- Email: [email protected]
Plasma-sprayed hydroxyapatite (HA) coatings with flame-spheroidized feedstock: microstructure and mechanical properties
Biomaterials 21 (2000) 1223}1234
Plasma-sprayed hydroxyapatite (HA) coatings with #ame-spheroidized feedstock: microstructure and mechanical properties S.W.K. Kweh , K.A. Khor *, P. Cheang School of Mechanical & Production Engineering, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore School of Applied Science, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore Received 23 July 1999; accepted 26 November 1999
Abstract Flame-spheroidized feedstock, with excellent known heat transfer and consistent melting capabilities, were used to produce hydroxyapatite (HA) coatings via plasma spraying. The characteristics and inherent mechanical properties of the coatings have been investigated and were found to have direct and impacting relationship with the feedstock characteristics, processing parameters as well as microstructural deformities. Processing parameters such as particle sizes (SHA: 20}45, 45}75 and 75}125 lm) and spray distances (10, 12 and 14 cm) have been systematically varied in the present study. It was found that the increase of particle sizes and spray distances weakened the mechanical properties (microhardness, modulus, fracture toughness and bond strength) and structural stability of the coatings. The presence of inter- and intralamellar thermal microcracks, voids and porosities with limited true contact between lamellae were also found to degrade the mechanical characteristics of the coatings, especially in coatings produced from large-sized HA particles. An e!ort was made to correlate the e!ects of microstructural defects with the resultant mechanical properties and structural integrity of the plasma-sprayed hydroxyapatite (HA) coatings. The e!ects of di!erent heat treatment temperatures (600, 800 and 9003C) on the mechanical properties of the coatings were also studied. It was found that a heat treatment temperature of 8003C does enhance the microhardness and elastic modulus of the coatings signi"cantly (P(0.05) whereas a further increment in heat treatment temperature to 9003C did not show any discernable improvements (P'0.1). The elastic response behaviour and fracture toughness of both the as-sprayed and heat-treated HA coatings using Knoop and Vickers indentations at di!erent loadings have been investigated. Results have shown that the mechanical properties of the coatings have improved signi"cantly despite increasing crack density after heat treatment in air. Coatings produced from the spheroidized feedstock of 20}45 lm (SHA 20}45 lm) sprayed at a stand-o! distance of 10 cm were found to possess the most favourable mechanical properties. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Hydroxyapatite; Knoop hardness; Elastic modulus; Fracture toughness; Bond strength; Microstructure; Plasma spray
1. Introduction Mechanical evaluation of plasma-sprayed bioactive calcium phosphate coatings, especially hydroxyapatite coatings, on bioinert metallic substrates have brought worldwide attention in both orthopaedic and dental applications where the demands of operational stresses of the coatings are stringently required. Such applications, however, require extensive understanding of the microstructure and mechanical properties of the coatings
* Corresponding author. Tel.: #65-799-5526; fax: #65-791-1859. E-mail address: [email protected] (K.A. Khor).
which is currently not available [1]. The determination of the mechanical properties such as Knoop hardness, elastic modulus, fracture toughness and bonding strength is therefore essential and crucial for the assessment of the service behaviour and performance of the bioceramic coatings. Apart from the biocompatibility aspects that render its usefulness, other features like hardness, density, cohesiveness and toughness of the coatings are essential requirements for an appealing satisfactory coating. Past research has noted that plasma spray parameters as well as the feedstock morphology and characteristics have direct in#uences on the mechanical and microstructural characteristics of the coatings [2}5]. Mechanical
0142-9612/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 9 9 ) 0 0 2 7 5 - 6
1224
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
properties varied accordingly as each processing factor varied. Various techniques such as bending, ring, tensile and ultrasonic tests have been developed for evaluating the elastic properties, microhardness and fracture toughness of the coatings. However, such techniques require freestanding samples which are di$cult to obtain [1]. Other methods that employ small samples or coating/substrate system are rather complex which makes acquisition of data rather impossible [6]. In the current study, a simpli"ed Knoop indentation technique that was developed by Marshall et al. [7] was employed to investigate the microhardness and elastic properties of the coatings produced with #ame-spheroidized HA powders (SHA). The advantage of this method is that the values of Young's modulus (E) and microhardness (H) can be obtained simultaneously with the use of small specimen [6]. The fracture toughness of the coatings with di!erent particle sizes and spray distances was also investigated using the conventional Vickers indentation technique, whereby the crack length propagation and dimension were crucially determined. Tanaka [8] has reported that a direct relational e!ect exists, which governs the fractal dimension of cracks to the fracture toughness of ceramic materials. It was found that the indentation fracture toughness increased with increasing fractal dimension of the propagated cracks. The e!ects of di!erent heat treatment temperatures (600, 800 and 9003C) on the mechanical properties of the coatings were investigated. It was found that the properties, which were directly related to their microstructural characteristics, improved signi"cantly (P(0.05) with increasing heat treatment temperatures till 8003C, but deteriorated with increasing spray distance and particle size. Coatings produced from spheroidized HA feedstock of 20}45 lm (SHA 20}45 lm) and at a spray distance of 10 cm were found to possess superior and favourable mechanical properties, especially after being heat treated at 8003C for 1 h in air.
2. Materials and methods 2.1. Feedstock and coating production The raw HA powder was prepared by reacting orthophosphoric acid with calcium hydroxide in a stoichiometric mole concentration [9]. The HA slurry was then spray dried into "ne HA powders using the Ohkawara LT-8 spray dryer. The spray-dried HA powders (SDHA) were then sieved into three size ranges (20}45, 45}75 and 75}125 lm) for subsequent #ame spraying in order to spheroidize the powders. Flame spraying, using the Miller FP 73 #ame torch, was done in distilled water in order to trap the spheroidized powders in water. Oxygen and acetylene were used as combustion
Table 1 Plasma spraying parameters Main arc gas (argon) Auxiliary gas (helium) Net energy Powder feed rate Spraying distances Coating thickness
50 psi (345 kPa) 30 psi (207 kPa) 12 kW 15 g/min (4 rpm) 10 cm; 12 cm; 14 cm 150 lm
gases. A computerized closed-loop-controlled powder feed hopper was used to deliver the powders accurately. The HA coatings were produced via plasma spraying using a robot-controlled 100 kW direct current plasma torch (SG-100 Miller Thermal Inc., USA) equipped with an advanced computerized closed-loop powder feeder system. Argon was used as the main plasma forming gas while helium forms the auxiliary gas. A typical plasma spraying condition can be seen in Table 1. 2.2. Post-treatment of spheroidized HA (SHA) coatings The as-sprayed coatings were heat treated for an hour at temperatures of 600, 800 and 9003C. The heat treatment was performed in air in a high-temperature furnace (Carbolite HTF-18, UK) with no gas #ow using heating and cooling ramps of 53C/min. 2.3. Mechanical characterization Detailed metallographic preparation of the samples was done prior to the indentation tests. The Isomet 2000 Precision Saw (Buehler, USA) was used to wafer-slice the specimens into appropriate sizes for hot mounting, which were subsequently polished using a variable-speed dualfunctional grinding-polishing machine (Ecomet 6, Buehler, USA). Microhardness Knoop and Vickers indentations were performed on the cross-sections of the coatings using the Matsuzawa MXT-70 microhardness indenter. The microhardness as well as the elastic modulus (E) of the coatings were determined by using the following equation: b b aH " ! , a a E
(1)
where b /a denotes the indent diagonal after elastic recovery during indentation; b/a is the ratio of the known Knoop indenter dimensions or geometry (1/7.11); a is a constant, having a value of 0.45, as described in previous literature [6,7,10]. H is the Knoop microhardness value obtained from the indentation, while E is the elastic modulus of the coatings. Test loads of 300 and 500 gf with a holding time of 15 s were used in the investigation in accordance with ASTM E384. A sample size of at least
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
10 was adopted to evaluate the H and E values, and an average value with standard deviation (S.D.) determination was collected. The fracture toughness of both the as-sprayed and heat-treated coatings was evaluated based on the equation shown below, which is related to the median cracks [11,12]: K "0.016(E/H ) (P/c ), (2) '! where K denotes the fracture toughness (MPa m ); '! H (GPa) the Vickers hardness; E (GPa) the elastic modulus obtained from Eq. (1); P the applied indentation load (kgf) and c the radial crack length (mm). The tensile adhesion test was performed in accordance with ASTM C633 using an Instron 4302 (50 kN load cell) universal test machine. The coated stubs were adhesively glued together with coupling stubs using a cold-cured 3M-DP460 epoxy adhesive glue to investigate the mechanical bond strength. The crosshead speed used was 1 mm/min. 2.4. Coating characterization A scanning electron microscope (Cambridge Stereo Scan 360 SEM) was used to study the surface morphology and microstructural characteristics of the plasmasprayed coatings. Phase analysis of the spheroidized powders and coatings were performed on a Philips MPD 1880 X-ray di!ractometer (XRD) using Cu K radiation a at 40 kV and 30 mA. The scanning range (2h) was from 20 to 603 at a scan speed of 13 min\ with a step size of 0.023. Relative crystallinity (C ) of the HA coatings was assessed and determined by comparing the main peak intensity, that is the [2 1 1] peak, of the coatings with the main [2 1 1] peak intensity of the assumed reference HA standard. Commercial HA powder (Kyoritsu, Japan) was used as the reference standard for the calculation as it does not show any line-broadening e!ect, and thus it could serve as the standard of 100% crystallinity [13]. The following equation is as shown: A C (%)" ;100%, (3) A
where C is the relative crystallinity of the measured HA coatings; A the integrated area intensity of the
[2 1 1] peak of HA powders or coatings and A de
notes the integrated area intensity of the [2 1 1] peak of the assumed HA standard. 2.5. Statistical analysis The data for each investigation (hardness, elastic modulus, bonding strength, fracture toughness and relative crystallinity) were presented as mean$SD (standard deviation). The signi"cance of the results was
1225
determined by the paired or unpaired comparison t-test. P values of less than 0.05 (2-tailed) were considered as signi"cant. Marginal signi"cance (MS) is de"ned at 0.05(P(0.1 and non-signi"cance (NS) is de"ned at P'0.1.
