Water disinfection usually requires expensive chemicals or equipment. Chlorination is a common disinfection method, although it is not able to inactivate all pathogens. High concentrations of residual chlorine also cause an unpleasant taste and smell in drinking water. As an alternative, photocatalysis and photoelectrochemical treatment has a high disinfection potential in drinking water by using solid catalysts, such as titanium dioxide. Highly reactive hydroxyl radical generated during the process serves as the main oxidant, capable of inactivating a wide range of microorganisms. In this study, we proposed a novel comparison between Gram-positive and gram-negative microorganisms. An immobilized TiO2 film promoted higher efficiency in water disinfection processes. The treatment effectively inactivated Escherichia coli and Staphylococcus aureus bacterial microorganisms in a shorter period than other alternative methods.
Alternative technologies for water disinfection aimed at human consumption have been extensively investigated (Bekbölet & Araz 1996; Butterfield et al. 1997; Christensen et al. 2003). Disinfection processes include chlorine dioxide, ozonation, advanced filtration, germicidal ultraviolet light (UV) and radiation. Chlorination is one of the most important disinfection methods although it is not able to inactivate all pathogens. Nonetheless, high concentrations of residual chlorine cause an unpleasant taste and smell in drinking water. Chemical disinfection may also lead to harmful by-product formation, such as chloroform and other trihalomethanes, haloacetic acids and chlorite (Richardson & Postigo 2012).
Even though alternative disinfection processes may be effective, many of these require expensive chemicals or equipment to produce disinfectants. Disinfection with UV, for example, applies shortwave radiation (<280 nm) which requires expensive equipment associated with increased energy utilization.
Photocatalysis is a viable option with high disinfection potential in drinking water with the use of solid catalysts, such as titanium dioxide (TiO2). The development of photocatalysis has been the focus of considerable attention in recent years with photocatalysis being used in a variety of products across a broad range of research areas, including environmental and energy-related fields. The photocatalytic properties of certain materials have been used to convert solar energy into chemical energy to oxidize or reduce materials and to degrade pollutants and inactivate bacteria.
In previous studies, a semiconductor had been utilized as a photocatalyst, thus inducing a series of reductive and oxidative reactions on its surface (Dunlop et al. 2002). When TiO2 is irradiated with light with a wavelength that is equal or greater than its band gap, electromagnetic radiation is absorbed and electrons are promoted from the valence band to the conduction band. The conduction band electron (e−cb) reduces oxygen to O2• and the generated holes in the valence band (h+vb) create hydroxyl radical and H2O2, hence removing electrons from absorbed oxidizable species or reacting with OH− or H2O to form more hydroxyl radical (•OH). However, the absence of electron scavengers allows the recombination of photoexcited electron with the valence band hole in nanoseconds simultaneously with heat dissipation. An electron scavenger is vital for successful functioning of photocatalysis (Cho et al. 2004; Chong et al. 2010).
The recommended approach to prevent recombination of electron/hole pairs is to apply an electric field enhancement as an electron scavenger. This method, named photoelectrochemical treatment, consists of an electric field that removes conduction band electrons from a TiO2 semiconductor to a counter electrode. As a result, photochemical efficiency for the generation of holes is increased.
Highly reactive hydroxyl radical (•OH) serves as the main oxidant, thus capable of inactivating a wide range of microorganisms, including Gram-negative and Gram-positive bacteria (Paspaltsis et al. 2006). Hydroxyl radicals have a high oxidation potential, second only to fluorine. They are much more reactive than chlorine, and are thus proposed as an alternative oxidant for disinfection.
In this study, we propose a novel comparison between Gram-positive and Gram-negative microorganisms during photocatalysis and photoelectrochemical treatment. An immobilized TiO2 photocatalyst was used to disinfect water. Previous studies with powdered TiO2 were extremely challenging, since powder removal from disinfected water requires additional post-treatment steps. Therefore using an immobilized TiO2 in the form of a film can solve such a shortcoming and promote higher efficiency in water disinfection processes.
MATERIALS AND METHODS
Escherichia coli and Staphylococcus aureus growth
E. coli CCT 1457 and S. aureus ATCC 9144 cells were grown under aerobic conditions at 28–30 °C overnight in 10.0 mL of nutrient broth (tryptone 5.0 g/L; yeast extract 2.5 g/L; glucose 1 g/L). The nutrient broth was centrifuged at 400 g for 10 min and the bacterial pellet was resuspended in 5.0 mL of Na2SO4 (200 mg/L). 1.0 mL of this suspension was added to the reactor containing 250 mL of Na2SO4.
Photocatalytic treatment was conducted with 0.5 g of TiO2 (VETEC) added to 250 mL of Na2SO4 into a reactor. Each bacterial species was added separately to the reactor. Aliquots were removed at intervals of 10 min and subject to plate count agar (PCA) growth. After overnight incubation at 35 °C, the colonies were counted and calculated regarding survival.
