Preparation and characterization of single crystalline anatase TiO₂ films on LSAT (001) by MOCVD

Abstract

TiO₂ thin films with anatase structure have been prepared on [LaAlO₃]_0.3[SrAl_0.5Ta_0.5O₃]_0.7 (LSAT) (001) substrates by metalorganic chemical vapor deposition (MOCVD) in the substrate temperature range of 500–650 °C. Tetrakis-dimethylamino titanium (TDMAT) is used as the organometallic source and oxygen as oxidant. Structural and optical properties of the films as well as the epitaxial mechanism have been investigated by X-ray diffraction (XRD), transmission electron microscopy (TEM) and optical transmittance spectra. The measurements and analyses show that the TiO₂ film grown at 550 °C exhibits the best crystallinity with anatase structure. The obtained TiO₂ film is a single-crystalline epitaxial film with no twins and very few defects. The heteroepitaxial relationship is determined as TiO₂ (001) ‖ LSAT (001) with TiO₂ [100] ‖ LSAT 〈100〉. The average transmittance of the film deposited at 550 °C in the visible wavelength is about 84% with an optical band gap of about 3.27 eV.

---

1. Introduction

Recently, titanium dioxide (TiO₂) films have attracted remarkable attention due to their inherent properties such as a large refractive index, a high dielectric constant¹ and a high photochemical activity.² These unique properties have aroused much interest in the development of applications including photocatalysis,^3,4 dye-sensitized solar cells (DSSCs),⁵ gas sensors,⁶ field effect transistors (FETs),⁷ high-k gate insulators⁸ and transparent electronics.⁹ Among the three typical polymorphs of TiO₂, anatase (tetragonal) with an energy gap of 3.2 eV is one of the most commonly investigated materials for photocatalysis¹⁰ and transparent conductivity.^9,11 Certain applications, such as transparent conductors or photoelectrochemical cells, require thin films of TiO₂ that exhibit a specific crystal structure, orientation or morphology. Thus, a wide range of oxide substrates have been used to grow TiO₂ thin films, such as LaAlO₃ (LAO),^12,13 SrTiO₃ (STO),^13–16 α-Al₂O₃ ¹⁷ and Y-stabilized ZrO₂.^14,18 Epitaxial TiO₂ films have been grown by pulsed laser deposition (PLD),^12,13 molecular beam epitaxy (MBE),^15,19 and metalorganic chemical vapor deposition (MOCVD).¹⁶ Single-crystal [LaAlO₃]_0.3[SrAl_0.5Ta_0.5O₃]_0.7 (LSAT) has a cubic perovskite structure with a lattice parameter of a = 0.3868 nm and it is suitable as a substrate to grow epitaxial anatase TiO₂ films with square lattice symmetry (a = 0.3785 nm). S. Yamamoto et al. reported the deposition of TiO₂ films on LSAT (001) by PLD.¹⁴ P. Fisher et al. studied the morphology of TiO₂ films deposited on LSAT (001) by MBE.¹⁵ However, to the best of our knowledge, seldom reports have been published on the structural and optical properties of epitaxial anatase TiO₂ films grown on LSAT (001), particularly by the means of MOCVD. MOCVD is suitable for production of thin films due to its advantages, such as easy control of deposition rate and film composition, and relatively large throughput up to commercial volume over other film fabrication techniques.

In the present work, high quality single crystalline anatase TiO₂ thin films with no twins have successfully been prepared on LSAT (001) by MOCVD. The structural and optical properties of the films as well as the epitaxial relationships between the substrate and the anatase TiO₂ film were investigated in detail. Preparation of high quality anatase TiO₂ heteroepitaxial films will lay a good foundation for improving the doping efficiency of TiO₂ materials and performance of TiO₂-base semiconductor devices.

