Synthesis, Characterization, and Bacterial Fouling-Resistance Properties of Polyethylene Glycol-Grafted Polyurethane Elastomers

Despite advances in material sciences and clinical procedures for surgical hygiene, medical device implantation still exposes patients to the risk of developing local or systemic infections. The development of efficacious antimicrobial/antifouling materials may help with addressing such an issue. In this framework, polyethylene glycol (PEG)-grafted segmented polyurethanes were synthesized, physico-chemically characterized, and evaluated with respect to their bacterial fouling-resistance properties. PEG grafting significantly altered the polymer bulk and surface properties. Specifically, the PEG-grafted polyurethanes possessed a more pronounced hard/soft phase segregated microstructure, which contributed to improving the mechanical resistance of the polymers. The better flexibility of the soft phase in the PEG-functionalized polyurethanes compared to the pristine polyurethane (PU) was presumably also responsible for the higher ability of the polymer to uptake water. Additionally, dynamic contact angle measurements evidenced phenomena of surface reorganization of the PEG-functionalized polyurethanes, presumably involving the exposition of the polar PEG chains towards water. As a consequence, Staphylococcus epidermidis initial adhesion onto the surface of the PEG-functionalized PU was essentially inhibited. That was not true for the pristine PU. Biofilm formation was also strongly reduced.


Introduction
Polymeric materials have a prominent place in biomedical applications, due to their broad range of physico-chemical properties that can be tailored to fit a wide plethora of applications [1][2][3]. Segmented polyurethanes (PUs) are among the most important classes of biomedical polymers, mainly due to their excellent hemocompatibility and unique mechanical properties deriving from the presence of hard segment-rich and soft segment-rich domains in a phase-separated microstructure. Such hard/soft-phase segregation in the polymer permits the combination of elastomeric properties that are typical of rubbers, with high mechanical resistance properties typical of thermoplastic materials [4]. A large variety of biomedical-grade PUs with different compositions, such as Biomer®, Pellethane®, and Cardiothane, are on the market, and their applications cover mainly cardiovascular devices, including central venous catheters and heart valves, but also artificial organs, scaffolds for tissue engineering, and wound dressings [5].
Despite the benefits of using PUs for intravascular devices manufacturing, complications are still associated with their use. Especially, these materials do not protect patients from the risk of developing localized or systemic bloodstream infections [6,7]. For this reason, considerable efforts have been polymer was also shown to be able to bind antibiotics [44] and complex antiseptic metal ions [45], thanks to the presence of a reactive carboxylic group per repeat unit. Such a group was now exploited for the grafting of PEG by a Steglich reaction mediated by di-cyclohexylcarbodiimide and 4dimethylaminopyridine (DMAP) activation. Specifically, three PEG:PEUA molar ratios (2:1, 3:1, and 5:1) were employed during synthesis, and an excess of PEG was chosen to reduce the probability of PEUA cross-linking by the bi-functional PEG. Among the three obtained PEG-functionalized polymers, the one resulting from a 2:1 PEG:PEUA molar ratio was found to be insoluble in organic solvents, and it also precipitated during functionalization, suggesting a high degree of polymer cross-linking. In contrast, a good degree of solubility in all common organic solvents (tetrahydrofurane, dimethyl formamide, and dimethyl sulfoxide) was shown by the polymer obtained from a 5:1 molar ratio while the polymer obtained with a 3:1 molar ratio showed intermediate behavior, resulting in only partial solubility in some solvents, specifically in di-methylformamide.
Nuclear magnetic resonance spectroscopy was used to evaluate the success of polymer synthesis and functionalization with PEG. In Figure 2, the 13 C-NMR spectra of pure PEUA and PEUA-PEG obtained with a 5:1 molar ratio are reported. Specifically, three PEG:PEUA molar ratios (2:1, 3:1, and 5:1) were employed during synthesis, and an excess of PEG was chosen to reduce the probability of PEUA cross-linking by the bi-functional PEG. Among the three obtained PEG-functionalized polymers, the one resulting from a 2:1 PEG:PEUA molar ratio was found to be insoluble in organic solvents, and it also precipitated during functionalization, suggesting a high degree of polymer cross-linking. In contrast, a good degree of solubility in all common organic solvents (tetrahydrofurane, dimethyl formamide, and dimethyl sulfoxide) was shown by the polymer obtained from a 5:1 molar ratio while the polymer obtained with a 3:1 molar ratio showed intermediate behavior, resulting in only partial solubility in some solvents, specifically in di-methylformamide.
