Hydrodesulphurization Research Articles

Heterogeneously catalyzed coal hydroliquefaction: Screening of catalysts and characterization

This project is centred upon the study of the activities of different catalysts in one-step coal liquefaction processes. A series of alumina supported catalysts was prepared by multistep impregnation, including a conventional CoMo/Al2O3 and other preparations containing Zn as a second promoter, and the alumina was acidified with different fluorine contents. These catalysts were extensive physicochemically characterized and their hydrodesulphurization (HDS) and hydrogenation (HYD) activities were tested using a model compound (thiophene). The results obtained indicate that partial substitution of Co by Zn does not affect the HDS and HYD activities and that fluorination diminishes these activities because of textural changes of the carrier. The prepared catalysts were tested in coal liquefaction and their activities were compared to those of cheap iron containing dispersion catalysts such as red mud, Fe2O2 aerosol, and Cottrell and pyrite ashes (by-products of the sulphuric acid industry). Most of the liquefaction tests were carried out in a magnetically stirred laboratory autoclave for 1 h, at 400 rpm, in the 673 K (400°C) − 748 K (475°C) temperature range, at 17 ± 1 MPa operating pressure, with two different solvents; tetralin and anthracene oil, and using mainly 2/1 and 3/1 solvent/coal ratios. The liquefaction products were fractionated into oils, asphaltenes and preasphaltenes, using pentane, toluene and tetrahydrofurane (THF) as extractive solvents. Prior to the study and comparison of the catalyst activities and exploration of the influence of various operating conditions (temperature, tetralin/coal ratio, type of solvent and operating pressure) on catalytic and non-catalytic coal liquefaction was performed. The coal used in most of the experiments was a Spanish subbituminous A coal. Additionally in order to contrast results a comparison of all the catalysts was carried out with a standard high volatile (HV) bituminous coal (140 FR 36) from the European Centre for Coal Specimens Stichting Steenkoolbank Nederland (SBN). Supported catalysts present higher activities than iron-based catalysts. Among these catalysts, red mud proved to be the most active. Catalytic experiments using anthracene oil as solvent and CoZnMo/fluorinated alumina catalysts present maximum yields indicating that Zn as second promoter and carrier acidification to be beneficial in coal liquefaction with solvents similar to those used in real plants.

A unified theory of periodic and promotion effects in transition metal sulphide hydrodesulphurization catalysts

The dependence of the hydrodesulphurization (HDS) activity of monometallic transition metal sulphides for thiophenic species is an electronic effect: the quality of the active site determines the rate of the rate-limiting step in the HDS process. It is shown that the strength of the interaction between the metal d electrons and sulphur 3p electrons,I, correlates well with catalytic activity.I depends strongly on the oxidation state of the transition metal. The synergistic promotion effect in Co-Mo-S and Ni-Mo-S is explained in terms of electron transfer from Co(Ni) to Mo. This corresponds to the removal of σ metal-sulphur antibonding electrons from Co(Ni), greatly increasing their intrinsic catalytic activity. It follows that Co and Ni are the active elements in these mixed sulphides.

Hydrodesulphurization of thiophene by transition metal sulphides. A molecular orbital topology study

Transition metal sulphides have the ability to catalyze the hydrodesulphurization (HDS) process of thiophene and related compounds in the presence of hydrogen. We present the results of scattered-wave calculations on model catalyst clusters and catalyst—thiophene complexes, which indicate that pπ bonding, between the sulphur atoms in the catalyst and the sulphur and carbon atoms in the thiophene ring, is responsible for binding the thiophene molecule to the catalyst in the initial stages of the HDS process.

Mo-USY zeolites for hydrodesulphurization: II. Surface properties of sulphided catalysts and activity for thiophene hydrodesulphurization

Molybdena-containing ultrastable (USHY) zeolite catalysts were prepared using different molybdenum precursors and activation procedures and the sulphided catalysts were characterised by nitric oxide and pyridine adsorption. X-ray photoelectron spectroscopy and tested in the hydrodesulphurization (HDS) of thiophene at 3 MPa pressure and a reaction temperature of 623 K. MoUSY catalyst exhibited decreasing conversions as a function of time-on-stream due to coking while thiophene conversion over MoNaY remained constant. No relationship between dispersion and HDS or hydrogenation selectivity was found. Differences between hydrogenation selectivities over MoUSY and MoNaY catalysts were attributed to the low concentration of support bound molybdenum species in the oxide form of the latter catalyst.

