surface & coatings technology - city university of hong … adhesion of biologically active...

Surface & Coatings Technology 265 (2015) 222–234

Contents lists available at ScienceDirect

Surface & Coatings Technology

j ourna l homepage: www.e lsev ie r .com/ locate /sur fcoat

Graded metal carbon protein binding films prepared by hybrid cathodicarc — Glow discharge plasma assisted chemical vapor deposition

Mohammed Ibrahim Jamesh a,b, R.L. Boxman a,c, Neil J. Nosworthy a, I.S. Falconer a, Paul K. Chu b,⁎,Marcela M.M. Bilek a, Alexey Kondyurin a, R. Ganesan a, David R. McKenzie a,⁎a Applied and Plasma Physics, School of Physics (A28), University of Sydney, Sydney, NSW 2006, Australiab Department of Physics and Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, Chinac Electrical Discharge and Plasma Lab, Tel Aviv University, POB 39040, Tel Aviv 69978, Israel

⁎ Corresponding authors. Tel.: +852 34427724; fax: +E-mail addresses: [email protected] (P.K. Chu), D

(D.R. McKenzie).

http://dx.doi.org/10.1016/j.surfcoat.2014.11.0250257-8972/© 2014 Published by Elsevier B.V.

a b s t r a c t
a r t i c l e i n f o
Article history:Received 11 September 2014Accepted in revised form 8 November 2014Available online 15 November 2014

Keywords:Plasma polymerizationProtein immobilizationGraded interlayerThin film adhesionCathodic arc deposition

Graded composite layers containingmetal and plasma polymer components were deposited using a cathodic arcin conjunction with plasma immersion ion implantation. Using a bias potential throughout, pure metal wasdeposited initially using the cathodic arc alone and then acetylene was added to the process to increase thefraction of the plasma polymerized carbon film. To test adhesion, the substrate and filmwere strongly deformedby folding the substrate inward and outwardwith a small radius of curvature. Strong adhesion between themetalsurface and the deposited layers was achieved by the use of the graded layers as inferred from the SEM observa-tions of the deformation region. Strong adhesion of biologically active protein molecules to the surface of thegraded layer was confirmed by detergentwashing and colorimetric enzyme activity assays. These characteristicssuggest that the coatings may be suitable for cardiovascular stent applications.

© 2014 Published by Elsevier B.V.

1. Introduction

Immobilizing proteins on surfaces by means of covalent bindingwhile retaining their function over a long period of time is increasinglyin demand for many biomedical and biochemical applications such asbiosensors and medical implants [1–6]. For implantable medical deviceapplications, immobilizing proteins by utilizing physical interactionsrenders the proteins vulnerable to detachment and undesirable confor-mation changes which affect their bioactivity can lead to adverse re-sponses such as inflammation, clotting, activation of cellular immuneresponse, excessive fibrosis, or rejection. During heart transplantationusing cardiopulmonary bypass heart-lung machines, patients exhibitan inflammatory response which is characterized by increased expres-sion of at least ten leukocyte cluster-of-differentiation (CD) antigens[7]. Coronary artery disease is primarily caused by the inflammation ofcoronary vesselwalls [8] and a study on coronary stents provided strongevidence that increased inflammation can arise from the use of barestents [9]. To minimize these inflammatory responses, attempts havebeen made to cover surfaces which are exposed to blood with proteinmolecules such as albumin derived from human blood plasma.However, these attempts have frequently been unsuccessful [10]. The

852 [email protected]

lack of success may be attributed to difficulties in achieving strongprotein binding or causing denaturation (conformational change) ofthe molecule. Maintaining the native conformation of an immobilizedbiomolecule is critical in situations where the active site involvesamino acid residues that are proximate due to the molecule's energeti-cally favored folded state in aqueous biological environments. In suchcases, varying of conformation state of a molecule can adversely affectthe biorecognition process [11].

Plasma modification technologies have been found useful to im-prove surface properties [11,12] such as imparting a covalent proteinbinding capability [1,13–16], corrosion resistance [17,18], biocompati-bility and antimicrobial [19] properties of biomaterials.

Surface energy is an important characteristic that determines thestrength of physical adsorption and the degree of retention of nativeconformation of protein molecules. Highly hydrophilic surfaces suchas polyethylene glycol do not bind proteins [20] whereas hydrophobicsurfaces adsorb proteins relatively strongly but typically induce changesin their conformation. Hydrophilic surfaces, while desirable to preserveprotein function [21], have low protein binding affinity [22]. Bilek et al.[1] demonstrated a plasma ion-assisted processes which can enablecovalent binding to hydrophilic surfaces. The plasma modification wasshown to generate mobile unpaired electrons associated with freeradicals and these were shown to be the mechanism for the formationof covalent bonds with biomolecules. Yin et al. [2] demonstrated onesuch carbon plasma deposition process on a stainless steel (SS) sub-strate which provided a robust interface for covalent immobilization

http://crossmark.crossref.org/dialog/?doi=10.1016/j.surfcoat.2014.11.025&domain=pdf

http://dx.doi.org/10.1016/j.surfcoat.2014.11.025

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.surfcoat.2014.11.025

http://www.sciencedirect.com/science/journal/02578972

223M.I. Jamesh et al. / Surface & Coatings Technology 265 (2015) 222–234

of tropoelastin while Bax et al. [23,24] demonstrated a pure implanta-tion process to covalently immobilize a range of extra-cellular-matrixproteins to PTFE. The immobilized tropoelastin protein molecules en-hanced the attachment of endothelial cells and improved theirproliferation.

Amorphous carbon films are well-known as being biocompatible[25–29] and hence it is an attractive candidate for covalent immobiliza-tion of protein molecules [2,15,30,31]. Liu et al. [25] studied that amor-phous carbon film stimulates lower inflammatory reaction and higherosteoblast viability. Wei et al. [26] studied that amorphous carbon filmpromotes the human endothelial ECV304 cell growth. Ma et al. [27]studied that amorphous carbon film suppressed the viable monocyte/macrophages attachment. Kwok et al. [28] studied that silver dopedamorphous carbon film improves the hemocompatibility and anti-bacterial properties. Dowling et al. [32] reported that after 4 weeks,amorphous carbon film implant had no macroscopical adverse effectson the cortical bone or muscular tissue of sheep under in vivo condi-tions. Zolynski et al. [33] reported that after 28 weeks, amorphouscarbon film implant enables the healing of bone fracture under in vivoconditions in the human body. Waterhouse et al. [15] reported thatamorphous carbon film covalently immobilizes protein molecules andreduces thrombogenicity. Yin et al. [2] reported that amorphous carbonfilm covalently immobilizes protein molecules. However, in the case ofmetallic materials such as stainless steels, adhesion between carbon-based plasma polymers and the metal is a major problem because ofthe different thermal expansion coefficients between the twomaterials.Yin et al. [30] deposited a carbon-based plasma polymer film using agraded interface between the metal and carbon containing polymerwhich displayed robust covalent immobilization of protein as well asexceptionally strong adhesion (between 18 and 26 MPa) to the under-lying metal substrate.

