This content has been downloaded from IOPscience. Please scroll down to see the full text.
Download details:
IP Address: 140.113.38.11
This content was downloaded on 28/04/2014 at 22:45
Please note that terms and conditions apply.
Investigation of the Galvanic Effect between RuN Barriers and Cu Seed Layers
View the table of contents for this issue, or go to the journal homepage for more 2011 Jpn. J. Appl. Phys. 50 121803
(http://iopscience.iop.org/1347-4065/50/12R/121803)
Investigation of the Galvanic Effect between RuN Barriers and Cu Seed Layers
Chia-Yang Wu, Wen-Hsi Lee, Shih-Chieh Chang1, and Ying-Lang Wang1
Department of Electrical Engineering, National Cheng Kung University, Tainan 701, Taiwan, R.O.C. 1Institute of Lighting and Energy Photonics, National Chiao Tung University, Hsinchu 30050, Taiwan, R.O.C. Received April 22, 2011; revised August 26, 2011; accepted September 4, 2011; published online November 24, 2011
In this study, we investigated the galvanic effect between the Cu metals and ruthenium nitride (RuNx) films that were deposited at various nitrogen (N2) gas flow rates in chemical mechanical polishing slurries. It was found that the galvanic corrosion of the RuNx films was inhibited with increasing N2gas flow ratio, whereas the galvanic corrosion of the Cu seed layers was enhanced. Electrochemical impedance spectroscopy showed that the galvanic corrosion resistance of RuNxincreased and that of the ruthenium oxide layer decreased as N2flow ratio increased. This was because the increase in the N content in the RuNx films inhibited the corrosion and oxidation of the Ru metals.
# 2011 The Japan Society of Applied Physics
1. Introduction
In Cu metallization, a barrier layer is necessary to prevent the Cu metal from diffusing into the dielectric layer.1–15) Materials used as barrier layers should not only be thin enough to reduce effective metal resistance, but should also have compatible adhesion between Cu and the dielectric layers.16–24)In addition, good step-coverage is necessary for
barrier deposition to form a uniform layer along the surfaces and corners of silicon wafers. Tantalum and tantalum nitride (Ta/TaNx) films have been widely used as diffusion barriers for Cu metallization owing to their excellent chemical and thermal stabilities, and the physical properties of TaNxfilms deposited at various nitrogen (N2) gas flow rates have been widely investigated.25–28)
With the decreasing feature size of microelectronic devices, the overall thickness of the Cu seed layer and barrier metal has to be reduced to achieve the Cu elec-troplating process. Ruthenium (Ru) has the potential to act as both the seed and barrier layers.29,30) However, the barrier property of pure Ru for Cu diffusion is worse than that of a Ta/TaNx film, and Cu diffusion occurs after annealing at 450C for 10 min in a Cu/Ru/Si structure.31) Therefore,
how to modify the Ru metal so that it can be plated directly is very important for Ru to replace Ta/TaNxas a Cu diffu-sion barrier layer. Several studies have shown that ruthenium nitride (RuNx) barrier layers have better diffusion barrier properties and thermal stability than layers of pure Ru metals.5–8) The N content of the RuNx films also affects their barrier properties and thermal stability. In addition to physical properties, the chemical behavior of the RuNxfilms in the chemical mechanical polishing (CMP) process should be considered to achieve a robust production line. During the CMP process, various defects are usually generated on the Cu lines because Cu metal is soft and easily corroded. Defects generated by chemical corrosion, galvanic corro-sion, and photocorrosion have been examined in previous studies.8–10)The galvanic effect is due to the difference in the electrochemical potential at the interface between the barrier and Cu in a slurry environment. Kondo et al. investigated the galvanic effect of metals such as tungsten (W), titanium, Ta, and their nitrides,8) Ernur et al. studied the effect of slurry composition on galvanic corrosion,9) and Tai et al. investigated the extent of galvanic corrosion between Cu
and different barrier materials.10) Moreover, the degree of galvanic corrosion between Cu and various barrier materials including their nitrides has been widely investigated. For example, Hung et al. reported that the galvanic corrosion of the TaNx films to Cu seed layers is retarded by N.32,33) However, to date there have been few studies about the galvanic effect between the RuNxfilms and Cu seed layers. To address this gap in the literature, in the present study, we investigate the galvanic effect between Cu seed layers and RuNx films deposited at various N2 gas flow rates in CMP slurries by electrochemical impedance spectroscopy (EIS),34–39) energy dispersive X-ray (EDX), and field-emission scanning electron microscopy (FESEM).
