In situ nanocrystalline HZSM-5 zeolites encaged heteropoly acid H3PMo12O40 and Ni catalyst for hydroconversion of n-octane

Chemical Engineering Science 62 (2007) 4469 – 4478 www.elsevier.com/locate/ces

In situ nanocrystalline HZSM-5 zeolites encaged heteropoly acid H3PMo12 O40 and Ni catalyst for hydroconversion of n-octane Lidong Chen, Xiangsheng Wang ∗ , Xinwen Guo, Hongchen Guo, Hai’ou Liu, Yongying Chen State Key Laboratory of Fine Chemicals, Department of Catalysis Chemistry and Engineering, Dalian University of Technology, Dalian 116012, China Received 6 December 2006; received in revised form 13 April 2007; accepted 11 May 2007 Available online 23 May 2007

Abstract The heteropoly acid H3 PMo12 O40 (PMo) and Ni in situ encaged in the secondary pore of nanocrystalline HZSM-5 zeolites was prepared from molybdenum oxide and phosphoric acid and nickel nitrate, in a slurry mixture of nanocrystalline HZSM-5 zeolite crystals and deionized water. Catalysts were characterized by ICP, FT-IR, XRD, 31 P MAS-NMR, ESR, Py-IR, NH3 -TPD, BET, SEM and TEM. PMo cannot enter the interior pores of the zeolite, but becomes encaged in the secondary pore of nanocrystalline HZSM-5 zeolite. The hydroconversion of n-octane over various catalysts was investigated in order to obtain light isomers of alkanes and aromatics products with high octane number. From the results presented in the paper, it is clear that the catalytic properties of the “in situ encaged” PMo–Ni and impregnated PMo–Ni properties are similar, and the number and distribution of BrZnsted and Lewis acid sites in these two catalysts are similar, too. It is concluded that both the aromatization of n-octane and the ability of producing light iso-alkanes are enhanced over the PMo and Ni ‘in situ encaged’ and impregnated catalysts than other catalysts. The effects of the acidity and temperature on the activity of n-octane hydroconversion over investigated catalysts are demonstrated. 䉷 2007 Elsevier Ltd. All rights reserved. Keywords: In situ encaging; 12-Molybdophosphoric acid; Nanocrystalline HZSM-5 zeolite; Hydroconversion; n-Octane; Aromatization; Ni

1. Introduction The hydroconversion catalytic processes to produce gasoline and diesel fuels are applied to a large range of petroleum fractions. The hydroconversion catalytic processes of long chain alkanes are of increasing importance for the petroleum industry due to the growing demand for light iso-alkanes and aromatics, which are required in reformulated gasoline as high octane number enhancement. Research octane number of n-octane is −17, while those of isopentane and aromatics are 92.3 and > 98, respectively. Therefore, the investigation of n-octane hydroconversion has attracted increasing interest (de Lucas et al., 2006; Yori et al., 2005; Chen et al., 2007). Heteropoly acids (HPA) have attracted much interest owing to their very strong acidity and redox properties. Several new industrial processes based on HPA catalysis have been

∗ Corresponding author.

E-mail address: [email protected] (X. Wang). 0009-2509/$ - see front matter 䉷 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.ces.2007.05.013

developed and commercialized in the last two decades (Kozhevnikov, 1998; Mizuno and Misono, 1998; Heravi et al., 2006; Misono, 2001). The Keggin type HPA are the most important in catalysis, PMo is the usual catalyst of choice because of higher oxidation potential compared with tungsten HPA. They are strong BrZnsted acid catalysts, and their acidity is stronger than that of conventional solid acids like zeolite and mixed oxides. The structure of a Keggin heteropolyanion is close to a sphere with a diameter of about 1 nm. If HPA could be formed like a “ship in a bottle” in a “cage” whose size is slightly larger than their anions, the formed HPA anions would not be able to diffuse out of the cage (Mukai et al., 2001). Recently, Sulikowski et al. (1996) reported that Keggin structured 12-tungstophosphoric acid (PW) could be encapsulated in the supercages of Y-type zeolite and this catalyst was active in isomerization and disproportionation of m-xylene in the gas phase. Mukai et al. (1997) have succeeded to immobilize PMo in a dealuminated Y-type zeolite crystal, through the esterification of acetic acid with ethanol, indicating its potential to become an alternative solid acid catalyst to conventional

