butene alkylation on sulfated alumina: Influence of sulfation condition on textural, structural and catalytic properties

butene alkylation on sulfated alumina: Influence of sulfation condition on textural, structural and catalytic properties

Applied Catalysis A: General 344 (2008) 107–113 Contents lists available at ScienceDirect Applied Catalysis A: General journal homepage: www.elsevie...

700KB Sizes 0 Downloads 18 Views

Applied Catalysis A: General 344 (2008) 107–113

Contents lists available at ScienceDirect

Applied Catalysis A: General journal homepage: www.elsevier.com/locate/apcata

Isobutane/butene alkylation on sulfated alumina: Influence of sulfation condition on textural, structural and catalytic properties Marina Yu. Smirnova *, Gleb A. Urguntsev, Artem B. Ayupov, Alexei A. Vedyagin, Gennady V. Echevsky Boreskov Institute of Catalysis, Prosp. Akad. Lavrentieva 5, Novosibirsk 630090, Russian Federation

A R T I C L E I N F O

A B S T R A C T

Article history: Received 6 December 2007 Received in revised form 4 April 2008 Accepted 4 April 2008 Available online 11 April 2008

Samples of sulfated alumina with different sulfate contents were prepared by treatment of g-alumina with an aqueous solution of a sulfate precursor (sulfuric acid or ammonium sulfate). Under conditions where no catalyst deactivation was observed, it was found that yield of C5+ products as well as C5–C7 products selectivity increase as the sulfate content increases up to 23 wt.% and remain practically constant with further increase in sulfate content. Catalysts deactivation was studied at higher olefin WHSV and lower isoparaffin/olefin ratio in the feed. Increasing the sulfate content up to 14 wt.% led to an increase in activity. A reduction in activity was observed when the sulfate content increased further. Samples with similar sulfate content obtained from different sulfate precursors exhibited similar butenes conversion and product distributions under both conditions. The differences in texture of sulfated alumina catalysts prepared from different sources of sulfate were determined by nitrogen adsorption. Different sulfate species, the formation of which is influenced by the source and sulfate content, were detected by DTG. Formation of sulfate-containing phases on the samples with sulfate content of more than 8 wt.% was found by X-ray diffraction (XRD). The appearance of these phases resulted in a significant drop in surface area and pore volume. ß 2008 Elsevier B.V. All rights reserved.

Keywords: Sulfated alumina Alkylation Isobutane Butene

1. Introduction Sulfate-promoted oxides are widely known as solid acid catalysts for hydrocarbon transformations [1–4]. Sulfated zirconia (SZ) has attracted a great deal of attention from researchers because it has the highest acid strength (=0  16.04) and activity. Sulfated alumina (SA) with =0  14.5 has been considered relatively unpromising as a solid acid due to its lower activity and acidity in comparison with most other sulfated oxides [5,6]. However, wide availability, heat stability and low price make it an attractive catalyst. Moreover, the phase transitions that are known for zirconia are not observed during preparation of SA. Although there is a large body of works devoted to the study of the dependence of SZ catalytic activity on its preparation history, determining the appropriate conditions for SA synthesis remains a major issue. Moreover, the influence of sulfate content on the physical and catalytic properties of SA samples has not been extensively investigated. It is well known that the optimum sulfate

* Corresponding author. Tel.: +7 383 3269589; fax: +7 383 3309827. E-mail addresses: [email protected] (M.Yu. Smirnova), [email protected] (G.A. Urguntsev), [email protected] (A.B. Ayupov), [email protected] (A.A. Vedyagin), [email protected] (G.V. Echevsky). 0926-860X/$ – see front matter ß 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2008.04.004

content for SZ does not exceed the amount required for monolayer surface coverage, and any further increase results in the reduction of catalytic activity [7]. The usual preparation method for catalysts based on zirconium, tin and titanium oxides includes treatment with 0.5 M sulfuric acid [2,8]. Synthesis of active SA samples using 2.5 M sulfuric acid was reported by Hino and Arata. However, the sulfate content of these SA samples after calcination was not given [6]. An SA sample with sulfate content much greater than the optimum sulfate content for SZ was prepared with 4.8 M sulfuric acid [10]. However, its catalytic activity was not measured [10]. No X-ray diffraction (XRD) peaks indicating the presence of a zirconium sulfate phase were observed for SZ samples with sulfate content higher than that required for formation of monolayer. To explain this, it was proposed that some sulfate groups could migrate into the bulk of the zirconia or be located in additional layers above the monolayer [7,9]. Furthermore, as predicted in [7], conversion of the SZ surface into zirconium sulfate should destroy catalytic activity. In contrast, appearance of the aluminium sulfate phase in SA samples did not lead to an appreciable change of their activity in n-butane isomerization [10]. It is reasonable to suspect based on references above that active SA catalysts could be prepared in wider range of sulfate concentrations than SZ catalysts, and different optimal sulfate content is needed for SA catalysts than for other sulfated oxides.

