Mn-modified CuO, CuFe2O4 and Î³-Fe2O3 three-phase strong

Subscriber access provided by Kaohsiung Medical University

Energy, Environmental, and Catalysis Applications

Mn-modified CuO, CuFe2O4 and #-Fe2O3 three-phase strong synergistic coexistence catalyst system for NO reduction by CO with wider active window Xiaobing Shi, Bingxian Chu, Fan Wang, Xiaoling Wei, Lixia Teng, Minguang Fan, Bin Li, Lihui Dong, and Lin Dong ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b13220 • Publication Date (Web): 29 Oct 2018 Downloaded from http://pubs.acs.org on October 29, 2018

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 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Keywords: three-phase coexisting, Mn-modified, NO reduction by CO, broad temperature window, in situ DRIFTS

1

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Mn-modified CuO, CuFe2O4 and γ-Fe2O3 three-phase strong synergistic coexistence catalyst system for NO reduction by CO with wider active window Xiaobing Shi,† Bingxian Chu,† Fan Wang,† Xiaoling Wei,† Lixia Teng,† Minguang Fan,† Bin Li,*,† Lihui Dong,*,†,‡ Lin Dong‡

†Guangxi

Key Laboratory of Petrochemical Resource Processing and Process Intensification Technology, School of Chemistry and Chemical Engineering, Guangxi University, Nanning 530004, PR. China ‡School

of the Environment, Jiangsu Key Laboratory of Vehicle Emissions Control, Center of Modern Analysis, Nanjing University, Nanjing 210093, PR China

Abstract A series of samples with precursor’s molar ratio of {KMn8O16}/{CuFe2O4}=0, 0.008, 0.010, 0.016, and 0.020 were successfully synthesized for selective catalytic reduction of NO by CO. The physicochemical properties of all the samples were studied in detail by combining the means of XPS, H2-TPR, SEM-Mapping, XRD, N2-physisorption (BET), NO + CO model reaction and in situ FT-IR techniques. The results show that three phases of γ-Fe2O3, CuFe2O4 and CuO, which have strong synergistic interaction, co-exist in this catalyst system, and different phases play a leading role in different temperature ranges. Mn species are highly dispersed in the three-phase coexisting system in the form of Mn2+, Mn3+, and Mn4+. Due to the strong interaction between Mn2+ and Fe species, a small amount of Cu2+ precipitates from CuFe2O4 and grows along CuO (110) plane which has better catalytic performance. Mn3+ can inhibit the conversion of γ-Fe2O3 to αFe2O3 at high temperature and then increases the high temperature activity. The synergistic effect between Mn4+ and the surfaces of three phases generates active oxygen species Cu2+-O-Mn4+ and Mn4+-O-Fe3+, which can be more easily reduced to some synergistic oxygen vacancies during the reaction. Furthermore, the formed synergistic oxygen vacancies can promote the dissociation of NO and are also propitious to the transfer of oxygen species. All of these factors make the appropriate manganese modified 2


Page 2 of 40

Page 3 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


three-phase coexisting system have better catalytic activity than manganese-free catalyst, making NO conversion rate reach 100% at around 250 °C and maintain to 1000 °C. Combining comprehensive analysis of various characterization results and in situ infrared, as well as XRD results in the equilibrium state, a new possible NO + CO model reaction mechanism was temporarily proposed to further understand the catalytic processes. Keywords: three-phase coexisting, Mn-modified, NO reduction by CO, broad temperature window, in situ DRIFTS

1. Introduction NOx species from fuel combustion can cause serious problems, such as lung bronchial disease, greenhouse effect, ozone depletion, acid rain, and photochemical smog. These can have an immeasurable impact on our living environment, health, and national economy. As the emission regulation of NOx becomes more and more stringent, the reduction of NOx has caused widespread concern around the world.1 Though the technology of selective catalytic reduction of NOx with NH3 has been widely used for NOx reduction, its disadvantages can not be ignored,2,3 such as air preheater blocks and NH3 leakage. The catalytic removal of NOx by CO is probably one of the most promising methods to meet the current requirements in various deNOx technologies, because the internal combustion engines using fossil fuels often produce the two gases at the same time, and this deNOx technology can achieve the purpose of eliminating two kinds of harmful gases at the same time.2 Many researchers have spent a lot of energy studying the catalytic performance and mechanism of various noble metal catalysts in the process of carbon monoxide decomposing NOx, such as Pd, Rh, and Pt supported on some carrier materials. These catalysts have high catalytic activity for the the reduction of NO by CO, but their scarcity, 3



high temperature instability, high price, and easy toxicity have greatly limited their applications. As a result, cheap alternative materials are becoming more and more concerned by researchers. In recent years, many studies have focused on the development of transition metals and rare earth complex oxides,2,3 especially cobalt, copper, nickel, and iron oxides which have been reported have NO + CO reactivity activities.3 In these transition metal oxides, iron-containing catalysts have been widely studied due to their low cost, extensive source, environmental friendliness, stability, and durability.4 Fe based materials provide promising results for various catalytic reactions. For example, Ce-Mn/TiO2 catalysts modified by iron have catalytic performance for the selective reduction of NO by CO.5 Previous literature reported that oxides containing some iron species provide catalytic activity in the NO elimination.6-8 Study9 reported that γ-Fe2O3 has better capacity to reduce NO than α-Fe2O3 due to the more acidic sites on the surface, and magnetic γ-Fe2O3 has a relatively strong magnetic adsorption capacity for NO molecules, thereby promoting chemical adsorption of NO on magnetic surfaces. It also has better magnetic susceptibility which is slightly reduced after being used, and the catalyst still has the potential for magnetic recycling and reuse,6 but the defect is that the thermal stability is worse and more or less converted to α-Fe2O3 under high temperature conditions, resulting in a decrease in catalytic activity.9 Magnetite Fe3O4 is a kind of ferromagnetic material with cubic antispine structure at room temperature, which has attracted considerable attention because of its electromagnetic properties and a reactive structure for the CO shift reaction.10 When Fe3O4 and γ-Fe2O3 coexist, they will exhibit better stability, magnetic properties and catalytic properties.11 However, one extremely important point cannot be ignored that the efficiency of mono iron oxides at low temperature (< 300 °C) is not high.11,12 Catalysts that contain CuO species have been identified as active substances for the reduction of NO by CO.12,13 but, it was found that single copper oxide still exhibited lower NO reduction activity at low temperatures.13,14 It is known that the possible double exchange behavior between Mn3+ and Mn4+ in manganese species can produce a large amount of surface active oxygen exhibiting unique catalytic redox properties, which has attracted a lot of attention in the study of NO reduced by CO model reaction.14

4


Page 4 of 40

Page 5 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In this work, we successfully used a simple new method to prepare a synergistic threephase coexistence catalyst system (γ-Fe2O3, CuFe2O4, and CuO modified by some Mn species). There is a strong synergy between these phases, and different phase species take turns to play a leading role in different temperature ranges, resulting in considerable catalytic performance for NO reduction by CO at low and high temperatures (100-1000 °C). This catalyst can effectively avoids the above disadvantages discussed above. Combining comprehensive analysis of various characterization results and in situ infrared, as well as XRD results in the equilibrium state, the reason for low temperature considerable activity, high temperature stability, and a new possible NO + CO model reaction mechanism in different temperature region were temporarily proposed.

5



2. Experimental 2.1 Preparation of catalysts CuFe2O4 magnetic nanoparticles precursor was obtained by the method similar to the typical preparation reported in the literature.15 The detailed synthesis process is described below: firstly, 2.5 mmol CuCl2·2H2O (0.426 g) and 5 mmol FeCl3·6H2O (1.351 g) were dissolved in ethylene glycol (40 mL); Subsequently, 1.0 g polyethylene glycol and 3.6 g CH3COONa were added to the mixture solution obtained above, and then the mixture was mixed at room temperature and magnetic stirring at a suitable speed for 1 h, the green mixture system will be observed; At last, a 100 mL Teflon-lined stainless-steel autoclave is used to fill the solution obtained, and then it was placed in an oven at 200 °C for 12 h. The resulting precipitated material was centrifuged and washed three times with deionized water and absolute ethanol, and then dried for 12 h at 80 °C. KMn8O16 precursor was obtained by a typical method,16 22.05 g manganese acetate tetrahydrate and 9.84 g potassium permanganate mixture was ground in an agate mill until homogeneous, and the powder was then stored in a beaker and heated to 80 °C for 4 h. After washing with deionized water to remove unreacted material, a black powder was obtained which was dried overnight at 80 °C. The two precursors were chosen according to different molar ratio, and they were grinded evenly in agate mortar, and then put in a muffle furnace and calcined for 10 h at 400 °C.15 A series of samples with precursor’s molar ratio of {KMn8O16}/{CuFe2O4}=0, 0.008, 0.010, 0.016, and 0.020 were labeled as 0MnFeCu, 0.008MnFeCu, 0.010MnFeCu, 0.016MnFeCu, 0.020MnFeCu, respectively. 2.2. Catalysts characterization The X'Pert PRO diffractometer (PANalytical, The Netherlands) using Cu/Ka radiation (λ=1.54060 Å) was used to obtain the Power X-ray diffraction (XRD) patterns of the samples. The scan voltage and current were set to 40 kV and 40 mA, the scan rate was 8 o·min-1,

and the scan range 2θ was 10 to 80o.

