You are on page 1of 9

Journal of Energy Chemistry 22(2013)795–803

Effect of mass transfer limitations on catalyst performance


during reduction and carburization of iron based
Fischer-Tropsch synthesis catalysts
Akbar Zamaniyana,b∗ , Yadollah Mortazavib , Abbas Ali Khodadadib, Ali Nakhaei Poura
a. Department of Natural Gas Conversion, Gas Research Division, Research Institute of Petroleum Industry (RIPI), Tehran 14665-137, Iran;
b. School of Chemical Engineering, University College of Engineering, University of Tehran, Tehran 11155-4563, Iran
[ Manuscript received September 2, 2012; revised October 8, 2012 ]

Abstract
Existence of intraparticle mass transfer limitations under typical Fischer-Tropsch synthesis has been reported previously, but there is no suitable
study on the existence of intraparticle diffusion limitations under pretreatment steps (reduction and activation) and their effect on catalytic
performance for iron based catalysts. In this study, Fe-Cu-La-SiO2 catalysts were prepared by co-precipitation method. To investigate the
intraparticle mass transfer limitation under reduction, activation and reaction steps, and its effect on catalytic performance, catalyst pellets with
different sizes of 6, 3, 1 and 0.5 mm have been prepared. All catalysts were calcined, pretreated and tested under similar conditions. The
catalysts were activated in hydrogen (5% H2 in N2 ) at 450 ◦ C for 3 h and exposed to syngas (H2 /CO = 1) at 270 ◦ C and atmospheric pressure
for 40 h. Afterwards, FTS reaction tests were performed for approximately 120 h to reach steady state conditions at 290 ◦ C, 17 bar and a feed
flow (syngas H2 /CO = 1) rate of 3 L/h (STP). Using small pellets resulted in higher CO conversion, FT reaction rate and C5+ productivity
as compared with larger pellets. The small pellets reached steady state conditions just 20 h after starting the reaction. Whereas for larger
pellets, CO conversion, FT reaction rate and C5+ productivity increased gradually, and reached steady state and maximum values after 120 h of
operation. The results illustrate that mass transfer limitations exist not only for FTS reaction but also for the reduction and carburization steps
which lead to various phase formation through catalyst activation. Also the results indicate that some effects of mass transfer limitations in
activation step, can be compensated in the reaction step. The results can be used for better design of iron based catalyst to improve the process
economy.
Key words
Fischer-Tropsch; mass transfer limitation; activation; catalyst; GTL

1. Introduction range of 200–300 ◦ C and pressure range of 10–60 bar. Iron


catalysts are commonly used due to their low costs in compar-
In the face of increasing oil prices and decreasing oil re- ison with other active metals [5].
serves, alternative routes for production of liquid fuels attract Various types of reactors (fixed-bed, fluidized-bed,
considerable attention in research and industrial application. ebulliating-bed and slurry-phase) differ with respect to the
Fischer-Tropsch synthesis (FTS) offers the possibility of the most suitable particle size of the catalyst used. The range of
conversion of synthesis gas to a mixture of linear hydrocar- catalyst sizes reflects the tradeoff between the desires to pro-
bons. Fuels produced with FTS are of a high quality due to a vide high catalyst utilization and maintain a manageable pres-
very low aromaticity and near zero sulfur content. New and sure drop across the reactor. If a fixed-bed mode of operation
stringent regulations may promote replacement or blending of is envisaged, the FT catalyst will generally consist of particles
conventional fuels by sulfur and aromatic-free FTS products with a few millimeters in size for reasons of pressure drop and
[1−4]. heat transfer. It can be easily envisaged that in this size, in-
The most common Fischer-Tropsch catalysts are group traparticle pore diffusion limitation may exist, as previously
VIII metals (Co, Ru and Fe). In conventional FTS processes, reported [6,7].
iron and cobalt based catalysts are known to be very effective Using a model that assumed first-order kinetics with
catalysts which are used commercially in the temperature respect to hydrogen, Zimmerman et al. [8] strongly suggested

Corresponding Author. Tel: +98-21-48252336; Fax: +98-21-44739716; Email: zamaniyana@ripi.ir

