Effect of lean-oxygen treatment on the adsorption and activity of zirconium phosphate @ Ce0.75Z0.25O2 for NH3-SCR deNOx

Catalysis Today 267 (2016) 47–55

Contents lists available at ScienceDirect

Catalysis Today journal homepage: www.elsevier.com/locate/cattod

Effect of lean-oxygen treatment on the adsorption and activity of zirconium phosphate @ Ce0.75 Z0.25 O2 for NH3 -SCR deNOx Jun Yu a , Zhichun Si a,∗ , Xuankun Li a , Lei Chen a,c , Xiaodong Wu a,b , Duan Weng a,b a

Graduate School at Shenzhen, Tsinghua University, Shenzhen 518055, China The Key Laboratory of Advanced Materials of Ministry of Education, School of Materials Science and Engineering, Tsinghua University, Beijing 100084, China c China development strategy institute for building materials industry, 100035, China b

a r t i c l e

i n f o

Article history: Received 11 September 2015 Received in revised form 14 January 2016 Accepted 14 January 2016 Available online 12 February 2016 Keywords: NH3 -SCR Lean-oxygen treatment Activity Adsorption Redox

a b s t r a c t As oxygen is one of the essential reactants in the standard SCR reactions, the ability of restoring and releasing oxygen is an important factor to determine the SCR activity of the catalyst. Ceria catalysts present an innate prospect in SCR reaction because of the facilitated redox cycles from Ce4+ to Ce3+ . In the present study, zirconium phosphate @ Ce0.75 Z0.25 O2 (ZP/CZ) catalysts were pre-treated by O2 , N2 and H2 to study the effects of lean-oxygen teatment on the structure, NH3 /NOx adsorption and activity of catalysts. The catalysts were characterized by activity test, H2 -TPR, XPS, DRIFT, NOx -TPD and kinetics study. The results showed that H2 treatment led to deeply reduced catalyst surface, resulting in mainly nitrites instead of nitrates adsorbed on catalyst in NO + O2 reaction and reduced Brønsted acidity of catalyst, which were responsible for the lowered deNOx activity of H2 treated catalyst. N2 treatment had only negligible influence on catalyst compared to O2 treated catalyst. Ammonium nitrate route was consolidated as possible NH3 -SCR reaction mechanism over ZP/CZ catalyst by kinetics study. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Nitrogen oxides (NO, NO2 and N2 O) deriving from the automobile exhausts and the flue gases, lead to various environmental problems, such as PM2.5 and acid rain. Selective catalytic reduction of NOx by ammonia (NH3 -SCR) has been proven to be one of the most promising deNOx technologies. The most widely employed NH3 -SCR catalyst in the industry is V2 O5 –WO3 (or MoO3 )/TiO2 which shows an excellent deNOx performance and stability in the temperature range of 300–400 ◦ C [1–3]. Transition-metal exchanged zeolites also present satisfactory NH3 -SCR performances for deNOx in a wide temperature range [4–6]. Recently, nontoxic CeO2 -based NH3 -SCR catalysts have received much attention due to their good hydrothermal stability and high deNOx efficiency which are derived from the good redox property of ceria [7–17]. Pure ceria shows rather poor SCR activity in the temperature range of 200–500 ◦ C [18]. The high active surface oxygen of ceria catalyst results in NH3 oxidation on the surface of the catalyst especially at high temperatures. The acidic components with low

∗ Corresponding author. Fax: +86 755 26036417. E-mail address: [email protected] (Z. Si). http://dx.doi.org/10.1016/j.cattod.2016.01.013 0920-5861/© 2016 Elsevier B.V. All rights reserved.

redox properties have been proved to be essential for enhancing the deNOx efficiency of ceria [9–13], which means that improving the ammonia adsorption and inhibiting the ammonia oxidation on ceria are key important of ceria-based NH3 -SCR catalyst. In our previous report [19–21], ZP/CZ catalyst shows over 80% NOx conversion at 250–450 ◦ C. Porous acidic ZP/CZ can reduce the strong interaction between phosphate and cerium in addition to introducing acid sites on surface of catalyst [20]. The mobility of surface lattice oxygen on ZP/CZ catalyst is essential to a catalyst with high NH3 -SCR activity at low temperatures, and NH4 NO3 is proved to be an important intermediate for NH3 -SCR reaction on ceria based catalyst [20]. After hydrothermal aging and sulfur aging, ZP/CZ catalyst still possessed similar performances with those of home made Cu-SAPO-34 and vanadium catalysts at higher temperatures. Ammonia activation and nitrite/nitrate formation depend on the surface/lattice oxygen of ceria, which can vary with real conditions because both oxidation and reduction conditions may occur in mobile exhaust. The surface oxygen has been reported to be highly active in the NH3 /NO oxidation reaction due to its higher mobility than lattice oxygen [22,23]. And the high mobility of lattice oxygen was essential for a catalyst of high NH3 -SCR activity [22,23]. If the catalyst was used in diesel exhaust deNOx , soot and hydrocarbon may deposit on catalyst. The removal of soot and hydrocarbon may cause the temporary reduction conditions on catalyst surface. On

