SO2-tolerant Selective Catalytic Reduction of NOx over Meso-TiO2

Subscriber access provided by Université de Strasbourg - Service Commun de la Documentation

Remediation and Control Technologies

SO2-tolerant Selective Catalytic Reduction of NOx over Meso-TiO2@Fe2O3@Al2O3 Metal-based Monolith Catalysts Lupeng Han, Min Gao, Jun-ya Hasegawa, Shuangxi Li, Yongjie Shen, Hongrui Li, Liyi Shi, and Dengsong Zhang Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.9b00435 • Publication Date (Web): 07 May 2019 Downloaded from http://pubs.acs.org on May 8, 2019

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38

Environmental Science & Technology

1

SO2-tolerant Selective Catalytic Reduction of NOx over Meso-

2

TiO2@Fe2O3@Al2O3 Metal-based Monolith Catalysts

3

Lupeng Han†§, Min Gao‡§, Jun-ya Hasegawa‡, Shuangxi Li†, Yongjie Shen†, Hongrui Li†, Liyi Shi†

4

and Dengsong Zhang†*

5

†Department

6

School of Materials Science and Engineering, Shanghai University, Shanghai, 200444, China.

7

‡Institute

8

§These

of Chemistry, College of Sciences, Research Center of Nano Science and Technology,

for Catalysis, Hokkaido University, Sapporo 001-0021, Japan.

authors contributed equally to this work.

9

1

ACS Paragon Plus Environment


Page 2 of 38

10

ABSTRACT

11

It is an intractable issue to improve the low-temperature SO2-tolerant selective catalytic reduction

12

(SCR) of NOx with NH3 because deposited sulfates are difficult to decompose below 300 oC. Herein,

13

we established a low-temperature self-prevention mechanism of mesoporous-TiO2@Fe2O3 core-shell

14

composites against sulfate deposition using experiments and density functional theory. The

15

mesoporous TiO2-shell effectively restrained the deposition of FeSO4 and NH4HSO4 because of weak

16

SO2 adsorption and promoted NH4HSO4 decomposition on the mesoporous-TiO2. The electron transfer

17

at the Fe2O3 (core)-TiO2 (shell) interface accelerated the redox cycle, launching the “Fast SCR”

18

reaction, which broadened the low-temperature window. Engineered from the nano- to macro-scale,

19

we achieved one-pot self-installation of mesoporous-TiO2@Fe2O3 composites on the self-tailored

20

AlOOH@Al-mesh monoliths. After the thermal treatment, the mesoporous-TiO2@Fe2O3@Al2O3

21

monolith catalyst delivered a broad window of 220-420 oC with NO conversion above 90% and had

22

superior SO2 tolerance at 260 oC. The effective heat removal of Al-mesh monolith catalysts restrained

23

NH3 oxidation to NO and N2O while suppressing the decomposition of NH4NO3 to N2O, and this led

24

to much better high-temperature activity and N2 selectivity. This work supplies a new point for the

25

development of low-temperature SO2-tolerant monolithic SCR catalysts with high N2 selectivity,

26

which is of great significance for both academic interests and practical applications.

27

2


Page 3 of 38


28

1. INTRODUCTION

29

Concerns about acid rain, PM2.5 pollution, and photochemical smog caused by excessive emissions

30

of nitrogen oxides (NOx) from plants and automobiles have motivated strong interests in controlling

31

NO emission.1-3 Selective catalytic reduction (SCR) of NO with NH3 over commercial V2O5-

32

WO3(MoO3)/TiO2 catalysts has been widely used for the abatement of NOx in power plants. However,

33

V-based catalysts have such drawbacks as narrow operation windows (300-400 °C), poor alkaline-

34

resistance, and biological toxicity of V2O5.4-6 Moreover, the formation of (NH4)2SO4/NH4HSO4 and

35

metal sulfates in SO2-containing flue gases can lead to the deactivation of catalysts by blocking and

36

destroying active sites. Because it is generally difficult to decompose sulfates under low

37

temperatures,7-9 it is of great significance to develop green SCR catalysts with wide temperature

38

windows and strong SO2 resistance at low temperatures.

39

Thus far, it has been relatively effective to improve the SO2 tolerance via doping/modification with

40

rare earth oxides or transition metal oxides. Sm doping improves the SO2 tolerance of MnOx-TiO2

41

catalysts because the formation of SO3 is inhibited via the favorable electron transfer from Sm2+ to

42

Mn4+.10 Ni- or Co-modified MnOx-CeO2 catalysts promote the formation of active bidentate nitrate

43

species that are not affected by SO2.11 CeO2 can be used as sacrificial sites to improve the SO2 tolerance

44

of the Mn-Ce/TiO2 catalyst because sulfates are preferentially generated on CeO2. Besides, ammonium

45

sulfates decompose more easily because of the decreased binding energy between NH4+ and sulfate

46

ions associated with the CeO2 doping.12 Furthermore, SiO2-modified CeO2/TiO2,13 Sn-modified

47

MnOx-CeO2,14 Nb- or Eu-modified Mn/TiO215,16 and Ho-modified Fe-Mn/TiO217 catalysts all show

48

improved SO2 tolerance. Despite these encouraging results, the activity of the catalysts mentioned

49

above actually decrease obviously in the presence of SO2.10-17 3



Page 4 of 38

50

Fe-based catalysts, such as Fe-W,18 Fe-V,19 and Fe-Ti,20 have attracted increasing attention because

51

of good medium-high temperature activity and N2 selectivity. Moreover, Fe-based catalysts show

52

satisfactory SO2 tolerance at high temperature (>300 oC) because of the promotion effect of the formed

53

FeSO4 (which serve as acid sites).21 It was found that the NH3-SCR reaction over the WOx/Fe2O3

54

catalyst follows the Eley–Rideal mechanism. Specifically, the active adsorbed NH3 species can react

