Article pubs.acs.org/Langmuir
Generic Top-Functionalization of Patterned Antifouling Zwitterionic Polymers on Indium Tin Oxide Yan Li,† Marcel Giesbers,† Marieke Gerth,† and Han Zuilhof*,†,‡ †
Laboratory of Organic Chemistry, Wageningen University, Dreijenplein 8, 6703 HB Wageningen, The Netherlands Department of Chemical and Materials Engineering, King Abdulaziz University, Jeddah, Saudi Arabia
‡
S Supporting Information *
ABSTRACT: This paper presents a novel surface engineering approach that combines photochemical grafting and surface-initiated atom transfer radical polymerization (SI-ATRP) to attach zwitterionic polymer brushes onto indium tin oxide (ITO) substrates. The photochemically grafted hydroxyl-terminated organic layer serves as an excellent platform for initiator attachment, and the zwitterionic polymer generated via subsequent SI-ATRP exhibits very good antifouling properties. Patterned polymer coatings can be obtained when the surface with covalently attached initiator was subjected to photomasked UV-irradiation, in which the C−Br bond that is present in the initiator was broken upon exposure to UV light. A further, highly versatile top-functionalization of the zwitterionic polymer brush was achieved by a strain-promoted alkyne−azide cycloaddition, without compromising its antifouling property. The attached bioligand (here: biotin) enables the specific immobilization of target proteins in a spatially confined fashion, pointing to future applications of this approach in the design of micropatterned sensing platforms on ITO substrates.
■
INTRODUCTION Initially developed as a transparent electrode for materials science, indium tin oxide (ITO) is now a solid substrate of increasing importance in the area of biomaterials due to its favorable combination of conductivity, transparency, and biocompatibility.1−3 A variety of possible applications, as a platform for cell cultivation4 or cell-based biosensing5 and as a substrate for immunosensors for disease diagnosis,6 have been demonstrated in recent publications. The surface characteristics of the ITO substrate can significantly influence the performance of the final biodevices. For such applications, manipulation of the chemical and physical properties of the surface is highly desirable, which can on the one hand inhibit unwanted interactions at the interface (e.g., nonspecific binding of the analyte) and on the other hand enable recognition between a surface-bound probe and its target biomolecule. This has become the focus of recent research.7−9 Over past decades, poly(ethylene glycols) (PEGs) are the most commonly studied antifouling materials to reduce nonspecific protein adsorption.10,11 However, PEG is a polyether that auto-oxidizes, especially in the presence of oxygen and metal ions as are found in many biochemically relevant solutions.12 As an alternative to PEG, zwitterionic polymers have attracted increasing interest in recent years. It was found that zwitterionic polymers, such as poly(sulfobetaine) (polySB) and poly(carboxybetaine) (polyCB), do not only show excellent ultralow fouling properties at body temperature but also exhibit a remarkable stability in aqueous media.13−15 Such zwitterionic polymer brushes, especially © 2012 American Chemical Society
polyCB, have subsequently been functionalized via EDC/ NHS and other conjugation techniques to couple specific bioaffinity to generic protein repellence. While these methods have been shown to work effectively,13,16,17 such conjugation requires several steps in many cases, and easy-to-use one-step approaches to achieve this are highly welcome. Zwitterionic polymers are usually coated onto surfaces by SIATRP, a “grafting-from” technique, which requires surfacebound initiators/catalysts.18,19 Prior studies have shown that organosilane and phosphonic acids reagents can introduce reactive groups on many metal oxide surfaces.20,21 However, while these groups offer some benefits, they also have a number of limitations. For example, silanes are prone to cleavage by hydrolysis under ambient conditions resulting in desorption of the organic layers.22 The poor solubility of phosphonic acids in many organic solvents and the requirement of extensive heating to form stable covalent bonds at the interface also limit their use as surface anchoring groups.23 Recently, we have found that covalently attached organic layers on ITO surfaces can be prepared by photochemical grafting of ITO with 1-alkenes.24 This photochemical approach is particularly attractive due to its significant advantages: mild reaction conditions, low cost of operation, and the possibility of high-throughput surface patterning by using optical masks. The obtained organic coatings also proved to be highly stable during prolonged exposure to aqueous solution, even at elevated temperatures. Received: June 2, 2012 Published: August 13, 2012 12509
