Formation and Antifouling Properties of Amphiphilic Coatings onPolypropylene FibersKiran K. Goli,† Orlando J. Rojas,‡,∥ and Jan Genzer*,§

†Department of Materials Science & Engineering, ‡Department of Forest Biomaterials, and §Department of Chemical & BimolecularEngineering, North Carolina State University Raleigh, North Carolina, United States∥Department of Forest Products Technology, Aalto University, FI-00076 Aalto, Espoo, Finland

*S Supporting Information

ABSTRACT: We describe the formation of amphiphilic polymeric assemblies via a three-step functionalization process appliedto polypropylene (PP) nonwovens and to reference hydrophobic self-assembled n-octadecyltrichlorosilane (ODTS) monolayersurfaces. In the first step, denatured proteins (lysozyme or fibrinogen) are adsorbed onto the hydrophobic PP or the ODTSsurfaces, followed by cross-linking with glutaraldehyde in the presence of sodium borohydride (NaBH4). The hydroxyl and aminefunctional groups of the proteins permit the attachment of initiator molecules, from which poly (2-hydroxyethyl methacrylate)(PHEMA) polymer grafts are grown directly through “grafting from” atom transfer radical polymerization. The terminalhydroxyls of HEMA’s pendent groups are modified with fluorinating moieties of different chain lengths, resulting in amphiphilicbrushes. A palette of analytical tools, including ellipsometry, contact angle goniometry, Fourier transform infrared spectroscopyin the attenuated total reflection mode, and X-ray photoelectron spectroscopy is employed to determine the changes inphysicochemical properties of the functionalized surfaces after each modification step. Antifouling properties of the resultantamphiphilic coatings on PP are analyzed by following the adsorption of fluorescein isothiocyanate-labeled bovine serum albuminas a model fouling protein. Our results suggest that amphiphilic coatings suppress significantly adsorption of proteins ascompared with PP fibers or PP surfaces coated with PHEMA brushes. The type of fluorinated chain grafted to PHEMA allowsmodulation of the surface composition of the topmost layer of the amphiphilic coating and its antifouling capability.

■ INTRODUCTIONTailoring surfaces with biocompatible and nonfouling charac-teristics represents one of the outstanding challenges in today’ssurface science.1 Most material surfaces in use are prone tononspecific adsorption of biomolecules; a widespread polymersuch as polypropylene (PP) is not exempt from suchphenomenon. Owing to its good mechanical strength, lowdensity, chemical resistance, thermal stability, and low cost, PPis widely used in textile and filtration industries, amongothers.2−6 Unfortunately, inherent hydrophobicity and poorbiocompatibility of PP limit its use in biological applicationsincluding biofiltration and bioseparation.2,5,7 In general, whensynthetic surfaces come into contact with biological milieu,adsorption of platelets and proteins occurs nearly instanta-neously leading to undesired effects, such as thrombus andplaque formation on medical implants, teeth, and dentalrestoratives, etc. Likewise, fouling of textile surfaces, contactlenses, kidney dialysis membranes or bioprocessing equipment

is ubiquitous. This phenomenon compromises the efficiency ofthe product and its functions.4,8−13 In particular, duringbiofiltration, proteins from blood and body fluids interactstrongly with hydrophobic surfaces causing pore blocking andbiofouling of PP supports. In turn, fluid flux is reduced whichleads to high trans-membrane pressures, increased operationalcost, and eventually device deterioration.5,14,15 Hence, control-ling bioadhesion of hydrophobic materials, including PP, toenhance biocompatibility is an important prerequisite in theirdeployment.7,16,17 To this end, numerous studies have beencarried out to tailor the surface properties of materials afterchemical and physical modifications.2,5 It has been establishedthat grafting surfaces with macromolecules of tailored chemical

Received: August 2, 2012Revised: September 25, 2012Published: September 27, 2012

Article

pubs.acs.org/Biomac

© 2012 American Chemical Society 3769 dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−3779

composition, length, and surface density is paramount inachieving biocompatilibity.14,18,19 The review by Edmondson etal.20 reported on the formation of polymer brushes on silicasurfaces via surface initiated polymerizations using variousroutes including free radical, cationic, anionic, and atomtransfer radical polymerizations (ATRPs). In addition, polymergrafting provides numerous opportunities to introduce newfunctional groups that can further enhance the functionality ofcoating layers.9

A generalized approach to prevent fouling of surfaces isrestricting initial protein adsorption, which necessitates theinstallation of protein-repellent functionalities. Surfaces graftedwith ethylene glycol, that is, oligo- or poly-ethylene glycol(OEG or PEG) are used nearly exclusively as benchmarks forprotein-repellent coatings.8,10,21 However, temperature insta-bility as well as sensitivity toward oxidation or auto-oxidationcatalyzed by transition metals limit the widespread applicationof such coatings.8,22 In addition, OEG or PEG chains containonly a limited number of accessible hydroxyl groups, whichrestrict further chemical modifications. These limitations havetriggered the development of novel protein-repellent materials.9

In particular, 2-hydroxyethyl methacrylate, sulfobetaine meth-acrylate, carboxybetaine methacrylate, 2-methacryloyloxy-ethylphosphorylcholine, poly[N-(2-hydroxypropyl) methacryla-mide], 3-sulfopropyl-methacrylate, and 2-(tert-butylamino)-ethyl methacrylate have been tested and proposed as efficientprotein-repellent coating polymers.4,9,23 In addition, severalresearch groups have demonstrated that effective antifoulingcoatings can be fabricated from amphiphilic polymeric materialsthat feature both PEO and fluorinated groups.24,25 Whereas themechanism behind this effectiveness is yet to be understoodcompletely, the general consensus is that amphiphilicity or thepresence of the two extreme chemistries (highly hydrophilic,i.e., PEG, and highly hydrophobic, i.e., fluorinated groups) isresponsible for the unusual enhanced performance.Numerous surface grafting methods have been employed to

generate polymeric coatings. Among the most efficient ones areanchoring ex situ synthesized polymers to the surface of interest(so-called “graf ting to” method) or generating polymeric graftsdirectly from surface-bound polymerization initiators (so-called“graf ting f rom” method). In both cases, the grafts are anchoredto the surface by either naturally present functional moieties onthe substrate or those that have been generated by surfacepreactivation.18 The preactivation methods typically involveirradiation by UV,26,27 electron beam,28 and ozone.29 Somesurfaces can also be modified by using flame,30 plasma,4 andcorona treatments.31 While relatively efficient in endowing thesurfaces with reactive groups, such “harsh” physical treatmentsmight compromise the mechanical properties of the substratesor supports.5 This is particularly problematic in the case ofrelatively thin fiber-based materials. To overcome thislimitation, we have recently proposed a novel route forachieving surface functionalization of hydrophobic materials(flat supports and polyolefin-based textile fibers) primed withdenatured proteins as a primary step for further modificationwith functional groups.32 Other researchers have developedalternative methods to immobilize proteins on hydrophobicsurfaces. For instance, Brynda et al.33 reported on the formationof protein multilayers consisting of human serum albumin(HSA) and heparin in their native states on polyethylene. Thefirst HSA layer was bound to the surface through hydrophobicinteractions, whereas successive layers were held by electro-static interactions. Protein multilayers were stabilized by cross-

