A

iw(wpsb©

K

1

wecipwso

tT

0d

Wear 262 (2007) 1337–1345

Tribological behaviour of Cl-implanted TiN coatingsfor biomedical applications

M.P. Gispert a, A.P. Serro a,b,∗, R. Colaco c, A.M. Botelho do Rego d,E. Alves e, R.C. da Silva e, P. Brogueira f, E. Pires g, B. Saramago a

a Centro de Quımica Estrutural, Complexo I, Instituto Superior Tecnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugalb Instituto Superior de Ciencias da Saude Egas Moniz, Quinta da Granja, Monte da Caparica, 2829-511 Caparica, Portugal

c Departamento de Engenharia de Materiais, Instituto Superior Tecnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugald Centro de Quımica-Fısica Molecular, Complexo I, Instituto Superior Tecnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal

e Instituto Tecnologico Nuclear, Estrada Nac. 10, 2686-953 Sacavem, Portugalf Departamento de Fısica, Instituto Superior Tecnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal

g Ceramed, Estrada do Paco do Lumiar, n◦22 Edifıcio Q, Campus do INETI, 1649-038 Lisboa, Portugal

Received 1 February 2006; received in revised form 15 November 2006; accepted 8 January 2007Available online 15 February 2007

bstract

The tribological behaviour of the prosthetic pair TiN coated stainless steel/ultra high molecular weight polyethylene (UHMWPE) may bemproved by chlorine-implantation of the TiN surface. Friction and wear were determined using a pin-on-disk apparatus and the wear mechanismsere investigated through scanning electron microscopy (SEM) and atomic force microscopy (AFM). Rutherford backscattering spectrometry

RBS) was used to determine the chlorine distribution profiles in the chlorine-implanted TiN coatings before and after the tribological experiments,hile X-ray photoelectron spectroscopy (XPS) was used to characterize chemically the same samples. Chlorine-implantation led to a significant

olymeric wear reduction when the lubricant was Hanks’ balanced salt solution (HBSS). If bovine serum albumin (BSA) was added to HBSS, atrong decrease of both friction and polymeric wear was observed for implanted and non-implanted TiN coatings. The former case was explainedy the formation of a titanium oxide layer on the TiN surface, while the latter derived from albumin adsorption.

2007 Elsevier B.V. All rights reserved.

eywords: Chlorine ion-implantation; TiN coating; UHMWPE; Friction; Wear; Albumin

to

Usoai

. Introduction

Osteoarthritis and rheumatoid arthritis are rheumatic diseaseshich often require substitution of natural joints by prostheses,

specially in what concerns hip and knee. The most commonause of failure and lack of durability of total hip prosthesess related with the generation of ultra high molecular weightolyethylene (UHMWPE) wear debris from the acetabular part,

hen sliding against the ceramic or the metallic ball which sub-

titutes the femoral head. The debris can lead to the inflammationf surrounding tissues and give rise to a process of bone resorp-

∗ Corresponding author at: Centro de Quımica Estrutural, Complexo I, Insti-uto Superior Tecnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal.el.: +351 21 841 9223; fax: +351 21 846 4455.

E-mail address: anapaula.serro@ist.utl.pt (A.P. Serro).

tmp

cmhr

043-1648/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.wear.2007.01.017

ion (osteolyse), which often results in the loosening of fixationf the acetabular and/or femoral parts.

In order to improve the tribological performance of theHMWPE components, several solutions have been proposed,

uch as the research of new formulations and processing meth-ds for the polymer [1–3], changes in sterilization methods tovoid oxidative degradation [4,5] or surface treatments and coat-ngs both on the polymeric and metallic parts [6–11]. Most ofhe studies focused on the effect of hard coatings or surface treat-

ents as ion implantation found a significant reduction of theolymer wear and metallic surface abrasion [8,9,12,13].

Titanium nitride (TiN) is one of the most studied ceramic

oatings due to its known biocompatibility [14,15]. Thisaterial leads to a significant increase in the metallic surface

ardness, helps in the protection against corrosion [16,17] andeduces the bacterial colonization [18]. It is also responsible for

1 ear 26

afl

ffpdTobiiafm

igptabapdn(b(ib(Ctcttobp

na

2

2

mtscPstt

dtwaa

atidwtp

dfvmi

tw

wr

2

2

pApvTcrs

2

bpatae

2

pt

338 M.P. Gispert et al. / W

significant decrease of the metal ion release to the biologicaluids [19].

