orthopedic nano diamond coatings: control of surface properties and their impact on osteoblast...

Orthopedic nano diamond coatings: Control of surfaceproperties and their impact on osteoblast adhesion andproliferation

Lei Yang,1 Brian W. Sheldon,1 Thomas J. Webster1,21Division of Engineering, Brown University, Providence, Rhode Island 029122Department of Orthopedics, Brown University, Providence, Rhode Island 02912

Received 28 February 2008; revised 16 June 2008; accepted 11 July 2008Published online 4 November 2008 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.32227

Abstract: The superior mechanical and tribological prop-erties of diamond coatings suggest their promise forimproving current orthopedic implants. Therefore, under-standing and controlling biological responses on diamondcoatings are important and necessary for their advance-ment in orthopedics. For this reason, the objective ofthe present study was to correlate surface properties ofdiamond coatings with osteoblast (OB) adhesion and pro-liferation. Diamond coatings on silicon of variable surfacefeatures (specifically, grain size, surface roughness andsurface chemistry) were fabricated by microwave plasmaenhanced chemical-vapor-deposition (MPCVD). Scanningelectron microscopy (SEM) as well as atomic force micros-copy (AFM) was applied for topographical characterizationand contact angles were measured to assess surface wett-ability. Results revealed that the grain size, surface rough-ness and wettability of diamond coatings can be controlledby adjusting H2 plasma in the MPCVD process. Further,

results showed enhanced OB adhesion on nanocrystallinediamond (ND) with grain sizes less than 100 nm whereasnanostructured diamond/amorphous carbon coatings(NDp) and microcrystalline diamond (MD) inhibited OBadhesion. H2 plasma treated ND (NDH) also promoted OBadhesion. Similarly, OB proliferated to a greater extent onND and NDH compared with MD and uncoated siliconcontrols. In summary, surface properties (including topog-raphy and chemistry) of diamond coatings can be con-trolled to either promote or inhibit OB functions, whichimplies that various forms of diamond coatings can beused to either support or inhibit bone growth in differentregions of an orthopedic implant. � 2008 Wiley Periodicals,Inc. J BiomedMater Res 91A: 548–556, 2009

Key words: diamond; osteoblast adhesion; proliferation;topography; surface chemistry; nanocrystalline; nanotech-nology

INTRODUCTION

Synthetic materials (mainly alloys and ultra highmolecular weight polymers) currently used as ortho-pedic replacements function sufficiently for only10–15 years after implantation. There are numerousreasons for the failure of bone implants, includingmechanical failure of the materials (through suchmechanisms as fracture, fatigue, wear, etc.), corro-sion and poor cytocompatibility properties (leadingto insufficient tissue bonding, toxicity, etc.). Forinstance, intense wear of articulating surfaces andadverse tissue reactions caused by such wear debrisoften results in the failure or loosening of theimplant and requires reoperation.1,2 Diamond-coatedmetals may offer a practical solution to this problem

because of the low friction coefficient and ultra-lowwear rate of these materials.3,4 In addition, diamondcoatings exhibit many other superior propertiesattractive for orthopedic implants, such as highchemical resistance, high fracture toughness andhigh bonding strength to various substrates.5–9 Fur-thermore, the feasibility of modifying diamond coat-ings with different macromolecules possibly enablesbetter clinical performance for orthopedic prostheticapplications.10,11

However, attempts to coat current orthopedicimplants with diamond require an improved under-standing and control of interactions between dia-mond and osteoblasts (OBs, bone-forming cells).Although Tang et al.12 provided preliminary quanti-tative and morphological evidence of cell interac-tions on diamond coatings, the biological responsesof OBs toward diamond coatings are still not clearand have not been sufficiently studied to date. Arecent investigation provided evidence that nano-crystalline diamond (ND)/amorphous carbon com-posite coatings are nontoxic to OB-like cells,13 but all

Correspondence to: T. J. Webster; e-mail: [email protected] grant sponsor: Coulter Foundation

� 2008 Wiley Periodicals, Inc.

the possible effects of diamond surface propertieswere not considered. Garguilo et al.14 correlated thesurface properties of diamond coatings (such ashydrophobicity) to the adsorption of fibrinogen (animportant protein for inflammation and inflamma-tory cell adhesion), but they did not report on subse-quent, long-term cell behaviors. Also, these studiesdid not specify the difference in cell functions ondiamond coatings of different grain sizes, which canbe easily changed during coating processes.

