Biomolecular Engineering 24 (2007) 438442 www.elsevier.

com/locate/geneanabioeng

Morphology and chemical characterization of Ti surfaces modied for biomedical applications

M. Lewandowska a,*, A. Roguska a, M. Pisarek a,b, B. Polak c, M. Janik-Czachor b, K.J. Kurzydowski a
a

Faculty of Materials Science and Engineering, Warsaw University of Technology, Woloska 141, 02-507 Warsaw, Poland b Institute of Physical Chemistry, Polish Academy of Sciences, Kasprzaka 44/52, 01-224 Warsaw, Poland c Department of Chemistry, Warsaw University of Technology, Noakowskiego 3, 00-664 Warsaw, Poland

Abstract The aim of the present work is to characterize in detail the chemical composition and morphology of titanium surfaces subjected to various environments. Modications consisted of exposure of Ti to acidic, alkaline or polymer solutions. Such modications result in chemical and/or morphological changes in the Ti surface. Special attention has been given to identifying the factors inuencing cell adhesion and growth. SEM examinations provided morphological characterization of the Ti samples. Surface analytical techniques such as AES or XPS combined with Ar+ ion sputtering allowed examination of the chemical properties of the Ti surface after chemical pretreatments and investigating the chemical composition of the Ti oxide layer. Raman spectroscopy investigations allowed determination of the crystalline phases of the Ti-oxide layers and characterization of the dextran-modied surface. The results show large differences in the morphology of Ti pretreated with different procedures whereas only minor differences in the chemistry of the surfaces were found. High-resolution Auger investigations have revealed that all the chemical modications of Ti surfaces resulted in the formation of a titanium oxide layer. XPS conrmed that TiO2 is the main component of the chemically modied Ti surface. The Raman spectroscopy investigations showed that the titanium surface with a dextran coating is rich in hydroxyl groups. All the surfaces investigated exhibit a hydrophilic character. The possible inuence of various surface features on surface biocompatibility is discussed. # 2007 Elsevier B.V. All rights reserved.
Keywords: Biomaterials; Titanium; Surface chemical treatment; Auger electron spectroscopy (AES); X-ray photoelectron spectroscopy (XPS); Raman spectroscopy

1. Introduction Titanium is known as a good biomaterial for various medical applications. In bulk form, it is used for the fabrication of implants (Pohler, 2000), whereas in the form of porous structures, it serves as a support for living cells (Spoerke et al., 2005). Biomedical applications of pure titanium are possible due to certain properties, such as relatively high mechanical strength and fatigue resistance (Rack and Qazi, 2006), good corrosion resistance and biocompatibility (Kasemo, 1983). Some of these properties, in particular the biological response of titanium, are strongly determined by the surface characteristicsits morphology, chemistry and physical properties. Surface properties may be changed by applying various surface
* Corresponding author. Tel.: +48 22 660 84 41; fax: +48 22 660 85 14. E-mail address: malew@inmat.pw.edu.pl (M. Lewandowska). 1389-0344/$ see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bioeng.2007.07.002

modications while the crucial bulk properties such as tensile strength and fatigue resistance remain unchanged. There are many physical and chemical methods that can be used in the functionalization or activation of a surface; a relevant review can be found elsewhere (Liu et al., 2004). Among these methods, chemical modications seem to be relatively simple and inexpensive. In addition, they may control the surface characteristics over a broad range. It has been discovered that the contact of cells with materials of different morphology and chemistry results in a modication of their shape and bioactivity of the cells (Anselme, 2000; Boyan et al., 1996). However, the morphological and chemical factors responsible for the stable growth of cells are still not well known. Bearing this in mind, the aim of the present work is to compare various chemical modications of a titanium surface in terms of chemical composition and morphology.

