Synergistic Effect of Mediated Electrochemical Reduction and Mediated Electrochemical Oxidation on NO Removal by Electro-Scrubbing

A combined MER (mediated electrochemical reduction) and MEO (mediated electrochemical oxidation) approach was examined for the first time for the efficient removal of NO by electro-scrubbing. The generation of a mediator (Ni(I) from Ni(II)(CN)42− in 9 M KOH) by electrochemical reduction was identified by the changes in the ORP value and potentiometric titration. The Ni(I) formation found to be 9% (4.3 mM) in 2 h electrolysis, which is decreased to 1.83 mM (4%) during addition of NO demonstrates NO removal occurred. The MER of NO at the cathodic half-cell by electro-scrubbing revealed the formation of NH3 and N2. The reductive removal efficiency of NO was almost 100% up to a gas flow rate of 0.25 g min−1. At the anodic half-cell, the Co(III) mediator from Co(II)SO4 in 5 M H2SO4 was generated electrochemically with concomitant conversion of NO to NO3− in the solution phase with a removal efficiency of 28%. The combination of MEO and MER with electro-scrubbing demonstrates 97% removal efficiency of NO up to 0.5 g min−1. The combined removal approach is more suitable for high concentrated NO with energy efficiency of 0.99 g/kWh compared to the individual MEO (0.221 g/kWh) and MER (0.769 g/kWh) approaches. © The Author(s) 2016. Published by ECS. This is an open access article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY, http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse of the work in any medium, provided the original work is properly cited. [DOI: 10.1149/2.0791614jes] All rights reserved.

Abundantly generated nitric oxide (NO) gas that includes power plant and steel plant etc., is decomposed by many technologies, such as wet scrubbing, catalytic oxidation, and selective catalytic reduction (SCR). [1][2][3] The reduction of NO leads a combination of cationic intermediates and products, N 2 O, NH 2 OH, N 2 , and NH 3 , in the solution phase. [4][5][6] The reduction products varied depending on the catalyst and experimental conditions: on TiO 2 in the colloid solutions to produce NH 3 ; 4 in heterogeneous catalytic reduction process with NH 3 as the reactant produced N 2 ; 7,8 and N 2 O formed on boron-doped graphene method. 9 At the same time, the oxidation of NO leads to the anionic products of NO 2 and NO 3 − . 10,11 Heterogeneous catalytic oxidation at the gas phase produced NO 2 , 12 but solution phase oxidation by ozone and H 2 O 2 produced the NO 3 − product. 1,13 Because of the catalytic reaction, NO is solubilized in an aqueous solution, otherwise it less soluble or insoluble in an aqueous solution. 14 Using this approach, many oxidant- 15,16 and reductant-17 containing wet-scrubbing processes have been adopted in NO removal. Note that the above methods require a continuous catalyst or reactant feed to make the process sustainable.
An electron considered to be a green catalyst and can be feed continuously to regenerate the spent catalyst sustainably. Using this approach, electrogenerated Ce(IV) and Mn(III) mediators have been used to oxidize NOx to NO 3 − in HNO 3 medium. 18,19 In a similar method, electrogenerated Ag(II) with a wet scrubber in the name of 'electro-scrubbing' have used NOx oxidation to NO 2 in a single stage scrubber and NO 3 − in a two stage scrubber column. 20 In the presence of SO 2 in NO oxidation using the electrogenerated Ag(II) mediator in highly concentrated nitric acid have shown NO 2 with 60 to 80% of NO 3 − formation. [21][22][23] Note that unless an additional stage scrubber column is used, the NO becomes NO 2 , which means NO becomes a greenhouse gas (NO 2 ). On the other hand, reduction produces nongreenhouse gas, such as NH 3 or N 2 . In electrochemical reduction method, the direct reduction of NO on a Pt electrode revealed N 2 O and NH 3 depending on the crystal phase and conditions. 24,25 NO reduction at the Au electrode and semiconductor electrode ZnO and In 2 O 3 produced N 2 as a product. 26,27 To minimize the over potential and electrode fouling, an indirect reduction or mediated electrochemical reduction (MER) method was developed and under this method, NO reduction was initiated using Co(III)(TMPyP)(H 2 O) 2 ] 5+ (cobalt tetrakis(N-methyl-2-pyridyl)porphine) by CV (cyclic voltammetry) and ESR (electron spin resonance) analysis 28 and they found a restric-tion of the pH due to dissociation of the mediator. A packed bubble column with Fe(II)(EDTA) was used to absorb NO, which was then reduced electrochemically by a dithionite mediator (in-situ oxidant) generated in the batch reactor in 0.1 M Na 2 SO 4 at pH 5.5, where found NH 3 and amidosulfonic acid are the product. 29 However, the effective reductive removal of NO using suitable mediator and electrolyte was not fully elucidated.
