Removal of NO With Fe(II)NTA Solution Catalyzed By The Carbon Treated With Ethylenediamine

: Fe( Ⅱ )NTA solution manifests a good performance in the simultaneous removal of sulfur 7 dioxide and nitric oxide. Activated carbon is used to catalyze the reduction of Fe(III)NTA to Fe( Ⅱ )NTA to 8 retain the ability of absorbing NO. Ethylenediamine(EDA) solution is capable of changing the physical 9 structure and chemical characteristics on the carbon surface to improve the catalytic capability of activated 10 carbon. The experiments suggest that the best treatment condition be immersing the carbon in 5.0 mol l -1 11 EDA solution for 6 h followed by being heated at 700 ℃ in N 2 for 4 h. The modification with EDA 12 increases the surface area and alkalinity on the carbon. The experiments also indicate that the removal 13 efficiency of nitric oxide catalyzed by the modified carbon is significantly improved compared with that of 14 the original one. 15


Introduction 18
The emissions of sulfur dioxide and nitric oxide are causing a series of environmental problems such as 19 acid rain, ozone layer destruction, photochemical smog, and even endangering human health(Toro et al. by chemical agents has not been applied commercially yet due to their high costs and the production of 29 large amounts of waste water. The SCR technology is most widely used in coal-fired power plants to 30 complete the reduction of nitric oxide in the flue gases. But this technology suffers the disadvantage of 31 high capital and operating costs ( Tang et al. 2020). Hence there is an urgent need to develop a low-cost, 32 easily-industrialized method of denitrification. The low solubility of NO in water can be avoided effectively by binding NO with complexants. Such 34 liquid-phase complex method has a broad industrial application prospect because it holds the advantages of 35 high denitrification efficiency, fast reaction rate and small equipment input. The approach that introduces 36 In the meantime, the SO2 existing in the gas stream also dissolves into the aqueous solution: 51 However, the oxygen coexisting in the flue gases may oxidize Fe(Ⅱ)(NTA)to Fe(III)(NTA)(Eq.(5)) during 54 the gas scrubbing. NO removal efficiency will decrease quickly as the operation proceeds due to the 55 To retain the NO removal efficiency, activated carbon can be used as a catalyst and the sulfite/bisulfite 58 ions produced by SO2 absorption into the aqueous solution act as reductants to regenerate Fe(Ⅱ)(NTA) -. 59 The mechanism of Fe(III)(NTA) catalytic reduction can be expressed as follows: 60 The Fe(III)NTA adsorbed on the carbon surface disintegrates into Fe(III) and NTA: 61 Electrochemical half-cell reduction potential of Fe(III)/Fe(II)(Eq(7)) exhibits that Fe(III) is a strong 63 oxidant and can be reduced to Fe(II) easily. 64 The net reaction for the regeneration of Fe(II) ions can be written as follow: 66 Fe(Ⅱ)(NTA) is regenerated by Fe(Ⅱ) coordinating with NTA(Eq. (9)). Therefore, the NO removal 68 efficiency can be sustained for a long time. 69 Besides, the NO coordinated with Fe(Ⅱ)(NTA) may be reduced to N2O by sulfite. 71 Nitrous oxide is also a harmful pollutant that contributes to the depletion of the ozone layer (Ravishankara 73 et al. 2009 According to the discussion above, this technology realizes not only the absorption and reduction of 78 nitric oxide but also the absorption and oxidation of sulfur dioxide.

79
Activated carbon plays an important role in the process of regeneration of Fe(Ⅱ)(NTA). The catalytic 80 activity of activated carbon is dependent on its surface characteristics. By transforming the physical 81 structure and chemistry characteristics on the carbon surface, the capabilities of activated carbon may be 82 improved to a certain extent. Wang  has been tried to treat the activated carbon of coconut to promote its catalytic capacity in the regeneration of 100 The coconut activated carbon purchased from Shanghai Activated Carbon Co., Ltd, was purged with 105 deionized water and then dried at 110 ℃ in a vacuum for 24 h. Carbon samples of 100 ~ 120 mesh were 106 prepared before being used as catalysts or modified with EDA solution. The carbon was dealt with EDA 107 solution in the following steps: Firstly, 5 g activated carbon was impregnated in 250 mL EDA solution for 108 several h at room temperature; secondly, after being filtrated, the carbon sample was put in a vacuum 109 drying oven at 110 ℃ for 12 h; thirdly, the samples obtained were heated under N2 atmosphere in a 110 furnace at set temperature for a few h. 111