6


Page 6 of 40

Page 7 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The adsorption-desorption isotherm of N2 at 77 K was obtained using a Micrometrics TriStar I 3020 analyzer. The pore distribution and specific surface area were calculated from the nitrogen sorption isotherms by the Barrett-Joyner-Halenda (BJH) and BrunauerEmmett-Teller (BET) methods, respectively. HITACHI S-3400N electron microscope (Hitachi Company of Japan) was used to measure the scanning electron microscope (SEM) images of the samples at 20 kV. Before the FESEM test, the samples were suspended in ethanol, dispersed by ultrasonic, and dripped onto aluminum sheet. Tecnai G2 F20 S-TWIN instrument (FEI Company America) with 200 kV acceleration voltage was used to obtain the transmission electron microscopy (TEM) images. The X-ray photoelectron spectra (XPS) of the samples were obtained by using ESCALAB 250Xi multifunctional imaging electron spectrometer (Thermo Fisher Company America) which used monochromatic Al Kα radiation (hν = 1486.6 eV) at the power level of 150 W. The electronic binding energy is calibrated on the basis of C1s (284.8 eV). The irradiated area and detection depth of the samples were 2 mm × 1 mm and 2-5 nm, respectively. The H2-TPR profiles of the samples were obtained by using FINESORB-3010 automatic chemical adsorption apparatus (Finetec Corporation). 20 mg sample was heated from room temperature to 110 °C (which was kept under these conditions for 1 h before analysis) under a flow rate of 50 mL·min−1 of N2, and then was switched to 10 mL·min−1 H2-Ar flow (7% H2 by volume) for 30 min after being cooled to room temperature in a N2 atmosphere. After that, the temperature was raised from room temperature to 900 °C (10 °C·min−1) and the thermal conductivity detector (TCD) was used to continuously analyze the H2 consumption. The Nicolet iS50 Fourier transform infrared spectrometer equipped with mercury cadmium telluride and cooled with liquid nitrogen, with a spectral resolution of 4 cm-1 (number of scans = 32), was used to obtain the in-situ diffuse reflectance infrared Fourier transform spectrum (DRIFTS) from 650 to 4000 cm-1. The catalyst powder was placed in a sample cell and pretreated with purified N2 for 1 h at 400 °C to eliminate adsorbed 7



impurities on the surface, and the catalyst background spectra at different target temperatures were collected during cooling to room temperature. Subsequently, the catalyst was respectively exposed to 10 mL·min−1 NO-He (5 vol% NO and 95 vol% He) or 10 mL·min−1 CO-He (10 vol% CO and 90 vol% He) or 10 mL·min−1 NO-He (5 vol% NO and 95 vol% He) + 10 mL·min−1 CO-He (10 vol% CO and 90 vol% He) gaseous flow until saturation. The FT-IR of CO, CO + NO and NO at each target temperature were collected at the heating rate of 5 °C·min−1 from 50 to 400 °C. Finally, subtracting the corresponding background reference spectra at various target temperatures to get the results. 2.3 Catalytic activity measurements The performance of the catalyst for NO reduced by CO was obtained under simulated reaction conditions. The reaction conditions include feed steam composed of He-NO (95 vol% He and 5 vol% NO) and He-CO (90 vol% He and 10 vol% CO), 10% H2O (only when used) and 1000 ppm SO2 (only when used), and make the flow rate of the mixed gas flowing through the sample constant at 150000 mL·g-1·h-1 (GHSV). Each sieved sample(15 mg, 40-60 mesh) was packed in quartz tube and pretreated at 110 °C in high purity N2 flow for 1 h to remove impurities. The reaction gas mixture is switched after the sample is cooled to room temperature. Two chromatographic columns (diameter = 3 mm, length = 1.75 m) and two thermal conductivity detectors (T = 100 °C) were used to analyze the catalytic reaction products. Data collection begins when the outlet concentration of the gas reaches a steady state. The NO conversion and N2 selectivity are calculated as follows:12 NO coversion 

N2

selectivity

[ NO ]in  [ NO ]out 100% [ NO ]in

(

2 * [ N 2 ]out ) 100% [ NO ]in  [ NO ] out

Convertion of NO to N 2  NO cov ersion  N 2 selectivity

8


Page 8 of 40

Page 9 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


3. Results and discussion 3.1. Catalytic performance

9



Fig. 1. The results of (a) NO conversion (%); (b) N2 selectivity (%); (c) conversion

Page 10 of 40

of

NO to N2 (%) over various catalysts as a function of reaction temperatures; (d) The NO conversion or CO to CO2 conversion of sample 0.010MnFeCu. Reaction conditions: 5% NO, 10% CO, and balanced He; SV = 150,000 mL·g−1·h−1. (e) conversion of NO to N2 (%) over catalyst 0.010MnFeCu at different GHSV= (30000, 150,000, 300000) mL·g−1·h−1; The NO conversion, N2 selectivity and the conversion of NO to N2 over a series of catalysts and the CO conversion on catalyst 0.010MnFeCu as a function of temperature are shown in Fig. 1. From Fig. 1(a), it can be seen that the NO conversion on all catalysts is almost the same at temperatures below 150 °C, and it is relatively low. When the temperature rises from 150 to 200 °C, the catalytic conversion of the more Mn-containing species is significantly higher than these of the 0MnFeCu, 0.008MnFeCu, indicating that the incorporation of the appropriate amount of Mn species further enhances the performance of these catalysts. As the temperature rises from 200 to 230 °C, the NO conversion increases dramatically, which may be because the surface of catalyst reaches the best stable catalytic state, and each reactant reacts quickly on the surface of catalyst. When the temperature continues to increase to 350 °C, the conversion of NO gradually increases to 100%. When the temperature eventually rises to 1000 °C, the NO conversion rate remains constant at 100%. This shows that the catalyst not only has a good NO conversion at low temperatures but also exhibits excellent stability at high temperatures. Fig. 1(b) shows the selectivity of N2 of all the samples. It was found that all catalysts show the similar curve, that N2 selectivity decreases sharply to a minimum with the temperature increasing from 150 to 200 °C, and then quickly rises again to 100% around 300 °C. However, 0.010MnFeCu catalyst has higher selectivity in the whole temperature region, which suggests that the appropriate amount of Mn manganese incorporation can further improve the selectivity of the products.17 From Fig. 1(c), it can be seen that when the temperature rises from 100 to 200 °C, the conversions of NO to N2 of 0MnFeCu, 0.008MnFeCu and 0.020MnFeCu catalysts keep at a lower value. However, those of 0.010MnFeCu and 0.016MnFeCu samples were much higher, which indicated that the appropriate amount of Mn manganese incorporation can improve the catalytic 10


Page 11 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


performance. As can be seen from Fig. 1(d), the conversion rate of NO and CO at 200220, 220-230 °C increased drastically compared to those at other temperature ranges, which may be due to the rapid formation of some active catalytic species on the catalyst surface. In addition, it can also be observed that the curve of the conversion of CO and NO are approximately the same, but their values rise to 50% and 100% respectively at 300 °C, which indicates that NO and CO reach a dynamic equilibrium on the catalyst. From Fig. 1(e), it can be known that when the space velocity is 30000 mL·g−1·h−1, the N2 generation rate reaches 100% at 230 °C and stabilizes to 1000 °C. Although the 100% NO conversion temperature increases with the increase of the space velocity, due to the high temperature activity stability of the catalyst, NO can be completely eliminated as long as the temperature is sufficiently high. In order to further explore the reasons for the differences in catalytic performance, we performed a series of characterizations of these samples and presented the corresponding results in the following sections. The effects of SO2, SO2 +H2O, and H2O on the activity of 0.010MnFeCu catalyst at 320 °C were also studied (shown in Fig. S1).The results show that the catalyst has a certain water and sulfur resistance

3.2. Texture characterization (N2-physisorption) The corresponding pore size distribution of these catalysts shows a narrow peak centered around 16-30 nm, which is consistent well with the pore distribution region of mesopore (Fig. S2), indicating that the samples have uniform mesoporous size distribution. The summarizing of the BET surface area, pore size and pore volume of these catalysts are shown in Table S1. It can be seen from Table S1 that the surface area and pore volume of the samples have no obvious change, which may indicate that the influence of the deposition of trace Mn oxides will be insignificant to the structure of all samples. Fig. S3 shows that the average size of the particles are about 10-30 nm. 3.3. XPS results

11



Fig. 2. (a) XPS results of O1s of 0MnFeCu, 0.010MnFeCu, 0.020MnFeCu and KMn8O16 catalysts; (b) XPS results of Cu2p of 0MnFeCu, 0.010MnFeCu and 0.020MnFeCu catalysts; (c) XPS results of Fe2p of 0MnFeCu, 0.010MnFeCu and 0.020MnFeCu catalysts; (d) XPS analysis of Mn2p in 0.010MnFeCu, 0.020MnFeCu catalyst and KMn8O16. The XPS spectra for O1s of all the samples are also shown in Fig. 2(a), three individual peaks of all samples were distinguished. The peak observed around 529.5 eV corresponds to the lattice oxygen O2- (denoted as Oβ). The second peak having higher binding energy with the position around 531.1 eV (denoted as Oα) is assigned to chemical adsorbed oxygen (O-, O22-), the binding energy occurring at 533.1 eV (denoted as Oγ) is contributed to the physical adsorbed oxygen (H2O, OH-).18,19 Surface adsorbed oxygen Oα is generally considered to be more reactive in oxidation because it has a higher mobility 12


