Oxidative dehydrogenation of alkanes over vanadium-magnesium-oxides

Oxidative dehydrogenation of alkanes over vanadium-magnesium-oxides

/ A PA LE IY D CP AT L SS I A: GENERAL ELSEVIER Applied Catalysis A: General 157 (1997) 105-116 Oxidative dehydrogenation of alkanes over vanadium...

639KB Sizes 0 Downloads 63 Views

Recommend Documents

Oxidative Dehydrogenation of Alkanes over V-based Catalysts: Influence of Redox Properties on Catalytic Performance
The activity and selectivity of the catalysts used in oxidative dehydrogenation of alkanes results from a combination of

Oxidative dehydrogenation of cyclohexane over heteropolymolybdates
The oxidation of cyclohexane in presence of molecular oxygen has been studied at 350 and 400°C over heteromolybdates cat

Oxidative dehydrogenation of ethane over LaF3CeO2 catalysts
A new series of very selective rare earth oxyfluoride LaF3CeO2 catalysts was prepared and used for the oxidative dehydro

Oxidative dehydrogenation of cyclohexane over vanadate catalysts
The oxidative dehydrogenation of cyclohexane was studied at 440 and 484°C over NdV04, Mg3(VO4)2, and a mixed VMg0 cata

Transient response studies of isobutane oxidative dehydrogenation over molybdenum catalysts
Kinetics and mechanism of isobutane oxidative dehydrogenation were studied over cobalt and nickel molybdate catalysts. T

Oxidative dehydrogenation of propane over chromium-loaded calcium-hydroxyapatite
Chromium-loaded hydroxyapatite catalysts Cr(x)/CaHAp (0.1 ≤ x ≤ 3.7 wt.% Cr) were prepared by ion-exchange and character

Oxidative Dehydrogenation of Ethane over a Perovskite-Based Monolithic Reactor
The oxidative dehydrogenation (ODH) of ethane has been investigated in a short-contact-time reactor consisting of a LaMn

Oxidative dehydrogenation of ethane over BaF2 promoted SmO3LaF3 catalysts
Oxidative dehydrogenation of ethane (ODE) was investigated on Sm2O3LaF3 and BaF2 promoted Sm2O3LaF3 catalysts. Under the




Applied Catalysis A: General 157 (1997) 105-116

Oxidative dehydrogenation of alkanes over vanadium-magnesium-oxides ~g

H.H. K u n g , M.C. K u n g lpatieff Laboratory, Center for Catalysis and Surface Science, Northwestern University, 2137 Sheridan Road, Evanston, IL 60208, USA


Vanadium-magnesium-oxides are among the most selective and active catalysts for the oxidative dehydrogenation of alkane. The selectivity for dehydrogenation depends strongly on the alkane, the Mg vanadate phase, the presence of modifiers, and to a lesser extent, the method of preparation. The results of studies on these variables are reviewed and discussed with respect to the current understanding of the nature of the active site and requirements for selective dehydrogenation. Keywords: Oxidative dehydrogenation; Alkane oxidation; Vanadium-magnesium-oxide;Selective oxidation of


1. Introduction There has been a strong interest to study the oxidative dehydrogenation of ethane, propane and butane because of the potential commercial interest to utilize alkane effectively. Ideally, for oxidative dehydrogenation, the formation of alkenes is the only reaction Eq. (1). In practice, however, nonselective oxidation to carbon oxides Eq. (2) and other products also occurs. CnH2n+2 + 12-O2 ---* CnH2n + H 2 0


Cnnzn+2 + ½(3n + 1)O2 ~


n C O 2 + (n + 1 ) H 2 0

The mixed oxides of vanadium and magnesium is the first system reported to show high selectivity for oxidative dehydrogenation of alkane [ 1,2]. Since the first detailed report in 1988 [1], many subsequent studies have resulted in substantially deeper understanding of the catalyst system. This paper reviews the current * Corresponding author. Fax: +1-708 4671018; e-mail: [email protected] nwu.edu. 0926-860X/97/$17.00 © 1997 Elsevier Science B.V. All rights reserved. PII S0926-860X(97)00028-8


H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116

understanding. Particular emphasis will be placed on studies that aim at increasing the understanding of the nature of the active sites and factors that determine selectivity for dehydrogenation.