2. Experimental details

The deposition of the TiO₂ films was carried out on the LSAT (001) substrates (double-face polished with a thickness of 0.5 mm, buying from Shanghai Daheng Optics and Fine Mechanics Co., Ltd. China) using a high vacuum MOCVD system with two separate gas flows. Prior to being loaded into the growth chamber, the substrates were cleaned successively with ethanol and deionized water in an ultrasonic bath and dried with pure nitrogen to prevent any contaminations on the surface. Commercially available tetrakis-dimethylamino titanium Ti[N(CH₃)₂]₄ (TDMAT) (purity 99.9999%, buying from Jiangsu Nata Opto-electronic Material Co., Ltd. China) was used as the organometallic (OM) source. Ultrahigh purity N₂ (9N) was used as the carrier gas for OM source. High purity O₂ (5N) was injected using a separate delivery line into the reactor as an oxidant. The temperature of the bubbler containing TDMAT was maintained at 20 °C with a pressure of 200 Torr. During the deposition, the growth pressure was kept at 10 Torr and the substrate temperatures (T_s) were 500–650 °C. The flow rates of TDMAT and O₂ were fixed at 15 and 70 sccm, respectively. The oxidation reaction of TDMAT, which takes place at above 500 °C, is given by

Ti(N(CH₃)₂)₄ + 15O₂ → TiO₂ + 8CO₂↑ + 2N₂↑ + 12H₂O↑

The out-of-plane epitaxial relationship of our samples was determined using a Rigaku D/MAX 2200PC X-ray diffractometer (XRD) with Cu Kα radiation. The Φ-scans were carried out with a X'Pert Pro MPD X-ray diffractometer with Cu Kα radiation to determine the in-plane epitaxial relationship. To further study the microstructure and epitaxial relationship of the samples, high-resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED) were performed with a Tecnai F30 transmission electron microscope operated at 300 kV on the cross-section of the sample prepared by gluing two pieces together with the thin films facing each other. The X-ray photoelectron spectroscopy (XPS) was employed to confirm the chemical composition of the prepared films using an ESCALAB MK II Multi-Technique Electron Spectrometer. The optical transmittance spectra were obtained using a TU-1901 double-beam UV-vis-NIR spectrophotometer in the wavelength range of 200–800 nm.

3. Results and discussion

XRD θ–2θ scans of the TiO₂ samples deposited at different substrate temperatures (T_s) are shown in Fig. 1(a)–(d), respectively. Besides a diffraction peak corresponding to the substrate LSAT (002), only another diffraction peak at about 38.2° corresponding to the anatase phase TiO₂ (004) can be observed in Fig. 1(a)–(d). This result indicates that the films deposited on LSAT (001) substrates are tetragonal anatase TiO₂ with a single orientation along c-axis. With increasing T_s from 500 to 650 °C, the location of the TiO₂ (004) peak does not change obviously, whereas the intensity first increases and then decreases. The full width at half maximum (FWHM) of peak TiO₂ (004) is approximately 0.42, 0.35, 0.52 and 0.62°, and the crystal size of the films calculated by the Scherrer formula using the FWHM is about 19.8, 23.8, 16.0, and 13.4 nm, corresponding to T_s of 500, 550, 600 and 650 °C, respectively. The film deposited at 550 °C has a minimum value of FWHM and a maximum value of crystal size, which reveals this sample has the best crystalline quality. These results imply that the substrate temperature significantly affects the structure of the TiO₂ films and the out-of-plane epitaxial relationship of the sample is TiO₂ (001) ‖ LSAT (001). So the following investigations are performed on the sample prepared at the optimized substrate temperature of 550 °C.

Fig. 1 XRD spectra of the TiO₂ samples deposited at different substrate temperatures of (a) 500 °C, (b) 550 °C, (c) 600 °C and (d) 650 °C.

