Cr-free Fe based catalysts for high-temperature water-gas shift reactions

Catalysis Today 210 (2013) 2–9

Contents lists available at SciVerse ScienceDirect

Catalysis Today
journal homepage: www.elsevier.com/locate/cattod

The review of Cr-free Fe-based catalysts for high-temperature
water-gas shift reactions
Dae-Won Lee a , Myung Suk Lee a , Joon Yeob Lee a , Seongmin Kim a , Hee-Jun Eom a ,
Dong Ju Moon c , Kwan-Young Lee a,b,∗
a

Department of Chemical and Biological Engineering, Korea University, Anam-dong, Seongbuk-gu, Seoul 136-713, Republic of Korea
Green School, Korea University, Anam-dong, Seongbuk-gu, Seoul 136-713, Republic of Korea
c
Clean Energy Research Center, Korea Institute of Science and Technology (KIST) 39-1, Hawolgok-dong, Seongbuk-gu, Seoul 136-791, Republic of Korea
b

a r t i c l e

i n f o

Article history:
Received 31 August 2012
Received in revised form
20 December 2012
Accepted 21 December 2012
Available online 20 February 2013
Keywords:
Water-gas shift reaction

High-temperature shift
Fe/Cr catalysts: Cr-free catalysts

a b s t r a c t
Since it was patented by Bosch and Wild at 1914, the Fe/Cr-based mixed oxide catalyst has been used for
water-gas shift reactions (WGSRs). Until the present, this catalyst has been used as the primary catalyst
for industrial high-temperature shift (HTS) reactions. However, because environmental concerns about
chromium elements were raised in the early 1980s, the replacement of chromium in HTS catalysts has
been intensely studied by many groups. These studies have contributed notable insights into HTS catalysis
using Fe-based oxides, especially about the reaction mechanism and functions of promoter elements. In
some cases, the potential of using a substituent metal previously neglected because of properties inferior
to those of chromium was rediscovered after noteworthy improvements were produced by combining
it with other metals in promoting the Fe-oxide catalyst. This paper reviews the recent studies of Cr-free
Fe-based HTS catalysts, especially focusing on the roles and functions of the non-chromium promoters
in the catalysts.
© 2013 Elsevier B.V. All rights reserved.

1. Water-gas shift reaction: general considerations
Water-gas shift reaction (WGSR) is a redox-type reaction to convert carbon monoxide and water vapor into carbon dioxide and
hydrogen (Eq. (1)), which was first discovered by an Italian physicist
Felice Fontana in 1780 [1].
CO (g) + H2 O (v) ↔ CO2 (g) + H2 (g) [ H 0 = −41.1 kJ/mol]

(1)

WGSR is now mostly associated with the steam reforming of
hydrocarbons (natural gas, petroleum gas, naphtha, gasoline, coals
and various types of biomass) to produce hydrogen for use in the
synthesis of ammonia and methanol and for the Fischer–Tropsch
process [2–4]. WGSR generates additional hydrogen using gases

remaining after steam reforming, which is generally used to
optimize the H2 /CO molar ratio optimal for the production of
(liquid) hydrocarbons in the Fischer–Tropsch process. In polymerelectrolyte membrane fuel cells (PEMFC) systems, it is used to
remove carbon monoxide, which poisons the electrode catalysts
[5,6]. Through WGSR, the carbon monoxide content is reduced from
10–15% to 0.5–1%, which is then further reduced to trace levels

∗ Corresponding author at: Department of Chemical & Biological Engineering,
Korea University, Anam-dong, Seongbuk-gu, Seoul 136-713, Republic of Korea.
Tel.: +82 2 3290 3299; fax: +82 2 926 6102.
E-mail address: (K.-Y. Lee).
0920-5861/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
/>
(<10–100 ppm) by preferential oxidation (PROX) or methanation
catalysis [7].
WGSR is reversible and moderately exothermic, yielding
41.1 kJ/mol of carbon monoxide. The equilibrium constant for
WGSR depends on the temperature, as expressed by Eq. (2),
which implies that the WGSR products (or H2 /CO ratio) decrease
with increasing temperature [1,8]. An equilibrium CO conversion
(XCO,eq. ) is determined from the equilibrium constant and gas composition using Eq. (3) [1].
Kp =
Kp =

yCO2 ,eq · yH2 ,eq
yCO,eq · yH2 O,eq

= exp

4577.8

− 4.33
T

(yCO2 ,in + yCO,in · XCO,eq )(yH2 ,in + yCO,in · XCO,eq )
[yCO,in (1 − XCO,eq )](yH2 O,in − yCO,in · XCO,eq )

(2)

(3)

yA,eq is the equilibrium mole fraction of component A at T; yA,in is
the mole fraction of component A in the reactant mixture; XCO,eq is
the equilibrium CO conversion at T; T is the reaction temperature
(where the equilibrium state is defined).
Because of this exothermic reversibility, WGSR is typically performed in two stages: a high-temperature shift (HTS, typically
370–400 ◦ C, 10–60 atm) and a low-temperature shift (LTS, ∼200 ◦ C,
10–40 atm) [1–4]. The units are connected via an inter-stage cooler.
HTS is characterized by fast kinetics, but the final CO conversion
is limited by equilibrium. In contrast, LTS undergoes slow kinetics, but the thermodynamic limitation is much less severe than


D.-W. Lee et al. / Catalysis Today 210 (2013) 2–9

that of HTS. By combining HTS and LTS in series and matching
both properly, adjustment of the final gas composition (H2 /CO
ratio) becomes feasible and free from thermodynamic limitations.
Because of the different corresponding reaction conditions and
environments, different catalysts are used in HTS compared to
an LTS process [1–3,9]. In most industrial WGSR processes, Fe/Crbased mixed oxides (Fe/Cr/Cu) and Cu/Zn/Al mixed oxides are used
as the HTS and LTS catalysts, respectively. In research, Co-based catalysts (Co/Mn, Co/Cu, Co/Mo) have attracted interest due to their

sulfur tolerance and high HTS activity [9–11]. Au-supported catalysts have been extensively studied for their high LTS activity
and potential stability in oxidative atmospheres [1,12–14]. PGM
(platinum-group metal) catalysts have recently gained attention
due to their broad applicability, covering both HTS and LTS [1,15].
However, most commercial applications still adopt Fe/Cr/Cu and
Cu/Zn/Al as working catalysts because of their high activity, durability and reasonable manufacturing cost.
Steam to carbon monoxide (S/C) ratio is one of the important
controlled factors to determine the performance of the WGSR. First,
the equilibrium CO conversion increases as the S/C ratio increases,
increasing the final H2 /CO ratio without increasing the catalyst load
or temperature. Steam is a mild oxidant that slows the reduction
of the component (metal) oxides in a WGSR catalyst, which largely
prevents the excessive reduction and activity loss of the catalyst
during the reaction. Steam also slows methanation (reaction formula (4) and (5)), which is an undesirable side reaction of WGSR
because it decreases the hydrogen concentration and causes a loss
of surface area because of its exothermicity.
CO (g) + 3H2 (g) ↔ CH4 (g) + H2 O (v) [ H 0 = −206.3 kJ/mol]
(4)

