Generation of active Co(III) and peroxodiphosphate by synergistic electrocatalytic system with phosphate and themediatorcobalt(II) and its degradationperformance

The promising synergistic electrocatalytic system of phosphate (PO4 3 ) with the mediator cobalt(II) (for short E-Co(II)-PO4 3 ) was employed to degrade cationic dye methylene blue (MB). The exploration in the electrocatalytic process revealed that the main intermediate active oxidation products were Co(III), accompanied with hydroxyl radicals and peroxodiphosphates (P2O8 4 ). Their synergistic electrocatalytic degradation rate to MB and total organic carbon (TOC) was up to 100 and 60% in 40 min, respectively, which was 5 times and 2.6 times that in a direct electrocatalytic system, correspondingly. The degradation process of the E-Co(II)-PO4 3 system on MB started with the bond being broken at the N-C junction of the MB molecule and intermediate active oxidation substances being generated, such as phenothiazine, 2-amino-5-(N-methylformamide) benzene sulfonic acid and N1,N1-dimethyl-1,4 diaminobenzene. Then, the intermediates were degraded into aniline, phenol and benzene sulfonic acid, and eventually decomposed into inorganic substances like CO2 and water. The electrocatalytic degradation mechanism of E-Co(II)-PO4 3 system on MB was the combination of indirect oxidation of the intermediate oxidants like Co(III), P2O8 4 and the hydroxyl radical with direct electrocatalysis on the platinum titanium electrode, where the electrocatalytic oxidation of Co(III)


GRAPHICAL ABSTRACT INTRODUCTION
Most of the organic pollutants (especially colored dye pollutants) in industrial wastewater are difficult to degrade, and it is hard to meet the standards through conventional treatment processes or their enhancement (Brillas ). Colored dyes are mutagenic, carcinogenic, and allergic (Dassanayake et al. ). Studies have shown that even wastewater containing only small amounts of dyes could seriously affect the growth of aquatic organisms (Sekar et al. ). Among them, most dyeing industries prefer to use cationic dyes. However, cationic dyes are more toxic than anionic dyes because of their aromatic ring structure of delocalized electron and special synthetic methods (Doltabadi et al. ). Therefore, how to efficiently degrade cationic dyes (such as rhodamine B (RhB), methylene violet (MV) and methylene blue (MB)) with low cost and high efficiency in industrial wastewater has become a challengeable issue (Cui et al. ).
There are several methods for the treatment of dye wastewater: ion exchange, flocculating sedimentation (Doltabadi et al. ), advanced oxidation technology (Dassanayake et al. ; Nidheesh et al. ), activated carbon adsorption (Doltabadi et al. ), and biodegradation (Dhall et al. ). Some of these physical technologies (such as activated carbon adsorption and flocculating sedimentation) have significant treatment efficacies, but they only transfer organic dyes from the liquid phase to solid for separation, which cause serious secondary pollution. The biological treatment method is not ideal for the treatment of dye wastewater, which has certain toxicity and is difficult to degrade. Advanced oxidation technology has drawn wide attention because some strong oxidizing active groups (such as hydroxyl radicals, sulfate radicals, etc) can oxidize and decompose organic pollutants to generate small molecules such as water and carbon dioxide (Nidheesh et al. ). Among them, the ozonation and Fenton methods have significant effects on the degradation of pollutants. However, they require continuously adding oxidants to the wastewater, which increases the treatment cost. At the same time, a large amount of sludge is produced in the Fenton oxidation process; the initial investment in ozonation is large and more undegradable substances may be generated during the degradation process (Wang et al. ). Although photocatalytic technologies possess great advantages in energy utilization, the chromaticity of dye wastewater is large, which prevents the light source entering; at the same time, ultraviolet photocatalytic energy consumption is huge. Visible light catalytic efficiency is low, which is hardly satisfied in engineering application (Dassanayake et al. ). All these have limited the application of photocatalytic technology in the field of printing and dyeing wastewater treatment.
