Synthesis variables effect on TiO 2 /polyester fabrics photoactivity and stability

TiO 2 -SiO 2 coatings on fabric samples were obtained by immersion of 9cm 2 of commercial polyester in a TiO 2 -SiO 2 suspension. This suspension was prepared by adding TiO 2 P25 to a silica matrix obtained from the hydrolysis of tetraethyl orthosilicate (TEOS). The aim of this work was to study the relationship between the immersion time of the fabric on the photocatalytic activity and stability of the coating against several cycles of use and washing. Obtained samples were characterized by scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS) and infrared spectroscopy (FTIR-ATR). Evaluation of modified fabrics was performed by self-disinfection tests under 250W/m 2 of simulated sunlight using E. coli as model bacteria. In general, all modified fabrics (except that obtained at 2h immersion time) inactivated bacteria in less than 120min and its regrowth was prevented for at least 24h after the disinfection tests. It was determined that low immersion time (3-4h) lead to active and stable fabrics to 3 cycles of use and washing, whereas higher immersion times (12h) lead to more active but unstable fabrics. activity of the times.


Introduction
A self-disinfecting fabric has the ability to inactivate pathogenic microorganisms such as E. coli because on its surface a photocatalyst such as TiO 2 is fixed and ready to be activated in the presence of UV light, thus leading to the production of oxidative species that attack the cell wall, and cause lysis and death of the microorganism [1][2][3][4][5][6]. The development of these fabrics, as a strategy to combat the spread of nosocomial infections [7][8][9][10][11], is still in the primary state due to difficulties to immobilize TiO 2 on the fabric surface, so as to keep it active and stable for several cycles of use and washing [12][13][14]. Different alternatives have been explored in order to overcome this issue. The use of bonding agents such as silica is one of the most attractive ideas due of its usability and low cost [15][16][17][18][19]. In this case, the linker has two functions: trap TiO 2 particles and strongly adhere them to the fabric, allowing UV light activation of the semiconductor. In 2008, Li et al. [20] studied the interaction between silica and certain fabrics such as cotton, linen, nylon and polyester, and determined that the adhesion of the coatings of SiO 2 in synthetic fabrics is mainly the result of thermodynamic affinities, dipole-dipole interactions and hydrogen bonds, while the adhesion over fabrics based on cellulose is produced by covalent bonds formed between the precursor groups of the silica and hydroxyl groups of the fabric. In 2010, Mejia et al. [21] reported that different fabric samples coated by the immersion of polyester in a suspension of TiO 2 -SiO 2 were sufficiently active to degrade gaseous methanol. They also found a relationship between the immersion time of the fabric and its photocatalytic activity. Despite of the progress made by these authors regarding immobilization of TiO 2 , it is still necessary to study the behavior of fabrics coated with TiO 2 -SiO 2 on self-disinfection processes and their stability during several cycles of use and washing. For this reason, the focus of this paper is to obtain self-disinfecting fabrics by immersion of polyester samples in TiO 2 -SiO 2 suspensions and determine the influence of immersion time on the activity and stability during three cycles of use and washing.

Modification of PE samples with TiO 2 -SiO 2
It was used one type of commercial polyester fabric (PE, polyester:cotton 65:35). Fabric samples of 9cm 2 were coated with TiO 2 -SiO 2 by a single immersion in the TiO 2 -SiO 2 suspension during a variable time between 2 to 12h, to adhere different amounts of TiO 2 -SiO 2 in the fabric. During immersion, the suspension is kept stirred at 100rpm to keep homogeneity and prevent precipitation of TiO 2 . Afterwards, the sample was removed from the suspension and dried in stove at 30±2°C for 15min. Finally, the TiO 2 -SiO 2 coated sample was soaked in distilled water and dried in oven at 100°C. The weakly adhered coating particles were detached from the fabric by washing with water under sonication (model LC-ULTRASONICS ELMA 30H) for 15min at 60Hz. The fixed TiO 2 -SiO 2 in the fabric was determined by difference in weight of the fabric before and after the coating process.