3. Results and discussion 3.1. Microstructural ewect on mechanical properties of SHA coatings Past research has shown that the microstructural deformities do possess a domineering signi"cance on the mechanical and structural strength of the coatings. In general, defect features such as voids, porosities and inter- and intralamellar microcracks have proven to be e!ective degrading agents to the mechanical properties of the coatings [14,15]. The "ne interlamellar porosity with limited true area of contact between lamella structures and the characteristic anisotropic e!ect of plasmasprayed coatings have also contributed to a decline in the mechanical strength [16]. McPherson [17] has noted the signi"cant dependence of interlamellar contact on the mechanical properties of the coatings, although the dominating and contributing e!ect of porosity, grain size, crystal structure and phase composition could not be neglected as well. In the current study, the e!ects of particle size and spraying distance in relation to their microstructures were critically reviewed. Figs. 1 and 2 show the crosssectional and planar views of two types of as-sprayed coatings (SHA 20}45 and 75}125 lm) at a particular spray distance of 10 cm. It was found that the SHA 20}45 lm coating has a much denser lamellar structure than SHA 45}75 and 75}125 lm coatings. Typical lamellar striations were seen in the coatings, indicating the characteristic nature of the deposition pattern in plasma spraying where layering and stacking of the coating occurs (Figs. 1b and 2b). However, a closer examination revealed that the larger particle size coatings (SHA 45}75 and 75}125 lm) possess numerous unmelted particles, cavities and macropores, whereas, in the SHA 20}45 lm coating, there is little or no signi"cant indication of the presence of cavities. Unlike the larger particle size coatings (Fig. 2a), a #atter and smoother surface pro"le as a result of the neatly stacked disc-like splats is observed in the SHA 20}45 lm coating (Fig. 1a). The presence of good interlamellar contact and minute amount of unmelted particles with the absence of macropores in the SHA 20}45 lm coating has resulted in the improvement of the mechanical strength and properties of the coating. With the increase in spraying distances, a di!erent phenomenon in morphology and microstructure was observed. Figs. 1 and 3 show the planar and cross-sectional views of as-sprayed SHA 20}45 lm coatings at di!erent
1226
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 1. (a) Planar view of SHA 20}45 lm coatings sprayed at a distance of 10 cm. (b) Cross-sectional view of SHA 20}45 lm coatings sprayed at a distance of 10 cm.
spray distances (10 and 14 cm). It was found that there is an increasing amount of porosities and unmelted particles with non-uniform deposition in coatings sprayed at larger spray distances (12 and 14 cm). In comparison, the amount of unmelted particles with irregular spills of deposition was relatively more in coatings sprayed at 12 and 14 cm than in coatings which were sprayed at 10 cm. Figs. 1a and 3a revealed the di!erences in morphological features. The decline in thermal energy density and the corresponding deposition e$ciency [18] as a result of the increase in spraying distance may have caused the particles to be deposited unmelted. In addition, such a process may also have hindered the spreading of the melted particles along the substrate, thus resulting in a &cluttered' deposition. The large distribution of the larger size HA particles will remain unmelted at a larger spray distance due to its higher heat capacity [5,19]. Morphological examination has found that the deterioration of splat formation has also resulted in the increase of
Fig. 2. (a) Planar view of SHA 75}125 lm coatings sprayed at a distance of 10 cm. (b) Cross-sectional view of SHA 75}125 lm coatings sprayed at a distance of 10 cm.
interconnected pores [20]. Such microstructural and morphological deformities as a result of the increase in spraying distance have led to the deterioration of the coating integrity and mechanical properties, which will be discussed in detail in the subsequent sections. The e!ect of heat treatment (600, 800 and 9003C) has shown marked improvements in the mechanical properties of the coatings. This was re#ected in the signi"cant rise in microhardness and modulus values (P(0.05) at a temperature of 8003C, which subsequently remains relatively constant at 9003C. This can be seen in Fig. 8. Such improvements were noted in the altered microstructural features of the coatings. Lamellar structures were denser and good interlamellar contact was evident despite the presence of minute amount of visible voids and interlamellar pores. However, SEM results have shown an increase in the microcracks density with the increase in heat treatment temperatures as shown in Fig. 4. The high thermal stresses as a result of the rapid cooling and solidi"cation during plasma spraying coupled with the
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
1227
Fig. 4. Cross-sectional view of heat treated SHA 20}45 lm coatings sprayed at a distance of 10 cm.
Fig. 3. (a) Planar view of SHA 20}45 lm coatings sprayed at a distance of 14 cm. (b) Cross-sectional view of SHA 20}45 lm coatings sprayed at a distance of 14 cm.
coe$cient of thermal mismatch between the HA coating (CTE&11.6;10\/K) and the Ti}6Al}4V substrate (CTE&8.5;10\/K) may have caused such phenomena [21]. Experimental results have shown that the presence of microcracks has little or no signi"cant e!ect on the mechanical properties as a whole. The increase in crystallinity and densi"cation as a result of thermal treatment may have masked and superceded the detrimental e!ect caused by thermal microcracks. It was reported that heat treatment could modify the coating microstructure and improve adhesion between coating and substrate, and hence, improve the mechanical properties (fracture toughness and adhesion strength) of the coatings [22,23]. Fig. 4 illustrates a characteristic heat-treated SHA 20}45 lm coating at a spray distance of 10 cm. 3.2. Relative crystallinity of SHA coatings Besides microstructural e!ects, the resultant structural stability is largely dependent on the phase content and crystallinity of the coatings. Fig. 5 shows an X-ray
Fig. 5. XRD Di!ractograms of (a) as-sprayed SHA 20}45 lm coatings sprayed at 10 cm; (b) heat-treated SHA 20}45 lm coatings at 8003C for 1 h.
di!ractogram of the as-sprayed and heat-treated state of the coatings using #ame-spheroidized powders of 20}45 lm (SHA 20}45 lm) sprayed at a stand-o! distance of 10 cm. It can be seen that the process of plasma spraying has resulted in the formation of amorphous calcium phosphate (ACP) coatings with low crystallinity level, which were indicated by the large broad peaks of low intensities (Fig. 5a). The extreme heating and cooling conditions can produce metastable phases such as tricalcium phosphate (TCP), tetracalcium phosphate (TTCP), oxyapatite (OHA) and calcium oxide (CaO) as a result of rapid cooling from a high temperature [20,24]. However, the e!ect of heat treatment at 8003C in a moisture-"lled environment as shown in Fig. 5b may have promoted
1228
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
a recrystallization reaction where the ACP, b-TCP and TTCP were reverted back to crystalline HA coupled with the reduction of the CaO phase, which can be postulated in the following reactions [25,26]: 2Ca (PO ) (TCP)#Ca P O (TTCP)#H O PCa (PO ) (OH) , (4) 2Ca (PO ) (TCP)#H O#CaOPCa (PO ) (OH) . (5) Research has shown that the amount of amorphousphase content of the plasma-sprayed HA coatings is dependent on the spraying and deposition conditions as well as its feedstock characteristics [27,28]. This in turn a!ects the mechanical properties and structural stability of the coatings. However, the ideal and desired crystalline HA phase was found to be less favourable in generating good mechanical properties. The production of coatings with good mechanical features could only be obtained from coatings with predominantly amorphous phase, which has undergone complete melting with well-#attened splats. Such well-melted splats, that form maximum contact between the substrate and adjoining layers, would elicit the adhesive and cohesive characteristics needed for the formation of coatings with good mechanical strength. The presence of crystalline HA, such as in SHA 75}125 lm coatings, would tend to produce coatings which are porous and mechanically weak in nature due to the limited amount of melting. Therefore, the importance of adequate melting is detrimental and necessary for the production of coatings with coherent microstructure, good structural strength and integrity. Coatings with an inherently poor microstructure, such as those found in coatings produced with larger size HA particles (SHA 45}75 and 75}125 lm) and coatings sprayed at larger spray distances (12 and 14 cm), will generally have poor mechanical properties, as such coatings contain numerous unmelted particles, pores and irregular build up of uneven melted splat [29]. Fig. 6a and b showed the e!ects of particle size and spray distance on the relative crystallinity (C ) of the predominant HA phase, calculated from Eq. (3), as a function of its hardness to modulus (H/E) ratio. A sample size of 10 was used in this investigation where the "nal average readings (mean$SD) were recorded. It was found that as the H/E ratio increases, the relative crystallinity level of HA increases as well with respect to the increase in spray distance and particle size. The term H/E may be considered as a brittleness index, where it represents the degree of elastic response in elastic}plastic materials [6]. Therefore, the higher the H/E ratio, the more brittle would the coatings be. In Fig. 6a, it can be seen that the relative crystallinity of the as-sprayed coatings for a particular particle size is very much lower than that of the heat-treated samples.
Fig. 6. (a) E!ect of particle size on the relative crystallinity of SHA 20}45 lm coatings as a function of the hardness to modulus (H/E) ratio. (b) E!ect of spray distance on the relative crystallinity of SHA 20}45 lm coatings as a function of the hardness to modulus (H/E) ratio.