Ion implantation is used to modify a large variety of sur-ace properties, such as the surface hardness, the resistance toriction, wear, fatigue, corrosion and oxidation. The mechanicalroperties of the underlying material are preserved, since theepth of penetration of the ions is generally lower than 1 �m.here are several studies about the effect of the implantationf hydrogen, helium, nitrogen, argon, oxygen or carbon ions iniomaterials used in joint prostheses [9–11]. The role of chlorineon in this type of applications was never investigated. However,t is known that when implanted on TiN coatings in cutting toolsnd forming dies, this ion gives rise to a significant decrease ofriction and wear in dry conditions, which was attributed to aechanism of self-lubrication [20,21].The objective of the present work was to assess if chlorine-

mplanted TiN coating on the metallic components is aood solution to improve the tribological performance ofrosthetic pairs. We investigate the tribological behaviour ofhese coatings deposited on stainless steel substrates rubbinggainst UHMWPE in lubricated conditions. As lubricants, theiological model fluid Hanks’ balanced salt solution (HBSS)nd solutions of albumin, the major protein present in theeriprosthetic fluid, in HBSS were used. Friction and wear wereetermined using a pin-on-disk apparatus and the wear mecha-isms were investigated through scanning electron microscopySEM) and atomic force microscopy (AFM) observations ofoth disks and pins. Rutherford backscattering spectrometryRBS) was used to determine the chlorine distribution profilesn the chlorine-implanted TiN coatings before and after the tri-ological experiments while X-ray photoelectron spectroscopyXPS) was used to characterize chemically the same samples.omparison studies were made with Ar-implanted TiN coatings

o elucidate the specificity of the chlorine ion. Argon ions werehosen due to the similarity in size with the chlorine ions. Sincehe surface wettability of the prosthetic materials is knowno affect its tribological behaviour [22], the surface energyf the substrates was determined as well as the interfacialehaviour of the various combinations lubricant/prostheticair.

We should point out that the citotoxicity studies which areecessary to qualify Cl-implanted TiN coating as a biomaterial,re presently under way.

. Experimental

.1. Materials

Disks of UHMWPE (CHIRULEN®, Poly Hi Solidur, Ger-any) with 7 cm of diameter were cut from sheet with 1 mm of

hickness. The counterfaces were pins of AISI 316L austenitic,tainless steel with 1 cm of diameter coated with TiN. The TiNoating was deposited by PVD—arc evaporation with a META-

LAS coating machine, at 300 ◦C, using a Ti cathode. Theamples were etched with titanium ions at 900 V bias voltageo remove any possible oxide on the surface. In the deposi-ion process the following parameters were used: 200 A cathode

Dimu

2 (2007) 1337–1345

eposition current; 100 V bias voltage and 1.4 × 10−2 mbar forhe nitrogen pressure. The thickness of TiN coating, measuredith a profilometer, varied between 1.2 and 1.6 �m and the aver-

ge roughness (Ra), measured by AFM for an analysed surfacerea of 225 �m2, did not exceed 30 nm.

TiN coated substrates were implanted with chlorine ions, withnominal fluence of 1 × 1017 cm−2 and energy 150 keV, at room

emperature. For comparison purposes, some substrates weremplanted with argon ions in the same conditions. In these con-itions the total energy density deposited in the TiN coatingas 2.4 kJ/cm2 at a rate of 0.43 W/cm2. The base pressure in

he implantation chamber was in the range of 10−5 Pa, and theressure during implantation was in the range of 10−3 Pa.

All substrates were cleaned in an ultrasonic bath with Extran®

iluted solution for 10 min and with distilled and deionised wateror 2 × 10 min and dried overnight at room temperature in aacuum oven. To test the possibility of water absorption by poly-eric substrates, a few were immersed in water for 2 h, but no

ncrease in weight was detected.Hank’s balanced salt solution (Sigma, ref. H8264) and solu-

ions of bovine serum albumin (BSA; Serva ref.11930) in HBSSith a concentration of 4 mg/mL were used as lubricants.The standard liquids used for measuring static contact angles

ere water (distilled and deionized) and diiodomethane (Merckef. 6053, doubly distilled under vacuum).

.2. Methods

.2.1. Tribological testsThe tribological experiments were carried out at room tem-

erature on a rotating pin-on-disk tribometer Wazau TRM1000.normal load of 67.5 N was applied which corresponds to a

ressure in the contact zone of 0.88 MPa. The tangential slidingelocity was 46 mm/s and the sliding distances were 1000 m.he wear rate of UHMWPE, whose density is 935 kg m−3, wasalculated following a weight loss measurement technique. Theesults are the average of, at least, three experiments for eachystem.

.2.2. RBSChlorine distribution profiles were measured by Rutherford

ackscattering spectrometry (RBS). All the RBS analyses wereerformed using 2 MeV He+ beams from a 2.5 MV Van de Graaffccelerator. The backscattered ions were detected by means ofwo silicon surface barrier detectors, placed at angles of 140◦nd close to 180◦ to the incoming beam direction, and havingnergy resolutions of 13 and 18 keV, respectively.

.2.3. SEM and AFMAnalysis of the surfaces by scanning electron microscopy was

erformed on Hitachi S2400 equipment. For SEM observations,he polymeric surfaces were previously covered with gold.

