- Email: [email protected]
(Pb,Sr)TiO3 thin films etching characteristics using inductively coupled plasma
Thin Solid Films 447 – 448 (2004) 688–692
(Pb,Sr)TiO3 thin films etching characteristics using inductively coupled plasma Gwan-Ha Kim, Kyoung-Tae Kim, Dong-Pyo Kim, Chang-Il Kim* School of Electrical and Electronics Engineering, Chungang University, 221, Huksuk-Dong, Dongjak-Gu, Seoul 156-756, South Korea
Abstract Etching characteristics of (Pb,Sr)TiO3 (PST) thin films were investigated using inductively coupled chlorine-based plasma system as functions of gas-mixing ratio, radio frequency power and direct current bias voltage. It was found that increasing of Ar content in gas mixture lead to sufficient increasing of etch rate and selectivity of PST to Pt. The maximum etch rate of PST film is 56.2 nmymin and the selectivity of PST film to Pt is 0.8 at Cl2 y(Cl2 qAr) of 20%. It was proposed that sputter etching is the dominant etching mechanism while the contribution of chemical reaction is relatively low due to low volatility of etching products. 䊚 2003 Elsevier B.V. All rights reserved. Keywords: Etching; (Pb,Sr)TiO3; Optical emission spectroscopy; Langmuir probe
1. Introduction There is an increased research activity to search appropriate thin film materials for dynamic random access memories (DRAM) and microwave dielectric applications. Among the various dielectric films, the (Ba,Sr)TiO3 (BST) thin film was noticed as the most promising material for the capacitor dielectric of future high-density DRAM because its high dielectric constant with good thermal stability and paraelectricity at normal operating temperature w1–3x. Although BST possesses a satisfactorily high dielectric constant, it was known that a post heat treatment at a high temperature was essential to obtain good electrical property. However, the heat treatment at high temperature can cause deleterious effects on an electrode, barrier metal and contact plug w4 x . Strontium titanate (ST) is one of the few titanates, which is cubic phase at room temperature. Tetragonal lead titanate (PT) films have higher dielectric constant than ST films at room temperature and higher crystallization temperature than BST. The addition of lead into ST makes its dielectric constant higher and the temperature of crystallization lower w5,6x. Therefore, (Pb,Sr)TiO3 (PST) thin film can be a promising material for the capacitor dielectric of future high-density gigabit *Corresponding author. Tel.: q82-2-820-5334; fax: q82-2-8129651. E-mail address: [email protected] (C.-I. Kim).
DRAM application because its high dielectric constant and paraelectricity at normal operating temperature. However, there is no report on the characteristics and mechanism of PST thin films in terms of plasma parameters including gas-mixing ratio, radio frequency (RF) power and direct current (DC) bias voltage. In this study, inductively coupled plasma etching system was used for PST etching because of its high plasma density, low process pressure and easy control bias power. The dry etching of the PST films was studied using Cl2 yAr gas chemistry by varying the concentration of the etch gases. Systematic studies were carried out as a function of the etching parameters, including the RF power and the DC bias voltage to the substrate. The Cl2 yAr plasmas were characterized by optical emission spectroscopy (OES) and Langmuir probe analysis. The volume densities of the plasma were monitored by a Langmuir probe and the changes of chemical composition in the chamber was analyzed with OES. 2. Experiment Thin films of (Pb0.6Sr0.4)TiO3 with excess Pb-acetate of 10 mol% were prepared using lead acetate trihydrate, acetate and titanium iso-propoxide as the starting materials. Acetic acid and 2-methoxyethanol were used as the solvent for the sol–gel method. PST precursor solutions were spin-coated on the Pt (120 nm)yTi (30
0040-6090/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2003.09.007
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
Fig. 1. The etch rate of PST as well as the selectivity of PST to Pt as a function of Cl2yAr mixing ratio.
nm)ySiO2 ySi substrate, and then pre-baked on a hot plate at 400 8C for 10 min to remove the organic materials. The pre-baked films were annealed at 650 8C for 1 h in oxygen atmosphere for crystallization. The final thickness of PST thin film was approximately 200 nm. The PST thin films were etched as a function of Cl2 y Ar plasma gas-mixing ratio. Then, PST thin films were etched as functions of RF power and DC bias voltage. Etch rates were measured by using a surface profiler (Tencor, a-step 500). The volume densities of the plasma were monitored by a Langmuir probe (HIDEN, ESP) and the changes of chemical composition in the chamber was analyzed with OES (SC TCEH PCM 420).
