- Email: [email protected]
Nitriding of aluminum by using supersonic expanding plasma jets
Vacuum 65 (2002) 403–408
Nitriding of aluminum by using supersonic expanding plasma jets Y. Andoa,*, S. Tobea, H. Taharab, T. Yoshikawab a
Ashikaga Institute of Technology, 268-1 Omae, Ashikaga, Tochigi 326-8558, Japan b Osaka University, 1-3 Machikaneyama, Toyonaka, Osaka 560-8531, Japan
Abstract For the purpose of developing a rapid direct nitriding process to produce aluminum nitride layer without melting or vaporizing of aluminum during operation, nitridings of aluminum and aluminum alloy by using supersonic expanding nitrogen plasma jets were carried out. Nitride layer could not be formed on the substrate in the case of nitrogen gas, though the supersonic expanding plasma jets had enough reactivity to form rapidly titanium nitride layer on the surface of the substrate. Many fibers were formed on the substrate in the case of hydrogen/nitrogen mixture gas. However, the fibers could not be confirmed as aluminum nitride and the surface of the substrate came to be rough during operation in this case. Since the same phenomenon occurred in the case of hydrogen/argon mixture working gas, it was thought that the roughness of the substrate during operation was due to hydrogen infiltration into the substrate. In the case of nitriding of aluminum–magnesium alloy, unlike the case of aluminum substrate use, nitride layer was formed on the substrate even on the condition that pure nitrogen was used as working gas. From these results, this process was found to have a high potential for nitriding to produce aluminum nitride. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Surface modification; Supersonic expanding plasma jet; Thermal plasma; Plasma nitriding; Aluminum nitride
1. Introduction Aluminum nitride ceramics are promising materials with good thermal conductivity, electrical insulation property and piezoelectric property. Therefore, they are expected as substrate and package for LSI and elements for surface acoustic wave generator. So far, since most of the products made from aluminum nitride have been produced by sintering, aluminum nitride has been produced as powders. However, recently, as the products of aluminum nitride ceramics become small and *Corresponding author. Fax: +81-284-62-9802. E-mail address: [email protected] (Y. Ando).
complicated, thin film type of aluminum nitride comes to play an important role. Though thermal chemical vapor deposition and reactive sputtering have been generally used as film deposition process, these processes have some problems, for example, low deposition rate, crack generation and peeling off. So that, development of a new aluminum nitride synthesis processes is expected, while, nitriding is the process that can create gradient material between films and substrates relatively easily. Direct nitriding of aluminum and carbothermal reduction of alumina has been successfully conducted recently [1], though it was thought to be difficult to do nitriding of aluminum due to rigid oxide layer on the aluminum surface.
0042-207X/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 2 - 2 0 7 X ( 0 1 ) 0 0 4 4 9 - 3
404
Y. Ando et al. / Vacuum 65 (2002) 403–408
Especially, the nitriding process using supersonic expanding plasma jets can be expected as high deposition rate and low temperature nitriding process [2]. The supersonic expanding plasma jets are the low temperature plasma jets with high chemically reactive particle density which are created by supersonic adiabatic expansion of thermal plasma jets. According to our previous study on plasma diagnostics of supersonic expanding plasma jets [5,6], supersonic nitrogen plasma jets were 1014 in electron number density and 2000 K in rotation temperature which was almost the same value as transition temperature. So, the nitriding of aluminum without melting substrate during operation can be conducted by using this plasma jets, though conventional rapid nitriding processes for aluminum using thermal plasma jets need melting of the substrate surface during operation in order to eliminate the rigid surface oxide layer. In our previous study [3,4], nitriding of titanium and some steels using supersonic expanding nitrogen plasma jets at 30 Pa in chamber pressure was carried out. Consequently, it was proved that the plasma jets had enough reactivity to form a hard and thick nitride layer on the surface of the sample by only a few minutes plasma jets irradiation in the case of the nitriding of the titanium. Besides, thick hard layer could be formed on the substrate of the sample by using hydrogen/nitrogen plasma jet even in the case of steels. In this study, in order to investigate the reactivity of the supersonic expanding nitrogen plasma jets for aluminum, nitriding of aluminum by the supersonic expanding plasma jets was carried out.
