Journal of Environmental Chemical Engineering 5 (2017) 54–62

Contents lists available at ScienceDirect

Journal of Environmental Chemical Engineering

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

Effect of CO2 co-feeding on the conversion of syngas derived from waste

to liquid fuel over a bi-functional Co/H-ZSM-5 catalyst
Michael Olawale Daramola* , Khuthadzo Matamela, Olawumi Oluwafolakemi Sadare
School of Chemical and Metallurgical Engineering, Faculty of Engineering and The Built Environment, University of the Witwatersrand, Private Bag X3, Wits
2050, Johannesburg, South Africa

A R T I C L E I N F O A B S T R A C T

Article history:
Received 25 July 2016 A bi-functional Co-HZSM-5 catalyst was synthesised by the Incipient Wetness Impregnation (IWI) of
Received in revised form 16 November 2016 cobalt onto H-ZSM-5 and effect of CO2-co-feeding on the performance of the catalyst during Fischer
Accepted 17 November 2016 Tropsch (F-T) was investigated. Physico-chemical property of the catalyst was checked with N2
Available online 18 November 2016 physisorption (for BET surface area), Transmission Electron Microscopy and Scanning Electron
Microscopy equipped with EDS. Effect of CO2 co-feeding was investigated at Gas Hourly Space Velocity
Keywords: (GHSV):1200 ml h
1 g
1; pressure: 15 bar, temperature: 250  C; and H2/CO ratio: 2.5; with a syngas
Bi-functional catalyst mixture containing 5% CO2. Pre-calibrated GCs were used to analyse the reaction products. The BET
Syngas
surface area, the pore volume and the pore size of the catalyst were 292 m2/g, 0.18 cm3/g and 2.83 nm,
CO2 co-feeding
respectively. TEM images showed that cobalt particles were well dispersed within the support. In the
Fisher-Tropsch
Gas-to-liquid presence of CO2 co-feeding, CO conversion was 73% and CO conversion was 53% without CO2 co-feeding.
Cobalt catalyst In addition, the co-fed CO2 was converted to hydrocarbons and a CO2 conversion of 63% was obtained. The
selectivity of the catalyst to methane was 53%, 34% to C2, <4% to C4-C5 and <10% to C6+ in the presence of
CO2 co-feeding. Co-feeding CO2 also affected the selectivity of the catalyst to olefin as well. Selectivity of
the catalyst to C3 decreased from 43% to 35% while C4 increased from 25% to 32% during CO2 co-feeding.
However, selectivity of the catalyst to C6+ olefin hydrocarbon increased from 9% to 15% during CO2 co-
feeding.
ã 2016 Elsevier Ltd. All rights reserved.

1. Introduction Gasoline (MTG) synthesis, and (iv) Methanol-to-Olefins (MTO)

synthesis [2]. Higher alcohol synthesis focuses on ethanol
Management of waste has been a crucial and delegated production, and has enjoyed much attention, but still offers a
responsibility, undertaken on behalf of the industry, public low productivity and poor selectivity. Fischer-Tropsch synthesis is
authorities by waste management specialists. Waste accumulation a surface-catalyzed polymerization reaction that involves conver-
has been known to create various problems to the environment sion of synthesis gas (CO and H2) derived from coal, natural gas,
and to the human health. These problems are aggravated when heavy oil or biomass into hydrocarbons consisting of paraffins,
waste accumulates at incorrect landsite [1]. However to solve olefins, alcohols and aldehydes Fischer-Tropsch (F-T) synthesis
environmental problems and ensure a healthy living, waste produces a wide range of mainly linear paraffinic hydrocarbons,
management is very essential. Therefore the need for environ- and the product distribution depends on the type of catalyst, the
mentally acceptable waste disposal methods and/or beneficiation process conditions and reactor configuration [2]. After F-T reaction,
has become a priority in developing countries like South Africa. the obtained long-chain linear paraffins are cracked further into
One way to beneficiate solid waste materials is to gasify them to short-chain hydrocarbons in fluidized-bed catalytic cracking (FCC)
syngas (major components are CO and H2). Syngas gas can be used units to maximize the realization of the desired product fractions
to produce a variety of chemicals and liquid fuels. Liquid fuels can [3]. Cobalt (Co) and Iron (Fe) are the two most metal catalysts
be made from syngas via four basic processes: (i) Higher alcohol employed in the commercial F-T synthesis [4,5]. These metal
synthesis, (ii) Fischer-Tropsch (F-T) synthesis, (iii) Methanol-to- catalysts can catalyze F-T synthesis at temperatures below 523 K,
also known as the Low-Temperature Fischer-Tropsch (LTFT) [3,4].
Co-based catalyst is more preferred at LTFT to Fe-based catalyst
while Fe-based catalyst is preferred at high-Temperature Fischer-
* Corresponding author.
E-mail address: michael.daramola@wits.ac.za (M.O. Daramola). Tropsch (HTFT) [6]. In addition, Co-based catalysts have high

http://dx.doi.org/10.1016/j.jece.2016.11.019
2213-3437/ã 2016 Elsevier Ltd. All rights reserved.
 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62 55

