Printer Friendly

Diamond-Like Carbon Coating for Corrosion Protection of Metallic Implants.

Introduction

The protection of metallic implants from in vivo corrosion has always received serious attention in the area of artificial organs development [1,2]. There are numerous implantable devices made with metals and alloys, the most popular being orthopedic fixation devices (like supporting bone plates and screws), articulating devices (like prostheses for hip joints, knee joints, elbow joints and shoulder joints) [3], cardiovascular stents [4], ureteral stents [5] and endosseous dental implants [6]. The high modulus and yield point coupled with ductility make metals inevitable for such applications. Certain surgically implantable metals and alloys such as low-carbon stainless steel, cobalt chromium alloy and titanium, having biocompatibility and corrosion resistance have been used conventionally.

The conventional metallic materials are well accepted by the medical professionals for implantation, and they showed high success rates in usage. However, the loadbearing metallic implants are always under the threat of failure, mainly due to corrosion [1,2]. All the surgically implantable metals undergo chemical or electrochemical dissolution at some finite rate, due to the complex and corrosive environment of the human body. The body fluid is aqueous electrolyte having dissolved oxygen and large amounts of sodium and chloride ions are present along with numerous ions, complexes and biomolecules. It poses intensive corrosive environment for a metal, which can attack even the most corrosion-resistant materials. Movement or stress associated with the functioning of the device will aggravate the problem, with increased amount of corrosion and/or wear debri [7].

The threat of implant corrosion is two-pronged: (i) It leads to mechanical defects and weakening of the implant, and eventually to the failure; (ii) The degradation products from the implant may elicit local and systemic responses [8]. Any failure of the implant may prove heavily taxing to the patient. In each orthopaedic implant failure, the patient faces the trauma of repeated surgeries and has to bear the additional expenses. The corrosion products can be the leaching ions, their complexes (through the local interaction with body fluid) and/or the dislodging particles. The local tissue response may range from mild edema to chronic inflammation and alteration in the local tissue structures [7,8].

The corrosion resistance of metals could be enhanced by bulk alloying, and more effectively by modifying the surface through ion implantation and coating [9]. Coatings of adherent ceramic layers through vapour deposition techniques have been proved more cost-effective and viable. Coatings for the corrosion protection of implants should have certain essential qualities like high hardness, low friction and biocompatibility to both blood and tissues. Only two materials have been identified to date to possess this particular combination of properties, namely Titanium Nitride (TiN) and Diamond-like Carbon (DLC) [10]. Among them, DLC is preferred because it contains only carbon and hydrogen and offers the best possible blood compatibility [10]. Apart from the material properties, the coverage of the coating over the implant surface should be conformal and free from cracks and pin-holes. The coating should be adherent and have low level of crystallinity (or preferably amorphous) to avoid cracking and peeling. These properties are dependent on the method of coating and the arrangements in the coating equipment. TiN and DLC coatings have been commercialized in certain products like coronary stents. Ample clinical evidences are available for the advantage of these coatings, especially in the case of DLC [11].

There are large number of reports related to DLC coatings endorsing their high mechanical strength, high blood compatibility, low coefficient of friction and high corrosion resistance. Interestingly, the term "Diamond-like Carbon" refers to a class of materials with varying ratios of carbon species existing in different hybridized states along with hydrogen and showing corresponding variation in the properties [3,11]. The approach of classifying DLC is detailed in the reviews and it has been identified that the class "amorphous hydrogenated carbon (a-C:H)" is ideal for biomedical applications [3, 10, 11]. This class (a-C:H) is characterized by the range of ratios of carbon species ([sp.sup.2] and [sp.sup.3] hybridization) and the hydrogen content. DLC coatings in this defined range satisfy the requirements for mechanical strength, biocompatibility and corrosion resistance. Coatings with good adhesion to the implant surface could be produced viably through techniques like vapour deposition, particularly the plasma enhanced chemical vapour deposition (PECVD) [10, 11]. However, it is difficult to have a standardization, as the material composition and properties tend to vary with the coating technique, precursors, process kinetics and the design of the coating system [10, 11]. It demands a careful designing of the coating system and implementation of sufficient level of automation to generate viable and repeatable coating of DLC, which is crucial for biomedical applications.

The present work demonstrates the development of diamond-like carbon coating in dedicated equipment for the corrosion protection of metallic implants. A semi-industrial scale batch coating system has been used in this study, based on plasma enhanced chemical vapour deposition (PECVD) technique using radiofrequency (RF) source. The equipment is fitted with specially designed liquid precursor injection and streamlined vapour flow mechanisms, along with plasma shield for improved plasma control. Large process space and process automation are additional features. The system ensures process viability and batch-to-batch repeatability. Coatings were prepared on implantable grade austenitic stainless steel substrates and characterized for chemical composition, thickness, micromorphology, micro hardness, modulus, adhesion, frictional properties and corrosion susceptibility. The capability of the dedicated equipment to generate diamond-like carbon coatings for the surface modification and protection of metallic implants is delineated.

Experimental

Equipment

The PECVD system used for the study has been exclusively designed and fabricated by M/s. Isytech (Lannion, France). It consisted of a stainless steel high vacuum chamber with radiofrequency (RF) power coupling through a flat-plate electrode and other fittings for precursor flow and plasma confinement. The schematic representation of the set-up is shown in Figure 1.

The chamber is in flat-end cylindrical geometry with 60 cm height and 45.5 cm diameter, having front opening facility for easy loading of the samples. Water cooling channels are integrated on the outer surface and appropriate ports for pumping, viewing and gas/vapour injection are provided. The RF electrode is a 300 mm dia water-cooled stainless steel plate, fitted flat onto the base, insulated with PTFE. It is connected to a water cooled RF generator (13.56 MHz, 3 KW, Huttinger AHC) through an automatic matching network of 5 KW rating. A DC bias probe is connected to measure the self-bias potential acquired on the electrode with respect to the plasma, which is one of the parameters deciding the coating quality.

