Microporous and Mesoporous Materials 118 (2009) 341–347

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials

journal homepage: www.elsevier .com/locate /micromeso

Effect of microwaves in the dealumination of mordeniteon its surface and acidic properties

María Dolores González a,b, Yolanda Cesteros a, Pilar Salagre a,*, Francisco Medina b,Jesús E. Sueiras b

a Facultat de Química, Universitat Rovira i Virgili, C/Marcellí Domingo s/n, 43007 Tarragona, Spainb Escola Tècnica Superior d’Enginyeria Química, Universitat Rovira i Virgili, Avenue Països Catalans, 26, 43007 Tarragona, Spain

a r t i c l e i n f o

Article history:Received 22 April 2008Received in revised form 2 September 2008Accepted 4 September 2008Available online 11 September 2008

Keywords:MicrowavesDealuminationMordeniteAcidityPorosity

1387-1811/$ - see front matter � 2008 Elsevier Inc. Adoi:10.1016/j.micromeso.2008.09.005

* Corresponding author. Tel.: +34 977559571; fax:E-mail address: pilar.salagre@urv.net (P. Salagre).

a b s t r a c t

Commercial mordenite was partially dealuminated in HCl medium by conventional heating or undermicrowaves by refluxing or autoclaving at 373 K at different times. Samples were characterized byAAS, XRD, N2 physisorption, FT-IR, 27Al NMR, NH3–TPD, and SEM techniques. The acidity of the dealumi-nated samples was also determined by testing them as catalysts in two acid catalysed reactions: theisomerization of styrene oxide to obtain b-phenylacetaldehyde, and the styrene oxide ring-opening reac-tion to give 2-ethoxy-2-phenylethanol. The use of microwaves, under autoclave or refluxing conditions,enhances dealumination, favours the later elimination of the Al extracted during washing, and affects thesurface and acidic characteristics of the resulting samples. All catalysts showed similar low amounts ofBrønsted acid sites. However, the catalyst treated under microwaves by autoclaving at shorter time(15 min) presented active acid centres with medium strength, and interestingly, lower amounts of strongLewis acid sites (responsible for deactivation) than the rest of catalysts. These characteristics explain thetotal conversion obtained for this catalyst for the styrene oxide ring-opening reaction.

� 2008 Elsevier Inc. All rights reserved.

1. Introduction

Dealumination of zeolites is an essential step in the synthesis ofa large number of commercial zeolite formulations. Barrer andMakki were the first to show that dealumination of zeolite frame-works could be achieved without loss of the zeolite structure, dem-onstrating that a natural zeolite clinoptilolite can be dealuminatedby treatment with hydrochloric acid [1].

Mordenite has been identified as suitable acid catalyst in sev-eral industrial processes, such as cracking and isomerization ofhydrocarbons [2]. This zeolite is comprised of two straight channeltypes: (i) larger channels, also called main channels, accessiblethrough twelve member oxygen rings with an opening of7.0 � 6.5 Å and (ii) smaller channels, often referred to as com-pressed channels, which include eight member oxygen rings with2.6 � 5.7 Å [3]. On the whole, mordenite catalysts undergo rapiddeactivation because of their uni-dimensional pore system withsmall side-pockets that are generally not accessible for reactantmolecules, and limit the free diffusion of intermediate and productmolecules. However, by applying dealumination methods, morde-nite has been successfully used in industrial processes such as theDOW’s process for cumene production [4], and the Shell’s process

ll rights reserved.

+34 977559563.

for hydroisomerization of linear alkanes to branched alkanes [5].A recent study showed that the dealumination of mordenite canimprove the reactant molecules diffusion through its uni-dimen-sional pores, and results in higher resistance to deactivation cata-lyst for the isomerization of n-hexane [6].

Framework dealumination of mordenite causes changes in bothpore system and acidity. Several authors reported that dealumina-tion of mordenite lead to an enlargement of pore sizes in the mainchannels as well as in the side-pockets, and/or an increase of porevolume and surface area in mesopores [7–15]. With respect to theacidity, the dealumination of mordenite involves the formation ofnew Lewis acid sites due to the presence of extra-framework alu-minum species (EFAl), such as AlO+, Al(OH)2+, AlðOHÞþ2 or AlO(OH),which imparts, via an inductive effect, stronger acidity to theremaining Brønsted acid sites [16–19].

