The effect of cobalt and zinc precursors on Co (10%)/Zn (x%)/TiO 2 (x=0, 5) Fischer-Tropsch catalysts

Article abstract of DOI:10.1016/j.molcata.2004.09.004

Co/Zn/TiO₂ catalysts prepared from zinc/cobalt nitrate produced larger Co particles and more wax (>C₁₆) when compared to the other catalyst combinations. A series of Co (10%)/Zn (x%)/TiO₂ (x = 0, 5) catalysts were prepared from different nitrate and acetate precursors. TPR and chemisorption techniques revealed that a mixture of cobalt precursors (cobalt acetate and cobalt nitrate) on titania were more easily reduced when compared to Co (10%)/TiO₂ catalysts prepared from either cobalt acetate or cobalt nitrate alone. By contrast, after addition of zinc, catalysts prepared from zinc acetate and cobalt nitrate had the most highly dispersed cobalt species when compared to catalysts prepared from the other combinations of nitrate and acetate precursors of zinc and cobalt. Mixed precursors of zinc and cobalt were also more active and had higher CO conversion in Fischer-Tropsch synthesis (FTS) when compared to catalysts prepared from: (i) both nitrate precursors or (ii) both acetate precursors of zinc and cobalt. However, the catalyst prepared from zinc nitrate and cobalt nitrate produced more wax (>C₁₆) when compared to the other catalysts studied (50% versus 12-32% wax). Data suggest that larger Co particles correlate with wax production as well as reduced CO conversion. CO chemisorption data correlated better with the FT activity data than did Co XPS data.

Full text of DOI:10.1016/j.molcata.2004.09.004

Journal of Molecular Catalysis A: Chemical 225 (2005) 137–142
The effect of cobalt and zinc precursors on
Co (10%)/Zn (x%)/TiO (x = 0, 5) Fischer–Tropsch catalysts
2
a,b
a,∗
Nobuntu N. Madikizela-Mnqanqeni , Neil J. Coville
a
Molecular Sciences Institute, School of Chemistry, University of the Witwatersrand, Johannesburg, 2050, South Africa
b
Sastech R&D, P.O. Box 1, Sasolburg, 9570, South Africa
Received 10 July 2004; accepted 1 September 2004
Available online 8 October 2004
Abstract
A series of Co (10%)/Zn (x%)/TiO
techniques revealed that a mixture of cobalt precursors (cobalt acetate and cobalt nitrate) on titania were more easily reduced when compared to
Co (10%)/TiO catalysts prepared from either cobalt acetate or cobalt nitrate alone. By contrast, after addition of zinc, catalysts prepared from
2
(x = 0, 5) catalysts were prepared from different nitrate and acetate precursors. TPR and chemisorption
2
zinc acetate and cobalt nitrate had the most highly dispersed cobalt species when compared to catalysts prepared from the other combinations
of nitrate and acetate precursors of zinc and cobalt.
Mixed precursors of zinc and cobalt were also more active and had higher CO conversion in Fischer–Tropsch synthesis (FTS) when
compared to catalysts prepared from: (i) both nitrate precursors or (ii) both acetate precursors of zinc and cobalt. However, the catalyst
prepared from zinc nitrate and cobalt nitrate produced more wax (>C16) when compared to the other catalysts studied (50% versus 12–32%
wax). Data suggest that larger Co particles correlate with wax production as well as reduced CO conversion. CO chemisorption data correlated
better with the FT activity data than did Co XPS data.
©
2004 Elsevier B.V. All rights reserved.
Keywords: Fischer–Tropsch catalysis; Zinc; Cobalt; Titania; XPS
1
. Introduction
result in the formation of small particles that were not easy
to reduce and were not active towards FTS. However, nitrate
precursors formed larger particles that reduced at lower tem-
peratures and were active in FTS.
Similar results were obtained when different cobalt pre-
cursors (nitrate, chloride, acetate and acetylacetonate) were
supported on silica SBA-15 and MCM-41 [2,4]. As with the
study on alumina, superior performance towards FT synthesis
was obtained for the cobalt nitrate precursor [2].
