Bioenergy systems for the future 7 catalysts for conversion of synthesis gas

Catalysts for conversion of
synthesis gas
V. Palma, C. Ruocco, M. Martino, E. Meloni, A. Ricca
University of Salerno, Salerno, Italy

Nomenclature
BTL
CTL
CNT
CSTR
DME
DMT
FT
FTS
GHSV
GTL
GNS
HTFTS
HTS
LTFTS
LTS
MMA
MMT
MTBE
MTO
MTP
NCNT
NP
NW-FS
O/P
OCNT

PEMFC
PROX
PSA
RWGS

biomass to liquid
coal to liquid
carbon nanotubes
continuous stirred tank reactor
dimethyl ether
dimethyl terephthalate
Fischer-Tropsch
Fischer-Tropsch synthesis
gas hourly space velocity
gas to liquid
graphene nanosheets
high-temperature Fischer-Tropsch synthesis
high-temperature shift (or high-temperature water-gas shift)
low-temperature Fischer-Tropsch synthesis
low-temperature shift (or low-temperature water-gas shift)
methyl methacrylate
million metric tons
methyl tert-butyl ether
methanol to olefin
methanol to paraffin
nitrogen-functionalized carbon nanotube
narrow pore
nanowires
olefins/paraffins
oxygen-functionalized carbon nanotube

proton exchange membrane fuel cell
preferential oxidation
pressure swing adsorption
reverse water-gas shift

Bioenergy Systems for the Future. />© 2017 Elsevier Ltd. All rights reserved.

7


218

Bioenergy Systems for the Future

SSA
STP
WGS
WP

7.1

specific surface area
standard temperature and pressure
water-gas shift
wide pore

Introduction

The synthesis gas (or simply syngas) is the reaction product of several transformation
processes (Fig. 7.1) such as reforming processes (Ghoneim et al., 2015), partial oxidation (Christian Enger et al., 2008) and gasification (Mahinpey and Gomez, 2016);

the chemical composition of syngas depends on the raw materials and the production
process used (Couto et al., 2013), however the main components are carbon oxides
(CO and CO2) and hydrogen (H2). Syngas is recognized as the most suitable raw material for manufacturing a wide range of chemicals and fuels and is therefore involved in
several catalytic processes (Fig. 7.1).
The wide use of the syngas, in recent decades, has been favored by the low cost of
fossil sources; however, the increasingly severe restrictions on CO2 emissions, the
growing worry of public opinion on climate changes, and the spread of alarming data
on the reserves of crude oil have pushed the interest toward the use of alternative
sources. The best alternative to the fossil fuels are the biomass, renewable materials
that contain considerable quantities of carbon, hydrogen, and oxygen, restorable by
photosynthetic reaction (Maschio et al., 1994). Certainly, there are many disadvantages in using biomass with respect the fossil fuel, the presence of contaminants, a
variable hydrogen to carbon ratio due to the composition of the sources, a low energy
density and high costs make them uncompetitive; however, nowadays, there are no
other real alternatives. Theoretically, all biological materials (both animal and vegetable) represent a biomass; however, only cheap materials and wastes are conveniently
converted into syngas, and wastes from wood processing, energy crops, agricultural
residues, by-products from processing of biological materials, municipal and sludge
wastes, and food industry wastes are normally used as raw materials for the syngas

Methanol synthesis

Raw materials
(natural gas,
coal,
biomass)

Ammonia synthesis,
other processes

Syngas

Fisher Tropsch
process (synthetic
fuels)

Fig. 7.1 Syngas—production and transformation.


Catalysts for conversion of synthesis gas

219

production. There exist two processes for converting the biomass in biofuel and biopower, the fermentation that produces mainly ethanol (biogas) and the thermochemical conversion to syngas. The variability of the sources requires a flexibility in
processing steps; however, the most of biomass contain a significant amount of water,
so a preliminary drying process is commonly performed before going to the conversion processes; alternatively, a hydrothermal processing directly degrades the biomass
(Tekin et al., 2014); similarly, the product gas stream, from thermochemical process
(pyrolysis, combustion, and gasification), contains many unwanted by-products to be
removed before going to next process (Kumar et al., 2009):
l

l

l

l

l

Particulate, removed according to the size by cyclone separators, wet scrubbers, electrostatic precipitators, and barrier filters
Alkali compounds, removed by barrier filters (Turn et al., 2001).
Nitrogen compounds, removed at high temperature with dolomites, Ni-based catalysts and

Fe-based catalysts (Lepp€alahti and Koljonen, 1995).
Sulfur compounds, removed by limestone, dolomite or calcium oxide
Tar compounds, removed directly in the gasifier with the use of specific catalysts (Han and
Kim, 2008) or alternatively in a separate reactor

The fate of the syngas depends on the process in which it is involved and the desired
final product (Fig. 7.1); this chapter wants to provide a general overview on the primary catalytic systems involved in the most widespread conversion processes of the
syngas, focusing on the results of the latest research. The main process is the methanol
synthesis that produces one of the most flexible chemical commodities and energy
sources (Ajay et al., 2014). Methanol is used as feedstock in synthesis of formaldehyde and acetic acid; additives in adhesive, foams, plywood subfloors, and windshield
washer; methyl tert-butyl ether (MTBE), a gasoline component; and dimethyl ether
(DME), a clean-burning fuel. Methanol is also used as additive into gasoline or as
vehicle fuel itself.
The Fischer-Tropsch process (FT) converts the syngas into a mixture of products
refined to synthetic fuels, lubricants, and petrochemicals (de Klerk, 2000); depending
from the sources, the overall process, from the raw material to the final product, is
named GTL (gas to liquid), CTL (coal to liquid), or BTL (biomass to liquid) (van
de Loosdrecht and Niemantsverdriet, 2013).
The Haber-Bosch process (Haber, 2002) allows to obtain ammonia by reacting
nitrogen with pure hydrogen, usually obtained from syngas by removal of carbon
monoxide with the water-gas shift reaction (Palma et al., 2016) eventually coupled
with methanation (R€
onsch et al., 2016), preferential oxidation or by a hydrogen
permselective membrane reactor (Piemonte et al., 2010), and CO2 sequestration.
Ammonia is an important commodity for the fertilizers industry, is the precursor of
urea and ammonium salts (nitrates and phosphates), of the nitric acid and polyamides.
The carbonylation processes allow to introduce the carbonyl group (CO) into
organic or inorganic substrate, by reacting with pure carbon monoxide, obtained from
syngas by reverse water-gas shift reaction (De Falco et al., 2013). Interesting examples of carbonylation are the Monsanto process (Paulik and Roth, 1968), that allows to
prepare the acetic acid by carbonylation of methanol and, the Mond process for the

extraction and purification of Nickel (Mond et al., 1890).


220

7.2

Bioenergy Systems for the Future

Fischer-Tropsch synthesis

The Fischer-Tropsch synthesis (FTS) is an important catalytic process used for the
conversion of syngas (derived from coal, natural gas, biomass, or other carboncontaining species) to hydrocarbons with different chain lengths. The product selectivity is strongly dependent on temperature, pressure, and catalyst choice
(Ghareghashi et al., 2013). Generally, low-temperature FTS is carried out over
Co-based catalyst in the temperature interval of 190–250°C and at 20–40 bar yielding
products with high average molecular weight (middle distillates and waxes). Conversely, iron-catalyzed process are driven at 340°C and 20 bar, with the aim of
obtaining short-chain hydrocarbons (fuels and petrochemicals) (Delparish and
Avci, 2016; Shin et al., 2013). In order to face the reaction exothermicity and avoid
hot spots or rapid catalyst deactivation, multitubular fixed bed, with external cooling
(for LTFTS), and slurry bubble column reactors (for HTFTS) are commercially
selected for the FTS process (Park et al., 2011). In the first case, considerable fraction
of the liquid reaction products has to be recycled to the reactor to remove the reaction
enthalpy, thus increasing pressure drops and making the reactor trickier to be operated
and less flexible to be scaled. On the other hand, in slurry reactors, the temperature is
uniform and pressure drops are low, being related to the hydrostatic pressure of liquid,
and the internal mass transfer limitations are ruled out by loading the catalyst as fine
powder. However, high aspect ratios (reactor height/diameter) and staging have to be
used to limit back-mixing phenomena. Also, particular attention has to be paid, especially during the size-scaling processes, both to the prevention of the catalyst attrition
and to the design of an efficient tool for the separation of the catalyst from the liquid
(Visconti et al., 2011). In addition, during FT synthesis, even though the reactants are

in the gas phase, the pores of the catalyst are filled with liquid products, and the diffusion rates in the liquid phase are typically three orders of magnitude slower than in
the gas phase; the increasing transport limitations may result in CO depletion and
lower C5 + selectivity. In a fixed-bed reactor, the selectivity problem can be solved
by using catalyst pellets where the catalytic material is deposited in a thin outer layer
(eggshell catalysts), while in a slurry reactor, the selectivity issue is faced by using
small catalyst particles (Liu et al., 2009). However, in the attempt of overcoming
the drawbacks of commercially available technologies for FT synthesis, different
technologies including the adoption of structured fixed-bed reactors, based on honeycomb monolith foams, knitted wires, or cross flow structures, and microchannel reactors have been recently proposed (Pangarkar et al., 2009; Twigg and Richardson,
2002; Cao et al., 2009). For example, monolithic catalysts assure low pressure drop,
high gas-liquid mass transfer rates in two-phase flow, the possibility of using high liquid and gas throughputs, and a good temperature control (Kapteijn et al., 2005; Hilmen
et al., 2001). For intensification of mass transfer between synthesis gas, liquid products and solid catalysts, alternative catalyst geometries like honeycombs, structured
packings, and foams have also been developed (Guettel et al., 2008a).
All group VII metals have noticeable activity for the hydrogenation of carbon monoxide to hydrocarbons. However, only ruthenium, iron, cobalt, and nickel have


