Hydroisomerization of a Refinery Naphtha Stream over Agglomerated

María Jesús Ramos , Juan Pedro Gómez , Fernando Dorado , Paula Sánchez , José Luis Valverde. Chemical Engineering Journal 2007 126 (1), 13-21...
0 downloads 0 Views 192KB Size
9050

Ind. Eng. Chem. Res. 2005, 44, 9050-9058

Hydroisomerization of a Refinery Naphtha Stream over Agglomerated Pd Zeolites Marı´a Jesu ´ s Ramos,† Juan Pedro Go´ mez,‡ Fernando Dorado,† Paula Sa´ nchez,† and Jose´ Luis Valverde*,† Departamento de Ingenierı´a Quı´mica, Facultad de Ciencias Quı´micas, Universidad de Castilla-La Mancha, Avenida Camilo Jose´ Cela s/n, 13071 Ciudad Real, Spain, and Centro Tecnologı´a Repsol-YPF, Ctra. Nnal. V, km 18, 28931 Mo´ stoles, Spain

Pd-mordenite (PdMOR), Pd-beta (PdBETA), and Pd-ZSM-5 (PdZSM-5) agglomerated with bentonite (Bent) were tested in the hydroisomerization of a C7-C8 fraction obtained by distillation of a real naphtha stream. Depending on the catalysts, the overall paraffins conversion obtained decreased in the following order: PdBETABent > PdMORBent > PdZSM-5Bent. Concerning the hydroisomerization products, the absence of methane and ethane revealed that hydrogenolysis did not contribute to cracking reactions. High aromatic and particularly benzene conversion was reached. Methylcyclopentane and cyclohexane were the typical products obtained during the hydrogenation of benzene over bifunctional catalysts, whereas methylcyclohexane was the main product in the hydrogenation of toluene. Some isomerization and ring-opening reactions were produced from these naphthenic compounds. Finally, the research octane number (RON) of the product stream was estimated. The increase of the RON was significantly higher in the catalysts based on agglomerated beta zeolite, because of the high amounts of multibranched isomers and aromatic compounds in the product stream, always satisfying the limits imposed by legislation. 1. Introduction Isomerization of paraffins is a process that frequently occurs in petroleum refinery schemes. In a conventional isomerization unit, low-molecular-weight paraffins are converted to isoparaffins in the presence of an acid catalyst. Isomerization is one of the several reactions occurring in the reforming of naphthas, which is undertaken to upgrade low-octane naphtha to a higheroctane effluent. Under the process conditions of reforming, other reactions could occur, such as aromatization (or dehydrocyclization) and dehydrogenation, along with some cracking.1 Environmental concerns have prompted legislation to limit the amount of total aromatics, and particularly benzene, in gasoline. Reductions of aromatics have a negative effect on the octane number that has to be compensated by other means.2 Isomerization reactions constitute the most important alternative due not only to the production of branched paraffins with high octane number but also to the fact that ring-opening reactions of aromatics compounds from reformer feedstocks are generated. For example, cyclohexane, a benzene precursor, can be rearranged over commercial paraffin isomerization catalysts to yield a mixture of branched paraffins. To develop isomerization reactions, acid catalysts are required: for example, zeolites.3-6 However, the performance of such catalysts has been confirmed almost exclusively by using either powders or tablet-shaped catalysts formed by simply compressing and hardening * To whom correspondence should be addressed. E-mail: [email protected]. Tel.:+34-926-29 53 00. Fax: +34926-29 53 18. † Universidad de Castilla-La Mancha. ‡ Centro Tecnologı´a Repsol-YPF.

zeolite powders. It is essential that catalysts be shaped for industrialization, improving the crush strength to prevent the catalyst from breaking down into powderlike materials in commercial use. Toward this end, materials such as clays, oxides, etc., have been used as binders. Information regarding the influence of the binder on the acidity and catalytic performance of zeolites is very important for the development of industrial catalysts. Some authors have studied the influence of different binders on the catalytic performance of zeolites.7-10 In previous works,5,6 for example, the advantages of bentonite as a binder were reported. Concerning hydroisomerization reactions, most of the research efforts have focused on hydrocarbon mixtures, mainly binary and ternary ones.3,4 Jime´nez et al.4 studied the hydroisomerization of a hydrocarbon feed containing n-hexane, n-heptane, and cyclohexane over zeolite catalysts. Tests involving binary mixtures of the three hydrocarbons revealed that n-heptane was the main source of cracking products. Gopal et al.3 reported the hydroisomerization of C5-C7 alkanes and simultaneous saturation of benzene over Pt/H-ZSM-12 and Pt-beta. Although Pt-beta was much more active and had better sulfur resistance, the capability of Pt/H-ZSM12 to provide a significantly higher yield of isomers made it an attractive catalyst. In contrast, the hydroisomerization of a real naphtha stream has not been widely studied. It is clear that branched paraffins are an excellent gasoline component, because their high octane number. Also, aromatic and naphthenic components contribute to the octane number, keeping branched paraffins under the specifications imposed by legislation. Therefore, it is highly desirable to increase the contribution of high-octane branched paraffins to the gasoline pool, keeping the amount of aromatic compounds under these specifications. Toward

10.1021/ie050765l CCC: $30.25 © 2005 American Chemical Society Published on Web 10/12/2005