Supported cobalt catalysts have been extensively used
in Fischer–Tropsch synthesis (FTS) when high molecular
weight products are desired [1–3]. The particle size of the
catalyst is one of the contributing factors that can affect the
selectivity as well as activity of supported cobalt catalysts.
One approach to the control of the ultimate particle size of
cobalt catalysts is by varying the cobalt precursors. Surpris-
ingly, only limited studies have been reported in which this
variable has been systematically explored.
Kraum and Baerns [5] performed a similar study on
cobalt catalysts supported on titania, zirconia and ce-
ria. Temperature-programmed reduction (TPR) studies con-
ducted on the Co/TiO2 catalysts prepared from a nitrate pre-
cursor, indicated the presence of one shoulder and one peak
Van de Loosdrecht et al. [3] studied the influence of or-
ganic and nitrate precursors on the particle size of alumina-
supported cobalt catalysts. Organic precursors were found to
◦
at 328 and 395 C in the TPR profiles, respectively, assigned
∗
3 4
to the stepwise reduction of Co O to Co [5]. However, when
Corresponding author. Tel.: +27 11 717 6738; fax: +27 11 717 6749.
E-mail address: ncoville@aurum.chem.wits.ac.za (N.J. Coville).
titania supported cobalt catalysts were prepared from organic
1
381-1169/$ – see front matter © 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2004.09.004

1
38
N.N. Madikizela-Mnqanqeni, N.J. Coville / Journal of Molecular Catalysis A: Chemical 225(200 5) 137–142
◦
precursors (cobalt acetate and cobalt oxalate), only one peak
was observed in the TPR profiles. The peak maximum was
at 387 C for the cobalt acetate precursor and at 383 C for
the cobalt oxalate precursor [5].
An increase in cobalt dispersion from 6.1 to 7.8% was also
observed when cobalt acetate as compared to cobalt nitrate
was added to titania [5]. Also, catalysts prepared from cobalt
acetate gave higher CO conversions and lower chain growth
when compared to catalysts prepared from nitrate precursors
catalysts were calcined at 300 C for 19 h after the final im-
pregnation step.
◦
◦
2
2
.2. Catalyst characterization
.2.1. X-ray photoelectron spectroscopy (XPS)
The percentage concentration of different species on the
surface of the material was determined, using XPS, recorded
with a Quantum 2000 Scanning ESCA microprobe (Physical
Electronics Company). Monochromised Al K␣ (1486.7 eV)
was used as the X-ray source.
(CO conversion of 26.2% versus 14.7%; alpha value changed
from 0.74 to 0.84).
An extensive TPR study of Co/SiO2 catalysts prepared
from different cobalt precursors (chloride, sulfate, nitrate, ac-
etate) also revealed that the counter-ion dramatically affected
the TPR profile [6].
Li and Coville [7] investigated the effect of cobalt nitrate,
cobalt acetate and cobalt chloride on the reducibility and cat-
alytic performance of boron modified cobalt FTS catalysts
supported on titania. Catalysts prepared using nitrate and ac-
etate precursors were found to reduce in two steps, whereas
catalysts prepared, using chloride precursors, reduced in one
step. Calcined acetate and nitrate catalysts showed higher ac-
tivity towards FTS whilst chloride catalysts exhibited lower
activity due to poisoning by chloride ions.
Different surface atoms observed include Co, Zn (where
applicable), Ti, O and C. Carbon was observed on all the sam-
ples and originated from the instrument background and not
from the catalysts. From the work done by Alstrup and co-
workers [11] the binding energies of the Co 2p3/2 peaks of the
clean cobalt metal are located at 778.5 and 793.6 eV. How-
ever, the Co 2p3/2 main peak of the oxidic cobalt is expected
at 780.3 eV with the Co 2p3/2–2p1/2 spin splitting equal to
15.1 eV [12].
