Applied Catalysis A: General 470 (2014) 250–260

Contents lists available at ScienceDirect

Applied Catalysis A: General

journal homepage: www.elsevier.com/locate/apcata

Review

Fischer–Tropsch synthesis: A review of the effect of CO conversion

on methane selectivity
Jia Yang a,b , Wenping Ma b , De Chen a , Anders Holmen a , Burtron H. Davis b,∗
a
Norwegian University of Science and Technology, NO-7491 Trondheim, Norway
b
Center for Applied Energy Research, University of Kentucky, 2540 Research Park Drive, Lexington, KY 40511, United States

a r t i c l e i n f o a b s t r a c t

Article history: Methane is the least desired product for the Fischer–Tropsch synthesis so that it is of paramount impor-
Received 1 July 2013 tance to reduce methane selectivity in the FT process. Despite numerous efforts devoted to the reduction
Received in revised form 30 October 2013 of methane selectivity, the effect of CO conversion on methane selectivity is still not well defined.
Accepted 31 October 2013
For cobalt and ruthenium based catalysts, methane selectivity generally decreases monotonically with
Available online 8 November 2013
increasing CO conversion within the range 20–80%, while on iron catalysts, the methane selectivity is
more influenced by its water–gas shift activity and potassium promotion. Methane selectivity remains
Keywords:
more or less constant at conversion lower than 70% for potassium promoted iron catalysts. Pressure and
Cobalt
Ruthenium temperature have a greater influence on methane selectivity for Co and Ru based catalysts. Pressure and
Iron temperature change the preference of the secondary reactions of primary olefins and tune methane selec-
Fischer–Tropsch synthesis tivity at different CO conversions. An increased extent of olefin readsorption may compete with methyl
Methane selectivity intermediates for surface sites and hence reduce methane selectivity. Water seems to play an important
Water effect role in determining the dependence of CH4 selectivity on CO conversion for the Co and Ru based catalysts
by either inhibiting the hydrogenation reaction or by increasing the amount of surface carbon for chain
growth. Choosing appropriate promoters and process conditions may reduce methane production.
© 2013 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
2. Effects of active components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
2.1. Co . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
2.2. Ru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
2.3. Fe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
2.4. Summary of the effect of active components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
3. Effects of process conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
3.1. Pressure effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
3.2. Temperature effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
3.3. Time on stream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
4. Possible mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256
4.1. Effect of readsorbed olefin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
4.2. Effect of water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
4.2.1. Water inhibiting hydrogenation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 257
4.2.2. Water increases the concentration of monomeric carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
5. Strategies to reduce methane selectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
6. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259

1. Introduction

Fischer–Tropsch synthesis (FTS) provides an alternative route

∗ Corresponding author. Tel.: +1 859 257 0251; fax: +1 859 257 0302. for the production of clean fuels from coal, natural gas and biomass.
E-mail address: burtron.davis@uky.edu (B.H. Davis). FTS is a stepwise polymerization reaction and the products in