Catalysts for conversion of synthesis gas

221

catalytic characteristics that allow considering them for commercial production. However, nickel catalysts, under practical conditions, produce too much methane; moreover, ruthenium is too expensive, and its worldwide reserves are insufficient
(Khodakov et al., 2007). Cobalt and iron were proposed as the first catalyst by Fischer
and Tropsch. Fe-based catalysts, despite being less expensive than Co ones ( Jin and
Datye, 2000; van Berge et al., 2000), strongly suffer for deactivation by coke and promote H2O formation by water-gas shift reaction (WGS). Sulfur content lower than 0.2
and 0.1 ppm are mandatory for normal operation of Fe- and Co-based catalysts,
respectively (Khodakov et al., 2007). Co catalysts are generally more resistant to attrition and are widely preferred for use in slurry-type reactors. Despite the industrialscale development of FTS process, the activity and stability of the catalyst need to
be improved. At that hand, the addition of suitable promoters and the selection of
proper supports provide a reasonable route for the enhancement of FTS catalyst performances. In the following sections, the performances of Co- and Fe-based catalysts
were reviewed and discussed.

7.2.1

Co-based catalysts

Al2O3, despite having lower surface area than SiO2 and TiO2, is commonly used as
support material for Co due to the strong metal-oxide interactions, its good mechanical
performances, and resistance to attrition (de la Osa et al., 2011; Prieto et al., 2009). On
the other hand, when Al2O3 is selected as support, it is mandatory to avoid the formation of hardly reducible cobalt aluminates that are responsible for activity reduction
( Jongsomjit et al., 2001). Iglesia et al. (Iglesia, 1997) found that for large cobalt particles supported on Al2O3, SiO2, and TiO2, the FT reaction rate depends on the number
of available cobalt surface atoms and that hydrocarbon selectivity is only slightly
affected by cobalt dispersion. A different phenomenon is observed for small cobalt
particles, having a strong impact on product selectivity: Particle size in the range
˚ improves olefinic products yields, while for smaller dimensions, C5 + selec60–80 A
tivity also decreased (Khodakov, 2009). The synthesis of Co-based catalysts
supported on alumina nanofibers was shown to assure a homogenous metal
particle-size distribution. After catalyst ultrasonication, more active cobalt particles
were generated, which marked improved C5 + selectivity lowering methane production. In addition, even at high reaction temperature and under much higher CO conversion (79%), a quite stable activity was observed at 230°C, 20 bar, and H2/CO ¼ 2
over 300 h of reaction (Liu et al., 2016). Flame spray pyrolysis technique was also
successfully employed for controlling catalytically active Co particles deposition
on Al2O3: a good catalytic activity was recorded at the above operative conditions
(Minnermann et al., 2013). Conversely, for Co/γ-Al2O3 catalysts, it was shown that
partial pores prefilling by incipient wetness impregnation of Al(NO3)3 resulted in catalyst similar to the eggshell systems ( Jacobs et al., 2016), which are able to decrease
C1–C4 light gas selectivity, improving, at the same time, C5 + selectivity.
Due to the need of limiting pressure drop and the consequent necessity of adopting
“big” catalyst pellets, low-temperature Fischer-Tropsch synthesis in industrial fixedbed reactors may suffer of strong intraparticle mass transport limitations, which are


222

Bioenergy Systems for the Future

known to result in decreased CO conversion rate and C5 + selectivity. Upon
decoupling the pellet diameter and the diffusive length, eggshell catalysts represent
an engineering solution for the intensification of the Fischer-Tropsch reactors. In this
regard, Fratalocchi et al. (Fratalocchi et al., 2015) showed that 600 μm pellets, with
catalytically active layers, 75 μm thick, grant a remarkable combination of high CO
conversion rate and high C5 + selectivity at 220–240°C, 25 bar, and H2/CO ratio of
1.73, thus resulting extremely interesting for operations in reactors 3–6 m long.
Concerning the impact of cobalt aluminate formation on catalyst activity, Moodley
et al. (Moodley et al., 2011) found an enhancement of aluminate content with water
partial pressure during FTS at 230°C, 10 bar, and H2/CO ¼ 1.5. The presence of water
is also regarded as one of the main causes of catalyst sintering during Fischer-Tropsch
process (Bezemer et al., 2010). Sintering mechanism for Co/Al2O3 catalysts in slurry
reactors was accelerated by the formation of intermediate surface cobalt-oxide species, and the deactivation was favored by increasing H2/CO ratios in syngas or by
the presence of even small amounts of water (Sadeqzadeh et al., 2013). Beside
sintering and aluminates formation, the activity drop commonly observed over FTS
catalyst is related to the deposition of carbonaceous species on catalyst surface.
The research of Pena et al. (Pen˜a et al., 2014) was focused on the identification of
the molecular structure of carbon species formed over a Co/Al2O3 catalyst in a slurry
reactor. Carbon adsorbed on spent catalyst was mainly constituted by α-olefins,
n-paraffins, branched alkanes/alkanes, aldehydes, and ketones, while carboxylic acids
were mostly detected at high water partial pressures. In particular, the increase in CO
conversion enhanced the isomerization of α-olefins favoring the formation of
branched alkanes/alkanes. Carbon species are probably nucleated on the cobalt particles and then migrate to alumina support and coke localized on the support showed
high reluctance toward hydrogenation.
The addition of promoters to Co/Al2O3 catalysts was shown on one hand, to prevent
Co-aluminate formation due to establishment of an intimate contact with the metal
(Nabaho et al., 2016a) and, on the other hand, to limit catalyst deactivation.
Park et al. (Park et al., 2012) investigated the impact of phosphorous addition to
Co/Al2O3 catalyst on deactivation induced by lumps formation in the presence of
water. γ-Al2O3, due to its hydrophilic properties, can undergo phase transformation

to pseudoboehmite (with low attrition resistance) in the presence of water vapor produced during FT reaction, forming fragmented fine catalyst powder. The deposition of
heavy hydrocarbons on these fragments causes the formation of aggregate catalyst
lumps. Conversely, phosphorous addition is able to suppress alumina hydrophilic
properties and reduce heavy hydrocarbon deposition on catalyst, thus preventing its
deactivation (Fig. 7.2). It was also reported (Tan et al., 2011) that the deposition of
small boron quantities on Co/γ-Al2O3 catalyst can hinder the deposition, nucleation,
and growth of resilient coke on catalyst surface, without affecting initial activity and
selectivity at 240°C and 20 bar.
The addition of noble metal promoters (Pd, Pt, Re, and Ru) was shown to improve
activity and stability of Co/Al2O3 catalysts (Ma et al., 2012). Improved CO conversion
was observed at 220°C, 22 bar, and H2/CO ¼ 2 over the promoted catalysts, with Pt
and Pd enhancing oxygenate formation and Re and Ru slightly decreasing it. At fixed


Catalysts for conversion of synthesis gas

223

Phosphorous—unmodified catalyst
Large phase transform
γ-Al2O3 in boehmite
Partial hydration
responsible of fine
powder

Emulsion of water and
hydrocarbons

Catalyst particles

Aggregated catalyst lump
(co-presence of water and hydrocarbons)

Liquid medium

Phosphorous—modified catalyst
Small phase transform
γ-Al2O3 in boehmite

Emulsion of water and
hydrocarbons
Catalyst particles
Liquid medium

No formation of aggregated catalysts

Fig. 7.2 Proposed mechanism for Co/Al2O3 catalyst deactivation in a slurry-phase reactor.