2.2.2. TPR
TPR experiments were performed on the Co/Zn (x)/TiO2
(x = 0, 1 and 5) catalysts. Sieved (1180–850 ␮m), calcined
The effect of mixing organic and nitrate precursors to pre-
pare Co/SiO2 catalysts using an impregnation method has
also been reported [8]. Mixed precursors displayed higher
activity towards FTS when compared to single-cobalt source
precursors. Large Co metal particles, resulting from nitrate
precursors, were proposed to assist in the reduction of small,
difficult to reduce particles that were obtained from organic
precursors through a hydrogen spillover mechanism [8].
These results have prompted us to investigate the effect of:
samples (100–200 mg) were placed in a quartz reactor and
degassed with nitrogen at 150 C. A reduction gas con-
◦
taining 5% hydrogen in argon was passed over the sam-
ple at a flow rate of 50 ml/min, while the temperature
◦
was linearly increased at a rate of 10 C/min from 27
to 950 C.
◦
2.2.3. Chemisorption analysis
H2 chemisorption was performed in a Micromeritics
(i) different single cobalt sources, (ii) mixed cobalt sources
ASAP 2010 instrument. Calcined samples were first reduced
and(iii)differentzincandcobaltsourcesonthepropertiesand
FT activity of a series of Co (10%)/Zn (x%)/TiO2 catalysts
◦
at a temperature of 250 C with 100% H2 for 16 h and then
evacuated under helium gas to remove all physisorbed hydro-
gen. Adsorption isotherms were extrapolated to zero pressure
to obtain chemisorption uptake. The following equation was
used to calculate dispersion:
(x = 0, 5). This publication continues on an investigation of
the effect of zinc on Co/TiO2 catalysts from our research
group as described in earlier papers [9,10].
ꢀVm/V
ꢁ
mol
D% =
Fs(100), (100)
2
. Experimental
W%/Wa
where Vm is the total volume of hydrogen chemisorbed, Vmol
the hydrogen molar volume, W% the percentage of cobalt by
2
.1. Catalyst preparation
weight, Wa the cobalt atomic weight, and Fs is the stoichiom-
etry factor (Fs = 2 for hydrogen).
Co (10%)/TiO2 catalysts were prepared by the incipient
wetness method, using TiO2 Degussa (P25) as the support.
Cobalt nitrate and cobalt acetate were either impregnated
separately to form a 10 wt.% Co/TiO2 catalyst or were co-
impregnated to form a nitrate/acetate (N/A) mixture in a 1:1
molarratioontheTiO2. Inanotherseriesofexperiments, vari-
ouscombinationsofcobaltandzincnitratesandacetateswere
used in the preparation of catalysts. In all the Co (10%)/Zn
2.2.4. pH measurement
The pH of different precursor solutions was determined
prior to impregnation. Before impregnation to the support
the cobalt ion concentration of the solutions from different
precursors was 1.42 mol/L and the zinc ion concentrations
were 0.638 mol/L. pH measurements were performed, using
a Yokogawa Electric Corporation pH meter, model PH 8211-
E. The pH meter was first calibrated, using buffer solutions
(
x%)/TiO2 (x = 1, 5) catalysts that were prepared, the zinc was
◦
always impregnated before the cobalt with drying (120 C for
1
9 h) in-between the impregnation and calcination steps. All

N.N. Madikizela-Mnqanqeni, N.J. Coville / Journal of Molecular Catalysis A: Chemical 225(200 5) 137–142
139
with pH 4, 7 and 9 and all measurements were carried out at
room temperature.
counter-ion. ThecounterioninfluencesthepHofthesolu-
tion and the bigger Zn particles formed from the nitrate
solution are associated with the pH value being far re-
moved from the zeta potential for TiO2 (ca. 6–6.5)[16].
The closer the pH value of a precursor solution to the to
thezetapotentialofDegussatitania, thelargertheamount
of impregnated metal dispersed on the surface.
