ISSN 00231584, Kinetics and Catalysis, 2014, Vol. 55, No. 4, pp. 481–491. © Pleiades Publishing, Ltd., 2014. Original Russian Text © E.N. Vlasova, I.V. Deliy, A.L. Nuzhdin, P.V. Aleksandrov, E.Yu. Gerasimov, G.I. Aleshina, G.A. Bukhtiyarova, 2014, published in Kinetika i Kataliz, 2014, Vol. 55, No. 4, pp. 506–516.
Catalytic Properties of CoMo/Al2O3 Sulfide Catalysts in the Hydrorefining of StraightRun Diesel Fraction Mixed with Rapeseed Oil E. N. Vlasovaa, c, *, I. V. Deliya, b, c, A. L. Nuzhdina, P. V. Aleksandrova, E. Yu. Gerasimova, c, G. I. Aleshinaa, and G. A. Bukhtiyarovaa a
Boreskov Institute of Catalysis, Siberian Branch, Russian Academy of Sciences, Novosibirsk, 630090 Russia b Novosibirsk State University, Novosibirsk, 630090 Russia c Research and Educational Center for Energy Efficient Catalysis at Novosibirsk National Research University Novosibirsk, 630090 Russia *email:
[email protected] Received December 23, 2013
Abstract—CoMo/Al2O3 sulfide catalysts varying in preparation method and Co/Mo ratio have been tested in the hydrorefining of a mixture of straightrun diesel fraction and rapeseed oil in a flow reactor at a temper ature of 340–360°C, a hydrogen pressure of 4.0–7.0 MPa, and a liquid hourly space velocity of 1–2 h–1. A comparison between catalysts prepared using citric acid (CoMo/Al2O31.5) and both citric and orthophos phoric acids (CoMoP/Al2O31.5) as promoters, with Co/Mo = 0.3 and 0.5, has demonstrated that the most active catalyst in hydrodesulfurization and hydrodenitrogenation is the phosphoruscontaining Co/Mo ≈ 0.5 sample. The addition of rapeseed oil to straightrun diesel fraction lowers the hydrodesulfurization and hydrodenitrogenation activities of the CoMo sulfide catalysts, irrespective of the method by which they were prepared. The fatty acid triglyceride conversion selectivity of these catalysts depends on the Co/Mo ratio and on reaction conditions: decreasing the Co/Mo ratio from 0.46 to 0.26, lowering the reaction temperature, and raising the hydrogen pressure and hydrogentofeedstock ratio increase the C18/C17 hydrocarbon ratio in the hydrogenated product. The addition of rapeseed oil improves the quality of the product; however, for attain ing the preset residual sulfur level in this case, the process needs to be conducted at a higher temperature than the hydrorefining of straightrun diesel fraction containing no admixture. DOI: 10.1134/S0023158414040144
INTRODUCTION The steadily increasing demand for environmen tally friendly motor fuels in the context of limited glo bal petroleum resources stimulates interest in develop ment and implementation of new catalytic technolo gies for processing renewable raw materials, including inedible vegetable oils, pyrolysis products of various kinds of biomass, etc. [1–9]. Diesel fuel production from renewable raw materials is presently based on processing of fatty acid triglycerides, which are the major component of vegetable oils and fats. The most widespread technology in this area is triglyceride transesterification with alcohols into methyl or ethyl esters of fatty acids, which received the name of first generation biofuels [1–6]. However, this product suf fers from a number of drawbacks and is mainly used as an additive in conventional diesel fuel [5, 10]. An alternative technology that has become popular in the last decade is hydrotreating of fatty acid triglyc erides to obtain an alkane mixture, which is referred to as secondgeneration biofuel, bio gas oil, or green die sel [2–8]. As distinct from the firstgeneration biofu
els, this product does not contain oxygen, has a large cetane index, is more stable, and is miscible with con ventional diesel fuel in any proportion, so it can be transported and stored using the existing infrastruc tures [5, 6]. Triglycerides can be industrially hydrotreated both in specialpurpose, newly designed units using only triglycerides as the feedstock and, when mixed with a dieselfuel petroleum distillate, in existing hydrorefin ing units [2–8]. In both cases, the process is con ducted over conventional Co(Ni)Mo/Al2O3 sulfide catalysts. Triglyceride hydrodeoxygenation occurs either via the removal of oxygen as water, with the number of carbon atoms in the hydrocarbon molecule remaining unchanged, or via the elimination of CO/CO2 molecules, yielding alkanes with a shorter chain than in the triglyceride’s acid [7, 9, 10]. One of the most important aspects of the simulta neous processing of vegetable oils and petroleum dis tillates is the effect of oxygencontaining compounds on the activity of catalysts in the hydrodesulfurization and hydrodenitrogenation of components of petro
481
482
VLASOVA et al.