55

directly with gaseous NO, and this leads to high SO2 tolerance because NOx does not need to

56

competitively adsorb onto the catalyst surface as nitrates or nitrites.22 Chen et al.23 found that

57

Fe2O3/MoO3 with specific layer-structured MoO3 can trap NH4+ from NH4HSO4, thereby promoting

58

the decomposition of NH4HSO4. We developed Fe2O3-promoted CeO2-WO3/halloysite catalysts which

59

prevented the irreversible bonding of SO2 with the active components.24 Currently, these newly-

60

developed Fe-based catalysts only achieve good SO2 tolerance at or above 300 oC.20-24 Superior SO2

61

tolerance at lower temperature (< 300 °C) is still challenging for Fe-based catalysts. Otherwise, the

62

adverse effects of the exothermic NH3-SCR reaction (Δ H 298 = -406.8 kJ/mol NO consumed) are

63

generally ignored. A high amount of released heat can generate hot spots in powder catalysts, and this

64

may lead to excessive oxidation of NH3 to NOx and also SO2 to SO3, which are associated with

65

decreased activity and selectivity. From academic and practical application viewpoints, it is worth

66

making efforts to develop low-temperature SO2-tolerant SCR monolith catalysts with good N2

67

selectivity and good heat-transfer properties.25



68

In this study, Fe2O3 encapsulated in mesoporous-TiO2 (m-TiO2@Fe2O3) was one-pot installed on

69

the self-tailored AlOOH@Al-mesh monolith substrates using titanate cross-linking agents. After

70

thermal treatments, the m-TiO2@Fe2O3@Al2O3 monolith catalyst showed a broad temperature

71

window and good N2 selectivity as well as superior SO2 tolerance at 260 oC. From detailed 4


Page 5 of 38


72

characterizations and density functional theory (DFT) calculations, the mechanisms of promoting

73

activity/N2 selectivity and self-prevention against SO2-poisoning were proposed.

74

2. MATERIALS AND METHODS

75

Self-tailored AlOOH@Al-mesh was synthesized according to the previous report.26 First, Al-mesh

76

was cut into pieces (8 cm×1 cm) and rolled into cylinders, each with a diameter of 8 mm. The Al-mesh

77

cylinders were pretreated with a 0.1 wt% NaOH aqueous solution and then washed several times with

78

deionized water. As-obtained Al-mesh was treated with vapor steam at 120 oC for 12 h, and

79

AlOOH@Al-mesh was obtained.

80

The m-TiO2@Fe2O3@Al2O3 Al-mesh monolith catalyst was prepared via a one-pot self-assembly

81

method using the titanate cross-linking molecule titanium bis(triethanolamine)diisopropoxide. A

82

certain amount of ferric nitrate was first dissolved in methanol. The titanate cross-linking agent was

83

then added to the solution in a certain molar ratio of Fe/Ti to form Fe3+-titanate complexes because Ti-

84

O-CH2-CH2-N-(CH2-CH2-OH)2 groups of titanate could chelate with Fe3+ ions. When the

85

AlOOH@Al-mesh was added to the solution of Fe3+-titanate using an incipient impregnation manner,

86

a silanization reaction took place between the alkoxy groups of Ti-O-CH-(CH3)2 and the surface

87

hydroxyl (-OH) groups of AlOOH@Al-mesh, leading to the formation of Al-O-Ti bonds. Some water

88

was then added to the system, and the rest Ti-O-CH-(CH3)2 and Ti-O-CH2-CH2-N-(CH2-CH2-OH)2

89

polymerized to produce Ti-O-Ti bonds. After drying at 100 oC for 2 h and calcining at 500 oC for 3 h

90

in air, the m-TiO2@Fe2O3@Al2O3 (Fe/Ti =4,3,2) monolith catalyst was obtained. The obtained

91

catalysts with different Fe/Ti mole ratios were denoted as m-TiO2@Fe2O3@Al2O3 (Fe/Ti =4,3,2)

92

monolith. The m-TiO2@Fe2O3/Al2O3 core-shell powder catalyst was prepared using commercial

93

AlOOH as a support using the same method for the m-TiO2@Fe2O3@Al2O3 monolith catalyst. 5



Page 6 of 38

94

The Fe2O3@Al2O3 monolith catalysts and Fe2O3/Al2O3 powder catalysts were prepared using an

95

incipient wetness impregnation method. Fe2O3 was loaded onto the AlOOH@Al-mesh composites or

96

AlOOH powders using an incipient wetness impregnation method with Fe(NO3)3·9H2O dissolved in

97

methanol. Some water was then added to the system, which was dried at 100 oC for 2 h. After

98

calcination at 500 oC for 3 h in air, the Fe2O3@Al2O3 monolith catalyst and Fe2O3/Al2O3 powder

99

catalyst were obtained. The catalysts that were pre-impregnated with 5 wt% NH4HSO4 were prepared

100

using an incipient wetness impregnation method followed by drying at 100 oC for 2 h.

101

The information about SCR activity tests, density functional theory (DFT) calculations, scanning

102

electron microscopy (SEM), transmission electron microscopy (TEM), high angle annular dark-field

103

scanning transmission electron microscopy (HAADF-STEM), X-ray powder diffraction (XRD),

104

Mössbauer spectra, N2 adsorption-desorption isotherms, X-ray photoelectron spectroscopy (XPS),

105

extended X-ray absorption fine structure (EXAFS), X-ray-absorption near-edge structures (XANES),

106

H2 temperature-programmed reduction (H2-TPR), NH3 temperature-programmed desorption (NH3-

107

TPD), NO + O2-TPD, SO2 + O2-TPD, SO2 + O2 + NH3-TPD, and NH4HSO4 temperature-programmed

108

decomposition (NH4HSO4-TPDC) and in situ diffuse reflectance infrared Fourier transform (DRIFT)

109

measurements can be found in the Supporting Information (SI).