2. Experimental Procedure
Blanket wafers were deposited with 200-nm-thick Cu seed layers using the ionized metal plasma (IMP) process on Si(100) substrates. In IMP deposition, inductively coupled Ar/N2plasma was used to ionize sputtered atoms. The RuNx barriers were deposited by direct-current (DC) magnetron sputtering at 10 W forward power from a Ru target (99.99%). The Ar flow rate was kept at 20 sccm, and the nitrogen flow rate was varied from 0 to 20 sccm, resulting in a process pressure of0:1 Torr. The nitrogen flow ratios ½N2=ðAr þ N2Þ were (1) 0, (2) 20, (3) 33, (4) 43, and (5)
50%. The gas purity was 99.9999% and the gas flow rate was controlled within0:1 sccm by mass flow controllers, which guaranteed repeatability. After the RuNx film deposition, a 200-nm-thick Cu seed layer was then deposited. In the corrosion analyses, the RuNx films deposited at various N2 flow ratios were used as the working electrodes (1 cm2), the Cu seed layers were used as the counterelectrodes (1 cm2), and Ag/AgCl was used as the reference electrode in a Cu CMP slurry, which contained aluminum oxide abrasive, surfactant, and hydrogen peroxide (5% H2O2) with a pH in the range of 6–7. In this case, impedance was not measured until open circuit potential (OCP) became stable, which could minimize the inaccuracies in the experiment. The potential difference was measured by electrically connecting the working electrode and the reference electrode to an electrometer, which is a high-impedance multimeter that can measure voltage, resistance, and current. The potential difference between the Cu seed layers and the RuNx films was obtained by subtracting the two potential differences measured using the Cu and RuNxelectrodes with respect to the Ag/AgCl reference electrode. The measurement was not
E-mail address: ylwang@tsmc.com DOI: 10.1143/JJAP.50.121803
conducted by connecting the the two electrodes directly. The N content of the RuNx films was determined using an X-ray photoelectron spectrometer (XPS) and the O ratio of RuNx films after 90 s immersion in the slurry was deter-mined by EDX. The electron binding energy of Ru was measured by XPS and the stoichiometric composition of the RuN films was measured by XPS. In addition, the cross-sectional profile and the thickness of the RuNxand Cu films before and after immersion in the slurry were obtained by FESEM (KLA-Tencor SEMVision). EIS is widely recog-nized as a powerful tool for the investigation of electro-chemical behaviors,34–39)and thus it was used to characterize the galvanic corrosion between the RuNx films and the Cu seed layers. Therefore, the RuN films were used as the working electrode and the Cu film as the counterelectrode. The OCP between the deposited RuNx film (1 cm2) and a platinum plate (1 cm2) was examined using a Princeton Applied Research PARSTAT 2273. The Nyquist plot was also obtained using a PARSTAT 2273 and the alternating current (AC) impedance behavior of the capacitor cells was analyzed. The impedance measurements were carried out at various potentials with a DC potential of 0 mV associated with OCP and a frequency range from 100 mHz to 400 kHz. The equivalent circuit was built up and simulated using software ZSimpWin version 3.1 with EIS data.
3. Results and Discussion
Figures 1(a)–1(c) show the microstructures of RuNx films in which the nitrogen content varied from 0 to 50%. The microstructures of the RuNx films were examined by SEM. It was found that pure Ru has a column structure. In the presence of nitrogen, the grain size decreased, the number of defects increased, and the structure transformed into RuNx. The composition of the RuN films was also measured by XPS. As shown in Table I, the N content of the RuN films increased significantly with N2 flow ratio.