leum fractions. It was demonstrated in a number of works that the addition of up to 10 wt % jatropha seed oil [11], rapeseed oil [12], or a spent edible oil [13, 14] does not decrease the activity of the NiMo/Al2O3 cat alyst in the hydrodesulfurization of dieselfuel petro leum distillates. According to the literature [15], the addition of sunflower oil to vacuum gas oil does not deactivate this catalyst either. In the hydrorefining of straightrun diesel fraction containing 5, 15, or 25 wt % vegetable oil, a decrease in the hydrodesulfurization activity of the NiMo/Al2O3 catalyst was observed at oil concentrations above 15 wt % [16]. The results obtained for different vegetable oils are in agreement: apparently, triglycerides and the prod ucts of their conversion exert no inhibiting effect on the hydrodesulfurization activity of the NiMo sulfide catalysts. The effect of triglyceride admixtures on the activity of СoMo/Al2O3 catalysts widely used in the hydrorefining of petroleum fractions has been investi gated to a much lesser extent. A study of the effect of palm oil on the activity of the commercial СoMo/Al2O3 catalyst in the hydrorefining of straight run diesel fraction demonstrated that the desulfuriza tion rate decreases as the palm oil concentration is increased from 0 to 5 wt % and remains unchanged at higher oil concentrations [17, 18]. There has been a comparative study of the quality of diesel fuels derived from straightrun diesel fraction and from the same fraction mixed with cottonseed oil [19], but no data concerning the effect of the oil on the hydrodesulfur ization activity of catalysts have been reported. In addition, the works cited above pay little attention to the effect of triglyceride additions on the hydrodeni trogenation activity of sulfide catalysts and do not consider how the properties of sulfide catalysts depend on the way they were prepared and on the composition of their active component (including the Co(Ni)/Mo ratio). According to presentday views, the active compo nent of hydrorefining catalysts is finely dispersed МоS2 particles with cobalt or nickel atoms occurring on their lateral faces [20]. The hydrodesulfurization activity of Co(Ni)Mo sulfide catalysts is enhanced in various ways. In particular, solutions containing orthophosphoric acid or chelating agents are employed in their preparation [21–23] and molybde num heteropoly compounds are used as the precursor [24–26]. More active CoMo catalysts for the hydrodesulfurization of diesel fractions can be pre pared by using solutions containing both orthophos phoric and citric acids [27, 28]. Here, we report the effect of rapeseed oil added to diesel fraction on the hydrodesulfurization and hydrodenitrogenation properties of CoMo/Al2O3 cat alysts differing in the way they were prepared an in Co/Mo ratio. The catalysts were prepared using solu
tions containing citric acid or both citric and ortho phosphoric acids along with molybdenum and cobalt compounds. EXPERIMENTAL Catalyst Preparation The support was γAl2O3 granules with a trefoil shaped cross section (Promyshlennye Katalizatory Co., Russia), whose size was 1.5 mm (Al2O31.5) or 1.2 mm (Al2O31.2). Catalysts were prepared by impregnating support granules with solutions contain ing various active component precursors. One of the solutions was prepared by dissolving appropriate amounts of ammonium paramolybdate (NH4)6Mo7O24 ⋅ 4H2O (analytical grade), cobalt ace tate Co(CH3COO)2 ⋅ 4H2O (pure grade), and citric acid C6H8О7 ⋅ H2O (reagent grade) (Vekton, Russia) in distilled water under continuous stirring [29]. In the preparation of the second solution, the starting com pounds were molybdenum oxide МоО3 (pure grade), orthophosphoric acid H3PO4 (specialpurity grade), citric acid C6H8О7 ⋅ Н2O (reagent grade) (Vekton, Rus sia), and cobalt hydroxide Co(OH)2 (95%, Aldrich, United States) [27]. The molybdenum concentration in the impregnating solution was adjusted so that the catalysts calcined at 550°C contained ~13 wt % Mo. The impregnated samples were dried in flowing nitro gen at 110°С