CO conversion (50%), Re and Ru improved CH4 and C5 + selectivity, whereas the
opposite effect was observed for Pt and Pd promoters. The latter metals also increased
2-C4 olefins selectivity and WGS activity of the final catalyst. On the other hand, Pt
addition had a negligible effect on C4 olefin isomerization. Concerning Co catalysts
supported on Al2O3 or SiO2 prepared via plasma technology (Chu et al., 2015), the
promotion by noble metals was shown to improve both cobalt dispersion and reducibility, thus enhancing the FT reaction activity.
Other porous supports (including SiO2, TiO2, activated carbon, and zeolite) are
usually selected for commercial Co-based catalysts (Lu et al., 2015; Shi et al.,
2012; Eschemann et al., 2015). However, the combination of two types of oxides
was found to improve the pore structure, cobalt dispersion, and reducibility. The effect
of alumina incorporation (0–3 wt%) into a Co/SiO2 catalysts on product gas distribution was investigated by Savost’yanov et al. (2017). Trace of alcohols and olefins were
only detected over the undoped catalyst and their content increase with alumina loading (Table 7.1). Over the 1 wt% sample, the molecular weight distribution became

narrower, increasing the C8–C25 fraction. A further reduction of SiO2/Al2O3 ratio
caused the opposite effect. Combustion synthesis method was employed for the preparation of Co/SiO2 and Co/SiO2-Al2O3 catalysts, and a strong impact of preparation
method on product gas distribution was observed (Ail and Dasappa, 2016): the innovative catalyst increased the yield to C6 + products at 230°C, 30 bar, and H2/CO ¼ 2.3,
resulting in the formation of long-chain hydrocarbons waxes (C24+) with respect to
the middle distillates (C10–C20), normally generated over impregnated catalysts.
However, the overall C6 + yield was further increased by Al2O3 addition, due to
the marked improvement (48%) in cobalt dispersion promoted by alumina.
Venezia et al. (Venezia et al., 2012) modified SiO2 support by TiO2 grafting and
observed an improvement in C5 + selectivity especially at high space velocities


224

Bioenergy Systems for the Future

C5+ product distribution in the FTS over Co/SiO2-Al2O3
catalysts at 210°C, 20 bar, and H2/CO ratio of 2
(Savost’yanov et al., 2017)
Table 7.1

Product distribution (wt%)
Paraffins

Olefins

Alcohols

Al2O3 content (wt%)

C5–C18

C19–C35

C35+

C5–C18

C5–C18

0
1
3

49.6
54.9
48.1

47.6
42.1
37.5

2.4
2.3
2.2

0.4
0.7
10.1

0.04

0.07
1.2

(GHSV ¼ 7200 hÀ1 at 210°C, 20 bar, and H2/CO ¼ 2). In addition, CoO oxide interaction with the doped support was enhanced, avoiding the particle mobility that can
lead to catalyst deactivation by sintering. A bimodal ZrO2-SiO2 support was selected
for jet fuel direct synthesis via FTS reaction with different 1-olefins as additives
(Li et al., 2016a). Olefins cofeeding effectively shifted the product distribution toward
jet fuel range, markedly suppressing CH4, CO2, and light hydrocarbons (C2–C4) formation. The large pores of the bimodal support, in fact, provided efficient pathways
for reactants conversion and products diffusion, while the newly formed small pores
assured high metal dispersion.
Rare-earth oxides, able to remarkably enhance catalyst reducibility, were shown to
be beneficial for improving long-chain hydrocarbons selectivity (Spadaro et al.,
2005). CeO2 addition (5 and 10 wt%) to Co/ZrO2 catalysts promoted the formation
of larger cobalt particles, inhibiting water reoxidation and Co particles aggregation
during the reaction and resulting in a better resistance toward deactivation (Zhang
et al., 2016a). Fig. 7.3 displays CO conversion with time onstream for doped and
undoped catalysts: CeO2-modified samples assured a good stability during 100 h of
reaction. However, the deactivation rate was affected by the amount of ceria added.
However, a rapid deactivation was observed in the initial stage of the reaction over the
10 wt% catalysts, caused by the initial smaller pores that were filled by the liquid
waxes produced during the FTS synthesis.
Carbon materials, including activated carbon, carbon nanotubes and nanofibers,
carbon spheres, and mesoporous carbon, are also reported as catalytic support for
Co, due to their several benefits with respect to the conventional oxides supports: these
materials display high purity, high mechanical strength and thermal stability, and
large surface area (Fu and Li, 2015). Moreover, having a hardly reducible surface,
Co particle reducibility can be improved. Carbon porous structure can also be properly
controlled in order to promote the cobalt dispersion (Ha et al., 2013). Dı´az et al. (2013)
carried out FT reaction over Co catalysts supported on carbon nanofibers prepared at
three different calcination temperatures (750°C, 600°C, and 450°C, denoted as Samples 1, 2, and 3, respectively). At 250°C, 20 bar, and H2/CO ratio of 2, the Samples 1

and 2, having a medium pore radius, displayed high catalytic activity without


Catalysts for conversion of synthesis gas

225

Fig. 7.3 CO conversion with
time onstream over Co/ZrO2
(dotted line), Co/(5)CeO2-ZrO2
(dash-dotted line), and Co/(10)
CeO2-ZrO2 (dashed line);
220°C, 20 bar, and H2/CO ¼ 2.

CO conversion (%)

100
80
60
40
20
0

0

10

20

30 40 50 60 70

Time-on-stream (h)

80

90 100

deactivation; however, the promotion of WGS and methanation reaction led to significant CO2 and CH4 production. The Sample 3, which was less active and suffered from
deactivation, showed an improvement in C5 + selectivity. Metal sintering was
observed over all the catalysts and, especially, over the Sample 3, due to its lower
structural order with respect to the other two catalysts. The performances of
Co-based catalysts supported on carbon nanotubes (CNT) and graphene nanosheets
(GNS) for FT reaction at 220°C, 18 atm, and H2/CO ¼ 2 were compared in order to
evaluate the effect of morphology and structure on catalyst stability (Karimi
et al., 2015).
The difference, in terms of SSA (Table 7.2), between the bare GNS and CNT can be
attributed to the nature and textural properties of graphene nanosheets. Higher porosity was also observed over the GNS samples, which can be related to the interlayer
spacing of them. After 480 h of reaction, a specific area reduction of 20% and 3%,
respectively, for the CNT and GNS catalysts, was observed. Moreover, the extent
of pore blockage for the CNT catalyst is higher than that of the GNS sample, as a consequence of its higher rates of sintering and clusters growth. During stability tests, dCo

Specific surface areas (SSA), porous volume (Vp),
and cobalt average crystallite sizes (dCO) for Co/CNT
and Co/GNS catalysts (Karimi et al., 2015)

Table 7.2

Sample

SSA (m2/g)

Vp (cm3/g)

dCo (nm)

CNT
Fresh Co/CNT
Used Co/CNT
GNS
Fresh Co/GNS
Used Co/GNS

497
372
298
848
602
586

1.034
0.765
0.428
2.2
1.46
1.23

—
8.6
10.08
—
7.8

8.8


226

Bioenergy Systems for the Future

increased for both the catalysts; however, a more significant crystal growth was
recorded over the Co/CNT catalyst. The latter sample displayed a drop of activity
of 15.8% in the first 120 h of reaction, while in the remaining period, it only drops
of 4.9%. Conversely, due to the lower extent of sintering phenomena and GNS hydrophilic properties that limit water deposition and cobalt reoxidation, only a conversion
reduction of about 2.7% in the first period and 1.2% between 120 and 480 h was
observed over the Co/CNT catalyst. The relationship between C5 + selectivity and
metal particle sizes for Co catalysts supported on carbon nanotubes and spheres
was also studied (Xiong et al., 2011). In order to limit water effect, measurements
were carried out at a low CO conversion (4%). The change in C5 + selectivity with
particle sizes (Fig. 7.4) can be explained considering that large Co particles promote
the formation of bridge-type adsorbed CO, more active and more easily dissociated,
which enhanced the chain growth and C5 + selectivity. Co supported over spheres,
however, displayed slightly higher C5 + selectivity for a similar sized particle. In fact,
over nonporous carbon spheres, all cobalt particles are dispersed on the outer surface
of the support. Therefore, the hydrogen concentration around the active cobalt particles is the same as that in the reactor, thus leading to less light hydrocarbon production.
Chernyak et al. (2016) investigated the effect of oxidation time (1–15 h) selected
during Co supported over carbon nanotubes preparation on their performances for FT
synthesis at 190°C, atmospheric pressure, and H2/CO ¼ 2. The sample oxidized for 9 h
displayed the highest CO conversion and yield to C5+ products. Selectivity to heavy
hydrocarbons is almost the same over the catalyst treated for 3, 9, and 15 h while
increased over the sample oxidized for 1 h.
Syngas mixture, especially when produced from coal or biomass, can be CO2rich, and carbon dioxide can significantly affect FT activity. Dı´az et al. (2014) studied the effect of CO2 cofeeding on the catalytic performances of a Co catalyst
supported over carbon nanofibers at 220–250°C, 20 bar, and H2/CO ¼ 2. Increasing

temperatures assure higher catalytic activity and the rate of undesired reactions
(WGS and methanation). However, once reaction temperature was fixed at 250°C,

100
99
C5+ selectivity (%)

Fig. 7.4 Influence on Co particle
size on C5 + selectivity over Co/
carbon nanotubes (square) and
spheres (triangle); 225°C, 8 bar,
and CO/H2 ratio of 0.5.

