Effect of hydroxyl group on global and local structures of ...

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy Materials Chemistry and Physics 129 (2011) 1071–1074

Contents lists available at ScienceDirect

Materials Chemistry and Physics journal homepage: www.elsevier.com/locate/matchemphys

Effect of hydroxyl group on global and local structures of hydrothermally grown KNbO3 nanorods B.K. Yun a , Y.S. Koo a , J.H. Jung a,∗ , M. Song b , S. Yoon b a b

Department of Physics, Inha University, Incheon 402-751, Republic of Korea Division of Nano Sciences and Department of Physics, Ewha Womans University, Seoul 120-750, Republic of Korea

a r t i c l e

i n f o

Article history: Received 27 October 2010 Received in revised form 17 May 2011 Accepted 24 May 2011 Keywords: Nanostructures Hydroxyl group Raman scattering Interstitial defect

a b s t r a c t We report the annealing temperature dependence of global and local structures of hydrothermally grown KNbO3 nanorods. With increasing annealing temperature, the amount of hydroxyl groups in the KNbO3 nanorods decreased and finally disappeared at about 800 ◦ C. Morphology of the nanorods seemed not to change significantly, however, X-ray intensity ratio between (0 2 2) and (2 0 0) planes, i.e., I0 2 2 /I2 0 0 , increased with decreasing lattice hydroxyl group contents. This result could be attributed to the fact that the hydroxyl groups were desorbed more effectively along the [0 1 1] direction than in other directions due to the elongated nanorod along the [0 1 1] direction. The frequencies of external and bending modes, besides stretching mode, showed red-shift with decreasing lattice hydroxyl group contents. This result implied that the lattice hydroxyl groups existed as interstitial defects near K ions. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Ferroelectric nanostructures, such as nanoparticles, nanowires and nanorods, have drawn considerable interest due to their scientific as well as technological importance [1–3]. The size effect inherent in ferroelectric nanostructures can modify the magnitude and direction of spontaneous electric polarization, the Curie temperature and the crystal structure of their ferroelectric bulk counterparts [4,5]. A spontaneous electric polarization and large piezoelectricity inherent in ferroelectric nanostructures can be utilized for ultra-high density non-volatile memory, nano piezoelectric actuator and transducer device applications [6,7]. Among the ferroelectric nanocrystals, KNbO3 nanorods have recently become one of the most important materials due additionally to their non-toxicity [8–10]. Until now, hydrothermal method has been frequently used for the synthesis of KNbO3 based nanostructures [11,12]. Due to the dehydrolysis nature of the hydrothermal process, hydroxyl groups could remain inside and on the surface of the nanocrystals. The remaining hydroxyl groups could act as defects [13]; which affect global and local structures, and ferroelectric properties of the KNbO3 based nanocrystals. Therefore, it is important to understand the characteristics of the hydroxyl group, such as amount and position, in KNbO3 based nanostructures.

In this paper, we investigated the effect of hydroxyl group on global and local structures of hydrothermally grown KNbO3 nanorods. With increasing annealing temperature, the amount of hydroxyl groups systematically decreased and then finally disappeared at about 800 ◦ C. By analyzing the global and local structures obtained from high-resolution X-ray diffraction and Raman scattering measurements, we will discuss the amount, position and nature of hydroxyl defects in the hydrothermally grown KNbO3 nanorods. 2. Experimental KNbO3 nanorods were synthesized by using a hydrothermal method [14]. The as-grown nanorods were annealed at selected temperatures for 12 h in air. Hereafter, we call the as-grown KNbO3 as ‘KNO-AG’ and the annealed KNbO3 at selected temperatures as ‘KNO-temperature’. Global structure and morphology of the KNO nanorods were characterized by using a high-resolution X-ray diffractometer (HR-XRD) with Cu K␣ radiation and a field emission scanning electron microscopy (FESEM), respectively. Hydroxyl groups in the KNO nanorods were characterized by using thermogravimetric (TG) analysis and Fourier transform infrared (FTIR) spectroscopy measurements. For FTIR spectroscopy measurement, the nanorods were mixed with KBr powder and dried at 120 ◦ C for 12 h. Micro Raman scattering measurement was used to characterize local structure of the KNO nanorods. The nanorods were excited by the 488 nm line of a DPSS laser, and focused to 1 ␮m by using a microscope objective lens (×100). The power of the excitation laser was about 1.5 mW to avoid laser heating.

