Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/51813230

Neutrondiffractionresidualstrainmeasurementsinnanostructuredhydroxyapatitecoatingsfororthopaedicimplants

Article·November2011

DOI:10.1016/j.jmbbm.2011.07.003·Source:PubMed

CITATIONS

13

READS

71

5authors,including:

Someoftheauthorsofthispublicationarealsoworkingontheserelatedprojects:

Developmentoffunctionally-gradedmaterialsforwheel-railcontactusinglasercladding.View

project

BenefitsofneutronscatteringincommercialapplicationsViewproject

R.Ahmed

Heriot-WattUniversity

88PUBLICATIONS913CITATIONS

SEEPROFILE

NadimulHaqueFaisal

RobertGordonUniversity

42PUBLICATIONS216CITATIONS

SEEPROFILE

MichaelE.Fitzpatrick

CoventryUniversity

201PUBLICATIONS1,749CITATIONS

SEEPROFILE

K.A.Khor

NanyangTechnologicalUniversity

325PUBLICATIONS10,197CITATIONS

SEEPROFILE

AllcontentfollowingthispagewasuploadedbyAnnaParadowskaon04December2014.

Theuserhasrequestedenhancementofthedownloadedfile.Allin-textreferencesunderlinedinblue

arelinkedtopublicationsonResearchGate,lettingyouaccessandreadthemimmediately.

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/jmbbm

Research paper

Neutron diffraction residual strain measurements innanostructured hydroxyapatite coatings fororthopaedic implants

R. Ahmeda,b,∗, N.H. Faisala,b, A.M. Paradowskac, M.E. Fitzpatrickd, K.A. Khore

a School of Engineering and Physical Sciences, Heriot-Watt University, Edinburgh, EH14 4AS, UKbCollege of Engineering, Alfaisal University, Riyadh 11533, Saudi Arabiac ISIS, Rutherford Appleton Laboratory, Didcot, OX11 0QX, UKdMaterials Engineering, The Open University, Milton Keynes, MK7 6AA, UKe School of Mechanical and Aerospace Engineering, Nanyang Technological University, Nanyang Avenue, 639798, Singapore

A R T I C L E I N F O

Article history:

Received 16 February 2011

Received in revised form

19 May 2011

Accepted 6 July 2011

Published online 4 August 2011

Keywords:

Hydroxyapatite coatings

Residual strain

Neutron diffraction

Plasma spray

Orthopaedic implants

A B S T R A C T

The failure of an orthopaedic implant can be initiated by residual strain inherent to

the hydroxyapatite coating (HAC). Knowledge of the through-thickness residual strain

profile in the thermally sprayed hydroxyapatite coating/substrate system is therefore

important in the development of a new generation of orthopaedic implants. As the coating

microstructure is complex, non-destructive characterization of residual strain, e.g. using

neutron diffraction, provides a useful measure of through thickness strain profile without

altering the stress field. This first detailed study using a neutron diffraction technique,

non-destructively evaluates the through thickness strain measurement in nanostructured

hydroxyapatite plasma sprayed coatings on a titanium alloy substrate (as-sprayed, heat

treated, and heat treated then soaked in simulated body fluid (SBF)). The influence of

crystallographic plane orientation on the residual strain measurement is shown to indicate

texturing in the coating. This texturing is expected to influence both the biological and

fracture response of HA coatings. Results are discussed in terms of the influence of heat-

treatment and SBF on the residual stress profile for these biomedical coatings. The results

show that the through thickness residual strain in all three coatings was different for

different crystallographic planes but was on average tensile. It is also concluded that the

heat-treatment and simulated body fluid exposure had a significant effect on the residual

strain profile in the top layers of HAC.c⃝ 2011 Elsevier Ltd. All rights reserved.

c

d

1. Introduction

The structure–property relationships of biomaterials such ashydroxyapatite {HA,Ca10(PO4)6(OH)2} play an important role

∗ Corresponding author at: School of Engineering and Physical Scien451 4722; fax: +44 0 131 451 3129.

E-mail address: R.Ahmed@hw.ac.uk (R. Ahmed).

1751-6161/$ - see front matter c⃝ 2011 Elsevier Ltd. All rights reservedoi:10.1016/j.jmbbm.2011.07.003

es, Heriot-Watt University, Edinburgh, EH14 4AS, UK. Tel.: +44 0 131

in the response of artificial implants (Leeuwenburgh et al.,

2008). HA has an almost identical chemical composition to

that of the mineral component of bone, and can be suc-

cessfully thermally sprayed on metallic orthopaedic implants

.

2044 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4

Nomenclature

dhkl Lattice interplanar spacingd0hkl Strain-free lattice interplanar spacing

Greek symbols

αCTE Coefficient of thermal expansionε Elastic strainσ Stress

Abbreviations

APS Air Plasma SprayAS As SprayedAS-HAC As Sprayed Hydroxyapatite CoatingCTE Coefficient of thermal expansionHT-HAC Heat-treated Hydroxyapatite CoatingHT-SBF-HAC Heat-treated Hydroxyapatite Coating im-

mersed in Simulated Body FluidHAC Hydroxyapatite CoatingHAp HydroxyapatiteHT Heat TreatedSBF Simulated Body FluidSEM Environmental Scanning Electron MicroscopyXRD X-ray Diffraction.

(Gross et al., 2004). It has excellent biocompatibility and isosteoconductive, allowing bone cells to grow on its surface(Sun et al., 2001; Williams, 2008). The dissolution and miner-alization behaviour of HA coatings are important factors in itsperformance (Cölfen, 2010; Boivin, 2007), and are affected byphase composition, crystalline nature, porosity, surface area,residual stress, surface roughness, morphology and crystalsize (Gross et al., 2004; Levingstone, 2008).

Clinical investigations have reported that the failure ofHA-coated orthopaedic implants can occur at the coat-ing/substrate interface (Gross et al., 2004). This failure is influ-enced by the residual stress, as it influences the cohesive andadhesive strength of the coating. Measurement of residualstress in thermal spray coatings is critical for modelling anddesigning improved orthopaedic implants. The influence ofcoating powder characteristics, spraying process, and spray-ing process parameters on the residual stress field of HA coat-ings has been the topic of previous investigations (e.g. Tsuiet al., 1998; Yang and Chang, 2001; Heimann et al., 2001; Yanget al., 2003; Yang and Chang, 2003; Stanic et al., 2004; Fioroet al., 2004; Fogarassy et al., 2005; Topic et al., 2006; Che et al.,2007; Renghini et al., 2008; Yang and Yang, 2008).