98
97
96
95
94
93
92

0

5

10

15 20 25 30 35 40
Cobalt particle size (nm)

45 50


Catalysts for conversion of synthesis gas

227

the presence of CO2 in the feed gas was demonstrated to affect the rate of catalytic
hydrogenation of CO and product distribution. H2/CO2, in fact, behaves as a mild
oxidizing agent on Co/CNFs under selected conditions. In the absence of CO, secondary catalytic activity decayed and methanation process raised a maximum. Therefore,
the decrease of CO conversion and C5 + selectivity with CO2 addition was attributed
to the lower reactivity of this component. CO2 also competed with Co for the adsorption on catalytic sites and C7–C20 hydrocarbon product distribution was shifted
toward lower-molecular-weight hydrocarbons by feeding higher amounts of CO2,
mainly caused by the easily desorption of the chains. Besides CO2, biomass-derived
syngas may contain different organic and inorganic impurities, including NH3. For Co
catalysts supported on Al2O3, TiO2, and SiO2, the addition of 10 ppmv of NH3 during
FT synthesis at 220°C, 19 bar, and H2/CO ¼ 2 caused a significant deactivation for
all supported cobalt catalysts, but the rate of deactivation was higher for the silicasupported catalysts relative to the alumina- and titania-supported catalysts
(Pendyala et al., 2016). Ammonia addition had a positive effect on product selectivity
(i.e., lower light gas products and higher C5 +) for alumina- and titania-supported
catalysts compared with ammonia-free conditions, whereas, the addition of ammonia
increased lighter hydrocarbon (C1–C4) products and decreased higher hydrocarbon
(C5 +) selectivity compared with ammonia-free synthesis conditions for the silicasupported catalyst. However, after H2 treatment, both titania- and alumina-based
catalysts were completely regenerated, while over Co/SiO2 samples, the loss of activity was irreversible, due to the formation of inactive cobalt-support compounds.
Co catalysts supported over manganese oxides were also employed in FT due to
their high yields toward light hydrocarbons and low CO2 and methane selectivity
(Zhou et al., 2015). However, Iqbal et al. (2016) found that the use of activated carbon
support for CoMnOx catalysts improved both the activity and selectivity to C2 +
hydrocarbons, while further lowering methane (from 22.1% to 7.0%) and carbon dioxide selectivity (from 37.0% to 20.4%) with respect to unsupported catalysts at 240°C,
6 bar, and CO/H2 ratio of 1.

In addition, carbon materials have been considered for the synthesis of hybrid
composites, such as carbon nanotubes-Al2O3 and carbon nanofibers-SiO2. In particular, carbon nanotubes enhanced the degree of reduction of Co, thus limiting C5 +
production (Zaman et al., 2009). Chernyak et al. (2015) also proved that the increase
in oxygen-containing groups on catalyst surface by H2O2 oxidation of Co/carbon
nanotubes-Al2O3 assured a redistribution of pore sizes, with smaller Co particles
and a reduction on C5 + yield.
Multifunctional catalysts, containing different types of active sites, have also
been proposed for FT synthesis. For example, zeolite, due to the acidity ascribable
to Al in the structure, promoted secondary reactions that include the formation of
lighter hydrocarbons, both aromatic and branched. In addition, depending on the
type of zeolite, it is possible to restrict the chain growth, thus yielding to lighter
hydrocarbons (Plana-Pallejà et al., 2016). Xing et al. (2015) developed a proper
preparation method to create hierarchical pores in a mesoporous zeolite for facilely
tuning the product distribution during Fischer-Tropsch synthesis. A series of catalysts was prepared by acid and basic leaching for different leaching times, showing


228

Bioenergy Systems for the Future

that, for a leaching time of 4 h, isoparaffin selectivity, at 260°C and 9.9 atm, reached
up to 52.4% and middle hydrocarbons become the main products due to the
optimized hydrocraking and isomerization function afforded by the hierarchical
zeolite structure. It was also demonstrated (Xing et al., 2016) that the employment
of hierarchically spherical (0.5–1 μm) Co-based zeolite catalysts, having aggregate
nanorods structure, improved isoparaffin and C5–11 hydrocarbons selectivity.
On the other hand, CH4 and C12 + selectivities were lower than those observed
for a commercial HZSM-5 supported Co catalyst (at 240°C, 10 bar, and H2/CO
ratio of 2). Lee et al. (2010) evaluated the effect of the temperature and time of
hydrogen treatment on the performances of Co-supported zeolite catalysts. The

sample treated at 500°C for 18 h (catalyst A) displayed finally dispersed metal
clusters inside zeolite pores. Increasing the temperature and the treating time, more
and more metal clusters inside zeolite cages migrated out of the pores and agglomerate into large, immobile aggregates at the external surface of zeolite crystals.
As a consequence, the sample A displayed low hydrogen chemisorption, showing a
more difficulty in olefins hydrogenation to the corresponding paraffins at 270°C
and H2/CO ratio of 2. As a result, the slow hydrogenation ability enhanced chain
propagation, thus improving C5 and higher hydrocarbon formation. A series of
Co catalysts supported on physically mixed ZSM-5/SBA-15 were tested for FTS
at 240°C and 20 bar (Wu et al., 2015). The composite-supported catalysts displayed
improved catalytic performances over the respective single material-supported
Co catalysts. In particular, the sample containing 20 wt% of ZSM-5 reached
the maximum CO conversion (90.6%, Table 7.3), the highest C5–C22 hydrocarbon
selectivities (70.0%), and the minimum formation of light hydrocarbons (13.3%
for CH4 and 7.0% for C2–C4 alkenes). Martı´nez et al. (2007) investigated the
performances of a hybrid catalyst prepared by mixing Co/SiO2 and zeolite. At
250°C, 20 bar, and H2/CO ¼ 2, zeolite promoted the cracking of C13 + long-chain
n-paraffin formed on the Co particles, mainly yielding to gasoline-range branched products. However, the accumulation of carbonaceous species caused a reduction in the latter products yield with time onstream. It was observed that the

Table 7.3 CO conversion (XCO) and selectivity results for the FTS
over Co/ZSM-5/SBA-15 catalysts; 240°C, 20 bar, and H2/CO 5 2
(Wu et al., 2015)
Hydrocarbon selectivity (%)
ZSM-5
(wt%)

XCO (%)

CH4

C2–C4

C5 +

C5–11

C12–22

C5–22

C23 +

0
10
20
30
50
100

90.0
90.2
90.6
87.9
84.9
70.5

23.4
21.4
13.3
13.8
18.7

18.5

11.5
10.0
7.0
8.8
9.4
12.0

65.1
68.8
79.7
77.3
71.9
69.6

33.9
33.1
32.0
33.6
34.7
23.9

26.6
29.3
38.0
35.2
30.7
34.1

60.5
62.4
70.0
68.8
65.4
58.0

4.7
6.4
9.7
8.5
6.5
11.6


Catalysts for conversion of synthesis gas

229

deactivation rate was little affected by zeolite acidity and increased with the zeolite
pore dimension. Coke molecules mainly comprised two- and three-ring aromatics in
large pore zeolites, while it was predominantly of paraffinic nature in the most stable HZSM-5. Aromatic coke is likely formed from light olefins produced in the FT
synthesis through consecutive oligomerization, cyclization, and dehydrogenation
reactions.
As observed above, the porous characteristics of the support significantly affect
activity and hydrocarbon selectivity in FT synthesis (Xiong et al., 2008). In this
regard, it was found (Wei et al., 2016) that three-dimensional mesoporous silica foams
not only provided cavities to suppress Co particles growth but also enhanced metals
dispersion and reducibility. In particular, foams having large- and open-pore structure
favored the occurrence of secondary reactions, leading to higher C5 + selectivity. At

210°C, 10 bar, and H2/CO ¼ 2, a stable behavior was observed. However, for higher
temperatures (250°C), cobalt sintering and the formation of Co-silica compounds caused rapid deactivation, and improvements in endurance performances were reached by
carbon coating or Al doping. It was also shown (Labuschagne et al., 2016) that Co
supported over porous SiC catalysts, calcined at 550°C, assured higher stability during
long-term tests than the Co/Al2O3 counterparts. Moreover, an aggressive acid washing
of the catalyst led to superior activities, even at high water partial pressures.
Noble metals addition is a well-known route to enhance activity of Co-based
Fischer-Tropsch catalysts, due to the improvement of cobalt-oxide reduction and
the increase in the active sites number (Das et al., 2003). den Otter et al. (2016) evaluated the effect of Pt addition to the activity of Co/γ-Al2O3 and Co/Nb2O5 catalysts
at 220°C. At 1 bar, the cobalt-weight-normalized activity of Co/γ-Al2O3 and Co/Nb2O5
was found to increase by a factor of 1.7 and 2.8 upon Pt promotion (Fig. 7.5, left). For
Co/γ-Al2O3, low activity and only slight influence of Pt promotion were observed at
1 bar. At 20 bar, no large influence of Pt promotion on the cobalt-weight-normalized
activity of Co/γ-Al2O3 was observed, whereas for Co/Nb2O5, a factor of 2.4 increase in
the activity per unit weight of cobalt was measured, without affecting C5+ selectivity
(Fig. 7.5, right). Similar catalytic activity for FT reaction was observed over PtCo and

20
1 bar

Activity (10–5molCO•gCO–1•s–1)

Activity (10–5molCO•gCO–1•s–1)

10
PtCo/γAl2O3

8
6

PtCo/Nb2O5

4
Co/γAl2O3

2
Co/Nb2O5

0
0

10

20

30
40
Time (h)

50

60

20 bar
PtCo/Nb2O5

15
Co/γAl2O3

10

PtCo/γAl2O3

5

Co/Nb2O5

0
0

20

40

60

80 100 120 140 160
Time (h)

Fig. 7.5 Cobalt-weight-normalized activity in Fischer-Tropsch synthesis at 1 bar, CO
conversion 1–5% (left), and 20 bar, CO conversion 21–34% (right), 220°C and H2/CO ¼ 2.


