journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
Durability evaluation of biopolymer coating on titanium alloy substrate J. Ryan Stanfieldn, Stacy Bamberg Departments of Mechanical Engineering and Bioengineering, University of Utah, 50 S Central Campus Dr., Salt Lake City, UT 84112, USA
art i cle i nfo
ab st rac t
Article history:
For this study, a commercially available phosphorylcholine (PC) polymer was applied to
Received 19 January 2014
Ti6Al4V ELI. A multivariate approach to design a statistically significant array of experiments
Received in revised form
was employed to evaluate and estimate optimization of PC-immobilization process factors.
11 March 2014
The seven process factors analyzed were (1) power level for RFGD plasma treatment, (2)
Accepted 13 March 2014
duration of plasma treatment, (3) concentration of PC solution used to coat samples, (4) rate at
Available online 21 March 2014
which samples were dipped in/out of the solution, (5) temperature for curing, (6) relative
Keywords:
humidity level during curing, and (7) duration of curing. Imaging and analysis of the coating
Titanium alloy
were done via fluorescence microscopy (FM), confirming the uniform coverage of PC polymer
Phospholipid polymer
on titanium substrate. The process factors were evaluated by three measured responses:
Durability
initial thickness, coating durability and degree of cross-linked coating, which were assessed by
Blood compatibility
FM, a spray test and extraction in IPA, respectively. Variations in PC solution concentration
Implantable medical devices
showed no impact on fouling resistance of the resultant coating. It was hypothesized that the PC-application process factors could be optimized to yield favorable outcomes in durability and degree of cross-linked coating responses. The resulting statistical model indicates that PC solution concentration, dip rate, and cure temperature are the three greatest singular effects on both durability and degree of cross-linking. In addition, plasma treatment of the substrate with O2 was effective in enhancing the degree of cross-linking of the polymer surface. & 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
Titanium and its alloys, in addition to stainless steels, are widely used in surgical implant applications. Cardiovascular treatments, such as vascular stents, heart valves, and ventricular assist devices (VADs), have continued to promote the development of surface modification methods in order to reduce platelet deposition on metallic surfaces (Liu et al., 2004). Anticoagulation and antiplatelet n
Corresponding author. Tel.: þ1 801 581 6441; fax: þ1 801 585 9826. E-mail address: ryan.stanfi[email protected] (J. Ryan Stanfield).
http://dx.doi.org/10.1016/j.jmbbm.2014.03.003 1751-6161/& 2014 Elsevier Ltd. All rights reserved.
pharmacologic therapies are frequently used on patients implanted with such apparatuses due to the risk of thrombosis and thromboembolism (Colli et al., 2007; John et al., 2008). To improve the thromboresistance of titanium alloys in contact with the cardiovascular system, various surface coatings have been developed such as diamond-like carbon (DLC), titanium nitride (TiN), heparin, silicone, and 2-methacryloyloxyethel phosphorylcholine (MPC) (Sin et al., 2005). Exceptional biocompatibility and hemocompatibility have been shown on
10
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
biomimetic surfaces synthesized from phospholipid polymers [phosphorylcholine (PC)] by reducing protein adsorption and platelet adhesion, as the PC head group is prevalent on the outside of biological cell membranes (Hayward and Chapman, 1984; Ishihara et al., 1990, 1998; Lewis 2000; Lewis et al., 2001). In addition to coatings, physical and chemical surface treatments have been utilized on titanium alloys including passivation, polishing, oxidation, vapor deposition, silanization, glow discharge plasma treatment, and ion implantation (Liu et al., 2004). Plasma treatment and ion implantation have been shown to increase the hardness and wear resistance of metallic substrates (Ueda et al., 2003). Moreover, plasma treatment has been shown to improve biocompatibility and increase adhesion of different materials (Yasuda and Gazicki, 1982). Radio frequency glow discharge (RFGD) plasma treatment is frequently used for surface cleaning, and only introduces ions and electrons, as opposed to bulk materials (Ueda et al., 2003). The deposited ions can act as functional groups to covalently attach other polymers or biomolecules (Chevallier et al., 2001). Covalent linkage of PC to a substrate is more stable than physically adsorbed PC. Thus, modification of a surface in order to promote the covalent connection of PC to a metallic substrate is highly desirable for long-term medical implant applications. Inorganic coatings on metallic surfaces, such as diamondlike carbon (DLC), titanium-oxides, and nitrides, which show evidence of mechanical and chemical stability, and high inertness, are employed to increase hemocompatibility of the base materials used in surgical devices. The hydrophobicity and low surface roughness exhibited on DLC-coated substrates have shown a higher proportion of albumin than fibrinogen in thrombogenic studies (Jones et al., 2000). However, DLC can be difficult to apply to the enclosed or undercut surfaces present on many implantable devices. Surface adhesion of DLC to substrates has been shown to be low without plasma pretreatment (Ozeki et al., 2010), as with MPC-coatings. Organic coatings are also used on metallic surfaces to improve hemocompatibility. PEO (Chen et al., 2005), PMEA (Gunaydin et al., 2002), PEG (Zhang et al., 2001), and other polymer coatings have been shown to reduce platelet adhesion and aggregation, and protein adsorption. However, durability, adhesion, the modification method, and efficacy impact the value of coatings, and must be considered on long-term surgical implants. While biomimetic organic polymers, such as PC, show promise for improvement of hemocompatibility, techniques to optimize immobilization of the material onto metals significant to implantable devices have not been widely evaluated for durability. Characterization of PC material coated on substrates has been reported using X-ray photoelectron spectroscopy (XPS) (Hayward et al., 1986), atomic force microscopy (AFM) (Clarke et al., 2001), scanning electron microscopy (SEM) (Lewis et al., 2002), and Fourier-transform infrared (FTIR) microscopy (Lewis et al., 2004). Furthermore, spectroscopic ellipsometry (SE) and fluorescence microscopy (FM) analysis techniques have been used to analyze the presence, distribution, thickness and swelling of PC films on titanium, polypropylene, silicone, and other materials (Tang and Lu, 2001; Wang et al., 2005). Using FM analysis, the intensity of the reflected light is shown to be proportional to the thickness of the PC
Fig. 1 – PC film thickness and FM intensity correlation as a function of solution concentration. film, as quantified by SE. The linear relationship is depicted in Fig. 1. The primary objective in this study was to examine the process of titanium alloy (Ti6Al4V) surface modification for hemocompatibility, and assess which variables had the greatest impact on durability and integrity of the surface coating. A simple surface modification process was developed implementing RFGD plasma treatment. The parameters for plasma treatment, surface coating and curing of PC polymer were varied for statistical analyses. A technique was developed to aggressively erode the surface coating and to investigate the removal of the coating. Additionally, other analyses were carried out on surface coating thickness, coverage, degree of cross-linking by FM, surface wettability, and surface coating biocompatibility via in vitro protein adsorption analyses. A commercially available software was implemented in order to create a feasible design of experiments (DOE) test matrix, and to model and analyze each response.
2.
Materials and methods
2.1.
Materials
Titanium alloy, Ti6Al4V extra low interstitial (ELI) as per ASTM standard F136, was procured (Titanium Metal Supply, Poway, CA), machined to an average surface roughness (Ra) of 400 nm, passivated per ASTM standard F86, and polished to Ra of 100 nm with 400, pumice, 6 μm, and 1 μm grade ceramic aluminum oxide radial bristle disks (3 M, St. Paul, MN). Digesil NC (RPM Technology, Reno, NV) was used as a silicone stripper post-polishing. A phospholipid copolymer, PC 1036, was obtained from Vertellus Biomaterials (Hampshire, UK). Rhodamine 6G (Sigma-Aldrich, St. Louis, MO) was used as a staining agent for fluorescent visualization.
2.2.
Surface pretreatment
After polishing, coupons were rinsed with distilled water and isopropyl alcohol (IPA), cleaned ultrasonically for 90 min in a 20 wt% phenol solution in dichloromethane (DCM), and solvent washed ultrasonically two times for 90 min each in DCM. The Ti6Al4V surfaces were pretreated by O2 plasma with RFGD (Plasma Technology Systems, Belmont, CA) for
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
1–5 min in a 0.3 T vacuum chamber. The applied plasma power ranged from 100 to 600 W at a frequency of 13.65 MHz. Control coupons did not undergo plasma pre-treatment.
2.3.
Phospholipid polymer application
Under ISO Class 8 air, the plasma-treated coupons were dipped at a rate from 1 to 19 mm/s in argon-blanketed 0.25– 2.25 wt% PC solution in absolute ethanol (EtOH). Then the solvent was allowed to evaporate for 20 min. The PC-coated coupons were cured in an environmental test chamber (Tenney, New Columbia, PA) for 4–16 h at 40–100 1C and 25–75% relative humidity (%RH), depending on the DOE test condition. The control coupons were dipped at a rate of 10 mm/s in 1.25 wt% PC solution and were cured for 4 h at 70 1C and 50% RH.
2.4.
Surface characterization
PC-coated titanium alloy surfaces were stained for 30 s in a solution of 0.02 wt% rhodamine 6G in distilled water, rinsed twice each for 30 s in distilled water, and dried at room temperature (RT) conditions for 2 h, as described by Wang et al. (2005). The stained coupons were imaged using a fluorescence microscope (Olympus, Center Valley, PA) and digital camera (Lumenera, Ottawa, ON). The images were analyzed with ImageJ software (National Institutes of Health). ImageJ was used to calculate the mean intensity value across all of the pixels in each image. Silicon wafers were used as a control for film thickness measurement using a variable angle spectroscopic ellipsometer (JA Woollam, Lincoln, NE) to show FM measurement correlation. Silicon wafers were coated and analyzed using the same methods as the Ti-alloy control coupons.
