Direct activation of CH4 to oxygenates and unsaturated hydrocarbons using N2O on Fe-modified zeolites

Article abstract of DOI:10.1016/j.molcata.2016.11.008

Direct activation of CH₄ to oxygenates and unsaturated light hydrocarbons was investigated using Fe-modified ZSM-5 and Ferrierite (FER) for a partial oxidation of CH₄ with N₂O oxidant. The amount of active α-oxygen sites and number of Bronsted acid sites on the Fe-modified zeolites were well correlated with CH₄ conversion rate and product distributions. The amount of α-oxygen sites was largely changed according to preparation method such as wet impregnation or ion-exchange of iron precursor and types of zeolites. A large number of Bronsted acid sites and α-oxygen sites on the Fe-modified FER revealed a higher oxygenate formation such as methanol and dimethyl ether (DME) with COx, and a larger number of strong acid sites on Fe-modified ZSM-5 was also responsible for a higher selectivity to light hydrocarbons by a successive dehydration of oxygenates formed. The different catalytic performances were verified through proper measurements of the amount and type of acidic sites as well as the α-oxygen sites measured by N₂O pulse chemisorption. The Fe-modified FER prepared by impregnation method possessed a larger amount of α-oxygen sites due to abundant Bronsted acid sites, which was responsible for a higher rate of CH₄ conversion to oxygenates with the help of N₂O decomposition on the α-oxygen sites originated from iron oxide nanoparticles.

Full text of DOI:10.1016/j.molcata.2016.11.008

Accepted Manuscript
Title: Direct activation of CH4 to oxygenates and unsaturated
hydrocarbons using N2O on Fe-modified zeolites
Author: Kyung Soo Park Jeong Hwa Kim So Hyun Park
Dong Ju Moon Hyun-Seog Roh Chan-Hwa Chung Soong Ho
Um Joon-Hwan Choi Jong Wook Bae
PII:
S1381-1169(16)30482-4
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2016.11.008
MOLCAA 10105
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
30-7-2016
29-10-2016
5-11-2016
Please cite this article as: Kyung Soo Park, Jeong Hwa Kim, So Hyun Park, Dong Ju
Moon, Hyun-Seog Roh, Chan-Hwa Chung, Soong Ho Um, Joon-Hwan Choi, Jong
Wook Bae, Direct activation of CH4 to oxygenates and unsaturated hydrocarbons
using N2O on Fe-modified zeolites, Journal of Molecular Catalysis A: Chemical
http://dx.doi.org/10.1016/j.molcata.2016.11.008
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. 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.

Submitted to the Journal of Molecular Catalysis A: Chemical
Direct activation of CH4 to oxygenates and unsaturated hydrocarbons
using N2O on Fe-modified zeolites
Kyung Soo Park1, Jeong Hwa Kim1, So Hyun Park1, Dong Ju Moon2, Hyun-Seog Roh3,
Chan-Hwa Chung1, Soong Ho Um1, Joon-Hwan Choi4, Jong Wook Bae1,*
1School of Chemical Engineering, Sungkyunkwan University (SKKU), Suwon, Gyeonggi-do,
440-746, Republic of Korea
2Clean Energy Research Center, Korea Institute of Science and Technology (KIST), 136-791
Seoul, Republic of Korea
3Department of Environmental Engineering, Yonsei University, 1 Yonseidae-gil, Wonju,
Gangwon 220-710, Republic of Korea
4Functional Ceramics Department, Korea Institute of Materials Science (KIMS), Changwon,
Gyeongnam 642-831, Republic of Korea
*Corresponding author (J.W. Bae): Tel.: +82-31-290-7347; Fax: +82-31-290-7272; E-mail
address: finejw@skku.edu

Graphical abstract
HIGHLIGHTS
 Direct activation of CH4/N2O to oxygenates was investigated on the Fe-modified
zeolites.
 Amount of α-oxygen and Bronsted acid sites significantly changed the product
distribution.
 Fe/FER showed a high activity due to a larger amount of α-oxygen and acidic sites.

ABSTRACT
Direct activation of CH4 to oxygenates and unsaturated light hydrocarbons was investigated
using Fe-modified ZSM-5 and Ferrierite (FER) for a partial oxidation of CH4 with N2O
oxidant. The amount of active α-oxygen sites and number of Bronsted acid sites on the Fe-
modified zeolites were well correlated with CH4 conversion rate and product distributions.
The amount of α-oxygen sites was largely changed according to preparation method such as
wet impregnation or ion-exchange of iron precursor and types of zeolites. A large number of
Bronsted acid sites and α-oxygen sites on the Fe-modified FER revealed a higher oxygenate
formation such as methanol and dimethyl ether (DME) with COx, and a larger number of
strong acid sites on Fe-modified ZSM-5 was also responsible for a higher selectivity to light
hydrocarbons by a successive dehydration of oxygenates formed. The different catalytic
performances were verified through proper measurements of the amount and type of acidic
sites as well as the α-oxygen sites measured by N2O pulse chemisorption. The Fe-modified
FER prepared by impregnation method possessed a larger amount of α-oxygen sites due to
abundant Bronsted acid sites, which was responsible for a higher rate of CH4 conversion to
oxygenates with the help of N2O decomposition on the α-oxygen sites originated from iron
oxide nanoparticles.
Key words: CH4 activation; N2O oxidant; Fe-modified zeolites; Acid sites; α-oxygen sites.

1. Introduction
A direct utilization of CH4, which is a main component of natural gas (or shale gas) and one
of important greenhouse gases (GHG), to useful chemicals or fuels has been widely studied
due to an increased demand of clean fuels from alternative resources [1] and to reduce the
emission of GHG. Compared to an indirect conversion of natural gas to useful chemicals
through various syngas production processes, a partial oxidation of CH4 with oxidants such as
oxygen or N2O to oxygenates as basic petrochemical feedstocks like methanol, formaldehyde,
dimethyl ether (DME) or formic acid seems to be one of the most attractive methods due to
its simple process scheme for direct conversion of CH4 [2-3]. Among those oxygenates, DME
was considered as one of clean fuels owing to its lower emission of air-pollutants by
producing 31,350 kJ/kg of heat [4], which also has similar physical properties with liquefied
petroleum gas (LPG). However, some drawbacks of a direct partial oxidation of CH4 to
oxygenates should be previously solved for commercial application by efficiently prohibiting
thermodynamically favorable and competitive combustion reaction by consecutive reaction
of formed oxygenated intermediates to fully combusted unuseful COx (CO and CO2). To
solve these shortcomings, many catalytic systems for direct methane activations have been
tested by using the active molybdate, vanadate or iron phosphate species supported on the
(meso)porous silica, which was applied for the production of formaldehyde mainly [5-7]. In
addition, some transition metal oxides or metal loaded zeolites have been also extensively
investigated due to their bifunctional activity of metallo-zeolite, which have not only solid
acid functionalities of the zeolite but also active sites for the activation of thermodynamically
stable light hydrocarbons to oxygenates or aromatic components [8-15]. Recently, substantial
attentions have been paid to develop proper catalytic systems by using transition metals such
as Fe, Co or Cu incorporated on the solid-acid ZSM-5 since these catalysts showed a superior
activity for a direct oxidation of CH4 to oxygenates even at a relatively low temperature with