4470

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

homogeneous catalysts. Kakuta and co-workers have successfully encapsulated the PMo and silicomolybdic acid species into the supercages of USY by using the hydrothermal technique Tran et al., 2005. Nowi´nska et al. (2003) reported that crystallization of mesoporous molecular sieves of SBA-3 structure from the gel containing HPA (pH < 2) resulted in the formation of new mixed mesostructures, which included the Keggin units in mesoporous material walls. Wang and co-workers conducted research on the hydroisomerization of n-heptane over hydrothermal treatment USY supported PW catalysts. They found that the PW-bearing catalysts showed much higher activity than the PW-free counterparts (Wang et al., 2004). Nanocrystalline zeolites are promising catalytic and adsorbent materials that have higher surface areas and reduced diffusion path lengths relative to conventional microcrystalline zeolite (Song et al., 2004). Nanocrystalline ZSM-5 zeolite exhibits higher activity, lower coke content and better stability as catalysts in the conversion of methanol to hydrocarbons (Sugimoto et al., 1987), in the oligomerization of ethylene (Yamamura et al., 1994), and in the FCC gasoline upgrading process (Zhang et al., 2004). They also exhibit increased selectivity of toluene conversion into cresol and decreasing coke formation relative to commercial ZSM-5 zeolite (Vogel et al., 2002). The size of the bulky anion exceeds the pore openings of medium-pore zeolites. Evidence is provided that the [PMo12 O40 ]3− Keggin unit of PMo with a diameter of 1 nm cannot enter the interior of ZSM-5 zeolite, the pore dimensions of which are 0.51 nm×0.56 nm and 0.53 nm×0.56 nm (Richter et al., 1991). Nanocrystalline HZSM-5 zeolite is thought to be an ideal material for the encapsulation of HPA molecules having a Keggin structure since they have the intergranular secondary pores. However, no data has been reported about using HPA/nanocrystalline HZSM-5 zeolite synthesized by in situ technique as a catalyst in the hydroconversion of alkanes. In this paper, for the first time, we investigated that PMo-Ni was in situ encaged in the secondary pores of nanocrystalline HZSM-5 zeolite and this catalyst was used for the hydroconversion of n-octane. The aromatization of n-octane and the ability of producing light iso-alkanes are enhanced over the PMo and Ni “in situ encaged” and impregnated in nanocrystalline HZSM-5 zeolite. The effects of the acidity and temperature on the activity of different catalysts in n-octane hydroconversion are investigated. 2. Experimental section 2.1. Catalyst preparation PMo was prepared according to a method previously described by Fournier (Rocchiccioli-Deltcheff et al., 1983) and the nanocrystalline HZSM-5 zeolite (the molar ratio of Si/Al was 14.2) was prepared according to the literature (Wang et al., 2000). The method reported by Mukai et al. (2001) is widely used for the in situ synthesis of PMo-Y. We modified this method to form PMo molecules in the secondary pores of nanocrystalline HZSM-5 zeolite. Nanocrystalline HZSM-5 of 10.0 g and MoO3

(99% purity) of 1.47 g were mixed in 70 g of deionized water. This mixture was stirred for 24 h at room temperature. Phosphoric acid (85% purity) of 0.099 g and nitric acid (65% purity) of 0.55 ml were added, and the obtained mixture was stirred at 85 ◦ C for 2 h. Then 0.99 g Ni(NO3 )2 · 6H2 O was added. The synthesized sample was dried at 320 ◦ C for 3 h. Then the catalyst IS-NiPMo-HZSM-5 is formed. For comparison, nanocrystalline HZSM-5 supported PMo and Ni was also prepared by the usual impregnation technique. The mixture of 1.70 g PMo and 0.99 g Ni(NO3 )2 · 6H2 O was dissolved in 20 ml deionized water. HZSM-5 zeolite of 10.0 g were added into the solution and steeped the mixture for 24 h. The slurry mixture was calcined at 320 ◦ C for 3 h. This catalyst is denoted by NiPMo–HZSM-5. The nanocrystalline HZSM-5 supported Ni (NiHZSM-5) was prepared by the usual impregnation technique. 2.2. Characterization FT-IR spectra of samples pressed with dried KBr into discs were recorded in the range of wave numbers 4000 − −400 cm−1 with a resolution of 4 cm−1 by using of an I Equinox55 Infrared spectrometer. X-ray diffraction patterns were taken on a D/max-2400 diffractometer through utilizing the Cu K radiation at 40 kV and 30 mA with a scanning rate of 2◦ / min. The ratio of silicon to aluminum in the zeolites was obtained on a Bruker SRS 3400 spectrometer. The bulk elemental composition of catalysts was examined by inductively coupled plasma (ICP) by using a Optima 2000DV. The 31 P MAS-NMR experiments were performed on a Varian Infinity-plus 400 spectrometer operating at a magnetic field strength of 9.4 T. The resonance frequencies at this field strength were 161.9 MHz for 31 P. A chemagnetics 5 mm tripleresonance MAS probe was employed to acquire all the spectra (with a spinning rate of 7 kHz). Phosphoric acid 85% was employed as external reference. The fresh samples (ca. 0.015 g) were transferred into ESR tubes. The catalysts were heated in situ of 340 ◦ C for 17 h in flowing hydrogen (30 ml min−1 ). The ESR spectra were recorded on a Bruker ESR 300 spectrometer. The spectra were recorded at liquid nitrogen. The g values of the paramagnetic species were determined by the use of Mn2+ dissolved in magnesium oxide. Acidic properties of catalysts were tested by the temperatureprogrammed desorption of ammonia (NH3 -TPD) by using a Quantachrome Chembet-3000 apparatus. Catalyst of 0.20 g was charged in the quartz tube, heated in situ of 320 ◦ C for 1 h in He flow, and cooled to 120 ◦ C for the saturation of the catalyst. NH3 -TPD profiles were obtained in the temperature range of 120–800 ◦ C with a constant heating rate of 10 ◦ C/ min. The samples of the IR spectra of chemisorbed pyridine (Py-IR) were pressed into self-supporting wafers. The wafers (0.010 g) were mounted in an infrared vacuum cell and degassed at 320 ◦ C for 4 h under vacuum ( 10−3 Pa), then cooled to 50 ◦ C and exposed to pyridine vapor for 0.5 h. The desorption was performed at 150, 250 and 350 ◦ C under vacuum. Pore volume and surface area were obtained on a Quantachrome Autosorb-1 using the BET method. Adsorption