Strong adhesion is desirable for biomedical implants such as cardio-vascular stents that undergo major deformations during surgical

WATER COOLING

CATHODE

TRIGGER3 Ω

FIELD COILS (Curren

SUBSTRATE HO

WATERCOOLING

Fig. 1. Schematic illustration

insertion. Graded layers can solve the adhesion problem between me-tallic materials and a carbon layer provided that the metal layer that isinitially deposited as part of the graded layer adheres well to themetal substrate. Cathodic arc deposition, a type of physical vapor depo-sition, is suitable for achieving well adheredmetal layers because of thehigh ion content of the depositing materials, thus enabling the metalions to be implanted into the metal surface in a process known as plas-ma immersion ion implantation and deposition (PIII&D). This type ofdeposition produces strong adhesion of the coating to the underlyingmetallic substrate [34–36]. A well adhered gradient metal-carbonlayer deposited using a cathodic arc process and enabling covalent bind-ing of biomolecules has not previously been reported. The aim of thiswork is to demonstrate that cathodic arc deposition combined withplasma immersion ion implantation (PIII), and acetylene plasmaassisted chemical vapor deposition (PACVD) with PIII provides an ad-herent graded composition coating with covalent binding capabilityfor the immobilization of biologically active proteins.

2. Methods and materials

2.1. Plasma deposition processes

Films containing Ti andCwere deposited in a cathodic arc depositionchamber (Fig. 1) fitted with a titanium cathode of 99.998% purity. Acurved magnetic solenoid filter of the type described by Aksenov et al.[37] was used to filter neutral atoms and macro-particles from theplasma stream. 10A currentwas applied to thefield coils of themagneticfilter, consisting of two layers of wire with n turns per meter. More de-tails about the chamber have been reported earlier [38–40]. Briefly,two power sources were used: (1) a low voltage and high current (setto 60 A) source to power a cathodic arc which generated the metalvapor plasma, and (2) a pulsed voltage source, set at −2 kV, 20 μspulse width, and 3 kHz pulse repetition frequency, for biasing the

PLASMA BEAM

t:10 A)

LDER

DC BIAS

ACETYLENE INLET

ANODE

of cathodic arc chamber.

(b) Pressure

50º

Coating

Substrate

Pressure

Substrate

Coating

64º

(a) Pressure

Pressure

Fig. 2. Schematic illustration of deformation by bending test (a) outward and (b) inward.

224 M.I. Jamesh et al. / Surface & Coatings Technology 265 (2015) 222–234

substrates and depositing carbon films by plasma assisted chemicalvapor deposition (PACVD) in a glow discharge. The glow dischargewas sustained in acetylene gas between a 80 mm diameter steel disksubstrate holder which acted as the cathode and the 30 cm diameterchamber, which was grounded and served as the anode. The pulsedpower supply was connected to the substrate feed through via an~2.5 m coaxial cable. The gas flow rates were regulated by mass flowcontrollers. Double side polished p-type silicon wafers (100) about300–360 μm thick and 316 L stainless steel (SS) foil (25 μm thick)were used as a substrates. Deposition was conducted at a base pressureof about 10−6 Torr.

The experimental parameters used for carbon films (C1–C5), Ti film(M1), Ti/C film (MC1), Ti/C graded film (G0) and Ti/C graded film witha top carbon layer (G1 and G2) are displayed in Table 1. Prior to deposi-tion, substrates were ultra-sonically cleaned with alcohol for 5 min andthen with distilled water for 5 min and then dried using compressedair. The Ti film M1 was deposited using the cathodic arc on a siliconwafer, without C2H2, for 5 min. During all of the arc depositions,−2 kVpulses were applied to the substrate holder. The MC1 Ti/C film was de-posited under the conditions noted for the M1 film, plus the additionof C2H2 gas with a flow rate of 80 sccm. For the graded coatings (G0,G1, andG2), the Tifilmwas deposited on stainless steel using the cathod-ic arcwith C2H2 introduced at different flow rates. The C2H2 gas flow ratewas increased in steps, starting at 0 and then 10, 20, 40, 60 and 80 sccmfor a period of 60 s at each pressure. In films G1 and G2, a C topcoat wasdeposited, after the graded Ti/C deposition, using the PACVD process at100 mTorr C2H2 for 5/10 min respectively. Carbon films were depositedusing only PACVD, i.e. without the arc. C1, C2 and C3 films were deposit-ed on silicon wafers for 11 min in 50, 100 and 150 mTorr of C2H2

gas, while C4/C5 carbon films were deposited on stainless steel for 5min/10 min in 100 mTorr.

2.2. Physical, chemical, and mechanical characterization

Raman scattering spectra were acquired in air using a LabRam HRspectrometer (Horiba Jobin Yvon) equipped with an Argon laser(green line at 514.5 nm) and an Olympus BX41 microscope. The chem-ical state of the coating was analyzed using X-ray photoelectron spec-troscopy (XPS) (ESCALAB220i-XL from Thermo Scientific). Crosssectional views and their corresponding EDX map on G1 Ti/C gradedfilms were examined using a Zeiss Ultraplus field emission scanningelectronmicroscope. The surface morphology of the C1, C2 and C3 sam-pleswere examined using a Zeiss EVO 50 scanning electronmicroscope.Atomic force microscopy (AFM) was employed in the tapping mode toexamine the morphology of the surface before and after plasma treat-ment using Pico SPM with WSxM software (Nanotec Electronica) [41].Bending tests were performed to determine the adhesion strength ofthree types of films: (1) PACVD carbon films (C4 and C5), (2) gradedTi/C without a carbon topcoat (G0) and (3) graded Ti/C with a topcoat(G1 and G2) on stainless steel substrates. The coated stainless steel

Table 1Experimental parameters used for the plasma deposition of Ti film, Ti/C film, Ti/C graded film,

