11) L. In-Seop, K. Hyoun-Ee and K. Suk-Young: Surf. Coat. Techn. 131 ... 22) I. S. Lee, C. N. Whang, G. H. Lee, F. Z. Cui and A. Ito: Nucl. Instr. Meth. Phys. Res.
Materials Transactions, Vol. 46, No. 2 (2005) pp. 228 to 235 #2005 The Japan Institute of Metals
Calcium Hydroxyapatite Thin Films on Titanium Substrates Prepared by Ultrasonic Spray Pyrolysis Vukoman Jokanovic and Dragan Uskokovic Institute of Technical Sciences of the Serbian Academy of Sciences and Arts, Knez Mihajlova 35/IV, 11000 Belgrade, Serbia and Montenegro Calciumhydroxyapatite thin films have been prepared by ultrasonic spray pyrolysis deposition on titanium substrate. The surface morphologies of the films formed at different deposition times were analysed by using SEM. The mechanism of the films formation was investigated in relation to the observed film morphology during the first stages of its formation, as well as to its consolidation to the final form. The chemical homogeneity of the synthesized films was determined by using EDS analysis, whereas the phase composition was determined by using XRD measurements and IR spectroscopy. The roughnesses and the thicknesses of the films were analysed by using surface profiler method. The continuous morphology films are obtained only for samples deposited more than 1.5 h. (Received July 22, 2004; Accepted December 10, 2004) Keywords: calciumhydroxyapatite, ultrasonic spray deposition, thin films, titanium substrate
1.
Introduction
Metallic implants based on alloys of cobalt and titanium are widely used in medicine due to their excellent mechanical properties and corrosion stability. Albeit these materials show the biocompatibility due to their inertness in living media, it is desirable to be covered with biologically active materials like hydroxyapatite that guides bone formation along a coating surface of the metallic implants. Realising this approach, a direct bone contact at the implant-bone interface is provided, supported by very good osteoinduction which is the main characteristic of calcium hydroxyapatite. This property might help in averting a long-term bone tissue resorption as well as in enhancing an early bone tissue formation at the surface of titanium alloy coatings in such a way that full weight bearing can be used much sooner after surgery. Acceleration of the bone formation in the initial stages of osteointegration leads to an improved implant fixation. Due to its biocompatibility and very good mechanical characteristics, titanium and its alloys are considered as the most universal materials for permanent implants like, for instance, dental implants of endosseous type.1–5) Hydroxyapatite-coated dental implants are usually recommended for the usage in maxilla areas, where bones are less dense.1–9) Such composites are widely used, because they have a capability to integrate directly into bone tissues through the processes of resorption and/or the processes of formation of a new bone on the surface of implant covered by calcium hydroxyapatite.10–12,14–16) A variety of methods17–27) can be used for the synthesis of calcium hydroxyapatite films, such as electrophoretic deposition,17) electrocrystallization,18) plasma spray deposition,19,20) flame spray deposition,19) magnetic field spattering,21) deposition by aging in fluids like SBF fluids,14) ion beam spattering,22,23) laser beam spattering,24) electron beam evaporation,25) sol-gel,26) aerosol-gel27) and hydrothermal synthesis.8,13) Most of these methods are expensive vacuumbased techniques, with problems related to the film homogeneity, require batch processing and usually high-vapourpressure chemicals or high purity targets as starting materials.
Major requirements related to the development of calcium hydroxyapatite coatings on metallic implants are the preparation of stoichiometric material with exact chemical and phase composition. The deposition of calcium hydroxyapatite on metal substrates by using ultrasonic spray pyrolysis can be considered as one of the methods for overcoming the drawbacks of plasma techniques, related to the problems of the poor control of the stoichiometry and phase composition of the deposited materials.28–35) Not only homogenous composition of coatings, but also a uniform distribution of the deposit along a metallic substrate might be achieved by using ultrasonic spray pyrolysis. Some of the demerits of the vacuum-based methods can be overcome by using wet chemical methods with homogenous solutions, but a control of the film thickness at the nanometer/micrometer scale is still very difficult. The method presented herein has a number of advantages in comparison with other methods, because it enables deposition of extremely homogenous coatings (chemical and phase properties) and control of their thicknesses. 2.