linking the amino groups of HSA and heparin withglutaraldehyde (GA). The same group has also demonstratedthe preparation of multilayers consisting of native fibrinogen(FIB) on polyethylene.34 Alternating layers of positivelycharged FIB and polyanions of a sodium salt of dextran sulfate(DS) were consecutively adsorbed by controlling the solutionpH, followed by cross-linking. The polyanion layers boundbetween cross-linked protein layers through electrostaticinteractions were washed out in PBS buffer.34

It should be stressed that attachment of synthetic polymersto proteins has been a topic of great recent interest. Few studieshave reported on the synthesis of protein−polymer conjugatesthrough the chemical reaction of reactive side chains of proteinresidues, including lysine or cysteine, with active sites of thepolymers. For example, Gao et al.35 reported on ATRP ofoligo(ethylene glycol) methyl ether methacrylate (OEGMA)from the C-terminus of green fluorescent protein (GFP), andPeeler et al.36 demonstrated the incorporation of initiators atspecific sites of GFP to polymerize oligo(ethylene oxide)monomethyl ether methacrylate monomer via ATRP. Finally,Depp et al.37 reported on the ATRP modification ofchymotrypsin with various polymers including poly(2-meth-acryloyloxyethyl phosphorylcholine), poly(N-2-hydroxypropyl-methacrylamide), and poly(monomethoxy-polyethyleneglycol-methacrylate). In all of these cases, polymer chains were graftedfrom protein terminus in the solution. Compared with theprevious work cited before, the methods proposed in this studyare unique not only because of the functionalization used(including the installation of amphiphilic properties on thesubstrate) but also because of the fact that the priming of thesurface involved adsorption of proteins after denaturation bymeans of urea or heat treatment. In fact, compared with nativeproteins, denatured proteins expose their hydrophobic domainsto the outer environment, thereby increasing the numberdensity of sites that can bind to hydrophobic substrates.38 Thereaction between amine groups of protein and aldehyde groupsof GA generates Schiff bases (cross-links) that are unstable andcan be cleaved in aqueous solutions.39 Therefore, we stabilizedthe cross-links by adding sodium borohydride (NaBH4) to theGA solution, which reduced the double bonds of the Schiff baseto secondary amine bonds.39 We further utilized functionalgroups in proteins to attach initiators, followed by subsequentgrowth of polymer brushes. Finally, another distinctive featureof the present work is that robust, highly resistant coating layersare demonstrated, a feature that is not likely to hold in otherreported systems, where layer stability under harsh solventconditions is not of concern.In summary, this article reports on the decoration of the

outer surfaces of protein coatings with polymerization initiators,which serve as active sites for “graf ting f rom” polymerization ofprotein-repellent poly(2-hydroxyethyl methacrylate)(PHEMA). Amphiphilic grafts are then generated byfluorinating PHEMA chains with selected fluorinating agents.The motivation of the postpolymerization via fluorinationreaction comes from our recent study, demonstrating that acoating with a single chemistry, such as ethylene glycol, is notsufficient to prevent the adsorption of a broad spectrum ofproteins and bio-organisms.40 Moreover, silicon substratescarrying amphiphilic coatings based on ethylene glycol andfluorinated chemistries act in concert to minimize proteinadsorption. For instance, they are effective in preventingadhesion of biospecies such as Ulva and Navicula, whichpossess affinity for hydrophobic and hydrophilic supports,

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793770

respectively.40 Concurrently with the functional PP fibers, weapply the same methodologies on reference flat surfacesconsisting of hydrophobized silicon wafers, which wereobtained after self-assembling of n-octadecyltrichlorosilane(ODTS). The latter experiments enable facile characterizationof the individual fabrication steps using sensitive analyticalmethods. Finally, we employ fluorescently tagged bovine serumalbumin probes to demonstrate that amphiphilic PHEMA-fluorinated coating layers installed on the preadsorbed proteinlayer endows PP fibers with effective antifouling characteristics.

■ MATERIALS AND METHODSDeionized water (DIW) (resistivity >16 MΩ cm) was produced usingthe Millipore water purification system. Silicon wafers (orientation[100]) were supplied by Silicon Valley Microelectronics. Phosphate-buffered saline (PBS) solution, HPLC-grade methanol, and toluenewere purchased from Fisher Scientific and used as received. ODTS wasobtained from Gelest (Morrisville, PA). Lysozyme (from chicken eggwhite, Mn = 14.3 kDa, isoelectric pH or pI = 11.3), fibrinogen (fromhuman plasma, Mn = 340 kDa, pI = 5.5), fluorescein isothiocyanate-labeled bovine serum albumin (FITC-BSA, Mn = 66 kDa), pyridine,CuCl (99.99%), CuCl2 (99.99%), 2,2′-bipyridine (99%), 2-bromopro-pionylbromide (2-BPB), 2-hydroxyethyl methacrylate (HEMA, 98%),trifluoroacetic anhydride (C4O3F6, F1), heptafluorobutyryl chloride(C3F7COCl, F3), pentadecafluoro-octanoyl chloride (C7F15COCl,F7), methanol, dichloromethane, GA, and NaBH4, were all purchasedfrom Sigma-Aldrich and were used as received. Tetrahydrofuran(THF) and triethylamine were purchased from Sigma-Aldrich andFluka and distilled from Na/benzophenone and calcium hydride,respectively.Preparation of Protein Solution and Deposition on Surfaces.