2. Experimental procedure As reported in our previous paper, experimental apparatus for nitriding consisted of vacuum chamber, plasma torch, gas supply system, power supply system and vacuum pump. Sample holder was placed at the axial center of the plasma jets to irradiate the center of the sample (this position will be indicated as ‘‘the center’’ in the following sentences) in the vacuum chamber. Pure nitrogen
Table 1 Components of Al–Mg alloy (wt%) Mg
Cr
Al
2.5
0.25
Bal
or hydrogen/nitrogen mixture gas was used as a working gas. Aluminum plate and aluminum– magnesium alloy plate were used as substrates. Components of aluminum–magnesium alloy were given in Table 1. The discharge power was 6 kW and nitriding time was 5 min. As a working gas, pure nitrogen and hydrogen/nitrogen mixture gas were used. Flow rate of nitrogen gases were 9.6SLM in both cases and the rate of hydrogen was 3.2SLM in the case of hydrogen/nitrogen mixture gas use. Surface conditions and surface components were investigated by optical microscope and X-ray diffraction, respectively. Nitriding temperature was measured by pyrometer and controlled by varying the irradiating distance, which was the distance between the head of the plasma torch and the surface of the substrate.
3. Results and discussion 3.1. Nitriding of aluminum First of all, the relation between irradiating distance and nitriding temperature in this experiment was confirmed. Fig. 1 shows a relation between radial distance from the center and temperature on the substrate in each experimental condition in the case of pure aluminum substrate and pure nitrogen working gas use. Since plasma jets had axial and radial variations in temperature, nitriding temperature decreased as the position on the sample was far from the center and the temperature decreased with increasing irradiating distance. Nitriding temperatures at the center and at the edge was 60 mm distant from the center fell down to 833 and 803 K, respectively, on the condition that irradiating distance was 350 mm though these temperatures 973 and 833 K in the case of 200 mm in irradiating distance.
Y. Ando et al. / Vacuum 65 (2002) 403–408
Fig. 1. Relation between radial distance from the center and nitriding temperature in each experimental condition.
405
of the sample. So, in order to promote nitriding rate, we tried nitriding of aluminum on the condition of shorter irradiating distance. However, surface hardening did not occur even in the case that the sample melted down during operation. From these results, it was proved that the supersonic expanding nitrogen plasma jets used in this study did not have enough reactivity to reduce the aluminum oxide layer though this plasma jets were enough reactive to form a thick titanium nitride layer for a few minutes in the case of nitriding of titanium. Then, nitriding of aluminum using supersonic expanding plasma jets by introducing reducer gas into the nitrogen plasma jets was studied as another rapid nitriding method to produce aluminum nitride using this plasma jets. As reducer gas, though hydrogen and carbon oxide (CO) which was used in carbothermal reduction of aluminum oxide [1] were effective, hydrogen was used because CO was toxic and had a tendency to cause carburizing. As for the reaction between irradiating distance and nitriding temperature, as shown in Fig. 3, the radial profiles of the temperature were steep in comparison with that in the case of pure nitrogen plasma jets use and the temperature at the center was E150 K higher than that in the case of the nitrogen plasma in each irradiating distance. In this experiment,
Fig. 2. X-ray diffraction pattern of the aluminum and the aluminum sample nitrided by nitrogen plasma jets.
Fig. 2 shows X-ray diffraction patterns of the non-nitrided aluminum sample and the sample nitrided on the condition that the irradiating distance was 250 mm. The upper one shows the pattern of the non-nitrided sample and the lower one shows that of the nitrided sample. Only aluminum peaks were detected in both patterns and aluminum nitride could not be confirmed in the pattern of the nitrided sample. Even on the condition that the irradiating distance was so close that the substrate melted down during operation, surface hardening did not occur and aluminum nitride layer could not be formed on the substrate
Fig. 3. Relation between radial distance from the center and nitriding temperature in each experimental condition.