dx.doi.org/10.1021/la3022563 | Langmuir 2012, 28, 12509−12517
Langmuir
Article
Figure 1. Schematic overview of the formation of top-functionalized, protein-repelling zwitterionic polymer brushes on an ITO substrate. room temperature (rt) for 2 h. The surfaces were removed and subsequently thoroughly cleaned by sonication in CH2Cl2 for 1 min and dried by a stream of Ar. Surface-Initiated Polymerization. SBMAA (0.9 g, 2.5 mmol) and 2,2′-bipyridine (0.16 g, 1.0 mmol) were dissolved in a mixture of isopropanol (6 mL) and water (2 mL) in a flask. The solution was sonicated for 3 min and was degassed for 30 min by purging with Ar. CuBr (0.07 g, 0.5 mmol) was added to a separate vial under argon (in a glovebox), which was closed by a septum. Subsequently, the initiatorcoated ITO surface and the CuBr was transferred into the flask containing reactants, under continuous flushing with Ar. The polymerization was carried out at rt for 6 h under Ar. Then the sample was removed and cleaned by sonication in warm water (∼60 °C) and further with ethanol for 5 min. Finally, the samples were dried under a stream of nitrogen. Photopatterning of Modified ITO via UV Irradiation. To obtain area-selective photocleavage of the alkyl bromide bond in the surface-bound initiator, a gold mask (see Figure 5) was placed between the modified ITO substrate and a quartz slide, and this was irradiated by two low-pressure mercury lamps (254 nm, 6.0 mW/cm2, Jelight, Irvine, CA). After 30 min of UV radiation, the ITO samples were washed with dichloromethane and absolute ethanol and dried by a stream of nitrogen. Strain-Promoted Cycloaddition on Zwitterionic Polymer. Replacement of the Br residue by reactive azide moieties was achieved by immerging the zwitterionic polymer-coated ITO substrate in an aqueous azide solution (0.5 M) for 15 h in a shaker bath, removing, and rinsing with water. The subsequent Cu-free click reaction was performed by submerging the azide-modified polymer coating in a solution of BCN−biotin (10 mM, CH3CN/H2O = 4:1) and then reacted for 15 h at rt in a shaker bath. Afterward, the substrate was thoroughly rinsed with PBS buffer and water to remove the physically adsorbed reactants. XPS Characterization. X-ray photoelectron spectroscopy was performed using a JPS-9200 photoelectron spectrometer (JEOL, Japan). Spectra were collected using monochromatic Al Kα X-ray radiation at 12 kV and 20 mA with an analyzer pass energy of 50 eV for wide scan and 10 eV for narrow scan. Because of the sample charging on semiconductor samples, a low-energy electron flood gun was used for charge compensation, and peak positions were assigned by referencing the methylene component of the C 1s peak to a binding energy of 285.0 eV. Atomic area ratios were determined after a baseline correction and normalizing the peak area ratios by the corresponding atomic sensitivity factors (1.00 for C1s, 1.80 for N1s, 2.93 for O1s, 1.68 for S2p, 14.8 for Sn3d5/2, and 13.4 for In3d5/2). Auger Elemental Mapping. Auger electron spectroscopy (AES) and scanning electron microscopy (SEM) were obtained using a field emission Auger microprobe (JAMP-9500F, JEOL Tokyo). The mapping images were plotted using the ∼380 eV for peak position of nitrogen element and 370 eV for the background (Figure S4, see Supporting Information). To obtain our Auger image, the selected region to be investigated (0.45 × 0.45 mm2) was scanned in a line-by-
Based on this prior work, the present paper describes the development of a convenient and effective approach for the micropatterning of top-functionalized antifouling coatings, and, in turn, proteins, onto an ITO substrate (Figure 1). Specifically, a photochemically attached, robust organic layer was used as an initiator platform for the growth of zwitterionic polymer via the SI-ATRP technique. The generated polymer brush is subsequently reacted with azide and top-functionalized in a highly versatile and easy-to-use fashion via a strain-promoted alkyne−azide cycloaddition (SPAAC) reaction,25 to allow spatially selective biointeractions (e.g., biotin−streptavidin binding), onto generally highly antifouling ITO substrates.