Copyright©2013, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. All rights reserved.
796 Akbar Zamaniyan et al./ Journal of Energy Chemistry Vol. 22 No. 5 2013

that transport limitations of H2 occur at particle with diame- respect to product yield and selectivity requires a deep under-
ters larger than 0.2 mm at temperature higher than 235 ◦ C with standing of catalyst behavior in reduction, carburization and
a fused iron ammonia synthesis catalyst. Post et al. [9] also reaction steps.
used first-order behavior and observed transport limitations of In this study, using different catalyst sizes, the existence
hydrogen at high temperatures (T >220 ◦ C, dP >0.4 mm) with of intraparticle pore diffusion limitations in porous iron based
a number of iron and cobalt based catalysts in a fixed-bed mi- FT catalysts under reduction, carburization (activation) and
cro reactor. Robert Becker et al. [10] showed computation- reaction steps and their effects on catalyst performance, have
ally that under typical Fischer-Tropsch conditions (200 ◦ C, been examined and described.
2.1 MPa), CO can also become diffusion limited. Intrin-
sic FT kinetic measurements should be performed with small 2. Experimental
(dP <0.2 mm) catalyst particles in order to eliminate diffusion
limitations of one of the reactants. Some of the most pertinent
information can be found in the results obtained by Post et al. 2.1. Catalyst synthesis
[9], who have demonstrated that the synthesis rate over a fused
iron catalyst which is essentially nonporous in the oxidic state, Most early FT catalysts were prepared with precipita-
increases with the decreasing particle size. This could be due tion techniques [5]. Alkali-promoted iron catalysts have been
to the diffusion limitations during synthesis reaction or to the applied industrially for Fischer-Tropsch synthesis for many
activation of the nonporous oxidic catalyst particles limited to years that have high water-gas shift activity, high selectivity to
a superficial layer only [9]. olefins and appear to be stable with a high H2 /CO ratio synthe-
There are number of studies on mass transfer limitations sis gas [3,11]. Here, Fe-Cu-La-SiO2 catalysts were prepared
and their effects during FTS, but to the best our knowledge by co-precipitation of Fe and Cu nitrates at constant pH of
there is no suitable study previously published on the exis- 7.5 to form porous Fe-Cu oxyhydroxide powder, which were
tence of intraparticle pore diffusion limitations under pretreat- promoted by La(NO3 )3 precursor and silica sol solution after
ment steps (reduction and carburization) and specially their treatment in air [12]. The nominal composition of the catalyst
effects on catalyst performance for iron based catalyst. The by weight ratio was tended to be 100Fe : 6Cu : 6La : 20SiO2 .
diffusion limitations result in an incomplete utilization of the To study the effect of intraparticle mass transfer limi-
catalyst particles and lead to changes in reactivity and selec- tations during reduction, activation and reaction steps and
tivity. It is of practical importance in the design of iron based their effects on catalyst performance, the catalyst pellets with
FT catalysts to obtain knowledge of various phases formed different sizes of 6, 3, 1 and 0.5 mm have been prepared as
during different pretreatments and under industrial FTS con- shown in Figure 1. All samples were calcined at 450 ◦ C in air
ditions. In addition of FT chemistry, an optimal operation with for 3 h.

Figure 1. Appearance photographs of the prepared catalyst pellets with different sizes. (a) 6 mm, (b) 3 mm, (c) 1 mm, (d) 0.5 mm
Journal of Energy Chemistry Vol. 22 No. 5 2013 797