selectivity to C5+ linear paraffin hydrocarbons (HCs) and low used as purge gas and carrier gas (for the GC), respectively,
selectivity to methanation and water gas shift (WGS) reactions, hydrogen and air were purchased from Afrox (Pty) South Africa. All
making them preferable catalysts for low temperature F-T these materials were used as-supplied without any further pre-
synthesis to produce long chain linear hydrocarbons [7]. treatment.
During the gasification of biomass waste, such as corn cob,
wood biomass, to mention a few, 5–12% CO2 is produced. 2.2. Methods
Separation of the CO2 from the gasification products to obtain
the syngas required for F-T synthesis requires additional equip- The catalyst was synthesized using Incipient Wetness Impreg-
ment, translating into additional costs. The CO2 could be co-fed nation (IWI) with Cobalt nitrate solution as the precursor for cobalt
with the syngas stream during F-T synthesis, thereby by-passing using modified technique reported by [4,5]. The solution was
the gas cleaning stage. However, CO2 co-feeding could have impregnated onto the H-ZSM-5 support; the H-form of ammonium
beneficial or detrimental effect on the F-T process, especially on the ZSM-5 zeolite obtained after the calcination of the ammonium
reactivity of the catalyst. Therefore, effect of CO2 co-feeding on the form ZSM-5 at 550  C for 5 h. Cobalt supported on silica catalyst
performance of F-T catalyst especially Co-based catalyst needs to (Co/SiO2) was prepared and evaluated at the same conditions as
be understood. Against this background, effect of CO2 co-feeding the Co/HZSM-5 catayst for comparison. Scheme of the catalyst
on the performance of a bi-functional Co/H-ZSM-5 catalyst during preparation is depicted in Fig. 1.
F-T synthesis was investigated in this study. The bi-functional The techniques employed for the characterization of the
catalyst, Co/H-ZSM-5, combined F-T synthesis and hydrocracking catalyst were: (i) N2 physisorption at 77 K to determine the BET
together in the same catalyst. The catalyst was synthesized via the surface area, pore volume and average pore size of the synthesized
Incipient Wetness Impregnation (IWI) method and the perfor- catalyst; (ii) transmission electron microscopy (TEM) that pro-
mance of the catalyst was tested using a fixed-bed reactor that was duces 2-D images to understand the dispersion and the particle
operated as a plug flow reactor (PFR) under steady-state and size of the dispersed Co particles; (iii) scanning electron
isothermal condition. The IWI method was chosen to ensure microscopy equipped with energy-dispersive X-ray (SEM/EDX)
moderate Co particle loading and good dispersion within the that produces 3-D images to understand the morphology and
support. The H-ZSM-5 zeolite was used as the support because of elemental composition of the catalyst; (iv) X-ray diffraction (XRD)
its high acidity and good hydrocracking activity. to provide information about the crystallinity and phase of the
catalysts; (v) Thermogravimetric analysis (TGA) to understand the
2. Materials and methods thermal stability of the prepared catalyst; and (vi) H2-Temperature
Programmed Reduction (TPR) to obtain the temperature at which
2.1. Materials the prepared catalyst will be completely reducible from in-active
oxide of the metal to active metal phase.
Cobalt (II) nitrate hexahydrate (Co(NO3)26H2O) salt, which was The Nitrogen physisorption experiment was carried using on
used as the source of Cobalt (Co) metal, was purchased from about 0.2 g of the catalyst sample on a BET set-up (model:
Sigma-Aldrich (Pty) South Africa. Ammonium form of Zeolite Micromeritics TriStar 3000 V6.05 A). The samples were degassed at
Socony-Mobil-5 (ZSM-5), from which the H-form of Socony-Mobil- 150  C before nitrogen adsorption experiment was conducted.
5 (H-ZSM-5) that served as the support for the catalyst was Transmission Electron Microscopy analysis was performed using
obtained, was purchased from Zeolyst (USA). Two syngas cylinders JEOL 100S FEI (from Field Electron and Ion (FEI) Corporate USA).
(one with a mixture of H2, CO and N2 and the other with a mixture About 0.1 mg of the support sample and the catalyst sample were
of H2, CO, CO2) were purchased from Afrox (Pty) South Africa to ultra-sonicated in methanol for 5 min, before placing a drop of the
mimic the syngas composition expected from the gasification of solution in a copper grid. The Cu-grid with the solution on it was
biomass. Nitrogen (N2, 99.99% purity) and Argon (Ar, 99.9% purity) ventilated and then fed into the TEM chamber for the analysis. X-