Therefore, in the present work, OB functions (specif-ically, adhesion and proliferation) were investigatedon diamond coatings andwere correlated to the adjust-able surface properties of diamond coatings, providingan insight toward controlling OB behaviors by manip-ulating diamond surfaces. Various nano/micron-crys-talline diamond coatings on silicon substrates wereproduced in this study by microwave plasmaenhanced chemical-vapor-deposition (MPCVD). Thecontrol of surface properties of such coatings wasachieved by adjusting H2 plasma during the MPCVDprocess. Four types of diamond coatings were created:ND, nanostructured diamond/amorphous carbonwith platelet grains (NDp), ND with H2 plasma treat-ment (NDH) and microcrystalline diamond (MD). Theimpact of diamond surface properties including topog-raphy (grain size and associated surface roughness)and surface chemistry (wettability and hydrogenterminations) on OB behaviors was then discussed.

MATERIALS AND METHODS

Diamond coatings

Diamond coatings created in this study were producedby MPCVD. The details of MPCVD and the mechanismsof diamond film growth can be found elsewhere.15 Differ-ent diamond coatings were deposited on a commercial sili-con wafer (phosphorus doped, (100) face polished, VirginiaSemiconductor, PA) with 0.6% methane, 5–25% hydrogengas and argon gas at 8008C and a chamber pressure of 111torr for 2 h. The silicon substrates were first seeded withnano diamond powder (Nanometer diamond powderNSD-A, particle size distribution 1–50 nm, GB Group, NY)distributed in methanol before film growth. Samples fabri-cated with different H2 concentrations were divided intodifferent groups as ND and MD, respectively. A specialtype of coating was grown without direct exposure toplasma and was termed NDp. Some ND coatings wereexposed to pure H2 plasma at 15 torr for 30 min immediatelyafter the deposition, and they were termed NDH. These fourtypes of diamond coatings are summarized in Table I.

Surface characterization of diamond coatings

Surface morphology of the diamond coatings wasobserved by a scanning electron microscope (SEM, LEO

1530VP, field-emission gun, Zeiss) at an accelerating volt-age of 4 kV. Raman spectroscopy (Renishaw Ramascope,laser excitations at 632 and 488 nm) was applied to iden-tify the chemical composition of NDp. Topography of thediamond coatings was further investigated by atomic forcemicroscopy (AFM, AutoProbe, Park Scientific Instrument).Silicon tips and AFM contact mode were used at a scan-ning speed of 0.5 Hz. For morphological images, the scanlengths were 2 lm for ND and 5 lm for micron-crystallinediamond. For the analysis of topographical parametersincluding peak-valley roughness (Rpv), root mean squareroughness (Rrms) and local geometric area (Ageo), the samescan length (5 lm) was used for all the samples and atleast three regions on the same sample were analyzed.

Wettability measurements

Contact angles on diamond coatings were measuredto assess the wettability and energetic states of diamond.Before measurements, the surfaces of the diamondcoatings were cleaned by soaking and then sonicating thesamples in a series of solutions including acetone, 100%ethanol, 70% ethanol, and deionized water. Acetone andethanol used in this procedure are widely applied for bio-material surface cleaning purposes,16 as such cleaningremoves the residual contaminants attached to the materialsurface by physico-absorption but does not change thechemical bonding state of surface. To identify the influenceof residual contaminants on wettability, several samplesbefore surface cleaning and after cleaning were measured.

