Health Articles

 

2pk.ir

J Sol-Gel Sci Technol Mixed matrix vanadium oxide catalytic nanocomposite membranefor styrene oxidation A. L. Ahmad • B. Koohestani • S. Bhatia •B. S. Ooi Received: 25 November 2011 / Accepted: 23 March 2012Ó Springer Science+Business Media New York 2013 Mesoporous nanocomposite membranes with vanadium oxide–carbon nanotubes (VxOy-CNTs) embed-ded in c-Al2O3 were successfully synthesized using the dip Catalytic membranes were first suggested by Sun in 1987 coating method. The membranes were evaluated for sty- Thereafter, Burggraaf [] indicated that c-alumina rene oxidation to determine the optimum styrene conver- membranes could be modified to become catalytic mem- sion and benzaldehyde selectivity. Several factors that branes. Ceramic membranes are technically important for influence the preparation of defect-free coatings, such as filtration and separation. They are also important for cat- the type of binder, the binder addition time and surface alytic reactions with other types of membranes because support treatments, were investigated. The physico-chem- they have unique characteristics, such as chemical stability, ical permeation properties of the membranes were char- long lifetimes, good defouling properties and higher acterized using scanning electron microscope, transmission operating temperatures in chemical and petrochemical electron microscope (TEM), X-ray Diffraction XRD, applications, compared to polymeric membranes Nitrogen adsorption (BET) and Thermogravimetric TGA.
Inorganic catalytic membranes are manufactured mainly Response surface methodology (RSM) was used to inves- from Al2O3, TiO2, ZrO2, SiO2 or mixtures of these com- tigate the effects of oxidant (H2O2) concentration, tem- pounds; of these materials, Al2O3 has received the most perature, contact time and catalyst loading on styrene research attention [–]. Furthermore, c-Al2O3 has been conversion and the selectivity of benzaldehyde. Based on widely studied for the preparation and modification of the RSM analysis, the optimal oxidation conditions inclu- catalytic membranes using the Sol–gel method because ded a reaction temperature of 45 °C, a differential pressure these types of membranes can be used for oxidation reac- of 1.5 bars, a molar ratio of H2O2: styrene of 1.5:1 and a tion purposes , ]. c-Al2O3 membranes are typically catalyst loading of 30 %. These conditions resulted in the supported on macro porous a-Al2O3 tubes or disks in the maximal styrene conversion of 25.6 and 84.9 % benzal- sol–gel method; they are dip or spin coated with a boehmite dehyde selectivity.
(AlOOH) precursor to improve conversion or selectivity].
Catalytic membrane  Nanocomposite However, when AlOOH is applied, the surface onto membrane  Vanadium oxide nanocomposite  Alumina  which c-Al2O3 is catalytically embedded must be free of cracks following calcination. Therefore, choosing a suit-able binder that can be burned off in the calcination stepwithout negative side effects on the catalyst or membraneis important. Although catalytic membranes with differentshapes, catalyst loadings and binders have been prepared A. L. Ahmad (&)  B. Koohestani  S. Bhatia  B. S. Ooi and tested in commercial ceramic support structures School of Chemical Engineering, Engineering Campus, –no studies have investigated the use of vanadium Universiti Sains Malaysia, Seri Ampangan, Nibong Tebal, S.P.S., 14300 Penang, Malaysia xOy) supported on multi-walled carbon nanotubes (MWCNTs) for styrene oxidation in catalytic membranes.
J Sol-Gel Sci Technol Due to their large surface area and outstanding thermal, (PEG) and polyvinyl alcohol (PVA) were obtained from chemical and mechanical stabilities, carbon nanotubes Merck Co., Germany, and a-alumina (AA-04) was pur- (CNTs) are promising for the development of advanced chased from Sumitomo Chemical Co., Japan.
composite materials that undergo catalytic oxidation reac-tions. As a result of their amazing properties, CNTs havebeen introduced into many host materials, including poly- 2.2 Catalyst and membrane preparation mers, metals and ceramics, to improve the overall proper-ties of CNT composite systems [ 2.2.1 Pretreatment of MWCNTs The growth of metal oxides on the walls of carbon nanotubes (CNTs) is a common technique for preparing Because MWCNTs prepared using the CVD method usu- composite materials [The use of vanadium oxide nano- ally contain carbonaceous or metallic impurities, an acid structures as metal oxide catalysts is of particular impor- treatment was employed to remove amorphous carbon and tance because vanadium oxides can be used for the partial metal-oxide impurities introduced during the process [ oxidation or dehydrogenation of alkanes to olefins [– Approximately 4 g of raw MWCNTs were pretreated and Therefore, blends of vanadium oxides and CNTs are oxidized with 400 ml of a 12 M HNO3 solution in a round expected to produce novel composite materials with bottom flask at a reflux temperature of 95 °C for 10 h. Prior enhanced chemical and physical properties.
to the oxidation process, the flask was placed in an ultra- Embedding carbon nanotubes and vanadium oxides into sonic water bath with a frequency of 37 kHz, a power of 150 watt and a temperature of 80 °C for 30 min to modify 2O3 poses several critical challenges that must be overcome to capture the full potential of CNT composites.
the oxygen-containing groups on the surface of the carbon The uniform dispersion of CNTs in the host matrix material nanotubes. The acid-treated MWCNTs were washed with is essential. As a result of van der Waals forces, CNTs are deionized water and ethanol to obtain a final solution pH of likely to form agglomerates or bundles instead of individ- 6-7 and then dried at 130 °C for 24 h.
ual tubes, and this can produce defects in compositemembranes. The second main challenge is associated withstrengthening the interaction between CNTs and the host 2.2.2 Preparation of CNTs-VxOy catalysts matrix to produce stable composite materials.
