- Email: [email protected]
Nano-porous TiN thin films deposited by reactive sputtering method
International Journal of Inorganic Materials 3 (2001) 1193–1196
Nano-porous TiN thin films deposited by reactive sputtering method Q. Fang a , *, J.-Y. Zhang b b
a Institute of Solid State Physics, Chinese Academy of Sciences, Hefei 230031, PR China Department of Electrical and Electronic Engineering, University College London, London WC1 E 7 JE, UK
Abstract Nano-porous TiN thin films deposited by a reactive sputtering process are reported. The effect of deposition parameters including sputtering power density, nitrogen partial pressure and deposition time on the thin film growth has been investigated. Crystallisation behaviour, chemical composition and microstructure of the deposited films were also investigated using X-ray diffraction, EDX and scanning electron microscopy. A test cell of Na(l) / beta’’ alumina / TiN was set up and tested at the temperature range of 600–8008C, in order to investigate the cell power density and the interfacial electrical property between the electrolyte and TiN electrode. The maximum power density of 0.20 W cm 22 could be achieved with a large electrode area of 30 cm 2 at 8008C. The effect of microstructure of the nanometer size thin film electrodes on the conductive property has been studied and discussed. 2001 Elsevier Science Ltd. All rights reserved. Keywords: TiN thin films; Reactive sputtering; Nano-porous; Columnar structure
1. Introduction Titanium nitride (TiN) is a versatile material using for surface hardening of cutting tools [1,2], high temperature electrodes [3–5] and diffusion barrier layers in semiconductor devices and ferroelectric memory capacitors [6,7], because of its outstanding properties such as high surface hardness, good corrosion resistance at room temperature and elevated temperatures, low friction coefficient and high electrical conductivity [8]. The alkali metal thermoelectric converter (AMTEC) is a system in which beta’’-alumina acts as a solid sodium ion conductive electrolyte. Sodium is used as the working medium, for direct conversion of heat into electricity [3–5,9]. The anode is normally liquid sodium itself, and the cathode usually is a porous refractory metal layer such as Mo or W. However, refractory metals show high initial power densities but degrade with strong reduction in power and efficiency after 10 to 1000 h, due to electrochemical reactions [8,10]. On the other hand, being a three-phase boundary, consisting of beta-alumina, sodium gas and electron conductor, the cathode has to supply the electrons to reduce the sodium ions, which goes through beta*Corresponding author. Present address: The Centre for Nanotechnology & Microengineering, University of Warwick, Coventry CV4 7AL, UK. Tel.: 144-2476-524-216; fax: 144-2476-418-922. E-mail address: [email protected] (Q. Fang).
alumina, to sodium gas. There are two contradictive requirements for the cathode materials — good conductivity and proper porosity. Therefore, the material and structure of the thin cathode layer play extremely important role in AMTEC-operation. Recently, the refractory compounds such as TiN, TiB 2 have been investigated as electrode layer for AMTEC, in order to improve the performance of AMTEC [3,4,8,9]. TiN films are usually manufactured by deposition with thermal chemical vapour deposition (CVD), plasma-enhanced CVD (PECVD) and physical vapour deposition. However, most of CVD methods require higher temperature to achieve the deposition, which sometimes causes thermal damage to the deposited films. Therefore, low substrate-temperature deposition methods such as sputtering have been used widely. However, the sputter deposition of alloys and compounds would be difficult, since the yields of various materials differ in a sputter process. Oxides, nitrides and other materials with highly volatile constituents will not sputter in an inert gas environment to produce stochiometric films [11]. In order to produce films incorporated with the required quantities of gaseous species, reactive sputtering is used in which the inert feed gas is mixed with a reactive gas. In general, reactive sputtering is characterised by two mechanisms in its operation. (a) The surface of a target (usually is pure metal) is cleared of the nitride, oxide or cabonide compound by a high ion bombardment flux at low partial
1466-6049 / 01 / $ – see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S1466-6049( 01 )00124-6
1194
Q. Fang, J.-Y. Zhang / International Journal of Inorganic Materials 3 (2001) 1193 – 1196
Table 1 The experimental conditions of TiN x reactive sputtering deposition Sputtering power (W) Distance between target and substrate (mm) System vacuum (mbar) N 2 /(Ar1N 2 ) (%) Sputtering time (min) Sputtering temperature (8C)
200–600 50 6310 23 0–100 30–300 25
pressures of reactive gas (N 2 , O 2 , etc.). (b) Increasing the partial pressure of reactive gas results in a target surface that is converted to the compound material. Optimising the partial pressures of reactive gas, sputtering power and substrate temperature to obtain the desired film properties can control the reactive deposition process. In this paper, the chemical composition, crystal structures of sputterdeposited TiN were studied. Its electrical conductivity and dependence on deposition condition also were investigated. In addition, the power densities of TiN sputter-deposited thin films as electrode for an AMTEC test cell and dependence on film structures will be discussed.