Statistical analysis has shown that the relative crystallinity of the as-sprayed coatings was signi"cantly lower (P(0.01) than the heat-treated coatings. For the assprayed coatings produced with 20}45 lm spheroidized powders (SHA 20}45 lm) sprayed at a distance of 10 cm, a relative crystallinity of about 12.6$2.4% was obtained. In comparison, a value of about 79$2.2% was achieved for the heat-treated SHA 20}45 lm coatings. The increase in crystalline regions as a result of the amorphous phase conversion may have contributed to the increase in relative crystallinity [29]. With the increase in particle size, an increase in relative crystallinity was observed with a corresponding increase in the H/E ratio. The amount of crystalline content was found to be predominantly higher (P(0.05) in the SHA 75}125 lm coatings than in SHA 20}45 lm coatings. This result reinforces the earlier comment that the presence of higher
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
1229
Fig. 7. Hardness and modulus of as-sprayed SHA coatings as a function of the spray distance.
crystallinity content found in the larger size HA coatings (SHA 75}125 lm) would well reduce the resistance to deformation and hence increase the H/E ratio (brittleness index). Therefore, such coatings were found to possess poor mechanical properties (microhardness, fracture toughness and bond strength) due to their poor interlamellar adhesion and cohesion abilities as a result of their poor microstructural characteristics. Unlike the SHA 75}125 lm coatings, the coatings produced with the smallest size range of 20}45 lm powders (SHA 20}45 lm) possess much superior mechanical properties (lower H/E ratio) although the relative crystallinity level is the lowest. It can be inferred that coatings with good mechanical properties are predominantly more amorphous in nature, whereby the mechanical and structural stability may not necessarily be derived from highly crystalline coatings. Tong et al. did related work on the e!ects of crystallinity on the stability of plasma-sprayed coatings, and their "ndings substantiated the above inference that the highly crystalline coatings would not lead to high-stability coatings. Instead, the microstructure of the coatings is more important when stability of the coatings is concerned [30]. An identical behavioural response was obtained with respect to its spray distance, as shown in Fig. 6b. It was found that coatings which were produced from a shorter spraying distance (10 cm) have a signi"cantly lower (P(0.05) crystallinity level and H/E ratio than coatings that were produced from a longer spraying distance (14 cm) of a particular particle size. Thus, it can be deduced that a shorter spraying distance would lead to
the formation of coatings which are highly dense and elastic in nature. An increase in the spray distance would result in a higher relative crystallinity level, but the mechanical properties would be compromised (higher H/E ratio). The increase in spray distance tends to produce a less coherently formed coating due to a dip in deposition e$ciency. As a result, the lack of good interlamellar contact coupled with the increase in the CaO content due to the increase in residence time of the powders in the plasma #ame may have resulted in the reduction of the microhardness, modulus and bond strength of the coatings. Therefore, an optimum spray distance of 10 cm would be ideal for the production of coatings with superior microstructural features and mechanical properties of a lower H/E ratio. 3.3. Microhardness & elastic modulus of SHA coatings The dependence of microhardness and elastic modulus on spray distances and particle sizes was determined on both as-sprayed and heat treated samples. An average of 15 indentations along the cross section of the coatings was taken for 5 sets of each sample. Fig. 7 illustrates the hardness and modulus relationship on as-sprayed samples using an indentation force of 300 gf. Results showed that the Knoop microhardness and elastic modulus values decrease with increasing spray distance and particle size. The highest Knoop microhardness (207.06$8.52 H ) and modulus (31.37$1.31 GPa) were found in the SHA 20}45 lm coating at a spray distance of 10 cm. However, as the spray distance increases, the
1230
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 8. E!ect of heat treatment on the hardness and modulus of SHA coatings.
hardness decreases to 138.46$6.92 H with a modulus of 12.33$1.37 GPa at the same particle size. Statistical analysis revealed that the hardness and modulus of coatings produced from a spray distance of 14 cm were significantly lower (P(0.05) than coatings produced from 10 cm. The increase in spray distance may have resulted in the formation of a less coherent structure with numerous mirostructural defects such as voids or porosities, and this has resulted in the reduction in the mechanical properties. The e!ects of particle size variation on hardness and modulus were evidently noted as well. It was found that coatings produced using smaller size particles (20}45 lm) would elicit signi"cantly higher (P(0.05) hardness and modulus values than coatings having 45}75 or 75}125 lm as feedstock. Ironically, the decline in mechanical properties in SHA 75}125 lm coatings was found to be trivial and marginally signi"cant (0.05(P(0.1) with respect to SHA 45}75 lm coatings, indicating that the use of larger size particles as feedstock to coatings will not deliver signi"cant improvements to the mechanical strength of the coatings in general. The increase in the indentation load (500 gf) was found to reduce the microhardness and modulus values of the coatings. A substantial drop of about 30$1.5% in hardness value and a corresponding drop of about 7$2.1% in modulus were noted in the SHA 20}45 lm coating at spray distance of 10 cm. The variation and dependency of Knoop hardness with load has long been reported and is in general agreement with the shortcoming of the Knoop indent [31,32]. It was found that post heat treatment does improve the hardness and modulus of the coatings. A sample size of 10 was used in each heat treatment temperature. Results
showed that the mechanical properties of the as-sprayed coatings (microhardness and modulus values) improved fairly (0.05(P(0.1) to a heat treatment temperature of 6003C. However, a signi"cant improvement of about 39$1.3% (P(0.05) was observed in the heat treated samples (SHA 45}75 lm coatings) at 8003C. A further increase in heat treatment temperature to 9003C did not show much improvement in the mechanical properties of the coatings. The properties at the former temperature have been found to be statistically indi!erent (P'0.1) to the properties at the latter temperature. Therefore, it can be concluded that a post thermal treatment at 8003C is more than su$cient for the enhancement of the assprayed condition of the coatings. Past research has shown that the e!ect of heat treatment causes densi"cation of the microstructure through the sintering action during heat treatment, which in turn results in the minimization of the porosity level [6]. On the other hand, defects formation in the form of intra- and interlamellar cracks with limited true contact between lamellae [16,17] was prominently noted as a result of a probable residual stress build-up during densi"cation. Such microstructural deformities, however, were found to have no e!ect on the overall enhancement of the mechanical properties in the coatings, as discussed in an earlier section. Fig. 8 shows the e!ect of heat treatment on the hardness and modulus of SHA coatings at a spray distance of 10 cm. The same trend of results holds true for the other spray distances. 3.4. Fracture toughness of SHA coatings Fig. 9 illustrates the fracture toughness of the assprayed SHA coatings as a function of particle size with
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
1231
Fig. 9. Fracture toughness of as-sprayed SHA coatings.
a spray distance of 10 cm. Fracture toughness was determined with "ve sets of specimens on the three types of coatings. An average mean plus standard deviation of about 15 indentations on each coating was recorded. It can be noted that there is a distinctive decreasing trend of fracture toughness as particle size increases. A relatively high fracture toughness (K ) value of 1.14$ '! 0.15 MPa m was obtained for SHA 20}45 lm coating. However, this value dropped to 0.63$0.23 MPa m for the large particle size (SHA 75}125 lm) coatings under an indentation load of 300 gf. The dip in fracture toughness as a result of particle size e!ects was signi"cantly noted (P(0.05) in the three types of coatings, indicating that the nature of fracture toughness was particle size dependent. Similarly, under an indentation load of 500 gf, an identical trend of decreasing fracture toughness was observed as the particle size was increased. However, within a particular size range, an increase in indentation load has resulted in the decline of fracture toughness. The strong in#uence of crack size on the fracture toughness of the coatings has been investigated and reported [33,34]. A large crack propagation that resulted from a high direct force impingement would cause a decrease in the fracture toughness, as depicted in Eq. (2). This would probably explain the above observation. According to Fig. 10, the e!ect of heat treatment (8003C for 1 h) has somewhat improved the coatings' fracture toughness. A substantial increment of about 10% in fracture toughness was noted in the SHA 20}45 lm coating under 300 gf loading. However, a similar trend of decreasing fracture toughness was noted with the increase in particle size. Increasing indentation
loading has the e!ect of increasing the crack propagation density through structural deformation, and hence reducing the fracture toughness. Results showed that the as-sprayed and heat-treated SHA 20}45 lm coating at a spray distance of 10 cm has the highest fracture toughness, which has shown better mechanical resilience than dense sintered hydroxyapatite, having a K value of '! about 1 MPa m [35]. 3.5. Tensile adhesion strength of SHA coatings Bond strength, which is a measure of the adhesion ability of the coating to the substrate, is an important property of thermal-sprayed coatings. The tensile bond strength, as shown in Fig. 11, was studied on as-sprayed coatings (n"10) in accordance with ASTM C633-79 using a crosshead speed of 1 mm/min. The e!ects of particle size and spray distance on the bonding strength were signi"cantly felt and observed. Results have revealed that smaller size range (SHA 20}45 lm) coatings exhibited a much higher adhesion strength (P(0.05) than larger size range coatings (SHA 75}125 lm) at a particular spray distance. The fully melted dense lamellar splats of the former may have contributed to its signi"cantly higher strength than the latter. The highest bond strength of about 23.22$0.68 MPa was obtained from SHA 20}45 lm coatings at a spray distance of 10 cm. However, with the increase of spray distance to 14 cm, the bond strength declined to 18.65$0.65 MPa. This could well mean that the decline in deposition e$ciency as a result of the increase in spray distance may have a!ected the adhesion performance to the substrate, thereby, reducing the bond strength of the coatings.
1232
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 10. Fracture toughness of heat-treated SHA coatings.
Fig. 11. Tensile bond strength as a function of the spray distance and particle size.
