Glycine Adsorption onto DLC and N-DLC Thin Films Studied by XPS and AFM∗

An understanding of protein adsorption to surfaces of materials is required for the control of biocompatibility and bioactivity. Amorphous carbon, commonly known as diamond-like carbon (DLC) is reported to have excellent biocompatibility. Hydrogenated amorphous-carbon thin films (DLC) and nitrogen doped a-C:H thin films (N-DLC) were prepared by plasma-enhanced chemical vapour deposition (PECVD). Glycine adsorption onto the surface of the films was investigated in order to aid in the elucidation of the mechanisms involved in protein adhesion. The physicochemical nature of the surfaces, before and after adsorption of glycine, was analysed using X-ray photoelectron spectroscopy (XPS) and atomic force microscopy (AFM). The XPS spectra highlighted a slight increase the ratio of sp/sp at low levels of N (5.4 atom %) whilst increasing the nitrogen dopant level (> 5.4 atom %) resulted in a decrease of the sp/sp ratio. Following exposure to solutions containing glycine, the presence of peaks at 285.0 eV, 399 eV and 532 eV indicated the adsorption of glycine to the surfaces with a quantitative change in the amount of C, N and O on the surfaces. Glycine was bound to the surface of the DLC films via both de-protonated carboxyl and protonated amino groups while, in the case of N-DLC gylcine was bound to the surface via anionic carboxyl groups and the amino group did not interact strongly with the surface. AFM analysis showed a change in surface roughness of the films with the ratio of rms values increasing following exposure to glycine. These results show that low levels of nitrogen doping in DLC enhances the adsorption of the amino acid, while, increased doping levels (> 5.4 atom %) led to a reduced adsorption, as compared to undoped DLC. Doping of DLC may allow control of protein adsorption to the surface. [DOI: 10.1380/ejssnt.2009.217]


I. INTRODUCTION
When a foreign material is placed within the human body, the surface of the material first makes contact with blood and reacts with the components which are mainly proteins and initial adsorption to the surface occurs in a matter of seconds.The response of the biological compounds and living cells towards any foreign material will depend on the surface characteristics of the material.Therefore, surface modification of implant materials is important for improved bioactivity and biocompatibility.
Amorphous carbon, commonly referred to as diamondlike carbon (DLC) is an excellent candidate for a biomaterial, due to its chemical inertness, high hardness, high wear and corrosion resistance, low frictional coefficient, high electrical resistivity, infrared-transparency, high refractive index and low surface roughness [1][2][3].There are many published research articles concerning the interaction of DLC coatings with blood and cells [4].It has been reported that DLC has excellent hemacompatibility and anti-thrombogenicity properties [5][6][7].Furthermore, it is reported that DLC has good biocompatibility in contact with blood monocytes which, control inflammatory reactions and can affect the osteointegration of implants.DLC coatings are reported to be non-toxic toward several living cell lines and no inflammatory response has been reported [8,9].All of these properties match well with the criteria of a good biomaterial coating for use on implants which come in contact with blood, such as orthopaedic, cardiovascular and artificial heart valves, contact lenses, stents or dentistry [10][11][12].
The properties of DLC may be modified by the introduction of dopant elements e.g.nitrogen [13].Nitrogen doped DLC (N-DLC) films are also good candidates for use as biocompatible coatings due to their chemical composition i.e. containing just carbon, nitrogen and hydrogen [14,15].The N-DLC coatings have comparable hemocompatibility with DLC coatings [4].It has been reported that N doping of DLC reduces the internal stress and poor adhesion which has been observed with undoped DLC films, however, N doping can reduce the hardness of the films [16,17].It has been reported that the cell adhesion percentage for nitrogen-doped DLC was enhanced more than 10% in comparison with that for a non-doped DLC [18].Experimental results have confirmed that DLC and N-DLC films possess good hemocompatibility compared with other metallic bio-materials such as NiTi alloy [15].In general DLC coatings modified with N or Si have been reported to perform better than the pure DLC coatings for inhibiting bacterial adhesion [16].
To date, only a few papers have been published concerning the attachment of amino acids onto to DLC surfaces.The current study is aimed at investigating the adsorption of the amino acid, L-glycine, on DLC and N-DLC films.Glycine (H 2 NCH 2 COOH) has been chosen because it is the simplest of the amino acids, it is ambivalent and exists as a zwiterion (pK a COOH = 2.35, pK b NH + 3 = 9.78) [19][20][21].The purpose of this study was to obtaining a better understanding of protein-surface interactions by investigating the interaction of a single amino acid to simplify the system given that proteins are complex and diverse molecules which are difficult to analyse [22].A range of DLC films were prepared using plasma enhanced chemical vapour deposition (PECVD).The attachment of L-glycine on the surfaces was analysed using X-ray photoelectron Spectroscopy (XPS) and Atomic Force Micro-  scope (AFM).

