- Email: [email protected]
Structural–mechanical and antibacterial properties of a soft elastic polyurethane surface after plasma immersion N2+ implantation
Materials Science and Engineering C 62 (2016) 242–248
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Structural–mechanical and antibacterial properties of a soft elastic polyurethane surface after plasma immersion N+ 2 implantation Ilya A. Morozov a,b,⁎, Alexander S. Mamaev c, Irina V. Osorgina b, Larisa M. Lemkina d, Vladimir P. Korobov d,e, Anton Yu Belyaev a, Svetlana E. Porozova e, Marina G. Sherban b a
Institute of Continuous Media Mechanics UB RAS, Academika Koroleva st. 1, 614013 Perm, Russia Perm State University, Bukireva st. 15, 614990 Perm, Russia Institute of Electrophysics UD RAS, Amundsen st. 106, 620016 Ekaterinburg, Russia d Institute of Ecology and Genetics of Microorganisms UB RAS, Golev st. 13, 614081 Perm, Russia e Perm National Research Polytechnic University, Komsomolsky av. 29, 614990 Perm, Russia b c
a r t i c l e
i n f o
Article history: Received 26 September 2015 Received in revised form 8 January 2016 Accepted 24 January 2016 Available online 26 January 2016 Keywords: Polyurethane Plasma immersion ion implantation Structural mechanical properties Bacterial adhesion Atomic force microscopy
a b s t r a c t The surface of elastic polyurethane treated by plasma immersion N+ 2 ion implantation at different fluences has been investigated. A folded surface structure is observed in all cases. Analysis has been performed to study the structural (roughness, steepness and fraction of folds, fractal characteristics), mechanical (stiffness, adhesion force between the AFM probe and the material) and wetting properties of surfaces. Under uniaxial stretching the cracks orthogonal to the axis of deformation and longitudinal folds are formed on the examined surfaces. After unloading the initial structure of the surface of deformed materials exposed to low fluences becomes smoother and does not recover, i.e. it has plastic properties. By contrast, the structure of the surfaces of materials subjected to high-fluence treatment recovers without visible changes and the cracks are fully closed. The study of Staphylococcus colonies grown on these materials has demonstrated significant reduction (from 3 to 5 times) in the vitality of bacteria on treated surfaces. This result was repeated on samples after 11 months of storage. Such antibacterial properties are primarily related to the structural changes of the surfaces accompanied by the increased hydrophilicity. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Metals, ceramics and polymers are widely used in modern medicine materials. One of the commonly used polymers for the manufacture of biomedical devices is polyurethane (PU). Depending on the composition and preparation techniques the mechanical properties of the PU can vary over a wide range from soft elastomers to stiff plastics. PU is suitable for creating catheters, cardioimplants, breast implants, interphalangeal prosthesis, etc. Currently, there are plenty of studies on surface modification of materials. These studies are intended to improve the quality of biomedical products (anti-bacterial properties, biocompatibility with living tissues). Surface modification can be done by different methods: mechanical and chemical techniques, plasma treatment, plasma immersion ion implantation, carbon sputtering [1–6], etc. The plasma immersion ion implantation (PIII) has some advantages. Modified layer is formed in the material itself, and there is no distinct
⁎ Corresponding author at: Institute of Continuous Media Mechanics UB RAS, Academika Koroleva st. 1, 614013 Perm, Russia. E-mail address: [email protected] (I.A. Morozov).
http://dx.doi.org/10.1016/j.msec.2016.01.062 0928-4931/© 2016 Elsevier B.V. All rights reserved.
boundary between the modified surface layer and the bulk material. Plasma treatment modifies the chemical structure and hydrophilicity of the surface, and this significantly influences protein adsorption [7]. Wang et al. [8] revealed that the PIII treatment of PU in acetylene medium affected platelet adhesion. Bax et al. [9] studied the influence of the PIII treatment of the PU surface on the adhesion behavior of tropoelastin — protein responsible for the adhesion and growth of tissue cells. It has been found that application of silver [10] or copper [11] ions improves the antimicrobial properties of the material. The influence of PIII on the structural–mechanical properties of the surface of PU is still poorly understood. Besides, in most works a relatively stiff PU (e.g., [9] — 35 MPa) is used. The PIII of soft elastic PU is a promising technique, which can also be used to improve the quality of movable, flexible medical devices. The mechanical compatibility between the material and the coating is important for the longevity of materials. It must satisfy operational conditions and loads. That is, the modified polymer layer must be strong enough to avoid delamination in the loading mode. The present work focuses on the structural and mechanical characteristics of the surface of elastic polyurethane subjected to the PIII treatment at different ion beam fluences. The effect of the treated surfaces on the adhesion of colonies of Staphylococcus epidermidis was examined.
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
The atomic force microscopy (AFM) in nanomechanical mapping regime was used to study the structure of the surface and its local mechanical properties. The formation of fractal folds was observed on the treated surfaces. With increasing treatment time the surface structure becomes more significantly ramified, its stiffness increases and an adhesion force between the AFM probe and the material is reduced. The properties of surfaces were studied in the deformed (stretched) and undeformed (unstretched) states and after dynamic loading. Depending on the fluence and consequently on the structural–mechanical properties of the surface, a 5-fold reduction in the viability of bacterial films formed on the surface of the materials was achieved. 2. Materials and methods 2.1. Preparation of polyurethane composition A polyurethane composition used in this study is a two-component system of a prepolymer (simple oligoether, mol. wt. 2300, based on polypropylene oxide and 2.4-toluene diisocyanate) and a hardener (aromatic diamine dissolved in the polyol) in a weight ratio of 100:46.2. The estimated amount of prepolymer was heated to 50 °C and vacuumed for 7 min. Then it was mixed with a hardener mixture heated to 60 °C. The mixture was stirred and then vacuumed again at 50 °C for 7 min. The vacuumed mixture was poured into a mold and solidified in a heat chamber at 100 °C for 20 h. Then the material was cooled at room temperature. The average thickness of polyurethane plates was 1.5 mm. The mechanical characteristics of the plates are stabilized in 1–2 weeks. 2.2. Surface treatment The samples were subjected to a plasma immersion ion implantation of nitrogen ions N+ 2 . A source of electrons with plasma cathode based on a glow discharge was used to generate plasma in the vacuum chamber. The chamber was filled with nitrogen at a rate of 50 ml/min, and the pressure of a working gas was set to 0.35 Pa. A beam of lowenergy electrons (up to 50 eV) with a diameter of 80 mm and a current of 0.5–0.6 A was formed in the double layer of space charge near the grid of plasma cathode. An electrically isolated holder of samples was located inside the vacuum chamber at a distance of 100 mm from the grid of electron emitting source. The temperature of the sample is measured using a thermocouple located on the holder. The sample was placed in the holder and closed by metal mesh with a cell size 0.6 × 0.6 mm. Pulsed DC (50 kHz) negative bias voltage with amplitude of 1 kV was applied to the holder. The plasma ions are generated by the electron beam and accelerated in the double layer of space charge region, which was created near the mesh. The average ion current density was 15 μA/cm2. The fluence F determined by the time of treatment t was: a) 2 · 1015 ions/cm2 (t = 20 s); b) 2 · 1016 ions/cm2 (t = 204 s); and c) 2 · 1017 ions/cm2 (t = 2040 s). The temperature of the first two samples did not exceed 80 °C, the last — 130 °C. These temperatures measured at the sample holder were less than the temperature of thermal degradation of PU — 170 °C. Analysis of the estimated distribution of the penetration depth of nitrogen ions (1 kV) into PU (density — 1.24 g/cm3, the elemental composition: \\NHCOO\\) performed in the program TRIM [12] by a standard SRIM Monte-Carlo simulation is shown the value of penetration depth around 10 nm. 2.3. Growing and analysis of biofilms Sterilized samples of 10 × 10 mm were placed in a glass vial with 2 ml of a suspension of bacteria S. epidermidis (33 GISK, Moscow) containing 108 CFU/ml (number of bacteria in 1 ml) for 48 h at 37 °C. At the end of the incubation period the plates were washed twice in