110

3. RESULTS AND DISCUSSION

111

3.1. Textural and structural properties of catalysts. As seen in Figure 1A, the m-TiO2@Fe2O3

112

composites were installed on the self-tailored AlOOH@Al-mesh substrates using titanate cross-linking

113

agents that have alkoxy (which condense with surface hydroxy groups of AlOOH) and hydroxyl/N-

114

containing (which chelate with Fe3+ ions) groups. After thermal treatments, the m-

115

TiO2@Fe2O3@Al2O3 core-shell monolith catalysts were obtained. The catalysts were shaped into 6


Page 7 of 38


116

cylinders, each with a diameter of 8 mm (Figure 1B); the catalysts had Al skeletons of 60 μm with a

117

regular cross structure (Figure S1). The m-TiO2@Fe2O3 composites were anchored on the surface of

118

Al2O3 nanosheets, which were derived from the AlOOH-nanosheet@Al-mesh via thermal treatment

119

(Figure S2). The m-TiO2@Fe2O3@Al2O3 monolith catalysts presented mesoporous characteristics

120

because the pyrolysis of titanate cross-linking agents produced H2, NH3, CH4, etc., thus creating

121

mesopores inside the TiO2 shell matrix (Figure 1C). An HAADF-STEM image and elemental maps of

122

Ti, Fe, and Al confirm that the m-TiO2@Fe2O3 composites were uniformly dispersed on the Al2O3

123

nanosheets (Figure 1D). From the XRD patterns (Figure S3) and Mössbauer spectra (Figure S4 and

124

Table S1), Fe2O3, FeAlO3, and Fe spinel phases were found over the m-TiO2@10Fe2O3@Al2O3

125

(Fe/Ti=3) monolith catalyst.27 The pore sizes of the 10Fe2O3@Al2O3 monolith and m-

126

TiO2@10Fe2O3@Al2O3 monolith catalysts with different Fe/Ti ratios were centered at ~4 nm (Figure

127

S5-S8). The specific surface area of the m-TiO2@10Fe2O3@Al2O3 monolith catalysts increased

128

compared to that of 10Fe2O3@Al2O3 monolith (Table S2), and this is likely because of the contribution

129

of an additional mesoporous TiO2 shell.

130

3.2. NO conversion, N2 selectivity, and SO2 tolerance of catalysts. Through tuning the Fe/Ti mole

131

ratio and Fe2O3 loading, the m-TiO2@Fe2O3@Al2O3 monolith with 10 wt% Fe2O3 and an Fe/Ti ratio

132

of 3 exhibits optimum activity; it delivers a broad window of 220-420 oC with NO conversion above

133

90% (Figure S9A and S9B). As seen in Figure 1E, the optimum m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3)

134

monolith catalyst shows a much broader temperature window than that of the 10Fe2O3@Al2O3

135

monolith catalyst. This indicates that there is a promoting effect of the m-TiO2@Fe2O3 composites on

136

both low-temperature and high-temperature activities. The optimum m-TiO2@10Fe2O3@Al2O3

137

(Fe/Ti=3) monolith catalyst shows stable NO conversion of ~94% for 100 hours while showing good 7



Page 8 of 38

138

reproducibility of activity (Figure S10). Similarly, the m-TiO2@10Fe2O3/Al2O3 powder catalysts with

139

different Fe/Ti mole ratios also show broader temperature windows than that of 10Fe2O3/Al2O3 (Figure

140

S11). It is notable that the N2 selectivity of the 10Fe2O3@Al2O3 monolith catalysts maintains 100%

141

from 90 to 450 oC, whereas the 10Fe2O3/Al2O3 powder catalysts shows decreased N2 selectivity above

142

210 oC (Figures S12 and 1F). The N2 selectivity of the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith

143

catalyst keeps at 100% below 360 oC and decreases with increasing the temperature (Figures S12 and

144

1F). The m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) powder catalysts show lower N2 selectivity than the m-

145

TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst within 150-450 oC (Figures S12 and 1F).

146

In the presence of SO2, the m-TiO2@10Fe2O3@Al2O3 monolith catalysts with different Fe/Ti ratios

147

exhibited enhanced low-temperature ( m-

162

TiO2@10Fe2O3/Al2O3 (Fe/Ti=3) powder > 10Fe2O3@Al2O3 monolith > m-TiO2@10Fe2O3@Al2O3

163

(Fe/Ti=3) monolith (Figure S16A). The formation of NO, N2O, and NO2 also decreases in the same

164

order (Figure S16B and S16C). The results indicate that the Al-mesh monolith catalysts show lower

165

NH3 oxidation activity than that of the Al2O3-supported powder catalysts. Meanwhile, the m-

166

TiO2@Fe2O3 composite decreases the NH3 oxidation ability compared to that of single Fe2O3. The Al-

167

mesh monolith catalysts restrain the NH3 oxidation to some extent because the better heat-transfer

168

property of Al-mesh can effectively remove the heat of exothermic reactions from the catalyst bed.

