Syngas conversion to ethanol over a mesoporous Cu/MCM-41 catalyst: Effect of K and Fe promoters

Article abstract of DOI:10.1016/j.apcata.2016.08.006

Transportation fuels such as ethanol can be obtained through thermochemical processing of biomass. Interest in the development of more selective catalysts for the conversion of biomass-derived syngas (H₂?+?CO) to ethanol is increasing in both academia and industry. In this work, we have evaluated the performances of K and Fe as metal promoters of a mesoporous Cu/MCM-41 catalyst and their effects on the product selectivity and especially on ethanol formation. The metal loading was 29?wt.% Cu, 2?wt.% Fe and 1.6?wt.%?K. The catalysts were tested at 300?°C, 20?bar and gas-hourly-space-velocities in the range of 1500–30000?ml_syngas/g_cat h; under these conditions the syngas conversion level was between 2 and 11%. The non-promoted Cu/MCM-41 catalyst showed interesting selectivity toward oxygenated compounds, mostly methanol. The addition of K as promoter increases the selectivity toward methanol even more, while the addition of Fe as promoter favors the formation of hydrocarbon compounds. When both K and Fe as promoters are incorporated into the Cu/MCM-41 catalyst, the reaction rate to oxygenated compounds is notably increased, especially for ethanol. The space time yield for ethanol for the Cu/MCM-41 catalyst is 0.3?×?10^?5 carbon-mol/g_cath which increases to 165.5?×?10^?5 carbon-mol/g_cath for the Cu-Fe-K/MCM-41 catalyst. From XPS analysis, the Cu-Fe-K/MCM-41 catalyst was found to have the following atomic composition: Cu_0.34Fe_0.08K_0.08Si_1.00. The promoting effect of both K and Fe, may be related to an increased reaction rate toward CO non-dissociation and CO-dissociation paths, respectively, which is beneficial for the ethanol formation. Further catalytic results, catalyst characterization and discussion of results are presented in this work.

Full text of DOI:10.1016/j.apcata.2016.08.006

Accepted Manuscript
Title: Syngas conversion to ethanol over a mesoporous
Cu/MCM-41 catalyst: effect of K and Fe promoters
Author: Luis Lopez Vicente Montes Henrik Ku sˇ ar Saul
Cabrera Magali Boutonnet Sven J a¨ r a˚ s
PII:
S0926-860X(16)30413-6
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2016.08.006
APCATA 15969
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
8-4-2016
2-8-2016
4-8-2016
Please cite this article as: Luis Lopez, Vicente Montes, Henrik Ku sˇ ar, Saul Cabrera,
Magali Boutonnet, Sven J a¨ r a˚ s, Syngas conversion to ethanol over a mesoporous
Cu/MCM-41 catalyst: effect of K and Fe promoters, Applied Catalysis A, General
http://dx.doi.org/10.1016/j.apcata.2016.08.006
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.

Syngas conversion to ethanol over a mesoporous Cu/MCM-41
catalyst: effect of K and Fe promoters
Luis Lopez*,a,b, Vicente Montes , Henrik Kušar , Saul Cabrera , Magali Boutonnet , Sven Järås
c
a
b
a
a
aKTH Royal Institute of Technology, Chemical Technology, Teknikringen 42, SE-100 44, Stockholm, Sweden.
bUMSA Universidad Mayor de San Andrés, Instituto de Investigaciones Químicas, Campus Universitario, La
Paz, Bolivia.
cOrganic Chemistry Department, University of Córdoba, ceiA3, Marie Curie Building, E-14014 Córdoba,
Spain.
Corresponding author: *lgln@kth.se
Graphical abstract
600
500
400
300
200
100
0
45
4
0
5
3
30
2
5
0
2
15
1
5
0
0
Cu/MCM-41
Cu-K/MCM-41
Cu-Fe/MCM-41 Cu-Fe-K/MCM-41
Highlights:



The effect K and Fe as metal promoters on mesoporous Cu/MCM-41 catalyst is studied.
K favors the formation of oxygenates and Fe favors the formation of hydrocarbons.
The combination of K and Fe on the Cu/MCM-41 catalyst notably increases the ethanol yield.

