Sulfide-based alkane dehydrogenation catalysts

Abstract

A catalyst for the dehydrogenation of alkanes to alkenes comprises a catalytically active material supported on a carrier, wherein the catalytically active material is a metallic sulfide (MeS) comprising Fe, Co, Ni, Cu, Mo or W or any combination of two or more metals selected from Pb, Sn, Zn, Fe, Co, Ni, Cu, Mo and W. The catalyst is regenerated in several steps. The dehydrogenation is carried out at a temperature between 450 and 650 C. and a pressure from 0.9 bar below ambient pressure to 5 bar above ambient pressure.

Claims

1. A regenerated catalyst for the dehydrogenation of alkanes to alkenes, said regenerated catalyst comprising a catalytically active material comprising a metallic sulfide (MeS) supported on a carrier, wherein the metallic sulfide (MeS) is a semiconductor, and wherein the catalyst has been regenerated in steps, wherein the steps for regeneration comprise oxidation in diluted air to avoid thermal runaway and to convert the sulfide into the corresponding sulfate and conversion back to the sulfide by reduction in dilute hydrogen comprising hydrogen sulfide.

2. The regenerated catalyst according to claim 1, wherein the metal of the metallic sulfide comprises Fe, Co, Ni, Cu, Mo or W or any combination of two or more metals selected from Pb, Sn, Zn, Fe, Co, Ni, Cu, Mo and W.

3. The regenerated catalyst according to claim 1, wherein the oxidation in dilute air is carried out at a temperature between 350 and 750 C.

4. The regenerated catalyst according to claim 1, wherein the carrier is treated with a dilute alkali compound and subsequently washed to remove acid sites.

5. The regenerated catalyst according to claim 4, wherein the dilute alkali compound is potassium carbonate or any other potassium compound.

6. A process for the dehydrogenation of alkanes to the corresponding unsaturated alkenes and hydrogen (H.sub.2) comprising contacting a feed gas comprising an alkane with a regenerated catalyst according to claim 1, said catalyst being a metallic sulfide supported on a carrier and comprising Fe, Co, Ni, Cu, Mo or W or a combination of two or more metals selected from Pb, Sn, Zn, Fe, Co, Ni, Cu, Mo and W, wherein the feed gas contains sulfur in an amount determined such that (a) the equilibrium reaction MeS+H.sub.2.Math.Me+H.sub.2S is shifted towards MeS throughout the reactor, and (b) the sulfur content is sufficient to avoid carbide formation throughout the reactor.

7. The process according to claim 6, wherein the dehydrogenation is carried out at a temperature between 450 and 650 C.

8. The process according to claim 6, wherein the dehydrogenation is carried out at a pressure from 0.9 bar below ambient pressure to 5 bar above ambient pressure.

9. The process according to claim 8, wherein the dehydrogenation is carried out at ambient pressure or at a pressure from 0.5 bar below ambient pressure up to ambient pressure.

10. The process according to claim 6, wherein the H.sub.2S/H.sub.2 ratio being from 10.sup.3 to 10.sup.1.

11. A process for the dehydrogenation of alkanes to the corresponding unsaturated alkenes and hydrogen (H.sub.2) comprising contacting a feed gas comprising an alkane with a catalyst for the dehydrogenation of alkanes to alkenes, said catalyst comprising a catalytically active material supported on a carrier, wherein the catalytically active material comprises a metallic sulfide (MeS), wherein the metallic sulfide (MeS) is a semiconductor, said catalyst being a metallic sulfide supported on a carrier and comprising Fe, Co, Ni, Cu, Mo or W or a combination of two or more metals selected from Pb, Sn, Zn, Fe, Co, Ni, Cu, Mo and W, wherein the feed gas contains sulfur in an amount determined such that (a) the equilibrium reaction MeS+H.sub.2.Math.Me+H.sub.2S is shifted towards MeS throughout the reactor, and (b) the sulfur content is sufficient to avoid carbide formation throughout the reactor.

12. The process according to claim 11, wherein the H.sub.2S/H.sub.2 ratio being from 10.sup.3 to 10.sup.1.

Description

EXAMPLE 1

(1) 13.9 g Cu(NO.sub.3).sub.2.2.5H.sub.2O is dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g).

(2) The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(3) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 2

(4) 13.9 g Cu(NO.sub.3).sub.2.2.5H.sub.2O is dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(5) The impregnation step is repeated one more time followed by drying and calcination.