the help of soft oxidants such as N2O and H2O2 [16-20]. Among them, Fe-ZSM-5 has been
widely investigated due to its higher activity for selective catalytic reduction (SCR) of N2O
pollutant with ammonia or light paraffins, which was known as one of the important
greenhouse gases emitted from automobiles [21-23]. In general, iron-incorporated zeolites
have been generally reported to have active sites for a selective N2O decomposition.
Although the structures of the active iron sites named as α-sites are still on debate, it has been
known to be formed as the reduced form after high temperature treatment [24]. While
decomposing N2O molecules on the active α-sites, two of iron species seem to be bridged
with adjacent oxygen atom originated from N2O to form Fe-O-Fe diiron complex (known as
α-oxygen sites) through equation (1). These α-oxygen sites are also known to have reactive
natures which are capable to oxidize the stable hydrocarbons such as methane and benzene to
methanol and phenol respectively with reversible redox properties by the following equation
(2) [25-27]. Based on the mechanisms, an activation of CH4 molecules on α-oxygen sites (Oα)
can proceed by a hydrogen abstraction mechanism with high reaction rate even at an ambient
temperature. The fragments formed by an activated CH4 have been reported to form methoxy
and hydroxyl groups while reacting with α-oxygen sites with a stoichiometric ratio of CH4:Oα
= 1:2 by the equation (2) [28]. A rate determining step has been known to be desorption step
of products [19,28], and an amount of active α-oxygen sites originated from Fe-O-Fe diiron
complexes has been also closely related with the amount of surface Bronsted acid sites on the
solid-acid zeolites as well [24,29,30].
(1) (FeII)α + N2O → (FeIII–O.-)α + N2
(2) CH4 + 2 (FeIII–O.-)α → (FeIII–OCH3)α + (FeIII–OH)α
In the present study, a direct synthesis of useful oxygenates such as methanol and DME as

well as C2 – C3 light hydrocarbons through an oxidative dehydrogenation of CH4 using N2O
oxidant was reported using Fe-modified zeolites to verify the roles of α-oxygen sites as well
as Bronsted acid sites, where the hybrid Fe-ferrierite catalytic systems have been scarcely
investigated in terms of product distributions as far as we know. The pristine ZSM-5 and
ferrierite (FER) zeolites were modified using iron precursor through two different preparation
methods such as a simple wet impregnation and ion-exchange in an aqueous solution to
elucidate the roles of amount of surface α-oxygen sites as well as Bronsted acid sites. The
number of surface Bronsted acid sites and amount of active α-oxygen sites on Fe-modified
zeolites simultaneously altered CH4 conversion rate and product distribution of oxygenates
and light hydrocarbons.
2. Experimental
2.1. Catalyst preparations and activity measurements
A solid acid zeolite of NH4-form FER was purchased from Zeolyst (Si/Al ratio of 9, CP-
914C), and H-form ZSM-5 was provided from Zeocat (Si/Al ratio of 28, PZ-50H). NH4-form
FER was previously calcined at 500 oC for 3 h under air atmosphere to make H-form FER. To
prepare Fe-modified zeolites (Fe/zeolites) such as the Fe/ZSM-5 and Fe/FER, the powder-
type zeolite was previously dispersed in deionized water, and a desired amount of iron
precursor of Fe(NO3)3•9H2O (Samchun, 98.5% purity) was added in the slurry at a fixed
7wt%Fe based on total weight of zeolite. After thoroughly mixing the above slurry in a rotary
o
evaporator for 1 h, vacuum evaporation was carried out at 60 C to prepare impregnated
Fe/zeolites. After drying overnight at 110 oC, the sample was calcined at 550 oC for 6 h under
air atmosphere with a heating rate of 1 oC/min. As-prepared catalysts are denoted as Fe/ZSM-
5(IMP) and Fe/FER(IMP), respectively. In addition, a liquid phase ion-exchange of the iron
precursor on H-form ZSM-5 and FER was also applied to prepare Fe-modified zeolites with a

fixed amount of Fe(NO3)3•9H2O precursor in an aqueous solution by vigorously stirring at 80
o
oC for 24 h. After washing and filtering, the sample was further dried at 110 C in an oven
overnight followed by calcining at 550 oC for 6 h in air environment with a heating rate of 1
oC/min. After the calcination, the powder was crushed in the particle size less than ~100 μm
to avoid an internal diffusion limitation on general heterogeneous catalysts [31-32]. The as-
prepared Fe/zeolites are denoted as the Fe/ZSM-5(IE) and Fe/FER(IE).
Catalytic activity for a direct CH4 oxidation to oxygenates and hydrocarbons was evaluated
in a fixed-bed tubular reactor with an outer diameter of 9 mm. Prior to activity test, 0.1 g of
Fe/zeolite was loaded in the reactor, and reduced under a flow of 5vol%H2 balanced with N2
at 700 oC for 3 h. After the reduction, the temperature was cooled down to room temperature,
and reactants with a molar ratio of CH4/N2O = 2.5 were fed into the reactor with an internal
o
standard gas of 5vol% He. The reaction was carried out at T = 280 C, P = 0.1 MPa, and
space velocity (SV) = 12000 L(total gas)/(kgcat∙h) for 10 h, and the catalytic activity at
steady-state after 5 h on stream was compared using an average values. Additional activity
test for a methanol conversion to oxygenates and hydrocarbons was further carried out in a
same fixed-bed tubular reactor with 0.1 g catalyst to verify activity and product distribution
of the conversion of methanol intermediates formed by direct CH4 activation from CH4/N2O.
o
Prior to the activity test, the as-prepared catalyst was reduced at 700 C for 3 h under a flow
of 5vol%H2/N2. Catalytic activity for methanol conversion was measured for 5 h using the
reactant of 10vol%N2O/He gas at a liquid methanol feed rate of 0.013 mL/min at the reaction
o
conditions of T = 290 C, P = 0.1 MPa, and SV = 12000 L/(kgcat∙h). The effluent gases from
the reactor were measured continuously by using a gas chromatograph (YL6500, Younglin).
Using a thermal conductivity detector (TCD) connected with a Carboxen 1000 packed
column, CH4, N2, H2, COx, and N2O were analyzed. The formed hydrocarbons and all
oxygenates by direct CH4 activation were also analyzed by using a flame ionization detector

(FID) connected with a HP PLOT Q capillary column. The rate of N2O conversion was
calculated using the variation of reactant concentration based on an internal standard gas of
He, and the rate of CH4 conversion and product distributions at steady-state for last 5 h on
stream was calculated based on total carbon balance.
2.2. Catalyst characterization
Powder X-ray diffraction (XRD) analysis of the fresh and used Fe/zeolites was performed to
verify the crystalline phases and variations of crystallinity before and after reaction using
X’Pert3 Powder (PANalytical) operated at 40 kV and 100 mA with Cu-Kα radiation of
0.15406 nm in the range of 5 - 35o at a scanning rate of 4o/min. The elemental composition of
Fe and Si/Al molar ratio on the fresh Fe/zeolites were also determined by X-ray fluorescence
(XRF) analysis using Bruker S4 operated at 60 kV and 150 mA. The surface morphology and
variation of crystallite sizes of iron oxides on the fresh and used Fe/FER(IMP) and
Fe/FER(IE) were further characterized using transmission electron microscopy (TEM) with a
TECNAI G2 instrument operated at an accelerating voltage of 200 kV.
Temperature programmed reduction with hydrogen (H2-TPR) was carried out using
BELCAT-M instrument to characterize reduction behaviors of the Fe/zeolites. For the TPR
analysis, 10vol%H2 balanced with Ar was used with 30 mg fresh sample which was
o
previously treated with He at a flow rate of 30 mL/min at 500 C for 1 h to remove any
contaminants and adsorbed water. After the pretreatment, the temperature was cooled down to
100 oC and kept for 0.5 h, and the temperature was subsequently increased up to 800 oC with
o
a ramping rate of 10 C/min. The effluent gases were passed through a molecular sieve to
remove water formed during TPR analysis and the amount of hydrogen consumption was
recorded using TCD. Temperature programmed desorption of ammonia (NH3-TPD) was
carried out using the same BELCAT-M instrument to characterize the amount of acid sites on

