Hydrothermal preparation of Fe-Zr catalysts for the direct conversion of syngas to light olefins

Article abstract of DOI:10.1039/c5ra27712d

Fe/Zr precursors were prepared by a one-step and two-step hydrothermal process respectively, following K promotion by incipient wetness impregnation, which then were used for the selective production of C₂-C₄ olefins via Fischer-Tropsch synthesis from syngas. The textural properties, phase structure, reduction behavior and the catalytic performance of the obtained catalyst samples were greatly affected by different preparation procedures. The T-Fe/Zr samples prepared by the two-step hydrothermal process caused a decrease in the surface area and easier reduction of iron species compared with that of the O-Fe/Zr samples prepared by the one-step hydrothermal process. The T-Fe/Zr samples before and after K promotion showed preferred orientation along the (-111) and (111) lattice planes of ZrO₂ in contrast to the ZrO₂ (101) lattice plane on O-Fe/Zr samples. For CO hydrogenation, all the catalysts exhibited high activity and good stability. The T-Fe/Zr-K catalysts with highly dispersed zirconia on the iron surface could prevent Fe-Zr mixed oxide formation by decreasing the interaction between iron and zirconia, leading to a slight increase in C₂-C₄ olefin selectivity and improvement on the product distribution.

Full text of DOI:10.1039/c5ra27712d

View Article Online
View Journal
RSC Advances
This article can be cited before page numbers have been issued, to do this please use: N. Chen, J.
Zhang, Q. Ma, S. Fan and T. Zhao, RSC Adv., 2016, DOI: 10.1039/C5RA27712D.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. This Accepted Manuscript will be replaced by the edited,
formatted and paginated article as soon as this is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/advances

Page 1 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
Hydrothermal preparation of FeꢀZr catalyst for the direct conversion of
syngas to light olefins
Ning Chen, Jianli Zhang*, Qingxiang Ma, Subing Fan, TianꢀSheng Zhao*
State Key Laboratory Cultivation Base of Natural Gas Conversion, College of Chemistry & Chemical
Engineering, Ningxia University, Yinchuan, 750021, China
Eꢀmail address: zhangjl@nxu.edu.cn; zhaots@nxu.edu.cn
Abstract
The Fe/Zr precursors were prepared by oneꢀstep and twoꢀstep hydrothermal process respectively,
following K promotion by incipient wetness impregnation, which then were used for the selective
production of C2ꢀC4 olefins via FischerꢀTropsch synthesis from syngas. The textural properties, phase
structure, reduction behavior and the catalytic performance of obtained catalyst samples were greatly
affected by different preparation procedures. The TꢀFe/Zr samples prepared by twoꢀstep hydrothermal
process caused decrease in the surface area and easier reduction of iron species compared with that of
OꢀFe/Zr samples prepared by oneꢀstep hydrothermal process. The TꢀFe/Zr samples before and after K
promotion showed preferred orientation along the (ꢀ111) and (111) lattice planes of ZrO2 in contrast to
the ZrO2 (101) lattice plane on OꢀFe/Zr samples. For CO hydrogenation, all the catalysts exhibited high
activity and good stability. The TꢀFe/ZrꢀK catalysts with highly dispersed zirconia on iron surface could
prevent FeꢀZr mixed oxide formation by decreasing the interaction between iron and zirconia, leading
to a slight increase in C2ꢀC4 olefin selectivity and improvement on the product distribution.
Keywords: FischerꢀTropsch synthesis; Syngas; Light olefins; Twoꢀstep hydrothermal preparation

RSC Advances
Page 2 of 24
View Article Online
DOI: 10.1039/C5RA27712D
1. Introduction
The FischerꢀTropsch synthesis (FTS), which is regarded as one of the most attractive processes to
convert syngas derived from coal, natural gas and biomass into liquid fuels and other valuable
chemicals such as light olefins and alcohols, has been the subject of renewed interest [1]. The direct
synthesis of light olefins from syngas via FTS as a nonꢀpetroleum based supply has attracted much
attention [1ꢀ4]. The catalyst research has progressed in recent years [1], especially on modified iron
catalysts, which have shown the great prospects for selective formation of light olefins from CO
hydrogenation and have been extensively studied due to its less expensive while adequate activity for
the water gas shift (WGS) reaction and high selectivity to light olefins [4ꢀ7]. However, the product
distribution ruled by AndersonꢀSchulzꢀFlory (AꢀSꢀF) distribution and the selectivity towards light
olefins need further improvement during CO hydrogenation.
The promotion effects of various transition metal promoters and supports on iron catalysts depend
strongly on the preparation methods and hence the structure of the catalysts [8ꢀ12]. The alkali metals,
especially potassium, have been used widely as promoters to improve the activity of iron catalysts and
to suppress the methane formation in the FTS [7, 8]. It is necessary to add certain amount of potassium
on iron catalysts for selective formation of light olefins in CO hydrogenation. In addition, the first step
in the FꢀT reaction on reduced group VIII metals is considered to be the adsorption and dissociation of
hydrogen and carbon monoxide [13]. Therefore, in contact with FTS, a structureꢀsensitive reaction [14],
the ability to produce light olefins is heavily dependent on the catalyst design, especially on the metal
surface of catalysts, i.e., the dispersion and active sites [15]. It is reported that zirconia possesses the
coexistence of acidity and basicity as well as reducibility and oxidizability of the surface, which has
widely been used as supports and additives in FTS and exhibits good chemical stability [16ꢀ20].

