- Email: [email protected]
Cathodic arc deposition with activated anode (CADAA) for preparation of in situ doped thin solid films
Vacuum 65 (2002) 433–438
Cathodic arc deposition with activated anode (CADAA) for preparation of in situ doped thin solid films Hirofumi Takikawa*, Keisaku Kimura, Ryuichi Miyano, Tateki Sakakibara Department of Electrical and Electronic Engineering, Toyohashi University of Technology, Tempaku-cho, Toyohashi, Aichi 441-8580, Japan
Abstract Cathodic arc deposition with an activated anode was developed for preparing doped thin solid films. The activated anode was a water-cooled crucible, and the material in it was evaporated and ionized by passing a partial arc current through the crucible. As an example, aluminum-doped zinc oxide (ZnO:Al) was synthesized by a Zn cathodic arc in an oxygen (O2) gas flow at 1.0 Pa. Al powder was used as a dopant precursor and placed in the crucible. The anodic plume plasma appears on the crucible anode, which is composed of cathode material of Zn and anode material of Al as well as a reactive gas of O2. Energy dispersive X-ray analysis revealed that the prepared-film contained Zn, Al and O. The ZnO:Al film on the glass substrate was transparent with a very strong X-ray diffraction peak of ZnO. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Cathodic arc deposition; Activated anode; ZnO film; Al doping; Film properties
1. Introduction Transparent conductive films have been widely used as transparent electrodes in many industrial optoelectronic devices such as photovoltaic cells, liquid-crystal displays, touch-screen displays, and window heaters [1,2]. These films can be prepared using pure metal oxides such as tin oxide (SnO2), indium oxide (In2O3), zinc oxide (ZnO), and cadmium stannates (Cd2SnO4, CdSnO3). However, in order to achieve the low resistivity and high stability, doped oxides, e.g., tin-doped indium oxide (In2O3:Sn, ITO; indium tin oxide), aluminum-doped zinc oxide (ZnO:Al), *Corresponding author. Tel.: +81-532-44-6727; fax: +81532-44-6727. E-mail address: [email protected] (H. Takikawa).
antimony-doped tin oxide (SnO2:Sb) and cadmium-doped tin oxide (SnO2:Cd), etc. are usually employed. Transparent films have been prepared by a variety of deposition techniques such as spray pyrolysis, chemical vapor deposition (CVD), evaporations, and sputtering. Cathodic vacuum arc deposition is a new way to prepare such films. Ben-Shalom et al. have prepared SnO2 by filtered vacuum arc deposition [2], and the present authors have prepared ZnO by shielded and non-shielded vacuum arc depositions [3–5]. However, there is no report with regard to in situ doping technique simultaneous with film preparation using cathodic arc deposition. In the present work, cathodic arc deposition with an activated anode (CADAA) is developed for the preparation of multi-element thin solid films such as ZnO:Al.
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 5 3 - 5
434
H. Takikawa et al. / Vacuum 65 (2002) 433–438
2. Apparatus and experiments In general, the cathode of a vacuum arc is very active and evaporates ionized materials by the cathode spot, whereas the anode is usually inert and merely works as the acceptor of the evaporated materials emitted from the cathode spot. However, the anode can be active in some cases, e.g., for higher arc current [6], for higher pressure [7], or when using a tiny electrode [8]. When the anode becomes active, the anodic arc plasma appears. To date, this plasma has been tested for preparing the thin films made of anode material [9–13]. In the present study, in order to prepare multielement film or doped film, a new system was developed which can simultaneously evaporate the cathode and anode materials. Cathodic arc deposition with an activated anode is depicted in Fig. 1. The Zn cathode (64 mm in diameter) was placed in the vacuum chamber (stainless steel; 200 mm in diameter and 300 mm in length). In order to steer the cathode spot, a permanent magnet was placed behind the cathode, and a protection plate was placed around the cathode, thus protecting the plasma expansion to the rear and also guiding the
magnetic field. The chamber was the main anode, which was grounded and electrically connected to the cathode through an external power supply. The main arc current Im flowed between the cathode and the main anode. The substrate was located 200 mm away from the cathode. A crucible hearth liner (molybdenum (Mo) cup; 20 mm in top diameter, 16 mm in bottom diameter, 10 mm in height) mounted in the pocket of a water-cooled copper (Cu) hearth was placed between the cathode and the substrate as shown in Fig. 1. The crucible was also electrically connected to the cathode through another external power supply. Thus, the crucible worked as another anode (subanode) by passing the current Ic between the cathode and the crucible. The outer part of the crucible hearth was insulated by ceramics and polytetrafluoroethylene (PTFE). The arc was ignited by a mechanical triggering unit. No macrodroplet-filtering system, in the present study, was employed since the number of droplets is relatively few and they would be small in a reactive cathodic arc with an O2 gas flow [4,14]. The experimental conditions for preparing ZnO:Al were as follows: in-crucible material, Al powder (150 mesh; average diameter, 105 mm); total arc current (=cathode current), 30 A; O2 gas flow rate, 25 ml/min; back pressure, 0.01 Pa; working pressure, 1.0 Pa. The films were prepared on a nonbiased substrate at room temperature. The deposition time was approximately 5 min, and the film thickness was 300–450 nm. The films were analyzed using an energy dispersive X-ray (EDX) in a scanning electron microscope (JEOL, JSM6300) for composition analysis, an X-ray diffractometer with a Cu Ka radiation (Rigaku, RINT2500) for crystalline structure, a double-beam spectrophotometer (Hitachi, 330) for transmittance, and a homemade four-point probe conductivity tester for electrical resistivity.