Page 12 of 40

Page 13 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


than lattice oxygen Oβ,20 it has been widely reported that the surface chemical oxygen has a positive effect on the catalytic effect of NO and CO reaction, and this mechanism may be attributed to not only CO can be oxidized, but also the adsorption of NO on the active site can be promoted.19,21 The higher Oα/O(α+β+γ) value means that there is more active oxygen on the surface of the catalyst.22,23 0.010MnFeCu sample has the highest content (0.46), which is in line with its better performance, and the other three samples are 0.07, 0.40, 0.32, respectively. According to the XPS spectrum, when proper trace manganese oxides were added to catalysts, chemical oxygen increased and then the redox performance was promoted, which can be consistent well with the results of the activities. In Fig. 2(b), the two main peaks at 933.6-933.9 and 952.7-953.7 eV are classified as Cu2p3/2 and Cu2p1/2, respectively.20 One can clearly find that the peaks of Cu2p3/2 and Cu2p1/2 have obvious vibration at the binding energies about 944 and 963 eV assigned to shakeup satellites, which are similar to the previous literatur,12,24-26 and the position and shape trend of the peak are also exactly the same as pure CuO. So, all above analysis shows that Cu species in all of our catalysts exist in the form of Cu2+. All Fe 2p spectra in Fig. 2(c) show two main peaks with binding energies of 710.7711.1 and 724.2-724.3 eV, which belong to Fe2p3/2 and Fe2p1/2, respectively. One can clearly find that two accompanying satellite peaks at the binding energies about 719.3719.2 and 732.7-732.8 eV are the characteristic of Fe3+ cations, which is quite similar to these report27-29 which showed Fe2p XPS spectra of γ-Fe2O3, Fe3O4, α-Fe2O3, and the shape trend and position of the peak is also exactly the same as pure Fe2O3. Therefore, iron species exist in the state of Fe3+ over the sample, and the valence state of surface iron can not be changed by doping trace manganese, which is highly consistent with the following XRD results. However, the binding energy of Fe2p in 0.010MnFeCu catalyst was a little higher than that of others, which may indicate that the higher interaction between Fe ions and neighbor high-valence Mn4+ species make Fe species in a more positively charged environment.22,30 As shown in Fig. 2(d), one can find that two distinct peaks centering at 641.5 and 653.5 eV can be clearly observed, which are attributed to Mn2p3/2 and Mn2p1/2, respectively. By performing peak fitting deconvolution, Mn 2p3/2 signal can be decomposed into three components at binding energy of 640.0, 642.2, and 643.5 eV with 13



three corresponding satellite peaks at 648.0, 653.3, 655.2 eV, respectively, which can be assigned to Mn2+, Mn3+ and Mn4+ species on the surface of 0.010MnFeCu catalyst, and similar to the previous report.19,21,31 As can be seen from the Fig. 2(d), both samples 0.010MnFeCu and 0.020MnFeCu contain newly formed Mn2+ species compared to KMn8O16, and 0.010MnFeCu has more Mn4+ and Mn2+. This may be because part of Mn2+ is generated due to

the interaction among iron, copper, and Mn species, in the process of being calcined. It is precisely because there are more Mn4+ in 0.010MnFeCu (Mn4+/(Mn4++Mn3++Mn2+) = 0.45) that the surrounding copper or iron are in a more positively charged environment, which is consistent with the shifting of their binding energy to high binding energy.

3.4. XRD and HRTEM results

14


Page 14 of 40

Page 15 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 3. (a) Powder XRD patterns of all samples: 0MnFeCu, 0.008MnFeCu, 0.010MnFeCu, 0.016MnFeCu, 0.020MnFeCu; (b) Partial enlargement XRD result; (c) The HRTEM images of 0.010MnFeCu sample. In order to further identify the detail crystallinity and the crystalline phase of the synthesized fresh samples, the XRD analysis was performed and the results of all the catalysts are showed Fig. 3(a). One can find that all of the samples show the similar peaks. Peaks appearing at 2θ = 18.5, 30.2, 35.6, 37.2, 43.0, 57.1, 62.8, 74.5, and 79.5o, matching well with the peaks for CuFe2O4 (JCPDS#25-0283), which are ascribed to (220), (311), (400), (107), (422), (511), (440), and (533) lattice planes of the cubic spinel structure, respectively. The small peaks can be clearly observed at 2θ = 32.5, 35.4, 35.5, 38.7, 38.9, 48.7, 58.3, 61.5, 66.2, and 68.1o, corresponding to the reflection of (110), (002), (11-1), (111), (200), (20-2), (11-3), (31-1), and (220) of CuO (JCPDS#48-1548), which may be derived from the decomposition CuFe2O4 during the thermal treatment process.32 Previous study7 shows that γ-Fe2O3 is an anti-spinel structure with all iron existing in the Fe3+ form, and it exists in tetrahedrally coordinated Fe3+ and octahedral Fe3+ structures and contains cationic vacancies. Although γ-Fe2O3, CuFe2O4 and Fe3O4 have a facecentered cubic structure and belong to a cubic system,28 the crystal structure is basically the same, but the difference is that γ-Fe2O3 has cationic vacancies,10 so, its XRD results are slightly different from those of the latter two. The difference between γ-Fe2O3 and CuFe2O4 can be clearly observed through partial enlarged drawing in Fig. 3(b). The results show that γ-Fe2O3 and CuFe2O4 coexist in our present samples. Meanwhile, it is interesting to find that the intensity of the characteristic peak for (110) of monoclinic CuO in 0.010MnFeCu sample is much higher than that of others. According to previous literatures,33,34 it is reported that Mn species is easily combined with Fe species and then weakens the interaction between Cu and Fe, leading the precipitation of CuO species, which are highly consistent with the results. Combining the analysis in Fig. S4 (seperate XRD result of 0.010MnFeCu sample ), it can be suggested that some Mn2+ promotes the precipitation of Cu2+ from CuFe2O4 and promotes CuO to grow along the CuO (110) (as showed in Fig. 3(c)) crystal plane during the heat treatment.35,36 With the increase of the manganese loading, some manganese substances 15



gather themselves to form tiny crystals or amorphous substances, which can not been observed from XRD results. So, it can be seen that the strength of CuO (110) gradually weakens. As reported previously,37 the catalytic activity of CuO nanoparticles with exposed active (110) facet for CO oxidation was significantly higher than that of the original nano-flower. It was also reported that the molecule NO preferentially binds to the oxygen atom pointing upward to the (110) plane.37,38 Based on the above XRD results, it can be concluded that three types of crystalline substances coexist in the present catalysts ( which can be further proved by Fig. 3(c) ), and their cations existing states are also highly consistent with XPS results discussed above. The 0.010MnFeCu sample has relatively more CuO (110) crystal plane (as showed in Fig. 3(c)), and these CuO species are closely related to the surrounding γ-Fe2O3 and CuFe2O4, which may produce a lot of defects accompanied by a large number of active oxygen species (O1s XPS results of 0.010MnFeCu sample showed in Fig. 2(a), and these active oxygen species can be easily reduced to oxygen vacancies) at the two phase contact surface, so it has higher activity. 3.5. SEM-Mapping results

16


Page 16 of 40

Page 17 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 4. SEM-Mapping of various elements in 0.010MnFeCu sample. In order to study the detailed distribution and morphology of various elements on the surface of the catalyst, SEM mapping analysis was used. Fig. 4 shows elemental maps of Fe, O, Mn, and Cu on the surface of 0.010MnFeCu sample. It can be clearly seen that Mn, Fe, O elements are highly dispersed on the surface of the catalyst, but Cu element is relatively concentrated in a part of the area, and in other parts is consistent with the distribution of other elements. From the above observations, it is tentatively believed that the Mn element is highly dispersed on the surface of the three phases, possibly in the form of microcrystals or indeterminate forms, which is consistent with the XRD analysis results. Some aggregated Cu element comes from CuO. In the regions where only iron and oxygen are distributed, these Fe species are derived from γ-Fe2O3 nanocrystals. In the region where Cu and Fe coexist, two elements exist in CuFe2O4 nanoparticles as Cu-O-Fe. These CuO species are closely surrounded by CuFe2O4 (can also be proved by the HRTEM images of 0.010MnFeCu sample showed in Fig. 3(c)). It can be seen from the figure that Mn element is highly uniformly dispersed, which indicates that some of the manganese species are doped on the surface of all phase. These manganese species may interact strongly with the surface cations, resulting in a large number of defect and a large number of active oxygen species (Cu2+-O-Mn4+, Cu2+-O-Mn3+, Mn4+-O-Fe3+ oxygen species be more easily reduced to Cu+-□-Mn3+, Cu+-□-Mn2+, Mn3+-□-Fe2+).17 The above SEM-Mapping result is well consistent with the XRD and XPS results, further proves the coexistence of three phases (γ-Fe2O3, CuFe2O4, CuO three-phase is close and not independent), which are difficult distinguished only by the XRD results. 3.6. TPR results and corresponding XRD results 17



