Thermoelectric oxide NaCo2O4 nanofibers fabricated by electrospinning

Thermoelectric oxide NaCo2O4 nanofibers fabricated by electrospinning

Materials Chemistry and Physics 99 (2006) 104–108 Thermoelectric oxide NaCo2O4 nanofibers fabricated by electrospinning Santi Maensiri ∗ , Wiwat Nuan...

360KB Sizes 0 Downloads 44 Views

Materials Chemistry and Physics 99 (2006) 104–108

Thermoelectric oxide NaCo2O4 nanofibers fabricated by electrospinning Santi Maensiri ∗ , Wiwat Nuansing Department of Physics, Faculty of Science, Khon Kaen University, Khon Kaen 40002, Thailand Received 22 May 2005; received in revised form 2 September 2005; accepted 4 October 2005

Abstract For the first time, sodium cobalt oxide (NaCo2 O4 ) nanofibers with diameter of ∼20–200 nm were prepared by electrospinning a precursor mixture of sodium acetate/cobalt acetate/PAN, followed by calcination treatment of the electrospun composite nanofibers. The sodium cobalt oxide nanofibers were characterized by TG–DTA, X-ray diffraction (XRD), Raman spectroscopy, scanning electron microscopy (SEM), and transmission electron microscopy (TEM). The NaCo2 O4 nanofibers calcined at 300 and 400 ◦ C were polycrystalline of ␥-Nax Co2 O4 phase having diameters of ∼20–60 nm with grain sizes of ∼5–10 nm, and whereas the nanofibers calcined at 800 ◦ C were single crystal having linked particles or crystallites with particle sizes of ∼60–200 nm. The results indicated a significant effect of calcination temperature on the crystalline phase and morphology of the nanofibers. © 2005 Elsevier B.V. All rights reserved. Keywords: Nanofiber; Electrospinning; Raman spectroscopy and scattering; Electron microscopy

1. Introduction Recently, there has been great interest in developing novel thermoelectric materials for thermoelectric energy conversion. This is widely recognized as a promising technology for both power generation in terms of waste heat recovery and cooling of various electronic devices employed in high technology [1,2]. Stability at high temperature is one of the required properties of such materials. Thermoelectric oxides have been regarded as potential candidates because, unlike conventional materials such as Bi2 Te3 , PbTe, and Si–Ge alloy, they are chemically and thermally stable and thus can be used at high temperatures without deterioration of their performance due to oxidation. Moreover, their production costs are comparatively low. Sodium–cobalt oxide, Nax Co2 O4 has been revealed to have potential as a thermoelectric material used in energy conversion and electronic devices. In general Nax Co2 O4 thermoelectric oxide shows three types of crystal structure depending on the x value; P3 type (␤phase, 1.1 ≤ x ≤ 1.2), P2 type (␥-phase, 1.0 ≤ x ≤ 1.4) and O3 type (␣-phase, 1.8 ≤ x ≤ 2.0) [3,4]. The fundamental structure of Nax Co2 O4 consists of Na and a CdI2 -type CoO2 sheet alternately stacked along the c-axis. The CoO2 is responsible for the electric conduction, whereas an insulating Na layer works as a



Corresponding author. Tel.: +66 43 202222–9x2248; fax: +66 43 202374. E-mail address: [email protected] (S. Maensiri).

