The effects of H2O on a vanadium-based catalyst for NH3-SCR at low temperatures: a quantitative study of the reaction pathway and active sites

Kuo Liu ab, Zidi Yan ad, Hong He acd, Qingcai Feng ab and Wenpo Shan *c
aState Key Joint Laboratory of Environment Simulation and Pollution Control, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China
bEditorial Office of Journal of Environmental Sciences, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China
cCenter for Excellence in Regional Atmospheric Environment, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen 361021, China. E-mail: wpshan@iue.ac.cn
dUniversity of Chinese Academy of Sciences, Beijing 100049, China

Received 11th July 2019 , Accepted 12th August 2019

First published on 12th August 2019


The influences of H2O on the adsorption amounts of NO, NO2, and NH3 on V2O5/WO3–TiO2 and the reaction pathway at low temperatures for selective catalytic reduction of nitrogen oxides by NH3 (NH3-SCR) were quantitatively obtained using the temperature programmed surface reaction as well as the transient response method (TRM). The active site distribution on V2O5/WO3–TiO2 was also clarified quantitatively. NOx adsorption took place on [V4+]–OH + [V5+]–O, [V5+]–OH, and [Ti4+]2–O, the surface concentrations of which were 24, 28 and 26 μmol g−1, respectively, whereas NH3 was activated on [V4+]–OH + [V5+]–O, [V5+]–OH, [Ti4+]2–O and W species, with concentrations of 24, 28, 26 and 274 μmol g−1, respectively. The V and W species did not distribute as a monolayer. The “nitrite path”, referring to the reaction between adsorbed NH3 and nitrite species to produce N2 and water (H2O), contributed to standard and fast SCR mechanisms, but the “NH4NO3 path”, referring to the reaction between the NO gas and NH4NO3 formed from the surface nitrates and adsorbed NH3 species to emit NO2, H2O and N2, did not occur at low temperatures without H2O. Both the “NH4NO3 path” and the “nitrite path” contributed to NH3-SCR with H2O in the gas feed. The presence of H2O increased the amounts of the adsorbed nitrates and ammonia at 150 °C, but hindered the reaction between the adsorbed NH3 and NO gas, lowering the SCR activity.


1 Introduction

Selective catalytic reduction of nitrogen oxides (NOx) by NH3 (NH3-SCR) is the dominant technology for NOx emission control.1 The low temperature activity of NH3-SCR catalysts is of great importance for the applications on both stationary sources such as industrial plants and mobile sources such as diesel vehicles.2,3 Vanadium (V)-based catalysts, especially V2O5–WO3/TiO2, have been commercialized and applied in NH3-SCR for decades due to their high NH3-SCR performance as well as their excellent resistance to SO2 and H2O poisoning.4–7 Although many efficient vanadium-free catalysts, such as zeolites8–11 and oxides,12–15 have been developed and even industrially applied, V-based catalysts are still irreplaceable for stationary applications and widely applied in diesel vehicles in developing countries.

Although the NH3-SCR mechanism on V-based catalysts has been investigated extensively, the main reaction mechanism at low temperatures is still a controversial subject. Some researchers found that NH3 tended to adsorb on the active sites of V-based catalysts, and then the adsorbed NH3 reacted with the adsorbed NO species, following the Langmuir–Hinshelwood (L–H) mechanism, such as on V2O5 supported by activated semi-coke (V2O5/ASC),16 a commercial V2O5 catalyst,17 and carbon–ceramic cellular monolith supported vanadium oxides.18 Another reaction pathway is activation of NO gas on the adsorbed NH3 species, forming a surface reaction intermediate that decomposes to N2 and H2O, which is called the Eley–Rideal (E–R) mechanism.19–22 Lian et al.20 reported the presence of the E–R mechanism on V2O5/CeTiOx and V2O5/CeO2 during NH3-SCR based on their DRIFT study results. By using kinetic modeling, Lei et al.21 found that NH3-SCR followed the E–R mechanism rather than the L–H mechanism on V2O5/AC below 473 K, either in the presence or absence of H2O. This mechanism was also suggested to take place on the V2O5 catalyst surface according to the density functional theory (DFT) calculation results by Soyer et al.22 Other researchers suggested that both L–H and E–R mechanisms contributed to NH3-SCR on V-based catalysts simultaneously, e.g., on V2O5–WO3/TiO2 (ref. 23 and 24) and V–Ce/Ti-PILC.25 During fast SCR, the reaction between NO and NH4NO3 (generated from NO2 and NH3) to form NH4NO2, which decomposes to N2 and H2O, was suggested to be an important step on V2O5–WO3/TiO2.26,27 Another mechanism is the reaction of NO with nitrates formed by NO2 adsorption to generate nitrites, which then combines the adsorbed NH3 to form NH4NO2 that subsequently decomposes to N2 and H2O.28 Therefore, it is important to investigate the NH3-SCR mechanism and the distribution of active sites over V2O5–WO3/TiO2.

Under practical conditions, H2O is always found to be present in diesel exhaust. H2O has been reported to deactivate V-based catalysts during NH3-SCR.29,30 It is widely accepted that competitive adsorption between H2O and NH3 molecules takes place on the surface of V-based oxides, and inhibits the reaction between NH3 and NO, e.g., on V2O5–WO3/TiO2–SiO2.31 Zhu et al.32 found that H2O could increase the amount of less-active NH4+ (NH3 adsorbed on Brønsted acid sites) species, and simultaneously, decrease the amount of more-active NH3 (NH3 adsorbed on Lewis acid sites) species, contributing to the inhibitory effect of H2O on NH3-SCR. Huang et al.33 found that H2O did not decrease the NO/NH3 adsorption over V2O5/AC, but enhanced the NH3 adsorption. Also, H2O hindered the reaction between the NH3 on Lewis acid sites and the NO gas over V2O5/AC, causing a decreased NH3-SCR performance in the presence of H2O. To date, the mechanism by which H2O inhibits the NH3-SCR reaction is still controversial and needs clarification. Moreover, quantitative information on the influence of H2O on NOx/NH3 adsorption amounts as well as the NH3-SCR mechanism on V-based catalysts at low temperatures is still lacking.

Considering the above-mentioned important issues still being debated for V-based catalysts, in this study, the reaction pathway of NH3-SCR as well as the active site dispersion at low temperatures over a traditional V2O5/WO3–TiO2 catalyst was quantitatively determined, and the influence of H2O on the adsorption amounts of the reactants in the feed including NO, NO2, and NH3 and the inhibitory mechanism of H2O on the NH3-SCR reaction pathway were also clarified. The results would be helpful for the understanding of the reaction mechanism and active sites, as well as the inhibition effects of H2O, on V-based catalysts for low-temperature SCR.

2 Experimental

2.1 Preparation of V2O5/WO3–TiO2

V2O5/WO3–TiO2 with a 3.5 wt% V2O5 loading was synthesized by an impregnation method. First, ammonium metavanadate (NH4VO3) was dissolved in an oxalic acid (H2C2O4) water solution. Then, DT-52 (a commercial TiO2 material containing 10 wt% WO3) obtained from CristalACTiV™ was added to the above solution, and after 1 h, under vigorous stirring, the solution was evaporated to dryness using a rotary evaporator. The obtained powder was dried at 110 °C and subsequently calcined in air at 500 °C for 3 h.