The batch reactor for the photoelectrochemical treatment consisted of a Pyrex glass tube 10.0 cm in diameter into which was immersed a 5.0 cm × 5.0 cm Ti/TiO2 electrode, nickel mesh counter electrode (8.5 cm in diameter) and Ag/AgCl reference electrode. The 1.4 V potential applied in the TiO2 electrode was controlled by a potentiostat (Figure 1).
The thermal electrodes were designed by heating a titanium metal film for 10 min at 400 °C in a furnace. The oxidized films were removed from the furnace in slow exhaustion and cooled down. According to Christensen et al. (2003), the performance of thermal electrodes depends on the treatment temperature. They identified a peak of UV-induced photocurrent temperatures in the range of 700–750 °C. In our experiments, a standard treatment temperature of 750 °C was used. This temperature ensures an oxide film constituted in rutile form (Gemelli & Camargo 2007; Lopes et al. 2012).
The photochemical effects of light sources with different wavelength ranges have a significant consequence on the photocatalytic reaction rate, depending on the types of photocatalysts. The crystalline ratio of anatase, at a wavelength of λ < 380 nm is sufficient for photonic activation (Herrmann 1999). The crystalline phase of rutile TiO2 can be activated with wavelengths of up to 400 nm, depending on the bandgap threshold for the type of rutile TiO2 used. In this study, both the experiments were irradiated by 2 × 8 W UV-A lamps (model F8T5/BLB, λ = 300–450 nm) placed centrally on top of the solution at a 10 cm distance from the liquid surface.
Preparation of samples for scanning electron microscopy
Aggregates in TiO2 were analyzed using a Zeiss DSM 940-A scanning electron microscope. The characterization of the crystalline structures of the TiO2 was obtained using a Siemens D5000 X-ray diffractometer, coupled to a texture goniometer and 40 kV, 30 mA copper tube.
Detection of bacterial cells after treatment
Samples of the bacterial suspension were collected from photocatalytic experiments at different intervals during 60 min. Serial dilutions were made in saline solution and 100 μL of each dilution was spread onto PCA, grown at 30 °C for 24 h. The number of colony forming units (cfu) determined the viable cell in the water sample after each exposure time. Three replicate plates were used at each sampling time.
RESULTS AND DISCUSSION
Characterization of the TiO2 particles
The surface of the TiO2 powder in suspension was observed using scanning electron microscopy (SEM) images. As shown in Figure 2, the suspension of TiO2 is constituted by aggregates of particles. The photocatalytic efficiency of TiO2 can change due to formation of aggregates that decrease external surface area.
The TiO2 solution used in this study presented different aggregate size, which affects the adsorption of E. coli to TiO2. Benabbou et al. (2007) have suggested that a TiO2 monolayer could be formed, completely covering the bacteria. Gogniat et al. (2006) showed that adsorption of bacteria on TiO2 aggregates is essential for the bactericidal effect of photocatalysis.
The TiO2 difractogram shows that TiO2 is present in the anatase phase only, as presented in Figure 3. This constitution is different from the P25 Degussa (80% anatase and 20% rutile) employed in many photocatalytic studies. The anatase phase is well known for its catalysis effectiveness.
Photocatalytic and photoelectrochemical disinfection of
The photocatalyst treatment inactivated all E. coli cells in 60 min, as shown in Figure 4. The UV irradiation did not cause inactivation of E. coli in the absence of TiO2, as well as when TiO2 was applied in the absence of UV irradiation. The OH radical has been considered the major cause for E. coli inactivation because the generated OH radical can attack bacterial cell walls (Cho et al. 2004). Also, the photogenerated hydroxyl groups from TiO2 lead to the photodecomposition of amino acids by the interaction with the carboxyl and amino groups (Tran et al. 2006).
The E. coli cells generate H2O2 continuously in aerobic growth, but the intracellular H2O2 is kept at low concentrations by the action of cellular defense mechanisms such as antioxidant enzymes. Nevertheless, when the E. coli cells enter environments where the extracellular H2O2 concentration exceeds 0.2 μmol/L, there is an influx of H2O2 towards E. coli (González-Flecha & Demple 1995; Cabiscol et al. 2000; Imlay 2013). Therefore, the H2O2 formed during the photocatalysis may cross membranes with ease, causing an oxidative stress, which can contribute to bacterial inactivation. Another mechanism of inactivation caused by H2O2 is peroxidation of membrane lipids, which promotes high disorder in the E. coli membrane (Maness et al. 1999; Pigeot-Rémy et al. 2012; Nie et al. 2014).
The effect of the photoelectrochemical treatment in E. coli was more efficient when the 1.4 V potential was applied to the UV-irradiated electrode. The absence of potential allows the electron/hole recombination process, decreasing the rate of inactivation, as shown in Figure 5.