0254-0584/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2005.10.004

charge reservoir to stabilize the crystal structure. Great attention has been paid to the P2 type ␥-Nax Co2 O4 because of its excellent thermoelectric properties at high temperature. A Nax Co2 O4 single crystal (x = 1) shows large in-plane thermoelectric power (S) and electrical conductivity (σ), and thus attains a figure of merit (Z = α2 σ/κ) comparable to that of Bi2 Te3 , where α is the Seebeck coefficient and κ the thermal conductivity [5,6]. Nevertheless, it is costly and difficult to prepare as large single crystals. Terasaki [7] reported that polycrystalline Nax Co2 O4 (x = 1) exhibits simultaneously high thermoelectric power and low resistivity (S ∼ 100 ␮V K−1 and ρ ∼200 ␮ cm) at 300 K. These thermoelectric properties are as good as those of the single crystal. Since then, Nax Co2 O4 polycrystalline ceramics have become hot spots in thermoelectric oxide materials research [8–14]. In recent years, one-dimensional (1D) nanostructured materials such as nanorods, nanowires or nanofibers have attracted much attention due to their importance for scientific interest and their significant potential to be used as nanodevices [15,16]. For thermoelectric applications, nanofibers of thermoelectric oxides will open the door to the exploration of a range of intriguing properties and applications associated with their 1D nanostructures. To the best of our knowledge, there have been no reports concerning the fabrication of NaCo2 O4 nanofibers. In this communication, we demonstrate for the first time the fabrication of thermoelectric oxide NaCo2 O4 nanofibers by calcination of the electrospun sodium acetate/cobalt acetate/polyacrylonitrile

S. Maensiri, W. Nuansing / Materials Chemistry and Physics 99 (2006) 104–108

105

(PAN) composite nanofibers. Electrospinning represents a simple and convenient method for preparing polymer fibers and ceramic fibers with both solid and hollow interiors that are exceptionally long in length, uniform in diameter ranging from several micrometers down to tens of nanometers, depending on the polymer and processing conditions [17–20]. 2. Experimental details 2.1. Materials preparation The starting chemicals used in this study were polyacrylonitrile (PAN) powder (Mw = 86 200, Mn = 22 600, Aldrich), N,N-dimethyl formamide (DMF) (Riedel, 99.8%), sodium acetate trihydrate (Riedel, 99–100%) and cobalt (III) acetate tetrahydrate (Fluka, >99%). A PAN solution of about 6.4 wt.% was first prepared by dissolving PAN powder in DMF and stirring for 2 h. Note that PAN solution had been used to prepare carbon-based nanofibers [21–24] and it was chosen here as a polymer template for NaCo2 O4 because it could be easily electrospun providing good nanofibers. In the preparation of a sodium acetate/cobalt acetate/PAN solution, 54.62 mg of sodium acetate trihydrate and 200 mg of cobalt (III) acetate tetrahydrate were dissolved in 4 ml of the prepared PAN solution. All the mixtures were vigorously stirred at room temperature before they became a homogeneous polymer solution. The electrospinning process was performed using the above mentioned polymer solutions: PAN solution and sodium acetate/cobalt acetate/PAN solution. The polymer solution was loaded into a plastic syringe equipped with a 23-gauge needle made of stainless steel. The needle was connected to a high-voltage supply (DEL High Voltage (0–100 kV), DEL Electronics Corp., USA). The solution was fed at a rate of 1.0 mL h−1 using a syringe pump (TERUMO Terufusion Syringe pump TE-331, Japan). A piece of flat aluminum foil was placed 10 cm below the tip of the needle, and used to collect the nanofibers. The voltage for electrospinning was 20 kV. The process was carried out at room temperature. The bulk sample of polycrystalline NaCo2 O4 used for the comparison through this article was fabricated by the solid-state reaction using Na2 CO3 (Fluka, 99%) and Co3 O4 (Nanostructured & Amorphous Materials, 99.5%) powders as starting materials. The mixed powders were calcined at 860 ◦ C for 12 h, and sintering of the disc compact was carried out in air at 920 ◦ C for 12 h.

