• Influence of furosemide initial concentration and pH on HC/ED degradation tests
To investigate the efficiency of the prototype reactor in the degradation of various concentrations of FUR, the first set of experiment were conducted with different initial concentrations (c0) of analyte (10, 30 and 50 mg/L) working at 20 bar of inlet pressure and 48 kHz of ED plasma frequency, both in flow and loop configurations (Fig. 3). Considering the high drug concentration, the pH of tap water (7.6) was corrected to 9 with a 10 M solution of NaOH to facilitate the FUR solubility. The pH variation during the treatments was also measured. In respect to the test performed with a starting FUR concentration of 10 mg/L, a complete degradation was observed after only 5 min of treatment in loop configuration, while the flow mode allowed to observe a degradation rate of 64%. For the sake of comparison, the same amount of FUR was also subjected to a degradation treatment under the effect of HC alone, without the contribute of ED plasma. During the non-hybrid treatment (HC alone), a 32% of degradation rate was observed in flow mode and a slight increase up to 43% and 46% was observed after 5 and 10 min of treatment, respectively. The flow mode and 5 min loop treatment allowed to obtain, respectively, a 57% and 98% degradation rate values during the treatment conducted on the 30 mg/L FUR solution, demonstrating the same reactor efficiency observed during the previous tests. The increase of initial concentration up to 50 mg/L slightly affected the degradation in flow (40%) but a near quantitative degradation of 95% was observed again after 5 min of treatment. Overall, a treatment time of 10 min ensured a complete degradation of FUR for each starting concentration value. In respect to the monitored pH during treatments (Fig. 3b), it was overall observed that as the initial concentration of FUR increased, the pH gradually decreased as a direct result of the greater amount of substrate available to be converted to organic acids under the extreme oxidative environment generated by hybrid HC/ED.
The observed results suggested that the pilot scale hybrid reactor maintained its efficiency even at very high starting FUR concentration values, which were chosen specifically to demonstrate the capabilities of the hybrid pilot scale reactor, even under the flow mode. The showed results demonstrate the actual chance of using the hybrid HC/ED technology to treat FUR-polluted water at industrial level in flow, considering that the FUR concentrations generally observed in hospital effluent or rivers/streams are in the range of µg/L 11, which is significantly lower than that used in the previously described experiments. However, it should be considered that the treatment of a real effluent could be affected or limited by the presence of additional contaminants or substances that can quench the radical oxidation.
• Influence of scavenger addition on HC/ED degradation tests of furosemide
To further test the efficiency of the HC/ED pilot scale reactor in presence of •OH scavengers or other pollutants, and to investigate the possible contribution of pyrolysis reactions in the degradation of contaminants, different degradation treatments were performed in presence of alcohols or an additional drug. As already demonstrated in a previous work [ref], the concentration of oxidizing compounds (UV light, H2O2, O3 and •OH) inside the recirculating wastewater can be enhanced by to the combination of the ED plasma with HC. Nevertheless, alcohols such as ethanol (EtOH) can scavenge •OH 38 Eq. (1) and a competition between the ED plasma-generated radicals and organic contaminants can take place, limiting their degradation rates.
$$C{H}_{3}C{H}_{2}OH+ {\bullet }\text{O}\text{H}\to {\bullet }\text{C}{H}_{2}C{H}_{2}OH+ {H}_{2}O$$
1
The relatively high EtOH volatility (vapor pressure = 55 mmHg) could facilitate its diffusion into cavitation bubbles during their generation, extinguishing the oxidative radical reactions that occurs at the gas(bubble)-liquid (bulk solution) interface. However, since there is no experimental evidence of EtOH diffusion into cavitation bubble, a 10 mg/L FUR solution was treated in presence of EtOH (3.5·10− 2 M and 7.0·10− 2 M) with an inlet pressure of 20 bar and an overall decrease in the degradation rate was observed for both flow and 5 min treatments (Fig. 4a).