48

J. Yu et al. / Catalysis Today 267 (2016) 47–55

the other hand, the burning regeneration of diesel particulate filter (DPF) by fuel may also lead to the reduction condition of SCR catalyst. Even the SCR-only system may be exposed to a rich feed under certain conditions such as cold-start period, extended vehicle idling, engine malfunction, or degraded upstream diesel oxidation catalysts (DOC). All of these scenarios raise questions regarding the potential impact of rich reducing agents on the performance of the SCR catalyst [24]. In the present study, ZP/CZ catalysts were treated by O2 , N2 and H2 at 500 ◦ C. And the effect oxygen-lean/rich treatment on the structure, adsorption and catalytic performances was studied. 2. Experimental 2.1. Materials and preparation method All the agents and materials (AR grade) for synthesizing the catalysts were from Aladdin Industrial Corporation, China. CeO2 –ZrO2 powder was synthesized by the precipitation method. An appropriate amount of cerium (III) chloride heptahydrate and zirconyl chloride octahydrate were dissolved in deionized water, and then ammonia solution (25%) was added dropwise until the pH of the mixed solution reached 9–10. The obtained mixture was aged at 60 ◦ C for 20 h, followed by spray drying at 200 ◦ C to get the spherical powders. After that, the obtained powders were calcined at 500 ◦ C for 3 h under atmosphere to get the CeO2 –ZrO2 powder. ZP/CZ catalyst was prepared by impregnating ammonia phosphate and zirconyl chloride octahydrate simultaneously on CeO2 –ZrO2 powder according to the method in reference [20]. ZP/CZ catalyst was treated by 1% H2 (v/v)/N2 , or pure N2 , or 10% (v/v) O2 /N2 flow (500 ml min−1 ) for 30 min at 500 ◦ C for 30 min, named as ZP/CZ-O2 , ZP/CZ-N2 and ZP/CZ-H2 respectively. 2.2. Characterizations The X-ray photoelectron spectroscopy (XPS) experiments were carried out on a PHI-QuanteaSXM system equipped with a monochromatic Al K␣ X-rays under UHV (6.7 × 10−8 Pa). The binding energies were calibrated internally by C 1s binding energy (BE) at 284.4 eV. Before experiment, each sample was pre-treated at 105 ◦ C in vacuum for 2 h. DRIFTS of NH3 -derived species on catalyst at 250 ◦ C arising from contact with NH3 were recorded in the range of 4000–650 cm−1 using a thermo Nicolet 6700 FTIR spectrometer. The sample was treated at 500 ◦ C in a 20% (v/v) O2 /N2 flow (100 ml min−1 ). Then, the sample was cooled down to 250 ◦ C, and subsequently purged by N2 (100 ml min−1 ) for 30 min for background collection. Afterthat, the gas mixture containing 1000 ppm NH3 in N2 (100 ml min−1 ) was flowed through the sample for 30 min to obtain the adsorbed samples. After N2 purging for 30 min, the spectra were collected. Temperature-programmed desorption of NH3 /NOx (NH3 -TPD, NO/NO2 -TPD) experiments were carried out in a quartz fixed bed reactor. 0.2 g of powder catalyst was preheated in flowing 5% O2 /N2 at 500 ◦ C for 30 min to remove any possible physic-adsorbed impurities. After cooling down to 100 ◦ C, the sample was exposed to a flow of 1000 ppm NO or NH3 + 5% O2 in N2 for 30 min, and then flushed by 500 ml min−1 N2 for 30 min to remove the physicaladsorbed molecules. After that, the catalyst was heated up to 500 ◦ C at a rate of 10 ◦ C min−1 . The NH3 /NOx concentrations were determined by Thermo Nicolet iS 10 FTIR spectrometer through a 2 m path-length gas sample cell at 120 ◦ C. H2 temperature-programmed reduction (H2 -TPR) was performed in a fixed-bed reactor with the effluent gases monitored using a quadrupole mass spectrometer (MS) (Omnistar 200).