The cylindrical process space above the electrode in the PECVD chamber is encompassed by a demountable stainless steel shield to confine the plasma. The arrangement for gas/vapour mixing and dispersion is inserted through the top of the shield to ensure uniform distribution and streamlined flow downwards. This helps to keep the plasma stable and uniform in a large volume, not less than a size of 30 cm diameter and 40 cm height. It was observed that the carbon coating is formed uniformly and reproducibly throughout this process space.

The precursors for the process (cyclohexane and tetra methyl silane) are taken in liquid form in metal containers and injected to the chamber after vaporizing through warming. The process gases and vapours are transported to the mixer through mass flow controllers.

Vacuum pumping of the chamber is done with turbo molecular pump (TMP, speed 1100 l/s) backed by a rotary pump (speed 100 l/min) which can give an ultimate vacuum of 2 x [10.sup.-6] mbar. Gate valve and automatic throttle valve with integrated regulator are fitted prior to the TMP to regulate the process pressure. The chamber pressure is monitored using a Bayard-Alpert Pirani Combination Gauge and the process pressure, by a Baratron capacitance manometer.

The system is fully automated with Siemens SIMATIC S7-300 universal controller and customised Induscreen Software. The operation of the system could be done from the touch-screen control panel, wherein the status of the chamber and accessories are graphically shown. Individual process steps (duration, RF power/bias, individual flow rates of vapours and gases and the process pressure) could be programmed and combined in sequence to make a full process recipe for automatic execution.

Materials

Two precursors were used in the process, in liquid form: tetra methyl silane (Aldrich ACS Reagent, NMR grade, 99.9%) and cyclohexane (Aldrich ACS Spectrophotometric grade, >=99%). The plasma reaction products of cyclohexane helps to obtain DLC, and those of TMS provides an amorphous silicon based interlayer to improve adhesion. Two gases, hydrogen (99.999 purity) and argon (99.998 purity) were provided for in situ plasma cleaning.

The substrates for coating were machined out of implantable grade SS316L metal rods. Fine grinding of the surface was done using SiC abrasive sheets of grit size 320 and 400, followed by rough polishing with 3 micron diamond paste and fine polishing with 1 micron diamond paste. The mean surface roughness (Ra) measured was 0.08 [micro]m. The polished samples were initially rinsed with deionised distilled water, cleaned twice ultrasonically in Extran (soap solution) with intermittent washes in deionised distilled water, and dried. They were then subjected to vapor degreasing.

Coating Process

The coating process was initiated after loading the samples onto the holders which are electrically fastened to the RF electrode. The chamber will be initially pumped down to base pressure (5 X [10.sup.-5] mbar). A desired process could be programmed by setting the RF bias, flow rates, duration and process pressure appropriately and executed automatically.

Plasma cleaning: Cleaning of the substrates to atomic level is important because the remnant impurities or oxides can drastically degrade the adhesion of the coating. The sample surfaces were cleaned in the plasma (as a part of the process recipe) initially in hydrogen and then in argon, at 100 sccm flow and a bias of 550V, for 10 min. Hydrogen removes oxidative impurities [12] and argon helps to clean contaminants through surface sputtering [13].

Coating steps: The coating process was programmed in continuation to plasma cleaning. A silicon-based inter-layer was deposited at first by admitting the vapours of tetra methyl silane, to enhance the adhesion of the diamond-like carbon coating onto the metal. The carbon layer was deposited seamlessly over the inter-layer using cyclohexane. The reactant flow rates, bias potential and process pressure decide the quality of the coating and hence these values were optimized by systematic trials.

Process optimization: Obtaining deposition at high rates is desired in commercial production to have viability. However, fast deposition will reduce the integrity and mechanical strength of coating. The rate of deposition, in turn, depends on the combination of flow rates, bias potential and process pressure. The in situ surface cleaning (plasma etching) of the substrate decides the adhesion of the coating and the silicon containing interlayer takes care of the thermal expansion mismatches of the substrate with the deposited carbon layer.

Carbon deposition trials were done by changing the flow rates (12-20 sccm), bias potential (400V-500V) and process pressure (1 X [10.sup.-2] - 5 X [10.sup.-2] mbar), and the deposition rate and microhardness values were measured in each case to optimize the combination of parameters. The in situ etching parameters (gas sequence and flow rate, bias and etching time) and the interlayer thickness were optimized by testing the adhesion strength in each combination of carbon deposition parameters.

The optimized coating process steps for 1 [micro]m thick coating on SS316L substrates were as follows: (i) etching in hydrogen, with 100 sccm at 550 V for 5 min followed by etching in argon at the same parameters, (ii) deposition of interlayer using TMS with 20 sccm at 450 V for 10 min, and (iii) deposition of carbon layer using cyclohexane with 20 sccm at 450 V for 20 min. These steps were programmed in sequence with inter-steps and ramping flow and without any discontinuity in bias, to keep the plasma active. The process pressure was kept stable at 2X10-2 mbar after the etching step. Coatings obtained in this optimized process were subjected to characterization.

Characterization

Physico-chemical Testing

The chemical characteristics of the coating were analysed using infrared spectroscopy (FTIR), X-ray Photoelectron Spectroscopy (XPS) and FT-Raman Spectroscopy. The crystallinity has been checked in X-ray Diffractometry (XRD).

The Fourier Transform Infrared analysis helps to identify the hydrocarbon bonds in diamond-like carbon coating. Thin translucent pellets of spectroscopic grade KBr (0.15 g quantity) were made and carbon coating was made to a thickness 500 nm at the experimental conditions (without any interlayer). Spectrum was recorded on Thermo-Nicolet 5700 spectrometer in transmission mode in the range 4000 - 400 [cm.sup.-1] at a resolution of 4 [cm.sup.-1]. Pellet of KBr alone was used to record the background spectra.

