Transparent and flexible photocatalytic film comprising organophosphonate-modified polysilsesquioxane-anchored titanium dioxide: hydroxy group ratio and organic substituent on phosphorous atoms

Transparent and flexible photocatalytic films have attracted considerable attention in recent years. We previously prepared a film with titanium dioxide (TiO2) and an anchor layer of phenylphosphonate-modified polysilsesquioxane (PhPPS-low), which had a phosphonate group and a phenyl substituent; this film exhibited transparency and flexibility. In this study, we reported the differences in the hydroxy group ratio on the phosphorous atoms and the presence or absence of phenylene moieties. Three organophosphonate-modified polysilsesquioxanes (APPS-low, APPS-high, and PhPPS-high) were synthesized. All photocatalytic films using APPS-high, APPS-low, and PhPPS-high exhibited photodegradation of methylene blue and photocatalytic bactericidal effects on Escherichia coli, and hydroxyl radical generation was confirmed. In particular, the photocatalytic film with PhPPS-high showed the highest photocatalytic ability. The transparent and flexible photocatalytic films composed of titanium oxide, organophosphonate-modified polysilsesquioxane, and poly(bisphenol A-co-epichlorohydrin) were prepared. The effects of hydroxy group ratio and organic substituent on phosphorus atoms in these films were evaluated by appearance and photocatalytic ability. Film using anchoring layer with APPS-low was formed large cracks, while films with other anchoring layers were formed no cracks. However, all films were formed small cracks after a 10-day durability test. All films showed photodegradation ability of methylene blue, photoinduced hydrophilicity, and photocatalytic bactericidal effects on Escherichia coli.


Introduction
Photocatalytic materials are being actively researched for applications as self-cleaning [1][2][3], environmental cleaning [4][5][6], and photocatalytic antibacterial materials [7]. In recent years, flexible photocatalytic materials, such as films, membranes, textiles, and papers, have attracted research interest [8][9][10][11]. In general, there are two methods to prepare a photocatalytic layer in flexible photocatalytic materials: (i) mixing or hybridization with polymers and (ii) deposition on a flexible substrate surface. The mixing method involves the direct mixing of photocatalysts and polymers, such as silicones [12][13][14][15] and organic polymers [16][17][18][19], while the hybridization method involves the in situ polymerization of photocatalyst and polymer precursors, such as alkoxysilanes [20][21][22] and organic monomers [16,23,24]. The transparency tends to be low due to light scattering at the interface of the photocatalytic particles and polymer, while the mechanical strength is often high owing to the toughness of the matrix polymers; thus, it can be used for applications that do not require transparency such as environmental purification filters. In contrast, transparent photocatalytic films are needed for materials that require transparency, such as flexible devices and curved glass. The deposition of a photocatalyst on a substrate can easily achieve high transparency [25,26]; however, adhesion between the layers of the photocatalyst and flexible substrate is crucial. If the adhesion is poor, durability and transparency are significantly reduced. Therefore, adhesion improvement is necessary for (i) insertion of an anchor layer between the photocatalytic layer and flexible substrate [27,28] and (ii) hydroxylation of the flexible substrate surface using an oxygen plasma [29]. Another technique involves embedding the photocatalyst onto silicone surface [30].
We have previously reported the preparation and properties of a transparent and flexible photocatalytic film composed of TiO 2 nanoparticles, phenylphosphonatemodified polysilsesquioxane (PhPPS-low), and poly(bisphenol A-co-epichlorohydrin) (PBE) layers [28]. The PBE layer exhibited high flexibility and transparency and contained hydroxyl groups. The PhPPS-low layer caught the TiO 2 nanoparticles by forming a highly hydrolysis-stable phosphatitanoxane (Ti-O-P) bond [31][32][33][34]; this separated the siloxane chains and acid substituent owing to the π-π interactions between phenylene moieties [35,36], showed an affinity to PBE [37], and protected PBE from TiO 2 . The influence of the organic moiety type and the hydroxy group ratio on the properties of the photocatalytic films needs to still be clarified. Herein, we reported the differences in the hydroxy group ratio on the phosphorus atom and the presence or absence of phenylene moieties (PhPPS and alkylphosphonate-modified polysilsesquioxane (APPS)). The photocatalytic ability of these films was evaluated by studying the photodegradation of methylene blue and the photocatalytic bactericidal effect on Escherichia coli (E. coli). Furthermore, changes after a 10-day weathering test were examined.