Prior to H2 -TPR experiment, 50 mg samples was treated with O2 (2 vol%)/He with a total flow rate of 50 ml min−1 at 500 ◦ C for 30 min, then cooled down to RT in the same atmosphere, and subsequently, flushed by 50 ml min−1 /He for 30 min to remove the physically adsorbed molecules. Finally, the reactor temperature was raised to 900 ◦ C at a constant heating rate of 10 ◦ C min−1 in H2 (5 vol.%)/He with a flow rate of 50 ml min−1 . H2 consumption during the experiment was monitored by MS. 2.2.1. Oxygen storage capacity (OSC) property tests Typically, 25 mg powders were loaded into a quartz tube reactor (i.d. = 1.0 cm) and a total gas flow rate of 100 ml min−1 was employed. The signals of the outlet gas were detected by an online quadrupole mass spectrometer (Omnistar 200). Prior to the OSC measurements, all the samples were first heated in 2% O2 /He at 500 ◦ C for at least 30 min. For OSC measurements, the sample was purged in pure He for 10 min to remove oxygen from the system and then exposed constantly to 4% CO/He at 400 ◦ C. Instantaneous CO2 concentration was monitored as a function of time, and the so-called total OSC was obtained by integrating the amount of CO2 formed during the first 2 min. 2.3. Activity measurement The activity measurement for NO reduction by ammonia was carried out in a fixed bed reactor made of quartz tube. Samples of 200 mg (0.1 ml) sieved to 40–80 mesh were used for evaluation. The reaction gas mixture consisted of 500 ppm NO, 500 ppm NH3 , 5% O2 , 5% H2 O and N2 as balance. The total flow of the gas mixture was 0.5 L min−1 at a gas hourly space velocity (GHSV) of 3 × 104 h−1 . 0.2 g oxidized catalysts (0.1 ml) were measured from various temperatures. The concentrations of nitrogen oxides and ammonia were measured at 120 ◦ C by a thermo Nicolet iS10 FTIR spectrometer equipped with a quartz tube (6 mm i.d.). The oxygenlean/rich catalyst samples were obtained by thermally treating the samples at 500 ◦ C in a 1%H2 (v/v)/N2 , or pure N2 , or 10% (v/v) O2 /N2 flow (500 ml min−1 ) for 30 min respectively. For kinetics study, a very high GHSV of 3 × 105 h−1 was adopted to eliminate the external diffusion effect. The samples with particles of mesh size 60–80 were used in order to rule out the internal diffusion effect. NO programmed oxidation (NO-TPO) tests were carried out using a similar method to NH3 -SCR experiment with 500 ppm NO and 5% O2 in N2 . The O2 shut-off test was carried out as follows: the oxygen supply was open when a steady SCR reaction (300–1800 s) was achieved at 250 ◦ C, and then was discontinued after 1000 s. The NH3 and NOx concentrations were analyzed by Nicolet iS10. The total flow of the gas mixture was 0.5 l min−1 at a gas hourly space velocity (GHSV) of 3 × 104 h−1 . 3. Results 3.1. Catalytic activities The NH3 -SCR activities of catalysts are shown in Fig. 1. From Fig. 1(a) and (b), above 80% NO conversions and 95% N2 selectivities are obtained by ZP/CZ-O2 catalyst in a broad temperature range of 230–450 ◦ C. N2 and H2 treatment reduces the NOx conversions and N2 selectivity at the whole testing temperature range, especieally at temperatures higher than 300 ◦ C. H2 treatment takes the most negative effect on the SCR activity of catalyst. As shown in Fig. 1(c) and (d), NO2 concentration becomes significant at 300 ◦ C and gradually increases with the increase in testing temperature. ZP/CZ-O2 catalyst presents over 97% N2 selectivity at temperatures lower than 400 ◦ C, and no N2 O is detected in this temperature range. However, NO2 and N2 O release significantly

J. Yu et al. / Catalysis Today 267 (2016) 47–55

49

Fig. 1. NH3 -SCR activities of the catalysts: (a) NOx conversions, (b) N2 selectivity, (c) NO2 concentrations and (d) N2 O concentrations.

increase when the catalysts were treated by N2 and H2 . From the following discussion in NO-TPO test (shown in Fig. 8), the NO2 release in NH3 -SCR reaction was mainly from the NO oxidation at high temperatures. The N2 O generation routes have been reported by numerous researchers, which were mainly ascribed to the deep cleavage of N H bond in NH3 [25,26] and NH4 NO3 decomposition [20,27,28]. However, the N2 O generation routes still need to be studied seriously due to the complex surface states of various catalysts. We ascribed to the NH4 NO3 decomposition because there always significantly nitrates deposit on ceria catalysts in NH3 -SCR reaction and the lean oxygen treatment will increase the N2 O selectivity of NH4 NO3 decomposition [20]. The results of NH3 oxidation was shown in Fig. 2. H2 treatment led to improved NH3 oxidation activity of catalyst which can be ascribed to the reduced acidity of catalyst (shown in Fig. 6): more NH3 adsorbed on ceria directly and be oxidized to N2 and NO. NO was the main NOx product of NH3 oxidation reaction, which became significantly at temperatures higher than 300 ◦ C. N2 O formed at temperatures higher than 450 ◦ C, suggesting that N2 O formation in NH3 -SCR reaction at temperatures lower than 450 ◦ C was from the NH4 NO3 decomposition. Ammonia oxidation on cerium sites at temperatures higher than 300 ◦ C may cause the shortage of ammonia for NOx reduction, leading to the N2 O and NO2 emission during NH3 -SCR process. In presence of ammonia, NO oxidation to NO2 on surface of catalysts may lead to the formation of ammonium nitrates (Eqs. (1) and (2)).