S. no. Sample ID Type of film Substrate C2H2 flowarc deposi

1. M1 Ti film Si 02. MC1 Ti/C film Si 803. G0 Ti/C graded film Stainless steel 0, 10, 20, 44. G1 Ti/C graded film with a carbon topcoat Stainless steel 0, 10, 20, 45. G2 Stainless steel 0, 10, 20, 46. C1 Carbon film Si7. C2 Si8. C3 Si9. C4 Stainless steel10. C5 Stainless steel

sample was bent outward and inward as shown in Fig. 2 to create twotypes of large deformations. The substrate was bent outward (Fig. 2a)using a sharp edge and bent inward (Fig. 2b). The bending deformationin all of these samples and the EDX spectrum of the deformed G2 sam-ple were examined using a Zeiss EVO 50 scanning electron microscope.The wettability of the stainless steel substrate and the G0 and G1 Ti/Cgraded films was measured using the sessile drop method by contactangle measurement equipment (Kruss DS10) with water and diiodomethane. The surface energy was calculated using the Owens–Wendt–Rabel–Kaelble method [42].

Ti/C graded film with a carbon topcoat and only carbon film.

rate during Ti cathodiction (sccm/min)

CVD of C2H2 Total duration (min)

Pressure (mTorr) Duration (min)

55

0, 60 & 80 60, 60 & 80 100 5 110, 60 & 80 100 10 16

50 11 11100 11 11150 11 11100 5 5100 10 10

(a)

10 μm 10 μm

10 μm

C1 carbon film C2 carbon film

C3 carbon film

(b)

(c)

Fig. 3. Surface morphology of (a) C1, (b) C2 and (c) C3 carbon films at 2000× magnification.


2.3. Protein binding test

Horseradish peroxidase (HRP) (Sigma P6782) was diluted to 50μg/mL in 10 mM sodium phosphate buffer (pH 7 PO4 buffer). Sampleswere incubated in the protein solution overnight with rocking in

(a)

C1

C3

G1

Fig. 4. XPS survey spectra of (a) C1 & C3 carbon films, (b) M1 Ti film an

75 mm sterile Petri dishes. After incubation, HRP was removed fromthe Petri dish and 15 mL of PO4 wash buffer was added and left for20min. Sampleswere then transferred to 50mL Falcon tubes containing40 mL of fresh PO4 buffer and rocked for an additional 20 min. This wasrepeated four times. To determine the binding strength, protein coated

(b)

(c)

M1

MC1

d MC1 Ti/C film and (c) G1 Ti/C graded film with a carbon topcoat.

Ti & C film on Si

Inte

nsity

(a.u

) sp3

sp2

Ti-C

C-O

(a)

Binding Energy (eV)

Ti & C graded + top carbon film on stainless steel

Inte

nsity

(a.u

)

sp3

sp2

(b)

C-O

281 282 283 284 285 287 286 288 289 Binding Energy (eV)

281 282 283 284 285 287 286 288 289

MC1 G1

Fig. 5. High resolution XPS spectra of C 1s obtained from (a) MC1 Ti/C film and (b) G1 Ti/C graded film with a carbon topcoat.


samples were incubated in a detergent 0.5% sodium dodecyl sulfate, in10 mM PO4 buffer for 1 h at room temperature. Samples were thenwashedwith PO4 buffer twomore times to remove the detergent beforetesting the surfaces for activity.

Ti interface

Coating

Stainless

substr

Bakelite

Ti/C

int

C

Ti

Fe In

ten

sity (a.u

)

0 0.2 0.4 0.6 0.8 1

2.5 µm

G1 g

Fig. 6. (a) Cross sectional view, (b) corresponding EDS map and (c) EDS

2.4. Protein activity test

The functionality of the enzymewas determined using a colorimetricassay. HRP protein coated samples (HRP substrate) were clamped

(b)CoatingC K series

Ba

ke

lite

Substrate

Ti interface

Ba

ke

lite

Substrate

Ba

ke

lite

1 µm

1 µm

1 µm

Substrate

Ti L series

Fe L series

steel

ate

graded

erface

C

Ti

Fe

(a)

(c)

1.2 1.4 1.6 1.8 2 µm

raded film

line scan spectrum on G1 Ti/C graded film with a carbon topcoat.


between two stainless steel plates separated by an O-ring (inner diame-ter 8 mm, outer diameter 11 mm), which sealed to the plasma-treatedsurface. One of the plates contained a 7 mm hole, through which theHRP substrate, TMB (3,3′,5,5′ tetramethylbenzidine, Sigma T0440 —

75 μL), was added. The solution was allowed to develop a blue color viathe oxidization of TMB by the oxygen radical generated during the hy-drolysis of H2O2 by HRP. After 30 s, 25 μL of oxidized TMB was removedand added to 50 μL of 2MHCl, producing a yellow diimine oxidation endproduct [43]. An additional 25 μL of unreacted TMB was then added,bringing the total volume to 100 μL. The absorbance of the liquid at450 nm was measured using a spectrophotometer. The HRP activitytest was carried out at room temperature (23 ± 2 °C) [44].

3. Results

The surface morphologies of the C1, C2 and C3 carbon films areshown in Fig. 3. The surface of the C1 and C2 films (50, 100 mTorr)was smoothwhereas the C3 (150mTorr) surfacewas rough. XPS surveyscan spectra of theC1 andC3 carbonfilms,M1 Tifilm,MC1Ti/Cfilm, andG1 Ti/C graded with a top carbon film are shown in Fig. 4. Fig. 4(a)shows a stronger C 1s peak from the C1 and C3 carbon films. A weakestC 1s peak and a stronger Ti 2p peak with a strongest O 1s peak was

(

(

400 μm

400 μm

(d)

(b)

C5 carbon film (lower magnification)


Deformed region having film with a

thin coating

Deformed region having film with a

thin coating

Before bending After ben

Coating

Substrate

T

(a)

Fig. 7. (a) Schematic diagram of bending coating outward; Surface morphology of (b, c) C4 andbend outward.

observed from theM1 Tifilm. However, theMC1 Ti/C film showed near-ly equal intensity of the Ti 2p, O 1s, and C 1s peaks. In contrast to theMC1 Ti/C film, the G1 Ti/C graded with a top carbon layer showed astronger C 1s peak without a Ti 2p peak. High resolution C 1s XPS spec-tra of the MC1 film and G1 film are shown in Fig. 5(a) and (b) respec-tively. The binding energy of the C 1s electrons emitted from the MC1filmwas 284.8 eV, while the C1s peak obtained from the G1 Ti/C gradedfilm with a top carbon layer appeared at a binding energy of 285.1 eV.The cross sectional view, corresponding EDS map and EDS line scanspectrum on G1 Ti/C graded with a top carbon film are shown inFig. 6. The thickness of the Ti/C interface was about 300 nm (Fig. 6c)and the thickness of the coating (Ti/C gradedwith a top carbon coating)was 798 nm (Fig. 6a).

