Regeneration of a Commercial Catalyst for the

Dehydrogenation

1

S. Kamal Masoudian Sepehr Sadighi

Research Article

Ali Abbasi Fathollah Salehirad Ahmad Fazlollahi

Regeneration of a Commercial Catalyst for the Dehydrogenation of Isobutane to Isobutene

Research Institute of the Petroleum Industry (RIPI), Catalysis and Nanotechnology Research Division, Tehran, Iran.

In the chemical and petrochemical industries, dehydrogenating of hydrocarbons to olefins as raw materials for the manufacture of various chemical products is an important economic issue. So, here, the redispersion and regeneration of a commercial Pt-Sn/c-Al2O3 catalyst utilized for the conversion of isobutane to isobutene was studied. First, in order to regenerate the deactivated commercial catalyst unloaded from a commercial reactor, coke burning under controlled conditions was carried out. Then, to redisperse the platinum on the surface of the catalyst, a HCl solution (injected through a syringe pump) and Cl2 were passed over the catalytic bed at specific flow rates and temperatures. Evaluation tests carried out over the regenerated catalyst prove that the regenerated catalyst has similar conversion, selectivity, and structural characteristics as compared to the fresh one. Keywords: Catalyst, Dehydrogenation, Isobutene, MTBE, Regeneration Received: February 09, 2013; revised: April 30, 2013; accepted: June 11, 2013 DOI: 10.1002/ceat.201300090

1

Introduction

Alkenes, with the general formula CnH2n, are a group of chemical materials with widespread use in the petrochemical industries [1]. With regard to this issue, dehydrogenation of hydrocarbons is an important commercial process due to the great demand for dehydrogenated hydrocarbons, especially olefins, as raw materials for the manufacture of various chemical products such as detergents, high-octane gasoline, pharmaceuticals, plastics, synthetic rubbers, and other well-known products [2]. An important chemical compound produced from the dehydrogenation process is methyl-tert-butyl ether (MTBE), which provides better combustion and also increases the octane number of gasoline [3]. MTBE is made by liquid-phase reaction of isobutene with methanol (MeOH), catalyzed by a strongly acidic macroreticulate ion exchange resin (e.g. Amberlyst 15, Lewatit SPC 118) in the temperature range from 40 to 100 °C [4]. To obtain isobutene for the production of MTBE, isobutane is dehydrogenated over a Pt-Sn/Al2O3 catalyst. It is well known that the deactivation and regeneration of catalysts are important factors influencing the reaction efficiency [5], and similar to other catalytic processes, the activity of the catalyst used for dehydrogenation of isobutene to isobutene should be preserved in a commercial plant, by the removal of contaminants through various washing steps and also by coking out of

– Correspondence: Dr. S. Sadighi ([email protected]), Research Institute of the Petroleum Industry (RIPI), Catalysis and Nanotechnology Research Division, West Blvd. Azadi Sport Complex, P.O. Box 14665-137, Tehran, Iran.

Chem. Eng. Technol. 2013, 36, No. 00, 1–7

the deactivated catalyst through continuous circulated regeneration [6]. Up to now, many researchers have focused on catalyst regeneration for heterogeneous catalytic reactions. Manga and Susu [7] studied the importance of maintaining dispersion during the regeneration of Pt-Sn/Al2O3 catalysts and its essential role in the activity during the cycle of the catalyst. Then, Kuntzel et al. [8] investigated the regeneration of hydrophobic zeolites with steam and they found that steam can be effective in desorption of organic solvents from the surface of the catalyst. Villacampa et al. [9] reported the results of the characterization and catalytic behavior of a co-precipitated Ni (30 %)/ Al2O3 catalyst during methane cracking and catalyst regeneration. In addition, Roh et al. [10] regenerated steam-reforming catalysts (Rh/CeO2–ZrO2) at low temperatures with O2 and found that the catalyst could be regenerated even at ambient temperature and that it could recover its initial activity after regeneration above 200 °C with 1 % O2. Monzón et al. [11] studied the redispersion of Pt-Sn/c-Al2O3 catalysts, and they found that the concentration of HCl had a significant effect on redispersion of the catalyst during regeneration. Moreover, Patience et al. [12] studied the regeneration of a cerium-doped FePO4 catalyst, which dehydrated glycerol to acrolein in the gas phase. It was found that this catalyst was easily regenerated by air, and the reaction rate was proportional to both the oxygen concentration and the quantity of carbon. Recently, regeneration of granular activated carbon by supercritical carbon dioxide and microwave radiation was investigated at varying temperatures and pressures, which showed the importance of the regeneration process carried out with state-of-the-art processes [13, 14]. In this research, we present laboratory-scale