Fig. 5. (a) H2-TPR profiles of these catalysts and the overlapped peaks of all catalysts fitted by Gaussian–Lorentzian curves; (b) XRD results for 0.010MnFeCu sample fully reduced by H2 at different temperatures; (c) Partial enlargement of (b). The redox performance is the key factor to influence the catalytic activity of NO reduction by CO. Therefore, the reductivity of all catalysts was evaluated by H2-TPR experiment, and the results are shown in Fig. 5. According to the literature,39 the first peak was attributed to the reduction of CuFe2O4 to Cu and Fe3O4, and the second one is ascribed to the reduction of CuO to Cu0 (as showed in Fig. 5(a)). However, studies40,41 reported that Cu-Fe spinel composite catalysts have similar compositions and TPRprofiles, which attributed the three reduction peaks from low temperature to high 18


Page 18 of 40

Page 19 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


temperature to the highly dispersed CuO to Cu, CuFe2O4 to Cu and Fe3O4, iron oxide to FeO or Fe, respectively. In order to understand the corresponding reduction peaks of different substances in these catalysts, here we have performed a sufficiently long time H2 reduction treatment of 0.010MnFeCu sample at different temperatures (showed in Fig. 5(b)), so that the reduction reaction reaches a steady state. Their XRD analysis results are shown in Fig. 5(b), it can been clearly seen that the corresponding peaks of CuO disappeared completely at first (reduced at 160 °C). At the same time, the strong metal Cu peak appeared, probably because the larger particles of CuO are completely reduced or reduced to smaller particles. Here, the H2-TPR peak can be attributed to the reduction of CuO to Cu. Subsequently, as the reduction temperature continues to increase (from 180 to 300 °C), the characteristic peak of γ-Fe2O3 gradually becomes smaller and eventually disappears, accompanying with the appearance of the Fe3O4 diffraction peak (Fig. 5(c)), because γFe2O3 is reduced to Fe3O4. As the temperature continuing rising (from 300 to 380 °C), the diffraction peak of CuFe2O4 decreases until disappears. Instead, the peak of Fe3O4 becomes a sharper peak (Fig. 5(c)), indicating that the two species (γ-Fe2O3, CuFe2O4) are reduced to form larger Fe3O4 particles. When the temperature reaches 800 °C, all the diffraction peaks disappear except for Cu and Fe metals. This is because that all substances are reduced to Cu0 and Fe0 at the high temperature. It is similar to previous literature,39,41 it can be proposed that the highly dispersed CuO is reduced firstly; and then, γ-Fe2O3 is reduced to Fe3O4; subsequently, CuFe2O4 is reduced to Fe3O4 and Cu; finally, Fe3O4 is reduced to FeO or Fe. The reduction peaks associated with manganese cannot be observed here, probably because Mn levels are too low to be observed. Based on the above facts, the overlapped peaks below 350 °C were tentatively fitted into three peaks by Gaussian–Lorentzian curves, as showed in Fig. 5(a). Three peaks located around 230, 270, and 300 °C, correspond to the reduction of high dispersed CuO to Cu, γ-Fe2O3 to Fe3O4, and CuFe2O4 to Fe3O4 and Cu, respectively, which could be confirmed by XRD of the catalysts gradually reduced by hydrogen. Their relative ration can be obtained by comparing the peak area of H2-TPR profiles, as showed in Table 1. 19



Page 20 of 40

From the data in the table, it can be seen that: (1) the peak area corresponding to the reduction of CuO is approximately 3 times that of γ-Fe2O3. This is due to the fact that the amount of Fe in the precursor is twice that of Cu, and there are three species (γ-Fe2O3, CuFe2O4, CuO) in the catalyst. Theoretically, the ratio of the amount of CuO and γ-Fe2O3 substances is 1:1, and the reduction of 1 mol CuO to Cu needs 1 mol H2, while the reduction of 1 mol γ-Fe2O3 to Fe3O4 will consume 1/3 mol H2. So, the area of the former is approximately three times of the latter, which further proves that our catalyst contains these three species. (2) the value of CuO/CuFe2O4 is the highest in 0.010MnFeCu sample, indicating that a proper proportion of the trace amount of manganese species loading promotes Cu2+ precipitation from CuFe2O4 to form CuO and grow along the (110) plane, which is consistent well with the XRD results in Fig. 3(a) and the HRTEM images showed in Fig. 3(c). (3) all the samples show the similar TPR profile (a small amount of manganese has a little effect on the relative content of each phase), except for 0.010MnFeCu shifting to a lower temperature, which is consistent well with its better catalytic performance. This may be because when an appropriate amount of manganese is incorporated, some of the newly added reactive oxygen species, such as Cu2+-O-Mn4+, Cu2+-O-Mn3+ and Mn4+-O-Fe3+, are formed on the surfaces of the phases, so that the catalyst is more easily to be reduced, make the reduction peak move to low temperature. Table 1 H2 consumption of each phase in various catalysts and the proportion of consumption. Sample/MnFeCu

0

0.008

0.010

0.016

0.020

3430

4255

4141

4434

3595

1167

1433

1323

1549

1200

9438

10221

8574

9707

8465

H2 (CuO)/H2 (CuFe2O4)

0.36

0.41

0.48

0.45

0.42

H2 (CuO)/H2 (γ-Fe2O3)

2.94

2.97

3.13

2.86

2.99

H2 consumption (CuO) (CuO → Cu0) H2 consumption (γ-Fe2O3) (γ-Fe2O3 → Fe3O4 ) H2 consumption (CuFe2O4) (CuFe2O4 → Cu0 + Fe3O4)

20


Page 21 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 6. H2 -TPR profiles of pure CuO, γ-Fe2O3 and 0.010MnFeCu samples. It has been widely accepted that the synergistic interactions among CuFe2O4, CuO, and γ-Fe2O3 can be appropriately reflected by their reduction behaviors. As reference, two pure samples of CuO and γ-Fe2O3 were also prepared (Fig. 6). Here, it can be seen that pure CuO has a strong peak located around 350 °C corresponding to the reduction of CuO to Cu, which is similar to report.42 One can clearly find that peaks ascribed to the reduction of CuO and γ-Fe2O3, occurring at 240, 270 °C in our samples, shift to lower temperature compared with those of the pure references, which shows strong interaction and synergistic effect between the highly dispersed oxide particles.43 Previous researcher reported2,44 that the oxygen species in Cu2+-O-Mn4+, Cu2+-O-Mn3+, Mn4+-O-Fe3+, Cu2+O-Fe3+ can be more easily reduced than Cu2+-O-Cu2+, Fe3+-O-Fe3+. So, we propose that there is a strong interaction among the three phases, and Cu2+-O-Fe3+ can be generated, as well as the incorporation of appropriate amount of Mn manganese can produce another active oxygen (Cu2+-O-Mn4+, Cu2+-O-Mn3+, Mn4+-O-Fe3+, Mn3+-O-Fe3+) on the surface of the different phases, which leads the reduction peaks shift to lower temperature, and further results in the excellent catalytic performance. This further proves that there is a 21



strong interaction among CuO, CuFe2O4 and γ-Fe2O3 three-phase coexistence catalyst system.

Through the above analysis (XRD, XPS, TPR, SEM-Mapping, HRTEM), it can be proposed that γ-Fe2O3, CuO and CuFe2O4 three phases coexist in the catalyst, and strong interaction and synergistic effect can generate among of them (Fig. 6), which is line with the good performance of the 0MnFeCu catalyst. Some loaded Mn species which are highly dispersed on the surface or between the contact interface of each phase can further interact with the cations of the surface of the three-phase coexisting catalyst, so as to produce more defects and eventually produce a large number of active oxygen that can be easily reduced to oxygen vacancies (confirmed by XPS and TPR results). As a result, 0.010MnFeCu catalyst has better catalytic activity than the 0MnFeCu catalyst. 3.7. CO or/and NO interaction with the catalysts (in situ FT-IR) and Possible low temperature reaction mechanism

Fig. 7. In situ FT-IR results of CO (10% in volume) adsorption on 0.010MnFeCu and 0MnFeCu catalysts.

22


Page 22 of 40

Page 23 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


To understand the adsorption and reduction performance of these catalysts, in situ infrared spectroscopy (FT-IR) tests of CO adsorption were performed at a series of temperatures, the corresponding results of 0.010MnFeCu and 0MnFeCu samples are given in Fig. 7. One can clearly see that the two samples had identical peaks at 2115 and 2173 cm1.