406
Y. Ando et al. / Vacuum 65 (2002) 403–408
irradiating distance was set to 300 mm in order to prevent melt down of the sample during operation. Figs. 4 and 5 show the appearance and the Xray diffraction pattern of the sample nitrided by hydrogen/nitrogen mixture plasma jets on the condition that the irradiating distance was 300 mm. In the case of using pure nitrogen as working gas, nitride layer was not formed on the substrate even on the condition that the substrate was melted down during operation. On the other
hand, many fibers were created on the substrate in the case of hydrogen/nitrogen mixture gas. This result suggested that these supersonic expanding plasma jets had enough chemical reactivity to destruct the aluminum oxide film preventing from nitriding of aluminum. However, aluminum nitride could not be observed in X-ray patterns of the nitrided sample and surface of the substrate came to be rough during operation in this case. The rough surface was formed due to generation of blisters during operation. The blisters appeared at the center of the substrate after the 3 min operation. The area of the blisters was expanded gradually toward the edge, and finally, the aluminum substrate was covered with the blisters. The blisters were thought to be generated by hydrogen absorption of aluminum substrate. Then, for the purpose of confirming whether the cause of blister generation was due to hydrogen absorption by aluminum substrate or not, hydrogen/argon mixture plasma jets irradiation to aluminum substrate was conducted. Consequently, almost the same phenomenon as that in the case of hydrogen/argon mixture plasma occurred and the aluminum substrate had covered with blisters after operation. Fig. 6 shows the appearance of the
Fig. 4. Appearance of the sample nitrided by hydrogen/ nitrogen mixture plasma jets.
Fig. 5. X-ray diffraction pattern of the aluminum sample nitrided by hydrogen/nitrogen mixture plasma jets.
Fig. 6. Appearance of the sample irradiated by Argon/nitrogen mixture plasma jets.
Y. Ando et al. / Vacuum 65 (2002) 403–408
407
hydrogen/argon mixture plasma jets irradiated sample. Thereby, it was thought that hydrogen infiltration into the substrate was dominant in this blister generation process. However, since these blisters were small and crowded in comparison with those in the case of hydrogen/nitrogen plasma jets, blisters generation was affected by chemical reaction between nitrogen and aluminum as well. 3.2. Nitriding of aluminum–magnesium alloy As the aluminum oxide removal method without addition of some reducer gas to plasma jets, introducing additives into aluminum substrate as reducer was thought to be effective. Next, rapid nitriding to obtain aluminum nitride without blister generation was tried by using aluminum– magnesium alloy plate as substrate. Figs. 7 and 8 show the appearance and the X-ray diffraction patterns of the aluminum–magnesium alloy sample nitrided on the same condition as the case of the nitriding of aluminum by pure nitrogen plasma jets. In this case, unlike the case of pure aluminum substrate, nitride layer was formed and surface of
Fig. 7. Appearance of the sample nitrided by hydrogen/ nitrogen mixture plasma jets.
Fig. 8. X-ray diffraction pattern of the aluminum–magnesium alloy sample nitrided by nitrogen plasma jets.
the substrate had become frosted on the substrate on the condition that irradiating distance was 250 mm, besides this phenomenon was observed even in the case of 350 mm in irradiating distance. According to the previous reports on direct nitriding to produce aluminum nitride [7,8], deoxidation of aluminum oxide surface layer by magnesium contented in aluminum occurred over 670 K and aluminum oxide was reduced. Then, chemical reaction between aluminum and nitrogen occurred and aluminum nitride layer was produced. These reactions could be expressed from the thermodynamic viewpoint by using the following equations. 3Mg þ 4Al2 O3 -3MgAl2 O4 þ 2Al
ð1Þ
3Mg þ Al2 O3 -MgO þ 2Al:
ð2Þ
Almost the same reactions might occur in our experiment. In this experiment, unlike the previous results in this study, the existence of aluminum nitride could be confirmed from not only the appearance of the nitrided sample but also the Xray diffraction pattern of the nitrided sample. From these results, it was proved that rigid aluminum oxide could be removed without melting or vaporization of aluminum by introducing reducer into aluminum substrate, besides, this process using supersonic expanding plasma jets was found to have a high potential for rapid direct nitriding process to produce aluminum nitride without melting and vaporization of aluminum.