2.2 Quantification and characterization

The reactant (including NO, NO2, and NH3) adsorption amounts were studied quantitatively at room temperature (30 °C) and the reaction temperature (150 °C) by the TRM. The experimental setup and the measuring principle were described in our previous study, and no obvious adsorption of NO, NO2 and NH3 was observed without catalysts.34 The residence time of the gas at the tube between the catalyst and detector was ∼1 s. 100 mg V2O5/WO3–TiO2 (40–60 mesh) was pretreated at 400 °C for 30 min in a 20 vol% O2/N2 flow before each adsorption experiment, and then the catalyst was cooled down to 30 or 150 °C in O2/N2. Afterwards, the catalyst was by-passed. After the reactant concentration was stable, the surface of V2O5/WO3–TiO2 was exposed to the gas mixture. The gas emitted at the outlet was monitored using an online FTIR spectrometer (NEXUS 670-FTIR). The peaks at 1302/1829, 1740–1748, and 1338/1728 cm−1 were used for detecting N2O3, N2O4, and N2O5, respectively,35,36 but none of them was observed. The adsorbed amount of NO, NO2, or NH3 (Nadsorbed) was obtained based on:
image file: c9cy01370a-t1.tif
where Nall is the overall amount of the reactant molecules introduced, and Aall and Aadsorbed are the areas of the FTIR peaks generated by Nall and Nadsorbed, respectively. The adsorption gas mixture was composed of either 500 ppm NH3 or 500 ppm NOx (NO2 or NO) in N2. The rate of gas flow was controlled to be 500 mL min−1. The adsorption experiment was conducted at least twice.

The amounts of weakly adsorbed species were determined by the following method. After NH3 or NOx adsorption, the catalyst was subjected to N2 purging, and the weakly adsorbed NH3 or NOx would desorb from the V2O5/WO3–TiO2 surface. Then, the catalyst surface was exposed to NH3 or NOx for the second time. The amount of weakly adsorbed molecules equaled that of NH3 or NOx adsorbed during the second adsorption procedure.

The TPSR was conducted with a 10 °C min−1 heating rate in a gas flow (500 mL min−1) containing 500 ppm NO/N2 with 0 vol% or 2 vol% H2O on 100 mg of the NO2 and NH3 pretreated V2O5/WO3–TiO2 catalyst (40–60 mesh). The gas concentration was recorded using an online FTIR spectrometer (NEXUS 670-FTIR) as well.

A N2 adsorption/desorption study employing a Quantachrome Quadrasorb SI-MP at −196 °C was applied to determine the surface area of V2O5/WO3–TiO2. V2O5/WO3–TiO2 was degassed at 300 °C for 5 h before the N2 physisorption. The surface area of V2O5/WO3–TiO2 was 60 m2 g−1.

3 Results and discussion

3.1 Adsorption of NO2

As shown in Fig. 1, introducing NO2 onto the V2O5/WO3–TiO2 surface led to a decrease in the NO2 concentration and an increase in the NO concentration at both 30 and 150 °C. Emission of N2O, N2O3, N2O4, or N2O5 was not observed. The molar ratio of the amount of the NO2 consumption (∼144 μmol g−1) to that of the NO formation (∼40 μmol g−1) was 3.6[thin space (1/6-em)]:[thin space (1/6-em)]1, as shown in Fig. 1. It was found by Gao et al.37 that NO2 could oxidize [V4+]–OH, producing V5+[double bond, length as m-dash]O and HNO2, according to their DFT results. Kantcheva et al.38 reported that both V5+[double bond, length as m-dash]O and V5+–OH on a vanadia–titania catalyst could adsorb NO2, forming V5+–NO3 and V4+–O–N5+O2, respectively. TiO2 could also adsorb NO2. Sivachandiran et al.39 pointed out that the molar ratio of NO2 adsorption, NO2 desorption after adsorption, and NO emission was 3[thin space (1/6-em)]:[thin space (1/6-em)]2[thin space (1/6-em)]:[thin space (1/6-em)]1 on TiO2 at room temperature. According to the literature cited above, it can be suggested that eqn (1)–(4) took place:
 
[V4+]–OH + NO2 ↔ [V5+]–O + HNO2(1)
 
[V5+]–O + NO2 ↔ [V5+]–ONO2(2)
 
2[V5+]–OH + 3NO2 ↔ 2[V5+]–ONO2 + NO + H2O(3)
 
3NO2 + [Ti4+]2–O ↔ 2([Ti4+]–ONO2) + NO(4)
where [V4+]–OH, [V5+]–O, [V5+]–OH, and [Ti4+]2–O are V4+ species, V5+ species, and lattice O2− species on TiO2 surface sites available for nitrate formation, respectively. V4+ and V5+ species in V2O5/WO3–TiO2 were proved to be present by other researchers.5 The surface nitrate species generated at 30 °C by NO2 adsorption (N*-NO3) was obtained as follows:
N*-NO3 = NNO2 consumptionNNO formation = (144–40) μmol g−1 = 104 μmol g−1
where NNO2 consumption and NNO formation are the amounts of NO2 consumed and NO formed, respectively.

At 150 °C, only a small amount of NO (∼6 μmol g−1) was evolved during NO2 adsorption, and almost all NO2 molecules adsorbed on the surface of V2O5/WO3–TiO2 formed surface nitrates, indicating that it is difficult for the forward reactions (eqn (3) and/or (4)) to take place, possibly due to the loss of OH groups or the reactivity of surface lattice oxygen ([Ti4+]2–O) at higher temperature. Also, some nitrates formed by eqn (3) and (4) might be weakly bonded, and cannot be formed at 150 °C, contributing to the low amount of NO evolved.


image file: c9cy01370a-f1.tif
Fig. 1 Adsorption of NO2/N2 (500 ppm) on V2O5/WO3–TiO2 at 30 and 150 °C.

3.2 Reactions between NO and surface nitrates

In order to calculate the amounts of nitrates remaining after N2 purging, NO2 adsorption was conducted for the second time to determine how many active sites on V2O5/WO3–TiO2 were released by N2 purging after NO2 adsorption. As shown in Fig. 2, the amounts of weakly adsorbed NO2 molecules removed by N2 were ∼20 μmol g−1 at 30 °C and negligible at 150 °C, indicating that the amounts of nitrates remaining were 104–20 = 84 and ∼54 μmol g−1, respectively. The subsequent exposure to NO brought about the evolution of NO2, and simultaneously, the consumption of NO, as shown in Fig. 3a. According to the results of NO adsorption on V2O5/WO3–TiO2 in Fig. 3b, no NO2 gas was evolved at either 30 or 150 °C, indicating that NO2 evolution shown in Fig. 3a did not originate from NO oxidation, but from the reaction between the NO gas and surface nitrates. It was reported that NO adsorption did not take place on W sites or V2O5 sites,40 and thus, it is possible that NO adsorption took place on TiO2. Surface nitrates, including bidentate, bridging, and monodentate nitrates, were observed on TiO2 in the NO adsorption process, either in the presence or absence of O2,41,42 and the following reactions might occur:
 
