Susceptibility to Hydrogen Absorption and Hydrogen Thermal Desorption of Titanium Alloys Immersed in Neutral Fluoride Solution under Applied Potential

Transcription

1 Materials Transactions, Vol. 49, No. 7 (28) pp to 1666 #28 The Japan Institute of Metals Susceptibility to Hydrogen Absorption and Hydrogen Thermal Desorption of Titanium Alloys Immersed in Neutral Fluoride Solution under Applied Potential Ken ichi Yokoyama 1; *, Katsutoshi Takashima 2 and Jun ichi Sakai 2;3 1 Department of Materials Science and Engineering, Kyushu Institute of Technology, Kitakyushu , Japan 2 Department of Materials Science and Engineering, Waseda University, Tokyo , Japan 3 Faculty of Science and Engineering, Waseda University, Tokyo , Japan The susceptibility to hydrogen absorption and hydrogen thermal desorption of titanium alloys has been investigated in a neutral 2.% NaF solution at 37 C under various applied cathodic potentials. Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys absorb hydrogen under less noble potentials than 1:5 V, 1:9 V and 1:4 V versus a saturated calomel electrode, respectively. The amounts of absorbed hydrogen of Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys under an applied potential of 2: V for 24 h are approximately 12, 1 and 11 mass ppm, respectively. Upon immersion in the 2.% NaF solution under an applied potential, the hydrogen thermal desorption behavior of Ti-.2Pd alloy differs from those immersed in acid fluoride solutions, whereas the hydrogen thermal desorption behaviors of Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys are similar to those immersed in acid fluoride solutions. The present results indicate that in a neutral fluoride solution, the effects of an applied potential on the hydrogen absorption and desorption behaviors of Ti-6Al-4V alloy are considerably smaller than those of commercial pure titanium reported previously, Ti-.2Pd and Ti-11.3Mo-6.6Zr-4.3Sn alloys. [doi:1.232/matertrans.mra2883] (Received March 4, 28; Accepted April 15, 28; Published June 4, 28) Keywords: titanium, corrosion, hydrogen embrittlement, fluoride, thermal desorption analysis 1. Introduction Hydrogen embrittlement of biomedical and dental devices made of titanium and its alloys is a problem in the presence of fluoride. 1 4) We have recently investigated environmental conditions for the hydrogen absorption of titanium alloys in fluoride solutions. 2 1) In acid fluoride solutions such as prophylactic agents, titanium and its alloys absorb substantial amounts of hydrogen, thereby causing pronounced degradation of the mechanical properties or fracture of titanium and its alloys. 2 8) In neutral fluoride solutions, the hydrogen absorption of titanium and its alloys is rarely confirmed, 2 4) although commercial pure titanium 3) and Ti-.2Pd alloy 8) fracture under a sustained tensile-loading test. For long-term immersion (above 1 h) in a neutral 2.% NaF solution, commercial pure titanium absorbs hydrogen; however, the hydrogen absorption of Ti-6Al-4V and Ti-11.3Mo-6.6Zr- 4.3Sn alloys has not yet been observed despite undergoing general corrosion. 9) In oral environments, the electrochemical potentials of titanium and its alloys sometimes shift to a less noble direction. For commercial pure titanium, the applied cathodic potential markedly enhances hydrogen absorption in neutral fluoride 1) and sodium chloride solutions. 11,12) When commercial pure titanium is immersed in a neutral 2.% NaF solution, hydrogen absorption occurs for a short time under less noble potential than 1:2 V versus a saturated calomel electrode (vs. SCE). 1) Under an applied potential of 2: V for 24 h, the amount of absorbed hydrogen is as high as 25 mass ppm. 1) For titanium alloys, however, the applied potential does not necessarily lead to hydrogen absorption. For instance, Ti-6Al-4V alloy does not absorb hydrogen in a neutral Na 2 SO 4 solution under an applied potential. 13) In *Corresponding author, yokken@post.matsc.kyutech.ac.jp neutral fluoride solutions, there are no reports on the effects of an applied potential on the hydrogen absorption behavior of titanium alloys. It is therefore necessary to confirm experimentally whether an applied potential induces hydrogen absorption in titanium alloys immersed in neutral fluoride solutions. The hydrogen thermal desorption behavior reflects the hydrogen state in various materials including hydrides, diffusive hydrogen, hydrogen trapped in defects and hydrogen in a solution; 14 23) hence, it becomes a clue for elucidating the mechanisms of hydrogen embrittlement. We have demonstrated some characteristics of the hydrogen desorption behavior of titanium and its alloys immersed in fluoride solutions. 