In detail, during the treatment performed with a 3.5·10− 2 M concentration of EtOH the degradation rate decreased by a value of 56% in flow mode and a value of 24% after 5 min, while the complete degradation was observed after 10 min of treatments. A scavenger concentration of 7.0·10− 2 M caused degradation rate reductions by values of 88% and 39% for the flow and 5 min treatments, respectively, with an 86% degradation rate after 10 min. However, despite the presence of scavengers, 5 min of treatment were sufficient to achieve higher degradation than observed in cavitation alone (43%). In respect to the pH variation of the water matrix, during the quenched experiments conducted with EtOH concentration values of 3.5·10− 2 M and 7.0·10− 2 M, a pH decreases to 7.2 and 6.8, respectively, was documented (Fig. 4b). On the contrary, during the unquenched test the minimum reached pH value was 8.6. The significant decrease in pH may be ascribe to the formation of acetic acid derived from a partial oxidation of EtOH via a two-step reaction: a first radical EtOH oxidation towards acetaldehyde by •OH attack 39 (Fig. 4c) followed by a further oxidation step of acetaldehyde supported by the extreme oxidant HC/ED environment to generate acetic acid. Two additional treatments were performed on tap water alone (pH = 9) in presence of EtOH and the final pH values after 10 min were 7.8 and 7.4 with starting EtOH concentrations of 3.5·10− 2 M and 7.0·10− 2 M, respectively. These results, in combination with an observed constant pH during a further test performed with only tap water (pH = 9), could confirm the contribute of EtOH oxidation previously hypothesized. The results of the experiments showed that EtOH acted as a •OH quencher during the treatments, limiting the radical oxidation mechanism. In addition, during the described treatments, a switch of ED plasma colour from the benchmark pink-purple (Fig. 4d) to a pale pink-blue-white (Fig. 4e) was observed. Under unperturbed conditions, both the HC and HC/ED plasma-generated •OH reach their excited state due to the energy provided by the electrical discharge (ED). Their subsequent relaxation determines the emission of UV light with a wavelength in the range of 287–309 nm (A2Σ+(v = 0, 1) → X2Π(v = 0))40, justifying the bright pink-purple plasma colour. In presence of EtOH, the decrease of •OH concentration also determines a decrease in the UV light emission, with a reduction of an additional source of oxidation. Despite the presence of EtOH, a partial degradation of the analyte has nevertheless been observed, probably due to the presence of HC/ED-generated H2O2 and O3 or, possibly, to the contribute of pyrolysis reaction which can occur during cavitational treatments 41,42 in the core of cavitation bubbles (Fig. 5a). To better underline a plausible contribution of pyrolysis reaction in the FUR degradation, additional tests were carried out in presence of tert-butyl alcohol (t-BuOH), as it has been demonstrated that it can diffuse into the cavitation bubbles during their generation and subsequent growth due to its volatility (vapor pressure = 46 mmHg). This phenomenon allows t-BuOH to exert its quenching effect, acting as a •OH scavenger both in the gas phase and in the gas-liquid interface of the cavitation bubble 43,44 as shown in Fig. 5a. Considering the colour variation of the emitted light observed during the cavitational treatments in presence of EtOH (Fig. 4d and 4e), a preliminary test was conducted in water with t-BuOH (3.5·10− 2 M) to investigate possible differences in the quenching activities of the two alcohols. In contrast of what observed in presence of EtOH, the addition of t-BuOH determined a variation of the benchmark light towards a more intense blue-white light (Fig. 5b). The extensive decrease of pink light could be due to the higher rate constant reaction of t-BuOH with •OH (1.08·1012 cm3 molecule− 1 s− 1) 45 than that of EtOH (0.4·1012 cm3 molecule− 1 s− 1)46, probably leading to a greater contribute of both the Hβ radical emitted light at 490 nm (blue region of the visible spectrum), determined in a previous work 25, and typical N2 emission between 400 and 440 nm (N2(C–B) transitions) observed during the generation of glow plasma in presence of water28. However, the complete understanding of the physicochemical phenomena that occurred during the described tests is difficult and requires further specific investigations.
