3.2.1. Photocatalytic activity regarding TCE
The results obtained from the evaluation of photocatalytic activity regarding TCE are listed in Table 2. Regarding TCE, Ag-N-TiO2 presented a lower activity when compared to the other photocatalysts, with a photocatalytic efficiency of 31.47% in UVA and no activity in visible light. Elsellami et al. (2017) researched TiO2 doping with Ag, verifying that an increase in this metal’s percentage diminished the photocatalytic efficiency because of an increased probability of electron-hole pair recombination. Similarly, the quantity of Ag in Ag-N-TiO2 may have led to this low efficiency.
Table 2
Photocatalyst efficiency in degrading 25 ppm TCE at a flow rate of 300 mL.min-1 under continuous flow
Photocatalyst | Radiation | Efficiency (%) | CO2 Selectivity (%) |
TiO2 PC 500 | UVA | 91.35 | 70.86 |
Visible | 40.57 | 28.57 |
Visible with filer | 0.72 | 0 |
N-TiO2 | UVA | 93.20 | 59.70 |
Visible | 22.10 | 65.56 |
Visible with filer | 5.86 | 118.20* |
Cu2O/N-TiO2 | UVA | 92.06 | 57.39 |
Visible | 36.46 | 9.32 |
Visible with filter | 12.64 | 13.05 |
*With low efficiency, the CO2 value produced is low and is mistakable with the baseline, which is already higher than previous experiments.
TiO2 PC500 achieved a 91.35% TCE conversion with UVA irradiation, with a CO2 selectivity of 70.86%, which agrees with past literature (Mohseni 2005). Moehseni (2005) had already demonstrated that TCE does not suffer photolysis with UVA but does with UVC radiation, which attributes the obtained results to heterogeneous photocatalysis. A TCE conversion under visible light was also observed, probably associated with the UVA in fluorescent lights (De Souza Lourenço et al., 2014). When a filter capable of removing emissions under 390 nm was used, there was no significant TCE conversion since the energy provided was insufficient to generate active sites and oxidant species.
The N-TiO2 photocatalyst had a TCE conversion close to TiO2 PC500, when irradiated with UVA, and slightly less when irradiated with visible light. The differential within the nitrogen material was that there was a higher CO2 selectivity within the visible spectrum, indicating lesser formations of more toxic chlorinated products. Even still, when utilizing visible light with a filter, N-TiO2 kept converting TCE, being associated to a new material band gap of 2.36 eV. Mitoraj and Kisch (2008) obtained a high efficiency with visible light regarding formic acid for N-TiO2 (80% degradation) (Mitoraj and Kisch 2008). However, they ran their tests in the liquid phase, where the photocatalyst was suspended. The current paper evaluated experiments in the gaseous phase, in which the photocatalyst was supported. However, in the absence of humidity, this may justify low efficiency and mineralization results obtained compared to the literature as mentioned above.
The Cu2O/N-TiO2 photocatalyst under UVA managed to achieve a TCE conversion similar to that of TiO2; however, under visible light with a filter, the material was more efficient than TiO2 PC500 and N-TiO2, which may be associated with the low band gap value (1.38 eV) and the n-p heterojunction between Cu2O and TiO2. Janczarek et al. (2018) had already verified a significant increase in TiO2 activity when this semiconductor was simultaneously modified with Cu2O and urea. However, in this paper, the target compound utilized 2-propanol and the efficiency was verified regarding acetone generation, which does not characterize full mineralization but a subproduct conversion.
Thus, evaluating only TCE conversion, the experiments were valid in that they compared distinct materials and the results obtained are favorable towards photocatalytic activity within the visible light spectrum.
3.2.2. Photocatalytic activity regarding a biological contaminant
Table 3 presents efficiencies calculated for each time there was a decrease in bacterial concentration in relation to the initial concentration in the presence of photocatalysts.
Table 3
Photocatalyst efficiency in eliminating E. coli.
