- Email: [email protected]
Etching characteristics of TaN thin film using an inductively coupled plasma
Surface & Coatings Technology 205 (2010) S333–S336
Contents lists available at ScienceDirect
Surface & Coatings Technology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s u r f c o a t
Etching characteristics of TaN thin film using an inductively coupled plasma Doo-Seung Um, Jong-Chang Woo, Chang-Il Kim ⁎ School of Electrical and Electrics Engineering, Chung-Ang University, Seoul 156-756, Republic of Korea
a r t i c l e
i n f o
Available online 11 August 2010 Keywords: TaN Etch Electrode ICP Diffusion barrier layer
a b s t r a c t In this study, TaN thin films used as a diffusion barrier layer and electrode material were etched with an inductively coupled plasma. The TaN thin film was deposited on a SiO2 layer by ALD. The dry etching mechanism of the TaN thin film was studied as a function of the BCl3/N2 gas mixing ratio, RF power, DC-bias voltage and process pressure. When the gas mixing ratio was BCl3 (14 sccm)/N2 (6 sccm), with the other conditions fixed, the highest etch rate was obtained. As the RF power and DC-bias voltage were increased, the etch rate of the TaN thin film was increased. By decreasing the process pressure, the etch rate of the TaN thin film was also increased. The chemical reaction on the surface of the etched TaN thin film was investigated by X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES). © 2010 Elsevier B.V. All rights reserved.
1. Introduction The core issue in current semiconductor technology, namely nanotechnology, has been intensively investigated. In particular, various attempts are being made to reduce the dimension of semiconductor devices using this technology. Research is underway to reduce the dimension of transistors and reduce the leakage current by replacing the SiO2 with high-k materials, and by using low dielectric constant insulator materials and low resistivity metal materials to reduce the RC delay between the metal layers [1–3]. One of the metals being studied as the nextgeneration material is copper, because its resistivity is 40% lower than that of aluminum and it has good attributes for electro-migration. However, copper has a fatal shortcoming in that it easily diffuses, so that a barrier layer to prevent its diffusion is essential. Metal nitrides, such as TaN, TiN and WN, have recently been studied as potential diffusion barrier layer materials and research into high-k and low-k materials is ongoing. TaN can be used as a good diffusion barrier layer for metal (Cu)/high-k insulator stacks and metal (Cu)/low-k insulator stacks, due to its favorable characteristics, such as its high melting point and hardness, mechanical stability and good conductivity [4–8]. Moreover, for gate stacks with an HfO2 insulator layer, the properties of the equivalent oxide thickness (EOT) can be improved by using a TaN/HfO2 stack to replace the poly-Si/SiO2 stack [1,9–13]. Besides, TaN materials have been used as the electrodes of MIM (metal–insulator– metal) capacitors [14,15]. To manufacture semiconductor devices, the electrode materials need to be easily etched in order to form (highly anisotropic and smooth sidewall) fine electrodes in MOSFETs (metal oxide semicon-
⁎ Corresponding author. E-mail address: [email protected] (C.-I. Kim). 0257-8972/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.surfcoat.2010.08.011
ductor field effect transistors). Therefore, we investigated the etching characteristics of TaN thin films and the surface reaction of TaN after etching. Four variables were studied in this experiment: the BCl3/N2 gas mixing ratio, the RF power (coil power), the DC-bias voltage governed by the electrode power and the process pressure (internal pressure). The chemical reactions on the surface of the TaN thin film were investigated using X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES) after the etching process.
2. Experimental details The dry etching of the TaN thin film was performed using an inductively coupled plasma etching system. Fig. 1 shows the inductively coupled plasma etching system and the plasma antenna in the system. The reactor consists of a cylindrical chamber with a diameter of 26 cm. The plasma antenna consisted of a 3.5-turn copper coil connected to a 13.56 MHz RF power supply located above a 24 mm-thick horizontal quartz window, which was separated from the inner space of the chamber. The bottom was connected to another 13.56 MHz asymmetric RF generator to control the bias voltage. The distance between the quartz window and substrate electrode was 9 cm. The substrate temperature was kept at 40° C by a chiller system. To keep the chamber free of impurities, it was evacuated to 10–6 Torr using a mechanical pump and a turbo molecular pump. The TaN thin film with a thickness of 100 nm was deposited on the SiO2 (10 nm)/Si-substrate by ALD (atomic layer deposition). The etching rate of the thin films was measured using a depth profiler (alpha-step 500, KLA Tencor) after the etching process. The etching rate is defined as the ratio of the total depth divided by the etching time (20 s). The chemical composition of the etched TaN thin film surface was evaluated using XPS (AXIS-HSi, Kratos) and AES (PHI 660, Perkin-Elmer).