for 8 h and at 200°C for 6 h. Physicochemical Properties of Catalysts The textural characteristics of the supports were determined on an ASAP2400 surface area and pore size analyzer (Micromeritics, United States) from low temperature (–196°C) nitrogen adsorption isotherms. Prior to being examined, the samples were kept in a vacuum at 150°C. The error in this method was ±10%. The catalysts were analyzed for molybdenum, cobalt, and phosphorus by inductively coupled plasma atomic absorption spectroscopy on an Optima 4300 DV atomic emission spectrometer (Perkin Elmer, United States). Xray diffraction patterns were obtained on an ARL X’TRA diffractometer (Thermo Fisher Scien tific, Switzerland) using CuKα (λ = 1.5418 Å) radiation (scanning step, 2θ = 0.05°; counting time per data point, 3 s). Highresolution transmission electron microscopy (HRTEM) studies were carried out on a JEM2010 microscope (JEOL, Japan) at an accelerating voltage of 200 kV and a resolution of 0.14 nm. Catalyst parti cles were deposited onto a copper grid from their eth anolic suspension pretreated in an ultrasonic disperser. The average size of the CoMoS nanoparticles and the average number of layers per particle were determined by examining at least 500 particles. KINETICS AND CATALYSIS
Vol. 55
No. 4
2014
CATALYTIC PROPERTIES OF CoMo/Al2O3 SULFIDE CATALYSTS
483
Table 1. Characteristics of the feedstock Heteroatom concentration
Density, g/cm3
Composition Diesel fraction Diesel fraction + 10 wt % rapeseed oil
0.852 0.858
Aromatic compounds*, wt %
S, μg/g
N, µg/g
O, wt %
mono
di
tri+
10330 9270
146 142
0.068 1.202
21.5 19.2
9.0 8.0
1.4 5.0
* Mono, bi, and polycyclic (tri+) aromatic hydrocarbons.
Catalytic Properties Sulfiding of the catalysts and determination of their catalytic activity in the hydrodesulfurization of diesel fraction were carried out in a flow reactor with an inner diameter of 16 mm and a length of 570 mm. The liquid feedstock and hydrogen were fed into the reac tor at its top. Diesel fraction was supplied using a Gil son305 liquid chromatography pump; hydrogen, using a Bronkhorst automated dosing system. The reactor was charged with catalyst granules (granule length of 4–5 mm, total volume of 10 mL) diluted with small silicon carbide particles (0.1–0.25 mm size frac tion) in a 1 : 2 ratio. The catalyst was sulfided with straightrun diesel fraction additionally containing 0.6 wt % sulfur as dimethyl disulfide (DMDS). The feedstock hourly space velocity was 2 h–1, the hydrogen tofeedstock ratio (ratio of the Н2 volume in m3 under standard conditions to the feedstock volume in m3) was 300, and the hydrogen pressure was 3.5 MPa. Sulfiding was performed in two steps, the first one at 240°C for 8 h and the second at 340°C for 6 h, raising the tem perature between the steps at a rate of 25°C/h. The feedstock was straightrun diesel fraction or the same fraction mixed with rapeseed. The properties of the feedstock are listed in Table 1. Rapeseed oil was cho sen for the reason that its dominant components are carboxylic acids containing 18 carbon atoms per mol ecule (oleic, linoleic, linolenic, and stearic acids), making it easier to carry out a reaction route anal ysis [7, 9]. According to the testing protocol, the first step was determination of the activity of the catalysts in the hydrodesulfurization of straightrun diesel fraction at 340°C, a hydrogen pressure of 3.5 MPa, a hydrogen tofeedstock volume ratio of 300, and a feedstock hourly space velocity of 2 h–1. Next, we studied the activity of the catalysts in the hydrorefining of diesel fraction mixed with rapeseed oil, varying the hydro gentofeedstock ratio, reaction temperature, hydro gen pressure, and feedstock hourly space velocity in the preset order. In the final step, we again estimated the activity of the catalysts in the hydrodesulfurization of diesel fraction under the firststep conditions. The testing duration was 14 h in the first step and 12 h in the other steps. The first sample for analysis was taken 5 h after the change in the reaction conditions; there after, the reaction mixture was sampled at 1h inter KINETICS AND CATALYSIS