Nowadays, electrochemical technologies have been increasingly used in the treatment of wastewater containing refractory organics because of their high degradation performance. Among these electrochemical technologies, electrocoagulation and electrochemical oxidation are commonly adopted. Electrocoagulation is a widely spread method that uses iron and aluminum electrodes to conduct the coagulation reaction and directly generate coagulants via anodic dissolution. Recent research has focused on enhancing the oxidation capabilities of iron-based electrocoagulation by adding different oxidants; for example, adding potassium ferrate(VI) (Elnakar & Buchanan ). Iron-based electrocoagulation has the advantages of the abundance and low price of iron and easy operation, while its disadvantage is the high proportion of total iron dissolved under acid or neutral condition.
The electrochemical oxidation is one of the advanced oxidation technologies. There are noteworthy advantages of electrocatalytic degradation on dyes, such as small space for the electrocatalytic device, low investment, easy operation with automation, wide application scope and high degradation efficiency. There are two types of electrocatalytic oxidation ways, direct oxidation and indirect. The direct electrochemical catalysis decomposes pollutants in water to achieve the degradation effect by anodic oxidation. In order to replace the expensive metal platinum electrode and boron doped diamond (BDD) electrode, with high electrochemical oxidation efficacy, the research in this area mainly focuses on electrode modification; for example, the modification of the titanium-based metal oxide electrode (Wang et al. ), carbon nanotube electrode (Yi et al. ) and three-dimensional electrode (Ghanbarlou et al. ). The issues for electrode modification mainly concern high oxidation efficiency, good stability, high oxygen evolution overpotential and low manufacturing cost. Indirect oxidation is also called mediated electrochemical oxidation (MEO), which is adopted to oxidize target pollutants by adding mediated ions in water. The main mechanism is that the mediated ions are oxidized at the anode to produce strongly oxidizing ions. In the process of oxidizing pollutants, the content of the mediated ions is not changed after redox (Govindan & Moon ) and favored by researchers (Chung & Park ; Costa et al. ; Soleymani & Moradi ). The common used mediated ions are silver ion (Ag þ ), cerium ion (Ce 3þ ), cobalt ion (Co(II)), phosphate (PO 4 3À ) and sulfate (SO 4 2À ). Among them, the reaction of Ag þ with chloride ions in water occurs easily and produces precipitates; Ce 3þ owns lower solubility at normal pH; PO 4 3À and SO 4 2À react slowly with the electrodes. These ions are difficult to employ to establish an efficient electrochemical oxidation system. However, Co(II) is highly soluble in water, highly oxidizing, and reacts fast with electrodes (Chung & Park ). Therefore, the synergistic electrocatalysis reaction system with PO 4 3À and Co(II) was selected in this study. Electrochemical oxidation is also called electrocatalysis because it only consumes electrical energy in the oxidation process. MEO can improve the energy utilization rate of direct electrocatalysis to a certain extent. Studies have shown that the MEO system could efficiently degrade a variety of refractory organic pollutants in strong acidic conditions (like concentrated sulfuric acid, concentrated nitric acid, and so on) (Lubis et al. ). However, the MEO system in conventional conditions has rarely been reported, which is not conducive for it to be popularized and put into use. Therefore, this study established a promising synergistic electrocatalysis reaction system with PO 4 3À and Co(II) (for short E-Co(II)-PO 4 3À ) and explored its oxidation mechanism and degradation performance.