the fresh Fe/zeolites. Before the NH3-TPD analysis, the sample was heated up to 250 oC with
a ramping rate of 10 oC/min and maintained at the same temperature for 2 h to eliminate any
contaminants and water adsorbed on the catalyst surfaces. After the pretreatment, the sample
o
was cooled down to 100 C, and NH3 gas with a purity of 99.999% was introduced on the
sample for 1 h at a flow rate of 50 mL/min followed by purging for 1.5 h to remove
physisorbed NH3 molecules. The sample was analyzed by heating from 100 to around 600 oC
at a ramping rate of 10 oC/min, and the amount of desorbed NH3 was analyzed by using TCD.
To compare relative amount of Bronsted and Lewis acid sites on the Fe/zeolites, Fourier-
transform infrared (FT-IR) spectroscopy analysis of an adsorbed pyridine (Py-IR) was carried
out using FT-IR spectrometer of Nicolet 6700 (Thermo fisher scientific). For the Py-IR
analysis, 30 mg of fresh sample was previously pelletized to a self-supported thin pellet, and
o
the temperature of FT-IR cell was raised up to 150 C in a vacuum atmosphere at a ramping
o
rate of 15 C/min to remove the adsorbed water and any contaminants. After the cell
temperature was stabilized at 150 oC, FT-IR background spectrum of the sample was taken at
a scanning speed of 32 scans/s with a resolution of 4 cm-1. After the collection of background
spectrum, 3 μL of pyridine was injected into the static vacuum system by using microsyringe
for the adsorption of pyridine to the pretreated sample pellet for 15 min. The physisorbed and
o
residual pyridine molecules was evacuated at 150 C for 0.5 h, and the Py-IR spectra of the
chemisorbed pyridine molecules were obtained on the fresh Fe/zeolites.
N2O titration was further carried out to measure the amount of active α-oxygen sites (Oα)
using AutoChem II 2920 instrument (Micrometrics). For the N2O titration analysis, 100 mg
o
sample was previously reduced at 700 C for 3 h under 10vol%H2 balanced with Ar, and
o
cooled down to 280 C. The pulses of 10 vol%N2O balanced with He were continuously
introduced up to a full saturation of active α-oxygen sites on the Fe/zeolites. The consumed
amount of N2O was calculated from the measured amount of the formed N2 using TCD.

Finally, the amount of the α-oxygen sites on the reduced Fe/zeolites was evaluated from the
accumulated amount of N2O consumption using a stoichiometry of N2O/Oα = 1.0.
Temperature programmed surface reaction (TPSR) of the used Fe/zeolites was carried out to
verify the amount of deposited coke precursors by using a BELCAT-M connected with gas
chromatography-mass spectrometer (GC-MS) (YL6900 GC/MSD, Younglin) equipped with a
deactivated fused silica capillary blank column. A 30mg of the used sample was loaded and
o
subsequently pretreated with 30 mL/min of He at 500 C for 1 h. After the pretreatment, the
o
sample was cooled down to 50 C and 10vol% H2 balanced with Ar was introduced by
o
o
increasing temperature up to 700 C with a ramping rate of 15 C/min for the complete
hydrogenation of the formed coke precursors to CH4. The effluent gases were passed through
molecular sieve for eliminating the formed H2O. The MS signals were collected in range of
amu = 2-50 per second, and the selected m/z values assigned to methane fragments (m/z = 14
and 15) was collected to exclude the MS signals from H2O fragments.
3. Results and Discussion
3.1. Physicochemical properties of the Fe-modified zeolites
The elemental compositions of the fresh Fe/zeolites measured by XRF are summarized in
Table 1. The Fe/zeolites prepared by impregnation method showed similar iron contents with
the values of 7.3 and 7.8wt% on the fresh Fe/FER(IMP) and Fe/ZSM-5(IMP), respectively.
However, since a larger amount of acid sites exists on the FER due to a lower Si/Al ratio of 9
than that of ZSM-5 of 28, the composition of iron species was found to be slightly higher on
the ion-exchanged Fe/FER than that of Fe/ZSM-5 with an iron content of 6.4wt% on the
Fe/FER(IE) and that of 4.8wt% on the Fe/ZSM-5(IE). The Si/Al ratio after Fe modification of
the zeolites was also found to be similar to parent zeolites in both FER and ZSM-5. With the
help of NH3-TPD and TEM analyses in following section, the dispersion of iron species on

the Fe/zeolites seems to be higher on the ion-exchanged Fe/ZSM-5(IE) and Fe/FER(IE) than
other Fe/zeolites, which was attributed to the adsorption character of iron precursor on the
Bronsted acid sites selectively, especially on the Fe/FER(IE) having a larger amount of
Bronsted acid sites as summarized in Table 1.
XRD patterns of the Fe/zeolites before and after reaction are displayed in Figure 1 and the
characteristic diffraction peaks of bare ZSM-5 and FER are also displayed in supplementary
Figure S1 which showed similar diffraction patterns with previously reported ones [33-35].
The each crystalline structure of the zeolites was preserved even after Fe modification and
reaction at 280 oC for 10 h. However, the characteristic diffraction peaks of iron species were
not clearly observed due to its higher dispersion on the ZSM-5 and FER possibly. In addition,
the XRD peak intensities of the characteristic diffraction peaks on the bare zeolites were
decreased after Fe modification, especially on the Fe/ZSM-5(IMP) and Fe/FER(IMP), due to
a larger amount of iron concentration as shown in Figure 1(a). Interestingly, the
characteristic diffraction peaks of the bare ZSM-5 and FER on the used Fe/zeolites were
preserved with relatively lower peak intensity as shown in Figure 1(b). This observation
suggests a possible structural deformation on the Fe/zeolites coming from a high temperature
o
reduction at 700 C or deposition of coke precursors during the reaction. These significant
decreases of the characteristic peak intensities of the FER and ZSM-5 can be also attributed
to a slight dissolution of aluminosilicate crystals in an acid iron nitrate precursor solution as
well [36]. In addition, the relative crystallinity of the fresh and used Fe/zeolites was further
calculated by comparing integrated peak areas of the characteristic diffraction peaks in the
range of 6 - 10o for the Fe/ZSM-5 and 8 - 10o for the Fe/FER. The variations of the
crystallinity of the Fe/zeolites also revealed a large deformation of the crystallinity of zeolites
after reaction. As summarized in Table 1, the decreased crystallinity of the Fe/zeolites before
and after reaction was found to be more significant on the Fe/ZSM-5 by showing the relative