Experimental Method
2.1 Fabrication of hydroxyapatite films Aqueous solution of stoichiometric amounts of Ca and P (Ca/P ¼ 1:67) in concentration of 0.065 M or 0.0162 M (mol/dm3 ) was used as the precursor for the synthesis of calcium hydroxyapatite films. Concentric nitric acid and urea (in equal molar ratio) were added into the solution in order to dissolve the precipitate, while vigorous stirring. An addition of thermo-hydrolyzable urea is necessary in order to obtain optimal pH in tubular furnace. Precipitation of calcium hydroxyapatite before adding nitric acid and urea is avoided by rigorous stirring of the solution of (NH4 )2 HPO4 , which is added dropwise into the solution of Ca(NO3 )2 . Titanium substrates high grade purity (TIKRUTAN RT12, Deutsche Titan GmBH, plate dimensions 10 mm 15 mm 1:5 mm), were treated within two phases: first, by grinding with SiC 600 in order to produce viable surface roughness, after which they were electrochemically treated with 8% solution H3 PO4 during 10 minutes by using voltage
Calcium Hydroxyapatite Thin Films on Titanium Substrates Prepared by Ultrasonic Spray Pyrolysis
of 20 V.2,8–10) Titanium substrates were then placed in a tubular furnace which was previously heated to 600 C. These conditions were selected as appropriate for the performed experiment from the calculations of the evaporation rate of water in aerosol droplets and the flow rate of aerosol droplets in the tubular reactor at given temperature. Ultrasonic atomizer Gapusol 9001, RBI with a transducer working on 1.7 MHz, sprayed the feeding mixture/precursor solution. The aerosol enters into the quartz tube (Heraues Rof 7/50) with interior temperature of 600 C, by using air gas as a carrier (flow rate of 0.66 l/s). The setup scheme can be found in Ref. 28. The substrate was horizontally placed on the bottom verge of the tube, at the distance 38 cm from its inlet. The residence time of the aerosol droplets before its collision with a substrate surface was 35 s. Deposition times were 0.05, 0.5, 1.5 and 5 h, respectively. 2.2 Characterisation Scanning electron microscopy, SEM (JEOL 5300), equipped with a semiautomatic image analyser (Videoplan, Kontron) has been used for analysing the morphology and homogeneity of the substrate coatings. Chemical homogeneity of coatings has been analysed on 4 samples. Treated elements have been: P, Ca, O and Ti. Values for O were very unreliable and were excluded from further consideration. EDS analysis was made by using Si (Bi) detector of X-rays and QX 2000 (Oxford Instruments, UK) connected with scanning electron microscopy and computer multi-channel analyser. X-rays from 0.5–20 keV with 10 eV per each channel were analysed within the performed measurements. Semi-quantitative analysis resulted in an average ratio for all the examined elements at different positions on the hydroxyapatite coating. Software package ZAF (Link Company), showing good results for elements with atomic number larger than 10 was used for comparing the measured intensities. Conditions of recording were: life-time = 100 s and window labels: O K (start: 0.38 keV; end: 0.64 keV; width chains: 14); P K (start: 1.84 keV; end: 2.16 keV; width chains: 17), Ca K (start: 3.50 keV; end: 3.86 keV; width chains: 19), TiK (start: 4.30 keV; end: 4.68 keV, width chains: 20). Roughness and film thickness were analysed by using ALFA STEP 500 Surface Profiler. Tip of a needle of Surface Profiler is sensitive to characteristic changes in the surface morphology. Jump of a needle between the uncovered part of substrate (zero point) and the coated surface defined the film thickness. The X-ray diffraction (XRD) method, (Philips PW 1050), with Cu-K1-2 radiation) was used for the phase analysis of calcium hydroxyapatite and the determination of the crystallite size and lattice parameters. Infrared spectroscopy, IR (PERKIN ELMER 983G) performed on a mixture of scrambled parts of the film and KBr pastille, covering the wave numbers ranging from 400 to 4000 cm1 , was used for the phase analysis as well. 3.