169

For the Al2O3-supported powder catalysts, some hot spots may be generated because inferior heat-

170

transfer accelerates the adverse NH3 oxidation. The m-TiO2@Fe2O3 composite decreases the NO and

171

N2O formation from NH3 oxidation, especially at high temperature (>300 oC), and this is favorable for

172

the NH3-SCR activity and N2 selectivity at high temperature compared to the corresponding activity

173

and selectivity of single Fe2O3 catalysts. Although the NH3 oxidation of m-TiO2@10Fe2O3/Al2O3

174

(Fe/Ti=3) is weaker than that of 10Fe2O3/Al2O3, the N2 selectivity of NH3-SCR for the former is worse

175

than that of the latter (Figure 1F), and this is likely because the inferior heat-removal ability of the

176

powder catalyst promoted the decomposition of NH4NO3 to N2O.29 The following in situ FTIR

177

experiments demonstrate that the m-TiO2@Fe2O3 composite promotes the adsorption of NH4+ species

178

and NO3- (bidentate nitrate) species more than single Fe2O3 does, and this increases the possibility of

179

NH4NO3 decomposition. By contrast, much less N2O formation over 10Fe2O3@Al2O3 and m-

180

TiO2@10Fe2O3@Al2O3 monolith catalysts can be attributed to the fact that the superior heat-removal

181

ability of Al-mesh effectively restrains the NH3 oxidation activity and NH4NO3 decomposition. 9



Page 10 of 38

182

The SO2-tolerant stability of the catalysts was tested at 260 oC, and the results are shown in Figure

183

1G. For the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith, NO conversion slightly decreases from

184

~96% to ~94% when 100 ppm SO2 was introduced into the feed, and then the conversion remains

185

stable for 24 hours. The NO conversion decreases to 92% and remains stable for 12 hours with an

186

increase in SO2 concentration to 300 ppm. With a further increase in the SO2 concentration to 500

187

ppm, NO conversion only decreases to 91% and keeps stable for another 12 hours. It is noteworthy

188

that the N2 selectivity remains at 100% during the whole SO2-tolerant stability test. By contrast, the

189

NO conversion of 10Fe2O3@Al2O3 monolith catalyst decreases from ~68% to ~46% after introducing

190

100 ppm SO2 to the feed and then remains stable for 24 hours. The NO conversion slightly increased

191

to ~49% with increases in the concentration of SO2 to 300 and 500 ppm, which is attributed to the

192

favorable effects of sulfation. Likewise, the m-TiO2@10Fe2O3/Al2O3 powder catalyst also shows

193

improved SO2-tolerant stability compared to that of the 10Fe2O3/Al2O3 powder catalyst at 260 oC

194

(Figure S17). These results indicate that the m-TiO2@Fe2O3 composite can indeed improve the SO2

195

tolerance of single Fe2O3 catalysts. The SO2-tolerant stability of the m-TiO2@10Fe2O3@Al2O3

196

(Fe/Ti=3) monolith catalyst was also tested at 240 oC (Figure S18). The activity decreases gradually

197

with the introduction of SO2, and this is due to more adverse effect of sulfation at lower temperatures.

198

Furthermore, the 10Fe2O3@Al2O3 monolith and m-TiO2@10Fe2O3@Al2O3 (Fe/T=3) monolith

199

catalysts showed good H2O tolerance (Figure S19). Under the co-existence of H2O and SO2, the m-

200

TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst shows more than 90% NO conversion within 270-

201

420 oC and good stability at 270 oC (Figure S20). It was reported that SiO2 increased the Brønsted acid

202

sites of Ce/TiO2 catalysts, and thus restrained the SO2 adsorption and the deposition of sulfates.13

203

Herein, we also prepared the m-SiO2@10Fe2O3@Al2O3 (Fe/Si=1) monolith catalyst using APTES as 10


Page 11 of 38


204

the coupling agent. However, the temperature window is much narrower than that of the m-

205

TiO2@Fe2O3@Al2O3 monolith catalysts (Figure S21A). Besides, the activitiy also decreases obviously

206

in the presence of 100 ppm SO2 (Figure S21B). The results indicate that the m-TiO2@Fe2O3

207

composites have stronger SO2 resistance than m-SiO2@Fe2O3 composites.

208

3.3. Insights into the improved low-temperature activity. The m-TiO2@Fe2O3 composite

209

improves the intrinsic activity of the catalyst (TOF at 100 °C) of 10Fe2O3@Al2O3 monolith from 3.1

210

×103 to 4.1×103, 6.5×103, and 7.6×103 for the m-TiO2@10Fe2O3@Al2O3 monolith with decreases in

211

the Fe/Ti mole ratio from 4 to 2 (Table S3). It is believed that the specific synergistic effects between

212

the Fe2O3 (core)-TiO2 (shell) interface are beneficial for improving the intrinsic activity. Therefore,

213

the structure-activity relationship of the catalysts was investigated in-depth in terms of redox properties,

214

adsorption-activation abilities, and the reaction mechanism.

215

Generally, chemisorbed oxygen species (denoted as Oα) are more active than the lattice oxygen

216

species (denoted as Oβ) in SCR reactions.30 Based on the O1s XPS results (Figure S22A), the m-

217

TiO2@10Fe2O3@Al2O3 monolith catalysts with different Fe/Ti ratios possess higher Oα/(Oα+Oβ) ratios

218

(86.4%-91.6%) compared to the 10Fe2O3@Al2O3 monolith catalyst (79.1%). Moreover, the O2-TPD

219

profiles (Figure S22B) show that more surface-active oxygen species are formed on the m-

220

TiO2@10Fe2O3@Al2O3 monolith catalysts than on the 10Fe2O3@Al2O3 monolith, and the surface-

221

active oxygen species increase with a decrease in the Fe/Ti ratio.31-33 These results indicate that the m-

222

TiO2@Fe2O3 composite enhances the oxidation capability. From the H2-TPR experiment, it can be

223

seen that reducible Fe2O3 decreased over the m-TiO2@10Fe2O3@Al2O3 monolith catalysts compared

224

to that over the 10Fe2O3@Al2O3 monolith catalyst (Figure S22C), and in particular, the m-

225

TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst shows the minimum reduction peak intensity,34 11



Page 12 of 38

226

indicating that there is some strong interaction between the Fe2O3 (core)-TiO2 (shell) interface. Such

227

a strong interaction leads to electron transfer between TiO2 and Fe2O3. Figures 2A and 2B show the