Fig. 1. Scheme for the synthesis of the Co/H-ZSM-5 catalyst. The same procedure was followed for the preparation of the Co/SiO2 catalyst, except the step that involves the
conversion of ammonium form-ZSM-5 to H-ZSM-5.
56 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62

ray diffraction was conducted using Bruker D2 phaser equipped enhanced uniform heat conduction and distribution in the reactor.
with a Lynxeye detector at 30 kV and the XRD patterns were Before catalyst loading, the reactor was cleaned by blowing
obtained by using a Co-Ka radiation (l = 1.79 Å) in 2u region pressurized-air through the reactor using an air-gun. This was
between 10 and 90 with a step size of 0.027. The SEM-EDX was done every time when loading the reactor. The SS-balls were
conducted using a Carl-Ziess Sigma Field-Emission-Scanning- cleaned by using an alcohol then drying them in an oven for about
Electron-Microscope (FE-SEM) equipped with Oxford X-act Ener- 20–30 min. After cleaning, the quartz wool was then placed on the
gy-Dispersion-Spectroscopy (EDS) detector. The system was frit-disc and ensured that it covered the whole disc surface area.
initially purged with Nitrogen gas for about 10 min to remove About 0.5 g of catalyst was loaded into the reactor. The catalyst
air and any other unwanted gases before the analysis. Thermog- was diluted with SS balls to enhance uniform heat transfer within
ravimetric analysis was performed on a TGA (model: TA the reactor and a thermocouple (PT-100) was inserted in the
InstrumentV20.9 Build 20/SDT Q600 equipment using a Universal middle of the reactor and connected to a temperature controller
V4.7A data handling analysis software). About 10–11 mg of the and indicator to indicate and control the reaction temperature. The
sample was analyzed within a temperature range of 25–850  C loaded reactor was then connected to the F-T experimental rig and
with a heating rate of 5  C min
1 under 100 cm3 min
1 air flow. The then pressurized. The experiment proceeded with the catalyst
H2-TPR was conducted on AutoChem II set-up (model: 2920 V3.05) activated in situ at 30 cm3 STP/min of H2 gas flow and at 330  C for
from 20  C to 900  C at a temperature ramp of 8  C/min using 5% 17–20 h at 8 bar, followed by cooling to room temperature under
ratio of H2 =Ar gas mixture, at a flow of 30 ml/min. the same H2 flow. A rate of 2  C/min and
2  C/min were used for
The performance of the synthesized bi-functional catalyst (Co/ heating and the cooling steps, respectively. The reactor system was
H-ZSM-5) was evaluated by using a plug flow reactor (PFR) (see equipped also with two traps. The first trap was used to trap wax
Fig. 2). The reactor consisted of stainless steel (SS) tubing and (Heavy Hydrocarbons) while the second trap was used to capture
Swagelok fittings. The used reactor was of length 250 mm, internal oil and water.
diameter (ID) 16 mm and an outer diameter (OD) 20 mm. A frit-disc The feed and reaction products from the reactor were analyzed
with about 1–2 mm diameter holes was fitted inside the reactor to with two pre-calibrated online GCs (Hewlett-Packard 5890
support the catalyst packing and to prevent the flushing away of equipped a column (11.5 in. 1/8 in.) packed with Carboxen-
the catalyst particles out of the reactor when operated at a very 1000,60/80 and a TCD; and a Packard-Bell 433 equipped with a
high flow rate of the feed. Stainless steel (SS) balls of about 2–5 mm column (2 m  1/8 in.) packed with Porapaq Q, 80–100 mesh and a
in diameter each was used to dilute the catalyst particles for FID). A pre-calibrated offline GC (Varian 3700 equipped with a

Fig. 2. Experimental setup for performance evaluation of the catalyst.

 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62 57

column (30 m  0.5 mm  5 mm) packed with Zebron, ZB-1 and a The XRD patterns depicted in Fig. 3 show that the catalyst
FID was used for the offline analysis of the reaction products. The particles were crystalline and this finding is in agreement with
Flame Ionization Detector (FID) used Nitrogen as a courier gas literature [5]. According to Lu et al. [12], ZSM-5 is orthorhombic in
while the Thermal Conductivity Detector (TCD) used Argon. To feed shape. Orthorhombic crystal structures possess Simple Cubic (SC),
the reactor, a two-way valve was used to switch between hydrogen Body-Centred (BC) and Face Centred (FC) lattice symmetry [13]. In
and synthetic gas during catalyst reduction and catalyst testing, the XRD patterns depicted in Fig. 3, the presence of cobalt metal
respectively. Most of the gaseous products (light Hydrocarbons) particles within the Co/H-ZSM-5 is shown with “black-diamond
were analyzed with the online GCs. The gas lines to the GCs were markers” on the pattern, (topmost XRD pattern in Fig. 3). At 2-u:
heated to 100  C to ensure isothermicity of the reaction. The FID- 36.4, 42.99, 70.01, and 77.51, there were no peaks diffracted for H-
GC gas outlet was connected to a flowmeter which was used to ZSM-5 sample. However after IWI of the cobalt onto the H-ZSM-5,
measure the flowrate of the gases. The flowmeter was connected to the resulting catalyst (Co/H-ZSM-5) showed peaks diffracted at the
a vent to purge out the gases from the system. The GCs were preset aformentioned 2-u, indicating that cobalt particles were success-
to analyze for an hour at an oven temperature of 200  C and cool to fully incorporated in the zeolite support.
room temperature for another hour. The GC signals were received Fig. 4 shows the SEM/EDX results of the Co/H-ZSM-5 catalyst.
through a transmission box connected to a computer. The CO The shape of the Co/H-ZSM-5 catalyst particles, as inferred from
conversion, the selectivity of the hydrocarbons in moles and the SEM image (Fig. 4 (top)) is irregular-rounded due to metallic
percentage selectivity (in terms of the total selectivity of all the nature of Cobalt. The image also shows agglomerates of the catalyst
hydrocarbon (paraffin and olefin)) were obtained using Eq. (1), particles. The EDX spectrum (Fig. 4 (botton)) shows that Cobalt
Eq. (2) and Eq. (3), respectively, as reported by [8–10]: (with electronegativity of about 1.88) was detected at about 0.8 kV
8h

i9 and 7 kV. In addition, the EDX results reveal an estimated
< NCO
bef ore reaction
NCO
af ter reaction NNN2
bef orereaction = experimental Co loading of about 25% in the catalyst. This value
COconversion ¼
N2
af ter reaction