2 8,1) The hydrogen desorption behavior of commercial pure titanium is sometimes changed by the hydrogen absorbing conditions. 1) However, fundamental data on the hydrogen desorption behavior of titanium and its alloys are still limited. The characteristics of the hydrogen desorption behavior of titanium alloys immersed in fluoride solutions under an applied potential should first be clarified. The purpose of the present study is to investigate the effects of an applied potential on the hydrogen absorption and thermal desorption behaviors of titanium alloys in a neutral fluoride solution by hydrogen thermal desorption analysis (TDA). In this article, we compare the hydrogen absorption and desorption behaviors of titanium alloys with those of commercial pure titanium reported previously. 1) 2. Experimental Procedures A.5-mm wire of an alpha titanium alloy (Ti-.2Pd; described previously 7,8) ) or an alpha-beta titanium alloy (Ti- 6Al-4V; described previously 4) ) and a.45-mm wire of a beta titanium alloy (Ti-11.3Mo-6.6Zr-4.3Sn; described previously 2,6) ) were cut into 5-mm-long specimens. The nominal

2 1662 K. Yokoyama, K. Takashima and J. Sakai Table 1 Chemical composition of tested titanium alloys (mass%). Ti C H O N Fe Pd Al V Mo Zr Sn Ti-.2Pd balance Ti-6Al-4V balance Ti-11.3Mo-6.6Zr-4.3Sn chemical composition of the wires is shown in Table 1. The specimens were carefully polished with a 6-grit SiC paper and ultrasonically washed in acetone for 5 min. The present experimental cell conditions were the same as those of our previous studies. 1) Cathodic polarization curves were measured (n ¼ 3) in 15 ml of a neutral 2.% NaF solution with a ph of 6.5 under aerated conditions at 37 1 C. The counter and reference electrodes used were a platinum electrode and a saturated calomel electrode (SCE), respectively. A potential of 2: V for 1 min was initially applied, and then scanned to a noble direction at a rate of 2 mv/min. An immersion test under a constant applied potential was performed in the same cell. The amount of desorbed hydrogen was measured by TDA for specimens subjected to the immersion test (n ¼ 1 to 3). The immersed specimens were cut at both ends and subjected to ultrasonic cleaning with acetone for 2 min. Subsequently, the specimens were dried in ambient air and then subjected to TDA, which was started 3 min after the removal of the specimens from the test solutions. A quadrupole mass spectrometer (ULVAC, Kanagawa, Japan) was used for hydrogen detection. Sampling was conducted at 3-s intervals at a heating rate of 1 C/h. The amount of desorbed hydrogen was defined as the integrated peak intensity. Titanium hydride on the surface of the specimens was examined using an X-ray diffractometer with Co K radiation of wavelength ¼ : nm in the 2 angle range from 1 to 9 operated at 4 kv and 2 ma. 3. Experimental Results The representative cathodic polarization curves of the tested titanium alloys in the neutral 2.% NaF solution under aerated conditions are shown in Fig. 1. The corrosion potentials of Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo-6.6Zr- 4.3Sn alloys obtained from the polarization curves were :3 V, :5 V and :4 V, respectively. The polarization curves of Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys indicate that a hydrogen evolution reaction in the 2.% NaF solution dominated between 1: V and 2: V. It appears that the hydrogen evolution reaction is mainly caused by a direct reduction of water, although there is also a reduction of hydrogen ions. In these potential ranges, the polarization behavior was almost the same irrespective of the titanium alloy types. At more noble potential ranges, a reduction of dissolved oxygen dominated, although the hydrogen reduction reaction very slightly occurred. The amounts of desorbed hydrogen of the tested titanium alloys immersed in the 2.% NaF solution for 24 h are shown as functions of the applied potential, as shown in Fig. 2. In the present study, the amount of hydrogen absorbed during the immersion test was calculated by subtracting the Potential, E / V vs. SCE Ti-6Al-4V.1 Ti-.2Pd Ti-11.3Mo-6.6Zr-4.3Sn Current density, i / ma cm Fig. 1 Cathodic polarization curves of tested titanium alloys in 2.% NaF solution. Desorbed hydrogen, C / mass ppm Ti-.2Pd Ti-6Al-4V Ti-11.3Mo-6.6Zr-4.3Sn Applied potential, E / V vs. SCE Fig. 2 Amounts of desorbed hydrogen obtained by TDA of tested titanium alloys immersed in 2.% NaF solution as functions of applied potential.

