Degradation of 4-Chlorophenol in Aqueous Solution by Sono-Electro-Fenton Process

Electro-Fenton (EF) and ultrasound radiation (US) have been of interest for the removal of chlorinated compounds from water. This study evaluates the effects of different parameters on sono-electro-Fenton (SEF) for degradation of 4-chlorophenol (4-CP) in an aqueous solution. This study uses pulsing US waves along with Pd-catalyzed EF to degrade contaminants in water while maintaining temperature. The usage of pulsing US waves along with Pd catalyzed EF to remove contaminants while maintaining temperature has not been reported previously. SEF ability to degrade 4-CP was compared with the performance of each process (EF and sonolysis) alone. Initial pH, current density, background electrolyte, Fe2+ concentration, Pd/Al2O3 catalyst concentration, US waves, and sonifier amplitude were optimized in a two electrode (Ti/mixed metal oxide or Ti/MMO) batch system. The degradation of 4-CP increased from 1.85% by US to 83% by EF to nearly >99.9% by coupled SEF. With US radiation under 70% amplitude and 1:10 ON/OFF ratio, the removal rate of 4-CP increased to 98% compared to 62% under EF alone within the first 120 min in the presence of 80 mg L−1 Fe2+, 16.94 mA cm−2 of current density, 1 g L−1 Pd/Al2O3 catalyst (10 mg Pd), and initial pH of 3. However, the degradation rate decreased after 120 min of treatment, and complete 4-CP removal was observed after 300 minutes. The sonolysis impacted the 4-CP removal under coupled SEF, mostly due to the contribution of mass transfer (micromixing), while radical formation was found to be absent under the conditions tested (20kHz). The pulsed US was found to increase the temperature by only 8.7°C, which was found not to impact the 4-CP volatilization or degradation. These results imply that low-level US frequency through pulses is a practical and efficient approach to support electro-Fenton reaction, improving reaction rates without the need for electrolyte cooling.