Previous report18 assigned them to gas-phase CO vibration, but they were also

attributed to the peaks of adsorbed CO.4 In order to clarify their assignment under the current conditions, we purposely conducted a blank control experiment showed in Fig. S5(b) (these two peaks have the identical intensity at the whole temperature ranges), the two peaks should be gas-phase CO. An interesting phenomenon can be observed, peak around 2105 cm-1 (which can be clearly distinguished from the peaks of the gas phase CO showed in Fig. S5(b)), appearing around 100 °C, gradually becoming stronger as the temperature rises,13 then gradually weakening until disappearing, was assigned to Cu+CO.45,46 Literature47 reported that CO was preferentially adsorbed Cu+, and Cu+-CO was very stable, so Cu+ species can be probed by CO. However, this peak from 0.010MnFeCu sample has disappeared completely at 200 °C, which is lower than that of 0MnFeCu sample. This may be because when a small amount of Mn species are incorporated, Cu2+-O-Mn4+, Cu2+-O-Mn3+, Mn4+-O-Fe3+ and other active oxygen species are generated at the contact interface of each phase,2,18 so that these active oxygen species can be more easily reduced by CO, and eventually Cu species being more easily reduced to Cu+ (Cu+-□-Mn3+, Cu+-□Mn2+, Mn3+-□-Fe2+), then to Cu0 (consist well with the TPR results), which is in line with better catalytic performance. For both samples, peaks observed at about 2341 and 2360 cm-1 correspond to gas phase CO2,48 of which the intensity increases rapidly at 150 °C. Because at 150 °C, there is a large amount of Cu+-CO which can rapidly combines with adjacent oxygen to form CO2.13,48

23



Page 24 of 40

Fig. 8. In situ DRIFTS results of NO (5% in volume) adsorption on 0MnFeCu and 0.010MnFeCu catalysts. In order to obtain an insight into the interaction between adsorbed NO and the surface components of the samples, the NO adsorption FT-IR spectra has been

carried out at

various temperatures from 25 to 400 °C on 0MnFeCu and 0.010MnFeCu catalysts. As shown in Fig. 8, for 0.010MnFeCu sample, peaks observed at 1004 and 1628 cm-1 are owned to the stretching mode of N=O from bridging bidentate nitrate, gradually decreasing with the increase of temperature and being lost at high temperature due to their poor stability.2,49 Bands appearing at 1205 and 1585 cm-1 correspond to symmetric and asymmetric vibration modes of chelating bidentate nitrate,2,49 increasing first and then decreasing slowly with raising the temperature up to 400 °C, in line with its relatively good stability compared with others. The peak at 1242 cm-1 is assigned to the vibration of chelating bidentate nitrite, having been reported by other researchers,13,50 which has better thermal stability than that of others. It is worth noticing that the peak located at 1277 cm-1 is larger than others at room temperature and decreases rapidly with increasing temperature and disappears completely below 200 °C due to its desorption or decomposition or transformation. The peak is assigned to the vibration of the linear nitrite and has the relatively poor stability.48,50 Bands appearing at 1551 and 1603 cm-1 correspond to monodentate nitrate and bridging monodentate nitrate, respectively.2 As the temperature raising, the first one decreases gradually and absolutely vanishes below 350 °C, the second shows a tendency to decrease in intensity but it never disappears even at 400 °C.2 24


Page 25 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The bands at 1844 and 1903 cm–1 are noticed for both samples, their intensity decreases gradually with temperature increase but never vanishes. The literatures4,51 have assigned the two peaks to the weakly adsorbed NO or gas phase NO, in order to get true information at our present condition, A blank control FT-IR test was tentatively conducted (without samples, the other conditions are exactly the same, shown in Fig S5(a)), from which we can accurately attribute them to the gas phase NO through the experimental results. It is interesting to find that 0MnFeCu sample has the similar vibration bands at 1903, 1844, 1630, 1608, 1587, 1545, 1279, 1249, 1211, 1006 cm–1 corresponding to gas phase NO (1903 and 1844 cm–1), bridging monodentate nitrates (1608 cm-1), linear nitrites (1279 cm–1), chelating bidentate nitrite (1249 cm-1), chelating bidentate nitrate (1211 and 1587 cm-1), monodentate nitrate (1545 cm-1), bridging bidentate nitrate (1630 and 1006 cm–1), respectively. In addition, compared with 0MnFeCu, it can be obviously noticed that all of the vibration bands of adsorbed NO species on the surface of 0.010MnFeCu catalyst except for two peaks at 1844 and 1903 cm –1 (both of the catalysts were surrounded by the same gas atmosphere) shift towards the lower wavenumbers. This phenomenon may be interpreted by the more strongly synergistic interaction among CuO, CuFe2O4 and γFe2O3 interfaces, which are promoted by some Mn species highly dispersed between these contact surfaces. So, it can cause the d-electron' back-donation from the metal cation to the antibonding orbital of NO, and then the N–O bond was weakened, which maybe account for its better performance.48 Comparing the change regulation of intensity of NOx adsorbed on the two samples, some interesting results can be obtained, the weakening degree of the peak intensity of nitrate and nitrite over 0.010MnFeCu is greater than that of 0MnFeCu, indicating that the decomposition, adsorption, desorption abilities of 0.010MnFeCu are better than those of 0MnFeCu catalyst.52 From the FT-IR spectra results discussed above, it can be concluded that: when the catalyst is placed in an atmosphere of NO, the oxygen species on its surface combine with NO and oxidize it to NOx species, this shows that the present catalyst has strong oxygen storage and release characteristics;50 As the temperature rises, some unstable NO begins to desorb, decompose or transform; When the temperature reaches 400 °C, most 25



of the NOx species has been desorbed or decomposed, except for stable nitrates;50 When a proper amount of Mn substance is added to the contact interface among CuO, γ-Fe2O3 and the surrounding CuFe2O4, some defects may occur due to strong interactions, so that more cations are in an unsaturated state (can adsorb a great of oxygen species, supported by XPS results), and thus have a stronger adsorbed ability on NO, and eventually make all infrared absorption peaks of NOx move toward lower wavenumbers.52

Fig. 9. In situ FT-IR results of NO and CO (5% NO and 10% CO in volume, respectively) co-adsorption on 0.010MnFeCu and 0MnFeCu catalysts. In order to better understand the mechanism of NO reduction by CO, the properties and relative quantities of surface species over 0.010MnFeCu and 0MnFeCu samples were studied by CO and NO co-adsorption in situ FT-IR technique under simulated reaction conditions, as shown in Fig. 9. One can clearly find that the interaction of CO and NO with samples produces several kinds of nitrates and nitrites species, which is similar to those results of NO adsorbed alone on the surface at room temperature. NO was preferentially adsorbed on the surface of the samples due to unpaired electrons, and then the adsorption of CO species on the surface of sample was inhibited, which is in line with the previous studies.50 However, some differences from those of the NO alone adsorption can be found, the intensities of all the NOx species decrease more quickly and they vanish completely at lower temperature, which demonstrate that some nitro-intermediate species

26


Page 26 of 40

Page 27 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


has reacted with CO. Peaks observed at 2341 and 2360 cm-1 correspond to gas phase CO2,48 which shows the oxidation of CO even at relatively low temperature. Here, a particularly different phenomenon should be noticed that when the temperature rises to 175 °C, all NOx species disappear completely, and two special peaks located about 1770-1790 and 2105 cm-1 begin to appear (the intensity of the

peak at 2105 cm-1

over 0.010MnFeCu is much higher than that of 0MnFeCu. This may be because when an appropriate amount of manganese is incorporated, some active oxygen species such as Cu2+-O-Mn4+, Cu2+-O-Mn3+, and Mn4+-O-Fe3+ (Fig. 2(d) showed that sample 0.010MnFeCu has more Mn4+ species and therefore can produce more Cu2+-O-Mn4+, Mn4+-O-Fe3+) can generate and then can be rapidly reduced, and some Cu+-□-Mn3+, Cu+-□-Mn2+, and Mn3+□-Fe2+ are immediately generated, and then immediately produced a large amount of Cu+-CO). As the temperature rises, their intensity increases sharply. When the temperature rises to 200 °C, their strength reaches the maximum value with the conversion of CO and NO beginning to increase rapidly (which is well in line with the convertion of NO and CO showed in Fig. 1(a, d)), and as the temperature continues to rise, their strength gradually decreases until disappears. Based on the current FT-IR analysis results and previous studies,53 the first peak can be attributed to the vibration of M-NO (M represents some cations), while it is assigned to Cu+-NO (the detailed FT-IR analysis in Fig. S6). The second one has been assigned to CO bonded to Cu+, which is generally accepted by many previous literatures.45,50 Another slightly different peak from the above two, located at 2237 cm-1, was attributed to the gas phase of N2O,48 which began to appear at 200 °C and gradually decreased until disappeared with the temperature increasing. This is highly consistent with the selectivity and catalytic activity of the catalyst showed in Fig. 1(b,c). We may intentionally propose that: at room temperature, NO preferentially adsorbs on some metal cations from the surface of the catalyst to cover some active sites, thereby inhibiting the adsorption of CO;50 With the gradual increase of temperature, part of NOx reacts directly with CO, and the other part desorbs or decomposes; Eventually, the NOx disappears completely at 175 °C, at this point, CO can directly react with surface active oxygen (Cu2+-O-Mn4+, Cu2+-O-Mn3+, Cu2+-O-Fe3+) to make the Cu2+ on the catalyst surface be reduced to Cu0 or Cu+ and generate a large number of oxygen vacancies (Cu+27



