Influence of concentration and order of aggregation of the active phases in V–Mo–O catalysts in the oxidative dehydrogenation of propane

Influence of concentration and order of aggregation of the active phases in V–Mo–O catalysts in the oxidative dehydrogenation of propane

Please download to get full document.

View again

of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.

Creative Writing

Publish on:

Views: 49 | Pages: 5

Extension: PDF | Download: 0

  Influence of concentration and order of aggregation of the active phasesin V–Mo–O catalysts in the oxidative dehydrogenation of propane Viviana Murgia a,c, *, Elsa M. Farfa´n Torres b,c , Juan C. Gottifredi a,c , Edgardo L. Sham a,c a Facultad de Ingenierı´ a, Universidad Nacional de Salta, Argentina b Facultad de Ciencias Exactas, Universidad Nacional de Salta, Argentina c  INIQUI – CONICET, Buenos Aires 177, 4.400 Salta, Argentina Available online 5 March 2008 Abstract The activation of alkanes by oxidative route is an alternative way to obtain products with greater added value. The mixed catalysts obtained byimpregnation of Mo and Von different supports conform a potentially attractive system to achieve dehydrogenation of propane. The activity andselectivity depend on the Mo/V ratio used. In this work, we have studied the effect of the concentration and the order of incorporation of the activephases on the catalytic behavior and the nature of the acid sites on the catalyst surface for this reaction. Catalysts with weight contents of 1.4 and2.8% of vanadium and/or 4 and 8% of molybdenum were prepared. The results show that for solids with low vanadium load the order of aggregation of the active phases does not modify the catalytic behavior. When vanadium load increases, greater conversion is observed whenmolybdenum is incorporated in the first place. This behavior can be related to the formation of Mo–V–O species. The catalytic properties are alsoinfluenced by the nature and strength of the acid sites on the surface. # 2008 Published by Elsevier B.V. Keywords:  Vanadium–molybdenum catalyst; Oxidative dehydrogenation; Propane 1. Introduction Selective oxidation reactions have been the subject of several studies to obtain derivatives which can be used inpetrochemical industries. Among the reactions which have notyet been industrially developed, dehydrogenation of lightparaffin could be an interesting choice, due to its many possibleoperative and economic advantages, in contrast with conven-tional dehydrogenation traditional processes. Unfortunately,the addition of oxygen also allows for competing combustionreaction of alkanes and alkenes to carbon oxides. Therefore,current research is focused primarily on the development of anactive ODH catalyst exhibiting high olefin selectivity.Previous studies have been focused on the behavior of vanadium oxide catalysts supported on silica and alumina [1]and molybdenum oxide supported on silica, alumina and titania[2] in the oxidative dehydrogenation of   n -butane. The resultsobtained have shown that, although molybdenum oxides areselective in the olefins formation, they are less active thanvanadium catalysts supported on alumina.According to such results, it is possible to argue thatinteraction between these oxides constitute an efficient systemfor this reaction. Other studies reported show that the catalyticbehavior depends on the used Mo/V ratio in supportedmolybdena–vanadia mixed oxide [3–9]. The catalytic proper-ties of these materials may also be influenced by the type andstrength of the acid sites on the surface [8].In this work, we have studied the influence of both theconcentrationandorderofmolybdenumorvanadiumprecursorsadditionon the catalytic behavior andthe nature of the acid siteson the catalyst in the oxidative dehydrogenation of propane. 