Materials Research Bulletin 46 (2011) 19–25

Mesoporous silica directly modified by incorporation or impregnation of someheteropolyacids: Synthesis and structural characterization

Alexandru Popa a,*, Viorel Sasca a, Erne E. Kiss b, Radmila Marinkovic-Neducin b,Ivanka Holclajtner-Antunovic c

a Institute of Chemistry Timisoara, Bl.Mihai Viteazul 24, 300223 Timisoara, Romaniab University of Novi Sad, Faculty of Technology, Cara Lazara 1, Novi Sad, Serbiac University of Belgrade, Faculty of Physical Chemistry, P.O. Box 47, 11158 Belgrade, Serbia

A R T I C L E I N F O

Article history:

Received 12 July 2010

Received in revised form 30 September 2010

Accepted 4 October 2010

Available online 12 October 2010

Keywords:

A. Inorganic compounds

B. Sol–gel chemistry

C. Raman spectroscopy

C. Thermogravimetric analysis

A B S T R A C T

The synthesis of heteropolyacids–mesoporous silica composites was carried out in acidic media by

impregnation and/or by direct incorporation of active phase. The effect of incorporation of

heteropolyacids (HPAs) species into organized mesoporous silica was studied by using non-ionic and

cationic surfactants. A comparison between direct incorporation of HPAs into mesoporous silica and

impregnation of HPAs on mesoporous silica was done. The structure and texture of H3PMo12O40 and

H4PVMo11O40 included on mesoporous silica were studied by XRD, SEM–EDS, FT-IR and Raman

spectroscopies and BET and pore size distribution. Thermal stability was determined by thermo

gravimetric analysis (TGA) and differential thermal analysis (DTA). FT-IR and Raman studies showed that

HPAs anions preserved their Keggin structure after incorporation or impregnation on mesoporous silica

support.

� 2010 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Materials Research Bulletin

journa l homepage: www.e lsev ier .com/ locate /mat resbu

1. Introduction

Heteropolyacids are promising acid, redox and bifunctionalcatalysts in homogeneous as well as in heterogeneous (liquid–solidor gas–solid) systems [1–5]. Especially Keggin type heteropolya-cids (HPAs) have been used extensively in acid-catalysed andoxidation reactions. Pure HPAs generally show low catalyticreactivity owing to their small surface area. In order to be moreeffective for catalytic reactions, HPAs are usually impregnated orincorporated on suitable porous materials with high surface area.Among the supports that have been mentioned in different studiesare: silica [6–8], active carbon [9], titania [10], polymeric materials[11,12] and molecular sieves [13–23].

Mesoporous materials became very attractive as supports forHPAs as they have very large surface area and a high pore volume,which could permit the easy entering of Keggin units into themesopores.

In the literature very few references have been reportedconcerning H3[PMo12O40] and H4[PMo11VO40] supported onmesoporous materials, majority of the studies have been focusedon investigation the most acidic HPA in the series, namely

* Corresponding author. Tel.: +40 256 491818; fax: +40 256 491824.

E-mail address: alpopa_tim2003@yahoo.com (A. Popa).

0025-5408/$ – see front matter � 2010 Elsevier Ltd. All rights reserved.

doi:10.1016/j.materresbull.2010.10.003

H3PW12O40. The main mesoporous materials used as carriers were:MCM-41 [13–17], SBA-15 [18–20], SBA-3 [21] and HMS [22,23].

Toufaily et al. [24] showed that the synthesis of orderedmesoporous silica materials can be performed using variouscationic and non-ionic surfactants and silica sources. They usedthree types of non-ionic surfactants and one type of ionicsurfactant. These materials exhibit structures with wormlikechannels, without a regular long-range periodicity, but withuniform channel diameters over a range comparable with theMCM-41 type materials. They prepared a solid catalyst by directincorporation of 12-tungstophosphoric acid H3PW12O40 (HPW)into organized mesoporous silica. The synthesis of functionalizedmesoporous silica was carried out in an acidic medium in thepresence of HPW, and the structuration of inorganic species wascontrolled either by a mixture of non-ionic and cationic surfactantor only by a non-ionic one.

The aim of this work was to study the direct incorporation ofHPAs H3PMo12O40 (HPM) and H4PVMo11O40 (HPVM) into meso-porous silica during the synthesis. The synthesis of mesoporoussilica containing HPAs was carried out in acidic media by using amixture of cationic and non-ionic surfactants, such as cetyltri-methylammonium bromide (C16TMABr) and Triton (TX-100) orTween 100.

In this study a comparison between direct incorporation ofHPAs into mesoporous silica and impregnation of HPAs on

A. Popa et al. / Materials Research Bulletin 46 (2011) 19–2520

mesoporous silica was done. The effect of the incorporation/impregnation of HPAs on the structure of mesoporous solids wasinvestigated by using different techniques.

