The present invention is directed to a process for the production of olefins by transhydrogenation.

There is a great impetus for developing processes which upgrade light olefins to high octane components, such as methyl-tert-butyl ether ( TBE) . Leaded gasoline has been phased out and the aromatics content of gasoline fuels has been restricted. However, the availability of olefin feedstocks such as isobutene for these ethers is limited. Therefore, processes for making these olefins from readily available feedstocks are sought. The process described herein is one such process and involves the conversion of available refinery feedstocks, such as isobutane, to products rich in isobutene. The combination of the high temperatures required for light paraffin dehydrogenation and the high endothermicity of such reactions has in the past been a challenge to their effective use, particularly in fixed-bed adiabatic reactors. Such reactors usually require preheating the reactor feed to a very high temperature in order to limit quenching of the dehydrogenation reaction. One solution to the problem of temperature drop involves splitting the catalyst bed within the reactor and providing interstage heating. The instant invention is provides a solution to this problem by coupling the endothermic reaction of paraffin dehydrogenation, with the hydrogenation of olefins which is an exothermic reaction. In particular, it has now been found that the transhydrogenation of cofed olefins, using hydrogen produced by the dehydrogenation of isobutane over a zeolite catalyst, particularly Pt/Sn-ZSM-5, is effective in increasing the yield and selectivity of isobutene. Light olefins such as propylene and ethene are generally employed as the cofeed. U.S. Pat. No. 4,546,204 (Parris) discloses a process in which ethylene and isobutane are contacted with a catalyst, giving isobutene and ethane in a transhydrogen-

ation reaction. Dehydrogenation catalysts suggested in Parris include heterogeneous catalysts such as supported and unsupported metals, metal oxides, and mixtures thereof, and homogeneous catalysts such as organometallic complexes. According to the invention, there is provided a process for producing, by transhydrogenation, an unsaturated analog of a feed compound containing an aliphatic moiety having 2 to 5 carbon atoms, and a saturated analog of a cofeed compound which is a light olefin, the process comprising contacting the feed compound and the cofeed compound with a non-acidic catalyst, said catalyst comprising a dehydrogenation metal and a non- acidic crystalline microporous material containing a metal modifier selected from Sn, In, Pb and Tl to dehydrogenate at least a portion of the feed compound and hydrogenate at least a portion of the cofeed compound.

One of the major advantages of the transhydrogenation process of the invention is that it reduces the magnitude of the temperature drop in the reaction zone, while also providing a higher per pass yield of desirable olefins.

The heat generated by transhydrogenation of olefins such as propylene greatly reduces the overall endothermicity of the dehydrogenation process. The transhydrogenation process also permits operation at lower temperatures and higher pressures. Olefins may be cofed into the middle of the catalyst bed.

Non-acidic intermediate pore size zeolites, such as ZSM-5, which contain noble metals such as Pt, and base metals such as tin, are found to be advantageous in transhydrogenation processes which operate in the absence of added hydrogen, due to their stability under low partial pressure hydrogen conditions. Metals on large pore zeolites or on amorphous supports age rapidly in the absence of cofed hydrogen invention whereas the ctalyst of the invention resists aging because the base metal is present within the zeolite and effectively protected from

bulky coke precursors. Noble metals such as Pt or Pd can tolerate low hydrogen pressure.

Moreover, solid, amorphous transalkylation catalysts are found to be more likely than crystalline microporous materials, such as zeolites, to result in side-reactions such as cracking which tend to accelerate in coking and/or aging of the catalyst, necessitating more frequent regeneration procedures.

Feed The feedstocks comprise at least one straight or branched chain aliphatic compound in which the aliphatic moiety has two to five carbon atoms. Other compounds effective in the instant invention include ethyl benzene, cumene and para-methyl ethyl toluene. In accordance with the invention, dehydrogenation of the aliphatic moiety occurs to yield the unsaturated analog. When the aliphatic moiety is substituted, the substituents can be aryls, substituted or unsubstituted. The class of reactants includes alkanes of 2 to 5 carbon atoms including propane, butane, isobutane, pentane and 2-methylbutane. Aromatics having aliphatic chains of 2 to 5 carbon atoms may also be used in the instant invention.

Light olefins are added to the reactor in the instant invention as a cofeed. Ethene and propene and mixtures thereof are the primary olefins used, although light FCC gas which contains impurities may also be used. Olefins are usually cofed in a ratio of from 0.05-0.5 moles olefin to 1 mole alkane. Light olefins may be combined with the feed and introduced into the reaction zone (usually an adiabatic reactor) at the top. Alternately, the catalyst bed may be split into two portions and the olefin cofeed may be added between the portions.