NO + [Ti4+]2–O ↔ [Ti4+]–ONO + [Ti3+]-□(5)
 
[Ti4+]–ONO + [Ti4+]2–O ↔ ([Ti4+]–O)2 = NO(bridging nitrate) + [Ti3+]-□(6)
 
([Ti4+]–O)2[double bond, length as m-dash]NO(bridging nitrate) ↔ [Ti4+][double bond, length as m-dash]O2[double bond, length as m-dash]NO(bidentate nitrate) + [Ti3+]-□ ↔ [Ti4+]–O–NO2(monodentate nitrate) + [Ti3+]-□(7)

image file: c9cy01370a-f2.tif
Fig. 2 Adsorption of NH3/N2 (500 ppm) or NO2/N2 (500 ppm) on V2O5/WO3–TiO2 saturated with NH3/N2 or NO2/N2 (500 ppm) and subsequently purged with N2 at 30 and 150 °C.

image file: c9cy01370a-f3.tif
Fig. 3 Adsorption of NO/N2 (500 ppm) (a) on V2O5/WO3–TiO2 saturated with NO2/N2 (500 ppm) at 30 and 150 °C and (b) on fresh V2O5/WO3–TiO2 at 30 and 150 °C.

The ratio of NO2 generation (101 μmol g−1) to NO reduction (48 μmol g−1) was about 2[thin space (1/6-em)]:[thin space (1/6-em)]1, and hence, the following reactions occurred:

 
NO + [V5+]–ONO2 ↔ [V5+]–ONO + NO2 ↔ [V4+]-□ + 2NO2(8)
 
NO + [Ti4+]–NO3 ↔ NO2 + [Ti4+]–ONO ↔ 2NO2 + [Ti3+]-□(9)
where [Ti3+]-□ denotes a Ti3+ site.

At 30 and 150 °C, after eqn (8) and (9) took place, the remaining nitrates could be calculated as follows:

30 °C: 84–48 = 36 μmol g−1

150 °C: 54–14 = 40 μmol g−1

At 150 °C, 40 μmol g−1 surface nitrates remained on the surface of V2O5/WO3–TiO2 after NO was introduced, indicating that most surface nitrates on V2O5/WO3–TiO2 were less reactive, and could not react with NO at low temperatures. It was reported that the NO gas reacted with the surface nitrates to generate nitrites, and subsequently, the nitrites generated will react with the NH3 molecules to generate H2O and N2.28,43 This reaction will be denoted as the “nitrite path” hereafter in the present manuscript. This reaction was regarded as an important step to form N2 during fast SCR on V2O5/WO3–TiO2.28 Similarly, this step was also suggested to be important in fast SCR on Fe- and Cu-based zeolites44,45 and on Ce–W mixed oxide (CeWOx).46 However, a large amount of surface nitrates could not react with NO on V2O5/WO3–TiO2 (Fig. 3a). Therefore, it was difficult for the reaction between the surface nitrates and the NO gas to occur at 150 °C, and it can be concluded that the “nitrite path” might be a slow step on V2O5/WO3–TiO2 during fast SCR.

3.3 NH3 adsorption

To elucidate the impact of the presorbed NO2 on the amount of NH3 adsorption, NH3 adsorption on both fresh and NO2-pretreated V2O5/WO3–TiO2 was conducted. As shown in Fig. 4a, the amount of the NH3 molecules adsorbed on NO2-saturated V2O5/WO3–TiO2 was ∼443 μmol g−1 at 30 °C, 65 μmol g−1 greater than the adsorption amount on fresh V2O5/WO3–TiO2 (∼378 μmol g−1). NO2 adsorption produced 84 μmol g−1 adsorbed nitrates after the N2 purging, as calculated in section 3.2, and it is inferred that the sites for adsorption of NH3 and some NO2 are different. No desorption of NO or NO2 was observed (Fig. 4a). Therefore, it can be concluded that the increase in the NH3 adsorption amount was due to the formation of NH4NO3 from the reaction between the adsorbed nitrates and the NH3 gas. NH3 reacted with some of the surface nitrates to generate NH4NO3, causing a 65 μmol g−1 increase in NH3 adsorption amount. Consequently, it is inferred that the ratio of NH3 to nitrate during NH4NO3 formation was 1[thin space (1/6-em)]:[thin space (1/6-em)]1.
 
[Mn+]–OH + NH3 ↔ [Mn+]–O–NH4(10)
 
[V5+]–ONO2 + [Mn+]–O–NH4 ↔ [Mn+]–O–NH4NO3–[V5+](11)
 
[Ti4+]–ONO2 + [Mn+]–O–NH4 ↔ [Mn+]–O–NH4NO3–[Ti4+](12)
where Mn+ and [Mn+]–OH are a metal site and a Brønsted acid site, respectively. A similar phenomenon was also observed at 150 °C. The amount of NH3 consumption on NO2-pretreated V2O5/WO3–TiO2 (∼223 μmol g−1) was about 55 μmol g−1 higher than that on the fresh V2O5/WO3–TiO2 (∼168 μmol g−1) at 150 °C. The increased amount (55 μmol g−1) was similar to that of the nitrates formed by NO2 adsorption (54 μmol g−1, see Fig. 1) at 150 °C, indicating that eqn (10)–(12) could take place at the reaction temperature.

image file: c9cy01370a-f4.tif
Fig. 4 (a) Adsorption of NH3/N2 (500 ppm) on fresh V2O5/WO3–TiO2 (VWTi) and on V2O5/WO3–TiO2 saturated with NO2/N2 (500 ppm) followed by N2 purging at 30 and 150 °C; (b) adsorption of NO/N2 (500 ppm) on V2O5/WO3–TiO2 treated with NO2/N2 (500 ppm) and subsequently with NH3/N2 (500 ppm) at 30 and 150 °C.

The amounts of weakly adsorbed NH3 at 30 and 150 °C were obtained using the experimental methods described in section 2.2. For NH3, the amounts removed at the second adsorption step were 191 ± 10 and 48 ± 5 μmol g−1 at 30 and 150 °C, respectively, as seen in Fig. 2. The overall NH3 adsorption amounts at 30 °C on the fresh and NO2-presorbed V2O5/WO3–TiO2 were ∼378 and ∼443 μmol g−1, respectively (Fig. 4a); therefore, the amount of NH3 remaining on the surface of V2O5/WO3–TiO2 after N2 purging can be calculated as follows:

Fresh V2O5/WO3–TiO2: 378 μmol g−1–191 μmol g−1 = 187 μmol g−1

NO2-presorbed V2O5/WO3–TiO2: 443 μmol g−1–191 μmol g−1 = 252 μmol g−1

At 150 °C, the adsorption amounts of NH3 on the fresh and NO2-pretreated catalysts were ∼168 and ∼223 μmol g−1, respectively, and thus, the amount of the adsorbed NH3 can be calculated as follows:

Fresh V2O5/WO3–TiO2: 168 μmol g−1–48 μmol g−1 = 120 μmol g−1

NO2-presorbed V2O5/WO3–TiO2: 223 μmol g−1–48 μmol g−1 = 175 μmol g−1.

