Formation of iodinated products in Fe (II)/peroxydisulfate (PDS) system

The transformation of iodide (I ) and hypoiodous acid (IO ) by Fe (II)-activated peroxydisulfate (PDS) in the presence of natural organic matter (NOM) was investigated. In the water of a reservoir in Shenzhen, four kinds of organic matter were detected: iodoform> diiodo-bromomethane> dibromo-iodomethane> dichloromethane. As pH increased from 3 to 10, I conversion rate decreased from 31.0% to 11.0%. There is basically no iodic acid (IO 3 ) generated under alkaline conditions, probably due to these organic matters. When PDS concentration is low, PDS can oxidize part of I to IO , and the oxidation of I into IO 3 reaction and continued oxidation reaction exist at the same time, so some IO 3 is generated in the system. As the proportion of Fe (II) increases, the generation rate of IO 3 gradually increases. When Fe 2þ/PDS is 1:5, 2:5, and 3:5, the conversion rates of IO 3 are 4.7%, 6.5%, and 8.4%, respectively. There are two main reasons for this: (i) the reaction of IO produced in the system with humic acid (HA) inhibits the further conversion of IO to IO 3 ; (ii) both HA and Iin the system can react with PDS, resulting in a reduction in the oxidation rate of I .

Yang et al. ). During oxidation, two pathways are conducted: (i) HOI reacts with natural organic matter (NOM) in water to produce a series of iodine by-products (I-DBPs), including IO À 3 and iodomethane, (ii) HOI is further oxidized to iodate (IO À 3 ) (Li et al. ; Dong et al. ). During the past decade, it has been investigated that I-DBPs are more toxic than chlorinated and brominated by-products, so there is an urgent need for a way to control iodine by-products in water bodies (Smith et  The oxidant can sequentially oxidize the I À to IO À and IO À 3 (Li et al. ). The IO À 3 will not react with organic matter to form iodine by-products, and the reaction rates of hypoiodine with different oxidants are different. Previous studies have investigated that peroxymonosulfate was able to oxidize I À , the second-order rate constant of reaction was 1.01 × 10 3 M À1 s À1 , while that for HOI was 1.08 × 10 2 M À1 s À1 (Li et al. ). Studies have shown that the concentration of iodized trihalomethanes increased with the concentration of I À in raw water (Hua et  indicates that the potential genotoxicity and cytotoxicity of I-DBP are usually several to hundreds of times higher than its chlorinated and brominated analogs. Peroxydisulfate (PDS), as an oxidant that is relatively stable at room temperature, has strong oxidizing properties and high solubility, and it has gradually been used in the environ-
Methanol, acetonitrile and acetic acid with purity >99% were purchased from Merck. All solutions were prepared using deionized water with 18.2 MΩ/cm Milli-Q water.

Formation of iodinated products
All the experiments were conducted in a 250 mL batch reactor at temperature 25 C. Add a pre-configured buffer solution containing 4 mM sodium tetraborate/potassium dihydrogen phosphate to a 250 mL iodine flask, add a calculated volume of KI standard solution, and prepare the concentration of KI required for the experiment. Use 0.1 mol/L H 2 SO 4 or 0.1 mol/L NaOH to adjust the pH of the prepared KI solution to the set value. The sampling amount is determined according to the amount of target substance required to determine the index, and a certain volume of sodium sulfite needs to be added immediately after sampling to terminate the reaction. After the end, the pH change is less than 0.2.
When measuring IO À 3 , add a calculated volume of Fe (II) to a 250 mL iodine flask, then add a calculated volume of PDS, and at the same time start timing sampling, take 1 mL and add it to the liquid phase containing 2,6-dichlorophenol and anhydrous sodium sulfite. In a vial, measure hypoiodic acid, take 5 mL and add it to an ion chromatography tube containing anhydrous sodium sulfite, and measure iodide ion and hypoiodic acid. In order to ensure the reliability of the experimental results, each group of experiments was repeated three times, and the error of the experimental results was controlled within 5%, and a blank experiment was set.
Analysis I À , IO À and IO À 3 I À , IO À 3 and IO À were measured by ion chromatography (ICS1500). The specific chromatographic conditions are a Dionex ICS1500 ion chromatograph (USA); isocratic leaching is used, the eluent is 9 mmol/L Na 2 CO 3 ; the flow rate of the eluent is 1 mL/min; the suppression method is conductivity suppression, which is caused by electrolytic water producing H þ , the suppressor is an ASRS-Ultra 2 mm external water anion suppressor; column temperature: 35 C; the detector is a conductivity detector; detector temperature is column (4 mm × 4 mm); IonPacAS9-HC type separation column (250 mm × 4 mm); injection volume is 1,000 μg/L; method detection limit is 0.3 μg/L. The detection of IO À used the method of reacting IO À with 2,6-dichlorophenol to produce 4-iodo-2,6-dichlorophenol, by measuring the amount of 4-iodo-2,6-dichlorophenol. The 4-iodo-2,6-dichlorophenol was measured by Waters high performance liquid chromatography under the conditions of: C18 column (4.6 mm × 250 mm × 5 μm), and the mobile phase is V(acetonitrile): V(water) ¼ 7:3, ultraviolet detection wavelength is 212 nm, flow rate is 1 mL/min, injection volume is 100 uL.