Abstract
Transportation fuels such as ethanol can be obtained through thermochemical processing of biomass. Interest
in the development of more selective catalysts for the conversion of biomass-derived syngas (H +CO) to
2
ethanol is increasing in both academia and industry. In this work, we have evaluated the performances of K
and Fe as metal promoters of a mesoporous Cu/MCM-41 catalyst and their effects on the product selectivity
and especially on ethanol formation. The metal loading was 29 wt.% Cu, 2 wt.% Fe and 1.6 wt.% K. The
catalysts were tested at 300 °C, 20 bar and gas-hourly-space-velocities in the range of 1500-30000
mlsyngas/gcat h; under these conditions the syngas conversion level was between 2 and 11%. The non-
promoted Cu/MCM-41 catalyst showed interesting selectivity toward oxygenated compounds, mostly
methanol. The addition of K as promoter increases the selectivity toward methanol even more, while the
addition of Fe as promoter favors the formation of hydrocarbon compounds. When both K and Fe as
promoters are incorporated into the Cu/MCM-41 catalyst, the reaction rate to oxygenated compounds is
notably increased, especially for ethanol. The space time yield for ethanol for the Cu/MCM-41 catalyst is
0
.3x10-5 carbon-mol/gcath which increases to 165.5x10-5 carbon-mol/gcath for the Cu-Fe-K/MCM-41 catalyst.
From XPS analysis, the Cu-Fe-K/MCM-41 catalyst was found to have the following atomic composition:
0.08Si1.00. The promoting effect of both K and Fe, may be related to an increased reaction rate
Cu0.34Fe0.08
K
toward CO non-dissociation and CO-dissociation paths, respectively, which is beneficial for the ethanol
formation. Further catalytic results, catalyst characterization and discussion of results are presented in this
work.
Keywords: copper, ethanol, mesoporous MCM-41, promoters, syngas.
1. Introduction
During the last few years, biofuels have become an important renewable fuel for the transportation sector.
Biofuels are gaseous or liquid fuels obtained from the processing of renewable resources such as biomass.
Biomass includes forest and agricultural biomass, industrial waste streams and municipal solid waste. Almost
all kinds of biomass can be converted to useful fuels via the thermochemical route [1], in which the primary
biomass is gasified to produce a mixture of gases know as synthesis gas (syngas) formed by hydrogen,
carbon monoxide and a minor amount of carbon dioxide. The syngas is thereafter catalytically converted to
the desired fuel, such as FT diesel, SNG (synthetic natural gas), ethanol and others using a heterogeneous
catalyst.
Ethanol grade fuel constitutes most of the biofuels consumed today. World ethanol production is forecast to
increase from about 100 billion liters in 2013 to 158 billion liters in 2023 [2]. The thermochemical approach
may become an important route for producing ethanol from biomass for large-scale production. The
gasification of biomass is the most studied technology for producing syngas, where some aspects inherent to
biomass are being evaluated, such as contaminants like chlorine, sulfur and alkali metals [1]. Several kinds of
catalysts have been tested for the conversion of syngas to ethanol and higher alcohols (C2+-OH) [3-5], among

which the Rh, Mo and Cu-based catalysts are the most promising up to date. Nevertheless, further research is
needed to gain insights into the conversion of syngas to ethanol, which will allow designing a more selective
and active catalyst.
It has been found that the catalyst support has an important role in the synthesis of ethanol and higher
alcohols. For example, silica support is beneficial for alcohol formation, which seems to be due to a different
surface composition of the adsorbed CO as compared to on other catalyst supports [3]. Silica meso-structured
2
support, such as MCM-41, is a novel material with high surface area (around 1000 m /g), which can be used
to improve the dispersion of the active sites. In previous works, we have used MCM-41 as support for a Rh-
based catalyst with interesting results [6, 7]. To the best of our knowledge, there are no other studies using
MCM-41 as support for Cu-based catalysts. In order to increase the selectivity to ethanol and higher alcohols
the Cu-based catalysts are promoted by, for example, alkali and transition metals [3-5]. Potassium, K, as a
promoter has been successfully used to improve the selectivity to higher alcohols over Cu/ZnO/Al
2
O
3
catalysts
[8, 9]. Another metal promoter with good results for alcohols formation is Fe, which facilitates the carbon chain
growth similarly as occurs in the Fischer-Tropsch synthesis [1]. Thus, a combination of K and Fe promoters
would be beneficial for both ethanol and higher alcohols formation.
The synthesis of higher alcohols is, to some extent, similar to the Fischer-Tropsch Synthesis (FTS) where
chain polymerization occurs to form mostly hydrocarbon compounds [10]. The FTS is in part described by the
model of Anderson-Shultz-Flory (ASF), defined by equation (1) [10, 11]:
푆푛
훼ꢀ−1(ꢁ ꢂ 훼)2
=
(1)
(2)
ꢀ
(1−ꢆ)ꢇ
푆
푛
푙ꢃ ꢄ ꢅ = 푙ꢃ
+ 푙ꢃ(훼)ꢃ
ꢀ
ꢆ
where S
n
is the selectivity to molecules with n carbon atoms and  is the chain growth propagation probability
(
independent of n). The value of  is obtained by least-square linear regression of the logarithmic form of the
ASF equation (2). The chain growth propagation probability  is directly related to the rate of chain
propagation and inversely to the rate of chain termination.
In this work, we have evaluated the catalytic performance of the following catalysts: Cu/MCM-41, Cu-
Fe/MCM-41, Cu-K/MCM-41 and Cu-Fe-K/MCM-41. Various characterization techniques (XRD,
N
2
physisorption, TEM, TPR, ATG and XPS) were employed to examine the physico-chemical properties of the
prepared catalysts. The catalytic experiments indicate that the incorporation of Fe-K into the Cu/MCM-41
catalyst favors the selectivity to oxygenated compounds and, more interestingly, the space time yield for
ethanol increases notably.
2. EXPERIMENTAL
2.1. Catalyst Preparation and Characterization Techniques