Subsequently, NO was introduced onto the surface of V2O5/WO3–TiO2, and no NO was consumed at 30 °C (Fig. 4b), indicating that no reaction took place during this process. At 150 °C, the amount of the adsorbed NH3 (175 μmol g−1) far exceeded that of the NO consumption (26 μmol g−1) (Fig. 4b), indicating that the adsorbed NH3 can partially participate in the SCR reaction at 150 °C.

3.4 NH3-SCR mechanism over V2O5/WO3–TiO2

The TPSR of the NH3 and NO2 pretreated V2O5/WO3–TiO2 at 30 °C in 500 ppm NO/N2 was also studied to clarify the reaction mechanism. As seen in Fig. 5, no obvious desorption of NO2, N2O or NH3 can be observed during the TPSR process. N2 formation during the TPSR was detected using an Agilent 6890N gas chromatograph with a TCD (TDX-01 column) intermittently. The gas sample was injected when the temperature of the catalyst was increased to ∼125 °C, and N2 was formed during the TPSR in NO/N2, indicating that the adsorbed NH3 had reacted with NO gas to form N2, and NH4NO3 formed on the surface also reacted with NO to produce N2 instead of decomposing to N2O directly. It was reported that the emission of N2O was negligible below 250 °C on V2O5/WO3–TiO2 either with or without H2O,27,47,48 in line with the results in the present work. Also, the reaction between NO gas and the formed NH4NO3 (eqn (13)) did not occur on V2O5/WO3–TiO2, since no NO2 was emitted during the TPSR.
 
NH4NO3 + NO → N2 + NO2 + 2H2O(13)

image file: c9cy01370a-f5.tif
Fig. 5 TPSR in NO/N2 (500 ppm) of V2O5/WO3–TiO2 saturated with NO2/N2 (500 ppm) at 30 °C and subsequently with NH3/N2 (500 ppm) at 30 °C.

In order to verify if all of the adsorbed NO2 and NH3 species have completely reacted with the NO gas during the TPSR, the adsorption of NO/NO2/NH3 and the reaction of NO with the adsorbed NO2/NH3 were conducted again. Before these experiments, the catalyst after the TPSR in NO/N2 was cooled down to 30 °C in N2, and then treated in O2/N2 for 20 min to supplement the lost oxygen. Afterwards, the NO/NO2/NH3 adsorption and the TPSR in NO/N2 of the NO2 and NH3 pre-treated catalyst were conducted. The results of the NO2/NH3 adsorption and the reaction of NO with the adsorbed NO2/NH3 species after the TPSR in NO/N2 were similar to those on the fresh catalyst, as shown in Fig. 6, indicating that the adsorbed NO2 and NH3 species on V2O5/WO3–TiO2 can completely react with NO during the TPSR.


image file: c9cy01370a-f6.tif
Fig. 6 (a) NO, NO2, and NH3 adsorption at 30 °C on used V2O5/WO3–TiO2, and (b) the TPSR in NO/N2 (500 ppm) of used V2O5/WO3–TiO2 saturated with NO2/N2 (500 ppm) at 30 °C and subsequently with NH3/N2 (500 ppm) at 30 °C. Used V2O5/WO3–TiO2 refers to the NO2 and NH3 pre-treated V2O5/WO3–TiO2 after the TPSR treatment in NO/N2.

The remaining NH3 amount on the NO2-saturated V2O5/WO3–TiO2 was ∼252 μmol g−1, as calculated in section 3.3, and the amounts of surface nitrates and NO consumption were ∼84 and ∼170 μmol g−1, respectively. Therefore, it can be suggested that ∼84 μmol g−1 NO2 can react with an equal amount of NO and twice the amount of NH3 according to the following equation:

 
NO + NO2 + 2NH3 → 2N2 + 3H2O(14)
which is a combination of eqn (1) and (2) and (15)–(20).
 
NO + [V5+]–ONO2 + [V4+]-□ ↔ 2[V5+]–ONO(15)
 
[Mn+]–NH3 + [V5+]–ONO → N2 + H2O + [Mn+] + [V5+]–OH(16)
 
[Mn+]–O–NH4 + [V5+]–ONO → N2 + H2O + [Mn+]–OH + [V5+]–OH(17)
 
NO + [Ti4+]–ONO2 + [Ti3+]-□ ↔ 2[Ti4+]–ONO(18)
 
[Mn+]–NH3 + [Ti4+]–ONO → N2 + H2O + [Mn+] + [Ti4+]–OH(19)
 
[Mn+]–O–NH4 + [Ti4+]–ONO → N2 + H2O + [Mn+]–OH + [Ti4+]–OH(20)
where Mn+ and [Mn+] denote a metal site and a Lewis acid site, respectively.

In this way, the remaining NH3 taking part in other reactions was 252–2 × 84 = 84 μmol g−1, and the remaining NO taking part in other reactions was 170–84 = 86 μmol g−1.

Since the ratio of the (remaining NO)/(remaining adsorbed NH3) was 84/86 ≈ 1/1, the following equation can be proposed:

 
4NO + 4(adsorbed NH3) + 2(O atom) → 4N2 + 6H2O(21)

Nitrite species were observed during NO adsorption on TiO2,41 and thus, eqn (21) might be a combination of eqn (5), (19) and (20).

 
NO + [Ti4+]2–O ↔ [Ti4+]–ONO + [Ti3+]-□(5)
 
[Mn+]–NH3 + [Ti4+]–ONO → N2 + H2O + [Mn+] + [Ti4+]–OH(19)
 
[Mn+]–O–NH4 + [Ti4+]–ONO → N2 + H2O + [Mn+]–OH + [Ti4+]–OH(20)

Many researchers reported that the E–R mechanism contributed to standard SCR on V-based catalysts19–22 following the equations:

 
NH3 + V5+[double bond, length as m-dash]O ↔ V4+–OH + NH2(22)
 
NH2 + NO → N2 + H2O(23)
where NH2 denotes a surface-adsorbed NH2 species.

Vuong et al.19 reported that it is not possible for nitrate to be formed by NO adsorption on V/Ce1−xTixO2, and thus, the E–R mechanism instead of the L–H mechanism occurred during standard SCR. However, since NO can be adsorbed on V2O5/WO3–TiO2, as shown in Fig. 3b, forming surface nitrates and nitrites,41 the L–H mechanism took place at low temperatures in this study. Analogously, Wang et al.16 reported that the L–H mechanism played a primary part in improving the NH3-SCR activity on V2O5/ASC at low temperatures. Valdés-Solís et al.18 also found that NO in the adsorbed phase participated in NH3-SCR by reacting with the adsorbed NH3.