228

fitted Fourier transform spectra of Fe K-edge EXAFS and the fitted magnitude spectra of k3-weighted

229

Fe K-edge EXAFS. The results of the EXAFS fittings show that the Fe phases of the 10Fe2O3@Al2O3

230

monolith and m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalysts are mainly in the forms of FeO

231

and hematite (Table S4). The bond lengths of Fe-O and Fe-Fe over the m-TiO2@10Fe2O3@Al2O3

232

(Fe/Ti=3) monolith are longer than those over the 10Fe2O3@Al2O3 monolith, indicating that the bond

233

strengths of Fe-O and Fe-Fe in the former are weaker than those in the latter. The results imply that

234

the strong interactions between iron oxides and TiO2 weaken the bond strength of iron oxides. From

235

the XANES results, it is found that the Fe-K pre-edge of m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3)

236

monolith shows higher energy (7115.5 eV) than that (7114.6 eV) of the 10Fe2O3@Al2O3 monolith

237

catalyst (Figure 2C), while the energy of the Ti-K XANES edge for the m-TiO2@10Fe2O3@Al2O3

238

(Fe/Ti=3) monolith is lower than that of the Fe-free m-TiO2@Al2O3 monolith catalyst (Figure 2D).

239

The results indicate the higher oxidation state of Fe and lower oxidation state of Ti over the m-

240

TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst compared to those of the 10Fe2O3@Al2O3

241

monolith and m-TiO2@Al2O3 monolith catalysts, respectively. These results evidence the electron

242

transfer from Fe2O3 to TiO2 for the m-TiO2@Fe2O3 composites. The Fe 2p and Ti 2p XPS results are

243

also evidences of the electron transfer from Fe2O3 to TiO2 in the m-TiO2@Fe2O3 composite, leading

244

to increased formation of the Fe3+ and Ti3+ species (Figures S23A and S23B). Such an electron transfer

245

from Fe2O3 to TiO2 accelerates the redox cycle of the SCR reaction, and this can enhance the low-

246

temperature activity.

247

DFT calculations were performed to investigate the interaction between m-TiO2 and Fe2O3. The 12


Page 13 of 38


248

anatase TiO2(101) surface was chosen as the computational model for m-TiO2. The dimer of Fe2O3,

249

namely (Fe2O3)2, was chosen to be the computational model. The most stable geometry of the (Fe2O3)2

250

cluster reported in ref. 35 was randomly put on four different adsorption sites (Figure S24) of the

251

TiO2(101) surface (i.e. O2c, Ti6c, O3c and Ti5c sites) to determine the most favorable adsorption structure.

252

The adsorption energy of (Fe2O3)2 cluster on the TiO2(101) surface is defined as Eb = E((Fe2O3)2) +

253

E(TiO2(101)) − ((Fe2O3)2/TiO2(101)), where E((Fe2O3)2) and E(TiO2(101)) denote the electronic

254

energy of (Fe2O3)2 cluster and TiO2(101) surface, respectively. The more positive values of adsorption

255

energy indicate the higher stability of total system. As shown in Figure 2E, (Fe2O3)2 can strongly

256

adsorb on the TiO2(101) surface with an adsorption energy of 1.59 eV. The Fe and O atoms of (Fe2O3)2

257

preferentially adsorb on the O2c and Ti6c sites of the TiO2(101) surface with an O2c−Fe bond length of

258

1.97 Å and Ti5c−O bond length of 2.15 Å. The charge transfer difference analysis shows that the

259

electron transfers from (Fe2O3)2 to the TiO2(101) surface with a value of 0.09 |e| (Figure 2F), which is

260

consistent with the experimental XPS results (Figure S23A and S23B). The electron depletion of

261

(Fe2O3)2 is mainly from the atoms localized at the interface between (Fe2O3)2 and TiO2(101), and no

262

strong change is found for the atoms far from the interface. Therefore, it is expected that the atoms

263

near the interface serve as the active sites for SCR reactions.

264

The adsorption-activation abilities of NH3 and NOx were investigated using the TPD technique and

265

in situ DRIFT spectra. NH3-TPD profiles (Figure S25A) and pyridine-IR spectra (Figure S25B) show

266

that there are more Brønsted acid and Lewis acid sites on the m-TiO2@10Fe2O3@Al2O3 monolith

267

compared to the 10Fe2O3@Al2O3 monolith catalyst.36-38 The in situ DRIFT spectra for the NH3

268

adsorption demonstrate that more NH3 and NH4+ species are adsorbed on the m-

269

TiO2@10Fe2O3@Al2O3 monolith than on the 10Fe2O3@Al2O3 monolith catalyst (Figure S25C).39-41 13



Page 14 of 38

270

As seen in NO + O2-TPD profiles (Figure S26A), the 10Fe2O3@Al2O3 monolith catalyst adsorbs less

271

monodentate species. On contrast, the TiO2@10Fe2O3@Al2O3 monolith catalyst improves the

272

adsorption of monodentate, bridging, and bidentate nitrates.42,43 DRIFT spectra of NO + O2 adsorption

273

also show that the m-TiO2@10Fe2O3@Al2O3 monolith catalyst promotes the adsorption of various

274

nitrate species, especially gaseous NO2 species that can promote the “Fast SCR” reaction (Figure

275

S26B).44-46 Moreover, the adsorption geometries of NH3 and NO on isolated (Fe2O3)2 and

276

(Fe2O3)2/TiO2(101) are compared in Figures 2G-2J by DFT calculations. The adsorption energy for

277

the small molecule on (Fe2O3)2/TiO2(101) is defined as Eb = E((Fe2O3)2/TiO2(101)) + E(mol) −

278

Emol@((Fe2O3)2/TiO2(101)), where mol denotes NH3, NO or SO2 molecule. NH3 and NO

279

preferentially adsorb on the Fe atoms of isolated (Fe2O3)2 with adsorption energies of 1.09 and 1.79

280

eV, respectively (Figures 2G and 2H). With the introduction of supported TiO2(101), the chemical

281

properties of (Fe2O3)2 are strongly affected, and the adsorption energy of NH3 increases to 1.90 eV.