Nuclear magnetic resonance spectroscopy was used to evaluate the success of polymer synthesis and functionalization with PEG. In Figure 2, the 13 C-NMR spectra of pure PEUA and PEUA-PEG obtained with a 5:1 molar ratio are reported.
As far as the 13 C-NMR spectrum of PEUA is concerned, the signal at 153 ppm was attributed to the urethanic C=O, the signal at 136 ppm to MDI-C 1 and MDI-C 4 , the signals at 120 ppm and 130 ppm to MDI-C 2 and MDI-C 3 , respectively, and the signal at 41 ppm to the CH 2 between the phenyl groups. The signals of the PPO soft phase were found at 18 ppm (methyl), 77 ppm (methine) and 73-75 ppm (methylene). The signals of DHMPA were at 46 ppm (quaternary carbon), 72 ppm (-CH 2 -), and ca. 20 ppm (-CH 3 ). Finally, at 175 ppm, the signal of the carboxylic group was observed. As far as the 13 C-NMR spectrum of PEUA-PEG is concerned, even if it was not possible to observe the peak related to the ester C=O, a new signal at 62 ppm was present, specifically the ester CH 2 of the linked PEG. Additionally, the peak of the DHMPA quaternary carbon shifted from 46 ppm to 50 ppm, presumably due to the different chemical environments. solubility in all common organic solvents (tetrahydrofurane, dimethyl formamide, and dimethyl sulfoxide) was shown by the polymer obtained from a 5:1 molar ratio while the polymer obtained with a 3:1 molar ratio showed intermediate behavior, resulting in only partial solubility in some solvents, specifically in di-methylformamide.
Nuclear magnetic resonance spectroscopy was used to evaluate the success of polymer synthesis and functionalization with PEG. In Figure 2, the 13 C-NMR spectra of pure PEUA and PEUA-PEG obtained with a 5:1 molar ratio are reported.  As far as the 13 C-NMR spectrum of PEUA is concerned, the signal at 153 ppm was attributed to the urethanic C=O, the signal at 136 ppm to MDI-C1 and MDI-C4, the signals at 120 ppm and 130 ppm to MDI-C2 and MDI-C3, respectively, and the signal at 41 ppm to the CH2 between the phenyl groups. The signals of the PPO soft phase were found at 18 ppm (methyl), 77 ppm (methine) and 73-75 ppm (methylene). The signals of DHMPA were at 46 ppm (quaternary carbon), 72 ppm (-CH2-), and ca. 20 ppm (-CH3). Finally, at 175 ppm, the signal of the carboxylic group was observed. As far as the 13 C-NMR spectrum of PEUA-PEG is concerned, even if it was not possible to observe the peak related to the ester C=O, a new signal at 62 ppm was present, specifically the ester CH2 of the linked PEG. Additionally, the peak of the DHMPA quaternary carbon shifted from 46 ppm to 50 ppm, presumably due to the different chemical environments.
In Figure 3A, the 1 H-NMR spectrum of PEUA is reported. The peaks at 9.6 and 10 ppm were attributed to urethane (NH), and the signal splitting was presumably due to the different linkers (PPO or DHMPA). The signals in the range of 7-8.0 ppm were attributed to the MDI aromatic ring, and the signal at 3.8 ppm was related to -CH2-between the two phenyls. The PPO methine and methylene groups showed a wide peak at 3.4 ppm when linked to the ether oxygen, and at 4.9 ppm when linked to the urethane NH. The signal at 3.5 ppm was attributed to the DHMPA methylene groups while the signal at 1.1 ppm was related to the methyl groups of DHMPA and PPO.
The signals of the PEG -CH2-groups in PEUA-PEG are in the same spectral range (ca. 3.6 ppm) as the protons of PPO ( Figure 3B). The yield of functionalization was calculated by the ratio of the intensity of the peak at 7 ppm (aromatic ring), and that of the peak at 3.6 ppm (methylene groups), Numeration of methylene bis-phenyl-diisocyanate (MDI) carbons in PEUA (A); 13 C-NMR spectrum of PEUA (B) and PEUA-PEG obtained with a 5:1 PEG:PEUA molar ratio (C).
In the inset of B, the magnification of the signal at 175 ppm, attributed to the PEUA carboxylic group, is reported.
In Figure 3A, the 1 H-NMR spectrum of PEUA is reported. The peaks at 9.6 and 10 ppm were attributed to urethane (NH), and the signal splitting was presumably due to the different linkers (PPO or DHMPA). The signals in the range of 7-8.0 ppm were attributed to the MDI aromatic ring, and the signal at 3.8 ppm was related to -CH 2 -between the two phenyls. The PPO methine and methylene groups showed a wide peak at 3.4 ppm when linked to the ether oxygen, and at 4.9 ppm when linked to the urethane NH. The signal at 3.5 ppm was attributed to the DHMPA methylene groups while the signal at 1.1 ppm was related to the methyl groups of DHMPA and PPO.
