Preparation of nanocrystalline γ-Al2O3 catalyst using different procedures for methanol dehydration

Available online at www.sciencedirect.comJURNLOFScienceDirectNATURALGASCHEMISTRYELSEVIERJourmal of Natural Gas Chemistry 20(2011)334 -38www.elsevler.com/ locatc/jngcPreparation of nanocrystalline ry-Al2O3 catalyst using differentprocedures for methanol dehydration to dimethyl etherAhmad Reza Keshavarz'，Mehran Rezaeil!2*，Fereydoon Yaripour31. Catalyst and Advanced Materials Research Laboraton, Chemical Enginering Department. Faculty of Engineering, Universityof Kashan, Kashan, Iran; 2. Instiue of Nanoscience and Nanotechnology, Universiry of Kashan, Kashan, Iran;3. Catalyst Research Group, Petrochemical Reearch & Technology Compary, NationalPetrochemical Company (NPC), P. O. Box 149650115, Tehran, Iran[Manuscript reeved September 27, 2010; revised November 23, 2010]Abstracttion method (sample B) and sol-gel method using sucrose and hexadecyltrimethyl ammonium bromide (CTAB) as templates (samples C andD, respectively). Textural and acidic properties of y-alumina samples are characterized by XRD, N2 adsorption-desorption and NH3-TPDtechniques.' Vapor-phase dehydration of methanol into dimethyl ether is carried out over these samples. Among them, sample C shows thehighest catalytic activity. NH-TPD analysis reveals that the sample with smaller crystallite size possesses higher concentration of mediumacidic sites and consequently higher catalytic activity. Thermal decomposition method leads to decrease in both surface arca and moderateacidity, therefore it is the cause of lower catalytic activity.Key wordsmethanol dehydratin; dimethyl ether; gama alumina; sol-gel1. Introductionalyst for this reaction. It has high surface area, excellentthermal stability, high mechanical resistance and catalytic ac-Dimethyl ether (DME) has been found to be an altema-tivity for DME formation due to its surface acidity. Re-tive diesel fuel because it has low NOx emission, near-zerocently, many methods have been applied to synthesize alu-smoke amounts and less engine noise compared with tradi-mina with a higher specific surface area and activity for DMEtional diesel fuels [1,2]. It can also be used to replace chlo-synthesis [10].rofluorocarbons (CFCs) which destroy ozone layer of the at-In the present work, the catalytic dehydration of methanolmosphere and used as an intermediate for producing manyto DME has been studied over nanocrysalline r-alumina pre-valuable chemicals such as lower olefins, methyl acetate,pared by four different methods. The effects of preparationdimethyl sulfate and liquified petroleum gas (LPG) alterma-method on the textural, acidic properties and catalytic activitytive. It is also used in power generation and as an aerosolof γ-alumina samples have been investigated.propellant, such as in hair spray and shaving cream, dueto its liquefaction property [3- -7]. Hence, there is a grow-2. Experimentaling demand to produce a large amount of DME to meet theglobal need.Dimethyl ether can be produced by methanol dehydra-2.1. Catalyst preparationtion over a solid- acid catalyst or direct synthesis from syn-gas by employing a hybrid catalyst, comprising a methanol2.1.1. Thermal decomposition methodsynthesis component and a solid acid catalyst [8]. Methanoldehydration to dimethyl ether is a potentially important pro-r-Al2O3 was obtained by the thermal decomposition ofcess and more favorable in the views of thermodynamics andtheb(中国煤化工-600°Cfor6hataheat-economy [9]. Commercially, r-Al2O3 is used as the cat-ingrsample A. .* Corresponding author. Tel: +98-361-5912469; E-mai: Rezaci@kashanu.ac.irHCNMHGThis work was supported by the Petrochemical Research & Technology Company of National Petrochemical Company in Iran.CopyrightO201 1, Dalian Institute of Chemical Physics, Chinese Academy of Sciences. All rights reserved.doi:i0. 1/100-93(10)60157-0Joumal of Natural Gas Chemistry Vol. 20 No.320113352.1.2. Precipitation methodventional flow apparatus, which included an on-line ther-mal conductivity detector (TCD). In a typical analysis, 0.25 gInitally, aluminium nitrate nanohydrate (ANN,of the sample is degassed at 500 °C under helium flow98.5 wt%) was dissolved in water under continuous stiring.