Degradation of Ibuprofen in flow-through system by the Electro-Fenton Process activated by two iron sources

The electrochemical degradation of ibuprofen (IBP) by electro-Fenton process has been studied in a flow-through system by evaluating the performance of two different iron sources, sacrificial cast iron anode and FeSO4 salt. The effect of operating conditions, including initial IBP concentration, cast iron anode location, initial FeSO4 concentration, applied current, the split current on the iron anode, solution pH, and flow rate on the efficacy of the process was evaluated. The sequence of the electrodes significantly influences ibuprofen removal. When using cast iron anode as iron source, placing the iron anode upstream achieved the best IBP removal rate. Split current of 3 mA applied on the iron anode out of 120 mA total current is the optimum current for remove 1 mg/L of IBP under a flow rate of 3 mL/min. There is a linear correlation between the applied current and the Fe2+ concentration in the FeSO4-system. The initial IBP concentration does not influence the rate of Fenton reaction. Flow rate influences the degradation efficiency as high flow rate dilutes the concentration of OH radicals in the electrolyte. FeSO4-system was less affected by the flow rate compared to the iron anode-system as the concentration of the Fe2+ was steady and not diluted by the flow rate. Both systems prefer acidic operation conditions than neutral and alkaline conditions. Iron-anode can be used as an external Fe2+ supply for the treatment for iron-free. These findings contribute in several ways to our understanding of the electro-Fenton process under flow conditions and provide a basis for how to design the reactor for the water treatment.


35
Electrochemical advanced oxidation processes (EAOPs) have been widely developed for environmental 36 remediation, especially for aqueous streams ). These processes 37 electrogenerate highly reactive species, such as hydroxyl radicals, which can effectively break down 38 aqueous pollutants such as toxic and persistent pesticides, organic synthesis dyes, pharmaceuticals and 39 personal care products (PPCPs), and industrial pollutants(H. Zhang  produce H2O2 on-site, which avoids the risks of transport, storage, and handling of H2O2; (2) generate 46 and regenerate Fe 2+ in-situ to increase the removal efficiency and decrease the sludge production; (3) 47 control the degradation kinetics to allow the mechanistic study  51 The degradation efficiency of EF process can be influenced by several parameters such as O2 supply, 52 stirring rate or liquid flow rate, temperature, pH, applied potential or current, nature and initial 53 concentration of organics, and concentration of iron catalyst(M. Zhou et al. 2007)(X. Zhang et al. 54 2009)(Su et al. 2019)(F. Yu et al. 2014). The influencing parameters can be separated into two 55 categories. One affects the degradation rate by directly influencing the generation of hydroxyl radicals, 56 and the other influences H2O2 production to affect the OH radicals production indirectly. 57 The pH of the electrolyte is a critical factor for maintaining the efficiency of the EF process. The optimal 58 pH range for the Electro-Fenton reaction is typically around 3(X. ) (Haber et al. 1934). At 59 this pH range, ferrous iron (Fe 2+ ) is in its highest oxidation state, which is necessary for the generation of 60 hydroxyl radicals. Operating under unexpected pH values can cause lower OH radical generation and 61 more iron sludge production. For example, previous research showed that the optimum of the pH is 3 62 on the removal of methyl parathion with 0. Then, the GF-(1:29) was used for the PDMS dampproof coating (Zhao et al. 2020). Since the mass of the 131 PDMS is 20 and 50, the electrodes with PDMS coating were marked as GF-(20) and GF-(50), respectively. 132

133
A vertical acrylic column was used as an electrochemical flow-through reactor with an inner diameter of 134 4.3 cm and length of 15 cm, including three sample ports at 5 cm, 9 cm, and 13 cm from the bottom of 135 the column. Electrodes were placed in parallel with respect to the flow direction at 4 cm, 7.7 cm, and 136 10.3 cm from the bottom of the column (Figure 1). Titanium rods were fabricated at electrode levels at 137 electrical connections to hold the electrodes in place and convey electrical current. 138