A schematic diagram representing the mechanism of crack forma-tion with possible delamination due to bending outward is shown inFig. 7(a). The surface morphology of three types of films on a stainlesssteel substrate after bending outward is shown in Fig. 7(b–k): PACVDcarbon films (C4 and C5); graded Ti/carbon without a carbon topcoat(G0) and with a topcoat (G1 and G2). Film delamination was observedin the deformed region of the C4 (Fig. 7b) and C5 (Fig. 7d) carbon filmsand a small piece of coating which appears to have broken loose wasobserved in the higher magnification views (Fig. 7c and e). On the

400 μm

c)

e)

400 μm

C4 carbon film (higher magnification)


Deformed region shows a small piece of

coating which appears to have broken loose

Deformed region shows a small piece of

coating which appears to have broken loose

ding

Coating

Substrate

Coating

Substrate

ensile stress

Cracks or delamination

After bending and then unbending

(d, e) C5 carbon films, (f, g) G1, (h, i) G2, and (j, k) G0 Ti/C graded films on stainless steel

(k)

400 μm

( j )

400 μm

400 μm

(i)

400 μm

(h)

400 μm

(g)

400 μm

(f) G1 Ti / C graded with a top carbon film (higher magnification)

G2 Ti / C graded with a top carbon film (higher magnification)

G0 Ti / C graded without a top carbon film (higher magnification)

G0 Ti / C graded without a top carbon film (lower magnification)

G2 Ti / C graded with a top carbon film (lower magnification)


Deformed region (undamaged coating) Deformed region (undamaged coating)

Coating Coating

50 μmSmall patches of delaminated area

Crack like pattern Deformed region with crack like pattern

and small patches of delaminated area

Deformed region with crack like pattern






50 μm

Small patches of delaminated area

Crack like pattern

Fig. 7 (continued).


other hand, a crack like pattern and small patches of delamination wereobserved throughout the deformed region from the G1 graded film(Fig. 7f and g), whereas a crack like pattern and small patches ofdelaminatedmaterial were observed froma fewplaces on the deformedregion in the G2 graded film (Fig. 7h and i). In contrast, neither delami-nation nor a crack-like pattern was observed from the G0 graded film(Fig. 7j and k). The surface morphology of the G2 film (Ti/C with carbontopcoat) on stainless steel bent outward and the EDS results acquiredfrom thedeformed region (Spectrum1: spot analysis on the delaminatedarea and spectrum 2: spot analysis on the non-delaminated area) areshown in Fig. 8. The spot analysis performed on the non-delaminatedarea showed a strong peak for carbon whereas the spot analysis on thedelaminated area showed peaks characteristic of the stainless steelsubstrate.

A schematic diagram which illustrates the mechanism of crack for-mation due to bending inward is shown in Fig. 9(a). The surface mor-phology of the three types of films on a stainless steel substrate afterthe inward bending test is shown in Fig. 9(b–k): PACVD carbon films(C4 and C5); graded Ti/carbon film without carbon topcoat (G0); andwith a topcoat (G1 and G2). Complete cleavage of the coating on acomparatively narrow area was observed on the C4 carbon film(Fig. 9b and c) and large patches of crack propagation over a largerarea were observed on the C5 carbon film (Fig. 9d and e). On theother hand, small patches of crack propagation in a comparativelywider areawere observed on theG1 graded film (Fig. 9f and g),whereassmall patches of crack propagation in a comparatively narrow areawere

observed on theG2 gradedfilm (Fig. 9h and i). In contrast, neither cleav-age nor cracking was observed on the G0 graded film (Fig. 9j and k).

The Raman spectrum acquired from the G1 Ti/C graded with top car-bon film is depicted in Fig. 10. The Raman spectrum can be deconvolutedinto two sub-bands: The C5/C7 odd ring band at 1548.15 cm−1 and theCsp3–Csp3 cluster band at 1354.46 cm−1.

The contact angle and surface energy of the bare stainless steel, G0Ti/C graded film without a carbon topcoat and G1 Ti/C graded filmwith a carbon topcoat are shown in Fig. 11. The water contact angle ofthe untreated stainless steel was about 88°, whereas those of G0 Ti/Cgraded film without a carbon topcoat and G1 Ti/C graded film with acarbon topcoat were about 65° and 68°, respectively (Fig. 11a). Theformamide contact angle was also reduced on the G0 and G1 gradedfilms compared to that of stainless steel. The surface energy of G0 andG1 graded filmswas higher (about 52mJ/m2) than that of the untreatedstainless steel (about 35 mJ/m2). AFM images of bare stainless steel, G0and G1 are shown in Fig. 12. The bare stainless steel and G0 wererelatively smooth whereas G1 with a top carbon film was rougher.

Fig. 13 shows the optical density arising from theHRP activity on theuntreated and plasma treated surfaces before and after SDS detergentwashing. The OD values observed for the treated surface without SDSwashing (mean 0.549)was significantly larger than that for the untreat-ed surface (mean 0.27, p b 0.0016), suggesting that the treated surfacewas either more densely covered with an active protein than theuntreated surface or that the molecular function was improved. Theratio of the OD value before and after SDS washing reveals that the G0

Small patches of delaminated area

Crack like pattern

Spectrum 1

Spectrum 2

Spectrum 1

Spectrum 2

200 μm 80 X magnification 50 μm


G2 Ti / C graded with a top carbon film (higher magnification) (a) (b)

Fig. 8. (a, b) Surface morphology of the G2 Ti/C graded filmwith a carbon topcoat on stainless steel bend outward and EDX analysis performed on the deformed region; spectrum 1: spotanalysis on the delaminated area and spectrum 2: spot analysis on the non-delaminated area.


plasma modified stainless steel surface retained approximately 77% ofthe active protein coverage on the surface in a functional state whereasafter SDS washing, 60% of the protein was retained in a functional state.Application of the test revealed that the difference between the meansof the samples before and after washing was not significant for thetreated samples (p b 0.13) but was significant for the untreated samples(p b 0.0036).