■
EXPERIMENTAL SECTION
Chemicals. All chemicals, unless otherwise noted, were of analytical grade and used as received. Milli-Q water with a resistivity above 18 MΩ·cm was used for surface cleaning. 5-Hexen-1-ol and Triton X-100 were purchased from Alfa Aesar; 1-hexadecene (99%) was obtained from Fluka and further purified by passing through as short silica column; copper(I) bromide (CuBr, 99.999%), [3(methacryloylamino)propyl]dimethyl(3-sulfopropyl)ammonium hydroxide inner salt (SBMAA, 96%), 2,2′-bipyridine (bipy), i-PrOH, NaN3, H2O2 (50%), NH3−H2O (25%), and fluorescein isothiocyanate-labeled bovine serum albumin (FITC-BSA) were purchased from Sigma-Aldrich; Cy3-stravavidin was from Invitrogen, USA, and biotinlabeled bicyclo[6.1.0]non-4-yn-9-ylmethyl carbamate (BCN-biotin) was from SynAffix B.V., The Netherlands. ITO Cleaning and Activation. The ITO substrates (25 × 25 mm2, Sigma-Aldrich, 8−12 Ω/sq) were first cleaned in ultrasonic baths using a solvent series with increasing polarity: CH2Cl2, acetone, and methanol for 5 min each. Then, the substrates were treated in an oxidizing bath of NH4OH:H2O2:H2O (1:1 ratio by volume of the 25% NH3 and the 50% H2O2 solutions) at 80 °C for 30 min to generate an abundant amount of −OH groups on the surface, then washed thoroughly with Milli-Q water, and dried under a stream of N2. Photochemical Attachment of Hydroxyl-Terminated Organic Layers. Prior to the photografting, 5-hexen-1-ol was placed into a specially designed quartz flask (see Supporting Information, Figure S1). Afterward, the freshly cleaned ITO substrate was immerged into the reactant, and the system was degassed by at least three consecutive pump−Ar purge cycles. UV irradiation was carried out by two lowpressure mercury lamps (254 nm, 6.0 mW/cm2, Jelight, Irvine, CA) at a fixed distance (∼0.5 cm) from the ITO slide for 4 h. After illumination, the substrate was removed from the reactor, rinsed with hexane and acetone, sonicated in ethanol for 5 min, and finally dried with nitrogen. Attachment of ATRP Initiator. The Br-initiator was subsequently attached onto the organically modified hydroxyl-terminated ITO surfaces via a reaction with 2-bromoisobutyryl bromide (0.34 g, 1.5 mmol) in dry dichloromethane (1 mL) containing Et3N (0.2 mL) at 12510
dx.doi.org/10.1021/la3022563 | Langmuir 2012, 28, 12509−12517
Langmuir
Article
Figure 2. Formation of covalently linked organic layers on ITO and subsequent SI-ATRP reaction to yield covalently linked zwitterionic polymer brushes. line fashion with resolution of 512 × 512 pixels, typically recording the entire map required 2 h. AFM Characterization. AFM topography image was obtained with an MFP3D AFM (Asylum Research, Santa Barbara, CA). Prior to the thickness measurements, the polymer-coated surfaces were immersed in pure water for 4 h at rt to fully swell the polymer. The imaging was performed in tapping mode in air using Olympus silicon cantilevers with a stiffness of 60 N/m. The 3D image was generated by the default MFP3D software. Static Water Contact Angle Measurements. The wettability of the modified surfaces was determined by automated static water contact angle measurements with the use of Krüss DSA 100 goniometer (volume of the drop of demineralized water is 3.0 μL). The reported value is the average of at least 3 droplets with the error of less than ±2°. Characterization of Protein Adsorption on Zwitterionic Polymer Coating. The ability of the zwitterionic polymer-coated ITO surface to resist the nonspecific adsorption of FITC-BSA was evaluated by fluorescence studies according to the procedure described by Ciampi et al.26 Briefly, ITO samples (1.5 × 1.5 mm2) were exposed for 1 h to protein solution (0.1% FITC-BSA, PBS buffer, pH 7.2). The samples were subsequently rinsed with Milli-Q water and then incubated with the elution buffer (0.3 M NaCl, 20 mM Na2HPO4, 2 mM EDTA, 1% Triton X-100, 1% β-mercaptoethanol) on an shaker table with slight shaking (70 rpm) overnight. Subsequently, the eluted FITC−BSA solution from the ITO surface was measured on an Edinburgh Instruments FLS900 fluorescence spectrometer. Samples were prepared in triplicate, and for each sample, a minimum of three repeated fluorescence measurements were obtained. In all cases, 494 nm was used as the excitation wavelength, and the emission spectrum was recorded from 505 to 700 nm. The integration of the emission spectrum was compared and normalized to that observed for the elution from a 1-hexadecene-grafted ITO surface. Adsorption of Cy3-Streptavidin and FITC-BSA on Patterned Biotinylated Surface. The biotinylated surface was incubated in a solution of Cy3-streptavidin (0.2 mg/mL) in PBS buffer for 1 h at rt. The substrate was subsequently rinsed 3 × in a 0.5% solution of Triton X-100 in PBS buffer to remove the nonspecifically bound streptavidin from the surface, followed by an additional rinse in DI water, and dried under a stream of nitrogen. Fluorescent images were measured with a confocal laser scanning microscope (Zeiss LSM 510 Meta), and the Cy3 dye was excited with at 533 nm. The adsorption of FITC-BSA on the biotinylated surface was similar to the protocols described for Cy3streptavidin, except rinsing was performed by using PBS buffer (pH = 7.2) without any detergent. As for the characterization of confocal laser scanning microscope, the FITC dye was excited with an argon ion laser at 488 nm.