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

4471

capacities for normal hexane (n–C6 ) and cyclohexane (c–C6 ) were also measured by the flow adsorption method. The adsorption capacity for 2,2,3-trimethylbutane (2,2,3-TMB) was used to characterize the secondary pores contained in the polycrystalline grains of nanoparticle ZSM-5 zeolite. The crystal size of HZSM-5 was determined using a JEM1200CEX transmission electron microscope (TEM) and JEOL JSM-6700F field emission scanning electron microscope (SEM). 2.3. Catalytic measurements The reaction was carried out in a continuous flow fixedbed reactor (diameter 10 mm). Reaction conditions were listed as the following: temperature 280–340 ◦ C, pressure 1.5 MPa, H2 /n-octane (mol ratio) 2.0, weight space velocity of n-octane 2.5 h−1 , the catalyst weight 1.68 g. The composition of the products was analyzed by a GC-8820 gas chromatograph equipped with FID and an OV-101 capillary column (50.0 mm × 0.25 mm). Before each catalytic measurement, the catalysts were reduced in flowing hydrogen (28 ml/min) at 1.5 MPa pressure and at the reaction temperature for 2.5 h. Catalytic performance was expressed using conversion of n-octane, aromatization ratio and i/n ratio of products, which were defined as follows: Conversion of n-octane (wt%) = (X0 − X1 ) × 100/X0 .

(1)

X0 and X1 are the weight percentage content of n-octane before and after the reaction. Aromatization ratio = [Xa /(X0 − X1 )] × 100.

(2)

Xa total concentration of aromatics (wt%), X0 and X1 are same (1). i/n = i(C5 + C6 + C7 )/n(C5 + C6 + C7 ).

(3)

i is total weight percentage content of C5 , C6 and C7 isoalkanes, n is total weight percentage content of C5 , C6 and C7 n-alkanes. 3. Results and discussion 3.1. Characterization of catalysts The TEM and SEM images of the nanoscale HZSM-5 zeolite are shown in Fig. 1. The crystal size of HZSM-5 zeolite is about 50 nm and its crystal size with the calculated particle size by XRD is shown in Table 1. Table 1 lists the results of pore volume and surface area of both nanocrystalline and microcrystalline ZSM-5. It can be seen that the total surface area of the calcined ZSM-5 samples increases as particle size decreases. Total surface areas of both calcined and as-synthesized samples were obtained from the nitrogen adsorption isotherms using the BET method (Song et al., 2004). The total surface area of the as-synthesized samples in which the internal surface is blocked by template molecules represents the external surface area of the zeolite

Fig. 1. TEM (1a) and SEM (1b) pictures of HZSM-5 zeolite.

sample (Camblor et al., 1998). The total surface area obtained from the calcined samples should contain contributions from both the internal and the external surfaces. The internal surface area obtained by taking the difference in total surface area of the calcined and the as-synthesized ZSM-5 samples is constant with a value of approximately 355 m2 /g (Song et al., 2004). The external surface area of nanocrystalline HZSM-5 approximately is 46 m2 /g. Only the total surface area of 366 m2 /g was obtained for 10 m samples. The adsorption characteristics features of the nanocrystalline and microcrystalline ZSM-5 zeolite are also shown in Table 1. There are significant differences in the physicochemical properties between these two kinds of crystals. Compared to the microcrystalline ZSM-5, the nanocrystalline ZSM-5 zeolite have higher n-C6 and c-C6 adsorption capacity. Moreover, its outstanding adsorption capacity for 2,2,3-TMB exhibits the existence of secondary pores whose openings are wide enough to allow free entry of relatively large molecules. The nanocrystal ZSM-5 with morphology of polycrystalline grains generates the secondary pores from inter crystal gaps in nanocrystalline ZSM-5. Elemental analysis for the IS-NiPMo–HZSM-5, NiPMo– HZSM-5, and NiHZSM-5 catalysts are shown in Table 2. It can be seen that the molybdenum loadings have different effects on the IS-NiPMo–HZSM-5 and NiPMo–HZSM-5, suggesting that the in situ encaging technique is better method compared with impregnation technique. Determined molar ratio for P:Mo (close to 1:12) also confirmed retention of the Keggin structure of the PMo in the nanocrystalline HZSM-5 zeolite. The X-ray diffraction of the pure PMo and different catalysts are revealed in Fig. 2. The pattern of the unpretreated PMo (Fig. 2, line 6) is typical for HPA (Damyanova et al., 2000; Rocchiccioli-Deltcheff et al., 1996). Upon pretreatment at 320 ◦ C, this range corresponds to the loss of crystallization water, and it is related to the formation of the anhydrous acid phase. The structure of the nanocrystalline HZSM-5 zeolite does not collapse upon loading with the strong acidic HPA. Apparent decrease of the lines intensity is due to the lowered amount of zeolite in the samples upon loading of Ni and

4472

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

Table 1 Characteristic of nanocrystalline and microcrystalline ZSM-5 zeolite Particle size

70–100 nm 10 m

BET area (as synthesized)a

46 N/A

BET area (calcined)b

401 366

Particle size (calculated)b

18 nm N/A

Differencec

355 N/A

Pore volume (cm3 /g)

0.73 0.61

Adsorption capacity (wt%) ( Wang et al., 1995) n − −C6

c − −C6

2, 2, 3-TMB

12.7 11.8

7.7 6.1

64 1.8

surface areas are provided in m2 /g. particle size of HZSM-5 measured by XRD (the Scherrer equation). c Difference = BET area (calcined)-BET area (as synthesized). a All

b The

Table 2 Elemental analysis (ICP) for the catalysts Catalyst

Al

Si

Mo

Ni

P

Ni/P/Mo (mol ratio)

IS-NiPMo–HZSM-5 NiPMo–HZSM-5 NiHZSM-5

2.86 2.82 3.24

44.12 44.31 49.36

6.19 4.99 N/A

1.07 0.94 1.35

0.18 0.15 N/A

0.30/0.09/1 0.32/0.093/1 N/A

1-HZSM-5 2-NiHZSM-5 3-NiPMo-HZSM-5 4-IS-NiPMo-HZSM-5 5-PMo(320) 6-PMo(25) 10

20

30 2 Theta (deg.)

