Effect of precursor on the performance of phosphate-modified γ-Al₂O₃ catalysts for the dehydration of methanol

Abstract

A series of PO₄³⁻/γ-Al₂O₃ monolith catalysts were prepared by incipient wetness impregnation of phosphate with different additives, including ammonia, triethanolamine (TEA), hexamethylenetetramine (HMT) and diethylenetriamine (MSDS). These catalysts were compared for their catalytic properties in a fixed-bed reactor at 1 atm, 250–350 °C and liquid hourly space velocity (LHSV) of 4 h⁻¹. The catalysts were characterized by nitrogen adsorption, X-ray diffraction (XRD) and temperature-programmed desorption of ammonia (NH₃-TPD). The results revealed that the catalytic activity of PO₄³⁻/γ-Al₂O₃ for methanol dehydration increased significantly as the content of phosphate increased to 7 wt%. However, when the phosphate content of PO₄³⁻/γ-Al₂O₃ was further increased to 10 wt%, the activity for methanol dehydration decreased. The catalytic activity of PO₄³⁻/γ-Al₂O₃ is found to be determined mainly by the amounts of weak and medium acid sites. At 7 wt% PO₄³⁻ loading, the proportion of weak and medium acid sites can be improved by addition of additives with high basicity to the impregnation process. The catalyst containing 7 wt% phosphate with added diethylenetriamine demonstrated the highest activity.

---

1 Introduction

In addition to being an important intermediate for production of many valuable chemicals such as olefins, methyl acetate and dimethyl sulfate, dimethyl ether (DME) has received global attention since it has great potential as a clean alternative fuel for diesel engines because its thermal efficiency is equivalent to traditional diesel fuel, it has much lower NOx and SOx emissions, near-zero smoke production and so on.^1,2 The production of DME via methanol dehydration is more favorable with regards to thermodynamics and economy.³

DME is obtained by methanol dehydration reactions using acid catalysts such as zeolite materials, γ-Al₂O₃ and acid-modified γ-Al₂O₃. Zeolites are characterized by their high acidity, however, such high acidity can result in significant coke formation and consequently fast deactivation.^4–8 Furthermore methanol can undergo secondary reactions to produce hydrocarbons at temperatures higher than 240 °C.⁹ γ-Al₂O₃ has been mostly employed due to its low price, easy availability and high stability⁸ but low activity at temperatures lower than 300 °C.⁹ Hence, the modification of γ-Al₂O₃ to increase its catalytic activity for methanol dehydration has become a very important step.

It has been reported that a 10 wt% Nb₂O₅-modified γ-Al₂O₃ catalyst exhibits a higher activity for methanol dehydration than the untreated γ-Al₂O₃ at 240–260 °C.⁸ Moreover, the sulfate ions incorporated into 10 wt% Nb₂O₅/γ-Al₂O₃ increased the amount of acid sites from 1.35 mmol g⁻¹ to 1.68 mmol g⁻¹, which resulted in higher catalytic activity at 240 °C. But the stability of sulfate catalysts was not reported. Khaleel¹⁰ found that Ti-modified alumina also showed higher catalytic activity and selectivity for methanol to DME at temperatures ≤220 °C. Yaripour and Baghaei¹¹ reported that the silica modified γ-Al₂O₃ catalysts also showed better performance for the methanol dehydration reaction, in which the sample with 6 wt% silica loading exhibited the best methanol conversion.

Mao¹ found that 10 wt% SO₄²⁻/γ-Al₂O₃ calcined at 550 °C exhibited the high selectivity and yield for DME from syngas. Yaripour and Baghaei¹¹ concluded that phosphorus-modified catalysts have shown better performance compared to the untreated γ-Al₂O₃ for methanol to DME. Panpranot¹² reported that amorphous AlPO₄ catalyst pretreated at 200–300 °C with steam exhibited higher methanol conversion than non-treated catalyst.

