Strain Relaxation in Sol−Gel Grown Epitaxial Anatase Thin Films - The

Anatase TiO2 thin films on LaAlO3 (LAO) substrates were epitaxially grown at a ... Spin-coating and an intermediate pyrolysis step were repeated 3 tim...
0 downloads 0 Views 220KB Size
J. Phys. Chem. C 2008, 112, 4205-4208

4205

Strain Relaxation in Sol-Gel Grown Epitaxial Anatase Thin Films Hyun Suk Jung,†,‡ Jung-Kun Lee,*,†,§ Jaegab Lee,‡ Bo Soo Kang,| Quanxi Jia,† and Michael Nastasi† Materials Physics and Applications DiVision, Los Alamos National Laboratory, Los Alamos, New Mexico 87545, Department of Mechanical Engineering and Materials Science, UniVersity of Pittsburgh, Pittsburgh, PennsylVania 15261, School of AdVanced Materials Engineering, Kookmin UniVersity, Jeongneung-dong, Seongbuk-gu, Seoul 136-702, Korea, and Analytical Engineering Center, Samsung AdVanced Institute of Technology, P.O. Box 111, Suwon 440-600, Korea ReceiVed: August 2, 2007; In Final Form: January 2, 2008

Anatase TiO2 thin films on LaAlO3 (LAO) substrates were epitaxially grown at a temperature as low as 350 °C using a simple sol-gel process. X-ray diffraction and high-resolution transmission electron microscopy showed that the anatase films have the epitaxial relationship of (001)TiO2||(001)LaAlO3. While the low-temperature growth of the anatase film yielded a residual strain, subsequent annealing at higher temperatures can remove the strain and recover the lattice parameters of a perfect anatase crystal. Measurements of the oxygen content in the anatase films by non-Rutherford elastic resonance scattering analysis suggest that the strain relaxation during higher temperature annealing is due to the incorporation of oxygen and the concomitant annihilation of oxygen vacancies.

Introduction The unique characteristics of TiO2 films such as the effective extraction of photogenerated carriers and photocatalytic activity under UV irradiation have made them very attractive for potential applications in solar energy conversion or photocatalytic decomposition of pollutants.1,2 Given the potential photochemical applications of anatase films, the key property that is responsible for excellent performance, is charge carrier transport, which is determined by the mobility. This in turn is critically dependent on the thin film’s crystal quality. For example, grain boundaries are detrimental to charge transport because photogenerated carriers scatter and/or are trapped at grain boundaries, which subsequently induce charge recombination and a decrease in the effective carrier concentration.3,4 Therefore, a high crystalline quality is a critical property that must be considered when new synthesis methods of anatase TiO2 films are being developed. Epitaxial anatase TiO2 films have been grown by a number of techniques such as molecular-beam epitaxy, pulsed laser deposition, and RF sputtering.5-7 While these synthesis methods are well-suited for basic research, they are not truly cost effective for large scale application; a cost-effective way to synthesize the epitaxial anatase films without a vacuum environment has not been available. Moreover, these vacuum-based growth methods have the disadvantage that the amount of residual hydroxyl component in the anatase films is negligible. Lei et al. reported the photocatalytic properties of epitaxial and polycrystalline anatase films prepared by RF sputtering8 and showed that the photocatalytic activity was low for both films regardless of the crystallinity. A low photocatalytic activity was ascribed to the vacuum growth process and the fact that the * Corresponding author. Tel.: (412) 648-3395; e-mail: [email protected]. † Los Alamos National Laboratory. ‡ Kookmin University. § University of Pittsburgh. | Samsung Advanced Institute of Technology.