Results and Discussion
3.1
Calculation of the substrate position and designing of the film morphology In order to obtain the ideal Ti substrate sheet position for
229
the maximum rate of the precipitation of the aerosol dropletsthat is precipitation that takes places over its whole volume almost immediately after the contact of the droplet with the substrate surface-velocity of the droplet flight must have been adjusted to the rate of its precipitation. The place in the furnace tube on which an aerosol droplet, due to its evaporation during its flight to it, reaches density that corresponds to the complete saturation of the precursor solution within the droplet, was marked as an ideal position for placing the substrate sheet and for deposition of calciumhydroxyapatite film. Equation A·1, given in Appendix,32) was used for determination of the precipitation rate that is equal to the rate of solvent evaporation. All of the values of the input variables must have been defined prior to that. Some of these quantities, such as the diffusion constant of the solvent in the gas phase (Dv ), were experimentally determined on the basis of the measurements of the distance at which maximal rate of film deposition and maximal rate of droplet precipitation occur over Ti substrate under the given rate of aerosol flow. The calculation was performed in such a way so that at first, distance from the tube inlet to the place of the largest material deposition within the tube was determined, was presupposed that this time should be approximately equal to the time of its precipitation, which as again equal to the typical time constant for diffusion of the solvent in the gas phase tdg . By following equation Dv ¼ R2 =tdg , where R is average radius of the aerosol droplets (R ¼ 1:6 mm), determined according the eq. A·2 (given in Appendix),28,29,32) the diffusivity of the solvent in the gas phase was obtained: Dv ¼ 1:2 1012 m2 /s. Since the droplet temperature during its evaporation was 100 C, partial vapour pressure in tube p1 ¼ 78:48 kPa, and on the droplet 98.1 kPa, molecular weight of the precursors 1 kg/mol, the evaporation rate was calculated as 1:2 1015 kg/s. Since an average droplet has 0:69 1015 kg in weight, it means that 14.8 s is needed for the evaporation of such droplet. During that time, a droplet moving with velocity of 0.11 m/s transverses the length of 38 cm. On the other hand, taking the ratio Dcr =Dr ¼ 90,32) (relative ratio of water vapour diffusivity within the gas phase and within an aerosol droplet (Nesic-Vodnik’s equation, Appendix 2, eq. A·3)32) under the assumptions that takes value 0.2 and 0.3 Rc (Rc is the critical droplet radius in the moment of precipitation) and that due to the existence of fine cylindrical pores, that did not come to the closing of the precipitated droplet’s surface during the drying as a valid one, and by placing the diffusion coefficients of water vapour in air Dv ¼ 1:2 1012 m2 /s in the given equation, a value for the diffusion coefficient in the precipitated layer Dcr ¼ 1:1 1010 m2 /s is obtained. Finally, by taking d ¼ 1 and 1 ¼ 0:8 (d is the relative weight concentration of the solvent vapour on the aerosol within furnace, whereby 1 is the relative weight concentration of solvent vapour within the furnace tube), it follows that the drying rate is between 1:8 1017 kg/s and 1 1017 kg/s, which corresponds to the drying time of 2.8–4.8 s. This implies that
230
V. Jokanovic and D. Uskokovic
traversed path of the droplet is 28–49 cm. Of course, with an advance in the drying process, the drying rate decreases. Considering the fact that the weight of the aerosol droplet depends on its density (which is directly dependent on the concentration of precursors comprised within), it becomes obvious that the time and the distance of the flight corresponding to the maximum concentration of precursors comprised within the aerosol droplets immediately before the deposition takes place, is dependent on the precursors concentration in the moment of the droplet formation as well. For the precursors’ concentrations of 0.065 M and 0.0162 M, as calculated in relation to hydroxyapatite, the calculated ideal positions for placing Ti sheets were: 38 and 40 cm from the position of aerosol input into tube, respectively. Therefore, for four times lower concentration of precursors, sheet that the deposition of hydroxyapatite film takes place upon, should, according to the calculation, be repositioned for 2 cm only.
Due to the small variations of the ideal distance of substrate for different concentration of precursors, it was possible to keep position of Ti substrate the some for both concentrations, without changing the aerosol flighting rate. If the sheet of Ti substrate is being placed on such a position, then viable conditions for the deposition will be set. Under such conditions, the aerosol droplets of maximal density will be hitting the target of substrate surface, adhering to it, and therefore, due to further acceleration of the process of precipitation conditioned by much higher sheet’s temperature comparing to the droplet temperature at the moment, the aerosol droplets will be having considerably limited smearing over the substrate’s surface and will be solidified in the forms of partially deformed spheres. The morphologies of calcium hydroxyapatite film after the deposition times of 1.5 h, as obtained from two solutions with different precursors’ concentrations 0.0162 M and 0.065 M (Figs. 1(a), (b) and 2(c)) affirm such a conclusion.
(b)
(a)
Fig. 1 Surface morphology of calcium hydroxyapatite films deposited 1.5 h for different precursor concentrations: (a) c ¼ 0:0162 M, (b) c ¼ 0:065 M.
(a)
(b)
(c)
(d)
Fig. 2
Surface morphology of the film deposited for different time for c ¼ 0:065 M (a) 0.05 h, (b) 0.5 h, (c) 1.5 h and (d) 5 h.