It was reported that W provides the active site for NH3 adsorption,32 V species (designated as [V5+]–O, [V4+]–OH, and [V5+]–OH) is mainly responsible for NH3 adsorption,38 and [V4+]–OH and TiO2 can provide the active sites for both NO2 and NH3 adsorption.39,40,49 In the following calculations, we designated the amounts of [V4+]–OH, [V5+]–OH, [V5+]–O, [Ti4+]2–O, and [W site] as α, β, γ, δ and ζ, respectively. The adsorption of one NO molecule (13 μmol g−1 at 30 °C) needs two [Ti4+]2–O sites to form one nitrate species, and thus, the amount of the [Ti4+]2–O sites was δ = 13 × 2 = 26 μmol g−1. Since the adsorption of one NO2 molecule needs one [V5+]–O or one [V4+]–OH site to form one nitrate species, and three NO2 molecules need two [V5+]–OH or one [Ti4+]2–O to form two nitrate species and one NO gas molecule, the overall amount of NO formation during NO2 adsorption was β/2 + δ, i.e., 40 μmol g−1 (Fig. 1), and the overall amount of NO2 consumption was α + 3 × β/2 + γ + 3 × δ, i.e., 144 μmol g−1 (Fig. 1). The adsorption of NH3 took place on one [V4+]–OH, one [V5+]–OH, one [V5+]–O, and one [W site] to form one adsorbed NH3, or on one [Ti4+]2–O to form two adsorbed NH3, so that the overall amount of NH3 adsorbed was α + β + γ + 2 × δ + ζ, i.e., 378 μmol g−1 (Fig. 4). Thus, the following equations can be obtained:

image file: c9cy01370a-t2.tif

After solving the equations, we can obtain: α + γ = 24, β = 28, δ = 26 and ζ = 274. It is difficult to determine the values of α and γ, since the molar ratio of V4+/V5+ was unknown. Therefore, the amounts of [V4+]–OH + [V5+]–O, [V5+]–OH, [Ti4+]2–O and [W site] species were 24, 28, 26 and 274 μmol g−1, respectively, corresponding to 0.241 ([V4+]–OH + [V5+]–O) nm−2, 0.281 [V5+]–OH nm−2, 0.261 [Ti4+]2–O nm−2, and 2.75 [W site] nm−2, respectively. It was reported that the monolayer coverage of V2O5 and tungsten oxide was achieved when the amounts of V2O5 and tungsten oxide in V2O5–WO3/TiO2 were below 6% and 8%, respectively.50,51 Since the amounts of vanadium oxide and tungsten oxide in V2O5/WO3–TiO2 in this study were 3.5 and 10 wt%, respectively, it can be inferred that, theoretically, the V species dispersed as a monolayer, but crystalline tungsten oxide was formed and could not disperse as a monolayer. The theoretical amount of the V species on the surface of the catalyst can be calculated to be 3.5% ÷ 184 = 190 μmol g−1. However, as we calculated above based on the adsorption results, the overall amount of the V species on the surface was α + γ + β = 24 + 28 = 52 μmol g−1, which was much lower than the theoretical amount of the V species (190 μmol g−1), indicating that the V species did not distribute as a monolayer. Although the amount of vanadium oxide (3.5 wt%) in this study was below the monolayer limit (6 wt%), the excessive tungsten species (10 wt%) on TiO2 could lead to the formation of clustered VOx or crystalline V2O5, as reported by Wang et al.47

At 150 °C, most of the [Ti4+]2–O sites could not adsorb NO2 to form nitrates and the NO gas (Fig. 1) since the nitrates formed on these sites were weak, and most of these sites might be spectators during NO2 adsorption at 150 °C. In addition, [V5+]–OH lost OH groups at 150 °C and was probably converted to [V5+]–O, adsorbing NO2 following eqn (2). Thus, [V4+]–OH and [V5+]–O mainly contributed to the adsorption of NO2. The [Ti4+]2–O sites were responsible for NO adsorption to generate nitrates39 during standard SCR. As shown in Fig. 4a, the amount of NH3 adsorption was ∼168 μmol g−1 at 150 °C, which was much less than that of the W sites (274 μmol g−1). It was reported that WOx facilitated the formation of the strongly adsorbed NH3 species, which were much more strongly adsorbed than those formed on the TiO2 or V2O5 species.52 Similarly, Zhu et al.32 found that the adsorbed NH3 on the Ti4+ sites in V2O5–WO3/TiO2 doesn't contribute to the NH3-SCR reaction. Therefore, it can be proposed that NH3 mainly adsorbed on the W species at 150 °C. In conclusion, the actual reaction sites involved 24 μmol g−1 [V4+]–OH + [V5+]–O, 28 μmol g−1 [V5+]–O (originating from [V5+]–OH), 24 μmol g−1 [Ti4+]2–O and 168 μmol g−1 [W sites].

Fig. 7 summarizes the active site distribution on V2O5/WO3–TiO2 and the standard and fast SCR mechanisms at 150 °C. The mechanism on V2O5/WO3–TiO2 does not include the reaction between NH4NO3 and NO (the “NH4NO3 path”) to produce N2, NO2, and H2O. As seen in Fig. 5, NO was consumed below 180 °C. Koebel et al.53 reported that NH4NO3 could decompose to NH3 and HNO3. Thus, it is possible that NH4NO3 on V2O5/WO3–TiO2 participated in the reaction by decomposing to NH3 and nitrates (reverse of eqn (11) and (12)), and then, one nitrate species reacted with one NO molecule to produce two nitrites, which would then react with NH3 to generate N2 (the “nitrite path”). Therefore, the “nitrite path” plays an important role in both fast and standard SCR mechanisms on V2O5/WO3–TiO2. The “nitrite path” has been reported by a lot of researchers during standard SCR. Bendrich et al.54 reported that the “nitrite path” occurred on Cu-CHA in standard SCR. On CeWOx (ref. 46) and a Mn/Fe–Ti spinel catalyst,55 the “nitrite path” has also been suggested to contribute to standard SCR. On V2O5/WO3–TiO2 during standard SCR, the nitrites might originate from the direct adsorption of NO on TiO2 (eqn (5)) or from the reaction between the nitrates formed by NO adsorption on TiO2 and the NO gas (eqn (8) and (9)), while during fast SCR, the nitrites formed mainly originated from the reaction between the nitrates generated by NO2 adsorption and the NO gas. As discussed in section 3.2, the reaction between surface nitrates and NO to form nitrites was a slow step on V2O5/WO3–TiO2, and thus, was the rate limiting step for fast SCR. Similarly, this step was also considered as the rate limiting step on Fe-ZSM-5 in fast SCR at low temperatures as studied by Grossale et al.56 It is possible that a greater amount of nitrite species was formed during fast SCR in comparison to that during standard SCR, since NO2 adsorption formed more surface nitrates than NO adsorption (Table 1), which might react with NO to form nitrite species, which is beneficial for NH3-SCR. This was one of the reasons why the fast SCR activity was higher than the standard SCR activity.


image file: c9cy01370a-f7.tif
Fig. 7 NH3-SCR mechanism on V2O5/WO3–TiO2 without H2O in the feed at 150 °C.
Table 1 Number of different active sites on V2O5/WO3–TiO2
Condition Temperature Sites for NO Sites for NO2 Sites for NH3
Without H2O 30 °C [Ti4+]2–O: 26 μmol g−1 [V4+]–OH + [V5+]–O: 24 μmol g−1 378 μmol g−1
[V5+]–OH: 28 μmol g−1
[Ti4+]2–O: 26 μmol g−1
150 °C [Ti4+]2–O: 24 μmol g−1 54 μmol g−1 168 μmol g−1
With H2O 30 °C 37 μmol g−1 91 μmol g−1 377 μmol g−1
150 °C 32 μmol g−1 88 μmol g−1 221 μmol g−1


3.5 NH3-SCR mechanism in the presence of H2O

The effect of H2O on the adsorption amounts of NO2, NO, and NH3 and the NH3-SCR mechanism on V2O5/WO3–TiO2 was also investigated. No N2O, N2O3, N2O4, or N2O5 emission was observed during the adsorption process. As shown in Fig. 8, the pre-adsorption of H2O did not affect the consumption of NO2 or NH3 at 30 °C. The ratio of the NO2 consumed to the NO gas generated was ∼3[thin space (1/6-em)]:[thin space (1/6-em)]1 (143[thin space (1/6-em)]:[thin space (1/6-em)]52) (Fig. 8a), pointing out that eqn (24) took place.
 