© 2013 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.cet-journal.com

2

S. Sadighi et al.

experiments carried out to regenerate a commercial Pt-Sn/ c-Al2O3 catalyst deactivated in an industrial plant during the hydrogenation of isobutane to isobutene. The results indicate acceptable characteristics, activity and selectivity of the regenerated catalyst in a laboratory-scale plant, compared to those of the fresh catalyst.

2

Experimental

2.1

Devices and Materials

2.1.1 Pilot Plant Device and Measuring Devices Simplified diagrams of the experimental devices used for the dehydrogenation of isobutane to isobutene and the regeneration of the deactivated catalyst are preFigure 2. Experimental Al2O3 catalyst. sented in Figs. 1 and 2, respectively. For each experiment, the reactor was loaded with 5 mL of catalyst with a diameter of 2–3 mm, diluted with an equal percentage of quartz particles. All experiments took place in a fixed-bed reactor with a bed height and diameter of about 40 and 18 mm, respectively. Moreover, to have a better distribution of the feed throughout the bed, the top and bottom of the catalytic bed were charged with 2.5 mm of quartz particles with the same diameter as mentioned above. To analyze the effluent stream of the reactor, a Hewlett Packard (HP) gas chromatograph (model 5890) was used. Moreover, the catalyst was analyzed by X-ray fluorescence (XRF) method using the EDX XR-300 and WDX 1410 devices manufactured by Link Analytical and Philips, respectively. The dis-

setup for the regeneration of industrially deactivated Pt-Sn/c-

persion of platinum on the Pt-Sn/c-Al2O3 catalyst was determined by a Philips X-ray diffractometer (model PW 1840) according to the UOP-905 standard procedure. To inject HCl solution as an oxychlorination agent during the regeneration process, a D-series syringe pump (model 260) licensed by Teledyne was used. HCl solution at a concentration of 5 mol L–1, used for oxychlorination of the spent catalyst, was purchased from Merck.

Figure 1. Experimental setup for the dehydrogenation reaction of isobutane.

www.cet-journal.com



Dehydrogenation

2.1.2 Catalyst In the present work, a commercial Pt-Sn/c-Al2O3 catalyst for the hydrogenation of isobutane to isobutene was used. The characteristics of the fresh catalyst are presented in Tab. 1. The type of the spent or deactivated catalyst studied and regenerated in this research was the same as the fresh commercial catalyst described above, but it was unloaded from the industrial reactor after completing its life cycle (∼2 years). The measured coke on the surface of the spent catalyst was equal to 5 wt %. Table 1. Characteristics and specifications of the Pt-Sn/c-Al2O3 catalyst studied. Ingredients of the catalyst [wt %] Pt

0.5

Sn

0.7

K

0.65

Mn

0.02

Fe

0.1

Ca

0.3

Na

0.05

Si

0.3

Cl

1.1–2a)

Surface area [m2 g–1]

Y

195

3 –1

Pore volume [cm g ]

55 ∼720

–3

Bulk density [kg m ]

m yiC4 H8 P × 100 m yiC4 H10 P

m yiC4 H10 F

m yiC4 H8 P × 100 m yiC4 H10 F

1

2

where S1) and Y are the total selectivity and conversion, respectively; m° is the molar flow rate of the feed or product streams; yiC4 H8 is the mole fraction of isobutene in the product; yiC4 H10 is the mole fraction of isobutane in the feed or product; and P and F represent the feed and product, respectively.