Urea (8 M concentration) was added to PBS buffer solutions andfiltered through a 0.2 μm filter, followed by the addition of protein toobtain lysozyme (LYS) and fibrinogen (FIB) concentrations of 1 and0.1 mg/mL, respectively. The solutions were allowed to stand at roomtemperature for 6 h to solubilize the proteins completely. Proteinadsorption was carried out at the isoelectric point (pI) of each protein(i.e., pH 11.3 or 5.5 for LYS and FIB, respectively); the solution pHwas adjusted by adding either 0.1 N HCl or NaOH dropwise. Sodiumazide (NaN3, 0.2%) was added to the buffer solution to prevent thegrowth of bacteria during the adsorption process.Melt-blown PP nonwoven sheets were obtained from The

Nonwovens Institute (North Carolina State University) and werecleaned with isopropanol prior to use. The average fiber diameter ofPP nonwoven melt-blown media was 5.15 ± 2.17 μm. Self-assembledmonolayers (SAMs) of ODTS on silicon wafers were preparedaccording to the procedure developed by Liu et al.41 The ODTS SAMswere employed to mimic the physicochemical characteristics of PP in aflat geometry. PP nonwoven and ODTS wafers were immersed inprotein solutions for 12 h at room temperature, then removed, rinsedwith PBS and DIW, and blow-dried with nitrogen gas. The resultantflat ODTS substrates were characterized by ellipsometry and contactangle measurements to determine the thickness and wettability,respectively, before and after protein adsorption. Cross-linking of theresultant protein layers on ODTS substrates and PP nonwovens wascarried out using the experimental conditions adopted from ourprevious work.32 Cross-linking of proteins in solutions was assessed byusing SDS-PAGE. (See the detailed information in the SupportingInformation.)Formation of Amphiphilic Polymer Brush Layers. The

accessible hydroxyl and amine groups of the protein’s hydrophilicfragments present over the free surface of the cross-linked proteinlayer were used to attach 2-BPB by following the conditions describedpreviously.42 ATRP of HEMA following the “grafting from” schemefrom the surface-bound 2-BPB initiator centers was carried out at 25°C for 9 h according to the method described by Arifuzzaman et al.40

The samples were sonicated vigorously for 5 min to remove anytrapped Cu catalysts. No traces of Cu were detected by XPS. The

PHEMA brushes were functionalized chemically by employingpostpolymerization modification (PPM) reactions to obtain P-(HEMA-co-fHEMA) amphiphilic grafts. Other research groups havein the past utilized the PPM route to tailor the chemical compositionof macromolecular grafts.43 Various fluorocarbons were used,including, trifluoroacetic anhydride (C4O3F6, F1), heptafluorobutyrylchloride (C3F7COCl, F3), and pentadecafluoro-octanoyl chloride(C7F15COCl, F7). The experimental conditions for the PPM wereadopted from our previous work.40 Because the lengths of the PHEMAbrushes are much smaller than the radius of curvature of the fibersurfaces (diameter = 5.15 ± 2.17 μm), we assumed that thephysicochemical properties of the brushes and its fluorinatedcounterparts were comparable to those grafted on flat substrates;that is, curvature effects were assumed to be negligible.

Ellipsometry. The thickness of bare ODTS SAMs before and afterthe formation of protein coated layers was determined using variableangle spectroscopic ellipsometry (VASE, J.A. Woollam). Specifically,ellipsometry was used to determine the thickness of the ODTS SAMand PHEMA brushes before and after the fluorination reaction toestablish the “degree of fluorination”. For each coating, the thicknesseswere measured in the dry state at three or more spots on eachspecimen and then averaged. Ellipsometric data were collected at anincidence angle of 75° for SAMs and protein coating layers and at anangle of 70° for PHEMA and amphiphilic polymers, using wavelengthsranging from 400 to 1100 nm in 10 nm increments. The thicknesseswere estimated using a Cauchy layer model assuming the index ofrefraction to be 1.544 for ODTS SAMs and 1.5440 for protein layers ontop of the ODTS SAMs. The thickness of PHEMA and P(HEMA-co-fHEMA) on protein coating layer was calculated using the methoddeveloped by Arifuzzaman et al.40

Contact Angle Measurements. Contact angles (θ) weremeasured with Rame−Hart contact angle goniometer (model 100-00) using DIW as the probing liquid at room temperature. Staticcontact angles were recorded by releasing an 8 μL droplet of DIW onthe surface. The advancing and receding contact angles, θA and θR,respectively, were measured by imaging the droplet at the tip of thesyringe and adding (θA) or withdrawing (θR) 4 μL of DIW. Contactangle measurements were carried out on at least three differentpositions over the surface and then averaged. In addition towettabilities, contact angle hysteresis (CAH), taken as θA − θR, wasevaluated to provide information about chemical heterogeneity of thesubstrates. A CAH of <10° indicated that the substrate coating wasrelatively uniform.

X-ray Photoelectron Spectroscopy. A Kratos Analytical AXISULTRA DLD X-ray photoelectron spectroscopy (XPS) instrumentemploying monochromated Al Kα radiation with charge neutralizationwas utilized to determine the chemical composition of fluorinatedPHEMA random copolymer grafts, that is, P(HEMA-co-fHEMA).Survey and high-resolution spectra were collected with pass energies of80 and 20 eV, respectively, using both electrostatic and magneticlenses for single angle spectra collection. The depth-dependentdistribution of the fHEMA units in P(HEMA-co-fHEMA) coatingswas assessed by angle-resolved XPS (AR-XPS), which was conductedat 90 and 30° takeoff angles using electrostatic lens to achieveenhanced angular resolution. The takeoff angle, defined as the anglebetween the sample and the detector probes, allowed penetrationdepths of ∼90 and ∼45 Å at 90 and 30°, respectively. The resultantdata were analyzed using the CasaXPS software package.

Fourier Transform Infrared Spectroscopy. Fourier transforminfrared (FTIR) spectroscopy spectra for PP nonwoven surfacesbefore and after functionalization were recorded using a Bio-Rad-Digilab FTS-3000 FTIR spectrometer equipped with crystalline ZnSein the ATR mode with continuous nitrogen purging. During thespectral collection, the nonwoven was pressed against the crystal undera uniform pressure of ∼700 psi using the micrometer pressure clamp.The spectra reported represent an average over 5 accumulations of 64scans with a resolution of 4 cm−1. The data were analyzed using theBio-Rad Win IR Pro software.

Atomic Force Microscopy. The surface topography of ODTSsubstrates before and after functionalization was examined using

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793771

Asylum Research MFP3D atomic force microscope (AFM) in air. TheAFM instrument was operated in tapping mode in AC mode using anAl-backside-coated Si probe with a force constant of ∼5 N/m and aresonance frequency in the range of 120−180 kHz. During AFMimaging, care was taken to keep the tip in the repulsive mode. Theroot-mean-square (RMS) surface roughness was calculated from theheight images using the MFP-3D software.Confocal Microscopy. To visualize the antifouling properties of

P(HEMA-co-fHEMA) coatings on PP nonwovens, we incubated thespecimens with 0.1 mg/mL FITC-BSA in PBS buffer at its pI of 4.9.After 12 h, the substrates were removed from the protein solution,washed thoroughly with PBS buffer, followed by rinsing with DIW anddried in a stream of nitrogen gas. Images were taken using a laserscanning confocal microscope (Zeiss, LSM710, magnification 20×).All samples were exposed to an excitation source of argon laser at 488nm, whereas the emission was collected in the range of 492−632 nm.