2.2. Materials characterization The as-spun sodium acetate/cobalt acetate/PAN composite nanofibers were subjected to thermogravimetric–differential thermal analysis (TG–DTA) using Pyris Diamond TG/DTA (Perkin Elmer Instrument, USA). This analysis was to determine the temperatures of possible decomposition and crystallization (or phase changes) of the nanofibers. The analyses were performed with a heating rate of 15 ◦ C min−1 in static air up to 1000 ◦ C. The as-spun sodium acetate, cobalt acetate/PAN composite nanofibers were calcined at 300, 400, and 800 ◦ C for 5 h in air in a box furnace (Lenton Furnaces, UK), using heating and cooling rates of 5 min ◦ C−1 . The characterization of the as-spun composite nanofibers and calcined nanofibers were assessed by X-ray diffraction (XRD) (PW3710 mpd control, The Netherlands), Raman spectroscopy (Jobin Yvon/Atago-Bussan T64000, France), scanning electron microscopy (SEM) (LEO 1450VP, UK), and transmission electron microscopy (TEM) (JEOL JEM 2010).

Fig. 1. TG–DTA curves of the electrospun sodium acetate/cobalt acetate/PAN composite fibers at a heating rate of 15 ◦ C min−1 in static air.

on the polymer side chain, which was confirmed by a dramatic weight loss in the TG curve over the corresponding temperature range. The decomposition of the main chain of PAN may result in the peak at 320 ◦ C. It is clear from the TG curve that all the PAN and organic groups of sodium acetate and cobalt acetate had been removed almost completely at 350 ◦ C, resulting in the metal oxide phase of NaCo2 O4 . This agrees with the XRD and Raman results, shown in Figs. 2 and 3, respectively. Fig. 2 shows the X-ray diffraction patterns of the calcined sodium acetate/cobalt acetate/PAN composite nanofibers and the sintered sample of NaCo2 O4 . By the comparison of the peaks to those of the powder diffraction pattern of NaCo2 O4 in the JCPDS card No. 27-0682 shows they were indexed as the ␥phase, although a small amount of unreacted Co3 O4 and some unknown phases were observed. Na content of the starting composition with x > 1.4 was necessary to obtain Nax Co2 O4 single phase free from impurities [10]. The precipitation of the Co3 O4 in the current study is therefore considered to be due to Na evaporation during the calcination. The results of Raman spectroscopy (Fig. 3) clearly indicate that NaCo2 O4 crystalline phase was formed after calcination of the as-spun sodium acetate/cobalt acetate/PAN composite nanofibers even at 300 ◦ C for 5 h. Three main Raman peaks labeled (1), (2) and (3) were observed for all the calcined

3. Results and discussion The results of simultaneous TG and DTA analyses of the asspun sodium acetate/cobalt acetate/PAN composite nanofibers are shown in Fig. 1. The DTA curve depicted a shallow endothermic peak at 210 ◦ C, which could be attributed to the loss of moisture and trapped solvent (water, DMF and carbon dioxide). The sharp exothermic peak between 250 and 350 ◦ C could be attributed to the loss of crystal water and the decomposition of acetate along with the degradation of PAN by a dehydration

Fig. 2. XRD patterns of: (A) NaCo2 O4 ceramic sample prepared by solid-state reaction; (B) NaCo2 O4 fibers calcined at 300 ◦ C; (C) NaCo2 O4 fibers calcined at 400 ◦ C; (D) NaCo2 O4 fibers calcined at 800 ◦ C.

106

S. Maensiri, W. Nuansing / Materials Chemistry and Physics 99 (2006) 104–108

Fig. 3. Raman spectra of: (A) PAN-DMF fibers; (B) sodium acetate/cobalt acetate/PAN composite fibers; (C) NaCo2 O4 fibers calcined at 300 ◦ C; (D) NaCo2 O4 fibers calcined at 400 ◦ C; (E) NaCo2 O4 fibers calcined at 800 ◦ C; (F) NaCo2 O4 ceramic sample prepared by solid-state reaction.