Fig. 2(a) and (b) shows the off-specular Φ-scans of the TiO₂ {101} planes (ψ = 68.3°) for the film prepared at 550 °C and the {104} planes (ψ = 45.0°) for the LSAT (001) substrate at a fixed position. Fig. 2(a) shows four diffraction peaks separated by 90° from each other corresponding to the anatase TiO₂ {101} planes, confirming the four-fold symmetry along c-axis of TiO₂ with a tetragonal structure. Four sharp diffraction peaks with the same Φ-angles as those of the TiO₂ {101} peaks are observed for substrate LSAT, indicating that the two lattices have the same coordinate system. These results reveal that the sample prepared in our study is single crystalline epitaxial with an in-plane orientation relationship of TiO₂ [100] ‖ LSAT 〈100〉. The lattice mismatch between the film and substrate is about 2.3%.

Fig. 2 Off-specular Φ-scans of (a) TiO₂ {101} and (b) LSAT {101} planes at a fixed position for the sample deposited at 550 °C.

XPS measurements was performed to confirmed the chemical elemental composition of our sample. The representative XPS survey spectrum for TiO₂ film grown at 550 °C is displayed in Fig. 3(a). The peaks of Ti 2s, O 1s, Ti 2p and C 1s core levels as well as O KLL Auger peak can be observed, without any other elements observed, indicating the formation of titanium oxide.²⁰ The C 1s peak is caused by the hydrocarbon contamination on the surface of the film. The peak positions have been calibrated using the C 1s peak at 284.6 eV as a reference due to the deviation originating from the charge accumulation at the surface of the film. Curves (b) and (c) show the Ti 2p and O 1s XPS spectra of the sample with deconvoluted and fit peaks, respectively. As can be seen in curve (b), the main sharp peak is located at the binding energy of around 458.4 eV, which is assignable to Ti 2p_3/2 of Ti⁴⁺ as observed in TiO₂ spectrum.²¹ In addition, a broad peak located at 464.2 eV with a peak-to-peak separation of 5.8 eV correspond to Ti 2p_1/2 of Ti⁴⁺, which is also in accord with TiO₂. No Ti³⁺ signals can be found at 457 eV, indicating the inexistence of Ti³⁺ on the surface of the thin film. Fig. 3(c) presents two fit peaks of O 1s and the spectrum consists of an intense main peak located at around 529.7 eV that is attributed to the Ti–O bonding in the TiO₂ film, and the other weak peak at about 531.8 eV which can be regard as the C–O or OH⁻ adsorbed species on the film surface.²² The stoichiometry of the films is determined by the relative areas of the fit peaks of O 1s (main peak at 529.7 eV) and Ti 2p with the correction of the elemental sensitivity factors. The atomic ratio of O and Ti is calculated and the value is about 2.0, which is accord with the stoichiometry of TiO₂.

Fig. 3 XPS spectra of the TiO₂ thin film grown at 550 °C.

Further structural details of the sample grown at 550 °C were studied by TEM observations. Fig. 4(a) shows the cross-sectional low magnification TEM image of the film–substrate interface from which the thickness of the film is measured to be about 47.7 nm. The HRTEM and SAED micrographs of the interface area between TiO₂ film and LSAT substrate are shown in Fig. 4(b). The incident electron beam is parallel to the [010] direction of the LSAT substrate. The SAED shows a typical single crystalline diffraction pattern and the diffraction spots of TiO₂ (004), TiO₂ (101), TiO₂ (103), LSAT (001), LSAT (101) and LSAT (200) planes are clearly visible. In the HRTEM micrograph, a uniform and ordered lattice array in the interface region can be seen clearly without obvious transitional atomic layer, and an obvious interface of the film and substrate is marked by the white line. It is revealed that the deposited TiO₂ film is an epitaxial film and the growth surface is TiO₂ (001). For the TiO₂ film portion, the lattice fringes corresponding to TiO₂ (004) and (101) exhibits an angular separation of 68.0°, which is in good accordance with the standard calculated value of 68.3°. For the substrate, the lattice fringes corresponding to LSAT (001) and (101) with an angular separation of 45.6° can be seen. The SAED and HRTEM analyses of the interface for the sample show a clear epitaxial relationship of TiO₂ (001) ‖ LSAT (001) which is consistent with the XRD results.