Page 3 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
Especially it has been proved to show good ability for CO hydrogenation, but is poor for hydrogenation
of olefin to corresponding paraffin with light hydrocarbons as the main products [16]. Our previous
studies showed that potassium promoted ZrO2 catalysts exhibited high ability in the production of
C1ꢀC4 hydrocarbons from CO hydrogenation with high olefin selectivity and low CO2 selectivity. Also,
the FeꢀZr bimetallic catalysts prepared by microwaveꢀhydrothermal method have shown great
advantages in increasing the activity and improving the product distribution during CO2 hydrogenation
[17]. However, for CO hydrogenation, the product spectrum over the above catalysts is still broad for
the direct synthesis of light olefins due to the strong chain propagation ability. Similar results are also
found on a FeꢀZr catalyst prepared by coꢀprecipitation method [18]. Furthermore, the strong
interactions between iron and zirconia, confirmed by previous studies on the reduction behavior of
Fe/ZrO2 and Fe/K/ZrO2 catalysts prepared by coꢀprecipitation [18, 19] and impregnation methods [15,
19, 20], significantly affect the surface states and the properties of the catalysts [18]. Based on the
advantages and prospects of zirconia catalyst in CO hydrogenation, the promotion effects of zirconia
additive on the structure and catalytic performance of FeꢀZr catalysts by different preparation methods
are desired.
Traditional preparation of FeꢀZr catalysts are generally by the coꢀprecipitation of iron and zirconia
precursors to form FeꢀZr mixed oxides [18,19] or using zirconia as support to form Fe/ZrO2 catalysts
by impregnation method [15, 19, 20] as discussed above. Also, the dispersion state of iron species, the
phase structure of ZrO2 and the interaction between iron and zirconia have proved to play important
roles in modifying the catalytic properties of FeꢀZr catalysts. In the present work, to further improve
the catalytic performance of FeꢀZr catalysts for CO hydrogenation to light olefins, we propose a new
strategy for designing FeꢀZr catalysts by a twoꢀstep hydrothermal process, following potassium

RSC Advances
Page 4 of 24
View Article Online
DOI: 10.1039/C5RA27712D
modification by impregnation method, in order to improve the dispersion of iron species and decrease
the interaction between iron and zirconia. Comparative study on the catalytic performance of the
catalysts prepared by twoꢀstep and oneꢀstep hydrothermal processes was conducted under the similar
catalyst compositions and reaction conditions. The experimental results indicated that the incorporation
of zirconia to iron catalyst by twoꢀstep methods was beneficial to further improve the product
distribution and olefin selectivity. which was also confirmed by the influence of K contents on the
activity of the catalysts. The relationship between the structure and properties of the catalysts was
discussed by means of the physicochemical property characterization.
2. Experimental
2.1 Catalyst preparation
Oneꢀstep FeꢀZr precursors with Fe2O3/ZrO2 molar ratio of 5:1, 3:1 and 1:1 were prepared by
microwaveꢀhydrothermal method using Fe(NO3)3·9H2O and ZrO(NO3)2·2H2O as raw materials and as
described in reference [17]. The process of twoꢀstep FeꢀZr catalyst preparation included the following
steps. The first step was to prepare Fe2O3 precursor using the same procedure as the preparation of
oneꢀstep FeꢀZr precursors. In the second step, the obtained iron precursor was firstly treated by
ultrasonic dispersion, which then was added to the mixed aqueous solutions of ZrO(NO3)2·2H2O and
urea (Zr/urea molar ratio=1/1), and reacted within a dynamic homogeneous reactor conducted at 160 ºC
for 4h. The Fe2O3/ZrO2 molar ratios were 5:1, 3:1 and 1:1 for different samples, respectively. After the
synthesis reaction, the FeꢀZr precipitation was washed and dried at 80 ºC overnight, and then calcined
in ambient air at 500 ºC for 3h. The obtained powders were impregnated with potassium carbonate
(Fe/K molar ratio=10:1, 20:1 and 30:1 for different samples) and dried at 110 ºC for several hours. The
catalysts with different Zr contents were labeled as Oꢀ5Fe/1Zr[5ꢀ1]ꢀK, OꢀFe/Zr[3ꢀ1]ꢀK and

Page 5 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
OꢀFe/Zr[1ꢀ1]ꢀK for oneꢀstep process, and TꢀFe/Zr[5ꢀ1]ꢀK, TꢀFe/Zr[3ꢀ1]ꢀK, TꢀFe/Zr[1ꢀ1]ꢀK for twoꢀstep
process, respectively. Finally, all the catalyst samples were granulated and sieved to obtain 20ꢀ40 mesh
for characterization and activity test.
2.2 Catalyst characterization
The nitrogen sorption experiments were performed at ꢀ196 ºC on a JWꢀBK132F type N2
physisorption equipment (Beijing JWGB Sci. & Tech. Co., Ltd). 200 mg sample was degassed under
vacuum at 300 ºC for 2h prior to measurement. The surface area of samples was determined using the
BrunauerꢀEmmettꢀTeller (BET) equation based on single point method.
Transmission electron microscopy (TEM) analysis was conducted on a HITACHI Hꢀ7650
microscope with an acceleration voltage of 80 kV to analyze the morphology of samples. HRTEM
instrument (JEMꢀ2100) equipped with an energy dispersive Xꢀray spectrometer (EDS) is also used to
observe the morphology and chemical composition of the samples. All the samples were ultrasonically
dispersed in anhydrous alcohol prior to the analysis.
Powder Xꢀray diffraction (XRD) measurements for all the fresh and spent catalysts were performed
on a Rigaku D/maxꢀ2200PC Xꢀray diffractometer with Cu Kα radiation (λ=1.5406 Å, 40 kV, 30 mA ) in
the 2θ angle ranging from 5° to 85° at scanning speed of 8°/min.
H2 temperatureꢀprogrammed reduction (H2ꢀTPR) measurement of the fresh catalysts was carried out
on a TPꢀ5000 chemisorption analyzer. Sample of 50 mg was pretreated with He (30 mL/min) at 300 ºC
for 1 h, and then cooled to room temperature. The sample was reduced by a reducing gas (30 mL/min)
composed of 5% H2 and 95% N2 as the sample was heated from room temperature to 800 ºC at a rate of
10 ºC /min. TPR profiles were recorded using thermo conductive detector.
The Xꢀray photoelectron spectra (XPS) for the fresh and used samples were recorded on a thermo

