Initiated Atom Transfer Radical Polymerization ... - Wiley Online Library

Full Paper

974

Summary: Surface-active microporous membranes were prepared from the poly(vinylidene fluoride)-graft-poly(2-(2bromoisobutyryloxy)ethyl acrylate) copolymer (PVDF-gPBIEA copolymer) by phase inversion in water. The PBIEA side chains could function as initiators for the atom transfer radical polymerization (ATRP) of 2-(N,N-dimethylamino)ethyl methacrylate on the membrane surfaces to give rise to the PVDF-g-PBIEA-ar-PDMAEMA membranes. N-alkylation with hexyl bromide and nitromethane gave rise to the quanternized PVDF-g-PBIEA-ar-QPDMAEMA membranes with polycation chains chemically tethered on the membrane surface, including the pore surfaces. The changes in the surface morphology and the surface chemical composition were confirmed by scanning electron microscopy and X-ray photoelectron spectroscopy. The scanning electron microscopy revealed that, in comparison to the pristine PVDF-g-PBIEA membranes, not only could the PVDF-gPBIEA-ar-QPDMAEMA membranes remove the Gramnegative bacterium Escherichia coli but also inhibited the

bacterial reproduction on the membranes to a significant extent.

PVDF-g-PBIEA and PVDF-g-PBIEA-ar-QPDMAEMA membranes after exposure to water-borne E. coli suspension for 24 h.

Surface-Initiated Atom Transfer Radical Polymerization on Poly(Vinylidene Fluoride) Membrane for Antibacterial Ability Guangqun Zhai,*1 Zhi L. Shi,2 En T. Kang,2 Koon G. Neoh2 1

Department of Materials Science and Engineering, Jiangsu Polytechnic University, Changzhou 213016, People’s Republic of China Fax: (þ86) 519 3290011; E-mail: [email protected] 2 Department of Chemical and Environmental Engineering, National University of Singapore, Kent Ridge, Singapore 119260, Singapore

Received: April 14, 2005; Revised: June 14, 2005; Accepted: June 15, 2005; DOI: 10.1002/mabi.200500079 Keywords: antibacterial; atom transfer radical polymerization (ATRP); membranes; PVDF; surface-active

Introduction Grafting of polymer chains is a strategy to impart the existing material surfaces with specific functionalities such as adhesion, biocompatibility, anti-fouling, and antistatic.[1,2] There are two approaches to chemically tether the polymer chains onto the substrate surfaces, viz. the reaction of telechelic polymers with the surfaces (grafting onto) and the surface-initiated polymerizations (grafting from).[3] The ‘‘grafting onto’’ approach can be readily conducted on reactive substrates with telechelic polymers. For examples, hydroxyl-terminated poly(ethylene glycol) (PEG) had been Macromol. Biosci. 2005, 5, 974–982

immobilized on anhydride-containing surface via esterification to improve the anti-fouling properties and blood compatibility of polymer membranes.[4] Epoxy-terminated PEG was grafted onto amine-containing polyethylene film via amidization.[5] The typical strategy for the ‘‘grafting from’’ approach is to immobilize the initiator moiety initially, followed by the subsequent polymerization of functional monomers to result in the surface-bound polymer chains. Ozone pretreatment is a general approach to activate the polymer substrates. The thermally labile peroxide groups were introduced onto the substrate surface, which are subsequently decomposed into free radicals to initiate

DOI: 10.1002/mabi.200500079

ß 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

975

Surface-Initiated Atom Transfer Radical Polymerization on Poly(Vinylidene Fluoride) Membrane for Antibacterial Ability

The PVDF-g-PBIEA copolymer and the PVDF-g-PBIEA microporous membranes were prepared as shown schematically in Figure 1.[20] 2-(N,N-dimethylamino)ethyl methacrylate (DMAEMA) was purchased from Aldrich Chemical Company, of Milwaukee, WI. The inhibitor in DMAEMA was removed by column chromatography. Copper (I) chloride (CuCl), copper (II) chloride (CuCl2), and the ligand for the ATRP, bipyridine (bPy), were also purchased from Aldrich Chemical Company. They were used as received. Fluorescien (Na salt) was purchased from Sigma Chemical Co. Peptone, yeast extract, agar, and beef extract were purchased from Oxoid. E. coli was obtained from American-type culture collection (ATCC DH5a). Surface Functionalization of the PVDF-g-PBIEA Membrane For the surface-initiated ATRP of DMAEMA on the PVDF-gPBIEA membranes, 2 ml of DMAEMA (12 mmol), 13 ml PVDF Main Chain

O3/O 2

O O

O OH O

O CH3

60 oC CH2=CH-C-O-CH 2-CH 2-O-C-C-Br BIEA Side Chain

(BIEA)