2.2. Catalyst characterization tion treatments for iron catalysts are H2 , CO, or synthesis gas
reduction [5].
Atomic absorption spectroscopy (Analyst 200 Perkin Here, the catalyst was activated in hydrogen (5% H2 in
Elmer) was used to determine the iron and copper con- N2 ) at 450 ◦ C for 3 h. The reactor temperature was then
tent of the catalyst. The contents of lanthanum and cooled down to 270 ◦ C under flowing hydrogen. Then, the
silica were determined by ICP (Wear Metal Analyzer catalyst was exposed to syngas (H2 /CO = 1) and remained un-
400 Perkin Elmer) and gravimetric methods, respectively. der this condition for about 40 h that is required in order to
The composition of the catalyst was determined to be form carbide phases. Afterwards, the reactor was pressurized
100.0Fe : 5.6Cu : 5.6La : 19.0SiO2. to 17 bar and the temperature was increased to 290 ◦ C for FTS
The surface area was calculated from the Brunauer- reaction tests. The pretreatment and reaction conditions were
Emmett-Teller (BET) equation, and pore volume and aver- selected based on previous works [3,11,15].
age pore diameter of the catalyst were determined by N2 ph-
ysisorption using a Micromeritics ASAP 2010 automated sys- 2.4. Experimental setup and test procedure
tem. A 0.5 g catalyst sample was degassed at 373 K for 1 h
and then at 573 K for 2 h prior to analysis. The analysis was All prepared catalysts were tested in an experimental set-
done using N2 adsorption at 77 K. up as shown in Figure 2. The feed stream was supplied from
XRD was used to determine the phase composition of high-pressure carbon monoxide and hydrogen cylinders using
the synthesized catalysts. The XRD pattern of the cata- individual mass flow controllers. The experiments were car-
lysts were collected using an X-ray diffractometer (Philips ried out using a stainless steel reactor (14.3 mm ID) placed in
PW1840 X-ray diffractometer) by monochromatized Cu/Kα a tubular triple zones furnace. The catalyst, without any di-
radiation (40 kV, 40 mA) using a step scan mode at a scan rate lution, was located between two layers of ceramic beads to
of 0.02o/s (2θ) from 10o −80o. XRD peak identification was homogenize and preheat the feed before it entered the cat-
performed by comparison with the JCPDS database software. alyst bed. For measuring and controlling the reaction tem-
perature, three J-type TICs (temperature indicators and con-
2.3. Catalyst reduction and activation trollers) were installed. The first thermocouple was located
before the catalyst bed in the preheater section whilst the sec-
The catalyst is synthesized in the form of a metal oxide, so ond and the third ones were inside and after the catalyst bed.
it must be subjected to an activation treatment to become ac- No temperature runaway was observed during the reaction
tive for Fischer Tropsch synthesis. The pretreatment for iron tests. The system pressure was controlled by a backpressure
based catalysts is not straightforward. The pretreatment en- regulator.
vironment has been reported to influence the initial catalytic The performances of all the catalysts were measured
activity significantly [13,14]. The common reported activa- under similar conditions. For this purpose, 2.7 g catalyst was

Figure 2. Experimental setup


798 Akbar Zamaniyan et al./ Journal of Energy Chemistry Vol. 22 No. 5 2013

loaded in the reactor. After catalyst reduction and activa- 9 nm, as determined by N2 adsorption (Micrometrics ASAP
tion the feed flow (syngas H2 /CO = 1) equal to 3 L/h (STP) 2010).
that provided the space velocity of 750 h−1 was conducted at Figure 3 shows XRD analysis for the catalyst samples.
17 bar and 290 ◦ C to start FTS reactions. The FTS test was The diffraction lines belong to magnetite (Fe3 O4 ) should be
took time about 120 h to reach steady state conditions. at about 2θ = 31o , 37o , 59o and 65o , and the diffraction lines
The reactor effluent was passed through a hot separa- belong to hematite (Fe2 O3 ) should be at about 2θ = 24o , 33o ,
tor operated at 80 ◦ C and cold separator operated at 0 ◦ C to 35o , 49o and 54o .
remove heavy and light liquid products, respectively. Non- Due to the existence and distribution of SiO2 , the inten-
condensable gases and feed were analyzed using an on-line sity of the hematite peaks was low. Figure 3 illustrates clearly
gas chromatograph (Varian CP 3800) equipped with a packed that the synthesized catalysts contained mainly magnetite and
column (3 m long) and a capillary column (Petrocol DH hematite phases.
100 m long) to determine CO conversion and the formation
of light hydrocarbon products. The liquid products were col-
lected and weighed from hot and cold separators every 24 h.

2.5. Reaction system and calculations

The overall reactions of Fischer-Tropsch synthesis can be


simplified as a combination of FT reaction and water-gas shift
(WGS) reaction as follows [16]:
CO+(1+m/2n)H2−→(1/n)Cn Hm +H2 O, rFTS (1)
Figure 3. XRD patterns of synthesized catalyst samples
CO+H2 O⇐⇒CO2 +H2 , rWGS (2)
where, n is the average carbon number and m is the average
number of hydrogen atoms in hydrocarbon products. Water is 3.2. FTS reaction tests
a primary product of FT reaction, and CO2 can only be pro-
duced by the water-gas shift reaction, so it can be concluded The results for CO conversion versus time during 120 h of
that: FTS reaction tests are presented in Figure 4. It can be seen that
RWGS = Rate of water-gas shift reaction (mol/(gFe ·h)) CO conversions for smaller catalyst size (0.5 mm and 1 mm)
= Rate of CO2 production (3) were higher than those of larger ones (3 mm and 6 mm). Also,
Thus the rate of FT reaction can be obtained by the fol- as demonstrated in Figure 4, the approach to steady state was
lowing formula: markedly different for different catalyst sizes. It should be
emphasized that the steady-state activity (and selectivity) of
RFTS = Rate of FTS (mol/(gFe ·h)) = RCO –RWGS
iron catalysts does not reached in minutes but rather over pe-
= RCO –RCO2 (4)
riods of several hours. The time required to reach steady state
RCO = Rate of CO consumption (mol/(gFe·h)) (5) conditions and maximum attainable CO conversion for these
catalyst sizes are also different. These results clearly indi-
CO conversion, CH4 selectivity and catalyst productivity
cated the existence of mass transfer limitations in FTS for the
are required to evaluate the catalyst performance which have
applied catalyst under operation conditions.
been calculated from the following formulas:
CO conversion = Mole of CO converted/Total CO inlet (6)
CH4 selectivity = Mole of CH4 produced/CO converted (7)
Productivity = W desired product (g)/[WFe (g)×time (h)] (8)
To ensure the reproducibility of the results, at least two
sets of data under similar conditions were obtained. The sta-
tistical analysis on the data was done and error lower than
±2.0% was obtained.

3. Results

3.1. Catalyst characterization

BET surface area of the catalyst was 107 m2 /g, the pore Figure 4. CO conversion versus time during FTS reaction for various catalyst
volume was 0.24 cm3 /g and the average pore diameter was sizes
Journal of Energy Chemistry Vol. 22 No. 5 2013 799

The time required to reach steady state and maximum


value for CO conversion was about 20 h for the 0.5 and 1 mm
catalysts. At this time, CO conversion was about 93% and
82% for the 0.5 and 1 mm catalysts, respectively, whereas at
the same point in time CO conversion for the 3 and 6 mm cat-
alyst pellets were 38% and 60%, respectively. It is surprising
and unexpected that CO conversion for the 6 mm catalyst was
higher than that for the 3 mm one. The following figures and
discussion may help to interpret this behavior.
As Figure 4 indicates, the time required for CO conver-
sion to reach steady state and maximum value was about 100 h
for the 6 mm catalyst. CO conversion for the 3 mm catalyst
increased continuously even after 120 h. Therefore, for the
same catalyst composition, the catalyst with smaller size re-
quires less activation time and shows higher CO conversion Figure 6. WGS reaction rate versus time during FTS for various catalyst
which in turn affects the economy of the process. sizes
The results for FTS rate of reaction versus time are pre-
The results for CO2 and CH4 selectivities for various cat-
sented in Figure 5. The behavior of various catalyst sizes, in
alyst sizes versus reaction time are presented in Figures 7 and
regard to the required time to achieve steady state, was rela-
8, respectively.
tively similar to Figure 4 only with a few differences. Simi-
WGS reaction is the likely source of CO2 production, so
lar to Figure 4, smaller catalysts showed higher FTS rates of
similar results to Figure 6 can be seen here. The 6 mm cat-
reaction as compared with the larger catalysts. In contrary alyst showed the highest CO2 and CH4 selectivities. All the
to Figure 4, the 0.5 and 1 mm catalysts indicated the same catalysts showed a maximum CH4 selectivity at around 20 h
FTS rate of reaction which in turn implies that they have sim-
ilar mass transfer limitation characteristics under the reaction
condition. Thus it can be concluded that mass transfer limi-
tations can be eliminated by catalysts with sizes in the range
of 0.5 mm around under the applied condition. The rate of
FTS reaction increased continuously even after 120 h for the
3 and 6 mm catalysts whereas it approached steady state and
maximum value for the 0.5 and 1 mm catalysts faster.
Figure 6 shows the results for WGS reaction rate. The
6 mm and 0.5 mm catalysts displayed the highest WGS reac-
tion rates. Similar to CO conversion and FTS rate of reac-
tion, the rate of WGS reaction for the 3 mm catalyst increased
continuously even after 120 h. On the other hand, WGS rate
reached steady state conditions more rapidly for the others.
The values of major parameters obtained from an analysis
of Figures 4 to 6 are summarized in Table 1. Figure 7. CO2 selectivity versus time during FTS for various catalyst sizes

Figure 5. FT reaction rate versus time during FTS for various catalyst sizes Figure 8. CH4 selectivity versus time during FTS for various catalyst sizes
800 Akbar Zamaniyan et al./ Journal of Energy Chemistry Vol. 22 No. 5 2013

Table 1. Major reaction parameters for various catalyst sizes


Catalyst size Maximum CO Time to reach Initial activity (CO Maximum FTS rate of Maximum WGS rate of
(nm) conversion (%) steady state (h) conversion at 20 h) (%) reaction (mol/(h·gFe )) reaction (mol/(h·gFe ))
0.5 94.3 20 93.1 0.0215 0.0167
1 82.7 20 77.3 0.0206 0.0158
3 70.5 − 47.3 0.0199 0.0135
6 73.3 100 57.2 0.0164 0.0167

of FTS reaction time. CH4 selectivities for 6 mm and 3 mm in catalyst performance versus time can be explained by the
catalysts decreased slowly after 20 h (i.e., the maximum differences in distribution of iron phases formed during the
value) running of FTS reaction. pretreatment which can be attributed to the differences in mass
Figure 9 shows C5+ productivity for various catalyst transfer limitations resulted from differences in catalyst size.
sizes. C5+ productivity increased versus time for all the cata-
lysts. The 0.5 mm and 1 mm catalysts exhibited about double
C5+ productivity as compared with the 6 mm and 3 mm cata-
lysts.

4. Discussion

The significant differences in CO conversion, the required


time to reach steady state, FTS reaction rate and WGS reaction
rate (Figures 4−9) imply the existence of mass transfer lim-
itations. The results can be interpreted by the effect of mass
transfer limitations on various steps, including reduction, car-
burization and reaction.
It has been reported that cobalt, nickel and ruthenium
remain in metallic state under FTS conditions [16], but the
composition of iron-based catalysts changes during Fischer- Figure 9. C5+ productivity versus time during FTS for various catalyst sizes
Tropsch synthesis. Several phases of iron are known in iron-
based catalysts subjected to FTS conditions. These include As mentioned earlier, the reduction and carburization pro-
metallic iron (α-Fe), iron oxides (hematite, α-Fe2 O3 ; mag- cesses are performed by hydrogen and syngas (H2 , CO), re-
netite Fe3 O4 , and Fex O), and five different forms of iron car- spectively. Bulk and surface changes in the speciation of iron
bides. The formation and composition of these iron phases de- during the pretreatment of FTS reaction on iron catalysts have
pend on the process conditions, catalyst activation and deacti- been studied and sequential bulk phase modifications during
vation, and catalyst composition [17−21]. There are a number the activation process from Fe2 O3 to Fe3 O4 to FeCx (iron car-
of studies on iron catalysts concerning the role of iron phases bides) have been reported [19−21]. The sequence of transfor-
in FT synthesis that indicate the carbide phases result in high mation for iron phases during the pretreatment and FTS can
FT activity whereas the magnetite phase has negligible cat- be illustrated schematically in Figure 10. It also addresses the
alytic activity toward FT reactions but may be active for WGS changes that occur during the exposure to H2 and syngas (H2
reaction [5,16,17,22]. Thus, it is expected that the differences and CO) environments.

Figure 10. Transformation sequence for iron phase during pretreatment and FTS [20]

The studies indicate that during the pretreatment and re- concentration profiles in reduction, carburization and reaction
action steps, Fe2 O3 crystallites are rapidly converted to Fe3 O4 steps. Thus, the combinations of various phases (hematite,
crystallites, but the conversion of Fe3 O4 crystallites to iron magnetite, carbides and Fe metal) are different for various cat-
metal and carbides is relatively slow and requires a number of alyst sizes. Figure 11 indicates that this is the case indeed. In
steps [19,20]. On the other hand, the diffusivities of H2 and this figure, the sharp peaks observed clearly at 2θ equal to 35o ,
CO are different, with hydrogen diffusivity being greater than 57o and 63o assign to Fe3 O4 , and the peaks at 2θ equal to 39o ,
that of CO. Also, different catalyst sizes result in different 41o , 43o , 44o , 70o and 78o assign to various types of iron car-
diffusion lengths. These parameters influence the H2 and CO bides. Also the peaks at 2θ equal to 24.3o, 33.3o, 35.8o, 40.8o,
Journal of Energy Chemistry Vol. 22 No. 5 2013 801

49.6o, 54.1o, 57.6o and 64.1o assign to Fe2 O3 . the color contrast, albeit in a qualitative manner depicts, the
Figure 12 traces the iron phase transformations of iron amount of each phase present with black color denoting high
catalyst from calcined catalyst through the pretreatments (re- content, gray denoting less content and white meaning the
duction and carburization) and reaction steps. In this figure, lowest content.

Figure 11. XRD patterns of the catalysts after reaction

Figure 12. Qualitative trace of iron phase transformations of iron catalyst from the calcined catalyst through pretreatments and reaction steps

After 3 h pretreatment of the calcined catalyst in the pres- be noted that the mass transfer limitation is highlight and im-
ence of H2 , hematite phase (Fe2 O3 ) was reduced partially to portant here, due to low diffusivity of CO, number of slow
Fe3 O4 and totally to metal Fe. Thus, a mixture of Fe2 O3 , chain steps, and different diffusion lengths. Thus the degree of
Fe3 O4 and Fe metal formed. In this step, the content of vari- carburization process and the amount of various phases will be
ous phases in the catalyst increased in the order of Fe metal, differed for different catalyst sizes. Therefore, it is expected
Fe2 O3 and Fe3 O4 , as shown in Figure 12. With regard to the that the extent of carburization (carbide phase formation) in
high diffusivity of H2 , it is expected that a similar combina- terms of catalyst size decreases in the order of 0.5, 1, 3 and
tion of the aforementioned phases formed for all the catalyst 6 mm. In other words, the amount of magnetite phase in terms
sizes during this step. of catalyst size increases in the order of 0.5, 1, 3 and 6 mm.
At carburization step after exposure to syngas for 40 h, Briefly, it is expected that the catalysts with larger diffusion
Fe3 O4 , Fe2 O3 and metal Fe were converted partially to car- length contain mainly magnetite and hematite phases, and less
bide phases. The reaction itself may promote the carburiza- carbide and active metal phases.
tion. Thus during the reaction step, some of other remaining Figure 4 shows that the 0.5 and 1 mm catalyst pel-
phases may be converted to carbide phases and lead to an in- lets activate rapidly and reach to maximum CO conversion
crease of carbide phase content as shown in Figure 12. It must value about 20 h after reaction starts. This indicates that these
802 Akbar Zamaniyan et al./ Journal of Energy Chemistry Vol. 22 No. 5 2013

catalysts have been reduced, carburized and the active phases Typically, steady-state activity reached only after several
have been formed well. On the other hand, CO conversion hours of reaction as a result of the slow transformations of
for the 3 mm pellet increases with time gradually. CO is the iron and iron oxides species to carbide phase at the surface
key to form carbide phases. As mentioned earlier, there are and in the bulk of iron crystallites [19,20]. As shown in Fig-
more mass transfer limitations under carburization step than ure 4, the 0.5 and 1 mm catalysts reach steady-state activ-
those under reduction one. As the diffusional lengths of the ity only after 20 h of the reaction. It demonstrates that due
3 mm catalysts are, on average, larger than those of the 0.5 to low mass transfer limitations, transformations to carbide
and 1 mm pellets, mass transfer limitation of CO can affect the phase take place rapidly. Whereas, this not the case for the
formation rate of carbide phases during carburization. It may 3 mm and 6 mm catalysts, which need much more time to
be concluded that the 3 mm catalyst has likely been reduced reach steady-state activity (Figures 4 and 5).
relatively well, whereas this is not the case for carburization. As WGS reaction is the source of CO2 production and
Due to similarity of the reaction and carburization steps, the with respect to previous figures and results, the behavior of
carburization process can be continued in parallel to the re- Figure 7 is known clearly and there is no need for additional
action, and the remaining iron oxides and metal sites can be discussion, but the results of Figure 8 need some discussion
converted to carbides phases in the presence of CO. Figure 11 as following:
shows that this is the case as the XRD patterns of the catalysts 1. Low CO conversion for the 6 mm catalyst as shown in
are relatively similar to each other after reaction. Figure 4, leads to high CH4 selectivity based on Equation 7.
CO conversion for the 6 mm catalyst shows similar be- 2. High methane selectivity can be explained by catalyst
havior, as mentioned in previous discussion. The increase rate surface wetting effect [23]. The methane formation is con-
of CO conversion for the 6 mm catalyst was less than that for sidered mainly due to methanation reaction at the dry catalyst
the 3 mm catalyst. Longer diffusion lengths for the 6 mm cat- surface. Evolution of the liquid product along the catalytic
alyst relative to the 3 mm one lead to more CO mass transfer bed is illustrated in Figure 13. The liquid product can be ac-
limitations and less formation of carbide phases during the re- cumulated in the made volume among the particles. For large
action testing. It is surprising that CO conversion for the 6 mm catalyst particles, the accumulated liquid may be low due to
catalyst is more than that of the 3 mm one. carrying by the gas stream. Also as shown in Figure 5, the rate
In spite of Figure 4, the 6 mm catalyst showed a lower of FT reaction is low for large particles due to incomplete car-
FT reaction rate than the 3 mm one (Figure 5). As mentioned burization. Thus, there is more dry catalyst surface, resulting
above, it is expected that the 6 mm catalyst contains mainly in more methane formation. A decrease in methane selectiv-
hematite and magnetite phases as compared with others. Re- ity for all the catalysts versus time may be explained by this
ferring to prior discussion, due to CO mass transfer limita- effect too. The dry catalyst surface can be covered by liquid
tions, longer diffusion lengths lead to incomplete carburiza- product with time, so methane formation decreases slightly
tion and active phase formation, not only in the carburization versus time for the catalysts with larger size. Also it can be
step but also during the reaction. Thus, for this catalyst, a low concluded that this not the case for the 0.5 and 1 mm sizes.
content of active phase (carbide and active metal) and more The methane selectivity reaches to steady state condition af-
magnetite phase (active phase for WGS) are the possible rea- ter 20 h for these catalysts. Here, the catalyst with small size
sons for a lower FT reaction rate and higher WGS reaction leads to good filling of the catalyst bed, resulting in that the
rate, respectively (Figures 5 and 6). catalyst surface is covered rapidly with liquid product. After
On the other hand, due to a lower FT reaction rate, less 20 h, there is no significant change in methane selectivity for
heat is released by the reaction, causing a lower bed temper- the 0.5 and 1 mm catalysts which can be explained by Equa-
ature for the 6 mm catalyst relative to the other catalysts. As tion 7.
the water gas shift reaction is an exothermic and equilibrium
limited reaction, low temperature leads to an increase in WGS
reaction rate too. Thus for the 6 mm catalyst, higher WGS
reaction rate relative to the other catalysts and higher CO con-
version relative to the 3 mm catalyst are expected. The 0.5 mm
catalyst leads to better filling of the catalyst bed, so the reactor
has better radial heat transfer performance, a smaller temper-
ature gradient, and less hot spot that cause high WGS reaction
rate.
Figure 6 indicated that WGS reaction rate increases with
time for all the catalysts. This is not the reason for increase Figure 13. Evolution of liquid product along the catalytic bed
of the amount of magnetite phase with time; because, as men-
tioned before, the magnetite phase decreases with time due to 3. High WGS reaction rate causes the usage of CO and
the continuation of carburization process in the reaction step. production of H2 , so H2 /CO increases. High H2 /CO leads to
The water concentration increases due to the increase of FT an increase of light products such as methane [24].
reaction rate for all the catalyst. This leads to an increase of 4. It has been reported that heat and mass transfer limi-
WGS reaction rate that is thermodynamically controlled. tations result in increases in methane formation [25]. As ex-
Journal of Energy Chemistry Vol. 22 No. 5 2013 803

