Scalable Synthesis of Self‐Disinfecting Polycationic Coatings for Hospital Relevant Surfaces

The prevention of microbial infections is a global challenge. Efficient antimicrobial coatings that rapidly kill microorganisms upon contact can help minimize their transmission. However, their scalable synthesis is challenging. This work demonstrates the scalable synthesis and characterization of self‐disinfecting nanofilms for the postmodification of hospital‐relevant surfaces. Their antimicrobial action is based on charge interactions between a supercharged cationic surface film and the negatively charged bacteria membrane. Photoinitiated bulk polymerization of an air‐dried [2‐(methacryloyloxy)ethyl]trimethylammonium chloride film on cotton (gowns), nitrile rubber (protective gloves), and glass surfaces (tables, screens) is used for their supercharging, and studied with streaming potential measurements. A 6 nm thick coating dominated by cationic quarternary amine groups is shown by a combination of spectroscopic imaging ellipsometry and X‐ray photoelectron spectroscopy. Antimicrobial in vitro evaluation of the coated surfaces demonstrates up to ≈4 log reductions in bacterial populations in less than 5 min. Confocal laser scanning microscopy and live‐dead staining confirm the surface‐induced killing of bacteria. The coating's range of compatible materials and its rapid bactericidal activity can combat the surface transmission of bacteria and may help to contain the spread of infectious diseases. Its synthesis in environmental conditions is promising for integration into industrial processes.


Introduction
Infectious diseases burden our society significantly by causing 20% of all fatalities worldwide. [1] Their prevention, especially in healthcare settings, remains a global challenge for humanity, particularly given the rise in antimicrobial resistance, The bacterial cell membrane is the main target for an antibacterial coating. The membrane's major components are amphiphilic and mostly negatively charged phospholipids and lipopoly saccharides. [7] These phospholipids self-assemble into a bilayer that separates the bacteria cytoplasm from their surroundings and is responsible for maintaining the osmotic pressure, the difference in ionic strength, and the transport of specific molecules in and out of the cell. [8] The hydrophilic phosphate groups on both sites of the membrane protect the hydrophobic alkyl chains in the interior from energetically unfavorable contact with water molecules. [9] Charge attractions between the negatively charged bilayer membrane and a positively charged surface may modify the curvature of the bilayer, leading to a loss in the barrier function of the membrane and, eventually, the death of the bacteria. Indeed, the most effective compounds for contact-killing coatings have been cationic. [10] However, to the best of our knowledge, the scalable synthesis of a biocompatible antibacterial coating with >log 4 reductions within less than 5 min for materials such as cotton, glass, and nitrile rubber, based on molecules that are inactive upon leaching, has not yet been achieved.
Various strategies for antimicrobial surface modification have been researched, including the adsorption of nanoparticles, covalent linking of polymers, and incorporation of small antimicrobial molecules are the most common. [11] For cotton and natural or latex rubber, the noncovalent integration of silver and copper nanoparticles or small molecular weight antibacterial molecules such as Triclosan allows antibacterial functionalization. [12] However, metallic nanoparticles are correlated with adverse effects on human health and negative environmental impacts. [13] Furthermore, the leaching of the small molecule antibiotics or metal nanoparticles from the treated surfaces can result in sub-lethal concentrations of these bioactive in the local environment, which may lead to the development of antibiotic resistance. [12h] Thus, the covalent attachment of the antibacterial functionality to surfaces is an attractive strategy to limit these adverse effects.
The covalent attachment of cationic moieties, amphiphilic zwitterionic polymers, or polymers functionalized with photosensitizers able to produce reactive oxygen species was reported to provide cotton fabrics with antibacterial properties. [14] More recently, through complex synthetic chemical steps, research groups could produce dense cationic and hydrophobic polymer brushes on various surfaces. [15] The use of protective group chemistry allows for achieving highly positive coatings on glass that showed a 100% antibacterial activity (beyond the limit of detection) within 5 min when bacteria are sprayed on the surface. [10a] However, the multiple-step synthesis under an inert atmosphere and the use of strong organic solvents limit their upscaling to commercial production. Hence, a cheap, simple, and scalable coating process is required that can be applied on various surfaces. Developing an antimicrobial coating process mainly based on water as a solvent, scalable to large surface areas, easy to implement in a production line, and applicable to existing high-touch surfaces is pivotal to limiting surface-borne pathogen transmission on a global scale.
Direct plasma coating of a hydrophobic cationic polymer allowed a simple one-step process to render glass antimicrobial. [16] An alternative approach for polycationic surface modification is through radical polymerization, widely used in industry and scalable to large quantities. [17] Monomers suitable for radical polymerization include those with a C=C double bond, allowing the addition of a monomer to the radical-containing chain by opening the π-bond and transferring the radical to the added monomer. [18] [2-(Methacryloyloxy)ethyl] trimethylammonium chloride (METAC) is a cationic monomer that fits these criteria. [19] However, the charge repulsions between growing polycationic polymers during the reaction in water compromise the resulting surface charge density that is required for fast antimicrobial action.