Identical adhesion properties were observed with the increase in particle size. At a spray distance of 10 cm, the bond strength of SHA 75}125 lm coatings exhibited a low value of 10.53$0.83 MPa. The large proportion of unmelted particles coupled with the corresponding increase in porosities in the coatings may have resulted in the degradation of the adhesion capability. Delamination of the above-mentioned coating caused by poorly formed lamellar structure and splats was also observed which may have contributed to the drop in adhesion strength.
Research has also indicated the inter-dependency of adhesion strength and crystallinity, which is interrelated with the particle size. It was reported earlier that a high level of crystallinity commonly associated with larger particle size coatings would tend to generate coatings with poorer mechanical properties (that is, coatings with low microhardness and elastic modulus values or high H/E ratio) as a result of its microstructural disabilities. Hence, a lower count of crystallinity level (as in SHA 20}45 lm coatings) would deliver coatings with higher adhesion strength [36]. Brossa et al. have rea$rmed the
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 12. Fracture surface of SHA 20}45 lm coating at a spray distance of 10 cm.
1233
ed from variations in inherent crystallinity level of the coatings, pore morphology, porosities and interlamellar structure. SHA 20}45 lm coating at a spray distance of 10 cm possesses the most favourable mechanical properties. However, the enhanced mechanical properties as a result of heat treatment have introduced into the coatings thermal defects in the form of inter- and intralamellar cracks. Such microstructural defects were found to have little or no e!ect on the overall mechanical strength of the coatings. The fracture toughness of the coatings possesses identical behavioural properties, which showed increasing toughness at increasing heat treatment temperature, but declining toughness with larger particle sizes. Mechanical properties were found to decrease with increasing test load.
Acknowledgements e!ect and dependency of the adhesion strength on the crystallinity of the coatings [37]. The fracture mode of all coatings consists of a mixture of cohesive and adhesive failure, with cohesive failure more commonly found in high strength coatings. The lack of good interlamellar contact and the presence of failure initiation sites within the coating [38] as a result of the increase in the amount of unmelted and porosities content in low strength coatings (SHA 75}125 lm coatings) might have promoted premature fracture at the glue}coating interface region, thus resulting in adhesive failure. A characteristic fractography of the SHA 20}45 lm coating is shown in Fig. 12.
4. Conclusion The mechanical properties of the as-sprayed and heattreated SHA coatings as a function of the spray distances and particle sizes were reviewed. It was found that the microhardness and elastic moduli improved with the e!ects of increasing heat treatment temperature, but deteriorated with increasing spray distances and particle sizes. A thermal treatment of 8003C was found to enhance better mechanical properties. However, no statistical signi"cance (P'0.1) in mechanical strength improvements was observed for higher heat treatment temperature (9003C). A shorter spray distance (10 cm) using smaller size feedstock (SHA 20}45 lm) would produce coatings with much more favourable (P(0.05) mechanical properties such as hardness, elastic modulus and fracture toughness and features than coatings produced at larger spray distance (12 or 14 cm) and particle size (SHA 75}125 lm). It was also noted that coatings with a low degree of crystallinity would tend to inherit better mechanical properties with a low H/E ratio, and vice versa. The changes in hardness and modulus values were associated with microstructural changes that result-
The continued support of the research grant JT ARC 4/96 is gratefully acknowledged.
References [1] Wallace JS, Ilavsky J. Elastic modulus measurements in plasma sprayed deposits. J Therm Spray Tech 1998;7(4):521}6. [2] Malmberg S, Heberlein J. E!ect of plasma spray operating conditions on plasma jet characteristics and coating properties. Proceedings of the Nat Therm Spray Conference, 1993. p. 451}6. [3] Vardelle M, Vardelle A, Fauchais P. Spray parameters and particle behaviour relationships during plasma spraying. J Therm Spray Tech 1993;2(1):79}92. [4] Cheang P, Khor KA. Thermal spraying of hydroxyapatite (HA) coatings: e!ects of powder feedstock. J Mater Proc Technol 1995;48:429}36. [5] Cheang P, Khor KA. Addressing processing problems associated with plasma spraying of hydroxyapatite coatings. Biomaterials 1996;17:537}44. [6] Leigh SH, Lin CK, Berndt CC. Elastic response of thermal spray deposits under indentation tests. J Am Ceram Soc 1997;80(8): 2093}9. [7] Marshall DB, Noma T, Evans AG. A simple method for determining elastic-modulus-to-hardness ratios using Knoop indentation measurements. J Am Ceram Soc 1982;65(10):C175}6. [8] Tanaka M. Fracture toughness and crack morphology in indentation fracture of brittle materials. J Mater Sci 1996;31:749}55. [9] Kweh SWK, Khor KA, Cheang P. The production and characterization of hydroxyapatite (HA) powders. J Mater Process Technol 1999;89}90:373}7. [10] Sung YM. Microindentation analysis of inviscid melt-spun ceramic "bres. J Mater Sci Lett 1997;16:1804}6. [11] Beshish GK, Florey CW, Worzala FJ. Lenling WJ. Fracture toughness of thermal spray ceramic coatings determined by the indentation technique. J Therm Spray Tech 1993;2(1):35}8. [12] Slosarczyk A, Bialoskorski J. Hardness and fracture toughness of dense calcium-phosphate-based materials. J Mater Sci: Mater Med 1998;9:103}8. [13] Wang BC, Chang E, Lee TM, Yang CY. Changes in phases and crystallinity of plasma-sprayed hydroxyapatite coatings under heat treatment: a quantitative study. J Biomed Mater Res 1995; 29:1483}92.
1234
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
[14] Shi KS, Qian ZY, Zhuang MS. Microstructure and properties of sprayed ceramic coating. J Am Ceram Soc 1988;71(11):924}9. [15] Lutz EH. Microstructure and properties of plasma ceramics. J Am Ceram Soc 1994;77(5):1274}80. [16] McPherson R, Cheang P. Elastic anisotropy of APS alumina coatings and its relationship to microstructure. High Perf Ceram Films Coatings 1991:277}90. [17] McPherson R. A review of microstructure and properties of plasma sprayed ceramic coatings. Surface Coat Technol 1989; 39/40:173}81. [18] Malmberg S, Heberlein J. E!ect of plasma spray operating conditions on plasma jet characteristics and coating properties. J Therm Spray Tech 1993;2(4):339}44. [19] Khor KA, Cheang P. Characterization of thermal sprayed hydroxyapatite powders and coatings. J Therm Spray Tech 1994; 3(1):45}50. [20] Liu DM, Chan HM, Wu JD. Plasma-sprayed HA coating: e!ect of di!erent calcium phosphate ceramics. J Mater Sci: Mater Med 1994;5:147}53. [21] Bauer G. Biochemical aspects of osseo-integration. In: Heink G, editor. Osseo integrated implants, vol. 1. Boca Raton, FL: CRC Press, 1990. p. 81}97. [22] Ji H, Marquis PM. E!ect of heat treatment on the microstructure of plasma-sprayed hydroxyapatite coating. Biomaterials 1993; 14(1):64}8. [23] Filaggi MJ, Pilliar RM, Coombs NA. Post-plasma-spraying heat treatment of the HA Coating/Ti}6Al}4V implant system. J Biomed Mater Res 1993;27:191}8. [24] Gross KA, Berndt CC. Thermal processing of hydroxyapatite for coating production. J Biomed Mater Res 1998;39:580}7. [25] Cao Y, Weng J, Chen JY, Feng JM, Yang ZJ, Zhang XD. Water vapour-treated hydroxyapatite coatings after plasma spraying and their characteristics. Biomaterials 1996;17(4):419}24. [26] McPherson R, Gane N. Structural characterization of plasmasprayed coatings. J Mater Sci: Mater Med 1995;6:327}34.
[27] Gross KA, Berndt CC, Herman H. Amorphous phase formation in plasma-sprayed hydroxyapatite coatings. J Biomed Mater Res 1998;39:407}14. [28] Chen J, Wolke JGC, Groot K. Microstructure and crystallinity in hydroxyapatite coatings. Biomaterials 1994;15(5):396}9. [29] Khor KA, Cheang P, Wang Y. The thermal spray processing of HA powders and coatings. JOM 1997;49(2):51}7. [30] Tong W, Chen J, Li X, Cao Y, Yang Z, Feng J, Zhang X. E!ect of particle size on molten states of starting powder and degradation of the relevant plasma-sprayed hydroxyapatite coatings. Biomaterials 1996;17(15):1507}13. [31] Voort GFV, Lucas GM. Microindentation hardness testing. Adv Mater and Proc 1998;9:21}5. [32] Quinn GD. Hardness testing of ceramics. Adv Mater Proc 1998;154(2):23}7. [33] Lawn BR, Evans AG, Marshall DB. Elastic/Plastic indentation damage in ceramics: the median/radial crack system. J Am Ceram Soc 1980;63(9}10):574}81. [34] Ponton CB, Rawlings RD. Dependence of the vickers indentation fracture toughness on the surface crack length. J Br Ceram Trans 1989;88:83}90. [35] Aoki H. Medical applications of hydroxyapatite. Tokyo: Ishiyaku EuroAmerica Inc., 1994. p. 290}1. [36] Yang CY, Liu RM, Wang BC, Lee TM, Chang E, Hang YS, Chen PQ. In vitro and in vivo mechanical evaluations of plasmasprayed hydroxyapatite coatings on titanium implants: the e!ect of coating characteristics. J Biomed Mater Res 1997; 37:335}45. [37] Brossa F, Cigada A, Chiesa R, Paracchini L, Consonni C. Adhesion properties of plasma sprayed hydroxyapatite coatings for orthopaedic prostheses. Biomed Mater Engng 1993; 3(3):127}36. [38] Berndt CC, Gross KA. Characteristics of hydroxyapatite biocoatings. Proceedings of the International Therm. Spray Conference & Expo, Orlando, USA, 1992. p. 465}70.