0926-860X/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.apcata.2013.10.061
 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
general follow an Anderson–Schulz–Flory distribution. The primary
objective is to produce as much transportation fuels as possible
[1]. Methane is the least desired product and the amount that is
produced needs to be reduced. The product selectivity is strongly
related to the process conditions. An increase in temperature shifts
the distribution towards the products with lower carbon numbers
on iron [2,3], ruthenium [4] and cobalt [4]. An increase in total pres-
sure usually shifts the product selectivity to heavier products [4].
Increasing H2 /CO ratios results in producing more amounts of light
hydrocarbons and lower olefins [2,3]. Less well defined is the prod-
uct selectivity with conversion at one set of reaction conditions.
This paper concentrates on the influence of CO conversion induced
by varying synthesis gas space velocity on the methane selectivity.
The effect of CO conversion on product distribution strongly
depends on the secondary reactions of the primary products, espe-
cially olefins. Secondary reactions of primary olefins may include
hydrogenation, reinsertion, cracking or hydrogenolysis and isomer-
ization. The dominant secondary reactions of the primary olefins
depend on the operating conditions and the catalyst. Cracking or
hydrogenolysis is enhanced by high temperature and is strongly
inhibited by high CO pressure and water partial pressure [5].
According to Bond [6], the initial step of hydrogenolysis is the loss
of hydrogen atoms with the formation of radicals. Generally a nega-
tive reaction order with H2 was found over Co and Fe catalysts. The
detailed mechanism is beyond the scope of this work. However,
higher temperature should favor the hydrogenolysis of hydrocar-
bons. Besides, Iglesia et al. [7], Madon et al. [8] and Komaya and
Bell [9] concluded that readsorption of an olefin and chain initi-
ation is the most important secondary reaction for Ru, Co and Fe
catalysts at typical Fischer–Tropsch conditions instead of hydro-
genation or hydrogenolysis reactions. Water has been reported to
inhibit secondary hydrogenation and hydrogenolysis [5,7]. The type
and extent of secondary reactions may change at different conver-
sion levels and hence cause a selectivity change.
A decrease in syngas space velocity increases the residence time
of reactants in the catalyst bed, and hence CO conversion increases.
At typical Fischer–Tropsch conditions, the C5+ selectivity increases
and CH4 selectivity decreases with increasing CO conversion for
the Co and Ru based catalysts. The olefin to paraffin ratio decreases
with increasing CO conversion. The increase in chain growth and
the decrease in olefin to paraffin ratio has been attributed to the
enhanced secondary reaction (readsorption and reinitiation) of ␣-
olefins at prolonged bed residence times [10]. Water may further
enhance the effect of ␣-olefin readsorption on chain growth. Water
has been reported to inhibit secondary hydrogenation of olefins
[7,11]. Therefore more ␣-olefin will be available for readsorption
and reentering the growing chain (instead of termination as paraf-
fin) at a higher water partial pressure (higher CO conversion).
However, the reason for the change of methane selectivity with CO
conversion is still not very well defined. In this paper, the depend-
ence of methane selectivity on bed residence time (CO conversion)
for typical Fischer–Tropsch Fe, Co and Ru catalysts will be reviewed
and the possible mechanisms to account for this will be discussed.
2. Effects of active components
The selectivity of FTS depends primarily on which transition
metal (Fe, Co or Ru) is used. Iron based FT catalysts are active for
the water–gas-shift (WGS) reaction and are therefore suitable for
low H/CO feed. On the other hand, the WGS activities on Co and Ru
catalysts are negligible under normal FTS conditions. Therefore the
product selectivity of Fe catalyst depends both on FT activity and
WGS activity, while Co and Ru depend mainly on their FT activities.
Therefore, the change of product distribution with CO conversion
is quite similar for both Co and Ru catalysts. Fe behaves differently
251
with increasing CO conversion. The methane selectivity trend with
CO conversion for Co, Ru and Fe are discussed respectively.
2.1. Co
Most of the studies on cobalt based catalysts report a decrease
in methane selectivity and a concomitant increase in C5+ selectiv-
ity with increasing CO conversions under typical FTS conditions
[12–17]. Holmen and coworkers [12,13] investigated Co catalysts
supported on Al2 O3 , SiO2 and TiO2 with and without Re promotion
in a fixed-bed reactor and observed a common trend with CO con-
version 10–50% range for all of the catalysts that were tested. Thus,
the support identity and Re promoter did not influence the depend-
ence of selectivity on the CO conversion. Davis and coworkers
[14,15] did thorough studies on the unpromoted and Ru promoted
Co catalysts on Al2 O3 support over a wide range of CO conversion
(12–90%) both in fixed-bed and slurry phase reactors. In the fixed-
bed reactor, using a 15%Co/Al2 O3 , the selectivity to CH4 decreased,
whereas the C5+ selectivity increased linearly with increasing CO
conversion within the range 35–90%. CO2 selectivity increased also
with increasing CO conversion due to the increased WGS reac-
tion at high water partial pressures. 1-Olefin content of the C2 –C5
fraction decreased and the 2-olefin content of C4 hydrocarbons
increased with increasing CO conversion. In the slurry phase exper-
iment (Fig. 1), CH4 and C5+ selectivities followed the same trends
as those observed in the fixed-bed reactor in the CO conversion
range 10–80%. The formation of methane decreased with increasing
CO conversion. However, at high conversions (>80%), CH4 and CO2
selectivity increased exponentially with increasing CO conversion.
The C5+ selectivity remained nearly unchanged for CO conversions
above 80%. When CO conversion exceeds 80%, part of Co was oxi-
dized as evidenced by XANES results [14], possibly by the high
water partial pressure (PH2 O /PH2 > 0.9), and cobalt oxide surfaces
may contribute to the enhanced WGS reaction and methane for-
mation reactions.
For a 10%Co/TiO2 catalyst tested in a continuously stirred
tank reactor (CSTR) at 493 K, 2.4 MPa, H2 /CO = 2 and GHSV
0.75–4 Nl gcat −1 h−1 , the methane selectivity again decreased
(15–7%) and C5+ selectivity increased (80–91%) with CO conversion
in the range 8–63% (CAER to be published data). The C2 –C4 selectiv-
ity decreased about 2% over the entire range. Therefore the increase
in C5+ selectivity is mainly due to the decrease in the methane
selectivity at higher CO conversion.
Modelling of transient experiments using isotopic labelled 13 CO
(methanation conditions) has revealed that two reaction pathways
via two different carbon pools for methane formation exist for
cobalt based catalysts [18,19]. Moreover, a recent study shows that
one reaction route is hydrogenation of carbon pool of HCHO* and
another is through decomposition of HCHO* to another carbon pool
namely CHx * which can both form methane and undergo chain
growth (Scheme 1) [19]. In order to reduce methane formation, the
first reaction route for methane formation should be minimized.
The relative importance of the two reaction routes depends on the
concentration of surface hydrogen. A low hydrogen concentration
Scheme 1. Possible reaction mechanism for methane formation [19].
252 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
Fig. 1. (a) Methane selectivity, (b) C5+ selectivity, (c) CO2 selectivity, (d) H2 /CO usage ratio as a function of CO conversion. (0.27%Ru–25%Co/Al2 O3 , 493 K, 1.4 MPa, H2 /CO = 2.1
and 0.3–15 Nl gcat −1 h−1 ) [14].
can reduce the contribution of reaction route I to methane for-
mation. The decrease in methane formation with increasing CO
conversion can be possibly explained by a decrease in surface H/CO
ratio due to an increase in H2 O concentration. The adsorption of CO
and the intermediates such as O* and OH* from H2 O decomposition
is relatively strong on the surface, which could reduce the adsorp-
tion of H. The effect of H2 O on the hydrogen surface concentration
is consistent with the observation that cracking or hydrogenolysis
is strongly inhibited by high CO and water partial pressures [5].
The relevance of the CO hydrogenation at methanation conditions
(483 K, 0.185 MPa, H2 /CO = 10) and FTS conditions was investigated
by den Breejen et al. [20]. They found that the activity and selec-
tivity trends with CO hydrogenation (483 K, 0.185 MPa, H2 /CO = 10)
agreed well with the trend found for FT conditions (483 K, 3.5 MPa,
H2 /CO = 2) and suggested that the results obtained at methanation
conditions are relevant for FTS.
So far, the data all showed a considerable decrease in methane
selectivity with increasing CO conversion. On average, methane
decreases about 1% when the CO conversion increases 10% for
cobalt catalysts on different supports with and without promo-
tion at a given set of reaction conditions (493–498 K, 1.4–2.4 MPa,
H2 /CO = 2). There were also exceptions. For a 12.4%Co/SiO2 cata-
lyst tested in a slurry phase reactor (Fig. 2), methane selectivity
decreased only slightly from 6% to 5.5% with CO conversion in the
range 22.1–77.7% at 483 K, 2.02 MPa and H2 /CO = 2 [21]. Similarly,
CH4 selectivity was found to decline slightly from 7.5% to 6.5% for a
0.2%Ru10%Co/TiO2 (T = 503 K, P = 2.03 MPa, H2 /CO = 2.0, CSTR) over
the conversion range 40–80% [22]. In another case, CH4 selectivity
remained more or less constant for a Co-Pt/Al2 O3 catalyst (T = 483 K,
P = 2.93 MPa, H2 /CO = 2.0, CSTR) over the conversion range 20–50%
[23]. The reason why methane selectivity did not change signifi-
cantly with CO conversion might be related to the type of support
and promoters.
2.2. Ru
With regards to methane selectivity as a function of CO conver-
sion, Ru behaves similarly to Co. A U.S. patent (1986) [24] showed
that the methane selectivity as a function of CO conversion follows
almost the same trend on 12%Co0.5%Re/TiO2 and 1%Ru/TiO2 cat-
alyst in the range of CO conversion of 20–80% at 473 K, 1.93 MPa,
H2 /CO = 2.15 (Fig. 3). The methane selectivity is about 2% higher on
Co than on Ru for a CO conversion of about 95%.
Iglesia and coworkers [5,10] examined the effect of bed res-
idence time on hydrocarbon selectivity for titania supported
ruthenium catalysts at 476–483 K, H2 /CO = 2.1, 0.55–0.6 MPa using
a fixed-bed reactor (Fig. 4). Product molecular weight increases
with increasing bed residence time. C5+ selectivity increases, while
CH4 selectivity decreases, as bed residence time and CO conversion
increase on the Ru/TiO2 catalysts. The absolute increase in C5+ selec-
tivity is more than the decrease in methane selectivity, suggesting
the decrease of C2 –C4 hydrocarbon selectivity also contributed to
higher C5+ selectivity.
The decrease of methane with CO conversion for ruthenium and
cobalt in Fig. 3 (1% for a CO conversion increase of 25%) seems to
be less significant as compared to the cobalt catalysts presented in
Fig. 1 (1% for a CO conversion increase of 10%). This is most likely
due to different operation temperatures. The cobalt catalyst was
tested at 493 K, and is 20 K higher than in the case of Fig. 3. An
increase in temperature in general results in more methane for-
mation for FTS. The level of the methane selectivity is higher at
relative high temperature (493 K); therefore there is more room
for improvement. This is supported by a recent work carried out at
CAER where a 2.5%Ru/NaY catalyst was tested at 493 K, 1.93 MPa,
H2 /CO = 2 using a CSTR (CAER unpublished data). Consistent with
what has been reported, the methane selectivity decreased and C5+
selectivity increased almost linearly with increasing CO conversion.
 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
Fig. 2. CO conversion, methane and CO2 selectivity (mol%C based) as a function of space time. (12.4%Co/SiO2 , 483 K, 2.02 MPa and H2 /CO = 2) [21].
Fig. 3. Selectivity of Co–Re and Ru as a function of CO conversion [24].
Fig. 4. Methane and C5+ selectivity (C%) as a function of bed residence time (CO
conversion). (Ru/TiO2 , 476 K, H2 /CO = 2.1, 0.6 MPa, 5–60% CO conversion) [10].
253
The magnitude of the decline is about 1% for every 10% increase in
CO conversion, similar to the cobalt catalyst tested at the same tem-
perature. This result demonstrates that how significant methane
selectivity changes with CO conversion depends on the reaction
temperature. At relatively low temperature (e.g. 473 K), the CH4
selectivity is generally low, increase CO conversion could decrease
methane selectivity marginally, and decrease in C2 –C4 hydrocarbon
contributes more to higher hydrocarbon formation. At relatively
high temperature (e.g. 493 K), methane selectivity is relatively high.
The enhanced C5+ selectivity mainly arises from the inhibition of
methane formation at higher CO conversion level.
2.3. Fe
The correlation of methane selectivity with CO conversion is not
straightforward for Fe based FT catalysts. The commercial iron cat-
alyst is a multicomponent system normally promoted with K to
improve the selectivity to higher molecular weight hydrocarbons
[25]. In addition, Fe catalysts are active for the WGS reaction and
can therefore also tune product selectivity. In general, the methane
selectivity is low compared to the Co and Ru tested at similar con-
ditions. With iron, methane selectivity remains low (e.g. 7C %) even
at the high temperature (∼340 ◦ C) of the Synthol process [26,27].
For a commercial supported iron catalyst (Fe/Cu/K/SiO2 ) tested
using a slurry reactor, the space velocity did not impact the prod-
uct distribution at 523 K, 1.5 MPa, H2 /CO = 0.67 and CO conversion
range 26–56% [28]. On a precipitated iron-silica catalyst promoted
with potassium, Raje et al. [29] found methane selectivity was
constant at 5–6% of hydrocarbons for a low alpha catalyst (with
atomic composition of 100Fe/4.6Si/0.36 K) using a CSTR and the
methane selectivity was similar for all the catalysts studied at low
carbon monoxide conversions (<50%) (Fig. 5). Above 50% carbon
monoxide conversion, the methane selectivity increased slightly
with increasing conversion. Similar results have been found on
a 100Fe/4.6Si/Cu/1.4 K (atomic ratio) catalyst [30]. Another study
also reported a case where an increase in syngas partial pressure
resulted in an increase in FT reaction rate (CO conversion) while
the methane selectivity remained constant for a potassium pro-
moted iron catalyst (513 K, H2 /CO = 2, 0.5–2.5 MPa) [17]. The lack
of dependence of methane selectivity on the CO conversion might
be due to the promotion effect from K or the WGS activity of iron.
K was known to modify the selectivity and is used regularly as a
promoter for iron catalysts [31,32]. In one of the previously men-
tioned cases (Fig. 5), the level of potassium had little impact on
methane production up to about 50% CO conversion and above this
level increasing potassium content slightly decreases the amount of
methane that is produced. It was noticed that the H2 /CO ratio varied
from 0.5 to 3 in the CO conversion range investigated (Fig. 5b), and
methane selectivity remained more or less constant when CO con-
version is lower than 70% (H2 /CO ratio 0.5–0.8). The WGS activity
254 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260