The potential dynamic curves were used to determine the corrosion effect between the Cu seed layers and the RuNx films, as shown in Fig. 2. It can be seen that the corrosion current densities of the Cu seed layers were slightly higher than those of the RuNxfilms, which implies that the intrinsic corrosion rates of the Cu seed layers were higher than those of the RuNxfilms. In addition, Fig. 3 shows that the OCP of the Cu seed layers was more negative than those of the RuNx films, which also indicates that the RuNxfilms were nobler than the Cu seed layers with regard to their activity in the slurry. (Note that each OCP was first compared with that of the Ag/AgCl reference electrode in OCP measurement.) We then obtained the potential difference between the Cu seed layers and RuNx films by subtracting the OCP measured in the previous OCP measurement.
In the electrochemical analysis, negative OCPs indicate that the Cu seed layers were the substrate that was more easily corroded. For RuNx films, the OCP became more positive with increasing nitrogen content, indicating that RuNxfilms with N doping were less corroded. This suggests that increasing the N content in RuNx films suppresses electron release owing to the high electronegativity of N atoms. Figure 4 shows that the binding energy of the Ru peaks (3d3=2: 284.2 eV, 3d5=2: 280.0 eV) was split, and some of these peaks shifted slightly to a higher value. This phenomenon is
Fig. 1. SEM cross-sectional images before slurry dipping of RuNxfilms deposited at N2gas flow ratios of (a) 0, (b) 20, and (c) 50%.
Table I. Atomic ratio of Ru and RuN films. Sputtering nitrogen ratio ½N2=ðN2þ ArÞ (%) Ru atomic ratio (%) N atomic ratio (%) 0 99.6 0.4 20 93.3 6.7 33 91.0 9.0 43 86.5 13.5 50 77.6 22.4 1E-5 -0.5 0.0 0.5 1.0 1.5 E (V vs Ag/AgCl) Log I (A) (1)Cu (2)Ru (3)RuN(20%) (4)RuN(33%) (5)RuN(43%) (6)RuN(50%) (2) (1) (3,4,5,6) 1E-4 1E-3
Fig. 2. Potential dynamic curves of the Cu seed layer and the RuNxfilms deposited at various N2gas flow ratios in the slurries.
C.-Y. Wu et al. Jpn. J. Appl. Phys. 50 (2011) 121803
due to the fact that the electrons in the Ru atoms were in different valence states, inducing the chemical shift. The increase in binding energy indicates that N in the RuNx films suppresses electron release and inhibits oxidation.
The galvanic corrosion rate of RuNxfilms is lower than the self-corrosion rate, indicating that the corrosion rate of RuNx films is decreased by the galvanic effect. In contrast, the galvanic corrosion rate of Cu is larger than the self-corrosion rate. The Bode plots in Fig. 5(a) show the effect of applied frequency on the impedance between the Cu seed layers and the RuNxfilms deposited at N2gas flow ratios from 0 to 50%. In this case, the RuNxfilms deposited at various N2flow rates were used as the working electrodes, the Cu films were used as the counterelectrodes, and all potentials were measured relative to that of the Ag/AgCl reference electrode. In the low-frequency region (0.1–1 Hz), the impedance decreased with increasing N2 content. In contrast, the impedance slightly increased with increasing N2 content in the high-frequency region (>100 kHz), as shown in Fig. 5(b).
Figure 6 shows the Nyquist plots of the Cu–RuNx elec-trochemical system in the CMP slurry. The Nyquist plots in Fig. 6(b) also show thatZimandZre both increased with
increasing N2 content in the high-frequency (20–400 kHz) region, which means that the RuNxfilms doped with a high N content were less corroded. Figure 6(a) shows the Cu– RuNxsystem measured from the middle-frequency region to the low-frequency region. The Nyquist plots of the Cu–RuNx electrochemical system can be fitted with two semicircles using ZSimpWin: a small semicircle in the high-frequency range (>20 Hz) followed by a large semicircle in the middle-low-frequency range (<20 Hz). The second semicircle reveals that both Zim and Zre decrease with increasing N2
content in the low-frequency (0.1–20 Hz) region, indicating that the Ru films oxidize more easily than the RuNx films.
Figure 7 shows the simulated equivalent circuit of the Cu–RuNx electrochemical system. In this circuit, the series and parallel combinations of resistance or capacitance elements indicate whether their representative reactions occur sequentially (in series) or simultaneously (in paral-lel).40)Rs is the bulk-solution resistance,Cdl is the
double-layer capacitance of the electrogenerated from the sur-face corrosion of the RuNx films, and Rcorr is the
corro-sion resistance (associated with the double layer). The chemical reaction of the RuNx films in the Cu–RuNx electrochemical system occurs through the following inter-mediate steps:
Fig. 3. Potential difference between the Cu seed layers and RuNxfilms measured with an electrometer.
