Preparation, Phase Stability and Characterization of Superconductive ...

CHINESE JOURNAL OF PHYSICS

VOL. 43, NO. 3-II

JUNE 2005

Preparation, Phase Stability and Characterization of Superconductive Cobalt Oxyhydrates C.-J. Liu,1, ∗ C.-Y. Liao,1 L.-C. Huang,1 C.-H. Su,1 W.-C. Hung,1 S. Neeleshwar,2 Y.-Y. Chen,2 and Chia-Jung Charlie Liu3 1

Department of Physics, National Changhua University of Education, Changhua 500, Taiwan, R.O.C. 2 Institute of Physics, Academia Sinica, Taipei, Taiwan, R.O.C. 3 Department of Chemistry, Medical Technology, and Physics, Monmouth University, West Long Branch, NJ 07764-1898, U.S.A. We have first used the aquesous KMnO4 solution to de-intercate and oxidize γ-phase of Na0.7 CoO2 and successfully form the superconducting phase cobalt oxyhydrates with onset Tc = 3.4 - 4.6 K. It is found that the phase stability of superconducting cobalt oxyhydrate phase is unstable with respect to the ambient humidity and vacuum but is reversible in terms of phase crystallinity. According to the thermogravimetric analyses (TGA) and x-ray diffraction (XRD) patterns, two relatively stable phases can be identified at relatively low temperatures as a result of dehydration of the cobalt oxyhydrate. In this paper, the phase stability of Nax CoO2 ·yH2 O obtained using Br2 in acetonitrile solution is also discussed. PACS numbers: 81.20.-n, 74.62.Bf, 74.70.-b

I. INTRODUCTION

The recent discovery of superconductivity [1] in hexagonal cobalt oxyhydrates Nax CoO2 ·yH2 O with Tc = 4–5 K has attracted great attention [2–6]. The discovery of CoO2 -based superconductors is a breakthrough in searching for novel non-cuprate superconductors. The parent material Na0.7 CoO2 is a potential thermoelectric material with large thermopower and high electrical conductivity [7]. After immersing the parent material in Br2 acetonitrile solution and washing with de-ionized water, the Na+ partially de-intercalate from the structure and consequently the water molecule is inserted between the CoO2 layer and the Na layer, resulting an elongation of the c-axis to ∼ 19.6 ˚ A, a key factor of enhancing 2D character for the superconductivity to occur [1]. There are four phases [8] in the related cobalt oxides: (1) α-Nax CoO2 (0.9 ≤ x ≤ 1, O3 phase); (2) α’-Na0.75 CoO2 (O’3 phase); (3)β-Nax CoO2 (0.55 ≤ x ≤ 0.6, P3 phase); and (4)γ-Nax Coy O2 (0.55 ≤ x/y ≤ 0.74, P2 phase). Na0.7 CoO2 belongs to the category of γ-phase. The P2 phase represents two CoO2 layers in the unit cell with the Na+ in the trigonal prismatic surroundings. The process of immersing γ-Na0.7 CoO2 in Br2 /CH3 CN solution followed by filtering and rinsing is generally considered as a chemical oxidation. In this paper, we report a novel route of preparing the superconducting cobalt oxyhydrates using aqueous KMnO4 solution as a de-intercalating and oxidizing agent. It is an alternative route to use oxidant KMno4

http://PSROC.phys.ntu.edu.tw/cjp

547

c 2005 THE PHYSICAL SOCIETY

OF THE REPUBLIC OF CHINA

548

PREPARATION, PHASE STABILITY AND CHARACTERIZATION . . .

VOL. 43

instead of using toxic Br2 /CH3 CN solution to form the superconducting phase of cobalt oxyhydrates. We also present a detailed study on the phase stability of cobalt oxyhydrates and find that the phase stability of KMnO4 -prepared samples seem to be more robust than the Br2 -prepared ones.