Concerning the acid strength, some authors^13,14 suggested (mostly based on NH₃-TPD profiles) that acid sites of medium strength are the most desirable from the viewpoint of DME selectivity, while strong acid sites do promote the formation of by-products at the expense of DME; others,¹¹ however, stated that acid sites of relatively weak and medium strength (again based on NH₃-TPD results) can also efficiently convert methanol to DME or even that weak acid sites are more effective for the selective DME synthesis.²

It is apparent from the above discussions that a systematic work is required in order to gain more insights into the influence of the acid property of the dehydration catalysts. For this purpose, we have prepared a series of PO₄³⁻/γ-Al₂O₃ catalysts in which the γ-Al₂O₃ acidity was systematically varied by different amount and types of additives. The acid properties of the samples have been measured by NH₃-TPD. The effects of phosphorus on the catalytic activity have been investigated. In addition, the activity was correlated with the amount of surface acid sites of these catalysts.

2 Experimental

2.1 Preparation of catalysts

2.1.1 Preparation of γ-Al₂O₃ supports. Alumina columns were prepared according to the procedure described in Liu et al.¹⁵ Typically, a mixture was produced in a blender to form homogeneous dough with component of pseudo-boehmite (100 g), methylcellulose (2 g), 68% nitric acid (10 g) and water (120 g). The dough was extruded through a die into columns with a diameter of 3 mm. Finally, the green columns were dried at 120 °C for 24 h and then calcined at 600 °C for 4 h.

2.1.2 Phosphate-modified Al₂O₃ catalysts. The PO₄³⁻/γ-Al₂O₃ samples with various phosphate contents were prepared by wet impregnation of γ-Al₂O₃ columns with an aqueous solution containing suitable amount of (NH₄)₂HPO₄, followed by drying at 110 °C overnight and then calcined at 600 °C in air for 5 h. The samples were designated as xPO₄³⁻/Al₂O₃, where x stands for the weight percentage of phosphate ion.

The second series of samples was prepared following the above procedure, but (NH₄)₂HPO₄ was substituted with 28% ammonium hydroxide and phosphoric acid. This series of samples was denoted A(y)-7PO₄³⁻/Al₂O₃, where A and y stand for the ammonium hydroxide and n(NH₃)/n(PO₄³⁻) ratios of impregnation solution, respectively.

The third type of catalysts was also prepared following the above procedure, but ammonium hydroxide was substituted with TEA, HMT and MSDS. This series of samples with TEA, HMT and MSDS were named T(3)-7PO₄³⁻/Al₂O₃, H(3)-7PO₄³⁻/Al₂O₃ and M(3)-7PO₄³⁻/Al₂O₃, respectively. 3 means the mole ratios of nitrogen atoms in additives to H₃PO₄.

2.2 Characterization of the supports and catalysts

Pore characteristics of the samples were obtained from adsorption–desorption isotherms of nitrogen at 77 K with a Micromeritics ASAP2000 apparatus. The specific surface areas were calculated using BET equation. The total pore volume was calculated from a single point on the nitrogen adsorption isotherm at the relative pressure of about 0.995.¹⁵

The XRD patterns of the prepared catalysts were measured from10° to 80° 2θ using an X-ray diffract meter (Rigaku Dmax-2000) with Cu-Kα radiation. Prior to the XRD analyses, the samples were grinded to sizes smaller than 0.1 mm.

The NH₃-TPD was performed in a Micromeritic Chemisorb2020 equipped with a thermal conductivity detector (TCD). Before adsorption, the sample was pretreated in high purity N₂ (30 cm³ min⁻¹) at 550 °C for 1 h. Then, it was saturated with 5% NH₃/He at 100 °C for 1 h (30 cm³ min⁻¹) and subsequently flushed with flowing He (30 cm³ min⁻¹) at 100 °C for 1 h to remove physisorbed NH₃. The NH₃-TPD was carried out from 100 to 600 °C at a constant heating rate of 5 °C min⁻¹.