Calcium Hydroxyapatite Thin Films on Titanium Substrates Prepared by Ultrasonic Spray Pyrolysis
From the presented micrographs, the agglomerates of very fine particles of spherical-spheroid shapes can be observed, which clearly shows that due to high density of the aerosol droplets in the moment of their contact with the substrate’s surface, the particles that were formed by their solidifications processes were of similar shapes, as well as that the film morphologies were similar, independent of the precursors concentration. 3.2
Surface morphology and mechanism of the film forming Analyses of the depth, morphology, particle size distribution and homogeneity of the chemical structure of calcium hydroxyapatite coatings on titanium substrates have shown that the coating already formed after 1.5 h. Observation of the deposit in periods from 0.05–0.5 h has shown that at first very small islands, corresponding to the size of the aerosol droplets wetted on the surface of substrate were formed. The islands were of irregular shapes, and thus were formed by the coalescence of grains, whereby the bigger grains grew on the account of the disappearance of the smaller ones. By using calculations for average droplet size (over 100 nm, corresponding to almost every real case) presented in the previous references,30–34) it was shown that all the droplets were of diameters much larger than the dimensions of the nuclei in embryo stage. Due to this fact, the classical initial step of coating forming was skipped and the initial stage of the deposition process was coalescence. At this stage, as is shown in Fig. 2(a), big and small islands/ clusters appeared together, isolated from each other and/or locally joined forming larger islands of irregular shapes. By prolonging the deposition process, further forming of the bigger islands by coalescence of smaller islands into bigger ones continued. This was caused by changing their morphology in agreement with the shape limitations given by the geometry of the coalescing elements/islands (Fig. 2(b)). As can be seen from the Fig. 2(b), channels among the islands are clearly present. At that stage, they were not occupied by intensive secondary nucleation yet. Calcium hydroxyapatite particles in form of ideal spheres size from 1 to 2 mm were noticed (insert on the top substrate surface) at the surface of substrate. The concentration of the needle particles was increasing at the positions of the deepest holes/channels, due to the opportunity of growth in certain preferred directions in the places on substrate with the largest voids. The needle particles had up to 5 mm in length. After 1.5 h the surface of titanium substrate was completely covered with calcium hydroxyapatite, as can be seen from Fig. 2(c). Particles of calcium hydroxyapatite were of submicron dimensions. If calciumhydroxyapatite particles solidified without previous wetting (insert in the Fig. 2(c)) dimensions of aerosol drops and the solution concentration within in the moment of precipitation determines the dimensions of the obtained particles. Spatial dimensions of crystallites were from 0.5–1 mm. The particles had different shapes: polygonal, rounded and needle-shaped. Some needleshaped particles almost 5 mm in diameter were found on places with low concentration of individual particles. The particles made by rapid coalescence of large number of small
231
particles were observed in the areas of larger concentration of individual particles. These particles were of polygonal shapes with the size up to 45 mm. Continuous formation of the layers of unequally sized particles over the whole surface of substrate was found on the film deposited with higher thickness (5 h of deposition time — Fig. 2(d)). Particles were agglomerated in about 10 mm characteristically sized blocks containing more layers and of the petal-like structures. The largest part of the coating was very homogenous with particles of up to 1 mm in size completely covering the substrate surface. 3.3 Film homogeneity Coating homogeneity was analysed by using EDS (microchemical analysis) and the obtained results are given in the Table 1. Based on the shown results, it is obvious that in the first layers (obtained in the initial phase of deposition, after 0.05 h), calcium hydroxyapatite deficient of calcium was obtained. In the initial stage of the film growth small islands clearly visible in Fig. 1 (sample 1) were found to have high contents of Ti (58.2–80.5 mass%) and relatively low concentrations of Ca (12.3–30.7 mass%) and P (7.1–14 mass%). The EDS analysis revealed in sample 2, the calcium hydroxyapatite layer slightly sufficient with Ca (Ca/P ¼ 1:70{1:73), whereby for the samples 3 and 4 calcium hydroxyapatite layers were slightly deficient with Ca. Based on the obtained results, it can be noticed that calcium hydroxyapatite in the starting phase of layer forming is slightly deficient of Ca. During this phase the formation of the needle-shaped particles becomes possible. These needleshaped calcium hydroxyapatite particles do not contain stoichiometric content of Ca and P. The starting phase is also related to the striving of the system towards decreasing of the surface energy. In general case, the oxide and hydrated layer of Ti-TiO2 and TiO2x OHx , respectively, might form on the surface of the titanium alloy at the temperature of thermal treatment used herein. In reaction with liquid precursor droplets these layers should cause a charging of the titanium surface and the adsorption of calcium ions by their hydroxyl groups.9) Relating to other investigations, it might be proposed that the hydrated TiO2x OHx surface probably causes the formation of new phases like CaTiO3 . Secondly, the abundant Ca2þ ions on the surface might cause an easy PO4 3 adsorption, followed by the formation of calcium phosphate nuclei. This might be a possible explanation for deficient calcium in the first layers on the titanium alloy substrate. Since larger grains have lower surface energy and are more stable, calcium and phosphate ions in the droplets in contact with TiO2x OHx layers should more easily precipitate and crystallize in such forms. In addition, the surface of titanium alloy with a number of calcium ions should be charged positively and adsorb negative ions, like hydroxyl groups. Also, the surface of TiO2x OHx may attract Ca ions and combine phosphate ions in TiO2x OHx HPO4 by covalent bonding.9) A slight decrease of Ca can be noticed during the next phases of the experiment. The sample 2 represents the
232
V. Jokanovic and D. Uskokovic Table 1
Specimen
Time of deposition, td (h)
EDS analysis of calcium hydroxyapatite films for c ¼ 0:065 M. Chemical composition (mass%)
Chosen snapshot Ti
1.