A Dimension D3100 with a Nanoscope IIIa controller from

igital Instruments was used for AFM measurements. The imag-

ng was performed in TappingMode using commercial tappingode etched silicon probes. The measurements were performed

nder ambient conditions.

ear 262 (2007) 1337–1345 1339

2

opr1aoasnaSnF2

2

awbtsf

sdt1a(ua

ssct

3

3

radaaFw1flss

FT

ve

crc

ttcsiws

3

Tthe theoretical RBS spectrum obtained with the simulation codeRUMP [28]. The best model derives from the description of theTiN layer with a Gaussian distribution of Cl ions according toFig. 3(b). The resulting chlorine concentration profile is shown in

M.P. Gispert et al. / W

.2.4. XPSThe XPS spectrometer used was a XSAM800 (KRATOS)

perated in the fixed analyser transmission (FAT) mode, with aass energy of 20 eV and the non-monochromatised Mg K� X-adiation (hν = 1253.7 eV). A current of 10 mA and a voltage of3 kV were used. Samples were analysed in ultra high vacuum,nd typical base pressure in the sample chamber was in the rangef 10−7 Pa. All sample transfers were made in air. Samples werenalysed at room temperature, at a take-off angle relative to theurface of 90◦. Spectra were collected and stored in 300 chan-els with a step of 0.1 eV and 60 s of acquisition by sweep, usingSun SPARC Station 4 with Vision software (KRATOS). A

hirley background was subtracted and curve fitting for compo-ent peaks was carried out using Gaussian–Lorentzian profiles.or quantification purposes, sensitivity factors were 1.80 for Tip, 0.66 for O 1s, 0.25 for C 1s, 1.58 for Ca 2p and 0.39 for P 2p.

.2.5. Wettability measurementsThe surface free energy of the counterface substrates γSV

s well as their dispersive, γdSV, and polar, γ

pSV, components

ere determined using the method of Owens and Wendt [23]ased on the experimental values of the static contact angles ofwo testing liquids: water and diiodomethane. The values of theurface tension components of the liquids, which are necessaryor the calculations, were taken from an earlier publication [24].

The static contact angles were measured through the ses-ile drop method using a micrometric syringe to generate therops inside a thermostatized chamber (25 ◦C) saturated withhe corresponding vapour. The images were recorded during800 s using an optical equipment described elsewhere [25], andnalysed using the Axisymmetric Drop Shape Analysis-ProfileADSA-P) program [26,27]. The values of the contact anglessed for the surface free energy calculations correspond to theverage of at least 10 experiments.

The surface tension of the biological model fluids was mea-ured as a function of time by the pendant drop method, using theame apparatus and software as well as the same experimentalonditions. The equilibrium values for the surface tension wereaken when no further time variation was detected.

. Results

.1. Tribological tests

The dynamical friction coefficients measured in the tests car-ied out using HBSS and HBSS + BSA are shown in Fig. 1. In thebsence of protein, the evolution of the friction coefficient wasifferent for the non-implanted and the Cl-implanted substrates,lthough the friction coefficient presented similar starting valuesnd reached identical values after a sliding distance of 1000 m.or the non-implanted surface, there was a running-in periodhere the friction coefficient increased rapidly during the first00 m and then tended asymptotically to a constant value, while

or the Cl-implanted surface, the friction coefficient increasedinearly with the sliding distance. Tests with the Ar-implantedurface led to the same type of behaviour of the non-implantedubstrates but the friction coefficient rose to significantly higher

FH

ig. 1. Friction coefficient vs. distance for the tribological pairs TiN/UHMWPE,iN(Cl)/UHMWPE and TiN(Ar)/UHMWPE in HBSS and HBSS + BSA.

alues than those obtained with the Cl-implanted substrates orven the non-implanted TiN.

The presence of albumin in the lubricant kept the frictionoefficient at a low, almost constant, value. Nevertheless, theole of the chlorine ion in the decrease of the friction coefficientould still be observed.

Fig. 2 shows the wear rate of the UHMWPE substrates afterhe tribological tests. No wear could be measured in the coun-erfaces. Although some dispersion of the results exist, it can belearly observed that, in the absence of protein, Cl-implantedubstrates led to the lowest wear rate of UHMWPE, while thosemplanted with Ar led to the highest wear rate. The polymericear against all counterfaces in the presence of protein was much

maller than the corresponding wear in HBSS.

.2. RBS

In Fig. 3(a), one experimental RBS spectrum of Cl-implantediN in the as-implanted condition is shown, superimposed to

ig. 2. Wear rate of UHMWPE rubbed against TiN, TiN(Cl) and TiN(Ar) inBSS and HBSS + BSA after 1000 m.

1340 M.P. Gispert et al. / Wear 26

Fig. 3. (a) RBS spectrum of TiN(Cl) before friction tests. RUMP simulation forthe model that best describes the experimental data can be seen superimposed(see text). The arrows mark the surface positions of the elements indicated, andthe steep ascent near channel 210 is due to the Fe of the substrate; (b) chlo-rine depth distribution as derived from the simulation in (a). The experimentaldistribution is very close to the theoretical prediction of the MC code SRIM-2s

FdT7o

Fsto

3

smaaiib

list(sHiwofo

3

itvdtwphosphate [31]. Comparison of the composition of TiN coat-

003, shown superimposed. The small shift observed is accounted for by someputtering during Cl-ion implantation.

ig. 3(b), and compares well with the near-Gaussian distributionerived theoretically with the SRIM-2003 simulation code [29].

he results allow to estimate that the thickness of the TiN film is10 nm and that the chlorine concentration reaches a maximumf 14 at.% at ∼90 nm depth, while at the surface it is 0.2 at.%.

icr

Fig. 4. SEM images of TiN (a)

2 (2007) 1337–1345

riction tests with HBSS and HBSS + BSA did not introduce anyignificant change in the chlorine distributions which indicateshat no detectable mass loss of the Cl-implanted TiN samplesccurred during the tribological tests.