Non-destructive through thickness residual strain mea-surements in thick coating/substrate systems is possible viathe high penetration depth achieved by the neutron diffrac-tion technique (Ahmed et al., 2007, 2008). Although neutrondiffraction has been used to investigate strain in the past(Millet et al., 2001; Ahmed et al., 2010) in these systems,through thickness evaluation of strains in the HA coat-ing/substrate system has not yet been realized. The primaryaim of the current study, which we believe is the first de-tailed investigation of its kind, was to non-destructively in-vestigate, using neutron diffraction, the relative changes in

Fig. 1 – SEM image of nanostructured hydroxyapatite(n-HAp) powder.

the through-thickness residual strain and stress profile inplasma sprayed HA coatings. These coatings were depositedon Ti-6Al-4V substrate and strain measurements were con-ducted in the as-sprayed, heat-treated, and heat-treated thenimmersed in simulated body fluid (SBF) conditions.

2. Materials and methods

2.1. Coated test specimen

Wet chemical synthesis was used to manufacture nanos-tructured hydroxyapatite (n-HAp) powder of spherical mor-phology with a porous structure as shown in Fig. 1. Thespray-dried HA powder was produced through rotary atom-ization of HA slurry (Lima et al., 2005; Xu et al., 2004). The ag-glomerates were typically 50 µm, and were annealed at 900 ◦Cfor 2 h prior to air plasma spraying. The wet chemical reac-tion yields needle shaped HA measuring 40 nm across andaround 200 nm in length as previously described elsewhere(Lima et al., 2005; Xu et al., 2004). Following spray drying theneedle shapemorphology is retained, and after annealing, theneedle shaped HA forms fine grains typically 100 nm is size.Air plasma spray coating of the HA powder was performed us-ing the process parameters listed in Table 1. The hydroxyap-atite coatings (HAC) were 220 µm thick and were deposited ona substrate disc of 30 mm diameter and 6 mm thick commer-cial titanium alloy (Ti-6Al-4V). The Ti-6Al-4V substrate waslightly grit blasted to prepare the surface for the coating de-position. The three conditions of plasma sprayed HA speci-mens prepared are listed in Table 2, which include as-sprayed(AS-HAC), heat-treated (HT-HAC), and heat-treated and thensoaked in simulated body fluid (HT-SBF-HAC).

Heat-treatment of the as-sprayed HA coated specimenwasdone at 600 ◦C for 3 h in an inert (argon gas) atmosphere.Simulated body fluid (SBF) is ametastable buffer solution. TheSBF was prepared according to the procedure described in Liet al. (2004) by dissolving reagent grade CaCl2, KH2PO4.3H2O,NaCl, KCl, MgCl2.6H2O, CaHCO3 and Na2SO4 in distilledwater. The heat-treated coated specimens were soaked in 60ml SBF for 6 h. After soaking, the specimens were removedfrom the solution, rinsed with distilled water and dried atroom temperature.

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4 2045

Table 1 – Thermal spray conditions and powders.

Spray gun Air Plasma Spray (APS)Spray material (powder) HydroxyapatiteMain gas (Ar) flow rate 37.7 l/minAuxiliary gas (He) flow rate 23.6 l/minCarrier gas (Ar) flow rate 18.8 l/minSpray distance 75 mmRobotic arm velocity 100 mm/sGun power 12.1 kW

Table 2 – Air plasma sprayed hydroxyapatite coatedspecimen details.

No. Hydroxyapatitecoating specimen type

Conditions

1 As-sprayed (AS-HAC) As-sprayed2 Heat-treated (HT-HAC) Heat treated at 600 ◦C for 3 h3 Immersed in simulated

body fluid afterheat-treatment(HT-SBF-HAC)

Heat treated at 600 ◦C for 3 hand immersed in 60 ml SBFfor 6 h

2.2. Microstructural characterization

Coating microstructure was investigated after fracturing thecoating to reveal coating porosity and splat morphology. Themicrostructure of the coatings was examined by scanningelectron microscopy (SEM). X-ray diffraction (XRD) analysiswas used to reveal the crystalline phase content in the HAcoatings. A Bruker AXS D8 ADVANCE X-ray diffractometeroperating at 40 kV and 40 mA with Cu–Kα radiation(λ = 0.1542 nm) was used for these measurements. Thegoniometer was run from 2θ = 5◦–120◦ with a step of 0.009◦

at 46.2 s per step.

2.3. Nanoindentation measurements

Nanoindentation experiments at 100 mN load were carriedout to obtain the elastic modulus and hardness using a cal-ibrated NanoTest R⃝ (MicroMaterials, UK) indenter equippedwith a Berkovich tip. For each loading–unloading cycle, load-ing and unloading lasted 15 s respectively, and a dwell time of5 s at each peak load was used. Ten measurements were per-formed on polished coating surfaces (AS-HAC and HT-HAC)and thirty on each cross-section (AS-HAC, HT-HAC and HT-SBF-HAC), which were distributed in 3 lines of 10 measure-ment points each, at 50 µm, 100 µm and 150 µm respectivelyfrom the coating’s surface. Measurement on the HT-SBF-HACsurface was not conducted as the polishing would remove theSBF affected area, leading to an incorrect representation ofthe SBF sample surface.

Similarly, twenty measurements were performed on eachsubstrate cross-section near the interface, distributed in twolines of 10 measurement points each, at 50 µm and 100 µmrespectively from the interface. The raw data were then usedto construct the force–displacement (P–h) curve and resultswere evaluated based on the Oliver and Pharr method (Oliverand Pharr, 1992).

2.4. Measurement of neutron diffraction residual strainand stress

Neutron diffraction strain measurements were performed atthe UK ISIS Facility, using the ENGIN-X strain measurementdiffractometer (Ahmed et al., 2007, 2008). This is a pulsedneutron diffractometer equipped with slits and collimatorsto achieve small gauge volumes (Daymond et al., 1997).The experiments were conducted in vertical scan mode tomeasure through thickness residual strain profile of thecoating–substrate system and a gauge volume height of theorder of 300–400 µm was used using a slit size of 10 mm ×

0.2 mm. The details of the vertical scan method are describedelsewhere (Ahmed et al., 2008). Strain measurements wereperformed at the centre of the specimen. To achieve athrough-thickness residual strain profile with high resolution,a partially-submerged beam was used for measurementsnear the coating surface, and a beam submerged in thecoating and substrate materials near the coating–substrateinterface. By using the fast vertical scan and careful levellingof specimen using theodolites, the coating surface and thecoating–substrate interface was located with a resolution ofabout 50 µm. This indicates the positional accuracy withwhich measurement points within the sample were located.When the gauge volume is partially-submerged, a smallergauge size is obtained than is defined by the aperture sizealone as detailed elsewhere (Edwards, 2003).

