Selective Oligomerization of Isobutene on Lewis ... - ACS Publications

Subscriber access provided by UNIV OF WINNIPEG

Research Note/Communication

Selective Oligomerization of Isobutene on Lewis Acid Catalyst: Kinetic Modeling Amitava Sarkar, Deepyaman Seth, Flora T.T. Ng, and Garry L. Rempel Ind. Eng. Chem. Res., Just Accepted Manuscript • DOI: 10.1021/ie501173z • Publication Date (Web): 19 Aug 2014 Downloaded from http://pubs.acs.org on August 20, 2014

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Industrial & Engineering Chemistry Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Industrial & Engineering Chemistry Research

Selective Oligomerization of Isobutene on Lewis Acid Catalyst: Kinetic Modeling

Amitava Sarkar†, Deepyaman Seth, Flora T. T. Ng*, and Garry L. Rempel Department of Chemical Engineering, University of Waterloo, ON, Canada, N2L 3G1

________________________________________________________________ * To whom corresponding should be addressed. E-mail: [email protected] †

Current address: Total E&P Canada Ltd., Calgary, AB, Canada

ACS Paragon Plus Environment


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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 50

Topical Heading: Reactors, Kinetics and Catalysis

Keywords:

Oligomerization;

Isobutene,

Kinetic

NiSO4 / γ − Alumina , Isooctenes; Selective dimerization;

2


model,

Page 3 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ABSTRACT The kinetics of oligomerization of isobutene was studied on a NiSO4 / γ − Alumina catalyst in a stirred batch autoclave at 50 – 90 °C at 2170 kPa with different concentration of isobutene. Experimental results revealed that the catalyst has high dimer selectivity and did not show any significant deactivation during the reaction. A generalized kinetic model based on Langmuir-Hinshelwood (LH) type reaction sequence was developed. The intraparticle diffusion effects inside the catalyst particle were correlated to the reaction rates and mass transfer rate between the catalyst particle and liquid phase. The developed intrinsic kinetic model with estimated model parameters was found to describe the experimental data accurately. The magnitude of the activation energies was found to be in the range of 13-27 kJ mol–1 which suggests that the oligomerization reaction proceeds via surface rearrangement as considered in the present LH-type kinetic model.

Introduction Recent developments in refining industry are majorly influenced by environmental issues, speciality product demand and cost effectiveness [1-4]. Due to the phase out of methyl- tertiary butyl ether for environmental protection, there is a need to have suitable replacement to enhance the octane number of gasoline[5]. New processing methodologies

3



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 50

for environmental friendly high octane gasoline formulation while meeting other clean fuel specifications can be based on: •

retrofitting an existing MTBE unit to the production of other alkyl ether like ethyl tertiary butyl ether (ETBE) and tertiary amyl ether (TAME).

•

alkylation of the isobutene with isobutane catalyzed by strong acids for producing paraffinic high octane alkylate blendstock.

•

production

of

isooctane

(2,2,4-trimethylpentane,

TMPA)

by

selective

dimerization of isobutene and subsequent hydrogenation of the produced isomeric dimers. While considering the alkyl ethers as the choice, it should be noted that the health and safety issues of ETBE and TAME are less understood and both ethers may suffer from the same environmental perception as MTBE. Although several alkylation processes have been developed using liquid acids like HF and H2SO4 as catalysts [6-9], but these processes have negative environmental impacts, provide potential risks on populated areas and suffer from intense environmental scrutiny and substantial drawbacks. Anhydrous HF is a highly toxic and corrosive liquid with a boiling point close to room temperature and can form a ground-hugging vapor cloud if released into atmosphere. Sulfuric acid is also corrosive and the major disadvantage of H2SO4 based process is high acid consumption in the alkylation process-almost one third of the total operating cost can be attributed to acid consumption [9]. Alkylation processes based on solid acids are not presently operated on an industrial scale [10] as a major limitation of solid-acid alkylation catalysts is that they are quickly fouled during reaction and are deactivated after several regenerations.

4


Page 5 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In order to overcome the above problems, resort could be made to use the surplus isobutene feedstock for the production of isooctane which has both research octane no (RON) and motor octane number of 100. Oligomerization of alkenes has resurfaced as a very promising technology for converting light olefinic fractions into aromatics free higher value gasoline fuel and middle distillates recently [1-2, 11-12]. Oligomerization has significant advantages over other conventional solutions (alkylation) as it has a greater flexibility on product distribution. Few conventional process configurations for isooctane units have been developed and licensed [13-16]. The isooctane processes mentioned above are based on conventional reactor technology and follow the general process sequence: selective dimerization-product separation-hydrogenation of dimers. The high exothermic heat of oligomerization needs to be removed to control the extent of oligomerization for improving dimer selectivity. In some cases polar compounds has been used as diluent to enhance the dimer selectivity. Since oligomerization is a consecutive reaction and the aim is to maximize the production of dimer, the reaction system is suitable for applying catalytic distillation (CD) technology. Both the oligomerization of isobutene and hydrogenation of olefinic oligomes are exothermic in nature and the oligomeric products can easily be separated by distillation, it would be beneficial to exploit the application of CD technology [17] for the production of isooctane. The selective dimerization of isobutene and subsequent in situ hydrogenation of the dimers can be carried out in a single distillation column while utilizing the reaction heat for the distillation process. Instantaneous separation of the dimers from the reaction mixture would result in increased dimer selectivity by

5



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

minimizing the formation of higher oligomers (i.e., trimer, tetramers, etc.) . Subsequent in-situ hydrogenation will directly result pure isooctane stream without need of any recycling or further downstream processing. The CD based isooctane process needs fewer pieces of process equipment and less energy than that of conventional processes mentioned earlier because of proper heat integration. Hence, both the capital and operating cost of a CD based process will be less than that of conventional processes. A new CD process for a one-step production of isooctane from isobutene and hydrogen has been developed recently [18]. Intrinsic kinetics of oligomerization of isobutene is required for the successful design and optimization of such CD based isooctane process. The production of oligomers higher than dimers, which have lower octane rating, will decrease the selectivity for the isooctene production. Thus the oligomerization catalyst should be highly selective for the dimerization reaction and at the same time, it should have high activity. The conventional industrial process used in the oligomerization of isobutene is generally carried out using 68% sulphuric acid at 80 °C, solid or liquid phosphoric acid or an acidic ion-exchange resin at 100 °C and about 20 atm pressure [19]. Since isobutene is a very reactive olefin, the oligomerization reaction can be promoted by almost any electrophilic catalyst which include sulphonated styrene–divinylbenzene copolymer [20], nickel substituted synthetic mica–montmorillonite [21], fluorinated alumina [22], AlH3/Davison SiO2 [23], pentasil zeolites [24], cation exchange resins [25], benzylsulfonic acid siloxane [26], Cr(V) or V(IV) or Zn(II) substituted HZSM-5 [27], perfluorinated resinsulfonic acids [28], H-mordenite [29], macroporous sulphonic acid resins [30], silica–alumina–nickel oxide gel [31], heteropolyacids [32], and sulfated TiO2