.3. SEM and AFM

SEM (Fig. 4) and AFM (Fig. 5) analysis of TiN coatedpecimens before and after Cl-implantation do not indicate anyodification of the surface morphology and roughness associ-

ted with the implantation process. Both surfaces present poresnd small droplets which are typical of TiN coatings and weredentified by other authors as being solid TiN [30] generated dur-ng the PVD process. After the tribological tests, the surfaces ofoth coatings did not show any significant difference.

Topographical changes of the UHMWPE surface after tribo-ogical experiments depend exclusively on the lubricant. SEMmages show that the typical grooved pattern of the polymericurfaces caused by the machining process disappears after theribological tests against TiN or Cl-implanted TiN in HBSSFig. 6). Instead, parallel scratches typical of two body abra-ive wear are observed in the worn region. After friction inBSS + BSA, vestiges of the original grooves are still visible

n the wear track of the UHMWPE surfaces (Fig. 7), since theear rate is much smaller in this case. In the presence of protein,nly the vestiges of delamination are present in the worn sur-aces and no abrasion scratches could be observed both in TiNr Cl-implanted TiN samples.

.4. XPS

Table 1 presents the elemental composition of TiN coat-ng and Cl-implanted TiN coating before and after the frictionests in HBSS. The samples analysed before friction were pre-iously incubated in HBSS. Calcium and phosphorus wereetected on all samples. The binding energy of Ca 2p3/2 pho-oelectrons was 347.9 ± 0.2 eV and of P 2p3/2 photoelectronsas 133.7 ± 0.2 eV which indicates the precipitation of calcium

ngs before and after chlorine implantation shows that nitrogenoncentration decreases, while oxygen increases and titaniumemains almost constant. Together with the XPS spectra for Ti

and TiN(Cl) (b) surfaces.

M.P. Gispert et al. / Wear 262 (2007) 1337–1345 1341

Fig. 5. AFM images of TiN (a) and TiN(Cl) (b) surfaces.

F t TiNp .

2tmT

cC

asi

FP

ig. 6. SEM images of UHMWPE surface after tribological experiments againsarallel grooves (indicated by the arrow) can be seen only beside the wear track

p of both TiN coatings shown in Fig. 8(a and b), this indicateshat, after implantation, part of TiN (455.4 eV) was oxidized

ainly to TiO2 (458.8 eV), although other intermediate oxidesinOx (x < 2) may be present.

After friction, Ti and N concentrations decreased while theoncentrations of O, C, P and Ca increased in both coatings.omparison of the XPS spectra for Ti 2p of both TiN (Fig. 8(a))

ara

ig. 7. SEM images of UHMWPE surface after tribological experiments against Tiarallel grooves indicated by the arrow remain inside the wear track while new delam

or TiN(Cl) in HBSS, being (b) the magnification of the wear track. The original

nd Cl-implanted TiN (Fig. 8(b)) shows that after friction thepectra kept the same pattern although the peaks became lessntense.

Fittings of O 1s peaks obtained before and after friction,

re shown in Fig. 9(a and b), for TiN and Cl-implanted TiN,espectively. A significant decrease in the metallic oxide peaknd a concomitant increase of the peaks attributed to phosphate

N or TiN(Cl) in HBSS + BSA, being (b) the magnification of the wear track.ination grooves appear.

1342 M.P. Gispert et al. / Wear 262 (2007) 1337–1345

Fw

awActedta

iCtm

r

TEt

TNOCPC

Fig. 9. Fitting of O 1s peaks obtained before and after friction in substrates TiN(a) and TiN(Cl) (b).

ig. 8. Comparison of the XPS spectra of Ti 2p obtained before and after frictionith substrates TiN (a) and TiN(Cl) (b).

nd organic oxygen occurred after friction, which is consistentith the increase in Ca and P concentrations also observed.lthough the presence of C is usually associated with organic

ontaminants, the increase in the C concentration detected forhe TiN surface may be attributed, at least partially, to the pres-nce of polyethylene traces resulting from polymeric transferuring friction. The fact that no polyethylene was identified inhe SEM images may signify that it probably exists in a vestigialmount.

The difference between the behaviour of TiN and Cl-mplanted TiN becomes evident when the ratios of O, P anda signals over the Ti signal are compared before and after fric-

ion. According to Fig. 10, these ratios increase with frictionore significantly in the absence of chlorine.It is interesting to refer that only a vestigial amount of chlo-

ine could be detected both, before and after friction, which

able 1lemental composition of TiN and Cl-implanted TiN before and after friction

ests in HBSS

TiN before TiN(Cl) before TiN after TiN(Cl) after

i 13.0 11.3 2.7 6.514.9 7.2 5.2 5.634.8 44.4 37.8 44.433.5 33.7 46.0 34.3

2.4 1.9 4.4 4.5a 1.4 1.5 3.9 4.8 Fig. 10. Ratios of the atomic concentrations of O, P and Ca relative to Ti found

on TiN and TiN(Cl) before and after the friction tests.