This work shows a simple and versatile method for synthesizing an antibacterial coating at very high charge density with contact-killing antibacterial properties. In this process, METAC films on silica, cotton, and nitrile are polymerized with UV light under dried conditions to maximize the METAC density and surface charge. Coating formation was characterized using spectroscopic imaging ellipsometry, contact angle measurements, X-ray photoelectron spectroscopy (XPS), and streaming potential analysis. Bactericidal activity was demonstrated in vitro against Gram-negative E. coli and the Gram-positive S. aureus bacteria with surface contact antibacterial assays. The results showed that the developed method for METAC coating combines scalable chemistry and ease of application with rapid contact-killing of bacteria on a time scale relevant for high-touch surfaces.

METAC Coating
The coating procedure for nitrile, cotton, and silica surfaces consisted of i) dip-coating an aqueous METAC monomer solution containing the photoinitiator lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP), followed by ii) drying for 5 min at 50 °C, iii) UV irradiation at 300 nm, and iv) washing with water for 5 min to remove potential nonbound compounds ( Figure 1A). The coating was covalently immobilized on nitrile films through the reaction of the METAC with the nitrile C=C double bonds. Cotton and Si surfaces required an additional pretreatment step before coating. The significant component of cotton is cellulose, a biopolymer consisting of β(1→4) linked d-glucose. [20] The cotton material was first soaked in the photoinitiator (1 wt% in water) for 12 h to ensure a high number of radicals under light irradiation at the cotton surface for the reaction with the METAC. The cotton radicals are most likely formed by abstracting hydrogen from a carbon in the glucose units in cellulose. [21] The Si surfaces were reacted for 18 h at 25 °C with a triethoxyvinylsilane, allowing its attachment by silane condensation. [22] The vinyl's C=C double bond could react with the METAC. Further details on the specific methods are given in the Experimental Section. For each surface, 20 individual samples were coated; this was required to allow all the measurements. It also proved the reproducibility of the process.
The coating protocol was optimized by studying the effect of the LAP content on the coating thickness and chemistry. The attachment of the METAC coating onto Si-wafer surfaces was first evaluated with contact angle measurements. The water contact angle on the vinylsilane-coated Si-wafers in the absence www.advmatinterfaces.de of METAC was 40.5° ± 2.9° (Figure 2A and Figure S2, Supporting Information). Upon coating with METAC, the contact angle was modified to 16.6° ± 0.3°, 12.2° ± 1.4°, and 22.1° ± 0.2° at 0.01, 0.2, and 0.5 wt % LAP, respectively. The initial decrease in the contact angle upon increasing the LAP concentration from 0.1 to 0.2 wt% in the METAC solution may result from a thicker surface coating with higher METAC density, resulting in higher hydrophilicity. The following decrease in contact angle at 0.5 wt % LAP indicates an inhomogeneous coating under these reaction conditions where parts of the hydrophobic vinylsilane-coated Si surface are exposed to the water droplet.
The thickness and homogeneity of the resulting METAC coating on the Si-wafer were further investigated by imaging ellipsometry ( Figure 2B-E). This method allows imaging of the thickness of the film at the sub-nanometer resolution, with a lateral resolution of about 1 µm 2 . The coating at 0.01 wt% LAP showed an average thickness of around 5.8 ± 0.4 nm that was relatively homogeneously distributed over the imaged area of at least 1 mm 2 . The corresponding fits of the ellipsometric parameters are shown in Figure S3 in the Supporting Information. It is hypothesized that for such a thin coating with a low initiator content, the polymer chains were most likely a randomly distributed copolymer of METAC with the surface-bound vinylsilane. Hence, rather than having a brush-like structure, the METAC chains may have multiple grafting points to the surface at high density, forming a dense film, which may explain the rather thin coating layer. Moreover, the homogeneity of this relatively thin coating may be advantageous for the antimicrobial protection of substrates.
At higher LAP content, the coating thickness on Si-wafers increased to 8.8 ± 0.3 nm at 0.2 wt% LAP and 11.3 ± 0.4 nm at 0.5 wt% LAP, see Figure 2B with the fits presented in Figures S4 and S5 in the Supporting Information. The corresponding ellipsometry images at 0.2 and 0.5 wt% LAP ( Figure 2D,E) Figure 1. A) Schematic representation of the coating process. The Si-wafer was pretreated with a vinylsilane, whereas the cotton was incubated with the photoinitiator for 12 h. Nitrile rubber did not require pretreatment. The surface was then dipped into an aqueous solution of 25 wt% METAC and 0.01 wt% LAP for 20 s. The surfaces were dried in an oven at 50 °C for 15 min, followed by UV irradiation at 7.13 kW m −2 . Potential residual nonbound components were washed off the surface with water. Details are presented in the Experimental Section. B) Chemical structure of the surfaces and the coating compounds with images of the materials. The functionalities that allow the polymerization and binding of METAC to the surface are highlighted. A more detailed reaction mechanism is presented in Figure S1 in the Supporting Information.
www.advmatinterfaces.de demonstrated more irregularity in the coating's surface, most significantly for the 0.5 wt% LAP sample, with a thickness varying between 0 and 18.5 nm. The effect of LAP content on the thickness and the presence of defects agrees with the observed variations in contact angle ( Figure 2A). The morphological differences in the coating at increased initiator content may be due to greater chain-polydispersity from increased initiation and termination effects. [18] This may lead to a less uniform surface coverage with voids due to fewer crosslinking points of the shorter chains with the surface, compared to longer chains.
