Surface Modification of Titanium Utilizing a Repassivation Reaction ...

Materials Transactions, Vol. 43, No. 12 (2002) pp. 3005 to 3009 Special Issue on Biomaterials and Bioengineering c
2002 The Japan Institute of Metals

Surface Modification of Titanium Utilizing a Repassivation Reaction in Aqueous Solutions Takao Hanawa1 , Sachiko Hiromoto1 , Katsuhiko Asami2 , Hidemi Ukai3 and Koichi Murakami3 1

Biomaterials Center, National Institute for Materials Science, Tsukuba 305-0047, Japan Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan 3 Ishikawajima-Harima Heavy Industries Co., Ltd., Yokohama 235-8501, Japan 2

The surface modification of titanium utilizing repassivation reaction was attempted. The original surface oxide film on titanium plates were mechanically removed under immersion in water and Hanks’ solutions with pH 5.0 and 7.4 followed by repassivation or regeneration of surface oxide film in the same solutions. Then the surface-modified titanium plates were immersed in Hanks’ solution (pH 7.4) up to 604.8 ks. Their surface was analyzed with scanning electron microscopy with energy dispesive X-ray spectroscopy, X-ray photoelectron spectroscopy, and X-ray diffractometry, and the mass gain was measured using a microbalance, to confirm the performance of the modification. It was found that more calcium phosphate was precipitated on specimens modified in Hanks’ solutions in comparison with that in water. Calcium and phosphate ions contained in the surface oxides modified in Hanks’ solutions had effect for the titanium surface to adsorb more calcium and phosphate ions than specimens repassivated in water. In addition, more calcium phosphate was precipitated on titanium modified in pH 5.0 2− solution than that in pH 7.4 solution because more H2 PO− 4 and/or HPO4 ions exist in the surface oxide regenerated in pH 5.0 Hanks’ solution than in pH 7.4 Hanks’ solution. It is suggested that a solution with proper pH and ion concentration for the modification should give higher efficiency for the modification of titanium surface. (Received April 8, 2002; Accepted June 12, 2002) Keywords: titanium, surface modification, calcium ion, phosphate ion, repassivation, apatite

1. Introduction Various surface treatments of titanium have been attempted to improve its bone conductivity. The most popular process is plasma-spraying of calcium phosphate ceramics, e.g. hydroxyapatite, on titanium. Also, pulsed laser-deposition technique is applied to the coating of apatite on titanium.1, 2) On the other hand, apatite coating on titanium is possible in situ in aqueous solutions containing calcium and phosphate ions, where electrochemical techniques3, 4) are applied to the coating. The specific surface property of titanium that forming calcium phosphate on itself in aqueous solutions5–11) makes it easy to precipitate calcium phosphate. The first stage of the formation of calcium phosphate is the adsorption of phosphate ions. The passivation and microdissolution of titanium in biological systems were discussed and the significance of adsorption of phosphate ions by the passive film was pointed out.12, 13) The corrosion resistance of titanium is attributed to its surface oxide which is rapidly formed in water and works as a passive film. The film is readily regenerated even if destroyed.14) The surface film consists of non-stoichiometric TiO2 and is amorphous or of low crystallinity.15) Therefore, titanium has two noteworthy characteristic properties; calcium phosphate formation and fast repassivation in aqueous solutions. In this regard, interesting results are obtained.16) When the surface oxide film is mechanically removed in Hanks’ solution, phosphate ions are preferentially taken up in a regenerated surface film during regeneration and components of the regenerated film becomes titanium oxide and titanium oxyhydroxide containing titanium phosphate.16) Calcium ions and also phosphate ions are adsorbed by the film after regeneration and calcium phosphate or calcium titanium phosphate is

eventually formed at the outermost surface. Ions constituting Hanks’ solution other than calcium and phosphate were absent from the surface oxide. This calcium phosphate formation during and after repassivation process is available for surface modification of titanium. In this study, we attempted to modify titanium surface just by simple abrasion and immersion in Hanks’ solutions without organic species, and evaluated the surface-modified titanium with various techniques. 2. Experimental Methods A flowchart of experimental process is shown in Fig. 1. 2.1 Surface modification of titanium Commercially pure titanium plate was used as the starting material. The composition of the material is listed in Table 1. The plates were metallographically polished with SiC paper Polished and washed titanium plate Surface modification

, Abraded in water, pH 5.0, and 7.4 Hanks solutions

Weighed with microbalance

, Immersed in pH 7.4 Hanks solution Evaluation of the performance XPS

Dried in desiccator

Weighed with microbalance

Heated

XRD

Observation with SEM/EDS

Fig. 1 Flowchart of experimental process.

