Reaction behaviors of CH3CN catalytic combustion over CuCeOx-HZSM-5 composite catalysts: The mechanism of enhanced N2 selectivity

Article abstract of DOI:10.1016/j.apcata.2019.117373

In this paper, the reaction behaviors of CH₃CN catalytic combustion over CuCeO_x-HZSM-5 catalysts prepared by the physical mixing and wet impregnation methods were investigated. The experimental results had revealed that the composite catalysts exhibited excellent degradation efficiency and high N₂ selectivity: the Cu-Ce oxides ensured the high activity, and the introduction of HZSM-5 limited the formation of NO_x. The involved reaction mechanism was then studied by an in situ DRIFT analysis. It could be deduced that different reaction pathways would occur over CuCeO_x and HZSM-5. The catalytic reaction over CuCeO_x mainly proceeded via an oxidation pathway. The CN group could be converted to isocyanate species [NCO(a)] which would be further oxidized to NO_x. Over HZSM-5, CH₃CN was mainly hydrolyzed into ammonia. For the CuCeO_x-HZSM-5 composite catalysts, the over-oxidation of NCO(a) and NH₃ was inhibited, and an internal selective catalytic reduction (SCR) reaction could occur, thereby greatly improving the N₂ selectivity.

Full text of DOI:10.1016/j.apcata.2019.117373

Journal Pre-proof
Reaction behaviors of CH3CN catalytic combustion over
CuCeOx-HZSM-5 composite catalysts: The mechanism of enhanced N2
selectivity
Yuxing Wang (Investigation) (Writing - original draft), Qingji Ying
(Investigation), Yaoyu Zhang (Formal analysis), Yue Liu
(Supervision) (Writing - review and editing), Zhongbiao Wu
(Resources)<ce:contributor-role>Writing - review and editing)
PII:
S0926-860X(19)30528-9
DOI:
https://doi.org/10.1016/j.apcata.2019.117373
APCATA 117373
Reference:
To appear in:
Applied Catalysis A, General
Received Date:
Revised Date:
Accepted Date:
29 October 2019
4 December 2019
6 December 2019
Please cite this article as: Wang Y, Ying Q, Zhang Y, Liu Y, Wu Z, Reaction behaviors of
CH3CN catalytic combustion over CuCeOx-HZSM-5 composite catalysts: The mechanism of
enhanced N2 selectivity, Applied Catalysis A, General (2019),
doi: https://doi.org/10.1016/j.apcata.2019.117373

This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2019 Published by Elsevier.

Reaction behaviors of CH3CN catalytic combustion over CuCeOx-HZSM-5
composite catalysts: The mechanism of enhanced N2 selectivity
Yuxing Wang1, Qingji Ying1, Yaoyu Zhang1, Yue Liu1* yueliu@zju.edu.cn and
Zhongbiao Wu1,2
1Department of Environmental Engineering, Zhejiang University, 866 Yuhangtang
Road, Hangzhou, 310058, P. R. China.
2Zhejiang Provincial Engineering Research Center of Industrial Boiler & Furnace
Flue Gas Pollution Control, 866 Yuhangtang Road, Hangzhou, 310058, P. R. China.
*Corresponding author: Tel: +86 571 87953088; Fax: +86 87953088;

Graphical abstract
Highlights




High N2 selectivity was achieved on CuCeOx-HZSM-5 composite catalysts.
The introduction of HZSM-5 enhanced the hydrolysis of CH3CN and NCO(a).
The enhanced N2 selectivity could be attributed to the internal SCR reaction.
The synergism between CuCeOx and HZSM-5 facilitated the internal SCR reaction.
Abstract:
In this paper, the reaction behaviors of CH3CN catalytic combustion over CuCeOx-
HZSM-5 catalysts prepared by the physical mixing and wet impregnation methods

were investigated. The experimental results had revealed that the composite catalysts
exhibited excellent degradation efficiency and high N2 selectivity: the Cu-Ce oxides
ensured the high activity, and the introduction of HZSM-5 limited the formation of
NOx. The involved reaction mechanism was then studied by an in situ DRIFT
analysis. It could be deduced that different reaction pathways would occur over
CuCeOx and HZSM-5. The catalytic reaction over CuCeOx mainly proceeded via an
oxidation pathway. The CN group could be converted to isocyanate species [NCO(a)]
which would be further oxidized to NOx. Over HZSM-5, CH3CN was mainly
hydrolyzed into ammonia. For the CuCeOx-HZSM-5 composite catalysts, the over-
oxidation of NCO(a) and NH3 was inhibited, and an internal selective catalytic
reduction (SCR) reaction could occur, thereby greatly improving the N2 selectivity.
Keywords: Acetonitrile; CuCeOx; HZSM-5; Catalytic combustion; N2 selectivity
1 Introduction
Volatile organic compounds (VOCs) have received widespread attentions as major
precursors of ozone and secondary organic aerosols [1]. As one category of highly
toxic VOCs, the emission of cyanide-containing waste gases, such as acetonitrile and
acrylonitrile, have become a serious threaten to human health [2]. Incineration and
catalytic combustion are the most promising methods for eliminating VOCs.
However, the incineration of cyanide-containing gas would normally lead to high
levels of secondary pollution (NOx) emissions and high energy costs [3, 4]. As such,

catalytic combustion has been recognized as one of the most suitable methods for
cyanide-containing gas abatement due to its low operating temperature and less
secondary pollution [5].
Currently, there are still few reports focused on the catalytic combustion of cyanide-
containing gases. The noble metal catalysts had showed excellent low-temperature
activities, but their N2 selectivity was not satisfactory. For instance, Pt/Al2O3 catalysts
had exhibited a total HCN conversion of 95% at temperatures above 250 °C, while
the N2 selectivity was found to be less than 30% [6]. Adsorbed NCO species have
been proposed to be important intermediates for the oxidation of CN groups, which
could be rapidly converted to NOx and N2 [7]. Similar research results were presented
for the catalytic combustion of CH3CN [8]. Moreover, due to the superiorities in
terms of their economic efficiency, transition metal oxide catalysts were also
employed for cyanide-containing compound catalytic combustion in several previous
studies [4, 9, 10]. Although superior activities were obtained, NOx formation at high
temperatures was still a problem.
Meanwhile, cyanide-containing compounds could hydrolyze over some catalysts,
including TiO2, Al2O3 and HZSM-5 [11]. For example, the hydrolysis of acrylonitrile
proceeds over zeolite catalysts to form acetic acid and NH3 [12]. However, these
abovementioned catalysts possessed an insufficient deep oxidation ability. This fact
motivated the design of a composite catalyst possessing both a high activity and a
good N2 selectivity, in which (1) one component comprised materials such as
transition metal oxides with a high activity for the deep oxidation of a cyanide-

