Microstructure, bioactivity and osteoblast behavior of ... · Raman spectroscopy revealed that the zirconia coating was composed of monoclinic zirconia which was stable at low temperature,

Acta Biomaterialia 6 (2010) 990–1000

Contents lists available at ScienceDirect

Acta Biomaterialia

journal homepage: www.elsevier .com/locate /ac tabiomat

Microstructure, bioactivity and osteoblast behavior of monoclinic zirconiacoating with nanostructured surface

Guocheng Wang a,b, Fanhao Meng a,b, Chuanxian Ding a,b, Paul K Chu c, Xuanyong Liu a,b,c,*a Key Laboratory of Inorganic Coating Materials, Chinese Academy of Sciences, Shanghai 200050, PR Chinab Shanghai Institute of Ceramics, Chinese Academy of Sciences, Shanghai 200050, PR Chinac Department of Physics & Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong

a r t i c l e i n f o a b s t r a c t

Article history:Received 17 June 2009Received in revised form 27 September2009Accepted 28 September 2009Available online 1 October 2009

Keywords:Monoclinic zirconiaNanostructurePlasma-sprayingBioactivityOsteoblasts

1742-7061/$ - see front matter � 2009 Acta Materialdoi:10.1016/j.actbio.2009.09.021

* Corresponding author. Address: Key Laboratory oChinese Academy of Sciences, Shanghai 200050, PR C

E-mail address: [email protected] (X. Liu).

A monoclinic zirconia coating with a nanostructural surface was prepared on the Ti–6Al–4V substrate byan atmospheric plasma-spraying technique, and its microstructure and composition, as well as mechan-ical and biological properties, were investigated to explore potential application as a bioactive coating onbone implants. X-ray diffraction, transmission electron microscopy, scanning electron microscopy andRaman spectroscopy revealed that the zirconia coating was composed of monoclinic zirconia whichwas stable at low temperature, and its surface consists of nano-size grains 30–50 nm in size. The bondstrength between the coating and the Ti–6Al–4V substrate was 48.4 ± 6.1 MPa, which is higher than thatof plasma-sprayed HA coatings. Hydrothermal experiments indicated that the coating was stable in awater environment and the phase composition and Vickers hardness were independent of the hydrother-mal treatment time. Bone-like apatite is observed to precipitate on the surface of the coating after soak-ing in simulated body fluid for 6 days, indicating excellent bioactivity in vitro. The nanostructured surfacecomposed of monoclinic zirconia is believed to be crucial to its bioactivity. Morphological observationand the cell proliferation test demonstrated that osteoblast-like MG63 cells could attach to, adhere toand proliferate well on the surface of the monoclinic zirconia coating, suggesting possible applicationsin hard tissue replacements.

� 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Plasma-sprayed hydroxyapatite (HA) coatings are commonlyand clinically used in hard tissue replacement due to their chemi-cal similarity to the mineral components of bones and hard tissuesin mammals [1–3]. However, the clinical use of plasma-sprayedhydroxyapatite coatings is plagued by their low crystallinity andpoor bonding strength on titanium alloys [4–6]. The low crystallin-ity gives rise to fast dissolution of the hydroxyapatite coatings dur-ing contact with human body fluids, subsequently shortening theirlifetime, whereas the poor bonding strength results in delamina-tion, causing safety concerns. It is thus important to explore newbioactive coatings to improve tissue integration and control thefriction at the interface of the implants [7].

Chemical and dimensional stability, mechanical strength,toughness, and Young’s modulus similar to that of stainless steelalloys make zirconia an excellent ceramic biomaterial for use asa femoral head [8]. In addition, zirconia has also been used as a sec-ond phase to improve the bonding strength and fracture toughness

ia Inc. Published by Elsevier Ltd. A

f Inorganic Coating Materials,hina. Tel.: +86 21 52412409.

of HA coatings [9–11]. Zirconia shows morphological fixation withthe surrounding tissues without producing any chemical or biolog-ical bonding when implanted [12]. In our previous work, a zirconiafilm with nanostructured surface prepared by cathodic arc deposi-tion was demonstrated to be bioactive [13]. It was also reportedthat zirconia gel with tetragonal or monoclinic structure exhibitedhigher apatite-forming ability in SBF fluids than amorphous gel[14]. A zirconia coating fabricated by micro arc oxidation could alsoinduce apatite precipitation on its surface in modified simulatedbody fluids [15]. Our previous work showed that plasma-sprayedcalcia-stabilized zirconia coating had good bioactivity, which wasdependent on the content of the monoclinic phase in the coating[16]. The phase composition and microstructure of the zirconiacoating should be related to the fabrication technique and process,which may give rise to a difference in the bioactivity.

Zirconia exists in three crystalline phases: monoclinic, tetrago-nal and cubic. Tetragonal and cubic zirconia possess superiormechanical properties but undergo low temperature degradation(LTD) in water or water vapor [17,18]. LTD can reduce the mechan-ical strength and service life of the zirconia-based materials [19,20]and it is thought to be the main reason for the failure of zirconiaartificial joint balls [21]. LTD, which depends on the microstructureand fabrication process, is accelerated by micro-cracks, high rough-

ll rights reserved.

http://dx.doi.org/10.1016/j.actbio.2009.09.021mailto:[email protected]://www.sciencedirect.com/science/journal/17427061http://www.elsevier.com/locate/actabiomat