CH3

PVDF-g-PBIEA O O Microporous Membrane by Phase Inversion

PVDF-g-PBIEA Copolymer

Surface-Active PVDF-g-PBIEA Membrane ATRP of DMAEMA


PVDF-g-PBIEA-ar-PDMAEMA Membrane C6H13Br/CH 3NO2 ++++

+ + + + ++ + + ++ + + + ++ ++ + + + ++ + + +++ + ++ + + ++ + + + + + + ++ ++ + + ++ + + + ++++++++ ++++ + + + + + + + + + + + ++ ++ + +


+

+ ++ + ++ + + ++


www.mbs-journal.de

Materials


Macromol. Biosci. 2005, 5, 974–982

Experimental Part


the radical graft copolymerization in a conventional fashion. For example, the polyurethane film was pretreated by ozone to generate peroxide groups on the surface, which could initiate free radical graft copolymerization to produce the surface-tethered PEG chains.[6] As a membrane material, poly(vinylidene fluoride) (PVDF) has been extensively studied in the past decades for its application in ultrafiltration and microfiltration membrane,[7,8] protein adsorption, immobilization and separation,[9,10] wasterwater treatment,[11,12] proton conduction,[13,14] and stimuli-responsive controlled releases.[15] With the advances in controlled/‘‘living’’ radical polymerization techniques, atom transfer radical polymerization (ATRP) has been readily utilized to prepared well defined copolymer based on PVDF copolymers and membranes. For instances, Mayer et al. reported a simple strategy to prepare PVDF graft copolymer via ATRP initiated by the secondary C–F moieties along the main chains.[16] On the other hand, 4-vinylbenzl chloride (VBC) was graft copolymerized on the PVDF membrane surface after exposure to g-ray irradiation, where the surface-bonded benzyl chloride moieties can function as an efficient initiator for the ATRP of styrene to produce the PVDF-g-PVBCar-PS membranes.[17] Strategies have been developed to address the growing need for preparing antibacterial surface and for preventing the formation of biofilms among which, the incorporation of both the silver nanoparticles[18] and the quanternary ammonium groups[19] have been intensively studied for their efficacy. Obviously, the manipulation of polymer is much more facile than that of inorganic nanoparticles, especially on the organic substrates. Recently, we have reported on the graft copolymerization of the ozone-pretreated PVDF with an ATRP-type inimer (initiator monomer), 2-(2-bromoisobutyryloxy)ethyl acrylate (BIEA), to prepare the poly(vinylidene fluoride)-graftpoly(2-(2-bromoisobutyryloxy)ethyl acrylate) (PVDF-gPBIEA) copolymer, which could function as a macroinitiator for arborescent macromolecules and as an active membrane substrate for surface-initiated ATRP.[20] This approach has offered exciting options for the macromolecular design at both molecular and surface levels. In this work, we proposed a strategy to obtain the antibacterial membranes based on this surface-active membrane. Microporous membrane with ‘‘active’’ or ATRP initiator-immobilized surfaces, including pore surfaces, were cast from PVDF-g-PBIEA solutions by phase inversion. Surface-initiated ATRP of tertiary amine-containing monomers from the microporous ‘‘surface-active’’ membrane, followed by N-alkylation or quanternization by alkyl halide, gave rise to arborescent polycation chains chemically bonded on the membrane surfaces. The antibacterial characteristics of the so-obtained membranes to the Gramnegative bacterium Escherichia coli were qualitatively analyzed.

Antibacterial PVDF-g-PBIEA-ar-QPDMAEMA Membrane

Figure 1. The schematic illustration of the preparation of the PVDF-g-PBIEA copolymers, fabrication, and surface functionalization of the PVDF-g-PBIEA microporous membranes. ß 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

976

G. Q. Zhai, Z. L. Shi, E. T. Kang, K. G. Neoh

of deionized water, and 1 piece of the PVDF-g-PBIEA MF membrane (about 0.03 g in weight) of about 1 cm 2 cm in area were added into a Pyrex1 test tube. A purified argon stream was introduced to degas the mixture for about 20 min. 45.6 mg of 2,20 -bPy (0.30 mmol), 12 mg of CuCl2 (0.09 mmol), and 13.4 mg of CuCl (0.14 mmol) were added to the solution. The polymerization reaction was allowed to proceed for predetermined time at room temperature. After the reaction, the membrane was washed with copious amount of doubly distilled water over a period of about 6 h, and was dried by pumping overnight under reduced pressure. The surfaceinitiated ATRP of DMAEMA on the PVDF-g-PBIEA membrane surface is illustrated schematically in Figure 1. For the N-alkylation of the PVDF-g-PBIEA-ar-PDMAEMA membrane, 0.5 ml hexyl bromide, 1.5 ml nitromethane were added into the test tube to totally immerse a piece of the PVDFg-PBIEA-ar-PDMAEMA membrane. The tube was sealed and allowed to react in a water bath at 35 8C for 48 h to produce the quanternized PVDF-g-PBIEA-ar-PDMAEMA (PVDF-gPBIEA-ar-QPDMAEMA) membrane. After the reaction, the PVDF-g-PBIEA-ar-QPDMAEMA membrane was washed with copious amount of methanol and doubly distilled water, and dried for the subsequent characterization and bioassay. The N-alkylation of the PVDF-g-PBIEA-ar-PDMAEMA membrane is also illustrated schematically in Figure 1.