Fig. 4 The cross-sectional (a) low magnification TEM, (b) HRTEM and SAED micrographs of the interface between TiO₂ film grown at 550 °C and LSAT substrate.

The good crystallinity and smooth morphology of the substrate contribute to the excellent epitaxial growth of TiO₂ films. During the film deposition, TiO₂ molecules are adsorbed by the LSAT substrate, and form nuclei on the substrate surface through the process of diffusion and condensation. The nuclei will constantly absorb molecules and steadily grow up into islands and expand into layer. Layer by layer repeatedly forms the well-crystalline epitaxial TiO₂ films. The lattice constant of the LSAT substrate is large than that of the anatase TiO₂, and the lattice mismatch along [100] direction is about 2.3%. Therefore, the anatase film epitaxially grown on the LSAT is under extension stress. In addition, the linear coefficient of thermal expansion of LSAT substrate and anatase TiO₂ along [100] orientation are about 10 × 10⁻⁶ K⁻¹ and 4.5 × 10⁻⁶ K⁻¹, respectively. So the defects of the anatase crystal seen in the HRTEM micrograph can be attributed to the lattice mismatch and thermal expansion mismatch between the TiO₂ film and the substrate. The critical thickness for the single crystal anatase film deposited on the LSAT substrate is about 8 nm estimated from the lattice mismatch. From the aspect of dynamic, the formation of misfit dislocation must overcome certain energy barrier, which makes the actual critical thickness considerably larger than the theoretical value, particularly when the substrates own a smooth surface. So the maximum possible thickness for our single crystal anatase TiO₂ film is much larger than 8 nm, maybe reaches 100 nm.

The optical transmittance spectra of the TiO₂ sample deposited at 550 °C and the bare LSAT substrate as a function of wavelength in the range of 200–800 nm are shown in Fig. 5. The average transmittance in the visible range of the sample is about 67%, and that of the LSAT substrate is about 79%. The absolute average transmittance for the film is about 84%. As a semiconductor with indirect band gap transition, the absorption coefficient (α) and optical band gap (E_g) have such a relationship: α(hv) = A(hv − E_g)², where h is the Planck's constant, ν is the frequency of the incident photon and A is a material dependent constant. The inset shows the plot of (αhν)^1/2 as a function of photon energy hv for the 550°C-deposited TiO₂ sample. The E_g can be determined by extrapolating the straight-line portion of this plot to the energy axis, and the value is about 3.27 eV, which is similar to that of TiO₂ samples annealed at 900 °C reported by D. J. Won et al.²³

Fig. 5 Optical transmittance spectra of the TiO₂ sample deposited at 550 °C and the bare substrate, with the plot of (αhν)^1/2 vs. hν for the sample deposited at 550 °C shown in the inset.

4. Conclusions

The anatase TiO₂ have been deposited on LSAT (001) substrates at 500–650 °C by the MOCVD method. The XRD θ–2θ analyses indicated that the film crystallinity was strongly influenced by the substrate temperature and the sample grown at 550 °C had the best crystalline quality. The XPS analyses showed that the deposited film was stoichiometric TiO₂. The XRD Φ-scans as well as the HRTEM and SAED measurements showed a clear epitaxial relationship of TiO₂ (001) ‖ LSAT (001) with TiO₂ [100] ‖ LSAT 〈100〉 for the sample grown at 550 °C. The average transmittance of the TiO₂ film in the visible range was about 84% with an optical band gap of 3.27 eV. The growth of high-quality single crystalline TiO₂ epitaxial films will lay a good foundation for the preparation of doped-TiO₂ materials and it can be widely applied in many fields such as transparent electronics devices and FETs.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (Grant No. 51472149).

