Você está na página 1de 9

Available online at www.sciencedirect.

com

Microporous and Mesoporous Materials 109 (2008) 485493 www.elsevier.com/locate/micromeso

Eects of niobium addition on ZSM-5 studied by thermal and spectroscopy methods


Ivoneide C.L. Barros a, Valdeilson S. Braga b, Daniel S. Pinto b, Julio L. de Macedo b, A. Dias b, S lvia C.L. Dias b,* Geraldo N.R. Filho c, Jose
a

ncias Exatas, Universidade Federal do Amazonas, Av. Gen. Rodrigo Octa vio Jorda Instituto de Cie o Ramos, 3000, 69077-000 Manaus, AM, Brazil b rio de Cata lise, Instituto de Qu mica, Universidade de Bras lia, Caixa Postal 4478, Bras lia, DF 70904-970, Brazil Laborato c ncias Exatas e Naturais, Universidade Federal do Para , Rua Augusto Correa 1, 66075-010 Bele m, PA, Brazil Centro de Cie Received 2 January 2007; received in revised form 25 May 2007; accepted 29 May 2007 Available online 12 June 2007

Abstract Incorporation of niobium into microporous molecular sieves brings up new properties for these materials. ZSM-5 supported catalysts containing 2, 5, 13 and 19 wt.% of niobium pentoxide were prepared by aqueous impregnation and characterized by the combination of thermal (TG/DTA) and spectroscopic methods such as FTIR, FT-Raman, DRIFTS, and 29Si and 27Al MAS and 29Si CP-MAS-NMR. In addition, X-ray diraction (XRD), inductively coupled plasma atomic emission spectroscopy (ICP-AES) and measurements of specic surface area by nitrogen adsorption (BET method) were carried out in order to obtain structural, textural and stability information about the solids. The samples were studied at dierent temperatures and heating time to optimize these parameters and to avoid a possible reduction of acidic properties by dehydroxylation. MAS-NMR and XRD results showed that the zeolite structure containing Nb did not undergo dealumination after thermal treatment, according to the presence of SiAOANb units on the zeolites structure in FTIR and Raman studies. Thermal analysis conrmed that Nb2O5 impregnation reduced zeolite ZSM-5 dehydroxylation and thus promoted thermal stabilization. Nitrogen adsorption isotherms by the BET method indicated that surface area and pore volume decreased with the increase of Nb loading on the calcined samples at 450 C. In addition, the nature of the zeolite active sites was investigated by spectroscopic and thermal analysis after pyridine adsorption. It was proposed that there are two dierent acidic sites (Brnsted and hydrogen bond site types) on the studied systems. 2007 Elsevier Inc. All rights reserved.
Keywords: ZSM-5 zeolite; Niobium pentoxide; NbZSM5; Thermal analysis; Acidity

1. Introduction Solid acid catalysts are the basis of several rening and petrochemical processes [13]. Niobium pentoxide is an interesting catalyst because of its high acidity and watertolerant properties. Processes in which niobium oxidebased materials are applied include: dehydration of alcohols, esterication, FischerTropsch synthesis, etc. [4].
Corresponding author. Tel.: +55 61 3307 2162; fax: +55 61 3368 6901. E-mail addresses: ibarros@ufam.edu.br (I.C.L. Barros), narciso@ ufpa.br (G.N.R. Filho), scdias@unb.br (S.C.L. Dias). URL: www.unb.br/iq/labpesq/qi/labcatalise.htm (S.C.L. Dias). 1387-1811/$ - see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.micromeso.2007.05.050
*

The preparation of solid materials containing niobium leads to bifunctional catalysts that can be used on processes of acid catalysis and adsorption where activity, stability, and selectivity depend on the characteristics of both acidic and metallic sites [5,6], as well as the pore size in the case of zeolites. Depending on the support properties and niobium content, dierent species of NbOx are presented. For basic support surfaces, only NbO6 is present. For acidic surfaces such as zeolites, there are NbO7 and NbO8 species as well as NbO6. The location of these units depends on the chemistry of the support hydroxyl group [7]. Microporous, such as Y [8] and ZSM-5 [9,10] zeolites, and mesoporous MCM-41

