key: cord-254638-f86irz06 authors: Sunday, Michael Oluwatoyin; Sakugawa, Hiroshi title: A simple, inexpensive method for gas-phase singlet oxygen generation from sensitizer-impregnated filters: Potential application to bacteria/virus inactivation and pollutant degradation date: 2020-07-23 journal: Sci Total Environ DOI: 10.1016/j.scitotenv.2020.141186 sha: doc_id: 254638 cord_uid: f86irz06 Abstract Airborne infectious diseases such as the new Coronavirus 2019 (COVID-19) pose serious threat to human health. Indoor air pollution is a problem of global environmental concern as well. Singlet oxygen (1O2) is a reactive oxygen species that plays important role in bacteria/virus inactivation and pollutant degradation. In this study, we found that commercially available filters typically deployed in air purifier and air conditioning units, impregnated with Rose Bengal (RB) as a 1O2 sensitizer, can be used for heterogeneous gas-phase generation of 1O2. It was confirmed that irradiation of the RB filter under oxygen gas stream produced 1O2, which was measured using furfuryl alcohol trapping method followed by HPLC analysis. It was also observed that the amount of 1O2 generated increases as the light intensity increased. Similarly, the sensitizer loading also positively influenced the 1O2 generation. The heterogeneous gas-phase generation of 1O2 can find potential applications in air purifier and air conditioning units for the purpose of bacteria/virus inactivation and/or pollutant degradation thereby improving indoor air quality. J o u r n a l P r e -p r o o f 3 pollutants (Latch et al., 2003; Liu et al., 2011; Kim et al., 2012; Xie et al., 2018; Blacha-Grzechnik et al., 2020) and inactivation of bacteria/virus (Banks et al., 1985; Dahl et al., 1987 Dahl et al., ,1988 Lenard et al., 1993; Nitzan et al., 1995; Wainwright et al., 1998; Usacheva et al., 2001; Villen et al, 2006; Bartusik et al, 2012; Costa et al., 2012; Felgenträger et al., 2014; Kim et al., 2020) . Photodynamic inactivation (PDI), which involves generating ROS such as 1 O 2 upon irradiation of a sensitizer, has been deployed to inactivate a wide range of bacteria and viruses. PDI occurs via two mechanisms, namely: Type I mechanism involving the generation of free radicals, and Type II mechanisms involving 1 O 2 . Type II mechanism involving generation of 1 O 2 has been shown to be the predominant mechanism of bacteria/virus inactivation (Costa et al., 2012) . Comprehensive reviews highlighting the various bacteria (Hamblin and Hasan, 2004; Liu et al., 2015) and viruses (Costa et al., 2012) inactivated by 1 O 2 are available in literature. Viruses including human immunodeficiency virus (HIV), influenza virus, Sendai virus (Lenard et al., 1993) , dengue virus (Huang et al., 2004) , vaccinia virus (Turner and Kaplan, 1968) , adenovirus (Schagen et al., 1999) , enterovirus (Wong et al., 2010) etc. have all been reported to be inactivated by 1 O 2 . In addition, it was showed that herpesvirus and influenza virus (enveloped type) and adenovirus and poliovirus (unenveloped type) were inactivated by 1 O 2 using a spray of J o u r n a l P r e -p r o o f The above-mentioned reports majorly involved homogeneous 1 O 2 generation. However, homogeneous generation in aqueous phase cannot be easily deployed to provide 1 O 2 needed in the gas phase for potential indoor air purification. 1 O 2 can also be generated from heterogeneous systems where a solid-state sensitizer, either in isolation or deposited by physical and/or chemical modification onto another solid substance, is irradiated in the presence of oxygen gas to generate 1 O 2, which flows along with the gas stream into a collecting solution where it reacts with a substrate. Reaction of the formed 1 O 2 with the selective substrate in solution provides evidence of 1 O 2 generation from the solid-gas heterogeneous system. In addition to substrate solution, involvement of 1 O 2 in the deactivation of bacteria and viruses in air has also been reported. Kim et al (2020) showed that 1 O 2 inactivated micro-organisms in air up to a distance of 10 -15 cm away from the source. Several researchers have presented different experimental set-ups based on this heterogeneous system and demonstrated 1 O 2 generation from them with further application of the generated 1 O 2 for substrate degradation or bacteria/virus inactivation (Bartusik et al., 2012a (Bartusik et al., , 2012b Zamadar et al., 2009; Aebisher et al., 2010; Carpenter et al., 2015; Zhao et al., 2014; Hettegger et al., 2015) . Nevertheless, the potential application of 1 O 2 for the purpose of improved air quality has not been well explored. One way of achieving that is to develop products that can generate 1 O 2 in the gas phase. Such products will help to inactivate micro-organisms in the indoor environment and may find applications as air purifiers in places like small rooms or offices, on automobiles, trains and even hand dryers. This will make a significant contribution towards cleaner indoor air by removing bacteria/virus or pollutants present in the air. In this study, we present a