■ RESULTS AND DISCUSSIONAdsorption of Denatured Proteins onto ODTS and PP

Surfaces. Figure 1 depicts schematically the formation offunctional coatings on PP fibers. Similar procedures wereapplied to flat ODTS reference surfaces. Detailed description ofthe experimental procedures leading to such coatings isprovided in the Materials and Methods. LYS (Mn = 14.3kDa, pI = 11.345) and FIB (Mn = 340 kDa, pI = 5.545) werechosen to prime the surface because of their different molecularweights and isoelectric points. The layer comprising adsorbeddenatured proteins was characterized by ellipsometry andcontact angle measurements. The dry thicknesses (H) of LYSand FIB layers over the ODTS substrate were 22.7 and 31.5 Å,which corresponded to protein coverages of 0.79 and 0.38,respectively. The coverage is defined here as the mass fractionof protein compared with that for a perfect monolayer (Ho),that is, coverage = H/Ho with Ho = [Mo/(ρNA)]

1/3; Mo is themolecular mass or molecular weight of the protein, ρ is thedensity of the protein (assumed 1 g/cm3), and NA is Avogadro’snumber. The calculated thicknesses of the FIB and LYS layersare 82.8 and 28.8 Å, respectively. The sizes of native, globularLYS and FIB proteins are 30 × 30 × 45 Å3 and 60 × 60 × 450Å3, respectively.1 The thickness of the protein coating layeradsorbed on hydrophobic ODTS is less than that for end-on orside-on orientations of native proteins. Therefore, it may beconcluded that the denaturation process leads to the rupture(disruption/unfolding) of protein’s native structure.11,46 In fact,the thickness of denatured/unfolded protein coating formed is

found to be equivalent to the length of a few amino acid sidechains that form loose random structures on the surface withhydrophobic amino acids interacting with the hydrophobicsurface and the hydrophilic residues dangling into the outermedium.32 The lower coverage of FIB relative to LYS cannot beexplained entirely by the difference in deposition solutionconcentrations.32 Hence, the difference may also be attributedto the complex elongated structure of FIB, which forms largeaggregates that prevent close packing of protein molecules.32 Incontrast, LYS is smaller than FIB and can pack more readily onthe surface.As reported in our previous work, protein coating stabilities

were improved significantly by introducing cross-links amongthe adsorbed protein molecules using well-known cross-linkingagent, GA, in the presence of NaBH4.

32 The cross-linkingreaction resulted in an increase in protein coating thickness of 3± 1 Å due to the incorporation of GA resulting from thereaction of hydroxyl and amine groups of protein with carbonylgroups of GA. The presence of a protein layer was furtherconfirmed by contact angle measurements. The static contactangle measured by DIW on the bare ODTS substrate was 108± 2°. After adsorption of FIB (or LYS), the contact angledecreased to 59 ± 3°, suggesting that hydrophilic amino acidswere exposed to the outer environment, whereas hydrophobicamino acids resided primarily at the ODTS surface. Waterdroplets placed on top of unmodified PP nonwoven sheetsrolled away, whereas the protein-modified PP nonwovenexhibited improved wettability; the wettability of the fibersindicated the attachment of protein layers to PP that rendersthe surface hydrophilic (see movies in the SupportingInformation). Cross-linking with GA did not affect thewettability of the denatured protein layer, as deduced fromthe contact angle measurements.

Formation of Amphiphilic Polymeric Brushes. Poly-merization initiators, 2-BPB, were grafted to the cross-linkedprotein coatings (XPS data available in the SupportingInformation) that served as reaction centers for the ATRP ofHEMA. The experimental details pertaining to the grafting andpolymerization of HEMA are described in the Materials andMethods section. The resulting PHEMA macromolecular graftsgrown from the initiators attached to LYS and FIB proteincoatings on flat ODTS SAMs had thickness of 63.0 ± 1.5 and63.5 ± 1.5 nm, respectively. Contact angle measurements

Figure 1. Schematic illustration depicting the steps leading to the formation of amphiphilic fiber mats. A polypropylene (PP) nonwoven sheet isexposed to the solution of denatured protein and subsequently cross-linked with glutaraldehyde (GA) and NaBH4. After depositing 2-bromopropinoyl bromide (2-BPB), poly(2-hydroxyethyl methacrylate) (PHEMA) brushes were formed by ATRP of 2-hydroxyethyl methacrylate(HEMA). Subsequent postpolymerization modification (PPM) with fluorinated agents resulted in P(HEMA-co-fHEMA) amphiphilic randomcopolymer grafts. Similar surface modification sequence was applied to ODTS surfaces.

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793772

performed after the polymerization revealed that the wettabilitywas reduced to 48 ± 2°, confirming the presence of addedhydrophilic moieties. Water droplets also wetted PP nonwovensheets grafted with PHEMA due to the presence of polarhydroxyl groups on the surface. (See the movies in theSupporting Information.) We have not determined directly themolecular weight of PHEMA brushes on the surfaces. Arigorous method to do so would entail degrafting the chainsfrom the supports and measuring their molecular weight by sizeexclusion chromatography. We have noted, however, thatPHEMA brushes grown under identical conditions from siliconwafers (albeit from a different initiator, i.e., 11-(2-bromo-2-methyl) propionyloxy-undecyl trichlorosilane (BMPUS)) ex-hibited thicknesses similar to those reported herein. From ourprevious work,47 we know that the molecular weight of thebrush (Mn) can be related to the dry thickness (H) via Mn =1200H, where Mn is in Da and h is in nm, for brushes havinggrafting density of 0.45 nm−2. One can thus use thisapproximate relationship to convert the dry brush thickness(∼63 nm) into molecular weight (∼75.6 kDa). An importantattribute of PHEMA chains is that the hydroxyl terminus ineach HEMA monomer can be further modified chemically.6 Inour work, we have derivatized PHEMA chains using PPMprotocols with various fluorination agents, including, trifluoro-acetic anhydride (C4O3F6, F1), heptafluorobutyryl chloride(C3F7COCl, F3), and pentadecafluoro-octanoyl chloride(C7F15COCl, F7), to produce random copolymers of P-(HEMA-co-fHEMA), where fHEMA denotes the HEMAsegment fluorinated with the respective fluorination agent (cf.Figure 2).We employed Fourier-transform infrared spectroscopy

measurements in the attenuated total reflectance mode

(FTIR-ATR) to verify the chemical composition of the coatingsafter each modification step. Figure 3 presents the IR spectrafor PP nonwoven sheets following their modifications with FIBand amphiphilic polymer brushes. Table 1 summarizes thecharacteristic IR vibrations found in PP and the functionalizedPP fibers. The presence of FIB on the PP surface is verified by

Figure 2. Schematic illustration of chemical modification of PHEMA brushes with trifluoroacetic anhydride (F1), heptafluorobutyryl chloride (F3),and pentadecafluoro-octanoyl chloride (F7) leading to the formation of amphiphilic random copolymers of P(HEMA-co-fHEMA), that is, PHEMA-F1, PHEMA-F3, and PHEMA-F7, respectively.