XPS characterized the coating surface chemistry. The nitrogen 1s XPS spectrum of the METAC coating prepared with 0.01 wt% LAP and its corresponding fit are shown in Figure 3A. The spectrum was deconvoluted into four contributions, corresponding to the four species of N present on the surface after crosslinking. [23] i) The prominent peak at about 401.7 eV corresponds to the trimethyl quaternary ammonium group. ii) The small peak at slightly higher energies (about 402.3 eV) may correspond to quaternary ammonium that has been crosslinked.
iii) The peak at lower energies (about 399.1 eV) is most likely signaling the presence of ternary amino groups. iv) The peak around 406.1 eV indicates the formation of NO bonds. The crosslinking under atmospheric conditions (in the presence of O 2 ) may explain the presence of the NO bonds. In addition, the generations of radicals during the polymerization process may have led to the oxidation of some species on the surface. However, 71.5% of the detected N species on the surface were trimethyl quaternary ammonium. In contrast, ternary amino groups, crosslinked quaternary ammonium, and NO groups contained 11.0%, 10.5%, and 7.0% N species, respectively (see Figure 3B). Hence, over 90% of the N atoms are found as cationic quaternary amines or as pH-dependent positively charged ternary amines, leading to the positive charge of the coating. The C 1s spectra of the layer synthesized with 0.01% LAP and its corresponding fit are shown in Figure 3C. The spectrum was deconvoluted into three contributions corresponding to i) the chemical shifts from CC (at 284.6 eV); ii) the CO and CN (at 286.2 eV); and iii) the C=O (at 288.8 eV). [24] The atomic ratio from the CO and CN contributions (about 65% of all C) was slightly higher than the theoretical ratio for METAC monomers (about 57% of all C). The contribution of LAP is not expected to account for the 8% discrepancy since it is present in low concentrations (0.01 wt%). The higher ratio of CO may be caused by partial oxidation during the crosslinking process, as discussed above for N.
The N 1s XPS spectra for the samples prepared with 0.2 and 0.5 wt% LAP are shown in Figure S6 in the Supporting Information. The wide range XPS spectra are shown in Figure S7 in the Supporting Information for the three LAP contents. They have the same features as the spectra taken for 0.01 wt% LAP, Figure 2. Effect of LAP concentration on the thickness and the hydrophilicity of the METAC coating on Si-wafers. A) The water contact angle on the Siwafers coated with vinylsilane, or METAC, at 0.01, 0.2, or 0.5 wt% LAP. B) The corresponding average coating thickness estimates were calculated from Delta and Psi fittings with varying angles of incidence between 60° and 80° and varying wavelengths between 360 and 1000 nm. The fits are presented in Figures S3, S4, and S5 in the Supporting Information. The thickness images were derived from ellipsometry parameters of the coatings produced with C) 0.01 wt%, D) 0.2 wt%, and E) 0.5 wt% LAP content. The increased termination effects at higher LAP concentrations result in a less uniform coating due to the lower number of bonds to the surface.

www.advmatinterfaces.de
indicating no significant difference in surface composition. The similarity between the spectra of the coatings synthesized with higher content of LAP compared to those of 0.01% LAP suggests that there is no significant increase in quaternary ammonium content on the surface (see Table S1, Supporting Information for the atomic ratio of N species in the different coatings); therefore, both a smaller number of longer polymer chains (created at lower initiator content) or a higher number of shorter chain (made at higher initiator contents) led to the same surface chemistry. This is in good agreement with the presence of the coating prepared at the three LAP concentrations, as imaged by ellipsometry. Considering that a more uniform METAC coating thickness was observed on Si-wafers with the lowest (0.01 wt%) LAP content while providing similar hydrophobicity and surface chemistry, this coating parameter was selected for further studies on nitrile rubber and cotton.
The surface ζ-potential of the METAC-coated and uncoated surfaces is presented in Figure 3D. The coated surfaces consistently showed a positive potential ranging from about +10 mV for cotton to +40 mV for nitrile and +50 mV for Sisurface, demonstrating successful METAC attachment on all the surfaces. The high positive charge density was likely only possible through polymerization in the air-dried bulk state. The coating drying reduced the water content, concentrating METAC and its counter-ions. The double layer forces arising between the positively charged molecules in aqueous environ-ments were significantly reduced in this state. The dielectric constant of air is smaller than that of water (ε air = 1.0 F m −1 and ε water = 80.4 F m −1 ), [25] leading to an approximately ninefold decrease in the Debye length in air. [26] Further, the ionic strength of the solution increases upon water evaporation, which also decreases the Debye length. This novel process allowed, to the best of our knowledge for the first time, the accumulation of METAC molecules from water at a very high density on the surface that were then covalently linked to the surface and other METAC monomers.