Measurements
Nuclear magnetic resonance (NMR) spectra were recorded on a JEOL Resonance JNM-ECP 500 spectrometer (JEOL, Japan) at 23 ± 1°C. The 31 P{ 1 H} NMR spectra were obtained at 202 MHz and calibrated using a thin glass tube with 8.5% phosphoric acid-D 2 O solution as an external standard. Surface observation was performed via scanning electron microscopy (SEM) using a Regulus 8230 (Hitachi High-Technologies, Japan) with an acceleration voltage of 0.5 kV and confocal laser scanning microscopy (CLSM) using a color laser 3D profile microscope VK-8510 (KEYENCE, Japan). Scanning electron microscopy-energy-dispersive X-ray spectrometry (SEM-EDX) images were obtained using a SU9000 field emission scanning electron microscope (Hitachi High-Technologies, Japan) with energy-dispersive X-ray spectroscopy. The Ti/Pt ratios of the photocatalytic films were observed via X-ray fluorescence (XRF) using a Supermini 200 (Rigaku, Japan) with a Pd X-ray source. XRF samples were sputtercoated with platinum using a JEC-3000FC (JEOL, Japan). The sputter-coating was operated at 20 mA for 40 s at 1.5 Pa. Fluorescence spectra were measured using an FP-8600 spectrophotometer (JASCO, Japan). For the photodegradation of methylene blue and fluorescence probe method using coumarin, a SUV-4 UV lamp (AS ONE, Japan) was used as the UV irradiator (5 mW cm −2 @254 nm, which was placed 2 cm from the film). The weathering tests of the photocatalytic films were conducted using a xenon weather meter NX75 (Suga Test Instruments, Japan). One cycle was defined as follows: using an irradiance of 60 W m −2 , temperature of 63°C, relative humidity of 0%, and time of 102 min followed by an irradiance of 60 W m −2 , temperature of 38°C, relative humidity of 55%, and time of 18 min; the weathering test consisted of 20 cycles and corresponded to 10 days of sunlight in Tokyo, Japan.

Preparation of organophosphonate-modified polysilsesquioxane
Poly(3-mercaptopropylsilsesquioxane) (2.6 g,~20 mmol), AIBN (66 mg, 0.4 mmol), and diethyl phosphonatecontaining allyl groups (30 mmol) were dissolved in tetrahydrofuran (THF, 10.4 g). After bubbling with N 2 gas (~70 mL min −1 ) for 30 min, the solution was refluxed for 24 h. The solution was evaporated and then dissolved in dichloromethane (10 mL). Me 3 SiBr (30 or 60 mmol) was then added to the solution and stirred for 24 h. The solution was evaporated to eliminate dichloromethane and bromoethane, which was formed by the reaction of diethyl phosphonate and Me 3 SiBr. Then, 50 mL of methanol was added to the viscous liquid, followed by reflux for 24 h. After removal of Me 3 Si(OMe) and methanol by evaporation, the residue was dissolved in methanol and added to THF with stirring. The white paste-like precipitate was collected and dried to obtain organophosphonate-modified polysilsesquioxanes.

Preparation of photocatalytic coating films
Photocatalytic coating films were prepared according to a previously reported method [28]. Table 1 lists the conditions of spin-coating and heating. A 10 wt.% PVA aqueous solution was coated and heated on a glass substrate. A 15 wt.% PBE-THF solution was coated and heated. A 5 wt.% organophosphonate-modified polysilsesquioxane-MeOH solution was coated and heated. Finally, the titania sol was coated and heated.

Degradation of methylene blue by the photocatalytic coating film
The photocatalytic coating film was soaked in a Petri dish containing 2.5 μmol L −1 methylene blue aqueous solution in the dark. It was followed by illumination with a 254 nm UV irradiator at 22 ± 3°C. The photodegradation ability was evaluated by measuring the absorbance of methylene blue via UV-Vis spectroscopy.
Photocatalytic sterilization against E. coli by the photocatalytic coating film Photocatalytic bactericidal experiments were performed according to the literature [40,41]. The Petri dish, plastic stage, photocatalytic coating film, filter paper, and poly(ethylene glycol terephthalate) (PET) films were washed, sterilized, and dried. Filter paper soaked with sterile ultrapure water was placed in a Petri dish to maintain constant humidity. The plastic stage was placed on the filter paper, and the photocatalytic coating film was placed on the plastic stage. A 20 µL aliquot of E. coli IAM12119 T suspension was added dropwise onto the photocatalytic coating film. The PET film was placed thereon to prevent the bacterial suspension from drying, and the entire Petri dish was covered with a wrap. The UVA irradiation intensity was adjusted to 0.25 mW/cm 2 . The photocatalytic reaction was performed for 3 h. After the photocatalytic reaction, the bacterial suspension on the photocatalytic coating film was collected and diluted 10 5 -, 10 6 -, and 10 7 -fold with phosphate-buffered saline. Then, 100 µL of each diluted bacterial suspension was spread on agar medium and incubated. The number of colonies formed by the culture was counted, and the survival rate of E. coli was calculated via colony forming units per mL (CFU/ mL) measurement using the following equation: Survival rate (%) = 100 × (number of bacterial colonies after photocatalytic reaction for 3 h × dilution factor)/ (number of colonies at 0 h).
This test was conducted three times, and the average value was used to determine the survival rate of E. coli.

Fluorescence probe method using coumarin
The photocatalytic coating film was soaked in a Petri dish containing 5.3 mmol L −1 cumarin aqueous solution in the dark, followed by illumination with a 254 nm UV irradiator at 22 ± 3°C. The solution was measured via fluorescence spectroscopy.

