Sol-derived Hydroxyapatite Ddip-coating of a Porous Ti6Al4V Powder Compact

A sintered porous Ti6Al4V powder compact with a mean pore size of 63 μm and an average porosity of 37±1% was dip-coated at soaking times varying between 1and 5-minute using a sol-derived calcium Hydroxyapatite (HA) powder. The coated compacts were heat-treated at 840 oC. The coating thickness was found to increase with increasing soaking time, from 1.87 μm at 1-minute soaking to 9 μm at 5-minute soaking on the average. It was shown that at increasing soaking times, the originally open pores started to close, while at low soaking times the Ti6Al4V particles were partially coated. The coating layer was shown to be nano porous and the depth of coating was observed to be relatively shallow: only few particles near the compact surface were HA-coated.


Introduction
The application of porous components of bio-compatible Ti and Ti 6 Al 4 V alloy is to provide bet-ter interaction with bone due to the higher degree of bone in-growth and body fluid transport through three-dimensional interconnected array of pores [1], leading to improved implant fixation. Furthermore, the relatively low elastic moduli of porous metals as compared with those of bulk metals reduce the extent of stress shielding causing implant loosening [2]. The sintering powder compacts and the space holder methods were previously applied successfu-lly to produce biocompatible open-cell Ti and Ti 6 A-l 4 V powder compacts and foams, respectively [3][4][5][6][7]. Both methods, the sintering powder compacts and the space holder methods, allow the direct near netshape fabrications of porous implant components having elastic modulus values comparable with that of natural bone, relatively homogeneous pore struc-tures and high level of porosities (30-70%) [3,5,6]. The porosity level of sintered Ti compacts suitable for the bone replacement was found around 30%, which was in accord with the proposed optimal por-osity for the ingrowths of new-bone tissues [5,8,9]. However, the compressive yield strength of sintered Ti compacts was shown to be lower than that of the human cortical bone due to the relatively low yield strength of Ti powder used [5]. The use biocompati-ble stronger Ti alloy powders such as Ti 6 Al 4 V, which is widely used in biomedical applications, increased the yield strength of powder compacts [3].
Ti is known to be a bioinert material and it ca-nnot directly bond to living bone directly [10]. Ca-lcium Phosphate (CaP) coating of porous surfaces including Ti mesh [11,12] and the sintered powder (beads) surface coatings [13,14] has beneficial effe-cts on the bone-generating properties, improving the implant fixation to bone. The aim of this study is to investigate the effect of dip-coating soaking time on the hydroxyapatite (HA) coating layer thickness of a Ti 6 Al 4 V sintered powder compact. The selected sol-derived powder dip-coating method is relatively simple, inexpensive and suitable for the mass prod-uction. Ti 6 Al 4 V powder compacts were prepared in house using a powder metallurgical-sintering route and originally developed for the hard tissue replace-ment applications such as spinal cages used in spinal surgery.

Materials and Method
The sintered powder compacts were prepared using atomized spherical Ti 6 Al 4 V alloy powders *corresponding author. E-mail: mustafaguden@iyte.edu.tr (Phelly Materials Company), complied with ASTM 1580-1 standard [15]. The powder particle size ra-nged between 100 and 200 µm with a mean particle size of 157 µm. The green powder compacts were compacted uniaxially at room temperature at 400 MPa, using a polyvinyl alcohol solution (5-10% by volume) as binder. The sintering of the compacts was conducted at 1200°C for 2h under high purity Ar atmosphere. The details of the compact prepara-tion are given in elsewhere [3]. The compact cont-ained 100% three-dimensional interconnected pores with pore sizes ranging between 16 and 200 µm, a mean pore size of 63 µm and an average porosity of 37±1%. The sintering of the needle-like a-phase (acicular alpha) of the starting powder resulted in the formation of so-called Widmanstätten microstr-ucture (α+β)   The elastic modulus and compressive yield strength varied between 4 and 6 GPa and 155 and 175 MPa, respectively [3], satisfying the streng-th and modulus requirements for certain hard-tissue replacements such as spinal cages used for bone ingrowth in patients with spinal trauma, tumors and degenerative diseases.
The HA powder for the dip coating solution was prepared according to the protocol developed by Maviş and Taş [16], which was essentially a mod-ified route of chemically precipitated calcium HA powder processing originally proposed by Hayek and Newesely in 1963 [17]. A 3 ml aliquot of 0.1 g/l methyl cellulose solution (99% pure, Sigma, St. Louis, MO) acting as dispersant was mixed with 1440 ml of deionized water. Thereafter, 0.152 mol of Ca(NO 3 ) 2 . 4H 2 O (99% pure, Merck, Darmsta-dt, Germany) and 0.091 mol of (NH 4 ) 2 HPO 4 (99% pure, Merck) were dissolved with above solution se-quentially (Ca/P=1.67). This mixing step resulted in formation of an opaque solution mixture. Finally, 115 ml of 28-30 vol% NH 4 OH (99% pure, Perform-ans, Turkey) was added to the final opaque solution and pH of the solution was kept around 10. Then, the solution was heated to the temperature range of 60-70°C on a hot plate for 90 min. The precipitates formed were vacuum filtered using Buchner funn-el, washed with distilled water for several times and then dried in an oven at 90°C for 6 h. Above proce-dure of HA powder processing was found to yield a single phase HA at relatively high temperatures, up to 1200-1300 °C [16,18].
Dip-coating of the cylindrical compacts, ~15 mm in diameter and ~10 mm in height, were performed in a dip coater at soaking times from 1 to 5-minute with one minute interval (dipping and removal rate of 100 mm/min). The compacts were heat-treated after dip-coating under Ar atmosphere (to prevent the oxidation of the Ti64 particles) at a temperature of 840 °C (2 h of peak soaking time), with a heating and cooling rate of 2 °C/min. The microscopic ana-lysis on the coated compact samples was performed using a Nikon Eclipse L150 optical microscope and a Philips XL30-SFEG scanning electron microsco-pe (SEM) with an Energy Dispersive X-ray (EDX). The polished cross-sections of epoxy-mounted coa-ted compact samples were etched with Kroll's rea-gent (3 cm 3 of HF and 6 cm 3 of HNO 3 in 100 ml of H 2 O) and used for the determination of the thickne-ss and depth of the coating. A Philips X'pert X-ray diffraction (XRD) instrument with Cu-Ka radiation was used for the phase determination.