2. The V-Mg-O system V205, being an acidic oxide, reacts readily with the basic MgO. Depending on the composition of the mixture, various magnesium vanadates can be formed. The phase diagram of this system [3,4] indicates that magnesium orthovanadate (Mg3(VO4)2), magnesium pyrovanadate (Mg2V207), and magnesium metavanadate (MgV206) are stable compounds. The structure of Mg3(VO4)2 is characterized by chains of edge-sharing MgO6 units linked together by isolated VO4 tetrahedra [5]. a-MgEV207 is made up of comer-sharing VO4 tetrahedra that form V207 units [6], and MgV206 is made up of metavanadate chains of edge-sharing VO5 units [7]. Various preparation techniques can be used to prepare Mg vanadates. Extended high temperature calcination of a mixture of V205 and MgO produces the Mg vanadate phase that corresponds to the stoichiometry of the mixture. For example, with a 1 : 3 ratio of V205 and MgO, although the initial phase observed was Mg pyrovanadate, the final stable phase was Mg orthovanadate [8]. Such solid state methods of preparation produce low surface area materials. High surface area materials can be prepared by impregnation of ammonium metavanadate or vanadium oxalate onto Mg(OH)2 [9], MgO [1,10], or Mg oxalate [11]. MgO can be derived from Mg carbonate, Mg hydroxide, or Mg oxalate [11]. After impregnation, calcination at 550°C produces high surface area, supported vanadates. The resulting material is not well-crystallized, as indicated by broad IR peaks of the solid [1]. Stoichiometric vanadates can be prepared by this method when mixtures of appropriate compositions are used. However, even with a stoichiometric mixture, pure phases can only be obtained by calcination at high temperatures (700°C or higher) [9]. Heating at lower temperatures results in incomplete transformation, and thus the presence of mixed phases [9,11,12]. The reaction between vanadium and magnesium precursors has been investigated using differential thermal analysis [9-11]. For a mixture containing 40 wt% V205, for example, heat evolution attributed to the formation of ot-Mg2V207 was observed at around 400°C, whereas that for Mga(VO4) 2 at around 600°C [9]. In all cases, in spite of the high vanadium oxide contents, there is no evidence of endothermic events that correspond to the melting of V205. This is consistent with the high reactivity between V205 and MgO. The absence of V205 supported on MgO is confirmed by spectroscopic techniques. IR and Raman spectra, as well as X-ray diffraction of the samples show only the presence of Mg vanadate phases [1,9,10]. In no case was there evidence of V205 crystallites or dispersed surface phase of O=VO3 units.

H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116


V - M g - O samples prepared by calcination to 600°C show EPR signals of V 4+ [9,10,13] that is present in small concentrations, possibly due to anion vacancy defects. 51V NMR spectra indicate the presence of V ions in different environments [13]. From the NMR parameters, it was concluded that there are four different surface V ions. These are isolated V ions with a tetrahedral coordination, present primarily in samples of low V contents. As the V content increases, V ions with a tetrahedral coordination that interact with other V ions appear. The third form is V ions in an octahedral environment. Again, with increasing V content, these ions are observed to interact with other V ions. These V ions are on the surface, because they interact with gaseous H20 and CO2. At V loadings higher than 5 wt%, V ions characteristic of those in Mg3(VO4) 2 appear. For a sample containing 20 wt% V205, the formation of Mg3(VO4)2 was readily observed with 51V NMR [14]. The ease of formation of this compound was attributed to the fact that during impregnation of the vanadium precursor onto MgO, MgO was transformed into Mg(OH)2. This resulted in occlusion of some vanadium ions in the bulk of the catalyst, which facilitated formation of Mg vanadate compounds [ 11].

