Introduction

Antifouling/biocompatible polymer surfaces and interfaces have attracted considerable attention because of their utility in membrane separation processes for water treatment, marine coatings and biomedical materials. Nonionic hydrophilic polymer surfaces with good antifouling/biocompatible properties have recently been developed.1 Numerous mechanisms have been proposed to explain the compatibility of such materials. For example, poly(2-methoxyethylacrylate) (PMEA) can suppress platelet adhesion remarkably well compared with poly(2-hydroxyethylmethacrylate) (PHEMA).2, 3 To clarify the origin of the excellent blood compatibility of PMEA, a previous study2 investigated the adsorption of proteins onto PMEA. The results indicated that the interactions between PMEA and proteins were weaker than those observed between PHEMA and proteins. The structure of water in the vicinity of polymer surfaces has been recognized as a fundamental factor for antifouling properties, as it plays a key role in preventing proteins and blood cells from directly interacting with the polymer interface.

Molecular dynamics (MD) calculations have provided useful information at the atomistic level in various fields,4, 5 including protein adsorption.6, 7, 8 To elucidate the correlation between antifouling/biocompatible properties and the behavior of water molecules at an interface, detailed physicochemical data are required. In particular, the adsorption free energy of proteins approaching a polymer surface is a key thermodynamic factor for evaluating their adsorption properties.9 From an engineering perspective, recent studies10, 11, 12 have shown that antifouling/biocompatible properties can be evaluated from the free energy profiles, which are generated from MD simulations in water, of amino acids approaching polymer repeat units. Specifically, the hydrophobic residue phenylalanine (Phe) can be conveniently used as a probe molecule because the profiles of Phe differ significantly depending on the nature of the repeat units.10 Simple strategies for assessing the interactions between hydrophobic probe molecules and polymer repeat units offer a convenient theoretical approach for initial screening of the antifouling/biocompatible properties of polymers. However, because of the high computational costs, it is generally difficult to acquire sufficient statistical samplings for the free energy calculations.13 However, if the free energy profiles of structurally simpler probe molecules, such as benzene, are in qualitative agreement with those of Phe, the computational cost for the initial screening of antifouling polymer repeat units would be reduced. Thus, structural simplification of the probe molecule could enable more sophisticated combinatorial screening. In this study, benzene was examined as a probe molecule for estimating antifouling properties using free energy calculations from MD simulations.

Materials and methods

Free energy profiles of benzene approaching the following five repeat units in an aqueous solution were calculated: m-phenylene isophthalamide (PA), polyethylene terephthalate (PET), polyvinyl alcohol (PVA), 2-hydroxyethylmethacrylate (HEMA), and 2-methoxyethylacrylate (MEA). The chemical structural formulas of the repeat units are shown in Figure 1. It should be noted that the three repeat units for PVA were applied in our simulations, while one repeat unit was used for the other four types of materials, similar to our previous studies.10, 11, 12 This procedure was used because the major objective of this work was to investigate whether Phe can be replaced with a benzene molecule as a simpler probe by observing the differences between the free energy profiles of benzene and Phe.

Figure 1
figure 1

Chemical structural formulae of (a) the polyethylene terephthalate repeat unit, (b) m-phenylene isophthalamide, (c) the hypothetical polyvinyl alcohol repeat unit, (d) 2-hydroxyethylmethacrylate and (e) 2-methoxyethylacrylate.

Full size image

The free energy profiles and potential of mean force values for a benzene molecule approaching each repeating unit were calculated from the summation of the difference in the Helmholtz free energy, as described in our previous studies.10, 11, 12 A schematic illustration of a benzene molecule approaching a repeat unit of PET is shown in Figure 2. In this work, non-bonding interactions were described with the 12-6 Lennard–Jones potential, in which the cutoff distance was set at 1.0 nm. The benzene molecules, polymer repeat units and water molecules were modeled with the consistent valence force field (CVFF),14 and the long-range electrostatics were considered in the particle mesh Ewald method.15 Each pair of benzene molecules and repeat units along with 512 water molecules were confined within boxes of dimensions ∼2.5 × 2.5 × 2.5 nm with densities set at 1 g cm−3. An NVT ensemble (the number of atoms as well as the volume and temperature of the box remained constant) was used in our simulations. A periodic boundary condition was applied in all three directions. The temperature was set at 293 K using a scaling method, and the initial velocity of each molecule was set according to the Boltzmann distribution. The duration of each MD step was 0.5 fs for the systems involving HEMA and MEA monomers and 1.0 fs for the systems involving the other repeat units. Potential of mean force was evaluated from MD simulations that were conducted for 3.2–10 ns, and the free energy profiles were used to assess the antifouling properties of the repeat units.

Figure 2
figure 2

Schematic illustration of a benzene molecule approaching a polyethylene terephthalate monomer. For clarity, surrounding water molecules are omitted. A full color version of this figure is available at Polymer Journal online.

Full size image

Results

The calculated free energy profiles of PA, PET, PVA, HEMA and MEA are shown in Figure 3. The horizontal axis represents the molecular distance between the centers of mass of the benzene molecule and the repeat unit. To investigate the variation of the calculated profiles as a function of the elapsed time, the profiles were also estimated by dividing the overall MD calculation into three equal time intervals. No significant fluctuations were observed over the simulation times; the profiles only varied within a range of ∼1 kJ mol−1, indicating the statistical accuracy of the obtained profiles. The profiles of PA and PET display a deeper and broader energetically stable region, whereas those of PVA, HEMA and MEA do not show a prominent minimum. The minimum values of the energy profiles increase in the order of PA<PET<PVA<HEMA<MEA. The profiles of PA and PET suggest that benzene has strong affinity for these repeat units, likely due to hydrophobic interactions with the phenyl groups in these materials, and remains in the vicinity of the respective polymeric surfaces. The profile of PVA suggests that it has intermediate antifouling properties, most likely because it is generally more hydrophilic than PA and PET. By contrast, the profiles of HEMA and MEA indicate that although benzene can approach these surfaces, it can separate from them more readily compared with the interactions between benzene and PA and PET. However, HEMA and MEA do not completely inhibit protein adsorption and cell adhesion,2 as confirmed by the shallow minimum observed in their respective profiles.

Figure 3
figure 3

Free energy profiles for a benzene molecule approaching a repeat unit of (a) m-phenylene isophthalamide, (b) polyethylene terephthalate, (c) polyvinyl alcohol, (d) 2-hydroxyethylmethacrylate and (e) 2-methoxyethylacrylate, in water at 293 K. A full color version of this figure is available at Polymer Journal online.

Full size image

To assess the availability of benzene as a probe molecule, the calculated minimum values of the energy profiles for benzene approaching the five types of materials studied were compared with those calculated for Phe approaching each repeat unit in previous studies,10, 11, 12 as shown in Figure 4. The minimum values associated with benzene and Phe increase in the same order: PA<PET<PVA<HEMA<MEA. In the systems involving repeat units of PA, PET, and PVA, the values associated with Phe were smaller than those associated with the benzene molecule. By contrast, comparable values were observed for systems involving HEMA and MEA. Because atomistic interactions are generally amplified for larger and bulkier molecules, the affinities between the repeat units and Phe are stronger than those for a benzene molecule, resulting in points that are more energetically stable in the profiles. However, it should be noted that the difference between the profiles of the different probe molecules is negligible when HEMA and MEA, which have stronger antifouling properties, are involved. The absolute differences and errors in the minimum values are worth investigating in future works. Regardless, the same trend was observed in both systems (that is, benzene and Phe) studied. Thus, these results confirm that the use of a structurally simpler probe molecule is a feasible and practical alternative approach for theoretically evaluating the antifouling properties of polymer repeat units.

Figure 4
figure 4

Corresponding minimum values of the free energy profiles associated with a benzene molecule or phenylalanine approaching a given repeat unit. HEMA, 2-hydroxyethylmethacrylate; MEA, 2-methoxyethylacrylate; PA, m-phenylene isophthalamide; PET, polyethylene terephthalate; PVA, polyvinyl alcohol.

Full size image

Discussion

A previous experimental study investigating the biocompatible/antifouling properties of PET, PHEMA and PMEA demonstrated that the amounts of adhered platelets decreased in the order of PET>PHEMA>PMEA.3 Another study2 reported that the amounts of proteins adsorbed onto PHEMA and PMEA were nearly equivalent. However, unlike PHEMA, PMEA can significantly decrease the amount of adhered platelets. These observations confirm the accuracy of our assessment of the antifouling properties, which is shown in Figure 4. However, it is unclear as to whether our calculated minimum values reflect the amount of adhered platelets and/or that of adsorbed proteins.

To examine the relationship between the hydrophilic and antifouling properties, the static water contact angles of PET,16, 17 PVA,18, 19 PHEMA2, 3 and PMEA2, 3 (Table 1) were considered. These data suggest that PMEA and PVA have similar hydrophilic properties, whereas PHEMA is the most hydrophilic. However, PMEA can suppress the adhesion of platelets more remarkably than PHEMA.2, 3 Considering that the calculated minimum values of the free energy profiles can be correlated with these experimental insights, as discussed above, it can be concluded that hydrophilic properties are a prerequisite for the inhibition of protein adsorption.12 A previous study2 suggested that the interactions between PMEA and proteins were weaker than those involving PHEMA. The microscopic behavior of water molecules in the vicinity of polymer materials can certainly be correlated with their antifouling properties. MD simulation is a powerful approach to investigating interatomic affinities in the vicinity of polymer surfaces.20 Hence, antifouling properties should be evaluated at the atomistic level using MD simulations in future works.

Table 1 Experimental static water contact angles of PET, PVA, PHEMA and PMEA at ambient temperature
Full size table

The objective of this work was to confirm the feasibility of performing initial screening of antifouling polymer materials with computational costs lower than those of our previous studies.10, 11, 12 The results obtained in this study strongly suggest that the affinity between a material and a probe molecule can be evaluated quantitatively, which can facilitate the efficient molecular design of antifouling materials.