2. Experimental A pure Ti (99.99) target (Leyboldt–Heraeus) was used for deposition of TiN thin films on flat glass and b‘‘-Al 2 O 3
tube of 25 mm diameter and 200 mm length by a DC magnetron reactive sputtering. The sputtering power was from 300 to 500 W in an Ar (99.99) / N 2 (99.99) environment. At first, the target surface and the reactor chamber were cleared of nitride or oxide and residual oxygen at a pure Ar feed gas, then increasing partial pressure of the reactive gas N 2 , a stochiometric TiN film can be produced at a proper nitrogen partial pressure. Crystallisation behaviour, chemical composition and microstructure of the deposited films were investigated using X-ray diffraction, EDX and scanning electron microscopy. A test cell of Na (l) / beta’’ alumina / TiN was set up and test at the temperature range of 400–8008C, in order to investigate the cell power density and the interfacial electrical property between the electrolyte and TiN electrode. The details of the test rig and processing can be found elsewhere [3,4,9]. The experimental conditions under which the sputtering deposition was carried out are listed in Table 1.
3. Results and discussion Fig. 1 shows SEM of TiN thin film sputtered on a glass (a) and on a beta’’-alumina tube (b), respectively. The film on the glass (a) shows a columnar structure, which are also found in Mo and TiB 2 sputter layers. The column size in the TiN film is about 50–80 nm. While the film on the
Fig. 1. SEM of TiN thin film sputtered on a glass (a) and on a beta’’-alumina tube (b).
Q. Fang, J.-Y. Zhang / International Journal of Inorganic Materials 3 (2001) 1193 – 1196
1195
Fig. 2. Current–voltage relationship in an AMTEC cell using TiN as electrode at different temperatures (a) and the corresponding power density–current curves (b).
Table 3 TiN x thin films deposited using various sputtering powers ( pN2 40% and 60 min) Sputtering power (W) Thickness (mm) Surface resistance (V / cm)
200 0.2 100
300 0.3 80
400 0.8 60
500 1.2 50
titanium content in the films is reached (N 2 /(Ar1N 2 ) (%).40), the absorbed spectra move towards slightly more reddish energies due to extra nitrogen in the films. The current–voltage relationship in an AMTEC cell using TiN thin film as electrode at different temperatures (a) and the corresponding power density–current curves (b) are shown in Fig. 2. It can be seen that a maximum current density of about 1.0 A cm 22 and a maximum power density of 0.2 W cm 22 could be achieved with 30 cm 2 electrode. It is noted that the good performance was achieved with thickness of the TiN film of 2–3 mm only. From our previous work [4], the optimal thickness of the TiN film, which is sputtered directly from a TiN target, is between 5 and 6 mm. The decrease of performance for a thinner electrode is due to higher sheet resistance. It is believed that the improvement of the AMTEC performance by the reactive sputtering TiN films rises from its lower porosity, lower sheet resistance and uniform columnar structure, which might benefit to electron transfer and letting a higher sodium vapour pressure looses in an electrochemical cell.