689
as etch product during the PST film etching with chlorine-based plasma chemistries w7x. But, due to the low volatility of SrCl2 compounds, the physical ion bombardment effect to enhance the ion assisted etch product desorption mechanism is required. The etch rate of the Pt decreases as the Cl2 yAr gas-mixing ratio increases. This means that sputtering by Ar ions is more effective than the chemical reaction by Cl radicals. Consequently, the selectivity of the PST film to Pt increases as the Cl2 yAr gas-mixing ratio increases as shown in Fig. 1. The maximum etch rate of PST film is 56.2 nmymin and the selectivity of PST film to Pt is 0.8 at Cl2 y(Cl2qAr) of 20%. Fig. 2 shows the effect of RF power on the etch rate of PST and Pt, and its etch selectivity to Pt. Cl2 y(Cl2q Ar) gas combination of 20% was used in the experiment and other process condition were the same as those shown in Fig. 1. As the RF power increases from 500 to 800 W, the etch rate of PST and Pt increases from 24.2 to 60.3 nmymin and from 31.5 to 95.3 nmymin, respectively. The increase of PST etch rate with the increase of RF power appears to be from the increase of reactive chlorine radicals and ion density with the increase of RF power. Meanwhile, the selectivity of PST to Pt seems to be constant at ;0.84. Fig. 3 shows the etch rate of PST and Pt, and the selectivity of PST to Pt as a function of DC bias voltage. Cl2 yAr gas mixture, RF power, chamber pressure and substrate temperature were Cl2 y(Cl2 qAr) of 20%, 700 V, 2 Pa, 25 8C, respectively. As the DC bias voltage increases from y100 to y250 V, the etch rate of PST and Pt increases from 35.9 to 70.4 nmymin and from 40.2 to 85.6 nmymin, respectively. On the other hand,
3. Results and discussion The PST films were etched as a function of the Cl2 y Ar gas-mixing ratio in an ICP etching system. Fig. 1 shows the etch rates of PST films and Pt, and selectivity of PST film to Pt as a function of the Cl2 yAr gasmixing ratio when total flow rate was maintained at 20 sccm. Other process conditions such as RF power, DC bias, chamber pressure and substrate temperature were also maintained at 700 W, y150 V, 2 Pa, 25 8C, respectively. As the gas-mixing ratio increases to 20% Cl2, the etch rate of the PST film increases. However, when the gas-mixing ratio exceeds more than 20% Cl2, the etch rate of the PST film decreases. This result confirms that PST film was etched not only by ion bombardment but also by chemical reaction. PbCl4 (boiling point: 50 8C), SrCl2 (boiling point: 1250 8C) and TiCl4 (boiling point: 136 8C) are expected to form
Fig. 2. The etch rate of PST as well as the selectivity of PST to Pt as a function of RF power.
690
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
Fig. 3. The etch rate of PST as well as the selectivity of PST to Pt as a function of DC bias voltage.
the selectivity of PST to Pt seems to be constant, since the ion bombardment effect is similar to PST and Pt etching. To understand the roles of physical and chemical effects, the data about the influence of Cl2 addition on volume densities of active particles are needed. For these purposes we used OES analysis. For the control of active species volume densities behavior in Cl2 yAr plasmas, we selected such emission maximums as atomic lines of Cl (452.6 nm), Ar (750.4 nm) and molecular band of Cl2 (307.4 nm). These maximums are frequently used for analytical purposes including optical emission actinometry to determine absolute densities of active particles. Fig. 4 illustrates variations of emission intensities of selected maximums as a function of Cl2 concentration in gas mixtures. It is evidently clear that the increasing of Ar content in Cl2 yAr mixture leads to a direct proportional increase of emission intensity of Ar atoms. Note that Ar (750.4 nm) line is characterized by maximal excitation threshold energy among all selected maximums and thus its excitation constant should be maximal sensitive to electron energy distribution function (EEDF) variations, especially in the region of highenergy ‘tail’. Therefore, OES data confirm our conclusion that changing of Ar fraction in gas mixtures does not lead to sufficient changes of multiplication of excitation constant and electron density. To our mind it is possible to assume that the same behavior will be obtained for Arq ions volume density. The reason is that ionization rate is to follow the on the behavior of Ar ground state atoms while the rate of decay of Arq ions due to diffusion to reactor walls is expected to be