40

50

Fig. 2. XRD spectra of the pure PMo and different catalysts.

PMo. For IS-NiPMo-HZSM-5 or NiPMo-HZSM-5 catalysts, no peak of the additional crystal structures (PMo or Ni) is found, so does NiHZSM-5. This phenomenon indicates that PMo is highly dispersed in the nanocrystalline HZSM-5 zeolite, as well as Ni. Fig. 3 shows the FT-IR spectra of the different catalysts samples. The main characteristic features of the Keggin structure are observed at 1063 cm−1 (as P–Oa), 962 cm−1 (as Mo–Od), 867 cm−1 (as Mo–Ob–Mo) and 790 cm−1 (as Mo–Oc–Mo) (Damyanova et al., 2000; Rocchiccioli-Deltcheff et al., 1996). Upon pretreatment at 320 ◦ C, the characteristic bands of PMo are still detected, but the intensities are somewhat changed. The characteristic peaks of PMo were detected both on the IS-NiPMo-HZSM-5 catalyst, although their intensities are somewhat low. As one of the main peaks that appeared at 1065 and 794 cm−1 was overlapped with the peak of framework of nanocrystalline HZSM-5 zeolite, and the spectrum of

IS-NiPMo-HZSM-5 catalyst was attempted (Fig. 3b, line 1). It shows that the Keggin structure is essentially preserved on the support after in situ encaging. The IR characteristic bands of the IS-NiPMo–HZSM-5 have some shifts compared with those of the PMo. It indicates that a strong chemical interaction occurs between the Mo–Od of PMo and the Al–OH/Si–OH groups of HZSM-5. The interaction is similar to that between PW and SiO2 reported by Hu’s group (Guo et al., 2000). The band at 962 cm−1 (Mo–Od) is almost completely disappeared and the intensity of the band at 867 cm−1 (Mo–Ob–Mo) is strongly decreased. This fact proves that PMo has been decomposed a little due to the deeper interaction with basic extra-framework Al–O species. The 31 P chemical shift () provides important information concerning the structure, composition and electronic states of HPA. Fig. 4 shows 31 P MAS-NMR spectra of samples. The chemical shift for 31 P at −3.47 ppm for PMo12 O3− 40 was correlated with the P–Oa bond strength [ (P–Oa)] (Concellón et al., 1998). The 31 P chemical shift of the PMo heteropolyanion indicates that the four groups of three edge-shared MoO6 octahedra (Mo3 O13 triad) protected the central atom. The spectra showed that IS-NiPMo-HZSM-5 sample presents only one line with maxima at −3.63 ppm (Fig. 4, line 2). This suggests that the structure of the encaged PMo keeps intact undegraded Keggin structure. The slight shift toward higher field can be caused by an increase of the interaction of the acid with the support during calcination, as it was the case for another PW on SiO2 (Levebyre, 1992; Joshi and Mukesh, 1997). The proton transfer according to the following equation may form the interacting species: + − H3 PMo12 O+ 40 ≡ M–OH → (≡ M–OH2 ) (H2 PMo12 O40 )

M = Si–OH or Al–OH.

(4)

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

a

1PMo

b

1063 cm-1 962 cm-1 867 cm-1 790 cm-1

4473

867 cm-1 2 IS-NiPMo-HZSM-5 962 cm-1 790 cm-1 1063 cm-1 3 NiPMo-HZSM-5

2 IS-NiPMo-HZSM-5 3 NiPMo-HZSM-5 4 NiHZSM-5 5 HZSM-5 1350

1200

1050 900 750 Wave number (cm-1)

600

2000

1600 1200 800 Wave number (cm-1)

400

Fig. 3. The FTIR spectra of the catalyst samples.

Fig. 4. 31 P MAS-NMR spectra of the catalyst samples.

The chemical shifts by 31 P NMR of NiPMo–HZSM-5 catalyst were observed with resolved signals at −3.50 and −1.02 ppm. The first one was attributed to finely dispersed PMo molecules, retaining the Keggin structure, and the second to a lacunary or unsaturated species, such as PMo11 and P2 Mo17 or P2 Mo12 and P2 Mo21 , which have similar lines in 31 P NMR solution spectra (Joshi and Mukesh, 1997). The following facts convincingly prove that PMo cannot penetrate the interior pore of the zeolite, but becomes encaged in the intergranular secondary pores of nanocrystalline HZSM-5 zeolite. (I) The inner pore dimension of the HZSM-5 (∼ 0.55 nm, Richter et al., 1991) is less than the size of PMo molecules (1.0 nm Mukai et al., 2001). (II) The 31 P NMR and FT-IR spectra of IS-NiPMo-HZSM-5 catalyst showed that the Keggin structure is essentially preserved on the support after in situ