Fig. 5. (a) Methane selectivity and (b) exit H2 /CO ratio as a function of carbon monoxide conversion on an iron-silica catalyst (543 K, H2 /CO = 0.67 and 1.3 MPa). ,
100Fe/4.6Si/0.36 K; , 100Fe/4.6Si/1.4 K; and ♦, 100Fe/4.6Si/2.2 K [29].

is significantly enhanced when CO conversion is higher than 80% as
evidenced by higher H2 /CO ratio for catalyst with 100Fe/4.6Si/1.4 K.
Methane selectivity increased significantly at the higher H2 /CO
ratio (1–3). The exact mechanism on how potassium reduces the
methane selectivity and increases the chain growth probability is
still not known. Attempts have been made to get better understand-
ing of the effect of alkali on the FTS selectivity. The formation of
methane over unpromoted and potassium-promoted bulk iron cat-
alysts applied in FTS was studied by dosing carbon monoxide pulses
in hydrogen [33]. CH4 response curves showed two formation path-
ways. The authors concluded that the presence of K reduces the
formation of CH4 by blocking the fast formation channel and by
establishing a new and slower reaction pathway without changing
the reaction pathway towards higher hydrocarbon [33]. It is likely
that the slow pathway is less dependent on the H2 /CO ratio or the
rate is so slow as compared to chain growth step that the effect
of H2 /CO ratio on the methane formation rate is less pronounced
over the conversion range 20–70%. Investigation on hydrogenol-
ysis of hydrocarbons has also shed some light on the mode of
action of alkali added to iron catalysts for FTS [6]. According to the
author, alkali’s action is not in promoting the formation of C C
bonds, but rather in maintaining the adsorbed intermediates in the
fully saturated state required for further chain growth. Similar to
potassium, other Group I alkali metal (Na, Rb, Cs) promoted iron cat-
alysts have the same methane trend as a function of CO conversion
in the range 20–80% [34]. In addition, except for the Li-promoted
and the unpromoted iron catalysts, the fraction of methane in the
hydrocarbon products does not depend upon the type of alkali
promoter.
Apart from alkali promoters, WGS reaction may also tune the
methane selectivity at different CO conversions, especially at con-
version levels higher than 80%. The methane selectivity increased
significantly with H2 /CO ratio in that range. In CO hydrogenation
experiments with the presence of CO2 , Dry [4] provided a relation
between the methane selectivity and the factor PH2 1/2 (PCO+CO2 )−1
at reactor inlet for alkaline-promoted fused iron catalysts in a high
temperature fluidized-bed reactor. Higher CO2 pressures in the feed
result in a decrease of the methane selectivity. CO2 may tune H2 /CO
ratio through the reverse WGS reaction at high temperatures.
To summarize, relatively low methane selectivity is observed
for potassium promoted iron catalysts. CO conversion has little
effect on methane selectivity in the CO conversion range lower
than 70%. When CO conversion is higher than 70%, the WGS reac-
tion becomes significant compared to FTS and the CH4 selectivity

Table 1
Summary of the methane selectivity trend with increasing conversion for Co, Ru and Fe catalysts.

Catalystsa Promoter Support Reactor T (K) Syngas pressure H2 /CO ratio SCH4 trend with SCH4 at ca. 50% References
(MPa) Xco (ca. 20–70%) conversionb

12Co/␥-Al2 O3 – ␥-Al2 O3 Fixed-bed 483 1.9 2.1 Decrease 9* [13]

15Co/␥-Al2 O3 – ␥-Al2 O3 Fixed-bed 493 1.9 2.0 Decrease 11.0* [15]
20Co0.5Re/␥-Al2 O3 Re ␥-Al2 O3 Fixed-bed 483 1.9 2.1 Decrease 9* [12]
12Co0.5Re/␥-Al2 O3 Re ␥-Al2 O3 Fixed-bed 483 1.9 2.1 Decrease 8.7* [13]
25Co0.48Re/␥-Al2 O3 Re ␥-Al2 O3 CSTR 493 2.2 2.1 Decrease 7.1 [35]
25Co0.27Ru/␥-Al2 O3 Ru ␥-Al2 O3 CSTR 493 2.2 2.1 Decrease 7.1 [14]
25Co0.5Pt/␥-Al2 O3 Pt ␥-Al2 O3 CSTR 493 2.2 2.1 Decrease 8.3 [35]
15Co0.5Pt/Al2 O3 Pt Al2 O3 CSTR 483 2.9 2.0 Constant 8.0 [23]
12Co/SiO2 – SiO2 Fixed-bed 483 1.9 2.1 Decrease 8.7* [13]
12.4Co/SiO2 – SiO2 CSTR 483 1.4 2.0 Decrease slightly 5.7* [21]
(0.5%)
12Co0.5Re/SiO2 Re SiO2 Fixed-bed 483 1.9 2.1 Decrease 8.5* [13]
12Co/TiO2 – TiO2 Fixed-bed 483 1.9 2.1 Decrease 9.0* [13]
12Co0.5Re/TiO2 Re TiO2 Fixed-bed 483 1.9 2.1 Decrease 9.0* [13]
10Co0.2Ru/TiO2 Ru TiO2 CSTR 503 1.7 2.0 Decrease slightly 7.6 [22]
(1%)
1%Ru/TiO2 – TiO2 Fixed-bed 473 1.9 2.2 Decrease 4.0* [24]
1.2%Ru/TiO2 – TiO2 Fixed-bed 476 0.6 2.1 Decrease 3.7* [10]
2.5%Ru/NaY – NaY CSTR 493 1.9 2.0 Decrease 3.7* CAER unpublished
data
100Fe/4.3Cu/4.1K/25SiO2 Cu, K – CSTR 523 1.5 0.67 Little variation 5.3 [28]
100Fe/4.6Si/K K – CSTR 543 1.3 0.7 Little variation 5.0–6.0 [29]
100Fe/4.6Si/Cu/1.4K Cu, K – CSTR 543 1.3 0.7 Little variation 4.5–5.4 [30]
Fe/K K – CSTR 513 0.5–2.5 2.0 Little variation – [17]
a
The concentration of cobalt and ruthenium based catalysts are wt%, the concentration of active metal of iron based catalysts are atomic composition of the catalysts.
b
*After the number reflect interpolated value at about 50% conversion level.
 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
Fig. 6. (a) Methane selectivity, (b) C2 –C4 selectivity, (c) C5+ selectivity (C%) and (d) C2 –C4 olefin to paraffin ratio as a function of CO conversion. (12%Co/SiO2 , 493 K, H2 /CO = 2
at 0.1 MPa and 1.01 MPa) [36].
increases rapidly due to higher H2 /CO ratios caused by the WGS
reaction.
2.4. Summary of the effect of active components
The methane selectivity trends with increasing conversion
(20–80%) were summarized in Table 1. The methane selectivity
decreases with CO conversion in general at conversion level below
80% for most of Co catalysts under different temperatures and
pressures. However, for some cobalt catalysts supported on silica,
methane selectivity depends very weakly on the CO conversion.
The methane selectivities at about 50% conversion were also pre-
sented in Table 1. The methane selectivity (at 50% conversion)
varied between 5.7% and 11% for cobalt-based catalysts with differ-
ent support and promoters. No clear correlation was found between
support materials or promoters with methane selectivity at 50%
conversion. Possible mechanisms for the change of methane selec-
tivity with CO conversion are discussed in Section 4.
For ruthenium-based catalysts, the methane selectivity
decreased with increasing conversion for all the catalysts reviewed
in conversion range 10–90%. The methane selectivity at 50%
conversion level was all at about 4%, much lower than cobalt based
catalyst, suggesting that with ruthenium catalysts, one can achieve
lower methane selectivity at similar productivity level.
Fe based FT catalysts are normally promoted with K, and Fe itself
is active for WGS reaction at high temperatures. The variation of
methane with CO conversion is usually small, within 2% at CO con-
versions lower than 60–70%. When CO conversion exceeds 70%,
WGS activity becomes higher and the methane selectivity increases
with increasing CO conversion due to higher H/CO ratio result-
ing from the enhanced WGS activity. The methane selectivity at
255
50% conversion level was between 4.5% and 6%, depending on the
amount of potassium. The iron catalysts were tested at much higher
temperature (513–543 K vs. 483–493 K) and the methane selectiv-
ity was comparable low as for ruthenium catalysts and much lower
than for cobalt catalysts.
3. Effects of process conditions
3.1. Pressure effect
Pressure has a profound effect on how product selectivity
changes with CO conversion. In a study of the effects of Mn pro-
motion on the activity and selectivity of a 12%Co/SiO2 catalyst
for Fischer–Tropsch synthesis using a fixed-bed reactor, Bell et al.
[36] showed that at 0.1 MPa, the selectivity for all products was
unaffected by CO conversion (5–50%) at 493 K, H2 /CO = 2 (Fig. 6).
However, the olefin to paraffin (O/P) ratio of the C2 –C4 fraction
decreased with increasing CO conversion due to increased hydro-
genation of the olefins with increasing space time. At 1.01 MPa,
the C2 –C4 olefins was claimed to be incorporated into the growing
hydrocarbon chain, leading to a decrease in C2 –C4 selectivity and
an increase in C5+ selectivity with increasing CO conversion.
Tavasoli et al. [37] studied the effect of bed residence time on
hydrocarbon selectivity at low pressure (0.1 MPa, H2 /CO = 2) for a
15%Co/␥-Al2 O3 catalyst using a fixed-bed reactor (Fig. 7). It was
found that the CO conversion level increases linearly with bed resi-
dence time up to 15% CO conversion. The CH4 selectivity decreases
while the C5+ liquid hydrocarbon selectivity increases as CO con-
version increases from 1.4% to 5%. Selectivity towards light gaseous
hydrocarbons, i.e., C2 –C4 decreases also with increasing CO con-
version within this range. Further increases in CO conversion from
256 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
Fig. 7. Methane, C2 –C4 and C5+ selectivity as a function of CO conversion [37].
(15%Co/␥–Al2 O3 , 493 K, H2 /CO = 2 and 0.1 MPa).
5 to 15% do not influence the methane, C2 –C4 or C5+ selectivity.
The authors attributed the higher CH4 selectivity at low conver-
sion (<5%) to the idea that methane is produced mainly at the very
beginning of the catalytic bed. It has been shown that co-feeding
water along with ethylene to H2 /CO or adding ethylene below
the reactor inlet significantly decreased the fraction of ␣-olefin
hydrogenated to the corresponding paraffin [7]. It was concluded
that water, whether indigenous or externally added, inhibits the
secondary hydrogenation of ␣-olefins during FTS. Similarly, the
hydrogenation of the methane precursor may be inhibited by water
at relatively high CO conversion levels (high water partial pressure).
At lower conversion levels (<5%), methane is quite high due to the
small amount of water present; when CO conversion is higher than
5%, a threshold amount of water is reached and therefore effectively
inhibits the hydrogenation reaction. Hence, all of the hydrocar-
bon selectivities do not change with CO conversion for conversions
higher than 5% in this case.
The examples indicate that at low pressure (atmospheric pres-
sure), secondary hydrogenation is dominant for readsorbed olefin,
and the product distribution does not change with CO conversion
except for extremely low CO conversion (<5%). At high pressure
(2 MPa), reinsertion is considered to be a dominant pathway for
readsorbed olefin and the products shifts to heavier hydrocarbon. In
addition, methane decreases significantly with increasing CO con-
version. The mechanism for this will be discussed later. Considering
the similarity of Ru and Co, both are good hydrogenation catalysts
and have low WGS activity, the selectivity trends with CO conver-
sion for Ru catalyst is expect to be similar to Co at both low and
high pressure. However, for iron catalysts, it is more complicated.
Investigations at Sasol [38] showed that with precipitated iron cat-
alysts at low temperatures, the H2 /CO ratio within the reactor was
the most important parameter which controlled the selectivity but
with fused catalysts at higher temperatures the CO2 partial pressure
appears to dictate the product selectivity.
3.2. Temperature effect
According to the Arrhenius equation, temperature influences
the reaction rate constant to a different extent, based on the
Fig. 8. Evolution of carbon monoxide conversion and methane selectivity over
Co/Al2 O3 catalyst in a slurry reactor (T = 493 K, 2 MPa, inlet H2 /CO = 2.2) [41].
magnitude of the activation energy. For example, an increase of the
temperature could change the preference for the secondary reac-
tions of the primary products of FTS and therefore change product
selectivity. With iron, methane selectivity remains low (e.g. 7C%)
even at the high temperature (∼340 ◦ C) of the Synthol process
[26,27]. However, methane selectivity is more sensitive to tem-
perature for Co and Ru catalysts due to their good hydrogenation
properties.
Everson et al. [39] studied the effect of space velocity on the
product distribution for a 0.5%Ru/Al2 O3 catalyst at 0.8, 1.2 and
1.6 MPa within the 498–577 K temperature range using a stirred
gas–solid reactor. At temperatures around 500 K and 525 K, the
methane and C5+ selectivities appear to be largely independent of
space velocity (and consequently conversion). With temperatures
higher than 550 K, increasing space velocity decreases conversion,
and hence the methane content decreases, and the C5+ fraction
increases with decreasing CO conversion. This is opposite to the