(30mL.min-) for Ih. After that, the sample is cooled toThe molar ratio of water to ANN was I. Then precipitation150 °C and then saturated with a mixture of helium and 2%was carried out by adding aqueous ammonia to the aboveammonia for 1.5 h. The sample is then purged with a he-stirred solution. The pH of solution was adjusted to aboutlium flow for 0.5 h to remove weakly and physically adsorbed7.5. After that, the precipitate was filtered and dried at 110°CNH3 on the surface of the catalyst. After this operation, theovernight. The final solid was then calcined in a flow ofsample is cooled to room temperature and then is heated atair from room temperature to 600 °C with a heating rate ofa rate of 10°C-min-l under a flow of helium carrier gas2 °C-min-', and kept at this temperature for 6 h. The obtained(40 mL:min- |) from 25 °C to 700 °C and the amount of am-catalyst is named as sample B.monia in the effluent is measured using TCD and recorded asa function of temperature.2.1.3. Sol-gel method via sucrose as template2.3. Catalytic performanceFirstly, aluminium isopropoxide as an aluminium precur-Vapor-phase dehydration of methanol was conducted insor (AIP, 99 wt%) and sucrose as template were dissolved sep-the Chemical Data Systems (CDS) unit. Figure 1 showsarately in water (molar ratios of AIP : H2O and sucrose :H2Oa simplified flow diagram of the CDS unit. Nitrogen, aswere 1:2). The above solutions were mixed together. Af-ter that, the mixture was peptized using nitric acid (10 wt%)the internal standard, was fed through a set of mass flowunder vigorous strring by carefully adjusting the pH value tocontrollers (Bronkhorst HI-TECH, EL-FLOW), and methanol5.5. The resulting mixture was aged at 80°C for 5h. Thewas pumped from a feed tank through a set of metering pumpssolid product was dried at 110 °C for 12 h in static air and cal-(ILSHIN Autoclave Co., Ltd.). Methanol and N2 were subse-cined in air at 600 °C for 6 h. The obtained catalyst is referredquently introduced into a preheater that was set at a tempera-ture of 250 °C. The temperature of the down stream effluentherein as sample C.was constantly maintained at temperatures above 150°C toavoid the possible condensation of water, methanol, or DME.2.1.4. Sol-gel method via CTAB as templateCatalytic activity was studied under steady state conditions ina fixed-bed reactor (stainlessness steel with an intermal diam-In a typical preparation, firstly aluminum isopropoxideeter of 2.7 mm and a length of 305 mm). The reactor was(AIP, 99 wt%) and hexadecytrimethyl ammonium bromide asheated by three individual furmaces located at top, middle andtemplate (CTAB, 99 wt%) were dissolved in water. The molarbottom sections.ratios of water to AIP and CTAB to AIP were chosen to be90 and 0.8, respectively. After that, the mixture was peptizedMFC: Mass fnow cortollerusing nitric acid (10 Wt%) under vigorous siring by carefullyBPR: Back pressure regulatoradjusting the pH value to 5.5. The mixture was aged at 80°CMFCTIC: Temperature indicatorfor 5 h. The solid product was dried at 110°C for 15 h andGC: Gas chromatograpbfinally calcined at 600°C for 6h. The obtained catalyst is|NitrogenPC: Personal computerreferred herein as sample D.TICI←n Fixed-bedreactorTIC2- 一TIC42.2. Characterization techniquesTIC3 +r ] MethanolThe BET surface area, the total pore volume and theTIC5. +mean pore diameter were measured using a N2 adsorption-EnOmm[cdesorption isotherm at liquid nitrogen temperature(- 196 °C)，PumpBPRusing a NOVA 2200 instrument (Quantachrome). Prior to the Figure 1. A siplied flow diagram of the chenical data sysems (CDS) unitadsorption-desorption measurements, all the samples were de-gassed at 200°C in aN2 flow for 16 h to remove the moisturePrior to the catalytic activity measurements, the sam-and other adsorbates.ples were crushed, sieved to 60- 120 mesh size, and thenThe X-ray diffraction (XRD) patterns of all the cal-treated in situ at a heating rate of 5 °C.min-' under N2 flowcined samples were recorded on a Philips X'Pert (40kV,(50 ml中国煤化工: to 220°C and kept at30 mA) X-ray diffractometer, using a Cu Ka radiation source this ter7 eric pressure. In a typi-YH(入= 1.542 A) and a mickel filter in the 20 range of 109-760.calexpcN M H G.120 mesb) was loadedThe acidity of the samples was measured via temperature-in the middle section of the reactor and methanol was pumpedprogrammed desorption of ammonia (NH3-TPD), using a with different weight hourly space velocities (WHSV = 1.75TPR/TPD 2900 instrument (Micromeritics, USA) with a con-and 11.6h~ ). Activity tests were conducted at 300 °C under336Ahmad Reza Keshavar et al./ Journal of Naural Gas Chemistry Vol. 20 No. 32011atmospheric pressure. The reaction products were analyzeddecreased