Characterization of the Polymeric Membranes The surface chemical composition of the membranes was analyzed by X-ray photoelectron spectroscopy (XPS) on a Kratos AXIS HSi spectrometer, under conditions similar to those reported in the earlier work.[20] The surface morphology of the membranes was studied by scanning electron microscopy (SEM) on a JEOL 6320 SEM. The membranes were mounted on the sample studs by means of double-sided adhesive tapes. A thin layer of palladium was sputtered onto the membrane surface prior to the SEM measurement. The measurements were performed at an accelerating voltage of 15 kV. The transmission Fourier transform infrared spectra of the PVDF-g-PBIEA copolymer, the PVDF-g-PBIEA-arPDMAEMA and PVDF-g-PBIEA-ar-QPDMAEMA membranes were recorded on Nicolet Avatar 370 spectrometer. The antibacterial characteristics to waterborne E. coli of both the pristine PVDF-g-PBIEA membrane and the PVDF-gPBIEA-ar-QPDMAEMA membrane were determined. The E. coli was cultivated under the condition similar to reported elsewhere.[21,22] E. coli was cultivated in 50 ml of a 3.1% yeast dextrose broth (containing 10 g l1 peptone, 8 g l1 beef extract, 5 g l1 sodium chloride, 5 g l1 glucose, and 3 g l1 yeast extract at a pH of 6.8) at 37 8C. The bacterial cell concentration was estimated from the optical density at 540 nm based on the standard calibration with the assumption that the optical density of 1.0 at 540 nm is equivalent to approximately 109 cells ml1. For the waterborne antibacterial assay, the E. coli-containing broth was centrifuged at 2 700 rpm for 10 min, and after the removal of the supernatant, the cells were washed twice with PBS (pH ¼ 7.0) and resuspended in PBS at a concentration of 107 cells ml1. The substrates were immersed in this suspension in a sterile Erlenmeyer flask. Macromol. Biosci. 2005, 5, 974–982

www.mbs-journal.de

The flask was then shaken at 200 rpm at 37 8C for 2 h. The substrates were then removed from the above flask and washed three times with sterile PBS and placed in Petri dishes. This was followed by the immediate addition of solid growth agar in yeast-dextrose broth, autoclaved, poured into a Petri dish, and dried under reduced pressure at room temperature overnight. The sealed Petri dishes were then incubated at 37 8C for 24 h. The substrates after the waterborne antibacterial tests were characterized by SEM. The sample fixation and preparation for SEM were as follows: the substrates were first washed with PBS right after incubation and 3 vol.-% glutaraldehyde in PBS was added for 5 h and stored at 4 8C. After that, the glutaraldehyde solution was removed and the substrates were washed with PBS, followed by step dehydration with 25, 50, 70, 95, and 100% ethanol for 10 min each. The substrates were then dried and sputter-coated with a thin film of platinum for SEM imaging. In another test, the sealed Petri dishes were incubated at 37 8C for 2 h, with other procedures similar to those mentioned above.

Results and Discussion The preparation and characterization of PVDF-g-PBIEA copolymer and its surface-active PVDF-g-PBIEA membrane had been reported earlier.[20] The chemical composition of the PVDF-g-PBIEA copolymer was determined by elemental analysis. The bulk graft concentration of the PVDF-g-PBIEA copolymer, defined as the number of BIEA repeat units per repeat unit of PVDF, or the ([–BIEA–]/ [–CH2CF2–])bulk ratio was about 0.05, based on a ([Br]/ [C])bulk ratio of 0.02, using the equation ([–BIEA–]/ [–CH2CF2–])bulk ¼ 2[Br]/([C] 9[Br]), where the factors 2 and 9 are introduced to account for the fact that there are 2 and 9 carbon atoms per repeat unit of PVDF and BIEA polymer, respectively. The average degree of the PBIEA side chain is estimated to be about 8, based on the peroxide content of around 104 mol g1 and the bulk graft concentration of the copolymer.[20] Figure 2(a) shows the SEM image, obtained at a magnification of 2 000 of the PVDFg-PBIEA membrane. The PVDF-g-PBIEA membrane exhibits a well defined porous structure with an average pore diameter of 1–2 mm. Figure 3 shows the XPS wide-scan, C 1s and Br 3d corelevel spectra of the PVDF-g-PBIEA membrane. The C 1s lineshape was curve-fitted with five peak components using the following strategy. The peak components of equal intensity with binding energies (BEs) of 286.0 and 290.5 eV are attributed to the –CH2– and –CF2– species, respectively, of the PVDF main chains. The peak components with BEs of about 284.6 and 288.8 eV are associated, respectively, with the neutral C–H and the O C–O species of the BIEA side chains. The peak component with the BE of about 286.5 eV arises from combined contributions of the C–O species and the C–Br terminus of the BIEA side chains.[20,23] On the other hand, the Br 3d core-level spectra were resolved with ß 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim


Figure 2. SEM images of (a) the PVDF-g-PBIEA membrane, the PVDF-g-PBIEA-ar-PDMAEMA membrane with a reaction time of (b) 1 min and (c) 5 min, respectively, and (d) the PVDF-g-PBIEA-arQPDMAEMA membrane.

two peak components assigned to Br 3d3/3 and Br 3d5/2, with a separation energy of 1.05 eV, respectively.[23] Surface enrichment of PBIEA chains occurs during phase inversion to give rise to a ([–BIEA–]/[–CH2CF2–])surface ratio of 0.07, based on the XPS-derived ([Br]/[C])surface ratio of 0.03. The BIEA repeat units on the PVDF-g-PBIEA membrane surface could function as ATRP initiators of specific monomers. In the preparation of antibacterial polymers, a typical strategy involves the copolymerization of monomers containing amine moieties, such as vinylpyridine[21] and DMAEMA,[24] followed by N-alkylation with alkyl halide. The ATRP of DMAEMA in aqueous or alcoholic media using a variety of initiators and catalyst systems had been extensively studied.[25–27] Surface-initiated ATRP of DMAEMA on some functionalized substrates were also developed as the surface-tethered PDMAEMA chains could impart specific functionalities to the substrate materials.[24,28,29] In this study, a very short ATRP time of 1 min with DMAEMA was allowed for the preparation of the PVDF-gPBIEA-ar-PDMAEMA membranes, with the intention of preventing the pores from being excessively filled by the PDMAEMA side chains. Due to the short reaction time and low initiator amount, the monomer conversion was relatively low. By comparing the weights of the membrane before and after the surface-initiated ATRP, the monomer conversion was estimated to be less than 5%. The absorption bands at the wave number of about 2 820 and 1 440 cm1, ascribed to the ns and ds modes of N–CH3 group, appeared on the FTIR spectrum of the PVDF-g-PBIEA-ar-PDMAEMA membrane, indicating Macromol. Biosci. 2005, 5, 974–982

www.mbs-journal.de

that DMAEMA has been incorporated into the membrane. The surface morphology and the surface chemical composition of the PVDF-g-PBIEA-ar-PDMAEMA membranes were probed by SEM and XPS, respectively. Figure 2(b) shows the SEM images, obtained at a magnification of 2 000, of the PVDF-g-PBIEA-ar-PDMAEMA membranes after 1-min ATRP of DAMEMA. It was observed that the surface morphology of the pristine PVDF-g-PBIEA has significantly been varied after the surface-initiated ATRP for 1 min. The pore density was reduced considerably after the surface grafting. On the other hand, the SEM image also indicated that the pore size seems to increase a little bit after the membrane is treated with DMAEMA. Since the surface-initiated ATRP of DMAEMA also occurred on the pore wall, and the PDMAEMA side chains produced from the pore walls may greatly be contributed to the filling of the membrane pores, partially accountable for the decrease in the pore density. Scanning electron microscopy results also showed that after 5 min surface-initiated ATRP of DMAEMA, a dense layer of PDMAEMA formed on the PVDF-g-PBIEA membrane surface, completely covering the membrane surfaces, and no porous structure was observed, because the PDMAEMA chains have completely filled the pore spaces, as shown in Figure 2(c). The high rate of the surfaceinitiated ATRP of DMAEMA on the PVDF-g-PBIEA membrane in an aqueous medium can be attributed to high spatial concentration of ATRP initiators on the grafted PBIEA chains, which allows the formation of an arborescent PDMAEMA architecture. On the other hand, it has been noted that the polar methacrylate monomers, ß 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

977

978


PVDF-g-PBIEA Membrane Wide WideScan Scan

(a)

F 1s

C 1s O 1s

O KLL

Intensity (Arb. Units)

Br 3d

0 200 400 600 800 1000 1200 C 1s (b) (-CH2-)PVDF

C-O and C-Br -CH-

O=C-O

282

284

286

288

(-CF2-)PVDF

292

290

Br 3d

(c) Br 3d 5/2

66

Br 3d 3/2

70 74 72 68 Binding Energy (eV)