408
Y. Ando et al. / Vacuum 65 (2002) 403–408
4. Conclusions In this study, the nitridings of aluminum and aluminum–magnesium alloy by using supersonic plasma jets were studied in order to develop a rapid direct nitriding process to produce aluminum nitride without melting and vaporization of aluminum during operation. The conclusions of this study were summarized as follows: (1) Nitride layer could not be formed on the surface of the aluminum substrate on the condition that pure nitrogen plasma jets were used though the nitrogen plasma jets had enough reactivity to form nitride layer on the surface in the case of titanium substrate. (2) By addition of hydrogen to nitrogen plasma jets, the aluminum nitride layer could be formed due to reduction of the rigid aluminum oxide surface layer by hydrogen or nitrogen hydride. However, surface of the substrate came to be rough because the aluminum absorbed hydrogen during operation. (3) By introducing magnesium as reducer into the aluminum substrate, aluminum nitride layer could be formed on the surface of the sample
without melting or vaporization of even on the condition that pure plasma jets were used. (4) From these results, this process was have a high potential for rapid without melting or vaporization of to produce aluminum nitride.
substrate nitrogen found to nitriding substrate
References [1] Komeya K, Mitsutani E, Meguro T. J Ceram Soc J 1993;101:377–82. [2] Yonezawa T, Tahara H, Andoh Y, Onoe K, Yoshikawa. Proceedings of the 20th International Symposium Space Technology and Science 1996:ISTS 96-a-3-05. [3] Andoh Y, Tobe S, Tahara H, Yoshikawa T. Proceedings of the United Thermal Spray Conference 1999. p. 234–9. [4] Ando Y, Tobe S, Tahara H, Yoshikawa T. Proceedings of the 14th International Symposium Plasma Chemistry 1999. p. 1185–90. [5] Tahara H, Uda N, Onoe K, Tsubakishita Y, Yoshikawa T. IEEE Trans Plasma Sci 1994;22:58–64. [6] Tahara H, Komiko K, Yonezawa T, Andoh Y, Yoshikawa T. IEEE Trans Plasma Sci 1996;24:218–25. [7] Kondoh K, Kimura J, Takeda Y. J Jpn Soc Powder Powder Metall 1999;46:801–10. [8] Kotani K, Ikeuchi K, Matsuda F. Quarterly J Jpn Weld Soc 1996;14:551–62.
Nitriding of aluminum by using supersonic expanding plasma jets Y. Andoa,*, S. Tobea, H. Taharab, T. Yoshikawab a
Ashikaga Institute of Technology, 268-1 Omae, Ashikaga, Tochigi 326-8558, Japan b Osaka University, 1-3 Machikaneyama, Toyonaka, Osaka 560-8531, Japan
Abstract For the purpose of developing a rapid direct nitriding process to produce aluminum nitride layer without melting or vaporizing of aluminum during operation, nitridings of aluminum and aluminum alloy by using supersonic expanding nitrogen plasma jets were carried out. Nitride layer could not be formed on the substrate in the case of nitrogen gas, though the supersonic expanding plasma jets had enough reactivity to form rapidly titanium nitride layer on the surface of the substrate. Many fibers were formed on the substrate in the case of hydrogen/nitrogen mixture gas. However, the fibers could not be confirmed as aluminum nitride and the surface of the substrate came to be rough during operation in this case. Since the same phenomenon occurred in the case of hydrogen/argon mixture working gas, it was thought that the roughness of the substrate during operation was due to hydrogen infiltration into the substrate. In the case of nitriding of aluminum–magnesium alloy, unlike the case of aluminum substrate use, nitride layer was formed on the substrate even on the condition that pure nitrogen was used as working gas. From these results, this process was found to have a high potential for nitriding to produce aluminum nitride. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Surface modification; Supersonic expanding plasma jet; Thermal plasma; Plasma nitriding; Aluminum nitride
1. Introduction Aluminum nitride ceramics are promising materials with good thermal conductivity, electrical insulation property and piezoelectric property. Therefore, they are expected as substrate and package for LSI and elements for surface acoustic wave generator. So far, since most of the products made from aluminum nitride have been produced by sintering, aluminum nitride has been produced as powders. However, recently, as the products of aluminum nitride ceramics become small and *Corresponding author. Fax: +81-284-62-9802. E-mail address: [email protected] (Y. Ando).
complicated, thin film type of aluminum nitride comes to play an important role. Though thermal chemical vapor deposition and reactive sputtering have been generally used as film deposition process, these processes have some problems, for example, low deposition rate, crack generation and peeling off. So that, development of a new aluminum nitride synthesis processes is expected, while, nitriding is the process that can create gradient material between films and substrates relatively easily. Direct nitriding of aluminum and carbothermal reduction of alumina has been successfully conducted recently [1], though it was thought to be difficult to do nitriding of aluminum due to rigid oxide layer on the aluminum surface.
