Controlled Preparation of Co₃O₄@porous-SiO₂ Nanocomposites for Fischer–Tropsch Synthesis

Abstract

Monodispersible Co₃O₄ nanoparticles were prepared via a facile solvothermal route using polyvinylpyrrolidone (PVP) as capping agent and the porous silica shell was then coated by means of the Stöber process to fabricate Co₃O₄@porous-SiO₂ (Co₃O₄@p-SiO₂) nanocomposites. The particle size of Co₃O₄ and porous silica shell thickness of Co₃O₄@p-SiO₂ nanocomposites could be easily controlled through changing the amount of PVP and tetraethoxysilane, respectively. The high resolution transmission electron microscopy results, together with the X-ray diffraction results, indicated that monodispersible Co₃O₄ nanoparticles were successfully prepared and uniformly encapsulated by porous silica. During the growth of silica shell, the PVP was trapped and dispersed in the silica shell. As a result, the porous silica shell was obtained after burning off the PVP and a positive correlation existed between the Brunauer–Emmett–Teller surface area of the porous silica shell and quantity of PVP in the original Co₃O₄ nanoparticles. Compared with the Co/SiO₂ reference catalyst, CO conversion of the Co@p-SiO₂ model catalyst was more stable and higher in a period of 240 h, and hydrocarbon selectivity towards C₅–C₁₈ fraction was also higher than that of the Co/SiO₂ catalyst. The results of analysis for the Co@p-SiO₂ catalyst showed that core@shell structure could maintain high dispersion of Co particles so as to provide higher number of Co active sites, and enhance selectivity towards C₅–C₁₈ fraction due to confined structure of porous channel in the silica shell.

Graphical Abstract

Schematic diagram of the procedure for synthesizing Co₃O₄@p-SiO₂ nanocomposites.

1 Introduction

Fischer–Tropsch (F–T) synthesis has attracted much research attention nowadays because of the severe oil crisis and the tight fuel specifications [1, 2]. One of the goals of modern F–T synthesis research is to design a catalyst capable of highest selectivity to the desired product without sacrificing activity and lifetime [3]. Cobalt F–T synthesis catalysts are preferred due to their low activity for the water–gas shift reaction, slow deactivation rates and highly selective to linear paraffin than iron catalysts [4]. Since cobalt is ~1,000 times more expensive than iron, therefore optimal design of a cobalt catalyst is essential for its utilization. Decreasing the Co particle size to increase the exposed surface areas per unit mass of the Co [5] and hindering Co particles from agglomerating during reaction would be reasonable pathway to achieve this goal. Although up to now, many methods were designed to prepare cobalt-based catalysts with different Co particle size, and morphologies as well. It had been difficult to translate these advances into successful application in F–T synthesis as to be the drawback of easily agglomeration and sintering of the Co particles under harsh catalytic reaction conditions. In order to prevent Co particles from agglomerating and decrease the deactivation rates of the cobalt catalysts, a novel cobalt catalyst with core@shell structure was preferable to be adopted, where a porous shell played dual roles as: permitting syngas molecules to pass through the channel freely and preventing Co cores from coming into direct contact each other and thus hindering Co cores from agglomerating. This confined structure enabled cobalt catalyst showing special catalytic performances and product distribution in F–T synthesis. For example, Tsubaki co workers [6–9] had prepared core@shell cobalt catalysts by coating H-ZSM-5 zeolite membranes on the Co/SiO₂ catalysts, which showed an excellent selectivity to the C₅–C₁₁ isoparaffins (gasoline distillates) benefiting from the confined reaction conditions constructed by H-ZSM-5 membranes. Moreover, the article of Trépanier et al. [10] has also showed that carbon nanotubes, when used as a cobalt catalyst support, allowed a better Co dispersion and minimized the Co metal phase interaction with the support. The Co particles located inside the tubes were more active than the ones on the outer surface of the CNT owing to confined effect. Thus, how to synthesize cobalt catalysts with confined structure via a facile method became a big challenge.

