- Email: [email protected]
ZSM-5 zeolites prepared by different methods
Microporous and Mesoporous Materials 87 (2005) 93–102 www.elsevier.com/locate/micromeso
Effect of thermal treatment on surface and bulk properties of Fe/ZSM-5 zeolites prepared by different methods Mohamed Mokhtar Mohamed
a,*
, Ibraheem Othman Ali b, N.A. Eissa
c
a
b
Department of Chemistry, Faculty of Science, Benha University, Benha, Cairo 650371, Egypt Department of Chemistry, Faculty of Science, Al-Azhar University, Nasr City, Cairo 1184, Egypt c Department of Physics, Faculty of Science, Al-Azhar University, Nasr City, Cairo 1184, Egypt Received 11 June 2005; received in revised form 23 July 2005; accepted 23 July 2005 Available online 9 September 2005
Abstract Fe/ZSM-5 samples (5 wt.% Fe) prepared by in situ incorporation using TPABr template under hydrothermal conditions (Fe– ZSM-5in), chemical liquid deposition (Fe–ZSM-5imp) and solid–solid (Fe–ZSM-5ss) interaction were characterized by N2 physisorption, TG/DSC, X-ray diffraction, FTIR spectroscopy, UV–Vis diffuse reflectance spectroscopy and 57Fe Mo¨ssbauer spectroscopy techniques. Calcination at 550 °C leads to almost complete removal of template that was associated with dislodgment of significant fraction of Fe to external positions as recognized for the in situ prepared sample (Fe–ZSM-5in). This sample showed an increase in lattice volume suggesting the presence of the majority of Fe ions in tetrahedral positions inside zeolite channels and offered as well the lowest crystallites size (75 nm) and maximum SBET (453 m2/g) between all samples. On the other hand, Fe–ZSM-5ssbef, resulting ˚ ), from solid–solid interaction and subjected to heat treatment in vacuum at 200 °C, measured the lowest mean pore radius (r; 23 A and pore volume (Vp; 0.3887 cm3/g), giving a hint about the probability of finding neutral iron oxide nanoparticles (a-Fe2O3) as a separate phase that has been validated by Mo¨ssbauer (IS = 0.3 mm/s, QS = 0.2 mm/s, Heff = 520 kOe) and UV–Vis (400 nm) investigations. This sample also demonstrated that the majority of Fe occupied framework positions beside a fraction identified as small oligonuclear oxo-iron ions (Fe3þ x –O; 290 nm). Interestingly, Fe–ZSM-5ssaft, resulting from solid–solid interaction and sub˚ ) values, reflecting the enforced location of Fe jected to air calcination at 550 °C, measured maximum Vp (0.6380 cm3/g) and r (39 A in this sample leading to an effective pore widening and thus a pronounced mesoporosity is attained. IR bands due to masT–O in ZSM-5 (1105 cm1) showed a shift to lower wave numbers (1059 cm1) following Fe incorporation reflecting the extent of exchanging Fe in this sample (Fe–ZSM-5ssaft), unlike the rest of the samples which showed splitting, which accounted for the presence of residual Al3+ beside Fe3+ ions in the same site. Mo¨ssbauer data of this sample confirmed the latter result and indicated the maximum lattice imperfection and showed as well the lowest degree of crystallinity. The Fe–ZSM-5impaft sample, subjected to heat treatment at 550 °C, showed a-Fe2O3 species where that heated at 110 °C presented the lowest SBET (361 m2/g). More correlations were evaluated and discussed on the effect of thermal treatment on the existence of various Fe species (either framework or nonframework), their electronic states and local structures. Ó 2005 Published by Elsevier Inc. Keywords: Fe–ZSM-5; In situ preparation; Impregnation; Solid–solid interactions; Characterization; Fe species; Mo¨ssbauer spectroscopy
1. Introduction
*
Corresponding author. Tel./fax: +966 025274264. E-mail address: [email protected] (M.M. Mohamed).
1387-1811/$ - see front matter Ó 2005 Published by Elsevier Inc. doi:10.1016/j.micromeso.2005.07.032
Zeolites with 10-membered oxygen rings normally possess a high siliceous framework structure. The 10-membered oxygen ring zeolites also possess other important characteristic properties including high
94
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
activity, high tolerance to coking and high hydrothermal stability. Among the family of 10-membered oxygen ring zeolites, the MFI type (ZSM-5) zeolite is probably the most useful one. It has been acknowledged that the tetrahedral surrounding could be stable for cations with 0.414 > q > 0.225 (q = rM/rO, where rM and rO are the radii of the cation and the oxygen ion, respectively). Cations that do not fulfill this criterion such as Fe3+ may also be introduced into the framework under specific synthesis conditions. The introduction into the framework of ZSM-5 atoms other than Si and Al can modify significantly the acid–base and catalytic properties of the zeolites. Some modification studies have been made with framework iron insertion substituting aluminium in the synthesis process [1–4]. Such isomorphous substitution of Fe either for Al or Si of zeolites has been reported to exhibit improved catalytic properties in a variety of chemical reactions. Among them are ethanol conversion to hydrocarbons, polymerization, isomerization, dehydrogenation, reduction of NO either by NH3 or hydrocarbons and selective oxidations [5– 11]. For most of these reactions, the catalytic activity is thought to be related to extra-framework iron species (iron oxides, dinuclear iron sites and oxohydroxides aggregates) formed during ion exchange and solid–solid treatments [12,13]; those acting either independently or in synergy with protonic-Fe(OH)Si sites (Bro¨nsted acid). Fe exchanged ZSM-5 zeolite exhibited a high and stable catalytic activity in several selective redox reactions of great importance: NO [14], N2O [15] decomposition as well as selective catalytic reduction of NO by hydrocarbons in an excess of O2 [16]. On the other hand, CuZSM-5 showed lower catalytic activity and less stability towards NO reduction due to the alteration of the copper ionic state on the catalyst surface and to dealumination consequences. Accordingly, Fe–ZSM-5 is subjected to numerous studies principally for correlating the structure of the active site(s) in various applications. Thus, an intense attention still attracting many researchers to study structure–activity relationship for each active Fe site can be revealed. From the analysis of the literature data briefly discussed above, it is however evident that the formation of framework and non-framework Fe–ZSM-5 is primarily connected with the conditions applied during preparation and to the applied thermal treatment that is far from being completely understood. In this essence, the coordination and oxidation changes of Fe3+ species when incorporated with ZSM-5 were followed before and after calcination for samples prepared by different methods, including in situ incorporation during synthesis, impregnation and solid–solid interactions in conjunction with their electronic states, surface properties and local structures.
2. Experimental 2.1. Materials The materials used were: silicic acid powder, sodium hydroxide pellets (A.R. 98%), aluminum sulfate (Al2 (SO4)3 Æ 16H2O Merck), iron nitrate (Fe(NO3)3 Æ 9H2O Merck), tetrapropylammonium bromide (TPABr, Fluka), n-propyl amine (Merck) and concentrated sulfuric acid. 2.1.1. Preparation of ZSM-5 The hydrogels of the following oxide molar compositions were prepared for the synthesis of ZSM-5 zeolite: 3:12Na2 O : 0:162Al2 O3 : 6:185SiO2 : 0:185Q : 100H2 O where Q is the tetrapropylammonium bromide template. A specific amount of NaOH was added to silicic acid in a small amount of H2O (40 ml) while stirring, followed by heating at 80 °C until a clear solution was reached. The TPABr was dissolved in a little amount of H2O (10 ml) with heating at 50 °C for 20 min. The solution of TPABr was added to that of sodium silicate solution whilst stirring for 15 min. In addition, 1 ml of n-propyl amine was added as a mobilizing agent to that mixture. The aluminum sulfate, on the other hand, was dissolved in a small amount of H2O (10 ml) while adding 0.05-ml concentrated H2SO4 with stirring until reaching a clear solution. To the latter solution, the combined solution of sodium silicate and TPABr was added followed by stirring for 30 min. The pH of the mixture was adjusted at 11 by using NaOH (0.1 M) and H2SO4 (0.1 M) solutions. Finally, the mixture was hydrothermally treated at 160 °C in an oil bath, using stainless steel autoclaves, for 6 days. The autoclaves were removed at the specified time from the oil bath and quenched immediately with cold water. The solid product was filtered and washed with distilled water until the pH of the filtrate dropped to 8. The products were dried at 110 °C for 10 h, then calcined at 550 °C for 6 h in an air oven. 2.1.2. Preparation of Fe–ZSM-5 2.1.2.1. Inclusion of Fe–ZSM-5 zeolite (built-in). The synthesis procedure is described in molar compositions as follows: 3:12Na2 O : 0:138Al2 O3 : 6:185SiO2 : 0:0103Fe2 O3 : 0:185Q : 100H2 O where Q is the tetrapropylammonium bromide. Aluminum sulfate was dissolved in a small amount of water (5 ml) followed by addition of 0.05-ml H2SO4. On the other hand, the soluble complex of sodium silicate was formed onto which the template tetrapropylammonium bromide (TPABr) was added with n-propyl amine
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