Contact angles on diamond surfaces were measuredthrough the sessile drop shape method under ambient con-ditions. A deionized water drop (3 lL) was placed on thediamond surface and its profile was recorded immediatelyby the drop shape analyzer (Easy Drop FM40, Kruss, Ger-many). Contact angles were then calculated by computersoftware (DSA1, Kruss, Germany) attached to the analyzer.At least five different spots on each sample were selectedfor measurements.

TABLE IFilm Growth Conditions for Diamond Coatings

Name

H2 Concentration(% Flow Rate)

for Film Growth

H2 PlasmaTreatment afterFilm Growth

NDp 20 –ND5 5 –ND10 10 –MD15 15 –MD20 20 –MD25 25 –NDH5 5 30 minNDH10 10 30 min

ND, nanocrystalline diamond; MD, microcrystalline dia-mond; NDp, grown without direct exposure to plasma;NDH, H2 plasma-treated coatings. The number indicatesthe percentage of H2 used for deposition.

ORTHOPEDIC NANO DIAMOND COATINGS 549

Journal of Biomedical Materials Research Part A

OB adhesion assay

Human OB (CRL-11372, ATCC) of population numbersbetween 10 and 15 were cultured in Dulbecco’s ModifiedEagle’s Medium (DMEM) supplemented with 10% fetalbovine serum (FBS, Cyclone, USA) and 1% penicillin/streptomycin (Cyclone, USA) until they reached 80–100%confluence, at which point they were lifted from tissue cul-ture polystyrene with trypsin-EDTA (Cyclone, USA). Thediamond coatings were cleaned as described above andwere sterilized by UV light exposure overnight before cellculture. Silicon wafers for diamond film growth were usedas controls.

For adhesion assays, OBs were seeded at a density of3500 cells/cm2 in DMEM supplemented with 10% FBS and1% penicillin/streptomycin. After seeding on the diamondcoatings, cells were cultured under standard cell cultureconditions (5% CO2, 95% humidified air, 378C) for 4 h. Atthat time, non-adherent cells were removed by rinsing inphosphate buffer saline (PBS) whereas adherent cells werefixed with 10% formalin (Fisher Scientific, USA) andstained by 40,6-diamidino-2-phenylindole (DAPI, SigmaAldrich, USA). Adherent cells were counted under a fluo-rescence microscope (FM, Axiovert 200M, Zeiss) usingstandard protocols.

OB proliferation assay

For proliferation assays, OBs were seeded at 3000 cells/cm2 in DMEM supplemented with 10% FBS and 1% peni-cillin/streptomycin under standard cell culture conditionsfor 1, 3, and 5 days. After these time periods, cells werefixed, stained and counted in a similar way as describedabove. Cell culture media was changed every other day.Silicon wafer and microscope glass cover slips (CG) etchedwith 0.1M NaOH were used as controls.

OB morphology by SEM

For SEM observation of OB morphology on different di-amond coatings, OBs on ND5 and MD20 after 48 h ofincubation were rinsed with 0.1M sodium cacodylatebuffer (pH 7.2; Fisher Scientific, USA), fixed with 2% glu-taraldehyde in 2% paraformaldehyde (Fisher Scientific,USA), then dehydrated in a series of alcohol solutions andcritically point dried. The samples were coated with goldbefore SEM (LEO 1530VP, Zeiss) analyses.

Statistical analysis

All cell experiments were conducted in triplicate andwere repeated at least three separate times. All results arepresented as mean 6 the standard deviation (SD) andanalysis of variance (ANOVA) was used to check statisticalsignificance between means.

RESULTS

Topographical evolution of diamond coatings

SEM and AFM images of the different diamondcoatings fabricated under the varied growth condi-tions are shown in Figure 1. ND5 and ND10revealed nearly equal-axial, spherical nano crystalli-tes with sizes of 20–80 nm. ND10 exhibited slightlylarger grain sizes than ND5 and a few square grainsless than 100 nm were observed. For MD10, MD15,and MD25, grain sizes increased dramatically from200 nm to 2 lm and the formation of large crystallo-graphical facets (e.g. (111) and (100) planes) as wellas consequently micron size tetrahedral or cubicgrain shapes were detected. Morphological resultsobtained from AFM confirmed this evolution of dia-mond grain sizes and shapes. SEM and AFM analysesof NDH5 and NDH10 (not shown here) revealed verysimilar morphologies to ND5 and ND10, respectively.NDp that was fabricated without direct exposure toplasma revealed a very different topography com-posed of nanometer size platelets (Fig. 2). The 30–100nm platelets were irregular in shape and some sharp,thorn-like edges were observed. The Raman spectraprovided evidence that NDp consisted of diamondand amorphous carbon (spectra not shown here).