486

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

and MCM-48 [11,12] materials are used as supports to the above mentioned oxides. However, there is still considerable disagreement over how to characterize active catalytic sites in micro and mesoporous materials. An important goal of the work in our group has been to develop a better understanding of the chemistry that occurs at the acidic sites in well-characterized zeolites (e.g., ZSM-5 [13], Y [14], TS-1 [15], beta [16], mordenite [17]) and MCM-41 [18]. NbZSM-5 has been prepared by several methods. Chang et al. [19] used aqueous impregnation of ammonium niobium oxalate and identied Nb5+ as the charge compensation cations. EPR studies indicated niobium incorporation on MFI structure using hydrothermal synthesis and the presence of SiAOASi and NbAOASi units on the silicalite network [20]. A structural stabilization was found when Nb was incorporated on Cu-ZSM-5 catalysts, leading to a decrease of Al framework elimination during thermal treatments [21]. Catalytic tests using NbZSM-5 showed its eciency on methanol to hydrocarbons synthesis [22], ethane oxygenation [19], m-xylene isomerization and n-heptane cracking [10]. Prasetyoko and co-workers showed that NbTS-1 acts as bifunctional catalysts (oxidative and acidic functions) in the transformation of 1-octene to 1, 2-octanediol through the formation of 1, 2-epoxyoctane [23]. Finally, CuANb/ ZSM-5 systems having Nb instead of Al in the MFI structure showed high activity at moderate temperatures for NO decomposition [24]. In the present paper, we have investigated the properties of H-[Nb]ZSM-5 samples with 2, 5, 13 and 19 wt.% Nb2O5. Acidity and structural characterizations were made by XRD, ICP-AES, TG/DTG, FTIR, DRIFTS, FT-Raman techniques, 27Al and 29Si magic angle spinning solid state NMR spectroscopy (MAS-NMR) with and without cross-polarization (CP-MAS-NMR). 2. Experimental 2.1. Materials NH4ZSM-5 zeolite (CBV5524G) was obtained in the ammonium form from Zeolyst International with the following characteristics: SiO2/Al2O3 mole ratio = 50, surface area = 425 m2 g1, and 0.05 wt.% Na2O NH4[NbO(C2O4)2(H2O)2](H2O)n was furnished by Companhia Brasileira de Metalurgia e Minerac a o (CBMM). Pyridine (99%, Vetec) was distilled over CaH2 (Merck) using a 12 in. Vigreux column. Cyclohexane (99.5%, Vetec) was dis molecular tilled over P2O5 (Merck) and stored over 3 A sieves (Aldrich) just prior to the experiments. All other reagents used in this work were analytical grade and used as supplied. 2.2. Preparation of the catalysts The catalysts were prepared using the aqueous impregnation procedure through adequate mixtures of

NH4ZSM-5 and ammonium niobium oxalate in water producing zeolites with 2, 5, 13 and 19 wt.% Nb2O5. Each system was kept stirring at 80 C until all of the water had evaporated. Next, each solid was dried in an oven at 100 C to eliminate the solvent from zeolite pores. The calcination step was done in a mue furnace using ambient air (deep-bed conditions) at 450 and 550 C (10 C min1) and time intervals between 2 and 8 h. 2.3. Elemental and XRD analyses Silicon, aluminum and niobium contents were analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES, Spectroame FVM3). Each sample was digested by LiBO2 in platinum crucibles at 950 C for 20 min, dissolved in a 2 mol L1 HCl solution at 40 C and then adequately diluted. X-ray powder diraction (XRD) analysis was conducted in a Rigaku X-ray diractometer (model D/MAX-2A/C) ) at 40 kV and 20 mA. A using CuKa radiation (1.5418 A 2h range from 5 to 50 was scanned at 2 min1. 2.4. Nitrogen adsorption isotherm measurements Surface area, pore volume and pore distribution were obtained by the BET method of nitrogen adsorption isotherms at 77 K (Quantachrome, NOVA 1200). Before the measurements, the samples were pre-treated at 300 C under dynamic vacuum. 2.5. Thermal analysis TG/DTG and DTA curves were obtained in a 2960 simultaneous TG-DSC (TA Instruments) using synthetic air (99.999%) as the purge gas (100 mL min1). The analyses were made from ambient temperature ($26 C) to 1500 C at 10 C min1. 2.6. Infrared measurements FTIR spectra were recorded on a Bruker Equinox 55 spectrometer equipped with a DTGS detector. FTIR transmittance measurements (256 scans with 4 cm1 resolution) were performed using a standard KBr pellet technique. After base adsorption experiments, samples were analyzed for the presence of Brnsted and Lewis acidities. Diuse reectance FTIR spectroscopy (DRIFTS) was carried out using a HARRICK diuse reectance accessory (The Praying MantisTM) and a high temperature chamber with KBr windows. The samples were dehydrated in situ at 450 C for 1 h in vacuum and then cooled to room temperature. All spectra were acquired as absorbance against an alignment mirror (256 scans with 4 cm1 resolution). FT-Raman spectra were recorded on a Bruker FRA 106/S module attached to a Bruker Equinox 55 spectrometer (256 scans and 2 cm1 resolution). The laser (Nd:YAG) excitation wavelength and laser power were

