- Email: [email protected]
Nitriding of zirconium cathode for arc-heater testing in air
Vacuum 85 (2010) 591e595
Contents lists available at ScienceDirect
Vacuum journal homepage: www.elsevier.com/locate/vacuum
Nitriding of zirconium cathode for arc-heater testing in air Ai Momozawa a, *, Sven Taubert a, Satoshi Nomura a, Kimiya Komurasaki a, Yoshihiro Arakawa b a b
Department of Advanced Energy, The University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa, Chiba 277 8561, Japan Department of Aeronautics and Astronautics, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113 8656, Japan
a b s t r a c t Keywords: Plasma Arc-heater Nitriding Oxidation Zirconium cathode Argoneair mixture gas Specific enthalpy
Zirconium cathodes with a nitrided surface layer are promising candidates for arc-heater tests in air because of their good erosion resistance and excellent physical properties. Nitriding of a zirconium cathode is performed by a microwave plasma generator. Nitrided cathodes yield a stable plasma flow of argoneair up to 80 A at a maximum flow rate of 4 l/min argon and 0.8 l/min air for a long duration. The specific input power corresponds to 11.8 MJ/kg. During arc-heater testing, nitriding is a governing reaction and golden zirconium nitride is formed at the tip of the cathode surface. The formation mechanism of a ceramic layer during microwave plasma generator treatment and arc-heater testing will also be discussed. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction For the investigation of Thermal Protection Systems (TPS) for re-entry vehicles, it is necessary to simulate the atmospheric reentry conditions [1,2]. An arc-heater is one of the promising plasma generators that can be used to simulate the re-entry environment in a ground test. In our previous research, we developed an arc-heater using a zirconium cathode with a thin oxidized zirconia surface and a premixed injection system [3,4]. A stable discharge was successfully maintained over 2 h without significant erosion of the zirconium cathode for an electric current of 40 A, which is about 3.5 MJ/kg of the specific enthalpy, and a high dissociation rate of the oxygen flow. For the simulation of a more severe atmospheric re-entry condition, high specific enthalpies are required at a high total pressure (Fig. 1). The specific enthalpy is a function of plasma temperature and is estimated as follows; assuming an isentropic expansion and chemically frozen flow through the nozzle, the total specific enthalpy h0 is conserved, which is expressed as:
Z h0 ¼
1 Cp dT þ u2 þ hchem 2
(1)
where Cp is the specific heat at constant pressure, u the flow velocity, hchem the chemical potential [5]. The final goal of our study is to produce an arc-heater that can achieve higher plasma temperatures by means of a higher input current, and which can produce a pure plasma flow without
* Corresponding author. Fax: þ81 4 7136 4030. E-mail address: [email protected] (A. Momozawa). 0042-207X/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.vacuum.2010.08.017
significant erosion of the cathode using dry air as the working gas. In the present work, we describe the development of a new cathode by oxidation or nitriding using a microwave plasma generator (MPG) and the present results of the arc-heater testing. 2. Experimental method and apparatus 2.1. Experimental procedure Zirconium cathode spikes were oxidized and nitrided by a microwave plasma generator (MPG). Arc-heater tests were performed using oxidized and nitrided cathodes with argoneoxygen and argoneair gas mixtures to determine the capacity for application in an arc-heater. X-ray diffraction (XRD) analyses (RINT-2500 V, Rigaku), using CuKa radiation, were conducted to determine the phases of the samples. The analyses of microstructures were performed with a scanning electron microscope (SEM, LEO 440i) and a field-emission scanning electron microscope (FE-SEM, S4200, Hitachi). 2.2. Microwave plasma generator (MPG) The microwave power supply of the MPG can provide up to 2 kW at a maximum current of 750 mA. The microwave generator runs at 2.42 GHz. The oxygen or nitrogen gas in the crossing area absorbs the energy from microwaves and is ionized. The cathode spike is also located in this section. The volume flow rate was set to 1 l/min of oxygen and nitrogen, the pressure to 133 Pa, and the input power to 600 W for different duration time. To ignite the arc and to avoid the contamination of the cathode surface, 1 l/min of argon gas flow was input for a minute at the beginning of the experiments.