600 1.25 45
beta’’-alumina tube (b) shows a cauliflower structure, which consist of a number of columns. This columnar structure of TiN films is important for letting the sodium gas pass through the cathode layers itself during an electrochemical processing of AMTEC. Table 2 shows a relationship of N 2 partial pressures and deposition results from the reactive sputtering of TiN x films. Table 3 shows a dependence of sputtering power on TiN film thickness and surface conductivity, which indicated that the thickness and surface conductivity of TiN film increase with a rising sputtering power. Table 2 shows that stochiometric TiN films with lower sheet resistance and higher film deposition rate were prepared by reactive sputtering (500 W and 60 min) when N 2 partial pressures is about 40%. It was also found that colours of the deposited TiN x films changed from metallic grey to gold and finally brownish red as the nitrogen partial pressure increases. According to the Drude model [12], these TiN x thin films colour variations can be explained by titanium metallic bonds in the film. As the metallic bonds in the film becomes weaker, the absorbed spectra move towards lower energies; when the minimum
4. Conclusions Nano-porous TiN thin films deposited by a reactive sputtering process are reported. The effect of deposition parameters including sputtering power density, nitrogen partial pressure and deposition time on the thin film growth has been investigated. The stochiometric TiN films with
Table 2 TiN x thin films deposited under various N 2 partial pressures at 500 W, 60 min N 2 /(Ar1N 2 ) (%) Film colour
0 grey
10 dark grey
20 yellow
40 gold
60 light-brown
Thickness (mm) Sheet resistance (V / cm) N / Ti (%) Crystal phases
0.9 50 0 a-Ti
1.0 80 30.561 a-Ti, d-TiN
1.15 60 85.562 d-TiN, TiN
1.3 45 9962 TiN
1.2 50 10362 TiN
100 brownishred 1.15 55 10562 TiN
1196
Q. Fang, J.-Y. Zhang / International Journal of Inorganic Materials 3 (2001) 1193 – 1196
lower sheet resistance and higher film deposition rate were prepared by reactive sputtering when N 2 partial pressures is about 40%. It was also found those colours of the TiN x films changed from metallic grey to gold and finally brownish red as the nitrogen partial pressure increases. The maximum current density in an AMTEC test cell using TiN thin film as electrode at 8008C is about 1.0 A cm 22 and a maximum power density of 0.2 W cm 22 can be achieved with an electrode of 30 cm 2 .
References [1] Illmann U, Ebert R, Reiße G, Freller H, Lorenz P. Thin Solid Films 1994;241(1–2):71–5. [2] Quinto DT. Int J Refract Met Hard Mater 1996;14(1–3):7–20.
¨ [3] Boßmann H-P, Fang Q, Knodler R, Harbach F. In: 26th. IECEC. Boston, vol. 5, 1991, pp. 481–6. ¨ [4] Fang Q, Knodler R. J Mater Sci 1992;27:6725–9. [5] Fang Q, Wendt H. J Appl Electrochem 1996;26:43–52. [6] Park K-C, Kim K-B, Raaijmaker IJ, Ngan K. J Appl Phys 1996;80:5674. [7] Sherman A. Jpn J Appl Phys 1991;30:3553. [8] Fang Q. Ph.D thesis (Doktor der Naturwissenschaften), Technische Hochschule Darmstadt, 1993. [9] Harbach F, Knoedler R, Bossmann H-P. Physik in unserer Zeit 1992;1:34. [10] Hunt TK, Weber N, Cole T. In: Bates JB, Farrington GC, editors, Fast ionic transport in solids, North-Holland, 1981, p. 263. [11] Chou TC, Adamson D, Mardinly J, Nieh TG. Thin Solid Films 1991;205:131. [12] Roquiny P, Bodart F, Terwagne G. Surf Coat Technol 1999;119:278–83.
Nano-porous TiN thin films deposited by reactive sputtering method Q. Fang a , *, J.-Y. Zhang b b
a Institute of Solid State Physics, Chinese Academy of Sciences, Hefei 230031, PR China Department of Electrical and Electronic Engineering, University College London, London WC1 E 7 JE, UK
Abstract Nano-porous TiN thin films deposited by a reactive sputtering process are reported. The effect of deposition parameters including sputtering power density, nitrogen partial pressure and deposition time on the thin film growth has been investigated. Crystallisation behaviour, chemical composition and microstructure of the deposited films were also investigated using X-ray diffraction, EDX and scanning electron microscopy. A test cell of Na(l) / beta’’ alumina / TiN was set up and tested at the temperature range of 600–8008C, in order to investigate the cell power density and the interfacial electrical property between the electrolyte and TiN electrode. The maximum power density of 0.20 W cm 22 could be achieved with a large electrode area of 30 cm 2 at 8008C. The effect of microstructure of the nanometer size thin film electrodes on the conductive property has been studied and discussed. 2001 Elsevier Science Ltd. All rights reserved. Keywords: TiN thin films; Reactive sputtering; Nano-porous; Columnar structure
1. Introduction Titanium nitride (TiN) is a versatile material using for surface hardening of cutting tools [1,2], high temperature electrodes [3–5] and diffusion barrier layers in semiconductor devices and ferroelectric memory capacitors [6,7], because of its outstanding properties such as high surface hardness, good corrosion resistance at room temperature and elevated temperatures, low friction coefficient and high electrical conductivity [8]. The alkali metal thermoelectric converter (AMTEC) is a system in which beta’’-alumina acts as a solid sodium ion conductive electrolyte. Sodium is used as the working medium, for direct conversion of heat into electricity [3–5,9]. The anode is normally liquid sodium itself, and the cathode usually is a porous refractory metal layer such as Mo or W. However, refractory metals show high initial power densities but degrade with strong reduction in power and efficiency after 10 to 1000 h, due to electrochemical reactions [8,10]. On the other hand, being a three-phase boundary, consisting of beta-alumina, sodium gas and electron conductor, the cathode has to supply the electrons to reduce the sodium ions, which goes through beta*Corresponding author. Present address: The Centre for Nanotechnology & Microengineering, University of Warwick, Coventry CV4 7AL, UK. Tel.: 144-2476-524-216; fax: 144-2476-418-922. E-mail address: [email protected] (Q. Fang).
alumina, to sodium gas. There are two contradictive requirements for the cathode materials — good conductivity and proper porosity. Therefore, the material and structure of the thin cathode layer play extremely important role in AMTEC-operation. Recently, the refractory compounds such as TiN, TiB 2 have been investigated as electrode layer for AMTEC, in order to improve the performance of AMTEC [3,4,8,9]. TiN films are usually manufactured by deposition with thermal chemical vapour deposition (CVD), plasma-enhanced CVD (PECVD) and physical vapour deposition. However, most of CVD methods require higher temperature to achieve the deposition, which sometimes causes thermal damage to the deposited films. Therefore, low substrate-temperature deposition methods such as sputtering have been used widely. However, the sputter deposition of alloys and compounds would be difficult, since the yields of various materials differ in a sputter process. Oxides, nitrides and other materials with highly volatile constituents will not sputter in an inert gas environment to produce stochiometric films [11]. In order to produce films incorporated with the required quantities of gaseous species, reactive sputtering is used in which the inert feed gas is mixed with a reactive gas. In general, reactive sputtering is characterised by two mechanisms in its operation. (a) The surface of a target (usually is pure metal) is cleared of the nitride, oxide or cabonide compound by a high ion bombardment flux at low partial
1466-6049 / 01 / $ – see front matter 2001 Elsevier Science Ltd. All rights reserved. PII: S1466-6049( 01 )00124-6
1194
Q. Fang, J.-Y. Zhang / International Journal of Inorganic Materials 3 (2001) 1193 – 1196
Table 1 The experimental conditions of TiN x reactive sputtering deposition Sputtering power (W) Distance between target and substrate (mm) System vacuum (mbar) N 2 /(Ar1N 2 ) (%) Sputtering time (min) Sputtering temperature (8C)
200–600 50 6310 23 0–100 30–300 25
pressures of reactive gas (N 2 , O 2 , etc.). (b) Increasing the partial pressure of reactive gas results in a target surface that is converted to the compound material. Optimising the partial pressures of reactive gas, sputtering power and substrate temperature to obtain the desired film properties can control the reactive deposition process. In this paper, the chemical composition, crystal structures of sputterdeposited TiN were studied. Its electrical conductivity and dependence on deposition condition also were investigated. In addition, the power densities of TiN sputter-deposited thin films as electrode for an AMTEC test cell and dependence on film structures will be discussed.