154
The iron concentration was determined by the 1-10 phenanthroline analytical method (Komadel and 155 Stucki 1988). For 156 measuring the total dissolved iron, a 0.5 mL sample taken from the sampling ports were filtered 157 by the 0.45 µm syringe filter, then mixed with 0.25 mL of 10% of hydroxylamine hydrochloride 158 to reduce all dissolved iron to ferrous iron. 1 mL of the acetic acid buffer (20 g acetic ammonia 159 and 25 mL acetic acid in 100 mL of water) was used to adjust the pH to 3 to 5, and 1 mL of the 160 1-10 phenanthroline monohydrate (1 g/L) was added subsequently to the sample. The solution was 161 measured at 510 nm wavelength using a UV-VIS spectrophotometer (SHIMADZU UV-1800). The 162 concentration of ferrous iron was measured with the same method without adding the hydroxylamine 163 hydrochloride. The pH was measured by a pH meter (Thermo Scientific). The removal of IBP in the flow-through system approximately followed the pseudo-first-order reaction 170 at the beginning of the experiment. The model is given by: 171 172 Where t is the reaction time (min), k is the rate constant (/min), and C0 and C are the concentration of 173 IBP (mg/L) at times of t=0 and t=t, respectively. 174 The total removed IBP (mg) was estimated by the following equation: 175

Equation 5
176 Where t is the reaction time (min), Ci is the initial concentration of IBP (mg/L) and C is the 177 concentration at t. 178 The charge efficiency (mg/C) was calculated by: 179 180 Where t is the reaction time (s), and I is the applied current (A). The generation of Fe 2+ can be changed by adjusting the split current on the sacrificial iron anode in the 250 electrochemical system. Experiments were conducted to analyze the removal rate of different initial 251 IBP concentration (1, 3, and 5 mg/L) under different split currents (1.6, 3, 6, and 10 mA)( Figure 6). 252 When the initial concentration of IBP was 1 mg/L, the ratio of IBP left in the electrolyte at 115 mins 253 was 69.2%, 83.2%, 80.5%, and 82.8% for split current of 1.6, 3, 6, and 10 mA, respectively. Increasing 254 initial IBP to 3 mg/L decreased the removal rate to 32.4%, 46.4%, 45.6%, and 37.4%. When the 255 concentration of IBP was further increased to 5 mg/L, the removal rate was 26.7%, 34.9%, 28.6%, 256 and 11.5%. The optimum removal rate was achieved at the split current of 3 mA for different initial 257 IBP values. of time and current. A higher split current generated more dissolved iron at sampling P3. There 262 was a sharp decrease when the electrolyte passed the anode, as most dissolved iron will be 263 oxidized to Fe 3+ after passing the anode. P1 has the minimum amount of dissolved iron because the 264 pH at the cathode is alkaline, leading to iron precipitation. Increasing dissolved iron in the 265 electrolyte can enhance the Fenton reaction, but excess iron can consume OH radicals and generate 266 more precipitation. Since the production of H2O2 should be similar under different split currents, the 267 optimum removal rate appears under the maximum OH radical production. The concentration of initial 268 IBP does not affect the generation of OH radicals. Thus, the optimum removal rate appears at 3 mA for 269 different initial IBP. anode. The removal rate was 24.5%, 46.1%, 46.6%, and 33.5% when the current was 60, 90, 120, and 276 150 mA, respectively. The best removal rate appeared at 120 mA with the highest k value (Table 1). The 277 IBP removal was normalized by charge in Figure 8b.  Previous results showed that the yield of H2O2 was low at the beginning of the experiment and took time 285 to stabilize at a low current. Hydrogen peroxide, as the precursor of the Fenton reaction, can directly 286 affect the removal rate. This explains why there is a delay in the degradation experiments at 60 mA and 287 90 mA and why the experiment at 60 mA requires a longer time to reach the steady state (Table 1). 