TiO2 SCR catalyst: Comparison of K2SO4, KCl and K2O

Chemical Engineering Journal 348 (2018) 637–643

Contents lists available at ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

Effect of different potassium species on the deactivation of V2O5-WO3/TiO2 SCR catalyst: Comparison of K2SO4, KCl and K2O

T

⁎

Ming Kong, Qingcai Liu , Jian Zhou, Lijun Jiang, Yuanmeng Tian, Jian Yang, Shan Ren, Jiangling Li Engineering Research Center for Energy and Environment of Chongqing, College of Materials Science and Engineering, Chongqing University, Chongqing 400044, China

H I GH L IG H T S

G R A P H I C A L A B S T R A C T

mechanisms of three po• Deactivation tassium species on V O -WO /TiO 2

• • •

5

3

2

catalyst are compared. Deactivation rate of potassium-poisoned catalysts follows KCl > K2O > K2SO4. The introduction of SO42− creates new sites and contributes to catalyst acidity. Cl− provides new ammonia adsorption sites but is useless for NO reduction.

A R T I C LE I N FO

A B S T R A C T

Keywords: SCR catalysts Deactivation Potassium species Acid sites Redox ability

The effect of different potassium species, including K2O, KCl and K2SO4, on the deactivation of V2O5-WO3/TiO2 catalysts were measured in the research. Potassium species were supplied by impregnating catalyst with corresponding KNO3, KCl and K2SO4 aqueous solution. Catalytic activity was also measured and the physical and chemical properties of fresh and poisoned catalysts were characterized by XRD, N2 physisorption, NH3-TPD, H2TPR, XPS and NH3-DRIFTS. Deactivation could be observed in the potassium-containing catalysts and the deactivation rate follows KCl > K2O > K2SO4. The surface chemisorbed oxygen was also reduced and the downward trend was in good accordance with the SCR activity. In addition, both the amount and stability of the Brønsted and Lewis acid sites dropped after potassium introduction, and the reducibility of surface active species decreased. The introduction of SO42− created new Brønsted acid sites and performed advantages to ammonia adsorption and NO reduction. Cl− could react with vanadia active sites to form -O-V-Cl bond, which provided some new ammonia adsorption sites, but these newly generated sites could not revitalize the adsorbed ammonia and were helpless to catalytic activity promotion. Besides, KCl presented the maximum impact on catalyst reducibility was another contribution to its largest activity decline. Finally, the probable poisoning mechanisms of different potassium species over V2O5-WO3/TiO2 catalysts were proposed.

1. Introduction Combustion of most fuels results in the emission of nitrogen oxide ⁎

Corresponding author. E-mail address: [email protected] (Q. Liu).

https://doi.org/10.1016/j.cej.2018.05.045 Received 27 November 2017; Received in revised form 16 April 2018; Accepted 7 May 2018 Available online 08 May 2018 1385-8947/ © 2018 Elsevier B.V. All rights reserved.

(NOx), which causes serious threat to the environment because it would create issues of acid rain and photochemical smog [1–3]. Selective catalytic reduction (SCR) of NOx with NH3 has been proven as one of

