Photosensitive cellulosic materials based on a covalently grafted phenosafranin-modified silsesquioxane analog for bactericidal applications

Light-triggered antimicrobial cellulose surfaces were obtained by the immobilization of a photosensitive phenosafranin dye (PSF) in a hybrid organic–inorganic silsesquioxane polymer applied on handsheets prepared from a standard bleached softwood pulp. These coatings were deposited by polycondensation of methyltriethoxysilane and an alkoxysilyl derivative of phenosafranin (TESPSF) obtained by the thiolene addition reaction and coupling of the succinic anhydride derivative with a primary amine group of PSF. TESPSF and coatings were characterized by advanced techniques in terms of chemical structure (1H, 13C, 29Si NMR, MS, ATR-IR), surface properties (SEM, EDX, water contact angles), and optical properties (UV, reflection light intensity, ISO brightness). The light-induced antimicrobial activity of sheets of paper coated with new materials showed the inhibition of the growth of the bacteria Staphylococcus aureus ATCC 6538 and Escherichia coli ATCC 8739. The cytotoxicity studies of modified cellulose surfaces were performed using erythrocyte lysis assays under both dark and light conditions exhibited no toxicity on erythrocytes. Thus, the new material did not reveal harmful effects on erythrocytes, regardless of the presence and absence of light.


Introduction
The prevalence of various infections especially in the context of global crisis caused by detrimental microorganisms contributes to the destruction of public health and it is the source of serious problems, particularly since the transmission of COVID-19.The Vol:. ( 1234567890) impact on public well-being is also connected with the spread of multidrug-resistant bacteria (Brown and Wright 2016;Pascuta and Vodnar 2022).Every day, materials employed in healthcare systems can accumulate infectious agents such as the coronavirus, vancomycin-resistant enterococci, and methicillin-resistant Staphylococcus aureus (MRSA).It is noteworthy, that certain microbes exhibit remarkable resilience with the capability to persist for over 90 days (Neely and Maley 2000).Thus, the shortcomings of antimicrobial medicines can be overcome by the integration of bactericidal materials, photodynamic inactivation and personal protective equipment (PPE) against infections to control the spread of diseases (Duan et al. 2022;Mahira et al. 2019;Nawab et al. 2022).Cellulosic materials with anti-pathogen coatings are still needed for food and medical packaging as well as frequently handled items like banknotes to avoid bacterial contamination (Amirabad et al. 2018;Jung et al. 2018).The conventional PPE can passively adsorb or physically block pathogens from entering the body but without the ability of their neutralization and biological or chemical deactivation (Kumar et al. 2021).Consequently, the presence of living bacteria or viruses could potentially lead to cross-contamination, that is a great challeng in terms of PPE recycling, disinfection, and disposal (Jabłońska-Trypuć et al. 2022).Nowadays, the bacterial photodynamic inactivation (PDI) system involves photosensitizers to generate reactive oxygen species (ROS) under light irradiation.These agents exhibit capability to eliminate pathogens with a high degree of efficiency (Tonon et al. 2021).ROS, including singlet oxygen ( 1 O 2 ), play a pivotal role in inducing oxidative stress, which in term leads to damage in both Gram-positive and Gram-negative bacteria (Youf et al. 2021;Raza et al. 2022).Thus, our research group is actively engaged in the development of antimicrobial fabrics.
To achieve this, we used a photochemical dyes i.e., phenosafranin chloride (PSF) that was immobilized on silicon supports (Rozga-Wijas et al. 2019).Various solid supports have been explored for immobilizing photosensitizers.These supports encompass silicone polymers, silica nanoparticles, magnetic nanoparticles, and porous silicon.The aforementioned materials serve as platforms for different photosensitizers such as rose bengal (Gurianov et al. 2022), porphyrins (Rychtarikova et al. 2012;Secret et al. 2013), phthalocyanines (Baigorria et al. 2018;Mapukata et al. 2021), methylene blue (Fernandez-Lodeiro et al. 2021;Piccirillo et al. 2009) to enhance the efficancy and application of photosensitizers in antimicrobial therapies (Spagnul et al. 2015).Besides, the surface of highly biocompatible silica particles can be easily modified through the grafting of photosensitizers, integrated with metallic nanoparticles, and gold nanorods to further augment the PDI efficacy of phthalocyanine dihydroxide, and methylene blue (Grüner et al. 2018).Additionally, nanocomposites with silica and silica-containing core-shell particles or less common silica nanofibers exhibited the potential to induce the PDI effect (Ma et al. 2022).Recently, crystal violet was immobilized onto commercial face mask filters as daylight-induced antibacterial and antiviral materials (Jeong et al. 2022).