3 Susceptibility to Hydrogen Absorption and Hydrogen Thermal Desorption of Titanium Alloys Immersed (a) 1 (b) 3 (c) Hydrogen desorption rate, R / mass ppm min V -1.7 V -1.6 V -1.8 V -1.5 V -1.4 V -1.3 V -1.2 V nonimmersed Hydrogen desorption rate, R / mass ppm min V -1.9 V -1.5 V nonimmersed -1.6 V -1.8 V Hydrogen desorption rate, R / mass ppm min V -1.8 V -1.5 V -1.4 V -1.3 V nonimmersed -1.2 V Temperature, T / C Temperature, T / C Temperature, T / C Fig. 3 Hydrogen thermal desorption curves of (a) Ti-.2Pd, (b) Ti-6Al-4V and (c) Ti-11.3Mo-6.6Zr-4.3Sn alloys immersed in 2.% NaF solution under various applied potentials. The numerical values in this figure indicate the applied potential. amount of hydrogen desorbed of a nonimmersed specimen, that is, the predissolved hydrogen content from the amount of desorbed hydrogen. The amounts of desorbed hydrogen of the nonimmersed Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo- 6.6Zr-4.3Sn alloys were 34, 96 and 14 mass ppm, respectively. The increments in the amount of desorbed hydrogen of Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys were observed below the applied potentials of 1:5 V, 1:9 V and 1:4 V, respectively. The amount of desorbed hydrogen, that is, absorbed hydrogen, increased with decreasing applied potential. Under an applied potential of 2: V, the amounts of absorbed hydrogen of Ti-.2Pd, Ti-6Al-4V and Ti- 11.3Mo-6.6Zr-4.3Sn alloys were approximately 12, 1 and 11 mass ppm, respectively. The amount of absorbed hydrogen of Ti-.2Pd alloy, compared with those of Ti-6Al- 4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys, varied widely. The hydrogen thermal desorption curves of the tested titanium alloys immersed in the 2.% NaF solution for 24 h under various applied potentials are shown in Figs. 3(a) to (c). In these figures, the hydrogen desorption curve of the nonimmersed specimen is also included. For Ti-.2Pd alloy, hydrogen desorption was sometimes observed widely ranging from 3 to 7 C. In contrast, for Ti-6Al-4V and Ti- 11.3Mo-6.6Zr-4.3Sn alloys, the temperature range of the hydrogen desorption was comparatively narrow. The intensity of the hydrogen desorption peak increased with decreasing applied potential. The hydrogen desorption peak of the tested titanium alloys often shifted at higher temperatures under the applied potentials without hydrogen absorption, despite the result that the amount of desorbed hydrogen did not increase compared with those of the nonimmersed specimens. The typical X-ray diffraction (XRD) pattern from the surface of Ti-.2Pd alloy immersed in the 2.% NaF solution under an applied potential of 2: V is shown in Fig. 4. Titanium hydride, TiH 2 (tetragonal; a ¼ :312 nm, c ¼ Intensity (arbitrary units) 1 2 :418 nm), was identified. For Ti-6Al-4V and Ti-11.3Mo- 6.6Zr-4.3Sn alloys, hydrides or corrosion products were not detected in the present immersion conditions. 4. Discussion Co Kα 2 Theta (degree) alpha Ti TiH 2 Fig. 4 Typical XRD patterns of surface of Ti-.2Pd alloy immersed in 2.% NaF solution under applied potential of 2: V. One important finding in the present study is that the applied cathodic potential causes Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys to absorb a substantial amount of hydrogen in the neutral 2.% NaF solution. In our previous study, 1) for commercial pure titanium, the critical potential for hydrogen absorption is approximately 1:2 V in a 2.% NaF solution for 24 h. As shown in Fig. 2, the critical potentials for the hydrogen absorption of Ti-.2Pd, Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys were approximately 1:5 V, 1:9 V and 1:4 V, respectively. It is likely that the susceptibility to the critical potential for hydrogen absorption in the 2.% NaF solution is in this 8 9