simple, inexpensive set-up, designed from commonly available materials, for gas-phase 1 O 2 generation. A filter material typically employed in the air purifier J o u r n a l P r e -p r o o f 5 and air conditioning units was impregnated with Rose Bengal and irradiated using a panel of LED lights. Furfuryl alcohol (FFA) in solution was used as a substrate to provide qualitative and quantitative evidence of 1 O 2 generation from the filter. The reagents and materials used are listed in supplementary information (S1). RB stock was prepared by dissolving 10 g RB in 150 g of the supplied gel. Two pieces of filters of the same dimension (L × W: 7 × 2.2 cm) were cut-out from the same filter material and their initial weights were obtained. RB was physically impregnated into one of the filters by completely immersing it in the RB-dissolved gel for 5 minutes. The gel was necessary to help impregnate RB in the filter because of the hydrophobic nature of the filter. The other filter was treated with a blank gel and used as the control filter. Both filters were dried overnight using cold air from an air drier. Treating the control filter with the blank gel helped to account for the contribution of the gel to the weight of the filters after drying. The RB-treated filter was further subjected to air blowing to remove any loosely adhered RB particles on the filter. The images of the RB-treated and control filters are presented in Fig S2. Thereafter, the amount of RB impregnated into the filter was determined gravimetrically from the difference in the weight of the filter before impregnation and weight after impregnation (dry, RB-impregnated filter). The amount of RB per area of filter (mg/ cm 2 ) was then calculated. To investigate the effect of sensitizer loading on 1 O 2 generation, different filters were impregnated with varying amounts of RB. This was done by firstly preparing serial dilutions (10, J o u r n a l P r e -p r o o f 6 with these dilute solutions ( Fig S2) . The amount of RB impregnated into the filters from treatment with different solutions of RB was also calculated. A schematic diagram of the set-up in this study is shown in Fig. 1a the glass column where the RB-impregnated filter had been attached was then covered by a hydrophobic PTFE filter which was glued to the edges of the glass column and secured tightly to the sides of the column using transparent tapes. The column was stoppered at the top using a silicon stopper with hole. A capillary tube was inserted into the glass column through the hole of the stopper. The tube delivered oxygen gas to the surface of the RB-impregnated filter at a rate of 0.1 -0.2 L/min. The stopper ensured that the top of the column was air-tight preventing any loss of oxygen through the top. The stability of this assembly and monitoring potential leakage or damage was ensured by immersing the assembly firstly in ultrapure water for about 5 minutes prior to immersion in the substrate solution. This preliminary immersion test showed that there was no leakage or leaching of RB into the solution. This rules out any possibility of aqueous phase contribution to 1 O 2 generation. The test also showed that at the immersion depth, the gas stream was sufficiently contacting the solution and was not just directly escaping into the air in the laboratory. This was necessary to ensure optimal contact between the 1 O 2 in the gas stream Journal Pre-proof J o u r n a l P r e -p r o o f 7 and the substrate solution. This shows that potential 1 O 2 exiting the filter along with oxygen flow will contact the reacting solution. The irradiation light was suspended above the glass column. The light was from a locally fabricated LED panel (Excel Co. Ltd., Fukuyama, Japan) consisting of 60 LED bulbs emitting white light with a power output of up to 144 W when the full irradiation mode is adopted. The spectrum of the irradiation light was obtained using a miniature spectrophotometer device Four mL solution of 100 µM FFA in D 2 O was employed as a substrate to demonstrate 1 O 2 generation from the irradiation of impregnated filter. 1 O 2 reacts with FFA to produce 6-hydroxyl-2H-pyran-3(6H)-one (6-HP-one) as a major product (Haag et al., 1984) . 1 O 2 arriving in the solution was monitored indirectly through the degradation of FFA and corresponding formation of 6-HP-one by HPLC analysis (Supplementary info. S2). Although 1 O 2 was not generated in solution in this experiment, but in the gas-phase from the irradiation of RB-treated filter, the determined photoformation rate (calculated in solution) is indicative of the amount of 1 O 2 arriving in the substrate solution. This is expected to be lesser than that generated from the filter due to quenching of some 1 O 2 as it travels through the pores of the filter, or by the gas stream before successfully contacting the substrate solution. Therefore, the obtained information on 1 O 2 generation from the filter can be regarded as an estimate. J o u r n a l P r e -p r o o f 8 The 1 O 2 in the oxygen gas stream passing through the filter and arriving in the substrate solution was monitored