M.Yu. Smirnova et al. / Applied Catalysis A: General 344 (2008) 107–113

108

The aim of this paper is to describe a detailed study of the influence of sulfation conditions (source and concentration of sulfate) on the textural, structural and catalytic properties of SA focusing in particular on differences from SZ. 2. Experimental 2.1. Catalysts preparation Commercial pseudoboehmite particles (Katalizator) of 0.5– 0.8 mm size were calcined at 550 8C for 2 h in order to prepare galumina. SA samples were prepared by treatment of the resultant g-alumina (surface area 229 m2 g1, pore volume 0.39 cm3 g1) with an aqueous solution of (NH4)2SO4 or H2SO4 (catalysts were referred as SN and SH, respectively). The support was immersed into a sulfate precursor solution with the desired concentration (the liquid/solid volume ratio was maintained at 3); the resultant slurry was dried at 120 8C for 12 h and calcined at 550 8C for 2 h. SA samples after calcination were denoted as SHc or SNc. The composition and preparation conditions of SA samples are presented in detail in Table 1. The SH-HT-32 sample of basic aluminium sulfate (3Al2O3 4SO38H2O) supported on g-alumina was hydrothermally synthesized from g-alumina (6.1 g) and 72 ml of 0.407 M H2SO4 solution. The hydrothermal treatment was carried out at 150 8C for 24 h. After the run the solid product was filtered and calcined in air at 550 8C for 2 h. After calcination this sample was denoted as SH-HT-32c. Sulfate content was measured as sulfur on Vario EL III CHNOS Elemental analyser. 2.2. Catalyst characterization High-resolution adsorption–desorption isotherms of N2 at 196 8C were taken in the 106 to 0.995 P/P0 range with the static volumetric instrument Autosorb-1-C (Quantachrome Instruments). The BET method was used to evaluate specific surface area. The pore size distribution (PSD) and micropore volume were calculated by the NLDFT method from the adsorption branch of the isotherm with a software package supplied with the Autosorb-1-C. X-ray diffraction patterns were recorded using a HZG-4c diffractometer with CuKa radiation (l = 1.54178 A˚) equipped with a graphite monochromator. Data were recorded in the 58 and 508 2u angle range with a step size of 0.028 2u and counting time of 3 s per step. Thermogravimetric analysis was performed on a Netzsch STA 449C instrument. The sample temperature was linearly raised to 1000 8C in a 30 ml/min air flow with a heating rate of 58/min. 2.3. Catalytic test Isobutane (>99% purity) and isobutane/n-butenes feed with isoparaffin/olefin (i/o) molar ratios of 4.5 and 9.5 were used for catalytic experiments. The liquid-phase alkylation reaction was Table 1 Composition and preparation conditions of SA samples Sample

Source of sulfate

C(SO4)0 (wt.%)

C(SO4)after calcination (wt.%)