II. EXPERIMENTAL

We have used a rapid heat-up procedure [7] to make the parent material γ-Na0.7 CoO2 to avoid the evaporation of Na in the heating process. The procedure to make the cobalt oxyhydrates is quite straight forward, being pretty much like a one-pot reaction. We first prepare the desired concentration of aqueous KMnO4 solution by dissolving the KMnO4 in 50 - 680 ml of de-ionized water. The γ-Na0.7 CoO2 powders of 0.5–1 g were then immersed and stirred in aqueous KMnO4 solution or Br2 /CH3 CN solution at room temperature for 5 days. The products were filtered and washed with de-ionized water several times. The powders were then stored in a chamber with a relative humidity of 98 % to avoid the loss of water content. This procedure has the advantage of mass production of superconductive cobalt oxyhydrates without resort to the highly toxic acetonitrile solution of bromine. Powder X-ray diffraction (XRD) patterns were obtained using a Shimadzu XRD-6000 diffractometer equipped with Fe Kα radiation. Thermogravimetric analysis (TGA) was carried out by using a Perkin Elmer Pyris 1 thermogravimetric analyzer in flowing O2 at the slowest rate of 0.1◦ C/min available to this instrument. Electrical resistivity as a function of temperature was measured using standard dc four-probe techniques. A commercial SQUID magnetometer (Quantum Design) was used to characterize the superconducting transition temperature of the samples.

III. RESULTS AND DISCUSSION

Fig. 1 shows the powder x-ray diffraction patterns of cobalt oxyhydrates obtained by immersing the parent material γ-Na0.7 CoO2 in aqueous KMnO4 solution with different molar ratios of KMnO4 relative to the sodium. All the reflection peaks are indexable based on a hexagonal lattice with space group P63 /mmc. The (002) reflection occurring at 2θ ≈ 11.3◦ (Fe Kα radiation) corresponds to the fully hydrated phase of c ≈ 19.6 ˚ A, being consistent with that for the superconducting Nax CoO2 ·yH2 O phase obtained by the Br2 /CH3 CN treatment and indicating the successful intercalation of water between the sodium and CoO2 layers. Note that the as-prepared 0.05X and 0.1X samples are a mixture A, which A and a non-hydrate phase of c ≈ 11.2 ˚ of a superconductive phase of c ≈ 19.6 ˚ will be discussed later. Fig. 2 shows the time evolution of XRD patterns for 0.4X sample. The broader and smaller peak intensity with time indicates the nature of its phase instability with respect to

VOL. 43

C.-J. LIU, C.-Y. LIAO, et al.

549

FIG. 1: The XRD patterns of fully hydrated phase of cobalt oxyhydrates obtained by treating the parent material γ-Na0.7 CoO2 in aqueous KMnO4 solution with different molar ratios of KMnO4 /Na. The 0.3X sample refers to KMnO4 /Na = 0.3.

the ambient environment. The TGA data (not shown) also indicate a slow weight decrease as a result of gradual loss of water in the ambient environment. This phase instability behavior of KMnO4 -prepared cobalt oxyhydrates is similar to that of Br2 /CH3 CN-prepared sample reported by Foo et al. [4]. It seems that the KMnO4 -prepared sample is more robust than the Br2 /CH3 CN-prepared one because it can keep the c ≈ 19.6 ˚ A phase for at least three days, whereas the Br2 /CH3 CN-prepared one would readily transform in a few hours to the intermediate phase of c ≈ 13.8 ˚ A with the characteristic (002) reflection at 2θ ≈ 16◦ (see Fig. 3). For resistivity measurements, it usually requires pressing the sample for characterization. It should be noted that the fully hydrated phase in the pressed pellet form could transform to the c ≈ 13.8 ˚ A phase more easily than the powder form. As shown in Fig. 3, the intermediate phase of c ≈ 13.8 ˚ A and non-hydrate phase of c ≈ ˚ 11.2 A can be achieved by heating at relatively low temperatures, which is very helpful in studying the 2D effects on their physical properties because all three phases have the same Na content and little change in the a-axis. Even though with the phase instability of the fully hydrated phase , the losing and gaining process of water is reversible and the phase crystallinity can be restored by storing the sample back in a chamber with sufficient humidity (see Fig. 2).