0.05
2.
0.5
3.
1.5
4.
5
Ca
Ca/P (mass%)
P
Ca/P (at.%)
1.
80.52
12.39
7.09
1.74
1.35
2.
73.14
17.65
9.21
1.91
1.48
3.
58.21
30.75
14.04
2.19
1.70
1.
31.95
46.73
21.32
2.19
1.70
2. 1.
29.03 14.6
49.00 56.7
21.98 28.7
2.29 1.97
1.73 1.58
2.
13.35
57.82
28.83
2.0
1.56
3.
4.02
64.93
31.24
2.07
1.61
4.
2.23
65.51
32.27
2.03
1.55
5.
1.99
66.54
31.47
2.11
1.61
1.
—
66.94
33.06
2.02
1.54
2.
—
67.12
32.88
2.04
1.56
3. 4.
— —
68.07 68.65
31.93 31.35
2.04 2.1
1.56 1.61
5.
—
69.60
30.40
2.29
1.75
6.
—
66.90
33.1
2.02
1.54
exception from this tendency (The content of the film was at this point slightly sufficient with Ca). However, such small variations in the composition might be neglected since they fall into the range of the experimental error of the measurement. In addition to this, since each aerosol droplet has the same chemical composition set to the stoichiometric formula of calcium hydroxyapatite, the spatial fluctuations of the composition, if not neglected, must only be related to the mechanism of the surface reaction between substrate and droplets. In accordance to being said, due to neglecting influence of the substrate surface on the mechanism of the film formation, the chemical composition of the deposited calcium hydroxyapatite coincided with the stoichiometrical formula in the final stage of the deposition (after 5 h). At this stage the grains of calciumhydroxyapatite comprising synthesised film with the average thickness between 5 and 12 mm had preferentially polygonal shape. 3.4 Surface roughness and film thickness The results of the measurements of surface roughnesses and film thicknesses of the synthesized calcium hydroxyapatite by using surface profiler method are given in the Table 2. The method used consisted of two steps: first, the roughness of the substrates was measured, and second, the film thickness was obtained as the difference between the previously measured substrate roughness and the film thickness after the deposition. From the Table 2, it can be seen that the film roughness decreased with an increase of the film thickness. At the extremely thin films (films deposited in very short times), roughness was identical with the substrate roughness. At the layers deposited in some longer times (0.5 h), the roughness was almost equal to the substrate roughness. This was due to the shape of the surface (in agreement with the EDS snapshots), that had a characteristic calcium hydroxyapatite islands separated by empty channels (as visible from
Table 2 Roughness and thickness of calcium hydroxyapatite films for c ¼ 0:065 M. Specimen
Time of deposition, td (h)
Film thickness, d (mm)
Average roughness, h (mm)
Maximum roughness, hm (mm)
1
0.05
—
—
—
2
0.5
—
—
—
3
1.5
5
10
15
4
5
12
3
8
the Fig. 2(b)). The processes of secondary nucleation were still not intensive enough at this stage of deposition. Despite that, partial fusion of the islands and the forming of a complete continuous layer were noticed. A coating with significant roughness was obtained with the sample of calcium hydroxyapatite deposited during 1.5 h, because islands, still present in the sample, did not merge yet, and after this time it was only the beginning of the process of the secondary nucleation of the film. Average roughness of these samples was 10 mm and the maximum roughness was 15 mm. In this case, the average film thickness was 5 mm. After the deposition in 5 h, the channels between islands were filled with material, which was deposited by the secondary nucleation. The average roughness of this layer was 3 mm, the maximum roughness 8 mm and the average film thickness 12 mm. 3.5 Phase analysis Phase analysis of the films was performed by using XRD and IR spectroscopy. The resulting XRD patterns presented in Fig. 3, show all the characteristic peaks correspondent to the obtained calcium hydroxyapatite (space group P63/m ). The relative intensities for all the characteristic diffraction maxima (211) I100 , (112) I60 , (300) I60 , (002) I40 , (213) I40 , (222) I30 , (202) I25 , (004) I20 and (210) I18 correspond to the relative intensities of diffraction maxima found in a standard
Calcium Hydroxyapatite Thin Films on Titanium Substrates Prepared by Ultrasonic Spray Pyrolysis
(211)
100
80
transmittance
(004)
(321)
(222)
(210)
40
(202)
(300)(112)
(002)
4
60
Intensity
233
3 2
20
1 0
10
20
30
40
50
60
4000
2θ
3500
3000
2500
2000
cm Fig. 3 X-ray diffraction of calcium hydroxyapatite film obtained by spray pyrolysis after 5 h of the deposition.