Fig. 2. NH3 conversion (black), NO2 (red), NO (blue) and N2 O concentrations (pink) in NH3 oxidation reactions over O2 -treated catalyst (full), H2 -treated (hollow) and N2 -treated (half-fill). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Decomposition of ammonium nitrates can result in the formation of N2 O (Eq. (3)). In our previous report [20], NH4 NO3 is proved to be an important intermediate for NH3 -SCR reaction on ZP/CZ catalyst when catalyst had high mobile surface lattice oxygen. Two distinct adsorption sites for NH3 and NOx result in a close contact between

50

J. Yu et al. / Catalysis Today 267 (2016) 47–55 Table 1 The XPS results of the samples. Sample

Ce content (%)

Ce3+ /(Ce3+ + Ce4+ ) (%)

ZP/CZ-O2 ZP/CZ-N2 ZP/CZ-H2

12 12 13

19 19 26

O/OT (%) OL /OT

OC /OT

OS /OT

60 58 56

28 29 27

12 13 15

results indicate that the redox property of catalyst is significantly reduced by non-oxygen treatment. Accordingly, the N2 O and NO2 release arises from the decomposition of nitrates and ammonium nitrates rather than ammonia oxidation. Highly reduced surface inhibits the decomposition of ammonium nitrates to N2 and H2 O with the release of NO2 and oxygen vacancy [20]. 3.3. XPS and OSC

Fig. 3. H2 -TPR profiles of catalysts.

ads-NH3 /NH4 + and ads-NO3 − /NO2 − species on the catalyst which can react with gaseous NO rapidly with the participation of active surface lattice oxygen and oxygen vacancy. From this point, the oxygen species on catalyst surface may be changed by N2 and H2 treatment, which will be consolidated by the following results. 2NO + O2 → 2NO2

(1)

2NO2 + 2NH3 + H2 O → 2NH4 NO3

(2)

NH4 NO3 → N2 O + 2H2 O

(3)

3.2. H2 -TPR H2 -TPR profiles of catalysts are shown in Fig. 3. ZP/CZ-O2 exhibits four broad reduction peaks at the temperature intervals of 240–1000 ◦ C. The low-temperature peak (411 ◦ C) is attributed to the reduction of surface oxygen related with O*–Cen+ and the high-temperature peaks (608, 664 and 822 ◦ C) are ascribed to the reduction of lattice oxygen from bulk CeO2 –ZrO2 solid solution [19]. N2 and H2 treatments significantly reduce the activity of surface oxygen as indicated by the disappearance of H2 consumption at low temperatures (<550 ◦ C). The reduction peak of lattice oxygen of bulk CeO2 –ZrO2 shifts from 664 and 822 ◦ C to 693–695 and 851 ◦ C when the catalyst was treated by non-oxygen conditions. These

Fig. 4 and Table 1 shows the XPS results of the BE of Ce 3d O 2p. The percentages of Ce3+ in Ce on the surface of catalysts were calculated by the method in Ref. [29]. The peak of Ce 3d spectra are resolved by 8 peaks. The concentration of Ce3+ in ceria can be determined from equation [Ce(III)] = Ce(III)/(Ce(III) + Ce(IV)), in which Ce(III) is defined as vI + uI and Ce(IV) is defined as v + vII + vIII + u + uII + uIII , where u and v correspond to Ce 3d3/2 and Ce 3d5/2 contributions at 882.7 and 900 eV. XPS spectra of Ce4+ 3d3/2 and Ce4+ 3d5/2 have strong characteristic satellites at 888.8 (vII ), 898.7 (vIII ), 907.6 (uII ) and 916.9 (uIII ). Ce3+ 3d3/2 and Ce3+ 3d5/2 show characteristic satellites at 885 (vI ) and 906 (uI ). The O1s XPS spectra can be deconvoluted into three distinct peaks: 1) one with lower binding energy assigned to lattice oxygen (OL ) on CeO2 , 2) chemisorbed oxygen (OC ), and 3) peak at the highest binding energy is due to surface oxygen by hydroxyl species and/or adsorbed water species (OS ) on the surface [30,31]. According to the XPS results in Fig. 4(a) and Table 1, the percentage of Ce3+ /Ce on catalysts treated by O2 , N2 and H2 , are 19%, 19% and 26% respectively, while the surface Ce content does not change obviously. N2 treatment results in the largest chemisorption oxygen on catalyst because part of lattice oxygen might release from the lattice sites to generate oxygen vacancy and chemisorbed oxygen when the catalyst were treated by N2 . However, H2 treatment generates more Ce3+ species on catalyst due to the deeply reduced catalyst surface. It can be inferred that oxygen dissociation adsorption on surface oxygen vacancies were the key step and followed by oxygen diffusion into the lattice oxygen vacancies, and then the

Fig. 4. XPS spectra of (a) Ce 3d and (b) O1s.