76

Figure 3. XPS wide-scan, C 1s and Br 3d spectra of the PVDF-gPBIEA membrane.

for examples, embracing DMAEMA,[30] 2-hydroxyethyl methacrylate (HEMA),[31] and poly(ethylene glycol) methacrylate (PEGMA),[32] show a remarkable solvent acceleration effect in aqueous or alcoholic ATRP processes. The preferable coordination between solvent molecules and ATRP catalyst might be accountable for such a universal phenomenon.[33,34] Figure 4(a) and (b) show, respectively, the XPS widescan and C 1s core-level spectra of the PVDF-g-PBIEA-arPDMAEMA membrane. The wide-scan spectrum is dominated by the C 1s, N 1s, F 1s, and O 1s signals. A Br 3d signal with BE of about 70 eV was also observed. The appearance of the N 1s signal at BE of about 400 eV indicates that the DMAEMA was graft copolymerized on the PVDF-gPBIEA membrane surfaces. The surface elemental ratio Macromol. Biosci. 2005, 5, 974–982

www.mbs-journal.de

of nitrogen to fluorine or the ([N]/[F])surface ratio, determined from their respective peak areas, is about 1.05. Thus, the surface graft concentration, defined as the number of DMAEMA repeat units per repeat unit of PVDF, or the ([–DMAEMA–]/[–CH2CF2–])surface ratio, can be estimated to be about 2.1 simply from the equation ([–DMAEMA–]/ [–CH2CF2–])surface ¼ 2 ([N]/[F])surface, where the factor 2 is introduced because each repeat unit of PVDF contains two fluorine atoms while each DMAEMA repeat unit contains one nitrogen atom only. The sampling depth of XPS, as a surface characterization technique, is reported to be about 7.5 nm for organic matrix.[35] Therefore, such a surface functionalization may not lead to a detectable variation in the elemental composition of the membrane bulk. The C 1s lineshape of the PVDF-g-PBIEA-arPDMAEMA membrane was curve-fitted with six peak components, adopting the following strategy. The major peak component with BE of about 284.6 eV is associated with the neutral C–H species. The peak components with BEs of 285.8 and 290.5 eVand of about equal intensity were assigned to the CH2 and CF2 species of the PVDF main chains.[20,23] The peak components with BEs of about 286.2 and 288.8 eV were attributed to the C–O species and the O C–O species, respectively, of the corresponding PBIEA and PDMAEMA graft chains. The peak component with BE of about 285.5 eV is attributed to the C–N species of the PDMAEMA side chains.[23] In comparison to that of the PVDF-g-PBIEA membrane, the intensity of the (CF2)PVDF species was reduced to a considerable extent. The molar ratio of the (CF2)PVDF species to the total carbon species, determined from the corresponding C 1s peak component spectral areas, was decreased from 0.28 for the pristine PVDF-g-PBIEA to 0.04 after 1-min of surface-initiated ATRP of DMAEMA, indicating that the PVDF matrix was intensively covered by the PDMAEMA layer. The ([–DMAEMA–]/ [–CH2CF2–])surface ratio could also be deduced from the C 1s core-level spectrum. The peak component of (CF2)PVDF is associated with the PVDF main chains, and the peak component of C–N is attributed to PDMAEMA graft chains. The ([C–N]/[CF2]PVDF)surface ratio is determined from their respective peak component spectral area to be about 6.0. Thus, the ([–DMAEMA–]/ [–CH2CF2–])surface is estimated from the equation([–DMAEMA–]/[–CH2CF2–])surface ¼ ([C–N]/[CF2]PVDF)surface/3 to be about 2.0. The factor of three is introduced to account for the fact that there are three C–N species per DMAEMA repeat unit and only one CF2 per repeat unit of PVDF. This result is very close to that from the ([N]/[F])surface ratio. Figure 4(c) and (d) show the XPS N 1s and Br 3d corelevel spectra of the PVDF-g-PBIEA-ar-PDMAEMA membrane. The N 1s spectrum was resolved into the C–N peak component and the C–Nþ peak component with the BEs of ß 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

979


PVDF-g-PBIEA-ar-QPDMAEMA Membrane

PVDF-g-PBIEA-ar-PDMAEMA Membrane

(a)

Wide scan F 1s

C 1s

C 1s

O 1s

O 1s Br 3d

0

(b)

200

Br 3d

N 1s

N 1s

Br 3p

400

600

800

1000

1200

0

200

400

Intensity (Arb. Units)

1200

284

286

288

CF2

CF2

290

292

282

284

286

288

N 1s

400

402

292

(g)

C-N

C-N +

398

290

C-N +

C-N

(d)