data (JCPDS No. 9-432). The lattice parameters of the obtained calcium hydroxyapatite were calculated by using Rietveld refinement method and Kolarie program package for all 9 major reflections. The obtained values were: c ¼ 0:6885 nm and a ¼ 0:9460 nm. Given values were close to lattice parameters found in literature (c ¼ 0:6884 nm and a ¼ 0:9368 nm).36,37) It was not possible to resolve calcium carbonate peaks and confirm its presence on basis of the XRD analysis only. Slight shifting of the characteristic diffraction picks corresponding to the planes (004) and (300) that might be caused by the presence of carbonate groups was not large enough. Shifting of these picks might be caused the composition nonstoichiometry as well. Therefore, IR spectroscopy was used to investigate the possible presence of the carbonate phase. The crystals of calcium hydroxyapatite were, however, the main constitutive phase within the deposited films. As can be seen from the Fig. 2(a), the deposition processes first took place at a certain distance from the inlet of the furnace tube where the precursors enclosed within droplets transformed into solid particles, whereby other droplets having higher viscosity and containing already formed sub-elements networked with substrate were networked with high speed at the moments of collision. The latter form of crystallites became nuclei for the further processes of crystallisation and the grain growth inside the film. IR spectroscopy was used as well in order to compare IR spectra for the samples synthesized by using different precursors concentrations (0.0162 M and 0.065 M), as well as for the identification of the type of the deposited carbonate hydroxyapatite. IR spectra of hydroxyapatite films, shown in Fig. 4, have the same structures for both used concentrations of precursors 0.0162 M (1-deposition time 1.5 h and 3-deposition time 5 h) and 0.065 M (2-deposition time 1.5 h and 4-deposition time 5 h) as well as for both deposition times. By comparing the IR spectra of the films deposited by using different concentrations of the precursors and different deposition times, the following results were derived: (1) Stretching OH vibration, originating from OH ions incorporated in the calciumhydroxyapatite crystal lattice, that corresponds to the band at 3540 cm1 , for the
1500
1000
500
-1
Fig. 4 IR spectrum of calcium hydroxyapatite films obtained by spray pyrolysis method for different precursor concentrations (0.0162 and 0.065 M) and different time of the deposition: 0.0162 M (curve 1; c ¼ 0:0162 M, t ¼ 1:5 h), (curve 2; c ¼ 0:065 M, t ¼ 1:5 h), (curve 3; c ¼ 0:0162 M, t ¼ 5 h) and (curve 4; c ¼ 0:065 M; t ¼ 5 h).
(2) (3)
(4)
(5) (6)
sample obtained by using precursor concentration c ¼ 0:0162 M and the deposition time t ¼ 1:5 h (Fig. 4.-1), whereby all the other samples exhibit the same band at 3561 cm1 . It is obvious that these values correspond to its minimum only, and that the same band covers stretching vibrations that OH group is typical of, corresponding to wave numbers 3440–3430 cm1 and 3744 cm1 that define vibrations of Hþ ion within associated water molecules at the surface of hydroxyapatite molecules. The band corresponding to bending OH vibration at 1635 cm1 has the same shape for all samples. The doublet band between 1400 and 1450 cm1 corresponds to the stretching CO3 2 vibrations that all the samples are typical of. This vibration is an evidence of the presence of carbonate hydroxyapatite in all the deposited samples. The bands corresponding to asymmetrical stretching PO4 3 vibrations exhibit certain differences among different samples. For the sample corresponding to c ¼ 0:0162 M and t ¼ 1:5 h, the band is found at 1101– 1022 cm1 ; for the sample corresponding to the same precursor concentration, but t ¼ 5 h, the band covers the range of 1115–1008 cm1 ; for the samples corresponding to precursor concentrations of 0.065 M, the band is found at 1108–1022 and 1108–1015 cm1 . Therefore, there is a tendency toward narrower bands when the deposition time becomes lower, although the differences are rather small and can easily be neglected. The bands corresponding to the stretching CO3 2 vibration are the same for all the samples. The band at 630 cm1 , that corresponds to the liberation OH vibrations remains at the same wave number for all samples. Its well defined position in the IR spectra confirms that all the synthesized samples comprise deposited carbonate hydroxyapatite of B-type as well as that within all samples preferential replacement of PO4 3 ions by CO3 2 ions occurs.38,39)
234
4.
