ABSTRACT

The increasing emergence of antimicrobial resistance and the development of new infective viral strains represent a constantly growing threat. Metal-based nanomaterials have emerged as promising tools in the fight against bacterial and viral infections; however, the release of metal nanoparticles/ions in clinical applications may cause undesired side effects (allergies, systemic toxicity), reducing their practical use in antimicrobial treatment. Moreover, the metal-based nanoparticles possess predominantly antibacterial effects, while their antiviral efficiency remains controversial. Thus, the development of metal-free strategies enabling combined antibacterial/antiviral properties is a significant challenge. Here, we report a strategy based on light irradiation of nitrogen-doped graphene acid (NGA) possessing dual photothermal and photodynamic modes of action. The antimicrobial activity is activated through a clinically approved near-infrared (NIR) light source, and both viral and bacterial spreading can be hampered on the coating irradiation on a scale of minutes (5 to 10 min). The developed metal-free strategy reduced 90.9% and 99.99% for S. aureus and P. aeruginosa, respectively, as well as 99.97% for murine hepatitis virus. Importantly, this research represents a significant advancement in the development of safe, metal-free, and effective antimicrobial treatments. NGA coatings are safe for skin, showing no sensitization or irritation, and offer significant potential for advanced antimicrobial treatments.

1 Introduction

According to the World Health Organization (WHO), antibiotic resistance is one of the greatest threats to global health, affecting everyone regardless of age or location [1]. The diminishing efficacy of existing antibiotics in combating infections such as pneumonia, tuberculosis, gonorrhea, and salmonellosis underscores the increasing challenge in managing these diseases [2]. Every year, 4.95 million people die due to infections associated with antibiotic resistance [3, 4]. Alarmingly, it is estimated that by 2050, there may be no effective antibiotics available unless new drugs are developed or discovered, emphasizing the urgency to explore alternative methods for tackling antibiotic-resistant pathogens [5]. This dreadful scenario has been likened to a “silent pandemic,” marked by an escalating prevalence of cases and resistant strains, posing a threat more severe than the SARS-COV-2 pandemic [6]. In response to these formidable risks, there is a pressing need to investigate alternative strategies beyond conventional antibiotics [7].

Antimicrobial coatings have emerged as a cutting-edge technology in the field of material science and healthcare [8, 9]. These coatings play a pivotal role in preventing the spread of infectious diseases by inhibiting the growth and survival of microorganisms on various surfaces and are thus an alternative strategy to overcome antibiotic resistance [10, 11].

Phototherapy (PT) is gaining more and more attention as an antimicrobial treatment strategy [12]. PTs are used in clinics for the treatment of different diseases, including cancer (together with immunotherapy) [13, 14] and infections [15-18]. In PT, the nanomaterials act as photosensitizers, able to convert light energy into heat (photothermal effect) and to produce reactive oxygen species, ROS, via photodynamic effect [19]. Although nanotechnology-based phototherapy has progressed due to its efficacy in antimicrobial treatment with minimal bacterial resistance, its practical use in clinical settings has been impeded by phototoxicity and restricted penetration of UV and visible light sources into tissues [15]. For these reasons, it is crucial to use nanomaterials that can be activated in the near-infrared (NIR) windows where most of the soft tissue is transparent. In this context, graphene oxide (GO), the water-dispersible form of graphene, is gaining more and more attention in the biomedical field [20, 21]. GO bears a wide variety of interesting properties, including generally good biocompatibility, versatile surface chemistry, high surface area, and the ability to overcome cell barriers, making it particularly interesting for drug delivery and tissue engineering. Additionally, GO is a powerful photosensitizer able to convert NIR light into heat and ROS efficiently, so it has been used in cancer and antimicrobial phototherapy [22-25]. Graphene oxide is also well explored in antimicrobial treatment. Here, its intrinsic antimicrobial properties [26-28], its photoactivity [29-32], or combination with metal-based nanoparticles [33-35] are frequently used. However, GO itself still suffers from two major handicaps, mainly related to its production. First, the harsh oxidation conditions during synthesis induce high batch-to-batch variations in size, chemical composition and defects [36]. Secondly, it is very difficult to control the surface chemistry during the oxidation process, leading to a randomization of the oxygenated groups present on GO’s surface [37, 38]. Thus, the development of any other graphene derivative with well-defined chemical composition and possessing dual photothermal and photodynamic properties, applicable in antibacterial/antiviral treatment, still represents a significant challenge.