Formation of I-DBPs in water
The determination of I-DBPs is carried out by combined use of gas and quality. Gas chromatographic conditions: electron capture detector, detector temperature is 300 C, inlet temperature is 200 C, program temperature is 40 C, hold for 5 min, to raise 1 C/min to 45 C, hold for 3 min, raise 10 C/min to 135 C, then raise 25 C/min to 220 C, inject 1 μL, splitless injection. Mass spectrometry conditions: EI ion source, ion source temperature 225 C, quadrupole temperature 150 C, chromatography-mass spectrometer connection temperature 170 C, electron energy 70 eV, solvent delay 3 min, scan mass range m/z 20-500, scan mode select ion monitoring (m/z).
Extraction conditions by gas mass spectrometry: MTBE is used as the extractant. At the end of the reaction, a 30 mL sample and a bottle with a Teflon cap are taken, 6 g of anhydrous sodium carbonate is added to shake, and a 3 mL pure MTBE solution is added to the sample with a pipette gun. In the test, shake for 20 minutes in a vortex shaker and stand still for 20 minutes. Take 1 mL of the upper layer organic solution in the sample bottle with the air quality for detection.

AOI
Absorbable organic halogen (AOI) was measured with a Multi X 2500 total organic halogen analyzer (Analytic Jena, Germany). An APU28 carbon column was used for sample adsorption. Add nitric acid stock solution to the sample, the pH value of the whole sample is less than 2, take a 50 mL sample in the pressure filter, and set the program.
After the sample passes through the APU28, the AOI in it is completely adsorbed on the APU28 carbon column. Then elute the ions in the carbon column with 25 mL sodium nitrate stock solution, take out the sample carbon column at the end of the desorption, and absorb the surface moisture on the filter paper; put the sample carbon column into the combustion tube, mineralize at 950 C, AOI is converted into HI gas; HI gas is dehydrated by concentrated sulfuric acid, and detected by high-sensitivity microcoulometric titration. In order to ensure reliable measurement results, 2-iodophenol is used as the standard substance for calibration.

Formation of I-DBPs in real water
The formation of I-DBPs is closely related to the type of organic matter in the water. In the actual water body, the organic matter contained in it is relatively complex, including multiple functional groups: benzene rings, amino groups, carboxyl groups, etc. The type of organic matter may be different from the products formed by PDS oxidation, but the possible paths and mechanisms for functional groups are similar. To investigate I-DBPs formation during oxidation of PDS by Fe (II) in real water, I À was spiked into the river water from a reservoir from Shenzhen. Six typical iodinated organic compounds (bromochloroiodomethane, monobromodiiodomethane, dibromomonoiodomethane, monochlorodiiodomethane, dichloromonoiodomethane, iodoform) were monitored.