MATERIALS AND METHODS
The calibration curve of methylene blue (MB) solution MB standard solution was prepared at concentrations of 1 mg L À1 , 2 mg L À1 , 4 mg L À1 , 8 mg L À1 and 13 mg L À1 ; the corresponding absorbance at each concentration was measured at its maximum absorption wavelength (665 nm). Using the absorbance as the ordinate and the MB concentration as the abscissa, a calibration curve was depicted to clarify the linear equation of the correlation between the absorbance at 665 nm (noted as Y) and MB concentration (noted as X) for the solution is Y ¼ 0.1631X þ 0.021 with an R of 0.9993 (see Figure SI-1 in Supplementary Material).

Electrocatalysis experiment
The electrocatalysis experimental setup is shown in Figure 1. In the experiment, a stainless steel electrode clip was used to connect the cathode and anode, which were linked up to the direct current (DC) regulated power supply (MS305D, Maisheng Co., Ltd, 2018) by wire. The volume of the electrolytic cell was 700 mL, and the electrode interval was 2 cm. The experimental procedures are demonstrated below.
Firstly, the following experimental reagents were transferred to the electrolytic cell: 25 mM electrolyte sodium sulfate (Na 2 SO 4 ), 2 mM cobalt sulfate, 3 mM sodium phosphate (Na 3 PO4) and 12.5 mg L À1 MB solution (400 mL); the pH was adjusted to 4 with sulfuric acid (0.1 M) and sodium hydroxide (0.1 M) via pH meter (STARTER3100, OHAUS, USA, 2017), the magnetic stirrer (800 r min À1 , 79-1, Jintan Dadi Automation Instrument Factory, 2017) was started; the DC power supply was switched on and the current density adjusted to 12 mA cm À2 to perform the electrocatalytic experiments.
During the electrocatalytic tests, the MB dye solution was sampled at the reaction time intervals of 5 min, 10 min, 15 min, 20, 30 and 40 min, and the absorbance of the corresponding sample measured; then the MB residual concentration (C e ) in the solution was obtained according to the calibrated standard curve shown in Figure SI-1. Thus, the degradation rate was achieved from the calculation (1-C e )/ C 0 , where C 0 was the initial concentration of MB. The degradation performance at different electrocatalytic systems, like Co(II)-3 mM Na 2 SO 4 , E-3 mM Na 2 SO 4 , E-3 mM Na 2 PO 4 , E-Co(II), E-Co(II)-PO 4 3À , were tested for obtaining an optimal electrocatalytic system. And then, the effect of different experimental factors on the optimal synergistic electrocatalysis efficacy were explored. The experimental factors were as follows: initial concentrations of Co(II) and PO 4 3À , current density, initial pH, and electrolyte concentration. Each group of experiments was repeated three times, and the three experimental data were averaged to obtain the MB degradation rate.

Shielding experiment and capture test
There are some intermediate active oxidation substances existing in the liquid phase of the novel indirect electrocatalytic system. The effect of the intermediate active oxidation products on MB can be masked by introducing a shielding agent into the system. The change of degradation rate can indirectly reveal the existence of intermediate products. In addition, based on the behaviors of the active oxygen species reacting with some capture agents producing complexes, the active oxygen species can be manifested by detecting the corresponding complex by introducing the matching capture agent into the system. TBA and sodium thiosulfate (Na 2 S 2 O 3 ) were used for shielding experiments to screen the free radicals and all active oxidation substances in the solution. DMPO and ammonia were employed to capture free radicals and Co(III), respectively. Here, it should be noted that the concentration of the shielding agent should satisfy the condition that it would not affect the electrode direct electrocatalysis efficacy for MB.

Types of intermediate oxidation products in E-Co(II)-PO 4 3À system
It was confirmed that the intermediate oxidation products, like free radicals and Co(III), existed in the E-Co(II)-PO 4 3À system by detecting the change of MB oxidation removal rate after introducing shielding agents such as TBA and Na 2 S 2 O 3 into the system. The results are shown in Figure 2.

Free radicals
TBA was introduced into the system for detecting hydroxyl radicals and sulfate radicals (Govindan & Moon ). As shown in Figure 2(a), after adding 40 mmol of TBA, the degradation efficiency of the system on MB decreased slightly (from 53% to 45% in 15 min), which indicated that a small amount of free radicals were generated during the degradation of MB in the E-Co(II)-PO 4 3À system. The degradation rate of MB was almost the same when the dosage of TBA increased to 100 mmol, which meant the oxidation ability of the free radicals in this electrocatalytic system was weak. With EPR spectrometer (EMXplus-10/12, Bruker, Germany, 2018) and DMPO as the capture agent, the type of free radicals were determined further. A weak hydroxyl radical atlas appeared in Figure 2(b), which proved the hydroxyl radical (·OH) existed in the system. Determination of the strong intermediate oxidation products Co(III) and peroxodiphosphates (P 2 O 8 4À ) As a strong reducing agent, Na 2 S 2 O 3 is often used to detect the presence of oxidation substances in solution. By adding 4 mmol Na 2 S 2 O 3 to the E-Co(II)-PO 4 3À system, the degradation rate of the system on MB decreased significantly (from 100% to 20% in 40 min), which indirectly indicated that the electrocatalytic system produced intermediate products in the liquid phase that possessed a strong oxidative degradation effect on MB. It was concluded that the main intermediate oxidation product in the system was Co(III), and a small amount of hydroxyl radicals and P 2 O 8 4À were also present, which was consistent with the results reported in the literature (Costa et al. ; Lubis et al. ).