A considerable number of zeolite dealumination techniqueshave been developed. Thus, for mordenite we found dealuminationstudies by treatment with steam or SiCl4 vapour at elevated tem-peratures or treatment with (NH4)SiF6, mineral acids (i.e., HCl,HNO3), organic acids (i.e., acetic acid, oxalic acid), F2, chelatingagents (i.e., EDTA), etc [4–27]. Conventional heating is used whenapplying temperature during dealumination.

Nowadays, microwave irradiation is being applied for the dry,synthesis, and cation-exchange of zeolites [28–30]. The use ofmicrowaves considerably decreases the preparation times, with

Table 1Chemical analyses results

Sample Si/Al (atomic ratio) Na2O (wt.%)

M 6.5 6.50R15 min 14.9 0.10R2 h 18.2 0.08A15 min 11.2 0.11A2 h 17.1 0.07MWR15 min 15.5 0.11MWR2 h 20.2 0.04MWA15 min 15.8 0.08MWA2 h 23.6 0.04

342 M.D. González et al. / Microporous and Mesoporous Materials 118 (2009) 341–347

the subsequent energy saving, and modifies the samples proper-ties. Therefore, microwave syntheses constitute valuable processesin Green Chemistry. There is no reference about the use of micro-waves for the dealumination of zeolites.

The aim of this work is to study the effect of using microwaves(in autoclave and by refluxing), during dealumination in HCl med-ium of commercial mordenite, on its resulting surface and acidicproperties. Commercial mordenite was also dealuminated in acidmedium by conventional heating (in autoclave and by refluxing)for comparison. All samples were characterized by a wide numberof techniques.

2. Experimental

2.1. Preparation of partially dealuminated mordenite samples

The starting material was a commercial Na-Mordenite (zeolyst,Si/Al = 6.5, CBV 10A Lot No. 1822-50), designated as M. We treated1 g of mordenite with 30 mL HCl 6 M for each preparation. Foursamples were heated under microwaves (Milestone ETHOS-TOUCHCONTROL equipped with a temperature controller), two in an auto-clave at 373 K for 15 min and 2 h (samples MWA15 min, MWA2 h,respectively), and the other two by refluxing at the same tempera-ture and times (samples MWR15 min and MWR2 h). Four more sam-ples were heated by autoclaving in a conventional oven at 373 Kfor 15 min and 2 h (samples A15 min, A2 h, respectively) or by tradi-tional refluxing at the same temperature and times (samples R15 min

and R2 h). After the acid treatment, samples were washed severaltimes with deionized water, and dried in an oven overnight.

2.2. Elemental analyses

Elemental analyses of the samples were obtained with a PhilipsPW-2400 sequential XRF analyzer with Phillips Super Q software.All measures were made in triplicate.

2.3. 27Al MAS NMR

27Al NMR spectra were obtained with a Varian Mercury Vx400 Mhz with a probe of 7 mm CPMAS at a frequency of400 MHz by spinning at 5 kHz. The pulse duration was 2 ls, the de-lay time was 5 s, and the chemical shift reference was high purityaluminium nitrate.

2.4. X-ray diffraction (XRD)

Powder X-ray diffraction patterns of the samples were obtainedwith a Siemens D5000 diffractometer using nickel-filtered Cu Karadiation. Samples were dusted on double-sided sticky tape andmounted on glass microscope slides. The patterns were recordedover a range of 2h angles from 5� to 40�. Crystalline phases wereidentified using the Joint Committee on Powder Diffraction Stan-dards (JCPDS) files (43-0171 corresponds to mordenite). Cellparameters were calculated from [200], [020] and [202] peaksusing a matching profile with WIN FIT 1.2 software. Crystallinitywas determined by comparing the sum of the peak areas of[150], [202], [350] and [402] (22–32� 2H) of the modifiedmordenites with respect to commercial Na-mordenite.

2.5. Nitrogen physisorption

BET areas were calculated from nitrogen adsorption isothermsobtained at 77 K using a Micromeritics ASAP 2000 surface analyzerwith a value of 0.164 nm2 for the cross-section of the nitrogen mol-ecule. Samples were pretreated in vacuum at 573 K for 6 h. Pore

volumes and surface areas of micropores and mesopores weredetermined from their isotherms using the Horvath–Kawazoemethod and the BJH method, respectively.