For the preparation of the catalyst support, the “atrane route” [34] was used to obtain a hexagonal
mesoporous silica, type MCM-41, where cetyltrimethylammonium bromide was used as the structural
directing agent. A characteristic of this method is the use of a hydrolysis-retarding agent such as
triethanolamine (TEA). TEA and tetraethyl orthosilicate were mixed to obtain silatrane complexes, which are
the precursors of Si. Further preparation procedures are described elsewhere [34,35]. The addition of Cu onto
MCM-41 was done by aqueous wetness impregnation using an excess of water under constant stirring for 3 h.
After impregnation, the sample was dried at 120 °C for 3 h and then calcined at 500 °C for 5 h to obtain the
Cu/MCM-41 catalyst. Subsequently, K was added to the Cu/MCM-41 catalyst following the same procedure
as for Cu addition and using the same thermal treatment (described above). Then, Fe was added to the Cu-
K/MCM-41 catalyst also following the same procedure as for the Cu addition and the same thermal treatment.
The metal precursors used were Cu(NO
3
)
2
∙3H
2
O, Fe(NO
3
)
3
∙9H
2
O and K
2
CO purchased from Sigma-Aldrich.
3
Powder X-ray diffraction (XRD) was performed in a Siemens D5000 instrument using Cu K- radiation
2=10-90º, step size=0.04º) equipped with a Ni filter and operated at 40 kV and 30 mA. The metal loading
was estimated by atomic absorption spectroscopy in a Perkin Elmer AAnalyst 200, for which the samples
were dissolved in HF (40%) for 48h at ambient temperature. N physisorption was carried out in a
(
2
Micromeritics ASAP 2000 instrument. The Brunauer-Emmett-Teller (BET) method was used to estimate the
specific surface area and the Barrett-Joyner-Halenda (BJH) method was used to estimate the pore size and
pore volume from the desorption isotherm. Temperature programmed reduction (TPR) was performed in a
Micromeritics Autochem 2910 instrument; a reducing gas mixture formed by 5% H
the catalyst sample at a flow of 50 ml/min while the sample temperature was increased by 5 °C/min up to 900
C. Thermogravimetric analysis (TGA) was carried out using a SETARAM STA (TG-DTA-DSC 1600 °C); a
continuous flow of He was passed through the sample during the analysis. The sample was heated up to 400
C at a rate of 5 °C/min. The differential thermogravimetric (DTG) curves were estimates from the TGA data.
2
/Ar was passed through
°
°
Transmission electron microscopy (TEM) images were collected using a JEOL JEM 1400 microscope. All
samples were mounted on 3 mm holey carbon copper grids. The particle size and distribution were estimated
by examination of more than 100 metal particles. X-ray photoelectron spectroscopy (XPS) data were recorded
using a Leybold-Heraeus LHS10 spectrometer as the main chamber, capable of operating down to 2∙10−9
Torr, which was equipped with an EA-200MCD hemispherical electron analyzer with a dual X-ray source
using AlKα (hv=1486.6 eV) at 120 W, 30 mA, with C(1s) as energy reference (284.6 eV). The powder samples
were gently pressed to form pellets of 4 mm × 4 mm and 0.5 mm thickness, followed by outgassing to a
pressure below 2∙10−8 Torr at 150 °C to remove chemisorbed volatile species. For TEM and XPS analyses the
catalyst samples were previously reduced in a flow of 50 ml/min H
kept at this temperature for 60 min.
2
and heated at 5 °C/min up to 370 °C and
2.2. Catalytic Testing
The catalytic testing was carried out in a stainless-steel, down-flow fixed bed reactor. The internal diameter of
the reactor was 8.3 mm and the system components have been described by Andersson et al. [12]. About 200
mg of the catalyst sample with a particle size of 160-250 m was charged into the reactor. Before the reaction,

the catalysts were reduced under pure hydrogen flowing at 50 ml/min and heated at a rate of 5 °C/min up to
70 °C and kept at this temperature for 60 min. Then, the reaction conditions were fixed to 300 °C and 20 bar.
3
The gas hourly space velocity (GHSV) was varied between 1500 and 30000 mlsyngas/gcat in order to maintain
the reaction conversion at less than 11% and thus avoid mass-transfer limitations. All the experiments were
made with a syngas ratio of H /CO = 2/1.
2
The reaction products were analyzed using a gas chromatograph (Agilent GC-7890A) connected on-line to
the catalytic reactor. A detailed description of the GC configuration and the analytical procedure can be found
elsewhere [12]. N
2
was mixed with syngas which acted as internal standard to quantify the smaller gases
) using a thermal conductivity detector. For the larger gaseous compounds, two flame
(
CH , CO and CO
4
2
ionization detectors (FIDs) were used, one for non-polar gases (hydrocarbon compounds) and the other for
the polar gases (oxygenated compounds). The internal normalization of corrected peak areas method was
used to quantify the hydrocarbon and oxygenated compounds in both FIDs [13]. From these results; the
syngas conversion, carbon-mole selectivity and carbon balance were calculated using the equations shown in
reference [12]. In all the experiments the carbon balance was in the range of 99.4-100%.
3
3
3
. RESULTS AND DISCUSSION
.1. Catalyst Characterization
.1.1. Powder X-ray diffraction (XRD)
The X-ray diffraction patterns of the prepared catalysts are shown in Figure 1. After Cu impregnation, the
signals observed at 35.6*2 and at 38.8*2 correspond to CuO when compared to the powder diffraction of
standard compounds reported in JCPDS (No. 05-661). The addition of K and Fe does not generate new
signals, this could probably be due, in part, to the small amounts used in the impregnation (3%wt. K and
1.6%wt. Fe) and, in part, be due to highly dispersed aggregates of K and Fe, resulting in no appreciable
signals using the XRD technique.
3
.1.2. N
2
phys is orption
Table 1 summarizes the results from N
2
physisorption for the prepared catalysts. It can be noted that the
mesoporous MCM-41 support has a large surface area. The high surface area of the MCM-41 support is due
to the small 1D channel porous structure which is interconnected by a thin wall thickness of around 0.8 nm
[14]. The pore size of the prepared MCM-41 corresponds to the mesoporous range, i.e. 2.9 nm (Table 1).
Moreover, when the metals (Cu, K and Fe) are impregnated onto MCM-41 the surface area decreases. This
suggests a structure degradation possibly due to 1D pores collapsing during the catalyst preparation. The
structure degradation of MCM-41 during catalyst preparation has been also observed by Landau et al. [15]. In
our catalysts, the structure degradation is less noticeable in the Cu/MCM-41, Cu-K/MCM-41 and Cu-Fe/MCM-
41 catalysts and more notable in the Cu-Fe-K/MCM-41 catalyst.
3
.1.3. X-ray photoelectron s pectros copy (XPS)

