- Email: [email protected]
Fischer–Tropsch synthesis over alumina supported cobalt catalyst: Effect of promoter addition
Accepted Manuscript Title: Fischer–Tropsch synthesis over alumina supported cobalt catalyst: Effect of promoter addition Author: Katsuya Shimura Tomohisa Miyazawa Toshiaki Hanaoka Satoshi Hirata PII: DOI: Reference:
S0926-860X(15)00032-0 http://dx.doi.org/doi:10.1016/j.apcata.2015.01.017 APCATA 15209
To appear in:
Applied Catalysis A: General
Received date: Revised date: Accepted date:
24-9-2014 28-11-2014 10-1-2015
Please cite this article as: K. Shimura, T. Miyazawa, T. Hanaoka, S. Hirata, FischerndashTropsch synthesis over alumina supported cobalt catalyst: Effect of promoter addition, Applied Catalysis A, General (2015), http://dx.doi.org/10.1016/j.apcata.2015.01.017 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Co-loaded only
La(2%)
an
us
cr
CO conversion / %
ip t
Graphical abstract
V(0.5%)
La(2%)+ V(0.5%)
Ac ce p
te
d
M
Metal promoter
1
Page 1 of 40
Highlights ・Mg, Ca, Sr, Ba, Y, La, Ce, Ti, V, Mn, Zn, Zr and Mo were examined as promoters.
Ac ce p
te
d
M
an
us
cr
・La-loading increased Co surface area, while V-loading improved TOF.
ip t
・Activity of Co/Al2O3 catalysts was largely increased by co-loading of La and V.
2
Page 2 of 40
Fischer–Tropsch synthesis over alumina supported cobalt catalyst: Effect of
cr
Katsuya Shimura,* Tomohisa Miyazawa, Toshiaki Hanaoka and Satoshi Hirata
ip t
promoter addition
Biomass Refinery Research Center, National Institute of Advanced Industrial Science and Technology (AIST),
us
Kagamiyama 3-11-32, Higashihiroshima, Hiroshima 739-0046, Japan
an
*Corresponding author
M
Katsuya Shimura, E-mail: [email protected], Tel.: +81-82-420-8292
d
Abstract
te
In order to improve the activity of Co/Al2O3 catalyst for FT synthesis, loading effect of metal promoter on
Ac ce p
Al2O3 support was examined. In this study, total 13 kinds of metal (Mg, Ca, Sr, Ba, Y, La, Ce, Ti, V, Mn, Zn, Zr and Mo) were employed as promoters and θ-Al2O3 prepared by the calcination of commercial boehmite was used as support. Co/Al2O3 catalysts loaded with various metal promoters were prepared by a sequential impregnation method. Loading of Mn, V and Mo on Al2O3 support decreased Co reducibility and surface area of Co metal, but increased intrinsic activity per active Co metal sties (i.e. TOF). On the other hand, metal additives other than Mn, V and Mo varied Co reducibility and Co surface area, but did not influence TOF. High overall activity was obtained over Co/Al2O3 catalysts loaded with rare earth elements (e.g. La and Ce), which was due to the high surface area of Co metal. We also examined the co-loading of La and V promoters 3
Page 3 of 40
on Al2O3 support. Co/Al2O3 catalysts loaded with both La and V (Co/La+V/Al2O3) showed higher activity than those loaded with either La or V. CO conversion rate over the best Co/La(2)+V(0.5)/Al2O3 catalyst was
ip t
1.6 times as high as that over bare Co/Al2O3 catalyst. Over the Co/La+V/Al2O3 catalyst, La and V promoters
cr
cooperatively improved Co reducibility, Co surface area and TOF, resulting in an increase of overall catalytic
us
activity.
M
an
Keywords: Fischer-Tropsch synthesis, Cobalt, Alumina, Metal promoter
1. Introduction
d
It is very important for the realization of a sustainable society to establish practical methods of producing
te
clean alternative fuels. Fischer-Tropsch (FT) synthesis, which can synthesize hydrocarbon mixtures from
Ac ce p
syngas, is an excellent way to produce liquid fuel, since various carbon sources (e.g. coal, natural gas and biomass) can be used as feedstock [1, 2]. Moreover, clean liquid fuels without containing any sulfur, nitrogen and aromatic compounds can be obtained by FT synthesis. Thus, FT synthesis has gained renewed attention as one of the production methods of liquid fuels and various researches such as development of catalysts and reactors are currently underway to enhance the efficiency of this reaction system. Among the potential catalysts for FT synthesis, such as Fe, Co, Ni and Ru, Co-based catalysts are advantageous to the practical application due to the relatively high activity and selectivity to long-chain liner hydrocarbons, high resistance toward deactivation, low activity for the competitive water-gas shift reaction and lower price than Ru [3–5]. It 4
Page 4 of 40
has been accepted that FT synthesis over Co catalyst would be a structure-insensitive reaction and the activity of Co catalyst would depend on the number of exposed Co metal sites on the catalyst surface [6–8]. Therefore,
ip t
Co was typically deposited on supports having high surface area (e.g. SiO2, Al2O3, TiO2 and carbon materials)
cr
in order to increase the number of active Co metal sites [1–5]. However, several research groups recently found that turnover frequency (TOF) monotonically decreased with decreasing Co particle size when Co
us
particle size was smaller than ca. 10 nm [9–11]. These reports indicate that the optimum Co particle size
an
would exist for FT synthesis. Thus, the precise control of Co particle size is very important for the development of highly active Co catalysts for FT synthesis.
M
Due to the high thermal stability and the strong resistance to attrition, alumina is frequently employed as
d
support for Co-based FT catalysts. The great ability of Al2O3 support to stabilize small metal clusters would
te
suppress the aggregation of Co metal particles during catalytic reaction and may extend the catalyst lifetime.
Ac ce p
However, the strong interaction of Al2O3 support with Co oxide suppresses the reduction of Co oxide at the same time. This means a decrease in the number of active Co metal sites and results in the formation of low active catalysts. Thus, small amount of noble metals (e.g. Pt [12–14], Re [12, 13, 15], Ru [12, 13, 16], Pd [13, 17], Ir [18], Au [19. 20] and Ag [19, 21]) is often added to Co/Al2O3 catalysts, since they can facilitate the reduction of Co oxide and increase the number of active Co metal sites, presumably by hydrogen dissociation and spillover from the promoter surface. However, noble metals are not suitable to the industrial application due to their high cost. In order to improve the Co reducibility of Co/Al2O3 catalysts, use of other methods, such as the structural control of Al2O3 support [22–25] and the improvement in the preparation method of 5
Page 5 of 40
Co/Al2O3 catalysts [11, 26–28], is preferable. It is also known that loading small amount of metal promoters on Al2O3 support sometimes increases the
ip t
activity of Co/Al2O3 catalyst for FT synthesis. Various metal promoters such as Zr [29–31], alkali earth
cr
elements [32–36], rare earth elements [37–40], and others (P [41], SiO2 [42], B [43], Ti [44]) were reported to increase the activity of Co/Al2O3 catalyst. It is generally accepted that these metal additives would improve
us
the interaction of Al2O3 support with Co oxide and thereby increase the Co reducibility and the number of
an
active Co metal sites. Although various metal promoters were found for Co/Al2O3 catalyst as described above, effect of each metal promoter on the structure of Co particles and the activity of Co/Al2O3 catalyst is not well
M
understood. This is because experimental conditions (e.g. type of Al2O3 support, pre-reduction condition of
d
catalyst, type of reaction vessel and reaction condition) were different among researchers and there were few
te
studies to compare several metal promoters at the same time. If several Co/Al2O3 catalysts modified with
Ac ce p
different metal promoters were evaluated in the same condition, we can obtain the better understanding for the loading effect of each metal promoter on Co/Al2O3 catalyst. In the present study, loading effect of metal promoter on Al2O3 support was examined in order to prepare highly active Co/Al2O3 catalysts for FT synthesis. Total 13 kinds of metal were employed as promoters and θ-Al2O3 prepared from commercial boehmite [45] was used as support. Impact of each metal promoter on the structure of Co particles and the activity of Co/Al2O3 catalyst was systematically studied. We also examined the co-loading of La and V promoters on the Al2O3 support and found that La and V additives could cooperatively increase the overall activity of Co/Al2O3 catalyst for FT synthesis. 6
Page 6 of 40
2. Experimental
ip t
2.1. Catalyst preparation
cr
Co/Al2O3 catalysts modified with various promoters (named Co/M(x)/Al2O3) were prepared by a sequential impregnation method. M and x represent the kind of metal additives and the loading amount (wt%),
us
respectively. First, M/Al2O3 was prepared from θ-Al2O3 (BET specific surface area: 84 m2 g-1) powders and an
an
aqueous solution of each metal promoter by an impregnation method. θ-Al2O3 was obtained by calcination of commercial boehmite (AlO(OH)·nH2O, Wako) at 1000 ˚C for 10 h [45]. The precursors of metal additives Mg(NO3)26H2O,
Ca(NO3)24H2O,
Sr(NO3)2,
Ba(NO3)2,
M
were
Y(NO3)36H2O,
La(NO3)36H2O,
d
Ce(NO3)36H2O, (NH4)2TiO(C2O4)2, NH4VO3, Mn(NO3)26H2O, Zn(NO3)26H2O, ZrO(NO3)22H2O, and
te
(NH4)6Mo7O244H2O. Loading amount of each metal was typically 1 wt%. After impregnation, the obtained
Ac ce p
powders were dried at 100 ˚C for 12 h, followed by calcination in air at 500 ˚C for 3 h. Then, Co (20 wt%) was loaded on M/Al2O3 by an impregnation method using Co(NO3)26H2O (99.5%, Wako) as Co precursor. Finally, the obtained powders were dried at 100 ˚C for 12 h and calcined in air at 400 ˚C for 3 h. 2.2. Characterization
Powder X-ray diffraction (XRD) pattern was recorded at room temperature on a Rigaku diffractometer RINT 2500 TTRIII using Cu Kα radiation (50 kV, 300 mA). The mean particle size of Co3O4 (dCo3O4) was calculated from the diffraction line at 2θ = 36.9 degree with the Scherrer equation. The obtained particle size of Co3O4 could be used to calculate that of Co metal (dCo) after hydrogen reduction pretreatment by the following 7
Page 7 of 40
formula (Eq. (1)) [14]. dCo ＝ 0.75 × dCo3O4
Eq. (1)
ip t
Temperature programmed reduction under H2 (H2-TPR) was carried out with BELCAT-B (BEL Japan Inc.). The calcined catalyst (0.10 g) was mounted in a quartz cell and heated up to 900 ˚C in a flow of 5% H2/Ar (30
cr
ml min-1). The heating rate was 10 ˚C min-1. The reduction degree of supported cobalt was calculated from the
us
amount of H2 consumption during H2 reduction pretreatment at 400 ˚C for 6 h. The effluent gas was passed
an
through a 5A molecular sieve trap to remove the produced water before reaching a thermal conductivity detector.
M
Hydrogen chemisorption experiments were performed on BELCAT-B. Before measurement, samples (0.15
d
g) were reduced at 400 ˚C for 6 h in a flow of 5% H2/Ar (15 ml min-1) and held at 400 ˚C for 1 h in a flow of
te
Ar (30 ml min-1) to desorb the residual chemisorbed hydrogen. After cooling the sample down to 100 ˚C in a
Ac ce p
flow of Ar, H2 chemisorption measurements were started. Corrected dispersion (Dcorr, Eq. (2)) and surface area of Co metal were calculated according to the method reported in the literature [46]. Eq. (2)
Transmission electron microscopy (TEM) images of the reduced and passivated Co/Al2O3 catalysts were recorded by a JEOL electron microscope (JEM-3000F, 300 kV) equipped with energy dispersive X-ray spectroscopy (EDS). In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) spectra were recorded with a JASCO FT/IR-4100 spectrometer equipped with a diffuse reflectance attachment and a MCT detector. The 8
Page 8 of 40
sample was placed in an infrared cell equipped with CaF2 windows. The spectra were obtained by collecting 64 scans at 4 cm-1 resolution. Before measurement, samples were reduced at 450 ˚C for 4 h in a flow of 5%
ip t
H2/Ar (40 ml min-1) and held at 450 ˚C for 0.5 h in a flow of N2 (40 ml min-1) to desorb the residual
cr
chemisorbed hydrogen. Subsequently, the system was cooled to 30 ˚C in a flow of N2 and CO was introduced. After adsorption of CO, the catalyst surface was purged with N2 to remove gaseous CO and then IR spectra
us
were recorded.
an
2.3. Typical procedures of catalytic reactions
FT synthesis was performed with a continuous stirred tank reactor in a similar way to the previous study [47].
M
Before reaction, catalyst (2.5 g) was in-situ reduced in a flow of H2 (40 ml min-1) at 400 ˚C and 6 h. Since the
d
Co reduction degree largely influences the number of active Co metal sites, catalysts should be reduced in the
te
optimum condition. In the present study, we roughly examined the effect of reduction condition on Co
Ac ce p
reduction degree using bare Co/Al2O3 catalyst. Co-based FT catalysts are typically reduced in H2 at 300–500 ˚C before reaction. In the present study, Co reduction degree of bare Co/Al2O3 catalyst increased from 42% to 64%, when reduction temperature increased from 300 to 400 ˚C. Therefore, reduction at 400 ˚C would be preferable to reduction at 300 ˚C. However, we could not increase the reduction temperature more than 400 ˚C due to the low performance of electric furnace of the reactor. Co reduction degree also depended on the reduction time. When reduction time was longer than 3 h, Co reduction degree was almost constant. From these results, we carried out the H2 reduction pretreatment at 400 ˚C for 6 h. After the reduction pretreatment, reactor was cooled down to room temperature and purged by N2 gas. Then, 9
Page 9 of 40
n-hexadecane (80 g) solvent was added. The reaction was carried out at 230 ˚C and 1.0 MPa for 8 h in a flow of synthetic gas (100 ml min-1). The syngas used for this reaction was obtained by the gasification of woody
ip t
biomass, followed by the gas purification and the composition adjustment [48]. Since the product gas obtained
cr
by the gasification of woody biomass contained impurities such as sulfur and tarry compounds, we removed them by passing the gas through scrubber and desulfurization tower [48, 49]. After gas cleaning, the
us
concentration of sulfur compounds decreased to less than 5 ppb [49]. Composition of this biomass-derived
an
syngas was confirmed by gas chromatography (GC) as follows: H2 (60.0%), CO (31.3%), CH4 (4.7%) and N2 (4.0%). The effluent gas after FT synthesis was analyzed by on-line GC. A thermal conductivity detector
M
(TCD) with a Porapak-Q column was used to analyze inorganic gases (H2, CO, CO2, CH4 and N2). Light
d
hydrocarbons (C1–C4) were analyzed by a flame ionization detector (FID) with a RT-QPLOT capillary
te
column. Hydrocarbons dissolved in the solvent and cooled in the trap were analyzed by GC-FID equipped
Ac ce p
with a UA-DX30 capillary column.