plained before, mass transfer limitations exist for large cata- References
lyst size. Also due to a low radial heat transfer coefficient,
heat transfer limitation exists too. [1] Guettel R, Turek T. Chem Eng Sci, 2009, 64(5): 955
With respect to previous figures and results, the behavior [2] Fox III J M. Catal Rev-Sci Eng, 1993, 35(2): 169
of Figure 9 is known clearly and there is no need to discuss [3] Pour A N, Housaindokht M R, Tayyari S F, Zarkesh J. J Nat Gas
further. Chem, 2010, 19(2): 107
[4] Dry M E. Appl Catal A, 1996, 138(2): 319
5. Conclusions [5] Van der Laan G P, Beenackers A A C M. Catal Rev-Sci Eng,
1999, 41(3-4): 255
The effects of pretreatment steps (reduction and carbur- [6] Shen W J, Zhou J L, Zhang B J. J Nat Gas Chem, 1996, 5(1):
ization) on the catalytic performance of iron based FTS cata- 59
lyst have been investigated. The observations reveal that mass [7] Shen W J, Zhou J L, Zhang B J. J Nat Gas Chem, 1996, 5(2):
transfer limitations exist during the reduction and carburiza- 107
tion in addition to the reaction step. [8] Zimmerman W H, Rossin J A, Bukur D B. Ind Eng Chem Res,
Based on the results of our studies, we conclude that the 1989, 28(4): 406
observed differences in activity-time behavior, reaction rates, [9] Post M F M, Van’t Hoog A C, Minderhoud J K, Sie S T. AIChE
selectivity and productivity, are not only for mass transfer J, 1989, 35(7): 1107
limitations in the reaction step, but also some of them refer [10] Robert Becker E, Pereira C J. Computer-Aided Design of Cata-
to differences in the distributions of iron metal, carbides and lysts. New York: Marcel Dekker, 1993. 199
oxides (hematite and magnetite) in the catalyst resulted from [11] Pour A N, Zare M, Zamani Y. J Nat Gas Chem, 2010, 19(1): 31
mass transfer limitations in the pretreatment step. The high [12] Zamaniyan A, Mortazavi Y, Khodadadi A A, Manafi H. Appl
CO conversion, high FT reaction rate and short time required Catal A, 2010, 385(1-2): 214
to reach steady state conditions are indicative of complete re- [13] Hao Q L, Bai L, Xiang H W, Li Y W. J Nat Gas Chem, 2009,
duction and carburization and a high content of active phase 18(4): 429
for small pellets. By contrast, a low FT reaction rate, high [14] Feyzi M, Mirzaei A A, Bozorgzadeh H R. J Nat Gas Chem,
WGS reaction rate, and long time required to reach steady 2010, 19(3): 341
state condition are indicative of incomplete carburization and [15] Pour A N, Shahri S M K, Zamani Y, Zamanian A. J Nat Gas
higher content of magnetite phase for large pellets. The cat- Chem, 2010, 19(2): 193
alysts with longer diffusion lengths contain mainly magnetite [16] Davis B H, Occelli M L. Fischer-Tropsch Synthesis, Catalysts
and hematite phases and less carbide and active metal. In addi- and Catalysis. Amsterdam: Elsevier Science Ltd, 2007. 381
tion to higher CO conversion and FT reaction rates, the small [17] Lox E S, Marin G B, De Grave E, Bussière P. Appl Catal A,
catalysts need much shorter activation times due to higher ex- 1988, 40(1-2): 197
tents of reduction and carburization, which in turn affects the [18] Shroff M D, Kalakkad D S, Coulter K E, Kohler S D, Harring-
process economy. ton M S, Jackson N B, Sault A G, Datye A K. J Catal, 1995,
Due to similarities in the reaction and carburization steps, 156(2): 185
the carburization process can be continued in parallel to the [19] Xu J, Bartholomew C H. J Phys Chem B, 2005, 109(6): 2392
reaction. Thus, some effects of mass transfer limitations from [20] Davis B H. Catal Today, 2009, 141(1-2): 25
the pretreatment can be compensated in the reaction step after [21] Bukur D B, Koranne M, Lang X S, Rao K R P M, Huffman G P.
a long time, so that the remaining iron oxides and metal sites Appl Catal A, 1995, 126(1): 85
can convert to carbides phases in the presence of CO. [22] Bukur D B, Nowicki L, Manne R K, Lang X S. J Catal, 1995,
Combining the information from this study with earlier 155(2): 366
works on catalyst design, it may provide a new perspective [23] Liu W, Hu J L, Wang Y. Catal Today, 2009, 140(3-4): 142
of better design of iron based catalyst to improve the process [24] Donnelly T J, Satterfield C N. Appl Catal, 1989, 52(1-2): 93
economy. [25] Zhan X D, Davis B H. Appl Catal A, 2002, 236(1-2): 149

You might also like