A. Film deposition
A silicon wafer was cleaned with a commercial cleaning product (Blazers substrate cleaner solution A (code BD481900) followed by solution B) and then dipped in acetone for 5 minutes, rinsed twice in distilled water and finally dried in a flow of nitrogen gas.DLC and N-DLC films were prepared by PECVD using a Diavac RF 13.56MHz PECVD system with negative electrode selfbias voltages set at 400 V. Deposition was carried out using a argon/acetylene mixture for the formation of DLC films under plasma glow discharge.Nitrogen gas was used introduced for N-DLC.The ratio for argon and acetylene was varied by altering the gas flow and different levels of N doping was achieved by altering the N gas flow.Prior to the deposition stage, an argon etch was used to clean surface substrates.A summary of the key deposition parameters used are given in Table I.
Stylus profilometry was used to determine film thickness (Dektak 8 Advanced Stylus Profiler Veeco Instruments Inc., USA).Automatic levelling was selected and system based software performed average step height (ASH) calculations (Table I).

B. Preparation of Glycine solution
Glycine (HOPKIN & WILLIAMS LTD) was dissolved in aqueous phosphate buffer saline (SIGMA) to give 0.01M solution at pH 7.4.The DLC and N-DLC coated samples were immersed in a Petri-dish containing of 25 cm 3 of glycine solutions (one separate Petri-dish for each sample).Incubation was carried out at 37 • C for 16 h in a shaker.Following incubation, samples were washed twice with distilled water and dried in a flow of nitrogen gas.

C. Characterisation of samples
XPS measurements were carried out using a KRATOS XSAM 800 equipped with an energy analyser.The X-ray source employed was an Mg Kα at 15 keV and 20 mA.The base pressure within the spectrometer during examination was ∼8×10 −10 bar.All spectra were referenced by setting the hydrocarbon C 1s peak to 285.0 eV to compensate for residual charging effects.

D. Surface roughness of samples using Atomic
Force Microscopy (AFM) The surface morphology of the films was analysed by AFM using a Dimension TM 3100 (Veeco Metrology group).All imaging was performed at room temperature in air.A root mean square surface roughness and particle sizes were derived directly from AFM height images.