RSC Advances
Page 6 of 24
View Article Online
DOI: 10.1039/C5RA27712D
scientific ESCALAB 250 spectrometer using an Al Kα Xꢀray source, where the base pressure of the
chamber was less than 2×10ꢀ8 Pa. The binding energies (BEs) were calibrated relative to adventitious
carbon using the C 1s peak at 284.8 eV.
2.3 Catalytic activity tests
The CO hydrogenation reaction was carried out on a microꢀtype fixedꢀbed reactor system (8 mm i.d,
400 mm in length). Before reaction, all the catalyst samples were reduced in situ using the syngas feed
(H2/CO=2) at 1000 hꢀ1, 0.1 MPa and 300 ºC for 4 h. The typical reaction conditions were as follows:
catalyst packing amount, 2 mL (20ꢀ40 mesh), H2/CO (molar ratio) =2, GHSV=1000 hꢀ1, T= 320 ºC and
p=1.5 MPa. The effluent gas was analyzed on an online gas chromatograph (GCꢀ9560ꢀI), with a 2 m
TDXꢀ01 packed column for C1 products analysis via TCD detector, and a 50 m Al2O3 capillary column
for C1ꢀC5 hydrocarbons products via FID detector, respectively. The liquid products were analyzed on
an offꢀline gas chromatograph (GCꢀ9560ꢀII), with a 2 m GDXꢀ 401 packed column for the aqueous
phase products analysis via TCD detector and a 30 m SEꢀ30 capillary column for the oil phase products
analysis via FID detector. The mass balance was based on carbon. The activity was expressed as the
steadyꢀstate CO conversion, and light olefin selectivity (wt.%) was calculated by olefin fraction in
overall hydrocarbon products and olefin to paraffin ratio (O/P) in the C2ꢀC4 fraction.
3. Results and discussion
3.1 Textural properties of the catalysts
The BET surface area, pore volume and average pore diameter of Fe/Zr catalysts are shown in Table
1. It could be seen that all the samples exhibited relatively low surface area compared with other
related studies [17, 20]. An increase in surface area and decrease in pore size were observed on all Zr
promoted samples compared with that of the zirconiaꢀfree catalysts. The samples prepared by the

Page 7 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
twoꢀstep hydrothermal method led to a decrease in BET surface area compared with that of oneꢀstep
FeꢀZr catalysts. The pore volume was decreased with increasing zirconia content upon oneꢀstep Fe/Zr
catalysts. This phenomenon was not obvious for the twoꢀstep Fe/Zr catalysts. Potassium promotion
caused slight decrease in BET surface area.
3.2 Morphology of the fresh catalysts
Fig.1 shows the TEM images of Fe2O3 and Fe/Zr samples after calcination at 500 ºC. It could be
seen that the asꢀprepared Fe2O3 samples with smooth surface were nearly spherical with average
particles size in about 150 nm as shown in Fig.1a. The use of ultrasonic dispersion condition in the first
step for Fe2O3 sample was to reduce the agglomeration of iron particles. No change in the particle size
distribution was observed after dispersion as shown in Fig.1b, suggesting that the particles exhibited
good mechanical stability, which was beneficial to the homogenous precipitation of ZrO2 on iron
surface during the second step preparation procedure. Separate zirconia and iron oxide particles or their
complex over OꢀFe/Zr[3ꢀ1] sample could be speculated when using oneꢀstep hydrothermal
precipitation and the particles became obviously inhomogeneous as shown in Fig.1c. The Fe2O3
particles with 1.45±0.05 Å d spacing corresponding to the (311) lattice plane of hematite (PDF#
33ꢀ0664) could be seen in Fig.1e. Zirconium oxide mainly exposed its (200) facet with a d spacing of
1.79±0.05 Å (PDF# 80ꢀ0784) (Fig.1e). Uniform zirconia particles about 20 nm mainly deposited on the
surface of iron oxide were clearly different from TꢀFe/Zr[3ꢀ1] sample in Fig.1(d). The homogeneous
dispersion of zirconia could efficiently improve the dispersion of iron active sites. As shown in Fig.1f,
the Fe2O3 particles exhibited 1.45±0.05 Å and 1.69±0.05 Å d spacing corresponding to the (300) and
(116) lattice planes of hematite on the sample TꢀFe/Zr[3ꢀ1] (PDF# 33ꢀ0664), respectively, and the ZrO2
(220) facet was observed as indicated by a d spacing of 1.81±0.05 Å (PDF# 72ꢀ1669). The EDS