alized with zwitterionic polymer brushes in a three-step procedure (Figure 2). First, hydroxyl-terminated modified ITO surfaces were obtained via the photochemical grafting of NH4OH/H2O2-treated ITO with 5-hexen-1-ol.24,27 After the UV irradiation for 4 h, according to the attenuation of the electrons emitted from the bulk ITO,24 the thickness of the hydroxyl-terminated organic layer is estimated to be 3.9 nm, which is considerably higher than the calculated length of single 5-hexen-1-ol molecule (∼0.75 nm determined by Chem3D), suggesting multilayer formation and the occurrence of secondary grafting. Similar multilayer formation has been observed before on ITO for substituted alkenes20 and has also been reported for the photografting of alkenes onto other −OH-terminated surfaces, such as glass,28 (acid-treated) silicon carbide,29 and TiO2.27 The narrow-scan XPS spectrum of C 1s region (Figure 3a) displays two main peaks at 285.0 and 286.3
Figure 3. High-resolution C 1s scan of (a) 5-hexen-1-ol grafted ITO surface and (b) after the attachment of a bromo-initiator.
eV, which correspond to the carbon atoms bound to oxygen (C−O) and carbon (C−C), respectively. The intensity ratio of the C−C/C−O is 5.1. Though previous study indicates covalent attachment of alkene on the ITO surface was achieved by one of the alkene carbons bound to the ITO substrate via a C−O−In (Sn) linkage, the signal corresponding to the C−O at the bottom interface may suffer considerable attenuation due to the presence of the thick organic layer, thus contributing a strongly diminished amount of C−O intensity. Therefore, the
■
RESULTS AND DISCUSSION Formation and Characterization of Zwitterionic Polymer-Coated ITO Surface. ITO surfaces were function12511
dx.doi.org/10.1021/la3022563 | Langmuir 2012, 28, 12509−12517
Langmuir
Article
mechanism of the grafting and the chemical structure of the organic layer are probably well consistent with the previously reported photografting on TiO2.27 This will then yield a repeating −(CH2)5−CROH− unit growing from the surface in which R = H or alkyl, depending on whether additional radical reactions have taken place on that alcohol carbon atom or not; ideally, this would correspond with the C−C/C−O ratio of 5.0, in good agreement with our experimental result of 5.1. We also find a small peak at 287.8 eV, likely corresponding to ketone or aldehyde groups. The presence of a small amount of aldehyde is probably due to H-abstraction from the terminal alcohol moieties during the termination of the grafting27 and was further confirmed by the XPS spectrum of the sample stored in air for 2 weeks (see Supporting Information, Figure S2). The accessibility of the surface-bound organic −OH groups can be easily probed by secondary modification. To this aim, a bromine-containing ATRP initiator was attached to the alcoholmodified ITO surfaces via a reaction with 2-bromoisobutyryl bromide. The static water contact angle of the resultant ITO substrate was increased from 69° to 75° after the 2 h reaction, suggesting the formation of more hydrophobic ester group. The presence of the bromoinitiator was also evident from the XPS characterization (Figure 3b). A small C 1s peak at 289.3 eV is attributed to the ester carbon, and a shoulder peak at 286.4 eV is due to the carbon contribution from both C−O and C−Br (may also contain aldehyde residue). The distinct Br 3d peak at about 70 eV (Figure S3, see Supporting Information) after the esterification further confirms the presence of surface-tethered bromoinitiators. Tentatively, we estimate from the thickness (3.9 nm) and the length of a repeating −(CH2)5−CROH− unit (0.5−0.7 nm, depending on the tilt angle) that there are 6−8 alcohol moieties available for further functionalization per chain. From the C/Br ratio obtained after attachment of the initiator (16.3), we estimate the number of bromoinitiator moieties to be 3−4 per chain, implying that clearly sufficient initiator moieties were attached for the SI-ATRP reaction. It is worth noting thatdespite a previous report of successful grafting of benzoyl chloride derivatives on ITO30the direct attachment of bromoinitiators on the native −OH groups of bare ITO is rather irreproducible in our hands and typically generates a much lower density of initiator on the surface. In contrast, the grafted 5-hexenol-derived organic layer gives a high density of alcohol groups at some distance from the bare ITO surface, which may facilitate the esterification with 2bromoisobutyryl bromide.31 PolySBMAA was then grafted from the bromoinitiatorcoated ITO surface via an ATRP reaction with SBMAA. After the polymerization, the water contact angle of the modified surface was below the detection limit of the equipment (