We have recently developed a 2D colloidal and biocompatible nitrogen-doped graphene acid (NGA) [39], which has a well-defined chemistry and colloidal stability [40]. In the current study, we report the preparation and characterization of an NGA-based antimicrobial coating (Figure 1) and benchmark it with a previously reported TiOx/AgOx plasma coating. Such a coating containing NGA embedded in a polyvinyl alcohol (PVA) matrix cross-linked with 1,4-boronic acid (DBA) shows excellent antibacterial and antiviral properties when irradiated with NIR light. To the best of our knowledge, it is the first metal-free strategy based on a compact carbon-polymer coating enabling light-induced antibacterial and antiviral treatment and thus, exceptional potential for use in medicinal practice. Both photothermal and photodynamic contributions have been proven in the antimicrobial mode of action. The long-term stability of the NGA PVA-DBA coating, repeated use after irradiation without a loss of efficiency, and no observed adverse effects on skin cells represent other added values opening the doors for clinical testing of this novel antimicrobial strategy. Staphylococcus aureus is the main pathogen responsible for different skin diseases, including atopic dermatitis, and its uncontrolled spreading onto the skin induces higher viral infection rates [41]. At the same time, Pseudomonas aeruginosa is generally detected with deep chronic ulcers [41]. In this work, we describe, for the first time in the field, the use of metal-free coatings with dual photothermal and photodynamic activity. The produced NGA coating exhibits good adhesive and wettability properties for a broad application on different substrates and reduced biofouling, high disinfectant properties against both viruses and bacteria, a long shelf life without losing its activity over time and irradiation cycles, and is activatable by bioinert NIR light and does not induce acute side effects in our skin models. We believe that these coatings may find application for the treatment of for example, external skin infections.

Schematic representation of individual steps towards the synthesis of NGA PVA-DBA coating and its use for light-induced antiviral/antibacterial treatment. After synthesis, NGA was embedded into the PVA-DBA matrix. The coating showed high antimicrobial activity against viruses and bacteria using 5 to 10 min of NIR irradiation.

1.1 Background and Novelty

Metal nanoparticles (e.g., Ag, Au, Cu, and Zn) as zero-valent and metal oxides have been widely studied for their antimicrobial properties [42-45]. Their disinfectant activity is based, among others, on the release of metal ions that interfere with the pathogens through various mechanisms [8, 46]. Direct contact-mediated killing occurs when nanoparticles attach to and penetrate the bacterial cell wall, causing membrane damage and leakage of cellular content leading to bacterial death [47]. The application of Ag- and Cu-nanoparticles can also trigger the formation of reactive oxygen species (ROS) that irreversibly damage the pathogen’s membrane, compromising their infectivity [35]. However, three important aspects are reducing the application potential of nano-metal-based coatings, including (i) apparent toxicity of Ag/Cu nanoparticles against human cells presenting some risk for humans [48-54], (ii) the release of metal ions from a coating (e.g., wound dressings, catheters, and surgical wires) into the biological fluid or environment [55-61], (iii) recently discovered resistance of bacteria to repeated application of metal nanoparticles [62-64]. Metal nanoparticles have also been studied in phototherapy. A gold standard in this field is nanosized titanium (TiO2), which is used as a disinfectant coating on frequently touched surfaces, including the hospital environment (the so-called advanced oxidation process AOP). For titanium to be active, it requires high energy UV light that is phototoxic for skin and eyes. Doped TiO2 with other metals such as Ag, Cu, or Co can be excited with visible light but with concerns about costs and side effects of the final coatings [65]. Thus, developing metal-free coatings with tunable antimicrobial activity would be particularly interesting for different medical scenarios. In this paper, we describe the use of a novel metal-free coating based on NGA as a powerful NIR activatable coating that exerts high activity against bacteria and viruses without exerting any acute toxicity to skin. Compared to metal-based antimicrobial coatings, our strategy possesses many advantages (Table 1). Firstly, the long shelf life (over 20 days) without any loss of activity is associated with the good stability of the NGA compared to other metal structures (e.g., Ag or TiO2) a surface oxidation and change of the catalytic site [66, 67] while their uncontrolled release from the antimicrobial coating is currently posing great concerns [68]. Secondly, there is an economic and green impact where graphene-based materials are generally more cost-effective due to their abundant raw material (graphite) and simpler production processes, while Ti and Ag extraction and processing are considered impactful to the environment and costly, and there is an imminent risk for their supply chain [69, 70]. Finally, while metal ions released, such as silver, can induce bacterial resistance [71], no resistance was observed following multiple sessions of phototherapy [16].

TABLE 1. Comparison of NGA and metal-based antimicrobial coatings.

Properties	NGA vs. metal coating
Antimicrobial activity	NGA based coatings can generate on-demand reactive oxygen species under NIR light irradiation with wide range of disinfectant properties
Clinical applications	Ag and TiO2 coatings are widely used in antimicrobial coatings while NGA it is still at a research level
Case studies	Phototherapy with NGA has an intrinsic limitation due to the light penetration so can be used only up to a few of centimeters
Catalytic mechanism	In metal-based antimicrobial coatings, the metal ions are constantly released from the surface. NGA act as a catalyst able to convert O2 in superoxide on demand for several cycles
Reduced toxicity	While metal ions may initiate sensitization and irritation, while graphene-based materials are considered to be generally safer [21]
Environmental impact	NGA is more environmentally friendly as it does not leach harmful metal ions into the environment, which can be a concern with metal coatings
Cost-effectiveness	NGA can be produced relatively inexpensively compared to some metals, making it a cost-effective option for large-scale applications

2 Results and Discussion

2.1 Characterization of NGA and NGA PVA-DBA Antimicrobial Coating

As an emerging photosensitive 2D material, we employed dually functionalized graphene (NGA) containing both out-of-plane functional groups (carboxyl) and in-plane heteroatoms (nitrogen doping) as a bio-inert and very efficient coating material with highly potent antibacterial and antiviral effects (Figure 1).