RSC Advances
Page 8 of 24
View Article Online
DOI: 10.1039/C5RA27712D
elemental composition signals showed the presence of Fe, Zr and O atoms on the catalyst surface for
the OꢀFe/Zr[3ꢀ1] and TꢀFe/Zr[3ꢀ1] samples (Fig.2). The surface Zr/Fe molar ratio in TꢀFe/Zr[3ꢀ1]
sample (Fig.2b) calculated from EDS was 1.71 times higher than that of the OꢀFe/Zr[3ꢀ1] sample
(Fig.2a).
3.3 Reduction behavior of the catalysts
The H2ꢀTPR profiles for FeꢀZr samples with different molar ratio are shown in Fig.3. The reduction
behavior of the samples was distinctly affected by preparation methods and zirconia content. All
samples mainly consisted of three peaks: a, b and c. The first hydrogen consumption peak (a) was in
the range of 300 ºC to 375 ºC and the second peak (b) was at around 400 ºC. Particularly in TꢀFe/Zr
samples, the peak (a) was more visible compared with OꢀFe/Zr and pure Fe2O3 samples. Meanwhile,
both peaks (a) and (b) shifted to higher temperatures with increasing zirconia content. It was clear that
the reduction peak of (a) and (b) shifted to high temperatures compared with the bulk αꢀFe2O3 sample
at 410 ºC. The third one (c) showed a broad reduction feature at temperatures above 550 ºC, lowered
than that of the αꢀFe2O3 sample at around 700 ºC. For the samples with the same Fe/Zr molar ratio, the
reduction peaks (b) and (c) for all OꢀFe/Zr catalyst samples were slightly higher than that of TꢀFe/Zr
catalyst samples. The iron species may exist in the following different states: (i) large particles or bulk
αꢀFe2O3; (ii) surface species formed by the interaction of iron oxide with zirconia [19, 21] and (iii)
small iron particles on catalyst surface. The peak (a) might be the reduction of small iron particles with
high dispersion on the surface of the catalysts. The peak (b) was for the bulk αꢀFe2O3 to Fe3O4.
Meanwhile, because the FeO is unstable and easily decomposes into metallic Fe and magnetite without
hydrogen consumption [15, 20, 22], the peak (c) at the range of 500 ºC to 725 ºC included the
following stages simultaneously: (i) Fe3O4 → (FeO) → αꢀFe and (ii) FeO → Fe3O4 → (FeO) → αꢀFe

Page 9 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
[22]. The reduction temperature of TꢀFe/Zr catalysts was lower than that of OꢀFe/Zr catalysts,
suggesting that the different dispersion states of iron oxide for the two kinds of catalysts resulted in
different interactions between iron and zirconia. For OꢀFe/Zr catalyst samples, the Fe3+ cations doped
in the zirconia lattice facilitated the formation of FeꢀZr mixed oxide, and then strengthened the
interaction between iron and zirconia during the oneꢀstep hydrothermal procedure [23], leading to the
decrease of reducibility of the catalysts and also the different phase structures of zirconia compared
with twoꢀstep hydrothermal process as shown in XRD measurement (see Fig.4).
3.4 Phase structure of the catalysts
The XRD patterns of the fresh samples are illustrated in Fig.4. The diffraction peaks assigned to
αꢀFe2O3 were detected for all the fresh samples. It was evident that the TꢀFe/Zr[1ꢀ1] sample before and
after potassium promotion were preferred orientation along the (ꢀ111) and (111) crystal faces of ZrO2,
compared with the ZrO2 (101) direction on OꢀFe/Zr[1ꢀ1] sample. For ZrO2, the (ꢀ111) and (111)
reflections belong to the stable monoclinic phase [23, 24]. The (101) reflection of the metastable
tetragonal phase were easily formed during the oneꢀstep precipitation process, which was stabilized due
to the incorporation of Fe3+ ions to ZrO2 to form Fe/Zr mixed oxides [24, 25]. For the used samples as
shown in Fig.5, Fe3O4 was found in all the K promoted samples after reaction (Fe/K=10:1), and iron
carbides FexC with low diffraction intensity was observed in all the used samples. The ZrO2 phases
remained high stable before and after reaction. For all the fresh and used catalyst samples, there was no
significant decrease in diffraction intensity after potassium promotion, and no diffractive peaks
associated with potassium carbonate were detected due possibly to its low concentrations and good
dispersion.
3.5 X-ray photoelectron spectroscopy (XPS) measurement

RSC Advances
Page 10 of 24
View Article Online
DOI: 10.1039/C5RA27712D
The surface composition of the samples calculated from XPS data is shown in Table 2. It was shown
that zirconia was enriched on the surface for all the fresh and used samples. However, at the same
Fe/Zr molar ratio, the surface content of zirconia in TꢀFe/Zr samples, as expected, were much higher
than that of OꢀFe/Zr samples, indicating that most of the zirconia precipitated equably on iron surface
during the second precipitation process as proved by TEM measurement. Although zirconia was known
not to pronouncedly interact with alkali metals [15], the electronic effect of potassium on zirconia for K
promoted samples, revealed by Zr (3d) XPS spectra in Fig.6, caused a slight increase in Zr/Fe molar
ratio on the surface of the catalysts. Potassium was enriched for all the samples after reaction compared
with that of the fresh samples.
The XPS spectra of Zr (3d) and Fe (2p) are shown in Fig. 6 and Fig. 7, respectively. It was obvious
that the preparation methods had great influence on the chemical environment of surface zirconia and
iron. The Zr (3d) peaks for TꢀFe/Zr[1ꢀ1] sample shifted to lower binding energy at 181.8 eV compared
with that of OꢀFe/Zr[1ꢀ1] catalyst at 182.1 eV as shown in Fig.6, indicating a higher electron density
and a weaker interaction between iron oxide and zirconia when using the twoꢀstep hydrothermal
method, which was also proved by the results of H2ꢀTPR and XRD measurements. The Zr (3d) peaks
shifted toward lower position with potassium promotion (Fe/K=10/1). It was reasonable that the
electronic effect of potassium increased the electron density around zirconia. The Fe (2p) binding
energy and the peak shape indicated that Fe species were characteristic of Fe3+ ions in Fe2O3 for all the
samples [26] (see Fig. 7a). The binding energy of the Fe (2p3/2) showed peaks at 710.9 eV and Fe (2p1/2
peaks were detected at around 724.5 eV, respectively. It was known that electrons in the outer Fe (3p)
were sensitive to the variation of electronic structures [27, 28]. For Fe (2p) region, a little decrease in
binding energy could be observed with increase in Zr content for both OꢀFe/Zr and TꢀFe/Zr samples.
)