production rate of • OH compared to traditional Fenton's method, and supports reduction of Fe 3+ to Fe 2+ at the cathode (Eq. 5) [1]. Although an in situ production of H2O2 can be hindered by the low solubility of oxygen in water [20,23,24], a Pd catalyst has been proven to enhance its production rate. Ultrasonic radiation has also demonstrated a great potential for various applications, including the intensification of chemical synthesis, cleaning, and water treatment [9][10][11]25]. Application of frequencies above 20 kHz causes growth of cavitation bubbles which become unstable after a number of cycles. Upon collapsing, each of the bubbles acts as a hotspot, generating energy to increase the temperature (up to 5,000 K) and pressure (up to 500 atm) with cooling rates as fast as 109 K/s. There are many parameters affecting the cavitation and bubble collapse process (sound wave frequency and intensity, external pressure and temperature, solvent characteristics, and presence of soluble gases), and, consequently, the impact of sonolysis on the degradation of the contaminants in aqueous solutions [8,13,26].
Sonolysis occurs through three reaction zones; the gas phase region inside the bubble (pyrolysis reactions), the interfacial region (reactions occurring in pressure/temperature gradients in aqueous phase), and the bulk solution [13,26,27]. The chemical effect of the applied acoustic field is the sonolysis of water molecules and thermal dissociation of oxygen molecules (when present), which produce different kinds of reactive species ( • OH, H • , O • and hydroperoxyl radicals (OOH • ) (example given in Eq. 6 where the sign ')))' denotes US power).
The physical effects of US irradiation are induced by cavitation bubbles (micro-jets and shockwaves) and by propagation of US waves through a liquid medium (streaming) [13]. Where surface instabilities occur, cavity collapse is very asymmetric and generates highspeed jets (micro-jets) of liquid causing the enhancement of the rate of mass transfer [28][29][30].
gas bubbles at the surface of electrode leading to removal of gas bubbles from the surface of electrode through the waves' vibration [29,31]. Several studies have combined US radiation with other advanced oxidation processes (AOPs), including using EF to remove chlorinated compounds from aqueous solutions [7,[32][33][34][35][36][37]. These coupled processes have resulted in an increase of contaminant removal in comparison with using each method separately. Trabelsi et al. (1996) used 500KHz US radiation along with EF (current density=68 Am −2 ) and showed that a total degradation of phenol within 20 minutes with no production of toxic intermediates is possible [38]. Yasman et al. used US radiation (20 KHz) along with EF mechanism to treat 2,4-dichlorophenxyacetic acid (2,4-D) and its derivative 2,4dichlorophenol (2,4-DCP). They accomplished almost 50% oxidation of 2,4-D solution (300 ppm) in only 60 seconds, while complete removal was achieved after 10 minutes [39]. Liang et al. investigated the impact of US on a Fenton-like reaction for 4-chlorophenol removal from water, and found that under tested conditions, 4-CP was completely decomposed within 2 min of ultrasonic irradiation when its initial concentration in the solution was 100 mg/l [28]. In another study, Mehmet et al. used an undivided electrolytic cell with a Pt anode and a 3-dimensional carbon-felt cathode to carry out EF and SEF oxidation for three contaminants, 2,4-dichlorophenoxyacetic acid (2,4-D), 4,6-dinitro-o-cresol (DNOC), and the synthetic azo dye azobenzene (AB). It was observed that synergistic effect between EF and US provides a higher degradation rate than that provided by the two techniques separately for 2,4-D and DNOC [6].
The objective of this study is to investigate oxidation of 4-CP by Pd-catalyzed EF process coupled with sonolysis using pulsed US waves. While the benefits of US pulse by borondoped diamond electrodes were reported [40], the application of pulsing US waves along with Pd-catalyzed EF reaction has not been reported. In this research, we examine the application of US waves at ON/OFF ratio of 0.1 (US was ON: 5.9 sec. and OFF: 59 sec). The performance of EF under different initial pH, Fe 2+ concentration, palladium (Pd) catalyst concentration, background electrolytes, and current densities was also tested. We also evaluated the simultaneous use of Fe anode in different application regimes as a practical, in situ supply of ferrous iron. The SEF tests were conducted under optimum conditions, contaminant removal by SEF process was compared with both EF and sonolysis, and the mechanisms involved were then evaluated.

Experimental setup
As shown in Figure 1, a one-liter acrylic cell with an 11.4 cm inner diameter and a 10 cm height was used as batch reactor. Two Ti-based mixed metal oxide meshes (Ti/MMO, IrO 2 /Ta 2 O 5 coating on titanium mesh type, 3N international, USA) with 3.6 cm in diameter, 1.8 mm thickness, and a surface area of 11.8 cm 2 were used as both anode and cathode. The distance between the electrodes was 6 cm. The synthetic contaminated groundwater was prepared by adding 4-CP to achieve a concentration of 200 ppm in the electrolyte (10 mM Na 2 SO 4 , NaHCO 3 or NaNO 3 ) with different Fe 2+ concentrations. An alternative to the addition of ferrous sulfate as the Fe 2+ source was using a cast iron anode with dimensions of 85×15×1.8 mm (length × width × thickness) to produce Fe 2+ in-situ. The current was split between Ti/MMO and iron anodes by a rheostat, where the current applied to the iron anode was calculated to supply a total of 80 ppm Fe 2+ (based on Faraday's law assuming the charge transfer between electrode surface and electrolyte is a 100% faradaic process). Fe 2+ production was tested for optimum EF in the following conditions: during the entire 240 min by applying 2.85 mA cm −2 to iron anode (with a surface area of 6.3 cm 2 ), during the first 30 minutes of treatment under 22.85 mA cm −2 , and following a 30-minute delay of treatment under 3.17 mA cm −2 supplied to the iron anode. The iron anode's ON/OFF periods were applied based on the estimated H 2 O 2 production, which reaches maximum values after 30 min.
The use of an iron anode in the electrochemical cell allowed for the generation of a wide range of Fe 2+ concentrations without relying solely on the naturally present amounts of Fe 2+ in the groundwater (concentration rarely exceeds 50 ppm). Sulfuric acid and sodium hydroxide were used to adjust the pH of the electrolyte. After adding the synthetic groundwater, defined Fe 2+ , and Pd/Al 2 O 3 catalyst concentrations, the cell was sealed and the solution was stirred at a rate of 180 rpm using a magnetic stirrer. As summarized in Table 1, the influence of different parameters on EF towards 4-CP transformation were tested.
During SEF and sonolysis tests, a sonifier (20 KHz Branson Ultrasonics Co.) with a 7.7 cm titanium horn was placed in the reactor and defined US amplitudes were applied in pulses. The SEF tests were conducted under 80 mg L −1 Fe 2+ , 10 mg L −1 Pd/Al 2 O 3 powder, initial pH of 3 and 16.94 mA cm −2 current density, with the amplitudes (%) of: 10, 30, 50, and 70 using ON/OFF pulse ratio of 0.1 and 10% to 30% with ON/OFF=0.2.

