Electrochemical conversion pathways and existing morphology of arsenic ( III ) in anode-cathode separated electrolytic cells

To explore the electrochemical conversion of arsenic at different voltages and pH, an open separated electrolytic cell with a platinum anode and a graphite cathode was selected for this paper. The form and concentration of arsenic in the anodic cell and cathodic cell were detected. Experimental results proved that at 40.0 V, As(III) in an acid electrolyte in the cathodic cell was firstly mainly reduced to AsH3 with trace As(0) as intermediate. As the electrolysis time arrived at 27 min, pH in the cathodic cell jumped suddenly from acidity to alkalinity, accompanied by the majority of the remaining As(III) converting to As(V) for an instant. As time went on, As(III) and As(V) remained almost unchanged at the ratio of 1:3, and the reduction of As(III) became extremely weak in the alkaline environment. When pH in the cathodic tank was adjusted to keep it acid, As(III) was eventually converted to AsH3. Compared with high voltage, at a low voltage of 1.0 V the cathode failed to achieve the potential of As(III) reduction and As(III) was eventually oxidized to As(V) in the acid catholyte. Electrochemical oxidation of As(III) in the open cathodic cell was likely caused by in-situ generation of peroxide from electrochemical reduction of O2. Theoretical support for electrochemical oxidation of As(III) on a carbon cathode in neutral and weak alkaline media is provided in this study.


GRAPHICAL ABSTRACT INTRODUCTION
Arsenic (As) contamination is widely recognized as a global health problem. The distribution of As(III) and As(V) in natural water depends on the redox conditions and pH of water, and As(III) is prevalent in anoxic groundwater.
Compared with As(V), As(III) is reported to have low affinity to the surface of various minerals and be difficult to remove, because it mainly exists as nonionic H 3 AsO 3 in natural water with pH <9. Nevertheless, As(V) can be easily adsorbed on solid surfaces and removed easily.
Since As(III) is more toxic (Mandal & Suzuki ), more mobile and more difficult for removal than As(V) (Raychoudhury et  electro-Fenton pro-oxidation of As(III) into As(V) and electrochemical reduction of As(III) into As(0) (Brusciotti & Duby ) are promising and convenient methods to pretreat As(III) from aqueous solution. Compared with surface water, remediation of polluted groundwater is extremely difficult and requires large financial resources.
Electrochemical technology is very suitable for remediation of arsenic contaminated groundwater, owning to the advantages of in-situ remediation, no need to add chemicals to the groundwater, no pollution and high efficiency (Lacasa et al. ). So, it is important to investigate the existing form and transform mechanism of As(III) during electrochemical experiments.
Various researchers have investigated the electrochemical reaction of As(III) using different electrodes at negative potentials in acid and basic media through potentiostatic methods (Dewalens et al. ). In early studies, As(III) at negative potentials was found to be reduced to As(0) and/ or AsH 3 , and the yields of As(0) and AsH 3 at negative potentials were significantly determined by the negative potentials, pH and cathode materials (Bejan & Bunce ; Cao et al. ). Low pH benefited the deposition of As(0) and the evolution of arsine (Smirnov et al. ).
Recently, the partial oxidation of As(III) at negative bias potentials has been reported. Ten per cent of As(III) was observed to transform into As(V) on TiO 2 electrode at negative bias potentials in an air saturated solution, and superoxide (O 2 À ) which was generated from the reduction of O 2 was speculated to be the only oxidant (Fei et al. ). Transformation of As(III) to As(V) was highly accelerated by the small quantities of H 2 O 2 produced from O 2 reduction under the alkaline conditions (pH 10.0-11.0) that were produced automatically around the cathodic elec-  The electrolysis experiments were conducted for 3, 6,9,12,15,18,21,24,27,30,35,40,45,50,55 and 60 minutes.
To determine As(III) and As(V) in the anolyte and the catholyte during the electrochemical process, 0.4 mL water samples were withdrawn by a 1-mL pipette at certain intervals, diluted and filtered through a 0.22-μm PTFE filter, then a high performance liquid chromatography-inductively coupled plasma mass spectrometer (HPLC-ICP-MS) was used to separate arsenic species and determine the concentration of As(III) and As(V) in the water samples with different reaction time.
To determine As(0) produced in the cathodic cell, four identical graphite rod electrodes were prepared and four sets of experiments were conducted for different times, specifically, the first experiment for 15 minutes, second for 30 minutes, third for 45 minutes, and fourth for 60 minutes.
Then electrodeposited As (0)  To measure pH of the anolyte and the catholyte during the electrochemical process, the electrode of a pH meter (Mettler Toledo, Switzerland) was inserted into the anolyte or the catholyte, and the pH values were read out at certain intervals.
To measure Eh of the anolyte and the catholyte at different reaction time, the electrode of an ORP (oxidationreduction potential) meter (Dapu, China) was inserted into the anolyte or the catholyte, the Eh data read out (namely the Eh at power-on) as the electrolysis was carried out for 3, 6,9,12,15,18,21,24,27,30,35,40,45,50,55 and 60 minutes respectively, then the power supply was immediately turned off and the Eh data read out at power-off, then the power was turned back on and continued electrolyzing As(III).