The signals of the PEG -CH 2 -groups in PEUA-PEG are in the same spectral range (ca. 3.6 ppm) as the protons of PPO ( Figure 3B). The yield of functionalization was calculated by the ratio of the intensity of the peak at 7 ppm (aromatic ring), and that of the peak at 3.6 ppm (methylene groups), from which the contribution of PPO was subtracted. The resulting yield of esterification was ca. 30% in the case of the polymer obtained with a 5:1 PEG:PEUA molar ratio, and 20% for the soluble portion of the polymer obtained with a 3:1 molar ratio.  In segmented polyurethanes, differential scanning calorimetry can allow for not only the determination of the polymer transition temperatures, but also the estimation of polymer hard/soft phase segregation. Specifically, information on polymer phase segregation can be extrapolated by the glass transition temperature (Tg), which, being characteristic of the soft phase of the polymer, depends on the soft phase mobility, and is affected by the interaction of the soft phase with the hard phase. An increase in the Tg value usually suggests low mobility of the soft phase, as a consequence of an increase in hard/soft phase mixing [46]. In Figure 4, as an example, the thermograms of PEUA and PEUA-PEG30 in cycles I and II of heating are reported, while in Table 1, the values of the glass transition temperature, and the variation of the specific heat (ΔCp) at Tg for all of the polymers are reported. The Tg value of free PPO, found in the literature [47], is also reported in Table 1 for comparison. In segmented polyurethanes, differential scanning calorimetry can allow for not only the determination of the polymer transition temperatures, but also the estimation of polymer hard/soft phase segregation. Specifically, information on polymer phase segregation can be extrapolated by the glass transition temperature (T g ), which, being characteristic of the soft phase of the polymer, depends on the soft phase mobility, and is affected by the interaction of the soft phase with the hard phase. An increase in the T g value usually suggests low mobility of the soft phase, as a consequence of an increase in hard/soft phase mixing [46]. In Figure 4, as an example, the thermograms of PEUA and PEUA-PEG 30 in cycles I and II of heating are reported, while in Table 1, the values of the glass transition temperature, and the variation of the specific heat (∆C p ) at T g for all of the polymers are reported. The T g value of free PPO, found in the literature [47], is also reported in Table 1 for comparison.  In the first cycle of heating, a flex, corresponding to the glass transition of the soft phase, and a broad endothermic band, corresponding to the melting of hard microdomains, can be observed in the thermograms of all polymers. Such an endothermic transition was no longer observable in the II cycle of heating. The Tg values of the PEG-functionalized polyurethanes were similar and lower than that of PEUA (Table 1). That suggests a greater phase segregation in the PEG-functionalized polyurethanes, compared to pristine PEUA. Presumably, PEG chains introduced in the hard phase of PEUA favored the cohesion of the hard domains by hindering the establishment of H-bonds among the urethane -NH groups of the hard phase and the ether oxygens of the soft phase. Additionally, some cross-linking between the polymer hard segments could occur, and further contribute to the segregation of the PEG-functionalized polymers. The variation in specific heat in correspondence with glass transition is also lower in the PEG-functionalized polymers compared to PEUA ( Table 1), suggesting that a lower amount of the soft phase is involved in the glass transition. Presumably, a higher cohesion of the hard micro-domains, or the partial cross-linking of hard segments in the PEGfunctionalized polymers can reduce the amount of the mobile portion of the soft phase. Such a phenomenon has already been described in the literature for segmented polyurethanes functionalized with carboxylate or sulfonate ions [48]. Table 1. Glass transition temperature (Tg) and variation of specific heat (ΔCp) at the glass transition temperature for PEUA and PEG-functionalized polyurethanes. The glass transition temperature of the free soft phase (PPO) is also reported, for comparison. (*) The subscript indicates that the PEUA:PEG molar ratio employed during functionalization. Indeed, it was not possible to determine the functionalization degree for this polymer, as it is insoluble in common solvents.