Plasma-sprayed hydroxyapatite (HA) coatings with #ame-spheroidized feedstock: microstructure and mechanical properties S.W.K. Kweh , K.A. Khor *, P. Cheang School of Mechanical & Production Engineering, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore School of Applied Science, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore Received 23 July 1999; accepted 26 November 1999
Abstract Flame-spheroidized feedstock, with excellent known heat transfer and consistent melting capabilities, were used to produce hydroxyapatite (HA) coatings via plasma spraying. The characteristics and inherent mechanical properties of the coatings have been investigated and were found to have direct and impacting relationship with the feedstock characteristics, processing parameters as well as microstructural deformities. Processing parameters such as particle sizes (SHA: 20}45, 45}75 and 75}125 lm) and spray distances (10, 12 and 14 cm) have been systematically varied in the present study. It was found that the increase of particle sizes and spray distances weakened the mechanical properties (microhardness, modulus, fracture toughness and bond strength) and structural stability of the coatings. The presence of inter- and intralamellar thermal microcracks, voids and porosities with limited true contact between lamellae were also found to degrade the mechanical characteristics of the coatings, especially in coatings produced from large-sized HA particles. An e!ort was made to correlate the e!ects of microstructural defects with the resultant mechanical properties and structural integrity of the plasma-sprayed hydroxyapatite (HA) coatings. The e!ects of di!erent heat treatment temperatures (600, 800 and 9003C) on the mechanical properties of the coatings were also studied. It was found that a heat treatment temperature of 8003C does enhance the microhardness and elastic modulus of the coatings signi"cantly (P(0.05) whereas a further increment in heat treatment temperature to 9003C did not show any discernable improvements (P'0.1). The elastic response behaviour and fracture toughness of both the as-sprayed and heat-treated HA coatings using Knoop and Vickers indentations at di!erent loadings have been investigated. Results have shown that the mechanical properties of the coatings have improved signi"cantly despite increasing crack density after heat treatment in air. Coatings produced from the spheroidized feedstock of 20}45 lm (SHA 20}45 lm) sprayed at a stand-o! distance of 10 cm were found to possess the most favourable mechanical properties. 2000 Elsevier Science Ltd. All rights reserved. Keywords: Hydroxyapatite; Knoop hardness; Elastic modulus; Fracture toughness; Bond strength; Microstructure; Plasma spray
1. Introduction Mechanical evaluation of plasma-sprayed bioactive calcium phosphate coatings, especially hydroxyapatite coatings, on bioinert metallic substrates have brought worldwide attention in both orthopaedic and dental applications where the demands of operational stresses of the coatings are stringently required. Such applications, however, require extensive understanding of the microstructure and mechanical properties of the coatings
* Corresponding author. Tel.: #65-799-5526; fax: #65-791-1859. E-mail address: [email protected] (K.A. Khor).
which is currently not available [1]. The determination of the mechanical properties such as Knoop hardness, elastic modulus, fracture toughness and bonding strength is therefore essential and crucial for the assessment of the service behaviour and performance of the bioceramic coatings. Apart from the biocompatibility aspects that render its usefulness, other features like hardness, density, cohesiveness and toughness of the coatings are essential requirements for an appealing satisfactory coating. Past research has noted that plasma spray parameters as well as the feedstock morphology and characteristics have direct in#uences on the mechanical and microstructural characteristics of the coatings [2}5]. Mechanical
0142-9612/00/$ - see front matter 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 1 4 2 - 9 6 1 2 ( 9 9 ) 0 0 2 7 5 - 6
1224
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
properties varied accordingly as each processing factor varied. Various techniques such as bending, ring, tensile and ultrasonic tests have been developed for evaluating the elastic properties, microhardness and fracture toughness of the coatings. However, such techniques require freestanding samples which are di$cult to obtain [1]. Other methods that employ small samples or coating/substrate system are rather complex which makes acquisition of data rather impossible [6]. In the current study, a simpli"ed Knoop indentation technique that was developed by Marshall et al. [7] was employed to investigate the microhardness and elastic properties of the coatings produced with #ame-spheroidized HA powders (SHA). The advantage of this method is that the values of Young's modulus (E) and microhardness (H) can be obtained simultaneously with the use of small specimen [6]. The fracture toughness of the coatings with di!erent particle sizes and spray distances was also investigated using the conventional Vickers indentation technique, whereby the crack length propagation and dimension were crucially determined. Tanaka [8] has reported that a direct relational e!ect exists, which governs the fractal dimension of cracks to the fracture toughness of ceramic materials. It was found that the indentation fracture toughness increased with increasing fractal dimension of the propagated cracks. The e!ects of di!erent heat treatment temperatures (600, 800 and 9003C) on the mechanical properties of the coatings were investigated. It was found that the properties, which were directly related to their microstructural characteristics, improved signi"cantly (P(0.05) with increasing heat treatment temperatures till 8003C, but deteriorated with increasing spray distance and particle size. Coatings produced from spheroidized HA feedstock of 20}45 lm (SHA 20}45 lm) and at a spray distance of 10 cm were found to possess superior and favourable mechanical properties, especially after being heat treated at 8003C for 1 h in air.
2. Materials and methods 2.1. Feedstock and coating production The raw HA powder was prepared by reacting orthophosphoric acid with calcium hydroxide in a stoichiometric mole concentration [9]. The HA slurry was then spray dried into "ne HA powders using the Ohkawara LT-8 spray dryer. The spray-dried HA powders (SDHA) were then sieved into three size ranges (20}45, 45}75 and 75}125 lm) for subsequent #ame spraying in order to spheroidize the powders. Flame spraying, using the Miller FP 73 #ame torch, was done in distilled water in order to trap the spheroidized powders in water. Oxygen and acetylene were used as combustion
Table 1 Plasma spraying parameters Main arc gas (argon) Auxiliary gas (helium) Net energy Powder feed rate Spraying distances Coating thickness
50 psi (345 kPa) 30 psi (207 kPa) 12 kW 15 g/min (4 rpm) 10 cm; 12 cm; 14 cm 150 lm
gases. A computerized closed-loop-controlled powder feed hopper was used to deliver the powders accurately. The HA coatings were produced via plasma spraying using a robot-controlled 100 kW direct current plasma torch (SG-100 Miller Thermal Inc., USA) equipped with an advanced computerized closed-loop powder feeder system. Argon was used as the main plasma forming gas while helium forms the auxiliary gas. A typical plasma spraying condition can be seen in Table 1. 2.2. Post-treatment of spheroidized HA (SHA) coatings The as-sprayed coatings were heat treated for an hour at temperatures of 600, 800 and 9003C. The heat treatment was performed in air in a high-temperature furnace (Carbolite HTF-18, UK) with no gas #ow using heating and cooling ramps of 53C/min. 2.3. Mechanical characterization Detailed metallographic preparation of the samples was done prior to the indentation tests. The Isomet 2000 Precision Saw (Buehler, USA) was used to wafer-slice the specimens into appropriate sizes for hot mounting, which were subsequently polished using a variable-speed dualfunctional grinding-polishing machine (Ecomet 6, Buehler, USA). Microhardness Knoop and Vickers indentations were performed on the cross-sections of the coatings using the Matsuzawa MXT-70 microhardness indenter. The microhardness as well as the elastic modulus (E) of the coatings were determined by using the following equation: b b aH " ! , a a E
(1)
where b /a denotes the indent diagonal after elastic recovery during indentation; b/a is the ratio of the known Knoop indenter dimensions or geometry (1/7.11); a is a constant, having a value of 0.45, as described in previous literature [6,7,10]. H is the Knoop microhardness value obtained from the indentation, while E is the elastic modulus of the coatings. Test loads of 300 and 500 gf with a holding time of 15 s were used in the investigation in accordance with ASTM E384. A sample size of at least
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
10 was adopted to evaluate the H and E values, and an average value with standard deviation (S.D.) determination was collected. The fracture toughness of both the as-sprayed and heat-treated coatings was evaluated based on the equation shown below, which is related to the median cracks [11,12]: K "0.016(E/H ) (P/c ), (2) '! where K denotes the fracture toughness (MPa m ); '! H (GPa) the Vickers hardness; E (GPa) the elastic modulus obtained from Eq. (1); P the applied indentation load (kgf) and c the radial crack length (mm). The tensile adhesion test was performed in accordance with ASTM C633 using an Instron 4302 (50 kN load cell) universal test machine. The coated stubs were adhesively glued together with coupling stubs using a cold-cured 3M-DP460 epoxy adhesive glue to investigate the mechanical bond strength. The crosshead speed used was 1 mm/min. 2.4. Coating characterization A scanning electron microscope (Cambridge Stereo Scan 360 SEM) was used to study the surface morphology and microstructural characteristics of the plasmasprayed coatings. Phase analysis of the spheroidized powders and coatings were performed on a Philips MPD 1880 X-ray di!ractometer (XRD) using Cu K radiation a at 40 kV and 30 mA. The scanning range (2h) was from 20 to 603 at a scan speed of 13 min\ with a step size of 0.023. Relative crystallinity (C ) of the HA coatings was assessed and determined by comparing the main peak intensity, that is the [2 1 1] peak, of the coatings with the main [2 1 1] peak intensity of the assumed reference HA standard. Commercial HA powder (Kyoritsu, Japan) was used as the reference standard for the calculation as it does not show any line-broadening e!ect, and thus it could serve as the standard of 100% crystallinity [13]. The following equation is as shown: A C (%)" ;100%, (3) A
where C is the relative crystallinity of the measured HA coatings; A the integrated area intensity of the
[2 1 1] peak of HA powders or coatings and A de
notes the integrated area intensity of the [2 1 1] peak of the assumed HA standard. 2.5. Statistical analysis The data for each investigation (hardness, elastic modulus, bonding strength, fracture toughness and relative crystallinity) were presented as mean$SD (standard deviation). The signi"cance of the results was
1225
determined by the paired or unpaired comparison t-test. P values of less than 0.05 (2-tailed) were considered as signi"cant. Marginal signi"cance (MS) is de"ned at 0.05(P(0.1 and non-signi"cance (NS) is de"ned at P'0.1.
