Science in China Series E: Technological Sciences © 2008

SCIENCE IN CHINA PRESS

Springer

www.scichina.com tech.scichina.com www.springerlink.com

Study on method and mechanism for simultaneous desulfurization and denitrification of flue gas based on the TiO2 photocatalysis ZHAO Yi†, ZHAO Li, HAN Jing, XU YongYi & WANG ShuQin School of Environmental Science & Engineering, North China Electric Power University, Baoding 071003, China

Based on the TiO2 photocatalysis mechanism, a new method of simultaneous desulfurization and denitrification from flue gas was proposed. Preparation of TiO2 photocatalyst, design of photocatalysis reactor and influencing factors for simultaneous removal of SO2 and NO, and removal mechanism of SO2 and NO were studied. After the optimal values of concentration of O2 in flue gas, the relative humidity of flue gas and the irradiation time in the photocatalysis reactor were used, the efficiencies of removal for SO2 and NO can be achieved above 98% and about 67%, respectively. According to the results of removal products analysis, the removal mechanism of SO2 and NO based on TiO2 photocatlysis can be put forward, namely, SO2 was oxidized to SO3 partly, the bulk of NO was oxidized to NO2, and both were removed by resorbing finally. TiO2 photocatalysis, flue gas, simultaneous desulfurization and denitrification, reaction mechanism

In order to protect air environment, technologies such as wet limestone-gypsum process, flue gas circulating fluidized bed (CFB) process and sea water desulfurization process and so on are adopted to remove SO2 from flue gas in power plant, among which wet limestone-gypsum technology is the most commonly used. For removal of NOx, selective catalytic reduction (SCR) or selective non-catalytic reduction (SNCR) is usually installed after the desulfurization equipment, namely fractional removal, in which the yard area and the operating cost will be increased. So this technology of combined removal of SO2 and NO is difficult to apply any more. To reduce the cost of flue gas purification, development of new technologies and equipment of simultaneous flue gas desulfurization and denitrification is necessary. TiO2 is the commonest photocatalyst due to its favorable chemical properties, high stability and low cost. Its interest in wastewater and air treatment has received much attention lately. Sig― nificant achievements in basic researches and applications have been made[1 3]. A cogent and Received November 19, 2006; accepted April 3, 2007 doi: 10.1007/s11431-008-0021-0 † Corresponding author (email: [email protected])

Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

successful application for TiO2 photocatalysis is the mineralization of undecomposed pollutants[1], showing that this technique has favorable application foreground and may be developed to an industrialized environmental treatment technology[4]. Nonetheless, the reports about removal of SO2 and NO from flue gas based on TiO2 photocatalytic oxidation is seldom, though, a few studies have been reported on NO at low concentration level. NO removal at 10―6 g·m−3 concentrations has been conducted by Takashi Ibusuki[5], whereas investigation of simultaneous flue gas desulfurization and denitrification using TiO2 has not been found. In this paper, the preparation of supported TiO2 photocatalyst was studied, and simultaneous removal experiments for high concentration of SO2 and NO were carried out at a self-designed photocatalysis reactor, using simulated flue gas. The factors affecting on the efficiency of simultaneous desulfurization and denitrification were investigated, the optimal conditions for simultaneous removal of SO2 and NO by this photocatalyst were determined. According to analysis of removal products, the mechanism of SO2 and NO removal by TiO2 photocatlysis is proposed. The theoretic and technical understructures for industrial application are provided.