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005 9051

this end, agglomerated catalysts based on mordenite, beta, and ZSM-5 zeolites were tested in the hydroisomerization of a C7-C8 stream obtained by distillation of a real naphtha stream. 2. Experimental Section 2.1. Catalyst Preparation. The parent zeolites mordenite (Si/Al ) 10.4), beta (Si/Al ) 13.0), and ZSM-5 (Si/Al ) 15.6) were supplied in the ammonium form by Zeolyst International. To obtain the acid form, the zeolites were calcined at 550 °C for 15 h. Bentonite was supplied by Aldrich Chemical Co. Mordenite, beta, and ZSM-5 samples in the acid form are denoted HMOR, HBETA, and HZSM-5, respectively. Zeolite (35 wt %) and bentonite (65 wt %) were mixed and suspended in water at 60 °C for 2 h. The suspension was then dried at 120 °C for 12 h. After grinding and sieving, particles with an average particle size of 0.75 mm were obtained. Finally, the agglomerated zeolite was calcined at 550 °C for 15 h. After the agglomeration process, the samples with mordenite and ZSM-5 zeolites were ion-exchanged with 0.6 N HCl (35 mL‚g-1), whereas for the samples with beta zeolite, the ion-exchange process was carried out three times with 1 M NH4Cl (30 mL‚g-1). Then, the samples were calcined again at 550 °C for 15 h. Metal incorporation was carried out by the impregnation technique: The sample was placed in a glass vessel and kept under vacuum at room temperature for 2 h in order to remove water and other compounds adsorbed on the zeolite. A known volume of an aqueous metal precursor solution [Pd(NO3)2] was then poured over the zeolite. The solvent was then removed by evaporation under vacuum. The metal concentration of the impregnating solution was calculated to obtain a final palladium content of 1 wt %. After the impregnation process, the catalysts were calcined at 450 °C for 4 h and reduced in situ under a hydrogen flow of 190 mL‚min-1‚g-1. The final, metal-incorporated catalysts are identified first with the symbol of the metal (Pd), followed by the designation of the zeolite (MOR, BETA, and ZSM-5). The suffix “Bent” (bentonite) is also added to indicate the nature of the binder. 2.2. Catalyst Characterization. Surface area was determined by using N2 as the sorbate at 77 K in a static apparatus (Micromeritics ASAP 2010 adsorptive and desorptive apparatus). The samples were evacuated under a vacuum of 5 × 10-3 Torr at 350 °C for 15 h. Specific total surface areas were calculated using the BET equation. Surface area measurements had an error of (3%. To quantify the total amount of metal incorporated into the catalysts, atomic absorption (AA) measurements were performance using a SpectrAA 220FS spectrophotometer. The error of these measurements was (1%. The total acid site density and the acid strength distribution of the catalysts were measured by temperature-programmed desorption of ammonia (TPDA), using a Micromeritics TPD/TPR 2900 analyzer. The samples were housed in a quartz tubular reactor and pretreated in flowing helium (g99.9990% purity) while being heated at 15 °C‚min-1 to the calcination temperature of the sample. After the catalysts had been reduced under a hydrogen flow, the samples were cooled to 180 °C and saturated for 15 min in an ammonia

Figure 1. Pilot-plant distillation column scheme (T ) thermocouple, R ) resistance, LD/D ) reflux ratio).

stream (g99.9990% purity). The catalysts were then allowed to equilibrate in a helium flow at 180 °C for 1 h. Next, ammonia was desorbed by heating at a rate of 15 °C‚min-1. Temperature and detector signals were simultaneously recorded. The total acidity was obtained by integration of the area under the curve. This curve was fitted using two peaks, which were classified as weak and strong acidity depending on the desorption temperature. The use of these peaks was not based on any peak assignment to a specific Bro¨nsted or Lewis acid sites, but it was a convenient way to categorize the acid strength distribution obtained by this method. The average relative error in the acidity determination was lower than 3%. The chemisorption measurements were carried out using a dynamic pulse technique with an argon flow of 50 mL‚min-1 and pulses of H2 (g99.9995% purity). To calculate the metal dispersion, an adsorption stoichiometry of Pd/H ) 1 was assumed.11 The apparatus used was the same as that described for TPDA. Dispersion measurements with H2 pulses were carried out at 60 °C to avoid the spillover phenomenon.12 Previously, the sample was pretreated by heating at 15 °C‚min-1 in argon flow to 250 °C and kept constant at this temperature for 20 min. Then, the sample was reduced in situ. Next, the hydrogen was removed by being placed in flowing argon for 30 min, at a temperature that was 10 °C higher than the reduction temperature. Finally, the sample was cooled to the experimental temperature in an argon gas flow. The dispersion measurements with H2 pulses had an error of (5%. 2.3. Distillation Procedure. A naphtha stream supplied by the company REPSOL-YPF (Madrid, Spain), containing n-paraffins, isoparaffins, olefins, aromatics, and naphtenes, was the feed to a pilot-plant distillation unit, the scheme of which is shown in Figure 1. Table 1lists the naphtha stream composition. The most abundant fraction was that with carbon atom numbers in