1000

C-N (CH2)PVDF O=C-O C-O

C-H

O=C-O

C-N C-O

396

800

(f)

(CH2)PVDF

(c)

600

C 1s C-H

282

(e)

F 1s

404

406

396

398

400

402

404

(h)

Br 3d Br-

Br-

C-Br C-Br

64

66

68

70

72

74

64

66

68

70

72

74

Binding Energy (eV) Figure 4. XPS wide-scan, C 1s, N 1s, and Br 3d spectra of the PVDF-g-PBIEA-arPDMAEMA membrane and the PVDF-g-PBIEA-ar-QPDMAEMA membrane.

about 399.5 and 402 eV, respectively.[23] The Br 3d spectrum was resolved into C–Br peak component and Br peak component, with BEs of about 70.5 and 67 eV, respectively.[23] For the N 1s lineshape, the neutral C–N species of the PDMAEMA side chains dominates the total nitrogen species. The molar ratio of the C–Nþ species to the total nitrogen species is estimated to be about 0.2. On the other hand, the molar ratio of the Br species to the total bromine species is estimated to be about 0.71, indicating that the terminal C–Br moieties of the PDMAEMA side chains could be readily involved in the N-alkylation of the tertiary Macromol. Biosci. 2005, 5, 974–982

www.mbs-journal.de

amine groups of the PDMAEMA chain, leading to the formation of the C–Nþ species and Br species. Hexylbromide is used for the N-alkylation of the PVDFg-PBIEA-ar-PDMAEMA membrane into the quanternized PVDF-g-PBIEA-ar-QPDMAEMA membrane, as an N-alkyl chain of six-carbon units in length and has been shown to be one of the most effective.[36] After the reaction, FTIR results indicated that a new absorption band at the wave number of about 1 475 cm1, suggesting that the tertiary amine species has been quanternized. The surface morphology and the surface chemical composition ß 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

980


of the PVDF-g-PBIEA-ar-QPDMAEMA membrane were analyzed by XPS and SEM, respectively. Figure 2(d) shows the SEM image, obtained at a magnification of 2 000, of the PVDF-g-PBIEA-ar-QPDMAEMA membrane. The pore density of the PVDF-g-PBIEA-arQPDMAEMA membrane was reduced significantly by the N-alkylation reaction of PDMAEMA side chains, because the hexyl groups were introduced into the each repeat unit of the PDMAEMA side chains. However, the average pore size only underwent a slight reduction after the N-alkylation reaction. Figure 4(e) and (f) show the XPS wide-scan and C 1s core-level spectra of the PVDF-g-PBIEA-arQPDMAEMA membranes. In comparison to that of the corresponding PVDF-g-PBIEA-ar-PDMAEMA membrane, the Br 3d signal has been significantly enhanced after the N-alkylation with hexylbromide. The surface elemental ratio, or the ([C]:[O]:[N]: [Br])surface ratio, of the PVDF-g-PBIEA-ar-QPDMAEMA membrane, is determined to be about 100:16.87:7.16:4.10. The significant increase in the surface elemental ratio of bromine to carbon, or the ([Br]/[C])surface ratio, indicated that the PDMAEMA side chains had been N-alkylated by hexyl bromide. The C 1s core-level spectrum of the PVDF-g-PBIEA-ar-QPDMAEMA membrane was curvefitted using the similar strategy as that of PVDF-g-PBIEAar-PDMAEMA membrane. After the N-alkylation reac-

tion, the molar ratio of the C–N species was increased from 0.25 for the PVDF-g-PBIEA-ar-PDMAEMA membrane to 0.30, confirming the occurrence of N-alkylation of PDMAEMA side chains. Figure 4(g) and (h) show, respectively, the N 1s and Br 3d core-level spectra of the PVDF-g-PBIEA-arQPDMAEMA membrane. They were resolved using the similar approach as that for the PVDF-g-PBIEA-arPDMAEMA membrane. After the N-alkylation, the C–Nþ peak component dominates the N 1s lineshape of the PVDF-g-PBIEA-ar-QPDMAEMA membrane, and the surface molar ratio of the C–Nþ species to the total nitrogen species is estimated to be about 0.80, significantly higher than that of the initial PVDF-g-PBIEA-ar-PDMAEMA membrane, indicating that most DMAEMA repeat units have been N-alkylated. The molar ratio of the Br species to the total bromine species is estimated to be about 0.84. The C–Br peak component results primarily from the chain termini of the PDMAEMA chains. The XPS results suggested that most of the DMAEMA repeat units in the side chains had been quanternized to polycations for the subsequent antibacterial measurement. The antibacterial assay to waterborne E. coli was carried out on both the pristine PVDF-g-PBIEA membrane and the PVDF-g-PBIEA-ar-QPDMAEMA membrane, and the results were revealed by SEM images. Figure 5 shows