M. Lewandowska et al. / Biomolecular Engineering 24 (2007) 438442

439

2. Experimental
For the investigations a pure titanium (Grade 2) was used in the form of a 6 mm rod cut into cylinders 2 mm in thickness. The samples were polished, degreased with detergent and rinsed in distilled water. Next, they were subjected to four different chemical surface modications: (1) etching in a piranha solution (98% H2SO4 + 30% H2O2 mixture, the volume ratio 1:1) at room temperature for 4 h (this pretreatment is assigned as PRT), (2) etching in piranha boiling solution (PBS) for 10 min, (3) soaking in 5.0 M NaOH aq. at 60 8C for 24 h, washing in deionized water at 40 8C for 48 h, heating with a rate of 5 8C/h up to 600 8C, annealing at this temperature for 1 h, and then cooling with a furnace (this pretreatment is assigned as SH, standing for sodium hydroxide), and (4) immersing in a water dextran solution (0.05 wt.%) for 10 min (assigned as D, standing for dextran). An unmodied titanium surface was used as a reference. Morphology of the modied titanium surfaces was examined by a scanning electron microscope (SEM, S-3500N, Hitachi). The chemical compositions of the modied surface layers were analyzed by Auger electron spectroscopy (AES) and X-ray photoelectron spectroscopy (XPS) (Briggs and Grant, 2003). For this purpose an Auger microprobe analyzer, a Microlab 350 (Thermo Electron) with XPS optional function was applied with a lateral resolution of about 20 nm for AES and several mm for XPS. The chemical state of the surface species was identied using the high-energy resolution of the Auger spectrometer (the energy resolution of the spherical sector analyzer is continuously variable between 0.6% and 0.06%) and of the XPS spectrometer (the maximum energy resolution is 0.83 eV). Appropriate standards for AES and XPS reference spectra were also used. XPS spectra were excited using Al Ka (hn = 1486.6 eV) radiation as a source. The measured binding energies were corrected referring to the energy of C 1s signal at 285 eV. An Avantage based data system was used for data acquisition and processing. The crystalline phases of the layers were identied by Raman spectroscopy. Spectra were recorded on a Nicolet Almega XR apparatus (laser: 532 nm; exposure time: 30 s; laser power:

25 mW; spectrograph aperture: 50 mm pinhole). In order to evaluate the wettability of the modied titanium surfaces, contact angle measurements were carried out at a constant room temperature. A given volume of DMEM (Dulbeccos Modied Eagle Medium) was placed on the surface in order to form a drop. Then, the contact angle was measured on both sides of the drop using the goniometric method.

3. Results Fig. 1 shows SEM images of titanium surfaces. In the asreceived state characteristic, oriented grooves resulting from grinding can be seen (Fig. 1a). The morphology of the polymercoated surfaces (Fig. 1b) is similar to that in the as-received state. The soaking in NaOH aq. solution (SH) results in a developed morphology similar to a honeycomb (Fig. 1c). The surface layer seems also quite porous. Pores of diameter 0.10.2 mm are well visible in the honeycomb islands, surrounded by elongated depressions and cracks, several mm long and about 0.5 mm wide. After etching in piranha solution at room temperature (PRT) or in boiling piranha solution (PSB) the morphology is quite different and less developed than that produced in the SH pretreatment. However, sub-microporosity effects on the piranha treated surfaces can also be seen, e.g. in boiling piranha solution a high population of shallow depressions about 0.51 mm in diameter is formed (Fig. 1d). In order to determine the chemical composition of the titanium surfaces before and after modication, an AES study was performed. The AES survey spectra taken from the surface

Fig. 1. SEM images of typical morphologies of Ti surface subjected to different chemical modications: (a) as-received, (b) after immersing in water dextran solution (D), (c) after NaOH and heat pretreatment (SH), and (d) after piranha pretreatment in boiling solution (PBS).

440

M. Lewandowska et al. / Biomolecular Engineering 24 (2007) 438442 Table 1 Ti 2p3/2 binding energies as measured from the corrected XPS spectra and the surface TiO compounds evaluated from deconvolution procedure Samples pretreatment As-received PRT Binding energy (eV), Ti 2p3/2 458.9 458.3 456.2 454.6 458.5 458.8 457.8 TiO2 Ti2O3 TiO2 Ti2O3 TiO TiO2 77% 23% Compound TiO2 67% 21% 12%

PBS SH

Fig. 2. High-resolution Auger spectra (energy analyzer resolution 0.1 eV within the energy range 350440 eV) taken at Ti surface chemically modied in different solutions. For comparison, Ti LMM reference spectra for metallic Ti and titanium oxide TiOx are given, as indicated in the gure.

of the materials analyzed indicate that Ti was oxidized on all the surfaces investigated (the signals from Ti and O are well distinguishable). Moreover, a benecial incorporation of Ca is also noted for the SH-modied surface, which apparently results from the preparation procedure. Fig. 2 presents a set of high-resolution Ti LMM spectra showing Ti LMM signals for the samples under investigation, as well as Ti metallic and TiO2 oxide reference spectra. Careful inspection of all of them conrms that there are Ti-oxides on all the surfaces investigated before and/or after the NaOH or piranha pretreatment. Comparison of these spectra does not reveal any very distinct difference in their shape suggesting that similar Ti-oxides are formed at the surface as a result of each of the pre-treatments. In order to get a better insight into the chemical state of titanium in the Ti-oxygen surface compounds, additional XPS measurements were performed on NaOH aq. and piranha treated pure Ti rods. Fig. 3 demonstrates the corresponding Ti 2p spectra before Ar+ ion sputtering. These results conrm that

TiO2 is the main component of the chemically modied Ti surface. Deconvolution of the Ti 2p signals suggests, however, that some lower Ti-oxides are also present in the pretreated samples. The stoichiometry of these oxides and their volume fractions, as estimated from the XPS data, are listed in Table 1. Contamination of the samples with carbon, in particular those pretreated with piranha, made the analysis difcult. In order to clean the surface, Ar+ ion bombardment was used. Analysis of spectra after the ion bombardment may provide some information about the stability of oxide layers. NaOH treated samples produced an oxide layer which was rather quite stable against Ar+ ion bombardment. The spectra did not change much after sputtering. Piranha-treated Ti shows a more complex XPS spectra. This may reect the presence of a surface oxide which is more prone to the modications introduced by Ar+ ion sputtering than that produced by the SH pretreatment. The crystallographic structure of the titanium oxide layers was analyzed by Raman spectroscopy. The results of these investigations are presented in Fig. 4. Raman spectra of the titanium surface before and after modication consist of a band at 143 cm1 which correspond to anatase. The spectra of the SH-modied surface consist additionally of two strong bands at 440 and 608 cm1 and a weaker band at 248 cm1, which indicate rutile to be present at the surface. The bands corresponding to anatase vary in intensity, which can be attributed to the different content of crystalline anatase phase in the surface layer.

Fig. 3. Ti 2p XPS spectra taken at Ti surface chemically modied in different solutions: as-received, after piranha pretreatment at room temperature (PRT), after piranha pretreatment in boiling solution (PBS), and after NaOH and heat pretreatment (SH).

Fig. 4. Raman spectra of Ti surface chemically modied in different solutions. A, anatase; R, rutile.

M. Lewandowska et al. / Biomolecular Engineering 24 (2007) 438442

441

Fig. 5. Raman spectrum of Ti surface covered with a discontinuous dextran coating. OH and CH2 signals from dextran-covered part of the surface are assigned. A, anatase signal from the uncovered part is well distinguishable.

Fig. 6. Diagram presenting the values of contact angle for unmodied and modied titanium surfaces.