Prior to the preparation of the antimicrobial coating, we first proceeded with the nanomaterial characterization. Analysis of NGA by x-ray photoelectron spectroscopy (XPS) revealed the dominant content of carbon (70%), along with 24% oxygen and 6% nitrogen (Figure 2a). Deconvolution of the C 1 s HR-XPS spectral region revealed that the most abundant component was centered at 286.3 eV, representing mainly sp3 CC carbons (in terms of elemental composition) and possible CO and CN configurations [40]. The high presence of sp3 carbons arises from the graphene carbons with bonded carboxylic groups. The component around 290.4 eV can be clearly assigned to carboxylic groups. The feature at 288.2 eV was assigned to lone CO bonds or possible amide configurations. The presence of CO and CO bonds was also evident from the deconvoluted O 1 s spectral region [40]. The infrared (IR) spectrum of NGA (Figure 2b) exhibited similar features to those of conventional organic carboxylic acids, particularly a broad OH absorption band in the stretch between 2700 and 3700 cm−1 due to excessive hydrogen bonding of the groups. The absorption bands at 3474 and 3240 cm−1 correspond to different Hbonding configurations. The strongest band in the fingerprint region at 1735 cm−1, typical of CO stretching, is neighboring the second strongest band at 1600 cm−1, which is assigned to asymmetric stretching of the carboxyl group. The bands of symmetric stretching vibrations at 1430 and CO stretching at 1230 cm−1 lie on the broader feature of skeletal CC vibrations [72]. Photodynamic activity of graphene-based material relies on their capacity to stabilize unpaired electrons (spin) on their surface. In order to obtain more information about the nature of the spin-containing defects in the structure of the NGA, we performed Electron Paramagnetic Resonance (EPR) measurements (Figure 2c). NGA exhibits typical EPR fingerprints, resulting in a sharp resonant line at B = 325 mT that gives g = 1.997. Furthermore, we did not observe any indication of a high-spin state (e.g., S > 1/2), thus no indication of dipolar components and zero-field splitting terms. Moreover, the EPR spectra did not show any detectable interaction with the 14 N (I = 1) atoms, indicative of the exclusive localization of spin-active sites at the carbon centres [73]. A high ID/IG value of 1.33 determined by Raman spectroscopy (Figure 2d) indicates the highly defective character of the graphene lattice of the NGA, along with the high level of sp3 functionalization by carboxyl groups, suggesting that it could be a good candidate for a photosensitizer [74].

XPS survey spectrum (a), FT-IR spectrum (b), CW X-band EPR spectrum (c), Raman spectrum (d) of NGA material.

Due to the good water dispersibility of the NGA, to produce a stable coating we decided to incorporate it into a polymer composite. In this context, the polymeric matrix should: (1) be biocompatible to avoid side effects on tissue (e.g., inflammation), (2) be adhesive and easy to apply, (3) be able to disperse the NGA well into the final coating, allowing the permeation of O2 in order to not hamper NGA’s photodynamic activity, (4) be hydrophilic in order to prevent biofouling. To produce a robust adhesive coating applicable on different surfaces, we incorporated NGA (3.5% in weight) into the PVA-DBA matrix, which represents an efficient adhesive on many surfaces including plastics and glass. By itself, PVA is a biocompatible water-soluble polymer used in different biomedical applications. Moreover, PVA-DBA hydrogels are easy to apply and show good hydrophilicity and antifouling capacity [75].

By our optimized protocol, we produced a coating with 0.27 mg/cm2 density and an average thickness of 3.3 ± 0.4 μm. The NGA PVA-DBA films are characterized by low roughness as shown by the profilometry analysis (Figure S1). Contact angle measurements of PVA-DBA and NGA PVA-DBA show good surface hydrophilicity with contact angles of 69.5° ± 3.0° and 69.3° ± 4.1° respectively (Table S1). Importantly, we did not observe any change in the surface hydrophilicity due to the NGA embedding into the matrix, which is important for avoiding absorption and biofouling [76].

Next, we performed detailed microscopic analysis of both pristine NGA and NGA PVA-DBA coating. The analysis of NGA by high-resolution transmission electron microscopy (HR-TEM) and scanning electron microscopy (SEM) showed that NGA comprised few-layered graphene flakes with lateral sizes around 800 nm (Figure 3a–c) with the SEM image with lateral size in the inset. Elemental chemical mapping of NGA confirmed the homogeneous coverage of the flakes with nitrogen and oxygen, coinciding with the spatial distribution of carbon (Figure 3d–f). The SEM images of NGA PVA-DBA coating revealed a quite homogeneous layered-like surface structure of the coating (Figure 3g,h), as already reported in the literature [75]. After irradiation, we observed a change in the morphology that might be due to the drying of the coating associated with the photothermal activity of the NGA flakes (Figure 3i,j).

(a, b) HR-TEM images of NGA (inset: SEM image of lateral size of NGA). (c) HAADF-STEM image of an NGA flake. EDS chemical mapping of NGA for (d) carbon, (e) nitrogen, and (f) oxygen. SEM images of (g, h) NGA PVA-DBA before irradiation, and (i, j) NGA PVA-DBA after irradiation.