2.6. FT-IR

Infrared spectra were recorded on a Bruker-Equinox-55 FT-IRspectrometer. The spectra were acquired by accumulating 32 scansat 4 cm�1 resolution in the range of 400–4000 cm�1. Samples wereprepared by mixing the powdered solids with pressed KBr disks ina ratio of 5:95, and dried in an oven overnight.

2.7. Temperature-programmed desorption-mass spectrometryexperiments (TPD)

The acid properties of the samples were characterized by NH3–TPD using a TPD/R/O 1100 Thermo Finnigan, equipped with a pro-grammable temperature furnace and TCD detector. The gas outletwas couplet to a quadrupole mass spectrometer Pfeiffer GSD300to identify the peaks. Experiments were performed with 3% NH3/He flowing through the sample which was previously activatedat 673 K for 1 h. The desorption of NH3 was made by flowing He20 cm3/min from room temperature to 1073 K at 5 K/min.

2.8. Scanning electron microscopy (SEM)-X-Ray microanalysis

This technique was used to observe the morphology and parti-cle sizes of the samples. Experiments were performed on ascanning electron microscope, JEOL JSM6400, operating at acceler-ating voltage of 15 kV, work distances of 15 mm, and magnifica-tions in the range 3700–50,000�.

2.9. Catalytic activity studies

Isomerization of styrene oxide (SO), and styrene oxide ring-opening reactions were carried out in the liquid phase at atmo-spheric pressure at room temperature. The catalytic experimentswere performed using 0.8 g of catalyst, 20 ml of solvent (tolueneor ethanol to favour SO isomerization or SO ring-opening, respec-tively) and 0.48 ml of styrene oxide. The reaction products, takenat 3 h of reaction, were analysed by GC on a Shimadzu GC-2010instrument equipped with a 30 m capillary column DB-1, coatedwith phenylmethylsilicon, and a FID detector.

3. Results and discussion

Table 1 shows the Si/Al ratio, and the Na2O weight percentageof the acid-treated mordenites compared to the commercial one.We observed an increase of the Si/Al ratio accompanied by a de-crease in the Na+ content, as expected, for all the treated samples.Longer treatments times, and the use of microwaves resulted in anincrease in the amount of Al removed, independently of using

M.D. González et al. / Microporous and Mesoporous Materials 118 (2009) 341–347 343

autoclave or refluxing methods. Therefore, there was a clear reduc-tion of time by using microwaves to achieve similar dealuminationdegrees than by conventional heating, especially under autogenouspressure. 27Al NMR spectra (not shown here) exhibited the pres-ence of non-tetrahedral Al for all the modified samples, confirmingthe presence of Al extra-framework for all of them.

The acid and heating conditions used here did not cause drasticchanges in the mordenite structure after acid treatment, althoughthere was some decrease in the crystallinity of the mordenitestructure, as deduced from the relative crystallinity values calcu-lated from XRD results for the dealuminated samples (Table 2).These values were similar to those reported by other authors whendealuminating mordenite with mineral acids [6,14]. This decreasein the crystallinity was slightly larger for the microwaved samples(Table 2), probably related to their higher dealumination (Table 1).XRD peaks appeared displaced to higher 2H values, and therefore,to lower interplanar distances for all treated samples with respectto commercial mordenite. This displacement was slightly higherfor the mordenites treated under microwaves (Fig. 1). This in-volved a variation of a, b and c parameters resulting in a decreaseof the unit cell volumes for all the acid-treated samples (Table 2).Whatever the dealumination method, the variation of the b param-eter was greater than that of a and c. This anisotropic contractionhas been attributed to the fact that Al atoms occupy non-equiva-lent sites and, during dealumination, some of them would be re-moved before other ones [14]. Such contraction (Si–O bond

Table 2Characterization of samples by XRD and FT-IR techniques

Sample XRD

Critallinity (%) a (Å) b (Å) c (Å)