282

However, the adsorption energy of NO slightly decreases but still has a large value of 1.65 eV (Figures

283

2I and 2J). Bader charge analysis shows that the direction of charge transfer is opposite for the cases

284

of NH3 and NO adsorption on the (Fe2O3)2/TiO2(101). The electrons are transferred from

285

(Fe2O3)2/TiO2(101) to adsorbed NO, whereas they are transferred from adsorbed NH3 to

286

(Fe2O3)2/TiO2(101) (Figures 2K and 2L). The hydrogen bond between an H atom of NH3 and the O2c

287

site results in the stretch of N-H bond and a redistribution of electrons on the (Fe2O3)2/TiO2(101)

288

surface, which also promotes the activation of N-H bonds in NH3. The negatively charged NO can

289

serve as a nucleophilic species that can react with positively charged NH3, resulting in a high NO

290

conversion at low temperature.

291

Figures 3A, 3A1, 3B, and 3B1 show in situ DRIFT spectra of the transient reactions between NO + 14


Page 15 of 38


292

O2 and pre-adsorbed NH3 at 150 °C over the 10Fe2O3@Al2O3 monolith and m-TiO2@10Fe2O3@Al2O3

293

(Fe/Ti=3) monolith catalysts. The two catalysts show the NH3 species (at 1227 and 1622 cm-1) and

294

NH4+ species (at 1685 and 1458 cm-1) after the NH3 adsorption.39,40 After introducing NO + O2, the

295

NH4+ and NH3 species over the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst are gradually

296

consumed and vanish within 10 min, and meanwhile the gaseous NO2 appears at 1615 cm-1 and

297

bidentate nitrate species appear at 1545, 1305, and 1246 cm-1.44-46 The formation of the gaseous NO2

298

species could trigger the “Fast SCR” reaction and improve the low-temperature activity. By contrast,

299

the NH4+ and NH3 species over the 10Fe2O3@Al2O3 monolith catalyst are not consumed within 20

300

min, whereas the bidentate nitrate species appear at 1539, 1304 and 1230 cm-1 after the introduction

301

of NO + O2. These results suggest that the m-TiO2@Fe2O3 composite improves the reactivity of the

302

NH3 and NH4+ species. Figures 3C, 3C1, 3D, and 3D1 show in situ DRIFT spectra of the transient

303

reactions between NH3 and pre-adsorbed NO + O2 at 150 °C over the 10Fe2O3@Al2O3 monolith and

304

m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalysts. The 10Fe2O3@Al2O3 monolith catalyst

305

shows bidentate nitrate species at 1539 and 1303 cm-1 after NO + O2 adsorption.44-46 After the

306

introduction of 500 ppm NH3, these nitrate species hardly change after 10 min, and this indicates the

307

inferior reactivity of bidentate nitrate species on the 10Fe2O3@ns-Al2O3 monolith catalyst. There is no

308

obvious appearance of the NH3 or NH4+ species with the continuous introduction of NH3, and this

309

indicates that the adsorbed bidentate nitrate species restrain the adsorption of ammonia species. After

310

the adsorption of NO + O2, the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst shows different

311

bidentate nitrate species at 1582 and 1546 cm-1 and the gaseous NO2 species at 1616 cm-1 in addition

312

to the same bidentate nitrate species (1304 cm-1) that adsorbed on the m-TiO2@10Fe2O3@Al2O3

313

monolith catalyst.44-46 After the introduction of 500 ppm NH3, the gaseous NO2 species and the weakly 15



Page 16 of 38

314

adsorbed bidentate nitrate species (1582 cm-1) gradually decrease; however, the other two kinds of

315

bidentate nitrate species (1546 cm-1 and 1304 cm-1) hardly change within 15 min. The results indicate

316

that the m-TiO2@Fe2O3 composite promotes the formation of gaseous NO2 and the weakly adsorbed

317

bidentate nitrate species, which can react with the activated NH3 species more easily. Moreover, the

318

NH4+ species (1419 cm-1) appears over the m-TiO2@10Fe2O3@Al2O3 monolith catalyst with the

319

continuous introduction of NH3,39 and this indicates that m-TiO2@10Fe2O3 promotes the adsorption

320

of ammonia species.

321

3.4. Insight into the enhanced SO2 tolerance at low temperature. The SO2 + O2-TPD and SO2 +

322

O2 + NH3-TPD experiments were implemented to analyze the sulfate species deposited on the catalysts.

323

For SO2 + O2-TPD, weakly adsorbed SO2 and metal sulfates both decrease over the m-

324

TiO2@10Fe2O3@Al2O3 monolith catalyst with a decrease in the Fe/Ti ratio (Figure S27).13 In terms

325

of SO2 + O2 + NH3-TPD, the amounts of deposited HN4HSO4 and metal sulfates both decrease over

326

the m-TiO2@10Fe2O3@Al2O3 monolith compared to those over the 10Fe2O3@Al2O3 monolith catalyst

327

(Figure S28).47 Moreover, the in situ DRIFT spectra of SO2 + O2 adsorption over the 10Fe2O3@Al2O3

328

monolith, m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith and m-TiO2@Al2O3 monolith were

329

investigated. As shown in Figure 4A, after the adsorption of SO2 + O2 at 30 oC, two bands at 1020 and

330

1645 cm-1 appear over the catalysts, and these correspond to surface sulfate species and H2O molecules,

331

respectively. The H2O molecule is generated from the reaction between SO2 and surface hydroxy.48

332

The amount of surface metal sulfates decreases over the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3)