Polymer
Tg (°C) −67 -In Figure 5, the thermogravimetric curves of the pristine and functionalized polymers are reported. The first derivative of the weight vs temperature is also reported. As it can be observed, PEUA-PEG1:2 (obtained with a PEG:PEUA molar ratio 2:1) was the most thermally stable polymer, presumably because of the higher crosslinking content. Pristine PEUA degraded mainly in one step from 250 to 400 °C, while the PEG-functionalized polymers showed two steps of degradation. The first one was attributed to PEG weight loss, and the second one to the degradation of the polyurethane backbone. By taking into consideration the weight of the PEUA repeat unit, a good In the first cycle of heating, a flex, corresponding to the glass transition of the soft phase, and a broad endothermic band, corresponding to the melting of hard microdomains, can be observed in the thermograms of all polymers. Such an endothermic transition was no longer observable in the II cycle of heating. The T g values of the PEG-functionalized polyurethanes were similar and lower than that of PEUA (Table 1). That suggests a greater phase segregation in the PEG-functionalized polyurethanes, compared to pristine PEUA. Presumably, PEG chains introduced in the hard phase of PEUA favored the cohesion of the hard domains by hindering the establishment of H-bonds among the urethane -NH groups of the hard phase and the ether oxygens of the soft phase. Additionally, some cross-linking between the polymer hard segments could occur, and further contribute to the segregation of the PEG-functionalized polymers. The variation in specific heat in correspondence with glass transition is also lower in the PEG-functionalized polymers compared to PEUA ( Table 1), suggesting that a lower amount of the soft phase is involved in the glass transition. Presumably, a higher cohesion of the hard micro-domains, or the partial cross-linking of hard segments in the PEG-functionalized polymers can reduce the amount of the mobile portion of the soft phase. Such a phenomenon has already been described in the literature for segmented polyurethanes functionalized with carboxylate or sulfonate ions [48]. Table 1. Glass transition temperature (T g ) and variation of specific heat (∆C p ) at the glass transition temperature for PEUA and PEG-functionalized polyurethanes. The glass transition temperature of the free soft phase (PPO) is also reported, for comparison. (*) The subscript indicates that the PEUA:PEG molar ratio employed during functionalization. Indeed, it was not possible to determine the functionalization degree for this polymer, as it is insoluble in common solvents.
Soft phase (PPO) −67 -In Figure 5, the thermogravimetric curves of the pristine and functionalized polymers are reported. The first derivative of the weight vs. temperature is also reported. As it can be observed, PEUA-PEG 1:2 (obtained with a PEG:PEUA molar ratio 2:1) was the most thermally stable polymer, presumably because of the higher crosslinking content. Pristine PEUA degraded mainly in one step from 250 to 400 • C, while the PEG-functionalized polymers showed two steps of degradation. The first one was attributed to PEG weight loss, and the second one to the degradation of the polyurethane backbone. By taking into consideration the weight of the PEUA repeat unit, a good agreement between the PEG weight loss and the 30% esterification yield determined by 1 H-NMR was found for PEUA-PEG 30 . That was not true for PEUA-PEG 20 , presumably because of the chemical heterogeneity between the polymer soluble portion submitted to 1 H-NMR, and the whole-polymer sample (soluble portion + insoluble portion) submitted to thermogravimetric analysis. agreement between the PEG weight loss and the 30% esterification yield determined by 1 H-NMR was found for PEUA-PEG30. That was not true for PEUA-PEG20, presumably because of the chemical heterogeneity between the polymer soluble portion submitted to 1 H-NMR, and the whole-polymer sample (soluble portion + insoluble portion) submitted to thermogravimetric analysis. Material surface hydrophilicity/hydrophobicity is a very important feature that affects bacterial adhesion to the material itself. In general, it has been recently shown that superhydrophobic and superhydrophilic surfaces can both prevent microbial adhesion [49]. Superhydrophobic surfaces are those resembling the lotus leaf, which has been shown to have a water contact angle that is higher than 150°, and self-cleaning properties, thanks to its hierarchical micro/nanostructured surface covered with a low surface energy waxy hydrophobic film [50]. On the other side, very hydrophilic surfaces have intrinsic antifouling properties, due to the formation of a dense layer of water molecules, which weakens the interaction between the bacterial cell surface and the material surface [29,51].
To evaluate the effect of PEG-grafting on polymer hydrophilicity, polymer swelling in water and the dynamic contact angle were evaluated. Since these analyses, as well as the subsequent mechanical and biological tests, were performed on polymer films obtained by solvent casting (see Experimental Section), only the soluble polymers PEUA and PEUA-PEG30 will be subsequently characterized.
In Figure 6A, the swelling curves for the two polymers are reported. Material surface hydrophilicity/hydrophobicity is a very important feature that affects bacterial adhesion to the material itself. In general, it has been recently shown that superhydrophobic and superhydrophilic surfaces can both prevent microbial adhesion [49]. Superhydrophobic surfaces are those resembling the lotus leaf, which has been shown to have a water contact angle that is higher than 150 • , and self-cleaning properties, thanks to its hierarchical micro/nanostructured surface covered with a low surface energy waxy hydrophobic film [50]. On the other side, very hydrophilic surfaces have intrinsic antifouling properties, due to the formation of a dense layer of water molecules, which weakens the interaction between the bacterial cell surface and the material surface [29,51].
To evaluate the effect of PEG-grafting on polymer hydrophilicity, polymer swelling in water and the dynamic contact angle were evaluated. Since these analyses, as well as the subsequent mechanical and biological tests, were performed on polymer films obtained by solvent casting (see Experimental Section), only the soluble polymers PEUA and PEUA-PEG 30 will be subsequently characterized.
In Figure 6A, the swelling curves for the two polymers are reported.
As it can be observed, PEUA-PEG 30 was more strongly hydrophilic than PEUA, reaching a swelling degree at an equilibrium of ca. 150%, compared to ca. 25% of PEUA. Such a big increase in polymer hydrophilicity after PEUA functionalization with PEG is surely due to the PEG intrinsic hydrophilic properties, but presumably also to the good hard/soft phase segregation of PEUA-PEG 30 , which makes the polymer soft phase free to move and interact with water molecules. This hypothesis was confirmed by the calculation of the Diffusion Coefficient (D) of water in each of the two polymeric matrices. Specifically, in a diffusional regime in which the Fick's first law is valid, D can be calculated by the following equation [52]: W tot W ∞ tot = 4t 1/2 D 1/2 Lπ 1/2 where W eq is the amount of water adsorbed at the equilibrium, W t is the amount of water that is adsorbed at the specific time t, and L is the polymer film thickness (which was assumed to be constant during swelling). This expression describing the transport behavior of the penetrant into the polymer can be applied only to the initial stages of swelling, i.e., up to a 60 % increase in the mass of the polymer (W t /W eq. ≤ 0.6) [53]. The values of D were obtained from the plot of the ratio of the swollen polymer mass at time t, and t = eq. (W t /W eq ), as a function of the ratio of the square root of time (t 1/2 ), by following the initial slope method ( Figure 6B). The following D values were extrapolated: D PEUA = (1.7 ± 0.1) × 10 −9 cm 2 · s −1 , and D PEUA-PEG30 = (5.0 ± 0.2) × 10 −9 cm 2 ·s −1 . The higher D value of water in PEUA-PEG 30 compared to PEUA suggests a higher ability of water molecules to diffuse into the polymer matrix, presumably due to a higher mobility of the hydrophilic soft phase achieved by the higher polymer hard/soft phase segregation.
Material surface hydrophilicity/hydrophobicity is a very important feature that affects bacterial adhesion to the material itself. In general, it has been recently shown that superhydrophobic and superhydrophilic surfaces can both prevent microbial adhesion [49]. Superhydrophobic surfaces are those resembling the lotus leaf, which has been shown to have a water contact angle that is higher than 150°, and self-cleaning properties, thanks to its hierarchical micro/nanostructured surface covered with a low surface energy waxy hydrophobic film [50]. On the other side, very hydrophilic surfaces have intrinsic antifouling properties, due to the formation of a dense layer of water molecules, which weakens the interaction between the bacterial cell surface and the material surface [29,51].
To evaluate the effect of PEG-grafting on polymer hydrophilicity, polymer swelling in water and the dynamic contact angle were evaluated. Since these analyses, as well as the subsequent mechanical and biological tests, were performed on polymer films obtained by solvent casting (see Experimental Section), only the soluble polymers PEUA and PEUA-PEG30 will be subsequently characterized.
In Figure 6A, the swelling curves for the two polymers are reported. Measurements of dynamic contact angle confirmed the higher hydrophilicity of PEUA-PEG 30 , compared to PEUA. In such an analysis, the surface hydrophilicity (or wettability) can be evaluated [54]. In Figure 7, as an example, the profile of the two immersion cycles for PEUA-PEG 30 is shown, and in Table 2, the values of the contact angles, in advancing and in receding, and the contact angle hysteresis, are reported. As observed in Table 2, in the first cycle of immersion, the two polymers showed similar θ av , suggesting a similar wettability. However, PEUA-PEG 30 showed a significantly lower θ rec than PEUA (35 • vs. 47 • ) and, as a consequence, a higher hysteresis.
Such a finding is presumably related to the surface chemical heterogeneity of PEUA-PEG 30 (greater than PEUA). Additionally, in the second immersion cycle, PEUA-PEG 30 showed a kinetic contact angle hysteresis, evidenced by the decrease of θ adv from 94 • to 83 • , suggesting a molecular rearrangement of the polymer surface on wetting, which presumably involved the exposition of the polar PEG chains towards water.
Langmuir firstly observed the flipping of the surfactant molecules, as polar head groups migrated towards hydrophilic environments [55]. It was, later demonstrated that such re-organization was driven by the minimization of the surface free energy at the interface [56]. compared to PEUA. In such an analysis, the surface hydrophilicity (or wettability) can be evaluated [54]. In Figure 7, as an example, the profile of the two immersion cycles for PEUA-PEG30 is shown, and in Table 2, the values of the contact angles, in advancing and in receding, and the contact angle hysteresis, are reported. As observed in Table 2, in the first cycle of immersion, the two polymers showed similar θav, suggesting a similar wettability. However, PEUA-PEG30 showed a significantly lower θrec than PEUA (35 ° vs 47 °) and, as a consequence, a higher hysteresis.   59 46 Such a finding is presumably related to the surface chemical heterogeneity of PEUA-PEG30 (greater than PEUA). Additionally, in the second immersion cycle, PEUA-PEG30 showed a kinetic contact angle hysteresis, evidenced by the decrease of θadv from 94 ° to 83 °, suggesting a molecular rearrangement of the polymer surface on wetting, which presumably involved the exposition of the polar PEG chains towards water.
Langmuir firstly observed the flipping of the surfactant molecules, as polar head groups migrated towards hydrophilic environments [55]. It was, later demonstrated that such reorganization was driven by the minimization of the surface free energy at the interface [56].
The two polymers also showed significantly different mechanical properties, as shown by INSTRON analysis. In Figure 8, the stress-strain curves of PEUA and PEUA-PEG30 are reported.  The two polymers also showed significantly different mechanical properties, as shown by INSTRON analysis. In Figure 8, the stress-strain curves of PEUA and PEUA-PEG 30 are reported.
As it can be observed, both polymers showed an elongation at a break of ca. 10. The Young Modulus, determined from the slope of the initial linear trend, was also similar for the two polymers (E PEUA = 0.9 MPa e E PEUA-PEG30 = 0.7 MPa). However, the trend of the two stress-strain curves was very different. Indeed, PEUA showed a Yield Point at an elongation of ca. 0.7, in correspondence with which a collapse of the material resistance was observed. In contrast, PEUA-PEG 30 showed a typical trend of elastomeric materials, in which the Yield Point was not really clear, and the stress increased with the elongation up to material break. The stress at break of PEUA-PEG 30 is also ca. one order of magnitude higher than that of PEUA (0.06 MPa). Overall, the mechanical properties of PEUA-PEG 30 are coherent, with a marked hard/soft phase segregation in this polymer as evidenced by DSC analysis. Presumably, the presence of strong chain-chain interactions hinders chain slippage, and make the material resistant to big deformations. In addition, the occurrence of chemical cross-linkages between the hard segments cannot be excluded, even if the solubility of the sample suggests a case of low cross-linking. Of course, if present, hard-hard segment cross-linkages can further justify the elastomeric properties of the polymer.  59 46 Such a finding is presumably related to the surface chemical heterogeneity of PEUA-PEG30 (greater than PEUA). Additionally, in the second immersion cycle, PEUA-PEG30 showed a kinetic contact angle hysteresis, evidenced by the decrease of θadv from 94 ° to 83 °, suggesting a molecular rearrangement of the polymer surface on wetting, which presumably involved the exposition of the polar PEG chains towards water.
Langmuir firstly observed the flipping of the surfactant molecules, as polar head groups migrated towards hydrophilic environments [55]. It was, later demonstrated that such reorganization was driven by the minimization of the surface free energy at the interface [56].
The two polymers also showed significantly different mechanical properties, as shown by INSTRON analysis. In Figure 8, the stress-strain curves of PEUA and PEUA-PEG30 are reported. As it can be observed, both polymers showed an elongation at a break of ca. 10. The Young Modulus, determined from the slope of the initial linear trend, was also similar for the two polymers (EPEUA = 0.9 MPa e EPEUA-PEG30 = 0.7 MPa). However, the trend of the two stress-strain curves was very As it can be noted, S. epidermidis can easily adhere to PEUA. Indeed, the presence of S. epidermidis colonies in the first stages of adhesion could be observed on the PEUA surface ( Figure 9A). In contrast, PEUA-PEG30 was essentially free from colonization ( Figure 9B). Such different initial bacterial adhesion rates were reflected in the biofilm formations on the two polymer surfaces. Indeed, heavy colonization of the PEUA surface was observed with the presence of large biofilm structures ( Figure 9C) after 24 hr incubation, while just a few sporadic adhering cells and small bacterial aggregates were present on PEUA-PEG30 ( Figure 9D). Such difference in biofilm formation was confirmed by the cell count, the number of CFU per surface unit (CFU/cm 2 ) being 4 × 10 8 and 7 × 10 5 for PEUA and PEUA-PEG30, respectively. The excellent ability of PEUA-PEG30 to reduce bacterial adhesion and biofilm formation is definitely related to its higher bulk and surface hydrophilicity which also presumably involve the exposition of PEG chains at the material/water interface, as shown by dynamic contact angle analysis. However, since the polymer does not completely inhibit bacterial adhesion, such a system should be used in combination with other anti-biofilm molecules, to maximize its performance [15,57,58]. Unexpectedly, the mechanical resistance of the PEG-functionalized polymer also improved because the presence of PEG, with benefits for the application. As it can be noted, S. epidermidis can easily adhere to PEUA. Indeed, the presence of S. epidermidis colonies in the first stages of adhesion could be observed on the PEUA surface ( Figure 9A). In contrast, PEUA-PEG 30 was essentially free from colonization ( Figure 9B). Such different initial bacterial adhesion rates were reflected in the biofilm formations on the two polymer surfaces.
Indeed, heavy colonization of the PEUA surface was observed with the presence of large biofilm structures ( Figure 9C) after 24 h incubation, while just a few sporadic adhering cells and small bacterial aggregates were present on PEUA-PEG 30 ( Figure 9D). Such difference in biofilm formation was confirmed by the cell count, the number of CFU per surface unit (CFU/cm 2 ) being 4 × 10 8 and 7 × 10 5 for PEUA and PEUA-PEG 30 , respectively.
The excellent ability of PEUA-PEG 30 to reduce bacterial adhesion and biofilm formation is definitely related to its higher bulk and surface hydrophilicity which also presumably involve the exposition of PEG chains at the material/water interface, as shown by dynamic contact angle analysis. However, since the polymer does not completely inhibit bacterial adhesion, such a system should be used in combination with other anti-biofilm molecules, to maximize its performance [15,57,58]. Unexpectedly, the mechanical resistance of the PEG-functionalized polymer also improved because the presence of PEG, with benefits for the application.
Functionalization of PEUA with PEG was achieved by a DCC-activated esterification of polymer carboxylic acids, by using DMAP as an accelerator [59]. Specifically, DMAP and DCC, in equimolecular amounts with respect to the polymer carboxyl groups, were added to a 5% (w/v) solution of PEUA in dichloromethane at 0 • C. After 20 min, PEG was added in molar excess (2:1, 3:1 and 5:1) with respect to PEUA to avoid polymer crosslinking. Temperature was then raised to 25 • C and the solution was left under stirring for 24 h. After the reaction, dicyclohexylurea precipitate during polymer functionalization was filtered. Then, the polymer was recovered by precipitation in hexane, and washed in water to eliminate the unreacted PEG. The polymer, whose repetitive unit is reported in Figure 1, was named PEUA-PEG X, where x is the functionalization degree.
Differential scanning calorimetry (DSC) was performed from −100 to +150 • C under N 2 flux by using a Mettler TA-3000 DSC apparatus (Mettler Toledo, Columbus, Ohio, US). The scan rate used for the experiments was 10 • C·min −1 and the sample weight was 6-7 mg. Thermo-gravimetric analysis (TGA) was carried out employing a Mettler TG 50 thermobalance at a heating rate of 10 • C·min −1 under N 2 flow in the temperature range 25-600 • C.
Swelling measurements of both the pristine and functionalized polyurethanes were performed at room temperature in water. For such experiments, circular test samples (1 cm in diameter and 100 µm in thickness) were obtained by cutting cast films of the polymers. Polymer films were obtained by dissolving the polymer in tetrahydrofuran (5% w/v) and layering such a solution on circular Teflon plates, 5 cm in diameter. After solvent evaporation at room temperature, polymer films were detached from the Teflon plates with the aid of tweezers. The polyurethane samples were then immersed in water. At the determined times, specimens were removed from the water and weighed, after removal of the excess of the solvent using filter paper. The analysis was repeated until a constant weight (equilibrium swelling) was reached. The swelling degree, SD, was calculated by applying the following equation: where W t is the weight of the sample after swelling at time t, and W 0 was the initial weight of the film. Five parallel swelling experiments were performed for each sample, and data were reported as the average value ± standard deviation. The dynamic contact angle was performed by using a Dynamic Contact Angle Analyzer Cahn DCA 312 (CAHN Instruments, Inc., South Carmenita Road, Cerritos, CA, USA), on the basis of the Wilhelmy balance method. Such a method consists of detecting by a balance the change in the weight of a thin vertical plate when it is brought in contact with a liquid. The detected weight change is a combination of the buoyancy and the wetting force, while the gravity force remains the same. Since the wetting force (F w ) is defined as: where γ lv is the liquid surface tension, p is the perimeter of the sample, and θ is the contact angle, the force change (F) detected by the balance is: where V is the volume of the displaced liquid, ∆ρ is the difference in density between the liquid and air, and g is the acceleration of gravity. Thus, as long as γ lv and the solid perimeter (p) are known, the contact angle value can be determined.
Contact angle measurements were performed with an immersion rate of 20 µm/s, carrying out two consecutive immersion cycles. In the second immersion cycle, the immerged area of the sample was greater than that of the first cycle, in order to verify the absence of the release of substances from the sample into water. The analyses were performed on samples obtained by layering the polymer on rectangular glass coverslide (1 cm × 2 cm) by solvent casting. When the sample was immersed into the liquid, the advancing contact angle (θ adv ) was recorded, while when the sample was emerging, the receding contact angle (θ rec ) was measured ( Figure 8). The difference between the advancing and the receding angles is called the contact angle hysteresis (H = θ adv − θ rec ), which mainly arises from surface roughness and/or chemical heterogeneity [60]. However, other factors, including liquid adsorption and retention, or molecular rearrangement on wetting, can contribute to hysteresis [61].
The mechanical properties of the pristine and functionalized polyurethanes were studied by tensile tests, which were carried out with an ISTRON 4502 instrument (Instorn Inc., Norwood, MA, USA). Measurements were performed on rectangular polymer films (50 mm long × 6 wide × 0.2 thick), by employing a 10 N load cell and a deformation rate of 50 mm·min −1 .

Bacterial Strains and Culture Medium
For an evaluation of the antifouling features of polymers, the oxacillin-resistant Staphylococcus epidermidis ATCC 35984, a strong exopolysaccharide producer [62], was routinely grown in tryptic soy agar (TSA) and tryptic soy broth (TSB). Biofilm production was assessed according to the protocol described by Francolini et al. [63].

Assessment of the Polymer Ability to Prevent Bacterial Adhesion and Biofilm Formation
To evaluate the effect of polyurethanes on bacterial adhesion and biofilm formation, test tubes were filled with 2.5 mL of a bacterial suspension (0.5 McFarland) and grown at 37 • C in TSB supplemented with 1% glucose, to promote massive exopolysaccharide production. One tube was used as a reference (control), while polymer discs (1 cm in diameter, 100 µm in thickness) were introduced into the other tubes. Tubes were incubated at 37 • C for 30 min to assess early bacterial adhesion, or for 24 h to assess late bacterial adhesion [64,65].
After incubation, bacterial adhesion and biofilm formation were assessed by scanning electron microscopy (SEM, Assing, Rome, Italy). Specifically, polymer discs were collected, washed twice with PBS (pH 7.4) to remove loosely adherent cells, fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) at room temperature for 30 min, dehydrated through graded ethanol, treated with hexamethyldisilazane for 20 min, and gold-sputtered for SEM observation.
Polymer discs incubated with bacterial suspension for 24 h (biofilm formation) were also submitted to colony forming unit (CFU) counts. Specifically, disks were put into test tubes with 10 mL of phosphate buffer, and sonicated for 5 min to remove the adherent cells. Six 10-fold dilutions were prepared, and three 10 µL aliquots of each dilution were plated onto TSA plates. Plates were then incubated at 37 • C for 18 h, and CFUs were counted and referred to the polymer surface unit (CFUs/cm 2 ).

Statistics
Analysis of variance comparisons were performed using Mini-Tab. Differences were considered significant for p < 0.05. Data are reported as means ± SD.

Conclusions
This study confirms the strong antifouling ability of polyethylene glycol materials, and shows the great potential for PEG-grafting to confer bacterial resistance properties to segmented polyurethanes. Indeed, the functionalization of a thermoplastic polyurethane with PEG resulted in a material with superior elastomeric properties, and the ability to prevent the adhesion of the Gram-positive S. epidermidis, a microbial pathogen that is commonly isolated in medical device-related infections. Additionally, since the developed antifouling material is intrinsically active, it does not exhaust its activity over time, and it could provide long-term protection, at least in principle. Under an applicative point of view, the PEG-functionalized polyurethane obtained in this study could find a role in several biomedical applications, spanning from intravascular medical device manufacturing to wound dressings, where it could be applied as an antifouling coating, or incorporated within the bulk structure.