Introduction

Polyethersulfones (PES) are amorphous thermoplastic polymers that are commonly used in medical applications due to their heat resistance, hydrolysis resistance, and biocompatibility.1 PES contains aromatic sulfone groups that impart this material with relatively high mechanical strength and a high glass transition temperature; ether linkages in this material provide flexibility to the polymer chain and lower the glass transition temperature.2 Due to its thermal stability, PES is mainly used as a replacement for polycarbonate when exposure to high temperatures is anticipated.3 For example, medical applications of PES include use in surgical tools and other devices that require autoclave sterilization.4

Several potential medical applications of PES have recently been described. For example, Schwenter et al.5 encapsulated fibroblasts and retinal pigment epithelial cells within polyethersulfone hollow fibers; the encapsulated primary human fibroblasts were shown to be useful for gene therapy-mediated delivery of human erythropoietin in a murine model. PES membranes are commonly used to prepare hemodialyzers for treatment of end-stage renal disease; examples include a high-flux dialyzer made of gamma-sterilized polyethersulfone (Bellco, Mirandola, Italy) and a low-flux dialyzer made of steam-sterilized polyethersulfone (Bellco).6 Unger et al.7 examined the use of hollow, porous PES fibers for use as a scaffold for tissue engineering. They noted that fibroblast, epithelial, keratinocyte, glial, and osteoblast cells adhered to and spread over the surfaces of the PES fibers. Adherence and growth of endothelial cells required coating of the fibers with either gelatin or fibronectin.7 Christopherson et al.8 examined proliferation and differentiation of neural stem/progenitor cells on laminin-coated electrospun PES fiber meshes with fiber diameters between 283 nm and 1452 nm. Lower amounts of cell aggregation and higher amounts of cell spreading were noted on PES fibers with larger diameters.8 Babaeijandaghi et al.9 described the use of elecrospun PES nanofiber scaffolds in wound healing. They showed that fibroblast proliferation on PES was similar to that on tissue culture polystyrene. In addition, they demonstrated improved collagen deposition and greater fibroblast maturation on PES than on a conventional dressing material, Vaseline (Unilever, Englewood Cliffs, NJ) gauze.

As noted by Gugumus,10 yellowing of PES associated with photooxidation takes place with outdoor exposure. A significant reduction in tensile strength is also associated with exposure. This phenomenon is associated with oxidation of the phenyl moieities in PES.10 The low photostability of PES may encumber the use of PES in wound dressings and external medical applications. For example, photodegradation of PES may result from use of germicidal ultraviolet (UV) lamps, use of UV sterilization procedures, or exposure to sunlight. The photoactivity of PES is a well-known phenomenon; for example, light exposure of PES may be utilized to initiate polymerization of hydrophilic monomers.11,12 The change in the optical absorbance of PES at a wavelength of 330 nm can also be used to develop dosimeters that evaluate solar UV exposure.13

Several approaches are available for mitigating the low photostability of PES.14 A benzotriazole UV absorber may be used to improve light stability; less brittleness and yellowing were noted after light exposure in light-exposed PES formulations that contained this UV absorber.15 Another approach for improving the photostability of PES involves depositing a conformal coating on the surface of PES by means of atomic layer deposition (ALD).16 Atomic layer deposition involves deposition of coatings using a layer-by-layer approach that involves alternating chemical reactions between gaseous precursor molecules and a solid surface. As noted by George, continuous, smooth, pinhole-free films can be prepared by driving reactions to completion for each reaction cycle.17 Use of appropriate processing parameters (e.g., precursor exposure times) may enable growth of conformal coatings on a variety of high-aspect-ratio nanoscale structures, including nanoporous structures.18 Atomic layer deposition is also known to protect other materials, including cellulosic paper, from degradation under UV exposure.19 It is important to note that during ALD on polymers, some ALD materials and process conditions will produce conformal coatings on the polymer surface, while other conditions will lead to precursor diffusion and reaction within the polymer subsurface region20,21 The extent of surface versus bulk reaction depends on the polymer structure as well as the detailed ALD conditions.22