□-Mn3+, Cu+-□-Mn2+, Cu+-□-Fe2+).18,45 When the temperature rises to 200 °C, a large amount of Cu+ (some Cu2+-O-Cu2+ are also reduced to Cu+, resulting in further increase in the content of Cu+) is generated on the surface of the catalyst, so that a certain amount of NO and a large amount of CO are adsorbed. Meanwhile, surface oxygen vacancies can weaken the N-O bond and promote the decomposition of the adsorbed NO species into N and O atoms,13 then O combines with the adjacent CO adsorbed on Cu+ to form CO2. Most of the N radicals combine with neighboring Cu+-NO molecules to form N2O, and the rest combined with each other to form N2, which is consistent well with the low N2 selectivity at 200 °C (Fig. 1(b)). Thereafter, as the temperature continues to rise to 300 °C, most of the Cu+ is reduced to Cu0 accompanied by the disappearance of Cu+-CO and Cu+NO. Due to the disappearance of Cu+-NO, most NO may be directly decomposed by oxygen vacancies to O and N, and two N atoms are combined into N2, which can be proved by the selectivity of N2 increasing sharply as shown in the Fig. 1(b). Finally, all the peaks disappeared at 300 °C, except for the CO, NO gas phase peaks. So, the FT-IR spectra results above 300 °C will not be showed. One can find several special phenomenon. 1, No COX absorption peaks can be seen throughout the temperature range, however, the peak of CO2 can be seen, which indicates that CO can quickly react with the catalyst and desorb, so that it does not occupy the reaction site (as showed in Fig. 7); 2, When the temperature rises to 175°C, all NOx species disappear completely, and two special peaks located about 1770-1790 (Cu+-NO) and 2105 cm-1 (Cu+-CO) begin to appear (as showed in Fig. 9); 3, When the temperature rises by about 300 °C, all the peaks disappear, except for the characteristic peak of the gas phase CO2 (as showed in Fig. 9). These phenomena cannot be found in previous literature, The mechanism in this paper is somewhat different from the above literature.24,6,12-14,18,44,45,49,50

3.8. Possible high temperature reaction mechanism over catalyst When the temperature rises to 300 °C, due to high temperature desorption or rapid reaction, all peaks vanish completely except for some peaks of gas phase species (CO, NO, CO2). Therefore, we can not get any information from FT-IR spectra results, but

28


Page 28 of 40

Page 29 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Cu2+ species have been reduced to Cu0 before 300 °C which can be proved by the complete disappearance of peaks for Cu+-CO. To further prove that the Cu species exist in the form of Cu0 on the surface of the catalyst under our present condition at higher temperature (300 °C), we collected the catalyst which had reacted for a long time at 400 °C and been swept by nitrogen to room temperature. The XRD results (Fig. S7(a, b)) showed that all the peaks assigned to CuO disappeared completely except peaks for γ-Fe2O3, accompanied by the appearance of some strong Cu peaks and Fe3O4 peaks, which further demonstrates

that some Cu2+

species may have been reduced to Cu0. The detailed discussion in the figure also proves this. Previous studies13,46 reported the relationship between Cu+ content and temperature in Cu-containing composite oxides, and it was found that when the temperature was lower than 300 °C, Cu+ was completely reduced to Cu0, which is in line with the results. The XRD results of samples having reacted long enough under 400, 600, 800 °C ( showed in Fig. S8) show that it is similar to the crystal phase species at 400 °C. This indicates that there may be no significant change in the active components of the catalyst in the temperature range of 400 to 1000 °C. So, here, it was tentatively believed that under high temperature conditions, γ-Fe2O3 and Fe3O4 may be catalytically active substances. It was reported that both Fe3O4 and γ-Fe2O3 had similar cubic dense stacked oxide ion arrays,11 which explains the interchangeability of the two compounds in redox and reduction, because these reactions have relatively little change on the whole structure. The conversion between Fe3O4 and γ-Fe2O3 is quite easy which is well consistent with the excellent high temperature performance of our catalyst,29 and it is also explained that Fe3O4 and γ-Fe2O3 phases are found coexisting in the process of all high temperature reaction.9 Studies54,55 have shown that the catalytic activity of γ-Fe2O3 is significantly higher than that of α-Fe2O3, which is similar with the study (Fig. 10). However, the thermal stability of α-Fe2O3 is better than that of γ-Fe2O3.56,57 So, it is also possible that part of γ-Fe2O3 is converted to α-Fe2O3 at high temperatures. The XRD results showed in Fig. 11 indicate that α-Fe2O3 is generated in 0MnFeCu sample when the temperature rises to 400 °C. In contrast, samples containing traces of Mn species do not have this corresponding diffraction peak. This may be because when the temperature is increased, the Mn species 29



can suppress the γ-Fe2O3 to α-Fe2O3 conversion, is well consistent with previous report58 in which reported that the Mn3+-doped γ-Fe2O3 displays superparamagnetic behavior at room temperature, and the conversion of γ-Fe2O3 to α-Fe2O3 at high temperatures can also be prevented. Combined with experimental results and literature, it can be inferred that Mn3+ inhibits the conversion of γ-Fe2O3 to α-Fe2O3. The possible high temperature reaction mechanism was proposed. When surface oxygen ions are removed by CO, accompanied by the formation of Fe2+, the polycrystalline transition of γ-Fe2O3 to Fe3O4 begins, and then the newly formed Fe2+ diffuses into the bulk phase and forms Fe3O4. Since the crystal structures of γ-Fe2O3 and Fe3O4 are very similar, the phase transition between them is actually a topological process, and needs no crystallographic changes, which makes the migration of Fe2+ need less energy. Therefore, γ-Fe2O3 can be easily reduced to Fe3O4. Similarly, the transition of Fe3O4 to γ-Fe2O3 starts when NO contacts with oxygen vacancy. The electrons are transferred from bulk Fe2+ to adjacent Fe3+, and finally to the surface to bind with NO after several transfers, so that NO is decomposed into N and O (N and adjacent N combine to form N2). At the same time, the newly generated Fe3+ expands from the body phase to the outside, leaving a cationic vacancy to form γ-Fe2O3.54,59,60

Fig. 10. Conversion of NO to N2 on γ-Fe2O3, α-Fe2O3, Fe3O4.

30


Page 30 of 40

Page 31 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 11. XRD results of 0.010MnFeCu and 0MnFeCu samples.

Fig. 12. NO conversion on CuO, 0.010MnFeCu, γ-Fe2O3 and the physical mixture. To further prove that different crystal phases in the strong synergistic coexistence catalyst system play dominant catalytic roles at different temperatures,61 we show diagram of the relationship between catalytic activity and temperature for different pure phases. It can be clearly seen from Fig. 12, when the temperature rises from 180 to 200 °C, the catalytic activity of pure CuO increases abruptly due to the excellent low temperature catalytic activity, and as the temperature continues to rise, its activity decreases sharply due to most Cu2+ being reduced to Cu0. However, under the same conditions, when the temperature is less than 250 °C, the catalytic conversion of pure 31



Fe2O3 is almost 0, and the catalytic activity gradually increases to 100% as the temperature continues to increase up to 600 °C due to the excellent high temperature catalytic activity. It is clear that the curves of the physical mixture of γ-Fe2O3, CuO, and CuFe2O4 are just simple summations of pure species. The difference from the above three is that the present sample not only has a low temperature catalytic activity like pure CuO, but also has high temperature stability such as γ-Fe2O3, and also completely better than physical mixing. Combined with the XRD, XPS, FT-IR, SEM-Mapping and in situ FT-IR analysis, it is believed that when the temperature is lower than 250 °C, the CuO species of which the surface contained Mn species in present three-phase coexisting catalyst play a dominant catalytic role, and most of the Cu2+ is reduced to Cu0 as the temperature increases to 300 °C. Meanwhile, Fe3O4 (CuFe2O4 was reduced to Fe3O4) begins to appear. When the temperature is higher than 400 °C (400-1000 °C), Fe3O4/γ-Fe2O3 becomes a highly active substance. A possible NO reduction by CO reaction mechanism (schematic diagram) is proposed in Fig. 13.

32


Page 32 of 40

Page 33 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 13. Possible reaction mechanism of NO reduction by CO on 0.010MnCuFe catalyst (schematic diagram). 4. Conclusions 33



The three phases of γ-Fe2O3, CuFe2O4 and CuO co-exist in our catalyst system, resulting in a large number of active oxygen in their interface, which can be reduced to oxygen vacancies by CO, resulting in a good activity of 0MnFeCu catalyst even without manganese. When a suitable amount of manganese is incorporated into the catalyst, the Mn2+ species interacts strongly with Fe species, allowing Cu2+ to precipitate from CuFe2O4 and preferentially grow along the CuO (110) plane surface, which has better catalytic performance. At high temperature (300-1000 °C), Mn3+ can suppress the transformation of γ-Fe2O3 to α-Fe2O3, which leads to a stable high temperature catalytic performance. Mn4+ on the surface of the catalyst generates reactive oxygen species, which can be more easily reduced to form some oxygen vacancies during the reaction. These factors eventually lead to the excellent catalytic performance of 0.010MnFeCu catalyst with a broad temperature window.