Photocatalyst | Radiation | Ci E. coli (CFU/mL) | E(1h) (%) | E(2h) (%) | E(3h) (%) | E(4h) (%) | E(5h) (%) |
Photolysis | UVA | 1.25 x 106 | * | 40 | 97 | 99 | 99 |
Photolysis | Vis | 9.33 x 105 | 12 | * | * | * | * |
TiO2 PC500 | UVA | 1.68 x 106 | 17 | 46 | 79 | 98 | |
Vis | 1.15 x 106 | * | * | * | 10 | 39 |
N-TiO2 | UVA | 5.50 x 105 | * | 59 | 97 | 99 | 99 |
Vis | 6.00 x 105 | * | * | * | 54 | 72 |
Cu2O/N-TiO2 | UVA | 1.83 x 106 | 27 | 49 | 93 | 99 | 100 |
Vis | 6.50 x 105 | 72 | 96 | 99 | 99 | 99 |
| | | E(0.25h) (%) | E(0.5h) (%) | E(0.75h) (%) | E(1h) (%) |
Ag-N-TiO2 | UVA | 1.32 x 106 | 25 | 38 | 70 | 100 |
Vis | 1.27 x 104 | 95 | 99 | 100 | 100 |
*There was an increase in E. coli concentration. Observation: 100% means total inactivation within the detection method threshold.
UVA photolysis inhibited bacterial growth after 180 min of 97% irradiation, which agrees with the literature (Bono et al. 2021; Foster et al. 2011; Pizarro and Orce 1988; Sunada et al. 1998). Visible light, however, did not cause any inhibiting effects on bacterial growth, which was already expected because this microorganism is found in the air (Choi et al. 2020; Sánchez et al. 2012; Wu et al. 2020).
TiO2 experiments confirmed that UVA-irradiated material showed photocatalytic activity regarding E. coli (Carp et al. 2004; Rojviroon and Sirivithayapakorn 2018), whilst visible light did not affect bacterial growth inhibition up to 180 min, which is in agreement with the material’s band gap value of 307 eV – UVA energy.
UVA tests with the N-TiO2 photocatalyst intended to maintain the initial E. coli concentration within the same order of magnitude as the other experiments (106 CFU/mL); however, an initial concentration of 5.50 x 105 CFU/mL was obtained on account of the variation between growth processes and bacterial separation. Thus, directly comparing these results with the ones obtained for TiO2 under UVA or for photolysis may induce a false sense of improved efficiency since literature has shown that the increase in the initial E. coli concentration decreased TiO2 activities (Rojviroon and Sirivithayapakorn 2018). Even still, N-TiO2 showed high bacterial inhibition (99%, t = 300 min) compared with the same experiment in the dark (58% after t = 300 min). Within the visible light spectrum, the photocatalyst led to some bacterial growth inhibition (73%, t = 300 min), which was already expected from the calculated band gaps. Nevertheless, bacterial growth was observed in the initial moments, which may lead to the hypothesis that bacteria may use nitrogen as a nutrient source for their metabolic reactions (Peterson et al. 2005), leading to a competition between this effect and heterogeneous photocatalysis.
By the end of the Cu2O/N-TiO2 UVA radiation experiment, a bacterial inhibition rate of 100% was observed, indicating a synergistic effect between photocatalysis and copper’s bactericidal effects (Maisetta et al. 2010; Valeria Prado et al. 2012). Bacterial concentration in relation to initial concentration always decreased, both in the dark and under radiation. Up to the second hour, the same relative activity was reported with and without irradiation. However, bacterial concentration declines faster from the second hour onwards when under photocatalyst irradiation. A plausible hypothesis is that in the dark, despite copper’s bactericidal effects, bacteria utilize nitrogen within the photocatalyst surface as a source of nutrients (Peterson et al. 2005), making it so the experiment’s concentration between 180–300 min remains almost constant. However, in the presence of light, this effect is overcome by the material’s photocatalytic activities, with an efficiency of 27% within the first hour and up to 93% in 3 hours, reaching 100% inactivity after 300 min. The material was more effective under the visible light spectrum, with 96% of bacterial inhibition in 120 min and 99% by the end of the experiment. High photocatalytic activity in the visible light spectrum may be directly related to its band gap value of 1.73 eV. Janczareket al. (2018) had already verified good photocatalytic activity with visible light using isopropanol as a chemical contaminant. Despite not dealing with bacterial activities, they propose two photocatalytic mechanisms for visible light which justify improvements in regards to UVA and which may be occurring in the current research: the first is the mechanism based on the n-p heterojunction between Cu2O and TiO2 sensitized by polytriazenes. Within this mechanism, the probability of electron-hole pairing would decrease on account of the fact photo-stimulated electrons would transfer between the Cu2O conduction band to the sensitized TiO2 conduction band, and holes would be transferred between the TiO2 valence band and the Cu2O valence band. As the lifetime of active sites is augmented, such is the improvement of the photo-degradation process. The second mechanism is based on the Z-scheme, in which photo-generated electrons with lower modified TiO2 reduction capacities recombine with holes with lower Cu2O accumulated oxidation capacities. Thus, electron/hole pairs with higher redox capacities may be kept, improving the photo-degradation process.