containing gas and (2) the other component possessed a hydrolysis functionality to
produce NH3 for reducing NOx via an internal SCR reaction cycle. In this work, Cu-
Ce mixed oxides were selected as the main oxidation component, since such catalysts
were widely reported for their excellent activities in CO oxidation and VOC catalytic
abatement, owing to the good reducibility induced by copper-cerium interactions [13-
15]. Additionally, HZSM-5 zeolites were employed herein as the component used to
facilitate the hydrolysis process due to their advantages of high specific surface areas,
abundant acidic sites and good thermal stabilities [16]. Moreover, transition metal-
modified zeolites have shown a satisfactory activity and N2 selectivity for NH3
selective oxidation and NH3-SCR reactions [17, 18].
In this paper, acrylonitrile (CH3CN) was selected as the target pollutant. In addition,
the CH3CN catalytic combustion performances (including the activity, CO2 yield and
N-containing byproduct yield) over CuCeOx-HZSM-5 composite catalysts
synthesized by physical mixing and wet impregnation methods were investigated.
After that, various characterization studies were performed to reveal the reaction
behaviors of CH3CN catalytic combustion and the intrinsic mechanism regarding the
structure-performance relationship.
2 Experimental
2.1 Materials
HZSM-5 zeolites were purchased from Tianjin Zeolite Plant of Nankai University
(China). The Ce(NO3)3·6H2O (AR), Cu(NO3)2·3H2O (AR) and citric acid (AR) were
purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd. (China).

Deionized water was used in all experiments.
2.2 Catalyst preparation
The CuCeOx catalyst was prepared via the following synthesis method.
Ce(NO3)3·6H2O and Cu(NO3)2·3H2O were used as the precursors of the copper and
cerium species, respectively, with a molar ratio Cu/(Cu+Ce) of 0.5. First, the two
precursors were added to a citric acid solution to achieve a molar ratio of the citric
acid/metal components of 1.3 and a molarity (Cu+Ce) of 0.6 M. The mixture was
continuously stirred for 4 h. Then, the solution was dried at 120 °C for approximately
20 h and calcined at 500 °C for 4 h in static air.
The CuCeOx-HZSM-5 composite catalysts were prepared by physical mixing and
traditional impregnation methods. CuCeOx+HZSM-5 (SiO2/Al2O3=18, 27, 300) was
prepared by mechanically mixing the CuCeOx catalyst and HZSM-5 zeolite at a
weight ratio of 1/1. CuCe/ZSM-5 (SiO2/Al2O3=27) was prepared by a traditional
impregnation method. The detailed procedures were as follows: HZSM-5 was
impregnated by the Ce(NO3)3·6H2O and Cu(NO3)2·3H2O mixture. Similarly, the
molar ratio of Cu/(Cu+Ce) was 0.5, and the mass ratio of CuCeOx and HZSM-5 was
1/1. The drying and calcining procedures were the same as the procedures used to
prepare CuCeOx.
2.3 Catalysts characterizations
The BET analysis was conducted using a nitrogen adsorption apparatus (JW-BK132F,
China) to obtain the specific surface areas, pore volume and pore size of the catalysts.
XRD was conducted using an X-ray diffractometer (model D/max RA, Rigaku Co.,
Japan; Cu Kα radiation, 0.154118 nm) to determine the crystal phases of the catalysts.

X-ray photoelectron spectroscopy (XPS) was performed with Al Kα X-rays (Thermal,
ESCALAB 250, USA) to determine the chemical states of the catalyst surface atoms.
Hydrogen temperature-programmed reduction (H2-TPR) was conducted using a
custom-made multipurpose adsorption instrument (TP-5089, Tianjin Xianquan Co.,
Ltd., China) equipped with a thermal conductivity detector (TCD) and a mass
spectrometer (Hiden Analytical Ltd., UK). For the H2-TPR experiments, 0.05 g of the
samples were pretreated in a flow of O2/He at 400 °C for 1 h, and then, the
temperature was cooled to 30 °C. At 30 °C, the He flow was switched to a flow of 6%
H2/N2, after which the temperature was increased from 30 °C to 800 °C at a rate of
10 °C/min. The temperature-programmed surface reaction of CH3CN (CH3CN-TPSR)
and NH3 (NH3-TPSR) was also carried out on a TP-5089 device. For the CH3CN-
TPSR experiments, 0.75 g of the samples were pretreated in a flow of O2/He at
400 °C for 1 h, and then, the temperature was cooled to 30 °C. At 30 °C, the O2/He
flow was switched to a flow of 600 ppm CH3CN (200 ppm NH3 for the NH3-TPSR),
5% O2 and He as a carrier gas (60 mL/min) for 1.5 h. Thereafter, the temperature was
increased from 30 °C to 700 °C at a rate of 5 °C/min, and the data were recorded with
a mass spectrometer. In situ DRIFT was conducted using a Nicolet 6700 FT-IR
spectrometer equipped with an MCT detector. The catalysts were pretreated in a flow
of He at 300 °C for 1 h in the cell and then cooled to the corresponding temperature
(100 °C, 250 °C or 350 °C) to obtain the background. Thereafter, the IR spectra were
collected at different times in a flow of 600 ppm CH3CN with 5% O2/He.
2.4 Activity measurements
The catalytic activity tests were performed in a fixed-bed microreactor at atmospheric