crystallinity from 35 to 15 on the Fe/ZSM-5(IMP), 36 to 18 on the Fe/ZSM-5(IE) than those
of the Fe/FER from 44 to 30 on the Fe/FER(IMP) and 72 to 37 on the Fe/FER(IE),
respectively. A large decrease of crystallinity on the Fe/ZSM-5 during the reaction can be
attributed to the preferentially deposition of iron nanoparticles on the outer surfaces of the
ZSM-5 due to a smaller amount of acid sites than the FER, which also caused the structural
deformation of zeolites. The variation of relative crystallinity of the Fe/zeolites after reaction
was found to be in the orders of Fe/FER(IMP) < Fe/FER(IE) < Fe/ZSM-5(IE) < Fe/ZSM-
5(IMP) with the respective values of 31.8, 48.6, 50, and 57.1, which suggests the significant
structural collapses of the Fe/zeolites after reaction especially on the Fe/ZSM-5 and the
trends are related with the coke deposition as discussed in the following TPSR results.
Since the present ZSM-5 having a higher Si/Al ratio than that of FER possesses a smaller
amount of acid sites with a lower ion-exchange capacity, an excess amount of iron species
can be preferentially deposited on the outer surfaces of the ZSM-5 in a form of iron oxide
nanoparticles [37,38]. These phenomena can be also responsible for a decreased catalytic
stability of the zeolite structures under the reaction condition due to a less possibility of metal
incorporation in the frameworks of zeolites [39], especially on the Fe/ZSM-5. The relative
crystallinity observed on the used Fe/ZSM-5(IMP) and Fe/ZSM-5(IE) with the respective
values of 15 and 18 was more significant than those of the used Fe/FER with the respective
values of 30 and 37 on the used Fe/FER(IMP) and Fe/FER(IE). Therefore, the Fe/ZSM-5 and
Fe/FER prepared by ion-exchange method showed a lower content of iron species than those
of the impregnation method because the iron precursor can be heterogeneously deposited on
the zeolite surfaces through an impregnation method. The Fe/ZSM-5(IE) revealed a less
amount of Fe loading compared to the Fe/FER(IE) due to a lack of ion exchanging capacity
with the protons attached on the Bronsted acid sites of ZSM-5 compared to FER. Therefore,
the deposited iron species seems to exist in cationic forms of iron species on the acid sites of

zeolites with a partial formation of crystalline iron oxide nanoparticles on the outer surface of
zeolites, and the size of iron nanoparticles seems to be much smaller on the Fe/zeolites
prepared by ion-exchange method which can be hardly detected by XRD analysis due to its
small particle size or overlapping with zeolite structures.
To verify the reduction behaviors of the Fe/zeolites, H2-TPR analysis was performed and the
o
reduction patterns are displayed in Figure 2. Hydrogen was consumed from 200 C and the
o
distinctive reduction peaks of iron oxides were observed at around 290 and 450 C with a
small peak at around 650 oC. All reduction patterns of the Fe/zeolites revealed the presence of
different iron species on the surfaces through different interactions. Since iron (III) nitrate
precursor was used for the modification of zeolites having the character of facile formation of
Fe3+ cations after decomposition, the Fe3+ species can exist in a cationic form or Fe3+-OH and
stabilized Fe2+ species on the acid sites [40]. Moreover, an ion-exchanged cationic iron
species can be easily aggregated on the acid sites to form iron oxide nanoparticles during a
calcination step as well [41]. Therefore, various states of iron species such as isolated Fe
cationic monomers, dimers, oligomers as iron oxo species or iron oxide nanoparticles can be
possibly formed due to an excess usage of iron precursor above an ion-exchange capacity of
the ZSM-5 or FER zeolites, which were responsible for the observation of multiple reduction
peaks on the Fe/zeolites [41-44]. As shown in Figure 2, all Fe/zeolites showed a shoulder
o
reduction peak in the range of 200 - 300 C which was attributed to the reduction of the
cationic iron species [40]. The larger peak intensity below 300 oC on the Fe/FER compared to
the Fe/ZSM-5 was attributed to a larger ion exchange capacity on the Fe/FER having a larger
amount of acid sites, especially on the Fe/FER(IMP). In addition, the Fe/ZSM-5 showed a
smaller intensity of reduction peaks than those of the Fe/FER and the distinguishable peaks at
around 400 - 600 oC can be attributed to the reduction of iron oxide nanoparticles formed on
the outer surfaces of zeolites [39,45], especially large on the Fe/FER(IMP) and lower

reduction temperature on the Fe/FER(IE). Therefore, the Fe/FER(IMP) showing larger
reduction peak intensity compared to the Fe/ZSM-5 revealed the well dispersion of iron
nanoparticles on the abundant acid sites of FER with a relatively stronger interaction than
Fe/FER(IE). The Fe/FER(IMP) was preferentially reduced in a wide range of temperatures
o
than the Fe/ZSM-5 without showing a higher reduction peak above 650 C, and it also leads
to a possible formation of homogeneously distributed iron oxide nanoparticles. In terms of
the formation of α-oxygen sites on the Fe/zeolites, the ion-exchanged Fe3+ species adsorbed
on the Bronsted acid site selectively [40] and the aggregated iron oxide nanoparticles can
suppress amounts of active α-oxygen sites, which seems to be dominant on the Fe/FER(IMP).
Based on the results of TPR and surface acidity measurements, the ion-exchanged cationic
iron species on the Bronsted acid sites and abundant formation of well-dispersed iron oxide
nanoparticles were responsible for a preferential formation of active α-oxygen sites as well,
which are in line with the results of N2O titration and NH3-TPD in the following section.
3.2. Catalytic activities on the Fe-modified zeolites
The rates of CH4 and N2O conversion with product distribution at steady-state using average
values are summarized in Table 2 and Figure 3. Although conversion of CH4 was observed
below 1% with a maximum conversion on the Fe/FER(IMP) due to a difficult activation of
CH4 in our reaction conditions, all Fe/zeolites revealed significantly different product
distributions. Catalytic activity in a gas-phase on the Fe/FER was found to be higher than that
of the Fe/ZSM-5 irrespective to the preparation methods. The rate of CH4 conversion was
found to be 3.11 and 2.42 mmol/(gcat∙h) on the Fe/FER(IMP) than Fe/FER(IE), respectively.
The highest rate of N2O conversion was also observed on the Fe/FER(IMP) with a value of
5.40 mmol/(gcat∙h), and the variations of the rates of CH4 and N2O conversion as shown in
Figure 3 were not significant during 10 h on stream on all Fe/zeolites catalysts which also

supports the insignificant deactivation of the catalysts. However, the significant changes of
product distributions were observed by showing a higher selectivity to oxygenates (mainly
methanol and DME with insignificant byproducts) and COx on the Fe/FER and by revealing
a higher selectivity to unsaturated C2-C3 hydrocarbons and oxygenates on the Fe/ZSM-5. For
example, the selectivity of CO with a value of 48.9% and that of oxygenates with 47.6%
(DME selectivity of 27.9%) were observed on the Fe/FER(IMP) and the similar values were
observed on Fe/FER(IE). However, a higher selectivity to light olefinic hydrocarbons such as
ethylene and propylene on the Fe/ZSM-5 was observed on the Fe/ZSM-5(IMP) with the
value of 42.2% and a large amount of oxygenates of 57.8% (DME selectivity of 37.6%) was
also observed instead of oxidized COx products with the similar values on the Fe/ZSM-5(IE).
Small amount of hydrogen was also produced from this reaction, and H2/CO ratio was found
to be around 0.20, which is far from a theoretical value of 2 for the possible methanol steam
reforming due to a small amount of H2O formed on the Fe/FER(IMP). Based on the observed
lower H2/CO ratio on the most active Fe/FER(IMP), CO was not mainly formed from the
steam reforming of methanol product and it seems to be originated from a direct combustion
of methane possibly with the help of active α-oxygen species.As shown in Figure 3, all of the
catalysts were deactivated slowly during the reaction of 10 h with increasing the selectivity to
oxygenates and decreasing hydrocarbons and COx products due to the possible aggregation
of iron nanoparticles during the reaction.
In addition, the catalytic activity changes of the Fe/FER(IMP) at different space velocities
are summarized in Table S1 and Figure S2 with time on stream. By increasing space
velocities from 6000 to 18000 L/(kgcat·h), the rate of CH4 and N2O conversions were
decreased from 6.23 to 2.27 mmol/(gcat∙h) and from 10.42 to 4.20 mmol/(gcat∙h) due to a short
residence time in a catalyst-bed. However, the oxygenates including DME and methanol were
gradually increased from 22.7 to 50.5% with an increase of space velocity by simultaneously