6


Page 6 of 50

Page 7 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


membrane [33]. However, most of the above mentioned catalyst system suffer from severe catalyst deactivation during reaction and some of these results dimer selectivity as low as 10% [34]; thus, unsuitable for application in a CD column reactor [17]. Although a large number of studies on oligomerization of isobutene on various catalyst systems have been reported in the literature, only few of them are related to the intrinsic kinetics and in particular detailed kinetic modeling. Like any polymerization reaction, oligomerization of isobutene involves chain initiation, propagation and termination. Depending on the nature of catalyst system used, the type of the intermediates during propagation can be different. Therefore different mechanisms are involved which suggests that a kinetic model derived for a particular catalyst system may not be suitable for use with a different catalyst system. Izquierdo et al. [30] studied the oligomerization of isobutene catalyzed by a macroporous styrene-divinylbenzene (SDVB) sulphonic acid resin in presence of some methanol and MTBE which were added to the system to study the influence of MTBE formation on the rate of dimerization. However, the concentrations of methanol and MTBE were not included in the developed kinetic model. It was found that the rate of oligomerization can be best described by a two-phase semi-empirical kinetic model implying the coexistence of a Langmuir-Hinshelwood (LH) mechanism and a modified Eley-Rideal (ER) one (first order with respect to isobutene concentration). Similarly Honkela et al. [35] studied the oligomerization of isobutene on a commercial SDVB sulphonic acid resin together with tert-butyl alcohol (2-methyl-2-propanol, TBA), to increases the dimer selectivity by reducing the activity of the sulfonic acid sites of the resin. LH-type kinetic models were developed for the dimerization and trimerization reactions only and the formation of

7



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

tetramers was ignored in the model development. It was found that the only model to represent the experimental data successfully was the one where dimers form using two active sites and trimers via dimers using three active sites. In spite of using a porous resin catalyst in the experiments, both Izquierdo et al. [30] and Honkela et al. [35] ignored the internal diffusion effects in the development of the kinetic model. Nickel (II) sulfate or nickel (II) chloride supported on γ − alumina has been used as the most successful heterogeneous catalysts for the oligomerization of lower olefins (e.g., ethylene and propylene) by several researchers [36-39]. It was found that nickel (II) sulfate supported on γ − alumina, having acidity of Lewis acid type only [18], is the most promising catalyst in terms of activity and dimer selectivity, lower cost, ease of preparation, negligible deactivation and ease of regeneration. The nickel loading and the reaction conditions influence the catalyst performance in terms of catalyst lifetime, activity or productivity, isobutene conversion and dimer selectivity. However, no intrinsic kinetic study has been reported so far in the literature for the oligomerization of isobutene on NiSO4 / γ − Al2O3 catalyst. The effect of system pressure and use of 2methylbutane as solvent on the kinetics of oligomerization of 1-butene on a

NiSO4 / γ − Al2O3 catalyst having 3 wt% nickel loading was studied by Huang [40]. It was found that when the system pressure is high enough for maintaining the butenes in liquid phase, the effect of any change in the system pressure on the oligomerization kinetics is insignificant. The effect of process parameters on catalyst performance and the kinetics of oligomerization of isobutene on the NiSO4 / γ − Al2O3 catalyst having 1.5 wt% nickel loading was presented by Xu [41]. It was found that at a stirring speed greater than 400 rpm, the external mass transfer effects are negligible while catalyst performance

8


Page 8 of 50

Page 9 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


remains constant. The internal diffusion inside the catalyst pores was ignored and a pseudo-homogeneous kinetic model was developed to describe the observed kinetics. The effect of nickel loading on the dimer selectivity and catalyst activity was also studied and it was found that higher dimer selectivity can be achieved by reducing the nickel loading of the catalyst. Thus a nickel loading of 1 wt% for the NiSO4 / γ − Al2O3 catalyst has been selected in the current research. Since the effect of system pressure, stirring speed and amount of solvent on the performance of the NiSO4 / γ − Al2O3 catalyst for oligomerization of isobutene is known [41]; a set of experiments for the oligomerization of isobutene was performed to determine the intrinsic kinetics where the reaction temperature and initial concentration of isobutene were varied only. The experiments were performed in a stirred batch autoclave equipped with a Harshaw falling catalyst basket. The liquid samples at times were collected in a 16-loop automated multiposition sampling valve and analyzed by an online GC equipped with a flame ionization detector. The purpose of the present research was to develop a detailed intrinsic kinetic model for the oligomerization of isobutene on a NiSO4 / γ − Al2O3 catalyst having 1 wt% nickel loading and to reveal the significance of internal diffusion on the observed kinetics. The model parameters will be validated by the experimental observations. The developed model can be utilized to describe the oligomerization kinetics in the process development and optimization of a CD based isooctane unit.

Experimental Details Materials: An in-house NiSO4 / γ − Al2O3 catalyst (1 wt% Ni, 204 m2 g–1 specific surface area, and 0.55 cm3 g–1 total pore volumes) made of a special type of distillation

9



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

packing was used in all the experiments. Nickel (II) sulfate hexahydrate, (Aldrich Chemical Company, Inc., 46, 790-1, 99.99+%) was used as the nickel precursor. The nickel loading of the catalyst was determined using XRF ( Oxford Lab-X 3000S Instruments, Inc.) after calibration with a physical mixture of ultra-fine γ − alumina powder and nickel (II) sulfate hexahydrate. Pure isobutene (Praxair Canada, Inc., IY 2.0FX, Grade 2.0, 99.0%) was used as reactant and HPLC grade 2-methylbutane (SigmaAldrich Inc., 270342, 99.5+%) was used as solvent in all experiments. Ultra-high purity nitrogen (99.5+%) gas supplied by Praxair Canada, Inc., was used as received.

Methods: The oligomerization experiments were carried out over a range of temperature and isobutene concentration as summarized in Table 1. All the experiments were performed in a 300 cm3 SS 316 bolted closure stirred batch reactor (Autoclave Engineers, Inc.) under constant nitrogen pressure (± 5 kPa). The details of the reactor setup and operating procedure are presented in elsewhere [18]. The reactor is equipped with a Harshaw falling catalyst basket (approximate volume 15 cm3), an axial down turbine impeller, an internal water cooling coil, a dip-tube for liquid-phase sampling, a thermowell containing a J-type thermocouple, and a dual action programmable temperature controller (±1 °C). A 120 VAC external furnace equipped with thermowell containing a J-type thermocouple provides energy for heating and the cooling coil equipped with solenoid valve controls the temperature of the reactor. A simplified schematic of the reactor set-up is presented in Figure 1. Before each experiment, the required amount of catalyst was calcined in dry air at 500 °C for 12 hours. The calcined catalyst stored in HPLC grade n-heptane under argon atmosphere, was then transferred into the Harshaw falling catalyst basket of the reactor

10


Page 10 of 50

Page 11 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


inside a glove bag maintained with argon environment. Required amount of 2methylbutane was added to the reactor and the reactor was then sealed within the glove bag followed by slow purging with argon. After removing the reactor from the glove bag, the reactor was purged slowly with nitrogen. The isobutene was transferred from the supply tank to the reactor with the help of a sampling bomb. The reactor, loaded with catalyst and reactant mixture, was set to the desired reaction temperature under constant stirring. Once the reactor temperature reached the desired set point, the reactor was pressurized with nitrogen to the desired reaction pressure following which the catalyst basket was dropped into the reaction mixture inside the reactor by a sudden switchoff/switch-on of the stirrer. The start of the reaction time was noted at the point when the catalyst was dropped into the reaction mixture. The liquid sample at times were collected in a 16-loop automated multiposition sampling valve (approximate volume of each loop is 0.25 cm3) located in a heated sample enclosure maintained at 250 °C and analyzed by an online GC (Perkin Elmer, Inc.) equipped with a HP-5 capillary column (J & W Scientific; length = 30 m; diameter = 0.32 mm; film thickness = 0.25 µm ) and a flame ionization detector. The flow connections between the reactor, the heated sample enclosure, and the GC were made of heated transfer line maintained at 250 °C to avoid any condensation in the line. A split ratio of 200:1 was used for all analysis. Integration of the GC results was performed with a PE Nelson 1020 GC Plus integrator (Perkin Elmer, Inc.). The carrier gas flow rate was 1 ml min–1. The oven temperature was initially kept at 45 °C for 5 minutes, then increased to 75 °C at a rate of 10 °C min–1 and finally ramped to 250 °C at a rate of 25 °C min–1. The detector temperature was set to 300 °C.

11



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The reaction product from isobutene oligomerization contains mainly unconverted isobutenes, isooctenes, dodecenes and hexadecenes. The GC peaks corresponding to 1TMP and 2-TMP in the product mixture was identified after matching retention time to that of pure compounds. No other isobutene dimer except 1-TMP and 2-TMP was found in the product oligomers. It is evident that obtaining the actual amount for each isomer in the groups of isobutene trimers and tetramers require identification of each component. As the primary objective of the research was related to isooctene and isooctane, any isomers of isobutene trimer and tetramer was not distinguished; instead, they were simply considered as a single pseudocomponent and referred to as trimers and tetramers in this study. The amount of trimers and tetramers was simply obtained by taking the sum of all the area percentages within a retention time range. For trimers, the retention times range is chosen between (13-15) min and for tetramers the above range is chosen as (16-18) min.

Qualitative Observations Reproducibility of the experiments were carried out by comparing the rate of reaction of isobutene at a reference condition (55 mol% isobutene in 2-methylbutane at 50 °C, 2170 kPa and 1000 rpm stirring speed) with different batches of catalyst. These replicates showed that the maximum variation in the rates was about ±2%. Experiments were performed with the spent catalyst, without exposing to air to check for any catalyst deactivation during oligomerization. After 60 minutes of reaction at a reference condition the contents of the reactor was vented off and fresh solvent and isobutene were loaded to the reactor using a sampling bomb and reaction was continued

12


Page 12 of 50

Page 13 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


at the reference condition for another 60 minutes. The variation of the reaction rate was found to be ±2%, which suggests no significant catalyst deactivation. Experimental results showed that the NiSO4 / γ − Al2O3 catalyst is active for oligomerization of isobutene and results in higher dimer selectivity compared to the catalyst with 1.5 wt% nickel loading [41]. A typical time-concentration profile at 50 °C for isobutene is presented in Figure 2 and a similar profile is shown in Figure 3 for the oligomerization products of the same reaction. The experiments have shown that the reaction mixture contains mainly solvent, monomers, dimers, trimers and very small amount of tetramers. This indicates that under normal reaction conditions not only the thermal and catalytic cracking is negligible but the production of oligomers larger than tetramers is also insignificant. Therefore, only dimerization, trimerization and tetramerization were considered in this study. The effect of reaction temperature on the conversion of isobutene in oligomerization and resulting dimer selectivity is shown in Figure 4. The conversion of isobutene was defined as the ratio of amount of isobutene disappears per unit volume at a particular time to the initial amount isobutene present per unit volume. The dimer selectivity was calculated as the ratio of total concentration of isobutene dimers (1-TMP and 2-TMP) produced to the total concentration of isobutene oligomers (dimers, trimers, tetramers and higher oligomers together) produced at a particular time. It is evident from Figure 4 that increasing temperature results higher conversion at the expense of lower dimer selectivity. The initial rate of oligomerization did not vary very much from 50 °C to 90 °C suggesting that the oligomerization reaction may be limited by internal diffusion. To describe the effect of temperature on the kinetics of oligomerization, an Arrhenius-type

13



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

of plot in the range of temperature studied experimentally is presented in Figure 5. The plot is almost linear suggesting no shift in the controlling mechanism from kinetic to mass transfer control. The activation energy estimated from this plot was found to be in the range of 27-28 kJ mol–1, which suggested that the kinetics was pore diffusion limited and intraparticle diffusion effects have to be considered in the development of intrinsic kinetic model.

Generalized Kinetic Model for the Oligomerization It is known that Lewis acid-catalyzed oligomerization of isobutene does not proceed via the carbocationic mechanism [34] . The active catalytic site responsible for the oligomerization of isobutene on NiSO4 / γ − Al2O3 catalyst was suggested to consist of a low valent nickel ion and an acid center [18]. Based on the commonly accepted mechanism for oligomerization of lower olefins on Lewis acid catalysts [34, 42], prior knowledge from kinetic experiments in the current study, a generalized reaction scheme based on Langmuir-Hinshelwood type of reaction sequence is considered. The generalized reaction scheme for the oligomerization of isobutene is outlined in Table 2. The activation of isobutene is assumed to initiate by the adsorption of the isobutene molecule to the active site of the catalyst. Two adjacently adsorbed isobutene species can react to form dimer which can either react with another adsorbed isobutene species to form trimer or can desorb from the catalyst surface, becoming a free molecule. In the same way, the adsorbed trimer species can react with another adsorbed isobutene species to form tetramer. It is known that dimerization of isobutene yields mainly a mixture of 1TMP and 2-TMP in a ratio of approximately 4:1 [43]. The formation of the above two dimers was confirmed by GC analysis in the present research and the presence of any

14


Page 14 of 50

Page 15 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


other type of isooctene dimer was not observed during the kinetic experiments. Thus the formations of above isomeric dimers are considered only. Although, the reaction of three adjacently adsorbed isobutene species to form trimers or four adjacently adsorbed isobutene species to form tetramers are theoretically possible, but less likely to occur from probability point of view. It was also found that the rate equation derived for trimers formation from simultaneous reaction of three isobutene molecules cannot successfully interpret the kinetic data [35]. Hence, these types of reactions are not considered. Similarly, isobutene tetramers can be produced either by the reaction of an adsorbed trimer with an adsorbed isobutene species or by the reaction of two adjacently adsorbed isobutene dimers species. As the reaction of two adjacently adsorbed isooctene species is less favourable sterically, the formation of tetramer is considered only by the reaction of an adsorbed trimer with an adsorbed isobutene species. In addition, the re-adsorption of the free dimers and trimers on the catalyst surface to form activated species is not considered due to the fact that the adsorption of these free dimers and trimers, if any, is much weaker than the adsorption of isobutenes [34]. The isobutene trimers and tetramers produced are observed to be more than one type. Hence, the corresponding isomeric oligomers having carbon number 12 and 16 are treated as single olefin pseudocomponent and expressed as trimer (i.e., C12= ) and tetramer (i.e., C16= ), respectively. The interactions between the adsorbed species are the rate limiting/determining steps (RDS). Therefore, steps (2-6) in Table 2 are treated as RDS. In the generalized reaction sequence considered in the present study, the adsorption and desorption steps are assumed to be fast enough for the quasi-equilibrium hypothesis to be applied. Hence, the

15



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

terms K 1 , K 7 , K 8 , K 9 and K10

Page 16 of 50

denote the corresponding adsorption equilibrium

constants. Thus

θ A−S = K1c Aθ S

(1)

θ A21−S = K 7 c A21θ S

(2)

θ A22 −S = K8c A22θ S

(3)

θ A3 − S = K 9 c A3θ S

(4)

θ A4 −S = K10c A4 θ S

(5)

A total balance of the active sites of the catalyst along with the pseudo-steady-state assumption for reaction intermediates and use of equations (1)-(5) results in

θS =

1 + K 1c A + K 7 c A21

1 + K 8 c A22 + K 9 c A3 + K 10 c A4

(6)

Thus the reaction rate for different species can be expressed as: rA21 = k 2θ A2 − S − k 4θ A− S θ A21 − S ≅ α 1c A2 − α 2 c A c A21

(7)

rA22 = k 3θ A2− S − k 5θ A− S θ A22 − S ≅ α 3 c A2 − α 4 c A c A22

(8)

rA3 = k 4θ A− S θ A21 − S + k 5θ A− S θ A22 − S − k 6θ A− S θ A3 − S ≅ α 2c Ac A21 + α 4c Ac A22 − α5c Ac A3

(9)

rA4 = k 6θ A− S θ A3 − S ≅ α 5 c A c A3

(10)

rA = −2rA21 − 2rA22 − 3rA3 − 4rA4

(11)

where, α1 , α 2 , α 3 , α 4 and α5 are the lumped rate constants corresponding to surface reaction step 2, 3, 4, 5, and 6, respectively, of the reaction sequence shown in Table 2.

Reactor Model for the Oligomerization 16


Page 17 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The reactor model used for analysis was similar to the model published previously for the liquid phase hydrogenation of isooctene

in a semibatch reactor [ 44 ] . A

schematic of the reaction system used for the model development is shown in Figure 6. Each experiment was modeled as a three-phase, unsteady state process. The entire reactor volume was divided into two parts-the vapor phase and the liquid phase containing dispersed/suspended solid catalyst particles. For a given stirrer speed, pressure and temperature, the mass transfer rate of the volatile components are calculated; as the ratio of this rate with the diffusion rate was much larger than unity, vapor-liquid equilibrium was assumed. The solid catalyst had an egg-shell type uniform distribution of the catalytically active layer ( about 1mm in average thickness close to 10% of the dimension of the catalyst particle ). For the given catalyst pellets, the ratio of the radius of these pellets to the depth of the region where reaction take places was >>1. Hence, the region of the catalyst where the reaction takes place can therefore be assumed to have rectangular co-ordinates. The dynamic reactor model consists of the material balance of each component in the vapor and liquid phases (except nitrogen). Since the vapor and liquid phases were in equilibrium, the mass transfer rate between these two phases was infinitely large and a material balance of the sum of each component in the two phases can be written as

d L (ni + n iV ) = − AN i dt

z =0

i = 1,...6

;

(12)

Component i varies from 1 to 6 corresponding to isobutene, 2-methylbutane, 1-TMP, 2= = TMP, C12 , and C16 , respectively. Since the amount of trimers and tetramers were small,

for simplification, the physical properties of C12= and C16= “pseudospecies” are

17



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 50

approximated to be same as 1-dodecene and 1-hexadecene, respectively. As N2 was continuously fed into the reactor to maintain constant pressure, the material balance for nitrogen was replaced by the following equation 6

VR =

∑ n iL i =1

ρ

L

7

+

∑n

V i

i =1

ρV

;

(13)

The numbers of moles of each component in the vapor and the liquid phases are related through the following equilibrium relationship

K

eq i

n iL n Vi = ; ∑ n Lj ∑ nVj

i = 1,...6

(14)

For the calculation of the vapor-liquid equilibrium constant, the vapor phase fugacity coefficients were calculated using the virial equation of state [45]. The liquid phase activity coefficients were calculated using the modified UNIQUAC equation [45], the parameters in which were estimated by the UNIFAC method [46]. The above system of equations can only be solved if the value of N i at z = 0 is specified. To obtain the values of N i , the concentration gradient at z = 0 must be known, or in other words, the concentration profile within the catalyst must be known. Determination of the concentration profile would require the solution of the following mass-balance equations within the catalyst particle

εP

∂c i ∂N i + = ri ρ S ; ∂t ∂z

i N i ≅ − Deff

∂c i ; ∂z

18

i = 1,...6

(15)

i = 1,...6

(16)


Page 19 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The value of the effective diffusivities were obtained from i D eff =

εP i Dm ; τP

i = 1,...6

(17)

Molecular diffusion coefficients of the mixture ( D mi ) were calculated from the binary diffusion coefficients and liquid-phase molar fractions. The binary diffusivities and the diffusivity in the mixture were estimated by the Wilke-Chang method [47]. The porosity and tortuosity values of the catalyst particles were approximated to the corresponding average values for aluminium oxide, i.e., 0.5 and 4.0 respectively [48]. Substituting i Equation (16) into (15) and considering Deff being independent of z results in the

following equation 2 ∂c i i ∂ ci εP − Deff = ri ρ S ∂t ∂z 2

i = 1,...6

(18)

The reaction rate terms must be supplied before the above system of equation can be solved. As the experiments were performed under isothermal conditions, the energy balances for bulk-phases were not needed. Furthermore, calculations showed that, under the experimental conditions applied, the maximum temperature difference inside the catalyst particle was reduced to an insignificant level (i.e., less than 0.6 °C). Hence, the liquid and solid bulk phase was assumed isothermal. Thus the energy balance for the catalyst particle was also omitted.

Solution of Model Equations and Parameter Estimation The solutions for the model equations were obtained using the method described in [44]. The mass balance of each component in the liquid phase is coupled with the mass balance equation in the catalyst phase through the term N i at z = 0 . The ordinary

19



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 50

differential equation (ODE) for the liquid phase along with the non-linear algebraic equations (NLAE) for vapor-liquid equilibrium, are solved independently of the partial differential equation (PDE) in the solid phase. In this case the time steps were sufficiently small so that the bulk liquid phase concentration at the beginning of each time step can be used as the boundary value for the solution of the PDE. The finite-volume method has been used for the discretization of the catalyst massbalance equations. As the intensity of reaction was very strong near the catalyst surface compared to that inside the pellet, closely spaced control volumes were required near the catalyst surface. A very fine mesh was adopted in this short region and a coarse grid was adopted in the interior of the catalyst particle. A total of 20-30 control volumes were sufficient to capture the system dynamics accurately. The arrangement of a representative mesh segment is shown in Figure 7. In the discretized equations, the superscripts denote the time step and the subscripts denote the discretization points. A fully implicit scheme has been adopted. Upon discretization, the following set of equations were obtained

(n )

L j +1 i

( ) + (n )

− niL ∆t

j

V i

( )

j +1

− niV ∆t

j

( )

i = 2 A Deff

j +1   L   n  (ci ) j +1 −  i ρ L    ∑ n Lj    1 j +1   ;  f1 θ 1 ∆z      

i = 1,...,6.

∑ (n )

L j +1 i

(ρ ) L

j +1

∑ (n ) +

 eq n iL  K i  ∑ nkL  

V i

(ρ ) V

j +1

(19)

j +1

j +1

=V R

 nV  = i V  ∑ n k 

(20)

j +1

;

i = 1,...6

For control volume k = 1

20


(21)

Page 21 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


εP

(ci )kj +1 − (ci )kj ∆tθ k ∆z

= (ri ρ S )k θ k ∆z; j +1

   i −  Deff    

(ci )kj ++11 − (ci )kj +1 ) fk +1 (θ + θ )∆z / 2 − (Deffi ) jfk+1 k k +1

(

j +1

( )

j +1 ci k

j  niL   L ρ   −  ∑ n Lj      θ k ∆z / 2    

i = 1,...,6

(22)

For control volumes k = 2,..., (m − 1)

εP

(ci )kj +1 − (ci )kj ∆t θ k ∆z

= (ri ρ S )k θ k ∆z; j +1

j +1 j +1 (ci )kj ++11 − (ci )kj +1 j +1 (c i )k − (ci )k −1  i ) fk +1 (θ + θ )∆z / 2 − (Deff ) fk (θ + θ )∆z / 2  k k +1 k −1 k 

 i −  Deff 

(

j +1

i = 1,...,6

(23)

For control volume k = m

εP

(ci )kj +1 − (ci )kj ∆tθ k ∆z 7

+

( ) i Deff

(

O.F . = ∑∑ ciexp − cical j i =1

j +1 fk

)

(ci )kj +1 − (ci )kj +−11 j +1 = (ri ρ S )k θ k ∆z; (θ k −1 + θ k )∆z / 2

2

i = 1,...,6

(24)

(25)

j

where, j denotes analyzed product samples and i is the component index. The minimization of equation (25) was done with the FMINCON routine in Matlab 6.5® which is a subspace trust region method and is based on interior-reflective Newton method [49, 50 ]. All the lumped reaction rate constants were assumed to be function of temperature according to the following form  E iapp R 

α i = α iref exp  −

1  − 1  T Tref 

   

(26)

21



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The reference temperature, Tref was chosen as 50 °C. Thus the estimated parameters were reaction rate constants at the reference temperature and corresponding apparent activation energies.

Results and Discussion The results of the oligomerization experiments which were used in the evaluation of the parameters of the kinetic model were carried out over a wide range of reaction temperature and initial concentration of isobutene as summarized in Table 1. Experiments were conducted at temperatures ranging from 50-90 °C at a constant pressure of 300 psig (2860 kPa absolute) and an agitation speed of 1000 rpm. 2Methylbutane was used as solvent in all experiments. The initial concentration of isobutene was in the range of 50 mol% to 75 mol%. The estimated rate constants along with the activation energies defined by equation (26) are presented in Table 3. In addition, values of the residual sum of squares (RSS) of all the data points and the residual root mean square (RRMS) are presented. Based on the estimated rate constants presented in Table 3, the concentration profiles of the species present during the oligomerization of isobutene were calculated. Figure 8 and Figure 9 present a typical concentration profile in an oligomerization experiment performed at 50 °C. The concentration profiles predicted by the model were found to be in good agreement with the experimental results. The shape of the dynamic model curves follows the experimental data points closely, indicating the accuracy of the developed model. Similar good fits of experimental and model-predicted concentration profiles were observed for other experiments also. Based on the total volume of liquid-phase and the slope of the concentration-time plot, the overall activity may be calculated. However, as

22


Page 22 of 50

Page 23 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


there is diffusion limitation, this rate is essentially the diffusion rate. Therefore, the actual catalyst activity must be calculated based on the kinetic parameters. Also, as most kinetic parameter estimation is based on the overall rate, comparing the overall rate from literatures with the intrinsic kinetic rate may be misleading. Figure 10 and Figure 11 showed the model predictions are in agreement with the experimental data obtained under various reaction conditions. It is known that exchange/rearrangement takes place among atoms and molecules that are adsorbed at different active sites on catalyst surface. The reason for this lies in the low activation energies for transport along the surface as compared to the high values for desorption into the bulk phases or for diffusion into the porous catalyst particle. Activation energies for surface diffusion of adsorbed species from one active site to another are frequently one-half or less of the activation energies for desorption into bulk phase or diffusion into solid phase [51]. As a result, the magnitude of activation energy for surface diffusion/migration reactions (i.e., reaction following LH-type kinetics) are generally one-half or even less than that of the reaction which follows either ER-type kinetics or in the situation where a species will be adsorbed to the surface prior to the reaction. Oligomerization of isobutene on Lewis acid catalyst was assumed to proceeds via LH-type kinetics involving rearrangement of the adsorbed surface species. The magnitude of the activation energies estimated from the developed model is found to be in the range of 13-27 kJ mol–1. Since the intraparticle diffusion effects have been considered in the development of the kinetic model, the estimated values justify the assumption of surface migration reactions.

23



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

It is interesting to note that although, for simplicity, the formations of isobutene oligomers were assumed to be irreversible as shown in Table 2, the resultant kinetic model can predict the experimental data fairly well. Huang [40] noted that when pure 1octene was used as the reactant, the product obtained after one hour of reaction was found to contain only the isomers of 1-octene along with some hexadecenes and no butene was detected in the product. This implied that the decomposition of dimers of C4 olefins is not possible under normal oligomerization reaction condition and provides strong support that the dimer formation shown in Table 2 is essentially irreversible. Hence, for the purpose of the kinetic modeling, the formation of isobutene dimers, trimers and tetramers were also assumed to be irreversible. Oligomerization of isobutene was severely diffusion limited inside the catalyst particles. The concentration profile of isobutene and product oligomers inside the catalyst particles after 120 minutes of reaction is shown in Figure 12. Model results showed the concentration of isobutene decreased with time whereas the concentration of oligomeric products increased. The concentration of isobutene

was much lower compared to the

surface concentration within a short distance from the surface. The concentration of products inside the catalyst particle was much higher than the surface concentration of oligomers. This is assumed to be the results of two phenomena-generations of oligomers within the catalyst particle by reaction and lower diffusivity of oligomers as compared to isobutene. As isobutene concentration rapidly drops due to consumption by reaction, most of the catalyst thickness remained unutilized. Since the concentration of oligomers is higher inside the catalyst particle, there is the potential of coke formation by reaction of

24


Page 24 of 50

Page 25 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


these higher oligomers if reaction temperature is increased. Hence, a very thin egg-shell distribution of a active catalyst layer on the support would be preferable .

Conclusions The NiSO4 / γ − Al2O3 catalyst was found to be active for the oligomerization of isobutene. No significant catalyst deactivation was observed during the study. Increasing the temperature favors higher isobutene conversion and lower dimer selectivity. Considering the intraparticle diffusion effects and dynamic reactor model, a LH-type of kinetic model was developed. Comparison of the experimental and calculated concentrations of all data points showed good agreement with experimental results, and the developed model with estimated model parameters was found to predict the kinetic behavior of the system (where results were not used for parameter estimation of the developed model) accurately. The magnitude of activation energies for oligomerization reaction suggests that the reaction proceeds via surface migration/rearrangement. The model prediction suggests that the oligomerization of isobutene is diffusion limited inside the catalyst particles, i.e., the rate of reaction as observed is limited by the diffusion of material inside the catalyst pellet The concentration profile of isobutene drops rapidly while that of oligomers escalates within a short distance from the surface. Hence, a very thin egg-shell distribution of active catalyst layer on the support has been suggested to avoid any possible coke formation. The results suggest that the reaction system for isobutene oligomerization can truly be represent by the kinetic model developed in the current study. Thus the developed kinetic model can be used for the purpose of the theoretical modeling and analysis of the catalytic distillation process for the production of isooctane.

25



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Acknowledgement Financial support from the National Sciences and Engineering Research Council of Canada is gratefully acknowledged.

Notations A = total external surface area of the catalyst, m2

c i = concentration of component i , mole m–3 or mole cm–3 d P = effective diameter of catalyst particle, m, cm, mm i Deff = effective diffusivity of component i , m2 s–1

Dmi = molecular diffusivity of component i in the mixture, m2 s–1 Eiapp = apparent activation energy corresponding to lumped rate parameter α i as defined in equation (7) - (11), J mol–1

f k = interface between (k − 1) th control volume and k th control volume k i = rate constant for reaction i at temperature T , K K i = adsorption equilibrium constant of component i

K ieq = vapor-liquid equilibrium constant of component i MON = motor octane number, dimensionless

niL = total number of moles of component i in the liquid phase niV = total number of moles of component i in the vapor phase N i = molar flux of component i in the solid phase, mole m–2 s–1 ri = rate of reaction of component i , mole s–1 kg cat–1

26


Page 26 of 50

Page 27 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


R = universal gas constant [ ≡ 8.314 J mol–1 K–1]

RON = research octane number, dimensionless T = temperature, °C or K

Tref = reference temperature [ ≡ 323.15 K] t = time variable, s or min or hr V R = volume of the reactor, m3 z = distance, m, cm, mm

Greek Letters α1 ,...,α 5 = lumped rate constants for oligomerization reactions as defined in equations (7) – (11), mole s −1 kg.catalyst −1 (m 3 mole−1 ) −2

α iref = rate constant for reaction i at reference temperature Tref , K

ε P = porosity of the catalyst particle

ρ L = density of the liquid phase, mole m–3 ρ V = density of the liquid phase, mole m–3 ρ S = bulk density of the catalyst, kg m–3 θ i = fraction of the active sites of catalyst surface covered by component i

θ1 ,....θ k = fraction of ∆ z τ P = tortuosity of the catalyst particle

Subscripts and Superscripts A = isobutene

A21 = 2,4,4-trimethyl-1-pentene A22 = 2,4,4-trimethyl-2-pentene 27



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A3 = isobutene trimers (single olefin pseudocomponent) A4 = isobutene tetramers (single olefin pseudocomponent) cal = calculated eq = equilibrium exp = experimental i = component index, reaction index

j = stage index, time step index, k = discretization point index G = gas phase L = liquid phase

n = number of components

R = reactor ref = reference temperature (323.15 K) S = solid phase P = catalyst particle

1-TMP = 2,4,4-trimethyl-1-pentene 2-TMP = 2,4,4-trimethyl-2-pentene TMPA = 2,2,4-trimethylpentane V = vapor phase

Literature Cited 1. Kriván, E., I. Valkai, and H. Jenő, The Oligomerization of Olefin Hydrocarbons

in Light FCC Naphtha on Ion Exchange Resin Catalyst. Chemical Engineering Transactions, The Italian Association of Chemical Engineering, 2012, 29, 985.

28


Page 28 of 50

Page 29 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2. Kriván, E., G. Marsi, and J. Hancsók, Investigation of the Oligomerization of

light olefins on ion exchange resin catalyst. Hungarian Journal of Industrial Chemistry, 2010, 38 (1), 53. 3. Coelho, A., Caeiro, G., Lemos, M. A. N. D. A., Lemos, F., Ribeiro, F. R., 1-

Butene oligomerization over ZSM-5 zeolite: Part 1 – Effect of reaction conditions. Fuel, 2013, 111, 449. 4. Bellussi G., Mizia F., Calemma V., Pollesel P., Millini R., Oligomerization of

olefins from Light Cracking Naphtha over zeolite-based catalyst for the production of high quality diesel fuel. Microporous and Mesoporous Materials,

2012, 164, 127. 5. Tsai, M. J., Kolodziej, R., and Ching, D. Consider new technologies to replace MTBE. Hydrocarbon Processing, 2002, February (2), 81. 6. Parkinson, G. Reviving up for alkylation. Chemical Engineering, 2001, 108 (1), 27. 7. Meister, J. M., Black, S. M., Muldoon, B. S., Wei, D. H., and Roeseler, C. M. Optimize alkylate production for clean fuels. Hydrocarbon Processing, 2000, 79 (5), 63. 8. Albright, L. F. Developing technology. Improving alkylate gasoline technology. CHEMTECH, 1998, 28(7), 46. 9. Albright, L. F. Updating alkylate gasoline technology, CHEMTECH, 1998, 28(6), 40. 10. Feller, A., and Lercher, J. A. Chemistry and technology of isobutane/alkene alkylation catalyzed by liquid and solid acids, in: Gates, B. C. and Knözinger, H.

29



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(Eds.), Advances in Catalysis, 48, Elsevier Inc., San Diego, CA, USA, 2004, 2297. 11. Corma, A. and S. Iborra, Oligomerization of Alkenes, in: Derouane, E. (Ed)., Catalysts for Fine Chemical Synthesis, Vol 4. Microporous and Mesoporous Solid Catalysts, John Wiley & Sons, Ltd., 2004. 125. 12. Process Economics Program Review, C3-C4 Oligomerization for Gasoline, 2009, 2009-14, IHS Chemical. 13. Vandersall, M. T. Review of isooctane processes as replacements for MTBE. Proceedings of 12th Brazilian Congress on Catalysis, Angra dos Reis, Rio de Janeiro, September 16 – 19, 2003, 1084. 14. Hunszinger P., Järvelin H., Purola V. M., Nurminen M., Khalil M. S., Birkhoff R. Start-up and operation of the first on-purpose isooctane unit at Alberta Envirofuels Inc., Edmonton, Canada. Presented at: NPRA 2003 Annual Meeting, San Antonio, TX, Mar 23-25, 2003, Paper AM-03-43. 15. Hunszinger P., Järvelin H., Nurminen M., and Birkhoff R. Case history: converting an MTBE unit to isooctane operation. Hydrocarbon Processing, 2003, 82 (9), 57. 16. Sloan, H. D., Birkhoff, R., Gilbert, M. F., Pyhalahti, A., and Nurminen, M. Isooctane production from C4’s as an alternative to MTBE. Presented at: NPRA 2000 Annual Meeting, San Antonio, TX, March 26 –28, 2000, Paper: AM-0034.8-11. 17. Ng, F. T. T., and Rempel, G. L. Catalytic distillation, in: Horváth, I. T. (Eds.), Encyclopedia of Catalysis, 2. Wiley-Interscience, New York, 2003, 477.