The oxidation of olefin compounds to equivalent epox- The multi-walled carbon nanotubes impregnated with ides or oxides is an important step in the manufacturing of vanadium oxide (CNTs-VxOy) catalysts were prepared large volumes of fine- and pharmaceutical-grade chemicals using an incipient wetness technique ]. In this tech- [For example, the oxidation of terminal alkenes, such nique, a saturated NH4VO3 aqueous solution (100 ml) was as styrene, is difficult and necessitates longer reaction times refluxed for 24 h to obtain an orange hydrolysis product.
[Styrene oxide (SO) and benzaldehyde are typically The suitable amount of HNO3-treated MWCNTs was manufactured by oxidizing styrene, with stoichiometric floated in the hydrolysis solution, stirred and refluxed at amounts of peracid as the oxidizing agent ]. Styrene 100 °C for 12 h. The suspension was then filtered, washed oxidation in the presence of H several times with deionized water and ethanol, and dried 2O2 is widely used in the production of two significant products, SO and benzalde- at 130 °C overnight. It was then calcified in an N2 atmo- hyde []. To the best of the authors' knowledge, no sphere at a heating rate of 3 °C/min for 5 h until a tem- studies have investigated benzaldehyde production with perature of 500 °C had been reached. This was followed by air heating at 250 °C for 3 h.
xOy-CNT as the catalyst and Al2O3-VxOy-CNT as the catalytic membrane.
2.2.3 Preparation of AlOOH sol (Boehmite sol) 2 Materials and methods Aluminum tri-sec-butoxide (Al(C3H7O)8) was hydrolyzedin an excess amount of deionized water (H2O:Al3? molar ratio of 113:1) at 85 °C; the solution was vigorously stirredfor 1 h in an open container until most of the alcohol MWCNTs, with a purity of C 98 wt%, an average outer evaporated. This was followed by peptization with dilute diameter of 60–100 nm and a length of 5–15 lm, were HNO3 (0.2:1 H?/Al3? mol) to form a stable colloidal sol.
purchased from Shenzhen Nanotech Port Co., China. Nitric The sol was maintained at 80 °C for 24 h under reflux acid (HNO3), aluminum tri-sec-butoxide Al(C3H7O)8, conditions. A boehmite sol, with a pH value of 3.43, was ammonium metavanadate (NH4VO3), polyethylene glycol

J Sol-Gel Sci Technol Fig. 1 Schematic diagram of liquid permeation membrane test rig 2.2.4 Preparation of AlOOH sol with CNTs-VxOy catalysts Table 1 Independent variables and their coded and actual values by addition of PVA following peptization used in the response surface study To prepare a suspension of c-Al2O3 sol embedded with CNTs-VxOy catalysts, 20 ml of boehmite sol was dilutedwith deionized water and then mixed with 0.04 g of CNTs- VxOy and 40 cc of a solution of 0.8 g of PVA and 100 g water. The prepared sol contained 10 % CNTs-VxOy. The Temperature (°C) mixture was then placed in an ultrasonic water bath with a frequency of 37 kHz, a power of 150 watt and a temper-ature of 85 °C for 30 min to disperse the catalyst bundles procedure described in Sect. to form a boehmite sol.
and agglomerates in the sol solution. After sonification, the A boehmite sol, with a pH value of 3.32, was obtained. To solution was stirred for 24 h.
prepare the suspension of c-Al2O3 sol with CNTs-VxOy The binders applied to prepare the sol solution with the catalysts, 22 cc of boehmite sol was diluted with deionized catalyst were polyethylene glycol (PEG) and polyvinyl water and mixed with 0.04 g of CNTs-VxOy.
alcohol (PVA) with average molecular weights of 10,000 To prepare a sol using a binder mixture, 4 g of equal and 30,000–70,000, respectively.
wt% PVA and PEG was mixed with deionized water fol-lowing the procedure described in this section.
2.2.5 Preparation of AlOOH sol with CNTs-VxOy catalysts by addition of PVA/PEG prior to peptization 2.2.6 Preparation and pre-treatment of a-alumina support for coating the catalytic membrane Enough PVA binder to prepare a final solution with 2 wt%PVA was stirred with deionized water at 90 °C for 10 h.
Homemade alumina support pellets were prepared from Alumium tri-sec-butoxide was then added following the alumina powder with a particle size of less than 250 lm.