Analysis
At specific times, 2 ml of solution were collected from the sampling port (located 2.4 cm from the bottom of the reactor, Figure 1) and were filtered through a 0.22 μm pore size syringe filter. 4-CP and phenol concentrations were measured by a 1200 Infinity Series HPLC (Agilent) equipped with a 1260 DAD detector and a Thermo ODS Hypersil C18 column (4.6 × 50 mm) with a 5 µm particle size. Mobile phase was a mixture of methanol, water, and glacial acetic acid (49:49:2) with a 1 mL min −1 flow rate. Detection wavelength was 254 nm. The retention time was 2.5 min for phenol and 4.34 min for 4-CP [41]. Total organic carbon (TOC) measurements were performed by a TOC analyzer, TOC-L CPH-CPN (Shimadzu, Japan), after sample filtration through 0.45 μm pore size filters (Millipore), and acidification (pH≤2) with concentrated HCl. The 4-CP removal % was calculated by (Eq. 7): where C 0 is the initial concentration of 4-CP (mg L −1 ) and C t is 4-CP concentration at time t during treatment (mg L −1 ). pH and dissolved oxygen (DO) were measured by pH meter and DO meter (Thermo Scientific). H 2 O 2 was measured at 405 nm on a Shimazu UV-Vis spectrometer after coloration with TiSO 4 . Benzoic acid (BA) was used as hydroxyl radical probe [4] and was measured along with p-hydroxybenzoic acid (pHBA) by a 1200 Infinity Series HPLC (Agilent) equipped with a 1260 DAD detector (Agilent) and a Thermo ODS Hypersil C18 column (4.6 × 50 mm) with a 5 µm particle size. The initial BA concentration was 5 mM. 1 mL samples were taken at the predetermined time intervals and samples were then filtered by the 0.22μm filter (PVDF). The treated samples were then quenched by 0.1 mL ethanol before the HPLC analysis. The mobile phase was a mixture of methanol and 0.1% phosphoric acid (20:80) with a 0.5 mL min −1 flow rate. The detection wavelengths for BA and p-HBA were 280 nm and 210 nm, respectively. The retention times for BA and p-HBA were 14.0 min and 5.0 min, respectively. Both BA and p-HBA were stable throughout the analytical process.  Figure 2a. Preliminary tests showed that pH=3 provides a higher removal percentage, which is in accordance with other studies [23,42]. The degradation of 4-CP via EF reaction follows zero-order kinetics, indicating that degradation is limited by the availability of the reactive • OH ( Table 2). The 4-CP degradation rate increased from 0.0004 min −1 in the absence of Fe 2+ to 0.0043 min −1 in the presence of 80 mg L −1 Fe 2+ . In the absence of Fe 2+ , only 11% of 4-CP was removed via (i) indirect hydrodechlorination at the Ti/MMO cathode, and/or (ii) Pd-catalyzed reduction processes [23]. The results show that an increase in Fe 2+ increases 4-CP degradation as it increases • OH concentration reactions.