The increase in surface charge varied in magnitude depending on the surface coated; for the flat and smooth Si-wafer surface, the increase in the ζ-potential was about +90 mV. The rougher nitrile surface showed a rise of +70 mV, and that for the porous cotton was +30 mV. These differences could have resulted from the modified streaming current at porous and rough surfaces. For such samples, the nonlaminar flow regime provoked by the surface irregularities can lead to difficulties in the ζ-potential analysis. [27] The ability to design highly charged surfaces may be of interest not only for antibacterial application, as we show in the next section, but also for the design of layer-by-layer deposition, as well as for any adsorption onto surfaces that is based on charge interactions. For example, highly negatively charged materials can efficiently remove heavy metals for water purification. [28]

Antimicrobial Properties of METAC-Coated Surfaces
The METAC coatings on Si-wafers, nitrile films, and cotton fabrics synthesized with 0.01 wt% LAP were selected based on their coating characterization to investigate their biological activity. The in vitro antibacterial activity was tested by inoculating their surfaces with bacterial suspensions of Gram-negative Escherichia coli (E. coli) Deutsche Sammlung von Mikroorganismen (DSM) 1103 and DSM 498 and Gram-positive Staphylococcus aureus (S. aureus) DSM 4910 strains. These bacterial strains were chosen since they represent Gram-negative and positive species and have been identified as leading pathogens in hospital-acquired infections. [29] The antimicrobial activity of the surfaces was evaluated after 5 min of contact using colony counting assays, which was the shortest possible time interval owing to sample handling (see the Experimental Section for details). Eliminating pathogens within such a short timeframe minimizes their spread in high-traffic environments. It would be more relevant than testing antimicrobial activity over 18 to 24 h, as recommended by the International Organization for Standardization (ISO) norms. [12h,30] Following 5 min of contact with the METAC-coated Siwafers, a 4.1 ± 0.7, 3.7 ± 0.7, and 3.8 ± 0.6 log reduction in viable populations of the E. coli strains DSM 1103 and 498, and the S. aureus strain DSM 4910 was measured compared to the uncoated Si-wafers ( Figure 4A). The death of the bacteria on the surface was confirmed by confocal laser scanning microscopy with live/dead staining ( Figure S8, Supporting Information). Additionally, the biofilm growth on the Si-wafers was quantified by crystal violet staining following 5 min of contact and 24 h incubation at 37 °C in Mueller Hinton Broth (MHB) culture media (see the Material and Methods Section in the Supporting Information). The S. aureus cells that may have adhered to the METAC-coated Si-wafers could not form a biofilm within 24 h of incubation. In contrast, the bacteria that adhered to the untreated Si-wafers formed a detectable biofilm layer ( Figure  S9, Supporting Information). This suggests that biofilm formation is prevented through rapid bacterial killing upon such short contact. On nitrile films, the METAC coating showed a 4.0 ± 0.7, 2.7 ± 1.0, and 3.2 ± 0.3 log reduction in viable populations of the same respective strains ( Figure 4B).
On cotton fabrics, the METAC coating caused a 4.4 ± 0.6; 1.2 ± 0.2; and 3.6 ± 0.8 log reduction in viable populations of E. coli strains DSM 4910, 498, and S. aureus strain 4910, respectively, as compared to the uncoated cotton fabrics ( Figure 4C). It is worth noting that bacteria extraction from nitrile films and cotton fabrics was limited; nearly half or more of the starting bacteria populations ("bacteria only" in Figure 4C) could no longer be detected after extraction following 5 min contact with untreated control nitrile and cotton for all the tested strains ("control" in Figure 4C). This likely occurred due to the rough surface of nitrile and cotton materials and the 3D structure of the cotton fiber networks, which offered better bacteria entrapment and larger surface area for their attachment and prevented their extraction.
The more potent antibacterial activity of the METAC coating on the Si-wafer compared to cotton and nitrile substrates can probably be attributed to their varying surface morphology and charge. Compared to flat Si surfaces, the roughness and porosity of cotton and nitrile surfaces can result in limited Each experiment was done in triplicate with triplicate readout (n = 9; ttest: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

www.advmatinterfaces.de
METAC film deposition and compromised UV-irradiation power density. This can lead to lower surface METAC densities and coating defects, benefiting bacterial colonization.