practically constant under the constant pressure conditions. As for Cl2 yAr plasma, the fact of primary interest is the constancy of chlorine atoms volume density up to 40% addition of Cl2. Effect of constancy of chlorine atoms volume density may be explained by increasing dissociation degree of Cl2 molecules during Ar addition. Analysis of data reported above allows making some conclusions concerning etching mechanism of PST thin films. It is evidently clear that more than twofold increasing of etching rate in gas mixtures up to 20% addition of Cl2 cannot be supported only by chemical mechanism, because volume densities of chlorine atoms have a tendency to decreasing. To our mind, the increasing of etching rate may be explained by two factors. First factor is connected with acceleration of physical sputtering of both main material such as PST film and surface layer of reaction products. The evident contribution of physical sputtering in etching process is confirmed by the fact that etching rate in pure Ar is sufficiently more than in Cl2 gases. Second factor represents the sequence of first one and connected with acceleration of chemical interaction through the increasing of reaction probability and destruction of metal– oxide bonds. In this case, extreme behavior of etching rate and decreasing of etching rate at Cl2 content less than 20% may be connected with decreasing and disappearing of chemical channel due to rapid decreasing of volume density of chemical active species. To investigate the influence of Cl2 yAr mixing ratio on mass content of plasma active species, we experimentally measured electron temperature and electron density as input parameters. Volume densities of charged particles in Cl2 yAr plasma were calculated during the
Fig. 4. Emission intensities of Cl (452.6 nm), Cl2 (307.4 nm) and Ar (750.4 nm) as a function of Cl2yAr mixing ratio.
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
691
Fig. 6. Volume densities of positive ions as a function of Cl2yAr mixing ratio. Fig. 5. Ionization rates of positive ions as a function of Cl2yAr mixing ratio.
simultaneous solution of balance quasi-stationary kinetic equations and quasi-neutrality equation. And, to determine volume densities of neutral particles, we used the combination of mass balance equation and the equation of chemical kinetics for chlorine atoms. Fig. 5 shows that the results of the rates of the processes supported formation of active particles in Cl2 y Ar plasma. Data show that addition of Cl2 leads to rapid increase of dissociation rate of Cl2 molecules. The reason is that Cl2 dissociation process is characterized by relatively low threshold energy (3.4 eV) and thus dissociation rate coefficient is not very sensitive to the changes of electron temperature. Therefore, increase of dissociation rate coefficient is overcompensated by the decreasing electron density and Cl2 dissociation rate increases monotonously. Note that the data of Fig. 5 reflect only the behavior of direct electron impact dissociation rate. Although dissociative ionization and dissociative attachment are considered as possible sources of chlorine atoms, the contribution of these processes is insignificant. As Cl2 content in Cl2 yAr plasma increases from 2 to 98%, the rate of dissociative attachment increases from 3.8=1016 to 4.4=1014 cmy3 sy1 while the rate of dissociative ionization changes in the scopes of 1.1=1016 –6.3=1014 cmy3 sy1. Therefore, our data show that electron impact dissociation of Cl2 molecules may be considered as the main source of chlorine atoms in plasma volume. This conclusion is in good agreement with the published result concerning both pure Cl2 plasma and Cl2 yAr mixture. The neutral particles kinetics indicate that the behavior of the volume density of Cl atoms follows the behavior of dissociation rate. Therefore, increasing of Ar content
in Cl2 yAr mixture leads to monotonous decreasing volume density of chlorine atoms. Corresponding data are represented in Fig. 6. As shown in Fig. 5, total ionization rate (as the sum of partial values) has a weak tendency to increasing in Ar-rich plasma because partial ionization rates of all components change monotonously. However, this effect is compensated by the increase of heterogeneous decay of positive ions through the increase of ion Bohm velocity and, as the result, total density of positive ions decreases in all the scopes of Ar content in Cl2 yAr mixture. Additionally, calculations of relative volume density of negative ions (bsnyne) give the values in the scopes of 0.1–0.01. At the same time, it was found that in high-pressure chlorine plasma (100 Pa), the value of relative density of negative ions exceeds 100. Therefore, our data confirms that at low pressures chlorine plasma transforms from high electronegative to moderately or low electronegative form. The data concerning the influence of Cl2 yAr mixing ratio on volume densities of positive ions are represented in Fig. 6. 4. Conclusion Etching characteristics of PST thin films were investigated using inductively coupled chlorine-based plasma system as functions of gas-mixing ratio, RF power and DC bias. It was found that increasing of Ar content in gas mixture lead to sufficient increasing of etching rate and selectivity of PST to Pt. The maximum etch rate of PST film is 56.2 nm miny1 and the selectivity of PST film to Pt is 0.8 at Cl2 y(Cl2qAr) of 20%. Due to the relatively low volatility of by-products formed during the etching by Cl2 yAr plasma, ion bombardment in addition to chemical reaction was required to obtain
692
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
high PST etch rates. It was proposed that sputter etching is the dominant etching mechanism while the contribution of chemical reaction is relatively low due to low volatility of etching products. References w1x R.B. Khnmankar, M.A. Kressley, M.R. Visokay, T. Moise, G. Xing, S. Nemoto, Y. Okuno, S.J. Fang, A.M. Wilson, J.F. Gaynor, T.Q. Hurd, D.L. Crenshaw, S. Summerfelt, L. Colombo, IEDM Dig. Technol. 4 (1997) 245.