Chemical Engineering Journal 348 (2018) 637–643

M. Kong et al.

the most effective technologies for NOx removal in the fuels-fired flue gas. Catalysts system, containing 0.5–3.0 wt% V2O5, 5–10 wt% WO3 (or MoO3) and TiO2 as the support, is the most widely applied on commercial for decades [4–7]. Vanadium is the active component and tungsten is the promoter that stabilizes the anatase titania, favors the spreading of vanadia on the catalyst surface and increases catalyst acidity. However, several issues concerning deactivation of the SCR catalysts are emergency to be considered. The catalyst unit usually works before flue gas purification equipment to satisfy temperature requirement. Therefore, large amount of fly ash and alkali (alkaline earth) metals in the flue gas would deactivate the catalyst [8–10]. Research have shown that alkali oxides and salts are the major components to deactivate SCR activity and the effect of alkali metals on V2O5-based catalysts has been widely studied [11–23]. Previous studies generally concluded that alkalis in the main group IA (K and Na) exhibit greater poisoning extent than that in the main group IIA (Ca and Mg). The deactivation of V2O5-WO3/TiO2 catalysts is usually caused by the neutralization of the BrØnsted acid sites, the decrease of surface chemisorption oxygen and the reduced reducibility of vanadium species. The way and state K+ is added to the catalyst also have an influence on its effect. Lei et al. [17] compared the deactivation of V2O5-WO3/TiO2 catalysts by KCl via three methods, wet impregnation, solid diffusion and vapor deposition. They found that the deactivation rate follows the order of vapor deposition ≫ solid diffusion > wet impregnation. Larsson et al. [23] compared the wet impregnation and the aerosol injection, they reported that the poisoning effect of wet impregnation was more serious than that of the aerosol injection. But Klimczak et al. [19] found that the wet impregnation and aerosol deposition of alkali metals caused the same deactivation behavior. However, in view of these studies, researchers primarily focused on comparison of different alkalis species [12,16,18–20] and different poisoning methods [17,19,23], or studies on single alkali metal [11,14,21,22]. Therefore, it was difficult to make a comparison or conclusion from these results due to the total different materials, preparation methods and properties characterizations. Based on this, the single alkali metal, potassium, was selected as the object in our research. Three main kinds of potassium species, K2O, KCl and K2SO4, were selected to be loaded on the catalysts. Wet impregnation method was used to deactivate catalyst due to the solubility of potassium in water-containing flue gas. The obtained poisoned catalysts were then characterized by the same measurements. It aimed at comparing the effect of different potassium species on the deactivation of V2O5-WO3/TiO2 catalysts at the same level and distinguishing their differences on the poisoning mechanism.

2.2. Catalyst characterization The powder X-ray diffraction (XRD) patterns were determined by Rigaku D/max-2500/PC diffractometer, operating at 40 kV and 40 mA using CuKα radiation. The BET surface area, pore size and pore volume of the samples were measured by N2 physisorption at 77 K using Micromeritics ASAP 2010 instrument. Temperature-programmed desorption of NH3 (NH3-TPD) and temperature-programmed reduction of H2 (H2-TPR) experiments were conducted on a chemisorption analyzer (AutoChem Ⅱ2920, Micromeritics Instrument). In a typical NH3-TPD experiment, 100 mg sample was pretreated in He (50 ml/min) for 1 h at 400 °C and then cooled down to 100 °C. Prior to temperature program desorption, the sample was exposed to gas mixture of 5% NH3 in He (50 ml/min) for 1 h and then purged in He until the TCD signal was stabilized. Finally, the sample was heated up to 400 °C (10 °C/min) under a flow of helium (50 ml/min) to record the NH3 TCD signal. For H2-TPR experiment, 100 mg of sample was placed in one arm of a U-shaped quartz tube. It was carried out under 5% H2 in Ar (50 ml/min) from 50 °C to 900 °C at a rate of 10 °C/min. X-ray photoelectron spectroscopy (XPS) profiles were acquired with a Thermo-Scientific system at room temperature using Al Kα radiation (1484.6 eV). The binding energy was referenced to the C 1 s line at 284.6 eV and peak deconvolution was performed using the Thermo Avantage v5.973 software. Diffuse reflectance infrared Fourier transform spectroscopy of NH3 (NH3-DRIFTS) was recorded on a Thermo Scientific Nicolet 6700 spectrometer, which was equipped with a Harrick IR cell and an MCT detector cooled by liquid N2. Prior to the experiment, catalysts were firstly purged at 350 °C for 1 h under N2 gas (total flow rate 100 ml/ min), and the background spectrum at desired temperature was collected NH3 was adsorbed at 50 °C with 500 ppm NH3/N2 flow for 1 h, and then flushed in N2 for 1 h. Before scanning, samples were kept for 30 min at the desired temperature and the DRIFT spectra were collected by accumulating 64 scans at a resolution of 4 cm−1. 2.3. Catalyst activity measurements SCR activity measurements were conducted in the fixed-bed reactor containing 200 mg of catalysts (40–60 meshes). The feed gas mixture contained 500 ppm NO, 500 ppm NH3, 4% O2, and N2 as the balance gas. The total gas flow rate was 500 ml/min. The outlet gas concentrations (NO, NOx and O2) were monitored by flue gas analyzer (MRU, Germany OPTIMA7), and the measurement was performed between 200 and 500 °C in intervals of 50 °C. Each data acquisition was preceded by equilibration for 20 min. The catalytic activities were evaluated by NOx conversion (%) according to the following equation:

2. Experimental 2.1. Catalyst preparation

NOx conversion(%) =

The fresh V2O5-WO3/TiO2 catalyst with 1 wt% V2O5 and 5 wt% WO3 was prepared by impregnation method. The ammonium metavanadate (NH4VO3) and ammonium paratungstate [(NH4)10(W12O41)·5H2O] were mixed in the oxalic solution of desired proportions, and commercial TiO2 (P25) was used as the precursor to obtain the slurry. The slurry was then stirred for 6 h, dried at 110 °C for 12 h and calcined at 500 °C for 5 h in air. The fresh catalyst (denoted as 1 V) was then grinded and sieved within 40–60 meshes for evaluation. The poisoned catalysts were prepared by wet impregnation with K2SO4, K2O (KNO3 was used as the precursor) and KCl solution, respectively. The potassium loadings were 0.5 wt%, 1.0 wt% and 2.0 wt% (denoted as KSx for K2SO4, KNx for K2O, and KCx for KCl, where x is the molar K/V ratio), corresponding to molar K/V ratios of 0.6, 1.2 and 2.4 respectively. The potassium-induced catalysts were then stirred for 6 h, dried at 110 °C for 12 h and calcined at 500 °C for 5 h in air. The samples with 1.0 wt% potassium loading were selected to do the characterization afterwards.

[NOx ] inlet −[NOx ] outlet × 100% [NOx ] inlet

(1)

3. Results and discussion 3.1. Catalytic activity Fig. 1 presents the catalytic activity of fresh and K+-poisoned catalysts with different potassium species and contents. The results show that fresh catalyst performs the best NO reduction efficiency that keeps above 90% in the temperature range of 200–450 °C. Introducing potassium deactivates the catalysts and the activity sequence follows the order of 1V > KS > KN > KC. The catalytic activity declines with increasing potassium contents and promotes with temperature elevating. When the K/V molar ratio is 0.6, K2O and K2SO4 perform negligible effect on NO reduction from 300 °C to 450 °C but presents severe deactivation below 300 °C. When the K/V molar ratio increases to 1.2, 638

Chemical Engineering Journal 348 (2018) 637–643

M. Kong et al.

100

1V KN0.6 KN1.2 KN2.4 KS0.6 KS1.2 KS2.4 KC0.6 KC1.2 KC2.4

60

1V

Intensity (a.u.)

NO Conversiton (%)

80

40

KC1.2

KS1.2

20

0 150

200

250

300

350

400

450

KN1.2

500

o

Temperature ( C)

0.0

Fig. 1. Catalytic activity of different K+-poisoned catalysts. Reaction conditions: 500 ppm NO, 500 ppm NH3, 4% O2, total flow gas 500 ml/min, GHSV = 60000 h−1.

0.2

0.4

0.6

0.8

1.0

Relative pressure (P/P 0) Fig. 3. N2 physisorption isotherms of different catalysts.

K2SO4-poisoned catalyst still keeps good activity between 300 °C and 400 °C and shows almost the same activity with KS0.6 sample from 200 °C to 400 °C, but the high-temperature stability deteriorates. This phenomenon can be attributed to the presence of SO42−, which provides some new active sites for NO reduction, but could react with NH3 to generate sulfate at high temperature and results in the activity deterioration. The catalytic activity and stability continue to decrease with K2SO4 content increasing to 2.0 wt%. Although KCl presents more serious deactivation effect than K2O on the V2O5/TiO2 catalysts, K2O performs considerable relationship to the potassium content, ie. the deactivation extent is more intensive as K2O content increasing from 1.0 wt% to 2.0 wt% comparing with KCl-poisoned catalysts.

Table 1 BET surface area, pore volume and pore size of different K+-poisoned catalysts. Catalyst

BET (m2/g)

Pore volume (cm3/g)

Pore size (nm)

1V KC1.2 KS1.2 KN1.2

81.3 78.9 75.4 77.6

0.31 0.25 0.25 0.24

10.03 12.55 12.99 12.21

decrease for the poisoned catalysts, but the pore size increases. It can be attributed to the blockage of catalyst pores resulting from potassium species, inhibiting the contact between reactant and catalyst active sites. The decreased BET surface area and increased pore size are disadvantageous for NO and NH3 adsorption on the catalysts. Considering the results of N2 physisorption and NO reduction, it seems that BET surface area is not the dominate reason for the catalyst deactivation.