X ray photoelectron spectroscopy (XPS) analysis of the coating surface was performed with SPECS GmbH Spectrometer (Phoibos 100, MCD Energy Analyser) using MgK alpha radiation (1253.6 eV). The analysis was done at the pressure [10.sup.-10] mbar. The spectrometer was calibrated using the photoemission lines of Ag ([3d.sup.3/2] of Ag corresponding to 367eV with reference to Fermi Level). Peaks were recorded with constant pass energy of 40 eV.

The Raman spectroscopy (FT Raman) is helpful in measuring the disorder of carbon in the coating, mainly the [sp.sup.2] bonds. DLC, generally gives the characteristic peaks around 1580-1600 [cm.sup.-1] (commonly known as G-peak, [sp.sup.2] mode) and around 1350 [cm.sup.-1] (commonly known as D-peak, disordered allowed zone edge mode). The coating was analysed in WITec alpha 300R Confocal Raman system. This uses a frequency doubled NdYAG laser of wavelength 532 nm which was polarized horizontally (in the x-direction). All measurements were performed using an UHTS300 (ultra-high throughput spectrometer) and back-illuminated CCD cameras. A 60x objective was used for imaging and measurements were carried out at an integration time of 3s.

The possibility of crystalline order in the film was tested using X-ray diffractometry (Bruker AXS, D8 Advance). 2 micron thick coatings made on the metal substrates were subjected to XRD at grazing incidence (2 degrees), with thin-film attachment. The range of diffraction angle was 10-90[degrees], at a speed of 1[degrees]/min.

Thickness and micromorphology

The 'Calotest' method is rather simple for thickness measurements and ideally suited for hard coatings. This is a destructive test in which a circular crater is made by abrasion and the size of the annular shaped exposed edge of the coating is measured. A rotating heavy metal ball smeared with an abrasive slurry is used for abrading the coating. The thickness is calculated through the geometrical relation connecting the dimensions of the annular ring and the diameter of the abrading sphere.

The carbon coatings obtained were tested in Calotest Compact machine (CSM Instruments) in which the abrasion was done with a steel ball of diameter 30 mm and ultra-fine diamond paste (average particle size of 0.1 im), at a speed of 1000 rpm for 3 min. Microscopic image was acquired digitally on a computer and calculations were done using software. The deposition rates were calculated by dividing the gross thickness with the deposition time.

The surface profiling of the coatings was done using Talysurf CLI 1000 (Taylor Hobson) 3D profiling system. This is a non-contact optical profilometry based on chromatic length aberration (CLA) technique in which the spectral dispersion of light beam over the surface is recorded and the data is interpolated to reconstruct as depth variation.

The CLA 300 gauge used for the test had a range 300 mm and resolution 10 nm. The scanning (or the sample stage motion) was done at a speed 200 mm/s with a sampling rate of 200 Hz, in an XY envelope of 500 mm length with 1 mm spacing. The Z axis measurement of the CLA gauge was calibrated using a secondary roughness standard (traceable to United Kingdom Accreditation Service). Line scan was extracted from the test data to calculate Average Roughness ([R.sub.a]), Root Mean Square Roughness ([R.sub.q]), and Maximum Profile Peak Height ([R.sub.p]) values. Area scan helped to construct the 3D image, to check the uniformity of the coating.

Micro hardness and modulus

The mechanical properties (microhardness and modulus) of the carbon coating were studied using CSM Micro-Combi Tester (CSM Instruments, USA). The coating surface was intended with Vickers diamond indenter tip (triangular pyramid, a = 136[degrees]) under a gradually increasing normal load (typically 100 mN/ min) at a desired depth. On reaching a preset depth, the load was held for 10 s and slowly removed allowing a partial or complete relaxation of the intended portion. The position of the indenter relative to the sample surface was continuously monitored with a differential capacitive sensor and plotted against the load. This was repeated in 3 x 3 matrix (each point at 250 mm separation) to have the average value. The Vickers' microhardness value and the Youngs' modulus were calculated from the load-displacement curves acquired. The indentation depth selection started from 0.2 mm, where the coating gives a meaningful value after recovery. The measurements were taken at depth intervals of 0.1 mm, down to 1 mm beyond which the substrate effects predominates.

Coating adhesion

Adhesion strength of the coating to the substrate (SS316L flat coupons) was investigated using micro-scratch method. The test has been performed in the Micro-Combi Tester, in the micro scratch test mode. A linearly progressive normal load was applied with a Rockwell (Diamond) indenter of 100 mm radius. The starting load was 0.05 N (default value) with a loading rate of 25 N/min. The final load and scratch length were so selected that total delamination occurs within the scratch span. Testing was performed over a scratch length of 4 mm at speed of 10 mm/min, with a final load value of 15N. The critical loads for first crack event (Lc1), first de-lamination (Lc2) and total delamination

(Lc3) were determined from the friction data (the friction coefficient and frictional force). As the friction values fluctuate due to the presence of particle debri and local irregularities, the critical loads were reconfirmed using the acoustic emission data recorded during the test. The failure events were verified using the penetration depth profile as well as with the help of microscopic images taken using the camera attachment at the respective positions of the events.

Frictional Properties

A pin-on-disc type tribometer was used for measuring the frictional properties of the DLC coatings. The machine (Model TR-20CL-M3, Ducom Instruments, India) consisted of a rotating sample (disc) holder and a plunger (connected to controller and sensor system) to mount the pin. The PC-based data acquisition system permitted the measurement of rotation speed, wear depth, wear rate, frictional force and temperature. A beam type load cell (strain gauge based) with maximum capacity of 5kg was available to impart frictional force up to 10 N.