Quantitative roughness analyses (Rpv and Rrms) ofthe diamond surfaces are listed in Table II. NDp,ND and NDH had relatively smaller Rpv values thanMD. NDp, ND and NDH also had similar Rrms val-ues and they were much less than those for MD (ca.25 nm vs. 70 nm). Generally, Rpv and Rrms valuesincreased considerably as the H2 concentrationincreased, which agrees with the topographical evo-lution observed by SEM. NDH had similar Rrms andRpv values to ND, confirming the fact that they hadsimilar microstructures.

Wettability of diamond coatings

Contact angles of all the diamond coatings with-out H2 plasma treatment are presented in Figure 3.Results showed that the ND group had smaller con-tact angles than the MD, indicating that the ND coat-ings were more hydrophilic. Results also showedthat NDp was even more hydrophilic than other dia-mond coatings.

A parameter of diamond coatings, the roughnessfactor s, was also plotted in Figure 3. It is defined as:

s ¼ Ageo

Aprojð1Þ

where Ageo and Aproj are geometric areas and pro-jected areas obtained from AFM local scans, respec-

550 YANG, SHELDON, AND WEBSTER


tively. One can tell that there was a good positivecorrelation between s and contact angles. s of siliconwas defined as unity because it is very smooth.

Figure 4 shows contact angles of ND and NDHbefore and after surface cleaning. For nano-crystal-line diamond coatings without H2 plasma treatment

Figure 1. SEM (left) and AFM (right) images of (a) ND5; (b) ND10; (c) MD15; (d) MD20, and (e) MD25. [Color figure canbe viewed in the online issue, which is available at www.interscience.wiley.com.]



(i.e. ND5 and ND10), there was no statistically sig-nificant change in contact angles before and aftercleaning. However, NDH5 and NDH10 demon-strated statistically significant increases in contactangles after cleaning. Moreover, when comparing allthe samples after cleaning, the samples with H2

plasma treatment (NDH) were more hydrophobicthan those without treatment (ND), although the dif-ference was slight. s for each sample was also plot-ted. As the cleaning process did not change the sur-face morphology and subsequent s, the variation ofcontact angles on the same diamond coating beforeand after cleaning was independent from s. Further-more, s for ND and NDH were not significantly dif-ferent (p > 0.1), thus, s did not necessarily reveal acorrelation with a change of contact angles betweenND and NDH.

OB adhesion on diamond coatings

Results of this study showed that ND5 and ND10enhanced OB adhesion compared with NDp andMD (Fig. 5). NDp decreased OB adhesion the most,indicating an opposite effect of inhibiting cell adhe-sion compared with ND. For H2 plasma-treated dia-mond coatings, no statistically significant differencewas observed in OB adhesion between ND andNDH (Fig. 6). However, more hydrophilic ND

showed a slightly better OB adhesion than NDHwhen comparing Figure 4 and Figure 6, though thedifference was not statistically significant.

OB proliferation on diamond coatings

Results from the OB proliferation study (Fig. 7)showed that ND coatings with/without H2 plasmatreatment (i.e. ND and NDH) promoted OB prolifer-ation to a greater extent compared with MD and sili-con controls after 3 days. OBs on NDH showedslightly better proliferation than that on ND. All thediamond coatings enhanced OB proliferation morethan silicon controls after 3 days.

