Full Paper Light-Induced CVD of Titanium Dioxide Thin Films I: Kinetics of Deposition** By Estelle Halary-Wagner, Tristan Bret, and Patrik Hoffmann* Titanium dioxide thin films are obtained by CVD on low temperature substrates (60±210 C) using perpendicular irradiation from a long pulse 308 nm excimer laser. Titanium tetra-isopropoxide is used as the precursor in an oxygen-containing atmosphere with a total pressure in the chamber of 10 mbar. An empirical law describing the growth rate as a function of the experimental parameters (substrate temperature and laser fluence) is derived. The deposited thickness is proportional to the number of photons and has an Arrhenius dependence on the substrate temperature. Low growth rates per pulse (less than 0.1 nm per pulse) obtained allow a good control of the deposited thickness, while the use of high laser repetition rates leads to high temporal growth rates (up to 100 nm min ±1 ). Keywords: Excimer Laser, Growth rate, Kinetics, Titanium dioxide 1. Introduction Light-induced (LI) CVD uses photon irradiation to promote or activate a CVD reaction. [1] Different kinds of light sources (lasers, lamps) and irradiation modes (perpen- dicular or parallel irradiation, focused or non-focused beams) have been proposed. [2] Among these, the use of excimer lasers (powerful pulsed light sources in the UV range) directly irradiating the substrate was shown to be a promising technique for the deposition of thin films at low substrate temperatures. In particular, a large number of oxides (TiO 2 , [3,4] Ta 2 O 5 , [5] SiO 2 , [6] ZrO 2 , [7] PbO, [3] ZnO, [8] Al 2 O 3 , [9] SnO 2 [10] ) have been deposited by this method, usually using an organometallic compound combined with an oxidant as the reactive species, and mainly using ArF (193 nm) or KrF (248 nm) lasers. We focus here on TiO 2 deposition from titanium tetra- isopropoxide (TTIP) in an oxygen-containing atmosphere induced by irradiation by a long pulse (250 ns) XeCl excimer laser (308 nm). This system is particularly interesting as; i) deposition is possible at substrate temperatures as low as room temperature, and deposition was shown even on polymer substrates, [11] ii) localized deposition in the irra- diated area is demonstrated with a lateral resolution of 10 lm, [12] and iii) crystallinity of the deposits can be tailored (to amorphous, anatase, or rutile phase) by varying the irradiation conditions. [13] In this paper, the kinetics of the deposition reaction are studied in detail and a complete empirical law, describing the growth rate as a function of the process parameters, is derived. The interpretation of this formula gives additional insights into the deposition mechanism, which includes both photolytic and thermal contributions. 2. Results The effects of the ªenergy input parametersº (substrate temperature T and irradiation parameters; laser fluence F , laser repetition rate f , and number of pulses N) on the deposited thickness are studied systematically. Deposits are obtained at all tested substrate temperatures (60±210 C) and all tested laser fluences (1±400 mJ cm ±2 ). 2.1. Derivation of the Empirical Equation for the Deposited Thickness Keeping the deposition time constant (t = 20 min), the variation of the deposited thickness was studied systemati- cally on glass substrates for five pulse repetition rates (f = 1 Hz, 5 Hz, 10 Hz, 15 Hz, and 20 Hz), three fluences (F = 100 mJ cm ±2 , 200 mJ cm ±2 , and 400 mJ cm ±2 ) and three substrate temperatures (T = 60 C, 135 C, and 210 C). 2.1.1. Influence of the Number of Pulses and Laser Repetition Rate The variation in the growth rate as a function of the number of pulses (obtained for different laser repetition rates) for the nine different conditions of substrate temper- ature and fluence (T ,F) is presented in Figure 1. The growth rate varies linearly with the number of pulses, independently Chem. Vap. Deposition 2005, 11, No. 1 DOI: 10.1002/cvde.200306303 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 21 ± [*] Dr. P. Hoffmann, Dr. E. Halary-Wagner, Dr. T. Bret Advanced Photonics Laboratory Ecole Polytechnique FØdØrale de Lausanne (EPFL) CH-1015 Lausanne-EPFL (Switzerland) E-mail: patrik.hoffmann@epfl.ch [**] This research was supported by grants from the Swiss National Science Foundation (project no. 20-59404.99). The authors gratefully acknowl- edge James A. DeRose for proofreading the manuscript.