The friction coefficient was evaluated in unlubricated sliding conditions at ambient temperature. Pin of 3 mm diameter was mounted over metal disc of 25 mm diameter and loaded with weight so as to have contact stress of 14 kg/[cm.sup.2]. The tests were conducted with speed of 60 rpm for a track diameter of 18 mm (sliding speed of 55 mm/sec) for a duration of 1 hour. Stainless steel (grade 316L) discs and pins polished to a mean surface roughness ([R.sub.a]) of 0.1 [micro]m were used for the study. Different combinations of SS316L pins and disc surfaces (bare and coated in both cases) were tested and the frictional data was recorded in each case.

Corrosion susceptibility

The corrosion susceptibility of the metal (substrate) surface, with and without the carbon coating, was assessed using cyclic potentio dynamic polarization test as per ASTM G-2129 (Standard Test Method for Conducting Cyclic Potentiodynamic Polarization Measurements to Determine the Corrosion Susceptibility of Small Implant Devices). The measurements were done in an electrochemical cell made with borosilicate glass having 1 litre capacity. Air tight inlets were provided for electrodes, gas bubbler and thermometer. The test sample served as the working electrode, which was enclosed in a holder. A standard calomel electrode (SCE), connected to the electrolyte through a bridge tube, was used as the reference electrode. The auxiliary electrode made of platinum supplying the current was positioned symmetrically relative to the working electrode (or the test sample). The cell was filled with phosphate buffered saline (PBS) and kept in a temperature controlled water bath at 37 [+ or -] 1 [degrees]C. High purity nitrogen gas was bubbled through the electrolyte at a rate of 150 cc/min. A potentiostat/galvanostat (Model GellAC, ACM Instruments, UK) interfaced with a computer, was used for voltage sweeping of the working electrode and the corresponding data logging.

The test samples were made as discs of 12 mm diameter and 4mm thickness and fitted onto the sample holder after rinsing with distilled water. The holder had a water-sealed cavity which exposes only the front surface of the sample to the electrolyte and provides electrical contact onto the back side. The resting potential [E.sub.r] was determined before starting the potentiodynamic scan in the selected rate. The starting potential ([E.sub.i]) and the final potential ([E.sub.f]) were set at cathodic and anodic range with reference to [E.sub.r]. Analysis of the polarization curves were done using Tafel slopes with the help of software, to determine [E.sub.corr] (corrosion potential), [I.sub.corr] (corrosion current) and [C.sub.R] (corrosion rate, in millimeters per year).

Results

Structure of the coating

The FTIR spectrum of the carbon coating deposited using cyclohexane over KBr is given in Figure 2. A sharp, prominent absorption is observed at 1384 [cm.sup.-1] corresponding to the C-C[H.sub.3] symmetric deformation peak. The small peak at 1585 [cm.sup.-1] indicates C=C ([sp.sup.2]) conjugated stretching [14]. In the region 2830-2990 [cm.sup.-1], where the absorption of C-H bonds are normally present, three distinct peaks at 2856, 2927 and 2958 [cm.sup.-1] are seen. These could be assigned to C[H.sub.3] ([sp.sup.3]) symmetric stretch, C[H.sub.2] ([sp.sup.3]) asymmetric stretch and C[H.sub.3] ([sp.sup.2]) asymmetric stretch, respectively [15]. The band centered at 3127 [cm.sup.-1] could be attributed to the moisture adsorbed by the KBr pellet surface. It is clear that the coating made with cyclohexane contains typical carbon-carbon bonds and hydrocarbon bonds.

The XPS survey spectrum gave a peak around 285 eV indicating the presence of carbon (the binding energy range of C1s). The part of the peak scanned in the range 282-292 eV is shown in Figure 3. The peak is broad, indicating the presence of multiple species of carbon. A deconvolution was done to find the ratios of [sp.sup.2] and [sp.sup.3] carbon species (positioned at 284.8 and 285.7eV) as presented in Figure 3. The quantities of these species were calculated as 79.1% and 20.9% respectively. This ratio is very typical with diamond-like carbon. Generally, a good diamond-like carbon should contain more than 20% [sp.sup.3] carbon [11].

The Raman spectrum of the coating is given in Figure 4, along with the spectra of diamond (the indenter tip of microhardness testing machine) and graphite ('highly oriented poly crystalline graphite' purchased from Sigma-Aldrich) for comparison. There are two prominent but broad peaks in the spectrum of the coating, at 1563 [cm.sup.-1] and 1335 [cm.sup.-1]. These are typical with amorphous diamond-like carbon films [16, 17].

Being a technique more sensitive to pi-bonds in carbon network, the Raman spectrum of carbon coating will give information primarily about [sp.sup.2] carbon. A single peak is reported at 1575 [cm.sup.-1] for single-crystal graphite, attributed to the [E.sub.2g2] mode (the G peak). The characteristic peak of microcrystalline graphite is at 1355 [cm.sup.-1] (the D peak) [17]. In the present sample, both G and D peaks appear. The possibility of the presence of free graphite or carbon with dangling bonds in the film has been ruled out, as the resistivity of the film is extremely high (results not included).

X-Ray diffraction spectrum gave only a high background noise, indicating that the coating is devoid of any crystalline order. The films are fully amorphous in nature.

Compiling the results of FTIR, XPS, FT-Raman and XRD, it could be understood that the coating obtained is "Amorphous Hydrogenated Diamond-Like Carbon (a-C:H)" [18].

Thickness and micromorphology

The substrates were coated in the system for different periods and the thicknesses were measured using Calotest machine to determine the rate of deposition. The average rate of the interlayer with TMS (i.e. amorphous silicon-carbon material) is 40 nm/min and that of the layer with cyclohexane (i.e. diamond-like carbon material) is 30 nm/min. Finally the optimized process steps for 1 micron was done and the coating was tested for thickness, the result of which is given in Figure 5. The thickness value obtained is 1.062 micrometer.