243
10 mM phosphate buffer (pH 7.2). The experiments were repeated three times. As a result, the areas formed by the one-cell thick bacteria layer (established by AFM methods) were found on the surface. The total number of bacteria on the surface and the amount of vital cells were investigated. In the first case, the analysis was carried out by optical microscopy (video microscope Hirox KH-7700, Japan). At least 10 pictures of size 332 × 443 μm (magnification × 210 times) were captured from each sample, and the fraction of dark areas occupied by bacteria to the total area of the image was calculated. The number of vital cells was determined by the MTS tetrazolium recovery level in the Cell Proliferation Assay system (“Promega”, USA) in accordance with the manufacturer's prescription. The optical density of MTS solutions was determined at 490 nm by a spectrophotometer PD-330 (Japan). Preparation and analysis of biofilms were performed after one week after surface treatment and then repeated (with the same old samples) after 11 months of storage samples in dark place at room temperature. 2.4. Spectroscopy The study of the Raman spectra was carried out on the multifunction spectrometer “SENTERRA” (Bruker) at a 532 nm wavelength of emitting laser. 2.5. Atomic force microscopy The properties of surfaces were studied using an atomic force microscope Dimension Icon (Veeco, USA) in a nanomechanical mapping mode (PeakForce QNM). Nanoindentation was carried out at each point of the surface with a frequency of 2 kHz [13]. The obtained force curves of interaction of the AFM-probe with the surface were analyzed. The following structural–mechanical properties of the material were studied: 1) surface height; 2) adhesion force between the tip and the sample; 3) depth of penetration of the probe into the material; and 4) material stiffness (elastic modulus E calculated in terms of the Derjaguin–Muller–Toporov (DMT) model). Depending on the hardness of the surface, the probes ScanAsyst-Air (by Bruker, a nominal tip radius R = 6 nm, stiffness k, calibrated by free thermal oscillations — 0.6 N/m) or HA-FM (by NT-MDT, R = 10 nm, k = 5.5 N/m) were used. The depth of indentation into the material was kept at a value of ~5 nm. In this case the applied probe-sample force was 2 (untreated PU), 10 (PIII 2 · 1015 ions/cm2), 15 (PIII 2 · 1016 ions/cm2), and 20 (PIII 2 · 1017 ions/cm2) nN. The probe radii used in DMTmodel were calibrated by measuring the known stiffness of polydimethylsiloxane samples (2.5 MPa). For measuring a tip-sample adhesion force, only soft probes (k = 0.6 N/m) were used, and the force applied to the surface was small. In this case, the indentation depth was b2 nm for untreated PU, and ~0 nm for PIII treated samples. A structural–mechanical analysis was carried out to study the samples in both undeformed and stretched static states. In the latter case the sample was stretched and fixed in a small device that was placed directly under the microscope scanner. 2.6. Mechanical dynamical tests An investigation of the surface of treated samples after dynamic mechanical loading has also attracted our interest. Samples with size of 10 × 4 mm were tested for 10 min in a dynamo-mechanical analyzer DMA/SDTA861e (Mettler Toledo, Switzerland) with a frequency of 30 Hz and a 10% strain. After loading, the samples were examined by AFM.
244
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
Fig. 1. AFM-height images: (a) — untreated PU; after PIII: (b) — 2 · 1015 ions/cm2; (c) — 2 · 1016 ions/cm2; and (d) — 2 · 1017 ions/cm2.
2.7. Wetting angle measurements The surface wettability was studied using drop shape analyzer DSA25E (Kruss GmbH) by sessile drop method. Water contact angle measurement was performed 10 times for each sample and then averaged.
3. Results and discussion The analysis of the results of mechanical dynamic tests did not reveal differences between the treated materials and the untreated PU. Apparently, this can be attributed to the fact that the thickness of the modified surface layer is small. The resulting Raman spectra also do not differ from each other for the same reason. The peak of 1314 cm− 1 (intensive enough in the pure PU) is not observed in the treated PUs. In addition, because of treatment (regardless the sample) the ratio of intensity of peaks of 1322 and 1268 cm−1 changed (for untreated polyurethane from 1.6 to 1.3), i.e. objectively the 1268 cm− 1 peak intensity increased. These changes, since they are the same for all treated samples, can be associated with the changes in surface defects of PU.
The measured water contact angle for untreated sample was 75.18 ± 0.04° and after PIII: 70.64 ± 0.54° — 2 · 1015 ions/cm2, 70.62 ± 0.30° — 2 · 1016 ions/cm2, 70.58 ± 0.26° — 2 · 1017 ions/cm2. The contact angle decreased after the treatment with low fluence and then remained almost unchanged for higher fluences. Decrease of contact angle means increase of surface hydrophilicity. The surfaces of the raw and treated PU are shown in Fig. 1. Hereinafter, in the left-hand corner of the image (Fig. 1) under the horizontal bar the length scale is given. To the right of the vertical bar the min/max and units of the measured value are indicated. To save the space, all the images are cut off vertically. The surface of the untreated sample is relatively smooth, without pronounced features (Fig. 1a). The surface of the samples subjected to treatment is covered with folds (marked by contours in Fig. 1b–d), and cracks (shown by arrows in Fig. 1b). The folds were contoured as follows. The AFM images were analyzed as “ordinary” digital images. Their contrast ratio was increased, and they were converted into black–white binary images (transition threshold was determined by the Otsu algorithm [14]). With increasing the dose of ions, cracks disappear (Fig. 1c), and the entire surface is covered with folds. In addition, there occur single stiff
Fig. 2. AFM images of stiffness of untreated PU (a), height (b) and stiffness (c) of one of the clusters on the PU surface after PIII 2 · 1016 ions/cm2.
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
“inclusions” of size 20 … 60 nm and clusters of these inclusions (marked by arrows in Fig. 1d) of 300 … 600 nm. The stiffness map of the surface of untreated PU (topography shown in Fig. 1a) is given in Fig. 2a. The average modulus of the untreated surface is 4 MPa. The surface and stiffness of one of the clusters on the treated surface are shown in Fig. 3b, c. The cluster stiffness is significantly higher than that of the surrounding surface — by ~ 5 times (measurement accuracy may be affected by local roughness). The mechanical properties of the treated surfaces are discussed in detail below. Factors for the appearance of folds are as follows. The flow of ions creates pressure that deforms the surface layer of the material, and simultaneously ion irradiation effects cause changes in the properties of the deformed layer change — it becomes stiffer. Upon completion of the treatment, the PU returns its initial undeformed state, folding the surface layer. Apart from pressure, the fold formation may also be caused by temperature. During the treatment the polymer with a forming layer is heated and expanded. The process of cooling is accompanied by inhomogeneous compression of the PU surface layer.
245
Table 1 Structural parameters of the surfaces. Treatment
ϕw, %
Dp
Rrms, nm
Untreated PU 2 · 1015 ions/cm2 2 · 1016 ions/cm2 2 · 1017 ions/cm2
– 31 40 51
– 1.47 1.57 1.62
9 18 20 48
The fraction ϕw of folds (the ratio of the area of the contours of folds to the entire image area) is given in Table 1. The bigger the fluence, the higher ϕw: an increase from 31% (PIII 2 · 1015 ions/cm2) to 51% (2 · 1017 ions/cm2). It has been established that the perimeter P and the area A of the fold contours are related by the fractal equation: P ~ ADp/2. The values of P(A) in double logarithmic coordinates with approximating lines (slopes are equal to Dp/2) are shown in Fig. 3. The fractal dimensions Dp are presented in Table 1. They increase with the treatment time, i.e. the structure of folds becomes more convoluted. The coefficient of determination R2 – the accuracy criterion for linear approximation lg(P)(lg(A)) – is also shown in Fig. 3. The relationship between the surface roughness and the size of the examined area in the AFM-image is written as vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u X ns 2 u 1 ns X Rrms ðsÞ ¼ t 2 zij −bzNs ; ns i¼1 j¼1
Fig. 3. Double logarithmic plots of area–perimeter relationships for: (a) — 2 · 1015 ions/cm2; (b) — 2 · 1016 ions/cm2; and (c) — 2 · 1017 ions/cm2.
Fig. 4. (a) — surface roughness versus observation area size (for AFM-images 5 × 5 μm); (b) — local slope distribution.
246
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
Fig. 5. AFM-images of the treated surfaces at 20% elongation: (a) — 2 · 1015 ions/cm2; (b) — 2 · 1016 ions/cm2; and (c) — 2 · 1017 ions/cm2. The axis of deformation is vertical.
where s is the size of the area with points ns × ns in the AFM-image; zij are the heights, observed in the given area; b z N |s is the average height. Thus, we select the areas s × s in the AFM images and calculate Rrms(s) for each area. The estimated roughness of the area of size s is the average of local roughness: b Rrms N. To calculate roughness 10 images 5 × 5 μm (512 × 512 points) were obtained for each sample. The results are shown in Fig. 4a. Note that for larger images various surface artifacts (scratches, contaminations, etc) fall into the observation area. The roughness of such artifacts is out of interest in the scope of the present work. The surface topography can be divided into flat spatial triangular elements. The angle between the normal to the triangle and the vertical axis (the angle θ in spherical coordinates) represents the local slope of the surface — the low values of θ correspond to areas that are close to the horizontal ones. Distribution of the angles θ for the images under consideration is shown in Fig. 4b. Fig. 4a shows that in all cases the roughness increases asymptotically to a certain value. The roughness of the surfaces with PIII 2 · 1015 ions/cm2 and 2 · 10 16 ions/cm 2 differs only slightly. The sample with 2 · 1017 ions/cm2 has the highest value of roughness. The values of surface roughness (for s = 5 μm) are given in Table 1. With the increase of fluence, the distribution of slope angles (Fig. 4b) expands and shifts to higher values, i.e. the edges of folds become steeper.
Under uniaxial deformation, the structure of the treated surfaces changes. Fig. 5 shows the images of samples stretched by 20%; the axis of stretching is vertical. Elongation causes the appearance of cracks on the treated surfaces. The initial surface structure of materials in the low-fluence regime (see. Fig. 1b, c) is smoothed during stretching, and new folds that are parallel to the axis of deformation (Fig. 5a, b) appear — the result of sample compression in the direction orthogonal to the load. Similar, but less pronounced, longitudinal folds are visible on the surface of the sample in the high-fluence regime (Fig. 5c). However, in this case the initial surface texture also remains. The mechanical properties of the material in open cracks are similar to those of the untreated PU. Measurements of the crack depth in the vicinity of its tip located in the samples elongated by 20% (Fig. 5) made it possible to estimate the thickness of the surface layer. This value is 6 … 9 nm, which is consistent with the results given by the program TRIM. More precise measurements of the layer thickness and establishment of a correlation with fluence are impossible due to the presence of folds. The increase in elongation causes further crack growth and increase in longitudinal fold concentration. The surface of the material elongated twice and then unloaded is shown in Fig. 6. The surface structure of unloaded materials with a low fluence (Fig. 6a, b) is significantly different from the initial one (Fig. 1b, c). The
Fig. 6. AFM-images of the surfaces after removal of 100% elongation: (a) — 2 · 1015 ions/cm2; (b) — 2 · 1016 ions/cm 2; and (c) — 2 · 1017 ions/cm2.
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
247
Table 2 Adhesion properties of bacteria.
Fig. 7. Distributions of tip-sample adhesion force (a) and surface stiffness (b).
stretch-induced vertical folds, as well as the traces of closed horizontal cracks, are visible. This is indicative of the plastic properties of these surfaces. At the same time, the surface of the material subjected to 2 · 1017 ions/cm2 recovered completely, and no signs of loading were observed. Similar behavior was observed under dynamical mechanical loading — the surface of materials in the low-fluence regime became smoother, and a great number of cracks occurred, whereas the surface of the PU treated with 2 · 1017 ions/cm2 remained without visible changes. It is important to note that, despite the fact that the surface changes after application of load, the coating continuity is preserved. No peeled surface layers and open cracks were revealed in the unloaded materials. Distributions of the tip-surface adhesion force and surface hardness are shown in Fig. 7. According to the results, the adhesion force (Fig. 7a) decreases with increasing fluence. The accuracy of AFM adhesion measurements is affected by surface roughness [15]. For the 2 · 1017 ions/cm2 treated material the distribution of adhesion shows a weakly pronounced maximum at ~1 nN. Comparison of the surface and the corresponding adhesion map revealed that such small adhesion values were obtained for areas with a sharp slope (fold edges). Thus, a maximum in the vicinity of 1 nN is the measurement error associated with a change in
Treatment
βall, %
βliv
γliv
Untreated PU 2 · 1015 ions/cm2 2 · 1016 ions/cm2 2 · 1017 ions/cm2
62 ± 3.7 27 ± 3 15 ± 1.5 13 ± 2.5
1.45 ± 0.15 0.86 ± 0.07 0.81 ± 0.02 0.30 ± 0.03
1.38 ± 0.18 0.82 ± 0.02 0.87 ± 0.02 0.29 ± 0.004
the contact area of the tip-surface. The dotted line on the adhesion curve corresponds to actual values. The stiffness of the examined surfaces increases with increasing fluence (Fig. 7b). The estimated values of Young's modulus are: 4 (untreated material, Fig. 3a), 39 (2 · 1015 ions/cm2), 52 (2 · 1016 ions/cm2), and 100 (2 · 1017 ions/cm2) MPa. Due to the small thickness of the modified surface layer, the measured stiffness is the characteristic of the polymer–surface layer system. The definition of stiffness of the modified layer itself (for example, using finite element modeling of the indentation of a hard film on a soft substrate) is complicated by the small thickness and the inhomogeneous (non-elastic) properties of the modified polymer depending the distance from the surface. Fig. 8 shows examples of optical and AFM images of Staphylococcus cells on the surface of the material treated with PIII 2 · 1017 ions/cm2. The results presented here have shown that the adhesion of bacteria to surfaces depends on the fluence. The fraction of the area occupied by cells βall and the number of vital cells βliv (measured after one week of PIII treatment) and γliv (measured and after 11 months of storage of PIII treated samples in dark place at room temperature) are given in Table 2. The values are reduced drastically during the transition from the raw PU to the material subjected to minimal treatment. The number of vital cells has changed only insufficiently for materials treated with a dose 2 · 1015–2 · 1016 ions/cm2. The total number of cells is reduced, i.e. the amount of dead bacteria on the surface is reduced. After PIII 2 · 1017 ions/cm2 treatment βliv has decreased more than twice in comparison with the previous value and by almost 5 times in comparison with the untreated PU. The results of second measurements of vital cells γliv on the same samples after 11 months of storage did not shown significant difference between measurements on fresh samples. In the work by [16], it is shown that the roughness of a silicon substrate has an effect on the adhesion and growth of colonies of Staphylococcus cells only when Rrms N 200 nm, but the surface energy has no significant effect on cell growth. On the other hand, PereraCosta et al. [17] have revealed that the surface texture affects the growth of bacterial films even with low roughness. The results (Table 2) are correlated with the surface structure (roughness, fractal properties, and fraction of folds), hydrophilicity and the adhesion force between the AFM probe and the material (Fig. 7a). In the work [18] was shown that the more hydrophilic surface the higher resistance to the bacterial adhesion. In comparison with the untreated PU, the hydrophilicity of the material treated with 2 · 1015 ions/cm2 increased, and therefore the main factors initially decreasing the number of vital cells are reduction of wetting and the formation of the folded structure on the surface. Further reduction of cells (for the samples with higher fluences) can be attributed primarily
Fig. 8. Optical (a) and AFM (b) images of bacteria on the surface treated with 2 · 1017 ions/cm2.
248
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
to the formation and growing of the folded structure, because the hydrophilicity remains unchanged. Another explanation is the formation of hard bumps (see. Fig. 2b, c) on the material surface. These bumps also hinder the consolidation of bacteria on the surface. Moreover, the surface energy of PIII treated polymers asymptotically decreases over time [19]. However, the number of vital cells on the old samples does not changed in comparison with the fresh ones. This can be another evidence that the surface structure is the main cause of the reduction of bacterial adhesion. 4. Conclusions The structural–mechanical properties of the surfaces of soft elastic polyurethane after plasma immersion ion implantation were studied. According to the AFM measurements, the thickness of the modified layer was 6–9 nm. The fractal folded structure is formed on the surface after the treatment. The surface roughness increases with increasing PIII treatment fluence, the folds become more sinuous (fractal dimension increases), and their fraction increases. The most rough and sinuous surface occurs after the 2 · 1017 ions/cm2 fluence treatment. Measurements of local mechanical properties show that the adhesion force between the probe and the surface decreases, and the surface stiffness increases with increasing treatment fluence. Under uniaxial deformation, the folds codirectional with the axis of stretching (due to compression in the transverse direction) and cracks are formed. The initial structure of the surface of materials treated with low fluence (2 · 1015–2 · 1016 ions/cm2) smoothes under stretching; after removal of the load the structure does not restore, the longitudinal folds and the traces of cracks remain, i.e. the surface exhibits plastic properties. At the same time, the surface of the material treated with 2 · 1017 ions/cm2 restores its initial structure after unloading, and the cracks closed without a trace. Similar pattern was observed in the study of materials subjected to dynamic loading. It should be noted that after removing the mechanical load in all the cases the coating continuity (delamination of the modified layer did not occur) was maintained. Investigation of Staphylococcus colonies grown on the surface of materials of interest shows a five-fold reduction in the number of vital cells in comparison with the untreated PU. Such differences are primarily due to the structural properties of the surfaces (the structure of the folds) supported by the increased hydrophilicity of the treated surface. These antibacterial properties remained unchanged at least after 11 months of surface treatment.
Acknowledgments This work was supported by the Russian Foundation for Basic Research (grants 14-04-00687, 13-01-96009_r_ural_a, 14-0896003_r_ural_a) and the Ministry of Education of Perm Region under agreement (S-26/632). References [1] T. Lu, Y. Qiao, X. Liu, Surface modification of biomaterials using plasma immersion ion implantation and deposition. A review, Interface Focus 2 (2012) 325–336. [2] A. Kurella, N.B. Dahotre, Surface modification for bioimplants: the role of laser surface engineering, J. Biomater. Appl. 20 (2005) 5–50. [3] B. Bhushan, B.K. Gupta, Handbook of Tribology: Materials, Coatings and Surface Treatments, McGraw-Hill, Inc., New York, 1991. [4] C.M. Cotell, Pulsed laser deposition and processing of biocompatible hydroxyapatite thin films, Appl. Surf. Sci. 69 (1993) 140–148. [5] A.C. Duncan, F. Weisbuch, F. Rouais, S. Lazare, C. Baquey, Laser microfabricated model surfaces for controlled cell growth, Biosens. Bioelectron. 17 (2002) 413–426. [6] E.N. Antonov, V.N. Bagratashvili, V.K. Popov, M.D. Ball, D.M. Grant, S.M. Howdle, C.A. Scotchford, Properties of calcium phosphate coatings deposited and modified with lasers, J. Mater. Sci. Mater. Med. 14 (2003) 151–155. [7] P. Alves, S. Pinto, H.C. de Sousa, M.H. Gil, Surface modification of a thermoplastic polyurethane by low-pressure plasma treatment to improve hydrophilicity, J. Appl. Polym. Sci. 122 (2011) 2302–2308. [8] J. Wang, P. Yang, H. Sun, J.Y. Chen, Y.X. Leng, G.J. Wan, N. Huang, Surface modification of medical polyurethane by acetylene plasma immersion ion implantation, 30th International Conference on Plasma Science 2003, p. 282 (Abstracts). [9] D.V. Bax, A. Kondyurin, A. Waterhouse, D.R. McKenzie, A.S. Weiss, M.M.M. Bilek, Surface plasma modification and tropoelastin coating of a polyurethane copolymer for enhanced cell attachment and reduced thrombogenicity, Biomaterials 35 (2014) 6795–6809. [10] H. Cao, X. Liu, F. Meng, P.K. Chu, Biological actions of silver nanoparticles embedded in titanium controlled by micro-galvanic effects, Biomaterials 32 (2011) 693–705. [11] W. Zhang, Y. Zhang, J. Ji, Q. Yan, A. Huang, P.K. Chu, Antimicrobial polyethylene with controlled copper release, J. Biomed. Mater. Res. A 83A (2007) 838–844. [12] http://www.srim.org/. [13] B. Pittenger, N. Erina, C. Su, Quantitative Mechanical Property Mapping at the Nanoscale with PeakForce QNM, Application Note, 1–12, Veeco Instruments Inc., 2010 [14] N. Otsu, A threshold selection method from gray-level histograms, IEEE Trans. Syst. Man Cybern. 9 (1979) 62–66. [15] D.-L. Liu, J. Martin, N.A. Burnham, Optimal roughness for minimal adhesion, Appl. Phys. Lett. 91 (2007) 043107. [16] H. Tang, T. Cao, X. Liang, A. Wang, S.O. Salley, J. McAllister, K.Y. Ng, Influence of silicone surface roughness and hydrophobicity on adhesion and colonization of Staphylococcus epidermidis, J. Biomed. Mater. Res. A 88A (2009) 454–463. [17] D. Perera-Costa, J.M. Bruque, M.L. González-Martín, A.C. Gómez-García, V. VadilloRodríguez, Studying the influence of surface topography on bacterial adhesion using spatially organized microtopographic surface patterns, Langmuir 30 (2014) 4633–4641. [18] Y.H. An, R.J. Friedman, Concise review of mechanisms of bacterial adhesion to biomaterial surfaces, J. Biomed. Mater. Res. 43 (1998) 338–348. [19] A. Kondyurin, M. Bilek, Ion Beam Treatment of Polymers: Application Aspects from Medicine to Space, second ed. Elsevier, 2014.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Structural–mechanical and antibacterial properties of a soft elastic polyurethane surface after plasma immersion N+ 2 implantation Ilya A. Morozov a,b,⁎, Alexander S. Mamaev c, Irina V. Osorgina b, Larisa M. Lemkina d, Vladimir P. Korobov d,e, Anton Yu Belyaev a, Svetlana E. Porozova e, Marina G. Sherban b a
Institute of Continuous Media Mechanics UB RAS, Academika Koroleva st. 1, 614013 Perm, Russia Perm State University, Bukireva st. 15, 614990 Perm, Russia Institute of Electrophysics UD RAS, Amundsen st. 106, 620016 Ekaterinburg, Russia d Institute of Ecology and Genetics of Microorganisms UB RAS, Golev st. 13, 614081 Perm, Russia e Perm National Research Polytechnic University, Komsomolsky av. 29, 614990 Perm, Russia b c
a r t i c l e
i n f o
Article history: Received 26 September 2015 Received in revised form 8 January 2016 Accepted 24 January 2016 Available online 26 January 2016 Keywords: Polyurethane Plasma immersion ion implantation Structural mechanical properties Bacterial adhesion Atomic force microscopy
a b s t r a c t The surface of elastic polyurethane treated by plasma immersion N+ 2 ion implantation at different fluences has been investigated. A folded surface structure is observed in all cases. Analysis has been performed to study the structural (roughness, steepness and fraction of folds, fractal characteristics), mechanical (stiffness, adhesion force between the AFM probe and the material) and wetting properties of surfaces. Under uniaxial stretching the cracks orthogonal to the axis of deformation and longitudinal folds are formed on the examined surfaces. After unloading the initial structure of the surface of deformed materials exposed to low fluences becomes smoother and does not recover, i.e. it has plastic properties. By contrast, the structure of the surfaces of materials subjected to high-fluence treatment recovers without visible changes and the cracks are fully closed. The study of Staphylococcus colonies grown on these materials has demonstrated significant reduction (from 3 to 5 times) in the vitality of bacteria on treated surfaces. This result was repeated on samples after 11 months of storage. Such antibacterial properties are primarily related to the structural changes of the surfaces accompanied by the increased hydrophilicity. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Metals, ceramics and polymers are widely used in modern medicine materials. One of the commonly used polymers for the manufacture of biomedical devices is polyurethane (PU). Depending on the composition and preparation techniques the mechanical properties of the PU can vary over a wide range from soft elastomers to stiff plastics. PU is suitable for creating catheters, cardioimplants, breast implants, interphalangeal prosthesis, etc. Currently, there are plenty of studies on surface modification of materials. These studies are intended to improve the quality of biomedical products (anti-bacterial properties, biocompatibility with living tissues). Surface modification can be done by different methods: mechanical and chemical techniques, plasma treatment, plasma immersion ion implantation, carbon sputtering [1–6], etc. The plasma immersion ion implantation (PIII) has some advantages. Modified layer is formed in the material itself, and there is no distinct
⁎ Corresponding author at: Institute of Continuous Media Mechanics UB RAS, Academika Koroleva st. 1, 614013 Perm, Russia. E-mail address: [email protected] (I.A. Morozov).
http://dx.doi.org/10.1016/j.msec.2016.01.062 0928-4931/© 2016 Elsevier B.V. All rights reserved.
boundary between the modified surface layer and the bulk material. Plasma treatment modifies the chemical structure and hydrophilicity of the surface, and this significantly influences protein adsorption [7]. Wang et al. [8] revealed that the PIII treatment of PU in acetylene medium affected platelet adhesion. Bax et al. [9] studied the influence of the PIII treatment of the PU surface on the adhesion behavior of tropoelastin — protein responsible for the adhesion and growth of tissue cells. It has been found that application of silver [10] or copper [11] ions improves the antimicrobial properties of the material. The influence of PIII on the structural–mechanical properties of the surface of PU is still poorly understood. Besides, in most works a relatively stiff PU (e.g., [9] — 35 MPa) is used. The PIII of soft elastic PU is a promising technique, which can also be used to improve the quality of movable, flexible medical devices. The mechanical compatibility between the material and the coating is important for the longevity of materials. It must satisfy operational conditions and loads. That is, the modified polymer layer must be strong enough to avoid delamination in the loading mode. The present work focuses on the structural and mechanical characteristics of the surface of elastic polyurethane subjected to the PIII treatment at different ion beam fluences. The effect of the treated surfaces on the adhesion of colonies of Staphylococcus epidermidis was examined.
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
The atomic force microscopy (AFM) in nanomechanical mapping regime was used to study the structure of the surface and its local mechanical properties. The formation of fractal folds was observed on the treated surfaces. With increasing treatment time the surface structure becomes more significantly ramified, its stiffness increases and an adhesion force between the AFM probe and the material is reduced. The properties of surfaces were studied in the deformed (stretched) and undeformed (unstretched) states and after dynamic loading. Depending on the fluence and consequently on the structural–mechanical properties of the surface, a 5-fold reduction in the viability of bacterial films formed on the surface of the materials was achieved. 2. Materials and methods 2.1. Preparation of polyurethane composition A polyurethane composition used in this study is a two-component system of a prepolymer (simple oligoether, mol. wt. 2300, based on polypropylene oxide and 2.4-toluene diisocyanate) and a hardener (aromatic diamine dissolved in the polyol) in a weight ratio of 100:46.2. The estimated amount of prepolymer was heated to 50 °C and vacuumed for 7 min. Then it was mixed with a hardener mixture heated to 60 °C. The mixture was stirred and then vacuumed again at 50 °C for 7 min. The vacuumed mixture was poured into a mold and solidified in a heat chamber at 100 °C for 20 h. Then the material was cooled at room temperature. The average thickness of polyurethane plates was 1.5 mm. The mechanical characteristics of the plates are stabilized in 1–2 weeks. 2.2. Surface treatment The samples were subjected to a plasma immersion ion implantation of nitrogen ions N+ 2 . A source of electrons with plasma cathode based on a glow discharge was used to generate plasma in the vacuum chamber. The chamber was filled with nitrogen at a rate of 50 ml/min, and the pressure of a working gas was set to 0.35 Pa. A beam of lowenergy electrons (up to 50 eV) with a diameter of 80 mm and a current of 0.5–0.6 A was formed in the double layer of space charge near the grid of plasma cathode. An electrically isolated holder of samples was located inside the vacuum chamber at a distance of 100 mm from the grid of electron emitting source. The temperature of the sample is measured using a thermocouple located on the holder. The sample was placed in the holder and closed by metal mesh with a cell size 0.6 × 0.6 mm. Pulsed DC (50 kHz) negative bias voltage with amplitude of 1 kV was applied to the holder. The plasma ions are generated by the electron beam and accelerated in the double layer of space charge region, which was created near the mesh. The average ion current density was 15 μA/cm2. The fluence F determined by the time of treatment t was: a) 2 · 1015 ions/cm2 (t = 20 s); b) 2 · 1016 ions/cm2 (t = 204 s); and c) 2 · 1017 ions/cm2 (t = 2040 s). The temperature of the first two samples did not exceed 80 °C, the last — 130 °C. These temperatures measured at the sample holder were less than the temperature of thermal degradation of PU — 170 °C. Analysis of the estimated distribution of the penetration depth of nitrogen ions (1 kV) into PU (density — 1.24 g/cm3, the elemental composition: \\NHCOO\\) performed in the program TRIM [12] by a standard SRIM Monte-Carlo simulation is shown the value of penetration depth around 10 nm. 2.3. Growing and analysis of biofilms Sterilized samples of 10 × 10 mm were placed in a glass vial with 2 ml of a suspension of bacteria S. epidermidis (33 GISK, Moscow) containing 108 CFU/ml (number of bacteria in 1 ml) for 48 h at 37 °C. At the end of the incubation period the plates were washed twice in
243
10 mM phosphate buffer (pH 7.2). The experiments were repeated three times. As a result, the areas formed by the one-cell thick bacteria layer (established by AFM methods) were found on the surface. The total number of bacteria on the surface and the amount of vital cells were investigated. In the first case, the analysis was carried out by optical microscopy (video microscope Hirox KH-7700, Japan). At least 10 pictures of size 332 × 443 μm (magnification × 210 times) were captured from each sample, and the fraction of dark areas occupied by bacteria to the total area of the image was calculated. The number of vital cells was determined by the MTS tetrazolium recovery level in the Cell Proliferation Assay system (“Promega”, USA) in accordance with the manufacturer's prescription. The optical density of MTS solutions was determined at 490 nm by a spectrophotometer PD-330 (Japan). Preparation and analysis of biofilms were performed after one week after surface treatment and then repeated (with the same old samples) after 11 months of storage samples in dark place at room temperature. 2.4. Spectroscopy The study of the Raman spectra was carried out on the multifunction spectrometer “SENTERRA” (Bruker) at a 532 nm wavelength of emitting laser. 2.5. Atomic force microscopy The properties of surfaces were studied using an atomic force microscope Dimension Icon (Veeco, USA) in a nanomechanical mapping mode (PeakForce QNM). Nanoindentation was carried out at each point of the surface with a frequency of 2 kHz [13]. The obtained force curves of interaction of the AFM-probe with the surface were analyzed. The following structural–mechanical properties of the material were studied: 1) surface height; 2) adhesion force between the tip and the sample; 3) depth of penetration of the probe into the material; and 4) material stiffness (elastic modulus E calculated in terms of the Derjaguin–Muller–Toporov (DMT) model). Depending on the hardness of the surface, the probes ScanAsyst-Air (by Bruker, a nominal tip radius R = 6 nm, stiffness k, calibrated by free thermal oscillations — 0.6 N/m) or HA-FM (by NT-MDT, R = 10 nm, k = 5.5 N/m) were used. The depth of indentation into the material was kept at a value of ~5 nm. In this case the applied probe-sample force was 2 (untreated PU), 10 (PIII 2 · 1015 ions/cm2), 15 (PIII 2 · 1016 ions/cm2), and 20 (PIII 2 · 1017 ions/cm2) nN. The probe radii used in DMTmodel were calibrated by measuring the known stiffness of polydimethylsiloxane samples (2.5 MPa). For measuring a tip-sample adhesion force, only soft probes (k = 0.6 N/m) were used, and the force applied to the surface was small. In this case, the indentation depth was b2 nm for untreated PU, and ~0 nm for PIII treated samples. A structural–mechanical analysis was carried out to study the samples in both undeformed and stretched static states. In the latter case the sample was stretched and fixed in a small device that was placed directly under the microscope scanner. 2.6. Mechanical dynamical tests An investigation of the surface of treated samples after dynamic mechanical loading has also attracted our interest. Samples with size of 10 × 4 mm were tested for 10 min in a dynamo-mechanical analyzer DMA/SDTA861e (Mettler Toledo, Switzerland) with a frequency of 30 Hz and a 10% strain. After loading, the samples were examined by AFM.
244
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
Fig. 1. AFM-height images: (a) — untreated PU; after PIII: (b) — 2 · 1015 ions/cm2; (c) — 2 · 1016 ions/cm2; and (d) — 2 · 1017 ions/cm2.
2.7. Wetting angle measurements The surface wettability was studied using drop shape analyzer DSA25E (Kruss GmbH) by sessile drop method. Water contact angle measurement was performed 10 times for each sample and then averaged.
3. Results and discussion The analysis of the results of mechanical dynamic tests did not reveal differences between the treated materials and the untreated PU. Apparently, this can be attributed to the fact that the thickness of the modified surface layer is small. The resulting Raman spectra also do not differ from each other for the same reason. The peak of 1314 cm− 1 (intensive enough in the pure PU) is not observed in the treated PUs. In addition, because of treatment (regardless the sample) the ratio of intensity of peaks of 1322 and 1268 cm−1 changed (for untreated polyurethane from 1.6 to 1.3), i.e. objectively the 1268 cm− 1 peak intensity increased. These changes, since they are the same for all treated samples, can be associated with the changes in surface defects of PU.
The measured water contact angle for untreated sample was 75.18 ± 0.04° and after PIII: 70.64 ± 0.54° — 2 · 1015 ions/cm2, 70.62 ± 0.30° — 2 · 1016 ions/cm2, 70.58 ± 0.26° — 2 · 1017 ions/cm2. The contact angle decreased after the treatment with low fluence and then remained almost unchanged for higher fluences. Decrease of contact angle means increase of surface hydrophilicity. The surfaces of the raw and treated PU are shown in Fig. 1. Hereinafter, in the left-hand corner of the image (Fig. 1) under the horizontal bar the length scale is given. To the right of the vertical bar the min/max and units of the measured value are indicated. To save the space, all the images are cut off vertically. The surface of the untreated sample is relatively smooth, without pronounced features (Fig. 1a). The surface of the samples subjected to treatment is covered with folds (marked by contours in Fig. 1b–d), and cracks (shown by arrows in Fig. 1b). The folds were contoured as follows. The AFM images were analyzed as “ordinary” digital images. Their contrast ratio was increased, and they were converted into black–white binary images (transition threshold was determined by the Otsu algorithm [14]). With increasing the dose of ions, cracks disappear (Fig. 1c), and the entire surface is covered with folds. In addition, there occur single stiff
Fig. 2. AFM images of stiffness of untreated PU (a), height (b) and stiffness (c) of one of the clusters on the PU surface after PIII 2 · 1016 ions/cm2.
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
“inclusions” of size 20 … 60 nm and clusters of these inclusions (marked by arrows in Fig. 1d) of 300 … 600 nm. The stiffness map of the surface of untreated PU (topography shown in Fig. 1a) is given in Fig. 2a. The average modulus of the untreated surface is 4 MPa. The surface and stiffness of one of the clusters on the treated surface are shown in Fig. 3b, c. The cluster stiffness is significantly higher than that of the surrounding surface — by ~ 5 times (measurement accuracy may be affected by local roughness). The mechanical properties of the treated surfaces are discussed in detail below. Factors for the appearance of folds are as follows. The flow of ions creates pressure that deforms the surface layer of the material, and simultaneously ion irradiation effects cause changes in the properties of the deformed layer change — it becomes stiffer. Upon completion of the treatment, the PU returns its initial undeformed state, folding the surface layer. Apart from pressure, the fold formation may also be caused by temperature. During the treatment the polymer with a forming layer is heated and expanded. The process of cooling is accompanied by inhomogeneous compression of the PU surface layer.