The obtained mesoporous materials were characterized by FT-IR and Raman spectrometry, X-ray diffraction at low angles, N2

adsorption–desorption measurements, thermal analysis and SEM–EDS analysis.

2. Materials and methods

2.1. Samples preparation

Molybdo phosphoric acid, H3[PMo12O40]�12H2O was purchasedfrom Merck. H4 [PMo11VO40]�12H2O was prepared by twomethods: Tsigdinos and hydrothermal method [25,26]. The siliconsource was tetraethoxysilane (TEOS) from Fluka. Two types of non-ionic surfactants were used in this study: polyethyleneglycol-4-tert-octylphenylether with 9–10 ethoxy groups (Triton X-100)from Fluka and polyethylene sorbitan–monostearate (Tween 60)from Merck. The cationic surfactant used was cetyltrimethylam-monium bromide C16H33(CH3)3 NBr (CTMABr) from Fluka. Sodiumfluoride NaF, from Fluka, was employed as mineralizing agent.

Organized mesoporous silica-included HPAs (denoted HPA-in-TX 100/CTMABr or Tween 60) with 15% HPA loading, wereprepared by the hydrolysis of tetraethyl orthosilicate using non-ionic and cationic surfactants. The procedure described by Toufailyet al. [24] was applied with some modifications. In the firstsynthesis we used a mixture of TX-100 and CTMABr surfactants,while in the second we used as surfactant only Tween 60.

The first synthesis was performed by sol–gel technique with thefollowing molar composition—1SiO2: 0.22 TX100: 0.04 CTMABr:0.04 NaF: 0.006 HPM: 168 H2O. In the first step, 7.4 g of TX-100 and0.82 g of CTMABr were dissolved in 160 ml distilled watercontaining 4 ml of hydrochloric acid (Hcl 37 wt.%). Then, after aclear solution was obtained, 11 g of TEOS was added and stirreduntil complete dissolution. Then a solution of 0.57 g HPM in 30 mlacidified (HCl) distilled water was added to the first solution andstirred for 4 h at room temperature. The solution was aged for 24 hat room temperature without stirring. A small amount of sodiumfluoride (0.1 g) was added then in order to promote the hydrolysisof TEOS. The solution was furthermore aged at 60 8C for 48 h. Thesolid product was filtered, washed with distilled water and dried inair for 6 h. Calcination for template surfactants removal wascarried out under air by increasing temperature from 25 to 350 8Cwith a rate of 2 8C/min and heating at 350 8C for 4 h.

The second synthesis was performed by the same procedure butwith Tween 60 as non-ionic surfactant. The following molarcomposition was used—1SiO2: 0.067 Tween 60: 0.04 NaF: 0.006HPM: 148 H2O.

Therefore, for the micellar system based on TX 100/CTMABr, theincorporation of HPAs was based on the cationic cosurfactant(CTMABr), while for the second system based on Tween 60, HPAswere introduced by interaction with the protons of the ethoxygroups of the non-ionic surfactant.

The mesoporous silica-supported HPAs were prepared byimpregnation of active phase (HPM and HPVM) on mesoporoussilica support obtained by the same sol–gel procedure as describedearlier, but without addition of the heteropolyacid. In this case, fora better organic surfactants removal, calcination was carried outunder air by increasing temperature from 25 to 550 8C with a rateof 2 8C/min and heating at 550 8C for 4 h. A solution of 15 wt.%HPAs was impregnated on the calcined samples of mesoporoussilica with TX 100/CTMABr and Tween 60, respectively. Theimpregnated samples were denoted as HPA on TX 100/CTMABr orHPA on Tween 60.

The structure and texture of HPM and HPVM included orimpregnated on organized mesoporous silica were studied by XRD,FT-IR and Raman spectrometry, low temperature nitrogenadsorption technique and scanning electron microscopy withenergy dispersive spectrometry (SEM–EDS).

2.2. Measurements of textural properties

Textural characteristics of the outgassed samples wereobtained from nitrogen physisorption using a Quantachromeinstrument, Nova 2000 series. The specific surface area SBET, meancylindrical pore diameters dp and adsorption pore volume VpN2

were determined. Prior to the measurements the samples weredegassed to 10�5 Pa at 250 8C. The BET specific surface area wascalculated by using the standard Brunauer, Emmett and Tellermethod on the basis of the adsorption data. The pore sizedistributions were calculated applying the Barrett–Joyner–Halenda (BJH) method to the desorption branches of the isotherms.The IUPAC classification of pores and isotherms were used in thisstudy.