3.2. Physical properties From the results of XRD patterns in Fig. 2, only anatase TiO2 were detected, indicating the finely dispersed or amorphous existence of V2O5, WO3 and all potassium species, on the surface of TiO2-support. Besides, all catalysts present similar N2 physisorption isotherms in Fig. 3 that belong to typical type IV with H1 hysteresis loops according to the IUPAC, which is dominant in ordered mesoporous materials with uniform cylindrical pores [24]. Introducing different potassium species don’t change the type of catalyst pores. The BET surface area, pore volume and pore size of fresh and K+-poisoned catalysts are listed in Table 1, and it is obvious that the BET surface area and pore volume

3.3. NH3-TPD and H2-TPR results Fig. 4 presents the NH3-TPD curves and corresponding NH3 desorption amount of fresh and K+-poisoned catalysts. The 1 V catalyst shows the largest NH3 desorption amount and the signal still hold until

Peak 1 ( 145 oC)

1V KC1.2 KS1.2 KN1.2

Peak 2 ( 250 oC)

KN1.2

1.5 1.32

3

Amount of NH 3 desorption (cm /g)

Intensity (a.u.)

Intensity (a.u.)

anatase TiO2

KS1.2

1.0 0.68 0.64 0.5

0.0

1V

0.55

KC1.2 KS1.2 KN1.2

Catalysts

KC1.2 1V

100 20

40

60

150

200

250

300

350

400

Temperature ( oC)

80

2

Fig. 4. NH3-TPD profiles and the NH3 desorption amount of fresh and K+poisoned catalysts.

Fig. 2. XRD patterns of different catalysts. 639

Chemical Engineering Journal 348 (2018) 637–643

M. Kong et al.

: 530.1eV

400 °C. After introduction of different potassium species, the NH3 desorption amount decreases distinctly, and each curve can be divided into two peaks, peak 1 at around 145 °C and peak 2 at around 250 °C, which are corresponding to the weakly and strongly bound ammonia. The weakly bound ammonia can be assigned to the BrØnsted acid sites and the strongly bound ammonia can be assigned to the Lewis acid sites. Among these three K+-poisoned catalysts, the sequence for the amount of the BrØnsted acid sites follows KS > KC > KN, whereas the Lewis acid sites follows KC > KS > KN, which are not correlated with the catalytic activity. The KCl-poisoned catalyst performs the worst catalytic activity but shows larger amount of weakly adsorbed ammonia than K2O-poisoned catalyst and the largest amount of strongly adsorbed ammonia. It is because the introduction of Cl− provides more ammonia adsorption sites (Cl-). However, combined with the results of SCR activity above, these sites are not active and useless for NO reduction. For the K2SO4-poisoned catalyst, although potassium occupies the acid sites, the introduction of SO42− supplies the lost sites, and it belongs to the BrØnsted acid sites, which is advantageous to the catalytic activity. Research has shown that, not only the acid sites are important to the catalysts, but the redox ability of vanadium species is also considered as another key property for SCR reaction [25]. Therefore, the reducibility of vanadium and tungsten species was measured and the results are presented in Fig. 5. Two apparent reduction peaks, centered at 481 °C and 775 °C, are observed on 1 V catalyst, which can be assigned to the reduction of V5+ to V3+ and W6+ to W0 respectively [26,27]. However, it is clear that the main reduction peaks of potassium-poisoned catalysts shift to higher temperature in the range of 585–590 °C and 830–856 °C. Simultaneously, an obvious shift of the onset point toward higher temperatures is also observed for poisoned catalysts. It means the reducibility of vanadium and tungsten species is weakened. This phenomenon can be explained by potassium entering into V2O5 and WO3 lattice to form strong chemical bond, which results in the decrease of lattice oxygen concentration or inhibits the release of lattice oxygen, leading to the vanadium and tungsten species harder to be reduced. In comparison, for these three K+-poisoned catalysts, the reduction peaks position of vanadium species doesn’t show much different changes, but present larger differences on the onset point and tungsten species in KCl-poisoned catalyst. This result reveals that KCl performs stronger poisoning on influencing the catalyst reducibility. It also can be contributed to the evident activity decrease of KCl-poisoned catalyst in our research.

O : 531.3eV

Intensity (a.u.)

