Topography and Surface Energy Dependent Calcium Phosphate

Heterogeneous nucleation and growth of calcium phosphate (CaP) on sol−gel derived TiO2 coatings was investigated in terms of surface topography and ...
0 downloads 0 Views 314KB Size
Download PDF
Langmuir 2006, 22, 8209-8213

8209

Topography and Surface Energy Dependent Calcium Phosphate Formation on Sol-Gel Derived TiO2 Coatings Mikael Ja¨rn,† Sami Areva,† Viljami Pore,‡ Jouko Peltonen,*,† and Mika Linden† Department of Physical Chemistry, A° bo Akademi UniVersity, Porthansgatan 3-5, FIN-20500 Turku, Finland, and Laboratory of Inorganic Chemistry, Department of Chemistry, UniVersity of Helsinki, P.O. Box 55, FI-00014 Helsinki, Finland ReceiVed April 10, 2006. In Final Form: June 29, 2006 Heterogeneous nucleation and growth of calcium phosphate (CaP) on sol-gel derived TiO2 coatings was investigated in terms of surface topography and surface energy. The topography of the coatings was derived from AFM measurements, while the surface energy was determined with contact angle measurements. The degree of precipitation was examined with scanning electron microscopy (SEM) and X-ray photoelectron spectroscopy (XPS). The precipitation of CaP was found to be dependent on both topography and surface energy. A high roughness value when combining the RMS roughness parameter Sq with the number of local maxima per unit area parameter Sds enhances CaP formation. The hydrophilicity of the coating was also found to be of importance for CaP formation. We suggest that the water contact angle, which is a direct measure of the hydrophilicity of the surface, may be used to evaluate the surface energy dependent precipitation kinetics rather than using the often applied Lewis base parameter.

Introduction Heterogeneous nucleation of minerals is an important phenomenon in many processes. In many industrial applications, it is important to prevent the formation of mineral scales. For example, in cooling water technology, desalination, paper industry, and oil production, scale formation is a serious problem. The surface precipitation (i.e., heterogeneous nucleation) of carbonates, sulfates, phosphates, and oxalates of alkaline earth metals is an often encountered problem.1 Also, the formation of calcium oxalate kidney stones is a serious problem.2 On the other hand, rapid nucleation of calcium phosphates may be desired in dental or orthopedic implants to enhance their integration to bone tissue, whereas undesirable nucleation occurring in the vascular system may result in arteriosclerosis. Thus, control over biomineralization (i.e., heterogeneous nucleation of minerals during tissue generation) is essential in modern medicine for developing new biomaterials as well as preventing undesired calcifications. Therefore, it is of vital interest to understand the relation between the physicochemical properties of the surface and the heterogeneous nucleation and growth of inorganic deposits.1-4 Many previous studies concerning the heterogeneous nucleation and growth of calcium phosphate (CaP) have been reported. Most of the studies have involved calcium phosphate growth on powders. Kinetic studies of nucleation and growth on dispersed particles and macromolecules from a wide range of supersaturated solutions have been performed using the constant composition method.5-10 Surface precipitation of calcium phosphate on planar * To whom correspondence should be addressed. E-mail: [email protected]. †A ° bo Akademi University. ‡ University of Helsinki. (1) Mineral scale formation and inhibition; Amjad, Z., Ed.; Plenum Press: New York, 1995. (2) Singh, R. P.; Gaur, S. S.; White, D. J.; Nanchollas G. H. J. Colloid Interface Sci. 1987, 118, 379. (3) Tarasevich, B. J.; Chusuei, C. C.; Allara, D. L. J. Phys. Chem. B 2003, 107, 10367. (4) Wu, W.; Zhuang, H.; Nancollas, G. H. J. Biomed. Mater. Res. 1997, 35, 93. (5) Wu, W.; Nancollas, G. H. Langmuir 1997, 13, 861. (6) Koutsoukos, P. G.; Nancollas, G. H. Colloids Surf. 1987, 28, 95. (7) Dalas, E.; Kallitsis, J. K.; Koutsoukos, P. G. Langmuir 1991, 7, 1822.

