3.  Surface Texturing
Surface modifications have been applied to metallic biomaterials in order to improve mechanical, chemical, and physical properties such as wear resistance, corrosion resistance, biocompatibility and surface energy, etc. For enhancing the mechanical retention between two surfaces, one or both surfaces are normally modified to increase effective surface area either by sand-blasting, shot-peening, or laser-peening method. Another distinct purpose of surface modification is found on implant surfaces for both dental and orthopedic applications to exhibit biological, mechanical and morphological compatibilities to receiving vital hard/soft tissue, resulting in promoting osseointegration [97]. Such modifications are, in general, divided into two categories: surface concave texturing and surface convex texturing. Surface concave textures can be achieved by either material removal from its surface layer by chemical or electrochemical action, or mechanical indentations (caused by sand-blasting, shot-peening, or laser-peening) [97]. On the other hand, surface convex textured surfaces can be formed by depositing certain types of particles by one of several physical or chemical depositing techniques (like CVD, PVD, plasma-spraying, etc.) or diffusion bonding [97]. If density and porosity of deposited particles can be appropriately controlled, a porous surface can be achieved, leading to successful bone ingrowth. Surface roughness measurement is one of the most frequently and easily employed methods to characterize the modified surfaces. Hence, alternation of surface roughness should also be discussed in association with surface modifications.
Biological survival, particularly longevity of biological adhesive joints, is often dependent on thin surface films. Surfaces and interfaces behave completely different from bulk properties, as previously discussed. The characteristics of a biomaterial surface govern the processes involved in biological response. Surface properties such as surface chemistry, surface energy, and surface morphology may be studied in order to understand the surface region of biomaterials [97]. The surface plays a crucial role in biological interactions for four reasons: (1) the surface of a biomaterials is the only part contacting with the bio-environment, (2) the surface region of a biomaterial is almost always different in morphology and composition from the bulk, (3) for biomaterials that do not release or leak biologically active or toxic substance, the characteristics of the surface governs the biological response (foreign material vs. host tissue), and (4) some surface properties such as topography affect the mechanical stability of the implant-tissue interface [98–102]. Like the interface, the surface has a certain characteristic thickness, (1) for the case when the interatomic reaction is dominant, such as wetting or adhesion, atoms within a depth of 100 nm (1,000 Å) will be important, (2) for the case of the mechanical interaction, such as tribology and surface hardening, since the elasticity due to the surface contact and the plastically deformed layer will be a governing area, the thickness of about 0.1–10 μm will be important, and (3) for the case when mass transfer or corrosion is involved, the effective layer for preventing the diffusion will be within 1–100 μm [98–102].
As described previously, controlled surface roughness (rugophilicity) plays an important role to enhance the osseointegration of titanium implants [103–105]. Compared to smooth surfaces, osteoblasts can be grown on rough surfaces, which were fabricated in various methods, as will be discussed in details in the following sections [106,107]. One of the most important manufacturing parameters of titanium implants is roughening of the surface for increasing the effective surface area of implant body adjacent to the bone interface, thereby improving the cell attachment, bone apposition and biomechanical stability of the implant [59,108–113].
Such important surfaces can be further modified or altered in a favorable fashion to accommodate, facilitate, or promote more biofunctionality and bioactivity in mechanical, chemical, electrochemical, thermal, or any combination of these methods.

3.1.  Sand-Blasting
Sand-blasting, as well as shot-peening (which will be discussed in the following section), has three purposes: (1) cleaning surface contaminants prior to further operation, (2) roughening surfaces to increase effective surface area (for example, under some circumstances, the effective surface area could be double than the original surface area), and (3) producing beneficial surface compressive residual stress [114]. As a result, such treated surfaces exhibit higher surface energy, indicating higher surface chemical and physical activities, and enhancing fatigue strength as well as fatigue life due to compressive residual stress [114].
In order to obtain satisfactory fixation and biofunctionality of biotolerated and bioinert materials, some of the mechanical surface alternation such as threaded surface, grooved surface, pored surface, and rough surface have been produced that promote tissue and bone ingrowth [115]. But so far, there is no report on suitable roughness to specific metallic biomaterials. In general, on the macroscopic level (>10 μm), roughness will influence the mechanical properties of the interface, the way stresses are distributed and transmitted, the mechanical interlocking of the interface, and the biocompatibility of biomaterials. On a smaller scale, surface roughness in the range from 10 nm to 10 μm may influence the interface biology, since it is of the same order in size as cells and large biomolecules [48]. Topographic variations of the order of 10 nm and less may become important because microroughness on this scale length consists of material defects such as grain boundaries, dislocation steps, and vacancies, which are known to be active sites for adsorption, and thus may influence the bonding of biomolecules to the implant surface. There is evidence that surface roughness on a micron scale allows cellular adhesion that alters the overall tissue response to biomaterials [48]. Microrough surfaces allow early better adhesion of mineral ions or atoms, biomolecules, and cells, form stronger fixation of bone or connective tissue, result in a thinner tissue-reaction layer with inflammatory cells decreased or absent, and prevent microorganism adhesion and plaque accumulation, when compared with the smooth surfaces [48].
Piattelli et al. [116] conducted a histological and histochemical evaluation in rabbits to study the presence of multinucleated giant cells (MGCs) at the interface with machined, sand-blasted (with 150 μm alumina media), and plasma-sprayed titanium implants. It was reported that (i) MGCs were not observed at any of the experimental times around machined and sand-blasted titanium surfaces; whereas (ii) MGCs were present at the interface with titanium plasma-sprayed implants at two weeks and two months, (iii) at four and eight weeks these cells tended to decrease in number, and (iv) an inflammatory infiltrate was not present in connection with the MGCs [116].
Although alumina (Al2O3) or silica (SiO2) particles are most frequently used as a blasting media, there are several different types of powder particles utilized as media [113]. Surface roughness modulates the osseointegration of orthopedic and dental titanium implants [113]. This process may cause the release of cytotoxic silicium or aluminium ions in the peri-implant tissue [113]. To generate a biocompatible roughened titanium surface, an innovative grid-blasting process using biphasic calcium phosphate (BCP) particles was developed by Citeau et al. [113]. Ti-6Al-4V discs were either polished, BCP grid-blasted, or left as-machined. BCP grid-blasting created an average surface roughness of 1.57 μm compared to the original machined surface of 0.58 μm. X-ray photoelectron spectroscopy indicated traces of calcium and phosphorus and relatively less aluminum on the BCP grid-blasted surface than on the initial titanium specimen. It was reported that (i) scanning electronic microscopy observations and measurement of mitochondrial activity (MTS assay) showed that osteoblastic MC3T3-E1 cells were viable in contact with the BCP grid-blasted titanium surface, (ii) MC3T3-E1 cells expressed alkaline phosphate (ALP) activity and conserved their responsiveness to bone morphogenetic protein BMP-2, and (iii) the calcium phosphate grid-blasting technique increased the roughness of titanium implants and provided a non-cytotoxic surface with regard to mouse osteoblasts [113]. Tribo-chemical treatment has been proposed to enhance the bond strength between titanium crown and resin base [117]. Using silica-coated alumina as a blasting media under relatively low pressure, silica layer is expected to remain on the blasted surface so that retention force is enhanced by silan-coupling treatment.
Although the recent development of investment materials and casting machines has enabled clinical applications of titanium in dentistry, there remain several problems to be solved. First of all, efficient finishing techniques are required. Titanium is known to be difficult to grind because of its plasticity, stickiness, low heat conductivity, and chemical reactivity at high temperatures [118,119]. Although blasting shows several advantages, there is evidence of adverse effects: (1) surface contamination, depending on type of blasting media, and (2) distortion of blasted workpiece, depending on blasting manner and intensity. Miyakawa et al. [120] studied the surface contamination of abraded titanium. Despite low grinding speeds and water cooling, the abraded surfaces were found to be contaminated by abrasive constituent elements. Element analysis and chemical bond state analysis of the contaminants were performed using an electron probe microanalyzer. X-ray diffraction of the abraded surface was performed to identify the contaminants. It was reported that (i) the contamination of titanium is related to its reactivity as well as its hardness, (ii) in spite of water cooling and slow-speed abrading, titanium surfaces were obviously contaminated, (iii) contaminant deposits with dimensions ranging from about 10 to 30 μm occurred throughout the surfaces, and (iv) the contaminant of titanium, although related to the hardness, resulted primarily from a reaction with abrasive materials, and such contamination could negatively influence titanium’s resistance to corrosion and its biocompatibility [120].
Normally, fine alumina particles (50 μm Al2O3) are recycled within the sand-blasting machine. Ceramics such as alumina are brittle in nature, therefore some portions of recycled alumina might be brittle-fractured. If fractured sand blasting particles are involved in the recycling media, it might result in irregular surfaces, as well as potential contamination. Using fractal dimension analysis [121–123], a sample plate surface was weekly analyzed in terms of topographic changes, as well as chemical analysis of sampled recycled Al2O3 particles. It was found that after accumulated use time exceeded 30 mins, the fractal dimension (DF) remained a constant value of about 1.4, prior to that it continuously increased from 1.25 to 1.4. By the electron probe microanalysis on collected blasting particles, unused Al2O3 contains 100% Al, whereas used (accumulated usage time was about 2,400 sec) particles contained Al (83.32 wt%), Ti (5.48), Ca (1.68), Ni (1.36), Mo (1.31), S (1.02), Si (0.65), P (0.55), Mn (0.49), K (0.29), Cl (0.26), and V (0.08), strongly indicating that used alumina powder was heavily contaminated, and a high risk for the next material surface to be contaminated. Such contaminants are from previously blasted materials having various chemical compositions, and investing materials as well [124].
There is evidence of surface contamination due to mechanical abrasive actions [125]. As a metallographic preparation, the surface needs to be mechanically polished with a metallographic paper (which is normally SiC-adhered paper) under running water [125]. It is worth mentioning here that polishing paper should be changed between different types of materials, and particularly when a dissimilar metal-couple is used for galvanic corrosion tests, such couple should not be polished prior to corrosion testing because both materials could become cross-contaminated. Hence, there are attempts to use TiO2 powder for blasting onto titanium material surfaces. It was reported that titanium surfaces were sand-blasted using TiO2 powder (particle size ranging from 45 μm, 45 μm-63 μm, and 63 μm-90 μm) to produce the different surface textures prior to fibroblast cell attachment [126].
In the rabbit tibia, CpTi implants, which were sand-blasted with 25 μm Al2O3 and TiO2 particles, were inserted in the rabbit tibia for 12 weeks [127]. Even though the amount of Al on the implant surface was higher than for the Al2O3-blasted implants compared to implants not blasted with Al2O3, any negative effects of the Al element were not detected [128], which is in contrast to those reported by Johansson et al. [127], who reported that Al release from Ti-6Al-4V implants was found to coincide with a poorer bone-to-implant over a three month period. It is possible that the lack of differences between TiO2-blasted and the Al2O3-blasted implants depends on lower surface concentrations of toxic Al ions than those reported by Johansson et al. [127]. Wang [129] investigated the effects of various surface modifications on porcelain bond strengths. Such modifications included Al2O3 blasting, TiO2 blasting, HNO3 + HF + H2O treatment, H2O2 treatment, and pre-oxidation in air at 600 °C for 10 min. Ti-porcelain couples were subjected to 3-point bending tests. It was concluded that TiO2 air abrasion showed the highest bond strength, which was significantly different from other surface treatments.
Recently, it was reported that sand-blasting using alumina particle caused a remarkable distortion on a Co-Cr alloy and a noble alloy [130,131]. It was estimated that the stress causing the deflection exceeded the yield strength of tested materials. It was also suggested that the sand-blasting should be done using the lowest air pressure, duration of blasting period, and particle size alumina in order to minimize distortion of crowns and frameworks. To measure distortion, Co-Cr alloy plates (25 mm long, 5 mm wide, 0.7 mm thick) were sand-blasted with Al2O3 of 125 μm. Distortion was determined as the deflection of the plates as a distance of 20mm from the surface. It was reported that (i) the mean deflections varied between 0.37 mm and 1.72 mm, and (ii) deflection increased by an increase in duration of the blasting, pressure, particle size, and by a decrease in plate thickness [130].

3.2.  Shot-Peening and Laser-Peening
Shot peening (which is a similar technique to sand-blasting, but has more controlled peening power, intensity, and direction) is a cold working process in which the surface of a part is bombarded with small spherical media called shot. Each piece of shot striking the material acts as a tiny hammer, imparting to the surface small indentations or dimples. In order for the dimple to be created, the surface fibers of the material must be yielded in tension. Below the surface, the fibers try to restore the surface to its original shape, thereby producing below the dimple a hemisphere of cold-worked material highly stressed in compression. Overlapping dimples (which are sometimes called forged dimples) develop an even layer of metal in residual compressive stress. Both compressive stresses and cold working effects are used in the application of shot peening in forming metal parts, called “shot forming” [132].
The laser peening technology is recently developed, claiming non-contact, no-media, and contamination-free peening method [132]. Before treatment, the workpiece is covered with a protective ablative layer (paint or tape) and a thin layer of water. High-intensity (5–15 GW/cm2) nanosecond pulses (10–30 ns) of laser light beam (3–5 mm width) striking the ablative layer generate a short-lived plasma which causes a shock wave to travel into the workpiece. The shock wave induces compressive residual stress that penetrates beneath the surface and strengthens the workpiece [133–136], resulting in improvements in fatigue life and retarding in stress corrosion cracking occurrence. Cho et al. laser-treated CpTi screws and inserted in right tibia metaphysics of white rabbits for 8 weeks [137]. It was reported that (i) SEM of laser-treated implants demonstrated a deep and regular honeycomb pattern with small pores, and (ii) eight weeks implantation, the removal torque was 23.58 N-cm for control machined and 62.57 N-cm for laser-treated implants. Gaggl et al. reported that (i) surfaces of laser-treated Ti implants showed a high purity with appropriate roughness for good osseointegration, and (ii) the laser-treated Ti had regular patterns of micropore with interval of 10–12 μm, diameter of 25 μm, and depth of 20 μm [138].
At the end of this section, it is necessary to summarize various techniques to measure and characterize the surface roughness. They include that (1) surface roughness can be measured using a profilometer with sharp edge stylus, which is a contact method, (2) atomic force microscopy can provide non-contact surface topography from which the surface roughness can be indirectly measured, and (3) fractal dimension analysis can be used to present the surface roughness in non-Euclidian dimension [121,124]. Recently, Hansson et al. [139] employed computer simulations to measure surface roughness. The lateral resolution was defined as the pixel size of a profiling system. A surface roughness was simulated by a trigonometric function with random periodicity and amplitude. The function was divided into an array of pixels simulating the pixels of the profiling system. The mean height value for each pixel was used to calculate the surface roughness parameters. It was found that the accuracy of all the surface roughness parameters investigated decreased with increasing pixel size. This tendency was most pronounced for mean slope and developed length ratio, amounting to about 80% of their true values for a pixel size of 20% of the true mean high-spot spacing. It was concluded that the lateral resolution of an instrument/method severely compromises the precision of surface roughness parameters which are measured for roughness features with a mean high-spot spacing less than five times the lateral resolution [139].