3. Oxidative dehydrogenation of alkane on V - M g - O catalysts V - M g - O catalysts of compositions ranging from low V content to one equivalent for the formation of Mg pyrovanadate (69 wt% V205) have been studied for the oxidative dehydrogenation of ethane [ 15], propane [9,10,16-18], butane [ 1,19], 2-methylpropane [15], and cyclohexane [20]. In general, addition of V to MgO increases the activity and the selectivity for dehydrogenation significantly. The selectivity for dehydrogenation is also much higher than for V205. The extent of enhancement, however, depends on the composition of the catalyst and, to a lesser extent, its pretreatment. Interestingly, the catalytic behavior also depends on the alkane. Among the alkanes, the oxidative dehydrogenation of propane is the most studied. Fig. 1 illustrates some of the data published in the literature for propane. It can be seen that similar catalytic behavior was observed by different investigators, when the selectivity for propene is plotted as a function of propane conversion. This similarity was obtained with catalysts of different V contents and MgO prepared with different precursors (Mg oxalate, carbonate, or hydroxide). As will be discussed later, these variables cause minor variations in the catalytic behavior. The apparent small variations are a result of the fact that the selectivity-conversion relationship in the oxidative dehydrogenation of propane is quite similar over Mg orthovanadate and Mg pyrovanadate [9,13,16,21]. Thus, although different amounts of Mg orthovanadate and pyrovanadate were present in catalysts of different preparations, their catalytic behavior did not differ significantly.


H.H. Kung, M.C, Kung/Applied Catalysis A: General 157 (1997) 105-116

80 X

70 •







-> 5o I-(J LU "~ 4 0 LU






Z 30



n" 2 0 13.

10 0













PROPANE CONVERSION, % Fig. 1. Illustrative catalytic behavior of V-Mg-O-<) catalysts in the oxidative dehydrogenation of propane. Source of data: ( x ) - ref. [17]; (+) - ref. [16]; ( V ) - ref. [9] (ex Mg hydroxide); ( l l ) - ref. [11] (ex Mg oxalate); ( 0 ) - ref. [9,20] (Mg pyrovanadate); (*) - ref. [11] (ex Mg carbonate); and ( A ) - ref. [20] (Mg orthovanadate).

The situation for other alkanes is different, as shown in Table 1. The selectivities for dehydrogenation are similar for ethane over both Mg orthovanadate and pyrovanadate, but the orthovanadate is much more selective for the dehydrogenation of butane and 2-methylpropane. In addition, at low conversions, the ratio of the rates of consumption of oxygen molecules to alkane is about two for ethane, propane, butane, 2-methylpropane, and cyclohexane on Mg orthovanadate (Fig. 2) [15]. On Mg pyrovanadate, this ratio is close to two for ethane and propane, but close to four for butane and 2-methylpropane. Since these ratios are the average number of oxygen molecules that react with each hydrocarbon molecule, they were termed average oxygen stoichiometry (AOS). To explain these data, as well as similar ones obtained with a VPO catalyst, it was proposed that in the formation of primary products in the oxidative dehydrogenation of alkane, there is a selectivity-determiningstep involving the reaction of an adsorbed alkyl [15]. (An adsorbed alkyl is the first reaction intermediate


H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116 Table 1 A comparison of the selectivity for alkenes over Mg orthovanadate and Mg pyrovanadate Alkane

C2H6a C3Hsc C4Hlo c i-C4H1oa

Mg orthovanadate

Mg pyrovanadate

Temp. (°C)

Alkane conv. (%)

Select. (%)


Temp. (°C)

Alkane conv. (%)

Select. (%)


540 541 540 500

5.2 6.7 8.5 8

24 64 65.9 b 64

2.1 2.1 2.5 2.1

540 505 500 502

3.2 7.9 6.8 6.8

30 61 31.8 b 25

2.1 2.1 3.9 4.1

Data from ref. [15]. The data for Mg orthovanadate were obtained with 40 VMgO, which contained Mg orthovanadate and MgO. b Sum of selectivity to butenes and butadiene. c Data from ref. [28]. d Average oxygen stoichiometry, see text.


~ 0.8





~0.5 "o 0.4

o.2! o.a!


















Hydrocarbon reacted/hydrocarbon feed Fig. 2. Relationship between rates of oxygen consumption and alkane consumption. Mg orthovanadate: ( l l ) ethane; ( A ) - propane; ( V ) - 2-methylpropane; (V) - butane; ( 0 ) - cyclohexane. Mg pyrovanadate: (+) ethane; (x) - propane; (*) - 2-methylpropane; and (JRI) - butane.

formed by abstraction of a H atom from an alkane molecule [12].) The number of surface VOx units, which presumably form the surface active site, that can effectively interact with the adsorbed alkyl determines the AOS. The effective size of such a unit is determined by both the size of the adsorbed hydrocarbon species as well as the rate of reoxidation of the vanadium active center. Thus, for Mg orthovanadate, because the VO4 units are isolated from each other (i.e. sufficiently far apart), adsorbed ethyl, propyl, or butyl species can only interact