3. Results and discussion
3.1. Characterization of Co/M/Al2O3 catalysts Co(20%)/M(1%)/Al2O3 catalysts, which were prepared by a sequential impregnation method, were characterized by XRD, H2-TPR, H2-chemisorption and TEM. First, Co/M/Al2O3 catalysts before H2 reduction pretreatment were analyzed by XRD. From XRD patterns of all Co/M/Al2O3 catalysts, only peaks assigned to θ-Al2O3 and Co3O4 were observed (Fig S1). Peaks of other Co compounds (e.g. CoO and Co aluminate) and 10
Page 10 of 40
those derived from metal additives were not observed. Particle sizes of Co3O4 and Co metal on all Co/M/Al2O3 catalysts were estimated from the diffraction line at 36.9 degree using the Scherrer equation and
ip t
Eq. (1). The results were listed in Table 1. Particle sizes of Co3O4 and Co metal of Co/M/Al2O3 catalysts were
cr
15.7–19.9 and 11.7–14.9 nm, respectively, and varied with the kind of metal promoters. The addition of Ca, Y, La, Ce, Ti and Zn had little influence on Co particle size. Loading of Mg, Sr, Ba, Mn, Zr and Mo slightly
us
decreased Co particle size. V additives largely decreased Co particle size. Difference in Co particle size is
an
attributed to the variation of metal-support interaction with the kind of metal promoters. It is general that Co dispersion increases and Co particle size decreases with increasing the extent of metal-support interaction.
M
This means that interaction of Al2O3 support with Co oxide is the strongest over Co/V/Al2O3 catalyst among
d
the examined samples.
te
Fig. 1 shows H2-TPR profiles of bare Co(20%)/Al2O3 and various Co(20%)/M(1%)/Al2O3 catalysts. It is
Ac ce p
generally accepted that peak observed at 300–400 ˚C corresponds to the reduction of Co3O4 to CoO and peaks at 400–600 ˚C correspond to the reduction of CoO to Co metal [3]. Similarly, peaks observed at 600 ˚C and higher temperatures should originate from the reduction of barely reducible Co species such as cobalt aluminate. Peak observed at 300–400 ˚C of all Co/M/Al2O3 catalysts was very similar to each other. This shows that loading metal promoters on Al2O3 support has little influence on the reduction of Co3O4 to CoO. However, peaks observed at more than 400 ˚C were largely influenced by the kind of metal promoters. When V, Mn, Mg and Mo were loaded on Al2O3 support, peak area at more than 600 ˚C became much larger than that of bare Co/Al2O3 catalyst (Fig. 1 (A)). This indicates that these metal promoters should suppress the 11
Page 11 of 40
reduction of Co oxide. On the other hand, loading of Ca and La clearly decreased the area of high temperature peaks, indicating that these metal additives could promote the reduction of Co oxide (Fig. 1 (B)). Results of
ip t
H2-TPR clearly showed that even low (1 wt%) loading of metal promoters on Al2O3 support could greatly
cr
influence the interaction of Al2O3 support with Co oxide and vary the Co reducibility of Co/Al2O3 catalyst as follows. The addition of V, Mn, Mg and Mo would strengthen the interaction of Al2O3 support with Co oxide.
us
The strong metal-support interaction would promote the solid-solid reaction of Al2O3 support with Co
an
particles during the thermal pretreatment and suppress the aggregation of Co particles, resulting in the formation of small Co particles with low reducibility. In contrast, La and Ca promoters would reduce the
M
surface defective sites on Al2O3 support and weaken the interaction of Al2O3 support with Co particles. The
d
weak metal-support interaction would promote the particle growth of Co during the thermal pretreatments,
te
resulting in the formation of relatively large Co particles with high reducibility.
Ac ce p
Reduction degree of Co (Dred), surface area of Co metal (SA) and corrected dispersion of Co metal (Dcorr) were measured by H2-TPR and H2 chemisorption. The results were listed in Table 1. Dred, SA and Dcorr of Co/M/Al2O3 catalysts were 40–87%, 0.21–1.37 m2 g-1 and 0.31–1.26%, respectively, and depended largely on the kind of metal additives. As estimated from the result of H2-TPR (Fig. 1), Co/Al2O3 catalysts loaded with V, Mn, Mg and Mo showed low Co reduction degree (40–54%), while Co/Al2O3 catalysts loaded with La and Ca showed high Co reduction degree (76–87%). Surface area of Co metal tended to increase with increasing Co reduction degree and reached a plateau when Co reduction degree was higher than 65% (Fig. 2 (A)). Among the examined samples, Co/La/Al2O3 catalyst showed the highest Co surface area (1.37 m2 g-1), while 12
Page 12 of 40
Co/V/Al2O3 catalyst showed the lowest Co surface area (0.21 m2 g-1). Dispersion of Co metal also depended on Co reduction degree. As Co reduction degree increased, Co dispersion first increased due to increase in the
ip t
number of surface Co metal sites (Fig. 2 (B)). The highest value was obtained over Co/Al2O3 catalyst having
cr
moderate reduction degree (64%). However, further increase of reduction degree slightly decreased Co dispersion. This is probably because the aggregation of Co particles would be promoted over Co/Al2O3
us
catalysts having high reduction degree during the thermal pretreatment.
an
Reduction degree of Co and surface area of Co metal were plotted against particle size of Co3O4 (Fig. 3). Co reduction degree tended to increase with an increase of Co3O4 particle size (Fig. 3 (A)). In other words,
M
reduction of large Co3O4 particles was easier than that of small Co3O4 particles. Co surface area first increased
d
until 18.8 nm and then decreased with increasing Co3O4 particle size (Fig. 3 (B)). If Co reduction degree of all
te
Co/Al2O3 catalysts is same, Co surface area should increase with decreasing Co particle size. However, as
Ac ce p
shown in Fig. 3 (A), the decrease of Co3O4 particle size also decreased Co reducibility and suppressed the formation of surface Co metal sites. Therefore, the highest Co surface area was obtained over Co/Al2O3 catalysts having moderate Co3O4 particle size due to the trade-off effect of particle size and reduction degree. In order to examine the distribution of Co particles on Al2O3 support, Co/Al2O3, Co/La(2)/Al2O3 and Co/V(0.5)/Al2O3 catalysts were analyzed by TEM. Two representative TEM images measured at low (Fig. 4 (A)–(D)) and high magnification (Fig. 4 (E)–(H)) were shown for each sample. In Fig. 4, black spots show the area having high concentration of Co. Co particle size, Co reduction degree and Co surface area of Co/La(2)/Al2O3 catalyst were comparable to those of Co/La(1)/Al2O3 catalyst (Table 2, entries 2, 3). 13
Page 13 of 40
Properties of Co particle on Co/V(0.5)/Al2O3 catalyst were also similar with those of Co on Co/V(1)/Al2O3 catalyst (Table 2, entries 4, 5). As shown in Fig. 4 (E)–(G), Co particles smaller than 20 nm were mainly
ip t
observed from TEM images of all samples. Although Co surface area measured by H2 chemisorption was
cr
largely varied with the kind of metal additives in the range of 0.2–1.4 m2 g-1, no clear differences in Co particle size and distribution were observed from their TEM images. Therefore, we suggest that influence of
us
metal promoter on Co particle size would be small. The variation of Co surface area with the kind of metal
an
promoter is mainly attributed to the differences in Co reduction degree of each catalyst. Distribution of Co particles and metal promoters on Al2O3 support was further analyzed by TEM-EDS. Figs.
M
S2 and S3 show TEM-EDS maps of Co/La(2)/Al2O3 and Co/V(0.5)/Al2O3 catalysts. For both samples,
d
distribution of Co particles on Al2O3 support was rather heterogeneous. The distribution of La and V promoter
te
was also inhomogeneous, although loading amount of these promotes was as small as 0.5–2 wt%. However, it
Ac ce p
is interesting to note that distribution of Co particles was similar with that of metal promotes. In other words, Co particles tended to exist in the area where concentration of metal promoter was high. This result indicates that metal promoters would contribute to the generation and dispersion of Co particles. As conclusion of this section, loading small amount of metal promoters modified the interaction of Al2O3 support with Co3O4 particles and varied Co3O4 particle size, Co reduction degree and Co surface area. Co reduction degree and Co surface area tended to increase with increasing Co3O4 particle size. Co/La/Al2O3 catalyst showed the highest Co surface area, while Co/V/Al2O3 catalyst showed the lowest Co reduction degree and Co surface area. 14
Page 14 of 40
3.2. Catalytic performance of Co/M/Al2O3 and structure-activity relationship
ip t
FT synthesis was performed on various Co/Al2O3 catalysts using a continuous stirred tank reactor at 230 ˚C
cr
and 1.0 MPa for 8 h. Fig. S4 shows time course of CO conversion and space time yield of C5+ products (STY of C5+) over bare Co/Al2O3 catalysts as the representative. Deactivation of catalysts was hardly observed for
us
all samples due to the short reaction time. However, prolonged reaction time should absolutely deactivate the
an
catalyst due to the structural changes of cobalt particles, such as sintering, surface reconstruction and carbization, as reported by various research groups [50, 51]. Therefore, we can only discuss the effect of metal
M
promoter on initial catalytic activity (8 h after the reaction started) in this paper, since catalytic performance
d
may be largely changed during a long time reaction test. We will examine the lifetime of various Co/M/Al2O3
te
catalysts and report the effect of each metal promoter on the catalyst stability in the future. However, it should
Ac ce p
be noted that in our previous study using similar reaction condition and reaction setup, the activity of Co/SiO2 and Co/Mn+Zr/SiO2 catalysts was almost constant for 1 week owing to the precise control of reaction temperature and the use of hexadecane solvent [52]. Thus, we expect that deactivation of the Co/M/Al2O3 catalysts may be quite slow in the present reaction condition and reaction setup. Table 3 shows CO conversion, product selectivity, STY of C5+ and TOF per surface Co metal atom on bare Co(20%)/Al2O3 and various Co(20%)/M(1%)/Al2O3 catalysts. TOF was calculated from the results of H2 chemisorption and catalytic reaction. CO conversion and STY of C5+ products over Co/M/Al2O3 catalysts were 33–55% and 126–200 mg g-1 h-1, respectively. Addition of La, Ce and Zr increased CO conversion and 15
Page 15 of 40
STY of C5+ products, while metal additives other than La, Ce and Zr decreased them. Among the catalysts listed in Table 3, Co/Ce/Al2O3 catalyst showed the highest CO conversion and STY of C5+ products.
ip t
Selectivities of CO2, CH4, C2–C4 and C5+ products over Co/M/Al2O3 catalysts were 2–5%, 6–10%, 4–10% and
cr
78–86%, respectively. Co/V/Al2O3 catalyst showed the highest C5+ selectivity, while Co/Zn/Al2O3 catalyst showed the lowest one. The similar results were also reported for Co/TiO2 catalyst [53]. Previously, our
us
research group studied the loading effect of metal promoters on Co/TiO2 catalyst and found that Co/V/TiO2
an
catalyst showed high C5+ selectivity and Co/Zn/TiO2 catalyst showed low C5+ selectivity [53]. However, we need to compare product selectivity at the same conversion level in order to understand the effect of metal
M
additives on product selectivity more properly, since product selectivity depends on CO conversion [54]. TOF
d
largely depended on the kind of metal additives. When Mg, Ca, Sr, Ba, Y, La, Ce, Ti, Zn and Zr were used as
te
additives, TOF was not influenced by the kinds of metal promoter and almost the same as that of bare
Ac ce p
Co/Al2O3 catalyst. In contrast, TOF of Co/Al2O3 catalysts loaded with V, Mn and Mo (0.36–0.65 s-1) was much higher than that of bare Co/Al2O3 catalyst (0.15 s-1). This result is not usual, because FT synthesis over Co catalysts is a structure-insensitive reaction in many cases [6–8]. However, a few research groups reported that Mn-loading increased TOF of Co/C [55] and Co/TiO2 [56, 57] catalysts. They propose that Mn2+ promoters would withdraw electrons from Co metal sites and vary the adsorption ability and the hydrogenation/dehydrogenation ability of Co metal sites [58]. The facts that V-loading decreased the number of active metal sites but increased the TOF were also reported for CO hydrogenation over Ru/Al2O3 [59], Ni/Al2O3 [60] and Rh/SiO2 [61] catalysts. From the 16
Page 16 of 40
analysis of reaction kinetics [59] and the FT-IR study [61], it was proposed that role of V promoter would be to promote the CO dissociation and the hydrocarbon chain growth. Thus, we also carried out the FT-IR study
ip t
in order to examine the effect of V promoter on CO adsorption. Fig. S5 shows FT-IR spectra of CO adsorbed on Co/Al2O3 and Co/V(0.5)/Al2O3 catalysts at room temperature. The IR spectrum of Co/Al2O3 catalyst exhibited a
cr
strong band at 2050 cm-1, which assigned to the linear adsorbed CO on surface Co metal atom. Very small bands
us
assigned to the bridge-bonded CO with metallic Co (2000–1800 cm-1) were also observed from the spectrum of
an
Co/Al2O3. Co/V(0.5)/Al2O3 catalyst also exhibited the clear FT-IR spectrum of CO adsorbed on Co metal, although V-loading largely disturbed the adsorption of H2 (Table 2). V-loading did not influence the position and intensity of
M
the linearly bonded CO band. However, the intensity of the bridge bonded CO band was clearly increased by
d
V-loading. From the result of IR study, we suggest that V promoter may influence the state and amount of the
te
adsorbed CO, resulting in an improvement of specific activity of Co/Al2O3 catalyst. Further study is now carried
Ac ce p
out to clarify the role of V promoter.