The present work developed a procedure that permitted the preparation of the Co₃O₄@porous-SiO₂ (Co₃O₄@p-SiO₂) nanocomposites with different particle sizes of Co₃O₄ core and thickness of porous silica shell via solvothermal route using polyvinylpyrrolidone (PVP) as capping agent. Compared with the conventional Co/SiO₂ catalyst prepared via precipitation method, the Co@p-SiO₂ catalyst, together with high dispersion of Co particles due to confined structure, could exhibit higher and more stable CO conversion and higher selectivity towards C₅–C₁₈ fraction in F–T synthesis.

2 Experimental Section

2.1 Nanocomposites Preparation

2.1.1 Preparation of the Co₃O₄@p-SiO₂ Nanocomposites

The synthesis procedure for the Co₃O₄@p-SiO₂ nanocomposites was illustrated in Scheme 1. In a typical synthesis, 7.09 g Co(NO₃)₂·6H₂O and 3.54 g PVP were dissolved in 800 mL ethanol solution under magnetic stirring at room temperature. The concentrations of the solution were calculated in order to obtain 20 % cobalt loadings in the final catalysts. After the solution turned clear, it was transferred into a Teflon-lined stainless autoclave with the total volume of 1,000 mL, which afterwards was loaded in an oven at 453 K for 4 h and then naturally cooled to room temperature. The suspension was again transferred to the beaker. 2,076 mL ethanol was added into the suspension. After that, 1,656 mL distilled water and 144 mL 25 % ammonia were subsequently added into the suspension. After magnetic stirring for 2 h, 20 mL tetraethoxysilane (TEOS) was added dropwise into the suspension during 12 h. Magnetic stirring was continued for 48 h. The solid products were collected by filtration, washed thoroughly with distilled water and ethanol, and dried in an oven at 333 K. The template was removed by calcination at 773 K for 6 h in flowing air. The prepared nanocomposites were denoted as Co₃O₄@p-SiO₂, which were selected as model catalysts for the subsequent F–T synthesis. Meanwhile, by controlling Co/PVP mass ratio, Co₃O₄@p-SiO₂ nanocomposites with different core sizes and thickness of silica shell were also prepared using same method.

Scheme 1

Schematic diagram of the procedure for synthesizing Co₃O₄@p-SiO₂ nanocomposites

2.1.2 Preparation of the Co₃O₄/SiO₂ Nanocomposites

The Co₃O₄/SiO₂ nanocomposites (cobalt content 20 %) were prepared as follows: Firstly, 60 mL distilled water and 60 mL NH₃·H₂O were dissolved in 648 mL ethanol solution in sequence with magnetic stirring at room temperature. After that, 31.8 mL TEOS was added dropwise into above solution. Magnetic stirring was continued for 2 h. The spherical SiO₂ support were collected by filtration, washed thoroughly with distilled water and ethanol, and dried in an oven at 333 K. Secondly, the Co₃O₄ nanoparticles were precipitated on the as-made SiO₂ support by adding a highly concentrated NH₃·H₂O solution to a dilute cobalt nitrate solution, After aging for 2 h, the precipitate was separated from the suspension by filtration, washed thoroughly with distilled water and ethanol. Finally, the products were dried at 333 K and calcined at 773 K for 6 h. The as-made nanocomposites were denoted as Co₃O₄/SiO₂, which were used as reference catalysts.

2.2 Nanocomposites Characterization

High resolution transmission electron microscopy (HRTEM) images were obtained on a Philips CM200FEG operating at 200 kV. The powder X-ray diffraction (XRD) patterns were recorded on a Bruker diffractometer using a CuKα radiation. The texture properties of nanocomposites were determined by N₂ adsorption/desorption isotherms at 77 K using a Micromeritics Tristar 3000 sorptometer. The pore size distribution and pore volume were determined by the BJH method. X-ray photoelectron spectroscopy spectra (XPS) were performed on a PerkinElmer PHI-5300 spectrometer. Hydrogen temperature programmed reduction (H₂-TPR) was carried out in a quartz microreactor heated by an electrical furnace. The reactor was loaded with 25 mg of calcined sample and heated at a rate of 10 K/min to 1,233 K with a gas consisting of 5 % H₂ in N₂. The gas flow rate was 60 mL/min. The mount of H₂ consumed by the sample was measured by analysis of the effluent gas with a thermal conductivity detector (TCD).