230

Bioenergy Systems for the Future

AuCo catalysts supported on Al2O3 (Nabaho et al., 2016b). However, a key factor to
consider for the commercial viability of these catalysts is their activity and selectivity
over extended time onstream. In particular, Au-promoted samples reduce less readily

than Pt-promoted samples with each oxidation-reduction regeneration cycle, which
implies that Au-Co/Al2O3 may be a less attractive option in commercial applications.
Pendyala et al. (2014) evaluated the effect of PtCo/Al2O3 catalyst particle (sieve) on
performances for Fischer-Tropsch synthesis. Four different size ranges were selected
(20–63, 63–106, 106–180, and 180–355 μm). The increase of catalyst sieve size was
accompanied by losses in CO conversion, except for the smallest sieve range. Such
losses can be related to the filling of the interior of the catalyst particles with heavy
waxes, thereby blocking catalytically active sites. Due to the presence of waxes, the
small particles tended to flocculate to larger clusters or were small enough to move at
the speed of the liquid. The effect of sulfur poisoning on PtCo/Al2O3 FT catalysts was
also studied (Barrientos et al., 2016). S clearly affected catalyst activity and stability at
210°C, 20 bar, and H2/CO ¼ 2.1, decreasing long-chain hydrocarbons selectivity. In
addition, sulfur significantly enhanced secondary hydrogenation of olefins. However,
such effects were clearly related to S content, and the poisoning is negligible at
moderately low sulfur coverages (0–250 ppm).
Au addition to Co/Al2O3 and Co/SiO2 catalysts was shown to improve Co particles
reduction and catalyst activity ( Jalama et al., 2011). This positive effect was found
to be similar to that observed with other noble metals such as Ru and Re. However,
as a drawback, Au loadings below or equal to 1 wt% on alumina and in the range
0.5–5 wt% on silica and titania increased methane selectivity. Pirola et al. (2014)
observed that the addition of low Ru or Pt loadings to Co/SiO2 catalysts improved both
CO conversion and total yield toward the desired products (C2 + species that exclude
CH4 and CO2) at 220°C, 20 bar, and H2/CO ¼ 2. This improvement is related on one
hand to the formation of a Ru1–yCoy solid solution and on the other hand to the generation of a PtCo intrametallic compound. Ru and Re were also added to Co/SiO2 catalysts prepared by conventional drying and high-temperature supercritical drying. The
innovative preparation method led to the production of a less-reducible cobalt silicate,
and the noble metal addition improved cobalt species reducibility and dispersion, partially suppressing the Co particles coverage by SiO2 layers. As a result, higher olefin
production was measured, while the C11 + selectivity was lower than that observed
over the conventional catalysts at 230°C, 20 bar, and H2/CO ¼ 2 (Iida et al., 2013).
Osakoo et al. (2013) prepared silica-supported catalysts by impregnation and
coprecipitation using a reverse micelle technique. Over latter sample, tested at

230°C, 5 bar, and H2/CO ¼ 2, lower Co3O4 particle sizes were measured, which
reduced methane and C2–C4 selectivities. However, the addition of Pd in the range
0–1 wt% increased methane formation and negatively affected CO conversion: the
best results were observed for the 0.2 wt% catalyst prepared by coprecipitation, which
displayed 34.8% of CO conversion and high mole fraction (0.38) of paraffins in the
gasoline range (C5–C9). The effect of the addition of different noble metals (Ag, Au,
and Rh) to Co/SiO2 catalysts was investigated in a work of Yan et al. (2011). Au and
Rh showed a promoting effect on the FT activity, whereas the addition of Ag had a
detrimental effect. Moreover, the addition of small amounts of Rh (0.1–0.5 wt%)


Catalysts for conversion of synthesis gas

231

improved CO conversion by 50% without affecting catalyst selectivity. For Co/TiO2
catalysts promoted by Ru, it was observed that the interaction in the bimetallic particles can reduce site blockage by carbonaceous species (Eschemann et al., 2016).
Similarly, for Re-promoted catalysts, the improved activity and stability are due to
the interaction of noble metal with cobalt. Noble metals were also shown to be more
active hydrogenation catalysts than Co, thus accelerating CO hydrogenation. The performances of a Co/CeO2 catalyst were compared with a Pt-promoted sample during
FTS at 220°C, 1 bar, and H2/CO ¼ 2 (Lorito et al., 2016). Similar activities were
observed in the two cases, as Pt was poorly active for CO hydrogenation under these
conditions. The improvement of Co dispersion promoted by Pt was probably responsible for the increase in methane and the decrease in propene formation.
In order to lower the catalyst price, the substitution of noble metals with cheaper
components was also proposed. For example, bimetallic Co-Ni catalysts showed
high activity for FT reaction, displaying high selectivity to gasoline-range hydrocarbons (C5–C12) (Calderone et al., 2011). Shimura et al. (2015) investigated the influence of Co/Ni ratio and impregnation sequence on the activity of Al2O3 supported
catalysts. Low Ni loading slightly increased CO conversion rate at nearly constant
C5 + selectivity, but excess Ni loading largely decreased CO conversion rate and
C5 + selectivity at 230°C, 10 bar, and H2/CO ¼ 1.91. The catalytic activity did not
depend on impregnation sequence, when Ni loading amount was low. However, when

Co was substituted with large amount of Ni, catalysts prepared by sequential impregnation method (Ni first and then Co) showed higher activity than those prepared by
coimpregnation method and reverse sequential impregnation method (Co first and
then Ni). These results indicate that catalysts with a Co-rich surface would be better
than those with a Ni-rich surface. The best catalyst was 19%Co-1%Ni/Al2O3 that
exhibited 1.6 times as high activity as 20%Co/Al2O3 catalyst prepared by a conventional impregnation method. Co-Cu catalysts supported on Al2O3 were also investigated for Fischer-Tropsch reaction at 220°C, 16 bar, and H2/CO ratio of 2 ( Jacobs
et al., 2009). As shown in Table 7.4, the increase in Cu loading decreased CO conversion, due to a poisoning of surface Co sites. At similar conversion levels, the growth in
Cu content slightly increased methane production reducing, at the same time, C5 +
yield. However, a further improvement in Cu loading led to a prohibitive increase
in methane selectivity (21.6% vs 9.2%) and precipitous drop in C5 + selectivity
(47.4% vs 80.6%).

Comparison of CO conversion (XCO) and product
selectivity (Si) over mono- and bimetallic Co/Al2O3 catalysts;
220°C, 16 bar, and H2/CO ratio of 2 ( Jacobs et al., 2009)

Table 7.4

Catalyst

XCO (%)

SCH4 (%)

SC5+ (%)

SCO2 (%)

15%Co/Al2O3
0.49%Cu-15%Co/Al2O3
15%Co/Al2O3

1.63%Cu-15%Co/Al2O3

47.8
50.6
28.7
29.9

8.9
9.9
9.2
21.6

80.6
76.6
81.6
47.7

0.82
0.83
0.67
1.51


232

Bioenergy Systems for the Future

Honeycomb monoliths have been developed as support for the cobalt phase by several researchers. CoRe/γ-Al2O3 monolithic and powder catalysts were tested for FT
reaction, demonstrating that, at similar methane selectivity, the structured catalysts
assure higher reaction rates. In fact, the reaction rate enhancement is most probably

caused by the advantageous mass transfer characteristics of the monolithic catalyst
(Guettel et al., 2008b). Moreover, for tests carried out between 210°C and 232°C,
at 25 bar, and H2/CO ratio of 2, higher C5–C18 liquid fractions and olefin/paraffin
ratios are obtained by conducting the FT reaction over the ceramic monolith catalyst
coated with the bimetallic catalyst. In addition, if significant wax formation was
observed with the packed particle bed, wax was not detected in the liquid products
for FT synthesis over monoliths (Liu et al., 2009). Ceramic foam, which could be
an alternative to ceramic monoliths, has also been extensively developed for the
FT reaction. The advantage of foam versus a straight-channel monolith is the high
degree of radial mixing, which improves reactant distribution and convective heat
transfer (Liu et al., 2014). Lacroix et al. (2011) compared the FT activity of two foam
catalysts (Co/SiC and Co/Al2O3) at 220°C, 40 bar, and H2/CO ¼ 2. At medium conversion (<50%), the two catalysts display similar C5 + selectivity, indicating that the
intrinsic selectivity between the two catalysts is close from each other. However, when
the CO conversion was increased to 70%, a significant difference in terms of the C5 +
selectivity was observed between the two catalysts, that is, 80% on the Co/SiC and
54% on the Co/Al2O3, which indicate that under severe FTS reaction conditions
the SiC seems to be more suitable support than alumina. Additional catalytic test conducted on a hybrid support, that is, Al2O3-coated SiC foam, again confirmed the high
C5 + selectivity under a similar severe reaction conditions in the presence of a SiC
structure underneath of the alumina layer that plays a role of heat disperser.