Page 11 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
Especially for OꢀFe/Zr[5ꢀ1] sample, the Fe (2p) binding energy was higher than those of other samples,
indicating that the electron density of surface iron atom declined due to the zirconia incorporated into
iron oxide to form stronger interaction between iron and zirconia at low zirconia content by the
oneꢀstep hydrothermal process. The binding energy of Fe (2p) shifting towards lower position was also
observed on K promoted samples. For the used samples, it was evident that the Zr (3d) and Fe (2p) and
signals were attenuated by carbon deposit as shown in Fig.7 and Fig.8b. The absence of the satellite
peak at approximately 718.9 eV was evidence for the transformation of Fe2O3 to Fe3O4 or iron carbides
(Fig.7b) [29]. No signal for carbidic carbon at around 283.3 eV was fitted in the form of iron carbide on
the catalyst surface (Fig.8), although a small amount of iron carbides was detected in the bulk structure
by XRD measurement as shown in Fig.5.
3.6 Catalytic performance
The catalytic performance of the catalysts for CO hydrogenation is summarized in Table 3. All
catalysts exhibited high activity (CO % > 90 %) and olefin selectivity (O/P > 4.5). The catalyst samples
prepared by the twoꢀstep hydrothermal process exhibited relatively low activity but higher selectivity
=
+
of C2 ꢀC4= and lower selectivity of C5 than that of the samples prepared by the oneꢀstep hydrothermal
process with Fe/Zr molar ratio of 3:1 and 1:1, indicating that the catalyst samples with high zirconia
content and highly dispersed Zr on iron surface were favorable to the enhancement in olefin selectivity
+
and the decrease in C5 selectivity, especially for the TꢀFe/Zr[1:1]ꢀK sample. The relatively weak
metalꢀsupport interaction and potassium enrichment on iron surface would promote the dissociation of
CO and the formation of iron carbide active phase in FTS [30], and simultaneously enhanced the
olefin/paraffin ratio and the chain growth probability [6]. However, it is worth mentioning that the
production of C5+ hydrocarbons was significantly inhibited over TꢀFe/ZrꢀK catalysts with increase of Zr

RSC Advances
Page 12 of 24
View Article Online
DOI: 10.1039/C5RA27712D
content. The twoꢀstep hydrothermal process caused a higher Zr/Fe and K/Fe ratio on iron surface
compared with that of oneꢀstep method (see Table 2), which then promoted the disperse of active iron
species to inhibit the chain propagation ability and improve the olefin selectivity. Although the CO
conversion on TꢀFe/ZrꢀK catalysts was lower than that of the OꢀFe/ZrꢀK catalysts, there were no great
differences in CO2 selectivity on both catalysts.
The effect of potassium content on olefin selectivity over the two catalyst samples is shown in Fig.9.
The olefin/paraffin (O/P) ratios of TꢀFe/Zr[1ꢀ1]ꢀK catalyst were higher than that of OꢀFe/Zr[1ꢀ1]ꢀK
catalysts under the same Fe/K ratios and decreased with the reduce in potassium contents. It was
obvious that the TꢀFe/Zr[1ꢀ1]ꢀK catalysts showed better property in the improvement of olefin
selectivity than OꢀFe/Zr[1ꢀ1]ꢀK catalysts. As shown in Fig.10, the two catalyst samples exhibited high
activity and good stability with 168 h time on stream, although a slight decrease in activity was
observed for both of the two samples. Also, the CO conversion of TꢀFe/Zr[1ꢀ1]ꢀK catalyst (Fe/K=20:1)
was slightly lower than that of OꢀFe/Zr[1ꢀ1]ꢀK catalyst due to a large amount of zirconia precipitated
on iron surface. As shown in Fig.11, the TꢀFe/Zr[1ꢀ1]ꢀK catalyst (Fe/K=10:1) showed higher olefin
selectivity than that of OꢀFe/Zr[1ꢀ1]ꢀK catalyst, especially preferably to enhance the ethene selectivity,
suggesting that the secondary hydrogenation reaction was inhibited on the samples prepared by the
twoꢀstep process, due to the poor hydrogenation ability of a large amount of ZrO2 deposited on iron
surface [16], as confirmed by forementioned TEM and XPS measurement. As a whole, the present
experiment results reveal that the catalytic performance of FeꢀZr catalysts can be greatly improved by
different catalyst preparation protocols. The resulting surface structure modification of FeꢀZr catalysts
is related to the improvement in the olefin selectivity and product distribution in FTS.
4. Conclusions