J Sol-Gel Sci Technol The alumina pellets, which had diameters of 20 mm and containing 0.1 g/l of particles. If particles have a low Zeta thicknesses of 2 mm, were dried and calcinated at 1,000 °C potential value, then there is no force to prevent particle for 3.5 h. The a-alumina disk was then cleaned with aggregation, and there is dispersion instability. The divid- deionized water in an ultrasonic bath to remove unbounded ing line between stable and unstable aqueous dispersions is particles. This was repeated several times until the deion- generally considered to be ?30 or -30 mV. For Al(OH)3, ized water had changed from its initial milky white color to the isoelectric point, the point at which the Zeta potential is a clear solution. Next, the a-alumina disk was cleaned in a zero, occurs at a pH of 7.5 [ mixture of 2-propanol, ethanol, and deionized water in avolumetric ratio of 1:2:2 in an ultrasonic bath with the 2.4 Styrene oxidation: experimental set-up and analysis same conditions. This was repeated several The modified support was then dried at 130 °C for 5 h. The Styrene oxidation and liquid permeation were conducted introduced functional group (-OH) on the surface of the using the experimental set-up shown in the schematics membrane helped to achieve a better coating of CNT-VxOy diagram (Fig. ). The membrane disks were sealed with embedded in c-Al2O3 under acidic conditions.
silicon O-rings in a custom stainless steel permeation cell, A thin layer of CNTs-VxOy-alumina was deposited onto which was placed inside a jacketed heater. The liquid the support using the dip coating technique. The support permeation test rig consisted of four sections: (a) a feed was dipped in the sol for 12 s and then dried at ambient atomizer; (b) a preheater; (c) a reaction chamber; and (e) a temperature for 12 h to obtain a dry gel layer. The dried gel product collector. Prior to the start of an experiment, the layer was calcinated at 550 °C for 3 h to produce a sup- membrane was pretreated at 150 °C under a nitrogen flow ported CNTs-VxOy-alumina membrane.
of 100 ml/min for 1 h to remove all moisture and impuri-ties. Nitrogen gas was used as the carrier gas. N2 gas car- 2.3 Zeta potential and particle size of Al2O3 ried the atomized liquid mixture, which consisted ofstyrene, acetone and hydrogen peroxide, at different tem- To prepare stable colloids, the Zeta potential of the Al2O3 peratures and pressures. To provide pressure differences, sol was evaluated using the laser Doppler velocimetry the support side was drained using a vacuum pump.
technique at different pH values. The pH of the samples The flow rate of the carrier gas streams was measured was controlled using HNO3. The Zeta potential and particle with a mass flow meter GFM (Model FM17, AALBORG, size measurements were performed in dilute suspensions USA). The feed side pressure, which was typically in the Fig. 2 Effect of ultrasonic treatment on coating of membrane a surface d coating with AlOOH ? PVA ? PEG without any treatment on of support before treatment, b surface of support after treatment, support, e coating with AlOOH ? PVA after support treatment, c coating with AlOOH ? PVA without any support treatment, f coating with AlOOH ? PVA ? PEG after support treatment J Sol-Gel Sci Technol Table 2 Zeta potential and mean particle size sol Al2O3 varied between the high level (?1) and the low level (-1),with the center points coded at level 25 and the two outer points corresponding to an a value of 2. The design con- sisted of 6 center points, 8 axial points and 16 fact points,rendering a total of 30 runs, which were used to analyze the experimental data. These data were then used to optimize PVA added to Al2O3 sol after the operating conditions. The measured responses were styrene conversion and benzaldehyde selectivity. The PVA added to sol before mathematical model chosen from the CCD had the highest PVA ? PEG added to sol polynomial order in which the insignificant conditions before peptization were aliased. The design factors and levels involved in theexperiment are shown in Table To improve efficiency, the responses can be related to selected factors by linear or quadratic models. A quadratic range of 100–400 cm3 min-1, was adjusted by regulating model, which includes the linear model, has been reported the carrier gas flow through the atomizer. The vacuum pressure on the permeate side was maintained at 35 mm Hg, and the temperature was maintained at 4 °C using cooling water to condense the solvent. A GC (Agilent technologies, Model 7890A GC system), equipped with where g is the predicted response; X long 9 0.32 mm ID 9 0.25 lm film thickness) and an FID i and Xj are the vari- detector, was used to analyze the feed and permeate 0 is the constant coefficient; bj, bjj and bij are the interaction coefficients of the linear, quadratic and second- streams. After the system reached steady-state, permeate order terms, respectively; and e quality was calculated based on peaks, as determined by i is the error (Eq. the gas chromatograph (GC).