trend that has been observed previously. The authors attributed
this to the hydrogenolysis of higher hydrocarbon at high temper-
ature (>550 K) and low CO coverage. It has been reported that
secondary hydrogenolysis to produce methane may occur on FT
catalysts [9,40]. However, hydrogenolysis decreases strongly with
increasing CO and H2 O pressures [10] and for reactor temperatures
lower than 548 K. It seems in this particular case, hydrogenolysis
becomes the dominant secondary reaction for readsorbed olefins at
temperature higher than 550 K and hence CH4 selectivity increases
with increasing CO conversion at prolonged bed residence time.
3.3. Time on stream
Methane selectivity also changes with time on stream as the
catalyst deactivates. Methane selectivity in general increased with
decreasing CO conversion as the catalyst deactivates with time-
on-stream for both iron [3,28,42] and cobalt catalysts [15,41,43].
This is consistent with the trend for varying space velocity. In one
particular case, methane was reported to increase from about 8%
to 15% as CO conversion decreased from 60% to about 30% for a
Co/Al2 O3 catalyst using a slurry phase reactor at 493 K, 2 MPa and
H2 /CO = 2.2 as shown in Fig. 8 [41].
4. Possible mechanisms
As discussed before, methane is the least desired products in
the FTS. It is of paramount importance to understand why methane
selectivity changes in order to reduce methane production. There
 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
are several explanations for the methane selectivity trend with
increased bed residence time (increasing CO conversions) which
can be mainly categorized into two groups: the effect of readsorbed
olefin and the effect of water.
4.1. Effect of readsorbed olefin
Decreased concentration of surface methyl chains by competi-
tive adsorption with olefins
Iglesia and coworkers [5,10] found on a Ru/TiO2 catalyst that
methane selectivity decreases with increasing CO conversion; how-
ever, the methane chain termination probability ␤ is independent
of bed residence time (Fig. 9a). The methane termination proba-
bility was defined as the ratio of the rate of H addition to surface
methyl group and the rate of monomer addition to methyl group,
as shown in Eq. (1).
kH,1 [CH3 ∗][H∗] kH,1 [H∗]
ˇH,1 = = (1)
kp,1 [CH3 ∗][CHx ∗] kp,1 [CHx ∗]
It was suggested that the observed decrease in CH4 selectivity as
bed residence time increases was due to decreases in the concen-
tration of surface methyl group [CH3 *] because changes in surface
monomer [CHx *] or hydrogen concentrations [H*] would lead to
a similar decrease in ˇH,1 [10]. Similar effects, methane selectiv-
ity decrease and methane termination probability ˇH,1 , remaining
constant was observed when olefins were added to the H2 /CO feed
(Fig. 9b), suggesting the olefins compete with surface methyl for
available chain growth sites. Fig. 9a and b show that the bed resi-
dence time and ethylene co-feeding had similar effect on methane
selectivity and methane termination probability. This suggests that
at prolonged bed residence time, the extent of olefin readsorption
increases and the synthesized olefins function the same as co-fed
olefins, may compete the adsorption sites with methyl groups and
therefore cause the decrease in methane selectivity. The extent of
␣-olefin readsorption increases with the bed residence time due to
diffusion limited removal of the products [7], enhanced solubility
[44–47] or enhanced physisorption of the products [9].
This mechanism is indirectly supported by ethylene co-feeding
studies to Fischer–Tropsch catalysts. Decreases in methane selec-
tivity upon co-feeding ethylene have been reported for Fe [48,49],
Co [50,51] and Ru catalysts [51–55]. Snel and Espinoza [49] reported
a 50% decrease in methane selectivity when co-feeding 10 vol%
ethylene with syngas for a carbon containing iron calcium cata-
lyst. The authors ascribed the decreased methane to the scavenging
effect of olefin for surface H. Hutchings et al. [50] found a 40%
decrease in methane upon co-feeding syngas with 11% ethylene to
a Co/MnO catalyst. They disagreed with the explanation of scaveng-
ing surface H based on the fact that ethylene mainly incorporated
into growing chains instead of being hydrogenated (consumption
H) at typical reaction conditions and the CO conversion did not
change with co-feeing of ethylene, since a decrease in surface H
coverage would lead to a decrease in overall activity. On the other
hand, reports of a constant [53,56] or increased [48] CH4 selec-
tivity with ethylene co-feeding were presented. The increase of
CH4 selectivity in the latter case is possibly due to hydrogenol-
ysis of ethylene at a much higher reaction temperature of 743 K
[48]. If the olefin can compete for surface sites with the methyl
chain, the readsorbed ␣-olefin would occupy more surface sites and
hence decrease the surface concentration of methyl chain initiators,
and therefore decrease the methane formation. The mechanism
involving the readsorption of alkenes is complex and will require
additional experimental studies.
Alkene reincorporation is considered from two viewpoints. In
the first, we may consider those alkenes the alkene that has des-
orbed from the catalyst particle within which it was formed and
has joined the liquid phase surrounding the catalyst particle. The
257
addition of a 14 C-labeled alkene to the liquid phase will allow one
to define the extent of readsorption and chain initiation for the bulk
alkenes. This has been done by adding alkenes between carbon
numbers two and nineteen [e.g., 46, 58–60]. These studies show
that alkene hydrogenation is more rapid than alkene reincorpo-
ration to initiate chain growth. Thus, alkene reincorporation once
it becomes a component of the solvent system to initiate a new
chain growth makes a small contribution to the overall produc-
tion of products. In the second, the alkene has been considered
to readsorb and initiate chain growth prior to exiting the catalyst
particle (e.g., 5, 8, 10 of this paper). Deuterium tracer studies indi-
cate that when product holdup effects due to vapor–liquid effects
are eliminated, the readsorption and chain initiation by alkenes
is minimal. Thus, alkene reincorporation has been utilized in this
review both from a historical aspect and from the small contribu-
tion that it may make on impacting the amount of methane that is
produced.
4.2. Effect of water
The effect of water on the Fischer–Tropsch activity and
product selectivity of cobalt based catalysts has been studied
extensively [13,22,60–73]. Adding water decreases methane selec-
tivity, increases C5+ selectivity and increases product olefinicity
[13,60,71,74]. The effect is reversible at relatively low water
concentration; however, at high water pressure irreversible deac-
tivation usually occurs. The exact mechanism of the effect of water
is still controversial.
First of all, transport limited arrival and removal of CO can
be excluded for being responsible for the selectivity change [60].
Steady-state isotopic transient kinetic analysis shows that the CO
inventory on cobalt FT catalysts is essentially constant over a wide
range of industrially relevant syngas pressures, indicating near sur-
face saturation with CO. Water cofeeding does not significantly
affect the CO inventory even at syngas pressures of a few bars.
Water therefore cannot compete effectively with CO for the cobalt
surface [64]. The in situ infrared spectroscopic studies lead to more
or less a similar conclusion: water does not influence the den-
sity or structure of adsorbed CO intermediates or the number of
exposed Co atoms [60]. Moreover, based on kinetic isotopic inves-
tigation using D2 O, no new pathways are claimed to be introduced
by the presence of water in H2 /CO reactant streams [60]. On the
other hand, water seems to selectively inhibits irreversible chain
termination by hydrogen addition and secondary hydrogenation
pathways possibly by competitive adsorption [11]. Another view
is that water increases the amount of surface monomeric carbon
without affecting its reactivity to products [64]. These latter two
mechanisms are discussed in detail below.
4.2.1. Water inhibiting hydrogenation
Hydrogenation of propene to propane has been studied as a
model reaction for secondary hydrogenation of olefins in the FTS.
It was shown that the olefins will to a larger extent readsorb and
participate in further chain growth rather than being hydrogenated
to paraffin under FTS conditions [11].
The influence of H2 O on secondary reactions of FTS on cobalt
and ruthenium catalysts was reported by Iglesia et al. [7]. It was
shown that co-feeding water (15 mol%) along with ethylene to
H2 /CO or adding ethylene below the reactor inlet significantly
decreased the fraction of ␣-olefin that is hydrogenated to the cor-
responding paraffin. Likewise, Hall et al. [75] observed a strong
decrease of secondary hydrogenation of added ethylene with the
addition of 1.6% water vapor for a Zr/Fe/Al2 O3 catalyst. Clearly,
water, whether indigenous or externally added, inhibits the sec-
ondary hydrogenation of ␣-olefins in the FTS. The exact mechanism
258 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260