Analytical methods and calculations
A CS300 electrochemical workstation (Koster, China) with a standard three-electrode system was used to detect As(0) produced in the cathodic cell. One graphite rod electrode before and after use as cathode for the electrochemical conversion of As(III) served as a working electrode, the other graphite rod was used as a counter electrode, and a saturated calomel electrode (SCE) worked as the reference one ( Figure 1(b)).
The separation of arsenic species and the detection of The calculation formula of current efficiency is where η is current efficiency, m 0 is the actual product quality (g), m is mass of product (g) obtained according to Faraday's law, I is current intensity (A), t is power-on time (h), k is the electrochemical equivalent (g/(A·h)), F is Faraday constant (96,485 C/mol), M is molar mass of arsenic element (74.9216 g/mol), n is the gains and losses of electron.

RESULTS AND DISCUSSION
The transformation of As(III) at high constant voltage

As(III) transformation in anodic cell
The concentration of As(III) and As(V) in the anodic tank with different electrolytic time was determined, as shown in Figure 2(a). Before 24 min, As(III) concentration decreased rapidly, while As(V) increased. At 24 min, the transformation from As(III) to As(V) was stopped and only As(V) remained in solution. The conversion process of As(III) was proved to match a zero-order reaction kinetics equation: where, C t (mg/L) was the concentration of As(III), t(min) was electrolysis time.
With the electrolytic time increasing, pH in the anodic cell was decreased (Figure 2 (Figure 2(b)). The conversion from As(III) to As(V) in the anodic cell was likely electro-catalyzed by platinum hydroxide formed on the electrode surface (Equations Pt(OH) þ As(OH) 3 ! Pt(OH)As(OH) 3 (4) PtOAs(OH) 3 ! Pt þ OAs(OH) 3 The current with different electrolytic time at 40.0 V is shown in Figure 3(a). As 40.0 V voltage was applied to electrochemical transformation of As(III), the current value which was shown on display screen of the constant voltage DC source was 110 mA at the beginning, and then the current increased with electrolysis time. As the electrolysis time reached 60 min, the current increased to 210 mA. In the process of constant voltage electrolysis, pH in the anodic cell decreased even lower, and pH in the cathodic cell increased to strong alkalinity, which improved the conductivity of the electrolyte and caused the current to increase. The relationship between the transformation percentage of As(III) into As(V) and the current efficiency in the anodic cell is shown in Figure 3(b). As the electrolysis time was between 12 to 15 min, the current efficiency reached the maximum of 2.3%.

As(III) transformation in cathodic cell
The concentration of As(III) and As(V) in the cathodic cell with 40.0 V voltage at different electrolytic time is shown in Figure 4(a) and 4(b). Before 27 min, As(III) concentration was slowly decreased from 40.453 mg/L to 29.823 mg/L, mainly converted into AsH 3 with trace As(0) (Figure 4(a)).
And this reduction process was in accord with the zeroorder kinetics equation: C t ¼ À0:405t þ 41:341 R 2 ¼ 0:968 (7) where C t (mg/L) was the concentration of As(III), t (min) was electrolysis time.
After 27 min, As(V) was presented in the cathodic cell suddenly, and As(III) decreased rapidly from 19.968 mg/L at 28 minutes to 8.259 mg/L at 30 minutes. After 30 min, The relationship between the transformation percentage of As(III) into As(V) and the current efficiency in anodic cell.