2.6 Physical characterization 2.5 Design of experiment (DOE) The prepared catalysts and membranes were analyzed DOE was used to study the effects of four factors, tem- using a transmission electron microscope (TEM) (Philips perature, differential pressure, catalyst concentration, and CM12 machine equipped with DOCU version 3.2 image ratio of styrene to oxidant, on styrene conversion and analysis systems) at an accelerating voltage of 80 kV and a benzaldehyde selectivity. The range of the study was scanning electron microscope (SEM) (SUPRA TM 35vp chosen based on preliminary studies and a literature review Zeiss instrument equipped with a W-Tungsten filament) to [The central composite design (CCD), with a full investigate the morphology and microstructure of the factorial, was selected using the Design Expert software samples. Nitrogen adsorption (Micromeretics ASAP 2020) (Version 6.0.6, Stat-Ease Inc., MN, USA). Each factor was was used to measure the BET surface area, pore volumeand pore size of the catalyst. The thermal stabilities of thecatalyst and the unsupported membrane were studied by Table 3 The surface area and pore volume of raw materials and thermal gravimetric analysis (PERKIN ELMER TGA7) in air at a heating rate of 10 °C/min from room temperature to Pore volume (cm3/g) a final temperature of 800 °C. A Malvern zetasizer nano between 17.00 and series analyzer (Nano-zs) was used to study the dielectric properties and sizes of the alumina sol.
MWCNTs acid treated 3 Result and discussion PVA added to sol Al2O3 after Figure b show the surface area of the membrane support PVA added to sol Al2O3 before before and after modification with solvent under ultrasonic conditions (described in Sect. , respectively. As shown PVA ? PEG added to sol Al2O3 in these figures, the surface after treatment had more uniform before peptization step particles than with untreated one because of removing

J Sol-Gel Sci Technol unbounded particle in an ultrasonic condition. As seen and isdefinable from the figures, it is clearly understood that thesurface treated with solvent under ultrasonic condition has farbetter and uniformed particles, it shows no unbounded parti-cles, as a result it proves the unique action of the solvent in anultrasonic treatment environment rather and comparing withthe untreated ones due to the removal of the unbounded par-ticles which is the result of the action of the solvent treatedwith homogeneous coated solution, now and therefore anexcellent adsorption site for the acidic solution was adopted,where for and on the supported surface without any treatmentno such result was observed. Figure c, d show the mem-branes coated with sol (AlOOH ? PVA) and (AlOOH ?PVA ? PEG), respectively, on the support surface withoutany treatment. In both samples, binder was added to the solprior to peptization. As shown in the SEM images, the coatingwas not homogenous and had many holes on the surface.
However, for the solvent-treated support, the solution coatingwas homogenous and smooth, as shown in Fig. e, f. Theresults show that the functional group introduced onto themembrane surface after solvent treatment provided excellentadsorption sites for the acidic sol.
Table shows the Zeta potential and mean particle size of the sol Al2O3 with and without binder under different pHconditions. Based on the results, the sol alumina withoutany binder was stable, but when PVA was added afterpeptization, the stability of the sol decreased, as shown bythe reduced Zeta potential of 15.2 mV. However, whenbinder was added to the sol prior to peptization, a morestable sol was obtained as a result of steric repulsion of thesol by the adsorbing polymers. Conversely, the higherpolymer contents contributed by the binder caused mem-brane defects related to shrinkage during the calcinationprocess required to produce an optimum membrane.
The specific surface area and pore volume of the catalyst and the unsupported membrane were analyzed usingnitrogen adsorption at -197°C. The physical characteris-tics of the raw MWCNTs, the acid-treated MWCNTs, theCNTs-VxOy, and three different types of unsupportedmembranes are listed in Table . Based on this analysis, weconcluded that the catalyst and all unsupported membraneshad meso-pores because the pores were larger than 2 nm.
The significant increase in the surface area of acid-treatedMWCNTs can be explained by the opening of pores by theMWCNTs, which makes the internal surface more acces-sible. However, after impregnation, the surface area of theprepared catalysts dropped drastically. This decrease insurface area and pore volume clearly indicates that theVxOy complexes were impregnated inside the MWCNTs.
Conversely, mixing the CNTs-VxOy with alumina sols prepared with different membrane support coating methods b–d adsorption isotherms of unsupported sample

J Sol-Gel Sci Technol Fig. 4 SEM images of CNT-VxOy catalytic membrane TEM image membrane PVA ? PEG add to sol before peptization SEM images of of a MWCNTs, b acid-treated MWCNTs, c the CNTs-VxOy, g membrane, PVA add to sol after peptization h membrane, PVA add d unsupported membrane PVA add to sol after peptization, e unsup- to sol before peptization i membrane, PVA ? PEG add to sol before ported membrane PVA add to sol before peptization, f unsupported the PVA, PEG and boehmite particles occurred duringpeptization. The dispersion of binder inside the solprovided a homogenous surface area. However, whenPVA was added to the sol after peptization, particles couldnot disperse uniformly, increasing the porosity of themembrane.