M.P. Gispert et al. / Wear 26

Table 2Surface free energy, �SV, and their dispersive and polar components, γd

SV andγ

pSV, respectively, for TiN and Cl-implanted TiN, before friction

�SV γdSV γ

pSV

TT

ac

3

caTct

Hspta

F(

4

tvrmiondt

liispfb

i

iN 47 ± 2 33 ± 1 14 ± 2iN(Cl) 43 ± 2 35 ± 1 8 ± 2

grees with the RBS estimation of 0.2 at.% for the chlorineoncentration at the surface.

.5. Wettability measurements

The surface free energy as well as their dispersive and polaromponents obtained from the measured values of the contactngles of water and diiodomethane, for TiN and Cl-implantediN, are given in Table 2. Ion implantation reduced the polaromponent of the surface free energy and, as a consequence, theotal surface free energy.

The time dependence of the contact angle of HBSS andBSS + BSA on TiN and on Cl-implanted TiN surfaces is

hown in Fig. 11(a and b), respectively. Chlorine-implanted TiN

resents lower contact angles with HBSS. For both surfaces,he addition of the protein increases the variation of the contactngles with time.

ig. 11. Time dependence of the contact angle of HBSS (a) and of HBSS+BSAb) on TiN and TiN(Cl).

wtptUctmptbtot

4

tmtmi

nta(o

iadne

2 (2007) 1337–1345 1343

. Discussion

A number of authors have reported an improvement of theribological properties of TiN films after implantation witharious ions [32–34]. However, the mechanisms which areesponsible for this improvement depend, not only on theethod of deposition of TiN, but also on the species and the

mplantation parameters. While argon and carbon implantationf PVD-type TiN coatings mainly result in increased hard-ess values, nitrogen implantation was found to increase orecrease the hardness of TiN depending on the implantationreatment.

Furthermore, extensive work on the improvement of tribo-ogical properties by chlorine implantation of TiN coatingsndicates that chemical, instead of physical, modifications responsible for the reduction of wear and friction inevere dry conditions [35]. A self-lubrication mechanism wasroposed based on the role of Cl-atoms favoring in situormation of oxygen-deficient oxides that have shear deforma-ility.

The tribological pairs focused on the reported studiesnvolved essentially two metallic surfaces where one or bothere covered by TiN films. A different problem is faced when

he TiN rubs against UHMWPE. This coating is known torotect the metallic heads against corrosion and simultaneouslyo reduce friction and wear when articulating against anHMWPE bearing surface. However, the efficiency of the TiN

oating depends on the finishing of the surface. As deposited,he TiN surfaces are slightly rough due to the formation of

elted TiN droplets (Figs. 4 and 5) and, only after a polishingrocess, an adequate roughness is achieved [30]. An alternativeo the polishing treatment that may damage the coating, maye provided by ion implantation. In this work, chlorine implan-ation of the TiN coating was found to decrease the wear ratef the polymer in lubricated conditions and an interpretation ofhis behaviour is attempted.

.1. Surface modification by ion-implantation

The results of surface analysis by SEM and AFM confirmhe findings of other authors [36] that Cl-implantation does notodify physically the TiN coating. Aizawa et al. [20] claimed

hat no structural change was seen after Cl-implantation whicheans that chlorine atoms must be present as interstitial atoms

nto the TiN lattice.After Cl-implantation, part of TiN was oxidized to tita-

ium oxide which should be responsible for the decrease in theotal surface energy. Simultaneously, a new O 1s peak appearsfter chlorine implantation at 534.1 eV which is typical of H2OFig. 9(b)) and may be responsible for the lower contact anglef HBSS on Cl-implanted TiN.

The development of a superficial oxide layer on the Cl-mplanted TiN is due to the known oxidizing capacity of chlorine

toms. The oxidation process of TiN to titanium oxides wasetermined to occur in several stages [37]. Depending on theitrogen and oxygen partial pressures and the temperature, sev-ral intermediate oxides may be formed from TiO to TiO2.

1 ear 26

Aafi

4

dTptvladctp

TtHsa

atssmdbtariaHe

oiOdfspdipTmlottf

4

tstit

metbeY

c

Attt

CcoflttaoHSr

5

stUCcrciootbr

344 M.P. Gispert et al. / W

ccording to the XPS analysis, the Cl-implanted TiN samplesre mostly covered by TiO2 which is in agreement with thendings of other authors [38].