2.3. XRD analysis

Powder X-ray diffraction data were obtained with a XD 8Advanced Bruker diffractometer using the Cu Ka radiation in therange 2u = 0.5–58 at low angles and 2u = 5–608. The scanning wasmade with a step size of 0.02 and a step time of 2 s.

2.4. Surface characterization by FT-IR and Raman spectroscopies

The FT-IR absorption spectra were recorded with a Jasco 430spectrometer (spectral range 4000–400 cm�1, 256 scans, andresolution 2 cm�1) using KBr pellets.

Micro-Raman spectra were recorded on a DXR Raman Micro-scope (Termo Scientific). The 532 nm line of a diode-pumped solidstate high brightness laser was used as the exciting radiation andthe power of illumination at the sample surface was 10 mW.Collection of the scattered light was made through an Olympusmicroscope with infinity-corrected confocal optics, 25 mm pinholeaperture and standard working distance objective 50� and an1800 lines/mm grating, resolution 4 cm�1. Acquisition time was10 s with 10 scans. The laser spot diameter on the sample was1 mm. Thermo Scientific OMNICTM software was used for spectracollection and manipulation.

2.5. Thermal analysis

Thermal analysis was carried out using a TGA/SDTA 851-LF1100 Mettler apparatus. The samples with mass of about 100 mgwere placed in alumina crucible of 900 ml. The measurementswere performed in dynamic air atmosphere with the flow rate of50 mL min�1, in the temperature range of 25–700 8C with a heatingrate of 10 8C min�1.

2.6. SEM–EDS analysis

Microstructure characterization of the catalyst particles wascarried out with a JEOL JSM 6460 LV instrument equipped with anOXFORD INSTRUMENTS EDS analyser. Powder materials weredeposited on adhesive tape fixed to specimen tabs and then ionsputter coated with gold.

3. Results and discussion

All the N2 adsorption–desorption isotherms of the parentmesoporous silica with TX 100/CTMABr and Tween 60, respective-

[()TD$FIG]

Mesoporous silica with Tween 60

0

100

200

300

400

500

600

700

10.80.60.40.20

Relative pressure, p/po

Vo

lum

e ad

sorb

ed (

cc/g

)

Adsorption

Desorption

HPM in Tween 60

0

100

200

300

400

500

600

10.80.60.40.20

Relative pressure, p/po

Vo

lum

e ad

sorb

ed, c

c/g

Adsorption

Desorption

a

b

Fig. 2. Nitrogen adsorption–desorption plot of mesoporous silica with Tween 60 (a)

and HPM included in Tween 60 (b).

[()TD$FIG]

Mesoporous silica with TX100/CTMABr

0

100

200

300

400

500

600

10.80.60.40.20Relative pressure, p/po

Vo

lum

e ad

sorb

ed (

cc/g

)

Adsorption

Desorption

HPM in TX100/CTMABr

0

100

200

300

400

500

600

700

10.80.60.40.20

Relative presure, p/po

Vo

lum

e ad

sorb

ed, c

c/g

AdsorptionDesorption

a

b

Fig. 1. Nitrogen adsorption–desorption plot of mesoporous silica with TX 100/

CTMABr (a) and HPM included in TX 100/CTMABr (b).

Table 1Textural properties of mesoporous silica and silica included/impregnated HPAs.

Sample Specific

surface

area (m2/g)

Pore volume

BJHDes

(cc/g)

Average pore

diameter

BJHDes (nm)

Mesoporous silica with

TX 100/CTMABr

842 0.88 3.7

Mesoporous silica with

Tween 60

878 1.01 3.6

HPM in TX100/CTMABr 857 0.93 3.7

HPM in Tween 60 762 0.87 3.7

HPVM in TX100/CTMABr 849 0.91 3.7

HPVM in Tween 60 764 0.87 3.7

HPM on TX100/CTMABr 692 0.62 3.1

HPM on Tween 60 724 0.67 3.2

HPVM on TX100/CTMABr 704 0.63 3.1

HPVM on Tween 60 730 0.68 3.3

A. Popa et al. / Materials Research Bulletin 46 (2011) 19–25 21

ly, and HPM or HPVM included/impregnated on mesoporous silicashow a typical adsorption curve of type IV (Figs. 1 and 2Figs. 1a, band 2 a,b). The specific surface area, pore volume and porediameter determined from the isotherms using the BJH method aregiven in Table 1.

For parent mesoporous silica and HPM or HPVM included onmesoporous silica is evidenced an obvious hysteresis loop at arelative pressure of p/p0 = 0.4–0.8. For HPAs impregnated onmesoporous silica composites a narrow hysteresis loop areobserved (not shown).

The pore size distribution curves of parent mesoporous silicawith TX 100/CTMABr and Tween 60, respectively, have narrowpore size distribution within mesopore range with a maximum at37 A and 36 A, respectively.