333

monolith catalyst with a decrease in the Fe/Ti ratio, and this indicates that the m-TiO2@Fe2O3

334

composite can restrain sulfate deposition. With an increase in adsorption temperature, the adsorbed

335

surface sulfates and H2O molecules gradually decrease, whereas some stable sulfates are still present 16


Page 17 of 38


336

on the 10Fe2O3@Al2O3 monolith catalyst above 150 oC (Figure S29A). However, surface sulfates and

337

H2O molecules gradually decrease and vanish below 300 oC over the m-TiO2@10Fe2O3@Al2O3

338

(Fe/Ti=3) monolith and m-TiO2@Al2O3 monolith (Figures S29B and S29C). It is speculated that only

339

some surface sulfates formed on the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith and m-

340

TiO2@Al2O3 monolith; however, more bulk sulfates are generated on 10Fe2O3@Al2O3 monolith, and

341

the decomposition of these species is difficult. The in situ DRIFT of SO2 + O2 + NH3 adsorption was

342

also performed over the 10Fe2O3@Al2O3 monolith, m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith

343

and m-TiO2@Al2O3 monolith catalysts (Figure S30). The results show that the amounts of surface

344

sulfite and HSO4- species are much less over the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith

345

catalyst than those over the 10Fe2O3@Al2O3 monolith catalyst.39,40,48,49 The TGA was used to further

346

investigate the deposited amount of sulfates on the 10Fe2O3@Al2O3 monolith and m-

347

TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalysts after reaction in the presence of SO2. As seen in

348

Figure 4B, the weight loss below 300 oC is ascribed to the adsorbed water, the surface hydroxyl groups,

349

and the decomposition of NH4HSO4, whereas the weight loss above 300 oC is attributed to the

350

decomposition of FeSO4 on the basis of the TGA results of NH4HSO4 and FeSO4·7H2O decomposition

351

(Figure S31 and S32).11,50,51 It is noteworthy that the weight loss of the m-TiO2@10Fe2O3@Al2O3

352

(Fe/Ti=3) monolith (0.22%) is lower than that of the 10Fe2O3@Al2O3 monolith (0.36%) in the range

353

of 300-650 °C, and this suggests that there is less formation of FeSO4 on the m-TiO2@10Fe2O3@Al2O3

354

(Fe/Ti=3) monolith catalyst.

355

Moreover, the surface sulfates were investigated by analyzing the valence states of surface Fe over

356

the 10Fe2O3@Al2O3 monolith and m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalysts after

357

reaction (at 210, 240, 270, and 300 oC) in the presence of SO2. As seen in Figure 4C (Figure S33A and 17



Page 18 of 38

358

S33B), after reaction at 210 oC in the presence of SO2, the Fe3+ fractions of the two catalysts decrease

359

from 55.2% and 65.2% for the fresh samples (Figure S23A) to 18.1% and 53.1%, respectively, because

360

the formation of FeSO4 decreases the Fe3+ fraction. When the reaction temperature is increased, the

361

Fe3+ fractions of both catalysts increase, and this is likely because the formation of FeSO4 is restrained

362

at higher temperatures. It is noteworthy that the Fe3+ fractions of the m-TiO2@10Fe2O3@Al2O3

363

(Fe/Ti=3) monolith are always higher than that of the 10Fe2O3@Al2O3 monolith at the same reaction

364

temperature, and this indicates that the m-TiO2@Fe2O3 composite shows good protection against

365

FeSO4 deposition. Moreover, the Fe3+ fractions and the percent concentration of surface SO42- were

366

investigated for the 10Fe2O3@Al2O3 monolith and m-TiO2@10Fe2O3@Al2O3 monolith catalysts with

367

different Fe/Ti ratios after reaction at 260 oC in the presence of SO2, and the results are shown in Figure

368

4D (Figure S34). Compared to the 10Fe2O3@Al2O3 monolith catalyst, larger Fe3+ fractions and lower

369

amount of SO42- species are generated over the m-TiO2@10Fe2O3@Al2O3 monolith catalyst with a

370

decrease in the Fe/Ti ratio, and this is further evidence that the m-TiO2@Fe2O3 composite restrains

371

sulfate deposition.

372

To explain why the m-TiO2@Fe2O3 composite restrains sulfate deposition, the SO2 adsorption

373

energy on the TiO2-shell and the Fe2O3-core of the m-TiO2@Fe2O3 composite were investigated using

374

DFT calculations. Figures 4E and 4F show the adsorption geometries of SO2 on TiO2 (101) and

375

(Fe2O3)2/TiO2(101). Clearly, SO2 adsorbs weakly on the TiO2(101) surface with an adsorption energy

376

of 0.21 eV and easily desorbs from the TiO2 surface. Although the SO2 adsorption energy on

377

(Fe2O3)2/TiO2(101) increases to 0.86 eV, the small adsorption energy causes SO2 to be easily replaced

378

by the adsorption of NO and NH3 because of the higher adsorption energy of NH3 (1.90 eV) and NO

379

(1.65 eV) on (Fe2O3)2/TiO2(101). The results of DFT calculations also show that SO2 is not easily 18


Page 19 of 38


380

adsorbed on the m-TiO2 shell and the Fe2O3 core of the m-TiO2@Fe2O3 composite. Therefore, the

381

deposition of sulfate is largely restrained on the m-TiO2@Fe2O3 composite.