3. Results and discussion 3.1. Microstructural ewect on mechanical properties of SHA coatings Past research has shown that the microstructural deformities do possess a domineering signi"cance on the mechanical and structural strength of the coatings. In general, defect features such as voids, porosities and inter- and intralamellar microcracks have proven to be e!ective degrading agents to the mechanical properties of the coatings [14,15]. The "ne interlamellar porosity with limited true area of contact between lamella structures and the characteristic anisotropic e!ect of plasmasprayed coatings have also contributed to a decline in the mechanical strength [16]. McPherson [17] has noted the signi"cant dependence of interlamellar contact on the mechanical properties of the coatings, although the dominating and contributing e!ect of porosity, grain size, crystal structure and phase composition could not be neglected as well. In the current study, the e!ects of particle size and spraying distance in relation to their microstructures were critically reviewed. Figs. 1 and 2 show the crosssectional and planar views of two types of as-sprayed coatings (SHA 20}45 and 75}125 lm) at a particular spray distance of 10 cm. It was found that the SHA 20}45 lm coating has a much denser lamellar structure than SHA 45}75 and 75}125 lm coatings. Typical lamellar striations were seen in the coatings, indicating the characteristic nature of the deposition pattern in plasma spraying where layering and stacking of the coating occurs (Figs. 1b and 2b). However, a closer examination revealed that the larger particle size coatings (SHA 45}75 and 75}125 lm) possess numerous unmelted particles, cavities and macropores, whereas, in the SHA 20}45 lm coating, there is little or no signi"cant indication of the presence of cavities. Unlike the larger particle size coatings (Fig. 2a), a #atter and smoother surface pro"le as a result of the neatly stacked disc-like splats is observed in the SHA 20}45 lm coating (Fig. 1a). The presence of good interlamellar contact and minute amount of unmelted particles with the absence of macropores in the SHA 20}45 lm coating has resulted in the improvement of the mechanical strength and properties of the coating. With the increase in spraying distances, a di!erent phenomenon in morphology and microstructure was observed. Figs. 1 and 3 show the planar and cross-sectional views of as-sprayed SHA 20}45 lm coatings at di!erent
1226
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 1. (a) Planar view of SHA 20}45 lm coatings sprayed at a distance of 10 cm. (b) Cross-sectional view of SHA 20}45 lm coatings sprayed at a distance of 10 cm.
spray distances (10 and 14 cm). It was found that there is an increasing amount of porosities and unmelted particles with non-uniform deposition in coatings sprayed at larger spray distances (12 and 14 cm). In comparison, the amount of unmelted particles with irregular spills of deposition was relatively more in coatings sprayed at 12 and 14 cm than in coatings which were sprayed at 10 cm. Figs. 1a and 3a revealed the di!erences in morphological features. The decline in thermal energy density and the corresponding deposition e$ciency [18] as a result of the increase in spraying distance may have caused the particles to be deposited unmelted. In addition, such a process may also have hindered the spreading of the melted particles along the substrate, thus resulting in a &cluttered' deposition. The large distribution of the larger size HA particles will remain unmelted at a larger spray distance due to its higher heat capacity [5,19]. Morphological examination has found that the deterioration of splat formation has also resulted in the increase of
Fig. 2. (a) Planar view of SHA 75}125 lm coatings sprayed at a distance of 10 cm. (b) Cross-sectional view of SHA 75}125 lm coatings sprayed at a distance of 10 cm.
interconnected pores [20]. Such microstructural and morphological deformities as a result of the increase in spraying distance have led to the deterioration of the coating integrity and mechanical properties, which will be discussed in detail in the subsequent sections. The e!ect of heat treatment (600, 800 and 9003C) has shown marked improvements in the mechanical properties of the coatings. This was re#ected in the signi"cant rise in microhardness and modulus values (P(0.05) at a temperature of 8003C, which subsequently remains relatively constant at 9003C. This can be seen in Fig. 8. Such improvements were noted in the altered microstructural features of the coatings. Lamellar structures were denser and good interlamellar contact was evident despite the presence of minute amount of visible voids and interlamellar pores. However, SEM results have shown an increase in the microcracks density with the increase in heat treatment temperatures as shown in Fig. 4. The high thermal stresses as a result of the rapid cooling and solidi"cation during plasma spraying coupled with the
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
1227
Fig. 4. Cross-sectional view of heat treated SHA 20}45 lm coatings sprayed at a distance of 10 cm.
Fig. 3. (a) Planar view of SHA 20}45 lm coatings sprayed at a distance of 14 cm. (b) Cross-sectional view of SHA 20}45 lm coatings sprayed at a distance of 14 cm.
coe$cient of thermal mismatch between the HA coating (CTE&11.6;10\/K) and the Ti}6Al}4V substrate (CTE&8.5;10\/K) may have caused such phenomena [21]. Experimental results have shown that the presence of microcracks has little or no signi"cant e!ect on the mechanical properties as a whole. The increase in crystallinity and densi"cation as a result of thermal treatment may have masked and superceded the detrimental e!ect caused by thermal microcracks. It was reported that heat treatment could modify the coating microstructure and improve adhesion between coating and substrate, and hence, improve the mechanical properties (fracture toughness and adhesion strength) of the coatings [22,23]. Fig. 4 illustrates a characteristic heat-treated SHA 20}45 lm coating at a spray distance of 10 cm. 3.2. Relative crystallinity of SHA coatings Besides microstructural e!ects, the resultant structural stability is largely dependent on the phase content and crystallinity of the coatings. Fig. 5 shows an X-ray
Fig. 5. XRD Di!ractograms of (a) as-sprayed SHA 20}45 lm coatings sprayed at 10 cm; (b) heat-treated SHA 20}45 lm coatings at 8003C for 1 h.
di!ractogram of the as-sprayed and heat-treated state of the coatings using #ame-spheroidized powders of 20}45 lm (SHA 20}45 lm) sprayed at a stand-o! distance of 10 cm. It can be seen that the process of plasma spraying has resulted in the formation of amorphous calcium phosphate (ACP) coatings with low crystallinity level, which were indicated by the large broad peaks of low intensities (Fig. 5a). The extreme heating and cooling conditions can produce metastable phases such as tricalcium phosphate (TCP), tetracalcium phosphate (TTCP), oxyapatite (OHA) and calcium oxide (CaO) as a result of rapid cooling from a high temperature [20,24]. However, the e!ect of heat treatment at 8003C in a moisture-"lled environment as shown in Fig. 5b may have promoted
1228
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
a recrystallization reaction where the ACP, b-TCP and TTCP were reverted back to crystalline HA coupled with the reduction of the CaO phase, which can be postulated in the following reactions [25,26]: 2Ca (PO ) (TCP)#Ca P O (TTCP)#H O PCa (PO ) (OH) , (4) 2Ca (PO ) (TCP)#H O#CaOPCa (PO ) (OH) . (5) Research has shown that the amount of amorphousphase content of the plasma-sprayed HA coatings is dependent on the spraying and deposition conditions as well as its feedstock characteristics [27,28]. This in turn a!ects the mechanical properties and structural stability of the coatings. However, the ideal and desired crystalline HA phase was found to be less favourable in generating good mechanical properties. The production of coatings with good mechanical features could only be obtained from coatings with predominantly amorphous phase, which has undergone complete melting with well-#attened splats. Such well-melted splats, that form maximum contact between the substrate and adjoining layers, would elicit the adhesive and cohesive characteristics needed for the formation of coatings with good mechanical strength. The presence of crystalline HA, such as in SHA 75}125 lm coatings, would tend to produce coatings which are porous and mechanically weak in nature due to the limited amount of melting. Therefore, the importance of adequate melting is detrimental and necessary for the production of coatings with coherent microstructure, good structural strength and integrity. Coatings with an inherently poor microstructure, such as those found in coatings produced with larger size HA particles (SHA 45}75 and 75}125 lm) and coatings sprayed at larger spray distances (12 and 14 cm), will generally have poor mechanical properties, as such coatings contain numerous unmelted particles, pores and irregular build up of uneven melted splat [29]. Fig. 6a and b showed the e!ects of particle size and spray distance on the relative crystallinity (C ) of the predominant HA phase, calculated from Eq. (3), as a function of its hardness to modulus (H/E) ratio. A sample size of 10 was used in this investigation where the "nal average readings (mean$SD) were recorded. It was found that as the H/E ratio increases, the relative crystallinity level of HA increases as well with respect to the increase in spray distance and particle size. The term H/E may be considered as a brittleness index, where it represents the degree of elastic response in elastic}plastic materials [6]. Therefore, the higher the H/E ratio, the more brittle would the coatings be. In Fig. 6a, it can be seen that the relative crystallinity of the as-sprayed coatings for a particular particle size is very much lower than that of the heat-treated samples.
Fig. 6. (a) E!ect of particle size on the relative crystallinity of SHA 20}45 lm coatings as a function of the hardness to modulus (H/E) ratio. (b) E!ect of spray distance on the relative crystallinity of SHA 20}45 lm coatings as a function of the hardness to modulus (H/E) ratio.