Characterization of photocatalytic films
Photocatalytic coating films were composed of a TiO 2 layer, organophosphonate-modified polysilsesquioxane layer, PBE layer, PVA layer, and glass substrate. When the PVA layer of this coating film was removed by immersion in water, transparent and flexible photocatalytic free-standing films were obtained, as shown in Fig. 1. However, owing to risks such as film breakage, all measurements were performed using coated films.
The amount of TiO 2 on the photocatalytic films was calculated using XRF. To compare the amount of TiO 2 , platinum sputtering was performed, and the amount of platinum was used as a standard. The Ti/Pt ratios of photocatalytic films with APPS-low, APPS-high, and PhPPShigh were 2.5, 1.4, and 1.2, respectively. The thickness of the TiO 2 layer increased in the order APPS-low > APPShigh > PhPPS-high.
The surface morphologies of the photocatalytic films were observed by SEM (Fig. 2). Fresh film with APPS-low showed large cracks, while no cracks were observed in fresh films with APPS-high and PhPPS-high. Cracks were found in the absence of TiO 2 via SEM-EDX (Fig. S1). After a 10day weathering test, photocatalytic films with APPS-high and PhPPS-high formed cracks; hence, the film surface condition changed over time.

Properties of photocatalytic films
The photocatalytic ability of these films was evaluated through the photodegradation test of methylene blue using UV irradiation at 254 nm and the photocatalytic bactericidal effect on E. coli via UVA irradiation at 315-400 nm. Figure 3 shows the absorbance of the methylene blue supernatant at 664 nm. The photodegradation ability of the fresh photocatalytic films was in the order of PhPPS-high > APPS-high > APPS-low. The PhPPS-low a 42 49 9 a This data is from our previous work (ref. [28]) photodegradation ability of the films was approximately maintained even after the weathering test. Figure 4 and Table 3 show the survival rates of E. coli on the photocatalytic films. The photocatalytic sterilization against E. coli on fresh films was weaker than that on the commercial TiO 2 -coated glass TKC-304. The photocatalytic sterilization against E. coli on the films after the weathering test improved. These films also exhibited photoinduced hydrophilicity upon irradiation with 254 nm UV (Fig. S2). Methylene blue was oxidatively degraded by active radicals (•OH, •O 2 − ) and/or H 2 O 2 [42][43][44], and E. coli degradation proceeded via oxidative damage to DNA, proteins, and cell membranes by •OH, •O 2 − , and H 2 O 2 [7,45,46]. Hence, the increase in photocatalytic ability after the weathering test was potentially related to an increase in the number of hydroxyl radicals generated.
To investigate the improved photocatalytic bactericidal effect on E. coli, hydroxyl radical generation was observed using the umbelliferone fluorescence probe method, as shown in Fig. 5a [47][48][49]. As the amount of hydroxyl radicals generated increased, the emission intensity of umbelliferone increased. Figure 5b shows the emission intensity of umbelliferone; however, the emission intensities of the films after the weathering test was weaker than those of the fresh films. Therefore, the improvement in the photocatalytic bactericidal ability of the film after the weathering test was likely due to other causes and not the number of hydroxyl radicals. The lifetime of hydroxyl radicals was extremely short (1 ns), and the proximity of TiO 2 and E. coli was responsible for their catalytic bactericidal ability [50]. Elahifard et al. proposed that P-OH tended to trap the cell walls of E. coli [51]. We presumed that small cracks formed during the weathering test (Fig. 2), i.e., P-OH in the anchor layer  made it easier for bacteria to adhere to the photocatalytic film, although no experimental evidence is provided.
As per these results, the surface morphologies and photocatalytic ability of films differed by the organic substituent and hydroxy group ratio on the phosphorus atom. The photocatalytic film with PhPPS-high showed the highest photocatalytic ability.

Conclusion
Transparent and flexible photocatalytic films were prepared, and the differences in the hydroxy group ratio on    5 a Formation of umbelliferone through the reaction of coumarin with hydroxyl radical and b intensity of 455 nm emission emitted from umbelliferone excitated at 332 nm phosphorus atoms and the presence or absence of phenylene moieties were studied. The deposited amount of TiO 2 in the photocatalytic films increased in the order APPS-low > APPS-high > PhPPS-high determined via XRF. Fresh film with APPS-low showed large cracks, and all films after the weathering tests showed the formation of small cracks. The photocatalytic ability of the photocatalytic film before/after the weathering test had the photodegradation ability of methylene blue. The photocatalytic bactericidal ability of E. coli on the photocatalytic film after the weathering test was improved compared with that of the fresh photocatalytic film. Using the umbelliferone fluorescence probe method, the amount of hydroxyl radicals generated in the photocatalytic film after the weathering test decreased compared to that of the fresh film. Therefore, we presumed that contact between TiO 2 and bacteria was promoted by the formation of small cracks during the weathering test and the attraction of the cell walls of E. coli by P-OH. According to these results, the photocatalytic films with a high hydroxy group ratio exhibited good photocatalytic ability; in particular, the photocatalytic film with PhPPS-high showed the highest photocatalytic bactericidal ability against E. coli. Funding Open access funding provided by Tokyo University of Science.

Compliance with ethical standards
Conflict of interest The authors declare no competing interests.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/.