Fig. 4. XPS spectra of the as-deposited pure Ru and RuN films.
0.1 0 500 1000 1500 2000 2500 Ru RuN(20%) RuN(33%) RuN(43%) RuN(50%)
Bode Impedance |Z| (Ohmic)
Frequency (Hz) (a) 1 10 100 1000 10000 100000 0.1 0 500 1000 1500 2000 2500 Ru RuN(20%) RuN(33%) RuN(43%) RuN(50%)
Bode Impedance |Z| (Ohmic)
Frequency (Hz) (a) 1 10 100 1000 10000 100000 100 250 300 350 400 450 500 Ru Ru(20%) Ru(33%) Ru(43%) Ru(50%)
Bode Impedance |Z|(Ohmic)
Frequency (Hz) (b)
1000 10000 100000
Fig. 5. (a) Bode plots of the Cu–RuN electrochemical system in the CMP slurry with the RuNxfilms deposited at N2gas flow ratios from 0 to 50%, and (b) magnification of the high-frequency region.
Ru! Ru4þþ 4e ð1Þ Ru4þþ 2O2! RuO2: ð2Þ In eq. (2), Cox is the capacitance of the RuNx oxidization
layer, and Rox is the resistance of the RuNx oxidization
layer. On the other hand, the two semicircles shown in Fig. 5 indicate the transient surface oxidation of the RuNx films. The first semicircle (high-frequency region) indicates the continuous corrosion of the RuNx films, and the second semicircle (middle-low-frequency region) indicates the Ru metal oxides in the peroxide-based solutions.
The values of each component in the circuit of Fig. 7 are obtained by fitting the Nyquist spectra in Fig. 6, and they are summarized in Table II, where CPEdl is the impedance of
the constant phase element (CPE) of the electrogenerated double layer from surface corrosion of the RuNxfilms, and CPEox is the impedance of the RuNx oxidation layer. The
capacitances in the system can be transferred from CPE using41–43) Cdl¼ ðQdl RcorrÞð1=nÞ Rcorr ; ð3Þ Cox¼ðQ ox RoxÞð1=nÞ Rox : ð4Þ
Consequently, Rcorr increases with the increase in N2
content, whileRox decreases, as shown in Table II. This is
because the RuNxfilms doped with high N content are more resistant to oxidization and are less corroded than those doped at lower N content.
Figure 8 shows the oxygen content of the RuNx films, which decreases as N2 flow rate decreases because RuNx film’s oxidation is retarded and the double-layer capacitance of the electrogenerated double layer is restricted. This demonstrates that RuNx films doped at high N content are more resistant to oxidization because the N element holds the electrons strongly. Therefore, increasing the N ratio in the RuNx films decreases the galvanic corrosion rate of
Fig. 7. Proposed equivalent circuit diagram showing the electrochemical characteristics of RuNxgalvanic corrosion in the CMP slurry solution, whereRsis the bulk-solution resistance,Cdlis the double-layer capacitance, Rcorris the corrosion resistance (associated with the double layer),Coxis the oxidization-layer capacitance, andRoxis the oxidization resistance of RuNx. 0 0 100 200 300 400 500 600 700 Ru RuN(20%) RuN(33%) RuN(43%) RuN(56%) Zim Zre (a) 500 1000 1500 2000 2500 0 0 100 200 300 400 500 600 700 800 Ru RuN(20%) RuN(33%) RuN(43%) RuN(56%) Zim Z re (b) 50 100 150 200 250 300 350 400
Fig. 6. (a) Nyquist plots of the Cu–RuN electrochemical system in the CMP slurry with the RuNxfilms deposited at N2gas flow ratios from 0 to 50%, and (b) magnification of impedance close to zero in the high-frequency region.