2.3 Catalysts Test

F–T synthesis was carried out at 2 Mpa, 1,000 h⁻¹ (the gas flow rate: 33.3 mL/min) and a H₂/CO ratio of 2 in a stainless-steel fixed-bed reactor (i.d. = 10 mm) with a catalyst loading of 2 mL. The catalyst was reduced in a flow of hydrogen at 673 K for 10 h and then naturally cooled to room temperature before switching to syngas. Data were taken at steady state after 24 h on-stream. Wax was collected with a hot trap and the liquid products were collected in a cold trap. The gas effluents were analyzed on-line by using Carbosieve-packed column with TCD. The gas hydrocarbons were analyzed on-line using Porapack-Q column with FID. Oil and wax were analyzed offline in OV-101 capillary columns. 5 % N₂ was added to syngas as an internal standard. The carbon balance and mass balance were 100 ± 5 %.

3 Results and Discussion

3.1 Synthesis and Characterization of Co₃O₄@p-SiO₂ Nanocomposites

PVP is an amphiphilic polymer and usually employed as stabilizer for the preparation of various metal nanoparticles [11–15]. As reported previously, the Co₃O₄ nanoparticles with different particle size could be prepared via solvothermal route using PVP as stabilizer [16]. The PVP was chosen as stabilizer for the following reasons. The hydroxyl groups located on the surface of prepared Co₃O₄ nanoparticles, the hydrogen bond could form between PVP and Co₃O₄ nanoparticles [18]. By adsorbing PVP onto the surface of Co₃O₄ nanoparticles, this PVP layer could help to stabilize Co₃O₄ nanoparticles and avoid aggregation. It has been known that the coating of the host materials by PVP was very effective for the subsequent silica shell formation [19]. Figure 1a showed the XRD pattern of the Co₃O₄ nanoparticles prepared via solvothermal methods. It was found that all the diffraction peaks matched Co₃O₄ well with standard JCPDS data (42–1,467). After coating with porous silica shell, the Co₃O₄@p-SiO₂ nanocomposites, as shown in Fig. 1b, exhibited similar diffraction peaks to that of the Co₃O₄ nanoparticles, suggesting that the Co₃O₄ phase was retained. However, it was found that the diffraction peaks of Co₃O₄@p-SiO₂ nanocomposites had relatively low intensity, because of the shielding effect of the porous silica shell to the detecting X-rays [17]. Moreover, the Co₃O₄/SiO₂ reference nanocomposites, as shown in Fig. 1c, also exhibited similar diffraction peaks to that of the Co₃O₄ nanoparticles, indicating that cobalt phase was also present in the form of Co₃O₄ spinel.

Fig. 1

XRD patterns of a Co₃O₄ nanoparticles; b Co₃O₄@p-SiO₂ nanocomposites; c Co₃O₄/SiO₂ nanocomposites

Seen from Fig. 2a–c, it could be clearly observed that the PVP-stabilized Co₃O₄ nanoparticles were monodispersible and uniform. By controlling the amount of PVP, the Co₃O₄ nanoparticles with different size could be prepared. Moreover, these PVP-stabilized Co₃O₄ nanoparticles favored uniform growth of silica shell owing to interaction between PVP and Si–OH groups. To fabricate Co₃O₄@p-SiO₂ nanocomposites, the stöber process would be applied to coat the silica onto the PVP-stabilized Co₃O₄ nanoparticles, which acted as seeds or nuclei for silica shell formation. Furthermore, in order to ensure uniform coating and to avoid formation of individual silica particles, the reaction conditions had to be properly controlled. For example, a lower TEOS concentration was indispensable of preventing aggregation of the particles during the growth of the initial silica shell [20]. Furthermore, by varying the amount of PVP and TEOS, Co₃O₄@p-SiO₂ nanocomposites with different particle size of Co₃O₄ and silica shell thickness could also be prepared. Seen from Fig. 2d–i, monodispersible and uniform silica coating of Co₃O₄@p-SiO₂ nanocomposites could be clearly observed in HRTEM photographs. It was remarkable that nearly perfect core@shell morphology was obtained, with rather uniform size distributions. Such results demonstrated the high efficiency of this procedure, despite its simplicity, as compared to other silica-coating techniques previously reported [21–25].