3.3.  Chemical, Electrochemical, and Thermal Modifications
There are several experimental results on chemical, electrochemical, thermal, and combinations of these with regard to altering the Ti surface to facilitate better surface chemical, mechanical, and biological reactions. Endo [140] treated NiTi in 30% NHO3, then heated at 400 °C for 0.75 h, and NHO3 treatment, followed by boiling in water for 6–14 h. The variously treated NiTi surfaces were tested for dissolution resistance in bovine serum. It was found that (i) those stems thermally treated were found to have significantly lower metal ion release due to stable rutile oxide (TiO2) formation, (ii) human plasma fibronection (an adhesive protein) was covalently immobilized onto an alkylaminosilane derivate of NiTi substrate with glutaraldehyde, and (iii) the XPS spectra suggested that gamma-aminopropyltriethoxysilane (γ-APS) was bonded to the surface through metallosiloxane bonds (Ti-O-Si) formed via a condensation reaction between the silanol end of γ-APS and the surface of the hydroxyl group, with a highly cross-linked siloxane network formed after heat treatment of the silanized surface at 100 °C. Based on these findings, it was concluded that human plasma fibronectin was immobilized at the surface, and significantly promoted fibroblast spreading, suggesting that this chemical modification offers an effective means of controlling metal/cell interactions [140]. In study done by Browne et al. [141], hip replacement stems manufactured from the Ti-6Al-4V alloy were surface-treated and tested for dissolution resistance in bovine serum. Specimens were degreased in 1,2-dichloroethane vapor and surface treated in one of four ways: (1) 35% nitric acid for 10 min-typical commercial treatment, (2) 35% nitric acid for 16 h and rinsed in distilled water, (3) thermal heating in a furnace for 0.75 h at 400 °C, and (4) 35% nitric acid, then aged in boiling distilled water in a silica beaker for various times, 6, 8, 10, and 14 h. It was found that thermal treatment and aging of surface oxides promote the formation of dense rutile structure. This is effective in reducing metal ion dissolution (up to 80%), particularly in the early stages of implantation where the stem surface is equilibrating with its surroundings. This benefit is further enhanced on rough surfaces with an increased surface area. It was, therefore, concluded that (i) the thermal treatment and aging of the surface oxides are important with respect to cementless and porous implants, and (ii) such treatments could be incorporated in commercial manufacturing procedures to reduce the risk of metal dissolution being a contributory factor towards revision surgery [141].
Krozer et al. [142] investigated the possible influence of an amino-alcohol solution on machined Ti surface properties. Screw-shaped CpTi implants and CpTi studs were used. They were rinsed (1) in running deionized water for 2 min, (2) NaCl solution for 2 min followed by deionized water washing, and (3) rinsed in 5% H2O2 for 2 min followed by deionizd water washing, and rinsed in deionized water for 2 min. The amino-alcohol solution was supplied to the sample surfaces, and four methods were used in order to remove the adsorbed alcohol molecules. It was shown that (i) rinsing in water, saline solution, and 5% H2O2 did not remove the amino-alcohol from the surface; however (ii) exposure to ozone produced by using a commercial mercury lamp in ambient air resulted in complete removal of the adsorbed amino-alcohol, and (iii) the presence of such a film most likely prevents re-integration to occur at the implant-tissue interface in vivo [141]. In study done by Rupp et al. [142], CpTi was first blasted with 354–500 μm large grits, followed by (1) HCl/HF/HNO3 etching, (2) HCl/H2SO4 etching, (3) HCl/H2SO4/HF/ oxalic acid + neutralized, and (4) HCl/H2SO4/HF/oxalic acid + oxidized. It was reported that the Ti modifications which shift very suddenly from a hydrophobic (high surface contact angle) to a hydrophilic (low surface contact angle) state adsorbed the highest amount of immunologically assayed fibronectin [143]. This is suggesting that microtexturing greatly influenced both the dynamic wettability of Ti implant surfaces during the initial host contact and the initial biological response of plasma protein adsorption.
MacDonald et al. [144] investigated the microstructure, chemical composition, and wettability of thermally, and chemically modified Ti-6Al-4V disks, and correlated the results with the degree of adsorption between the radiolabeled fibronectin and Ti-6Al-4V alloy surface and subsequent adhesion of osteoblast-like cells. It was found that (i) heating either in pure oxygen or atmosphere resulted in an enrichment of Al and V within the surface oxide, (ii) heating (in pure oxygen or atmosphere) and hydrogen peroxide treatment, both followed by butanol treatment, resulted in a reduction in content of V, but not in Al, (iii) heating (oxygen/atm) or hydrogen peroxide treatment resulted in a thicker oxide layer and a more hydrophilic surface when compared with chemically-passivated controls (in 40% NHO3); however, the post-treatment with butanol resulted in a less hydrophilic surface than heating or hydrogen peroxide treatment alone, and (iv) the greatest increases in the adsorption of radiolabeled fibronectin following treatment were observed with hydrogen peroxide/butanol-treated samples followed by hydrogen peroxide/butanol and heat/butanol, although binding was only increased by 20–40% compared to untreated control. These experiments with radiolabeled fibronectin indicated that enhanced adsorption to the glycoprotein was more highly correlated with changes in chemical composition, reflected in V content and decrease in the V/Al ratio, than with changes in wettability. It was, therefore, concluded that an increase in the absolute content of Al and/or V, or in the Al/V ratio is correlated with an increase in the fibronectin-promoted adhesion of an osteoblast-like cell line [144]. Li et al. [145] modified the surface of CpTi (grade 2) implants by the micro-arc oxidation, operated under voltage ranging from 190, 230, 270, 350, 450 and 600 V to form a porous layer. It was found that (i) with increasing voltage, the roughness (from 0.3 to 2.5 μm) and thickness (from 1 μm to 15 μm) of the film increased, and (ii) the TiO2 phase changed from anatase to rutile. The micro-arc oxidation was carried out in an aqueous electrolyte with calcium acetate monohydrate and calcium glycerophosphate in deionized water. During the micro-arc oxidation, it was found that (i) Ca and P ions were incorporated into the oxide layer, (ii) the in vitro cell responses were also dependent on the oxidation condition, and (iii) with increasing voltage, the alkaline phosphatase activity increased, while the cell proliferation rate decreased. Preliminary in vivo tests of the micro-arc oxidation-treated specimens on rabbits showed a considerable improvement in their osseointegration capacity as compared to the un-modified CpTi implant [145].
The surface bioactivity of titanium was investigated after water and hydrogen plasma immersion ion implantation (PI3) by Xie et al. [146]. PI3 method excels in the surface treatment of components possessing a complicated shape such as medical implants. In addition, water and hydrogen plasma immersion ion implantation has been extensively studied as a method to fabricate silicon-on-insulator substrates in the semiconductor industry, and so it is relatively straightforward to transfer the technology to the biomedical field. Water and hydrogen were plasma-implanted into titanium sequentially. It was found that (i) after incubation in simulated body fluids for cytocompatibililty evaluation in vitro, bone-like hydroxyapatite was found to precipitate on the (H2O + H2) implanted samples, while no apatite was found on titanium samples plasma implanted with water or hydrogen alone, and (ii) human osteoblast cells were cultured on the (H2O+H2)-implanted titanium surface and they exhibited good adhesion and growth. It was, accordingly, suggested that plasma immersion ion implantation is a practical means to improve the surface bioactivity and cytocompatibility of medical implants made of titanium [146].
Rohanizadeh et al. [147] investigated methods of preparing different types of titanium oxide (TiO2) and their effects on apatite deposition and adhesion on titanium surfaces. CpTi discs were subjected to the following treatments: (1) heat treatment at 750 °C; (2) oxidation in H2O2 solution followed by heat treatment; (3) dipping in rutile/gelatin slurry; and (4) dipping in anatase/gelatin slurry. Surface-treated Ti discs were immersed in a supersaturated calcium phosphate solution to allow apatite deposition. It was shown that (i) the percentage of area covered by deposited apatite was highest in sample discs which were dipped in an anatase/gelatin slurry, compared to the other groups, (ii) apatite deposited on Ti discs pretreated in H2O2 solution demonstrated the highest adhesion to the titanium substrate, and (iii) the surface treatment method affects the type of TiO2 layer formed (anatase or rutile) and affects apatite deposition and adhesion on the Ti surface [147].

3.4.  Coating
The coating layer is not only required to exhibit an expected function, depending on its original specific aims, but it is also important to notice that the coating layer is only functional if it adheres well to the metal substrate and if it is strong enough to transfer all loads. Coated substrate possesses at least two layers and one intermediate interface. If such coupled is subjected to stressing, although the strain field should be assumed to be a continuum, the stress field of the couple exhibits a discrete one due to differences in modulus of elasticity, as discussed in the previous section for mechanical compatibility. This discrete stress field results in interfacial stress, and if the interfacial stress is higher than the bonding strength, the couple can be debonded or delaminated, causing the structural integrity to no longer be maintained.

3.4.1.  Carbon, Glass, Ceramic Coating
The surface of Ti-6Al-4V has been modified by ion beam mixing a thin carbon film [148]. XPS analysis showed that after mixing, the surface film consists essentially of a Ti compound containing (Ti, O, and C), TiO2, Ti, and C. The composition of the surface modified film determined by Rutherford backscattering spectrometry is approximately Ti0.5O0.3C0.2 and its thickness is about 200 μm. It was also reported that after three months immersion in a simulated body fluid, the growth of calcium phosphate species containing both HPO4− and H2PO4− (probably CaHP4 and Ca(PO4)2) have been observed [148]. The corrosion resistance and other surface and biological properties of NiTi were enhanced using carbon plasma immersion ion implantation and deposition (PI3). Poon et al. [149] mentioned that either an ion-mixed amorphous carbon coating fabricated by plasma immersion ion implantation and deposition or direct carbon PI3 can drastically improve the corrosion resistance and block the out-diffusion on Ni from the metal. The tribo-logical tests showed that the treated surfaces are mechanically more superior and cytotoxicity tests revealed that both sets of plasma-treated samples favored adhesion and proliferation of osteoblasts [149]. With regard to potential toxicity of Ni, this is one of methods to prevent or shield the Ni element to diffuse out from NiTi surface. There is another way to achieve the similar outcome by selectively leaching out Ni from the NiTi surface layer by chemically etching the NiTi surface in mixed acid aqueous solution of HF + HNO3 + H2O (1:1:3 by volume) [66].
Bioactive glass (BAG) is a bioactive material with a high potential as implant material. Reactive plasma spraying produces a feasible BAG-coating for Ti-6Al-4V dental implants. It was shown that (i) the coating withstands, without any damage, an externally generated tensile stress of 47 MPa, and (ii) adhesion testing after two months of in vitro reaction in a simulated body fluid showed that coating adhesion strength decreased by 10%, but the implant was still adequate for load-bearing application [150].
Saiz et al. [151] evaluated the in vitro response in simulated body fluid of silicate glass coating on Ti-6Al-4V. Glasses belonging to the SiO2-CaO-MgO-Na2O-K2O-P2O5 system were used to prepare 50–70 μm thick coatings by employing a simple enameling technique. It has been found that (i) coatings with silica content lower than 60 wt% are more susceptible to corrosion and precipitate carbonated HA on their surface during in vitro tests; however (ii) these coatings have a higher thermal expansion than the metal, (iii), after 2 month in simulated body fluid, crack grows in the coating, reaches the glass/metal interface and initiates delamination, and (iv) glasses with silica content higher than 60wt% are more resistant to corrosion and have lower thermal expansion, and these coatings do not crack, but such glasses with silica do not precipitate apatite even after two months in simulated body fluid [151]. Lee et al. [152] prepared calcium-phosphate, apatite-wollastonite (CaSiO3) (1:3 by volume fraction) glass ceramic, apatite-wollastonite (1:1) glass ceramic, and bioactive CaO-SiO2-B2O3 glass ceramic coatings by the dipping method. Coated and uncoated Ti-6Al-4V screws were inserted into the tibia of 18 adult mongrel male dogs for 2, 4 and 8 weeks. It was found that (i) at 2, 4, and 8 weeks, the extraction torque of these ceramic-coated screws was significantly higher than the corresponding insertion torque, and (ii) strong fixation was observed even at two weeks in all three coatings except CaO-SiO2-B2O3 glass ceramic coating [152].

3.4.2.  Hydroxyapatite Coating
Enhancement of the osteoconductivity of Ti implants is potentially beneficial to patients since it shortens the medical treatment time and increases the initial stability of the implant. To achieve better osteoconductivity, apatite [Ca10-x(HPO4)x(PO4)6-x(OH)2-x] coating has been commonly employed on the Ti implant surface [152].
Hydroxyapatite is a major mineral component in animal and human bodies [153]. It has been used widely not only as a biomedical implant material but also as biological chromatography supports in protein purification and DNA isolation. Spherical HA ceramic beads have recently been developed that show improvements in mechanical properties and physical and chemical stability. These spherical ceramic beads are typically 20–80 μm in size. There are advantages to reducing the granule size of the spherical HA material: (1) the smaller the granule size, the higher the specific surface area and the higher the bonding capacity, (2) theoretically, the specific surface area (i.e., surface area per volume) is proportional to 6/d, where d is the diameter of the spherical granule, (3) in addition, the mechanical properties of a packed column can be improved by reducing the granule size, resulting in more contacting surface areas, and thereby greater frictional forces between granules, and (4) furthermore, a uniform pack is expected to have a homogeneous pore distribution [154]. The porosity and the specific surface areas of the HA material can be controlled by changing the morphology of the granules, for example, the solid spheres and doughnuts. Hence, Luo et al. [155] introduced a new method to produce spherical HA granules ranging in size from 1 to 8 μm with controlled morphology. This method involves an initial precipitation followed by a spray-drying process, which is the controlling step, to produce granules with various structures. It was reported that by adjusting the operating parameters (e.g., atomization pressure) and starting slurry (e.g., concentration), the hollow or solid spheres and doughnut-shaped HA were fabricated.
BSE (bovine spongiform encephalopathy) or ‘mad cow disease’ could have been caused by animal-feed contaminated with human remains a controversial theory. Accordingly, although bovine-derived hydroxyapatite was used as a semi-natural HA [154], the articlers reviewed here are limited to those published before the BSE issue became to be addressed and received a public attention.
With the growing demands of bioactive materials for orthopedic as well as maxillofacial surgery, the utilization of hydroxyapatite (HA, with Ca/P = 1.67) and tri-calcium phosphate (TCP, with Ca/P = 1.50) as fillers, spacers, and bone graft substitutes has received great attention during the past two to three decades, primarily because of their biocompatibility, bioactivity, and osteoconduction characteristics with respect to host tissue. Porous hydroxyapatite granules with controlled porosity, pore size, pore size distribution, and granule size were fabricated using a drip-casting process. Granules with a wide range of porosity from 24 to 76 vol.%, pore size from 95 to 400 μm, and granular sizes from 0.7 to 4 mm can be obtained. This technique allows the fabrication of porous granules with desirable porous characters simulating natural bone architecture, and is expected to provide advantages for biomedical purposes [155].