S334
D.-S. Um et al. / Surface & Coatings Technology 205 (2010) S333–S336
Fig. 2. The etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the BCl3/N2 gas mixing ratio.
TaN thin film mainly depends on the chemical etching by chlorine radicals in the BCl3/N2 plasma. 3.2. The effect of the RF power Fig. 3 shows the effect of the RF power (coil power) on the etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR, while the other conditions were kept fixed (total gas flow rate: 20 sccm, BCl3:N2 = 3:7, DC-bias voltage: − 200 V, process pressure: 2 Pa, substrate temperature: 40° C). As shown in Fig. 3, as the RF power was increased from 400 W to 700 W, the etch rate of the TaN thin film increased to 154 nm/min, while the etch selectivity of the TaN thin film to SiO2 decreased from 5.5 to 3.7. The etch selectivity of the TaN thin film to PR remained in the range from 0.27 to 0.33 as the RF power was varied. Increasing the RF power increased the number of reactive radicals and ion current density in the BCl3/N2 plasma [19], which contributed to the increase in the etch rate of the TaN thin film. It is thought that the etch rate of the TaN thin film was affected by the RF power [20–22]. 3.3. The effect of the DC-bias voltage Fig. 1. Schematic diagram of ICP etcher apparatus.
3. Results and discussion 3.1. The effect of the gas mixing ratio The etching of the TaN thin film was studied as a function of the BCl3/N2 gas mixing ratio at an RF power of 500 W, a DC-bias voltage of −200 V, a process pressure of 2 Pa, and a substrate temperature of 40° C. The total gas flow rate was 20 sccm. Fig. 2 shows the etch rate of the TaN thin film and the selectivity of the TaN to SiO2 and PR as a function of the BCl3/N2 gas mixing ratio. The highest etch rate of the TaN thin film was 110 nm/min at BCl3 (30%)/N2 (70%). The etch selectivity of the TaN thin film to the SiO2 thin film showed a similar pattern to the etch rate of the TaN thin film. The etch selectivity of the TaN thin film to PR varied between 0.01 and 0.5. Tables 1 and 2 show the melting points, the boiling points and the Gibb's free energy potentials, respectively [13,16,17]. This data can be explained by the fact that the main by-product in the etching of the TaN thin film is TaCl5 [13,16,17]. Oh et al. showed that the addition of N2 to BCl3 increases the number of Cl radicals resulting from the recombination reduction between B and Cl [18]. It is thought that the etch rate of the
Fig. 4 shows the etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the DC-bias voltage by electrode power while the other conditions were fixed (total gas flow rate: 20 sccm, BCl3:N2 = 3:7, RF power: 500 W, process pressure: 2 Pa, substrate temperature: 40° C). As shown in the figure, the etch rate of the TaN thin film increased dramatically to 169 nm/min as the DC-bias Table 1 Melting points and boiling points of the etch products at atmospheric pressure. Materials
Melting point (°C)
Boiling point (°C)
Ta TaN TaCl5 [g]
3007 3090 216
5458 – 239
Table 2 Gibb's free energy of the volatile etch by-products in the Ta-based material.
TaN
Reaction (g)
ΔG°f (kJ/mol)
TaCl3 TaCl4 TaCl5
− 313.17 − 540.657 − 709.287
D.-S. Um et al. / Surface & Coatings Technology 205 (2010) S333–S336
S335
Fig. 3. The etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the RF power.