Figure 3. FTIR spectra collected in transmission mode (from bottomto top) on PP nonwoven sheet (black), PP-sheet coated withdenatured fibrinogen (FIB) layer (dark magenta), PP-FIB fiber withPHEMA brushes (orange), and PP-FIB-PHEMA modified withvarious fluorinated agents (green), that is, F1, F3, or F7, as describedin the text. (See the Supporting Information for FTIR data of LYS.)

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793773

the appearance of peaks located between 1700 and 1550 cm−1

assigned to amide I (CO stretching) and amide II bands(C−N stretching and N−H bending) and between 3420 and3250 cm−1 that originate from amide A (N−H stretching) andO−H stretching. The observed bands are featureless and broad,due to the overlap of various underlying component bands ofproteins.47−49 In PHEMA-grafted PP nonwoven mats, a newpeak appears at 1720 cm−1 that corresponds to the COstretching vibration of the ester group present in PHEMA. Inaddition, peaks at 1250 and 1080 cm−1 are attributed to C−Ostretching and O−H deformation of the C−O−H groups,respectively. The peak at 1150 cm−1 also arises from the C−O−C stretching of the PHEMA. The significant increase in theintensity and broadness of the band at 3500−3100 cm−1 isattributed to the characteristic −OH stretching of PHEMA,which verifies the grafting of PHEMA to PP surface. The

disappearance of hydroxyl stretches at 3500−3100 cm−1 inspecimens containing PHEMA-F1, PHEMA-F3, and PHEMA-F7 attached to the FIB-modified PP nonwovens is associatedwith the attachment of F1, F3, and F7 to PHEMA. Thederivatization of hydroxyl groups of PHEMA with fluorinatedgroups is further evidenced by the appearance of a characteristicstretch at 1800 cm−1 corresponding to formation of thecarbonyl group CH−COO−CF (fluorinated ester) and thecharacteristic −CF2- stretches at 1214 and 1150 cm−1.40,42,50

Fluorination of PHEMA brushes leads to increase in the drythickness of the grafted layer built on ODTS flat supports withprimer protein layer. A nonuniform PPM reaction may affectthe surface topography; therefore, AFM was used to image thesurface of samples before and after derivatization withfluorinating agents. Figure 4 presents typical AFM scans ofLYS-grafted PHEMA (top row) and FIB-grafted PHEMA(bottom row) before and after coupling reactions. ThePHEMA surface appears relatively smooth on both proteinsupports, revealing that ATRP of HEMA produces uniformlydistributed PHEMA grafts. Derivatization of PHEMA withfluorinated agents results in an increased surface roughness. Inaddition, the surfaces become rougher with increasing size ofthe fluorinated agent. Whereas roughness may affect thecharacterization of surfaces using methods that are suited forflat samples, we put the effect of roughness aside temporarilyand return to it after presenting the results of ellipsometry,angle-resolved XPS (AR-XPS), and contact angle measure-ments.The top portion of Figure 5 summarizes the thickness data of

PHEMA-based films prepared on LYS- and FIB- basedcoatings, on ODTS SAMs before and after fluorination withvarious agents. The ellipsometry data indicate that the absolutethickness after fluorination increases with increasing size of thefluorinating agent. The thickness increase due to fluorination

Table 1. IR Vibrations in PP and Functionalized PP FiberMaterials

wavenumber(cm−1) group/functionality class compound

2990−2850 −CH3 and −CH2− in PP PP1380−13701475−14501465−14403420−3250 −OH in alcohol and NH stretch

(amide A)FIB, PHEMA

1700−1550 CO and −NH2 in primary andsecondary amides

FIB

1740−1720 CO stretch in aldehydes, esters PHEMA, FIB1240−1070 C−O−C in ethers PHEMA1786 CO in CH−COO-CF F1, F3 and F71680 CO in CH−COO-CH F1, F3 and F7,

PHEMA1300−1000 C−F in fluoro compounds F1, F3 and F7

Figure 4. Tapping mode atomic force microscope (AFM) topography images taken from ODTS after functionalization with PHEMA and P(HEMA-co-fHEMA) with three different mesogens in the dry state. The PHEMA samples were fabricated utilizing “grafting from” polymerization fromdenatured lysozyme (top row) and fibrinogen (bottom row) coating on ODTS SAMs using methodological steps described in the text.

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793774

was used to obtain information about the amount of fHEMA inthe P(HEMA-co-fHEMA) copolymer by using a simple modeldeveloped by Arifuzzaman et al. (See detailed model in theSupporting Information.)40 The resulting chemical composi-tions in the P(HEMA-co-fHEMA) copolymers are plotted inthe bottom portion of Figure 5. The data in Figure 5 reveal thatthe amount of F3 inside the PHEMA brush is highest, whereasthat of F7 is the lowest. Several factors can contribute to theobserved size dependence, including, (i) larger steric hindranceimposed on the F7 molecules relative to F3 and (ii) highertendency of F7 to cluster due to strong interactions leading tolarger complexes that may not penetrate deep inside the brush.We attribute the lower loading of F1 inside the brush relative toF3 due to the different reaction conditions employed incoupling F1 relative to attaching F3 or F7 to PHEMA (seeMaterials and Methods section).Although ellipsometry offers insight into the overall loading

of fluorinated agents inside the brush, it does not provideinformation about their spatial distribution throughout thePHEMA grafts. Contact angle and AR-XPS measurements canreveal key information on the distribution of fluorinated agentat the surface and in the subsurface region of the P(HEMA-co-fHEMA) brush. Therefore, to investigate the spatial distribu-tion of fHEMA in P(HEMA-co-fHEMA) developed on LYS-coated ODTS, we employed AR-XPS at two different takeoffangles, 90 and 30°, whose probing depths are ∼90 and ∼45 Å,respectively. The discrete points in Figure 6 summarize atomicconcentrations of carbon (C), oxygen (O), fluorine (F), andsilicon (Si) as a function of takeoff angles for PHEMA brushesproduced on LYS proteins before and after modification withfluorinated agents. The solid lines in Figure 6 represent the

predicted elemental concentrations corresponding to uniformdistributions of fHEMA fluorinating agents, that is, assumingthat all hydroxyl groups of HEMA are coupled uniformly withthe respective fluorinated agents. The agreement between thepredicted and measured elemental compositions for bothangles reveals that the distribution of the fluorinated agentsinside the PHEMA brush is relatively uniform within the first90 Å.Contact angle measurements provide information about the

physicochemical properties of the surface with the amphiphiliccoatings. Figure 7 summarizes the changes in the advancing

(θA, squares) and receding (θR, circles) DIW contact angles ofPHEMA brushes grown from LYS and FIB supports after PPMwith fluorinating agents. The DIW contact angle over aPHEMA grafted surface is ∼48°, whereas the contact angles of−CF2− and densely packed −CF3 groups are ∼114 and ∼119°,respectively.40,51 From the data, coupling fluorinating agents to

Figure 5. (top) Ellipsometric dry thickness of parent PHEMA brushes(orange) and the corresponding thickness increase after fluorination ofPHEMA with various fluorinating agents (green). The experimentswere performed on two types of denatured protein-based substrates:lysozyme (LYS) and fibrinogen (FIB). (bottom) Corresponding molefraction of the fluorinated HEMA segments in P(HEMA-co-fHEMA)brushes, as determined from the ellipsometric data using a simpleanalytical model described in the Supporting Information.