2.5.
agitated for 24 h. Any PC polymer chains that had not been cross-linked through the cure process would theoretically wash away during this operation. After the IPA agitation, coupons were dried for 4 h in a convection oven at 40 1C. The coupons were stained, imaged and analyzed as per the above process. FM intensity values of extracted and non-extracted coupons were compared to find a net change in value after the IPA agitation.
2.7.
Surface wettability and protein adsorption
Prior to FM analysis, the static contact angle of water on a selection of coated and uncoated coupons was measured with a contact angle goniometer (190-F1; Ramé-Hart Instrument Co., Succasunna, NJ). Five PC-coated samples were prepared similar to those previously described as control coupons with the exception of PC solution concentration. The manufacturersupplied hardware and software were utilized to calculate the contact angle of a 2 mL droplet of distilled water with five measurements per sample. In order to evaluate the validity of different PC solution concentrations, surface protein adsorption on coated and uncoated polyethylene terephthalate (PET) samples was gaged by a double antibody binding assay. PET strips were prepared similar to control coupons, but used both 0.25 and 2.25 wt% PC solutions in EtOH. Samples were incubated in human plasma (ProteoGenex, Culver City, CA) for 10 min or 24 h, washed four times in phosphate-buffered saline (PBS; BD Biosciences, San Jose, CA), incubated in anti-human fibrinogen (Sigma-Aldrich) for 30 min, washed four times in PBS, incubated in anti-goat IgG peroxidase conjugate (SigmaAldrich) for 30 min, washed four times in PBS, and incubated in o-phenylenediamine dihydrochloride (Sigma-Aldrich) for 10 min. Absorbance was read and recorded spectrophotometrically at 450 nm.
Surface coating durability 2.8.
Initial work to characterize coating adhesion and durability by means of a standardized adhesion test (ASTM D3359) was unfruitful; therefore a spray test was developed to aggressively erode a coating from a metallic substrate. In order to hydrate the PC-coated surface prior to the durability test, coupons were soaked in sterile water for 24 h. A spray gun (Binks, Glendale Heights, IL) was mounted 5 mm from the surface and the coupon was sprayed with sterile water in four distinct locations for 2, 20, 30, and 60 min per location. An input air pressure on the spray gun of 414 kPa (60 psi) produced a force of 0.3 N perpendicular to the coupon surface. Coupons were then stained, imaged, and analyzed, as per the above process. With the ImageJ software the area of the removed PC was measured at all four locations per coupon. Spray time vs. area was plotted and a logarithmic trend line was fit to each coupon data set. The area under the logarithmic curve was integrated in order to define a metric for coating durability, where lower area implied higher durability.
2.6.
11
Degree of surface coating cross-linking
Because the PC polymer is soluble in IPA, as well as EtOH, a selection of coupons was placed in HPLC-grade IPA and
Design of experiments
JMPs Statistical Discovery Software (SAS, Cary, NC) was used to create a randomized quarter fraction screening design for seven experimental factors. The DOE test matrix required 38 runs to be carried out for each of the three responses, so 114 coupons were prepared with the above process. The range for each experimental factor is shown in Table 1. Additionally, 30 control coupons were prepared with the process outlined above, and are tabulated below. Note that the control coupons were not subjected to surface modification via plasma treatment.
3.
Results
3.1.
Surface characterization
We show the linear correlation between solution concentration and film thickness in Fig. 1, and the corresponding relationship with FM intensity values. Similar to the results of Wang et al., surface coating uniformity and morphology can be analyzed via the FM images presented in Fig. 2, where solution concentration and dip rate have been varied. Fig. 3a
12
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
Table 1 – Experimental factors for multivariate statistical testing with associated ranges and control values. Var
Factors
Low
High
Control
Units
A B C D E F G
Plasma time Plasma power Dip rate PC solution concentration Cure temperature Cure time Cure environment
1 100 1 0.25 40 4 25
5 600 19 2.25 100 16 75
– – 10 1.25 70 4 50
min W mm/s wt% 1C h %RH
Fig. 2 – Uniformity and morphology of surfaces coated with varying solution concentrations and extraction rates. Top row: 1.25 wt% PC in EtOH at (a) 1, (b) 10, and (c) 19 mm/s extraction rates. Bottom row: 10 mm/s extraction rate at (a) 0.25, (b) 1.25, and (c) 2.25 wt% PC in EtOH solution. Scale bars are 1 mm.
displays the mean intensity value for each group of coupons processed under distinct coating conditions, where FM intensities are proportional to both dip rate and solution concentration. The JMPs Statistical Discovery Software built a model based on the results from the mean intensity value data for each of the test conditions from the DOE matrix. The actual by predicted plot from the model is shown in Fig. 4a. The model further estimates the singular and coupled effects on the FM intensity, which correspond to thickness. A Pareto chart containing the t-ratios (standardized effect) for each of the seven factors on FM intensity is displayed in Fig. 5a, with the highest two singular effects being solution concentration and dip rate. The statistical significance of each effect is tested by comparing the mean square against an estimate of the experimental error. In this case two effects, solution concentration and dip rate, have P-values less than 0.05 (po0.0001), indicating that they are significantly different from zero at the 95.0% confidence level. Those effects described as positive increase the FM intensity or thickness value.
3.2.
Coating durability
A photograph and the corresponding FM images of a titanium alloy coupon coated with PC solution and sprayed for durability are shown in Fig. 6. Fig. 7 displays a series of spray
time vs. area points for the control coupons with logarithmic trend lines. Samples that yielded trend lines below the controls were projected to be more durable, while trend lines above were predicted to be less durable than the controls. The actual by predicted plot from the model for coating durability is shown in Fig. 4b. The model further estimates the singular and coupled effects on the coating durability. The Pareto chart containing the t-ratios for each of the seven factors on coating durability is displayed in Fig. 5b, with the highest two singular effects being solution concentration and dip rate. In this case, three effects, solution concentration, dip rate, and solution concentration coupled with cure temperature, have statistical significance with P-values less than 0.05. An optimal process to maximize the coating durability was predicted from the model and is presented in Table 2.
3.3.
Coating cross-linking
Fig. 3b displays the mean intensity value for each group of extracted coupons processed under distinct coating conditions, where the resulting FM intensities are proportional to dip rate, but not solution concentration. The mean FM intensity for the control coupons decreased by 25% after IPA-based extraction; the intensities for all other groups shown in Fig. 3 decreased by an average of 21% post-extraction.
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
13
The actual by predicted plot from the model for degree of cross-linking is shown in Fig. 4c. The model further estimates the singular and coupled effects on the degree of cross-linking. The Pareto chart containing the standardized effects for each of the seven factors is displayed in Fig. 5c, which shows that no singular effect has significant impact on the degree of cross-linking. However, one coupled effect, solution concentration with cure temperature, has statistical
significance with a P-value less than 0.01. An optimal process to maximize the degree of cross-linking was predicted from the model and is presented in Table 2.
Fig. 3 – Mean fluorescence microscopy image intensities for varying PC solution concentrations and dip rates. (a) Titanium alloy coupons coated, cured, and imaged, and (b) titanium alloy coupons coated, cured, extracted in IPA, and imaged.
Fig. 5 – Pareto charts display standardized effect (t-ratio) value along the horizontal axis for all seven singular factors as identified in the tables for (a) surface characterization, (b) durability response, and (c) degree of cross-linking response.
Fig. 4 – Statistical fit models for (a) surface characterization, (b) durability response, and (c) degree of cross-linking response show actual by predicted values for each. Horizontal dotted mean lines, angled solid regression lines, and 95% dotted confidence curves.
14
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
Fig. 6 – Titanium alloy coupon after undergoing the spray test with corresponding FM images. Beginning in the upper left-hand corner and going clockwise the four locations are sprays for increasing time durations of 2, 20, 30, and 60 min.
Table 2 – Statistical predictions for process optimization for both coating durability and degree of cross-linking. Var
Factor
Durability
Cross-link
A B C D E F G
Plasma time Plasma power Withdrawal rate Solution concentration Cure temp Cure time Cure humidity
Low Low High High High High High
High High Low Low Low High High
Fig. 7 – Area (pixels) by time (min) for measured spray locations on a sample batch of control coupons with logarithmic trend lines. Area under the logarithmic curve represents durability response value, where lower areaunder-the-curve values represent higher durability.
3.4.
Surface wettability and protein adsorption
The wettability of each surface was measured five times per sample on modified and unmodified Ti6Al4V coupons using a static contact angle goniometer. The contact angle of a 2 mL droplet on unmodified coupons was 87721, while for coupons modified with 0.25 and 2.25 wt% PC solution, the angles were 64721 and 60731, respectively. The amount of fibrinogen adsorbed on the PC-coated and uncoated PET strips was measured by the absorbance of the reacted o-phenylenediamine dihydrochloride at 450 nm. The amount of adsorbed fibrinogen on the PET samples is shown in Fig. 8. The average absorbance of the uncoated PET was much higher than those of the coated samples, and the
Fig. 8 – Fibrinogen adsorption on uncoated and PC 1036 coated PET samples.
reduction of fibrinogen adsorption was 85–87% for samples coated with 0.25 or 2.25 wt% PC solution. This considerable difference was verified by student t-tests comparing the coated and uncoated samples (po0.0001).
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
4.