592
A. Momozawa et al. / Vacuum 85 (2010) 591e595
Fig. 1. Re-entry environment of the re-entry vehicles. Fig. 3. Zirconium cathode spike and copper socket of a newly designed cathode.
2.3. Arc-heater configuration The arc-heater configuration is shown in Fig. 2. This laboratory model is primarily made of copper. The anode is cooled by water. The newly developed cathode has a diameter of 6 mm to allow application of higher electric currents. A water-cooled copper socket was used to improve cooling (Fig. 3). The copper socket compensates poor thermal conductivity of zirconia [4] and reduces erosion of the cathode spike. The working principle of the arc-heater is explained by Matsui et al. [3,4]. The working gas can consist of dry air, argon and/or pure oxygen. According to experimental requirements, the working gas can be mixed in different ratios. The current varies between 40 and 90 A. The voltage is between 20 and 40 V, depending on the distance between anode and cathode, on the working gas and its pressure, and on the accretion point of the arc. 3. Results 3.1. Oxidation and nitriding of zirconium cathode spike by the MPG In order for the zirconium cathode to be applied for arc-heater testing with argoneair working gas, a refractory oxide or nitride
Fig. 2. Structural design of arc-heater.
Fig. 4. Zirconium cathode spikes: (a) before MPG treatment, (b, c) after oxidation, and (d) after nitriding.
layer is required to be formed on the cathode surface. Fig. 4 shows the cathodes before and after the oxidation and nitriding by MPG. 3.1.1. Oxidation of zirconium cathode spike Oxidized zirconium (zirconia) surface yielded a black or white surface. Only a black-colored zirconia cathode successfully created a plasma flow at the arc-heater testing. The cathode surface color is mainly determined by the depth of the copperezirconium connection (Fig. 5), if other experimental conditions are identical. The depth of the copperezirconium connection significantly influences the
Fig. 5. Two different conditions of copper socket and zirconium cathode spike connection.
A. Momozawa et al. / Vacuum 85 (2010) 591e595
Fig. 6. X-ray diagrams of oxidized and nitrided cathode surface by MPG.
593
3.1.2. Nitriding of zirconium cathode spike The nitrided zirconium surface is dark greyeblack (Fig. 5). According to Adachi et al., zirconium nitride has a good electrical conductivity and is also applicable to the arc-heater testing [9]. The XRD pattern of the nitrided cathode is expected to fit the pattern of the nitrides. However, they do not fit the reflection patterns of the nitrides, but showed the same patterns of zirconia phases (Fig. 6). We investigated the oxidized and nitrided layers using SEM. Fig. 7 shows the SEM micrographs of the cross-sections of the samples. After 30 min of treatment, the oxidized layer was much thicker: The oxidized layer is 25 mm thick and the nitride layer 8 mm, respectively. The nitride layer thickness does not seem to exceed over 20 mm even with longer duration of nitriding over 60 min. Furthermore, the texture and the color of the ceramic layers were different (see Fig. 6). The results suggest that nitriding occurs, although the zirconia structure appears to be maintained. 3.2. Arc-heater testing operating with air
Fig. 7. SEM micrographs of the cross-sections of oxidized (left) and nitrided (right) cathode by MPG.
heat transfer from the spike surface. In the case of a good connection, the spike is exposed to lower temperatures and white-colored zirconia is formed. On the other hand, a loose connection yields higher temperatures and hence, black-colored zirconia is formed. According to previous reports [6e8], this “blackened” zirconia occurs at higher temperatures and forms sub-stoichiometric ZrO2-x. This kind of zirconia has non-zero electrical conductivity. Its ability to ignite an arc is attributed to the creation of a plasma flow.