The strain was obtained from the shift in individual HAand Ti-6Al-4V diffraction peaks for the coating and substratematerials using a single peak fitting routine (Withers et al.,2000), and also Rietveld refinement. The General StructureAnalysis System (GSAS) software (Larson and von Dreele,2004) was used in the data analysis. The time-of-flight (TOF)data was normalized by the incoming flux distribution. Twodiffraction patterns, one for each 90◦ detector bank, weregenerated from the TOF data, and simultaneously analysedby the Rietveldmethod with GSAS. The crystallographic spacegroup (SG) P63/m with a lattice parameter of a = 9.424 Å andc = 6.884 Åwas used in this calculation for the hydroxyapatite(HA) phase. The fitting parameters of the refinement suchas wR and χ2 were in the range of 0.75–1.2 and 0.85–1.15,respectively, confirming good fit to the diffracting pattern.The peaks chosen for the strain analysis within each coatingmaterial were such that they had no or minimal overlap withother peaks. The strain free lattice parameter (d0hkl) for thecoating material was obtained by removing the coating fromthe substrate and crushing the deposit to form a powder.This powder was then put in a vanadium tube and its latticeparameter measured. The strain free lattice parameter for thesteel substrate was measured at the surface of the uncoatedsubstrate disc. The direct elastic strain in the material in themeasured direction was calculated as in Eq. (1):

εhkl =dhkl − d0hkl

d0hkl(1)

where dhklis the measured interplanar spacing and d0hkl is thestress-free interplanar spacing for the material. The elasticstress was calculated from these strain values using theappropriate elastic modulus.

2046 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4

Fig. 2 – Unpolished coating surface and fractured cross-section: (a) AS-HAC, (b) HT-HAC, and (c) HT-SBF-HAC. Boxed sectionindicate fractured cross-section of semimolten HAp granule.

3. Results

3.1. Coating microstructure

Scanning electron microscopic images of all the three HAcoatings are presented in Fig. 2. The sprayed surfaces (Fig. 2(i))reveal splats resulting from the impact of molten and semi-molten particles. There was no major difference in surfacemorphology between the two HA coatings (AS-HAC and HT-HAC) surface and cross-sections, except for the HT-SBF-HAC,where the surface as well as the fractured specimens showsmore detailed features. Somemicrocracks were also observedon the surface (Fig. 2(i)). Significant coating roughness can beseen in all the three coatings. Fractured specimens (Fig. 2(ii))show a qualitatively higher porosity for the AS-HAC andHT-HAC on the fractured surface.

3.2. X-ray diffraction analysis

The XRD spectrum (Fig. 3) of the as-received HA powdershowed the presence of a dominant HA phase. Fig. 3 summa-rizes the crystallographic planes and dhkl-spacing identified

a

b

c

d

Fig. 3 – XRD pattern of hydroxyapatite: (a) spray powder,(b) as-sprayed coating (AS-HAC), (c) heat-treated coating(HT-HAC), and (d) heat-treated and soaked in simulatedbody fluid coating (HT-SBF-HAC).

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4 2047

Fig. 4 – Nanoindentation results (hardness and elasticmodulus) of APS (HAC and Ti-6Al-4V substrate) coatedspecimens; [CS: coating surface, C: through thicknesscoating section, S: substrate section near interface, Ec, Es:elastic modulus of coating or substrate respectively, Hc, Hs:hardness of coating or substrate respectively, note:nanoindentation tests were not done for coating surface ofHT-SBF-HAC specimen].

5×10

510

6

Neu

tron

cou

nts

/ Å

a

b

1.5 2 2.5

d _ Spacing (Å)

Fig. 5 – Neutron diffraction pattern: (a) crushed AS-HACpowder used as d0hkl (i.e. stress-free interplanar spacing forthe coating), and (b) near the coating–substrate interface inthe as-sprayed hydroxyapatite coating (AS-HAC).

through X-ray diffraction analysis of the hydroxyapatite coat-ings.

3.3. Nanoindentation testing

The results of the nanoindentation testing are presented inFig. 4. Each data point in Fig. 4 represents an average of10 measurements. The results indicate the upper layers ofthe AS-HAC coating have higher hardness than the HT-HACcoating, with little or no hardness variation in the Ti-6Al-4Vsubstrate. The measurements of elastic modulus indicate theupper layers of the HT-HAC coating have higher values thanthe AS-HAC coating with relatively similar elastic modulus forall Ti-6Al-4V substrates.

3.4. Neutron diffraction measurement of residual strainand stress

A typical neutron diffraction spectrum is shown in Fig. 5,showing the diffraction pattern in the vicinity of thecoating–substrate interface in the AS-HAC coating, compared

Fig. 6 – Neutron diffraction residual strain profile of HAcoatings (for planes: 022 and 030) and Ti-6Al-4V substrate[single peak fitting routine].

Fig. 7 – Neutron diffraction residual strain profile of HAcoatings (for planes: 211 and 213) and Ti-6Al-4V substrate[single peak fitting routine].

with the crushed AS-HAC coating powder used to providethe d0hkl stress-free interplanar spacing. Individual diffractionpeaks corresponding to crystallographic planes (022), (030),(211), and (213) were used to measure the residual strain ineach of the three HA coatings, whereas diffraction peaks cor-responding to crystallographic planes (102) and (103) wereused to measure the residual strain in the titanium alloy(Ti-6Al-4V) substrate. Figs. 6 and 7 summarize the residualstrain measurement results obtained from individual diffrac-tion peaks, the x-axis of which is plotted on a log-scaleto highlight the differences in coating–substrate strain. Thetrends of residual strain for diffraction peaks correspondingto crystallographic planes (022) and (213) were largely com-pressive, whereas (030) and (211) were largely tensile.

After analysing individual diffraction peaks results, resi-dual strain results in the three coatings (as shown in Fig. 8)were analysed using Rietveld refinement. Fig. 9 represents thecorresponding residual stress measurement. The averagingof strain across the four planes, although not shown here,also gave similar results to the Rietveld refinement (Figs. 8and 9). An average value of elastic modulus (Fig. 4) wasused where measurement depth location of residual strainand elastic modulus did not match. These results indicatethat the average coating and substrate residual strain/stressvaried significantly with the coating conditions. The average

2048 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4

Fig. 8 – Neutron diffraction residual strain profile of HAcoatings and Ti-6Al-4V substrate [Rietveld refinementmethod].

Fig. 9 – Neutron diffraction residual stress profile of HAcoatings and Ti-6Al-4V substrate [Rietveld refinementmethod].

residual strain/stress (see Figs. 8 and 9) in the HT-HAC and HT-SBF-HAC coatings changes from tensile near surface towardscompressive near the interface, whereas the average residualstrain and residual stress in the AS-HAC coating is tensile.

4. Discussion

4.1. Coating morphology

It was noted that the coating showed different surface fea-tures (Figs. 2 and 10) from those in the surface of the start-ing feedstock nanostructured HA (n-HAp) powder (Fig. 1). Themicroscopic investigation demonstrated splat morphologieswith distinct columnar grain structure (Fig. 10), and revealingpores, voids, inter-splat areas and some cracks in splats withlittle or no fragmentation in the coating splats. The columnargrains observed in coatings indicate a degree of texturing inthe coating microstructure (Inagaki and Kameyama, 2007).