3. Results and discussions Fig. 1. Schematic illustration of cathodic arc deposition with activated anode (CADAA) apparatus. Cathode and crucible sub-anode is located in the chamber (main anode). Crucible sub-anode is heated by passing partial arc current.
3.1. Plasma characteristics A photograph of CADAA plasma is shown in Fig. 2. The main and crucible currents, Im and Ic ;
H. Takikawa et al. / Vacuum 65 (2002) 433–438
435
Fig. 2. Photograph of CADAA plasma, showing bright anodic plume plasma as well as cathode spot jet plasma on left side.
Fig. 4. Spectra emitted from anodic plume plasma formed above the crucible anode (Ic ¼ 5 A) (a) and spectra emitted from same position without anodic plume plasma (Ic ¼ 0 A) (b).
Fig. 3. Main arc voltages Vm between cathode and main anode (chamber) and crucible voltage Vc between cathode and crucible (activated anode).
were 25 and 5 A, respectively. At the left side of the picture, the cathodic plasma jet is seen. Above the crucible, bright anodic plume plasma is clearly seen, which is very similar to the single footpoint that appeared on the chamber anode surface [6]. The anodic plume plasma was not observed for Ic ¼ 0 A. Fig. 3 shows the main arc voltage Vm between the cathode and main anode (vacuum
chamber), and the crucible voltage between the cathode and the crucible anode. They are absolute values. The main arc voltage was from 10 to 15 V. On the other hand, the crucible voltage Vc was comparably higher, being from 40 to 60 V. At a lower Ic of o5 A, the crucible voltage was fluctuated considerably since the anodic plume plasma was formed intermittently and was unstable. The spectra radiating from the anodic plume plasma were measured using a monochromator in the visual range. A result for Ic ¼ 5 A is shown in Fig. 4(a). For comparison, the spectra radiating from the same position without the anodic plume plasma (Ic ¼ 0 A) are also shown in Fig. 4(b). For
H. Takikawa et al. / Vacuum 65 (2002) 433–438
436
Fig. 5. Substrate temperature during film preparation.
Fig. 6. Deposition rate as a function of crucible current.
those without anodic plume, only the spectra of Zn were identified. On the other hand, for Ic ¼ 5 A, spectra of Zn, Al, O2 and O+ 2 were identified. The results revealed that the crucible worked as an activated anode. 3.2. Film properties The substrate temperature, which was measured by thermocouple during film preparation, increased with the preparation time, as shown in Fig. 5. The rising rate of the substrate temperature was higher when the crucible current was higher. This indicates that the anodic plume plasma has a substrate-heating function. The deposition rate of the film as a function of the crucible current is shown in Fig. 6. At a lower crucible current, the deposition rate was lower than that for the case without anode activation (Ic ¼ 0 A). This is considered to be due to arc instability at a lower crucible current. As the crucible current increased, so did the deposition rate. This is because the cathodic arc plasma, namely the ions emitted from the cathode, was conducted toward the substrate as a result of establishing the current route with dense plasma between the cathode and the crucible anode. Composition of the film measured using EDX is shown in Fig. 7. It is found that as the crucible current increased, the atomic composition of Al also increased. At Ic ¼ 7 A, Al composition was approximately 4 at%.
Fig. 7. Film composition analyzed with an EDX.