The results from the colony counting assay demonstrated the formation of rapid antibacterial METAC coatings on Si-wafer, nitrile films, and cotton fabrics. In this regard, the minimal inhibitory concentration (MIC) and minimal bactericidal concentration (MBC) values for METAC as a monomer in solution were observed to be ≥40 mg mL −1 against all three tested strains, indicating its weak antimicrobial activity when not incorporated in the coating ( Figure S10 and Table S2, Supporting Information). The antibacterial activity of METAC polymers with a molecular weight of 87.4 kDa in solution, reported in the literature, with an MIC of 3.70 and 1.23 mg mL −1 for E. coli and S. aureus, respectively, [31] lies in between that of the monomer and the 2D surface. The increase in antibacterial potency with increasing dimensionality from 0D monomers to 1D polymers to 2D surfaces has been similarly observed when comparing the antimicrobial activity of cationic nanoparticles in solution to the identical particles immobilized on a surface. [10c] This dependence on dimensionality indicates that the mechanism is not based on the permeation of small molecules into the bacterial cytosol and their chemical interaction with various biochemical pathways but rather on the physical interaction between the outer components of the bacteria, such as the membrane, together with the embedded proteins, and the positively charged surface of the materials. Previous antimicrobial studies of cationic polymers showed the necessity to have a combination of hydrophobic and cationic groups to achieve the antibacterial activity. [33] This combination of functionalities allows a synergistic action of i) their electrostatic attraction to the negatively charged bacteria surface, followed by ii) the interpenetration of hydrophobic polymer chains into the bacterial membrane. [32] Such a mechanism falls short in explaining the antibacterial activity of the METAC coatings as they are hydrophilic. The mechanism of antimicrobial action is most likely only driven by electrostatic forces. To further investigate the electrostatic nature of the killing mechanism, the coating's antimicrobial activity was evaluated when extracting bacteria with buffers of varied ionic strength. Antibacterial activity of the METAC-coated Si-wafers against DSM 1103 strain persisted when extracted with MilliQ water or 10 mm Tris buffer supplemented with 0.1 wt% NaCl. However, the antimicrobial activity was almost entirely canceled at 0.9 wt% NaCl in Tris (see Figure S11, Supporting Information). The increase of the ionic strength screens the electrostatic interactions between the surface and bacteria. This observed loss of antibacterial activity further supports the hypothesis that the killing mechanism is based on the electrostatic interaction between the METAC coating and the negatively charged bacterial membrane. This mechanism agrees with the increased antibacterial activity reported for surfaces with an increased positive charge density. [34] Electrostatic interaction between surfaces and lipid vesicles has also been suggested as a main driving force for vesicle rupture in supported bilayer formation. [35] The proposed mechanism is depicted in Figure 5.
As the proposed bactericidal mechanism of the METAC coatings implies nonspecific contact-killing activity, the bacterial attachment to the surface over time may lead to the loss of the antibacterial activity of the surface. To test this, the METACcoated Si-wafers were subjected to repeated bacterial exposures, with a washing step using Tris buffer or ethanol in between the exposures. In this context, ethanol was selected for its everyday use as a surface disinfectant. While the METAC-coated Si-wafers showed rapid contact killing during the first bacterial exposure, these surfaces were antimicrobial inactive after the bacteria were applied the second time, regardless of whether the wafers were washed with Tris-HCl buffer ( Figure S12, Supporting Information) or ethanol ( Figure S13, Supporting Information) between the applications. Most importantly, treatment with ethanol itself did not appear to affect the antimicrobial activity of METAC-coated Si-wafers ( Figure S14, Supporting Information). It should be noted that the antibacterial assays performed here were done with high bacterial loads (6 × 10 5 and 2 × 10 5 CFU cm −2 for E. coli and S. aureus strains, respectively) compared to the 8-10 CFU cm −2 found on toilet door handles, being the most contaminated hospital surfaces. [36] The observed rapid bactericidal activity of METAC coatings toward such large bacteria loads and prevention of their biofilm formation is particularly attractive when considering applications of these coatings on nitrile-based single-use medical gloves or cotton-based garments.
The proposed antibacterial mechanism may allow selectivity toward bacteria over mammalian cells and limit the potential cytotoxicity of the coating due to the more substantial negative charge on the bacteria membranes compared to those of mammalian cell membranes. [37] The coating's biocompatibility was evaluated according to Deutsches Institut für Normung (DIN) European Standard (EN) ISO 10993-5. [38] This assay allowed for assessing the biocompatibility of METAC-coated cotton textile regarding the leachate in artificial sweat. It is based on the relative protein content measured by optical density at 570 nm www.advmatinterfaces.de (OD570) corresponding to cell growth. The leachate of the coating was characterized as biocompatible with <4% inhibition of cell growth. This value is significantly lower than the acceptable limit of 30% inhibition ( Figure S16, Supporting Information). The low cytotoxicity of the leachate suggests that the presence of METAC there is low and any amount that may have leached is nontoxic. This demonstrates that the METAC-coated cotton from this study to is suitable for clothing and use on health workers' gowns.
The designed METAC coating could find broad applications thanks to the simplicity, versatility, and high potential for scalability of its synthesis method. In addition to surfaces containing C=C double bonds and CH bonds that react with the radical METAC group, [21b] the METAC may also be attached to nonorganic surfaces having OH groups with the method shown for Si surfaces. Future applications of the METAC coating synthesized by methods presented here may include polyurethane and wooden surfaces, hence covering most of the high-touch surface chemistries. The combination of its versatility and rapid antimicrobial activity makes it a promising candidate for widespread applications and industrialization as an antimicrobial coating for high-touch surfaces.