Reduction of Fe(III)NTA 112
The experiments to test the catalytic capability of the carbon samples in the reduction of Fe(III)NTA where A stands for absorbency and C for Fe(II) concentration (10 -3 mol L -1 ). 125 Fe(III) concentration was computed from the difference between total iron and Fe(II

Combined removal of NO and SO2 132
The schematic diagram of the experimental apparatus for the simultaneous removal of NO and SO2 is 133 shown in Fig. 1 The absorption was performed in a packed tower (18 mm i.d., 1000 mm long) and the 134 Fe(II)NTAregeneration was carried out in a fixed-bed (20 mm i.d.) reactor packed with 20 g activated 135 carbon of 20-40 mesh. The temperature of the absorber and regeneration tower was controlled at 50 ℃ 136 by the jackets through which water from thermostatic baths was recycled. Five hundred milliliter 137 Fe(II)NTAsolution together with measured amount of Na2SO3 was added into the circulation tank. The 138 pH was controlled at 5.5 using NaOH (1.0 mol/L) solution by a THORNTON M300 pH/ORP transmitters 139 as well as a pH-electrode in the course of the experiment. The absorber was operated with a continuous 140 gas stream feeding at 270 ml min -1 from the bottom and a continuous scrubbing solution feeding at 25 ml 141 min -1 from the top. The absorbent discharging from the packed tower was fed into the circulation tank. Hence this set-up is feasibly and conveniently operated to monitor the NO and SO2 removal efficiency. 158

