investigations are reported on the effects of hydrogen annealing in ferroelectric films [4,6–8], but only a little work is done in CMR materials. In this paper, the annealing effects on La0.67Ba0.33MnO3 bulk ceramic samples in flowing 95% Ar:5% H2 mixed gas (AH) were studied. Upon annealing at 700 1C for different time, both the metal–insulator transition temperature, TP and the AC magnetic susceptibility decreased first, then increased, finally decreased again, in contrast to the previous report on the annealing effects in N2 with Ti oxygen getter [1]. We attribute this anomalous behavior to the combined effects of the oxygen loss and the Barium ion deficiency introduced by hydrogen annealing.

2. Experimental The bulk ceramic samples of La0.67Ba0.33MnO3 were synthesized using a conventional solid state reaction method. The stoichiometric amounts of La2O3, BaCO3 and MnO2 were mixed, ground and fired at 1300–1400 1C repeatedly. Finally the obtained powders were pelletized and fired at 1500 1C in air for another 10 h followed by furnace cooling. The average grain size of the as-prepared sample is about 5 mm by the SEM observation. In order to study the annealing effects, the asprepared bulk La0.67Ba0.33MnO3 samples (sample A) were annealed at 700 1C under flowing (200 ml/ min) AH for 3 h (sample B), 12 h (sample C), 36 h (sample D), and 60 h (sample E), respectively. For comparison, another as-prepared sample was annealed at 950 1C in flowing N2 (99.999%) with carbon powder as oxygen getter for 10 h (sample F obtained). Then samples E and F were further annealed together in flowing oxygen at 300 1C for 4 h, producing sample H and G, respectively. All the samples were characterized by X-ray diffraction (XRD) with Cu Ka radiation. The resistivity was measured by the standard four-probe method. AC magnetic susceptibility was measured by a

lock-in technique. X-ray photoelectron spectra (XPS) were acquired with a PHI15300/ESCA system. Al Ka radiation (1484.6 eV) was used as the source and the C 1s peak was used as a reference.

3. Results and discussion The XRD patterns indicated that the asprepared and all the annealed La0.67Ba0.33MnO3 samples have the cubic perovskite structure with almost the same lattice parameter of a ¼ 0:3906 nm: Within the XRD sensitivity, no other secondary phase was detected. The temperature dependent AC susceptibility w is shown in Fig. 1. The ferromagnetic transition for all the samples starts at almost the same temperature (T CA ¼ 351:7 K), but the transition width sensitively depends on the annealing time in AH. For the as-prepared sample A the transition is very sharp. After annealed for 3 h in AH, the susceptibility of sample B has a tail at high temperature. At low temperature, the susceptibility nearly

C

60 Susceptibility (arb.unit/g)

22

A B

0 200 400 Temperature (K)

D

40

E 20 C 0 0

100

200 300 Temperature (K)

400

500

Fig. 1. Temperature dependence of AC susceptibility for La0.67Ba0.33MnO3 samples of as-prepared (sample A), and annealed in AH at 700 1C for 3 h (sample B), 12 h (sample C), 36 h (sample D), and 60 h (sample E). The inset shows the AC susceptibility of sample C on an expanded scale.

ARTICLE IN PRESS Y. Cui et al. / Journal of Magnetism and Magnetic Materials 297 (2006) 21–25

equals to that of sample A. For sample C, the susceptibility amplitude is greatly decreased and the susceptibility curve shows three stages, without any indication of saturation even at the lowest temperature measured (see the inset of Fig. 1). Surprisingly, upon further annealing in AH (sample D), the susceptibility did not decrease any more. Instead, both the susceptibility amplitude and the transition width are nearly recovered to the level of the as-prepared sample (sample A). Finally if the sample D is annealed in AH longer, the susceptibility decreases again (sample E). But different from the behavior of sample C, the susceptibility of sample E gradually increases with decreasing temperature, approaching saturation smoothly at low temperatures. Fig. 2 shows the change in resistivity of different samples with temperature. Every sample exhibits a resistivity peak at its respective temperature T Pi (i ¼ A2E). Within the whole range of temperature we investigated, the resistivity is increased with annealing time in AH. From sample A to E, the peak resistivity increased by more than four orders. For as-prepared sample A, the insulator– metal transition temperature T PA ¼ 350:3 K is very close to its paramagnetic–ferromagnetic

103

E

Resistivity (Ωcm)

102 D 101 C 100 B 10-1

10-2

A 100

200 300 Temperature (K)

400

Fig. 2. Variation of resistivity with temperature for samples A–E. The insulator–metal transition temperature is marked by a short vertical line.

23

transition temperature T CA ; consistent with the double exchange mechanism. The striking result is that not following the initial decreasing tendency from sample A to C, the insulator–metal transition temperature increased from sample C to D with subsequent annealing in AH, in accordance with the AC magnetic susceptibility measurements shown in Fig. 1. Apparently, annealing the sample in AH would reduce the oxygen content as well as the ratio of Mn4+ and Mn3+, which leads to the decrease in T C (T P ) as well as AC magnetic susceptibility and increase in resistivity monotonically as previously reported [1], but in contrast with our observations. The absence of the signal near 3700 cm1 on the Raman spectra of La0.67Ba0.33MnO3 (not shown here), corresponding to a polar hydroxyl [OH] bond stretching [4], precludes one possibility that during the annealing in AH, the hydrogen was incorporated into the lattice as observed in SrTiO3 crystals [9] and Pb(Zr,Ti)O3 ferroelectric films [4]. Another possibility is that upon continuous loss of oxygen during annealing in AH, the ratio of Mn2+ and Mn3+ increased after the ratio of Mn4+ and Mn3+ decreased to zero, which recovered the ferromagnetism, just as observed in the electron doped (La, Ce)MnO3 compound [10,11]. Unfortunately, there is little change in the Mn 2p1/2,3/2 XPS spectra of our La0.67Ba0.33MnO3 compound before and after annealing in AH as seen from Fig. 3. However, there appeared extra shoulders on the higher binding energy side of the main Ba 3d3/2,5/2 peaks. At present we could not index the extra peaks, yet they at least imply that something happened on the Ba ions in addition to the oxygen loss during the annealing in AH. In order to clarify this, following the same annealing process as La0.67Ba0.33MnO3 mentioned above we annealed La0.67Ca0.33MnO3 and La0.67Sr0.33MnO3 bulk samples in AH at 700 1C. It was found that both the T C and T P decreased monotonically with annealing time, which evidenced that our anomalous observations are related to the change of the state of Ba ions. For comparison, another as-prepared La0.67Ba0.33MnO3 sample was annealed in flowing N2 at 950 1C for 10 h with carbon powder as an oxygen getter [2] (sample F). We believe that the