Characterization of fabric samples
Samples were analyzed by scanning electron microscopy (SEM) by using a FEI Quanta FEG TM model 650 provided with a BS-SE-detector and an energy dispersive spectrometer (EDS), under high vacuum (6x10 -4 Pa) and vacuum (10-130Pa) conditions, resolution of 6 to 100000x, voltage of 200V to 30kV and peak current continuously adjustable ≤200nA. Infrared spectroscopy (ATR-FTIR) was done using an Attenuated Total Reflectance cell (ATR) equipped with a ZnSe crystal to avoid damage of the polyester. The equipment used to perform spectroscopy was a Bruker Tensor 27 FTIR model supplemented with Bruker ATR Platinum.

Evaluation of photocatalytic activity of TiO 2 -SiO 2 coated samples Bacterial suspension.
The bacterial strain used as model microorganism was E. Coli ATCC 11229 (E. coli) provided by Microbiologics® Inc. This microorganism was preserved in cryobanks at -70°C. Two bacterial colonies from cryobank were inoculated into 5mL of nutrient medium Luria Bertani (LB) and incubated for 8h at 35°C with constant aerobic agitation of 100rpm. Then, the incubated culture broth was added to 25mL of fresh LB and remained in growth for 15h at the same temperature and stirring conditions until reaching the stationary growth phase. Subsequently, 1mL of the resulting culture was centrifuged at 3000rpm for 15min, as a result a biomass pellet was obtained to be twice washed with sterile saline solution (0.85g NaCl diluted in 100mL of distilled water). Finally, the washed biomass pellet was suspended in 1mL of sterile water, reaching a concentration of approximately 10 7 colony forming units (CFU)/mL.

Fabrics Self-disinfection tests.
Duplicate modified PE samples, were contaminated with 100mL of bacterial suspension with a concentration of 10 7 CFU/mL (C 0 ). Then, samples were irradiated at 250W/m 2 in a solar simulation chamber (SUNTEST CPS+, with temperature control, radiation between 300-800nm with 7% of photons emitted in the UV-A to B ranges and no photons emitted in UV-C range). Samples were withdrawn from the chamber at different times up to 120min to track their photodisinfection efficiency. Quantification of microorganisms in each sample (C) was conducted by the technique of serial dilutions, sowing on plate count agar (Merck) and plate count by microdroplet [22]. Additionally, the following control tests were performed: first, disinfection test without irradiation, second, disinfection test with the polyester fabric unmodified, third, disinfection test with the polyester fabric coated with SiO 2 .
Stability of modified PE samples. The selfdisinfection test of each sample was performed 3 times in a row (reuses) in order to determine the stability of the coating. Before each reuse, the coated fabric was washed with distilled water to remove residual organic material from the preceding disinfection. Moreover, the percentage of detachment TiO 2 -SiO 2 from modified PE samples (% Detachment) during the first use was determined. The following formula was used: %Detachment = ((W b -W a ) /W b )*100, where W b and W a mean weight before and after first use, respectively.

Effective disinfection time (EDT 24 ).
After each self-disinfection test it was determined the EDT 24 for all the modified PE samples. In this case, the samples that were subjected to disinfection tests are stored in dark for 24h. After this time regrowth microorganisms on the sample is quantified. EDT 24 determination was used as an indication of any microorganism regrowth after 24h of no light irradiation after the photocatalytic treatment [5,6]. Figure 1 shows SEM micrographs of: a) unmodified PE sample and b) to d) modified PE sample with 4h immersion time in different approaches. By comparing the a) and b) micrographs it is observed that the modification of PE sample produces a thick and cracked coating of TiO 2 -SiO 2 . The c) and d) micrographs shows that TiO 2 -SiO 2 coating is also amorphous and porous. It was determined in the SEM-EDS analysis that this coating is composed by a TiO 2 /Si mass ratio of 2.26 which is similar to that used in the synthesis of the coating. Figure 2 presents FTIR-ATR spectra of modified PE samples with TiO 2 -SiO 2 to different immersion times. All spectra have the same peaks with different intensities. In general the bands shown in IR spectra are attributed to both fabric functional groups and coating species. First, the bands at 1713cm -1 , and 1300-1600cm -1 are associated to vibrations of saturated esters due to carbonyl groups (-C=O), and -C-H bonds associated to the benzene ring [2], respectively. Second, there are bands at 810 and 1095cm −1 ascribed to symmetric vibration of Si-O-Si and to asymmetric stretching of Si-O-Si, respectively [23][24][25]. The band at 1254cm -1 is associated with disorder modes of SiO 2 amorphous layers [25]. The band at 720cm -1 is attributed to TiO 2 [4,24]. Additional bands are observed in the ranges 3000-3800, 1610-1650, 850-990cm -1 . The band located in the first range corresponds to fundamental stretching vibration of the hydroxyl groups. These hydroxyl groups derived from adsorbed water (3400-3500cm -1 ), silanols linked to molecular water through hydrogen bonds (3540cm -1 ), and Si and OH pairs mutually linked by hydrogen bonds and internal silanol (3660cm -1 ) [26]. The second range is assigned to the deformation of molecular water. There is a contribution near to 1650cm -1 which may be due to residual ethanol. The third range is associated with silanol bonds vibration (Si-OH) [25,26]. The Peak in the 2800-2980cm -1 region is associated to C-H stretching vibrations of CH 3 and CH 2 attributed to the PE sample structure or contamination [17].  On the other hand, Figure 2 shows that the intensity of both TiO 2 and Si-O-Si symmetric vibration peaks increases with immersion time but the intensity of Si-O-Si asymmetric stretching peak is similar for all analyzed samples. These results indicate that the increase of immersion time causes an increment in the disorder modes of the SiO 2 network that it is associated to increase in the TiO 2 peak. Figure 3 shows the amount of adhered TiO 2 -SiO 2 to each PE sample as a function of the immersion time. An increase in the amount of adhered coating on fabric is observed when increasing immersion time. There was a raise in adhesion between 4 and 5.5h in relation to others immersion times.