composite nanofibers. These three peaks were assigned, respectively, as E2g at 478.1–482.4 and 520.9–522.2 cm−1 , A1g at 685.5–692.9 cm−1 , for the space group of P63 /mmc [25–28]. The nanofibers calcined at 300 ◦ C showed Raman peaks at 483.5, 522.2, and 692.2 cm−1 while Raman peaks at 482.4, 520.9, and 692.8 cm−1 were detected for the nanofibers calcined at 400 ◦ C. In the case of the nanofibers calcined at 800 ◦ C, there were four Raman peaks observed at 478.1, 520.9, 613.1 and 685.5 cm−1 . The extra peak at 613.1 cm−1 labeled (*) was assigned as E2g [25–28]. The Raman results are consistent with

the results obtained from the bulk NaCo2 O4 sample prepared by solid-state reaction, in which Raman peaks were observed at 486.6, 524.9, and 695.3 cm−1 . From Fig. 3, the Raman-active mode in the NaCo2 O4 corresponds to a frequency of ωR ∼ 685–695 cm−1 , which possibly arises from a symmetrical stretching mode of the Co–O vibrational unit. This mode is very sensitive to any disorder in the oxygen sublattice resulting from thermal and/or grain size-induced non-stoichiometry [29,30]. The influence of the microstructure of NaCo2 O4 nanofibers on the shape of the Raman spectrum was revealed by the increases in asymmetry and the line broadening. The former is attributed to reduction of the photon lifetime in the nanocrystalline regime, whereas the latter can be described by the dependence of its half width upon the grain size [29–32]. The broader the Raman-active mode, the smaller the grain size. Fig. 4 illustrates the non-woven nanofiber morphology revealed by scanning electron microscopy (SEM). Despite some noticeable beads, the surface of as-spun sodium acetate/cobalt acetate/PAN composite nanofibers appears quite smooth due to the amorphous nature of the sodium acetate/cobalt acetate/PAN composite nanofibers (Fig. 4A). The diameter of the as-spun composite nanofibers ranges from 100 to 200 nm. The NaCo2 O4 nanofibers obtained after calcination at 300 ◦ C (Fig. 4B) exhibited shrinkage which reduced diameters of the nanofibers to less than 100 nm. The decomposition of PAN and the removal of the CH3 COO group of sodium acetate and cobalt acetate molecules is attributed to nanofiber shrinkage. After calcination at 400 ◦ C (Fig. 4C), the diameters of the nanofibers became even smaller (∼30–80 nm). This was thought to be caused by the

Fig. 4. SEM micrographs of: (A) sodium acetate/cobalt acetate/PAN composite fibers; (B) NaCo2 O4 fibers calcined at 300 ◦ C; (C) NaCo2 O4 fibers calcined at 400 ◦ C; (D) NaCo2 O4 fibers calcined at 800 ◦ C; (E) a single aligned NaCo2 O4 fiber on Si substrate calcined at 400 ◦ C.

S. Maensiri, W. Nuansing / Materials Chemistry and Physics 99 (2006) 104–108

107

Fig. 5. TEM bright field images with corresponding selected-area electron diffraction (SAED) of: (A) NaCo2 O4 nanofibers calcined at 300 ◦ C; (B) NaCo2 O4 nanofibers calcined at 400 ◦ C; (C) NaCo2 O4 nanofibers calcined at 800 ◦ C.

complete removal of organic molecules and the development of NaCo2 O4 crystals. Interestingly, the nanofibers formed a structure of linked particles or crystallites after calcination at 800 ◦ C (Fig. 4D). These changes in the morphology are related to a dramatic change in crystal structure as observed in electrospun vanadium pentoxide nanofibers [33]. In addition, Li et al. method [34] has been applied for the preparation of aligned nanofibers which could be collected over a gap formed between two strips of silicon substrate. They could be transferred onto other substrates for various applications. The SEM micrograph of a single aligned sodium acetate/cobalt acetate/PAN composite nanofibers, calcined in air for 5 h at 400 ◦ C, is shown in Fig. 4E. The success of preparing aligned thermoelectric oxide nanofibers suggests new opportunities in exploration of a range of unique properties (e.g. electrical, thermal, and mechanical properties) and applications associated with their 1D nanostructures. The morphology and structure of the calcined sodium acetate/cobalt acetate/PAN composite nanofibers were further investigated by transmission electron microscopy (TEM). Fig. 5 shows TEM bright field images with corresponding selectedarea electron diffraction (SAED) of the composite nanofibers calcined in air for 5 h at different temperatures. The nanofibers calcined at 300 ◦ C were ∼60 nm in diameter with a small grain size of <5 nm (Fig. 5A) while smaller diameter of ∼20 nm with larger grain size of ∼5–10 nm was obtained for the nanofibers