Results
The synthesized HA precursor powder used for dip-coating solution yielded stable dip-coating sol-ution against the sedimentation as previously noted [16]. Initially, the powder was heat treated at various temperatures until 1200 °C for 2 h in order to dete-rmine the effect of temperature on the crystallinity of the powder. XRD spectra of the sol-derived HA powder as function of heat treatment temperature is shown in Figure 1. The phase composition of solderived powder was identified as low crystalline HA at 90 °C. The heat-treatment at higher temperatur-es however improves the crystallinty of the powd-er and the powder remains single-phase until abo-ut relatively high temperatures, 1200 °C, as noted previously in refs. [16,18]. The powder XRD data shown in Figure 1 are agreed closely with the hexa-gonal structure of calcium HA (JCPDS-4-932). The hexagonal unit cell parameters of the powder heattreated at 840 °C are calculated as; a=0.949 nm and c= 0.695 nm. The HA powder coated compacts was however heat treated at a temperature of 840 °C, be-low the α+β phase transition temperature [19]. The transmission electron micrograph of the synthesized sol-derived HA powder heat-treated at 840 °C is sh-own in Figure 2. The particles in Figure 2 are mostly elliptical in shape with the sizes of 100-200 nm.
The schematic of a coated powder compact sam-ple is shown in Figure 3(a). In the SEM analysis of the coating layer, the pictures were taken from the cylindrical compact sample surface from the region A of Figure 3 Figures 3(c) and (d). At 4-minute soaking, the HA coating starts to fill the originally open pores between sintered particles (Fi-gure 3(e)). The pores are almost completely closed when the soaking time increases to 5-minute ( Figure  3(f)). Large cracks are also seen on the coating la-yer particularly near the sintering necks as marked by arrow in Figure 3(f). In all dip-coated compact samples, the coating started from the sintering necks  of the particles and proceeded to the uncoated regi-ons. Therefore; as stated earlier, the coating layer is thicker at the particle sintering necks as marked with white arrows in Figure 4, a similar observation was previously noted in a Ti-bead coated Ti implant [20]. Accompanying extensive microscopic observations on the sintering necks revealed that the coating thi-ckness differences between the sintering necks and the particle surface and the extent of coating cracks at the sintering necks decreased as the soaking time increased from 1 to 3-minute.
The crystal size of the heat-treated powder was microscopically determined from the coated and heat-treated compact samples. The SEM micrograph of a coated (3-minute soaking) and heat-treated com-pact sample surface is shown in Figure 5(a). The coa-ting layer in Figure 5(a) clearly proves the formation of a nanoporous coating layer on Ti 6 Al 4 V particles following the heat treatment process at 840 °C. The pore size of the coating layer is comparable with HA particle size, in 100 nm's sizes. The EDX analysis on the coating layer after heat treatment at 840 °C gives a Ca and P ratio of 1.67 as depicted in Figure 5(b).
The variation of the thicknesses of the coating layer as function soaking time was determined mi-croscopically from Ti 6 Al 4 V particles adjacent to the coating layer. For that, the coated compact samples were epoxy impregnated and then cut perpendicular to the cylinder flat surface (A in Figure 3(a)) thro-ugh mid-section. Figures 6(a-e) show sequentially the coating layers on the particles after 1,2,3,4 and 5-min soaking times. The particle, coating layer and    epoxy layer are marked with arrows in Figure 6(a). As the soaking time increases the coating layer th-ickness increases as seen in Figures 6(a-e). The cr-acks in the coating layer are clearly seen in the mi-crograph of Figure 6(d) (marked with arrows). It is also noted that the coating layer is delaminated from the particle surface, partly arising from the attack of the chemical etching reagent used. In Figure 6(b), the Widmanstätten microstructure of the sintered Ti 6 Al 4 V particles, composing β lathes (bcc and rich in V)  and α platelets (hcp and rich in Al) is also seen.
In spite of the thicker coating layer formation near sintering necks, the coating layer thickness on the particles is considered to be relatively homoge-nous. The variation of HA coating layer thickness as function of soaking time is shown in Figure 7. At least 8 measurements were taken for each soaking time. The coating thickness shown in Figure 7 inc-reases with increasing soaking time, from 1.87 µm at 1-minute soaking to 2.95 µm at 2-minute, 4.15 µm at 3-minute, 5.66 µm at 4-minute and 9 µm at 5-minute soaking time on the average.