Plasma-sprayed HA-coated devices demonstrated wide variability in the percentage of the HA coating remaining on the stems. Porter et al. [156] reported that (i) the coating was missing from a substantial portion of a stem after only about six months of implantation, and (ii) many ultrastructural features of the bone bonded to the HA coatings on these implants from human subjects were comparable to those on HA-coated devices implanted in a canine model. The geometric design and chemical compositions of an implant surface may have an important part in affecting early implant stabilization and influencing tissue healing.
The influence of different surface characteristics on bone integration of titanium implants was investigated by Buser et al. [54]. Hollow-cylinder implants with different surfaces were placed in the metaphyses of the tibia and femur in six miniature pigs. After three and six weeks, the implants with surrounding bone were removed and analyzed in undecalcified transverse sections. The histologic examination revealed direct bone-implant contact for all implants. However, the morphometric analyses demonstrated significant differences in the percentage of bone-implant contact, when measured in cancellous bone. It was reported that (i) electropolished, as well as the sand-blasted and acid pickled (medium grit, HF/HNO3) implant surfaces, had the lowest percentage of bone contact with mean values ranging between 20 and 25%, (ii) sand-blasted implants, with a large grit, and titanium plasma-sprayed implants demonstrated 30–40% mean bone contact, and (iii) the highest extent of bone-implant interface was observed in sand-blasted and acid attacked surfaces (large grit; HCl/H2SO4) with mean values of 50–60%, and hydroxyapatite (HA)-coated implants with 60–70%. It was therefore concluded that the extent of the bone-implant interface is positively correlated with an increasing roughness of the implant surface. Moreover the morphometric results indicated that (i) rough implant surfaces generally demonstrated an increase in bone apposition compared to polished or fine structured surfaces, (ii) the acid treatment with HCl/H2SO4 used for sand-blasted with large grit implants has an additional stimulating influence on bone apposition, (iii) the HA-coated implants showed the highest extent of bone-implant interface, and (iv) the HA coating consistently revealed signs of resorption. It was suggested that sand-blasting and chemical etching with HCl/H2SO4 as well as HA coating, seemed to be the most promising alternatives to titanium implants with smooth or titanium plasma-sprayed surfaces [54].
Souto et al. [157] investigated the corrosion behavior of four different preparations of plasma-sprayed hydroxyapatite (HA) coatings (50 μm and 200 μm) on Ti-6Al-4V substrates in static Hank’s balanced salt solution through DC potentiodynamic and AC impedance EIS techniques. Because the coatings are porous, ionic paths between the electrolytic medium and the base material can eventually be produced, resulting in the corrosion of the coated metal. It was concluded that significant differences were found in electrochemical behavior for similar nominal thicknesses of HA coatings obtained under different spraying [157]. Filiaggi et al. [158] reported that evaluations of an HA-coated Ti-6Al-4V implant system using a modified short bar technique for interfacial fracture toughness determination revealed relatively low fracture toughness values. Using high resolution electron spectroscopic imaging, evidence of chemical bonding was revealed at the plasma-sprayed HA/Ti-6Al-4V interface, although bonding was primarily due to mechanical interlocking at the interface [158]. The modulus of elasticity, residual stress and strain, bonding strength, and microstructure of the plasma-sprayed hydroxyapatite coating were evaluated on Ti-6Al-4V substrate with and without immersion in Hank’s balanced salt solution. It was reported that (i) the residual stress and strain, modulus of elasticity, and bonding strength of the plasma-sprayed HA coating after immersion in Hank’s solution were substantially decreased, and (ii) the decayed modulus of elasticity and mechanical properties of HA coatings were caused for by the degraded interlamellar or cohesive bonding in the coating due to the increased porosity after immersion that weakens the bonding strength of the coating and substrate system. It was also suggested that the controlled residual stress and strain in the coating might promote the long-term stability of the plasma-sprayed HA-coated implant [159]. Yang et al. [160] investigated the effect of titanium plasma-sprayed (TPS) and zirconia (ZrO2)-coated titanium substrates on the adhesive, compositional, and structural properties of plasma-sprayed HA coatings. Apatite-type and α-tricalcium phosphate phases were observed for all HA coatings. The coating surfaces appeared rough and melted, with surface roughness correlating to the size of the starting powder. It was found that (i) no significant difference in the Ca/P ratio of HA on Ti and TPS-coated Ti substrates was observed, (ii) however, the Ca/P ratio of HA on ZrO2-coated Ti substrate was significantly increased, (iii) interfaces between all coatings and substrates were observed to be dense and tightly bound, except for HA coatings on TPS-coated Ti substrate interface; however (iv) an intermediate TPS or ZrO2 layer between the HA and Ti substrate resulted in a lower adhesive strength as compared to HA on Ti substrate [160].
HA can be admixed with Ti powder. 80HA-20Ti powder and 90HA-10Ti powder (by weight) were mechanically mixed and dry-pressed and heat treated at 1,100 °C for 30 min in vacuum (less than 6 × 10−3 Pa). Heat treatment of HA specimens in vacuum resulted in the loss of hydroxyl groups, as well as the formation of a secondary β-tricalcium phosphate phase. It was concluded that the in vacuo heat treatment process completely converted the metal-ceramic composites to ceramic composites [161]. Moreover, using air plasma spraying and vacuum plasma spraying methods, HA and a mixture of HA + Ti were deposited on Ti-6Al-4V. It was reported that a higher adhesion was obtained with vacuum plasma spraying rather than with air plasma spraying [162].
Lee et al. [163] employed the electron-beam deposition method to obtain a series of fluoridated apatite coatings. The fluoridation of apatite was aimed to improve the stability of the coating and elicit the fluoride effect, which is useful in the dental restoration area. Apatite fluoridated at different levels was used as initial evaporants for the coatings. It was observed that (i) the as-deposited coatings were amorphous, but after heat-treated at 500 °C for 1 h, the coatings crystallized well to an apatite phase without forming any cracks, (ii) the adhesion strengths of the as-deposited coatings were about 40 MPa, and after heat treatment at 500 °C, it decreased to about 20 MPa; however the partially fluoridated coating maintained its initial strength. It was also reported that (i) the osteoblast-like cells responded to the coatings in a similar manner to the dissolution behavior, (ii) the cells on the fluoridate coatings showed a lower proliferation level compared to those on the pure HA coating, and (iii) the alkaline phosphatase activity of the cells was slightly lower than of on the pure HA coating [163]. Kim et al. [164] deposited HA and fluoridated HA films on CpTi (grade 2). HA sol was prepared by mixing P(C2H5O)3 and distilled water, and fluoridated HA sol was prepared by NH4F in P containing solution (like HA by replacing the OH group with F ions) and their films were deposited on CpTi (grade II). The mixture was stirred at room temperature for 72 hours. Dipping CpTi into the solutions was performed at 500 °C for 1 h. It was concluded that (i) the coatings layers were dense, uniform, and had a thickness of approximately 5μm after heat treatment at 500 °C, (ii) the fluoridated HA layer showed much lower dissolution rate (in 0.9% NaCl as physiological saline solution) than pure HA, suggesting the tailoring of solubility with F-incorporation within the apatite structure, (iii) the osteoblast-like MG63 and HOS cells grew and proliferated favorably on both coatings and pure Ti, and (iv) especially, both coated Ti exhibited higher alkaline phosphatase expression levels as compared to non-coated Ti, confirming the improved activity and functionality of cells on the substrate via the coatings [164].
There are several studies done on effects of post-spray coating heat-treatment on mechanical and structural integrities of the coated film. A post-plasma-spray heat treatment (at 960 °C for 24 h followed by slow furnace cooling) was performed to enhance chemical bonding at the metal/ceramic interface, and hence improve the mechanical properties. It was found that any improvements realized were lost due to the chemical instability of the coating in a moisture-laden environment, with a concomitant loss in bonding properties, and this deterioration appeared to be related to environmentally assisted crack growth as influenced by processing conditions [165]. HA coatings on Ti-6Al-4V were annealed at 400 °C in air for 90 h, and were evaluated as post-coating heat treatment. It was found that (i) the oxide species TiO2, Al2O3, V2O5, V2O3 and VO2 were present on both as-coated and as heat-treated samples, and (ii) the fatigue resistance of the substrate was found to be significantly reduced by the heat treatment, due to the stress relief [166]. Ti-6Al-4V substrate was heated at 25°, 160°, and 250 °C, followed by cooling in air or air + CO2 gas during operating of the HA plasma coating. It was reported that (i) when residual compressive stress was 17 MPa the bonding strength was 9 MPa, while the bonding strength continuously reduced to 3 MPa for a residual compressive stress of about 23 MPa, and (ii) the compressive residual stress weakened the bonding at the interface of the HA and the Ti-substrate [167], due to remarkable mechanical mismatching.
As has been reviewed in the above, a plasma spray method has been widely accepted as the apatite coating method because it gives tight adhesion between the apatite coating and Ti. However, the plasma spray method employs extremely high temperatures (10,000–12,000 °C) during the coating process. Unfortunately, it results in potentially serious problems including (1) an alteration of structure, (2) formation of apatite with extremely high crystallinity, and (3) long term dissolution and the accompanying debonding of the coating layer. To reduce these drawbacks associated with the conventional plasma spray method, a high viscosity flame spray method was developed. Although the high viscosity flame spray method employs lower temperature compared with the plasma spray method, it is still 3,000 °C, which is adequate to alter the crystal structure and formation of apatite with extremely high crystallinity. To avoid shortcomings in the apatite coating methods using high temperatures, many alternative room temperature coating methods were studied extensively including ion beam sputtering, dipping, electrophoretic deposition, and electrochemical deposition [168]. Mano et al. [168] developed a new coating method called blast coating, using common sand-blasting equipment at room temperature. Blast-apatite coated CpTi implants were inserted in tibia of rats for 1, 3, and 6 weeks. It was found that (i) the apatite coating adhered tightly to the Ti surface even after the 6 week implantation, (ii) the implants cause no inflammatory response, showing strong bone response and much better osteoconductivity compared with the uncoated CpTi implant, (iii) the new bone formed on the surface of coated implants was thinner compared with that formed on the surface of Ti implant. Therefore, the blast-apatite implant has a good potential as an osteoconductive implant material [168].
Plasma-sprayed HA coatings on commercial orthopedic and dental implants consist of mixtures of calcium phosphate phases, predominantly a crystalline calcium phosphate phase, hydroxyapatite, and an amorphous calcium phosphate with varying ratios. Alternatives to the plasma-spray method are being explored because of some of its disadvantages, as mentioned before. Moritz et al. [169] developed a two-step (immersion and hydrolysis) method to deposit an adherent apatite coating on titanium substrate. First, titanium substrates were immersed in acidic solution of calcium phosphate, resulting in the deposition of a monetite (CaHPO4) coating. Second, the monetite crystals were transformed to apatite by hydrolysis in NaOH solution. Energy dispersive spectroscopy had revealed the presence of calcium and phosphorous elements on the titanium substrate after removing the coating using tensile or scratching tests. It was reported that (i) the average tensile bond of the coating was 5.2 MPa and cohesion failures were observed more frequently than adhesion failures, (ii) this method has the advantage of producing a coating with homogenous composition on even implants of complex geometry or porosity, and (iii) this method involves low temperatures and, therefore, can allow the incorporation of growth factors or biogenic molecules [169].
A novel method to rapidly deposit bone apatite-like coatings on titanium implants in simulated body fluid (SBF) has been proposed by Han et al. [170]. The processing was composed of two steps: micro-arc oxidation of titanium to form titania (TiO2) films, and UV-light illumination of the titania-coated titanium in SBF. It was reported that (i) the micro-arc oxidation films were porous and nanocrystalline, with pore sizes varying from 1 to 3 μm and grain sizes varying from 10–20 to 70–80 nm; the predominant phase in titania films changed from anatase to rutile, and the bond strength of the films decreased from 43.4 to 32.9 MPa as the applied voltage increased from 250 to 400 V, (ii) after UV-light illumination of the films in SBF for 2 h, bone apatite-like coating was deposited on the micro-arc oxidation film formed at 250 V, and (iii) the bond strength of the apatite/titania bilayer was about 44.2 MPa [170].
The technique was further developed using the sol-gel method [164] to coat Ti surface with hydroxyapatite (HA) films [171]. The coating properties, such as crystallinity and surface roughness, were controlled and their effects on the osteoblast-like cell responses were investigated. The obtained sol-gel films had a dense and homogeneous structure with a thickness of about 1μm. It was found that (i) the film heat-treated at higher temperature had enhanced crystallinity (600 °C > 500 °C > 400 °C), while retaining similar surface roughness, (ii) when heat-treated rapidly (50 °C/min), the film became quite rough, with roughness parameters being much higher (4–6 times) than that obtained at a low heating rate (1 °C/min), and (iii) the dissolution rate of the film decreased with increasing crystallinity (400 °C > 500 °C > 600 °C), and the rougher film had slightly higher dissolution rate. The attachment, proliferation, and differentiation behaviors of human osteosarcoma HOS TE85 cells were affected by the properties of these films. It was further reported that (i) on the films with higher crystallinity (heat treated over 500 °C), the cells attached and proliferated well, and expressed alkaline phosphatase and osteocalcin to a higher degree as compared to the poorly crystallized film (heat treated at 400 °C), and (ii) on the rough film, the cell attachment was enhanced, but the alkaline phosphatase and osteocalcin expression levels were similar as compared to the smooth films [171].
The sol-gel method was favored due to the chemical homogeneity and fine grain size of the resultant coating, and the low crystallization temperature and mass-producibility of the process itself. The sol-gel-derived HA and TiO2 films, with thicknesses of about 800 and 200 nm, adhered tightly to each other and to the CpTi (grade 2) substrate. It was reported that (i) the highest bond strength of the double layer coating was 55 MPa after heat treatment at 500 °C due to enhanced chemical affinity of TiO2 toward the HA layer, as well as toward the Ti substrate, (ii) human osteoblast-like cells, cultured on the HA/TiO2 coating surface, proliferated in a similar manner to those on the TiO2 single coating and on the CpTi surfaces; however (iii) alkaline phosphatase activiy of the cells on the HA/TiO2 double was expressed to a higher degree than on the TiO2 single coating and CpTi surfaces, and (iv) the corrosion resistance of Ti was improved by the presence of the TiO2 coating, as confirmed by a potensiodynamic polarization test [172]. Sol-gel-derived TiO2 coatings are known to promote bone-like hydroxyapatite formation on their surfaces in vitro and in vivo. Hydroxyapatite integrates into bone tissue. In some clinical applications, the surface of an implant is simultaneously interfaced with soft and hard tissues, so it should match the properties of both. Ergun et al. [173] studied the chemical reactions between hydroxylapatite (HA) and titanium in three different kinds of experiments to increase understanding the bond mechanisms of HA to titanium for implant materials. HA powder was bonded to a titanium rod with hot isostatic pressing. Interdiffusion of the HA elements and titanium was found in concentration profiles measured in the electron microprobe. Titanium was vapor-deposited on sintered HA discs and heated in air; perovskite (CaTiO3) was found on the HA surface with Rutherford backscattering and X-ray diffraction measurements. Powder composites of HA, titanium, and TiO2 were sintered at 1,100 °C; again, perovskite was a reaction product, as well as β-Ca3(PO4)2, from a decomposition of the HA. Based on these findings, it was reported that (i) chemical reactions and interdiffusion between HA and TiO2 during sintering resulted in chemical bonding between HA and titanium; thus, cracks and weakness at HA-titanium interfaces probably result from mismatch between the coefficients of thermal expansion of these materials, and (ii) HA composites with other ceramics and different alloys should lead to better thermal matching and better bonding at the interface [173].