indirectly by following the peaks of FFA degradation and 6-HP-one formation. Monitoring 6-HP-one formation confirms that the degradation of FFA is due to chemical reaction with 1 O 2 arriving from the filter into the substrate solution. A typical plot showing the degradation of FFA and corresponding formation of 6-HP-one is shown in Fig. 2a . Furthermore, the degradation of FFA due to 1 O 2 followed a first order kinetics. The first order plot of FFA degradation during irradiation of RB-treated filter and untreated (blank) filter is shown in Fig. 2b . Irradiation of the RB-treated filter caused a significant degradation of FFA compared to the blank filter. The value obtained for the blank filter was similar to that obtained when the FFA solution was directly irradiated with oxygen flowing into the solution in the absence of the filter assembly. This suggests that the blank filter had no significant contribution to 1 O 2 formation and that the RB-treated filter was solely responsible for the formed 1 O 2 . In addition, there was no leaching of RB into solution during the irradiation. This is because the hydrophobic PTFE filter shielded the RB-treated filter from directly contacting the solution. This is a very important precaution because RB is highly soluble (100 mg/mL) in water. Therefore, any contact between the RB-treated filter and the solution will encourage a massive leaching or dissolution of RB. Under such scenario, there will be aqueous phase contribution to 1 O 2 generation. No leaching of RB, in part or whole, was observed during our experiments. This shows that the only source of The solvent isotope effect and the influence of 1 O 2 scavengers are two important ways of demonstrating the generation and involvement of 1 O 2 in a particular reaction. To confirm the presence of 1 O 2 in the gas stream arriving in the substrate solution, the solvent isotope effect was (Haag et al, 1984) for the reaction of 1 O 2 with FFA in H 2 O, the 100 µM FFA concentration only made an insignificant 4% contribution to the deactivation rate constant of 2.5 × 10 5 s -1 for 1 O 2 in H 2 O. This shows that the solvent isotope effect is mainly due to collision with the solvents, without any significant contribution from the physical quenching of 1 O 2 by FFA. In addition, the effect of 1 O 2 scavenger on the degradation of FFA in D 2 O was studied using NaN 3 as a scavenger. NaN 3 is an efficient scavenger of 1 O 2 with a second order rate constant of Journal Pre-proof J o u r n a l P r e -p r o o f 7.8 × 10 8 M -1 s -1 (Wilkinson and Brummer, 1981) . The degradation of FFA in D 2 O was observed to reduce significantly in the presence of 1 mM NaN 3 compared to its absence (FFA + D 2 O, no NaN 3 ) (Fig. 3) . This further confirms the generation of 1 O 2 . The results of the solvent isotope effect and NaN 3 scavenger experiment clearly confirm that 1 O 2 is present in the gas stream arriving in the substrate solution, and it is responsible for the degradation of FFA. To investigate how light intensity can influence the generation of 1 O 2 , the intensity of the irradiation light was varied by employing the full-power mode and half-power mode of the irradiation light and/or increasing the distance between the filter and irradiation light. A plot of the amount of 1 O 2 detected in solution, during the irradiation of RB-treated filter, as a function of light intensity is shown in Fig. 4 . It is observed that the amount of 1 O 2 generated increased as the light intensity increases. The sharp increase between 1196 and 1861 μmole s -1 cm -2 suggest a regime of light intensity where higher production of 1 O 2 is observed. Based on the present experimental set-up, the distance between the irradiation light and filter is 11 cm. In an application where the light source is closer to the filter, the light intensity will be much higher and 1 O 2 generation can be expected to be much higher. The effect of the amount of RB adsorbed by the filter on 1 O 2 production was investigated by Table 1 . It is observed that 1 O 2 generation increased as the amount of RB adsorbed on the filter increased. The maximum solubility of RB in the gel was 10 g RB in 150 g of gel. This was the same solution used as the undiluted stock in Table 1 . Therefore, the effect of amount of adsorbed RB on 1 O 2 generation could not be studied beyond this point. Using the undiluted stock as the highest RB loading possible, the amount of 1 O 2 reaching the solution was calculated to be 12.8 μmole s -1 of 1 O 2 per mole of RB. It should be noted that the amount of 1 O 2 generated on the surface of the filter may be several orders of magnitude higher than the value reported here due to the heterogeneous distribution of 1 O 2 generation. It has been demonstrated for aqueous solutions that 1 O 2 generation in the DOM microenvironment could be up to three orders of magnitude higher than in the bulk solution due to physical quenching as the 1 O 2 diffuses into the bulk solution (Latch et al., 2006; Grandbois et al., 2008) . A similar explanation could be adopted here where the 1 O 2 generated on