382

To study the self-prevention function of the m-TiO2@Fe2O3 composite, we also investigated the

383

SCR performance of catalysts pre-impregnated with 5 wt.% NH4HSO4. As seen in Figure 4G, below

384

200 oC, the activity of the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith pre-impregnated with

385

NH4HSO4 is very low because NH4HSO4 covers the active sites of the catalyst. The activity increases

386

quickly with an increase in temperature to 300 oC, and this indicates that NH4HSO4 gradually

387

decomposes. The NO conversion is 98.6% at 300 oC, and this is comparable to that (99.2% at 300 oC)

388

for the fresh m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst. By contrast, NO conversion of

389

the 10Fe2O3@Al2O3 monolith catalyst pre-impregnated with NH4HSO4 is only 40.6% at 270 oC and

390

75.6% at 300 oC, both of which are much lower than the corresponding values (79.4% at 270 oC and

391

88.8% at 300 oC) over the fresh 10Fe2O3@Al2O3 monolith catalyst. It is noteworthy that NO

392

conversion of both catalysts pre-impregnated with NH4HSO4 is near 100% at a higher temperature,

393

and this is because of the promoting effect of sulfation that results from the residual SO42- species.28

394

These results indicate that the m-TiO2@Fe2O3 composite may promote NH4HSO4 decomposition,

395

which could effectively restrain NH4HSO4 deposition during NH3-SCR in the presence of SO2.

396

NH4HSO4-TPDC experiments (Figure S35) demonstrate that NH4HSO4 decomposition occurs more

397

readily on the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith compared to the 10Fe2O3@Al2O3

398

monolith catalyst. This is likely because of the presence of strong electron interactions between HSO4-

399

and the Brønsted acid sites from m-TiO2, which impairs the binding between NH4+ and HSO4-.

400

Moreover, the in situ DRIFT spectra of the transient reactions between SO2 + NH3 and pre-adsorbed

401

NO + O2 at 150 °C were investigated for the 10Fe2O3@Al2O3 monolith and m-TiO2@10Fe2O3@Al2O3 19



Page 20 of 38

402

(Fe/Ti=3) monolith catalysts (Figure S36A and S36B). For the 10Fe2O3@Al2O3 monolith catalyst,

403

with the introduction of SO2 + NH3, the pre-adsorbed bidentate nitrate species (1531 and 1302 cm-1)

404

hardly change within 15 min, and this indicates the inferior reaction efficiency of the bidentate nitrates.

405

It is noteworthy that no sulfate species are generated with the introduction of SO2, and this indicates

406

that the adsorbed bidentate nitrate species restrain the adsorption of SO2. The m-

407

TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) monolith catalyst shows the presence of the gaseous NO2 species

408

(1619 cm-1) and weakly adsorbed bidentate nitrate species (1581 cm-1), and the amounts of these

409

gradually decrease with the introduction of SO2 + NH3. Moreover, the NH4+ species (1427 cm-1) appear

410

over the m-TiO2@10Fe2O3@Al2O3 monolith catalyst after introducing SO2 + NH3. These results

411

indicate that the reactivity of gaseous NO2 and weakly adsorbed bidentate nitrate species is not affected

412

by the introduction of SO2 because the m-TiO2@Fe2O3 composite restrains sulfate deposition and

413

ensures the adsorption and reaction of NH4+ (with NO2 and active bidentate nitrate species). The in

414

situ DRIFT spectra of the transient reactions between SO2 + NO + O2 and pre-adsorbed NH3 at 150 °C

415

were also investigated over the 10Fe2O3@Al2O3 monolith and m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3)

416

monolith catalysts (Figure S37A and S37B). The pre-adsorbed NH4+ and NH3 species are still inactive

417

over the 10Fe2O3@Al2O3 monolith catalyst, which is the same observation as in the absence of SO2.

418

With the continuous introduction of SO2 + NO + O2, a great number of sulfates are deposited, and

419

these are associated with the formation of bidentate nitrate species. By contrast, a much lower amount

420

of sulfates are deposited on the m-TiO2@10Fe2O3@Al2O3 (Fe/Ti=3) catalyst. Although the lower

421

amount of deposited sulfates cause the adsorbed NH3 species to be inactive, the m-TiO2@Fe2O3

422

composite maintains a certain reactivity of the adsorbed NH4+ because of the largely reduced sulfate

423

deposition.39,40,48,52-54 20


Page 21 of 38


424

3.5. Activity/Selectivity-promotion and SO2-resistant mechanisms. The activity/selectivity-

425

promotion and SO2-resistant mechanisms of the m-TiO2@Fe2O3 composite on NOx reduction in the

426

absence/presence of SO2 are proposed (Figure 5). There is a strong interaction at the Fe2O3 (core)-TiO2

427

(shell) interface in the m-TiO2@Fe2O3 composite, and this results in favorable electron transfer from

428

Fe2O3 to TiO2. Thus, more Fe3+ and Ti3+ are generated, and these enhance the redox cycle as follows:

429

Fe3+ + Ti3+ ↔ Fe2+ + Ti4+. The synergistic effect between Fe2O3 and TiO2 promotes the formation of

430

O-ads and NO+ads, thereby leading to the formation of gaseous NO2, which launches the “Fast SCR”

431

reaction with active ammonia species.55,56 Moreover, the m-TiO2@Fe2O3 composite also promotes the

432

formation of weakly adsorbed bidentate nitrate species that can react with activated NH3 species more

433

easily. Therefore, the m-TiO2@Fe2O3@Al2O3 monolith catalyst exhibits good low-temperature

434

activity. For the Fe2O3@Al2O3 monolith catalyst, electron transfer only happens as Fe3+↔ Fe2+. Also,

435

the redox ability is not so strong, and thus, it only generates some inactive bidentate nitrate species

436

(NO3-). Therefore, inferior low-temperature activity is obtained because of the “Standard SCR”

437

reaction between NO3- and NH4+.

438

In the presence of SO2, the activity is decreased because of the sulfidation of active Fe species and

439

deposition of NH4HSO4 over the Fe2O3@Al2O3 monolith catalyst at low temperature (

SO2-tolerant Selective Catalytic Reduction of NOx over Meso-TiO2

Recommend Documents