(6) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 3

(7) 24 g Fe(NO.sub.3).sub.3.9H.sub.2O is dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(8) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 4

(9) 24 g Fe(NO.sub.3).sub.3.9H.sub.2O is dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(10) The impregnation step is repeated one more time followed by drying and calcination.

(11) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 5

(12) 24 g Fe(NO.sub.3).sub.3.9H.sub.2O and 0.34 g KNO3 are dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(13) The impregnation step is repeated one more time followed by drying and calcination.

EXAMPLE 6

(14) 16.5 g FeSO.sub.4.7H.sub.2O and 0.9 g KHSO.sub.4 are dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(15) The impregnation step is repeated one more time followed by drying and calcination.

EXAMPLE 7

(16) 17.3 g Ni(NO.sub.3).sub.2.6H.sub.2O is dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(17) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 8

(18) 17.4 g Co(NO.sub.3).sub.2.6H.sub.2O is dissolved in 37.5 g water. The solution is used to impregnate 50 g of support (pv=1 ml/g).

(19) The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(20) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 9

(21) 9 g MoO.sub.3 is dissolved in 37.5 g of NH.sub.4OH solution. The resulting solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(22) The sample is then washed in 100 ml of a 2% KNO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 10

(23) 17.5 g H.sub.2WO.sub.4 is dissolved in 37.5 g of NH.sub.4OH solution. The resulting solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(24) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board).

(25) The sample is filtered and dried overnight at 100 C.

EXAMPLE 11

(26) 14.5 g SnCl.sub.2.2H.sub.2O is dissolved in 37.5 g of NH.sub.4OH solution. The resulting solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(27) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 12

(28) 18 g Zn(NO.sub.3).sub.2.6H.sub.2O is dissolved in 37.5 g of a NH.sub.4OH solution. The resulting solution is used to impregnate 50 g of support (pv=1 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(29) The sample is then washed in 100 ml of a 2% K.sub.2CO.sub.3 solution for 1 hour (rolling board). Afterwards the sample is washed two times with 200 ml water (one hour each, rolling board). The sample is filtered and dried overnight at 100 C.

EXAMPLE 13

(30) 7.9 g MoO.sub.3 is dissolved in 37.5 g NH.sub.4OH (25%). This solution is used to impregnate 50 g of support (pv=0.75 ml/g). The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

(31) The sample is then impregnated with 7.7 g Co(NO.sub.3).sub.2.6H.sub.2O and 0.8 g KNO.sub.3 dissolved in 37.5 g water. The sample is rolled for 1 hour, dried overnight at 100 C. and calcined at 500 C. for 2 hours (4 hours heating ramp).

EXAMPLE 14

(32) Results, Cu-Based Catalysts

(33) These catalysts have been tested previously as reported in U.S. Pat. No. 3,275,705, according to which a temperature of between 649 and 660 C. was used. The present tests were run at a much lower temperature using a dilute mixture of 10% propane in nitrogen.

(34) The catalyst prepared according to Example 1 was tested after being reduced and sulfided. The result for Cu was an activity of 72 Nl/h/kg cat of propylene at a temperature of 600 C. At 560 C., the activity was 35 Nl/h/kg cat corresponding to an activation energy of 1.13 eV.

(35) The results reported in U.S. Pat. No. 3,275,705 are up to 320 Nl/h/kg cat. A considerable thermal cracking is expected at this temperature. If the results of the present testing are extrapolated to 660 C., an activity of 190 Nl/h/kg cat is obtained. However, this is not an initial activity. The test results reported in U.S. Pat. No. 3,275,705 are of a short duration and therefore reporting an initial activity and no lifetime. Besides that, the Cu load of the catalyst prepared according to Example 1 could easily be increased.

(36) X-ray powder diffraction of the spent sample showed digenite, Cu.sub.1.78S, with an average crystallite size of 18 nm.

EXAMPLE 15

(37) Results, Fe-Based Catalysts

(38) The iron catalysts made according to Example 3 start out as an oxide and are slowly reduced and activated during the first propane dehydrogenation (PDH) test. After the second test it is just regenerated and then exposed to the propane-containing gas. The FeSO.sub.4 is reduced to FeS, a CO.sub.2 peak is noted, and during the next 12 hours the activity increases, whereupon it slowly declines. This cycle is repeated twice. Integration of the CO.sub.2 peak around 75 hours gave 0.3 g of carbon corresponding to 6% on the catalyst, and further 3% of the converted carbon ended up on the catalyst. The activity of the catalyst after 100 hours was 265 Nl/h/kg cat of propylene. The spent catalyst showed the presence of FeS with a crystallite size of 10 nm.