OB morphology on diamond coatings

SEM observations of OB on ND5 and MD20 after48 h are shown in Figure 8. As shown, the cell den-sity on ND5 [Fig. 8(a)] was higher than that onMD20 [Fig. 8(b)], which agrees with the results pre-viously presented for the OB proliferation tests.Although OBs spread well on both diamond coat-ings, extending of OB cell membranes on ND5 waslarger than on MD20. Comparing the filopodia ofOBs also suggested a better adhesion of cells onND5 as more radially distributed filopodia wereobserved on ND5 than on MD20.

DISCUSSION

Control of surface topography of diamond coatings

Topographical characterization studies exhibited apositive correlation between grain size and H2 flowrate/concentration in the MPCVD process. Nearlyequal-axial, round nano crystallites were formedwith H2 concentrations below 10% and the grain sizewas a few tens of nanometers. As the H2 concentra-

Figure 2. SEM (left) and AFM (right) images of NDp. [Color figure can be viewed in the online issue, which is availableat www.interscience.wiley.com.]

TABLE IISurface Roughness of Diamond Coatings

Name Rpv (nm) Rrms (nm)

NDp 277 6 136 23.1 6 5.1ND5 352 6 274 20.4 6 3.6ND10 394 6 200 26.9 6 1.7MD15 906 6 423 76.2 6 2.4MD20 755 6 145 63.9 6 4.1MD25 857 6 207 70.8 6 3.5NDH5 240 6 75 23.6 6 5.5NDH10 250 6 55 29.9 6 4.0



tion reached 15% and above, the growth of NDresulted in micrometer size tetrahedral or cubicshape diamond grains. The continuous increase ingrain size from the nano to micron range withrespect to the rise of H2 concentration can be attrib-uted to the suppression effect of H2 on secondarynucleation. This suppression results in the survivaland growth of primary diamond crystals into themicron size.15 This suppression effect also explainsthe trend that Rpv and Rrms values graduallyincreased within each group of diamond as more H2

was added to the MPCVD process, the grains on thesurface grew larger and consequently made thesurface more rough.

Importantly, the discussion above implies thepossibility that topographical features of diamondcoated on silicon (including, but no limited to, grain

size and surface roughness) can be simply controlledby H2 concentration during the MPCVD process.

Variation of wettability of diamond coatings

Information of surface wettability and surfaceenergy (although not directly) can be provided bycontact angles. The definition of contact angles is:

cos u ¼ gSV � gSLgLV

ð2Þ

where u is contact angle, gSV is interfacial tensionbetween the solid and vapor, gSL is the interfacialtension between the solid and liquid, and gLV is theinterfacial tension between the liquid and vapor.

Figure 4. Contact angles on ND and NDH before and after removal of surface contaminants. Data 5 mean 6 SD; N 5 5;*p < 0.01 compared with NDH5 before cleaning and ND5 after cleaning; **p < 0.001 compared with NDH10 before clean-ing; **p < 0.1 compared with ND10 after cleaning. Roughness values of each coating was also plotted.

Figure 3. Contact angles and roughness values for diamond coatings without H2 plasma treatment. Data 5 mean 6 SD;for contact angles: N 5 5; *p < 0.01 compared with NDp; **p < 0.01 compared with ND5 and ND10; #p < 0.05 comparedwith ND10.



However, the topographical features of rough surfa-ces can make the geometric surface area greater thanthe projected surface area (i.e. s greater than 1) andconsequently result in a variation of gSV as well asgSL.

17 If the geometric surface area is greater, gSVand gSL will decrease accordingly, which indicatesthat the difference between them is (i.e. gSV 2 gSL)minimized and gLV becomes crucial to determineu.18 Therefore, the value of cos u actually becomesvery small, suggesting a large u. This can explain thepositive correlation between s and contact angles inFigure 3 (i.e. a high s corresponded to a high contactangle).

However, s does not show a consistent correlationwith contact angles in Figure 4, implying other fac-tors may be influencing contact angles. Because s forND and NDH were not significantly different, theincrease in contact angles from ND to NDH (bothafter cleaning) cannot be negatively correlated to s.However, the increase can be attributed to a varia-tion of surface chemical bonding introduced by H2

plasma treatment. On one hand, H2 plasma treat-ment can induce strong C��H bonds by chemisorp-tion, which decreases gSV.

