Alumina membrane coated activated carbon: a novel strategy to enhance the mechanical properties of a solid catalyst

Abstract

This paper presents a novel strategy to enhance the mechanical properties of a solid catalyst. By employing a slip-casting method, an alumina membrane can be prepared on the surface of activated carbon particles uniformly. Significantly, the mechanical properties test indicates that the attrition index of activated carbon decreases from 12.6% to 2.9% due to the alumina membrane coating.

---

Introduction

Stability is one of the most important properties of solid catalysts besides activity or selectivity.^1,2 For example, the mechanical failure of the catalyst, rather than the loss of its activity, is often the cause for process shutdowns and catalyst replacement.³ For many reactors, especially fluidized or moving bed reactors, the catalyst is inevitably subjected to mechanical stress due to interparticle collisions and bed-to-wall impacts. This mechanical stress leads to a gradual degradation of the individual catalyst particles, which is often termed attrition.^4,5 As a result, the generated fines will bring downstream contamination, and even affect the density distribution in the reactor.

Currently, the mechanical properties of a solid catalyst are improved through optimization of granulation/pelletization process.³ Pelletization will influence not only the mechanical strength of the catalyst, but also the catalyst activity, surface area and pores distribution.⁶ For example, for pelletization by molding, the porosity of catalyst decreases with increasing the molding pressure. On the other hand, the high molding pressure may reduce the catalyst grain surface defects (such as space, dislocations and grain boundaries) which will make the catalyst activity decreased. Subsequent heat treatment after pelletization is a very complex process,⁷ in which oxide of low value is oxidized into high value, and combined water and other sinters are removed, and the new oxide crystal is formed. At the same time, owing to primary particles bonding, integration and cross-linking, some form of secondary structure is formed at high temperature. Therefore the improvement in mechanical strength of a catalyst is achieved at the cost of loss of catalyst activity.

Sol–gel derived porous alumina granular particles have shown excellent mechanical properties including crushing strength and attrition resistance.^8,9 In this work, we studied a novel strategy to enhance mechanical properties of solid catalyst by coating a thin sol–gel derived alumina membrane on the surface of the porous activated carbon particles. Activated carbon is a typical catalyst or support, which is widely used in the chemical process such as acetylene hydrochlorination,¹⁰ NO_x reduction¹¹ and Fisher–Tropsch synthesis.¹² With high mechanical strength of the sol–gel derived nanostructured alumina membrane, we expected that the coated alumina membrane would enhance the mechanical properties significantly but has no effect on the micropore structure and density of activated carbon. Since the coated alumina membrane has pore size much larger than the pores of the activated carbons, we expected that the coating would have no effect on gas diffusion and chemical reaction characteristics of the activated carbon catalyst. This paper reports synthesis and properties of the alumina membranes coated activated carbon catalysts.

Experimental

Activated carbon (AC) particles were prepared from coconut-shell (particle size: 450–900 μm). Alumina membrane coated activated carbon (AC@alumina) samples were prepared by the sol–gel slip-casting method. Boehmite sol of 0.56 M in aluminum concentration was synthesized by the Yoldas process.^13,14 In brief, the precursor aluminum tri-sec-butoxide (ALTSB) was first hydrolyzed in water at 85 °C and stirred for 1 h. The resulting slurry with AlOOH precipitates was peptized with HNO₃, at an HNO₃ to AlOOH molar ratio of 0.07. The sol was refluxed overnight for more than 12 h at 100 °C, followed by exposure to air at 90 °C for 2 h to evaporate the remaining alcohol. Then, AC particles were immersed in the prepared boehmite sol at room temperature for 10 min. The resulting samples were dried at 40 °C with high humidity for 24 h and calcined at 450 °C for 4 h with a heating and cooling rate of (40 °C h⁻¹). The thickness of the alumina membranes on the AC particles can be controlled by repeated immersion/drying/calcination described above. For comparison, pure alumina was prepared by drying the boehmite sol followed by calcination at 500 °C for 4 h.

The morphology of the materials synthesized was studied by scanning electron microscopy (SEM, TM-1000 and ultra-55). X-ray diffraction (XRD) was conducted using an X' Pert MPD X-ray diffractometer equipped with Cu Kα radiation at 40 kV and 30 mA. The microporous and mesoporous structure of the particles was analyzed using nitrogen adsorption porosimetry (Micromeritics, ASAP 2020). The surface area was calculated by applying the BET adsorption model. The HK method was used to determine the average pore size of samples. The pore volume was determined as the liquid equivalent of the volume adsorbed at the relative pressure close to 1 on the adsorption isotherm. The micropore volume was calculated by the t-plot method, and the external surface area was obtained by subtracting the external surface area from the BET surface area. The macropore structure of the particles was studied by mercury porosimeter (Micromeritics, AutoPore IV 9510).