Atomic ratios obtained from the XPS technique are included in Table 1. The Cu/MCM-41 catalyst has a Si/Cu
ratio of 2.8, which is slightly increased to 2.9 when K is incorporated into the catalyst (Cu-K/MCM-41). When
Fe is incorporated (Cu-Fe/MCM-41) the Si/Cu is increased to 7.0, which may suggest that Fe atoms cover
some Cu particles or that some Cu atoms migrate from the outside to the core of the catalyst particle. Finally,
when both K and Fe are incorporated (Cu-Fe-K/MCM-41) the Si/Cu ratio is remains at 2.9, suggesting that K
favors the dispersion of Cu on the catalyst’s surface. This can be related to some reports suggesting that K
+
stabilizes some Cu sites, presumably Cu species [16, 17].
Figure 3 shows the Cu2p-spectra of the prepared catalysts. The presence of the metal promoters do not
modify the oxidation state of copper since the characteristics peaks, Cu2p3/2 (932.5 eV) and Cu2p1/2 (952.2
0
+1
eV), remain in the same position in all the catalysts. These signals can be attributed to either Cu or Cu [18],
which can be related to the active sites for the catalytic conversion of syngas, and in particular for the
methanol synthesis [19]. There are not signals of any satellite in the range 940-945 eV, indicating that CuO is
not present in the prepared catalysts [18]. The binding energy for each element in the catalysts is: Si=103.2
eV, O= 532.5-532.7 eV, Cu=932.4-932.6 eV, K=293.0-293.2 eV and Fe=711.3 eV.
3
.1.4. Trans mis s ion electron micros copy (TEM)
Figure 4 shows the TEM images of the prepared catalysts. From these images it can be noted that the
incorporation of Fe and K onto the Cu/MCM-41 catalyst does not drastically change the morphology of the
catalyst. This can be related to the low amount of Fe and K used in the catalyst preparation. Moreover, the
average of the metal particle size is about 5 nm for the Cu/MCM-41, Cu-K/MCM-41 and Cu-Fe/MCM-
41catalysts and it is slightly increased to 6 nm for the Cu-Fe-K/MCM-41 catalyst (see Table 1). The average of
the particle size takes in account the contribution of all the metals present in the catalysts.
3
.1.5. Temperature-programmed reduction (TPR)
Figure 5 shows the TPR profiles of the prepared catalysts. The reduction temperature is centered at around
45 °C with a shoulder at 220 °C for the Cu/MCM-41 catalyst. The shoulder in TPR may indicate the formation
2
of strong metal-support interactions, which affect the reduction of the metal oxide [20], also reported for
mesoporous catalysts [21]. When K is added, a single reduction peak is observed which is decreased to
around 210 °C (Cu-K/MCM-41). It seems that the presence of K partially eliminates or weakens some metal-
support interactions between Cu and the MCM-41 support. In the Cu-Fe/MCM-41 and Cu-Fe-K/MCM-41
catalysts a reduction peak centered at 335 °C and 350 °C, respectively, is observed. This reduction peak can
be attributable to the reduction of oxidized species of Fe, possibly Fe
2
O
3
 Fe
3
O [22]. In both catalysts a
4
shoulder at 295 °C is observed which can be attributed to the reduction of CuO, thus CuO in Fe-containing
catalysts is reduced at higher temperature than for the Cu/MCM-41 catalyst. The latter may be due to some
electronic effects induced by the presence of Fe.
3
.1.6. Thermogravimetric analys is (TGA)
The thermograms of the prepared catalysts are shown in Figure 6. The mass loss in the range of 20-152 C
can be attributed to the process of elimination of physisorbed water (dehydration) from the surface of the

catalysts. This mass loss represents about 3.0-5.9 wt.%. A large tail of this process is observed in all the
catalysts; it can be ascribed to the elimination of surface hydroxyl groups (dehydroxylation) heterogeneously
distributed on the catalyst surface [23]. The mass loss of the dehydroxylation is about 1.5-2.5 wt.%. Besides
the dehydration/ dehydroxylation processes no further changes are observed in the catalysts.
3.2. Catalytic activity
3
.2.1. Effect of the metal promoters on the product s electivity trends
In the catalytic activity experiments, all catalysts were tested using the same syngas ratio, H
, the product selectivity trends obtained from the mesoporous catalysts are shown. The products are divided
into oxygenated compounds (alcohols, aldehydes, carbon dioxide and acetates) and hydrocarbon compounds
alkanes and alkenes). If we first analyze the Cu/MCM-41 catalyst, it can be observed that the selectivity to
2
/CO=2. In Figure
7
(
oxygenated compounds is higher than the selectivity to hydrocarbon compounds. Reports, however, using
Cu-based catalysts supported on non-porous silica show the opposite selectivity trends [24, 25] as compared
to our results. Some characteristics of the prepared Cu/MCM-41 catalyst could affect to the product selectivity,
such as: i) structured catalyst with an average pore size of 3.2 nm (Table 1) and ii) some metal-support
interactions as suggested by the TPR analysis (Figure 5). In a previous work it was found that a mesoporous
catalyst (Rh/MCM-41) can concentrate water vapor in the pore structure [6, 7]. That could also be the case for
the Cu/MCM-41 catalyst. Thus, the particular product selectivity obtained for the Cu/MCM-41 catalyst could be
related to both water vapor concentration and some metal-support interactions. The concentration of water
vapor occurs due to the small pore size in the mesoporous catalysts [6, 7]. Accordingly, we believe that the
product distribution is less affected by the water vapor concentration in the promoted catalysts since they
present collapsed pores (Figure 2) and thus minor or negligible water vapor concentration could be expected.
On the other hand, since the experiments were carried out at low syngas conversion we expect minor or
negligible catalyst deactivation effects on the catalytic results. Also, a minor effect on the product distribution
due to the degree of syngas conversion could be expected since the experiments were done at similar syngas
conversion (2-11%).
a) Effect of K as promoter
Incorporation of K into the Cu/MCM-41 catalyst changes the product selectivity as can be seen in Figure 7
(
Cu-K/MCM-41), where the oxygenated compounds are an even higher fraction and represent 92% of all the
products. It is known that K is a good promoter for the formation of oxygenated compounds and this is related
+
to two main functions: i) stabilizing the active sites, presumably Cu species [16, 17]), which can be related to
the lower reduction temperature of the Cu-K/MCM-41 catalyst compared to the Cu/MCM-41 catalyst as can be
seen in Figure 5 and ii) blocking acid sites thus decreasing the hydrocarbons formation [26]. In addition, it was
observed by TPR analysis (see section 3.1.5.) that the metal-support interactions in the Cu/MCM-41 catalyst
are weakened or eliminated by the presence of K, which suggests the metal-support interactions have a minor
effect on the formation of oxygenated compounds. Thus, the higher selectivity to oxygenated compounds