III. RESULTS AND DISCUSSION
The DLC and N-DLC films were prepared with thicknesses ranged from 175 to 225 nm.In case of N-DLC, small amount of nitrogen content, has not significant affects on the film growth rates, whilst, observed that film thickness was increased with increasing nitrogen content.The details of the preparation conditions of the films are given in Table I  was used to analyze the surface composition films specifically for carbon, oxygen and nitrogen, before and after exposure to glycine.The XPS spectra for DLC films showed the main peaks for C 1s at 285 eV, O1s at 531 eV and N 1s at 400 eV (for N-DLC samples).Analysis of the C1s peak verified carbon for the DLC film with full width-half maximum (FWHM) of 1.6 eV, which was de-convoluted into three Gaussian peaks.The first was centred at 284.7 eV which is attributed to sp 2 carbon (C=C), the second was centred at 285.2 eV is belong to sp 3 bulk carbon (C-C) and the third peak was centred at ∼287.1 eV corresponding to (CO) bond (see Fig. 1).The atomic percent of carbon was determined to be 92.4%.
In the case of N-DLC, de-convolution of the C 1s band gave four peaks centred at 284.8 eV, 285.5 eV, 286.6 eV and 287.2 eV which correspond to sp 2 hybridized carbon  (C=C), sp 3 hybridized carbon (C-C), carbon nitrogen bonds (C-N), and the bonding states of C to O, respectively, (Fig. 1) [22].The main difference to be noted is the C 1s band was slightly shifted toward lower binding energy with broadening of the FWHM from 1.6 eV to ∼2.4 eV, following nitrogen doping.
Figure 2 shows the N 1s band at ∼400 eV following nitrogen doping with FWHM around (2.3 eV).Deconvolution of this band yields three peaks centred at ∼399 eV, ∼400 eV and ∼401 eV which, can be attributed to C-N, N=C and (N-N or N-O) bonds respectively.The N 1s band area increased with increasing partial pressure of (N 2 ) in the chamber, during film formation.Furthermore, the ratio of C=N to C-N was increased with increasing nitrogen flow.
It also can be seen that N 1s band are slightly shifted towards lower binding energy with increasing nitrogen content, which can be assigned to nitrogen bonded to carbon (Table II).This was attributed to formation of C=C and C=N bonds with increasing nitrogen flow rate.Accordingly, the sp 3 /sp 2 ratios of the DLC and N-DLC films can be determined from the relative intensities of the C1s spectra of the two groups of photoelectrons separated by a curve-fitting technique.The peak at a lower binding energy (∼284.7 eV) is attributed to the sp 2 carbon hybridizations, while that at a high binding energy (∼285.2) is attributed to the sp 3 carbon configurations.The sp 3 /sp 2 ratio of DLC is obtained to be ∼54%, appropriating into account the area ratio of the sp 3 to sp 2 band.Thus, changes in the sp 3 /sp 2 ratio were observed depending on the level of (N 2 ) doping. Figure 3 shows the sp 3 /sp 2 ratio versus the measured nitrogen atom %.A slight increase was observed on going from 0 to 5.4% N-atom and then the ratio decreased with increasing nihttp://www.trogen content to 44% in case of N3 sample.This due to nitrogen will form double bonds with carbon in the film structure.This correlates with the results of other researchers [23,24].The other main peak at around 531.4 eV observed in the DLC, corresponds to O 1s.The de-convolution of this band gives two sub peaks, which assigned to O=C and O-C bonds.The presence of nitrogen in the film leads to shift the spectrum toward lower binding energy due to formation of N-O bond (Fig. 4).This result is in agreement with other results [25][26][27].
Following exposure to glycine, changes in the carbon, oxygen and nitrogen content of the films was observed.Figure 5(A) shows the de-convolution of the C 1s band folloing the adsorption of glycine onto the DLC film with broader of FWHM (2.3 eV).Five major peaks were observed i.e.; the lower binding energy is centred at 284.5 eV and corresponds to the acetate group (COOC) [28]; the second peak is observed at ∼285.3 eV which attributed to aliphatic carbon (C-C) in the glycine molecule [20]; the peak at 286.2 eV correlates with the C-N bond [29]; the peak at 287.1 eV is due to (C-O); and the peak at 288.5 eV is due to carbonyl (C=O) group [28,30,31].Also small peak was observed at ∼284.8 eV could be attributed to C=C of the sample.
In case of the N1 sample (Fig. 5(B)) the C 1s band is de-convoluted to five major peaks centred at 284.For the N3 sample, the FWHM was 2.0 eV and peaks were centred at 284.5 eV, 285.3 eV, 286.4 eV, 287.4 eV and 288.6 eV following adsorption of glycine (Fig. 5(D).The appearance of band at ∼287.1 eV indicates the presence of C-O bond which proves the interaction of carboxyl group of glycine with the DLC and N-DLC surfaces.On other hand, slight lower shift of peak at ∼284.5 eV with reduced of the band area at 284.8 eV are good evidence that happened a changes on the surface structures followed by adsorption of glycine.The area of the sub peak at 284.5 eV which corresponds to the ester group (O=COC) is significantly decreased with high nitrogen content of the films (Table III).Also small peak appeared at ∼284.8 could be remains of sp 2 bond which comes from substrate, and this peak area well increased with increasing of nitrogen content.As well, researchers reported that nitrogen doping DLC, gives rise to an increase in the polarity of the surface and high surface tension with lower contact angle [32].These suggest that a high ratio of nitrogen in the film surface leads to a decrease in the adsorption of glycine.
The increase in intensity of the N 1s peak following the adsorption of glycine (as compared between Figs. 6 and 2), suggests that glycine is bound to the surfaces of DLC and N-DLC films.The N1s peak revealed three types of nitrogen-binding state, corresponding to the carbonnitrogen(C-N) bond, deprotonated (NH 2 ) and protonated (NH + 3 ) amine groups.Figure 6(A) shows the deconvoluted N 1s peak of the DLC following adsorption of glycine.The peaks are centred at 399.4 eV (C-N), 400.3 eV (NH 2 ) and 402.4 eV for (NH + 3 ) which is characteristic for the zwitterionic glycine (with FWHM 2.9 eV for N1s band).
For N-DLC samples, also the de-convolution of the N 1s band yielded three peaks of N1s, centred at ∼398.8 eV, ∼400.4 eV, and ∼402.4 eV (for N1 and N2) samples with FWHM value of 2.9 eV.Whilst in case of N3 sample, de-convolution of the N 1s band yielded three bands centred at 398.8 eV, 400.3 eV and 401.7eV with FWHM equal to 2.5 eV.The intensity of the amino group band at ∼400 eV indicates the adsorption of glycine onto the surfaces through a carboxyl group.For the DLC sample, the band location at 399.4 eV is maybe due to the interaction of some amino groups (NH 2 ) with the surface of the sample as compared to the N-DLC samples which appears at around 398.8 eV.The dominating high binding energy peak at 402.4 eV is evidence that the glycine was adsorbed onto surface via carboxyl group, also high intensity of this band could be assigned to formation of multilayer (Fig. 9) [32].In the case of N2 the intensity of peak at 402.4 eV is reduced, whereas there is a shift of the NH + 3 band to 401.7 eV for the N3 sample (Fig. 6(D)) and a decrease of band area (Table III).This indicates that the amount of glycine adsorption is reduced for the N3 sample [33].This is clear evidence that glycine adsorbs predominantly in anionic form (COO − ) in the case of nitrogen doped DLC [33].These results are agreed with our previous investigation, that absence of Raman shift at 3104 cm −1 which is belongs to protonated amine group [34].In addition, the nonexistence of a binding energy shift for both NH 2 and NH + 3 components in the XPS spectra for glycine adsorption suggests that the amino group does not interact strongly with the surfaces.
Following the adsorption of glycine, the XPS spectra (Fig. 7) showed the bands of the O1s centred at 532.1 eV for DLC, 531.9 eV for (N1), 531.4 eV for (N2) and 531.0 eV for (N3), with significant shifting toward higher binding energy for the N-DLC.Results are agree with Hasselstrom et al who reported that the presence of carboxylic carbon atom led to shift of the O1s peak toward higher binding energy with a similar shift for the methyl group and a partially shift found for methyl amine in the C1s peak [28].
Figure 7, shows the de-convolution of the O 1s peak for glycine adsorbed onto DLC and N-DLC samples.The peaks are centred at ∼531.5 eV and ∼532.1 eV and can be attributed to C=O and C-O respectively (Table III).The binding energy position observed at 532.1 eV is in good agreement with the results of Lofgren et al. which confirmed that glycine is interacted with the surface in zwitterionic form.According to literature, the shifting of the O 1s toward higher bending energy is confirmed following multilayer adsorption of glycine onto surface, and the somewhat asymmetric oxygen peak with the central binding energy position is in good agreement that glycine was bonded by a deprotonated carboxyl group [32].
The surface morphology of the DLC and N-DLC films was analysed by AFM.The roughness of a surface is generally reported as a rms (root mean square) value.From Fig. 8, we can see that, DLC surface is flat with rms value of ∼0.405 nm.Whilst, the rms, slight increased with increasing of nitrogen content (Table IV).These results are in agreement with Tian et al., who reported that nitrogen doped DLC film causes increase of surface roughness of the samples [35].
The AFM image analysis showed that both rms and R a values increased following the adsorption of glycine onto DLC.Also after attachment of glycine, the differences in morphology and roughness of N-DLC surfaces were detectable.Results from Table IV and Fig. 8 showed that significant changes in the rms values, after attachment of glycine [16,33].These results agreed with researches which reported on the adsorption of amino acids onto surfaces [36].From AFM images which showed in Fig. 8, confirms non-uniform distribution of glycine on the surfaces.
From the results, we can propose a model of glycine adsorption on to DLC and N-DLC samples where by the amino acid is chemically bonded to the surfaces mainly via the anionic carboxyl group (COO − ) and the amino group does not play a significant role in the initial monolayer formation [34].Whilst amino group could be contribute via hydrogen bonds which occurs between (NH + 3 and COO − ) groups to form a multilayer of adsorbed glycine onto surfaces (Fig. 9).Our results show that a low level of nitrogen doping (< 5.4%) in DLC led to an increase in the adsorption of the amino acid, while, increased doping levels (> 5.4%) resulted in a reduction in the amount of adsorption, as compared to undoped DLC.Therefore, doping of DLC may allow control of protein adsorption to the surface.