Conclusions
Biocompatible antimicrobial polymer coatings were prepared by photoinitiated radical polymerization of dried cationic METAC monomer films from water on Si-wafer, glass, nitrile, and cotton surfaces. The in-depth physicochemical characterization of the modified Si-wafers demonstrated the covalent attachment of METAC onto the surface and the formation of coatings around 6 nm in thickness. The METAC to initiator ratio was optimized to form thin, uniform surface coatings. The coating's main characteristic is its very high positive charge density, having a ζ-potential up to 90 mV higher than the noncoated surface, owing to the high density of quaternary ammonium groups.
The high surface ζ-potential of the coatings was correlated with in vitro microbiological studies that showed rapid elimination of Gram-positive (S. aureus) and Gram-negative (E. coli) bacterial strains within 5 min of contact with all the coated materials. The antibacterial mechanism was shown to be most likely contact-killing. The electrostatic interactions between the negatively charged bacterial membrane and the positively charged coating are thought to attract the bacteria to the METAC-coated surfaces and induce curvature modifications to their membrane, leading to their loss of barrier function and, eventually, death.
The METAC coatings' versatility in terms of substrate choice, ease of production, the potential for scalability, and fast antibacterial activity makes them an attractive candidate for protecting high-touch surfaces by rendering them rapidly antimicrobial. Standard measures to reduce the risk of transmitting microorganisms from healthcare workers to patients and between patients include applying personal protective equipment (PPE) such as gloves, long-sleeved gowns, and aprons. Given the high versatility and antimicrobial efficacy of METAC coatings, future trials should test the impact of METAC-coated PPE compared to standard of care on healthcare-associated infections in hospital settings (e.g., tertiary care intensive care units with a high prevalence of multidrug resistant bacteria). Such coatings could help reduce the transmission of pathogenic microorganisms and thus minimize the burden of hospital-acquired infections for improved patient outcomes.

Experimental Section
Nitrile Rubber Film Preparation and Coating: A method published earlier by Lenko et al. is adapted to prepare nitrile rubber films by a precrosslinking step with ZnO and a post-crosslinking step with a trithiol and a photocrosslinker. [39] 9.8 g of hydrogenated nitrile butadiene rubber (BASF, India) was mixed with 0.2 g of ZnO (Sigma, Steinheim, Germany). The mixture was sonicated (Lab500 NexTgen Ultrasonic platform, SinapTec, Lezennes, France) until the ZnO aggregates were dispersed and the suspension had a slightly milky appearance. The mixture was mixed at 50 °C for 1 h and left cool down to 25 °C. The mixture was decanted to separate the sedimented ZnO. 0.05 g of trimethylolpropane tris(3-mercaptopropionate) (trithiol) (Sigma, Steinheim, Germany) and 0.05 g diphenyl(2,4,6-trimethylbenzoyl) phosphine oxide (Sigma, Steinheim, Germany) was added to the mixture and agitated in the dark for 1 h. The nitrile films were formed by dip coating the ceramic former in a coagulant mix of 2.5% wt Pluronic F-127 (Sigma, Steinheim, Germany) and 30% wt CaNO 3 (Roth, Karlsruhe, Germany), drying the former at 50 °C for 10 min, then dipping the former in the nitrile mix described above at 25 °C for 20 s, followed by H 2 O at 25 °C for 10 s to wash away the excess nitrile from the film. The nitrile film on the former was dried at 50 °C for 10 min.
A solution of 25 wt% METAC was prepared by diluting the commercial 75 wt% METAC solution (Sigma, Steinheim, Germany) with deionized (DI) water. LAP (Sigma, Steinheim, Germany) was added at defined concentrations. The film was formed by dipping the former in this solution. The dipping of the former was repeated before crosslinking the METAC with a UV lamp emitting at 300 nm (Rayonet, Middletown, Connecticut, USA) for 15 min at a power of 7.13 kW m −2 . The film was washed with excess water at 25 °C for 5 min. The films used for control experiments were produced similarly without dipping in the METAC/LAP solution.
Coating of Cotton Fabrics: The cotton fabric was provided by Livinguard AG (Chams, Switzerland). The fabrics were cut to 1 × 1 cm 2 and were soaked in a 1 wt% LAP (Sigma, Steinheim, Germany) aqueous solution overnight (12 h) in the dark. The fabric was then dip-coated with a 25 wt% METAC in water solution (Sigma, Steinheim, Germany) containing 0.01 wt% LAP for 10 s; and dried at 50 °C for 5 min. The dipping procedure was repeated two times before the fabrics were exposed to a UV lamp emitting at 300 nm (Rayonet, Middletown, Connecticut USA) for 15 min at a power of 7.13 kW m −2. The fabrics were then rinsed with excess water for a minute to remove any residual METAC or initiator that was not covalently bound to the surface.