1V

KC1.2

538 537 536 535 534 533 532 531 530 529 528 527 526 Binding energy (eV) Fig. 6. XPS spectrum of O1s over different catalysts.

vanadium and tungsten species are influenced along with the addition of different potassium species, which is correlated to the surface oxygen. Therefore, the state of surface oxygen was analyzed and the concentration ratio of chemisorbed oxygen (Oα/Oα + Oβ) was calculated due to its crucial function in the catalytic oxidation reaction [28]. The results are shown in Fig. 6. The O1s peaks could be fitted into two peaks assigning to lattice oxygen (denoted as Oβ) at 530.0–530.4 eV and surface chemisorbed oxygen (denoted as Oα) at 531.1–531.5 eV [20]. It can be obviously seen that the Oα/(Oα + Oβ) ratio decreases after potassium loading, and the order is shown as follows: 1V > KS1.2 > KN1.2 > KC1.2, which is in accordance with the variation of SCR activity. These changes can be explained by the formation of strong bonds between surface oxygen center and potassium, resulting in the reducibility decline of surface species. 3.5. NH3-DRIFTS Researchers believed that potassium could neutralize the acid sites to deactivate the catalysts, so DRIFTS measurements of NH3 adsorption were carried out to clarify the effect of different potassium species on the acid sites over V2O5-WO3/TiO2 catalysts, and the results are presented in Fig. 7. The broad band in the range of 3100–3400 cm−1, 1100–1300 cm−1 and peaks at ∼1604 cm−1 belong to NH3 adsorption on the Lewis acid sites, whereas peaks at around 3058 cm−1, 1844 cm−1, 1697 cm−1 and 1454 cm−1 can be assigned to NH4+ on the BrØnsted acid sites [29,30]. It is clear that intensity of the main peaks in 1 V catalyst that assigned to the BrØnsted acid sites (3058, 1844 and 1454 cm−1) and the Lewis acid sites (3100–3400 and 1100–1300 cm−1) are more intense than that in K+-poisoned catalysts, indicating all kinds of potassium species influence ammonia adsorption on both the BrØnsted and Lewis acid sites. However, it is interesting that the peaks of poisoned catalysts at 1697 cm−1 (assigned to the symmetric bending vibrations of NH4+ on the BrØnsted acid sites) and 1604 cm−1 (assigned to the asymmetric bending vibrations of N-H bond in NH3 on Lewis acid sites) become more intense than that of fresh catalyst due to the introduction of Cl− and SO42−, but they perform only a bit function to the ammonia adsorption. Besides, with temperature elevating, the peaks become less intense for all samples, especially for the poisoned catalysts. It means the stability of the adsorbed ammonia on the poisoned catalysts is weakened. In the case of 1 V catalyst, the signals of both the BrØnsted- and Lewis-bonded ammonia could be detected until 350 °C, and the similar phenomenon could also be observed in KS1.2 and KC1.2 samples, but in KN1.2 sample the signals almost vanish. This result is coincident with the NH3-

Considering the H2-TPR results above, the redox properties of W6+ W0

onset point

775

481 ~370

1V

Intensity (a.u.)

590 ~460

KN1.2

830

585 833

856 ~456

588

KS1.2 ~487

KC1.2

200

300

400

500

600

700

800

KS1.2

KN1.2

3.4. XPS of O1s

V5+ V3+

O

900

Temperature (oC) Fig. 5. H2-TPR profiles of fresh and K+-poisoned catalysts with different potassium species. 640

Chemical Engineering Journal 348 (2018) 637–643

M. Kong et al.

(a) 1.4

(b) 1.4

+

Br nsted-NH4

1V

+

KS1.2

Lewis-NH3

Lewis-NH3

Br nsted-NH4

Lewis-NH3

Lewis-NH3

o

350 C o

350 C

Intensity (a.u.)

Intensity (a.u.)

o

250 C

o

250 C

o

150 C

o

100 C

o

150 C o

100 C o

50 C

o

50 C

Lewis-NH3

Lewis-NH3

0.0 4000

3500

3000

2500

2000

1500

0.0 4000

1000

-1

3500

3000

2500

Wavenumber (cm ) (c) 1.4

KC1.2

Br nsted-NH4

Lewis-NH3

2000

1500

1000

-1

Wavenumber (cm )

(d) 1.4

+

KN1.2

Lewis-NH3

+

Br nsted-NH4 Lewis-NH3

o

350 C

Lewis-NH3

o

350 C o

Intensity (a.u.)

Intensity (a.u.)