Discussion
For this study, a commercially available PC-polymer (PC 1036) was applied to Ti6Al4V ELI. A multivariate approach to design a statistically significant array of experiments was employed to evaluate and estimate optimization of PC-immobilization process factors. The seven process factors analyzed were power level for RFGD plasma treatment, duration of plasma treatment, concentration of PC solution used to coat samples, rate at which samples were dipped in/out of the solution, temperature for curing, relative humidity level during curing, and duration of curing. The responses were coating durability and degree of cross-linked coating, which were assessed by spray testing and extraction in IPA, respectively. Imaging and analysis of the coating were done via FM. Evaluation on a set of samples by staining and imaging with fluorescence microscopy showed that all coupons had complete coverage and uniform distribution of the PC 1036 polymer. The average pixel intensity across the images was proportional to the solution concentration and the dip rate. The efficacy of the cross-link characterization test was demonstrated with the average pixel intensity of FM images decreasing by 21% post-extraction in IPA. While the FM intensity of the control coupons dropped by 25% postextraction in IPA, the coupons that underwent surface pretreatment with plasma and were processed with the same dip rate and solution concentration conditions as the controls only dropped by 7% post-extraction, validating the benefit of the RFGD plasma treatment relative to the degree of coating cross-linking. The results from the spray testing did not exhibit a significant benefit to the durability of the coating with plasma treatment. Indeed, no significant difference was seen between durability data for coupons processed with and without plasma treatment under the same solution concentration and dip rate conditions. This may be attributed to the type of bond formed by the PC polymer and the substrate. The surface of the substrate is a thin titanium oxide layer formed during passivation, rather than the titanium alloy. Employing a solvent-evaporation technique to apply the polymer with a suspended solution directly to an oxide layer likely produces a noncovalent or intermolecular bond. Thus, inclusion of RFGD plasma treatment has little impact on the adhesion, and therefore durability of the applied biopolymer film. A protein adsorption study on PET samples shows that solution concentration had no impact on fouling resistance of the coating. Further statistical analysis of the four groups of coated PET samples showed that the longer 24-hour incubation period did not affect the performance of the PC 1036 polymer coatings. The lack of change in biocompatibility due to the solution concentration is due to the consistency of the PC 1036 polymer itself. It has been previously reported that the ratio of the PC head group to the hydrophobic chain is related to the biocompatibility of the surface, and the interaction between the hydrophobic chain and the substrate had a great influence on fouling resistance (Ueda et al., 1992). By changing the solution concentration this ratio remains unaffected (i.e. the phosphorylcholine to hydrophobic chain ratio is constant in PC 1036).
15
The solution concentration did show an impact on the wettability of the surface. As reported countless times in the literature, the addition of PC increased the wettability, or decreased the contact angle, of the substrate, which has been correlated with a decrease in protein adsorption (Iwasaki and Saito, 2003; Ye et al., 2009b). The contact angle was measured on Ti6Al4V, and not on PET, where the biocompatibility was quantified. The two substrates vary enough to impact the morphology, adhesion, and wettability of the applied PC, which could account for solution concentration having an impact on wettability but not on biofouling. The whole factorial model in Fig. 4 explains a significant proportion of the variation in both the original surface characterization and the coating durability because the regression line and 95% confidence curves cross the sample mean, which is further quantified by the low p-values for each of the models. The two fit models for coating thickness, by means of FM intensity, and durability are dominated by the effects of solution concentration and dip rate, as can be seen from the standardized effects in Fig. 5. Furthermore, the plasma power and time are two factors that lie beneath statistical significance, as interpreted by the statistical model. The statistical fit model for the degree of cross-linking (p 0.1), which only explains 30% of the variation, is principally governed by the synergistic effects of solution concentration and cure temperature. While the fit model reports that the degree of cross-linking is inversely affected by solution concentration, cure temperature, and dip rate, all singular variables have no statistical significance on the cross-linking response. Thus, the optimization of the PC-coating and immobilization process will not be equivalent for both durability and degree of cross-linking. The original hypothesis or intention of this study was to optimize the process for both responses. From the statistical model, optimization of the degree of cross-linking results in a negative consequence on the coating durability twice as great as the resulting consequence would be if the optimization were reversed. Accordingly, it is recommended that optimization be performed for durability response. Coatings to improve the hemocompatibility of medical devices implanted for extended periods of time continue to be studied and evaluated, but the stability and durability of PC coatings on titanium substrates have not been qualified or quantified. A durable biomimetic coating will extend the hemocompatibility benefits of phosphorylcholine-based polymers immobilized on implantable titanium medical devices. Furthermore, optimizing the durability response holds lower risk than that of the degree of cross-linking, owing to the statistical fact that more variability was explained by the durability model than the degree of cross-linking model. While a significant database of statistical information was generated during this study, it is not without limitations. Planar coupons were used for PC-coating, immobilization, durability, and degree of cross-linking qualitative analysis. Processing and analyzing the coupons was relatively simple and repeatable; however, the wide, flat surface geometry is not directly applicable to a wide array of implantable medical devices for which a durable, biomimetic coating may be advantageous. While the responses from durability
16
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
and cross-linking tests are maintained, the effect of hard ( 901) edges and three-dimensional surface geometry on the distribution and immobilization of PC on titanium is not qualified in this study. No direct or indirect effects of blood and biological matter were considered during the evaluation of durability response of PC on titanium. Employment of a biomimetic material similar to blood may alter the statistical fit model for all responses. A small number of samples were analyzed using PBS as the working fluid for the spray tests. The PBS did wear away the PC coating at a faster rate than that of water, but the rate did not impact the overall statistical response of the model, and using it would have only decreased the time necessary for testing in a proportional manner while simultaneously increasing the wear on the spray device. Also, durability was studied by spraying sterile water perpendicular to the PC-coated surface under a steady-state condition. Realistic effects of pulsatile flow with varying shear stresses on the surface are not modeled here. The spray technique was developed as a method of eroding the PC polymer from the titanium substrate in a time-efficient manner. It is recommended that in order to evaluate and model the outcome of responses an elaborate test set-up could be designed such that the surfaces are subject to pulsatile shear flows and assessed at longer-term intervals for a more complete biological erosion system response. Our primary interest is in PC-coated Ti substrates exposed to fluid flow; however, Kyomoto et al. (2007a, 2007b, 2008a, 2008b) coated polyethylene surfaces with MPC and exposed them to mechanical wear in a ball-and-cup hip-joint simulator, and found increased wear resistance to polymeric surfaces coated with photo-irradiated MPC. Further, they reasoned that the degree of cross-linking of the MPC on polymer surface is directly proportional to the resistance to wear. Our study examines the impact of process factors on degree of crosslinking and durability as separate responses, which does not contradict the conclusion that cross-linked biopolymer is proportional to wear resistance. Despite the limitations, this study shows a multivariate statistical fit model to determine the process factors affecting durability of a biomimetic polymer coating on a metallic substrate. The power and efficiency of a statistically based design of experiments provide an advantage when testing several variables simultaneously. Implementation of silanated zwitterionic surface modifiers has been shown to increase the stability of covalently and chemically bonded organic films on inorganic surfaces (Matinlinna et al., 2004; Ye et al., 2009a). However, the use of silane coupling agents is not presently pursued. Loss of control of PC coating coverage and uniformity may result from additional process steps due to surface modification through silanization (Ye et al., 2010, 2013). It is desired to analyze the effects of RFGD plasma treatment without the addition of a silane material in order to model the improvement and optimization of a pre-existing process that was not subject to plasma treatment. It is proposed that the integration of plasma treatment to a dip coating process can be optimized in order to improve coating durability and degree of cross-linking, which could improve long-term reliability and biocompatibility of a modified Ti6Al4V surface.
5.
Conclusions
Ti-alloy surfaces were modified with a phospholipid polymer, PC 1036, by varying process factors, and a statistical model was built to qualify the impact of each variable on coating durability and degree of cross-linking. FM imaging demonstrated successful coverage of the titanium surface with PC. The resulting fit model showed the criticality of PC solution concentration, dip-application rate, and cure temperature for durability of the surface coating. No single factor displayed a statistically significant effect on the degree of cross-linking. Resistance to bio-fouling was not affected by changes in solution concentration. This statistical model may prove useful in the realistic application of PC to implantable medical devices.
r e f e r e n c e s
Chen, H., Zhang, Z., Chen, Y., Brook, M.A., Sheardown, H., 2005. Protein repellant silicone surfaces by covalent immobilization of poly(ethylene oxide). Biomaterials 26, 2391–2399. http://dx. doi.org/10.1016/j.biomaterials.2004.07.068. Chevallier, P., Castonguay, M., Turgeon, S., Dubrulle, N., Mantovani, D., McBreen, P.H., Wittmann, J.C., Laroche, G., 2001. Ammonia RF-plasma on PTFE surfaces: chemical characterization of the species created on the surface by vapor-phase chemical derivatization. J. Phys. Chem. B 105, 12490–12497. http://dx.doi.org/10.1021/jp011607k. Clarke, S., Davies, M.C., Roberts, C.J., Tendler, S.J.B., Williams, P. M., Lewis, A.L., O’Byrne, V., 2001. Atomic force microscope and surface Plasmon resonance investigation of polymer blends of poly ([2-(methacryloyloxy) ethyl] phosphorylcholine-co-lauryl methacrylate) and poly(laurylmethacrylate). Macromolecules 34, 4166–4172. http://dx.doi.org/10.1021/ma0016681. Colli, A., Verhoye, J.P., Leguerrier, A., Gherli, T., 2007. Anticoagulation or antiplatelet therapy of bioprosthetic heart valves recipients: an unresolved issue. Eur. J. Cardiothorac. Surg. 31, 573–577. http://dx.doi.org/10.1016/j.ejcts.2007.02.002. Gunaydin, S., Farsak, B., Kocakulak, M., Sari, T., Yorgancioglu, C., Zorlutuna, Y., 2002. Clinical performance and biocompatibility of poly (2-methoxyethylacrylate)-coated extracorporeal circuits. Ann. Thorac. Surg. 74, 819–824. http://dx.doi.org/10. 1016/S0003-4975(02)03796-7. Hayward, J.A., Chapman, D., 1984. Biomembrane surface as model for polymer design: the potential for heamocompatibility. Biomaterials 5, 135–142. http://dx.doi. org/10.1016/0142-9612(84)90047-4. Hayward, J.A., Durrani, A.A., Lu, Y., Clayton, C.R., Chapman, D., 1986. Biomembranes as models for polymer surfaces. IV. ESCA analyses of a phosphorylcholine surface covalently bound to hydroxylated substrates. Biomaterials 7 (252–253), 256–258. http://dx.doi.org/10.1016/0142-9612(86)90045-1. Ishihara, K., Aragaki, R., Ueda, T., Watanabe, A., Nakabayashi, N., 1990. Reduced thrombogenicity of polymers having phospholipids polar groups. J. Biomed. Mater. Res. 24, 1069–1077. http://dx.doi.org/10.1002/jbm.820240809. Ishihara, K., Nomura, H., Mihara, T., Kurita, K., Iwasaki, Y., Nakabayashi, N., 1998. Why do phospholipids polymers reduce protein adsorption?. J. Biomed. Mater. Res. 39, 323–330. http://dx.doi.org/10.1002/(SICI)1097-4636(199802)39:2%3c323:: AID-JBM21%3e3.0.CO;2-C. Iwasaki, Y., Saito, N., 2003. Immobilization of phosphorylcholine polymers to Ti-supported vinyldimethylsilyl monolayers and reduction of albumin adsorption. Coll. Surf. B: Biointerfaces 32, 77–84. http://dx.doi.org/10.1016/S0927-7765(03)00147-4.
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
John, R., Kamdar, F., Liao, K., Colvin-Adams, M., Boyle, A., Joyce, L., 2008. Improved survival and decreasing incidence of adverse events with the HeartMate II LVAD as bridge-totransplant therapy. Ann. Thorac. Surg. 86, 1227–1235. http:// dx.doi.org/10.1016/j.athoracsur.2008.06.030. Jones, M.I., McColl, I.R., Grant, D.M., Parker, K.G., Parker, T.L., 2000. Protein adsorption and platelet attachment and activation, on TiN, TiC, and DLC coatings on titanium for cardiovascular applications. J. Biomed. Mater. Res. A 52, 413–421. http://dx.doi. org/10.1002/1097-4636(200011)52:2%3c413::AID-JBM23%3e3.0. CO;2-U. Kyomoto, M., Moro, T., Konno, T., Takadama, H., Yamawaki, N., Kawaguchi, H., Takatori, Y., Nakamura, K., Ishihara, K., 2007a. Enhanced wear resistance of modified cross-linked polyethylene by grafting with poly(2-methacryloyloxyethyl phosphorylcholine). J. Biomed. Mater. Res. A 82, 10–17. http:// dx.doi.org/10.1002/jbm.a.31134. Kyomoto, M., Moro, T., Konno, T., Takadama, H., Kawaguchi, H., Takatori, Y., Nakamura, K., Yamawaki, N., Ishihara, K., 2007b. Effects of photo-induced graft polymerization of 2methacryloyloxyethyl phosphorylcholine on physical properties of cross-linked polyethylene in artificial hip joints. J. Mater. Sci.: Mater. Med. 18, 1809–1815. http://dx.doi.org/10. 1007/s10856-007-3063-0. Kyomoto, M., Moro, T., Miyaji, F., Hashimoto, M., Kawaguchi, H., Takatori, Y., Nakamura, K., Ishihara, K., 2008a. Effect of 2methacryloyloxyethyl phosphorylcholine concentration on photo-induced graft polymerization of polyethylene in reducing the wear of orthopaedic bearing surface. J. Biomed. Mater. Res. A 86, 439–447. http://dx.doi.org/10.1002/jbm.a.31511. Kyomoto, M., Moro, T., Miyaji, F., Konno, T., Hashimoto, M., Kawaguchi, H., Takatori, Y., Nakamura, K., Ishihara, K., 2008b. Enhanced wear resistance of orthopaedic bearing due to the cross-linking of poly (MPC) graft chains induced by gammaray irradiation. J. Biomed.. Mater. Res. Part B: Appl. Biomater. 84, 320–327. http://dx.doi.org/10.1002/jbm.b.30874. Lewis, A.L., 2000. Phosphorylcholine-based polymers and their use in the prevention of biofouling. Coll. Surf. B: Biointerfaces 18, 261–275. http://dx.doi.org/10.1016/S0927-7765(99)00152-6. Lewis, A.L., Cumming, Z.L., Goreshi, H.H., Kirkwood, L.C., Tolhurst, L.A., Stratford, P.W., 2001. Crosslinkable coatings from phosphorylcholine-based polymer. Biomaterials 22, 99–111. http://dx.doi.org/10.1016/S0142-9612(00)00083-1. Lewis, A.L., Tolhurst, L.A., Stratford, P.W., 2002. Analysis of a phosphorylcholine-based polymer coating on a coronary stent pre- and post-implantation. Biomaterials 23, 1697–1706. http://dx.doi.org/10.1016/S0142-9612(01)00297-6. Lewis, A.L., Willis, S.L., Small, S.A., Hunt, S., O’Byrne, V., Stratford, P. W., 2004. Drug loading and elution from a phosphorylcholine polymer coated coronary stent does not affect long-term stability of the coating in vivo. Biomed. Mater. Eng. 14, 355–370. Liu, X., Chu, P.K., Ding, C., 2004. Surface modification of titanium, titanium alloys, and related materials for biomedical applications. Mater. Sci. Eng.: R: Rep. 47, 49–121. http://dx.doi. org/10.1016/j.mser.2004.11.001. Matinlinna, J.P., Laajalehto, K., Laiho, T., Kangasniemi, I., Lassila, L.V.J., Vallittu, P.K., 2004. Surface analysis of Co–Cr–Mo alloy and Ti substrates silanizated with trialkoxysilanes and silane mixtures. Surf. Interface Anal. 36, 246–253. http://dx.doi.org/ 10.1002/sia.1681.
17
Ozeki, K., Masuzawa, T., Hirakuri, K.K., 2010. The wear properties and adhesion strength of the diamond-like carbon film coated on SUS, Ti and Ni–Ti with plasma pre-treatment. Biomed. Mater. Eng. 20, 21–35. http://dx.doi.org/10.3233/ BME-2010-0612. Sin, D.C., Kei, H.L., Miao, X., 2005. Surface coatings for ventricular assist devices. Expert Rev. Med. Devices 6, 51–60. http://dx.doi. org/10.1586/17434440.6.1.51. Tang, Y., Lu, J.R., 2001. Swelling of zwitterionic polymer films characterized by spectroscopic ellipsometry. Macromolecules 34, 8768–8776. http://dx.doi.org/10.1021/ma010476i. Ueda, M., Silva, M.M., Otani, C., Reuther, H., Yatsuzuka, M., Lepienski, C.M., Berni, L.A., 2003. Improvement of tribological properties of Ti6Al4V by nitrogen plasma immersion ion implantation. Surf. Coat. Technol. 170, 408–410. http://dx.doi. org/10.1016/S0257-8972(03)00137-3. Ueda, T., Oshida, H., Kurita, K., Ishihara, K., Nakabayashi, N., 1992. Preparation of 2-methacryloyloxyethyl phosphorylcholine copolymers with alkyl methacrylates and their blood compatibility. Polym. J. 24, 1259–1269. http://dx.doi. org/10.1295/polymj.24.1259. Wang, J., Bartlett, J.D., Dunn, A.C., Small, S., Willis, S.L., Driver, M. J., Lewis, A.L., 2005. The use of rhodamine 6G and fluorescence microscopy in the evaluation of phospholipid-based polymeric biomaterials. J. Microsc. 217, 216–224. http://dx.doi. org/10.1111/j.1365-2818.2005.01453.x. Yasuda, H., Gazicki, M., 1982. Biomedical applications of plasma polymerization and plasma treatment of polymer surfaces. Biomaterials 3, 68–77. http://dx.doi.org/10.1016/0142-9612(82) 90036-9. Ye, S.H., Johnson, C.A., Woolley, J.R., Snyder, T.A., Gamble, L.J., Wagner, W.R., 2009a. Covalent surface modification of a titanium alloy with a phosphorylcholine-containing copolymer for reduced thrombogenicity in cardiovascular devices. J. Biomed. Mater. Res. A 91, 18–28. http://dx.doi.org/10. 1002/jbm.a.32184. Ye, S.H., Johnson, C.A., Woolley, J.R., Oh, H.-I., Gamble, L.J., Ishihara, K., Wagner, W.R., 2009b. Surface modification of a titanium alloy with a phospholipid polymer prepared by a plasma-induced grafting technique to improve surface thromboresistance. Coll. Surf. B: Biointerfaces 74, 96–102. http://dx.doi.org/10.1016/j.colsurfb.2009.06.032. Ye, S.H., Johnson, C.A., Woolley, J.R., Murata, H., Gamble, L.J., Ishihara, K., Wagner, W.R., 2010. Simple surface modification of a titanium alloy with silanated zwitterionic phosphorylcholine or sulfobetaine modifiers to reduce thrombogenicity. Coll. Surf. B: Biointerfaces 79, 357–364. http://dx.doi.org/10.1016/j.colsurfb.2010.04.018. Ye, S.H., Jang, Y.S., Yun, Y.H., Shankarraman, V., Woolley, J.R., Hong, Y., Gamble, L.J., Ishihara, K., Wagner, W.R., 2013. Surface modification of a biodegradable magnesium alloy with phosphorylcholine (PC) and sulfobetaine (SB) functional macromolecules for reduced thrombogenicity and acute corrosion resistance. Langmuir 29, 8320–8327. http://dx.doi. org/10.1021/la401341y. Zhang, F., Kang, E.T., Neoh, K.G., Wang, P., Tan, K.L., 2001. Surface modification of stainless steel by grafting of poly (ethylene glycol) for reduction in protein adsorption. Biomaterials 22, 1541–1548. http://dx.doi.org/10.1016/ S0142-9612(00)00310-0.