Table II. Element values of equivalent circuit in Fig. 6 required for the best fitting of impedance spectra in Fig. 5. Rs() Rcorr () CPEdl Rox() CPEox Q n Cdl(pF) Q n Cox(mF) RuN (0%) 21 260 4:00 1012 0.952 1.41 1982 2:01 109 0.750 0.0583 RuN (20%) 17 275 4:50 1012 0.946 1.40 1678 2:00 109 0.720 0.0257 RuN (33%) 25 289 3:80 1012 0.951 1.31 1486 1:88 109 0.710 0.0245 RuN (43%) 22 305 4:22 1012 0.944 1.25 1165 1:71 109 0.725 0.0239 RuN (50%) 21 310 4:15 1012 0.942 1.18 864 1:73 109 0.710 0.0196 C.-Y. Wu et al. Jpn. J. Appl. Phys. 50 (2011) 121803
RuNx and the reduction rate of the Cu seed layers. The extent of galvanic corrosion for the Cu seed layers and the RuNx films thus depends on their potential difference. Furthermore, according to Stern and Geary,44) and Stern45) the following equation shows the mathematical relationship between corrosion resistance and corrosion current for polarization data.
Icorr¼
a b
2:303Rcorrðaþ bÞ
; ð5Þ
whereIcorrandRcorrare the corrosion current and corrosion
resistance, respectively, and a andb are the anodic Tafel
slope and cathodic Tafel slope, respectively. According to our previous experiments, self-corrosion occurs in the TaNx films;31,32) it decreases with increasing N2 flow ratio. The results of the impedance measurements and the potential differences between the Cu seed layers and the RuNx films measured using an electrometer system show that the corrosion current decreases from 1.034 to 0.859 mA when the N2 flow ratio for RuNx deposition increases from 0 to 50%, as shown in Fig. 8. This result is consistent with the results of our previous investigation, which showed that the galvanic corrosion current is inhibited by the N element of TaNx films.31,32) The parameters in impedance spec-troscopy show that a larger Rcorr and smaller Rox indicate
that the RuNx metal was more resistant to corrosion or formation of the RuO2 layer.Cox is related to the formation
of RuO2.
The proposed mechanism for the corrosion of the Cu seed layers and the RuNxfilms in acidic slurry is shown in Fig. 9. For Cu films, the overall extent of corrosion is dominated by the intrinsic corrosion in the slurry. In this study, the intrinsic Cu corrosion rate was about 38.67 nm/min and the galvanic corrosion rate of the Cu films was about 111– 200 nm/min, which depended on the N content of the RuNx films, as shown in Table III. When the N2gas flow ratio was increased for RuNxdeposition, the galvanic corrosion of the RuNxfilms was suppressed owing to the electronegativity of the N element, whereas the Cu galvanic corrosion was enhanced because the Cu metal received fewer electrons from the RuNxfilm.
4. Conclusions
The purpose of this study is to investigate the galvanic effect between the Cu metals and RuNx films. The N content of RuNx films influenced not only the physical film properties, but also the chemical activity in the CMP process. To obtain a robust Cu metallization process, the N content of the RuNx barriers should be optimized to achieve a compromise between the barrier properties and chemical stability. The mechanism of galvanic corrosion in the Cu–RuNx electro-chemical system was investigated, and a computer simula-tion using EIS data produced an equivalent circuit of the surface reaction with regard to the relationship of the RuNx films with the Cu seed layers. The corrosion resistance increased with the N2 flow ratio for RuNx deposition, but the opposite results were found for the ruthenium oxide resistance associated with ruthenium oxide capacitance. This demonstrates that the N element effectively inhibits the oxidation of the Ru metals and decreases the galvanic corrosion rate of the RuNx films. In this study, we thus conclude that increasing the N content in the RuNx films suppresses the electron release, because the N element has high electronegativity, similarly to fluorine and oxygen. The chemical shift of the RuNxelectron shows that the N content makes RuNxmore resistant to oxidation.
Acknowledgments
This work was supported by the National Science Council of Taiwan (Grant No. NSC 99-2622-E-006-034-CC3). The authors thank National Cheng Kung University, Tainan, 0 0 10 20 30 40 50 60 70 80 90 100
N2 flow ratio (N 2/N2+Ar)
O Atomic % 0.84 0.86 0.88 0.90 0.92 0.94 0.96 0.98 1.00 1.02 1.04 Corrosion current ( mA/cm 2 ) 10 20 30 40 50
Fig. 8. Relationship for RuNxfilms deposited at various N2gas flow rates between EDX of O element ratio after a 90 s dip and corrosion current in Cu–RuNxsystem.