as a mobilizing agent. On the other hand, the preparation of Fe(NO3)3 solution was carried out. A simultaneous addition of iron nitrate solution and sodium silicate solution into the aluminum sulfate was achieved, followed by stirring for 30 min. Sulfuric acid (0.1 M) or sodium hydroxide (0.1 M) adjusted the pH to 11. The amorphous gel formed was allowed to age for 1 h at room temperature. The reaction mixture was transferred to 300-ml stainless steel autoclaves and maintained in an oil bath at 160 °C under autogeneous pressure. The autoclave was removed from the oil bath after 6 days and quenched in cold water for product identification. The solid products were separated by filtration. Excess alkali was thoroughly washed with water repeatedly until the pH of washing liquid was close to 8, and the products were dried in an oven at 120 °C for 10 h, then calcined at 550 °C for 6 h to remove the template. This sample was referred to as Fe–ZSM-5in. The color of this sample was pale brown. 2.1.2.2. Preparation of Fe–ZSM-5 by CLD (chemical liquid deposition). The iron oxide in zeolites was also prepared by impregnation of ZSM-5 with an aqueous solution of iron nitrate to achieve a loading of 5 wt.% iron oxide (0.0082 molar ratio). The volume of solution used was that necessary to completely wet the zeolite sample. The reaction temperature was at 80 °C for 4 h by using temperature controller type REX-P 90. After impregnation, the water was slowly allowed to evaporate at 110 °C till dryness, and then the precursor sample was either dried at the same temperature for 6 h or calcined at 550 °C in an air oven for 6 h. The samples were referred to as Fe–ZSM-5impbef and Fe–ZSM-5impaft, respectively. The color of the samples was reddishbrown. 2.1.2.3. Preparation of Fe–ZSM-5 by solid–solid interaction. The ZSM-5 sample was degassed at 200 °C for 3 h to remove physisorbed water molecules prior to being physically mixed with iron nitrate (5 wt.% iron oxide (0.0082 as molar ratio) at room temperature in a glove bag filled with nitrogen gas (99% purity). The mechanically mixed sample was stirred overnight in a current of pure nitrogen flowing at a rate of 30 cm3/min. The obtained sample either heated at 200 °C under a reduced pressure of 105 Torr for 4 h or calcined at 550 °C in air for 6 h was referred to as Fe–ZSM-5ssbef and Fe– ZSM-5ssaft, respectively. The color of the Fe–ZSM-5ssbef sample was reddish-brown, whereas the Fe–ZSM-5ssaft sample obtained a brown color that converted to reddish-brown with time. 2.2. Experimental techniques The X-ray diffractograms of various zeolitic samples were measured by using a Philips diffractometer (type
95
PW 3710). The patterns were run with Ni-filtered copper ˚ ) at 30 kV and 10 mA with a radiation (k = 1.5404 A scanning speed of 2h = 2.5°/min. The crystal sizes of the prepared materials were determined using the Scherrer equation. The instrumental line broadening was measured using a LaB6 standard. The crystallinity of the prepared samples was calculated using the ratio of the sum of the areas of the most intense peaks for Fe–ZSM-5 samples (2h = 20–25°) to that of the same peaks for the standard (Na–ZSM-5 Mobil chemicals) and multiplying by 100. FT-IR spectra of the samples were recorded with a JASCO single beam FT-IR 5300-spectrometer with 50 co-added scans at 2 cm1 resolution. All IR measurements were carried out at room temperature using KBr technique. Optical spectra in the UV–Vis region were recorded using diffuse reflectance technique. The spectrometer was a Shimadzu reflectospectrometer type UV-2200A. The reflectance white standard was BaSO4. Thermogravimetric and differential scanning calorimetry (TG/DSC) were carried out using Shimadzu-50 thermal analyzer units. The sensitivity of TG and DSC measurements was 0.01 mg and 25 lv, respectively. In each run, about 10 mg of uncalcined sample was heated from room temperature to 1000 °C at a heating rate of 10 °C/min in a current of N2 flowing at a rate of 30 ml/min. The nitrogen adsorption isotherms were measured at 196 °C using a conventional volumetric apparatus. The specific surface area was obtained using the BET method. The micropore volume and the external surface area were obtained from the t-plot method. The Mo¨ssbauer spectra (MS) were recorded at room temperature using a constant acceleration spectrometer. The source activity was 3.70 GB9 of 57Co in Co matrix. Iron metal was used as a reference for the isomer shift (IS). The MS were fitted using a program based on the distribution of hyperfine magnetic field (HMF) and quadrupole splitting (QS).
3. Results and discussion 3.1. Surface texture The different surface characteristics of various investigated zeolitic materials were determined from N2 adsorption/desorption isotherms conducted at 77 K. These characteristics include BET-surface area (SBET), total pore volume (Vp), mean pore radius (r), surface area derived from Vlt plots (St), surface area of micropores (Sl), external surface area (Sext), surface area of wide pores (Swid), volume of micropores ðV lp Þ, volume of wide pores ðV wid p Þ and microporosity percentages (Table 1). The N2 adsorption/desorption isotherms for
96
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
Table 1 Some Surface characteristics of ZSM-5 and iron containing ZSM-5 investigated adsorbents; prepared by different methods, heating at 300 °C under a reduced pressure of 105 Torr Sample
SBET (m2/g)
St (m2/g)
V total p (cm3/g)
r ˚) (A
Sl (m2/g)
Sext (m2/g)
Swid (m2/g)
V lp (cm3/g)
V wid p (cm3/g)
Microporosity (%)
C-const.
ZSM-5 Fe–ZSM-5in Fe–ZSM-5ssaft Fe–ZSM-5ssbef Fe–ZSM-5impaft Fe–ZSM-5impbef
429 453 408 422 437 361
420 461 371 391 416 329
0.4166 0.6551 0.6380 0.3887 0.5430 0.3624
24 36 39 23 31 25
422 284 246 353 326 286
74 100 69 63 88 54
7 169 162 69 111 36
0.4054 0.4090 0.3843 0.3249 0.4045 0.2866
0.0112 0.2461 0.25375 0.06377 0.13846 0.07576
97 62 60 84 75 79
21 21 26 10 8 10
Note: BET-surface area (SBET); total pore volume (Vp); mean pore radius (r); surface area derived from V1t plots (St); surface area of micropores (Sl); external surface area (Sext); surface area of wide pores (Swid); volume of micropores ðV lp Þ; volume of wide pores ðV wid p Þ and microporosity percentages.
Fe–ZSM-5 samples (not shown) were typically type 1 [17] in nature. Although the validity of the BET model for highly microporous materials is questionable, the surface area derived from this model in the adapted pressure range P/P0 = 0.01–0.1 can still be used for comparative purposes [18]. At high P/P0, some deviation from type 1 isotherm was observed for some samples along with desorption hysteresis. This may be due to intracrystalline mesopores depicted from Vlt plots (Fig. 1) for all Fe containing ZSM-5 samples; except Fe–ZSM-5ssaft, those constructed depended on the value of C-constant in the BET equation. The existence of a downward deviation in the latter sample is indicative of the domination of narrow pores. Inspection of the results in Table 1 reveals that: (i) the values of SBET and St are close to each other for various investigated zeolitic samples and thus justify the correct choice of standard t curves used in pore analysis and rather indicate the absence of the ultra-micropores in these materials. The SBET of Fe–ZSM-5impbef measured the lowest specific area between all samples that comprises 16% from that of the parent. This can be due to the occurrence of some Fe species inside zeolite pores and thus blocking them. (ii) The values of external surface area of various samples constitute high values (>10%) of the SBET pointing to the porous nature of these materials and rather suggest the appreciable presence of mesopores that concurrently emphasized from ˚ ). The latter finding was conhigh r values (24–39 A firmed by both the slope position rendered for samples instead of the almost horizontal saturation plateau (characteristics of an ideal microporous structure) and upward deviations depicted for majority of the samples (Fig. 1). An increase in the mesopores area of some samples (see Sext in Table 1) could be due to the consequence of extraction of heteroatoms (Al, Fe) to non-framework positions. (iii) The specimen designated as Fe–ZSM-5ssaft measured the maximum Vp and r. This might indicate the enforced location of Fe species in the pores of this particular sample leading to an effective pore widening. Conversely, the Fe–ZSM-5ssbef sample measured the
˚ ), giving a hint about the opportulowest r value (23 A nity of finding Fe species as a separate phase and thus indicates a lower value of Vp comparatively. (iv) The percentage microporosity in Fe–ZSM-5ssbef measured the highest value (84%) between all samples, as rather confirmed from Vlt plot of the sample that verified only the presence of narrow pores as a consequence of the existence of downward deviations. (v) The observed increase in the BET surface area of Fe–ZSM-5in (5%) when compared with ZSM-5 can be attributed either to the creation of new pores due to departure of NOx compounds during the thermal dissociation of iron nitrate or to the formation of Fe2O3 species outside zeolite pores probably as a result of the template removal during thermal treatment. The present result might recommend using Fe–ZSM5in in surface driven applications, since this sample is shown to withstand surface accessibility (an increase of 5% in SBET, Vp max and r max) and contain both microporous and mesoporous characters. This result indicates as well the enforced location of Fe species in the pores of ZSM-5in specimen and specifically in compensation positions (tetrahedral sites) as will be visualized later by XRD, IR and Mo¨ssbauer investigations. 3.2. Material bulk characteristics 3.2.1. Thermal analysis Fig. 2 shows TG and DSC curves of uncalcined ZSM-5 synthesized using TPABr while being heated under an argon atmosphere. The weight loss occurs in three steps: the first between 25 and 400 °C due to desorption of water either present on external or internal surfaces and comprises a weight loss of 0.8%, the second step between 420 and 520 °C, due to thermal decomposition of TPA+ ions occluded in various types of pores and comprises a weight loss of 6.4% and the third, between 520 and 900 °C that comprises 0.3% weight loss and is ascribed to elimination of H2O during condensation of surface OH groups and/or removal of chemically bound H2O. DSC, on the other hand, shows
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
97
1. 5 300
1. 0
TG-DSC (f)
0.0 300
(e)
200
608
Enthalpy/µv.s/mg
100
0.0 -1.0 -2.0 -3.0 -4.0 -5.0 470
200
TG/%
DTG/%/min
0.0
5.0 0.0
-2
100
-1.0 -4
0.0
-2.0 (d)
200
-6 -3.0
Vacc (g/ml)
-8 25 100
200
400
600
800
1000
Temperature/ºC Fig. 2. TGA/DSC curves of synthesized ZSM-5 zeolite.
0.0 400 (c)
300 200 100 0.0 400
(b)
300 200
the 608 peak to elimination of H2O during condensation of surface OH groups is unlikely but indeed could be due to thermal decomposition of TPA+ molecules occluded in narrow pores [19]. This is confirmed from lowering the total mass loss of TPA+ (6.4%) than that of the calculated value (11.7%) [20] assuming four TPA+ cations per unit cell in a pure ZSM-5 crystal. Likewise, similar results can be obtained from the DTG curves for the other ZSM-5 zeolite and may correspond to the relatively strong interactions between anionic inorganic framework and cationic organic species located at different sites of the ZSM-5 framework [21].