The test for attrition resistance of the sample was carried out according to the standard rotating drum method.¹⁵ Tumbling the particles in a rotating tube and measuring the proportion of fines (less than 400 μm) generated is used as a measure of attrition resistance. The weight loss of particles calculated by the following equation was used as attrition index:

Results and discussion

Fig. 1(a)–(c) show SEM images of activated carbon before and after coating with alumina membrane for one time. Comparing Fig. 1(a) with Fig. 1(b), it is evident that a uniform layer of membrane is formed on the surface of activated carbon particle. The thickness of membrane is ca. 550 nm (Fig. 1(c)). XRD patterns of the prepared samples are shown in Fig. 1(d). The characteristic peaks located at 45.6 and 67.0° can be found and corresponds to γ-Al₂O₃ although some of them overlap partly by those of AC.¹⁶ The results indicate that a thin layer of γ-Al₂O₃ membrane has been synthesized successfully on activated carbon by slip-casting method. In addition, the thickness of alumina membrane can be controlled by repeated coating. For example, the thickness is ca. 753 nm after two-time coating and 999 nm for three-time coating (ESI, Fig. S1†).

Fig. 1 SEM images of (a) a representative view of activated carbon, (b) a representative view of activated carbon coated by alumina membrane, (c) cross-sectional view of activated carbon coated by alumina membrane, and (d) XRD patterns of the samples.

The nitrogen adsorption–desorption isotherms of activate carbon before and after coating of alumina, and pure alumina, are shown in Fig. 2. It can be seen that the N₂ isotherm for AC is typical type I, characteristic of microporous materials. The isotherms for alumina clearly show a hysteresis loop, characteristic of mesoporous materials.¹⁷ The isotherms for the AC@alumina sample resemble that for the AC without coating. This is because of the small quantity of the alumina membrane which contributes negligibly to the nitrogen adsorption on AC@alumina sample. The results also imply that the coating of alumina membrane does not affect adsorption of nitrogen on activated carbon. With the increase of alumina thickness, the hysteresis loop becomes evident indicating the increase of alumina content in the sample (ESI, Fig. S2†).

Fig. 2 N₂ adsorption isotherms of the samples.

The BET surface area, pore volume, and average pore size for the samples are summarized in Table 1. It can be seen that activated carbon has a BET surface area of 810 m² g⁻¹, much higher than that of alumina. The pore volume and micropore area of activated carbon are 0.481 cm³ g⁻¹ and 573.1 m² g⁻¹, respectively. It is clear that these values that quantify the microstructure of the AC@alumina lie between those of the pure activated carbon and pure Al₂O₃ because only activated carbon contributes to the micropore of the sample. In addition, the average pore size of AC@alumina is 2.5 nm, almost equal to that of pure activated carbon. All these results indicate that the cover of the alumina membrane does not affect the microstructure of activated carbon.

Table 1 Textural and mechanical properties of the samples

	AC	Alumina	AC@alumina
BET surface area (m^2 g^−1)	810.0	251.4	727.7
Pore volume (cm^3 g^−1)	0.481	0.343	0.462
Avg. pore diameter (nm)	2.4	5.5	2.5
Micropore area (m^2 g^−1)	573.1	0	443.5
Avg. macropore diameter (nm)	176.6	—	130.9
Macropore volume (mL g^−1)	0.1125	—	0.0994
Attrition index (%)	12.6	2.9	2.9

---

In addition, mercury porosimetry was used to study difference in the pore structure of the samples and to determine pore sizes ranging from a few nanometers to several micrometers, as given in Fig. 3. The corresponding pore size distribution is also included. It is evident that the curves of two materials are similar but their pore size distribution different. There are some pores with size centered at ca. 1000 nm on AC compared with that on AC@alumina. As we known, mercury is a strongly hydrophobic liquid for almost all substances. The pore size and volume of materials can be obtained by immersing the sample under a quantity of mercury and then increasing the pressure of the mercury. With the applied pressure increasing, mercury penetrates into a gradually increasing proportion of pores with decreasing size and consequently the total amount of mercury intruded increases. Due to the coatings of alumina membrane, the mercury can not enter into the pore on AC@alumina if the pressure is not high enough, that is, the pore with larger size can not be measured. This is the reason why the peak at ca. 1000 nm is absent on the pore distribution of AC@alumina. Importantly, this also implies that alumina membrane is uniform and can protect activated carbon.