Page 13 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
The different zirconia addition procedures for the preparation of FeꢀZr catalysts have significant
influence on the physical and chemical properties of iron catalysts. The TꢀFe/Zr samples prepared by
twoꢀstep hydrothermal process caused different phase structures of ZrO2, weakened the interaction
between iron and zirconia and increased the reducibility of bulk Fe2O3 phase compared with the
OꢀFe/Zr samples by oneꢀstep hydrothermal process. Further promotion by potassium decreased the
BET surface area of the catalyst samples. Increasing zirconia content in TꢀFe/ZrꢀK catalysts was
beneficial to disperse iron active sites by small and uniform zirconia particles distributed on iron
surface. The enrichment of zirconia and potassium reduced the secondary hydrogenation ability to
improve the olefin selectivity and suppressed the formation of heavy hydrocarbons during CO
hydrogenation.
Acknowledgment
The authors gratefully acknowledged the financial supports from the Key Project of Natural Science
Foundation of Ningxia (NZ13010) and the National Natural Science Foundation of China (21366025).
References
[1] H. M. Torres Galvis and K. P. de Jong, ACS. Catal., 2013, 3, 2130ꢀ2149.
[2] H. M. Torres Galvis, A. C. J. Koeken, J. H. Bitter, T. Davidian, M. Ruitenbeek, A. Iulian Dugulan
and K. P. de Jong, Catal.Today, 2013, 215, 95ꢀ102.
[3] O. O. James, B. Chowdhury, M. A. Mesubi and S. Maity, RSC. Adv., 2012, 2, 7347ꢀ7366.
[4] H. M. Torres Galvis, J. H. Bitter, C. B. Khare, M. Ruitenbeek, A. Iulian Dugulan and K. P. de Jong,
Science, 2012, 335, 835ꢀ838.
[5] ꢁ. ÖzkaraꢀAydınoğlu, Ö. Ataç, Ö. F. Gül, ꢁ. Kınayyiğit, S. ꢁal, M. Baranak and Đ. Boz, Chem. Eng.
J., 2012,181-182, 581ꢀ589.

RSC Advances
Page 14 of 24
View Article Online
DOI: 10.1039/C5RA27712D
[6] G .C. Wang, K. Zhang, P. Liu, H. T Hui and Y. S. Tan, J. Ind. Eng.Chem., 2013, 19, 961ꢀ965.
[7] H. L. Wang, Y. Yang, J. Xu, H. Wang, M. Y. Ding and Y. W. Li, J. Mol. Catal. A: Chem., 2010,
326, 29ꢀ40.
[8] J. B. Li, H. F. Ma, H. T. Zhang, Q. W. Sun, W. Y. Ying and D. Y. Fang, Reac. Kinet. Mech. Catal.,
2014, 112, 409ꢀ423.
[9] S. H. Kang, J. W. Bae, P. S. S. Prasad, S. J. Park, K. J. Woo and K. W. Jun, Catal. Lett., 2009, 130,
630ꢀ636
[10] S. Li, A. Li, S. Krishnamoorthy and E. Iglesia, Catal. Lett., 2001, 77, 197ꢀ205.
[11] S. K. Das, S. Majhi, P. Mohanty and K. K. Pant, Fuel Proc. Technol., 2014, 118, 82ꢀ89.
[12] A. A. Mirzaei, S. Vahid and H. O. Torshizi, J. Nat. Gas Sci. Eng., 2013, 15, 106ꢀ117.
[13] R. Snel, Catal. Rev., 1987, 29, 361ꢀ445.
[14] R. A. Van Santen, Acc. Chem. Res., 2009, 42, 57ꢀ66.
[15] F. R. van den Berg, M. W. J. Crajé, A. M. van der Kraan and J. W. Geus, Appl. Catal. A: Gen.,
2003, 242, 403ꢀ416.
[16] W. Zhang, R. Gao, C. Su and Y. Yin, Stud. Surf. Sci. Catal., 1993, 75, 2793ꢀ2796.
[17] X. J. Su, J. L. Zhang, S. B. Fan, Q. X. Ma and T. S. Zhao, RSC Adv., 2015, 5, 80196ꢀ80202.
[18] F. J. Berry, S. Jobsen and M. R. Smith, Hyperfine Interact., 1989, 46, 607ꢀ611.
[19] K. Chen, Y. Fan, Z. Hu and Q. Yan, Catal. Lett., 1996, 36, 139ꢀ144.
[20] M. AlꢀDossary and J. L. G. Fierro, Appl. Catal. A: Gen., 2015, 499, 109ꢀ117.
[21] X. Gao, J. Shen, Y. Hsia and Y. Chen, J. Chem. Soc. Faraday Trans., 1993, 89, 1079ꢀ1084.
[22] W. K. Jozwiak, E. Kaczmarek, T. P. Maniecki, W. Ignaczak and W. Maniukiewicz, Appl. Catal. A:
Gen., 2007, 326, 17ꢀ27