M 100 18.13(3) 20.48(2) 7.51(3)R15 min 76 18.10(2) 20.15(2) 7.46(3)R2 h 75 18.18(2) 20.25(3) 7.46(3)A15 min 73 18.12(3) 20.21(2) 7.47(3)A2 h 75 18.18(2) 20.25(3) 7.46(3)MWR15 min 67 18.10(2) 20.16(3) 7.46(3)MWR2 h 68 18.08(2) 20.19(3) 7.45(3)MWA15 min 70 18.07(3) 20.13(2) 7.46(3)MWA2 h 68 18.05(2) 20.11(3) 7.45(3)

a Calculated from XRD patterns.b Frequencies of the main asymmetric stretch (t1) and the main symmetric stretch (t

Fig. 1. XRD patterns for the samples: (a) M, (b

length: 1.62 Å vs. 1.69 Å for Al–O) has been generally accountedfor, in steam-dealuminated zeolites, by the migration of siliconatoms and filling of the structural defects created by the departureof Al ions (cicatrization effect), [14,31], but also due to the presenceof silanol groups [32]. The contraction of the unit cell volume wasslightly higher for the mordenites heated under microwaves, andincreased at higher Si/Al ratio.

Infrared spectra of dealuminated mordenites and the startingmordenite were taken since it is well known that the symmetricand asymmetric stretching frequencies of the T–O bond (T = Si,Al) increases when the aluminium content is lower [33]. This isdue to the increase of the strength of the T–O bond when the Alcontent decreases (the Si–O bond is shorter than the Al–O bondand Al has lower electronegativity than Si). The FT-IR spectra ofthe acid-treated mordenites showed a shift to higher values ofthe symmetric and asymmetric stretching frequencies of the T–Obond (T = Si, Al) (Table 2). This confirmed the removal of alumin-ium atoms from the framework.

Besides, the acid-treated samples showed higher BET areas,lower micropore area/non-micropore area ratios, and higher porevolumes than commercial mordenite (Table 3). These results canbe associated to the loss of aluminum in the mordenite structurewhich lead to higher mesoporosity, and therefore, to higher surfaceareas. This is in agreement with the results reported by otherauthors [6,14,27]. Interestingly, this tendency appeared moremarked for the microwaved samples. The appearance of hysteresis

FT-IR Frequency of bands (cm�1)b

Unit cell volume (Å3)a m1 m2

2791 1068 6292722 1087 6392747 1093 6462737 1084 6352747 1093 6442722 1090 6382721 1093 6442713 1091 6412705 1093 643

2) due to the T–O bond (T = Si, Al).

) MWA2 h, (c) MWR2 h, (d) A2 h and (e) R2 h.

Table 3Characterization of samples by nitrogen physisorption

Sample BET area(m2/g)

micropore area/non-microporearea ratio

Volume pore(cm3/g)

M 303 8.9 0.059R15 min 384 5.7 0.109R2 h 412 5.1 0.110A15 min 376 6.1 0.096A2 h 397 5.3 0.106MWR15 min 388 5.1 0.102MWR2 h 409 5.2 0.121MWA15 min 419 5.9 0.110MWA2 h 430 4.9 0.120

344 M.D. González et al. / Microporous and Mesoporous Materials 118 (2009) 341–347

loops in the adsorption isotherm on the acid-treated mordenitesconfirmed the formation of mesopores by the acid treatment.

Scanning electron microscopy was used to monitor the mor-phologies and sizes of the particles of the acid-treated mordenitescompared to commercial mordenite. Fig. 2 shows the micrographsobtained for several samples. Commercial mordenite exhibitedhomogeneous rounded particles with sizes in the range 100–300 nm (Fig. 2a). Acid treatment under heating resulted in an

Fig. 2. Scanning electron micrographs of samples: M

agglomeration of the particles (Fig. 2b–e), which presented heter-ogeneous sizes ranging from 100 to 3700 nm.