Characterization of carbon samples 159
FT-IR was used to analyze the functional groups on the surface of activated carbon by KBr 160 compression method(O'reilly and Mosher 1983) and the point of zero charge(pHpzc) was determined by 161 mass titration.. An ASAP2020 surface Analyzer(Micromeritics Co. USA) was used to measure the 162 specific surface area of activated carbon with nitrogen as adsorption medium at 77K. The total surfaces of 163 the carbon samples were calculated by BET method. The surface area and volume of mesopores were 164 obtained by BJH method, and the micropores of which were computed by t-plot method. The content of 165 acidic and basic functional groups on the surface of activated carbon were measured by Boehm 166 titration (Boehm 1994). XPS was characterized by an ESCALAB 250 electron spectrometer from Thermo 167 Corporation with 300 W AlKα radiation at the base pressure of 3×10 -9 mbar. 174 To explore the effect of EDA concentration on the catalytic capacity of the activated carbon, six 175 carbon samples were impregnated in EDA solutions with a concentration of 3.0 mol L -1 , 4.5 mol L -1 , 5.0 176 mol L -1 , 6.0 mol L -1 , 7.5 mol L -1 and 10.5 mol L -1 , respectively, for 6 h at ambient temperature. And then 177 they were heated in N2 at 800 ℃ for 4 h. The prepared samples were used to catalyze the reduction of 178 44.06 and 41.07%, respectively. As a result, the best EDA concentration for the carbon modification is 183 The reason for the improvement in the catalytic ability of the carbon samples treated with EDA 185 solution may be given according to the change of its surface characteristics. 186 The carbon samples were detected by FTIR to analyze their surface chemistry. The band exhibited at 2919 cm -1 is ascribed to the stretching vibration of the hydrocarbon single bond C-H. 191 The carbonyl absorption peak from lactonic and carboxyl is exhibited at 1625 cm -1 . The peak at 1089 192 cm -1 is due to the phenolic -OH group and C-O group. It can be seen from Fig.3   The molar percentages based on the peak resolution are illustrated in Table 1. The data listed in Table 1  202 indicates that the molar percentage of graphitic carbon on the carbon surface increases from 70.07% to 203 71.70% after being modified with EDA solution. The percentage of the carbon in phenolic, alcohol or 204 ether groups decreases slightly (from 8.97% to 8.47%) after the carbon treated by EDA solution. The 205 carbon in carbonyl or quinine groups reduces imperceptibly from 7.41% to 7.22%. The C in lactonic or 206 nitrogenous group drops from 5.16% to 4.75%.
The molar percentage of carboxyl decreases sharply 207 from 2.31% to 0.92%.
XPS C 1s spectra depicts that the EDA treatment ameliorates the π structure on 208 the carbon.   The XPS O 1s spectra depicted in Fig. 8  The concentrations of the surface functional groups determined by Boehm titration listed in Table 3  227 may also account for the improvement of the carbon catalytic ability. It can be seen that the total acidic 228 groups and basic groups have been changed greatly after being modified with EDA solution. The total 229 basic groups on the modified carbon are significantly raised while the total acidic groups are reduced 230 compared to those on the original one. For instance, the total basic groups on the original carbon are only 231 4.02×10 -4 mol g -1 but those on the one immersed in 6.0 mol L -1 EDA solution rise to 7.18×10 -4 mol g -1 . 232 The phenolic hydroxyl on this modified carbon is reduced by 81.25% compared with that on the 233 unmodified one. The carboxylic also drops greatly from 1.50 ×10 -4 mol g -1 to 0.11×10 -4 mol g -1 . And the 234 amount of lactonic on this modified carbon is nearly one half times that on the original one. The physical 235 characteristics of the carbon samples listed in Table 4 suggest that the modification with EDA solution 236 gives rise to the increase in total surface area, mesopore area and micropore area on the carbon surface. 237 For instance, the total surface area on the original carbon is 779 m 2 g -1 , but that on the one impregnated in 238 5.0 mol L -1 EDA solution is 813 m 2 g -1 . The micropore area on this modified carbon increases by 26 m 2 239 g -1 compared with that on the original one. The reason for the magnification of both SBET and Smic is that 240 ethylenediamine can etch the activated carbon and remove the ash in the pores. Furthermore, 241 ethylenediamine as well as the acidic groups on the carbon surface decompose at high temperature, which 242 may bring about an increase in the pore structure and surface area on the carbon surface. The 243 modification with EDA solution not only increases the basic functional groups on the carbon surface but 244 also amplifies the pore structure of the carbon, which helps to adsorb Fe(III)NTA and disintegrate it into 245 Fe 3+ and NTA 3-, accelerating the reduction of Fe(III). Therefore, the carbon modified with EDA is 246 superior to the original one as a catalyst in the regeneration of Fe(III)NTA. 247   Table 3 also manifests that the total basic groups goes up gradually while the total acid groups go 248 down slightly as the EDA concentration increases from 3.0 mol L -1 to 6.0 mol L -1 . This may be because 249 more ethylenediamine is adsorbed onto the activated carbon and reacts with the acidic groups on the carbon 250 surface as EDA concentration increases, which is beneficial to the formation of basic groups at high 251 temperature. 252 Table 3  The data in Table 4 reveals that the total surface area and micropore area magnify by 10 m 2 /g and 17 254 m 2 /g, respectively with the EDA concentration rising from from 3.0 mol L -1 to 5.0 mol L -1 . However, 255 when the EDA concentration increases from 5.0 mol L -1 to 6.0 mol L -1 , the total surface area and micropore 256 area decrease by 8 m 2 /g and 10 m 2 /g, respectively. Therefore, the SBET of the carbon will increase with the 257 EDA concentration because more pores will be produced due to the reaction bwtween EDA and carbon 258 when the samples are calcined at high temperature. However, as the EDA concentration increases over 5 259 mol L -1 , SBET decreases because some micropores will be transformed into mesopores and macropores due 260 to the violent reaction between carbon and EDA. 261 SBET and Smic of the carbon soaked in 5.0 mol L -1 EDA solution are bigger than those of the carbon 263 soaked in 6.0 mol L -1 EDA solution and the basicity of the former is slightly weaker than that of the latter. 264 Greater SBET and Smic are conducive to the catalytic ability of the carbon in the regeneration of Fe(II)NTA. 265 Therefore, the carbon soaked in 5.0 mol L -1 EDA solution exhibits stronger catalytic ability than the carbon 266 soaked in 6.0 mol L -1 EDA solution. SBET of the carbon soaked in 6.0 mol L -1 EDA solution is slightly 267 smaller than that of the carbon soaked in 3.0 mol L -1 EDA solution. But the basicity of the carbon soaked 268 in 6.0 mol L -1 EDA solution is much stronger than that of the one soaked in 3.0 mol L -1 EDA solution. At 269 a pH below the isoelectric point of the carbon, the carbon is positively charged and will adsorb 270 preferentially anionic species(Rodriduez-Reinoso 1998). The higher the pHpzc, the greater the positive 271 charge density on the carbon surface, which is favorable for the adsorbability of the anionic NTA and 272 sulfite on activated carbon. Thus, the reduction of Fe(III)NTA is benefited. The basic groups on the 273 carbon samples play a more important role than their physical structure. Therefore, the carbon soaked in 274 6.0 mol L -1 EDA solution gets a higher Fe(III)NTA conversion than the one immersed in 3.0 mol L -1 EDA 275 solution . 276