CO (g) + H2 (g) ↔

1
1
CH4 (g) + CO2 (g) [ H 0 = −247.3 kJ/mol]
2
2
(5)

For the reasons presented above, an appropriate amount of extra
steam is usually added to the gas stream for the WGS (water-gas

shift) reactor.
In addition to the S/C ratio, the reduction factor (or R factor),
defined as the contents of reductive gases (CO and H2 ) relative to
the contents of non-reductive gases (CO2 and H2 O) (Eq. (6)), is used
by several researchers as the measure of the ‘reductive’ nature of
reactant gases in WGSR [16–22].
R=

PCO + PH2
PCO2 + PH2 O

(6)

In practice, the R factor is used to predict whether the reactant
gas causes over-reduction of Fe3 O4 (to FeO or Fe) in a Fe-based HTS
catalyst. It is generally known that the gas does not cause overreduction when R is below 1.2, but over-reduction occurs when R
is above 1.6 [18,19,22].
The WGS reactant gas composition varies widely based on the
choice of feedstock and reforming conditions (S/C ratio, temperature, pressure, w/f). For example, the theoretical composition of
biomass-reformed gas (40% H2 , 44% CO and 16% CO2 ) [23] is quite
different from that of typical steam methane reformate (56.7% H2 ,
10% CO, 6.7% CO2 and 26.7% H2 O) [24]. The variation in the R factor of HTS reactant gases can be observed in Table 1 [18,23,25–33],
2 and 3. In general, the R factor increases from 0.3–1.0 to 1.0–1.6
as the syn gas H2 content increases when using natural gas as the
feedstock for steam reforming (CH4 has the highest H/C ratio among
hydrocarbons), and the addition of extra steam is required to prevent over-reduction of Fe3 O4 by decreasing R factor of the reformed

3

Table 1

R factors adopted in various studies.
Catalyst

R factor (S/C ratioa )

Reference

Ni
Cu/Mo/Fe
Cr/Fe
Cu/Th/Fe
Cr/Fe
Cu/Al/Fe
Cu/Al/Fe
Cu/Al/Fe
Cu/Al/Fe
Ni/Al/Fe
V/Fe
Cu/Al/Fe

0.65 (5.00)
0.81 (2.00)
0.83 (6.00)
0.86 (6.00)
1.00 (6.00)
1.00 (1.00)
1.17 (1.00)
1.17 (1.00)
0.70 (2.00)
1.58 (2.50)

1.00 (6.00)
1.00 (6.00)

[25]
[23]
[18]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[19]
[33]

a

1.17 (1.00)
2.33 (2.00)
1.40 (4.00)

2.33 (2.00)

Steam/CO ratio.

gas. However, lower extra-steam addition is desirable to reduce
operational costs [20]. The amount of extra-steam should be carefully determined, because the performance of a WGS catalyst would
change if applied under different R factors.
2. Fe/Cr-based HTS catalysts

2.1. General background for Fe/Cr-based HTS catalysts
Iron–chromium (ferrochrome, Fe/Cr) oxide was first patented as
a WGS catalyst by Bosch and Wild in 1914 [34]. Although nearly
a century has passed, Fe/Cr is still utilized as a primary catalyst
for HTS because of its reasonable activity and durability in most
applications.
The pre-activated catalyst is generally composed of 87–95%
Fe2 O3 (ferric oxide, mineral name: hematite), 5–10% Cr2 O3
(chromium (III) oxide) and miscellaneous other components, such
as CuO, Co2 O3 and/or MgO [3,9,35,36]. The Fe/Cr catalyst is usually synthesized through the base-catalyzed co-precipitation of
Fe2 (SO4 )3 and Cr2 (SO4 )3 using Na2 CO3 [35]. The residual sulfate
ions should be carefully washed to avoid producing hydrogen sulfide during activation (pre-reduction) and reaction, which poisons
the LTS catalyst downstream of the HTS catalyst bed [35]. After
calcination, the major catalyst phase is a Fe2 O3 –Cr2 O3 mixture, in
which Cr2 O3 is occasionally incorporated into the ␣-Fe2 O3 lattice
[9,35]. The calcined catalyst should be pre-reduced before its use
in the reaction, through which Fe2 O3 is turned into its catalytically active Fe3 O4 (ferric ferrous oxide, mineral name: magnetite)
phase. The pre-reduction is usually performed at 315–460 ◦ C using
reactant gas (syn gas) [3]; however, to prevent over-reduction into
FeO or metallic Fe, the R factor of the reactant gas is adjusted to
approximately 1.0 by adding extra steam [18,20].
In plant operation, the commercial Fe/Cr catalysts generally
decrease the carbon monoxide output from 10% to15% in feed flow
to 2–3% [2,35]. The activity of the Fe/Cr catalyst is improved by the
addition of a promoter element, such as Cu, decreasing the activation energy [37] and increasing selectivity (i.e., inhibiting the
methanation [35]) of the catalysis. Selected test results for Fe/Cr
catalysts are listed in Table 2 [38–42].
The WGSR mechanism of Fe/Cr catalysts is typically understood
as a redox-type mechanism, which has been interpreted in two
ways (Fig. 1): (i) the regenerative (Rideal–Eley type) and (ii) the

associative (Langmuir–Hinshelwood type) mechanisms [3,9]. The
former is often perceived as more suitable for Fe/Cr catalysts [3].
The regenerative mechanism is usually facilitated by the exchange
of electrons between Fe2+ and Fe3+ in the octahedral site of magnetite during WGS catalysis [29,43].
The Fe/Cr catalyst is used in a process for an average of 3–5
years without exchanging with fresh catalyst [3,44]. The activity
decrease is mostly due to the thermal sintering of the magnetite


4

D.-W. Lee et al. / Catalysis Today 210 (2013) 2–9

Table 2
HTS performances of Fe/Cr and Fe/Cr/Cu catalysts.
Catalyst
(mol ratio)

S/C ratioa
S/G ratiob

R factor

GHSVc (h−1 )

Press. (bar)

Temp: CO Conv.
[Equil. CO Conv.]

Reference

Fe/Cr
(91/9)

S/C = 3.5
S/G = 0.4

1.4

60,000

1

400 ◦ C: 68% [97%]
450 ◦ C: 75% [95%]

[38]

Fe/Cr
(91/9)

S/C = 3.5
S/G = 3.5

0.3d

60,000

1

400 ◦ C: 65% [97%]
450 ◦ C: 77% [95%]

[39]

Fe/Cr
(93/7)

S/C = 4.8
S/G = 0.7

0.3d

3000

20

360 ◦ C: 89% [89%]

[40]

Fe/Cr/Cu
(89/9/2)

S/C = 3.5
S/G = 3.5

0.3d

60,000

1

400 ◦ C: 79% [97%]
450 ◦ C: 82% [95%]

[38]

Fe/Cr/Cu
(90/8/2)

S/C = 2.0
S/G = 1.0d

0.9d

10,000

10

380 ◦ C: 79% [87%]

[41]

Fe/Cr/Cu
(89/9/2)

S/C = 3.5
S/G = 3.5

0.3d

60,000

1

400 ◦ C: 79% [97%]
450 ◦ C: 83% [95%]