pressure. To investigate the effect of each component on the activity and selectivity of
the composite catalysts, 0.625 g of the individual catalysts or 1.25 g of the composite
catalysts (40-60 mesh) were loaded in a quartz tube reactor. The flow rate of the
reactant gas was 1000 mL/min, which corresponds to W/F=0.0375 or 0.075 gs/mL.
The reactant gas consisted of ca. 600 ppm CH3CN, 5% O2 and N2 as a balance gas.
The inlet and outlet concentration of CH3CN were analyzed online by a gas
chromatograph with an FID, while CO2 was analyzed with a nickel conversion
furnace and an FID. The concentration of NO and NO2 were analyzed with an
infrared gas analyzer (MGA5, MRU Co., Germany); the concentration of N2O was
analyzed by another flue gas analyzer (G200, Geotech Co., UK); the effluent stream
containing NH3 was first absorbed in a 0.01 M H2SO4 solution. Then, the
concentration of NH3 was determined by Nessler’s reagent spectrophotometry.
The N2 selectivity was not directly determined, as N2 is not detectable by infrared gas
analyzers. It is feasible to evaluate the N2 selectivity by calculating the yield of N-
containing byproducts (NH3, NO, NO2 and N2O). The conversion of CH3CN and the
yield of CO2, NH3, NO, NO2 and N2O were defined as follows:
Conversion of CH3CN = ([CH3CN]in − [CH3CN]out) / [CH3CN]in ×100%
Yield of CO2 = [CO2]out / 2 [CH3CN]in × 100%
Yield of NOx = ([NO]out + [NO2]out + 2[N2O]out) / [CH3CN]in ×100%
Yield of NH3 = [NH3]out / [CH3CN]in × 100%
3 Results and discussions
(1)
(2)
(3)
(4)
3.1 Catalytic performances

The catalytic degradation efficiencies and yield of the products (including CO2, NH3
and NOx) for CH3CN catalytic combustion over different samples were presented in
Fig. 1. It was found that CuCeOx had the highest activity, and CH3CN could be
completely converted at 225 °C. The activities of CuCeOx+HZSM-5 and CuCe/ZSM-
5 were lower than that of the CuCeOx sample. In addition, no conversion of CH3CN
could be observed on pure HZSM-5 until 350 °C was reached. The CO2 yield profile
depicted in Fig. 1b shows that the variations of the CO2 yield fitted well with the
results of the activity. The CO2 yield of all the samples except pure HZSM-5 could
approach 90% at temperatures above 250 °C.
The NOx yield of all the samples is shown in Fig. 1c. Over the CuCeOx sample, the
yield of NOx rapidly increased when the temperature increased to greater than 200 °C.
For instance, a NOx yield of more than 40% was achieved at 275 °C. Thereafter, the
yield of NOx was slowly increasing with increased temperatures. After introducing
HZSM-5, the NOx yield of the CuCeOx+HZSM-5 and CuCe/ZSM-5 samples
decreased dramatically. For the CuCe/ZSM-5 sample, the NOx yield was less than
10% within the temperature range of 150-400 °C. As shown in Fig. 1d, NH3 could
form over CuCeOx at lower temperatures, while the NH3 formation over HZSM-5
could be detected at higher temperatures, which may be due to HZSM-5 having quite
a low activity. The yield of NH3 over CuCeOx and CuCeOx+HZSM-5 displayed a
maximum value at 200 °C. From 200 to 250 °C, the yield of NH3 decreased, and the
yield of NOx increased simultaneously. It could be speculated that the oxidation of
NH3 could be one of the sources of NOx. However, the decreased NH3 emissions did

not lead to an evident increase in the NOx formed over the CuCe/ZSM-5 sample.
Additionally, the stability test of the CuCe/ZSM-5 sample was carried out at 250 °C
for 18 h, as illustrated in Fig. S1, which confirmed the very good stability of the
catalyst.
3.2 Physical property analysis
Table 1 shows the textural properties of the different samples, including the BET
surface area, pore volume and pore size. It could be observed that CuCeOx (41.3
m2/g) possessed a mesoporous structure, whose pores were much larger than those of
the HZSM-5-related samples. In addition, the CuCe/ZSM-5 and CuCeOx+HZSM-5
catalysts had higher BET surface areas compared to the CuCeOx sample, owing to the
HZSM-5 support having a much larger surface area (370.3 m2/g) than the CuCeOx
sample. Compared with the physically mixed sample, CuCe/ZSM-5 had a lower
surface area and smaller pore volume, because more metal species migrated and
accumulated inner the HZSM-5 microchannel [19]. Fig. 2 illustrated the XRD
patterns of all the samples. For the CuCeOx sample, the diffraction peaks located at
28.5°, 33.1°, 47.5° and 56.3° could be ascribed to CeO2 with a fluorite-like structure
(JCPDS 81-0792). In addition, the two diffraction peaks located at 35.6° and 38.8°
were assigned to CuO (JCPDS 89-8599). For HZSM-5, the diffraction peaks located
at 23.0°, 23.9°, 24.3°, 26.9°, 29.9°, 45.0° and 45.6° could be ascribed to the
characteristic MFI-type framework of HZSM-5 (JCPDS 79-2401) [20, 21]. For the
composite catalysts, the peaks corresponding to CeO2 and CuO could be clearly
observed. And the MFI-type framework of HZSM-5 was preserved, indicating that

metal loading will not change the structure of the HZSM-5 support.
3.3 Reducibility and surface properties
The H2-TPR experiments were conducted for the CuCeOx-HZSM-5 catalysts to
evaluate the reducibility of various samples. As shown in Fig. 3, the CuCeOx and
CuCeOx+HZSM-5 samples exhibited two reduction peaks at the lower temperature
range. Based on the previous studies [22-24], the peak located at 150-160 °C could be
attributed to finely dispersed Cu species strongly interacting with CeO2, and the peak
located at ~200 °C could be attributed to larger CuO clusters on the surface of CeO2.
For CuCe/ZSM-5, the first peak located at ~200 °C could be attributed to the
reduction of larger CuO clusters that are less associated with CeO2 or isolated Cu2+
ions, and the other peak at approximately 250 °C could be assigned to the reduction of
large CuO particles [25, 26]. It is well known that Cu2+ resulted from the Cu ions
occupying the ion-exchange sites of HZSM-5 [21]. The strong interaction between the
Cu species and the HZSM-5 support would weaken the Cu-Ce interactions, leading to
the lower reducibility of the CuCe/ZSM-5 sample. For pure HZSM-5, no obvious
reduction peak was observed within the temperature range investigated, suggesting
the absence of reducible oxygen species. The H2-TPR results were in good
accordance with the activity results.
To further investigate the surface atomic states, XPS characterization was then carried
out. Fig. 4A shows the Cu 2p core level spectra for the CuCeOx-HZSM-5 samples.
For all the samples, the Cu 2p3/2 peak located at ~933.5 eV and the corresponding
shake-up satellites located at 941-943.4 eV confirmed the appearance of bulk-like