w2x S. Yamamichi, P.Y. Lesaicherre, H. Yamaguchi, K. Takemura, S. Sone, H. Yabuta, K. Sato, T. Tamura, K. Nakajima, S. Ohnishi, K. Tokashiki, Y. Hayashi, Y. Miyasaka, M. Yoshida, H. Ono, IEEE Trans. Elect. Dev. 44 (1997) 1076. w3x M. Kiyotoshi, K. Eguchi, Electrochem. Soc. Proc. 144 (1997) 1063. w4x H.J. Chung, S.J. Chung, J.H. Kim, S.I. Woo, Thin Solid Films 394 (2001) 213. w5x D.H. Kang, J.H. Kim, J.H. Park, K.H. Yoon, Mat. Res. Bull. 36 (2001) 265. w6x H.J. Chung, S.I. Woo, J. Vac. Sci. Technol. B 19 (2001) 275. w7x David R. Lide (Ed.), CRC Handbook of Chemistry and Physics, CRC Press LLC, Washington, DC, 1998, pp. 4–66.
(Pb,Sr)TiO3 thin films etching characteristics using inductively coupled plasma Gwan-Ha Kim, Kyoung-Tae Kim, Dong-Pyo Kim, Chang-Il Kim* School of Electrical and Electronics Engineering, Chungang University, 221, Huksuk-Dong, Dongjak-Gu, Seoul 156-756, South Korea
Abstract Etching characteristics of (Pb,Sr)TiO3 (PST) thin films were investigated using inductively coupled chlorine-based plasma system as functions of gas-mixing ratio, radio frequency power and direct current bias voltage. It was found that increasing of Ar content in gas mixture lead to sufficient increasing of etch rate and selectivity of PST to Pt. The maximum etch rate of PST film is 56.2 nmymin and the selectivity of PST film to Pt is 0.8 at Cl2 y(Cl2 qAr) of 20%. It was proposed that sputter etching is the dominant etching mechanism while the contribution of chemical reaction is relatively low due to low volatility of etching products. 䊚 2003 Elsevier B.V. All rights reserved. Keywords: Etching; (Pb,Sr)TiO3; Optical emission spectroscopy; Langmuir probe
1. Introduction There is an increased research activity to search appropriate thin film materials for dynamic random access memories (DRAM) and microwave dielectric applications. Among the various dielectric films, the (Ba,Sr)TiO3 (BST) thin film was noticed as the most promising material for the capacitor dielectric of future high-density DRAM because its high dielectric constant with good thermal stability and paraelectricity at normal operating temperature w1–3x. Although BST possesses a satisfactorily high dielectric constant, it was known that a post heat treatment at a high temperature was essential to obtain good electrical property. However, the heat treatment at high temperature can cause deleterious effects on an electrode, barrier metal and contact plug w4 x . Strontium titanate (ST) is one of the few titanates, which is cubic phase at room temperature. Tetragonal lead titanate (PT) films have higher dielectric constant than ST films at room temperature and higher crystallization temperature than BST. The addition of lead into ST makes its dielectric constant higher and the temperature of crystallization lower w5,6x. Therefore, (Pb,Sr)TiO3 (PST) thin film can be a promising material for the capacitor dielectric of future high-density gigabit *Corresponding author. Tel.: q82-2-820-5334; fax: q82-2-8129651. E-mail address: [email protected] (C.-I. Kim).