0042-207X/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 4 2 - 2 0 7 X ( 0 1 ) 0 0 4 4 9 - 3
404
Y. Ando et al. / Vacuum 65 (2002) 403–408
Especially, the nitriding process using supersonic expanding plasma jets can be expected as high deposition rate and low temperature nitriding process [2]. The supersonic expanding plasma jets are the low temperature plasma jets with high chemically reactive particle density which are created by supersonic adiabatic expansion of thermal plasma jets. According to our previous study on plasma diagnostics of supersonic expanding plasma jets [5,6], supersonic nitrogen plasma jets were 1014 in electron number density and 2000 K in rotation temperature which was almost the same value as transition temperature. So, the nitriding of aluminum without melting substrate during operation can be conducted by using this plasma jets, though conventional rapid nitriding processes for aluminum using thermal plasma jets need melting of the substrate surface during operation in order to eliminate the rigid surface oxide layer. In our previous study [3,4], nitriding of titanium and some steels using supersonic expanding nitrogen plasma jets at 30 Pa in chamber pressure was carried out. Consequently, it was proved that the plasma jets had enough reactivity to form a hard and thick nitride layer on the surface of the sample by only a few minutes plasma jets irradiation in the case of the nitriding of the titanium. Besides, thick hard layer could be formed on the substrate of the sample by using hydrogen/nitrogen plasma jet even in the case of steels. In this study, in order to investigate the reactivity of the supersonic expanding nitrogen plasma jets for aluminum, nitriding of aluminum by the supersonic expanding plasma jets was carried out.
2. Experimental procedure As reported in our previous paper, experimental apparatus for nitriding consisted of vacuum chamber, plasma torch, gas supply system, power supply system and vacuum pump. Sample holder was placed at the axial center of the plasma jets to irradiate the center of the sample (this position will be indicated as ‘‘the center’’ in the following sentences) in the vacuum chamber. Pure nitrogen
Table 1 Components of Al–Mg alloy (wt%) Mg
Cr
Al
2.5
0.25
Bal
or hydrogen/nitrogen mixture gas was used as a working gas. Aluminum plate and aluminum– magnesium alloy plate were used as substrates. Components of aluminum–magnesium alloy were given in Table 1. The discharge power was 6 kW and nitriding time was 5 min. As a working gas, pure nitrogen and hydrogen/nitrogen mixture gas were used. Flow rate of nitrogen gases were 9.6SLM in both cases and the rate of hydrogen was 3.2SLM in the case of hydrogen/nitrogen mixture gas use. Surface conditions and surface components were investigated by optical microscope and X-ray diffraction, respectively. Nitriding temperature was measured by pyrometer and controlled by varying the irradiating distance, which was the distance between the head of the plasma torch and the surface of the substrate.
3. Results and discussion 3.1. Nitriding of aluminum First of all, the relation between irradiating distance and nitriding temperature in this experiment was confirmed. Fig. 1 shows a relation between radial distance from the center and temperature on the substrate in each experimental condition in the case of pure aluminum substrate and pure nitrogen working gas use. Since plasma jets had axial and radial variations in temperature, nitriding temperature decreased as the position on the sample was far from the center and the temperature decreased with increasing irradiating distance. Nitriding temperatures at the center and at the edge was 60 mm distant from the center fell down to 833 and 803 K, respectively, on the condition that irradiating distance was 350 mm though these temperatures 973 and 833 K in the case of 200 mm in irradiating distance.
Y. Ando et al. / Vacuum 65 (2002) 403–408
Fig. 1. Relation between radial distance from the center and nitriding temperature in each experimental condition.