3. Results and discussion ∗ Corresponding author. Tel.: +82 32 860 7659; fax: +82 32 872 7562. E-mail address: [email protected] (J.H. Jung). 0254-0584/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.matchemphys.2011.05.059

TG and FTIR spectroscopy measurements show that the amount of hydroxyl groups in the KNO nanorods decreased with increasing

Author's personal copy 1072

B.K. Yun et al. / Materials Chemistry and Physics 129 (2011) 1071–1074

Fig. 1. (a) Thermogravimetry and (b) Fourier transform infrared absorption spectra of the KNO nanorods.

annealing temperature and finally disappeared at about 800 ◦ C. Fig. 1(a) shows TG result of the KNO nanorods. The total weight loss of about 1.1%, up to 800 ◦ C, seems to occur through three different mechanisms. A gradual weight loss occurs over a wide temperature range from 50 to 600 ◦ C, while a sudden loss occurs in a narrow temperature range from 300 to 400 ◦ C. The former and the latter weight loss may be attributed to the desorption of surface-adsorbed and lattice hydroxyl groups, respectively. Wada et al. reported similar TG results in hydrothermally grown BaTiO3 nanoparticles [15]. Based on the particle size dependence in a TG analysis, they showed that the amount of surface-adsorbed hydroxyl groups increased with decreasing particle size while that of lattice hydroxyl groups remained nearly the same irrespective of particle size. Another sudden weight loss occurs in a narrow temperature range, i.e., at about 700 ◦ C. Since K ion is reported to evaporate near 800 ◦ C in bulk KNO [16], we attribute the third weight loss in the KNO nanorods to the volatilization of K ion. Fig. 1(b) shows absorption spectra of the annealed KNO nanorods near the stretching mode of hydroxyl group. To completely remove interference due to absorbed H2 O, we measured the absorption spectra in vacuum just after drying the nanorod. The KNO-AG shows a broad and strong absorption spectrum. On the other hand, the KNO-300 clearly shows two absorption peaks centered at about 3350 and 3500 cm−1 . With a further increase in annealing temperature, for example, the KNO-600 shows an absorption peak centered at about 3450 cm−1 and the KNO-800 shows negligible absorption. (The absorption spectra of KNO700 are nearly the same as those of KNO-800.) It is well known that the absorption spectra due to surface-adsorbed hydroxyl groups are quite broad because the absorption occurred at many surface sites, and those due to lattice hydroxyl groups are narrow and located at about 3500 cm−1 [15,17]. Based on these facts and the results shown in Fig. 1(a), the absorption peak of surface-adsorbed hydroxyl groups should become narrower and show a frequency shift with increasing annealing temperature, as shown in Fig. 1(b). The absorption peak of lattice hydroxyl groups should start to decrease above 300 ◦ C with negligible frequency shift. In KNO-300, the absorption of lattice hydroxyl

Fig. 2. (a) XRD patterns of the KNO nanorods near 2 = 45◦ (left) and 56◦ (right). (b) Annealing temperature dependence of volume (open squares), and intensity ratio between (0 2 2) and (2 0 0) planes, i.e., I0 2 2 /I2 0 0 (solid squares). In (b), the dash-dot line represents the volume of a KNO single crystal.