19 On the other hand, thesurface atomic reconstruction and formation of sp2

bonds triggered by H2 plasma can also decreasegSV.

14 Both of these processes will increase contactangles. The statistically significant increase in contactangles for both NDH5 and NDH10 after cleaningcompared with those before cleaning (notice that sdoes change after cleaning) indicates that surfacecontaminants induced by physico-absorption inMPCVD may also contribute to a variation of surfacewettability. Clearly, an important suggestion fromthis fact is that cleaning or sterile procedures caninduce different surface properties mediating celladhesion.

Again, the results above suggest that variations ins, surface chemical bonds and physico-absorbed con-taminants are all closely related to the concentrationand treatment of H2 plasma. In this sense, wettabil-

ity and energetic states of diamond surfaces can becontrolled by adjusting H2 plasma during theMPCVD process.

OB adhesion and proliferation ondiamond coatings

Most importantly, this study showed that NDwith grain sizes less than 100 nm promoted OB ad-hesion compared with MD. In addition to grain sizedifferences, it is important to emphasize that NDand MD have very different topographical featuresthat also influence surface wettability and surfaceenergetics. As ND had lower contact angles thanMD, it is not surprising that more hydrophilic surfa-ces led to better OB adhesion since many proteinsimportant for cell attachment prefer to adsorb to ahydrophilic surface.20,21 On the other hand, betterOB adhesion on the ND surface could be attributedto the enhanced surface area and energetics on nano-structured surfaces, which also influenced proteinadsorption and improved cell performance.22 How-ever, the adherent cell density was still lower onthese samples than on silicon controls. Therefore,other approaches (such as the use of various func-tional groups from specific proteins) are suggestedto further promote OB adhesion. Fortunately, surfacefunctionalization on nano-crystalline diamond is fea-sible and several researchers have already success-fully functionalized nano-crystalline diamond coat-ings with BMP-2 and DNA.10,11 OB adhesion on NDand NDH demonstrated that H2 plasma treatmentdoes not influence OB adhesion although there is asmall change in hydrophilicity on diamond surfaces.This is important to the surface functionalizationstudies mentioned above because the plasma treat-ment provides H-terminations that are suitable forvarious linkages to other molecules while maintain-ing the cell density on the diamond coatings.

Interestingly, OB adhesion tests revealed oppositeresults on ND and NDp. ND and NDp had differentsurface properties (surface chemistry, topographyand wettability), although they had similar surfaceroughness (Rpv and Rrms). ND is mainly composed

Figure 5. OB adhesion on ND, NDp, and MD. Data 5mean 6 SD; N 5 3; *p < 0.01 compared with NDp; **p <0.05 compared with MD15, MD20, and MD25.

Figure 6. OB adhesion on ND and NDH. Data 5mean 6 SD; N 5 3; *p < 0.01 compared with others.



of sp3 carbon, but NDp is a mixture of diamond andamorphous carbon (the latter is mainly sp2 carbon).However, unlike diamond-like-carbon (DLC) thatalso contains large portions of amorphous carbon orsp2 bonded carbon,23 NDp exhibited very poor OBadhesion, even though the surface was very hydro-philic. As NDp was fabricated without direct expo-sure to the H2-Ar-CH4 plasma, the diamond andamorphous carbon may have fewer H-terminationsor C��H bonds on the surface compared with theDLC coatings cited above. This difference in surfacechemistry could alter the availability of active sitesthat attract molecules and proteins, which may haveresulted in poor OB adhesion. The special topogra-phy of NDp (such as edge interspacing) may havealso affected cell attachment.24 However, the exactreasons behind this are not clear at this point andneed to be the focus of further evaluation. The inhib-itory effect of NDp on OB adhesion suggests a possi-ble method to reduce OB adhesion behavior onselected portions of orthopedic implants (such asrotating components) by creating similar diamondcoatings (which also possess excellent wear proper-ties).