7.2.2

Fe-based catalysts

Iron catalysts are efficiently employed for olefin synthesis via Fischer-Tropsch reaction for several reasons, which includes high selectivity, low methanation activity,
availability, low price, and lower sensitivity to poisons (Torres Galvis and de Jong,
2013). Iron catalysts can be also efficient for utilization of syngas produced from coal
and biomass that have lower hydrogen content (H2/CO < 2). In fact, Fe catalysts promote WGS reaction, providing additional H2. Moreover, according to the promoters
and the operative conditions used, Fe also assures high flexibility, modulating the
selectivity to alkenes or branched hydrocarbons (Cano et al., 2017). α-Al2O3, SiO2,

and carbon materials are commonly employed as support substances for Fe-based catalysts. Such weakly interactive supports were selected due to their capability in promoting iron species activation for the formation carbides, which improved FT
performances. Generally, fresh Fe catalysts mainly contain α-Fe2O3, which can be
converted to different reduced iron species (such as Fe3O4, FeO, or metallic iron)
under H2 atmosphere. Furthermore, the reduced Fe3O4 is transformed continually
to metastable FeO phase or reduced directly to metallic iron. It is generally accepted
that the strong interaction of metal support in the supported iron catalysts may stabilize the metastable FeO. Under CO or syngas atmosphere, these reduced iron species


Catalysts for conversion of synthesis gas

233

Fe

80

Fe/Al2O3

80

CO
H2

Conversion (%)

40

20

0

40

H2

CO

40

160
80
120
Time on stream (h)

200 40

80
120
160
Time on stream (h)

Conversion (%)

60

60

20

0

200

Fig. 7.6 Activity and stability of Fe and Fe/Al2O3 catalysts; 260°C, 15 bar, and H2/CO ¼ 0.67.

could be converted to different types of iron carbides, which are deemed as active
phases for FTS (Ding et al., 2015).
In order to highlight the impact of metal-support interaction on activity and stability for FT synthesis, the performances of an Fe bulk catalyst were compared with
an Fe/Al2O3 sample, prepared by a combination of coprecipitation and spray-driedmethod (Xu et al., 2016). As shown in Fig. 7.6, Fe catalyst had higher FTS activity
and deactivated slowly with time onstream, whereas the CO conversion of Fe/Al2O3
catalyst is stable or even increases slowly. Apparently, incorporation of Al2O3 into
iron catalyst decreases the catalyst activity but improves the catalyst stability. The
lower activity of Al2O3 incorporated catalyst is probably due to the strong
Fe-Al2O3 interaction, which inhibited the reduction and the carburization of iron
species.
Xu et al. (2016) proposed an Fe-α-Al2O3 catalyst modification in view of limiting
their poisoning problem, which is an issue problem in FT synthesis. The addition of
sulfur to the catalyst significantly decreased activity, due to the inhibition of CO
dissociation and the limited formation of iron carbide phases. The presence of sulfur
also suppressed the formation of C5 + hydrocarbons and shifted the products to
C2–C4 hydrocarbons. At the same time, the olefin to paraffin (O/P) ratio of the
C2–C4 hydrocarbons decreased with increasing S/Fe molar ratio (operative conditions, 200°C, 20 bar, and H2/CO ¼ 1). However, the resistance to sulfur poisoning
of the Fe/α-Al2O3 catalyst was found to improve with increasing reaction temperature
from 310°C to 350°C.
Several authors have shown that FTS rates of Fe-based catalysts increase with
K and Cu addition and that the best performances can be achieved when both


234

Bioenergy Systems for the Future

K and Cu are present. Such species, in fact, promote the formation of Fe carbide with
higher dispersion (Li et al., 2002). By investigating the effect of calcination temperature, in the interval 250–650°C, on the performances of Fe/Cu/K/Al2O3 catalysts for
FTS, it was found that high calcination temperature can result in more Al2O3 into iron
oxide lattice, which further enhanced the metal-support interaction and led to the separation of iron oxide and CuO. Moreover, it weakened the promotional effect of CuO
and K2O, thereby severely suppressing the reduction and carburization of the catalyst
(Wan et al., 2007).
A beneficial effect of Na addition to Fe-based catalysts was also observed: the
alkali metal enhanced iron carbidization, suppressing the secondary olefin hydrogenation and increasing the chain growth probability (Torres Galvis et al., 2013). Cheng
et al. evaluated the effect of sodium addition on the activity of iron catalysts supported
on alumina, silica, CMK-3, and carbon nanotubes for Fischer-Tropsch reaction.
Sodium promotion, as displayed in Table 7.5, leads to higher olefin-to-paraffin ratios
for both light and long-chain hydrocarbons, lower methane selectivity, and higher
chain growth probability. The selectivity effects were more pronounced for the catalysts supported by carbon nanotubes. The impact of sodium promotion was less significant on silica and CMK-3 supported catalysts.
Very little influence of sodium promotion on the catalytic performance was
observed in the alumina-supported iron catalysts because of formation of sodium aluminates, having a relative small effect for Fischer-Tropsch synthesis. In more inert

Table 7.5 Activity of Na-promoted Fe-based catalysts for FTS
at 300°C, 20 bar, and H2/CO 5 2 (Cheng et al., 2015)
Product distribution (%, CO2-free)

Sample

XCO
(%)

Olefin/paraffins
ratio C2–C4

CH4

C2–C4
olefins

C2–C4
paraffins

C5+

Fe/CNT
0.1Na/Fe/CNT
0.3Na/Fe/CNT
0.5Na/Fe/CNT
Fe/SiO2
0.1Na/Fe/SiO2
0.3Na/Fe/SiO2
0.5Na/Fe/SiO2
Fe/Al2O3
0.1Na/Fe/Al2O3
0.3Na/Fe/Al2O3
0.5Na/Fe/Al2O3
Fe/CMK-3
0.1Na/Fe/CMK-3
0.3Na/Fe/CMK-3
0.5Na/CMK-3

76.4
75.3
59.6
50.6

26.1
14.2
19.2
18.5
51.4
55.3
65.7
70.1
42.3
24.5
10.9
10.1

1.3
5.5
5.6
5.7
1.9
4.4
5.9
6.1
1.6
1.4
1.9
2.6
0.9
5.4
5.8
6.9

9.6
3.4
3.4
3.8
10.2
7.1
5.4
4.9
8.2
7.4
5.7
5.1
10.1
5.6
4.2
4.8

17.9
18.1
16.7
17.0
19.5
18.3
19.2
19.4
16.8
16.5
15.9
18.2
16.1

23.7
14.4
24.1

14.2
3.3
3.0
3.0
10.5
4.2
3.3
3.2
10.8
11.8
8.5
7.1
18.9
4.4
2.5
3.5

58.3
75.2
76.9
76.2
59.8
70.4
72.1
72.5
64.2

64.3
69.9
69.6
54.9
66.3
78.9
67.6


Catalysts for conversion of synthesis gas

235

supports such as silica and CMK-3, the optimum sodium contact could be in the range
0.3–0.5 Na/Fe. Fe/CNT demonstrates very strong effect of sodium on the selectivity
already at the small amounts of promoter (0.1 Na/Fe) with further decrease in the
activity at the higher sodium content (Cheng et al., 2015). The performances of a catalyst consisting of SiO2 nanowires and highly dispersed Fe2O3 (denoted NW-FS) were
compared with the FTS activity of an industrial spherical Fe2O3/SiO2 catalyst (denoted indus-FS) at 320°C and H2/CO ¼ 1. NW-FS sample exhibited a high selectivity
for light olefins, especially for ethylene in the Fischer-Tropsch synthesis. This was
because of the highly dispersed Fe2O3 and low diffusion resistance of its open structure. The C2–C4 olefin/paraffin ratio was 3.3, which was higher than that of indus-FS,
equal to 1.9 (Chen et al., 2014a). Sudsakorn et al. (2001) investigated the effect of the
addition of small concentration of precipitated SiO2 on the performances of a spraydried Fe catalyst. SiO2 addition decreased catalyst particle density, thus resulting in
lower attrition resistance. The best results, in terms of high active surface area, good
dispersion in the slurry, and high attrition resistance, were measured for a SiO2 loading
of 10 wt%. The effect of the activation atmospheres on the FTS activity of a γ-Fe2O3
catalyst supported on SBA-15 was investigated in a work of Perez De Berti et al.
(2016). The experimental results, carried out at 330°C, 20.3 bar, and H2/CO ¼ 2,
showed that activation with pure H2 produces a catalyst more active and less selective
to methane. In order to explain the different catalytic properties, the schematic representation shown in Fig. 7.7 was proposed. Activation in H2/CO would occur following
a “shrinking-core” model. Instead, pure H2 would lead to expose suddenly the surface

of Fe3O4 to a carburizing mix (H2/CO). This situation would produce a great number
of iron carbide nuclei. The Fe oxide nanoparticles in the catalysts would have great
number of iron carbide “nodules” with smaller size if pure H2 is used as activation
atmosphere. Therefore, a larger number of sites for CO adsorption and dissociation
and shorter diffusion paths would be obtained, and the catalyst will be more active.

Iron carbides

Activation

Iron carbides
nodules
FTS

CO
:H

2

γ-Fe2O3
γ-Fe2O3

Iron carbides

Fe3O4

Fe3O4

Fe3O4

Fe3O4
FTS

H

2

Precursor
Precursor

Fe3O4

Fig. 7.7 Schematic representation of an Fe catalyst supported on SBA-15 and reduced under
different atmospheres.


236

Bioenergy Systems for the Future

Li et al. (2015) evaluated the effect of alkali metals on the FTS activity of an Fe
catalyst supported over SiO2 at 260°C, 15 bar, and H2/CO ¼ 2. The amount of CH4
gradually declined with increasing the periodic orders of alkali metals, while the
amount of ethane increased. Alkali metals significantly decreased the surface concentration of hydrogen, inhibited the hydrogenation capability, and improved the CO
adsorption and C–C coupling reaction with increasing the periodic orders of alkali
metals. These results were being ascribed to the increasing basic strength of alkali
oxides in the order of Li2O < Na2O < K2O < Rb2O.
Copper is typically selected as a promoter for Fe catalyst to facilitate the reduction
of iron oxides, while small amounts of potassium are commonly added to promote the
formation of iron carbides.