groups is quite obvious. In KNO-AG and KNO-600, on the other hand, the absorptions of lattice hydroxyl groups are not obvious due to the broad spectrum and the higher-frequency shifted spectrum originated from the surface-adsorbed hydroxyl groups. However, we believe that the absorption of lattice hydroxyl groups decreases with annealing temperature and nearly disappears above 600 ◦ C. Global structural properties of the KNO nanorods seem to follow the evolution of hydroxyl groups and K ion with annealing temperature. In Fig. 2(a), we show XRD patterns of the annealed KNO nanorods near the (0 2 2) and (2 0 0) planes, i.e., 2 = 45◦ (left), and near the (1 1 3) and (1 3 1) planes, i.e., 2 = 56◦ (right). To precisely determine the structural parameters, we scanned 2 angles with a very high resolution of 0.002◦ . In the wide XRD patterns, we did not observe secondary phases in the annealed KNO nanorods (not shown). With increasing annealing temperature, especially above 300 ◦ C, i.e., the temperature at which lattice hydroxyl group starts to desorb, all peaks become narrow. At 700 ◦ C, i.e., the temperature near which K ion starts to volatilize, the peak positions abruptly shift to higher angles. In Fig. 2(b), we show annealing temperature dependences of volume (open squares) and the intensity ratio between the (0 2 2) and (2 0 0) planes, i.e., I0 2 2 /I2 0 0 (solid squares). Since hydroxyl groups could result in an increase in lattice constants [13], we expected a shrink in volume upon the desorption of the lattice hydroxyl groups. Probably due to the small amount of hydroxyl groups in our KNO-AG, however, the volume of the annealed KNO does not change significantly. On the other hand, the volume of KNO sharply shrinks above 600 ◦ C, probably due to the volatilization of K ion as has been observed in K-deficient KNO single crystals [18]. While there is little change in volume below 600 ◦ C, the intensity ratio I0 2 2 /I2 0 0 remains the same below 300 ◦ C but starts to increase above 300 ◦ C, at which temperature the lattice hydroxyl groups start to desorb. We notice that the KNO-AG grows along [0 1 1] direction and its zone axis is parallel to [1 0 0] direction [11,14]. We think that the increase in I0 2 2 /I2 0 0 might be related to the morphology of the nanorods. The desorption of the lattice hydroxyl group occurs at the whole nanorods, which results in more

Author's personal copy B.K. Yun et al. / Materials Chemistry and Physics 129 (2011) 1071–1074

1073

Fig. 3. SEM images of (a) KNO-AG, (b) KNO-600 and (c) KNO-800.

to the K ion could be strongly affected by their contents. Such a scenario could also explain the annealing temperature dependences of the bending and stretching frequency shifts. Since the bond angle of O–Nb–O is determined by the ionic radius of K ion, the amount of interstitial hydroxyl groups might also change the angle of O–Nb–O, hence the frequency. Since the length of Nb–O is not significantly affected by K ion, the amount of interstitial hydroxyl groups only slightly affect the Nb–O bond length, hence the frequency. If K ion starts to volatilize, there should be some chemical changes near the K ion to satisfy the charge neutrality condition. Such changes might cause the hardening of the external phonon frequency as observed in Fig. 4(b).

Intensity (arb. units)

(a) KNO-800 KNO-700 KNO-600 KNO-300 KNO-AG 0 Relative Raman Shift (cm-1)