film surface lacks hydroxyl groups that produce highly reactive hydroxyl radicals. These findings suggest that photocatalytic activity improvement in anatase films can be obtained from synthesis methods that produce epitaxial material and surface hydroxyl groups. A chemical solution synthesis process is an alternative approach to vacuum requiring techniques for the epitaxial growth of anatase TiO2 films. Recently, epitaxial anatase was grown in a highly pressurized water-based solution by the dissolution and reprecipitation of the Ti component.9 However, this process has to progress under high pressure (≈30 atm). In the present paper, we report the growth of epitaxial anatase films in ambient atmosphere using a sol-gel method. We further investigated the effect of processing conditions on lattice strain and its relaxation. Experimental Procedures A TiO2 sol consisting of titanium-alkoxide and ethanol was prepared as follows. Titanium isopropoxide (TTIP, Aldrich, 97%) was dissolved in dry ethanol (Aldrich, 99.9%), and the resulting solution was partially hydrolyzed by adding a mixture of distilled water, nitric acid, and ethanol. The molar ratio of TTIP/H2O/HNO3 was 1:4:0:04. A final sol was spun on LaAlO3 (001) substrates at 3000 rpm for 30 s. The films were then annealed at 350 °C for 5 min to decompose organics in the films. Spin-coating and an intermediate pyrolysis step were repeated 3 times until the film thickness was approximately 150 nm. Consequently, the TiO2 films were then annealed at temperatures in the range of 350 to 900 °C in wet O2 or forming gas (5% H2-95% Ar). Results and Discussion X-ray diffraction (XRD) measurements were used to investigate the crystal structure of the sol-gel grown films. Figure 1a shows XRD diffraction patterns of 350 °C annealed TiO2 films on the LaAlO3 (001) substrate. Only the anatase (004)

10.1021/jp076194n CCC: $40.75 © 2008 American Chemical Society Published on Web 02/23/2008

4206 J. Phys. Chem. C, Vol. 112, No. 11, 2008

Figure 1. (a) XRD θ-2θ scan results of 350 °C annealed TiO2 film on LaAlO3 and (b) XRD Φ scan results for (101) TiO2 and (101) LaAlO3.

Jung et al.

Figure 2. (a) HRTEM image of TiO2 film on LaAlO3 and (b) selected area diffraction image of the TiO2 film on LaAlO3.

peak was observed in the θ-2θ scan, suggesting that the film is a single phase TiO2 with a preferential c-axis orientation. Figure 1b shows the Φ scans of (101) TiO2 and (101) LaAlO3. The epitaxial nature of TiO2 on LaAlO3 is clearly seen from those measurements. Consistent with previous reports, the epitaxial relationship between anatase-TiO2 and LaAlO3 can be described as10,11

(001)TiO2||(001)LaAlO3 [100]TiO2||[100]LaAlO3 In Supporting Information 1, the full width at half-maximum (fwhm) of the rocking curve from the (004) peak of the 350 °C annealed anatase film was 1.9°. The average value of fwhm of peaks from Φ scans of (101) TiO2 was 2.3°, as compared to the average value of 0.95° for LaAlO3. The epitaxial quality of the sol-gel-based anatase film is comparable to or a little bit lower than those deposited from hydrothermal methods, which demonstrates that the simple sol-gel process is another alternate to fabricating the epitaxial anatase films at a low temperature.5-7,9 Figure 2a shows the high-resolution transmission electron microscopy (HRTEM) image of the 350 °C annealed TiO2 film. The interface between the film and the substrate is flat and shows no secondary phase. The lattice image of the HRTEM micrograph confirms that the anatase films grew epitaxially on the LAO substrates, in agreement with XRD. Corresponding selective area diffraction (SAD) for the 350 °C annealed TiO2 film shows two sets of diffraction patterns. The distinctive

Figure 3. XRD θ-2θ scan results of TiO2 films on LaAlO3, which were annealed at 350, 500, 700, 800, and 900 °C.

diffraction spots marked with subscripts A and L represent the diffraction patterns from the anatase TiO2 and LAO phases, respectively. This observation indicates that the anatase film is epitaxially related to LAO.6,10 It is also noted that the diffraction spots from anatase are sharp and distinguished, which implies the high crystallinity of the sol-gel grown anatase films. Figure 3 presents the change in the (004) peaks of the anatase films, which were annealed at various temperatures from 350 to 900 °C. The peak position of the (004) plane in the 350 °C