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

487

1064 nm and 126 mW, respectively, and the signal was detected by a liquid N2 cooled Ge detector. 2.7. Pre-adsorption of pyridine Gas phase pyridine (Py) adsorption was conducted simultaneously for all zeolite samples accordingly to the procedure described by Ghesti and co-workers [14]. Platinum crucibles loaded with the samples ($20 mg) were placed in a shallow porcelain plate and inserted into a glass reactor adapted to a tubular furnace (Model F21135 Thermolyne). The catalysts were dehydrated in dried N2 (100 mL min1) at 300 C for 1 h, cooled to 100 C, and then gaseous pyridine diluted in N2 was allowed to pass through the samples for 1 h. The temperature was held at 100 C under N2 for 2 h to remove any physically adsorbed pyridine. After that, the samples were analyzed by desorption using TG/DTA (from 26 to 800 C at 10 C min1) and FTIR. 2.8. Solid state NMR measurements NMR experiments were performed at 7.05 T in a Varian Mercury Plus NMR spectrometer equipped with a 7 mm Varian probe and zirconia rotors. 27Al MAS spectra were recorded at 78.172 MHz, speed of 6 kHz, pulse duration of 1.0 ls (10), a recycle delay of 0.4 s and 6000 scans. 3 The spectra were referenced to AlH2 O6 (0 ppm). 29Si MAS spectra were recorded at 59.595 MHz, speed of 3 kHz, pulse duration of 7.5 ls (90), a recycle delay of 20 s and 500 scans. 29Si CP-MAS spectra were obtained under the same conditions and contact time of 1000 ls. The 29Si spectra were indirectly referenced to kaolin (91.5 ppm). 3. Results and discussion 3.1. Structural analysis Chemical and surface analyses of the samples by ICPAES and BET, respectively, are given in Table 1. Specic surface area and pore volume obtained by nitrogen adsorption isotherms for parent ZSM-5 zeolite show an increase on specic surface area after calcination at 450 C from 364.5 to 389.6 m2 g1 and from 0.228 to 0.243 cm3 g1,

respectively. This increase is related to residual water and template molecules presented on zeolite channels before calcination. When Nb2O5 is added to HZSM-5, at rst, it is observed a decrease on the surface area and pore volume for the 2 wt.% sample due to the blockage of some active sites of zeolite. Maxima values for surface area and pore volume are obtained for 5 wt.% Nb2O5 where these species can contribute for the surface area of the support. A significant decrease of specic surface area and pore volume is observed in calcined samples having more than 13 wt.% of Nb2O5. These trends are associated with the fact that at this loading, Nb2O5 can block zeolite ZSM-5 channels and cavities, and act as a modifying agent of the zeolite structure, besides the formation of monolayer of Nb2O5 over HZSM-5. Nb2O5/ZSM-5 samples showed a preservation of the zeolitic structure after impregnation and activation procedures as conrmed by XRD (Fig. 1). Apart from a small loss of crystallinity in relation to the parent ZSM-5, the strongest three peaks between 22.5 and 25.0 (2h) were almost coincident in all XRD patterns, with no peaks related to Nb2O5 crystalline phases or to metal oxide clusters. Studies of phase transition of pure Nb2O5 nH2O at 450 C by Braga and co-workers [25] showed the appearance of a structure with low crystallinity and characteristic reections of TTANb2O5. In our samples, a peak at 28.5 is observed for ZSM-5 with 19% of Nb2O5 after calcination at 550 C which does not appear in HZSM-5 diraction prole, and is assigned to Nb2O5 reection. The peak at

Fig. 1. XRD patterns of the samples: (a) NH4ZSM-5, (b) Nb(2)NH4ZSM-5, (c) Nb(5)NH4ZSM-5, (d) Nb(13)NH4ZSM-5 and (e) Nb(19)NH4ZSM-5.

Table 1 Chemical and physical properties of the zeolite samples treated at temperature of 450 C Sample NH4ZSM-5 HZSM-5 Nb(2)HZSM-5 Nb(5)HZSM-5 Nb(13)HZSM-5 Nb(19)HZSM-5
a b

(Si/Al)totala 22.5 22.5 25.9 23.9 22.5 22.4

Nb (%)a 0 0 1.4 3.5 9.1 13.3

Nb2O5 (%)a 0 0 2 5 13 19

Surface area (m2 g1)b 364.5 389.6 379.3 391.9 362.4 305.1

Pore volume (cm3 g1)b 0.228 0.243 0.229 0.256 0.237 0.208

Calculated by ICP-AES. Calculated by BET.

488

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

2h = 28.5 (XRD analysis) for Nb(19)HZSM-5 was observed only for the sample calcined at 550 C because, at this temperature, Nb2O5 reection could be observed although this is not possible for the sample calcined at 450 C. In addition, the higher Nb content on zeolite, the higher monolayer coverage and the more evident Nb2O5 reection. Prazetyoko et al. [23] reported a splitting of the diraction line around 2h = 29 for ammonium niobium oxalate after calcination at 550 C, indicating the sample as crystalline niobium pentoxide phase with orthorhombic structure (TANb2O5). Ko and Weissman [26] indicated that pure Nb2O5 nH2O calcined at 450 C presents a TT-phase of low crystallinity (d- and c-Nb2O5) composed of distorted octahedral and bipyramidal pentagons. At 600 C, niobium oxalate decomposition is complete. This behavior is strengthened by XRD analysis of Nb2O5 supported on silicaalumina [25], where the formation of a crystalline phase only occurs at 800 C (a mixture of T, M and HANb2O5 phases). Supported Nb2O5 is much more resistant to the formation of crystalline phases. The lower extension of Nb2O5 phase transformation is related to the degree of interaction with the support. In our samples, Nb2O5 crystalline T-phase was achieved at higher temperature than pure Nb2O5 due to a strong interaction between zeolite ZSM-5 and niobium pentoxide [26]. In addition, the formation of crystalline phases on supported materials is also dependent on loading. At lower concentration (below the monolayer), Nb2O5 is well dispersed and even at high temperatures it does not show any transition phase [25,27]. On the other hand, at higher loadings (over the monolayer) the increase of Nb2O5 amount makes the diusion on the surface easier and crystallization may occur at lower temperatures. 3.2. Thermal analysis Figs. 2 and 3 show the TG curves of NH4ZSM-5 and modied samples Nb(X)NH4ZSM-5 (X = 2, 5, 13 and 19 wt.% Nb2O5) before calcination, respectively. NH4ZSM-5 thermo-decomposition shows a weight loss of