405
of the sample. So, in order to promote nitriding rate, we tried nitriding of aluminum on the condition of shorter irradiating distance. However, surface hardening did not occur even in the case that the sample melted down during operation. From these results, it was proved that the supersonic expanding nitrogen plasma jets used in this study did not have enough reactivity to reduce the aluminum oxide layer though this plasma jets were enough reactive to form a thick titanium nitride layer for a few minutes in the case of nitriding of titanium. Then, nitriding of aluminum using supersonic expanding plasma jets by introducing reducer gas into the nitrogen plasma jets was studied as another rapid nitriding method to produce aluminum nitride using this plasma jets. As reducer gas, though hydrogen and carbon oxide (CO) which was used in carbothermal reduction of aluminum oxide [1] were effective, hydrogen was used because CO was toxic and had a tendency to cause carburizing. As for the reaction between irradiating distance and nitriding temperature, as shown in Fig. 3, the radial profiles of the temperature were steep in comparison with that in the case of pure nitrogen plasma jets use and the temperature at the center was E150 K higher than that in the case of the nitrogen plasma in each irradiating distance. In this experiment,
Fig. 2. X-ray diffraction pattern of the aluminum and the aluminum sample nitrided by nitrogen plasma jets.
Fig. 2 shows X-ray diffraction patterns of the non-nitrided aluminum sample and the sample nitrided on the condition that the irradiating distance was 250 mm. The upper one shows the pattern of the non-nitrided sample and the lower one shows that of the nitrided sample. Only aluminum peaks were detected in both patterns and aluminum nitride could not be confirmed in the pattern of the nitrided sample. Even on the condition that the irradiating distance was so close that the substrate melted down during operation, surface hardening did not occur and aluminum nitride layer could not be formed on the substrate
Fig. 3. Relation between radial distance from the center and nitriding temperature in each experimental condition.
406
Y. Ando et al. / Vacuum 65 (2002) 403–408
irradiating distance was set to 300 mm in order to prevent melt down of the sample during operation. Figs. 4 and 5 show the appearance and the Xray diffraction pattern of the sample nitrided by hydrogen/nitrogen mixture plasma jets on the condition that the irradiating distance was 300 mm. In the case of using pure nitrogen as working gas, nitride layer was not formed on the substrate even on the condition that the substrate was melted down during operation. On the other
hand, many fibers were created on the substrate in the case of hydrogen/nitrogen mixture gas. This result suggested that these supersonic expanding plasma jets had enough chemical reactivity to destruct the aluminum oxide film preventing from nitriding of aluminum. However, aluminum nitride could not be observed in X-ray patterns of the nitrided sample and surface of the substrate came to be rough during operation in this case. The rough surface was formed due to generation of blisters during operation. The blisters appeared at the center of the substrate after the 3 min operation. The area of the blisters was expanded gradually toward the edge, and finally, the aluminum substrate was covered with the blisters. The blisters were thought to be generated by hydrogen absorption of aluminum substrate. Then, for the purpose of confirming whether the cause of blister generation was due to hydrogen absorption by aluminum substrate or not, hydrogen/argon mixture plasma jets irradiation to aluminum substrate was conducted. Consequently, almost the same phenomenon as that in the case of hydrogen/argon mixture plasma occurred and the aluminum substrate had covered with blisters after operation. Fig. 6 shows the appearance of the
Fig. 4. Appearance of the sample nitrided by hydrogen/ nitrogen mixture plasma jets.
Fig. 5. X-ray diffraction pattern of the aluminum sample nitrided by hydrogen/nitrogen mixture plasma jets.
Fig. 6. Appearance of the sample irradiated by Argon/nitrogen mixture plasma jets.
Y. Ando et al. / Vacuum 65 (2002) 403–408
407
hydrogen/argon mixture plasma jets irradiated sample. Thereby, it was thought that hydrogen infiltration into the substrate was dominant in this blister generation process. However, since these blisters were small and crowded in comparison with those in the case of hydrogen/nitrogen plasma jets, blisters generation was affected by chemical reaction between nitrogen and aluminum as well. 3.2. Nitriding of aluminum–magnesium alloy As the aluminum oxide removal method without addition of some reducer gas to plasma jets, introducing additives into aluminum substrate as reducer was thought to be effective. Next, rapid nitriding to obtain aluminum nitride without blister generation was tried by using aluminum– magnesium alloy plate as substrate. Figs. 7 and 8 show the appearance and the X-ray diffraction patterns of the aluminum–magnesium alloy sample nitrided on the same condition as the case of the nitriding of aluminum by pure nitrogen plasma jets. In this case, unlike the case of pure aluminum substrate, nitride layer was formed and surface of
Fig. 7. Appearance of the sample nitrided by hydrogen/ nitrogen mixture plasma jets.