Effect of impregnation time 277
The duration of the carbon impregnated in EDA solution is a vital factor influencing the effect of 278

289
The data in Table 5 reveals that the total acidic groups decrease from 0.80×10 -4 mol g -1 to 0.60×10 -4 290 mol g -1 as the impregnation time prolongs from 4 h to 6 h. But if the impregnation time extends to 8 h, the 291 amount of acidic functional groups is almost unchanged. The total basic groups increase with the 292 impregnation time because the reaction between carbon and EDA is of benefit to the enhancement of the 293 basicity of the carbon. The physical characteristics shown in Table 6 indicate that SBET decreases from 294 816 m 2 g -1 to 705 m 2 g -1 with the extension of the impregnation time from 4 h to 8 h.. This may be 295 because the long reaction time between carbon and EDA turns some micropores into mesopores and 296

macropores. 297
Though the carbon soaked in the EDA solution for 4 h has the biggest SBET, it exhibits the weakest 298 catalytic ability in the regeneration of Fe(II)NTAbecause it holds the smallest basic groups on its surface. 299 The basic groups on the carbon soaked in the EDA solution for 8 h are slightly more than those on the one 300 immersed in the EDA solution for 6 h but the latter holds bigger SBET than the former. Therefore, the 301 carbon soaked in the EDA solution for 6 h can obtain a higher Fe(III)NTA conversion than the one soaked 302 in the EDA solution for 8 h. Both the physical structure and surface chemistry of the activated carbon 303 determine its catalytic ability in the reduction of Fe(III)NTA jointly. 304

Effect of calcination temperature 307
To explore the effect of calcination temperature on the catalytic capacity of the activated carbon, four 308 carbon samples were heated in N2 for 4 h at 600, 700, 800, and 850 ℃, respectively after having been 309 impregnated in 5.0 mol L -1 EDA solution for 6 h at ambient temperature. The prepared samples were used 310 to catalyze the reduction of Fe(III)NTA at 70 ℃. The conversions of Fe(III)NTA presented in Fig. 7 prove 311 that the best calcination temperature for the carbon modification is 700 ℃. After 90 min's reaction, the 312 Fe(III)NTA conversion obtained increases from 51.01 to 53.62% as the calcination temperature is raised 313 from 600 to 700 ℃. However, when the temperature rises further to 800 ℃, the Fe(III)NTA conversion 314 drops to 50.24%.  Table 7 presents the chemical functional groups of the carbon samples impregnated in 5.0 mol L -1 319 EDA solution for 6 h followed by being carbonized for 4 h at 600, 700, and 800 ℃, respectively. It can 320 be seen that the total acidic groups and the total basic groups are almost unchanged when calcinating the 321 carbon samples at 600 and 700 ℃. The total basic groups decrease from 7.20 ×10 -4 mol g -1 to 7.01×10 -4 322 mol g -1 when the calcining temperature rises to 800 ℃. This may be because the acidic functional groups 323 on the carbon surface such as carboxyl, lactone and phenolic hydroxyl have been decomposed sufficiently 324 above 600 ℃ and if the calcination temperature increases to 800 ℃, some basic functional groups begin 325 to decompose. The physical characteristics illustrated in Table 8 exhibit that the total surface area, 326 mesopore area and micropore area of the carbon samples increase gradually with the calcination 327 temperature rising from 600 to 800 ℃ because more micropores and mesopores are produced due to the 328 reaction between EDA and carbon proceeding more violently at higher temperature. 329 In spite of its biggest SBET, the carbon calcined at 800 ℃ gets the lowest Fe(III)NTA conversion 330 because it holds less basic groups than the other two samples. The carbon calcined at 700 ℃ holds similar 331 chemistry characteristics with the carbon calcined at 600 ℃, the former is superior to the latter as a catalyst 332 in the regeneration of Fe(II)NTAbecause the former owns larger surface area. 333

Effect of calcination time 336
The effect of calcination time on the catalytic capability of activated carbon should also be 337