Page 15 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
[23] P. D. L. Mercera, J. G. van Ommen, E. B. M. Doesburg, A. J. Burggraaf and J. R. H. Ross, Appl.
Catal., 1991, 71, 363ꢀ391.
[24] P. D. L. Mercera, J. G. van Ommen, E. B. M. Doesburg, A. J. Burggraaf and J. R. H. Ross, Appl.
Catal., 1990, 57, 127ꢀ148.
[25] P. D. L. Mercera, J. G. van Ommen, E. B. M. Doesburg, A. J. Burggraaf and J. R. H. Ross, Appl.
Catal., 1991, 78, 79ꢀ96.
[26] S. D. Qin, C. H. Zhang, J. Xu, Y. Yang, H. W. Xiang and Y. W. Li, Appl Catal A: Gen., 2011, 392,
118ꢀ126.
[27] H. Y. Suo, S. G. Wang, C. H. Zhang, J. Xu, B. S. Wu, Y. Yang, H. W. Xiang and Y. W. Li, J.
Catal., 2012, 286, 111ꢀ123.
[28] Y. Liu, F. Chen and Y. Zhang, Reac. Kinet. Mech. Cat., 2015, 114, 433ꢀ449.
[29] J. P. Baltrus, J. R. Diehl, M. A. Mcdonald and M. F. Zarochak, Appl. Catal., 1989, 48, 199ꢀ213.
[30] M. L. Cubeiro, H. Morales, M. R. Goldwasser, M. J. PérezꢀZurita, F. GonzálezꢀJiménez and
C.Urbina de N, Appl. Catal. A: Gen., 1999, 189, 87ꢀ97.

RSC Advances
Page 16 of 24
View Article Online
DOI: 10.1039/C5RA27712D
Figures and Table captions
Fig.1 TEM images of the samples
Fig.2 EDS spectra of the samples
Fig.3 H2ꢀTPR profiles of the samples
Fig.4 XRD patterns of the fresh samples
Fig.5 XRD patterns of the used samples
Fig. 6 Zr (3d) XPS spectra of the samples
Fig. 7 Fe (2p) XPS spectra of the catalysts (a) fresh and (b) used samples
Fig.8 C (1s) XPS spectra of the used samples
Fig.9 Effect of K content on light olefin selectivity over Fe/Zr[1ꢀ1]ꢀK samples
Fig.10 Stability of the catalysts with time on stream
Fig.11 Effect of preparation process on C2ꢀC4 olefin selectivity
Table 1 Textural property of the fresh catalyst samples
Table 2 Surface composition of different samples
Table 3 Catalytic activity of the catalysts for CO hydrogenation

Page 17 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
Figures
Fig.1 TEM images of the samples
(a) Fe2O3ꢀbefore dispersion; (b) Fe2O3ꢀafter dispersion; (c) OꢀFe/Zr[3ꢀ1]; (d) TꢀFe/Zr[3ꢀ1]; (e)
HRTEM image of OꢀFe/Zr[3ꢀ1]; (f) HRTEM image of TꢀFe/Zr[3ꢀ1]

RSC Advances
Page 18 of 24
View Article Online
DOI: 10.1039/C5RA27712D
Fig.2 EDS spectra of the samples
(a) OꢀFe/Zr[3ꢀ1]; (b) TꢀFe/Zr[3ꢀ1]
c
b
b
c
a
a
O-Fe/Zr[5-1]
O-Fe/Zr[3-1]
b
c
a
O-Fe/Zr[1-1]
b
c
b
a
c
a
b
T-Fe/Zr[5-1]
T-Fe/Zr[3-1]
c
a
b
c
T-Fe/Zr[1-1]
Fe2O3
a
150 200 250 300 350 400 450 500 550 600 650 700 750 800 850
T(℃)
Fig.3 H2-TPR profiles of the samples

Page 19 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
●
ZrO2
T-Fe/Zr[1-1]-K
T-Fe/Zr[1-1
(-111)
●
(111)
●
●
]
T-Fe/Zr[3-1]
T-Fe/Zr[5-1]
(101)
●
●
●
O-Fe/Zr[1-1]
O-Fe/Zr[3-1]
O-Fe/Zr[5-1]
Fe2O3
0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
2-Theta(o)
Fig.4 XRD patterns of the fresh samples
●
▼
ZrO2
FexC
●
●
▼
T-Fe/Zr[1-1]-K
T-Fe/Zr[3-1]-K
●
▼
●
T-Fe/Zr[5-1]-K
●
O-Fe/Zr[1-1]-K
O-Fe/Zr[3-1]-K
●
●
O-Fe/Zr[5-1]-K
Fe3O4
0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
2-Theta(o)
Fig. 5 XRD patterns of the used samples

RSC Advances
Page 20 of 24
View Article Online
DOI: 10.1039/C5RA27712D
182.1 eV
181.8 eV
O-Fe/Zr[1-1]-K-Used
O-Fe/Zr[1-1]-K
O-Fe/Zr[1-1]
O-Fe/Zr[3-1]
O-Fe/Zr[5-1]
T-Fe/Zr[1-1]-K-Used
T-Fe/Zr[1-1]-K
T-Fe/Zr[1-1]
T-Fe/Zr[3-1]
T-Fe/Zr[5-1]
174
176
178
180
182
184
186
188
190
192
194
Binding Energy(eV)
Fig. 6 Zr (3d) XPS spectra of the samples
710.9 eV
710.5 eV
(a)
O-Fe/Zr[5-1]
O-Fe/Zr[3-1]
O-Fe/Zr[1-1]
O-Fe/Zr[1-1]-K
T-Fe/Zr[5-1]
T-Fe/Zr[3-1]
T-Fe/Zr[1-1]
T-Fe/Zr[1-1]-K
700
710
720
730
740
Binding Energy (eV)
710.3 eV
(b)
O-Fe/Zr[1-1]-K
T-Fe/Zr[3-1]-K
T-Fe/Zr[1-1]-K
700
710
720
730
740
Binding Energy (eV)
Fig.7 Fe (2p) XPS spectra of the catalysts (a) fresh and (b) used samples

