Effective utilizations of liquid-phase plasma for remediation of non-degradable organic pollutants in water

2.1 Chemicals

Fluorene (analytical grade) was purchased from Aladdin. Commercial titanium dioxide (P25, 80% anatase, 20% rutile, BET area, ≈50 m²/g, Degussa) was used as a photocatalyst, and the iron salt FeSO₄·7H₂O (≥ 99%, Sigma–Aldrich Co. LLC.) was used for the Fenton reaction. All other organic and inorganic reagents used were of analytical grade and were used as purchased without further purification.

2.2 Water remediation

2.2.1 LPP treatment

Experiments were performed at room temperature and atmospheric pressure. Flu was dissolved in methanol to prepare a 500 mg/L stock solution because of its inferior solubility, and it was diluted to 10 ppm for all experiments. A 50 mL volume of Flu solution was treated, and plasma discharge in water solution was produced using a bipolar pulse power supply. Tungsten rods were used as electrodes and insulated by an alumina tube. The gap distance between the electrodes and electrode protrusion were set at 1 mm to concentrate the energy. The applied frequency was varied at 10, 20, and 30 kHz, and the pulse width was controlled at 1, 2, and 3 μs. The solutions were mixed using a magnetic stirrer to obtain a homogeneous plasma active field during the experiments. All experiments were conducted at least three times to ensure reproducibility, including the LPP/TiO₂ and LPP/Fe²⁺ experiments. A schematic illustration of the LPP is shown in Figure 1.

Figure 1:

Schematic illustration of experimental setup for liquid-phase plasma and image of plasma reaction field

2.3 Extraction

Samples for HPLC were prepared as follows: A 50 mL volume of the treated solution from the LPP reactor was transferred into a separatory funnel with PTFE caps. Methylene chloride (15 mL) was used to rinse the reactor and transfer the rinsed solvent to the separatory funnel. The separatory funnel was sealed and shaken for 1 min with periodic venting to release excess pressure. The mixture was allowed to equilibrate for 10 min. Subsequently, 10 milliliters of dichloromethane was separated and dried using nitrogen. The extract was dissolved in 2 mL of acetonitrile for HPLC and then filtered through PTFE filters prior to analysis. All samples were stored in brown sample vials in a refrigerator at 4 °C until analysis is performed.

For FTIR analysis, 50 mL of treated contaminated water from the reactor was transferred to a separatory funnel. Next, 10 mL of dichloromethane was added to the separatory funnel. The separatory funnel was agitated for 1 min and left to stand for 10 min. Subsequently, the dichloromethane layer was separated from the water and then dried using nitrogen gas. The extraction was repeated for three times and then combined. The extract was dissolved in dichloromethane (100 μL) and deposited on potassium bromide.

2.4 Analysis

The electrical characteristics of the LPP were monitored using an oscilloscope (Tektronix, MSO 3014) equipped with a voltage probe (Tektronix, P6015A) and a current probe (Tektronix, TCP0020). The optical characteristics of the plasma were recorded using an optical emission spectrometer (OES) (Ocean Insight, FLAME-T-XR1). HPLC (Waters, Waters e2695) was employed to measure the concentration of Flu after treatment, and a photodiode array detector (PDA) (Waters, 2998 PDA) was used to record the absorption spectra at 264 nm. A Sunfire C18 column (Waters, 4.6 mm × 250 mm × 5 μm) was used to perform separation. The mobile phase included acetonitrile as component A and water (5% acetonitrile) as component B. Isocratic elution was performed for 5 min using acetonitrile/water (4:6) (v/v), followed by linear gradient elution to 100 acetonitrile for more than 25 min at a flow rate of 1 mL min^-1. The samples were injected into the HPLC system using a microsyringe, and the injection volume was 10 μL. FTIR (Thermo Fisher Scientific, iS50) was employed to characterize the chemical bonds of the samples. The transmittance mode was used, and the spectra were obtained from 400 to 4000 cm^-1 with a resolution of 1 cm^-1. The production of hydrogen peroxide was determined using the titanium sulfate colorimetric method based on a reaction with titanyl ions in the presence of sulfuric acid, which appears yellow [15]. A UV/Vis spectrometer (UV-1900i, Shimadzu) was used to measure the absorbance of the liquid at 405 nm to determine the generated H₂O₂.

3.1 Electrical characteristics of LPP

Plasma was generated in the liquid owing to the production of bubbles, which emerged because of Joule heating and electrolysis [16]. Because the current passing through the small surface area of the electrodes generated thermal energy, the liquid between the electrodes was heated to its boiling point, and bubbles were generated. Furthermore, electrolysis occurred because of the ionic current, which generated hydrogen and oxygen gases around the electrode tips. As shown in region (a) of Figure 2, only 0.07 A of the ionic current, which is the consequence of the electrolysis of water, flowed in a short period at the beginning of the pulse. When suitable conditions were attained, electron avalanche occurred, and plasma was generated. Simultaneously, a significant increase in current and a significant decrease in voltage at point (b) were indicated, as shown in Figure 2. The current increased until 5.38 A, and the discharge current predominated over the ionic current in region (c) of Figure 2. After the plasma, the voltage source became a storage capacitor.