After an arc-heater test, both oxidized and nitrided cathode spikes showed the metallic golden color of zirconium nitride, i.e., ZrN was formed on the surface during the experiments [10e12]. White phase was also formed at the edge of the cathode spike. Though it was impossible to identify this phase, it might be zirconium oxy-nitride or zirconia. Several experiments with different electric currents and gas flow rates were carried out to determine the suitability for application in an arc-heater with reactive gas flow. A newly designed oxidized cathode is able to withstand 70 A at a gas flow rate of 2 l/min argon and 0.2 l/min air (air rate: 10 vol.%) without any significant erosion [13]. However, in the cases of the higher electric current input and the higher flow rate, the operation time was limited to minutes as severe cathode erosion occurred shortly after ignition. On the other hand, a stable discharge for a long duration time, more than 20 min, was possible with a nitrided cathode. Significantly less erosion was observed in this case. Fig. 8 and Fig. 9 show cross-sectional SEM micrographs and the XRD patterns of the oxidized and nitrided cathodes after arc-heater testing. The experimental conditions are tabulated in Table 1. Note that the operation time of the oxidized cathode is not identical to that of nitrided cathode because of the severe erosion of the former. As a result of experiments, we concluded the nitrided cathode resists up to 80 A at a maximum flow rate of 4 l/min argon and 0.8 l/ min air over 30 min and provides a stable arc discharge. The specific input power corresponds to 11.8 MJ/kg.
Fig. 8. SEM micrographs of the cross-sections of oxidized (left) and nitrided (right) cathode after arc-heater testing.
594
A. Momozawa et al. / Vacuum 85 (2010) 591e595 Table 2 Physical properties of ZrO2 and ZrN.
Melting point, K Work function Heat conductivity, W m1 K1 (at room temperature) Current density,a A/mm2 (at melting temperature) a
Fig. 9. X-ray diagrams of oxidized and nitrided cathode surface after arc-heater testing.
ZrO2
ZrN
3000 3.96 4.0 2.4
3700 2.92 21.9 380
Applying RichardsoneDushman formula.
Nitriding governs the reaction at the tip of the cathode spike, which has the highest temperature, whereas oxy-nitriding and/or oxidation seem to occur at the lower temperature area of the spike. Dissociated oxygen and nitrogen atoms also diffuse into the zirconium cathode spike at the same time. The existence of lemonyellow colored Zr2N2O and white Zr7N4O8, at the ceramic and metal boundary after arc-heater testing indicates this reaction. 4.2. Zirconium nitride as cathode spike
4. Discussion The experimental results indicated that a nitrided zirconium cathode is highly suitable for the arc-heater testing operating in air. It is necessary to understand the reaction mechanisms of nitriding during MPG treatment and arc-heater testing for optimization.
4.1. Reaction mechanism 4.1.1. Nitriding by MPG As discussed in Section 3.1.2, the XRD pattern of the nitrided cathode closely fitted with the patterns of zirconia phases (Fig. 6), although the nitriding does in fact occurs. It is believed that the zirconia structure was formed because of the high reactivity of dissociated oxygen with zirconium and the presence of residual oxygen. According to Lyapin [14], zirconium actively adsorbs oxygen at temperatures over 500 K. Residual oxygen, O2, H2O and other impurities are present inside the nitrogen gas supply and in the residual air of the MPG chamber after evacuation. These minute amount of impurities significantly influence the reaction. Oxygen dissociates at lower energies than nitrogen and then reacts strongly with zirconium and forms ZrO2-x. With increasing temperature, an active nitrogen atom is incorporated at a vacancy or replaces an oxygen site of ZrO2-x during nitriding. This idea is supported by observing the XRD results (Fig. 6). The peaks of the nitrided cathode are shifted to lower angles compared to those of oxidized samples, which means that nitrogen atoms are incorporated at zirconia vacancy sites. This “ZrO2 structure type” zirconium nitride may be stable under the MPG process conditions at high temperatures and the low pressure of 133 Pa. Further investigation of the correlations of the crystal structures of zirconia, zirconium oxy-nitrides, and zirconium nitride would help to better understand the mechanisms involved. 4.1.2. Reaction during arc-heater testing According to the experimental results, it seems that the governing reaction on the cathode spike depends on the surface temperature. Table 1 Experimental conditions of arc-heater testing. Oxidized Operation time, min Input power, current Maximum flow rate argon/air, l/min Air: argon ratio Plenum pressure, atm.