The acidity of all the partially dealuminated samples was eval-uated by NH3–TPD. Fig. 3 shows the NH3–TPD thermograms of sev-eral representative samples whereas Table 4 depicts the TPDdesorption temperature maxima obtained for all mordenites. TheNH3–TPD profile of Na-Mordenite only presented one peak withlow intensity at 475 K [34,35]. This peak has been assigned toammonia weakly held or physically adsorbed on the mordenite[35]. On the other hand, the partially dealuminated samplesshowed two non-symmetrical NH3–TPD peaks (Fig. 3): one med-ium–high intense peak (peak 1) at 413–513 K, and a lower–med-ium intense peak (peak 2) at higher desorption temperatures(778–843 K). This is in agreement with the NH3–TPD thermogramsreported by other authors for dealuminated zeolites [36]. Thepeaks of the NH3–TPD thermograms corresponding to the samplesdealuminated under refluxing had less intensity than those of theautoclaved samples (Table 4, Fig. 3). Therefore, the refluxed sam-ples have less acidity than the autoclaved ones. The use of atmo-spheric pressure favours more efficiently than autogeneouspressure the later elimination during washing of the Al removedfrom the zeolitic structure.

(a), MWA2 h (b), A2 h (c), MWR2 h (d) and R2 h (e).

Fig. 3. NH3–TPD thermograms of several dealuminated samples: (a) A15 min, (b) R15 min, (c) MWA15 min, (d) MWR15 min and (e) MWA2 h.

M.D. González et al. / Microporous and Mesoporous Materials 118 (2009) 341–347 345

In the literature, the assignment of the low temperature peak(peak 1) appears to be controversial [36]. This peak has been attrib-uted to the release of NH3 hydrogen-bound to NHþ4 cations (read-sorption of NH3) [37] but also to weakly acid silanol groups(Brønsted acidity) [38]. The possibility that weak Lewis acid siteswere responsible for this peak has also been considered [39–41].It has been suggested that some extra-framework alumina species,such as AlðOHÞþ2 and Al(OH)2+ could be sorption sites of a weak Le-wis acid character [39,40]. The non-symmetrical shape of this peakin our samples allowed us to think that this peak was formed formore than one contribution.

Mordenites dealuminated under refluxing had a less intensepeak 1 than the autoclaved ones (Table 4, Fig. 3). The increase of

Table 4NH3–TPD results for the samples

Samples NH3 TD (K)a

Peak 1 Peak 2

M 475 (l) –R15 min 433 (m) 793 (m)R2 h 433 (m) 803 (h)A15 min 435 (h) 813 (m)A2 h 423 (h) 835 (h)MWR15 min 413 (m) 788 (m)MWR2 h 413 (m) 783 (h)MWA15 min 513 (h) 843 (l)MWA2 h 418 (h) 778 (m)

a TD, Maxima of NH3 desorption temperature peaks; (l), low intense peak; (m),medium intense peak; (h), high intense peak.

the treatment time did not affect the intensity and the maximumof desorption temperature of peak 1 for the refluxed mordeniteswhereas for the autoclaved samples, we observed a decrease inthe desorption temperature maximum at longer treatment times(2 h) (Table 4). The use of microwaves affects in a different waywhen using refluxing or autoclaving preparation methods. Forthe refluxed samples (MWR15 min, MWR2 h), and for the mordenitetreated in autoclave for 2 h (MWA2 h) we saw a lower desorptiontemperature of peak 1 than the corresponding samples conven-tionally heated. However, sample MWA15 min, treated under micro-waves in autoclave for 15 min, presented a maximum of peak 1 athigher desorption temperature than sample A15 min, treated in con-ventional autoclave for 15 min. Taking into account all these vari-ations observed between the samples, we believe that peak 1 canbe assigned, on the whole, to weak acid sites (Brønsted and/or Le-wis) associated to the dealumination procedure. Interestingly,microwaves led to weaker acidity than conventionally heated sam-ples, except for sample MWA15 min (Table 4, Fig. 3), which showedmedium acidity (higher desorption temperature). In this case, theshorter time used during dealumination was not enough to favoura later efficient elimination of the extracted Al (which was higherthan for sample A15 min (Table 1)), that probably remained as ex-tra-framework AlðOHÞþ2 and Al(OH)2+ species. An inductive effectbetween both, these species and the silanols present, can explainthe higher desorption temperature of peak 1 observed for this sam-ple. This medium acidity, achieved by using microwaves at shortertimes during the acid treatment, will play an important role on thecatalytic results, as commented below.