Fig. 3. DTG curves of non-activated samples: (a) Nb(2)NH4ZSM-5, (b) Nb(5)NH4ZSM-5, (c) Nb(13)NH4ZSM-5 and (d) Nb(19)NH4ZSM-5.

6.02% below 115 C, attributed to water molecules physically adsorbed, followed by a loss of 1.11% between 115 and 247 C due to chemically adsorbed water, and ammonia from 247 up to 522 C (loss of 1.07%). The systems containing niobium show four dierent weight loss stages. The rst one, between 25 and 137 C, refers to the removal of physically adsorbed water (endothermic process). The second, between 116 and 214 C, is due to chemically adsorbed water. The third stage, between 190 and 441 C can be related to residual water occluded in zeolite pores or to ammonia from both parent materials (zeolite and niobium sources), besides the elimination of CO and CO2. The small loss at 600 C is connected to the oxidation of pyrolytic carbon by oxygen in the synthetic air atmosphere [28]. The fourth stage (exothermic process), between 395 and 655 C, refers to the decomposition of niobium oxalate (fragments of carbon monoxide and dioxide) [28]. A possible transition phase at about 1400 C was assigned according to DTA curves for the zeolite and Nb2O5 and conrmed by XRD analysis of the catalysts after TG/ DTA runs (not shown). Although, this peak may be of low intensity, it is reproducible for the other samples containing Nb at and over 19 wt.% and it was already observed for other supported catalysts [25,27]. The decomposition of ammonium niobium oxalate can be represented by the following NH4 NbOC2 O4 2 H2 O2 H2 On O2 ! xNH3
D

y CO2 wH2 O zNb2 O5

Fig. 2. DTG (a) and DTA (b) curves for NH4ZSM-5 zeolite from room temperature to 1500 C.

In addition, a small weight loss occurs between the forth stage and the transition phase, which may be related to zeolite ZSM-5 dehydroxylation. For NH4ZSM-5, this dehydroxylation begins at 522 C reaching 1.79% weight loss. The samples containing 2, 5 and 13 wt.% Nb2O5 showed lower losses (0.59, 0.38 and 0.16%, respectively) and a slight increase (0.29%) is observed for 19 wt.% Nb2O5. This indicates that impregnation of 215 wt.% Nb2O5 on ZSM-5 decreases its dehydroxylation and thus promotes thermal stabilization. Similar behavior occurs

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

489

to Nb/Cu-ZSM-5, where Nb acts as active part on the thermal stabilization [21,24]. Although thermal analysis point to 620 C as the temperature for complete removal of water and organic residues, other temperatures were studied in order to avoid a possible reduction of acidic properties. Thermal treatments at 450, 500 and 550 C from 2 to 12 h for all samples indicated that at 450 C/8 h the complete removal of ammonia occurs, and only 0.44% of organic residues remain in Nb(13)HZSM-5 and 0.52% in Nb(19)HZSM-5. At 550 C/2 h, no organic residue is present in the samples. 3.3. FTIR, DRIFTS and raman studies FTIR spectra of niobium modied samples calcined at 450 C and 550 C are presented in Fig. 4. The band at 1225 cm1 is assigned to external asymmetric stretching vibration of four chains of ve member rings disposed around a double helix. The band at 1103 cm1 is assigned to internal asymmetric stretching of SiAOAT bonds. A band at about 798 cm1 is related to symmetric stretching of external bonds between tetrahedra and at 547 cm1 there is a vibration sensitive to the zeolite topology, caused by external bonds of double ve member rings. At last, vibrations of internal bonds (TAO) of SiO4 and AlO4 tetrahedra are located near 454 cm1 [29]. Vibrations related to NbAO stretching are usually observed between 600 and 950 cm1 [9,12,3032]. For the Nb(X)HZSM-5 samples (Fig. 4), the spectra indicate a shoulder around 948 cm1, which becomes more intense for Nb(13)HZSM-5 and Nb(19)HZSM-5. The samples before calcinations also have this shoulder. Vibrations around 949 cm1 can have some contributions and among them, a band near 960 cm1 is due to the disturbance of SiAO by the presence of a metallic ion in a nearby environment [9,12]. Thus, this band is used to indicate the incorporation of metals in the framework [3032]. After thermal activation (450 C), the band at 1400 cm1 related to NAH stretching disappears in all samples. Fig. 5 shows DRIFTS spectra of HZSM-5 and Nb(19)HZSM-5 calcined at 450 C in the hydroxyl region. A band

