Development of photoactivable glycerol-based coatings containing quercetin for antibacterial applications

Abstract

The development of new antibacterial coatings (against Escherichia coli and Staphylococcus aureus) using a natural dye, quercetin, according to a green chemistry process was investigated. Quercetin was used as both a photosensitizer and antibacterial agent. The synthesized material was developed according to a cationic photopolymerization process under light irradiation. The photosensitizing mechanism involving quercetin and an iodonium-based cationic photoinitiator was described for the first time according to steady state photolysis and fluorescence experiments. The resulting coatings showed excellent adhesion on a stainless steel plate as demonstrated by nanoindentation and scratch tests, with a high thermal stability up to 375 °C. Finally, a primary investigation was conducted to assess the antibacterial properties of the glycerol-derived coatings against Escherichia coli and Staphylococcus aureus under light illumination. Electron paramagnetic resonance spectroscopy confirmed the generation of reactive oxygen species, such as singlet oxygen, which is responsible for inhibiting bacteria proliferation.

---

1. Introduction

Contamination by microorganisms is of great concern in many fields such as health care products, medical devices, hospital surfaces/furniture, or surgery equipments. This problem is associated with an annual mortality of many thousands of people in the US and a constant increase in health-care costs.^1,2 To solve the problem of bacteria resistance, the development of new antimicrobial systems in the biomedical industry for fighting infections has attracted considerable attention. The main strategies are based on the decrease of the capacity of bacteria to attach to a surface,^3–7 the lethal contact^8–10 which induces the biochemical death of bacteria, the biocide leaching^11–15 and finally the disinfection of surfaces using coatings that produce reactive oxygen species^16–19 (ROS) upon light activation, also known as photodynamic inactivation (PDI). In this latter method, the light excitation of a photosensitizer (PS) generates ROS which subsequently cause cell death. In this context, PSs do not have to penetrate the bacterium or even come in contact with the cell to be effective. Such antibacterial surfaces that work according to the PDI principle could help to reduce the transmission of multi-resistant microorganisms,^20–23 which is of great importance in hospital hygiene, and can offer a constant prevention of microorganism adhesion and proliferation on any surface.²⁴ Some investigations have already been performed essentially with porphyrins and phthalocyanine derivatives,^25,26 hydroxyethyl Michler's ketone,²⁷ toluidine blue,²⁸ eosin Y,^29,30 methylene blue^28,31,32 or rose bengal.^30,33 However, the use of natural dyes instead of the aforementioned synthetic dyes is an interesting environmental challenge. To our knowledge, there is no investigation on the synthesis of antibacterial coatings using natural dyes which could act as a “bacteria killer” according to the PDI process. This lack of studies prompted us to develop new antibacterial coatings based on a natural dye. Thus, the originality of this study relies on two points: (i) the use of a natural flavonoid dye, quercetin, which could be used as a reactive agent with dual functionality, i.e. an antibacterial promoter and a photosensitizer under light activation, and (ii) the photochemical studies of (quercetin/iodonium salt) photoinitiating system (used for the initiation of the polymerization and the synthesis of the coatings) which have never been investigated yet.

Quercetin (3,5,7,3′,4′-pentahydroxyflavone) is one of the most abundant flavonol-type flavonoids^34–36 found in many common fruits and vegetables (apple, grape, lemon, tomato, onion, etc.), beverages (tea, red wine), olive oil, and propolis from the bee hives. Quercetin has many health-promoting effects,^37–39 including anti-inflammatory and anti-allergic effects, as well as the protection against cardiovascular health and cancer risk. In addition, it has been reported that quercetin enhances the antiviral activity³⁹ owing to its strong antioxidant action and can be used as a skin antioxidant protector against UV radiation.^{35,36,40–42}

This investigation presents a simple route to efficiently synthesize an antibacterial coating derived from quercetin and glycerol monomer, using environmental-friendly method (photochemistry process), followed by a complete study of the (quercetin/iodonium salt) photoinitiating system to perform the synthesis of the quercetin-containing glycerol coatings.

The first part of this study focuses on the photochemical properties of the photoinitiating system (quercetin/iodonium salt) using steady state fluorescence and electrochemistry experiments. In the second part, the thermal and mechanical properties of the quercetin derivated coatings were characterized by thermo gravimetric analysis, nanoindentation and scratch tests. In a third part, the antibacterial properties of the coatings against Escherichia coli and Staphylococcus aureus were evaluated and correlated with the ability of the residual quercetin embedded in the film to photogenerate singlet oxygen.

2. Results and discussion

2.1 Synthesis of the glycerol triglycidyl ether (GTE)

This monomer was synthesized using a two-step procedure. In the first step, glycerol was converted to triallyl glycerol by reacting it with allyl bromide and sodium hydride in dry DMF via a S_N2 reaction. After purification, the structure of the molecule was confirmed by ¹H nuclear magnetic resonance (NMR) spectroscopy with the appearance of allylic signals at 5.91, 5.27 and 5.16 ppm, respectively (Fig. 1A). The double bonds of triallyl glycerol were subsequently epoxidized using m-CPBA in CH₂Cl₂. The reaction was monitored by ¹H NMR spectroscopy. Fig. 1B shows the total disappearance of the ethylenic signals along with the appearance of signals related to epoxide moieties at 3.14, 2.78 and 2.59 ppm, respectively. The ¹H NMR spectrum of glycerol triglycidyl ether appears more complex than that of triallyl glycerol. This is due to the non-stereospecificity of m-CPBA epoxidation, which leads to the formation of isomers.

Fig. 1 ¹H NMR spectra of (A) triallyl glycerol and (B) glycerol triglycidyl ether.