100 0.0 400 (a)
300 200 100 0.0 0.0
5
10
15
20
P/Po Fig. 1. V1t plots of (a) synthesized ZSM-5, (b) Fe–ZSM-5in, (c) Fe– ZSM-5impaft, (d) Fe–ZSM-5impbef, (e) Fe–ZSM-5ssaft and (f) Fe–ZSM5ssbef.
four endothermic peaks at 100, 470, 608 and 875 °C. The third weight loss which occurred in the 520–900 °C range (TG curve) includes the two endothermic peaks positioned at 608 and 875 °C. Accordingly, ascribing
3.2.2. Crystalline composition XRD patterns of Fe–ZSM-5 samples are displayed in Fig. 3 together with parent ZSM-5 synthesized from TPABr template. The XRD patterns of all iron-containing ZSM-5 showed the typical spectrum of MFI structure. The samples were well crystalline, and a few of a-Fe2O3 phase which were hardly recognized as small ˚ ), 34.1 (2.39 A ˚ ), 42.3 particles at 2h = 24.8 (3.69 A ˚ ˚ (2.26 A) and 54.7 (1.71 A) agreed with those for aFe2O3 phase with d-spacings = 3.67, 2.42, 2.20 and ˚ , respectively [22]. The diffraction lines of ZSM1.69 A 5 around 2h = 45.6, 30.3 and 24.0 of pattern a are singlet, whereas those of b, c (45.6 and 24.0), d (45.6, 30.3), e (45.6) and f (45.6, 30.3 and 24.0) are doublets. This difference corresponds to the change of crystal system following Fe incorporation, namely singlet peak orthorhombic and doublet monoclinic system, respectively [23].
98
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
cies (2h = 33–36°). On the contrary, an increase in the lattice volume of this sample (Fe–ZSM-5in) is obtained and however suggests the presence of some of the Fe inside zeolite channels specifically in tetrahedral positions. The value of average particles size, calculated by Scherrer equation, for the line at 2h = 23° of ZSM-5 crystallites in Fe–ZSM-5in was 75 nm, which was the lowest between all samples. This demonstrated that incorporating Fe in ZSM-5 effectively prevents particles agglomeration, allowing the materials to maintain their dispersion. This gives a clue on the anticipated strong interaction rendered between Fe and ZSM-5 in evolution of the Fe–ZSM-5in sample. Fe–ZSM-5ssaft presented the lowest degree of crystallinity between all samples and offered as well the largest particles diameter, as confirmed from r of this sample that measured ˚ ). This observed decrease in crysthe largest value (39 A tallinity may be, in part, due to the enhanced absorption of X-ray due to Fe cations. Combining the results obtained from XRD and BET investigations, one may suggest that a-Fe2O3 species might be deposited on the outermost surface layers of the zeolite blocking some pores in Fe–ZSM-5impbef, Fe–ZSM-5ssaft, and Fe–ZSM-5ssbef samples, where it is probably present outside the channels in Fe–ZSM-5in and Fe–ZSM-5impaft.
double α –Fe2O 3
(f)
(e)
Intensity/a.u.
(d)
(c)
(b)
(a) 4
10
20
30
40
50
60
2θ/degree Fig. 3. X-ray diffraction patterns of (a) ZSM-5, (b) Fe–ZSM-5in, (c) Fe–ZSM-5impbef, (d) Fe–ZSM-5impaft, (e) Fe–ZSM-5ssaft and (f) Fe– ZSM-5ssbef.
Table 2 presents lattice parameters, lattice volume and crystallinity percentages of all samples. Given that the strongest lines of ZSM-5 (2h = 22–25°) are slightly enhanced in Fe–ZSM-5in, hence, the diffusion of Fe3+ did not proceed into compensating positions in the ZSM-5 framework. Accordingly, this sample showed weak lines at 2h = 42.3° and 54.7° indicative of aFe2O3. One must emphasize that the latter lines are not the dominant ones used to appear for the same spe-
3.2.3. FTIR spectroscopy IR spectra of the zeolite lattice vibration modes and the corresponding Fe containing ones are depicted in Fig. 4 in the mid (2000–400 cm1) range. All spectra showed a typical ZSM-5 structure associated with minor changes as a result of Fe incorporation. The bands correspond to mas(T–O) at 1105 and 1227 cm1 in ZSM-5 showed shifts to lower wave numbers following Fe incorporation such as in Fe–ZSM-5in to be at 1045 and 1220 cm1, respectively. This indeed suggests the good incorporation of Fe species inside ZSM-5 zeolite channels, most likely in compensating positions. This sample also shows a shoulder at 1000 cm1 probably ascribed to the Si–O–Si linkages used to measure the extent of Si takeover in vacant sites left by Al [24], giving an evidence on exchanging framework of Si by Fe as well. Of particular importance, the mas T–O band at 1105 cm1, in parent ZSM-5, showed splitting in
Table 2 Effect of different preparation methods and thermal treatments of Fe–ZSM-5 samples on crystallites size, crystallinity and unit cell parameters ˚) ˚ )3 Sample D (nm) Cell parameters (A V (A Crystallinity (%)
ZSM-5 Fe–ZSM-5in Fe–ZSM-5impaft Fe–ZSM-5impbef Fe–ZSM-5ssaft Fe–ZSM-5ssbef
86.15 75.68 86.58 89.75 125.68 89.48
a
b
c
20.705 20.333 19.154 20.248 19.993 19.968
20.238 19.933 20.003 20.214 20.729 20.122
13.396 14.051 13.800 13.441 13.246 13.892
5613.43 5694.71 5287.27 5501.18 5489.72 5581.76
79 100 69 92 50 73
457
550 440
797 1059
(f)
1642
1219 1116 1073
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
440
1073
(e)
457
1059
(d)
545 435
1045 (b)
445
1227
1105
Absorbance/a.u.
(c)
1000
(a)
2000
1000
400
Wavenumber/cm-1 Fig. 4. FT-IR spectra of (a) ZSM-5, (b) Fe–ZSM-5in, (c) Fe–ZSM5ssaft, (d) Fe–ZSM-5ssbef, (e) Fe–ZSM-5impaft and (f) Fe–ZSM-5impbef.
Fe–ZSM-5ssbef, Fe–ZSM-5impaft and Fe–ZSM-5impbef into 1073, 1059 and 1073 cm1, respectively, besides the permanent existence of the one at 1116 cm1. This could be due to the presence of residual Al3+ nearby Fe3+ in the same site. Based on the values of the percentages of the ionic character of Fe–O and Al–O linkages that equal 54 and 64, respectively, a more covalent character of the former is indeed expected that have a superior electronegativity (1.64 vs 1.47) as well. This could give a hint about varying the relaxation effects of both
99
linkages and thus an expected higher wave numbers is more corresponding to the Fe–O linkages, i.e., 1116 cm1. On the other hand, the Fe–ZSM-5ssaft sample showed only one band at 1059 cm1 that was lower than that at 1105 cm1 (parent), indicating the higher exchange of Fe to be in compensating positions in this sample. 3.2.4. UV–Vis diffuse reflectance spectroscopy Diffuse reflectance spectra of all Fe–ZSM-5 samples (Fig. 5) were recorded in order to obtain some information about their electronic properties. Absorption bands at 210, 290 and 370 nm were observed for all Fe containing ZSM-5 samples. A band at 400 nm was also viewed for all Fe–ZSM-5 samples apart from Fe–ZSM-5impbef and Fe–ZSM-5ssaft samples. The charge transfer bands (Fe3+ O) at 210 (k < 300 nm) and 290 nm are typically attributed to isolated Fe3+ species, either tetrahedrally coordinated in the zeolite framework or with high coordination [12]. A broad band at 400 nm characteristic of Fe3+ ions present as a-Fe2O3 nanoparticles at the external surface of the zeolite crystals (P400 nm) was depicted. The band at 370 nm that can be assigned to iron ions in small oligonuclear clusters [25] was enhanced in Fe–ZSM-5impbef and Fe–ZSM-5ssaft probably at the expense of that at 400 nm. This might be due to dislodgment of framework Fe3+ ions and even a certain degree of Fe association. The presence of a broad band at 400 nm in Fe–ZSM-5in, Fe–ZSM-5impaft and Fe– ZSM-5ssbef indicates that a-Fe2O3 particles are formed as a result of partial removal of Fe from framework positions. The enhancement of this band, to be as a distinct one in Fe–ZSM-5in, is very much associated with the template removal during thermal treatments. The absence of this band in other samples (Fe–ZSM-5impbef and Fe–ZSM-5ssaft) did not necessarily signify the nonexistence of such species but most probably correlated with decreased particles size of these species. It is noteworthy mentioning that the color of the samples is still in contact even following acid leaching (HCl, 0.1 M, 3 h), except Fe–ZSM-5ssbef sample that suffers a little fainting. This indicates that the majority of Fe species are in tetrahedral positions or in highly dispersed state inside zeolite channels. 3.2.5. Mo¨ssbauer spectroscopy The MS of various Fe–ZSM-5 samples measured at room temperature are shown in Fig. 6. All spectra were fitted into a broad singlet paramagnetic state, except Fe– ZSM-5ssbef that showed a symmetric sextet of low intensity together with a broad singlet. The presence of an unresolved doublet (singlet) indicates that a fraction of Fe ions in [Fe, Al] MFI is exhibiting a faster electron relaxation time due to an enhanced spin–spin interaction between Fe ions as a result of reduced Fe–Fe ion distances [26]. Analysis of MS parameters (Table 3) of
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
370
100
Fe2O3
290
400
210
(f)
Fe3+-O
Absorbance/a.u.