at 3732 cm1 is assigned to isolated or terminal silanol groups. Another strong band at 3595 cm1 is related to Brnsted acidic sites of AlAO(H)ASi bridges. Two weak bands at 3683 and 3655 cm1 can be related to the presence of amorphous materials. Both spectra present a similar number of bands, but a shift from 3732 to 3723 cm1 is related to isolated silanol groups. In addition, a decrease of the intensity of the two strongest bands is observed when the amount of Nb increases (not shown), in agreement with the decrease in the amount of isolated silanol groups and Brnsted acidic sites [23,33]. These results imply an interaction between Nb and support through SiAOANb bonds. Fig. 6 shows Raman spectra for Nb2O5, HZSM-5 and Nb(X)HZSM-5 samples calcined at 450 C. A strong band around 500700 cm1 and a broad shoulder between 850 and 1000 cm1 displayed by Nb2O5 are presented on the modied Nb(X)HZSM-5 samples. As reported by Wachs and co-workers, they can be related to: (i) terminal Nb@O bond (8501000 cm1), (ii) NbO4 on low coverage surface (980990 cm1), (iii) NbO5 on intermediate coverage

Fig. 5. DRIFTS spectra in the OH region of dried zeolites: (a) HZSM-5 and (b) Nb(19)HZSM-5.

Fig. 4. FTIR spectra of zeolite in its dierent forms: (a) Nb(19)NH4ZSM5 after impregnation; (b) Nb(19)HZSM-5 activated at 450 C; (c) parent zeolite (NH4ZSM-5) and (d) HZSM-5 activated at 450 C.

Fig. 6. FT-Raman spectra of the samples activated at 450 C: (a) Nb2O5; (b) Nb(2)HZSM-5; (c) Nb(5)HZSM-5; (d) Nb(13)HZSM-5 and (e) Nb(19)HZSM-5.

490

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

surface (930950 cm1); (iv) NbO6 on high coverage surface (630650 cm1); and (v) NbAO stretching at 500 700 cm1 [21,34]. Nb@O bonds are presented as a highly distorted octahedral (NbO6), while NbAO bonds belong to slightly distorted octahedra, NbO7 and NbO8 species [34]. Microporous niobium silicalite (AM-11) [35] exhibits a strong band at 687 cm1 and weaker bands between 100 and 300 cm1, similar to nenadkevichite (titanium niobium silicalite), which displays bands at 668 and 226 cm1. For both systems, the bands are associated to NbO6 octahedra. The higher Nb content on zeolite, the more pure the Nb2O5 spectrum is, which indicates Nb2O5 above monolayer coverage. On the contrary of what happens to the Nb2O5 spectrum, the modied Nb(X)HZSM-5 spectra (Fig. 6) display two controversial bands: one around 380 cm1 and another weak band at 294 cm1, both found for niobium silicalite NbS-1 [9]. A decrease in the number of acid sites was observed for the parent zeolite and for the niobium modied samples when they were treated at 550 C for 8 h, probably due to dehydroxylation and or dealumination of zeolite. 3.4.
29

Si and

27

Al MAS-NMR studies

The NH4ZSM-5 spectrum presents an intense peak at 107 ppm (not shown) and a shoulder at %110 ppm, besides a weak signal at 100 ppm. The lines at 107 and 110 ppm are attributed to Si(0Al) environment and the line at 100 ppm to Si(1Al) units. The split at %110 ppm can be assign to Si atoms in non equivalent crystallographic sites, which can cause dierent chemical splits for the same Si(nAl) environment [30,36].

There is a correspondence between samples in relation to the number of peaks, shifts and intensities. 29Si MASNMR spectra shows at least two regions with superimposed resonance lines, d = 96 up to 98 ppm and d = 100 ppm, a precise signal at d = 107 ppm, and an additional shoulder at d % 111 ppm. According to Rocha and co-workers [36], it is dicult to ascribe new superimposed signals because there is no systematic study reported in the literature about the relation between the chemical shift of 29Si NMR and the number of Nb polyhedra coordinated to a Si tetrahedron. A behavior similar to TS-1 is expected, where a systematic shift to lower eld is observed when the number of Ti polyhedra bound to a Si tetrahedron increases. Thus, it was determined that the peaks at d = 107 and 111 ppm are related to Si (4Si, 0Al, 0Nb) environment. A resonance at %100 ppm is attributed to Si (3Si, 1Al) or Si (3Si, 1Nb) environments. The signal around 98 ppm is tentatively assigned to structural defect sites (SiOH and/or SiOM+, M = Na+, TPA+ Tetrapropylammonium) or Si(2Si, 2Nb) or even to Si(3Si, 1Nb) after the insertion of Nb2O5 on the zeolite structure, where the niobium cation can replace framework aluminum sites. Due to the superimposition of signals observed in all 29 Si MAS NMR spectra of modied samples, 1H-29Si NMR cross-polarization (CP) was used to detect an increase of signal intensity (and even those weak signals that can not be detected without CP) for Si atoms at defective SiOH sites (hydroxyl nests generated by Al removal) inside the zeolite structure [9,37,38]. In that way, it was possible to assign a strong increase in signal intensities between 96 and 98 ppm in all CP-MAS spectra of activated and non-activated samples, in addition to a Si(0Al) peak of the parent structure (Figs. 7 and 8).