Fig. 3 Mercury porosimetry curves for the samples.

The SEM images in Fig. 1(a) and (b) clearly show presence of macroporous voids on the surface of activated carbon particles. The volume and size of the macroporous voids, determined by mercury porosimetry, are also given in Table 1. As shown, coatings of alumina membrane decreases pore volume and diameter of the macroporous voids. This indicates formation of alumina membrane on the inner surface of macroporous voids, which is confirmed by SEM seen in Fig. 4. According to the previous results,¹⁸ the diameter of boehmite colloidal particles in the sol is about 108 nm, which is much bigger than that of the micropores of activated carbon. Thus, the particles can only enter into the macropores of activated carbon in the preparation.

Fig. 4 SEM images of the external surface of activated carbon coated by alumina membrane.

Interestingly, the attrition index of activated carbon particles decreases from 12.6% to 2.9% after coated by alumina membrane, according to the mechanical properties test (Table 1), which is equal to that of pure alumina. The indexes of activated carbons with two-time coating and three-time coating are both 2.9%. In addition, the samples show good reusability. The attrition index keeps constant after several times measurements. This result indicates that the coating of inorganic membrane on solid catalyst can enhance its mechanical properties efficiently. This is because the sol–gel derived alumina has a high mechanical strength due to its nanostructure with crystallites well bound together. So the coating of the γ-Al₂O₃ membrane protects and strengthens the activated carbon particles.

Conclusions

In summary, the alumina membrane can be synthesized on the surface of activated carbon uniformly by slip-casting method. This strategy can make the attrition of activated carbon decrease from 12.6% to 2.9% and gives a novel way to enhance the mechanical properties of solid catalyst. In addition, the coatings of alumina membrane have little effect on the microstructure of activated carbon.

Acknowledgements

Financial supports from National Natural Science Foundation of China (21176211, 21376209) and the Program for Zhejiang Leading Team of S&T Innovation (2013TD07), are gratefully acknowledged. The authors also thank Prof. Y. S. Lin from Arizona State University, USA, for his valuable discussion.

Notes and references

1. J. Hagen, Industrial Catalysis: A Practical Approach, WILEY-VCH Verlag GmbH & Co., 2006.
2. D. F. Wu, J. C. Zhou and Y. D. Li, AIChE J., 2007, 53, 2618–2629.
3. J. Regalbuto, Catalyst Preparation Science and Engineering, Taylor and Francis Group, LLC, 2007.
4. R. Boerefijn, N. J. Gudde and M. Ghadiri, Adv. Powder Technol., 2000, 11, 145–174.
5. A. Thon and J. Werther, Appl. Catal., A, 2010, 376, 56–65.
6. Y. D. Li, X. M. Li and L. Zhang, Catal. Today, 1999, 51, 73–84.
7. B. M. Fedorov, V. I. Nekhoroshev and I. A. Zhukov, Kinet. Catal., 1992, 33, 151–156.
8. G. Buelna and Y. S. Lin, Microporous Mesoporous Mater., 1999, 30, 359–369.
9. G. Buelna, Y. S. Lin, L. Liu and J. D. Litster, Ind. Eng. Chem. Res., 2003, 42, 442–447.
10. M. Conte, A. F. Carley, G. Attard, A. A. Herzing and C. J. Hutchings, J. Catal., 2008, 257, 190–198.
11. F. Gonçalves and J. L. Figueiredo, Appl. Catal., B, 2004, 50, 271–278.
12. W. P. Ma, E. L. Kugler and D. B. Dadybujor, Energy Fuels, 2007, 21, 1832–1842.
13. Z. M. Wan and Y. S. Lin, J. Catal., 1998, 174, 43–51.
14. B. E. Yoldas, Am. Ceram. Soc. Bull., 1975, 54, 286–288.
15. Standard Test Method for Attrition and Abrasion of Catalysts and Catalyst Carriers, ASTM D 4058-1996.
16. R. Kousar, D. H. Kim, B. Y. Yu, H. P. Ha, S. H. Kim and J. Y. Byun, Appl. Catal., A, 2012, 423, 176–184.
17. K. S. W. Sing, D. H. Everett, R. A. Haul, W. L. Moscou, R. A. Pierotti, J. Rouquérol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603–619.
18. W. P. Yang, S. S. Shyu, E. S. Lee and A. C. Chao, Sep. Sci. Technol., 1996, 31, 1327–1343.

---

Footnote

† Electronic supplementary information (ESI) available: Some other characterization results. See DOI: 10.1039/c5ra23769f

---