Therefore, a degradation test on a 10 mg/L FUR solution was performed in the presence of t-BuOH (3.5·10− 2 M) at 20 bar. Overall, a decrease of FUR degradation rate was observed compared to the treatments in presence of EtOH. In detail, the degradation rate with t-BuOH in flow mode (29%) was similar to that observed in presence of EtOH (28%), but a 17% decrease in the extent of degradation was observed for both 5 and 10 min of loop treatment, respectively, in respect to the test performed with EtOH (Fig. 6a). Despite the results could demonstrate the higher efficiency of t-BuOH in the •OH radicals quenching than EtOH, foaming was observed inside the reaction chamber after 1 minute of loop treatment (Figure S6), probably due to the microscopic phase separation at cluster level (i.e coexistence of water-rich clusters and organic cosolvent-rich clusters) generated during acoustic cavitational treatments carried out in presence of a binary water-organic solvent mixture47. Under acoustic cavitation, water molecules promote self-association of organic molecules as a balance of interactions controlling the microscopic structure in the solution. The presence of foam caused a reduction of the cavitation intensity (reduction of cavitation plume length along the reaction chamber) and the consequent reduction of the ED plasma extension as a direct consequence, contrary to what observed during the flow mode treatment and the preliminary test with t-BuOH previously described (Fig. 5b). Due to this, to confirm the faster reaction between t-BuOH and •OH (k = 1.08·1012 cm3 molecule− 1 s− 1) than EtOH (0.4·1012 cm3 molecule− 1 s− 1), an additional treatment was performed lowering the inlet pressure to 15 bar to reduce foaming and to avoid possible misleading in the interpretation of the data. The results obtained at 15 bar were not significantly different to the degradation rates observed at 20 bar (Fig. 6a), confirming both the contribute of only t-BuOH scavenging (and not to foaming) in decreasing degradation efficiency at 20 bar, and the faster t-BuOH quenching reactivity than EtOH. In respect to the final pH values reached at the end of the performed treatments (Fig. 6b), in presence of t-BuOH were observed slight pH decreases (8.1 at 20 bar and 8.3 at 15 bar) with respect to the unquenched treatment (8.6). However, the addition of t-BuOH may lead to the formation of different by-products compared to benchmark experiment, thereby explaining the slight further decrease in pH. In addition, considering that t-BuOH is resistant to oxidation due to its tertiary alcohol structure, the conclusive pH values from tests conducted with this scavenger did not exhibit a significant decrease, in contrast to the observed trend in the presence of EtOH.
Overall, the results obtained in the presence of radical scavengers show that the mechanism of radical oxidation at the interface between gas and liquid cavitation bubbles during HC/ED treatments is the major contributor. However, considering the high degradation rate observed despite the presence of EtOH and t-BuOH, a possible contribution of the pyrolysis reaction cannot be excluded. In this regard, further studies should be conducted in the future to gain a deeper understanding of the hybrid HC/ED innovative technology.
In view of a real industrial application of the hybrid HC/ED technology, a further experiment was conducted also in presence of another antibiotic (the metronidazole (MNZ)) to assess the efficiency of the treatment in the simultaneous degradation of two different pollutant APIs. For this purpose, a 5 L solution with a total drug loading of 20 mg/L (10 mg/L of FUR + 10 mg/L of MNZ) was treated at 20 bars for a maximum treatment time of 10 min (Fig. 7a). The treatment performed in flow mode allowed to observe a FUR and a MNZ degradation rates of 52% and 43%, respectively, while the 5 min loop test guaranteed a quantitative FUR degradation (99%) and a 93% MNZ abatement. After 10 min both the diuretic (FUR) and the antibiotic (MNZ) were completely degraded, demonstrating the maintenance of the HC/ED reactor efficiency even in the case of treatments conducted in the presence of two different contaminants for prolonged treatment time. However, the addition of MNZ slightly affected the FUR abatement during the flow experiment; indeed, the FUR degradation rate decreased from 64%, obtained in the treatment without MNZ, to 52%.