V. Jokanovic and D. Uskokovic
Conclusion
The ideal position in furnace tube—the one at which the density of aerosol droplets that hit the substrate is maximal, and thereby aerosol droplets preferentially solidify in the forms close to spherical—was evaluated for both concentrations of the starting solution. Films deposited for very short times (0.05–0.5 h) shown islands of calcium hydroxyapatite, obtained by the precipitation and the drying of aerosol droplets on the substrate surface. The spray pyrolysed sample did not get completely continuous film of calcium hydroxyapatite on titanium substrate after 1.5 h as well. However, the completely monolithic films of calcium hydroxyapatite were obtained after the treatment of 5 h. Roughness of calcium hydroxyapatite had the lowest value for the films synthesised during the longest deposition times (5 h): average roughness was 3 mm and maximal roughness 8 mm. Roughness of the films deposited within the periods of less than 1.5 h were equal to the substrate roughness. The films were of continuous morphology only for the samples deposited for 1.5 h and 5 h with their thicknesses of 5 mm and 12 mm, respectively. The composition of the synthesized thin films was nearly homogenous and close to the values corresponding to stoichiometrical formula calcium hydroxyapatite as obtained from EDS analysis. X-ray analysis and IR spectroscopy proved that carbonate calcium hydroxyapatite was deposited. IR spectroscopic analyses have proved that the phase composition of the film derived from both, different precursor solutions and both performed deposition times, is approximately the same. In both cases, carbonate hydroxyapatite of B type was detected as the deposited phase. Acknowledgements This work was supported by the Ministry of Science and Technology of the Republic of Serbia through the project No. 1431 — ‘‘Molecular designing of monolithic and composite materials’’. We would like to thank Prof. G. Borchardt and B. Jokanovic, Institut fur Mettalurgie, Techniche Universitat Clausthal, Termochemie und Mikrokinetik and M. Miljkovic and Z. Nedic for their support and cooperation in SEM, EDS and Surface profiler analysis and IR analysis. The authors thank Prof. M. Zlatanovic for his most kind and helpful comments.
REFERENCES 1) R. Z. Le Geros: Calcium phosphates in oral biology and medicine, (New York University, College of Dentistry, 1991) 114–121. 2) G. C. Engelmayr, E. Rabkin, F. W. H. Sutherland, F. J. Schoen, J. E. Mayer and M. M. Sacks: Biomaterials 26 (2005) 175–187. 3) W. R. Lacefield: Hydroxyapatite coatings, In: P. Ducheyne JE Lemons, editors. Bioceramics: Materials characteristics versus in vivo behaviour, 523, (Annals of the New York Academy of Sciences, 1988) 72–80. 4) H. M. Kim, F. Miyaji, T. Kokubo and T. Nakamura: J. Biomed. Mater. Res. 38 (1997) 121–127.
5) L. L. Hench: J. Am. Ceram. Soc. 81 (1998) 1705–1728. 6) L. L. Hench: J. Am. Ceram. Soc. 74 (1991) 1487–1510. 7) R. Z. LeGeros and J. P. LeGeros: Dense hydroxyapatite. In. L. L. Hench and J. Wilson editors, An introduction to bioceramics, Adv. Ser. Ceram. 1 (World Scientific Publishing Co. Pte. Ltd., London, Hong Kong, Singapore, 1998) 139–180. 8) M. C. Andrade, M. R. Travers Filgueiras and T. Ogasawara: J. Europ. Ceram. Soc. 22 (2002) 505–510. 9) B. Feng, J. Y. Chen, S. K. Qi, L. He, J. Z. Zhao and X. D. Zhang: Biomaterials 23 (2002) 173–179. 10) W. Q. Yan, T. Nakamura, K. Kawanabe, S. Nishigochi, M. Oka and T. Kokubo: Biomaterials 18 (1997) 1185–1190. 11) L. In-Seop, K. Hyoun-Ee and K. Suk-Young: Surf. Coat. Techn. 131 (2000) 181–186. 12) M. Aizawa, T. Hanazawa and K. Itanati: Key Eng. Mat. 191&192 (2000) 103–106. 13) H. Yao-Shan, E. Chang and L. Hok-Shin: Ceram. Int. 24 (1998) 7–12. 14) J. L. Ong, G. N. Raikar and T. M. Smoot: Biomaterials 18 (1997) 1271– 1275. 