Fig. 5. The etch rate of the TaN thin film and the selectivity of the TaN3 thin film to SiO2 and PR as a function of the process pressure.
voltage was increased from −100 V to − 250 V. The etch selectivity of the TaN thin film to SiO2 increased from 3.4 to 4.4 and the etch selectivity of the TaN thin film to PR increased from 0.31 to 0.41 as the DC-bias voltage increased up to −250 V. This is explained by the fact that etching of the TaN thin film was affected by the physical action induced by the electric field. And this physical action, ion bombardment by the electric field, is absolutely needed for the etching process of the TaN thin film in the BCl3/N2 plasma [20–24].
was increased from 1 Pa to 3 Pa. As the pressure decreases, the meanfree-path becomes longer, due to the falling gas flow rate in the chamber. As the mean-free-path increases, the probability of collisions and charge transfer decreases in the substrate sheath region [22]. We considered that the increase in the etch rate with decreasing
3.4. The effect of the process pressure Fig. 5 shows the effect of the process pressure on the etch rate of the TaN thin film and the etch selectivity of the TaN thin film to SiO2 and PR at the other conditions were fixed (total gas flow rate: 20 sccm, BCl3:N2 = 3:7, RF power: 500 W, DC-bias voltage: −200 V, substrate temperature: 40° C). As shown in Fig. 5, the etch rate of the TaN thin film increased with decreasing process pressure. The etch selectivity of the TaN thin film to the SiO2 thin film showed a similar pattern to the etch rate of the TaN thin film. The etch selectivity of the TaN thin film to PR increased from 0.33 to 0.52 as the process pressure
Fig. 4. The etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the DC-bias voltage.
Fig. 6. The XPS spectra of the Ta and N peaks obtained from the surface of the TaN thin films after etching in the BCl3-based plasmas. (a) Ta 4f peak, (b) N 1s and Ta 4p peaks.
S336
D.-S. Um et al. / Surface & Coatings Technology 205 (2010) S333–S336
various DC-bias voltages. The Cl peak is observed in Fig. 7 (a), but not in Fig. 7 (b). This result means that the DC-bias voltage influences the volatility of the TaClx compounds. We also observed a difference in the intensity of the oxygen peaks between Fig. 7 (a) and (b). The oxygen peak in Fig. 7 (b) is larger than that in Fig. 7 (a). This means that the TaClx compounds were removed during the etching process at a high DC-bias voltage. Therefore, many Ta–O bonds were formed in the atmosphere. The XPS and AES results demonstrate that a TaN thin film with a clean surface can be obtained by using a high DC-bias voltage. It can be concluded that the etching mechanism of the TaN thin film is highly dependent on the DC-bias voltage.
4. Conclusion In this study, the etching characteristics of TaN thin films were investigated as functions of the BCl3/N2 gas mixing ratio, RF power, DC-bias voltage and process pressure in an inductively coupled plasma system. The chemical reaction on the surface of the TaN thin film was investigated using XPS and AES analysis. The best etch rate was obtained with the BCl3 (6 sccm)/N2 (14 sccm) gas plasma. The etch rate was increased at an RF power of ~700 W, increased at a DCbias voltage of ~−250 V and decreased at a process pressure of ~1 Pa. The experiment performed by varying the DC-bias voltage showed that the ion bombardment is an important parameter in the etching process. In our experiment, the highest etch rate of the TaN thin film of 169 nm/min was obtained at a gas mixing ratio of BCl3 (6 sccm)/N2 (14 sccm), an RF power of 500 W, a DC-bias voltage of − 250 V, a pressure at 2 Pa, and a substrate temperature of 40° C.
References
Fig. 7. The AES spectra and surface composition obtained from the surface of the TaN thin films after etching in the BCl3-based plasmas. (a) DC-bias voltage of − 50 V, (b) DC-bias voltage of –200 V.