Coating of Silica Surfaces: The silicon wafer (Si-mat, Kaufering, Germany) or glass slide (Paul Marlenfeld GmbH, Lauda-Königshofen, Germany) was plasma treated for 1 min at 30 mW (Zepto, Diener, Ebhausen, Germany). They pretreated with a 20 mg mL −1 solution of triethoxyvinylsilane (Sigma, Darmstadt, Germany) in 99% pure ethanol (Thommen-Furler AG, Rüti bei Büren, Switzerland) for 18 h. The pretreated surface was then coated by repeated dipping into a solution containing 25 wt% METAC in water (Sigma, Steinheim, Germany) with 0.01 wt% LAP (Sigma, Steinheim, Germany) for 10 s and dried at 50 °C for 5 min steps. The UV crosslinking and washing was performed as described in the Cotton section above.
Ellipsometry: Scanning ellipsometry was performed between 60.0° and 80.0° using an Accurion EP4 imaging ellipsometer (Accurion, Goettingen, Germany) with a 5x objective, equipped with a white light Xenon arc lamp and filtered to 547.2 nm. Further spectroscopic ellipsometry was measured at angles of 65.0°, 72.5°, and 80.0°, covering www.advmatinterfaces.de the range around the critical angle of the Si substrate (about 76°). The wavelength was varied from 360 to 1000 nm using a Xenon arc lamp with a filter wheel. The ellipsometric images were acquired at an angle of 70° at a wavelength of 547.2 nm. The data were treated using the Accurion model software. The samples were modelled as a Si substrate covered by a 2.2 nm SiO 2 layer, on top of which resides the polymer layer. The METAC layer was described with a Cauchy model where n is the refractive index, A and B are the Cauchy parameters, and λ is the light's wavelength. The Cauchy parameters were set to A = 1.45 and B = 0.01, as previously used to model a dry organic layer. [40] The Si and SiO 2 were taken as described in the software database. The polymer layer's thickness was the only varied parameter and all others were maintained constant. Contact Angle Measurements: The static water contact angle was measured with a drop shape analyzer (DSA) 30S (Krüss, Germany). A drop of 50 µL was deposited on the surface of interest. Using the sessile drop settings, the contact angle was measured with the Krüss software. Each sample was measured in triplicate and the results reported are the average.
XPS: XPS measurements were carried out on an Axis Supra (Kratos Analytical) using an aluminum anode's monochromated Kα X-ray line. The pass energy was set to 40 eV with a step size of 0.15 eV. The samples were electrically insulated from the stage; therefore, a lowenergy electron flood gun was used for charge neutralization. The data are referenced at 284.8 eV using the aliphatic CC bound of the measured C 1s orbital.
The C 1s and N 1s spectra were fitted with Gaussian-Lorentzian peaks (30% Lorentzian) and a Shirley background was used. The area under each peak was integrated to obtain the atomic ratios.
Surface Charge: The surface ζ-potential was determined using a Surpass instrument (Anton Paar GmbH, Graz, Austria). The ζ-potential of the control and coated surfaces were measured at pH 5.6 using a 0.001 m KCl. Cotton samples and nitrile latex films were measured with a clamping cell and Si surfaces (glass slides) were measured with an adjustable gap cell (Anton Paar GmbH, Graz, Austria).
The ζ-potential of the porous cotton surface was derived from the measured streaming current with Equation (2) [41] where dI str /dΔp is the slope of the streaming current (I str ) against differential pressure (Δp), η is the dynamic viscosity of the buffer solution, ε is the dielectric constant of the buffer solution, ε 0 is the permittivity of vacuum, and L and A are the length and cross-section of the streaming channel, corresponding to 35 and 100 µm, respectively. The ζ-potential of the nonporous glass and nitrile surface was derived from the measured streaming potential with Equation (3) where dU str /dΔp is the slope of the streaming potential (U str ) against differential pressure (Δp), η is the dynamic viscosity of the buffer solution, ε is the dielectric constant of the buffer solution, ε 0 is the permittivity of vacuum, and κ B is the electrolyte conductivity.  14.80%, and 9.90%). L929 cell lines were incubated with extract dilutions for 68-72 h. After incubation, the bicinchoninic acid (BC) assay measured cytotoxicity as optical density at 570 nm (OD570). [42] Bacteria Used for In Vitro Experiments: The bactericidal activity of the coated and uncoated surfaces against Escherichia coli (E. coli) strains DSM 1103 and DSM 498 and Staphylococcus aureus (S. aureus) strain DSM 4910 (Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), Braunschweig, Germany) was investigated using i) a modified microdilution plating assay and ii) agar pouring assays. [43] Glycerol stocks of the bacteria were spread-plated on Mueller Hinton agar (MHA, Sigma-Aldrich, Buchs, Switzerland) and incubated overnight at 37 °C. A few of the emerging single colonies were then transferred to a 25 vol % dilution of MHB in 10 mm Tris-HCl buffer pH 7.4 (both from Sigma-Aldrich, Buchs, Switzerland) for overnight incubation at 37 °C and 180 RPM using an orbital shaker-incubator (ES-20, Biosan, Riga, Latvia). Afterward, 1 mL of the overnight culture was transferred to 9 mL of fresh 25 vol% MHB in Tris buffer and incubated for a further ≈3 h. Optical density measurements at 600 nm (OD600, Biochrom Ultrospec 10 Spectrometer, Holliston MA, USA) showed that the bacteria in this fresh inoculate were in their exponential growth phase for at least the first 4 h. Therefore, after the ≈3 h incubation, 4 mL of the culture was centrifuged at 3000 x g for 5 min. The pellet was resuspended in an equal volume of either 1 or 10 vol % MHB in Tris buffer for the surface or MIC assays, respectively. The OD600 of these bacterial dispersions was then measured to estimate the bacteria concentrations according to calibration curves. The calibration curves for each strain were constructed beforehand by diluting the bacterial dispersions to different OD600 values and quantifying the bacteria populations by the microdilution plating method described below.