The detection of Co(III)
Since the residence time of Co(III) at conventional pH is extremely short, it is difficult to measure directly. However, Co(III) can be captured by ammonia molecules to form a variety of stable cobalt-amine complexes, which have the maximum absorption peak within the wavelength range of 320-520 nm (Okubo et al. ). The existence of Co(III) was verified with analysis on the characteristics of the solution by UV-Vis full-scan spectroscopy (UV-5100, Shanghai Yuanxi Instrument Co., Ltd, 2019) and introducing 25% of ammonia (1 mL) as the capture agent. The results shown in Figure 2(c) displayed that there were two absorption peaks at the wavelength 350 and 517 nm when the reaction time was at 5 min, 15 min, 30 min. These two absorption peaks revealed the existence of Co(III), which was consonant with the result reported by Okubo et al.  whether MB kept decomposing after the E-Co(II)-PO 4 3À system was switched off. When the power was off for 10 min, the absorbance of the MB solution decreased from 1.400 to 1.393, which showed that P 2 O 8 4À may present in the E-Co(II)-PO 4 3À system. However, the decrease of the absorbance was unconspicuous, which may be caused by the reaction between the P 2 O 8 4À and Co(II), so that the P 2 O 8 4À concentration in the system was decreased (Hariharan & Maruthamuthu ).
In order to eliminate the interference of sodium sulfate electrolyte, the electrolyte sodium phosphate was employed in the E-Co(II)-PO 4 3À system instead of sodium sulfate. KIstarch test paper was adopted to detect the presence of intermediate oxidation products in the reaction solution when the power was off. It was found that the KI-starch test paper was blue, which indicated that the oxidation substances still existed in the solution. Co(III) and hydroxyl radicals have extremely short residence time (Liu et al. ), which makes detecting their presence difficult. Hence, it was inferred that the intermediate oxidation product was P 2 O 8 4À .
Therefore, the main intermediate oxidation products existing in the E-Co(II)-PO 4 3À system were Co(III), and a small number of hydroxyl radicals and P 2 O 8 4À .

The comparison of degradation effects of several different catalytic systems on MB
The degradation effect of several different electrocatalytic systems on MB is shown in Figure 3. Under the condition of no electricity, when Co(II) and phosphate ions were acting on MB at the same time, the MB degradation did not happen within 50 min. After adding the electrolytes (25 mM Na 2 SO 4 and 3 mM Na 3 PO 4 ) with direct electrochemical oxidation on MB by platinum plating on a titanium electrode for 30 min, the degradation rate was only 18%. In order to eliminate the influence of the Na 3 PO 4 electrolyte, 3 mM Na 3 PO 4 was replaced with 3 mM Na 2 SO 4 ; then, the MB degradation rate was 20%. To take Na 2 PO 4 as the electrolyte, the degradation efficiency of direct electrocatalysis was a bit higher than that by electrolyte Na 3 SO 4 . However, for the E-Co(II) system, the degradation rate of MB was up to 76.3% at 30 min. After the combination of the electrolytes PO 4 3À with the E-Co(II) system, the degradation rate grew to 91% within 30 min, and even reached 100% when the reaction time lengthened to 40 min. Therefore, in the process of MB degradation, the current utilization rate of direct electrochemical oxidation was low, and its degradation rate was less than 20%; however, the E-Co(II) system brought a much higher degradation rate (76.3%) and an improved current utilization rate compared with that of the direct electrochemical oxidation system. Furthermore, when the electrolyte Na 3 PO 4 was coupled with E-Co(II) to form the E-Co(II)-PO 4 3À synergistic electrocatalytic system, the most exciting result was that the oxidative degradation rate on MB was up to 100% within 40 min, which was 5 times of that of the direct electrochemical oxidation system with electrolytes Na 3 SO 4 . At the same time, the degree of mineralization of MB was explored by testing total organic carbon (TOC) residual concentration (measured by TOC analyzer, TOC-3000, Shanghai Yuanxi Instrument Co., Ltd, 2015) in the E-Co(II)-PO 4 3À electrocatalytic system and then compared with that in the direct electrocatalysis system (shown in Figure 3(b)). It was clearly shown that the TOC degradation rate reached 60% in the E-Co(II)-PO 4 3À system, 2.6 times that in the direct electrocatalytic system (23%) for 2.5 h, which demonstrated that the mineralization efficiency of the former was higher.

Comparison of degradation efficacy of E-Co(II)-PO 4 3À system on MB under different conditions
The electrocatalytic oxidation experiments of MB were conducted in the E-Co(II)-PO 4 3À system, and the influence of different factors below on the degradation efficacy of the system is presented in Figure 4(a)-4(e). The main influential factors were the concentration of Co(II), PO 4 3À and electrolyte, current density and initial pH.