Finishing photocatalytic research, Ag-N-TiO2 was the material which provided 100% bacterial inhibition within the shortest timeframe. Due to rapid bactericidal activity, experiments were repeated with smaller sample intervals (15 min). Silver has been approached as a TiO2 dopant because it acts as an electron trap and improves the semiconductor’s photodegradation process (Ismael 2020; Sacco and Vaiano 2020). Silver is also widely known for its bactericidal activity due to Ag+ ions, which permeate the bacterial membrane, as well as the damage caused to surrounding tissues and bacterial death (Flores-Lopez et al. 2021; López-Heras et al. 2015). Therefore, a synergistic effect between silver’s bactericide activity and heterogeneous photocatalysis was expected. Despite the high efficiency within a shorter timeframe, in the experiment conducted with UVA irradiation it was observed that the material proved to be more effective in the dark than when under radiation. Such data indicates that the material itself was already toxic to bacteria, yet somehow UVA radiation in the photocatalytic process was not as efficient as previously expected. On the other hand, photolysis experiments showed a notorious bacterial growth within the 1 h of irradiation. Therefore, in the Ag-N-TiO2, experiment conducted up to 1 hour of sampling, bacterial inhibition is not associated with photolysis. When only N-TiO2 was used, similar bacterial growth was observed; therefore it is possible that UVA radiation somehow modified Ag-N-TiO2, hindering silver’s natural bactericidal activities or even competing with them. In order to examine this hypothesis, the material was suspended in deionized water and remained under UVA contact at during a time period equal to the bactericidal test, then dried and put under XPS analysis. It was possible to observe that there appeared an additional titanium peak at 456.9 eV (regarding Ti2O3) and a silver one at 367.9 eV (regarding the Ag-O-Ti on the TiO2 surface) (Fig. 3S-d) (Atla et al. 2012). Possibly, when in the presence of radiation, Ag2O Ag+ ions react to TiO2 superficial hydroxyls, modifying Ti4+ sites into Ti3+ in the form of Ti2O3 (Atla et al. 2012; Samsuddin et al. 2017). Thus, in the experiment with bacteria, the photocatalyst simultaneously inhibits bacterial growth and modifies Ti and Ag structurally. Even with no irradiation, the material itself can promote 100% inhibition in 1h of contact, which is as effective as it gets when under irradiation. The experiment with visible light may corroborate the hypothesis that UVA radiation is modifying the material with silver. In the experiment irradiated with visible light, bactericidal efficiency was higher than in the system without any radiation at all. Even if initial bacterial concentrations differ between the UVA and visible light experiments, each experiment was always conducted with a control without irradiation, with the same initial concentration, allowing a direct comparison between activities with or without the irradiation of the photocatalyst under visible light spectrum. Within the first 15 minutes of contact with visible light, the photocatalyst presented an efficiency of 95%, whilst the photocatalyst left in the dark had a bacterial inhibition efficiency of 62%. Thus, photocatalysis under visible light happened, which is justified by the material’s band gap calculation of 1.38 eV obtained from DRS analysis. With this photocatalyst, total bacterial inhibition was reached in 45 minutes for irradiated and non-irradiated material alike, which was the best efficiency obtained in the smallest timeframe within all of the tested photocatalysts.