Fig. 9. Proposed mechanism of the galvanic effect between the Cu seed and the RuNxfilms.
Table III. Summary of intrinsic and galvanic corrosion rates of Cu films and RuNxfilms deposited at various N2gas flow ratios. Units are in nm/min. N2flow ratio (%) RuNx self-corrosion rate (nm/min) RuNxgalvanic corrosion rate (nm/min) Cu self-corrosion rate (nm/min) Cu galvanic corrosion rate (nm/min) 0 0.672 0.092 38.67 111.79 20 0.636 0.088 — 162.36 33 0.627 0.087 — 175.67 43 0.602 0.084 — 182.32 50 0.566 0.082 — 202.29
Taiwan, for technical support. National Cheng Kung University also assisted in meeting the publication costs of this article.
1) D. Edelstein, J. Heidenreich, R. Goldblatt, W. Cote, C. Uzoh, N. Lustig, P. Roper, T. McDevitt, W. Motsiff, A. Simon, J. Dukovic, R. Wachnik, H. Rathore, R. Schulz, L. Su, S. Luce, and J. Slattery:IEDM Tech. Dig., 1997,
p. 773.
2) P. C. Andricacos, C. Uzoh, J. O. Dukovic, J. Horkans, and H. Deligianni:
IBM J. Res. Dev.42 (1998) 567.
3) K. W. Chen, Y. L. Wang, L. Chang, S. C. Chang, F. Y. Li, and S. H. Lin:
Electrochem. Solid-State Lett.7 (2004) G238.
4) A. Beverina, H. Bernard, J. Palleau, J. Torres, and F. Tardif:Electrochem.
Solid-State Lett.3 (2000) 156.
5) A. Jindal and S. V. Babu:J. Electrochem. Soc.151 (2004) G709.
6) Y. Homma, S. Kondo, N. Sakuma, K. Hinode, J. Noguchi, N. Ohashi, H. Yamaguchi, and N. Owada:J. Electrochem. Soc.147 (2000) 1193.
7) S. Tamilmani, W. Huang, and S. Raghavan:J. Electrochem. Soc.153
(2006) F53.
8) S. Kondo, N. Sakuma, Y. Homma, and N. Ohashi:Jpn. J. Appl. Phys.39
(2000) 6216.
9) D. Ernur, S. Kondo, D. Shamiryan, and K. Maex:Microelectron. Eng.64
(2002) 117.
10) K. Tai, H. Ohtorii, S. Takahashi, N. Komai, H. Horikoshi, S. Sato, Y. Ohoka, Y. Segawa, M. Ishihara, Z. Yasuda, and T. Nogami:IEEE 2002 Int.
Interconnect Technology Conf., 2002, p. 194.
11) S. C. Chang, J. M. Shieh, J. Y. Fang, Y. L. Wang, B. T. Dai, and M. S. Feng:J. Vac. Sci. Technol. B22 (2004) 116.
12) W. T. Tseng, C. W. Liu, B. T. Dai, and C. F. Yeh:Thin Solid Films290
(1996) 458.
13) T. Hara, F. Togoh, T. Kurosu, K. Sakamoto, Y. Shioya, T. Ishimaru, and T. K. Doy:Electrochem. Solid-State Lett.4 (2001) G65.
14) B. Du and I. I. Suni:Electrochem. Solid-State Lett.8 (2005) 283.
15) S. C. Kurity, S. Seal, W. Fei, J. Ramsdell, V. H. Desai, Y. Li, S. V. Babu, and B. Wood:J. Electrochem. Soc.150 (2003) C36.
16) Y. L. Chin, B. S. Chiou, and W. F. Wu:Jpn. J. Appl. Phys.41 (2002) 3057.
17) W. F. Wu, C. C. Wu, K. L. Ou, and C. P. Chou:J. Electrochem. Soc.150
(2003) G83.
18) C. C. Chang, J. S. Chen, and W. S. Hsu:J. Electrochem. Soc.151 (2004)
746.