J. Yu et al. / Catalysis Today 267 (2016) 47–55

51

Table 2 The reaction orders with respect to NO (␣), NH3 (␤) and O2 (␥) at 250 ◦ C, 350 ◦ C and 450 ◦ C. Temp. (◦ C)

190 210 230

NO

NH3

O2

˛

Intercept

ˇ

Intercept



Intercept

0.842 0.858 0.983

−3.587 −2.409 −1.649

−0.173 −0.163 −0.150

3.573 4.028 4.532

0.217 0.253 0.325

1.329 1.470 2.151

3.4. NH3 and NOx adsorption/desorption

Fig. 5. The OSC properties of ZP/CZ-O2 , ZP/CZ-N2 and ZP/CZ-H2 at 400 ◦ C.

redox property of catalyst can be recovered. The catalyst treated by H2 resulted in too many lattice oxygen vacancies which needed too long diffusion path for the active oxygen. Therefore, the lattice oxygen of H2 -treated catalyst cannot be supplemented timely. This result will be consolidated by the OSC experiment (Fig. 5). The OSC results are shown in Fig. 5. Because the OSC data was obtained by CO + O2 reaction, the most ready available oxygen in NH3 -SCR was defined as equivalent to the OSC in the first ten seconds although this treatment may be not so accurate or in line with the actual situation. From Fig. 5, ZP/CZ-O2 and ZP/CZ-N2 presented the similar oxygen release rates (the slop = 0.025 in the first ten seconds) which were significantly higher than that of ZP/CZ-H2 catalyst (slop = 0.017). These results indicated that the oxygen release of ZP/CZ catalyst was significantly reduced by H2 treatment and cannot be recovered by O2 treated at 500 ◦ C. The activities of SCR catalysts have been proved to be controlled by the redox properties of catalysts, especially at low temperatures (<250 ◦ C) [2]. Oxygen vacancies always directly participate in the redox circles of SCR reactions [20]. Because cerium sites (Cen+ ) act as the only redox sites for the NH3 -SCR reaction on CeO2 –ZrO2 –PO4 3− catalyst. Catalyst owning facilitated transformation of Ce3+ ↔ Ce4+ will have the high NH3 -SCR activity. Therefore, it is difficult for the Ce3+ sites to participate in the NH3 -SCR reaction if they cannot be reoxidized.

Fig. 6 shows the DRIFT spectra of ammonia derived species on catalysts. From Fig. 6(a), all the spectra show the similar profiles, containing the bands of Brønsted acid sites (mainly −P OH and Zr OH) bonded NH4 + (v:3107 cm−1 ,  s : 1684 cm−1 ,  as : 1427 cm−1 ) and Lewis acid sites (mainly pyrophosphate, Zr4+ and Cen+ ) bonded NH3 (v: 3346 and 3271 cm−1 ,  as : 1601,  s :1336 and 1273 cm−1 ). Generally, areas of peaks represent the relative concentrations of Lewis and Brønsted acid sites on catalysts. It is found that H2 treatment leads to a significant decrease in acidity of catalyst, especially in the Brønsted acidity of catalyst. From Fig. 6(b), both NH3 on Lewis sites (3404 and 3276 cm−1 ) and NH4 + on Brønsted sites (1448 and 2700 cm−1 ) decreased by elevating temperature. However, the Brønsted site bonded NH4 + species were the main NH3 derived species on catalyst surface in the whole temperature region Table 2. The NH3 -TPD profiles were shown in Fig. 7. The acidity of catalysts were calculated as ZP/CZ-N2 (1996 mmol g−1 ) ≈ ZP/CZ-O2 (1985 mmol g−1 ) > ZP/CZ-H2 (1796 mmol g−1 ). It can be concluded that N2 treatment did not change the acidity of catalyst obviously, but H2 treatment led to a significant decrease in acid strength of catalyst due to the reduced Brønsted acid sites. NOx –TPD measurements were carried out, and the results are shown in Fig. 8. There is mainly NO2 desorbed from N2 /O2 -treated catalysts, while mainly NO is desorbed from the H2 -treated catalyst. This result proved that the decomposition of nitrates/nitrites is determined by the oxidation state of catalyst surface. The H2 treatment resultes in more oxygen vacancies containing extra electrons which can react with gasous NO to generate ads-NO− (Eq. (4)). AdsNO− can react with lattice oxygen (Ce4+ –O*) to generatre nitrite (Eq. (5)) and then nitrates (Eq. (6)). However, when the catalyst surface

Fig. 6. (a) DRIFT spectra of NH3 derived species on catalysts at 250 ◦ C and (b) on ZP/CZ-O2 catalyst at various temperatures.

52

J. Yu et al. / Catalysis Today 267 (2016) 47–55

Fig. 9. NO-TPO profiles of catalysts.