2. Experimental For the preparation of mixed oxides,  g -Al 2 O 3  (Aldrich, S  0  = 132 m 2  /g) previously calcined at 550  8 C, is impregnatedwith a dissolution of ammonium heptamolybdate to obtain 4 or8% of Mo. After drying, it is impregnated with an ammonium  Available online at Catalysis Today 133–135 (2008) 87–91* Corresponding autor at: INIQUI [CONICET], Buenos Aires 177, 4.400Salta, Argentina. E-mail addresses: (V. Murgia), Sham).0920-5861/$ – see front matter # 2008 Published by Elsevier B.V.doi:10.1016/j.cattod.2007.12.038  metavanadate dissolution to obtain a load of 1.4 or 2.8% of vanadium.The catalysts obtained are denominated Mo  x V  y  /Al 2 O 3 .Thesame process can be applied to obtain V  y Mo  x  /Al 2 O 3  solids, byinverting the order of impregnation.The solids are dried at 120  8 C during 4 h, with a heatingvelocity of 2  8 C/min, and calcined for 16 h at 550  8 C, at aheating velocity of 10  8 C/min.The specific surface area of catalysts is measured by meansof a Micromeritics Flow Sorb II 2003. The samples arepreviously degassed for 2 h at 200  8 C. Studies of XRD arecarried out in a Powder Diffractometer Rigaku Model Dmax-IIC, using CuK  a  radiation. Studies of surface acidity arecarried out through FTIR spectroscopy of adsorbed pyridine,using a Bruker IFS 88 spectrophotometer. Studies are madewith self-supported thin wafers of 20 mg. Pretreatment is donein vacuum for 30 min at 250  8 C. Once the reference spectrumhas been registered, a pyridine pulse is adsorbed at roomtemperature. Then, it is evacuated for 30 min to 80  8 C; 30 minat 150  8 C, and 30 min at 200  8 C, and the correspondingspectrum is registered in each case. Thermogravimetrical andthermodifferential DTA-TG analyses are carried out in aRigaku TAS 1000 unit. Experiments are done using 20 mg of sample in a static air atmosphere, with a limit temperature of 1000  8 C, and a heating velocity of 10  8 C/min. The solids arealso characterized by Raman spectroscopy, using a Bruker RFS100 spectrophotometer equipped with He–Ne ( l  = 632.8 nm)laser with an OLYMPUS DX-40 microscope. Resolution of spectra is 7 cm  1 , and their acquisition consists of a 30 saccumulation for each sample. Measures of catalytic activityare performed with a fixedbed microreactor.The dimensions of the catalytic bed high are set at 2 cm, and 25/30 mesh quartz isused as an upper and lower counter-bed. The mass of catalystused is 0.4 g. Measures are taken by varying the total flowbetween 150 and 250 mL/min, and the temperature of reactionbetween450and550  8 C.Molarcompositionsoffeedusedinthiswork are C 3 H 8  /O 2  /N 2 :20/15/65 and C 3 H 8  /O 2  /N 2 :4/8/88. Theanalyses of the products in the reactionare carried outusing twogas chromatographers connected in series. Light substances areanalyzed in a Shimadzu GC-3BT chromatographer, and heavysubstances in a Varian 3700 chromatographer. 3. Results and discussion Specific surface of solids is shown in Table 1.It can be observed that the order of aggregation of the oxidesdoes not modify the specific surface area of the resultingcatalyst; while for a fixed concentration of the first oxide, theload increase of the second one leads to a slight decrease of thisproperty.XRD patterns of calcined catalyst showed that all the solidsare amorphous. No crystalline phases of molybdenum orvanadium can be detected, within the technical sensitiveness.These results show that a good dispersion of the oxides on the g -Al 2 O 3  surface take place.All samples present a similar DTA profile, with a wideexothermic event associated with a monotonous weight lossranging from 4 to 4.5% between 90 and 900  8 C, whichcoincides with the observed behavior of the support. Between820 and 900  8 C, there appears a slight exothermic peak, relatedwith the initial weight loss attributed to evaporation of themolybdenum oxide.Studies of pyridine adsorption show that on V 1.4  /Al 2 O 3  andV 2.8  /Al 2 O 3  solids, only acid sites