245
Table 1 Structural parameters of the surfaces. Treatment
ϕw, %
Dp
Rrms, nm
Untreated PU 2 · 1015 ions/cm2 2 · 1016 ions/cm2 2 · 1017 ions/cm2
– 31 40 51
– 1.47 1.57 1.62
9 18 20 48
The fraction ϕw of folds (the ratio of the area of the contours of folds to the entire image area) is given in Table 1. The bigger the fluence, the higher ϕw: an increase from 31% (PIII 2 · 1015 ions/cm2) to 51% (2 · 1017 ions/cm2). It has been established that the perimeter P and the area A of the fold contours are related by the fractal equation: P ~ ADp/2. The values of P(A) in double logarithmic coordinates with approximating lines (slopes are equal to Dp/2) are shown in Fig. 3. The fractal dimensions Dp are presented in Table 1. They increase with the treatment time, i.e. the structure of folds becomes more convoluted. The coefficient of determination R2 – the accuracy criterion for linear approximation lg(P)(lg(A)) – is also shown in Fig. 3. The relationship between the surface roughness and the size of the examined area in the AFM-image is written as vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u X ns 2 u 1 ns X Rrms ðsÞ ¼ t 2 zij −bzNs ; ns i¼1 j¼1
Fig. 3. Double logarithmic plots of area–perimeter relationships for: (a) — 2 · 1015 ions/cm2; (b) — 2 · 1016 ions/cm2; and (c) — 2 · 1017 ions/cm2.
Fig. 4. (a) — surface roughness versus observation area size (for AFM-images 5 × 5 μm); (b) — local slope distribution.
246
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
Fig. 5. AFM-images of the treated surfaces at 20% elongation: (a) — 2 · 1015 ions/cm2; (b) — 2 · 1016 ions/cm2; and (c) — 2 · 1017 ions/cm2. The axis of deformation is vertical.
where s is the size of the area with points ns × ns in the AFM-image; zij are the heights, observed in the given area; b z N |s is the average height. Thus, we select the areas s × s in the AFM images and calculate Rrms(s) for each area. The estimated roughness of the area of size s is the average of local roughness: b Rrms N. To calculate roughness 10 images 5 × 5 μm (512 × 512 points) were obtained for each sample. The results are shown in Fig. 4a. Note that for larger images various surface artifacts (scratches, contaminations, etc) fall into the observation area. The roughness of such artifacts is out of interest in the scope of the present work. The surface topography can be divided into flat spatial triangular elements. The angle between the normal to the triangle and the vertical axis (the angle θ in spherical coordinates) represents the local slope of the surface — the low values of θ correspond to areas that are close to the horizontal ones. Distribution of the angles θ for the images under consideration is shown in Fig. 4b. Fig. 4a shows that in all cases the roughness increases asymptotically to a certain value. The roughness of the surfaces with PIII 2 · 1015 ions/cm2 and 2 · 10 16 ions/cm 2 differs only slightly. The sample with 2 · 1017 ions/cm2 has the highest value of roughness. The values of surface roughness (for s = 5 μm) are given in Table 1. With the increase of fluence, the distribution of slope angles (Fig. 4b) expands and shifts to higher values, i.e. the edges of folds become steeper.
Under uniaxial deformation, the structure of the treated surfaces changes. Fig. 5 shows the images of samples stretched by 20%; the axis of stretching is vertical. Elongation causes the appearance of cracks on the treated surfaces. The initial surface structure of materials in the low-fluence regime (see. Fig. 1b, c) is smoothed during stretching, and new folds that are parallel to the axis of deformation (Fig. 5a, b) appear — the result of sample compression in the direction orthogonal to the load. Similar, but less pronounced, longitudinal folds are visible on the surface of the sample in the high-fluence regime (Fig. 5c). However, in this case the initial surface texture also remains. The mechanical properties of the material in open cracks are similar to those of the untreated PU. Measurements of the crack depth in the vicinity of its tip located in the samples elongated by 20% (Fig. 5) made it possible to estimate the thickness of the surface layer. This value is 6 … 9 nm, which is consistent with the results given by the program TRIM. More precise measurements of the layer thickness and establishment of a correlation with fluence are impossible due to the presence of folds. The increase in elongation causes further crack growth and increase in longitudinal fold concentration. The surface of the material elongated twice and then unloaded is shown in Fig. 6. The surface structure of unloaded materials with a low fluence (Fig. 6a, b) is significantly different from the initial one (Fig. 1b, c). The
Fig. 6. AFM-images of the surfaces after removal of 100% elongation: (a) — 2 · 1015 ions/cm2; (b) — 2 · 1016 ions/cm 2; and (c) — 2 · 1017 ions/cm2.
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
247
Table 2 Adhesion properties of bacteria.
Fig. 7. Distributions of tip-sample adhesion force (a) and surface stiffness (b).
stretch-induced vertical folds, as well as the traces of closed horizontal cracks, are visible. This is indicative of the plastic properties of these surfaces. At the same time, the surface of the material subjected to 2 · 1017 ions/cm2 recovered completely, and no signs of loading were observed. Similar behavior was observed under dynamical mechanical loading — the surface of materials in the low-fluence regime became smoother, and a great number of cracks occurred, whereas the surface of the PU treated with 2 · 1017 ions/cm2 remained without visible changes. It is important to note that, despite the fact that the surface changes after application of load, the coating continuity is preserved. No peeled surface layers and open cracks were revealed in the unloaded materials. Distributions of the tip-surface adhesion force and surface hardness are shown in Fig. 7. According to the results, the adhesion force (Fig. 7a) decreases with increasing fluence. The accuracy of AFM adhesion measurements is affected by surface roughness [15]. For the 2 · 1017 ions/cm2 treated material the distribution of adhesion shows a weakly pronounced maximum at ~1 nN. Comparison of the surface and the corresponding adhesion map revealed that such small adhesion values were obtained for areas with a sharp slope (fold edges). Thus, a maximum in the vicinity of 1 nN is the measurement error associated with a change in
Treatment
βall, %
βliv
γliv
Untreated PU 2 · 1015 ions/cm2 2 · 1016 ions/cm2 2 · 1017 ions/cm2
62 ± 3.7 27 ± 3 15 ± 1.5 13 ± 2.5
1.45 ± 0.15 0.86 ± 0.07 0.81 ± 0.02 0.30 ± 0.03
1.38 ± 0.18 0.82 ± 0.02 0.87 ± 0.02 0.29 ± 0.004
the contact area of the tip-surface. The dotted line on the adhesion curve corresponds to actual values. The stiffness of the examined surfaces increases with increasing fluence (Fig. 7b). The estimated values of Young's modulus are: 4 (untreated material, Fig. 3a), 39 (2 · 1015 ions/cm2), 52 (2 · 1016 ions/cm2), and 100 (2 · 1017 ions/cm2) MPa. Due to the small thickness of the modified surface layer, the measured stiffness is the characteristic of the polymer–surface layer system. The definition of stiffness of the modified layer itself (for example, using finite element modeling of the indentation of a hard film on a soft substrate) is complicated by the small thickness and the inhomogeneous (non-elastic) properties of the modified polymer depending the distance from the surface. Fig. 8 shows examples of optical and AFM images of Staphylococcus cells on the surface of the material treated with PIII 2 · 1017 ions/cm2. The results presented here have shown that the adhesion of bacteria to surfaces depends on the fluence. The fraction of the area occupied by cells βall and the number of vital cells βliv (measured after one week of PIII treatment) and γliv (measured and after 11 months of storage of PIII treated samples in dark place at room temperature) are given in Table 2. The values are reduced drastically during the transition from the raw PU to the material subjected to minimal treatment. The number of vital cells has changed only insufficiently for materials treated with a dose 2 · 1015–2 · 1016 ions/cm2. The total number of cells is reduced, i.e. the amount of dead bacteria on the surface is reduced. After PIII 2 · 1017 ions/cm2 treatment βliv has decreased more than twice in comparison with the previous value and by almost 5 times in comparison with the untreated PU. The results of second measurements of vital cells γliv on the same samples after 11 months of storage did not shown significant difference between measurements on fresh samples. In the work by [16], it is shown that the roughness of a silicon substrate has an effect on the adhesion and growth of colonies of Staphylococcus cells only when Rrms N 200 nm, but the surface energy has no significant effect on cell growth. On the other hand, PereraCosta et al. [17] have revealed that the surface texture affects the growth of bacterial films even with low roughness. The results (Table 2) are correlated with the surface structure (roughness, fractal properties, and fraction of folds), hydrophilicity and the adhesion force between the AFM probe and the material (Fig. 7a). In the work [18] was shown that the more hydrophilic surface the higher resistance to the bacterial adhesion. In comparison with the untreated PU, the hydrophilicity of the material treated with 2 · 1015 ions/cm2 increased, and therefore the main factors initially decreasing the number of vital cells are reduction of wetting and the formation of the folded structure on the surface. Further reduction of cells (for the samples with higher fluences) can be attributed primarily
Fig. 8. Optical (a) and AFM (b) images of bacteria on the surface treated with 2 · 1017 ions/cm2.