As shown in Fig. the pore size distribution of the unsupported membrane prepared by adding PVA or PEG atdifferent times could be narrowly controlled. The mem-brane pore size distribution of the membrane prepared byadding PVA prior to peptization was narrower than that of Fig. 5 Weight loss curve for catalyst and unsupported membrane the membrane prepared by adding PVA to the sol afterpeptization. The pore size distribution of this membrane increased the specific surface area because Al2O3 possesses was narrow compared to that of the membrane prepared by a greater surface area and a more porous structure than adding the PVA and PEG mixture as a binder during VxOy. The addition of PVA and PEG prior to peptization peptization. The addition of binder molecules to the alu- resulted in a higher surface area for the unsupported mina sol prior to peptization resulted in a uniformly dis- membrane. Using a mixture of PVA and PEG as a binder tributed sol. Conversely, the addition of PEG increased the produced larger pore volumes than the use of PVA alone.
pore volume, pore diameter and BET surface area. As This result indicates that more hydrogen bonding among shown in the adsorption isotherms of Fig. d, monolayer

J Sol-Gel Sci Technol Fig. 6 XRD patterns ofunsupported membrane withdifferent condition ofpreparation formation was usually complete when the relative pressure the CNTs-VxOy catalyst inside the membrane was random.
reached 0.3 because the radius available for condensation However, the distribution of catalyst inside the membrane was decreased by the thickness of the monolayer or by prepared by adding PVA (Fig. or a mixture of PVA and approximately two molecular diameters. Based on the PEG (Fig. i) after peptization was much better than that results, the samples with relative pressures above 0.3 had inside the membrane prepared by adding PVA after pep- type IV hysteresis according to the IUPAC classification, indicating the presence of mesoporous materials related to The thermogravimetric weight loss curves for the CNT- cylindrical pores open at both ends ].
VxOy and unsupported membranes are shown in Fig. The TEM images (Fig. show that the acid-treated The thermal gravity curve in air for the CNT-VxOy catalyst MWCNTs had an open-ended structure and some defects shows that a major weight loss occurred at temperatures following acid treatment. The pentagonal defects and the greater than 500 °C, corresponding to the decomposition of curved dimensional area are important because they can the CNT template and indicating that 49 wt% was a result increase reactivity as a result of the tubes' structural of VxOy loading []. As shown, the higher weigh instability []. Figure c demonstrates that VxOy was ts associated with the residue were a result of the VxOy and successfully distributed on the surface and within the c-Al2O3 on the surface of the support.
CNTs. It can be noticed from the image that the VxOy For all the unsupported membranes, the initial losses of impregnated inside MWCNTs. The walls of MWCNTs are physically bound water and volatile materials were similar presented with randomly margins of dark and bright and occurred at temperatures below 150 °C. The weight regions. The bright margins represent CNTs, while dark losses at temperatures between 250 and 280 °C coincided margins (few nm) represent the vanadium oxide nanopar- with the evolution of H2O, CO2, and ethylene oxides, based ticles, which are intercalated between the CNTs layers.
on the decomposition of PVA or PEG, and weight losses at Figure d–f show the unsupported membranes prepared temperatures between 400 and 450 °C were due to the with different binders under different conditions. Figure d phase change of boehmite to c-Al2O3. This occurred at shows the membrane for which PVA was applied as a temperatures near 400 °C, and the losses of water were a binder after peptization of the Al2O3 sol. Figure f show result of reactions with the small amount of carbon residues the membranes for which PVA and a mixture of PVA and from the binder , PEG, respectively, were added to the sol Al2O3 prior to 2A1OOH ! c - A12O3 þ H2O: peptization. The distribution and homogeneity of particlesin the unsupported membrane prepared by adding PVA or a Figure shows the XRD patterns of the unsupported mixture of PVA and PEG to the sol Al2O3 prior to pepti- membrane prepared with the sol Al2O3 and different zation were much better than those of the membrane pre- binders in the presence of CNTs-VxOy. It was calcinated in pared by adding PVA after peptization. As shown in air at a heating rate of 3 °C/min and remained at 550 °C for Fig. large holes were present in the unsupported 3 h. The XRD patterns confirm that the composites were a membrane when PVA was added after peptization.
hybrid of c-Al2O3, VxOy and MWCNTs. The XRD results Scanning electronic microscope (SEM) images of the indicate that the mixed composition of the products CNTs-VxOy membrane supported on a-alumina are shown matched the patterns of different types of V2O5.H2O and in Fig. As shown in these figures, the orientation of 6VO2.5H2O. The peaks at 2h = 26.2° and 2h = 50.6° J Sol-Gel Sci Technol Table 4 The experimental data for the oxidation of styrene Amount of styrene of each experiment is 2 mol Amount of solvent 1,000 ml Table 5 ANOVA table for styrene oxidation Styrene conversion Benzaldehyde selectivity DF degree of freedom, CV coefficient of variation, std. Dev. standard deviation

J Sol-Gel Sci Technol Fig. 7 Response surface plotshowing the effect of catalystpercentage and differentialpressure, and their mutual effecton the conversion of styrene(a, b) and on benzaldehydeselectivity (c, d) and styreneoxide selectivity (e, f) correspond to the (110) and (020) diffraction planes of presence of c-Al2O3, VxOy and MWCNTs in the composite V2O5, respectively [].
membranes and indicate that the choice of binder did not The peaks at 2h = 23.5 and 2h = 26.3° correspond to affect the membrane properties.
MWCNTs. The peaks overlap with the V2O5 peak (110),corresponding to the reflection peak (002) associated withthe inner layer spacing of nanotubes (d002 = 0.338 nm).