The effects of HA coating and macrotexturing of Ti-6Al-4V was tested by Hayashi et al. [174,175]. Implants were inserted in dog’s femoral condyles. It was demonstrated that when grooved Ti implants were used, the addition of HA coating significantly improved the biological fixation. In addition, a grooved depth of 1 mm was found to give significantly better fixation than 2 mm. When compared to implants with traditional diffusion-bonded bead-coated porous surfaces, HA-coated grooved Ti implants were found to show better fixation at 4 weeks after implantation, but significantly inferior fixation at 12 weeks. Hence, it was concluded that while a groove depth of 1 mm was optimal in HA-coated and further grooved Ti implants, they are still inferior to bead-coated Ti implants with respect to longer-term fixation [174]. Two different groups of HA coated and uncoated porous Ti implants, 250–350 μm and 500–700 μm diameter beads, were press-fitted into femoral canine cancellous bone, for 12 weeks. It was reported that (i) the percentage of bone and bone index were higher in the HA coated implants, when comparing HA coated vs. uncoated implant in the 250–350 μm bead diameter group, (ii) comparing 500–700 μm, bone ingrowth and bone depth penetration were higher in HA coasted samples, and (iii) the HA-coating was an effective method for improving bone formation and ingrowth in the porous implants [176].
Several HA-coated and uncoated Ti-6Al-4V femoral endoprostheses were evaluated in adult dogs for 52 weeks. It was found that (i) histologic sections from the uncoated grooved implants showed no direct bone-implant apposition with no fibrous tissue interposition after up to 10 weeks, and (ii) the HA-coated grooved implants demonstrated extensive direct bone-coating apposition after five weeks [177]. HA-coated cylindrical plugs of CpTi, Cp-Ti screws, and partially HA-coated CpTi screws were inserted in tibia of adult New Zealand white rabbits for three months. The histological results demonstrated that although there were no marked differences in bony reaction at the cortical level to the different implant materials, HA-coating appeared to induce more bone formation in the medullary cavity. It was also noted that 3 months after insertion loss of coating thickness had occurred [178]. Threaded HA-coated CpTi implants were inserted in the rabbit tibial metaphysis. After six weeks and 1 year post-insertion, the semiloaded implants were histomorphometrically analyzed. It was reported that (i) there was more direct bony contact with the HA-coated implants after 6 weeks, and (ii) one year after insertion, there was significantly more direct bone-to-implant contact with the uncoated CpTi controls, suggesting that the HA coating does not necessarily improve implant integration over a long time period [179]. After six or 12 weeks, four goats were used for mechanical tests and three for histology. To cope with the severe femoral bone stock loss encountered in revision surgery, impacted trabecular bone grafts were used in combination with an HA-coated Ti stem. In this first experimental study, the results indicated that this technique had a high complication rate. However, it was shown that impacted grafts sustained the loaded stems and that incorporation of the graft occurred with a biomechanically stable implant. The technique allows gradual graft incorporation and stability, but more investigations are needed before its introduction into clinical practice [180]. Surface treatments, such as sand-blasting or deposit or rough pure Ti obtained by vacuum plasma spraying, bring about an increase of passive and corrosion current density values as a consequence of the increase of the surface exposed to the aggressive environment. It should, however, be outlined that in the case of rough Ti, only Ti ions are released instead of Al or V out of Ti-6Al-4V. The presence of deposits of HA by vacuum plasma spraying causes an increase of about one order of magnitude in the passive and corrosion current density of the metallic substrate [181].
There are several works in regard to mechanical properties of HA and bond strength of HA layer. Cook et al. [182] conducted interface shear strength tests using a transcortical push-out model in dogs after 3, 5, 6, 10 and 32 weeks on CpTi-coated Ti-6Al-4V and HA-coated Ti-6Al-4V. The mean values for interface shear strength increased up to 7.27 MPa for HA-coated implants after 10 weeks of implantation, while uncoated CpTi had 1.54MPa [182]. Yang et al. [183] reported that the direction of principal stresses were in proximity to and perpendicular to the spraying direction. The measured modulus of elasticity (MOE) of HA (16 GPa at maximum) was much lower than the theoretical value (i.e., 112 GPa). The denser, well-melted HA exhibited a higher residual stress (compressive, 10–15 MPa at maximum), as compared with the less dense, poor-melting HA. The denser coatings could be affected by higher plasma power and lower powder feed rate. It was also reported that the thicker 200 μm HA exhibited higher residual stress than that of the thinner 50 μm HA [183]. Mimura et al. [184] characterized morphologically and chemically the coating-substrate interface of a commercially available dental implant coated with plasma-sprayed HA, when subjected to mechanical environment. A thin Ti oxide film containing Ca and P was found at the interface on Ti-6Al-4V. When the implant was subjected to mechanical stress, a mixed mode of cohesive and interfacial fractures occurred. The cohesive fracture was due to separation of the oxide film from the substrate, while the interfacial fracture was due to the exfoliation of the coating from the oxide film bonded to the substrate. It was reported that (i) microanalytical results showed diffusion of Ca into the metal substrate, hence indicating the presence of chemical bond at the interface; however, (ii) mechanical interlocking seemed to play the major role in the interfacial bond [184]. Lynn et al. [166] conducted uniaxial fatigue tests (σmax/σmin stress ratio, R = −1, stress amplitude of 620 MPa, frequency of 50Hz) on blasted-Ti-6Al-4V with HA coating on Ti-6Al-4V with film thickness ranging from 0, 25, 50, 75, 100, and 150 μm. It was found that (i) samples with 150 μm were shown to have significantly decreased fatigue resistance, while coatings of 25–100 μm were found to have no effect on fatigue resistance, and (ii) HA coatings with 25–50 μm show no observable delamination during fatigue tests, while coatings with 75–150 μm thick were observed to spall following but not prior to the initiation of the first fatigue crack in the substrate [166].
There are several studies done on apatite-like formation without HA coating. Ti can form a bone-like apatite layer on its surface in simulated body fluid (SBF) when it is treated in NaOH. When pre-treated Ti is exposed to SBF, the alkali ions are released from the surface into the surrounding fluid. The sodium ions increase the degree of supersaturation of the soaking solution with respect to apatite by increasing pH. On the other hand, the released Na+ causes an increase in external alkalinity that triggers an inflammatory response and leads to cell death. Therefore, it would be beneficial to decrease the release of Na+ into the surrounding tissue. It was found that (i) the rate of apatite formation was not significantly influenced by a lower amount of Na+ ion in the surface layer, and (ii) Ti with the lowest content of Na+ could be more suitable for implantation in the human body (4 at.%) [185]. Jonášová et al. [186] pre-treated CpTi surfaces: (1) in 10 M NaOH at 60 °C for 24 h, and (2) etched in HCl under inert atmosphere of CO2 for 2hr, followed by 10 M NaOH treatment. It was found that Ti treated in NaOH can form hydroxycarbonated apatite after exposition in SBF. Generally, Ti is covered with a passive oxide layer. In NaOH this passive film dissolves and an amorphous layer containing alkali ions is formed. When exposed to SBF, the alkali ions are released from the amorphous layer and hydronium ions center into the surface layer, resulting in the formation of Ti-OH groups in the surface. The acid etching of Ti in HCl under inert atmosphere leads to the formation of a micro-roughened surface, which remains after alkali treatment in NaOH. It was shown that the apatite nucleation was uniform and the thickness or precipitated hydroxycarbonated apatite layer increased continuously with time. The treatment of Ti by acid etching in HCl and subsequently in NaOH, is a suitable method for providing the metal implant with bone-bonding ability [186]. Wang et al. [187] reported that bone-like apatite was formed on Ti-6Al-4V surfaces pre-treated in NaOH solution immersed in SBF, while no apatite was formed on untreated Ti-6Al-4V. The increase in electrical resistance (by EIS) in the outermost surface of pre-treated Ti-6Al-4V indicated apatite nucleation. With an increase in immersion time in SBF, islands of apatite were seen to grow and coalesce on pre-treated Ti-6Al-4V under SEM. It was also mentioned that the growth of apatite corresponded to the increase in electrical resistance of the surface layer [187].

3.4.3.  Ca-P Coating
As we have just reviewed the evidence showed a formation of an apatite-like film containing Ca and P ions when Ti was pre-treated in NaOH, followed by immersion in SBF (simulated body fluid) or PBL (phosphate buffered liquid). Clinically, it was found that for Ti an increase in oxide thickness and an incorporation of Ca and P were found in placed Ti implants which were in the shape of screws, and had been osseointegrated in patients’ jaws for times ranging from 0.5 to 8 years [30]. These provide “hindsight” knowledge, where later on people will use this knowledge to modify the Ti surfaces.
Formation of calcium-phosphate on alkali- (10 M NaOH at 60 °C for 24 h) and heat-treated (alkali-treated then at 600 °C for 1 hfollowed by furnace cooling) CpTi was tested. Samples were immersed in the revised SBF with the same HCO3− concentration level as in human blood plasma. It was reported that (i) electron diffraction for the precipitates revealed that octacalcium phosphate, instead of HA, directly nucleates from amorphous calcium-phosphate, (ii) the octacalcium phosphate (OCP) crystals continuously grew on the Ti surfaces rather than transforming to AP, and (iii) calcium titanate (CaTiO3) was identified by electron diffraction [188]. Hanawa et al. [189] characterized surface films formed on titanium specimens which were immersed in electrolyte solutions (pH: 4.5, 5.1, 7.4) at 37 °C for 1 h, 1 day, 30 days and 60 days by XPS, FTIR-RAS Fourier transform infrared reflection absorption spectroscopy to understand the reaction between Ti and inorganic ions. For comparison, the surfaces of Ti-6Al-4V and NiTi were also characterized. XPS data revealed that calcium phosphates (CP) were naturally formed on these specimens. In particular, compared with the CP formed on the Ti alloys, the CP formed on Ti immersed for 30 days in the solution with pH 7.4 was more like HA. It was also reported that (i) the compositions of the CP formed on the specimens changed with the immersion time and the pH value of the solution (the solution with pH 7.4 was Hank’s balanced solution without organic species, solution with pH 4.5 was the artificial saliva without sodium sulphide and urea), and (ii) a CP similar to AP is naturally formed on Ti in a neutral electrolyte solution in 30 days [189].
Calcium phosphate (Ca-P) coatings have been applied onto titanium alloy prostheses to combine the strength of the metals with the bioactivity of Ca-P. It has been clearly shown in many publications that a Ca-P coating accelerates bone formation around the implant. However, longevity of the Ca-P coating for an optimal bone apposition onto the prosthesis remains controversial. Barrère et al. [190] evaluated biomimetic bone-like carbonate apatite (BCA) and octacalcium phosphate (OCP) coatings which were deposited on Ti-6Al-4V samples to evaluate their in vitro and in vivo dissolution properties. The coated plates were soaked in α-MEM for 1, 2, and 4 weeks, and they were analyzed by Back Scattering Electron Microscopy (BSEM) and by Fourier Transform Infra Red spectroscopy (FTIR). Identical coated plates were implanted subcutaneously in Wistar rats for similar periods. BSEM, FTIR, and histomorphometry were performed on the explants. In vitro and in vivo, a carbonate apatite (CA) formed onto OCP and BCA coatings via a dissolution-precipitation process. It was reported that (i) in vitro, both coatings dissolved overtime, whereas in vivo BCA calcified and OCP partially dissolved after 1 week, and (ii) thereafter, OCP remained stable. This different in vivo behavior can be attributed to (1) different organic compounds that might prevent or enhance Ca-P dissolution, (2) a greater reactivity of OCP due to its large open structure, or (3) different thermodynamic stability between OCP and BCA phases. It was, therefore, concluded that these structural and compositional differences promote either the progressive loss or calcification of the Ca-P coating, and might lead to different osseointegration of coated implants [190]. Barrère et al. [191] studied the nucleation and growth of a calcium phosphate (Ca-P) coating deposited on titanium implants from simulated body fluid (SBF), using atomic force microscopy (AFM) and environmental scanning electron microscopy (ESEM). Forty titanium alloy plates were assigned into two groups: a smooth surface group having a maximum roughness Rmax < 0.10 μm, and a rough surface group with an Rmax < 0.25 μm. Titanium samples were immersed in SBF concentrated by five (SBF × 5) from 10 min to 5 hours and examined by AFM and ESEM. It was observed that scattered Ca-P deposits of approximately 15 nm in diameter appeared after only 10 min of immersion in SBF × 5, and (ii) these Ca-P deposits grew up to 60–100 nm after 4 h on both smooth and rough Ti-6Al-4V substrates. It was found that (i) a continuous Ca-P film formed on titanium substrates, (ii) a direct contact between the Ca-P coating and the Ti-6Al-4V surface was observed, and (iii) the Ca-P coating was composed of nanosized deposits and of an interfacial glassy matrix, which might ensure the adhesion between the Ca-P coating and the Ti-6Al-4V substrate. The Ca-P coating detached from the smooth substrate, whereas the Ca-P film extended onto the whole rough titanium surface over time. In the case of rough Ti-6Al-4V, the Ca-P coating evenly covered the substrate after immersion in SBF × 5 for 5 h. Accordingly, it was suggested that (i) the heterogeneous nucleation of Ca-P on titanium was immediate and did not depend on the Ti-6Al-4V surface topography, and (ii) the further growth and mechanical attachment of the final Ca-P coating strongly depended on the surface, for which a rough topography was beneficial [191].
The bioconductivity of a new biomedical Ti-29Nb-13Ta-6Zr alloy was achieved by a combination of surface oxidation and alkaline treatment [192]. It was reported that (i) immersion in a protein-free SBF and fast calcification solution led to the growth of CP phase on the oxidized and alkali-treated (10 M NaOH at 40 °C for 24 h) alloy, and the new bioconductive surface was still harder than the substrate, (ii) oxidation at 400 °C × 24 h led to the formation of a hard layer, (iii) the oxides are mainly composed of TiO2, with small amounts of Nb2O5 and ZrO2; an oxygen diffusion layer exists beneath the surface oxide layer, and (iv) a titanate layer forms on the pre-oxidized surface after alkali treatment, and growth of a layer of Ca-P has been successfully induced on the titanate layer by immersing in SBF [192]. Bogdanski et al. [193] coated NiTi with CP by dipping in oversaturated CP solution. Since polymorphonuclear neutrophil grannucytes (PMN) belong to the first cells which will adhere to implant materials, the apoptosis of isolated human PMN after cell culture on non-coated and CP-coated SME NiTi was analyzed by light and scanning electron microscopy and flow cytometry. It was found that (i) in contrast to PMN adherent to non-coated TiNi, the apoptosis of PMN adherent CP-coated samples was inhibited, and (ii) cell culture media obtained from cultured leukocytes with CP-coated were able to transfer the apoptosis inhibiting activities to freshly isolated PMN [193]. The compositions of the surface and the interface of calcium phosphate ceramic (CP) coatings electrophoretically deposited and sintered on CpTi and Ti-6Al-4V were evaluated before and after four weeks immersion in a simulated physiological solution. In the CP coating-metal interface, it was mentioned that (i) the phosphorus diffused beyond the Ti oxide layer, resulting in the depleted phosphorus in the ceramic adjacent to the metal, and (ii) the surface of the ceramic, however, was substantially unchanged [194].