4 1664 K. Yokoyama, K. Takashima and J. Sakai order: commercial pure titanium 1) > Ti-11.3Mo-6.6Zr- 4.3Sn alloy > Ti-.2Pd alloy > Ti-6Al-4V alloy. However, upon immersion in the 2.% NaF solution for a long period (more than 1 h), the hydrogen absorption of commercial pure titanium occurs under a more noble potential such as a corrosion potential of :6 V. 9) Thus, the critical potentials for the hydrogen absorption of the titanium alloys presently tested are possibly confined to the case of a short immersion time. It is considered that there is a close connection between the electrochemical behavior and the hydrogen absorption behavior. As shown in Fig. 1, the cathodic polarization behavior in the 2.% NaF solution of the titanium alloys presently tested is similar to that of commercial pure titanium reported previously. 1) For the same applied potential, it is assumed that the amount of hydrogen evolution is the same irrespective of the titanium alloy types, because the amount of hydrogen evolution generally depends on current densities. However, the amount of absorbed hydrogen greatly varied with the titanium alloy types tested even for the same applied potential (the same current density). For example, under an applied potential of 2: V, the order of hydrogen absorption according to the amount by the titanium alloys is as follows: commercial pure titanium (25 mass ppm) 1) > Ti-.2Pd alloy (12 mass ppm), Ti-11.3Mo-6.6Zr-4.3Sn alloy (11 mass ppm) > Ti-6Al-4V alloy (1 mass ppm). Accordingly, the hydrogen absorption behavior of the titanium alloys immersed in the 2.% NaF solution does not necessarily correspond to the electrochemical behavior, although a large amount of hydrogen evolution readily leads to hydrogen absorption. The electrochemical behavior and corrosion resistance of titanium and its alloys in fluoride solutions have been investigated by several researchers ) In fluoride solutions, the corrosion resistance of Ti-6Al-4V alloy is similar to or lower than that of commercial pure titanium, 25,28,32) whereas the corrosion resistance of Ti-Pd alloys is higher than that of commercial pure titanium. 28,29,33) However, in the 2.% NaF solution, the effects of an applied cathodic potential on the hydrogen absorption behavior of Ti-6Al-4V alloy were considerably smaller than those of commercial pure titanium, 1) Ti-.2Pd and Ti-11.3Mo-6.6Zr-4.3Sn alloys, as shown in Fig. 2. This result indicates that corrosion resistance does not necessarily dominate the hydrogen absorption behavior. Furthermore, the results are consistent with that the hydrogen absorption resistance of Ti-6Al-4V alloy is higher than that of commercial pure titanium immersed in acid and neutral fluoride solutions without an applied potential. 4) In Na 2 SO 4 solutions, 13) the hydrogeninduced environment-assisted cracking behavior of Ti-6Al- 4V alloy can be explained from the thermodynamically stable chemical species in the potential-ph diagram (Pourbaix diagram 34) ) of the titanium/water system, but is not explained by those of the aluminum/water and vanadium/water systems. In fluoride solutions, however, the hydrogen absorption behavior of titanium and its alloys is not always explained by the potential-ph diagram of the titanium/water system, as indicated in our previous studies. 2 1) With regard to the hydrogen absorption behavior of titanium and its alloys in fluoride solutions, there may be a specific cause. These findings also indicate that the hydrogen absorption behavior cannot be estimated only from the electrochemical behavior. In addition to the electrochemical behavior, the surface conditions of the titanium alloys including their surface oxide films should be considered. The surface of titanium and its alloys is covered mainly with TiO 2 thin films. For Ti-6Al-4V alloy, surface oxide films contain small amounts of Al 2 O 3 and V 2 O 5. 