found over the Cu-K/MCM-41 catalyst may be mostly due to the promoting effect of K and less caused by
metal-support interactions.
b) Effect of Fe as promoter
The incorporation of Fe into the Cu/MCM-41 catalyst changes the product selectivity as can be seen in Figure
7
(Cu-Fe/MCM-41), where the selectivity to hydrocarbon compounds increases from 16% to 54%. While the K
promoter favors formation of oxygenated compounds, Fe promoter favors hydrocarbon formation. It is known
that Fe is a good catalyst for the Fischer-Tropsch synthesis, where the formation of hydrocarbons is preferred.
It is believed that the hydrocarbon formation occurs via CO dissociation, which can be either unassisted or H-
assisted, where radicals C* (or *CH) are formed on the catalyst’s surface [1]. The radicals then react to form
C-C bonds, which after hydrogenation steps produce hydrocarbon compounds. From the XPS analysis (see
Table 1) a smaller proportion of Cu was observed on the surface of the Cu-Fe/MCM-41 catalyst than on the
Cu/MCM-41 catalyst, which might reduce the formation of oxygenated compounds and favor hydrocarbon
formation. The presence of Fe increases the syngas conversion. In order to keep the syngas conversion to
less than 11%, it was necessary to increase the gas hourly space velocity from 1500 mlsyngas/gcath for
Cu/MCM-41 to 31000 mlsyngas/gcath for Cu-Fe/MCM-41. From TPR analysis, some active sites in the Cu-
Fe/MCM-41 catalyst have higher activation energy than the Cu-K/MCM-41 catalyst (Figure 5).
c) Effects of Fe and K addition
The product selectivity obtained when both K and Fe promoters are incorporated into the Cu/MCM-41 catalyst
is shown in Figure 7. It can be observed that the selectivity to hydrocarbon compounds (52%) is similar to the
selectivity to oxygenated compounds (48%) after 20 hours on stream. It was noted previously that K as
promoter favors oxygenated compounds formation, possibly via CO non-dissociation [3], while Fe as promoter
favors hydrocarbon formation, possibly via CO dissociation [1]. Hence, when K and Fe promoters are present
both CO dissociation and CO non-dissociation will take place. In addition, CO insertion could also take place
over the Fe promoter and oxygenated compounds can be formed via the so-called enol mechanism [1].
Moreover, there is no consensus about which mechanism is the dominant [3-5], and further studies are
needed to clarify the reaction mechanism for Cu-promoted catalysts. The TPR analysis of the Cu-Fe-K/MCM-
41 catalyst suggests some metal-support interactions and higher activation energy than for Cu/MCM-41 and
Cu-Fe/MCM-41 catalysts (see Figure 5). In the next section, the product distribution between the oxygenated
products will be analyzed and how the combination of both K and Fe promoters can be a good strategy for
boosting the selectivity to alcohols.
3.2.2. Selectivities to alcohols and other oxygenated compounds
The product distribution between the oxygenated compounds is shown in Figure 8. The oxygenated
compounds are methanol, higher alcohols (C2+-Alcohols), acetates (methyl and ethyl acetate) and others
(
aldehydes and acetic acid). It can be noted that methanol is the dominant product (>98%) for the Cu/MCM-41
and Cu-K/MCM-41 catalysts. When Fe is incorporated into Cu/MCM-41, the methanol selectivity decreases to