SH-5 SH-10 SH-20 SH-30 SH-40 SN-20 SN-40

H2SO4 H2SO4 H2SO4 H2SO4 H2SO4 (NH4)2SO4 (NH4)2SO4

5 10 20 30 40 20 40

4.5 8.3 14.2 22.9 33.6 15.6 29.2

carried out in an autoclave with agitation at the reaction temperature of 50 8C. Each SA sample was activated ‘‘ex-situ’’ at 550 8C for 1 h in air before starting the reaction runs. After activation catalyst (2 g) was loaded into an autoclave filled with argon and then isobutane was charged. After that, isobutane/nbutenes feed maintained in the liquid phase under argon pressure was fed to the autoclave. To compare the catalytic parameters of catalysts without deactivation, the reaction was carried out under mild conditions: 25 g of isobutane was charged into the autoclave, feed with i/o of 9.5 was used and the WHSV of butenes was kept at 0.1 h1. To observe catalyst deactivation, we loaded 13 g of isobutane into the autoclave and used a feed with i/o of 4.5. The WHSV of butenes in this case was 0.3 h1. The composition of the reaction product was determined by gas chromatography with a DB-1 capillary column. The reaction mixture from the autoclave was sampled periodically with a syringe specially designed for sampling liquid hydrocarbons under pressure. The following parameters were determined: conversion of butenes (X C4 ¼, wt.%), total yield of C5+ products divided by the amount of butene reacted (Y C5þ , wt.%), selectivity to C5–C7 and C8 products (SelC5 C7 ; SelC8 , wt.%) and selectivity to trimethylpentanes (SelTMPs, wt.%). The average value over the total catalytic run for each parameter was calculated for experiments carried out under mild conditions. 3. Results 3.1. X-ray diffraction data Fig. 1A presents XRD patterns of SH samples before calcination. No peaks are observed for SH-10. The pattern of SH-20 exhibits only the characteristic diffraction peaks of aluminium sulfate hydrate (the patterns of SH-10 and SH-20 are not shown in Fig. 1). Besides these, peaks corresponding to the basic aluminium sulfates appear in patterns of SH-30 and SH-40. The XRD patterns of calcined SA samples with various sulfate content are shown in Fig. 1B. All samples but SH-10c contain anhydrous aluminium sulfate. The patterns of SH-30c, SH-40c and SN-40c exhibit an additional diffraction peak of basic aluminium sulfate not observed in the XRD patterns of SH-30 and SH-40. As can be seen from Fig. 1, the content of sulfate phases increases with the increase of sulfate amount in SA samples. However, none of the basic aluminium sulfates registered in SH-30 and SH-40 is detected in the XRD patterns of the calcined samples. 3.2. Influence of source and sulfate content on surface area, pore volume and porous structure of SA samples N2 adsorption–desorption isotherms of SHc samples with different sulfate contents have the H3 type of hysteresis loop [11], which is associated with slit-shaped pores (that is quite probable for this support) or with ink-bottle pores. This means that the desorption branch of the isotherm is not equilibrium and should not be used for porosity analysis. To get the correct PSD, the adsorption branch is analyzed with the NLDFT-adsorption kernel [12]. Specific surface area and pore volume for SA samples depending on sulfate content and sulfate precursor are shown in Table 2. The increase in sulfate content up to the amount required for a half of monolayer results in a pore volume decrease of SH-10c but has no effect upon its surface area (the sulfate monolayer content for our g-alumina calculated on the basis of four sulfate groups per square nanometer is 14.6 wt.%). We suggest that for this sample a homogeneous distribution of sulfate over the support surface is realized. Further increases in sulfate content lead to a decrease in

M.Yu. Smirnova et al. / Applied Catalysis A: General 344 (2008) 107–113

109

Fig. 1. XRD powder patterns of SA samples before calcination (A) and calcined at 550 8C (B).

Fig. 2. PSD of SA samples with different sulfate content (A) and source of sulfate (B).

both the pore volume and surface area for samples treated either by ammonium sulfate or by sulfuric acid. The following processes, which are induced by increasing the sulfate concentration in a treating solution, could explain this: the appearance of sulfatecontaining phases (as evidenced by XRD; see Section 3.1), the growth of their crystal size and increase in their content on the support surface. It should be noted that SH-20c and SN-20c samples have similar surface area and sulfate content. The surface area of SN-40c is similar to that of SH-30c, but its sulfate content is significantly greater.

Changes in the porous structure of alumina due to sulfation are shown in Fig. 2. Fig. 2A presents the PSD of SHc samples depending on sulfate content when sulfuric acid is used as a source of sulfate. It is clear that an increase in sulfate content leads to a narrowing of the PSD and shifts its maximum towards a smaller pore diameter. In addition, a dramatic drop in PSD intensity is observed for SH40c. We think that sulfate-containing phases with poor dispersion could be the reason for pore blocking or the transformation of some mesopores into micropores. This is confirmed by a more abrupt decrease in the pore volume than in the surface area for SH-40c sample and an increase in the Vmicro/V ratio for SH-40c and SN-40c samples over the ratio for SH-30c and SN-20c correspondingly. It is noteworthy that the porosity of the samples treated with sulfuric acid differs more from that of the initial alumina than the porosity of the samples treated with ammonium sulfate (Fig. 2B). These results give evidence of a more uniform distribution of sulfate on the surface of SNc samples than on surface of SHc samples. Thus, the influence of various sulfate precursors on the texture of SA catalysts is different. Application of sulfuric acid as a source of sulfate leads to dissolution of the support followed by deposition of aluminium sulfate and other sulfate-containing phases on the surface of the alumina during the drying stage. In contrast, the interaction between the alumina and ammonium sulfate occurs only during calcination.