The line profile and constructed 3D surface image of the coating, recorded using the optical profilometer, is given in Figure 6. The length of the profile is 500 mm, with a resolution 10 nm. The 3D image constitutes an XY envelope of 500 mm length each with 1 mm spacing. The roughness parameters of [R.sub.a], [R.sub.q] and [R.sub.p] are inscribed in the figure. The [R.sub.a] (average roughness) value is 10.2 nm, against the underlying substrate surface roughness of 82.5 nm.

The image shows a highly homogeneous surface without any prominent growth features or defects. Generally pores are common with thin film coatings, but in the present case, only a few pores are identified. The maximum profile peak height ([R.sub.p]) is 28.2 nm. Previous reports showed that the [R.sub.a] values achievable in DLC made in RF-assisted CVD are comparable with the roughness of the substrate, and the best smoothness obtained was in the range 20-30 nm [19].

Micro hardness and modulus

The microhardness measured at various indentation depths are plotted in Figure 7. It shows decreasing values when going deeper into the coating, from >4800 HV at 0.2 micron to <1000 HV at 1 micron. This variation could be anticipated, considering the inherent limitations in the microhardness measurement by vertical indentation [20]. At indentation depths close to the interface, the substrate effects come into play and the values obtained may not be true for a hard coating. The near-surface values will be more close to the true value of the coated material but the measurements are less reliable because the recovery after indentation is appreciable compared to the measuring depth. This problem in estimating the correct microhardness of thin coatings has been elaborated by many researchers [20] and the reasonable way of measurement has been suggested based on the deformation volume around the indenter. The calculation of hardness of a coating is related the deformation volumes of the coating material and that of the substrate with respect to the total plastic deformation volume. Empirically, the depth of indentation should be such that the plastic deformation zone of indentation is within the coating.

DLC is at least 4 times harder than SS316L in Vicker's scale as per the reported data, and hence in the present case, the hardness at a depth of 25% of the total coating thickness (i.e. at 0.25 [micro]m in a 1 [micro]m coating) is taken as a reasonable estimate. This value, as marked in Figure 7, is 4095 HV (~40 GPa), which is in the higher side of typical hardness for DLC [21].

The modulus corresponding to these measurements, calculated by the software, is given in Figure 8. The initial high value (479 GPa) reduces fast till a depth of 0.3 [micro]m and stabilizes to an average of 261 GPa. The modulus value at 0.25 [micro]m depth is 339 GPa.

Coating adhesion

The outcomes of the micro scratch test on the standard 1 [micro]m DLC coating are given in Figure 9. It contains data related to friction and acoustic emission on the scratch tip, along with images of the scratch track corresponding to the coating failure events. The friction coefficient and frictional force follow the same track with several bumps and peaks, as the scratch progresses. These could be attributed to the presence of particle debri. Identification of the true coating failure event could be done with the help of the acoustic data which represent the vibrations emitted during failures. The reasonable estimates of load values at the first crack (Lc1) and first delamination (Lc2) could be taken as 2.24 N and 2.98 N as marked in the traces in Figure 9. Similarly, the load value for full delamination (Lc3) could be identified as 10.22 N. A reconfirmation of the failure events was done by observing the corresponding points on the scratch track through optical imaging. The images corresponding to the first crack, first delamination and full delamination are given in the right panel of Figure 9.

The failure load values obtained are comparable with that reported for DLC in literature.

Frictional Properties

The study of the frictional properties of the coating on pin-on-disc tribometer was performed mainly considering the applications in articulating implants, like total hip and total knee prostheses wherein low-friction surfaces are preferred. In the test, the load and speed conditions were selected corresponding to the applications, but in dry conditions. Different combinations like uncoated disc versus bare pin, coated disc versus bare pin and coated disc versus coated pin have been tried. The average friction coefficients recorded in each case are represented in Table 1. It could be seen that the friction coefficient of the DLC coating with DLC as counterface is 0.09, which is a remarkably low value [22].

Corrosion susceptibility

In the cyclic potentiodynamic polarization test, the resting potential [E.sub.r] was monitored for one hour, before starting the potentiodynamic scan in the positive (forward) direction. The starting potential ([E.sub.i]) was set at 100 mV negative to [E.sub.r], the scan rate was selected as 0.167 mV/s and the final potential ([E.sub.f]) was set as 500 mV positive to [E.sub.r]. The resultant polarization curve of the DLC coated steel surface is given in Figure 10. The Tafel slopes are marked as dotted lines in the cathodic and anodic regions. The converging point of these slopes represented [E.sub.corr] (the corrosion potential) and [I.sub.corr] (the corrosion current) values. The corrosion rate, [C.sub.R] (the rate of metal dissolution) in millimeters per year, was estimated by conducting analysis on triplicate samples and averaging.

The samples coated with 1 [micro]m thick DLC gave a corrosion rate of 4.675 X [10.sup.-5] mils/yr. The uncoated sample (polished SS-316L of the same batch metal) had a corrosion rate of 1.824 X [10.sup.-3] mils/yr (curves not shown), which by itself, is very low among other metals [2]. Coating with diamond like carbon reduces corrosion rate of SS-316L implants by 2 orders of magnitude.

The protective efficiency (percentage) of the coating could be determined from the polarization curve data by means of the equation

[P.sub.i] = 100 (1 - [I.sub.corr] / [i.sup.o.sub.corr])

where [i.sub.corr] and [i.sup.o.sub.corr] indicate the corrosion current densities in the presence and absence of the coating, respectively [23, 24]. As per the data obtained from figure 10, the protective efficiency could be calculated as

[P.sub.i] = 100 {1 - (9.313 x [10.sup.-8] / 3.634 x [10.sup.-6])} = 97.44%

Discussion

The highlight of this study is that the semi-industrial scale PECVD-based carbon coater is capable of providing good quality DLC coating appropriate for modifying and protecting metallic implants.