Nitrided >20
3 12 kW, 60 A 3/0.4 0.13 0.45
3/0.8 0.26 0.6
The nitrided cathode was found to be more resistant to the attack by dissociated or ionized oxygen and consequently, is more resistant to erosion during arc-heater testing. This good resistance is attributed to the strong covalent bonding of zirconium nitride, whereas ZrO2-x has weak ionic bonding. The oxy-nitride layer and the oxy-nitriding itself seem to retard the oxygen diffusion. Zirconium nitride has better physical properties as a cathode than zirconia [9,15]. As mentioned previously, to obtain the high specific enthalpy needed for re-entry simulations, it is necessary to attain a higher plasma temperature. With its high melting point, superior work function, and thermal conductivity, zirconium nitride can theoretically provide higher current density than ZrO2, which allows applications to apply a higher input current, resulting in a higher specific enthalpy (Table 2). 5. Conclusions Oxidation and nitriding of zirconium cathode by a microwave plasma generator (MPG), followed by arc-heater testing, were successfully performed in an argoneair flow. The following results were obtained: (1) The zirconium cathode spike with a nitrided surface is highly suitable as cathode material for arc-heater testing in air; a stable discharge for a long duration is possible. (2) The nitrided cathode spike showed excellent erosion resistance during arc-heater testing compared to the oxidized cathode. (3) During arc-heater testing, nitriding is a governing reaction at the tip of the cathode spike and golden ZrN is formed at the surface, whereas oxy-nitriding and/or oxidation seem to occur at the edge area. (4) Nitrided cathode resists up to 80 A at a maximum flow rate of 4 l/min argon and 0.8 l/min air and provides a stable arc discharge. The specific input power corresponds to 11.8 MJ/kg. References [1] Birkman MA. J Propul Power 1996;12:1011e7. [2] Auweter-Kurtz M, Kurtz HL, Laure S. J Propul Power 1996;12:1053e61. [3] Matsui M, Ikemoto T, Takayanagi H, Komurasaki K, Arakawa Y. Generation of atomic oxygen flows by an arcjet using a zirconium hollow cathode. Adv Appl Plasma Sci 2005;5:55e60. [4] Matsui M, Komurasaki K, Arakawa Y. Sensitivity enhancement of laser absorption spectroscopy for atomic oxygen measurement in microwave air plasma. Vacuum 2008;83(1):21e2. [5] Matsui M, Komurasaki K, Herdrich G, Auweter-Kurtz M. AIAA J 2005;9: 2060e4. [6] Ackermann RJ, Garg SP, Rauh EG. J Am Ceram Soc 1977;60(7e8):341e4.
A. Momozawa et al. / Vacuum 85 (2010) 591e595 [7] Predel B. LandolteBörnstein group IV, vol. 5I. Berlin: Springer-Verlag; 1997. p. 1e4. [8] Schneider F. PhD thesis. Germany: Justus-Liebig-Universität Giessen; 2004. [9] Adachi J, Kurosaki K, Uno M, Yamanaka S. J Alloys Compounds 2005;399: 242e4. [10] Badwal SPS, Bannister MJ, Hannink RHJ. Science and technology of zirconia V. Lancaster: Technomic Publishing Company Inc.; 1993.
595
[11] Predel B. LandolteBörnstein group IV, vol. 5H. Berlin: Springer-Verlag; 1997. p. 1e3. [12] Okamoto H. J Phase Equilibria Diffusion 2006;27:551. [13] Momozawa A, Taubert S, Nomura S, Komurasaki K, Arakawa Y. Zirconium cathode for arc-heaters operating with air. Adv Appl Plasma Sci 2009;7:43e6. [14] Lyapin A. The initial stages of the oxidation of zirconium. PhD thesis, Germany: Universität Stuttgart; 2005. [15] Williams WS. JOM 1998;6:62e6.