Photocatalytic activity and coating loss of modified PE samples: effect of immersion time
In contrast, the adhesion of the coating tends to stabilize for immersion time nearby to 12h. This stabilization indicates that adhesion limit of TiO 2 -SiO 2 in PE samples is nearby to 25mgcm -2 . Figure 4 shows the relationship between immersion time and photocatalytic activity of modified PE samples. It is observed that after 3h of immersion time the modified PE samples can inactivate E. coli (10 7 CFU/mL) and achieve effective disinfection in times less than 120min under 250W/m 2 of simulated sunlight irradiation. Furthermore, the photocatalytic activity increases with the amount of adhered coating and this one increases with time of immersion as previously observed, although this tendency is not notable between samples with 4 and 5.5h immersion time which have a similar photocatalytic activity. In all cases the EDT 24 was equal to the inactivation time shown in Figure 4 indicating that the inactivation of E. coli was complete. Inactivation of E. coli and effective disinfection of samples was due to TiO 2 presence in the samples, which was observed by SEM-EDS. Since the nature of TiO 2 P25 was not modified as shown in IR analysis, it is expected that when irradiated with UV light below λ<385nm (UV-A to B range) in the presence of water or air this semiconductor generates oxidizing radical species (OH*, O 2 * -) able to degrade the cell wall of microorganisms [5,12,27]. These processes of radicals formation occurs only in the surface of coating because the TiO 2 particles located inside the SiO 2 matrix of TiO 2 -SiO 2 coating do not have access to the incident light. In this case the TiO 2 particles are not photoactivated [28]. Therefore, due to different morphologies, as observed in Figure 1, it is expected that semiconductor coverage do not allow the photocatalytic activation of the entire coating on the fabrics. An unexpected behavior is observed in modified PE samples with 5.5 and 12h immersion time since it was found a low difference in TiO 2 -SiO 2 adhesion ( Figure 3) but a significant difference in photocatalytic activity compared to other samples. This behavior evidences that the photocatalytic activity of modified PE samples may depend on the TiO 2 /SiO 2 ratio in the coating surface. Mejia et al. [21] showed that samples with immersion times up to 11h increase in %TiO 2 loading compared with lower immersion times. Here, a different result was observed for samples with 4 and 5.5h immersion time that show a higher difference in TiO 2 -SiO 2 adhesion ( Figure 4) but a negligible difference in photocatalytic activity as compared to higher immersion times. These samples have different amounts of TiO 2 -SiO 2 as shown in Figure 3, but  have a similar amount of TiO 2 in the coating surface. This indicates that increasing the immersion time cause a variation in the adhesion of the TiO 2 -SiO 2 and a variation in the TiO 2 /SiO 2 ratio of the coating surface, as well as it was observed by Selishchev et al. [29]. Another determining factor on the sample photoactivity is the adsorbed water on the coating because, as it explain above, the water is necessary for the production of oxidants species (Equation 2) that cause bacterial inactivation. FTIR-ATR analysis showed that the adsorbed water increases with the immersion time. Therefore, the increase of the photoactivity with immersion time can also be associated to the increment in the adsorbed water amount. Table 1 shows the percentage of detached coating from the modified PE samples during the first use. The samples with 2, 3 and 4h had negligible losses but samples with 5.5 and 12h immersion times have considerable losses. This indicates that increasing the immersion time causes a decrease in the adhesion strength of the aggregates located on the coating surface, possibly because it reaches the adhesion limit of TiO 2 -SiO 2 in PE samples, as previously discussed. Such a decrease in the resistance of the coating may be associated with the increased vibrations related with disordered modes of amorphous SiO 2 observed by FTIR-ATR analysis. Photocatalytic stability of modified PE sample Figure 5 shows the results of first, second and third use of the modified PE samples with 4h immersion time. Control test indicate that both the unmodified PE sample under irradiation and the modified PE sample in obscurity have no photoactivity, therefore, the photocatalytic activity of the modified PE samples with 4h immersion time in the three uses was caused by TiO 2 presence. An important difference is observed between the result of first and second use which could be associated to a low detachment of TiO 2 as shown in Table 1. A different result is observed between second and third use where it is not shown variation on photoactivity indicating that the TiO 2 -SiO 2 coating did not suffer significant changes during the evaluation and washing procedures. The similarity in the results presented in Figure 5 indicates that the fabrics are stable at least until the third use.
This stability is related with -Si-O-Si-groups from SiO 2 present in the coating as observed in IR analyses, these groups have the function of trapping the TiO 2 and keep it immobile over the fabric surface. Several authors have mentioned that the modified fabric without SiO 2 shows TiO 2 detachment and indeed a total loss of photoactivity caused by reuse of modified sample [19]. Accordingly, the detached coating from modified PE sample between the first and second use could be TiO 2 deposited in the coating surface that wasn't trapped by SiO 2 matrix

Conclusion
The use of a SiO 2 matrix obtained by the hydrolysis of TEOS, as a binder between the commercial polyester and TiO 2 allows the modified fabric to remain active and stable during self-disinfection tests. TiO 2 -SiO 2 /PE samples inactivate an E. coli inoculum (10 7 CFU/mL) in times of 120min under constant irradiation of 250W/m 2 . The immersion time of the fabric in the TiO 2 -SiO 2 suspension is directly related to the self-disinfecting activity of the TiO 2 -SiO 2 /PE. High immersion times (12h) lead to highly active but unstable samples, whereas smaller immersion times (2-4h) lead to less active but stable fabrics during three cycles of use and washing.