250 C o

250 C o

150 C

o

150 C o

100 C

o

100 C o

o

50 C

50 C Lewis-NH3

0.0 4000

3500

3000

2500

2000

1500

Lewis-NH3

1000

0.0 4000

3500

-1

Wavenumber (cm )

3000

2500

2000

1500

1000

-1

Wavenumber (cm )

Fig. 7. NH3-DRIFTS spectra of fresh and K+-poisoned catalysts with different potassium species collected from 50 °C to 350 °C.

TPD profiles because of the introduction of Cl− and SO42−. Cl− performs properties of the BrØnsted and Lewis acid sites, which is advantageous to ammonia adsorption, but the adsorbed ammonia can hardly be activated and have no contribution to participate in the NO reduction reaction. It is the pseudo acid sites. SO42− belongs to the BrØnsted acid and it shows promotion performance to catalytic activity at low temperature.

Then, the produced HCl react with V]O to form Cl-V-O-H, which can proceed to react with KCl. Finally, Cl-V-O-K and -V-O-K are generated. It results in the deactivation of active sites. The newly formed Cl- bond also exhibits ammonia adsorption ability on the catalyst, but it is not active to revitalize the adsorbed ammonia. For the K2SO4-poisoned catalyst, K+ occupies the place of H+ on VOH to deactivate the catalyst. However, a new -OH bond can be generated on the introducing SO42−, and it provides the new sites for ammonia adsorption and activation to react with NO, which is the reason for the inconspicuous deactivation on K2SO4-poisoned catalyst. Therefore, although potassium causes the deactivation of V2O5WO3/TiO2 catalysts, the way and extent of different potassium species worked over the catalysts are totally different.

3.6. Deactivation mechanism of different potassium species According to the above analysis, poisoning mechanism of different potassium species on V2O5-WO3/TiO2 catalysts is presented in Fig. 8. In the case of K2O poisoning, K+ combines with both the Brønsted acid sites (V–OH) and the Lewis acid sites (V]O) to form -V-O-K, and simultaneously generates H2O. This reaction results in the active site inactive for NH3 adsorption, and then catalytic activities decline. For the KCl-poisoned catalyst, the V-OH bond firstly react with KCl to form -V-O-K and HCl. Generally, the newly formed HCl could adsorb on the V active sites to form -OH groups, which has been confirmed in previous researches [31,32]. The reactions of HCl and vanadium oxides can be described as following:

VO2 + 2HCl → V(OH)2 Cl2

(2)

V(OH)2 Cl2 → VOCl2 + H2 O

(3)

V2O5 + 2HCl → V2O3 (OH)2 Cl2

(4)

V2O3 (OH)2 Cl2 → VO2 Cl2 + H2 O

(5)

V2O5 + 2HCl → 2V(OH)2 Cl

(6)

4. Conclusion The deactivation mechanisms of V2O5-WO3/TiO2 catalysts by different potassium species were investigated, compared, and analyzed in this work. KCl, K2SO4 and K2O all presented deactivation effect on the catalysts, but the deactivation rate was totally different. The catalytic activity order followed 1 V > KS > KN > KC. K2SO4 performed the lightest poisoning to the catalyst due to the introduction of SO42−, but its high-temperature stability was weakened because the probable reaction between SO42− and adsorbed ammonia. The BET surface area decreased after introducing potassium, but it wasn’t the main reason responsible for catalytic deactivation. Moreover, in the case of poisoned catalysts, both the amount and stability of the Brønsted and Lewis acid sites dropped, and the reducibility of surface active species decreased. The surface chemisorbed oxygen was also reduced and the downward 641

Chemical Engineering Journal 348 (2018) 637–643

M. Kong et al.

NH3 H O

O

K2 O

O V O V O

K

K

O

O

K

H O

+ H2O

O V O V O

O

KCl

O V O V O

O

O + HCl

O V O V O NH3

O

H

K

O

O

V

O

V

O

KCl

O

K

K

O

O

V

O

Cl

Cl

V

+ HCl

O

NH3 NH3

H O O

V

O O

V

NH3

K

O

O

K2SO4

O

V

O O

V

OH O

+ S6+ O O O

Fig. 8. Schematic diagram of K2O, KCl and K2SO4 poisoning mechanism on the V2O5–WO3/TiO2 SCR catalysts.

trend was in good accordance with the SCR activity. The introduction of SO42− was advantageous to the weakly adsorbed ammonia and performed the same function as the Brønsted acid, while Cl− was good for both weakly and strongly adsorbed ammonia. However, ammonia adsorbed on the Cl- bond was inactive and had no contributions to NO reduction. What’s more, KCl performed maximum impact on the catalyst reducibility, which was another reason for its worst SCR activity. Finally, the deactivation mechanisms of different potassium species on V2O5–WO3/TiO2 catalysts were suggested.