Fig. 2 shows the coating’s surface, which was mainly com-posed of splats with some semi-molten HAp granules, includ-ing surface connected macropores, and surface connectedcracks in splats, which was consistent with previously re-ported results for plasma sprayed HAp coatings (e.g. Tsuiet al., 1998; Yang et al., 2003; Chung and Chang, 2005; Li et al.,

Fig. 10 – SEM image of fractured HT-SBF-HAC coating:magnified view of the boxed area in Fig. 2(c)(ii) showingzone separated at coating surface and fractured coatingcross-section indicating (i) top layer columnar structure,(ii) un-molten part of the feedstock, and (iii) globular porein splat.

2004). The pores form as a result of poor bonding between ad-jacent splats, whereas microcracking arises from shrinkage ofthe splat during quenching and subsequent differential ther-mal contraction between substrate and coating (Tsui et al.,1998). Typical morphology changes of the HT-SBF-HAC surfacesplats induced by SBF (in vitro incubation for 6 h) shown inFigs. 2 and 10, indicate that the splat has dissolved into theSBF (diffusion and dissolution), which was consistent withpreviously reported results for plasma sprayed HAp coatings(Li et al., 2004).

The XRD spectrum (Fig. 3) analysis revealed the presenceof a dominant HA phase (Ca10(PO4)6(OH)2, JCPDS No. 9–432).However, comparing the spectra for all three coatings(AS-HAC, HT-HAC and HT-SBF-HAC) showed improvedcrystallinity of the HA phases compared to the spray powder,which was consistent with previously reported results forplasma sprayed HAp coatings where SBF interaction leadto crystal growth in the near-surface region (Yang et al.,2003; Chung and Chang, 2005; Renghini et al., 2008). Thephase of the spray dried HA is crystalline HA. Duringplasma spraying however negligible amounts of tri-calciumphosphate, tetra-calcium phosphate, amorphous calciumphosphate and calcium oxide can form as indicated in Fig. 3.

4.2. Nanoindentation testing

A recent investigation (Karlis et al., 2010) reported the findingsof nanoindentation analysis of amorphous and crystallinephases in HA coatings. This investigation indicated that thehardness of amorphous and crystalline HA phases rangedfrom 1.5 ± 0.3 GPa and 3.0 to 7.7 GPa, respectively. The averagehardness measured in the current investigation lies withinthis range (Fig. 4), between 2 and 5 GPa. When comparingthe surface hardness of AS-HAC and HT-HAC, one shouldexpect a higher hardness for the latter due to the improvedcrystallinity of the HT coatings (Fig. 3). However, despite thescatter in hardness values, the trend is reversed in Fig. 7,which is due to the larger number of micro-cracks observed

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4 2049

on the surface of the polished HT specimen. Through-thickness hardness measurements of the three specimensindicate very little difference except for the near-surface HT-SBF-HAC measurement, which is attributed to the expectedcrystal growth due to SBF interaction in the near-surfaceregion (Yang et al., 2003; Chung and Chang, 2005; Renghiniet al., 2008).

Heat treatment of thermal spray coatings is expected toimprove inter-splat bonding, and hence the elastic modulus,the extent of which depends upon the heat treatmentconditions (Ahmed et al., 2008). Similarly, improvement inHA crystallinity can also increase the elastic modulus. Inthe current investigation, there was an increase in averageelastic modulus, especially in the surface and near surfaceregions, which is attributed to the changes in both thecrystallinity and inter-splat bonding. The average values ofthese modulus measurements are broadly in agreement withthe previously reported average modulus values of 48 ± 6 GPain the amorphous and 100–120 GPa in the crystalline region(Karlis et al., 2010). Previous investigations have indicatedthat the modulus of thermally-sprayed material is about1/3 of the bulk material (Ahmed et al., 2008). Single crystalstudies on HA using nano-indentation have indicated valuesof 125–150 GPa (Viswanath et al., 2007; Zamiri and De, 2011).Based on the 1/3 assumption the coating deposit is expectedto have an average modulus of 50 GPa, which relates moreclosely to the values quoted for the amorphous phase (Karliset al., 2010). As the coating exhibited good crystallinity, themeasured values range from 60 to 120 GPa (Fig. 4). Themodulus of the coating will however be highly dependentnot only on the coating process and deposition parameters,but also the hardness and modulus of the start powder andmethod of modulusmeasurement. Previously reported valuesof indentation modulus shown in Table 3 indicate a muchbroader scatter of values which range from 5 to 100 GPa dueto these factors (Yang and Chang, 2001; Heimann et al., 2001;Millet et al., 2001; Yang et al., 2003; Yang and Chang, 2003;Stanic et al., 2004; Fioro et al., 2004; Fogarassy et al., 2005;Topic et al., 2006; Che et al., 2007; Renghini et al., 2008; Yangand Yang, 2008; Jakani et al., 2007).

4.3. Coating residual strain and residual stress profile

A review of published literature on the residual stressvalues of HA coatings summarized in Table 3 indicatesthat the residual stress values in HA coatings are verymuch dependent upon the measurement method. X-raydiffraction values of measured residual stress using thesin2 ψ technique for HA coatings deposited in a range ofcoating thicknesses and deposition conditions, indicate thatthe measured stress is always compressive at the coatingsurface (Table 3). As discussed later, this is inconsistent withthe mechanisms of residual stress generation on the basisof coefficient of thermal expansion (CTE) difference and alsothe role of quenching stresses. The values quoted from thesynchrotron X-ray method largely indicate tensile residualstress values varying through the coating thickness. Althoughboth conventional and synchrotron X-ray investigationsprovide non-destructive measures of residual stress, thedifference in reported values is subtle, and can originate from

three sources i.e. (i) measurement method: the conventionalsin2 ψ technique does not require d0 spacing evaluationfor stress measurement whereas synchrotron X-ray andneutron measurements require d0; (ii) the diffraction planeused for stress measurement, as there can be intergranularstress partitioning that affects the results; and (iii) depth ofmeasurement e.g. as the penetration depth of conventionalX-rays is small (typically 6 µm in HA (Heimann et al.,2001)), surface roughness effects and sample preparatione.g. grinding or polishing can influence the measured stressvalue.