346 M.D. González et al. / Microporous and Mesoporous Materials 118 (2009) 341–347

With respect to peak 2, it had higher intensity for the samplesdealuminated at longer times (2 h) (Table 4). Since commercialmordenite did not present peak 2, we can assume that this secondpeak can be related to some strong Lewis acid centres associated toextra-framework Al insoluble species generated from the zeoliticframework during dealumination. Interestingly, peak 2 was less in-tense for the samples prepared under microwaves, especially byautoclaving (Fig. 3c,d) than for the samples prepared by conven-tional heating (Fig. 3a,b). This seems to confirm that the use ofmicrowaves allows us a better elimination during washing of theAl extracted during dealumination resulting in less amounts ofstrong Lewis acid centres.

In order to try to determine the nature of the acidity of the cen-tres observed by NH3–TPD, we tested the samples as catalysts intwo reactions catalysed by different acid sites: the isomerizationof styrene oxide to obtain b-phenylacetaldehyde, which is mainlycatalysed by Brønsted acids sites, and the styrene oxide ring-open-ing to give 2-ethoxy-2-phenylethanol, which is catalysed by bothBrønsted and Lewis acid sites [34].

Table 5 shows the catalytic activity of all samples for both reac-tions. All catalysts showed similar yield values to b-phenylacetal-dehyde (PA), slightly higher than commercial mordenite, for thestyrene oxide isomerization. Since Brønsted acid sites catalyse thisreaction, as commented above, we can assume that the dealumi-nated samples have similar low amounts of Brønsted acid sites.

However, in the styrene oxide ring-opening reaction, catalystspresented different catalytic behaviour. There are two importantfeatures to remark: on the one hand, the samples treated at shortertimes showed higher conversion and selectivity to EPE values thanthose treated at longer times, independently of the dealuminationmethod used; on the other, the samples dealuminated undermicrowaves exhibited higher conversion and higher selectivity toEPE values than the samples dealuminated under conventionalheating. Thus, MWA15 min yield much higher conversion and muchhigher selectivity to 2-ethoxy-2-phenylethanol (EPE), followed bysample MWR15 min, than the rest of catalysts (Table 5).

The higher conversion and higher selectivity to EPE observed forsample MWA15 min can be explained by the presence of active acidcentres (Lewis and Brønsted) with medium strength (peak 1) to-gether with the existence of very low amounts of strong Lewis acidcentres (peak 2), observed by NH3–TPD (Fig. 3c). At higher treat-ment times, samples have weaker acid centres (peak 1) and higheramounts of strong acid centres (peak 2) (Table 4). Strong acid cen-tres are responsible for deactivation, as previously reported [34]. Infact, we observed a higher deactivation with time (not shown here)for the catalysts acid-treated at longer times (2 h), which presentedhigher amounts of these strong acid centres, as commented above.

Table 5Catalytic activity for the two acid catalysed reactions

Catalysts SOisomerizationa

SO ring-openingb

% PA yield %Conversion

% PAselectivity

% EPEselectivity

M 12 15 – –R15 min 42 36 44 56R2 h 41 32 51 49A15 min 36 49 46 54A2 h 54 43 48 52MWR15 min 43 73 16 84MWR2 h 42 50 34 66MWA15 min 50 100 4 96MWA2 h 44 46 40 60

PA, b-phenylacetaldehyde; EPE, 2-ethoxy-2-phenylethanol.a Reaction time: 3 h, solvent: toluene.b Reaction time: 3 h, solvent: ethanol.

If we compare the catalytic activity results of catalysts R15 min andMWR15 min, prepared by refluxing under conventional heating andunder microwaves, respectively, we observe that they have similaramounts of weak acid centres (Fig. 3b and d) but sample MWR15 min

has lower amounts of strong acid sites, explaining its higher con-version value. The less amounts of these strong acid sites presentin the microwaved samples (especially at shorter times), explainthe higher conversion values observed for these catalysts.

In this work, we report for the first time that the use of micro-waves during dealumination of mordenite by acid treatment, notonly enhances dealumination by decreasing considerably the timeneeded to achieve similar dealumination degrees than by conven-tional heating methods but also favours the later elimination of theAl extracted from the zeolite structure resulting in partially dealu-minated mordenites with less strong acid sites than those conven-tionally dealuminated. In the literature, the mechanism ofmicrowave heating of zeolite has been studied. Several authors re-ported that the heating proceeds in two steps; in the first step thehydrated zeolite absorbs microwaves through its adsorbed water,and in the second step the heated zeolite directly microwaves[42–43]. Therefore, the different heating process which involvesthe use of microwaves allow us to dealuminate mordenite in lesstime obtaining materials with different acidic properties, withtheir subsequent potential use in catalysis.