G. Wang et al. / Acta Biomaterialia 6 (2010) 990–1000 991

ness and pores [17–19]. Generally, monoclinic zirconia is more sta-ble at room temperature than cubic and tetragonal zirconia. Thecalculated energy vs. volume data at absolute zero temperatureconfirmed the higher stability of the monoclinic phase [22]. There-fore, some researchers have made attempts to prepare the densepolycrystalline monoclinic ZrO2 components. Cutler et al. [23]found that annealing at the cubic stability temperature and rapidcooling were crucial to the fabrication of dense undoped mono-clinic zirconia. Samples fabricated by this method exhibited a frac-ture toughness in the range 3.7–6.0 MPa m1/2, a Young’s modulusof about 175 GPa, and a thermal expansion (8.7 � 10�6 C�1) closerto those of Ti alloy than hydroxyapatite [23]. Yttria- and ceria-sta-bilized zirconia coatings with the tetragonal or cubic phase havebeen deposited by plasma-spraying and are widely used in thermalprotection [24–26]. However, to our knowledge, preparation andcharacterization of plasma-sprayed monoclinic zirconia coatingswithout a stabilizer have been seldom reported. Because mono-clinic zirconia is a stable phase at low temperature (99.9%) were purchasedfrom Farmeiya Advanced Materials Co. Ltd. (Jiujiang, China). Toimprove the flowability, the powders were sintered at 1450 �C for6 h in air, mechanically ground, and then spheroidized by a plasmajet. Details of the spheroidizing process can be found in our previouspaper [27]. After spheroidization, zirconia powders 2–20 lm in sizewere obtained. Ti–6Al–4V substrates with dimensions of 10 �10 � 2 mm were ultrasonically cleaned in ethanol and deionizedwater and then grit-blasted with alumina grit. An atmospheric plas-ma-spraying system (Sulzer Metco, Switzerland) was utilized to de-posit the coatings. Argon (35 slpm) and hydrogen (12 slpm) wereused as the primary and auxiliary plasma forming gases, respec-tively. The powder feeding rate was about 20 g min�1 using argon(3.5 slpm) as the carrier gas. The plasma arc current and voltage were620 A and 68 V. The spraying distance was fixed at 120 mm.

The surface morphology of the coating was examined by fieldemission scanning electron microscopy (FE-SEM, JSM-6700F, JEOL,Japan). The crystalline phase of the coating was identified using aSiemens D5000 diffractometer with Cu Ka1 (k = 1.54056 Å) irradia-tion with a scanning step of 0.02. Scans were obtained from 15� to85� at 2� per minute. Micro Raman spectroscopy (Lamb 1B, DilorInc., France) with an excitation wavelength of 632.8 nm was em-ployed to detect the surface phase composition of the coating.The tensile bond strength between coating and substrate was mea-sured in accordance with ASTM C-633-79. For this test, coatings ofapproximately 380 lm in thickness were sprayed on Ti–6Al–4Vrods of 25.4 mm in diameter. A thin layer of E-7 adhesive glue withtensile fracture strength of over 70 MPa was applied. The tensilebonding strength was measured using a universal testing machine(Instron-5592, SATEC, USA) and five samples were tested indepen-dently. The results were reported as means ± standard deviation(SD). The nano-hardness and elastic modulus of the coating weredetermined on three randomly selected polished coating samples

using MTS Nano Indenter� XP (MTS Cooperation, Nano InstrumentInnovation Center, TN, USA). At least seven indents were made oneach coating sample. The final results of nano-hardness and elasticmodulus of the coating were the average values of all individualindentations. Vickers indentation (Wilson–Wolpert Tukon2100B,Instron, Norwood, MA) was used to assess crack formation andpropagation around the indentation; this test was repeated threetimes at separate files on the polished coating surface. In order toevaluate its hydrothermal stability, the coating was hydrother-mally treated in an autoclave at 134 �C for different time periods.Afterwards, the phase composition was determined by XRD andVickers hardness was measured by Vickers indentation; at least10 indents were made.

Before transmission electron microscopy (TEM) observation, theas-sprayed coating samples with a thickness of 1 mm was attachedto a tripod polisher with glue, ground, and polished from both sidesto a thickness of about 30 lm. Afterwards, the thickness of the coat-ing specimen was further reduced using a low-angle ion-thinningprecise ion polishing system (PIPS). TEM examination was carriedout using a JEM-2010 TEM at an accelerating voltage of 200 kV.

2.2. Bioactivity evaluation

After ultrasonic washing in ethanol and rinsing with deionizedwater, the zirconia samples were immersed in a simulated bodyfluid (SBF) at 36.5 �C without stirring to investigate the bioactivity.The SBF solution was prepared according to the method proposedby Kokubo and Takadama [28]. After immersion for 2, 6 and12 days, the samples were removed from the SBF solution, washedwith deionized water, and then dried at 40 �C. In order to investi-gate the variation in the Ca and P ion concentrations in the SBFsolution, the coating samples were immersed in 25 ml of the SBFsolution for 1, 2, 4, 6, 8, 10 and 12 days without stirring. The Caand P concentrations in the SBF solution after immersion for vari-ous time periods were detected by inductively coupled plasmaoptical emission spectroscopy (ICP-OES, AX, Varian, USA). The sur-face morphologies of the coating soaked in the SBF solution for 2, 6and 12 days were evaluated by field emission scanning electronmicroscopy under secondary electron imaging (FE-SEM, JSM-6700F, JEOL, Japan). The composition of the deposits on the surfacewas determined by energy-dispersive X-ray spectrometry (EDS) inthe electron probe (EPMA, JAX-8100, Japan).

2.3. Cytocompatibility evaluation

2.3.1. Cell cultureThe osteoblast-like cell line MG63 (Cells Resource Center of

Shanghai Institute for Biological Science, Shanghai, China) wasseeded on the surface of zirconia coatings to evaluate the cytocom-patibility. The cells were cultured at 37 �C in a humidified atmo-sphere of 5% CO2 in air, in 75 cm2 flasks (Corning Incorporated,USA) containing 10 ml of a-minimum essential medium (a-MEM) (Minimum Essential Medium alpha-Medium, Gibco, Invitro-gen, Inc.), 10% fetal calf serum (FCS) (Excell Biology Inc., SouthAmerica), 2 mM L-glutamine (Hyclone, USA), 1% antimicrobial ofpenicillin, and streptomycin (Antibiotic–Antimycotic, Gibco, Invit-rogen Corporation). The medium was changed every third day andfor the subculture, the cell monolayer was washed twice withphosphate-buffered saline (pH 7.4, Gibco, Invitrogen) andincubated with trypsin–ethylenediaminetetraacetic acid (EDTA)solution (0.25% trypsin, 1 mM EDTA (Chemicon International)) for5–10 min at 37 �C to detach the cells. The effect of trypsin was theninhibited by adding the full medium at room temperature. Thecells were centrifuged (900 rev min�1, 28 �C, 6 min) and resus-pended in the complete medium for reseeding onto the coatingsurface.

992 G. Wang et al. / Acta Biomaterialia 6 (2010) 990–1000

2.3.2. Morphology observationA 1 ml cell suspension with a cell density of 1 � 104 cell ml�1

was added into each well containing the coating samples withdimensions of 10 � 10 � 2 mm3. Two independent experimentswere carried out. The first one was for laser scanning confocalmicroscopy observation (LSCM) and the second one for SEM obser-vation. After culturing for 1, 2, 5, 24 and 48 h, the cell layers wererinsed with PBS (phosphate-buffered saline) three times. The cellswere fixed in a 3% glutaric dialdehyde diluent for 20 min at roomtemperature followed by three rinses with PBS. The cells were per-meabilized with 0.2% (v/v) Triton X-100 (Amresco, USA) for 4 minat room temperature followed by three rinses with PBS. The cellswere stained with rhodamine phalloidin (invitrogen detectiontechnologies, USA) at room temperature for 40 min followed bythree rinses with PBS and then stained with DAPI (Chemical Inter-national) for 5 min. The cytoskeletal actin and cell nuclei were ob-served with scanning laser confocal microscopy (LSCM, ZeissLsm510 meta).

Samples with dimensions of 10 � 10 � 2 mm3 sterilized byc-irradiation were put into the 24-well culture plates (Costar,USA). The l.0 ml cell suspension with a cell density of 1 �105 cell ml�1 was added into each well. The culture plate was trans-ferred gently to a 37 �C incubator. The incubation time durationswere 3, 7 and 11 days. After each time point, the samples were ta-ken out and rinsed with a phosphate-buffered saline solution (pH7.2, PBS) twice to remove unattached cells. All the samples werefixed with 2.5% glutaraldehyde solution in a sodium cacodylate buf-fer (pH 7.4, Gibco, Invitrogen) for 30 min after removal from theculture plate after each incubation time. Prior to SEM observation,the samples were dehydrated in a grade ethanol series (30, 50,75, 90, 95 and 100v/v%) for 10 min, respectively, with final dehydra-tion conducted in absolute ethanol twice followed by drying in thehexamethyldisilizane (HMDS) ethanol solution series.

Fig. 1. Secondary electron image of the plasma-sprayed monoclinic zirconia coatingtaken by SEM: (a) 3000� and (b) 80,000�.

2.3.3. Cell proliferation and vitalityCell proliferation and vitality were determined using alamar-

BlueTM assay (AbD Serotec Ltd., UK) measuring the accumulativemetabolic activity. Polystyrene with the same dimensions as thecoatings cut from tissue culture plate were used as control. Fivesamples were tested for each culturing time period to improve datastatistics. At the end of each culturing time, the culture mediumwas removed and 1 ml fresh medium with 5% alamarBlueTM wasadded to each well. After incubation for 5 h, 100 ll of the culturemedium was transferred to a 96-well plate for measurement.Accumulation of reduced alamarBlueTM in the culture mediumwas determined by an enzyme labeling instrument (BIO-TEK, ELX800) at extinction wavelengths of 570 and 600 nm. The operationprocedures and calculation of cell proliferation or viability of cellsfollowed strictly the instruction of alamarBlueTM assay.

Fig. 2. Backscattering electron image of the cross-section of the plasma-sprayedmonoclinic zirconia coating taken by EPMA.

3. Results and discussion

3.1. Microstructure and phase composition

The surface morphology of the plasma-sprayed zirconia coatingshown in Fig. 1a indicates that the coating possesses a surfaceroughness (Ra) of about 6.48 ± 0.37 lm. Under the higher magnifi-cation (Fig. 1b), a nanostructure can be observed on the coatingsurface. The size of the grains in the nanostructured surface wasestimated to be in the range of 30–50 nm according to 5 microgra-phies which were taken at random. Formation of the nanostruc-tured surface is believed to stem from the high-speed coolingrate in the plasma-spraying process. The cross-sectional view ofthe coating discloses an obvious lamellar structure (Fig. 2). Noobvious gaps between the coating and substrate can be observed.

The bonding strength between the coating and Ti alloy substratemeasured by ASTM C-633-79 was 48.4 ± 6.1 MPa. It is higher thanthat between plasma-sprayed HA coatings and Ti alloy, which weregenerally less than 20 MPa [29,30]. The higher bonding strengthmay be related to the close match in the thermal expansion coeffi-cients between the monoclinic zirconia and Ti alloy, which havebeen reported to be 8.7 � 10�6 �C�1 [23] and 8.8 � 10�6 �C�1 [31],respectively. The images of fracture surface after bond strength

Fig. 3. Fracture surface images of the plasma-sprayed monoclinic zirconia coatingafter bonding strength test (co-cohesion, ad-adhesion).

Fig. 5. XRD patterns of (a) plasma spheroidized powders and (b) plasma-sprayedmonoclinic zirconia coating.


test are shown in Fig. 3. From Fig. 3, it can be seen that both adhe-sive failure (ad, failure at the interface between the coating and thesubstrate) and cohesive failure (co, failure at the interface betweencoating layers) appeared during the bond strength evaluation. Itindicates that the measured bonding strength is a combination ofadhesive and cohesive strength.