X-ray diffraction patterns of ZnO and ZnO:Al films prepared on borosilicate glass substrates are shown in Fig. 8. A very strong diffraction peak of ZnO (0 0 2) was detected, indicating that the films were highly oriented to the c-axis normal to the substrate. The peak of ZnO (0 0 2) of ZnO:Al was stronger than that of ZnO (Ic ¼ 0 A). No diffraction peaks of Al or Al2O3 were detected. Transmittance spectra of the films are shown in Fig. 9. It was found that the ZnO:Al film was more transparent than the ZnO film in the visual region. For Ic ¼ 0; 5, and 7 A, absorption appeared in the infrared region, which indicates higher carrier density.
H. Takikawa et al. / Vacuum 65 (2002) 433–438
437
Fig. 10. Resistivity of ZnO and ZnO:Al films as a function of crucible current.
4. Conclusions Fig. 8. X-ray diffraction patterns of ZnO and ZnO:Al films prepared on glass substrate.
Fig. 9. Transmittance of ZnO and ZnO:Al films prepared on glass substrate, measured using substrate transmittance as a background reference.
Resistivity of the film is shown in Fig. 10 as a function of the crucible current. The lowest resistivity of approximately 2 103 O cm was observed for the film ZnO (Ic ¼ 0 A). ZnO:Al film had a higher resistivity. However, as the amount of Al dope was increased, the resistivity of ZnO:Al film tended to decrease. Optimization of the doping quantity is required.
A novel system of cathodic arc deposition with an activated anode (CADAA) was developed for manufacturing multi-element thin solid films or doped films. The method consists of simple mechanism of flowing the partial arc current to the crucible sub-anode. In the present article, an aspect of anodic plume plasma generated above the crucible anode was presented. We then show that the in-crucible material is evaporated, and that the cathodic and in-crucible materials and ambient gas are excited in the plasma. As an example, ZnO:Al film was fabricated on a nonbiased substrate at room temperature. The ZnO:Al film then proved to have superior properties on crystal orientation, transmittance and electrical conduction than the nondoped ZnO film. Evaporation from the crucible can easily be controlled by regulating the crucible current, implying that the composition or doping amount in the film can be controlled. Another feature of this novel method is the possibility that mixed multi-elements can be used as in-crucible material. The method is not limited to the preparation of doped ZnO film but is potentially applicable to synthesize a variety of compound thin films with multi-elements or dopants. In addition, many variations in design configuration are possible based on the use of both cathode and anode vapors of the arc plasma, including the
438
H. Takikawa et al. / Vacuum 65 (2002) 433–438
employment of a filtering system of cathodic macrodroplets. Acknowledgements This work was partly supported by a Grant-inAid for Scientific Research (B) and a Grant-in-Aid for Encouragement of Young Scientists from The Japan Society for the Promotion of Science, and Itoh Optical Industrial Co., Ltd. References [1] Hartnagel HL, Dawar AL, Jain AK, Jagadish C. Semiconducting transparent thin films. Bristol and Philadelphia: Institute of Physics Pub, 1995. [2] Ben-Shalom A, Kaplan L, Boxman RL, Goldsmith S, Nathan M. Thin Solid Films 1993;236:20.
[3] Takikawa H, Kimura K, Miyano R, Sakakibara T. Trans Mater Res Soc Jpn 2000;25:345. [4] Takikawa H, Kimura K, Miyano R, Sakakibara T. Thin Solid Films 2000;377–378:74. [5] Miyano R, Kimura K, Izumi K, Takikawa H, Sakakibara T. Vacuum 2000;59:159. [6] Kimblin CW. J Appl Phys 1969;40:1744. [7] Takikawa H, Fujishima T, Sakakibara T. Trans IEE Jpn 1994;114-A:123 [in Japanese]. [8] Grissom JT, Newton JC. J Appl Phys 1974;45:2885. [9] Dorodnov AM, Kuznetsov AN, Petrosov VA. Sov Tech Phys Lett 1979;5:418. [10] Ehrich H. J Vac Sci Technol A 1988;6:134. [11] Ehrich H, Hasse B, Mausbach M, Muller . KG. IEEE Trans Plasma Sci 1990;18:895. [12] Mausbach M, Ehrich H, Muller . KG. J Vac Sci Technol B. 1993;11:1909. [13] Benstetter G. IEEE Trans Plasma Sci 1996;24:1389. [14] Takikawa H, Matsui T, Sakakibara T, Bendavid A, Martin PJ. Trans IEE Jpn 1999;119-A:1243 [in Japanese].