0.5

Average pore radius [Å]

(Fig. 1), the reactor was initially loaded with the Pt-Sn/c-Al2O3 catalyst. Then, the nitrogen gas at a gas hourly space velocity (GHSV) of 600 h–1 was sent through the catalytic bed from the top to the bottom. Simultaneously, the temperature of the bed was increased to 450 °C during 1.3 h, and the reactor was held at this condition for 30 min. Then, the nitrogen was substituted with hydrogen at a GHSV of 270 h–1, and the bed temperature suddenly increased to 530 °C. This condition was held for 1 h to activate the catalyst. After that, isobutane was fed to the reactor at a GHSV of 540 h–1, and the hydrogen flow was manipulated to set the hydrogen-to-hydrocarbon ratio (H2/HC) as equal to 0.5. Finally, the temperature of the reactor was ramped to 575 °C during 40 min. After reaching steady-state operation (about 2 h), with a sampling time of about 2 h, the outlet stream of the reactor was sent to the gas chromatograph, and this sampling procedure was conducted for at least 80 h. Before using these data to calculate selectivity and conversion, it was necessary to perform a carbon balance for each point. If a reasonable value was not provided (± 2%) this point was rejected. Then, the selectivity and conversion of each point was calculated according to the gas chromatography results as follows: S

Properties of the catalyst

3

a) Pertains to the regenerated catalyst.

2.2.2 Regeneration Method 2.1.3 Feed The purity of the isobutane feed used in this research was close to that of the commercial feed, as shown in Tab. 2. Table 2. Analysis of the isobutane feed used. Components

Value [mol %]

C3H8 i-C4H10 C4H10

2.2

0.6 96.8 2.6

Methodology

In order to regenerate the spent or deactivated commercial catalyst (Fig. 2), the reactor was initially loaded with it. Dry air at a GHSV of 1500 h–1 was sent over the catalytic bed from the top of the reactor to the bottom. Simultaneously, the temperature of the bed was increased to 510 °C during 2 h, and the reactor was held at these conditions for 1 h to achieve complete coke burning. After that, the HCl solution at a liquid hourly space velocity (LHSV) of 0.9 h–1 was injected into the reactor using a syringe pump. Promptly, Cl2 gas at a GHSV of 36 h–1 was added to the inlet line of the reactor. The mixture of dry air, Cl2, and HCl was passed through the catalytic bed for 2 h. Then, the HCl solution and Cl2 streams were cut off, and the dry air flowed to the reactor for 30 min at 510 °C. Finally, the heater of the reactor was turned off, but the dry air flowed through the catalytic bed to reach ambient temperature. Then,

2.2.1 Evaluation method According to the recipe of the catalyst licensor, to evaluate the conversion, selectivity and activity of the commercial catalyst


– 1)

List of symbols at the end of the paper.


www.cet-journal.com

4

S. Sadighi et al.

the regenerated catalyst could be evaluated according to the methodology described in Sect. 2.2.1.

3

Results and Discussion

Selectivity and conversion time profiles for the fresh Pt-Sn/ c-Al2O3 catalyst are shown in Fig. 3. As expected, these variables are stable versus the time of stream (TOS). The nearly constant profiles of these factors versus the TOS reveal the stability of the fresh catalyst for processing of isobutane to isobutene. The dehydrogenation product obtained from the fresh catalyst is presented in Tab. 3, designated as Cfresh (column 2). The equilibrium conversion of the dehydrogenation of isobutane to isobutene was calculated by the minimization of the Gibbs free energy using the Aspen plus software (AspenTech, 2006). The results show that the molar equilibrium conversion of isobutane under the described reaction conditions is about

Figure 3. Activity and selectivity of the fresh hydrogenation catalyst versus time. Table 3. Analysis of the dehydrogenation product. Catalyst component

Caa)

Cfresh

Cbb)

59 %. It is concluded that the conversion of isobutane reported in Fig. 3 is reasonably lower than the equilibrium value. Following the investigation of the fresh catalyst, the regeneration of an industrially spent catalyst with the same composition (Pt-Sn/c-Al2O3) was studied. It is obvious that, during the hydrogenation process in an industrial reactor, the activity of the fresh catalyst slowly decreases with the process time, due to the deposition of carbon and the agglomeration of the noble metal particles on the surface of the catalyst. As previously mentioned, to regenerate the deactivated catalyst, the deposited coke must first be removed by thermal treatment of the spent catalyst in an oxidizing atmosphere. Such an operation may promote further agglomeration of the noble metals and may change the metal particle surface or may lead to phase separation. Therefore, to redisperse the agglomerated noble metals, the coke-free catalyst is treated with halogen compounds, e.g. Cl2, HCl, and 1,2-dichloroethane, in the presence of H2O and an oxidizing agent, i.e. dry air, at elevated temperatures [15]. In this work, for a first attempt, a sample of the spent catalyst was heated with a controlled amount of dry air to burn the coke from the surface of the catalyst. This was continued until the oxygen content of the effluent gas stream from the reactor was the same as that of the inlet stream. After this step, the catalyst was simultaneously contacted with Cl2 and dry air to redisperse the platinum on the surface of the regenerated catalyst after reduction; it was then tested for isobutane dehydrogenation. Although the chlorine content of the catalyst was 1.7 wt %, the obtained results showed that the regenerated catalyst was not improved in comparison to the spent catalyst. In a second attempt, the Cl2 agent was replaced with 1,2-dichloroethane, and it was again observed that the regenerated catalyst (chlorine content of 2.5 wt %) was not successful in isobutane dehydrogenation. Eventually, the platinum of the coke-free catalyst was redispersed using a controlled amount of a mixture of Cl2 and HCl solution at the elevated temperature of 510 °C (Tab. 4). The evaluation results, discussed later, proved that this regenerated catalyst could show approximately the same activity as the fresh catalyst.

[mol %]

Table 4. Components used for oxychlorination.

C6+

0.113

0.563

0.070

CH4

1.538

2.168

2.001

C3H8

1.009

1.415

1.321

C3H6

0.731

0.860

0.877

i-C4H10

46.365

43.328

43.054

n-C4H10

1.948

1.981

1.996

1-C4H8

0.395

0.468

0.435

i-C4H8

44.263

45.308

45.282

trans-C4H8

2.015

2.141

2.072

cis-C4H8

1.623

1.662

1.681

C2H4

–

0.193

0.127

C2H6

–

0.699

1.086

a) After 55 h. b) After 85 h.

www.cet-journal.com

Gas phase [mol %] Cl2

HCl

H2 O

Air

H2O/Cl–

1.31

3.7

40.6

54.4

10.97

As mentioned before, the first and second catalysts regenerated using chlorine and 1,2-dichloroethane did not show the desired performance compared to the fresh catalyst. Foger and Jaeger [15] reported the presence of chlorine/support surface complexes [Pt(IV)Clx] during chlorine treatment of a Pt/c-Al2O3 catalyst. They suggested that platinum halide complexes exhibited a strongly anionic character, and they were expected to interact readily with c-Al2O3. Lieske et al. [16] found the existence of four different species containing Pt(IV), i.e., a(PtO2) and b(PtO2) as chlorine-free oxides, and [Pt4+(OH)xCly] and [Pt4+OxCly] as chlorine-containing complexes. Their finding indicated that redispersion of the noble metal was associated