Figure 5. SEM images of the PVDF-g-PBIEA membrane (a and b) and the PVDF-g-PBIEA-arQPDMAEMA (c and d) membrane after 2- and 24-h exposure to waterborne E. coli suspensions, respectively. Macromol. Biosci. 2005, 5, 974–982

www.mbs-journal.de



the SEM images of the PVDF-g-PBIEA membrane and the PVDF-g-PBIEA-ar-QPDMAEMA membranes after a 2- and 24-h exposure to waterborne E. coli suspension, respectively. Figure 5(a) and (c) show that after 2-h exposure to the waterborne E. coli suspension, there are a comparable amount of bacteria E. coli adhered on the membrane surfaces, indicating that the PVDF-g-PBIEA and PVDF-gPBIEA-ar-QPDMAEMA membrane have comparable adhesion characteristics to bacteria E. coli in this study. Figure 5(b) and (d) show the SEM images of PVDF-gPBIEA membrane and PVDF-g-PBIEA-ar-QPDMAEMA membrane after exposure to waterborne E. coli suspension for 24 h, respectively. For the PVDF-g-PBIEA membrane, E. coli was densely and uniformly distributed throughout the membrane surface. However, for the PVDF-g-PBIEAar-QPDMAEMA membrane, a significant decease in the concentration of E. coli cells was observed and the bacteria were sparsely distributed on the membrane surface, confirming the antibacterial characteristics of the PVDF-gPBIEA-ar-QPDMAEMA membrane. Since the two membranes exhibit the comparable adhesion ability to the bacteria, the SEM results implied that the quanternized PDMAEMA chains inhibited partially the reproduction of the bacteria E. coli on the membrane surface. Besides, it also indicated that the quanternized PDMAEMA chains may inhibit the growth of the bacteria to a significant extent as well, since the bacteria on the PVDF-g-PBIEA-arQPDMAEMA membrane surface exhibit a much smaller size than those on the PVDF-g-PBIEA-ar-PDMAEMA membrane surface. These SEM results indicated that the QPDMAEMA chains on the membrane surface, including the pore surfaces, are accountable for the antibacterial ability of the PVDF-g-PBIEA-ar-QPDMAEMA membrane. Although the exact mechanism of the antibacterial action of quanternary ammoniums to Gram-negative bacteria is underfurther investigation, it is, to a great extent, attributed to enhance the cell permeability and disrupt cell membranes, finally leading to the destruction of the bacterial body.[37–39] In addition, a closer observation on Figure 5(b) and (d) also reveals that the E. coli cell has a typical size of 2–3 mm in length, considerably larger than the average pore size of the PVDF-g-PBIEA-ar-QPDMAEMA membrane in this study. It might imply that the PVDF-g-PBIEA-arQPDMAEMA membrane might help filter E. coli out and inhibit their reproduction spontaneously, which, however, needs further investigation on the size exclusion effect for the verification.

Conclusion The PVDF-g-PBIEA-ar-PDMAEMA microporous membrane was obtained by the surface-initiated ATRP of DMAEMA on the surface-active PVDF-g-PBIEA microMacromol. Biosci. 2005, 5, 974–982

www.mbs-journal.de

porous membrane. The pore size and surface morphology can be controlled by changing the ATRP time. The surface grafted PDMAEMA chains were transformed into the QPDMAEMA side chains by the N-alkylation with hexyl bromide. SEM images show that the surface modification did not destroy the porous structure of the membrane. XPS results suggested that the PDMAEMA side chains were quanternized into the polycationic chains to a great extent. The antibacterial assay of PVDF-g-PBIEA-arQPDMAEMA membrane to waterborne E. coli on the membranes showed that quanternized PDMAEMA chains can not only inhibit the growth of the bacteria on the membrane surface but also break the bacterial body into fragments, and filter the bacteria out from the suspension directly.