The cross-sectional TEM micrograph and selected area diffraction(SAD) pattern of the coating are depicted in Fig. 4. In comparisonwith Fig. 1b, a significant increase in the grain size is observed. Thebright field image shows that the interior columnar grains are about100–200 nm in diameter. The size difference in the interior and exte-rior grains is caused by the different heating processes. The interiorgrains are continuously heated by the plasma jet while the exteriorones cool quickly after plasma-spraying. The same phenomenonwas observed from plasma-sprayed TiO2 coatings [32]. The SAD pat-tern indicates that the grains in the as-sprayed coating were com-posed of monoclinic zirconia, as shown in the inset.

The XRD patterns acquired from the zirconia powders and zir-conia coating are exhibited in Fig. 5. Both the powders and coatingwere composed of monoclinic zirconia. However, the diffractionpeaks from the coating were broader in comparison with thosefrom the powders. This may be caused by the formation of someamorphous phase and nanosized grains during plasma-spraying.The Raman spectrum of the as-sprayed coating (Fig. 6) shows thatall the bands correspond to those of monoclinic zirconia reportedin Ref. [33], further proving that the coating was composed ofmonoclinic zirconia. In the dissertation of Kim [34], it was also

Fig. 4. Cross-sectional TEM micrograph and (inset) SAD pattern of the plasma-sprayed monoclinic zirconia coating.

reported that plasma-sprayed zirconia coating fabricated frompure zirconia powders exhibited the monoclinic form.

3.2. Mechanical properties and hydrothermal stability

The hardness (H) and elastic modulus (E) of the plasma-sprayedmonoclinic zirconia measured by nano-indentation are listed in

Fig. 6. Raman spectrum acquired from the plasma-sprayed monoclinic zirconiacoating.

Table 1Elastic modulus (E) and hardness (H) of the zirconia coating in comparison with3 mol.% Y2O3–ZrO2 [35] and 12.8 mol.% CaO–ZrO2 [16].

MonoclinicZrO2 coating

3 mol.% Y2O3–ZrO2 coating

12.8 mol.%CaO–ZrO2 coating

Phase composition M T C + 8.7% MElastic modulus E (GPa) 139.72 ± 24.96 136.41 ± 25.95 178.53 ± 27.31Hardness H (GPa) 7.60 ± 1.77 9.20 ± 2.98 13.05 ± 2.92H/E 0.055 0.067 0.073

Note: M, monoclinic phase; T, tetragonal phase; C, cubic phase.


Table 1. For comparison, the 3 mol.% Y2O3–ZrO2 coating composedof the tetragonal phase [35] and 12.8 mol.% CaO–ZrO2 coating com-posed of mainly the cubic phase [16] are summarized. All tests onthese two coatings were performed under the same condition asthe monoclinic zirconia coating. The term H/E represented the de-gree of elastic response in elastic–plastic materials [36,37]. It wastaken as a useful indicator of a given incapacity of materials to ab-sorb impact energy, that is to say, materials with lower H/E value

Fig. 7. Typical SEM micrographs showing the crater created after Vickers inden-tation: (a) monoclinic coating, (b) 3 mol.% Y2O3–ZrO2 coating and (c) 12.8 mol.%CaO–ZrO2 coating (500 g force, 15 s).

Fig. 8. XRD patterns of the plasma-sprayed monoclinic zirconia coating (a) beforeand after hydrothermal treatment at 134 �C for (b) 24 h, (c) 72 h and (d) 144 h.

Fig. 9. Vickers hardness of the plasma-sprayed monoclinic zirconia coating afterhydrothermal treatment at 134 �C for various time periods.

Fig. 10. Ca and P concentrations in SBF solution after immersion for various timeperiods (one measurement was taken).


possess stronger capacity to dissipate impact energy [38]. Table 1shows that the elastic modulus and nano-hardness of the plas-ma-sprayed monoclinic zirconia coating are 139.72 ± 24.96 and7.60 ± 1.77 GPa, respectively. In comparison with the 3 mol.%Y2O3–ZrO2 and 12.8 mol.% CaO–ZrO2 coatings, the plasma-sprayedmonoclinic zirconia coating has the lowest H/E value, indicatingthat it was more resistant to exterior energy than the other twocoatings. The typical morphology observed by SEM from the inden-

Fig. 11. SEM micrographs of the plasma-sprayed monoclinic zirconia coatingimmersed in SBF solution for (a) 2 days, (b) 6 days and (c) 12 days.

tation craters formed by the Vickers indenter on the plasma-sprayed monoclinic zirconia coating and the reference coatingsare displayed in Fig. 7. After Vickers indentation, no obvious crackswere formed around the impression on the plasma-sprayed mono-clinic zirconia coating surface while spallation was observed on the12.8 mol.% CaO–ZrO2 coating, further indicating that the plasma-sprayed monoclinic coating may have better toughness.

In addition, the stability of the plasma-sprayed monoclinic zir-conia coating in water was investigated. The XRD patterns andVickers hardness of the coating before and after hydrothermaltreatment in a 134 �C autoclave are shown in Figs. 8 and 9, respec-tively. Fig. 9 illustrates that the Vickers hardness was lower thanthe nano-hardness, as shown in Table 1. It was influenced by themicrostructure of the coating, such as pores and micro-cracks.Moreover, both the phase composition (Fig. 8) and hardness(Fig. 9) were independent of the hydrothermal treatment time.Therefore, it can be inferred that the plasma-sprayed monocliniczirconia coating is stable in a water environment.

3.3. Bioactivity evaluation

The bioactivity of the plasma-sprayed monoclinic zirconia coat-ing was investigated by examining precipitation of apatite on the

Fig. 12. XRD patterns of the plasma-sprayed monoclinic zirconia coating immersedin SBF solution for (a) 2 days, (b) 6 days and (c) 12 days.