Supporting Information Water and sulfur resistance of 0.010 sample; XRD patterns; N2 adsorption–desorption isotherms of all the samples and pore size distribution of the samples; TEM images; In situ FT-IR results

Acknowledgements This work was supported by National Nature Science Foundation of China (Nos. 21507014, 21663006, 21763003), the Program for Science and Technology Development Plan of Nanning (No. 20163146)

References: (1) Liang, Z.; Ma, X.; Lin, H.; Tang, Y. The energy consumption and environmental impacts of SCR technology in China. Applied Energy 2011, 88, 1120-1129. (2) Xiong, Y.; Yao, X.; Tang, C.; Zhang, L.; Cao, Y.; Deng, Y.; Gao, F.; Dong, L. Effect of COpretreatment on the CuO–V2O5/γ-Al2O3 catalyst for NO reduction by CO. Catalysis Science & Technology 2014, 4, 4416-4425. (3) Dong, L.; Zhang, B.; Tang, C.; Li, B.; Zhou, L.; Gong, F.; Sun, B.; Gao, F.; Dong, L.; Chen, Y. Influence of CeO2 modification on the properties of Fe2O3–Ti0.5Sn0.5O2 catalyst for NO reduction by CO. Catalysis Science & Technology 2013, 4, 482. 34


Page 34 of 40

Page 35 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(4) Wang, L.; Cheng, X.; Wang, Z.; Zhang, X.; Ma, C. NO reduction by CO over iron-based catalysts supported by activated semi-coke. The Canadian Journal of Chemical Engineering 2017, 95, 449-458. (5) Shen, B.; Liu, T.; Zhao, N.; Yang, X.; Deng, L. Iron-doped Mn-Ce/TiO2 catalyst for low temperature selective catalytic reduction of NO with NH3. Journal of Environmental Sciences 2010, 22, 1447-54. (6) Li, J.; Wang, S.; Zhou, L.; Luo, G.; Wei, F. NO reduction by CO over a Fe-based catalyst in FCC regenerator conditions. Chemical Engineering Journal 2014, 255, 126-133. (7) Qu, W.; Chen, Y.; Huang, Z.; Gao, J.; Zhou, M.; Chen, J.; Li, C.; Ma, Z.; Chen, J.; Tang, X. Active Tetrahedral Iron Sites of γ-Fe2O3 Catalyzing NO Reduction by NH3. Environmental Science & Technology Letters 2017, 4, 246-250. (8) Fan, Z.; Shi, J.; Gao, C.; Gao, G.; Wang, B.; Niu, C. Rationally Designed Porous MnOx–FeOx Nanoneedles for Low-Temperature Selective Catalytic Reduction of NOx by NH3. ACS Applied Materials & Interfaces 2017, 9, 16117-16127. (9) Capone, S.; Manera, M. G.; Taurino, A.; Siciliano, P.; Rella, R.; Luby, S.; Benkovicova, M.; Siffalovic, P.; Majkova, E. Fe3O4/γ-Fe2O3 Nanoparticle Multilayers Deposited by the Langmuir–Blodgett Technique for Gas Sensors Application. Langmuir 2014, 30, 1190-1197. (10) Nica, S.; Nica, V.; Grigoras, V. C.; Varganici, C.; Popovici, D.; Hulubei, C.; Ioan, S. Influence of two structural phases of Fe3O4 and γ-Fe2O3 on the properties of polyimide/iron oxide composites. Polymer International 2015, 64, 1172-1181. (11) Daniel, P.; Shylin, S. I.; Lu, H.; Tahir, M. N.; Panth Fer, M.; Weidner, T.; M Ller, A.; Ksenofontov, V.; Tremel, W. The surface chemistry of iron oxide nanocrystals: surface reduction of γ-Fe2O3 to Fe3O4 by redox-active catechol surface ligands. Journal of Materials Chemistry C 2018, 6, 326 - 333. (12) Dasireddy, V. D. B. C.; Likozar, B. Selective catalytic reduction of NOx by CO over bimetallic transition metals supported by multi-walled carbon nanotubes (MWCNT). Chemical Engineering Journal 2017, 326, 886-900. (13) Zhang, L.; Yao, X.; Lu, Y.; Sun, C.; Tang, C.; Gao, F.; Dong, L. Effect of precursors on the structure and activity of CuO-CoO x /γ-Al2O3 catalysts for NO reduction by CO. Journal of Colloid & Interface Science 2017, 509, 334-345. (14) Liu, T.; Yao, Y.; Wei, L.; Shi, Z.; Han, L.; Yuan, H.; Li, B.; Dong, L.; Wang, F.; Sun, C. Preparation and Evaluation of Copper–Manganese Oxide as a High-Efficiency Catalyst for CO Oxidation and NO Reduction by CO. The Journal of Physical Chemistry C 2017, 121, 12757-12770. (15) Zhuang, Y.; Gao, W.; Yu, Y.; Wang, J. Facile fabrication of three-dimensional porous CuFe2O4 cages as highly efficient and recyclable heterogeneous catalyst. Materials & Design 2017, 130, 294-301. (16) Dharmarathna, S.; King Ondu, C. K.; Pahalagedara, L.; Kuo, C.; Zhang, Y.; Suib, S. L. Manganese octahedral molecular sieve (OMS-2) catalysts for selective aerobic oxidation of thiols to disulfides. Applied Catalysis B: Environmental 2014, 147, 124-131.

35



(17) Zha, K.; Cai, S.; Hu, H.; Li, H.; Yan, T.; Shi, L.; Zhang, D. In Situ DRIFTs Investigation of Promotional Effects of Tungsten on MnOx -CeO2 /meso-TiO2 Catalysts for NOx Reduction. The Journal of Physical Chemistry C 2017, 121, 25243-25254. (18) Wang, L.; Cheng, X.; Wang, Z.; Ma, C.; Qin, Y. Investigation on Fe-Co binary metal oxides supported on activated semi-coke for NO reduction by CO. Applied Catalysis B: Environmental 2017, 201, 636-651. (19) Shen, Q.; Zhang, L.; Sun, N.; Wang, H.; Zhong, L.; He, C.; Wei, W.; Sun, Y. Hollow MnOx-CeO 2 mixed oxides as highly efficient catalysts in NO oxidation. Chemical Engineering Journal 2017, 322, 4655. (20) Shan, W.; Liu, F.; He, H.; Shi, X.; Zhang, C. A superior Ce-W-Ti mixed oxide catalyst for the selective catalytic reduction of NOx with NH3. Applied Catalysis B: Environmental 2012, 115-116, 100106. (21) Yan, L.; Liu, Y.; Zha, K.; Li, H.; Shi, L.; Zhang, D. Scale–Activity Relationship of MnOx -FeOy Nanocage Catalysts Derived from Prussian Blue Analogues for Low-Temperature NO Reduction: Experimental and DFT Studies. ACS Applied Materials & Interfaces 2017, 9, 2581-2593. (22) Zhang, Y.; Qin, Z.; Wang, G.; Zhu, H.; Dong, M.; Li, S.; Wu, Z.; Li, Z.; Wu, Z.; Zhang, J.; Hu, T.; Fan, W.; Wang, J. Catalytic performance of MnOx–NiO composite oxide in lean methane combustion at low temperature. Applied Catalysis B: Environmental 2013, 129, 172-181. (23) Li, J.; Liang, X.; Xu, S.; Hao, J. Catalytic performance of manganese cobalt oxides on methane combustion at low temperature. Applied Catalysis B: Environmental 2009, 90, 307-312. (24) Zhang, W.; Quan, B.; Lee, C.; Park, S.; Li, X.; Choi, E.; Diao, G.; Piao, Y. One-Step Facile Solvothermal Synthesis of Copper Ferrite–Graphene Composite as a High-Performance Supercapacitor Material. ACS Applied Materials & Interfaces 2015, 7, 2404-2414. (25) Zhang, Q.; Wang, H.; Ning, P.; Song, Z.; Liu, X.; Duan, Y. In situ DRIFTS studies on CuO-Fe2O3 catalysts for low temperature selective catalytic oxidation of ammonia to nitrogen. Applied Surface Science 2017, 419, 733-743. (26) Hussain, S.; Hussain, S.; Waleed, A.; Tavakoli, M. M.; Wang, Z.; Yang, S.; Fan, Z.; Nadeem, M. A. Fabrication of CuFe2O4/α-Fe2O3 Composite Thin Films on FTO Coated Glass and 3-D Nanospike Structures for Efficient Photoelectrochemical Water Splitting. ACS Applied Materials & Interfaces 2016, 8, 35315-35322. (27) Sun, G.; Dong, B.; Cao, M.; Wei, B.; Hu, C. Hierarchical Dendrite-Like Magnetic Materials of Fe3O4, γ-Fe2O3, and Fe with High Performance of Microwave Absorption. Chemistry of Materials 2011, 23, 15871593. (28) Lu, J.; Jiao, X.; Chen, D.; Li, W. Solvothermal Synthesis and Characterization of Fe3O4 and γ-Fe2O3 Nanoplates. The Journal of Physical Chemistry C 2009, 113, 4012-4017.