as follows: sampleA > sample B > sample C >by an on-line gas chromatograph (GC) of Varian CP-3800,sample D. The y-Al2O3 prepared by the addition of CTAB as .equipped with a PoraPlot-Q-HT column to separate the reac-template exhibited the smallest crystallite size among the sam-tion products.ples. Therefore it can be concluded that the templates stronglyaffected the textural properties.3. Results and discussionTextural characterstics of catalysts were listed in Table 1.As described in Table 1, the surface area of samples decreasedThe XRD patterns of the samples (Figure 2) clearly in-as follows: sample D > sample C > sample B > sample A.dicate that the catalysts exhibit the typical 7-phase. Table 1It can be observed that both r-alumina samples D and C pre-presents the crystallite sizes of various samples calculated bypared in the presence of the surfactant showed higher surfaceDeby-Scherer formula. Crystallite sizes of various catalystsarea than the other two catalysts samples A and B. This showsthat the using of template increases the surface area. It meansthat template prevents more intimate contact and aggregationamong alumina particles during preparation. Moreover, su-crose and CTAB acted as chelating agent. During calcining,the chelated complex decomposed, and the produced gasesprevent agglomeration and helped to form pores and fine par-ticles with high surface area in the final products [11,12]. It(2)was also observed that sample D has higher surface area thansample B. On the other hand, the influence of CTAB on the3)surface area is more dominant compared to the effect of su-crose, which is possibly due to the stronger ability of CTABto facilitate the process of template formation. Probably, the1030506070cationic charge of CTAB micelles favors their adsorptions20/(0on aluminium hydroxides species during sol-gel preparation.Figure 2. XRD pttrms of 7r-Al2O3 samples prepared by (1) thermal decom-Therefore CTAB causes a lower degree of solid aggregation.position, (2) precipitation, (3) sol-gel via sucrose as template and (4) sol-gelIt is also possible that CTAB decomposes into more and largervia CTAB as templatemolecules than sucrose during combustion process.Table 1. Physicochemical properties of the prepared 7r-Al2O3 samplesSurface area Pore volume Pore diameter ParticleCrysalliteAcidity (molNH 2)TgPCSample(m2-g-1) (cm3.g-)(nm)size (nm)A__ size (nm)卢wweak & modecrate strong___ peak1 peak 2A1920.448.98.56.32.420.0300.0722324242413.26.4.53.350.0350.0522044182920.547.45.4:2.320.0390.040201406375_0.6116.24.3.1.360.0470.58097389: Determined by BET area; b Delermined by XRD resuts; e Parial sintering facor; Ts: Temperature of desorption maximaThe theoretical particle sizes are also calculated from sur-face area, assuming spherical particles, by the following equa-ψ=(( DBETtion:DxRD,6000DBET =ρxSThe experimental data showed that ψ increased for the sam-ples A and B prepared without surfactant. This is due to severewhere, DBET is the equivalent particle diameter (nm), P is thesintering of their primary crystals.density of the material (g.cm~ "), and S is the specific surfaceAlthough the specific surface area is one of the most im-area (m2.g) [13]. It can be observed that the equivalent par-portant parameters, it must be taken into account that, some-ticle diameter decreased with the addition of surfactant. Thistimes, there is no direct relationship between catalyst activityobservation confirmed the positive effect of surfactant addi-and the physical properties of the catalyst such as surface area.tion in decreasing the particle size. As a result, it can be .Such predictions can be validated by the aid of chemisorptionseen that the particle sizes of samples C and D are smallermeasurements. Here the number of catalytically active sur-than those of samples A and B. Based on aforementioned dis-face中国煤化工re-programmeddesorp-cussion, template prevents intimate contact and aggregationtion-TPD measurements areamong alumina particles during preparation. Hence, particleperfoMYHCNMHGgthandtheamountsofsize decreases in this case. The ψ as a factor is used to reflectacid sites on catalyst surface. Desorption peaks in the rangethe partial sintering extent of the primary crysallites and it isof 180-250°C, 260- -330°C, and 340- -500°C in the NH3-calculated by the following equation:TPD profiles are commonly attributed to NH3 that has beenJournal of Narural Gas Chemisty Vol. 20No. 32011337chemisorbed on weak, moderate and strong acid