surfaces has mainly been studied from solutions with a relatively high degree of supersaturation.11-15 Heterogeneous nucleation of calcium phosphate is generally believed to be initiated by the adsorption of calcium ions onto negatively charged surface sites.7,8,13,16 TiO2 surfaces are good model surfaces for such studies in the sense that the isoelectric point of titania is about 6, for both anatase and rutile, which means that the surface carries a net negative charge in aqueous solutions at pH > 6.16 Furthermore, the solubility of TiO2 is low, which makes the interpretation of the results more straightforward. A fast nucleation and growth of calcium phosphate on titania has previously been associated with a high Lewis base surface tension parameter of the substrate.4,17 The surface topography at the nanometer level has also been shown to influence the calcium phosphate formation in simulated body fluid.18 This phenomenon was related to the charge density and the topographical matching of the titania surface and CaP crystal size found in bone.19-21 Peltola et al.22 suggested that a certain distance distribution between surface heterogeneities favored calcium phosphate precipitation. A procedure for enhanced quantification and specification of surfaces by utilizing a comprehensive set of roughness parameters (8) Spanos, N.; Koutsoukos, P. G. J. Mater. Sci. 2001, 36, 573. (9) Combes, C.; Fre`che, M.; Rey, C. J. Mater. Sci. 1999, 10, 231. (10) Dalas, E.; Chrissanthopoulos, A. J. Cryst. Growth 2003, 255, 163. (11) Tanahashi, M.; Kokubo, T.; Matsuda, T. J. Biomed. Mater. Res. 1996, 31, 243. (12) Wen, H. B.; Wolke, J. G. C.; de Wijn, J. R.; Liu, Q.; Cui, F. Z.; de Groot, K. Biomaterials 1997, 18, 1471. (13) Kokubo, T. Acta Mater. 1998, 7, 2519. (14) Nooney, M. G.; Campbell, A.; Murrell, T. S.; Lin, X. F.; Hossner, L. R.; Chusuei, C. C.; Goodman, D. W. Langmuir 1998, 14, 2750. (15) Barrere, F.; Snel, M. M. E.; van Blitterswijk, C. A.; de Groot, K.; Layrolle, P. Biomaterials 2004, 25, 2901. (16) Viitala, R.; Jokinen, M.; Peltola, T.; Gunnelius, K.; Rosenholm, J. B. Biomaterials 2002, 23, 3073. (17) Wu, W.; Nancollas, G. H. J. Colloid Interface Sci. 1998, 199, 206. (18) Jokinen, M. Ph.D. Thesis, Abo Akademi University, 1999. (19) Jokinen, M.; Patsi, M.; Rahiala, H.; Peltola, T.; Ritala, M.; Rosenholm, J. B. J. Biomed. Mater. Res. 1998, 4, 295. (20) Peltola, T.; Jokinen, M.; Rahiala, H.; Pa¨tsi, M.; Heikkila¨, J.; Kangasniemi, I.; Yli-Urpo, A. J. Biomed. Mater. Res. 2000, 51, 200. (21) Peltola, T.; Paldan, H.; Moritz, N.; Areva, S.; Korventausta, J.; Jokinen, M.; Narhi, T.; Happonen, R. P.; Yli-Urpo, A. Key Eng. Mater. 2002, 218-220, 207. (22) Peltola, T.; Jokinen, M.; Rahiala, H.; Pa¨tsi, M.; Heikkila¨, J.; Kangasniemi, I.; Yli-Urpo, A. J. Biomed. Mater. Res. 2000, 51, 200.

10.1021/la060966+ CCC: $33.50 © 2006 American Chemical Society Published on Web 08/19/2006