.2. Friction and wear in HBSS

Friction and wear tests in HBSS revealed significantifferences between the non-implanted and the Cl-implantediN. Fig. 1 shows that for the Cl-implanted TiN the running-ineriod is much longer than for the non-implanted TiN andhe Ar-implanted TiN. The friction coefficient reached similaralues for the non-implanted and the chlorine-implanted TiN,ower than those obtained with the Ar-implanted TiN. Sincergon and chlorine atoms have approximately the same size, theifference in the friction coefficients suggests that the effect ofhlorine-implantation must elapse from chemical transforma-ions rather than from physical changes due to the implantationrocess.

SEM micrographs of worn surfaces of UHMWPE againstiN and Cl-implanted TiN are similar leading to the conclusion

hat the wear mechanisms are essentially abrasion in both cases.owever, measurements of the volume of material removed

how that the wear rate of the polymer decreased when slidinggainst Cl-implanted TiN.

Since the chlorine implantation did not affect the roughnessnd the morphology of the TiN coating, the improvement ofhe tribological performance of the Cl-implanted TiN coatinghould be attributed to the presence of the titanium oxideurface layer. The hardness of TiO2 is typically ∼1000 HV,uch smaller than that of TiN, ∼2500 HV. Although the

ifference between the hardness of the UHMWPE and that ofoth TiN and TiN(Cl) counterfaces is large, the fact is that, sincehe wear mechanism of the polymer is predominantly two-bodybrasion, a softer counterface will tend to result in a lower wearate of the polymer. The softening of the TiN through chlorinemplantation is opposite to the hardening effect conventionallyssociated to physical modification by ionic implantation.owever, other authors have already reported the same

ffect [35].XPS analysis of both coatings shows that the main effect

f friction on the chemical composition of the surface is thencrease of calcium phosphate deposition. The increase of the

1s peak characteristic of phosphate is associated with theecrease of Ti–O peak. This modification is more importantor the non-implanted TiN coating which leads to the conclu-ion that the presence of chlorine partially inhibits the calciumhosphate precipitation. Calcium phosphate deposition happensue to the supersaturation of HBSS in the pH range of 5–10n relation to several phosphate species, such as octacalciumhosphate, dicalcium phosphate dihydrate or hydroxyapatite.he kinetics of this precipitation during the friction experi-ents must depend on the flash temperature which should be

ower for the Cl-implanted TiN. The longer running-in period

bserved in Fig. 1 for these samples may, in turn, be relatedo the slower precipitation of calcium phosphate upon theitanium oxide layer which covers the Cl-implanted TiN sur-ace.

A

2

2 (2007) 1337–1345

.3. Role of protein in the lubricant

The decrease of the friction coefficient when BSA is addedo the lubricant is evident in Fig. 1 while the decrease in wear ishown in Fig. 2. Although the effect of the protein is so importanthat it attenuates the differences between non-implanted and ion-mplanted TiN coatings, the Cl-implanted TiN coating still leadso the lowest friction and wear.

The presence of albumin in the lubricant resulted in the for-ation of an adsorbed protein layer on all surfaces, which is

videnced by the significant decrease in the contact angle withime, observed in Fig. 11. The decrease of the contact angle maye correlated with the decrease in the solid/lubricant interfacialnergy (γSL) associated with the adsorption process, through theoung equation [39]:

os θ = γS − γSL

γL(1)

ssuming the surface tension of the solid, γS, as a constant,he decrease of γSL together with the decrease of the surfaceension of the lubricant containing protein (γL) with time leado the increase of cos θ.

Radiolabelling measurements of albumin adsorption on TiN,l-implanted TiN and UHMWPE during contact with protein-ontaining HBSS (unpublished results) confirmed the presencef an albumin layer on these surfaces. Several authors [25,40]ound that this adsorbed layer protects the surface of the tribo-ogical pairs leading to the reduction of the interaction betweenhe solid surfaces. This enhances the boundary component ofhe lubrication, as previously mentioned by Widmer et al. [40],nd, as a consequence, reduces friction and wear. The changef the wear mechanism from abrasive, when the lubricant isBSS, to adhesive when BSA is added to HBSS is evident fromEM images of the worn surfaces of UHMWPE (Figs. 6 and 7,espectively).

. Conclusions

The tribological behaviour of the prosthetic pair TiN coatedtainless steel/UHMWPE may be improved through implan-ation of chlorine atoms in the TiN coating. Wear of theHMWPE component was greatly reduced when sliding againstl-implanted TiN coated substrates, in HBSS, while the frictionoefficient increased more slowly with the sliding distance. Weareduction may be attributed to the substitution of the hard TiNounterface by a titanium oxide layer with lower hardness whichs a less wear-aggressive for the polymer surface. The behaviourf the friction coefficient was attributed to the slow precipitationf calcium phosphate. When albumin was added to HBSS, fric-ion and wear decreased significantly due to protein adsorptionut still, the Cl-implanted TiN coating led to the best tribologicalesults.

cknowledgements

This study was financially supported by the programme POCI010 under project POCI/SAU-BMA/55493/2004. M. Gispert

ear 26

wadtH

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[(1805) 65–87.