(e)
(d)
(c)
(b)
(a)
200
250
300
350
400
450
500
nm Fig. 5. UV–Vis spectra of (a) ZSM-5, (b) Fe–ZSM-5in, (c) Fe–ZSM5impaft, (d) Fe–ZSM-5impbef, (e) Fe–ZSM-5ssaft and (f) Fe–ZSM-5ssbef.
former samples revealed that Fe3+ are in fourfold coordination state and are homogeneously well distributed providing an evidence for their location in framework T sites. The decreased IS values (>0.3 mm/s) of the samples highly emphasize the above assignment. In addition, the quality of the fitting of the data is not good enough to nullify the presence of other possibilities for Fe to have other configurations. On the other hand, the existence of Fe3+ in octahedral coordination in Fe–ZSM-5ssbef, besides the tetrahedral one; characteristics of Fe3+ in tetrahedral sites (IS = 0.25, QS = 0.2 mm/s), was found to have the parameters of a-
Fig. 6. Room temperature Mo¨ssbauer spectra of (a) Fe–ZSM-5in, (b) Fe–ZSM-5impaft, (c) Fe–ZSM-5impbef, (d) Fe–ZSM-5ssaft and (e) Fe– ZSM-5ssbef.
Fe2O3 species (IS = 0.3, QS = 0.2 mm/s, Heff = 520 kOe) [27]. Based on the value of Heff, particles size close to 2.0 nm can be estimated from these superparamagnetic iron oxide particles; the so-called nanoparticles. It can be seen that a maximum lattice imperfection was perceived for Fe3+ in Fe–ZSM-5ssaft sample when Fe occupied internal surface (tetrahedral sites) as conceived from the LW data that were the biggest between all samples. These were indeed responsible for the decreased crystallinity of this sample (50%) that offered as well the largest crystallites size between all samples, as illustrated from XRD investigations. Thus, one can presume that during template burning, Fe3+ occupying tetrahedral positions in Fe–ZSM-5ssaft can suffer
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
101
Table 3 State of Fe in all Fe–ZSM-5 samples as characterized from Mo¨ssbauer spectroscopy Samples
Isomer shift (IS) (mm/s)
Fe–ZSM-5in Fe–ZSM-5impaft Fe–ZSM-5impbef Fe–ZSM-5ssaft
0.239 0.261 0.26 0.23
Fe–ZSM-5ssbef Doublets Sextets Field 518.6a
0.25 0.37
a
Quadruple splitting (QS) (mm/s)
Full width at half maximum (LW) (mm/s)
0.2 0.2 0.2 0.2
0.74 0.65 0.63 0.74
0.2 0.2
0.56 0.29
518.6 is the mean hyperfine field; Heff.
dislodgment from the structure. This suggests that a fraction of framework Fe has migrated to extra-framework positions during calcinations that induce a great loss in zeolite water, allowing Fe ions to move freely and closer to other Fe ions, in both framework and non-framework positions. The precipitation of neutral ion compounds such as a-Fe2O3 in Fe–ZSM-5ssbef that was not thermally activated, could be due to the consequence of forming Bro¨nsted acid sites [28] which resulted from freeing H+ ions produced from H2O and indeed not compensated by NO3 ions [Fe(NO3)3 ! Fe3+ + 3NO3]. However, this compensation process takes place during elevation of temperature more easily probably for increasing the ions mobility. The higher relative weights of singlet components in all Fe–ZSM-5 samples in comparison with the corresponding Fe–ZSM-5ssbef could be an evidence for the higher iron oxide dispersion in the former samples. Interestingly, Fe–ZSM-5ssbef lost only 7% of its pore volume. This could indicate that the majority of a-Fe2O3 are present in the outer surface of ZSM-5 crystals. On the other hand, the Fe–ZSM-5ssaft indicates a gain in Vp comprising 50% compared with ZSM-5, reflecting the accessibility of ZSM-5 sample in opening some pores during Fe incorporation, affecting severely the crystallinity of this sample. The increase in quadrupole splitting of Fe–ZSM-5impbef can be interpreted as an increase in electron spin interaction caused by the movement of extra-framework Fe closer to other iron in both framework and extra-framework positions. The increase in quadrupole splitting is also likely due to the presence of disturbed octahedrally coordinated Fe ions in the extra-framework positions. Nevertheless, since the average isomer shift of this sample (0.25 mm/s) is close to that of Fe–ZSM-5in (0.24 mm/s), the majority of iron ions (>80%) are still tetrahedrally coordinated in isolated framework positions.
nated in the MFI framework. This is evidenced from the reported increase in lattice volume and decreased crystallites size of this material that is taken as a criterion for the devoted strong interaction aroused between Fe and ZSM-5. This sample showed the largest SBET between all samples, thus it can be chosen in surface driven applications. Fe3+ ions did not completely exchange Al3+ in the structure but rather showed the existence of both ions and possibly in the same sites. Hardly detected nanoparticles of a-Fe2O3 were depicted in this sample as confirmed from XRD and UV–Vis investigations in addition to small amounts of small clusters of Fe3þ x O species. Fe species depicted on Fe–ZSM-5imp either calcined or not were varied in the sense that Fe–ZSM-5impbef did not indicate a-Fe2O3 particles in the electronic spectra and rather measured the lowest SBET (361 m2/g). This did not nullify the presence of the latter species that could be of smaller sizes to be detected by the UV–Vis technique. Most probably, these species are located in the pores of this sample and thus could be responsible for the devoted decrease in the specific surface area of this sample. This could give a criterion about the dislodgment of Fe3+ from framework to extraframework as a function of the thermal effect, which in its turn facilitates the migration of Fe3+ as isolated a-Fe2O3 through water elimination or with the interaction between resid¨ nsted sites [Fe(OH)Si]. ual BrO Calcination of the as-synthesized Fe–ZSM-5ssaft material presented more or less typical Fe species as Fe–ZSM-5ssbef except the absence of a-Fe2O3 species that has been ascertained for the latter by Mo¨ssbauer and UV–Vis investigations. Surface texture of both samples presented varied properties in the sense that Fe–ZSM-5ssbef owned high microporosity (84%) where the other acquired the maximum mesoporosity (40%) between all samples.
4. Conclusion
References
The majority of Fe cations in the as-synthesized Fe–ZSM-5in zeolites calcined at 550 °C are present as paramagnetic Fe3+ ions that are tetrahedrally coordi-
[1] A. Ribera, I.W.C.E. Arends, S. de Vries, J. Perez-Ramirez, R.A. Sheldon, J. Catal. 195 (2000) 287. [2] J.B. Taboada, A.R. Overweg, P.J. Kooyman, I.W.C.E. Arends, G. Mul, J. Catal. 231 (2005) 56.
102
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
[3] J. Perez-Ramirez, G. Mul, F. Kapteijn, J.A. Moulijn, A.R. Overweg, A. Domenech, A. Ribera, I.W.C.E. Arends, J. Catal. 207 (2002) 113. [4] M. Schwidder, M.S. Kumar, K. Klementiev, M.M. Pohl, A. Bruckner, W. Grunert, J. Catal. 231 (2005) 314. [5] M.M. Mohamed, J. Mol. Catal. 200 (2003) 301. [6] M.M. Mohamed, F. Abdel-Hai, J. Mol. Catal. 211 (2004) 199. [7] P.B. Venuto, Micropor. Mater. 2 (1994) 297. [8] Md. Uddin, T. Komatsu, T. Kashima, J. Catal. 150 (1994) 439. [9] R.Q. Long, R.T. Yang, J. Catal. 207 (2002) 224. [10] X. Feng, W.K. Hall, Catal. Lett. 41 (1996) 45. [11] M.M. Mohamed, N.A. Eissa, Mater. Res. Bull. 38 (2003) 1993. [12] J. Perez-Ramirez, J. Catal. 227 (2004) 512. [13] E.J.M. Hensen, Q. Zhu, M.M.R.M. Hendrix, A.R. Overweg, P.J. Kooyman, M.V. Sychev, R.A. Van Santen, J. Catal. 221 (2004) 560. [14] H.Y. Chen, W.M.H. Sachtler, Catal. Today 42 (1998) 73. [15] G.I. Panov, Cattech 4 (2000) 18. [16] J. Perez-Ramirez, E.V. Kondratenko, Chem. Commun. (2003) 2152. [17] K.S.W. Sing, D.H. Everett, R.A.W. Haul, I. Moscou, R.A. Pierotti, J. Rouqueroi, T. Siemieniewsk, Pure Appl. Chem. 57 (1985) 603.
[18] F. Rouquerol, J. Rouquerol, K.S.W. Sing, in: F. Schutz, K.S.W. Sing, J. Weitkamp (Eds.), Handbook of Porous Materials, vol. 1, Wiley-VCH, Weinhiem, 2002, p. 250. [19] M.M. Mohamed, T.M. Salama, I. Othman, G.A. El-Shobaky, Appl. Catal. A 279 (2005) 23. [20] T.M. Salama, M.M. Mohamed, I. Othman, G.A. El-Shobaky, Appl. Catal. A 286 (2005) 85. [21] M.M. Mohamed, T.M. Salama, I. Othman, E. Abdalla, Micropor. Mespor. Mater., in press. [22] J. Zhao, Z. Feng, F.E. Huggins, N. Shah, G.P. Huffman, I. Wener, J. Catal. 148 (1994) 194. [23] E.L. Wu, S.L. Lawton, D.H. Olson, A.C. Rohrman Jr., G.T. Kokotailo, J. Phys. Chem. 83 (1979) 2777. [24] S. Bordiga, A. Damin, G. Berlier, F. Bonino, G. Ricchiardi, A. Zecchina, C. Lamberti, Int. J. Mol. Sci. 2 (2001) 167. [25] G. Lehmann, Z. Phys. Chem. Neue. Folge 27 (1970) 279. [26] C. Minchev, H. Huwe, T. Tsoncheva, D. Paneva, M. Dimitrov, I. Mitor, M. Froba, Micropor. Mesopor. Mater. 81 (2005) 333. [27] C.D.-M. Assou, J.L. Dormann, O. Gorochov, D. Svoronos, Hyperfine Interact. 28 (1986) 915. [28] X. Feng, W.K. Hall, J. Catal. 166 (1997) 386.