[6] G.T. Went, L.J. Leu, A.T. Bell, Quantitative structural analysis of dispersed vanadia species in TiO2 (anatase)-supported V2O5, J. Catal. 134 (1992) 479–491. [7] J.P. Chen, R.T. Yang, Role of WO3 in mixed V2O5–WO3/TiO2 catalysts for selective catalytic reduction of nitric oxide with ammonia, Appl. Catal. A 80 (1992) 135–148. [8] J. Einvall, S. Albertazzi, C. Hulteberg, A. Malik, F. Basile, Investigation of reforming catalyst deactivation by exposure to fly ash from biomass gasification in laboratory scale, Energy Fuels 21 (2007) 2481–2488. [9] S.A. Benson, J.D. Laumb, C.R. Crocker, J.H. Pavlish, SCR catalyst performance in flue gases derived from subbituminous and lignite coals, Fuel Process. Technol. 86 (2005) 577–613. [10] R. Khodayari, C.U.I. Odenbrand, Regeneration of commercial TiO2–V2O5–WO3 SCR catalysts used in bio fuel plants, Appl. Catal. B 30 (2001) 87–99. [11] H. Kamata, K. Takahashi, C.U.I. Odenbrand, The role of K2O in the selective reduction of NO with NH3 over a V2O5 (WO3)/TiO2 commercial selective catalytic reduction catalyst, J. Mol. Catal. A: Chem. 139 (1999) 189–198. [12] J.P. Chen, R.T. Yang, Mechanism of poisoning of the V2O5/TiO2 catalyst for the reduction of NO by NH3, J. Catal. 125 (1990) 411–420. [13] O. Kröcher, M. Elsener, Chemical deactivation of V2O5/WO3–TiO2 SCR catalysts by additives and impurities from fuels, lubrication oils, and urea solution. I. Catalytic studies, Appl. Catal. B 75 (2008) 215–227. [14] L. Lietti, P. Forzatti, G. Ramis, Potassium doping of vanadia/titania de-NOxing catalysts: surface characterization and reactivity study, Appl. Catal. B 3 (1993) 13–35. [15] L. Lisi, G. Lasorella, S. Malloggi, Single and combined deactivating effect of alkali metals and HCl on commercial SCR catalysts, Appl. Catal. B 50 (2004) 251–258. [16] D. Nicosia, I. Czekaj, O. KrÖcher, Chemical deactivation of V2O5/WO3–TiO2 SCR catalysts by additives and impurities from fuels, lubrication oils and urea solution. Part II. Characterization study of the effect of alkali and alkaline earth metals, Appl. Catal. B 77 (2008) 228–236. [17] T.Y. Lei, Q.C. Li, S.F. Chen, Z.Y. Liu, Q.Y. Liu, KCl-induced deactivation of V2O5–WO3/TiO2 catalyst during selective catalytic reduction of NO by NH3: comparison of poisoning methods, Chem. Eng. J. 296 (2016) 1–10. [18] L. Chen, J.H. Li, M.F. Ge, The poisoning effect of alkali metals doping over nano V2O5–WO3/TiO2 catalysts on selective catalytic reduction of NOx by NH3, Chem. Eng. J. 170 (2011) 531–537. [19] M. Klimczak, P. Kern, T. Heinzelmann, M. Lucas, P. Claus, High-throughput study of the effect of inorganic additives and poisons on NH3-SCR catalysts-Part I: V2O5–WO3/TiO2 catalysts, Appl. Catal. B 95 (2010) 39–47.