3006

T. Hanawa, S. Hiromoto, K. Asami, H. Ukai and K. Murakami Table 1 Chemical composition of commercial pure titanium. Composition (mass%)

Fe

Mn

Si

Cr

Ni

Cu

0.001

0.001

0.004

0.002

0.006

0.0025

Sn

C

N

O

H

Ti

0.0017

0.0075

0.003

0.035

0.0016

balance

Table 2

Na 0.0022

Ion concentrations of Hanks solution without organic species. Ion

Concentration/kmol m−3

Na+

1.42 × 10−1

K+

5.81 × 10−3

Mg2+

8.11 × 10−4

Ca2+

1.26 × 10−3

Cl−

1.45 × 10−1

HPO2− 4

7.78 × 10−4

SO2− 4 CO2− 3

8.11 × 10−4 4.17 × 10−3

and finished with #1200 SiC paper. The plates were ultrasonically rinsed in acetone for 900 s and followed by rinsing in ethyl alcohol for 900 s and drying with a stream of high purity nitrogen gas (> 99.999%). The titanium plates thus prepared were abraded and immersed in deionized water (pH 6.9) and Hanks’ solutions without organic species and with ion concentrations similar to extracellular fluid. The ion concentrations of the Hanks’ solutions are summarized in Table 2. As for Hanks’ solutions, two types solutions were prepared. The pH of one solution was 7.4 just after preparation, and the other one was adjusted to pH 5.0, by addition of 0.1 kmol m−3 hydrochloric acid. The specimens were abraded with #800 SiC attached to a polytetrafluoroethylene rod in water and pH 7.4 Hanks’ solutions at 310 K and kept in the same solutions for 300 s in the same solutions. 2.2 Microbalance After being weighed with a microbalance and their dimensions being measured, the surface-modified titanium specimens were immersed in pH 7.4 Hanks’ solution in a sealed polytetrafluoroethylene vessels at 310 K up to 2.59 Ms (30 d). The solution was changed to fresh one every two days. As some of the present authors have shown in previous reports,5–8) no precipitate was formed in the solution and on the wall of bottle. After immersion, the specimens were rinsed in deionized water, dried and kept in a desiccator until measurements. For mass gain measurements, specimens were dried in the desiccator for 172.8 ks. The mass gain was evaluated from the mass difference before and after immersion for 2.59 Ms in the pH 7.4 Hanks’ solution.

Fig. 2 Mass gain of titanium abraded in water and Hanks’ solutions with pH 5.0 and 7.4 followed by immersion in pH 7.4 Hanks’ solution at 310 K for 2.59 Ms.

identified with energy dispersive X-ray spectrometory (EDS). 2.4 XPS Surface analysis of specimens immersed in pH 7.4 Hanks’ solution for 3.6 ks (1 h), 86.4 ks (1 d), and 604.8 ks (7 d) was performed with X-ray photoelectron spectroscopy (XPS) using an electron spectrometer (SSI, SSX100). All binding energies given in this paper are relative to the Fermi level and all spectra were excited with the monochromatized Al Kα line (1486.61 eV). The spectrometer was calibrated against Au 4f7/2 (binding energy, 84.07 eV) and Au 4f5/2 (87.74 eV) of pure gold and Cu 2p3/2 (932.53 eV), Cu 2p1/2 (952.35 eV), and Cu Auger L3 M4,5 M4,5 line (kinetic energy, 918.65 eV) of pure copper. The energy values were based on published data.17) The take-off angle for photoelectron detection was 35◦ from the surface of the specimen. The composition and thickness of the surface oxide were estimated, according to the methods of Asami et al.18, 19) Empirical data20–22) and theoretically calculated datum23) of the relative photoionization cross-sections were used for the quantification. 2.5 XRD after heat treatment After immersion in the pH 7.4 Hanks’ solution for 604.8 ks, some of the surface-modified specimens were heat-treated at 873 K for 1800 s under a reduced pressure of 100 Pa to increase the crystallinity of the reactant on titanium. The heattreated specimens were analyzed using X-ray diffractometry (XRD). Two each specimen were prepared for XPS and XRD analysis for each condition, and three specimens were also prepared for mass gain measurement to confirm data reproducibility. 3. Results

2.3 SEM and EDS The surfaces of titanium specimens polished in water and surface-modified in pH 7.4 Hanks’ solution and immersion in Hanks’ solution for 604.8 ks (7 d) were observed after using scanning electron microscopy (SEM) and a precipitate was

3.1 Mass gains The mass gains of surface-modified titanium by immersion in the Hanks’ solution after repassivation pretreatment are shown in Fig. 2. The mass gains of titanium for the spec-

Surface Modification of Titanium Utilizing a Repassivation Reaction in Aqueous Solutions

3007

Fig. 3 Surface of titanium immersed in pH 7.4 Hanks’ solution at 310 K for 604.8 ks after polished in water (a) and titanium polished in pH 7.4 Hanks’ solution (b).