CuO. In addition, the peak located at ~935.8 eV could be ascribed to Cu2+ that was
bound to weak oxygen species [27, 28]. Particularly, for CuCe/ZSM-5, the peak
located at ~935.8 eV could be ascribed to isolated Cu2+ that was tetrahedrally
coordinated in the HZSM-5 framework. It could be found that more Cu2+ species
existed on the CuCe/ZSM-5 sample, which confirmed the strong interaction between
the Cu species and the support. For the CuCeOx and CuCeOx+HZSM-5 samples, the
Cu 2p3/2 peak located at 932.5 eV could be ascribed to Cu+ [29]. Additionally, the
presence of Cu+ indicated the strong interaction between CuO and CeO2 [13], which
was consistent with the high reducibility of the CuCeOx and CuCeOx+HZSM-5
samples, as shown by the H2-TPR results [30].
The Ce 3d core level spectra are shown in Fig. 4B. The curves of the Ce 3d spectra
were composed of ten peaks, and the “v” with “u” peaks represented the spin orbit
split 3d5/2 and 3d3/2 states, respectively. The doublets (u0, u′) and (v0, v′) were assigned
to Ce3+, while the other peaks were assigned to Ce4+ [31, 32]. The O 1S core level
spectra was shown in Fig. 4C. The peak (α) at ~529.6 eV could be ascribed to the
lattice oxygen (Olαtt) of the metal oxides. In addition, the peak (β) at 530.6-531.5 eV
could be ascribed to the adsorbed oxygen (Oads) [19, 33]. For the CuCeOx sample, the
peak (γ) at ~532.7 eV could be attributed to the oxygen in the hydroxyl groups [34].
Meanwhile, the peak (γ) could be attributed to the oxygen of the HZSM-5 support of
the composite catalysts [19, 35]. The surface Ce3+/Ce4+ and Oads/Olatt molar ratios
were calculated from the Ce 3d and O 1S curve-fittings, respectively, and the result is
shown in Table 1. The Ce3+/Ce4+ ratios decreased in the order of CuCeOx+HZSM-5

(0.176) ≈ CuCeOx (0.172)＞CuCe/ZSM-5 (0.138), and the relative abundance of
Oads followed the same order. In general, the abundant Ce3+ species and Oads will be
beneficial to the reducibility, thereby promoting the CH3CN degradation reactivity
[36]. Besides the catalytic activity, N2 selectivity was also very important to CH3CN
catalytic combustion. Therefore, the reaction process was investigated in the
following section to determine the relative mechanism of the enhanced N2 selectivity
over the composite catalysts.
3.4 TPSR of CH3CN oxidation
As shown in Fig. 5, the oxidation behaviors of CH3CN and the formation of the N-
containing products over all the samples were further investigated by the TPSR
technique. Over the CuCeOx sample, the concentration of CH3CN began to decrease
at temperatures above 200 °C, and N2, NH3 and NOx subsequently formed. N2O was
the main byproduct within the temperature range of 200 to 300 °C, while at higher
temperatures, large amounts of NO formed at elevated temperatures. Compared to
CuCeOx, the composite samples showed an evident decrease in the CH3CN
concentration at higher temperatures, suggesting their lower activity. However, it was
found that N2 was the main product that formed over the two composite samples.
Obviously, the introduction of HZSM-5 limited the formation of NOx and improved
the N2 selectivity over the composite catalysts, which was consistent with the results
of the activity tests. Additionally, more NO emissions formed over CuCeOx+HZSM-5
at temperatures over than 400°C compared to the impregnated sample. For the pure
HZSM-5 sample, the rapid oxidation of CH3CN was detected at temperatures greater

than 350 °C, and small amounts of NOx formed. To further disclose the reaction
mechanism responsible for the different N-containing product formation behaviors, in
situ DRIFT measurements were then conducted.
3.5 DRIFT Measurements
3.5.1 CH3CN + O2 coadsorption over HZSM-5
Fig. 6 shows the DRIFT spectra of the adsorption of 600 ppm CH3CN with 5% O2
over HZSM-5 for various times and at different temperatures. At 100 °C (see Fig.
6a), due to the adsorption of CH3CN, a series of IR bands emerged, and their
intensities grew gradually. The bands at 2278 and 2306 cm-1 could be ascribed to the
νCN FR doublet of the CH3CN molecules interacting with strong H-bonding. In
addition, the band at 2326 cm-1 could be attributed to the νCN FR doublet of the
CH3CN molecules coordinated on the Al3+ cations, and its resonance peak was
overlapped with the band at 2278-2306 cm-1. The band at 2939 cm-1 could be assigned
to CH3 stretching [37]. The sources of the bands at 1286 and 1184 cm-1 were
ambiguous and needed further identification. In addition, several weak bands at 1695,
1666 and 1585 cm-1 could also be detected and may be derived from the hydrolysis of
acetonitrile. According to a previous study [12], CH3CN was hydrolyzed by a two-
step process: acetamide formed first and was then rapidly hydrolyzed to release
ammonia and acetic acid. As such, the bands at 1666 and 1585 cm-1 could be ascribed
to the C=O stretching and NH2 bending of acetamide, respectively [38]. In addition,
the band at 1695 cm-1 was attributed to the adsorbed molecules of acetic acid [39]. As
the temperature increased, as shown in Fig. 6b and 6c, the absorbance of the bands at
2278, 2306, 2326, and 2939 cm-1 decreased, which was due to the limited adsorption