2
.3. CO hydrogenation
A system of three stainless steel fixed bed reactors was
used to evaluate the activities and selectivities of catalysts
◦
(
b) Co/TiO2: XPS data for Co/TiO2 samples prepared from
towards FTS [13]. Catalysts were calcined at 300 C for
different Co precursors were recorded. A slight increase
1
9 h in an oven prior to loading in the reactors. Catalyst
(approximately 20%) in the amount of cobalt species dis-
(
2.0 ml–1.5 g) was loaded into each reactor. Catalysts were
◦
persed on the titania support was observed when cobalt
acetate precursor salts were used in the catalyst prepa-
ration instead of cobalt nitrate salts (Table 1). Similar
results were obtained by Kraum and Baerns [5]. The use
of cobalt acetate as a precursor thus resulted in the for-
mation of a more highly dispersed cobalt surface species.
A significant increase in the surface cobalt species was
also observed when mixed precursor salts of cobalt were
co-impregnated on the titania support. The amount of
cobalt increased from 2.07 to 2.98% (Table 1). These
results show that mixed acetate and nitrate precursors of
cobalt improves the dispersion of cobalt on TiO2. Sun et
al. observed similar results when mixed cobalt precursor
salts were employed in the preparation of Co/SiO2 [8].
Correlation was also established when considering the
pH of different solutions prior to impregnation versus the
isoelectric point of Degussa titania. The closer the pH of
solution to the isoelectric point of the support, the more
the cobalt is dispersed on the support (Tables 1 and 2).
c) Co (10%)/Zn (5%)/TiO2catalysts: Table 1 shows the XPS
data recorded on Co (10%)/Zn (5%)/TiO2 samples in
which the source of cobalt and zinc were varied. Cobalt
was always impregnated after the addition of zinc to the
titania support. In this set of experiments when the zinc
and cobalt precursors were varied, the amount of cobalt
dispersed on the surface (3.08–3.66%) was found to be
less dependent on the zinc counterion than on the cobalt
counterion used (3.08 and 3.12 for acetate and 3.66 and
3.49 for nitrate ions; Table 1).
then reduced in situ under a flow of pure hydrogen at 250 C
for 16 h at a GHSV of 500 h . After reduction, the reactor
temperatures were decreased to 220 C followed by the in-
−
1
◦
troduction of synthesis gas (10% Ar, 60% H2, 30% CO) at a
−
1
pressure of 8 bar and a GHSV of 400 h . No additional gases
were co-fed into the synthesis gas. Flow rates were controlled
by Brookes flow meters and needle valves. The catalysts were
allowed to stabilise over a period of 120 h and the mass bal-
ance was then performed over a period of 100–120 h.
The method used for mass balance calculations is similar
to the one used by Nijis and Jacobs [14]. The calculations
are based on carbon. The amount of carbon entering the re-
actor is equal to the amount of carbon reacted to form prod-
ucts plus unreacted carbon leaving the reactor, i.e. %mass
balance = (A + B)/C × 100, where A = mol carbon unreacted,
B = mol carbon reacted and C = mol carbon entering the re-
actor.
A calibration gas containing hydrogen, carbon monoxide,
carbon dioxide, methane, ethane and ethylene was used to
calibrate the TCD and FID data. Response factors obtained
were converted to molar compositions for different compo-
nents, using the method described by Scanlon and Wills [15].