(39) The 12% iron catalyst was made from sulfates, which during the start-up were reduced to sulfides. This gave a quite stable activity of 160 Nl/h/cat of propylene.

(40) In the low sulfur gas, the activity declined to 40 Nl/h/cat of propylene. Thermodynamics indicate that iron carbide, Fe.sub.3C, would be the stable phase provided that the H.sub.2S level is below 100 ppm. Indeed, X-ray powder diffraction indicated that iron was in the carbide phase, Fe.sub.3C, with a crystallite size of 100 nm.

(41) Sun et al. (Chem. Eng. J. 244, 145-151 (2014)) examined a sulfated iron catalyst supported on alumina at 560 C. and found an initial activity of 70 Nl/h/cat. After 6 cycles of regeneration it had declined to 50 Nl/h/cat. A study on isobutane by Wang et al. of Fe supported on silica at 560 C. gave an initial activity of 17 Nl/h/cat. The condition was quite close to equilibrium.

EXAMPLE 16

(42) Results, Ni-Based Catalysts

(43) The nickel catalyst prepared according to Example 7 showed a high activity and to some extent an activity dependent on the activity on the sulfur level. The final activity was 330 Nl/h/cat of propylene at 600 C. At 560 C., an activity of 140 Nl/h/cat was found. This corresponds to an activation energy of 1.3 eV. This is a remarkably high activity, found close to the melting point of NiS. The tests with gas 4 were conducted under circumstances where a phase transition in the NiS system could take place. The phase diagram shows some complexity above 800 K. Conversion of propane into hydrogen and propene will change the H.sub.2S/H.sub.2 ratio and thereby change the stability for a certain phase.

(44) After regeneration, the spent catalyst showed a NiO phase with a crystallite size of 17 nm by X-ray powder diffraction. For Nickel, Wang et al. found an activity of 15 Nl/h/kg cat of isobutylene at 560 C.

EXAMPLE 17

(45) Results, Co-Based Catalysts

(46) The cobalt catalyst was made according to Example 8. It was started up in a low sulfur gas. However, after a number of phase transitions, the activity gradually increased. In the low sulfur gas, the state of Co is most like CoS, or Co.sub.9S.sub.8 should be formed. In a high sulfur concentration, Co.sub.3S.sub.4 may be formed. Cobalt sulfide has a homogeneity range. It is noted that CoSO.sub.4 reacts with propane, giving CO.sub.2. About 3% of the converted carbon ends up on the catalyst.

(47) The final activity at 600 C. was 170 Nl/h/kg cat of propylene. At 560 C., the activity was 70 Nl/h/kg cat corresponding to an activation energy of 0.7 eV. The regenerated spent catalyst showed the presence of a Co.sub.3O.sub.4 with a crystallite size of 10 nm.

(48) In connection with dehydrogenation of isobutane over a CoS catalyst, Chinese publications by Wang et al. have reported 42 and 15 Nl/h/kg cat of isobutylene, respectively. In the present experiment, sulfur was added as ammonium sulfate to the Co/Al.sub.2O.sub.3 catalyst, whereas Wang et al. also used co-feeding with H.sub.2S.

EXAMPLE 18

(49) Results, Mo-Based Catalysts

(50) The Mo-based catalyst was made according to Example 9. During the tests, the carbide-sulfide interface was challenged. For up to 140 hours, the experiments were run with a low (100 ppm) sulfur level. A rapidly declining activity is seen, probably due to formation of molybdenum carbide. At 150 hours, the sulfur level is increased and the sulfide phase is stabilized along with the activity. The cracking activity is above the background level for the sulfide/carbide catalyst. The maximum activity was 60 Nl/h/cat propene at 600 C.

(51) The spent catalyst was examined using X-ray powder diffraction. No molybdenum phases were seen, although the chemical analysis showed a molybdenum content of 7.25 wt %.