Influence of initial Co(II) concentration
The effect of initial Co(II) concentration on the MB degradation of the E-Co(II)-PO 4 3À electrocatalytic system is shown in Figure 4(a). From Figure 4(a), the oxidation ability of the E-Co(II)-PO 4 3À electrocatalytic system was enhanced significantly with the increase in Co(II) concentration. When the Co(II) concentration was set as 0 mM, 0.2 mM, 0.5 mM, 1, 2 and 4 mM, the MB degradation rate of the electrocatalytic system was 18%, 36%, 51%, 67%, 94% and 100% accordingly at 30 min. It was clear that the increasing of Co(II) concentration facilitated the yield per unit time of the strong oxidation substance -Co(III), which significantly enhanced the electrocatalytic performance of the E-Co(II)-PO 4 3À system (Lubis et al. ). At the same time, it could be seen that the degradation efficiency of this system was doubled at 0.2 mM of Co(II) compared to direct electrocatalysis. When the Co(II) concentration was up to 4 mM, the degradation efficacy of the E-Co(II)-PO 4 3À system was five times that of direct electrocatalysis. This phenomenon indicated that the cobalt ion dielectric system had an enhancing effect on the MB degradation rate and the current utilization efficiency.

Effect of initial PO 4 3À concentration
At different PO 4 3À concentrations, the degradation efficacy of the E-Co(II)-PO 4 3À system on MB is presented in Figure 4(b  (Hariharan & Maruthamuthu ), which enhanced the oxidation ability of the system. However, the increase in the PO 4 3À concentration caused the formation of more Co 3 (PO 4 ) 2 precipitates on the cathode, which reduced the Co(II) in the liquid phase, leading to a decrease in the oxidation efficiency of the system. These two possible reasons resulted in a slight effect of the increasing PO 4 3À concentration on the oxidation efficiency of the E-Co(II)-PO 4 3À system.

Effect of initial pH value
The effect of initial pH value on the MB degradation efficacy of E-Co(II)-PO 4 3À system is shown in Figure 4(c). The degradation efficiency of the system increased gradually with the decrease in pH value. When the pH was 4, it reached the optimum electrocatalytic efficacy. At the initial pH value 2, 4, 6, 7, 8 and 10, the electrocatalytic degradation rate on MB of the E-Co(II)-PO 4 3À system was 100%, 91%, 58%, 50%, 42% and 25% accordingly at 30 min. The results demonstrated that the degradation efficacy of the system gradually decreased with the growth in pH value. The main reason was that Co(II) existed in the form of Co(OH) 4 2À in alkaline solution, which reduced the oxidation performance of the system, while the acidic condition can inhibit side reactions (like the electrolysis of water) and the hydrolysis of Co(II). In addition, when the initial pH value was 4, 6, and 8, there were some black precipitates adhering to the cathode surface. This situation was even worse when pH was 6. In addition, the pH value of the cathode surface was higher than that of the solution in the electrocatalysis process, and a certain amount of Co(OH) x and Co 3 (PO 4 ) 2 precipitates were formed on the cathode surface (Chaplin ), which caused the decline in the degradation efficiency of the E-Co(II)-PO 4 3À system.

Effect of current density
Degradation efficacy of E-Co(II)-PO 4 3À system on MB at different current densities is shown in Figure 4(d). When the current density was set as 5 mA cm À2 , 10 mA cm À2 , 12 mA cm À2 , 14 mA cm À2 and 18 mA cm À2 , the degradation rate of the system was 32%, 57%, 67%, 76%, and 88% accordingly after 20 min of electrocatalytic reaction. The data showed that the increase in current density accelerated  the degradation of MB, and its current utilization rate showed a nonlinear decrease. This was because, as the current density increased, the following changes occurred in the system: first, the generation rate of Co(III) sped up, which improved the oxidation efficacy of the system; second, the direct electrocatalytic action of the electrode was enhanced; third, the side reactions (like the electrolysis of water) in the system was strengthened, which led to an increased energy consumption (see Table 1). The increasing current density of the electrodes improved the degradation efficacy of the E-Co(II)-PO 4 3À system, and more energy was consumed accordingly. Considering the effects of energy consumption and degradation efficacy together, 12 mA cm À2 was chosen to be the optimal current density in the E-Co(II)-PO 4 3À system.