Therefore, Ni(II)(CN) 4 2− was taken for homogeneous mediator generation in a cathodic half-cell in 9 M KOH (its high equilibrium constant value (2 × 10 31 ) facile high pH restriction 30 ) electrolyte with scrubber column for the MER of NO. Subsequently, electrolytic generation and quantification of electrogenerated Ni(I) was optimized and its MER of NO with wet scrubbing process was analyzed. Similar to the cathodic half-cell setup, an anodic half-cell setup was used with electrogenerated Co(III) in 5 M H 2 SO 4 . The removal efficiency and reduction products were monitored using a real time online gas FTIR analyzer. A combination of MEO and MER of NO was performed using the sequential feed of NO by electro-scrubbing. The efficiency of NO removal was monitored by the different gas flow rate toward industrial applications. In addition, GC was used for gas phase N 2 analysis, and the solution phase products were analyzed by nitrate detection and solution phase FTIR to confirm the products and propose a reaction pathway.

Experimental
Chemicals.-K 3 [Fe(III)(CN) 6 ] was purchased from Aldrich and used without further purification. KOH (99.8%), Co(II)SO 4 , and KMnO 4 were obtained from Junsei Chemical Co., Ltd. (Japan). Sulfuric acid (95%) was supplied by Samchun Chemical Co., Ltd. (South Korea). Gas cylinders of NO (concentration 99.5%) and Ar (μmol/mol) were supplied by Inter Gas, Korea. The silver mesh electrode was purchased from 4scientific, USA. The Ti mesh and Pt-coated Ti mesh electrodes were purchased from Wesco Electrodes and Systems (South Korea). All solutions were prepared using reverse osmosis purified water (Human Power III plus, South Korea) with a resistivity of 18 M -cm. 4 2− preparation.-Ni(II)(CN) 4 2− was prepared using the procedure reported elsewhere. 31   was cooled slowly until a mass of thin orange platelets appeared. The resulting complex was filtered rapidly, washed with cold alcohol, recrystallized in ethanol, dried in a vacuum desiccator, and stored in an air-tight brown bottle.

Ni(I) and Co(III) mediators generation and the electro-scrubbing
setup.-The electrolytic cell employed had a flow-through divided cell configuration, as reported elsewhere, 32 with an additional scrubber column setup, as shown in Figure 1. The cell had cathode and anode areas of 50 cm 2 and an electrolyte volume of 500 ml. A scrubber column (40 cm high and 5.5 cm (i.d)) packed with 1 cm 2 of the Teflon tubes as packing material was attached to the top of the each anolyte and catholyte tanks, which is already attached with the flow through electrolytic divided cell. The scrubbing system was composed of an air supply system, a scrubbing solution (Ni(II)/Ni(I) in KOH solution), a scrubbing reactor column, and an FTIR gas analyzer system (MIDAC Corporation, USA) equipped with a data logger. The Ni(I) and Co(III) formation experiments were conducted using a catholyte containing 9 M KOH and 0.05 M Ni(II)(CN) 4 2− in a 1 L glass tank and an anolyte containing 5 M H 2 SO 4 and 0.05 M Co(II)SO 4 in another 1 L glass tank, which were both connected to a flow through electrolytic divided (by Nafion 324 from DuPont, USA) cell membrane. The anolyte and catholyte were circulated continuously using magnetic pumps (Pan, World Co., Ltd, Taiwan) through the anode and cathode compartments, respectively, at a rate of 2 L min −1 . The electrolysis experiments were conducted using constant current mode by a constant current density of 25 mA cm −2 (by a D.C. power supply from Korea Switching Instruments).