4. Discussion

The surface morphology (Fig. 3) of the carbon films deposited at 50mTorr (C1) and 100mTorr (C2) C2H2were smooth, whereas the surfacewas rough at 150mTorr (C3) C2H2. This could be possibly due to the for-mation of clusters in the plasma [45]. The XRD spectrum (not providedhere) did not show any characteristic peaks which reveals that the car-bon films were amorphous whereas the XPS spectra (Figs. 4 and 5)show characteristic peaks and provide much more information aboutthe carbon films. The weakest C 1s XPS peak (Fig. 4b) observed fromthe M1 Ti film may be due to atmospheric contaminants. The stronger

Ti 2p peak (Fig. 4b) with the strongest O 1s peak was observed fromthe M1 Ti film, and suggests that the Ti was oxidized. However, theMC1 Ti/C film showed nearly equal intensity (Fig. 4b) of the Ti 2p, O1s, and C 1s peaks, which were possibly due to the formation of Ti–Cbonding along with Ti–O bonding. In contrast to the MC1 Ti/C film(Fig. 4b), the G1 Ti/C gradedwith a carbon topcoat (Fig. 4c) had a stron-ger C 1s peakwithout a Ti 2p peak, suggesting that the top surface of theG1 film was composed of carbon, without titanium. Fitting with CASAXPS software indicated that the C 1s XPS spectra of the MC1 film(Fig. 5a) was composed of 4 sub-peaks and the G1 film (Fig. 5b) of 3sub-peaks. The MC1 film XPS spectrum had a peak at 283 eV, whichcould be due to bonding between Ti and C. The C 1s XPS spectra ofboth the MC1 (Fig. 5a) and G1 (b) films contained the peak for the C–O structure, indicating the presence of an oxide layer on the surface.MC1 and G1 films had binding energies at 285.1 and 285.2 eV respec-tively. These suggest that the carbon films contained sp3 (diamond-like) bonding in the amorphous carbon phase. The MC1 Ti/C film andG1 Ti/C graded film with a carbon film topcoat had binding energies of284.3 and 284.6 eV respectively. This suggests that the carbon film


formedwas graphite-like, i.e. containing carbon atoms in sp2 hybridiza-tion. The sp3 amorphous carbon phasewasmore pronounced for the G1film (66.3%) than for the MC1 film (49.5%). The EDS line scan spectrum(Fig. 6c) of the G1 film revealed that a Ti/C graded interface was formedbetween the stainless steel substrate and coating.

As stents require well adhered coatings that are stable after beingextensively deformed. Wire stents are extensively deformed at the in-tersections of their wires during their expansion when the stent isinserted into an artery. Outward bending produces tensile stress in thecoating and cracks, and possibly delamination. An acceptable outcomeis a film which might crack on bending, but without delamination, andthe cracks close upon un-bending. In contrast to the C4 and C5 carbonfilms (Fig. 7b–e) which, after bending and unbending contained only athin coating after the film mostly delaminated, the G1 and G2 filmsand graded Ti/C with a carbon topcoat (Fig. 7f–i) had only a few delam-ination patches, suggesting that the adhesion of the carbon filmwas im-proved by the Ti/C graded interface. Nevertheless, the Ti/C graded layerof theG0film did not delaminate (Fig. 7j and k). The EDS spot analysis ofthe deformed G2 film (Fig. 8) showed that the well adhered area wascomposed of carbon whereas the delaminated area was composed ofstainless steel, i.e. the substrate. This observation suggests that the de-lamination occurred only at the Ti/C interface with the substrate. This

Before bending After be

Coating

Substrate Substrate

Compre

(a)


Deformed region (Complete cleavage of coating on a comparatively narrow area)

(

(

400 μm

400 μm

(d)

(b)


Deformed region (large patches of crack propagation in a comparatively bigger area)

Fig. 9. (a) Schematic diagramof bending coating inward; surfacemorphology of (b, c) C4 and (dinward.

further suggests that the carbon topcoat adheredwell to the Ti/C gradedlayer but failure occurred at the interface between the stainless steelsubstrate and the Ti/C graded layer. That theG0 film did not delaminate,while the G1 and G2 films did delaminate at the stainless steel/ Ti/C in-terface, suggests that deposition of the C topcoatweakened the stainlesssteel/ Ti/C interface, perhaps by adding more force at the interface.

Bending inward compressively stresses the film, inducing upwardexpansion and possible delamination. The C4 carbon film (Fig. 9b andc) was completely cleaved, whereas large patches of cracks, but nocleavage, were observed on the C5 carbon film (Fig. 9d and e). Thesesuggest that adhesion of the C5 carbon film was slightly better thanthat of the C4 carbon film. In contrast to the C4 and C5 carbon films(Fig. 9b–e), the graded G1 film, (Fig. 9f and g) had few crack patches,suggesting that the adhesion strength of the carbon film was improvedby the Ti/C graded interface, whereas the graded G2 film with a carbontopcoat (Fig. 9h and i) had only a few crack patches, and thesewere con-fined to a narrow region. This suggests that the adhesion strength of theG2 film is better than that of the G1 film. Nevertheless the G0 film andgraded Ti/C without a carbon topcoat (Fig. 9j and k) had no cracks. G0had no cracks, and G1 and G2 had few cracks. Together with the previ-ous sentence and other discussed evidence (bending outward test), thisagain suggests that the addition of the C topcoatweakened the stainless

nding After bending and

then unbending

Coating

Coating

Substrate

ssive stress

Cracks


400 μm

c)

e)

400 μm


Deformed region (Complete cleavage of coating on a

comparatively narrow area)

Deformed region (large patches of crack propagation in a

comparatively bigger area)

, e) C5 carbon films, (f, g) G1, (h, i) G2, and (j, k) G0 Ti/C graded films on stainless steel bend

Deformed region (small patches of crack propagation in a

comparatively wider area)

Deformed region (small patches of crack propagation in a comparatively wider area)

G1 Ti / C graded with a top carbon film (higher magnification)


(k)

400 μm

( j )

400 μm

400 μm

(i)

400 μm

(h)

400 μm

(g)

400 μm

(f )

Deformed region (small patches of crack propagation in a

comparatively narrow area)


Small patches of crack propagation in a

comparatively narrow area


Deformed region


Deformed region (undamaged coating)

Coating


Deformed region (undamaged coating)

Coating

Fig. 9 (continued).