Effect of thermal treatment on surface and bulk properties of Fe/ZSM-5 zeolites prepared by different methods Mohamed Mokhtar Mohamed
a,*
, Ibraheem Othman Ali b, N.A. Eissa
c
a
b
Department of Chemistry, Faculty of Science, Benha University, Benha, Cairo 650371, Egypt Department of Chemistry, Faculty of Science, Al-Azhar University, Nasr City, Cairo 1184, Egypt c Department of Physics, Faculty of Science, Al-Azhar University, Nasr City, Cairo 1184, Egypt Received 11 June 2005; received in revised form 23 July 2005; accepted 23 July 2005 Available online 9 September 2005
Abstract Fe/ZSM-5 samples (5 wt.% Fe) prepared by in situ incorporation using TPABr template under hydrothermal conditions (Fe– ZSM-5in), chemical liquid deposition (Fe–ZSM-5imp) and solid–solid (Fe–ZSM-5ss) interaction were characterized by N2 physisorption, TG/DSC, X-ray diffraction, FTIR spectroscopy, UV–Vis diffuse reflectance spectroscopy and 57Fe Mo¨ssbauer spectroscopy techniques. Calcination at 550 °C leads to almost complete removal of template that was associated with dislodgment of significant fraction of Fe to external positions as recognized for the in situ prepared sample (Fe–ZSM-5in). This sample showed an increase in lattice volume suggesting the presence of the majority of Fe ions in tetrahedral positions inside zeolite channels and offered as well the lowest crystallites size (75 nm) and maximum SBET (453 m2/g) between all samples. On the other hand, Fe–ZSM-5ssbef, resulting ˚ ), from solid–solid interaction and subjected to heat treatment in vacuum at 200 °C, measured the lowest mean pore radius (r; 23 A and pore volume (Vp; 0.3887 cm3/g), giving a hint about the probability of finding neutral iron oxide nanoparticles (a-Fe2O3) as a separate phase that has been validated by Mo¨ssbauer (IS = 0.3 mm/s, QS = 0.2 mm/s, Heff = 520 kOe) and UV–Vis (400 nm) investigations. This sample also demonstrated that the majority of Fe occupied framework positions beside a fraction identified as small oligonuclear oxo-iron ions (Fe3þ x –O; 290 nm). Interestingly, Fe–ZSM-5ssaft, resulting from solid–solid interaction and sub˚ ) values, reflecting the enforced location of Fe jected to air calcination at 550 °C, measured maximum Vp (0.6380 cm3/g) and r (39 A in this sample leading to an effective pore widening and thus a pronounced mesoporosity is attained. IR bands due to masT–O in ZSM-5 (1105 cm1) showed a shift to lower wave numbers (1059 cm1) following Fe incorporation reflecting the extent of exchanging Fe in this sample (Fe–ZSM-5ssaft), unlike the rest of the samples which showed splitting, which accounted for the presence of residual Al3+ beside Fe3+ ions in the same site. Mo¨ssbauer data of this sample confirmed the latter result and indicated the maximum lattice imperfection and showed as well the lowest degree of crystallinity. The Fe–ZSM-5impaft sample, subjected to heat treatment at 550 °C, showed a-Fe2O3 species where that heated at 110 °C presented the lowest SBET (361 m2/g). More correlations were evaluated and discussed on the effect of thermal treatment on the existence of various Fe species (either framework or nonframework), their electronic states and local structures. Ó 2005 Published by Elsevier Inc. Keywords: Fe–ZSM-5; In situ preparation; Impregnation; Solid–solid interactions; Characterization; Fe species; Mo¨ssbauer spectroscopy
1. Introduction
*
Corresponding author. Tel./fax: +966 025274264. E-mail address: [email protected] (M.M. Mohamed).
1387-1811/$ - see front matter Ó 2005 Published by Elsevier Inc. doi:10.1016/j.micromeso.2005.07.032
Zeolites with 10-membered oxygen rings normally possess a high siliceous framework structure. The 10-membered oxygen ring zeolites also possess other important characteristic properties including high
94
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
activity, high tolerance to coking and high hydrothermal stability. Among the family of 10-membered oxygen ring zeolites, the MFI type (ZSM-5) zeolite is probably the most useful one. It has been acknowledged that the tetrahedral surrounding could be stable for cations with 0.414 > q > 0.225 (q = rM/rO, where rM and rO are the radii of the cation and the oxygen ion, respectively). Cations that do not fulfill this criterion such as Fe3+ may also be introduced into the framework under specific synthesis conditions. The introduction into the framework of ZSM-5 atoms other than Si and Al can modify significantly the acid–base and catalytic properties of the zeolites. Some modification studies have been made with framework iron insertion substituting aluminium in the synthesis process [1–4]. Such isomorphous substitution of Fe either for Al or Si of zeolites has been reported to exhibit improved catalytic properties in a variety of chemical reactions. Among them are ethanol conversion to hydrocarbons, polymerization, isomerization, dehydrogenation, reduction of NO either by NH3 or hydrocarbons and selective oxidations [5– 11]. For most of these reactions, the catalytic activity is thought to be related to extra-framework iron species (iron oxides, dinuclear iron sites and oxohydroxides aggregates) formed during ion exchange and solid–solid treatments [12,13]; those acting either independently or in synergy with protonic-Fe(OH)Si sites (Bro¨nsted acid). Fe exchanged ZSM-5 zeolite exhibited a high and stable catalytic activity in several selective redox reactions of great importance: NO [14], N2O [15] decomposition as well as selective catalytic reduction of NO by hydrocarbons in an excess of O2 [16]. On the other hand, CuZSM-5 showed lower catalytic activity and less stability towards NO reduction due to the alteration of the copper ionic state on the catalyst surface and to dealumination consequences. Accordingly, Fe–ZSM-5 is subjected to numerous studies principally for correlating the structure of the active site(s) in various applications. Thus, an intense attention still attracting many researchers to study structure–activity relationship for each active Fe site can be revealed. From the analysis of the literature data briefly discussed above, it is however evident that the formation of framework and non-framework Fe–ZSM-5 is primarily connected with the conditions applied during preparation and to the applied thermal treatment that is far from being completely understood. In this essence, the coordination and oxidation changes of Fe3+ species when incorporated with ZSM-5 were followed before and after calcination for samples prepared by different methods, including in situ incorporation during synthesis, impregnation and solid–solid interactions in conjunction with their electronic states, surface properties and local structures.
2. Experimental 2.1. Materials The materials used were: silicic acid powder, sodium hydroxide pellets (A.R. 98%), aluminum sulfate (Al2 (SO4)3 Æ 16H2O Merck), iron nitrate (Fe(NO3)3 Æ 9H2O Merck), tetrapropylammonium bromide (TPABr, Fluka), n-propyl amine (Merck) and concentrated sulfuric acid. 2.1.1. Preparation of ZSM-5 The hydrogels of the following oxide molar compositions were prepared for the synthesis of ZSM-5 zeolite: 3:12Na2 O : 0:162Al2 O3 : 6:185SiO2 : 0:185Q : 100H2 O where Q is the tetrapropylammonium bromide template. A specific amount of NaOH was added to silicic acid in a small amount of H2O (40 ml) while stirring, followed by heating at 80 °C until a clear solution was reached. The TPABr was dissolved in a little amount of H2O (10 ml) with heating at 50 °C for 20 min. The solution of TPABr was added to that of sodium silicate solution whilst stirring for 15 min. In addition, 1 ml of n-propyl amine was added as a mobilizing agent to that mixture. The aluminum sulfate, on the other hand, was dissolved in a small amount of H2O (10 ml) while adding 0.05-ml concentrated H2SO4 with stirring until reaching a clear solution. To the latter solution, the combined solution of sodium silicate and TPABr was added followed by stirring for 30 min. The pH of the mixture was adjusted at 11 by using NaOH (0.1 M) and H2SO4 (0.1 M) solutions. Finally, the mixture was hydrothermally treated at 160 °C in an oil bath, using stainless steel autoclaves, for 6 days. The autoclaves were removed at the specified time from the oil bath and quenched immediately with cold water. The solid product was filtered and washed with distilled water until the pH of the filtrate dropped to 8. The products were dried at 110 °C for 10 h, then calcined at 550 °C for 6 h in an air oven. 2.1.2. Preparation of Fe–ZSM-5 2.1.2.1. Inclusion of Fe–ZSM-5 zeolite (built-in). The synthesis procedure is described in molar compositions as follows: 3:12Na2 O : 0:138Al2 O3 : 6:185SiO2 : 0:0103Fe2 O3 : 0:185Q : 100H2 O where Q is the tetrapropylammonium bromide. Aluminum sulfate was dissolved in a small amount of water (5 ml) followed by addition of 0.05-ml H2SO4. On the other hand, the soluble complex of sodium silicate was formed onto which the template tetrapropylammonium bromide (TPABr) was added with n-propyl amine
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