45% while the selectivity to higher alcohols is increased to about 33%. Furthermore, when both K and Fe
promoters are incorporated the selectivity to higher alcohols represents around 49% of all the oxygenated
compounds. This indicates that the combination of these promoters with different roles, K promoting the CO
non-dissociation and Fe promoting the CO dissociation, is favorable for the formation of higher alcohols.
More interesting is the increasing selectivity toward ethanol. The ethanol selectivity related to all the studied
alcohols (C1-5OH) changes drastically by the action of the metal promoters. Low ethanol selectivity is found in
the Cu/MCM-41 and Cu-K/MCM-41 catalysts, being 1% and 2%, respectively, while the ethanol selectivity in
the Cu-Fe/MCM-41 catalyst increases to 30%. The combined Fe and K promoters in the Cu-Fe-K/MCM-41
catalyst even boost the ethanol selectivity to 41%. In addition to the ethanol selectivity being enhanced,
syngas conversion is also increased, which in turn increases the ethanol yield as can be observed in Table 2.
-
5
The ethanol yield for the Cu-Fe-K/MCM-41 catalyst achieves the value of 165.5x10 C mol/ gcat h, which is
notably higher than that of the non-promoted Cu/MCM-41 catalyst (0.3x10-5 C mol/ gcat h). Accordingly, it
seems that the surface composition of the Cu-Fe-K/MCM-41 catalyst is suitable for ethanol formation, the
atomic composition in this catalyst (from XPS analysis) is Cu0.34Fe0.08K0.08Si1.00. Furthermore, the ethanol
formation could possibly even be increased by varying the final surface composition of Cu, Fe and K. The
latter is a current subject of study in our group and future results will be published.
3.2.3. Anderson–Schulz–Flory (ASF) distribution
The ASF plots for alcohols and hydrocarbon compounds are shown in Figure 9. The shapes of the curves can
give qualitative information on the alcohol and hydrocarbon compounds formation; the more linear curve
indicates a kind of polymerization mechanism via C
of the carbon number (n). It can be observed that the hydrocarbon curve (HC’s, dotted lines) for the Cu/MCM-
1 catalyst is quite similar to the corresponding curve of the Cu-K/MCM-41 catalyst. Also, the Cu-Fe/MCM-41
1
monomers with chain growth probability () independent
4
catalyst shows a HC’s curve similar to that of the Cu-Fe-K/MCM-41 catalyst. This suggests that two kinds of
reaction mechanisms could proceed; one when Cu and K are present and another when Cu-Fe and Cu-Fe-K
are present. The first mechanism favors the oxygenated compounds formation, especially methanol, as was
also observed from the selectivity trends (Figures 7 and 8) which may be related to the CO non-dissociation
and/or CO insertion mechanisms. The second mechanism favors the formation of hydrocarbon compounds
(
Figures 7 and 8) and may be related to the CO dissociation mechanism. However, further detailed studies
are needed to clarify the reaction mechanism for alcohols formation over the Cu-Fe-K/MCM-41 catalyst. We
believe that the various reaction mechanisms could take place in parallel resulting in a beneficial combination
of CO non-dissociation, CO dissociation and CO insertion for the formation of alcohols, and particularly for
ethanol formation.

4. Conclusions
The catalytic performance of a mesoporous Cu/MCM-41 catalyst and the effect of K and Fe promoters were
evaluated for the conversion of syngas at 300 °C, 20 bar and gas-hourly-space-velocities in the range of
1500-30000 mlsyngas/gcat h, under these conditions the syngas conversion level was between 2 and 11%. The
non-promoted Cu/MCM-41 catalyst produces mostly oxygenated compounds, among which methanol is the
main product. When K is incorporated into the Cu/MCM-41 catalyst, the selectivity to methanol is further
increased. On the other hand, when Fe is incorporated into the Cu/MCM-41 catalyst, the selectivity to
hydrocarbon compounds is favored. The action of both K and Fe as promoters may be related to their effect
on the reaction mechanisms. K as promoter seems to enhance the CO non-dissociation and CO insertion
mechanisms to form oxygenated compounds as main products, and Fe as promoter probably enhances the
CO dissociation mechanism to form hydrocarbon compounds. When both K and Fe are simultaneously
present in the Cu/MCM-41 catalyst, it seems that both kinds of mechanisms (CO dissociation and non-
dissociation) can take place and favor the formation of higher alcohols (C2+-OH). The reaction rate is also
notably increased when both promoters are incorporated into the Cu/MCM-41 catalyst, the space time yield of
ethanol is increased from 0.3x10-5 carbon-mol/gcath to 165.5x10-5 carbon-mol/gcath. From XPS, the atomic
surface composition in the Cu-Fe-K/MCM-41 catalyst is Cu0.34Fe0.08K0.08Si1.00.
Acknowledgements
The Swedish International Development Cooperation Agency (SIDA) is gratefully acknowledged for financial
support.
References
[
1] R. Suarez Paris, L. Lopez, J. Barrientos, F. Pardo, M. Boutonnet, S. Jaras, Chapter 3 Catalytic conversion of
biomass-derived synthesis gas to fuels, in: Catalysis: Volume 27, The Royal Society of Chemistry, 2015, pp.
2-143.
2] E. Voegele, OECD-FAO outlook report predicts growth in ethanol production, in: Ethanol Producer
6
[
Magazine, July 18, 2014, Accessed date: 2016-01-15 (http://ethanolproducer.com/articles/11264/oecd-fao-
outlook-report-predicts-growth-in-ethanol-production).
[
3] M. Gupta, M.L. Smith, J.J. Spivey, Heterogeneous Catalytic Conversion of Dry Syngas to Ethanol and
Higher Alcohols on Cu-Based Catalysts, ACS Catal., 1 (2011) 641-656.
4] V. Subramani, S.K. Gangwal, A review of recent literature to search for an efficient catalytic process for
the conversion of syngas to ethanol, Energy & Fuels, 22 (2008) 814-839.
5] J.J. Spivey, A. Egbebi, Heterogeneous catalytic synthesis of ethanol from biomass-derived syngas,
Chemical Society Reviews, 36 (2007) 1514-1528.
6] L. Lopez, J. Velasco, S. Cabrera, M. Boutonnet, S. Järås, Effect of syngas conversion and catalyst reduction
[
[
[
temperature in the synthesis of ethanol: concentration of water vapor in mesoporous Rh/MCM-41 catalyst,
Catalysis Communications, 69 (2015) 183-187.
[
7] L. Lopez, J. Velasco, V. Montes, A. Marinas, S. Cabrera, M. Boutonnet, S. Järås, Synthesis of Ethanol from
Syngas over Rh/MCM-41 Catalyst: Effect of Water on Product Selectivity, Catalysts, 5 (2015) 1737.
8] K.J. Smith, R.B. Anderson, The higher alcohol synthesis over promoted Cu/ZnO catalysts, The Canadian
Journal of Chemical Engineering, 61 (1983) 40-45.
9] I. Boz, M. Sahibzada, I.S. Metcalfe, Kinetics of the Higher Alcohol Synthesis over a K-promoted
CuO/ZnO/Al2O3 Catalyst, Industrial & Engineering Chemistry Research, 33 (1994) 2021-2028.
[
[