Statistical analysis has shown that the relative crystallinity of the as-sprayed coatings was signi"cantly lower (P(0.01) than the heat-treated coatings. For the assprayed coatings produced with 20}45 lm spheroidized powders (SHA 20}45 lm) sprayed at a distance of 10 cm, a relative crystallinity of about 12.6$2.4% was obtained. In comparison, a value of about 79$2.2% was achieved for the heat-treated SHA 20}45 lm coatings. The increase in crystalline regions as a result of the amorphous phase conversion may have contributed to the increase in relative crystallinity [29]. With the increase in particle size, an increase in relative crystallinity was observed with a corresponding increase in the H/E ratio. The amount of crystalline content was found to be predominantly higher (P(0.05) in the SHA 75}125 lm coatings than in SHA 20}45 lm coatings. This result reinforces the earlier comment that the presence of higher
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
1229
Fig. 7. Hardness and modulus of as-sprayed SHA coatings as a function of the spray distance.
crystallinity content found in the larger size HA coatings (SHA 75}125 lm) would well reduce the resistance to deformation and hence increase the H/E ratio (brittleness index). Therefore, such coatings were found to possess poor mechanical properties (microhardness, fracture toughness and bond strength) due to their poor interlamellar adhesion and cohesion abilities as a result of their poor microstructural characteristics. Unlike the SHA 75}125 lm coatings, the coatings produced with the smallest size range of 20}45 lm powders (SHA 20}45 lm) possess much superior mechanical properties (lower H/E ratio) although the relative crystallinity level is the lowest. It can be inferred that coatings with good mechanical properties are predominantly more amorphous in nature, whereby the mechanical and structural stability may not necessarily be derived from highly crystalline coatings. Tong et al. did related work on the e!ects of crystallinity on the stability of plasma-sprayed coatings, and their "ndings substantiated the above inference that the highly crystalline coatings would not lead to high-stability coatings. Instead, the microstructure of the coatings is more important when stability of the coatings is concerned [30]. An identical behavioural response was obtained with respect to its spray distance, as shown in Fig. 6b. It was found that coatings which were produced from a shorter spraying distance (10 cm) have a signi"cantly lower (P(0.05) crystallinity level and H/E ratio than coatings that were produced from a longer spraying distance (14 cm) of a particular particle size. Thus, it can be deduced that a shorter spraying distance would lead to
the formation of coatings which are highly dense and elastic in nature. An increase in the spray distance would result in a higher relative crystallinity level, but the mechanical properties would be compromised (higher H/E ratio). The increase in spray distance tends to produce a less coherently formed coating due to a dip in deposition e$ciency. As a result, the lack of good interlamellar contact coupled with the increase in the CaO content due to the increase in residence time of the powders in the plasma #ame may have resulted in the reduction of the microhardness, modulus and bond strength of the coatings. Therefore, an optimum spray distance of 10 cm would be ideal for the production of coatings with superior microstructural features and mechanical properties of a lower H/E ratio. 3.3. Microhardness & elastic modulus of SHA coatings The dependence of microhardness and elastic modulus on spray distances and particle sizes was determined on both as-sprayed and heat treated samples. An average of 15 indentations along the cross section of the coatings was taken for 5 sets of each sample. Fig. 7 illustrates the hardness and modulus relationship on as-sprayed samples using an indentation force of 300 gf. Results showed that the Knoop microhardness and elastic modulus values decrease with increasing spray distance and particle size. The highest Knoop microhardness (207.06$8.52 H ) and modulus (31.37$1.31 GPa) were found in the SHA 20}45 lm coating at a spray distance of 10 cm. However, as the spray distance increases, the
1230
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 8. E!ect of heat treatment on the hardness and modulus of SHA coatings.
hardness decreases to 138.46$6.92 H with a modulus of 12.33$1.37 GPa at the same particle size. Statistical analysis revealed that the hardness and modulus of coatings produced from a spray distance of 14 cm were significantly lower (P(0.05) than coatings produced from 10 cm. The increase in spray distance may have resulted in the formation of a less coherent structure with numerous mirostructural defects such as voids or porosities, and this has resulted in the reduction in the mechanical properties. The e!ects of particle size variation on hardness and modulus were evidently noted as well. It was found that coatings produced using smaller size particles (20}45 lm) would elicit signi"cantly higher (P(0.05) hardness and modulus values than coatings having 45}75 or 75}125 lm as feedstock. Ironically, the decline in mechanical properties in SHA 75}125 lm coatings was found to be trivial and marginally signi"cant (0.05(P(0.1) with respect to SHA 45}75 lm coatings, indicating that the use of larger size particles as feedstock to coatings will not deliver signi"cant improvements to the mechanical strength of the coatings in general. The increase in the indentation load (500 gf) was found to reduce the microhardness and modulus values of the coatings. A substantial drop of about 30$1.5% in hardness value and a corresponding drop of about 7$2.1% in modulus were noted in the SHA 20}45 lm coating at spray distance of 10 cm. The variation and dependency of Knoop hardness with load has long been reported and is in general agreement with the shortcoming of the Knoop indent [31,32]. It was found that post heat treatment does improve the hardness and modulus of the coatings. A sample size of 10 was used in each heat treatment temperature. Results
showed that the mechanical properties of the as-sprayed coatings (microhardness and modulus values) improved fairly (0.05(P(0.1) to a heat treatment temperature of 6003C. However, a signi"cant improvement of about 39$1.3% (P(0.05) was observed in the heat treated samples (SHA 45}75 lm coatings) at 8003C. A further increase in heat treatment temperature to 9003C did not show much improvement in the mechanical properties of the coatings. The properties at the former temperature have been found to be statistically indi!erent (P'0.1) to the properties at the latter temperature. Therefore, it can be concluded that a post thermal treatment at 8003C is more than su$cient for the enhancement of the assprayed condition of the coatings. Past research has shown that the e!ect of heat treatment causes densi"cation of the microstructure through the sintering action during heat treatment, which in turn results in the minimization of the porosity level [6]. On the other hand, defects formation in the form of intra- and interlamellar cracks with limited true contact between lamellae [16,17] was prominently noted as a result of a probable residual stress build-up during densi"cation. Such microstructural deformities, however, were found to have no e!ect on the overall enhancement of the mechanical properties in the coatings, as discussed in an earlier section. Fig. 8 shows the e!ect of heat treatment on the hardness and modulus of SHA coatings at a spray distance of 10 cm. The same trend of results holds true for the other spray distances. 3.4. Fracture toughness of SHA coatings Fig. 9 illustrates the fracture toughness of the assprayed SHA coatings as a function of particle size with
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
1231
Fig. 9. Fracture toughness of as-sprayed SHA coatings.
a spray distance of 10 cm. Fracture toughness was determined with "ve sets of specimens on the three types of coatings. An average mean plus standard deviation of about 15 indentations on each coating was recorded. It can be noted that there is a distinctive decreasing trend of fracture toughness as particle size increases. A relatively high fracture toughness (K ) value of 1.14$ '! 0.15 MPa m was obtained for SHA 20}45 lm coating. However, this value dropped to 0.63$0.23 MPa m for the large particle size (SHA 75}125 lm) coatings under an indentation load of 300 gf. The dip in fracture toughness as a result of particle size e!ects was signi"cantly noted (P(0.05) in the three types of coatings, indicating that the nature of fracture toughness was particle size dependent. Similarly, under an indentation load of 500 gf, an identical trend of decreasing fracture toughness was observed as the particle size was increased. However, within a particular size range, an increase in indentation load has resulted in the decline of fracture toughness. The strong in#uence of crack size on the fracture toughness of the coatings has been investigated and reported [33,34]. A large crack propagation that resulted from a high direct force impingement would cause a decrease in the fracture toughness, as depicted in Eq. (2). This would probably explain the above observation. According to Fig. 10, the e!ect of heat treatment (8003C for 1 h) has somewhat improved the coatings' fracture toughness. A substantial increment of about 10% in fracture toughness was noted in the SHA 20}45 lm coating under 300 gf loading. However, a similar trend of decreasing fracture toughness was noted with the increase in particle size. Increasing indentation
loading has the e!ect of increasing the crack propagation density through structural deformation, and hence reducing the fracture toughness. Results showed that the as-sprayed and heat-treated SHA 20}45 lm coating at a spray distance of 10 cm has the highest fracture toughness, which has shown better mechanical resilience than dense sintered hydroxyapatite, having a K value of '! about 1 MPa m [35]. 3.5. Tensile adhesion strength of SHA coatings Bond strength, which is a measure of the adhesion ability of the coating to the substrate, is an important property of thermal-sprayed coatings. The tensile bond strength, as shown in Fig. 11, was studied on as-sprayed coatings (n"10) in accordance with ASTM C633-79 using a crosshead speed of 1 mm/min. The e!ects of particle size and spray distance on the bonding strength were signi"cantly felt and observed. Results have revealed that smaller size range (SHA 20}45 lm) coatings exhibited a much higher adhesion strength (P(0.05) than larger size range coatings (SHA 75}125 lm) at a particular spray distance. The fully melted dense lamellar splats of the former may have contributed to its signi"cantly higher strength than the latter. The highest bond strength of about 23.22$0.68 MPa was obtained from SHA 20}45 lm coatings at a spray distance of 10 cm. However, with the increase of spray distance to 14 cm, the bond strength declined to 18.65$0.65 MPa. This could well mean that the decline in deposition e$ciency as a result of the increase in spray distance may have a!ected the adhesion performance to the substrate, thereby, reducing the bond strength of the coatings.
1232
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 10. Fracture toughness of heat-treated SHA coatings.
Fig. 11. Tensile bond strength as a function of the spray distance and particle size.
Identical adhesion properties were observed with the increase in particle size. At a spray distance of 10 cm, the bond strength of SHA 75}125 lm coatings exhibited a low value of 10.53$0.83 MPa. The large proportion of unmelted particles coupled with the corresponding increase in porosities in the coatings may have resulted in the degradation of the adhesion capability. Delamination of the above-mentioned coating caused by poorly formed lamellar structure and splats was also observed which may have contributed to the drop in adhesion strength.