35,36) Our previous study 4) suggested that the corrosion product of Na 3 AlF 6 deposited on the surface of Ti-6Al-4V alloy often acts as a barrier to hydrogen absorption in an acid fluoride solution, whereas such effects of the corrosion product of Na 3 TiF 6 are not observed. The corrosion product of Na 3 AlF 6 is also detected by XRD analysis on the surface of Ti-6Al-4V alloy immersed in the neutral 2.% NaF solution without an applied potential for a long-term immersion, 9) although it was not detected in the present study. Aluminum in the surface oxide films of Ti-6Al-4V alloy might be associated with the high resistance to hydrogen absorption in neutral fluoride solutions. The effects of aluminum on the hydrogen absorption behavior of titanium and its alloys should also be examined. In addition to the surface conditions, Tal-Gutelmacher and Eliezer 37) reported that the amount of absorbed hydrogen of Ti-6Al-4V alloy depends on its microstructure. The effects of microstructure, particularly the distribution forms of the alpha and beta phases, on the hydrogen absorption behavior probably play important roles. When the amount of absorbed hydrogen exceeds a few hundred mass ppm, the degradation of the mechanical properties, i.e., hydrogen embrittlement, of titanium alloys generally occurs. 38 4) In the present study, the amount of absorbed hydrogen is the mean of the absorbed hydrogen of the entire specimen. Upon absorbing hydrogen at around room temperature, most of the absorbed hydrogen exists as hydrides in the surface layer of the specimen for commercial pure titanium and alpha titanium alloys, 41 45) whereas hydrogen diffuses toward the center of the specimen for alpha-beta or beta titanium alloys. 2,4,45) From our previous study, 45) for commercial pure titanium of.5-mm-diameter specimen, the local hydrogen concentration in the surface layer of the specimen is often one order of magnitude larger than the mean amount of absorbed hydrogen. Even if the mean amount of absorbed hydrogen is small, hydrogen embrittlement possibly occurs. In addition, the amount of hydrogen is not necessarily the sole criterion for hydrogen embrittlement; the hydrogen state in materials must also be considered ) Consequently, it is likely that titanium alloys absorb sufficient hydrogen resulting in hydrogen embrittlement in the 2.% NaF solution under an applied potential. In usual oral environments, nonetheless, extreme potential shift such as in the present experimental conditions may rarely occur. It seems that at least Ti-6Al-4V alloy does not degrade in relation to hydrogen absorption in neutral oral environments containing fluoride, because the critical potential for the hydrogen absorption of the alloy is as low as 1:9 V. For Ti-.2Pd alloy, the origin of the scatter of the amount of absorbed hydrogen is unknown. In an acid fluoride solution without an applied potential, 7) the amount of

5 Susceptibility to Hydrogen Absorption and Hydrogen Thermal Desorption of Titanium Alloys Immersed 1665 absorbed hydrogen of Ti-.2Pd alloy is also scattered. It seems that the amount of absorbed hydrogen of Ti-.2Pd alloy is scattered readily. The hydrogen thermal desorption behavior depends strongly on the hydrogen absorbing conditions for Ni-Ti superelastic alloy ) The hydrogen desorption temperature and peak profiles are changed sensitively by different types of immersing solution such as fluoride solutions, 47,49) methanol solutions 48) and ethanol solutions. 5,51) The hydrogen desorption behavior of commercial pure titanium immersed in NaF solutions under an applied potential is not always consistent with that immersed in acid fluoride solutions without an applied potential. 