(
3
. Results and discussion
3
.1. XPS analysis
(
a) Zn/TiO2: XPS data were collected on Zn/TiO2 samples
prepared from both zinc acetate and zinc nitrate. Zinc
acetate (pH 5.68) gives more surface zinc than does zinc
nitrate (pH 2.76; an increase of about 20%), indicating
that the size of the zinc particles is affected by the zinc
Nevertheless, the catalysts prepared from cobalt ni-
trate and zinc acetate precursor had the highest amount
of cobalt dispersed on the surface when compared to
other counterions. These results show that the amount
Table 1
XPS results showing the effect of changing cobalt and zinc precursors on Co (10%)/Zn (5%)/TiO2 catalysts
Cobalt and zinc precursor salts
%Ti 2p
%Co 2p3/2
%Zn 2p3/2
%Co/%Ti
%Co/%Zn
%Co/(%Ti + %Zn)
Co/TiO2
Cobalt acetate
Cobalt acetate + cobalt nitrate (1:1)
Cobalt nitrate
18.6
18.3
24.2
2.59
2.98
2.07
0
0
0
0.139
0.163
0.086
–
–
–
13.93
16.28
8.56
Co/Zn (5%)/TiO2
Cobalt acetate and zinc acetate
Cobalt acetate and zinc nitrate
Cobalt nitrate and zinc acetate
Cobalt nitrate and zinc nitrate
19.9
18.5
18.1
25.8
3.08
3.12
3.66
3.49
2.3
0.155
0.169
0.202
0.135
1.34
2.04
3.13
6.35
13.87
15.57
18.99
13.26
1.54
1.17
0.55

1
40
N.N. Madikizela-Mnqanqeni, N.J. Coville / Journal of Molecular Catalysis A: Chemical 225(200 5) 137–142
Table 2
Cobalt Chemisorption measurements
Cobalt and zinc precursor salts
2
2
pH data
Cobalt dispersion (%)
Metallic surface area (m /g) sample
Metallic surface area (m /g) metal
Co/TiO2
Cobalt acetate
Cobalt acetate + cobalt nitrate (1:1)
Cobalt nitrate
6.69
6.41
2.99
1.82
2.21
1.35
1.19
1.43
0.91
11.92
14.43
9.15
Co/Zn (5%)/TiO2
Cobalt acetate + zinc acetate
Cobalt acetate + zinc nitrate
Cobalt nitrate + zinc acetate
Cobalt nitrate + zinc nitrate
2.23
2.31
2.62
1.98
1.48
1.59
1.73
1.33
14.84
15.96
17.34
13.39
of cobalt in the surface is affected by more than one
factor.
show any significant effect on the reducibility of cobalt.
ThedataaresimilartothatreportedbyKraumandBaerns
[5], but with different absolute values due to the different
reaction/measurement conditions used. Sun et al. [8] also
The presence of zinc in the catalyst is anticipated to im-
prove the dispersion of cobalt by forming a ‘sandwich
layer’ between the cobalt and the support [10]. In addi-
tion, cobalt dispersion is affected by the cobalt counterion
used in the preparation of catalyst as well as the pH of the
solution prior to impregnation. Cobalt nitrate precursors
are expected to form bigger particles when compared
to their organic counterparts. Further, formation of big
cobalt particles on top of well dispersed zinc, is expected
to improve the stablity of cobalt particle.
Changes were also observed in the amount of zinc de-
tected on the surface due to the changes in the zinc
and cobalt precursor salts. In particular: (i) zinc acetate
yielded more surface zinc for similar cobalt salts than did
zinc nitrate and (ii) cobalt nitrate gave a lower yield of
surface zinc with cobalt nitrate.
3
+
obtained similar results for the reducibility of Co to
2+
Co when working with Co/SiO2 catalysts. However, a
vast difference in the second cobalt reduction step was
observed in the work reported by Sun et al., which was
attributed to a metal support interaction [8].
When the nitrate precursor was combined with the ac-
etateprecursormoreeasilyreduciblespeciesareobtained
(Fig. 1b). These results correlate with the results obtained
by Sun et al. on Co/SiO2 catalysts [8]. The ease in the
reducibility of cobalt prepared from mixed precursors
was attributed to a spillover effect. Bigger particles re-
sulting from nitrate precursors were thought to assist in
the reduction of smaller particles prepared from acetate
precursors [8].
The observed increase in cobalt dispersion when cobalt
acetate is used as a precursor, is associated with smaller
cobalt particles, as found for Co on SiO2 [8]. This is also
consistent with smaller zinc particles generated from ac-
etate rather than nitrate. Cobalt oxide particles are ex-
pected to be more stable and zinc is also known to im-
prove the stability of cobalt [12,17–20].