lattice hydroxyl groups desorbing along [0 1 1] direction; hence the increase in I0 2 2 /I2 0 0 . Morphology of the KNO nanorods is robust irrespective of the hydroxyl group contents. Fig. 3 shows SEM images of (a) KNO-AG, (b) KNO-600 and (c) KNO-800. (SEM image of KNO-700 is nearly the same as that of KNO-800.) As shown in Fig. 3(a), the KNO-AG nanorods have various aspect ratios. The width ranges from 60 to 250 nm and the length ranges from 150 nm to 1.1 ␮m. Morphology of the KNO-600 is similar to that of the KNO-AG. Morphology of the KNO-800 seems to change slightly, but appears similar to that of the KNO-AG. Local structural properties of the KNO nanorods seem to be strongly affected by the amount, position and nature of the lattice hydroxyl group. In Fig. 4(a), we show the Raman scattering spectra of the annealed KNO nanorods. Raman scattering spectra of the KNO-AG show typical phonon peaks of KNO single crystal with orthorhombic structure [19]. With increasing annealing temperature, there is no significant change in overall spectrum but there is successive narrowing of phonon width. For example, the full width at half maximum (FWHM) of the phonon located at about 836 cm−1 for KNO-AG is estimated to be 60 cm−1 , while the FWHM for KNO600 (KNO-800) is 50 (40) cm−1 . In Fig. 4(b), we show the annealing temperature dependence of phonon frequencies for an external mode located at about 280 cm−1 , a bending mode at about 598 cm−1 and a stretching mode at about 836 cm−1 . To clarify the shift of phonon frequency, we show the relative frequency shift by subtracting the phonon frequency of KNO-AG from that of the annealed KNO. The phonon peaks show red-shift up to about 500 ◦ C and then show blue-shift above 500 ◦ C. We note that the phonon frequency of the external mode shows significant annealing temperature dependence, while that of the stretching mode shows only a slight dependence. Atomic motions of external, bending and stretching modes characterize the relative motion of K ions with respect to NbO6 octahedra, the bending motion of O–Nb–O bonds and the stretching motion of Nb–O bonds, respectively [20,21]. If the hydroxyl group (OH) in KNO substitutes lattice oxygen (O), the phonon frequencies of the external, bending and stretching modes will be modified due to the heavier mass of OH than that of O and the weaker attractive Coulomb forces of Nb–OH and K–OH than those of Nb–O and K–O [17]. Irrespective of amount of hydroxyl groups, however, the stretching mode does not show any frequency shift. In addition, the external and bending modes soften with decreasing amount of lattice hydroxyl groups, i.e., above 300 ◦ C, and then they harden with the volatilization of K ion above 600 ◦ C. These experimental results imply that the lattice hydroxyl groups in the KNO should exist as interstitial defects rather than substitutional defects [22]. If the hydroxyl groups mainly locate near K ions, the phonons related

200

400 600 800 Raman Shift (cm-1)

1000

10 (b) 0 -10

Stretching Bending

-20

External 200

400 600 Annealing Temperature (oC)

800

Fig. 4. (a) Raman scattering spectra of the KNO nanorods. (b) Annealing temperature dependence of relative Raman shifts for the external mode located at about 280 cm−1 (solid circles), bending mode located at about 598 cm−1 (solid squares) and stretching mode located at about 836 cm−1 (solid diamonds).

Author's personal copy 1074

B.K. Yun et al. / Materials Chemistry and Physics 129 (2011) 1071–1074

4. Conclusions We investigated global and local structures of annealed KNbO3 nanorods through XRD, SEM and Raman scattering measurements. On annealing, the weight of the KNbO3 nanorods decreased due to the desorption of surface-adsorbed hydroxyl groups over a wide temperature range, i.e., 50–600 ◦ C, lattice hydroxyl groups over a narrow temperature range, i.e., 300–400 ◦ C and the volatilization of K ion near 700 ◦ C. The desorption of the hydroxyl groups had no significant effect on the unit cell volume, probably due to the small amount of the hydroxyl groups in the KNbO3 nanorods (less than 1.1% of the total mass). However, an increase in intensity ratio between (0 2 2) and (2 0 0) planes, i.e., I0 2 2 /I2 0 0 was observed. This result is attributed to the fact that the hydroxyl groups disappear more effectively along [0 1 1] direction than in other directions, since the elongated direction of the KNbO3 nanorods is parallel to the [0 1 1] direction. With the desorption of hydroxyl groups, we observed a red-shift of phonon frequencies for external and bending modes, but not for stretching mode. With the volatilization of K ion, on the other hand, we observed a blueshift of phonon frequencies for the external and bending modes. Based on these results, we suggest that the hydroxyl groups might reside near K ions as interstitial defects rather than substitutional defects. Acknowledgements This work was supported by the Korea Research Foundation Grant funded by the Korean Government (MOEHRD) (KRF-2008313-C00253). M.S. and S.Y. were supported by the Korea Research