30


Page 30 of 50

Page 31 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


18. Sarkar, Amitava A Catalytic Distillation Process for One-step Production of Isooctane from Isobutene – Process Development, Modeling and Analysis.. Ph.D. Thesis, University of Waterloo, ON, Canada, 2005. 19. Olah, G., and Molnár, A. Hydrocarbon Chemistry. John Wiley & Sons, Inc., Hoboken, New Jersey, USA, 2003, 723. 20. Haag, W. O. Oligomerization of isobutylene on cation exchange resins, Chemical Engineering Progress Symposium Series, 1967, 63 (73), 140. 21. Bercik, P. G., Metzger, K. J., and Swift, H. E. Oligomerization of C3 – C4 olefins using a novel nickel-aluminosilicate catalyst. Industrial and Engineering Chemistry Product Research Development, 1978, 17(3), 214. 22. Kerkhof, F. P. J. M., Oudejans, J. C., Moulijn, J. A., Matulewicz, E. R. A. Structure and activity of fluorinated alumina. 2. Nature of the active site for 2methylpropene oligomerization. Journal of Colloid and Interface Science, 1980, 77(1), 120. 23. Slaugh, L. H. Alkene oligomerization process. U. S. Patent, Patent No. 4367361, 1982. 24. Minachev, K. M., and Kondrat‫׳‬ev, D. A. The properties and use in catalysis of zeolites of the pentasil type. Russian Chemical Reviews, 1983, 52 (12), 1113. 25. O'Connor, C.T., Kojima, M., and Schumann, W. K. The oligomerization of C4 alkenes over cationic exchange resins. Applied Catalysis, 1985, 16, 193. 26. Saus, A., and Schmidl, E. Benzyl sulfonic acid siloxane as a catalyst: oligomerization of isobutene. Journal of Catalysis, 1985, 94, 187.

31



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

27. Kustov, L. M., Kondrat‫׳‬ev, D. A., Krasieva, T. B., Borovkov, V. Y., Kazanskii, V. B., and Minachev, K. M. Effect of metal oxide modifying additives on the catalytic properties of high-silicon zeolites of the pentasil type in transformations of C3 – C4 olefins. Kinetics and Catalysis, 1989, 30 (1), 145. 28. Bucsi, I., and Olah, G. A. Oligomerization of 2-methylpropene and transformation of 2,4,4-trimethyl-2-pentene over supported and unsupported perfluorinated resinsulfonic acid catalysts. Journal of Catalysis, 1992, 137, 12. 29. Ngandjui, L. M. T., and Thyrion, F. C. Kinetic study of the oligomerization of isobutene over H-mordenite. Chemical Engineering and Processing, 1992, 31, 1. 30. Izquierdo, J. F., Vila, M., Tejero, J., Cunill, F., and Iborra, M. Kinetic study of isobutene dimerization catalyzed by a macroporous sulphonic acid resin. Applied Catalysis A: General, 1993, 106, 155. 31. Brunelli, M., Castelvetro, W., Bellussi, G., and Peratello, S. Catalyst and process for oligomerizing olefins. US Patent, Patent No. 5510555, 1996. 32. Sanati, M., Hörnell, C., and Järås, S. G. The oligomerization of alkenes by heterogeneous catalysts, in Spivey, J. J. (Ed.), Catalysis, 1999, 14, The Royal Society of Chemistry, Cambridge, U.K., 236. 33. Mantilla, A., Tzompantzi, F., Ferrat, G., López-Ortega, A., Romero, E., OrtizIslas, E., Gómez, R., and Torres, M. Room temperature olefins oligomerization over sulfated titania. Chemical Communications, 2004, 13, 1498. 34. Skupińska, J. Oligomerization of olefins to higher oligomers. Chemical Reviews,

1991, 91, 613.

32


Page 32 of 50

Page 33 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


35. Honkela, M. L., and Krause, A. O. I. Kinetic modeling of the dimerization of isobutene. Industrial & Engineering Chemistry Research, 2004, 43(13), 3251. 36. Cai, T., Zang, L., Qi, A., Wang, D., Cao, D., and Li, L. Propene oligomerization catalyst derived from nickel sulfate supported on alumina, Applied Catalysis,

1991, 69, 1. 37. Cai, T. Studies of a new alkene oligomerization catalyst derived from nickel sulfate. Catalysis Today, 1999, 51, 153. 38. Chauvin, Y., Commereuc, D., Hugues, F. and Thivolle-Cazat, J. Nickel-based heterogenous catalysts for olefin oligomerization. I. Support and anion effects. Applied Catalysis, 1988, 42, 205. 39. Sohn, J. R. Catalytic activities of nickel containing catalysts for ethylene dimerization and butane isomerization and their relationship to acidic properties. Catalysis Today, 2002, 73, 197. 40. Huang, C. Catalytic Distillation: experimental study and theoretical modeling. Ph.D. Thesis, University of Waterloo, ON, Canada, 1999. 41. Xu, Yongqiang, Application of catalytic distillation for the oligomerization of isobutylene. MASc. Thesis, University of Waterloo, ON, Canada, 2002. 42. O'Connor, C.T., and Kojima, M. Alkene oligomerization. Catalysis Today, 1990, 6(3), 329. 43. Whitemore, F. C. Mechanism of the polymerization of olefins by acid catalysts. Industrial and Engineering Chemistry, 1934, 26, 94. 44. Sarkar, A, Seth D, Ng, F.T.T., Rempel, G.L. Kinetics of Liquid-Phase Hydrogenation of Isooctenes on a Pd/γ-alumina Catalyst , 2006, 52(30) 1142.

33



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 34 of 50

45. Prausnitz J., Anderson T., Grens E., Eckert C., Hsieh R., O’Connell J. Computer calculations for multicomponent vapor-liquid and liquid-liquid equilibria. Prentice Hall, Inc., Upper Saddle River, USA, 1980. 46. Fredenslund, A., Gmehling, J. Rasmussen, P. Vapor-liquid equilibria using UNIFAC: a group contribution method. Elsevier Scientific Publishing Co., Amsterdam, 1977. 47. Reid, R. C., Prausnitz, J. M., Sherwood, T. K. The properties of gases and liquids. McGraw-Hill, New York, USA, 1987. 48. Satterfield, C. N. Mass Transfer in Heterogeneous Catalysis. MIT Press, Cambridge, USA, 1970. 49. Coleman, T. F., Li, Y. On the convergence of reflective Newton methods for large-scale

nonlinear

minimization

subject

to

bounds.

Mathematical

Programming, 1994, 67(2), 189. 50. Coleman, T. F., Li, Y. An interior, trust region approach for nonlinear minimization subject to bounds. SIAM Journal on Optimization, 1996, 6, 418. 51. Somorjai, G. A. Chemistry in Two Dimensions: Surfaces. Cornell University Press, New York, 1981, 29.

34


Page 35 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 1. Summary of reaction condition studied experimentally in the oligomerization of isobutene in a stirred batch reactor. 2-methylbutane was used as solvent in all experiments.