Fig. 2

HRTEM photographs of a–c different particle sizes of Co₃O₄ nanoparticles [uniform 7.09 Co(NO₃)₂·6H₂O was provided under different PVP mass from d (PVP = 14.17 g), e (PVP = 7.09 g) to f (PVP = 3.54 g)]; d–f Co₃O₄@p-SiO₂ nanocomposites with different particle sizes of Co₃O₄ nanoparticles [uniform 20 mL TEOS was provided under different PVP mass from d (PVP = 14.17 g), e (PVP = 7.09 g) to f (PVP = 1.77 g)]; g–i Co₃O₄@p-SiO₂ nanocomposites with different thicknesses of porous silica shell [uniform 3.54 g PVP was provided under different TEOS from d (TEOS = 20 mL), e (TEOS = 40 mL) to f (TEOS = 60 mL)]

Since the major contribution of surface area and porous structures results from PVP, it is therefore necessary to correlate the quantity of PVP in the reacting system with final specific surface area of Co₃O₄@p-SiO₂ nanocomposites. Figure 3 showed the Brunauer–Emmett–Teller (BET) surface area of Co₃O₄@p-SiO₂ nanocomposites after calcinations at various PVP/Co mass ratios. It could be clearly observed that the BET surface area of the Co₃O₄@p-SiO₂ nanocomposites increased with increasing PVP/Co mass ratio from 0.25 to 2.0. This result indicated that a positive correlation existed between the BET surface area of the porous silica shell and quantity of PVP.

Fig. 3

The BET surface area of Co₃O₄@p-SiO₂ nanocomposites after calcinations at various PVP/Co mass ratios

In addition, Fig. 4 also showed HRTEM images of the Co₃O₄@p-SiO₂ model catalysts and Co₃O₄/SiO₂ reference catalysts before and after reaction, respectively. Seen from Fig. 4a, the fresh Co₃O₄@p-SiO₂ catalysts showed nearly perfect core@shell morphology, Co₃O₄ nanoparticles were completely coated by porous silica. After F–T reaction, the HRTEM image of the spent Co@p-SiO₂ model catalysts (see Fig. 4b; Table 1) illustrated that the core@shell structure was perfectly preserved due to higher hydrothermal stability of porous silica shell. The coated Co particles did not agglomerate owing to protection of porous silica shell during the reaction. However, the HRTEM image of the fresh Co₃O₄/SiO₂ catalysts prepared via a controlled coprecipitation of cobalt nitrate with ammonia showed the fluffy Co₃O₄ nanoparticles were synthesized onto outer surface of spherical SiO₂ support (see Fig. 4c; Table 1). For the spent Co/SiO₂ catalysts, it was clear that serious aggregation of the Co particles were observed on the external surface of spherical SiO₂ support (see Fig. 4d), indicating that small Co particles located on outer surface of the nonporous spherical SiO₂ support were easier to agglomerate into large Co clusters without protection.

Fig. 4

HRTEM photographs of a fresh Co₃O₄@p-SiO₂ nanocomposites, b spent Co@p-SiO₂ model catalysts, c fresh Co₃O₄/SiO₂ nanocomposites, d spent Co/SiO₂ reference catalysts

Table 1 Textural properties of fresh and spent catalysts

In Fig. 5a, it was not found Co_2p or Co_(LMM) peaks in XPS spectra of Co₃O₄@p-SiO₂ nanocomposites, whereas XPS spectra was sensitive only to the Co ions localized at the surface and subsurface layer of the nanocomposites (depth of analysis ~6 nm) [26]. This result indicated that the Co₃O₄ nanoparticles did not located on the surface of porous silica shell and were completely coated by porous SiO₂. Whereas, in the case of Co₃O₄/SiO₂ nanocomposites, Co_2p or Co_(LMM) peaks were presented in XPS spectra in Fig. 5b, suggesting that the Co₃O₄ nanoparticles located on the external surface of nonporous silica spheres.