[
10] C.H. Bartholomew, R.J. Farrauto, Fundamentals of Industrial Catalytic Processes, New Jersey, Second
ed., Wiley-Interscience, 2006.
11] R. Andersson, M. Boutonnet, S. Jaeras, Correlation patterns and effect of syngas conversion level for
product selectivity to alcohols and hydrocarbons over molybdenum sulfide based catalysts, Appl. Catal., A,
17-418 (2012) 119-128.
12] R. Andersson, M. Boutonnet, S. Järås, On-line gas chromatographic analysis of higher alcohol synthesis
products from syngas, Journal of Chromatography A, 1247 (2012) 134-145.
13] Chapter 15 Quantitative Analysis By Gas Chromatography Measurement of Peak Area and Derivation of
[
4
[
[
Sample Composition, in: G. Georges, L.G. Claude (Eds.) Journal of Chromatography Library, Elsevier,
Amsterdam, Oxford, New York, Tokyo 1988, pp. 629-659.
[
(
[
14] S. Biz, M.L. Occelli, Synthesis and Characterization of Mesostructured Materials, Catalysis Reviews, 40
1998) 329-407.
15] M.V. Landau, S.P. Varkey, M. Herskowitz, O. Regev, S. Pevzner, T. Sen, Z. Luz, Wetting stability of Si-
MCM-41 mesoporous material in neutral, acidic and basic aqueous solutions, Microporous and Mesoporous
Materials, 33 (1999) 149-163.
[
16] G.R. Sheffer, T.S. King, Potassium's promotional effect of unsupported copper catalysts for methanol
synthesis, Journal of Catalysis, 115 (1989) 376-387.
17] L.E.Y. Nonneman, V. Ponec, On the problem of the activity in methanol synthesis of supported,
unpromoted copper catalysts, Catalysis Letters, 7 (1990) 213-217.
18] N. Pauly, S. Tougaard, F. Yubero, Determination of the Cu 2p primary excitation spectra for Cu, Cu2O
and CuO, Surface Science, 620 (2014) 17-22.
[
[
[
[
19] G. Ertl, H. Knözinger, J. Weitkamp, Handbook of Heterogeneous Catalysis, VCH, Weinheim, 1997.
20] S.D. Robertson, B.D. McNicol, J.H. De Baas, S.C. Kloet, J.W. Jenkins, Determination of reducibility and
identification of alloying in copper-nickel-on-silica catalysts by temperature-programmed reduction, Journal
of Catalysis, 37 (1975) 424-431.
[
21] A. Martínez, G. Prieto, The Application of Zeolites and Periodic Mesoporous Silicas in the Catalytic
Conversion of Synthesis Gas, Topics in Catalysis, 52 (2009) 75-90.
22] S.V. Sirotin, I.F. Moskovskaya, B.V. Romanovsky, Synthetic strategy for Fe-MCM-41 catalyst: a key factor
for homogeneous or heterogeneous phenol oxidation, Catalysis Science & Technology, 1 (2011) 971-980.
23] B. Pawelec, J.L.G. Fierro, Chapter 4 Applications of thermal analysis in the preparation of catalysts and in
[
[
catalysis, in: E.B. Michael, K.G. Patrick (Eds.) Handbook of Thermal Analysis and Calorimetry, Elsevier Science
B.V., 2003, pp. 119-190.
[
2
[
24] L.E.Y. Nonneman, V. Ponec, Promotion and support effects in syngas reactions, Catalysis Letters, 7 197-
03.
25] A. Gotti, R. Prins, Basic Metal Oxides as Cocatalysts for Cu/SiO2Catalysts in the Conversion of Synthesis
Gas to Methanol, Journal of Catalysis, 178 (1998) 511-519.
26] A.-M. Hilmen, M. Xu, M.J.L. Gines, E. Iglesia, Synthesis of higher alcohols on copper catalysts supported
on alkali-promoted basic oxides, Applied Catalysis A: General, 169 (1998) 355-372.
[

Cu-Fe-K/MCM-41
Cu-Fe/MCM-41
Cu-K/MCM-41
Cu/MCM-41
MCM-41
5
15
25
35
45
55
65
75
2
 (degrees)
Figure 1. XRD patterns of the prepared mesoporous catalysts.
5
4
3
00
00
00
1.8
1.5
1.2
0.9
0.6
0.3
0.0
200
1
00
0
0.0
0.2
0.4
0.6
P/Po
0.8
1.0
0
5
10 15 20 25 30 35 40 45 50 55 60 65 70 75
Pore diameter (nm)
MCM-41
Cu/MCM-41
Cu-K/MCM-41
Cu-Fe/MCM-41
Cu-Fe-K/MCM-41
Figure 2. BJH pore size distribution and (inset) N
2
-isotherm of the prepared catalysts.