calcined at 400 ◦ C (Fig. 5B). The diffraction patterns of the two nanofibers showed ring patterns, revealing their nanocrystalline structure. According to the diffraction patterns in Fig. 5A and B, lattice constants obtained from the diffraction images coincide with those of polycrystalline ␥-Nax Co2 O4 (JCPDS card No. 270682): the lattice constants measured agree with those of (0 0 2), (0 0 4), (1 0 0), (1 0 3), (1 0 6), and (1 1 0) planes. Fig. 5C clearly shows linked particles or crystallites in the nanofibers calcined at 800 ◦ C. The sizes of particles varied from 60 to 200 nm. The diffraction pattern of these nanofibers (insert in Fig. 5C) shows a strong superstructure, revealing their single crystalline structure. Insightful details of an electron diffraction study of Nax Co2 O4 have been recently reported elsewhere [27,35], and was not further investigated in our study. 4. Conclusion Nanofibers of thermoelectric oxide NaCo2 O4 with diameters of ∼20–200 nm were prepared by using the electrospun sodium acetate/cobalt acetate/PAN composite nanofibers as precursor through calcination treatment. The XRD, Raman spectroscopy and SAED results suggested the formation of ␥-Nax Co2 O4 in the sodium acetate/cobalt acetate/PAN composite nanofibers calcined at 300–800 ◦ C. Accomplishing of this work has implied the possibility to electrospin NaCo2 O4 thermoelectric oxide into non-woven nanofiber mats and uniaxially aligned nanofibers.

108

S. Maensiri, W. Nuansing / Materials Chemistry and Physics 99 (2006) 104–108

We believe these electrospun nanofibers may have potential applications in thermoelectric nanodevices. The thermoelectric properties (electrical and thermal properties) of the electrospun NaCo2 O4 nanofibers are under investigation. Acknowledgements The authors would like to thank the Department of Chemistry for providing TG–DTA facilities, the Department of Biology for providing SEM facilities, and the Science and Technology Service Center (Chiang Mai University) for providing TEM facilities. We thank Mr. Prasit Thongbai for providing the bulk NaCo2 O4 ceramic sample. This work is supported by the Research Division, Khon Kaen University. References [1] D.M. Rowe, CRC Handbook of Thermoelectrics, CRC Press LLC, Florida, 1995. [2] G.S. Nolas, J. Sharp, H.T. Goldsmid, Thermoelectrics, in: Basic Principles and New Materials Developments, Springer-Verlag, Berlin, 2001. [3] C. Fouassier, G. Matejka, J.-M. Reau, P. Hagenmuller, J. Solid State Chem. 6 (1973) 532. [4] M. Jansen, R. Hoppe, Z. Anorg. Allg. Chem. 408 (1974) 104. [5] I. Terasaki, Y. Sasago, K. Uchinokura, Phys. Rev. B 56 (1997) R12685. [6] H. Yakabe, K. Kikuchi, I. Terasaki, Y. Sasago, K. Uchinokura, Proceedings of the 16th International Conference on Thermoelectrics, 1997, IEEE Piscataway, 1998, p. 523. [7] I. Terasaki, Proceedings of the 18th International Conference on Thermoelectrics, 1999, IEEE Piscataway, 2000, p. 569. [8] T. Katawa, Y. Iguchi, T. Itoh, K. Takahara, Phys. Rev. B 60 (1999) 10584. [9] S. Tajima, T. Tani, S. Isobe, K. Koumoto, Mater. Sci. Eng. B 86 (2001) 20. [10] W. Shin, N. Izu, I. Matsubata, N. Murayama, Jpn. J. Appl. Phys. 42 (2003) 7013.