Fig. 9. Bed residence time and ethylene olefin cofeed effects on chain termination probability and methane carbon selectivity (1.2 wt% Ru/TiO2 , 476 K, H2 /CO = 2.1, 0.6 MPa,
5–60% CO conversion) (a) bed residence time effect (5–60% CO conversion) and (b) ethylene cofeed study (<8% CO conversion) [10].

on how water decreases hydrogenation is not clear. Competitive
adsorption with surface H might be a plausible reason; however,
dissociative adsorption of water should increase surface H concen-
tration.
4.2.2. Water increases the concentration of monomeric carbon
Steady-state isotopic transient kinetic analysis reveals that
water increases the amount of surface monomeric carbon [64]. This
increased surface concentration of monomeric carbon is caused
by an acceleration of the CO dissociation rate without a matching
reactivity increase in the subsequent hydrocarbon synthesis steps.
Based on their proposed model, the rate expression for methane
formation was obtained as shown in Eq (2).
1 is small and therefore there is little variation of the water partial
SCH4 = ∗
 ∗
 (2)
1 + (kg C /kt,1 ) 2 + (kg C /kt,n )
where kg is the rate constant for chain growth, kt,1 and kt,n is the
termination rate constant for C1 and Cn respectively.  C* is the site
coverage of chain growth monomer.
According to this expression methane selectivity will decrease
as  C* increases if the ratios of rate constants within the hydrocar-
bon synthesis network remain constant (Fig. 10).
Whether by inhibition of hydrogenation or by increasing the car-
bon monomer and thereby enhancing chain growth, water seems
to play a major role in controlling methane selectivity at different
CO conversion levels.
Moreover, it has been shown that external added water reduces
methane selectivity and removes the dependence of methane
selectivity on CO conversion [76], suggesting the effect of increasing
the amount of water may level off (Fig. 11). It may be that the con-
tribution of the indigenous water to the total water partial pressure
pressure within this conversion range. The work, on the other hand,