: NCO
bef ore reaction ; is two times the value reported by Sartipi et al. [5] for IWI of the
cobalt-solution on H-ZSM-5 support because a two-step impreg-
 100%
nation method suggested by Sartipi et al. [5] to enhance catalyst
ð1Þ loading in microporous materials such as H-ZSM-5, was used in the
synthesis.
   The TEM images depicted in Fig. 5 reveal that the Co particles
NHC;i trxn were successfully dispersed within the H-ZSM-5. Fig. 5(a) depicts
Selectivity ðmolesÞ ¼ DW
ð2Þ

rCo the H-ZSM-5 zeolite support before the impregnation; while
Fig. 5(b) shows that the as-prepared catalyst has Co3O4 embedded
within the support (as indicated with the arrows). By using the
Nparaf f in;i TEM images (Fig. 5) and the XRD patterns (Fig. 3) and following the
%Selectivity ¼  100 ð3Þ
Ntotal
paraf f in produced procedure of Guo et al. [14], the size of the metal grains could be
estimated to be about 8  3 nm. The TEM images confirmed that
Where NCO, is the number of mole of CO (mol), and NN2 is the
cobalt oxide particles were successfully impregnated onto the
number of mole of N2 (mol). NHC, is the number of mole of
supports and that the Co particles were uniformly dispersed within
hydrocarbon (HC) (mol); trxn, the reaction time (hour); DW, is the
the support. These observations agree well with the observations
catalyst weight (g); and i, is the component. Nparaffin, is the number
reported by Mohanty et al. [15] and Sartipi et al. [5] for their
of mole of paraffin (mol); Ntotal-paraffin, is the total number of mole
catalysts prepared using the same method.
of paraffin produced (mol).
The TEM images depicted in Fig. 6 show the morphology of the
traditional Co/SiO2 catalyst used for comparison in this study. The
3. Results and discussion
images show that the Co particles were well dispersed within the
support and the average size of the Co metal particles in the
3.1. Catalyst preparation and characterization
catalyst could be estimated to be about 15  0.5 nm. In addition,
the EDS analysis conducted on the catalyst sample revealed a
The N2 physisorption experiment revealed BET surface area,
theoretical Co loading of about 20%. For Co/SiO2, average particle
pore volume and average pore size of the catalyst to be 292 m2/g,
size of 16 nm and 8–35 nm were reported by Sartipi et al. [16], and
0.18 cm3/g and 2.83 nm, respectively (see Table 1). In addition, the
Ernest [17], respectively, for Co particles dispersed within their
results of the N2 physisorption, as presented in Table 1, show that
SiO2 supported catalysts. Therefore, the average particle size of the
the BET surface area, the pore volume and the pore size of the
Co particles obtained in this study for the synthesized Co/SiO2 is in
support (H-ZSM-5) were 333 m2/g, 0.23 cm3/g and 3.03 nm,
good agreement with literature. In addition, the BET surface area of
respectively. The reduction in the BET surface area (from
about 229 cm2/g and pore volume of about 0.9 cm3/g, obtained for
333 m2/g to 292 m2/g), pore volume (from 0.23 cm3/g to
the catalyst agree with the results reported by Sartipi [4].
0.18 cm3/g) and the average pore size (from 3.03 nm to 2.83 nm)
The TGA plot shown in Fig. 7 reveals that the Co/H-ZSM-5
of the H-ZSM-5 support after IWI indicates that Co particles were
catalyst was thermally stable up to 500  C. It can be observed from
successfully impregnated onto the support. From the value of the
Fig. 7 that at temperatures below 200  C, a rapid weight loss of
average pore size of the catalyst in Table 1, it could be concluded
about 5–7% was observed for the catalyst, the support and the
that the synthesized catalyst consisted of mesopores according to
ammonium-form ZSM-5, resulting in a steep slope of the thermal
the definition of mesoporous solids [11].

Table 1
BET surface area, pore volume and pore size of the catalyst and the support.