4. Conclusions

The use of microwaves for the dealumination of commercialmordenite affects the surface and acidic properties of the resultingpartially dealuminated samples. Microwaves lead to faster dealu-mination than conventional heating by autoclaving as well as byrefluxing.

Dealumination of the samples was observed by XRD, N2 physi-sorption, IR, elemental analyses, 27Al MAS NMR, and SEM tech-niques since the acid-treated samples showed higher Si/Al molarratio, lower cell volume, higher BET area, and lower microporearea/non-micropore area ratio than commercial mordenite. Thesevariations appeared more pronounced for the microwaved sam-ples, confirming the effect of using microwaves on the surfacecharacteristics of the mordenite samples.

NH3–TPD thermograms presented two non-symmetricaldesorption peaks for all the partially dealuminated samples. Thefirst desorption peak has been related to the presence of weak–medium acid centres, which are a contribution of silanol groups(Brønsted acidity) together with AlðOHÞþ2 or Al(OH)2+ Lewis speciesgenerated during dealumination. The second peak has been associ-ated to strong Lewis acid centres due to extra-framework insolublealuminium species also formed during dealumination.

Samples dealuminated under refluxing have less acidity thanthe autoclaved ones due to the best elimination, during washing,of the Al extracted at atmospheric pressure.

The use of microwaves leads to lower acidity, due to the higherdealumination produced followed by a more efficient eliminationof Al during washing, except for the sample prepared under auto-clave at shorter time. Interestingly, in this sample the presence ofacid sites with medium strength contributes decisively to the cat-alytic activity results.

The partially dealuminated samples showed similar lowamounts of Brønsted acid sites, as deduced from the catalytic resultsobtained for the isomerization of styrene oxide. Interestingly, fromthe catalytic results obtained for the styrene oxide ring-openingreaction, we observed that when the mordenite was dealuminatedat shorter time (15 min) under microwaves, higher conversion val-ues were obtained. The presence of active acid centres with mediumstrength together with the lower amounts of stronger Lewis acid

M.D. González et al. / Microporous and Mesoporous Materials 118 (2009) 341–347 347

sites, responsible for deactivation, observed for catalyst MWA15 min

explains its total conversion for the styrene oxide ring-openingreaction.

Acknowledgments

The authors are grateful for the financial support of the Minis-terio de Educación Ciencia and FEDER funds (CTQ2005-02384/PPQ).

References

[1] R.M. Barrer, M.B. Makki, Can. J. Chem. 42 (1964) 1481.[2] I.E. Maxwell, Catal. Today 1 (1987) 385.[3] D.W. Breck, Zeolite Molecular Sieves, Wiley, New York, 1974, p. 122.[4] M.M. Olken, J.M. Garces, in: von Ballmoos, J.B. Higgings, M.M.J. Treacy (Eds.),

Proceedings of 9th International Zeolite Conference, Butterworth-Heinemann,Boston, MA, 1993, Montreal, Canada, 1992, p. 559.

[5] G.J. Lee, J.M. Garces, G.R. Meima, M.J.M. Van der Aalst, US Patent 32517, 1989.[6] N. Viswanadham, M. Kumar, Micropor. Mesopor. Mater. 92 (2006) 31.[7] F. Raatz, C. Marcill, E. Freund, J. Chem. Soc. Faraday Trans. 79 (1983) 2299.[8] J. Nagano, T. Eguchi, T. Asanuma, M. Nakayama, N. Nakamura, E.G. Derouance,

Micropor. Mesopor. Mater. 33 (1999) 249.[9] S. van Donk, A. Broersma, O.L.J. Gijzeman, J.A. Van Bokhoven, J.H. Bitter, K.P. De

Jong, J. Catal. 204 (2001) 272.[10] B.L. Meyers, T.H. Fleisch, G.J. Ray, J.T. Miller, J.B. Hall, J. Catal. 110 (1998) 82.[11] M.J.A. van Tromp, M.T. Garriga Oostenbrink, J.H. Bitter, K.P. de Jong, D.C.