The process is found to be repeatable and viable. Repeatability is the assurance of maintaining the specifications of a product from batch to batch, and viability refers to the economy of production output with respect to cost and time. In common practice, it is difficult to maintain the plasma stable because of the pumping turbulence and non-uniformity of the vapour flow, which in turn, alters the coating quality. The specially designed shield part and the vapor flow system ensures the stability of the plasma. It protects from pumping turbulence and maintains a vertical streamlined flow in the process space. This mechanism, in conjunction with computer control, makes the process repeatable.

The quality of DLC is inversely related to the rate of deposition. Increasing the rate by increasing the flow rate or increasing the pressure level will degrade the properties of the coating. In the present case, optimum values of flow rate, bias potential and chamber pressure were determined through trials. It gives an acceptable rate of formation of the coating with prescribed quality. Generally for biomedical applications, DLC of 1-4 [micro]m is preferred. The process time for obtaining 1 [micro]m coating in the present study is 40 minutes. The process chamber is designed for batch coating so that large number of pieces (implants) could be loaded at a time. Therefore, the coating system and the process developed is ideal for industrial production. The process uses liquid precursors which adds to the safety, economy and convenience.

In the chemical composition analysis, FTIR and FT-Raman confirmed the presence of C=C, C-H, C[H.sub.2] and C[H.sub.3] carbon-hydrogen bondings. XPS revealed that both [sp.sup.2] and [sp.sup.3] carbon species are present, with the former in large quantity. XRD showed the amorphous nature of the coating obtained. The XRD results, along with the lack of conductivity of the coating, rule out the presence of graphitic content. Thus, the coating structure could be envisaged as a network of [sp.sup.2] and [sp.sup.3] hybridised carbon with the dangling bonds attached to hydrogen. As per the DLC classification, this is "Amorphous Hydrogenated Diamond-Like Carbon (a-C:H)"[18]. The [sp.sup.2] and [sp.sup.3] content (79.1% and 20.9% respectively) obtained in XPS is typical with 'a-C:H' obtainable in PECVD process.

Generally in the Raman spectrum of DLC films, both G and D peaks appear. The former is due to the bond stretching of all pairs of [sp.sup.2] atoms in both rings and chains, and the latter is due to the breathing modes of rings [16]. There is a T peak reported at 1060 [cm.sup.-1] due to C-C [sp.sup.3] vibrations which appears only during UV excitation [16]. The broadening observed in the G and D peaks in amorphous diamond like carbon depends on the size of clusters, their distribution, chemical environment and stress in the film. The existence of the D peak is considered as the evidence for condensed benzene rings in amorphous hydrogenated carbon films. The G peak is not necessarily composed of the [E.sub.2g2] mode of graphite alone but also of [sp.sup.2] C=C stretch vibrations [17]. The interpretation of the G and D peaks is complicated in DLC unless done in conjunction with some other technique, and hence Raman analysis appears useful only for qualitative assessment.

The coating obtained in the present case is formed out of the vapor phase decomposition and reaction of cyclohexane in the plasma driven by radio frequency (RF) power. The different dissociation channels the corresponding energies for cyclohexane (c-[C.sub.6][H.sub.12]) have been reported, as identified through thermal analysis. In the temperature range 1000-1300K, c-[C.sub.6][H.sub.12] can dissociate in steps to (3[C.sub.2][H.sub.4]), (2[C.sub.3][H.sub.6]), ([C.sub.4][H.sub.6] + [C.sub.2][H.sub.4] + [H.sub.2]), (c[C.sub.6][H.sub.11] + H) ([C.sub.2][H.sub.4] + *[C.sub.4][H.sub.8]* and further to *[C.sub.4][H.sub.8]* ! 2[C.sub.2][H.sub.4]) or (c[C.sub.6][H.sub.10] + [H.sub.2]) [25]. However, in the present case of PECVD, the dissociation pathways are difficult to envisage. Optical emission techniques are less efficient in detecting the emissions from excited carbon-hydrogen species. Moreover, the depositing species are controlled by local parameters over the sample surface, unlike the gas phase species in the plasma.

The nano-level smoothness of the coating ([R.sub.a] value of 10.2 nm) is notable feature which is highly useful for cardiovascular devices. Even highly polished bare metal implants have roughness in the range of 0.1 [micro]m which may lead to adherence of blood components and act as foci to thrombosis. The nanolevel smoothness of DLC along with inherent blood compatibility provide a non-thrombotic surface [10]. A few models of DLC-coated coronary stents are commercially available.

The mechanical properties of the obtained coating are also remarkable. The estimated indentation hardness is 4095 HV (~40 GPa), where as the reported Vickers hardness values for a-C:H are typically around 3500 HV [3]. The corresponding modulus value is 339 GPa, which will guard against the cracking initiated by internal stresses. Normally, DLC develop an internal stress as high as 10GPa due to mismatch in thermal expansion coefficient with the metal substrate. The stresses at the interface will lead to coating peel off from the substrate. It is known to be extremely difficult to make a stable DLC coating on stainless steel [3]. In the present case, the interlayer made with TMS takes care of the peel-off problem and imparts very high adhesion strength.

The friction coefficient obtained in the pin on disc test (0.9 for DLC to DLC) is the lowest known for any surface. This is particularly relevant in articulating implants. The friction between the counter-faces is a factor of concern in total hip and total knee prostheses. Higher friction may lead to tissue necrosis due to frictional heat, formation of debris in the surroundings and higher corrosion of the implant surface. The surface smoothness, microhardness, modulus and adhesion strength together makes the coating useful for the load-bearing articulating implants.

The corrosion rates determined using potentiodynamic polarization studies indicated that the coating reduces the corrosion of the implantable grade stainless steel surface by 2 orders of magnitude. The threat of leaching metal ions from implants due to corrosion has been articulated in many studies [26]. The ions can either accumulate in tissues near the implant or may be transported to other parts of the body. A case study of 20 stainless steel hip arthroplasties in the human body after 10-13 years showed a considerably higher metallic concentration in the body fluid in comparison with controls without implant. This included excess Ni in blood (0.51 ig/L), plasma (0.26 ig/L) and urine (2.24 ig/L), and excess Cr level in plasma (0.19 ig/L). A failed Ti-6Al-4V total knee replacements after 57 months showed excess concentration of Ti in the serum (135.57 ig/L) than in the control. Also, cobalt toxicity from the degrading metal-on-metal total hip replacements has been reported [27]. The scope of a protective coating like DLC is evident from the current study.

Conclusion

It was possible to obtain adherent and conformal carbon coating over implantable metal surface using the semi-industrial scale coating equipment based on PECVD. The equipment has special arrangements for vapor/gas flow in the process space and plasma shield accessory for the homogeneity and stability of the plasma. The availability of large process area and the use of liquid precursors are other features. This, along with the provision for automated process execution, ensures viability and repeatability of the process. At optimized conditions, the process time for obtaining 1 micrometer coating is 40 min. This is an optimal thickness for coating metallic implants.

The characterization of the coating using FTIR, XPS, FT-Raman and XRD, showed that the coated material is "Amorphous Hydrogenated Diamond-Like Carbon (a-C:H class)". The coating is found to be conformal with nanometer level smoothness, wherein the average roughness is 10.2 nm. The microhardness estimated is 4095 HV (~40 GPa), which is far higher than that of typical a-C:H coatings. The corresponding modulus is 339 GPa. The values indicate that the coating is capable of resisting indentations and can withstand internal and interfacial stresses. A high adhesion strength of 10.22 N is seen at full delamination (Lc3) of the coating.

The coating with itself as counter-face in pin-on-disc test gave a friction coefficient of 0.09, which is a remarkably low value. The corrosion rate of metal with 1 im coating measured through potentiodynamic polarization, is 4.675 X [10.sup.-5] mils/yr, indicating a reduction in corrosion by 2 orders of magnitude compared to bare metal. The corrosion protection efficiency is calculated as 97.44%.

In summary, the coating system is capable of providing adherent and mechanically strong DLC (a-C:H class) coating which can protect metallic implants from in vivo corrosion.

Acknowledgement

The author would like to express thanks to Dr. G.S. Bhuvaneshwar, Dr. H.K.Varma and Mr. C.V. Muraleedharan (SCTIMST, Trivandrum), Dr. G. Mohan Rao (IISc, Bangalore) and Mr.Yvon Sampeur and Mr.Fabrice Oge (Isytech, France) for helping to plan and execute this study. Technical support extended by Mr. Sajin Raj and Mr. Willi Paul is also gratefully acknowledged.

References

[1.] M. Sivakumar, S.Rajeswari, Investigations of failures in stainless steel orthopaedic implant devices: Pit induced stress corrosion cracking, J. Mater. Sci. Lett., 11, 1039-1042 (1992).

[2.] M.Sivakumar, U. Kamachi Mudali, S. Rajeswari, Investigation of failures in stainless steel orthopaedic implant devices: Fatigue failure due to improper fixation of a compression bone plate, J. Mater. Sci. Lett., 13, 142-145 (1994).

[3.] S.A. Catledge, V. Thomas and Y.K. Vohra, Nanostructured diamond coatings for orthopaedic applications, Woodhead Publ Ser Biomater., 105-150 (2013). doi:10.1533/ 9780857093516.2.105.

[4.] J.J. Wykrzykowska, R.P. Kraak, S.H. Hofma, Bioresorbable Scaffolds versus Metallic Stents in Routine PCI, N Engl J Med., 376, 2319-2328 (2017), DOI: 10.1056/ NEJMoa1614954.

[5.] P.S. Kallidonis, I.S. Georgiopoulos, I.D. Kyriazis, S. Kontogiannis, A.M. Al-Aown, Liatsikos EN, The full metallic double-pigtail ureteral stent: Review of the clinical outcome and current status, Indian J Urol, 31, 8-14 (2015).

[6.] C. C. Wachesk, V. J. Trava-Airoldi, N. S. Da-Silva, A. O. Lobo and F. R. Marciano, The Influence of Titanium Dioxide on Diamond-Like Carbon Biocompatibility for Dental Applications, J. Nanomaterials, 2016, Article ID 8194516, 7 pages (2016). doi:10.1155/2016/8194516.

[7.] U. Kamachi Mudali, T.M. Sridhar and Baldev Raj, Corrosion of bio implants, Sadhana 28, 601-637 (2003).

[8.] L.G. Mercuri. The TMJ Concepts patient fitted Total Temporomandibular Joint Reconstruction Prosthesis in Total Temporomandibular Joint Reconstruction, Oral and Maxillofacial Surgery Clinics of North America, 12, 73-91 (2000).

[9.] X. Liu, P. K. Chu, C. Ding, Surface modification of titanium, titanium alloys, and related materials for biomedical applications, Materials Science and Engineering: R: Reports, 47, 49-121 (2004).

[10.] Roy RK and Lee KR, Biomedical Applications of Diamond-Like Carbon Coatings: A Review, J Biomed Mater Res Part B: Appl Biomater, 83B: 72-84, (2007).

[11.] G. Dearnaley and J.H. Arps, Biomedical applications of diamond-like carbon (DLC) coatings: A review, Surface & Coatings Technology, 200, 2518-2524, (2005).

[12.] N. Korner, E. Beck, A. Dommann N. Onda, J. Ramm, Hydrogen plasma chemical cleaning of metallic substrates and silicon wafers, Surface and Coatings Technology, 76-77 Pt 2, 731-737 (1995).