[42]

a
b
c
d

Steam/CO ratio.
Steam/gas ratio.
“Wet gas” base.
Estimated from the data given in the paper.

phase, but in plant operation, increasing the reaction temperature
compensates for this decrease [45]. Additionally, the Fe/Cr catalyst
is seldom deactivated by sulfur poisoning, unlike LTS catalysts (e.g.,
Cu/Zn/Al) [45].
2.2. Catalyst structure of Fe/Cr catalyst: role of Cr
For many years, the chromium in the Fe/Cr HTS catalysts has
been predominantly recognized as a stabilizer to prevent the thermal sintering of Fe3 O4 and loss of surface area of the catalyst.
In a fresh state, Fe3 O4 /Cr2 O3 has a much higher specific surface

area (40 m2 /g for Fe3 O4 /8 wt.% Cr2 O3 ) than un-promoted Fe3 O4
(8 m2 /g) [46]. Fe3 O4 /Cr2 O3 also exhibits much slower magnetite
sintering than un-promoted Fe3 O4 [9]. Chinchen et al. insisted
that as the reaction progresses, discrete Cr2 O3 grains grow and
become dispersed over Fe3 O4 domains, thereby blocking the thermal agglomeration of Fe3 O4 particles [47,48].
A different opinion is that Cr3+ ions enter into the inverse-spinel
lattice of Fe3 O4 and form a solid solution. Robbins et al. found that
Cr3+ ions dissolve into the Fe3 O4 lattice and occupy the octahedral site and the displaced Fe2+ and Fe3+ ions (from the octahedral
sites) are transferred to the tetrahedral sites [49]. Edwards et al.

claimed that the dissolved Cr3+ is enriched at the surface region of
the catalyst and that the Cr-enriched surface shell, being more thermodynamically stable than the Fe-rich core, reduces ion diffusion
and sintering effects [50]. Natesakhawat et al. reported that the
Cr3+ in Fe/Cr was oxidized to Cr6+ during WGS catalysis [28]. The
Cr3+ ↔ Cr6+ oxidation–reduction cycle was expected to enhance
the redox rate of magnetite and promote the WGS activity of the
catalyst.
Today, it is generally understood that chromium acts as both a
textural (preventing thermal sintering) and functional (enhancing
redox efficiency) promoter in Fe/Cr HTS catalysts.
2.3. The necessity of substituting chromium with other elements
Hexavalent chromium (found in chemical compounds containing Cr+6 ) is a strong carcinogen, threatening human health and the
environment [51]. Exposure through inhalation and drinking water
causes cancer and serious damage to human organs and skin. In
contrast, trivalent chromium (Cr3+ ) has very low toxicity and is a
nutrient for the human body [51].
Concerns about the environmental hazard and toxicity of hexavalent chromium have been raised since the early 1920s. Since

Fig. 1. Redox-type WGSR mechanisms over magnetite-based catalysts.

D.-W. Lee et al. / Catalysis Today 210 (2013) 2–9

5

Table 3
HTS performances of Cr-free Fe-based catalysts.
Catalyst
(mol ratio)

S/C ratioa
S/G ratiob

R factor

GHSVc
[WHSV]

Press. (bar)

Temp.: CO Conv.
[Equil. CO Conv.]

Reference

Fe/Al/Cu/Ce
(89/8/2/1)d

S/C = 10.0
S/G = 1.0

1.0d

7000 h−1

20

350 ◦ C: 92%d [95%]

[61]

Fe/Al/Ce
(96.3/3.4/0.3)

S/C = 3.8
S/G = 1.0

0.6d

6000 h−1 d
(3000 h−1 , dry gas base)

1

350 ◦ C:
53% (after aging) [95%]

[62]

Fe/Al

(91/9)

S/C = 1.0
S/G = 0.1

1.0d

[0.060 m3 /gcat /h]

1

400 ◦ C: 25% [77%]

[28]

Fe/Al/Cue
(87/9/4)

S/C = 1.0
S/G = 0.1

1.0d

[0.060 m3 /gcat /h]

1

400 ◦ C: 46% [77%]

[28]

Fe/Al/Cuf
(77/8/15)

S/C = 1.0
S/G = 0.1

1.2d

[0.060 m3 /gcat /h]

1

400 ◦ C: 57% [63%]

[30]

Fe/Ni
(80/20)

S/C = 3.7
S/G = 0.6

1.2

[0.060 m3 /gcat /h]

1

400 ◦ C: 70% [78%]

[20]

Fe/Ni
(80/20)

S/C = 2.6
S/G = 0.3

2.0

[0.025 m3 /gcat /h]

1

400 ◦ C: 50% [65%]
Selectivity > 89.2%

[21]

Fe/Ni
(67/33)

S/C = 2.6
S/G = 0.3

2.0

[0.025 m3 /gcat /h]

1

400 ◦ C: 64% [65%]
Selectivity > 88.8%

[21]

Fe/Ni/Cs
(66/31/3)

S/C = 2.6
S/G = 0.3

2.0

[0.075 m3 /gcat /h]

1

400 ◦ C: 61% [65%]
Selectivity = 100%

[66]

Fe/Ni/Zn
(64/31/5)

S/C = 2.6
S/G = 0.3

2.0

[0.030 m3 /gcat /h]

1

400 ◦ C: 65% [65%]
Selectivity > 97%

[67]

Fe/Ni/Alg
(37/22/41)

S/C = 3.0
S/G = 0.4

1.4d

14,500 h−1 d
(10,000 h−1 , dry gas base)

1

400 ◦ C: 54% [57%]

[32]

Fe–Ni/Ce–Zr
(11–10/53–26)d

S/C = 3.7
S/G = 0.6

1.1d

15,600 h−1 d
(10,000 h−1 , dry gas base)

1

400 ◦ C: 75% [70%]
Slight methanation

[68]

a
b
c
d
e
f
g

Steam/CO ratio.
Steam/gas ratio.
“Wet gas” base.
Estimated from the data given in the paper.
Prepared with co-precipitation.
Prepared with sol–gel method.

Prepared by solution-spray plasma method.

the U.S. National Research Council published the general guidelines for chromium compound risk assessments in 1983, the EPA
has published many practical guidelines for the identification and
assessment of hexavalent chromium [52]. The Occupational Health
and Safety Administration (OSHA) under the U.S. Department of
Labor enforced strict regulations regarding worker exposure to
hexavalent chromium in several industries [53]. In Europe, the
recently published European Restriction of Hazardous Substances
(RoHS) banned the use of six hazardous materials, including hexavalent chromium, in all electronic–electrical equipment [54]. It is
only a matter of time before these regulations are expanded to cover
entire industries.
Returning to the Fe/Cr HTS catalyst, the chromium species in
a fresh Fe/Cr catalyst is usually Cr+3 (Cr2 O3 ), which is much less
toxic than Cr+6 . The Cr+6 content is low, but workers must take
precautions when handling the catalyst throughout the span of
the operation. Moreover, Cr+6 is water-soluble and is leached from
the catalyst by condensed steam or cold water, which could be a
threat to the environment, even with minimal disposal [51]. There
are several possibilities for producing hexavalent chromium during the manufacturing of the catalyst. For instance, some of the
Cr+3 ions that were not precipitated can be oxidized into Cr+6 when
the catalyst is calcined at high temperature with the mineral base
(Na+ ) present in the precipitates [51]. The motivation for replacing
chromium with other elements mostly lies in these environmental
and health concerns [55]. In addition to WGSR, the issue of replacing chromium is being similarly discussed in the development of
catalysts for fatty alcohol production (the hydrogenation of fatty
esters) [56,57].