G1 Ti/C graded film with a carbon topcoat

Fig. 10. Raman spectrum of G1 Ti/C graded film with a carbon topcoat.


steel/ Ti/C interface. When an atomic impurity such as Ti is introducedinto the interface in contact with carbon, some amorphization andhigh internal stress may be generated to affect the adhesion strengthof a film. Nevertheless, in this present study, the plasma immersiontechnique is used to generate heat and carbon graphitization thusallowing the internal stress to relax. As a result, strong adhesion isachieved on the G0 film. For the Ti/C graded sample without a carbontopcoat, the adhesion strength of the carbon film is improved by theTi/C graded interface (G1 and G2 films). Hence, G0 film is the mostpromising coating with respect to mechanical properties.

The Raman spectrum (Fig. 10) of G1 graded film can bedeconvoluted into two sub-bands: The C5/C7 odd ring band at1548.15 cm−1 and the Csp3–Csp3 cluster band at 1354.46 cm−1 [46].In general, the bands at 1548.15 cm−1 and at 1354.46 cm−1 can beassigned for G and D bands respectively. However, it has to be empha-sized, that Raman spectrum can be significantly up- or downshifted bystress and that observed Raman bands and peaks do not always corre-spond to what was originally assigned. The deposited G1 graded filmis obviously not very soft and not plastic when considering bendingtest results (Figs. 7 and 9). Therefore, the Raman spectrummust be con-sidered as up-shifted by compressive stress. The broader band observedat 1548.15 cm−1 then cannot correspond to a G band, but mainly toC5/C7 odd ring bands caused by ion irradiation. The observed1354.46 cm−1 band has here also to be considered as stress up-shifted. Therefore, it will not correspond to an A-edge graphenicmode, but to an up-shifted disordered Csp3–Csp3 cluster band

(normally at ~1330 cm−1). In present case, there is only few hexagonalcyclic sp2 rings left (destroyed), and consequently no significantgraphenic A edge band can be observed and the lower side band ofthe C5/C7 band at 1548.15 cm−1 will merely correspond to danglingCsp2–Csp3 bonds (known to exist at 1470 cm−1 when not stressshifted). The C5/C7 odd ring band and the Csp3–Csp3 cluster band

(b)

(a)

Fig. 11. (a) Contact angle and (b) the polar component of surface energy of the stainlesssteel substrate (SS), and G0 Ti/C graded filmwithout a carbon topcoat and G1 Ti/C gradedfilm with a carbon topcoat.

(a)

(b)

(c)

G0 film

G1 film

Stainlesssteel

Fig. 12. AFM images: (a) Bare stainless steel, (b) G0 Ti/C graded film without a carbontopcoat and (c) G1 Ti/C graded film with a carbon topcoat.

Fig. 13. Optical density measurements from the HRP activity assay before and after 0.5%SDS detergent cleaning on the untreated and G0 plasma treated stainless steel surfaces.


observed from the Raman spectrum (Fig. 10) confirms the formation ofsp2/sp3 bonding in G1 graded film and further supports the XPS result(Fig. 5b).

The contact angle measurements (Fig. 11a) showed that the plasmamodified surface was more hydrophilic than the bare stainless steelsubstrate. The more hydrophilic surface and higher surface energy ob-served from the graded films (Fig. 11) suggest that the biocompatibilityof the surface was improved and the surface could better retain thefunction of immobilized protein than the untreated stainless steel [1].The AFM images (Fig. 12) showed a smother surface for the graded G0film than the rougher G1 topcoat surface. However there was no signif-icant difference between the contact angle and surface energy of thesetwo films. Biocompatibility and protein adhesion of a film surface notonly depend on the surface energywhich is a part of surface passivationaspects but also depend on stoichiometric distribution and geometricaltopography. A hydrophilic surface can adsorb water molecules prefer-entially and may form an interfacial hydrated complex providingroom for protein adsorption. Hence, both the G0 and G1 films havinghydrophilic surfaces can better retain the functions of the immobilizedprotein. However, the hydrophilic/hydrophobic characteristics areonly a part of the surface passivation aspects. In contrast to a uniformand dense amorphous carbon surface, a less dense carbon surfacecontaining different types of edges can be easily oxidized. Hence, theuniform and dense amorphous carbon surface, G1 graded film withcarbon topcoat, and less dense carbon surface, G0 graded film withoutthe carbon topcoat, can exhibit different surface passivation. Therefore,the G1 graded filmwith the carbon topcoat and G0 graded filmwithoutthe carbon topcoat show a different degree of biocompatibility and alsoin the ability of retaining the functionality of the immobilized protein.

The binding strength of the untreated and plasmamodified surfacesfor proteinmolecules adsorbed by incubation in a protein containing so-lutionwas investigated to determinewhether the proteinmolecules arephysically or covalently bound and whether bioactivity is retained.

Bioactivity is essential inmany applications of surface immobilized pro-teins. The G0 plasmamodified surface providedmuch better bioactivity(Fig. 13) of the immobilized protein, than the untreated stainless steel.The OD (Fig. 13) observed with treated surfaces was significantly largerthan with untreated surfaces, showing that the treated surfaces weremore densely covered with protein which retained its conformation


than the untreated surface. Bilek et al. [1], Bax et al. [24] and Yin et al. [2,30] used SDS washing to demonstrate covalent immobilization of pro-tein molecules on plasma treated surfaces. The SDS washing protocolused in these literature studies was more aggressive than that usedhere — their SDS (1–2% w/v) solution was heated to 70–90 °C. Underthese conditions SDS washing can completely elute proteins fromeven very hydrophobic surfaces such as polytetrafluoro ethylene(PTFE). This shows that the mechanism for SDS-resistant attachmentobserved on their more hydrophilic plasma treated surfaces than PTFEcannot be due to the hydrophobic interaction. While the SDS protocolused in this present work was not aggressive as used in the literaturestudies, statistical testing showed that the means of the protein activityon the treated samples were not significantly different before and afterthe SDS washing whereas the means of the protein activity on the un-treated samples before and after SDS washing were highly significantlydifferent (p b 0.0013). We propose that the demonstrated covalentbinding capability of plasma CVD deposited carbon from an acetyleneprecursor applies equally well to the carbon we have deposited by ourplasmaCVDprocess sincewe have used the same precursor. The advan-tage of our less aggressive SDS protocol was that it was aggressiveenough to distinguish clearly between the weaker protein binding onan untreated surface from the stronger binding on our treated surfacewhile preserving the function of the protein so that we can test for itsfunctionality at the same time [13,44]. The analogy between our workand that of Yin et al. [30,31] is particularly strong. Yin et al. demonstrat-ed that acetylene plasma deposition produces carbon topcoat having avery large density of attachment sites for strong, robust covalentimmobilization of a dense monolayer of protein that retained about80% of the deposited protein after aggressive SDS washing [2].