2. Experimental A pure Ti (99.99) target (Leyboldt–Heraeus) was used for deposition of TiN thin films on flat glass and b‘‘-Al 2 O 3
tube of 25 mm diameter and 200 mm length by a DC magnetron reactive sputtering. The sputtering power was from 300 to 500 W in an Ar (99.99) / N 2 (99.99) environment. At first, the target surface and the reactor chamber were cleared of nitride or oxide and residual oxygen at a pure Ar feed gas, then increasing partial pressure of the reactive gas N 2 , a stochiometric TiN film can be produced at a proper nitrogen partial pressure. Crystallisation behaviour, chemical composition and microstructure of the deposited films were investigated using X-ray diffraction, EDX and scanning electron microscopy. A test cell of Na (l) / beta’’ alumina / TiN was set up and test at the temperature range of 400–8008C, in order to investigate the cell power density and the interfacial electrical property between the electrolyte and TiN electrode. The details of the test rig and processing can be found elsewhere [3,4,9]. The experimental conditions under which the sputtering deposition was carried out are listed in Table 1.
3. Results and discussion Fig. 1 shows SEM of TiN thin film sputtered on a glass (a) and on a beta’’-alumina tube (b), respectively. The film on the glass (a) shows a columnar structure, which are also found in Mo and TiB 2 sputter layers. The column size in the TiN film is about 50–80 nm. While the film on the
Fig. 1. SEM of TiN thin film sputtered on a glass (a) and on a beta’’-alumina tube (b).
Q. Fang, J.-Y. Zhang / International Journal of Inorganic Materials 3 (2001) 1193 – 1196
1195
Fig. 2. Current–voltage relationship in an AMTEC cell using TiN as electrode at different temperatures (a) and the corresponding power density–current curves (b).
Table 3 TiN x thin films deposited using various sputtering powers ( pN2 40% and 60 min) Sputtering power (W) Thickness (mm) Surface resistance (V / cm)
200 0.2 100
300 0.3 80
400 0.8 60
500 1.2 50
titanium content in the films is reached (N 2 /(Ar1N 2 ) (%).40), the absorbed spectra move towards slightly more reddish energies due to extra nitrogen in the films. The current–voltage relationship in an AMTEC cell using TiN thin film as electrode at different temperatures (a) and the corresponding power density–current curves (b) are shown in Fig. 2. It can be seen that a maximum current density of about 1.0 A cm 22 and a maximum power density of 0.2 W cm 22 could be achieved with 30 cm 2 electrode. It is noted that the good performance was achieved with thickness of the TiN film of 2–3 mm only. From our previous work [4], the optimal thickness of the TiN film, which is sputtered directly from a TiN target, is between 5 and 6 mm. The decrease of performance for a thinner electrode is due to higher sheet resistance. It is believed that the improvement of the AMTEC performance by the reactive sputtering TiN films rises from its lower porosity, lower sheet resistance and uniform columnar structure, which might benefit to electron transfer and letting a higher sodium vapour pressure looses in an electrochemical cell.