Dehydrogenation

only with formation of the [Pt(IV)Clx] surface complexes occurring during the oxychlorination process. Therefore, the failure of those regeneration attempts may be due to the formation of surface oxides of types a(PtO2) or b(PtO2), not resulting in small Pt particles during the reduction process. It is believed that, by using the mixture of Cl2 and HCl solution in the oxychlorination process, the formation of chlorine-containing complexes of type [Pt2+(OH)xCly] or [Pt2+OxCly] is facilitated. These two surface complexes cause the activity of the regenerated catalyst to reach almost the same level as that of the fresh catalyst. The presence of water in the oxychlorination reaction is expected to help form the chlorine surface complexes more readily. Furthermore, a 20:1 molar ratio of water to chloride has been reported for successful oxychlorination operation of a naphtha-reforming catalyst, which has a higher chlorine content than the dehydrogenation catalyst [2]. In Fig. 4, the selectivity and conversion corresponding to the last regenerated catalyst, using the process described in Sect. 2.2.2, are sketched versus the TOS. From this figure, it is confirmed that this regenerated catalyst can acceptably preserve its selectivity and conversion, similar to the fresh catalyst, but there are a few fluctuations in the selectivity and conversion profiles, as discussed later. To have a better judgment, the mole percentages of the components in the product stream of the regenerated catalyst are also shown in Tab. 3, designated as Ca* and Cb*. It is observed that the profile of the components in the product is close to that of the fresh catalyst, and the yield of isobutene is even a little higher than in the case of the fresh catalyst. Furthermore, after 85 h, the regenerated catalyst still preserved its performance character, without considerable changes in the by-products, such as C6+, CH4, C3H8, 1-C4H8, and cis-C4H8.

Figure 4. Activity and selectivity of the regenerated hydrogenation catalyst versus time.

For the fresh catalyst, the halogen content was about 1.1 wt % of the catalyst (see Tab. 1), to avoid excessive side reactions (mainly cracking) resulting in gas formation and decreasing selectivity. The components resulting from cracking reactions of the regenerated catalyst (columns 3 and 4 in Tab. 3) are slightly higher in amount than those of the fresh catalyst (column 2 in Tab. 3), implying the high chlorine content of the regenerated catalyst. This is confirmed by the chlorine analysis of both the fresh and regenerated catalysts shown


5

in Tab. 1. Furthermore, Fig. 4 illustrates that the selectivity of the regenerated catalyst is slightly lower than that of the fresh one. This is due to a somewhat higher cracking ability of the regenerated catalyst, leading to the slightly higher conversion of the catalyst (see Figs. 3, 4 and Tab. 3). The fluctuations shown in the selectivity and conversion profiles of the regenerated catalyst (Fig. 4) may be due to a loss of chlorine content of the catalyst during the dehydrogenation reaction. In Tab. 5, the dispersion of platinum on the fresh, spent, and regenerated catalysts is presented. It can be seen that the spent catalyst unloaded from the commercial reactor has a lower platinum dispersion in comparison with the fresh one. This phenomenon decreases the isobutene yield, and after about 2 years, it becomes cost effective to shut down the reactor in order to exchange the old catalyst with a regenerated one. But from this table it is revealed that the regenerated catalyst has a higher platinum dispersion than the fresh catalyst. The values revealed in Tab. 5 can prove the effective redispersion of platinum on the Pt-Sn/c-Al2O3 catalyst after performing the regeneration process described in this research. Table 5. Dispersion of Pt on the catalysts studied. Catalyst

4

Dispersion [%]

Fresh

79

Spent

75

Regenerated

86

Conclusions

In this paper, laboratory-scale experiments for the regeneration of deactivated Pt-Sn/c-Al2O3 were presented. This catalyst was utilized for the dehydrogenation of isobutane to isobutene in a commercial-scale reactor. In order to regenerate the spent catalyst, dry air at a controlled GHSV and temperature was passed through the catalytic bed to achieve complete coke burning. After this step, oxychlorination was immediately carried out with a mixture of Cl2 gas and HCl solution, at a specific GHSV and LHSV, respectively, in the presence of a controlled amount of air under supervised regeneration conditions. The evaluation test results obtained from the regenerated catalyst after reduction showed that the activity of the commercial spent Pt-Sn/c-Al2O3 catalyst was comparable to that of the fresh one, proving the effectiveness of the presented methodology. Additionally, the product analysis showed that the regenerated catalyst after 85 h of testing under commercial operating conditions still had comparable activity to the fresh catalyst. These results are surely important for reusing the deactivated catalyst in a commercial plant, with a great effect on the economy of the process. Moreover, reusage decreases the environmental issue created by disposal of the waste catalyst. The authors have declared no conflict of interest.


www.cet-journal.com

6

S. Sadighi et al.

Symbols used C C*

[ml %] [mol %]

mF mp S yiC4 H8 yiC4 H10 Y

[mol h–1] [mol h–1] [–] [mol %] [mol %] [–]

mole fraction of the products mole fraction of the products for the regenerated catalyst molar flow rate of the feed molar flow rate of the product total selectivity mole fraction of isobutene mole fraction of isobutane total conversion

References [1] M. Mohagheghi, G. Bakeri, M. Saeedzad, Chem. Eng. Technol. 2007, 30 (12), 1721. [2] T. Imai, C. W. Hung, US Patent 4 438 288, 1984. [3] M. Quiroga, M. R. Capeletti, N. Figoli, U. Sedran, Appl. Catal., A 1999, 177, 37. [4] A. Rehfinger, U. Hoffmann, Chem. Eng. Technol. 1990, 13, 150.

www.cet-journal.com

[5] C. Xu, Chem. Eng. Technol. 2007, 30 (7), 880. [6] E. Schwarz, K. Rock, J. Byeseda, R. Pehler, presented at the AIChE 1992 Annual Meeting, Miami 1992. [7] N. H. Manga, A. A. Susu, Chem. Eng. Technol. 1996, 19, 263. [8] J. Küntzel, R. Ham, T. Melin, Chem. Eng. Technol. 1999, 22 (12), 991. [9] J. I. Villacampa, C. Royo, E. Romeo, J. A. Montoya, P. Del Angel, A. Monzóna, Appl. Catal., A 2003, 252, 363. [10] H. S. Roh, A. P. Yong Wang, D. L. King, Catal. Lett. 2006, 110, 1. [11] A. Monzón, T. F. Garetto, A. Borgna, Appl. Catal., A 2003, 248, 279. [12] G. S. Patience, Y. Farrie, J. F. Devaux, J. L. Dubois, Chem. Eng. Technol. 2012, 35 (9), 1699. [13] A. Heidari, M. N. Lotfollahi, H. Baseri, Chem. Eng. Technol. 2013, 36 (2), 1. [14] L. Guocheng, W. Limei, W. Xiaoyu, L. Libing, L. Zhaohui, H. Wenhui, Int. J. Chem. React. Eng. 2012, 10 (1). DOI: 10.1515/1542-6580.3117 [15] K. Foger, H. Jaeger, J. Catal. 1985, 92, 64. [16] H. Lieske, G. Lietz, H. Spindler, J. Volter, J. Catal. 1983, 81 (1), 8.



Dehydrogenation

Research Article: The regeneration of deactivated Pt-Sn/c-Al2O3 catalyst utilized for the dehydrogenation of isobutane to isobutene in a commercialscale plant is presented. The activity of the commercial Pt/c-Al2O3 catalyst is recovered after the regeneration process and the selectivity of the regenerated catalyst for the production of isobutene is similar to that of the fresh one.

7

Regeneration of a Commercial Catalyst for the Dehydrogenation of Isobutane to Isobutene S. K. Masoudian, S. Sadighi*, A. Abbasi, F. Salehirad, A. Fazlollahi Chem. Eng. Technol. 2013, 36 (䊏), XXX … XXX DOI: 10.1002/ceat.201300090



www.cet-journal.com