Protein plasma treating the substrate surface promoted endothelialcell attachment and proliferation. Hence the layer deposited usingacetyleneplasma in this study provided a platform for the robust attach-ment of functional protein molecules. This indicates that physisorptioncannot be responsible for the robust protein attachment observed onthe plasma treated surfaces (G0, Fig. 13) and that a covalent bond be-tween an active group on the carbon surface and an exposed sidechain of the protein was formed. The ability to covalently immobilizeproteins on a hydrophilic surface is a key advance that retains proteinconformation and bioactivity. Bilek et al. [1] proposed that the forma-tion of a free radical density n0 of about 6 × 1025 m−3 on the plasmatreated surface is responsible for the covalent binding. The plasmamodified surface generates layers having unpaired electronswhich pro-vide a universal protein binding platform and yields sufficient electronmobility within the layer. The unpaired electrons present in the freeradicals can covalently link with amino acid residues present in theprotein molecules. However, the addition of elements that suppressthe π conjugation in carbon structures diminishes the mobility of theunpaired electrons. Bilek et al. [1] observed this phenomenon that dra-matically reduction of covalent binding capability in plasma treatedsamples after the addition of oxygen, hydrogen or stainless steel inclu-sions. In some polymers like polydimethylsiloxane (PDMS), absence ofcovalent binding capability is observed because of the presence ofhigh concentrations of silicon. Hence, it is inferred that the covalentbinding capability is directly proportional to the concentration ofunpaired mobile electrons on carbon sites. Rapid covalent binding ofprotein molecules with a higher cell density on the plasma modifiedsurface was observed by Tran et al. [42]. Gan et al. [47] observed cova-lent binding of proteinmoleculeswith 100% coverage on plasma treatedsurfaces. Waterhouse et al. [15] observed good covalent binding ofprotein molecules with low thrombogenicity on plasma treated stain-less steel coronary stents. Ho et al. [44] observed improved binding ofprotein molecules while its function on plasma treated surface isretained. Nosworthy et al. [13] observed good covalent binding of pro-tein molecules with better retention of conformation on plasma treatedsurfaces. These results suggest that a high degree of covalent binding ofprotein, which retained its functional state, was achieved on a carbon

containing plasma modified stainless steel surface (G0, Fig. 13). G0film offers good protein adhesion as well as considered as the mostpromising coating with respect to mechanical properties. The gradedTi–C interface provided strong adhesion between the metal surfaceand deposited layers and was essential to avoid coating failure whenmechanical stresswas applied to the substrate. The outcomewas a coat-ing system that can provide covalent binding of protein to a metal stentfor cardiovascular applications. Such a stent could be functionalizedwith a suitable protein to enhance the biocompatibility.

5. Conclusion

It was demonstrated that the deposition of a Ti/C composite gradedfilm by cathodic arc deposition combined with plasma immersion ionimplantation and the subsequent deposition of an amorphous carbonfilm by PACVD produced a coating which bonded organic molecules.The molecules retained their functionality after washing. The gradedcoating was shown to have good adhesion even when subjected tomechanical stress.

Acknowledgments

We acknowledge funding from the Australian Research Council,Hong Kong Research Grants Council (RGC), General Research Funds(GRF) No. 112212, and City University of Hong Kong Strategic ResearchGrant (SRG) No. 7004188. The authors acknowledge the facilities andscientific and technical assistance of the Australian Microscopy &Microanalysis Research Facility at the Australian Centre for Microscopy&Microanalysis at the University of Sydney aswell as Dr. CenkKocer, DrPatrick Trimby and Mr. David beech for their assistance.

References

[1] M.M.M. Bilek, D.V. Bax, A. Kondyurin, Y. Yin, N.J. Nosworthy, K. Fisher, A.Waterhouse,A.S. Weiss, C.G. dos Remedios, D.R. McKenzie, Proc. Natl. Acad. Sci. U. S. A. 108 (2011)14405–14410.

[2] Y. Yin, S.G. Wise, N.J. Nosworthy, A. Waterhouse, D.V. Bax, H. Youssef, M.J. Byrom,M.M.M. Bilek, D.R. McKenzie, A.S. Weiss, M.K.C. Ng, Biomaterials 30 (2009)1675–1681.

[3] G.S. Wilson, R. Gifford, Biosens. Bioelectron. 20 (2005) 2388–2403.[4] F. Lisdat, F.W. Scheller, Anal. Lett. 33 (2000) 1–16.[5] J.S. Seo, S. Lee, C.D. Poulter, J. Am. Chem. Soc. 135 (2013) 8973–8980.[6] S. Petersen, A. Strohbach, R. Busch, S.B. Felix, K.P. Schmitz, K. Sternberg, J. Biomed.

Mater. Res. B Appl. Biomater. 102 (2014) 345–355.[7] S. Lal, R. Lui, L. Nguyen, P. Macdonald, G. Denyer, C. dos Remedios, Proteomics 4

(2004) 1918–1926.[8] G.J. Wilson, G. Nakazawa, R.S. Schwartz, B. Huibregtse, B. Poff, T.J. Herbst, D.S. Baim,

R. Virmani, Circulation 120 (2009) 141–149.[9] T. Inoue, R. Sohma, T. Miyazaki, Y. Iwasaki, I. Yaguchi, S. Morooka, Am. J. Cardiol. 86

(2000) 1057–1062.[10] G. Ohqvist, G. Settergren, S. Lundberg, Scand. J. Thorac. Cardiovasc. Surg. 15 (1981)

257–262.[11] M.M.M. Bilek, D.R. McKenzie, Biophys. Rev. 2 (2010) 55–65.[12] X. Liu, P. Chu, C. Ding, Mater. Sci. Eng. R: Rep. 47 (2004) 49–121.[13] N.J. Nosworthy, A. Kondyurin, M.M.M. Bilek, D.R. McKenzie, Enzym.Microb. Technol.

54 (2014) 20–24.[14] D.V. Bax, R.S. Tipa, A. Kondyurin, M.J. Higgins, K. Tsoutas, A. Gelmi, G.G.Wallace, D.R.