Influence of different Fe
In order to evaluate the mechanism of 4-CP removal, EF was performed in the presence of two different concentrations of tert-butyl ( • OH scavenger) (Eq. 8) [1]. A significant change in 4-CP removal was observed after 60 minutes of treatment in the absence and presence of the radical scavenger; the 4-CP zero-order decay rate in the absence and presence of tertbutyl are 0.0041 min −1 and 0.0016 min −1 , respectively. Also, changes in degradation rate during 60 minutes of treatment in the presence of tert-butyl were negligible. This indicates that • OH are the primary reactive species responsible for 4-CP degradation via electro-Fenton reaction supported by Pd catalyst under conditions tested (approx. 75%) [1].
In order to continuously supply the system Fe 2+ and delay the redox of ferrous ion to ferric ion, an iron electrode was used instead of externally adding Fe 2+ [31,43]. In addition, the capability to maintain in situ Fe 2+ sources are valuable for the treatment of groundwater with low natural Fe 2+ content. Iron anodes have multiple advantages: (i) Fe 2+ is continuously released from the sacrificial iron anode, (ii) manipulating the current density controls the electrolytic production of Fe 2+ , (iii) by reversing the polarity of iron electrode, the generation of Fe 2+ can be prevented or suppressed [31,44,45], and iron anode reactions demand less energy [46]. Figure 2b shows the comparison of the system's performance for 4-CP removal in the presence of iron anode and external Fe 2+ addition. The EF system is most effective (>99.9% removal) when the iron anode was operating for the first 30 min of testing because Fe 2+ is continuously produced in situ by anodic corrosion. The correlation between Pd/Al 2 O 3 concentration and H 2 O 2 production has previously been proven [23].

Pd catalyst-The
The influence of catalyst concentration was also evaluated based on TOC removal in addition to 4-CP transformation and decay (Figure 3b). The decay in TOC during the treatment indicates the total mineralization of the parent compound (4-CP) and its oxidation byproducts. Similar to 4-CP decay, higher Pd concentrations increase TOC removal rates. However, the overall TOC removal is limited (% with 10 mg Pd L −1 ) indicating that 4-CP transforms into other dissolved organic compounds (e.g., phenol) within the first six hours of treatment. Prolonging the treatment to 10 hours removed up to 85% of TOC since • OH are continuously generated during EF, causing total mineralization of 4-CP and its byproducts.
In addition to Pd concentration, we also tested the influence of Pd support type on the overall degradation efficiency. Although Pd/Al 2 O 3 is a commercial catalyst extensively investigated for catalytic oxidation of volatile organic carbons (VOCs), we tested Pd on active carbon (Pd/C) as alternative catalyst support. While both catalysts support the 4-CP removal (Figure 3c), the transformation pathways significantly differ. Based on the control experiments (without current), 59.71% of 4-CP was removed from the solution in 240 min when Pd/C was applied while 4-CP concentration decay was only 3.23% when Pd/Al 2 O 3 was used.
Absorption rate of Pd/AL 2 O 3 is lower than Pd/C, which is consistent with other studies [42]. This indicates that Pd/C supports 4-CP sorption over EF reaction and, although removal rate and efficiency is significant, Pd/C is not suitable catalyst for 4-CP degradation via EF reaction.
Because of the intrinsic properties of Al, such as its low standard reduction potential, high abundance, high reactivity, stability, and inexpensiveness, Pd/Al 2 O 3 was used as a catalyst type in all experiments. Although current efficiency (calculated based on Faraday's law) indicates that utilization of 16.94 mA cm −2 is significantly less compared to 10.16 mA cm −2 , we conducted all our tests under 16.94 mA cm −2 . This is based on the effects of current density on the degradation profile of phenol, a 4-CP degradation byproduct that was analyzed during the treatment (Figure 4b). Since the main focus of this study was to evaluate the feasibility of using US to enhance electro-Fenton reaction, further analysis is needed to identify all oxidation byproducts. Charge transferred between electrode and electrolyte was calculated. Then, at each time point, removal efficiency was divided by the charge to produce Figure 4c. Figure 4c shows removal efficiency of 4-CP per charge versus time, which are 45%, 46%, and 38% for 5.08 mA cm −2 , 10.16 mA cm −2 and 16.94 mA cm −2 respectively. As can be seen in the figure, i=16.94 mA cm −2 has lower removal per charge transferred, meaning that at i=16.94 mA cm −2 some charges transferred are not being used in the 4-CP oxidation reactions.