1) Similarly, in the present study, the hydrogen desorption behavior of Ti-.2Pd alloy immersed in the 2.% NaF solution under an applied potential differed from those immersed in acid fluoride solutions without an applied potential, 7) despite the same amount of desorbed hydrogen. In the case of immersion in the 2.% NaF solution under an applied potential, the hydrogen desorption peak was often wide, and hydrogen desorption from room temperature to 3 C was not observed. Our previous study 45) indicated that hydrogen desorbed at low temperatures corresponds to the dissociation of hydride, whereas hydrogen desorbed at high temperatures is associated with hydrogen that diffuses into the specimen after hydride dissociation. Hydride formation was confirmed in the present study, as shown in Fig. 4. The possible reason for the variation of the hydrogen desorption behavior is that the distribution or morphology of hydride in the surface layer of the specimen is often changed by hydrogen absorbing conditions. 52) On the other hand, the hydrogen desorption behaviors of Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys immersed in the 2.% NaF solution under an applied potential are similar to those immersed in acid fluoride solutions reported previously. 2,4,6) Eliezer et al. 53) demonstrated that the hydrogen desorption of Ti-6Al-4V alloy is observed as a single peak from 3 C for gas-phase charging, whereas it is observed as a single peak from 5 C for electrochemical charging. Moreover, the hydrogen desorption behavior of a solution-annealed -21S titanium alloy (Ti-15Mo-3Al-3Nb-.3Si; single beta-titanium phase) is due to the difference between gas-phase charging and electrochemical charging. 54) To change the hydrogen desorption behavior of these alloys, a marked change in the hydrogen charging condition may be needed. As shown in Fig. 3, the hydrogen desorption peak often shifted at higher temperatures, although the amount of desorbed hydrogen did not increase compared with those of nonimmersed specimens. The possible reason for this is that surface conditions including the oxide films of the specimens are changed by immersion in the 2.% NaF solution. The relationships between surface conditions and hydrogen desorption behavior have been described in our previous article. 45) 5. Conclusions We have demonstrated that an applied cathodic potential causes titanium alloys to absorb substantial amounts of hydrogen in a neutral fluoride solution. The effects of an applied potential on the critical potential for hydrogen absorption and the amount of absorbed hydrogen of Ti- 6Al-4V alloy immersed in a neutral fluoride solution are considerably smaller than those of commercial pure titanium, Ti-.2Pd and Ti-11.3Mo-6.6Zr-4.3Sn alloys. It is likely that the hydrogen thermal desorption behavior of Ti-.2Pd alloy depends on hydrogen absorbing conditions, whereas those of Ti-6Al-4V and Ti-11.3Mo-6.6Zr-4.3Sn alloys are hardly affected by hydrogen absorbing conditions. The present results imply that the hydrogen absorption behavior of titanium alloys immersed in fluoride solutions is related to the surface conditions including surface oxide films rather than an electrochemical factor. Acknowledgements This study was supported in part by a Grant-in-Aid for Young Scientists (B) ( ) and a Grant-in-Aid for Scientific Research (C) (185672) from the Ministry of Education, Culture, Sports, Science and Technology, Japan. REFERENCES 1) M. H. O. Könönen, E. T. Lavonius and J. K. Kivilahti: Dent. Mater. 