DRAM application because its high dielectric constant and paraelectricity at normal operating temperature. However, there is no report on the characteristics and mechanism of PST thin films in terms of plasma parameters including gas-mixing ratio, radio frequency (RF) power and direct current (DC) bias voltage. In this study, inductively coupled plasma etching system was used for PST etching because of its high plasma density, low process pressure and easy control bias power. The dry etching of the PST films was studied using Cl2 yAr gas chemistry by varying the concentration of the etch gases. Systematic studies were carried out as a function of the etching parameters, including the RF power and the DC bias voltage to the substrate. The Cl2 yAr plasmas were characterized by optical emission spectroscopy (OES) and Langmuir probe analysis. The volume densities of the plasma were monitored by a Langmuir probe and the changes of chemical composition in the chamber was analyzed with OES. 2. Experiment Thin films of (Pb0.6Sr0.4)TiO3 with excess Pb-acetate of 10 mol% were prepared using lead acetate trihydrate, acetate and titanium iso-propoxide as the starting materials. Acetic acid and 2-methoxyethanol were used as the solvent for the sol–gel method. PST precursor solutions were spin-coated on the Pt (120 nm)yTi (30
0040-6090/04/$ - see front matter 䊚 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2003.09.007
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
Fig. 1. The etch rate of PST as well as the selectivity of PST to Pt as a function of Cl2yAr mixing ratio.
nm)ySiO2 ySi substrate, and then pre-baked on a hot plate at 400 8C for 10 min to remove the organic materials. The pre-baked films were annealed at 650 8C for 1 h in oxygen atmosphere for crystallization. The final thickness of PST thin film was approximately 200 nm. The PST thin films were etched as a function of Cl2 y Ar plasma gas-mixing ratio. Then, PST thin films were etched as functions of RF power and DC bias voltage. Etch rates were measured by using a surface profiler (Tencor, a-step 500). The volume densities of the plasma were monitored by a Langmuir probe (HIDEN, ESP) and the changes of chemical composition in the chamber was analyzed with OES (SC TCEH PCM 420).
689
as etch product during the PST film etching with chlorine-based plasma chemistries w7x. But, due to the low volatility of SrCl2 compounds, the physical ion bombardment effect to enhance the ion assisted etch product desorption mechanism is required. The etch rate of the Pt decreases as the Cl2 yAr gas-mixing ratio increases. This means that sputtering by Ar ions is more effective than the chemical reaction by Cl radicals. Consequently, the selectivity of the PST film to Pt increases as the Cl2 yAr gas-mixing ratio increases as shown in Fig. 1. The maximum etch rate of PST film is 56.2 nmymin and the selectivity of PST film to Pt is 0.8 at Cl2 y(Cl2qAr) of 20%. Fig. 2 shows the effect of RF power on the etch rate of PST and Pt, and its etch selectivity to Pt. Cl2 y(Cl2q Ar) gas combination of 20% was used in the experiment and other process condition were the same as those shown in Fig. 1. As the RF power increases from 500 to 800 W, the etch rate of PST and Pt increases from 24.2 to 60.3 nmymin and from 31.5 to 95.3 nmymin, respectively. The increase of PST etch rate with the increase of RF power appears to be from the increase of reactive chlorine radicals and ion density with the increase of RF power. Meanwhile, the selectivity of PST to Pt seems to be constant at ;0.84. Fig. 3 shows the etch rate of PST and Pt, and the selectivity of PST to Pt as a function of DC bias voltage. Cl2 yAr gas mixture, RF power, chamber pressure and substrate temperature were Cl2 y(Cl2 qAr) of 20%, 700 V, 2 Pa, 25 8C, respectively. As the DC bias voltage increases from y100 to y250 V, the etch rate of PST and Pt increases from 35.9 to 70.4 nmymin and from 40.2 to 85.6 nmymin, respectively. On the other hand,
3. Results and discussion The PST films were etched as a function of the Cl2 y Ar gas-mixing ratio in an ICP etching system. Fig. 1 shows the etch rates of PST films and Pt, and selectivity of PST film to Pt as a function of the Cl2 yAr gasmixing ratio when total flow rate was maintained at 20 sccm. Other process conditions such as RF power, DC bias, chamber pressure and substrate temperature were also maintained at 700 W, y150 V, 2 Pa, 25 8C, respectively. As the gas-mixing ratio increases to 20% Cl2, the etch rate of the PST film increases. However, when the gas-mixing ratio exceeds more than 20% Cl2, the etch rate of the PST film decreases. This result confirms that PST film was etched not only by ion bombardment but also by chemical reaction. PbCl4 (boiling point: 50 8C), SrCl2 (boiling point: 1250 8C) and TiCl4 (boiling point: 136 8C) are expected to form
Fig. 2. The etch rate of PST as well as the selectivity of PST to Pt as a function of RF power.