Reduction degree of Co and surface area of Co metal were important factors to determine the activity of Co catalyst. Thus, influence of these parameters on CO conversion and TOF was examined. Results were shown in Figs. 5 and 6. Overall, CO conversion did not show any clear correlation with Co reduction degree (Fig. 5 (A)). CO conversion did not also depend on Co surface area (Fig. 6 (A)). However, except for Co/Al2O3 catalysts loaded with V, Mn, Mo and Ca, CO conversion tended to increase with increasing Co reduction degree and Co surface area. Low activity of Co/Ca/Al2O3 catalyst is probably due to the poisoning effect of Ca ions [62, 63]. The different tendency of Co/Al2O3 catalysts loaded with V, Mn and Mo is 17
Page 17 of 40
attributed to their high TOF, as listed in Table 3. TOF largely depended on Co reduction degree and Co surface area. When Co reduction degree was higher than 55%, TOF was not varied with reduction degree (Fig. 5 (B)).
ip t
However, TOF of Co/M/Al2O3 catalysts having low Co reduction degree, such as Co/V/Al2O3, Co/Mn/Al2O3
cr
and Co/Mo/Al2O3 catalysts, was much higher than that of bare Co/Al2O3 catalyst. Increase of TOF with decreasing Co reduction extent was reported by Lee and co-workers [64]. They proposed that the electronic
us
effect of subsurface Co oxide may influence the high TOF. Similarly, TOF was not varied with Co surface
an
area, when CO surface area was higher than 0.7 m2 g-1 (Fig. 6 (B)). However, TOF of Co/M/Al2O3 catalysts having low Co surface area, such as Co/V/Al2O3, Co/Mn/Al2O3 and Co/Mo/Al2O3 catalysts, was much higher
M
than that of bare Co/Al2O3 catalyst. Results of Fig. 6 were consistent with those of Fig. 5, since Co/M/Al2O3
d
catalysts with lower Co reduction degree tended to show lower Co surface area (Fig. 2 (A)). In the present
te
study, we could not show clear reasons why addition of V, Mn and Mo largely increased TOF. However, we
Ac ce p
tentatively propose that high TOF of Co/Al2O3 catalysts loaded with V, Mn and Mo is due to not the size of Co particles but the electronic effect of unreduced Co oxide and metal additives. This is because Co particle size of Co/Al2O3 catalysts prepared in the present study is in the range of 11.7–14.9 nm (Table 1) and TOF of supported Co catalysts is generally constant for Co particle size larger than 10 nm [9–11]. Results of characterization and reaction tests showed that the examined metal additives could be divided into two categories: V, Mn and Mo decreased Co reduction degree and Co surface area, but increased TOF. Metal additives other than V, Mn and Mo varied Co reduction degree and Co surface area, but did not influence TOF. 18
Page 18 of 40
3.3. Catalytic performance and physicochemical property of Co/La+V/Al2O3
ip t
The overall activity of Co/Al2O3 catalyst depends on the number of surface Co metal sites and the TOF.
cr
Improvement of both parameters is necessary for an increase of overall catalytic activity. As described in section 3.1 and 3.2, La-loading increased Co surface area, while V-loading improved TOF. Thus, co-loading of
us
La and V promoters on Al2O3 support was examined. Results of FT synthesis over Co/Al2O3, Co/La/Al2O3,
an
Co/V/Al2O3 and Co/La+V/Al2O3 catalysts were shown in Table 4. CO conversion and STY of C5+ products over Co/La/Al2O3 catalysts were 50–54% and 188–192 mg g-1 h-1, respectively, and a little larger than those of
M
bare Co/Al2O3 catalyst (Table 4, entries 1–3). CO conversion and STY of C5+ products over Co/V/Al2O3
d
catalysts were 37–53% and 145–192 mg g-1 h-1, respectively (Table 4, entries 4, 5). As V-loading amount
te
increased, CO conversion and STY of C5+ products first increased until 0.5% and then decreased. CO
Ac ce p
conversion and STY of C5+ products over Co/La+V/Al2O3 catalysts were 55–77% and 208–279 mg g-1 h-1, respectively, and higher than those of Co/Al2O3, Co/La/Al2O3 and Co/V/Al2O3 catalysts (Table 4, entries 6–9). The highest activity was obtained over Co/La(2)+V(0.5)/Al2O3 catalyst. CO conversion and STY of C5+ products over this catalyst were 1.6 times higher than those of bare Co/Al2O3 catalyst. Product selectivities of CO2, CH4, C2–C4 and C5+ products over Co/La+V/Al2O3 catalysts were 2–4%, 7–9%, 5–8% and 81–85%, respectively. Co-loading of La and V had little impact on product selectivity. TOF of Co/La+V/Al2O3 catalysts (0.27–0.61 s-1) was higher than that of Co/La/Al2O3 catalysts (0.14–0.15 s-1) but lower than that of Co/V/Al2O3 catalysts (0.65–0.68 s-1). As V-loading amount increased, TOF of Co/La+V/Al2O3 catalysts 19
Page 19 of 40
increased. However, TOF of Co/La+V/Al2O3 catalysts decreased with increasing La-loading amount. Results of Table 4 clearly show that co-loading of La and V is effective for an increase in the overall activity of
ip t
Co/Al2O3 catalyst (see also Figs. 5 and 6).
cr
Co/La+V/Al2O3 catalysts were characterized by XRD, H2-TPR, H2 chemisorption and TEM-EDS in order to clarify the reason for their high catalytic activity. Fig. 7 shows H2-TPR profiles of Co/Al2O3, Co/La/Al2O3,
us
Co/V/Al2O3 and Co/La+V/Al2O3 catalysts. Spectra of Co/La/Al2O3 and Co/V/Al2O3 catalysts showed that the
an
area of high temperature peaks more than 600 ˚C was decreased by La-loading but increased by V-loading. H2-TPR profiles of Co/La+V/Al2O3 catalysts were relatively similar with those of Co/V/Al2O3 catalysts.
d
V-loading amount decreased Co reducibility.
M
Increase of La-loading amount increased Co reducibility of Co/La+V/Al2O3 catalysts, while increase of
te
Co particle size was determined by XRD (Table 2). Co3O4 particle sizes of Co/La/Al2O3, Co/V/Al2O3 and
Ac ce p
Co/La+V/Al2O3 catalysts were 18.8–19.1, 15.7–18.7 and 16.1–18.6 nm, respectively. Co3O4 particle size of Co/V/Al2O3 catalysts decreased with increasing V-loading amount and Co/V(1)/Al2O3 catalyst showed the smallest size (Table 2, entries 4, 5). However, co-loading of La suppressed the decrease of Co3O4 particle size by V-loading to some extent (Table 2, entries 6–9). Reduction degree of Co, surface area of Co metal and corrected dispersion of Co metal were also listed in Table 2. When La-loading amount of Co/La/Al2O3 catalyst was increased from 0 to 2 wt%, Co reduction degree increased from 64% to 82% but Co dispersion decreased from 1.21% to 1.06% (Table 2, entries 1–3). Co surface area of Co/La/Al2O3 catalyst increased with increasing La-loading amount until 1% and then became constant. Co reduction degree, Co surface area and 20
Page 20 of 40
Co dispersion of Co/V/Al2O3 catalysts were 40%, 0.21–0.29 m2 g-1 and 0.35–0.48%, respectively. V-loading largely decreased these three parameters. Co reduction degree, Co surface area and Co dispersion of
ip t
Co/La+V/Al2O3 catalysts were higher than those of Co/V/Al2O3 catalysts but lower than those of Co/La/Al2O3
cr
catalysts (see also Figs. 2 and 3). Results of Table 2 show that Co/La+V/Al2O3 catalysts have properties intermediate between Co/La/Al2O3 and Co/V/Al2O3 catalysts. Therefore, we propose the reasons for high
us
catalytic activity of Co/La+V/Al2O3 as follows: La-loading is effective for an improvement in Co reducibility
an
and Co surface area. V-loading is effective for an increase of TOF. Over Co/La+V/Al2O3 catalysts, these positive effects are obtained at the same time, resulting in an increase of the overall catalytic activity.
M
Co/La+V/Al2O3 catalysts were also analyzed by TEM. As shown in Fig. 4 (H), Co particles smaller than 20
d
nm were mainly observed from TEM images. TEM images of Co/La+V/Al2O3 catalyst resembled those of
te
Co/Al2O3, Co/La/Al2O3 and Co/V/Al2O3 catalysts, indicating that co-loading of La and V had a small
Ac ce p
influence on the distribution of Co particles on Al2O3 support. Fig. 8 shows the TEM-EDS maps of Co/La(2)+V(0.5)/Al2O3 catalyst. La and V promoters were not homogeneously dispersed on A2O3 support. However, distribution of La and V promoters was similar to that of Co particles, as in the case of Co/La(2)/Al2O3 and Co/V(0.5)/Al2O3 catalysts (Figs. S2 and S3). This means that Co particles tend to be formed on the area where concentrations of La and V are high. We suggested that over Co/La(2)+V(0.5)/Al2O3 catalyst, the good contact of La and V promoters with Co particles would effectively improve the overall activity. La promoter did not influence the specific activity. However, La-loading could increase the number of active Co metal sites, since La promoter weakened the interaction of Al2O3 support 21
Page 21 of 40
with Co oxide and greatly increased the Co reducibility. In contrast, V-loading could largely improve the specific activity probably due to the promotion of CO activation, although V promoter also strengthened the
ip t
interaction of Al2O3 support with Co oxide and decreased the Co reducibility (i.e. number of active Co metal
cr
sites). The good contact of La and V promoters with Co particles enabled to improve both the number and
us
property of active Co metal sites, resulting in an enhancement of overall catalytic activity.
an
Conclusions
Total 13 kinds of metals (Mg, Ca, Sr, Ba, Y, La, Ce, Ti, V, Mn, Zn, Zr and Mo) were examined as additive of
M
Al2O3 support. Impact of each metal promoter on the structure of Co particles and the activity of Co/Al2O3
d
catalyst for FT synthesis was studied in details. Among the examined metal promoters, V, Mn and Mo
te
decreased Co reducibility and surface area of Co metal. However, TOF of Co/Al2O3 catalysts loaded with V,
Ac ce p
Mn and Mo was much higher than that of bare Co/Al2O3 catalyst. In contrast, metal additives other than V, Mn and Mo varied Co reducibility and surface area of Co metal without influencing the TOF. High overall activity for FT synthesis was obtained, when rare earth elements such as La and Ce were used as additives. Activity of Co/Al2O3 catalyst was further increased by loading of both La and V promoters on the Al2O3 support. The best catalyst was Co/La(2)+V(0.5)/Al2O3 and CO conversion rate over this catalyst was 1.6 times higher than that over bare Co/Al2O3 catalyst. Over the Co/La+V/Al2O3 catalyst, La and V promoters cooperatively increased Co surface area and TOF, resulting in an increase of overall catalytic activity.
22
Page 22 of 40
Acknowledgment. This work was partially supported by a research grand from the Japan Prize Foundation. The authors are deeply grateful to Mr. Hideyuki Yokoyama and Ms. Tomomi Monzen for their experimental
Ac ce p
te
d
M
an
us
cr
ip t
assistance.
23
Page 23 of 40
References [1] E.F. Sousa-Aguiar, F.B. Noronha, A. Faro, Jr., Catal. Sci. Technol. 1 (2011) 698–713.
ip t
[2] R. Luque, A.R. de la Osa, J.M. Campelo, A.A. Romero, J.L. Valverde, P. Sanchez, Energy Environ. Sci. 5
cr
(2012) 5186–5202. [3] A.Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 107 (2007) 1692–1744.
us
[4] Q. Zhang, J. Kang, Y. Wang, ChemCatChem 2 (2010) 1030–1058.
an
[5] H. Jahangiri, J. Bennett, P. Mahjoubi, K. Wilson, S. Gu, Catal. Sci. Technol. 4 (2014) 2210–2229. [6] S.W. Ho, M. Houalla, D.M. Hercules, J. Phys. Chem. 94 (1990) 6396–6399.
M
[7] B.G. Johnson, C.H. Bartholomew, D.W. Goodman, J. Catal. 128 (1991) 231–247.
d
[8] E. Iglesia, S.L. Soled, R.A. Fiato, J. Catal. 137 (1992) 212–224.
te
[9] G.L. Bezemer, J.H. Bitter, H.P.C.E. Kuipers, H. Oosterbeek, J.E. Holewijn, X. Xu, F. Kapteijn, A.J. van
Ac ce p
Dillen, K.P. de Jong, J. Am. Chem. Soc. 128 (2006) 3956–3964. [10] G. Prieto, A. Martínez, P. Concepción, R. Moreno-Tost, J. Catal. 266 (2009) 129–144. [11] J. Park, Y. Lee, P.R. Karandikar, K. Jun, K. Ha, H. Park, Appl. Catal. A 411–412 (2012) 15–23. [12] G. Jacobs, T.K. Das, Y. Zhang, J. Li, G. Racoillet, B.H. Davis, Appl. Catal. A 233 (2002) 263–281. [13] W. Ma, G. Jacobs, R.A. Keogh, D.B. Bukur, B.H. Davis, Appl. Catal. A 437–438 (2012) 1–9. [14] D. Schanke, S. Vada, E.A. Blekkan, A.M. Hilmen, A. Hoff, A. Holmen, J. Catal. 156 (1995) 85–95. [15] T.K. Das, G. Jacobs, P.M. Patterson, W.A. Conner, J. Li, B.H. Davis, Fuel 82 (2003) 805–815. [16] A. Kogelbauer, J.G. Goodwin, Jr., R. Oukaci, J. Catal. 160 (1996) 125–133. 24
Page 24 of 40
[17] G. Jacobs, W. Ma, P. Gao, B. Todic, T. Bhatelia, D.B. Bukur, S. Khalid, B.H. Davis, Top Catal. 55 (2012) 811–817.
ip t
[18] D.C. Cronauer, G. Jacobs, L. Linganiso, A.J. Kropf, J.W. Elam, S.T. Christensen, C.L. Marshall, B.H.
cr
Davis, Catal. Lett. 141 (2011) 968–976.
[19] G. Jacobs, M.C. Ribeiro, W. Ma, Y. Ji, S. Khalid, P.T.A. Sumodjo, B.H. Davis, Appl. Catal. A 361
us
(2009) 137–151.
an
[20] K. Jalama, N.J. Coville, H. Xiong, D. Hildebrandt, D. Glasser, S. Taylor, A. Carley, J.A. Anderson, G.J. Hutchings, Appl. Catal. A 395 (2011) 1–9.
M
[21] T. Jermwongratanachai, G. Jacobs, W. Ma, W.D. Shafer, M.K. Gnanamani, P. Gao, B. Kitiyanan, B.H.
te
165–180.
d
Davis, J.L.S. Klettlinger, C.H. Yen, D.C. Cronauer, A.J. Kropf, C.L. Marshall, Appl. Catal. A 464–465 (2013)
Ac ce p
[22] H. Xiong, Y. Zhang, S. Wang, J. Li, Catal. Commun. 6 (2005) 512–516. [23] S. Rane, Ø. Borg, J. Yang, E. Rytter, A. Holmen, Appl. Catal. A 388 (2010) 160–167. [24] J. Sun, C. Xing, H. Xu, F. Meng, Y. Yoneyama, N. Tsubaki, J. Mater. Chem. A 1 (2013) 5670–5678. [25] J. Zhang, J. Chen, J. Ren, Y. Sun, Appl. Catal. A 243 (2003) 121–133. [26] J. van de Loosdrecht, M. van der Haar, A.M. van der Kraan, A.J. van Dillen, J.W. Geus, Appl. Catal. A 150 (1997) 365–376. [27] J.W. Bae, Y. Lee, J. Park, K. Jun, Energy Fuels 22 (2008) 2885–2891. [28] C. Liu, J. Li, Y. Zhang, S. Chen, J. Zhu, K. Liew, J. Mol. Catal. A 363–364 (2012) 335–342. 25
Page 25 of 40
[29] F. Rohr, O.A. Lindvåg, A. Holmen, E.A. Blekkan, Catal. Today 58 (2000) 247–254. [30] B. Jongsomjit, J. Panpranot, J.G. Goodwin Jr., J. Catal. 215 (2003) 66–77.