IV. CONCLUSIONS
Glycine interaction onto DLC and N-DCL was studied using XPS and AFM.XPS is a powerful tool for the analysis of adsorption of bio-molecules to surfaces and has allowed a detailed elucidation of the mechanism of glycine adsorption to DLC and N-DLC films.Doping the DLC films with nitrogen above (5.4atom %) resulted in a decrease of the sp 3 /sp 2 ratio.The XPS analysis indicated that the glycine was bonded to DLC and N-DLC mainly via the carboxylate group and the amino group did not play a significant role in the adsorption of glycine on to samples which contain (> 5.4 nitrogen atom %) .These results show that low level of nitrogen doping (< 5.4%) in DLC led to an increase in the adsorption of the amino acid, while, increased doping levels (>5.4 N atom %) resulted in a reduction in the amount of adsorption, as compared to undoped DLC.From AFM images confirmed that non-uniform adsorption of glycine on the surfaces.However, for both DLC and N-DLC, multilayer adsorption occurs via hydrogen bonding between the COO − and the NH + 3 groups of the glycine.Doping of DLC with particular elements may allow better control of protein adsorption to the surface which is important for biomedical applications.

FIG. 1 :
FIG.1: XPS spectra of C 1s for DLC and N-DLC films before attachment of glycine.

FIG. 2 :
FIG.2: XPS spectra of N 1s for N-DLC samples before attachment of glycine.

FIG. 4 :
FIG. 4: XPS spectra O 1s showing shift in binding energy following N doping.

TABLE I :
PECVD process parameters employed for the deposition DLC films.

TABLE II :
XPS analysis before and after glycine adsorption.

TABLE III :
The measurement C 1s , O 1s and N 1s spectra can be resolved into their spectral components through curve fitting.The best fit parameters (mean values) are given, as well as the signal emitters, in their chemical environments.

TABLE IV :
Roughness values (RMS and Ra) for surface of samples before and after attachment of glycine.RMS (or Rq) is the roughness based on a least square calculated with the best fit of the height points.Ra is obtained by a logarithm which measures the average deviation between the peaks and values from the mean line of the surfaces.± represent the values calculated standard deviation.