decreasing the selectivity to hydrocarbons and combusted COx products due to a facile
oxidation of intermediates at a longer residence time selectively on the active α-oxygen sites.
The effects of reaction temperatures to catalytic activity were also investigated on the
Fe/FER(IMP) at different temperatures from 240 to 360 oC and the results are summarized in
Table S2 and Figure S3 with time on stream (h). As the temperatures were increased, the rate
of conversions of CH4 and N2O were dramatically increased from 0.18 to 48.20 mmol/(gcat∙h)
and from 1.22 to 145.38 mmol/(gcat∙h), respectively. The combustion reaction of CH4 to form
o
COx products was preferential at high temperature above 300 C. At a lower temperature of
240 oC, even though the amount of reacted methane was quiet around 0.18 mmol/(gcat∙h), the
oxygenates including DME and methanol (95.7%) and hydrocarbons (4.3%) were mainly
o
formed without COx formation. Above the temperature of 260 C, CO was formed with the
selectivity of 55.0% based on total carbons and the selectivity to oxygenates was dramatically
decreased to 42.4%. At a higher temperature above 300 oC, CO2 was a main oxidized product
and COx selectivity was increased to 62.2% whereas the selectivity to oxygenates was
suppressed to 32.7% with the significantly enhanced rates of CH4 and N2O conversion with
the values of 7.18 to 14.21 mmol/(gcat∙h). The formed hydrocarbons were found to be mainly
C2-C3 olefins, which seems to be attributed to the active α-oxygen species originated from a
radical decomposition of N2O through a possible oxidative coupling of methane. At the much
o
higher temperature of 360 C, the fully combusted COx selectivity was increased to 88.3%
with a lower selectivity to hydrocarbons and oxygenates. The effects of feed composition of
CH4/oxidant were also reported by using nonthermal plasma at ambient conditions [46,47],
and the direct methane conversion to oxygenates was largely altered according to molar ratios
of CH4/N2O at fixed flow rate of the reactants. By increasing a molar composition of N2O,
the conversion of CH4 was increased while that of N2O was decreased and the selectivity to
oxygenates such as methanol and formaldehyde was decreased whereas the amount of COx

was increased dramatically as well. Therefore, the abundant presence of active α-oxygen
species with a relatively higher decomposition rate compared with those of CH4 can be
responsible for a higher selectivity to fully oxidized COx formation instead of oxygenates.
From the results of N2O titration as summarized in Table 1, the rate of N2O conversion was
not directly proportional to total number of α-oxygen sites on the Fe/zeolites, however, the
consumed amount of N2O was related to the number of those sites, which showed a higher
value on the Fe/FER than Fe/ZSM-5. During the reaction, the reactive oxygen species by
N2O decomposition can be adsorbed on the partially reduced iron nanoparticles which were
named as α-oxygen sites. Since this α-oxygen can activate the adsorbed CH4 molecule, CH4
can be dissociated to form methoxy and hydroxyl species on the oxidized Fe3+ sites, and the
methoxy species can be migrated to Si sites by forming methanol selectively on the Bronsted
acid sites [19,28]. These reaction schemes were proposed by in-situ FT-IR characterization
for the reaction of CH4 and N2O on the Fe/MFI catalyst previously [48]. The adsorbed
CH3OH intermediates on the acid sites can be further reacted with another methoxy group or
methanol to DME by successive dehydration reaction as well [49]. Therefore, not only the
formation of methoxy groups on the α-oxygen sites but also the amount of hydroxyl groups
of the Bronsted acid sites seems to be important for the formation of methanol and DME
from an adsorbed CH4 and N2O. The amount of surface hydroxyl groups, which were mainly
originated from Bronsted acid sites of the Fe/zeolites, is also important for a formation of α-
oxygen sites [24,29,30]. By changing the location of iron species during catalyst preparation,
the amount of unoccupied Bronsted acid sites can be adjusted by controlling the size of iron
nanoparticles and the amount of α-oxygen sites as well. In addition, an excess dehydration of
the oxygenates on the Fe/ZSM-5 seems to be responsible for forming olefinic hydrocarbons
through a well-known MTO reaction pathway, and the dehydration of oxygenates to olefinic
hydrocarbons can occur on the strong Bronsted acid sites which finally transformed to coke

precursors. From the results of slight catalyst deactivation with time on stream as shown in
Figure 3, it was observed that the selectivity to olefinic hydrocarbons and COx was steadily
decreased. This activity variation seems to be attributed to coke formation on the strong acid
sites by lessening the dehydration activity of oxygenates to hydrocarbons through a well-
known carbon pool mechanisms [28-30,49]. Therefore, the strength and amount of the
surface hydroxyl groups associated with Bronsted acid sites on the Fe/zeolites can change the
rate of CH4 conversion and product distribution by altering dehydration activity of the
oxygenated intermediates on the Fe/zeolites. For a brief comparaision with present Fe/FER
catalytic system with others, the characteristics and activities of various heterogeneous
catalysts to utilize CH4 and N2O (or H2O2) oxidant to form oxygenates selectively are
summarized in Table 3, and a higher formation rate of oxygenates such as DME and
methanol was observed on the present Fe/FER(IMP).
3.3. Active sites for a direct CH4 activation with different product distribution
To confirm the proposed reaction schemes and different catalytic performances during direct
activation of CH4 with the help of N2O, the α-oxygen sites and the amount of acid sites with
the types such as Bronsted and Lewis sites on the Fe/zeolites were quantitatively measured by
the characterizations of N2O titration, NH3-TPD and Py-IR analysis. The results of NH3-TPD
and Py-IR analysis are displayed in Figure 4 and Figure 5, and the results are summarized in
Table 1. As shown in Figure 4, three major desorption peaks of NH3 on the Fe/zeolites were
o
observed at maximum peak temperatures at around 180, 400 and 500 C. These desorption
peaks can be assigned to the weak, strong acid sites and water desorption by a formation of
water from surface hydroxyl groups of zeolite frameworks [41,42,50-55]. The much larger
peak intensity of the weak acid sites was observed on the Fe/FER than the Fe/ZSM-5 due to a
lower Si/Al ratio of FER than ZSM-5 with intrinsically larger amount of weak acid sites on

the FER [39,53,54]. The number of weak and strong acid sites was further calculated by
deconvoluting the broad desorption peaks of NH3, and the results are summarized in Table 1.
By comparing with the acid sites of the unmodified bare zeolites, the number of weak acid
sites afrer Fe modification on the zeolites was increased and vice versa of the strong acid sites.
The respective values of weak and strong acid sites on the bare ZSM-5 were found to be of
0.145 and 0.594 mmol NH3/g and those values of 0.541 and 0.400 mmolNH3/g on the FER.
The decreased amount of the strong acid sites after Fe modification with the respective values
of 0.417 and 0.375 mmolNH3/g on the Fe/ZSM-5(IMP) and Fe/ZSM-5(IE) and those of 0.362
and 0.253 mmolNH3/g on the Fe/FER(IMP) and Fe/FER(IE) was attributed to the selective
ion-exchange of iron nanoparticles on the strong acid sites [52-54]. In addition, an increased
amount of the weak acid sites with the respective values of 0.206 and 0.255 mmolNH3/g on
the Fe/ZSM-5(IMP) and Fe/ZSM-5(IE) and those of 0.545 and 0.554 mmolNH3/g on the
Fe/FER(IMP) and Fe/FER(IE) seems to be attributed to the newly formed interaction of iron
oligomers with the protons of the Bronsted acid sites of zeolites.
A relative amount and type of the acid sites of the Fe/zeolites were further characterized by
Py-IR and the absorption peaks of pyridine molecules are displayed in Figure 5. Three
characteristic absorption bands of pyridine molecules were clearly observed by the adsorption
of pyridine on Bronsted (B), Lewis (L) and combined acid sites of the Fe/zeolites. The
absorption band located at around 1450 cm-1 was attributed to Lewis acid sites for C-C
stretching vibration mode of the adsorbed pyridine complex. The absorption band at around
1540 cm-1 was attributed to C-C stretching vibration of the pyridinium ions adsorbed on the
Bronsted acid sites, and the peak at around 1490 cm-1 is assigned to the interaction of
pyridine molecules on both Lewis and Bronsted acid sites as well [55]. To verify separate
amounts of Bronsted and Lewis acid sites on the Fe/zeolites, the integrated area of the each
peak at 1540 cm-1 was divided by that of 1450 cm-1 which is denoted as B/L ratio, and the

