Biosynthesis and Functionalization of Bacterial Cellulose Membranes with Cerium Nitrate and Silver Nanoparticles

This work aimed to synthesize bacterial cellulose (BC) membranes functionalized with Ce(NO3)3 and silver nanoparticles (AgNPs) heat-treated in oven and autoclave. The AgNPs were characterized by TEM and the membranes were characterized by their rehydration capacity, Thermogravimetric Analysis (TGA), Scanning Electron Microscopy with Dispersive Energy Spectroscopy (SEM/EDS), Wide-angle X-ray Diffractometry (XRD), Atomic Absorption Spectroscopy (AAS), antimicrobial activity and cytotoxicity. The BC membranes were white to light yellow or brown indicating functionalization with AgNps with 20 nm. XRD detected a decrease in the degree of crystallinity in functionalized membranes. The autoclaved samples exhibited the largest porosity and good rehydration capacity. AAS showed Ag presence in all samples. The best antibacterial activity was achieved for the membrane’s oven treated with 40 ppm of AgNPs. Among the functionalized membranes, the most recommended to be used as burn dressing is the Ag40A and Ag20CN1.1A since they showed good porosity, good antimicrobial activity, and better suitable cell viability.


Introduction
Accidents with burns have become a global public health problem. When the burn is not fatal, it causes deconfiguration, prolonged hospitalization and/or disability of the patient 1 . Hospital infections are among the most important side effects and have the potential to cause serious complications. Burn wounds provide an ideal medium for bacterial proliferation and an entry portal into the bloodstream. The loss of the epithelial barrier, hypermetabolic/hypercatabolic states, and immunosuppression predispose burned patients to infections 2 . During the last decade, bacterial cellulose (BC) has attracted great attention for its wide range of biomedical and tissue engineering applications 3 . It can be used as a temporary skin substitute in the treatment of difficult wounds, burns and ulcers or as durable support required for tissue engineering. BC consists of a translucent and gelatinous film, formed by cellulose nano and microfibrils and as such BC is one of three nanoscale-sized, that can be produced by several microorganisms, such as Komagataeibacter, Agrobacterium, Rhizobium, Sarcin, Alcaligenes. The chemical structure of BC is composed of (1-4) D-glucopyranose chains with β-type glycosidic bonds 4 . The physical and mechanical properties of BC membranes are related to their well-spaced nano and microfibrils, thus creating an extensive surface area with spacings that allow the retainment of a large amount of liquid, while maintaining a high degree of order. The hydrogen bonds between these fibrous units stabilize their structure and offer a high mechanical resistance 5 . Based on its recent tested clinical performance, and in accordance with several other studies 4,6 on the properties of this biomaterial, BC can be considered an ideal material for high quality dressings 6 . However, BC itself has no antibacterial activity to prevent wound infection. In this sense, great efforts have been devoted to the development of antibacterial BC membranes, especially those containing silver nanoparticles (AgNPs) 7 . Silver and its compounds have known antibacterial and antifungal activity. Yet, at the nanoscale, this action is amplified 8 . According to Li et al. 9 , thermic treatment of silver, besides aiding in the functionalization, can induce the growth of AgNPs, that is, silver in the form of nanoparticles. Cerium is a metal with bacteriostatic properties and is effective in the form of cerium nitrate (Ce(NO 3 ) 3 ) against a wide range of bacteria. It has potent antimicrobial action and low toxicity to mammalian cells. It is believed that cerium nitrate exerts a protective effect against post-burn immunosuppression caused by lipidprotein complexes 10 . In this context, this work proposes the synthesis of BC membranes with different concentrations of Ce(NO 3 ) 3 and AgNPs, submitting them to different thermal treatments, and characterization techniques (physical-chemical, thermal, antibacterial activity and cytotoxicity).

Methodology
The process of synthesis and characterization of functionalized BC membranes submitted to different heat treatments can be more clearly described in the schematic illustration shown in Figure 1.

Synthesis and purification of BC membranes
The microorganism used was the bacterium Komagataeibacter hanseni ATCC 23769, preserved in a refrigerator. For its activation and cultivation, mannitol medium (MM) was used, containing C 6 H 14 O 6 (20 g·L -1 ), soy peptone (5 g·L -1 ), yeast extract (5 g·L -1 ), Na 2 HPO 4 (2.7 g·L -1 ) and C 6 H 8 O 7 (1.15 g·L -1 ), autoclaved at 121 °C for 20 min. The cells were activated in 500 mL Erlenmeyer flasks containing 100 mL of MM, incubated at 30 o C under static conditions for 2 days. After this time, the inoculum was transferred to the culture medium at a ratio of 20% with optical density (O.D.) between 0.15 and 0.19. This step was carried out in 250 mL Erlenmeyer flasks with 50 mL of culture medium and/or in 96-well plates with 200 μL of culture medium, both incubated in an oven at 30 °C under static condition for 12 days for the formation of hydrated BC. The synthesized BC was separated from the culture liquid, washed with distilled water, and purified with a 0.1 M NaOH solution at 80 °C for 60 min to remove the bacterial cells and other impurities. Then, the pellicles were washed with distilled water until reached a of pH 7, and sterilized at 121 °C for 15 min.

Functionalization of BC membranes
The solution containing 40 ppm of AgNPs was purchased from Diver Saúde, Curitiba, Brazil. Ce(NO 3 ) 3 also purchased from Neon, São Paulo, Brazil as Ce(NO 3 ) 3 ·6H 2 O, 99% A.P. The solutions containing Ce(NO 3 ) 3 and AgNPs were functionalized into BC membranes in different concentrations, based on the commercial drug formulation Dermacerium® ( Table 1). The membranes were manually pressed with absorbent paper to remove excess of water to enhance the absorption of the AgNPs and Ce(NO 3 ) 3 after immersion in the respective solutions for 24 h, rotating every two hours to ensure the total absorption of the substances by the membranes 11 . The membranes were then submitted to two different thermal treatments, aiming to establish the functionalization between the Ag and the BC hydroxyls 11 . The first one was performed in an oven at 120 ºC (O) and the second one in an autoclave at 121 ºC (A), both for two hours.

Lyophilization of BC
The membranes were lyophilized to remove all water without collapsing the pores. The membranes were previously frozen and dehydrated by the lyophilization process using Terroni LT 1000 Lyophilizer for 24 h and stored in a desiccator for further characterization analyses.

Transmission Electron Microscopy (TEM)
TEM analysis (FEI Tecnai G 2 F20, equipped with EDS system) of AgNPs was performed by using two drops of the concentrated colloidal dispersions, which were carefully dripped over copper grids (300 mesh) coated with carbon. The grids were dried overnight at room conditions before being analyzed by the microscope operating at 200 kV. The software ImageJ (public domain). The software ImageJ was used to evaluate the images and determine the particle size.

Rehydration capacity
The membranes were lyophilized and weighed to obtain the dried mass (dm). Then, the membranes were immersed in 200 mL of deionized water and kept at room temperature up to 96 h. The mass of the rehydrated BC (rm) was determined after 2, 24, 48, 72 and 96 h. Before weighing, the excess water was drained for 60 seconds. The water reabsorption capacity was obtained by equation 1 12 . .

Thermogravimetric analysis (TGA)
TGA analyses were performed to determine the thermal degradation behavior (thermal stability), mass loss, based on the onset temperature for degradation (T onset ) and the maximum degradation temperature (T max ) of the membranes. TG curves were obtained using the equipment TGA-Q50 (TA Instruments). Samples were heated from 25 to 1000 ºC at a heating rate of 10 ºC min -1 under an oxidizing atmosphere.

Scanning electron microscopy with dispersive energy spectroscopy (SEM/EDS)
Scanning electron microscope JEOL (model JSM-6390LV) was used to acquire the images of membrane surfaces to evaluate their microstructural characteristics. Samples were fixed on metal supports and covered with thin gold layer. Electron beam and X-rays were emitted toward the samples and the reflected signals were detected by a solid-state silicon detector for a point elemental analysis.

Wide-angle X-ray diffractometry (DRX)
X-ray diffraction was measured with an X-ray diffractometer (Model XRD8 Advance, Bruker AXS). The diffractograms were recorded using CuK-a radiation (k = 1.5418Å), 40 kV and 25 mA = 1000 W. Samples were scanned from 5 to 60° at a scan speed of 1°/min. The X´Pert HighScore Plus software was used to analyze the diffractograms and determine the crystallinity degree (Xc) of the membranes. The Xc was determined as a function of the areas corresponding to the crystalline peaks and the amorphous halo (peak extended under the crystalline peaks) 13

Xc 100
Ac Aa where Ac represents the sum of the areas of the crystalline peaks and Aa corresponds to the area of the amorphous halo.

Atomic absorption spectroscopy (AAS)
The amount of Ag contained in the functionalized BC membranes was quantified by AAS using the equipment Varian model 55B SpectrAA. Samples were previously dried in an oven at 105 ºC for 2 h. The mass of each sample was determined in an analytical balance, and the samples were digested with 10 mL of 50% (v/v) HNO 3 in a digestor block for 2 h at 105 °C. After the digestion process, the samples were cooled to room temperature, and centrifuged at 3500 rpm for 20 min for phase separation. Then, the supernatant was collected and the Ag content in the samples was measured in an AAS at wavelength λ = 328.1 nm, specific for Ag. Dilutions with HNO 3 50% v/v were necessary for the samples that extrapolated the reading ranges. After measuring the silver content in the solutions, the obtained values were divided by the individual mass of the samples, and the results were presented in mg/g.

Antibacterial activity
The antibacterial activity of the BC membranes was assessed using the disc diffusion technique of Kirby and Bauer 14  negative) and Staphylococcus aureus (Gram-positive) were used in the assays. These bacteria were pre-cultured for 24 h at 37 °C, in trypticasein soybean broth (TSB). The microorganisms tested were diluted with 0.9% of a sterile saline solution to achieve a concentration of about 0.5 of the McFarland standard. Thereafter, the bacteria suspension was evenly distributed on Petri dishes with Müller Hilton (MH) medium using a swab to form a layer of microorganisms. The neat and functionalized BC membranes were cut into 6 mm diameter discs and placed over the bacterial growth media. The inhibition zone test was used to evaluate the antimicrobial activity 15 .

Cytotoxicity
Mouse fibroblasts (L929 strain), multiplied in Dulbecco's Modified Eagle Medium (DMEM, supplemented with 15% fetal bovine serum) were used to evaluate the cytotoxicity of the pure and functionalized BC membranes. The biomaterials were autoclaved and placed in DMEM medium for 24 h. In parallel, 9x10 3 cells/well were placed in 96-well plates and incubated for 24 h at 37 °C / 5% CO 2 . After that, metal solutions were poured over the adhered cells, and incubated again at 37 °C / 5% CO 2 for 1, 3 and 7 days. The membranes were then removed, and cells were washed with phosphate buffered saline (PBS) to remove dead or non-adherent cells. Cytotoxicity test was performed by the MTS method 16 . The optical density reading was performed in the plate reader at 490 nm to calculate the relative cell viability.

Statistical analysis
The antibacterial and cytotoxicity assays were performed in triplicate, and statistical analysis was performed using one-way ANOVA followed by the Tukey test for multiple comparisons, with 95% confidence. The analyses were performed using the Minitab17 software.

Transmission electron microscopy (TEM)
In this study, TEM was applied to determine the particle size of the AgNPs, as shown in the images displayed in Figure 2. Particles with an average size of 20 nm were observed, as well as the presence of some agglomerates. EDS analysis of a single particle showed a predominance of Ag, as expected. The presence of Cu in the analysis might be due to the grid, while the Cl and Si might be attributed to some compounds added to the formulation used by the company (Diver Saúde) to synthesize the colloidal silver.

Characterization of BC membranes
As shown in Figure 3, macroscopically homogeneous membranes were obtained with colours ranging from white, for the pure BC, to brownish for the membranes containing AgNPs and Ce(NO 3 ) 3 . Darker shades were visualized by increasing the concentration of AgNPs. Color changes in a depended fashion with the concentration of AgNP and Ce(NO 3 ) 3 and are a visual indication that the antibacterial substances were incorporated into the membranes. Li et al. 9 used UV-Visible absorption to correlate the concentration of AgNPs with the variation in coloration from bright yellow to deep yellow as well as to brown.

Rehydration capacity
From Figure 4, it is possible to observe that the neat BC membrane reabsorbed 97.17% of water in the first 2 h of test, presenting slightly increase over time. These results demonstrate the efficiency of such membrane as wound dressing absorber. According to the Biocel user's manual, BC curatives may be kept sterile and dry. When applied, they should be hydrated to better adhere to the lesion, facilitating skin's suitability.  volume occupied by the AgNPs and Ce(NO 3 ) 3 inside the pores. The membrane submitted to the thermal treatment in an oven in the presence of only AgNPs showed a slightly slower water absorption rate in the first 48 h. However, the final rehydration capacity was basically the same for samples Ag40A and Ag40O after 96 h. In general, the functionalized BC membranes prepared with the highest concentrations of Ce(NO 3 ) 3 (Ag20CN2.2 and Ag40CN2.2) showed a lower rehydration capacity when compared to the neat BC and Ag40 samples. These results suggest that the presence of Ce(NO 3 ) 3 in BC membranes contributed for the reducing of the pore sizes or even clogging them, reducing the pathways for water absorption. This hypothesis is supported by comparing the rehydration capacity of the samples containing 2.2 and 1.1% of Ce(NO 3 ) 3 . The results shown in Figure 5 for such samples demonstrate that the water absorption by the membranes increased when the Ce(NO 3 ) 3 content was lowered. It is important to notice that the membranes submitted to a thermal treatment in the autoclave and in the oven presented a rehydration capacity with opposite behavior by increasing the concentration of AgNPs. Figures 6a and 6b show the TG curves for the functionalized BC membranes submitted to the oven and autoclave thermal treatment, respectively. Temperature data for the beginning The rehydration capacity of the functionalized membranes decreased in comparison with the neat BC (Fig. 5). These results suggest that the antibacterial compounds (AgNPs and Ce(NO 3 ) 3 ) might have been incorporated in the membranes through their attachment to the hydroxyl groups of BC. Based on this hypothesis, the membrane should have fewer free hydroxyls than the neat BC, reducing its affinity with water. In other words, the presence of AgNPs or Ce(NO 3 ) 3 in the membranes contributed to the reduction of hydrogen bonds between the water and the OH groups, decreasing the space available for the water diffusivity toward the core of the membranes. Molina et al. 17 reported that any factor that reduces the pore size of the BC membrane will have a direct effect on the transport mechanism in the hydrogel. The thermal treatment can collapse the pores of the membranes, hindering the rehydration of the BC membranes. Thus, the findings obtained in this study suggest that the rehydration behavior of the BC membranes might be influenced by the of degradation (T onset ), maximum degradation temperature (T max ), mass losses and residues are shown in Table 2. All BC membranes presented the same characteristic degradation profile with three mass loss events. The first stage of mass loss occurred between room temperature (≅ 30 ºC) and 150 ºC, referring to the water loss of the samples 18 . The second, and more significant stage, occurred with a T onset2 of 284 ºC and a T max2 of 301 ºC for neat BC, and is attributed to cellulose degradation 18 . For the functionalized BC membranes, T max2 was determined between 302 °C and 375 °C. The third stage represents the degradation of carbonaceous residues that can extend from 400 to 600 °C 19 . This third event was observed in this work in the range between 310 °C and 491 °C. Considering that the analysis was performed under an oxidizing atmosphere, the residual mass may indicate the presence of inorganic material in the sample 7 , suggesting the incorporation of the metallic elements in the BC membrane.

Thermogravimetric analysis (TGA)
It was also possible to observe that in the membranes with a higher concentration of the functionalizing agents, the percentage of residue increased, reinforcing this relation. Another fact also observed was that the highest residual percentages were found for the samples submitted to the thermal treatment in the oven, pointing this as the better functionalization strategy due to the greater amount of residue. Figure 7 shows the image obtained by scanning electron microscopy (SEM) analysis of the upper surface of the neat  BC membrane (magnification of 10,000x). As observed, the neat BC presents a dense network of nanofibrils organized in a random arrangement, forming a highly porous structure that allows a high liquid retention. Similar microstructure was reported in the literature by Fischer et al. 11 and Godinho et al. 20 . The diameter of the fibrils of the neat BC membrane, measured with the ImageJ software, is between 40 and 96 nm. Yan et al. 21 produced BC membranes from Komagataeibacter hanseni with fibrils in the same size range. Nevertheless, the nanofibers showed a high aspect ratio and, consequently, a fibrils network of BC with different morphological characteristics. SEM images of the functionalized BC membranes oventreated are shown in Figure 8. From the micrographies, it is not possible to observe the fibrils, since the oven treatment collapsed the pores, and the membranes presented a highdensity surface. The micrography of the Ag40O sample clearly shows that the AgNPs (white spots) are well dispersed in the BC membrane. However, although it is not possible to observe the particles in the micrographs of the oventreated samples containing AgNPs and Ce(NO 3 ) 3 , these were confirmed by EDS, which detected the presence of Ce and Ag in all samples.

Scanning electron microscopy with dispersive energy spectroscopy (SEM/EDS)
From SEM images (magnification of 10,000x) of the autoclaved functionalized BC membranes (Fig. 9), it is noticeable that the fibrillar structure was preserved in all the membranes. However, according to the data listed in Table 3, there was an increase in the fibrils diameter with the incorporation of AgNPs, although mainly when Ce(NO 3 ) 3 was used. These results suggest that fibrils might be coated by the compounds functionalized in the membranes. Fischer et al. 11 also observed a coating of BC fibrils by the incorporation of such substances. EDS analysis of sample Ag40CN2.2A (Fig 9) demonstrated that the incorporation of AgNPs and Ce(NO 3 ) 3 were higher in the autoclaved BC membranes than those submitted to the thermal treatment in the oven. Such findings are evidences that the antibacterial compounds were incorporated in the BC membranes. Shiny regions are observed on the fibrils surface in the micrographs of the membranes containing AgNPs and Ce(NO 3 ) 3 , and might be assigned to the presence of the metallic compounds used to functionalize the membranes. Barud et al. 22 also observed the presence of shiny points onto the surface of BC fibrils, and assigned them to the presence of AgNPs obtained from the hydrolytic decomposition of silver complexes with triethanolamine.

Wide-angle X-ray diffractometry (XRD)
In the XRD pattern of neat and functionalized BC membranes ( Figure 10) this was evidenced by crystalline peaks at diffraction angles at 2 θ = 14.55° (101), 16.88° (101) and 22.73° (002), all typical of Cellulose I 23 The different concentrations of AgNPs and Ce(NO 3 ) 3 with different thermal treatments did not show differences in the reflective angle and spacing "d" of the typical BC membrane (Table 4). However, the degree of crystallinity (Xc) of the membranes decreased by about 5.4% with the addition of AgNPs and Ce(NO 3 ) 3 simultaneously, whereas a slight increase of Xc was observed for the membranes containing only Ag (Ag40O and Ag40A). The presence of Ag was confirmed by the diffraction peaks at 38.3° and 46.5° (indicated by the arrows in Fig. 7a) in the membrane containing only AgNPs and thermally treated in an oven (sample Ag40O). These two diffraction peaks at 38.3° and 46.5° can be attributed to metallic silver corresponding to the crystallographic planes (111) and (200), respectively 24 . It was not possible to observe the characteristic peaks of AgNPs and Ce(NO 3 ) 3 in the other membranes due to the low concentration of these substances in the samples. The decrease of crystallinity of functionalized BC membranes can be associated with the cleavage of hydrogen bonds by inclusion of metal elements in the crystallites. This also observed by Casaburi et al. 23 in your work, when studied BC samples treated with NaOH.

Atomic absorption spectroscopy (AAS)
The results of the Ag concentrations obtained through the AAS technique are presented in Figure 11. The concentrations obtained for the samples Ag40O and Ag40A were higher because these samples contain only AgNPs at the highest concentration. The other samples had Ag levels ranging from 0.02 to 1.37 mg/g, proving that the AgNPs were functionalized into the BC membranes. In a general way, it is possible to conclude that the membranes thermally treated in the autoclave presented more AgNPs than those treated in the oven. These results might be a consequence of the porous structure of the autoclaved membranes, which have a higher surface area available for the attachment of such metallic compound.  It was found that there was no significant difference between treatments with AgNPs and Ce(NO 3 ) 3 in oven for both microorganisms. However, when the effect of autoclaving treatment on the different organisms tested was evaluated, a significant difference was observed for the different treatments.

Antibacterial activity
The largest inhibition was observed against the bacterium Staphylococcus aureus (Gram-positive) regardless of the thermal treatment. Some authors report that the antibacterial action of AgNPs against Staphylococcus aureus are higher than Gram-negative bacteria 25 . Generally, Gram-negative bacteria are more resistant to antibacterial agents than Gram-positive bacteria due to the complexity of the cell membrane of Gram-negative, which hinders the entry and, consequently, the diffusion of the AgNPs into these microorganisms 26 . Based on studies with individual strains from different cultures, Tomacheski et al. 27 reported that S. aureus is more resistant to AgNPs in comparison to to E. coli. In this way, the authors showed that the bactericidal efficiency of the nanoparticles does not depend only on the structure of the bacterial membrane. Therefore, it is not possible to generalize the antibacterial effect of AgNPs based on results of only one strain, since different bacteria exhibit different sensitivities as a function of the structures attached to the cell wall. However, some mechanisms of toxicity are suggested: a) the contribution of free silver ions followed by disruption of ATP production and DNA replication (AgNPs act on the phosphate of DNA molecules resulting in the inactivation of their replication); b) formation of oxygen reactive species by the ions and AgNPs. These molecules react with the sulfhydryl -SH groups of the cell wall to form R-S-S-R blocking respiration, and causing cell death; c) direct damage of the cell membrane by AgNPs, affecting the permeability of the membrane 27 .
The membrane thermally treated in the oven and containing only AgNPs showed the largest inhibition zone (13 mm and 17 mm for P. aeruginosa and S. aureus, respectively) and, consequently, the highest antibacterial efficiency. This behavior might be explained in terms of the microstructure of such membranes. The sample Ag40 (oven) presented a collapsed fibrillar structure, while sample Ag40 (autoclave) formed a porous fibrillar microstructure. So, the nanostructures incorporated in sample Ag40 (oven) might be concentrated on the surface of the membranes, whereas in sample Ag40 (autoclave), the nanostructures are supposed to be incorporated inside the porous of the membrane as well. In this situation, the diffusion of the nanoparticles toward the bacteria media from the autoclaved sample might be slower than from the sample treated in the oven. The fact that the membranes presented antibacterial activity is an evidence of the presence of AgNP and Ce(NO 3 ) 3 in the samples.