Research has also indicated the inter-dependency of adhesion strength and crystallinity, which is interrelated with the particle size. It was reported earlier that a high level of crystallinity commonly associated with larger particle size coatings would tend to generate coatings with poorer mechanical properties (that is, coatings with low microhardness and elastic modulus values or high H/E ratio) as a result of its microstructural disabilities. Hence, a lower count of crystallinity level (as in SHA 20}45 lm coatings) would deliver coatings with higher adhesion strength [36]. Brossa et al. have rea$rmed the
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
Fig. 12. Fracture surface of SHA 20}45 lm coating at a spray distance of 10 cm.
1233
ed from variations in inherent crystallinity level of the coatings, pore morphology, porosities and interlamellar structure. SHA 20}45 lm coating at a spray distance of 10 cm possesses the most favourable mechanical properties. However, the enhanced mechanical properties as a result of heat treatment have introduced into the coatings thermal defects in the form of inter- and intralamellar cracks. Such microstructural defects were found to have little or no e!ect on the overall mechanical strength of the coatings. The fracture toughness of the coatings possesses identical behavioural properties, which showed increasing toughness at increasing heat treatment temperature, but declining toughness with larger particle sizes. Mechanical properties were found to decrease with increasing test load.
Acknowledgements e!ect and dependency of the adhesion strength on the crystallinity of the coatings [37]. The fracture mode of all coatings consists of a mixture of cohesive and adhesive failure, with cohesive failure more commonly found in high strength coatings. The lack of good interlamellar contact and the presence of failure initiation sites within the coating [38] as a result of the increase in the amount of unmelted and porosities content in low strength coatings (SHA 75}125 lm coatings) might have promoted premature fracture at the glue}coating interface region, thus resulting in adhesive failure. A characteristic fractography of the SHA 20}45 lm coating is shown in Fig. 12.
4. Conclusion The mechanical properties of the as-sprayed and heattreated SHA coatings as a function of the spray distances and particle sizes were reviewed. It was found that the microhardness and elastic moduli improved with the e!ects of increasing heat treatment temperature, but deteriorated with increasing spray distances and particle sizes. A thermal treatment of 8003C was found to enhance better mechanical properties. However, no statistical signi"cance (P'0.1) in mechanical strength improvements was observed for higher heat treatment temperature (9003C). A shorter spray distance (10 cm) using smaller size feedstock (SHA 20}45 lm) would produce coatings with much more favourable (P(0.05) mechanical properties such as hardness, elastic modulus and fracture toughness and features than coatings produced at larger spray distance (12 or 14 cm) and particle size (SHA 75}125 lm). It was also noted that coatings with a low degree of crystallinity would tend to inherit better mechanical properties with a low H/E ratio, and vice versa. The changes in hardness and modulus values were associated with microstructural changes that result-
The continued support of the research grant JT ARC 4/96 is gratefully acknowledged.
References [1] Wallace JS, Ilavsky J. Elastic modulus measurements in plasma sprayed deposits. J Therm Spray Tech 1998;7(4):521}6. [2] Malmberg S, Heberlein J. E!ect of plasma spray operating conditions on plasma jet characteristics and coating properties. Proceedings of the Nat Therm Spray Conference, 1993. p. 451}6. [3] Vardelle M, Vardelle A, Fauchais P. Spray parameters and particle behaviour relationships during plasma spraying. J Therm Spray Tech 1993;2(1):79}92. [4] Cheang P, Khor KA. Thermal spraying of hydroxyapatite (HA) coatings: e!ects of powder feedstock. J Mater Proc Technol 1995;48:429}36. [5] Cheang P, Khor KA. Addressing processing problems associated with plasma spraying of hydroxyapatite coatings. Biomaterials 1996;17:537}44. [6] Leigh SH, Lin CK, Berndt CC. Elastic response of thermal spray deposits under indentation tests. J Am Ceram Soc 1997;80(8): 2093}9. [7] Marshall DB, Noma T, Evans AG. A simple method for determining elastic-modulus-to-hardness ratios using Knoop indentation measurements. J Am Ceram Soc 1982;65(10):C175}6. [8] Tanaka M. Fracture toughness and crack morphology in indentation fracture of brittle materials. J Mater Sci 1996;31:749}55. [9] Kweh SWK, Khor KA, Cheang P. The production and characterization of hydroxyapatite (HA) powders. J Mater Process Technol 1999;89}90:373}7. [10] Sung YM. Microindentation analysis of inviscid melt-spun ceramic "bres. J Mater Sci Lett 1997;16:1804}6. [11] Beshish GK, Florey CW, Worzala FJ. Lenling WJ. Fracture toughness of thermal spray ceramic coatings determined by the indentation technique. J Therm Spray Tech 1993;2(1):35}8. [12] Slosarczyk A, Bialoskorski J. Hardness and fracture toughness of dense calcium-phosphate-based materials. J Mater Sci: Mater Med 1998;9:103}8. [13] Wang BC, Chang E, Lee TM, Yang CY. Changes in phases and crystallinity of plasma-sprayed hydroxyapatite coatings under heat treatment: a quantitative study. J Biomed Mater Res 1995; 29:1483}92.
1234
S.W.K. Kweh et al. / Biomaterials 21 (2000) 1223}1234
[14] Shi KS, Qian ZY, Zhuang MS. Microstructure and properties of sprayed ceramic coating. J Am Ceram Soc 1988;71(11):924}9. [15] Lutz EH. Microstructure and properties of plasma ceramics. J Am Ceram Soc 1994;77(5):1274}80. [16] McPherson R, Cheang P. Elastic anisotropy of APS alumina coatings and its relationship to microstructure. High Perf Ceram Films Coatings 1991:277}90. [17] McPherson R. A review of microstructure and properties of plasma sprayed ceramic coatings. Surface Coat Technol 1989; 39/40:173}81. [18] Malmberg S, Heberlein J. E!ect of plasma spray operating conditions on plasma jet characteristics and coating properties. J Therm Spray Tech 1993;2(4):339}44. [19] Khor KA, Cheang P. Characterization of thermal sprayed hydroxyapatite powders and coatings. J Therm Spray Tech 1994; 3(1):45}50. [20] Liu DM, Chan HM, Wu JD. Plasma-sprayed HA coating: e!ect of di!erent calcium phosphate ceramics. J Mater Sci: Mater Med 1994;5:147}53. [21] Bauer G. Biochemical aspects of osseo-integration. In: Heink G, editor. Osseo integrated implants, vol. 1. Boca Raton, FL: CRC Press, 1990. p. 81}97. [22] Ji H, Marquis PM. E!ect of heat treatment on the microstructure of plasma-sprayed hydroxyapatite coating. Biomaterials 1993; 14(1):64}8. [23] Filaggi MJ, Pilliar RM, Coombs NA. Post-plasma-spraying heat treatment of the HA Coating/Ti}6Al}4V implant system. J Biomed Mater Res 1993;27:191}8. [24] Gross KA, Berndt CC. Thermal processing of hydroxyapatite for coating production. J Biomed Mater Res 1998;39:580}7. [25] Cao Y, Weng J, Chen JY, Feng JM, Yang ZJ, Zhang XD. Water vapour-treated hydroxyapatite coatings after plasma spraying and their characteristics. Biomaterials 1996;17(4):419}24. [26] McPherson R, Gane N. Structural characterization of plasmasprayed coatings. J Mater Sci: Mater Med 1995;6:327}34.
[27] Gross KA, Berndt CC, Herman H. Amorphous phase formation in plasma-sprayed hydroxyapatite coatings. J Biomed Mater Res 1998;39:407}14. [28] Chen J, Wolke JGC, Groot K. Microstructure and crystallinity in hydroxyapatite coatings. Biomaterials 1994;15(5):396}9. [29] Khor KA, Cheang P, Wang Y. The thermal spray processing of HA powders and coatings. JOM 1997;49(2):51}7. [30] Tong W, Chen J, Li X, Cao Y, Yang Z, Feng J, Zhang X. E!ect of particle size on molten states of starting powder and degradation of the relevant plasma-sprayed hydroxyapatite coatings. Biomaterials 1996;17(15):1507}13. [31] Voort GFV, Lucas GM. Microindentation hardness testing. Adv Mater and Proc 1998;9:21}5. [32] Quinn GD. Hardness testing of ceramics. Adv Mater Proc 1998;154(2):23}7. [33] Lawn BR, Evans AG, Marshall DB. Elastic/Plastic indentation damage in ceramics: the median/radial crack system. J Am Ceram Soc 1980;63(9}10):574}81. [34] Ponton CB, Rawlings RD. Dependence of the vickers indentation fracture toughness on the surface crack length. J Br Ceram Trans 1989;88:83}90. [35] Aoki H. Medical applications of hydroxyapatite. Tokyo: Ishiyaku EuroAmerica Inc., 1994. p. 290}1. [36] Yang CY, Liu RM, Wang BC, Lee TM, Chang E, Hang YS, Chen PQ. In vitro and in vivo mechanical evaluations of plasmasprayed hydroxyapatite coatings on titanium implants: the e!ect of coating characteristics. J Biomed Mater Res 1997; 37:335}45. [37] Brossa F, Cigada A, Chiesa R, Paracchini L, Consonni C. Adhesion properties of plasma sprayed hydroxyapatite coatings for orthopaedic prostheses. Biomed Mater Engng 1993; 3(3):127}36. [38] Berndt CC, Gross KA. Characteristics of hydroxyapatite biocoatings. Proceedings of the International Therm. Spray Conference & Expo, Orlando, USA, 1992. p. 465}70.