11 (1995) ) K. Kaneko, K. Yokoyama, K. Moriyama, K. Asaoka, J. Sakai and M. Nagumo: Biomaterials 24 (23) ) K. Yokoyama, K. Kaneko, Y. Miyamoto, K. Asaoka, J. Sakai and M. Nagumo: J. Biomed. Mater. Res. A 68A (24) ) K. Yokoyama, T. Ogawa, K. Asaoka and J. Sakai: Corros. Sci. 47 (25) ) K. Yokoyama, T. Ogawa, K. Asaoka and J. Sakai: Mater. Sci. Eng. A 384 (24) ) T. Ogawa, K. Yokoyama, K. Asaoka and J. Sakai: Biomaterials 25 (24) ) K. Yokoyama, T. Ogawa, K. Asaoka and J. Sakai: J. Alloy. Compd. 4 (25) ) K. Yokoyama, T. Ogawa, K. Asaoka and J. Sakai: Mater. Sci. Eng. A 419 (26) ) K. Yokoyama, T. Ogawa, K. Asaoka and J. Sakai: J. Biomed. Mater. Res. B 78B (26) ) K. Takashima, K. Yokoyama, K. Asaoka and J. Sakai: J. Alloy. Compd. 431 (27) ) T. Fukuzuka, K. Shimogori, H. Satoh and F. Kamikubo: Desalination 31 (1979) ) T. Nosetani, M. Shimono, Z. Tanabe, S. Sato, H. Hirose, K. Onda and Y. Yamaguchi: Sumitomo Light Met. Tech. Rep. 15 (1974) ) T. Haruna, M. Hamasaki and T. Shibata: Mater. Trans. 1 (25) ) W. Y. Choo and J. Y. Lee: Metall. Trans. A 13A (1982) ) K. Ono and M. Meshii: Acta Metall. Mater. 4 (1992) ) H. Chatbi, M. Vergnat and G. Marchal: Appl. Phys. Lett. 64 (1994) ) A. Takasaki, Y. Furuya, K. Ojima and Y. Taneda: J. Alloy. Compd. 224 (1995) ) A. Turnbull, R. B. Hutchings and D. H. Ferriss: Mater. Sci. Eng. A 238 (1997) ) S. W. Smith and J. R. Scully: Metall. Mater. Trans. A 31A (2) ) M. Nagumo, M. Nakamura and K. Takai: Metall. Mater. Trans. A 32A (21) ) K. Takai and R. Watanuki: ISIJ Int. 43 (23) ) T. Nishiue, Y. Kaneno, H. Inoue and T. Takasugi: J. Alloy. Compd. 364 (24) ) F. G. Wei and K. Tsuzaki: Metall. Mater. Trans. A 37A (26)

6 1666 K. Yokoyama, K. Takashima and J. Sakai 24) H. Mimura and Y. Miyagawa: Jpn. J. Dent. Mater. Dev. 15 (1996) ) Y. Oda, E. Kawada, M. Yoshinari, K. Hasegawa and T. Okabe: Jpn. J. Dent. Mater. Dev. 15 (1996) ) L. Reclaru and J.-M. Meyer: Biomaterials 19 (1998) ) M. Nakagawa, S. Matsuya, T. Shiraishi and M. Ohta: J. Dent. Res. 78 (1999) ) M. Nakagawa, S. Matsuya and K. Udoh: Dent. Mater. J. 2 (21) ) M. Nakagawa, S. Matsuya and K. Udoh: Dent. Mater. J. 21 (22) ) H. H. Huang: Biomaterials 23 (22) ) H. H. Huang: Electorchim. Acta 47 (22) ) N. Schiff, B. Grosgogeat, M. Lissac and F. Dalard: Biomaterials 23 (22) ) M. Nakagawa, Y. Matono, S. Matsuya, K. Udoh and K. Ishikawa: Biomaterials 26 (25) ) M. Pourbaix: Atlas of Electrochemical Equilibria in Aqueous Solutions, (NACE International, TX, 1966). 35) M. Ask, J. Lausmaa and B. Kasemo: Appl. Surf. Sci. 35 (1989) ) R. N. S. Sodhi, A. Weninger and J. E. Davies: J. Vac. Sci. Technol. A 9 (1991) ) E. Tal-Gutelmacher and D. Eliezer: J. Alloy. Compd (25) ) G. A. Lenning, C. M. Craighead and R. I. Jaffee: Trans. AIME 2 (1954) ) C. R. McKinsey, M. Stern and R. A. Perkins: Trans. ASM 5 (1958) ) D. N. Williams: J. Inst. Metals. 91 (1962) ) I. I. Phillips, P. Poole and L. L. Shreir: Corros. Sci. 12 (1972) ) I. I. Phillips, P. Poole and L. L. Shreir: Corros. Sci. 14 (1974) ) Z. F. Wang, C. L. Briant and K. S. Kumar: Corrosion 54 (1998) ) Z. F. Wang, C. L. Briant and K. S. Kumar: Corrosion 55 (1999) ) T. Ogawa, K. Yokoyama, K. Asaoka and J. Sakai: J. Alloy. Compd. 396 (25) ) K. Yokoyama, S. Watabe, K. Hamada, J. Sakai, K. Asaoka and M. Nagumo: Mater. Sci. Eng. A 341 (23) ) K. Yokoyama, K. Kaneko, K. Moriyama, K. Asaoka, J. Sakai and M. Nagumo: J. Biomed. Mater. Res. A 65A (23) ) K. Yokoyama, T. Ogawa, K. Asaoka, J. Sakai and M. Nagumo: Mater. Sci. Eng. A 36 (23) ) K. Yokoyama, K. Kaneko, K. Moriyama, K. Asaoka, J. Sakai and M. Nagumo: J. Biomed. Mater. Res. A 69A (24) ) T. Ogawa, K. Yokoyama, K. Asaoka and J. Sakai: Mater. Sci. Eng. A 393 (25) ) T. Ogawa, K. Yokoyama, K. Asaoka and J. Sakai: Mater. Sci. Eng. A 422 (26) ) Z. A. Foroulis: Boshoku Gijutsu 29 (198) ) D. Eliezer, E. Tal-Gutelmacher, C. E. Cross and Th. Boellinghaus: Mater. Sci. Eng. A 433 (26) ) D. Eliezer, E. Tal-Gutelmacher, C. E. Cross and Th. Boellinghaus: Mater. Sci. Eng. A 421 (26) 2 27.