Table 2 Specific surface area and pore volume of SAc samples depending on source of sulfate and sulfate content Sample

SBET (m2 g1)

Pore volume V (cm3 g1)

Vmicro (cm3 g1)

Vmicro/ V (%)

g-Al2O3 SH-10c SH-20c SH-30c SH-40c SN-20c SN-40c

229 228 199 155 120 209 162

0.39 0.34 0.32 0.28 0.14 0.36 0.26

0.0023 0.0109 0.0070 0.0048 0.0034 0.0054 0.0058

0.6 3.2 2.2 1.7 2.4 1.5 2.2

110

M.Yu. Smirnova et al. / Applied Catalysis A: General 344 (2008) 107–113

Fig. 4. DTG profiles of SH-HT-32 sample before calcination and calcined at 550 8C.

Fig. 3. DTG profiles of SA samples before calcination (A) and calcined at 550 8C (B). Note: DTG profiles of SA samples in Fig. 3A are moved apart along Y-axis for clarity.

3.3. Thermogravimetric analysis DTG profiles of SA samples before calcination are shown in Fig. 3A. The low temperature peak centered at 110–142 8C is assigned to water removal. Higher temperature peaks up to 1000 8C are associated with the decomposition of sulfate species. Only the peak of the most strongly bonded sulfate species centered at 930 8C is observed in the profile of SH-5. The increase in the sulfate content results in the shift of this peak towards lower temperatures, as is seen for SH-10 sample, and this peak broadening for the SH-20 sample. Such thermal decomposition agrees well with the literature [9,10]. The peaks at 650–680 8C, 730–760 8C and 806–820 8C caused by decomposition of sulfate-

containing phases emerge in the profiles of SH-30 and SH-40. The peak with the maximum at 805–820 8C corresponds to aluminium sulfate decomposition. This phase was identified by the DTG profile of Al2(SO4)318H2O, which is not shown in Fig. 3. The DTG profiles of SH-30c and SH-40c shown in Fig. 3B exhibit a significant reduction in the intensity of peak at 650–690 8C corresponding to weakly bonded sulfate species, while the magnitude and position of peaks decomposing above 700 8C change slightly. It should be mentioned that the decomposition profile of weakly bonded sulfate species is similar for SN-40c and SH-40c but quite different for SN-20c and SH-20c. The DTG profile of SN-20c is shifted towards higher temperatures than the SH-20c profile. As discussed earlier, some phases of basic aluminium sulfates are observed in the XRD patterns of SH-30 and SH-40. These phases disappear from XRD patterns after calcination and another phase of basic aluminium sulfate emerges. Based upon these results, it can be proposed that basic aluminium sulfates undergo structural transformations during calcinations, and these transformations are reflected in the DTG profiles of samples with high sulfate content. Note that similar DTG profiles were obtained for SA samples with high sulfate content in ref. [10]. The appearance of a peak at 630 8C was associated with amorphous multilayer sulfate decomposition [10]. Calcination can induce decomposition of crystalline basic aluminium sulfates followed by amorphization or/and transformation into phase with different composition. To confirm this, the SH-HT-32 sample of basic aluminium sulfate (3Al2O34SO38H2O) supported on g-alumina was prepared. This phase is one of basic aluminium sulfates which we observed in SH-30 and SH-40 samples. Its thermal behavior before and after calcination at 550 8C is presented in Fig. 4. The peak at 850 8C is observed in both profiles, while the peak at 430 8C disappears after calcination. Calcination leads to an essential reduction in the intensity of XRD peaks corresponding to 3Al2O34SO38H2O (Fig. 4, overlapping mode). However, no other phases are detected in SH-HT-32 and SH-HT-32c samples. 3.4. Catalytic activity It is well known that the chemistry of paraffin/olefin alkylation is complex. This reaction is accompanied by a broad range of reactions with different demands in catalyst acidity. Focusing only on the observed olefin conversion does not allow estimation of