The performances of promoted and unpromoted Fe-containing catalysts prepared
by the sonochemical method and supported on SiO2 were compared with the FT activity of traditional impregnated catalysts at 250°C, 20 bar, and H2/CO ¼ 2 (Comazzi
et al., 2017). Catalytic activity was higher in sonochemically prepared catalysts over
impregnated samples with the same amount of active metal and promoters; in particular, an increase of about 5 times was observed for the 10 wt% Fe sample. All the
sonicated samples had lower selectivity to methane; in particular, methane selectivity
is 6 times lower for the 10 wt% Fe catalyst, while the 30 wt% Fe and the KCu promoted catalysts displayed a decrease in methane selectivity of about 50% with respect
to the impregnated samples. The measured selectivity to CO2 is lower for the sonicated unpromoted catalysts, while the sample containing Cu and K presented the
highest selectivity value toward carbon dioxide. All the catalysts synthesized by
the sonochemical method showed higher selectivity to > C7 species with respect to
the impregnated samples. The improved reactant conversion and selectivity values
observed over the innovative catalysts can be attributed to the production of nanostructured materials with better surface and morphological properties. Cano et al.
(2017) evaluated the performances of Fe catalysts supported on SiO2 and SBA-15,
doped with Cu and K, at 270°C, 10 bar, and H2/CO ¼ 2. The results of the catalytic
tests performed showed firstly the importance of porosity in silica supports for
FTS catalysts, since Fe/SBA-15 showed a higher activity, major chain growth formation of the products and more selectivity to olefins than the Fe/SiO2 catalyst. Moreover, the effect of K and Cu as promoters showed that the addition of K can enhance
the catalytic activity and favor the selectivity to olefins. In fact, it was observed that
the presence of K and Cu in the vicinity of Fe over the support surface creates new
active sites, with the creation of conductor interfaces with different electronic density
distribution. On the other hand, the addition of Cu apparently increases the stability of
the catalysts. As a result, the best performance in the FT reaction was obtained with
FeCu/SBA-15 and FeK/SBA-15. In view of investigating poisoning phenomena, the
effect of H2S on activity and stability of an Fe/SiO2 catalyst promoted by K and Cu
was investigated in a CSTR between 230°C and 270°C, at 13 bar, and for a H2/CO
ratio ranging between 0.67 and 0.77 (Ma et al., 2016). Adding 0.1 ppm H2S for
72 h led to a small deactivation; for example, the deactivation rate based on average
percentage loss in CO rate per day at 270°C was 0.28%; increasing the H2S level from
0.2 to 1.0 ppm linearly increased the deactivation rate from 0.57% to 4.6%. The H2S


Catalysts for conversion of synthesis gas

237

limit for the iron catalyst at which nearly zero deactivation rate can be achieved was
determined to be 50 ppb. The cofeeding of different levels of H2S also altered the
product selectivities. Adding sulfur for about 400 h was found to gradually decrease
CH4 selectivity and increase C5 + selectivity. The added sulfur improved the selectivities of the secondary reactions of olefins and the WGS reaction even though the rates
for these declined. Moreover, the Fe/S ratio decreased dramatically from 13.5 to 6.0
when the temperature was increased from 230°C to 270°C. The results suggest that
sulfur poisoning of the Fe catalyst was exacerbated at lower temperatures. Todic
et al. (2016) investigated the effect of process conditions on the FT activity of an
industrial FeKCu/SiO2 catalyst in a stirred tank slurry reactor, finding a reduction
of methane production and increase of C5 + products by decreasing temperature (from
260°C to 220°C), inlet H2/CO ratio (2:0.67), and/or increasing pressure (8–25 bar).
Moreover, overall selectivity toward methane and C5 + did not show significant
changes with variations in residence time. For an Fe-Mn-K-SiO2 catalyst tested at
262°C and 5 bar (Ding et al., 2013), it was found that activation in higher H2/CO ratio
promoted the reduction of α-Fe2O3 to Fe3O4, whereas decreasing H2/CO ratio facilitated the formation of iron carbides on the surface of magnetite formed and surface
carbonaceous species. During the FT synthesis reaction, the catalyst reduced in lower
H2/CO ratio presented higher catalytic activity, which may be attributed to the formation of iron carbides (especially χ-Fe5C2) on the surface layers, providing the more
active sites for FT synthesis. The surface carbonaceous species formed had a negligible effect in keeping the FT synthesis activity and stability. In addition, pretreatment
in higher H2/CO ratio facilitated the product distributions shifting toward lowermolecular-weight hydrocarbons.
The performances of a coprecipitated Fe-Cu-K catalysts were compared with the
results achieved after the employment of silica and alumina as structural promoters
(Rafati et al., 2015). The doubly promoted Fe-Cu-K-Si-Al catalyst achieved higher
CO and CO2 conversions than the Fe-Cu-K catalyst and singly promoted Fe-CuK-Al and Fe-Cu-K-Si catalysts at 350°C, 10 bar, and H2/CO ¼ 2. The CO and CO2
conversions of the syngas with 54% H2/10% CO/29% CO2/7% N2 over the doubly
promoted catalyst were 88.3% and 25.2%, respectively, compared with 81.8% and
18.5% for the Fe-Cu-K catalyst. In this case, the C5 + selectivity of the doubly promoted catalyst was 71.9%, which was slightly lower than 75.5% for the Fe-Cu-K catalyst. The CO2 was converted to hydrocarbons using the doubly promoted catalyst
when the CO2/(CO + CO2) ratio was higher than 0.35 for H2-balanced syngas at

H2/(2CO + 3CO2) ¼ 1.0 and 0.5 for H2-deficient syngas at H2/(2CO + 3CO2) ¼ 0.5.
The increase of hydrogen content in the syngas increased the methane selectivity
at the expense of decrease in the liquid hydrocarbon selectivity.
Zn addition to Fe-based catalysts was shown to enhance light olefins selectivity
and to decrease carbon selectivity (Gao et al., 2016). Ning et al. (2013) studied the
effect of K and Cu addition to FeZn catalysts for FTS at 230°C, 16 bar and
H2/CO ¼ 2.4. The CO conversion of the unpromoted sample decreases with time
onstream. The addition of Cu the latter catalyst did not show any improvement. However, the simultaneous presence of Cu and K assured a more evident increasing CO
conversion during activity test. In contrast with the FeZn catalyst, Cu increased the


238

Bioenergy Systems for the Future

CH4 selectivity, while K decreased it. The CH4 selectivity of the four catalysts is varied in the order of Cu/FeZn > FeZn > KCu/FeZn % K/Fe. The sequence of CO2 selectivity is K/FeZn % KCu/FeZn > Cu/FeZn % FeZn. In fact, the water-gas shift activity
was increased upon Cu addition onto pure iron catalyst (Martinelli et al., 2014). In
order to improve the reactivity of CO2, potassium was added to an Fe catalyst containing Zn and Cu. K-promoted iron catalysts, at 220°C, 30 bar, and H2/CO ¼ 1,
favored CO2 adsorption, thus enhancing selectivity toward middle distillates. CO2
also had a key role in preventing the CO shift to CO2, thus improving the overall economy of the conversion process and avoiding a net CO2 production. Interestingly, upon
increasing the K loading, the CO conversion rate is decreased, both in the presence and
in the absence of CO2, possibly as a result of the very strong CO adsorption on the
catalytic surface.
Beside silica and alumina, other high surface area oxide, including ZrO2, TiO2,
MnO, and MgO, have been selected as supports for Fe-FT catalysts. By varying
the pretreatment temperature in hydrogen atmosphere of Fe/ZrO2 catalysts, different
activity was observed during FTS at 320°C, 20 bar, and H2/CO ¼ 2 (Al-Dossary and
Fierro, 2015). The pretreatment temperature affected the particle sizes of iron oxide,
and the FT activity of the catalysts was strongly affected by this dimension. In particular, the CO conversion rate during FTS increases as a function of the increase
in particle size, reaching a maximum value for mean Fe particle size of approximately

7 nm, obtained after pretreatment at 900°C. Qing et al. (2011) showed the beneficial
effects of ZrO2 incorporation into Fe/SiO2 catalysts for FT reaction. Fig. 7.8 displays
CO conversion and iron carbide content as a function of TOS at 280°C, 1 bar, and

Fe5C2 content (%)

50
40
30
20
10
90
CO conversion (%)

Fig. 7.8 CO conversion and iron
carbides content for Fe/ZrO2
(continuous line), Fe/SiO2
(dashed line), and Fe/SiO2ZrO2
catalysts (dotted line) during FT
reaction at 280°C, 1 bar, and
H2/CO ¼ 1.