EXAMPLE 19

(52) Results, CoMo Based Catalysts

(53) 5 g of a catalyst prepared as described in Example 14 was tested. The test lasted for 3 weeks and included 10 regenerations and tests with temperatures from 520 to 640 C. The sulfur level was not in any case below 1000 ppm. Thus carbide formation could be excluded.

(54) The starting activity was 260 Nl/h/cat and the final activity was 210 Nl/h/kg cat at 600 C. At other temperatures the corresponding maximum activities were 104 Nl/h/cat at 580 C., 100 Nl/h/cat at 560 C. and 42 Nl/h/cat at 520 C. This corresponds to an activation energy of 0.27 eV.

(55) At the lower temperatures, the activity declined slower than at the high temperature. Runs of up to 64 hours were done, during which a decline in the activity from 1.31% to 0.57% C.sub.3H.sub.6 was seen.

(56) Table 4 below indicates that a relation exists between the band gap energy and the activity, but also to some extent between the activation energy and the band gap, in particular for the lower activities. This could make sense when taking into consideration that, for high activities, other mechanisms than available electrons/holes could give rate limitations, such as simple transport limitations. Catalysts based on iron and nickel and possibly also on cobalt and copper will have sufficient activity for operation in the Oleflex process. In case of operation in the Catofin process they may be less suitable. A pre-reduction may be required in order to avoid loss of propane in the reduction of sulfate to sulfide. However, the activity for iron and nickel is quite high, so possibly less metal and thereby less sulfate may be required.

(57) TABLE-US-00004 TABLE 4 Summary of results Activity Activation Band gap energy Compound Nl/h/kg cat energy (eV) eV Cu.sub.2S 70 1.1 1.2-1.5 CoS 100 1.2 0.9 Co.sub.3S.sub.4 170 1.3 1.15 FeS 260 ~0 NiS 330 1.3 0.4 Ni.sub.3S.sub.2 100 0.5 ~0.4 CoMo 260 0.3 ? MoS.sub.2 60 ~0 1.3-1.4

(58) The catalysts are deactivated slowly by carbon deposition and therefore need to be regenerated, just like the commercially available catalysts based on platinum or chromium oxide. The regeneration takes place by combustion in dilute air, i.e. 1% O.sub.2 and 99% N.sub.2, at 560-600 C.

(59) Regeneration of Fe, Co, Ni and Cu sulfides using N.sub.2 with 1% O.sub.2 will lead formation of the corresponding sulfate, at least for Fe and Co, whereas Ni and Cu will most likely form oxides, which has been confirmed by X-ray powder diffraction of the spent samples. The Mo and W sulfides are converted into oxides. In this connection, a possible loss of active material as a volatile oxide (MoO.sub.3) has to be taken into account. In order to conserve sulfur on the catalyst, regeneration should start at 400 C. followed by a carbon removal at 600 C.

(60) Re-sulfidation using a mixture of H.sub.2 and H.sub.2S in N.sub.2 has often been carried out prior to the dehydrogenations tests. In some cases, this has been omitted, whereby initially a large formation of CO.sub.2 was noted. Some sulfates, in particular MnSO.sub.4, have shown a good selectivity for oxidative dehydrogenation of propane.

(61) The regeneration takes the catalyst through two phase transition stages, from sulfide to sulfate or oxide and back again to sulfide. The phase transitions involve not only structural transformations, but also volume changes. Thus, in the case of NiS to NiSO.sub.4 an expansion of 250% takes place. It is expected that sintering/dispersion of the system will reach steady state after a number of regenerations.

(62) During dehydrogenation, some carbon is deposited on the catalyst, resulting in a slow deactivation. Dehydrogenation takes place for some hours followed by catalyst regeneration in N.sub.2 containing 1% O.sub.2. This is typically followed by a sulfidation or a direct return to dehydrogenation. In this case, a direct reaction between sulfate and propane takes place, resulting in a large CO.sub.2 formation.

(63) The carrier is treated with a dilute alkali compound and subsequently washed to remove acid sites. Preferably the dilute alkali compound is potassium carbonate or any other potassium compound.

(64) In the experiments, the carrier has been dipped in a dilute potassium carbonate solution followed by a two-step wash in demineralized water, resulting in a potassium content of 0.15 wt %. Acid sites have been removed, but not necessarily all of them. The results indicate a pressure influence on the carbon formation, and they also indicate that carbon formation takes place from propylene, not propane. Furthermore, the results indicate that there is a complete carbon removal during regeneration.