248
I.A. Morozov et al. / Materials Science and Engineering C 62 (2016) 242–248
to the formation and growing of the folded structure, because the hydrophilicity remains unchanged. Another explanation is the formation of hard bumps (see. Fig. 2b, c) on the material surface. These bumps also hinder the consolidation of bacteria on the surface. Moreover, the surface energy of PIII treated polymers asymptotically decreases over time [19]. However, the number of vital cells on the old samples does not changed in comparison with the fresh ones. This can be another evidence that the surface structure is the main cause of the reduction of bacterial adhesion. 4. Conclusions The structural–mechanical properties of the surfaces of soft elastic polyurethane after plasma immersion ion implantation were studied. According to the AFM measurements, the thickness of the modified layer was 6–9 nm. The fractal folded structure is formed on the surface after the treatment. The surface roughness increases with increasing PIII treatment fluence, the folds become more sinuous (fractal dimension increases), and their fraction increases. The most rough and sinuous surface occurs after the 2 · 1017 ions/cm2 fluence treatment. Measurements of local mechanical properties show that the adhesion force between the probe and the surface decreases, and the surface stiffness increases with increasing treatment fluence. Under uniaxial deformation, the folds codirectional with the axis of stretching (due to compression in the transverse direction) and cracks are formed. The initial structure of the surface of materials treated with low fluence (2 · 1015–2 · 1016 ions/cm2) smoothes under stretching; after removal of the load the structure does not restore, the longitudinal folds and the traces of cracks remain, i.e. the surface exhibits plastic properties. At the same time, the surface of the material treated with 2 · 1017 ions/cm2 restores its initial structure after unloading, and the cracks closed without a trace. Similar pattern was observed in the study of materials subjected to dynamic loading. It should be noted that after removing the mechanical load in all the cases the coating continuity (delamination of the modified layer did not occur) was maintained. Investigation of Staphylococcus colonies grown on the surface of materials of interest shows a five-fold reduction in the number of vital cells in comparison with the untreated PU. Such differences are primarily due to the structural properties of the surfaces (the structure of the folds) supported by the increased hydrophilicity of the treated surface. These antibacterial properties remained unchanged at least after 11 months of surface treatment.
Acknowledgments This work was supported by the Russian Foundation for Basic Research (grants 14-04-00687, 13-01-96009_r_ural_a, 14-0896003_r_ural_a) and the Ministry of Education of Perm Region under agreement (S-26/632). References [1] T. Lu, Y. Qiao, X. Liu, Surface modification of biomaterials using plasma immersion ion implantation and deposition. A review, Interface Focus 2 (2012) 325–336. [2] A. Kurella, N.B. Dahotre, Surface modification for bioimplants: the role of laser surface engineering, J. Biomater. Appl. 20 (2005) 5–50. [3] B. Bhushan, B.K. Gupta, Handbook of Tribology: Materials, Coatings and Surface Treatments, McGraw-Hill, Inc., New York, 1991. [4] C.M. Cotell, Pulsed laser deposition and processing of biocompatible hydroxyapatite thin films, Appl. Surf. Sci. 69 (1993) 140–148. [5] A.C. Duncan, F. Weisbuch, F. Rouais, S. Lazare, C. Baquey, Laser microfabricated model surfaces for controlled cell growth, Biosens. Bioelectron. 17 (2002) 413–426. [6] E.N. Antonov, V.N. Bagratashvili, V.K. Popov, M.D. Ball, D.M. Grant, S.M. Howdle, C.A. Scotchford, Properties of calcium phosphate coatings deposited and modified with lasers, J. Mater. Sci. Mater. Med. 14 (2003) 151–155. [7] P. Alves, S. Pinto, H.C. de Sousa, M.H. Gil, Surface modification of a thermoplastic polyurethane by low-pressure plasma treatment to improve hydrophilicity, J. Appl. Polym. Sci. 122 (2011) 2302–2308. [8] J. Wang, P. Yang, H. Sun, J.Y. Chen, Y.X. Leng, G.J. Wan, N. Huang, Surface modification of medical polyurethane by acetylene plasma immersion ion implantation, 30th International Conference on Plasma Science 2003, p. 282 (Abstracts). [9] D.V. Bax, A. Kondyurin, A. Waterhouse, D.R. McKenzie, A.S. Weiss, M.M.M. Bilek, Surface plasma modification and tropoelastin coating of a polyurethane copolymer for enhanced cell attachment and reduced thrombogenicity, Biomaterials 35 (2014) 6795–6809. [10] H. Cao, X. Liu, F. Meng, P.K. Chu, Biological actions of silver nanoparticles embedded in titanium controlled by micro-galvanic effects, Biomaterials 32 (2011) 693–705. [11] W. Zhang, Y. Zhang, J. Ji, Q. Yan, A. Huang, P.K. Chu, Antimicrobial polyethylene with controlled copper release, J. Biomed. Mater. Res. A 83A (2007) 838–844. [12] http://www.srim.org/. [13] B. Pittenger, N. Erina, C. Su, Quantitative Mechanical Property Mapping at the Nanoscale with PeakForce QNM, Application Note, 1–12, Veeco Instruments Inc., 2010 [14] N. Otsu, A threshold selection method from gray-level histograms, IEEE Trans. Syst. Man Cybern. 9 (1979) 62–66. [15] D.-L. Liu, J. Martin, N.A. Burnham, Optimal roughness for minimal adhesion, Appl. Phys. Lett. 91 (2007) 043107. [16] H. Tang, T. Cao, X. Liang, A. Wang, S.O. Salley, J. McAllister, K.Y. Ng, Influence of silicone surface roughness and hydrophobicity on adhesion and colonization of Staphylococcus epidermidis, J. Biomed. Mater. Res. A 88A (2009) 454–463. [17] D. Perera-Costa, J.M. Bruque, M.L. González-Martín, A.C. Gómez-García, V. VadilloRodríguez, Studying the influence of surface topography on bacterial adhesion using spatially organized microtopographic surface patterns, Langmuir 30 (2014) 4633–4641. [18] Y.H. An, R.J. Friedman, Concise review of mechanisms of bacterial adhesion to biomaterial surfaces, J. Biomed. Mater. Res. 43 (1998) 338–348. [19] A. Kondyurin, M. Bilek, Ion Beam Treatment of Polymers: Application Aspects from Medicine to Space, second ed. Elsevier, 2014.