In the current study, TiO2 coatings were deposited on PES membranes using atomic layer deposition to improve photostability and maintain biocompatibility of the PES membranes. TiO2 thin-film surfaces are associated with low cytotoxicity are utilized in variety of human health applications, including sunscreens, orthopedic implant coatings, and dental implant coatings23–25 Atomic layer deposition has been used to grow titanium oxide thin films on a variety of high-aspect-ratio nanostructures, including nanoparticles and nanotube thin-film structures.26–28 Low-temperature atomic layer deposition has been demonstrated with tetrakis-dimethyl-amido titanium (TDMAT) on several substrates; for example, Kaariainen et al.29 deposited a 50-nm TiO2 coating on polymethyl methacrylate (PMMA) using low-temperature atomic layer deposition.30 Because atomic layer deposition at low temperature is associated with slow deposition rates, modifications to the atomic laser deposition process (e.g., plasma assisted atomic layer deposition) may be used to perform low-temperature depositions at faster rates than thermal atomic layer deposition.31 The goals of this study were to determine: (I) if TiO2 can be coated onto PES membranes at relatively low temperatures using TDMAT and water as precursors, (II) whether the TiO2 coating protects the PES membrane from UV damage, and (III) whether the TiO2 coating has an effect on macrophage cell viability.

Materials and Methods

PES membranes (pore size = 0.8 μm, diameter = 47 mm) for coating and controls were obtained from a commercial source (Whatman, Maidstone, UK). These membranes exhibit visible circular pores and good sample-to-sample uniformity. Per the manufacturer, the as-obtained materials are hydrophilic, are stable in alkaline pH, exhibit smooth surfaces, and exhibit low protein binding.32 TiO2 deposition on the PES membranes was performed in a custom-built hot wall viscous flow tube reactor. The reactor, TiO2 filmstoichiometry, and TiO2 film characterization have been previously described.16 TDMAT and water were used as the precursor and reactant for TiO2 coating growth, respectively. TDMAT was purchased from a commercial source (Strem Chemicals, Newburyport, MA). The precursor cell was maintained at 27°C and the precursor was dosed into the reactor with an argon carrier gas. Argon was obtained from a commercial source (National Welders, Charlotte, NC) and was passed through a Drierite Air Purifier (W.A. Hammond Drierite Co. Ltd., Xenia, OH) to remove water. Deionized (DI) water was obtained from a DI water system (Millipore, Billerica, MA); it was used as the reactant and was maintained at a temperature of 27°C in a stainless steel bubbler during coating growth. Water was introduced into the reactor without an argon carrier gas.

The samples were placed in a custom-made quartz sample holder during the ALD deposition. Five 3 cm × 3 cm-area positions were available for holding samples in the quartz sample holder. These positions were numbered one (pointing toward the flow) to five (pointing toward the pump); position one was the first to receive doses from the source lines. During ALD processing, pressure was maintained at 2 Torr. Cycles consisted of a 10-s argon purge, a 5-s TDMAT dose, a 10-s argon purge, and a 5-s water dose. Prior to ALD processing, the substrates were introduced into the reactor, heated in vacuum (5 × 3 × 10−6 Torr) to 100°C, and dried for 3 min. Optimal processing times for purging and dosing were determined using a TC100 quadrupole mass spectrometer (QMS) (Leybold-Inficon, East Syracuse, NY). Scans were conducted over a range of 50 Da. Every point was measured with a dwell time of 16 ms. Because of the range and the capability of the QMS, data points could only be obtained every 2 s.

Thickness measurements of TiO2 coated-silicon were obtained using an Auto EL ellipsometer (Rudolph Technologies Inc., Flanders, NJ); an incidence angle of 75° was used in this study. Three measurements were made at random spots on each sample to assess the uniformity of the TiO2 coating. A film thickness of 21 nm was obtained from a sample that was prepared with 100 ALD cycles. Scanning electron microscopy (SEM) was performed on the using a JEOL 6400 cold-field emission scanning electron microscope (JEOL Ltd., Tokyo, Japan); this instrument was equipped with an energy dispersive x-ray (EDX) spectrometer attachment, a Link Pentafet detector (Link Analytical, Redwood City, CA), and a 4Pi Universal Spectral Engine pulse processor (4Pi Analytical, Hillsborough, NC).

