Reactions of hydroconversion of linear alkanes underlie the most important processes used in petrochemical industry and petroleum refinery. For example, hydroisomerization of n-alkanes of light gasoline fractions into their corresponding branched forms is conducive to improvement of octane number of straight-run gasolines, as well as of low-temperature performance of diesel fuels and hydrocarbon base oils of lubricants [1–3].

In recent years, diesel fuel consumption has been increasing steadily, especially in cold climates [4]. Production of fuels with improved low-temperature performance relies on several technical practices and catalytic processes. One of the most common practices consists in increasing the fractional composition content of lighter components. This prevents high-molecular-weight hydrocarbons from getting into marketable fuels but at the same time causes depletion of the raw materials base of production. From this perspective, use of catalytic processes seems more attractive. Selective hydrocracking results in long-chain n-alkane splitting into two smaller molecules, leading to reduction in the cloud point and pour point [5]. Hydroisomerization allows improving the low-temperature characteristics of the feedstock via skeletal isomerization of long-chain alkanes with formation of branched isomers [6]. In both cases catalytic hydroconversion enables reducing the content of n-paraffinic hydrocarbons in the feedstock [7, 8]. Apart from that, the hydroisomerization route allows maintaining higher yields of diesel fuel with improved low-temperature characteristics and minimizing the yield of gaseous hydrocarbons.

Catalysts used in industrial hydrodewaxing and hydroisomerization processes typically comprise expensive platinum group metals (hydrogenating components) and silicoaluminophosphates (acidic components) [9–11]. However, noble metal-based catalysts have extremely poor resistance to catalytic poisons, in particular, low sulfur tolerance [12]. Accordingly, considerable research efforts are being directed now at development of noble metal-free catalysts more tolerant to poisoning. For example, widely used components of hydrotreating catalysts are transition metal sulfides which afford sufficient catalytic activity in hydrogenation/dehydrogenation reactions while comparing favorably with noble metals in terms of their cost [13–15].

Different types of zeolites, zirconium oxides, natural clays, and mesoporous aluminosilicates find use as acidic components of bifunctional catalysts [16–18]. The best performance in hydroisomerization reactions is exhibited by SAPO-11 silicoaluminophosphates [19]. However, in some cases the cracking function of the catalyst is useful alongside the isomerization function. This applies, e.g., to the case of adjusting the fractional composition content of the hydrogenation product while increasing the gasoline yield from heavy feedstock [20, 21]. Various approaches were tested for tuning the selectivity of hydroconversion process, e.g., via changing the number of hydrogenating active sites [22], the type of the hydrogenation component [23], and the type/amount of acid component, specifically zeolite [24–27]. Use of zeolite-containing transition metal sulfide-based catalysts typically results in the dominance of cracking reactions [28, 29].

Some studies demonstrated the predominant deactivation of bifunctional noble metal-based catalysts in the cracking reactions run in the presence of nitrogen-containing compounds in the feedstock [12, 30, 31]. This suggests that processing of nitrogen-containing feedstock or use of nitrogen-containing additives to the prepared feedstock may provide a means for tuning the selectivity of the hydroconversion process. Insufficiently studied for bifunctional transition metal sulfide-based catalysts, this approach is of both scientific and industrial importance.

Herein, we report on the synthesis and study of catalysts based on comparatively less expensive sulfides and ZSM-5, Beta, and Y zeolites as applied to n-hexadecane hydroconversion reaction with the view to assessing the prospects for tuning the process selectivity by adding nitrogen-containing compounds to the feedstock.

EXPERIMENTAL

Samples of routinely used commercial zeolites NH4NaY (Si/Al = 2.5), Beta (Si/Al = 30), and ZSM-5 (Si/Al = 30) served as acidic additives. Composite catalyst supports were produced as extrudate granules with a diameter of 1 mm and a length of 2 mm using pseudoboehmite (Sasol Company, Germany) as a binder and nitric acid as a peptizing agent. The extrudates were dried at 80–120°C for 6 h and then calcined at 550°C for 4 h. After calcination the composite support contained 40 wt % NH4NaY, Beta, or ZSM-5 and 60 wt % γ-Al2O3.