[33] A. Bao, K. Liew, J. Li, J. Mol. Catal. A 304 (2009) 47–51.
us
[34] A. Bao, J. Li, Y. Zhang, J. Nat. Gas Chem. 19 (2010) 622–627.
cr
[32] Y. Zhang, H. Xiong, K. Liew, J. Li, J. Mol. Catal. A 237 (2005) 172–181.
ip t
[31] H. Xiong, Y. Zhang, K. Liew, J. Li, J. Mol. Catal. A 231 (2005) 145–151.
an
[35] A.R. de la Osa, A.D. Lucas, J.L. Valverde, A. Romero, I. Monteagudo, P. Coca, P. Sánchez, Catal. Today 167 (2011) 96–106.
M
[36] A.R. de la Osa, A.D. Lucas, A. Romero, J.L. Valverde, P. Sánchez, Fuel 90 (2011) 1935–1945.
d
[37] J.S. Ledford, M. Houalla, A. Proctor, D.M. Hercules, L. Petrakis, J. Phys. Chem. 93 (1989) 6770–6777.
te
[38] S. Vada, B. Chen, J.G. Goodwin, Jr., J. Catal. 153 (1995) 224–231.
Ac ce p
[39] Z. Cai, J. Li, K. Liew, J. Hu, J. Mol. Catal. A 330 (2010) 10–17. [40] S. Zeng, Y. Du, H. Su, Y. Zhang, Catal. Commun. 13 (2011) 6–9. [41] J.W. Bae, S. Kim, Y. Lee, M. Lee, K. Jun, Catal. Commun. 10 (2009) 1358–1362. [42] X. Sun, X. Zhang, Y. Zhang, N. Tsubaki, Appl. Catal. A 377 (2010) 134–139. [43] K.F. Tan, J. Chang, A. Borgna, M. Saeys, J. Catal. 280 (2011) 50–59. [44] S. Park, S. Kim, M.H. Woo, J.W. Bae, K. Jun, K. Ha, Appl. Catal. A 419–420 (2012) 148–155. [45] K. Shimura, T. Miyazawa, T. Hanaoka, S. Hirata, J. Mol. Catal. A 394 (2014) 22–32. [46] R.C. Reuel, C.H. Bartholomew, J. Catal. 85 (1984) 63–77. 26
Page 26 of 40
[47] Y. Liu, T. Hanaoka, T. Miyazawa, K. Murata, K. Okabe, K. Sakanishi, Fuel Process. Technol. 90 (2009) 901–908.
ip t
[48] T. Hanaoka, T. Miyazawa, M. Nurunnabi, S. Hirata, K. Sakanishi, J. Jpn. Inst. Energy 90 (2011)
cr
1072–1080.
[49] T. Hanaoka, Y. Liu, K. Matsunaga, T. Miyazawa, S. Hirata, K. Sakanishi, Fuel Process. Technol. 91
us
(2010) 859–865.
an
[50] N.E. Tsakoumis, M. Rønning, Ø. Borg, E. Rytter, A. Holmen, Catal. Today 154 (2010) 162–182. [51] A.M. Saib, D.J. Moodley, I.M. Ciobîcă, M.M. Hauman, B.H. Sigwebela, C.J. Weststrate, J.W.
M
Niemantsverdriet, J. van de Loosdrecht, Catal. Today 154 (2010) 271–282.
d
[52] T. Miyazawa, T. Hanaoka, K. Shimura, S. Hirata, Catal. Commun. 57 (2014) 36–39.
te
[53] K. Shimura, T. Miyazawa, T. Hanaoka, S. Hirata, Appl. Catal. A 460–461 (2013) 8–14.
Ac ce p
[54] S. Storsæter, Ø. Borg, E.A. Blekkan, A. Holmen, J. Catal. 231 (2005) 405–419. [55] G.L. Bezemer, P.B. Radstake, U. Falke, H. Oosterbeek, H.P.C.E. Kuipers, A.J. van Dillen, K.P. de Jong, J. Catal. 237 (2006) 152–161.
[56] F. Morales, D. Grandjean, A. Mens, F.M.F. de Groot, B.M. Weckhuysen, J. Phys. Chem. B 110 (2006) 8626–8639.
[57] T.E. Feltes, L. Espinosa-Alonso, E. de Smit, L. D’Souza, R.J. Meyer, B.M. Weckhuysen, J.R. Regalbuto, J. Catal. 270 (2010) 95–102. [58] F. Morales, E. de Smit, F.M.F. de Groot, T. Visser, B.M. Weckhuysen, J. Catal. 246 (2007) 91–99. 27
Page 27 of 40
[59] T. Mori, A. Miyamoto, N. Takahashi, M. Fukagaya, T. Hattori, Y. Murakami, J. Phys. Chem. 90 (1986) 5197–5201.
ip t
[60] R. Karaselcuk, A.I. Isli, A.E. Aksoylu, Z.I. Onsan, Appl. Catal. A 192 (2000) 263–271.
cr
[61] X. Mo, J. Gao, N. Umnajkaseam, J.G. Goodwin Jr., J. Catal. 267 (2009) 167–176.
[62] C.M. Balonek, A.H. Lillebø, S. Rane, E. Rytter, L.D. Schmidt, A. Holmen, Catal. Lett. 138 (2010) 8–13.
Ac ce p
te
d
M
an
[64] J. Lee, D. Lee, S. Ihm, J. Catal. 113 (1988) 544–548.
us
[63] A.H. Lillebø, E. Patanou, J. Yang, E.A. Blekkan, A. Holmen, Catal. Today 215 (2013) 60–66.
28
Page 28 of 40
Figure captions Fig. 1. H2-TPR profiles of Co/Al2O3 and Co/M(1)/Al2O3 catalysts.
ip t
Fig. 2. Influence of Co reduction degree on (A) Co surface area and (B) Co dispersion.
cr
Fig. 3. Influence of Co3O4 particle size on (A) Co reduction degree and (B) Co surface area.
Fig. 4. TEM images of (A and E) Co/Al2O3, (B and F) Co/La(2)/Al2O3, (C and G) Co/V(0.5)/Al2O3 and (D
us
and H) Co/La(2)+V(0.5)/Al2O3 catalysts.
an
Fig. 5. Influence of Co reduction degree on (A) CO conversion and (B) TOF. Fig. 6. Influence of Co surface area on (A) CO conversion and (B) TOF.
M
Fig. 7. H2-TPR profiles of Co/Al2O Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 catalysts.
Ac ce p
te
d
Fig. 8. TEM-EDS maps of Co/La(2)+V(0.5)/Al2O3 catalyst.
29
Page 29 of 40
Table 1
Structural analysis of Co/Al2O3 and Co/M(1)/Al2O3 catalysts. PSa / nm
Metal ion (M)
Co3O4
Co metal
Dredb / %
SAc / m2 g-1
Dcorr d / % 1.21
－
19.4
14.6
64
1.18
2
Mg
17.9
13.4
50
0.72
3
Ca
19.8
14.9
87
1.30
4
Sr
18.2
13.7
72
1.13
5
Ba
18.5
13.9
75
1.12
0.98
6
Y
19.0
14.2
64
1.23
1.26
7
La
18.8
14.1
76
1.37
1.17
8
Ce
19.5
14.6
64
1.15
1.17
9
Ti
19.4
14.5
58
0.99
1.12
10
V
15.7
11.7
40
0.21
0.35
11
Mn
17.9
13.4
48
0.50
0.67
12
Zn
19.9
14.9
63
1.00
1.03
13
Zr
18.5
13.9
66
1.18
1.16
14
Mo
17.8
0.25
0.31
ip t
1
us
Entry
0.98
cr
1.02
an M
13.3
54
b
c
Particle size determined by XRD. Reduction degree of Co. Surface area of Co metal.
d
Corrected dispersion of Co metal.
Ac ce p
te
d
a
Table 2
0.93
Structural analysis of Co/Al2O3, Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 catalysts. Dredb
SAc
Dcorr d
Co metal
/%
/ m2 g-1
/%
19.4
14.6
64
1.18
1.21
0
18.8
14.1
76
1.37
1.17
2
0
19.1
14.3
82
1.33
1.06
0
0.5
18.7
14.0
40
0.29
0.48
5
0
1
15.7
11.7
40
0.21
0.35
6
1
0.5
18.6
13.9
57
0.87
0.98
7
1
1
16.1
12.1
42
0.34
0.52
8
2
0.5
18.3
13.7
69
1.04
0.99
9
2
1
17.2
12.9
48
0.51
0.69
Entry 1 2 3 4
PSa / nm
La
V
/ wt%
/ wt%
Co3O4
0
0
1
a
Particle size determined by XRD. b Reduction degree of Co. c Surface area of Co metal.
d
Corrected dispersion of Co metal. 30
Page 30 of 40
Table 3 Catalytic performance of Co/Al2O3 and Co/M(1)/Al2O3 in FT synthesis. Metal ion (M)
CO conv. a
STY(C5+) b
Product selectivity / %
-1
TOF
-1
/%
CO2
CH4
C2–C4
C5+
/ mg g h
/ s-1
－
48
4
8
6
82
179
0.15
2
Mg
33
4
7
5
84
126
0.17
3
Ca
40
2
7
10
81
144
0.11
4
Sr
44
3
8
7
82
163
0.14
5
Ba
46
4
9
8
79
162
0.15
6
Y
46
4
8
9
79
162
0.14
7
La
50
4
6
7
83
188
0.14
8
Ce
55
4
7
7
82
200
0.18
9
Ti
38
3
7
5
85
147
0.14
10
V
37
3
7
4
86
145
0.65
11
Mn
48
4
9
6
81
173
0.36
12
Zn
43
4
10
8
78
153
0.16
13
Zr
49
2
8
7
83
184
0.15
14
Mo
35
5
8
6
81
127
0.51
us
an
M
b
CO conversion. Space time yield of C5+ products.
te
d
a
ip t
1
cr
Entry
Table 4 Entry
La
V
CO conv. a
STY(C5+) b
Product selectivity / %
/ wt%
/%
CO2
CH4
C2–C4
C5+
/ mg g h
/ s-1
0
0
48
4
8
6
82
179
0.15
1
0
50
4
6
7
83
188
0.14
2
0
54
4
8
9
79
192
0.15
0
0.5
53
3
10
7
80
192
0.68
0
1
37
3
7
4
86
145
0.65
6
1
0.5
67
4
8
7
81
244
0.29
7
1
1
55
3
9
5
83
208
0.61
8
2
0.5
77
3
8
8
81
279
0.27
9
2
1
59
2
7
6
85
225
0.43
2 3 4 5
-1
-1
TOF
/ wt%
1
a
Ac ce p
Catalytic performance of Co/Al2O3 Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 in FT synthesis.
b
CO conversion. Space time yield of C5+ products.
31
Page 31 of 40
(B)
(A)
Co/Al 2O3
Co/Al 2O3
Co/Ce/Al 2O3
ip t
Co/Y/Al 2O3
H2 consumption / a.u.
Co/Mo/Al 2O3 Co/Mg/Al 2O3
Co/Zr/Al 2O3
cr
Co/Ti/Al 2O3
Co/Sr/Al 2O3
Co/Ba/Al 2O3
us
H2 consumption / a.u.
Co/Zn/Al2O3
Co/Mn/Al 2O3 Co/V/Al 2O3
Temperature / ˚C
Co/Ca/Al 2O3
100 200 300 400 500 600 700 800 900
M
100 200 300 400 500 600 700 800 900
an
Co/La/Al 2O3
Temperature / ˚C
Ac ce p
te
d
Fig. 1. H2-TPR profiles of Co/Al2O3 and Co/M(1)/Al2O3 catalysts.
32
Page 32 of 40
(A) La
0.5
Co/M/Al2O3
0
V
40
50
60
70
80
cr
Co/La+V/Al2O3
ip t
1
90
us
Co surface area / m2 g-1
1.5
Co reduction degree / % (B)
an
1.5
M
1
0.5
Co/La+V/Al2O3
d
Co dispersion / %
La
Co/M/Al2O3
V
te
0
40
50
60
70
80
90
Ac ce p
Co reduction degree / %
Fig. 2. Influence of Co reduction degree on (A) Co surface area and (B) Co dispersion.
33
Page 33 of 40
90 Co/La+V/Al2O3 Co/M/Al2O3
80
La
ip t
70 60 50 V
40 30 16
17
18
19
20
21
us
15
cr
Co reduction degree / %
(A)
Co3O4 particle size / nm
Co/La+V/Al2O3 Co/M/Al2O3
M
1
La
0.5 V
d
Co surface area / m2 g-1
1.5
an
(B)
te
0
Ac ce p
15
16
17
18
19
20
21
Co3O4 particle size / nm
Fig. 3. Influence of Co3O4 particle size on (A) Co reduction degree and (B) Co surface area.
34
Page 34 of 40
(B)
us
cr
ip t
(A)
100 nm
an
100 nm
(C)
Ac ce p
te
d
M
(D)
100 nm
100 nm
35
Page 35 of 40
(F)
us
cr
ip t
(E)
20 nm
an
20 nm
(H)
Ac ce p
te
d
M
(G)
20 nm
20 nm
Fig. 4. TEM images of (A and E) Co/Al2O3, (B and F) Co/La(2)/Al2O3, (C and G) Co/V(0.5)/Al2O3 and (D and H) Co/La(2)+V(0.5)/Al2O3 catalysts.
36
Page 36 of 40
80 70 60 50 V 0.5% Mn
40 30 V 1%
Ca
Mo
20 50
60
70
80
90
us
40
ip t
Co/La+V/Al2O3 Co/M/Al2O3
cr
CO conversion / %
(A)
0.8 V 0.5% V 1%
TOF / s-1
0.6
Co/La+V/Al2O3
Mo
0.4
Mn
Co/M/Al2O3
M
(B)
an
Co reduction degree / %
te
0
d
0.2
40
50
60
70
80
90
Ac ce p
Co reduction degree / %
Fig. 5. Influence of Co reduction degree on (A) CO conversion and (B) TOF.
37
Page 37 of 40
(A)
Co/La+V/Al2O3 Co/M/Al2O3
70
50
Mn
V 0.5% V 1%
40
Ca
30
Mo
20 0.5
1
1.5
us
0
ip t
60
cr
CO conversion / %
80
2
-1
(B)
0.8 V 0.5%
Mo
0.4
Co/La+V/Al2O3 Co/M/Al2O3
M
TOF / s-1
0.6 V 1%
an
Co surface area / m g
Mn
0
d
0.2
0.5
1
1.5
Co surface area / m2 g-1
Ac ce p
te
0
Fig. 6. Influence of Co surface area on (A) CO conversion and (B) TOF.
38
Page 38 of 40
Co/Al2O3 Co/La(1)/Al2O3
ip t
H2 consumption / a.u.
Co/La(2)/Al2O3
Co/V(1)/Al2O3
us
Co/La(1)+V(0.5)/Al 2O3
cr
Co/V(0.5)/Al2O3
Co/La(1)+V(1)/Al 2O3
an
Co/La(2)+V(0.5)/Al 2O3
M
Co/La(2)+V(1)/Al 2O3
100 200 300 400 500 600 700 800 900
d
Temperature / ˚C
Ac ce p
te
Fig. 7. H2-TPR profiles of Co/Al2O3, Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 catalysts.