Generally, after HPM or HPVM incorporation in mesoporoussilica matrix, the pore volume of samples decreased with theconcentration of active phase and also the surface area decreasedwith HPAs loading. A different behaviour is observed for HPM andHPVM included on mesoporous silica with TX 100/CTMABr, as BETsurface and pore volume had almost the same values as the parentmesoporous silica ones (Table 1). A possible explanation for thisunexpected behavior is that the combination of non-ionicsurfactant TX-100 and cationic cosurfactant CTMABr could leadto a very high dispersion of active phase during sol–gel synthesis,and no evidently blockage of the pores by HPA particles takesplace. Toufaily et al. [24] had obtained also higher specific surfacearea for HPW–silica composite prepared by incorporation of 12-tungstophosphoric acid into silica than for parent mesoporoussilica. They suggest that after introduction of HPW into silica, avery regular arrangement of channels with very homogeneousopenings are obtained.

For both HPAs impregnated on mesoporous silica with TX 100/CTMABr and Tween 60, respectively, textural parameters havelower values than for included HPAs ones.

The pore size distribution curves of HPM and HPVM included onmesoporous silica with TX 100/CTMABr and Tween 60, respective-ly, have one maximum within mesopore range at approximatelythe same values as in the case of pure mesoporous silica.

For the HPAs impregnated on both types of mesoporous silica itcould be observed a decrease of average pore diameter togetherwith a significant decrease of specific surface area and pore volume(Table 1).

The surface area and pore size diameter of impregnated HPAsdecreases with HPAs loading due to the partial blockage of themesopores of the support by HPAs particles.

The XRD patterns at low angles for the initial mesoporous silicashow a broad diffraction peak below 2.08 (2u) for both materialsprepared with TX 100/CTMABr and Tween 60, respectively (Fig. 3).The broad peak is centred at 46.0 A for materials prepared with TX

[()TD$FIG]

7654321

4

3

2

1

(1) Mesoporous silica with Tween 60 (2) HPM included in Tween 60 (3) HPM supported on Tween 60 (4) HPVM supported on Tween 60

Inte

nsity

(a.

u.)

2 Theta (º)

Fig. 3. X-ray diffraction pattern of mesoporous silica–heteropolyacids composites.

[()TD$FIG]

400600800100012001400160018002000

1636

1616

595

785

864

965

1064

574

465

800962

108065

4

3

2

1

(1) Tween60 (2) HPM in Tween60 (3) HPM on Tween60 (4) HPVM on Tween60 (5) HPM (6) HPVM

Tra

nsm

ittan

ce, a

.u.

Wavenumber, cm-1

Fig. 5. FT-IR spectra of mesoporous silica with Tween 60–heteropolyacids

composites.

A. Popa et al. / Materials Research Bulletin 46 (2011) 19–2522

100/CTMABr and at 62.7 A for materials prepared Tween 60. Forthe two HPM and HPVM included or impregnated on mesoporoussilica the diffraction peaks at low angles are presented but withdiminished intensity and slightly shifted to higher u values. Thediffraction peak of both included and impregnated HPA appears tohave almost the same intensity. The broad peaks for HPM andHPVM on/in mesoporous silica with Tween 60 composites arecentred at 58.9 A for HPM in Tween 60, 59.3 A for HPM on Tween60 and at 60.1 A for HPVM on Tween 60. It can be asserted that thelong-range order of mesoporous silica is decreased even for aloading of 15 wt.% HPAs.

From XRD measurements at large angles (2u = 5–608) distinctsharp Bragg reflections characteristic for HPAs are not detected forHPAs included or impregnated in/on mesoporous silica (notshown). This is an indication that HPM or HPVM anions aredispersed at the molecular level in the silica matrix and that phasesegregation in silica–HPAs composites under the applied condi-tions does not take place.

The presence of the Keggin anion in organized mesoporoussilica–HPA composites, was confirmed by FT-IR analysis. ThePMo12O40

3� Keggin ion structure consists of a PO4 tetrahedronsurrounding by four Mo3O13 formed by edge-sharing octahedra.[()TD$FIG]

400600800100012001400160018002000

1616

570

595

785

864965

1064

6

5

(1) TX100 (2) HPM in TX100 (3) HPM on TX100 (4) HPVM on TX100 (5) HPM (6) HPVM

43

21

462

800

962

1082

1638

Tra

nsm

ittan

ce, a

.u.

Wavenumber, cm-1

Fig. 4. FT-IR spectra of mesoporous silica with TX 100/CTMABr–heteropolyacids

composites.

These groups are connected with each other by corner-sharingoxygen. This structure gives rise to four types of oxygen, beingresponsible for the fingerprint bands of Keggin anion between1200 cm�1 and 700 cm�1.