References

1. K. Fröhlich, M. Ťapajna, A. Rosová, E. Dobročka, K. Hušeková, J. Aarik and A. Aidla, Electrochem. Solid-State Lett., 2008, 11, G16–G21.
2. D. C. Hurum, A. G. Agrios, K. A. Gray, T. Rajh and M. C. Thurnauer, J. Phys. Chem. B, 2003, 107, 4545–4549.
3. A. L. Linsebigler, G. Lu and J. T. Yates Jr, Chem. Rev., 1995, 95, 735–758.
4. A. Fujishima, T. N. Rao and D. A. Tryk, J. Photochem. Photobiol., C, 2000, 1, 1–21.
5. B. Liu and E. S. Aydil, J. Am. Chem. Soc., 2009, 131, 3985–3990.
6. N. Savage, B. Chwieroth, A. Ginwalla, B. R. Patton, S. A. Akbar and P. K. Dutta, Sens. Actuators, B, 2001, 79, 17–27.
7. M. Katayama, S. Ikesaka, J. Kuwano, H. Koinuma and Y. Matsumoto, Appl. Phys. Lett., 2008, 92, 132107.
8. M. Kadoshima, M. Hiratani, Y. Shimamoto, K. Torii, H. Miki, S. Kimura and T. Nabatame, Thin Solid Films, 2003, 424, 224–228.
9. Y. Furubayashi, T. Hitosugi, Y. Yamamoto, K. Inaba, G. Kinoda, Y. Hirose, T. Shimada and T. Hasegawa, Appl. Phys. Lett., 2005, 86, 252101.
10. L. Kavan, M. Grätzel, S. E. Gilbert, C. Klemenz and H. J. Scheel, J. Am. Chem. Soc., 1996, 118, 6716–6723.
11. N. Yamada, T. Hitosugi, N. L. H. Hoang, Y. Furubayashi, Y. Hirose, S. Konuma, T. Shimada and T. Hasegawa, Thin Solid Films, 2008, 516, 5754–5757.
12. H. Sakama, G. Osada, M. Tsukamoto, A. Tanokura and N. Ichikawa, Thin Solid Films, 2006, 515, 535–539.
13. R. J. Kennedy and P. A. Stampe, J. Cryst. Growth, 2003, 252, 333–342.
14. S. Yamamoto, T. Sumita, T. Yamaki, A. Miyashita and H. Naramoto, J. Cryst. Growth, 2002, 237–239, 569–573.
15. P. Fisher, O. Maksimov, H. Du, V. D. Heydemann, M. Skowronski and P. A. Salvador, Microelectron. J., 2006, 37, 1493–1497.
16. G. S. Herman, Y. Gao, T. T. Tran and J. Osterwalder, Surf. Sci., 2000, 447, 201–211.
17. C. N. Yeh, Y. M. Chen, C. A. Chen, Y. S. Huang, D. S. Tsai and K. K. Tiong, Thin Solid Films, 2010, 518, 4121–4125.
18. A. Lotnyk, S. Senz and D. Hesse, Thin Solid Films, 2007, 515, 3439–3447.
19. X. Weng, P. Fisher, M. Skowronski, P. A. Salvador and O. Maksimov, J. Cryst. Growth, 2008, 310, 545–550.
20. M. Oku, K. Wagatsuma and S. Kohiki, Phys. Chem. Chem. Phys., 1999, 1, 5327–5331.
21. S. Tominaka, Y. Tsujimoto, Y. Matsushita and K. Yanaura, Angew. Chem., Int. Ed., 2011, 50, 7418–7421.
22. Z. Lu, C. Yip, L. Wang, H. Huang and L. Zhou, ChemPlusChem, 2012, 77, 991–1000.
23. D. J. Won, C. H. Wang, H. K. Jang and D. J. Choi, Appl. Phys. A, 2001, 73, 595–600.

---