[13.] B.O. Aronsson, J. Lausmaa, and B.Kasemo, Glow discharge plasma treatment for surface cleaning and modification of metallic biomaterials. J. Biomed. Mater. Res., 35, 49-73 (1997).

[14.] C. Degang, L. Fanxiu, Y. Rang, Y. Ruizeng and S. Qingbiao, Effect of Laser Irradiation on the Optical Properties of Diamond-Like Carbon, Chinese Phys. Lett. 12 87 (1995).

[15.] Z. Sun, Y. Sun, O. Yang, X. Wang and Z. Zheng, IR spectral investigation of the pyrolysis of polymer precursor to diamond-like carbon, Surf. Coat. Technol., 79 108-112 (1996).

[16.] A.C. Ferrari, Determination of bonding in diamond-like carbon by Raman spectroscopy, Diamond and Related Materials 11 1053 (2002).

[17.] J. Schwan, S. Ulrich, V. Batori, and H. Ehrhardt, and S.R.P. Silva, Raman spectroscopy on amorphous carbon films, J. Appl. Phys., 80, 440 (1996).

[18.] C.Donnet, Recent Progress on the Tribology of Doped Diamond-like and Carbon Alloy Coatings: A Review, Surf. Coat. Tech., 100-101, 180 (1998).

[19.] D.Tischler, Z.Budinska, P.Vleak, I.Stipanek, Nanoindentation test for DLC coating analysis Materials, Engineering 18, 105 (2011).

[20.] S.J. Bull and D.S. Rickerby, New developments in the modelling of the hardness and scratch adhesion of thin films, Surf. Coat. Technol., 42, 149-164 (1990).

[21.] A. Grill, Review of the tribology of diamond-like carbon, Wear, 168, 143-153 (1993).

[22.] J.A. Heimberg, K.J. Wahl and I.L. Singer, Superlow friction behavior of diamond-like carbon coatings: Time and speed effects, Appl Phys Lett, 78, 2449-2451 (2001).

[23.] Y.J. Yu, J.G. Kim, S.H. Cho, J.H. Boo, Plasma-polymerized toluene films for corrosion inhibition in microelectronic devices. Surf Coat Technol, 162, 161 (2003).

[24.] S-H Lee, J-G Kim, H-W Choi and K-R Lee, Microtensile strain on the corrosion performance of diamond-like carbon coating, J Biomed Mater Res 85A, 808 (2008).

[25.] J.H. Kiefer, K.S. Gupte, L.B. Harding, S.J. Klippenstein, Shock tube and theory investigation of cyclohexane and 1-hexene decomposition, J. Phys Chem A. 113, 13570 (2009).

[26.] Y. Okazaki, E. Gotoh, Comparison of metal release from various metallic biomaterials in vitro, Biomaterials, 26, 1121 (2005).

[27.] X. Mao, A.A. Wong and R.W. Crawford, Cobalt toxicity--an emerging clinical problem in patients with metal-on-metal hip prostheses?, Med. J. Austr., 194 649-651 (2011).

Manoj Komath

Biomedical Technology Wing, Sree Chitra Tirunal Institute for Medical Sciences and Technology, Poojappura, Trivandrum

695012, India

Received 15 December 2017; Accepted 30 December 2017; Published online 31 December 2017

* Coresponding author: Dr. Manoj Komath E-mail: manoj@sctimst.ac.in

Caption: Figure 1: Schematic diagram of the PECVD chamber

Caption: Figure 2: FTIR of the carbon coating

Caption: Figure 3: The C1s peak obtained in the X-ray Photoelectron Spectroscopy (XPS) of the surface of the carbon layer. The [sp.sup.2] and [sp.sup.3] carbon species contents are estimated by the deconvolution of the peak

Caption: Figure 4: The Raman spectrum taken on the surface of the coating (A). The spectra of diamond crystal (B) and highly oriented polycrystalline graphite (C) are shown for comparison

Caption: Figure 5: Thickness measurement of the carbon coating programmed for 1 micron, using Calotest. The abraded crater image is seen with the measurement window and the test parameters

Caption: Figure 6: The line profile of the coated sample and the corresponding numerical values, along with the constructed 3D image

Caption: Figure 7: Depth variation of Vickers microhardness in 1 micron coating. The value at 0.25 microns (marked) is taken as the reasonable estimate for the coating hardness

Caption: Figure 8: Modulus variation across of 1 micron coating at various depths. The value becomes almost stable below 0.3 microns, the average of which is represented by the line

Caption: Figure 9: The results of the scratch test, with traces of friction data and acoustic emission data in the left panel. The critical loads for first crack (Lc1), first delamination (Lc2) and full delamination (Lc3) are marked by comparing the traces. The optical images corresponding to the three events of are given in the right panel

Caption: Figure 10: Polarization curve obtained in potentiodynamic polarization study. The Tafel slopes are marked with sloping dotted lines. The horizontal dotted line corresponds to the rest potential
Table 1: Friction coefficient values of sliding pin on disc

Surface                      Material           Friction
                                               Coefficient
Base    Contact       Base         Contact

Metal    Metal    Polished disc    Bare pin       0.363
Metal     DLC      Coated disc     Bare pin       0.137
DLC       DLC      Coated disc    Coated pin      0.09
COPYRIGHT 2017 Society for Biomaterials and Artificial Organs
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Original Article
Author:Komath, Manoj
Publication:Trends in Biomaterials and Artificial Organs
Date:Oct 1, 2017
Words:8043
Previous Article:Collagen Based 3D Model for In Vitro Study of Osteogenesis.
Next Article:Cationically Modified Solid Lipid Nanoparticles for Intestinal Delivery of an Antiplatelet Agent, Clopidogrel: Preparation and Characterization.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters