Direct grafting of anti-fouling polyglycerol layers to steel and other technically relevant materials

Highlights

• Grafting of hyperbranched polyglycerol onto steel, aluminum, and silicon surfaces was achieved in a single step without additional reagents.

• Linear layer growth permits control of thickness by reaction time.

• Protein repulsion correlates with the layer thickness.

• Experiments with E. coli and Acinetobacter baylyi prove the antifouling behavior of these coatings.

Abstract

Direct grafting of hyperbranched polyglycerol (PG) layers onto the oxide surfaces of steel, aluminum, and silicon has been achieved through surface-initiated polymerization of 2-hydroxymethyloxirane (glycidol). Optimization of the deposition conditions led to a protocol that employed N-methyl-2-pyrrolidone (NMP) as the solvent and temperatures of 100 and 140 °C, depending on the substrate material. In all cases, a linear growth of the PG layers could be attained, which allows for control of film thickness by altering the reaction time. At layer thicknesses >5 nm, the PG layers completely suppressed the adhesion of albumin, fibrinogen, and globulin. These layers were also at least 90% bio-repulsive for two bacteria strains, E. coli and Acinetobacter baylyi, with further improvement being observed when the PG film thickness was increased to 17 nm (up to 99.9% bio-repulsivity on silicon).

Introduction

Biofouling, i.e. the settlement and growth of microorganisms on surfaces, has a considerable impact on the functionality and stability of materials in contact with water, such as the machines used in paper production, the food industry, and cooling towers, as well as water purification membranes, ships’ hulls, or deep water sensors [1], [2]. Bacterial adhesion influences the function of implants or catheters in a negative way, typically causing inflammations or infections [3], [4]. Corrosion, increased flow resistance, and clogging often result in increased operating costs and downtimes for technical systems.

In recent years, many approaches to prevent such interactions have been investigated [5]. While for certain applications, biocides are a suitable approach, other, less invasive strategies are more appropriate for different situations (e.g. to avoid poisoning of patients or the induction of resistances in bacteria). One of these less invasive strategies is based on the prevention of protein and bacteria adsorption using surface modification with a layer of protein resistant materials [6]. The aim of this strategy is to create bio-resistant surfaces, that are not recognized by bio-entities such as cells or proteins. In biomedical applications, these coatings can be used to reduce protein adsorption and bacteria adhesion while maintaining the mechanical and physical properties of the underlying substrate.

Since the early 1980s, surface-bound polyethyleneglycol (PEG, (CH₂CH₂O)_n) has been known to suppress or at least reduce the adsorption of proteins and cells through the formation of hydrogels [7]. This observation led to the development of self-assembled monolayers (SAMs) on gold based on thiolates bearing oligoethyleneglycol (OEG) or oligoethyleneglycol monomethylether (MeOEG) head groups, which are resistant to the adsorption of a variety of proteins and even reduce cell adhesion [8], [9]. These systems are still considered the benchmarks and are applied widely in bio-molecular research [9], [10], [11], [12], [13]. A serious drawback of OEG-based surface layers is that they are prone to autoxidation, in vivo digestion, and thermal decomposition [14], [15]. Furthermore, when exposed to light, the degradation accelerates [16], [17]. This instability limits their use for the modification of medical surfaces such as implants and catheters [14], [15], [18], [19], [20]. Alternative SAM systems have also been found to be protein-resistant, such as phosphorylcholines [21], [22], polycarboxybetaines [23], sulfobetaines [24], [25], oligopropylenesulfoxides [26], and a series of carbohydrate-based compounds [27], such as mannitol [28] and galactose terminated layers [29], but none of them turned out to be as bio-repulsive for a broad range of proteins as were the OEG-based monolayers.

Branched molecules have been predicted to prevent the adsorption of proteins more efficiently than linear molecules at comparable coverage [30], [31]. Therefore, coatings of branched molecules with anti-fouling properties attracted considerable interest for biomedical applications [32]. The multiple branched, or hyperbranched, versions of the OEG molecules are the polyglycerols (PG, Fig. 1) [30], [33], [34]. These molecules can be formed easily using the ring-opening polymerization of 2-hydroxymethyloxirane (glycidol) initiated by nucleophiles (‘Nuc⁻’ in Fig. 1) [35], [36]. In spite of the structural similarity, these molecules are significantly more stable than the OEG-based ones (e.g. the PG surfaces can even be sterilized at high temperatures) [37], [38]. The reasons for this improved thermal and radiative stability are still unknown to date.

In recent years, PG moieties were used in thiolate-based SAM systems to render gold surfaces protein-repulsive [38], [39], [40]. Similar strategies have been used to deposit preformed PG moieties onto different materials, such as the use of polylysine as anchoring group for titanium surfaces [41], triethoxysilane modified compounds on glass surfaces [42] or electrostatic forces for PG derivatives carrying an amino-terminated tether [43].

Several groups realized that the nucleophile in the polymerization reaction does not need to be a discrete molecule but can also be a surface bound species. Such a reaction results in the growth of covalently bound PG onto the solid surface, a process known as grafting [44]. Typically, first a primer layer is deposited which exposes suitable nucleophilic groups (e.g. amino or alkoxide groups), from which the grafting process is initiated [45], [46], [47]. On polymer surfaces, nucleophilic sites have been generated by oxygen plasma treatment [48], while on certain oxide surfaces, the exposed hydroxy groups have been used after activation through deprotonation [49], [50].

In this paper, we wish to describe the growth of hyperbranched polyglycerol layers on the native oxide layers of steel, aluminum, and silicon (Fig. 2) without the use of any chemical activator. The aim of this project was to simplify the existing protocols for the grafting of PG layers onto solid surfaces, particularly for technically relevant materials (the native silica layer on silicon served as a model for glass and many ceramic materials). To the best of our knowledge, the grafting of PG layers on steel has not been performed to date, although the enormous potential of antifouling layers for steel is apparent because it is the most common material for biomedical equipment, e.g. catheters or implants.

As the only pre-treatment step, the substrates were cleaned by oxygen plasma or piranha solution (H₂SO₄/H₂O₂) to remove residues, such as grease or adhesive polymers. Next, the grafting of the PG layers was directly performed on these surfaces. The resulting layers were characterized using ellipsometry, contact angle goniometry, and infrared reflection absorption spectroscopy (IRRAS). The relation between the layer thickness and the resistance to the adsorption of different bovine proteins (serum albumin, γ-globulin, and fibrinogen) was subsequently determined using ellipsometry. Substrates showing the complete suppression of protein adsorption were later used to study the adhesion of E. coli (GFP) and Acinetobacter baylyi as models for real-world fouling processes.