Adv. Theor. Appl. Mech., Vol. 5, 2012, no. 1, 1-10 A Plasma Emission Controller for Reactive Magnetron Sputtering of Titanium Dioxide Films Raad A. Swady DMPS, College of Arts & Sciences, University of Nizwa, Oman albdeery@unizwa.edu.om Mahmood K. Jasim DMPS, College of Arts & Sciences, University of Nizwa, Oman mahmoodkhalid@unizwa.edu.om Abstract Films of TiO 2 were synthesized at room temperature on glass substrates by dc reactive sputtering from an unbalanced magnetron using a vacuum system with low pumping speed. The unbalanced magnetron source was located in an enclosed volume of a getter box, within a vacuum chamber, into which oxygen gas was admitted. The flow of the reactive gas was controlled by monitoring the emission line of metallic titanium. Observation of the spectral line emission from sputtered titanium provided a fast response time for oxygen injection. Optical emission monitoring demonstrated a systematic control for the reactive deposition process by shifting the operational output signal to a stable position at which the sputtering rate and oxygen consumption were well controlled at each metal line set point. Optimization was assessed by measuring the absorption and refractive index.
2 R. A. Swady and M. K. Jasim Keywords: Reactive sputtering, Unbalanced magnetron, Plasma emission, Oxygen Consumption, Titanium dioxide, Absorption, Refractive index I Introduction The operation of a reactive process requires the delivery of material at a constant rate to the growing film surface. This can be achieved by using reactive magnetron sputtering. Magnetron sputtering operates at gas pressures which are sufficiently low to allow good transfer of material from cathode target to the substrate but are consistent with those needed to deliver reactive gas at sufficient rate from the residual atmosphere. They provide an easily controlled, constant-rate source for elemental sputtering, however, when operated at high rates in a low speed pumped enclosure they give rise to a serious instability, preventing the formation of optimum material. The rate of reaction on the growing film surface depends on the energy available, either in the depositing species or as substrate temperature [1]. As a dielectric film TiO 2 has been intensively investigated due to its high refractive index (2.25-2.60), high transmission in the visible and near infrared region, chemical stability, hardness, and photo catalysis for water cleavage [2-7]. Consequently, it offers a wide range of various industrial applications such as optical coatings and protective layers for very large-scale integrated circuits, surface acoustic wave gas sensor system, photocatalytic applications, and dye-sensitized solar cells [7-10]. The optical properties of TiO 2 are clearly affected by the deposition technique. The substrate temperature and the degree of film oxidation play an important role in identifying the stoichiometry or composition [2]. Optical losses in single dielectric layers as well as in multi-layers were attributed to scattering, and the change in film deposition process from thermal evaporation to sputtering tended to produce films with considerably low losses [6]. II Deposition Procedure A deposition process was used which had already demonstrated that aluminum metal films could be made of greater purity [11]. It was known that observation of the flow of reactive gas necessary to maintain a certain partial pressure could be of considerable use in predicting the required conditions to give stoichiometric materials [12,13]. The apparatus is illustrated in Figure 1. The magnetron was mounted in an open box within the vacuum chamber, which was evacuated and the sputtering of the
Plasma emission controller 3 metal initiated in a given pressure of argon. The box was semi-sealed with a sliding lid and the conditions for creating TiO x set by admitting oxygen in response to a signal from a Plasma Emission Monitor (PEM). After a period of time, sufficient to achieve stable conditions, the substrate was exposed within the box and the film was deposited. Sufficient argon leaked into the box from the surrounding vacuum chamber to maintain the sputtering atmosphere, the reactive gas being fed directly into the box itself. The magnetron current, magnetron potential, and argon gas pressure were 4 A, 360 V, and 5 mtorr respectively. The ion current density striking the growing films and the corresponding bias voltage were measured using a 1 cm 2 probe in place of the substrate. III Plasma Emission Monitor The pressure instability encountered during reactive magnetron sputtering of titanium dioxides in a low pumping speed vacuum system has already been investigated by manually adjusting the oxygen flow [14]. The appearance of hysteresis was observed when the emission intensity of titanium, magnetron potential, and total system