690
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
Fig. 3. The etch rate of PST as well as the selectivity of PST to Pt as a function of DC bias voltage.
the selectivity of PST to Pt seems to be constant, since the ion bombardment effect is similar to PST and Pt etching. To understand the roles of physical and chemical effects, the data about the influence of Cl2 addition on volume densities of active particles are needed. For these purposes we used OES analysis. For the control of active species volume densities behavior in Cl2 yAr plasmas, we selected such emission maximums as atomic lines of Cl (452.6 nm), Ar (750.4 nm) and molecular band of Cl2 (307.4 nm). These maximums are frequently used for analytical purposes including optical emission actinometry to determine absolute densities of active particles. Fig. 4 illustrates variations of emission intensities of selected maximums as a function of Cl2 concentration in gas mixtures. It is evidently clear that the increasing of Ar content in Cl2 yAr mixture leads to a direct proportional increase of emission intensity of Ar atoms. Note that Ar (750.4 nm) line is characterized by maximal excitation threshold energy among all selected maximums and thus its excitation constant should be maximal sensitive to electron energy distribution function (EEDF) variations, especially in the region of highenergy ‘tail’. Therefore, OES data confirm our conclusion that changing of Ar fraction in gas mixtures does not lead to sufficient changes of multiplication of excitation constant and electron density. To our mind it is possible to assume that the same behavior will be obtained for Arq ions volume density. The reason is that ionization rate is to follow the on the behavior of Ar ground state atoms while the rate of decay of Arq ions due to diffusion to reactor walls is expected to be
practically constant under the constant pressure conditions. As for Cl2 yAr plasma, the fact of primary interest is the constancy of chlorine atoms volume density up to 40% addition of Cl2. Effect of constancy of chlorine atoms volume density may be explained by increasing dissociation degree of Cl2 molecules during Ar addition. Analysis of data reported above allows making some conclusions concerning etching mechanism of PST thin films. It is evidently clear that more than twofold increasing of etching rate in gas mixtures up to 20% addition of Cl2 cannot be supported only by chemical mechanism, because volume densities of chlorine atoms have a tendency to decreasing. To our mind, the increasing of etching rate may be explained by two factors. First factor is connected with acceleration of physical sputtering of both main material such as PST film and surface layer of reaction products. The evident contribution of physical sputtering in etching process is confirmed by the fact that etching rate in pure Ar is sufficiently more than in Cl2 gases. Second factor represents the sequence of first one and connected with acceleration of chemical interaction through the increasing of reaction probability and destruction of metal– oxide bonds. In this case, extreme behavior of etching rate and decreasing of etching rate at Cl2 content less than 20% may be connected with decreasing and disappearing of chemical channel due to rapid decreasing of volume density of chemical active species. To investigate the influence of Cl2 yAr mixing ratio on mass content of plasma active species, we experimentally measured electron temperature and electron density as input parameters. Volume densities of charged particles in Cl2 yAr plasma were calculated during the
Fig. 4. Emission intensities of Cl (452.6 nm), Cl2 (307.4 nm) and Ar (750.4 nm) as a function of Cl2yAr mixing ratio.