This study also showed that ND with grain sizeless than 100 nm (i.e. ND and NDH) enhanced OBproliferation compared with the MD and the siliconcontrol after 3 days. This enhancement can be attrib-uted to better OB adhesion on ND and NDH. SEMobservations revealed improved spreading of OBmembranes on ND compared with MD. This obser-vation also supports findings of less proliferation ofOBs on MD.24 The proliferation results on NDH5were slightly better than ND5, indicating that the H2

plasma treatment does not inhibit OB proliferation

on nano-crystalline diamond coatings but possiblypromotes it. However, OB proliferation on ND andNDH was still lower than on the controls, so surface

Figure 7. OB proliferation on ND, NDH, and MD. Data 5 mean 6 SD; N 5 3; *p < 0.05 compared with MD20; **p <0.05 compared with ND5, NDH5, and MD20; ***p < 0.01 compared with ND5, NDH5, and MD20; #p < 0.05 comparedwith MD20; ##p < 0.01 compared with ND5, and NDH5.

Figure 8. SEM images of OB morphology on (a) ND5 and(b) MD20 of incubation for 48 h.



functionalization may also be needed to furtherenhance OB proliferation on diamond coatings. Thetrend that ND enhances OB proliferation is consistentwith the recent studies by Amaral et al.22 However,this study provides a clearer fundamental understand-ing of whyND enhances OB proliferation.

CONCLUSIONS

Surface properties of diamond coatings (includinggrain size, roughness, wettability and chemistry) areclosely related to and can be controlled by the H2

plasma in the MPCVD process. The correlationbetween OB functions and proper surface propertiesof diamond was established in this study for the firsttime. ND can promote OB adhesion and prolifera-tion, but MD and nanostructured diamond/amor-phous carbon inhibited these cell functions. The fea-sibility of controlling OB functions by adjusting sur-face properties of diamond coatings opens anapproach to further improve orthopedic implant effi-cacy. For instance, ND with grain size less than 100nm can be applied to the stem of a total hip implantwhere bone tissue bonding is highly desirable,whereas nanostructured diamond/amorphous car-bon can be used on the articulating surfaces of hipimplants where bone growth is unnecessary butwear-resistance is crucial. In addition, the investiga-tion of H2 plasma-treated diamond coatings showedvariations of surface properties but a consistency inOB functions compared with the coatings withoutH2 plasma treatment, which can be fundamental andcritical information for the further functionalizationof diamond coatings for orthopedic applications.

The authors thank Abhishek Kothari for assistance inMPCVD, Anthony W. McCormick for SEM training, andQunyang Li for acquiring AFM data.

References

1. Kobayashi A, Bonfield W, Kadoya Y, Yamac T, FreemanMAR, Scott G, Revell PA. The size and shape of particulatepolyethylene wear debris in total joint replacements. ProcInst Mech Eng H: J Eng Med 1997;211:11–15.

2. Pohler OEM. Failures of metallic orthopedic implants. In:Becker WT, Shipley RJ, editors. ASM Handbook. Vol. 11: Fail-ure Analysis and Prevention. Ohio, USA: ASM International;1986. p 670–694.

3. Hayward IP, Singer IL. The tribological behavior of diamondcoatings. In: Messier R, Glass JT, Bulter JE, Roy R, editors.Proceedings of the Second International Conference on NewDiamond Science Technology, Pittsburgh, PA: MaterialsResearch Society; 1986. p 785–789.

4. Lappalainen R, Selenius M, Anttila A, Konttinen YT, Santa-virta SS. Reduction of wear in total hip replacement prosthe-ses by amorphous diamond coatings. J Biomed Mater Res BAppl Biomater 2003;66:410–413.

5. Spear KE. Diamond-ceramic coating of the future. J AmCeram Soc 1989;72:171–191.

6. Drory MD, Gardinier CF, Speck JS. Fracture toughness ofchemically vapor-deposited diamond. J Am Ceram Soc1991;74:3148–3150.