results are summarized in Table 1. The B/L ratio was also calculated on the unmodified bare
zeolites and the values were found to be 8.3 and 74.0 on the ZSM-5 and FER, respectively.
The B/L ratio was decreased after Fe modification on both zeolites by showing the values of
3.6 and 1.7 on the Fe/ZSM-5(IMP) and Fe/ZSM-5(IE) and those of 32.8 and 21.5 on the
Fe/FER(IMP) and Fe/FER(IE), respectively. From the Py-IR results, the number of Bronsted
acid sites on the Fe/FER(IMP) was found to be larger than other Fe/zeolites and the catalytic
activity was found to be higher on the Fe/FER(IMP) having a larger amount of Bronsted acid
sites because Bronsted acid sites are known to be most active sites for methanol dehydration
reaction [52-54]. Therefore, a larger amount of weak Bronsted acid sites with abundant α-
oxygen sites can be responsible for a higher methane conversion rate on the Fe/FER and a
lower conversion rate seems to be attributed to a lower amount of Bronsted acid sites with
strong acidity on the Fe/ZSM-5 through the MTO reaction pathway which can lead to form
severe coke precursors as well [56].
The decomposition activity of N2O as a mild oxidizing agent can be strongly affected by
total amount of α-oxygen sites on the Fe/zeolites. Therefore, the quantity of active α-oxygen
sites was measured by a pulse N2O chemisorption and the amount of α-oxygen sites is
summarized in Table 1. The much smaller amount of α-oxygen sites was observed on the
Fe/ZSM-5(IMP) with the value of 1.8 x 1019 molecules/g than the Fe/FER(IMP) with 10.3 x
1019 molecules/g. In addition, a smaller amount of α-oxygen sites was also observed on the
Fe/ZSM-5(IE) with the value of 0.7 x 1019 molecules/g than that of 10.0 x 1019 molecules/g
on the Fe/FER(IE), and the Fe/FER showed almost ten times larger amount of α-oxygen sites
than the Fe/ZSM-5. In general, the active α-oxygen sites can be formed by redox cycles of
two iron cations through a transformation of Fe3+  Fe2+, which are attached to ion-
exchanged iron sites closely bridged with an oxygen atoms by a decomposition of N2O [51].
Since FER zeolite has a shorter distance and easy accessibility between ion-exchanged α-iron

cations than that of ZSM-5, the shorter distance between ion-exchanged iron cations on the
FER may increase a probability of active α-oxygen formation. Furthermore, α-iron cations
are located on the pore channel of 10-membered ring with a pore size of 0.42 x 0.54 nm,
whereas the ZSM-5 has slightly larger pore size of 0.54 x 0.56 nm of 10-membered ring with
zigzag pore structures intersecting the straight pore channel with a longer distance between
ion-exchanged iron cations [17,50,51]. These positive structural properties of the FER can
enhance the decomposition rate of N2O [38], and it can enhance a rate of CH4 activation
possibly. However the active α-oxygen sites on the Fe/ZSM-5 have not been intimately faced
due to the characteristic zigzag pore structures of the ZSM-5, and N2O decomposition seems
to be difficult on the Fe/ZSM-5 compared to the Fe/FER.
Even though a loading amount of iron precursor was found to be lower on the Fe/ZSM-5(IE)
and Fe/FER(IE) with the respective values of 4.8 and 6.4 wt% than those of 7.8 and 7.3 wt%
on the Fe/ZSM-5(IMP) and Fe/FER(IMP), the amount of α-iron sites was found to be larger
on the Fe/FER than the Fe/ZSM-5 especially on the Fe/zeolites prepared by impregnation
method. From the results of N2O titration, the amount of active α-oxygen sites was slightly
larger on the Fe/FER(IMP) than the Fe/FER(IE) as well. As shown in Figure 6, TEM images
of the selected Fe/FER(IMP) and Fe/FER(IE) after reaction showed a significant aggregation
of the iron oxide nanoparticles (magnified images in the inset figures). The fresh Fe/FER(IE)
showed well-dispersed iron nanoparticles with a particle size of below 10 nm because of the
selective adsorption of iron precursor to the Bronsted acid sites, however, the iron oxide
nanoparticles were significantly aggregated on the used Fe/FER(IE) with a particle size above
50 nm. Interestingly, the iron oxide nanoparticles with a heterogeneous size distribution in the
range of 10 – 50 nm in size were clearly observed on the fresh Fe/FER(IMP). On the used
Fe/FER(IMP), the aggregation of iron nanoparticles was found to be more significant than
that of the Fe/FER(IE), which suggests a less number of iron nanoparticles blocks the active

Bronsted acid sites compared with the Fe/FER(IE). This observation can explain a slightly
higher selectivity to DME on the Fe/FER(IMP) than Fe/FER(IE) due to largely exposed
active acid sites for methanol dehydration to DME [39,49,53,54,57,58]. In addition, a larger
iron oxide nanoparticles on the Fe/FER(IMP), which are active sites by forming α-oxygen
sites, can enhance the fully oxidized COx products through a facile oxidation of intermediates,
especially on the Fe/FER compared to the Fe/ZSM-5. In addition, the deposited coke
precursors can also change the catalytic activity and product distribution by blocking the
active sites such as α-oxygen sites and Bronsted acid sites mainly. Even though a filamentous
coke formation was not clearly observed even in the magnified TEM images on the used
Fe/FER, the extent of coke formation was confirmed by hydrogenation of cokes through
TPSR. As shown in Figure 7, TPSR experiment showed that the deposited coke precursors
were hydrogenated to form CH4 under a flow of H2, which was verified by peaks of the m/z
values of 15 and 14 on the used Fe/ZSM-5 and Fe/FER. The amount of coke formation was
much larger on the used Fe/ZSM-5 than the Fe/FER irrespective to the preparation methods,
which revealed the coke deposition also altered the product distribution and the rate of CH4
conversion as well. The formed coke precursors seem to be much dominant on the Fe/ZSM-5
having stronger acid sites due to a facile reaction pathway through MTO reaction. The
amount of coke deposition on acid sites and α-oxygen sites was well correlated with catalytic
performances, and we proposed a possible reaction pathway and schemes based on the
product distribution with respect to acid sites and α-oxygen sites in Figure 8 and Scheme 1.
3.4. Proposed reaction pathway for the formation of oxygenates and hydrocarbons
To further verify the effect of acidic properties and α-oxygen sites on the Fe/ZSM-5 and
Fe/FER for a different product distribution, methanol conversion reaction to oxygenates and
hydrocarbons was further carried out under an environment of N2O flow with a feeding of