Fig. 10. Methane selectivity as a function of surface carbon coverage (Co–Re catalyst) [64].
 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
Fig. 11. The effect of CO conversion on the selectivity of CH4 and C5+ in the presence of added water () compared with runs without water addition () for 12%Co–0.5%Re/Al2 O3
at 483 K, 20 MPa and H2 /CO = 2.1 [76].
confirms the view that water indeed plays an important role in
reducing methane selectivity.
5. Strategies to reduce methane selectivity
(1) Run the reaction at relatively low temperature and high pres-
sure which still allows reasonable productivity.
(2) Run the reaction at relative high conversion but still lower than
the limit for significant water–gas-shift reaction to occur; oper-
ate at conversion levels where significant catalyst deactivation
should be avoided.
(3) Co-feeding certain amounts of light olefin, optimizing the reac-
tion conditions to minimize hydrogenation and hydrogenolysis
of fed olefin.
(4) Co-feeding certain amount of water which does not cause sig-
nificant catalyst deactivation.
(5) Add suitable promoters, K promotion is widely used for iron
based catalyst; however, for the cobalt based catalyst, K
promotion causes significant catalytic activity decreases. Mn
promotion may both increase catalytic activity and decrease
methane selectivity for cobalt based catalyst.
6. Summary
How methane selectivity changes with CO conversion depends
on the identity of the catalyst and process conditions. The methane
selectivity trend is very similar for cobalt and ruthenium. At a typi-
cal Fischer–Tropsch operation window (2 MPa and H2 /CO = 2), on
cobalt and ruthenium based catalysts, CH4 selectivity normally
decreases with increasing conversion up to about 80%. A further
increase in CO conversion may cause a significant increase in
methane selectivity due to enhanced WGS reaction and/or cata-
lyst deactivation due to high water partial pressure. Fe based FT
catalysts are normally promoted with K, and Fe itself is active for
WGS reaction at high temperatures. Since K is presumed to inhibit
CH4 formation on Fe catalyst, and low H2 /CO ratio feed is used,
the methane selectivity for the Fe based catalyst is comparable to
that of the Ru catalyst but much lower than the cobalt based cat-
alyst even though the reaction temperature for the Fe catalyst is
much higher. Because of these, the change in methane selectivity
with CO conversion for the Fe catalyst mainly depends on both K
and WGS activity. The variation of methane with CO conversion is
usually small, within 2% at CO conversions lower than 60–70% due
to the presence of the K promoter. When CO conversion exceeds
70%, WGS activity becomes high regardless of the amount of K in
the catalysts and the methane selectivity increases with increasing
259
CO conversion due to higher H/CO ratio resulting from the WGS
activity.
Temperature and pressure have a great influence on methane
selectivity for Co and Ru catalysts. Pressure and temperature
change the preference of secondary reactions for primary olefins. At
temperatures greater than 550 K, hydrogenolysis of high molecular
weight hydrocarbons can occur, which can dramatically increase
CH4 selectivity. Thus methane selectivity increases with increas-
ing CO conversion (longer residence time). On the other hand, an
increase in pressure tends to change the extent of reinsertion and
hydrogenation of the olefins. At atmospheric pressure, methane
and C2 –C4 hydrocarbon selectivity does not change with increas-
ing CO conversion. However, at elevated pressure, readsorption
and reinsertion of olefins appears to be enhanced compared to
the hydrogenation reaction, causing the methane selectivity to
decrease with increasing CO conversion.
Possible mechanisms for the methane selectivity trends with
CO conversion were discussed. Primary olefin may compete with
methyl groups for surface sites. Enhanced readsorption of olefin at
longer bed residence time may cause a decrease in methyl interme-
diates and hence result in a decrease of methane formation. Water
plays an important role in determining methane selectivity, in par-
ticular for the Co and Ru catalysts. Different mechanisms regarding
roles of water have been proposed:
(1) Water adsorbs competitively with H and therefore reduces
methane formation.
(2) Water increases the coverage of surface carbon inventory (chain
growth monomer), and therefore increases the chain growth
by adding the carbon monomer relative to methane formation.
Hence, methane decreases with increasing water pressure.
Acknowledgements
The authors are very grateful for financial support from the
Research Council of Norway. CAER researchers would like to thank
the Commonwealth of Kentucky. The authors would also like to
thank Robert O’Brien for providing the data of exit H2 /CO ratio for
Fig. 5b.
References
[1] A. de Klerk, Energy & Environmental Science 4 (2011) 1177–1205.
[2] R.A. Dictor, A.T. Bell, Journal of Catalysis 97 (1986) 121–136.
[3] T.J. Donnelly, C.N. Satterfield, Applied Catalysis 52 (1989) 93–114.
[4] M.E. Dry, in: J.R. Anderson, M. Boudart (Eds.), Catalysis-Science and Technology,
vol. 1, Springer-Verlag, New York, 1981, pp. 160–255.
260 J. Yang et al. / Applied Catalysis A: General 470 (2014) 250–260
[5] R.J. Madon, S.C. Reyes, E. Iglesia, Journal of Physical Chemistry 95 (1991)
7795–7804.
[6] G.C. Bond, Catalysis by Metals, Academic Press, London, 1962.
[7] E. Iglesia, S.C. Reyes, R.J. Madon, S.L. Soled, in: H.P.D.D. Eley, B.W. Paul (Eds.),
Advances in Catalysis, Academic Press, San Diego, 1993, pp. 221–302.
[8] J. Madon Rostam, E. Iglesia, C. Reyes Sebastian, Non-Flory Product Distribu-
tions in Fischer–Tropsch Synthesis Catalyzed by Ruthenium, Cobalt, and Iron,
Selectivity in Catalysis, American Chemical Society, Washington, DC, 1993, pp.
383–396.
[9] T. Komaya, A.T. Bell, Journal of Catalysis 146 (1994) 237–248.
[10] E. Iglesia, S.C. Reyes, R.J. Madon, Journal of Catalysis 129 (1991) 238–256.
[11] C. Aaserud, A.-M. Hilmen, E. Bergene, S. Eric, D. Schanke, A. Holmen, Catalysis
Letters 94 (2004) 171–176.
[12] O. Borg, S. Eri, E.A. Blekkan, S. Storsaeter, H. Wigum, E. Rytter, A. Holmen, Journal
of Catalysis 248 (2007) 89–100.
[13] S. Storsæter, Ø. Borg, E.A. Blekkan, A. Holmen, Journal of Catalysis 231 (2005)
405–419.