The μ-Raman spectrum of NGA PVA-DBA coating is reported in Figure S2. The component analysis evidenced the presence of Raman vibrations of NGA (D and G bands at 1348 and 1596 cm−1, respectively) together with the vibrations of the PVA-DBA matrix (2922 CH vibration, 1611 CC and CCH vibration and 1320 cm−1 associated with the 5-ring B ester breathing mode) [77, 78]. Moreover, the component spectrum of NGA PVA-DBA does not display any appreciable shift that could be associated with a chemical change. The latter evidence, together with the high uniformity of the component analysis, strongly suggests that NGA is homogeneously embedded into the PVA-DBA cross-linked polymer matrix while the presence of the 2D nanomaterial does not alter the matrix structure.

2.2 Photothermal and Photodynamic Properties of NGA PVA-DBA Antimicrobial Coating

In evaluating the photoactivity of NGA PVA-DBA coatings, experiments were conducted to assess their capability to generate chemical radicals (photodynamic therapy, PDT) or heat (photothermal therapy PTT). Comparisons were made with PVA-DBA and indocyanine green (ICG) as controls, revealing a significantly higher photooxidation activity in the presence of NGA compared to ICG (Figure 4a). Additionally, pseudo-first-order decay analysis indicated enhanced degradation rates for NGA PVA-DBA compared to controls. From that, we found a kobs of 0.099 ± 0.001, 0.071 ± 0.005, and 0.0238 ± 0.0005 s−1 for ICG PVA-DBA, NGA PVA-DBA, and PVA-DBA, respectively. Considering that in the PVA-DBA matrix there are no photoactive materials, we can associate the DHE degradation with the photo-bleaching of the probe. Additionally, when the coatings were stored at RT for a week, we found that ICG lost almost all its photodynamic activity, while NGA PVA-DBA displayed a similar activity to freshly prepared (Figure 4b). Note that ICG is a standard molecule used for bioimaging at the clinical level and as a PDT agent for anticancer therapy [79, 80]. More recently, ICG has also been applied for PDT of bacterial infections [81]. The use of ICG with Hydrosun light has already been reported as PDT for oral mucosa infections against periodontitis and peri-implants [82]. It is worth noting that in most cases, ICG is used in solution as a photosensitizer, where its activity can be enhanced due to the high uptake in cells. Additionally, ICG, like many other molecular photosensitizers, shows a short shelf life at room temperature. We can conclude that NGA PVA-DBA coating under NIR light can produce ROS with a similar activity to ICG but with a significantly longer shelf life.

(a) Photooxidation kinetics of DHE using PVA-DBA, NGA PVA-DBA, and ICG PVA-DBA using NIR light; (b) NGA PVA-DBA and ICG PVA-DBA ROS photooxidation ability of DHE (10 min NIR light exposure) after aging; (c) ΔT measured during NIR light irradiation of NGA PVA-DBA coating; (d) ΔT of NGA PVA-DBA sample during cycles of NIR irradiation (5 min) and cooling (5 min).

Graphene family materials are additionally well-known for their photothermal activity—the ability to convert light into heat—and their applicability in PTT [29-32]. In this context, we wanted to assay the PTT activity of the prepared films (Figure 4c). The presence of NGA elicited a high PTT activity, reaching a plateau after 5 min of NIR light exposure with a ΔT of 14°C, whereas PVA-DBA and the glass slide display an increase of only 1°C and 0.2°C, respectively. Furthermore, we estimated the PTT efficiency of the films, finding 4.4% ± 0.3% and 0.17% ± 0.05% for NGA and PVA-DBA and PVA-DBA, respectively. The temperature to effectively kill bacteria should reach at least 58°C, where cells reach an apoptotic/necrotic state [83]. However, this would pose a limit for PTT approaches since this high temperature would cause irreversible damage to mammalian cells [84]. More recently, it was reported in E**scherichia coli that a temperature between 45°C and 55°C can turn bacteria from a live to a pre-apoptotic state, already effective for bacteria and sensibly enhancing the wound healing process [83]. We found that the average T we measured onto the surface during light irradiation reached 44°C; this, together with ROS formation (Figure 4a,b) can be enough for surface disinfection. Finally, we assayed the ability of NGA PVA-DBA coatings to be reused for multiple irradiation sessions (Figure 4d). We found that a similar ΔT > 13°C was maintained for up to 20 cycles. Considering periodontitis, the rejection of the implant occurs in 10% of the cases [85]. Meanwhile, 41.5% of the patients with catheter-related bloodstream infections experienced recurrent infections [86]. Our developed coating can diminish infections in clinical settings. Indeed, NIR light has a penetration depth of a few cm into soft tissues and can be used in different clinical settings, including skin infection, to avoid the spread of pathogens. Additionally, while PDT relies on the reaction of the photosensitizer with molecular oxygen, PTT is an oxygen-independent mechanism. Thus, the combination of PDT and PTT with a single light source is ideal for sterilization of the surface, even in biofilm-contaminated implants where oxygen concentration is typically low [87]. Similar metal-free strategies have been reported for the treatment of infection using polydopamine and ICG as PTT and PDT agents, respectively [88, 89]. In our case, we showed that NGA, as a single element, acts as a dual-mode photosensitizer using an NIR light source, producing catalytically and efficiently heat and superoxide radicals.