Anyway, considering the very high initial drugs loading (20 mg/L) which deviates significantly from the actual concentrations of the two drugs generally found in several effluents (µg/L), the slight decrease in FUR abatement observed during the flow treatment can be considered insignificant in the case of treatments performed on real drug-polluted wastewater. In addition to the scavenging effect exerted by organic substances such as alcohols or other drug pollutants, certain inorganic substances commonly present in wastewater and/or in drinking water can also quench •OH. Among the common inorganic ions present in tap water (e.g. Ca2+, Mg2+, Cl−, SO42−, HCO3− etc), variation of bicarbonate concentration could induce scavenging effect because of the competitive reactions of carbonate, bicarbonate, and FUR with •OH (Eqs. 2 and 3) 33.
$$C{O}_{3}^{2-}+ \bullet OH\to \bullet C{O}_{3}^{-}+O{H}^{-}$$
2
$$HC{O}_{3}^{-}+ \bullet OH\to \bullet C{O}_{3}^{- }+ {H}_{2}O$$
3
Considering the variable average concentration of HCO3− in the range of 200–500 mg/L observed in tap waters analysed in different Italian cities (Table S2), a degradation treatment was performed with the addition of bicarbonate to investigate the reproducibility of the degradation process as a function of geographic water composition in case of drinking water treatments. The starting HCO3− concentration of the available tap water (257 mg/L Turin, Italy) was increased up to 402 mg/L (2.4·10− 3 M) but no scavenger effect has been observed during the treatments due to both low HCO3− concentration and low rate constant (8.5·109 cm3 mol− 1s− 1)48 of the reaction of bicarbonate ion with hydroxyl radicals.
• UV and UV/H2O2 degradation tests for furosemide
In order to compare the efficiency of the pilot scale HC/ED prototype reactor with other AOPs, additional lab-scale batch treatments were performed on a 0.2 L solution of FUR with a starting concentration of 10 mg/L and a pH value of 9, under UV irradiation alone (390 nm, average intensity of 137 mW/cm2) or under combined UV/H2O2 (1:100 FUR:H2O2 molar ratio). A preliminary prolonged degradation test of 3 hours was conducted under UV irradiation only, which allowed to reduce the starting antibiotic concentration by a value of 88%. Moreover, in contrast to what observed during the HC/ED treatments (Figure S7), the HPLC analysis clearly revealed the formation of a by-products (Figure S8) during the UV degradation test, probably due to the different degradation pathways of UV and HC/ED treatments or to the incapability of UV light in the degradation of these compouds. Subsequently, UV-irradiated solutions were treated for 5 and 10 min to directly compare the results obtained with the pilot scale HC/ED reactor. Those experiments revealed a FUR degradation rate of 11% and 18% after 5 and 10 min of treatment time, respectively, confirming the suffering of FUR to photochemical degradation 49,50. The addition of H2O2 allowed an increase of the UV alone degradation rates up to 22% and 35% due to the additional •OH radicals generated by the UV-induced homolytic bond cleavage of H2O2. As shown in Fig. 8a and 8b, the degradation rates observed for both UV and combined UV/H2O2 lab-scale treatments conducted for 5 and 10 min were lower than the analogous treatments (HC, HC/ED, HC/ED + EtOH and HC/ED + t-BuOH) performed at pilot scale, overall. In addition, further UV and UV/H2O2 treatments were tried on a 0.5 L FUR solution with a starting concentration of 10 mg/L for a total time of 5 min. The 2.5-fold scale up of lab scale treatments exceeded the effective operating limit, as only a 1% degradation was observed in both the case of UV alone and in the combined process with H2O2. In detail, the lab-scale findings demonstrated even lower treatment efficiency compared to the pilot-scale operation with the use of HC as the sole treatment agent (43% of degradation rate), indicating the reactor's remarkable effectiveness even when employing a non-hybrid approach. The comparison of the three technologies revealed how the hybrid HC/ED can be considered a promising technology for its scalability at industrial level considering the higher degradation rates in treating large volumes of polluted wastewater, even in a non-hybrid approach (HC) and in presence of radical scavenger compounds.