Fig. 5 shows the Raman spectrum of a Ti surface with a dextran coating. Bands from OH groups (3500 cm1) and CH2 groups (2900 cm1) indicate that dextran is present at the surface. However, there is also a clear distinguishable band corresponding to anatase (143 cm1) in the spectrum. This implies that the polymer lm is not homogenously distributed at the surface, and remains discontinuous leaving some areas not covered at all. The contact angle was measured using a DMEM solution, whose chemical composition is similar to human body uids. The values of contact angle obtained are presented in Fig. 6. The results clearly indicate that SH and PBS pretreatment led to a higher contact angle than that for the other pretreated samples. The samples subjected to these modications are characterized by sub-micro- and microporosity, which apparently inuence the wettability of the titanium surface. However, it should be noted that wettability is an average surface property measured on the macroscopic scale. The lowest value of contact angle (458) is observed for the samples with a dextran coating. 4. Discussion It is generally accepted that rough, textured and porous surfaces are able to stimulate cell attachment, differentiation

and formation of the extracellular matrix (Boyan et al., 1996). Moreover, an appropriate surface roughness can produce benecial mechanical interlocking at the initial adhesion stage and aid in further cell adhesion (der Brugge et al., 2002). The morphology of NaOH or piranha treated surfaces can be generally described as microscopically rough due to the grinding, and as porous on the sub-micrometric scale due to chemical treatment (etching and oxidation). However, after etching in PRT or in PBS, the morphology is quite different and in fact less developed than that produced in the SH pretreatment. The differences in surface morphology after these chemical surface modications may be caused by different chemical reactions leading to oxide formation. In the case of piranha solution, the oxidizing factor is H2O2, whereas for alkaline solution is H2O (corrosion under water decomposition, see Kaesche, 1979). The kinetics of these reactions vary depending on temperature and environment. Thus, the different topography of the surface may result from various oxide thickness. Recently, more attention has been paid to the role of surface chemistry in the behavior of cells. Surface components inuence the response of cells to biomaterials (Tambasco de Oliveira and Nanci, 2004; Feng et al., 2003). In the present investigations Auger electron spectroscopy has shown that all the chemical modications of Ti surfaces resulted in formation of a titanium oxide layer. XPS conrmed that TiO2 is the main component of the chemically modied Ti surface. However, some lower Ti-oxides are also present for the samples pretreated in the piranha solution at room temperature or in the NaOH solution. More detailed discussion is given elsewhere (Lewandowska, 2007). The other factor which can inuence cell behavior and growth is the crystal structure of TiO2. TiO2 may occur in three polymorphic forms: anatase, brookite and rutile. Rutile is thermodynamically the most stable modication, whereas anatase and brookite can undergo irreversible exothermic transition to rutile over a wide range of temperatures (Diebold, 2003). It is generally accepted that the natural surface oxide lm of titanium is anatase. Raman spectroscopy investigations showed that after alkali and subsequent heat treatment the surface oxide is a mixture of rutile and anatase. This may be a result of the heat treatment applied. Etching in piranha solution (PRT and PBS) leaves anatase unchanged. Some authors suggest that oxide structure plays an important role in ion release into the physiological uid (Lee et al., 2000). The different levels of ion release change the environmental uid condition and further inuence the reaction between cell and biomaterial. Rutile is denser with a closer packed structure. Thus, ion diffusion will be more difcult than for anatase and the corrosion resistance of rutile will be greater. Analysis of the Raman spectroscopy data showed that the titanium surface covered with dextran coating is rich in hydroxyl groups. However, the polymer coating is not uniform, but occurs locally at the surface (see the prominent Raman signal from anatase, Fig. 5). It is known from the literature that cell adhesion is generally better on hydrophilic surfaces. Good wettability of biomater-