Cathodic arc deposition with activated anode (CADAA) for preparation of in situ doped thin solid films Hirofumi Takikawa*, Keisaku Kimura, Ryuichi Miyano, Tateki Sakakibara Department of Electrical and Electronic Engineering, Toyohashi University of Technology, Tempaku-cho, Toyohashi, Aichi 441-8580, Japan
Abstract Cathodic arc deposition with an activated anode was developed for preparing doped thin solid films. The activated anode was a water-cooled crucible, and the material in it was evaporated and ionized by passing a partial arc current through the crucible. As an example, aluminum-doped zinc oxide (ZnO:Al) was synthesized by a Zn cathodic arc in an oxygen (O2) gas flow at 1.0 Pa. Al powder was used as a dopant precursor and placed in the crucible. The anodic plume plasma appears on the crucible anode, which is composed of cathode material of Zn and anode material of Al as well as a reactive gas of O2. Energy dispersive X-ray analysis revealed that the prepared-film contained Zn, Al and O. The ZnO:Al film on the glass substrate was transparent with a very strong X-ray diffraction peak of ZnO. r 2002 Elsevier Science Ltd. All rights reserved. Keywords: Cathodic arc deposition; Activated anode; ZnO film; Al doping; Film properties
1. Introduction Transparent conductive films have been widely used as transparent electrodes in many industrial optoelectronic devices such as photovoltaic cells, liquid-crystal displays, touch-screen displays, and window heaters [1,2]. These films can be prepared using pure metal oxides such as tin oxide (SnO2), indium oxide (In2O3), zinc oxide (ZnO), and cadmium stannates (Cd2SnO4, CdSnO3). However, in order to achieve the low resistivity and high stability, doped oxides, e.g., tin-doped indium oxide (In2O3:Sn, ITO; indium tin oxide), aluminum-doped zinc oxide (ZnO:Al), *Corresponding author. Tel.: +81-532-44-6727; fax: +81532-44-6727. E-mail address: [email protected] (H. Takikawa).
antimony-doped tin oxide (SnO2:Sb) and cadmium-doped tin oxide (SnO2:Cd), etc. are usually employed. Transparent films have been prepared by a variety of deposition techniques such as spray pyrolysis, chemical vapor deposition (CVD), evaporations, and sputtering. Cathodic vacuum arc deposition is a new way to prepare such films. Ben-Shalom et al. have prepared SnO2 by filtered vacuum arc deposition [2], and the present authors have prepared ZnO by shielded and non-shielded vacuum arc depositions [3–5]. However, there is no report with regard to in situ doping technique simultaneous with film preparation using cathodic arc deposition. In the present work, cathodic arc deposition with an activated anode (CADAA) is developed for the preparation of multi-element thin solid films such as ZnO:Al.
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 5 3 - 5
434
H. Takikawa et al. / Vacuum 65 (2002) 433–438
2. Apparatus and experiments In general, the cathode of a vacuum arc is very active and evaporates ionized materials by the cathode spot, whereas the anode is usually inert and merely works as the acceptor of the evaporated materials emitted from the cathode spot. However, the anode can be active in some cases, e.g., for higher arc current [6], for higher pressure [7], or when using a tiny electrode [8]. When the anode becomes active, the anodic arc plasma appears. To date, this plasma has been tested for preparing the thin films made of anode material [9–13]. In the present study, in order to prepare multielement film or doped film, a new system was developed which can simultaneously evaporate the cathode and anode materials. Cathodic arc deposition with an activated anode is depicted in Fig. 1. The Zn cathode (64 mm in diameter) was placed in the vacuum chamber (stainless steel; 200 mm in diameter and 300 mm in length). In order to steer the cathode spot, a permanent magnet was placed behind the cathode, and a protection plate was placed around the cathode, thus protecting the plasma expansion to the rear and also guiding the
magnetic field. The chamber was the main anode, which was grounded and electrically connected to the cathode through an external power supply. The main arc current Im flowed between the cathode and the main anode. The substrate was located 200 mm away from the cathode. A crucible hearth liner (molybdenum (Mo) cup; 20 mm in top diameter, 16 mm in bottom diameter, 10 mm in height) mounted in the pocket of a water-cooled copper (Cu) hearth was placed between the cathode and the substrate as shown in Fig. 1. The crucible was also electrically connected to the cathode through another external power supply. Thus, the crucible worked as another anode (subanode) by passing the current Ic between the cathode and the crucible. The outer part of the crucible hearth was insulated by ceramics and polytetrafluoroethylene (PTFE). The arc was ignited by a mechanical triggering unit. No macrodroplet-filtering system, in the present study, was employed since the number of droplets is relatively few and they would be small in a reactive cathodic arc with an O2 gas flow [4,14]. The experimental conditions for preparing ZnO:Al were as follows: in-crucible material, Al powder (150 mesh; average diameter, 105 mm); total arc current (=cathode current), 30 A; O2 gas flow rate, 25 ml/min; back pressure, 0.01 Pa; working pressure, 1.0 Pa. The films were prepared on a nonbiased substrate at room temperature. The deposition time was approximately 5 min, and the film thickness was 300–450 nm. The films were analyzed using an energy dispersive X-ray (EDX) in a scanning electron microscope (JEOL, JSM6300) for composition analysis, an X-ray diffractometer with a Cu Ka radiation (Rigaku, RINT2500) for crystalline structure, a double-beam spectrophotometer (Hitachi, 330) for transmittance, and a homemade four-point probe conductivity tester for electrical resistivity.
3. Results and discussions Fig. 1. Schematic illustration of cathodic arc deposition with activated anode (CADAA) apparatus. Cathode and crucible sub-anode is located in the chamber (main anode). Crucible sub-anode is heated by passing partial arc current.
3.1. Plasma characteristics A photograph of CADAA plasma is shown in Fig. 2. The main and crucible currents, Im and Ic ;
H. Takikawa et al. / Vacuum 65 (2002) 433–438
435
Fig. 2. Photograph of CADAA plasma, showing bright anodic plume plasma as well as cathode spot jet plasma on left side.
Fig. 4. Spectra emitted from anodic plume plasma formed above the crucible anode (Ic ¼ 5 A) (a) and spectra emitted from same position without anodic plume plasma (Ic ¼ 0 A) (b).
Fig. 3. Main arc voltages Vm between cathode and main anode (chamber) and crucible voltage Vc between cathode and crucible (activated anode).
were 25 and 5 A, respectively. At the left side of the picture, the cathodic plasma jet is seen. Above the crucible, bright anodic plume plasma is clearly seen, which is very similar to the single footpoint that appeared on the chamber anode surface [6]. The anodic plume plasma was not observed for Ic ¼ 0 A. Fig. 3 shows the main arc voltage Vm between the cathode and main anode (vacuum
chamber), and the crucible voltage between the cathode and the crucible anode. They are absolute values. The main arc voltage was from 10 to 15 V. On the other hand, the crucible voltage Vc was comparably higher, being from 40 to 60 V. At a lower Ic of o5 A, the crucible voltage was fluctuated considerably since the anodic plume plasma was formed intermittently and was unstable. The spectra radiating from the anodic plume plasma were measured using a monochromator in the visual range. A result for Ic ¼ 5 A is shown in Fig. 4(a). For comparison, the spectra radiating from the same position without the anodic plume plasma (Ic ¼ 0 A) are also shown in Fig. 4(b). For
H. Takikawa et al. / Vacuum 65 (2002) 433–438
436
Fig. 5. Substrate temperature during film preparation.
Fig. 6. Deposition rate as a function of crucible current.
those without anodic plume, only the spectra of Zn were identified. On the other hand, for Ic ¼ 5 A, spectra of Zn, Al, O2 and O+ 2 were identified. The results revealed that the crucible worked as an activated anode. 3.2. Film properties The substrate temperature, which was measured by thermocouple during film preparation, increased with the preparation time, as shown in Fig. 5. The rising rate of the substrate temperature was higher when the crucible current was higher. This indicates that the anodic plume plasma has a substrate-heating function. The deposition rate of the film as a function of the crucible current is shown in Fig. 6. At a lower crucible current, the deposition rate was lower than that for the case without anode activation (Ic ¼ 0 A). This is considered to be due to arc instability at a lower crucible current. As the crucible current increased, so did the deposition rate. This is because the cathodic arc plasma, namely the ions emitted from the cathode, was conducted toward the substrate as a result of establishing the current route with dense plasma between the cathode and the crucible anode. Composition of the film measured using EDX is shown in Fig. 7. It is found that as the crucible current increased, the atomic composition of Al also increased. At Ic ¼ 7 A, Al composition was approximately 4 at%.
Fig. 7. Film composition analyzed with an EDX.