Cytotoxicity
A preliminary cytotoxicity assay is one of the most important evaluations of the biological properties of the biomaterials prior to in vivo assessment 28 . According to the ISO 10993-5 29 , if the cell viability is greater than 70% of the control group, then the material is considered as noncytotoxic. In the cytotoxicity assay performed in this work (Figure 13), it was possible to observe that Ag40A and Ag40O were the samples that presented the best effect on cell viability, revealing a similar behavior regardless the thermal treatment. For these membranes, the cell viability was maintained during the first 3 days of the test. However, the functionalized BC membranes with AgNPs e Ce (NO 3 ) 3 presented a negative effect over the cell viability over the 3 days of evaluation. These findings suggest that Ce(NO 3 ) 3 is more cytotoxic than AgNPs in the range of concentrations used in the assays. The obtained results indicated that the membranes functionalized with AgNPs could induce a slight decrease in the viability of the cells tested only after 7 days. Although cell viability remained in the range of 70%, considering the standard deviation for both samples, the statistical analysis showed that this difference is statistically significant in relation to the BC standard. Zhang et al. 30 explain that the reason for this slight cytotoxicity may be related to the transport of AgNPs into the cell via endocytosis. The AgNPs can be found in the mitochondria and in the nucleus, but their cytotoxic effect depends on the size and their surface modification. Ramenzoni et al. 31 investigated the influence of cerium chloride on fibroblast cells at different concentrations of cerium and demonstrated a positive response when fibroblast cells were used. Nevertheless, the authors concluded that this effect depends on concentration, time exposure, and cell culture. They emphasized that different cell types may also react differently when exposed to cerium, exhibiting various effects of cytotoxicity. Therefore, the implementation of topical application of cerium to accelerate wound healing merits further investigation. Correlating the data obtained from the BC microstructure, cell viability and antibacterial activity, the BC membranes that stood out were those autoclaved and prepared with only AgNPs (Ag40A) and prepared with 20 ppm of AgNP and 1.1% of Ce(NO 3 ) 3 (sample Ag20CN1.1A). These two membrane compositions might be recommended to be used as wound dressing, since they showed good porosity, good antibacterial activity and adequate cell viability on the first day of analysis.

Conclusions
Based on the characterizations performed for the membranes with different formulations, it can be concluded that the autoclaved samples Ag40 and Ag20CN1.1 are the most promising for application in the treatment of burnt skin. The BC membranes thermally treated in the autoclave exhibited a porosity higher than those treated in the oven. This characteristic allows a better gas exchange at the wound site, and good rehydration capacity, which are factors considered extremelly important, since the dressings should be rehydrated prior to application over the skin to improve adhesion. The autoclaved membranes prepared with the solution containing only 40 ppm of AgNPs and with 20 ppm and 1.1% of Ce(NO 3 ) 3 1.1% (Ag20CN1.1) showed the highest antibacterial efficacy. Nevertheless, for the sample Ag20CN1.1, it is necessary to change the dressing constantly due to the rapid lowering of cell viability. On the other hand, the autoclaved membrane containing only AgNPs is more advantageous to be used in the treatment of burnt skins, since there is no need for constant dressing change. Figure 13. Graph of % of cell viability over time for the samples thermally treated in the oven and in the autoclave. *significant differences with respect to the BC group (control) on the respective days of exposure (or incubation).