The Keggin structure of bulk HPAs is identified by an IRspectrum containing the main vibrations at 1064 cm�1, 965 cm�1,864 cm�1, 785 cm�1 assigned to the stretching vibrations nas P–O,nas Mo55Ot, nas Mo–Oc–Mo and nas Mo–Oe–Mo [27,28]. These bandsare preserved on the mesoporous silica–HPA composites, but theyare broadened and partially overlapped because of the strongabsorption bands of silica (1090, 800 and 462 cm�1) (Fig. 4).

The introduction of heteropolyacids into the mesoporous silicamatrix slightly influenced the structure of resulted composite(Figs. 4 and 5). The vibration band at ca. 1090 cm�1 assigned tonas(Si–O–Si) is decreased to 1082 cm�1 by incorporation of HPAsinto the structure of the silica. The band at ca. 962 cm�1 present inthe spectrum of included and impregnated HPAs samples can beassigned to the nas Mo55Ot stretching vibration. The bands at 800and 462 cm�1 can be assigned to ns(Si–O–Si) and d(Si–O–Si) bonds,respectively [6].

The bands of HPAs included on mesoporous silica in the 1300–400 cm�1 region are partially or completely overlapped by thebands of the silica matrix. The band assigned to the P–Oasymmetric stretching vibration at 1064 cm�1 is completelyoverlapped by the strong band at 1090 cm�1 of the support.

So, two IR bands of pure HPAs are not overlapped by silica bandsin the spectra of included or impregnated HPM and HPVM. Thesebands appeared at 962 and 570 cm�1, which can be assigned to thenas Mo55Ot stretching vibration and to the bending vibration d (O–P–O) of the pure HPAs spectra.

The characteristic bands observed in the low wave numberrange of Raman spectra for pure HPM and HPVM heteropolyacidsare caused by dominantly deformation Mo–O–Mo vibrations andof the entire framework. The other characteristic bands of thespectrum in the range of 900–1100 cm�1 are attributed to thevalence vibrations of the individual M55O groups, the PO4 and thepulsation vibrations of all 12 Mo55O groups.

The Raman spectrum of bulk HPM gives the main bands at 996,983, 882, 603, and 246 cm�1 which are assigned to symmetric (nas)and asymmetric (nas) vibrations of terminal oxygen nas (Mo–Ot)and nas (Mo–Ot), of corner shared bridged oxygen nas (Mo–Ob–Mo),of edge shared bridged oxygen nas (Mo–Oc–Mo) and of oxygen inthe central tetrahedron nas (P–Oa) [27,28]. The Raman band at603 cm�1 is assigned as a combined stretching and bending motion

[()TD$FIG]

60050040030020010085

90

95

100

-4

-2

0

2

-0.0100

-0.0075

-0.0050

-0.0025

0.0000

0.0025

TG

DTG

DTA

DT

G, m

g s

-1

Mas

s, %

Temperature, ºC

Fig. 7. TG-DTG and DTA curves of bulk HPM heteropolyacid.

A. Popa et al. / Materials Research Bulletin 46 (2011) 19–25 23

of the Mo–Ob–Mo bonds of Mo3O13groups, while band at 246 cm�1

is assigned to the symmetric stretch of M–Oa bonds [27,28].The Raman spectrum of silica does not show any vibration band

in this wavenumber region. So, on the spectra of the included orimpregnated samples, only the characteristic bands of pure HPMand HPVM heteropolyacids are observed.

Raman spectra of pure HPAs and HPM and HPVM included in orimpregnated on mesoporous silica with Tween 60 are shown inFig. 6.

In the Raman spectra of HPM and HPVM included orimpregnated on mesoporous silica with Tween 60, the presenceof the strongest bands ns(Mo–Ot) at 989.9 cm�1 (HPM) and at1001.8 cm�1 (HPVM) and ns(Mo–Oa) at 244 cm�1 (HPM) and at240 cm�1 (HPVM) confirm the presence of Keggin anion included/impregnated on the composites. Spectra of HPVM has bands whichare splitted and shifted due to replacement of Mo with V.

As the Keggin unit could be mainly characterized by thestretching mode assigned to metal–oxygen terminal, it can beassumed that the Keggin structure is well preserved in/on themesoporous silica support, after its incorporation or impregnation.This observation is in agreement with FT-IR results.

The main processes observed during the thermal treatment ofparent acids are: the hydrated or crystallisation water eliminationin several steps, the decomposition of the anhydrous acids byconstitutive water removal (all accompanied by endothermic[()TD$FIG]

020040060080010001200

4

3

2

1

(1) HPM (2) HPM in Tween 60 (3) HPM on Tween 60 (4) Tween 60

Inte

nsity

, a.u

.