Available online at www.sciencedirect.com
www.elsevier.com/locate/jmbbm
Research Paper
Durability evaluation of biopolymer coating on titanium alloy substrate J. Ryan Stanfieldn, Stacy Bamberg Departments of Mechanical Engineering and Bioengineering, University of Utah, 50 S Central Campus Dr., Salt Lake City, UT 84112, USA
art i cle i nfo
ab st rac t
Article history:
For this study, a commercially available phosphorylcholine (PC) polymer was applied to
Received 19 January 2014
Ti6Al4V ELI. A multivariate approach to design a statistically significant array of experiments
Received in revised form
was employed to evaluate and estimate optimization of PC-immobilization process factors.
11 March 2014
The seven process factors analyzed were (1) power level for RFGD plasma treatment, (2)
Accepted 13 March 2014
duration of plasma treatment, (3) concentration of PC solution used to coat samples, (4) rate at
Available online 21 March 2014
which samples were dipped in/out of the solution, (5) temperature for curing, (6) relative
Keywords:
humidity level during curing, and (7) duration of curing. Imaging and analysis of the coating
Titanium alloy
were done via fluorescence microscopy (FM), confirming the uniform coverage of PC polymer
Phospholipid polymer
on titanium substrate. The process factors were evaluated by three measured responses:
Durability
initial thickness, coating durability and degree of cross-linked coating, which were assessed by
Blood compatibility
FM, a spray test and extraction in IPA, respectively. Variations in PC solution concentration
Implantable medical devices
showed no impact on fouling resistance of the resultant coating. It was hypothesized that the PC-application process factors could be optimized to yield favorable outcomes in durability and degree of cross-linked coating responses. The resulting statistical model indicates that PC solution concentration, dip rate, and cure temperature are the three greatest singular effects on both durability and degree of cross-linking. In addition, plasma treatment of the substrate with O2 was effective in enhancing the degree of cross-linking of the polymer surface. & 2014 Elsevier Ltd. All rights reserved.
1.
Introduction
Titanium and its alloys, in addition to stainless steels, are widely used in surgical implant applications. Cardiovascular treatments, such as vascular stents, heart valves, and ventricular assist devices (VADs), have continued to promote the development of surface modification methods in order to reduce platelet deposition on metallic surfaces (Liu et al., 2004). Anticoagulation and antiplatelet n
Corresponding author. Tel.: þ1 801 581 6441; fax: þ1 801 585 9826. E-mail address: ryan.stanfi[email protected] (J. Ryan Stanfield).
http://dx.doi.org/10.1016/j.jmbbm.2014.03.003 1751-6161/& 2014 Elsevier Ltd. All rights reserved.
pharmacologic therapies are frequently used on patients implanted with such apparatuses due to the risk of thrombosis and thromboembolism (Colli et al., 2007; John et al., 2008). To improve the thromboresistance of titanium alloys in contact with the cardiovascular system, various surface coatings have been developed such as diamond-like carbon (DLC), titanium nitride (TiN), heparin, silicone, and 2-methacryloyloxyethel phosphorylcholine (MPC) (Sin et al., 2005). Exceptional biocompatibility and hemocompatibility have been shown on
10
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
biomimetic surfaces synthesized from phospholipid polymers [phosphorylcholine (PC)] by reducing protein adsorption and platelet adhesion, as the PC head group is prevalent on the outside of biological cell membranes (Hayward and Chapman, 1984; Ishihara et al., 1990, 1998; Lewis 2000; Lewis et al., 2001). In addition to coatings, physical and chemical surface treatments have been utilized on titanium alloys including passivation, polishing, oxidation, vapor deposition, silanization, glow discharge plasma treatment, and ion implantation (Liu et al., 2004). Plasma treatment and ion implantation have been shown to increase the hardness and wear resistance of metallic substrates (Ueda et al., 2003). Moreover, plasma treatment has been shown to improve biocompatibility and increase adhesion of different materials (Yasuda and Gazicki, 1982). Radio frequency glow discharge (RFGD) plasma treatment is frequently used for surface cleaning, and only introduces ions and electrons, as opposed to bulk materials (Ueda et al., 2003). The deposited ions can act as functional groups to covalently attach other polymers or biomolecules (Chevallier et al., 2001). Covalent linkage of PC to a substrate is more stable than physically adsorbed PC. Thus, modification of a surface in order to promote the covalent connection of PC to a metallic substrate is highly desirable for long-term medical implant applications. Inorganic coatings on metallic surfaces, such as diamondlike carbon (DLC), titanium-oxides, and nitrides, which show evidence of mechanical and chemical stability, and high inertness, are employed to increase hemocompatibility of the base materials used in surgical devices. The hydrophobicity and low surface roughness exhibited on DLC-coated substrates have shown a higher proportion of albumin than fibrinogen in thrombogenic studies (Jones et al., 2000). However, DLC can be difficult to apply to the enclosed or undercut surfaces present on many implantable devices. Surface adhesion of DLC to substrates has been shown to be low without plasma pretreatment (Ozeki et al., 2010), as with MPC-coatings. Organic coatings are also used on metallic surfaces to improve hemocompatibility. PEO (Chen et al., 2005), PMEA (Gunaydin et al., 2002), PEG (Zhang et al., 2001), and other polymer coatings have been shown to reduce platelet adhesion and aggregation, and protein adsorption. However, durability, adhesion, the modification method, and efficacy impact the value of coatings, and must be considered on long-term surgical implants. While biomimetic organic polymers, such as PC, show promise for improvement of hemocompatibility, techniques to optimize immobilization of the material onto metals significant to implantable devices have not been widely evaluated for durability. Characterization of PC material coated on substrates has been reported using X-ray photoelectron spectroscopy (XPS) (Hayward et al., 1986), atomic force microscopy (AFM) (Clarke et al., 2001), scanning electron microscopy (SEM) (Lewis et al., 2002), and Fourier-transform infrared (FTIR) microscopy (Lewis et al., 2004). Furthermore, spectroscopic ellipsometry (SE) and fluorescence microscopy (FM) analysis techniques have been used to analyze the presence, distribution, thickness and swelling of PC films on titanium, polypropylene, silicone, and other materials (Tang and Lu, 2001; Wang et al., 2005). Using FM analysis, the intensity of the reflected light is shown to be proportional to the thickness of the PC
Fig. 1 – PC film thickness and FM intensity correlation as a function of solution concentration. film, as quantified by SE. The linear relationship is depicted in Fig. 1. The primary objective in this study was to examine the process of titanium alloy (Ti6Al4V) surface modification for hemocompatibility, and assess which variables had the greatest impact on durability and integrity of the surface coating. A simple surface modification process was developed implementing RFGD plasma treatment. The parameters for plasma treatment, surface coating and curing of PC polymer were varied for statistical analyses. A technique was developed to aggressively erode the surface coating and to investigate the removal of the coating. Additionally, other analyses were carried out on surface coating thickness, coverage, degree of cross-linking by FM, surface wettability, and surface coating biocompatibility via in vitro protein adsorption analyses. A commercially available software was implemented in order to create a feasible design of experiments (DOE) test matrix, and to model and analyze each response.
2.
Materials and methods
2.1.
Materials
Titanium alloy, Ti6Al4V extra low interstitial (ELI) as per ASTM standard F136, was procured (Titanium Metal Supply, Poway, CA), machined to an average surface roughness (Ra) of 400 nm, passivated per ASTM standard F86, and polished to Ra of 100 nm with 400, pumice, 6 μm, and 1 μm grade ceramic aluminum oxide radial bristle disks (3 M, St. Paul, MN). Digesil NC (RPM Technology, Reno, NV) was used as a silicone stripper post-polishing. A phospholipid copolymer, PC 1036, was obtained from Vertellus Biomaterials (Hampshire, UK). Rhodamine 6G (Sigma-Aldrich, St. Louis, MO) was used as a staining agent for fluorescent visualization.
2.2.
Surface pretreatment
After polishing, coupons were rinsed with distilled water and isopropyl alcohol (IPA), cleaned ultrasonically for 90 min in a 20 wt% phenol solution in dichloromethane (DCM), and solvent washed ultrasonically two times for 90 min each in DCM. The Ti6Al4V surfaces were pretreated by O2 plasma with RFGD (Plasma Technology Systems, Belmont, CA) for
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
1–5 min in a 0.3 T vacuum chamber. The applied plasma power ranged from 100 to 600 W at a frequency of 13.65 MHz. Control coupons did not undergo plasma pre-treatment.
2.3.
Phospholipid polymer application
Under ISO Class 8 air, the plasma-treated coupons were dipped at a rate from 1 to 19 mm/s in argon-blanketed 0.25– 2.25 wt% PC solution in absolute ethanol (EtOH). Then the solvent was allowed to evaporate for 20 min. The PC-coated coupons were cured in an environmental test chamber (Tenney, New Columbia, PA) for 4–16 h at 40–100 1C and 25–75% relative humidity (%RH), depending on the DOE test condition. The control coupons were dipped at a rate of 10 mm/s in 1.25 wt% PC solution and were cured for 4 h at 70 1C and 50% RH.
2.4.
Surface characterization
PC-coated titanium alloy surfaces were stained for 30 s in a solution of 0.02 wt% rhodamine 6G in distilled water, rinsed twice each for 30 s in distilled water, and dried at room temperature (RT) conditions for 2 h, as described by Wang et al. (2005). The stained coupons were imaged using a fluorescence microscope (Olympus, Center Valley, PA) and digital camera (Lumenera, Ottawa, ON). The images were analyzed with ImageJ software (National Institutes of Health). ImageJ was used to calculate the mean intensity value across all of the pixels in each image. Silicon wafers were used as a control for film thickness measurement using a variable angle spectroscopic ellipsometer (JA Woollam, Lincoln, NE) to show FM measurement correlation. Silicon wafers were coated and analyzed using the same methods as the Ti-alloy control coupons.
2.5.
agitated for 24 h. Any PC polymer chains that had not been cross-linked through the cure process would theoretically wash away during this operation. After the IPA agitation, coupons were dried for 4 h in a convection oven at 40 1C. The coupons were stained, imaged and analyzed as per the above process. FM intensity values of extracted and non-extracted coupons were compared to find a net change in value after the IPA agitation.
2.7.
Surface wettability and protein adsorption
Prior to FM analysis, the static contact angle of water on a selection of coated and uncoated coupons was measured with a contact angle goniometer (190-F1; Ramé-Hart Instrument Co., Succasunna, NJ). Five PC-coated samples were prepared similar to those previously described as control coupons with the exception of PC solution concentration. The manufacturersupplied hardware and software were utilized to calculate the contact angle of a 2 mL droplet of distilled water with five measurements per sample. In order to evaluate the validity of different PC solution concentrations, surface protein adsorption on coated and uncoated polyethylene terephthalate (PET) samples was gaged by a double antibody binding assay. PET strips were prepared similar to control coupons, but used both 0.25 and 2.25 wt% PC solutions in EtOH. Samples were incubated in human plasma (ProteoGenex, Culver City, CA) for 10 min or 24 h, washed four times in phosphate-buffered saline (PBS; BD Biosciences, San Jose, CA), incubated in anti-human fibrinogen (Sigma-Aldrich) for 30 min, washed four times in PBS, incubated in anti-goat IgG peroxidase conjugate (SigmaAldrich) for 30 min, washed four times in PBS, and incubated in o-phenylenediamine dihydrochloride (Sigma-Aldrich) for 10 min. Absorbance was read and recorded spectrophotometrically at 450 nm.
Surface coating durability 2.8.
Initial work to characterize coating adhesion and durability by means of a standardized adhesion test (ASTM D3359) was unfruitful; therefore a spray test was developed to aggressively erode a coating from a metallic substrate. In order to hydrate the PC-coated surface prior to the durability test, coupons were soaked in sterile water for 24 h. A spray gun (Binks, Glendale Heights, IL) was mounted 5 mm from the surface and the coupon was sprayed with sterile water in four distinct locations for 2, 20, 30, and 60 min per location. An input air pressure on the spray gun of 414 kPa (60 psi) produced a force of 0.3 N perpendicular to the coupon surface. Coupons were then stained, imaged, and analyzed, as per the above process. With the ImageJ software the area of the removed PC was measured at all four locations per coupon. Spray time vs. area was plotted and a logarithmic trend line was fit to each coupon data set. The area under the logarithmic curve was integrated in order to define a metric for coating durability, where lower area implied higher durability.
2.6.
11
Degree of surface coating cross-linking
Because the PC polymer is soluble in IPA, as well as EtOH, a selection of coupons was placed in HPLC-grade IPA and
Design of experiments
JMPs Statistical Discovery Software (SAS, Cary, NC) was used to create a randomized quarter fraction screening design for seven experimental factors. The DOE test matrix required 38 runs to be carried out for each of the three responses, so 114 coupons were prepared with the above process. The range for each experimental factor is shown in Table 1. Additionally, 30 control coupons were prepared with the process outlined above, and are tabulated below. Note that the control coupons were not subjected to surface modification via plasma treatment.
3.
Results
3.1.
Surface characterization
We show the linear correlation between solution concentration and film thickness in Fig. 1, and the corresponding relationship with FM intensity values. Similar to the results of Wang et al., surface coating uniformity and morphology can be analyzed via the FM images presented in Fig. 2, where solution concentration and dip rate have been varied. Fig. 3a
12
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
Table 1 – Experimental factors for multivariate statistical testing with associated ranges and control values. Var
Factors
Low
High
Control
Units
A B C D E F G
Plasma time Plasma power Dip rate PC solution concentration Cure temperature Cure time Cure environment
1 100 1 0.25 40 4 25
5 600 19 2.25 100 16 75
– – 10 1.25 70 4 50
min W mm/s wt% 1C h %RH
Fig. 2 – Uniformity and morphology of surfaces coated with varying solution concentrations and extraction rates. Top row: 1.25 wt% PC in EtOH at (a) 1, (b) 10, and (c) 19 mm/s extraction rates. Bottom row: 10 mm/s extraction rate at (a) 0.25, (b) 1.25, and (c) 2.25 wt% PC in EtOH solution. Scale bars are 1 mm.
displays the mean intensity value for each group of coupons processed under distinct coating conditions, where FM intensities are proportional to both dip rate and solution concentration. The JMPs Statistical Discovery Software built a model based on the results from the mean intensity value data for each of the test conditions from the DOE matrix. The actual by predicted plot from the model is shown in Fig. 4a. The model further estimates the singular and coupled effects on the FM intensity, which correspond to thickness. A Pareto chart containing the t-ratios (standardized effect) for each of the seven factors on FM intensity is displayed in Fig. 5a, with the highest two singular effects being solution concentration and dip rate. The statistical significance of each effect is tested by comparing the mean square against an estimate of the experimental error. In this case two effects, solution concentration and dip rate, have P-values less than 0.05 (po0.0001), indicating that they are significantly different from zero at the 95.0% confidence level. Those effects described as positive increase the FM intensity or thickness value.
3.2.
Coating durability
A photograph and the corresponding FM images of a titanium alloy coupon coated with PC solution and sprayed for durability are shown in Fig. 6. Fig. 7 displays a series of spray
time vs. area points for the control coupons with logarithmic trend lines. Samples that yielded trend lines below the controls were projected to be more durable, while trend lines above were predicted to be less durable than the controls. The actual by predicted plot from the model for coating durability is shown in Fig. 4b. The model further estimates the singular and coupled effects on the coating durability. The Pareto chart containing the t-ratios for each of the seven factors on coating durability is displayed in Fig. 5b, with the highest two singular effects being solution concentration and dip rate. In this case, three effects, solution concentration, dip rate, and solution concentration coupled with cure temperature, have statistical significance with P-values less than 0.05. An optimal process to maximize the coating durability was predicted from the model and is presented in Table 2.
3.3.
Coating cross-linking
Fig. 3b displays the mean intensity value for each group of extracted coupons processed under distinct coating conditions, where the resulting FM intensities are proportional to dip rate, but not solution concentration. The mean FM intensity for the control coupons decreased by 25% after IPA-based extraction; the intensities for all other groups shown in Fig. 3 decreased by an average of 21% post-extraction.
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
13
The actual by predicted plot from the model for degree of cross-linking is shown in Fig. 4c. The model further estimates the singular and coupled effects on the degree of cross-linking. The Pareto chart containing the standardized effects for each of the seven factors is displayed in Fig. 5c, which shows that no singular effect has significant impact on the degree of cross-linking. However, one coupled effect, solution concentration with cure temperature, has statistical
significance with a P-value less than 0.01. An optimal process to maximize the degree of cross-linking was predicted from the model and is presented in Table 2.
Fig. 3 – Mean fluorescence microscopy image intensities for varying PC solution concentrations and dip rates. (a) Titanium alloy coupons coated, cured, and imaged, and (b) titanium alloy coupons coated, cured, extracted in IPA, and imaged.
Fig. 5 – Pareto charts display standardized effect (t-ratio) value along the horizontal axis for all seven singular factors as identified in the tables for (a) surface characterization, (b) durability response, and (c) degree of cross-linking response.
Fig. 4 – Statistical fit models for (a) surface characterization, (b) durability response, and (c) degree of cross-linking response show actual by predicted values for each. Horizontal dotted mean lines, angled solid regression lines, and 95% dotted confidence curves.
14
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
Fig. 6 – Titanium alloy coupon after undergoing the spray test with corresponding FM images. Beginning in the upper left-hand corner and going clockwise the four locations are sprays for increasing time durations of 2, 20, 30, and 60 min.
Table 2 – Statistical predictions for process optimization for both coating durability and degree of cross-linking. Var
Factor
Durability
Cross-link
A B C D E F G
Plasma time Plasma power Withdrawal rate Solution concentration Cure temp Cure time Cure humidity
Low Low High High High High High
High High Low Low Low High High
Fig. 7 – Area (pixels) by time (min) for measured spray locations on a sample batch of control coupons with logarithmic trend lines. Area under the logarithmic curve represents durability response value, where lower areaunder-the-curve values represent higher durability.
3.4.
Surface wettability and protein adsorption
The wettability of each surface was measured five times per sample on modified and unmodified Ti6Al4V coupons using a static contact angle goniometer. The contact angle of a 2 mL droplet on unmodified coupons was 87721, while for coupons modified with 0.25 and 2.25 wt% PC solution, the angles were 64721 and 60731, respectively. The amount of fibrinogen adsorbed on the PC-coated and uncoated PET strips was measured by the absorbance of the reacted o-phenylenediamine dihydrochloride at 450 nm. The amount of adsorbed fibrinogen on the PET samples is shown in Fig. 8. The average absorbance of the uncoated PET was much higher than those of the coated samples, and the
Fig. 8 – Fibrinogen adsorption on uncoated and PC 1036 coated PET samples.
reduction of fibrinogen adsorption was 85–87% for samples coated with 0.25 or 2.25 wt% PC solution. This considerable difference was verified by student t-tests comparing the coated and uncoated samples (po0.0001).
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
4.