Fig. 7.

29

Si MAS-NMR (1) and 1H-29Si CP-MAS NMR spectra (2) of zeolite: (a) NH4ZSM-5 and (b) HZSM-5.

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

491

Fig. 8.

29

Si MAS-RMN (1) and 1H-29Si CP-MAS NMR spectra (2) of zeolite: (a) Nb(19)NH4ZSM-5 and (b) Nb(19)HZSM-5.

Thermal treatments of zeolites can lead to considerable dealumination of zeolitic structure [39]. To evaluate the dealumination extent, Si/Al ratios were determined through 29Si MAS-NMR data and the Loewenstein Eq. (2) for samples before and after thermal treatments
4 Si=Al R4 m0 I4;n =Rm0 n=4I4;n

All 27Al MAS-NMR spectra (not shown) exhibit a resonance line at 55 5 ppm corresponding to the signal of tetrahedral aluminum species. Another very weak signal at %0 ppm is assigned to extra framework aluminum in an octahedral coordination. 3.5. Strength, distribution and total number of sites In order to determine the total number of acidic sites, the desorption prole of pyridine was analyzed by TG/DTA. Two phenomena were identied by DTA curves, an endothermic peak between 130 C and 318 C, and an exothermic peak between 420 C and 735 C. These peaks were associated to weaker (n1) and stronger acidic sites (n2), respectively. According to Table 3, HZSM-5 activated at 450 C/8 h presents the highest number of acidic sites (0.46 mmol g1) desorbed between 200 and 800 C, and this number decreases as the Nb2O5 content increases, reaching 0.38 mmol g1 for Nb(19)HZSM-5. Samples activated at 550 C/8 h have a decrease in the total number of acidic sites. This can be related to higher temperature/time exposition producing dehydroxylation and/or dealumination of zeolite. Samples treated at 450 C/8 h and 550 C/2 h showed a similar behavior under desorption conditions. Nonetheless, the second desorption for Nb(X)HZSM-5 samples display maxima at lower temperatures in relation to the parent HZSM-5, indicating a lower stability for pyridinium ion. It can be concluded that calcination treatments at 450 C/8 h and 550 C/2 h were good methods to activate the materials. Calcinations at 550 C/ 8 h promoted a considerable loss of acidity. In addition, results from Table 3 indicate that the addition up to 13% Nb2O5 contributes to stability of the structure under

m 4 and 0 6 n 6 4 2

Table 2 shows the results of Si/Al molar ratio for the niobium modied samples before and after calcination. All the spectra were deconvolved using a Gaussian distribution considering four contributions for silanol groups. The parent zeolite (NH4ZSM-5, Si/AlFr = 23.7) undergoes dealumination after thermal treatment (HZSM-5, Si/AlFr = 25.9). For niobium modied samples, even after calcinations, almost identical values are observed for Si/Al ratio. The addition of niobia on ZSM-5 promotes a decrease of Al atoms eliminated from the framework during thermal treatments [21].

Table 2 Total and framework Si/Al ratios calculated by ICP-AES and 29Si NMR, respectively Sample (Si/Al)totala 22.5 26.0 24.0 22.5 22.4 23.7 24.0 24.5 24.9 24.8
29

(Si/Al)Frb Non-activated Activated 25.9 24.1 23.9 24.5 24.9

ZSM-5 Nb(2)ZSM-5 Nb(5)ZSM-5 Nb(13)ZSM-5 Nb(19)ZSM-5


a b

Calculated by ICP-AES. Framework Si/Al ratios calculated by

Si NMR deconvolution.

492

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

Table 3 Total number of acidic sites determined by TG/DTG for zeolites treated at 450 C and 550 C Sample HZSM-5 Nb(5)HZSM-5 Nb(13)HZSM-5 Nb(19)HZSM-5 HZSM-5 Nb(5)HZSM-5 Nb(13)HZSM-5 Nb(19)HZSM-5 HZSM-5 Nb(19)HZSM-5 Activation 450 C/8 h n1 (120400 C) 0.02 0.07 0.06 0.06 0.03 0.08 0.04 0.06 0.02 0.01 n2 (400800 C) 0.44 0.35 0.34 0.31 0.41 0.37 0.35 0.32 0.38 0.29 nT (mmol g1)a 0.46 0.43 0.40 0.38 0.45 0.45 0.40 0.39 0.40 0.30 nT (mmol g1)b 0.46 0.36 0.36 0.35 0.42 0.41 0.38 0.36 0.39 0.32

550 C/2 h

550 C/8 h

nT = n1 + n2. a Py desorbed from 120 to 800 C. b Py desorbed from 300 to 800 C.