4 Model fitting and analysis of variance (ANOVA) Most peaks also had low intensities, indicating that theV2O5–MWCNTs nanocomposite had a microcrystal struc- The experimental data related to the oxidation of styrene to ture ]. The XRD patterns at 2h = 46.2° and 67° cor- benzaldehyde using supported membranes and an appli- respond to c-Al2O3 ]. For all the preparation methods, cation of PVA ? PEG prior to peptization are given in similar XRD patterns were found. These results confirm the Table The ANOVA results show that the reaction of

J Sol-Gel Sci Technol Fig. 8 Scheme styreneoxidation styrene with H2O2 is best described by a quadratic poly- result of data variation in the fitted model [The lack- nomial model. The quadratic models for styrene conversion of-fit F-values of 3.14 for styrene conversion and 2.59 for (Eq. ) and benzaldehyde selectivity (Eq. ) are given benzaldehyde selectivity are within the acceptable error Figure displays the effect of catalyst loading on con- Styrene ðconversion %Þ ¼ 11:68 þ 3:70A  7:25B version and product selectivity during oxidation in the þ 3:05C  3:00D  0:43A2 catalytic membrane. As expected, increasing catalyst þ 1:49B2 þ 2:36C2 þ 1:60D2 loading (i.e., the amount of catalyst by wt% on the surface  6:33AB þ 3:54AC þ 0:57AD of membrane) from 10 to 30 % resulted in higher styrene  3:93BC þ 0:59BD þ 1:17CD conversion because there was more contact between the precursor and the catalyst. Conversely, benzaldehydeselectivity and styrene oxidation decreased at higher cata- Benzaldehyde ðselectivity %Þ ¼ 64:47  2:02A  2:63B lyst loadings. This occurred because the product could be converted to by-products, according to the mechanism þ 1:33A2  3:04B2 proposed in Fig. , at longer contact times between the þ 0:027C2  2:45D2 benzaldehyde or styrene oxide and the catalyst. Figure þ 1:67AB  9:82AC shows the scheme formation mechanism of the six þ 0:63AD þ 13:89BC observed products. As shown in Fig. benzaldehyde was þ 2:16BD  5:18CD produced through direct oxidative cleavage of the styrene double bond or through nucleophilic attack of the styrene Table shows that both the styrene conversion and oxide [catalytic reaction was solely caused by benzaldehyde selectivity terms were less than 0.05 of the VxOy, as the alumina particles, acid wash MWCNTs, alu- probability of the F values, indicating significance of the mina membrane and mix matrix alumina membrane model results. The integrity of the model fit was quantified include MWCNTs had no effect on styrene conversion. No using the determination coefficient (R2) [In this case, sign of reaction was detected for the membrane and par- the R2 values for styrene conversion (0.9014) and benz- ticles in the presence of H2O2, demonstrating that the aldhyde selectivity (0.9032) show that only 9.86 % of reaction is merely catalyzed by VxOy and MWCNTs can styrene oxidation and 9.68 % of benzaldehyde selectivity provide channel for shorter contact time of reactant with were not explained by the model. This discrepancy was the J Sol-Gel Sci Technol Fig. 9 Response surface plotshowing the effect of oxidantand differential temperature,and their mutual effect on theconversion of styrene(a, b) andon benzaldehyde selectivity(c, d) and styrene oxideselectivity (e, f) Styrene conversion and product selectivity at various The effects of H2O2 concentration on the conversion of differential pressures (i.e., the pressure between the catalyst styrene and product selectivity are presented in Fig. and the permeate) are presented in Fig. At higher dif- When the H2O2: styrene ratio was increased from 1.5:1 to ferential pressures, styrene conversion rates were relatively 4:1, the styrene conversion obviously decreased. This could lower with improved benzaldehyde and styrene oxide have been a result of the H2O2 shielding effect caused by selectivity. This result also shows that extended reaction the formation of styrene in the H2O2 emulsion. When the times between the product and the catalyst produced amount of H2O2 was increased, the shielded styrene by-products such as benzoic acid and benzene, which molecules were prevented from contacting the catalyst decreased the selectivity of both benzaldehyde and styrene surface, therefore reducing the reaction rate. This effect was more pronounced at low temperatures because at low J Sol-Gel Sci Technol Fig. 10 FTIR spectra of polystyrene Table 6 Limitations adjusted to reach the optimal value for styrene membrane surface, as shown by the FTIR results in oxidation to benzaldehyde Table shows the optimum differential pressure (delta p), temperature, catalyst and oxidant concentration condi-tions for styrene oxidation and benzaldehyde selectivity.
To validate the model, a series of experiments was con- ducted under the following predetermined conditions: a reaction temperature of 45 °C, a differential pressure of 1.5 bar; an H2O2: styrene molar ratio of 1.5:1 and a catalyst concentration of 30 %. An optimum styrene conversion of 25.6 % and a benzaldehyde selectivity of 84.9 % were temperatures, the emulsion was relatively more stable. At high H2O2 concentrations and low temperatures, the stableH2O2 emulsion on the catalyst surface prevented the further Nanocomposite membranes were successfully prepared by reaction of styrene oxides; as a result, benzaldehyde embedding VxOy and CNTs into a c-Al2O3 matrix using selectivity was low. Because of the instability of the the dip coating method. The structure and morphology of emulsion and the further reaction of styrene oxides with the the catalytic membranes synthesized on the alumina sup- catalyst to produce benzaldehydes at high temperatures, port were extremely sensitive to the pretreatment method.
benzaldehyde selectivity increased.