X-ray diffraction patterns of ZnO and ZnO:Al films prepared on borosilicate glass substrates are shown in Fig. 8. A very strong diffraction peak of ZnO (0 0 2) was detected, indicating that the films were highly oriented to the c-axis normal to the substrate. The peak of ZnO (0 0 2) of ZnO:Al was stronger than that of ZnO (Ic ¼ 0 A). No diffraction peaks of Al or Al2O3 were detected. Transmittance spectra of the films are shown in Fig. 9. It was found that the ZnO:Al film was more transparent than the ZnO film in the visual region. For Ic ¼ 0; 5, and 7 A, absorption appeared in the infrared region, which indicates higher carrier density.
H. Takikawa et al. / Vacuum 65 (2002) 433–438
437
Fig. 10. Resistivity of ZnO and ZnO:Al films as a function of crucible current.
4. Conclusions Fig. 8. X-ray diffraction patterns of ZnO and ZnO:Al films prepared on glass substrate.
Fig. 9. Transmittance of ZnO and ZnO:Al films prepared on glass substrate, measured using substrate transmittance as a background reference.
Resistivity of the film is shown in Fig. 10 as a function of the crucible current. The lowest resistivity of approximately 2 103 O cm was observed for the film ZnO (Ic ¼ 0 A). ZnO:Al film had a higher resistivity. However, as the amount of Al dope was increased, the resistivity of ZnO:Al film tended to decrease. Optimization of the doping quantity is required.
A novel system of cathodic arc deposition with an activated anode (CADAA) was developed for manufacturing multi-element thin solid films or doped films. The method consists of simple mechanism of flowing the partial arc current to the crucible sub-anode. In the present article, an aspect of anodic plume plasma generated above the crucible anode was presented. We then show that the in-crucible material is evaporated, and that the cathodic and in-crucible materials and ambient gas are excited in the plasma. As an example, ZnO:Al film was fabricated on a nonbiased substrate at room temperature. The ZnO:Al film then proved to have superior properties on crystal orientation, transmittance and electrical conduction than the nondoped ZnO film. Evaporation from the crucible can easily be controlled by regulating the crucible current, implying that the composition or doping amount in the film can be controlled. Another feature of this novel method is the possibility that mixed multi-elements can be used as in-crucible material. The method is not limited to the preparation of doped ZnO film but is potentially applicable to synthesize a variety of compound thin films with multi-elements or dopants. In addition, many variations in design configuration are possible based on the use of both cathode and anode vapors of the arc plasma, including the
438
H. Takikawa et al. / Vacuum 65 (2002) 433–438
employment of a filtering system of cathodic macrodroplets. Acknowledgements This work was partly supported by a Grant-inAid for Scientific Research (B) and a Grant-in-Aid for Encouragement of Young Scientists from The Japan Society for the Promotion of Science, and Itoh Optical Industrial Co., Ltd. References [1] Hartnagel HL, Dawar AL, Jain AK, Jagadish C. Semiconducting transparent thin films. Bristol and Philadelphia: Institute of Physics Pub, 1995. [2] Ben-Shalom A, Kaplan L, Boxman RL, Goldsmith S, Nathan M. Thin Solid Films 1993;236:20.
[3] Takikawa H, Kimura K, Miyano R, Sakakibara T. Trans Mater Res Soc Jpn 2000;25:345. [4] Takikawa H, Kimura K, Miyano R, Sakakibara T. Thin Solid Films 2000;377–378:74. [5] Miyano R, Kimura K, Izumi K, Takikawa H, Sakakibara T. Vacuum 2000;59:159. [6] Kimblin CW. J Appl Phys 1969;40:1744. [7] Takikawa H, Fujishima T, Sakakibara T. Trans IEE Jpn 1994;114-A:123 [in Japanese]. [8] Grissom JT, Newton JC. J Appl Phys 1974;45:2885. [9] Dorodnov AM, Kuznetsov AN, Petrosov VA. Sov Tech Phys Lett 1979;5:418. [10] Ehrich H. J Vac Sci Technol A 1988;6:134. [11] Ehrich H, Hasse B, Mausbach M, Muller . KG. IEEE Trans Plasma Sci 1990;18:895. [12] Mausbach M, Ehrich H, Muller . KG. J Vac Sci Technol B. 1993;11:1909. [13] Benstetter G. IEEE Trans Plasma Sci 1996;24:1389. [14] Takikawa H, Matsui T, Sakakibara T, Bendavid A, Martin PJ. Trans IEE Jpn 1999;119-A:1243 [in Japanese].