550


VOL. 43

FIG. 2: The time evolution of XRD patterns for the 0.4X sample. The reflection peaks become smaller and broader with time in the ambient environment, indicating phase crystallinity deteriorates with time, but this c ≈ 19.6 ˚ A phase persists at least for 3 days and good phase can be restored by storing the deteriorated sample back in a wet chamber.

FIG. 3: The time evolution of XRD patterns for the Br2 /CH3 CN-prepared cobalt oxyhydrate. In a few hours, the fully hydrated phase readily converts to the lower hydrate phase of c ≈ 13.8 ˚ A with the characteristic (002) reflection at 2θ ≈ 16◦

The fully hydrated phase is also unstable under vacuum. As shown in Fig. 4, the c ≈ 13.8 ˚ A phase gradually evolves with disappearance of the c ≈ 19.6 ˚ A phase upon evacuation. This result implies that one has to take precaution when characterization of the physical properties on this material is carrier out using the closed-cycle refrigerators which usually require evacuating the sample chamber. Due to the phase instability, it makes one

VOL. 43


551

FIG. 4: The evolution of XRD patterns for the Br2 /CH3 CN-prepared Nax CoO2 ·yH2 O under vacuum. The c ≈ 19.6 ˚ A phase crystallinity gradually deteriorates accompanied by the appearance of the c ≈ 13.8 ˚ A phase.

˚ phase after low T measurements. wonder whether the material still keep the c ≈ 19.6 A Fig. 5 shows the XRD patterns of one of the Br2 /CH3 CN-prepared samples before and after resistivity measurements. The c ≈ 19.6 ˚ A phase has obviously transformed to the c ≈ ˚ 13.8 A phase after resistivity measurements. To improve this situation, we have designed a method by sealing the resistivity measuring device with water-containing crucible in it to protect the sample from losing water. As shown in Fig. 6, it turns out quite successful in keeping the c ≈ 19.6 ˚ A phase. Fig. 7 shows the temperature dependence of magnetization for both the Br2 /CH3 CNprepared and KMnO4 -prepared (0.3X) samples. Both samples show a onset superconducting transition temperature Tc,onset , defined as the magnetization beginning to decrease in magnitude, is approximately 4.6 K. But the diamagnetic signal occurs at a slightly lower temperature for Br2 /CH3 CN-prepared sample (∼ 3.7 K) than the 0.3X one (∼ 4.1 K). Fig. 8 shows the temperature dependence of magnetization for the 0.05X sample. The onset Tc is 3.4 K and 4.5 K for the as-prepared powders and the same sample but stored back in a wet chamber for 10 days, respectively. This result is consistent with the powder XRD results shown in Fig. 9. The as-prepared powders of 0.05X sample exists in two phases, i.e., the fully hydrated phase of c ≈ 19.6 ˚ A and the non-hydrate phase of c ≈ 11.2 ˚ A. After a hydration process by storing the mixed phase of 0.05X sample in a wet chamber for 4

552


VOL. 43

FIG. 5: The XRD patterns for the Br2 /CH3 CN-prepared Nax CoO2 ·yH2 O before and after resistivity measurements. The c ≈ 19.6 ˚ A phase has converted to the c ≈ 13.8 ˚ A phase after measurements.

FIG. 6: The XRD patterns for the fully-hydrated cobalt oxyhydrates before and after resistivity measurements. The c ≈ 19.6 ˚ A phase still remains after measurements.