2.3 Antimicrobial Activity of NGA PVA-DBA Coatings

The final objective of our study was the assessment of the antimicrobial activity of NGA PVA-DBA coatings against viruses and bacteria, thus exploiting the dual photothermal and photodynamic properties of the coating. The antiviral capacity of NGA PVA-DBA, PVA-DBA, and non-coated control glass slides was performed according to ISO-21702 over 2 h with a murine hepatitis virus (MHV) system. The plaque-forming units concentration (pfu/mL; see Figure 5a) was calculated as part of each antiviral assay (N = 3–6, n = 2) as requested in the ISO-21702 standard described. On average, a pfu/mL concentration of 70,278 ± 17,599 was recovered on glass slides after immediate washout (see Control Figure 5a). The immediate washout, by which the virus suspension is contacted for < 30 s within a glass slide sandwich, represents the control to determine how many MHV viruses survive and can be retrieved from the non-coated glass slides, since this treatment is performed for samples after irradiation and incubation as well (Control within Figure 5a). The calculated control pfu/mL concentration served as a reference MHV count for further sample assessments. The contacting time of 2 h on control glass samples, even without coating in darkness, resulted in a percentage viral reduction of 87.8% ± 1.8% with a respective pfu/mL count of 8,450 ± 896. Irradiation with light for 5 min even further reduced the MHV recoveries to 450 ± 142 pfu/mL. NGA PVA-DBA revealed an excellent antiviral capacity of up to 99.97% ± 0.04% when irradiated for only 5 min (15 ± 5 pfu/mL). In the absence of irradiation, the antiviral activity of graphene family materials has been reported in different studies [6, 90]. However, most of them report the ability of the 2D carbon materials in suspension to block viral entry or reduce infectivity once inside cells. In our case, we embedded NGA in a PVA-DBA coating so the interaction between the nanomaterial and the virus was limited due to the coating structure. Additionally, we proved that NGA PVA-DBA can generate ROS and increase the local temperature under NIR-light irradiation (Figure 4), and the combination of these two effects leads to a very high antiviral activity reducing the titer of 4 log compared to the control (Figure 5a).

Antimicrobial assessment of NGA PVA-DBA. (a) The antiviral assessment was performed according to ISO-21702 over 2 h with a murine hepatitis virus (MHV) system w/o (green bars “-L”) and with 5 min irradiation (red bars “+ L”). The Control represents the immediate washout. All NGA-coated antiviral and control assessments were conducted (N = 3–6, n = 2) at least in triplicates. (b, c) The assay was performed by incubating bacterial suspension with coating samples for 2 h, followed by removal of suspension and washing. The collected suspension and washing solution were pooled and plated on agar plates for colony counting. AgOx/TiOx samples are used as positive controls. Statistical analysis was conducted by a post hoc Tukey test assessing the mean value and variance from non-irradiated control versus irradiated data. p-values < 0.05 were reported with *, respectively p-values < 0.01 with **, p-values < 0.001 with ***.

The antibacterial activity of the NGA PVA-DBA coated glass slides was evaluated as described in the Materials and Methods. AgOx/TiOx coatings were used as positive controls as they possess antimicrobial activity through reactive oxygen species (ROS) and release a negligible amount of antimicrobial metal ions such as Ag+ due to the plasma processing [91]. Setting the viable bacteria in the initial bacterial culture on glass as 100% (7.6 × 104 CFU/mL for S. aureus and 4.3 × 106 CFU/mL for P. aeruginosa), NGA PVA-DBA coatings reduced viable S. aureus by 69% without light irradiation (Figure 5b). NGA PVA-DBA did not significantly impact the P. aeruginosa viability without light (Figure 5c). Organoboron compounds for example, boric acid and boronic moieties, have been reported to increase susceptibility in Methicillin-Resistant S. aureus (MRSA) being able to inhibit β-lactamase, an enzyme highly associated with bacterial resistance [92]. Additionally, boronic species have been reported to inhibit the NorA efflux pump on MRSA, preventing antibiotic efflux and bacterial recovery [93]. This can explain the higher antimicrobial activity of the NGA PVA-DBA without light for S. aureus compared to P. aeruginosa. However, with light irradiation, the viable S. aureus and P. aeruginosa were dramatically reduced by 90.9% and 99.99%, respectively. Thus, with NIR-light irradiation, NGA PVA-DBA samples enhanced their antibacterial activity against both S. aureus and P. aeruginosa due to the photo-stimulated generation of ROS and heat. Besides, the long-term stability of the PVA-DBA NGA coating remains uncertain, especially in harsh biofluids (presence of proteins, salts, and enzymes), raising concerns about its durability and sustained efficacy. In our study, the crosslinking reaction of the PVA chains through the boronic ester might be subjected to hydrolysis over days. Additionally, while initial results are promising, there is a possibility that their effectiveness in tackling biofilms may diminish due to the short lifetime of the superoxide generated during the irradiation treatment. On the other hand, the generation of radicals can retard biofilm formation, extending the lifetime a—crucial aspect in catheter infections. For these reasons, while use in internal implants would need further engineering of the coating, we believe that this technology can be particularly efficient for non-permanent devices such as catheters, wound dressing, or tooth aligners.

2.4 Skin Sensitization/Irritation With NGA PVA-DBA

To assess an exposure scenario with skin, artificial sweat (AFS) extracts of NGA PVA-DBA coating on human skin, an acute in vitro KeratinoSens skin sensitization and irritation assay was performed (see Figure 6a,b) of serially diluted coating extracts with and without irradiation according to OECD guidelines No. 442D.

In vitro skin irritation and skin sensitization potential assessment of the NGA coating extracts. KeratinoSens skin sensitization & irritation assay was performed according to OECD guidelines No. 442D. NGA PVA-DBA were compared to control glass slides non-coated extracts, and coated glass slides without active compound (PVA-DBA;) in serial dilution ranging from 0.3125% to 10.0% artificial sweat extract within the KeratinoSens medium (a). In addition, a known skin sensitizer Ethylene glycol dimethyl acrylate (EGDMA) in the concentration range of 1–1000 μM was included (b). The black dotted line indicate a skin irritation threshold of below 70% viability and the red dotted line a skin sensitization above 1.5 fold induction (EC1.5) compared to the solvent control. NGA DNA-PVA extracts and controls with and without illumination revealed no skin irritation (≥ 70% viability) and sensitization (< 1.5 fold induction) effects according to OECD guidelines No. 442D (a). The NIR light irradiation for 5 min (“+L”) was conducted directly on the cells in the assay plates after extract addition and did not result in a reduced irritation or sensitization as illustrated in NGA PVA-DBA (+L). Graphene acid coating experiments (N = 3–6, n = 3), and positive controls (N = 4, n = 2–3) were conducted in triplicates.

All samples, including NGA PVA-DBA glass slides (0.3125%–10.0% AFS), PVA-DBA coated glass slides (0.3125%–10.0% AFS), and control glass slides (0.3125%–10.0% AFS) revealed neither a skin irritation effect (cell viability of all samples > 70%) nor a skin sensitization effect (fold induction ≤ 1.5×) in the applied AFS extraction concentration range. The black dotted line in Figure 6a,b indicates the viability threshold of 70%, whereas the red dotted line indicates the sensitization threshold of 1.5 fold induction, which needs to be reached to classify the sample as an irritant or sensitizer. A slight but non-significant tendency to lower cell viability and lower fold induction (reduced skin sensitization) was observed within irradiated samples (triangles, Figure 6a) in comparison to non-irradiated samples (squares, Figure 6a). The validity of the assay was shown with the positive control Ethylene glycol dimethyl acrylate (EGDMA) which revealed a skin irritation (301 μM IC30; 389 μM IC50) and a skin sensitization effect (132 μM EC1.5) in compliance with OECD guidelines No. 442D (Figure 6b). Graphene acid coating experiments (N = 3–6, n = 3) and positive controls EGDMA (N = 4, n = 2–3) were conducted in triplicates. Although NGA PVA-DBA did not show any acute effect on advanced cell models, it is worth noting that prior to their use, a deeper toxicological profile, considering also their potential long-term effects, should be performed. Indeed, on one hand, boronate crosslinked coatings exhibit notable stability [94, 95] due to the dynamic covalent bonds formed between boronic acids and diols attributed to the rehybridization of boron from sp2 to sp3 [96] Consequently, these coatings are less prone to degradation and swelling. On the other hand, the intrinsic corrosive nature of the biological fluids (e.g., blood, mucus, saliva) characterized by high salinity and/or the presence of proteins, might enhance the hydrolysis of the boronic ester ligands and favor the coating biotransformation. Regarding the NGA, generally, graphene-based materials are considered quite biocompatible; size, structure, and functionalization play a pivotal role in their interaction with cells [21, 97]. Additionally, while most of the carbon nanomaterials are biodegradable, there is a lack of knowledge about NGA persistence in the environment.

3 Conclusion

In conclusion, our study successfully demonstrates that NGA can function as a dual-mode (PDT and PTT) and metal-free NIR photoactive coating with exceptional longevity and a high number of irradiation cycles. It shows significant activity against coronavirus (MHV), gram-positive (S. aureus), and gram-negative (P. aeruginosa) bacteria without causing acute skin irritation or sensitization, marking a paradigm shift in surface disinfection. Key achievements of this study include:

1. Discovering and understanding the mode of action of NIR-NGA-based PDT and PTT.
2. Demonstrating efficacy in surface disinfection without skin irritation or sensitization.
3. Developing an easy and versatile production method for the coating.

4 Experimental Section/Methods

4.1 Chemical Reagents

Graphite, fluorinated polymer > 61 wt.% F—GF, (Millipore Sigma), sodium azide 99%–NaN3 (Millipore Sigma), dimethylformamide pure-DMF (Lach:ner), absolute ethanol—EtOH (Penta), nitric acid 65%–HNO3 (Lach:ner), acetone 99% (Millipore Sigma). Ultrapure water (18 MΩ cm) was used for all solutions.

PVA (MW: 9000—10 000) Merck, Benzene-1,4-diboronic acid Merck, 4-(1,2-Dihydroxyethyl)benzene-1,2-diol (DHE) Merck.

4.2 NIR Irradiation

Hydrosun w575 with a wIRA filter was used as a NIR irradiation source. Irradiation experiments were conducted with a working distance of 37 cm, ensuring the following irradiance composition: Vis (590–780 nm) 46.04 mW/cm−2 (26.59%), NIR (780–1400 nm) 126.53 mW/cm−2 (73.09%), IR (> 1400 nm) 0.55 mW/cm−2 (0.32%) [98].

4.3 Nitrogen-Doped Graphene (NGA) Preparation

The synthesis of NGA consists of two steps. In the first step, nitrogen-doped graphene (NG) is prepared; 50 g of fluorographite was dispersed in a glass flask in 1500 mL of DMF and stirred for 72 h, then sonicated for 24 h. Then 150 g of NaN3 was added to the previous mixture, transferred to a spherical flask, and left stirring and heating at 130°C for 3 days with a condenser in the hood. After the end of the reaction, the sample was washed in falcons with DMF (3×), acetone (3×), ethanol (3×), distilled water (3×), and hot distilled water (2×), using centrifugation (14 000 rcf). In the second step of the synthesis, NG is oxidized to form nitrogen-doped graphene acid (NGA); 10 g of the previously prepared N-doped graphene was treated with 45% HNO3 for 48 h at 100°C in a glass flask with the condenser. After completion of the reaction, the product was purified by washing with filtration (Whatman cellulose membrane filter, pore size 0.2 μm); 5 times with hot distilled water, 5 times with distilled water, and purified by dialysis (dialysis tubing cellulose membrane, 14 kDa cutoff).

4.4 Embedding of NGA Into PVA-DBA Matrix

The matrix was prepared according to Nishiyabu et al. with slight modifications [75]. NGA stock dispersion (5 mg/mL) was prepared in Milli-Q water, sonicated in a bath for 30 min, and stored at room temperature. This stock dispersion was sonicated for 5 min before each use. A 249 mg of PVA (0.026 mmol) were dissolved in 6.9 mL Milli-Q Water at 60°C, stirring continuously for at least 2 h. Afterward, 6.6 mL of EtOH was added, and either 1.5 mL of water (PVA) or NGA stock dispersion (NGA PVA). As a crosslinking stock solution, 170 mg of DBA (1.03 mmol) were dissolved in 25 mL EtOH. The polymeric solutions (PVA or NGA PVA) and the DBA solution were mixed in a 75:25 ratio, respectively (PVA 1.3 μM, DBA 10 μM, NGA 0.375 mg/mL). This mixture is then used to drop-cast the films onto the desired surfaces. Before coating, all glass slides of varying sizes were cleaned via sonication for 5 min in Milli-Q water and dried with tissue paper. The prepared PVA-NGA-DBA or PVA-DBA solution was then distributed 40 and 17 μL on 15 and 10 mm diameter round coverslips (Novoglas Coverslips 01–0015/05 and 01–0010/05), respectively (22 μL/cm2). After drying, the same amount of DBA solution was added. Finally, after this second drying, the glass slides were washed twice by dipping 10 times into Milli-Q and 10 times into EtOH; this step was performed twice. Afterward, the coating was brushed, dried with paper, and left to dry.

4.5 Characterization of NGA and NGA/PVA-DBA Coating

A profilometer (Dektak XT, Bruker) with a 12.5 μm stylus radius was used to determine the thickness and estimate the roughness of the coatings. Coating density was determined by measuring the weight of coverslips with a known surface area before and after coating using a microbalance. Raman measurements were performed using an alpha300R WITec confocal Raman spectrometer with a 532 nm laser excitation source, 5 mW, and 0.5 acquisition time, 600 g/mm. Spectra have been recorded with a 50× magnification, mapping a surface 50 × 50 μm with an xy resolution of 1 μm. After recording, spectra have been processed with WITec Control 6.1 using baseline subtraction and multicomponent analysis. The contact angle was measured in static mode by depositing distilled water drops (2 μL) with a Krüss (DSA25) device. FTIR spectra were recorded on an iS5 FTIR spectrometer (Thermo Nicolet) using the Smart Orbit ZnSe ATR accessory. Briefly, a droplet of water dispersion of the relevant material was placed on the ZnSe crystal and dried. The spectra were then acquired by summing 52 scans using a nitrogen gas flow through the ATR accessory. ATR and baseline correction were applied to the collected spectra.

The NGA sample was characterized by transmission electron microscopy (TEM) using a JEOL 2100 TEM instrument (JEOL) with an accelerating voltage of 200 kV. High-resolution transmission electron microscopy (HR-TEM) was performed with a HR-TEM TITAN 60–300 microscope with an X-FEG type emission gun, operating at 300 kV. The coating morphology was examined through scanning electron microscopy (SEM) JEOL 7900F microscope (JEOL, Japan), with an accelerating voltage of 5 kV. Fourier transformed infra-red (FTIR) spectra were obtained with an iS5 FTIR spectrometer by Thermo Nicolet, which included a Smart Orbit ATR accessory featuring a ZnSe crystal. For this, 20 μL of the sample dispersed in distilled water was applied to the ZnSe crystal and allowed to air-dry. X-ray photoelectron spectroscopy (XPS) measurements were performed using a Nexsa G2 spectrometer from Thermo Fisher Scientific, equipped with an Al Kα radiation source. The resulting data were analyzed with Avantage software. Raman spectroscopic data were acquired using a DXR Raman microscope. This process involved employing a 780 nm excitation line from a diode laser. EPR spectra were collected on an x-band (∼9.14–9.17 GHz) spectrometer, JEOL JES-X-320, equipped with a variable He temperature set-up ES-CT470 apparatus. The experimental temperature was set to 78 K. The quality factor (Q) was kept above 650. High purity quartz tubes (Suprasil, Wilmad, ≤ 0.5 OD) and the accuracy of the g-values was determined by comparison with a Mn2+/MgO standard (JEOL standard). The microwave power was set to 1.0 mW to avoid any power saturation effects. A modulation width of 0.35 mT and a modulation frequency of 100 kHz were used. EPR spectra were collected with a time constant of 30 ms and a sweep time of 2 min with two accumulations to improve the signal-to-noise ratio.