450 C/8 h and 550 C/2 h, without signicant loss of acidity. Although two sites have been distinguished by TG/DTA on the studied samples, the amount of these sites for the samples under study is dierent from the reported values from preliminary CalAd (calorimetryadsorption studies of pyridine in liquid phase) that showed to be: 0.53 mmol (n1) of weak hydrogen bond sites (DH = 8.6 3.8 kcal/ mol) and 0.0415 mmol (n2) of strong Brnsted sites (DH = 42.1 0.8 kcal/mol) [13]. On the other hand, the data presented in Table 3 show the opposite trend, i.e., n1 < n2. This dierence can be attributed to the fact that gas phase adsorption experiments were carried out at 100 C while CalAd experiments were at 25 C. Auroux and Datka [40], through microcalorimetry experiments using pyridine on mordenite, identied that at temperatures higher than room temperature, the pyridine molecule reacts with any available acidic sites regardless of their acidities. This indicates that thermal desorption analysis can lead to an overestimation of the amount of strong sites. Drago and co-workers [13] showed that for a ZSM-5 zeolite, most of the tetrahedral Al sites make hydrogen bonds with pyridine. Finally, Corma [6] indicates that pyridine adsorbs strongly even to the weak acidic sites without specicity due to its strong basicity. HZSM-5 has a framework aluminum density of 0.66 mmol g1 calculated from elemental analysis and about 0.46 mmol g1 (Table 3) of the total acidic sites calculated by thermogravimetry. This lower value suggests that not all Al sites are accessible to pyridine or capable of generating Brnsted sites [13]. The eect of calcination procedures can be followed by considering that at lower temperatures (450 C), the ratio of pyridine adsorbed by framework Al was 0.70, decreasing to 0.61 when activated at 550 C/8 h. The highest decrease was observed with Nb(19)HZSM-5 (0.46) under the same calcinations conditions, indicating that the access to the acidic sites can be blocked by Nb2O5 species. FTIR spectra (Fig. 9) display four bands in the pyridine region (1624 1440 cm1). Drago and co-workers observed

Fig. 9. FTIR spectra of zeolite treated at 450 C with adsorbed pyridine: (a) HZSM-5; (b) Nb(5)HZSM-5; (c) Nb(13)HZSM-5 and (d) Nb(19)HZSM-5.

the corresponding bands after calorimetric experiments using liquid pyridine as the probe molecule for ZSM-5 (Si/Al = 36) calcined at 450 C/24 h/vacuum [13]. Similar behavior is observed for all studied samples, indicating that gaseous pyridine and liquid pyridine (not shown) access Brnsted sites (1543 and 1491 cm1) and hydrogen bond sites (1444 cm1). In addition, on pure HZSM-5 and Nb(X)HZSM-5 samples, the rst peak (%120400 C) is associated with an endothermic phenomenon (DTA curves) for physical desorption of pyridine, which corroborates the FTIR results for hydrogen bond sites. 4. Conclusions Thermal analysis of supported ZSM-5 containing 2, 5, 13 and 19 wt.% of niobium pentoxide indicated two exothermic peaks: one near 600 C attributed to oxidative decomposition of niobium oxalate and another at 1375 C related to HANb2O5 phase transition. Nitrogen adsorption isotherms (BET method) revealed that both specic surface area and pore volume decrease when the amount of niobium pentoxide increases, reaching a maxi-

I.C.L. Barros et al. / Microporous and Mesoporous Materials 109 (2008) 485493

493

mum value of 5 wt.%. Above that, niobium pentoxide can be blocking the channels and cavities of ZSM-5. 29Si CPMAS-NMR indicated the presence of defected silanol groups at 98 ppm in addition to Si(0Al) and Si(1Al or 1Nb) at 110, 107 and 100 ppm, respectively. These results corroborate that the addition of Nb avoids structural dealumination after thermal treatments. Gas phase pyridine adsorption and FTIR studies showed that all samples present Brnsted and hydrogen bonding sites. The inuence of calcination temperature on acid sites indicated that parent zeolite (HZSM-5) has the highest acidity when treated at 450 C, decreasing at 550 C, from 2 up to 8 h. FTIR, XRD and FT-Raman results indicate the presence of niobium pentoxide monolayer on the surface of ZSM5 for amounts lower than 13 wt.% of Nb2O5. Acknowledgments To CNPq, CAPES/PQI, FINATEC, Finep-CTPetro, Finep-CTInfra, UnB-IQ, UnB-IG, UFAM-ICE, FAPDF/ SCDT/CNPq and CBMM (donation of niobium reagents). References
sticas, Propiedades y Aplicaciones [1] G. Giannetto, Zeolitas: Caracter n Tecnolo gica, Caracas, 1990. Industriales, Editorial Innovacio [2] D.W. Breck, Zeolite Molecular Sieves: Structure Chemistry and Use, John Wiley & Sons Inc., New York, 1974. [3] J.B. Higgins, in: P.J. Heaney, C.T. Prewitt, G.V. Gibbs (Eds.), Silica: Physical Behavior, Geochemistry and Materials Applications, Book Crafters Inc., Washington DC, 1994. [4] (a) M.A. Abdel-Rehim, A.C.B. Santos, V.L.L. Camorim, A.C. Faro Jr., Appl. Catal. A 305 (2006) 211; (b) M.I. Sairre, E.S. Bronze-Uhle, P.M. Donate, Tetrahedron Lett. 46 (2005) 2705; (c) F.M.T. Mendes, C.A.C. Perez, F.B. Noronha, M. Schmal, Catal. Today 101 (2005) 45. [5] (a) W.E. Farneth, R.J. Gorte, Chem. Rev. 95 (1995) 615; hman, B. Ganemi, E. Bjo (b) L.O. O rnbom, K. Rahkamaa, R.L. Keiski, J. Paul, Mater. Chem. Phys. 73 (2002) 263. [6] A. Corma, Chem. Rev. 95 (1995) 559. [7] V. Rakic, V. Dondur, U. Mioc, D. Jovanovic, Top. Catal. 19 (2002) 241. [8] K. Tanabe, Catal. Today 78 (2003) 65. [9] A.M. Prakash, L. Kevan, J. Am. Chem. Soc. 120 (1998) 13148. [10] A.O. Florentino, M.J. Saeki, P. Cartraud, P. Magnoux, M. Guisnet, m. Nova 20 (1997) 9. Qu