Sol-Gel Grown Epitaxial Anatase Thin Films annealed film is far away from 37.70°, which is expected for perfect anatase single crystals. The shift in the XRD peak position for the 350 °C annealed film indicates that the lowtemperature sol-gel processed film possesses a shorter out-ofplane lattice parameter due to an in-plane tensile stress.12-16 Further annealing of the anatase films at higher temperatures relaxes the strain. The experimentally measured out-of-plane d-spacings as a function of annealing temperature are plotted in the inset of Figure 3. Lattice mismatch between TiO2 and LaAlO3, difference in thermal expansion coefficient, and pyrolysis and densification of organic groups during the burn-out process may all cause the decrease in the d-spacing along the out-of-plane direction. Since anatase (a ) 0.378 nm) and LaAlO3 (a ) 0.379 nm) have very similar lattice parameters, the small lattice mismatch (-0.26%) between them should have a very small contribution to the lattice parameter changes shown in Figure 3.17 Since the lattice constant of LaAlO3 (a ) 0.379 nm) is slightly larger than that of anatase (a ) 0.378 nm), the lattice mismatch between LaAlO3 and anatase produces an in-plane tensile strain of 0.26%, which is smaller than what was observed in this research. Furthermore, the strain caused by the lattice mismatch should not be a function of annealing temperature.18 Therefore, the strain relaxation that is observed in Figure 3 is not associated with the epitaxial stress that is produced by the lattice mismatch of films and substrates. The other factors affecting the strain state of the epitaxial films are the changes in chemical composition and thermal stress. The residual thermal stress of the films is determined by the difference in the thermal expansion coefficient as follows:19

σ ) Ef(Rf - Rs)∆T where Ef is Young’s modulus of TiO2, and Rf and Rs are the thermal expansion coefficients of the anatase phase and LAO, respectively. Since Ef is 230 GPa, and the thermal expansion coefficients (R) of TiO2 and LaAlO3 are 8.6 and 9.2 ppm/°C, the expected residual stress for the 350 °C annealed anatase film is only 46 MPa.19-21 Given Poisson’s effect, this corresponds to a residual strain of less than 0.03%. Therefore, the residual stress, induced by the difference in the thermal expansion coefficient, is most likely not responsible for the significant decrease in the d-spacing observed in the 350 °C annealed film. Furthermore, pyrolysis and densification in the sol-gel process derived film can lead to the evolution of inplane tensile stress. If this is the case, the increase in the annealing temperature should reduce the out-of-plane lattice parameter since the pyrolysis and densification are more vigorous at higher temperatures.22,23 However, contrary to this expectation, the out-of-plane lattice parameter of our films increased with an increase in the annealing temperature. The in-plane tensile stress almost disappeared after 900 °C annealing. To explore the origin of the strain state, the oxygen content of the films was measured. Figure 4 shows non-Rutherford elastic resonance scattering (NRERS) spectra of 350 and 900 °C annealed TiO2 films using 16O(a,a)16O at 3.05 MeV.24 The energy of the incident He+ beam was controlled to be resonant with oxygen slightly below the surface to eliminate surface effects. Previously, NRERS analysis was successfully used to measure the oxygen non-stoichiometry of YBCO films within less than 1%.25 The almost identical shape of the Ti and La peaks for the 350 and 900 °C annealed films, as shown in Figure 4, illustrates that the change in the surface roughness and the formation of internal pores and cracks were minimized in this study. However, there is a clear change in the oxygen peak

J. Phys. Chem. C, Vol. 112, No. 11, 2008 4207

Figure 4. Spectra of non-Rutherford elastic resonance scattering for 350 °C (red dotted line) and 900 °C (black solid line) annealed TiO2 films on LaAlO3.

intensity. The oxygen content of the 900 °C annealed anatase film is higher than that of the 350 °C annealed one as is plotted in the inset of Figure 4. The oxygen contents in each sample were evaluated using RUMP.26 The calculated difference in the oxygen content was approximately 1.9%. This NRERS analysis demonstrates that the lower temperature annealed films have a significant oxygen deficiency. We believe that the thermal decomposition process of the residual organic materials in the sol-gel grown films develops a transient oxygen deficient state in ambient atmosphere and the creation of oxygen vacancies. In TiO2, the oxygen vacancy is accompanied by an increase in the local electron concentration rather than the formation of a cation vacancy, leading to the shrinkage of the lattice parameter.27 This may be the reason as to why the 350 °C annealed film possessed an in-plane tensile stress. Munoz et al. have shown the concurrent generation of Ti3+ ions and oxygen vacancies during the thermal decomposition of TiO2 gels using Rietveld refinement of XRD results.28 To confirm the presence of the oxygen vacancy, we also measured X-ray photoemission spectra to confirm the presence of oxygen vacancies in a 350 °C annealed anatase film. As shown in Supporting Information 2, the binding energy of Ti 2p3/2 from the 350 °C annealed film was 458.2 eV, lower than 458.5 eV from the 900 °C annealed one. According to Endle et al.29 and Jiang et al.,30 the Ti 2p3/2 peak shifts to lower binding energy, which results from the presence of Ti3+ ions. Since Ti3+ ions are generated with oxygen vacancies, the lower binding energy shift supports the presence of oxygen vacancies in the 350 °C annealed film. We also measured the in-plane lattice parameters of the anatase films. An increase in the annealing temperature decreased the in-plane lattice parameter slightly and got it very close to that of ideal anatase TiO2. Since the lattice parameter of the c-axis of the 900 °C annealed films is also that of ideal anatase TiO2, this clearly shows that the strain of as-deposited anatase films is relaxed when defects such as oxygen vacancy are annihilated by introducing the oxygen at high temperatures. Figure 5 compares the lattice strain in epitaxial TiO2 films that were annealed in forming gas or O2 at 900 °C. In contrast to the fully oxidized anatase film, the (004) peak position of the films annealed in a reducing atmosphere is quite close to the

4208 J. Phys. Chem. C, Vol. 112, No. 11, 2008

Jung et al. References and Notes

Figure 5. XRD θ-2θ scan results of TiO2 films on LaAlO3, which were annealed in a forming gas (red line) and O2 gas (black line) at 900 °C.

350 °C annealed film. Given that the forming gas treatment significantly increased the concentration of oxygen vacancies in oxide materials, the results in Figure 5 demonstrate that the strain states of our epitaxial films by the sol-gel process are controlled by oxygen vacancies. Conclusion The epitaxial growth of anatase films was achieved at a temperature of 350 °C by using a simple sol-gel method. The residual strain in these films was attributed to the generation of oxygen vacancies during the burn-out process. Higher temperature annealing in oxygen relieved the residual strain. The relaxation of strain could be associated with the annihilation of oxygen vacancies during the annealing process, as evidenced by the non-Rutherford elastic resonance scattering analysis. Acknowledgment. The Los Alamos portion of this work was supported by Laboratory Directed Research and Development and DOE Office of Science funds of the Los Alamos National Laboratory. Kookmin University is supported by the ERC Program (Center for Materials and Processes of SelfAssembly) of MOST/KOSEF (R11-2005-048-00000-0). This work was also supported by the Seoul R&BD Program. Supporting Information Available: X-ray diffraction data of TiO2 film and XPS images of TiO2 films. This information is available free of charge via the Internet at http://pubs.acs.org.