Page 21 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
284.79 eV
O-Fe/Zr[1-1]-K
T-Fe/Zr[1-1]-K
280
282
284
286
288
290
Binding Energy (eV)
Fig.8 C (1s) XPS spectra of the used samples
T-Fe/Zr[1-1]-K
O-Fe/Zr[1-1]-K
5
4
3
2
1
10
20
30
Fe/K molar ratios
Fig.9 Effect of K content on light olefin selectivity over Fe/Zr[1-1]-K samples
Reaction conditions: H2/CO = 2, 320 ºC, 1.5MPa and 1000 hꢀ1

RSC Advances
Page 22 of 24
View Article Online
DOI: 10.1039/C5RA27712D
100
95
Fe/K=20:1
O-Fe/Zr[1-1]-K
T-Fe/Zr[1-1]-K
90
85
80
48
72
96
120
144
168
Time on Stream (h)
Fig. 10 Stability of the catalysts with time on stream
Reaction conditions: H2/CO = 2, 320 ºC, 1.5MPa and 1000 hꢀ1.
8
O-Fe/Zr[1-1]-K
T-Fe/Zr[1-1]-K
7
6
5
4
3
2
1
0
C4=/C40
C2=/C20
C3=/C30
Fig.11 Effect of preparation process on C2-C4 olefin selectivity
Reaction conditions: H2/CO=2, 320 ºC, 1.5MPa, 1000 hꢀ1, TOS=120 h.
Tables

Page 23 of 24
RSC Advances
View Article Online
DOI: 10.1039/C5RA27712D
Table 1 Textural property of the fresh catalyst samples
Catalyst
BET surface
area A / ( m2·gꢀ1 )
15.12
Pore volume
v/( cm3·gꢀ1 )
0.13
Average pore
size d / nm
34.07
17.98
22.03
9.56
a
Fe2O3
OꢀFe/Zr[5ꢀ1]
OꢀFe/Zr[3ꢀ1]
OꢀFe/Zr[1ꢀ1]
TꢀFe/Zr[5ꢀ1]
TꢀFe/Zr[3ꢀ1]
TꢀFe/Zr[1ꢀ1]
TꢀFe/Zr[1ꢀ1]ꢀK b
42.26
0.19
32.32
0.18
55.43
0.13
26.37
0.12
18.45
10.15
15.90
20.18
33.12
0.08
29.30
0.12
22.81
0.11
a: Before dispersion. b: Fe/K molar ratio=10:1.
Table 2 Surface composition of different samples
Surface atom content (mol.%) a
Samples
Fe
Zr
K
O
C
Zr/Fe
Fe/K
OꢀFe/Zr[5ꢀ1]
OꢀFe/Zr[3ꢀ1]
OꢀFe/Zr[1ꢀ1]
OꢀFe/Zr[1ꢀ1]ꢀK
OꢀFe/Zr[1ꢀ1]ꢀKb
TꢀFe/Zr[5ꢀ1]
16.34
12.81
10.00
5.34
8.15
/
53.36
53.04
47.74
39.97
34.48
47.92
40.84
38.32
46.36
9.42
22.14
21.10
29.45
41.46
45.64
30.19
42.29
42.82
27.81
85.82
0.50
1.02
1.28
1.70
2.23
1.06
2.16
4.94
4.95
5.80
/
13.05
12.80
9.09
/
/
/
/
4.15
1.29
4.06
9.05
6.77
0.60
10.65
5.33
11.25
11.53
15.69
17.74
1.94
/
/
TꢀFe/Zr[3ꢀ1]
/
/
TꢀFe/Zr[1ꢀ1]
3.17
/
/
TꢀFe/Zr[1ꢀ1]ꢀK
TꢀFe/Zr[1ꢀ1]ꢀKb
3.59
4.51
2.48
0.80
0.14
0.34
a: Calculated from the peak area of XPS spectra; b: Used samples
Table 3 Catalytic activity of the catalysts for CO hydrogenation
CO
Sel. C mol/%
Product wt /%
Samples a
O/P
=
=
0
0
+
Conv.%
CH4
CO2
CH4
C2 ꢀC4
C2 ꢀC4
C5
OꢀFe/Zr[5ꢀ1]ꢀK
OꢀFe/Zr[3ꢀ1]ꢀK
OꢀFe/Zr[1ꢀ1]ꢀK
TꢀFe/Zr[5ꢀ1]ꢀK
TꢀFe/Zr[3ꢀ1]ꢀK
TꢀFe/Zr[1ꢀ1]ꢀK
97.23
97.47
97.98
95.28
96.04
90.03
13.22
12.65
13.01
14.55
13.46
14.18
20.83
28.76
20.76
22.29
25.75
20.11
19.00
20.37
18.79
21.45
20.50
19.99
36.52
37.89
38.99
38.37
40.87
43.84
7.14
7.56
8.11
7.68
7.59
8.38
37.34 5.11
34.18 5.01
34.11 4.80
32.49 5.00
31.04 5.38
27.79 5.23
Reduction conditions: H2/CO=2, 0.1 MPa, 1000hꢀ1, 300 ºC for 4 h.
Reaction conditions: H2/CO=2, 320 ºC, 1.5MPa, 1000 hꢀ1, TOS=48h.
a: Fe/K molar ratio =10:1.

RSC Advances
Page 24 of 24
View Article Online
DOI: 10.1039/C5RA27712D
Graphical abstract
Two-step hydrothermal preparation of Fe/Zr-K catalysts could improve the olefin selectivity and
product distribution by reducing the secondary hydrogenation ability and suppressing the
formation of heavy hydrocarbons during CO hydrogenation.