Raman shift, cm-1

020040060080010001200

4

3

2

1

(1) HPVM (2) HPVM in Tween 60 (3) HPVM on Tween 60 (4) Tween 60

Inte

nsity

, a.u

.

Raman shift, cm-1

a

b

Fig. 6. Raman spectra of pure heteropolyacids and mesoporous silica–

heteropolyacids composites: (a) HPM in/on Tween 60 and (b) HPVM in/on

Tween 60.

effects) and finally the crystallisation process of constitutive oxidesaccompanied by exothermic effects [29,30].

In the region of the hydrated water elimination, the DTA curve(Fig. 7) of bulk HPM shows three endothermic peaks at 72, 105 and116 8C, which may be assigned to bonded water from the crystalhydrates with different number of water molecules [29]. From TGand DTG curves it could be observed the loss of water ofcrystallisation processes accompanied by a considerable weightloss. In the temperature range from 150 to 350 8C for HPM andfrom 150 to 300 8C for HPVM, no weight changes or thermal effectswere observed. The final process evidenced by an exothermic peakat 430 8C is assigned to the crystallisation of constitutive oxides:MoO3 and phosphate oxides.

For both heteropolyacids HPM and HPVM included orimpregnated in/on mesoporous silica from TG, DTG and DTAthermal curves, one can see a different behaviour in comparisonwith pure HPAs in the temperature range corresponding to theelimination of the hydrated water (Figs. 8 and 9).

The first endothermic effect (109.5 8C) is due to the additivethermal effects of the desorbed water from silica surface and to theloss of the first part of the HPA crystallisation water. The secondendothermic effect appears at 224 8C and is due to the loss of thesecond part of the crystallisation water. The loss of the hydratedwateriscompletedat280–300 8C, i.e.attemperatureswhichoverrunthose of the unsupported HPA with 100–120 8C, which is due to theincreased hydrophilic features of the mesoporous silica support.

[()TD$FIG]

800700600500400300200100

86

88

90

92

94

96

98

100

102

-6

-4

-2

0

-0.030

-0.025

-0.020

-0.015

-0.010

-0.005

0.000

Mas

s, %

Temperature, ºC

TG

DTA

DTG

DT

G, m

g s

-1

Fig. 8. TG-DTG and DTA curves of HPM on mesoporous silica with TX 100/CTMABr.

[()TD$FIG]

500400300200100

86

88

90

92

94

96

98

100

102

800700600-6

-4

-2

0

-0.025

-0.020

-0.015

-0.010

-0.005

0.000

DTG

DTA

TG

Mas

s, %

Temperature, ºC

DT

G, m

g s

-1

Fig. 9. TG-DTG and DTA curves of HPM on mesoporous silica with Tween 60.

[()TD$FIG]

Fig. 11. SEM micrographs of mesoporous silica with Tween 60.[()TD$FIG]

Fig. 12. SEM micrographs of HPVM on mesoporous silica with TX 100/CTMABr.

A. Popa et al. / Materials Research Bulletin 46 (2011) 19–2524

In the temperature range 400–700 8C a slow and continuousloss of sample weight are proceeding, owing to the departure ofwater molecules of the HPAs and probably to remaining traces ofthe organic surfactant used for mesoporous silica. The lack of aclear delimitation between the processes of water release – as itwas a continuous loss of hydrated and constitutional water – isprobably due to porous texture of mesoporous silica, which couldcause a delay (or even a blocking) of water elimination. For thesample HPM on TX 100/CTMABr an exothermic peak could beobserved on DTA curve at a temperature higher (680 8C) than thatcorresponding to bulk HPM (430 8C). This exothermic peak couldbe assigned to the decomposition of HPAs and crystallisation ofconstitutive oxides resulted after decomposition. In the case ofHPM or HPVM impregnated on mesoporous silica with Tween 60this exothermic peak could not be observed until 700 8C (Fig. 9).Therefore immobilization on mesoporous silica obviouslyincreases the thermal stability of the Keggin structures incomparison with their parent bulk heteropolyacids.

Electron microscopic studies were performed for mesoporoussilica–HPAs composites using SEM mode. The micrographs ofparent mesoporous silica with TX 100/CTMABr and Tween 60,respectively, and mesoporous silica–heteropolyacids compositesregistered at low magnification (500–10,000�) are displayed inFigs. 10–13.

The SEM image shows that mesoporous silica with TX 100/CTMABr-HPAs composites are composed of spherical particleswith an average diameter of approximately 1–5 mm. The surface

[()TD$FIG]

Fig. 10. SEM micrographs of mesoporous silica with TX 100/CTMABr.

morphology of mesoporous silica–heteropolyacids composites ispractically identical to that of the parent silica.