of a Lewis type are present,which are characteristic of the  g -Al 2 O 3  support. Fig. 1 showsthe spectra of adsorbed pyridine on V 1.4  /Al 2 O 3 . The bandsobserved at 1453, 1493, 1577 and 1622 cm  1 correspond topyridine attached to Lewis acid sites [10,11].As it can be observed, the more the temperature of evacuation increases, the significantly lower the intensity of absorption becomes. The intensity of 1616 cm  1 band is notsignificantly altered, but the position of the maximum is shiftedtoward higher frequencies. This could be related with twodifferent types of Lewis acid sites on the catalyst surface [12].Figs. 2 and 3 show the spectra of FTIR of Py adsorbed on theV 1.4 Mo  x  /Al 2 O 3  solids, after evacuation during 30 min at 80  8 C(a), 150  8 C (b), and 250  8 C (c).Pyridine-related to Lewis centers bands are observed. Theband located at 1540 cm  1 , associated with adsorbed pyridineon Bro¨nsted acid sites, can also be detected for both V 1.4 Mo 4  / Al 2 O 3  and V 1.4 Mo 8  /Al 2 O 3 . For V 1.4 Mo 8  /Al 2 O 3  catalyst thisband shows the highest intensity when the temperature of evacuation is 80  8 C [13]. This band vanishes after theevacuation at 150  8 C for V 1.4 Mo 4  /Al 2 O 3 , which indicates thatthese are relatively weak acid sites. For V 1.4 Mo 8  /Al 2 O 3 , greateracid strength is evident due to a permanence of thischaracteristic band after evacuation at the same temperature. Table 1Specific surface of   g -Al 2 O 3 , and catalystsSample  S  0  (m 2  /g) g -Al 2 O 3  132V 1.4  /Al 2 O 3  133V 1.4 Mo 4  /Al 2 O 3  109V 1.4 Mo 8  /Al 2 O 3  105V 2.8 Mo 4  /Al 2 O 3  96V 2.8 Mo 8  /Al 2 O 3  94Mo 8 V 1.4  /Al 2 O 3  104Mo 8 V 2.8  /Al 2 O 3  92Mo 4  /Al 2 O 3  130Fig. 1. FTIR spectra of pyridine adsorbed on V 1.4  /Al 2 O 3  after evacuationduring 30 min at: (a) 80  8 C and (b) 150  8 C. V. Murgia et al./Catalysis Today 133–135 (2008) 87–91 88  Pyridine studies show that the order of aggregation of the oxidedoes not modify the acidity properties of the studied catalyst.Fig. 4 shows the Raman spectra for V 1.4 Mo  x .The band centered near at 1033 cm  1 , spectrum (a), isassigned to the symmetrical stretching mode of the terminalbond V O of isolated monomer groups [3,5,14].The spectrum for V 1.4 Mo 4  /Al 2 O 3  shows a wide, asymmetricband, with a maximum intensity at 939 cm  1 . This band isattributed to the stretching mode of bridges M–O–M bonds inpolymeric surface species of molybdenum or vanadium oxides[3,5]. In the spectrum (c), the band at 847 cm  1 is attributed toantisymmetrical stretching modes of Mo–O–Mo bonds of superficial molybdenum oxides in different configurations(dimmers and olygommers).Bibliographic references show that the bands located at1000 cm  1 are related with the antisymmetrical stretching of terminal groups M O of monomeric and polymeric surfacespecies of vanadium and molybdenum [3,5,8,14]. The fact thatthis band is shifted at 956 cm  1 may involve an additionaldistortion of the species of molybdenum superficial oxideswhich result from the interaction with vanadium cations [3].Fig. 5 shows the Raman spectra obtained for Mo 4 V  x  /Al 2 O 3 .The band located at 1039 and 911 cm  1 , spectrum (a),corresponds to the vibration modes of Mo O and Mo–O–Mo, respectively.The spectrum for Mo 4 V 1.4  /Al 2 O 3  shows a wide band at934 cm  1 . This band presents greater definition as theconcentration of vanadium increases, and a splitting can beobserved with maxima at 939 and 830 cm  1 . The presence of aband at 770 cm  1 could be attributed to V–O–Mo vibrations.This band corresponds to polymolybdovanadate species, whichshows that there is direct interaction of supported species whenmolybdenum is incorporated in the first instance [3,5].Catalytic results show that an increase in the total flow leadsto an extremely low decrease on propane