liquid methanol and the results are summaized in supplementary Table S3. For a methanol
conversion to hydrocarbons by MTO reaction pathway, methanol can be previously
dehydrated to DME, and it can be transfomred to a broad range of hydrocarbons including
light olefins such as C2 - C4 hydrocarbons through carboon pool mechanisms [59-61]. In
addition, methanol conversion with stability and product distribution are strongly depending
on the surface acidic properties such as the number of acid sites and their strengths [60,61].
As shown in supplementary Table S3, the bare ZSM-5 and FER showed a similar conversion
of methanol with a value of 87.7% on the ZSM-5 and 87.4% on the FER. However, a
significantly different product distribution was observed by showing a higher hydrocarbon
selectivty of 51.3% on the ZSM-5 and a higher DME selectivity of 99.8% on the FER. Based
on the acidic properties of the Fe/zeolites as summarized in Table 1 from NH3-TPD and Py-
IR, the observed higher conversion of methanol and DME selectivity on the bare FER can be
attributed to a larger amount of Bronsted acid sites, and vice versa on the ZSM-5 by showing
a higher selectivity to hydrocarbons due to a large amount of strong acid sites. On the ZSM-5,
conversion of methanol seems to follow a well-known auto-catalyzed reaction on the strong
acid sites through a consecutive transformation of DME to hydrocarbons [59]. However, after
Fe modification on the zeolites, methanol conversion was decreased due to a suppressed
amount of Bronsted acid sites, which was more significant on the Fe/ZSM-5(IE) and
Fe/FER(IE) than the Fe/ZSM-5(IMP) and Fe/FER(IMP). This can be attributed to a well
distributed iron oxide nanoparticles on the acid sites of the zeolites through an ion-exhange as
confirmed by TEM analysis. Interestingly, the selectivity to hydrocarbons on the Fe/ZSM-5
was significantly suppressed due to a selective adsorption of iron species on the strong acid
sites as well [52-54,59]. However, the product distribution showed that the selectivity to
oxygenates was above 90% on all Fe-modified zeolites. The selectivity to oxygenates can be
largely increased from 48.7 % on the bare ZSM-5 up to 97% with a small amount of

hydrocarbon formation on the Fe/ZSM-5 due to selective adsorption of α-oxygen sites from
iron on the stong acid sites. These results also revealed that the reaction followed a different
reaction pathway compared to a general methanol dehydration by introducting α-oxygen sites
for methanol conversion. The reaction mainly undergoes the dehydration of methanol to
DME by prohibiting the formation of carbon pools from a succesive dehydration of DME on
the Fe/ZSM-5. Therefore, the formation of coke precursors on the strong acidic sites by the
succesive dehydration of DME to hydrocarbons through carbon pool mechanisms can be
largely prohibited by the assistance of α-oxygen sites. The differences of the converted
amount of N2O was clearly observed on all Fe/zeolites which suggests the α-oxygen site can
efficiently take part in CH4 activation reaction. The α-oxygen sites can also form CO product
by decomposition of intermediates through a partial oxidation reaction, especially on the
Fe/FER having an abundant active α-oxygen sites. Compared with a direct conversion of CH4
with N2O, a relative amount of CO formed was found to be quite small during methanol
dehydration. However, hydrocarbon selectivity was still higher on the on the Fe/ZSM-5 than
the Fe/FER. Generally, COx molecules can be formed by a partial oxidation of methane or
surface intemediates such as adsorbed hydrocarbons or oxygenates [52,53]. Therefore, the
observed larger amount of CO on the Fe/FER than the Fe/ZSM-5 during a direct activation of
CH4 with N2O can be attributed to a partial oxidation of the adsorbed intermediates due to an
abundant presence of actvie α-oxygen sites on the Fe/FER. Based on the addition reaction
results, the reaction pathway of CH4 activation through N2O decomposition with a different
product distribution on the Fe/zeolites are schematically presented in Figure 8, and the
reaction scheme is also displayed in Scheme 1. The Fe/ZSM-5 having a small amount of α-
oxygen sites with strong Bronsted acid sites showed lower CH4 and N2O conversion with a
higher selectivity to olefinic hydrocarbons and DME with a little bit higher amount of coke
deposition due to an easy consecutive reaction of methanol or DME intermediates to olefinic

hydrocarbons on the strong acid sites. However, the Fe/FER having a larger amount of the α-
oxygen sites with a large amount of weak Bronsted acid sites can increase the rate of CH4 and
N2O conversion with a higher selectivity to oxygenates and COx products. This was mainly
attributed to a higher formation of methoxy groups with a large amount of Bronsted acid sites
as hydroxyl supplier and a simultaneous oxidation of the adsorbed methoxy or methanol
intermediates to fully oxidized COx on active α-oxygen sites of the Fe/FER as summarized in
Scheme 1.
In summary, the activation of CH4 under N2O decomposition was much easier on the
Fe/FER due to the presence of a larger amount of active α-oxygen sites and weak Bronsted
acid sites, especially on the Fe/FER(IMP). However, the product distribution such as DME,
methanol and hydrocarbons was strongly depending on the amount and strength of Bronsted
acid sites through the proposed reaction pathway in Figure 8. The Fe/ZSM-5 having a large
amount of strong acid sites seems to be responsible for a selective production of olefinic
hydrocarbons and DME through a consecutive dehydration reaction of methanol (or DME) to
unsaturated C2-C3 hydrocarbons. Therefore, the Fe/FER having abundant Bronsted acid sites
seems to be a proper catalytic system to obtain oxygenates such as methanol and DME with a
higher rate of CH4 conversion compared to the Fe/ZSM-5. This simple controlling method of
α-oxygen sites and Bronsted acid sites can be applied to obtain a higher rate of CH4
conversion and selectivity to oxygenates for a direct CH4 activation with N2O decomposition.
4. Conclusions
For a direct activation of CH4 under N2O decomposition to oxygenates such as methanol
and DME as well as light olefinic hydrocarbons on the Fe-modified ZSM-5 and FER, the
amount of active α-oxygen sites and weak Bronsted acid sites are crucial factors to alter CH4
conversion and product distribution as well. A facile activation of CH4 with N2O

decomposition was found to be more significant on the Fe/FER prepared by impregnation
method due to an abundant presence of active α-oxygen sites as well as weak Bronsted acid
sites. The Fe/ZSM-5 having a large amount of strong acid sites was also responsible for a
higher selectivity to unsaturated C2-C3 hydrocarbons and DME with a little bit higher coke
deposition. However, the Fe/FER having a large amount of the weak Bronsted acid sites was
found to be proper for obtaining oxygenates such as methanol and DME. Therefore, the main
controlling variables for catalytic performances during direct CH4 activation with the help of
N2O decomposition can be the amount of active α-oxygen sites with a different distribution of
iron oxide nanoparticles and the Bronsted acid sites with their strength, which were strongly
altered by preparation methods.
Acknowledgements
This work was also financially supported by an institutional program grant (2E26570-16-
037) from the Korean Institute of Science and Technology (KIST) and by Fundamental
Research Program of the Korea Institute of Materials Science (PNK4310). The authors would
like to acknowledge the financial support from the National Research Foundation of Korea
(NRF) grant funded by the Korea government (NRF-2014R1A1A2A16055557 and NRF-
2016M3D3A1A01913253). This work was supported by the National Research Council of
Science and Technology (NST) through Degree and Research Center (DRC) Program (2015).
This work was financially supported by the R&D Center for Valuable Recycling (Global-Top
R&D Program) of the Ministry of Environment of Korea (Project No. RE201606017).