Methods such as ultrasonic cleaning, modifying surface The styrene conversion rates and product selectivity at functional groups and adding a binder could be employed different reaction temperatures are shown in Fig. As to prepare uniform membranes. The coated membranes had expected for low temperatures, the benzaldehyde and sty- a smooth, homogeneous surface layer, and no cracking or rene oxide selectivity were higher, and the styrene con- delaminating phenomena were observed between the version rates were lower. In the range of 45–65 °C, styrene underlying membrane and the coating layer. The catalytic conversion increased as the temperature increased. This contactor membrane oxidized styrene to benzaldehyde at a occurred because the increased temperatures caused more styrene conversion rate of 25 % and a benzaldehyde styrene oxide and benzaldehyde molecules to decompose selectivity of 85 % under optimum conditions, which to other by-products. Further increasing the temperature included a temperature of 45 °C, an H2O2:styrene molar ([65 °C) caused a sudden flux reduction. This was due to ratio of 1.5:1, a differential pressure of 1.5 bar and a cat- the polymerization of styrene to polystyrene ] on the alyst loading of 30 %.
J Sol-Gel Sci Technol The authors gratefully acknowledge the Minis- 21. Xu JF et al (2002) Nonlinear optical transmission in VOx nano- try of Science, Technology and Innovation (Malaysia) for providing tubes and VOx nanotube composites. Appl Phys Lett 81(9):1711– financial support (2008 science fund 03-01-05-SF0401) for this study.
22. Weckhuysen BM, Keller DE (2003) Chemistry, spectroscopy and the role of supported vanadium oxides in heterogeneous catalysis.
Catalysis Today 78(1–4 SPEC.):25–46 23. Lutz JT, Grayson M, Eckroth D, Bushey GJ, Eastman CI, Klingsberg A, Spiro L (eds) (1980) Kirk–Othmer Encyclopedia 1. Sun YM, Khang SJ (2002) A catalytic membrane reactor: its of Chemical Technology, vol 9, 3rd edn.Wiley, NewYork, performance in comparison with other types of reactors. Ind Eng Chem Res 29(2):232–238 24. Coperet C, Adolfsson H, Sharpless KB (1997) A simple and 2. Uhlhorn RJR, Huits In't Veld MHBJ, Keizer K, Burggraaf AJ efficient method for epoxidation of terminalalkenes. Chem (1989) First international conference on inorganic membranes.
Montpellier, France, pp 323–328.
25. Swern D (1971) OrganicPeroxides, vol 2. Wiley Interscience, 3. Saracco G (1998) Structured catalysts and reactors. Marcel Dekker, New York, pp 463–500 26. Wang L et al (2005) Direct synthesis, characterization of 4. Fain DE (1994) Inorganic membranes: the new industrial revo- Cu-SBA-15 and its high catalytic activity in hydroxylation of lution, ICIM394, July 10–14, 1994 phenol by H2O2. J Mol Catal A: Chem 230(1–2):143–150 5. Han SH et al (2008) Control of pore characteristics in carbon 27. Selvaraj M, Song SW, Kawi S (2008) Epoxidation of styrene over molecular sieve membranes (CMSM) using organic/inorganic mesoporous Zr-Mn-MCM-41. Microporous Mesoporous Mater hybrid materials. Desalination 233(1–3):88–95 6. Rahman MA et al (2011) Development of a catalytic hollow fibre 28. Al-Ajlouni AM et al (2008) Kinetic studies on phenylpho- membrane micro-reactor for high purity H2 production. J Membr sphopolyperoxotungstates catalyzed epoxidation of olefins with Sci 368:116–123 hydrogen peroxide. J Mol Catal A: Chem 287(1–2):159–164 7. Tsuru T (2008) Nano/subnano-tuning of porous ceramic membranes 29. Hulea V, Dumitriu E (2004) Styrene oxidation withH2O2 over for molecular separation. J Sol-Gel Sci Technol 46(3):349–361 Ti-containing molecular sieves with MFI, BEA and MCM-41 8. Julbe A, Farrusseng D, Guizard C (2001) Porous ceramic mem- topologies. Appl Catal A 277(1–2):99–106 branes for catalytic reactors—overview and new ideas. J Member 30. Datsyuk V et al (2008) Chemical oxidation of multiwalled carbon Sci 181(1):3–20 nanotubes. Carbon 46(6):833–840 9. Dong J et al (2008) Microporous inorganic membranes for high 31. Chen X et al. (2007) Carbon nanotube-induced preparation of temperature hydrogen purification. J Appl Phys 104(12): vanadium oxide nanorods: application as a catalyst for the partial oxidation of n-butane. Mater Res Bull 42:354–361 10. Bottino A et al (2005) Vapour phase oxidation of toluene in 32. Du X et al (2009) Influences of pH value on the microstructure V/Al2O3-TiO2 catalytic reactors. Catal Today 99(1–2):171–177 and phase transformation of aluminum hydroxide. Powder 11. Alfonso MJ, Mene´ndez M, Santamarı´a J (2000) Vanadium-based Technol 192(1):40–46 catalytic membrane reactors for the oxidative dehydrogenation of 33. El-Masry MH, Sadek OM, Mekhemer WK (2004) Purification of propane. Catal Today 56(1–3):247–252 raw surface water using electro-coagulation method. Water Air 12. Xia C et al (1996) Boehmite sol properties and preparation of Soil Pollut 158(1):373–385 two-layer alumina membrane by a sol-gel process. J Membr Sci 34. Guin D, Baruwati B, Manorama SV (2005) A simple chemical synthesis of nanocrystalline AFe2O4 (A = Fe, Ni, Zn): An effi- 13. Alfonso MJ, Mene´ndez M, Santamarı´a J (2002) Oxidative cient catalyst for selective oxidation of styrene. J Mol Catal A: dehydrogenation of butane on V/MgO catalytic membranes.