39
Page 39 of 40
Co
La
V
Ac ce p
te
O
d
M
an
Al
us
cr
ip t
50 nm
Fig. 8. TEM-EDS maps of Co/La(2)+V(0.5)/Al2O3 catalyst
40
Page 40 of 40
S0926-860X(15)00032-0 http://dx.doi.org/doi:10.1016/j.apcata.2015.01.017 APCATA 15209
To appear in:
Applied Catalysis A: General
Received date: Revised date: Accepted date:
24-9-2014 28-11-2014 10-1-2015
Please cite this article as: K. Shimura, T. Miyazawa, T. Hanaoka, S. Hirata, FischerndashTropsch synthesis over alumina supported cobalt catalyst: Effect of promoter addition, Applied Catalysis A, General (2015), http://dx.doi.org/10.1016/j.apcata.2015.01.017 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Co-loaded only
La(2%)
an
us
cr
CO conversion / %
ip t
Graphical abstract
V(0.5%)
La(2%)+ V(0.5%)
Ac ce p
te
d
M
Metal promoter
1
Page 1 of 40
Highlights ・Mg, Ca, Sr, Ba, Y, La, Ce, Ti, V, Mn, Zn, Zr and Mo were examined as promoters.
Ac ce p
te
d
M
an
us
cr
・La-loading increased Co surface area, while V-loading improved TOF.
ip t
・Activity of Co/Al2O3 catalysts was largely increased by co-loading of La and V.
2
Page 2 of 40
Fischer–Tropsch synthesis over alumina supported cobalt catalyst: Effect of
cr
Katsuya Shimura,* Tomohisa Miyazawa, Toshiaki Hanaoka and Satoshi Hirata
ip t
promoter addition
Biomass Refinery Research Center, National Institute of Advanced Industrial Science and Technology (AIST),
us
Kagamiyama 3-11-32, Higashihiroshima, Hiroshima 739-0046, Japan
an
*Corresponding author
M
Katsuya Shimura, E-mail: [email protected], Tel.: +81-82-420-8292
d
Abstract
te
In order to improve the activity of Co/Al2O3 catalyst for FT synthesis, loading effect of metal promoter on
Ac ce p
Al2O3 support was examined. In this study, total 13 kinds of metal (Mg, Ca, Sr, Ba, Y, La, Ce, Ti, V, Mn, Zn, Zr and Mo) were employed as promoters and θ-Al2O3 prepared by the calcination of commercial boehmite was used as support. Co/Al2O3 catalysts loaded with various metal promoters were prepared by a sequential impregnation method. Loading of Mn, V and Mo on Al2O3 support decreased Co reducibility and surface area of Co metal, but increased intrinsic activity per active Co metal sties (i.e. TOF). On the other hand, metal additives other than Mn, V and Mo varied Co reducibility and Co surface area, but did not influence TOF. High overall activity was obtained over Co/Al2O3 catalysts loaded with rare earth elements (e.g. La and Ce), which was due to the high surface area of Co metal. We also examined the co-loading of La and V promoters 3
Page 3 of 40
on Al2O3 support. Co/Al2O3 catalysts loaded with both La and V (Co/La+V/Al2O3) showed higher activity than those loaded with either La or V. CO conversion rate over the best Co/La(2)+V(0.5)/Al2O3 catalyst was
ip t
1.6 times as high as that over bare Co/Al2O3 catalyst. Over the Co/La+V/Al2O3 catalyst, La and V promoters
cr
cooperatively improved Co reducibility, Co surface area and TOF, resulting in an increase of overall catalytic
us
activity.
M
an
Keywords: Fischer-Tropsch synthesis, Cobalt, Alumina, Metal promoter
1. Introduction
d
It is very important for the realization of a sustainable society to establish practical methods of producing
te
clean alternative fuels. Fischer-Tropsch (FT) synthesis, which can synthesize hydrocarbon mixtures from
Ac ce p
syngas, is an excellent way to produce liquid fuel, since various carbon sources (e.g. coal, natural gas and biomass) can be used as feedstock [1, 2]. Moreover, clean liquid fuels without containing any sulfur, nitrogen and aromatic compounds can be obtained by FT synthesis. Thus, FT synthesis has gained renewed attention as one of the production methods of liquid fuels and various researches such as development of catalysts and reactors are currently underway to enhance the efficiency of this reaction system. Among the potential catalysts for FT synthesis, such as Fe, Co, Ni and Ru, Co-based catalysts are advantageous to the practical application due to the relatively high activity and selectivity to long-chain liner hydrocarbons, high resistance toward deactivation, low activity for the competitive water-gas shift reaction and lower price than Ru [3–5]. It 4
Page 4 of 40
has been accepted that FT synthesis over Co catalyst would be a structure-insensitive reaction and the activity of Co catalyst would depend on the number of exposed Co metal sites on the catalyst surface [6–8]. Therefore,
ip t
Co was typically deposited on supports having high surface area (e.g. SiO2, Al2O3, TiO2 and carbon materials)
cr
in order to increase the number of active Co metal sites [1–5]. However, several research groups recently found that turnover frequency (TOF) monotonically decreased with decreasing Co particle size when Co
us
particle size was smaller than ca. 10 nm [9–11]. These reports indicate that the optimum Co particle size
an
would exist for FT synthesis. Thus, the precise control of Co particle size is very important for the development of highly active Co catalysts for FT synthesis.
M
Due to the high thermal stability and the strong resistance to attrition, alumina is frequently employed as
d
support for Co-based FT catalysts. The great ability of Al2O3 support to stabilize small metal clusters would
te
suppress the aggregation of Co metal particles during catalytic reaction and may extend the catalyst lifetime.
Ac ce p
However, the strong interaction of Al2O3 support with Co oxide suppresses the reduction of Co oxide at the same time. This means a decrease in the number of active Co metal sites and results in the formation of low active catalysts. Thus, small amount of noble metals (e.g. Pt [12–14], Re [12, 13, 15], Ru [12, 13, 16], Pd [13, 17], Ir [18], Au [19. 20] and Ag [19, 21]) is often added to Co/Al2O3 catalysts, since they can facilitate the reduction of Co oxide and increase the number of active Co metal sites, presumably by hydrogen dissociation and spillover from the promoter surface. However, noble metals are not suitable to the industrial application due to their high cost. In order to improve the Co reducibility of Co/Al2O3 catalysts, use of other methods, such as the structural control of Al2O3 support [22–25] and the improvement in the preparation method of 5
Page 5 of 40
Co/Al2O3 catalysts [11, 26–28], is preferable. It is also known that loading small amount of metal promoters on Al2O3 support sometimes increases the
ip t
activity of Co/Al2O3 catalyst for FT synthesis. Various metal promoters such as Zr [29–31], alkali earth
cr
elements [32–36], rare earth elements [37–40], and others (P [41], SiO2 [42], B [43], Ti [44]) were reported to increase the activity of Co/Al2O3 catalyst. It is generally accepted that these metal additives would improve
us
the interaction of Al2O3 support with Co oxide and thereby increase the Co reducibility and the number of
an
active Co metal sites. Although various metal promoters were found for Co/Al2O3 catalyst as described above, effect of each metal promoter on the structure of Co particles and the activity of Co/Al2O3 catalyst is not well
M
understood. This is because experimental conditions (e.g. type of Al2O3 support, pre-reduction condition of
d
catalyst, type of reaction vessel and reaction condition) were different among researchers and there were few
te
studies to compare several metal promoters at the same time. If several Co/Al2O3 catalysts modified with
Ac ce p
different metal promoters were evaluated in the same condition, we can obtain the better understanding for the loading effect of each metal promoter on Co/Al2O3 catalyst. In the present study, loading effect of metal promoter on Al2O3 support was examined in order to prepare highly active Co/Al2O3 catalysts for FT synthesis. Total 13 kinds of metal were employed as promoters and θ-Al2O3 prepared from commercial boehmite [45] was used as support. Impact of each metal promoter on the structure of Co particles and the activity of Co/Al2O3 catalyst was systematically studied. We also examined the co-loading of La and V promoters on the Al2O3 support and found that La and V additives could cooperatively increase the overall activity of Co/Al2O3 catalyst for FT synthesis. 6
Page 6 of 40
2. Experimental
ip t
2.1. Catalyst preparation
cr
Co/Al2O3 catalysts modified with various promoters (named Co/M(x)/Al2O3) were prepared by a sequential impregnation method. M and x represent the kind of metal additives and the loading amount (wt%),
us
respectively. First, M/Al2O3 was prepared from θ-Al2O3 (BET specific surface area: 84 m2 g-1) powders and an
an
aqueous solution of each metal promoter by an impregnation method. θ-Al2O3 was obtained by calcination of commercial boehmite (AlO(OH)·nH2O, Wako) at 1000 ˚C for 10 h [45]. The precursors of metal additives Mg(NO3)26H2O,
Ca(NO3)24H2O,
Sr(NO3)2,
Ba(NO3)2,
M
were
Y(NO3)36H2O,
La(NO3)36H2O,
d
Ce(NO3)36H2O, (NH4)2TiO(C2O4)2, NH4VO3, Mn(NO3)26H2O, Zn(NO3)26H2O, ZrO(NO3)22H2O, and
te
(NH4)6Mo7O244H2O. Loading amount of each metal was typically 1 wt%. After impregnation, the obtained
Ac ce p
powders were dried at 100 ˚C for 12 h, followed by calcination in air at 500 ˚C for 3 h. Then, Co (20 wt%) was loaded on M/Al2O3 by an impregnation method using Co(NO3)26H2O (99.5%, Wako) as Co precursor. Finally, the obtained powders were dried at 100 ˚C for 12 h and calcined in air at 400 ˚C for 3 h. 2.2. Characterization
Powder X-ray diffraction (XRD) pattern was recorded at room temperature on a Rigaku diffractometer RINT 2500 TTRIII using Cu Kα radiation (50 kV, 300 mA). The mean particle size of Co3O4 (dCo3O4) was calculated from the diffraction line at 2θ = 36.9 degree with the Scherrer equation. The obtained particle size of Co3O4 could be used to calculate that of Co metal (dCo) after hydrogen reduction pretreatment by the following 7
Page 7 of 40
formula (Eq. (1)) [14]. dCo ＝ 0.75 × dCo3O4
Eq. (1)
ip t
Temperature programmed reduction under H2 (H2-TPR) was carried out with BELCAT-B (BEL Japan Inc.). The calcined catalyst (0.10 g) was mounted in a quartz cell and heated up to 900 ˚C in a flow of 5% H2/Ar (30
cr
ml min-1). The heating rate was 10 ˚C min-1. The reduction degree of supported cobalt was calculated from the
us
amount of H2 consumption during H2 reduction pretreatment at 400 ˚C for 6 h. The effluent gas was passed
an
through a 5A molecular sieve trap to remove the produced water before reaching a thermal conductivity detector.
M
Hydrogen chemisorption experiments were performed on BELCAT-B. Before measurement, samples (0.15
d
g) were reduced at 400 ˚C for 6 h in a flow of 5% H2/Ar (15 ml min-1) and held at 400 ˚C for 1 h in a flow of
te
Ar (30 ml min-1) to desorb the residual chemisorbed hydrogen. After cooling the sample down to 100 ˚C in a
Ac ce p
flow of Ar, H2 chemisorption measurements were started. Corrected dispersion (Dcorr, Eq. (2)) and surface area of Co metal were calculated according to the method reported in the literature [46]. Eq. (2)
Transmission electron microscopy (TEM) images of the reduced and passivated Co/Al2O3 catalysts were recorded by a JEOL electron microscope (JEM-3000F, 300 kV) equipped with energy dispersive X-ray spectroscopy (EDS). In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) spectra were recorded with a JASCO FT/IR-4100 spectrometer equipped with a diffuse reflectance attachment and a MCT detector. The 8
Page 8 of 40
sample was placed in an infrared cell equipped with CaF2 windows. The spectra were obtained by collecting 64 scans at 4 cm-1 resolution. Before measurement, samples were reduced at 450 ˚C for 4 h in a flow of 5%
ip t
H2/Ar (40 ml min-1) and held at 450 ˚C for 0.5 h in a flow of N2 (40 ml min-1) to desorb the residual
cr
chemisorbed hydrogen. Subsequently, the system was cooled to 30 ˚C in a flow of N2 and CO was introduced. After adsorption of CO, the catalyst surface was purged with N2 to remove gaseous CO and then IR spectra
us
were recorded.
an
2.3. Typical procedures of catalytic reactions
FT synthesis was performed with a continuous stirred tank reactor in a similar way to the previous study [47].
M
Before reaction, catalyst (2.5 g) was in-situ reduced in a flow of H2 (40 ml min-1) at 400 ˚C and 6 h. Since the
d
Co reduction degree largely influences the number of active Co metal sites, catalysts should be reduced in the
te
optimum condition. In the present study, we roughly examined the effect of reduction condition on Co
Ac ce p
reduction degree using bare Co/Al2O3 catalyst. Co-based FT catalysts are typically reduced in H2 at 300–500 ˚C before reaction. In the present study, Co reduction degree of bare Co/Al2O3 catalyst increased from 42% to 64%, when reduction temperature increased from 300 to 400 ˚C. Therefore, reduction at 400 ˚C would be preferable to reduction at 300 ˚C. However, we could not increase the reduction temperature more than 400 ˚C due to the low performance of electric furnace of the reactor. Co reduction degree also depended on the reduction time. When reduction time was longer than 3 h, Co reduction degree was almost constant. From these results, we carried out the H2 reduction pretreatment at 400 ˚C for 6 h. After the reduction pretreatment, reactor was cooled down to room temperature and purged by N2 gas. Then, 9
Page 9 of 40
n-hexadecane (80 g) solvent was added. The reaction was carried out at 230 ˚C and 1.0 MPa for 8 h in a flow of synthetic gas (100 ml min-1). The syngas used for this reaction was obtained by the gasification of woody
ip t
biomass, followed by the gas purification and the composition adjustment [48]. Since the product gas obtained
cr
by the gasification of woody biomass contained impurities such as sulfur and tarry compounds, we removed them by passing the gas through scrubber and desulfurization tower [48, 49]. After gas cleaning, the
us
concentration of sulfur compounds decreased to less than 5 ppb [49]. Composition of this biomass-derived
an
syngas was confirmed by gas chromatography (GC) as follows: H2 (60.0%), CO (31.3%), CH4 (4.7%) and N2 (4.0%). The effluent gas after FT synthesis was analyzed by on-line GC. A thermal conductivity detector
M
(TCD) with a Porapak-Q column was used to analyze inorganic gases (H2, CO, CO2, CH4 and N2). Light
d
hydrocarbons (C1–C4) were analyzed by a flame ionization detector (FID) with a RT-QPLOT capillary
te
column. Hydrocarbons dissolved in the solvent and cooled in the trap were analyzed by GC-FID equipped
Ac ce p
with a UA-DX30 capillary column.