(b) Co (10%)/Zn (5%)/TiO2: We have previously reported
on reducibility studies performed in Co (10%)/Zn
(x%)/TiO2 (x = 0, 1, 5) for catalysts prepared from nitrate
A change in the %Co/%Zn ratio is observed when chang-
ing from catalysts prepared from acetate precursors to
catalysts prepared from nitrate precursors (Table 1).
These results suggest that when nitrate precursors are
used most of the zinc oxide is covered by cobalt oxide
[9,10]. Thus, the source of cobalt plays a role in exposing
zinc on the surface of the catalyst (Table 1).
3
.2. Catalyst reduction study
Much work has been reported on the reducibility of sup-
ported cobalt catalysts [5,8,9]. Reduction of cobalt takes
place in two reduction steps, which corresponds to the re-
duction of Co3 to Co and reduction of Co to metallic
+
2+
2+
cobalt, respectively [9].
Fig. 1. TPR profiles showing the effect of adding different precursors of
cobalt to generate Co/TiO2 catalysts, (a) cobalt acetate, (b) cobalt acetate
and cobalt nitrate (1:1 molar ratio) and (c) cobalt nitrate.
(
a) Co/TiO2catalysts: Both cobalt acetate and cobalt nitrate
(Fig. 1a and c) gave similar TPR profiles and did not

N.N. Madikizela-Mnqanqeni, N.J. Coville / Journal of Molecular Catalysis A: Chemical 225(200 5) 137–142
141
Table 3
Summary of the FT results
Metal precursors
Co Ac.
24.5
0.43
Co Ac. + Co nitr.
Co nitr.
29.9
Co Ac. + Zn Ac.
Co Ac. + Zn Nitr.
Co Nitr. + Zn Ac.
Co Nitr. + Zn Nitr.
CO conversion (%)
Activity (mmol/g/s)
36.4
0.96
46.1
1.30
51.0
1.58
54.6
1.72
42.6
0.89
0.64
Selectivities (% by mass)
C1
28.5
23.7
13.9
13.6
19.6
23.4
16.3
14.7
15.3
31.8
◦
15.1
9.9
8.5
9.7
56.8
26.4
11.8
21.4
27.7
12.8
−1
20.2
16.7
22.5
12.7
27.5
14.5
13.4
18.4
21.3
32.4
◦
10.9
9.2
12.8
16.6
50.6
C2–C4
C5–C11
C12–C15
C16+
◦
−1
Reaction conditions: calcinations, 300 C, 1 atm, 19 h; reduction, 250 C, 1 atm, 500 h , 100% H2 and synthesis, 220 C, 8 bar, 400 h , 200 h, 2H2:1CO.
precursors and the results for x = 0 and 1% revealed no
differences in the TPR profiles [9,10]. For the 5% loaded
zinc, a combination of zinc acetate and cobalt nitrate re-
sulted in the formation of easily reducible cobalt species
when compared to zinc and cobalt salts with the same
ions (zinc nitrate and cobalt nitrate, or zinc acetate and
cobalt acetate) or the catalyst prepared from zinc nitrate
and cobalt acetate [9]. These results correlate well with
the results obtained from the XPS technique.
The cobalt acetate precursor gave a cobalt catalyst which
has higher methane selectivity and lower wax selectiv-
ity when compared to the nitrate counterions, a result
in agreement with earlier studies on TiO2 [5] (Fig. 2;
Table 3). These results are also similar to those ob-
served by Sun et al. on SiO2 [8]. It is thus clear that
the Co counter-ion (via the solution pH) plays an im-
portant role in influencing particle size and consequently
the FT activity of the Co catalysts. The FT results cor-
relate with cobalt reducibility, particle size as well as
dispersion as revealed by XPS, TPR and chemisorption
techniques. Hydrogen spillover reactions during reduc-
tion, whenmixedprecursorsaltsareemployed, havebeen
suggested to be a contributing factor for the observed cat-
alytic behavior [8].