1 1.1

Experiment Preparation of supported photocatalyst

A mixed solution prepared by mixing (NH4 )2 TiF6 and H3 BO3 with a ratio of 1:3 was added into the beaker to submerse the quartz sand particles completely, the beaker, covered with a glass surface, was heated at 40℃ for 45 h [6]. The quartz particles were cleaned by water and dried naturally, and then calcinated at 450 ℃ for 0.5 h. with a temperature gradient of 10℃·min −1 . Hence, the supported photocatalyst was obtained. (NH4 )2 TiF6 used in the experiment is chemical grade, and H3BO3 and absolute alcohol are analytical grade, the particle size of quartz sand particles is 4―6 mm, and its purify is greater than 90%.The absorber is a special bubbling reactor. 1.2

Experimental apparatus

Figure 1 shows a schematic diagram of the self-designed experimental equipment, where 9 shows the key apparatus which is a photocatalysis reactor covered with stainless steel. A UV lamp covered by a quartz tube is vertically placed at the center of this reactor. There is a 2 cm gap from the inner wall of the reactor to the outer one of the lamp, which is filled with supported TiO2 catalyst.

Figure 1 Scheme of experimental apparatus. 1, SO 2 cylinder; 2, NO cylinder; 3, N 2 cylinder; 4, O 2 cylinder; 5, flowmeters; 6, gas blender; 7, water vapour generator; 8, compressor; 9, photocatalysis reactor; 10, absorber; 11, dryer; 12, flue gas analyzer. ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

269

The UV lamp placed in the photocatalysis reactor is 40 W of power and 253.7 nm of main wavelength (Atlantic Ultraviolet in USA). SO2 , NO, N2 and O2 (1, 2, 3 and 4 in Figure 1) were adjusted by flowmeter (5 in Figure 1), and mixed in a gas blender (6) by different flow rate, thus the simulated flue gas was obtained. The humidity of the simulated flue gas was controlled by adjusting the injection time of water vapor that was produced with a generator (7). Simulated flue gas entered photocatalysis reactor (9) from the total flowmeter (5), the inlet concentrations of SO2 and NO were determined before of the experiment. While UV lamp was opened, the experiments were beginning. The reacted flue gas was absorbed by a absorber (10), the concentrations of SO2 and NO coming from drier (11) were continuously measured with a flue gas analyzer (12). The removal efficiencies were calculated by comparing the concentrations of SO2 and NO after and before experiments. 1.3

Analysis items and methods

An MRU95/3 CD flue gas analyzer, made in Germany was used to determine the concentrations of SO2 and NO at inlet and outlet. A scanning electron microscope (SEM, KYKY-2800B type, Scientific Instrument Corporation of the Chinese Academy of Sciences) was used to observe the micro-structures of fresh catalyst and spent catalyst，and an X-ray energy spectrometer (EDS, Vantage DIS type, Thermo NORAN Company in USA) was used to analyze the surface compositions of fresh catalyst and spent catalyst. To study the mechanism of desulfurization and denitrification simultaneously by using TiO2 photocatalyst, removal products were analyzed with chemical analysis. The contents of SO 24− and SO32− were determined by barium chromate photometry[7]. The content NO −2 was determined by N-(1-naphthyl)-ethylenediamine photometry[8], and that of NO3− by the process of the reduction of zinc powder.

2 2.1

Results and discussion Effect of O2 on removal efficiency

According to the content of O2 in flue gas, 10% O2 in the simulation flue gas was used in this experiment. The comparison experiments of effect of O2 on the removal efficiencies of SO2 and NO were conducted in presence and absence of O2 . The experimental conditions were as follows. Injection time of water vapor: 15 min; reaction temperature: room temperature; simulated flue gas flow: 0.064 m3·h−1; concentration of SO2: 804 mg·m−3; concentration of NO: 1256 mg·m−3. The variation of the efficiencies for SO2 and NO removals as a function of irradiation time is shown in Figure 2. The results show that the removal efficiencies of SO2 in presence and absence of O2 are 98% and 23%, respectively, and 67% and 15% respectively for NO, indicating that the presence of O2 in flue gas is a crucial factor for photocatalytic oxidation of SO2 and NO. O2 is the captor for the photoproduced electron of the conduction band of TiO2, and can prevent efficiently the recombination of the electrons and the positive holes. At the same time, the electrons are captured by O2 to produce some kinds of active radical such as ·OH, O2−, ·O which can remove SO2 and NO by oxidation[9]. It can be considered that the electrons transport from the 270

ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

catalyst to O2 adsorbed on the surface of catalyst is the crucial step of the photocatalysis oxidation. Nevertheless, the adsorption capacity of O2 lies on strongly flue gas humidity, O2 will be not adsorbed on the surface of TiO2 if flue gas is completely dry[10]. Hence, the higher removal efficiency of SO2 may be related to physical adsorption on the surface of the catalyzer[11].