8210 Langmuir, Vol. 22, No. 19, 2006

was reported recently.23 This approach should apply very well for the evaluation of the role of topography on nucleation. However, it is to be expected that both the morphology and the chemical properties of the surface have a cooperative influence on any surface precipitation reactions, and very few if any studies have been carried out where the combined effect of these parameters has been systematically studied. For example, if the previously described relations are valid, a surface having a suitable surface roughness for precipitation but exhibiting a low Lewis base surface tension parameter may be expected not to induce surface nucleation and vice versa. In this paper, the combined effect of surface topography and the effect of surface energy on calcium phosphate formation have been studied. The studied materials were sol-gel derived TiO2 coatings. The topography of the materials was derived from AFM measurements. We have recently demonstrated that by utilizing a versatile set of roughness parameters, the topography dependent functionality of a surface may be specified.23 The results suggest that both the topography and the surface energy govern the nucleation of calcium phosphate. Experimental Procedures Preparation of the Sol-Gel Samples. Commercially pure (c.p.) titanium was used as a substrate material for the experiments. Titanium was ground by silicon carbide papers having 320 and 1000 grits and cleaned thoroughly with acetone followed by ultrasonic washing in acetone and ethanol for 5 + 5 min before dipping into the sol. Some coatings were also prepared on microscope glass slides. The titania coatings were prepared by the sol-gel dip coating technique as described previously,24,25 but with slight modifications. Briefly, the sol was prepared by dissolving tetraisopropylorthotitanate [Ti(OCH(CH3)2)4] into absolute ethanol (solution I). Solution II was prepared by mixing ethyleneglycolmonoethyl ether (C2H5OCH2CH2OH), deionized water, and 1 M hydrochloric acid (37%) with absolute ethanol. Solutions I and II were mixed at room temperature with vigorous stirring. The clear sol was kept at 0 °C during aging and the dip coating process. The coating procedure started after 24 h of sol aging, and the number of coating layers was three. After deposition of each layer, the substrates were sintered at various temperatures and time periods given in the text and subsequently cleaned in an ultrasonic bath with acetone and ethanol (5 + 5 min) and finally dried in air. The uniformity of the coatings was generally good, with little or no cracks. Characterization. Atomic Force Microscopy (AFM) Measurements and Image Analysis. The AFM images were recorded with a Nanoscope IIIa AFM (Digital Instruments, Santa Barbara, CA). All the images were recorded in tapping mode using silicon cantilevers with a resonance frequency between 250 and 300 kHz. The scan rate was typically 0.7-2 Hz. The free tapping amplitude was 70-100 nm for the high kinetic energy tapping measurements. The damping ratio rsp () Asp/A0) controlling the level of forced damping was chosen by tuning the set-point amplitude, Asp. All images (512 pixels × 512 pixels) were measured in air without filtering. The microscope was placed on an active vibration isolation table (MOD-1M, JRS Scientific Instruments, Switzerland), which was further placed on a massive stone table to eliminate external vibrational noise. The Scanning Probe Image Processor (SPIP, Image Metrology, Denmark) software was used for the roughness analysis of the images.26 A set of roughness parameters has been developed and standardized for versatile characterization of various surface properties in three dimensions.27 The root mean square (RMS) roughness Sq is the most widely used amplitude roughness parameter that actually gives the (23) Peltonen, J., Ja¨rn, M.; Areva, S.; Linden, M.; Rosenholm, J. B. Langmuir 2004, 20, 9428. (24) Jokinen, M.; Pa¨tsi, M.; Rahiala, H.; Peltola, T.; Ritala, M. J. Biomed. Mater. Res. 1998, 42, 295. (25) Kajihara, K.; Yao, T. Sol-Gel Sci. Technol. 1998, 12, 185. (26) Image Metrology. The Scanning Probe Image Processor, SPIP, User’s and Reference Guide; Image Metrology: Copenhagen, 2001.

Ja¨rn et al. standard deviation of height. Surface skewness Ssk describes the asymmetry of the height distribution. A skewness value equal to 0 represents a Gaussian-like surface. Negative values of Ssk refer to a surface porous sample (i.e., the valleys dominate over the peak regimes). Correspondingly, the local maxima dominate over the valleys for Ssk > 0. Surface kurtosis Sku gives a measure for the sharpness of the surface height distribution. A Gaussian value for this parameter is 3.0, and much smaller values indicate a very broad (heterogeneous) height distribution, whereas values much larger than 3.0 refer to a surface with almost quantized height values. The number of local maxima per unit area is given by the spatial parameter Sds. Besides the number, also the form of the local maxima (summits) is of central interest. Two hybrid parameters describe the form of the summits: the mean summit curvature, Ssc, and the RMS value of the surface slope, Sdq. Most of the previous parameters contribute to the effective surface area: the absolute height difference and the number and form of local maxima, among others. A measure for the effective surface area with respect to the projected area is given in percent by the surface area ratio parameter Sdr. A more thorough description of the parameters was given in a recent paper.23 The reported roughness values in this study are mean values from five to 10 measurements. Contact Angle Measurements. The contact angle of water, ethylene glycol (EG, Merck purity 99.5%), and diiodomethane (DIM, Sigma purity 99%) against the studied substrate was measured optically by using a Kru¨ss contact angle instrument. The Milli Q filtration system (Millipore Corp.) was used for water purification. The drop volume was about 1 µL. The reported values are mean values of 10 measurements. The surface energy components of the solid substrate were calculated with the van Oss-Chaudhury-Good method. X-ray Diffraction (XRD). Film crystallinity was examined with a Bruker D8 Advance X-ray diffractometer at a grazing incidence mode using Cu KR radiation. The average crystallite sizes of anatase and rutile were determined according to the Scherrer equation using the peak position and full width at half-maximum (fwhm) data of the integrated reflections from anatase [101] (at 25.4° 2θ)28 and rutile [110] (at 27.84° 2θ).29 The amount of rutile in the coatings was calculated by the following semiempirical equation:30 rutile % ) 100 - 100/(1 + 1.33Irutile/Ianatase)