Fig. 7. NH3 -TPD profiles of catalysts.

4. Discussion is deeply reduced by H2 treatment, only nitrites can be formed. Therefore, only NO desorbs from ZP/CZ-H2 . 3+

Ce

− 䊐 + NO → Ce

4+

−

− NO

(4)

Ce4+ − NO− + Ce4+ − O∗ → Ce4+ − NO2 − + Ce3+ − 䊐

(5)

Ce4+ − NO2 − + Ce4+ − O∗ → Ce4+ − NO3 − + Ce3+ − 䊐

(6)

where Ce4+ –O* is an active oxygen on ceria and Ce3+ − 䊐 is an oxygen vacancy. Fig. 9 shows the NO-TPO profiles of catalysts. It is interesting that the catalysts treated by O2 , N2 and H2 , showed equally non active in NO + O2 reaction, and only obtained no less than 10% conversion at 450C. However, these three catalysts all presented remarkable NOx adsorption (shown in Fig. 5), consolidating that NOx deposited on ZP/CZ catalyst will poison the catalyst if the NOx cannot react with NH3 /NH4 + immediately. This conclusion was obtained in our previous report [20]. Compared with the NO2 generation in deNOx activity test, it can be deduced that NO2 generation in the NH3 -SCR reaction was mainly from the NO oxidation.

4.1. Kinetics Fig. 10(a) shows the effect of changing inlet NO concentration on the NOx conversion rate. At all temperatures studied, the reaction rate varies linearly with NO concentration, and the intercepts are negative. NO reaction order was close to 1 when the reaction temperatures are higher than 230 ◦ C. Both the slope and the intercept increase with increasing temperature. From a mechanistic point of view, a value of 1 indicates the reaction of NO from the gas phase. Therefore, the lower value than 1 obtained for the low-temperature range used in this work suggets that NO may participate in the reaction, at least partially, after being adsorbed on the catalyst surface (Langmuir-Hinshelwood mechanism) [32]. Alternatively, a negative intercept can also be attributed to the nitrate storage on catalyst surface. The controlling step of NH3 -SCR reaction at low temperatures is the decomposition of ammonia nitrate intermediate. The effect of variations in inlet ammonia concentration on the SCR rate is shown in Fig. 10(b). It is clear from the data that the NH3 -SCR reaction is inhibited by the adsorption of ammonia; the NOx conversion rate increases if less ammonia is fed, even at low ammonia concentrations. The degree of rate inhibition decreases as the temperature is increased, ascribting to the fast adsorption

Fig. 8. NOx -TPD over catalysts: (a) NO2 generations and (b) NO generations.

J. Yu et al. / Catalysis Today 267 (2016) 47–55

53

Fig. 10. Dependence of NO conversion rate on NO (a), NH3 (b) and O2 (c) concentration at 190–230 ◦ C.

and desorption of ammonia. Apparent reaction order of zero for ammonia was reported due to the full coverage of active centers by NH3 molecules [32]. However, in this work, the reaction order in ammonia decreases from approximately −0.173 at 190 ◦ C to −0.15 at 230 ◦ C; for comparison, Stevenson et al. [33] observed an ammonia order of −0.55 at 350 ◦ C over H-ZSM-5, while Shan et al. [31] calculated an ammonia order of zero over vanadium catalyst at 125–400 ◦ C. The slightly negative ammonia order of ZP/CZ catalyst indicates that part of adsorbed NO or O2 , which will be inhibited by NH3 , paritcipated in the SCR reaction at lower temperatures over ZP/CZ catalyst. Furthermore, linear fits to these data give positive intercepts, which indicates NO adsorbed and storaged on catalyst when no NH3 was let in. Fig. 10(c) shows how changes in oxygen concentration affect the rate of NOx conversion. Between 0.5% and 3.0% oxygen, there is an approximate linear relationship between oxygen concentration and the NOx conversion rate. Linear fits to these data give positive intercepts, which indicate lattice oxygen participates in the SCR reaction when no gasous O2 is let in. The reaction order in the oxygen concentration obtained in this work varies between 0.217 and 0.325 which are very similar with the data in reference [32]. This result suggests that the influence of O2 gas is far from being negligible. It is generally assumed that the role of oxygen is to regenerate the oxidation state of the active phase (such as Ce3+ ) via a redox mechanism or to improve the adsorption of NO [34].