Effect of initial pH
The oxidation effect of PDS at different pH is different. At the same time, pH also affects the type of active iodine produced by PDS oxidizing iodine ions, and the type of active iodine is greatly different in the degree of difficulty of being reacted by the oxidant. The pH affects the ability of PDS to oxidize I À , the presence of Fe (II) in the experimental system, and the rate of PDS oxidation of active iodine. Therefore, the types and amounts of I-DBPs produced under different pH conditions will be quite different.
The change of formation of I À transformation with pH in the Fe (II)/PDS system is illustrated in Figure 2. The ratio of the Fe (II) transformation to PDS was 4:5 as pH was increased from 3 to 10. As the pH increases, the total amount of AOI gradually increases and then slowly decreases, IO À 3 gradually decreases as the pH increases, I À gradually increase as the pH increases and the concentration of IO À is barely detectable in the solution. For IO À 3 , 31% of the I À are converted to IO À 3 at pH 3. While in organic-free systems, the IO À 3 generation rate is 56% and gradually decreases at pH ¼ 4, 5, 6, and 7. Especially, there is basically no IO À 3 generated under alkaline conditions, probably due to the organic matter. When the system generates IO À 3 , it is quickly captured and cannot be oxidized. For I À , the conversion efficiency of I À gradually decreases with increasing pH. As the pH increases from 3 to 9, the remaining amount of I À rises from 57% to 98%.
Compared with the absence of organic matter, the conversion efficiency of I À is also greatly reduced at the same pH. This is because the organic substances in the system react with sulfate radicals generated by Fe (II)/PDS, and the free radicals participating in the oxidation reaction of I À decrease, so that the conversion rate of I À decreases.
No IO À was detected in the system with organic matter. This is because a large amount of organic matter causes a part of the IO À produced by the reaction to be captured and a part of it to be oxidized to IO À 3 . It can be seen from the iodine conservation curve that the total iodine amount of I À plus IO À 3 plus AOI is close to the total amount of iodine, basically reaching iodine conservation.   of 20 μM, and I À initial concentration of 10 μM. The ratio of the Fe (II) transformation to PDS was 4:5 at pH 3. When the concentration of the PDS is low, PDS can oxidize part of I À to IO À , and the oxidation of I À into IO À 3 reaction and the continued oxidation reaction exist at the same time, so some IO À 3 is generated in the system. However, due to the limited concentration of PDS, IO À that has not been oxidized to IO À 3 can quickly react with organic matter to form I-DBPs. The concentration of IO À 3 increases with the increase of the PDS concentration, and it stabilizes at 100 μM, indicating that the higher the PDS concentration, the more thorough the oxidation reaction that occurs, and the higher the I conversion rate, the saturation reaches 100 μM. For I À , the higher the PDS concentration, the lower the iodine ion concentration, indicating that more I À are oxidized. The sum of I À and IO À 3 concentration after 40 μM is almost unchanged, indicating that the PDS concentration has a greater influence on the first step of the iodine oxidation reaction, and has little effect on the conversion of IO À to I-DBPs. The law of AOI decreases with the increase of oxidant concentration, which is the largest at 60, indicating that under this condition, secondary iodine easily reacts with organic matter to form iodine byproducts. Figure 4 shows the effect of Fe (II)/PDS on I À transformation in the Fe (II)/PDS system, at a constant PDS dosage of 100 μM, I À initial concentration of 10 μM at pH 3. As the proportion of Fe (II) increases, the generation rate of IO À 3 gradually increases. When Fe 2þ /PDS is 1:5, 2:5, and 3:5, the conversion rates of IO À 3 are 4.7%, 6.5%, and 8.4, respectively. When Fe 2þ /PDS reaches 5:5, about 33% of I À are oxidized to IO À 3 . When Fe 2þ /PDS reaches 6:5, 36% of the I À are converted to IO À 3 , indicating that at Fe 2þ /PDS ratio greater than 4:5, a large amount of I À are converted into IO À in the experimental system, and then further converted into IO À 3 , and when Fe 2þ /PDS is small, the activation generates fewer sulfate radicals and the ability to oxidize I À is insufficient, so the generation rate of IO À 3 is small. The conversion rate of I À is similar to that of IO À 3 . It gradually increases with the increase of Fe 2þ /PDS, and then tends to be stable. It can be seen from the figure that when Fe 2þ /PDS is 6:5, the residual amount of iodide ion is the smallest at 48%, and the conversion rate is the largest.

Effect of initial Fe (II)/PDS molar ratio
When the Fe 2þ /PDS ratio is 1:5, the residual amount of iodide ion reaches 73%.
The AOI in the experimental system gradually increases with the increase of the proportion of ferrous iron and then decreases. When the Fe 2þ /PDS ratio is 3:5, the AOI generation rate is 30% at the maximum, and the Fe 2þ /PDS ratio  is 4:5, 5:5 and 6:5. The generation rates of AOI are 15.9%, 15.7% and 14.0%.

Effect of NOM
The effect of NOM on I À transformation in the Fe(II)/PDS system was examined in the presence of commercial humic acid (HA; used as a representative NOM), at pH 3. As can be seen in Figure 5, as the concentration of HA increases, the AOI in the system gradually increases.
When the HA concentration reaches 1 mg/L, the AOI reaches the maximum. When the HA concentration in the experimental system is increased again, AOI concentration remains basically unchanged. The amount of IO À 3 produced gradually decreases with increasing HA concentration. There are two main reasons for this: (i) the reaction of IO À produced in the system with HA inhibits the further conversion of IO À to IO À 3 ; (ii) both HA and I À in the system can react with PDS, resulting in a reduction in the oxidation rate of I À .
From this we can conclude that the total organic carbon of the water body affects the generation of I-DBPs in the water. It is not that the higher the TOC, the easier it is for the water body to produce I-DBPs. We can also control I-DBPs by controlling the total organic carbon product formation.

CONCLUSIONS
This study investigated the Fe (II) activated PDS process under different pH, Fe (II)/PDS, PDS dosage and NOM conditions. Based on the experimental results obtained, the following conclusions can be obtained: (1) The experiment results verified the generation of AOI in the case of adding NOM to the Fe (II)/PDS oxidation system. It was found that under acidic conditions, as the AOI formation rate is larger, I À production is more.
(2) Under alkaline conditions, although the AOI generation rate is small, the conversion rate of iodide ions is small.
When the concentration of oxidant in the experimental system is increased, the AOI generation amount gradually increases and then decreases.
(3) The AOI generation increases with the proportion of Fe (II). When the ratio is 3:5, the generation rate of AOI is at most 30%. The generation of AOI gradually increases with the increase of HA concentration, and when the concentration is 1 mg/L, the AOI concentration is the largest.

DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.