The possible mechanisms of minimization of prosthesis-derived bone growth inhibitors by shielding of the metal and reduction of the associated metal dissolution was investigated by Ducheyne et al. [195]. Ti, Al, and V release rates were determined in vitro for Ti-6Al-4V alloy both with and without a CP coating. Surfaces were passivated in 40% HNO3 at 55 °C for 20 min. Calcium phosphate ceramic was electrophoretically deposited and subsequently vacuum sintered at 925 °C for 2 h. Immersion for 1, 2, 4, 8, and 16 weeks in Hank’s balanced salt solution, simulated physiological solution with 1.5 mM DS-EDTA. It was found that (i) the CP-coated specimens contained no measurable amounts of Ti, (ii) the Al ion solution around the CP-coated specimen was significantly greater than the concentration around the non-coated specimens; however (iii) Al did not increase significantly with time, at least up to 4 weeks immersion. The CP coating produced a significant increase of biological fixation, yet at the same time a greater Al release into solution, calling into question the value of calcium phosphate ceramic coating in shielding adverse metal passive dissolution to enhance bone growth [195].
Similarly to HA coating, there are various (chemical, electrochemical, and physical) processes for Ca-P coating available. The ions were implanted in sequence, first Ca and then P, both at a dose of 1017 ions/cm2 at beam energy of 25 keV on CpTi (grade 2). The corrosion tests were done in SBF at 37 °C. It was concluded that (i) the CpTi surface subjected to two-step implantation of Ca+P at a dose of 1017 ions/cm2 becomes amorphous, (ii) the implantation of Ca+P ions increases the corrosion resistance in SBF exposure at 37 °C for up to 3,200 h, (iii) during exposures to SBF, CP precipitates form the implanted as well as non-implanted samples, (iv) the CP precipitates have no effect on the corrosion resistance, and (v) under the conditions of the applied examination, the biocompatibility of Ti subjected to the two-step implantation of Ca+P ions was similar to that of untreated sample [196]. CpTi (grade 2) were surface-modified by anodic spark discharge anodization and a thin layer (ca. 5 μm) of amorphous TiO2 containing Ca and P - Ti/AM. Some of the Ti/AM samples were further modified by hydrothermal treatment to produce a thin outermost (ca. 1 μm) of HA (Ti/AM/HA). It was reported that (i) non-anodized vacuum annealed and hydrofluoric acid etched samples, used as control material, showed good bone adsorption producing Ti excellent surface properties, (ii) anodization at a voltage of 275 V that produced a crack free amorphous TiO2 film containing Ca and P (Ti/AM) provided good results for cytocompatibility, important morphological characteristics (micropores without crack development) but presented the lowest bone apposition probably due to the degradation of amorphous TiO2 film, and (iii) hydrothermal treatment at 300 °C for 2 h that produced a sub-micrometer layer of HA crystals (ca. 1 μm) upon the amorphous TiO2 film (Ti/AM/HA) gave rise to the highest bone apposition only at 8 weeks [81].
Coating by a RF magnetron sputter technique for the production of thin Ca-P coatings can be produced that vary in Ca/P ratio as well as structural appearance. Jansen et al. [197] studied the effect of non-coated titanium and three different Ca-P sputtered surfaces on the proliferation and differentiation (morphology and matrix production) of osteoblast-like cells. It was found that (i) proliferation of the osteoblast-like cells was significantly higher on non-coated than on Ca-P coated samples, (ii) on the other hand, more mineralized extracellular matrix was formed on the coated surfaces, and (iii) TEM confirmed that the cells on the coated substrates were surrounded by extracellular matrix with collagen fibres embedded in crystallized needle shaped structures. On the basis of these findings, it was concluded that (i) the investigated Ca-P sputter coatings possess the capacity to activate the differentiation and expression of osteogenic cells, and (ii) bone formation proceeds faster on Ca-P surfaces than on Ti substrates. Further, it was noticed that this bone stimulating effect appeared to be independent of the Ca-P ratio of the deposited coatings [197].
Hayakawa et al. [198] prepared four types of Ti implants: (1) as-blasted with titania (<150 μm), (2) sintered with Ti beads with 50–150 μm diameter, (3) blasted with IBDM (ion beam dynamic mixing) Ca-P coating, and (4) multilayered sintered implants with IBDM Ca-P coating. The Ca-P coating was rapid heat-treated with infrared radiation at 700 °C. The implants were inserted into the trabecular bone of the left and right femoral condyles of 16 rabbits, for 2, 3, 4, and 12 weeks. It was reported that (i) histological evaluation revealed new bone formation around different implant materials after already 3 weeks of implantation, (ii) after 12 weeks, mature trabecular bone surrounded all implants, and (iii) the combination of surface geometry and Ca-P coating benefits the implant-bone response during the healing phase [198].
Biomimetic deposition and electrochemical deposition in solution were used for CP deposition on sintered CpTi bead-cylinders (1,300 °C for 2 h under vacuum). The supersaturated solution for CP deposition was prepared by dissolving NaCl, CaCl2, Na2HPO4·2H2O in distilled water buffered to pH 7.4. Biomimetic deposition was done by immersing samples in solution at 37 °C for 1, 2, 4, and 6 days, while electrochemical deposition was performed at −1.8 V (vs. Hg2SO4/Hg reference) for 2 h at 37 °C. It was reported that (i) a pre-coating alkaline treatment (5 mol/L NaOH at 60 °C for 5 h) is necessary to obtain a uniform coating layer on the inner pore surfaces when the biomimetric deposition is used, (ii) electrochemical deposition is more efficient and less sensitive to the conditions of the Ti surfaces compared to the biomimetric depositions; however (iii) the electrochemical deposition produces less uniform and thinner coating layers on the inner pore surfaces compared with the biomimetric deposition, and (iv) the crystal structure of the deposited Ca-P is octacalcium phosphate regardless of the deposition methods [199]. Heimann et al. [200] deposited hydroxyapatite and duplex hydroxyapatite + titania bond coat layers onto Ti-6Al-4V substrates by atmospheric plasma spraying at moderate plasma enthalpies. From as-sprayed coatings and coatings incubated in simulated body fluid, electron-transparent samples were generated by focused ion beam excavation. It was found that (i) adjacent to the metal surface a thin layer of amorphous calcium phosphate (CP) was deposited, (ii) after in vitro incubation of duplex coatings for 24 weeks, Ca-deficient defect apatite needles precipitated from amorphous CP, and (iii) during incubation of hydroxyapatite without a bond coat for 1 week diffusion bands were formed within the amorphous CP of 1–2 μm width parallel to the interface metal/coating, presumably by a dissolution-precipitation sequence [200]. Aluminoborosilicate glass + HA powder were coated on CpTi at 800–900 °C. It was found that (i) the precipitates were Ca-deficient carbonate with low crystallinity, (ii) both morphologies and composition of the precipitates in vivo were similar to those in vitro, and (iii) the HA particles on the surface of the composite act as nucleation sites for precipitation in physiologic environments, whereas the glass matrix is independent of it [201].
For the same purpose as done for the HA-coating, a Ca-P coated layer is needed to post-coating heat treatment. Lucas et al. [202] investigated the post-deposition heat treatments of ion beam as-sputtered coatings by varying the time and temperature. Ca-P coatings, deposited using a hydroxyapatite-fluorapatite target, received the post-deposition heat treatments as follows: 150, 300, 400, 500, and 600 °C for either 30 min or 60 min, followed by furnace cooling to room temperature. It was found that (i) the average bond strengths for the coatings heated to 500 °C (30 min), 600 °C (30 min) and 600 °C (1 h) were 40.1 MPa, 13.8 MPa and 9.1 MPa, respectively. The 500 °C heat treatment for 30 minute resulted in an HA-type coating without reducing the tensile bond strength of the coating. Thus, temperature and time are critical parameters in optimizing coating properties [202], suggesting that the diffusion process must be significantly involved.
Gan et al. [203] evaluated the interface shear strength of Ca-P thin films applied to Ti-6Al-4V substrates using a substrate straining method - a shear lag model. The Ca-P films were synthesized using sol-gel methods from either an inorganic or organic precursor solution. Strong interface bonding was demonstrated for both film types. It was reported that (i) the films were identified as non-stoichiometric hydroxyapatite, but with different Ca/P ratios, (ii) the Ca-P films were 1–1.5 μm thick, and (iii) the shear strength was approximately 347 and 280 MPa for inorganic and organic route-formed films, respectively [203]. NiTi was coated with CP by dipping in oversaturated CP solution, with layer thickness (5–20 μm). The porous nature of the layer makes it mechanically stable enough to withstand both the shape memory transition upon cooling and heating and also strong bending of materials (superelasticity). The adhence of human leukocytes and platelets to the AP layer was analyzed in vivo. By comparison, in non-coated SME-NiTi, it was reported that leukocytes and platelets showed a significantly increased adhesion to the coated NiTi [204]. Fend et al. [205] pre-treated CpTi: by immersing in boiling Ca(OH)2 solution for 30–40 min to have only Ca-added pre-calcification, immersing in 20% H3PO4 solution at 85–95 °C for 30 min to have only P-added pre-phosphatization, and immersed these pre-calcified CpTi in supersaturated CP solution at 37 °C for 1 week. It was found that (i) the osteoblast amount and activity on the surfaces containing Ca are higher than those on the surface containing sole P ions, (ii) Ca2+ ion sites on the material surfaces favor protein adsorption, such as fibronectin or vitronectin as ligands of osteoblast, onto the surface due to positive electricity, chemical, and biological function, and (iii) on the AP surfaces, Ca2+ ions are the active sites of the osteoblast adhesion and also promote the cell adhesion on PO43– ion sites [205].
As we have seen above, there are many evidence indicating that, in vitro tests, calcium phosphate is precipitated on such surfaces when they are immersed in a simulated physiological solution, suggesting a main reason for excellent biocompatibility (e.g., [181]). On the other hand, there are evidence suggesting that, in vivo, after even seven day implantation of such modified Ti in rat bone, it was found that Ca/P ratio reduced less than 1 (i.e., Ca-depletion) [206].

3.4.4.  Composite Coating
Bioactive calcium phosphate (CaP) coatings were produced on titanium by using phosphate-based glass (P-glass) and hydroxyapatite (HA), and their feasibility for hard tissue applications was addressed in vitro by Kim et al. [207]. P-glass and HA composite slurries were coated on Ti under mild heat treatment conditions to form a porous thick layer, and then the micropores were filled in with an HA sol-gel precursor to produce a dense layer. The resultant coating product was composed of HA and calcium phosphate glass ceramics, such as tricalcium phosphate (TCP) and calcium pyrophosphate (CPP). It was reported that the coating layer had a thickness of approximately 30–40 μm and adhered to the Ti substrate tightly, (ii) the adhesion strength of the coating layer on Ti was as high as about 30 MPa, (iii) the human osteoblastic cells cultured on the coatings produced by the combined method attached and proliferated favorably, and (iv) the cells on the coatings expressed significantly higher alkaline phosphatase activity than those on pure Ti, suggesting the stimulation of the osteoblastic activity on the coatings [207]. Maxian et al. [208] evaluated the effect of amorphous calcium phosphate and poly-crystallized (60% crystalline) HA coatings on bone fixation of smooth and rough (Ti-6Al-4V powder sprayed) Ti-6Al-4V implants after four and 12 weeks of implantation in a rabbit trascortical femoral model. Histological evaluation of amorphous versus poorly crystallized HA coatings showed significant differences in bone apposition and coating resorption that were increased within cortical compared to cancellous bone. The poorly crystallized HA coatings showed the most degradation and least bone apposition. Mechanical evaluation, however, showed no significant differences in push-out shear strengths. Significant enhancement in interfacial shear strengths for bioceremic coated, as compared to uncoated implants, was seen for smooth-surfaced implants (3.5–5 times greater) but not for rough-surfaced implants at four and 12 weeks. Based on these results, it was suggested that once early osteointegration is achieved biodegradation of a bioactive coating should not be detrimental to the bone/coating/implant fixation [208]. Plasma sprayed HA coatings on titanium alloy substrates have been used extensively due to their excellent biocompatibility and osteoconductivity. However, the erratic bond strength between HA and Ti alloy has raised concern over the long-term reliability of the implant. Accordingly, Khor et al. developed HA/yttria-stabilized-zirconia (YSZ)/Ti-6Al-4V composite coatings that possess superior mechanical properties to conventional plasma sprayed HA coatings [209]. Ti-6Al-4V powders coated with fine YSZ and HA particles were prepared through a unique ceramic slurry mixing method. The composite powder was employed as feedstock for plasma spraying of the HA/YSZ/Ti-6Al-4V coatings. The influence of net plasma energy, plasma spray standoff distance, post-spray heat treatment on microstructure, phase composition, and mechanical properties were investigated. It was found that (i) coatings prepared with the optimum plasma sprayed condition showed a well-defined splat structure, (ii) HA/YSZ/Ti-6Al-4V solid solution was formed during plasma spraying, which was beneficial for the improvement of mechanical properties, (iii) the microhardness, modulus of elasticity, fracture toughness, and bond strength increased significantly with the addition of YSZ, and (iv) post-spray heat treatment at 600 °C and 700 °C for up to 12 h was found to further improve the mechanical properties of coatings [209]. Yttria stabilized zirconia (YSZ) is often used as reinforcement for many ceramics because it has the merits of high strength and enhanced toughening characteristics during crack-particle interactions [210–213].
Yamada et al. [214] utilized the Cullet method for which (1) the mixture of HA powder and glass frits are sintered at 900–1000 °C for from 5 to 10 min to prepare well homogenized coating powder, whereas the conventional method is just mixing and not sintering, and (2) the time of etching treatment, through which the bioactive surface is formed using the mixed solution of HNO3 and HF, is relatively short (within 1 min) compared to the conventional method. Through this method, functionally gradient HA/Ti composite implants were successfully fabricated with higher quality compared with the conventional method [214]. Suzuki et al. [215] coated titanium dioxide onto silicone substrates by radio-frequency sputtering. It was reported that silicone coating with titanium dioxide enhanced the breakdown of peroxynitrite by 79%. Titanium dioxide-coated silicone inhibited the nitration of 4-hydroxy-phenylacetic acid by 61% compared to aluminum oxide-coated silicone and 55% compared to uncoated silicone. Titanium dioxide-coated silicone exhibited a 55% decrease in superoxide compared to uncoated silicone, and a 165% decrease in superoxide compared to uncoated polystyrene [215]. Titanium dioxide-coated silicone inhibited IL-6 production by 77% compared to uncoated silicone. Based on these findings, it was concluded that the anti-inflammatory properties of titanium dioxide can be transferred to the surfaces of silicone substrates [215].