Isobutene concentration, mol%

50, 55, 60, 65, 75,80

Temperature, °C

50, 70, 90

System pressure, kPa

2170

Stirring speed, rpm

1000

35



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 2. Proposed reaction sequence and corresponding rate expression in the oligomerization of isobutene. A: isobutene, A21: 1-TMP; A22: 2-TMP, A3: trimers; A4: tetramers; S: active site on catalyst surface;

θ i : fraction of active catalyst site

covered by species i ; RDS: rate determining step.

Step

Reaction

Rate

1

k1 → A+ S ←   A − S k−1

2

k2 A − S + A − S →  A21 − S + S

RDS

3

k3 A − S + A − S →  A22 − S + S

RDS

4

k4 A21 − S + A − S →  A3 − S + S

RDS

5

k5 A22 − S + A − S →  A3 − S + S

RDS

6

k6 A3 − S + A − S →  A4 − S + S

RDS

7

k7 → A21 + S ←    A21 − S k− 7

k1cAθ S − k−1θ A−S

k 7 c A21θ S − k −7θ A21 − S

36


Page 36 of 50

Page 37 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


8

k8 → A22 + S ←    A22 − S k−8

k8 c A22θ S − k −8θ A22 − S

9

k9 → A3 + S ←    A3 − S k−9

k 9 c A3θ S − k −9θ A3 − S

10

k 10 →  A4 + S ←  A4 − S k−10

k10 c A4θ S − k −10θ A4 − S

Table 3. Estimated model parameters for the oligomerization of isobutene. Values of the RSS of all the data points and the RRMS are also presented. The reference temperature for the evaluation of rate constants was chosen as 50 °C.

Parameter

Estimated Value −2

 m3  mole  ; α1 , × ( s )( kg . cat )  mole 

1.496 × 10−9

−2

3 α 2 , mole ×  m  ; ( s )( kg . cat )  mole 

5.6 × 10−10

−2


4.21 × 10−10

−2

 m3  mole  ; α4, × ( s )( kg . cat )  mole 

1.471 × 10−9

−2


1.35 × 10−10

E1 , J mole−1 ;

22.7 × 103

E2 , J mole−1 ;

26.9 × 103

37



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

E3 , J mole−1 ;

25.4 × 103

E4 , J mole−1 ;

19.9 × 103

E5 , J mole−1 ;

13.6 × 103

RSS , mole m −3 ;

656.23

RRMS , mole m −3 ;

126.29

38


Page 38 of 50

Page 39 of 50

Figure 1.

A schematic of the experimental setup used in the oligomerization of

isobutene.

6.4E-03 experimental data point

Isobutene concentration, mol cm−3

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


6.2E-03

6.0E-03

5.8E-03

5.6E-03

5.4E-03

5.2E-03 0

15

30

45

60

Time, min

39


75

90

105


Figure 2. A typical concentration-time profile for the oligomerization of isobutene in a stirred batch reactor – reaction temperature: 50 °C; system pressure: 2170 kPa (300 psig); stirring speed: 1000 rpm; amount of catalyst: 2.11 g.

2.5E-04 1-TMP trimers

2.0E-04

Concentration, mol cm− 3

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 40 of 50

tetramers 2-TMP

1.5E-04

1.0E-04

5.0E-05

0.0E+00 0

15

30

45

60

Time, min

40


75

90

105

Page 41 of 50

Figure 3. The concentration-time profile of 1-TMP, 2-TMP, trimers and tetramers during the oligomerization of isobutene in a stirred batch reactor – reaction temperature: 50 °C; system pressure: 2170 kPa (300 psig); stirring speed: 1000 rpm; amount of catalyst: 2.11 g.

90

Conversion

Selectivity

80 Conversion and Dimer selectivity, %

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


70 60 50 40 30 20 10 0 50

70 Temperature, °C

41


90


Figure 4. Effect of reaction temperature on the conversion of isobutene and dimer selectivity during the oligomerization of isobutene – isobutene concentration: 55 mol%; system pressure: 2170 kPa (300 psig); stirring speed: 1000 rpm.

2.5

log (initial rate, min mol− 1 g catalyst− 1)

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 42 of 50

2.0 y = -3.449x + 11.535 R² = 0.9614 1.5

1.0

0.5 2.7

2.85

3

1000/T, K−1

42


3.15

Page 43 of 50

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Effect of temperature on the oligomerization of isobutene: Arrhenius-type of plot in the temperature range studied experimentally – system pressure: 2170 kPa (300 psig); stirring speed: 1000 rpm.

43



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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 44 of 50

Figure 6. A schematic of the stirred batch reaction system used for model development.

f1 1 2

3

fm+1

fk+1

fk k

k+1

m

θk∆z z=0

Figure 7. Arrangement of a general numerical grid used to discretize the catalyst mass balance equations. At the surface of the catalyst, z = 0.

44


Page 45 of 50

7500

6000

Concentration, mole m− 3

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Isobutene, expt 4500

Isobutene simu Isopentane, expt Isopentane, simu

3000

1500

0 0

1500

3000

4500

Time, s

45


6000

7500


Figure 8: The concentration-time profile for the liquid-phase oligomerization of isobutene – temperature: 50 °C, system pressure: 2170 kPa (300 psig); stirring speed: 1000 rpm; amount of catalyst 2.68 g.

500

1-TMP, expt 1-TMP, simu 400

Concentration, mole m− 3

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 46 of 50

2-TMP, expt 2-TMP, simu Trimer, expt

300

Trimer, simu Tetramer, expt Tetramer, simu

200

100

0 0

1500

3000

4500

Time, s

46


6000

7500

Page 47 of 50

Figure 9. The concentration-time profile for the liquid-phase oligomerization of isobutene – temperature: 50 °C, system pressure: 2170 kPa (300 psig); stirring speed: 1000 rpm; amount of catalyst 2.68 g.

700 1-TMP 2-TMP

−3

600

Predicted concentration, mole m

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


500 400 300 200

2

R (1-TMP) = 0.92 2 R (2-TMP) = 0.93

100 0 0

100

200

300

400

500

600

700

7500 Isobutene Isopentane 6000

4500

3000

1500 1500

2

R (Isobutene) = 0.96 2 R (Isopentane) = 0.99 3000

4500

6000

Experimental concentration, mole m

47


−3

7500


Figure 10. Comparison of the experimental and model predicted concentration of different component present in the oligomerization of isobutene.

200

Trimers

Predicted concentration, mole m− 3

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 48 of 50

Tetramers 150

100

R2 (Trimers) = 0.92 R2 (Tetramers) = 0.85

50

0 0

50

100

Experimental concentration, mole

48


150

m− 3

200

Page 49 of 50

Figure 11. Comparison of the experimental and model predicted concentration of trimers and tetramers present in the oligomerization of isobutene.

6000

Isobutene 1-TMP

5000

2-TMP

Concentration, mole m−3

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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Trimers

4000

Tetramers

3000

2000

1000

0 0.00

0.15

0.30

Distance from catalyst surface, mm

49


0.45


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 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 12. Simulated concentration profile for isobutene and oligomeric products inside catalyst particle after 7200 s of reaction at 50 °C and 2170 kPa (300 psig).

50


Page 50 of 50

Selective Oligomerization of Isobutene on Lewis ... - ACS Publications

Recommend Documents