and the catalytic conversion of CH3CN at high temperatures. At the same time, it was
found that the absorbance of the bands at 1666 and 1585 cm-1 owing to acetamide
initially increased and then decreased. Additionally, the absorbance of the band (1695
cm-1) attributed to acetic acid was observed to increase rapidly with increasing
temperatures. All these abovementioned results indicated the increased rate of the
CH3CN hydrolysis reaction at higher temperatures. In addition, the appearance of two
broad bands with rapidly increasing intensities at 3384 and 3294 cm-1 with increasing
+
temperatures further confirmed it, which were attributed to adsorbed NH4 and NH3
molecules, respectively [5, 40]. Moreover, the intensities of the bands at 1286 and
1184 cm-1 were observed to decrease initially at 250 °C and then increase at 350 °C,
which was accompanied by their positions shifting to 1288 and 1176 cm-1,
respectively. Due to the ammonia and acetic acid formed from the hydrolysis reaction
at high temperatures, these two bands may be related to the formation of ammonia
and acetic acid. To further identify these two bands, the DRIFT spectra of HZSM-5
exposed to 600 ppm NH3 at 250 °C was then obtained. As shown in Fig. S2, similar
IR bands of adsorbed NH3 located at 1288 and 1176 cm-1 could be detected. Based on
the literatures [41, 42], both bands appearing at high temperatures in Fig. 6 could be
attributed to the symmetric bending vibrations of the N-H bonds in coordinately
adsorbed NH3. Bands located at 1286 and 1184 cm-1 appeared as soon as CH3CN
adsorbed at low temperatures, and these bands were possibly due to the CN groups
interacting with H-bonding. Therefore, the absorbance of the bands located at 1286
and 1184 cm-1 decreased initially at an elevated temperature as less CH3CN adsorbed.

At 350 °C, new faint bands appeared at 1520 and 2202 cm-1, which could be ascribed
to amide (-NH2) species and weakly adsorbed NO, respectively [42].
3.5.2 CH3CN + O2 coadsorption over CuCeOx
Fig. 7 shows the DRIFT spectra of the adsorption of CH3CN in the presence of 5% O2
over CuCeOx for various reaction times and at different temperature. As shown in Fig.
7a, the bands at 2258 and 2293 cm-1 could be ascribed to the νCN FR doublet of
CH3CN interacting with Lewis acid sites. The bands at 2931 and 2999 cm-1 could be
assigned to CH3 stretching [37]. The cleavage of the C-H bond of CH3CN led to the
formation of CH2CN species, and the band at 2120 cm-1 proved the formation of
CH2CN species [43]. In the 1150-1660 cm-1 frequency range, a series of bands at
1556, 1529, 1471, 1427, 1207 and 1178 cm-1 appeared, which could be attributed to
the formation of acetamide monoanions [CH3CONH(a)] owing to the interaction
between CH3CN and the surface hydroxyl groups [43-45]. In addition, the bands at
1653 and 3330 cm-1 were due to the C=O vibration of CH3COO(a) and νNH of
CH3CONH(a) [43], respectively. When the temperature increased to 250 °C, as shown
in Fig. 7b, the band at 2120 cm-1 disappeared, and the absorbance of the bands at
1653, 1529, 1471, 1207 and 1178 cm-1 significantly decreased. New bands appeared
at 2170, 1602, 1344, 1234, 2360 and 2331 cm-1. The band at 2170 cm-1 could be
ascribed to NCO ionically bounded to metal cation [46]. Additionally, the band at
1602 cm-1 was attributed to the formation of NO2 species [47] or bridged nitrate
species [48], indicating the formation of NOx. Furthermore, the band at 1344 cm-1
could be assigned to bidentate carbonates [49], as well as the emission of CO2 (the
bands located at 2360 and 2331 cm-1) [50]. These results verified that the catalytic

oxidation reaction occurred over the CuCeOx sample. The NH3 observed on catalytic
performance may result from the hydrolysis of CH3CONH(a), and the weak band at
1234 cm-1 could be ascribed to NH3 coordinated on the Lewis acid sites [51]. By
further increasing the temperature to 350 °C, all the peaks attributed to the
CH3CONH(a) groups vanished, and the intensity of the new bands at 1375 cm-1 and
in the 1560-1450 cm-1 range, which are attributed to surface nitrate species [42, 52]
dramatically increased, suggesting that a large amount of NOx would be released at
high temperatures. As such, it could be concluded that the over oxidation of NCO(a)
was the main cause of NOx formation. In addition, the band at 2306 cm-1 could also be
ascribed to NCO(a) species [53].
3.5.3 DRIFT of CH3CN + O2 coadsorption over the composite catalysts
To determine the mechanism of the improved N2 selectivity over the CuCeOx-HZSM-
5 composite catalysts, DRIFT spectra of the coadsorption of CH3CN with 5% O2 over
CuCeOx+HZSM-5 and CuCe/ZSM-5 were obtained at 250 °C. As illustrated in Fig.
8a, the bands at 2322 and 2293 cm-1 could ascribed to adsorbed CH3CN molecules.
Additionally, the bands attributed to adsorbed CH3CONH2 (1664 cm-1), CH3COOH
(1695 cm-1), NHx species (3384, 3294 and 3273 cm-1), CH3CONH(a) (1558, 1427 and
1178 cm-1) and carbonates (1344 cm-1) could also be detected on CuCeOx+HZSM-5,
and these bands corresponded to the bands (1666, 1695, 3384 and 3294 cm-1)
observed for the HZSM-5 sample and the bands (1556, 1427, 1178 and 1344 cm-1)
observed for the CuCeOx sample. In addition, the bands at 1543 and 1439 cm-1 could
also be assigned to CH3CONH(a) [45]. Compared to the CuCeOx spectrum, the band
at 2170 cm-1 attributed to NCO(a) nearly vanished, and the band attributed to HNCO