Figure 6. XPS atomic surface concentration obtained at two differenttakeoff angles for carbon, oxygen, fluorine, and silicon, collected fromPHEMA brushes and PHEMA brushes fluorinated with F1, F3, or F7grown from LYS supports.

Figure 7. Advancing (squares) and receding (circles) contact anglesmeasured using deionized water (DIW) from P(HEMA-co-fHEMA)brushes prepared by fluorinating PHEMA brushes with F1, F3, and F7,as described in the text. The experiments were performed on two typesof denatured protein-based substrates: lysozyme (LYS) and fibrinogen(FIB).

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793775

PHEMA units result in an increase in DIW contact angles withincreasing the size of the fluorinating modifier. There is noconsiderable effect of the protein support layer on the contactangle data. Using the Cassie−Baxter model, an estimate of theexpected contact angle can be obtained for a system featuringuniformly distributed fluorinated units within HEMA buildingblocks, θmix.

52 By assuming that the areal fractions of HEMAand −CF3 are approximately equal, one arrives at a value ofθmix ≈ 82−85°. From the data, the θA of PHEMA-F1 is veryclose to the expected value corresponding to a uniformlydistributed HEMA and F1 units. In contrast, PHEMA-F7samples display θA in the range of ∼110−115°, which may beattributed to the surface segregation of fluorinated molecules atthe top of PHEMA brush. However, AR-XPS experiments ruleout considerable surface segregation of the fluorinated moietieson the surface. One possible reason for the increased contactangle can be associated with the roughness on the surface. Thisis supported by the observed increase in roughness of thesamples with increasing size of the fluorinating agent.Furthermore, roughness affects wettability on both the macroand microscales. On the macroscale, the wettability is describedby the Wenzel law. Here roughness contributes to wettability ina nonmonotonous manner; wettability increases with increasingroughness for θ ≈ <90°, whereas it decreases for θ ≈ >90°.Concurrently, the CAH also increases. From the data in Figure7, it is unlikely that macroscopic roughness affects thewettability in our specimens. On the microscale, increasingroughness leads to decreasing wettability; this effect is oftenaccompanied by a decrease in CAH. Whereas it is tempting torelate the wettability data and the surface roughness alone,molecular orientation (including possible surface reconstruc-tion) of chemical moieties present on the surface may alsocontribute to the observed behavior.Wang and coworkers53 demonstrated that the surface energy

of surfaces comprising semifluorinated moieties depends on thein-plane organization of those groups, which, in turn, varieswith increasing molecular packing. Longer fluorocarbon“fingers” featuring a larger number of −CF2− repeat unitspacked better and produced surfaces with lower surfaceenergies and lower wettabilities (i.e., higher DIW contactangles). Specifically, Wang et al. reported that surfaces withlonger fluorocarbon side chains, that is, −(CF2)6− or longer,produced highly oriented smectic B structures, whose top layermainly consisted of −CF3 units.53 On the basis of thesefindings, we suggest that whereas there is very little orientationof the fluorinated moieties in PHEMA-F1 and PHEMA-F3,PHEMA-F7 may feature some degree of self-organizationamong the F7 modifiers on the surface. Figure 8 presentspictorially the proposed molecular structure of the P(HEMA-co-fHEMA) brushes. We thus propose that in addition tosurface roughness the observed decrease in wettability ofPHEMA-F7 compared with PHEMA-F1 and PHEMA-F3 mayalso be associated with increasing fraction of −CF3 units andtheir better packing on the surface of P(HEMA-co-fHEMA)grafts. The movies in the Supporting Information providefurther evidence that surfaces featuring the F1 and F3 moietiesexhibit a relatively uniform distribution of the fluorinatingagents (drop pinning). In contrast, coatings prepared by PPMusing F7 have a large concentration of −CF3 groups on thesurface resulting in water droplets rolling off of the surface. Themolecular orientation of the fluorinating agents on the surfaceswill affect the performance of such amphiphiles, as will bedemonstrated below.

We have recently demonstrated that surfaces coated withPHEMA brushes as well as those54 made by modifyingPHEMA brushes with various fluorinating agents on silicasubstrates, some of which are presented in this article, areeffective in minimizing the adsorption of proteins.40 Here weshow that PP nonwoven sheets modified with the methodologyshown in Figure 1 exhibit protein fouling that is smaller thanthat of bare PP or PHEMA-modified PP. After surfacemodification, the functionalized PP nonwovens were incubatedin a solution of FITC-BSA (0.1 mg/mL concentration, pH 4.3)for 12 h at room temperature. The modified fibers were imagedusing laser scanning confocal microscope. Confocal microscopyimages of unmodified PP and LYS- and FIB-coated PPnonwoven sheets modified with amphiphilic coatings areshown in Figures 9 and 10, respectively. The images revealthat control PP, protein-coated PP, and PHEMA-coated PP arestained with FITC-BSA. However, the amphiphilic coatingsexhibit a strong reduction in fluorescein intensity, indicating alower deposition of the FITC-BSA and hence better antifoulingproperties. There is relatively little difference among theindividual coatings formed from fluorinated agents of differentlengths; however, it appears that PHEMA-F7 exhibits a largeradsorption of FITC-BSA relative to the other two amphiphiles.This behavior is likely associated with the presence of a largernumber of hydrophobic −CF3 groups on the surface ofPHEMA-F7, as previously discussed. We note that although thefluorescence microscopy measurements were carried out underidentical conditions for all specimens, it is not possible todetermine conclusively the absolute amount of adsorbedprotein. More quantitative studies will have to be performed,using multiple proteins and various protein concentrations, toquantify the effect of the type of the fluorinated modifier onantifouling properties of the coating. Nevertheless, whereasmore quantitative studies will have to be performed to quantifythe effect of the type of the fluorinated modifier on antifoulingproperties of the coating, our results reveal clearly thatcombining PHEMA and fluorine-based chemistries leads tocoating layers that inhibit significantly adsorption of proteinsrelative to bare hydrophobic or PHEMA-modified systems.