[11] (a) X. Gao, I.E. Wachs, M.S. Wong, J.Y. Ying, J. Catal. 203 (2001) 18; (b) L.J. Burcham, J. Datka, I.E. Wachs, J. Phys. Chem. B 103 (1999) 6015. [12] I. Nowak, Colloids Surf. A 241 (2004) 103. [13] R.S. Drago, S.C. Dias, M. Torrealba, L. Lima, J. Am. Chem. Soc. 119 (1997) 4444. [14] (a) S.C.L. Dias, J.L. Macedo, J.A. Dias, Phys. Chem. Chem. Phys. 5 (2003) 5574; (b) G.F. Ghesti, J.L. Macedo, V.C.I. Parente, S.C.L. Dias, J.A. Dias, Micropor. Mesopor. Mater. 100 (2007) 27. [15] R.S. Drago, S.C. Dias, M. McGilvray, A.L.M.L. Mateus, J. Phys. Chem. B 102 (1998) 1508. [16] J.L. Macedo, G.F. Ghesti, M. Paulo, S.C.L. Dias, J.A. Dias, Micropor. Mesopor. Mater., submitted for publication. [17] J.L. Macedo, S.C.L. Dias, J.A. Dias, Micropor. Mesopor. Mater. 72 (2004) 119. [18] M. Paulo, J.L. Macedo, F.A.C. Garcia, J.C.M. Silva, J.A. Dias, S.C.L. Dias, Micropor. Mesopor. Mater., submitted for publication. [19] Y.F. Chang, G.A. Somorjai, H. Heinemann, J. Catal. 154 (1995) 24. [20] J.P. Osegovic, R.S. Drago, J. Phys. Chem. B 104 (2000) 147. [21] I. Nowak, M. Ziolek, Chem. Rev. 99 (1999) 3603. [22] P.T. Wierzchowski, L.W. Zatorski, Catal. Lett. 9 (1991) 411. [23] D. Prasetyoko, Z. Ramli, S. Endud, H. Nur, Mater. Chem. Phys. 93 (2005) 443. [24] I. Sobczak, P. Decyk, M. Ziolek, M. Daturi, J.-C. Lavalley, L. Kevan, A.M. Prakash, J. Catal. 207 (2002) 101. [25] V.S. Braga, J.A. Dias, S.C.L. Dias, J.L. Macedo, Chem. Mater. 17 (2005) 690. [26] E.I. Ko, J.G. Weissman, Catal. Today 8 (1990) 27. [27] V.S. Braga, F.A.C. Garcia, J.A. Dias, S.C.L. Dias, J. Catal. 247 (2007) 68. [28] H. Langbein, G. Wolki, Thermochim. Acta 264 (1995) 67. [29] M.A. Ali, B. Brisdon, W.J. Thomas, Appl. Catal. A 252 (2003) 149. [30] J.B. Nagy, Z. Gabelica, E.G. Derouane, P.A. Jacobs, Chem. Lett. (1982) 2003. [31] M.W. Anderson, O. Terasaki, T. Oshuna, A. Philippou, S.P. Mackay, A. Ferreira, J. Rocha, S. Lindin, Nature 367 (1994) 347. [32] M. Ziolek, Catal. Today 78 (2003) 543. [33] J. Kotrla, L. Kubelkova, C.-C. Lee, R.J. Gorte, J. Phys. Chem. B 102 (1998) 1437. [34] J.-M. Jengh, I.E. Wachs, Chem. Mater. 3 (1991) 100. [35] J. Rocha, P. Branda o, Z. Lin, A.P. Esculcas, A. Ferreira, M.W. Anderson, J. Phys. Chem. 100 (1996) 14978. [36] J. Rocha, P. Branda o, A. Phillippou, M.W. Anderson, Chem. Commun. (1998) 2687. [37] G. Engelhardt, D. Michel, High-resolution Solid-State NMR of Silicates and Zeolites, John Wiley & Sons Inc., New York, 1987. [38] J.B. Nagy, Z. Gabelica, E.G. Derouane, Chem. Lett. (1982) 1105. [39] C.A. Fyfe, G.C. Gobbi, G.J. Kennedy, Chem. Lett. (1983) 1551. [40] A. Auroux, J. Datka, Appl. Catal. A 165 (1997) 473.

Você também pode gostar