4.6 ROS Detection

DHE was employed to assess which ROS species are produced by the coatings and if the ROS species production can be induced by irradiation with a NIR light source (namely Hydrosun w575). Stock solutions of 10 μM of DHE were prepared in Milli-Q water. The assays were performed in 6-well plates. The samples on 21 × 26 mm cover slips were placed into the 6-well plates, and 3 mL of the desired probe solution was added to each well. For ambient light assessments, 100 μL from each well was transferred into a 96-well plate and stored in the dark until the next sampling time point. The first sampling was at time point 0–30 s of adding the probe, and then sampling occurred every 2 min to see the probe degradation over time. After the desired assay time, the 96-well plate was measured with a BertholdTech Mithras2 plate reader at the excitation wavelength of 355 nm and emission wavelength of 420 nm for DHE, and at 404 nm excitation. The same setting was used for NIR light irradiation. The cyclic stability of the ROS generation was assessed with the DHE probe at irradiation intervals of 5 min. Sampling was done before and after each round of irradiation, and after each irradiation the DHE solution was discarded; the coated glass was washed with Milli-Q water before adding fresh 10 μM DHE solution and sampling time point 0 for the next cycle of irradiation.

4.7 Photothermal Activity

All samples were placed on an empty petri dish on pipette tips resting on parafilm to ensure the least amount of heat transfer through contact. To calculate the thermal efficacy, the glass slides were positioned so that the coating was face down. One cycle consisted of irradiation with the NIR lamp for 10 min and a cooling period of 10 additional minutes, after which another irradiation was started. This was repeated for up to 20 cycles. During this assessment, a thermal camera FLIR Systems Thermovision A40 recorded the temperature changes on all samples. A simultaneous irradiation of NGA samples, PVA-DBA matrix without the NGA particles, and of the glass itself was performed. The thermal data were analyzed using FLIR Tools software, calculating the average of the temperature for each glass substrate. The photothermal [99] efficiency of the films was calculated according to the equation:

η = I film I lamp = Cm ∆ T P lamp St × 100 % , $$ \upeta =\frac{{\mathrm{I}}_{\mathrm{film}}}{{\mathrm{I}}_{lamp}}=\frac{\mathrm{Cm}\Delta T}{{\mathrm{P}}_{\mathrm{lamp}}\mathrm{St}}\times 100%, $$ (1)

where η is the photothermal conversion efficiency, Ifilm is the heat generated by the light on the surface of the film. Ilamp is the heat of the lamp reaching the surface of the film. C is the specific heat capacity of the glass (0.84 J g−1 K−1), m is the weight of the coated glass, ∆T is the change in glass temperature values, Plamp is the power of the light irradiation (200 mW/cm2 according to the producer), S is the area of the glass and t is the irradiation time.

4.8 Determination of Antibacterial Activity

To analyze the antibacterial activity of the fabricated coatings, S. aureus ATCC 6538 and P. aeruginosa ATCC15442 were used. A single bacterial colony was picked from a plate counter agar (PC-Agar) plate and added to 5 mL of a solution containing 30 wt.% Tryptic soy broth (TSB). The suspensions were incubated at 37°C and 160 rpm overnight. The bacterial culture was diluted with 30 wt% TSB to above 106 colony-forming units (CFU)/mL and further grown for 1.5 h to obtain exponentially growing cells, which were then diluted with phosphate-buffered saline (PBS, Sigma P4417) to 105 CFU/mL for both S. aureus and P. aeruginosa. A 200 μL of the prepared bacterial suspensions were loaded onto the disks and incubated for 2 h at 37°C without shaking. For NIR treatment, the samples were irradiated with Hydrosun w575 with a wIRA filter for 10 min, using untreated samples as a comparison. The bacterial suspension was removed from the coupon surface, and the coupons were washed twice with 400 μL PBS. The suspension was pooled with the wash solution, and serial dilutions of 1:10 were performed, followed by subsequent plating on PC-Agar. The plates were incubated overnight at 37°C. An automatic colony counter (Scan300Interscience) was used for colony counting.

4.9 Preparation of AgOx/TiOx

As a positive control for the antibacterial activity, a nanostructured combination of silver oxide and titanium oxide was prepared by low-pressure plasma technology. The full process was carried out in the same plasma chamber, a pilot-scale reactor located in Empa St. Gallen that allows plasma cleaning, material deposition, and oxidation. Briefly, the sample consists of silver oxide nano islands on nonstoichiometric titanium oxide with an overall thickness of around 60 nm. To prepare the sample, first titanium and then silver targets were sputtered (i.e., using the magnetron sputtering technique), followed by plasma oxidation in an Ar/O2 environment. In the discussion, we refer to this material as AgOx/TiOx. This combination of metal oxides is activated, producing ROS when H2O and O2 molecules reach the surface. Because of the plasma oxidation of Ag, ion rele