Contents lists available at ScienceDirect
Vacuum journal homepage: www.elsevier.com/locate/vacuum
Nitriding of zirconium cathode for arc-heater testing in air Ai Momozawa a, *, Sven Taubert a, Satoshi Nomura a, Kimiya Komurasaki a, Yoshihiro Arakawa b a b
Department of Advanced Energy, The University of Tokyo, 5-1-5 Kashiwanoha, Kashiwa, Chiba 277 8561, Japan Department of Aeronautics and Astronautics, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113 8656, Japan
a b s t r a c t Keywords: Plasma Arc-heater Nitriding Oxidation Zirconium cathode Argoneair mixture gas Specific enthalpy
Zirconium cathodes with a nitrided surface layer are promising candidates for arc-heater tests in air because of their good erosion resistance and excellent physical properties. Nitriding of a zirconium cathode is performed by a microwave plasma generator. Nitrided cathodes yield a stable plasma flow of argoneair up to 80 A at a maximum flow rate of 4 l/min argon and 0.8 l/min air for a long duration. The specific input power corresponds to 11.8 MJ/kg. During arc-heater testing, nitriding is a governing reaction and golden zirconium nitride is formed at the tip of the cathode surface. The formation mechanism of a ceramic layer during microwave plasma generator treatment and arc-heater testing will also be discussed. Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction For the investigation of Thermal Protection Systems (TPS) for re-entry vehicles, it is necessary to simulate the atmospheric reentry conditions [1,2]. An arc-heater is one of the promising plasma generators that can be used to simulate the re-entry environment in a ground test. In our previous research, we developed an arc-heater using a zirconium cathode with a thin oxidized zirconia surface and a premixed injection system [3,4]. A stable discharge was successfully maintained over 2 h without significant erosion of the zirconium cathode for an electric current of 40 A, which is about 3.5 MJ/kg of the specific enthalpy, and a high dissociation rate of the oxygen flow. For the simulation of a more severe atmospheric re-entry condition, high specific enthalpies are required at a high total pressure (Fig. 1). The specific enthalpy is a function of plasma temperature and is estimated as follows; assuming an isentropic expansion and chemically frozen flow through the nozzle, the total specific enthalpy h0 is conserved, which is expressed as:
Z h0 ¼
1 Cp dT þ u2 þ hchem 2
(1)
where Cp is the specific heat at constant pressure, u the flow velocity, hchem the chemical potential [5]. The final goal of our study is to produce an arc-heater that can achieve higher plasma temperatures by means of a higher input current, and which can produce a pure plasma flow without
* Corresponding author. Fax: þ81 4 7136 4030. E-mail address: [email protected] (A. Momozawa). 0042-207X/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.vacuum.2010.08.017
significant erosion of the cathode using dry air as the working gas. In the present work, we describe the development of a new cathode by oxidation or nitriding using a microwave plasma generator (MPG) and the present results of the arc-heater testing. 2. Experimental method and apparatus 2.1. Experimental procedure Zirconium cathode spikes were oxidized and nitrided by a microwave plasma generator (MPG). Arc-heater tests were performed using oxidized and nitrided cathodes with argoneoxygen and argoneair gas mixtures to determine the capacity for application in an arc-heater. X-ray diffraction (XRD) analyses (RINT-2500 V, Rigaku), using CuKa radiation, were conducted to determine the phases of the samples. The analyses of microstructures were performed with a scanning electron microscope (SEM, LEO 440i) and a field-emission scanning electron microscope (FE-SEM, S4200, Hitachi). 2.2. Microwave plasma generator (MPG) The microwave power supply of the MPG can provide up to 2 kW at a maximum current of 750 mA. The microwave generator runs at 2.42 GHz. The oxygen or nitrogen gas in the crossing area absorbs the energy from microwaves and is ionized. The cathode spike is also located in this section. The volume flow rate was set to 1 l/min of oxygen and nitrogen, the pressure to 133 Pa, and the input power to 600 W for different duration time. To ignite the arc and to avoid the contamination of the cathode surface, 1 l/min of argon gas flow was input for a minute at the beginning of the experiments.