˚ phase grows at the expense of c ≈ 11.2 ˚ days, the superconducting c ≈ 19.6 A A phase. The shifting of the onset Tc should be a consequence of the change in the weighting of magnetic contribution from the nonsuperconducting dehydrate phase of c ≈ 11.2 ˚ A. Note that the hydration process is not expected to change the sodium content of the 0.05X sample. It has been a controversial issue whether the Tc of the fully hydrated phase of

VOL. 43


553

FIG. 7: Zero-field cooled dc magnetization of the fully-hydrated Nax CoO2 ·yH2 O. The applied field is 20 Oe and 50 Oe for samples prepared by aqueous KMnO4 and Br2 /CH3 CN solution, respectively.

FIG. 8: Zero-field cooled dc magnetization of the KMnO4 -prepared 0.05X sample. The onset Tc shifts to ∼ 4.5 K after the as-prepared powders are stored in a wet chamber for 10 days, however, the superconducting volume fraction is still much less that the 0.3X sample.

Nax CoO2 ·yH2 O is associated with the sodium content [2, 9]. Our results on the 0.05X sample might shed some light on this controversial issue, which will be pursued further.

554


VOL. 43

˚ FIG. 9: The XRD patterns for the KMnO4 -prepared 0.05X sample. The superconducting c ≈ 19.6 A phase grows at the expense of the nonsuperconducting c ≈ 11.2 ˚ A phase as the as-prepared powders are stored in a wet chamber for 4 days. IV. SUMMARY

In summary, We have first used the aqueous KMnO4 solution to prepare superconductive cobalt oxyhydrates instead of using Br2 in acetonitrile solution. The phase instability of superconductive cobalt oxyhydrates suggests the necessity of carefully handling the sample for accurate characterization, particularly those need to pump the sample down during measurements. It is suggested that the phase crystallinity should be checked before and after physical measurements to make sure the sample is still in a good condition.

Acknowledgments This work is supported by the National Science Council of ROC, grant no. NSC 92-2112-M-018-005.

References Electronic address: [email protected] [1] K. Takada, H., Sakurai, E. Takayama-Muromachi, F. Izumi, R. A. Dilanian, and T. Sasaki, Nature 422, 53 (2003). [2] R. E. Schaak, T. Klimczuk, M. L. Foo, and R. J. Cava, Nature 424, 527 (2003). [3] J. Cmaidalka, A. Baikalov, Y. Y. Xue, R. L. Meng, and C. W. Chu, Physica C 403, 125 (2004). [4] M. L. Foo, R. E. Schaak, V. L. Miller, T. Klimczuk, N. S. Rogado, Y. Wang, G. C. Lau, C. Craley, H. W. Zandbergen, N. P. Ong, and R. J. Cava, Solid State Commun. 127, 33 (2003). ∗

VOL. 43


555

[5] H. D. Yang, J.-Y. Lin, C. P. Sun, Y. C. Kang, K. Takada, T. Sasaki, H. Sakurai, and E. Takayama-Muromachi, cond-mat/0308031 (2003). [6] C.-J. Liu, C.-Y. Liao, L.-C. Huang, C.-H. Su, S. Neeleshwar, Y.-Y. Chen, and C.-J. C. Liu, Physica C 416, 43 (2004). [7] T. Motohashi, E. Naujalis, R. Ueda, K. Isawa, M. Karppinen, and H. Yamauchi, Appl. Phys. Lett. 79, 1480 (2001). [8] M. Mikami, M. Yoshimura, Y. Mori, and T. Sasaki, Jpn. J. Appl. Phys. 41, L777 (2002). [9] C. J. Milne, D. N. Argyriou, A. Chemseddine, N. Aliouane, J. Veira, and D. Alber, condmat/0401273. [10] W. Koshibae, K. Ysutsui, and S. Maekawa, Phys. Rev. B 62, 6869 (2000). [11] W. Koshibae and S. Maekawa, Phys. Rev. Lett. 87, 236603 (2001).