2.2 Photophysical fingerprint of quercetin

Fig. 2 displays the normalized absorption and fluorescence spectra of quercetin in acetonitrile. The low energy side of the absorption spectrum is dominated by a distinctive band at 370 nm with moderate intensity (ε_max ∼ 19 000 M⁻¹ cm⁻¹).

Fig. 2 Normalized absorption and fluorescence spectra of quercetin in acetonitrile. Note: Raman scattering peaks of acetonitrile are indicated by asterisks in the fluorescence spectrum of quercetin.

The fluorescence spectrum shows a maximum located at 535 nm leading to a significant Stokes shift of approximately 8350 cm⁻¹. Such a large Stokes shift is a clear indication of a significant electronic change between ground and excited state. Moreover, quercetin is a very low emissive chromophore that exhibits a fluorescence quantum yield of less than 10⁻³ in acetonitrile. Similarly to the 3-hydroxyflavone^43,44 and its 4′-substituted derivatives,⁴³ the fluorescence of quercetin⁴⁴ stems from the radiative deactivations of two relaxed excited species which are in fast equilibrium at singlet excited state (Scheme 1).

Scheme 1 Schematic representation of the various radiative deactivation pathways of quercetin at S₁ state (FC stands for: Franck Condon state).

The first emitting species (blue emitting one) corresponds typically to an internal charge transfer (ESICT) from catechol to pyrone groups. Subsequent to this latter electronic relaxation, an internal proton transfer process occurs from the 3-hydroxyl function to the carbonyl oxygen atom, giving rise to the emission of an internal proton transfer (ESIPT) tautomer. Such a new fluorescent species exhibits a benzopyrylium-like configuration. It should be noted that the fluorescence spectra of ESICT and ESIPT are strongly overlapped in such manner that only one emission band is observed for quercetin. However, we will see hereafter that the presence of iodonium salts will lead to a specific quenching of the ESIPT species.

2.3 Photosensitizing properties of quercetin

Steady state photolysis experiments in acetonitrile were first performed to demonstrate the photosensitizing role of quercetin in the presence of iodonium salts. Fig. 3 displays the absorbance time profile at 370 nm of a solution of quercetin in the presence of iodonium salts. The kinetics relative to a reference solution of quercetin without a photoacid generator is also depicted to highlight the effects. In absence of iodonium salts, the absorption spectrum of quercetin remains constant upon irradiation whereas the presence of photoacid generator clearly induces the rapid photobleaching of quercetin. For instance, the intensity of the longest wavelength absorption band of quercetin collapses by 97% after 17 min irradiation. Fig. S2† shows the evolution of the UV-visible spectra of quercetin alone and quercetin/Iod.

Fig. 3 Plots of the absorbance monitored at 370 nm as a function of the irradiation time for the two solutions of quercetin (3.5 × 10⁻⁵ M in acetonitrile non-degassed) without (A) and with (B) iodonium salts ([Iod] = 8.6 × 10⁻⁴ mol L⁻¹) (λ_irr: 365 nm, I = 70 mW cm⁻²).

A reaction between excited quercetin and iodonium salts is therefore demonstrated. Interestingly, such a reaction leads to the concomitant photogeneration of H⁺ which can be detected using an acid indicator, such as rhodamine B^45,46 (RhB). As shown in Fig. 4A, the continuous irradiation of a solution of quercetin mixed with iodonium salts and RhB leads to the progressive growth of an absorption band in the 480–650 nm range corresponding to the acid form of RhB (RhBH⁺). The photogeneration of H⁺ is observed only when quercetin and iodonium salts are mixed together. This clearly confirms the photosensitizing properties of quercetin. Moreover, the amount of H⁺ detected during the irradiation of a solution that was initially N₂-saturated is globally equivalent to that of a non-degassed solution, as illustrated in Fig. 4B. It is clear that the presence of dissolved oxygen in acetonitrile ([O₂]_diss ∼ 10⁻³ M (ref. 47)) does not have any inhibiting effects on the photoreaction between the excited quercetin and the iodonium derivative. This suggests that the photosensitizing process occurs mainly from the singlet excited states of quercetin.

Fig. 4 (A) Evolution of the absorption spectrum of quercetin (5 × 10⁻⁵ M in non-degassed ACN) in presence of Ph₂I⁺, PF₆⁻ (1.5 × 10⁻³ M) during irradiation at 365 nm (I = 10 mW cm⁻²). RhB (1 × 10⁻⁵ M) is added as acid indicator. Inset: evolution of the absorbance of RhBH⁺ at 550 nm. (B) Evolution of the absorbance of the protonated RhBH⁺ during irradiation of a non-degassed and a N₂-saturated solution.

This assumption is corroborated by the fluorescence quenching of quercetin in the presence of increasing amounts of iodonium salts (see Fig. 5). The fluorescence spectrum clearly shows a decrease in intensity. It should be emphasized that this quenching mainly affects the ESIPT band which is located in the 530–680 nm range. As a consequence, after addition of iodonium salts, the relative contribution of ESIPT band to the total fluorescence spectrum drops drastically in such manner that the low intensive ESICT band can be detected in the 460–500 nm range. Noteworthy, the excitation spectrum collected from this latter emission band matches the absorption band of quercetin.

Fig. 5 Evolution of the fluorescence spectrum of quercetin upon gradual addition of Iod in acetonitrile (λ_ex: 395 nm). Raman scattering peaks of acetonitrile are indicated by asterisks.

According to the strong electron-accepting characteristics of iodonium, it is proposed that the fluorescence quenching of quercetin can be mainly ascribed to an efficient photoinduced electron transfer (PeT) from the ESIPT state of quercetin to the photoacid generator. The free energy associated to this proposed mechanism can be estimated according to the Rehm–Weller equation:⁴⁸

ΔG_eT = E_ox − E_red − E_ESIPT + C

The following approximations were made: (i) the coulombic part of the stabilization energy (C) is negligible, (ii) the energy of ESIPT state can be estimated by E_ESIPT ≈ 1/2hc (ν_abs + ν_fluo) leading to a value of ca. 2.84 eV; and (iii) E_ox and E_red correspond to the oxidation potential of the quercetin and the reduction potential of the diphenyliodonium salt (E_red = −0.2 V vs. SCE⁴⁷). The oxidation potential of quercetin in acetonitrile was measured by cyclic voltammetry. According to the cyclic voltammogram of the dye (see Fig. S3†), the first oxidation wave, which is irreversible, exhibits a half-wave potential of about 1.1 V vs. SCE. As a consequence, ΔG_eT has an estimated value of ca. −1.54 eV in acetonitrile which is consistent with a largely exergonic PeT process.

Therefore, quercetin, as a natural dye, constitutes an interesting candidate that can promote acid photogeneration in the presence of iodonium salts. This photosensitizing process will be addressed to produce quercetin-derivative coatings by cationic photopolymerisation.

2.4 Synthesis and characterization of the quercetin-derivative coatings

The photochemical reactivity of the photosensitive formulation containing glycerol triglycidyl ether (GTE)/quercetin/iodonium was evaluated by Real-Time Fourier Transform Infrared (RT-FTIR) spectroscopy by monitoring the decrease of the epoxy group at 910 cm⁻¹ upon irradiation (Fig. 6A). The band associated with the epoxy group decreases concomitantly with the growth of a new absorbance at 1080–1100 cm⁻¹ (Fig. 6B). This latter band confirms the formation of the polyether network, and leads to the appearance of a new band between 3100 and 3600 cm⁻¹ due to the –OH group of the polyether network (Fig. S4†). These results demonstrate the photopolymerization of GTE. As shown in Fig. 6A, the final epoxy conversion is approximately 75% after 1200 s of irradiation. The coating is totally take-free. An optical image of the coating deposited on a stainless steel plate is displayed on Fig. 7.

Fig. 6 (A) Photopolymerization kinetic of GTE under light activation and (B) FTIR spectra of the coating during the 1200 s of irradiation. Xe lamp, intensity = 70 mW cm⁻², film thickness = 100 μm.

Fig. 7 Optical image of the stainless steel substrate (left) and the quercetin-derivative coating deposited on a stainless steel substrate.

2.5 Thermogravimetric analyses

Thermal stability of the coatings was studied by thermogravimetric analysis (TGA). The results are displayed in Fig. 8. The coating reached 50% weight loss at 375 °C.

Fig. 8 TGA thermogram of the photoinduced quercetin-derivative coating.

2.6 Mechanical analyses of the coatings

The hardness of the glycerol-derived coatings was first determined relative to a standard set of pencil leads. Considering that the pencil hardness is always above 7H, it means that no macroscopic scratches were observed up to a hardness of 7H. To highlight more precisely the hardness of the coatings, further investigations, i.e. nanoindentation and scratch tests, were performed.

2.6.1 Nanoindentation. The loading and unloading curves during the nanoindentation tests were superimposed, showing that the coatings exhibit a viscoelastic behavior at room temperature, indicating that the polymer is in a rubber-like state at room temperature. The elastic modulus and the hardness, as measured by the Oliver and Pharr method⁴⁹ was approximately 189 ± 18 MPa (mean ± standard deviation) and 12 ± 1 MPa, respectively.

2.6.2 Scratch tests. Regarding the scratch tests, a comparison of the height profiles before and after scratch tests reveals the complete recovery of the material, confirming the rubber-like behavior (Fig. 9). The lack of bulging on the height profile during and after scratching indicates the apparition of brittle fracture or delamination of the coating. Moreover, the optical images of the sample after scratching show no residual imprint, confirming that neither brittle fracture nor delamination occurs. Hence, there is a good adhesion to the substrate and the good resistance of the coatings to brittle fracture.

Fig. 9 Height profile before, during and after scratch. A complete recovery of the sample after scratch and the absence of bulge in the profile during scratch evidence the rubber like behavior of the coatings and its very good adhesion and resistance to brittle fracture.

2.7 Fluorescence of the coatings

After irradiation of the quercetin derived formulation (900 s), unreactive quercetin molecules are left inside the coating, as observed by UV-vis spectrometry (Fig. 10A, dot line) and its epifluorescence (Fig. 10B). The remaining fluorescence of quercetin can be used to generate reactive oxygen species (ROS) able to eliminate bacteria, as it will be demonstrated in the following section.

Fig. 10 (A) Evolution of the UV-vis spectra of the quercetin derived coating after 900 s of irradiation. Solid line = before irradiation, dot line = after irradiation. Xe lamp, intensity = 70 mW cm⁻². Formulation was sandwiched under 2 glass plates. (B) Epifluorescence of quercetin located into the polymer coating after 900 s of irradiation. Inset = fluorescence spectrum of the embedded quercetin into the polymer film after polymerization.

2.8 Antibacterial properties

Prior to investigate the antibacterial effect of the quercetin derivative-coatings, the ability of quercetin to inhibit bacterial proliferation in solution was investigated against Gram-negative (E. coli) and Gram-positive (S. aureus) bacteria after 2 h and 6 h, with and without light illumination. Initially, 6.5 × 10⁶ CFU mL⁻¹ of each bacteria was introduced in solution at t = 0 h.

First, a reference study referring to the light irradiation against the two bacteria strains was performed (Fig. 11 and S5†). The results show that the irradiation does not have any influence on the proliferation of E. coli and S. aureus alone. The two strains proliferate in the same way with or without irradiation. A second reference study was performed in the presence of DMSO, because quercetin is insoluble in water. According to these results, DMSO does not affect the bacterial activity because the decrease in CFU is not significant regardless of the stains used with or without light illumination (Fig. 11 and S5†).

Fig. 11 Influence of the incubation time (0, 2 h, 6 h) and the irradiation on the growth of (A) E. coli and (B) S. aureus. [Quercetin] = 500 mM.

The last interesting results are the effect of light irradiation on the proliferation of bacteria. The antibacterial effect of quercetin was then evaluated in DMSO, under light activation. The results show that irradiated quercetin only leads to an antibacterial effect with S. aureus. Indeed, the development of Gram-negative bacteria (E. coli) is neither inhibited by the illumination nor affected by the presence of quercetin in the dark. On the contrary, the irradiated quercetin solutions allow the total death of S. aureus after 2 h of incubation. The antibacterial effect against S. aureus remains efficient even after 6 h. The variation in quercetin killing efficacy at different exposure durations was statistically significant (p value < 0.05 on comparing percentage survival). In the absence of irradiation, more than half of bacteria are always alive after 6 h of incubation with S. aureus.

Some studies⁵⁰ on the inactivation of bacteria demonstrate that the photosensitizers do not need to penetrate the cell membrane or even to come into contact to be effective. Indeed, if sufficient quantities of singlet oxygen can be generated near the outer membrane of the bacteria, it will lead to its damage. However, only Gram-positive bacteria (S. aureus) are affected by this oxygen singlet process. Indeed, the lipopolysaccharide (LPS) coatings of the cell wall of the Gram-negative bacteria offers some protection from the toxic effects of exogenous agents.^51,52 In addition to possibly forming a structural barrier to penetration, this outer membrane may form a chemical trap for singlet oxygen; it is composed of unsaturated fatty acids and proteins, which are compounds known to react chemically with singlet oxygen.^53,54

As a result, some strains that fail to produce a large portion of the LPS displayed greater sensitivity to exogenous singlet oxygen than the strains with this capability.⁵³ Most Gram-positive bacteria (S. aureus) lack this protective structure analogous to the Gram-negative LPS and the outer membrane in which it is anchored. This can explain the 99% inhibition of the proliferation of S. aureus after 2 h of incubation. ROS and particularly singlet oxygen probably diffuses readily through the relatively open structure of the peptidoglycan layer of the S. aureus cell wall to react with the vital target.

In this study, Electron Paramagnetic Resonance (EPR) spectroscopy was used to monitor the ROS generation ability of quercetin upon photo-illumination (Fig. 12). Amount of 2,2,6,6-tetramethylpiperidine (TEMP), a diamagnetic molecule, was used to capture oxygen radical species by yielding a paramagnetic product, the nitroxide radical TEMPO.^55–57 As shown in Fig. 12, the EPR spectral signal of three lines of equal intensity, which is attributed to the TEMPO nitroxide radical, was observed when an oxygen-saturated solution of quercetin was irradiated in the presence of TEMP at room temperature. During illumination of quercetin solution, the intensity of the EPR signal increases gradually, indicating the formation of reactive oxygen species like singlet oxygen.

Fig. 12 A) ESR spectra obtained under light irradiation of quercetin in tert-butylbenzene in aerated conditions after 240 s. The increase of the ESR spectrum is ascribed to TEMPO creation. Light intensity = 270 mW cm⁻². (B) Evolution of the generated TEMPO radicals in solution during light activation.

According to these results, it is likely that quercetin, which was introduced and immobilized in the coatings, could generate reactive oxygen species like singlet oxygen to inhibit the proliferation of bacteria on stainless steel substrates. Mechanistic interpretations^46,53,58 have already demonstrated that the resulting reactive species, such as singlet oxygen could be generated by these immobilized photosensitizers on the material surface, followed by its diffusion, resulting in the damage of the bacteria envelope. Interestingly, the diffusion of quercetin through the synthesized coatings was not observed until 6 h of incubation. The ability of the quercetin derivative-coatings (deposited on stainless steel substrates) to inhibit the bacteria adhesion/proliferation was investigated with and without the light activation (Fig. 13). A quantitative and well-tried method in biological studies for quantifying the total biofilm population was used, as reported by many antibacterial investigations.^10,59,60

Fig. 13 Comparison of the antibacterial properties of quercetin-derivative coatings against (A) E. coli and (B) S. aureus with and without light activation after 2 h and 6 h of incubation.

In this case, 15 × 10⁶ CFU mL⁻¹ of each bacteria was introduced and placed in contact with the coatings at t = 0 h. According to Fig. 13, the antibacterial effect is clearly more efficient against S. aureus than for E. coli. Upon light illumination, the number of CFU S. aureus drastically decreases, whereas it increases significantly on the stainless steel substrate (Fig. S6†). After 2 h and 6 h of incubation, 99% inhibition of S. aureus adhesion was observed. On the other hand, without illumination of the coatings, the adhesion/proliferation of S. aureus remains important: the number of CFU per cm² increases from 1.5 × 10⁶ (2 h of incubation) to 10⁷ after 6 h (Fig. 13). This later result demonstrates that light is essential for leading to an antibacterial effect. For E. coli, the antibacterial property of the coating is not efficient. Indeed, whatever the conditions used (illumination or not), the number of CFU increase after 2 h and 6 h of incubation. These results are in accordance with the results obtained in solution.

This experiment highlights the efficiency of reactive oxygen species, which are generated from the illumination of the quercetin-derivative coatings, to strongly reduce the adherence/proliferation of S. aureus on stainless steel supports.

3. Conclusions

This study examined the synthesis of antibacterial coatings derived from a natural dye (quercetin) and a glycerol triglycidyl ether monomer. The ability of quercetin to be used both as a photosensitizer of iodonium salt and as an efficient antibacterial agent (even incorporated into a coating) against S. aureus was demonstrated for the first time. For this purpose, the photochemical behavior of quercetin in the presence of iodonium salts was described by steady state and fluorescence experiments, demonstrating (i) the efficiency of this system to be used as a photoinitiating couple and (ii) its ability to generate H⁺ photoacid for the cationic photopolymerization. The resulting coatings synthesized under light illumination showed very good adherence properties to the stainless steel substrates and a high thermal stability up to 375 °C.

The synthesized coatings containing quercetin led, under light activation, to an inhibition of S. aureus proliferation of 99% after 2 h and 6 h of incubation. This new coating could be used successfully to avoid bacteria proliferation and be deposited onto disposable paramedical devices, such as clamp or scalpels, which can be used one-time for a few hours.

4. Experimental

4.1 Materials

Glycerol (≥99%), sodium hydride (97%), 3-chloroperbenzoic acid (≤77%, m-CPBA), quercetin (≥95%, Qr), 2,2,6,6-tetramethyl-1-piperidinyloxy (98%, TEMPO), 2,2,6,6-tetramethylpiperidine (≥99%, TEMP), rhodamine B (≥95%, RhB) and anhydrous DMF were purchased from Sigma-Aldrich. Iodonium 4-(2-methylpropyl)phenyl-hexafluorophosphate (Iod) was purchased from Badische Anilin und Soda Fabrik (BASF). Allyl bromide (99%) was acquired from Alfa Aesar. Petroleum ether, ethyl acetate (analytical grade) and dichloromethane (synthesis grade) were obtained from Carlo Erba. All chemicals were used as received. Table 1 lists the chemical structure of the compounds used in this study.

Table 1 Chemical structure of the compounds used in this study

Compound	Function	Structure
Glycerol	Monomer precursor
Quercetin	Photosensitizer and antibacterial agent
Iodonium 4-(2-methylpropyl)phenyl-hexafluorophosphate (Iod)	Cationic photoinitiator
2,2,6,6-Tetramethylpiperidine (TEMP)	Singlet oxygen scavenger
2,2,6,6-Tetramethyl-1-piperidinyloxy (TEMPO)	Determination of ^1O2
Rhodamine B (RhB)	Acid indicator

---

4.2 Synthesis of triallyl glycerol

Triallyl glycerol was synthesized using a procedure reported elsewhere. Glycerol (5 g, 54.3 mmol) and NaH (6.84 g, 285 mmol) were dissolved in 50 mL of anhydrous DMF at 0 °C. After 10 min, allyl bromide (17 mL, 285 mmol) was added and the solution was stirred for 3 h at room temperature. The mixture was then poured into water and extracted with CH₂Cl₂. The organic phase was concentrated and the residue was purified on a silica gel column (petroleum ether–EtOAc 1 : 1.5 as eluent). The evaporation of the solvent and subsequent drying under high vacuum led to 8.272 g (yield = 72%) of triallyl glycerol. The product structure was confirmed by ¹H NMR. ¹H NMR (CDCl₃) δ (ppm) 5.91 (m, 3H), 5.27 (m, 3H), 5.16 (m, 3H), 4.15 (dt, J = 5.7, 1.4 Hz, 2H), 4 (dt, J = 5.6, 1.4 Hz, 4H), 3.69 (m, 1H), 3.54 (m, 4H).

4.3 Synthesis of glycerol triglycidyl ether (GTE)

Triallyl glycerol (5 g, 23.6 mmol) was dissolved in 50 mL of CH₂Cl₂. m-CPBA (24.42 g, 142 mmol), dissolved in 75 mL of CH₂Cl₂ was added dropwise over a one hour and the solution was stirred overnight at room temperature. After filtering the precipitate, the solution was concentrated to approximately 50 mL and placed at −20 °C. The precipitate was then filtered and the operation was repeated 3 times until no more precipitate appeared. Finally, the solution was passed through a silica gel pad, evaporated and dried under vacuum, yielding 4.93 g of glycerol triglycidyl ether (yield = 80%). The total conversion of double bonds to epoxide was confirmed by ¹H NMR spectroscopy. ¹H NMR (CDCl₃) δ (ppm) 3.90 (dt, J = 11.7, 2.7 Hz, 1H), 3.78 (dd, J = 11.6, 2.6 Hz, 2H), 3.71 (m, 1H), 3.59 (m, 5H), 3.39 (m, 2H), 3.14 (m, 3H), 2.78 (t, J = 4.6 Hz, 3H), 2.59 (m, 3H).

4.4 Photopolymerization procedure

For the cationic photopolymerization, 8 mg of Iod (4 wt% with respect to the epoxy monomer), 5 mg of Qr (2.5 wt% with respect to the epoxy monomer) were dissolved into epoxy monomer formulation (GTE, 200 mg) containing 100 μL of acetone. The kinetics of photopolymerization were followed by real time Fourier transform infrared spectroscopy (RT-FTIR) using a Thermo-Nicolet 6700 instrument. The liquid samples were applied to a BaF₂ chips by means of calibrated wire-wound applicator. The thickness of the UV-curable film was evaluated at 100 μm. The RT-FTIR analyses were carried out under air conditions. Samples were irradiated at room temperature, by means of a Lightningcure LC8-03 lamp from Hamamatsu, equipped with a xenon lamp (200 W) coupled with a flexible light guide. The end of the guide was placed at a distance of 4 cm. The maximum UV light intensity at the sample position was found to be 70 mW cm⁻². The photopolymerization was monitored by the disappearance of the epoxy function of the GTE monomer at 910 cm⁻¹. The decreases of epoxy function at 910 cm⁻¹ and the increase of polyether at 1080–1100 cm⁻¹ demonstrated the efficiency of the photopolymerization. Conversion rate was calculated with the followed equation (eqn (1)):

Epoxy conversion (%) = 100 × (A₀ − A_t)/A₀ (1)

A₀ represents the area at t = 0 s and A_t represents the area at time t.

4.5 Synthesis of the glycerol-derivative coatings

Prior to the deposition of formulation on stainless steel plate, the latter was intensively cleaned with ethanol and toluene. The photosensitized formulation containing GTE, Iod and Qr (100 μL) were then deposited and spin-coated (2000 rpm during 3 s) on the dried stainless steel plates. Both sides of stainless steel plate were irradiated at a distance of 4 cm at room temperature using a Lightningcure LC8-03 lamp from Hamamatsu, equipped with a xenon lamp (200 W) coupled with a flexible light guide. The irradiation time was fixed to 900 s per side with an intensity of 70 mW cm⁻².

4.6 UV-visible absorption and fluorescence measurements

The absorption measurements were carried out using a Perkin-Elmer Lambda 2 spectrometer. Steady-state fluorescence spectra were collected from a FluoroMax-4 spectrofluorometer. All emission spectra were spectrally corrected.

4.7 Fluorescence microscopy

An inverted microscope IX73 from Olympus equipped with a 75 W Xe Lamp housing was used. The excitation and emission light is filtered with a fluorescence mirror unit (U-FUN from Olympus) associating a band pass filter centered at 365 nm (BP360-370), a dichroic mirror (DM410) and a long pass filter (BA420IF).

4.8 Cyclic voltammetry

The oxidation potential (E_ox) of quercetin was measured by cyclic voltammetry⁶¹ using a computer-controlled Radiometer Voltalab 6 potentiostat with a three-electrode single compartment cell. The working electrode was a platinum disk. A saturated calomel electrode (SCE) used as a reference was placed in a separate compartment. Measurements were performed at 300 K, in N₂-degassed acetonitrile with a constant concentration (0.1 M) of (n-Bu)₄BF₄. Ferrocene was used as the internal reference.

4.9 Pencil hardness

The hardness of the coating was determined relative to a standard set of pencil leads. The pencil hardness measurement begins with the lowest pencil and continues up the scale to determine the maximum hardness able to scratch the surface of the glycerol-derived coatings (method: ASTMD3363-74, 2000). The surface hardness is determined by scratching the leads across the coating at a controlled angle of 45°, the value given are the lowest grade of pencil that could induce a scratch on the coating surface. The pencil hardness was measured using a no. 553 pencil hardness tester (Yasuda Seiki Seisakusho LTD.). Pencils were supplied by Staedtler Mars Lumograph 100 (Germany).

4.10 Nanoindentation and scratch tests

Nanoindentation and scratch tests were carried out on the coatings deposited on a stainless steel substrate using a Nano Indenter G200 (Agilent Technologies) with a Berkovich tip (Micro Star Technologies). Twenty nanoindentation tests per sample were performed. Samples were loaded and unloaded at constant strain rate (0.05 s⁻¹) using the Continuous Stiffness Measurement (CSM) method until an indentation depth of 1 μm was reached. The unloading stage was performed after a hold load plateau of 300 s to exhibit the viscous behavior. Twenty scratch tests were performed; face forward, with an increasing load from 0.1 to 100 mN for a scratching distance of 500 μm. The distance between two scratches was fixed to 500 μm.

4.11 Thermogravimetric analyses (TGA)

10 mg of the glycerol-derivative films were introduced into aluminum pans and analyzed using a Setaram Setsys Evolution 16 thermobalance by heating the samples at a rate of 15 °C min⁻¹ from 20 to 800 °C under argon atmosphere.

4.12 Antibacterial properties of the coatings

The initial adhesion assays were performed using two strains of bacteria, namely E. coli ATCC25922 and S. aureus ATCC6538 on the quercetin derivative-coatings. Before the in vitro antibacterial tests, the bacterial strains were grown aerobically overnight in Luria–Bertani broth at 37 °C with stirring. Overnight cultures of E. coli and S. aureus grown in Luria–Bertani broth were diluted to an optical density (OD 600 nm) of 0.05 in sterile LB broth. At this point, the stainless steel supports and stainless steel substrates with coatings (1.5 cm × 1.5 cm) were immersed in the culture. The corresponding vials were placed on a slantwise rotating wheel to avoid the sedimentation of bacteria, incubated for different times (2 h and 6 h) and shaken at 150 rpm to allow initial adhesion to occur (INFORS AG-CH 4103, Bottmingen-Basel, Switzerland). During the incubation time, some samples were illuminated with 4 lamps (intensity = 170 μmol m⁻² s⁻¹). The emission spectrum of the lamp is given in Fig. S1.† After adhesion, the samples were rinsed seven times with sterile saline solution (NaCl, 0.9% w/v) to remove any non-adherent cells.

Colonized native and treated samples were then transferred to 2 mL sterile saline (solution A) and vortexed vigorously for 30 s. The samples were then transferred to 2 mL sterile saline (solution B) and sonicated in a Branson 2200 sonicator for 3 min. Samples were transferred once more to 2 mL sterile saline (solution C) and vortexed vigorously for 30 s. Suspensions A, B and C were pooled, serially diluted and plated on PCA medium for viable counting. The cells removed during these three phases represent the loosely attached biofilm population. A 100 μL volume of the detached viable bacteria solution was introduced onto the surface of a plate count agar plate. The process was repeated through a succession of 24 pre-dried substrates. Finally, the total bacterial adhesion was determined by a counting of the CFUs, after overnight static incubation of the agar plates at 37 °C. Each experiment was conducted four times. The levels of adhesion are given as the number of cells per square centimeter.

4.13 Antibacterial properties of quercetin in solution

The reference solution (bacteria/DMSO) and the quercetin solution in DMSO (160 μM) containing bacteria were placed on a slantwise rotating wheel to avoid the sedimentation of bacteria, and incubated for 2 h and 6 h at room temperature under illumination or not. The volume of DMSO corresponds to 3% v/v of the total bacteria solution. The process is the same as described in the previous paragraph. The suspensions were pooled and diluted serially. A 100 μL volume of the viable bacteria solution was introduced onto the surface of a plate count agar plate.

Finally, the number of viable bacteria was determined by a counting of the CFUs, after overnight statically incubation of the agar plates at 37 °C. Each experiment was done four times.

4.14 Statistical analysis

All values corresponding to the anti-adherence property of E. coli and S. aureus are expressed as mean ± standard deviation. Statistical analysis was performed using Student's t-test to calculate the significance level of the data. Differences were considered statistically significant at p < 0.05. Ten samples per group were evaluated.

Acknowledgements

We would like to thank CNRS institute and University of Paris-Est Creteil (UPEC) for financial support, Léon Preira for the cutting process of stainless steel substrates and Séna Hamadi for the TGA analysis.

References

1. R. M. Klevens, J. R. Edwards, C. L. J. Richards, T. C. Horan, R. P. Gaynes, D. A. Pollock and D. M. Cardo, Public Health Rep., 2007, 122, 160.
2. M. F. Sampedro and R. Patel, Infect. Dis. Clin., 2007, 21, 785.
3. I. P. Parkin and R. G. Palgrave, J. Mater. Chem., 2005, 15, 1689.
4. J. Genzer and K. Efimenko, Biofouling, 2006, 22, 339.
5. G. Cheng, Z. Zhang, S. Chen, J. D. Bryers and S. Jiang, Biomaterials, 2007, 28, 4192.
6. G. M. Manecka, J. Labrash, O. Rouxel, P. Dubot, J. Lalevée, S. A. Andaloussi, E. Renard, V. Langlois and D. L. Versace, ACS Sustainable Chem. Eng., 2014, 2, 996.
7. R. Poupart, A. Haider, J. Babinot, I. K. Kang, J. P. Malval, J. Lalevée, S. A. Andalloussi, V. Langlois and D. L. Versace, ACS Biomater. Sci. Eng., 2015, 1, 525.
8. M. Ignatova, K. Starbova, N. Markova, N. Manolova and I. Rashkov, Carbohydr. Res., 2006, 341, 2098.
9. M. Ignatova, Z. Petkova, N. Manolova, N. Markova and I. Rashkov, Macromol. Biosci., 2012, 12, 104.
10. S. El Habnouni, V. Darcos, X. Garric, J.-P. Lavigne, B. Nottelet and J. Coudane, Adv. Funct. Mater., 2011, 21, 3321.
11. C. Lorenzini, A. Haider, I.-K. Kang, M. Sangermano, S. Abbad-Andalloussi, P.-E. Mazeran, J. Lalevée, E. Renard, V. Langlois and D.-L. Versace, Biomacromolecules, 2015, 16, 683.
12. T. T. Ruckh, R. Oldinski, D. A. Carroll, K. Mikhova, J. D. Bryers and K. C. Popat, J. Mater. Sci.: Mater. Med., 2012, 23, 1411.
13. J. Lu, M. A. Hill, M. Hood, D. F. Greeson, J. R. Horton, P. E. Orndorff, A. S. Herndon and A. E. Tonelli, J. Appl. Polym. Sci., 2001, 82, 300.
14. D. Chung, S. E. Papadakis and K. L. Yam, Int. J. Food Sci. Technol., 2003, 38, 165.
15. M. L. W. Knetsch and L. H. Koole, Polymers, 2011, 3, 340.
16. M. J. Niedre, A. J. Secord, M. S. Patterson and B. C. Wilson, Cancer Res., 2003, 63, 7986.
17. X. Ragàs, L. P. Cooper, J. H. White, S. Nonell and C. Flors, ChemPhysChem, 2011, 12, 161.
18. X. Ragas, M. Agut and S. Nonell, Free Radical Biol. Med., 2010, 49, 770.
19. R. Cahan, R. Schwartz, Y. Langzam and Y. Nitzan, Photochem. Photobiol., 2011, 87, 1379.
20. V. Decraene, J. Pratten and M. Wilson, Curr. Microbiol., 2008, 57, 269.
21. V. Decraene, J. Pratten and M. Wilson, Infect. Contr. Hosp. Epidemiol., 2008, 29, 1181.
22. S. Ismail, S. Perni, J. Pratten, I. Parkin and M. Wilson, Infect. Contr. Hosp. Epidemiol., 2011, 32, 1130.
23. V. Decraene, J. Pratten and M. Wilson, J. Gen. Appl. Microbiol., 2008, 54, 195.
24. M. Wilson, Infect. Contr. Hosp. Epidemiol., 2003, 24, 782.
25. A. Felgentrager, T. Maisch, A. Spath, J. A. Schroder and W. Baumler, Phys. Chem. Chem. Phys., 2014, 16, 20598.
26. S. Yano, S. Hirohara, M. Obata, Y. Hagiya, S. Ogura, A. Ikeda, H. Kataoka, M. Tanaka and T. Joh, J. Photochem. Photobiol., C, 2011, 12, 46.
27. K. W. Oh, H.-M. Choi, J. M. Kim, J. H. Park and I. S. Park, Text. Res. J., 2014, 84, 808.
28. M. N. Usacheva, M. C. Teichert and M. A. Biel, Lasers Surg. Med., 2001, 29, 165.
29. G. E. Cohn and H. Y. Tseng, Photochem. Photobiol., 1977, 26, 465.
30. F. Freire, A. C. Costa, C. A. Pereira, M. Beltrame Junior, J. C. Junqueira and A. O. Jorge, Laser. Med. Sci., 2014, 29, 949.
31. M. J. Bovis, S. Noimark, J. H. Woodhams, C. W. M. Kay, J. Weiner, W. J. Peveler, A. Correia, M. Wilson, E. Allan, I. P. Parkin and A. J. MacRobert, RSC Adv., 2015, 5, 54830.
32. P. V. Araujo, K. I. Teixeira, L. D. Lanza, M. E. Cortes and L. T. Poletto, Acta odontologica latinoamericana, 2009, 22, 93.
33. Y. Guo, S. Rogelj and P. Zhang, Nanotechnology, 2010, 21, 065102.
34. M. G. L. Hertog and P. C. H. Hollman, Eur. J. Clin. Nutr., 1996, 50, 63.
35. M. E. Inal, A. Kahraman and T. Koken, Clin. Exp. Dermatol., 2001, 26, 536.
36. P. Sestili, A. Guidarelli, M. Dachà and O. Cantoni, Free Radical Biol. Med., 1998, 25, 196.
37. J. P. Brown, Mutat. Res., 1980, 75, 243.
38. M. A. Read, Am. J. Pathol., 1995, 2, 235.
39. E. Ohnishi and H. Bannai, Antiviral Res., 1993, 22, 327.
40. A. Russo, R. Acquaviva, A. Campisi, V. Sorrenti, C. Di Giacomo, G. Virgata, M. L. Barcellona and A. Vanella, Cell Biol. Toxicol., 2000, 16, 91.
41. R. Casagrande, S. R. Georgetti, W. A. Verri Jr, D. J. Dorta, A. C. dos Santos and M. J. V. Fonseca, J. Photochem. Photobiol., B, 2006, 84, 21.
42. B. Choquenet, C. Couteau, E. Paparis and L. J. M. Coiffard, J. Nat. Prod., 2008, 71, 1117.
43. A. P. Demchenko, K.-C. Tang and P.-T. Chou, Chem. Soc. Rev., 2013, 42, 1379.
44. P. K. Sengupta and M. Kasha, Chem. Phys. Lett., 1979, 68, 382.
45. M. Jin, J. Xie, J.-P. Malval, A. Spangenberg, O. Soppera, D.-L. Versace, T. Leclerc, H. Pan, D. Wan, H. Pu, P. Baldeck, O. Poizat and S. Knopf, J. Mater. Chem. C, 2014, 2, 7201.
46. M. Condat, P.-E. Mazeran, J.-P. Malval, J. Lalevée, F. Morlet-Savary, E. Renard, V. Langlois, S. Abbad Andalloussi and D.-L. Versace, RSC Adv., 2015, 5, 85214.
47. J.-P. Fouassier and J. Lalevée, Photoinitiators for Polymer Synthesis: Scope, Reactivity and Efficiency, Wiley-VCH Verlag GmbH and Co KGaA, Weinheim, 2012.
48. D. Rehm and A. Weller, Isr. J. Chem., 1970, 8, 259.
49. Q. Kan, W. Yan, G. Kang and Q. Sun, J. Mech. Phys. Solids, 2013, 61, 2015.
50. T. Maisch, J. Baier, B. Franz, M. Maier, M. Landthaler, R.-M. Szeimies and W. Bäumler, Proc. Natl. Acad. Sci. U. S. A., 2007, 104, 7223.
51. K. E. Sanderson, T. MacAlister, J. W. Costerton and K. J. Cheng, Can. J. Microbiol., 1974, 20, 1135.
52. T. A. Dahl, W. R. Midden and P. E. Hartman, Photochem. Photobiol., 1988, 47, 357.
53. T. A. Dahl, W. R. Midden and P. E. Hartman, Photochem. Photobiol., 1987, 46, 345.
54. R. C. Straight and J. D. Spikes, in Singlet O₂, ed. A. A. Frimer, CRC Press, Inc., Boca Raton, Fla, 1985, vol. IV.
55. M. Hakala-Yatkin and E. Tyystjärvi, Biochim. Biophys. Acta, Bioenerg., 2011, 1807, 243.
56. D. Bellus, H. Lind and J. F. Wyatt, J. Chem. Soc., Chem. Commun., 1972, 1199.
57. Z. Markovic, B. Todorovic-Markovic, D. Kleut, N. Nikolic, S. Vranjes-Djuric, M. Misirkic, L. Vucicevic, K. Janjetovic, A. Isakovic, L. Harhaji, B. Babic-Stojic, M. Dramicanin and V. Trajkovic, Biomaterials, 2007, 28, 5437.
58. C. Ringot, V. Sol, R. Granet and P. Krausz, Mater. Lett., 2009, 63, 1889.
59. D. J. Balazs, K. Triandafillu, P. Wood, Y. Chevolot, C. v. Delden, H. Harms, C. Hollensteind and H. J. Mathieu, Biomaterials, 2004, 25, 2139.
60. P. J. Eginton, H. Gibson, J. Holah, P. S. Handley and P. Gilbert, J. Ind. Microbiol., 1995, 15, 305.
61. J.-P. Malval, C. Chaimbault, B. Fischer, J.-P. Morand and R. Lapouyade, Res. Chem. Intermed., 2001, 27, 21.

---

Footnote

† Electronic supplementary information (ESI) available: Emission spectrum of the lamp used during the different incubation times (Fig. S1), evolution of the UV-visible spectra of quercetin alone and quercetin/Iod in ACN during irradiation (Fig. S2), cyclic voltammetry of quercetin in ACN (Fig. S3), evolution of the –OH bond from the polyether network during the 1200 s of irradiation (Fig. S4), influence of the incubation time (0, 2 h, 6 h), without illumination, on the growth of (A) E. coli and (B) S. aureus with and without DMSO (Fig. S5) and evolution of the adhesion/proliferation of E. coli and S. aureus on the stainless steel substrate (Fig. S6). See DOI: 10.1039/c5ra25267a

---