Cu/MCM-41
9
32.5
Cu-K/MCM-41
Cu-Fe/MCM-41
Cu-Fe-K/MCM-41
9
52.2
9
15 920 925 930 935 940 945 950 955 960 965 970
Binding Energy (eV)
Figure 3. Cu2p-spectra of the prepared catalysts.
ꢀꢁꢂ ꢂ-ꢃꢄ
ꢀ-ꢅꢁꢂ ꢂ-ꢃꢄ
ꢀ
-ꢆꢇꢁꢂ ꢂ-ꢃꢄ
ꢀ-ꢆꢇ-ꢅꢁꢂ ꢂ-ꢃꢄ
Figure 4. TEM images of the prepared catalysts.

Cu-Fe/MCM-41
Cu/MCM-41
Cu-K/MCM-41
Cu-Fe-K/
MCM-41
1
30 170 210 250 290 330 370 410 450
Temperature (°C)
Figure 5. TPR profiles of the prepared catalysts.
0
1
2
3
4
5
MCM-41
Cu-K-MCM-41
Cu-Fe-K-MCM-41
Cu-MCM-41
Cu-Fe-MCM-41
,
-
,
,
,
,
,
,
,
-
,
-
,
-
,
-
,
-
,6
,
-
7
8
9
,
-
,
-
20
140 260 380
Temp. (C)
-
10
,
2
0
80
140
200
260
320
380
,
,
Temperature (C)
Figure 6. TGA and DTG profiles of the prepared catalysts.

Cu/MCM-41
Cu-K/MCM-41
Cu-Fe/MCM-41 Cu-Fe-K/MCM-41
1
00
90
80
70
60
50
40
30
20
10
0
Hydrocarbons
Oxygenates
Conversion
0
0
55 11 00 11 55 22 00 2 50 3 05 3 15 0 4 01 5 45 050 555 1 06 0 1 56 5 0 705 7510 80158520
Time on stream (h)
Figure 7. Product selectivity trends obtained from the prepared catalysts. Reaction conditions: syngas H /CO
2
ratio: 2/1, 300 °C, 20 bar and a GHSV of 1500 mlsyngas/gcat h for the Cu/MCM-41 and Cu-K/MCM-41 catalysts
and 30000 mlsyngas/gcat h for the Cu-Fe/MCM-41 and Cu-Fe-K/MCM-41 catalysts.
Cu/MCM-41
Cu-K/MCM-41
Cu-Fe/MCM-41 Cu-Fe-K/MCM-41
1
00
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
Methanol
C -Alcohols
2
+
Acetates
Others
0
55 11 00 15 20 2 50 3 50 31 50 41 05 45 0 50 5 551060156 50 7 05 71 50 8 10 5 8 25 0
Time on stream (h)
Figure 8. Oxygenates and alcohol selectivities obtained from the prepared catalysts. Reaction conditions as
specified in Figure 7.

5
Cu/MCM-41 Cu-K/MCM-41 Cu-Fe/MCM-41 Cu-Fe-K/MCM-41
OH's OH's
,
,
4
3
2
1
0
1
2
3
HC's
HC's
OH's
HC's
OH's
-
-
-
HC's
0
11 22 33 44
5
61 72 83 49 10 1 11 1 22 133 144 15 1 16 1 27 138 14 9 20 21
Carbon number (n)
Figure 9. ASF plots for alcohols (OH’s) and hydrocarbons (HC’s) for the prepared catalysts. Reaction
conditions as specified in Figure 7.

Table 1. Physico-chemical properties of the prepared catalysts.
Average
metal
particle size
by TEM
Metal loading (%)
by AAS
BET
surface
Average
pore
size
Atomic ratios by XPS
Pore
Catalyst
volume
area
3
(
cm /g)
Cu
K
Fe Si/Cu Cu/K Cu/Fe Fe/K
2
(
m /g)
(nm)
(
nm)
MCM-41
710
375
0.49
0.24
2.9
3.2
-
-
-
-
-
-
-
-
-
-
-
-
-
Cu/MCM-41
4.8
28.89
2.8
Cu-K/MCM-41
208
200
72
0.26
0.32
0.27
5.1**
6.6**
5.0*
5.0*
6.1*
29.79 1.53
26.82
-
2.9
7.0
2.9
13.0
-
-
-
-
Cu-Fe/MCM-41
Cu-Fe-K/MCM-41
-
2.28
4.3
5.3
14.8**
29.30 1.65 2.07
11.7
2.2
*
Average of all the metals. ** Average of the interparticle space.
Table 2. Space time yield (STY) of alcohols and other oxygenated compounds over the mesoporous
catalysts. Reaction conditions as indicated in Figure 7.
-
5
Conv.
%)
STY (10 C mol/ gcat h)
OH -C OH Acetates* Others**
Catalyst
(
C
1
OH
C
2
C
3
4
Cu/MCM-41
2.2
2.8
27.3
42.9
0.3
0.8
0.0
0.1
0.0
0.0
0.2
0.0
Cu-K/MCM-41
Cu-Fe/MCM-41
Cu-Fe-K/MCM-41
10.0
4.9
307.2
145.4
159.2
165.5
61.4
83.3
46.4
53.0
101.8
59.4
*
Methyl acetate and ethyl acetate. ** Propanal, acetaldehyde and acetic acid.