Fig. 5. ESR spectra of the catalyst samples: (a) IS-NiPMo-HZSM-5, under H2 reduce at 340 ◦ C for 17 h; (b) IS-NiPMo–HZSM-5 (Mn2+ in MgO (powder)); (c) NiPMo–HZSM-5, under H2 reduce at 340 ◦ C for 17 h; (d) NiHZSM-5, under H2 reduce at 340 ◦ C for 17 h.

encaging, these facts prove that PMo can be formed in the intergranular secondary pores and avoid coordinating with basic extra-framework Al–O species. The commercial ZSM-5 zeolite is not suitable for the support of HPA because they do not have intergranular secondary pores. (III) The XRD of the samples indicates that PMo is highly dispersed into the nanocrystalline HZSM-5 zeolite. The ESR spectra of catalysts in Fig. 5 are obtained after reduction under H2 at 340 ◦ C. It can be seen from the figure, the spectra of NiHZSM-5 catalyst are composed of two signals: the basic field signal leads to the formation of Ni0 , which the Ni2+ is reduced. The few Ni2+ signals show that Ni2+

4474

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

Scheme 1. Reduction of a polyoxometalate represented schematically by Mo(O)–O–Mo(O)—with hydrogen.

75

IS-NiPMo-HZSM-5 NiPMo-HZSM-5 NiHZSM-5 HZSM-5

Signal (mV)

60

45

30

15

0 200

300

400 Temperature (oC)

500

600

Fig. 6. The NH3 –TPD profiles of the different catalysts.

cannot completely reduce in our experiment. Apparently, the spectra of IS-NiPMo–HZSM-5 and NiPMo–HZSM-5 catalysts are composed of two signals: a normal magnetic field signal leads to the formation of Mo5+ which was also confirmed by ESR spectroscopy. The g ⊥ and g values of the ESR spectra are 1.937 and 1.992, respectively, which matches the literature data (Kogan et al., 1999). Mo6+ is diamagnetic and has no ESR signals. The second signals at g = 2.25 leads to the formation of Ni2+ which was also confirmed by ESR spectroscopy, the Ni0 signal cannot be observed at signals after reduction under H2 at 340 ◦ C for 17 h. When the reduced time is increased to 56 h, a decrease in the intensity of the Mo5+ signal is observed, however, the Ni0 signal cannot be observed just the same. The Mo6+ is reducing in preference to Ni2+ since their oxidation–reduction potential was different.Importantly, in this reaction mode the substrate may also be hydrogen. Thus, the reduction of polyoxometalates by hydrogen (Scheme 1) has been reported by Kogan et al. (1999). The acid properties of different catalysts were examined by the NH3 -TPD and Py-IR. Fig. 6 shows the NH3 -TPD of different catalysts. In the spectra of IS-NiPMo–HZSM-5 catalyst, desorption peaks appeared at 260 and 480 ◦ C, in addition to the sharp peaks at 590 ◦ C. Okumura et al. (2005) considered that the sharp peak emerging at 590 ◦ C has possibly originated from the fragment of CO2 , not from NH3 . In contrast, the species

of desorption at the broad peak at 590 ◦ C may be attributed to the fragment of CO2 . Compared with other catalysts, the concentration of weak acid sites for the IS-NiPMo–HZSM-5 catalyst has changed little, however, the concentration of strong acid sites increased. The change of concentration of acid sites could be attributed to the proton transfer between PMo and Si–OH/Al–OH of nanocrystalline HZSM-5 zeolite. When the proton transfer happens, the protonation and dehydration as reported by Kozhevnikov et al. (1996) may be formed (Mastikhin et al., 1990). It also indicates that the state between PMo and the Al–OH/Si–OH groups of nanocrystalline HZSM-5 zeolite was complex, which was confirmed by the FT-IR and 31 P MASNMR spectra. Py-IR has shown that pyridine molecule can serve to determine the concentration of BrZnsted and Lewis acid sites on solid catalysts. Table 3 displays the total acid and acidic type distribution of catalysts. The NiHZSM-5 catalyst exhibits fewer BrZnsted and more Lewis acid sites than HZSM-5. The change of Lewis acid sites may be caused by ion-exchange of HZSM-5 with Ni, or by blocking of Ni particles on pores during catalyst pretreatment. Although PMo is BrZnsted acid, PMo and Ni loaded HZSM-5 nanocrystalline zeolite has more Lewis acid sites and less BrZnsted acid sites in comparison with the original HZSM-5 zeolite. The dealumination of HZSM-5 zeolite would proceed during loading of HPA, leading to formation of extra-framework octahedral aluminum. Several researchers, such as Katada et al. (2004), reported this henomenon. 3.2. Activity of the catalysts in hydroconversion of n-octane The hydroconversion of n-octane is carried out to test the activity of the catalysts. Fig. 7 shows the effect of temperature on conversion of n-octane. The introduction of PMo and Ni improves the n-octane conversion of nanocrystalline HZSM-5 zeolite. Especially at 280 ◦ C, the conversion of n-octane over the introduction PMo and Ni is higher than 90.5%, while the conversion of n-octane for the HZSM-5 is 47.1%. The effect of the temperature on the n-octane conversion for the tested catalysts follows the sequence: HZSM-5 > NiHZSM-5 > NiPMo-HZSM-5 ≈ IS-NiPMo-HZSM-5. Fig. 8 shows the effect of reaction temperature on the aromatization ratio. The aromatics are thermodynamically favored over the corresponding alkanes (Smiešková et al., 2004). As can be seen from the figure, the reaction temperature affects the aromatization ratio. When the reaction temperature increases