M.Yu. Smirnova et al. / Applied Catalysis A: General 344 (2008) 107–113

111

Table 3 Catalytic properties of SA depending on sulfate content and source of sulfate. Conditions: T = 50 8C, i/o in the feed = 9.5, olefin WHSV = 0.1 h1, mcat = 2 g, TOS = 4 h Sample

SH-5c

SH-10c

SH-20c

SN-20c

SH-30c

SH-40c

SN-40c

X C4¼ (wt.%) Y C5þ (wt.%) SelC5 C7 (wt.%) SelC8 (wt.%) SelTMPs (wt.%) TMP/DMH

0 0 0 0 0 0

98.8 157 6 84 76 9.5

99.2 188 19 64 42 1.9

99.5 185 21 69 48 2.3

98.6 217 35 49 22 0.8

95.7 215 31 52 28 1.2

98.2 218 41 46 22 0.9

either the alkylation activity of various catalysts or the degree to which side reactions, such as cracking and oligomerization, proceed [13]. Contribution of the desired reaction, which produce TMPs, can be estimated by the alkylate yield per olefin. Yields higher than 204% (the theoretical yield corresponding to the stoichiometry of the i-C4/C4 alkylation reaction) as well as high selectivity to C5–C7 products are the result of high cracking activity; yields lower than theoretical can be explained by the contribution of oligomerization [13,14]. The catalytic parameters for SA samples depending on source and sulfate content are shown in Table 3. Since alkylation experiments were carried out at a low olefin WHSV and high i/o ratio in the feed, catalysts deactivation during catalytic run was not observed. A high level of butenes conversion (more than 95%) was observed for all samples with sulfate content upwards of 5 wt.%. The SH-5c sample was inactive in isobutane/butene alkylation. Both the alkylate yield and C5–C7 products selectivity increase with sulfate content up to 23 wt.% and remain practically constant with further increase of the sulfate content. The low TMP/DMH ratio observed for samples with a high sulfate content can be explained by the high cracking activity of these catalysts because cracking of more branched TMPs occurs to a greater extent than cracking of DMHs [13]. SA samples with similar sulfate content obtained from different sulfate precursors exhibit equal total yields of C5+ products and similar product distributions. Corma et al. [13] reported that catalyst sites with a certain acid strength are vital for alkylation reaction. The weakest acid sites show activity only in oligomerization. Alkylation is catalyzed by sites with medium–strong acidity, whereas cracking occurs only with the strongest acid sites. Based on these statements, we conclude that an increase in sulfate content leads to an increase in acid sites strength of our SA samples. Under severe reaction conditions (low i/o ratio, high butenes WHSV), our SA catalysts undergo deactivation during the catalytic

Fig. 5. Change of butenes conversion with TOS obtained under severe conditions (i/o in the feed = 4.5, olefin WHSV = 0.3 h1).

run. Catalysts with similar catalytic parameters under mild conditions exhibit different behaviors during deactivation. Butenes conversion drops with time on stream (TOS) (Fig. 5). An increase in activity is observed with an increase in sulfate content up to 14 wt.%; then the activity decreases with further increase in sulfate content. Sample SH-40c exhibits a drop in butenes conversion identical to that for SH-10c. The evolution of the reaction products with TOS is shown in Figs. 6 and 7. For all SHc samples except SH10c, a decrease in cracking products (Fig. 6) and an increase in TMPs selectivity (Fig. 7) are observed with TOS. These results can be explained considering that the strongest acid sites, which are responsible for cracking, are the first to be poisoned [13].

Fig. 6. Change of C5–C7 products selectivity with TOS obtained under severe conditions.

Fig. 7. Change of TMPs selectivity with TOS obtained under severe conditions.