sites, re-spectively [14]. NH3-TPD of samples is shown in Figure 3.0F @WHSV=1.75h"Temperatures of desorption maxima peaks (Ta) and the acid-8088 WHSV-11.66 h"'ity content of the catalysts are summarized in Table1. Theamounts of weak and moderate acid sites decreased as fol-60lows: sample D > sampleC > sample B > sample A. It isclearly observed that the total amounts of weak and moderate40[acid sites of samples D and C are more than those on the othersamples, and the thermal decomposition of the boehmite pre-20cursor produced a catalyst (sample A) with the lowest moder-ate acid sites. Also among these sarmples, sample D has thehighest proportion of weak and moderate acidic sites.Figure 4. Methanol conversion to DME over ~-Al2O3 samples prepared bythermal decomposition, precipitation, sol-gel via sucrose as template and sol-(4)/gel via CTAB as template in two levels of WHSV3)y4. Conclusions/(2)一This work shows that the textural and acidic properties of7-alumina are greatly influenced by the preparation method.From the catalytic activity and acidity results, it can be con-(1)/cluded that the 7-Al2O3 with the addition of cationic surfac-tant presented a better catalytic performance for the dehydra-tion of methanol to dimethyl ether (DME) in terms of activity00200400500and stability in a wide range of WHSV from 1.75to 11.66 h- .Temperature(C)Furthermore, catalyst with more moderate acidic sites showedFigure 3. NH3-TPD profiles of 7y-Al2O3 samples prepared by (1) thermalhigher activity for this reaction. As a final result, the addi-decomposition, (2) precipitation, (3) sol-gel via sucrose as template and (4)tion of surfactant decreased the crystallite size and increasedsol-gel via CTAB as templatethe surface area. Therefore the acessibility of medium acidicsites increased the amount of favorable active site for the de-Figure 4 shows the catalytic performance for methanolhydration of methanol to DME.dehydration over ry-Al2O3 catalysts at 300°C and WHSV of1.75 and 11.66 h- 1 under steady-state conditions. As indi-Acknowledgementscated in Figure 4, sample A prepared by thermal decomposi-The authors wish to acknowledge Petrochemical Research &tion method shows the lowest catalytic activity among all theTechnology Company of National Petrochemical Company in Iransamples. It can also be observed that samples B, C and Dfor their financial support of this study.show approximately equivalent catalytic activity. The exper-imental results clearly showed that a low WHSV of 1.75 h-1Referencesis not suitable for catalyst screening. Therefore, the cat-alytic actities are investigated under a higher space veloc-[1] Fleisch T H, Basu A, Gradassi M J, Masin J G. Stud Surf Sciity (WHSV=11.66h-). It is noted that the selectivity toCatal, 1997, 107: 117DME over these catalysts is almost 100% under these reaction[2] Semelsberger T A, Borup R L, Greene H L. J Power Sources,conditions. Comparison between activities under different2006, 156(2): 497WHSV shows that the activities of all samples decreased with[3] Vishwanathan V, Jun K W, Kim J W, Roh H S. Appl Catal A,increasing WHSV. Besides, the catalytic activity decreased as2004, 276(1-2): 251follows: sample D > sample C > sample B > sample A.4]CaiGY,LiuZM.ShiRM,HeCQYangLX,SunCL,ChangY J. Appl Catal A, 1995, 125(1): 29Both samples D and C prepared in the presence of surfac-[5] Xu M T, Goodman D W, Batacharyya A. Appl Catal A, 1997,tant showed higher activity than samples A and B. In partic-ular, the sol-gel method with the addition of CTAB produces[6中国煤化工，Kim MI, Y0S,0pa more active catalyst (sample D) than the others. Compari-son of catalytic activity, crystallite size and moderate acidity[7]HCNMHGun K w. Cal Lem, 204,04of samples confirmed that the catalyst with smaller crystallite96(1-2): 23size gives more favorable active sites for methanol dehydra-[8]FeiJH,HouZY,ZhuB,LouH,ZhengXM.APpplCatalA,tion to dimethyl ether.2006, 304(): 49338Ahmad Reza Keshavarz et al/ Journal of Natural Gas Chemistry Vol. 20 No.32011[9] Yaripour F, Baghaei F, Schmidt I, Perregaard J. Catal Commun,[12] Wu Y I, Bandyopadhyay A, Bose s. Mater Sci Eng A, 2004,2005, 6(2): 147380(1-2): 349[10] KimS M, Lee Y J, Bae J W, Potdar H s, Jun K W. Appl Catal [13] Rezaei M, Alavi s M, Sahebdelfar s, Yan z F. Powder Technol,A, 2008, 348(1): 1132006, 168(2): 59[11] Aguado J, Escola J M, Castro M C, Paredes B. Microporous [14] Arena F, Dario R, Parmaliana A. Appl Catal A, 1998, 170(1):Mesoporous Mater, 2005, 83(1-3): 18112中国煤化工MYHCNMHG