pressure was due to the increase in the energy of bombardment caused by the increase in the mean-free-path. 3.5. Analysis by XPS and AES In order to analyze the etching mechanism of the TaN thin films in detail, the etched surfaces were examined by XPS and AES. Fig. 6 shows the XPS narrow scan spectra of the Ta 4f, Ta 4p, and N 1s peaks in the surface compositions for the as-deposited films and TaN thin films etched using the BCl3/N2 plasma. Fig. 6 (a) shows the variation of the Ta 4f peaks and Fig. 6 (b) shows the variation of the N 1s peaks and the decrease in the intensity of the Ta 4p peak after the etching process. The variation of the Ta 4f and N 1s peaks can be explained as follows: the Ta–N bonds were broken by the chemical reaction of Ta with the Cl radicals on the surface of the TaN thin film. The etching of the TaN thin film in the BCl3 plasma results in the formation of TaClx radicals on its surface. As the DC-bias voltage increases, the number of Ta–N bonds on the TaN surface is reduced and the number of Ta–O bonds is increased due to sputtering. Fig. 7 shows the AES spectra of the surface of the TaN thin film after etching. Fig. 7 (a) and (b) shows the spectra of the film etched at
[1] M.H. Shin, M.S. Park, N.E. Lee, J.Y. Kim, C.Y. Kim, J.H. Ahn, J. Vac. Sci. Technol. A 24 (2006) 1373. [2] B. Cheng, M. Cao, P.V. Vande, W. Greene, H. Stork, Z. Yu, J.C.S. Woo, IEEE Trans. Electron Devices 46 (1) (1999) 261. [3] J.R. Shi, S.P. Lau, Z. Sun, X. Shi, B.K. Tay, H.S. Tan, Surf. Coat. Technol. 138 (2001) 250. [4] J.C. Lin, C. Lee, J. Electrochem. Soc. 146 (9) (1999) 3466. [5] Q. Xie, X.P. Qu, J.J. Tan, Y.L. Jiang, M. Zhou, T. Chen, G.P. Ru, Appl. Surf. Sci. 253 (2006) 1666. [6] J. Liu, J. Bao, M. Schambeg, W.C. Kim, P.S. Ho, R. Laxman, J. Vac. Sci. Technol. A 23 (4) (2005) 1107. [7] A. Arranz, C. Palacio, Surf. Interface Anal. 29 (2000) 653. [8] A. Furuya, E. Soda, M. Shimada, S. Ogawa, Jpn. J. Appl. Phys. 44 (10) (2005) 7430. [9] B.H. Lee, R. Choi, L. Kang, S. Gopalan, R. Nieh, K. Onishi, Y. Jeon, W.J. Qi, C. Kang, J.C. Lee, IEDM 39 (2000) 2.6.1. [10] V.N. Bliznetosov, L.K. Bera, H.Y. Soo, N. Balasubramanian, R. Kumar, G.Q. Lo, W.J. Yoo, C.H. Tung, L. Linn, IEEE Trans. Semicond. Manuf. 20 (2) (2007) 143. [11] V. Bliznetsov, R. Kumar, L.K. Bera, L.W. Yip, A. Du, T.E. Hui, Thin Solid Films 504 (2006) 140. [12] S.K. Yang, H.H. Kim, B.H. O, S.G. Lee, E.H. Lee, S.G. Park, S.P. Chang, J.G. Lee, H.Y. Song, J. Kor. Phys. Soc. 51 (2007) S198. [13] W.S. Hwang, J. Chen, W.J. Yoo, V. Bliznetsov, J. Vac. Sci. Technol. A 23 (4) (2005) 964. [14] T.K. Kang, C.T. Chang, C.L. Lin, W.F. Wu, Microelectron. Eng. 86 (2009) 1994. [15] T.K. Kang, C.W. Chen, C.L. Lin, W.F. Wu, Jpn. J. Appl. Phys. 47 (2008) 5374. [16] J. Chen, W.J. Yoo, Z.Y. Tan, Y. Wang, D.S.H. Chan, J. Vac. Sci. Technol. A 22 (2004) 1552. [17] J. Tonotani, T. Iwamoto, F. Sato, K. Hattori, S. Ohmi, H. Iwai, J. Vac. Sci. Technol. B 21 (2003) 2163. [18] C.S. Oh, T.H. Kim, K.Y. Lim, J.W. Yang, Semicond. Sci. Technol. 19 (2004) 172. [19] K.J. Park, K.H. Kim, W.M. Lee, H. Chae, I.S. Han, H.D. Lee, Trans. Electr. Electron. Mater. 10 (2) (2009) 35. [20] M.S. Shin, S.W. Na, N.E. Lee, J.H. Ahn, Thin Solid Films 506–507 (2006) 230. [21] H.J. Lee, B.S. Kwon, H.W. Kim, S.I. Kim, D.G. Yoo, J.H. Boo, N.E. Lee, Jpn. J. Appl. Phys. 47 (8) (2008) 6960. [22] J.K. Jung, W.J. Lee, Jpn. J. Appl. Phys. 40 (3A) (2001) 1408. [23] M.H. Shin, S.W. Na, N.E. Lee, T.K. Oh, J. Kim, T. Lee, J. Ahn, Jpn. J. Appl. Phys. 44 (7B) (2005) 5811. [24] A. Matsutani, H. Ohtsuki, F. Koyama, Jpn. J. Appl. Phys. 42 (12A) (2003) L1414.