During the electro-scrubbing experiments, the anolyte and catholyte solution were pumped separately using an additional pump ( Fig. 1) into the scrubber column at flow rate of 3 L min −1 . NO gas and Ar mixtures, which were obtained by the controlled mixing of air and NO gas using mass flow controllers (MFCs; model: 1179A13CS1BK-S, MKS Co. Ltd., USA), were introduced to the bottom of the scrubber at a set gas flow rate. The anolyte and catholyte scrubbing solutions were introduced to the top of the each scrubber counter to the gas flow at a constant liquid flow rate (3 L min −1 ). The scrubbing solutions were re-circulated through the electrochemical cell to regenerate Ni(I) and Co(III). Before starting the NO removal experiment, the electrochemical cell was first operated until the Ni(I) and Co(III) conversion reached a steady state, and the scrubbing solution was then pumped into the scrubber column. All the experiments are performed at room temperature 20 ± 2 • C.

Ni(I) and Co(III) analysis.-Ni(I)
concentrations were determined by redox titration against a standard KMnO 4 or Fe(III)(CN) 6 (0.001 M) solution using an ORP electrode (EMC 133, 6 mm Pt sensor electrode and Ag/AgCl reference electrode containing a gel electrolyte, iSTEK, USA) connected to an iSTEK multimeter (pH-240L, USA). In a similar manner, Co(III) was determined using Fe(II)SO 4 (0.001 M) as the titrant. The initial reduction potential of Ni(II) was approximately −170 mV, which then decreased as electrolysis progressed at a constant current density to reach approximately −800 mV. At the same time, the initial potential of Co(II) was around 600 mV, which then increased to approximately 1300 mV during electrolysis. The reduction and oxidation efficiencies are defined as follows: Oxidation efficiency (%) = [Co(III)] [Co(II)] × 100 [2] where Ni(I) and Co(III) are the concentrations of the electrolytically formed Ni(I) complex and Co(III) ion. Ni(II) and Co(II) are the initial concentrations of the Ni(II) complex and Co(II) ion precursor, respectively.
Analysis.-The aqueous reaction solution sample withdrawn at each defined interval during NO removal was analyzed by attenuated total reflectance -FTIR (ATR-FTIR, Thermo scientific, Nicolet iS5, USA) using a 2 μL drop of the reaction sample on the diamond prism to examine the products formed in solution. For GC analysis, gas samples (0.5 μl) were withdrawn via gas syringe (model 10R-GP-GT) from exit gas of cathodic scrubber column (Fig. 1) after NO reduction and analyzed by GC using a HP 6890. The molecular sieve-13X (length 1.6 m × 2.1 mm diameter with 13 × 60/80 molecular sieve mesh made by stainless steel tubing) was used for gas chromatographic separations (Sigma Aldrich). The GC oven temperature was set 200 • C and TCD (thermal conductivity detector) temperature was fixed to 250 • C. Helium was used as carrier at a flow rate of 2 ml min −1 .
Nitrate analysis by HACH.-During the NO removal process, a 10 ml aliquot sample of the anolyte solution was drawn from the anode or anolyte chamber, in which the required amount of salicylic acid (0.2 ml reagent) was added and the solution was neutralized by adding NaOH (pH 6.8). The resulting amber color solution was then analyzed directly using a HACH colorimeter (DR-2800, U.S.A) using 410 nm.
Energy analysis.-The total energy that needs to generate Ni(I)(CN) 4 3− and Co(III), and NO removal were calculated based on the literature, 32 as given in Eq. 3, which is equal to total supplied energy to produce the Ni(I)(CN) 4 3− and Co(III), and NO removal processes.