Discussion
For this study, a commercially available PC-polymer (PC 1036) was applied to Ti6Al4V ELI. A multivariate approach to design a statistically significant array of experiments was employed to evaluate and estimate optimization of PC-immobilization process factors. The seven process factors analyzed were power level for RFGD plasma treatment, duration of plasma treatment, concentration of PC solution used to coat samples, rate at which samples were dipped in/out of the solution, temperature for curing, relative humidity level during curing, and duration of curing. The responses were coating durability and degree of cross-linked coating, which were assessed by spray testing and extraction in IPA, respectively. Imaging and analysis of the coating were done via FM. Evaluation on a set of samples by staining and imaging with fluorescence microscopy showed that all coupons had complete coverage and uniform distribution of the PC 1036 polymer. The average pixel intensity across the images was proportional to the solution concentration and the dip rate. The efficacy of the cross-link characterization test was demonstrated with the average pixel intensity of FM images decreasing by 21% post-extraction in IPA. While the FM intensity of the control coupons dropped by 25% postextraction in IPA, the coupons that underwent surface pretreatment with plasma and were processed with the same dip rate and solution concentration conditions as the controls only dropped by 7% post-extraction, validating the benefit of the RFGD plasma treatment relative to the degree of coating cross-linking. The results from the spray testing did not exhibit a significant benefit to the durability of the coating with plasma treatment. Indeed, no significant difference was seen between durability data for coupons processed with and without plasma treatment under the same solution concentration and dip rate conditions. This may be attributed to the type of bond formed by the PC polymer and the substrate. The surface of the substrate is a thin titanium oxide layer formed during passivation, rather than the titanium alloy. Employing a solvent-evaporation technique to apply the polymer with a suspended solution directly to an oxide layer likely produces a noncovalent or intermolecular bond. Thus, inclusion of RFGD plasma treatment has little impact on the adhesion, and therefore durability of the applied biopolymer film. A protein adsorption study on PET samples shows that solution concentration had no impact on fouling resistance of the coating. Further statistical analysis of the four groups of coated PET samples showed that the longer 24-hour incubation period did not affect the performance of the PC 1036 polymer coatings. The lack of change in biocompatibility due to the solution concentration is due to the consistency of the PC 1036 polymer itself. It has been previously reported that the ratio of the PC head group to the hydrophobic chain is related to the biocompatibility of the surface, and the interaction between the hydrophobic chain and the substrate had a great influence on fouling resistance (Ueda et al., 1992). By changing the solution concentration this ratio remains unaffected (i.e. the phosphorylcholine to hydrophobic chain ratio is constant in PC 1036).
15
The solution concentration did show an impact on the wettability of the surface. As reported countless times in the literature, the addition of PC increased the wettability, or decreased the contact angle, of the substrate, which has been correlated with a decrease in protein adsorption (Iwasaki and Saito, 2003; Ye et al., 2009b). The contact angle was measured on Ti6Al4V, and not on PET, where the biocompatibility was quantified. The two substrates vary enough to impact the morphology, adhesion, and wettability of the applied PC, which could account for solution concentration having an impact on wettability but not on biofouling. The whole factorial model in Fig. 4 explains a significant proportion of the variation in both the original surface characterization and the coating durability because the regression line and 95% confidence curves cross the sample mean, which is further quantified by the low p-values for each of the models. The two fit models for coating thickness, by means of FM intensity, and durability are dominated by the effects of solution concentration and dip rate, as can be seen from the standardized effects in Fig. 5. Furthermore, the plasma power and time are two factors that lie beneath statistical significance, as interpreted by the statistical model. The statistical fit model for the degree of cross-linking (p 0.1), which only explains 30% of the variation, is principally governed by the synergistic effects of solution concentration and cure temperature. While the fit model reports that the degree of cross-linking is inversely affected by solution concentration, cure temperature, and dip rate, all singular variables have no statistical significance on the cross-linking response. Thus, the optimization of the PC-coating and immobilization process will not be equivalent for both durability and degree of cross-linking. The original hypothesis or intention of this study was to optimize the process for both responses. From the statistical model, optimization of the degree of cross-linking results in a negative consequence on the coating durability twice as great as the resulting consequence would be if the optimization were reversed. Accordingly, it is recommended that optimization be performed for durability response. Coatings to improve the hemocompatibility of medical devices implanted for extended periods of time continue to be studied and evaluated, but the stability and durability of PC coatings on titanium substrates have not been qualified or quantified. A durable biomimetic coating will extend the hemocompatibility benefits of phosphorylcholine-based polymers immobilized on implantable titanium medical devices. Furthermore, optimizing the durability response holds lower risk than that of the degree of cross-linking, owing to the statistical fact that more variability was explained by the durability model than the degree of cross-linking model. While a significant database of statistical information was generated during this study, it is not without limitations. Planar coupons were used for PC-coating, immobilization, durability, and degree of cross-linking qualitative analysis. Processing and analyzing the coupons was relatively simple and repeatable; however, the wide, flat surface geometry is not directly applicable to a wide array of implantable medical devices for which a durable, biomimetic coating may be advantageous. While the responses from durability
16
journal of the mechanical behavior of biomedical materials 35 (2014) 9 –17
and cross-linking tests are maintained, the effect of hard ( 901) edges and three-dimensional surface geometry on the distribution and immobilization of PC on titanium is not qualified in this study. No direct or indirect effects of blood and biological matter were considered during the evaluation of durability response of PC on titanium. Employment of a biomimetic material similar to blood may alter the statistical fit model for all responses. A small number of samples were analyzed using PBS as the working fluid for the spray tests. The PBS did wear away the PC coating at a faster rate than that of water, but the rate did not impact the overall statistical response of the model, and using it would have only decreased the time necessary for testing in a proportional manner while simultaneously increasing the wear on the spray device. Also, durability was studied by spraying sterile water perpendicular to the PC-coated surface under a steady-state condition. Realistic effects of pulsatile flow with varying shear stresses on the surface are not modeled here. The spray technique was developed as a method of eroding the PC polymer from the titanium substrate in a time-efficient manner. It is recommended that in order to evaluate and model the outcome of responses an elaborate test set-up could be designed such that the surfaces are subject to pulsatile shear flows and assessed at longer-term intervals for a more complete biological erosion system response. Our primary interest is in PC-coated Ti substrates exposed to fluid flow; however, Kyomoto et al. (2007a, 2007b, 2008a, 2008b) coated polyethylene surfaces with MPC and exposed them to mechanical wear in a ball-and-cup hip-joint simulator, and found increased wear resistance to polymeric surfaces coated with photo-irradiated MPC. Further, they reasoned that the degree of cross-linking of the MPC on polymer surface is directly proportional to the resistance to wear. Our study examines the impact of process factors on degree of crosslinking and durability as separate responses, which does not contradict the conclusion that cross-linked biopolymer is proportional to wear resistance. Despite the limitations, this study shows a multivariate statistical fit model to determine the process factors affecting durability of a biomimetic polymer coating on a metallic substrate. The power and efficiency of a statistically based design of experiments provide an advantage when testing several variables simultaneously. Implementation of silanated zwitterionic surface modifiers has been shown to increase the stability of covalently and chemically bonded organic films on inorganic surfaces (Matinlinna et al., 2004; Ye et al., 2009a). However, the use of silane coupling agents is not presently pursued. Loss of control of PC coating coverage and uniformity may result from additional process steps due to surface modification through silanization (Ye et al., 2010, 2013). It is desired to analyze the effects of RFGD plasma treatment without the addition of a silane material in order to model the improvement and optimization of a pre-existing process that was not subject to plasma treatment. It is proposed that the integration of plasma treatment to a dip coating process can be optimized in order to improve coating durability and degree of cross-linking, which could improve long-term reliability and biocompatibility of a modified Ti6Al4V surface.
5.
Conclusions
Ti-alloy surfaces were modified with a phospholipid polymer, PC 1036, by varying process factors, and a statistical model was built to qualify the impact of each variable on coating durability and degree of cross-linking. FM imaging demonstrated successful coverage of the titanium surface with PC. The resulting fit model showed the criticality of PC solution concentration, dip-application rate, and cure temperature for durability of the surface coating. No single factor displayed a statistically significant effect on the degree of cross-linking. Resistance to bio-fouling was not affected by changes in solution concentration. This statistical model may prove useful in the realistic application of PC to implantable medical devices.
r e f e r e n c e s
Chen, H., Zhang, Z., Chen, Y., Brook, M.A., Sheardown, H., 2005. Protein repellant silicone surfaces by covalent immobilization of poly(ethylene oxide). Biomaterials 26, 2391–2399. http://dx. doi.org/10.1016/j.biomaterials.2004.07.068. Chevallier, P., Castonguay, M., Turgeon, S., Dubrulle, N., Mantovani, D., McBreen, P.H., Wittmann, J.C., Laroche, G., 2001. Ammonia RF-plasma on PTFE surfaces: chemical characterization of the species created on the surface by vapor-phase chemical derivatization. J. Phys. Chem. B 105, 12490–12497. http://dx.doi.org/10.1021/jp011607k. Clarke, S., Davies, M.C., Roberts, C.J., Tendler, S.J.B., Williams, P. M., Lewis, A.L., O’Byrne, V., 2001. Atomic force microscope and surface Plasmon resonance investigation of polymer blends of poly ([2-(methacryloyloxy) ethyl] phosphorylcholine-co-lauryl methacrylate) and poly(laurylmethacrylate). Macromolecules 34, 4166–4172. http://dx.doi.org/10.1021/ma0016681. Colli, A., Verhoye, J.P., Leguerrier, A., Gherli, T., 2007. Anticoagulation or antiplatelet therapy of bioprosthetic heart valves recipients: an unresolved issue. Eur. J. Cardiothorac. Surg. 31, 573–577. http://dx.doi.org/10.1016/j.ejcts.2007.02.002. Gunaydin, S., Farsak, B., Kocakulak, M., Sari, T., Yorgancioglu, C., Zorlutuna, Y., 2002. Clinical performance and biocompatibility of poly (2-methoxyethylacrylate)-coated extracorporeal circuits. Ann. Thorac. Surg. 74, 819–824. http://dx.doi.org/10. 1016/S0003-4975(02)03796-7. Hayward, J.A., Chapman, D., 1984. Biomembrane surface as model for polymer design: the potential for heamocompatibility. Biomaterials 5, 135–142. http://dx.doi. org/10.1016/0142-9612(84)90047-4. Hayward, J.A., Durrani, A.A., Lu, Y., Clayton, C.R., Chapman, D., 1986. Biomembranes as models for polymer surfaces. IV. ESCA analyses of a phosphorylcholine surface covalently bound to hydroxylated substrates. Biomaterials 7 (252–253), 256–258. http://dx.doi.org/10.1016/0142-9612(86)90045-1. Ishihara, K., Aragaki, R., Ueda, T., Watanabe, A., Nakabayashi, N., 1990. Reduced thrombogenicity of polymers having phospholipids polar groups. J. Biomed. Mater. Res. 24, 1069–1077. http://dx.doi.org/10.1002/jbm.820240809. Ishihara, K., Nomura, H., Mihara, T., Kurita, K., Iwasaki, Y., Nakabayashi, N., 1998. Why do phospholipids polymers reduce protein adsorption?. J. Biomed. Mater. Res. 39, 323–330. http://dx.doi.org/10.1002/(SICI)1097-4636(199802)39:2%3c323:: AID-JBM21%3e3.0.CO;2-C. Iwasaki, Y., Saito, N., 2003. Immobilization of phosphorylcholine polymers to Ti-supported vinyldimethylsilyl monolayers and reduction of albumin adsorption. Coll. Surf. B: Biointerfaces 32, 77–84. http://dx.doi.org/10.1016/S0927-7765(03)00147-4.