Contents lists available at ScienceDirect
Surface & Coatings Technology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s u r f c o a t
Etching characteristics of TaN thin film using an inductively coupled plasma Doo-Seung Um, Jong-Chang Woo, Chang-Il Kim ⁎ School of Electrical and Electrics Engineering, Chung-Ang University, Seoul 156-756, Republic of Korea
a r t i c l e
i n f o
Available online 11 August 2010 Keywords: TaN Etch Electrode ICP Diffusion barrier layer
a b s t r a c t In this study, TaN thin films used as a diffusion barrier layer and electrode material were etched with an inductively coupled plasma. The TaN thin film was deposited on a SiO2 layer by ALD. The dry etching mechanism of the TaN thin film was studied as a function of the BCl3/N2 gas mixing ratio, RF power, DC-bias voltage and process pressure. When the gas mixing ratio was BCl3 (14 sccm)/N2 (6 sccm), with the other conditions fixed, the highest etch rate was obtained. As the RF power and DC-bias voltage were increased, the etch rate of the TaN thin film was increased. By decreasing the process pressure, the etch rate of the TaN thin film was also increased. The chemical reaction on the surface of the etched TaN thin film was investigated by X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES). © 2010 Elsevier B.V. All rights reserved.
1. Introduction The core issue in current semiconductor technology, namely nanotechnology, has been intensively investigated. In particular, various attempts are being made to reduce the dimension of semiconductor devices using this technology. Research is underway to reduce the dimension of transistors and reduce the leakage current by replacing the SiO2 with high-k materials, and by using low dielectric constant insulator materials and low resistivity metal materials to reduce the RC delay between the metal layers [1–3]. One of the metals being studied as the nextgeneration material is copper, because its resistivity is 40% lower than that of aluminum and it has good attributes for electro-migration. However, copper has a fatal shortcoming in that it easily diffuses, so that a barrier layer to prevent its diffusion is essential. Metal nitrides, such as TaN, TiN and WN, have recently been studied as potential diffusion barrier layer materials and research into high-k and low-k materials is ongoing. TaN can be used as a good diffusion barrier layer for metal (Cu)/high-k insulator stacks and metal (Cu)/low-k insulator stacks, due to its favorable characteristics, such as its high melting point and hardness, mechanical stability and good conductivity [4–8]. Moreover, for gate stacks with an HfO2 insulator layer, the properties of the equivalent oxide thickness (EOT) can be improved by using a TaN/HfO2 stack to replace the poly-Si/SiO2 stack [1,9–13]. Besides, TaN materials have been used as the electrodes of MIM (metal–insulator– metal) capacitors [14,15]. To manufacture semiconductor devices, the electrode materials need to be easily etched in order to form (highly anisotropic and smooth sidewall) fine electrodes in MOSFETs (metal oxide semicon-
⁎ Corresponding author. E-mail address: [email protected] (C.-I. Kim). 0257-8972/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.surfcoat.2010.08.011
ductor field effect transistors). Therefore, we investigated the etching characteristics of TaN thin films and the surface reaction of TaN after etching. Four variables were studied in this experiment: the BCl3/N2 gas mixing ratio, the RF power (coil power), the DC-bias voltage governed by the electrode power and the process pressure (internal pressure). The chemical reactions on the surface of the TaN thin film were investigated using X-ray photoelectron spectroscopy (XPS) and Auger electron spectroscopy (AES) after the etching process.
2. Experimental details The dry etching of the TaN thin film was performed using an inductively coupled plasma etching system. Fig. 1 shows the inductively coupled plasma etching system and the plasma antenna in the system. The reactor consists of a cylindrical chamber with a diameter of 26 cm. The plasma antenna consisted of a 3.5-turn copper coil connected to a 13.56 MHz RF power supply located above a 24 mm-thick horizontal quartz window, which was separated from the inner space of the chamber. The bottom was connected to another 13.56 MHz asymmetric RF generator to control the bias voltage. The distance between the quartz window and substrate electrode was 9 cm. The substrate temperature was kept at 40° C by a chiller system. To keep the chamber free of impurities, it was evacuated to 10–6 Torr using a mechanical pump and a turbo molecular pump. The TaN thin film with a thickness of 100 nm was deposited on the SiO2 (10 nm)/Si-substrate by ALD (atomic layer deposition). The etching rate of the thin films was measured using a depth profiler (alpha-step 500, KLA Tencor) after the etching process. The etching rate is defined as the ratio of the total depth divided by the etching time (20 s). The chemical composition of the etched TaN thin film surface was evaluated using XPS (AXIS-HSi, Kratos) and AES (PHI 660, Perkin-Elmer).