TiO2-coated and uncoated PES membranes were assayed for potential toxicity using the MTT assay (CellTiter 96 Non-Radioactive Cell Proliferation Assay; Promega, Madison, WI). TiO2-coated PES membranes and uncoated PES membranes were sterilized with a 254-nm UV source (CL-1000; UVP, Upland, CA). All the membranes were exposed to UV light for 30 min on each side to ensure that both sides were sterilized. After completion of ultraviolet sterilization, the membranes were cut to fit the bottom of the wells using a hole punch and placed into 96-well plates. Non-UV-sterilized PES samples were cut and handled in a sterile environment; these materials did not undergo UV sterilization. RAW 264.7 cells (American Type Culture Collection, Manassas, VA), obtained from a mouse leukemic monocyte-macrophage cell line, were added to the 96-well plates at a cell density of 2 × 105 cells/mL. The cells were subsequently incubated for either 24 h or 48 h. MTT medium was prepared by adding dye solution at a 15:100 ratio to media as specified by the kit. To obtain a total volume of 115 μL, 15 μL of MTT dye was added to each well. The plates were then incubated under cell culture conditions for 3 h. After each incubation period, the solubilization solution/stop mix was added and the plates were incubated again for 1 h at 37°C. The wells were mixed and the contents were then transferred to duplicate wells in a 96-well plate for absorbance measurements. Absorbance was measured at λ = 570 nm (reference wavelength = 650 nm) using a 96-well OPTIMax plate reader (Molecular Devices, Sunnyvale, CA). Cells spiked with 5% dimethyl sulfoxide (DMSO) served as a positive control. The cells grown on the bottom of standard tissue culture polystyrene (TCPS) wells were used as negative controls. The data were normalized to the TCPS controls and were expressed as percent viability. All of the experimental and control materials were tested for interference with MTT dye against TCPS and did not cause detectable interference with the assay. The results from each data set were analyzed using Prism 4.0 statistical software (GraphPad Inc., La Jolla, CA). Results were expressed as mean ± standard deviation (SD). Statistical differences between control cells and other treatment groups, i.e., non-UV-sterilized PES, UV-sterilized PES, UV-sterilized TiO2-coated PES, and DMSO (positive cytotoxicity control) were assessed using one-way analysis of variance, with a Bonferroni post hoc test. Each experiment was repeated at least three times with each assay in duplicate. A p value of less than 0.05 was considered to be statistically significant.

RESULTS AND DISCUSSION

The size of the pores in TiO2-coated PES was noted to be smaller than that in unmodified PES; this difference in pore size was evident from the SEM images (Fig. 1a). The pore size was estimated to be approximately 350 nm for the coated sample. The uncoated samples had a nanoporous structure ranging from approximately 400 nm–900 nm, which was embedded within a larger microporous structure. SEM images revealed small cracks throughout the structure of UV-treated PES (Fig. 1d). Cracking may be due to chain scission. Yamashita et al.33 as well as Rivaton and Gardette34 have described UV degradation of PES in detail; they noted that this process involves chain scissions as well as formation of oxidative species, which are mainly of low molecular weight. Most of the oxidative species are created from photodegradation of diphenylethersulfone moieties and cleavage of aromatic rings.

Fig. 1
figure 1

SEM images of PES: (a) 21-nm layer of TiO2 on PES with smaller pores at 5 kV; (b) uncoated, non-UV-treated PES at 5 kV with visible fine pore structure; (c) uncoated non-UV-treated PES at 20 kV to emphasize large pores; and (d) uncoated UV-treated PES with visible crack formations (arrows). Scale bar at lower right of each image = 10 μm. The underlying pore structure of this material had pores of varying sizes. The larger pore structure developed cracks possibly due to photooxidative degradation during UV irradiation

Full size image

EDX analysis showed that titanium was present on the PES surfaces coated with TiO2 (Fig. 2a). No elemental differences were noted between the UV-treated and untreated PES surfaces (Fig. 2b, c). The UV-sterilized samples visibly changed color from a white color to a yellow color after UV exposure for 30 min on each side (Fig. 3). The TiO2-coated PES membrane did not show any color change after UV exposure. Rivaton and Gardette34 suggested that yellowing of PES is due to recombination of phenyl radicals.