Figure 2:

Current-voltage waveforms and generation stages of liquid-phase plasma

3.2 Degradation of Flu by LPP

Figure 3 shows the results for Flu removal from aqueous solutions with varying frequencies and pulse widths. The removal efficiency increased with the frequency and pulse width, and the highest degradation efficiency of Flu by the LPP after 10 min of treatment time was 85.4% at 30 kHz and 3 μs. Increasing the frequency and pulse width can increase the energy of the electron, resulting in frequent bombardment between the electron and water and the generation of more active species such as •OH, H₂O₂, and O₃.

Figure 3:

Removal efficiency of fluorene by liquid-phase plasma for varying discharge conditions

It was assumed that the nonlinear increase in the removal efficiency with the increase in the frequencies and pulse widths was due to the interaction among active species. The conversion of the produced species can be represented as follows [17]:

Hence, a slightly reduced removal efficiency was observed at 30 kHz and 1 μs. In addition, excess input energy can generate oxidant species that can react with active species, thereby resulting in less oxidation in Flu.

3.3 Degradation of Flu by LPP/TiO₂

The effect of TiO₂ on the degradation of Flu by LPP was investigated. The optimal operating conditions were a frequency of 30 kHz and a pulse width of 3 μs. However, the experiments were performed at 20 kHz and 2 μs because of the instability of the equipment under the former conditions. As shown in Figure 4, even though the samples with 0.05 g/L of TiO₂ did not indicate significant differences, the samples with 0.1 and 0.2 g/L of TiO₂ indicated improved performances. The degradation efficiency improved from 44.2% to 67.7% when the sample was treated for 15 min with 0.2 g/L of photocatalyst. The improved removal efficiency was primarily attributed to the conversion of surplus UV light to active species. UV radiation was generated by the abrupt deceleration of electrons, thermal radiation of ions, and recombination processes within the plasma [18]. By radiating UV light from the plasma, electron–hole pairs were generated on the surface of the added TiO₂, and they participated in redox reactions, thereby resulting in active species. The changed use of UV light effectively degraded Flu.

Figure 4:

Removal efficiency of fluorene by liquid-phase plasma combined with photocatalysis

3.4 Degradation of Flu by LPP/Fe²⁺

For the same reason as mentioned above, the operating conditions were set at a frequency of 20 kHz and a pulse width of 2 μs. Figure 5 shows the degradation efficiency of Flu by LPP combined with the Fenton treatment. The addition of the catalyst affected the degradation efficiency of Flu in the LPP reactor. The degradation efficiency improved from 44.2% to 72.4% when the sample was treated for 15 min with 10 mg/L of iron salt. This enhancement was primarily attributed to the effective use of H₂O₂ generated by plasma. The amount of H₂O₂ as a function of the treatment time is shown in Figure 6. As indicated by Equation 1, a hydroxyl radical is formed by the reaction between iron ions and hydrogen peroxide, which is known as the Fenton reaction. Compared with other studies, the association of the LPP with the Fenton reaction indicated significant benefits, since fewer additional chemical reagents were used and effective Applications of the active species were possible.

Figure 5:

Removal efficiency of fluorene by liquid-phase plasma combined with Fenton reaction

Figure 6:

Concentrations of generated H2O2 based on treatment time

3.5 Degradation pathway of Flu under LPP conditions

FTIR was used to identify the functional groups of the Flu decomposition intermediates to estimate the degradation pathway. The positions in the 600–900 cm^-1 region are characteristic of the number of adjacent hydrogen atoms on the aromatic ring, and the peak at 737 cm^-1 in Figure 7 declined, implying the degradation of Flu [19]. Meanwhile, the band centered at 3250–3600 cm^-1 increased significantly after treatment, which corresponded to the formation of hydroxyl groups [20]. By contrast, new peaks were observed at 803, 1030, 1097, and 1261 cm^-1. These peaks were not considered because they originated from the methyl alcohol used in the stock solutions.

Figure 7:

Infrared spectra of fluorene before and after plasma treatment

Optical emission spectroscopy analysis was performed under plasma discharge in water to observe the active species. As shown in Figure 8, the major active species were atomic hydrogen (λ = 662 nm), atomic oxygen (λ = 777 nm), and hydroxyl radicals (λ = 309 nm). The inset in Figure 8 shows the variation in the amount of hydroxyl radicals based on the discharge conditions. The quantity of hydroxyl radicals increased with the frequency and pulse width.

Figure 8:

Optical emission spectrum of plasma in discharge condition at 30 kHz and 3 μs

The results of FTIR and OES analyses suggested that the presence of hydroxyl radicals whose oxidation-reduction potential was 2.8 V, which is even higher than that of ozone (2.07 V) and hydrogen peroxide (1.8 V), contributed significantly to the decomposition of Flu in the aqueous solution.

Author Contributions

Conceptualization, S. H. Lee; Methodology, U. J. Kim; Software, U. J. Kim; Formal Analysis, U. J. Kim; Investigation, U. J. Kim; Resources, S. H. Lee; Data Curation U. J. Kim; Writing-Original Draft Preparation, U. J. Kim; Writing-Review & Editing, S. H. Lee; Visualization, U. J. Kim; Supervision, S. H. Lee; Project Administration, S. H. Lee; Funding Acquisition, S. H. Lee.