References
[1] J.H. Lunsford, Catal. Today 63 (2000) 165-174.
[2] A.B.M.L. Rahman, M. Kumashiro, T. Ishihara, Catal. Commun. 12 (2011) 1198-1200.
[3] M.J. Brown, N.D. Parkyns, Catal. Today 8 (1991) 305-335.
[4] N. Yuqin, C. Zhenghua, H. Xiuhua, J. Tao, P. Junxuan, L. Yu, Z. Xia, T. Yisheng, J. Nat.
Gas. Chem. 9(2000) 59-67.
[5] K.J. Zhen, M.M. Khan, C.H. Mak, K.B. Lewis, G.A. Somorjai, J. Catal. 94 (1985) 501-
507.
[6] H.F. Liu, R.S. Liu, K.Y. Liew, R.E. Johnson, J.H. Lunsford, J. Am. Chem. Soc. (1984),
106, 4117-4121.
[7] X. Wang, Y. Wang, Q. Tang, Q. Guo, Q. Zhang, H. Wan, J. Catal. 217 (2003) 457-467.
[8] R. Polniser, M. Stolcova, M. Hronec, M. Mikula Appl. Catal. A:Gen. 400 (2011) 122-130.
[9] H. Launay, S. Loridant, A. Pigamo, J.L. Dubois, J.M.M. Millet, J. Catal. 246 (2007) 390-
398.
[10] S. Chempath, A.T. Bell, J. Catal. 247 (2007) 119-126.
[11] S. Mansouria, O. Benlounes, C. Rabia, R. Thouvenot, M.M. Bettahar, S. Hocine, J. Mol.
Catal. A: Chem. 379 (2013) 255-262.
[12] N.V. Beznis, B.M. Weckhuysen, J.H. Bitter, Catal. Lett. 136 (2010) 52-56.
[13] M.F. Fellah, I. Onal, J. Phys. Chem.C 114 (2010) 3042-3051.
[14] J.P. Tessonier, L. Benoit, S. Rigolet, M.J. Ledoux, C. Pham-Huu, Appl. Catal. A: Gen.
336 (2008) 79-88.
[15] J.P. Tessonier, L. Benoit, M.J. Ledoux, C. Pham-Huu, Catal. Commun. 8 (2007) 1787-
1792.
[16] C. Hammond, M.M. Forde, M.H.A. Rahim, A. Thetford, Q. He, R.L. Jenkins, N.
Dimitratos, J.A. Lopez-Sanchez, N.F. Dummer, D.M. Murphy, A.F. Carley, S.H. Taylor, D.J.

Willock, E.E. Stangland, J. Kang, H. Hagen, C.J. Kiely, G.J. Hutchings, Angew. Chem. Int.
Ed. 51 (2012) 5129-5133.
[17] P. Pantu, B. Boekfa, B. Sunpetch, J. Limtrakul, Chem. Eng. Comm. 195 (2008) 1477-
1485.
[18] A. Heyden, N. Hansen, A.T. Bell, F.J. Keil, J. Phys. Chem. B 110 (2006) 17096-17114.
[19] E.V. Starokon, M.V. Parfenov, L.V. Pirutko, S.I. Abornev, G.I. Panov, J. Phys. Chem. C
115 (2011) 2155-2161.
[20] J. Xu, R.D. Armstrong, G. Shaw, N.F. Dummer, S.J. Freakley, S.H. Taylor, G.J.
Hutchings, Catal. Today 270 (2016) 93-100.
[21] Y.F. Chang, J.H. McCarty. Catal. Lett. 34(1995) 163-177.
[22] B.R. Wood, J.A. Reimer, A.T. Bell, M.T. Janicke, K.C. Ott, J. Catal. 224 (2004) 148-155.
[23] C. Sang, B.H. Kim, Carl R.F. Lund, J. Phys. Chem. B 109 (2005) 2295-2301.
[24] E.V. Starokon, K.A. Dubkov, L.V. Pirutko, G.I. Panov, Top. Catal. 23 (2003) 137-143.
[25] L. Kiwi-Minsker, D.A. Bulushev, A. Renken, J. Catal. 219 (2003) 273-285.
[26] G.I. Panov, A.K. Uriarte, M.A. Rodkin, V.I. Sobolev, Catal. Today 41 (1998) 365-385.
[27] P.J. Smeets, J.S. Woertink, B.F. Sels, E.I. Solomon, R.A. Schoonheydt, Inrog. Chem. 49
(2010) 3573-3583.
[28] M.V. Parfenov, E.V. Starokon, L.V. Pirutko, G.I. Panov, J. Catal. 318 (2014) 14-21.
[29] L.V. Pirutko, V.S. Chernyavsky, A.K. Uriarte, G.I. Panov, Appl. Catal. A: Gen. 227
(2002) 143-157.
[30] P.P. Notte, Top. Catal. 13 (2000) 387-394.
[31] R.J. Madon, M. Boudart, Ind. Eng. Chem. Fundam. 21 (1982) 438-447.
[32] A.Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 107 (2007) 1692-1744.
[33] D. Verboekend, R. Caicedo-Realpe, A. Bonilla, M. Santiago, J. Peerez-Ramirez, Chem.
Mater. 22 (2010) 4679-4689.

[34] F. Zhang, X. Chen, J. Zhuang, Q. Xiao, Y. Zhong, W. Zhu, Catal. Sci. Technol. 1 (2011)
1250-1255.
[35] X. Cheng, J. Wang, J. Guo, J. Sun, Y. Long, Chem. Phys. Chem. 7 (2006) 1198-1202.
[36] M.S. Batista, M. Wallau, E.A. Urquieta-Gonzalez, Brazilian J. Chem. Eng. 22 (2005)
341-351.
[37] J. Perez-Ramírez, J. Catal. 227 (2004) 512-522.
[38] L.Li, Q. Shen, J. Li , Z. Hao, Z.P. Xu, G.Q. Max Lu, Appl. Catal. A: Gen. 344 (2008)
131-141.
[39] S.H. Kang, J.H. Ryu, J.H. Kim, I.H. Jang, A.R. Kim, G.Y. Han, J.W. Bae, K.S. Ha,
Energy Fuels 26 (2012) 6061-6069.
[40] M. Mihaylov, E. Ivanova, N. Drenchev, K. Hadjiivanov, J. Phys. Chem. C 114 (2010)
1004-1014.
[41] L.J. Lobree, I.C. Hwang, J.A. Reimer, A.T. Bell, J. Catal. 186 (1999) 242-253.
[42] E.M. El-Malki, R.A. van Santen, W.M.H. Sachtler, J. Phys. Chem. 103 (1999) 4611-
4622.
[43] A. Guzman-Vargas, G. Delahay, B. Coq, Appl. Catal. B: Environ. 42 (2003) 369-379.
[44] H.Y. Chen, W.M.H. Sachtler, Catal. Today 42 (1998) 73-83.
[45] M.H. Jeong, D.H. Lee, J.W. Bae, Int. J. Hydrogen Energy 40 (2015) 2613-2620.
[46] Sk. Mahammadunnisa, P.M.K. Reddy, Ch. Subrahmanyam, RSC Adv. 4 (2014) 4034-
4036.
[47] Sk. Mahammadunnisa, K. Krushnamurty, Ch. Subrahmanyam, Catal. Today 256 (2015)
102-107.
[48] B.R. Wood, J.A. Reimer, A.T. Bell, M.T. Janicke, K.C. Ott, J. Catal. 225 (2004) 300-
306.
[49] K.S. Ha, Y.J. Lee, J.W. Bae, Y.W. Kim, M.H. Woo, H.S. Kim, M.J. Park, K.W. Jun,