(1)

Precipitation Experiments. The supersaturated calcium phosphate solution was prepared by dissolving reagent grade chemicals of NaCl, CaCl2‚2 H2O (Fluka), and Na2HPO4 (Fluka) into Milli Q purified water. The molar concentration of calcium was 3.0 mM, phosphate was 1.2 mM, and sodium chloride was 150 mM. The pH was buffered at 7.4 with tris(hydroxymethyl)aminomethane and 1 M HCl. The solutions were stable for a period of at least 3 weeks. The solution was filtered through 0.22 µm filters before use. The sol-gel prepared titania coatings were cut into pieces having dimensions of 1 cm × 1 cm. The coatings were ultrasonically cleaned in acetone, ethanol, and water for 5 + 5 + 5 min. The films were then dried in an oven at 40 °C for about 4 h before immersion in 20 mL of supersaturated calcium phosphate solution to give a surface area to solution volume ratio of 0.1. The solutions were kept in polyethylene bottles covered with a tight lid and placed in a shaking water bath at 25 °C for a period up to 14 days. At the end of the experiment, the coatings were washed with Milli Q water and dried at room temperature prior to scanning electron microscopic (SEM) and XPS analysis. Scanning Electron Microscopy (SEM). The surface structure was studied by scanning electron microscope, SEM (LEO type (27) Stout, K. J.; Sullivan, P. J.; Dong, W. P.; Mainsah, E.; Luo, N.; Mathia, T.; Zahouani, H. The deVelopment of methods for the characterization of roughness on three dimensions; Publication No. 15178 EN of the Commission of the European Communities: Luzemburg, 1994. (28) International Centre for Diffraction Data (ICDD), Card 21-1272. (29) International Centre for Diffraction Data (ICDD), Card 21-1276. (30) PC-APD Automated powder diffraction software operation manual, 4th ed.; April 1992; Ch. 11, p 11.

Calcium Phosphate Formation on TiO2 Coatings

Langmuir, Vol. 22, No. 19, 2006 8211

Figure 1. X-ray diffractograms of TiO2 samples processed at 500 °C for 1 h (sample A) and 700 °C for 1 h (sample B). Stereoscan 360, UK) equipped with a Link Inca 300 (Oxford Instruments, Great Britain)) EDS unit, which was used for the elemental analysis. The samples were coated with gold prior to analysis. X-ray Photoelectron Spectroscopy (XPS). The composition of the outermost surface of the samples was studied with X-ray photoelectron spectroscopy (XPS) using a Physical Electronics Quantum 2000 instrument equipped with a monochromatic Al KR X-ray source. An operating power of 25 W was used with a spot diameter of 1 mm. An electron flood gun and a low energy ion gun were used for charge compensation. The detector position was at an angle of 45° in relation to the sample surface.

Results and Discussion Surface Characterization. TiO2 Film Crystallinity. The crystallinity of the TiO2 films was determined by X-ray diffraction, XRD. Generally, films calcined at 300 °C were amorphous, calcinations at 500 °C resulted in anatase films (calcination time 10 min) or mixed anatase-rutile films (calcination time 1 h or longer), and films calcined at 700 °C were mainly (about 80%) in the rutile form. A longer calcination time leads to the formation of larger crystallites, as observed by the decrease in the measured full widths at half-maximum of the anatase and rutile reflections. The diffractograms of two different coatings (A and B) are presented in Figure 1. Sample A was calcined at 500 °C for 1 h, resulting in a coating of about 65% anatase and 35% rutile. The crystallite sizes were roughly 31 and 13 nm for anatase and rutile, respectively. Sample B was calcined at 700 °C for 1 h, which resulted in a rutile amount of about 80%. The crystallite sizes were 69 nm (anatase) and 65 nm (rutile). TiO2 Film Topography. Typical AFM images of the two different TiO2 samples in Figure 1 (A and B) are presented in Figure 2. Roughness parameters calculated from the AFM images are given in Table 1. In Figure 2, the rutile coating (sample B) looks rougher than the anatase coating (sample A), even if the height scales (Zrange) are almost equal. The general roughness difference is evidenced by the Sq parameter. Note, however, that any roughness value is dependent on the length scale of the measurement and that the roughness values reported in this paper are based on 3 µm × 3 µm AFM images. Sample A is composed of smaller particles than Sample B, which appears, besides smaller RMS roughness, as a higher density of summits (Sds). Single crystals cannot be observed from the AFM images, rather aggregates of