H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116

with one surface V O 4 unit. If one assumes that each V O 4 unit supplies a certain number of oxygen atoms to react with an adsorbed hydrocarbon molecule, then the reaction of these alkanes would all show the same AOS. From the experimental data, the number of oxygen atoms is four (i.e. AOS equals two). On the other hand, the surface of Mg pyrovanadate contains V207 groups, which are pairs of cornersharing VO4 units. The close proximity of two VO4 units makes it possible for larger alkyl groups to interact with both units. The result is that twice the number of oxygen atoms become available to react with each larger hydrocarbon molecule. Thus, the AOS for butane on Mg pyrovanadate is twice that for ethane. This model can be extended to explain the data on Mg metavanadate (MgV206) [ 12]. Mg metavanadate is made up of chains of edge-sharing VO5 units. Therefore, it would be possible for larger molecules, such as butane, to interact with more than one VO5 unit. The AOS would be high, and the catalyst would show low selectivity for dehydrogenation in the oxidation of these molecules. Indeed, the selectivity is only 14% at 13% butane conversion [12]. On the other hand, its selectivity for dehydrogenation in the oxidation of the smaller propane molecule was nearly as high as for Mg orthovanadate and pyrovanadate [18]. The slightly poorer performance of Mg metavanadate for propane conversion than the other two vanadates may be understood in terms of the ease of replenishment of lattice oxygen. In the geometric model described above, it is implicitly assumed that each VO4 unit would supply a fixed number of oxygen atoms to react with the alkyl species adsorbed on it. Clearly, this is a simplistic assumption. Nonetheless, one expects that this number depends primarily on the environment of the VO4 unit, i.e., how easily can oxygen anion vacancies be generated at this site, and how fast can the vacancies be replenished by migration of oxygen ions from the nearby positions in the lattice, relative to the surface residence time of the hydrocarbon species. At present, there is little information about these rate processes. But one might expect that they would be related to the redox properties of the cations in the vanadate. Thus, if the vanadate is composed of a more easily reducible cation, it would be able to form oxygen anion vacancies more easily and migration of oxygen anions in the lattice is also more rapid. These would lead to a larger AOS at the VO4 site, and the catalyst would form more oxidized products (on the average). That is, the selectivity for oxidative dehydrogenation would be lower. Indeed, this has been observed [22,23]. When oxidative dehydrogenation of butane was investigated over vanadates ofMg, Sm, Nd, Zn, Cr, Eu, Ni, Cu, and Fe, there is a general correlation between selectivity for dehydrogenation with the reduction potential of the cation: the more easily reducible the cations are, the less selective the catalyst is (Table 2). The only notable exception is Cr, which may be due to the presence of Cr 6+ on the surface. Not all cations that are difficult to be reduced form selective catalysts. In fact, Mg is the only alkali earth or alkaline cation that forms selective catalysts. Neither Ca, Ba, nor K form selective catalysts [19,24]. This is because except for Mg, all the other alkali earth and alkaline cations form carbonates that are stable under

H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116


Table 2 Correlation between reduction potential of the cation in orthovanadate and selectivity for oxidative dehydrogenation of butane Cation

Reduction potential (V)

Butane conv. (%)

Dehydrog. select. (%)

Mg Nd Sm Zn Cr Eu Ni Cu Fe

-2.40 2.30 -2.30 -0.76 -0.42 -0.35 -0.26 +0.32 +0.77

5.2 8.5 7.9 6.6 5.5 5.5 7.3 6.9 6.5

62 62 59 41 17 41 18 4 17

reaction conditions. Thus, under reaction conditions, the vanadate decomposes into V205 and the corresponding carbonate.