S334
D.-S. Um et al. / Surface & Coatings Technology 205 (2010) S333–S336
Fig. 2. The etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the BCl3/N2 gas mixing ratio.
TaN thin film mainly depends on the chemical etching by chlorine radicals in the BCl3/N2 plasma. 3.2. The effect of the RF power Fig. 3 shows the effect of the RF power (coil power) on the etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR, while the other conditions were kept fixed (total gas flow rate: 20 sccm, BCl3:N2 = 3:7, DC-bias voltage: − 200 V, process pressure: 2 Pa, substrate temperature: 40° C). As shown in Fig. 3, as the RF power was increased from 400 W to 700 W, the etch rate of the TaN thin film increased to 154 nm/min, while the etch selectivity of the TaN thin film to SiO2 decreased from 5.5 to 3.7. The etch selectivity of the TaN thin film to PR remained in the range from 0.27 to 0.33 as the RF power was varied. Increasing the RF power increased the number of reactive radicals and ion current density in the BCl3/N2 plasma [19], which contributed to the increase in the etch rate of the TaN thin film. It is thought that the etch rate of the TaN thin film was affected by the RF power [20–22]. 3.3. The effect of the DC-bias voltage Fig. 1. Schematic diagram of ICP etcher apparatus.
3. Results and discussion 3.1. The effect of the gas mixing ratio The etching of the TaN thin film was studied as a function of the BCl3/N2 gas mixing ratio at an RF power of 500 W, a DC-bias voltage of −200 V, a process pressure of 2 Pa, and a substrate temperature of 40° C. The total gas flow rate was 20 sccm. Fig. 2 shows the etch rate of the TaN thin film and the selectivity of the TaN to SiO2 and PR as a function of the BCl3/N2 gas mixing ratio. The highest etch rate of the TaN thin film was 110 nm/min at BCl3 (30%)/N2 (70%). The etch selectivity of the TaN thin film to the SiO2 thin film showed a similar pattern to the etch rate of the TaN thin film. The etch selectivity of the TaN thin film to PR varied between 0.01 and 0.5. Tables 1 and 2 show the melting points, the boiling points and the Gibb's free energy potentials, respectively [13,16,17]. This data can be explained by the fact that the main by-product in the etching of the TaN thin film is TaCl5 [13,16,17]. Oh et al. showed that the addition of N2 to BCl3 increases the number of Cl radicals resulting from the recombination reduction between B and Cl [18]. It is thought that the etch rate of the
Fig. 4 shows the etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the DC-bias voltage by electrode power while the other conditions were fixed (total gas flow rate: 20 sccm, BCl3:N2 = 3:7, RF power: 500 W, process pressure: 2 Pa, substrate temperature: 40° C). As shown in the figure, the etch rate of the TaN thin film increased dramatically to 169 nm/min as the DC-bias Table 1 Melting points and boiling points of the etch products at atmospheric pressure. Materials
Melting point (°C)
Boiling point (°C)
Ta TaN TaCl5 [g]
3007 3090 216
5458 – 239
Table 2 Gibb's free energy of the volatile etch by-products in the Ta-based material.
TaN
Reaction (g)
ΔG°f (kJ/mol)
TaCl3 TaCl4 TaCl5
− 313.17 − 540.657 − 709.287
D.-S. Um et al. / Surface & Coatings Technology 205 (2010) S333–S336
S335
Fig. 3. The etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the RF power.
Fig. 5. The etch rate of the TaN thin film and the selectivity of the TaN3 thin film to SiO2 and PR as a function of the process pressure.
voltage was increased from −100 V to − 250 V. The etch selectivity of the TaN thin film to SiO2 increased from 3.4 to 4.4 and the etch selectivity of the TaN thin film to PR increased from 0.31 to 0.41 as the DC-bias voltage increased up to −250 V. This is explained by the fact that etching of the TaN thin film was affected by the physical action induced by the electric field. And this physical action, ion bombardment by the electric field, is absolutely needed for the etching process of the TaN thin film in the BCl3/N2 plasma [20–24].
was increased from 1 Pa to 3 Pa. As the pressure decreases, the meanfree-path becomes longer, due to the falling gas flow rate in the chamber. As the mean-free-path increases, the probability of collisions and charge transfer decreases in the substrate sheath region [22]. We considered that the increase in the etch rate with decreasing
3.4. The effect of the process pressure Fig. 5 shows the effect of the process pressure on the etch rate of the TaN thin film and the etch selectivity of the TaN thin film to SiO2 and PR at the other conditions were fixed (total gas flow rate: 20 sccm, BCl3:N2 = 3:7, RF power: 500 W, DC-bias voltage: −200 V, substrate temperature: 40° C). As shown in Fig. 5, the etch rate of the TaN thin film increased with decreasing process pressure. The etch selectivity of the TaN thin film to the SiO2 thin film showed a similar pattern to the etch rate of the TaN thin film. The etch selectivity of the TaN thin film to PR increased from 0.33 to 0.52 as the process pressure
Fig. 4. The etch rate of the TaN thin film and the selectivity of the TaN thin film to SiO2 and PR as a function of the DC-bias voltage.
Fig. 6. The XPS spectra of the Ta and N peaks obtained from the surface of the TaN thin films after etching in the BCl3-based plasmas. (a) Ta 4f peak, (b) N 1s and Ta 4p peaks.
S336
D.-S. Um et al. / Surface & Coatings Technology 205 (2010) S333–S336
various DC-bias voltages. The Cl peak is observed in Fig. 7 (a), but not in Fig. 7 (b). This result means that the DC-bias voltage influences the volatility of the TaClx compounds. We also observed a difference in the intensity of the oxygen peaks between Fig. 7 (a) and (b). The oxygen peak in Fig. 7 (b) is larger than that in Fig. 7 (a). This means that the TaClx compounds were removed during the etching process at a high DC-bias voltage. Therefore, many Ta–O bonds were formed in the atmosphere. The XPS and AES results demonstrate that a TaN thin film with a clean surface can be obtained by using a high DC-bias voltage. It can be concluded that the etching mechanism of the TaN thin film is highly dependent on the DC-bias voltage.
4. Conclusion In this study, the etching characteristics of TaN thin films were investigated as functions of the BCl3/N2 gas mixing ratio, RF power, DC-bias voltage and process pressure in an inductively coupled plasma system. The chemical reaction on the surface of the TaN thin film was investigated using XPS and AES analysis. The best etch rate was obtained with the BCl3 (6 sccm)/N2 (14 sccm) gas plasma. The etch rate was increased at an RF power of ~700 W, increased at a DCbias voltage of ~−250 V and decreased at a process pressure of ~1 Pa. The experiment performed by varying the DC-bias voltage showed that the ion bombardment is an important parameter in the etching process. In our experiment, the highest etch rate of the TaN thin film of 169 nm/min was obtained at a gas mixing ratio of BCl3 (6 sccm)/N2 (14 sccm), an RF power of 500 W, a DC-bias voltage of − 250 V, a pressure at 2 Pa, and a substrate temperature of 40° C.
References
Fig. 7. The AES spectra and surface composition obtained from the surface of the TaN thin films after etching in the BCl3-based plasmas. (a) DC-bias voltage of − 50 V, (b) DC-bias voltage of –200 V.