Fig. 4 EDS spectrum obtained from titanium immersed in pH 7.4 Hanks’ solution at 310 K for 604.8 ks after polished in pH 7.4 Hanks’ solution.

imen abraded in Hanks’ solutions were significantly large in comparison with that abraded in water. In addition, the mass gain of the specimen repassivated in pH 5.0 Hanks’ solution was about twice as large as that in pH 7.4 Hanks’ solution. 3.2 Precipitate Surfaces of titanium polished in water and surfacemodified were observed as Fig. 3. A precipitate was observed on titanium polished in Hanks’ solution while nothing was observed on titanium polished in water. Calcium and phosphorus were detected from the precipitate using EDS as shown in Fig. 4. 3.3 XPS In XPS analysis, carbon was detected on all specimens. From the C 1s electron region spectra, it was concluded that none of the specimens contained carbonate because no peak was detected at an energy region of 289–290 eV where carbonate should give a C 1s peak.24) Therefore, carbon detected by XPS can be ascribed to so-called contaminant carbon. Carbon was not contained in the quantification in this study. Apart from carbon, in specimens abraded in Hanks’ solutions, phosphorus and calcium, along with titanium and oxygen, were detected, but other elements contained in the

Fig. 5 XPS atomic concentration of calcium and phosphorus on titanium abraded in water and Hanks’ solutions with pH 5.0 and 7.4 followed by immersion in pH 7.4 Hanks’ solution at 310 K for 3.6 ks (1 h), 86.4 ks (1 d), and 604.8 ks (7 d).

Hanks’ solution were undetected. The main oxidic state of titanium was Ti4+ state. The relative concentrations of titanium, oxygen, calcium, and phosphorus in the surface region of specimens, were normalized by taking the gross amount of these elements as 100 atomic percent. Figure 5 shows the relative concentration of calcium and phosphorus detected from the surface-modified titanium specimens after immersion in Hanks’ solution for 3.6 ks, 86.4 ks, and 604.8 ks. After 86.4-ks and 604.8-ks immersion, calcium and phosphorus are detected more on specimens abraded in Hanks’ solutions than that abraded in water. The P 2p peak showed that phosphorus existed as calcium phosphate because those binding energies are in agreement with that of a calcium phosphate.5–8) The binding energy of Ca 2p3/2 electron from the specimen was 347.6 eV, indicating that calcium exists as Ca2+ .25) Therefore, it can be said that calcium and phosphorus exist as a calcium phosphate where 2− phosphate exists as PO3− 4 and HPO4 .

3008

T. Hanawa, S. Hiromoto, K. Asami, H. Ukai and K. Murakami

precipitation of apatite-like substance on the surface-modified titanium is much faster than that on the titanium polished just in water. Therefore, the surface of titanium was modified by abrasion and the following immersion in Hanks’ solutions for the precipitation of apatite. The mechanism of the surface modification of titanium can be explained as follows. On repassivation process after abrasion or depassivation in Hanks’ solution, the last dehydration process is responsible for the reaction of phosphate ion as explained in a previous paper.16) Phosphate ions possibly collaborate with the dehydration for the anodic tetravalent reac2− tions. Phosphate ions exist as H2 PO− 4 and HPO4 at the pH employed in this study. Therefore, the reaction is described as follows:

Fig. 6 X-ray diffraction patterns of hydroxyapatite powder and those obtained from titanium abraded in Hanks’ solutions with pH 5.0 and 7.4, followed by immersion in pH 7.4 Hanks’ solution at 310 K for 2.59 Ms, and heat-treatment at 873 K for 1800 s.

