Plasma-Enhanced Synthesis of Bactericidal Quaternary Ammonium

8 finish, 0.30 in. thickness) sheet was purchased from McMaster-Carr ..... a result of the unidirectional, mechanical polishing process used during ma...
0 downloads 0 Views 1MB Size
Download PDF
Langmuir 2008, 24, 8583-8591

8583

Plasma-Enhanced Synthesis of Bactericidal Quaternary Ammonium Thin Layers on Stainless Steel and Cellulose Surfaces Soujanya N. Jampala,*,† M. Sarmadi,†,‡ E. B. Somers,§ A. C. L. Wong,†,§ and F. S. Denes†,| Materials Science Program, Departments of EnVironment, Textiles & Design and of Biological Systems Engineering, and Food Research Institute, UniVersity of WisconsinsMadison, Madison, Wisconsin 53706 ReceiVed February 6, 2008. ReVised Manuscript ReceiVed June 5, 2008 We have investigated bottom-up chemical synthesis of quaternary ammonium (QA) groups exhibiting antibacterial properties on stainless steel (SS) and filter paper surfaces via nonequilibrium, low-pressure plasma-enhanced functionalization. Ethylenediamine (ED) plasma under suitable conditions generated films rich in secondary and tertiary amines. These functional structures were covalently attached to the SS surface by treating SS with O2 and hexamethyldisiloxane plasma prior to ED plasma treatment. QA structures were formed by reaction of the plasmadeposited amines with hexyl bromide and subsequently with methyl iodide. Structural compositions were examined by electron spectroscopy for chemical analysis and Fourier transform infrared spectroscopy, and surface topography was investigated with atomic force microscopy and water contact angle measurements. Modified SS surfaces exhibited greater than a 99.9% decrease in Staphylococcus aureus counts and 98% in the case of Klebsiella pneumoniae. The porous filter paper surfaces with immobilized QA groups inactivated 98.7% and 96.8% of S. aureus and K. pneumoniae, respectively. This technique will open up a novel way for the synthesis of stable and very efficient bactericidal surfaces with potential applications in development of advanced medical devices and implants with antimicrobial surfaces.

Introduction A high incidence of microbial contamination and infections is a major concern in all existing and evolving technologies of medicine and biology. Nosocomial1 and device-associated2 infections have been reported to be the leading causes of mortality. The propensity toward infection is directly related to bacterial colonization and biofilms on surfaces. Staphylococcus aureus, commonly found on human skin and hair, colonizes on intravascular devices, catheters, sutures, and orthopedic devices.3 Another common bacterium, Klebsiella pneumoniae, causes serious epidemic and endemic nosocomial infections including pneumonia and urinary tract (cystitis), wound, burn, and intraabdominal infections.4 The permanent need to combat infections drives interest in preventive approaches rather than curative therapies by designing surfaces that resist bacterial adhesion and growth and decrease the potential for biofilm development. Cationic compounds such as quaternary ammonium (QA) salts are commonly deployed as disinfectants and are known to be effective against a wide variety of Gram-positive and Gramnegative bacteria,5 but the use of these antibacterial agents is limited due to residual toxicity and irritation to skin. To overcome the limitations of residual toxicity and to increase the lifetime of antibacterial activity, various researchers have synthesized * To whom correspondence should be addressed. E-mail: soujanya_ [email protected]. Phone: (630) 320 4128. Fax: (630) 320 4519. † Materials Science Program. ‡ Department of Environment, Textiles & Design. § Food Research Institute. | Department of Biological Systems Engineering.

(1) Weinstein, R. A. Emerging Infect. Dis. 1998, 4, 416–420. (2) Mayhall, C. G. Hospital Epidemiology and Infection Control, 3rd ed.; Lippincott Williams & Wilkins: Philadelphia, 2004; p 2060. (3) Costerton, J. W.; Stewart, P. S.; Greenberg, E. P. Science 1999, 284, 1318– 1322. (4) Jarvis, W. R.; Munn, V. P.; Highsmith, A. K.; Culver, D. H.; Hughes, J. M. Infect. Control 1985, 6, 68–74. (5) Gilbert, P.; Moore, L. E. J. Appl. Microbiol. 2005, 99, 703–715.

antibacterial polymers with organic cations,6 and these structures are often present as copolymers.7,8 Polymeric bactericides are reported to be more potent than their monofunctional counterparts.9 Another alternative is to covalently immobilize the surface with an antimicrobial compound that does not leach into the system.10–12 Other architectures with QA moieties such as selfassembled monolayers,13 polyelectrolyte multilayers,14 dendrimers,15 and long-chain amphiphiles10,16 have also been investigated. Some of the QA-based bactericidal structures consist of pyridinium head groups. Kawabata and Nishiguchi17 first studied systems containing poly(N-benzyl-4-vinylpyridinium halide) and found them to be highly effective against Gram-positive bacteria. Tiller and co-workers11,18 immobilized N-alkylated poly(4vinylpyridinium) on glass slides, nanoparticles, and other commodity polymers. The polycationic surfaces were shown to kill both airborne and waterborne bacteria. UV-induced surface graft copolymerization was used to graft 4-vinylpyridine groups (6) Ohta, S.; Misawa, Y.; Miyamoto, H.; Makino, M.; Nagai, K.; Shiraishi, T.; Nakagawa, Y.; Yamato, S.; Tachikawa, E.; Zenda, H. Biol. Pharm. Bull. 2001, 24, 1093–1096. (7) Li, G.; Shen, J.; Zhu, Y. J. Appl. Polym. Sci. 1996, 62, 2247–2255. (8) Kenawy, E.; Mahmoud, Y. G. Macromol. Biosci. 2003, 3, 107–116. (9) Kenawy, E.-R.; Worley, S. D.; Broughton, R. Biomacromolecules 2007, 8, 1359–1384. (10) Abel, T.; Cohen, J. I.; Engel, R.; Filshtinskaya, M.; Melkonian, A.; Melkonian, K. Carbohydr. Res. 2002, 337, 2495–2499. (11) Tiller, J. C.; Liao, C.; Lewis, K.; Klibanov, A. M. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 5981–5985. (12) Dong, B.; Manolache, S.; Somers, E. B.; Wong, A. C. L.; Denes, F. S. J. Appl. Polym. Sci. 2005, 97, 485–497. (13) Hayward, J. A.; Chapman, D. Biomaterials 1984, 5, 135–142. (14) Li, Z.; Lee, D.; Sheng, X.; Cohen, R. E.; Rubner, M. F. Langmuir 2006, 22, 9820–9823. (15) Chen, C. Z.; Beck-Tan, N. C.; Dhurjati, P.; Van Dyk, T. K.; LaRossa, R. A.; Cooper, S. L. Biomacromolecules 2000, 1, 473–480. (16) Haldar, J.; Kondaiah, P.; Bhattacharya, S. J. Med. Chem. 2005, 48, 3823– 3831. (17) Kawabata, N.; Nishiguchi, M. Appl. EnViron. Microb. 1988, 54, 2532– 2535. (18) Tiller, J. C.; Liao, C.; Lewis, K.; Klibanov, A. M. Biotechnol. Bioeng. 2002, 49, 465–471.

10.1021/la800405x CCC: $40.75  2008 American Chemical Society Published on Web 07/22/2008

8584 Langmuir, Vol. 24, No. 16, 2008

on polymeric surfaces, and subsequent alkylation rendered them lethal on contact with Escherichia coli.19 Another category of QA compounds contain cationic nitrogen in alkyl chains. Carbohydrate-based substrates were covalently bound to cationic agents with lipophilic adjuncts to impart antimicrobial properties.10 The polyammonium units were obtained by reacting 1,4diazobicyclo[2.2.2]octane with haloalkanes. Coatings based on poly(ethyleneimine) were developed by adjusting their hydrophobicity and positive charge. The polymeric chains were alkylated to increase the number of QA groups attached to a wide range of surfaces.20,21 Most of the methods involve wet-bench chemical reactions, and depending on the type of substrates used, the surfaces need to be functionalized prior to the grafting or synthesis of QA groups. Plasma-enhanced surface functionalization is a relatively cost-effective and environmentally friendly route to modify both inorganic and organic surfaces in a structure- and functionalitycontrolled fashion without altering the bulk properties of a material. The nature of plasma-deposited films is independent of the chemical nature of the substrate and can be deposited on almost any solid substrate, including hard and soft materials. However, the adhesion of the deposited plasma layer depends on the chemical and morphological surface properties. The reactive sites generated on the surface can then initiate subsequent in situ or ex situ surface derivatization reactions for covalent attachment of desired molecules (low or high molecular weight). Previous efforts in our group12,22,23 have successfully demonstrated the grafting of other antibacterial moieties such as PEG and silver using plasma depositions or plasma-mediated crosslinking. The aim of this study is to provide a simple and efficient way to functionalize stainless steel (SS) and cellulose-based filter paper with QA groups. SS is a preferred material for certain implants due to its mechanical properties and excellent corrosion resistance. Filter paper was used in this study as a reference for porous cellulosic materials used in medical textiles. In the present work, we report novel bottom-up chemical synthesis of QA groups using cold plasma technology. The substrate surfaces were functionalized using ethylene diamine (ED) plasma-polymerized films attached covalently to the SS surface via an intermediate layer deposited in O2 and hexamethyldisiloxane (HMDSO) plasma. The synthesis of QA groups was carried out by subsequent ex situ reaction of an ED plasma-deposited film with hexyl bromide and further methylated in methyl iodide. The bactericidal properties of modified surfaces were investigated against Grampositive S. aureus and Gram-negative K. pneumoniae.

Experimental Methods Materials. A SS (type 316, no. 8 finish, 0.30 in. thickness) sheet was purchased from McMaster-Carr (Chicago, IL) and cut into circular stamps of 1 in. diameter. The stamps were washed in hot alkaline detergent (Micro; International Products, Trenton, NJ) for 30 min followed by multiple rinses with acetone and deionized water and air-drying at room temperature. Whatman grade 5 filter papers of 1 in. diameter were used. All the reagents were obtained from Sigma-Aldrich (St. Louis, MO) and were used without further purification. Oxygen and argon were used as plasma gases and were supplied by Linde Gas. (19) Cen, L.; Neoh, K. G.; Kang, E. T. Langmuir 2003, 19, 10295–303. (20) Lin, J.; Qiu, S.; Lewis, K.; Klibanov, A. M. Biotechnol. Prog. 2002, 18, 1082–1086. (21) Lin, J.; Murthy, S. K.; Olsen, B. D.; Gleason, K. K.; Klibanov, A. M. Biotechnol. Lett. 2003, 25, 1661–1665. (22) Jiang, H.; Manolache, S.; Wong, A. C.; Denes, F. S. J. Appl. Polym. Sci. 2006, 102, 2324–2337. (23) Denes, A. R.; Somers, E. B.; Wong, A. C. L.; Denes, F. S. U.S. Patent 6,096,564, 2000; U.S. Patent 19,990,525, 2000.