3.1.4
Background electrolyte-Background electrolytes affect EF performance because they improve the solution conductivity and can either support or hinder the efficiency of EF reactions. Figure 5 shows the concentration of 4-CP over time in the presence of 10 mM NaNO 3 , 10 mM Na 2 SO 4 and 10 mM NaHCO 3 . The least 4-CP removal appears in the presence of 10 mM NaHCO 3 (k=0.0029 min −1 ) while the most removal is the system containing 10 mM Na 2 SO 4 (k=0.0043 min −1 ). The possible effects of inorganic anions on the electro-Fenton reaction include: (i) effect of ionic strength, (ii) complexation or precipitation of iron species, (iii) scavenging of • OH and formation of less reactive inorganic radicals, and (iv) oxidation including these inorganic radicals [48].
In EF systems, NaHCO 3 suppresses performance since it is not a strong electrolyte but, more importantly, because HCO 3 − acts as a • OH scavenger [29,49]. Although scavenging leads to formation of carbonate radicals, the reduction potential is less (E=1.5V) than that of • OH (E=2.43 V), meaning that the general oxidative activity in the system depletes. Further, in the pH range used in the study, a Fe(CO 3 ) complex is expected to be the most kinetically active Fe 2+ species for the H 2 O 2 activation [50].

Optimization of ultrasound (US) amplitude-We applied sonification in
pulses to avoid the need to cool the system thus increasing the practicality of the US use [40]. Optimum EF parameters (200 mg L −1 4-CP as an initial concentration, 80 mg L −1 Fe 2+ , 16.94 mA cm −2 of current density, 10 mg L −1 Pd/Al 2 O 3 catalyst and initial pH of 3) were used to identify the best sonifier amplitude. Figure 6a

Ultrasound (US) impact on 4-CP removal-Comparison of performance
under EF, US and SEF. In order to evaluate the impact of US on 4-CP degradation, we conducted sonolysis of 4-CP performed under optimum amplitude (70%) and ON/OFF ratio of 0.1. As presented in Figure 7a, application of US alone removed only 1.85% of 4-CP while SEF and EF achieved removal of >99.9% and 83%, respectively. Besides 4-CP decay, SEF shows favorable phenol decay profile (Figure 7b). Garbellini et al. found that the degradation kinetics of volatile organic compounds supported in sonicated electrochemical tests are due to the increase of mass transport, minimization of the electrode fouling, and the combined generation of • OH [40]. Here, we evaluated the sonolysis impact on degradation during sonolyisis assisted electro-Fenton of 4-CP in the means of: 1) H 2 O 2 and • OH formation, 2) changes in temperature, and 3) possible contribution of micromixing [7,28,35]. During EF, US and SEF, we tested the temperature changes, DO, H 2 O 2 and • OH (Table 3).

US impact on reactive species formation.:
The negligible change in 4-CP concentration during sonolysis indicates that under tested conditions there is no cavitation bubble impact (pyrolysis and radical formation), and that the temperature increase (+8.7°C) has minimal impact on 4-CP volatilization. The H 2 O 2 production was limited (up to 0.3 mg/L at 180 min comparing to 9.9 mg/L during EF) with negligible • OH formation (confirmed by BA decrease and pHBA formation). Previous studies show that sonolytic systems with a higher frequency than that used in this study have proven to be effective in improving organic compound removal efficiency [1,40,51].
The application of US along with electrochemical process slightly increased formation of H 2 O 2 to 1 mg/L at 180 min when Pd catalyst was absent compared to 0.3 mg/L with only sonolysis. When the Pd catalyst was applied, the H 2 O 2 formation reached 9.8 mg/L 180 min following the kinetics of EF (b), after which it declined rapidly to 7.8 mg/L. It was also found that the production of • OH based on reaction with BA was negligible when solely applied through sonolysis and coupled with electrochemical processes.  Table 3 indicate that enhancement of EF by sonolysis relies on processes other than enhancement of H 2 O 2 and • OH formation, and relies on loss due to volatilization during temperature increase.
Impact of temperature changes.: During SEF tests, we monitored the temperature changes and evaluated the impact of temperature on the 4-CP removal since we applied pulsed sonification in order to increase practicality of the applied coupled treatment and avoid the use of a cooling bath [53][54][55]. Although temperature increase had a negligible impact of 4-CP degradation/volatilization sonolysis, temperature increase has been found to have an impact on rates of reactions via Fenton process [38].

US contribution to mass transfer.:
The results indicate that there is no impact of US on overall 4-CP removal in SEF due to increases in reactive species production and temperature changes. This was expected, because higher frequencies are responsible for reactive species formation and pulses allowed the temperature control. Monnier et al. found that micromixing through acoustic cavitation was more important at 20 kHz than at higher frequencies (540 kHz or 1 MHz) [7,56], while acoustic streaming has more impact at higher frequencies [57]. Besides temperature increase and chemical effects of violent bubble collapse, studies have shown that the collapse in the electrolyte-containing particles larger than 150 µm during US can produce shock waves and micro-jets, which cause powders to act as sonochemical catalysts through chemical, physical or combined mechanisms [26,30,40]. Micromixing causes significant increase of the mass transport (accelerating undissolved solute and impurity particles to several hundred meters per second), which enhances the rates of reactions [7,26,28,30,40,49,56]. Due to the presence of Pd powder and applied frequency of US (20kHz), it is reasonable to suggest that the micro-jetting is the primary cause of the enhancement of 4-CP removal under SEF compared to EF tested in this study. Our results indicate that the production of H 2 O 2 (primary Pd catalyzed reaction) was not enhanced by US application although impacts of micro-jets include the dispersion of the catalysts and increase of the reactive area [7,28,58]. The improved fluid mixing by applied US is found to impact the rate of different chemical reactions even under mixing (as applied here) [11,25,51], and is suggested as a main impact of US on the SEF removal of 4-CP under tested conditions in this study. For example, Jordens et al. found that the best micromixing was with a 24 kHz probe and high power intensities [57] which is in agreement with 4-CP profile (Figure 7). Further investigations are needed to confirm the micromixing impacts on SEF and evaluate the parameters to enhance this effect on contaminant removal.

CONCLUSION
In this study, the performance of EF, US and SEF processes on 4-CP degradation were compared. Fe 2+ concentration, Pd/Al 2 O 3 catalysts concentration, initial pH and current density, and background electrolyte were evaluated in batch EF system. Optimum values were selected and SEF tests were performed with a 20KHz US instrument wave amplitude. Sonolysis of the contaminant was performed under optimum amplitudes. Optimum operating conditions for EF process are 80 mg L −1 Fe 2+ , 10 mg L −1 Pd/Al 2 O 3 , initial pH 3, 10 mM Na 2 SO 4 and 16.94 mA cm −2 current density. Under these conditions, removal of 4-CP within 180 minutes was highest by SEF (>99.9%), followed by EF (83%) and US only (1.85%). Since radical formation was found to be absent under the conditions tested (20kHz), the sonolysis impacted the 4-CP removal probably due to increasing the mass transfer. The US increased the temperature by 8.7°C which did not impact the 4-CP volatilization or degradation. These results show that low level US frequency through pulses could be used to support electro-Fenton reaction by improving the reaction rates without the need for the electrolyte cooling.