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

4475

Table 3 The acid type distribution of different catalysts 150–350 ◦ C

Catalyst

IS-NiPMo–HZSM-5 NiPMo–HZSM-5 NiHZSM-5 HZSM-5

 350 ◦ C

Total

B

L

L/(B + L)

B

L

L/(B + L)

B

L

L/(B + L)

0.73 0.66 0.67 0.99

0.42 0.39 0.48 0.31

0.37 0.37 0.42 0.24

0.56 0.57 0.37 0.40

0.32 0.27 0.45 0.13

0.36 0.32 0.55 0.25

1.29 1.23 1.04 1.39

0.74 0.66 0.93 0.44

0.36 0.35 0.47 0.24

The all surface areas are provided in m2 /g, L /B = 0.084 cm−2 mmol g−1 /0.059 cm−2 mmol g−1 (Datka, 1981).

Conversion of n-octane (wt%)

100

80

IS-NiPMo-HZSM-5 NiPMo-HZSM-5 NiHZSM-5 HZSM-5

60

40

280

290

300

310

320

330

340

Reaction temperature (oC) Fig. 7. Effect of temperature on conversion of n-octane.

25

Aromatization ratio

20

IS-NiPMo-HZSM-5 NiPMo-HZSM-5 NiHZSM-5 HZSM-5

15

10

5

0 280

290

300 310 320 o Reaction temperature ( C)

330

340

Fig. 8. Effect of temperature on aromatization ratio of n-octane.

from 280 to 340 ◦ C, the increases of the aromatization ratio of IS-NiPMo–HZSM-5, NiPMo–HZSM-5, NiHZSM-5 and HZSM-5 catalysts are 21.7, 20.4, 17.6 and 11.9 wt%,

respectively. The IS-NiPMo–HZSM-5 and NiPMo-HZSM-5 catalysts show the higher ability of aromatization than the other catalysts. The results indicate that the introduction of PMo and Ni in the nanocrystalline HZSM-5 zeolite can improve the ability aromatization of n-octane over HZSM-5. Table 4 manifests the distribution of products from the n-octane hydroconversion. It can be seen from the table that the main aromatization products over these catalysts are C7 , C8 and C9 aromatics. They account for about 85% of the total of aromatics over these catalysts. The other aromatic products are mainly benzene and C9+ . These aromatics contribute the main high octane number to gasoline in China. They are ideal at tempering components for gasoline if they are controlled to certain content in gasoline. Fig. 9 shows the influence of temperature on iso-to-n (i/n) ratio of the products. I/n ratio are the parameter to expound the ability of catalyst to produce light iso-alkanes. It can be seen from the figure that the reaction temperature affects i/n ratio remarkably. At 280–340 ◦ C, i/n ratio for these catalysts increase as the temperature rises. The transformation of hydrocarbons over acid catalysts such as HZSM-5 zeolite proceeds through carbonium/carbenium ion chemistry (Kazansky, 1999). Since thermodynamic data indicate that branched isomers are favored at low temperature it is recommended that the isomerization of linear carbonium ion occurs at low temperature. However, the reaction mechanism involves endothermic steps, and from the point of view of kinetics, high reaction temperature is required. It indicates the kinetics factors dominate the isomerization. At 340 ◦ C, i/n ratio for IS-NiPMo–HZSM5, NiPMo–HZSM-5, NiHZSM-5 and HZSM-5 are 2.88, 2.56, 2.39 and 1.72, respectively. The introduction of PMo and Ni in the nanocrystalline HZSM-5 zeolite exhibits better ability of producing light iso-alkanes than any other catalyst. To the best of our knowledge, the method is used for the in situ synthesis of HPA/nanocrystalline HZSM-5 zeolite catalyst is unique. It is generally proposed that on acid catalysts the aromatization of alkanes occurs through protolysis of alkane, cracking of carbonium ion to alkane and alkene, oligomerization of alkenes, cyclization of oligomerized products and formation of aromatics from cyclic rings by hydrogen transfer (Smiešková et al., 2004). The mechanism of aromatics formation suggested by Guisnet et al. (1992) in the reaction is mainly

4476

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

Table 4 The distribution of products from hydroconversion of n-octane Catalyst Conversion (wt%)

Components

IS-NiPMo–HZSM-5 96.33

NiPMo–HZSM-5 96.56

NiHZSM-5 96.72

HZSM-5 95.94

Alkanes

C1 − −C3 iso − −C4 n − −C4 iso − −C5 n − −C5 iso − −C6 n − −C6 iso − −C7 n − −C7 iso − −C8 C8+

3.02 10.61 19.77 16.52 7.57 8.28 1.54 1.90 0.15 1.90 0.18 2.16 0.29 2.36 8.20 8.57 2.55 0.76 23.29 2.88

3.40 8.81 20.25 17.04 8.98 9.11 1.77 1.76 0.14 1.18 0.12 1.95 0.37 2.25 7.94 8.38 2.30 0.81 22.45 2.56

3.27 7.90 20.64 17.51 10.27 8.49 1.69 3.03 0.18 2.08 0.62 1.93 0.41 1.16 5.62 7.17 3.69 1.06 19.34 2.39

3.62 11.37 20.17 15.80 12.30 7.73 2.51 3.74 0.42 1.69 0.54 1.65 0.64 0.54 3.23 5.23 3.49 1.27 14.34 1.79

Naphthenes Alkenes Aromatics

C6 C7 C8 C9 C9+

Aromatization ratio i/n ratio

FH (hydrogen flow rate) = 28.5 cm3 min−1 ; FRH (n-octane flow rate) = 0.10 cm3 min−1 ; catalyst 1.68 g; system pressure 1.5 MPa; pre-heating temperature of catalyst 340 ◦ C; reaction temperature 340 ◦ C; TOS 10 h.

3.0 IS-NiPMo-HZSM-5 NiPMo-HZSM-5 NiHZSM-5 HZSM-5

i/n ratio

2.5

2.0

1.5

1.0

0.5 280

290

300 310 320 Reaction temperature (oC)

330

340

Fig. 9. Effect of temperature on i/n ratio on products.

as following:

Consulting the mechanism, the aromatics ratio and light iso-alkanes depends on the concentration of the intermediates, which relies on the concentration of acid and the cooperation of the Lewis acid sites (Ni as the acid center) and the BrZnsted acid sites. At low L/(B + L) ratio for the HZSM-5 catalysts, the originally formed carbonium ions species are insufficiently transform from Lewis to BrZnsted acid sites. Although the L/(B + L) ratio of NiHZSM-5 catalyst is higher than those of other catalysts, a decrease of BrZnsted acid sites diminishes the probability of forming alkyl, alkenyl and naphthenic carbonium ions intermediate species. The IS-NiPMo–HZSM-5 and NiPMo–HZSM-5 catalysts have higher L/(B + L) ratio and more concentration Lewis acid sites. On these two catalysts, n-octane reaction pathways include the dehydrogenation of C8 H18 to form C8 H16 on Lewis acid sites and the cracking of C8 H18 to form C4 H10 and C4 H8 on BrZnsted acid sites. Secondary products include alkanes, alkenes, naphthenes, etc. The concentration of various intermediates reaches the maximum by means of the synergistic effect of BrZnsted and Lewis acid sites. The aromatization ratio may increase with increasing carbonium ions intermediates, and so does the probability H+

L acid or Ni C8H18 C8H18+

H+

H+

H+ C8H16

H+

C3-C5 Alkenes L acid or Ni

C6-C10 Naphthenes

C6-H10 Aromatics

C6-C10 Alkenes

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

of carbonium ions isomerization. Hence, i/n ratios increase accordingly. C6 – C9 aromatics are formed via dehydrocycliza+ tion reactions of C+ 6 – C9 alkenes on Lewis acid sites, which are formed in rapid BrZnsted acid-catalyzed oligomerizationcracking cycles. The catalytic difference between the encaged PMo–Ni and the impregnated PMo–Ni is little since their acid amount and an L/(B + L) ratio does not have much difference. We can get the conclusion that the improved activity of n-octane hydroconversion over the IS-NiPMo–HZSM-5 and NiPMo–HZSM-5 catalysts can be attributed to the increase of acidic concentration and the synergistic effect of BrZnsted and Lewis acid sites. 4. Conclusion Encagement of molybdenum species in the secondary pore of nanocrystalline HZSM-5 zeolite was accomplished by using the in situ synthesizing technique. PMo cannot penetrate the pore system of the HZSM-5 zeolite, but becomes encaged in the secondary pores of nanocrystalline HZSM-5 zeolite. The improved activity of n-octane hydroconversion over PMo and Ni in situ encaged and impregnated in the secondary pores of nanocrystalline HZSM-5 zeolite can be attributed to the increase of acidic concentration and the synergistic effect of BrZnsted and Lewis acid sites. Acknowledgment This project was supported by China Petroleum and Chemical Corporation (104002) and by Program for New Century Excellent Talents in University (NCET-04-0268). References Camblor, M.A., et al., 1998. Characterization of nanocrystalline zeolite Beta. Microporous and Mesoporous Materials 25, 57–74. Chen, L.D., et al., 2007. Hydroconversion of n-octane over nanoscale HZSM-5 zeolite promoted by 12-molybdophosphoric acid and Ni. Catalysis Communications 8, 423–426. Concellón, A., et al., 1998. Molybdophosphoric acid adsorption on titania from ethanol–water solutions. Journal of Colloid and Interface Science 204, 256–267. Damyanova, S., et al., 2000. Thermal behavior of 12-molybdophosphoric acid supported on zirconium-loaded silica. Chemistry of Materials 12, 501–510. Datka, J., 1981. Dehydroxylation of NaHY zeolite studied by infrared spectroscopy. Journal of Chemical Society Faraday Transactions 77, 2877–2881. de Lucas, A., et al., 2006. Kinetic model of the n-octane hydroisomerization on PtBeta agglomerated catalyst: influence of the reaction conditions. Industrial & Engineering Chemistry Research 45, 978–985. Guisnet, M., et al., 1992. Aromatization on short chain alkanes on zeolite catalysts. Applied Catalysis A: General 89, 1–30. Guo, Y.H., et al., 2000. Microporous polyoxometalates POMs/SiO2 : synthesis and photocatalytic degradation of aqueous organocholorine pesticides. Chemistry of Materials 12, 3501–3508. Heravi, M.M., et al., 2006. 12-Molybdophosphoric acid: a recyclable catalyst for the synthesis of Biginelli-type 3,4-dihydropyrimidine-2(1H)-ones. Catalysis Communications 7, 373–376. Joshi, M.V., Mukesh, D., 1997. Transfer hydrogenation of nitro compounds with heteropoly acid catalysts. Journal of Catalysis 168, 273–277.