HA coatings with titania addition were produced by the high velocity oxy-fuel spray process by Li et al. [216]. It was found that (i) the addition of TiO2 improves the MOE, fracture toughness, and shear strength of high velocity oxy-fuel sprayed HA-based coatings, (ii) the incorporation of the secondary titania phase is found to have a negative effects on the adhesive strength of high velocity oxy-fuel sprayed HA coatings, (iii) the titania is found to inhibit the decomposition of HA at evelated temperatures lower than 1,410 °C, at which point the mutual chemical reaction occurs, and (iv) a small amount of TiO2 added into high velocity oxy-fuel sprayed HA coatings with less than 20 vol% is therefore recommended for strengthening of HA-coatings [216]. Lu et al. [217] fabricated a two-layer hydroxyapatite (HA)/HA + TiO2 bond coat composite coating (HTH coating) on titanium by the plasma spraying technique. The HA + TiO2 bond coat (HTBC) consists of 50 vol% HA and 50 vol% TiO2 (HT). The as-sprayed HT coating consists mainly of crystalline HA, rutile TiO2 and amorphous Ca-P phase, but the post-spray heat treatment at 650 °C for 120 min effectively restores the structural integrity of HA by transforming non-HA phases into HA [216]. It was found that there exists interdiffusion of the elements within the HTBC, but no chemical product between HA and TiO2, such as CaTiO3 was formed. The toughening and strengthening mechanism of HTBC is mainly due to TiO2 as obstacles resisting cracking, and the reduction of the near-tip stresses resulting from stress-induced microcracking [217]. Ng et al. [218] mimicked bio-mineralization of bone by applying an initial TiO2 coating on Ti-6Al-4V by electrochemical anodization in two dissimilar electrolytes, followed by the secondary calcium (CaP) coating, subsequently applied by immersing the substrates in a simulated body fluid (SBF) with three times concentration (SBF × 3). Electrochemical impedance spectroscopy (EIS) and DC potentiodynamic polarization assessments in SBF revealed that the anodic TiO2 layer is compact, exhibiting up to a four-fold improvement in in vitro corrosion resistance over unanodised Ti-6Al-4V. X-ray photoelectron spectroscopy analysis indicates that the anodic Ti oxide is thicker than air-formed ones with a mixture of TiO2–x compound between the TiO2/Ti interfaces. The morphology of the dense CaP film formed, when observed using scanning electron microscopy, is made up of linked globules 0.1–0.5 μm in diameter without observable delamination. It was also found that (i) the calculated Ca:P ratios of all samples (1.14–1.28) are lower than stoichiometric hydroxyapatite (1.67), and (ii) a duplex coating consisting of a compact TiO2 with enhanced in vitro corrosion resistance and bone-like apatite coating can be applied on Ti-6Al-4V by anodization and subsequent immersion in SBF [218].
Knabe et al. [219] investigated the effects of novel calcium titanium, calcium, titanium zirconium phosphates suitable for plasma spraying on CpTi substrate on the expression of bone-related genes and proteins of human bone-derived cells, and compared the effects to that on native Ti and HA-coated Ti. Test materials were acid etched and sand-blasted, plasma-sprayed HA, and sintered CaPO4 with Ti, Zr, TiO2, and ZrO2. Human bone-derived cells were grown on these surfaces for 3, 7, 14, and 21 days, counted and probed for various mRNAs and proteins. It was reported that (i) all surfaces significantly affected cellular growth and the temporal expression of an array of bone-related genes and proteins, (ii) at 14 and 21 days, cells on sintered displayed significantly enhanced expression of all osteogenetic mRNAss, and (iii) surfaces of 55CaO·20TiO2·31P2O5 and CaTi4(PO4)6 had the lost effect on osteoblastic differentiation inducing a greater expression on an array of osteogenetic markers than recorded for cells grown on HA, suggesting that these novel materials may possess a higher potency to enhance osteogenesis [219]. Shtansky et al. [220] performed a comparative investigation of multicomponent thin films based on the systems Ti-Ca-C-O-(N), Ti-Zr-C-O-(N), Ti-Si-Zr-O-(N) and Ti-Nb-C-(N). TiC0.5 + 10%CaO, TiC0.5 + 20%CaO, TiC0.5 + 10%ZrO2, TiC0.5 + 20%ZrO2, Ti5Si3 + 10%ZrO2, TiC0.5 + 10%Nb2C and TiC0.5 + 30%Nb2C composite targets were manufactured by means of self-propagating high-temperature synthesis, followed by DC magnetron sputtering in an atmosphere of argon or in a gaseous mixture of argon and nitrogen. The biocompatibility of the films was evaluated by both in vitro and in vivo experiments. The in vitro studies involved the investigation of the proliferation of Rat-1 fibroblasts and IAR-2 epithelial cells on the tested films, morphometric analysis and actin cytoskeleton staining of the cells cultivated on the films. In vivo studies were fulfilled by subcutaneous implantation of Teflon plates coated with the tested films in mice and analysis of the population of cells on the surfaces. It was reported that (i) the films showed high hardness in the range of 30–37 GPa, significant reduced modulus of elasticity, low friction coefficient down to 0.1–0.2, and low wear rate in comparison with conventional magnetron-sputtered TiC and TiN films, (ii) no statistically significant differences in the attachment, spreading, and cell shape of cultured IAR-2 and Rat-1 cells on the Ti-Ca-C-O-(N), Ti-Zr-C-O-(N) (TiC0.5+10%ZrO2 target), Ti-Si-Zr-O-(N) films and the uncoated substrata was detected, and (iii) the adhesion and proliferation of cultured cells in vitro was perfect at all investigated films. Based on these findings, it was concluded that the combination of excellent mechanical properties with non-toxicity and biocompatibility makes Ti-Ca-C-O-N, Ti-Zr-C-O-N, and Ti-Si-Zr-O-N films promising candidates as tribological coatings to be used for various medical applications like total joint prostheses and dental implants [220]. Von Walter et al. [221] introduced a porous composite material, named “Ecopore”, and described in vitro investigation of the material and its modification with fibronectin. The material is a sintered compound of rutile TiO2 and the volcanic silicate perlite with a macrostructure of interconnecting pores. In an in vitro model, human primary osteoblasts were cultured directly on Ecopore. It was reported that human osteoblasts grew on the composite as well as on samples of its single constituents, TiO2 and perlite glass, and remained vital, but cellular spreading was less than on tissue culture plastic. To enhance cellular attachment and growth, the surface of the composite was modified by etching, functionalization with aminosilane and coupling of fibronectin, resulting in greatly enhanced spreading of human osteoblasts. It was therefore concluded that (i) Ecopore is non-toxic and sustains human osteoblasts growth, cellular spreading being improvable by coating with fibro-nectin, and (ii) the composite may be usable in the field of bone substitution [221].
Different biomaterials have been used as scaffolds for bone tissue engineering. Rodrigues et al. [222] characterized biomaterial composed of sintered (at 1,100 °C) and powdered hydroxyapatite and type I collagen (both of bovine origin) designs for osteoconductive and osteoinductive scaffolds. Collagen/HA proportions were 1/2.6 and 1/1 by weight, with particle sizes ranging from 200 to 400 μm. X-ray diffraction and infrared spectroscopy showed that the sintered (1,100 °C) bone was composed essentially of HA with minimum additional groups as surface calcium hydroxide, surface and crystal water, free carbon dioxide, and possibly brushite. It was reported that osteoblasts adhered and spread on both the HA particle surface and the collagen fibers, which seemed to guide cells between adjacent particles [222], suggesting that this biocomposite can be considered as ideal for its use as scaffold for osteoconduction and osteoinduction. Cheng et al. [223] prepared electrochemically a bovine serum albumin (BSA) protein-containing AP coating on a HA coated Ti-6Al-4V. It was reported that (i) the method resulted in a 70-fold increase in BSA inclusion compared to simple adsorption, and was subsequently released by a slow mechanism (15% loss over 70 h), and (ii) thus, this technique provides an efficient method of protein incorporation at physiological stem, with a potential for sustained release of therapeutic agents, as may be required for metallic implant fixation [223].
Redepenning et al. [224] prepared another type of biocomposite coatings containing brushite (CaHPO4·2H2O) and chitosan by electrochemical deposition. The brushite/chitosan composites were converted to hydroxyapatite/chitosan composites in aqueous solutions of sodium hydroxide. The coatings ranged from about 1 to 15% chitosan by weight. It was mentioned that qualitative assessment of the coatings showed adhesion significantly improved over that observed for electrodeposited coatings of pure HA [224].

3.4.5.  TiN Coating
In spite of their high strength, low density, and good corrosion resistance, the usefulness of Ti alloys in general engineering components is frequently limited by their poor wear resistance. If the alloy surface is subjected to conditions of sliding or fretting, adhesive wear can rapidly lead to catastrophic failure unless appropriate surface engineering is carried out. In order to combat modest contact loads, several surface treatments are commercially available, such as plasma nitriding or PVD coating with TiN. Titanium nitride is known for its high surface hardness and mechanical strength. It was also reported that the dissolution of Ti ions is very low [225]. As for dental implants, they are comprised of various components. The implant abutment part (the mucosa penetration part) is exposed in the oral cavity, and so plaque and dental calculus easily adhere on it. Removal of the plaque and dental calculus is necessary to obtain a good prognosis throughout the long term maintenance of the implant. Based on this background, Kokubo et al. [226] prepared CpTi (grade 1) samples by polishing with #2000grit paper, or buff-polishing with 6 μm diamond emulsion paste, followed by a 0.1% HF acid solution for 10 s to clean the surface, then treated in N2 atmosphere of 1 atm at 850 °C for 7 h (N2 flow rate: 50 L/min). It was reported that (i) the nitrided layer about 2 μm thick composed of TiN and Ti2N was formed on Ti by a gas nitriding method, and the dissolved amount of Ti ion in SBF (simulated body fluid) was as low as the detectable limit of ICP-MS (Inductively Coupled Plasma Mass Spectroscopy), and that the 1% lactic acid showed no significant difference from Ti [226]. SBF, in genral, consists of Na+ (142.0), K+ (5.0), Mg2+ (1.5), Ca2+ (2.5), Cl−(148.8), HCO3− (4.2), HPO4− (1.0). H.P. Na (142.0), K (5.0), Mg (1.5), Ca (2.5), Cl (103.0), HCO3 (13.5), and HPO4 (1.0).
Surface topography and chemistry have been shown to be extremely important in determining cell-substrate interactions and influencing cellular properties such as cell adhesion, cell-cell reactions, and cytoskeletal organization [227]. The cell-substrate interaction of primary hippocampal neurones with thin films of TiN was studied in vitro. TiN films of different surface chemistries and topographies were deposited by pulsed DC reactive magnetron sputtering and closed field unbalanced magnetron sputter ion plating to result in TiN thin films with similar surface chemistries, but different topographical features. It was reported that (i) primary hippocampal neurones were found to attach and spread to all of the TiN films, (ii) neuronal network morphology appeared to be more preferential on the nitrogen rich TiN films, and also reduced nanotopographical features, (iii) at early time points of one and four days in vitro primary hippocampal neurones respond to the presence of interstitial nitrogen rather than differences in nanotopography; however (iv) at seven days more preferential neutronal network morphology is formed on TiN thin films with lower roughness values and decreased size of topographical features [227]. Bull et al. [228] studied the use of thin titanium interlayers to promote the adhesion of TiN coatings on a range of substrates. For thin interlayers, an interstitially hardened titanium layer is formed at the interface, resisting the interfacial crack propagation. However, at a critical interlayer thickness, the surface contaminants are completely dissolved in the interfacial layer, and depositing any further titanium leads to an overall softer interfacial layer which offers less resistance to crack propagation, and delamination can easily take place. For this reason, failure is observed within the interlayer for thick interlayers, whereas it occurs at the interlayer/substrate interface for thinner interlayers. Another contribution to the enhanced adhesion comes from the reduction in coating stresses in the interfacial region due to the presence of a soft compliant layer, which was examined by changing the hardness of the interlayer deposited before coating deposition. It was concluded that (i) softer interlayers do not lead to improved adhesion performance in most cases, and (ii) it appears that the best adhesion results from a hard interlayer that leads to ductile failure at the coating/substrate interface, rather than the brittle failure observed due to the presence of oxide films [228].
Mechanical-electrochemical interactions accelerate corrosion in mixed-metal modular hip prostheses. These interactions can be reduced by improving the modular component machining tolerances or by improving the resistance of the components to scratch or fretting damage. Wrought Co-alloy (Co-Cr-Mo) is known to have better tribological properties compared to the Ti-6Al-4V alloy. Thus, improving the tribological properties of this mixed-metal interface should center on improving the tribological properties of Ti-6Al-4V. It was mentioned that (i) the nitrogen-diffusion-hardened Ti alloy samples had a more pronounced grain structure, more nodular surface, and significantly higher mean roughness values than the control Ti-6Al-4V, (ii) the nitrided Ti-6Al-4V samples also exhibited at least equivalent corrosion behavior and a definite increases in surface hardness compared to the control Ti-6Al-4V samples, and (iii) fretting can be reduced by decreasing micromotion or by improving the tribological properties (wear resistance and surface hardness) of the material components at this interface [229]. The corrosion behavior of the titanium nitride-coated TiNi alloy was examined in 0.9% NaCl solution by potentiodynamic polarization measurements and a polarization resistance method [230]. XPS spectra showed that the titanium nitride film consisted of three layers, a top layer of TiO2, a middle layer of TiNx (x > 1), and an inner layer of TiN, which agreed very well with results obtained by Oshida et al. [231]. The passive current density for the titanium nitride-coated alloy was approximately two orders of magnitude lower than that of the polished alloy in the potential range from the free corrosion potential to +500 mV (vs. Ag/AgCl). Pitting corrosion associated with breakdown of the coated film occurred above this potential. The polarization resistance data also indicated that the corrosion rate of the titanium nitride-coated alloy at the corrosion potential (+50 to +100mV) was more than one order of magnitude lower than that for the polished alloy. It was concluded that the corrosion rate of TiNi alloy can be reduced by more than one order of magnitude by titanium nitride coating, unless the alloy is highly polarized anodically in vivo [230]. Bordji et al. [232] prepared Ti alloys treated by: (1) glow discharge nitrogen implantation (1017 atoms cm−2), (2) plasma nitriding by plasma diffusion treatment, and (3) deposition of TiN layer by plasma-assisted chemical vapour deposition additionally to plasma diffusion treatment. A considerable improvement was noticed in surface hardness, especially after the two nitriding processes. A cell culture model using human cells was chosen to study the effect of such treatments on the cytocompatibility. The results showed that Ti-5Al-2.5Fe alloy was as cytocompatible as the Ti-6Al-4V alloy, and the same surface treatment led to identical biological consequences on both alloys. It was concluded that (i) after the two nitriding treatments, cell proliferation and viability appeared to be significantly reduced and the SEM revealed somewhat irregular surface states; however (ii) osteoblast phenotype expression and protein synthesis capability were not affected [232]. Goldberg et al. [233] utilized the plasma vapor deposition technique, by which the samples were placed into a vacuum chamber and sputtered to remove the oxide film, followed by depositing a 200 nm thick interlayer of titanium to enhance the coating/substrate interface. Alternating layers of TiN and AlN were deposited until a coating thickness of approximately 5 μm was produced. The mechanical and electro-chemical behavior of the surface oxides of Co-Cr-Mo and Ti-6Al-4V alloys during fracture and repassivation play an important role in the corrosion of the taper interfaces of modular hip implants. These corrosion properties were investigated in one group of Co-Cr-Mo and Ti-6Al-4V alloy samples passivated with nitric acid and another group coated with TiN/AlN coating. It was found that (i) Co-Cr-Mo had a stronger surface oxide and higher interfacial adhesion strength, making it more resistant to fracture than Ti-6Al-4V, (ii) if undistributed, the oxide on the surface of Ti-6Al-4V significantly reduced dissolution currents at a wider range of potential than Co-Cr-Mo, making Ti-6Al-4V more resistant to corrosion, (iii) the TiN/AlN coating had higher hardness and modulus of elasticity than Co-Cr-Mo and Ti-6Al-4V. It was much less susceptible to fracture, had higher interfacial adhesion strength, and was a better barrier to ionic diffusion than the surface oxides on Co-Cr-Mo and Ti-6Al-4V [233].