M.P. Gispert et al. / W

as awarded with a research grant of the same project. A.P. Serrocknowledges the Portuguese Foundation for Science for post-octoral grant SF RH/BPD/5666/2001. The authors are gratefulo Dra. Virginia Chu for the thickness measurements and to Polyi Solidur who kindly offered the UHMWPE.

eferences

[1] T.M. Wright, T. Fukubayashi, A.H. Burstein, The effect of carbon fiberreinforcement on contact area, contact pressure, and time-dependent defor-mation in polyethylene tibial components, J. Biomed. Mater. Res. 15 (1981)719–730.

[2] S. Li, E.G. Howard, Process of manufacturing ultra high molecular weightpolyethylene shaped particles, US Patent 5,037,928, 1991.

[3] O.K. Muratoglu, C.R. Bragdon, D.O. O’Connor, M. Jasty, W.H. Harris,A novel method of cross-linking ultra-high-molecular-weight polyethy-lene to improve wear, reduce oxidation, and retain mechanical properties:recipient of the 1999 HAP Paul Award, J. Arthopl. 16 (2) (2001) 149–160.

[4] H. McKellop, F.W. Shen, B. Lu, P. Campbell, R. Salovey, Effect of steril-ization method and other modifications on the wear resistance of acetabularcups made of ultra-high molecular weight polyethylene—a hip-simulatorstudy, J. Bone Joint Surg. Am. 82A (12) (2000) 1708–1725.

[5] L.S. Bargmann, B.C. Bargmann, J.P. Collier, B.H. Currier, M.B. Mayor,I. Current status of the bearing surface debate: current sterilization andpackaging methods for polyethylene, Clin. Orthop. Relat. Res. 369 (1999)49–58.

[6] F. Alonso, J.J. Ugarte, D. Sansom, J.L. Viviente, J.I. Onate, Effects of ionimplantation on Ti-6Al-4V on its frictional behaviour against UHMWPE,Surf. Coat. Technol. 83 (1996) 301–306.

[7] J.I. Onate, M. Comin, I. Braceras, A. Garcia, J.L. Viviente, M. Brizuela,N. Garagorri, J.L. Peris, J.I. Alava, Wear reduction effect on ultra-high-molecular-weight polyethylene by application of hard coatings andion implantation on cobalt chromium alloy, as measured in a kneewear simulation machine, Surf. Coat. Technol. 142–144 (2001) 1056–1062.

[8] E.Y. Gutmanas, I. Gotman, PIRAC Ti nitride coated Ti–6Al–4V headagainst UHMWPE acetabular cup–hip wear simulator study, J. Mater.Sci.-Mater. Med. 15 (4) (2004) 327–330.

[9] D.S. Xiong, Z.M. Jin, Tribological properties of ion implanted UHMWPEagainst Si3N4 under different lubrication conditions, Surf. Coat. Technol.182 (2–3) (2004) 149–155.

10] L. Torrisi, A.M. Visco, A. Valenza, Ion implantation on ultra high molecularweight polyethylene (UHMWPE) for medical prosthesis, Rad. Ef. Def.Solid. 158 (9) (2003) 621–623.

11] A. Pichat, L.M. Rabbe, J. Rieu, A. Rambert, C. Chabrol, M. Robelet,Effect of ion implantation on titanium alloy/polyethylene and 316L stain-less steel/polyethylene friction couples running in joint prostheses, Surf.Coat. Technol. 45 (1–3) (1991) 15–22.

12] L.P. Ward, C. Subramanian, K.N. Strafford, T.P. Wilks, Sliding wear stud-ies of selected nitride coatings and their potential for long-term use inorthopaedic applications, in: Proceedings of the Institution of MechanicalEngineers Part H, J. Eng. Med. 212 (H4) (1998) 303–315.

13] Y. Tamura, A. Yokoyama, F. Watari, T. Kawasaki, Surface properties andbiocompatibility of nitrided titanium for abrasion resistant implant materi-als, Dent. Mater. J. 21 (4) (2002) 355–372.

14] J. Narayan, W.D. Fan, R.J. Narayan, P. Tiwari, H.H. Stadelmaier, Diamond,diamond-like and titanium nitride biocompatible coatings for human-bodyparts, Mater. Sci. Eng. B 25 (1994) 5–10.

15] J. Vanraay, P.M. Rozing, C.A. Vanblitterswijk, R.M. Vanhaastert, H.K.Koerten, Biocompatibility of wear-resistant coatings in orthopedic-surgery

in vitro testing with human fibroblast cell-cultures, J. Mater. Sci.-Mater.Med. 6 (1995) 80–84.

16] R. Hubler, A. Cozza, T.L. Marcondes, R.B. Souza, F.F. Fiori, Wear andcorrosion protection of 316-L femoral implants by deposition of thin films,Surf. Coat. Technol. 142 (2001) 1078–1083.

[

2 (2007) 1337–1345 1345

17] M.T. Raimondi, R. Pietrabissa, The in vivo wear performance of pros-thetic femoral heads with titanium nitride coating, Biomaterials 21 (2000)907–913.

18] B. Groessner-Schreiber, A. Neubert, W.D. Muller, M. Hopp, M. Griepen-trog, K.P. Lange, Fibroblast growth on surface-modified dental implants:an in vitro study, J. Biomed. Mater. Res., Part A 64A (2003) 591–599.