4.2. O2 shut-off expriments The dynamic behavior of the NO and NH3 conversion upon changes in the O2 inlet were investigated at 250 ◦ C, and the reresults are shown in Fig. 11. From Fig. 11(a), when the O2 inlet is open, the NO outlet concentration immediately decreases due to the occurrence of the SCR reaction at 250 ◦ C. When the O2 is shut off, the NO concentration gradually reaches a new balance. Compared with our preveous results [20], lattice oxygen may participate in the SCR reaction via “ammonium-nitrate route” in which lattice oxygen reacts with adsNO to form ads-NH4 NO3 (Eqs. (4) and (6)); and then ads-NH4 NO3 reacts with gasous NO and NH3 to generate N2 and H2 O (Eq. (7)). Therefor, a delay for NO concentration to reach a new balance when O2 was shut off, was the consumption of ads-NH4 NO3 , which was discussed deeply in Ref. [20]. 2NH3 + 2NO + NH4 NO3 → 3N2 + 5H2 O

(7)

From Fig. 11(b), when the O2 inlet is open, NH3 outlet concentration gradually decreases following the NO concentration closely, indicating that ammonia storaged on catalyst as ammoniumnitrate-like species. The disparity in NO conversion over ZP/CZ-H2 catalyst compared with N2 and O2 treated catalysts, which is enlarged when oxygen is shut off, suggests that the participation of lattice oxygen in NH3 -SCR reaction can greatly improve the NOx conversion. It is noted that, in the absence of oxygen, the NH3 and

54

J. Yu et al. / Catalysis Today 267 (2016) 47–55

Fig. 11. NO concentration (a) and NH3 concentration (b) in the O2 shut-off test over catalysts at 250 ◦ C.

NO steady-state levels differs, in line with the following stoichiometry of the SCR reaction: 4NH3 + 4NO + O2 → 4N2 + 6H2 O

(8)

4NH3 + 6NO → 5N2 + 6H2 O

(9)

The fact that the NO concentration does not reach the new steady-state value immediately, can be tentatively explained by considering that either adsorbed or catalyst lattice oxygen is available for the reaction so that the SCR reaction occurs via the faster Reaction (8) instead of Reaction (9). When the catalyst was treated by H2 , catalyst surface is deeply reduced, leading to lower amount of lattice oxygen unusable in SCR reaction. Therefore, in the absence of oxygen, the activity of ZP/CZ-H2 is obviously lower than ZP/CZ-N2 and ZP/CZ-O2 . 4.3. Effects on H2 , N2 and O2 treatment on performances of catalyst Because NH3 /NH4 + and NOx species adsorbed on different sites (phosphate derived acid sites for NH3 /NH4 + adsorption and cerium sites for NOx adsorption) which were clarified by DRIFT in Ref. [20]. Nitrates were stable at low temperatures. The rate determination step of NH3 -SCR reaction at low temperatures may be the movement of NH3 /NH4 + from phosphate to ceria. The acid strengths of three catalysts (In Fig. 4, Lewis acid sites bonded NH3 located at 1273 cm−1 , Brønsted acid sites bonded NH4 + located at 1427 cm−1 ) were similar. Therefore, the different pretreatments only had negligible impact on low-temperature performance. However, the ammonia adsorption and movement on catalyst surface at high temperatures were easy to achieve. Because NH3 /NH4 + and NOx species adsorbed on different sites, it is difficult to form such unstable species like NH4 NO2 . The reduced redox property of catalyst treated by H2 resulted in reduced NH3 /NH4 + activation ability of catalyst and formation of nitrites rather than nitrates. Because nitrites were very unstable, they were inclined to selfdecomposition rather than to react with NH3 /NH4 + . Therefore, catalyst which have higher acidity and adsorb more nitrates can own high NOx conversions at high temperatures. 5. Conclusions ZP/CZ catalyst was pre-treated by O2 , N2 and H2 to study the effects of lean-oxygen treatment on adsorption and catalytic performances of catalyst. The results showed that H2 treatment led to highly reduced catalyst surface, resulting in two consequences: (1)