conversion.Theresultsobservedinbothofthefeedratiosusedshowthat,underhigheroxidizingconditionsC 3 H 8  /O 2  /N 2 :4/8/88,thesolidspresent greater conversions. Consequently, a decrease can beobserved in the selectivity of the dehydrogenation products.Tables 2 and 3 present the results obtained for C 3 H 8  /O 2  / N 2 :20/15/65 feed ratio and 200 mL/min total flow. For this feedratio an increase in propene selectivity is observed due to theless oxidative atmosphere used. This also contributes to avoidthe formation of hot spots in the reactor. In ODH reactions theenthalpy of CO 2  production corresponds to about   520 kcal/ mol at 500  8 C while for propene production is about  37 kcal/ mol at the same temperature. Then, if CO 2  formation isdecreased, the risk of reactor run-away is minimized. Fig. 2. FTIR spectra after pyridine adsorption for V 1.4  /Al 2 O 3  after evacuationduring 30 min at: (a) 80  8 C, (b) 150  8 C and (c) 250  8 C.Fig. 3. FTIR spectra of pyridine adsorbed on V 1.4 Mo 8  /Al 2 O 3  after evacuationduring 30 min at: (a) 80  8 C, (b) 150  8 C and (c) 250  8 C.Fig. 4. Raman spectra for (a) V 1.4 Mo 0  /Al 2 O 3 , (b) V 1.4 Mo 4  /Al 2 O 3  and (c)V 1.4 Mo 8  /Al 2 O 3 .Fig. 5. Raman spectra for: (a) Mo 4 V 0  /Al 2 O 3 , (b) Mo 4 V 1.4  /Al 2 O 3  and Mo 4  /V 2.8  / Al 2 O 3 . V. Murgia et al./Catalysis Today 133–135 (2008) 87–91  89  Thereisa slight influence of thecatalyticbehaviorduetothechanges in the order of aggregation of the active phases, forV 1.4 Mo  x  /Al 2 O 3 . However, there is a tendency to increase theconversion when Mo is incorporated in the first instance as canbe observed in Table 2.An increase in the concentration of Mo for V 1.4 Mo  x  /Al 2 O 3 leads to a conversion increase; while olefins selectivity reachesa maximum for V 1.4 Mo 4  /Al 2 O 3 . Fig. 6 shows these results forall the studied temperatures.For V 2.8 Mo  x  /Al 2 O 3 , the solids are more active when Mo isincorporated in the first instance, which confirms the tendencyobserved for V 1.4 Mo  x  /Al 2 O 3 . Table 3 shows these results.At a temperature higher than 470  8 C, a maximumhydrocarbon conversion is reached. This level has a directcorrespondence with the total O 2  consumption, and the resultsobtained show a kinetic effect which is due to the limitation of this reactive.Experimental measures could show overlapping betweenthis effect and those effects related with the structure of thesolids.BycomparingtheresultsobtainedforV  x Mo  y  /Al 2 O 3 ,itcanbeobservedthatastheconcentrationofvanadiumbecomesgreater,conversion also increases with an olefins selectivity loss.Aggregation of the second phase leads to a decrease in thespecific surface area, and a better definition of Raman spectra.As previously discussed, spectra for V 1.4 Mo 4  /Al 2 O 3  andMo 4 V 1.4  /Al 2 O 3  are significantly similar, with a maximumintensitywhich couldbe associated with bothspecies M OandM–O–M. The profile observed in the respective spectrum,together with the increase of the second oxide load, produce agreater resolution of these bands.According to the results obtained from the characterizationof the solids and the catalytic activity, we can conclude that theorder ofaggregation ofthe phasesdoes notmodify the natureortheacidityofthedispersedspeciesonthesurfaceofthealuminafor V 1.4 Mo  x  /Al 2 O 3 –Mo  x V 1.4  /Al 2 O 3  catalysts.For V 1.4 Mo  x  /Al 2 O 3  series, conversion increases when theconcentration of molybdenum oxide phase becomes greater;while selectivity reaches a maximum for V 1.4 Mo 4  /Al 2 O 3 .Studies of pyridine adsorption show that incorporation of molybdenum produces acid sites of a Bro¨nsted type, which canbe more easily detected when there is greater concentration of molybdenum oxide.Production of this type of acid centers involves modificationin the coordination environment of the surface species, thus itaffects catalytic behavior.In the studied catalysts, the