The most of surface treatments such as plasma nitriding or PVD coating with TiN are, however, carried out in the solid state and the depth of coating or hardening is restricted by low diffusivity. The diffusion coefficient of nitrogen in Ti is more than a thousand times slower than that in steels due to different crystalline structures. The packing factor of Ti (HCP) is 74%, while that of steel (BCC) is 68%, so that steel has more spaces available for diffusing species. In order to achieve the depth of hardening necessary to withstand the subsurface Hertzian stresses induced by heavy rolling contact, it is necessary to alloy the Ti surface in the molten state. The necessary depth of surface hardening can readily be achieved in this way by laser melting the surface in the presence of interstitial alloying elements such as carbon, oxygen, and nitrogen. Of these, nitrogen has been found to provide the best balance between increased hardness and decreased ductility, and can easily be added by laser gas nitriding. The Hertzian compressive stress in the substrate was increased to 1.36 GPa [234]. Pure iron has allotropic phase transformations: the first one is 910 °C between α-BCC and γ-FCC, and the second one is 1,390 °C between γ-FCC and δ-BCC. While investigating the deformation mechanism of transformation superplastivity, Oshida observed that the transformation front behaves as if sime-liquid due to loosing a clear crystalline electorn diffraction pattern even both α-BCC and γ-FCC are still in solid state. Based on this finding, the “semi-liquid trans-formation front” model was proposed. One of various applications using transformation superplasticity is a nitridation of metallic materials if they have an allotropic phase transformation temperature. It was demonstrated that CpTi was succeesfully nitrided when CpTi was heated and cooled repeatedly passing the β-transus temperature (between 800 °C and 930 °C) for several times in nitrogen gas filled chamber [231]. Ion implantation, diffusion hardening, and coating are surface modification techniques for improving the wear resistance and surface hardness of Ti alloy surfaces [235–239]. A Ti-6Al-4V sample was diffusion-hardened in a nitrogen atmosphere for 8 h at 566 °C and argon or nitrogen quenched to room temperature. The nitrogen-diffusion-hardened Ti-6Al-4V had TiN and TiNO complexes at the immediate surface and sub-surface layers. The diffusion-hardened samples also had a deeper penetration of oxygen compared to regular Ti-6Al-4V samples [240].
As briefly described in the above, Oshida et al. [231,241] applied a TiN coating onto CpTi substrate prior to porcelain firing to develop a new method to control the excessive oxidation. The bonding strength of porcelain to metals depends on the oxide layer between the porcelain and the metal substrate. Oxidation of a metal surface increases the bonding strength, whereas excessive oxidation decreases it. Titanium suffers from its violent reactivity with oxygen at high temperatures that yield an excessively thick layer of TiO2, and this presents difficulties with porcelain bonding. The oxidation kinetics of titanium simulated to porcelain firing was investigated, and the surface nitridation of CpTi as a process of controlling the oxidation behavior was evaluated. Nitrided samples with the arc ion plating PVD process and un-nitrided control CpTi were subjected to oxidation simulating of Procera porcelain with 550, 700, and 800 °C firing temperatures for 10 min in both 1 and 0.1 atmospheric air. The weight difference before and after oxidation was calculated, and the parabolic rate constant, Kp (mg2/cm4/s), was plotted against inverse absolute temperature (i.e., in an Arrhenius plot). Surface layers of the samples were subjected to x-ray and electron diffraction techniques for phase identifications. Results revealed that both nitrided and un-nitrided samples obey a parabolic rate law with activation energy of 50 kcal/mol. In addition, the study shows that nitrided CpTi had a Kp about 5 times lower than the un-nitrided CpTi, and hence the former needs 2.24 times longer oxidation time to show the same degree of oxidation. Phase identification resulted in confirming the presence of TiO2 as the oxide film in both groups, but with 1–2 μm thickness for the un-nitrided CpTi and 0.3–0.5 μm thickness for nitrided samples. Therefore, it can be concluded that nitridation of titanium surfaces can be effective in controlling the surface oxide thickness that might ensure satisfactory bonding with porcelain [231]. Oshida et al. [241] evaluated CpTi substrates subjected to porcelain firing and bond strengths under three-point bending mode (span length: 15mm; crosshead speed: 0.5 mm/min). Experimental variables included surface treatments of CpTi and porcelain firing schedules. Variables for the surface treatments were: (1) sandblasting, (2) mono- and triple-layered nitridation, and (3) mono-layered chrome-doped nitridation. Variables for the porcelain firing schedule included (4) bonding agent application, (5) bonding agent plus gold bonding agent application, and (6) Procera porcelain application. Statistically, all of them exhibited no significant differences. Hence, we employed two further criteria: (i) the minimum bond strength should exceed the maximum porcelain strength per se, and (ii) the CpTi substrate should not be heated close to the beta-transus temperature. After applying these criteria, it was concluded that mono-layered nitridation and mono-layered application of chrome-doped nitridation on both (with and without) sand-blasted and non-sand-blasted surfaces were the most promising conditions for a successful titanium-porcelain system [241]. It seems that an alloy which has the properties of titanium and is relatively inexpensive would be a very good material for surgical purposes. These requirements could be met, for example, by stainless steel coated with a firmly adhering non-porous titanium film. Głuszek et al. [242] coated 316L (18Cr-8Ni-2Mo with low carbon content) stainless steel with Ti or TiN by ion plating. The galvanic effects for the galvanic couples 316L/Ti, 316L/Ti-coated 316L, 316L/TiN-coated 316L were studied in Ringer’s solution. It was concluded that (i) both Ti and TiN coatings were non-porous, (ii) Ti served as an anode in the couple 316L/Ti, whereas for the other two couples, the coatings were the cathodes; however (iii) the dissolution rate of 316L stainless steel in these couples was smaller than expected owing to a strong polarization of the coatings [242].
For hip prostheses, the coupling between the metallic femoral head and the polymeric acetabular cup is normally used. Biotribiological phenomena contribute principally to the clinical failure of the prosthesis. In the metal-polymer coupling, the problems consist of biotribological wear, creep of the UHMWPE (ultra high molecular weight polyethylene), and fretting corrosion of the metal femoral head. Cyclic stress exceeding the fatigue resistance of UHMWPE produces surface microcracks and particulates that can migrate into the tissue of the host implant. This fretting-wear debris causes local irritation, proliferation of fibrous tissue, and necrosis of bone. Minimizing the wear is critically important for maintaining the long life of the femoral prosthesis. On the other hand, titanium alloys are susceptible to fretting corrosion; this susceptibility can be reduced via surface treatments. UHMWPE was gamma-ray sterilized. This sterilization technique results in a cross-linking of the polymer, which enhances its wear resistance. Tribio-logical behavior of N2-implanted and nonimplanted titanium alloys coupled with UHMWPE were studied using pin-on-flat tests, according to ASTM F 732-82 in bovine serum. The results show that while the non-implanted titanium alloy and the titanium with N2 on UHMWPE resulted in high final wear values, titanium implanted with O2 generates a wear value less than that obtained for polyethylene against 316L stainless steel. Ti-6Al-4V implanted with chromium exhibited the lowest wear. Hardness values of the implanted material corresponded to the wear rates, which assist in determining optimal elements for implantation. Implantation of certain elements may increase the surface activity, resulting in more adherent oxide layers that also increase wettability [100].

3.4.6.  Ti Coating
Lee et al. [243] conducted the in vivo study to evaluate the behavior and mechanical stability in implants of three surface designs, which were smooth surface Ti, rough Ti surface by plasma spray coating, and alkali and heat-treated. The implants were inserted transversely in a dog thighbone and evaluated at 4 weeks of healing. At four weeks of healing after implantation in bone, it was found that (i) the healing tissue was more extensively integrated with an alkali and heat-treated Ti implant than with the implants of smooth surface and/or rough titanium surfaces, (ii) the bone bonding strength (pull-out force) between living bone and implant was observed by a universal testing machine, (iii) the pull-out forces of the smooth surface Ti, plasma spray coated Ti, and alkali and heat-treated Ti implants were 235, 710, and 823 N, respectively, and (iv) histological and mechanical data demonstrated that appropriate surface design selection can improve early bone growth and induce an acceleration of the healing response, thereby improving the potential for implant osseointegration. In order to improve the biocompatibility of functional titanium-based alloys, Sonoda et al. [244] investigated pure titanium coatings on Ti-6Al-4V alloy by sputtering. More high quality thin film and higher growth rate were obtained by the sputtering with a DC source than with an RF source. After the cleaning method was established, the effect of sputtering on the thickness of the film was investigated with DC sputtering. It was concluded that the growth rate of sputtered titanium film was proportional to the applied electric power, and the orientation of the film highly depended on the heating temperature of the substrate [244]. Sonoda et al. [245] further applied this technique to the complete denture base of the Ti-6Al-4V alloy fabricated by superplastic forming. The base was attached to the substrate holder and cooled by water or heated at 417 °C. It was reported that the film deposited on the heated base was superior to that on the cooled one in smoothness, glossiness, uniformity, and covering of the fine cracks [245].
Histologically, Ti has been demonstrated to be a highly biocompatible material on account of its good resistance to corrosion, absence of toxic effects on macrophages and fibroblasts, and lack of inflammatory response in peri-implant tissues [245–249]. Ti endosseous dental screws with different surfaces (smooth Ti, Ti plasma-sprayed, alumina oxide sand-blasted and acid-etched, zirconia (ZrO2) sand-blasted and acid-etched) were implanted in femura and tibiae of sheep for 14 days to investigate the biological evolution of the peri-implant tissues and detachment of Ti debris from the implant surfaces in early healing. Implants were not loaded. It was reported that (i) all samples showed new bone trabeculae and vascularized medullary spaces in those areas where gaps between the implants and host bone were visible, (ii) in contrast, no osteogenesis was induced in the areas where the implants were initially positioned in close contact with the host bone, (iii) the threads of some screws appeared to be deformed where the host bone showed fractures, and (iv) Ti granules of 3–60 μm were detectable only in the peri-implant tissues of Ti plasma sprayed implants both immediately after surgery and after 14 days, thus suggesting that this phenomenon may be related to the friction of the Ti plasma spray coating during surgical insertion [250].
The use of porous coated implants for long-term biological fixation has been receiving an enthusiastic response, especially when the patients are younger and more active [251,252]. The application of Ti plasma-sprayed coatings to Ti-6Al-4V orthopedic implants results in a dramatic decrease in high-cycle fatigue performance. It was noted that the better bonding of the plasma sprayed and heat-treated implants results in a lower high-cycle fatigue strength. As with conventional sintered porous coatings, the application of a coating that contains defects serves as the crack initiator of the high cycle fatigue. It was also mentioned that the addition of the post-coating heat treatment to improve coating bonding strength resulted in a further reduction in the high cycle fatigue strength, most likely due to a higher frequency of bonding sites between the coating and substrate, and a more intimate metallurgical bond at those sites [253].
Recently, there was a growing interest in Ti plasma sprayed overcoats as a viable alternative to sintered bead or diffusion-bonded fiber metal surfaces, since the inherent roughness of such coatings is believed to favor the osteointegration of the bone [254,255]. Surface treatment plays an important role in the corrosion resistance of Ti. The cement, in spite of having reduced electrical conductivity in comparison to metal, is an ionic transporter, and therefore capable of participating in the corrosion process. The crevice corrosion at the metal-cement interface was studied by Reclaru et al., who reported that in the case of plasma spray surfaces, a process of diffusion of Ti particles in the electrolyte could accompany the crevice corrosion [256].
Xue et al. [257] modified plasma-sprayed titanium coatings by an alkali treatment. The changes in chemical composition and structure of the coatings were examined by SEM and AES. The results indicated that (i) a net-like microscopic texture feature, which was full of the interconnected fine porosity, appeared on the surface of alkali-modified titanium coatings, (ii) the surface chemical composition was also altered by alkali modification, and (iii) a sodium titanate compound was formed on the surface of the titanium coating and replaced the native passivating oxide layer. The bone bonding ability of titanium coatings were investigated using a canine model. The histological examination and SEM observation demonstrated that more new bone was formed on the surface of alkali-modified implants, and grew more rapidly into the porosity. It was therefore concluded that (i) the alkali-modified implants appose directly to the surrounding bone, (ii) in contrast, a gap was observed at the interface between the as-sprayed implants and bone, (iii) the push-out test showed that alkali-modified implants had higher shear strength than as-sprayed implants after 1 month of implantation, and (iv) an interfacial layer, containing Ti, Ca, and P, was found to form at the interface between bone and the alkali-modified implant [257].
Borsari et al. [258] developed a new implant surface with the purpose of avoiding as much stress shielding as possible, and thus prolong the prosthesis lifespan, and investigated the in vitro effect of this new ultra-high roughness and dense Ti (Ra = 74 μm) in comparison with medium (Ra = 18 μm) and high (Ra = 40 μm) roughness and open porous coatings, which were obtained by vacuum plasma spraying. MG63 osteoblast-like cells were seeded on the tested materials and polystyrene, as control, for three and seven days. It was reported that (i) akaline phosphatase activity had lower values for high roughness surfaces than medium and ultra-high rough surfaces, (ii) procollagen-I synthesis reduced with increasing roughness, and the lowest data was found for the ultra-high rough surface, (iii) all tested materials showed significantly higher Interleukin-6 levels than those of polystyrene at both experimental times, and (iv) the new ultra-high roughness and dense coating provided a good biological response, even though, at least in vitro, it behaved similarly to the coatings already used in orthopedics [258]. The bone response to different titanium plasma-sprayed implants was evaluated in the trabecular femoral condyles of 10 goats by Vercaigne et al. [259]. These implants were provided with three different titanium plasma-sprayed coatings with a Ra of 16.5, 21.4, and 37.9 μm, respectively. An Al2O3 grit-blasted implant with a Ra of 4.7 μm was used as a control. After an implantation period of three months, the implants were evaluated histologically and histomorphometrically. Only one implant was not recovered after the evaluation period. It was reported that (i) most of the implants showed a different degree of fibrous tissue alternating with direct bone contact, (ii) complete fibrous encapsulation of the implant was observed in some of the sections, and no signs of delamination of the plasma-sprayed coating was visible, (iii) no significant differences in bone contact were measured between the different types of implants, (iv) hismorphometrical analysis revealed significantly higher bone mass close to the implants (0–500 μm) for treated implants placed in medial femoral condyle and implants placed in the lateral condyle, and (v) at a distance of 500–1500 μm no difference in bone mass measurements between the different implants was observed [259]. Ong et al. [260] in vivo evaluated the bone interfacial strength and bone contact length at the plasma sprayed HA and Ti plasma sprayed implants. Non-coated Ti implants were used for control. Cylindrical coated or non-coated implants were implanted in the dogs’ mandibles. Loading of the implants was performed at 12 weeks after implantation. At 12 weeks after implantation (prior to loading) and one year after loading, implants were evaluated for interfacial bone-implant strength and bone-implant contact length. It was found that (i) no significant differences in interfacial bone-implant strength for all groups at 12 weeks after implantation and after one year loading in normal bone were found; however (ii) bone contact length for HA implants was significantly higher than the Ti plasma sprayed and Ti implants for both periods tested, and (iii) Ti plasma sprayed implants exhibited similar pull-out strength compared to the HA implants [260].