Foundation Grant funded by the Korean Government (MOEHRD) (2009-0073752). References [1] A. Schilling, D. Byrne, G. Catalan, K.G. Webber, Y.A. Genenko, G.S. Wu, J.F. Scott, J.M. Gregg, Nano Lett. 9 (2009) 3359. [2] W.S. Yun, J.J. Urban, Q. Gu, H. Park, Nano Lett. 2 (2002) 447. [3] I.I. Naumov, L. Bellaiche, H. Fu, Nature 432 (2004) 737. [4] J.E. Spanier, A.M. Kolpak, J.J. Urban, I. Grinberg, L. Ouyang, W.S. Yun, A.M. Rappe, H. Park, Nano Lett. 6 (2006) 735. [5] L. Louis, P. Gemeiner, I. Ponomareva, L. Bellaiche, G. Geneste, W. Ma, N. Setter, B. Dkhil, Nano Lett. 10 (2010) 1177. [6] W. Lee, H. Han, A. Lotnyk, M.A. Schubert, S. Senz, M. Alexe, D. Hesse, S. Baik, U. Gosele, Nat. Nanotechnol. 3 (2008) 402. [7] M. Alexe, C. Harnagea, D. Hesse, J. Electroceram. 12 (2004) 69. [8] G. Suyal, E. Colla, R. Gysel, M. Cantoni, N. Setter, Nano Lett. 4 (2004) 1339. [9] Y. Nakayama, P.J. Pauzauskie, A. Radenovic, R.M. Onorato, R.J. Saykally, J. Liphardt, P. Yang, Nature 447 (2007) 1098. [10] Y. Saito, H. Takao, T. Tani, T. Nonoyama, K. Takatori, T. Homma, T. Nagaya, M. Nakamura, Nature 432 (2004) 84. [11] A. Magrez, E. Vasco, J.W. Seo, C. Dieker, N. Setter, L. Forro, J. Phys. Chem. B 110 (2006) 58. [12] Y. Hu, H. Gu, Z. Hu, W. Di, Y. Yuan, J. You, W. Cao, Y. Wang, H.L.W. Chan, Cryst. Growth Des. 8 (2008) 832. [13] L. Qi, B.I. Lee, P. Badheka, L.-Q. Wang, P. Gilmour, W.D. Samuels, G.J. Exarbos, Mater. Lett. 59 (2005) 2794. [14] B.K. Yun, Y.S. Koo, J.H. Jung, M. Song, S. Yoon, Physica B 405 (2010) 4866. [15] S. Wada, T. Suzuki, T. Noma, Jpn. J. Appl. Phys. 34 (1995) 5368. [16] U. Flückiger, H. Arend, H.R. Oswald, Am. Ceram. Soc. Bull. 56 (1977) 575. [17] T. Noma, S. Wada, M. Yano, T. Suzuki, J. Appl. Phys. 80 (1996) 5223. [18] K. Kudo, K. Kakiuchi, K. Mizutani, T. Fukami, K. Hoshikawa, J. Cryst. Growth 267 (2004) 150. [19] A.M. Quittet, M.I. Bell, M. Krauzman, P.M. Raccah, Phys. Rev. B 14 (1976) 5068. [20] M.N. IIiev, M.V. Abrashev, J. Raman Spectrosc. 32 (2001) 805. [21] N. Kolev, C.L. Chen, M. Gospodinov, R.P. Bontchev, V.N. Popov, A.P. Litvinchuk, M.V. Abrashev, V.G. Hadjiev, M.N. IIiev, Phys. Rev. B 66 (2002) 014101. [22] S. Kapphan, G. Weber, Ferroelectrics 37 (1981) 673.