From the micrographs of Tween 60-HPAs composites registeredat low magnification it can be seen that composites formirregularly shaped larger assembles composed of primary silicaparticles. The average diameter of the primary silica particles is inthe range of 0.5–1 mm.

Additionally, HPAs distribution on samples surface wasanalysed by EDS method. The chemical composition of silicon

[()TD$FIG]

Fig. 13. SEM micrographs of HPVM on mesoporous silica with Tween 60.

A. Popa et al. / Materials Research Bulletin 46 (2011) 19–25 25

from mesoporous silica and of Mo, V and P elements ofheteropolyacids from mesoporous silica–heteropolyacids compo-sites were obtained by this technique. The EDS method wasperformed as point analysis on thin particles and it was acquiredover several domains with 10 mm � 10 mm dimensions, or less, onthe same particle. The analysis was repeated on different particlesof the same batch in order to ensure the reproducibility of theobtained results.

Microanalytical data of EDS analysis show that the molybde-num and phosphorous (15 wt.% HPM in/on mesoporous silica) andmolybdenum, phosphorous and vanadium (15 wt.% HPVM in/onmesoporous silica) content distribution is homogeneous and closeto stoichiometric values.

In the case of HPM included in mesoporous silica with TX 100/CTMABr composite the content of Mo as wt.% is 9.01 (stoichio-metric value is 9.46 wt.%), while P content could not be detected(stoichiometric value is 0.25 wt.%). For the same compositeprepared by impregnation of HPM, the content of Mo is8.65 wt.% (stoichiometric value is 9.46 wt.%), while P contentcould not be detected.

For HPVM included in mesoporous silica with TX 100/CTMABrthe content of Mo is 8.68 wt.% (stoichiometric value is 8.88 wt.%), Pcontent could not be detected (stoichiometric value is 0.19 wt.%)and V content is 0.32 (stoichiometric value is 0.42 wt.%). For thesame composite prepared by impregnation of HPVM, the contentof Mo is 8.45 wt.%, V content is 0.30 wt.%, while P content could notbe detected.

In the case of HPM included in mesoporous silica with Tween 60composites the content of Mo is 9.38 wt.% (stoichiometric value is9.46 wt.%), while P content could not be detected (stoichiometricvalue is 0.25 wt.%).

For HPVM included in mesoporous silica with Tween 60 thecontent of Mo is 8.30 wt.% (stoichiometric value is 8.88 wt.%), Pcontent could not be detected (stoichiometric value is 0.19 wt.%)and V content is 0.35 (stoichiometric value is 0.42 wt.%).

For the same composites prepared by impregnation of HPM orHPVM, the content of Mo and V are lower than in the case ofcomposites prepared by incorporation of HPAs.

It could be observed that impregnated samples exhibit a higherdeviations of Mo and V concentration values from the stoichio-metric ones, while mesoporous silica with included HPAs haveconcentration values closer to the stoichiometric ones. In the latercase it could be supposed that active phase was better homoge-nized by incorporation inside the composites pores than byimpregnation at the surface of the support.

In order to check the correctness of the values for Mo and Vcomposition, the composition values of the support elements,respectively, silicon and oxygen were analysed.

EDS analysis of silica included/impregnated HPM and HPVM onmesoporous silica with TX 100-CTMABr and Tween 60, respec-tively, shows that the concentration of Si is varying between 36.4and 42.7 wt.%, with a mean value of 38.86 (stoichiometric value is39.67 wt.%), while the O content is varying between 51.1 and54.2 wt.%, with a mean value of 53.30 (stoichiometric value is50.6 wt.%).

4. Conclusions

In this study a procedure is described for direct incorporation ofKeggin-type heteropolyacids into ordered mesoporous silica byusing a mixture of cationic and non-ionic surfactants, such ascetyltrimethylammonium bromide (C16TMABr) and Triton (TX-100) or Tween 100.

After incorporation or impregnation, the heteropolyacid anionspreserved their Keggin structure on the surface of mesoporous

silica–heteropolyacid composites and form finely dispersedheteropolyacid species. The mesoporous silica–heteropolyacidcomposites exhibit differential pore size distribution in themesoporosity range. It can be asserted that the long-range orderof mesoporous silica is decreased even for loading of 15 wt.%heteropolyacids.

The favourable effects of heteropolyacid incorporation onmesoporous silica is the increasing of pore volume and specificsurface area, which renders the silica–heteropolyacid compositesappropriate for heterogeneous catalysis. The mesoporous silica–heteropolyacid composites are thermally more stable than theparent acids, due to the strong anion-support interaction.