Catalvst

The catalyst for the catalytic dehydrogenation of the invention comprises a non-acidic composition including a hydrogenation/dehydrogenation metal and a non-acidic, crystalline microporous material containing a metal modifier selected from Sn, In, Tl and Pb. A suitable catalyst is further disclosed in U.S. Patent No. 4,886,926.

The hydrogenation/dehydrogenation metal can be any Group VIII noble metal. Preferably the metal may be Pt or Pd, most preferably Pt. The catalyst may also contain Ir and Rh. The amount of hydrogenation/dehydrogenation metal in the catalyst can range from 0.01 to 30 wt.% and preferably from 0.02 to 10 wt.% of the catalyst.

The metal modifier content of the catalyst can range from 0.01 to 20 wt.%, preferably 0.1 to 10 wt.%, of the crystalline microporous material.

The crystalline microporous material is preferably an intermediate pore zeolite having a constraint index of 1- 12, such as ZSM-5, ZSM-11, ZSM-22, ZSM-23 and ZSM-35. Most preferably the zeolite is ZSM-5 modified with Sn. The zeolite is non-acidic and preferably has an Al content less than 0.5 wt% and more preferably less than 0.2 wt%.

The crystalline microporous material of the instant invention, can contain other elements including boron, iron, chromium and gallium. These elements, can be present in amounts ranging from 0 to 10 wt.%.

The catalyst prefeably includes a silica binder.

Reaction Conditions

Catalytic dehydrogenation conditions include pressures varying from subatmospheric, to atmospheric to greater than atmospheric. Preferred pressures range from 0.1 atmosphere to atmospheric (10 to 100 kPa) . However, pressures up to 500 psig (3550 kPa) can be employed. The dehydrogenation is conducted at elevated temperatures ranging from 450°C to 700°C, and most preferably from 500 to 600°C. At reactor

hydrogen feed inlet ratios of zero, as generally exist in this invention, there is still a hydrogen partial pressure in the reactor because hydrogen is a by-product of dehydrogenation. The liquid hourly space velocity is 0.1 to 50, preferably 0.2 to 10. Under these conditions, the catalytic dehydrogenation of the invention exhibits reduced selectivity for hydrogenolysis and for isomerization.

Dehydrogenation, in the instant invention, is generally conducted in the absence of added hydrogen. However, diluents inert to conditions of catalytic dehydrogenation, such as nitrogen and methane, may be added.

The invention will now be more particularly described with reference to the Examples and the accompanying drawing which is a graph illustrating the relationship of isobutene yield and the hydrogenation of propene over time in the process of Example 3.

Example 1

Initially, a non-modified tin-free Pt/high silica ZSM- 5 catalyst was used to dehydrogenate isobutane at 550°C. Dilution of isobutane with an equimolar amount of He produced a 27.1wt% of isobutene. When the He was replaced with an equimolar amount of ethylene, the normalized yield of isobutene dropped to less than 0.4 wt%. Resubstitution of He for ethylene resulted in an enhanced yield of isobutene, which increased with time. After 0.6 hours, the isobutene yield was 4.6 wt%, and after an additional hour, the yield reached 7.4wt%.

The observation above is consistent with the known inhibiting effect of olefins on Pt-catalyzed dehydrogenations. Since isobutene and n-butene inhibit dehydrogenations to some extent, it is not unexpected that ethylene, which bonds more strongly to Pt, would inhibit even more severely, especially at the very high concentrations employed in this experiment.

In a subsequent experiment, a 50 wt% silica-bound Pt/Sn-ZSM-5 catalyst containing 0.62 wt% Pt, 1.44 wt% Sn, 0.14wt% Al, and 0.07 wt% Na, was divided into two 0.5g beds separated by vycor. Isobutane was introduced at the top of the reactor, while the olefin was cofed with the isobutane at the top of the bed. Reactor effluents were monitored by on-line gas chromatography. Regenerations were conducted in flowing hydrogen at 60 psig (515 kPa) and 500°C for 4-24 hours. The catalyst was also added at the top of the bed. One reason for the use of the Pt/Sn-ZSM-5 catalyst is that modification with tin is known to lessen olefin inhibition. Additionally, an equimolar amount of propylene was used as cofeed, instead of ethylene. At 500°C and about 1 hour on stream, a 9.9 wt% yield (normalized) of isobutene was observed. Furthermore, exactly 1.0 mole of propane was produced per mole of isobutene, indicating complete scavenging of H_ by propylene. Under these conditions, no net temperature drop was expected. The reaction was in fact expected to be slightly exothermic. Rapid aging was observed with the normalized yield of isobutene dropping to less than 2wt% after 7 hours on stream. Inhibition by propylene was also found to be reversible. After additional experimental work at various temperatures, removal of propylene led to a gradual increase in isobutane yield at 554°C over a period of 17 hours, climbing from 4.8 wt% after 1 hour to 8.7 wt% after 5 hours, and 16.7 wt% after 17 hours.