3.4.7.  Titania Film Coating
The high corrosion resistance and good biocompatibility of Ti and its alloys are due to a thin passive film that consists essentially of TiO2. There is increasing evidence, however, that under certain conditions, extensive Ti release may occur in vivo. An ion-beam-assisted sputtering deposition technique has been used to deposit thick and dense TiO2 films on Ti and stainless steel surfaces. A higher electrical film resistance, lower passive current density, and lower donor density (in order of 1015 cm−3) have been measured for sputter-deposited oxide film on Ti in contrast to the naturally formed passive oxide film on Ti (donor density in the order of 1020 cm−3). It was found that (i) the coated surface exhibited improved corrosion resistance in phosphate buffered saline, and (ii) the improved corrosion protection of the sputter-deposited oxide film can be explained by a low defect concentration and, consequently, by a slow mass transport process across the film [261]. The National Industrial Research Institute of Nagoya has established technology for forming a functional gradient Ti-O oxide film on Ti-6Al-4V by the reactive DC sputtering vapor deposition method. The overall film thickness in the experiment was 3 μm, and the Vickers hardness of the surface was 1,500 (vs. CpTi: 200–300) [262].
The conditions for obtaining titanium dioxide from the substrates titanium tetra-chloride and oxygen, and applying this to a surgical 316L stainless steel by PACVD (Plasma Assisted Chemical Vapor Deposition) were determined. It was established that during the process, Ti dioxide anatase is created. During exposure of the 316L stainless steel with the Ti dioxide coating, in Ringer’s solution, it was found that (i) protective properties of this coating improved, (ii) Ti dioxide covering increased the resistivity of 316L stainless steel to pitting corrosion and general corrosion, and (iii) any damage or partial removal of the coating did not cause an increased galvanic corrosion of the substrate [263].
Wollastonite (CaSiO3) ceramic was studied as a medical material for artificial bones and dental roots because of its good bioactivity and biocompatibility [264,265]. Lui et al. [266] prepared wollastonite/TiO2 composite coating using plasma spraying technology onto Ti-Al6-4V substrate. The composite coating revealed a lamellar structure with alternating wollastonite coatings and TiO2 coating. In the case of composite coatings, the primarily crystalline phases of the coatings were wollastonite and rutile, indicating wollastonite and TiO2 did not react during the plasma spraying process. It was found that (i) the mean Vickers microhardness of the coatings increased with an increase in the content of TiO2. Wollastonite/TiO2 composite coatings were soaked in SBF to examine their bioactivity, (ii) a carbonate-containing HA layer was formed on the surface of the wollastonite and composite coatings (wollastonite/TiO2: 7/3) soaked in SBF, while a carbonate-containing HA layer was not formed on the surface of the TiO2 and composite (with wollastonite/TiO2: 3/7) coatings after immersion, and (iii) a silica-rich layer appeared at the interface of the carbonate-containing HA and wollastonite and composite (7/3 with wollastonite/TiO2) coatings. The cytocompatibility study of osteoblasts seeded onto the surface indicated that the addition of wollastonite promoted the proliferation of osteoblasts [266].
Haddow et al. [267] investigated the effects of dip rate, sintering temperatures, and time on the chemical composition of the titania films, their physical structure and thickness, and adherence to a silica substrate. In order to produce films, the sol-gel method was employed. By this method, films which can be mimiced as closely as possible the natural oxide layer that is found on titanium. CpTi (grade IV) isopropoxide was dissolved in isopropanol to form the starting sol. Due to the ease of hydrolysis of this sol, a chelating agent, diethanolamine, was added. A small amount of water was added to the solution of alkoxide to partially polymerize the Ti species. Thin surface films of titania have been deposited onto glass substrates. These films are to be used as substrates in an in vitro model of osseointegration. If titania has been deposited onto glass substrates, the use of low dipping rates prevents cracking in the films, irrespective of the subsequent firing time or temperature. Firing at higher temperature (600 °C) produces predominantly glass films and mimic closely, in chemistry, the natural oxide layer that is formed on Ti implants. Refinement of the dipping set-up, or the use of dilute solutions, may result in the production of thinner films. Thinner films will almost certainly be crack-free after firing, and may result in less organic products being trapped in the film during the firing process. It was mentioned that (i) these data are important to use the titania films to develop an in vitro model to study the phenomenon of osseointegration, and (ii) coatings may be deposited onto a wide range of materials; this would be particularly beneficial, enabling the study of osseointegration by TEM and photon-based spectroscopies [267].
Sato et al. [268] used the sol-gel processing to coat Ti substrates with HA, TiO2, and poly(DL-lactic-glycolic acid). Coatings were evaluated by cytocompatibility testing with osteoblast-like cells (or bone-forming cells). The cytocompatibility of the HA composite coatings was compared to that of a traditional plasma-sprayed HA coating. Results showed that (i) osteoblast-like cell adhesion was promoted on the novel HA sol-gel coating compared to the traditional plasma-sprayed HA coating, and (ii) hydrothermal treatment of the sol-gel coating improved osteoblast-like cell adhesion [268]. Wang et al. [269] treated an amorphous titania gel layer on the CpTi surface after the Ti surface was treated with a H2O2/HCl solution at 80 °C. The thickness of the gel layer increased almost linearly with the period of the treatment. It was found that (i) a subsequent heat treatment above 300 °C transformed gradually the amorphous gels to the anatase crystal structure, and the rutile started to appear after heat treatment at 600 °C, and (ii) similar to the sol-gel derived titania gel coatings, titania gel layers exhibited in vitro apatite deposition ability after the gel layers exceeded a minimum thickness (0.2 μm) and was subsequently heated in a proper temperature range (400–600 °C) [269].
Titania (5–20 mol%) was mixed with pure HA or HA containing Ag2O (10–20 mol%) and was heated at 900 °C for 12 h. The sintered samples were found to contain mainly tricalcium phosphate (β-TCP). Enhanced TCP formation with impurity was observed with TiO2-Ag2O addition. An in vitro solubility study in phosphate buffer at physiological conditions showed the resorbale nature of these materials. It was also mentioned that the gradually functional material structure was formed by spreading a TiO2-Ag2O mixture on the surface of the HA green compact and heating at 900 °C [270].

3.5.  Porosity Controlled Surface and Texturing
As a result of coating titanium surfaces, uncontrolled surface porosity is produced. As reviewed in the previous section, not only coated material’s property but also porosity per se attributes the favorable osseointegartion. There are several researches and methods proposed to control surface porosity.
Void metal composite (VMC) is a porous metal developed to fix a prosthesis to bone by tissue ingrowth. The material is made by techniques which produce structures with controlled porosity, density, and physical properties. The ability to produce a range of structures creates porosities to study the effects of pore size, shape, and density on bone/metal interface strength. Ti-6Al-4V is the metal of choice for VMC. It was selected for its corrosion resistance, good mechanical properties, low density, and good tolerance by body tissue. Structures with spherical pore size ranging from 275 to 650 μm, and have been fabricated with up to 80% theoretical densities. The optimum structure for attachment strength seems to be a pore size of 450 μm and 50% theoretical density [271]. The control porosity can be achieved by blasting with alumina or titania [272].
The topography of titanium implants is of importance with respect to cellular attachment. Chung et al. [273] examined the topographies of three as-received implant systems (Nobelpharma, Swede-Vent, and Screw-Vent), followed by thermal (700 °C for 240 min) and anodic oxidation (70 V in 1M acetic acid solution) of the fixtures. Fixtures were self tapped into freshly sacrificed swine rib bone. It was found that (i) thermal and anodic oxidation, as well as implantation shear stress, had no effect on topography, and (ii) the growth of oxides and implantation shear stress had no affect on topography [272].
Petronis et al. [274] developed a model system for studying cell-surface interactions, based on microfabricated cell culture substrates. Porous surfaces consisting of inter-connecting channels with openings of subcellular dimensions are generated on flat, single crystal, silicon substrates. Channel size (width, depth), distribution, and surface coating can be varied independently and used for systematic investigation of how topographical, chemical, and elastic surface properties influence cell or tissue biological responses. Model porous surfaces have been produced by using two different microfabrication methods. Submicron-sized channels with very high depth-to-width aspect ratios (up to 30) have been made by using electron beam lithography and anisotropic reactive ion etching into single-crystal silicon. Another method uses thick-resist photolithography, which can be used to produce channels wider than 1 μm and with depth-to-width aspect ratios below 20 in an epoxy polymer [274]. Xiaoxiong et al. [275] created pit with controlled pit density and geometry to exhibit porosity controlled surfaces. The pit initiation process on CpTi in bromide solution was investigated by means of surface analysis. The results showed that the titanium surface film formed by anodic polarization in bromide solution was mainly TiO2. Prior to the pit initiation, Br ions were absorbed and accumulated on localized spots of the TiO2 film, forming bromine nuclei containing mostly TiBr4. The bromine nuclei grew into the critical nuclei when the film was in the critical state of breaking down. The depth of the critical nuclei was equal to or less than 3 nm. The concentration of bromine in the critical nuclei was the critical surface concentration. It was the requisite condition for pit initiation that the concentration of bromine in bromine nuclei reached critical surface concentration. It was mentioned that, in the system of titanium/bromide solution, the critical surface concentration was 25–35 wt% and was independent of the temperature and concentration of the solution [275].
The in vitro mineralization of osteoblast-like cells on CpTi with different surface roughness was examined. CpTi discs were polished through 600 grit SiC paper (grooved), polished through 1 μm diamond paste (smooth), or sand-blasted (rough). The discs were cleaned, acid passivated and UV sterilized (254 nm, 300 μW/cm2). Osteoblast-like cells were harvested from rat pups and were cultured. The cultures were grown for 6 or 12 days in media supplemented with 5 mM β-glycerophosphate. It was found that (i) in vitro mineralization responses were independent of surface roughness, and (ii) Alizarin red staining indicated small zones of mineralization on all surfaces, indicating that surface topography is known to affect osteoblast-like cell activities [272].
Ungersboeck et al. [276] investigated five types of limited contact dynamic compression plates of CpTi with different surface treatments and electropolished stainless steel limited contact dynamic compression plates. The surface roughness parameters and chemical surface conditions were determined and checked for probable surface contamination. After an implantation period of 3 months on long sheep bones, the soft tissue adjacent to the plates was evaluated histomorphometrically. The difference in roughness parameters was statistically significant for most surface conditions. It was reported that (i) a correlation was found between the surface roughness of the implants and the thickness of the adjacent soft tissue layer, (ii) the thinnest soft tissue reaction layer with a good adhesion to the implant surface was observed for the titanium anodized plates with coarse surface (20% nitric acid at 60 °C for 30 min), and (iii) smooth implants, in particular the electropolished stainless steel plates, induced statistically significant thicker soft tissue layers [276]. Larsson et al. [277] investigated the bone formation around titanium implants with varied surface properties. Machined and electropolished samples with and without thick anodically formed surface oxide were prepared and inserted in the cortical bone of rabbits (1, 3, and 6 weeks). It was found that (i) light microscopic morphology and morphometry showed that all implants were in contact with bone and had a large proportion of bone within the threads at six weeks, (ii) the electro-polished implants, irrespective of anodic oxidation, were surrounded by less bone than the machined implants after one week, (iii) after six weeks the bone volume, as well as the bone-implant contact, were lower for the merely electropolished implants than for the other three groups, and (iv) a high degree of bone contact and bone formation are achieved with titanium implants which are modified with respect to oxide thickness and surface topography; however, the result with the smooth (electropolished) implants indicates that a reduction of surface roughness, in the initial phase, decreases the rate of bone formation in rabbit cortical bone [277].
Thelen et al. [278] investigated mechanics issues related to potential use of a recently developed porous titanium material for load-bearing implants. This material may have advantages over solid Ti for enhancing the bone-implant interface strength by promoting bone and soft tissue ingrowth, and for reducing the bone-implant modulus mismatch, which can lead to stress shielding. It was mentioned that (i) simple analytic models provide good estimates of the elastic properties of the porous Ti, and (ii) the moduli can be significantly reduced to decrease the mismatch between solid Ti and bone, achieving the mechanical compatibility proposed by Oshida [279]. The finite element simulations show that bone ingrowth will dramatically reduce stress concentrations around the pores [278]. Takemoto et al. [280] prepared porous bioactive titanium implants (porosity of 40%) by a plasma-spray method and subsequent chemical and thermal treatments of immersion in a 5 M aqueous NaOH solution at 60 °C for 24 h, immersion in distilled water at 40 °C for 48 h, and heating to 600 °C for 1 h. It was reported that compression strength and bending strength were 280 MPa (0.2% offset yield strength 85.2 MPa) and 101 MPa, respectively. For in vivo analysis, bioactive and nontreated porous titanium cylinders were implanted into 6 mm diameter holes in rabbit femoral condyles. It was found that (i) the percentage of bone-implant contact (affinity index) of the bioactive implants was significantly larger than for the nontreated implants at all post-implantation times (13.5 versus 10.5, 16.7 versus 12.7, 17.7 versus 10.2, 19.1 versus 7.8 at 2, 4, 8, and 16 weeks, respectively), and (ii) the percentage of bone area ingrowth showed a significant increase with the bioactive implants, whereas with the nontreated implants it appeared to decrease after four weeks (10.7 versus 9.9, 12.3 versus 13.1, 15.2 versus 9.8, 20.6 versus 8.7 at 2, 4, 8, and 16 weeks, respectively), suggesting that porous bioactive titanium has sufficient mechanical properties and biocompatibility for clinical use under load-bearing conditions [280].

3.6.  FoaMed Metal
The foam-shaped materials exhibit a continuously connected ligamented, reticulated open-cell geometry having a duodecahedronal cell shape. The foam-shaped material is in a density ranging from 1 to 20% theoretical, and a cell density of 10 to 50 cells per linear inch, with material density and cell density independently variable. Foamed materials are presently produced in a wide range of plastics, metals, and composites having either solid or tubular ligaments. The metals include aluminum, nickel, copper, silver, zinc, leads, tin, magnesium, and stainless steel [281]. Unfortunately, although the technology for fabricating foamed metal is available, there is no report found on foamed titanium yet.