3.4 XRD The X-ray diffraction patterns of specimens after immersion in Hanks’ solution for 604.8 ks followed by heattreatment at 873 K are shown in Fig. 6. There are peaks at approximately 32◦ (2θ ) on XRD patterns obtained from surface-modified specimens. The positions of these peaks were in accord with that of strong peaks originating from hydroxyapatite whereas no peak should be given by titanium metal around 32◦ . Therefore, the precipitate formed on specimens in Hanks’ solution can be deduced as apatite. The relative peak height at 32◦ for specimen pretreated in pH 5.0 Hanks’ solution is larger than that in pH 7.4. The peaks at about 35◦ , 38.5◦ , and 40◦ for specimens are (100), (002), and (101) diffraction peaks of hcp Ti, respectively. 4. Discussion Judging from the electron binding energies by XPS, calcium and phosphorus in precipitate on surface-modified titanium exist mainly as calcium phosphate. Calcium phosphate precipitate was observed with SEM/EDS. In addition, XRD showed that the precipitates are crystallized to apatite by heat treatment at 873 K. Peaks at 25◦ and 47◦ were undetected in the surface-modified titanium and only single peak was detected at 32◦ because of low crystallinity of apatite formed on titanium. Increasing the crystallinity, these peaks are usually observed. Therefore, the mass gain of the specimen in Hanks’ solution was apparently caused by the precipitation of apatite or its precursor. In addition, a precipitate containing calcium and phosphorus were observed with SEM/EDS. The

− TiO(OH)2 + 2H2 PO− 4 → TiO(H2 PO4 )2 + 2OH

(1)

− TiO(OH)2 + HPO2− 4 → TiO(HPO4 ) + 2OH .

(2)

Reaction (1) is dominant in pH 5.0 Hanks’ solution because H2 PO− 4 exists more in the pH 5.0 solution than in the pH 7.4 solution in which reaction (2) should be dominant. The regeneration of the film finishes in a short period. After the reactions (1) and (2), calcium and phosphate ions are adsorbed by the surface during immersion in the solution after the repassivation process because repassivation process proceeds by the reaction with water and finishes within a short time of the order of millisecond. This calcium and phosphate adsorption process is also important for the modification because calcium phosphate precipitation is accelerated with calcium ions in the surface layer. The mass gain of titanium abraded in pH 5.0 Hanks’ solution is about twice as large as that in pH 7.4 Hanks’ solution (Fig. 2). This is in accordance with the results of XRD, that is, apatite is formed more on titanium when abraded in pH 5.0 Hanks’ solution rather than that in pH 7.4 Hanks’ solution (Fig. 6). However, no difference in amount of calcium and phosphorus quantified by XPS between in specimens abraded in pH 5.0 and pH 7.4 Hanks’ solutions (Fig. 5). The XRD intensity reflects the total amount of apatite-like deposit while XPS analyzes the surface region up to about 10 nm of depth. Theoretical density of hydroxyapatite, Ca10 (PO4 )6 (OH)2 , is 3.16 × 103 kg m−3 . Estimating from the mass gain (Fig. 2), the thicknesses of deposits on specimens are about 25 nm on the water-abraded to 360 nm on the pH 5.0 solution-abraded. The XPS result that the ratio [Ca]/[P] increases with immersion time suggests the reaction (1) will proceed further as follows: TiO(OH)2 + TiO(H2 PO4 )2 → 2TiO(HPO4 )2 + 2H2 O. (3) Neither calcium nor phosphorus ion exists in the surface oxide film on titanium abraded in water without further immersion in a Hanks’ solution16) while they exist on titanium abraded in Hanks’ solution. Furthermore, mass gain for the water-abraded specimen is remarkably small. These results suggest the surface film regenerated in water is compact and suppress further reaction. Incorporation of calcium and phosphorus in the surface impedes the growth of compact surface film and oxide promotes further adsorption of calcium and phosphate ions to the oxide. Low pH will also inhibit compact surface film formation.

Surface Modification of Titanium Utilizing a Repassivation Reaction in Aqueous Solutions