greatest olefins selectivity isobtained with those solids containing both types of acid centerson the surface, with moderate acid strength.The more concentration of molybdenum there is, the morethe strength of Bro¨nsted acid sites increases, which is favorableto achieve greater olefin–surface interaction, and therefore, acomplete degradation of combustion products.Forsolids,V 2.8 Mo  x  /Al 2 O 3 –Mo  x V2.8/Al 2 O 3 ,adecreaseinthespecific surface area can be observed, as well as greater propaneconversion when Mo is incorporated in the first instance. Theband at 770 cm  1 for Mo 4 V 2.8  /Al 2 O 3  could be due to thepresence of polymolybdovanadate species, which are respon-sibleforthegreatestcatalyticactivityobservedinthesecatalysts. Table 2Influence of the order of aggregation of the active phase V 1.4 Mo  x  /Al 2 O 3 Temperature ( 8 C) V 1.4 Mo 4  Mo 4 V 1.4  V 1.4 Mo 8  Mo 8 V 1.4 % Conversion % Selectivity % Conversion % Selectivity % Conversion % Selectivity % Conversion % Selectivity470 8.7 75.7 9.6 73.8 13.7 65.1 12.2 69.0500 14.6 67.1 15.7 65.3 20.4 54.5 21.4 55.4550 28.5 53.0 29.9 51.9 29.2 47.6 30.4 48.9Table 3Influence of the order of aggregation of the active phase V 2.8 Mo  x  /Al 2 O 3 Temperature ( 8 C) V 2.8 Mo 4  Mo 4 V 2.8  V 2.8 Mo 8  Mo 8 V 2.8 % Conversion % Selectivity % Conversion % Selectivity % Conversion % Selectivity % Conversion % Selectivity470 17.1 53.8 21.7 54.4 18.5 47.7 25.2 39.0500 26.9 46.3 27.9 46.2 26.2 41.6 26.6 41.0550 29.3 47.7 29.6 48.4 29.4 43.8 28.3 43.5Fig. 6. % Propene selectivity and % conversion for V 1.4 Mo  x  /Al 2 O 3 . V. Murgia et al./Catalysis Today 133–135 (2008) 87–91 90  4. Conclusions For low concentration of vanadium, the order of aggregationof the active phases does not affect catalytic behavior, whichmeans that the nature of the surface species is no substantiallymodified by this difference in the method of preparation.For a greater concentration of vanadium, the order of incorporationoftheactivephasesleadstoalterationofcatalyticproperties, thus greater conversion can be observed whenmolybdenum is incorporated in the first instance. Species Mo–V–O which is formed may be the responsible for this behavior.Incorporation of molybdenum modifies the superficialamorphous phases with an involvement of activity andselectivity of the catalyst. It can be observed that the catalyticproperties of this type of materials are influenced by the natureand strength of the acid sites on the surface. References [1] V. Murgia, E. Sham, J.C. Gottifredi, M. Farfa´n Torres, Latin Am. Appl.Res. 34 (2004) 75.[2] E. Sham, E.M. Farfa´n Torres, V. Murgia, L. Romero, J.C. Gottifredi, XVISimposio Iberoamericano de Cata´lisis, Cartagena, Colombia, 23–28thAug., vol.II, (1998), p. 1285.[3] M.A. Ban˜ares, S.J. Khatib, Catal. Today 96 (2004) 251.[4] H. Dai, A.T. Bell, E. Iglesia, J. Catal. 221 (2004) 491.[5] E. Heracleous,M. Machli,A.A. Lemonidou, I.A.Vasalos,J.Mol. Catal.AChem. 232 (2005) 29.[6] M.A. Ban˜ares, M.V. Martı´nez-Huerta, X. Gao, J.L.G. Fierro, I.E. Wachs,Catal. Today 61 (2000) 295.[7] P. Botella, J.M. Lo´pez Nieto, A. Dejoz, M.I. Va´zquez, A. Martı´nez-Arias,Catal. Today 78 (2003) 507.[8] A. Dejoz, J.M. Lo´pez Nieto, F. Ma´rquez, M.I. Va´zquez, Appl. Catal. AGen. 180 (1999) 83.[9] M.C. Kaezer Franca, R.A. da Silva San Gil, J.G. Eon, Catal. Today 78(2003) 105.[10] S. Riseman,F.Massoth,G.MurallDhar, E.Eyring,J. Phys.Chem.86(10)(1982) 1760.[11] A. Corma, V. Forne´s, E. Ortega, J. Catal. 92 (1985) 284.[12] S. Bagshaw, R. Cooney, App. Spectrosc. 50 (3) (1996).[13] M. Machli, E. Heracleous, A.A. Lemonidou, Appl. Catal. A Gen. 236(2002) 23.[14] P. Kornelak, F. Mizukami, A. Weseluchs-Birczynsca, L. Proniewicz,G. Djega-Mariadassou, A. Bialas, M. Najbar, Catal. Today 90(2004) 103. V. Murgia et al./Catalysis Today 133–135 (2008) 87–91  91
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks