Photocatalysis of Thin Films of TiO2 on Al2O3 Substrates 1, Timothy Luttrell2, Matthias Batzill2 David Turbay Department of Physics, Brown

University1, University of South Florida2

Introduction and Motivation
Titanium dioxide (TiO2) has grown to be one of the most promising photocatalysts in recent years because of its extensive applications in renewable and clean (Fig. 1) [3] energy. The rise in demand for these new energies has driven an increase in research on metal oxides and their properties. We were interested in growing the rutile structure of TiO2 because it has a lower excitation energy of 3.0 eV compared to anatase (3.2 eV) and, therefore, has better activity in the visible portion of the spectrum [2]. It has been shown that Al2O3 (sapphire) substrates are conducive to epitaxial rutile growth [1]. In this study, we measured the photocatalytic activity of thin films of TiO2 on r-Al2O3 (1 -1 0 2) substrates. We used pulse laser deposition (PLD) and molecular beam epitaxy (MBE) to grow the films. The films were characterized using x-ray photoelectron spectroscopy (XPS), and atomic force microscopy (AFM). Photoactivity was measured via the decomposition of methyl orange on the films surface using a UV/VIS spectrophotometer. The decomposition of this organic compound is driven by oxidation-reduction reactions on the surface of the TiO2 film (fig. 1). From this, we calculated the charge carrier diffusion length and compared it to that of anatase. The goals of the project were to grow epitaxial films of rutile TiO2, and compare the carrier diffusion lengths of anatase and rutile to help explain the higher activity of anatase. With this thickness, we were able to find the rate of deposition to determine the thickness of all subsequent films noting that there is an approximate linear relationship between thickness and time of deposition. We did further characterization with the AFM which provided high-quality small scale images. The absorption of the methyl orange was measured using a spectrophotometer, and the change in this quantity gave us a direct measure of the photocatalytic activity of the film. We analyzed the dependence of photoactivity on film thickness and determined the charge carrier diffusion length by fitting the data to the one dimensional model of diffusion length from the bulk to the surface:
Fig. 4

d C C0 d A e dt

(Eq. 2)
Fig. 5

In equation 2, A is an initial condition constant, d is the thickness of the film, and is the average mean free path of the charge carriers. The left hand side quantity gives us a direct measure of the photocatalytic activity.

Data & Results

For 10 minutes of deposition, we were able to estimate the thickness of the film to be 4.40.8nm using equation 1.

Characterization
We obtained the following AFM images of the substrate and films:
(a)
Fig. 2(a): A one micron noncontact resolution image of a clean sapphire substrate. The atomic steps are very clear.

UV- Catalysis
The spectrophotometer gave us the following plots . We can quantify the rate of decomposition by measuring the area under each curve.
Fig. 3

The slope of the fitted lines (Fig. 4) provided a comparable measurement of the photoactivity of the film, so a relationship between thickness and photoactivity was established. In figure 5, photoactivity was plotted against film thickness along with data previously found by another party in this lab. The data was fit with a function of the same form of Equation 2 and the fitting parameters gave us the appropriate values for the carrier diffusion length for both TiO2 structures. The diffusion length for anatase is 6.397 nm greater than that of rutile.

Experimental Procedure
Sapphire substrates were cleaned in an ultrasonic bath of acetone and ethanol, 10 minutes each. For film growth, we used PLD and MBE, which have shown to have very similar deposition properties with TiO2. Because of time constraints, we found the MBE via electron beam evaporization more the more feasible method with no significant drawbacks. The films grown by MBE had a constant temperature of approximately 600C, oxygen pressure of 2.0x10-6 Torr (background 2.0x10-8 Torr), and constant electron flux. XPS scan for the fist film grown by MBE was used to estimate the thickness of the film using the following equation:
(b)

Conclusion
Using the AFM and XPS, we were able to see that we were able to grow epitaxial thin films of TiO2 on sapphire (1 -1 0 2) using pulse laser deposition and molecular beam epitaxy. We were not able to get conclusive empirical evidence confirming the growth of rutile using x-ray diffraction (XRD) or reflective high-energy electron diffraction (RHEED) techniques. However, noting that activity was substantially lower than that of anatase, and that the AFM images were flat and ordered, we can speculate rutile (0 1 1) was grown, as this is the structure that grows on r-Al2O3 (1 -1 0 2) [5]. Furthermore, we saw that the charge carrier diffusion length was much higher in anatase. This suggests that the higher photoactivity of anatase can be attributed to anatase having a larger photoactive region than rutile does i.e. it is more probable for charge carriers from deeper inside the bulk to diffuse to the surface and set off photoreactions.
References: [1] M.R. Bayati, Sh. Joshi, R. Molaei, R.J. Narayan, J. Narayan, Structureproperty correlation in epitaxial (200) rutile films on sapphire substrates, Journal of Solid State Chemistry, Volume 187, March 2012, Pages 231-237 [2] Sin-iti Kitazawa, Yeongsoo Choi, Shunya Yamamoto, Tetsuya Yamaki, Rutile and anatase mixed crystal TiO2 thin films prepared by pulsed laser deposition, Thin Solid Films, Volume 515, Issue 4, 5 December 2006, Pages 1901-1904 [3] http://www.titanpe.com/library/kb2502.htm [4] Nix, Roger. Introduction to Surface Chemistry, Sec. 5.1. http://www.chem.qmul.ac.uk/surfaces/scc/scat5_1.htm [5] Bharat Jalan, Roman Engel-Herbert, Joel Cagnon, and Susanne Stemmer, Growth modes in metal-organic molecular beam epitaxy of TiO[sub 2] on r-plane sapphire, J. Vac. Sci. Technol. A 27, 230 (2009)

Fig. 2(a): A one micron image of the 4.4 nm film. We can see that its smooth from the line spectrum.

(c)
Fig. 2(a): A one micron image of 13.2 nm film. We can see that its smooth from the line spectrum. Fig. 3: The spectrum measured by the spectrophotometer for the 26.4 nm film. Each curve represents a reading taken every thirty minutes.

I Ti I Al

I Ti ,o e I Al ,o e

Ti Al

dz dz

(Eq. 1)

(d)
Fig. 2(a): A one micron image of 26.4 nm film. We can see that its smooth from the line spectrum.

IX,o and IX are the initial and final XPS intensities (respectively) of Ti or Al, d is the thickness of the film, and X is the inelastic mean free path of Ti or Al [4].

The area under each curve was computed using a data analysis software. A correlation between irradiation time and area was observed for each film. The results are shown in Figure 4.

REU Poster Symposium August 1, 2012 - Funding provided by NSF Grant DMR-1004873