4 R. A. Swady and M. K. Jasim pressure were plotted against oxygen flow rate. Such a hysteresis loop reveals the presence of a surplus of reactive gas which could not be pumped from the deposition chamber. Figures 2a and 2b show the hysteretic effect. The admission of the reactive gas in a reactive sputtering process is commonly done through control of the flow with a simple mechanical valve or an electronic feedback
Plasma emission controller 5 device which can indicate the magnitude of that flow. In high rate reactive magnetron sputtering conditions are often created where the process is unstable, switching between metallic and oxide sputtering modes without allowing access to the intermediate point, which is the required transition mode. Control of the partial pressure of the reactive gas with rapid feedback can prevent this instability. This feedback process can best be done by observation of the spectral line emission of the gas or the sputtering metal; this is called Plasma Emission Monitor (PEM). The same apparatus, deposition procedure stated earlier, and process control were used to deposit reactively sputtered TiN films from a planar magnetron and reactively sputtered thin films of TiN from an unbalanced magnetron at low target potentials utilizing a resistance-heated filament [15,16]. In this deposition process an emission signal of the titanium spectral line at 451 nm was used as a process control. Figure 3 shows a schematic diagram for the feedback loop control performed by PEM. The optical filter used was a narrow band filter of transmission of 35% at 451 nm when analyzed by a spectrophotometer. IV Results and Discussion The rate of oxygen consumption and corresponding absorption as a function of the Ti line set point for reactively sputtered TiO x films is shown in Figure 4. Below a Ti line set point of 16% the PEM is not operational due to a rapid increase in the
6 R. A. Swady and M. K. Jasim reactive gas pressure and the effect on pumping. In a very narrow region of the Ti line set point of 16%, an optimum transparent TiO 2 film with absorption of less than 1 and high refractive index of 2.32 at 633 nm was deposited at a flow rate of 25 sc cm. The thickness of this film was 42 nm. The absorption (1-R-T) of the films at 400 nm was measured using a scanning spectrophotometer model (323 UV-VIS-NIR). The thickness and refractive index were determined by ellipsometry. 100 100 Oxygen Consumption Sccm 80 60 40 20 80 60 40 20 0 0 0 20 40 60 80 100 Ti Line Set point % Figure 4. The dependence of Oxygen Consumption and Absorption of TiO 2 Films on Ti Line Set Point Absorption % As the sputtering rate increases both the films absorption and flow rate increase reaching a maximum of 90 % and 82 sc cm respectively at a Ti line set point of 73%. This indicates the incorporation of more oxygen atoms in the growing films. When the sputtering rate is further increased the incorporation of metallic titanium dominates and the absorption decreases implying the disappearance of the Ti x O x phase due to the decrease of oxygen content in the film which eventually reaches zero at a Ti line of 100%. To get more insight in to the control process of a Ti line signal on the reactive gas consumption during deposition and corresponding optimization, a correlation between oxygen consumption and film absorption is required; this correlation is plotted in Figure 5 for films shown in Figure 4.
Plasma emission controller 7 As can be clearly seen an optimum titanium dioxide film is synthesized for the lowest consumption of reactive gas within the control range of the plasma emission signal of the sputtered titanium. It can also be concluded that maximum absorption occurs at the maximum consumption of oxygen. Figure 5 reveals the fine control obtained by matching the input signal with the required injection of oxygen at each percentage of the titanium input signal as characterized by the deposition process. It was demonstrated that control of unstable reactive sputtering processes can be achieved by modulating the admission of the reactive gas through the spectral line emission of the gas or the sputtering metal in the magnetron plasma [17, 18]. It was shown that a measurement of the flow of the gas into the film-making process can allow prediction of film properties and provide an indication of the efficiency of the reaction kinetics [17]. An automatic control of the reactive magnetron sputtering conditions of TiO x films was attained by monitoring the emission line signal of metallic Ti and using it as a feedback to operate the gas admission system [19]. Comparing the refractive index of an optimum film (n=2.32) prepared in the getter box with other references it can be concluded that this value is competitive for films deposited at ambient temperature in a low pumping speed vacuum system. Values of n at wavelengths of 600 nm and 650 nm were reported to be 2.27 and 2.28 respectively for films deposited at a magnetron current of 6 A onto a glass substrate at room temperature [20]. A value of n of 2.41 at 650 nm for a TiO 2 film grown on a 300 ºC was obtained. However, consideration of the power or current levels and substrate heating used by (Ref. 20), it reveals that the improvement of optical properties of TiO 2 films prepared in the getter box is due to the effect of the unbalanced magnetron. This