19] J. Fisher, X.Q. Hu, T.D. Stewart, S. Williams, J.L. Tipper, E. Ingham, M.H.Stone, C. Davies, P. Hatto, J. Bolton, M. Riley, C. Hardaker, G.H. Isaac, G.Berry, Wear of surface engineered metal-on-metal hip prostheses, J. Mater.Sci.-Mater. Med. 15 (3) (2004) 225–235.

20] T. Aizawa, T. Akhadejdamrong, A. Mitsuo, Self-lubrication of nitrideceramic coating by the chlorine ion implantation, Surf. Coat. Technol. 177(2004) 573–581.

21] T. Akhadejdamrong, T. Aizawa, M. Yoshitake, A. Mitsuo, T. Yamamoto, Y.Ikuhara, Self-lubrication mechanism of chlorine implanted TiN coatings,Wear 254 (7–8) (2003) 668–679.

22] A. Borruto, G. Crivellone, F. Marani, Influence of surface wettability onfrition and wear tests, Wear 222 (1998) 57–65.

23] D.K. Owens, R.D. Wendt, Estimation of the surface free energy of poly-mers, J. Appl. Polym. Sci. 13 (1969) 1741–1747.

24] A.P. Serro, A.C. Fernandes, B. Saramago, Dynamic interfacial behaviourof bovine serum albumin solutions on titanium surfaces, Colloids Surf. A:Physicochem. Eng. Asp. 125 (1997) 209–219.

25] M.P. Gispert, A.P. Serro, R. Colaco, B. Saramago, Friction and wear mech-anisms in hip prosthesis: comparison of joint materials behaviour in severallubricants, Wear 260 (1–2) (2006) 149–158.

26] P. Cheng, D. Li, L. Boruvka, Y. Rotenberg, A.W. Neumann, Automation ofAxisymetric Drop Shape Analysis for measurements of interfacial tensionsand contact angles, Colloids Surf. 43 (1990) 151–197.

27] P. Cheng, A.W. Neumann, Computational evaluation of axisymetric dropshape analysis-profile (ADSA-P), Colloids Surf. 62 (1992) 297–306.

28] L.R. Doolittle, A semiautomatic algorithm for Rutherford backscatteringanalysis, Nucl. Instrum. Methods B 15 (1986) 227–231.

29] J.F. Ziegler, SRIM-2003, Nucl. Instrum. Methods B 219–220 (2004)1027–1036.

30] B. Derbyshire, J. Fisher, D. Dowson, C.S. Hardaker, K. Brummit,Wear of UHMWPE sliding against untreated, titanium nitride-coatedand ‘Hardcor’-treated stainless steel counterfaces, Wear 181–183 (1995)258–262.

31] S. Hiromoto, K. Kano, Y. Suzuki, K. Asami, A. Chiba, T. Hanawa, Surfacecharacterization and anodic polarization of nitrogen-ion-implanted nickel-free Co–Cr–Mo alloy, Mater. Trans. 46 (2005) 1627–1632.

32] A.J. Perry, V. Valvoda, D. Rafaja, D.L. Williamsom, B.D. Sartwell, On theresidual-stress and picostrucutre of titanium nitride films. 1. Implantationwith argon or krypton, Surf. Coat. Technol. 54 (1–3) (1992) 180–185.

33] F. Seidel, H.R. Stock, P. Mayr, Carbon, nitrogen and oxygen implantationinto TiN coatings, Surf. Coat. Technol. 108–109 (1998) 271–275.

34] R.R. Manory, C.L. Li, C. Fountzoulas, J.D. Demaree, J.K. Hirvonen, R.Nowak, Effect of nitrogen ion-implantation on the tribological propertiesand hardness of TiN films, Mater. Sci. Eng. A 253 (1998) 319–327.

35] T. Aizawa, A. Mitsuo, T. Akhadejdamrong, Chemical modification oftitanium nitride films via ion implantation, Mater. Trans. 44 (10) (2003)1962–1970.

36] T. Aizawa, A. Mitsuo, S. Yamamoto, T. Sumitomo, S. Muraishi, Self-lubrication mechanism via the in situ formed lubricious oxide tribofilm,Wear 259 (2005) 708–718.

37] I.G. Polyakova, T. Hubert, Thermal stability of TiN thin films investigatedby DTG/DTA, Surf. Coat. Technol. 141 (2001) 55–61.

38] T. Akhajdenung, T. Aizawa, M. Yoshitake, A. Mitsuo, Feasibility studyof self-lubrication by chlorine implantation, Nucl. Instrum. Methods Phys.Res. 207 (2003) 45–54.

39] T. Young, An essay on the cohesion of fluids, Philos. Trans. R. Soc. 95

40] M.R. Widmer, M. Heuberger, J. Voros, N.D. Spencer, Influence of poly-mer surface chemistry on frictional properties under protein lubricationconditions: implications for hip-implant design, Tribol. Lett. 10 (2001)111–116.