■ CONCLUSIONSWe have developed a simple and versatile method offunctionalizing hydrophobic surfaces by physisorbing denaturedproteins and cross-linking them. An appeal of the proposed

Figure 8. Schematic illustration depicting the spatial distribution of F1,F3, and F7 fluorinating species inside P(HEMA-co-fHEMA) brushes.

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793776

method is that it can be applied to various hydrophobicsurfaces; for example, flat hydrophobic surfaces as well as PPfibers were successfully used as demonstration of the proposedmethod. In addition, the properties of the coating layers werefound to be similar for two of the proteins tested, LYS and FIB.After priming the surfaces with a protein coating, they weresubsequently used to graft hydrophilic polymer brushes byusing surface-initiated polymerization from the proteins,followed by PPM reactions. Specifically, we preparedPHEMA macromolecular grafts on top and carried out PPMwith three fluorinated species. The resulting amphiphilic

coatings were tested for their ability to inhibit biomoleculeadsorption from solution. The fluorinated PHEMA amphiphiliclayers exhibited better antifouling characteristics than “bare”PHEMA brushes.The ability to endow PP fibers with functional layers without

resorting to harsh physical treatments such as plasma or coronamay be utilized in preparing filtration or separation platforms,whose function can be adjusted conveniently by choosing anappropriate chemistry of a polymer brush and eventuallyenhanced by employing a PPM reaction.

Figure 9. Merged optical microscopy and fluorescence microscopy images collected from fibers exposed to FITC-labeled BSA solutions. The fibersincluded bare polypropylene (PP), PP after coating with denatured layer of lysozyme (PP-LYS), and PP-LYS fibers after “grafting from”polymerization of HEMA (PP-LYS-PHEMA) and after fluorination of PHEMA with F1, F3, and F7 (PP-LYS-PHEMA-Fx, bottom row).

Figure 10. Merged optical microscopy and fluorescence microscopy images collected from fibers exposed to FITC-labeled BSA solutions. The fibersincluded bare polypropylene (PP), PP after coating with denatured layer of fibrinogen (PP-FIB), and PP-FIB fibers after “grafting from”polymerization of HEMA (PP-FIB-PHEMA) and after fluorination of PHEMA with F1, F3, and F7 (PP-FIB-PHEMA-Fx, bottom row).

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793777

Depending on the demands required from the coating layers,the application of amphiphilic coatings needs to be justified, forexample, when compared with single chemistries such asPHEMA, poly(ethylene glycol) methacrylate, various zwitter-ionic monomers, or hydroxypropyl methacrylamide. The datapresented here and in previous work by others demonstrateclearly that amphiphilic coatings produce surfaces, whoseperformance is oftentimes better than those made of singlecomponent chemistries.21,24,25,43,55

■ ASSOCIATED CONTENT*S Supporting InformationAnalytical model used to convert the ellipsometry thicknessdata into the mole fraction of the fluorinated HEMA segmentsresulted in P(HEMA-co-fHEMA) amphiphilic brushes. It alsoprovides information on immobilization of initiators on proteinprimer surfaces using XPS. The cross-linking of proteins in thepresence of GA and NaBH4 in solutions analyzed by SDS-PAGE is also reported. In addition, the growth of amphiphilicpolymer brushes on LYS-coated PP nonwovens and FIB-coatedODTS flat surfaces was also provided. The video shows thewettability of water on parent PP fiber mats and thefunctionalized surfaces. This material is available free of chargevia the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*E-mail: jan_genzer@ncsu.edu. Phone: +1-919-515-2069.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe thank The Nonwovens Institute at NC State University forsupporting this work. Partial support from the Office of NavalResearch (Grant No. N00014-07-1-0258) is also acknowledged.We thank Professor Orlin Velev (NC State University) forallowing us to use the confocal microscope in his lab. We thankDr. Eva Johannes from the Cellular and Molecular ImagingFacility (CMIF) for help with confocal microscopy experi-ments. Laser Imaging and Vibrational Spectroscopy Facilitylocated at North Carolina State University in the Departmentof Chemistry is acknowledged for help with FTIR spectroscopy.We also acknowledge the Chapel Hill Analytical and Nano-fabrication Laboratory (CHANL) at University of NorthCarolina for allowing us to use XPS and AFM equipment.This material is available free of charge via the Internet athttp://pubs.acs.org.

■ REFERENCES(1) Nakanishi, K.; Sakiyama, T.; Imamura, K. J. Biosci. Bioeng. 2001,91, 233−244.(2) Yang, Y.-F.; Wan, L.-S.; Xu, Z.-K. J. Membr. Sci. 2009, 326, 372−381.(3) Jang, J.; Go, W.-S. Fibers Polym. 2008, 9, 375−379.(4) Zhao, J.; Shi, Q.; Luan, S.; Song, L.; Yang, H.; Shi, H.; Jin, J.; Li,X.; Yin, J.; Stagnaro, P. J. Membr. Sci. 2011, 369, 5−12.(5) Chung, T. C.; Lee, S. H. J. Appl. Polym. Sci. 1996, 64, 567−575.(6) Hu, M.-X.; Yang, Q.; Xu, Z.-K. J. Membr. Sci. 2006, 285, 196−205.(7) Yang, Y.-F.; Wan, L.-S.; Xu, Z.-K. J. Membr. Sci. 2009, 337, 70−80.(8) Wyszogrodzka, M.; Haag, R. Langmuir 2009, 25, 5703−5712.(9) Zhang, Z.; Wang, J.; Tu, Q.; Nie, N.; Sha, J.; Liu, W.; Liu, R.;Zhang, Y.; Wang, J. Colloids Surf., B 2011, 88, 85−92.