Jampala et al. Preparation of Surface Layers with QA Groups. The overall scheme of functionalizing SS and filter paper surfaces with positively charged QA groups is shown in Figure 1. Plasma Surface Functionalization. Plasma treatments were carried out in a custom-built capacitively coupled parallel plate reactor described previously.24 In a typical experiment, the reactor was cleaned with oxygen plasma (300 mTorr, 300 W, 10 min) prior to the experiments. The clean SS stamps were placed symmetrically on the powered (lower) electrode, and the reactor was evacuated to its base pressure level (1 mTorr). The desired working pressure of the gases was established using the gas metering valve and pump throttle valve. The plasma glow discharge was then ignited by a radio frequency (rf) power source and sustained for the desired length of time. At the end of the deposition, the chamber was pumped to the base pressure followed by repressurizing it to atmospheric conditions using argon gas. The substrates were then removed and stored in a desiccator until subsequent analysis. SS substrates were pretreated with oxygen plasma followed immediately by HMDSO plasma deposition. The residual gases were pumped out, and ED plasma films were deposited. The oxygen and HMDSO plasma treatments facilitate the adhesion of ED plasmapolymerized films on SS. The ED plasma films deposited directly without the intermediate HMDSO plasma layer delaminated significantly in water. The gas pressure, rf power, and treatment time used for plasma depositions are as follows: oxygen plasma, 300 mTorr, 300 W, 5 min; HMDSO plasma, 200 mTorr, 200 W, 1 min; ED plasma, 100 mTorr, 100 W, 10 min. The optimal conditions selected for the ED plasma polymerization process were studied previously and are reported elsewhere.24 Filter paper was treated directly with ED plasma without the intermediate layer. The hydroxyl groups in the cellulose backbone are reactive functional groups and can be used for covalent attachment of ED plasma film. The paper substrates were placed on the powered electrode and treated with ED plasma at a 100 mTorr gas pressure, 50 W, and 10 min. Quaternization of Plasma-Polymerized Films. The plasma-treated substrates were immersed in a 30 mL solution of tert-amyl alcohol containing 10 vol % hexyl bromide and 0.2 g of KOH and were stirred for at least 12 h at 70 °C.25 The substrates were removed, thoroughly rinsed with methanol, and dried in a vacuum oven. The substrates were then placed in a sealed glass vial containing 10 vol % methyl iodide in tert-amyl alcohol maintained at 50 °C for 6 h.20 The substrates were rinsed with methanol and water and vacuumdried. Quaternized SS stamps and filter paper were stored in a sealed Petri dish before further analysis. Surface Analysis. Electron Spectroscopy for Chemical Analysis (ESCA). Surface layers were characterized by ESCA on a PerkinElmer Physical Electronics Phi 540 small-area spectrometer (Mg KR source, 15 kV, 300 W, 45° takeoff angle, Palo Alto, CA) with a concentric hemispherical energy analyzer (resolution 0.5 eV). A pass energy of 89.45 eV was used to obtain the survey spectra and elemental composition. High-resolution scans over a binding energy (BE) range of 20 eV were performed with a 35.75 eV pass energy. High-resolution peaks were resolved into individual Gaussian peaks associated with various chemical states of an element and curvefitted with the nonlinear least-squares method using AugerScan software, version 2.4.2 (RBD Enterprises). The surface-charge-origin BE shifts in the spectra were calibrated by sputter coating substrates with gold in a Desk II system sputter coater (pressure 50 mTorr, sputtering current 45 mA, 2 s, Denton Vacuum Inc., Morristown, NJ) and were corrected on the basis of Au 4f7/2 and Au 4f5/2 peaks at 84 and 87.7 eV, respectively. All BE assignments were done using the database for polymeric compounds.26 Fourier Transform Infrared Spectroscopy (FTIR). An ATI-Mattson RS-1 IR (Madison, WI) instrument was used for FTIR measurements performed under a nitrogen blanket. Data were acquired in the (24) Jampala, S. N.; Manolache, S.; Sarmadi, M.; Denes, F. S. J. Appl. Polym. Sci. 2008, 107, 1686–1695. (25) No¨ding, G.; Heitz, W. Macromol. Chem. Phys. 1998, 199, 1637–1644. (26) Beamson, G.; Briggs, D. High-Resolution XPS of Organic Polymers: The Scienta ESCA 300 Database; John Wiley & Sons Ltd.: Chichester, U.K., 1992.

Bactericidal QA Layers on SS and Cellulose Surfaces

Langmuir, Vol. 24, No. 16, 2008 8585

Figure 1. Scheme of functionalization with QA groups on SS and filter paper substrates.