As(III) was no longer transformed to As(V), As(III) and
As(V) existed dynamically at 1:3 ratio, and total concentration of As(III) and As(V) in solution remained almost unchanged; however, the reduction of As(III) was weakened extremely.
Meanwhile, in the cathodic cell pH jumped from acidity to alkalinity between 27 minutes and 30 minutes  Figure 4(d).
Surprisingly, it was found that the Eh of the catholyte at pH jump from 5 to 12 was higher than the stable fields of As(III)/As(V), which indicated that As(III) should be   (8)), thus decreasing the energy cost for water detoxification.
It has been demonstrated that the O 2 reduction reaction occurs via two pathways, which are 'direct' four-electron reduction and 'series' reduction.
The 'direct' pathway is a one-step four-electron reduction, as illustrated in Equations (9) and (10) In contrast, during the 'series' reduction process, O 2 is first reduced to ·O 2 À , as shown in Equation (11), the reaction is followed by further reduction step to H 2 O 2 (in acidic media, Equation (12)) or HO 2 -(in basic media, Equation (15)), then reduction to H 2 O (Equations (13) and (16)) or disproportionation (Equations (14) and (17)). H 2 O 2 easily transforms into HO 2 in alkaline media according to its pKa value of 11.7.
In acidic media: In basic media:  In the cathodic cell, H 2 evolution, O 2 reduction, NO 3 À reduction (Equations (18) and (19)), As(III) reduction and As(III) oxidation were the main reactions. Before 27 min, the solution still remained strongly acid, As(III) was converted to As(0) and AsH 3 at a rate of 0.405 mg L À1 min -1 ; it was reported that the reduction of O 2 to H 2 O 2 at pH <3 was relatively weak, which could be associated with two side reactions on the carbon cathode, namely, the reduction of H 2 O 2 to H 2 O and the production of H 2 (Soltani et al.

).
As a consequence of low yield of H 2 O 2 in acid electrolyte, As(III) oxidation to As(V) by H 2 O 2 was so weak that it could not be detected before 27 min. The distribution of arsenic species depends on the redox conditions and solution pH; As(III) is easily reduced in acid environment and easily oxidized in alkaline environment (Sharma & Sohn ). Between 27 minutes and 30 minutes pH in the cathodic cell jumped from 3.49 to 10.59 (Figure 4(c)), correspondingly the Eh significantly dropped, the reduction rate of As(III) sharply dropped to a very low level. As pH jumped from acidity to alkalinity, the production rate of H 2 O 2 on the graphite increased dramatically, furthermore, H 2 O 2 might be converted to HO 2 -, HO· and ·O 2 in basic electrolyte (Equations (20) and (21) in the catholyte was abruptly oxidized to As(V), which was similar to the result of Luo et al. () who found that an alkaline medium was more favorable to the efficiency of phenol degradation than an acid medium with carbon black as cathode.
After 30 min, As(III) did not convert to As(V) any more, and the ratio of As(III) to As(V) remained 1:3 all the time (Figure 4(a)). This could be ascribed to the significant decomposition of H 2 O 2 in alkaline solution (Wang & Wang ). It was reported that H 2 O 2 was relatively stable at pH <9. However, above pH 9, the decomposition rate of H 2 O 2 increased with growing pH, temperature (especially more than 23 C) and reaction time. The self-decomposition of H 2 O 2 at high pH was attributed to two factors: one factor was the catalytic effect of the container walls and the reagent impurities, and the other was that the anion, HO 2 -, played an important role in the base catalyzed decomposition of H 2 O 2 (Equation (22)) (Qiang et al. ).
A voltammogram analysis of the cathode electrode (graphite rod electrode) before and after use for the electrochemical conversion of As(III) for 60 minutes is shown in Figure 5. The oxidation peak from As(0) to As(III) appeared at 0.25 V, which indicated that As(0) was produced from the reduction of As(III) and deposited on the graphite cathode in the cathodic cell.
As the electrolysis time was 60 minutes, the yield of AsH 3 which was guided into the AgNO 3 solution and detected by the spectrophotometer was 5.999 mg/L (Figure 4(b)).
Within the first 30 minutes the yield of AsH 3 was 5.060 mg/L, then AsH 3 was only increased by 0.939 mg/L within the next 30 minutes. As (0) (25)) might be bonded covalently to As atoms (Equation (24)) to produce AsH 3 (Equation (26)). H 3 AsO 3 þ 3e À ! As þ 3OH À (deposited on the cathode) As þ 3H* ! AsH 3 ↑ The relationship between the generation percentage of As(0), AsH 3 or As(V) which were all from electrochemical conversion of As(III) and the current efficiency in cathodic cell is shown in Figure 6. As the electrolysis time was between 27 and 30 min, pH in the cathodic cell suddenly jumped from acidity to alkalinity, accompanied with the majority of the remaining As(III) converted to As(V) for an instant, and the current efficiency of As(III) into As(V) reached a maximum of 15.8%.
The As(III) transformation in cathodic cell with pH less than 1.5 The As(III) transformation in the cathodic cell with pH adjusted below 1.5 is shown in Figure 7. The Eh measured was À150 mV, As(III) was mainly converted to arsine gradually, with trace As(0) as intermediate products; after 110 min, As(III) no longer existed, and no As(V) was detected during the whole process. As(III) was oxidized with more difficulty in acid solution than that in basic media (Sharma & Sohn ), a high proton concentration in acid solution promoted H 2 evolution and reduced the current efficiency (Qiang et al. ), and the reduction of oxygen to H 2 O 2 was relatively weak in acid solution compared with that in basic solution (Leng et al. ). As a result, the conversion rate from As(III) to As(V) by H 2 O 2 was very low and not detectable in the solution with quite high initial As(III) concentration of 40 mg/L.