9052

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005

Table 1. Molar Composition of the Naphtha Stream number of carbon atoms

n-paraffins

isoparaffin s

5 6 7 8 9 10 11 12

0.04 4.82 6.71 5.47 2.32 0.57 0.07 0.01

0.01 3.81 6.97 9.81 4.98 1.83 0.40 0.00

the range between 7 and 8. To define a specific stream to develop a hydroisomerization process, a distillation process was carried out. It obtained three main fractions: C5-C6 and C7-C8 fractions and a third fraction consisting of hydrocarbons with carbon atom numbers equal to or higher than 9. To estimate the boiling temperature of each fraction, a HYSYS application (by AspenTech) was developed. In this way, the distillation temperatures of each of the three fractions were found to be 59-98, 98-127, and > 127 °C, respectively. The corresponding ASTM and TBP curves calculated using HYSYS are shown in Figure 2. The batch distillation procedure was as follows: The naphtha stream was placed in the reboiler. The temperature was increased to reach the boiling point of the liquid feed. At first, all of the vapor phase was condensed and returned to the column. Once the steady state was achieved, a liquid phase was obtained as the distillate (reflux ratio of 0.4). 2.4. Catalytic Experiments. The hydroisomerization reactions of the C7-C8 fraction were carried out in an Autoclave Engineers (BTRS-Jr.) microreactor that consisted of a tubular stainless steel reactor with vertical placing and downward flow. The liquid feed was performed by a HPLC pump. A back-pressure regulator valve allowed high-pressure experiments. Experimental conditions were as follows: weight of catalyst, 1.5 g; temperature, 250-390 °C; total pressure, 10 bar; WHSV, 10 gC7-C8‚h-1‚gzeolite-1; and H2/hydrocarbon molar ratio, 14. All data were collected after 1 h on stream. Reaction products were analyzed with an HP 5890 Series II gas chromatograph equipped with a flame ionization detector and an automatic valve for continuous analysis. The reactor effluent stream was sent for analysis through a heated line (about 180 °C) to the automatic valve. The gas chromatograph was equipped with a capillary column (Supelco Petrocol DH50.2, 0.2-mm i.d. and 50-m

Figure 2. ASTM (9) and TBP (2) curves obtained by HYSYS.

molar composition (%) olefins naphtenes 0.00 0.00 0.88 0.25 0.94 0.11 0.00 0.01

0.12 4.83 10.50 9.63 3.81 1.32 0.00 0.00

aromatics

total

0.00 0.90 6.91 9.17 2.22 0.49 0.03 0.06

0.17 14.36 31.97 34.33 14.27 4.32 0.50 0.08

Table 2. Characterization Data of the Raw Materials and the Three Catalysts acidity values (mmolNH3‚gcat-1) surface area (m2‚gcat-1)

D H2 (%)

total

bentonite HMOR HBETA HZSM-5 PdMORBent

37 560 636 412 228 (3.7%)a

24.4

0.038 0.996 0.626 0.573 0.378

PdBETABent

233 (2.5%)a

catalyst

PdZSM-5Bent 163 (3.0%)a

weak (T, °C)b

0.038 (274) 0.164 (312) 0.129 (273) 0.085 (275) 0.127 (291) 0.082c 25.3 0.240 0.090 (292) 0.070c 15.6 0.199 0.055 (304) 0.054c

strong (T, °C)b 0.832 (479) 0.497 (353) 0.488 (390) 0.251 (422) 0.291c 0.150 (369) 0.174c 0.144 (385) 0.171c

a Deviation from the theoretical surface area value (%). b Desorption ammonia temperature (°C). c Predicted acidity value calculated from the contribution of the nonagglomerated zeolite and the binder.

length). Results from a reproduced experiment showed that conversion and isomer selectivity had an error of (4%. 3. Results and Discussion 3.1. Catalyst Characterization. The characterization data of the raw materials and the three catalysts used in this work are reported in Table 2. The low acidity value of the bentonite, even not considering strong acidity, is remarkable. Table 2 shows the predicted values of weak and strong acidities calculated from the contributions of the raw materials (zeolite and bentonite). Higher values of the weak acid density of the catalysts were observed than were predicted from the contributions of the raw materials, and the opposite effect was observed for the strong acidity. No blocking of the zeolite channels by the metal or by the bentonite was observed: the experimental values of the surface area were in good agreement with the theoretical ones (based on the calculated deviation from the theoretical surface area values). The decrease of the number of strong acid sites could be due to solid-state ion exchange between the zeolite protons and the clay sodium cations during the calcination that follows the incorporation of the acid function.13-15 The Na+ cations are also weak acid sites and alter the density of these sites.5,6 Moreover, the acid treatment with HCl during the catalyst preparation (especially for ZSM-5) could produce partial dealumination of zeolite, the aluminum remaining in the form of extraframework aluminum (EFAL) species. In a previous work,5 the influence of the binder on catalyst properties (acidity, catalytic performance, etc.) was extensively studied. From the metal dispersion values (Table 2), it could be possible to claim that the metal particles are located mainly on the external surface of the zeolite crystals. The average diameter of the palladium particles16,17 was

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005 9053 Table 3. Molar Composition (%) of the C7-C8 Distillate Stream (Feed) number of carbon atoms

n-paraffins

isoparaffin s

olefins

naphtenes

aromatics

total

6 7 8

2.75 17.86 8.25

0.87 11.31 20.27

0.00 0.13 0.00

3.88 19.40 9.56

0.73 4.99 0.00

8.23 53.69 38.08

about 47-74 Å, too high for them to be located inside the main channels of the zeolites.5 3.2. Hydroisomerization of the C7-C8 Stream. Different experimental runs were carried out with three different zeolites: mordenite, beta, and ZSM-5, each agglomerated with bentonite. The level of overall paraffins conversion was changed by varying the reaction temperature. The composition of the C7-C8 stream, obtained by distillation of the real naphtha stream, is provided in Table 3. Thermal conversion of the C7-C8 distillate stream, less than 1 mol %, can be neglected at the experimental conditions used. The overall paraffins conversion described in this work was defined as the conversion of linear hydrocarbons, i.e., conversion of n-heptane and n-octane. For the catalysts examined, the overall paraffins conversion obtained (Figure 3) decreased in the following order: PdBETABent > PdMORBent > PdZSM-5Bent.5,6 In all cases, conversion increased with increasing reaction temperature. The low overall paraffins conversion obtained with sample PdZSM-5Bent could be related to its low acid site density as well as its low metal dispersion value. It is interesting to compare the overall paraffins conversion for PdBETABent and PdMORBent. For similar metal dispersion values (Table 2), the activity of beta zeolite was higher than that of mordenite zeolite. Because the numbers of metal sites were similar in the two catalysts, the acidity should be responsible for the different activities. However, because conversion is related to the acid site density,18 PdMORBent should be more active than PdBETABent because of the high acidity of the former (Table 2). Nevertheless, the opposite effect was observed. Both the zeolite acidity (number and strength distribution) and the pore structure play important roles in the overall paraffins conversion. Mordenite has a onedimensional pore structure with side pockets, whereas beta consists of intergrown linear channels of 12-MR and tortuous channels with intersections. It has been reported19,20 that only one-third to two-thirds of the acid sites from mordenite are accessible to alkanes. Carvill