3. Results and discussion
3.1. Characterization of Co/M/Al2O3 catalysts Co(20%)/M(1%)/Al2O3 catalysts, which were prepared by a sequential impregnation method, were characterized by XRD, H2-TPR, H2-chemisorption and TEM. First, Co/M/Al2O3 catalysts before H2 reduction pretreatment were analyzed by XRD. From XRD patterns of all Co/M/Al2O3 catalysts, only peaks assigned to θ-Al2O3 and Co3O4 were observed (Fig S1). Peaks of other Co compounds (e.g. CoO and Co aluminate) and 10
Page 10 of 40
those derived from metal additives were not observed. Particle sizes of Co3O4 and Co metal on all Co/M/Al2O3 catalysts were estimated from the diffraction line at 36.9 degree using the Scherrer equation and
ip t
Eq. (1). The results were listed in Table 1. Particle sizes of Co3O4 and Co metal of Co/M/Al2O3 catalysts were
cr
15.7–19.9 and 11.7–14.9 nm, respectively, and varied with the kind of metal promoters. The addition of Ca, Y, La, Ce, Ti and Zn had little influence on Co particle size. Loading of Mg, Sr, Ba, Mn, Zr and Mo slightly
us
decreased Co particle size. V additives largely decreased Co particle size. Difference in Co particle size is
an
attributed to the variation of metal-support interaction with the kind of metal promoters. It is general that Co dispersion increases and Co particle size decreases with increasing the extent of metal-support interaction.
M
This means that interaction of Al2O3 support with Co oxide is the strongest over Co/V/Al2O3 catalyst among
d
the examined samples.
te
Fig. 1 shows H2-TPR profiles of bare Co(20%)/Al2O3 and various Co(20%)/M(1%)/Al2O3 catalysts. It is
Ac ce p
generally accepted that peak observed at 300–400 ˚C corresponds to the reduction of Co3O4 to CoO and peaks at 400–600 ˚C correspond to the reduction of CoO to Co metal [3]. Similarly, peaks observed at 600 ˚C and higher temperatures should originate from the reduction of barely reducible Co species such as cobalt aluminate. Peak observed at 300–400 ˚C of all Co/M/Al2O3 catalysts was very similar to each other. This shows that loading metal promoters on Al2O3 support has little influence on the reduction of Co3O4 to CoO. However, peaks observed at more than 400 ˚C were largely influenced by the kind of metal promoters. When V, Mn, Mg and Mo were loaded on Al2O3 support, peak area at more than 600 ˚C became much larger than that of bare Co/Al2O3 catalyst (Fig. 1 (A)). This indicates that these metal promoters should suppress the 11
Page 11 of 40
reduction of Co oxide. On the other hand, loading of Ca and La clearly decreased the area of high temperature peaks, indicating that these metal additives could promote the reduction of Co oxide (Fig. 1 (B)). Results of
ip t
H2-TPR clearly showed that even low (1 wt%) loading of metal promoters on Al2O3 support could greatly
cr
influence the interaction of Al2O3 support with Co oxide and vary the Co reducibility of Co/Al2O3 catalyst as follows. The addition of V, Mn, Mg and Mo would strengthen the interaction of Al2O3 support with Co oxide.
us
The strong metal-support interaction would promote the solid-solid reaction of Al2O3 support with Co
an
particles during the thermal pretreatment and suppress the aggregation of Co particles, resulting in the formation of small Co particles with low reducibility. In contrast, La and Ca promoters would reduce the
M
surface defective sites on Al2O3 support and weaken the interaction of Al2O3 support with Co particles. The
d
weak metal-support interaction would promote the particle growth of Co during the thermal pretreatments,
te
resulting in the formation of relatively large Co particles with high reducibility.
Ac ce p
Reduction degree of Co (Dred), surface area of Co metal (SA) and corrected dispersion of Co metal (Dcorr) were measured by H2-TPR and H2 chemisorption. The results were listed in Table 1. Dred, SA and Dcorr of Co/M/Al2O3 catalysts were 40–87%, 0.21–1.37 m2 g-1 and 0.31–1.26%, respectively, and depended largely on the kind of metal additives. As estimated from the result of H2-TPR (Fig. 1), Co/Al2O3 catalysts loaded with V, Mn, Mg and Mo showed low Co reduction degree (40–54%), while Co/Al2O3 catalysts loaded with La and Ca showed high Co reduction degree (76–87%). Surface area of Co metal tended to increase with increasing Co reduction degree and reached a plateau when Co reduction degree was higher than 65% (Fig. 2 (A)). Among the examined samples, Co/La/Al2O3 catalyst showed the highest Co surface area (1.37 m2 g-1), while 12
Page 12 of 40
Co/V/Al2O3 catalyst showed the lowest Co surface area (0.21 m2 g-1). Dispersion of Co metal also depended on Co reduction degree. As Co reduction degree increased, Co dispersion first increased due to increase in the
ip t
number of surface Co metal sites (Fig. 2 (B)). The highest value was obtained over Co/Al2O3 catalyst having
cr
moderate reduction degree (64%). However, further increase of reduction degree slightly decreased Co dispersion. This is probably because the aggregation of Co particles would be promoted over Co/Al2O3
us
catalysts having high reduction degree during the thermal pretreatment.
an
Reduction degree of Co and surface area of Co metal were plotted against particle size of Co3O4 (Fig. 3). Co reduction degree tended to increase with an increase of Co3O4 particle size (Fig. 3 (A)). In other words,
M
reduction of large Co3O4 particles was easier than that of small Co3O4 particles. Co surface area first increased
d
until 18.8 nm and then decreased with increasing Co3O4 particle size (Fig. 3 (B)). If Co reduction degree of all
te
Co/Al2O3 catalysts is same, Co surface area should increase with decreasing Co particle size. However, as
Ac ce p
shown in Fig. 3 (A), the decrease of Co3O4 particle size also decreased Co reducibility and suppressed the formation of surface Co metal sites. Therefore, the highest Co surface area was obtained over Co/Al2O3 catalysts having moderate Co3O4 particle size due to the trade-off effect of particle size and reduction degree. In order to examine the distribution of Co particles on Al2O3 support, Co/Al2O3, Co/La(2)/Al2O3 and Co/V(0.5)/Al2O3 catalysts were analyzed by TEM. Two representative TEM images measured at low (Fig. 4 (A)–(D)) and high magnification (Fig. 4 (E)–(H)) were shown for each sample. In Fig. 4, black spots show the area having high concentration of Co. Co particle size, Co reduction degree and Co surface area of Co/La(2)/Al2O3 catalyst were comparable to those of Co/La(1)/Al2O3 catalyst (Table 2, entries 2, 3). 13
Page 13 of 40
Properties of Co particle on Co/V(0.5)/Al2O3 catalyst were also similar with those of Co on Co/V(1)/Al2O3 catalyst (Table 2, entries 4, 5). As shown in Fig. 4 (E)–(G), Co particles smaller than 20 nm were mainly
ip t
observed from TEM images of all samples. Although Co surface area measured by H2 chemisorption was
cr
largely varied with the kind of metal additives in the range of 0.2–1.4 m2 g-1, no clear differences in Co particle size and distribution were observed from their TEM images. Therefore, we suggest that influence of
us
metal promoter on Co particle size would be small. The variation of Co surface area with the kind of metal
an
promoter is mainly attributed to the differences in Co reduction degree of each catalyst. Distribution of Co particles and metal promoters on Al2O3 support was further analyzed by TEM-EDS. Figs.
M
S2 and S3 show TEM-EDS maps of Co/La(2)/Al2O3 and Co/V(0.5)/Al2O3 catalysts. For both samples,
d
distribution of Co particles on Al2O3 support was rather heterogeneous. The distribution of La and V promoter
te
was also inhomogeneous, although loading amount of these promotes was as small as 0.5–2 wt%. However, it
Ac ce p
is interesting to note that distribution of Co particles was similar with that of metal promotes. In other words, Co particles tended to exist in the area where concentration of metal promoter was high. This result indicates that metal promoters would contribute to the generation and dispersion of Co particles. As conclusion of this section, loading small amount of metal promoters modified the interaction of Al2O3 support with Co3O4 particles and varied Co3O4 particle size, Co reduction degree and Co surface area. Co reduction degree and Co surface area tended to increase with increasing Co3O4 particle size. Co/La/Al2O3 catalyst showed the highest Co surface area, while Co/V/Al2O3 catalyst showed the lowest Co reduction degree and Co surface area. 14
Page 14 of 40
3.2. Catalytic performance of Co/M/Al2O3 and structure-activity relationship
ip t
FT synthesis was performed on various Co/Al2O3 catalysts using a continuous stirred tank reactor at 230 ˚C
cr
and 1.0 MPa for 8 h. Fig. S4 shows time course of CO conversion and space time yield of C5+ products (STY of C5+) over bare Co/Al2O3 catalysts as the representative. Deactivation of catalysts was hardly observed for
us
all samples due to the short reaction time. However, prolonged reaction time should absolutely deactivate the
an
catalyst due to the structural changes of cobalt particles, such as sintering, surface reconstruction and carbization, as reported by various research groups [50, 51]. Therefore, we can only discuss the effect of metal
M
promoter on initial catalytic activity (8 h after the reaction started) in this paper, since catalytic performance
d
may be largely changed during a long time reaction test. We will examine the lifetime of various Co/M/Al2O3
te
catalysts and report the effect of each metal promoter on the catalyst stability in the future. However, it should
Ac ce p
be noted that in our previous study using similar reaction condition and reaction setup, the activity of Co/SiO2 and Co/Mn+Zr/SiO2 catalysts was almost constant for 1 week owing to the precise control of reaction temperature and the use of hexadecane solvent [52]. Thus, we expect that deactivation of the Co/M/Al2O3 catalysts may be quite slow in the present reaction condition and reaction setup. Table 3 shows CO conversion, product selectivity, STY of C5+ and TOF per surface Co metal atom on bare Co(20%)/Al2O3 and various Co(20%)/M(1%)/Al2O3 catalysts. TOF was calculated from the results of H2 chemisorption and catalytic reaction. CO conversion and STY of C5+ products over Co/M/Al2O3 catalysts were 33–55% and 126–200 mg g-1 h-1, respectively. Addition of La, Ce and Zr increased CO conversion and 15
Page 15 of 40
STY of C5+ products, while metal additives other than La, Ce and Zr decreased them. Among the catalysts listed in Table 3, Co/Ce/Al2O3 catalyst showed the highest CO conversion and STY of C5+ products.
ip t
Selectivities of CO2, CH4, C2–C4 and C5+ products over Co/M/Al2O3 catalysts were 2–5%, 6–10%, 4–10% and
cr
78–86%, respectively. Co/V/Al2O3 catalyst showed the highest C5+ selectivity, while Co/Zn/Al2O3 catalyst showed the lowest one. The similar results were also reported for Co/TiO2 catalyst [53]. Previously, our
us
research group studied the loading effect of metal promoters on Co/TiO2 catalyst and found that Co/V/TiO2
an
catalyst showed high C5+ selectivity and Co/Zn/TiO2 catalyst showed low C5+ selectivity [53]. However, we need to compare product selectivity at the same conversion level in order to understand the effect of metal
M
additives on product selectivity more properly, since product selectivity depends on CO conversion [54]. TOF
d
largely depended on the kind of metal additives. When Mg, Ca, Sr, Ba, Y, La, Ce, Ti, Zn and Zr were used as
te
additives, TOF was not influenced by the kinds of metal promoter and almost the same as that of bare
Ac ce p
Co/Al2O3 catalyst. In contrast, TOF of Co/Al2O3 catalysts loaded with V, Mn and Mo (0.36–0.65 s-1) was much higher than that of bare Co/Al2O3 catalyst (0.15 s-1). This result is not usual, because FT synthesis over Co catalysts is a structure-insensitive reaction in many cases [6–8]. However, a few research groups reported that Mn-loading increased TOF of Co/C [55] and Co/TiO2 [56, 57] catalysts. They propose that Mn2+ promoters would withdraw electrons from Co metal sites and vary the adsorption ability and the hydrogenation/dehydrogenation ability of Co metal sites [58]. The facts that V-loading decreased the number of active metal sites but increased the TOF were also reported for CO hydrogenation over Ru/Al2O3 [59], Ni/Al2O3 [60] and Rh/SiO2 [61] catalysts. From the 16
Page 16 of 40
analysis of reaction kinetics [59] and the FT-IR study [61], it was proposed that role of V promoter would be to promote the CO dissociation and the hydrocarbon chain growth. Thus, we also carried out the FT-IR study
ip t
in order to examine the effect of V promoter on CO adsorption. Fig. S5 shows FT-IR spectra of CO adsorbed on Co/Al2O3 and Co/V(0.5)/Al2O3 catalysts at room temperature. The IR spectrum of Co/Al2O3 catalyst exhibited a
cr
strong band at 2050 cm-1, which assigned to the linear adsorbed CO on surface Co metal atom. Very small bands
us
assigned to the bridge-bonded CO with metallic Co (2000–1800 cm-1) were also observed from the spectrum of
an
Co/Al2O3. Co/V(0.5)/Al2O3 catalyst also exhibited the clear FT-IR spectrum of CO adsorbed on Co metal, although V-loading largely disturbed the adsorption of H2 (Table 2). V-loading did not influence the position and intensity of
M
the linearly bonded CO band. However, the intensity of the bridge bonded CO band was clearly increased by
d
V-loading. From the result of IR study, we suggest that V promoter may influence the state and amount of the
te
adsorbed CO, resulting in an improvement of specific activity of Co/Al2O3 catalyst. Further study is now carried
Ac ce p
out to clarify the role of V promoter.