Where V and I are the average cell voltage and current that used for electrolysis; t is the electrolysis time (h). V e and V NO are the volume of electrolyte and volume of NO gas treated, respectively. The energy consumption was reported in kWhm −3 . In order to compare the present method of NO removal process with internationally accepted energy standard, the energy efficiency of NO removal was derived through a continuous flow operation process, 33 which was developed for NO removal using dielectric barrier discharge (DBD) method, as given in the Eq. 4.
Where FR is the gas flow rate (L/min) and FC is the initial feed concentration (ppm) used in the experiments and RE is removal efficiency of pollutant (NO). The gas constant for NO gas (30/22.4) at NTP (1 atm with 293 K) has been used to convert the gas mole into liquid. W is the applied current density to generate mediator and remove NO gas. Fig. 2A shows ORP (oxidation/reduction potential) and reduction efficiency variation of Ni(II)(CN) 4 2− during electrolysis in a 9 M KOH solution. The ORP value reached −800 mV in 120 min and increased steadily in further electrolysis times ( Fig. 2A  curve a) indicating the low oxidation state of Ni(II), in this case Ni(I) formation. 34 Ni(I) formation as a measure of the reduction efficiency, which was determined as mentioned in the Experimental section, increased with increasing electrolysis time to 9% (4.3 mM) for a 120 min duration ( Fig. 2A curve b). Once the NO gas was introduced into the electrogenerated Ni(I)/Ni(II)-containing solution, the reduction efficiency of Ni(II) decreased to 1.83 mM (3%) and was varied slightly as long as NO was purged into the solution, which indicates that a reaction between Ni(I) and NO occurred, called mediated reduction. The Ni(II) reduction efficiency increased reached approximately 7% (3.4 mM) once the NO was stopped, showing that the process is sustainable, 34 and Ni(I) can be generated without additional chemicals. As reported in the literature, NO is relatively insoluble in water, 14 but if scrubbed through a 9 M KOH solution, 100 ppm of the feed NO absorbed nearly 10 ppm in 15 min (Fig. 2B curve a) and decreased to 0 ppm in 20 min revealed absorption in a 9 M KOH solution. A similar trend was observed in a 9 M KOH solution with electrolysis (without 0.05 M Ni(II)(CN) 4 2− ) ( Fig. 2B curve b) with a slight increase in absorption, 15 ppm in 5 min, followed by a decrease to 0 ppm in 15 min, indicating that the DER (direct electrochemical reduction) of NO occurred in a short time and a lower concentration of NO or no DER is occurred. At the same time, in the presence of 0.05 M Ni(II)(CN) 4 2− or electrogenerated Ni(I), the 0 ppm found in the outlet means 100% removal efficiency of NO (Fig. 2B curve c), which shows NO reduction follows the MER process.

Electrogeneration of Ni(I) and NO reductive removal optimization.-
In addition, the feed concentration variation of NO resulted in 100% removal efficiency at 100 ppm from the beginning of the NO feed to the studied time ( Fig. 3A curve a). The NO removal efficiency decreased to 62% at the feed NO of 150 ppm (Fig. 3A curve b) and at feed of 200 ppm (Fig. 3A curve c) tells a saturation has been reached around 150 ppm feed concentration. Almost complete saturation oc- curred in the NO removal efficiency at a 200 ppm feed. This decrease in removal efficiency of NO at a high feed concentration might be due to the unavailability of Ni(I) in the electrolyte solution, which was confirmed by the Ni(I) concentration variation during NO removal at various feed NO concentrations, as shown in Fig. 3B. The Ni(I) concentration increased during 100 and 150 ppm of NO removal (Fig. 3B  curve a and b). On the other hand, no change in Ni(I) concentration at 200 ppm feed of NO (Fig. 3B curve c). In addition, the effect of the gas flow rate was performed to understand the limit of the gas flow rate, which is depicted in Fig. 3C. At a low gas flow rate of 0.25 g min −1 , almost 100% NO removal efficiency was observed (Fig. 3C curve a). At the same time, the NO removal efficiency decreased to 75% at a 150 ppm feed (Fig. 3C curve b) and 50% at 200 ppm feed (Fig. 3C curve c) NO concentration, as expected, indicating the unavailability of Ni(I).