19) A. A. Istratov and E. R. Weber:J. Electrochem. Soc.149 (2002) G21.
20) K. M. Latt, Y. K. Lee, S. Li, T. Osipowicz, and H. L. Seng:Mater. Sci.
Eng. B84 (2001) 217.
21) E. Kolawa, J. S. Chen, J. S. Reid, P. J. Pokela, and M. A. Nicolet:J. Appl.
Phys.70 (1991) 1369.
22) W. F. Wu, K. L. Ou, C. P. Chou, and C. C. Wu:J. Electrochem. Soc.150
(2003) G83.
23) M. Stavrev, C. Wenzel, A. Moller, and K. Drescher:Appl. Surf. Sci.91
(1995) 257.
24) S. M. Rossnagel and H. Kim:J. Vac. Sci. Technol. B21 (2003) 2550.
25) J. C. Lin and C. Lee:J. Electrochem. Soc.146 (1999) 3466.
26) A. E. Kaloyeros, X. Chen, T. Stark, S. C. Seo, G. G. Peterson, H. L. Frisch, B. Arkles, and J. Sullivan:J. Electrochem. Soc.146 (1999) 170.
27) N. D. Cuong, D.-J. Kim, B.-D. Kang, C. S. Kim, K.-M. Yu, and S.-G. Yoon:J. Electrochem. Soc.153 (2006) G164.
28) S. C. Chang, Y. L. Wang, and T. C. Wang:IEEE Trans. Semicond. Manuf.
17 (2004) 384.
29) T. P. Moffat, M. Walker, P. J. Chen, J. E. Bonevich, W. F. Egelhoff, L. Richter, C. Witt, T. Aaltonen, M. Ritala, M. Leskela¨, and D. Josella:
J. Electrochem. Soc.153 (2006) C37.
30) O. Chyan, T. N. Arunagiri, and T. Ponnuswamy:J. Electrochem. Soc.150
(2003) C347.
31) R. Chan, T. N. Arunagiri, Y. Zhang, O. Chyan, R. M. Wallace, M. J. Kim, and T. Q. Hurdc:Electrochem. Solid-State Lett.7 (2004) G154.
32) C. C. Hung, Y. S. Wang, W. H. Lee, S. C. Chang, and Y. L. Wang:
Electrochem. Solid-State Lett.10 (2007) H127.
33) C. C. Hung, Y. S. Wang, W. H. Lee, S. C. Chang, and Y. L. Wang:
Electrochem. Solid-State Lett.10 (2007) D100.
34) J. E. Garland, C. M. Pettit, and D. Roy:Electrochim. Acta49 (2004) 2623.
35) M. J. Walters, J. E. Garland, C. M. Pettit, D. S. Zimmerman, D. R. Marr, and D. Roy:J. Electroanal. Chem.499 (2001) 48.
36) C. Gabrielli: in Physical Electrochemistry: Principles, Methods, and Applications, ed. I. Rubinstein (Dekker, New York, 1995).
37) D. D. Macdonald:Electrochim. Acta51 (2006) 1376.
38) V. R. K. Gorantla, K. A. Assiongbon, S. V. Babu, and D. Roy:
J. Electrochem. Soc.152 (2005) G404.
39) V. R. K. Gorantla, S. B. Emery, S. Pandija, S. V. Babu, and D. Roy:Mater.
Lett.59 (2005) 690.
40) V. R. K. Gorantla, K. A. Assiongbon, S. V. Babu, and D. Roy:
J. Electrochem. Soc.152 (2005) G404.
41) L. Larabi, O. Benali, S. M. Mekelleche, and Y. Harek:Appl. Surf. Sci.253
(2006) 1371.
42) E. Barsoukov and J. R. Macdonald: Handbook of Impedance Spectroscopy (Wiley, Hoboken, NJ, 2005) p. 37.
43) M. E. Orazem, P. Shukla, and M. A. Membrino:Electrochim. Acta47
(2002) 2027.
44) M. Stern and A. L. Geary:J. Electrochem. Soc.104 (1957) 56.
45) M. Stern: Corrosion14 (1958) 61.
C.-Y. Wu et al. Jpn. J. Appl. Phys. 50 (2011) 121803