288 The production of Fe 2+ at the iron anode is the same when the total current varies because 291 the split current is the same. Increasing current enhanced water electrolysis on the Ti/MMO anode 292 and the 2e-ORR at the cathode. More H2O2 can be generated for the Fenton reaction. However, 293 increasing anodic oxygen production can oxidize more Fe 2+ to produce iron precipitation. Thus, the 294 amount of dissolved iron at P2 that can be supplied to the Fenton reaction decreases with increasing 295 current ( Figure 9). Previous research also showed that the electrogeneration of H2O2 decreased when 296 the current was over the optimum value because the decomposition reaction was enhanced. Thus, the 297 optimum current is 120 mA to achieve the best removal rate.  respectively. The removal rate decreased with the increase in flow rate. A similar trend was shown for 304 the total removal of IBP under different flow rates. The total removal of IBP decreased from 0.22 mg to 305 0.09 mg and 0.08 mg when the flow rate increased from 3 to 5 and 7 mL/min, respectively. The higher 306 flow rate has a lower k value and needs less time to reach the steady state (Table 2). Increasing the flow 307 rate shortens the retention time for oxygen on the cathode, which can decrease the reaction rate for 308 the 2e-ORR. A decrease in H2O2 concentration will lower the generation of OH radicals as well. 309 Increasing the flow rate also dilutes the concentration of OH radicals in the electrolyte. Thus, the 310 removal rate will decrease when the initial concentration of IBP remains constant. 311 Table 2 The pseudo-first-constant value, the time required to reach the steady state, the delay time and correlation coefficient The production of Fe 2+ on the iron anode remains the same since the current split is constant, but the 317 increased flow rate dilutes the dissolved iron in the electrolyte, which decreases the concentration of 318 the dissolved iron at P3 (Figure 11). However, the iron content at P2 remains similar, mainly because the 319 concentration of anodic oxygen decreased as well with the flow increasing. was 82.7%, 74.3%, 46.3%, and 11.6% when the pH in the influent was 2, 3, 7, and 11, respectively. The 325 removal rate is better under acidic conditions and decreases with the pH increase. Figure 12b shows the 326 pH values from different sampling ports at 60 mins which do not vary significantly with time. pH value 327 decreased at the anode and increased at the cathode due to the water electrolysis. Thus, the samples 328 from P2 and P3 are acidic, and those from P1 are alkaline when the electrolyte is neutral. Water 329 electrolysis is not the dominating reaction at the GF cathode, so pre-acidification can effectively control 330 the pH value at the cathode. The pH from sampling P1 is slightly higher than the value from P2 and P3 331 when the initial pH is 2 and 3. Ti/MMO anode significantly reduces the alkaline electrolyte pH due to the 332 strong water electrolysis reaction. 333 334 Figure 13 concentration of a) dissolved iron and b) Fe 2+ under different pH of the electrolyte 335 pH value has a significant effect on the state of the iron in the electrolyte. Figure 13 shows 336 the concentration of total dissolved iron and Fe 2+ in the electrolyte at different sampling ports and 337 times. The dissolved iron at the Fe anode (P3) is mainly ferrous. The concentration of total dissolved iron 338 is slightly higher than Fe 2+ due to the presence of Fe 3+ . This difference is not apparent because most of 339 the Fe 3+ precipitated. The concentration of dissolved iron decreases with the pH increase. There is 340 minimal dissolved iron in the alkaline electrolyte. It is worth noting that the iron concentration is 341 extraordinarily high when pH is 2 due to the iron dissolution at a very low pH. The Fenton reaction 342 requires dissolved iron as the catalyst to produce OH radicals. Thus, the removal rate is high under pH of 343 2 and 3. Although the modified GF generated more H2O2 under alkaline conditions, the removal rate 344 remained low without enough dissolved iron. 345 the pseudo-first-order reaction kinetics, then reached the steady state. This significant decrease at the 354 beginning is due to the strong Fenton reaction. Iron existed in the electrolyte in the form of Fe 2+ , and no 355 precipitation was observed during this period. The optimum removal rate (86%) under 60 mA appears at 356 the initial Fe 2+ concentration of 4 mg/L (Figure 14a). Reducing the initial Fe 2+ content 357 to 2 mg/L reduces the degradation to 60%. Not all the H2O2 can be activated to OH radicals when 358

Degradation of ibuprofen by FeSO4 as the iron source
Fe 2+ is insufficient. Raising Fe 2+ content to 5.4 mg/L and 10.9 mg/L lowers the removal rate to 359 80% and 70%. More Fe 2+ supplies a stronger catalytic cycle of the Fenton reaction. However, excessive 360 Fe2+ in the electrolyte can compete with IBP for ROS, such as OH radicals. We also observed that more 361 iron precipitate was produced on the cathode when the Fe 2+ concentration increased. The deposit might 362 inhibit the generation of H2O2 at the cathode. Increasing the applied current can change the optimum Fe 363 value for the removal rate. When the current was 90 mA, the optimum Fe 2+ content changed to 10.8 364 mg/L with a removal rate of 98%, compared to 79% at 5.4 mg/L and 69% at 14.3 mg/L. The optimum Fe 365 concentration continues to rise to 14.3 mg/L under 120 mA and 21.8 mg/L under 200 mA. The optimum 366 removal rate under 60 mA is lower than the others because insufficient H2O2 production leads to a lower 367 OH radical production. 368 369 370 Figure 15 Correlation between initial Fe 2+ content and current 371 Increasing current can enhance the reaction of water electrolysis and 2e-ORR to produce more anodic 372 O2 and H2O2. However, O2 can oxidize Fe 2+ to Fe 3+ , which competes with the Fenton reaction for Fe 2+ . At 373 the same time, the residual H2O2 can consume OH radicals to generate HO2·, which exhibits a low 374 oxidation power compared with OH radicals (Equation 10) and is relatively unreactive to organic 375 matters (Bielski et al. 2009). Thus, with the increase of applied current, the concentration of Fe must 376 increase to guarantee the optimal OH radicals' production for the removal of IBP. 377 The optimal Fe concentration for different currents (ranging from 60 mA to 200 mA) under 1 mg/L of IBP 378 is summarized in Figure 15. A positive linear correlation exists between the optimum Fe content and the 379 current. An equation to calculate the optimal Fe concentration (y, mg/L) under a specifically applied 380 current, y=0.   Figure 17a demonstrates the effect of current on the removal 404 of 1 mg/L of IBP at the flow rate of 3 mL/min. The optimum removal rate is 97% at 115 mins with 405 a current of 120 mA. Increasing the current to 200 mA or decreasing the current to 90 mA inhibited the 406 removal rate to 33% and 60%, respectively. The same trend was observed at IBP concentration of 3 407 mg/L and 5 mg/L. The optimum removal rate appears at 120 mA under different initial ibuprofen 408 content. This result matches the previous conclusion. Changing current can affect the electrochemical 409 reactions in the system. As a stable contaminant, IBP does not react with H2O2 and O2 straightly. Thus, 410 changing current cannot directly affect the removal rate. The optimum removal rate appears when the 411 maximum production of OH radicals is achieved. IBP does not affect the OH radical generation ( Figure  412 17d).  Figure 18 shows the totally removed IBP from the system in 115 mins and the charge efficiency under 416 different current conditions. The total removed IBP increased with increasing of the initial IBP. OH 417 radicals were more inclined to react with IBP when the concentration of IBP in the electrolyte was 418 increasing. The removal rate was low for the high initial concentration, but the quantity (in mass) of 419 removal was high.  Figure 19 analysis the correlation between iron content and initial concentration of IBP. 427 The removal rate at 10.8 mg/L of Fe 2+ is 87.5%, 62.7%, and 52.2% at initial IBP of 1, 3, and 5 mg/L. The 428 optimum removal rate appears under 1 mg/L of IBP. Increasing current does not change this trend. The 429 removal rate decreases with the increase of the initial concentration of IBP at the fixed Fe content. The 430 lower removal rate of IBP at high concentration is attributed to the generation of degradation by-431 products that compete with the targeted contaminants for OH radicals. The previous study shows that 432 the degradation of IBP by the EF process is mainly because of the destruction of the propionic acid group 433 to produce 4-isobutylacetophenone (Yuan, Gou, et al. 2013). OH radicals prefer to attack 4-434 isobutylacetophenone because of the stronger conjugated effect from the carboxyl group and benzene. 435 Therefore, the higher initial concentration can produce more by-products than the lower one, which will 436 compete with the removal of IBP. The highest removal rate appears at 14.3 mg/L of Fe 2+ , which is the 437 optimum Fe 2+ value under 120 mA. There is no correlation between the iron content and the 438 concentration of initial IBP. conditions, the total dissolved iron did not fluctuate much with time. There was a significant drop under 449 pH 7 after 30 mins due to the pH change around the electrode. As the pH stabilized with time, the 450 concentration of the dissolved iron became stable. Iron concentration in pH 11 is below the detection 451 limit for the spectrophotometer because of the formation of iron precipitation. Classic Fenton uses 452 soluble iron (Fe 2+ or Fe 3+ form) to guarantee the highest process efficiency. Fe 2+ can be quickly oxidized 453 to Fe 3+ , and most of the Fe 3+ will produce ferric hydroxide sludge when the pH is greater than 4. Thus, 454 the optimum pH for the Fenton reaction is between 2.8 to 3. Although the previous study showed that 455 the modified GF has better performance under alkaline conditions, the Fenton reaction can be greatly 456 inhibited. Under the optimum pH, more H2O2 can be activated to OH radicals for the degradation 457 reaction. All experiments were conducted under 120 mA with the initial IBP of 1 mg/L and Fe of 14.3 mg/L. The 464 removal of IBP was 97%, 91%, 89%, and 8% at flow rates of 3, 6, 8, and 10 mL/min at 115 mins. 465 Increasing the flow rate can dilute the concentration of H2O2 and OH radicals, but the concentration of 466 IBP and Fe 2+ remains the same. Thus, increasing the flow rate can decrease the removal rate. The total 467 removed IBP from the system in 115 mins increases with the flow rate increase from 3 to 8 mL/min then 468 decreases when the flow rate keeps increasing. The concentration of OH radicals is sufficient to react 469 with most of the IBP in the system when the flow rate is less than or equal to 8 mL/min. Because the 470 high flow rate will carry more IBP through the system at the same period, the high flow rate has a high 471 value of total removed IBP. However, when the flow rate keeps increasing to 10 mL/min, the 472 concentration of H2O2 is too low to generate enough OH radicals for the system. The removal rate and 473 the total removed IBP are then very low. The FeSO4 system is less sensitive to the flow rate compared to 474 the iron anode system. 475

476
This research evaluated the performance of two iron sources (cast iron anode and FeSO4 salt) on the 477 removal of ibuprofen by electro-Fenton process in the flow-through system. Experimental results 478 indicate that cast iron anode located at the bottom of the electrode's sequence achieve the best 479 performance for the ibuprofen removal. The optimum split current on iron anode is 3 mA when the total 480 current is 120 mA. Lower split current cannot generate sufficient Fe 2+ for the Fenton process but higher 481 split current can increase the iron precipitation. There is a linear correlation between the applied 482 current and the concentration of the concentration of Fe 2+ in the FeSO4-system. However, there is no 483 correlation between concentration of Fe 2+ and IBP concentration or current and initial IBP 484 concentration. Both systems have an optimum applied current at which the accumulated H2O2 485 concentration is maximum. Increasing flow rate can decrease the IBP removal rate due to the decreasing 486 of the oxygen retention time for the H2O2 production and the dilution of H2O2 concentration. However, 487 compared to FeSO4-system, flow rate has more effect on cast iron system because it dilutes the Fe 2+ 488 concentration as well. Both systems prefer to operate under acidic conditions because the Fe 2+ will be 489 less susceptible to be generate iron precipitation. Cast iron works as an external Fe 2+ for the electro-490 Fenton process, which can be used for the iron-free water conditions or the water without enough Fe 2+ . 491 FeSO4 has better resistance to the flow rate changing. It can be used for the high flow rate system. These 492 findings contribute in several ways to our understanding of the electro-Fenton process under flow 493 conditions and provide a basis for how to design the reactor for the water treatment. 494