as a mobilizing agent. On the other hand, the preparation of Fe(NO3)3 solution was carried out. A simultaneous addition of iron nitrate solution and sodium silicate solution into the aluminum sulfate was achieved, followed by stirring for 30 min. Sulfuric acid (0.1 M) or sodium hydroxide (0.1 M) adjusted the pH to 11. The amorphous gel formed was allowed to age for 1 h at room temperature. The reaction mixture was transferred to 300-ml stainless steel autoclaves and maintained in an oil bath at 160 °C under autogeneous pressure. The autoclave was removed from the oil bath after 6 days and quenched in cold water for product identification. The solid products were separated by filtration. Excess alkali was thoroughly washed with water repeatedly until the pH of washing liquid was close to 8, and the products were dried in an oven at 120 °C for 10 h, then calcined at 550 °C for 6 h to remove the template. This sample was referred to as Fe–ZSM-5in. The color of this sample was pale brown. 2.1.2.2. Preparation of Fe–ZSM-5 by CLD (chemical liquid deposition). The iron oxide in zeolites was also prepared by impregnation of ZSM-5 with an aqueous solution of iron nitrate to achieve a loading of 5 wt.% iron oxide (0.0082 molar ratio). The volume of solution used was that necessary to completely wet the zeolite sample. The reaction temperature was at 80 °C for 4 h by using temperature controller type REX-P 90. After impregnation, the water was slowly allowed to evaporate at 110 °C till dryness, and then the precursor sample was either dried at the same temperature for 6 h or calcined at 550 °C in an air oven for 6 h. The samples were referred to as Fe–ZSM-5impbef and Fe–ZSM-5impaft, respectively. The color of the samples was reddishbrown. 2.1.2.3. Preparation of Fe–ZSM-5 by solid–solid interaction. The ZSM-5 sample was degassed at 200 °C for 3 h to remove physisorbed water molecules prior to being physically mixed with iron nitrate (5 wt.% iron oxide (0.0082 as molar ratio) at room temperature in a glove bag filled with nitrogen gas (99% purity). The mechanically mixed sample was stirred overnight in a current of pure nitrogen flowing at a rate of 30 cm3/min. The obtained sample either heated at 200 °C under a reduced pressure of 105 Torr for 4 h or calcined at 550 °C in air for 6 h was referred to as Fe–ZSM-5ssbef and Fe– ZSM-5ssaft, respectively. The color of the Fe–ZSM-5ssbef sample was reddish-brown, whereas the Fe–ZSM-5ssaft sample obtained a brown color that converted to reddish-brown with time. 2.2. Experimental techniques The X-ray diffractograms of various zeolitic samples were measured by using a Philips diffractometer (type
95
PW 3710). The patterns were run with Ni-filtered copper ˚ ) at 30 kV and 10 mA with a radiation (k = 1.5404 A scanning speed of 2h = 2.5°/min. The crystal sizes of the prepared materials were determined using the Scherrer equation. The instrumental line broadening was measured using a LaB6 standard. The crystallinity of the prepared samples was calculated using the ratio of the sum of the areas of the most intense peaks for Fe–ZSM-5 samples (2h = 20–25°) to that of the same peaks for the standard (Na–ZSM-5 Mobil chemicals) and multiplying by 100. FT-IR spectra of the samples were recorded with a JASCO single beam FT-IR 5300-spectrometer with 50 co-added scans at 2 cm1 resolution. All IR measurements were carried out at room temperature using KBr technique. Optical spectra in the UV–Vis region were recorded using diffuse reflectance technique. The spectrometer was a Shimadzu reflectospectrometer type UV-2200A. The reflectance white standard was BaSO4. Thermogravimetric and differential scanning calorimetry (TG/DSC) were carried out using Shimadzu-50 thermal analyzer units. The sensitivity of TG and DSC measurements was 0.01 mg and 25 lv, respectively. In each run, about 10 mg of uncalcined sample was heated from room temperature to 1000 °C at a heating rate of 10 °C/min in a current of N2 flowing at a rate of 30 ml/min. The nitrogen adsorption isotherms were measured at 196 °C using a conventional volumetric apparatus. The specific surface area was obtained using the BET method. The micropore volume and the external surface area were obtained from the t-plot method. The Mo¨ssbauer spectra (MS) were recorded at room temperature using a constant acceleration spectrometer. The source activity was 3.70 GB9 of 57Co in Co matrix. Iron metal was used as a reference for the isomer shift (IS). The MS were fitted using a program based on the distribution of hyperfine magnetic field (HMF) and quadrupole splitting (QS).
3. Results and discussion 3.1. Surface texture The different surface characteristics of various investigated zeolitic materials were determined from N2 adsorption/desorption isotherms conducted at 77 K. These characteristics include BET-surface area (SBET), total pore volume (Vp), mean pore radius (r), surface area derived from Vlt plots (St), surface area of micropores (Sl), external surface area (Sext), surface area of wide pores (Swid), volume of micropores ðV lp Þ, volume of wide pores ðV wid p Þ and microporosity percentages (Table 1). The N2 adsorption/desorption isotherms for
96
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
Table 1 Some Surface characteristics of ZSM-5 and iron containing ZSM-5 investigated adsorbents; prepared by different methods, heating at 300 °C under a reduced pressure of 105 Torr Sample
SBET (m2/g)
St (m2/g)
V total p (cm3/g)
r ˚) (A
Sl (m2/g)
Sext (m2/g)
Swid (m2/g)
V lp (cm3/g)
V wid p (cm3/g)
Microporosity (%)
C-const.
ZSM-5 Fe–ZSM-5in Fe–ZSM-5ssaft Fe–ZSM-5ssbef Fe–ZSM-5impaft Fe–ZSM-5impbef
429 453 408 422 437 361
420 461 371 391 416 329
0.4166 0.6551 0.6380 0.3887 0.5430 0.3624
24 36 39 23 31 25
422 284 246 353 326 286
74 100 69 63 88 54
7 169 162 69 111 36
0.4054 0.4090 0.3843 0.3249 0.4045 0.2866
0.0112 0.2461 0.25375 0.06377 0.13846 0.07576
97 62 60 84 75 79
21 21 26 10 8 10
Note: BET-surface area (SBET); total pore volume (Vp); mean pore radius (r); surface area derived from V1t plots (St); surface area of micropores (Sl); external surface area (Sext); surface area of wide pores (Swid); volume of micropores ðV lp Þ; volume of wide pores ðV wid p Þ and microporosity percentages.
Fe–ZSM-5 samples (not shown) were typically type 1 [17] in nature. Although the validity of the BET model for highly microporous materials is questionable, the surface area derived from this model in the adapted pressure range P/P0 = 0.01–0.1 can still be used for comparative purposes [18]. At high P/P0, some deviation from type 1 isotherm was observed for some samples along with desorption hysteresis. This may be due to intracrystalline mesopores depicted from Vlt plots (Fig. 1) for all Fe containing ZSM-5 samples; except Fe–ZSM-5ssaft, those constructed depended on the value of C-constant in the BET equation. The existence of a downward deviation in the latter sample is indicative of the domination of narrow pores. Inspection of the results in Table 1 reveals that: (i) the values of SBET and St are close to each other for various investigated zeolitic samples and thus justify the correct choice of standard t curves used in pore analysis and rather indicate the absence of the ultra-micropores in these materials. The SBET of Fe–ZSM-5impbef measured the lowest specific area between all samples that comprises 16% from that of the parent. This can be due to the occurrence of some Fe species inside zeolite pores and thus blocking them. (ii) The values of external surface area of various samples constitute high values (>10%) of the SBET pointing to the porous nature of these materials and rather suggest the appreciable presence of mesopores that concurrently emphasized from ˚ ). The latter finding was conhigh r values (24–39 A firmed by both the slope position rendered for samples instead of the almost horizontal saturation plateau (characteristics of an ideal microporous structure) and upward deviations depicted for majority of the samples (Fig. 1). An increase in the mesopores area of some samples (see Sext in Table 1) could be due to the consequence of extraction of heteroatoms (Al, Fe) to non-framework positions. (iii) The specimen designated as Fe–ZSM-5ssaft measured the maximum Vp and r. This might indicate the enforced location of Fe species in the pores of this particular sample leading to an effective pore widening. Conversely, the Fe–ZSM-5ssbef sample measured the
˚ ), giving a hint about the opportulowest r value (23 A nity of finding Fe species as a separate phase and thus indicates a lower value of Vp comparatively. (iv) The percentage microporosity in Fe–ZSM-5ssbef measured the highest value (84%) between all samples, as rather confirmed from Vlt plot of the sample that verified only the presence of narrow pores as a consequence of the existence of downward deviations. (v) The observed increase in the BET surface area of Fe–ZSM-5in (5%) when compared with ZSM-5 can be attributed either to the creation of new pores due to departure of NOx compounds during the thermal dissociation of iron nitrate or to the formation of Fe2O3 species outside zeolite pores probably as a result of the template removal during thermal treatment. The present result might recommend using Fe–ZSM5in in surface driven applications, since this sample is shown to withstand surface accessibility (an increase of 5% in SBET, Vp max and r max) and contain both microporous and mesoporous characters. This result indicates as well the enforced location of Fe species in the pores of ZSM-5in specimen and specifically in compensation positions (tetrahedral sites) as will be visualized later by XRD, IR and Mo¨ssbauer investigations. 3.2. Material bulk characteristics 3.2.1. Thermal analysis Fig. 2 shows TG and DSC curves of uncalcined ZSM-5 synthesized using TPABr while being heated under an argon atmosphere. The weight loss occurs in three steps: the first between 25 and 400 °C due to desorption of water either present on external or internal surfaces and comprises a weight loss of 0.8%, the second step between 420 and 520 °C, due to thermal decomposition of TPA+ ions occluded in various types of pores and comprises a weight loss of 6.4% and the third, between 520 and 900 °C that comprises 0.3% weight loss and is ascribed to elimination of H2O during condensation of surface OH groups and/or removal of chemically bound H2O. DSC, on the other hand, shows
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
97
1. 5 300
1. 0
TG-DSC (f)
0.0 300
(e)
200
608
Enthalpy/µv.s/mg
100
0.0 -1.0 -2.0 -3.0 -4.0 -5.0 470
200
TG/%
DTG/%/min
0.0
5.0 0.0
-2
100
-1.0 -4
0.0
-2.0 (d)
200
-6 -3.0
Vacc (g/ml)
-8 25 100
200
400
600
800
1000
Temperature/ºC Fig. 2. TGA/DSC curves of synthesized ZSM-5 zeolite.
0.0 400 (c)
300 200 100 0.0 400
(b)
300 200
the 608 peak to elimination of H2O during condensation of surface OH groups is unlikely but indeed could be due to thermal decomposition of TPA+ molecules occluded in narrow pores [19]. This is confirmed from lowering the total mass loss of TPA+ (6.4%) than that of the calculated value (11.7%) [20] assuming four TPA+ cations per unit cell in a pure ZSM-5 crystal. Likewise, similar results can be obtained from the DTG curves for the other ZSM-5 zeolite and may correspond to the relatively strong interactions between anionic inorganic framework and cationic organic species located at different sites of the ZSM-5 framework [21].