Fig. 13. SEM micrographs of the surface of the polished monoclinic zirconia coatingimmersed in SBF solution for 28 days.


surface in the SBF solution. The Ca and P concentrations in the SBFsolution after immersion for a period of time are shown in Fig. 10.It can be seen that the Ca and P concentrations in the SBF solutiondecreased with immersion time, implying that the plasma-sprayedmonoclinic zirconia coating had the ability to induce precipitationof Ca and P compounds onto the surface in the SBF solution. TheXRD patterns of the coatings soaked in SBF for 2, 6 and 12 daysare displayed in Fig. 11. The secondary peak (26�) of the HA crys-talline phase can be observed from the XRD pattern of the coatingsoaked in SBF for 6 days. With increasing immersion time, it be-came more intense. Because the primary peak (2h = 32�) of thecrystalline HA is in close proximity to the [111] peak of the second-ary strong peak (2h = 31.5�) of the monoclinic zirconia phase, it isdifficult to discern them from each other. The corresponding sur-face morphologies of the coatings immersed in SBF solution for 2,6 and 12 days are depicted in Fig. 12. The surface of the coating

Fig. 14. LSCM images of MG63 cells cultured on the surface of the plasma-sprayed monstained with rhodamine phalloidin (red) and the nucleus is stained with DAPI (blue); thethe references to colour in this figure legend, the reader is referred to the web version o

soaked in SBF for 6 days was covered by many newly formed HAparticles (Fig. 12b) but none can be found on the surface of thecoating soaked in SBF for 2 days (Fig. 12a). After immersion inSBF for 12 days, the surface was completely covered by HA parti-cles and many tortoise-shell-like micro-cracks are clearly seendue to drying (Fig. 12c). The EDS results in the inset of each SEMmicrograph indicate that the newly formed particles are a cal-cium–phosphorus compound which is consistent with XRD results(Fig. 11). According to phase analysis (Fig. 11) and the morphologyobservation (Fig. 12), it can be concluded that the monoclinic zir-conia coating is able to induce hydroxyapatite formation and thuspossesses good bioactivity. The bioactivity was independent of thecoating thickness, but was strongly dependent on the character ofthe coating surface. In this work, the SBF immersion test was alsoperformed on the polished coating whose nanostructured surfacewas removed. Apatite does not form on the surface of the polished

oclinic zirconia coating for (a) 1 h, (b) 2 h, (c) 5 h, (d) 24 h and (e) 48 h. (F-actin isdensity of the cell suspension is 1 � 104 cell ml�1) Bar = 20 lm. (For interpretation of

f this article.)


zirconia coating soaked in SBF for 28 days, as shown in Fig. 13. Theresults confirm that the nanostructured surface is necessary for theobserved bioactivity.

The bioactivity of the monoclinic zirconia coating is believed tobe related to its phase composition and the nanostructured surface.During immersion in SBF, water molecules react with zirconia dis-sociating to form surface hydroxyl groups. The quantity and natureof the surface hydroxyl groups depend on the crystalline structureof zirconia [39,40]. Although the monoclinic phase has the lowestbulk energy among the three zirconia polymorphs [41], the surfaceenergy of the monoclinic phase is higher [42]. In contact withwater molecules, there is a higher tendency to lower its energyby dissociative adsorption of water, which is exothermic. Theadsorption enthalpies of half monolayer H2O coverage onthe monoclinic zirconia surface is �142 kJ mol�1 whereas that onthe tetragonal surface is �90 kJ mol�1 [43,44], indicating the high-er reactivity of the monoclinic zirconia surface. Accordingly, thesurface concentration of hydroxyl groups on the monoclinic ZrO2surface is higher than that on tetragonal ZrO2 [40]. It has been re-ported that the OH concentration is 6.2 molecule nm�2 on mono-clinic zirconia with the BET surface area of 19 m2 g�1, which ismuch higher than 3.5 molecule nm�2 for tetragonal zirconia with

Fig. 15. SEM pictures of MG63 cells cultured on the plasma-sprayed monoclinic zirconsuspension is 1 � 104 cell ml�1).

a BET surface area of 20 m2 g�1. Moreover, in the former case, thedominant species are tri-bridged hydroxyl groups which have beenshown to be more acidic and in the latter case, there is a combina-tion of bi-bridged and terminal hydroxyl groups. The higher aciditycaused by (Zr)3OH on the monoclinic zirconia surface makes it eas-ier for proton to be donated, forming negatively charged (Zr)3O�.Pettersson et al. [45] have reported that the isoelectric point ofmonoclinic zirconia without dopant is 6.4, indicating that themonoclinic zirconia surface should be negatively charged in SBFat a pH value of 7.4. The negatively charged surface naturally at-tracts calcium ions from the SBF solution. Deposition of calciumions ensues, followed by the crucial step of hydroxyapatite nucle-ation on the surface of the bioactive implant [46]. Deposition ofCa2+ is then followed by the arrival of HPO42�, forming calciumphosphate.

The nanostructured surface is crucial to the in vitro bioactivity.In our previous work, we confirmed that the nanostructured sur-faces of plasma-sprayed TiO2 coatings and cathode arc depositedZrO2 films possess better ability to induce apatite formation thanconventional surfaces [32,13]. In this work, apatite cannot precip-itate on the polished monoclinic zirconia coating after the nano-structured surface structure has been removed, indicating that

ia coating for (a) 1 h, (b) 2 h, (c) 5 h, (d) 24 h and (e) 48 h (the density of the cell


the nanostructure is a necessary condition to achieve enhancedbioactivity. Viitala et al. [47] have suggested that a nanostructuredtopography strongly affects the negatively charged surface in thatit increases the amount of OH per unit area, and so a surface com-posed of finer nanocrystalline particles has a higher surface chargedensity than larger ones [48].