[14] W. Ma, G. Jacobs, Y. Ji, T. Bhatelia, D.B. Bukur, S. Khalid, B.H. Davis, Topics in
Catalysis 54 (2011) 757–767.
[15] D.B. Bukur, Z. Pan, W. Ma, G. Jacobs, B.H. Davis, Catalysis Letters 142 (2012)
1382–1387.
[16] F.G. Botes, Industrial & Engineering Chemistry Research 48 (2009) 1859–1865.
[17] F.G. Botes, J.W. Niemantsverdriet, J. van de Loosdrecht, Catalysis Today (2013).
[18] H.A.J. van Dijk, J.H.B.J. Hoebink, J.C. Schouten, Topics in Catalysis 26 (2003)
111–119.
[19] J. Yang, Y. Qi, J. Zhu, Y.-A. Zhu, D. Chen, A. Holmen, Journal of Catalysis (2013),
http://dx.doi.org/10.1016/j.jcat.2013.05.018.
[20] J.P. den Breejen, P.B. Radstake, G.L. Bezemer, J.H. Bitter, V. Frøseth, A. Holmen,
K.P. de Jong, Journal of the American Chemical Society 131 (2009) 7197–7203.
[21] J. Li, G. Jacobs, T. Das, Y. Zhang, B. Davis, Applied Catalysis A: General 236 (2002)
67–76.
[22] J. Li, G. Jacobs, T. Das, B.H. Davis, Applied Catalysis A: General 233 (2002)
255–262.
[23] J. Li, X. Zhan, Y. Zhang, G. Jacobs, T. Das, B.H. Davis, Applied Catalysis A: General
228 (2002) 203–212.
[24] C.H. Mauldin, U.S. Patent, 4,568,663, Exxon Research And Engineering Co., Feb.4
(1986).
[25] M.E. Dry, Catalysis Today 6 (1990) 183–206.
[26] J.C. Hoogendoorn, J.M. Salomon, British Chemical Engineering 2 (1957), pp.
238–244; 308–312; 368–373; 418–419.
[27] A.P. Steynberg, R.L. Espinoza, B. Jager, A.C. Vosloo, Applied Catalysis A: General
186 (1999) 41–54.
[28] D.B. Bukur, S.A. Patel, X. Lang, Applied Catalysis 61 (1990) 329–349.
[29] A.P. Raje, R.J. O’Brien, B.H. Davis, Journal of Catalysis 180 (1998) 36–43.
[30] R.J. O’Brien, B.H. Davis, Catalysis Letters 94 (2004) 1–6.
[31] M.E. Dry, G.J. Oosthuizen, Journal of Catalysis 11 (1968) 18–24.
[32] L. Tau, H.A. Dabbagh, T.P. Wilson, B.H. Davis, Applied Catalysis 56 (1989)
95–106.
[33] B. Graf, H. Schulte, M. Muhler, Journal of Catalysis 276 (2010) 66–75.
[34] W. Ngantsoue-Hoc, Y. Zhang, R.J. O’Brien, M. Luo, B.H. Davis, Applied Catalysis
A: General 236 (2002) 77–89.
[35] W. Ma, G. Jacobs, R.A. Keogh, D.B. Bukur, B.H. Davis, Applied Catalysis A: General
437/438 (2012) 1–9.
[36] A. Dinse, M. Aigner, M. Ulbrich, G.R. Johnson, A.T. Bell, Journal of Catalysis 288
(2012) 104–114.
[37] A. Tavasoli, A. Khodadadi, Y. Mortazavi, K. Sadaghiani, M.G. Ahangari, Fuel
Processing Technology 87 (2006) 641–647.
[38] M.E. Dry, Product R&D 15 (1976) 282–286.
[39] R.C. Everson, E.T. Woodburn, A.R.M. Kirk, Journal of Catalysis 53 (1978)
186–197.
[40] E.W. Kuipers, C. Scheper, J.H. Wilson, I.H. Vinkenburg, H. Oosterbeek, Journal of
Catalysis 158 (1996) 288–300.
[41] A.Y. Khodakov, W. Chu, P. Fongarland, Chemical Reviews 107 (2007)
1692–1744.
[42] L. Bai, H.W. Xiang, Y.W. Li, Y.Z. Han, B. Zhong, Fuel 81 (2002) 1577–1581.
[43] S. Zarubova, S. Rane, J. Yang, Y. Yu, Y. Zhu, D. Chen, A. Holmen, ChemSusChem
4 (2011) 935–942.
[44] H. Schulz, K. Beck, E. Erich, Fuel Processing Technology 18 (1988) 293–304.
[45] L.M. Tau, H.A. Dabbagh, B.H. Davis, Energy Fuels 4 (1990) 94–99.
[46] W. Zimmerman, D. Bukur, S. Ledakowicz, Chemical Engineering Science 47
(1992) 2707–2712.
[47] E.W. Kuipers, I.H. Vinkenburg, H. Oosterbeek, Journal of Catalysis 152 (1995)
137–146.
[48] J. Barrault, C. Forquy, V. Perrichon, Journal of Molecular Catalysis 17 (1982)
195–203.
[49] R. Snel, R.L. Espinoza, C1 Molecular Chemistry 1 (1986) 349.
[50] G. Hutchings, R. Copperthwaite, M. Riet, Topics in Catalysis 2 (1995)
163–172.
[51] E. Iglesia, R. Madon, US Patent 4,754,092 A, Exxon Research And Engineering
Company (1988).
[52] C.J. Kim, US Patent 4,547,525 A, Exxon Research And Engineering Co. (1985).
[53] S.R. Morris, R.B. Moyes, P.B. Wells, R. Whyman, Journal of Catalysis 96 (1985)
23–31.
[54] D.S. Jordan, A.T. Bell, Journal of Physical Chemistry 90 (1986) 4797–4805.
[55] Y. Kobori, H. Yamasaki, S. Naito, T. Onishi, K. Tamaru, Journal of the Chemical
Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases 78
(1982) 1473–1490.
[56] A.A. Adensina, R.R. Hudgins, P.L. Silveston, Applied Catalysis 62 (1990) 295.
[57] L.-M. Tau, H.A. Dabbagh, B.H. Davis, Catalysis Letters 7 (1990) 141.
[58] B. Shi, R.J. O’Brien, S. Bao, B.H. Davis, Journal of Catalysis 199 (2001)
202.
[59] B. Shi, G. Jacobs, D. Sparks, B.H. Davis, Fuel 84 (2005) 1093.
[60] S. Krishnamoorthy, M. Tu, M.P. Ojeda, D. Pinna, E. Iglesia, Journal of Catalysis
211 (2002) 422–433.
[61] K.F. Hanssen, E.A. Blekkan, D. Schanke, A. Holmen, G.F. Froment, K.C.
Waugh, Transient and steady-state studies of the effect of water on cobalt
Fischer–Tropsch catalysts, Studies in Surface Science and Catalysis, Elsevier,
1997, pp. 193–202.
[62] M. Rothaemel, K.F. Hanssen, E.A. Blekkan, D. Schanke, A. Holmen, Catalysis
Today 38 (1997) 79–84.
[63] A.M. Hilmen, D. Schanke, K.F. Hanssen, A. Holmen, Applied Catalysis A: General
186 (1999) 169–188.
[64] C.J. Bertole, C.A. Mims, G. Kiss, Journal of Catalysis 210 (2002) 84–96.
[65] W. Conner, T.K. Das, G. Jacobs, J. Li, B.H. Davis, American Institute of Chemical
Engineers (2003) 2694–2700.
[66] G. Jacobs, T.K. Das, P.M. Patterson, J. Li, L. Sanchez, B.H. Davis, Applied Catalysis
A: General 247 (2003) 335–343.
[67] G. Jacobs, P.M. Patterson, T.K. Das, M. Luo, B.H. Davis, Applied Catalysis A:
General 270 (2004) 65–76.
[68] A.K. Dalai, T.K. Das, K.V. Chaudhari, G. Jacobs, B.H. Davis, Applied Catalysis A:
General 289 (2005) 135–142.
[69] T.K. Das, W.A. Conner, J. Li, G. Jacobs, M.E. Dry, B.H. Davis, Energy Fuels 19 (2005)
1430–1439.
[70] G. Jacobs, T.K. Das, J. Li, M. Luo, P.M. Patterson, B.H. Davis, Preprints/American
Chemical Society, Division of Petroleum Chemistry 50 (2005)
174–177.
[71] Ø. Borg, S. Storsæter, S. Eri, H. Wigum, E. Rytter, A. Holmen, Catalysis Letters
107 (2006) 95–102.
[72] T.K. Das, X. Zhan, J. Li, G. Jacobs, M.E. Dry, B.H. Davis, Studies in Surface Science
and Catalysis 163 (2007) 289–314.
[73] G. Jacobs, T.K. Das, J. Li, M. Luo, P.M. Patterson, B.H. Davis, Studies in Surface
Science and Catalysis 163 (2007) 217–253.
[74] S. Storsæter, Ø. Borg, E.A. Blekkan, B. Tøtdal, A. Holmen, Catalysis Today 100
(2005) 343–347.
[75] W.K. Hall, R.J. Kokes, P.H. Emmett, Journal of the American Chemical Society 82
(1960) 1027–1037.
[76] S. Storsæter, C. Aaserud, A.-M. Hilmen, O.A. Lindvåg, E. Bergene, D. Schanke, S.
Eri, A. Holmen, Fuel Chemistry Division Preprints 47 (2002) 159.