ARTICLE IN PRESS Y. Cui et al. / Journal of Magnetism and Magnetic Materials 297 (2006) 21–25

24

La 3d

Ba 3d

Mn 2p

XPS

D D

A D

A

A 820

840

860

770

780

790

BE (eV)

Binding Energy (eV)

800

640

650

660

BE (eV)

Fig. 3. XPS spectra of La 3d, Ba 3d, and Mn 2p for samples A and D.

70 A 60 Susceptibility (arb. unit /g)

H 50

G F

40 30 20 E 10 0 100

200 300 Temperature (K)

400

Fig. 4. AC magnetic susceptibility vs. temperature for sample A, E, F, G, and H. The annealing history of all the samples was described in the text.

main consequence of the nitrogen annealing is the loss of oxygen [1]. Then sample F together with sample E was annealed in flowing oxygen at 300 1C for 4 h. Thus sample G and sample H were obtained, respectively. The temperature dependence of AC susceptibility for samples G and H was plotted in Fig. 4. It is interesting to note that both the paramagnetic–ferromagnetic transition temperature and the amplitude of AC susceptibility for sample H are nearly completely recovered to the values of the as-prepared sample after oxygen annealing at such a low temperature of

300 1C. But there is only a little change in AC susceptibility of sample G compared to sample F, and the difference in AC susceptibility between sample G and the as-prepared sample is clear after the same oxygen annealing process, which further proved that the effect of annealing in AH is more than oxygen reduction, and other effects should be considered together. We ascribed the anomalous behavior of La0.67Ba0.33MnO3 sample during annealing in AH to the deficiency of Ba ions introduced by the segregation of some unknown Ba-rich phase. At the initial stage of the annealing in AH, the effect of oxygen loss is dominant. As the result, the reduction of Mn4+ content causes the decrease in the AC susceptibility and the increase in resistivity (from sample A to C as shown in Figs. 1 and 2) [1]. The oxygen diffusion out of the different stoichiometric well-crystallized La0.67Ba0.33MnO3 grains is very slow, so that the distribution of oxygen content inter- and intra- a grain is very inhomogeneous and the oxygen content is also grain dependent. This may be the reason for the existence of several stages on the AC susceptibility vs. temperature curves (samples B and C in Fig. 1). With more and more loss of oxygen due to annealing in AH, the increasing instability of the active Ba ions in La0.67Ba0.33MnO3 compound drives the segregation of an unknown Ba-rich phase on the surface of the grain or at the grain boundary. The Ba ion deficiency induced increase in Mn4+ content compensates for the decrease in Mn4+ content due to the oxygen loss, strengthening the ferromagnetism and metallicity of the sample (sample D in Figs. 1 and 2). With the aid of

ARTICLE IN PRESS Y. Cui et al. / Journal of Magnetism and Magnetic Materials 297 (2006) 21–25

the Ba vacancies, both the out-diffusion and the in-diffusion of the oxygen might become faster (sample H in Fig. 4). The increasing Ba ion vacancies and the segregated impurity phase enhanced the scattering of the charge carriers, resulting in the increase in the resistivity for sample D. Perhaps either the size or the volume of the segregated phase is so small that the XRD did not detect the impurity within its capability.

4. Conclusion In summary, the annealing effects of the La0.67Ba0.33MnO3 bulk ceramic samples in flowing AH were studied. Both the amplitude of the AC susceptibility and the insulator–metal transition temperature T P were not decreased monotonically with annealing time. The anomalous behavior is explained by the combined effects of the Ba ion vacancies resulting from the segregation of small sized unknown Ba-rich phase and the oxygen loss.

25

Acknowledgment We acknowledge the financial support from National 973 Project and Tsinghua University Natural Science Foundation.

References [1] H.L. Ju, et al., Phys. Rev. B 51 (1995) 6143. [2] W. Cai, et al., Phys. Stat. Sol. (a) 187 (2001) 529. [3] L. Ranno, et al., J. Phys.: Condens. Matter 8 (1996) L33. [4] S. Aggarwal, et al., Appl. Phys. Lett. 73 (1998) 1973. [5] K. Xiong, et al., Appl. Phys. Lett. 85 (2004) 2577. [6] H. Miki, et al., Jpn. J. Appl. Phys. 36 (Part 1) (1997) 1132. [7] T. Sakado, et al., Jpn. J. Appl. Phys. 37 (Part 1) (1998) 2565. [8] K. Takata, Appl. Phys. Lett. 78 (2001) 664. [9] S. Kapphan, et al., Ferroelectrics 25 (1980) 585. [10] P. Mandal, S. Das, Phys. Rev. B 56 (1997) 15073. [11] C. Mitra, et al., Phys. Rev. B 67 (2003) 092404.