112

M.Yu. Smirnova et al. / Applied Catalysis A: General 344 (2008) 107–113

It should be mentioned that similar butenes conversions and product distributions are observed for SA samples with similar sulfate contents, but are prepared from different sulfate precursors. 4. Discussion As was proposed in Section 1, different optimal sulfate content is needed for SA catalysts than for other sulfated oxides. Therefore, in the present study the concentration of sulfate in SA samples is varied over a wide range: from 4.5 to 33.6 wt.% (0.3–2.3 monolayer). Increasing the sulfate content up to 14 wt.% led to an increase in catalytic activity. Further increase of the sulfate concentration resulted in decreasing of catalytic activity and faster deactivation. Thus, the optimal sulfate concentration for our SA catalysts as well as for SZ is close to the amount required to form a monolayer. We already mentioned that SZ samples completely loose their activity when the zirconium sulfate phase is formed [7]. In our case, the sulfate-containing phases were not detected in the inactive SH-5c and low-activity SH-10c sample however were present in small amounts in the most active SH-20c. Further increase in the sulfate concentration led to the growth of the amount of sulfate phases and was accompanied by a decrease in the surface area and pore volume. These results pose a question: how does the presence of the sulfate phases affect the catalytic activity of the SA samples, and what is their role in the formation of the active sites? First, it is necessary to figure out what is the structure of the SA active sites. The active site models suggested for SZ and SA in the literature consist of monosulfate or polysulfate species bonded to the oxide surface [2,15–17]. As the SH-5c was absolutely inactive, it is natural to conclude that single sulfate groups cannot be considered as active sites. The catalytic activity appeared when the sulfate content exceeded the half of monolayer concentration and reached a maximum for SH-20c sample, whose sulfate concentration approached the monolayer. One can assume that such sulfate concentrations lead to reconstruction of support surface and formation of a ‘‘surface sulfate phase’’. Such phase might correspond to the surface of a 3D sulfate-containing phase or could be formed on the support surface as a 2D phase. Aluminium sulfate and basic aluminium sulfate were the bulk sulfate-containing phases detected by XRD. It is known that Al2(SO4)3 is not active in acid-catalyzed reactions [6]. It is clear that formation of this catalytically inert phase in large amounts and its deposition on the support surface led to a decrease in catalytic activity. 2Al2O34SO3xH2O is another crystalline phase formed in appreciable amounts on the samples with high sulfate concentrations. The comparison of the XRD data for SH-30c, SH-40c and SN40c with their catalytic activities allows us to conclude that this phase is either catalytically inert or has negligible activity. Thus, catalytic activity of SA is not caused by a surface of crystalline sulfate-containing phases. In addition to the crystalline phases, amorphous or finecrystalline phases, which cannot be detected by XRD, may exist on the surface of SA samples. The complex shape of the DTG profiles of SA samples with high sulfate content indicates that such phases may be present. The experimental results do not contradict the hypothesis that the active sites could be formed on the surface of the amorphous sulfate-containing phases. However, the potential structural instability of an amorphous phase gives ground to anticipate another route for active sites formation. The following route can be proposed: the decomposition of the sulfate-containing phases occurred during calcination of SA samples producing dispersed particles of

alumina. Such nanocrystals with many surface defects may be the location of active sites formation. It seems that similar route of formation catalytically active sites occurs in studies [18,19], where SZ samples were prepared by thermolysis of zirconium sulfate. At this moment, we are not able to determine unambiguously the active sites’ location (either the surface of bulk amorphous phase or the surface of the support). However, there is no doubt that sulfate-containing phases affect the formation of the active sites. This conclusion is based on the fact that while SH-10c and SH-40c have similar activity, many more cracking products are obtained with SH-40c, indicating that its acid sites are stronger. The different sulfate species formed on SA samples depending on the source and sulfate content were detected by DTG. We tried to correlate, on a qualitative basis, the strength of acid sites with the bonding energy of sulfates to the support surface. Putting together the DTG profiles of SH-5c and SH-10c samples with their catalytic properties, we have concluded that the most strongly bonded sulfate species are responsible for the formation of sites with low acid strength. Increasing the sulfate content results in a reduction of bonding energy and leads to an increasing strength of acid sites. From comparing the DTG profiles and catalytic properties of SH-20c and SN-20F, it is clear that samples with similar catalytic parameters have different thermal behaviors in the low temperature region. These data allow us to suggest that moderately bonded sulfate species produce the strongest acid sites. These conclusions are in a good agreement with some of the literature. According to ref. [9], sulfates with relatively low interaction energy induce the catalytic activity of SZ in isomerization of n-butane. However, the ‘‘surface sulfate’’, which was the most strongly bonded sulfate species, was found to produce the activity of SA in this reaction [10]. Weakly bonded sulfate species in large quantities were observed in the DTG profiles of our SA samples with the highest sulfate content. Sulfate species with low thermal stability were associated with weak acidity in ref. [10]. In our case, we suggest that such DTG profiles reflect structural transformations of sulfate-containing phases and DTG peaks may not be directly associated with sulfates species producing active sites. In addition to the effect of sulfate concentration on physical and catalytic properties of SA, the influence of the sulfate precursor on these properties has been studied in this paper. As noted above, the catalytic properties and contents of crystalline phases are similar for SA samples prepared from different sulfate precursors. However, some differences are observed in texture (surface area, pore size distribution), especially for the samples with high sulfate content. The sulfate concentration of SN-40c is close to that of SH-40c, whereas the surface area of the former sample is as high as that of SH-30c. Differences in textural properties are the result of support dissolution followed by deposition of sulfate phases during preparation of SH samples. In the case of SN samples, dissolution does not occur, and sulfate is distributed more uniformly. Note that samples from different sulfate precursors have different thermal behaviors. Therefore, it was concluded that their phase composition, including amorphous phases, is different. However, in spite of some differences in texture and phase composition, the formation of active sites of similar concentrations and strengths occurs on samples treated with different sulfate precursors. This is confirmed by the fact that they have similar catalytic parameters both under conditions where no catalyst deactivation occurs and under those favoring deactivation. Thus, the sulfate concentration is a more important factor during SA synthesis than the type of sulfating agent.