the surface of the filter may undergo significant quenching before arriving in the substrate solution. This quenching may be caused by the filters (both treated filters and hydrophobic PTFE filter) as the gas stream containing 1 O 2 migrates from the irradiated side of the filter to the substrate solution. Also, quenching by ground-state molecular oxygen may also arise. Therefore, the quantity of singlet oxygen reported here can only be considered as a minimum. Nevertheless, in the real applications, the filters will be directly exposed to light, while air is circulated or brought into contact with the filter. Under such circumstances, the amount of generated on the surface of the filter is >> than 1 O 2 migrating further away from the filter into the air. Therefore, contaminated air will encounter a large amount of 1 O 2 on the surface of the filter that will be significant towards bacteria/virus inactivation. (Schweitzer and Schmidt, 2003) while that of ·OH ranges between 0.01 -1s (Crosley et al., 2017) . The lifetimes of both ·OH and 1 O 2 may be lower depending on the nature of other substances present. Hence, these values can be considered to be their upper limit values at ground-levels in the atmosphere. Nevertheless, the use of RB as sensitizer, which has a high 1 O 2 quantum yield (0.76) (Wilkinson et al., 1992) , can ensure relatively high production of 1 O 2 . Therefore, more 1 O 2 can be available for bacteria/virus inactivation in 1 O 2 -based air purifiers compared to ·OH in ·OH-based systems. In addition, the system proposed in this study makes use of simple, cheap and commercially-available LED light sources emitting white light in the visible region, in-line with the λmax of RB (550 nm). Such light source is relatively simple, easier to handle and maintain. The findings from this study can contribute significantly to the development of 1 O 2 -based air purifiers with potential applications in small or large indoor environments in offices, at homes or even on automobiles and trains. Such products are necessary to contribute to improved indoor air quality considering the outbreak of airborne viral infections such as the present COVID-19. Although this is a preliminary study, the findings from this study have led to the development of a prototype where additional parameters necessary for its deployment are being studied. Generator Device Serving as a Point Selective Sterilizer The cytotoxic and photodynamic inactivation of micro-organisms by Rose Bengal Bacteria inactivation by a singlet oxygen bubbler. Identifying factors controlling the toxicity of 1 O 2 bubbles Generating 1 O 2 bubbles: A new mechanism for gas-liquid oxidations in water Efficient generation of singlet oxygen by perylene diimide photosensitizers covalently bound to conjugate polymers Gas-phase photolytic production of hydroxyl radicals in an ultraviolet purifier for air and surfaces Pure 1 O 2 cytotoxicity for bacteria Comparison of photodynamic action by Rose Bengal in gram-positive and gram-negative bacteria Inactivation of Ebola virus and Middle East respiratory syndrome coronavirus in platelet concentrates and plasma by ultraviolet C light and methylene blue plus visible light 1 O 2 generation in porphyrin-doped polymeric surface coating enables antimicrobial effects of Staphylococcus aureus Microheterogeneous concentrations of singlet oxygen in natural organic matter isolate solutions Singlet oxygen in surface waters Photodynamic treatment of adenoviral vectors with visible light: An easy and convenient method for viral inactivation Physical mechanism of generation and deactivation of singlet oxygen Photoinactivation of vaccinia virus with rose bengal Comparison of the methylene blue and toluidine blue photobactericidal efficacy against gram-positive and gram-negative microorganisms Aerosol and Surface Stability of SARS-CoV-2 as Compared with SARS-CoV-1 Solar water disinfection by photocatalytic singlet oxygen production in heterogeneous medium Photobactericidal activity of phenothiazinium dyes against methicillin-resistant strains of Staphylococcus aureus Quantum yields for the photosensitized formation of the lowest electronically excited singlet state of molecular oxygen in solution Methylene bluemediated photodynamic inactivation as a novel disinfectant of enterovirus 71 A mechanistic kinetic model for singlet oxygen mediated self-sensitized photo-oxidation of organic pollutants in water 1 O 2 delivery through the porous cap of a hollowcore fiber optic device 1 O 2 generation on porous superhydrophobic surfaces: Effect of gas flow and sensitizer wetting on trapping efficiency Photocatalytic air cleaners and materials technologies -Abilities and limitations We are grateful to Yokota Kogyo Shokai Co. Inc., Hiroshima, Japan for providing research funding for this study. We also appreciate Prof. Kazuhiko Takeda of the Graduate School of Biosphere Science, Hiroshima University, for the advice rendered during this study.Journal Pre-proof J o u r n a l P r e -p r o o f The authors declare that there is no conflict of interest regarding this publication.