If we can select a solution with a proper pH and ion concentration, calcium phosphate precipitation could be further increased. This surface modification technique can be applied to the surface modification of titanium under machining with a lubricant containing calcium and phosphate ions. 5. Conclusions The surface modification of titanium was attained by surface film regeneration treatment in Hanks’ solutions with pH 5.0 and 7.4. The performance of modified surface was assessed by immersion in Hanks’ solution, followed by mass gain measurements, SEM/EDS, XPS, and XRD and the following conclusions are drawn: (1) During the film regeneration process, uptake of phosphate and calcium ions into the regenerated surface film occurs in Hanks’ solutions. Incorporation of phosphate and calcium ions in the surface films continues during the following immersion in the pH 7.4 Hanks’ solution. (2) Uptake rates of phosphate and calcium ions and total mass gains depend on the composition of initially regenerated surface film which is determined by the environment where the original air-formed film has been removed and new passive film is regenerated. (3) Uptake rates of phosphate and calcium ions depend on immersion time, that is, uptake of phosphate ions is fast in the early stage and slows down, while concentration of calcium ions increases slowly but catches up. (4) Total mass gains are highest when the initial film regeneration is done in the pH 5.0 Hanks’ solution, and those are about half and 1/13 for specimens film-regenerated in the pH 7.4 Hanks’ solution and water, respectively, in comparison with those film-regenerated in the pH 5.0 Hanks’ solution. (5) The molar ratio of calcium ion to phosphate ion is lower than unity but approaches to unity with the increase of immersion time. (6) Calcium ions are combined with phosphate ions, and form apatite or apatite-like calcium phosphate. (7) This surface modification technique can be applied to

3009

the surface modification of titanium under machining with a lubricant containing calcium and phosphate ions. REFERENCES 1) C. K. Wang, J. H. Lin, C. P. Ju, H. C. Ong and R. P. H. Chang: Biomaterials 18 (1997) 1331–1338. 2) F. J. Garcia-Sanz, M. B. Mayor, J. L. Arias, J. Pou, B. Leon and M. Perez-Amor: J. Mater. Sci. Mater. Med. 8 (1997) 861–865. 3) S. Ban and S. Maruno: Biomaterials 19 (1998) 1245–1253. 4) J. S. Chen, H. Y. Juang and M. H. Hon: J. Mater. Sci. Mater. Med. 9 (1998) 297–300. 5) T. Hanawa and M. Ota: Biomaterials 12 (1991) 767–774. 6) T. Hanawa: The Bone-Biomaterial Interface. Ed. J. E. Davies (University of Toronto Press, Toronto, 1991) Chap. 4, pp. 49–61. 7) T. Hanawa and M. Ota: Appl. Surf. Sci. 55 (1992) 269–276. 8) T. Hanawa, O. Okuno and H. Hamanaka: J. Japan Inst. Metals 56 (1992) 1168–1173. 9) J. L. Ong, L. C. Lucas, G. N. Raikar, R. Connatser and J. C. Gregory: J. Mater. Sci. Mater. Med. 6 (1995) 113–119. 10) P. Li and P. Ducheyne: J. Bomed. Mater. Res. 41 (1998) 341–348. 11) A. G. Wever, J. Veldhuizen, H. J. deVries, D. R. Busscher, A. Uges and J. R. van Horn: Biomaterials 19 (1998) 761–769. 12) K. E. Healy and P. D. Ducheyne: J. Biomed. Mater. Res. 26 (1992) 319– 338. 13) K. E. Healy and P. D. Ducheyne: Biomaterials 13 (1992) 553–561. 14) D. F. Williams: Biocompatibility of Clinical Implant Materials, Vol. I, Ed. D. F. Williams (CRC Press, Boca Raton, FL, 1981) pp. 9–44. 15) E. J. Kelly: Mod. Aspect. Electrochem. 14 (1982) 319–424. 16) T. Hanawa, K. Asami and K. Asaoka: J. Biomed. Mater. Res. 40 (1998) 530–538. 17) K. Asami: J. Electron Spectrosc. 9 (1976) 469–478. 18) K. Asami, K. Hashimoto and S. Shimodaira: Corros. Sci. 17 (1977) 713–723. 19) K. Asami and K. Hashimoto: Corros. Sci. 24 (1984) 83–97. 20) K. Hashimoto, M. Kasaya, K. Asami and T. Masumoto: Corros. Eng. 26 (1977) 445–452. 21) K. Asami, M. S. De S´a and V. Ashworth: Proc. 6th European Symposium on Corrosion Behavior of Amorphous Ni–Cr–P–B alloys, 1985, pp. 769–778. 22) K. Asami, S. C. Chen, H. Habazaki, A. Kawashima and K. Hashimoto: Corros. Sci. 31 (1990) 727–732. 23) J. H. Scofield: J. Electron Spectrosc. 8 (1976) 129–137. 24) J. S. Hammond, J. W. Holubka, J. E. DeVries and R. A. Dickie: Corros. Sci. 21 (1981) 239–253. 25) C. D. Wagner: Practical Surface Analysis, 2nd ed., Eds. D. Briggs and M. P. Seah (Wiley, New York, 1990) pp. 595–634.