Figure 2. AFM images (3 µm × 3 µm) of TiO2 samples processed at (A) 500 °C for 1 h (sample A), Z-range 95.8 nm and (B) 700 °C for 1 h (sample B), Z-range 103.8 nm. The x/y/z ratio in the figure is 1:1:2. Table 1. Roughness Parameter Values for the Samples Shown in Figure 2 sample

A

B

Sq (nm) Ssk Sku Sds (µm-2) Sdr (%)

8.9 0.35 3.93 499 0.53

16.2 -0.32 3.52 75 1.85

particles. The differences in kurtosis (Sku) are small, both values being close to that of a Gaussian-like surface. The skewness (Ssk) values, however, demonstrate a clear difference between the samples. For sample A, the height asymmetry is caused by the dominating high regions, whereas the surface of sample B is dominated by valleys. The Sdr parameter can be used to calculate the r-value in the well-known Wenzel equation.31 According to Wenzel, the relation between the roughness dependent measured contact angle θm and Young’s contact angle θY corresponding to an ideally flat surface may be written as

cos θm ) r cos θY,

(2)

where r denotes the ratio between the real and the projected surface area of the sample. The roughness parameter Sdr may be used to calculate r from

r ) 1 + Sdr/100

(3)

The Sdr parameter of samples A and B gives r-values of 1.0053 and 1.0185. This means that the roughness induced increase of the surface area is very marginal for these samples and that the application of the Wenzel equation in the contact angle studies is therefore not necessary. Surface Energy Determinations. Lifshitz-van der Waals (σSVLW) and Lewis acid-base (σSV- and σSV+) components were calculated according to the van Oss-Chaudhury-Good (vOCG) (31) Wenzel, R. W. Ind. Eng. Chem. 1936, 28, 988.

8212 Langmuir, Vol. 22, No. 19, 2006

Ja¨rn et al.

Figure 3. Plot of total surface energy and Lewis base surface energy parameter vs water contact angle for a series of titania coatings.

method.32 According to the vOCG formalism, the surface energy components can be calculated from the contact angle values of three liquids with known properties (water, ethylene glycol, diiodomethane) according to

(1 + cos ΘSL)γLV ) 2(xγLVLWσSVLW + xγLV+σSV- + xγLV-σSV+ ) (4) The total surface energy, σSVtot, can then be obtained according to

σSVtot ) σSVLW + σSVAB ) σSVLW + 2xσSV+σSV-

Figure 4. SEM photographs of samples A and B after 14 days in supersaturated CaP solution.

(5)

In Figure 3, the Lewis base parameter and the total surface energy are plotted against the contact angle of water for a series of titania coatings. The variation in total surface energy is quite small between the coatings. However, the differences in the base parameter derived from the vOCG equation are significant. The base component gives a linear relationship to the water contact angle, which indicates that the interaction with water gives rise to the difference in the base parameter. The base component of the surface energy has previously been shown to correlate with the kinetics of calcium phosphate formation; a high value of the base parameter leads to a faster surface precipitation of calcium phosphaste.4,17 However, we find it peculiar that the base parameter would become dominant for very hydrophilic surfaces, which should contain a high number of Lewis acidic Ti-OH groups. We will return to this point in more detail next when we discuss the CaP precipitation results, but we note that the values obtained for the base parameter may be erroneous, due to the inherent sensitivity of the model when the LW component of the total surface energy is very dominant, as in the present case. Surface Precipitation of Calcium Phosphate on TiO2 Films. The ability of the TiO2 films to nucleate CaP on their surface was determined by immersing the films into a supersaturated CaP solution for 14 days. This is a commonly applied time frame in in vitro determinations of biomineralization capabilities of biomaterials.33 The coatings were then analyzed by SEM and XPS. As representative examples, SEM images of samples A and B exposed to a supersaturated CaP solution for 14 days are shown in Figure 4. A thick layer of CaP had formed on sample (32) Van Oss, C. J.; Chaudhury, M. K.; Good, R. J. Chem. ReV. 1988, 88, 927. (33) Areva, S.; Peltola, T.; Sa¨ilynoja, E.; Laajalehto, K.; Linden, M.; Rosenholm, J. B. Chem. Mater. 2002, 14, 1614.