3NO2 + H2O ↔ 2HNO3 + NO(24)

image file: c9cy01370a-f8.tif
Fig. 8 Adsorption of (a) NO2/N2 (500 ppm) and (b) NH3/N2 (500 ppm) on V2O5/WO3–TiO2 pretreated using 2 vol% H2O, and (c) adsorption of NO/N2 (500 ppm) and 2 vol% H2O on fresh V2O5/WO3–TiO2 at 30 and 150 °C.

However, at 150 °C, the NO2 adsorption amount (∼88 μmol g−1) increased compared with that on fresh V2O5/WO3–TiO2 (∼60 μmol g−1), and no NO was evolved, indicating that eqn (24) did not occur during this process, due to the loss of H2O at higher temperatures (150 °C). The presence of H2O increased the amount of adsorbed NO2 at the reaction temperature, possibly due to the production of OH groups on V4+ from the decomposition of H2O and the promotion of the reaction in eqn (1). Also, the amount of NH3 adsorption at 150 °C grew by ∼53 μmol g−1, indicating that H2O could improve the adsorption of NH3 at 150 °C (Table 1). Many researchers57,58 reported that competitive adsorption between NH3 and H2O took place, leading to a decreased amount of NH3 adsorption, and further, lower NH3-SCR activity. However, in this study, the amount of the adsorbed NH3 was promoted on V2O5/WO3–TiO2. Similarly, Giraud et al.59 found that H2O boosted the amount of the adsorbed NH3 on V2O5/WO3/TiO2 at 473 K. Huang et al.33 also reported the promotion of NH3 adsorption by H2O on V2O5/AC, due to the increase in Brønsted acid sites. It is possible that adsorbed H2O covers the surface of V2O5/WO3–TiO2 at 150 °C, and NO2 and NH3 mainly adsorb on H2O molecules, modifying the adsorption amount and strength of the reactants. Therefore, the promotion of NH3 adsorption by H2O in the present study could be due to the increase in Brønsted acid sites. The active sites for NO2 and NH3 activation with H2O at 30 °C were partially different from those in the absence of H2O, since NO2 and NH3 adsorbed on H2O in the presence of water.

The amounts of weakly adsorbed NO2 and NH3 molecules that could be removed by N2 purging were also calculated. For NO2, the amount that could be removed was negligible at 30 °C, whereas for NH3, the amounts removed were 130 ± 12 and 41 ± 5 μmol g−1 at 30 and 150 °C, respectively. The overall NH3 adsorption amounts on the H2O pretreated catalyst were ∼377 and ∼221 μmol g−1 at 30 and 150 °C (Fig. 8), therefore, the amounts of NH3 remaining on the surface of V2O5/WO3–TiO2 after N2 purging can be calculated as follows:

30 °C: 377 μmol g−1–130 μmol g−1 = 247 μmol g−1

150 °C: 221 μmol g−1–41 μmol g−1 = 180 μmol g−1.

At 30 °C, the amount of NH3 adsorption on the NO2-pretreated catalyst was ∼478 μmol g−1, and thus, the amount of adsorbed NH3 can be calculated as follows:

478 μmol g−1 – 130 μmol g−1 = 348 μmol g−1.

However, the presence of H2O facilitated NO adsorption on V2O5/WO3–TiO2 at both 30 and 150 °C (Fig. 8c), and the overall NO adsorption amounts were ∼37 and ∼32 μmol g−1 at 30 and 150 °C, respectively. However, the amounts that could be removed by N2 purging were 34 ± 3 and 30 ± 2 μmol g−1 at 30 and 150 °C, respectively, as shown in Fig. 9. Most of these NO molecules could be removed at either 30 or 150 °C, and were weakly adsorbed (Fig. 9), indicating that NO adsorbed weakly on the H2O-containing V2O5/WO3–TiO2 catalyst, and H2O could facilitate the weak adsorption of NO. It is well known that the weakly adsorbed species on the fresh catalyst cannot take part in the catalytic reaction, and therefore, it can be inferred that most of the adsorbed NO molecules on the H2O pretreated catalyst cannot contribute to the NH3-SCR reaction.


image file: c9cy01370a-f9.tif
Fig. 9 (a) Adsorption of NH3/N2 (500 ppm) or NO2/N2 (500 ppm) on V2O5/WO3–TiO2 saturated with H2O, and subsequently with NH3/N2 or NO2/N2 (500 ppm) followed by N2 purging at 30 and 150 °C; (b) adsorption of NO/N2 (500 ppm) and H2O (2 vol%) on V2O5/WO3–TiO2 saturated with NO and H2O followed by N2 purging at 30 and 150 °C.

The TPSR in NO + H2O of V2O5/WO3–TiO2 pretreated in H2O, NO2 and NH3 successively at 30 °C was also performed to clarify the impact of H2O on the NH3-SCR mechanism, and the results are shown in Fig. 10. The NO consumption and NH3 and NO2 production amounts were 189, 93, and 32 μmol g−1, respectively, and the numbers of surface NO2 and NH3 presorbed were ∼88 (Fig. 8a) and ∼348 μmol g−1, respectively, demonstrating that the surface adsorbed NO2 and NH3 species reacted with the NO gas in accordance with the ratios of NO/NH3 of 1/1 (101/101) (eqn (21)), NO/NH4NO3/NH3 of 1/1/1 (56/56/56), and NO/NH4NO3/(NO2 evolved) of 1/1/1 (32/32/32) (eqn (13)). In the case of NO/NH4NO3/NH3 1/1/1 (56/56/56), the following reactions might occur:


image file: c9cy01370a-f10.tif
Fig. 10 TPSR of pretreated V2O5/WO3–TiO2 in NO/N2 (500 ppm) and 2% H2O. The pretreatment conditions: H2O until saturation, NO2/N2 (500 ppm) until saturation and then NH3/N2 (500 ppm) until saturation at 30 °C.

Formation and decomposition of NH4NO3:

 
NH3 + HNO3 ↔ NH4NO3(25)

Reaction between NO and HNO3:

 
HNO3 + NO + * ↔ HNO2 + *-ONO(26)

Adsorption of NH3:

 
NH3 + # ↔ #-NH3(27)

Reactions between adsorbed NH3 and nitrite species:

 
HNO2 + #-NH3 → N2 + 2H2O + #(28)
 
*-ONO + #-NH3 → N2 + H2O + *-OH + #(29)
where * and # denote the active sites on V2O5/WO3–TiO2 for NO2 adsorption to form nitrites and for NH3 adsorption, respectively, which might be identical to the active sites in the absence of H2O, or might be new active sites modified by the H2O or OH molecules.

Fig. 11 summarizes the NH3-SCR mechanism with H2O in the feed. Eqn (13), the reaction between NH4NO3 and the NO gas to generate N2, H2O and NO2 gas (the “NH4NO3 path”), is absent from the mechanism in the absence of H2O, as discussed in section 3.4. However, in the presence of H2O, both the “NH4NO3 path” and the “nitrite path” were present. The role of the “NH4NO3 path” was probably not significant in standard SCR on V2O5/WO3–TiO2 since the number of nitrates generated by NO adsorption was significantly smaller than that generated by NO2 adsorption. Also, a large amount of NH3 (93 μmol g−1) started to desorb below 100 °C (Fig. 10). As Zhu et al.32 reported, H2O increased the amount of the less-active NH4+ species (NH3 on Brønsted acid sites) at the expense of the more-active NH3 species (NH3 on Lewis acid sites) on V2O5/WO3–TiO2, inhibiting NH3-SCR. Therefore, it can be suggested that the desorbed NH3 shown in Fig. 10 arises from NH3 on the new Brønsted acid sites produced by H2O which is less reactive during NH3-SCR, and preferentially desorbed rather than participating in the reaction. In the TPSR process in NO with H2O, a maximal NO consumption occurred at ∼135 °C. However, this temperature was lower when H2O was absent in the feed (∼120 °C), demonstrating that H2O hindered the reaction between the adsorbed NH3 and the NO gas. Analogously, Huang et al.33 suggested that on V2O5/AC, the presence of H2O hindered the reaction between NO and the NH3 molecules activated on the Lewis acid sites. Arnarson et al.60 investigated the standard SCR reaction over a vanadia titania model catalyst at low temperatures using DFT calculations and found that H2O prohibited the reoxidation of the reduced site and hindered the standard SCR. The reasons above can decrease the SCR activity with H2O in the feed.


image file: c9cy01370a-f11.tif
Fig. 11 NH3-SCR mechanism on V2O5/WO3–TiO2 with H2O in the feed at 150 °C.

4 Conclusions

In this study, the NH3-SCR reaction pathway at low temperatures, as well as the active site dispersion, over a V2O5/WO3–TiO2 catalyst was quantitatively determined, and the inhibition effects of H2O on the NH3-SCR reaction were also clarified.

On the fresh V2O5/WO3–TiO2 catalyst at low temperatures, the “nitrite path” contributed to standard and fast SCR mechanisms. The nitrite species originated from either the direct adsorption of NO on the catalyst or the reaction between NO gas and nitrates formed from NOx adsorption. During fast SCR, the reaction between NO gas and nitrates to produce nitrites might be the rate limiting step. However, the “NH4NO3 path” was absent without H2O in the gas feed. An increase in the amount of NH3 and NO2 adsorption at 150 °C was observed in the presence of H2O. Both the “NH4NO3 path” and the “nitrite path” were present in the presence of H2O on V2O5/WO3–TiO2. H2O inhibited the reaction between the adsorbed NH3 molecules and the NO gas, lowering the SCR activity. The amount of NH3 adsorbed significantly exceeded that of NOx, and the amounts of [V4+]–OH + [V5+]–O, [V5+]–OH, [Ti4+]2–O and [W site] species were 24, 28, 26, and 274 μmol g−1 on V2O5/WO3–TiO2, respectively.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the National Key R&D Program of China (2017YFC0211101), the Key Project of National Natural Science Foundation (21637005) and the National Natural Science Foundation of China (21872168, 51822811).