Materials BET surface area (m2/g) Pore volume (cm3/g) Average pore size (nm)
H-ZSM-5 333 0.23 3.03
Co/H-ZSM-5 292 0.18 2.83
58 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62

Fig. 3. XRD Patterns of the Support and the synthesized Co/H-ZSM-5 catalyst.

curve. The steady increase in weight loss from room temperature the reducibility of an heterogeneous catalyst. In this study, TPR was
to 100  C could be attributed to the loss in moisture content within conducted on the Co/H-ZSM-5 catalyst in the presence of H2. The
the catalyst and some hydration water. At a temperature between TPR was conducted from 50  C to 900  C in 5% of H2/Ar. Fig. 8
100  C and 150  C, the weight loss could be attributed to the depicts the H2-TPR profile of the Co/H-ZSM-5 catalyst. The onset
burning off of the loose organic matter within the catalyst. These for Co/H-ZSM-5 started at about 180  C and the reduction ended at
organic matter might have been introduced during the calcination about 650  C after peaking at A–D, resulting in a broader overall
of the as-prepared catalyst after the synthesis. At a temperature of peak. At peak labelled A in Fig. 9, tricobalt tetraoxide (Co3O4) was
150  C, a residual weight of 93 wt.%, 94 wt.% and 95 wt.% was reduced to Cobalt (II) oxide (CoO) at a temperature of about 272  C.
recorded for H-ZSM-5, ZSM-5 and Co/H-ZSM-5, respectively, The Cobalt (II) oxide was then reduced at 302  C to elemental
corresponding to a weight loss of 7 wt.% for H-ZSM-5, 6 wt.% for Cobalt (Co), which was then fully reduced to Co metal at 320  C (as
ZSM-5 and 5 wt.% for Co/H-ZSM-5. As the temperature increased shown with Peak C), thereby confirming the sequential conversion
between 150  C and 200  C the weight loss increased as well. of Co3O4 to CoO and CoO to Co [18]. At peak D the elemental Cobalt
However at temperatures above 200  C, a slow weight loss was binds itself to the zeolite support, forming Co/H-ZSM-5 at peak D.
observed. Hydrogen was highly consumed throughout the reduction process
After the calcination and during the catalyst synthesis, the (180–650  C). The large volume of hydrogen consumed at high
cobalt particles exist as cobalt oxide, and hence reduction is temperatures results in high oxidation states of the cobalt metal
required to obtain the active elemental Cobalt within the catalyst. and hence strong interaction of the cobalt metal with the support.
Temperature programmed reduction (TPR) experiment is usually This occurrence could be indicated by the overlapping of peak C
conducted in the presence of a reducing agent like H2 to investigate and D [19–21]. Wang et al. [19] have reported the high interaction
 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62 59

Fig. 4. Mophological and elemental composition of Co/H-ZSM-5. SEM image (top); EDX results (bottom).

of metal-support in their catalysts as well. In addition, the high the paths in the reduction of the Co3O4 to Co metal in the presence
reduction temperature of metals could be attributed to the high of H2 as the reducing agent could be proposed as follows:
acidity of the support materials, and this corroborates the results Co3O4 ! CoO ! Co.
obtained in this study since the zeolite support is acidic [22]. The
reduction profile also reveals the reduction of another oxide at 3.2. Effect of CO2-cofeeding on the performance of the Co/H-ZSM-5
500  C (see Peak E) which could be alumina because zeolite is an catalyst
aluminous-silicate material. Previous studies have shown that
alumina is reducible at a temperature 500  C [23–25]. Table 2 Table 3 shows the results obtained from the investigation of the
provides a summary of the TPR results obtained in this study, and effect of CO2 co-feeding on the performance of the bi-functional
Co/H-ZSM-5 catalyst. For comparison, the performance of the bi-
functional Co/H-ZSM-5 catalyst was compared with that of the
traditional Co/SiO2 catalyst (also synthesized for comparison) in
the presence of CO2 co-feeding under similar conditions (see
Table 3). The traditional catalyst was prepared via IWI using the
modified synthesis method reported by Guo et al. [14] with an
expected theoretical loading of about 20 wt.% of the Cobalt metal.
A CO conversion of 53.8% was obtained for Co/H-ZSM-5
catalyzed reaction at 250  C, at a Gas Hourly Space Velocity
(GHSV) of 1200 ml h
1 g
1, at a pressure of 15 bar and at a H2/CO
ratio of 2.5 without CO2. When a syngas mixture containing 5% of
CO2 was used, the CO conversion increased to 73.8%. Also the CO2
was hydrogenated to hydrocarbons with a conversion of 63.6% (see
Table 3). Recently, Yao et al. [26] reported CO conversion of about
20% in the presence of 5% of CO2 co-feeding. In addition, the
Fig. 5. TEM images of the synthesized bi-functional Co/H-ZSM-5 catalyst. (a) H- authors reported about 0.4% CO2 conversion at 200  C, at a pressure
ZSM-5 support and (b) Co/H-ZSM-5.
60 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62

it could be speculated that, in the presence of CO2 co-feeding the

presence of CO2 enhanced the rate at which CO was converted to
products, thereby resulting in a higher CO conversion than the
conversion in the absence of CO2 co-feeding.
Fig. 9 depicts the products of the Co/H-ZSM-5 catalyzed
reaction for a syngas mixture of H2/CO ratio of 2.5 with 0% CO2 and
with 5% CO2 in the feed. It can be observed from Fig. 9, that the
presence of CO2 resulted in a minor or no change in paraffin
selectivity. In the presence of CO2 co-feeding, the selectivity of the
bi-functional Co/H-ZSM-5 catalyst was 53% to methane, 34% to C2,
<4% to C4–C5 and <10% to C6+. These values are more or less the
same as the selectivity to these hydrocarbons (HCs) in the absence
of CO2 co-feeding. Yao et al. [26] reported that paraffin selectivity,
Fig. 6. TEM images of the prepared commercial catalyst. (a) SiO2 support and (b) in particular methane, increased with a decrease in CO partial
Co/SiO2.
pressure. In this study the CO partial pressure was kept constant
while the partial pressure of N2 was altered with the introduction
of 20 bar, at a flowrate of 3600 GHSV and at a H2/CO ratio of 2 for
of CO2 gas (changed from 15% to 10% while the CO2 was 5%).
the Co catalyzed reaction. However in the absence of CO2 co-
Therefore, the results obtained in this study are in agreement with
feeding, Yao et al. [26] reported a CO conversion of 14.6%, a value
the results of Yao et al. with regard to the selectivity to paraffin, in
lower than the value obtained in this study (see Table 3). Therefore,
particular methane [26]. In addition, it was observed that CO2 co-

Fig. 7. Thermogravimetric analysis of the prepared Co/H-ZSM-5 catalyst and the support.

Fig. 8. Temperature programme reduction (TPR) profile of the synthesized Co/H-ZSM-5 catalyst.
 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62 61

Fig. 9. Change in Olefin & Paraffin Selectivity in the absence of CO2 (a) and in presence of CO2 (b). Experimental conditions: H2/CO: 2.5; Temperature: 250  C; Pressure: 15 bar;
GHSV: 1200 ml h
1 g
1; CO2: 5%.

production of paraffin remained the same throughout the reaction

Table 2
duration for Co/H-ZSM-5 catalyzed F-T reaction.
Summary of the TPR of the Co/H-ZSM-5 catalyst.
From Table 3, CO conversion of 73.8% and 70.4% were obtained
Peak Reduction Temperature ( C) Phase for Co/H-ZSM-5 and Co/SiO2 catalysts, respectively, in the presence
A 272 Co3O4 of CO2 co-feeding, making the conversion from the Co/H-ZSM-5
B 302 CoO catalyst to be about 3% higher than that of the traditional Co/SiO2
C 320 Co catalyst. Fig. 10 shows that the selectivity of the bifunctional Co/H-
D 327 Co/H-ZSM-5
E 617 Al2O3
ZSM-5 catalyst to methane was 53% (see Fig. 10(a)). In the presence
of CO2 co-feeding, the CO2 conversion of the bi-functional catalyst
was 63.6% and the traditional Co/SiO2 catalyst displayed CO2
conversion of 75.3%. Most of the CO2 that reacted contributed
feeding influenced the selectivity of the catalyst to olefins as well.
mainly to olefin production; hence a high selectivity of 76% to C6+
The selectivity of C3 olefin decreased from 43% to 35% while C4
olefin was obtained for the Co/SiO2 catalyzed reaction (see
olefin increased from 25% to 32%. Furthermore, the C6+ olefin
Fig. 10(b)), contributing to the CO2 conversion of 75.3%. On the
increased from 9% to 15% when the syngas was co-fed with CO2.
other hand, CO2 reacted when Co/H-ZSM-5 catalyst was used,
However, Yao et al. reported that if the ratio of CO2 to CO in the CO2
yielded almost uniform distribution of C2–C4 olefins.
co-feeding is more less than 50% for a Cobalt catalyzed reaction, the
Comparing the selectivity of the catalysts, Fig. 10(a), (b) shows
CO2 will behave as an inert gas and takes no part in the reaction
that the presence of CO2 in the feed increased the selectivity of Co/
[26]. In contrary to this finding, results from this study show that in
H-ZSM-5 catalyst to olefin with Co/H-ZSM-5 producing more
the presence of CO2 and at a CO2/CO ratio is more less than 50%,
middle olefin hydrocarbons. In addition, the selectivity of Co/SiO2
hydrogenation of CO2 occurred. Therefore, the increase in the
catalyst to lower and middle olefin increased in the presence of CO2
selectivity of the catalyst to higher olefin from 9% to 15% in the
co-feeding, with a slight increase in the paraffin as well. Therefore,
presence of CO2 could be attributed to the speculation that when
for Co/SiO2 catalyzed reaction in the presence of 5% CO2, both C1 to
most of the CO has been hydrogenated to methane and ethane, and
C6+ hydrocarbon selectivity increased with the catalyst displaying
the remaining CO is not enough to produce C3+ paraffin, the
more selective to C6+ olefin. On the contrary, for Co/H-ZSM-5
reaction opted for CO2 to react, thereby resulting in more olefins
catalyzed F-T reaction, an increase in selectivity to olefin resulted
since the CO2 reaction favours olefin production. In addition, Yao
in a decrease or no change in selectivity to paraffin hydrocarbons.
et al. [26] reported a selectivity of 15% to methane in the presence
The CO2 co-feeding only influenced the selectivity of the catalysts
of CO2, but the authors did not separate the hydrocarbons (HCs)
to olefins. Consequently, the selectivity of the Co/H-ZSM-5 catalyst
formed during the reaction, but grouped them together as C2+
to olefins increased while its selectivity to paraffin remained the
hydrocarbons for which 85% selectivity was reported.
same or slightly decreased. Water formation (as a result of water-
It could be concluded that CO2 co-feeding enhances the catalyst
gas-reaction in the process) increased from 0.5 to 1.5 g in the
selectivity to intermediate olefin carbon number and also the CO
absence of CO2 co-feeding to about 2–3.5 g (by weight) for Co/SiO2
conversion. The enhanced CO conversion in the presence of CO2 is
catalyzed reaction in the presence of CO2 co-feeding, contributing
present can be attributed to olefin formation since no change in the
to the higher CO2 conversion observed for Co/SiO2 catalyst. The
paraffin selectivity was observed. Therefore, under similar con-
findings from this study agree well with the results of Dorner et al.
ditions, a syngas with or without CO2 content will offer the same
[27], where water production due to Water-Gas-Shift (WGS)
paraffin selectivity with a higher CO conversion. However, when
reaction in the presence of CO2 was reported.
CO2 was introduced, more olefins were produced but the