In Vitro Surface Antibacterial Activity: After centrifugation, the bacteria were resuspended in 1 vol % MHB in Tris buffer pH 7.4. The concentration was adjusted to 3 × 10 7 CFU per mL for E. coli and 1 × 10 7 CFU mL −1 for S. aureus. 1 cm 2 piece of METAC-treated Si-wafers, nitrile, or cotton was placed in sterile glass vials, and 20 µL droplets of bacteria culture were applied onto the materials. Corresponding untreated materials were used as controls. For nitrile and Si-wafers, a sterile piece of low-density polyethylene (LDPE) film (ROTILABO, 80 µm thick, Carl Roth GmbH, Karlsruhe, Germany) was placed on top of the droplet to spread the bacteria on the surfaces evenly ( Figure S15, Supporting Information). After a contact time of 5 min, bacteria were extracted for quantification by submerging the materials in 2 mL of 10 mm Tris-HCl buffer pH 7.4 (referred to as extraction buffer) and sonicating the solutions for 10 s using Vibra-CellTM 500 W 20 kHz (Sonics and Materials, Inc., Newtown CT, USA) with a 3 mm microtip at 35% amplitude (1 s pulse, 1 s pause). Sonication was necessary for cotton samples to release bacteria trapped inside the cotton fiber networks, even in control samples. The use of sonication has been previously reported to extract bacteria from cotton fabrics. [15b] In additional experiments on the effect of the ionic strength of the extraction medium on the process, extractions were also carried with MilliQ water and Tris buffer supplemented with 0.1 or 0.9 wt% NaCl. The bacteria in the extraction buffer were immediately quantified by microdilution plating and agar pouring assays (see below). As a control, the same 20 µL bacteria droplets were placed on the bottom of the glass vials in the absence of the materials, covered by an LDPE film for 5 min, and extracted by sonication (referred to as "just bacteria"). The log reduction is calculated by dividing the surviving population on METAC-coated materials by the mean of their corresponding uncoated control and is presented as the mean and its standard deviation. To correctly represent experimental variance on a log scale, the Student's t-test was also performed on the log differences between the data sets and their corresponding controls' means (and variances). Each value is an average of three independent experiments.
Agar Pouring and Microdilution Plating Assays: To quantify the viable bacterial populations after contact with the material, 0.5 mL fractions of the extracted media were transferred to separate petri dishes in triplicate and mixed with about 15 mL of warm (≈55 °C) MHA. After www.advmatinterfaces.de cooling, the agar plates were then incubated for 2 d at 37 °C and the CFU was quantified from the different colonies formed on the plates. The low detection limit of this agar pouring method was complemented with the microdilution plating assay, similar to previous studies, [44] allowing quantification of higher bacterial concentrations. The extracted media were serially diluted tenfold with 10 mm Tris-HCl buffer pH 7.4, containing 0.9 wt% NaCl, three times. 10 µL from each of the resulting dilutions were plated in triplicates on MHA. After overnight incubation at 37 °C, the colonies were counted.
Determination of MIC and MBC of METAC in Solution: After centrifugation, the bacteria were resuspended in 10 mm Tris-HCl pH 7.4 with 10 vol % MHB and the concentration was adjusted to 2 × 10 5 CFU mL −1 for all bacteria strains. In a transparent 96-well plate, bacteria cultures were added to 10 vol % MHB containing METAC at different concentrations in triplicate, with final 1 × 10 5 CFU mL −1 bacteria content and METAC concentrations from 0.625 to 160 mg mL −1 . The cultures were incubated at 37 °C with continuous shaking at 180 RPM for 24 h and the optical density of the well contents at 600 nm was measured using Spark 10 M Spectrometer (TECAN, Grödig, Austria). Results are presented as the average and standard deviation from the triplicate measurements. The concentration of METAC at which no changes in the OD600 were observed was stated as the MIC. Additionally, after the 24 h incubation in the spectrometer, the contents of the triplicate wells were pooled together and quantified using microdilution plating. The concentration of METAC in which no surviving bacteria was detected signifies the MBC.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.