The transformation of As(III) at low constant voltage
To contrast the conversion pathways of As(III) in cathodic cell at high and low voltage, 1.0 V of low voltage was imposed to electrolyze As(III) in an anode-cathode divided electrolytic tank. 100 ml of 0.1 M HNO 3 solution containing 8 mg/L As(III) (pH 1.12, Eh 512 mV) was added to the anodic cell and the cathodic cell respectively.
As(III) was completely oxidized into As(V) in the anodic cell on the 8th day (Figure 8(a)). In the cathodic cell As(III) was completely converted into As(V) on the 37th day (Figure 8  higher than that at high voltage,which caused As(III) to be slowly oxidized into As(V) in the catholyte with 8 mg/L total arsenic. The Eh of the original As(III) solution was 512 mV. As 1.0 V voltage was imposed, the Eh of the anolyte gradually rose at first, then remain unchanged at 1,100 mV after 8 d; the Eh in the cathodic cell also increased at the beginning, then kept the same at 600 mV after 12 d. The generation of H 2 O 2 with low yield from O 2 reduction reaction in the pH range from 1.12 to 1.37 might cause higher Eh value in the catholyte than that in the original As(III) solution.

CONCLUSIONS
The electrochemical conversion process and form of As(III) at different voltages and pH was explored in a separated oxidation and reduction environment which was provided by a bi-electrolytic tank separated by a salt bridge. In the anodic compartment, 30 ml of 40 mg/L As(III) on Pt electrode at 40.0 V voltage was completely oxidized into As(V) before 24 min at the speed of 1.558 mg L -1 min -1 by HPLC-ICP-MS. In the cathodic compartment, As(III) was mainly reduced to arsine with As(0) as intermediate on a graphite cathode at the speed of 0.405 mg L À1 min À1 before 27 min. Between 27 minutes and 30 minutes, pH in the cathodic cell jumped from acidity to alkalinity, three quarters of the remaining As(III) in the cathodic cell was transformed into As(V) suddenly. After 30 min, As(III) was not converted to As(V) in the alkaline catholyte, meanwhile the reduction of As(III) in the alkaline catholyte became extremely weak. As pH in cathodic tank was adjusted to keep it acid, As (III) was eventually converted to arsine.
Compared with 40.0 V of high voltage, at 1.0 V the cathode failed to achieve the potential of As(III) reduction and As(III) was completely oxidized to As(V) very slowly in the catholyte in the pH range from 1.12 to 1.37. An Eh-pH track diagram of an arsenic containing catholyte on a graphite cathode was produced.