600-4000 cm-1 wavenumber region with 250 scans at a resolution of 0.4 cm-1. Since plasma-induced modification processes involve the top 100 Å surface and FTIR reveals information deeper than 1 µm, the unmodified bulk will dominate the IR signatures. Therefore, potassium bromide pellets pressed with a die (International Crystal Laboratories, Garfield, NJ) and Carver Laboratory (Wabash, IN) press were kept in the plasma reactor for film deposition. For quaternized surfaces, KBr powder was ground on the surface of the functionalized samples and then pressed. Contact Angle Measurement. Static contact angles of deionized water were measured on plasma-deposited and quaternized surfaces at ambient temperature using a Rame´-Hart model 100-00 goniometer (Mountain Lakes, NJ). A microliter syringe is used to dispense 5 µL water droplets on the surface. Contact angles were measured on the opposite edges of at least 10 drops and averaged. Atomic Force Microscopy (AFM). The surface topography of unmodified and modified SS surfaces was evaluated in contact mode by a Pico Scan atomic force microscope (Molecular Imaging Inc., Tempe, AZ) with a scanner area of 6 µm × 6 µm, a scan rate of 5147.9 nm/s, 512 data points per line, and a silicon nitride nanoprobe. The surface roughness was compared on the basis of the mean absolute deviation values. Assays for Bactericidal Efficacy. SS Substrates. Unmodified SS and modified SS were evaluated for their ability to kill attached bacteria. The unmodified SS stamps were sterilized in an autoclave for 30 min prior to the experiments. S. aureus (ATCC no. 6538) and K. pneumoniae (ATCC no. 4352) were grown overnight in 5 mL of trypticase soy broth (TSB; Becton Dickinson, Sparks, MD) at 37 °C. In the case of K. pneumoniae, the culture was centrifuged at 3000 rpm for 10 min, and after the removal of supernatant, the cells were washed twice with 0.01 M phosphate-buffered saline (PBS; pH 7.2) and resuspended in PBS to avoid nutrient carryover in subsequent dilutions. Three 10-fold dilutions of the suspension in PBS were done to achieve initial bacterial concentrations of about 105 to 106 colony-forming units (cfu)/mL. A 60 µL inoculum of either S. aureus or K. pneumoniae was added onto each substrate in a Petri dish, and the sample was incubated at ambient temperature

for 0 and 24 h. For the 0 h samples, the SS stamps were immediately placed in centrifuge tubes containing 10 mL of PBS and sterilized glass beads. The stamps were vortexed for 30 s to remove the bacteria from the surfaces. The resulting suspension was diluted (10-fold or 0-fold) before 100 µL was plated on brain heart infusion agar (BHA) and incubated at 30 °C for 48 h. The number of colonies on the agar plates was counted and expressed as cfu per sample and represented the viable bacteria originally present on the unmodified and modified SS surfaces. The 24 h samples were kept in a closed humidified box to prevent death or injury of bacteria from desiccation. The bacteria counts were performed as described above. Two SS stamps were used in an individual experiment, and each experiment was repeated at least twice with samples from different batches of plasma treatments and quaternization reactions. The detection limit of our assay is 50 cfu/sample. Filter Paper Substrates. Unmodified and modified filter papers were tested for their bactericidal efficacy in accordance with American Association of Textile Chemists and Colorists (AATCC) Test Method 100-2004. The standard protocol was scaled down for the smaller sample sizes of 1 in. diameter, and they were incubated with the bacterial suspension at room temperature instead of 37 °C. Data were analyzed with Student’s t test with differences considered statistically significant at p < 0.05. Antibacterial activity was measured by the “percentage inactivation” of a bacterial species on the surface. At a given time t, inactivation is determined as the ratio of the difference in bacterial counts of control and treated surfaces to the control surface counts.

Results Synthesis of Surfaces with QA Groups. The SS surface was functionalized with ED plasma to generate highly cross-linked macromolecular layers with reactive nitrogen groups, and subsequently cationic QA groups were grafted ex situ. The inertness and lack of any chemical bonds for covalent coupling on SS make it difficult to directly deposit a stable organic film. Prior to the functionalization with ED plasma, a pretreatment

8586 Langmuir, Vol. 24, No. 16, 2008

Figure 2. Survey ESCA spectra of SS treated sequentially with O2 plasma + HMDSO plasma (p-HMDSO), ED plasma (p-ED), n-hexyl bromide (C6-SS), and methyl iodide (C6+C1-SS).

step involving O2 and HMDSO plasma was devised to increase the adhesion of the plasma-polymerized film on the metallic surface. A low-pressure and high-power treatment in O2 plasma is known to thoroughly clean the SS surface27 and ensure good adhesion through active surface sites. HMDSO plasma deposition was carried out immediately following O2 plasma deposition to form an organosilicon film that acts as an intermediate layer and binds covalently to the SS substrate. The ED plasma film is deposited on the HMDSO plasma intermediate layer such that a bioactive structure is stabilized. The films deposited without this plasma pretreatment step of O2 and HMDSO plasma delaminated completely in water. Hence, plasma deposition in this study resulted in covalently attached coatings that are resistant to delamination or leaching in aqueous solution. Amine functionalities deposited via ED plasma polymerization were alkylated using ex situ single-phase reaction with n-hexyl bromide in tert-amyl alcohol.20,25 The surfaces were further methylated in excess methyl iodide to produce QA salts that are attached to the surface. The overall process of N-hexyl substitution and subsequent methylation is referred as “quaternization” in the following sections. Plasma-Enhanced Modification of Surfaces. Survey ESCA spectra for modified SS at each step are compared in Figure 2. SS surfaces treated sequentially with O2 and HMDSO plasma (p-HMDSO in the figure), ED plasma (p-ED), n-hexyl bromide (C6-SS), and methyl iodide (C6+C1-SS) are investigated. The spectra indicate carbon (from core level C 1s), nitrogen (N 1s), oxygen (O 1s), silicon (Si 2p), and iodine (I 3d5/2) on the surface of the films. For quantitative analysis, all elemental concentrations were normalized to carbon. The surfaces treated with O2 plasma and successively with HMDSO plasma have peaks relating to the BE of C 1s, O 1s, and Si 2p. The absence of any major peak assignable to iron or chromium originating from SS attests the complete coverage of the substrates by 1 min of HMDSO plasma treatment. The high-resolution (HR) scan (Figure 3a) was fitted into chemical linkages corresponding to *C-Si at 284.3 eV, (27) Korzec, D.; Rapp, J.; Theirich, D.; Engemann, J. J. Vac. Sci. Technol., A 1994, 12, 369–378.

Jampala et al.

*Si-C at 101.5 eV, *SiOx at 99.8 eV, and *O-Si at 532.0 eV. These functional groups and corresponding plasma-generated free radicals are responsible for covalent attachment of the subsequent ED plasma-polymerized layer deposited on SS. The ED plasma film was found to consist of carbon, oxygen, and nitrogen functionalities. The plasma-deposited film chemistry depends on RF-plasma-electron-induced molecular fragmentation and subsequent recombination of these fragments into a macromolecular matrix on the surface. The N/C ratio is 0.6 and is less than the theoretical value of the precursor, suggesting that the chemical structures deposited in the plasma environment are significantly different from those of conventional polymers from the monomers. The oxygen incorporated in the films appears from the post plasma oxidation processes initiated by plasmagenerated radicals under open laboratory conditions and the background oxygen in the reactor at a 1 mTorr base pressure. The C 1s peak in Figure 3b was fitted into a quadramodal pattern with components positioned at 285.0, 285.6, 286.7, and 288.0 eV corresponding to functionalities of types *C-C and *CHx, carbon singly bonded to nitrogen (*C-NH2, *C-NH-C, *C-NdC, etc.), carbon bonded to nitrogen (*CdN, *CdNdC) and oxygen (*C-OH, *C-O-C), and carbon doubly bonded to oxygen (Cs*CdO, NHxs*CdO), respectively. The HR spectrum of N 1s is fitted with nitrogen from amine (secondary and tertiary) functionalities at 399 and 400.1 eV for *NHxsCdO and *NdC. There were no peaks found corresponding to nitroso, nitro, or nitrate groups, indicating that the nitrogen functionalities are not affected by post plasma oxidation processes. Also, it was difficult to infer the distribution of primary, secondary, and tertiary amine groups from the ESCA spectrum. The O 1s spectrum has peaks associated with the carbonyl group in amides (531.5 eV) and *O-C linkages (533 eV). Quaternization of Plasma-Modified Surfaces. The success of the plasma-enhanced attachment of QA groups on the SS surface can be ascertained by comparing the ESCA spectra of the films before and after the grafting process. There was no peak assignable to bromine from the N-hexyl substitution step as the anion was precipitated out with potassium hydroxide in the reaction. The HR C 1s spectrum of hexylated SS (C6-SS) in Figure 3c was broad, and another peak at 286.2 eV attributed to *C-N+ was added in curve-fitting. In the N 1s core level spectrum, *N+-C was identified at a BE above 400 eV. This confirms the reaction between surface amines and alkyl halides resulted in formation of QA groups. However, the reaction is not very specific, and a mix of tertiary and quaternary groups is formed. Hence, the addition of another alkylation step with methyl iodide improved the yield of cationic nitrogen surface sites. The peak area ratios of N+ moieties in C 1s and N 1s HR spectra increased after methylation. The presence of iodine peaks from various core levels in the survey spectrum of sample C6+C1-SS confirms the reaction between amine groups on the modified surface and methyl iodide. A doublet from I 3d5/2 and I 3d3/2 at 619.6 and 631.0 eV is attributed to the presence of I-. The I 3d5/2 peak in Figure 3d was used to determine the BE shifts associated with surface charging. The QA structures and I- are stable even after the sample is washed with water. The peak areas of *C-N+ increased from 15% of total carbon coverage in C6-SS to 28% in C6+C1-SS. Also, the decrease in the concentration of *C-N was observed in quaternized SS. The N 1s peak was fitted into three peaks with *+N-C centered at 401.8 eV in addition to *NHxsCdO and *NdC linkages. ESCA analyses of quaternized SS samples after being washed in an ultrasonicator were similar to those prior to washing and confirm the stability of polycationic structures on SS.

Bactericidal QA Layers on SS and Cellulose Surfaces

Langmuir, Vol. 24, No. 16, 2008 8587

Figure 3. HR ESCA spectra of SS treated sequentially with O2 plasma + HMDSO plasma (p-HMDSO), ED plasma (p-ED), n-hexyl bromide (C6-SS), and methyl iodide (C6+C1-SS).