592
A. Momozawa et al. / Vacuum 85 (2010) 591e595
Fig. 1. Re-entry environment of the re-entry vehicles. Fig. 3. Zirconium cathode spike and copper socket of a newly designed cathode.
2.3. Arc-heater configuration The arc-heater configuration is shown in Fig. 2. This laboratory model is primarily made of copper. The anode is cooled by water. The newly developed cathode has a diameter of 6 mm to allow application of higher electric currents. A water-cooled copper socket was used to improve cooling (Fig. 3). The copper socket compensates poor thermal conductivity of zirconia [4] and reduces erosion of the cathode spike. The working principle of the arc-heater is explained by Matsui et al. [3,4]. The working gas can consist of dry air, argon and/or pure oxygen. According to experimental requirements, the working gas can be mixed in different ratios. The current varies between 40 and 90 A. The voltage is between 20 and 40 V, depending on the distance between anode and cathode, on the working gas and its pressure, and on the accretion point of the arc. 3. Results 3.1. Oxidation and nitriding of zirconium cathode spike by the MPG In order for the zirconium cathode to be applied for arc-heater testing with argoneair working gas, a refractory oxide or nitride
Fig. 2. Structural design of arc-heater.
Fig. 4. Zirconium cathode spikes: (a) before MPG treatment, (b, c) after oxidation, and (d) after nitriding.
layer is required to be formed on the cathode surface. Fig. 4 shows the cathodes before and after the oxidation and nitriding by MPG. 3.1.1. Oxidation of zirconium cathode spike Oxidized zirconium (zirconia) surface yielded a black or white surface. Only a black-colored zirconia cathode successfully created a plasma flow at the arc-heater testing. The cathode surface color is mainly determined by the depth of the copperezirconium connection (Fig. 5), if other experimental conditions are identical. The depth of the copperezirconium connection significantly influences the
Fig. 5. Two different conditions of copper socket and zirconium cathode spike connection.
A. Momozawa et al. / Vacuum 85 (2010) 591e595
Fig. 6. X-ray diagrams of oxidized and nitrided cathode surface by MPG.
593
3.1.2. Nitriding of zirconium cathode spike The nitrided zirconium surface is dark greyeblack (Fig. 5). According to Adachi et al., zirconium nitride has a good electrical conductivity and is also applicable to the arc-heater testing [9]. The XRD pattern of the nitrided cathode is expected to fit the pattern of the nitrides. However, they do not fit the reflection patterns of the nitrides, but showed the same patterns of zirconia phases (Fig. 6). We investigated the oxidized and nitrided layers using SEM. Fig. 7 shows the SEM micrographs of the cross-sections of the samples. After 30 min of treatment, the oxidized layer was much thicker: The oxidized layer is 25 mm thick and the nitride layer 8 mm, respectively. The nitride layer thickness does not seem to exceed over 20 mm even with longer duration of nitriding over 60 min. Furthermore, the texture and the color of the ceramic layers were different (see Fig. 6). The results suggest that nitriding occurs, although the zirconia structure appears to be maintained. 3.2. Arc-heater testing operating with air
Fig. 7. SEM micrographs of the cross-sections of oxidized (left) and nitrided (right) cathode by MPG.
heat transfer from the spike surface. In the case of a good connection, the spike is exposed to lower temperatures and white-colored zirconia is formed. On the other hand, a loose connection yields higher temperatures and hence, black-colored zirconia is formed. According to previous reports [6e8], this “blackened” zirconia occurs at higher temperatures and forms sub-stoichiometric ZrO2-x. This kind of zirconia has non-zero electrical conductivity. Its ability to ignite an arc is attributed to the creation of a plasma flow.