McKenzie, A.S. Weiss, M.M.M. Bilek, Acta Biomater. 8 (2012) 2538–2548.[15] A. Waterhouse, S.G. Wise, Y.B. Yin, B.C. Wu, B. James, H. Zreiqat, D.R. McKenzie, S.S.

Bao, A.S. Weiss, M.K.C. Ng, M.M.M. Bilek, Biomaterials 33 (2012) 7984–7992.[16] M.M.M. Bilek, Appl. Surf. Sci. 310 (2014) 3–10.[17] M.I. Jamesh, G. Wu, Y. Zhao, D.R. McKenzie, M.M.M. Bilek, P.K. Chu, Corros. Sci. 82

(2014) 7–26.[18] G. Wu, X. Zhang, Y. Zhao, J.M. Ibrahim, G. Yuan, P.K. Chu, Corros. Sci. 78 (2014)

121–129.[19] Y. Zhao, M.I. Jamesh, W.K. Li, G. Wu, C. Wang, Y. Zheng, K.W.K. Yeung, P.K. Chu, Acta

Biomater. 10 (2014) 544–556.[20] G.P. Lopez, B.D. Ratner, C.D. Tidwell, C.L. Haycox, R.J. Rapoza, T.A. Horbett, J. Biomed.

Mater. Res. 26 (1992) 415–439.[21] M. Karlsson, J. Ekeroth, H. Elwing, U. Carlsson, J. Biol. Chem. 280 (2005)

25558–25564.[22] D. Kiaei, A.S. Hoffman, T.A. Horbett, J. Biomater. Sci. Polym. Ed. 4 (1992) 35–44.[23] D.V. Bax, D.R. McKenzie, A.S. Weiss, M.M.M. Bilek, Biomaterials 31 (2010)

2526–2534.[24] D.V. Bax, D.R. McKenzie, A.S. Weiss, M.M.M. Bilek, Acta Biomater. 5 (2009)

3371–3381.

http://refhub.elsevier.com/S0257-8972(14)01013-5/rf0005












































[25] B. Liu, T.F. Zhang, B.J. Wu, Y.X. Leng, N. Huang, Surf. Coat. Technol. (2014).[26] C. Wei, W.-J. Pan, M.-S. Hung, Surf. Coat. Technol. 224 (2013) 8–17.[27] W.J. Ma, A.J. Ruys, R.S. Mason, P.J. Martin, A. Bendavid, Z. Liu, M. Ionescu, H. Zreiqat,

Biomaterials 28 (2007) 1620–1628.[28] S.C.H. Kwok, W. Zhang, G.J. Wan, D.R. McKenzie, M.M.M. Bilek, P.K. Chu, Diam. Relat.

Mater. 16 (2007) 1353–1360.[29] A. Grill, Diam. Relat. Mater. 12 (2003) 166–170.[30] Y.B. Yin, K. Fisher, N.J. Nosworthy, D. Bax, S. Rubanov, B. Gong, A.S. Weiss, D.R.

McKenzie, M.M.M. Bilek, Plasma Process. Polym. 6 (2009) 658–666.[31] Y. Yin, N.J. Nosworthy, H. Youssef, B. Gong, M.M.M. Bilek, D.R. McKenzie, Thin Solid

Films 517 (2009) 5343–5346.[32] D.P. Dowling, P.V. Kola, K. Donnelly, T.C. Kelly, K. Brumitt, L. Lloyd, R. Eloy, M. Therin,

N. Weill, Diam. Relat. Mater. 6 (1997) 390–393.[33] P.W.K. Zolynski, A. Kaluzny, Z. Has, P. Niedzielski, S. Mitura, J. Chem. Vap. Depos. 4

(1996) 232–239.[34] R.L. Boxman, V. Zhitomirsky, B. Alterkop, E. Gidalevich, I. Beilis, M. Keidar, S.

Goldsmith, Surf. Coat. Technol. 86 (1996) 243–253.[35] I. Zukerman, V.N. Zhitomirsky, G. Beit-Ya'Akov, R.L. Boxman, A. Raveh, S.K. Kim, J.

Mater. Sci. 45 (2010) 6379–6388.[36] V.N. Zhitomirsky, I. Grimberg, L. Rapoport, R.L. Boxman, N.A. Travitzky, S. Goldsmith,

B.Z. Weiss, Surf. Coat. Technol. 133–134 (2000) 114–120.

[37] S.I.V.I.I. Aksenov, V.G. Padalka, V.E. Strel'Nitskii, V.M. Khoroshikh, Sov. Phys. Tech.Phys. 25 (1980).

[38] D.R. McKenzie, D. Muller, B.A. Pailthorpe, Z.H. Wang, E. Kravtchinskaia, D. Segal, P.B.Lukins, P.D. Swift, P.J. Martin, G. Amaratunga, P.H. Gaskell, A. Saeed, Diam. Relat.Mater. 1 (1991) 51–59.

[39] P.C.T. Ha, D.R. McKenzie, M.M.M. Bilek, E.D. Doyle, D.G. McCulloch, P.K. Chu, Surf.Coat. Technol. 200 (2006) 6405–6408.

[40] P.C.T. Ha, D.R. McKenzie, M.M.M. Bilek, S.C.H. Kwok, P.K. Chu, B.K. Tay, Surf. Coat.Technol. 201 (2007) 6734–6736.

[41] I. Horcas, R. Fernández, J.M. Gómez-Rodríguez, J. Colchero, J. Gómez-Herrero, A.M.Baro, Rev. Sci. Instrum. 78 (2007).

[42] C.T.H. Tran, A. Kondyurin, S.L. Hirsh, D.R. McKenzie, M.M.M. Bilek, J. R. Soc. Interface9 (2012) 2923–2935.

[43] P.D. Josephy, T. Eling, R.P. Mason, J. Biol. Chem. 257 (1982) 3669–3675.[44] J.P.Y. Ho, N.J. Nosworthy, M.M.M. Bilek, B.K. Gan, D.R. McKenzie, P.K. Chu, C.G. dos

Remedios, Plasma Process. Polym. 4 (2007) 583–590.[45] D.R. McKenzie, Rep. Prog. Phys. 59 (1996) 1611.[46] S. Neuville, Carbon Structure Analysis With Differentiated Raman Spectroscopy, LAP

Academic Publishing, Saarbrucken Germany, 2014. 1–176 (ISBN-13: 978-3-659-48909-9).

[47] B.K. Gan, A. Kondyurin, M.M.M. Bilek, Langmuir 23 (2007) 2741–2746.











































surface & coatings technology - city university of hong … adhesion of biologically active...

Documents