From energy dispersive X-ray spectroscopy (EDS) analysis, itcould be observed that impregnated samples exhibit a higherdeviation of Mo and V concentration values from the stoichiomet-ric ones, while mesoporous silica included heteropolyacids haveconcentration values closer to the stoichiometric ones. In the latercase it could be supposed that active phase was better homoge-nized by incorporation inside the composites pores than byimpregnation at the surface of the support.

Acknowledgements

These investigations were partially financed by the RoumanianMinistry of Education and Research, grant CNCSIS no. 94 GR and bythe Serbian Ministry of Sciences, grants 142024 and 142047.

References

[1] N. Mizuno, M. Misono, Chem. Rev. 98 (1998) 199.[2] I.V. Kozevnikov, Catal. Rev. Sci. Eng. 37 (1995) 311.[3] F. Cavani, Catal. Today 41 (1998) 73.[4] Y. Ono, in: J.M. Thomas, K.I. Zamaraev (Eds.), Perspectives in Catalysis, 1992, p.

431, Blackwell, London.[5] A. Popa, V. Sasca, J. Halasz, Appl. Surf. Sci. 255 (2008) 1830.[6] C. Rocchiccioli-Deltcheff, M. Amirouche, M. Fournier, R. Frank, J. Catal. 138 (1992)

445.[7] E.F. Kozevnikova, I.V. Kozevnikov, J. Catal. 224 (2004) 164.[8] A. Popa, V. Sasca, E.E. Kis, R. Marinkovic-Neducin, M.T. Bokorov, Ivanka Holclajt-

ner-Antunovic, Mater. Chem. Phys. 119 (3) (2010) 465.[9] I.V. Kozevnikov, Catal. Lett. 27 (1994) 187.

[10] S. Damyanova, J.L.G. Fierro, Chem. Mater. 10 (3) (1998) 871.[11] R. Dziembaj, A. Malecka, Z. Piwowarska, A. Bielanski, J. Mol. Catal. A: Chem. 112

(1996) 423.[12] V.D. Monopoli, L.R. Pizzio, M.N. Blanco, Mater. Chem. Phys. 108 (2008) 331.[13] A. Molnar, T. Beregszaszi, A. Fudala, P. Lentz, J.B. Nagy, Z. Konya, I. Kiricsi, J. Catal.

202 (2001) 379.[14] B.E. Ali, J. Tijani, M. Fettouhi, M.E. Faer, A.A. Arfaj, Appl. Catal. A: Gen. 283 (2005)

185.[15] R.R. Jermy, A. Pandurangan, Appl. Catal. A: Gen. 295 (2005) 185.[16] P.A. Jalil, N. Tabet, M. Faiz, N.M. Hamdan, Z. Hussain, Appl. Catal. A: Gen. 257

(2004) 1.[17] K. Nowinska, W. Kaleta, Appl. Catal. A: Gen. 203 (2000) 91.[18] P.M. Rao, A. Wolfson, S. Kababya, S. Vega, M.V. Landa, J. Catal. 232 (2005) 210.[19] A. Lapkin, B. Bozkaya, T. Mays, L. Borello, K. Edler, B. Crittenden, Catal. Today 81

(2003) 611.[20] P.M. Rao, M.V. Landau, A. Wolfson, A.M. Shapira-Tchelet, M. Herskowitz, Micro-

por. Mesopor. Mater. 80 (2005) 43.[21] K. Nowinska, R. Formaniak, W. Kaleta, A. Waclaw, Appl. Catal. A: Gen. 256 (2003)

115.[22] S. Damyanova, I. Dimitrov, R. Mariscal, J.L.G. Fierro, L. Petrov, I. Sobrados, Appl.

Catal. A: Gen. 256 (2003) 183.[23] F. Marme, G. Coudurier, J.C. Vedrine, Micropor. Mesopor. Mater. 22 (1998) 151.[24] J. Toufaily, M. Soulard, J.-L. Guth, J. Patarin, L. Delmonte, T. Hamieh, M. Kodeih, D.

Naoufal, H. Hamad, Colloids Surf. A: Phys. Eng. Aspects 316 (2008) 285.[25] G.A. Tsigdinos, C.J. Hallada, Inorg. Chem. 7 (1968) 437.[26] F. Kern, St. Reif, G. Emig, Appl. Catal. 150 (1997) 143.[27] A.J. Bridgeman, Chem. Phys. 287 (2003) 55.[28] C. Rocchiccioli-Deltcheff, M. Fournier, R. Frank, R. Thouvenot, Inorg. Chem. 22

(1983) 207.[29] V. Sasca, M. Stefanescu, A. Popa, J. Therm. Anal. Cal. 56 (1999) 569.[30] A. Popa, V. Sasca, M. Stefanescu, E.E. Kis, R. Marinkovic-Neducin, J. Serb, Chem.

Soc. 71 (3) (2006) 235.