Vol. 55
No. 4
2014
vals. The residual sulfur and nitrogen contents, the amounts of aromatic compounds, and the density and fractional makeup of hydrogenated products were esti mated from analysis data for three samples taken 10, 11, and 12 h after the beginning of the current testing step. Analysis of the Feedstock and Hydrogenated Products The hydrogenated products were analyzed in order to determine their oxygen, sulfur, nitrogen, and monocyclic and polycyclic arene contents and their fractional makeup, density, and С18/C17 hydrocarbon ratio. The total sulfur content of the feedstock and that of the hydrogenated products were measured on a LabX 3500SCl energydispersive Xray fluorescence analyzer (Oxford Instruments, United Kingdom). Micro amounts of nitrogen and sulfur in the products were measured on an ANTEK 9000NS nitrogen/sul fur analyzer (Antek Instruments L.P., United States) using the ASTM D 5762 and ASTM D 5453 standard procedures. The fractional makeup of samples was determined by the simulated distillation method via the ASTM 2887 standard procedure on an Agilent 6890N chromatograph (Agilent Technologies, United States) with a DB1 capillary column (length of 10 m, inner diameter of 0.53 mm, film thickness of 2.65 μm) using the Chemstation SimDis software. Aromatic compounds in the resulting diesel fuel were quantified via procedure 391/395 approved by the Institute of Petroleum (European Standard EN 12916) on a Varian ProStar chromatograph (Varian B.V., United States) fitted with a refractometric detector. Table 2. Characteristics of the alumina supports Granule Support size*, mm Al2O31.5 Al2O31.2
1.5 1.2
SBET, m2/g 208 235
Average Pore volume, pore diame cm3/g ter, Å 0.68 0.79
132 134
* Distance between the base and the trefoil triangle in the cross section of the granule.
484
VLASOVA et al.
The density of hydrogenated products was deter mined using a Densito 30PX densimeter (Mettler Toledo, Switzerland) via the ASTM D 4052 procedure. Cetane index (CI) was calculated according to the ASTM D 4737 procedure using the formula CI = 45.2 + 0.0892(T10 – 215) + 0.131(T50 – 260) + 0.0523(T90 – 310) (1) + 0.901B(T50 – 260) – 0.420B(T90 – 310) + 4.9 × 10–4(T10 – 215)2– 4.9 × 10–4(T90 – 310)2 + 107B + 60B2, where T10, T50, and T90 are the final boiling points for 10, 50, and 90% of the fraction (°C) and B = e–3.5(D – 0.85) – 1 (D is the density of the mixture at 15°C (g/cm3)). The total oxygen content of the reaction mixture before and after hydrorefining was determined using a Vario EL Cube CHNSO analyzer (Elementar Analy sensysteme GmbH, Germany). The measure of catalytic activity in hydrodesulfur ization was the reaction rate constant k calculated via a firstorder rate equation: k(n − 1) (2) = 1n−1 − 1n−1 , W [S] [S0 ] where W is the feedstock hourly space velocity ((m3 feed) (m3 catalyst)–1 h–1), [S] and [S0] is the sulfur content of the hydrogenated products and feedstock (wt %), respectively, and n = 1.6. RESULTS AND DISCUSSION Physicochemical Properties The textural characteristics of the alumina supports are presented in Table 2. The list of catalysts and their chemical compositions are given in Table 3. The cata lysts differ in their chemical composition and Co/Mo ratio. Their Mo content is 12.8–13.5 wt %. The 0.5СoMo/Al2O31.5 catalyst was prepared by impreg nating the support with a solution containing ammo nium paramolybdate, cobalt acetate, and citric acid. The 0.5СoMoP/Al2O31.5 and 0.3СoMoP/Al2O31.5 samples were prepared using МоО3, orthophosphoric acid, cobalt hydroxide, and citric acid. The catalysts supported on 1.5mm alumina granules were used to study the effect of rapeseed oil on their hydrodesulfu
rization and hydrodenitrogenation activities. Based on the results of testing the catalysts in the hydrorefining of the rapeseed oil + straightrun diesel fraction mix ture, we optimized the composition of the impregnat ing solution for preparing the 0.5СoMoP/Al2O31.2 catalyst. The latter was used to obtain a diesel fuel con taining