100 0.0 400 (a)
300 200 100 0.0 0.0
5
10
15
20
P/Po Fig. 1. V1t plots of (a) synthesized ZSM-5, (b) Fe–ZSM-5in, (c) Fe– ZSM-5impaft, (d) Fe–ZSM-5impbef, (e) Fe–ZSM-5ssaft and (f) Fe–ZSM5ssbef.
four endothermic peaks at 100, 470, 608 and 875 °C. The third weight loss which occurred in the 520–900 °C range (TG curve) includes the two endothermic peaks positioned at 608 and 875 °C. Accordingly, ascribing
3.2.2. Crystalline composition XRD patterns of Fe–ZSM-5 samples are displayed in Fig. 3 together with parent ZSM-5 synthesized from TPABr template. The XRD patterns of all iron-containing ZSM-5 showed the typical spectrum of MFI structure. The samples were well crystalline, and a few of a-Fe2O3 phase which were hardly recognized as small ˚ ), 34.1 (2.39 A ˚ ), 42.3 particles at 2h = 24.8 (3.69 A ˚ ˚ (2.26 A) and 54.7 (1.71 A) agreed with those for aFe2O3 phase with d-spacings = 3.67, 2.42, 2.20 and ˚ , respectively [22]. The diffraction lines of ZSM1.69 A 5 around 2h = 45.6, 30.3 and 24.0 of pattern a are singlet, whereas those of b, c (45.6 and 24.0), d (45.6, 30.3), e (45.6) and f (45.6, 30.3 and 24.0) are doublets. This difference corresponds to the change of crystal system following Fe incorporation, namely singlet peak orthorhombic and doublet monoclinic system, respectively [23].
98
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
cies (2h = 33–36°). On the contrary, an increase in the lattice volume of this sample (Fe–ZSM-5in) is obtained and however suggests the presence of some of the Fe inside zeolite channels specifically in tetrahedral positions. The value of average particles size, calculated by Scherrer equation, for the line at 2h = 23° of ZSM-5 crystallites in Fe–ZSM-5in was 75 nm, which was the lowest between all samples. This demonstrated that incorporating Fe in ZSM-5 effectively prevents particles agglomeration, allowing the materials to maintain their dispersion. This gives a clue on the anticipated strong interaction rendered between Fe and ZSM-5 in evolution of the Fe–ZSM-5in sample. Fe–ZSM-5ssaft presented the lowest degree of crystallinity between all samples and offered as well the largest particles diameter, as confirmed from r of this sample that measured ˚ ). This observed decrease in crysthe largest value (39 A tallinity may be, in part, due to the enhanced absorption of X-ray due to Fe cations. Combining the results obtained from XRD and BET investigations, one may suggest that a-Fe2O3 species might be deposited on the outermost surface layers of the zeolite blocking some pores in Fe–ZSM-5impbef, Fe–ZSM-5ssaft, and Fe–ZSM-5ssbef samples, where it is probably present outside the channels in Fe–ZSM-5in and Fe–ZSM-5impaft.
double α –Fe2O 3
(f)
(e)
Intensity/a.u.
(d)
(c)
(b)
(a) 4
10
20
30
40
50
60
2θ/degree Fig. 3. X-ray diffraction patterns of (a) ZSM-5, (b) Fe–ZSM-5in, (c) Fe–ZSM-5impbef, (d) Fe–ZSM-5impaft, (e) Fe–ZSM-5ssaft and (f) Fe– ZSM-5ssbef.
Table 2 presents lattice parameters, lattice volume and crystallinity percentages of all samples. Given that the strongest lines of ZSM-5 (2h = 22–25°) are slightly enhanced in Fe–ZSM-5in, hence, the diffusion of Fe3+ did not proceed into compensating positions in the ZSM-5 framework. Accordingly, this sample showed weak lines at 2h = 42.3° and 54.7° indicative of aFe2O3. One must emphasize that the latter lines are not the dominant ones used to appear for the same spe-
3.2.3. FTIR spectroscopy IR spectra of the zeolite lattice vibration modes and the corresponding Fe containing ones are depicted in Fig. 4 in the mid (2000–400 cm1) range. All spectra showed a typical ZSM-5 structure associated with minor changes as a result of Fe incorporation. The bands correspond to mas(T–O) at 1105 and 1227 cm1 in ZSM-5 showed shifts to lower wave numbers following Fe incorporation such as in Fe–ZSM-5in to be at 1045 and 1220 cm1, respectively. This indeed suggests the good incorporation of Fe species inside ZSM-5 zeolite channels, most likely in compensating positions. This sample also shows a shoulder at 1000 cm1 probably ascribed to the Si–O–Si linkages used to measure the extent of Si takeover in vacant sites left by Al [24], giving an evidence on exchanging framework of Si by Fe as well. Of particular importance, the mas T–O band at 1105 cm1, in parent ZSM-5, showed splitting in
Table 2 Effect of different preparation methods and thermal treatments of Fe–ZSM-5 samples on crystallites size, crystallinity and unit cell parameters ˚) ˚ )3 Sample D (nm) Cell parameters (A V (A Crystallinity (%)
ZSM-5 Fe–ZSM-5in Fe–ZSM-5impaft Fe–ZSM-5impbef Fe–ZSM-5ssaft Fe–ZSM-5ssbef
86.15 75.68 86.58 89.75 125.68 89.48
a
b
c
20.705 20.333 19.154 20.248 19.993 19.968
20.238 19.933 20.003 20.214 20.729 20.122
13.396 14.051 13.800 13.441 13.246 13.892
5613.43 5694.71 5287.27 5501.18 5489.72 5581.76
79 100 69 92 50 73
457
550 440
797 1059
(f)
1642
1219 1116 1073
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
440
1073
(e)
457
1059
(d)
545 435
1045 (b)
445
1227
1105
Absorbance/a.u.
(c)
1000
(a)
2000
1000
400
Wavenumber/cm-1 Fig. 4. FT-IR spectra of (a) ZSM-5, (b) Fe–ZSM-5in, (c) Fe–ZSM5ssaft, (d) Fe–ZSM-5ssbef, (e) Fe–ZSM-5impaft and (f) Fe–ZSM-5impbef.
Fe–ZSM-5ssbef, Fe–ZSM-5impaft and Fe–ZSM-5impbef into 1073, 1059 and 1073 cm1, respectively, besides the permanent existence of the one at 1116 cm1. This could be due to the presence of residual Al3+ nearby Fe3+ in the same site. Based on the values of the percentages of the ionic character of Fe–O and Al–O linkages that equal 54 and 64, respectively, a more covalent character of the former is indeed expected that have a superior electronegativity (1.64 vs 1.47) as well. This could give a hint about varying the relaxation effects of both
99
linkages and thus an expected higher wave numbers is more corresponding to the Fe–O linkages, i.e., 1116 cm1. On the other hand, the Fe–ZSM-5ssaft sample showed only one band at 1059 cm1 that was lower than that at 1105 cm1 (parent), indicating the higher exchange of Fe to be in compensating positions in this sample. 3.2.4. UV–Vis diffuse reflectance spectroscopy Diffuse reflectance spectra of all Fe–ZSM-5 samples (Fig. 5) were recorded in order to obtain some information about their electronic properties. Absorption bands at 210, 290 and 370 nm were observed for all Fe containing ZSM-5 samples. A band at 400 nm was also viewed for all Fe–ZSM-5 samples apart from Fe–ZSM-5impbef and Fe–ZSM-5ssaft samples. The charge transfer bands (Fe3+ O) at 210 (k < 300 nm) and 290 nm are typically attributed to isolated Fe3+ species, either tetrahedrally coordinated in the zeolite framework or with high coordination [12]. A broad band at 400 nm characteristic of Fe3+ ions present as a-Fe2O3 nanoparticles at the external surface of the zeolite crystals (P400 nm) was depicted. The band at 370 nm that can be assigned to iron ions in small oligonuclear clusters [25] was enhanced in Fe–ZSM-5impbef and Fe–ZSM-5ssaft probably at the expense of that at 400 nm. This might be due to dislodgment of framework Fe3+ ions and even a certain degree of Fe association. The presence of a broad band at 400 nm in Fe–ZSM-5in, Fe–ZSM-5impaft and Fe– ZSM-5ssbef indicates that a-Fe2O3 particles are formed as a result of partial removal of Fe from framework positions. The enhancement of this band, to be as a distinct one in Fe–ZSM-5in, is very much associated with the template removal during thermal treatments. The absence of this band in other samples (Fe–ZSM-5impbef and Fe–ZSM-5ssaft) did not necessarily signify the nonexistence of such species but most probably correlated with decreased particles size of these species. It is noteworthy mentioning that the color of the samples is still in contact even following acid leaching (HCl, 0.1 M, 3 h), except Fe–ZSM-5ssbef sample that suffers a little fainting. This indicates that the majority of Fe species are in tetrahedral positions or in highly dispersed state inside zeolite channels. 3.2.5. Mo¨ssbauer spectroscopy The MS of various Fe–ZSM-5 samples measured at room temperature are shown in Fig. 6. All spectra were fitted into a broad singlet paramagnetic state, except Fe– ZSM-5ssbef that showed a symmetric sextet of low intensity together with a broad singlet. The presence of an unresolved doublet (singlet) indicates that a fraction of Fe ions in [Fe, Al] MFI is exhibiting a faster electron relaxation time due to an enhanced spin–spin interaction between Fe ions as a result of reduced Fe–Fe ion distances [26]. Analysis of MS parameters (Table 3) of
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
370
100
Fe2O3
290
400
210
(f)
Fe3+-O
Absorbance/a.u.