442

M. Lewandowska et al. / Biomolecular Engineering 24 (2007) 438442

ials surface is an important factor in obtaining chemical interaction with physiological uids (Altankov and Groth, 1994). It should be noted here that all surfaces under investigation exhibit a hydrophilic character. The samples with a dextran coating show the best wettability among the materials investigated. The lowest value of contact angle observed for this modication may be ascribed to the presence of hydroxyl groups at the surface (compare Fig. 5). 5. Conclusions In this study, pure titanium was subjected to various surface modications and examined in terms of morphology, chemical characteristics and wettability. The results have shown that: (1) Etching in alkaline or acid solutions results in signicant changes in surface morphology, whose characteristic feature is the presence of sub-microporosity. (2) Chemical composition of etched surfaces does not differ signicantlyTi-oxides (mostly TiO2) are formed at the surface as a result of each of the pretreatments. (3) Annealing at 600 8C of a NaOH treated surface leads to the partial transition of the anatase phase to a more stable rutile, which may have some impact on the corrosion resistance of the modied material. (4) Dextran modication by immersing to its water solution allows one to obtain a surface rich in hydroxyl groups with higher wettability, but this coating is not uniform. Further studies are required in order to determine all microstructural and physico-chemical factors responsible for cell response and to optimize the modication procedures. Acknowledgments Surface characterizations were performed using a Microlab 350 located at the Physical Chemistry of Materials Center of the Institute of Physical Chemistry, PAS, and of the Faculty of Materials Science and Engineering, WUT. This work was nancially supported by the Polish Ministry of Science and Higher Education (grant No 3 T08A 01829), the Institute of

Physical Chemistry, PAS, and by the Foundation for Polish Science through a generous fellowship for M. Pisarek.

References
Altankov, G., Groth, T., 1994. Reorganization of substratum-bound bronectin on hydrophilic and hydrophobic materials is related to biocompatibility. J. Mater. Sci. Mater. Med. 5, 732737. Anselme, K., 2000. Osteoblast adhesion on biomaterials. Biomaterials 21, 667681. Boyan, B.D., Hummert, T.W., Dean, D.D., Schwartz, Z., 1996. Role of material surfaces in regulating bone and cartilage cell response. Biomaterials 17, 137146. Briggs, D., Grant, J.T., 2003. Surface Analysis by Auger and X-ray Photoelectron Spectroscopy. IM Publications, Chichester (Chapters 2 and 3). Diebold, U., 2003. The surface science titanium dioxide. Surf. Sci. Rep. 48, 53229. Feng, B., Weng, J., Yang, B.C., Qu, S.X., Zhang, X.D., 2003. Characterization of surface oxide lms on titanium and adhesion of osteoblast. Biomaterials 24, 46634670. Kaesche, H., 1979. Die Korrosion der Metalle. Springer-Verlag, New York (Chapter 10.4). Kasemo, B., 1983. Biocompatibility of titanium implants: surface science aspects. J. Prosthet. Dent. 49, 832. Lee, T.M., Chang, E., Yang, C.Y., 2000. A comparison of the surface characteristics and ion release of Ti6Al4V and heat-treated Ti6Al4V. J. Biomed. Mater. Res. 50, 499511. Lewandowska, M., Pisarek, M., Roguska, A., Kurzydowski, K.J., Janik-Czachor, M., 2007. SEM, scanning Auger and XPS characterization of chemically pretreated Ti surfaces intended for biomedical applications Mater. Chem. Phys. 104, 9397. Liu, X., Chu, P.K., Ding, C., 2004. Surface modication of titanium, titanium alloys, and related materials for biomedical applications. Mater. Sci. Eng. R 47, 49121. Pohler, O.E.M., 2000. Unalloyed titanium for implants in bone surgery. Injury Int. J. Care Injured 31, S-IX-13. Rack, H.J., Qazi, J.I., 2006. Titanium alloys for biomedical applications. Mater. Sci. Eng. C 26, 12691277. Spoerke, E.D., Murray, N.G., Huanlong, L., Brinson, L.C., Dunand, D.C., Stupp, S.I., 2005. A bioactive titanium foam scaffold for bone repair. Acta Biomater. 1, 523533. Tambasco de Oliveira, P., Nanci, A., 2004. Nanotexturing of titanium-based surfaces upregulates expression of bone sialoprotein and osteopontin by cultured osteogenic cells. Biomaterials 25, 403413. der Brugge, P.J., Dieudonne, S., Jansen, J.A., 2002. Initial interaction of U2OS cells with noncoated and calcium phosphate coated titanium substrates. J. Biomed. Mater. Res. 61, 399407.