[11] M. Ito, T. Nagira, D. Furumoto, S. Katsuyama, H. Nagai, Scripta Mater. 48 (2003) 403. [12] G. Peleckis, T. Motohashi, M. Karppinen, H. Yamauchi, Appl. Phys. Lett. 83 (2003) 5416. [13] A. Satake, H. Tanaka, T. Ohkawa, T. Fujii, I. Terasaki, Appl. Phys. Lett. 96 (2004) 931. [14] D. Prabhakaran, A.T. Boothroyd, R. Coldea, N.R. Charnley, J. Cryst. Growth 271 (2004) 74. [15] J. Hu, M.Y. Ouyang, C.M. Lieber, Nature 399 (1999) 48. [16] J.R. Heath, P.J. Kuekes, R.S. Williams, Science 280 (1998) 717. [17] D.H. Renaker, I. Chun, Nanotechnology 7 (1996) 216. [18] A. Frenot, I.S. Chronakis, Curr. Opin. Colloid Interf. Sci. 8 (2003) 64. [19] Z.H. Huang, Y.Z. Zhang, M. Kotaki, S. Ramakrishna, Compos. Sci. Technol. 63 (2003) 2223. [20] D. Li, Y. Xia, Adv. Mater. 16 (2004) 1151. [21] Y. Wang, S. Serrano, J.J. Santiago-Aviles, Synth. Met. 138 (2003) 423. [22] H. Hou, D.H. Reneker, Adv. Mater. 16 (2004) 69. [23] S.F. Fennessey, R.J. Farris, Polymer 45 (2004) 4217. [24] C. Kim, S.-H. Park, J.-I. Cho, D.-Y. Lee, T.-J. Park, W.-J. Lee, K.-S. Yang, J. Raman Spectrosc. 35 (2004) 928. [25] M.N. Ilieve, A.P. Litvinchuk, R.L. Meng, Y.Y. Sun, J. Cmaidalka, C.W. Chu, Physica C 402 (2004) 239. [26] Y.G. Shi, Y.L. Liu, H.X. Yang, C.J. Nie, R. Jin, J.Q. Li, Phys. Rev. B 70 (2004) 052502. [27] H.X. Yang, C.J. Nie, Y.G. Shi, H.C. Yu, S. Ding, Y.L. Liu, D. Wu, N.L. Wang, J.Q. Li, Solid State Commun. 134 (2005) 403. [28] K. Takada, M. Osada, F. Izumi, H. Sakurai, E. Takayama-Muromachi, T. Sasaki, Chem. Mater. 17 (2005) 2034. [29] H. Richter, Z.P. Wang, L. Ley, Solid State Commun. 39 (1981) 625. [30] P. Parayanthal, F.H. Pollak, Phys. Rev. Lett. 52 (1984) 1822. [31] M. Fuji, S. Hayashi, K. Yamamoto, Appl. Phys. Lett. 57 (1990) 2692. [32] W.H. Weber, K.C. Hass, J.R. McBride, Phys. Rev. B 48 (1993) 178. [33] P. Viswanathamurthi, N. Bhattarai, H.Y. Kim, D.R. Lee, Scripta Mater. 49 (2003) 577. [34] D. Li, Y. Wang, Y. Xia, Nano Lett. 3 (2003) 1167. [35] H.W. Zandbergenm, M. Foo, Q. Xu, V. Kumar, R.J. Cava, Phys. Rev. B 70 (2004) 024101.