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
691
Fig. 6. Volume densities of positive ions as a function of Cl2yAr mixing ratio. Fig. 5. Ionization rates of positive ions as a function of Cl2yAr mixing ratio.
simultaneous solution of balance quasi-stationary kinetic equations and quasi-neutrality equation. And, to determine volume densities of neutral particles, we used the combination of mass balance equation and the equation of chemical kinetics for chlorine atoms. Fig. 5 shows that the results of the rates of the processes supported formation of active particles in Cl2 y Ar plasma. Data show that addition of Cl2 leads to rapid increase of dissociation rate of Cl2 molecules. The reason is that Cl2 dissociation process is characterized by relatively low threshold energy (3.4 eV) and thus dissociation rate coefficient is not very sensitive to the changes of electron temperature. Therefore, increase of dissociation rate coefficient is overcompensated by the decreasing electron density and Cl2 dissociation rate increases monotonously. Note that the data of Fig. 5 reflect only the behavior of direct electron impact dissociation rate. Although dissociative ionization and dissociative attachment are considered as possible sources of chlorine atoms, the contribution of these processes is insignificant. As Cl2 content in Cl2 yAr plasma increases from 2 to 98%, the rate of dissociative attachment increases from 3.8=1016 to 4.4=1014 cmy3 sy1 while the rate of dissociative ionization changes in the scopes of 1.1=1016 –6.3=1014 cmy3 sy1. Therefore, our data show that electron impact dissociation of Cl2 molecules may be considered as the main source of chlorine atoms in plasma volume. This conclusion is in good agreement with the published result concerning both pure Cl2 plasma and Cl2 yAr mixture. The neutral particles kinetics indicate that the behavior of the volume density of Cl atoms follows the behavior of dissociation rate. Therefore, increasing of Ar content
in Cl2 yAr mixture leads to monotonous decreasing volume density of chlorine atoms. Corresponding data are represented in Fig. 6. As shown in Fig. 5, total ionization rate (as the sum of partial values) has a weak tendency to increasing in Ar-rich plasma because partial ionization rates of all components change monotonously. However, this effect is compensated by the increase of heterogeneous decay of positive ions through the increase of ion Bohm velocity and, as the result, total density of positive ions decreases in all the scopes of Ar content in Cl2 yAr mixture. Additionally, calculations of relative volume density of negative ions (bsnyne) give the values in the scopes of 0.1–0.01. At the same time, it was found that in high-pressure chlorine plasma (100 Pa), the value of relative density of negative ions exceeds 100. Therefore, our data confirms that at low pressures chlorine plasma transforms from high electronegative to moderately or low electronegative form. The data concerning the influence of Cl2 yAr mixing ratio on volume densities of positive ions are represented in Fig. 6. 4. Conclusion Etching characteristics of PST thin films were investigated using inductively coupled chlorine-based plasma system as functions of gas-mixing ratio, RF power and DC bias. It was found that increasing of Ar content in gas mixture lead to sufficient increasing of etching rate and selectivity of PST to Pt. The maximum etch rate of PST film is 56.2 nm miny1 and the selectivity of PST film to Pt is 0.8 at Cl2 y(Cl2qAr) of 20%. Due to the relatively low volatility of by-products formed during the etching by Cl2 yAr plasma, ion bombardment in addition to chemical reaction was required to obtain
692
G.-H. Kim et al. / Thin Solid Films 447 – 448 (2004) 688–692
high PST etch rates. It was proposed that sputter etching is the dominant etching mechanism while the contribution of chemical reaction is relatively low due to low volatility of etching products. References w1x R.B. Khnmankar, M.A. Kressley, M.R. Visokay, T. Moise, G. Xing, S. Nemoto, Y. Okuno, S.J. Fang, A.M. Wilson, J.F. Gaynor, T.Q. Hurd, D.L. Crenshaw, S. Summerfelt, L. Colombo, IEDM Dig. Technol. 4 (1997) 245.
w2x S. Yamamichi, P.Y. Lesaicherre, H. Yamaguchi, K. Takemura, S. Sone, H. Yabuta, K. Sato, T. Tamura, K. Nakajima, S. Ohnishi, K. Tokashiki, Y. Hayashi, Y. Miyasaka, M. Yoshida, H. Ono, IEEE Trans. Elect. Dev. 44 (1997) 1076. w3x M. Kiyotoshi, K. Eguchi, Electrochem. Soc. Proc. 144 (1997) 1063. w4x H.J. Chung, S.J. Chung, J.H. Kim, S.I. Woo, Thin Solid Films 394 (2001) 213. w5x D.H. Kang, J.H. Kim, J.H. Park, K.H. Yoon, Mat. Res. Bull. 36 (2001) 265. w6x H.J. Chung, S.I. Woo, J. Vac. Sci. Technol. B 19 (2001) 275. w7x David R. Lide (Ed.), CRC Handbook of Chemistry and Physics, CRC Press LLC, Washington, DC, 1998, pp. 4–66.