(1) O’Regan, B.; Gra¨tzel, M. Nature (London, U.K.) 1991, 353, 737. (2) Fujishima, A.; Honda, K. Nature (London, U.K.) 1972, 238, 37. (3) Matsui, H.; Tabata, H.; Hasuike, N.; Harima, H.; Mizobuchi, B. J. Appl. Phys. 2005, 97, 123511. (4) Yamamoto, S.; Sumita, T.; Sugiharuto, A.; Miyashita, A.; Naramoto, H. Thin Solid Films 2001, 401, 88. (5) Murakami, M.; Matsumoto, Y.; Nakajima, K.; Makino, T.; Segawa, Y.; Chikyow, T.; Ahmet, P.; Kawasaki, M. Appl. Phys. Lett. 2001, 78, 2664. (6) Park, B. H.; Huang, J. Y.; Li, L. S.; Jia, Q. X. Appl. Phys. Lett. 2002, 80, 1174. (7) Jellison, G. E., Jr.; Boatner, L. A.; Budai, J. D.; Jeong, B.-S.; Norton, D. P. J. Appl. Phys. 2003, 93, 9537. (8) Lei, M.; Sakae, T.; Yoichi, K.; Misao, I.; Shoichi, T.; Kenji, K. Appl. Surf. Sci. 2004, 238, 125. (9) Zhang, Z. J. Mater. Res. 2005, 20, 292. (10) Huang, J. Y.; Park, B. H.; Jan, D.; Pan, X. Q.; Zhu, Y. T.; Jia, Q. X. Philos. Mag. A 2002, 82, 735. (11) Gorbenko, O. Y.; Samoilenkov, S. V.; Graboy, I. E.; Kaul, A. R. Chem. Mater. 2002, 14, 4026. (12) Huang, W.; Zhu, J.; Zeng, H. Z.; Wei, X. H.; Zhang, Y.; Li, Y. R. Appl. Phys. Lett. 2006, 89, 262506. (13) Desfeux, R.; Bailleul, S.; Da Costa, A.; Prellier, W.; Haghiri-Gosnet, A. M. Appl. Phys. Lett. 2001, 78, 3681. (14) Li, H.; Roytburd, A. L.; Alpay, S. P.; Tran, T. D.; SalamancaRiba, L.; Ramesh, R. Appl. Phys. Lett. 2001, 78, 2354. (15) Shinde, S. R.; Ramesh, R.; Lofland, S. E.; Bhagat, S. M.; Ogale, S. B.; Sharma, R. P.; Venkatesan, T. Appl. Phys. Lett. 1998, 72, 3443. (16) Dimakis, E.; Iliopoulos, E.; Tsagaraki, K.; Adikimenakis, A.; Georgakilas, A. Appl. Phys. Lett. 2006, 88. (17) Sasahara, A.; Droubay, T. C.; Chambers, S. A.; Uetsuka, H.; Onishi, H. Nanotechnology 2005, 16, 18. (18) Lin, Y.; Wang, H.; Haeley, M. E.,; Foltyn, S. R.; Jia, Q. X.; Collis, G. E.; Burrell, A. K.; McCleskey, T. M. Appl. Phys. Lett. 2004, 85, 3426. (19) Shibata, T.; Irie, H.; Hashimoto, K. J. Phys. Chem. B 2003, 107, 10696. (20) Chakoumakos, B. C.; Schlom, D. G.; Urbanik, M.; Luine, J. J. Appl. Phys. 1998, 83, 1979. (21) Inaba, H.; Hayashi, H.; Suzuki, M. Solid State Ionics 2001, 144, 99. (22) Scherer, G. W. J. Sol.-Gel Sci. Technol. 1997, 8, 353. (23) Kozuka, H.; Takenaka, S.; Tokita, H.; Hirano, T.; Higashi, Y.; Hamatani, T. J. Sol.-Gel Sci. Technol. 2003, 26, 681. (24) Tesmer, J. R.; Nastasi, M. Handbook of Modern Ion Beam Materials Analysis; Materials Research Society: Pittsburgh, PA, 1995. (25) Zhai, H. Y.; Zhang, Z. H.; Chu, W. K. Appl. Phys. Lett. 2001, 78, 649. (26) Ziegler, J. F.; Biersack, J. P.; Littmark, U. The Stopping and Range of Ions in Solids; Pergamon Press, Inc.: New York, 1985. (27) Eastman, J. A. J. Appl. Phys. 1994, 75, 770. (28) Munoz, E.; Boldu, J. L.; Andrade, E.; Novaro, O.; Bokhimi, X. J. Am. Ceram. Soc. 2001, 84, 392. (29) Endle, J. P.; Sun, Y.-M.; White, J. M.; Ekerdt, J. G. J. Vac. Sci. Technol., A 1998, 16, 1262. (30) Jiang, N.; Zhang, H. J.; Bao, S. N.; Shen, Y. G.; Zhou, Z. F. Physica B 2004, 352, 118.