After an arc-heater test, both oxidized and nitrided cathode spikes showed the metallic golden color of zirconium nitride, i.e., ZrN was formed on the surface during the experiments [10e12]. White phase was also formed at the edge of the cathode spike. Though it was impossible to identify this phase, it might be zirconium oxy-nitride or zirconia. Several experiments with different electric currents and gas flow rates were carried out to determine the suitability for application in an arc-heater with reactive gas flow. A newly designed oxidized cathode is able to withstand 70 A at a gas flow rate of 2 l/min argon and 0.2 l/min air (air rate: 10 vol.%) without any significant erosion [13]. However, in the cases of the higher electric current input and the higher flow rate, the operation time was limited to minutes as severe cathode erosion occurred shortly after ignition. On the other hand, a stable discharge for a long duration time, more than 20 min, was possible with a nitrided cathode. Significantly less erosion was observed in this case. Fig. 8 and Fig. 9 show cross-sectional SEM micrographs and the XRD patterns of the oxidized and nitrided cathodes after arc-heater testing. The experimental conditions are tabulated in Table 1. Note that the operation time of the oxidized cathode is not identical to that of nitrided cathode because of the severe erosion of the former. As a result of experiments, we concluded the nitrided cathode resists up to 80 A at a maximum flow rate of 4 l/min argon and 0.8 l/ min air over 30 min and provides a stable arc discharge. The specific input power corresponds to 11.8 MJ/kg.
Fig. 8. SEM micrographs of the cross-sections of oxidized (left) and nitrided (right) cathode after arc-heater testing.
594
A. Momozawa et al. / Vacuum 85 (2010) 591e595 Table 2 Physical properties of ZrO2 and ZrN.
Melting point, K Work function Heat conductivity, W m1 K1 (at room temperature) Current density,a A/mm2 (at melting temperature) a
Fig. 9. X-ray diagrams of oxidized and nitrided cathode surface after arc-heater testing.
ZrO2
ZrN
3000 3.96 4.0 2.4
3700 2.92 21.9 380
Applying RichardsoneDushman formula.
Nitriding governs the reaction at the tip of the cathode spike, which has the highest temperature, whereas oxy-nitriding and/or oxidation seem to occur at the lower temperature area of the spike. Dissociated oxygen and nitrogen atoms also diffuse into the zirconium cathode spike at the same time. The existence of lemonyellow colored Zr2N2O and white Zr7N4O8, at the ceramic and metal boundary after arc-heater testing indicates this reaction. 4.2. Zirconium nitride as cathode spike
4. Discussion The experimental results indicated that a nitrided zirconium cathode is highly suitable for the arc-heater testing operating in air. It is necessary to understand the reaction mechanisms of nitriding during MPG treatment and arc-heater testing for optimization.
4.1. Reaction mechanism 4.1.1. Nitriding by MPG As discussed in Section 3.1.2, the XRD pattern of the nitrided cathode closely fitted with the patterns of zirconia phases (Fig. 6), although the nitriding does in fact occurs. It is believed that the zirconia structure was formed because of the high reactivity of dissociated oxygen with zirconium and the presence of residual oxygen. According to Lyapin [14], zirconium actively adsorbs oxygen at temperatures over 500 K. Residual oxygen, O2, H2O and other impurities are present inside the nitrogen gas supply and in the residual air of the MPG chamber after evacuation. These minute amount of impurities significantly influence the reaction. Oxygen dissociates at lower energies than nitrogen and then reacts strongly with zirconium and forms ZrO2-x. With increasing temperature, an active nitrogen atom is incorporated at a vacancy or replaces an oxygen site of ZrO2-x during nitriding. This idea is supported by observing the XRD results (Fig. 6). The peaks of the nitrided cathode are shifted to lower angles compared to those of oxidized samples, which means that nitrogen atoms are incorporated at zirconia vacancy sites. This “ZrO2 structure type” zirconium nitride may be stable under the MPG process conditions at high temperatures and the low pressure of 133 Pa. Further investigation of the correlations of the crystal structures of zirconia, zirconium oxy-nitrides, and zirconium nitride would help to better understand the mechanisms involved. 4.1.2. Reaction during arc-heater testing According to the experimental results, it seems that the governing reaction on the cathode spike depends on the surface temperature. Table 1 Experimental conditions of arc-heater testing. Oxidized Operation time, min Input power, current Maximum flow rate argon/air, l/min Air: argon ratio Plenum pressure, atm.