(e)
(d)
(c)
(b)
(a)
200
250
300
350
400
450
500
nm Fig. 5. UV–Vis spectra of (a) ZSM-5, (b) Fe–ZSM-5in, (c) Fe–ZSM5impaft, (d) Fe–ZSM-5impbef, (e) Fe–ZSM-5ssaft and (f) Fe–ZSM-5ssbef.
former samples revealed that Fe3+ are in fourfold coordination state and are homogeneously well distributed providing an evidence for their location in framework T sites. The decreased IS values (>0.3 mm/s) of the samples highly emphasize the above assignment. In addition, the quality of the fitting of the data is not good enough to nullify the presence of other possibilities for Fe to have other configurations. On the other hand, the existence of Fe3+ in octahedral coordination in Fe–ZSM-5ssbef, besides the tetrahedral one; characteristics of Fe3+ in tetrahedral sites (IS = 0.25, QS = 0.2 mm/s), was found to have the parameters of a-
Fig. 6. Room temperature Mo¨ssbauer spectra of (a) Fe–ZSM-5in, (b) Fe–ZSM-5impaft, (c) Fe–ZSM-5impbef, (d) Fe–ZSM-5ssaft and (e) Fe– ZSM-5ssbef.
Fe2O3 species (IS = 0.3, QS = 0.2 mm/s, Heff = 520 kOe) [27]. Based on the value of Heff, particles size close to 2.0 nm can be estimated from these superparamagnetic iron oxide particles; the so-called nanoparticles. It can be seen that a maximum lattice imperfection was perceived for Fe3+ in Fe–ZSM-5ssaft sample when Fe occupied internal surface (tetrahedral sites) as conceived from the LW data that were the biggest between all samples. These were indeed responsible for the decreased crystallinity of this sample (50%) that offered as well the largest crystallites size between all samples, as illustrated from XRD investigations. Thus, one can presume that during template burning, Fe3+ occupying tetrahedral positions in Fe–ZSM-5ssaft can suffer
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
101
Table 3 State of Fe in all Fe–ZSM-5 samples as characterized from Mo¨ssbauer spectroscopy Samples
Isomer shift (IS) (mm/s)
Fe–ZSM-5in Fe–ZSM-5impaft Fe–ZSM-5impbef Fe–ZSM-5ssaft
0.239 0.261 0.26 0.23
Fe–ZSM-5ssbef Doublets Sextets Field 518.6a
0.25 0.37
a
Quadruple splitting (QS) (mm/s)
Full width at half maximum (LW) (mm/s)
0.2 0.2 0.2 0.2
0.74 0.65 0.63 0.74
0.2 0.2
0.56 0.29
518.6 is the mean hyperfine field; Heff.
dislodgment from the structure. This suggests that a fraction of framework Fe has migrated to extra-framework positions during calcinations that induce a great loss in zeolite water, allowing Fe ions to move freely and closer to other Fe ions, in both framework and non-framework positions. The precipitation of neutral ion compounds such as a-Fe2O3 in Fe–ZSM-5ssbef that was not thermally activated, could be due to the consequence of forming Bro¨nsted acid sites [28] which resulted from freeing H+ ions produced from H2O and indeed not compensated by NO3 ions [Fe(NO3)3 ! Fe3+ + 3NO3]. However, this compensation process takes place during elevation of temperature more easily probably for increasing the ions mobility. The higher relative weights of singlet components in all Fe–ZSM-5 samples in comparison with the corresponding Fe–ZSM-5ssbef could be an evidence for the higher iron oxide dispersion in the former samples. Interestingly, Fe–ZSM-5ssbef lost only 7% of its pore volume. This could indicate that the majority of a-Fe2O3 are present in the outer surface of ZSM-5 crystals. On the other hand, the Fe–ZSM-5ssaft indicates a gain in Vp comprising 50% compared with ZSM-5, reflecting the accessibility of ZSM-5 sample in opening some pores during Fe incorporation, affecting severely the crystallinity of this sample. The increase in quadrupole splitting of Fe–ZSM-5impbef can be interpreted as an increase in electron spin interaction caused by the movement of extra-framework Fe closer to other iron in both framework and extra-framework positions. The increase in quadrupole splitting is also likely due to the presence of disturbed octahedrally coordinated Fe ions in the extra-framework positions. Nevertheless, since the average isomer shift of this sample (0.25 mm/s) is close to that of Fe–ZSM-5in (0.24 mm/s), the majority of iron ions (>80%) are still tetrahedrally coordinated in isolated framework positions.
nated in the MFI framework. This is evidenced from the reported increase in lattice volume and decreased crystallites size of this material that is taken as a criterion for the devoted strong interaction aroused between Fe and ZSM-5. This sample showed the largest SBET between all samples, thus it can be chosen in surface driven applications. Fe3+ ions did not completely exchange Al3+ in the structure but rather showed the existence of both ions and possibly in the same sites. Hardly detected nanoparticles of a-Fe2O3 were depicted in this sample as confirmed from XRD and UV–Vis investigations in addition to small amounts of small clusters of Fe3þ x O species. Fe species depicted on Fe–ZSM-5imp either calcined or not were varied in the sense that Fe–ZSM-5impbef did not indicate a-Fe2O3 particles in the electronic spectra and rather measured the lowest SBET (361 m2/g). This did not nullify the presence of the latter species that could be of smaller sizes to be detected by the UV–Vis technique. Most probably, these species are located in the pores of this sample and thus could be responsible for the devoted decrease in the specific surface area of this sample. This could give a criterion about the dislodgment of Fe3+ from framework to extraframework as a function of the thermal effect, which in its turn facilitates the migration of Fe3+ as isolated a-Fe2O3 through water elimination or with the interaction between resid¨ nsted sites [Fe(OH)Si]. ual BrO Calcination of the as-synthesized Fe–ZSM-5ssaft material presented more or less typical Fe species as Fe–ZSM-5ssbef except the absence of a-Fe2O3 species that has been ascertained for the latter by Mo¨ssbauer and UV–Vis investigations. Surface texture of both samples presented varied properties in the sense that Fe–ZSM-5ssbef owned high microporosity (84%) where the other acquired the maximum mesoporosity (40%) between all samples.
4. Conclusion
References
The majority of Fe cations in the as-synthesized Fe–ZSM-5in zeolites calcined at 550 °C are present as paramagnetic Fe3+ ions that are tetrahedrally coordi-
[1] A. Ribera, I.W.C.E. Arends, S. de Vries, J. Perez-Ramirez, R.A. Sheldon, J. Catal. 195 (2000) 287. [2] J.B. Taboada, A.R. Overweg, P.J. Kooyman, I.W.C.E. Arends, G. Mul, J. Catal. 231 (2005) 56.
102
M.M. Mohamed et al. / Microporous and Mesoporous Materials 87 (2005) 93–102
[3] J. Perez-Ramirez, G. Mul, F. Kapteijn, J.A. Moulijn, A.R. Overweg, A. Domenech, A. Ribera, I.W.C.E. Arends, J. Catal. 207 (2002) 113. [4] M. Schwidder, M.S. Kumar, K. Klementiev, M.M. Pohl, A. Bruckner, W. Grunert, J. Catal. 231 (2005) 314. [5] M.M. Mohamed, J. Mol. Catal. 200 (2003) 301. [6] M.M. Mohamed, F. Abdel-Hai, J. Mol. Catal. 211 (2004) 199. [7] P.B. Venuto, Micropor. Mater. 2 (1994) 297. [8] Md. Uddin, T. Komatsu, T. Kashima, J. Catal. 150 (1994) 439. [9] R.Q. Long, R.T. Yang, J. Catal. 207 (2002) 224. [10] X. Feng, W.K. Hall, Catal. Lett. 41 (1996) 45. [11] M.M. Mohamed, N.A. Eissa, Mater. Res. Bull. 38 (2003) 1993. [12] J. Perez-Ramirez, J. Catal. 227 (2004) 512. [13] E.J.M. Hensen, Q. Zhu, M.M.R.M. Hendrix, A.R. Overweg, P.J. Kooyman, M.V. Sychev, R.A. Van Santen, J. Catal. 221 (2004) 560. [14] H.Y. Chen, W.M.H. Sachtler, Catal. Today 42 (1998) 73. [15] G.I. Panov, Cattech 4 (2000) 18. [16] J. Perez-Ramirez, E.V. Kondratenko, Chem. Commun. (2003) 2152. [17] K.S.W. Sing, D.H. Everett, R.A.W. Haul, I. Moscou, R.A. Pierotti, J. Rouqueroi, T. Siemieniewsk, Pure Appl. Chem. 57 (1985) 603.
[18] F. Rouquerol, J. Rouquerol, K.S.W. Sing, in: F. Schutz, K.S.W. Sing, J. Weitkamp (Eds.), Handbook of Porous Materials, vol. 1, Wiley-VCH, Weinhiem, 2002, p. 250. [19] M.M. Mohamed, T.M. Salama, I. Othman, G.A. El-Shobaky, Appl. Catal. A 279 (2005) 23. [20] T.M. Salama, M.M. Mohamed, I. Othman, G.A. El-Shobaky, Appl. Catal. A 286 (2005) 85. [21] M.M. Mohamed, T.M. Salama, I. Othman, E. Abdalla, Micropor. Mespor. Mater., in press. [22] J. Zhao, Z. Feng, F.E. Huggins, N. Shah, G.P. Huffman, I. Wener, J. Catal. 148 (1994) 194. [23] E.L. Wu, S.L. Lawton, D.H. Olson, A.C. Rohrman Jr., G.T. Kokotailo, J. Phys. Chem. 83 (1979) 2777. [24] S. Bordiga, A. Damin, G. Berlier, F. Bonino, G. Ricchiardi, A. Zecchina, C. Lamberti, Int. J. Mol. Sci. 2 (2001) 167. [25] G. Lehmann, Z. Phys. Chem. Neue. Folge 27 (1970) 279. [26] C. Minchev, H. Huwe, T. Tsoncheva, D. Paneva, M. Dimitrov, I. Mitor, M. Froba, Micropor. Mesopor. Mater. 81 (2005) 333. [27] C.D.-M. Assou, J.L. Dormann, O. Gorochov, D. Svoronos, Hyperfine Interact. 28 (1986) 915. [28] X. Feng, W.K. Hall, J. Catal. 166 (1997) 386.