Figure 3. Overall paraffins conversion (mol %) versus reaction temperature for the three catalysts PdMORBent, PdBETABent, and PdZSM-5Bent.

et al.21 have argued that, in one-dimensional zeolites, the catalyst activity is controlled by the accessibility to the active sites, because of the effect of “single-file diffusion”. Actually, the overall paraffins conversion was lower for PdMORBent than for PdBETABent despite the higher acid site density of the former. Similar results were obtained by Gopal et al.3 for the hydroisomerization of C5-C7 alkanes over Pt/H-beta. The threedimensional structure of beta zeolite allows easy access to a greater number of active sites compared to onedimensional zeolites. In recent papers,5,6 we have reported the influence of the presence of a binder in some zeolites. The catalytic performance of agglomerated beta zeolite was influenced by the presence of extraframework aluminum (EFAL) species, which were responsible for the high conversion obtained with this catalyst, even higher than that obtained with mordenite and ZSM-5 zeolites. The presence of EFAL species could increase the catalytic activity in beta zeolites. These species could interact with the structural Bro¨nsted acid sites, enhancing their acid strength through a synergetic effect and, consequently, making them much more active for the hydroisomerization. EFAL species were introduced by the binder into the zeolite structure.5,6 The molar product distribution, at 50 mol % of overall paraffins conversion, is reported in Table 4. Whereas the overall paraffins conversion was changed by varying the reaction temperature, the product distribution was a function only of the level of conversion, regardless of the reaction temperature. The same results have been observed by other authors.22 Over all catalysts, the same types of products were obtained, including C1-C6 hydrocarbons (considered as cracking products obtained from the β-scission of C7 and C8 hydrocarbons), products with 7-8 carbon atoms (monobranched, dibranched, and tribranched isomers and the corresponding linear paraffins), and naphthenic and aromatics compounds. C9 or higher products were not observed. The absence of methane and ethane revealed that hydrogenolysis did not contribute to the cracking reaction. It is assumed that a catalyst is ideal when the metallic function is present in sufficient excess amount, meaning that the reactions in the acid sites can be considered as the rate-limiting step.23 Yet, if the metal content is too high, secondary reactions that are exclusively catalyzed by the metallic function (e.g., hydrogenolysis) would become the main reactions.24 In a previous work,25 the influence of the palladium and platinum loading on the hydroisomerization of n-octane over agglomerated beta zeolite was studied. It was observed that the yield of octane isomers increased with the hydrogenating/acid balance, then remaining constant at a palladium content higher than 1 wt %. In the latter case, the isomerization reaction over the acid sites was the limiting step of the n-octane transformation, as the yield of octane isomers did not depend on the metal content. The presence of aromatics compounds (benzene and toluene) in the C7-C8 distillate stream was observed

9054

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005

Table 4. Composition (mol %), at 50 mol % of Overall Paraffins Conversion, for the Products Obtained in the Hydroisomerization of the C7-C8 Distillate Stream catalyst product

feed

PdMORBenta

PdBETABentb

PdZSM-5Bentc

C1 + C2 C3 iso-C4 n-C4 iso-C5 n-C5 2,3-DMC4 2-MC5 3-MC5 n-C6 2,2-DMC5 2,4-DMC5 2,2,3-TMC4 3,3-DMC5 2-MC6 2,3-DMC5 3-MC6 n-C7 2,2-DMC6 2,5-DMC6 2,4-DMC6 3,3-DMC6 2,3,4-TMC5 2-M-3-EC5 2,3-DMC6 2-MC7 4-MC7 3,4-MC6 3-MC7 3-EC6 n-C8 naphthenes MCP (methylcyclopentane) CH (cyclohexane) DMCPs (dimethylcyclopentanes)d MCH (methylcyclohexane) TMCPs (trimethylciclopentanes)e DMCHs (dimethylciclohexanes)f CH conversion (mol %) MCH conversion (mol %) aromatics benzene toluene benzene conversion (mol %) aromatics conversion (mol %)

0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.36 0.47 2.75 0.11 0.29 0.03 0.11 3.97 1.25 5.55 17.86 1.40 1.66 1.25 0.00 0.14 0.97 0.68 6.42 1.69 0.47 4.33 1.26 8.25

0.00 6.11 8.66 3.27 3.36 0.54 0.14 0.37 0.32 1.75 0.06 0.44 0.03 0.10 2.93 0.96 4.33 12.68 1.34 1.82 1.04 0.34 0.04 0.64 0.77 2.68 1.93 0.47 3.91 3.50 5.07

0.00 7.43 14.38 5.08 4.35 0.97 0.13 0.59 0.46 1.85 0.14 0.41 0.05 0.15 2.24 1.26 4.31 8.59 1.26 1.18 1.01 0.42 0.22 0.80 1.14 1.08 1.29 0.63 2.44 3.14 4.26

0.00 12.86 9.27 4.71 3.40 3.49 0.10 0.35 0.32 1.56 0.07 0.16 0.00 0.18 2.13 0.66 2.98 10.51 1.02 1.42 0.82 0.25 0.07 0.65 0.62 4.10 1.46 0.34 3.27 2.91 5.20

1.39 2.49 6.53 12.87 3.33 6.24 -

1.71 1.60 7.15 10.06 2.53 5.43 35.74 21.83

2.06 1.70 7.55 8.78 3.16 3.65 31.72 31.78

1.06 1.63 5.27 10.77 1.96 4.19 34.54 16.32

0.73 4.99 -

0.07 0.69 90.88 86.86

0.06 1.64 92.25 70.42

0.00 0.25 100 95.56

a For PdMORBent, reaction temperature (°C) ) 390, overall paraffins conversion (mol %) ) 42.81. b For PdBETABent, reaction temperature (°C) ) 370, overall paraffins conversion (mol %) ) 49.03. c For PdZSM-5Bent, reaction temperature (°C) ) 390, overall paraffins conversion (mol %) ) 40.16. d DMCP ) 1,1-dimethylcyclopentane, 1,2-trans-dimethylcyclopentane, 1,3-trans-dimethylcyclopentane, 1,2cis-dimethylcyclopentane, 1,3-cis-dimethylcyclopentane. e TMCP ) 1-trans-2-cis-4-trimethylcyclopentane, cis,trans,cis-1,2,3-trimethylcyclopentane. f DMCH ) cis-1,3-dimethylcyclohexane, trans-1,4-dimethylcyclohexane, trans-3-ethylcyclopentane, cis-3-ethylcyclopentane, trans-2-ethylcyclopentane, 1-ethyl-1-methylcyclopentane, trans-1,2-dimethylcyclohexane.

(Table 4). Environmental concerns have prompted legislation to limit the amount of total aromatics, particularly benzene, in gasoline.2,3 The specifications are less than 1% (v/v) of benzene and no more than 35% (v/v) of aromatic compounds. The reduction of aromatics will have a negative impact on gasoline octane ratings. To satisfy the environmental specifications, the total hydrogenation of benzene could be achieved, keeping the rest of the aromatic compounds under the limit imposed by legislation. The catalyst PdBETABent seems to be the most suitable catalyst because it reduced the aromatic content under the specifications in addition to allowing a considerable octane number of the product to be maintained, as will be discussed later. Methylcyclopentane (MCP) and cyclohexane (CH) are the typical products obtained during the hydrogenation of benzene over bifunctional catalysts.2,3,26 Benzene hydrogenation occurs as a monofunctional reaction on

metals via three subsequent hydrogenation steps, leading to the formation of CH, which, in the presence of acid sites, produces MCP by isomerization.26 In Table 4, the molar distributions of MCP and CH are presented. The amount of CH in the reaction product was lower than the amount in the C7-C8 stream used as the feed, and the opposite effect was observed for MCP. Figure 4 shows the MCP/CH ratio versus the reaction temperature for the three catalysts used in this work. In accordance with the thermodynamic equilibrium, MCP is produced in higher quantities during the hydroisomerization.3 In fact, the MCP/CH ratio increased with the reaction temperature. The highest MCP/CH ratio was obtained with the sample PdBETABent. The synergetic effect caused by EFAL species over the acid sites of beta zeolite5,6 would contribute to the high production of MCP from the isomerization reaction of CH. This fact

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005 9055 Table 5. Percentages of Cracking Products Obtained at 50 mol % of Overall Paraffins Conversion catalyst product

PdMORBenta

PdBETABentb

PdZSM-5Bentc

C4/(C3 + C5) iso-C4/n-C4 iso-C5/n-C5 C3/C5 C3/C4

1.2 2.6 6.2 1.6 0.5

1.5 2.8 4.5 1.4 0.4

0.7 2.0 1.0 1.9 0.9

a For PdMORBent, reaction temperature (°C) ) 390, overall paraffins conversion (mol %) ) 42.81. b For PdBETABent, reaction temperature (°C) ) 370, overall paraffins conversion (mol %) ) 49.03. c For PdZSM-5Bent, reaction temperature (°C) ) 390, overall paraffins conversion (mol %) ) 40.16.

Figure 4. MCP/CH ratio versus reaction temperature for the three catalysts PdMORBent, PdBETABent, and PdZSM-5Bent.

is a consequence of beta’s three-dimensional framework, which does not cause steric hindrances. PdZSM-5Bent yielded the lowest amount of MCP (Table 4), suggesting a high capacity to support ringopening reactions, namely, (i)endocyclic rupture resulting in n-hexane, 2-methylpentane, and 3-methylpentane or (ii) exocyclic C-C rupture resulting in cyclopentane formation plus methane.27 Neither cyclopentane nor methane were observed as products, indicating that the ring-opening reaction of MCP would yield n-hexane and the monobranched isomers (2-methylpentane and 3-methylpentane). However, over the catalyst PdZSM-5Bent, low amounts of these products were observed (Table 4). Among the cracking products from C6 hydrocarbons, it is possible to obtain 2C3 (depropylation), C2 + C4 (deethylation), and C1 + C5 (demethylation).27 The high amount of propane when PdZSM-5Bent was used and the lack of methane and ethane in the products would confirm the cracking of C6 hydrocarbons toward propane. On the other hand, toluene hydrogenation yields methylcyclohexane (MCH).28,29 The rearrangement of MCH in terms of isomerization reactions leads to dimethyl- (1,1-, 1,2-cis and -trans, and 1,3-cis and -trans) and ethylcyclopentanes (DMCP and EtCP, respectively). Isomerization of MCH to DMCP and EtCP is assumed to take place on the acid sites of the zeolite, whereas the dehydrogenation of toluene is performed by a metal function.30 The MCH composition in the product was lower than that in the feed (Table 4). The opposite effect was observed for the composition of DMCP when PdMORBent and PdBETABent were used as catalysts. The presence of these naphthenic compounds would contribute to an increase in the octane number in the product. However, the DMCP molar composition in the product obtained using the catalyst PdZSM-5Bent was lower than that observed in the feed, indicating the occurrence of ring-opening reactions from DMCP.31 As a consequence, branched C7 hydrocarbons (2-methylhexane, 3-methylhexane, 2,2-dimethylpentane, 2,3-dimethylpentane, 2,4-dimethylpentane, and 3,3-dimethylpentane) that underwent successive cracking reactions, yielding mainly C3 and C4 hydrocarbons, were obtained. On the other hand, the intrinsic reactivity of MCH on acid catalysts might be higher than that of CH, because MCH has a tertiary carbon.32 However, the MCH conversion over PdZSM-5Bent was lower than that observed for CH (Table 4). A similar trend was observed when PdMORBent was used as the catalyst. In the case of PdBETABent, the conversions of both

MCH and CH were nearly the same. The molecular shapes suggest that the diffusivity of MCH inside the pores might be lower than that of CH. Therefore, the rate of MCH disappearance was probably controlled by its own diffusivity in the pores of ZSM-5, to a greater extent than that observed for CH conversion.32 The diffusional constraints of the one-dimensional framework of mordenite would cause the same effects. Nevertheless, both MCH and CH were converted at the same level when PdBETABent was used as the catalyst, because of the lower diffusional limitations presented in the beta zeolite framework. Other naphthenic compounds such as dimethylcyclohexanes (DMCHs) and trimethylcyclopentanes (TMCPs) were detected as final products, in a lower concentration than that observed in the feed. These compounds likely undergo ring-opening reactions,31 yielding branched octane isomers. Furthermore, because the concentration of branched octane isomers was lower in the products than in the feed, one can deduce that these compounds underwent cracking reactions. Concerning the cracking products, mostly C3, C4, and C5 hydrocarbons were detected. C3 and C4 products could result from the classical n-heptane cracking mechanism. This was not the case for C5 and C6 because complementary C2 and C1 fragments were not observed. Cracking products with three, four, and five carbon atoms could be also obtained from n-octane cracking. Table 5 reports the percentages of different cracking products at 50 mol % of overall paraffins conversion. Over PdMORBent and PdBETABent, the cracking pathway toward C4 was favored (high values of C4/(C3 + C5) ratio). In contrast, over PdZSM-5Bent, more cracking to C3 + C5 was found. Higher iso-C4/nC4 and iso-C5/n-C5 ratios were obtained for PdMORBent and PdBETABent, which could result only from the cracking of di- and tribranched isomers. The low ratios obtained for PdZSM-5Bent was a consequence of cracking from dibranched isomers.33 To confirm the main cracking pathway occurring over each catalyst, C3/ C5 and C3/C4 ratios were compared. The fact that the concentrations of C3 and C5 hydrocarbons were almost equal would indicate that they were the result of primary cracking of either linear or branched C8 carbenium ions.18 Therefore, n-octane cracking was favored over catalysts PdMORBent and PdBETABent. The contribution of classical n-heptane cracking was estimated as the molar yield of C3 and an equimolar yield of C4,34 as can be observed with sample PdZSM-5Bent. Because the purpose of the hydroisomerization process is to achieve an increase in the octane number of the gasoline, it is interesting to note the formation of multibranched isomers from the C7-C8 distillate stream

9056

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005

Figure 6. Estimated research octane number (RON) as a function of overall paraffins conversion for the three catalysts PdMORBent, PdBETABent, and PdZSM-5Bent.

Figure 5. Multibranched isomers (mol % obtained from the total content of mono and multibranched) versus overall paraffins conversion for the three catalysts used: (a) percentage of C7 multibranched isomers and (b) percentage of C8 multibranched isomers.

used as the feed. Figure 5 shows the percentages of C7 (Figure 5a) and C8 (Figure 5b) multibranched isomers in the total branched isomers versus the overall paraffins conversion obtained for the three catalysts. These values in the C7-C8 distillate stream were 16% for C7 multibranched isomers and 32% for C8 multibranched ones. The percentages of both isomers increased with increasing overall paraffins conversion, with this increase being greater when PdBETABent was used as the catalyst. The high amount of multibranched isomers obtained with the sample PdMORBent at low overall paraffins conversion is remarkable; however, when PdBETABent was used as the catalyst, an increase in the overall paraffins conversion led to a higher amount of multibranched isomers.35 The three-dimensional structure of beta facilitates a rapid diffusion of branched products and avoids their interaction with other sites that could cause their cracking.4 However, in mordenite (with a one-dimensional pore system), paraffinic intermediates might have longer lifetimes on strong acid sites and more possibilities to undergo successive rearrangement and cracking reactions.36 To compare the performance of the different catalysts used in this work, the octane number of the C5+ fraction in the product was estimated. Figure 6 shows this estimation versus the overall paraffins conversion. The RON of the mixture was evaluated from the products of the volume fractions of the individual C5 and higher hydrocarbons and their corresponding RONs; the contributions of all of the compounds were summed to obtain the overall value.3 In this way, the octane

number of the C7-C8 distillate stream (feed) was found to be 43.7. In practice, octane numbers do not blend linearly. To accommodate this fact, complex blending calculations employing blending octane numbers as opposed to the values for pure hydrocarbons are routinely employed. In general, the blending octane numbers are greater than the corresponding pure octane numbers, e.g., MCP has a higher research octane number than CH, but the blending octane numbers of both MCP and CH are similar. Therefore, the formation of both products from benzene should be acceptable. This method for estimating the RON value of mixtures is likely to be different from the real RON value, with the latter generally being higher.3 PdBETABent yielded a higher amount of multibranched isomers, which lead to higher RON values than those corresponding to linear alkanes. However, the contribution of the aromatic compounds to the RON of PdBETABent was more important. In fact, the aromatic conversion obtained with this catalyst was lower than that obtained with the other two catalysts (Table 4), keeping the aromatic content under the imposed specifications by legislation. Hence, the multibranched isomers, and mainly the aromatic products, contributed to the high RON value reported for sample PdBETABent, making it an attractive industrial catalyst for use in the hydroisomerization of real naphtha streams. 4. Conclusions Because branched paraffins are an excellent gasoline component, the hydroisomerization of a C7-C8 stream obtained by distillation of a real naphtha stream was carried out. The overall paraffins conversion described in this work was defined as the conversion of linear hydrocarbons, i.e., conversion of n-heptane and n-octane. Regarding the zeolitic supports, the overall paraffins conversions obtained with the different catalysts decreased in the following order: PdBETABent > PdMORBent > PdZSM-5Bent. In all cases, conversion increased with increasing reaction temperature. Despite the higher acid site density of other catalysts, the highest overall paraffins conversion was obtained with the catalyst PdBETABent. Regardless of the catalyst, the same types of products were obtained, including C1-C6 hydrocarbons (consid-

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005 9057

ered as cracking products obtained from the β-scission of C7 and C8 hydrocarbons), products with 7-8 carbon atoms (monobranched, dibranched, and tribranched isomers and the corresponding linear paraffins), naphtenes, and aromatic compounds,. The absence of methane and ethane indicated the absence of the hydrogenolysis reaction. High aromatics and particularly benzene conversions were obtained. For all of the catalysts, the aromatic content was kept under legislated gasoline specifications. n-Octane cracking was favored over samples PdMORBent and PdBETABent, whereas n-heptane cracking favored in catalyst PdZSM-5Bent. The greatest increase of the RON value was achieved when sample PdBETABent was used as the catalyst, because of the high contributions of multibranched isomers and aromatic products.

Acknowledgment Financial support from Ministerio de Ciencia y Tecnologı´a (CICYT) of Spain (Projects PPQ2001-1195-C0201) and Consejerı´a de Ciencia y Tecnologı´a de la Junta de Comunidades of the Region of Castilla-La Mancha (Project PBI-05-038) is gratefully acknowledged. Literature Cited (1) Huss, A., Jr.; Harandi, M. N.; Esteves, D. J.; Dovedytis, D. J.; Del Rossi, K. J. Combined paraffin isomerization/ring opening process for C5+ naphtha. U.S. Patent 5,334,792, 1994. (2) Arribas, M. A.; Ma´rquez, F.; Martı´nez, A. Activity, selectivity, and sulfur resistance of Pt/WOx-ZrO2 and Pt/Beta catalysts for the simultaneous hydroisomerization of n-heptane and hydrogenation of benzene. J. Catal. 2000, 190, 309. (3) Gopal, S.; Smirniotis, P. G. Pt/H-ZSM-12 as a catalyst for the hydroisomerization of C5-C7 n-alkanes and simultaneous saturation of benzene. Appl. Catal. 2003, 247, 113. (4) Jime´nez, C.; Romero, F. J.; Rolda´n, R.; Marinas, J. M.; Go´mez, J. P. Hydroisomerization of a hydrocarbon feed containing n-hexane, n-heptane and cyclohexane on zeolite-supported platinum catalysts. Appl. Catal. 2003, 249, 175. (5) De Lucas, A.; Valverde, J. L.; Sa´nchez, P.; Dorado, F.; Ramos, M. J. Influence of the binder on the n-octane hydroisomerization over palladium-containing zeolite catalysts. Ind. Eng. Chem. Res. 2004, 43, 8217. (6) De Lucas, A.; Valverde, J. L.; Sa´nchez, P.; Dorado, F.; Ramos, M. J. Hydroisomerization of n-octane over platinum catalysts with or without binder. Appl. Catal. 2005, 282, 15. (7) Devadas, P.; Kinage, A. K.; Choudhary, V. R. Effect of silica binder on acidity, catalytic activity and deactivation due to coking in propane aromatisation over H-gallosilicate (MFI). Stud. Surf. Sci. Catal. 1998, 113, 425.

(12) Bond, G. C.; Mallat, T. Studies of Hydrogen Spillover. J. Chem. Soc., Faraday Trans. 1981, 77, 1743. (13) Choudhary, V. R.; Devadas, P.; Kinage, A. K.; Guisnet, M. Influence of binder on the acidity and performance of H-gallasilicate (MFI) zeolite in propane aromatisation. Appl. Catal. 1997, 162, 223. (14) Romero, M. D.; Calles, J. A.; Rodrı´guez, A.; De Lucas A. Acidity modification during the agglomeration of ZSM-5 with montmorillonite. Microporous Mesoporous Mater. 1997, 9, 221. (15) Yoon, K. B.; Park, Y. S. Visual observation of salt migration between zeolite and clay in the solid state. J. Chem. Soc., Chem. Commun. 1993, 522. (16) Can˜izares, P.; De Lucas, A.; Valverde, J. L.; Dorado, F. n-Butane hydroisomerization over Pd/HZSM-5 catalysts. Palladium loaded by impregnation. Ind. Eng. Chem. Res. 1998, 37, 2592. (17) Can˜izares, P.; De Lucas, A.; Dorado, F.; Dura´n, A.; Asencio, I. Characterization of Ni and Pd supported on H-mordenite catalysts: Influence of the metal loading method. Appl. Catal. 1998, 169, 137. (18) Zhang, W.; Smirniotis, P. G. Effect of zeolite structure and acidity on the product selectivity and reaction mechanism for n-octane hydroisomerization and hydrocracking. J. Catal. 1999, 182, 400. (19) Kinger, G.; Vinek, H. n-Nonane hydroconversion on Ni and Pt containing HMFI, HMOR and HBEA. Appl. Catal. 2001, 218, 139. (20) Makarova, M. A.; Wilson, A. E.; van Liemt, B. J.; Mesters, C. M. A. M.; De Winter, A. W.; Williams, C. Quantification of Bro¨nsted acidity in mordenites. J. Catal. 1997, 172, 170. (21) Carvill, B. T.; Lerner, B. A.; Adelman, B. J.; Tomczak, D. C.; Sachtler, W. M. H. Increased catalytic activity caused by local destruction of linear zeolite channels: Effect of reduction temperature on heptane conversion over platinum supported in Hmordenite. J. Catal. 1993, 144, 1. (22) Chica, A.; Corma, A. Hydroisomerization of Pentane, Hexane, and Heptane for improving the octane number of gasoline. J. Catal. 1999, 187, 167. (23) Weitkamp, J. Hydrocracking, cracking and isomerization of hydrocarbons. Erdoel, Kohle-Erdgas-Petrochem. 1978, 31, 13. (24) Giannetto, G. E.; Perot, G. R.; Guisnet, M. Hydroisomerization and hydrocracking of n-alkanes. Ind. Eng. Chem. Prod. Res. Dev. 1986, 25, 481. (25) De Lucas, A.; Sa´nchez, P.; Dorado, F.; Ramos, M. J.; Valverde, J. L. Effect of the metal loading in the hydroisomerization of n-octane over beta agglomerated zeolite based catalysts. Appl. Catal. 2005, in press. (26) Simon, L. J.; van Ommen, J. G.; Jentys, A.; Lercher, J. A. Sulfur-tolerant Pt-supported catalysts for benzene hydrogenation. II. Influence of cation exchange level for Pt/MOR-based catalysts. J. Catal. 2001, 203, 434. (27) Issaadi, R.; Garin, F.; Eddine Chitour, C.; Maire, G. Catalytic behaviour of combined palladium-acid catalysts: Use of Al and Zr-pillared montmorillonite as supports. Part I. Reactivity of linear, branched and cyclic hexane hydrocarbons. Appl. Catal. 2001, 207, 323. (28) Osaki, T.; Tanaka, T.; Tai, Y. Hydrogenation of C60 on alumina-supported nickel and thermal properties of C60H36. Phys. Chem. Chem. Phys. 1999, 1, 2361.

(8) Wu, X.; Alkhawaldeh, A.; Anthony, R. G. Investigation on acidity of zeolites bound with silica and alumina. Stud. Surf. Sci. Catal. 2002, 143, 217.

(29) Lylykangas, M. S.; Rautanen, P. A.; Krause, A. O. I. Liquidphase hydrogenation kinetics of multicomponent aromatic mixtures on Ni/Al2O3. Ind. Eng. Chem. Res. 2002, 41, 5632.

(9) Chen, W. H.; Huang, S. J.; Lai, C. S.; Tsai, T. C.; Lee, H. K.; Liu, S. B. Effects of binder, coking and regeneration on acid properties of H-mordenite during TPD reaction. Res. Chem. Intermed. 2003, 29, 761.

(30) Belatel, H.; Al-Kandari, H.; Al-Khorafi, F.; Katrib, A.; Garin, F. Catalytic reactions of methylcyclohexane (MCH) on partially reduced MoO3. Appl. Catal. 2004, 275, 141.

(10) Rough, S. L.; Wilson, D. I.; Bayly, A.; York, D. Tapping characterization of high shear mixer agglomerates made with ultrahigh viscosity binders. Powder Technol. 2003, 132, 249.

(31) McVicker, G. B.; Daage, M.; Touvelle, M. S.; Hudson, C. W.; Klein, D. P.; Baird, W. C., Jr.; Cook, B. R.; Chen, J. G.; Hantzer, S.; Vaughan, D. E. W.; Ellis, E. S.; Feeley, O. C. Selective ring opening of naphthenic molecules. J. Catal. 2002, 210, 137.

(11) Rodrı´guez, F.; Rodrı´guez, I.; Moreno, C.; Guerrero, A.; Lo´pez, J. D. Pt catalysts supported on activated carbons (I). Preparation and characterization. J. Catal. 1986, 99, 171.

(32) Namba, S.; Sato, K.; Fujita, K.; Jim, J. H.; Yashima, T. Shape selective cracking of octane in the presence of another hydrocarbon on HZSM-5. Stud. Surf. Sci. Catal. 1986, 28, 661.

9058

Ind. Eng. Chem. Res., Vol. 44, No. 24, 2005

(33) Lugstein, A.; Jentys, A.; Vinek, H. Hydroisomerization and cracking of n-octane and C8 isomers on Ni-containing zeolites. Appl. Catal. 1999, 176, 119. (34) Blomsma, E.; Martens, J. A.; Jacobs, P. A. Reaction mechanisms of isomerization and cracking of heptane on Pd/HBeta zeolite. J. Catal. 1995, 155, 141. (35) Lee, J.-K.; Rhee, H.-K. Effect of metal/acid balance in Ptloaded large pore zeolites on the hydroisomerization of n-hexane and n-heptane. Korean J. Chem. Eng. 1997, 14, 451.

(36) Chao, K.-J.; Wu, H.-C.; Leu, L. J. Hydroisomerization of light normal paraffins over series of platinum-loaded mordenite and beta catalysts. Appl. Catal. 1996, 143, 223.

Received for review June 28, 2005 Revised manuscript received September 14, 2005 Accepted September 15, 2005 IE050765L