(10) Grafahrend, D.; Heffels, K.-H.; Beer, M. V.; Gasteier, P.; Moller,M.; Boehm, G.; Dalton, P. D.; Groll, J. Nat. Mater. 2011, 10, 67−73.(11) Lu, J. R.; Su, T. J.; Thirtle, P. N.; Thomas, R. K.; Rennie, A. R.;Cubitt, R. J. Colloid Interface Sci. 1998, 206, 212−223.(12) Haynes, C. A.; Norde, W. J. Colloid Interface Sci. 1995, 169,313−328.(13) Haynes, C. A.; Sliwinsky, E.; Norde, W. J. Colloid Interface Sci.1994, 164, 394−409.(14) Howarter, J. A.; Youngblood, J. P. J. Colloid Interface Sci. 2009,329, 127−132.(15) Wang, C.; Feng, R.; Yang, F. J. Colloid Interface Sci. 2011, 357,273−279.(16) Houmadi, S.; Ciuchi, F.; De Santo, M. P.; De Stefano, L.; Rea, I.;Giardina, P.; Armenante, A.; Lacaze, E.; Giocondo, M. Langmuir 2008,24, 12953−12957.(17) Malmsten, M. Colloids Surf., B 1995, 3, 297−308.(18) Liu, Z.-M.; Xu, Z.-K.; Wang, J.-Q.; Yang, Q.; Wu, J.; Seta, P. Eur.Polym. J. 2003, 39, 2291−2299.(19) Guo, W.; Tang, X.; Xu, J.; Wang, X.; Chen, Y.; Yu, F.; Pei, M. J.Polym. Sci., Part A: Polym. Chem. 2011, 49, 1528−1534.(20) Edmondson, S.; Osborne, V. L.; Huck, W. T. S. Chem. Soc. Rev.2004, 33, 14−22.(21) Gan, D.; Mueller, A.; Wooley, K. L. J. Polym. Sci., Part A: Polym.Chem. 2003, 41, 3531−3540.(22) Estephan, Z. G.; Schlenoff, P. S.; Schlenoff, J. B. Langmuir 2011,27, 6794−6800.(23) Rodriguez-Emmenegger, C.; Brynda, E.; Riedel, T.; Houska, M.;Subr, V.; Alles, A. B.; Hasan, E.; Gautrot, J. E.; Huck, W. T. S.Macromol. Rapid Commun. 2011, 32, 952−957.(24) Keszler, B.; Kennedy, J. P.; Ziats, N. P.; Brunstedt, M. R.; Stack,S.; Yun, J. K.; Anderson, J. M. Polym. Bull. 1992, 29, 681−688.(25) Krishnan, S.; Ayothi, R.; Hexemer, A.; Finlay, J. A.; Sohn, K. E.;Perry, R.; Ober, C. K.; Kramer, E. J.; Callow, M. E.; Callow, J. A.;Fischer, D. A. Langmuir 2006, 22, 5075−5086.(26) Huang, J.; Murata, H.; Koepsel, R. R.; Russell, A. J.;Matyjaszewski, K. Biomacromolecules 2007, 8, 1396−1399.(27) Yamamoto, K.; Miwa, Y.; Tanaka, H.; Sakaguchi, M.; Shimada,S. J. Polym. Sci., Part A: Polym. Chem. 2002, 40, 3350−3359.(28) Krause, B.; Voigt, D.; Hauβler, L.; Auhl, D.; Munstedt, H. J.Appl. Polym. Sci. 2006, 100, 2770−2780.(29) Yao, F.; Fu, G.-D.; Zhao, J.; Kang, E.-T.; Neoh, K. G. J. Membr.Sci. 2008, 319, 149−157.(30) Strobel, M.; Branch, M. C.; Ulsh, M.; Kapaun, R. S.; Kirk, S.;Lyons, C. S. J. Adhes. Sci. Technol. 1996, 10, 515−539.(31) Elsabee, M. Z.; Abdou, E. S.; Nagy, K. S. A.; Eweis, M.Carbohydr. Polym. 2008, 71, 187−195.(32) Goli, K. K.; Rojas, O. J.; Ozcam, A. E.; Genzer, J.Biomacromolecules 2012, 13, 1371−1382.(33) Brynda, E.; Houska, M. J. Colloid Interface Sci. 1996, 183, 18−25.(34) Brynda, E.; Houska, M. Macromol. Rapid Commun. 1998, 19,173−176.(35) Gao, W.; Liu, W.; Christensen, T.; Zalutsky, M. R.; Chilkoti, A.Proc. Natl. Acad. Sci. 2010, 107, 16432−16437.(36) Peeler, J. C.; Woodman, B. F.; Averick, S.; Miyake-Stoner, S. J.;Stokes, A. L.; Hess, K. R.; Matyjaszewski, K.; Mehl, R. A. J. Am. Chem.Soc. 2010, 132, 13575−13577.(37) Depp, V.; Alikhani, A.; Grammer, V.; Lele, B. S. Acta Biomater.2009, 5, 560−569.(38) Sawada, S.-ichi; Sasaki, Y.; Nomura, Y.; Akiyoshi, K. ColloidPolym. Sci. 2010, 289, 685−691.(39) Migneault, I.; Dartiguenave, C.; Bertrand, M. J.; Waldron, K. C.BioTechniques 2004, 37, 790−802.(40) Arifuzzaman, S.; Ozcam, A. E.; Efimenko, K.; Fischer, D. A.;Genzer, J. Biointerphases 2009, 4, FA33−FA44.(41) Liu, X.; Goli, K. K.; Genzer, J.; Rojas, O. J. Langmuir 2011, 27,4541−4550.(42) Balachandra, A. M.; Baker, G. L.; Bruening, M. L. J. Membr. Sci.2003, 227, 1−14.(43) Galvin, C. J.; Genzer, J. Prog. Polym. Sci. 2012, 37, 871−906.

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793778

(44) Wang, Y.; Lieberman, M. Langmuir 2003, 19, 1159−1167.(45) Salloum, D. S.; Schlenoff, J. B. Biomacromolecules 2004, 5,1089−1096.(46) Malmsten, M. J. Colloid Interface Sci. 1994, 166, 333−342.(47) Farris, S.; Song, J.; Hunag, Q. J. Agric. Food Chem. 2010, 58,998−1003.(48) Ekblad, T.; Andersson, O.; Tai, F.-I.; Ederth, T.; Liedberg, B.Langmuir 2009, 25, 3755−3762.(49) Kong, J.; Yu, S. Acta Biochim. Biophys. Sin. 2007, 39, 549−559.(50) Huang, W.; Kim, J.-B.; Bruening, M. L.; Baker, G. L.Macromolecules 2002, 35, 1175−1179.(51) Nishino, T.; Meguro, M.; Nakamae, K.; Matsushita, M.; Ueda,Y. Langmuir 1999, 15, 4321−4323.(52) Genzer, J.; Efimenko, K. Biofouling 2006, 22, 339−360.(53) Wang, J.; Mao, G.; Ober, C. K.; Kramer, E. J. Macromolecules1997, 30, 1906−1914.(54) Genzer, J.; Arifuzzaman, S.; Bhat, R. R.; Efimenko, K.; Ren, C.-lai; Szleifer, I. Langmuir 2012, 28, 2122−30.(55) Krishnan, S.; Weinman, C. J.; Ober, C. K. J. Mater. Chem. 2008,18, 3405−3413.

Biomacromolecules Article

dx.doi.org/10.1021/bm301223b | Biomacromolecules 2012, 13, 3769−37793779