4. Reduction of V-Mg-O catalysts The model described above assumes that the proximate source of oxygen that reacts with the adsorbed hydrocarbon species is surface lattice oxygen. Gaseous oxygen participates only after adsorption in other parts of the catalyst and then migration through the lattice to the active site. At present, no isotopic labelling experiment has been performed to confirm this assumption. On the other hand, indication that this is the case is provided by pulse reaction experiments and electrical conductivity measurements. When pulses of butane were passed over Mg3(VO4)2, selective production of dehydrogenation products was obtained (Fig. 3) [25]. The selectivity was comparable to the value for steady state reactions, and was maintained until an equivalent of about three monolayers of lattice oxygen was removed. When a similar experiment was performed over a much less selective catalyst of Ni3(VO4)2, a similar observation was obtained. In this case, the selectivity for dehydrogenation products was much lower, consistent with the values in the steady state reaction

[25]. The electrical conductivities of Mg orthovanadate, pyrovanadate, and metavanadate all show zeroth order dependence on oxygen partial pressure at reaction temperatures, which suggests that there is little adsorbed oxygen species on these solids [26,27]. However, upon exposure to propane or hydrogen, reduction of the solid occurs rapidly, and correspondingly the electrical conductivity increases. Likewise, exposure of a reduced solid to oxygen results in rapid decrease in electrical conductivity, showing that the solid is reoxidized rapidly. These observations point to the fact that the Mars and van Krevelen mechanism applies to the propane oxidation reaction on these solids [26]. In addition, Mg pyrovanadate was


H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116

100 90 80'

o~ 70 0



, v






(9 CO

50 O) C~ 0 ¢(9

40 30 20 10 0 0

I 5








% Reduction (Surface Equivalent) Fig. 3. Selectivity for dehydrogenation in pulse reaction of butane over V-Mg-O as a function of the degree of reduction of the catalyst. Data from ref. [25].

found to undergo reduction and reoxidation at rates faster than Mg orthovanadate and Mg metavanadate, which could be correlated to its higher oxidative dehydrogenation selectivity observed in that study [26,27].

5. Modified V-Mg-O The catalytic properties of V - M g - O catalysts can be modified by the presence of other elements, use of supports, and use of different precursors. The presence of K adversely affects the catalytic performance [28]. As shown in Table 3, samples prepared with MgO that was obtained by precipitation with KOH showed lower selectivity for dehydrogenation than samples using ammonium carbonate for precipitation. The difference was due to traces of K left in the samples after preparation. By XPS analysis, it was shown that K was preferentially segregated onto the surface of Mg vanadates. Deliberate impregnation of 1 wt% K onto a

H.H. Kung, M. C. Kung /Applied Catalysis A: General 157 (1997) 105-116


Table 3 Effect of K on the selectivity for oxidative dehydrogenationa Catalyst


Conv. (%)

Dehydrog. select. (%)

Mg3(VO4)z ex (NH4)2CO3 Mg3(VO4)/ex KOH Mg3(VO4)z ex (NI4_4)2CO3 Mg3(VO4) 2 ex KOH V - M g - O (40 wt% V20~) V-Mg--O (40 wt% V205)+1 wt% K

Butane Butane Propane Propane Butane Butane

8.5 7.2 6.7 3.5 27 20

65,9 [email protected] 63.6 53.6 61 53

a Data taken from ref. [28].

sample also lowered the selectivity (Table 3). The presence of K also affected the rate of formation of Mg orthovanadate from a mixture of MgO and (NH4)VO3. After an identical heat treatment, a sample without K formed Mg orthovanadate readily, whereas a sample with a trace amount of K formed a mixture of Mg orthovanadate and Mg pyrovanadate [9,28]. A similar deleterious effect of K has also been observed by others [29]. Modification by incorporation of Mo into Mg orthovanadate has also been studied [30]. It is possible to substitute Mo into the lattice position of V to form compounds of the formula Mg3-xVa-2xMo2xO8, which retains the structure of Mg orthovanadate when x<0.03. At higher concentration of vacancies, the crystalline lattice rearranged into a new structure, that of Mga.sVMoO8. The catalytic properties of these compounds, or their "self-supporting" mixtures are similar to Mg orthovanadate, except that the selectivity for dehydrogenation is slightly lower, and there are more cracked products, a result of the higher acidity due to the presence of Mo ions. The effect of surface acidity has also been observed for vanadia catalysts on different supports [14]. In most of the preparations of V - M g - O catalysts, very high calcination temperatures were avoided in order to obtain catalysts of high surface areas. As a result, it is likely that more than one crystallographic phases are present in the working catalyst. The effect of co-existence of phases was investigated, and it was found that the presence of o~-Mg2V207 or MgO enhances the selectivity for dehydrogenation of propane over Mg3(VOa)a [18]. A similar observation was reported in another study [11]. It was found that the two V-Mg-O catalysts with the highest selectivity for dehydrogenation contained a mixture of two phases, both present in significant amounts: MgO and Mg orthovanadate in one, and Mg orthovanadate and pyrovanadate in another. At present, the reason behind this phenomenon is not understood. Variations in the catalytic properties due to the preparation method has been investigated [11]. Catalysts prepared by impregnating different vanadium precursors (ammonium metavanadate (AV) or vanadium oxalate (VO)) onto Mg oxalate or MgO prepared from Mg oxalate, carbonate, or hydroxide showed somewhat different selectivities for dehydrogenation (Fig. 1). One observation