Fig. 8. X-ray diffraction pattern of the aluminum–magnesium alloy sample nitrided by nitrogen plasma jets.
the substrate had become frosted on the substrate on the condition that irradiating distance was 250 mm, besides this phenomenon was observed even in the case of 350 mm in irradiating distance. According to the previous reports on direct nitriding to produce aluminum nitride [7,8], deoxidation of aluminum oxide surface layer by magnesium contented in aluminum occurred over 670 K and aluminum oxide was reduced. Then, chemical reaction between aluminum and nitrogen occurred and aluminum nitride layer was produced. These reactions could be expressed from the thermodynamic viewpoint by using the following equations. 3Mg þ 4Al2 O3 -3MgAl2 O4 þ 2Al
ð1Þ
3Mg þ Al2 O3 -MgO þ 2Al:
ð2Þ
Almost the same reactions might occur in our experiment. In this experiment, unlike the previous results in this study, the existence of aluminum nitride could be confirmed from not only the appearance of the nitrided sample but also the Xray diffraction pattern of the nitrided sample. From these results, it was proved that rigid aluminum oxide could be removed without melting or vaporization of aluminum by introducing reducer into aluminum substrate, besides, this process using supersonic expanding plasma jets was found to have a high potential for rapid direct nitriding process to produce aluminum nitride without melting and vaporization of aluminum.
408
Y. Ando et al. / Vacuum 65 (2002) 403–408
4. Conclusions In this study, the nitridings of aluminum and aluminum–magnesium alloy by using supersonic plasma jets were studied in order to develop a rapid direct nitriding process to produce aluminum nitride without melting and vaporization of aluminum during operation. The conclusions of this study were summarized as follows: (1) Nitride layer could not be formed on the surface of the aluminum substrate on the condition that pure nitrogen plasma jets were used though the nitrogen plasma jets had enough reactivity to form nitride layer on the surface in the case of titanium substrate. (2) By addition of hydrogen to nitrogen plasma jets, the aluminum nitride layer could be formed due to reduction of the rigid aluminum oxide surface layer by hydrogen or nitrogen hydride. However, surface of the substrate came to be rough because the aluminum absorbed hydrogen during operation. (3) By introducing magnesium as reducer into the aluminum substrate, aluminum nitride layer could be formed on the surface of the sample
without melting or vaporization of even on the condition that pure plasma jets were used. (4) From these results, this process was have a high potential for rapid without melting or vaporization of to produce aluminum nitride.
substrate nitrogen found to nitriding substrate
References [1] Komeya K, Mitsutani E, Meguro T. J Ceram Soc J 1993;101:377–82. [2] Yonezawa T, Tahara H, Andoh Y, Onoe K, Yoshikawa. Proceedings of the 20th International Symposium Space Technology and Science 1996:ISTS 96-a-3-05. [3] Andoh Y, Tobe S, Tahara H, Yoshikawa T. Proceedings of the United Thermal Spray Conference 1999. p. 234–9. [4] Ando Y, Tobe S, Tahara H, Yoshikawa T. Proceedings of the 14th International Symposium Plasma Chemistry 1999. p. 1185–90. [5] Tahara H, Uda N, Onoe K, Tsubakishita Y, Yoshikawa T. IEEE Trans Plasma Sci 1994;22:58–64. [6] Tahara H, Komiko K, Yonezawa T, Andoh Y, Yoshikawa T. IEEE Trans Plasma Sci 1996;24:218–25. [7] Kondoh K, Kimura J, Takeda Y. J Jpn Soc Powder Powder Metall 1999;46:801–10. [8] Kotani K, Ikeuchi K, Matsuda F. Quarterly J Jpn Weld Soc 1996;14:551–62.