15) L. C. Lucas, W. R. Lacefield, J. L. Ong and R. Y. Whitehead: Coll. Surf. A: Physchem. Eng. Asp. 77 (1993) 141–147. 16) P. Royer and C. Rey: Surf. Coat. Techn. 45 (1991) 171–177. 17) A. Stoch, A. Broek, G. Kmita, J. Stoch, W. Jastrzbski and A. Rakowska: J. Mol. Struct. 596 (2001) 191–200. 18) M. Shirkhanzadeh: Nanostruct. Mater. 4 (1994) 677–684. 19) L.-G. Yu, K. A. Khor, H. Li and P. Cheang: Biomaterials 24 (2003) 2695–2705. 20) L. In-Seop, P. Jong-Chul and L. Young-Hee: Surf. Coat. Techn. 171 (2002) 252–256. 21) G. Bikulius, V. Burokas, A. Martuien and E. Matulionis: Surf. Coat. Techn. 172 (2003) 139–143. 22) I. S. Lee, C. N. Whang, G. H. Lee, F. Z. Cui and A. Ito: Nucl. Instr. Meth. Phys. Res. Sec. B: Beam Interaction with Materials and Atoms 206 (2003) 522–526. 23) C. X. Wang, Z. Q. Chen, L. M. Guan, M. Wang, Z. Y. Liu and P. L. Wang: Nucl. Instr. Meth. Phys. Res. Sec. B: Beam Interactions with Materials and Atoms 179 (2001) 364–372. 24) J. L. Arias, M. B. Mayor, J. Pou, Y. Leng, B. Leo´n and M. Pe´rez-Amor: Biomaterials 24 (2003) 3403–3408. 25) L. In-Seop, K. Dong-Hwan, K. Hyoun-Ee, J. Young-Chul and H. Chong-Hyun: Biomaterials 23 (2002) 609–615. 26) H. Ming-Fa, P. Li-Hsiang and C. Tsung-Shune: Mater. Chem. Phys. 74 (2002) 245–250. 27) M. Manso-Silva´n, M. Langlet, C. Jime´nez, M. Ferna´ndez and J. M. Martı´nez-Duart: J. Europ. Ceram. Soc. 23 (2003) 243–246. 28) V. Jokanovic´, Dj. Janac´kovic´, A. M. Spasic´ and D. Uskokovic´: Mater. Trans. JIM 37 (1996) 627–635. 29) V. Jokanovic´, Dj. Janac´kovic´ and D. Uskokovic´: Ultrason. Sonochem. 6 (1999) 157–169. 30) V. Jokanovic´, Dj. Janac´kovic´ and D. Uskokovic´: J. Nanostruct. Mater. 12 (1999) 349–352. 31) Dj. Janac´kovic´, V. Jokanovic´, Lj. Kostic´-Gvozdenovic´, Lj. Zivkovic´ and D. Uskokovic´: J. Mater. Res. 11 (1996) 1706–1716. 32) G. L. Messing, S. C. Zhang and G. V. Jayanthi: J. Am. Ceram. Soc. 76 (1993) 2707–2726. 33) G. V. Jayanthi, S. C. Zhang and G. L. Messing: J. Aerosol Sci. Technol. 19 (1993) 478–490. 34) V. Jokanovic, I. Nikcevic, B. Dacic and D. Uskokovic: J. Cer. Pocess. Research 5 (2004) 157–162. 35) K. Okuyima and W. L. Lenggero: Chem. Eng. Sci. 58 (2003) 537–547. 36) I. Nikcevic, V. Jokanovic, M. Mitric, Z. Nedic, D. Makovec and D. Uskokovic: J. Sol. St. Chem. 177 (2004) 2565–2574. 37) JCPDS File No. 9-432 (HA), International Center for Diffraction Data. 38) R. A. Nyquist and R. O. Kagel: Infrared Spectra of Inorganic Compounds, (3800–45 cm1 ), (Academic Press, New York and London, 1971) 493. 39) K. Nakamoto: Infrared and Raman Spectra of Inorganic and Coordination Compounds, (JohnWiley & Sons, New York, Chichester, Brisbane, Toronto, 1978).
Calcium Hydroxyapatite Thin Films on Titanium Substrates Prepared by Ultrasonic Spray Pyrolysis
where d ¼ 2R average aerosol droplet diameter, surface tension of precursor, f frequency of ultrasonic atomizer and precursor density. The drying rate of aerosol droplet is given by:32)
Appendix The rate of droplet evaporation is given by:32) dm 4RDv M p1 pd ¼ ; dt Rg T1 Td
235
ðA:1Þ
where m is mass of droplet, t evaporation time, p1 and T1 ambient vapour pressure and temperature of the reactor, pd and Td vapour pressure and temperature at the droplet surface, M molecular mass of the gas, Rg gas constant, R average droplet radius and Dv the diffusivity of solvent vapour in air inside of furnace tube. The average aerosol droplet diameter is given as follows:28,29,32) 1 3 d¼ ; ðA:2Þ f 2
dm ¼ dt
4Rc Dv ½d 1 ; Dv 1þ Dcr Rc
ðA:3Þ
where dm=dt is a drying rate, Rc droplet radius at time of precipitation, Dcr the diffusivity of vapour through the precipitate layer, the thickness of the crust, 1 i d the mass concentration of solvent bulk and surface vapour, respectively and Dv the diffusivity of solvent in air inside of furnace tube.