Table 3
Effect of CO2 co-feeding on the rate of conversion Co/H-ZSM-5 catalyzed F-T synthesis.

Catalyst H2/CO (%) GHSV (ml h
1 g
1) Temperature ( C) Pressure (bar) % CO Conversion % CO2 Conversion
CO2
Co/H-ZSM-5 2.5 0 1200 250 15 53.7 –
Co/H-ZSM-5 2.5 5 1200 250 15 73.8 63.6
Co/SiO2 2.5 5 1200 250 15 70.4 75.3
62 M.O. Daramola et al. / Journal of Environmental Chemical Engineering 5 (2017) 54–62

Fig. 10. Olefin and Selectivity in the presence of C O2. (a) for Co/H-ZSM-5, and (b) for Co/SiO2 catalyst. Experimental conditions: H2/CO: 2.5; Temperature: 250  C; Pressure:
15 bar; GHSV: 1200 ml h
1 g
1; CO2: 5%.

4. Conclusions [6] L.P. Dancuart, A.P. Steynberg, Fischer-Tropsch based GTL technology: a new
process, Stud. Surf. Sci. Catal. 163 (2007) 379–399.
[7] A.Y. Khodakov, W. Chu, P. Fongarland, Advances in the development of novel
Effect of CO2 co-feeding on the performance of a bifunctional cobalt Fischer-Tropsch catalysts for synthesis of long-chain hydrocarbons and
Co/H-ZSM-5 catalyst and the traditional Co/SiO2 catalyst during F-T clean fuels, Chem. Rev. 107 (5) (2007) 1692–1744.
synthesis was evaluated in this study. The effect of CO2 co-feeding [8] J. Price, An Investigation into Novel Bimetallic Catalysis for Use in the FT
Reaction, University of the Witwatersrand, Johannesburg, South Africa, 1994
was performed under the process conditions investigated in this (PhD Thesis).
study. Conversion and selectivity results obtained when CO2 was [9] A.M. Motchelaho, Iron and Cobalt Catalyst Supported on Carbon Nanotubes for
present in feed were compared to the results obtained in the Use in the Fischer-Tropsch Synthesis, University of the Witwatersrand,
Johannesburg, South Africa, 2011 (PhD Thesis).
absence of CO2. Also the performance of the bi-functional catalyst
[10] M. Moyo, Cobalt and Iron Supported on Carbon Spheres Catalysts for Fischer-
was compared to that of a traditional catalyst (also prepared in this Tropsch Synthesis, University of the Witwatersrand, Johannesburg, South
study for comparison) in terms of, which products are favoured as Africa, 2012 (PhD Thesis).
[11] M. Che, J. Vedriné, Characterization of Solid Materials and Heterogeneous
well as how the conversion of CO was affected. The co-fed CO2 was
Catalyst-from Structure to Surface Reactivity, in: M. Che, J. Vedriné (Eds.), 2nd
converted to hydrocarbons and a CO2 conversion of 63% was ed., Wiley-VCH Verlag GmbH & CO. KGaA, Weinheim, Germany, 2012.
obtained for the bi-functional catalyst. In terms of selectivity, both [12] R. Lu, H. Hejin Tangbo, Q. Wang, S. Xiang, Properties and characterization of
syngas (with CO2 co-feeding and without CO2 co-feeding) gave modified HZSM-5 zeolites, J. Nat. Gas Chem. 12 (2016) 56–62.
[13] E. Smith, Basic Physical Chemistry – The Route to Understanding, Imperial
similar product selectivity to paraffins. The bi-functional catalyst College Press, London, 2012.
was 53% selective to methane, 34% to C2, <4% to C3–C5 and less [14] C. Guo, Y. Wu, J. Zhan, Study on bimodal mesoporous Co/SiO2 catalysts for the
than 10% selective to C6+ HCs. It was observed that co-feeding CO2 Fischer-Tropsch synthesis, React. Kinet. Mech. Catal. 109 (2) (2013) 497–508.
[15] P. Mohanty, K.K. Pant, J. Parikh, D.K. Sharma, Liquid fuel production from
affected the olefin selectivity as well, for most olefin HCs selectivity syngas using bifunctional CuO-CoO-Cr 2O3 catalyst mixed with MFI zeolite,
increased. However, the C6+ increased from 9% to 15% when the Fuel Process. Technol. 92 (3) (2011) 600–608.
syngas was co-fed with CO2. It is noteworthy to mention that the [16] S. Sartipi, M. Alberts, V.P. Santos, M. Nasalevich, J. Gascon, F. Kapteijn, Insights
into the catalytic performance of mesoporous H-ZSM-5-supported cobalt in