pressure was due to the increase in the energy of bombardment caused by the increase in the mean-free-path. 3.5. Analysis by XPS and AES In order to analyze the etching mechanism of the TaN thin films in detail, the etched surfaces were examined by XPS and AES. Fig. 6 shows the XPS narrow scan spectra of the Ta 4f, Ta 4p, and N 1s peaks in the surface compositions for the as-deposited films and TaN thin films etched using the BCl3/N2 plasma. Fig. 6 (a) shows the variation of the Ta 4f peaks and Fig. 6 (b) shows the variation of the N 1s peaks and the decrease in the intensity of the Ta 4p peak after the etching process. The variation of the Ta 4f and N 1s peaks can be explained as follows: the Ta–N bonds were broken by the chemical reaction of Ta with the Cl radicals on the surface of the TaN thin film. The etching of the TaN thin film in the BCl3 plasma results in the formation of TaClx radicals on its surface. As the DC-bias voltage increases, the number of Ta–N bonds on the TaN surface is reduced and the number of Ta–O bonds is increased due to sputtering. Fig. 7 shows the AES spectra of the surface of the TaN thin film after etching. Fig. 7 (a) and (b) shows the spectra of the film etched at
[1] M.H. Shin, M.S. Park, N.E. Lee, J.Y. Kim, C.Y. Kim, J.H. Ahn, J. Vac. Sci. Technol. A 24 (2006) 1373. [2] B. Cheng, M. Cao, P.V. Vande, W. Greene, H. Stork, Z. Yu, J.C.S. Woo, IEEE Trans. Electron Devices 46 (1) (1999) 261. [3] J.R. Shi, S.P. Lau, Z. Sun, X. Shi, B.K. Tay, H.S. Tan, Surf. Coat. Technol. 138 (2001) 250. [4] J.C. Lin, C. Lee, J. Electrochem. Soc. 146 (9) (1999) 3466. [5] Q. Xie, X.P. Qu, J.J. Tan, Y.L. Jiang, M. Zhou, T. Chen, G.P. Ru, Appl. Surf. Sci. 253 (2006) 1666. [6] J. Liu, J. Bao, M. Schambeg, W.C. Kim, P.S. Ho, R. Laxman, J. Vac. Sci. Technol. A 23 (4) (2005) 1107. [7] A. Arranz, C. Palacio, Surf. Interface Anal. 29 (2000) 653. [8] A. Furuya, E. Soda, M. Shimada, S. Ogawa, Jpn. J. Appl. Phys. 44 (10) (2005) 7430. [9] B.H. Lee, R. Choi, L. Kang, S. Gopalan, R. Nieh, K. Onishi, Y. Jeon, W.J. Qi, C. Kang, J.C. Lee, IEDM 39 (2000) 2.6.1. [10] V.N. Bliznetosov, L.K. Bera, H.Y. Soo, N. Balasubramanian, R. Kumar, G.Q. Lo, W.J. Yoo, C.H. Tung, L. Linn, IEEE Trans. Semicond. Manuf. 20 (2) (2007) 143. [11] V. Bliznetsov, R. Kumar, L.K. Bera, L.W. Yip, A. Du, T.E. Hui, Thin Solid Films 504 (2006) 140. [12] S.K. Yang, H.H. Kim, B.H. O, S.G. Lee, E.H. Lee, S.G. Park, S.P. Chang, J.G. Lee, H.Y. Song, J. Kor. Phys. Soc. 51 (2007) S198. [13] W.S. Hwang, J. Chen, W.J. Yoo, V. Bliznetsov, J. Vac. Sci. Technol. A 23 (4) (2005) 964. [14] T.K. Kang, C.T. Chang, C.L. Lin, W.F. Wu, Microelectron. Eng. 86 (2009) 1994. [15] T.K. Kang, C.W. Chen, C.L. Lin, W.F. Wu, Jpn. J. Appl. Phys. 47 (2008) 5374. [16] J. Chen, W.J. Yoo, Z.Y. Tan, Y. Wang, D.S.H. Chan, J. Vac. Sci. Technol. A 22 (2004) 1552. [17] J. Tonotani, T. Iwamoto, F. Sato, K. Hattori, S. Ohmi, H. Iwai, J. Vac. Sci. Technol. B 21 (2003) 2163. [18] C.S. Oh, T.H. Kim, K.Y. Lim, J.W. Yang, Semicond. Sci. Technol. 19 (2004) 172. [19] K.J. Park, K.H. Kim, W.M. Lee, H. Chae, I.S. Han, H.D. Lee, Trans. Electr. Electron. Mater. 10 (2) (2009) 35. [20] M.S. Shin, S.W. Na, N.E. Lee, J.H. Ahn, Thin Solid Films 506–507 (2006) 230. [21] H.J. Lee, B.S. Kwon, H.W. Kim, S.I. Kim, D.G. Yoo, J.H. Boo, N.E. Lee, Jpn. J. Appl. Phys. 47 (8) (2008) 6960. [22] J.K. Jung, W.J. Lee, Jpn. J. Appl. Phys. 40 (3A) (2001) 1408. [23] M.H. Shin, S.W. Na, N.E. Lee, T.K. Oh, J. Kim, T. Lee, J. Ahn, Jpn. J. Appl. Phys. 44 (7B) (2005) 5811. [24] A. Matsutani, H. Ohtsuki, F. Koyama, Jpn. J. Appl. Phys. 42 (12A) (2003) L1414.