3
.3. Chemisorption results
(
a) Co/TiO2: The use of cobalt nitrate as a cobalt catalyst
precursor gave a cobalt dispersion of 1.35% (Table 2).
When acetate precursors were used in the preparation of
cobalt catalyst supported on titania, the amount of cobalt
dispersedonthe surface increased by approximately25%
(
b) Co (10%)/Zn (5%)/TiO2: Catalysts prepared from mixed
precursor salts produced from diffferent metals (i.e. zinc
and cobalt) were more active and have a higher CO con-
version than catalysts prepared from single-precursors
of zinc and cobalt (Table 3). The observed catalytic be-
havior of Co/Zn/TiO2, when compared to Co/TiO2 cat-
alysts, could be attributed to the effect of zinc forming
a ‘sandwich’ layer between cobalt and the support on
the dispersion and stability of cobalt crystalline particles
(Table 2). A combination of cobalt nitrate and cobalt
acetateaddedtoTiO2 resultedinanincreaseindispersion
(
± 39%) when compared to cobalt nitrate impregnated on
titania (Table 2). These results correlate with the results
obtained from the TPR technique where easily reducible
cobalt oxide species were detected when mixed cobalt
precursor salts were used.
[10].
(
b) Co (10%)/Zn (5%)/TiO2: In general the dispersion data
showed a small range of values that correlated with the
XPS data, except for the cobalt nitrate/zinc nitrate sam-
ple. The highest dispersion was also obtained when zinc
acetate was impregnated with cobalt nitrate (Table 2).
These results also show a correlation with the reducibil-
ity results obtained when using the TPR technique [9].
However, exposure of zinc oxide particles on the sur-
face was found to have detrimental effects with in-
3
.4. FTS
(
a) Co/TiO2: Catalysts prepared from the mixed cobalt pre-
cursor salts resulted in a higher CO conversion and a
higher catalytic activity (Table 3) when compared to cat-
alysts prepared from single-cobalt precursors. However,
the cobalt catalyst prepared from the nitrate precursor
resulted in higher selectivity towards the production of
wax and lower methane selectivity when compared to the
other Co/TiO2 catalysts (Fig. 2; Table 3).
Fig. 2. Effect on hydrocarbon selectivity of different cobalt precursors on
Co/TiO catalysts.
2

1
42
N.N. Madikizela-Mnqanqeni, N.J. Coville / Journal of Molecular Catalysis A: Chemical 225(200 5) 137–142
3
4
. Theuseofdifferentcobaltandzincprecursorsaltsresulted
inanincreaseintheamountofzincexposedonthesurface.
. Catalysts prepared from mixed precursors of zinc and
cobalt were more active and had higher CO conversion
when compared to catalysts prepared from both nitrate
precursors of zinc and cobalt and both acetate precursors
of zinc and cobalt.
5
. Exposure of zinc on the surface of Co/Zn/TiO2 has a detri-
mental effect on the catalytic behavior of the catalysts.
More methane was produced and this was accompanied
by less wax production.
6
. Nitrate precursors of zinc and cobalt gave better results in
terms of wax selectivity when compared to the combina-
tion of all other precursors.
Fig. 3. Effect on the percentage hydrocarbon selectivity of adding different
cobalt and zinc precursors on Co/Zn (5%)/TiO2 catalysts.
creased methane selectivity and decreased amounts of
high molecular weight products [10]. An increase in the
amount of zinc (as revealed by XPS data) resulted in
more methane production and less wax production.
Catalysts prepared from zinc acetate and cobalt acetate
showed the highest methane selectivity as well as the
most zinc exposed on the surface (Table 3). A similar be-
havior is also observed in catalysts prepared from zinc ni-
trate and cobalt acetate. Catalysts prepared from nitrates
of zinc and cobalt have the lowest methane selectivities
and the lowest amount of zinc on the surface (Table 3;
Acknowledgements
We wish to thank Sastech, THRIP, NRF and the University
for financial support, and the CSIR for performing the XPS
analysis.