In our research, we evaluate the effectiveness of PDI systems on the surface of cellulose fibers modified with a PSF photosensitizer entrapped or grafted to the siloxane-silsesquioxane network.PSF, being a phenazine compound, generates ROS that induce bacterial cell death (Bolton 1991;Imato et al. 2020).It is a promising photo-drug for photodynamic therapy (da Silva Junior et al. 2023).PSF exhibits a notable quantum yield for generating singlet oxygen ( 1 O 2 ) and demonstrates a favourable triplet quantum yield (Φ T = 0.2-0.42)as elucidate by (Broglia et al. 2006).Recently, phenazines were used as biological probes (Sen et al. 2019), and the biotin-PSF conjugate was reported as a photosensitive compound for targeted therapy and imaging (Blauz et al. 2021).Furthermore, a cellulose polymer offers a suitable platform for the development of new materials that can be readily coated with bactericidal agents improving modern technologies to support the limitation of the spread of detrimental microbes (Koshani et al. 2021;Seidi et al. 2022).Wang investigated electrospun diacetate microfibers that contained protoporphyrin IX (Fadavi et al. 2019;Wang et al. 2021).Chen made a substantial contribution by introducing an antimicrobial film based on cellulose 2,3-dialdehyde with an incorporated β-cyclodextrin/curcumin complex as a photosensitizer (Chen et al. 2021).
Recently, we reported the use of PSF as a potential photoactive pharmaceutical compound against clinical strains of S. aureus MRSA and E. coli (Rozga-Wijas et al. 2021).Our findings suggested that cationic PSF-polyhedral octasilsesquioxane (PSF-POSS) conjugates used in low concentrations could serve as promising photosensitizers against bacteria after green light irradiation for 5 min.For the purpose of augmenting the PSF-POSS applications in antibacterial personal protective systems, a hybrid material comprising PSF and siloxane-silsesquioxane polymer was utilized to cover a cellulose sheet.The silsesquioxane network acted as an amphiphilic nanocarrier of the cationic PSF photosensitizer and was selected to be a biocompatible composite for the purpose of changing the surface of cellulose sheets to give them particular bactericidal properties upon ultraviolet-visible light irradiation.We investigated two sol-gel networks based on methyltriethoxysilane (MTES) and PSF as well as the alkoxysilyl derivative of PSF.The molecule of a new photosensitizer (TESPSF) serves as a precursor for the sol-gel process and combines photodynamic properties with the reactivity of alkoxysilyl groups.TESPSF was designed for the development of innovative biocompatible materials capable of generating ROS species under light irradiation.The primary photochemical/photophysical mechanisms are initiated by the activation of the photosensitizer (PSF) through light absorption, leading to its transition into an excited state ( 3 PSF*), able to interact with molecular oxygen ( 3 O 2 ) and forming singlet oxygen ( 1 O 2 , type II reaction) through every transfer process (Miksa 2022).Microbiological studies were evaluated against reference strains of bacteria, S. aureus ATCC 6538 and E. coli ATCC 8739, by adaptation of a method with green light irradiation (λ Em.max = 522 nm, 7.2-7.5 mW/cm 2 ) for 15 and 30 min.All experiments were performed in a starch solution to utilize the cellulose surface in accordance with the procedure described for the MTES in water emulsion (Ganicz and Rozga-Wijas 2021).

Preparation of the antibacterial coating mixtures
The method of applying alkoxysilanes on handmade paper sheets was reported in our previous work (Ganicz and Rozga-Wijas 2021).Briefly, 45 g of a 10% polyvinyl alcohol solution (MOWIOL 8-88) was mixed with 45 mL of distilled water and homogenized at room temperature for 5 min at 23,000 rpm.Then, solutions containing 150 mL (134.25 g, 0.7529 mol) of MTES and 60 g of a 2% Sulforocanol L-327 were added to produce the emulsion with an alkoxysilane concentration of 47.37%.In a separate vessel, 10 g of wheat starch and 0.5 g of NaOH were dissolved in 250 mL of demineralized water and heated at 60 °C for 1 h.After cooling to room temperature, the resulting starch solution was homogenized with the alkoxysilane emulsion in a 1:1 (v/v) ratio and then the parent emulsion was divided into several 50 mL samples.E1 samples were prepared by adding PSF solutions via a microsyringe to the parent samples (Table 1s in data repository) at dye concentration from 0.01 to 10 mmol/L (as listed in Table 1) and were used immediately after their preparation.The E2 samples were prepared by adding 1, 5, 15, 25, 50 and 350 µL of a TESPSF solution (0.054 g, 8.14 10 -5 mol in 1 mL THF) separately to the parent emulsion, mixing for 15 min and applying to a blank sheet of paper.The volumes of solutions and concentrations of the photosensitizer in the E2 sol in the range of 0.016-5.6mmol/L are given in (Table 2 s in data repository).

Coating mixtures application
Coating mixtures were applied to the surfaces of the paper samples using an automatic coater (Control Coater, TUL, Lodz, Poland) and a standard Mayer rod No. 20 (K-bar) at a speed of 4 cm/s, yielding a wet film thickness of 50 µm.The coated papers were dried using a laboratory rotary drum dryer (Type 89, Mechanika Praha, Czech Republic) at a temperature of 100 °C ± 2 °C for 3.5 min.The handsheets were conditioned as indicated by the (ISO 187: 1990) standard procedure (24 h at 25 °C and 50% of air moisture).The arithmetic average weights of the base handsheets and coated handsheets are listed in (Tables 3 s and 4 s in a data repository).

Analytical methods
The 1 H, 13 C and 29 Si NMR spectra were run on Bruker AVANCE III 500 MHz spectrometer equipped with 5 mm inverse broadband dual channel probe head with Z-gradients, operating at 500.13, and 125.77, and 99.36 MHz for 1 H, 13 C and 29 Si respectively.All measurements were made at 295 K and the temperature was stabilized with a Bruker BCU 05 cooling system controlled by a VTU 3200 unit.Spectra were acquired and processed with TopSpin 3.6 Bruker software.Solid state 29 Si and 13 C-NMR spectra were run with a DSX 500 Bruker spectrometer.Spectra were acquired with cross-polarization, at 79.506 and 100.627MHz respectively, applying 90-µs pulses, 6-s pulse delay, and 3-ms contact time, with samples in 4.0-mm zirconia rotors spinning at 8 kHz.For all 1 H, 13 C and 29 Si NMR spectra, the chemical shift values were referenced externally to TMS (δ = 0.00 ppm).
High-resolution mass spectrometry (HRMS) analyses were run using a Synapt G2-Si mass spectrometer (Waters) equipped with a quadrupole-time-offlight mass analyzer.Methanol was used as a solvent.The measurements were performed in a positive ion mode with the desolvation gas flow at 500 L/h and capillary voltage set to 2500 V with a flow rate of 100 L/min.The measurement results were processed using the MassLynkx: 4.1 software (Waters) incorporated with the instrument.
Matrix-assisted laser desorption ionization time of flight (MALDI-TOF) measurements were performed using a Voyager Elite instrument (PerSeptive Biosystems, Framingham, MA, USA) instrument equipped with a pulsed N2 laser operating at 337 nm.Mass spectra were obtained in a linear mode with an accelerating voltage of 20 kV.Samples were prepared from THF solutions with 1,8-dihydroxy-10H-antracen-9-one (dithranol) as a matrix, and sodium trifluoroacetate (NaTFA) as a cationizing agent.
Fourier transforms infrared (FTIR) measurements were performed on Thermo Scientific Nicolet 6700 FT-IR instrument with Golden Gate ATR, (Attenuated Total Reflectance) Axion accessory with detector DTGS using a resolution of 2 cm −1 at a rate of 4 scans per second.The baseline was manually adjusted for all the spectra.The scanning electron spectroscopy (SEM) images were obtained using Jeol 6010 (Tokyo, Japan) connected to a secondary electron detector and X-ray energy dispersive spectrometer (EDS).The samples were placed on a double-sided conductive tape on an aluminum substrate and covered with a thin layer of 10 nm gold, using a Baltec SCD 050 Sputter Coater apparatus.Several images were obtained from various parts of the sample, to assure the reproducibility of the final image taken as representative of the entire handsheet, as well as the silica EDS elemental analysis.
The water contact angles were measured using a PGX goniometer (Testing Machines, Inc., New Castle, DE, USA) with the (TAPPI T 458:2004 2004) standard static method.The values shown in the tables and graphs are the arithmetic means of 3 measurements of randomly selected samples of each paper type.Studies of water absorbency were performed using Cobb test (ISO 535:2014(ISO 535: 2014)), conducted for 120 s.VIS reflection spectra and ISO brightness of the coated and uncoated samples were recorded using X-Rite SpectroEye spectrometer in accordance with the (ISO 2470-1:2016 2016) standard for samples conditioned as indicated by the (ISO 187:1990(ISO 187: 1990) ) procedure.
The evaluation of the antibacterial activity using the absorption method The studies were performed according to the procedure described in the BSI Standards Publication (BSI, BS EN ISO 20743:2021 2021) (BS EN ISO 20743:2021 2021) with some modifications.Two bacterial strains Staphylococcus aureus ATCC 6538 and Escherichia coli ATCC 8739 were used to estimate their biological activity using the absorption method.The bacterial glycerol stock solution was plated on the nutrient agar (NA) plate and incubated at 37 ℃ for 24 h.The plates containing bacteria were stored at 5-10 °C and used within one week from the date of preparation.Bacterial colonies were then transferred to 20 mL of NB in a 100 mL Erlenmeyer flask and incubated at 37 ℃ for 24 h with shaking.Then, 0.4 mL of the overnight inoculum (1 × 10 8 CFU/mL to 3 × 10 8 CFU/mL) was added to 20 mL of NB in a 100 mL Erlenmeyer flask and further incubated at 37 ℃ for 3 h ± 1 h with shaking.The bacterial cultures were diluted with the same broth to obtain an optical density corresponding to 1.0 × 10 5 CFU/mL to 3.0 × 10 5 CFU/mL using McFarland's nephelometer.Subsequently, 60 µL of the prepared bacterial suspension was applied to the discs coated with the appropriate compounds with a mass of 0.12 g ± 0.015 g and to the control discs (untreated).The discs were then incubated at 37 ℃ and irradiated with green light (λ Em.max = 522 nm, 7.2-7.5 mW/cm 2 ) for 15 and 30 min.After the illumination, the discs were placed in the neutralizing solution (at approximately pH about 1, 6, or 10) and shaken 5 times for 5 s.A series of dilutions were then made in NB medium and placed on NA plates.Incubation was carried out for 48 h at the appropriate temperature and the microorganisms grown on the plates were counted.Experiments in the dark were carried out simultaneously at the same time (15 and 30 min) and temperature.

Erythrocyte lysis assay
Erythrocytes were isolated from 5.0 mL fresh sheep blood by centrifugation for 10 min at 1000 × g and washed three times with 0.9% NaCl.Then, the cell suspension with a final concentration of 10 8 CFU/ mL in a volume of 400 µL was added to each well of the 12-well microtiter plate with discs covered with appropriate compounds and control discs (untreated) with a mass of 0.12 g ± 0.015 g.To estimate the relative hemolytic potential of tested probes, the appropriate controls, i.e., 0% lysis in PBS and 100% erythrocyte lysis in 4% Triton X-100 were used.Plates with samples were incubated for 1 h at 37 °C then centrifuged for 10 min at 1000 × g to separate the unlysed erythrocytes.Finally, the absorbance (A) was measured spectrophotometrically at 450 nm.The hemolysis percentage was calculated according to the equation presented by Sharma and Khan (2016).The % hemolysis = [(A450 of test compound treated sample-A450 of PBS buffer treated sample)/(A450 of 4% Triton X-100 treated samples-A450 of PBS buffer treated sample)] × 100.The tests were carried out in the dark and after 15 and 30 min of irradiation.

Statistical analysis
All antibacterial activity experiments were performed in triplicates, in three independent experimental sets.The graphics and the data were constructed and analyzed statistically by SD-VIS reflection spectra and ISO brightness measurements that were performed on 3 randomly selected points on 3 randomly selected samples of a given type.The concentration of coatings and the final data were reported as an arithmetic average of all 9 measurements.

Characterization of a new photosensitive precursor (TESPSF) of the sol-gel process
A new photosensitive sol-gel precursor, TESPSF had good dispersion properties in a silsesquioxane network due to hydrolyzable silyl groups attached to a thioether chain and a hydrophilic PSF molecule.TESSA was obtained by thiolene addition of MPTES to allylsuccinic anhydride in toluene in the presence of AIBN as a radical initiator.TESPSF was synthesized by a reaction of primary amino groups of PSF with the anhydride function of TESSA in DMF.The high signal in the MALDI spectrum described the alkoxysilyl-substituted PSF photosensitizer that forms a sodium ion [TESPSF + Na] + at m/z 685.9.The lower signal was assigned to the protonated ion [TESPSF + H] + at m/z 663.8 (calc.mass 663.26726 for C 34 H 43 N 4 O 6 SiS), (Fig. 2a).The isotopic distribution is identical to the theoretical distribution as shown in (Fig. 2b and c).Dimers that formed as a result of condensation with MPTS due to the high reactivity of ethoxysilyl groups were observed at higher m/z (Rozga-Wijas et al. 2010).
The identity of TESSA and TESPSF compounds was confirmed by 1 H, 13 C and 29 Si NMR spectra, and the assignment of shifts was made on the basis of 2D spectra for 1 H-1 H and 1 H-13 C nuclei (listed in the experimental section).Selected spectra are shown in (Figs. 3 and 4).The structure of the TESSA molecule is confirmed by two overlapping triples from 2.47 to 2.55 ppm attributed to the thioether bridge CH 2 -S-CH 2 , two multiplets at 2.66 and a 3.09 ppm attributed to the nonequivalent proton CH 2 number 8 assigned to the succinyl anhydride ring, and a singlet at 3.11 ppm ascribed to the CH methine proton, which is overlapped with proton 8a in the 1 H NMR spectrum but easily found in the COSY and HSQC spectra.
Finally, the signals proving the presence of alkoxysilyl groups needed for further reactions are found as a triplet and quartet at 1.21 and 3.81 ppm in the 1 H spectrum, and at 18.3 and 58.4 ppm respectively in the 13 C spectrum.The Fig. 4 illustrates the 1 H-13 C HSQC NMR spectra in the range up to 4.2 ppm before and after binding with PSF.A significant difference is between CH 2 protons No. 6 and 8 and CH proton No. 7, shifted from 3.11 to 2.78 ppm at the same carbon position assigned to 40.5 ppm.The ring-opening process removes the non-equivalence of protons number 6 from the five-membered anhydride structure.In the TESSA spectrum, we observed them at 1.75 and 2.01 ppm, and after the reaction was completed, they approached to 1.73 ppm.The resonance signals from 5.80 to 8.3 ppm correspond to the aromatic protons and the NH amide proton (8.25 ppm) in TESPSF.

The fabrication of light-induced antibacterial cellulose handsheets
The cellulosic handsheets were prepared from commercial bleached soft-wood pulp (BSK) using a laboratory paper machine according to the industrial standard ISO 5269-1 (2005).No additional pulp milling, no other mechanical or chemical fiber modification methods were performed, therefore resulting sheets were very porous and relatively mechanically resistant.Silica-starch emulsions based on MTES were prepared from the PSF hydrochloride solution E1 and the TESPSF solution E2.The photosensitive dye was dissolved in the sol and formed a homogeneous emulsion that uniformly coated the surface of the cellulose sheets according to the procedure described earlier (Ganicz and Rozga-Wijas 2021).The process of manufacturing antibacterial cellulose materials is illustrated in Scheme 2.
The surface concentration of PSF on modified papers was calculated based on the concentration of PSF in the starting siloxane sol and the weight of the paper cover listed in Table 1 and Table 5s in data repository.Besides, the basic building block of the silsesquioxane network is composed of a repeating unit of MeSi(O-)OH with a mass of 76.13.
The hydrolysis of a trialkoxysilane leads to silanetriol (Pietschnig and Spirk 2016), which can be stabilized by bifunctional secondary and tertiary amines (e.g., bipyridine or piperazine) to maximize the possible interaction sites (Prabusankar et al. 2004).Therefore, we expected that PSF endowed with two primary amine groups will form stable adducts with the silanol (SiOH), siloxide (SiO-), and siloxane (Si-O-Si) surface groups (Scheme 3).The solid-state 29 Si and 13 C NMR spectra were taken to illustrate the chemical structure of the covering on the cellulose fibers.For this purpose, the siloxane-starch emulsion E2.4 loaded with TESPSF was maintained at room temperature for 30 days and then centrifuged.The precipitate was washed several times with methanol, dried, and heated to 110 °C under reduced pressure.The 29 Si NMR spectrum of the light violet powder shows a change in the silicon shift from − 45.47 ppm characteristic of TESPSF to − 56.17 and − 66.01 ppm for the coating (Fig. 5a  including e.g., linear, cyclic, ladder, cage, and incomplete cage. The reflection FTIR analysis was performed to compare the handsheets before and after coating the cellulose fibers with the siloxane sol (Fig. 5c).The ATR-FTIR spectrum of native paper displays the classical bands with maxima wavenumbers at 3340 cm −1 (OH stretching), and signals at following frequencies 1325, 1015, 1094 and 1163 cm −1 are assigned to sv(C-O), v(C-H) bending, and sv(C-O and C-C) mode, respectively (Kizil et al. 2002;Meng et al. 2015).The methyl silsesquioxane-modified paper, both with and without PSF, indicated two new bands at 1270 and 770 cm −1 assigned to sv(Si-CH 3 ) and v(Si-CH 3 ) bending, confirming the formation of the silsesquioxane polymer by the polycondensation reaction (Siuda et al. 2019).Moreover, signals in the range of (1127-1004 cm −1 ) were attributed to sv s (Si-O-Si) and sv as (Si-O-C) modes in polymers, but they overlapped with the bands of cellulose vibrations.Absorption bands characteristic of the PSF dye are obscured in the FTIR spectrum of modified handsheets due to the low concentration of PSF.The molar ratio of PSF to the alkoxysilyl group (PSF/CH 3 SiO) ranges from 8.0 × 10 -3 to 8.0 × 10 -6 .Thus, the concentration of the methylsilyl group was three orders of magnitude higher than PSF, and the surface concentration of dye molecules increased from 9.0 × 10 -7 to 1.0 × 10 -3 mol per m 2 for E1 and from 6.0 × 10 -7 to 3.0 × 10 -4 mol per m 2 for E2 (Table 1).
X-ray energy dispersive spectrometer (EDS) analyses provided images of the elemental dispersion of carbon, oxygen, and silicon on cellulose coated with PSF embedded in a siloxane material.The measurement of selected samples (Fig. 6) shows that silicon was well dispersed on the modified surface E1 and E2 of the cellulose sheets.The EDS elemental analysis allowed us to estimate the amount of silicon and carbon in the silsesquioxane layer.However, it is not possible to determine the amount of photosensitizer anchored into the inorganic structure because nitrogen is not detectable.A decrease in the amount of carbon was observed for the E1.4 and E2.5 materials compared to the pure cellulose sheet due to covering of the cellulose surface with methylsilsesquioxane groups CH 3 Si(O-) 3 (see Table 1).

Surface properties
The surface of the cellulose fibers was coated with the silsesquioxane gel and starch, then dried using a rotary press machine and applying elevated temperature in the range of 95-100 °C to obtain the uniform coating.The coating was not a continuous layer but formed a cluster of spherical nanoparticles on the surface of the fibers after using aqueous emulsions of MTES (Ganicz and Rozga-Wijas 2021).The surfaces of E1/S and E1.1-E1.5 papers modified with PSF were examined by SEM (Fig. 6b and Fig. 3s in data repository).The photomicrographs did not The presence of PSF and its grafting in the silsesquioxane network was confirmed (apart from the visible color) by the measurement of relative intensities of reflected light, ISO brightness, and microbiological tests under light irradiation.The first experiment shows the dependence of the intensity of the reflected light on the type of covering of the paper sheet.Dye-loaded cellulose sheets coated with E1 and E2 emulsions show the maximum absorption at wavelengths of 527 and 537 nm, respectively (Fig. 7a and b) which correspond to the absorption of native PSF in methanol (529 nm) (Blauz et.al 2021).Studies of polymer-coated sheets with a surface concentration of PSF at 0.1 and 1.17 mmol/m 2 for E1 and 0.01 and 0.32 mmol/m 2 for E2 samples indicate a strong convergence of pre-and post-Cobb measurements.
Comparisons of the reflected light curves allowed us to assess the high stability of the immobilized PSF dye in the silsesquioxane coating.The polycondensation of alkoxysilanes can occur and causes internal voids due to the formation of branched and cyclic structures and incomplete condensation, leaving free SiOH groups to interact with the dye (Sato et al. 2022;Catauro and Ciprioti 2021).We indicated in the 29 Si NMR spectrum (Fig. 5b), that approximately 20% of the Si atoms remain substituted on a free hydroxyl group corresponding to T 2 units at − 56.17 ppm.It is likely that PSF is readily located in these spaces because of electrostatic interactions or hydrogen bonding, and then becomes blocked there during the drying process.It should be noted that the sol into which the photosensitizer is introduced is a polymer (polysiloxane-polyols) which has caused multi-point electrostatic interactions due to the repeating units of (-O)Si(OH) (Nakai et al. 2017).The second parameter, ISO brightness, decreased with increasing PSF  7) clearly proved that the main factor changing the ISO brightness of the samples had a strong absorption of PSF/TESPSF at a maximum wavelength of 520-540 nm that increased with the surface concentration of both compounds.
In our previous studies we proved that the ethoxysilanes solution was a very effective hydrophobization agent leading to the materials with strong surface hydrophobicity that improved barrier properties as measured by Cobb test (up to 250% better than uncoated paper) (Ganicz and Rozga-Wijas 2021).Unfortunately, the addition of PSF and TESPSF to the mixtures strongly decreases the effect of hydrophobicity on the surface.For example, samples with a small amount of PSF indicated the water contact angles over 90° due to the low concentration of a dye on the surface.For higher surface concentrations of PSF than 0.009 mmol/m 2 the measured contact angles were below 90° and decreased with an increasing its concentration.The addition of a very low concentration of TESPSF caused a slight decrease in contact angles below 90°, but they were still in the range of 90-82° at higher concentrations of TESPSF, and water drops were swallowed after 2-3 min (Table 1).
Photoinactivation of gram-positive and gram-negative bacteria and erythrocyte lysis assay of the modified cellulose surfaces The antibacterial properties of the modified paper were examined against both Gram-negative (E.coli) and Gram-positive (S. aureus) bacteria with exposure samples to green light for 15 and 30 min.Firstly, it appears necessary to show the differences between paper sheets left in a bacteria environment in the darkness and after irradiation to demonstrate no inhibition of bacterial growth in the dark conditions, as shown in (Fig. 8a and b).As an example, the number of bacteria was compared for E1.4 sheets with a PSF density of 0.1 mmol/m 2 and E2.4 sheets with a density of 0.04 mmol/m 2 .In the darkness, living bacteria remain on the surface, unchanged.After 15 min of green light irradiation, all S. aureus on E1.4 and E2.4 were killed, while the number of living E. coli Inhibition of the growth of S. aureus and E. coli bacteria was enhanced by increasing the photosensitizer concentration on the surfaces of E1 and E2 sheets.The handsheets E2 despite a lower concentration of PSF on the surface showed better antibacterial effectiveness compared to E1 (Fig. 9a  and b).Thus, the complete killing of S. aureus was observed after 15 min irradiation of E2.4 at a concentration of 0.04 mmol/m 2 , and the number of bacteria decreased by two orders at a concentration of 0.01 mmol/m 2 .However, the E1.4 sample reveals the complete killing of S. aureus after 15 min irradiation of the surface at a concentration of 0.1 mmol/m 2 .The E. coli bacteria were less sensitive to the ROS effect generated by PSF from the silsesquioxane network included in the cellulose surface, and test results for the E2 sample were illustrated in (Fig. 9c).For both E1 and E2 surfaces, 100% inhibition was achieved after extending the irradiation time to 30 min.
The photodynamic deactivation of bacteria was observed in E2 samples with covalently bound photosensitizer molecules, as well as in E1 including the silsesquioxane polymer with a dye-complex.In both cases, PSF molecules were well dispersed and strongly fixed with the inorganic silica matrix, generating singlet oxygen under irradiation using green light.The effectiveness of bacteria-killing was proportional to the concentration of PSF, so the singlet oxygen was the cell-deactivating factor (Rozga-Wijas et al. 2021).The greater efficiency of the E2 surface with shorter irradiation time and a lower PSF concentration resulted from the greater homogeneity of the cellulose coverage.
Moreover, the erythrocyte lysis assay was performed to evaluate the toxicity of modified sheets of paper on red blood cells and is described in Section "Erythrocyte lysis assay".Hemolysis was observed only in the case of disc E1.5 (the highest PSF concentration of 1.17 mmol/m 2 in the E1 sheet series) before and after irradiation using green light for 15 and 30 min, and it amounted to 8.78%, 9.05% and 13.25%, respectively.Studies with the rest of the tested modified sheets of paper showed no hemolysis in the dark conditions, after 15 and 30 min of irradiation (Fig. 10).

Conclusions
For the first time, new photosensitive cellulose materials were obtained as a result of their coverage by a silsesquioxane network containing hybrid PSF photosensitizers using the sol-gel method.These materials were very active in inhibiting the growth of S. aureus ATCC for samples of paper sheets with PSF Fig. 10 The effect of modified sheets of papers on erythrocytes.The results are presented as a percentage of hemolysis of red blood cells.Lack of bars in tested papers means no hemolysis.Triton X-100 was taken at 100%, and saline at 0% hemolysis.Data are presented as mean ± SD Vol.: (0123456789) molecules that were covalently bound or complex with the silsesquioxane network at concentrations of 0.04 and 0.1 mmol/m 2 , respectively, after green light irradiation (λ Em.max = 522 nm).Bactericidal studies against S. aureus and E.coli strains on the surface of modified sheets showed inhibition of the growth bacteria after 15 or 30 min of irradiation with green light.It is crucial, that the negative effect of photosensitizer-modified paper handsheets on red blood cells did not occur.Therefore, reported materials can be used in the design of self-disinfecting PPE and may also be tested in a medical dressing for drug carriers and wound healing persistently infected with clinical strains of S. aureus MRSA.Moreover, commercially available organofunctional trialkoxysilanes are inexpensive, and a TESPSF multifunctional sol-gel precursor easily forms homogeneous aqueous emulsions with polyols due to supramolecular interactions involving hydrogen and electrostatic bonds.

Fig. 5 a
Fig. 5 a The 29 Si NMR spectrum of the TESPSF in CD 3 OD; b 29 Si MAS NMR spectrum of the silica-starch E2.4 gel; c The FTIR spectrum of the handsheets of E1.5 coated and uncoated papers

Fig. 6
Fig. 6 The X-ray energy dispersive spectrometer (EDS) analyses of PSF-silica cellulose materials images display the elemental dispersion of C, O, and Si (yellow zone) at the micro-

Table 1
The concentration of a photosensitizer immobilized in the silsesquioxane network on the cellulose surfaces a Natural sheet b Sol coated sheet without PSF c Starch coated sheet with PSF d Not measured because of the very fast swallowing of water drops Vol:.(1234567890)