H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116

was that for different preparations, the ratio of various crystalline phases differed. For example, for catalysts containing about 18-19% V205, the crystalline phases detected on the sample prepared with VO and MgO (ex hydroxide) was mostly Mg orthovanadate with a small amount of MgO. For the sample prepared with AV and MgO (ex carbonate), somewhat more crystalline MgO was detected, and for the one prepared with AV and MgO (ex oxalate), nearly equal amounts of Mg orthovanadate and MgO were detected. Since there might be synergistic effect of the mixed phases, as described in the paragraph above, it is not surprising that these samples behaved differently as a catalyst. The authors did not interpret their results as due to presence of mixed phases [1l]. Instead, they reported an interesting correlation that the higher the binding energy of the Ols peak in XPS, the higher is the selectivity for dehydrogenation. They suggested that this indicates the role of nucleophilicity of the oxygen ions on their tendency to preferentially promote C-H bond abstraction. Modification of V-Mg-O by small amounts of Nb, Cr, or Sm only resulted in catalysts of lower dehydrogenation selectivities [29]. This is not surprising, in view of the fact that the selectivity correlates inversely with the ease of reduction of the cations, and Mg ion is among the most difficult to be reduced. Catalysts selective for dehydrogenation can be prepared by impregnating vanadia onto Mg-silicate [31]. As expected, because of the strong acid-base interaction between MgO and V205, compounds of Mg vanadate were detected in samples of higher V contents. Concurrently, the catalytic activity and selectivity for dehydrogenation also increased. Since the test reaction was propane oxidation, for which selective dehydrogenation could be obtained with Mg orthovanadate and Mg pyrovanadate, samples containing a wide range of V contents were found to be quite selective. Attempts to support the V - M g - O catalyst on A1203 and TiO2 were made [32]. In all cases, a mixture of Mg orthovanadate, Mg pyrovanadate, and vanadia unassociated with Mg was found to be present on the support. The relative amounts of the different phases depended on the strength of interaction of V ions with the support, with isolated surface VOx species observed when the interaction is strong. These isolated surface VOx species are also active sites selective for dehydrogenation. Finally, a Mg vanadium aluminophosphate (MgVAPO-5) has been prepared and tested to be selective for dehydrogenation [33]. However, it is not yet certain that under reaction conditions, the Mg and V ions remain as zeolite framework ions.

6. Conclusion

Recent activities in the preparation and characterization of the V-Mg-O catalysts for the oxidative dehydrogenation of alkane have advanced the understanding of this catalytic system. It is now understood that a desirable feature of

H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116


this catalytic system is the strong interaction of MgO with V 2 0 5 , resulting in the formation of Mg vanadates. In these compounds, the isolated or small VOx units formed are desirable for high dehydrogenation selectivity, because they can supply only a limited number of oxygen atoms for reaction with adsorbed hydrocarbon species. While some other supports could also form isolated VOx units, they could do so only at low V concentrations. The fact that the selectivity for dehydrogenation depends strongly on the hydrocarbon is very interesting but not well understood. It appears that a better understanding of this phenomenon would be very informative towards the design of highly selective catalysts, and improvement of the V-Mg-O system.