Koningsberger, J. Catal. 190 (2000) 209.[12] N.S. Nesterenko, F. ThibaultStarzyk, V. Montouillout, V.V. Yuschenko, C.

Fernandez, J.P. Gilson, F. Fajula, I.I. Ivanova, Micropor. Mesopor. Mater. 71(2004) 157.

[13] K.H. Lee, B.H. Ha, Micropor. Mesopor. Mater. 23 (1998) 211.[14] S. Moreno, G. Poncelet, Micropor. Mater. 12 (1997) 197.[15] D. Vergani, R. Prins, H.W. Kouwenkoven, Appl. Catal. A: Gen. 163 (1997) 71.[16] P.O. Fritz, J.H. Lunsford, J. Catal. 118 (1989) 85.

[17] M.H.W. Sonnemans, C. den Heijer, M. Crocker, J. Phys. Chem. 97 (1993) 440.[18] A. Corma, Chem. Rev. 95 (1995) 559.[19] R.A. Van Santen, G.J. Kramer, Chem. Rev. 95 (1995) 637.[20] M. Sawa, M. Niwa, Y. Murakami, Zeolites 12 (1995) 175.[21] M.J. Van Niekerk, J.C.Q. Fletcher, C. O’Connor, J. Catal. 138 (1992) 150.[22] G.J. Hutchings, A. Burrows, C. Rhodes, C.J. Kely, R. McClung, J. Chem. Soc.

Faraday Trans. 93 (1997) 3593.[23] Y. Hong, V. Gruver, J.J. Fripiat, J. Catal. 150 (1994) 421.[24] W.O. Haag, R.M. Lago, US Patent 4 326 994, 1982.[25] Z.M. Magrioti Noronha, J.L. Fontes Monteiro, P. Gélin, Micropor. Mesopor.

Mater. 23 (1998) 331.[26] M. Müller, G. Harvey, R. Prins, Micropor. Mesopor. Mater. 34 (2000) 135.[27] K.H. Chung, Micropor. Mesopor. Mater. 111 (2008) 544.[28] H.M. Kingston, S.J. Haswell, Microwave-Enhanced Chemistry Fundamentals,

Sample Preparation and Applications, American Chemical Society,Washington, DC, 1997.

[29] Y. Kuroda, T. Okamoto, R. Kumashiro, Y. Yoshikawa, M. Nagao, Chem.Commun. (2002) 1758.

[30] M.D. Romero, G. Ovejero, M.A. Uguina, A. Rodríguez, J.M. Gómez, Catal.Commun. 5 (2004) 154.

[31] J. Scherzer, J.M. Bass, J. Catal. 28 (1973) 115.[32] P.E. Eberly, C.N. Kimberlin, A. Voorhies, J. Catal. 22 (1971) 419.[33] B. Imelik, J.C. Vedrine, Catalyst Characterization, Physical Techniques for Solid

Materials, Plenum Press, New York, 1994.[34] I. Salla, O. Bergadà, P. Salagre, Y. Cesteros, F. Medina, J.E. Sueiras, T. Montanari,

J. Catal. 232 (2005) 239.[35] M. Niwa, N. Kertada, Catal. Surv. Jpn. 1 (1997) 215.[36] F. Lónyi, J. Valyon, Micropor. Mesopor. Mater. 47 (2001) 293.[37] H. Igi, N. Katada, M. Niwa, in: M.M.J. Treacy, B.K. Marcus, M.E. Bisher, J.B.

Higgins (Eds.), Proceedings of the 12th International Zeolite Conference,Materials Research Society, Warrendale, PA, 1999, p. 2643.

[38] N.Y. Topsoe, K. Pedersen, E.G. Derouane, J. Catal. 70 (1981) 41.[39] N.R. Meshram, S.G. Hegde, S.B. Kulkarni, Zeolites 6 (1986) 434.[40] G.L. Woolery, G.H. Kuehl, H.C. Timken, A.W. Chester, J.C. Vartuli, Zeolites 19

(1997) 288.[41] H.G. Karge, V. Dondur, J. Phys. Chem. 94 (1990) 765.[42] T. Ohgushi, S. Komarneni, A.S. Bhalla, J. Porous Mater. 8 (2001) 23.[43] T. Ohgushi, Mg. Nagae, J. Porous Mater. 10 (2003) 139.