Chem 242(1–2):26–31 Chem Eng J 90(1–2):131–138 35. Box G (1987) Statistics for experiments. An introduction to 14. Alfonso MJ et al (1999) Oxidative dehydrogenation of propane on design, data analysis and model building. Wiley, New York V/Al2O3 catalytic membranes. Effect of the type of membrane and 36. Donohue MD, Aranovich GL (1999) A new classification of reactant feed configuration. Chem Eng Sci 54(9):1265–1272 isotherms for Gibbs adsorption of gases on solids. Fluid Phase 15. Teixeira M et al (2011) Improving propyne removal from pro- pylene streams using a catalytic membrane reactor-a theoretical 37. Wildgoose GG, Banks CE, Compton RG (2006) Metal nano- study. J Membr Sci 375(1–2):124–133 particles and related materials supported on Carbon nanotubes: 16. Deshayes AL, Miro´ EE, Horowitz GI (2006) Xylene isomeriza- methods and applications. Small 2(2):182–193 tion in a membrane reactor: part II. Simulation of an industrial 38. Iijima S (1991) Helical microtubules of graphitic carbon. Nature reactor. Chem Eng J 122(3):149–157 17. Thostenson ET, Ren Z, Chou T-W (2001) Advances in the sci- 39. Serp P, Corrias M, Kalck P (2003) Carbon nanotubes and ence and technology of carbon nanotubes and their composites: a nanofibers in catalysis. Appl Catal A 253(2):337–358 review. Compos Sci Technol 61(13):1899–1912 40. Ahmad A et al. (2010) Synthesis of vanadium oxide nanotubes 18. Sharma A, Tripathi B, Vijay YK (2010) Dramatic improvement via an ultrasonic method. J Sol-Gel Sci Technol 56(3):1–6 in properties of magnetically aligned CNT/polymer nanocom- 41. Mishra R, Rao KJ (2000) Thermal and morphological studies of posites. J Membr Sci 361(1–2):89–95 binary and ternary composites of poly(vinylalcohol) with alumina 19. Neri G et al. (2010) In: Malcovati P et al. (eds) A novel approach and zirconia. Ceram Int 26(4):371–378 for the preparation of metal oxide/CNTs composites for sensing 42. Van Gestel T et al (2002) Alumina and titania multilayer mem- applications, in sensors and microsystems, Springer Netherlands, branes for nanofiltration: preparation, characterization and chemical stability. J Membr Sci 207(1):73–89 20. Ledoux MJ et al (2001) High-yield butane to maleic anhydride direct 43. Wu G-M et al (2008) Investigation on properties of V2O5– oxidation on vanadyl pyrophosphate supported on heat-conductive MWCNTs composites as cathode materials. J Sol-Gel Sci materials: b-SiC, Si3N4, and BN. J Catal 203(2):495–508 Technol 46(1):79–85 J Sol-Gel Sci Technol 44. Han Y-F et al (2007) Effects of different types of c-Al2O3 on the 47. Hulea V, Dumitriu E (2004) Styrene oxidation with H2O2 over activity of gold nanoparticles for CO oxidation at low-tempera- Ti-containing molecular sieves with MFI, BEA and MCM-41 tures. J Phys Chem C 111(7):3163–3170 topologies. Appl Catal A 277(1–2):99–106 45. Kshirsagar AC, Singhal RS (2007) Optimization of starch oleate 48. Jang BN, Wilkie CA (2005) The thermal degradation of poly- derivatives from native corn and hydrolyzed corn starch by styrene nanocomposite. Polymer 46(9):2933–2942 response surface methodology. Carbohydr Polym 69(3):455–461 46. Gao YL et al (2007) Investigation of the effects of food con- stituents on Bacillus subtilis reduction during high pressure andmoderate temperature. Food Control 18(10):1250–1257

Source: http://www.2pk.ir/pedex/upload/article/1406010336.pdf