3. Studies on Cr-free Fe-based HTS catalysts
Note: The HTS activities of Cr-free and Cr-containing catalysts

are listed in Tables 2 and 3. It was not straightforward to compare
a catalyst with another in activity, because each catalyst had been
tested in a different reaction condition (S/C ratio, R factor, w/f, temperature, etc.). So the activities were listed in a table format with
full reaction conditions provided.
3.1. Early studies: 1980–1990
Chinchen first tried using a Cr-free Fe-based HTS catalyst to
replace chromium with an element capable of forming a spinel
structure in iron oxide without unduly diluting the catalytic activity [58]. The chrome replacements were chosen according to ionic
size and oxidation number [28,58]. Among the candidates, Ca, Ce
and Zr were capable of forming spinel structures with Fe. Fe/Ce and
Fe/Zr showed higher surface areas than the commercial Fe/Cr catalysts, but their specific activities (activity per total catalyst weight)
were lower than those of commercial catalysts. A similar attempt
by Rethwisch and Dumesic was also unsuccessful. They tried using
Zn(II) and Mg to replace Cr, but the activities of Fe/Zn and Fe/Mg
were 30 times lower than that of magnetite [59]. It was argued
that the element (Zn or Mg) displaced all of the Fe2+ ions from
the octahedral site of inverse spinel lattice, preventing Fe2+ /Fe3+
redox transfers, which are the driving force of the regenerative
mechanism for magnetite catalysis.
Rethwisch et al. dispersed un-promoted magnetite over a
graphite support (Fe3 O4 /C) to enhance the catalytic turnover rate


6

D.-W. Lee et al. / Catalysis Today 210 (2013) 2–9

and decelerate the thermal agglomeration of magnetite [60]. The
catalyst initially showed high activity, which then decreased after
a few hours. The magnetite agglomerated during the reaction, but

the authors argued that the decline in activity was due to the constriction of pores within magnetite clusters, which was driven by
the surface hydroxylation of magnetite under wet atmospheres.
The activity recovered to some degree after the catalyst was dehydroxylated under a dry carbon monoxide atmosphere.
3.2. Fe/Al-based catalysts: Al/Ce or Al/Cu as replacements for Cr:
1995–2010
The first promising results for developing Cr-free Fe-based
catalysts were arguably those of Ladebeck and Kochloefl [61],
who replaced Cr in a Fe/Cu/Cr catalyst with Al/Ce. The resulting
Fe/Cu/Al/Ce catalyst showed activity superior to that of a commercial catalyst. Since then, Al has been studied more intensively than
any other element as a replacement for chromium in Fe-based HTS
catalysts. From now, it will be reviewed in its proper chronological
order.
Araújo and Rangel proved that the activity promoted by Al
becomes more prominent when Cu is included in the magnetite
texture [33]. Under low steam-to-gas ratios (S/G = 0.4; estimated
R factor = 0.9), the Fe/Al/Cu catalyst was similar in HTS activity
but better in selectivity (i.e., methanation suppression) compared
to a commercial Cr-containing catalyst. The authors argued that
Al/Cu promotes the formation of the magnetite phase during prereduction and stabilizes the phase against further reduction. They
deemed that Cu acted as a textural promoter rather than a functional promoter (creating or promoting activity by affecting the
electronic properties of the major active species [23]). However,
the thermal stability of the Fe/Al/Cu catalyst was not within the
scope of the study.
Liu et al. studied an Al/Ce-promoted Fe catalyst (Fe/Al/Ce), which
was given a proprietary name, NBC-1 [62,63]. The basic idea was
to adopt ␥-Fe2 O3 (maghemite) as the backbone of the catalyst,
which was thought to be more effective than ␣-Fe2 O3 in incorporating promoter elements, by utilizing the vacant sites of an
imperfect spinel structure [64]. From the context, it is inferred
that the authors regarded Al and Ce both as textural promoters
for the magnetite phase. The authors claimed that the catalyst was

active and thermo-resistant, by showing that it was comparable to
a commercial Fe/Cr catalyst in both HTS activity and specific surface area measured after high-temperature aging (530 ◦ C, 15 h). The
achievement would have been more promising if the catalyst had
been tested under more reducible conditions. (The R factor of the
reaction gas is estimated to be 0.6, which is much lower than the
conventional value, 1.0.)
Regarding the Fe/Al/Cu HTS catalyst, the Ozkan group have
published several noteworthy studies in the last decade [28–31],
providing a systematic understanding of the catalyst.
Using in situ XRD and TPR studies, the group proved that Al
played a role similar to that of Cr, inhibiting the thermal growth of
the magnetite phase and stabilizing the phase against further reduction to FeO and Fe [28]. However, the XPS studies indicated that Al
did not promote the redox rate of the iron oxide because there was
no change in its oxidation state (Al+3 ) during WGS catalysis. Unlike
Al, Cr changed its oxidation state (+3 ↔ +6) during catalysis, which
was thought to allow it to act as a functional promoter in promoting
the WGS activity of the Fe catalyst. Similarly to the results of Araújo
and Rangel’s study [33], the HTS activity was greatly enhanced
when Al and Cu were included in the catalyst together as Fe promoters (Table 3). Cu was considered a functional promoter for the Fe/Al
catalyst because TPR analysis indicated that Cu greatly enhanced
the reducibility of the iron oxide [28]. The authors claimed that
there are two ways for Cu to participate in the catalysis: (1) the Cu

Fig. 2. BET surface areas of Fe/Ni catalysts. FNxxyy denotes a Fe/Ni catalyst with Fe
and Ni in xx and yy wt.%, respectively; “Aged” implies that the catalyst was aged
with reaction for 3 h under 400 ◦ C, H2 (56.7%), CO (10%), CO2 (6.7%), H2 O (26.7%) and
WGSV = 0.025 m3 /gcat /h.

species serves as an electronic (functional) promoter and promotes
the redox rate of the catalysis and (2) the excluded Cu species is

present on the catalyst surface, which is reduced to metallic Cu during the reaction and provides additional active sites, similar to Cu
in the Cu/Zn/Al catalyst in LTS reactions. However, because metallic
Cu is very prone to thermal sintering, it is desirable to incorporate
all Cu species into an iron oxide structure and form a perfect solid
solution [29]. The sol–gel preparation of the catalyst fulfilled this
objective. In the author’s following papers [29,30], it was proven
that Cu was uniformly distributed over the iron oxide matrix when
the catalyst was prepared using the sol–gel method at pH 9 with
iron acetylacetonate as the Fe precursor (others in nitrates), and
C2 H5 OH/NaOH was used as the solvent/precipitant [29]. Such uniformity was not obtained using the conventional co-precipitation
method. The sol–gel method is more advantageous in that it allows
the formation of ␥-Fe2 O3 to be induced by adjusting the Fe2+ /Fe3+
ratio in the precursor solution and the aging time. The TPR and XPS
measurements confirmed that ␥-Fe2 O3 helps incorporate the promoter elements (Al and Cu) into the iron oxide structure to form
a uniform solid solution [29]. The authors further improved the
preparation method using propylene oxide as the gelation agent,
which improves the HTS activity and stability for the Fe/Al/Cu catalyst [31].
3.3. Fe/Ni-based catalysts: HTS catalysis of Fe/Ni/Zn and Fe/Ni/Cs
under high-R-factor conditions: 2009–2011
Ni has been perceived as unsuitable for use as a component of
WGS catalysts because it is easily reduced under WGS conditions
and manifests high methanation activity [65].
However, our research group has found that Ni is capable of
forming a solid solution with iron oxide, producing a Fe/Ni catalyst
that exhibits reasonable HTS activity even under highly reducible
conditions (R factor = 2) if promoted by another appropriate element [20,21,66,67]. Ni/Fe catalysts were prepared by conventional
co-precipitation, which produced inverse spinel NiFeO4 after calcination in air at 500 ◦ C. The inclusion of Ni increased the surface
area of the fresh catalyst, but the effect was drastically diminished
when the catalyst is aged in HTS reaction (Fig. 2), so it is technically improper to refer to Ni as a textural promoter. Ni can instead
be referred to as a functional promoter: under a high R factor of 2,

the Ni/Fe (67/33 in mol% (Table 3) or 66/34 in wt.% [21]) catalyst
showed high initial CO conversion (64%), close to the equilibrium
value (65%), whereas the commercial Fe/Cu/Cr catalyst showed
only 50% conversion [21]. TPR measurements confirmed the


D.-W. Lee et al. / Catalysis Today 210 (2013) 2–9

Fig. 3. HTS activities of Cs-promoted Fe/Ni catalysts; FN: Fe/Ni (66/34 in wt.%),
xCsFN: x wt.%-Cs impregnated Fe/Ni (66/34 in wt.%); H2 (56.7%), CO (10%), CO2 (6.7%)
and H2 O (26.6%), R factor = 2; 400 ◦ C; WHSV = 0.075 m3 /gcat /h [66].

Ni-enhanced redox rate of iron oxides [67]. Even with a high R factor (R = 2) and at high temperature (400 ◦ C), the catalyst managed
to maintain its initial activity for over 11 h. However, part of the catalyst was reduced to FeNi3 (awaruite) during the reaction, which
is attributed at least in part to the methanation side reaction [21].
Methane was produced from both CO and CO2 [21], from which the
selectivity of the HTS reaction over the Fe/Ni catalyst is estimated
as 85–90% (Table 3).
The problem of low selectivity, that is, the occurrence of methanation, was overcome by promoting the Fe/Ni catalyst with cesium
[66] or zinc [67].
By impregnating Cs on the Fe/Ni catalyst, the HTS activity was
greatly enhanced and the methanation was effectively restrained
(Fig. 3) [66]. The catalysts were tested under a weight hour space
velocity (WHSV = 0.075 m3 /gcat /h) three times higher than that
used in the previous study (Table 3). Because of the increase in
WHSV, the CO conversion of un-promoted Fe/Ni (NF, in Fig. 3) was
almost halved to 32%. Under such adverse conditions, the Fe/Ni/Cs
catalysts (3.9CsNF and 6.0CsNF, in Fig. 3) showed near-equilibrium
CO conversion (63% and 61%) with almost 100% selectivity. Based
on CO2 -TPD analysis, the improvement of the catalytic performance

was attributed to the increase in the number of weakly basic sites
by Cs promotion, on which the formate-intermediated associative
mechanism was thought to progress.
Zn promotion also enhanced the HTS activity of the Fe/Ni catalyst [67]. Zn was co-precipitated with Fe and Ni to form a solid
solution of (Zn,Ni)Fe2 O4 inverse spinel species. The Zn-promoted
Fe/Ni (Fe/Ni/Zn) showed near-equilibrium CO conversion with
excellent methanation restraint (selectivity over 98%, Table 3),
which was similar to the previous Cs-promoted Ni/Fe catalyst. The
catalyst showed very stable performance, maintaining its activity
over 15 h. However, Zn promotion is thought to be inferior to Cs
promotion in terms of activity enhancement because such a level
of activity was obtained under a WHSV of 0.035 m3 /gcat /h, whereas
Fe/Ni/Cs achieved a similar level of activity under a nearly doubled
WHSV of 0.075 m3 /gcat /h (Table 3). Zn performs the role of functional promoter for the Fe/Ni catalyst very well: first, Zn prevents
the reduction or disintegration of the inverse spinel phase during
reaction. Through time-dependent XRD analysis, it was found that
FeNi3 was formed from the disintegration of an unstable, incomplete layer of zinc–nickel ferrite near the catalyst surface. When
the unstable layer was used up, the core crystal of (Zn,Ni)Fe2 O4 was
intact, and the reduced phase (FeNi3 ) did not grow further throughout the rest of reaction [67]. Second, Zn enhanced the reducibility
of the catalyst, promoting CO oxidation with lattice oxygen, which

7

leads to an increase in the WGS rate and selectivity. The improved
reducibility of Fe/Ni/Zn was confirmed by H2 -TPR and CO-TGA measurements [67].
Watanabe et al. showed that combining Fe/Ni with Al resulted
in excellent HTS activity without significant methanation [32]
(Table 3). The authors prepared the Fe/Ni/Al catalyst with the
solution-spray plasma technique to produce Fe/Ni species welldispersed on the hollow Al2 O3 sphere. During the reaction, the
Fe/Ni species were partially reduced to Ni–Fe alloy (FeNi3 ), which

the authors noted was a crucial species in suppressing hydrogen
adsorption and CO methanation. The catalyst showed the best performance in terms of HTS activity and methanation suppression
when the Fe/(Fe + Ni) atomic ratio was between 0.5 and 0.8.
The authors developed this idea into dispersing Fe/Ni species
on the mesoporous CeO2 –ZrO2 support prepared by the “hardtemplate method” using KIT-6 as a template material [68]. The
purpose of this study was also to minimize methanation over
the Fe/Ni species. The basic ideas were, first, to improve Ni (or
Fe–Ni) dispersion using a support with a large specific surface area,
and second, to use a reducible oxide support that improves the
transfer rate of lattice oxygen (in order to suppress methanation
and increase the selectivity). Both requirements were simultaneously satisfied by impregnating Fe/Ni species on the mesoporous
CeO2 –ZrO2 prepared by the hard template method. The catalyst
showed improved thermal stability, HTS activity and methanation
suppression compared to the catalyst prepared using conventional,
co-precipitated CeO2 –ZrO2 support. In particular, the formation
of FeNi3 in a highly dispersed state over Fe–Ni/CeO2 –ZrO2 (hard
template) led to a more effective suppression of methanation.
3.4. Other noteworthy studies: 1998–2011
Costa et al. studied the use of Th as a replacement for Cr in
˚
Fe/Cr/Cu catalysts [26]. Because the ionic radius of Th4+ (0.94 A)
˚ Th4+ was not incoris considerably larger than that of Fe3+ (0.69 A),
porated into the iron oxide matrix; instead, it went to the surface,
forming a segregated phase. However, the presence of Th resulted
in the formation of smaller iron oxide particles and hindered the
thermal sintering of the particles. In addition, although Th was
present at the surface, it stabilized the magnetite phase against
deeper reduction. Hence, Th can be categorized as a textural promoter for Fe-based HTS catalysts. Except for its presence on the
surface, Th is almost identical to Al in its characteristics and role as
a promoter. Like Al, its activity is also largely enhanced by the use

of Cu as a co-promoter. The author claimed that the Fe/Th/Cu catalyst is more active than a commercial Fe/Cr/Cu catalyst at 370 ◦ C,
S/G = 0.6 (S/C = 6) and R factor = 0.8.
Júnior et al. tried using V(IV) (vanadium) as a chrome replacement [19]. In this study, vanadium-doped magnetite was prepared
by heating sol–gel-prepared, iron (III)–vanadium (IV) hydroxoacetate under nitrogen. Because magnetite was produced directly
with this method, pre-reduction was not needed when using this
catalyst in the HTS reaction. Vanadium was located mainly on the
surface as V2+ and V5+ species. There is some doubt about the claim
that vanadium acted as a textural promoter for the magnetite phase
because the specific surface area of V-doped magnetite was already
small (25–28 m2 /g) in the fresh state and the difference from that of
un-doped magnetite was marginal. However, the vanadium stabilized Fe3+ and increased the activity and selectivity of the magnetite
phase, implying that it acts as a functional promoter.
Martos et al. studied Fe/Mo(VI)/Cu as a Cr-free HTS catalyst, using
the oxidation-reduction method to prepare the catalyst [23]. Mo6+
was incorporated perfectly into the magnetite structure due to its
˚ compared to Fe3+ (0.69 A).
˚ Although
smaller ionic radius (0.62 A)
the catalyst was prepared without thermal calcination, the specific surface area of Fe/Mo was quite small (32 m2 /g). However, a


8

D.-W. Lee et al. / Catalysis Today 210 (2013) 2–9

linear relationship was found between the Mo content and BET
area, which implies that Mo is a textural promoter for the magnetite phase. However, the activity enhancement and stabilization
of the magnetite phase were obtained when Mo was paired with
Cu, which was very similar to the cases of Fe/Al/Cu and Fe/Th/Cu
described previously.

Boudjemaa et al. [69] examined the influence of acid-base
properties on Cr-free Fe-based catalysts in HTS reaction using
in situ DRIFT measurements as a major analytical tool. Hematite(later in reaction, magnetite-) supported SiO2 , TiO2 and MgO
were used as the catalysts, and the order of activity was related
to the basicity of the materials (measured by the activity in
isopropanol dehydrogenation): Fe2 O3 /MgO Fe2 O3 /TiO2 > Fe2 O3
(not supported) Fe2 O3 /SiO2 . The high activity of Fe2 O3 /MgO was
explained in terms of a formate-intermediated associate mechanism, in which the carbonyl species adsorbed on Fe reacts with
hydroxyl groups on the Fe–MgO interface to produce formate intermediates. It was proposed that the decomposition rate of formate
species governs the overall reaction rate, which is facilitated by the
weaker metal-oxygen bond in the basic oxides.
Mahadevaiah et al. incorporated Fe into the CeO2 crystalline
network, and the resultant catalyst exhibited impressive performances under LTS and HTS conditions [70]. It is generally accepted
that CeO2 is a good redox material for the regenerative mechanism in WGS catalysis. In CeO2 , the lattice oxygen easily interacts
with adsorbed CO (turning into CO2 ), and the depleted oxygen site
is restored with the release of H2 from H2 O. The WGS activity is
further augmented if CO adsorption is promoted by another element. The noble metals (Pt, Pd, Rh) are usually chosen for such
purpose [71], but the authors used Fe instead, partially substituting the CeO2 with Fe to form a solid solution of Ce1−x Fex O2−ı .
As a result, Fe not only promoted CO adsorption in WGS catalysis but also enhanced the oxygen storage capacity by synergetic
redox interaction between Ce4+ /Ce3+ and Fe3+ /Fe2+ [70]. Nearequilibrium CO conversion was achieved using Ce0.67 Fe0.33 O1.835
at temperatures above 450 ◦ C, and the activity was expanded to
the LTS region (∼285 ◦ C) when Pt was co-doped inside the catalyst
(Ce0.67 Fe0.33 Pt0.02 O1.785 ).
Another example of fixing Fe onto a non-magnetite structure was
suggested by the work by Sun et al., in which the perovskite structure was utilized as a catalyst matrix for Fe [72]. The basic idea
was as follows: in the LaFeO3 perovskite structure, La3+ lowers the
binding energy of oxygen in FeO6 octahedra, promoting the transfer rate of lattice oxygen (␣-oxygen) to the adsorbed CO (possibly
on Fe) to enhance the WGS rate in a regenerative (redox) mechanism. In addition, the authors adopted a general method to increase
the redox property of the perovskite catalyst, substituting another
cation (Ce4+ ) for the cations in LaFeO3 . Because of the restriction

in ionic radius, Ce4+ is incorporated exclusively into the A-site (i.e.,
La3+ ), making the perovskite structure non-stoichiometric, which
is higher in oxygen storage capacity and thermal stability than
stoichiometric LaFeO3 . Hence, the non-stoichiometric La0.9−x Cex FeO3
catalyst is more active in HTS reactions than the LaFeO3 catalyst.
In addition, its activity at temperatures above 550 ◦ C was higher
than those of commercial Fe/Cr/Cu catalysts operating at 450 ◦ C.
The accommodation of Ce4+ in this catalyst was limited to low values (x = 0.2; above 0.2, the perovskite became unstable); however,
a small amount of CeO2 was always found in the catalysts. It was
claimed that the segregated CeO2 phase also enhances the WGS
activity via its intrinsic redox property.
4. Summary and future perspectives
We have summarized the results of previous studies of Cr-free
Fe-based HTS catalysts, especially those addressing the promoters
used to replace Cr in Fe-based HTS catalysts.

(1) First, a promoter element should form a solid solution with
iron oxides or at least be located in the surface layer in a
well-dispersed state. In the Fe-based WGS catalysts, the role
of promoter is divided into textural and functional roles. A
textural promoter, whether it exists as individual crystallite or
fuses into iron oxide lattices, enhances the catalyst microstructure (surface area, porosity, grain size) and behaves as a barrier
for thermal growth of iron oxide crystallites (i.e., thermal sintering). A functional promoter enhances the redox rate of the
catalyst with its own redox activity or by facilitating the redox
cycle of iron oxide. Regardless of role, it is more desirable for
a promoter to form a homogeneous solid solution with iron
oxides.
(2) Chromium in a commercial Fe/Cr or Fe/Cr/Cu HTS catalyst
acts as a textural and functional promoter. Among the chrome
replacement promoters, Al and Th have functionalities as textural promoters, preventing thermal agglomeration and excessive

reduction of the magnetite phase. Ce and Cu function as functional promoters for the magnetite or ‘promoted’ magnetite (e.g.,
Fe/Cr, Fe/Al, Fe/Th) phases, improving the redox properties of
active species, which in turn increases the intrinsic WGS activity
of the catalyst.
(3) To date, there is no known single elemental promoter that
plays a dual role (textural and functional) with a promoting
functionality comparable to that of chromium. In developing
chrome replacement promoters, the most effective strategy
is to combine more than two non-chromium elements (usually one textural and one functional), in which the matching
between elements is very crucial. For instance, Zn is not a proper
promoter for iron oxide for itself; it increases the specific surface area but does not increase the WGS activity [23]. However,
Zn plays a prominent role as a functional promoter if paired
with Ni in promoting iron oxide, which not only improves the
WGS activity but also hinders methanation [67].
(4) In general, magnetite-based WGS catalysis follows a regenerative mechanism. Hence, it is used to produce a good result
when dispersing magnetite (or promoted magnetite species) over
reducible oxide support with superior lattice oxygen mobility. Cerium oxides and their derivatives can be used for such
a purpose.
(5) The use of base promoters, such as alkaline or alkaline-earth
metals, promotes the formate-intermediated associative mechanism in WGS catalysis. This approach could exhibit synergy
when combined with other functional/textural promoters for
promoting iron oxide catalysts.
(6) The immobilization of Fe into a crystalline matrix is sometimes
effective in stabilizing the active species (FeOx ) against thermal sintering or excessive reduction. The use of cerium-based
perovskite is a good example.

Presently, the development of Cr-free HTS catalysts is an important, ongoing topic in the catalyst industry. We have introduced
some studies about Cr-free “Fe-based” HTS catalysts, but the major
developmental trend involves noble metal catalysts, which exhibit
high activity in a compact catalyst bed and maintain this activity even in oxidative atmospheres [55]. These catalysts are quite

promising in specific fields, such as automobile applications. However, Fe-based catalysts are still advantageous in terms of material
cost, which is highly desirable for reducing the operation costs
of hydrogen plants [73]. Thus, the need for reasonably priced
HTS catalysts is providing an impetus for continuous studies
of Cr-free Fe-based catalysts. Based on this review, these studies should focus on developing well-matched (non-chromium)
promoter groups and crystalline matrixes for the active iron
species.


D.-W. Lee et al. / Catalysis Today 210 (2013) 2–9

Acknowledgements
This work was supported by the Human Resources Development
of the Korea Institute of Energy Technology Evaluation and Planning (20114010203050) grant funded by the Korea government
Ministry of Knowledge Economy.
This work was supported from GTL-FPSO project (KIST,
2M29760) funded by the Korean government, Ministry of Knowledge Economy (Project number: MKE2011T100200023) and
Industries (DSME, KOGAS, JNK Heaters, Hy-lok Korea).
Dr. Dae-Won Lee is supported by a Korea University grant.
References
[1] D. Mendes, A. Mendes, L.M. Madeira, A. Iulianelli, J.M. Sousa, A. Basile, AsiaPacific Journal of Chemical Engineering 5 (2010) 111.
[2] L. Lloyd, D.E. Ridler, M.V. Twigg, in: M.V. Twigg (Ed.), Catalyst Handbook, 2nd
ed., Wolfe Publishing Ltd., London, 1989, pp. 283–292.
[3] K. Kochloefl, in: G. Etrl, H. Knozinger, J. Weitkamp (Eds.), Handbook of Heterogeneous Catalysis, vol. 4, Wiley-VCH, Weinheim, 1997, pp. 1831–1843.
[4] C.N. Satterfield, Heterogeneous Catalysis in Industrial Practice, 2nd ed.,
McGraw-Hill, New York, 1991, pp. 442–446.
[5] J.J. Baschuk, X.G. Li, International Journal of Energy Research 25 (2001) 695.
[6] X. Cheng, Z. Shi, N. Glass, L. Zhang, J. Zhang, D. Song, Z-S. Liu, H. Wang, J. Shen,
Journal of Power Source 165 (2) (2007) 739.
[7] E.D. Park, D. Lee, H.C. Lee, Catalysis Today 139 (4) (2009) 280.

[8] J.M. Moe, Chemical Engineering Progress 58 (2) (1962) 33.
[9] D.S. Newsome, Catalysis Reviews: Science and Engineering 21 (2) (1980) 275.
[10] G.J. Hutchings, R.G. Copperthwaitet, F.M. Gottschalk, R. Hunter, J. Mellor, S.W.
Orchard, T. Sangiorgio, Journal of Catalysis 137 (2) (1992) 408.
[11] J.R. Mellor, R.G. Copperthwaite, N.J. Coville, Applied Catalysis 164 (1–2) (1997)
69.
[12] D. Andreev, V. Idakiev, T. Tabakova, A. Andreev, R. Giovanoli, Applied Catalysis
A: General 134 (2) (1996) 275.
[13] V. Idakiev, Z.-Y. Yuan, T. Tabakova, B.-L. Su, Applied Catalysis A: General 281
(1–2) (2005) 149.
[14] D. Tibiletti, A. Amieiro-Fonseca, R. Burch, Y. Chen, J.M. Fisher, A. Goguet, C.
Hardacre, P. Hu, D. Thompsett, Journal of Physical Chemistry B 109 (47) (2005)
22553.
[15] G. Jacobs, L. Williams, U. Graham, G.A. Thomas, D.E. Sparksand, B.H. Davis,
Applied Catalysis 252 (1) (2005) 66–74.
[16] J.W. Lywood, M.V. Twigg, Eur. Patent Appl., 0361648, 1990.
[17] C. Rhodes, G.J. Hutchings, A.M. Ward, Catalysis Today (1995) 43.
[18] C. Rhodes, B.P. Williams, F. King, G.J. Hutchings, Catalysis Communications 3
(2002) 381.
[19] I.L. Júnior, J.M.M. Millet, M. Aouine, M.C. Rangel, Applied Catalysis A: General
283 (2005) 91.
[20] J.Y. Lee, D.-W. Lee, K.-Y. Lee, Y. Wang, Catalysis Today 146 (2009) 260.
[21] J.Y. Lee, D.-W. Lee, Y.-K. Hong, K.-Y. Lee, International Journal of Hydrogen
Energy 36 (2011) 8173.
[22] G.K. Reddy, P. Boolch, P.G. Smirniots, Journal of Catalysis 282 (2011) 258.
[23] C. Martos, J. Dufour, A. Ruiz, International Journal of Hydrogen Energy 34 (10)
(2009) 4475.
[24] Y.S. Oh, H.S. Roh, K.W. Jun, Y.S. Baek, International Journal of Hydrogen Energy
28 (2003) 1387.
[25] S.H. Kim, S.W. Nam, T.H. Lim, H.I. Lee, Applied Catalysis B: Environmental 81

(2008) 97.
[26] J.L.R. Costa, G.S. Marchetti, M.C. Rangel, Catalysis Today 77 (2002) 205.
[27] A.L.C. Pereira, G.J.P. Berrocal, S.G. Marchetti, A. Albornoz, A.O. de Souza, M.C.
Rangel, Journal of Molecular Catalysis A: Chemical 281 (2008) 66.
[28] S. Natesakhawat, X. Wang, L. Zhang, U.S. Ozkan, Journal of Molecular Catalysis
A: Chemical 260 (2006) 82.
[29] L. Zhang, X. Wang, J.M.M. Millet, P.H. Matter, U.S. Ozkan, Applied Catalysis A:
General 351 (2008) 1.
[30] L. Zhang, J.M.M Millet, U.S. Ozkan, Applied Catalysis A: General 357 (2009)
66.