Notes and references

  1. G. Busca, L. Lietti, G. Ramis and F. Berti, Appl. Catal., B, 1998, 18, 1 CrossRef CAS.
  2. W. Shan and H. Song, Catal. Sci. Technol., 2015, 5, 4280 RSC.
  3. J. Li, H. Chang, L. Ma, J. Hao and R. T. Yang, Catal. Today, 2011, 175, 147 CrossRef CAS.
  4. G. He, Z. Lian, Y. Yu, Y. Yang, K. Liu, X. Shi, Z. Yan, W. Shan and H. He, Sci. Adv., 2018, 4, eaau4637 CrossRef CAS PubMed.
  5. A. Marberger, D. Ferri, M. Elsener and O. Krocher, Angew. Chem., Int. Ed., 2016, 55, 11989 CrossRef CAS PubMed.
  6. J. K. Lai and I. E. Wachs, ACS Catal., 2018, 8, 6537 CrossRef CAS.
  7. T. Xu, X. Wu, X. Liu, L. Cao, Q. Lin and D. Weng, J. Environ. Sci., 2017, 57, 110 CrossRef PubMed.
  8. A. M. Beale, F. Gao, I. Lezcano-Gonzalez, C. H. F. Peden and J. Szanyi, Chem. Soc. Rev., 2015, 44, 7371 RSC.
  9. U. Deka, I. Lezcano-Gonzalez, B. M. Weckhuysen and A. M. Beale, ACS Catal., 2013, 3, 413 CrossRef CAS.
  10. R. Oord, I. C. ten Have, J. M. Arends, F. C. Hendriks, J. Schimidt, I. Lezano-Gonzalez and B. M. Weckhuysen, Catal. Sci. Technol., 2017, 7, 3851 RSC.
  11. A. Wang, Y. Wang, E. D. Walter, N. M. Wahton, Y. Guo, G. Lu, C. H. F. Peden and F. Gao, Catal. Today, 2019, 320, 91 CrossRef CAS.
  12. F. Liu, Y. Yu and H. He, Chem. Commun., 2014, 50, 8445 RSC.
  13. X. Mou, B. Zhang, Y. Li, L. Yao, X. Wei, D. Su and W. Shen, Angew. Chem., Int. Ed., 2012, 51, 2989 CrossRef CAS PubMed.
  14. Z. Duan, J. Liu, J. Shi, Z. Zhao, Y. Wei, X. Zhang, G. Jiang and A. Duan, J. Environ. Sci., 2018, 65, 1–7 CrossRef PubMed.
  15. Z. Huang, X. Gu, W. Wen, P. Hu, M. Makkee, H. Lin, F. Kapteijn and X. Tang, Angew. Chem., Int. Ed., 2013, 52, 660 CrossRef CAS PubMed.
  16. J. Wang, Z. Yan, L. Liu, Y. Chen, Z. Zhang and X. Wang, Appl. Surf. Sci., 2014, 313, 660 CrossRef CAS.
  17. B. Yun and M. Kim, Appl. Therm. Eng., 2013, 50, 152 CrossRef CAS.
  18. T. Valdés-Solís, G. Marbán and A. B. Fuertes, Ind. Eng. Chem. Res., 2004, 43, 2349 CrossRef.
  19. T. H. Vuong, J. Radnik, J. Rabeah, U. Bentrup, M. Schneider, H. Atia, U. Armbruster, W. Grünert and A. Brückner, ACS Catal., 2017, 7, 1693 CrossRef CAS.
  20. Z. Lian, F. Liu and H. He, Ind. Eng. Chem. Res., 2014, 53, 19506 CrossRef CAS.
  21. Z. Lei, B. Han, K. Yang and B. Chen, Chem. Eng. J., 2013, 215-216, 651 CrossRef CAS.
  22. S. Soyer, A. Uzun, S. Senkan and I. Onal, Catal. Today, 2006, 118, 268 CrossRef CAS.
  23. S. Xiong, X. Xiao, Y. Liao, H. Dang, W. Shan and S. Yang, Ind. Eng. Chem. Res., 2015, 54, 11011 CrossRef CAS.
  24. W. Zhao, S. Dou, Q. Zhong, L. Wu, Q. Wang and A. Wang, Russ. J. Phys. Chem. A, 2017, 91, 2489 CrossRef CAS.
  25. J. Cheng, Y. Song, Q. Ye, S. Cheng, T. Kang and H. Dai, Mol. Catal., 2018, 445, 111 CrossRef CAS.
  26. C. Ciardelli, I. Nova, E. Tronconi, D. Chatterjee and B. Bandl-Konrad, Chem. Commun., 2004, 2718 RSC.
  27. I. Nova, C. Ciardelli, E. Tronconi, D. Chatterjee and B. Bandl-Konrad, Catal. Today, 2006, 114, 3 CrossRef CAS.
  28. I. Nova, C. Ciardelli, E. Tronconi, D. Chatterjee and M. Weibel, AIChE J., 2009, 55, 1514 CrossRef CAS.
  29. X. Wang, A. Shi, Y. Duan, J. Wang and M. Shen, Catal. Sci. Technol., 2012, 2, 1386 RSC.
  30. L. Gan, J. Chen, Y. Peng, J. Yu, T. Tran, K. Li, D. Wang, G. Xu and J. Li, Ind. Eng. Chem. Res., 2018, 57, 150 CrossRef CAS.
  31. K. Zheng, G. Yang, W. Shen, Q. Xu, F. Hu and Z. Li, Water, Air, Soil Pollut., 2018, 229, 109 CrossRef.
  32. M. Zhu, J.-K. Lai, U. Tumuluri, Z. Wu and I. E. Wachs, J. Am. Chem. Soc., 2017, 139, 15624 CrossRef CAS.
  33. Z. Huang, Z. Liu, X. Zhang and Q. Liu, Appl. Catal., B, 2006, 63, 260 CrossRef CAS.
  34. K. Liu, H. He, Y. Yu, Z. Yan, W. Yang and W. Shan, J. Catal., 2019, 369, 372 CrossRef CAS.
  35. R. H. Kagann and A. G. Maki, J. Quant. Spectrosc. Radiat. Transfer, 1984, 31, 173 CrossRef CAS.
  36. K. I. Hadjiivanov, Catal. Rev.: Sci. Eng., 2000, 42(1&2), 71 Search PubMed.
  37. X. Gao, X. Du, Y. Jiang, Y. Zhang, Z. Luo and K. Cen, J. Mol. Catal. A: Chem., 2010, 317, 46 CrossRef CAS.
  38. M. Kantcheva, V. Bushev and D. Klissurski, J. Catal., 1994, 145, 96 CrossRef CAS.
  39. L. Sivachandiran, F. Thevenet, P. Gravejat and A. Rousseau, Appl. Catal., B, 2013, 142-143, 196 CrossRef CAS.
  40. J. A. Odriozola, H. Heinemann, G. A. Somorjai, J. F. Garcia De, La Banda and P. Pereira, J. Catal., 1989, 119, 71 CrossRef CAS.
  41. K. Hadjiivanov and H. Knözinger, Phys. Chem. Chem. Phys., 2000, 2, 2803 RSC.
  42. G. Ramis, G. Busca, V. Lorenzelli and P. Forzatti, Appl. Catal., 1990, 64, 243 CrossRef CAS.
  43. I. Nova, C. Ciardelli, E. Tronconi, D. Chatterjee and M. Weibel, Top. Catal., 2007, 42-43, 43 CrossRef.
  44. M. P. Ruggeri, A. Grossale, I. Nova, E. Tronconi, H. Jirglova and Z. Sobalik, Catal. Today, 2012, 184, 107 CrossRef CAS.
  45. A. Grossale, I. Nova, E. Tronconi, D. Chatterjee and M. Weibel, Top. Catal., 2009, 52, 1837 CrossRef CAS.
  46. K. Liu, F. Liu, L. Xie, W. Shan and H. He, Catal. Sci. Technol., 2015, 5, 2290 RSC.
  47. C. Wang, S. Yang, H. Chang, Y. Peng and J. Li, Chem. Eng. J., 2013, 225, 520 CrossRef CAS.
  48. X. Li, X. Wu, T. Xu, D. Weng, Z. Si and R. Ran, Chin. J. Catal., 2016, 37, 1340 CrossRef.
  49. L. Huang, K. Zha, S. Namuangruk, A. Junkaew, X. Zhao, H. Li, L. Shi and D. Zhang, Catal. Sci. Technol., 2016, 6, 8516 RSC.
  50. M. A. Vuurman, I. E. Wachs and A. M. Hirt, J. Phys. Chem., 1991, 95, 9928 CrossRef CAS.
  51. T. Machej, J. Haber, A. M. Turek and I. E. Wachs, Appl. Catal., 1991, 70, 115 CrossRef CAS.
  52. F. Giraud, C. Geantet, N. Guilhaume, S. Loridant, S. Gros, L. Porcheron, M. Kanniche and D. Bianchi, J. Phys. Chem. C, 2015, 119, 15401 CrossRef CAS.
  53. M. Koebel, G. Madia and M. Elsener, Catal. Today, 2002, 73, 239 CrossRef CAS.
  54. M. Bendrich, A. Scheuer, R. E. Hayes and M. Votsmeier, Appl. Catal., B, 2018, 222, 76 CrossRef CAS.
  55. S. Yang, F. Qi, S. Xiong, H. Dang, Y. Liao, P. K. Wong and J. Li, Appl. Catal., B, 2016, 181, 570 CrossRef CAS.
  56. A. Grossale, I. Nova, E. Tronconi, D. Chatterjee and M. Weibel, J. Catal., 2008, 256, 312 CrossRef CAS.
  57. S. Ding, F. Liu, X. Shi, K. Liu, Z. Lian, L. Xie and H. He, ACS Appl. Mater. Interfaces, 2014, 7, 9497 CrossRef PubMed.
  58. W. Shan, F. Liu, H. He, X. Shi and C. Zhang, Chem. Commun., 2011, 47, 8046 RSC.
  59. F. Giraud, J. Couble, C. Geantet, N. Guilhaume, S. Loridant, S. Gros, L. Porcheron, M. Kanniche and D. Bianchi, J. Phys. Chem. C, 2018, 122(43), 24619 CrossRef CAS.
  60. L. Arnarson, H. Falsig, S. B. Rasmussen, J. V. Lauritsen and P. G. Moses, J. Catal., 2017, 346, 188 CrossRef CAS.

This journal is © The Royal Society of Chemistry 2019