Residual strain and stress results presented in Figs. 6–9indicate that, with all other factors constant, the measuredstrain is dependent upon the measurement plane, which isindicative of a textured coating. Residual stress measure-ment planes which are almost parallel to the c-axis of theHCP structure ((030) & (211)) indicate tensile residual strain,whereas planes ((022) & (213)) which are inclined to thec-axis indicate compressive values for all measured speci-mens. Similarly, the magnitude of measured strain values ishigher in the former planes. This dependency of measuredvalues on plane is not unusual on the basis of volume con-servation and Poisson’s ratio. However, it does indicate thatwhen a single peak fitting analysis is used e.g. in the sin2 ψ

technique in X-ray measurements (Table 3), the results willbe sensitive to the selected peak, and hence reported com-pressive stress values in HA are not unusual. Texturing inthermal spray coatings in general can originate from quench-ing of lamellae and plastic deformation of un-molten coatingparticles (Iordanova and Forcey, 1997). Thermophysical andthermomechanical characteristics of coating particles duringcoating deposition therefore control the degree of texturingin thermal spray coatings. Plasma sprayed HA coatings inparticular are known to generate varying degrees of texturein (002), (004) and (211) peaks (Christopher and Craig, 1995;Tong et al., 1996). A recent study has indicated that it is pos-sible to enhance the texture of HA coatings during plasmaspraying, as each major crystal face of HA i.e. (100) face(a-face) and (001) face (c-face) of HA possess different prop-erties such as protein adsorption and corrosivity in saliva (In-agaki and Kameyama, 2007). The analysis of multiple peakspresented in this investigation therefore highlights this de-pendency of texture, which can form during HA coating for-mation in thermal spraying, on residual strain. This texturingand its influence on residual strain will not only influence thebiological response (Inagaki and Kameyama, 2007), but alsoits fracture behaviour, which will improve in grains that areoriented such they carry a compressive residual stress. Formacro-mechanical property evaluations, an averaged value ofstrain e.g. as indicated in Fig. 8 can be used to interpret themacro-strain the coated HA specimen. Such detailed strainanalysis is currently not possible using other techniques e.g.sin2 ψ where a single peak is measured or semi-destructivetechniques such as hole-drilling or curvature where an aver-aged stress is evaluated. Even for neutron measurements, al-though this represents a powerful method of measurement,complications do arise, as in most cases the gauge volumecontains diffraction peaks from both the coating and sub-strate materials (Fig. 5), which overlap and reduce the accu-racy of Rietveld peak refinement.

2050 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4

Tabl

e3

–Li

tera

ture

resu

lts

ofre

sid

ual

stre

ssin

ther

mal

lysp

raye

dH

Aco

atin

gs.

Res

idual

stress

mea

surem

ent

method

Coa

ting

pro

cess

HA

coating

thickn

ess

(µm

)

Elas

ticm

odulusof

HA

coating(E

c,GPa

),m

easu

rem

ent

method

HA

crys

tal

plane

(hkl)

Subs

tratedetails,

Dim

ension

(l×

w×

t),m

m

Res

idual

stress

inHA

coating(σR,M

Pa)

+ve

:ten

sile

-ve:

com

press

ive

Com

men

tson

residual

stress

pro

file

ofco

ating

andsu

bstrate

Ref.

Neu

tron

diffrac

tion

APS

200,

150,

100,

40–

(213

)Ti-6A

l-4V

,451

×30

×5

Cou

ldnot

bemea

sured

Res

idual

stresses

insu

bstrate

werebo

thco

mpressivean

dtensile

in(101

)plane

(Milletet

al.,

2001

)

APS

220

–(211

)Ti-6A

l-4V

,30mm

dia.,

6mm

thick

Res

idual

micro

strain:

Abo

ut0to

+15

00Thro

ugh

thickn

esstensile

pro

file

inco

ating;

Res

idual

strain

insu

bstratewere

compressiveat

interfac

ein

(102

)&(103

)plane

(Ahmed

etal.,20

10)

X-ray

diffrac

tion

APS

180

±20

16.2

(3-p

ointbe

nding)

(621

)Ti-6A

l-4V

,25.4mm

dia.,

3.3mm

thick

Attopsu

rfac

e:−17

.4,

−20

.2,−

22.3,−

18.8,

−18

.0,−

26.6

Res

idual

stress

chan

gedwith

temperature

ofsu

bstratean

dco

olingco

nditions

(Yan

gan

dChan

g,20

01)

APS

160–

270

5.34

±1.83

,5.30

±1.15

(300

)Ti-6A

l-4V

,50

×20

×2,

25mm

dia.,70

mm

thick

Attopsu

rfac

e:−20

±2

(as-sp

raye

d);

−26

±2

(incu

bation

for24

h);

−29

±2(incu

bation

for

4wee

k);

Com

pressiveresidual

stress

increa

sedwithincu

bation

time

( Heiman

net

al.,20

01)

APS

190

±20

24.1

±1.7,

21.2

±1.5,

15.6

±1.7

–Ti-6A

l-4V

,22

×17

×3.3,

Attopsu

rfac

e:−44

.2(0

wee

k);−

31.8

(1wee

k);

−25

.8(2

wee

k);−

25.1

(3wee

k);−

21.2

(4wee

k);

Com

pressiveresidual

stress

dec

reas

edwithincu

bation

time

(Yan

get

al.,

2003

)

APS

180

±20

16.2,2

4.1(3-p

oint

bending)

(621

)Ti-6A

l-4V

,22

×17

×3.3,

25.4

mm

dia.,70

mm

thick

Attopsu

rfac

e:−18

.6,

−17

.7,−

17.0,−

16.8,

−16

.3,−

15.6,−

15.5,

−15

.0,−

14.7

Com

pressiveresidual

stress

dec

reas

edwithth

ickn

essof

coatingper

dep

ositionpas

s

(Yan

gan

dChan

g,20

03)

APS

260

±30

22.5,2

2.9,

23.2

(3-p

ointbe

nding)

(621

)Ti-6A

l-4V

,50

×15

×3.3

Attopsu

rfac

e:−39

.7,

−35

.6,−

56.5

Com

pressiveresidual

stress

(Chungan

dChan

g,20

05)

Slurry

dipping/

sintering

25,1

00–

(222

),(522

)ZrO

2tough

ened

Al 2O3

ceramic

Attopsu

rfac

e:0to

−20

;Atbo

ttom

laye

rs:0

to11

0Com

pressiveresidual

stress

intoplaye

rsan

dtensile

inbo

ttom

laye

rs

(Cheet

al.,

2007

)

APS

50,2

0024

.8(as-sp

raye

d),17

.3(4-w

eekim

mersion

),(3-p

ointbe

nding)

–Ti-6A

l-4V

,22

×17

×3.3

For50µm:−

21.0,−

17.3,

−13

.0,

For20

0µm:−

25.2,

−20

.5,−

15.8

Com

pressiveresidual

stress

for

stress

alon

g0◦

,along45

◦an

dalon

g90

◦to

thesp

raying

direc

tion

,res

pec

tive

ly

( Yan

gan

dYan

g,20

08)

APS

200,

150,

100,

40(213

)Ti-6A

l-4V

,451

×30

×5

200µm

thick:

−12

±815

0µm

thick:

−27

±610

0µm

thick:

−26

±740µm

thick:

−13

±9

Com

pressivein

coating

(Milletet

al.,

2001

)

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4 2051

Tabl

e3

(con

tinued

)

Res

idual

stress

mea

surem

ent

method

Coa

ting

pro

cess

HA

coating

thickn

ess

(µm

)

Elas

ticm

odulusof

HA

coating(E

c,GPa

),m

easu

rem

ent

method

HA

crys

tal

plane

(hkl)

Subs

tratedetails,

Dim

ension

(l×

w×

t),m

m

Res

idual

stress

inHA

coating(σR,M

Pa)

+ve

:ten

sile

-ve:

com

press

ive

Com

men

tson

residual

stress

pro

file

ofco

ating

andsu

bstrate

Ref.