Nitrided >20
3 12 kW, 60 A 3/0.4 0.13 0.45
3/0.8 0.26 0.6
The nitrided cathode was found to be more resistant to the attack by dissociated or ionized oxygen and consequently, is more resistant to erosion during arc-heater testing. This good resistance is attributed to the strong covalent bonding of zirconium nitride, whereas ZrO2-x has weak ionic bonding. The oxy-nitride layer and the oxy-nitriding itself seem to retard the oxygen diffusion. Zirconium nitride has better physical properties as a cathode than zirconia [9,15]. As mentioned previously, to obtain the high specific enthalpy needed for re-entry simulations, it is necessary to attain a higher plasma temperature. With its high melting point, superior work function, and thermal conductivity, zirconium nitride can theoretically provide higher current density than ZrO2, which allows applications to apply a higher input current, resulting in a higher specific enthalpy (Table 2). 5. Conclusions Oxidation and nitriding of zirconium cathode by a microwave plasma generator (MPG), followed by arc-heater testing, were successfully performed in an argoneair flow. The following results were obtained: (1) The zirconium cathode spike with a nitrided surface is highly suitable as cathode material for arc-heater testing in air; a stable discharge for a long duration is possible. (2) The nitrided cathode spike showed excellent erosion resistance during arc-heater testing compared to the oxidized cathode. (3) During arc-heater testing, nitriding is a governing reaction at the tip of the cathode spike and golden ZrN is formed at the surface, whereas oxy-nitriding and/or oxidation seem to occur at the edge area. (4) Nitrided cathode resists up to 80 A at a maximum flow rate of 4 l/min argon and 0.8 l/min air and provides a stable arc discharge. The specific input power corresponds to 11.8 MJ/kg. References [1] Birkman MA. J Propul Power 1996;12:1011e7. [2] Auweter-Kurtz M, Kurtz HL, Laure S. J Propul Power 1996;12:1053e61. [3] Matsui M, Ikemoto T, Takayanagi H, Komurasaki K, Arakawa Y. Generation of atomic oxygen flows by an arcjet using a zirconium hollow cathode. Adv Appl Plasma Sci 2005;5:55e60. [4] Matsui M, Komurasaki K, Arakawa Y. Sensitivity enhancement of laser absorption spectroscopy for atomic oxygen measurement in microwave air plasma. Vacuum 2008;83(1):21e2. [5] Matsui M, Komurasaki K, Herdrich G, Auweter-Kurtz M. AIAA J 2005;9: 2060e4. [6] Ackermann RJ, Garg SP, Rauh EG. J Am Ceram Soc 1977;60(7e8):341e4.
A. Momozawa et al. / Vacuum 85 (2010) 591e595 [7] Predel B. LandolteBörnstein group IV, vol. 5I. Berlin: Springer-Verlag; 1997. p. 1e4. [8] Schneider F. PhD thesis. Germany: Justus-Liebig-Universität Giessen; 2004. [9] Adachi J, Kurosaki K, Uno M, Yamanaka S. J Alloys Compounds 2005;399: 242e4. [10] Badwal SPS, Bannister MJ, Hannink RHJ. Science and technology of zirconia V. Lancaster: Technomic Publishing Company Inc.; 1993.
595
[11] Predel B. LandolteBörnstein group IV, vol. 5H. Berlin: Springer-Verlag; 1997. p. 1e3. [12] Okamoto H. J Phase Equilibria Diffusion 2006;27:551. [13] Momozawa A, Taubert S, Nomura S, Komurasaki K, Arakawa Y. Zirconium cathode for arc-heaters operating with air. Adv Appl Plasma Sci 2009;7:43e6. [14] Lyapin A. The initial stages of the oxidation of zirconium. PhD thesis, Germany: Universität Stuttgart; 2005. [15] Williams WS. JOM 1998;6:62e6.