[1] Y. Uyama, K. Kato, Y. Ikada, Adv. Polym. Sci. 1998, 137, 1. [2] K. Kato, E. Uchida, E. T. Kang, Y. Uyama, Y. Ikada, Prog. Polym. Sci. 2003, 28, 209. [3] B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677. [4] F. Q. Nie, Z. K. Xu, X. J. Huang, P. Ye, J. Wu, Langmuir 2003, 19, 9889. [5] K. Holmberg, F. Tiberg, M. Malmsten, C. Brink, Colloids Surf. A 1997, 123, 297. [6] Y. G. Ko, Y. H. Kim, K. D. Park, H. J. Lee, W. K. Lee, H. D. Park, S. H. Kim, G. S. Lee, D. J. Ahn, Biomaterials 2001, 22, 2115. [7] A. Bessieres, M. Meireles, R. Coratger, J. Beauvillain, V. Sanchez, J. Membr. Sci. 1996, 109, 271. [8] M. L. Yeow, Y. T. Liu, K. Li, J. Appl. Polym. Sci. 2004, 92, 1782. [9] J. Mueller, R. H. Davis, J. Membr. Sci. 1996, 116, 47. [10] C. W. Sutton, C. H. Wheeler, U. Sally, J. M. Corbett, M. J. Dunn, Electrophoresis 1997, 18, 424. [11] K. Jian, P. N. Pintauro, R. Ponangi, J. Membr. Sci. 1996, 117, 117. [12] F. A. Banat, J. Simandl, J. Membr. Sci. 1999, 163, 333. [13] S. Hietala, S. Holmberg, M. Karjalainen, J. Nasman, M. Paronen, R. Serimaa, F. Sundholm, S. Vahvaselka, J. Mater. Chem. 1997, 7, 721. [14] B. Mattsson, H. Ericson, L. M. Torell, F. Sundholm, J. Polym. Sci., Part A: Polym. Chem. 1999, 37, 3317. [15] T. Tarvainen, B. Svarfvar, S. Akerman, J. Savolainen, M. Karhu, P. Paronen, K. Jarvinen, Biomaterials 1999, 20, 2177. [16] J. F. Hester, P. Banerjee, Y. Y. Won, A. Akthakul, M. H. Acar, A. M. Mayes, Macromolecules 2002, 35, 7652. [17] S. Holmberg, P. Holmlund, C. E. Wilen, T. Kallio, G. Sundholm, F. Sundholm, J. Polym. Sci., Part A: Polym. Chem. 2002, 40, 591. [18] J. H. Dai, M. L. Bruening, Nano Lett. 2002, 2, 497. [19] M. Yoshinari, Y. Oda, T. Kato, K. Okuda, Biomaterials 2002, 22, 2043. [20] G. Q. Zhai, E. T. Kang, K. G. Neoh, Macromolecules 2004, 37, 7240. [21] L. Cen, K. G. Neoh, E. T. Kang, Langmuir 2003, 19, 10295. [22] Z. L. Shi, K. G. Neoh, E. T. Kang, Biomaterials 2005, 26, 501. ß 2005 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

981

982


[23] J. F. Moulder, W. F. Stickle, P. E. Sobol, K. Bomben, ‘‘The Handbook of X-ray Photoelectron Spectroscopy’’, 2nd edition, Perkin-Elmer Corporation (Physical Electronics), Wellesley, MA, USA 1992, pp. 216–217. [24] S. B. Lee, R. R. Koepsel, S. W. Morley, K. Matyjaszewski, Y. J. Sun, A. J. Russell, Biomacromolecules 2004, 5, 877. [25] Y. Q. Shen, S. P. Zhu, F. Q. Zeng, R. Pelton, Macromolecules 2000, 33, 5399. [26] S. B. Lee, A. J. Russell, K. Matyjaszewski, Biomacromolecules 2003, 4, 1386. [27] Y. H. Ma, Y. Q. Tang, N. C. Billingham, S. P. Armes, A. L. Lewis, Biomacromolecules 2003, 4, 864. [28] B. Zhao, W. J. Brittain, Macromolecules 2000, 33, 8813. [29] S. Creutz, P. Teyssie, R. Jerome, Macromolecules 1997, 30, 6. [30] X. Bories-Azeau, S. P. Armes, Macromolecules 2002, 35, 10241.

Macromol. Biosci. 2005, 5, 974–982

www.mbs-journal.de

[31] K. L. Robinson, M. A. Khan, M. V. de Paz Banez, X. S. Wang, S. P. Armes, Macromolecules 2001, 34, 3155. [32] M. M. Ali, D. H. Sto¨ver, Macromolecules 2004, 37, 5219. [33] S. Perrier, D. M. Haddleton, Macromol. Symp. 2002, 182, 261. [34] N. V. Tsarrvsky, T. Pintauer, K. Matyjaszewski, Macromolecules 2004, 37, 9768. [35] K. L. Tan, L. L. Woon, H. K. Wong, E. T. Kang, K. G. Neoh, Macromolecules 1993, 26, 2832. [36] J. C. Tiller, C. J. Liao, K. Lewis, A. M. Klibanov, Proc. Natl. Acad. Sci. 1998, 11, 5981. [37] C. Z. S. Chen, N. C. Beck-Tan, P. Dhurjati, T. K. van Dyk, R. A. LaRossa, S. L. Cooper, Biomacromolecules 2000, 1, 473. [38] S. Block, ‘‘Disinfection, Sterilization, and Preservation’’, 3rd edition, Lea & Febiger, Philadelphia, PA 1983. [39] C. J. Waschinski, J. Tiller, Biomacromolecules 2005, 6, 235.