The photocatalyst treatment killed all E. coli cells in 60 min, while photoelectrochemical treatment killed around 50% of cells in 60 min.
Compared with nano-metal Ag, which is commonly applied as antimicrobial particles, TiO2 is more cost effective. Titanium oxide presents higher efficiency (around 1 h) to exhibit antibacterial effects whereas Ag particles take 24 h to display their antimicrobial effects (Sondi & Salopek-Sondi 2004). Since TiO2 acts as a photocatalyst during disinfection processes, unlike other agents, the antibacterial effects of TiO2 will most likely not be decreased with cell inactivation and the consumption of itself.
Photocatalytic and photoelectrochemical disinfection of
The initial concentration of bacteria (106 cfu/mL) decreased about 50% in 60 min after UV light and TiO2 power systems treatment, while non-significant inactivation was observed when using only UV light or TiO2, as indicated in Figure 6.
The results showed that the photoelectrochemical inactivation was effective for S. aureus when applied potential showed a significant disinfection rate (Figure 7). Most cells were completely killed by the photoelectrochemical treatment in 20 min. According to Chung et al. (2009), S. aureus is killed by the detachment of the cell wall from the cell membrane, while E. coli suffers from changes in nucleoid pattern.
Inactivation of Gram-positive and Gram-negative species by photocatalytic and photoelectrochemical treatment
The comparison between inactivation rates of Gram-positive and Gram-negative species showed that the two strains are affected in different ways regarding photocatalyst and photoelectrochemical treatment, probably owing to different structure membrane and the adsorption to catalyst.
Studies also showed that Gram-negative pathogen was more sensitive to photocatalytic disinfection than Gram-positive pathogen (Backhaus et al. 2010). The variations are due to differences in cell surface structures (cell wall and cell membrane) between Gram-negative bacteria and Gram-positive bacteria. The cell walls of Gram-negative bacteria are thinner (around 12 nm) but triple-layered, as there is an inner membrane, a thin peptidoglycan layer, and an outer membrane. Gram-positive bacteria have a much thicker peptidoglycan layer (around 60 nm), but no outer membrane (Foster et al. 2011). Conversely, some studies report Gram-negative bacteria to be more resistant than Gram-positive bacteria to the antibacterial effects of TiO2 (Kangwansupamonkon et al. 2009).
Wang et al. (2013) also observed that the relation between the ratio of inactivating bacteria or actual antimicrobial activity by TiO2 depends on the initial bacterial concentration and bacterial population during inactivation. Such dependency occurs because when the initial bacteria population is too high, many dead cells and mineralization products are generated during the reaction process. These intermediate products of photocatalysis have the function of protecting live bacteria, and the factor likely to be contributing to the loss of antibacterial activity is competition between these intermediates and bacteria for •OH radicals. Hence, the concentration of radical •OH in the bacteria surface is decreased. When the initial bacteria concentrations are low, the protection and competition are significantly reduced, resulting in major disinfecting effectiveness (Rincón & Pulgarin 2004).
First, an efficient inactivation of the bacterial cells occurs when there is close contact between the microorganisms and the TiO2 catalyst (Benabbou et al. 2007; Foster et al. 2011). Thus, the study presents two situations. When the TiO2 power is used, it is adsorbed because the bacteria are bigger than TiO2 particles. Conversely, when the thermal electrode is used, the electrode surface adsorbs the bacteria.
In the photocatalytic treatment, the TiO2 power wraps E. coli cells interacting with electron/holes and its reactive species. In the photoelectrochemical treatment, electrostatic interaction ensures the contact of the electrode surface (positively charged) with the outer membrane (negatively charged). However, the low surface area of the electrode (10 cm2) leads to a decrease of the inactivation rate, mainly owing to the mass transfer problems. Moreover, it is known that phosphates adsorb strongly on TiO2 surface (Christensen et al. 2003; Selcuk 2010), suggesting that the phosphates, constituents of the outer membrane, might have contributed to delay in the E. coli inactivation rates. This hypothesis became clearer when compared with the findings of the photoelectrochemical treatment to S. aureus, which showed efficient inactivation in 20 minutes. S. aureus cells also can be adsorbed to the electrode surface by electrostatic interaction electrode/membrane, but the wall cell of S. aureus (Gram positive) is composed of a peptidoglycan thick layer, which may be affected by reactive species, causing damage to membrane permeability.
The photocatalysis and photoelectrochemical treatment effectively inactivated E. coli and S. aureus bacterial microorganisms. Photoelectrochemical treatment was more efficient than photocatalysis in S. aureus due to adsorption issues in the electrode surface when a potential was applied; whereas photocatalysis was more efficient than photoelectrochemical treatment in E. coli due to mass transfer limitations in the latter case. Both treatments were ineffective without UV-light. The proposed technology proved itself prone to practical applications on a laboratory scale. Further studies on scaled-up experiments can potentially provide cost-effective water treatment technologies based on the established experiments.