3.4. Cytocompatibility evaluation

The cytoskeleton of osteoblast-like MG63 cells seeded on theplasma-sprayed monoclinic zirconia coating was observed by laserscanning confocal microscopy (LSCM) after filamentous actin (F-actin) and nuclei staining. The monoclinic zirconia coating sup-ports continuous cellular growth from 1 to 48 h, as shown inFig. 14. After culturing for 1 h, the cells on the coating surface ex-hibit a round shape (Fig. 14a), showing that cell contact and attach-ment involve gravitation/sedimentation to the coating surfacewhereupon physical and biochemical forces act to close the cell–surface gap [49]. Afterwards, some adherent cells spread over thesurface of the coating and the red halo at the periphery of the cells,especially the spread ones, become brighter (Fig. 14b–d). Addition-ally, some cells are observed to connect to the cell walls (green ar-row) at the mitosis phase, indicating that the cells proliferateduring this period. After culturing for 48 h, the cells show a mul-ti-polar spindle morphology with a well-organized cytoskeletonstructure, as illustrated in Fig. 14e. Some large cells spanning a longvertical distance cannot be fully imaged by LSCM and they appearto be cut off. Hence, only the cell nucleus (blue) is shown but thewhole cell skeleton (red) cannot be observed. The surface rough-ness is an important factor influencing the cellular response[50,51]. In this study, the zirconia coating has micron-sized surfaceroughness (Ra = 6.48 ± 0.37 lm), which may contribute to the en-hanced osteoblast-like cell attachment and proliferation. Thesefindings are consistent with other studies. Ramaswamy et al. [51]have reported that Ca3ZrSi2O9 ceramic samples having a surface

Fig. 16. SEM photographs of MG63 cells cultured on the plasma-sprayed monoclinicsuspension is 1 � 105 cell ml�1).

roughness of 6.80 ± 0.77 lm possess good ability to support theadhesion and growth of osteoblasts.

The SEM micrographs of the osteoblast-like MG63 cells seededon the plasma-sprayed monoclinic zirconia coating are displayedin Fig. 15. After seeding for 1 h, the cells on the zirconia coating ex-hibit a round morphology (Fig. 15a), which is consistent with theLSCM observation. A number of secretory vesicles are observedon the surface and many filopodia extend from the MG63 cells asshown in Fig. 15a. After seeding for 2 h, the cells become largerand flatter. The filopodia become more elongated and extend toall directions (Fig. 15b). After seeding for 5 h, the cells enlargeand flatten obviously (Fig. 15c) and are in close contact with thecoating via the extended filopodia. However, at this time point,not all the cells flatten to the same level and some cells seem totransform slowly as shown in the inset of Fig. 11c. With further in-crease in the culturing time to 24 h, all the cells completely flatten(Fig. 15d). After culturing for 48 h, the cells appear to be more elon-gated and thicker (Fig. 15e). The results indicate that the MG63cells can attach and adhere well to the plasma-sprayed monocliniczirconia coating. Cell attachment and adhesion are the first phaseof cell/material interaction and the efficacy and quality of this firstphase will influence the capability of the cells to proliferate anddifferentiate upon contact with the implant [52].

Fig. 16 displays the morphology of the MG63 cells seeded on thecoating surface for 3–11 days showing higher cell densities. Afterseeding for 3 days, the number and size of cells increase, indicatingthat cells grow and proliferate well on the surface (Fig. 16a). Cellconfluence appears and a sheet-like layer is formed on the coatingafter the cells have been seeded for 7 days, as shown in Fig. 16b.The cells seeded on the coating for 11 days had a more elongatedmorphology and cover the whole surface (Fig. 16c). The results ofcell proliferation tests performed on the monoclinic coating andpolystyrene control displayed in Fig. 17 indicate that osteoblastswere viable on the zirconia coating surface as well as on the poly-styrene surface. No significant difference in the vitality can be ob-

zirconia coating for (a) 3 days, (b) 7 days and (c) 11 days (the density of the cell

Fig. 17. Percentage alamarBlueTM reduction for MG63 cells for different culturingperiods on the plasma-sprayed monoclinic zirconia coatings and polystyrenecontrol (the density of the cell suspension is 1 � 105 cell ml�1).


served from the osteoblasts cultured on the coating and control,thereby corroborating that the zirconia coating possesses goodcytocompatibility.

The good cytocompatibility of the plasma-sprayed monocliniczirconia coating was possibly related to its surface nanostructure.It was reported that nanosized topography improved cell attach-ment, adhesion and proliferation on biomaterials [53,54]. Websteret al. [55] have proposed that the critical grain size for osteoblastadhesion is 49–67 nm for Al2O3 and 32–56 nm for TiO2. The surfaceof the plasma-sprayed monoclinic zirconia coating is constructedby grains with the size range of 30–50 nm, which appears suitablefor cell attachment and viability.

4. Conclusions

Monoclinic zirconia coatings were fabricated by atmosphericplasma-spraying. The grains in the outermost layer of the coatingwere 30–50 nm in size, whereas those in the interior were100–200 nm. The plasma-sprayed monoclinic zirconia coatingexhibited better toughness than the zirconia coating with stabilizer,and bonded tightly to Ti–6Al–4V substrate. Phase composition andhardness of the monoclinic zirconia coating had no evident changesafter hydrothermal treatment in an autoclave at 134 �C for 168 h,indicating its stability in water. After immersion in SBF for 6 days,the surface of the monoclinic zirconia coating was covered by apa-tite, thus demonstrating good bioactivity. The nanostructured sur-face and monoclinic phase are two important factors for theexcellent bioactivity. The in vitro cell culture experiments revealedthat the monoclinic zirconia coating supported cell attachmentand adhesion on its surface and then enhanced cell proliferation.The results suggest that the monoclinic zirconia coating is poten-tially useful in hard tissue replacement due to its good bioactivityand biocompatibility, high bonding strength with titanium alloys,and high stability in an aqueous environment.

Acknowledgements

This work was jointly supported by National Basic Research Fundunder Grant 2005CB623901, Shanghai Science and Technology R&DFund under Grant 0952nm04400, 07QH14016, 07JC14057,0852nm03300 and 08ZR1421600, Shanghai-Unilever Research andDevelopment Fund 09520715200, National Natural Science Founda-

tion of China 30700170, and Hong Kong Research Grants Council(RGC) General Research Funds (GRF) CityU 112307.

Appendix A. Figures with essential colour discrimination

Certain figures in this article, particularly Figs. 3 and 14, are dif-ficult to interpret in black and white. The full colour images can befound in the on-line version, at doi:10.1016/j.actbio.2009.09.021.

References

[1] Suchanek W, Yoshimura M. Processing and properties of hydroxyapatite-basedbiomaterials for use as hard tissue replacement implants. J Mater Res1998;13:84–117.

[2] Geesink RGT, De Groot K, Klein C. Chemical implant fixation usinghydroxyapatite coatings. Clin Orthop Relat Res 1987;225:147–70.

[3] Jarcho M, Kay JF, Gumaer KI, Doremus RH, Drobeck HP. Tissue, cellular andsubcellular events at a bone-ceramic hydroxyapatite interface. J Bioeng1977;1:79–92.

[4] Radin SR, Ducheyne P. Plasma spraying induced changes of calcium phosphateceramic characteristics and the effect on in vitro stability. J Mater Sci MaterMed 1992;3:33–42.

[5] Sun L, Berndt CC, Khor KA, Cheang HN, Gross KA. Surface characteristics anddissolution behavior of plasma sprayed hydroxyapatite coating. J BiomedMater Res 2002;62:228–36.

[6] Chean P, Khor KA. Addressing processing problems associated with plasmaspraying of hydroxyapatite coating. Biomaterials 1996;17:537–44.

[7] Kim BK, Bae HE, Shim JS, Lee KW. The influence of ceramic surface treatmentson the tensile load strength of composite resin to all-ceramic coping materials.J Prosthet Dent 2005;4:357–62.

[8] Chevalier J. What future for zirconia as a biomaterial? Biomaterials2006;27:535–43.

[9] Gu YW, Khor KA, Pan D, Cheang P. Activity of plasma sprayed yttria stabilizedzirconia reinforced hydroxyapatite/Ti–6Al–4V composite coatings in simulatedbody fluid. Biomaterials 2004;25:3177–85.

[10] Chiu CY, Hsu HC, Tuan WH. Effect of zirconia addition on the microstructuralevolution of porous hydroapatite. Ceram Int 2007;33:715–8.

[11] Chou BY, Chang E. Plasma-sprayed hydroxyapatite coating on titanium alloywith ZrO2 second phase and ZrO2 intermediate layer. Surf Coat Technol2002;153:84–92.

[12] Marti A. Inert bioceramics (Al2O3, ZrO2) for medical application. Injury Int JCare Injured 2000;31:S-D33–6.

[13] Liu XY, Huang AP, Ding CX, Chu PK. Bioactivity and cytocompatibility ofzirconia (ZrO2) films fabricated by cathodic arc deposition. Biomaterials2006;27:3904–11.

[14] Uchida M, Kim HM, Kokubo T, Tanaka K, Nakamura T. Structural dependenceof apatite formation on zirconia gels in a simulated fluid. J Ceram Soc Jpn2002;110:710–5.

[15] Yan YY, Han Y. Structure and bioactivity of micro-arc oxidized zirconia films.Surf Coat Technol 2007;201:5692–5.

[16] Wang GC, Liu XY, Ding CX. Phase composition and in-vitro bioactivity ofplasma sprayed calcia stabilized zirconia coatings. Surf Coat Technol2008;202:5824–31.

[17] Kim DJ. Influence of aging environment on low temperature degradation oftetragonal zirconia alloys. J Eur Ceram Soc 1997;17:897–903.

[18] Castkova K, Hadraba H, Cihlar J. Hydrothermal ageing of tetragonal zirconiaceramics. Ceram-Silikaty 2004;48:85–92.

[19] Swab JJ. Low temperature degradation of Y-TZP materials. J Mater Sci1991;26:6706–14.

[20] Kenner GH, Pasco WD, Frakes JT, Brown SD. Mechanical properties of calciastabilized zirconia following in vivo and in vitro aging. J Biomed Mat Res Symp1975;6:63–6.

[21] Clarke IC, Manaka M, Green DD, William P, Pezzotti G, Kim YH, et al. Currentstatus of zirconia used in total hip implants. J Bone Joint Surg2003;85A(supplement 4):73–83.

[22] Dewhurst JK, Lowther JE. Relative stability structure and elastic properties ofseveral phase of pure zirconia. Phys Rev B 1998;57(2):745–7.

[23] Cutler RA, Reynolds JR, Jones A. Sintering and characterization ofpolycrystalline monoclinic tetragonal and cubic zirconia. J Am Ceram Soc1992;75(8):2173–83.

[24] Appiano S, Vincenzini P. Durability of zirconia thermal barrier coatings. J MaterSynth Process 1993;1:17–24.

[25] Brandon JR, Taylor R. Thermal properties of ceria and yttria partially stabilizedzirconia thermal barrier coatings. Surf Coat Technol 1989;39(/40):143–51.

[26] Raghavan S, Wang H, Porter WD, Dinwiddie RB, Mayo MJ. Thermal propertiesof zirconia Co-doped with trivalent and pentavalent oxides. Acta Mater2001;49(1):169–79.

[27] Liu XY, Ding CX. Characterization of plasma sprayed wollastonite powder andcoatings. Surf Coat Technol 2002;153:173–7.

[28] Kokubo T, Takadama H. How useful is SBF in predicting in vivo bonebioactivity. Biomaterials 2006;27:2907–15.

[29] Khor KA, Cheang P, Wang Y. Plasma spraying of combustion flame spheroidizedhydroxyapatite (HA) powders. J Therm Spray Technol 1998;7:254–60.


[30] Tsui YC, Doyle C, Clyne TW. Plasma sprayed hydroxyapatite coatings ontitanium substrates. Part 1: Mechanical properties and residual stress levels.Biomaterials 1998;19:2015–29.

[31] Gu YW, Khor KA, Cheang P. In vitro studies of plasma-sprayed hydroxyapatite/Ti–6Al–4V composite coatings in simulated body fluid (SBF). Biomaterials2003;24:1603–11.

[32] Liu XY, Zhao XB, Fu RKY, Ho JPY, Ding CX, Chu PK. Plasma-treatednanostructured TiO2 surface supporting biomimetic growth of apatite.Biomaterials 2005;26:6143–50.

[33] Phillippi CM, Mazdiyasni KS. Infrared and Raman spectra of zirconiapolymorphs. J Am Ceram Soc 1971;54:254–8.

[34] Kim YM. Electron microscopy study of the microstructures in the oxygen-deficient pure zirconia and yttrium doped zirconia. PhD Thesis Texas Univ,Austin, TX (USA).

[35] Wang GC, Liu XY, Gao JH, Ding CX. In vitro bioactivity and phase stability ofplasma-sprayed nanostructured 3Y-TZP coatings. Acta Biomaterialia2009;5:2270–8.

[36] Kweh SWK, Khor KA, Cheang P. Plasma-sprayed hydroxyapatite (HA) coatingswith flame-spheroidized feedstock: microstructure and mechanicalproperties. Biomaterials 2000;21:1223–34.

[37] Leigh SH, Lin CK, Berndt CC. Elastic response of thermal spray deposits underindentation tests. J Am Ceram Soc 1997;80(8):2093–9.

[38] Lawn BR, Howes VR. Elastic recovery at hardness indentations. J Mater Sci1981;16:2745–52.

[39] Korhonen ST, Calatayud M, Krause AOI. Stability of hydroxylated (�111) and(�101) surfaces of monoclinic Zirconia: a combined study by DFT and infraredspectroscopy. J Phys Chem C 2008;112:6469–76.

[40] Jung KT, Bell AT. The effects of synthesis and pretreatment conditions on thebulk structure and surface properties of zirconia. J Mol Catal A Chem2000;163:27–42.

[41] Christensen A, Carter EA. First-principles study of the surfaces of zirconia. PhysRev B 1998;58:8050–64.

[42] Iskandarova IM, Knizhnik AA, Rykova EA, Bagatur’yants AA, Potapkin BV,Korkin AA. First-principle investigation of the hydroxylation of zirconia andhafnia surfaces. Microelectron Eng 2003;69:587–93.

[43] Ushakov SV, Navrotsky A. Direct measurements of water adsorption enthalpyon hafnia and zirconia, surface enthalpy. Appl Phys Lett 2005;87:164103.

[44] Radha AV, Bomati-Miguel O, Ushakov SV, Navrotsky A, Tartaj P. Surfaceenthalpy, enthalpy of water adsorption, and phase stability in nanocrystallinemonoclinic zirconia. J Am Ceram Soc 2009;92:133–40.

[45] Petterson A, Marino G, Pursiheimo A, Rosenholm JB. Electrosteric stabilizationof Al2O3, ZrO2, and 3Y–ZrO2 suspensions: effect of dissociation and type ofpolyelectrolyte. J Colloid Interf Sci 2000;228:73–81.

[46] Svetina M, Ciacchi LC, Sbaizero O, Meriani S, De Vita A. Deposition of calciumions on rutile (110): a first principles investigation. Acta Mater2001;49:2169–77.

[47] Viitala R, Jokinen M, Peltola T, Gunnelius K, Rosenholm JB. Surface propertiesof in vitro bioactive and non-bioactive sol–gel derived materials. Biomaterials2002;23:3073–86.

[48] Vayssieres L, Chaneac C, Trone E, Joliver JP. Size tailoring of magnetite particlesformed by aqueous precipitation: an example of thermodynamic stability ofnanometric oxide particles. J Colloid Interf Sci 1998:205–12.

[49] Anselme K, Linez P, Bigerelle M, Maguer L, Hardouin P, Hildebrand HF, et al.The relative influence of the topography and chemistry of TiAi6V4 surface onosteoblastic cell behavior. Biomaterials 2000;21:1567–77.

[50] Anselme K, Bigerelle M, Noel B, Dufresne E, Judas D, Iost A, et al. Qualitativeand quantitative study of human osteoblast adhesion on materials withvarious surface roughnesses. J Biomed Mater Res 2000;49:155–66.

[51] Ramaswamy Y, Wu CT, Hummel AV, Combes V, Grau G, Zreiqat H. Theresponse of osteoblast, osteoclast and endothelial cells to zirconiummodified calcium-silicate-based ceramic. Biomaterials 2008;29:4392–402.

[52] Liu XM, Lim JY, Donahue HJ, Dhurjati R, Mastro AM, Vogler EA. Influence ofsubstratum surface chemistry/energy and topography on the human fetalosteoblastic cell line hFOB 1.19: phenotypic and genotypic responses observedin vitro. Biomaterials 2007;28:4535–50.

[53] Anselme K. Review: osteoblast adhesion on biomaterials. Biomaterials2000;21:667–81.

[54] Wan Q, Wang Y, Liu ZM, Qu X, Han BX, Bei JZ, et al. Adhesion and proliferationof OCT-1 osteoblast-like cells on micro- and nano-scale topography structuredpoly(L-lactide). Biomaterials 2005;26:4453–9.

[55] Webster TJ, Siegel RW, Bizios R. Osteoblast adhesion on nanophase ceramics.Biomaterials 1999;20:1221–7.

Microstructure, bioactivity and osteoblast behavior of monoclinic zirconia coating with nanostructured surfaceIntroductionMaterials and methodsFabrication and characterization of coatingsBioactivity evaluationCytocompatibility evaluationCell cultureMorphology observationCell proliferation and vitality

Results and discussionMicrostructure and phase compositionMechanical properties and hydrothermal stabilityBioactivity evaluationCytocompatibility evaluation

ConclusionsAcknowledgementsFigures with essential colour discriminationReferences

Documents

Microstructure, bioactivity and osteoblast behavior of ... · Raman spectroscopy revealed that the zirconia coating was composed of monoclinic zirconia which was stable at low temperature,