Fig. 5

XPS spectra of the a Co₃O₄@p-SiO₂ nanocomposites and b Co₃O₄/SiO₂ nanocomposites

3.2 Reducibility of the Co₃O₄@p-SiO₂ Nanocomposites

Hydrogen temperature programmed reduction is a very convenient technique for studying the reduction behaviors of catalysts qualitatively. In some case it is also possible to obtain useful information about the degree of interaction of the supported phase with the support from the reduction profiles of supported oxides. Figure 6 showed the H₂-TPR profiles for the Co₃O₄@p-SiO₂ and Co₃O₄/SiO₂ nanocomposites. For the Co₃O₄@p-SiO₂ nanocomposites (see Fig. 6a), the first peak was apparent and this was attributed to the reduction of Co₃O₄ to CoO. However, the first peak was significantly shifted to higher temperature than that of the Co₃O₄/SiO₂ nanocomposites. This result could be attributed to the protective effects of porous silica shell which inhibited H₂ to diffuse along the porous channels to access Co₃O₄ nanoparticles [27, 28]. Moreover, the second peak was also gradually shifted to higher temperatures and this was attributed to the reduction of CoO to metallic Co⁰. This result indicated that the reduction of CoO to metallic Co⁰ was difficult owing to H₂ diffusion limitation in porous channels. A reduction pattern in the temperature range of about 820–1,015 K was observed for the Co₃O₄@p-SiO₂ nanocomposites. Such a high reduction temperature might be assigned to the reduction of cobalt silicate species formed during the H₂-TPR experiment by reaction of CoO with Si–OH groups [26]. However, for the Co₃O₄/SiO₂ nanocomposites (see Fig. 6b), two peaks were apparent and these were attributed to the reduction of Co₃O₄ to CoO, which then reduced at lower temperature to metallic Co⁰. Significantly increasing area of the second peak indicated that the CoO particles were much easier to be reduced to active Co species. Thus first and second peaks were shifted to lower temperatures due to the lower interaction between CoO and Si–OH groups and higher contact area with H₂. As to the reduction of cobalt silicate species, a small reduction peak in the temperature range of about 595–695 K was also observed for the Co₃O₄/SiO₂ nanocomposites. Thus, the reducibility of the Co₃O₄/SiO₂ nanocomposites was higher than that of the Co₃O₄@p-SiO₂ nanocomposites, as shown in Table 1.

Fig. 6

H₂-TPR profiles of the Co₃O₄@p-SiO₂ and Co₃O₄/SiO₂ catalysts

3.3 FTS Performances

The performances of the Co@p-SiO₂ and Co/SiO₂ catalysts for F–T synthesis were evaluated in a fixed-bed reactor. Figure 7 depicted the variations of CO conversion with the time on-stream (TOS) under the same reaction temperature (503 K). CO conversion of the Co@p-SiO₂ catalyst was more stable and higher than that of Co/SiO₂ catalyst. In the case of the Co@p-SiO₂ catalyst, CO conversion decreased slightly from 71.6 to 56.1 % in a period of 240 h, whereas, the Co/SiO₂ catalyst showed a quick drop in CO conversion from 61.7 to 30.5 %. It is well known that dispersion and reducibility of the cobalt catalyst play an important role in catalytic activity. Although the Co@p-SiO₂ catalyst had the lower reduction degree than the Co/SiO₂ catalyst, it maintained a high dispersion of Co particles and number of active sites due to the protection of porous silica shell (see Fig. 4b; Table 1), which inhibited the agglomeration of Co particles during the reaction process. Therefore, more Co active sites were available for the CO hydrogenation. However, Co particles of the Co/SiO₂ catalyst agglomerated easily among them during the reaction without the protection of porous silica shell (see Fig. 4d), thus decreasing the number of active sites and catalytic activity. To conclude, it is known that the CO conversion not only related to the reducibility of catalysts, but also to the dispersion of Co active sites. A high and stable CO conversion of the Co@p-SiO₂ catalyst could therefore be attributed to a combination of high Co dispersion and appropriate reducibility in comparison to that of the Co/SiO₂ catalyst.