4477

Katada, N., et al., 2004. Dealumination of proton form mordenite with high aluminum content in atmosphere. Microporous and Mesoporous Materials 75, 61–67. Kazansky, V.B., 1999. Adsorbed carbocations as transition states in heterogeneous acid catalyzed transformations of hydrocarbons. Catalysis Today 51, 419–434. Kogan, V., et al., 1999. Polyoxometalates as reduction catalysts: deoxygenation and hydrogenation of carbonyl compounds. Angewandte Chemie International Edition 38, 3831–3834. Kozhevnikov, I.V., 1998. Catalysis by heteropoly acids and multicomponent polyoxometalates in liquid-phase reactions. Chemical Reviews 98, 171–198. Kozhevnikov, I.V., et al., 1996. Study of catalysts comprising heteropoly acid H3 PW12 O40 supported on MCM-41 molecular sieve and amorphous silica. Journal of Molecular Catalysis A: Chemical 114, 287–298. Levebyre, F., 1992. 31 P MAS NMR study of H3 PW12 O40 supported on − silica: formation of (SiOH+ 2 )(H2 PW12 O40 ). Journal of Chemical Society Chemical Communications 10, 756–757. Mastikhin, V.M., et al., 1990. 1 H and 31 P MAS NMR studies of solid heteropolyacids and H3 PW12 O40 supported on SiO2 . Journal of Molecular Catalysis 60, 65–67. Misono, M., 2001. Unique acid catalysis of heteropoly compounds (heteropolyoxometalates) in the solid state. Chemical Communications 13, 1141–1152. Mizuno, N., Misono, M., 1998. Heterogeneous catalysis. Chemical Reviews 98, 199–218. Mukai, S.R., et al., 1997. Preparation of encaged heteropoly acid catalyst by synthesizing 12-molybdophosphoric acid in the supercages of Y-type zeolite. Applied Catalysis A: General 165, 219–226. Mukai, S.R., et al., 2001. Key factors for the encapsulation of Keggin-type heteropoly acids in the supercages of Y-type zeolite. Chemical Engineering Science 56, 799–804. Nowi´nska, K., et al., 2003. Heteropoly compounds incorporated into mesoporous material structure. Applied Catalysis A: General 256, 115–123. Okumura, K., et al., 2005. The active and reusable catalysts in the benzylation of anisole derived from a heteropoly acid. Journal of Catalysis 234, 300–307. Richter, M., et al., 1991. Specific modification of the external surface of ZSM-5 zeolite by 12-tungsilicic acid. Journal of Chemical Society Faraday Transactions 87, 1461–1466. Rocchiccioli-Deltcheff, C., et al., 1983. Vibrational investigations of polyoxometalates 2 Evidence for anion–anion interactions in molybdenum (VI) and tungsten (VI) compounds related to the Keggin structure. Inorganic Chemistry 22, 207–216. Rocchiccioli-Deltcheff, C., et al., 1996. Catalysis by 12-molybdophosphates 1: catalytic reactivity of 12-molybdophosphoric acid related to its thermal behavior investigated through IR, Raman, polarographic and X-ray diffraction studies: a comparison with 12-molybdosilicic acid. Journal of Catalysis 164, 16–27. Smiešková, A., et al., 2004. Aromatization of light alkanes over ZSM-5 catalysts. Influence of the particle properties of the zeolite. Applied Catalysis A: General 268, 235–240. Song, W., et al., 2004. Synthesis characterization and adsorption properties of nanocrystalline ZSM-5. Langmuir 20, 8301–8306. Sugimoto, M., et al., 1987. Correlation between the crystal size and catalytic properties of ZSM-5 zeolite. Zeolites 7, 503–507. Sulikowski, B., et al., 1996. Novel “ship-in-the-bottle” a type catalyst: evidence for encapsulation of 12-tungstophosphoric acid in the supercage of synthetic faujasite. Catalysis Letters 39, 27–31. Tran, M.H., et al., 2005. Hydrothermal synthesis of molybdenum oxide catalyst: heteropoly acids encaged in US-Y. Applied Catalysis A: General 287, 129–134. Vogel, B., et al., 2002. The synthesis of cresol from toluene and N2 O on H[Al]ZSM-5: minimizing the product diffusion limitation by the use of small crystals. Catalysis Letters 79, 107–112. Wang, X., et al., 1995. In: Proceedings of International Symposium on Zeolite in China, Nanjing, 1995, p. 148.

4478

L. Chen et al. / Chemical Engineering Science 62 (2007) 4469 – 4478

Wang, X. et al., 2000. Ultrafine particulate 5-membered ring zeolite. C.N. Patent, 1,240,193. Wang, J., et al., 2004. Hydroisomerization of n-heptane over modified USYsupported H3 PW12 O40 catalysts: effect of hydrothermal treatment for USY. Journal of Industrial and Engineering Chemistry 10, 454–459. Yamamura, M., et al., 1994. Synthesis of ZSM-5 zeolite with small crystal size and its catalytic performance for ethylene oligomerization. Zeolites 14, 643–649.

Yori, J.C., et al., 2005. Hydroisomerization-cracking of n-octane on heteropolyacid H3 PW12 O40 supported on ZrO2 , SiO2 and carbon effect of Pt incorporation on catalyst performance. Applied Catalysis A: General 286, 71–78. Zhang, P.Q., et al., 2004. Characterization of modified nanoscale ZSM-5 zeolite and its application in the olefins reduction in FCC gasoline. Catalysis Letters 92, 63–68.