obsereved performance might have been influenced by the type of Fischer-Tropsch Synthesis, ChemCatChem 6 (2014) 142–156.
support used for the catalysts. Such influence was not considered [17] B. Ernst, A. Bensaddik, L. Hilaire, P. Chaumette, A. Kiennemann, Study on a
in the analysis of the catalyst performance. Future work should cobalt silica catalyst during reduction and Fischer-Tropsch reaction: in situ
EXAFS compared to XPS and XRD, Catal. Today 39 (1998) 329–341.
consider the investigation of the influence of catalyst support and [18] X. Xie, W. Shen, Morphology control of cobalt oxidenanocrystals for promoting
also the long-term operational stability of the bi-functional their catalytic performance, Nanoscale 1 (2009) 50–60.
catayst. [19] W.J. Wang, H.Y. Lin, Y.W. Chen, Carbon monoxide hydrogenation on cobalt/
zeolite catalysts, J. Porous Mater. 12 (1) (2005) 5–12.
[20] J.H. Martens, H.F.J. van’T Blik, R. Prins, Characterization of supported cobalt and
Acknowledgements cobalt-rhodium catalysts II. Temperature-programmed reduction (TPR) and
oxidation (TPO) of Co/TiO2 and Co -Rh/TiO2, J. Catal. 97 (1) (1986) 200–209.
The authors acknowledge the financial support of the National [21] S. Galvagno, C. Crisafulli, R. Maggiore, G.R. Tauszik, A. Giannetto, Temperature-
programmed reduction: metal-support interaction on supported
Research Foundation (NRF) South Africa for KM. Also, MOD monometallic Ru and Cu catalysts, J. Therm. Anal. 30 (1985) 611–618.
appreciates the financial assistance of iNkwazi through NWU/Wits [22] L. Rodriquez-Gonzalez, F. Hermes, M. Bertmer, E. Rodriquez-Castellon, A.
research collaboration. Jimenez-Lopez, The acid properties of H-ZSM-5 as studied by NH3-TPD and
27
Al-MAS-NMR spectroscopy, Appl. Catal. A: Gen. 328 (2) (2007) 174–182.
[23] C. Li, Y.W. Chen, Temperature-programmed-reduction studies of nickel oxide/
References alumina catalysts: effects of the preparation method, Thermochim. Acta 256
(2) (1995) 457–465.
[1] A.R. Phale, Environmental Impact and Waste Management of Used Tyres in The [24] D. Vermaire, The preparation of WO3/TiO2 and WO3/Al2O3 and
RSA, (2005) (Accessed 13 June 2016). characterization by temperature-programmed reduction, J. Catal. 116 (2)
[2] R.A. Dagle, J.A. Lizarazo-Adarme, V.L. Dagle, M.J. Gray, J.F. White, D.L. King, D.R. (1989) 309–317.
Palo, Syngas conversion to gasoline-range hydrocarbons over Pd/ZnO/Al 2O3 [25] M.C. Álvarez-Galván, V.A. de la Peña O’shea, J.L.G. Fierro, P.L. Arias, Alumina-
and ZSM-5 composite catalyst system, Fuel Process. Technol. 123 (2014) 65–74. supported manganese- and manganese-palladium oxide catalysts for VOCs
[3] S. Ali, N. Mohd Zabidi, D. Subbarao, Correlation between Fischer-Tropsch combustion, Catal. Commun. 4 (5) (2003) 223–228.
catalytic activity and composition of catalysts, Chem. Cent. J. 5 (1) (2011) 1–8. [26] Y. Yao, D. Hildebrandt, D. Glasser, Fischer-Tropsch synthesis using H2/CO/CO2
[4] S. Sartipi, J.E. Dijk, J. Van Gascon, F. Kapteijn, Toward bifunctional catalysts for syngas mixtures over a cobalt catalyst, Ind. Eng. Chem. Res. 49 (2010) 11061–
the direct conversion of syngas to gasoline range hydrocarbons: H-ZSM-5 11066.
coated Co versus H-ZSM-5 supported Co, Appl. Catal. A: Gen. 456 (2013) 11–22. [27] R. Dorner, D. Hardy, F. Williams, Influence of gas feed composition and
[5] S. Sartipi, K. Parashar, M. Makkee, J. Gascon, F. Kapteijn, Breaking the Fischer– pressure on the catalytic conversion of CO2 to hydrocarbons using a traditional
Tropsch synthesis selectivity: direct conversion of syngas to gasoline over cobalt-based Fischer-Tropsch catalyst, Energy Fuels 23 (2009) 4190–4195.
hierarchical Co/H-ZSM-5 catalysts, Catal. Sci. Technol. 3 (3) (2013) (p. 572).