600 1.25 45
beta’’-alumina tube (b) shows a cauliflower structure, which consist of a number of columns. This columnar structure of TiN films is important for letting the sodium gas pass through the cathode layers itself during an electrochemical processing of AMTEC. Table 2 shows a relationship of N 2 partial pressures and deposition results from the reactive sputtering of TiN x films. Table 3 shows a dependence of sputtering power on TiN film thickness and surface conductivity, which indicated that the thickness and surface conductivity of TiN film increase with a rising sputtering power. Table 2 shows that stochiometric TiN films with lower sheet resistance and higher film deposition rate were prepared by reactive sputtering (500 W and 60 min) when N 2 partial pressures is about 40%. It was also found that colours of the deposited TiN x films changed from metallic grey to gold and finally brownish red as the nitrogen partial pressure increases. According to the Drude model [12], these TiN x thin films colour variations can be explained by titanium metallic bonds in the film. As the metallic bonds in the film becomes weaker, the absorbed spectra move towards lower energies; when the minimum
4. Conclusions Nano-porous TiN thin films deposited by a reactive sputtering process are reported. The effect of deposition parameters including sputtering power density, nitrogen partial pressure and deposition time on the thin film growth has been investigated. The stochiometric TiN films with
Table 2 TiN x thin films deposited under various N 2 partial pressures at 500 W, 60 min N 2 /(Ar1N 2 ) (%) Film colour
0 grey
10 dark grey
20 yellow
40 gold
60 light-brown
Thickness (mm) Sheet resistance (V / cm) N / Ti (%) Crystal phases
0.9 50 0 a-Ti
1.0 80 30.561 a-Ti, d-TiN
1.15 60 85.562 d-TiN, TiN
1.3 45 9962 TiN
1.2 50 10362 TiN
100 brownishred 1.15 55 10562 TiN
1196
Q. Fang, J.-Y. Zhang / International Journal of Inorganic Materials 3 (2001) 1193 – 1196
lower sheet resistance and higher film deposition rate were prepared by reactive sputtering when N 2 partial pressures is about 40%. It was also found those colours of the TiN x films changed from metallic grey to gold and finally brownish red as the nitrogen partial pressure increases. The maximum current density in an AMTEC test cell using TiN thin film as electrode at 8008C is about 1.0 A cm 22 and a maximum power density of 0.2 W cm 22 can be achieved with an electrode of 30 cm 2 .
References [1] Illmann U, Ebert R, Reiße G, Freller H, Lorenz P. Thin Solid Films 1994;241(1–2):71–5. [2] Quinto DT. Int J Refract Met Hard Mater 1996;14(1–3):7–20.
¨ [3] Boßmann H-P, Fang Q, Knodler R, Harbach F. In: 26th. IECEC. Boston, vol. 5, 1991, pp. 481–6. ¨ [4] Fang Q, Knodler R. J Mater Sci 1992;27:6725–9. [5] Fang Q, Wendt H. J Appl Electrochem 1996;26:43–52. [6] Park K-C, Kim K-B, Raaijmaker IJ, Ngan K. J Appl Phys 1996;80:5674. [7] Sherman A. Jpn J Appl Phys 1991;30:3553. [8] Fang Q. Ph.D thesis (Doktor der Naturwissenschaften), Technische Hochschule Darmstadt, 1993. [9] Harbach F, Knoedler R, Bossmann H-P. Physik in unserer Zeit 1992;1:34. [10] Hunt TK, Weber N, Cole T. In: Bates JB, Farrington GC, editors, Fast ionic transport in solids, North-Holland, 1981, p. 263. [11] Chou TC, Adamson D, Mardinly J, Nieh TG. Thin Solid Films 1991;205:131. [12] Roquiny P, Bodart F, Terwagne G. Surf Coat Technol 1999;119:278–83.