Figure 2 Effect of the presence of O2 on removal efficiencies of SO2 and NO.

2.2

Effect of irradiation time on NO removal efficiency

For the purpose of determination of the optimum irradiation time, experiments were done. The experimental conditions were as follows. Injection time of water vapor: 15 min; reaction temperature: room temperature; simulated flue gas flow: 0.064 m3·h−1; concentrations of NO are 628, 942, 1256 and 1569 mg·m−3, respectively. The variation of the efficiencies for SO2 and NO removal as a function of irradiation time from 20 to 120 min is shown in Figure 3.

Figure 3 Effect of irradiation time on removal efficiency of NO.

It can be seen from Figure 3 that the removal efficiency for NO markedly increased with the irradiation time from 20 to 100 min. At an irradiation time of 100 min, the greatest removal efficiency was obtained and remained almost constant between 100 and 120 min. These variations are considered to be due to formation of water vapour resulting from evaporation of condensation water on the catalyzer surface, because of an increase of temperature of the reactor as the irradiation time. The hydroxy (OH) offered by water molecules trapped holes to produce more hydroxyl radicals (·OH), and photocatalytic oxidation rate of NO was accelerated, the removal efficiency was enhanced. Thus, the optimum irradiation time was 100 min for removal of NO and SO2. ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

271

2.3

Effect of flue gas humidity on NO removal efficiency

Flue gas humidity in coal-fired flue gas is about 4%―10%. As mentioned above, an increase of flue gas humidity is beneficial for removal of NO in the process of photocatalysis, and yet, the photocatalyst activity might be inhibited if the flue gas humidity exceeds a certain limitation. The variation of the efficiencies for SO2 and NO removal was investigated as a function of flue gas humidity, and results are shown in Figure 4. In this experiment, the concentrations of SO2 and NO were 804 and 1256 mg·m−3, respectively, other experimental conditions were the same as those mentioned above.

Figure 4 Effect of flue gas humidity on removal efficiency of SO2 and NO.

The results show that the efficiency for NO removal increased with the injection time of water vapor, the highest value was obtained at 15 min, and then, the efficiency slowly decreased. It is indicated that the photocatalysis activity was lowered down little by little. It can be seen from Figure 4 that it is very important to control the humidity for the photocatalysis reaction of NO[2] because of the close relationship between rate of photocatalysis oxidation in gas-phase and the amount of water vapor in the gaseous mixture. In addition, since SO2 is dissolved in water easily, the removal efficiency of SO2 is not affected unless the amount of vapor varies remarkably. 2.4

Experimental results of simultaneous desulfurization and denitrification

Five parallel experiments of the simultaneous desulfurization and denitrification were carried out under the optimal experimental conditions, namely, irradiation time: 100 min; injection time of water vapor: 15 min; the content of O2: 10%; reaction temperature: room temperature; simulated flue gas flow: 0.064 m3·h−1. The results in Table 1 show higher removal efficiencies and high reproducibility; this can offer a good basis for industrial applications of this technology. Table 1 the results of simultaneous desulfurization and denitrification under optimal conditions Items SO2 concentration (mg·m−3) SO2 removal efficiency (%) NO concentration (mg·m−3) NO removal efficiency (%)

3 3.1

1 1622 98 568 67

2 1642 99 600 68

3 1635 98 572 66

4 1656 97 510 65

5 1041 98 550 69

mean 98 67

Mechanism of SO2 and NO removal SEM and EDS analysis of TiO2 photocatalyst

The scanned SEM pictures of supported TiO2 photocatalyst are shown in Figure 5, in which A 272

ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

and B are enlarged to 10000 and 1000 times, respectively. It is shown from A and B that particles of TiO2 are basically sphere, uniformly distributed on quartz sand But local aggregation appears, with particle size ranging from 500 nm to1 μm. This may result from aggregation of small umpty nanometer particle. The structure of membrane is denser. EDS energy spectrum of the fresh and spent supported TiO2 photocatalyst are shown in Figures 6 and 7, it can be seen that quartz sand, namely supporter is covered by substantial Ti, and relative contents of Si and Ti are close. It is demonstrated that TiO2 membrane is uniformly supported on the surface of quartz sand by using liquid phase deposit in this experiment. It is found from Figure 7 that sulfur species appear on the surface of spent photocatalyst, which may be sulfate. Peaks of Fe and Cl in Figure 7 may result from the corrosion of stainless steel reactor by SO2 during the reaction process.

Figure 5 SEM pictures of supported TiO 2 photocatalyst.

Figure 6 Energy spectrum on the surface of fresh supported TiO2 photocatalyst.

Figure 7 Energy spectrum on the surface of spent supported TiO2 photocatalyst.

ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

273

3.2

Analysis of removal products

Six samples of removal products were determined by chemical methods (Table 2). SO 24− , SO32− , NO3− and NO −2 were detected out in the absorber, of which SO42- and NO3- were of a high proportion. It is shown adequately that a rapid photocatalytic oxidation of SO2 and NO in the reactor occurred, in which partial SO2 was oxidized to SO3 and the most of NO was oxidized to NO2 and NO3. The oxidants probably are some hydroxyl radicals and active oxygen species resulted from TiO2 photocatalysis. The potentiality of removal of SO2 and NO from flue gas by TiO2 catalytic oxidation and solution resorption is verified. Table 2 Ions concentrations in the absorbera) Items

1

2

3

4

5

6

−4

−1

0.688

0.694

0.189

0.412

0.313

0.960

−4

−1

0.214

0.128

0.094

0.117

0.187

0.687

−4

−1

NO (10 mg·L )

29.385

20.356

36.722

26.071

21.329

20.417

NO −2 (10−4 mg·L−1)

0.196

0.161

0.374

0.200

0.161

0.019

SO

2− 4

SO

2− 3 − 3

(10 mg·L )

(10 mg·L )

a) The above data with the blank value having been taken out.

3.3

Mechanism of SO2 removal by TiO2 photocatalysis

It is reported by Luo et al.[11] from their experiment about adsorption behavior of flue gas desulfurization using TiO2 that when spent TiO2 was desorbed by heating, 98% of SO2 was escaped according to FTIR spectroscopy, demonstrating that the removal of SO2 is to be mainly physical adsorption, resulting from Van der Waals force between SO2 molecule and molecules absorbed on TiO2 surface[12,13]. At the same time, it is also found that there was trace amount of SO3 on spent TiO2 surface. This chemical absorption may be brought by chemical reaction between SO2 molecules and molecules on TiO2 surface[14]. As for SO2-O2-N2 system[15], SO2 molecules produce single state (1SO2) and triplet state (3SO2) upon UV irradiation. The latter produces SO3 by the reaction, 3SO2+O2→SO3+O directly, and there is deactivation of 3SO2[16]. The forming reaction of SO3 does not proceed due to the absence of 3SO2 without UV irradiation. Results from X-ray Photoelectron Spectroscopy (SPS) show that a broad-band absorption ranging from 400―600 nm stands for the electron transition from the valence band to the conduction band, namely character band. Nevertheless, the response in 400― 600 nm is related to surface property and ultramicro characteristic of TiO2, which is called surface band[17]. According to experimental results of XPS, dissymmetry of O1s peak indicates different oxygen bonding states on the surface of TiO2, the stronger peak in 529.5 eV belongs to O1s peak of Ti-O-Ti in the crystal lattice, the weaker peak in 532.2 eV corresponds to O1s peak of the surface ≡ Ti-OH hydroxyl[18]. While reaction temperature increases, the quantity of surface hydroxyl might disappear, and the rate of photocatalytic oxidation declines. According to our experiments, Figure 7 and Table 2 show that the great mass of SO2 in simulated flue das undergoes the conversion from SO2 to SO3, the remainder is absorbed by absorber. So, high removal efficiency for SO2 is obtained. Combined our investigation with other ― results[11 18], the mechanism of SO2 removal by TiO2 photocatalysis is inferred as follows:

274

ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

TiO 2 + hν → e − + h + , O 2(g) → O 2(ads) → 2O(ads) , − O(ads) + e − → O(ads) + h + → O*(ads) ,

O*(ads) + H 2 O(ads) → 2 ⋅ OH (ads) , ⋅OH (ads) + SO 2(ads) → HOSO 2(ads) , HOSO 2(ads) + O*(ads) → ⋅OH (ads) + SO3(ads) , SO3(ads) → SO3(g) . The key step of reaction rate is the reaction between ·OH and SO2[17]. 3.4

Mechanism of NO removal by TiO2 photocatalysis

Dalton et al.[19] studied the surface reaction of NO and TiO2 by Raman spectrometer and XPS. The results show that the anatase TiO2 has better crystalline defects and higher catalytic activity than rutile TiO2, so NO can be converted by the former to nitrate efficiently. Experimental results reported by Hashimoto et al.[20] show that some active oxygen species such as super oxide (O2−) were generated, when TiO2 was irradiated in the presence of O2, which reacted with NOx to nitrate. From Figure 2, though N species are not detected on the surface of TiO2, substantial NO3− exist in the removal products according to chemical analysis (Table 2) because NO is difficult to dissolve in water, and absorbed only by the oxidizing of ·OH or NO −2 . Reaction equations are presumed as follows: − TiO 2 + hν → TiO*2 (ecb + h +vb ), − eeb +O 2(ads) → O −2(ads) , − h +vb + OH (ads) → ⋅OH (ads) ,

NO(g) + 2 ⋅ OH (ads) → NO 2(ads) + H 2 O(ads) , − + NO 2(ads,g) + ⋅OH → NO3(ads) + H (ads) , − NO(ads) + O −2 → NO3(ads) ,

3NO 2 + 2OH − → 2NO3− + NO + H 2 O, [HNO3 ](ads) → HNO3(aq) .

It can be seen from equations mentioned above that NO would be oxidized efficiently, if enough hydroxyl radicals and super oxides were adsorbed on the surface of TiO2.

4

Conclusions

(1) Supported TiO2 photocatalyst was prepared by using liquid phase deposit. According to SEM and EDS the raletive contents of Si and Ti are close. It is demonstrated that TiO2 membrane is uniformly supported on the surface of quartz sand. Supported TiO2 are basically sphere, uniformly distributed on quartz sand, but local aggregation appears, its particle size ranges from 500 nm to1 μm, and the structure of membrane is denser. (2) The experiments of simultaneous removal of SO2 and NO were carried out at a ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276

275

self-designed experimental system using prepared supported TiO2 photocatalyst. The results show that irradiation time, injection time of water vapor, the content of O2, reaction temperature and simulated flue gas flow are main affecting factors for simultaneous desulfurization and denitrification. The efficiencies of removal for SO2 and NO can be achieved 98% and 67%, respectively, under optimum experimental conditions. (3) The results of five parallel experiments of the simultaneous desulfurization and denitrification show the higher removal efficiencies for SO2 and NO and high reproducibility; this can offer a basis for industrial applications of this technology. (4) According to SEM and EDS analysis, and combined with the results of removal products with chemical analysis, the mechanism of simultaneous desulfurization and denitrification based on TiO2 photocatalysis can be inferred as: the most of NO was rapidly oxidized to NO2 and NO3, and partial SO2 was oxidized to SO3, those oxides and SO2 were absorbed by absorber to form SO 24− , SO32− , NO3− and NO −2 . 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