Reduction degree of Co and surface area of Co metal were important factors to determine the activity of Co catalyst. Thus, influence of these parameters on CO conversion and TOF was examined. Results were shown in Figs. 5 and 6. Overall, CO conversion did not show any clear correlation with Co reduction degree (Fig. 5 (A)). CO conversion did not also depend on Co surface area (Fig. 6 (A)). However, except for Co/Al2O3 catalysts loaded with V, Mn, Mo and Ca, CO conversion tended to increase with increasing Co reduction degree and Co surface area. Low activity of Co/Ca/Al2O3 catalyst is probably due to the poisoning effect of Ca ions [62, 63]. The different tendency of Co/Al2O3 catalysts loaded with V, Mn and Mo is 17
Page 17 of 40
attributed to their high TOF, as listed in Table 3. TOF largely depended on Co reduction degree and Co surface area. When Co reduction degree was higher than 55%, TOF was not varied with reduction degree (Fig. 5 (B)).
ip t
However, TOF of Co/M/Al2O3 catalysts having low Co reduction degree, such as Co/V/Al2O3, Co/Mn/Al2O3
cr
and Co/Mo/Al2O3 catalysts, was much higher than that of bare Co/Al2O3 catalyst. Increase of TOF with decreasing Co reduction extent was reported by Lee and co-workers [64]. They proposed that the electronic
us
effect of subsurface Co oxide may influence the high TOF. Similarly, TOF was not varied with Co surface
an
area, when CO surface area was higher than 0.7 m2 g-1 (Fig. 6 (B)). However, TOF of Co/M/Al2O3 catalysts having low Co surface area, such as Co/V/Al2O3, Co/Mn/Al2O3 and Co/Mo/Al2O3 catalysts, was much higher
M
than that of bare Co/Al2O3 catalyst. Results of Fig. 6 were consistent with those of Fig. 5, since Co/M/Al2O3
d
catalysts with lower Co reduction degree tended to show lower Co surface area (Fig. 2 (A)). In the present
te
study, we could not show clear reasons why addition of V, Mn and Mo largely increased TOF. However, we
Ac ce p
tentatively propose that high TOF of Co/Al2O3 catalysts loaded with V, Mn and Mo is due to not the size of Co particles but the electronic effect of unreduced Co oxide and metal additives. This is because Co particle size of Co/Al2O3 catalysts prepared in the present study is in the range of 11.7–14.9 nm (Table 1) and TOF of supported Co catalysts is generally constant for Co particle size larger than 10 nm [9–11]. Results of characterization and reaction tests showed that the examined metal additives could be divided into two categories: V, Mn and Mo decreased Co reduction degree and Co surface area, but increased TOF. Metal additives other than V, Mn and Mo varied Co reduction degree and Co surface area, but did not influence TOF. 18
Page 18 of 40
3.3. Catalytic performance and physicochemical property of Co/La+V/Al2O3
ip t
The overall activity of Co/Al2O3 catalyst depends on the number of surface Co metal sites and the TOF.
cr
Improvement of both parameters is necessary for an increase of overall catalytic activity. As described in section 3.1 and 3.2, La-loading increased Co surface area, while V-loading improved TOF. Thus, co-loading of
us
La and V promoters on Al2O3 support was examined. Results of FT synthesis over Co/Al2O3, Co/La/Al2O3,
an
Co/V/Al2O3 and Co/La+V/Al2O3 catalysts were shown in Table 4. CO conversion and STY of C5+ products over Co/La/Al2O3 catalysts were 50–54% and 188–192 mg g-1 h-1, respectively, and a little larger than those of
M
bare Co/Al2O3 catalyst (Table 4, entries 1–3). CO conversion and STY of C5+ products over Co/V/Al2O3
d
catalysts were 37–53% and 145–192 mg g-1 h-1, respectively (Table 4, entries 4, 5). As V-loading amount
te
increased, CO conversion and STY of C5+ products first increased until 0.5% and then decreased. CO
Ac ce p
conversion and STY of C5+ products over Co/La+V/Al2O3 catalysts were 55–77% and 208–279 mg g-1 h-1, respectively, and higher than those of Co/Al2O3, Co/La/Al2O3 and Co/V/Al2O3 catalysts (Table 4, entries 6–9). The highest activity was obtained over Co/La(2)+V(0.5)/Al2O3 catalyst. CO conversion and STY of C5+ products over this catalyst were 1.6 times higher than those of bare Co/Al2O3 catalyst. Product selectivities of CO2, CH4, C2–C4 and C5+ products over Co/La+V/Al2O3 catalysts were 2–4%, 7–9%, 5–8% and 81–85%, respectively. Co-loading of La and V had little impact on product selectivity. TOF of Co/La+V/Al2O3 catalysts (0.27–0.61 s-1) was higher than that of Co/La/Al2O3 catalysts (0.14–0.15 s-1) but lower than that of Co/V/Al2O3 catalysts (0.65–0.68 s-1). As V-loading amount increased, TOF of Co/La+V/Al2O3 catalysts 19
Page 19 of 40
increased. However, TOF of Co/La+V/Al2O3 catalysts decreased with increasing La-loading amount. Results of Table 4 clearly show that co-loading of La and V is effective for an increase in the overall activity of
ip t
Co/Al2O3 catalyst (see also Figs. 5 and 6).
cr
Co/La+V/Al2O3 catalysts were characterized by XRD, H2-TPR, H2 chemisorption and TEM-EDS in order to clarify the reason for their high catalytic activity. Fig. 7 shows H2-TPR profiles of Co/Al2O3, Co/La/Al2O3,
us
Co/V/Al2O3 and Co/La+V/Al2O3 catalysts. Spectra of Co/La/Al2O3 and Co/V/Al2O3 catalysts showed that the
an
area of high temperature peaks more than 600 ˚C was decreased by La-loading but increased by V-loading. H2-TPR profiles of Co/La+V/Al2O3 catalysts were relatively similar with those of Co/V/Al2O3 catalysts.
d
V-loading amount decreased Co reducibility.
M
Increase of La-loading amount increased Co reducibility of Co/La+V/Al2O3 catalysts, while increase of
te
Co particle size was determined by XRD (Table 2). Co3O4 particle sizes of Co/La/Al2O3, Co/V/Al2O3 and
Ac ce p
Co/La+V/Al2O3 catalysts were 18.8–19.1, 15.7–18.7 and 16.1–18.6 nm, respectively. Co3O4 particle size of Co/V/Al2O3 catalysts decreased with increasing V-loading amount and Co/V(1)/Al2O3 catalyst showed the smallest size (Table 2, entries 4, 5). However, co-loading of La suppressed the decrease of Co3O4 particle size by V-loading to some extent (Table 2, entries 6–9). Reduction degree of Co, surface area of Co metal and corrected dispersion of Co metal were also listed in Table 2. When La-loading amount of Co/La/Al2O3 catalyst was increased from 0 to 2 wt%, Co reduction degree increased from 64% to 82% but Co dispersion decreased from 1.21% to 1.06% (Table 2, entries 1–3). Co surface area of Co/La/Al2O3 catalyst increased with increasing La-loading amount until 1% and then became constant. Co reduction degree, Co surface area and 20
Page 20 of 40
Co dispersion of Co/V/Al2O3 catalysts were 40%, 0.21–0.29 m2 g-1 and 0.35–0.48%, respectively. V-loading largely decreased these three parameters. Co reduction degree, Co surface area and Co dispersion of
ip t
Co/La+V/Al2O3 catalysts were higher than those of Co/V/Al2O3 catalysts but lower than those of Co/La/Al2O3
cr
catalysts (see also Figs. 2 and 3). Results of Table 2 show that Co/La+V/Al2O3 catalysts have properties intermediate between Co/La/Al2O3 and Co/V/Al2O3 catalysts. Therefore, we propose the reasons for high
us
catalytic activity of Co/La+V/Al2O3 as follows: La-loading is effective for an improvement in Co reducibility
an
and Co surface area. V-loading is effective for an increase of TOF. Over Co/La+V/Al2O3 catalysts, these positive effects are obtained at the same time, resulting in an increase of the overall catalytic activity.
M
Co/La+V/Al2O3 catalysts were also analyzed by TEM. As shown in Fig. 4 (H), Co particles smaller than 20
d
nm were mainly observed from TEM images. TEM images of Co/La+V/Al2O3 catalyst resembled those of
te
Co/Al2O3, Co/La/Al2O3 and Co/V/Al2O3 catalysts, indicating that co-loading of La and V had a small
Ac ce p
influence on the distribution of Co particles on Al2O3 support. Fig. 8 shows the TEM-EDS maps of Co/La(2)+V(0.5)/Al2O3 catalyst. La and V promoters were not homogeneously dispersed on A2O3 support. However, distribution of La and V promoters was similar to that of Co particles, as in the case of Co/La(2)/Al2O3 and Co/V(0.5)/Al2O3 catalysts (Figs. S2 and S3). This means that Co particles tend to be formed on the area where concentrations of La and V are high. We suggested that over Co/La(2)+V(0.5)/Al2O3 catalyst, the good contact of La and V promoters with Co particles would effectively improve the overall activity. La promoter did not influence the specific activity. However, La-loading could increase the number of active Co metal sites, since La promoter weakened the interaction of Al2O3 support 21
Page 21 of 40
with Co oxide and greatly increased the Co reducibility. In contrast, V-loading could largely improve the specific activity probably due to the promotion of CO activation, although V promoter also strengthened the
ip t
interaction of Al2O3 support with Co oxide and decreased the Co reducibility (i.e. number of active Co metal
cr
sites). The good contact of La and V promoters with Co particles enabled to improve both the number and
us
property of active Co metal sites, resulting in an enhancement of overall catalytic activity.
an
Conclusions
Total 13 kinds of metals (Mg, Ca, Sr, Ba, Y, La, Ce, Ti, V, Mn, Zn, Zr and Mo) were examined as additive of
M
Al2O3 support. Impact of each metal promoter on the structure of Co particles and the activity of Co/Al2O3
d
catalyst for FT synthesis was studied in details. Among the examined metal promoters, V, Mn and Mo
te
decreased Co reducibility and surface area of Co metal. However, TOF of Co/Al2O3 catalysts loaded with V,
Ac ce p
Mn and Mo was much higher than that of bare Co/Al2O3 catalyst. In contrast, metal additives other than V, Mn and Mo varied Co reducibility and surface area of Co metal without influencing the TOF. High overall activity for FT synthesis was obtained, when rare earth elements such as La and Ce were used as additives. Activity of Co/Al2O3 catalyst was further increased by loading of both La and V promoters on the Al2O3 support. The best catalyst was Co/La(2)+V(0.5)/Al2O3 and CO conversion rate over this catalyst was 1.6 times higher than that over bare Co/Al2O3 catalyst. Over the Co/La+V/Al2O3 catalyst, La and V promoters cooperatively increased Co surface area and TOF, resulting in an increase of overall catalytic activity.
22
Page 22 of 40
Acknowledgment. This work was partially supported by a research grand from the Japan Prize Foundation. The authors are deeply grateful to Mr. Hideyuki Yokoyama and Ms. Tomomi Monzen for their experimental
Ac ce p
te
d
M
an
us
cr
ip t
assistance.
23
Page 23 of 40
References [1] E.F. Sousa-Aguiar, F.B. Noronha, A. Faro, Jr., Catal. Sci. Technol. 1 (2011) 698–713.
ip t
[2] R. Luque, A.R. de la Osa, J.M. Campelo, A.A. Romero, J.L. Valverde, P. Sanchez, Energy Environ. Sci. 5
cr
(2012) 5186–5202. [3] A.Y. Khodakov, W. Chu, P. Fongarland, Chem. Rev. 107 (2007) 1692–1744.
us
[4] Q. Zhang, J. Kang, Y. Wang, ChemCatChem 2 (2010) 1030–1058.
an
[5] H. Jahangiri, J. Bennett, P. Mahjoubi, K. Wilson, S. Gu, Catal. Sci. Technol. 4 (2014) 2210–2229. [6] S.W. Ho, M. Houalla, D.M. Hercules, J. Phys. Chem. 94 (1990) 6396–6399.
M
[7] B.G. Johnson, C.H. Bartholomew, D.W. Goodman, J. Catal. 128 (1991) 231–247.
d
[8] E. Iglesia, S.L. Soled, R.A. Fiato, J. Catal. 137 (1992) 212–224.
te
[9] G.L. Bezemer, J.H. Bitter, H.P.C.E. Kuipers, H. Oosterbeek, J.E. Holewijn, X. Xu, F. Kapteijn, A.J. van
Ac ce p
Dillen, K.P. de Jong, J. Am. Chem. Soc. 128 (2006) 3956–3964. [10] G. Prieto, A. Martínez, P. Concepción, R. Moreno-Tost, J. Catal. 266 (2009) 129–144. [11] J. Park, Y. Lee, P.R. Karandikar, K. Jun, K. Ha, H. Park, Appl. Catal. A 411–412 (2012) 15–23. [12] G. Jacobs, T.K. Das, Y. Zhang, J. Li, G. Racoillet, B.H. Davis, Appl. Catal. A 233 (2002) 263–281. [13] W. Ma, G. Jacobs, R.A. Keogh, D.B. Bukur, B.H. Davis, Appl. Catal. A 437–438 (2012) 1–9. [14] D. Schanke, S. Vada, E.A. Blekkan, A.M. Hilmen, A. Hoff, A. Holmen, J. Catal. 156 (1995) 85–95. [15] T.K. Das, G. Jacobs, P.M. Patterson, W.A. Conner, J. Li, B.H. Davis, Fuel 82 (2003) 805–815. [16] A. Kogelbauer, J.G. Goodwin, Jr., R. Oukaci, J. Catal. 160 (1996) 125–133. 24
Page 24 of 40
[17] G. Jacobs, W. Ma, P. Gao, B. Todic, T. Bhatelia, D.B. Bukur, S. Khalid, B.H. Davis, Top Catal. 55 (2012) 811–817.
ip t
[18] D.C. Cronauer, G. Jacobs, L. Linganiso, A.J. Kropf, J.W. Elam, S.T. Christensen, C.L. Marshall, B.H.
cr
Davis, Catal. Lett. 141 (2011) 968–976.
[19] G. Jacobs, M.C. Ribeiro, W. Ma, Y. Ji, S. Khalid, P.T.A. Sumodjo, B.H. Davis, Appl. Catal. A 361
us
(2009) 137–151.
an
[20] K. Jalama, N.J. Coville, H. Xiong, D. Hildebrandt, D. Glasser, S. Taylor, A. Carley, J.A. Anderson, G.J. Hutchings, Appl. Catal. A 395 (2011) 1–9.
M
[21] T. Jermwongratanachai, G. Jacobs, W. Ma, W.D. Shafer, M.K. Gnanamani, P. Gao, B. Kitiyanan, B.H.
te
165–180.
d
Davis, J.L.S. Klettlinger, C.H. Yen, D.C. Cronauer, A.J. Kropf, C.L. Marshall, Appl. Catal. A 464–465 (2013)
Ac ce p
[22] H. Xiong, Y. Zhang, S. Wang, J. Li, Catal. Commun. 6 (2005) 512–516. [23] S. Rane, Ø. Borg, J. Yang, E. Rytter, A. Holmen, Appl. Catal. A 388 (2010) 160–167. [24] J. Sun, C. Xing, H. Xu, F. Meng, Y. Yoneyama, N. Tsubaki, J. Mater. Chem. A 1 (2013) 5670–5678. [25] J. Zhang, J. Chen, J. Ren, Y. Sun, Appl. Catal. A 243 (2003) 121–133. [26] J. van de Loosdrecht, M. van der Haar, A.M. van der Kraan, A.J. van Dillen, J.W. Geus, Appl. Catal. A 150 (1997) 365–376. [27] J.W. Bae, Y. Lee, J. Park, K. Jun, Energy Fuels 22 (2008) 2885–2891. [28] C. Liu, J. Li, Y. Zhang, S. Chen, J. Zhu, K. Liew, J. Mol. Catal. A 363–364 (2012) 335–342. 25
Page 25 of 40
[29] F. Rohr, O.A. Lindvåg, A. Holmen, E.A. Blekkan, Catal. Today 58 (2000) 247–254. [30] B. Jongsomjit, J. Panpranot, J.G. Goodwin Jr., J. Catal. 215 (2003) 66–77.
[33] A. Bao, K. Liew, J. Li, J. Mol. Catal. A 304 (2009) 47–51.
us
[34] A. Bao, J. Li, Y. Zhang, J. Nat. Gas Chem. 19 (2010) 622–627.
cr
[32] Y. Zhang, H. Xiong, K. Liew, J. Li, J. Mol. Catal. A 237 (2005) 172–181.
ip t
[31] H. Xiong, Y. Zhang, K. Liew, J. Li, J. Mol. Catal. A 231 (2005) 145–151.
an
[35] A.R. de la Osa, A.D. Lucas, J.L. Valverde, A. Romero, I. Monteagudo, P. Coca, P. Sánchez, Catal. Today 167 (2011) 96–106.