Fig. 2
figure 2

EDX spectra confirming presence of (a) TiO2 layer on PES, and similar spectra of (b) UV-treated uncoated PES, and (c) non-UV-treated uncoated PES. Prominent Ti peaks across the sample surface indicate the layer was deposited uniformly along the underlying nanoporous structure. No elemental composition differences were detectable between UV-treated and non-UV-treated PES

Full size image
Fig. 3
figure 3

Color difference in uncoated PES cutouts before and after 30 min UV exposure on each side. The diameter of each cutout is approximately 10 cm. Discoloration indicated that PES is susceptible to a 254-nm germicidal UV source. UV irradiation likely cut chain bonds in the polymer and produced free radicals, which reacted with available oxygen to produce a color change

Full size image

The UV-treated PES without TiO2 coatings exhibited the most toxicity of the analyzed groups, with the lowest cell viability using the MTT assay (Fig. 4). After 24 h or 48 h incubation of macrophages with UV-sterilized PES, cell viability was 51% and 18%, respectively, of control cells. In contrast, TiO2-coated PES exposed for 30 min to UV treatment were not cytotoxic; the cell viability of TiO2-coated PES was not significantly different from that of controls. After 24 h or 48 h incubation of macrophages with TiO2-coated PES, cell viability was 107% and 89%, respectively, of control cells. Nonsterilized (no UV exposure) PES exhibited no significant differences in cell viability compared to control cells; viability was 89% after 24 h exposure and 74% after 48 h exposure.

Fig. 4
figure 4

Macrophage cell viability of different PES surfaces and controls expressed as a percentage of the cell and media control. DMSO (5%) was added to cells as a positive control treatment. Percentages for surviving cells were calculated by dividing the raw absorbance values for each treatment by the average absorbance from all corresponding control cells. The results are expressed as mean ± SD from three independent experiments. UV-treated PES without TiO2 showed the lowest cell viability, while TiO2-coated PES was not significantly different from controls. This indicated that UV-exposed PES is cytotoxic and that TiO2 coatings may protect PES polymer from UV damage and cytotoxicity. Asterisks indicate a statistically significant difference (p < 0.05) between the mean for a given PES treatment group and the mean for control cells

Full size image

Our results were consistent with the general low cytotoxicity of TiO2 surfaces. UV irradiation of PES resulted in the polymer becoming highly cytotoxic to macrophages, which was likely related to the above-mentioned changes in surface physical and chemical properties. In contrast, TiO2-coated PES surfaces were well tolerated by macrophages. Although in our case the material had a noticeable change in color, our results do not distinguish whether the toxicity coincides with the amount of color change. The lack of noticeable color change in the UV-treated TiO2-coated PES suggests that there were few defects in the TiO2 coating. The cytotoxicity results also suggest that the TiO2 coating contains few defects because pinholes or other coating defects may allow the PES membrane to become damaged upon UV exposure.

Conclusions

In summary, an ALD-grown TiO2 thin film was shown to be an effective noncytotoxic UV blocker for PES. The coated polymer did not change in color, show physical damage, or show cracking after exposure to UV light. These results indicate that a uniform TiO2 coating can protect UV-sensitive polymer surfaces from UV sources such as germicidal lamps and sunlight. Future studies will be needed to determine what types of physical and chemical changes (e.g., what types of chemical leachates) are responsible for the cytotoxic response in macrophages and whether this response occurs across other cell types. The oxygen content and surface roughness of TiO2-coated PES membranes after ultraviolet light exposure will also be considered35 It should also be noted that the anatase phase of TiO2 and to a lesser extent the rutile phase of TiO2 generate radicals under UV light illumination, which are associated with antimicrobial activity36 For example, Ahn et al.36 noted that UV light-irradiated anatase titanium oxide had significant activity against Streptococcus anguinis. The photocatalytic activity of atomic layer deposition-grown anatase TiO2 coatings on porous ceramic or porous metal substrates may be used for sterilization of surfaces (e.g., hospitals and veterinary clinics), reduction of wound dressing infections, and other healthcare applications37,38 For the low temperatures used in this study, ALD-grown TiO2 is predominantly amorphous, which will have reduced photocatalytic activity compared to anatase or rutile phases.39 Luo et al.,40 Sotto et al.,41 and Liu and Li42 described the use of PES in wastewater treatment.40–42 For example, Luo et al.40 and Sotto et al.41 showed lower fouling of TiO2 nanoparticle-modified PES by organic materials (e.g., humic acids). In another study, Luo et al.43 showed that TiO2 nanoparticle-modified PES had good resistance to fouling by polyethylene glycol-5000, a hydrophobic material. Another potential application for TiO2-coated PES involves use in direct methanol fuel cells; for example, Prashanta and Park44 showed that TiO2 nanoparticle-modified PES had an improvement water/methanol selectivity that correlated with an increase in the TiO2 weight percentage.