36


Page 36 of 40

Page 37 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(29) Zhang, X.; Niu, Y.; Meng, X.; Li, Y.; Zhao, J. Structural evolution and characteristics of the phase transformations between α-Fe2O3, Fe3O4 and γ-Fe2O3 nanoparticles under reducing and oxidizing atmospheres. Crystengcomm 2013, 15, 8166-8172. (30) Liu, X.; Jin, Z.; Lu, J.; Wang, X.; Luo, M. Highly active CuO/OMS-2 catalysts for low-temperature CO oxidation. Chemical Engineering Journal 2010, 162, 151-157. (31) Li, Y.; Wan, Y.; Li, Y.; Zhan, S.; Guan, Q.; Tian, Y. Low-Temperature Selective Catalytic Reduction of NO with NH3 over Mn2O3 -Doped Fe2O3 Hexagonal Microsheets. ACS Applied Materials & Interfaces 2016, 8, 5224-5233. (32) Dhanda, R.; Kidwai, M. Magnetically separable CuFe2O4 /reduced graphene oxide nanocomposites: as a highly active catalyst for solvent free oxidative coupling of amines to imines. RSC Advances 2016, 6, 53430-53437. (33) Ding, M.; Qiu, M.; Wang, T.; Ma, L.; Wu, C.; Liu, J. Effect of iron promoter on structure and performance of CuMnZnO catalyst for higher alcohols synthesis. Applied Energy 2012, 97, 543-547. (34) Ding, M.; Qiu, M.; Liu, J.; Li, Y.; Wang, T.; Ma, L.; Wu, C. Influence of manganese promoter on coprecipitated Fe–Cu based catalysts for higher alcohols synthesis. Fuel 2013, 109, 21-27. (35) Lu, X.; Bian, W.; Min, C.; Fu, Z.; Zhang, Q.; Zhu, H. Cation distribution of high-performance Mnsubstituted ZnGa2O4 microwave dielectric ceramics. Ceramics International 2018, 44, 10028-10034. (36 Liang, X.; Zhong, Y.; Zhu, S.; He, H.; Yuan, P.; Zhu, J.; Jiang, Z. The valence and site occupancy of substituting metals in magnetite spinel structure Fe3−xMxO4 (M = Cr, Mn, Co and Ni) and their influence on thermal stability: An XANES and TG-DSC investigation. Solid State Sciences 2013, 15, 115-122. (37) Peng, H.; Liu, Y.; Guo, Y.; Zhang, J.; Zhang, L.; Zhou, S.; Xu, X.; Liu, W.; Zhang, N.; Wang, X. Treating CuO nano‐flowers by H2O2: a novel and facile strategy to prepare high performance CuO nano‐sheets with exposed (110) facets. Chemcatchem 2016, 8, 3714-3719. (38) Zhao, Y.; Shi, H.; Chen, M.; Teng, F. Splitting growth of novel CuO straw sheaves and their improved photocatalytic activity due to exposed active {110} facets and crystallinity. CrystEngComm 2014, 16, 2417-2423. (39) Yang, S.; Su, W.; Lin, S. D.; Rick, J.; Cheng, J.; Liu, J.; Pan, C.; Liu, D.; Lee, J.; Chan, T.; Sheu, H.; Hwang, B. Preparation of nano-sized Cu from a rod-like CuFe2O4: Suitable for high performance catalytic applications. Applied Catalysis B: Environmental 2011, 106, 650-656. (40) Faungnawakij, K.; Kikuchi, R.; Fukunaga, T.; Eguchi, K. Catalytic hydrogen production from dimethyl ether over CuFe2O4 spinel-based composites: Hydrogen reduction and metal dopant effects. Catalysis Today 2008, 138, 157-161. (41) Lin, X.; Zhang, Y.; Yin, L.; Chen, C.; Zhan, Y.; Li, D. Characterization and catalytic performance of copper-based WGS catalysts derived from copper ferrite. International Journal of Hydrogen Energy 2014, 39, 6424-6432.

37



(42) Perin, G.; Fabro, J.; Guiotto, M.; Xin, Q.; Natile, M. M.; Cool, P.; Canu, P.; Glisenti, A.Cu@LaNiO 3 based nanocomposites in TWC applications. Applied Catalysis B: Environmental 2017, 209, 214-227. (43) Chi, G.; Shen, B.; Yu, R.; He, C.; Zhang, X. Simultaneous removal of NO and Hg0 over Ce-Cu modified V2O5 /TiO2 based commercial SCR catalysts. Journal of Hazardous Materials 2017, 330, 83-92. (44) Sun, C.; Tang, Y.; Gao, F.; Sun, J.; Ma, K.; Tang, C.; Dong, L. Effects of different manganese precursors as promoters on catalytic performance of CuO-MnOx/TiO2catalysts for NO removal by CO. Physical Chemistry Chemical Physics 2015, 17, 15996-16006. (45) Xiong, Y.; Yao, X.; Tang, C.; Zhang, L.; Cao, Y.; Deng, Y.; Gao, F.; Dong, L. Effect of COpretreatment on the CuO–V2O5/γ-Al2O3 catalyst for NO reduction by CO. Catalysis Science & Technology 2014, 4, 4416-4425. (46) Qi, L.; Yu, Q.; Dai, Y.; Tang, C.; Liu, L.; Zhang, H.; Gao, F.; Dong, L.; Chen, Y. Influence of cerium precursors on the structure and reducibility of mesoporous CuO-CeO2 catalysts for CO oxidation. Applied Catalysis B: Environmental 2012, 119-120, 308-320. (47) Wan, H.; Wang, Z.; Zhu, J.; Li, X.; Liu, B.; Gao, F.; Dong, L.; Chen, Y. Influence of CO pretreatment on the activities of CuO/γ-Al2O3 catalysts in CO+O2 reaction. Applied Catalysis B: Environmental 2008, 79, 254-261. (48) Yao, X.; Gao, F.; Yu, Q.; Qi, L.; Tang, C.; Dong, L.; Chen, Y. NO reduction by CO over CuO–CeO2 catalysts: effect of preparation methods. Catalysis Science & Technology 2013, 3, 1355-1366. (49) Liu, T.; Qian, J.; Yao, Y.; Shi, Z.; Han, L.; Liang, C.; Li, B.; Dong, L.; Fan, M.; Zhang, L. Research on SCR of NO with CO over the Cu0.1La0.1Ce0.8O mixed-oxide catalysts: Effect of the grinding. Journal of Molecular Catalysis A Chemical 2016, 430. (50) Liu, L.; Liu, B.; Dong, L.; Zhu, J.; Wan, H.; Sun, K.; Zhao, B.; Zhu, H.; Dong, L.; Chen, Y. In situ FTinfrared investigation of CO or/and NO interaction with CuO/Ce0.67Zr0.33O2 catalysts. Applied Catalysis B: Environmental 2009, 90, 578-586. (51) Qi, G.; Yang, R. T.; Chang, R. MnOx-CeO2 mixed oxides prepared by co-precipitation for selective catalytic reduction of NO with NH3 at low temperatures. Applied Catalysis B: Environmental 2004, 51, 93106. (52) Deng, C.; Huang, Q.; Zhu, X.; Hu, Q.; Su, W.; Qian, J.; Dong, L.; Li, B.; Fan, M.; Liang, C. The influence of Mn-doped CeO2 on the activity of CuO/CeO2 in CO oxidation and NO + CO model reaction. Applied Surface Science 2016, 389, 1033-1049. (53) Kantcheva, M. FT-IR spectroscopic investigation of the reactivity of NOx species adsorbed on Cu/ZrO and CuSO/ZrO catalysts toward decane. Applied Catalysis B: Environmental 2003, 42, 89-109. (54) Wei, X.; Zhou, Y.; Li, Y.; Shen, W. Polymorphous transformation of rod-shaped iron oxides and their catalytic properties in selective reduction of NO by NH3. RSC Advances 2015, 5, 66141 - 66146.

38


Page 38 of 40

Page 39 of 40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(55) Zhao, N.; Ma, W.; Cui, Z.; Song, W.; Xu, C.; Gao, M. Polyhedral Maghemite Nanocrystals Prepared by a Flame Synthetic Method: Preparations, Characterizations, and Catalytic Properties. Acs Nano 2009, 3, 1775-80. (56) Yen, F. S.; Chen, W. C.; Yang, J. M.; Hong, C. T. Crystallite Size Variations of Nanosized Fe2O3 Powders during γ- to α-Phase Transformation. Nano Letters 2002, 2, 245-252. (57) Azadmanjiri, J.; Simon, G. P.; Suzuki, K.; Selomulya, C.; Cashion, J. D.Phase reduction of coated maghemite (γ-Fe2O3 ) nanoparticles under microwave-induced plasma heating for rapid heat treatment. J. Mater. Chem. 2012, 22, 617-625. (58) Lai, J.; Shafi, K. V. P. M.; Loos, K.; Ulman, A.; Lee, Y.; Vogt, T.; Estournès, C. Doping γ-Fe2O3 Nanoparticles with Mn(III) Suppresses the Transition to the α-Fe2O3 Structure. Journal of the American Chemical Society 2003, 125, 11470-11471. (59) Gallagher, K. J.; Feitknecht, W.; Mannweiler, U. Mechanism of Oxidation of Magnetite to γ-Fe2O3. Nature 1968, 217, 1118-1121. (60) F, J.; JC, J.; M, W.; AV, C.; A, H.; PG, B. Synthesis of ordered mesoporous Fe3O4 and gamma-Fe2O3 with crystalline walls using post-template reduction/oxidation. Journal of the American Chemical Society 2006, 128, 12905-9. (61) Yun, L.; Li, Y.; Zhou, C.; Lan, L.; Zeng, M.; Mao, M.; Liu, H.; Zhao, X. The formation of CuO/OMS2 nanocomposite leads to a significant improvement in catalytic performance for NO reduction by CO. Applied Catalysis A: General 2017, 530, 1-11.

39



130x56mm (96 x 96 DPI)


Page 40 of 40

Mn-modified CuO, CuFe2O4 and Î³-Fe2O3 three-phase strong

Recommend Documents