346
The data listed in Table 9 depicts that the total acidic groups and the total basic groups on the surface 347 of activated carbon change little as the calcination time prolongs from 3 h to 5 h. But the physical 348 characteristics shown in Table 10 indicates that when the calcination time is extended from 3 h to 4 h, the 349 total surface area and micropore area increase from 782 m 2 g -1 and 674 m 2 g -1 to 800 m 2 g -1 and 702 m 2 g -1 , 350 respectively. This is because appropriate extension of the calcination duration is favorable for the 351 formation of micropores resulted from the reaction between carbon and EDA. The total surface area 352 decreases to 791 m 2 g -1 if the calcination time prolongs further to 5 h. The reason may be that excessive 353 calcination time leads to the transformation of some mesopores to macropores. The Fe(III)NTA 354 conversions they have got are in accordance with the sequence of their physical structures. Therefore, the 355 best calcination time is selected to be 4 h. 356

Simultaneous removal of NO and SO2 359
The experimental schematic apparatus of the simultaneous removal of NO and SO2 with Fe(II)NTA -360 solution as well as the Fe(II)NTAregeneration catalyzed by the raw carbon and the modified carbon is 361 shown in Fig. 1. The modified carbon was treated with EDA solution under the optimal condition 362 discussed previously. In the experiment, both of the absorber and the regeneration reactor are controlled 363 at 50°C. The absorption solution is a mixture of 500 mL 0.05 mol L -1 Fe(II)NTAand 0.04 mol L -1 364 Na2SO3. The concentrations of SO2 and NO in the gas inlet are 1800 ppm and 580 ppm, respectively. 365 O2 in the simulated flue gas is 5.0% (volume). The gas flow rate is 270 ml min -1 and that of the scrubbing 366 solution is controlled by a peristaltic pump at 25 ml min -1 . The pH value of the solution in the circulation 367 tank is controlled around 5.5. The experimental results are shown in Fig. 9. 368 It can be seen from Fig. 9 that the removal efficiency of NO decreases from 100% to about 35.9% 369 after 2.4 h run due to the consumption of Fe(II)NTA -. At this time, the Fe(II)NTAregeneration is started. 370 Obviously, the NO removal efficiency regain quickly in these two operations after the regeneration process 371 begins. But there is evident difference between two operations. The NO removal efficiency rises to 372 about 81.3% in 1 h and then begins to decrease gradually under the catalysis of the unmodified carbon. 373 After 58 h operation, the NO removal efficiency drops to 57.9% and is maintained at 52.1-54.4%. As a 374 contrast, when Fe(II)NTAregeneration is catalyzed by the modified carbon, the NO removal efficiency 375 reaches to 100% in about 15 min and is maintained at 100 % for 1.2 h. Thirty hours later, the NO removal 376 efficiency decreases to 78.4% and fluctuates between 73.7% and 77.3% in the run. Besides, there is no 377 SO2 detected in the outlet gas by FTIR in the whole process. The explanation for such phenomenon is as 378 follows. Fe(II)NTAis gradually oxidized to Fe(III)NTA by oxygen, leading to the reduction in NO removal 379 efficiency. After the Fe(II)NTAregeneration starts, Fe(III)NTA is reduced by sulfite/bisulfite to 380 (2) The carbon surface characterization demonstrates that the treatment with EDA solution gives rise to an 398 evident increase in the basic groups and obvious decrease in acidic groups on the carbon surface. The 399 BET results prove that the modification also brings about a slight increase in the surface area. And these 400 changes are favorable for the improvement of the catalytic activity of the activated carbon in the generation 401 of Fe(II)NTA -. The catalytic ability of activated carbon in the Fe(II)NTAgeneration relies on its physical 402 structure and surface chemistry. The surface chemistry plays more important role than its physical 403 structure in determining the catalytic capability of carbon. 404 (3) The modified coconut activated carbon can achieve a much higher NO removal efficiency than the 405 unmodified coconut activated carbon. Therefore, this modification with EDA solution is an effective way 406 to enhance the catalytic ability of the activated carbon in the simultaneous removal of NO and SO2 with 407    C1s spectra of the original carbon and the modi ed carbon Figure 5 O1s spectra of the original carbon and the modi ed carbon Effect of calcination temperature on the catalytic performance of activated carbon NO removal coupled with Fe(II)NTA-regeneration catalyzed by activated carbon