8 R. A. Swady and M. K. Jasim provides a plasma beam that leaks on to the substrate bombarding the growing film and kinetically activating the nucleation during deposition. During films deposition, the ion current density and bias voltage were 7 ma/cm 2 and 15 Volts respectively. Films prepared by reactive evaporation (RE), ion assisted deposition (IAD) and ion beam sputtering (IBS) resulted in refractive indices at 625 nm of 2.27, 2.43, and 2.49 respectively [2]. The effect of ion bombardment during film growth is to increase the density of the films and subsequently enhancing the optical properties [21]. The influence of deposition parameters such as partial pressure of oxygen, rate of deposition, and substrate temperature on the optical properties of TiO 2 was also investigated [22]. These films were deposited using reactive electron beam evaporation of TiO, Ti 2 O 3, and TiO 2 in a neutral and ionized oxygen atmosphere. Films deposited at substrate temperature of 175 ºC had refractive indices of 2.25 and 2.24 in neutral and ionized oxygen respectively. V Conclusions It has been demonstrated that a plasma emission monitor PEM can control the reactive magnetron sputtering in a low pumping speed vacuum system in conjunction with a getter box. The automatic control of the reactive sputtering using the Ti spectral line signal from the discharge plasma as a feedback, resulted in the growth of optimum films with competitive optical qualities, characterized by very low absorption and high refractive index. Acknowledgments The authors wish to express their sincere thanks to the referees, Editorial Board, as well as to Dr. Emil Minchev President of Hikari Ltd Managing Editor of Advances in Theoretical and Applied Mechanics Journal. References [1] R. P. Howson, H. A. Ja fer and A. G. Spencer, Thin Solid films, 193/194, (1990) 127. [2] J. J. Cuomo, S. M. Rossnagel, and H. R. Kauffman, Handbook of ion beam processing Technology, Park Ridge, New Jersey, USA, 1990, Ch. 19, P. 395.
Plasma emission controller 9 [3] W. T. Pawlewicz, D. D. Hays, and P. M. Martin, Thin Solid Films, 73 (1980)169. [4] R. P. Howson, K. Suzuki, C. A. Bishop, and M. I. Ridge, Vacuum, 34 (1984) 291. [5] J. M. Bennett, E. Pelletier, G. Albrand, J. P. Borgongo, B. Lazarides, C. K. Carniglia, R. A. Schmell, T. H. Allen, T. Tuttle-Heart, K. H. Gunther, and A. Saxer, Appl. Opt., 28 (1989) 3303. [6] H. K. Pulker, Thin Solid Films, 34 (1976) 343. [7] C. H. Heo, S. B. Lee and J. H. Boo, Thin Solid Films, 475 (2005) 183 [8] W. P. Jakubik, Molecular and Quantaum Acoustics, 27 (2006) 133 [9] T. Sakai, Y. Kuniyoshi, W. Aoki, S. Ezoe, T. Endo, and Y. Hoshi, Jpn. J. Appl. Phys. 47 (2008) 6548 [10] M. Chigane, M. Watanabe, M. Izaki, I Yamaguchi, and T. Shinagawa, Electrochem. Solid-State Lett., V 12, Issue 5, (2009) E5 [11] R. P. Howson and A. G. Spencer, Proc.32 nd Technical Conference of the Society of Vacuum Coaters-St. Louis (Society of Vacuum Coaters), 1989, P. 40. [12] R. P. Howson, H. Barankova, and A. G. Spencer, Thin Solid Films, 196 (1991) 315. [13] B Hichwa, and G. Caskey, Surf Coat Technol V. 33 Nov 1987, Pap Presented at the 14 th Int. Conf on Metall Coat Part 2, San Diego, CA, USA, Mar 23-27 1987 P. 393. [14] R. A. Swady, Caledonian Journal of Engineering, 5, 1(2009) 6. [15] Raad. A. Swady and R. P. Howson, ion plasma assisted techniques (IPAT), Brussels, Belgium, 1990, 346-351. [16] R. A. Swady, H. A. Ja fer, and R. P. Howson, Vacuum, 44, ¾ (1993) 297.
10 R. A. Swady and M. K. Jasim [17] R. P. Howson, R. A. Swady, A. G. Spencer, and H. Barankova, IPAT (ion plasma assisted techniques), Brussels, Belgium,,1990, 340-345. [18] G. S. Chen, C. C. Lee, h. Niu, W. Huang, R. Jann, and T. Schutte, Thin Solid Films, 516, 23 (2008) 8473. [19] K. Zakrzewska, A. Brudnik, M. Radecka, and W. Posadowski, Thin Solid Films, 343-344 (1999) 152. [20] M. Georgson, A. Roos, and C. G. Ribbing, J. Vac. Sci. Technol., A9 (1991) 2191. [21] J. J. Mcnally, G. A. Al-Jumaily, and J. R. McNeil, J. Vac. Sci. Technol., A4 (1986) 437. [22] K. N. Rao, OPt. Eng. 41, 9 (2002) 2357. Received: May, 2011