M
[36] A.R. de la Osa, A.D. Lucas, A. Romero, J.L. Valverde, P. Sánchez, Fuel 90 (2011) 1935–1945.
d
[37] J.S. Ledford, M. Houalla, A. Proctor, D.M. Hercules, L. Petrakis, J. Phys. Chem. 93 (1989) 6770–6777.
te
[38] S. Vada, B. Chen, J.G. Goodwin, Jr., J. Catal. 153 (1995) 224–231.
Ac ce p
[39] Z. Cai, J. Li, K. Liew, J. Hu, J. Mol. Catal. A 330 (2010) 10–17. [40] S. Zeng, Y. Du, H. Su, Y. Zhang, Catal. Commun. 13 (2011) 6–9. [41] J.W. Bae, S. Kim, Y. Lee, M. Lee, K. Jun, Catal. Commun. 10 (2009) 1358–1362. [42] X. Sun, X. Zhang, Y. Zhang, N. Tsubaki, Appl. Catal. A 377 (2010) 134–139. [43] K.F. Tan, J. Chang, A. Borgna, M. Saeys, J. Catal. 280 (2011) 50–59. [44] S. Park, S. Kim, M.H. Woo, J.W. Bae, K. Jun, K. Ha, Appl. Catal. A 419–420 (2012) 148–155. [45] K. Shimura, T. Miyazawa, T. Hanaoka, S. Hirata, J. Mol. Catal. A 394 (2014) 22–32. [46] R.C. Reuel, C.H. Bartholomew, J. Catal. 85 (1984) 63–77. 26
Page 26 of 40
[47] Y. Liu, T. Hanaoka, T. Miyazawa, K. Murata, K. Okabe, K. Sakanishi, Fuel Process. Technol. 90 (2009) 901–908.
ip t
[48] T. Hanaoka, T. Miyazawa, M. Nurunnabi, S. Hirata, K. Sakanishi, J. Jpn. Inst. Energy 90 (2011)
cr
1072–1080.
[49] T. Hanaoka, Y. Liu, K. Matsunaga, T. Miyazawa, S. Hirata, K. Sakanishi, Fuel Process. Technol. 91
us
(2010) 859–865.
an
[50] N.E. Tsakoumis, M. Rønning, Ø. Borg, E. Rytter, A. Holmen, Catal. Today 154 (2010) 162–182. [51] A.M. Saib, D.J. Moodley, I.M. Ciobîcă, M.M. Hauman, B.H. Sigwebela, C.J. Weststrate, J.W.
M
Niemantsverdriet, J. van de Loosdrecht, Catal. Today 154 (2010) 271–282.
d
[52] T. Miyazawa, T. Hanaoka, K. Shimura, S. Hirata, Catal. Commun. 57 (2014) 36–39.
te
[53] K. Shimura, T. Miyazawa, T. Hanaoka, S. Hirata, Appl. Catal. A 460–461 (2013) 8–14.
Ac ce p
[54] S. Storsæter, Ø. Borg, E.A. Blekkan, A. Holmen, J. Catal. 231 (2005) 405–419. [55] G.L. Bezemer, P.B. Radstake, U. Falke, H. Oosterbeek, H.P.C.E. Kuipers, A.J. van Dillen, K.P. de Jong, J. Catal. 237 (2006) 152–161.
[56] F. Morales, D. Grandjean, A. Mens, F.M.F. de Groot, B.M. Weckhuysen, J. Phys. Chem. B 110 (2006) 8626–8639.
[57] T.E. Feltes, L. Espinosa-Alonso, E. de Smit, L. D’Souza, R.J. Meyer, B.M. Weckhuysen, J.R. Regalbuto, J. Catal. 270 (2010) 95–102. [58] F. Morales, E. de Smit, F.M.F. de Groot, T. Visser, B.M. Weckhuysen, J. Catal. 246 (2007) 91–99. 27
Page 27 of 40
[59] T. Mori, A. Miyamoto, N. Takahashi, M. Fukagaya, T. Hattori, Y. Murakami, J. Phys. Chem. 90 (1986) 5197–5201.
ip t
[60] R. Karaselcuk, A.I. Isli, A.E. Aksoylu, Z.I. Onsan, Appl. Catal. A 192 (2000) 263–271.
cr
[61] X. Mo, J. Gao, N. Umnajkaseam, J.G. Goodwin Jr., J. Catal. 267 (2009) 167–176.
[62] C.M. Balonek, A.H. Lillebø, S. Rane, E. Rytter, L.D. Schmidt, A. Holmen, Catal. Lett. 138 (2010) 8–13.
Ac ce p
te
d
M
an
[64] J. Lee, D. Lee, S. Ihm, J. Catal. 113 (1988) 544–548.
us
[63] A.H. Lillebø, E. Patanou, J. Yang, E.A. Blekkan, A. Holmen, Catal. Today 215 (2013) 60–66.
28
Page 28 of 40
Figure captions Fig. 1. H2-TPR profiles of Co/Al2O3 and Co/M(1)/Al2O3 catalysts.
ip t
Fig. 2. Influence of Co reduction degree on (A) Co surface area and (B) Co dispersion.
cr
Fig. 3. Influence of Co3O4 particle size on (A) Co reduction degree and (B) Co surface area.
Fig. 4. TEM images of (A and E) Co/Al2O3, (B and F) Co/La(2)/Al2O3, (C and G) Co/V(0.5)/Al2O3 and (D
us
and H) Co/La(2)+V(0.5)/Al2O3 catalysts.
an
Fig. 5. Influence of Co reduction degree on (A) CO conversion and (B) TOF. Fig. 6. Influence of Co surface area on (A) CO conversion and (B) TOF.
M
Fig. 7. H2-TPR profiles of Co/Al2O Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 catalysts.
Ac ce p
te
d
Fig. 8. TEM-EDS maps of Co/La(2)+V(0.5)/Al2O3 catalyst.
29
Page 29 of 40
Table 1
Structural analysis of Co/Al2O3 and Co/M(1)/Al2O3 catalysts. PSa / nm
Metal ion (M)
Co3O4
Co metal
Dredb / %
SAc / m2 g-1
Dcorr d / % 1.21
－
19.4
14.6
64
1.18
2
Mg
17.9
13.4
50
0.72
3
Ca
19.8
14.9
87
1.30
4
Sr
18.2
13.7
72
1.13
5
Ba
18.5
13.9
75
1.12
0.98
6
Y
19.0
14.2
64
1.23
1.26
7
La
18.8
14.1
76
1.37
1.17
8
Ce
19.5
14.6
64
1.15
1.17
9
Ti
19.4
14.5
58
0.99
1.12
10
V
15.7
11.7
40
0.21
0.35
11
Mn
17.9
13.4
48
0.50
0.67
12
Zn
19.9
14.9
63
1.00
1.03
13
Zr
18.5
13.9
66
1.18
1.16
14
Mo
17.8
0.25
0.31
ip t
1
us
Entry
0.98
cr
1.02
an M
13.3
54
b
c
Particle size determined by XRD. Reduction degree of Co. Surface area of Co metal.
d
Corrected dispersion of Co metal.
Ac ce p
te
d
a
Table 2
0.93
Structural analysis of Co/Al2O3, Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 catalysts. Dredb
SAc
Dcorr d
Co metal
/%
/ m2 g-1
/%
19.4
14.6
64
1.18
1.21
0
18.8
14.1
76
1.37
1.17
2
0
19.1
14.3
82
1.33
1.06
0
0.5
18.7
14.0
40
0.29
0.48
5
0
1
15.7
11.7
40
0.21
0.35
6
1
0.5
18.6
13.9
57
0.87
0.98
7
1
1
16.1
12.1
42
0.34
0.52
8
2
0.5
18.3
13.7
69
1.04
0.99
9
2
1
17.2
12.9
48
0.51
0.69
Entry 1 2 3 4
PSa / nm
La
V
/ wt%
/ wt%
Co3O4
0
0
1
a
Particle size determined by XRD. b Reduction degree of Co. c Surface area of Co metal.
d
Corrected dispersion of Co metal. 30
Page 30 of 40
Table 3 Catalytic performance of Co/Al2O3 and Co/M(1)/Al2O3 in FT synthesis. Metal ion (M)
CO conv. a
STY(C5+) b
Product selectivity / %
-1
TOF
-1
/%
CO2
CH4
C2–C4
C5+
/ mg g h
/ s-1
－
48
4
8
6
82
179
0.15
2
Mg
33
4
7
5
84
126
0.17
3
Ca
40
2
7
10
81
144
0.11
4
Sr
44
3
8
7
82
163
0.14
5
Ba
46
4
9
8
79
162
0.15
6
Y
46
4
8
9
79
162
0.14
7
La
50
4
6
7
83
188
0.14
8
Ce
55
4
7
7
82
200
0.18
9
Ti
38
3
7
5
85
147
0.14
10
V
37
3
7
4
86
145
0.65
11
Mn
48
4
9
6
81
173
0.36
12
Zn
43
4
10
8
78
153
0.16
13
Zr
49
2
8
7
83
184
0.15
14
Mo
35
5
8
6
81
127
0.51
us
an
M
b
CO conversion. Space time yield of C5+ products.
te
d
a
ip t
1
cr
Entry
Table 4 Entry
La
V
CO conv. a
STY(C5+) b
Product selectivity / %
/ wt%
/%
CO2
CH4
C2–C4
C5+
/ mg g h
/ s-1
0
0
48
4
8
6
82
179
0.15
1
0
50
4
6
7
83
188
0.14
2
0
54
4
8
9
79
192
0.15
0
0.5
53
3
10
7
80
192
0.68
0
1
37
3
7
4
86
145
0.65
6
1
0.5
67
4
8
7
81
244
0.29
7
1
1
55
3
9
5
83
208
0.61
8
2
0.5
77
3
8
8
81
279
0.27
9
2
1
59
2
7
6
85
225
0.43
2 3 4 5
-1
-1
TOF
/ wt%
1
a
Ac ce p
Catalytic performance of Co/Al2O3 Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 in FT synthesis.
b
CO conversion. Space time yield of C5+ products.
31
Page 31 of 40
(B)
(A)
Co/Al 2O3
Co/Al 2O3
Co/Ce/Al 2O3
ip t
Co/Y/Al 2O3
H2 consumption / a.u.
Co/Mo/Al 2O3 Co/Mg/Al 2O3
Co/Zr/Al 2O3
cr
Co/Ti/Al 2O3
Co/Sr/Al 2O3
Co/Ba/Al 2O3
us
H2 consumption / a.u.
Co/Zn/Al2O3
Co/Mn/Al 2O3 Co/V/Al 2O3
Temperature / ˚C
Co/Ca/Al 2O3
100 200 300 400 500 600 700 800 900
M
100 200 300 400 500 600 700 800 900
an
Co/La/Al 2O3
Temperature / ˚C
Ac ce p
te
d
Fig. 1. H2-TPR profiles of Co/Al2O3 and Co/M(1)/Al2O3 catalysts.
32
Page 32 of 40
(A) La
0.5
Co/M/Al2O3
0
V
40
50
60
70
80
cr
Co/La+V/Al2O3
ip t
1
90
us
Co surface area / m2 g-1
1.5
Co reduction degree / % (B)
an
1.5
M
1
0.5
Co/La+V/Al2O3
d
Co dispersion / %
La
Co/M/Al2O3
V
te
0
40
50
60
70
80
90
Ac ce p
Co reduction degree / %
Fig. 2. Influence of Co reduction degree on (A) Co surface area and (B) Co dispersion.
33
Page 33 of 40
90 Co/La+V/Al2O3 Co/M/Al2O3
80
La
ip t
70 60 50 V
40 30 16
17
18
19
20
21
us
15
cr
Co reduction degree / %
(A)
Co3O4 particle size / nm
Co/La+V/Al2O3 Co/M/Al2O3
M
1
La
0.5 V
d
Co surface area / m2 g-1
1.5
an
(B)
te
0
Ac ce p
15
16
17
18
19
20
21
Co3O4 particle size / nm
Fig. 3. Influence of Co3O4 particle size on (A) Co reduction degree and (B) Co surface area.
34
Page 34 of 40
(B)
us
cr
ip t
(A)
100 nm
an
100 nm
(C)
Ac ce p
te
d
M
(D)
100 nm
100 nm
35
Page 35 of 40
(F)
us
cr
ip t
(E)
20 nm
an
20 nm
(H)
Ac ce p
te
d
M
(G)
20 nm
20 nm
Fig. 4. TEM images of (A and E) Co/Al2O3, (B and F) Co/La(2)/Al2O3, (C and G) Co/V(0.5)/Al2O3 and (D and H) Co/La(2)+V(0.5)/Al2O3 catalysts.
36
Page 36 of 40
80 70 60 50 V 0.5% Mn
40 30 V 1%
Ca
Mo
20 50
60
70
80
90
us
40
ip t
Co/La+V/Al2O3 Co/M/Al2O3
cr
CO conversion / %
(A)
0.8 V 0.5% V 1%
TOF / s-1
0.6
Co/La+V/Al2O3
Mo
0.4
Mn
Co/M/Al2O3
M
(B)
an
Co reduction degree / %
te
0
d
0.2
40
50
60
70
80
90
Ac ce p
Co reduction degree / %
Fig. 5. Influence of Co reduction degree on (A) CO conversion and (B) TOF.
37
Page 37 of 40
(A)
Co/La+V/Al2O3 Co/M/Al2O3
70
50
Mn
V 0.5% V 1%
40
Ca
30
Mo
20 0.5
1
1.5
us
0
ip t
60
cr
CO conversion / %
80
2
-1
(B)
0.8 V 0.5%
Mo
0.4
Co/La+V/Al2O3 Co/M/Al2O3
M
TOF / s-1
0.6 V 1%
an
Co surface area / m g
Mn
0
d
0.2
0.5
1
1.5
Co surface area / m2 g-1
Ac ce p
te
0
Fig. 6. Influence of Co surface area on (A) CO conversion and (B) TOF.
38
Page 38 of 40
Co/Al2O3 Co/La(1)/Al2O3
ip t
H2 consumption / a.u.
Co/La(2)/Al2O3
Co/V(1)/Al2O3
us
Co/La(1)+V(0.5)/Al 2O3
cr
Co/V(0.5)/Al2O3
Co/La(1)+V(1)/Al 2O3
an
Co/La(2)+V(0.5)/Al 2O3
M
Co/La(2)+V(1)/Al 2O3
100 200 300 400 500 600 700 800 900
d
Temperature / ˚C
Ac ce p
te
Fig. 7. H2-TPR profiles of Co/Al2O3, Co/La/Al2O3, Co/V/Al2O3 and Co/La+V/Al2O3 catalysts.
39
Page 39 of 40
Co
La
V
Ac ce p
te
O
d
M
an
Al
us
cr
ip t
50 nm
Fig. 8. TEM-EDS maps of Co/La(2)+V(0.5)/Al2O3 catalyst
40
Page 40 of 40