Acknowledgements Support of this work by the US Department of Energy, Basic Energy Sciences, Division of Chemical Sciences is gratefully acknowledged.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26]

M.A. Chaar, D. Patel, M.C. Kung and H.H. Kung, J. Catal., 105 (1987) 483. H.H. Kung, M.A. Chaar, US Patent 4772319 (1988). G.M. Clark and R. Morley, J. Solid State Chem., 16 (1976) 429. V.A. Matveevicheva, Z.I. Ezhkova, B.E. Zaitsev and A.G. Lyubarskii, Russ. J. Phys. Chem., 43 (1969) 143. N. Krishnamachari and C. Calvo, Canad. J. Chem., 49 (1971) 1629. R. Gopal and C. Calvo, Canad. J. Chem., B30 (1974) 2491. H. Ng and C. Calvo, Canad. J. Chem., 50 (1972) 3619. G.M. Clark and R. Morley, J. Solid State Chem., 16 (1976) 429. D. Siew Hew Sam, V. Soenen and J.C. Volta, J. Catal., 123 (1990) 417. J. Hanuza, B. Je~owska-Trzebiatowska and W. Oganowski, J. Molec. Catal., 29 (1985) 109. A. Corma, J.M. Ltpez Nieto and N. Paredes, J. Catal., 144 (1993) 425. D. Patel, M.C. Kung, H.H. Kung, in: M.J. Phillips, M. Ternan (Eds.), Proceedings of the 9th International Congress on Catal., Chem. Institute of Canada, 1988, p. 1554. O.B. Lapina, A.V. Simakov, V.M. Mastikhin, S.A. Veniaminov and A.A. Shubin, J. Molec. Catal., 50 (1989) 55. T. Blasco, J.M. Ltpez Nieto, A. Dejoz and M.I. V~zquez, J. Catal., 157 (1995) 271. P.M. Michalakos, M.C. Kung, I. Jahan and H.H. Kung, J. Catal., 140 (1993) 226. M.A. Chaar, D. Patel and H.H. Kung, J. Catal., 109 (1988) 463. R. Wang, M. Xie, E L i and C. Ng, Catal. Lett., 24 (1994) 67. X. Gao, E Ruiz, Q. Xin, X. Guo and B. Delmon, J. Catal., 148 (1994) 56. A. Corma, J. M. L6pez Nieto, N. Paredes, A. Dejoz, I. Vazquez, in: V. Cortts Corberfin, S. Vic Bell6n (Eds.), New Developments in Selective Oxidation II, Elsevier, Amsterdam, 1994, p. 113. M.C. Kung and H.H. Kung, J. Catal., 128 (1991) 287. K. Seshan, H.M. Swaan, R.H.H. Smits, J.G. van Ommen, J.R.H. Ross, in: G. Centi, E Trifirb (Eds.), New Developments in Selective Oxidation, Elsevier, Amsterdam, 1990, p. 505. O.S. Owen and H.H. Kung, J. Molec. Catal., 79 (1993) 265. O.S. Owen, M.C. Kung and H.H. Kung, Catal. Lett., 12 (1992) 45. D. Patel, P.J. Andersen and H.H. Kung, J. Catal., 125 (1990) 130. O.S. Owen, Ph.D. Thesis, Northwestern University, 1992. V. Soenen, J.M. Herrmann and J.C. Volta, J. Catal., 159 (1996) 410.


H.H. Kung, M.C. Kung/Applied Catalysis A: General 157 (1997) 105-116

[27] A. Guerrero-Ruiz, I. Rodriguez-Ramos, J.L.G. Fierro, V. Soenen, J.M. Herrmann, J.C. Volta, New Developments in Selective Oxidation by Heterogeneous Catalysis, Stud. Surf. Sci. Catal. 72 (1992) 203. [28] M.C. Kung and H.H. Kung, J. Catal., 134 (1992) 668. [29] R.X. Valenzuela, E.A. Mamedov and V. Cort6s Corber~, React. Kinet. Catal. Lea., 55 (1995) 213. [30] W.D. Harding, H.H. Kung, V.L. Kozhevnikov and K.R. Poeppelmeier, J. Catal., 144 (1993) 597. [31] A. Corma, J.M. L6pez Nieto, N. Paredes and M. P6rez, Appl. Catal. A, 97 (1993) 159. [32] X. Gao, Q. Xin and X. Guo, Appl. Catal. A, 114 (1994) 197. [33] P. Concepci6n, J.M. L6pez Nieto and J. P6rez-Pariente, Catal. Lett., 28 (1994) 9.