(2264 cm-1) appeared, which indicated the migration of NCO(a) from CuCeOx to
HZSM-5. A similar phenomenon was observed over the Cu-ZSM-5 catalyst [53].
Moreover, the band at 1602 cm-1 disappeared, which indicated a decrease in the
number of NO2 species or bridged nitrate species formed over CuCeOx, suggesting
that less NOx formed. For the CuCe/ZSM-5 sample shown in Fig. 8b, the bands
attributed to adsorbed CH3CONH2 (1662 and 1585 cm-1), CH3COOH (1695 cm-1),
NHx species (3380 and 3273 cm-1), CH3CONH(a) (1558, 1456, 1421, 1371 and 1163
cm-1) and HNCO (2264 cm-1) could also be observed. It is found that the catalytic
reaction occurring over CuCe/ZSM-5 was similar to that occurring over
CuCeOx+HZSM-5, and the latter reaction was more intense, since the intensity of the
IR bands due to CO2 (2360 and 2390 cm-1) were relatively stronger compared with
that of the band attributed to adsorbed CH3CN. Moreover, the relatively stronger IR
bands attributed to the hydrolysis intermediates indicated that the hydrolysis reaction
was more readily to happen over the CuCeOx+HZSM-5 sample due to its stronger
acidity compared to CuCe/ZSM-5 sample (see Fig. S3).
3.6 Mechanism of CH3CN catalytic decomposition
Based on the results of the activity tests, the characterization studies (especially the
DRIFT study) and previous studies [3, 5, 11], the CH3CN decomposition mechanism
over the different catalysts were concluded. It is proposed that the CH3CN
decomposition that occurred over HZSM-5 and CuCeOx followed different reaction
pathways. The hydrolysis of CH3CN was the predominant reaction that occurred over
the HZSM-5 sample. During the hydrolysis reaction, the protonation of CH3CN and

water nucleophilic attack proceed [12]. CH3CONH2 was the only hydrolysate in the
first step of CH3CN hydrolysis. Subsequently, CH3CONH2 rapidly hydrolyzed to
form NH3 and CH3COOH. The reaction pathway (hydrolysis pathway) over HZSM-5
could be described as follows:
CH3CN + H2O → CH3CONH2
(5)
(6)
CH3CONH2 + H2O → NH3+ CH3COOH
However, little NH3 emissions were detected during the activity tests at temperatures
below 300 °C. A possible reason for the minor NH3 emissions could be attributed to
the poor redox capacity of HZSM-5 inhibiting the further oxidation of the hydrolysis
products. Therefore, the hydrolysis reactions were limited at low temperatures. At
temperatures higher than 350 °C, the hydrolysis reaction extent increased greatly over
the HZSM-5 sample, in which CH3COOH and NH3 were further oxidized via the
following reaction pathways:
NH3 +O(a) → N2 or NOx + H2O
(7)
(8)
CH3COOH + O(a) → CO2 + H2O
Differing from the hydrolysis process that occurred over the HZSM-5 sample,
CH3CN would first interact with the surface hydroxyl groups on CuCeOx to form
CH3CONH(a) groups [45]. The CH3CONH(a) species could be further hydrolyzed to
ammonia at low temperature. However, as the temperature increased, the hydrolysis
reaction was limited and CH3CONH(a) species were rapidly oxidized to NCO(a)
species which could be further oxidized to form large amounts of NOx. Therefore,
NH3 yield reached a maximum value at 200 °C over CuCeOx and CuCeOx+HZSM-5

samples. In addition, it has been widely accepted that NCO species can react with
NO, O2 or NO+O2 to produce N2 at high temperature [54], which explained that the
amount of NOx formed over CuCeOx increased slowly when the temperature was
increased from 275 to 400 °C. The reaction pathway (oxidation pathway) over
CuCeOx can be described as follows:
CH3CN + OH(a) → CH3CONH(a)
(9)
CH3CONH(a) + O(a) → NCO(a) + CO2 + H2O
(10)
(11)
NCO(a) +O(a) → N2 or NOx + CO2
As for the CuCeOx-HZSM-5 composite catalysts, both hydrolysis and oxidation
reactions would happen. In addition to these two pathways, NOx formed from the
oxidation of NCO(a), which could react with the NH3 formed by the hydrolysis of
CH3CN via an internal SCR reaction, thereby improving the N2 selectivity.
Furthermore, the NCO(a) groups could migrate to HZSM-5 to form HNCO, and
HNCO would be subsequently hydrolyzed into CO2 and NH3 [55], which could also
participate in the internal SCR reaction to reduce NOx. In addition, the NH3-TPSR
results (see Fig. S4) revealed that the introduction of HZSM-5 could inhibit the over-
oxidation of NH3 to form NOx and improve the N2 selectivity. For the composite
catalysts, HZSM-5 could provide abundant Brønsted acid sites. According to
literatures [56, 57], the strong Brønsted acidity could suppress the over-oxidation of
NH3 and thereby increase the accessibility for the reduction of nitrogen oxides.
Therefore, it could be concluded that both internal SCR reaction and selective
catalytic oxidation (SCO) of NH3 to N2 could effectively occur over the composite

catalysts to enhance the N2 selectivity. All the reaction pathways regarding the
enhanced N2 selectivity were listed below:
NCO(a) + H+ → HNCO
(12)
(13)
(14)
(15)
HNCO + H2O → NH3 + CO2
NH3 + NOx + O(a) → N2+ H2O (SCR)
NH3 + O(a) → N2+ H2O (SCO)
The activity tests (see Fig. 1) indicated that CuCe/ZSM-5 maintained a good N2
selectivity in a broad temperature range, while the CuCeOx+HZSM-5 sample
generated some NOx emissions at high temperatures. This fact may be attributed to
the interaction between CuCeOx and HZSM-5 in the physically mixed sample being
weaker compared to that in the impregnation sample. Over the CuCeOx+HZSM-5
sample, NCO(a) may not migrate to the surface of HZSM-5 in time and was rapidly
oxidized to NOx at high temperatures. Furthermore, the better reducibility of this
sample (see the H2-TPR results in Fig. 3) may also led to the over-oxidation of
ammonia to release NOx at high temperatures. Overall, it could be concluded that the
Cu-Ce mixed oxides could ensure the CH3CN catalytic decomposition activity, while
the high redox ability of the Cu-Ce mixed oxides would lead to the over-oxidation of
nitrogen-containing intermediates [NCO(a) and NH3], which would release large
amounts of NOx and lower the N2 selectivity. Additionally, the introduction of HZSM-
5 would enhance the hydrolysis of CH3CN and NCO(a), thereby inhibiting the over-
oxidation of NCO(a) and improving the N2 selectivity via internal SCR reactions. In
addition, the strong acidity of HZSM-5 could also suppress the over-oxidation of NH3

to NOx at an elevated temperature, promoting the NH3 SCO reaction over the
composite catalysts. The relative mechanism of the enhanced N2 selectivity over the
composite catalysts was presented in Fig. 9. Such a mechanism could be further
verified by performing activity tests over the CuCeOx sample that was physically
mixed with HZSM-5 using different SiO2/Al2O3 ratios. As shown in Fig. 10, it was
found that the introduction of HZSM-5 at lower silica contents would lead to less NOx
formation. HZSM-5 zeolites with lower silica contents possessed more Brønsted acid
sites, which would enhance the inhibition effect on NH3 over-oxidation [57]. In
addition, such improved acidity would also promote the hydrolysis of CH3CN and
NCO(a).
4 Conclusion
In this paper, the CH3CN catalytic combustion performances were investigated over
CuCeOx and CuCeOx-HZSM-5 composite catalysts. The catalytic performances
indicated that the CuCeOx-HZSM-5 composite catalysts showed satisfactory catalytic
activities and formed less NOx, and CH3CN could be converted completely at 250 °C
with a NOx yield of less than 10%. Compared to pure CuCeOx sample, it is found that
the introduction of HZSM-5 significantly limited the formation of NOx and improved
the N2 selectivity. In particular, the CuCe/ZSM-5 sample showed an excellent N2
selectivity over a boarder temperature range. Based on the in situ DRIFT analysis, it
could be concluded that the CH3CN decomposition occurred over CuCeOx followed
an oxidation pathway (CH3CN → CH3CONH(a) →NCO(a) groups → N2 /NOx,

CO2), while the CH3CN decomposition that occurred over HZSM-5 proceeded via a
hydrolysis pathway (CH3CN → CH3CONH2 →CH3COOH, NH3). For the CuCeOx-
HZSM-5 composite catalysts, the promotional effect of HZSM-5 introduction on the
N2 selectivity can be explained by an internal SCR mechanism, in which the NOx
derived from the over-oxidation of NCO(a) on CuCeOx could be reduced by the NH3
formed by the CH3CN hydrolysis reaction that occurred over HZSM-5. In addition,
the migration of the NCO(a) intermediate from CuCeOx to HZSM-5 improved the
hydrolysis of HNCO to NH3, which would also inhibit the oxidation of NCO(a).
Moreover, the NH3-TPSR experiment showed that the introduction of strongly acidic
HZSM-5 could also inhibit the over-oxidation of NH3 at high temperatures.
Compared to the physically mixed sample, CuCe/ZSM-5 exhibited a better N2
selectivity owing to the stronger interaction between CuCeOx and HZSM-5, which
was beneficial to the hydrolysis reaction and internal SCR reaction.
Author Contributions
Yuxing Wang: Investigation, Writing- Original draft preparation. Yue Liu:
Supervision, Writing- Reviewing and Editing. Qingji Ying: Investigation. Yaoyu
Zhang: Formal analysis. Zhongbiao Wu: Resources, Writing- Reviewing and Editing.

Acknowledgements
The authors acknowledge for the financial support of National Natural Science
Foundation of China (No. 21577122) and the Program for Zhejiang Leading Team of
S&T Innovation (No. 2013TD07).

Reference
[1] W. Wu, B. Zhao, S. Wang, J. Hao, Journal of environmental sciences. 53 (2017) 224-237.
[2] R. Zhang, D. Shi, N. Liu, B. Chen, L. Wu, L. Wu, W. Yang, Catalysis Today. 258 (2015) 17-27.
[3] T. Nanba, S. Masukawa, J. Uchisawa, A. Obuchi, Journal of Molecular Catalysis A: Chemical.
276 (2007) 130-136.
[4] R. Zhang, P. Li, R. Xiao, N. Liu, B. Chen, Applied Catalysis B: Environmental. 196 (2016) 142-
154.
[5] R. Zhang, D. Shi, N. Liu, Y. Cao, B. Chen, Applied Catalysis B: Environmental. 146 (2014) 79-
93.
[6] H. Zhao, R.G. Tonkyn, S.E. Barlow, B.E. Koel, C.H.F. Peden, Applied Catalysis B:
Environmental. 65 (2006) 282-290.
[7] P. Dagaut, P. Glarborg, M. Alzueta, Progress in Energy and Combustion Science. 34 (2008) 1-46.
[8] G. Karakas, A. Sevinc, Catalysis Today. 323 (2019) 159-165.
[9] C. Du, H. Chen, X. Zhao, H. Zong, X. Zhou, Catalysis Communications. 123 (2019) 27-31.
[10] Y. Hu, J. Liu, J. Cheng, L. Wang, L. Tao, Q. Wang, X. Wang, P. Ning, Applied Surface Science.
427 (2018) 843-850.
[11] O. Kröcher, M. Elsener, Applied Catalysis B: Environmental. 92 (2009) 75-89.
[12] L.A.M.M. Barbosa, R.A. van Santen, Journal of Catalysis. 191 (2000) 200-217.
[13] S. Sun, D. Mao, J. Yu, Z. Yang, G. Lu, Z. Ma, Catalysis Science & Technology. 5 (2015) 3166-
3181.
[14] G. Zhou, H. Lan, T. Gao, H. Xie, Chemical Engineering Journal. 246 (2014) 53-63.

[15] T. Tsoncheva, G. Issa, T. Blasco, M. Dimitrov, M. Popova, S. Hernández, D. Kovacheva, G.
Atanasova, J.M.L. Nieto, Applied Catalysis A: General. 453 (2013) 1-12.
[16] P. Yang, S. Zuo, R. Zhou, Chemical Engineering Journal. 323 (2017) 160-170.
[17] A.C. Akah, G. Nkeng, A.A. Garforth, Applied Catalysis B: Environmental. 74 (2007) 34-39.
[18] T. Zhang, J. Liu, D. Wang, Z. Zhao, Y. Wei, K. Cheng, G. Jiang, A. Duan, Applied Catalysis B:
Environmental. 148-149 (2014) 520-531.
[19] Z. Zhu, G. Lu, Z. Zhang, Y. Guo, Y. Guo, Y. Wang, ACS Catalysis. 3 (2013) 1154-1164.
[20] Q. Dai, W. Wang, X. Wang, G. Lu, Applied Catalysis B: Environmental. 203 (2017) 31-42.
[21] L. Pang, C. Fan, L. Shao, K. Song, J. Yi, X. Cai, J. Wang, M. Kang, T. Li, Chemical Engineering
Journal. 253 (2014) 394-401.
[22] G. Avgouropoulos, T. Ioannides, H. Matralis, Applied Catalysis B: Environmental. 56 (2005) 87-
93.
[23] T. Caputo, L. Lisi, R. Pirone, G. Russo, Applied Catalysis A: General. 348 (2008) 42-53.
[24] C. Wang, Q. Cheng, X. Wang, K. Ma, X. Bai, S. Tan, Y. Tian, T. Ding, L. Zheng, J. Zhang, X. Li,
Applied Surface Science. 422 (2017) 932-943.
[25] S. Fan, J. Xue, T. Yu, D. Fan, T. Hao, M. Shen, W. Li, Catalysis Science & Technology. 3 (2013)
2357.
[26] W. Huang, Y. Liu, Z. Wu, Catalysis Communications. 122 (2019) 58-62.
[27] I. Platzman, R. Brener, H. Haick, R. Tannenbaum, Journal of Physical Chemistry C. 112 (2008)
1101-1108.
[28] M. Piumetti, S. Bensaid, T. Andana, N. Russo, R. Pirone, D. Fino, Applied Catalysis B:
Environmental. 205 (2017) 455-468.

[29] P. Liu, E.J. Hensen, Journal of the American Chemical Society. 135 (2013) 14032-14035.
[30] M. Konsolakis, Applied Catalysis B: Environmental. 198 (2016) 49-66.
[31] R. Leppelt, B. Schumacher, V. Plzak, M. Kinne, R. Behm, Journal of Catalysis. 244 (2006) 137-
152.
[32] D. Mrabet, A. Abassi, R. Cherizol, T.-O. Do, Applied Catalysis A: General. 447-448 (2012) 60-66.
[33] Q. Zou, Y. Zhao, X. Jin, J. Fang, D. Li, K. Li, J. Lu, Y. Luo, Applied Surface Science. 494 (2019)
1166-1176.
[34] T. Gu, Y. Liu, X. Weng, H. Wang, Z. Wu, Catalysis Communications. 12 (2010) 310-313.
[35] X. Dong, W. Shen, Y. Zhu, L. Xiong, J. Shi, Advanced Functional Materials. 15 (2005) 955-960.
[36] P. Dutta, S. Pal, M.S. Seehra, Y. Shi, E.M. Eyring, R.D. Ernst, Chemistry of Materials. 18 (2006)
5144-5146.
[37] G. Busca, T. Montanari, M. Bevilacqua, E. Finocchio, Colloids and Surfaces A: Physicochemical
and Engineering Aspects. 320 (2008) 205-212.
[38] C.-C. Chuang, W.-C. Wu, M.-X. Lee, J.-L. Lin, Physical Chemistry Chemical Physics. 2 (2000)
3877-3882.
[39] M.A. Hasan, M.I. Zaki, L. Pasupulety, Applied Catalysis A: General. 243 (2003) 81-92.
[40] L. Wang, W. Li, G. Qi, D. Weng, Journal of Catalysis. 289 (2012) 21-29.
[41] R.Q. Long, R.T. Yang, Journal of Catalysis. 194 (2000) 80-90.
[42] L. CHEN, J. LI, M. GE, Environmental science & technology. 44 (2010) 9590-9596.
[43] A. ABOULAYT, C. BINET, J. LAVALLEY, Journal of The Chemical Society-faraday
Transactions. 91 (1995) 2913-2920.
[44] L.F. Liao, C.F. Lien, D.L. Shieh, F.C. Chen, J.L. Lin, Physical Chemistry Chemical Physics. 4

(2002) 4584-4589.
[45] M. Tamura, A. Satsuma, K.-i. Shimizu, Catalysis Science & Technology. 3 (2013) 1386.
[46] Y. Yeom, B. Wen, W. Sachtler, E. Weitz, Journal Of Physical Chemistry B. 108 (2004) 5386-5404.
[47] R. Jin, Y. Liu, Z. Wu, H. Wang, T. Gu, Chemosphere. 78 (2010) 1160-1166.
[48] W. Yao, X. Wang, Y. Liu, Z. Wu, Applied Surface Science. 467-468 (2019) 439-445.
[49] G. Zhou, H. Liu, K. Cui, A. Jia, G. Hu, Z. Jiao, Y. Liu, X. Zhang, Applied Surface Science. 383
(2016) 248-252.
[50] G. RAMIS, G. BUSCA, V. LORENZELLI, Materials Chemistry and Physics. 29 (1991) 425-435.
[51] L. Zhang, H. He, Journal of Catalysis. 268 (2009) 18-25.
[52] S.M. Lee, S.C. Hong, Applied Catalysis B: Environmental. 163 (2015) 30-39.
[53] F. SOLYMOSI, T. BANSAGI, Journal of Catalysis. 156 (1995) 75-84.
[54] H. He, C. Zhang, Y. Yu, Catalysis Today. 90 (2004) 191-197.
[55] H.H. Ingelsten, R. Matarrese, M. Skoglundh, Topics in Catalysis. 52 (2009) 1921-1924.
[56] M. Jabłońska, A. Król, E. Kukulska-Zajac, K. Tarach, L. Chmielarz, K. Góra-Marek, Journal of
Catalysis. 316 (2014) 36-46.
[57] K. Góra-Marek, K.A. Tarach, Z. Piwowarska, M. Łaniecki, L. Chmielarz, Catalysis Science &
Technology. 6 (2016) 1651-1660.