9

[31] P. Gawade, B. Mirkelamoglu, B. Tan, U.S. Ozkan, Journal of Molecular Catalysis
A: Chemical 321 (2010) 61.
[32] K. Watanabe, T. Miyao, K. Higashiyama, H. Yamashita, M. Watanabe, Catalysis
Communications 10 (2009) 1952.
[33] G.C. Araújo, M.C. Rangel, Catalysis Today 62 (2000) 201.
[34] C. Bosch, W. Wild, Canada Patent 153,379, 1914.
[35] W. Ruettinger, O. Ilinich, in: S. Lee (Ed.), in: Encyclopedia of Chemical
Processing, Taylor & Francis, 2006.
[36] L. Lloyd, D.E. Ridler, M.V. Twigg, in: M.V. Twigg (Ed.), Catalyst Handbook, 2nd
ed., Wolfe Publishing Ltd., London, 1989, pp. 293–295.
[37] C. Rhodes, G. Hutchings, Physical Chemistry Chemical Physics 5 (2003) 2719.
[38] G.K. Reddy, P.G. Smirniotis, Catalysis Letters 141 (1) (2011) 27.
[39] A. Khan, P. Chen, P. Boolchand, P.G. Smirniotis, Journal of Catalysis 253 (2008)
91.
[40] V. Idakiev, D. Mihajlova, B. Kunev, A. Andreev, Reaction Kinetics and Catalysis
Letters 33 (1) (1987) 119.
[41] C. Martos, J. Dufour, A. Ruiz, International Journal of Hydrogen Energy 34 (2009)

4475.
[42] G.K. Reddy, K. Gunasekera, P. Boolchand, J. Dong, P.G. Smirniotis, Journal of
Physical Chemistry C 115 (2011) 7586.
[43] T. Popa, G. Xu, T.F. Barton, M.D. Argyle, Applied Catalysis A: General 379 (2010)
15.
[44] L. Lloyd, D.E. Ridler, M.V. Twigg, in: M.V. Twigg (Ed.), Catalyst Handbook, 2nd
ed., Wolfe Publishing Ltd., London, 1989, pp. 302–304.
[45] L. Lloyd, D.E. Ridler, M.V. Twigg, in: M.V. Twigg (Ed.), Catalyst Handbook, 2nd
ed., Wolfe Publishing Ltd., London, 1989, pp. 304–306.
[46] F. Domka, A. Basinska, R. Fieldcrow, Surface Technology 18 (1983) 275.
[47] G.C. Chinchen, R.H. Logan, M.S. Spencer, Applied Catalysis 12 (1984) 89.
[48] G.C. Chinchen, R.H. Logan, M.S. Spencer, Applied Catalysis 12 (1984) 93.
[49] M. Robbins, G.K. Wertheim, R.C. Sherwood, D.N.E. Buchanan, Journal of Physics
and Chemistry of Solids 32 (1971) 717.
[50] M.A. Edwards, D.M. Whittle, C. Rhodes, A.M. Ward, D. Rohan, M.D. Shannon,
G.J. Hutchings, C.J. Kiely, Physical Chemistry Chemical Physics 4 (2002) 3902.
[51] C. Pellerin, S.M. Booker, Environmental Health Perspectives 108 (2000) a402.
[52] P.C. Grevatt, Toxicological Review of Hexavalent Chromium, U.S. Environmental Protection Agency, 1998.
[53] U.S. Department of Labor, Hexavalent Chromium, OSHA 3373-10, 2009.
[54] Official Journal of the European Union 46 (L37) (2003) 19–37.
[55] C. Ratnasamy, J.P. Wagner, Catalysis Reviews 51 (2009) 325.
[56] Y. Hattori, K. Yamamoto, J. Kaita, M. Matsuda, S. Yamada, Journal of the American Oil Chemists Society 77 (12) (2000) 1283.
[57] J. Ladebeck, T. Regula, Studies in Surface Science and Catalysis 121 (1999) 215.
[58] G.C. Chinchen, Eur. Patent Appl. 0062410, 1982.
[59] D.G. Rethwisch, J.A. Dumesic, Applied Catalysis 21 (1986) 97.
[60] D.G. Rethwisch, J. Phillips, Y. Chen, T.F. Hayden, J.A. Dumesic, Journal of Catalysis
91 (1985) 167.
[61] J. Ladebeck, K. Kochloefl, Studies in Surface Science and Catalysis 91 (1995)
1079.
[62] Q. Liu, W. Ma, R. He, Z. Mu, Catalysis Today 106 (2005) 52.

[63] H. Jin, Q. Liu, Z. Mu, L. Diao, X. Zhang, W. Ma, Chin. Patent ZL95 121834.4, 1999.
[64] M.L. Kundu, A.C. Sengupta, G.C. Maiti, B. Sen, S.K. Ghoshi, V.I. Kuznetsov, G.N.
Kustova, E.N. Yurchenko, Journal of Catalysis 112 (1988) 375.
[65] A. Andreev, V. Idakiev, D. Mihajlova, D. Shopov, Applied Catalysis 22 (1986)
385.
[66] J.Y. Lee, D.-W. Lee, M.S. Lee, K.-W. Lee, Catalysis Communications 14 (2011) 37.
[67] M.S. Lee, J.Y. Lee, D.-W. Lee, D.J. Moon, K.-W. Lee, International Journal of Hydrogen Energy 37 (2011) 11218.
[68] K. Watanabe, T. Miyao, K. Higashiyama, H. Yamashita, M. Watanabe, Catalysis
Communications 12 (2011) 976.
[69] A. Boudjemaa, C. Daniel, C. Mirodatos, M. Trari, A. Auroux, R. Bouarab, Comptes
Rendus Chimie 14 (2011) 532.
[70] N. Mahadevaiah, P. Singh, B.D. Mukri, S.K. Parida, M.S. Hedge, Applied Catalysis
B: Environmental 108–109 (2011) 117.
[71] T. Blunesin, R.J. Gorte, G.W. Graham, Applied Catalysis B: Environmental 15
(1998) 107.
[72] Y. Sun, S.S. Hla, G.J. Duffy, A.J. Cousins, D. French, L.D. Morpeth, J.H. Edwards,
D.G. Roberts, International Journal of Hydrogen Energy 36 (2011) 79.
[73] FreedomCAR and Fuel Partnership, Hydrogen Production Roadmap:
Technology Pathways to the Future (2009): www1.eere.energy.gov/
hydrogenandfuelcells/pdfs/h2 production roadmap.pdf