FTIR. The surface chemistries of plasma-treated SS surfaces that were subsequently quaternized are compared in Figure 4. All the assignment to the peaks was done using the available literature.28,29 The spectrum of the film deposited in HMDSO plasma indicates the presence of Si-O- and Si-C-based structures. The major and strong regions identified are 720-780 (28) Bellamy, L. J. The Infra-Red Spectra of Complex Molecules, 2nd ed.; Wiley: London, 1958; p 425. (29) Coates, J. In Encyclopedia of Analytical Chemistry; Meyers, R. A., Ed.; John Wiley & Sons Ltd.: Chichester, U.K., 2000; pp 10815-10837.

cm-1 characteristic of Si(CH3)- end groups, 800-850 cm-1 unique to Si-CH3 and Si(CH3)2- stretching vibrations of the Si-C linkage, 1100-1180 cm-1 related to Si-O stretching vibrations, 1260 cm-1 representing CH3 rocking in Si(CH3)x, and 1400-1430 cm-1 corresponding to the deformation mode of (CH)x. The relatively high intensity of the 800-850 cm-1 band zone indicates the presence of a cross-linked structure and that Si(CH3) plasma-generated fragments played a significant role in the formation of the surface-deposited macromolecular surfaces.

8588 Langmuir, Vol. 24, No. 16, 2008

Jampala et al.

Figure 5. AFM images (scan area 6 µm × 6 µm, z-scale 200 nm) of unmodified, O2 and HMDSO plasma-treated (p-HMDSO), ED plasmatreated (p-ED), and quaternized (C6+C1-SS) SS surfaces.

Figure 4. FTIR spectra of O2 and HMDSO plasma-treated (p-HMDSO), ED plasma-treated (p-ED), and quaternized (C6+C1-SS) surfaces. Table 1. Water Contact Angles for Unmodified and Modified SS SS sample

treatment

contact angle (deg)

unmodified p-HMDSO p-ED C6-SS C6+C1-SS

O2 plasma and HMDSO plasma ED plasma N-hexylated methylated

60 ( 2 109 ( 1 11 ( 1 70 ( 8 51 ( 8

ED plasma-deposited films were found with predominant characteristic bands at 3440 cm-1 (N-H secondary amine stretching), 1629 cm-1 (N-H secondary amine deformation), 1105 cm-1 (C-N), and 810 cm-1 (N-H wagging). The absence of two bands usually associated with primary amine groups reveals that the plasma fragmentation and recombination form amine groups only in the form of secondary and tertiary amine functionalities. The band in the range of 3300-3600 cm-1 in the ED plasma was broad and can account for absorption from amides in addition to amines. Weak absorptions from hydrocarbons around 2900 cm-1 were also found. The spectrum of the surfaces after quaternization (N-hexylation and subsequent methylation) consists of strong peaks at 2920 and 1450 cm-1 associated with stretching and deformation vibrations of -CH2- groups in the hexyl chain grafted on the surface. A new strong band at 1250 cm-1 is attributed to QA bonds. All the amine-based signatures from ED plasma-deposited films decreased in intensity, indicating the conversion of amine groups into QA groups. Surface Wettability. A summary of the water contact angles measured on SS surfaces along the plasma-mediated grafting process of QA groups is given in Table 1. The HMDSO plasma film makes the surface extremely hydrophobic with a contact angle of 109°. However, after deposition of the ED plasma film, the surfaces are rendered very hydrophilic due to the presence

of amine and hydroxyl functionalities on the surface. The grafting of hexyl chains increased the contact angle, generating hydrophobic surface sites. Further methylation and formation of QA groups slightly reduced the contact angle, and this change can be explained by the existence of cationic nitrogen atoms attached to the surface. Various researchers30,31 have correlated surface free energies and macroscopic properties such as wettability to bacterial adhesion and biofilm formation. However, bacterial adhesion is dependent on many factors including hydrophobic/ hydrophilic interactions and electrostatic interactions with the cell wall of bacteria. Surface Topography. AFM images of unmodified and modified SS substrates are presented in Figure 5. The unmodified SS exhibits parallel grooves as a result of the unidirectional, mechanical polishing process used during manufacturing. The presence of even sharper or higher grooves is also obvious in the images of oxygen plasma-treated SS substrates. An apparent “polishing” effect can be noted in the 2D image probably due to the surface oxidation/ablation generated by the oxygen and HMDSO plasma species (p-HMDSO). The ED plasma exposure induces a dense, granular-type deposition on SS surfaces both in the valleys and on the peaks (p-ED). It appears that the EDbased plasma species recombine into granular structures on the SS sample surfaces, which is controlled by the distribution of nanoscale surface topographies. Quaternization of ED-modified surfaces amplifies the formation of separated, individual structures, leading to less dense and higher particle dimension morphological formations (C6+C1-SS). Bactericidal Efficacy of Quaternized SS Surfaces. The bactericidal activity of the modified substrates was measured with S. aureus and K. pneumoniae. The assay used here determines the number of viable bacteria remaining after incubation of the substrate surface in contact with a suspension of bacteria for 24 h followed by removal of bound cells by vortexing with glass beads. Due to large differences in water contact angles and surface roughness that may influence antibacterial activity, it was (30) Terada, A.; Yuasa, A.; Kushimoto, T.; Tsuneda, S.; Katakai, A.; Tamada, M. Microbiology 2006, 152, 3575–3583. (31) Kugler, R.; Bouloussa, O.; Rondelez, F. Microbiology 2005, 151, 1341– 1348.

Bactericidal QA Layers on SS and Cellulose Surfaces

Langmuir, Vol. 24, No. 16, 2008 8589

Figure 6. Bactericidal activity against S. aureus of SS surfaces treated successively with O2 plasma + HMDSO plasma (p-HMDSO), ED plasma (p-ED), n-hexyl bromide (C6-SS), and methyl iodide (C6+C1-SS). ND* ) not detectable.

Figure 8. Influence of the contact time on the bactericidal efficacy of quaternized SS. ND* ) not detectable. Figure 7. Bactericidal activity against K. pneumoniae of SS surfaces treated successively with O2 plasma + HMDSO plasma (p-HMDSO), ED plasma (p-ED), n-hexyl bromide (C6-SS), and methyl iodide (C6+C1SS).

necessary to investigate the activity of surfaces at all the steps in the grafting process: O2 and HMDSO plasma treatment, ED plasma treatment, N-hexylation, and methylation. The bacterial numbers of unmodified SS at 0 and 24 h and that of modified SS at 0 h were found to be similar. Quaternized surfaces (C6+C1SS) exhibited high bactericidal activity against S. aureus. No viable cells were recovered after 24 h of incubation, and at least a 3-log decrease was observed (Figure 6). The N-hexylated surfaces without subsequent methylation also reduced the bacteria counts by 91%. The killing effect increased with an increase in the concentration of positive nitrogen groups (as determined from ESCA) on the surfaces. Hence, the additional step of reacting surfaces with methyl iodide significantly improved the bactericidal efficacy of the surfaces. The bacteria numbers on unmodified and HMDSO and ED plasma-treated SS were similar, and the changes in surface chemistry after plasma treatments did not affect the survival of bacteria on SS substrates. The modified surfaces were not as effective in killing K. pneumoniae (Figure 7) as S. aureus. The quaternized surfaces with and without methylation reduced the bacteria by 98% and 87.5%, respectively. A time-dependent experiment was performed to determine the activity at various intervals over the course of a 24 h period. Samples were incubated for 0, 1, 2, 4, 6, 12, and 24 h, and the influence of the contact time is shown in Figure 8. The quaternized surfaces incubated with S. aureus showed bactericidal properties in the first hour, and bacteria were not detectable by 4 h. The surfaces reached their optimal activity against K. pneumoniae between 6 and 9 h. Bacterial cell counts on the agar plates were

constant after 24 and 48 h, suggesting that the cationic structures on the surfaces caused death and not injury of the cell. Hence, a contact time of 6 h was enough to kill the bacteria. Bactericidal Efficacy of Quaternized Filter Paper. The grafting process of QA groups was applied to commercial filter paper (Whatman grade 5). ESCA spectra shown in Figure 9 reveal the linkages in unmodified and modified paper surfaces. The unmodified paper has a cellulose structure and consists of *C-C, *C-O, *CdO, and *C(O)dO. The aliphatic and carboxyl groups originate from surface contamination and uronic acids, respectively. ED plasma-treated and quaternized filter paper surfaces have structures similar to those of the modified SS surfaces. The ESCA C 1s peak of the ED plasma film on paper was fitted with a quadramodal pattern consisting of *C-C, *C-O, *CsN/CdN, and *CdO. The presence of *C-N+ (Figure 9c) in quaternized paper confirms the success of the QA grafting procedure on filter paper. Figure 10 shows the bactericidal activity of quaternized filter paper. The N-hexylated ED plasma-treated and subsequently methylated surfaces showed bactericidal action, reducing the concentration of S. aureus by 98.7%. This confirms the applicability of the technique to high surface area porous materials. The efficacy of the bactericidal action was found to be 96.8% for K. pneumoniae. The decrease in the efficacy of the modified paper against S. aureus and K. pneumoniae in comparison to that of SS surfaces can be explained by the porous nature of the substrates. The interstitial spaces in filter paper are large enough for bacteria to penetrate and get trapped along the thin fibers. Under RF low-density plasma chemistry conditions, coating of the interior structures of the pore surfaces results in a nonuniform (e.g., island-type) deposition. Plasma-enhanced coating of porous structures or deep cavities results in uniform cavity-surface

8590 Langmuir, Vol. 24, No. 16, 2008

Jampala et al.

Figure 10. Bactericidal activity of unmodified and quaternized filter paper against S. aureus and K. pneumoniae.

Figure 9. HR ESCA spectra from the C 1s core level of (a) unmodified paper, (b) ED plasma-treated paper, and (c) quaternized paper with hexyl bromide and methyl iodide.

coatings usually by using high-density plasma (ECR, helicon, inductive) environments. The unmodified surface areas might be responsible for a less efficient antibacterial activity.