75
60
45
30
15
0
0

20 40 60

80 100 120 140 160 180 200


Catalysts for conversion of synthesis gas

239

H2/CO ¼ 2. It is known that the interconversion between iron oxides and carbides is
reversible upon the FT environment. At high H2O and CO2 partial pressures, iron carbide will be oxidized to Fe3O4, which leads to the deactivation of iron-based FTS
catalysts. Accordingly, the Fe3O4 could be recarburized to iron carbides when the
CO partial pressure was high. For FeZr catalysts, high iron carbide content after activation ensured high initial activity; thus, a large amount of H2O would be produced,
which in turn oxidized the iron carbides. Consequently, the iron carbide content
decreased gradually with TOS, which deactivated the catalyst. Meanwhile, the deactivation caused by the deposition of inactive carbonaceous compounds on the catalysts’ surfaces cannot be ruled out. However, the χ-Fe5C2 content of FeSi catalysts
increased slightly in the initial stage and then became stable. Therefore, the amount
of iron carbide did not change significantly, and no obvious change in the CO conversion level of FeSi catalysts was observed. On the other hand, the reduction and
carburization ability of FeSi catalysts were enhanced by ZrO2 addition: Probably,
more iron carbide would be formed in FeZrSi catalysts, and these species would be
more stable in terms of H2O oxidation than FeSi catalysts during FTS. As a result,
the χ-Fe5C2 content of Fe/SiO2ZrO2 catalysts increases gradually with TOS and
the highest CO conversion was recorded. Zhang et al. (2016b) compared the performances of conventional Fe-Mn catalysts with the FT activity of Fe2O3@MnO2 coreshell catalysts at 280°C, 20 bar, and H2/CO ¼ 1. The latter catalysts enhanced catalytic
performances, especially in C5+ hydrocarbon selectivity. In fact, Mn promoter can
accelerate the dissociation of CO and thus enhanced the concentration of active intermediates for chain growth. Moreover, compared with the pure Fe2O3 (Mn-free) catalyst, the selectivity toward C5 + hydrocarbons over Fe2O3@MnO2 catalyst was
increased from 44.6 to 66.6 wt%. Meanwhile, the undesired CH4 was decreased from
16.8 to 8.9 wt%.
Carbon materials have also been selected as suitable supports for Fe-based FT catalysts. Abbaslou et al. (2010) found that the iron oxide particles on a wide pore (WP)
carbon nanotubes support were larger than those supported on narrow pore (NP)
nanotubes. As a consequence, during FT tests carried out at 275°C, 20 bar, and
H2/CO ¼ 2, the activity of the NP catalyst (%CO conversion of 30) was 2.5 times that

of WP (%CO conversion of 12). In addition, the Fe catalyst supported on WP nanotube
was more selective toward lighter hydrocarbons with a methane selectivity of 41%
compared with that of NP sample with methane selectivity of 14.5%. Deposition of
metal particle on the carbon nanotubes with narrow pore size resulted in more active
and selective catalyst due to higher degree of reduction and higher metal dispersion.
FT reaction was also studied at 340°C, 25 bar, and H2/CO ¼ 1 over iron oxide
nanoparticles supported on untreated oxygen-functionalized carbon nanotubes
(OCNTs) and nitrogen-functionalized CNTs (NCNTs), as well as thermally treated
OCNTs (Chew et al., 2016). An activity loss for iron nanoparticles supported on
untreated OCNTs was observed, originated from severe sintering and carbon encapsulation of the iron carbide nanoparticles under reaction conditions. Conversely, the
sintering of the iron carbide nanoparticles on thermally treated OCNTs and untreated
NCNTs during reaction was far less pronounced. The presence of more stable surface
functional groups in both thermally treated OCNTs and untreated NCNTs is assumed


240

Bioenergy Systems for the Future

Table 7.6 FT performances of the impregnated Fe (A) and FeK
catalyst (B) and the KFe catalysts prepared by redox method (C);
300°C, 20 bar, and H2/CO 5 1 (Duan et al., 2016)
Catalyst

A

B

C

CO conversion (%)
CO2 selectivity (%)
CH selectivity (%)
CH4 selectivity (%)
C5+ selectivity (%)
α
Olefin/paraffin ratio

25.4
46.0
50.3
35.0
14.1
0.43
1.8

26.0
39.7
56.2
27.0
18.2
0.50
2.9

28.8
27.6
70.4
19.5
26.9
0.58

3.7

to be responsible for the less severe sintering of the iron carbide nanoparticles during
reaction. As a result, no activity loss for iron nanoparticles supported on thermally
treated OCNTs and untreated NCNTs was observed, which even became gradually
more active under reaction conditions. The activity of a K-promoted iron/carbon nanotubes composite, prepared by a redox method, was compared with the performances
of two Fe catalysts, prepared by impregnation, unpromoted or promoted by K (Duan
et al., 2016). Table 7.6 presents a comparison of the FT selectivity of the three catalysts under similar CO conversions. The catalyst prepared by the redox method
exhibited much higher selectivity of hydrocarbons (i.e., 70.4%C), ascribable to the
smaller size iron nanoparticles and higher degree of carbidization. Moreover, the latter
sample exhibited much lower CH4 selectivity and improved C5 + selectivity than the
impregnated catalysts, suggesting higher chain growth probability. Such parameter, in
fact, increases with Fe particle sizes and K addition favors chain growth (α). Moreover, the innovative catalyst displayed the highest olefin-to-paraffin ratio. The effect
of three alkali metal promoters (Li, Na, and K) on the catalytic performances of Fe
catalysts supported on carbon nanotubes for Fischer-Tropsch reaction at 275°C,
8 bar, and H2/CO ratio of 2 was studied in a work of Xiong et al. (2015). The addition
of alkali promoters led to an increase in crystallite size of the iron oxide and decreased
surface areas, as compared with the unpromoted Fe/CNT catalyst. The presence of Na
and K promoters slightly hindered catalyst reducibility by increasing the reduction
temperature of the iron oxide, while the potassium-promoted catalyst showed the most
pronounced effect, and no effect was observed for Li. The sodium- and potassiumpromoted catalysts were found to decrease the methane selectivity, increase the olefin
production, and shift the product selectivity to higher-molecular-weight hydrocarbons
during FTS. Furthermore, Na and Li greatly increased the CO conversion, while the
addition of K suppressed the activity. As a result, the catalyst promoted by Na resulted
in the largest increase in FTS reactivity compared with Li and K. It was also reported
(Li et al., 2016b) that K addition to Fe catalysts supported over graphite promoted the
reduction behavior and enhanced the selectivity to liquid hydrocarbons significantly at
260°C, 20 bar, and H2/CO ¼ 1. Products were mainly composed of C4–C10 α-olefins
with little methane, whose distributions changed with time onstream. The α-olefins in

Catalysts for conversion of synthesis gas

241

the liquid phase reaction media promoted the selectivity of C5 + distillate up to near
90% while suppressing the formation of lighter hydrocarbons with higher CO
conversion.
Ma et al. (2007) investigated the effect of K addition to an Fe-Cu-Mo catalyst
supported on activated carbon. Temperatures ranging between 260°C and 300°C,
20.8 bar, H2/CO ¼ 0.9, and the promotion by 0.9 wt% of potassium improved both
FT and WGS activity, while an opposite trend was observed for the 2 wt% of K.
The potassium promoter significantly suppresses formation of methane and methanol
and shifts selectivities to higher-molecular-weight hydrocarbons (C5 +) and alcohols
(C2–C5). Meanwhile, the potassium promoter changes paraffin and olefin distributions. At least for carbon numbers of 25 or less, increasing the K level to 0.9 wt%
greatly decreases the amount of n-paraffins and internal olefins (i.e., those with the
double bond in other than the terminal positions) and dramatically increases branched
paraffins and 1-olefins, but a further increase in the K level shows little additional
improvement. The addition of potassium changes the effect of temperature on the
selectivity to oxygenate. In the absence of K, oxygenate selectivity decreases with
temperature. However, when K is present, the selectivity is almost independent of
the temperature.
Many examples dealing with the use of Fe catalysts combined with zeolites are
available in recent literature. For FT synthesis at 280°C, 10 bar, and H2/CO ¼ 1, it
was observed (Yoneyama et al., 2005) that, before adding zeolite, to Fe, FTS products
mainly contained normal paraffins with long chain from C1 to C16. After adding the
zeolite, heavy hydrocarbons disappeared, and light hydrocarbons from C1 to C10 rich
in isoparaffins were produced. Methane selectivity of Fe hybrid catalyst was very low,
compared with Co hybrid catalyst at the same conditions, as Fe FTS catalyst had low
CH4 selectivity at higher temperature such as zeolite’s best reaction temperature.

These results indicated that the hybrid catalysts containing Fe FTS catalysts and
H-ZSM-5 for producing isoparaffins at one-step reaction were very effective.
Baranak et al. (2013) evaluated the influence of preparation method on the FT activity
of ZSM-5 supported iron catalysts. Zeolite-supported catalysts were synthesized by
using incipient wetness impregnation method, and hybrid catalyst was prepared by
physical admixing of ZSM-5 and base iron. At 280°C, 19 bar, and H2/CO ¼ 2, all catalysts displayed a CO conversion higher than 40%; the selectivity toward C5–C11
hydrocarbons of catalyst prepared by impregnation method was determined to be
50%–74%. The selectivity of the hybrid catalyst toward the same fraction was about
45%. No wax was detected in the products during the FT process using zeolitesupported iron catalysts. However, the impregnated catalyst displayed a stable behavior for 260 h of time onstream without any activity loss. In addition, the choice of low
acidity ZSM-5 support lead to lower selectivity for the light hydrocarbon and high
selectivity for gasoline-range components. The effect of Si/Al ratio, which influenced
support acidity, in ZSM-5 for FT reaction over Fe-based catalysts was also studied
(Plana-Pallejà et al., 2016). In fact, zeolite acidity is responsible for the cracking of
heavy hydrocarbons, and the formation of aromatics through oligomerization, cyclization, and dehydrogenation of primary short olefins. In particular, the increment in
acidic sites (low Si/Al ratios) induced the formation of more complex aromatic