Fig. 7

Change in CO conversion with TOS for the Co@p-SiO₂ and Co/SiO₂ catalysts (reaction conditions: H₂/CO = 2, GHSV = 1,000 h⁻¹, P = 2 MPa, T = 503 K)

The hydrocarbon selectivity of both Co@p-SiO₂ and Co/SiO₂ catalysts were summarized in Table 2 under the different reaction temperatures (503 and 513 K). The data showed that selectivity towards both CH₄ and C₅–C₁₈ fraction of the Co@p-SiO₂ catalyst was higher than that of the Co/SiO₂ catalysts. Whereas, the Co/SiO₂ catalyst was more selective towards C₁₉₊ fraction while its selectivity towards both CH₄ and C₅–C₁₈ fraction were lower. Reuel and Batholomew [29] reported that high CH₄ selectivity observed on the Co/SiO₂ catalyst with narrow pore was likely to be attributed to the presence of either unreduced cobalt species or the small cobalt particles which produced relatively high CH₄ fraction at the expenses of the C₁₉₊ fraction than large cobalt particles. In our case, a part of unreduced CoO in the Co@p-SiO₂ catalyst, which was discussed in H₂-TPR session, resulted in high CH₄ selectivity. Moreover, it is known that large Co particles lead to more selective for long-chain hydrocarbons under the same cobalt loadings [30, 31]. Khodakov et al. [27] had shown that large Co particles were more active in F–T synthesis and produced higher selectivity towards long-chain hydrocarbons than small Co particles. In addition, confined effect of the porous channel could inhibit the growth of long-chain hydrocarbons, thus increasing the selectivity towards C₅–C₁₈ fraction in the Co@p-SiO₂ catalyst. The results of product selectivity analysis for Co@p-SiO₂ catalysts showed that high CH₄ selectivity and low selectivity towards C₁₉₊ fraction of the Co@p-SiO₂ catalyst resulted from partly unreduced CoO species and the small Co particles. It was also reasonable to suggest that the confined structure of porous channel in the Co@p-SiO₂ catalyst could enhance product selectivity towards C₅–C₁₈ fraction. Whereas, high selectivity towards both CH₄ and C₁₉₊ fraction of the Co/SiO₂ catalyst could be ascribed to high reducibility and formation of the large Co clusters (see Fig. 4d), respectively. This result in turn led to decreasing of C₅–C₁₈ selectivity.

Table 2 Catalytic performances of catalysts for F–T synthesis

4 Conclusions

Our work showed that Co₃O₄ nanoparticles could be synthesized via a facile solvothermal route using PVP as capping agent. Novel core@shell Co₃O₄@p-SiO₂ nanocomposites with different Co₃O₄ particle sizes and silica shell thickness had been subsequently prepared via the Stöber process. CO conversion of the Co@p-SiO₂ model catalyst was more stable and higher than that of the Co/SiO₂ reference catalyst due to confined structure of porous silica shell, which could maintain high Co dispersion and number of active sites during harsh reaction condition. Selectivity towards both CH₄ and C₅–C₁₈ fraction of the Co@p-SiO₂ catalyst were higher than that of the Co/SiO₂ catalyst due to more unreduced CoO and confined effect of porous channels, respectively. Compared with the Co@p-SiO₂ catalyst, the large Co clusters formed during reaction owing to agglomeration among small Co particles was in favor of enhancing selectivity towards C₁₉₊ fraction for the Co/SiO₂ catalyst.