276

Hoffmann M R, Martin S T, Choi W, et al. Environmental applications of semiconductor photocatalysis. Chem Rev, 1995, 95(1): 69―96 Zhu J Y, Zeng Y Y, Kuang D B, et al. Gas-phase photocatalytic of NOx under weak intensity UV irradiation. Natur Sci (in Chinese), 2001, 40(6): 31―34 Yu J C, Lin J, Kwok R W M. Enhanced photocatalytic activity of Ti1−xVxO2 solid solution on the degradation of acetone. J Photoch Photobio A, 1997, 111: 199―203 Romero M, Blanco J, Sanchez B, et al. Solar photocatalytic degradation of water and air pollutants: Challenges and perspectives. Solar Energy, 1999, 66: 169―182 Ibusui T, Takeuchi K. Removal of low concentration nitrogen oxides through photoassisted heterogeneous catalysis. J Mol Catal, 1994, 88: 93―102 Zhao W K, Fang Y L. Synthesis of photocatalytic activity TiO2 thin films at low-tempeture. Acta Phys Chem (in Chinese), 2002, 18(4): 368―371 National Environmental Protection Agency. Redaction Council of Analysis Methods of Water and Wastwater, Monitoring Methods of Water and Wastwater. Beijing: China Environmental Science Press, 2002. 164 National Environmental Protection Agency. Redaction Council of Analysis Methods of Water and Wastwater, Monitoring Methods of Water and Wastwater. Beijing: China Environmental Science Press, 2002. 271 Xu A W, Liu H Q, Li Y Q. Study on the photocatalytic oxidation degradation of NOx in the air. Chem J Chinese U (in Chinese), 2000, 21: 1252―1256 Bickley R I, Stone F S. Photoadsorption and photocatalysis at rutile surfaces. J Catal, 1973, 31(31): 389―397 Luo Y Q, Li D J, Huang Z. Preparation of titanium dioxide particles and properties for flue gas desulfurization. Environ Sci (in Chinese), 2003, 24(1): 147―151 Yang R T, Li W B, Chen N. Reversible chemisorption of nitric oxide in the presence of oxygen on titania and titania modified with surface sulfate. Appl Catal A-Gen, 1998, 169: 215―225 Lin Y S, Deng S G. Analysis of liquid chromatography with nonuniform crystallite particles. Aiche J, 1990, 36(10): 1569― 1576 Wang Y, Mohammed S O, Lavelley J C, et al. FTIR study of adsorption and reaction of SO2 and H2S on Na/SiO2. Appl Catal B-Environ, 1998, 16: 279―290 Shang J, Xu Z L, Du G Y, et al. Studies on photocatalytic oxidation reaction SO2 over TiO2. Chem J Chinese U, 2000, 21: 1299―1300 Tang X Y. Atmosphere Environmental Chemistry. Beijing: High Education Press, 1990. 124―125 Campostrini R, Carturan G, Palmisano L. Sol-gel derived anatase TiO2 morphology and photoactivity. Mater Chem Phys, 1994, 38(3): 277―283 Sanchez J, Augustynski J. X-ray photoelectron spectroscopic study of the interaction of various anions with the oxide-covered titanium metal. J Electroanal Chem, 1979, 103(26): 423―426 Dalton J S, Janes P A, et al. Photocatalytic oxidation of NOx gases using TiO2: A surface spectroscopic approach. Enviro Pollu, 2002, 120: 415―422 Hashimoto K, Wasada K, Toukai N, et al. Photocatalytic oxidation of nitrogen monoxide over titanium (VI) oxide nanocrystals large size areas. J Photoch Photobio A, 2000, 136: 103―109

ZHAO Yi et al. Sci China Ser E-Tech Sci | Mar. 2008 | vol. 51 | no. 3 | 268-276