Discussion While there are many studies examining how to synthesize QA groups using wet chemistry reactions, top-down film deposition techniques have been less frequently investigated. The plasma-enhanced route to graft QA groups on metallic and cellulosic substrates provides an efficient method of synthesis that can be employed potentially on any solid porous or nonporous substrate including polymers, ceramics, and metals. The grafting procedure of reactive amine groups developed in this study was done in 30 min, comparable to other methods of functionalization.11,32 Perhaps the most important observation from antibacterial assays is that the modified SS and paper surfaces are highly effective in killing S. aureus and K. pneumoniae. Our results show that QA-grafted surfaces are more effective against S. aureus (32) Zhang, H.; Ru¨he, J. Macromolecules 2003, 36, 6593–6598.

than K. pneumoniae. Previous work20 done with glass slides derivatized with n-hexylpoly(ethyleneimine) exhibited 90% kill against S. aureus. The surfaces decreased the number of Pseudomonas aeruginosa and E. coli cells by 97% and 96%, respectively. Most surface-bound QA groups have been demonstrated with reduction percentages between 90% and 99%.10,11,19 The outermost surface of bacterial cells universally carries a net negative charge. We hypothesize that the surface-bound QA groups have a mode of action similar to that of cationic antibacterial agents. The cationic QA groups have a high binding affinity for bacterial cells and rapidly adsorb on the bacterial surface by electrostatic and hydrophobic interactions. These can then diffuse through the cell wall, bind to the cytoplasmic membrane, and cause disruption. The release of protons and eventual precipitation of the cell contents are attributed to structural disorganization of the cell.5,33 Other theories relating to QA groups suggest reaction of cationic agents with phospholipids as a cause of membrane distortion.34,35 The cationic agents were also reported to alter the PMF (proton-motive force), affecting cell metabolism.36 The efficacy of QA groups in disrupting cell membranes depends on the positive charge and hydrophilic-lipophilic balance. N-hexylation of the ED plasma-deposited film not only makes the surface hydrophobic but also converts secondary and tertiary amine groups in the plasma film into permanently cationic QA groups. By further extending the hypothesis that concentration is a key issue in lethal effects, the number of QA groups on the surfaces was boosted by methylating the N-hexylated plasmadeposited film, which indeed enhanced the overall bactericidal (33) Salton, M. R. J. J. Gen. Physiol. 1968, 52, 227–252. (34) Corbal, J. P. S. Can. J. Microbiol. 1991, 38, 115–123. (35) McDonnell, G.; Russell, A. D. Clin. Microbiol. ReV. 1999, 12, 147–179. (36) Russell, A. D.; Chopra, I., Understanding Antibacterial Action and Resistance, 2nd ed.; Ellis Horwood: Chichester, England, 1996.

Bactericidal QA Layers on SS and Cellulose Surfaces

activity of substrates with similar surface areas. In our work, the SS surface deposited with ED plasma and subsequently quaternized killed at least 99.9% of the S. aureus cells inoculated on the surface, and this capability is essential in combating microbial biofilms. Any bacterial cell that remains on the surface of a device adheres and embeds itself into the extracellular polymeric matrix, eventually causing serious infections. Thus, surfaces based on preventive strategies can be beneficial. The filter paper is pure cellulose and a natural polymer. We have also demonstrated that the same grafting procedure can be used for polymeric substrates. By changing the plasma deposition conditions, the ED plasma film can be laid directly onto polymeric surfaces without the need for an intermediate layer for covalent attachment. The modified surfaces were more biocidal toward Grampositive S. aureus than Gram-negative K. pneumoniae. It is known that Gram-positive bacteria have a cell wall composed of a thick porous peptidoglycan layer, which allows antibacterials to diffuse through the cell wall, but Gram-negative bacteria have an additional outer membrane with narrow restrictive channels that functions as a barrier to foreign molecules such as antibacterial agents. This makes the binding to the cytoplasm extremely slow and is a reason for the long contact times needed for Gramnegative bacteria with QA-bound surfaces.37 Resistance development by pathogens is the crucial limitation of existing antimicrobial agents. Hydrophobic cations such as QA groups can cause resistance via microbial MDR (multidrug resistance) pumps. Lewis and Klibanov38 reported cations attached (37) Nikaido, H. Science 1994, 264, 382–388. (38) Lewis, K.; Klibanov, A. M. Trends Biotechnol. 2005, 23(7), 343–348. (39) Jampala, S. N., Ph.D. Thesis, University of Wisconsin-Madison, 2007.

Langmuir, Vol. 24, No. 16, 2008 8591

to polymeric or macromolecular chains are not subject to efflux by MDRs. Surfaces modified with N-hexylpoly(vinylpyridinium) were examined against mutant strains of S. aureus. The grafting of QA groups in our study is done in a macromolecular matrix of a plasma-deposited film, and a similar argument against development of bacterial resistance can be extended here.

Conclusions We have demonstrated the development of surface layers that kill bacteria on contact using cold plasma techniques. The successful grafting of QA groups on SS and filter paper surfaces rendered them bactericidal against S. aureus and K. pneumoniae. Also, the importance of having a high concentration of N+ groups to achieve high potency to bacteria was realized. The efficacy of QA groups depends on the length of the alkyl chain and has been reported elsewhere.39 The surfaces with hexyl chains were most potent against S. aureus, and the ones reacted with dodecyl bromide exhibited maximum efficacy against K. pneumoniae. Hence, nonreleasing surface immobilization of QA groups in high densities substantially increases the bactericidal activity and resolves the disadvantages of conventional QA compounds used in aqueous solutions. Acknowledgment.WegratefullyacknowledgeSorinManolache at the University of WisconsinsMadison for assistance with the plasma experiments and Prof. Karen Leonas at the University of Georgia, Athens, for helpful suggestions. This work was supported by Hatch Fund S1026. LA800405X