Example 2 In the experiments of this example, the olefin being cofed was introduced into the middle of the catalyst bed, where some hydrogen had already formed by dehydrogenation of isobutane. In these experiments, the catalyst, a silica-bound Pt/SnZSM-5 (as described in Example 1) was divided into two 0.5 g beds separated by vycor. Isobutane was passed through the entire reactor at 100 cc/min, while

the olefin was introduced between the two beds. The results obtained are shown in Table 1, below.

TABLE 1

Isobutane 1 Dehydrogenation with Transhydrogenation

Temperature Percent %

Cofeed Ratio ^β C Isobutene Transhydroαenation

None 500 25.1 wt% 0

Propene 0.3 (a) 500 35.0 wt% (b) 61.0 Wt% (C)

None 460 15.7 wt% 0

Propene 0.3 (a) 460 26.4 wt% (b) 80.3 Wt% (c)

Ethylene 0.3 (a) 460 29.4 wt% (b) 83.2 Wt% (d)

None (e) 0.4 (a) 550 40.4 wt% 0

Propene (e) " 550 48.7 wt% (b) 62.7 Wt% (C)

Ethylene (e) •' 550 50.6 wt% (b) 62.0 wt% (d)

(a) Mole ratio olefin/isobutane

(b) Normalized yield

(c) Determined from moles of propane formed relative to butenes produced (d) Determined from moles of ethane formed relative to butenes produced (e) Isobutane flow = 50 cc/min.

Table 1 illustrates that significant amounts of transhydrogenation were observed for both ethylene or propene cofeeds. The degree of transhydrogenation was measured by the amount of cofed olefin saturated per mole of isobutane dehydrogenated to isobutenes. The accuracy of these values was confirmed by direct determination of the molecular hydrogen produced. The decrease in hydrogen observed agreed quite well the amount of propane produced by transhydrogenation of propene.

Example 3 In this example lower ratios of olefins were used so that all the hydrogen generated would not be consumed. In these experiments, 20 cc/min propene was added between the two catalyst beds, at 550"C and an isobutane flow of 50 cc/min. At this higher temperature, normalized isobutene yields ranged from 49 wt% to 38 wt% over a period of 108 hours, with transhydrogenation of propene accounting for about 46 wt% of the isobutene produced. The results obtained are shown graphically in the Figure.

Example 4

In this approach, a mixture of 18 cc/min propene and

100 cc/min isobutane was passed over both beds of the same reactor used in Example 3. The results obtained are shown in Table 2, below:

TABLE 2

Isobutane Dehydrogenation with Propene Cofeed

Isobutene % % C_

Hours(a) Y Yiieelldd ( Cb. ) Transhydrogenation(c) Hvdrogenation(d) 0 0..33 3 333..33 wwtt%% 47.8 wt% 85.7 wt%

5 32.4 wt% 47.3 wt% 86.2 wt%

10 31.4 wt% 48.4 Wt% 85.8 wt%

16 30.6 wt% 49.7 wt% 85.0 wt%

22 29.8 wt% 50.8 Wt% 84.3 wt% 2 288 2 288..77 wWtt%% 52.2 wt% 83.2 wt%

38 27.4 Wt% 52.1 Wt% 82.0 Wt%

48 26.0 Wt% 53.6 wt% 80.1 wt%

(a) After 5 days on stream and regeneration in 60 psig H2 at 500°C for 4 hours.

(b) Normalized yield

(c) Determined from mole ratio of propane formed relative to butenes produced.

(d) Fraction of propene transhydrogenated to propane.

Under the conditions of Table 2, approximately half of the heat required to dehydrogenate isobutane was recovered by transhydrogenation of cofed propene.

In a second experiment, after catalyst regeneration as described below, 20 cc/min ethylene was cofed with 50 cc/min isbutane at 550 ^β C. Immediately prior to ethylene addition, the isobutene yield was 39.9 wt%. Upon addition of ethylene, the normalized yield of isobutene increased to 48.0 wt%, with 64.1 wt% transhydrogenation to ethylene. At this high ethylene concentration, significant aging was observed. Thus, 3 hours later, the normalized isobutene yield decreased to 38.4 wt%, while the extent of transhydrogenation increased to 74.4 wt%. At 6 hours, the isbutene yield was 27.8 wt%, with transhydrogenation to ethane 82.5 wt%.

The aging observed due to cofeeding of olefins was found to be reversible by treatment with flowing hydrogen at 60 psig (515 kPa) and 500°C for 15 hours. In one study, the normalized isobutene yield had dropped from 33.3 wt% to 19.9 wt% over a period of 4 days, but after hydrogen regeneration, a yield of 33.5 wt% was again obtained. Hydrogen regeneration was used several times throughout these examples.