CoMo catalysts were prepared by single-dip incipient wetness impregnation of the support with the solution of active component precursors: phosphomolybdic acid (H3PMo12O40·18H2O), cobalt carbonate hydroxide hydrate [CoCO3∙mCo(OH)2∙nH2O], and chemically pure citric acid. For the synthetic procedure, see [32]. After impregnation the samples were dried at 60°C for 2 h, at 80°C for 2 h, and at 110°C for 6 h. The estimated content of molybdenum(VI) oxide in the catalysts was 15 wt %, and that of cobalt(II) oxide, 3.5 wt %. The active metal loadings were chosen based on the published data and on our own research results [14, 33, 34]. The content of the metals in the catalysts synthesized was monitored using an EDX-800HS Shimadzu X-ray fluorescence analyzer.

The textural characteristics of the composite supports were determined by low-temperature nitrogen adsorption on a Quantachrome Autosorb-1 surface area and pore size analyzer. The specific surface area was calculated using the Brunauer–Emmett–Teller (BET) model in the relative pressure range p/p0 = 0.05–0.3. The total pore volume and pore size distribution were determined from the desorption branch of the isotherm using the Barrett–Joyner–Halenda model. The micropore characteristics were evaluated by the t-plot method.

For analyzing the acidic properties of the materials was used the method of temperature-programmed desorption of ammonia (NH3–TPD). The experiments consisted of three stages. In the first (preparation) stage the adsorbed water was removed from the pores at a temperature of 120°C in a helium flow for 5 h (heating rate 10°C min–1). The second stage involved ammonia adsorption from a mixture of 10 vol % NH3 in helium (gas flow rate 30 cm3 min–1) at a temperature of 60°С (heating rate 10°C min–1) for 30 min, whereupon the unbound NH3 was removed at a temperature of 60°C in a helium flow for 60 min. In the third stage, NH3 was desorbed in a helium flow with heating to 750°C (heating rate 8°C min–1) and holding at this temperature for 45 min.

The catalytic properties of the bifunctional catalysts were examined in a laboratory fixed-bed flow reactor. A 1.5 cm3 volume of the catalyst (0.25–0.50 mm particles) was loaded into the isothermal region of the reactor. Gas phase sulfidation of the CoMo catalysts was carried out at 400°C and 1 MPa in a H2S/H2 flow (10/90 vol %) for 4 h. As a model feedstock for evaluating the catalyst performance in hydroisomerization reaction was used a mixture of n-hexadecane (3 wt %) in n-heptane as a solvent. Under noncatalytic treatment conditions the selected substrates were not subjected to chemical transformations. The catalytic activity was analyzed under the following conditions: temperature range 280–360°C, 1.5 MPa H2, weight hourly space velocity (WHSV) 1–3 h–1, and volume ratio of H2 to feedstock 150 nL/L.

Product identification was performed on a gas chromatograph with a Shimadzu GCMS-QP2010 quadrupole mass detector equipped with a flame ionization detector and a 30 m × 0.5 mm × 0.5 μm column (OV-101 nonpolar phase, dimethylpolysiloxane as stationary phase). Analysis conditions: detector temperature 250°C, vaporizer temperature 250°C; temperature program: isothermal for 2 min at 110°C, then heating to 250°C at 5°C/min. Flow rate of the carrier gas (helium) was 3 mL/min, and velocity, 30 cm/s. For achieving conversion in a steady state, the hydroconversion process was run for at least 8 h under constant reaction conditions.

The products were identified on a Shimadzu single quadrupole GCMS-QP2010 gas chromatograph-mass detector equipped with a 100 m × 0.25 mm × 0.5 μm nonpolar Agilent DB-Petro column (dimethylpolysiloxane as stationary phase). Analysis conditions: vaporizer temperature 250°C, detector temperature 200°C; temperature program: isothermal for 10 min at 140°C, then heating to 290°C at 5°C/min, then holding isothermally at 290°C for 20 min. The carrier gas (helium) flow rate was 3 mL min–1, and linear velocity, 30 cm/s.

The conversion and isomerization selectivity were calculated using the following equations:

$${X_{iso}}{\rm{ = }}{{iso{\rm{ - hexadecanes}}} \over {{\rm{initial}}\,n{\rm{ - hexadecane}}\,}},$$
$$Se{l_{iso}}{\rm{ = }}{{iso{\rm{ - hexadecanes}}} \over {{\rm{consumed}}\,n{\rm{ - hexadecane}}}}.$$

The reaction rate constant for hydroisomerization was determined by the equation:

$${k_{iso}} = - {F \over m}\ln (1 - {X_{iso}}),$$

where kiso is the pseudo-first-order reaction constant for n-hexadecane hydroisomerization, mol g–1 h–1; Xiso, n-hexadecane conversion, %, to C16 isomers; F, reactant (n-hexadecane) flow, mol/h, and m, weight of the catalyst, g.