The online FTIR gas analyzer can monitor simultaneously whether additional gas coming from the scrubbing column during the reductive removal of NO, here NH 3 Instead, NH 3 was observed at the scrubber exit, as shown in Figs. 4A and 4B curves b-d. At a 100 ppm NO feed, 75 ppm of NH 3 was observed in 5 min, which was then reduced to 45 ppm in 20 min and became constant thereafter (Fig. 4A curve b). At a 150 ppm feed NO, the initial NH 3 concentration was 162 ppm, which reduced to almost 100 ppm in 30 min and remained constant thereafter (Fig.  4A curve c). Similarly, the initial NH 3 concentration found 200 ppm and was reduced during the removal time to 75 ppm in 60 min (Fig.  4A curve d). The decreasing trend of the formed NH 3 concentration with the removal time might be due to the additional reaction that forms an additional product. NH 3 and NO in the aqueous phase leads to the formation of N 2 , 4,35 as shown in Reaction 9, and an additional reaction might have started after certain level of NH 3 formation.
This is what occurred in the present case. The initial NH 3 concentration was similar to the feed concentration of NO, and the NH 3 concentration began to decrease depending on the concentration of NH 3 , which was further confirmed by NH 3 formed by the gas flow rate variation (Fig. 4B curve b-d). A difference in the initial NH 3 concentration was observed depending on the feed flow rate, but later, there was no change in the NH 3 concentration, which means no change in the feed NO concentration. To monitor for N 2 , separate GC experiments were performed during NO reduction, as shown in Fig.  4A curve e. In the case of 100 ppm feed NO, the N 2 concentration increased to 45 ppm in 20 min and remained constant, which confirms the formation of N 2 with NH 3 , as shown in Reaction 2. For confirmation, the gas phase FTIR were shown in Fig. 5A. The inbuilt (provided by the MIDAC instrument library) standard gas FTIR spectra for NO 36 and NH 3 shown in Fig. 5A curve a and b for comparison. The direct NO feed to the online FTIR shows two strong stretching peaks (1899.2-1905 cm −1 ), 36 which is the region fixed to monitor removal of NO, behavior of N-O along with moisture peaks (Fig. 5A curve c). Fig. 5A curve d shows outcome FTIR spectrum during removal of NO by MER, where N-H bending peaks (989.9-995 cm −1 ), which is the region fixed to monitor NH 3 gas, found that is similar to the Fig. 5A curve b spectrum along with moisture peaks confirms the NH 3 formation. At the same time, the two strong stretching peaks for N-O completely vanished confirms the removal of NO. In addition, the solution phase ATR-FTIR of before and after the NO removal process revealed no change or an additional peak (Fig. 5B curve a and b) for N-O stretching region or N-H stretching region, which explains why the NH 3 formed was not in solution due to the high pH (14) of 9 M KOH. Beyond pH 11, NH 4 OH becomes NH 3 gas and escapes from solution. 37

Electrogeneration of Co(III) and NO oxidative removal optimization.-Although
NOx oxidation occurs by the MEO process, the use of Co(III) in a 5 M H 2 SO 4 solution is new. Similar to NO reduction, oxidation was performed under different conditions and is presented in Fig. 6A. No absorption was observed except for 3 ppm in a short time at 5 to 20 min (Fig. 6A curve a). Similar behavior was found for the DEO (direct electrochemical oxidation) of NO (without 0.05 M Co(II)SO 4 ( Fig. 6A curve b) at 15 min and reached 0 ppm in 20 min, which confirms that no DEO of NO had occurred. Almost 28% of the NO was removed in the presence of electrogenerated Co(III) in 5 M H 2 SO 4 (Fig. 6A curve c), which confirms that NO oxidation is influenced by MEO, as evidenced by the Co(III) concentration variations during the addition of NO. Fig. 6B shows 35% (18 mM) of Co(III) formed in 2 h electrolysis, which is decreased to 10 mM (20%) in 1 h after the NO was injected into the scrubbing column ( Fig. 6B curve a) that confirms the reaction between Co(III) and NO gas. NO 2 is monitored simultaneously along with the NO concentration during oxidative removal, and N 2 O and N 2 O 5 were monitored through the batch run mode (offline mode) after the NO removal experiments. Fig. 7 shows no N 2 O 5 , N 2 O, and NO 2 during the NO removal process ( Fig. 7 curve a-c). Previous studies of NO oxidative removal by electrogenerated Ag(II) revealed 60 to 80% removal obtained depending on the conditions, 20,23 but NO 2 was found as reaction product in first scrubber. In the present case, using electrogenerated Co(III), 28% removal efficiency was obtained with no NO 2 in the first scrubber exit. The solution was analyzed for NO 3 − using the HACH method and the results obtained showed an increase in the NO 3 − concentration from 62 ppm to 135 ppm with increasing oxidative removal time ( Fig.  7 curve d) up to a 60 min duration. Although the oxidative removal efficiency of NO was 28% by electrogenerated Co(III) in a 5 M H 2 SO 4 medium, there was no NO 2 greenhouse gas generated, but NO 3 − was observed at the first stage electro-scrubber itself. Fig. 3C curve a, 100% removal of NO was found until a 0.25 g min −1 gas flow rate. To increase the removal efficiency of NO at a higher gas flow rate, a combined removal approach, a combination of MEO and MER processes was attempted and the obtained results were presented in Fig. 8. 100 ppm of NO with a 0.5 g min −1 gas flow rate was injected into the bottom of the anolyte scrubber, which contained 30% Co(III) (15 mM) in a 4 M H 2 SO 4 solution. The exit gas was injected sequentially into the bottom of the catholyte scrubber, where 3.8 mM (8%) of Ni(I) in 9 M KOH was generated at the time of gas injection. A 28% removal efficiency was observed using the Co(III) MEO of NO ( Fig. 8 curve a) and the remaining 72% NO was removed completely by the Ni(I) mediator at the cathodic scrubber ( Fig. 8 curve b). These results show that a combination of MEO and MER makes NO removal possible at a high gas flow rate due to the additional removal process. Energy for mediator generation and NO removal.-Energy consumption for Ni(I)(CN) 4 3− and Co(III) generation, and NO removal were calculated using the Eq. 3 and presented in Table I. As shown in Table I, generation of only Ni(I)(CN) 4 3− needs 0.121 kWhm −3 , but the both Co(III)(18 mM) and Ni(I)(CN) 4 3− (4.3 mM) mediators can be generated by the same energy when operate the both half-cells together. The NO removal by using only MER (cathodic half-cell) is 0.060 kWhm −3 , but more than a half of the energy (0.031 kWhm −3 ) is reduced if use the MEO (anodic half-cell) with MER (cathodic halfcell). Table I also contains the energy efficiencies of Ni(I)(CN) 4 3− and Co(III) generation, and NO removal processes that are calculated using the Eq. 4. The energy needed to remove NO through individual MEO and MER processes found to be 0.221 g/kWh and 0.769 g/kWh respectively (Table I middle column). At the same time, the NO removal energy efficiency become higher 0.99 g/kWh (Table I middle column) when combined the MEO & MER processes, which is almost 10 times lower than the energy efficiency obtained for NO removal by nanosecond pulsed discharge (12 g/kWh) 38 and non-thermal discharge (∼20 g/kWh) 39 methods. Noteworthy here that the high installation cost of nanosecond pulsed discharge and non-thermal discharge methods makes the present electrochemical (installation cost is very less) approach more economical. Moreover, no electrochemical methods reported for energy efficiency of NO removal explains the present attempt is a comparatively efficient method in concerning the energy efficiency.

Conclusions
The MEO and MER combined removal of NO by electroscrubbing process is very efficient. The electrogeneration of Ni(I) from Ni(II)(CN) 4 2− was sustainable and was used for NO reduction for the first time by electro-scrubbing. The reduction of NO using Table I. Energy consumption (kWhm −3 ) and energy efficiency (g/kWh) during mediator generation and NO removal by electro-scrubbing.