Synch

rotron

radiation

(high

energy

X-ray

)

APS

60–7

010

8(213

)Ti-6A

l-4V

,10

×10

×1.5

Ti-6A

l-4V

(biaxial

in-p

lanemethod

):σ11

=−62

±14σ22

=

1±

17

Res

idual

stress

inco

ating

could

not

bemea

sured;

Res

idual

stresses

insu

bstrate

surfac

ewereco

mpressive

(Stanic

etal.,

2004

)

APS

120

––

Ti-6A

l-4V

with

Ti-co

atingan

d40

%Ti+

60%HA

coating

atinterfac

e

HA:0

–+30

;Subs

trate:

0–−13

0Thro

ugh

thickn

esstensile

pro

file

inco

ating;

Res

idual

stress

insu

bstratewere

compressive

(Fioro

etal.,

2004

)

APS

120

±10

4.9

–Ti-6A

l-4V

,20

×10

×2.7

Ave

rage

+25

.4in

HA

coating;

Ave

rage

−76

.5in

Ti-6A

l-4V

;

Thro

ugh

thickn

esstensile

&co

mpressivepro

file

inco

ating;

Res

idual

stress

insu

bstrate

wereco

mpressiveat

interfac

e

(Fog

aras

syet

al.,20

05)

APS

50,8

0,10

0–

–Ti-6A

l-4V

,Fo

r50µm:0

to+10

00Fo

r80µm:0

to+50

0Fo

r10

0µm:0

to+50

0

Thro

ugh

thickn

esstensile

pro

file

inco

ating;

Thinner

coatings

had

higher

tensile

stress

values

( Ren

ghini

etal.,20

08)

Finiteelem

ent

method

APS

120

±10

4.9

–Ti-6A

l-4V

,20

×10

×2.7

Ave

rage

+21

.7in

HA

coating;

Ave

rage

−83

.2in

Ti-6A

l-4V

;

Thro

ugh

thickn

esstensile

pro

file

inco

ating;

Res

idual

stress

insu

bstratewere

compressiveat

interfac

e

(Fog

aras

syet

al.,20

05)

Numerical

method

APS

100–

1000

0.5–

5.5GPa

(ten

sion

,co

mpressionan

dflex

uraltes

t)

–Ti-6A

l-4V

,Mildstee

lFo

rTi-6A

l-4V

subs

trate:

21±

2;36

±3;

18±

1;38

±4;

41±

6Fo

rmild

stee

lsubs

trate:

19±

2;26

±3;

17±

1;28

±4;

29±

4

Thro

ugh

thickn

esstensile

pro

file

inco

ating;

Res

idual

strain

insu

bstratewere

chan

gingfrom

compressiveto

tensile

(Tsu

ietal.,

1998

)

Curvature

method

APS

180

±20

100GPa

–Ti-6A

l-4V

,20

×20

×2

−12

0.8,

−11

7.7,

−56

±10

Chan

gedwithSB

Fincu

bation

time

(Top

icet

al.,

2006

)

Materialr

emov

almethod

APS

260

±30

22.8

(3-p

ointbe

nding)

–Ti-6A

l-4V

,50

×15

×3.3

Attopsu

rfac

e:−44

.0,

−36

.9,−

53.1

Com

pressiveresidual

stress

(Chungan

dChan

g,20

05)

Ram

anpiezo

-sp

ectros

copy

APS

––

–Ti-6A

l-4V

+10

0(spraye

din

air);

+60

(spraye

din

vac.)

Tensile

residual

stress;T

ensile

residual

stress

(Sergo

etal.,

1997

)

2052 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4

As was seen in Figs. 8 and 9, the measurements of theaverage residual strain/stress profile in the coatings variedsignificantly with the coating conditions. The average resid-ual strain and residual stress in the HT-HAC and HT-SBF-HAC coatings changes from tensile (near the surface) tocompressive (near the interface), whereas the average resid-ual strain/stress in the AS-HAC coating is slightly tensile.These residual strain values illustrate the combined effect ofthermal mismatch based on the difference in the coefficientof thermal expansion (αCTE) of the coating/substrate system,the role of quenching stress in individual lamellae, and theeffect of phase transformations (Tsui et al., 1998; Fogarassyet al., 2005; Ahmed et al., 2007, 2008). These factors are dis-cussed next on the basis of physical mechanisms dictatingthe residual stress behaviour.

1. The role of deposition and heat-treatment temperature. Macroresidual stress occurs due to the differences in the CTE ofthe coating and substrate materials. Different approacheshave been adapted in the published literature to model theinfluence of this mechanism of stress generation, which varyfrom the bending Almen test to mathematical and numericalmodels (Gill, 1993). A simplified mathematical approachutilizes the mathematical equation for the evaluation of thisresidual stress (σ) component using Eq. (2) (Gill, 1993; Ahmedet al., 2008):

σc,s = Ec,sxl

− αCTEc,s1T

(2)

where1 Ec,s and αCTE represent the elastic modulus andcoefficient of thermal expansion (CTE) of the coating layerand substrate materials. The term x/l represents the netstrain due to constrained contraction and 1T represents thetemperature change. This will lead to higher macro-tensilestress in the HA coatings as2 αCTEc > αCTEs (Soliman andWaheed, 1999). This is consistent with the averaged residualstrain and stress values in the HA coatings (Figs. 8 and9). Similarly during the heat-treatment of HA coatings thistensile residual stress will become more pronounced (Eq. (2)for 1T = 600 ◦C) which also is consistent with the observedbehaviour of strain comparison between the AS-HAC and HT-HAC coatings (Figs. 8 and 9).

Based upon Eq. (2), the average residual stress in thecoating and substrate materials can be approximated as+2764 MPa and −55 MPa respectively, for the case of a HAcoated specimen cooling from 600 ◦C to room temperatureduring the heat-treatment. A similar approximation on thebasis of a finite element model predicts an average stressof +2000 MPa and −15 MPa, respectively. The predictedaverage residual stress in the coatingmaterial is higher, but inbroad agreement with the experimentally measured coatingstress near the coating surface. Stress values predicted bythe models however are not very accurate as they assumeperfect bonding at the coating/substrate interface and donot consider any phase changes and changes in bondingmechanism which occur during the heat-treatment.

1 Subscripts c and s relate to the materials properties for thecoating layer and substrate, respectively.

2 αCTE of sintered HA: 11.5 × 10−6 K−1 and αCTE of Ti-6Al-4V:8.9 × 10−6 K−1 (www.matweb.com).

2. The role of quenching stress. The quenching stress, whichappears within individual splats, is caused by the constrainedcontraction of the solidifying splat at it cools, since itscontraction is constrained by the underlying lamella ormaterial. This again is a function of the temperature ofindividual splats. As the inter-splat bonding improves duringthe heat-treatment, quenching stresses are expected toreduce, however the strain and stress results indicate that theCTE influence on the residual strain and stress profile wasmore dominant than the expected reduction in quenchingstress.

3. Phase transformation, crystallinity and crystallographic planes.As indicated in Fig. 3, there was a higher degree of crys-tallinity after the heat-treatment and also further changesafter SBF immersion. Although the influence of thesetransformations is appreciable between different planes ofstrain measurement (Figs. 6–9) when compared between theHT-HAC and HT-SBF-HAC samples, the overall influence onthe averaged strain and stress values (Figs. 8 and 9) is small.Improvement in crystallinity after the SBF treatment is ex-pected to cause strain changes based upon the expectedvolume changes between the amorphous and crystallinestructures. Quantitative measure of these changes was notconsidered in this investigation.

4.4. Role of pre-existing residual stress on HA coatingdegradation and failure

There is no simple relationship between coating degradation,failure and total stress distribution during application of HAcoated implants. Since the HA coating dissolves faster onelevated areas or those subjected to higher levels of loading(Gross et al., 2004), the combined loading stress (σL) andpre-existing residual stress (σR) can affect the response ofthe materials significantly. Considering a similar level ofloading in implants, the magnitude and nature (compressiveor tensile) of pre-existing residual tensile stress in the implantcoating–substrate system is expected to influence the coatingdegradation and failure. The work presented here providesthe ability to non-destructively measure strain in thick HAcoating–substrate systemswith a view to apply this techniqueto clinical studies, where implants retrieved after the clinicaltrials can be non-destructively measured for residual strain,which otherwise would be difficult using conventional X-raytechniques. Similarly, the findings reported here indicate thatresidual strain is sensitive to peak selection and providesthe ability to comprehensively measure strain in actualimplants as part of the clinical trials using both Rietveld andsingle peak criteria. The biological performance of implantscan therefore be more dependent on strain measured onparticular planes instead of residual strain averaged acrossa number of diffraction peaks.

5. Conclusions

(a) The measured values of residual strain were sensitiveto the plane of strain measurement which clarifies theparadox in reported values of strain in the publishedliterature and clarifies the influence of coating textureand residual strain measurement method on measuredresidual strain.

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4 2053

(b) Compressive residual strain was measured along (022) &(213) planes whereas a tensile strain was measured along(030) & (211) planes. This texturing is expected to influenceboth the biological and fracture response of HA coatings.

(c) Post-treatment of coatings such as HT-HAC and HT-SBF-HAC had a varying degree of influence on the residualstrain in individual planes.

(d) Comparison of the residual average strain and residualstress profile between the AS-HAC, HT-HAC and HT-SBF-HAC coatings indicates that the average tensile residualstrain in the heat-treated coating increases following heattreatment and immersion in SBF.

(e) The degree of crystallinity in HA coatings increase afterthe heat-treatment and soaking in simulated body fluidwith a corresponding change in modulus and hardnessvalues.

Acknowledgements

The authors acknowledge the provision of beam time atthe STFC ISIS Facility, (experiment number RB910205) forthe neutron diffraction measurements. MEF is supportedby a grant through The Open University from The Lloyd’sRegister Educational Trust, an independent charity workingto achieve advances in transportation, science, engineeringand technology education, training and research worldwidefor the benefit of all.

R E F E R E N C E S

Ahmed, R., Faisal, N.H., Knupfer, S.M., Paradowska, A.M., Fitz-patrick, M.E., Khor, K.A., Cizek, J., 2010. Neutron diffrac-tion residual strain measurements in plasma sprayed nanos-tructured hydroxyapatite coatings for orthopaedic implants.Mater. Sci. For. 652, 309–314.

Ahmed, R., Yu, H., Stewart, S., Edwards, L., Santisteban, J.,2007. Residual strain measurements in thermal spray cermetcoatings via neutron diffraction. ASME J. Tribol. 129, 411–418.

Ahmed, R., Yu, H., Stoica, V., Edwards, L., Santisteban, J.R., 2008.Neutron diffraction residual strain measurements in post-treated thermal spray coatings. Mater. Sci. Eng. A 498, 191–202.

Boivin, G., 2007. The hydroxyapatite crystal: a closer look.Medicographia 29, 126–132.

Che, L., Gotoh, M., Yang, L., Hirose, Y., 2007. Residual stress stateof hydroxyapatite graded coating on ceramic substrate. JCPDS-International Centre for Diffraction Data (ISSN: 1097-0002)150–155.

Christopher, M.R., Craig, D.A., 1995. Crystallite orientationand anisotropic strains in thermally sprayed hydroxyapatitecoatings. Biomaterials 16, 691–696.

Chung, Y.C., Chang, E., 2005. Measurement of residual stressesin plasma-sprayed hydroxyapatite coating on titanium alloy.Surf. Coat. Technol. 190, 122–131.

Cölfen, H., 2010. Biomineralization: a crystal-clear view. Nat.Mater. 9, 960–961.

Daymond, M.R., Bourke, M.A.M., VonDreele, R.B., Clausen, B.,Lorentzen, T., 1997. Use of Rietveld refinement for elasticmacrostrain determination and for evaluation of plastic strainhistory from diffraction spectra. J. Appl. Phys. 82, 1554–1562.

Edwards, L., 2003. Near-surface stress measurement usingneutron diffraction. In: Fitzpatrick, M.E., Lodini, A. (Eds.),Analysis of Residual Stress using Neutron and SynchrotronRadiation, vol. 233. Taylor & Francis, London.

Fioro, F., Albertini, G., Girardin, E., Giuliani, A., Manescu, A.,Rustichelli, F., 2004. Neutron and synchrotron radiation non-destructuve methods for the characterisation of materials fordifferent applications. J. Alloys Compounds 382, 39–45.

Fogarassy, P., Cofino, B., Millet, P., Lodini, A., 2005. Residualstress in hydroxyapatite coating: nonlinear analysis and high-energy synchrotron measurements. IEEE Trans. Biomed. Eng.52, 1161–1166.

Gill, S.C., 1993. Residual stress in plasma sprayed deposits, Ph.D.Thesis, Gonville and Caius College, Cambridge University, UK.

Gross, K.A., Walsh, W., Swarts, E., 2004. Analysis of retrievedhydroxyapatite-coated hip prostheses. J. Therm. Spray Tech-nol. 13, 190–199.

Heimann, R.B., Graßmann, O., Zumbrink, T., Jennissen, H.P.,2001. Biomimetic processes during in vitro leaching ofplasma-sprayed hydroxyapatite coatings for endoprostheticapplications. Mat. -Wiss. U. Werkstofftech 32, 913–921.

Inagaki, M., Kameyama, T., 2007. Phase transformation of plasma-sprayed hydroxyapatite coating with preferred crystallineorientation. Biomaterials 28, 2923–2931.

Iordanova, I., Forcey, K.S., 1997. Texture and residual stresses inthermally sprayed coatings. Surf. Coat. Technol. 91, 174–182.

Jakani, S., Citterio, H., Millet, P., Lodini, A., 2007. Neutrondiffraction study of hydroxyapatite crystallites on implant. Eur.Cells Mater. (ISSN: 1473-2262) 13, 45.

Karlis, A.G., Saber-Samandari, S., Heemann, K.S., 2010. Evaluationof commercial implants with nanoindentation defines futuredevelopment needs for hydroxyapatite coatings. J. Biomed.Mater. Res. Part B: App. Biomater. 93B, 1–8.

Larson, A.C., von Dreele, R.B., 2004. General structure analysissystem. Los Alamos National Laboratory Report LAUR 86-748.

Leeuwenburgh, S.C.G., Jansen, J.A., Malda, J., Dhert, W.A.,Rouwkema, J., van Blitterswijk, C.A., Kirkpatrick, C.J., Williams,D.F., 2008. Trends in biomaterials research: an analysis ofthe scientific programme of the World Biomaterials Congress.Biomaterials 29, 3047–3052.

Levingstone, T.J., 2008. Optimisation of Plasma Sprayed Hydroxya-patite Coatings. Ph.D. Thesis, Dublin City University, Ireland.

Li, H., Khor, K.A., Cheang, P., 2004. Thermal sprayed hydroxyap-atite splats: nanostructures, pore formation mechanisms andTEM characterization. Biomaterials 25, 3463–3471.

Lima, R.S., Khor, K.A., Li, H., Cheang, P., Marple, B.R., 2005. HVOFspraying of nanostructured hydroxyapatite for biomedicalapplications. Mater. Sci. Eng. A 396, 181–187.

Millet, P., Girardin, E., Braham, C., Lodini, A., 2001. Residualstress analysis in implant coatings using X-ray and neutrondiffraction techniques. J. Neutron Res. 9, 443–448.

Oliver, W.C., Pharr, G.M., 1992. An improved technique fordetermining hardness and elastic modulus using load anddisplacement sensing indentation experiments. J. Mater. Res.7, 1564–1583.

Renghini, C., Girardin, E., Fomin, A.S., Manescu, A., Sabbioni,A., Barinov, S.M., Komlev, V.S., Albertini, G., Fiori, F., 2008.Plasma sprayed hydroxyapatite coatings from nanostructuredgranules. Mater. Sci. Eng. B 152, 86–90.

Sergo, V., Sbaizero, O., Clarke, D.R., 1997. Mechanical and chemicalconsequences of the residual stresses in plasma sprayedhydroxyapatite coatings. Biomaterials 18, 477–482.

Soliman, H.M., Waheed, A.F., 1999. Effect of differential thermalexpansion coefficient on stresses generated in coating. J.Mater. Sci. Technol. 15, 457–462.

Stanic, V., Carradó, A., Fiori, F., Albertini, G., Buslaps, T., 2004.Characterisation of microstructure and residual stresses inhydroxyapatite coatings on titanium prostheses. J. NeutronRes. 12, 117–122.

Sun, L., Berndt, C.C., Gross, K.A., Kucuk, A., 2001. Materialfundamentals and clinical performance of plasma-sprayedhydroxyapatite coatings: a review. J. Biomed. Mater. Res. 58,570–592.

2054 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 0 4 3 – 2 0 5 4

Tong, W., Chen, J., Li, X., Feng, J., Cao, Y., Yang, Z., Zhang, X.,1996. Preferred orientation of plasma sprayed hydroxyapatitecoatings. J. Mater. Sci. 31, 3739–3742.

Topic, M., Ntsoane, T., Heimann, R.B., 2006. Microstructuralcharacterisation and stress determination in as-plasmasprayed and incubated bioconductive hydroxyapatite coatings.Surf. Coat. Technol. 201, 3633–3641.

Tsui, Y.C., Doyle, C., Clyne, T.W., 1998. Plasma sprayedhydroxyapatite coatings on titanium substrates part 1:mechanical properties and residual stress levels. Biomaterials19, 2015–2029.

www.matweb.com.Viswanath, B., Raghavan, R., Ramamurty, U., Ravishankar, N.,

2007. Mechanical properties and anisotropy in hydroxyapatitesingle crystals. Scr. Mater. 57, 361–364.

Williams, D.F., 2008. On the mechanisms of biocompatibility.Biomaterials 29, 2941–2953.

Withers, P.J., Johnson, M.W., Wright, J.S., 2000. Neutron strainscanning using a radially collimated diffracted beam. PhysicaB 292, 273–285.

Xu, J.L., Khor, K.A., Dong, Z.L., Gu, Y.W., Kumar, R., Cheang,P., 2004. Preparation and characterization of nano-sizedhydroxyapatite powders produced in a radio frequency (rf)thermal plasma. Mater. Sci. Eng. A 374, 101–108.

Yang, Y.C., Chang, E., 2001. Influence of residual stress on bondingstrength and fracture of plasma-sprayed hydroxyapatitecoatings on Ti-6Al-4V substrate. Biomaterials 22, 1827–1836.

Yang, Y.C., Chang, E., 2003. The bonding of plasma-sprayedhydroxyapatite coatings to titanium: effect of processing,porosity and residual stress. Thin Solid Films 444, 260–275.

Yang, Y.C., Chang, E., Lee, S.Y., 2003. Mechanical properties andYoung’s modulus of plasma-sprayed hydroxyapatite coatingon Ti substrate in simulated body fluid. J. Biomed. Mat. Res.A 67, 886–899.

Yang, Y.C., Yang, C.Y., 2008. The influence of residual stresson the shear strength between the bone and plasma-sprayed hydroxyapatite coatings. J. Mater. Sci.: Mater. Med. 19,1051–1060.

Zamiri, A., De, S., 2011. Mechanical properties of hydroxyapatitesingle crystals from nanoindentation data. J. Mech. Behav.Biomed. Mater. 4, 146–152.