RESULTS AND DISCUSSION

Table 1 presents the results of the catalytic tests. The catalysts containing Beta and ZSM-5 zeolites as acidic additives showed high cracking activity, while isomerization products were almost completely absent. Throughout the 280–320°C temperature range examined the hexadecane cracking efficiency exceeded 95%. Hence, the catalysts tested in this study can find application in heavy feedstock processing with the aim of obtaining incremental yields of light components in commercial products.

Table 1. Results of the catalytic tests for hydroconversion of n-hexadecanea
Full size table

For inhibition the cracking reactions and increasing the selectivity of the isomerization process we prepared CoMo/Al2O3–ZSM-5(5%) catalyst with a lower content (5 wt %) of acidic additive (ZSM-5), since acid sites of the catalyst are specifically responsible for cracking reactions. However, a decrease in the ZSM-5 content in the catalyst did not cause any noticeable increase in selectivity of isomerization. The cracking route prevailed in the hydroconversion of n-hexadecane over CoMo/Al2O3–ZSM-5(5%) catalyst. A sufficiently high selectivity of isomerization at moderate cracking activity was achieved only with the use of the NH4NaY zeolite-based catalyst. While exhibiting lower overall activity, this catalyst provided a close to 10% isomerization efficiency at 320°C at the cracking efficiency not exceeding 30%. The differences in the catalytic properties between the catalysts tested are directly related to their physicochemical properties.

Figure 1 shows the low-temperature nitrogen adsorption–desorption isotherms of the composite supports synthesized. These are Type IV isotherms which are typical of mesoporous materials. H3-type hysteresis loops are indicative of slit-like pores characteristic of mesoporous alumina [35]. Table 2 describes the textural properties of the samples obtained. All the samples were characterized by the same mesopore size of ~12 nm, usually exhibited by alumina produced from pseudoboehmite, which suggests reproducibility of the preparation conditions. The micropore areas and volumes determined for the samples containing 40 wt % zeolite additive corresponded to the textural characteristics of the original zeolites, also presented in Table 2. The largest specific surface area and micropore volume were exhibited by the NH4NaY zeolite-containing sample. The textural characteristics of the Al2O3–ZSM-5(5%) sample were most similar to those of conventional mesoporous γ-Al2O3, which finding can be explained by a low content of the additive and limitations of the t-plot method [36]. Overall, no significant differences in the textural properties between the supports synthesized were revealed.

Fig. 1.
figure 1

Nitrogen adsorption–desorption isotherms at 77 K for the supports: (a) Al2O3–NH4NaY(40%), (b) Al2O3–Beta(40%), (c) Al2O3–ZSM-5(40%), and (d) Al2O3–ZSM-5(5%).

Full size image
Table 2. Textural characteristics of the catalyst supports and zeolite additives
Full size table

Figure 2 demonstrates the NH3–TPD curves for the supports tested. All the curves exhibited three distinct segments with maxima near 220, 450, and 700°C, corresponding to acid sites of different strengths. Quantitatively evaluation of the acidic properties of the samples synthesized (see Table 3) revealed close total acidities for all the supports, except for the sample added with the smallest zeolite amount, as reasonably expected. The samples containing zeolite ZSM-5 were characterized by a large number of medium-temperature acidic sites.

Fig. 2.
figure 2

NH3–TPD curves for the supports: (1) Al2O3–NH4NaY(40%), (2) Al2O3–Beta(40%), (3) Al2O3–ZSM-5(40%), and (4) Al2O3–ZSM-5(5%).

Full size image
Table 3. Acidic properties of the supports synthesized
Full size table

High cracking activity exhibited by the Beta- and ZSM-5 zeolite-containing catalysts may be attributed to high acidity of the zeolites and to small size of micropores, which caused diffusion limitations, thereby extending the contact time between the substrate molecules and the acidic active sites within the crystals [10, 37, 38]. A decrease in the amount of the zeolite added did not cause major increase in selectivity toward isomerization reaction, since the number of the medium-temperature acidic active sites enhancing the cracking reactions decreased insignificantly. The decrease in the total acidity of the support added with a smaller zeolite amount was not proportional to the amount of the additive. This may be due to a more efficient distribution of a small amount of the zeolite additive, affording additional acid sites via zeolite-binder interaction [39, 40].