journal of the mechanical behavior of biomedical materials 35 (2014) 9–17
John, R., Kamdar, F., Liao, K., Colvin-Adams, M., Boyle, A., Joyce, L., 2008. Improved survival and decreasing incidence of adverse events with the HeartMate II LVAD as bridge-totransplant therapy. Ann. Thorac. Surg. 86, 1227–1235. http:// dx.doi.org/10.1016/j.athoracsur.2008.06.030. Jones, M.I., McColl, I.R., Grant, D.M., Parker, K.G., Parker, T.L., 2000. Protein adsorption and platelet attachment and activation, on TiN, TiC, and DLC coatings on titanium for cardiovascular applications. J. Biomed. Mater. Res. A 52, 413–421. http://dx.doi. org/10.1002/1097-4636(200011)52:2%3c413::AID-JBM23%3e3.0. CO;2-U. Kyomoto, M., Moro, T., Konno, T., Takadama, H., Yamawaki, N., Kawaguchi, H., Takatori, Y., Nakamura, K., Ishihara, K., 2007a. Enhanced wear resistance of modified cross-linked polyethylene by grafting with poly(2-methacryloyloxyethyl phosphorylcholine). J. Biomed. Mater. Res. A 82, 10–17. http:// dx.doi.org/10.1002/jbm.a.31134. Kyomoto, M., Moro, T., Konno, T., Takadama, H., Kawaguchi, H., Takatori, Y., Nakamura, K., Yamawaki, N., Ishihara, K., 2007b. Effects of photo-induced graft polymerization of 2methacryloyloxyethyl phosphorylcholine on physical properties of cross-linked polyethylene in artificial hip joints. J. Mater. Sci.: Mater. Med. 18, 1809–1815. http://dx.doi.org/10. 1007/s10856-007-3063-0. Kyomoto, M., Moro, T., Miyaji, F., Hashimoto, M., Kawaguchi, H., Takatori, Y., Nakamura, K., Ishihara, K., 2008a. Effect of 2methacryloyloxyethyl phosphorylcholine concentration on photo-induced graft polymerization of polyethylene in reducing the wear of orthopaedic bearing surface. J. Biomed. Mater. Res. A 86, 439–447. http://dx.doi.org/10.1002/jbm.a.31511. Kyomoto, M., Moro, T., Miyaji, F., Konno, T., Hashimoto, M., Kawaguchi, H., Takatori, Y., Nakamura, K., Ishihara, K., 2008b. Enhanced wear resistance of orthopaedic bearing due to the cross-linking of poly (MPC) graft chains induced by gammaray irradiation. J. Biomed.. Mater. Res. Part B: Appl. Biomater. 84, 320–327. http://dx.doi.org/10.1002/jbm.b.30874. Lewis, A.L., 2000. Phosphorylcholine-based polymers and their use in the prevention of biofouling. Coll. Surf. B: Biointerfaces 18, 261–275. http://dx.doi.org/10.1016/S0927-7765(99)00152-6. Lewis, A.L., Cumming, Z.L., Goreshi, H.H., Kirkwood, L.C., Tolhurst, L.A., Stratford, P.W., 2001. Crosslinkable coatings from phosphorylcholine-based polymer. Biomaterials 22, 99–111. http://dx.doi.org/10.1016/S0142-9612(00)00083-1. Lewis, A.L., Tolhurst, L.A., Stratford, P.W., 2002. Analysis of a phosphorylcholine-based polymer coating on a coronary stent pre- and post-implantation. Biomaterials 23, 1697–1706. http://dx.doi.org/10.1016/S0142-9612(01)00297-6. Lewis, A.L., Willis, S.L., Small, S.A., Hunt, S., O’Byrne, V., Stratford, P. W., 2004. Drug loading and elution from a phosphorylcholine polymer coated coronary stent does not affect long-term stability of the coating in vivo. Biomed. Mater. Eng. 14, 355–370. Liu, X., Chu, P.K., Ding, C., 2004. Surface modification of titanium, titanium alloys, and related materials for biomedical applications. Mater. Sci. Eng.: R: Rep. 47, 49–121. http://dx.doi. org/10.1016/j.mser.2004.11.001. Matinlinna, J.P., Laajalehto, K., Laiho, T., Kangasniemi, I., Lassila, L.V.J., Vallittu, P.K., 2004. Surface analysis of Co–Cr–Mo alloy and Ti substrates silanizated with trialkoxysilanes and silane mixtures. Surf. Interface Anal. 36, 246–253. http://dx.doi.org/ 10.1002/sia.1681.
17
Ozeki, K., Masuzawa, T., Hirakuri, K.K., 2010. The wear properties and adhesion strength of the diamond-like carbon film coated on SUS, Ti and Ni–Ti with plasma pre-treatment. Biomed. Mater. Eng. 20, 21–35. http://dx.doi.org/10.3233/ BME-2010-0612. Sin, D.C., Kei, H.L., Miao, X., 2005. Surface coatings for ventricular assist devices. Expert Rev. Med. Devices 6, 51–60. http://dx.doi. org/10.1586/17434440.6.1.51. Tang, Y., Lu, J.R., 2001. Swelling of zwitterionic polymer films characterized by spectroscopic ellipsometry. Macromolecules 34, 8768–8776. http://dx.doi.org/10.1021/ma010476i. Ueda, M., Silva, M.M., Otani, C., Reuther, H., Yatsuzuka, M., Lepienski, C.M., Berni, L.A., 2003. Improvement of tribological properties of Ti6Al4V by nitrogen plasma immersion ion implantation. Surf. Coat. Technol. 170, 408–410. http://dx.doi. org/10.1016/S0257-8972(03)00137-3. Ueda, T., Oshida, H., Kurita, K., Ishihara, K., Nakabayashi, N., 1992. Preparation of 2-methacryloyloxyethyl phosphorylcholine copolymers with alkyl methacrylates and their blood compatibility. Polym. J. 24, 1259–1269. http://dx.doi. org/10.1295/polymj.24.1259. Wang, J., Bartlett, J.D., Dunn, A.C., Small, S., Willis, S.L., Driver, M. J., Lewis, A.L., 2005. The use of rhodamine 6G and fluorescence microscopy in the evaluation of phospholipid-based polymeric biomaterials. J. Microsc. 217, 216–224. http://dx.doi. org/10.1111/j.1365-2818.2005.01453.x. Yasuda, H., Gazicki, M., 1982. Biomedical applications of plasma polymerization and plasma treatment of polymer surfaces. Biomaterials 3, 68–77. http://dx.doi.org/10.1016/0142-9612(82) 90036-9. Ye, S.H., Johnson, C.A., Woolley, J.R., Snyder, T.A., Gamble, L.J., Wagner, W.R., 2009a. Covalent surface modification of a titanium alloy with a phosphorylcholine-containing copolymer for reduced thrombogenicity in cardiovascular devices. J. Biomed. Mater. Res. A 91, 18–28. http://dx.doi.org/10. 1002/jbm.a.32184. Ye, S.H., Johnson, C.A., Woolley, J.R., Oh, H.-I., Gamble, L.J., Ishihara, K., Wagner, W.R., 2009b. Surface modification of a titanium alloy with a phospholipid polymer prepared by a plasma-induced grafting technique to improve surface thromboresistance. Coll. Surf. B: Biointerfaces 74, 96–102. http://dx.doi.org/10.1016/j.colsurfb.2009.06.032. Ye, S.H., Johnson, C.A., Woolley, J.R., Murata, H., Gamble, L.J., Ishihara, K., Wagner, W.R., 2010. Simple surface modification of a titanium alloy with silanated zwitterionic phosphorylcholine or sulfobetaine modifiers to reduce thrombogenicity. Coll. Surf. B: Biointerfaces 79, 357–364. http://dx.doi.org/10.1016/j.colsurfb.2010.04.018. Ye, S.H., Jang, Y.S., Yun, Y.H., Shankarraman, V., Woolley, J.R., Hong, Y., Gamble, L.J., Ishihara, K., Wagner, W.R., 2013. Surface modification of a biodegradable magnesium alloy with phosphorylcholine (PC) and sulfobetaine (SB) functional macromolecules for reduced thrombogenicity and acute corrosion resistance. Langmuir 29, 8320–8327. http://dx.doi. org/10.1021/la401341y. Zhang, F., Kang, E.T., Neoh, K.G., Wang, P., Tan, K.L., 2001. Surface modification of stainless steel by grafting of poly (ethylene glycol) for reduction in protein adsorption. Biomaterials 22, 1541–1548. http://dx.doi.org/10.1016/ S0142-9612(00)00310-0.