2.3 Catalytic tests

The vapor phase dehydration reaction was carried out in a flow fixed-bed reactor (stainless-steel tube i.d., 10 mm; length, 240 mm). In an experiment, 4.0 g of catalyst (size, 40–60 mesh; volume, 9 ml) was loaded in the middle section of the reactor tube. Then, methanol was pumped with LHSV of 4 h⁻¹ by a set of metering pumps (ILSHIN Autoclave Co., Ltd.), preheated to temperature of 240 °C and introduced into the reactor. The reaction products were analyzed on line with a gas chromatograph (HP 4890D) equipped with a compacted column (Carboxen 1000, 3 m × 0.3 cm) connected to a TCD detector, where inorganic gases (CO and CO₂) were measured; and with a compacted column (Porapak-Q, 2.5 m × 0.3 cm) connected to a flame ionization detector to analyze organic compound (CH₃OH, CH₃OCH₃, light hydrocarbon).¹⁶ Methanol conversion (X) and product selectivity (S) were calculated as follows:

Methanol conversion:

DME selectivity:

3 Results and discussion

3.1 Characterizations of the samples

Table 1 summarizes the specific BET surface areas, pore volumes and pore diameters of the catalysts employed in this paper. The specific surface area and pore volume of γ-Al₂O₃ support are 325 m² g⁻¹ and 0.58 cm³ g⁻¹, respectively. The decrease of surface areas and pore volumes with the increase of PO₄³⁻ content may be caused by the formation of AlPO₄.

Table 1 Properties of the catalysts

Sample	SA (m^2 g^−1)	VP (cm^3 g^−1)	dP (nm)	Acidity^a (mmol g^−1)
				Weak^b	Moderate^c	Strong ^d	Total
a Determined by NH3-TPD.b Fitted peak located at 230 ± 20 °C in Fig. 2.c Fitted peak located at 290 ± 20 °C in Fig. 2.d Fitted peak located at 400 ± 30 °C in Fig. 2.
γ-Al2O3	325	0.58	7.1	0.15	0.22	0.31	0.68
3PO4^3−/Al2O3	317	0.55	6.9	0.16	0.28	0.40	0.84
7PO4^3−/Al2O3	307	0.50	6.5	0.18	0.32	0.60	1.10
10PO4^3−/Al2O3	264	0.40	6.1	0.18	0.34	0.62	1.14
A(1)-7PO4^3−/Al2O3	305	0.49	6.4	0.17	0.30	0.61	1.08
A(2)-7PO4^3−/Al2O3	307	0.50	6.5	0.19	0.34	0.59	1.12
A(3)-7PO4^3−/Al2O3	308	0.51	6.6	0.21	0.38	0.54	1.13
T(3)-7PO4^3−/Al2O3	304	0.48	6.3	0.18	0.31	0.60	1.09
H(3)-7PO4^3−/Al2O3	306	0.50	6.5	0.20	0.38	0.52	1.10
M(3)-7PO4^3−/Al2O3	310	0.52	6.7	0.31	0.38	0.48	1.17

---

The XRD patterns of the phosphorus-modified γ-Al₂O₃ samples are shown in Fig. 1. We can see that all the samples have a crystallite γ-Al₂O₃ phase (JCPDS no: 1-1307). These results indicated that the AlPO₄ in amorphous form highly dispersed on the surface of γ-Al₂O₃.

Fig. 1 XRD patterns of aluminium phosphate catalysts.

The surface acidic properties of the catalysts were characterized by NH₃-TPD, as shown in Fig. 2. In all cases, NH₃-TPD patterns are characterized by an intense desorption peak located at ca. 200 °C and abroad asymmetric decaying “tail” which extends up to ca. 600 °C. Qualitatively similar NH₃-TPD profiles are typically observed for different crystalline phases of γ-Al₂O₃.^12,17,18 In order to determine the acid strength and the amount of acid sites on catalyst surface, the NH₃-TPD profiles were deconvoluted into Gaussian peaks, on the assumptions that (a) the activation energy of desorption (E_d) is constant and independent of surface coverage, and (b) a normal distribution of E_d arises from a corresponding heterogeneity of sites distribution.^12,19 The fitting procedure showed that all experimental curves may be de-convoluted into three peaks, which located at 230 ± 20 °C (peak I), 290 ± 20 °C (peak II) and 400 ± 30 °C (peak III), respectively. Peaks I and II may be attributed to desorption of NH₃ from weak and medium acid sites, whereas peak III corresponds to strong acid sites.^17,20–22 The number of acid sites in these samples was calculated and is summarized in Table 1.

Fig. 2 The NH₃-TPD profiles obtained over the indicated Al₂O₃ catalysts. (A) NH₃-TPD patterns with a variation of phosphate contents (B) NH₃-TPD patterns with a variation of n(NH₄⁺)/n(PO₄³⁻) ratios (C) NH₃-TPD patterns with variation of additives.

For the xPO₄³⁻/Al₂O₃ samples, the number of the total acid sites was increased with phosphate content. Excessive phosphate loading cause the blockage of pores and significantly increase the strong acid sites (Table 1).

For the A(y)-7PO₄³⁻/Al₂O₃ samples, desorption peak was shifted to low desorption temperatures, suggesting that the acid strength of the samples decreased with the increase of the n(NH₃)/n(PO₄³⁻) ratios. Moreover, the number of weak and medium acid sites increase with the n(NH₃)/n(PO₄³⁻) ratios.

It is quite interesting to note that the amounts of weak and medium acid sites decreased as follows: M(3)-7PO₄³⁻/Al₂O₃ > A(3)-7PO₄³⁻/Al₂O₃ > H(3)-7PO₄³⁻/Al₂O₃ > T(3)-7PO₄³⁻/Al₂O₃. The sequence accords with the alkalinity of the additives (MSDS, pK_a = 10.10; ammonia, pK_a = 9.24; TEA, pK_a = 7.82 and HMT, pK_a = 5.15). This result suggests that the additive with high basicity would be in favour of the formation of highly dispersed AlPO₄.

3.2 Catalytic activity

The activity of catalysts as functions of reaction temperatures is shown in Fig. 3. For comparison, the equilibrium conversion curve for methanol dehydration to DME is also shown in Fig. 3.

Fig. 3 Catalytic test of catalyst with variation of (A) phosphate content (B) n(NH₃)/n(PO₄³⁻) ratios (C) additives.

As shown in Fig. 3A, the conversion of methanol increases with the increase of the PO₄³⁻ content at temperature below 300 °C. The acceleration of the catalytic performance was attributed to an increase of the acid sites.¹ Moreover, the 3PO₄³⁻/Al₂O₃ and 7PO₄³⁻/Al₂O₃ catalysts are 100% selective to DME within the range of temperatures studied here. Above 325 °C, the methanol conversion on 10PO₄³⁻/Al₂O₃ catalyst is lower than that of the other phosphate-modified γ-Al₂O₃ catalysts and the selectivity to DME for 10PO₄³⁻/Al₂O₃ decrease from 100% to 96%. This phenomenon can be attributed to blocking the active acid sites of γ-Al₂O₃ with phosphorus (Table 1). On the other hand, the high amount of strong acid sites do promote the formation of by-products at the expense of DME.^4,11

In the temperature region of 275 to 350 °C, the activity of catalysts with different n(NH₃)/n(PO₄³⁻) ratios and additive is improved in order of A(3)-7PO₄³⁻/Al₂O₃ > A(2)-7PO₄³⁻/Al₂O₃ > A(1)-7PO₄³⁻/Al₂O₃ and M(3)-7PO₄³⁻/Al₂O₃ > A(3)-7PO₄³⁻/Al₂O₃ > H(3)-7PO₄³⁻/Al₂O₃ > T(3)-7PO₄³⁻/Al₂O₃, respectively, which is in good agreement with the increasing order of the number of weak and medium acidic sites. Their selectivity of DME (not shown) is nearly 100%. This indicates that the number of weak and medium acidic sites plays an important role on the catalytic activity for methanol dehydration.¹¹

In order to verify above results, Fig. 4 depicts the methanol conversion as a function of the amount of weak and medium acid sites. It is clearly showing a linear relationship between these two parameters in reaction temperature region of 250–325 °C. This is in agreement with results of previous studies,¹¹ which reported the strength of weak and moderate acid sites is a crucial factor for catalytic activity of methanol dehydration.

Fig. 4 Correlation between the methanol conversion and acid sites.

3.3 Catalyst stability test

The stability of A(3)-7PO₄³⁻/Al₂O₃ and M(3)-7PO₄³⁻/Al₂O₃ catalyst were measured over a 50 h period. The operating temperature was 325 °C and the LHSV was 4 h⁻¹. Fig. 5 clearly showed that the methanol conversion remained constant for the whole period, which indicated that no noticeable deactivation of the two catalysts was occurring. This result revealed that the phosphate modified γ-Al₂O₃ had excellent stability for the synthesis of DME from methanol.

Fig. 5 Activity and stability of the PO₄³⁻/Al₂O₃ catalysts at 235 °C. (1) A(3)-7PO₄³⁻/Al₂O₃; (2) M(3)-7PO₄³⁻/Al₂O₃ LHSV = 4 h⁻¹.

4 Conclusions

A series of solid acid catalysts were prepared by impregnation method. The phosphate-modified γ-Al₂O₃ catalysts were used for the methanol dehydration reaction. The conversion of methanol was enhanced by PO₄³⁻ modification and the number of acid sites was increased by the modification. Furthermore, the proportion of weak and medium acid sites can be improved by high alkaline additives. The M(3)-7PO₄³⁻/Al₂O₃ catalyst exhibited the highest reactivity, selectivity and stability for the synthesis of DME from methanol. The high activity, selectivity and long time stability can be attributed to the high amount of weak and medium acid sites.

Acknowledgements

The authors are thankful for the financial support by Shanxi Province science and technology major projects (20111101013), the natural science foundation of Shanxi Province project (2009011011-4).

References

1. D. S. Mao, W. M. Yang, J. C. Xia, B. Zhang and G. Z. Lu, J. Mol. Catal. A: Chem., 2006, 250, 138–144.
2. L. Liu, W. Huang, Z. H. Gao and L. H. Yin, J. Ind. Eng. Chem., 2012, 18, 123–127.
3. S. M. K. Aboul-Fotouh, J. Fuel Chem. Technol., 2013, 41(9), 1077–1084.
4. L. W. Zhang, J. H. Wang, P. Wu, Z. Y. Hou, J. H. Fei and X. M. Zheng, Chin. J. Catal., 2010, 31, 987–992.
5. N. Khandan, M. Kazemeini and M. Aghaziarati, Appl. Catal., A, 2008, 349, 6–12.
6. Y. Tan, H. Xie, H. Cui, Y. Han and B. Zhong, Catal. Today, 2005, 104, 25–29.
7. K. C. Tokay, T. Dogu and G. Dogu, Chem. Eng. J., 2012, 184, 278–285.
8. D. H. Liu, C. F. Yao, J. Q. Zhang, D. Y. Fang and D. S. Chen, Fuel, 2011, 90, 1738–1742.
9. Y. Fu, T. Hong, J. Chen, A. Auroux and J. Shen, Thermochim. Acta, 2005, 434, 22–26.
10. A. Khaleel, Fuel, 2011, 90, 2422–2427.
11. F. Yaripour, F. Baghaei and I. Schmidt, Catal. Commun., 2006, 6, 542–549.
12. K. Lertjiamratn, P. Praserthdam, M. Arai and J. Panpranot, Appl. Catal., A, 2010, 378, 119–123.
13. A. R. Keshavarz, M. Rezaei and F. Yaripour, J. Nat. Gas Chem., 2011, 20, 334–338.
14. Z. Hosseini, M. Taghizadeh and F. Yaripour, J. Nat. Gas Chem., 2011, 20, 128–134.
15. L. S. Liu, Z. Y. Liu, J. L. Yang, Z. G. Huang and Z. H. Liu, Carbon, 2007, 45, 2836–2842.
16. R. M. Ladera, J. L. G. Fierro, M. Ojeda and S. Rojas, J. Catal., 2014, 312, 195–203.
17. S. S. Akarmazyana and P. Panagiotopouloua, Appl. Catal., B, 2014, 145, 136–148.
18. J. Khom-in, P. Praserthdam, J. Panpranot and O. Mekasuwandumrong, Catal. Commun., 2008, 9, 1955–1958.
19. F. Arena, R. Dario and A. Parmaliana, Appl. Catal., A, 1998, 170, 127–137.
20. S. M. Kim, Y.-J. Lee, J. W. Bae, H. S. Potdar and K.-W. Jun, Appl. Catal., A, 2008, 348, 113–120.
21. E. Kraleva, R. Palcheva, L. Dimitrov, U. Armbruster, A. Brückner and A. Spojakina, J. Mater. Sci., 2011, 46, 7160–7168.
22. G. R. Moradi, F. Yaripour and P. Vale-Sheyda, Fuel Process. Technol., 2010, 91, 461–468.

---