Relatively high selectivity toward isomerization reaction at moderate cracking activity achieved with the NH4NaY zeolite-containing catalyst (Table 4) may be due to its moderate acidity represented by equal amounts of Lewis and Brønsted acid sites, which favors the hydroisomerization reaction [41, 42]. Linear first-order reaction equation fits well our experimental results (Fig. 3). The observed rate constant of the isomerization reaction on the NH4NaY zeolite-based catalyst was estimated at 4.74×105 mol g–1 h–1 at 360°C, with 30% selectivity of the reaction toward isomers.

Fig. 3.
figure 3

Linear relationship between the n-hexadecane concentration and the conditional contact time for (a) hydrocracking and (b) hydroisomerization reaction.

Full size image
Table 4. Results of testing the NH4NaY zeolite-containing catalyst for hydroconversion of the feedstock in the presence of nitrogen-containing compoundsa
Full size table

With the view to improving the selectivity of n-hexadecane hydroisomerization in the presence of the NH4NaY-containing catalyst synthesized, we evaluated how it was influenced by the presence of nitrogen-containing compounds in the feedstock. Quinoline in amounts providing 50 and 150 ppm nitrogen concentration in the feedstock served as a nitrogen source. It is worth mentioning that quinoline was not detected among the hydrogenation products. Adding 50 ppm nitrogen to the feedstock caused reduction of the rate of the cracking reactions by nearly an order of magnitude, from 1.12×104 to 1.57×105 mol g–1 h–1 at 360°C (Fig. 4). This was due to predominant adsorption of quinoline on strong acid sites under the experimental conditions, which promoted the cracking reaction [43]. The resultant deceleration of the hydroisomerization reaction led to an almost threefold (from 23 to 61%) increase in selectivity toward isomerization. The beneficial effect produced by the inhibition is attributable to the fact that suppression of a part of the acid sites led to a balance between the acid sites and the hydrogenation sites that is optimal for the hydroisomerization reaction [44]. An increase in the nitrogen content in the feedstock to 150 ppm significantly inhibited the isomerization reaction as well, because nitrogen-containing bases were also adsorbed on weaker acid sites.

Fig. 4.
figure 4

Selectivity toward hydroisomerization reaction and the rates of hydrocracking and hydroisomerization reactions as functions of the quinoline amount added to the feedstock.

Full size image

CONCLUSIONS

Bifunctional zeolite-containing transition metal sulfide-based catalysts comprising different zeolite components in varied amounts were synthesized. The resultant materials were examined by low-temperature nitrogen adsorption and NH3–TPD method, and their catalytic activity for n-hexadecane hydroconversion reactions was tested.

In hexadecane hydroconversion on Beta and ZSM-5 zeolite-containing transition metal sulfide-based catalysts the cracking reactions dominated because of a high cracking activity due to a large number of strong acid sites and small zeolite micropores. Reduction in the content of the zeolite component to 5 wt % did not significantly affect the selectivity of the chemical transformations of n-hexadecane over the temperature range examined. The use of the bifunctional catalytic compositions based on Beta and ZSM-5 zeolites may be recommended for yielding incremental light components from heavy feedstock processing.

The NH4NaY zeolite-containing catalyst afforded relatively high (up to 30%) selectivity toward hydroisomerization reaction at moderate cracking activity due to large micropore sizes and moderate acidity of the additive. After further optimization the bifunctional composite catalysts based on NH4NaY type zeolites and transition metal sulfides will be useful in hydroisomerization and mild hydrocracking of middle distillate fractions aimed at improving their low-temperature properties.

By controlled addition of nitrogen-containing bases to the feedstock the selectivity toward hydroisomerization/hydrocracking reactions can be tuned to some extent via suppressing acid sites of the catalyst. This approach proved effective in our experiments with the sulfide-based bifunctional catalysts. Adding small (50 ppm) amounts of quinoline nitrogen allowed increasing the selectivity toward hydroisomerization reaction from 30 to 60% at a slight decrease of the observed reaction rate.