Effect of Na 2 SO 4 electrolyte
The electrolyte plays the role of electron transfer in the electrocatalytic system, and can be oxidized at the anode to generate various oxidation substances, which serves as a certain role of the mediated electrochemical oxidation (de Araujo et al. ). The degradation rate of the E-Co(II)-PO 4 3À system on MB at different concentrations of Na 2 SO 4 electrolyte showed a rapid increase (the fastest condition was at 25 mM) within 30 min, and stabilized at a peak level (100%) at 50 min (the slowest condition was also at 25 mM). In a range of Na 2 SO 4 concentration of 10 mM ∼ 25 mM, the system voltage declined, the side reactions weakened, and the MB degradation rate rose with the increase in the Na 2 SO 4 concentration; however, the amount of precipitates settling on the cathode surface increased, and the concentration of Co(III) and PO 4 3À in the solution reduced. At the same time, the proportion of PO 4 3À in the total amount of ions decreased with the increase in SO 4 2À concentration.
Therefore, the contact rate of Co(III) and PO 4 3À with the anode went down, the yield of Co(III) and P 2 O 8 4À dropped, and thus the oxidation efficacy of the system reduced. Thus, on the basis of the two possible reasons mentioned above, the degradation efficacy of the E-Co(II)-PO 4 3À system on MB increased dramatically then slowly tended to become stable with the increase in the Na 2 SO 4 concentration at a lower level (10 mM ∼ 25 mM). Similarly, when the Na 2 SO 4 concentration was higher (25 mM ∼ 75 mM), the effect of the increase in the electrolyte concentration on the oxidation efficiency of the E-Co(II)-PO 4 3À system was inconspicuous.
Degradation mechanism of E-Co(II)-PO 4 3À system on MB

Analysis of intermediate products in MB degradation process
The intermediate products being generated in the MB degradation process (measured by liquid mass spectrometer, Q Exactive, China, 2018) are shown in Figure 5. Samples were taken at the degradation times of 10, 30 and 60 min (noted as sample 1, sample 2 and sample 3, respectively . Then the intermediates were degraded to aniline, phenol and benzene sulfonic acid, and eventually was decomposed into inorganic substances such as CO 2 and water.  Figure 7. From Figure 7, there were two types of electrocatalysis degradation mechanisms happening in the E-Co(II)-PO 4 3À system. One was direct electrocatalytic action on platinumtitanium electrodes. Organic pollutants, Co(II) and PO 4 3À were anodized into the forms of small molecule organics, Co(III) and P 2 O 8 4À , respectively. Another mechanism was indirect electrocatalysis of Co(III), P 2 O 8 4À and hydroxyl radicals generated in the electrocatalytic process. The electrocatalytic oxidation ability of Co(III) was dominant in the E-Co(II)-PO 4 3À system. On the one hand, organic matter (such as MB) in the solution was swiftly decomposed into small molecules (such as phenothiazine) because of the  satisfy the chemical equilibrium. Therefore, the coupling system of PO 4 3À and E-Co(II) had a significant synergistic catalytic degradation efficacy on MB. In addition, a small number of hydroxyl radicals also oxidized the organic matter (such as MB) in the system (Figure 2(b)).
Therefore, there were two degradation mechanisms of the E-Co(II)-PO 4 3À system on MB; one was the direct electrocatalysis on the electrodes, and the other was the indirect oxidation of Co(III), P 2 O 8 4À , and hydroxyl radicals. Among them, the electrocatalytic oxidation of Co(III) was in dominion.

CONCLUSION
In this study, a novel synergistic electrocatalytic system (E-Co(II)-PO 4 3À ) was proposed and its degradation  performance was tested by adopting MB as a target pollutant. The removal rate of the E-Co(II)-PO 4 3À system on MB was as high as 100% in 40 min, which was 1.2 times that in E-Co(II) system and 5 times that in direct electrochemical oxidation. The TOC removal rate reached 60% in the E-Co(II)-PO 4 3À system, while that of the direct electrocatalytic system was only 23% correspondingly.
In the E-Co(II)-PO 4 3À system, the effect of the experimental factors on the MB removal rate was quite different. The three factors, like Co(II) concentration, current density and pH value had much stronger impact on degradation performance than these two, like the concentration of PO 4 3À and electrolyte. The main intermediate active oxidation products in the E-Co(II)-PO 4 3À system were Co(III), accompanied with a small amount of hydroxyl radicals and P 2 O 8 4À . With direct electrocatalysis on the platinum titanium electrode and indirect oxidation of these intermediate oxidation products, MB was eventually degraded into inorganic substances like CO 2 and water. Among them, the electrocatalytic oxidation of Co(III) played a leading role.