Figure 5. XPS spectra of samples A and B after 14 days in supersaturated CaP solution.

A, whereas no CaP was observed on sample B, which also was confirmed by the more surface sensitive XPS analysis (Figure 5). The other peaks in the XPS spectra correspond to adventitious carbon and some chlorine, which is a trace of NaCl from the supersaturated calcium phosphate solution. SEM-EDS analysis of sample A showed that the precipitated Ca/P ratio was about

Calcium Phosphate Formation on TiO2 Coatings

Langmuir, Vol. 22, No. 19, 2006 8213

hand corner to the upper left-hand corner of the window. Additionally, all coatings within the window were mainly of anatase form. No direct correlation between the crystal size and the precipitation was found. The reported Sq and Sds values are taken as a mean from five to 10 AFM images. For some coatings, the standard deviation of the roughness values was quite large, up to 20%. However, the trend is clearly seen from the figure. Thus, we believe that the good correlation observed between CaP precipitation kinetics and the combination of the water contact angle, which is a direct measure of the hydrophilicity of the surface, and surface roughness parameters may be a better explanation for the enhanced surface precipitation kinetics, which make previous assumptions correlating fast precipitation kinetics to a high Lewis base component of the surfaces doubtful.

Conclusion Figure 6. Plot of Sds × Sq as a function of the water contact angle for a series of titania coatings.

1.7, which indicates that the precipitated CaP is hydroxyapatite. This also indicates that the CaP nucleation and growth occurred at a much earlier stage since the initially precipitated CaP phase is amorphous CaHPO4, which is converted to hydroxyapatite with time as a consequence of dissolution-reprecipitation processes.33 Effect of Topography and Surface Energy on Precipitation. In a recent paper, we studied the influence of topography on CaP formation. The best correlation was found between the precipitated amount and the number of local maxima per unit area (Sds parameter) on the substrate.23 A high number (high Sds value) enhanced CaP formation. Here, the aim was to develop the analysis by considering not only the density (peak-to-peak distance) but also the height deviation of the local summits. Indeed, an even better correlation was found when the Sds parameter was multiplied by the Sq parameter. In Figure 6, the water contact angle is plotted against Sds × Sq for a series of titania coatings. The influence of both surface chemistry and topography on CaP formation is demonstrated in the figure. It was only the coatings within the shadowed window that nucleated CaP during a period of 14 days. It is worth mentioning that the Sq parameter alone gave no such correlation. An interesting result is that coatings with a low water contact angle but insufficient roughness precipitated no CaP. Respectively, coatings with a high roughness value but fairly high contact angles were equally inactive. Also, the amount of precipitated CaP increased when moving from the low right-

A study of heterogeneous precipitation of calcium phosphate on TiO2 coatings has been made. Our results suggest that a combination of both surface topography and surface energy controls the precipitation kinetics, and thus, the determination of these values is suggested as a useful toolbox for predicting fouling tendencies of surfaces. A hydrophilic coating is more favorable to initiate the formation of CaP. A high Lewis base parameter of the surface energy has previously often been associated with faster CaP precipitation. We found a linear relationship between the base component and the water contact angle. However, we find it peculiar that the base parameter would become dominant for very hydrophilic surfaces, which should contain a high number of Lewis acidic Ti-OH groups. Thus, we believe that the water contact angle, which is a direct measure of the hydrophilicity of the surface, may be a better explanation for the enhanced surface precipitation kinetics. A high roughness value when combining the RMS roughness parameter Sq with the number of local maxima per unit area parameter Sds enhances CaP formation. It is worth noting that the roughness values are characteristic of the image size. In this study, we analyzed the roughness values from 3 µm × 3 µm AFM images. Roughness parameters on another length scale would not necessarily give the same correlation to the precipitation of CaP, and this will be the subject of forthcoming studies. Acknowledgment. This work was carried out in the Shine Pro project of the Clean Surfaces technology program funded by the Finnish Funding Agency for Technology and Innovation (Tekes). LA060966+