M.Yu. Smirnova et al. / Applied Catalysis A: General 344 (2008) 107–113

113

5. Conclusions

Acknowledgements

In the present study, it was shown that SA samples in contrast with SZ are active in the presence of significant amounts of sulfatecontaining phases. It was found that the catalytic properties of SA samples in isobutane/butene alkylation depend slightly on the sulfate precursor and strongly on its concentration. The catalytic activity appeared when the sulfate content exceeded the half of monolayer concentration and reached a maximum for the sample, whose sulfate concentration approached the monolayer. It was concluded that single sulfate groups attached to the alumina surface do not produce active sites; active sites originate from the ‘‘surface sulfate phase’’. Further increase of the sulfate concentration resulted in decreasing of catalytic activity and led to the growth of the amount both of crystalline and amorphous sulfate phases. Crystalline sulfate-containing phases were inactive in alkylation and led to a reduction in surface area. Based on the catalytic parameters obtained under conditions where no catalyst deactivation was observed, we concluded that an increase in sulfate content leads to an increase in acid sites strength of our SA samples. These data and the fact that sulfate-containing phases were present in the most active sample led to conclude that the sulfate-containing phases play an essential role in the active sites formation.

The authors would like to thank Alexander V. Toktarev, Boreskov Institute of Catalysis, for helpful discussions.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

B.H. Davis, R.A. Keogh, R. Srinivasan, Catal. Today 20 (1994) 219. X. Song, A. Sayari, Catal. Rev. -Sci. Eng. 38 (1996) 329. K. Arata, Appl. Catal. A 146 (1996) 3. K. Arata, H. Matsuhashi, M. Hino, H. Nakamura, Catal. Today 81 (2003) 17. K. Satoh, H. Matsuhashi, K. Arata, Appl. Catal. A 189 (1999) 35. K. Arata, M. Hino, Appl. Catal. 59 (1990) 197. D. Faˇrcas¸iu, J.Q. Li, S. Cameron, Appl. Catal. A 154 (1997) 173. A. Corma, A. Martinez, C. Martinez, Appl. Catal. A 144 (1996) 249. A.F. Bedilo, K.J. Klabunde, J. Catal. 176 (1998) 448. T. Yang, T. Chang, C. Yeh, J. Mol. Catal. A 115 (1997) 339. K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Mouscou, R.A. Pierotti, J. Rouquerol, T. Siemieniewska, Pure Appl. Chem. 57 (1985) 603. M. Thommes, B. Smarsly, M. Groenewolt, P.I. Ravikovich, A.V. Neimark, Langmuir 22 (2006) 756. A. Corma, A. Martinez, C. Martinez, J. Catal. 149 (1994) 52. X. Xiao, J.W. Tierney, I. Wender, Appl. Catal. A 183 (1999) 209. M.L. Ferreira, E.H. Rueda, J. Mol. Catal. A 178 (2002) 147. J. Yang, M. Zhang, F. Deng, Q. Luo, D. Yi, C. Ye, Chem. Commun. (2003) 884. M. Hino, M. Kurashige, H. Matsuhashi, K. Arata, Thermoch. Acta 441 (2006) 35. E.E. Platero, M.P. Mentruit, Mater. Lett. 14 (1992) 318. Y. Huang, B. Zhao, Y. Xie, Appl. Catal. A 173 (1998) 27.