mainly nitrites instead of nitrates adsorbed on catalyst in NO + O2 reaction; (2) reduced Brønsted acidity of catalyst, which were responsible for the low deNOx activity of H2 reduced catalyst. N2 treatment had only negligible influence on catalyst compared to O2 -treated catalyst. Ammonium nitrate route was consolidated as possible NH3 -SCR reaction mechanism over ZP/CZ catalyst by kinetics study. Acknowledgments The authors would like to acknowledge the National Natural Science Foundation of China for financial support of Project 51202126. Moreover, we would also thank the financial support from Strategic Emerging Industry Development Funds of Shenzhen (JCYJ20120619152738634). References [1] G. Busca, M.A. Larrubia, L. Arrighi, G. Ramis, Catal. Today 107–108 (2005) 139. [2] L.J. Alemany, L. Lietti, N. Ferlazzo, P. Forzatti, G. Busca, E. Giamello, F. Bregani, J. Catal. 155 (1995) 117–130. [3] R.M. Heck, Catal. Today 53 (1999) 519. [4] A. Fritz, V. Pitchon, Appl. Catal. B 13 (1997) 1–25. [5] J.H. Kwak, R.G. Tonkyn, D.H. Kim, J. Szanyi, C.H.F. Peden, J. Catal. 275 (2010) 187–190. [6] D.W. Fickel, E. D’Addio, J.A. Lauterbach, R.F. Lobo, Appl. Catal. B 102 (2011) 441–448. [7] G. Qi, R.T. Yang, R. Chang, Appl. Catal. B 51 (2004) 93–106. [8] X. Gao, Y. Jiang, Y. Zhong, Z. Luo, K. Cen, J. Hazard. Mater. 174 (2010) 734–739. [9] (a) L. Chen, J. Li, M. Ge, R. Zhu, Catal. Today 153 (2010) 77–83; (b) S. Ding, F. Liu, X. Shi, K. Liu, Z. Lian, L. Xie, H. He, ACS Appl. Mater. Interfaces 7 (2015) 9497–9506. [10] L. Zhang, L. Li, Y. Cao, X. Yao, C. Ge, F. Gao, Y. Deng, C. Tang, L. Dong, Appl. Catal. B 165 (2015) 589–598. [11] L. Zhang, L. Li, Y. Cao, Y. Xiong, S. Wu, J. Sun, C. Tang, F. Gao, L. Dong, Catal. Sci. Technol. 5 (2015) 2251–2259. [12] K. Liu, F. Liu, L. Xie, W. Shan, H. He, Catal. Sci. Technol 5 (2015) 2290–2299. [13] Y. Liu, W. Yao, X. Cao, X. Weng, Y. Wang, H. Wang, Z. Wu, Appl. Catal. B 160 (2014) 684–691. [14] S. Yang, Y. Liao, S. Xiong, F. Qi, H. Dang, X. Xiao, J. Li, J. Phy. Chem. C 118 (2014) 21500–21508. [15] P. Maitarad, J. Han, D. Zhang, L. Shi, S. Namuangruk, T. Rungrotmongkol, J. Phy. Chem. C 118 (2014) 9612–9620. [16] X. Yao, L. Zhang, L. Li, L. Liu, Y. Cao, X. Dong, F. Gao, Y. Deng, C. Tang, Z. Chen, L. Dong, Y. Chen, Appl. Catal. B 150 (2014) 315–329. [17] Y. Shu, T. Aikebaier, X. Quan, S. Chen, H. Yu, Appl. Catal. B 150 (2014) 630–635. [18] L. Zhang, J. Pierce, V.L. Leung, D. Wang, W.S. Epling, The J. Phys. Chem. C 117 (2013) 8282–8289. [19] Z.C. Si, D. Weng, X.D. Wu, R. Ran, Z.R. Ma, Catal. Commun. 17 (2012) 146–149. [20] J. Yu, Z. Si, L. Chen, X. Wu, D. Weng, Appl. Catal. B 163 (2015) 223–232. [21] J. Yu, Z. Si, M. Zhu, X. Wu, L. Chen, D. Weng, J. Zou, RSC Adv. 5 (2015) 83594–83599. [22] L.J. Alemany, L. Lietti, N. Ferlazzo, P. Forzatti, G. Busca, E. Giamello, F. Bregani, J. Catal. 155 (1995) 117–130.

J. Yu et al. / Catalysis Today 267 (2016) 47–55 [23] Z. Wang, Z.P. Qu, X. Quan, Z. Li, H. Wang, R. Fan, Appl. Catal. B 134 (2013) 153–166. [24] Yang Zheng, Michael, P., Harold, Dan Luss Effects of CO, H2 and C3 H6 on Cu-SSZ-13 catalyzed NH3 -SCR, http://dx.doi.org/10.1016/j.cattod.2015.06. 028. [25] J. Wang, H. Cui, X. Dong, H. Zhao, Y. Wang, H. Chen, M. Yao, Y. Li, Appl. Catal. A 505 (2015) 8–15. [26] G. Qi, R.T. Yang, R. Chang, Appl. Catal. B 51 (2004) 93–106. [27] A. Grossale, I. Nova, E. Tronconi, D. Chatterjee, M. Weibel, J. Catal. 256 (2008) 312–322. [28] S. Suárez, J.A. Martín, M. Yates, P. Avila, J. Blanco, J. Catal. 229 (2005) 227–236.

55

[29] L. Xu, G. Guo, D. Uy, A.E. O’neill, W.H. Weber, M.J. Rokosz, R.W. McCabe, Appl. Catal. B 50 (2004) 113–125. [30] A.A. Battiston, J.H. Bitter, D.C. Koningsberger, J. Catal. 218 (2003) 163–177. [31] W.P. Shan, F.D. Liu, H. He, X.Y. Shi, C.B. Zhang, Appl. Catal. B 115–116 (2012) 100–106. [32] T. Valdés-Solís, G. Marbán, A.B. Fuertes, Ind. Eng. Chem. Res. 43 (2004) 2349–2355. [33] S.A. Stevenson, J.C. Vartuli, C.F. Brooks, J. Catal. 190 (2000) 228–239. [34] F. Kapteijn, L. Singoredjo, N.J.J. Dekker, J.A. Moulijn, Ind. Eng. Chem. Res. 32 (1993) 445.