Acknowledgements The authors wish to express their sincere appreciate for the financial support from the National Natural Science Foundation of China (No. 51274263 and No. 51204220) and Chongqing Science & Technology Commission (cstc2017shms-zdyfx0055). References [1] M. Shelef, Selective catalytic reduction of NOx with N-free reductants, Chem. Rev. 95 (1995) 209–225. [2] G. Busca, L. Lietti, G. Ramis, F. Berti, Chemical and mechanistic aspects of the selective catalytic reduction of NOx by ammonia over oxide catalysts: a review, Appl. Catal. B 18 (1998) 1–36. [3] V.I. Parvulescu, P. Grange, B. Delmon, Catalytic removal of NO, Catal. Today 46 (1998) 233–316. [4] L. Casagrande, L. Lietti, I. Nova, P. Forzatti, A. Baiker, SCR of NO by NH3 over TiO2supported V2O5–MoO3 catalysts: reactivity and redox behavior, Appl. Catal. B 22 (1999) 63–77. [5] L. Lietti, P. Forzatti, F. Bregani, Steady-state and transient reactivity study of TiO2supported V2O5–WO3 De-NOx catalysts: relevance of the vanadium–tungsten interaction on the catalytic activity, Ind. Eng. Chem. Res. 35 (1996) 3884–3892.

642

Chemical Engineering Journal 348 (2018) 637–643

M. Kong et al.

[26] C.Z. Wang, S.J. Yang, H.Z. Chang, Y. Peng, J.H. Li, Dispersion of tungsten oxide on SCR performance of V2O5-WO3/TiO2: acidity, surface species and catalytic activity, Chem. Eng. J. 225 (2013) 520–527. [27] S.J. Yang, C.Z. Wang, J.H. Li, et al., A novel magnetic Fe–Ti–V spinel catalyst for the selective catalytic reduction of NO with NH3 in a broad temperature range, Catal. Sci. Technol. 2 (2012) 915–917. [28] L.Q. Jing, Z.L. Xu, X.J. Sun, J. Shang, W.M. Cai, The surface properties and photocatalytic activities of ZnO ultrafine articles, Appl. Surf. Sci. 180 (2001) 308–314. [29] N.Y. Topsøe, H. Topsøe, J.A. Dumesic, Vanadia/titania catalysts for selective catalytic reduction (SCR) of nitric oxide by ammonia. I. Combined temperature programmed in situ FTIR and on-line mass spectroscopy studies, J. Catal. 151 (1995) 226–240. [30] N.Y. Topsøe, J.A. Dumesic, H. Topsøe, Vanadia/titania catalysts for selective catalytic reduction (SCR) of nitric oxide by ammonia. II. Studies of active sites and formulation of catalytic cycles, J. Catal. 151 (1995) 241–252. [31] S. Niksa, N. Fujiwara, A predictive mechanism for mercury oxidation of selective catalytic reduction catalysts under coal-derived flue gas, J. Air Waste Manage. 55 (2005) 1866–1875. [32] S. He, J. Zhou, Y. Zhu, Z. Luo, M. Ni, K. Cen, Mercury oxidation over a vanadia based selective catalytic reduction catalyst, Energy Fuels 23 (2009) 253–259.

[20] Q. Wan, L. Duan, J.H. Li, L. Chen, K.B. He, J, M, Hao, Deactivation performance and mechanism of alkali (earth) metals on V2O5–WO3/TiO2 catalyst for oxidation of gaseous elemental mercury in simulated coal-fired flue gas, Catal. Today 175 (2011) 189–195. [21] F. Castellino, A.D. Jensen, J.E. Johnsson, R. Fehrmann, Influence of reaction products of K-getter fuel additives on commercial vanadia-based SCR catalysts part I. Potassium phosphate, Appl. Catal. B 86 (2009) 196–205. [22] Y.J. Zheng, A.D. Jensen, J.E. Johnsson, J.R. Thogersen, Deactivation of V2O5-WO3TiO2 SCR catalyst at biomass fired power plants: elucidation of mechanisms by laband pilot-scale experiments, Appl. Catal. B 83 (2008) 186–194. [23] A.C. Larsson, J. Einvall, A. Andersson, M. Sanati, Targeting by comparison with laboratory experiments the SCR catalyst deactivation process by potassium and zinc salts in a large-scale biomass combustion boiler, Energy Fuels 20 (2006) 1398–1405. [24] K.A. Cychosz, R.G. Nicolas, J.G. Martines, M. Thommes, Recent advances in the textural characterization of hierarchically structured nanoporous materials, Chem. Soc. Rev. 46 (2017) 389. [25] F.S. Tang, B.L. Xu, H.H. Shi, J.H. Qiu, Y.N. Feng, The poisoning effect of Na+ and Ca2+ ion doped on the V2O5-WO3-TiO2 catalysts for selective catalytic reduction of NO by NH3, Appl. Catal. B 94 (2010) 71–76.

643