7. Azevedo AF, Corat EJ, Ferreira NG, Trava-Airoldi VJ. Wett-ability and corrosion tests of diamond films grown onTi6Al4V alloy. Surf Coatings Technol 2005;194:271–275.

8. Gowri M, Li H, Schermer JJ, van Enckevort WJP, ter MeulenJJ. Direct deposition of diamond films on steel using a three-step process. Diamond Relat Mater 2006;15:498–501.

9. Catledge SA, Fries MD, Vohra YK, Lacefield WR, Lemons JE,Woodard S, Venugopalan R. Nanostructured ceramics forbiomedical implants. J Nanosci and Nanotechnol 2002;2:293–312.

10. Steinmuller-Nethl D, Kloss FR, Najam-Ul-Haq M, Rainer M,Larsson K, Linsmeier C, Kohler G, Fehrer C, Lepperdinger G,Liu X, Memmel N, Bertel E, Huck CW, Gassner R, Bonn G.Strong binding of dioactive BMP-2 to nanocrystalline dia-mond by physisorption. Biomaterials 2006;27:4547–4556.

11. Yang WS, Auciello O, Butler JE, Cai W, Carlisle JA, Gerbi JE,Gruen DM, Knickerbocker T, Lasseter TL, Russell JN, SmithLM, Hamers RJ. DNA-modified nanocrystalline diamondthin-films as stable, biologically active substrates. Nat Mater2002;1:253–257.

12. Tang L, Tsai C, Gerberich WW, Kruckeberg L, Kania DR. Bio-compatibility of chemical-vapor-deposited diamond. Biomate-rials 1995;16:483–488.

13. Popov C, Kulisch W, Reithmaier JP, Dostalova T, Jelinek M,Anspach N, Hammann C. Bioproperties of nanocrystalline di-amond/amorphous carbon composite films. Diamond RelatMater 2007;16:735–739.

14. Garguilo JM, Davis BA, Buddie M, Kock FAM, Nemanich RJ.Fibrinogen adsorption onto microwave plasma chemicalvapor deposited diamond films. Diamond Relat Mater2004;13:595–599.

15. Gruen DM. Nanocrystalline diamond films. Annu Rev MaterSci 1999;29:211–259.

16. Webster TJ, Siegel RW, Bizios R. Osteoblast adhesion onnanophase ceramics. Biomaterials 1999;20:1221–1227.

17. Erbil HY, Demirel AL, Avc Y, Mert O. Transformation of asimple plastic into a superhydrophobic surface. Science2003;299:1377–1379.

18. Neinhuis C, Barthlott W. Seasonal changes of leaf surfacecontamination in beech, oak, and ginkgo in relation to leafmicromorphology and wettability. New Phytol 1998;138:91–98.

19. Hansen JO, Copperthwaite RG, Derry TE, Pratt JM. A tensio-metric study of diamond (111) and (110) faces. J ColloidInterf Sci 1989;130:347–358.

20. Khosroshahi ME, Tavakoli J, Mahmoodi M. The influence ofsurface chemistry and topography on the contact guidance ofMG63 osteoblast cells. J Mater Sci Mater Med 2007;18:705–714.

21. Wilson CJ, Clegg RE, Leavesley DI, Pearcy MJ. Mediation ofbiomaterial-cell interactions by adsorbed proteins: A review.Tissue Eng 2005;11:1–18.

22. Amaral M, Dias AG, Gomes PS, Lopes MA, Silva RF, SantosJD, Fernandes MH. Nanocrystalline diamond: In vitro bio-compatibility assessment by MG63 and human bone marrowcells cultures. J Biomed Mater Res A 2008;87:91–99.

23. Cui FZ, Li DJ. A review of investigations on biocompatibilityof diamond-like carbon and carbon nitride films. Surf Coat-ings Technol 2000;131:481–487.

24. Kunzler TP, Huwiler C, Drobek T, Voros J, Spencer ND.Systematic study of osteoblast response to nanotopographyby means of nanoparticle-density gradients. Biomaterials2007;28:5000–5006.



orthopedic nano diamond coatings: control of surface properties and their impact on osteoblast...

Documents