References
[1] E. Iglesia, Appl. Catal. A (1997) 161.
[2] J. Panpranot, S. Kaewkun, P. Praserthdam, J.G. Goodwin Jr., Catal.
Lett. 91 (2003) 95.
Fig. 3). The wax production (fraction >C ) is favoured
16
by nitrate counterions. This suggests that large Co par-
ticles favour chain propogation over chain termination
reactions.
Interestingly, the CO chemisorption data and not the XPS
data correlate with the catalyst activity. While this is con-
sistent with the general mechanism proposed in which
the amount of reduced Co determines the FT activity, it
also suggests that the secondary impact of Zn on the FT
reaction is important.
[3] J. Van de Loosdrecht, M. van der Haar, A.M. van der Kraan, A.J.
van Dillen, J.W. Geus, Appl. Catal. A: General 150 (1997) 365.
[4] A. Martinez, C. Lopez, F. Marquez, I. Diaz, J. Catal. 220 (2003)
86.
4
[
[
5] M. Kraum, M. Baerns, Appl. Catal. A: Gen. 186 (1999) 189.
6] E. van Steen, G.S. Sewell, R.A. Makhothe, C.M. Micklethwaite, H.
Manstein, M. de Lange, C.T. O’Connor, J. Catal. 162 (1996) 220.
[7] J. Li, N.J. Coville, Appl. Catal. A 181 (1999) 201.
[8] S. Sun, N. Tsubaki, K. Fujimoto, Appl. Catal. A: Gen. 202 (2000)
1
21.
9] N.N. Madikizela, N.J. Coville, J. Mol. Catal. A: Chem. 181 (2002)
29.
[
1
[
[
[
10] N.N. Madikizela-Mnqanqeni, N.J. Coville, Appl. Catal. A: Gen.
4
. Conclusion
(2004).
11] I. Alsstrup, I. Chorkendorff, R. Candia, B.S. Clausen, H. Topsoe, J.
Catal. 77 (1982) 397.
12] H.J. Thomas, M.N. Blanco, C.V. Caceres, N. Firpo, F.J. Gil Llam-
bias, J.J.G. Fierro, L. Agudo, J. Chem. Soc., Faraday Trans. 86
Our study allows for the following conclusions to be
drawn:
(
1990) 2765.
1
. The use of acetate precursor salts, to prepare Co/TiO2 cat-
alysts resulted in the synthesis of highly dispersed crys-
talline cobalt particles which were easy to reduce when
compared to their nitrate counter ions. However, these par-
ticles were less active towards FTS and resulted in higher
methane selectivity.
. Combinations of nitrate and acetate precursors used to
prepare Co/TiO2 catalysts resulted in better catalytic per-
fomance in terms of activity and CO conversion when
compared to the use of their mono precursors.
[
[
[
13] D.J. Duvenhage, N.J. Coville, Appl. Catal. A: Gen. 152 (1997) 43.
14] H.H. Nijs, P.A. Jacobs, J. Chromatogr. Sci. 19 (1981) 40.
15] J.T. Scanlon, D.E. Wills, J. Chromatogr. Sci. 23 (1985) 33.
[16] N.S. Apanos, P.G. Koutsoukos, J. Colloid Interface Sci. 241 (1999)
85.
[
17] A. Maezwa, Y. Okamoto, T. Imanaka, J. Chem. Soc., Faraday Trans.
3 (1987) 665.
8
2
[
18] T. Baird, K.C. Campbell, P.J. Holliman, R. Hoyle, D. Stirling, B.P.
Williams, J. Chem. Soc., Faraday Trans. 91 (1995) 3219.
[19] R.L. Chin, D.M. Hercules, J. Catal. 74 (1982) 121.
[20] B.R. Strohmer, D.M. Hercules, J. Catal. 86 (1984) 266.