Discussion
The synthesized sol-derived HA powder used in this study was found to be partially crystalline after drying at 90°C and were single-phase and crystalline until about 1200°C. Thermal treatment temperature following the coating should however be selected at a minimum level that still assures sufficient quality of HA film, in terms of crystallinity, film integrity, and adhesion to the substrate. Furthermore, the lo-wer sintering temperatures also avoid α+β phase transition occurring at 883 and 960 °C in titanium and titanium alloy, respectively [19]. This ensures the retention of the starting material microstructure; and hence allowing to controlling the microstructure development. Based on the above facts, the sintering of compacts coated with sol-derived and commerc-ial HA powders were performed at 840 °C, below α+β phase transition temperature.
The thicker deposit at the sintering necks is at-tributed to be due to capillary effects drawing the coating solution into this concave region. The sinter necks have a concave geometry with the radius of curvature of the concavity being small (Figure 4). Cracks on the coating layer are likely resulting from the greater residual stresses that develop within the-se thicker film regions. Future studies are therefore required to optimize the soaking-time as function of crack formation on the coating layer. It was previou-sly shown that the bone in-growth occurred between the delaminated calcium phosphate layer and the su-bstrate in a Ti implant coated with Ti beads; furth-ermore, the extent of bone in-growth or fill within the sintering necks increased in the presence of the calcium phosphate film [13]. Therefore, the cracks and delaminations formed at the sintering necks are believed to not significantly impair the bone in-gro-wth behavior of the studied compacts. Furthermore, epoxy-mounted coated compacts were sectioned lo-ngitudinally in order to determine the depth of the coating and the openness of the pores. In Figure 8, the cross-section of a coated sample at 5-minute so-aking time is shown near the corner of the sample (B in Figure 3(a)). The coating depth from the compact surface is seen in this figure relatively small (300 µm), proving that only two-three particles near the compact surface are coated. The same shallow dep-th of coating was also observed in compacts coated at different soaking times, proving that few Ti 6 Al 4 V particles near the compact surface were HA-coated.
The present results have shown that the dip-co-ating soaking time has a great effect on the film th-icknesses and the openness of pores particularly at the surface. The soaking times until about 3-minute are sufficient to form a relatively thick HA layer on Ti 6 Al 4 V particles. Nevertheless, the effects of cert-ain microstructural features including coating layer thickness, coating depth and the coating defects on bone in-growth properties of the powder compacts should be further assessed.

Conclusions
In this study, the effect of dip-coating soak-ing time on the HA coating layer thickness of a Ti 6 Al 4 V sintered powder compact was investigated. The sintered powder Ti 6 Al 4 V compacts were prepar-ed using a powder metallurgical-sintering route with a mean pore size of 63 µm and an average porosity of 37±1%. The porous Ti 6 Al 4 V powder compacts were successfully dip-coated using sol-derived nano size HA powder until about the soaking times of 3-min-ute by considering uniformity of the coating layer and openness of the initial pores. The coating started from the particle sintering necks and cracks on the coating layer particularly at the sintering necks were observed in all dip-coated compact samples. Incre-asing soaking times however resulted in closing the originally open pores, while at lower soaking times the particles were partially coated. Finally, the coat--ing layer was shown to be nano-porous, comprising nano size HA particles and pores.