Titania-alumina mixed oxides as supports for molybdenum hydrotreating catalysts

A series of coprecipitated TiO2-Al2O3 supports containing between 0 and 100% TiO2 were prepared and characterized by various methods. These solids were impregnated with molybdenum and tested in their sulphide form as catalysts in the hydrodesulphurization (HDS) of thiophene. The results showed that in the prepared mixed oxides supports, an interaction between TiO2 and Al2O3 exists and that the coprecipitated solid is different, in its surface properties, from a mechanical mixture of the two oxides. Acidity measurements show that Lewis and not Brønsted sites exist in the mixed oxides. The number of acid sites increases sharply at high TiO2 content. High-resolution electron microscopy shows that a better dispersion of MoS2 on the supports is obtained when the TiO2 content is increased. Activity measurements in the thiophene hydrodesulphurization reaction show that the intrinsic activity, expressed on a molybdenum atom basis, increases sharply at high TiO2 loadings in the mixed oxides supports. Comparison of the activity and acidity measurements point to a correlation between HDS activity and the number of acid sites on the support that affects the dispersion during the impregnation step and the reactivity of the MoS2 sites.

Cold reduced tubes from nickel-molybdenum stainless powder alloys : Y.V. Manegin et al. (Central Scientific Research Institute for Ferrous metallurgy, Russia.) Stal'. 1991, 76–79. (In Russian.)

Carbon-supported iron-molybdenum sulphide catalysts were characterized by means of Mössbauer spectroscopy at temperatures down to 4.2 K. Thiophene hydrodesulphurization (HDS) activity measurements were performed at 673 K in a flow microreactor operating at atmospheric pressure. The molybdenum content was 9.5 wt.-% whereas the iron content varied from 0.6 to 9.0 wt.-%. Sequential deposition (Molybdenum first) by pore-volume impregnation was employed to prepare oxidic catalyst precursors. The oxidic catalyst precursors were dried at 293 K in an air flow, followed by an additional hydrogen treatment up to 393 K. The type and relative particle sizes of the iron compounds present in the oxidic precursors and in the sulphided and reoxidized catalysts were determined by Mössbauer spectroscopy. It was demonstrated that after sulphidation for 4 h at 623 K, the composition of the sulphide catalyst depends on the iron content. Sulphided Fe-Mo/C catalysts contain a mixed “FeMoS” phase and “Fe-sulphide”. The former is responsible for the observed promoting effect toward thiophene HDS. From the temperature dependence of the resonant absorption areas, it was concluded that the iron atoms in the “FeMoS” phase are located at the surface of MoS2 microcrystals. The amount of “Fe-sulphide” present in the catalyst was found to increase with increasing iron content. This “Fe-sulphide” might partly cover the “FeMoS” phase, thus causing a decrease in the promoting effect.

Phosphorus promotion in nickel-molybdenum/alumina catalysts: model compound reactions and gas oil hydroprocessing

Three different preparation methods for Ni-Mo/Al 2O 3 catalysts containing between 0 and 7.0 wt.-% phosphorus are compared in thiophene hydrodesulphurization (HDS) and diisopropylbenzene (DIPB) hydrocracking reactions. A promotional effect of phosphorus is seen in thiophene HDS regardless of preparation method, with maximum activity at around 1 wt.-% phosphorus. P/Al 2O 3 catalysts (P = 0–7 wt.-% ) have little activity in DIPB hydrocracking indicating that phosphorus has very little effect on the Brønsted acidity of alumina. An increase in DIPB hydrocracking is observed for P-Ni-Mo catalysts with low phosphorus loadings, however, and so phosphorus does have an indirect effect on the Brønsted acidity of these catalysts. Impregnation of phosphorus prior to the metal species leads to the most active catalysts for both thiophene HDS and DIPB hydrocracking and this is explained in terms of a dual promotional effect of phosphorus; phosphorus increases both the dispersion and reducibility/sulphidibility of molybdate and increases the amount of nickel available to form the Ni-Mo-S active HDS phase by impeding nickel aluminate formation. One series of catalysts was investigated in the hydroprocessing of a Syncrude combined gas oil-fuel oil, and the same gas oil spiked with 5 vol.-% quinoline. The promotional effect of phosphorus in HDS was in good agreement with the model compound studies: 1 wt.-% phosphorus was the most active HDS catalyst in all cases. Furthermore, phosphorus loadings of between 1 and 3 wt.-% gave a product with lowest nitrogen content, and aromaticity. These loadings also gave the greatest sulphur removal and maintained good hydrogenation in the presence of quinoline, and gave the best quinoline hydrodenitrogenation, suggesting optimum surface acidity and hydrogenation properties. Optimum loadings of phosphorus (around 3 wt.-%) also improve the longevity of the catalyst by maintaining high surface area during the hydroprocessing reaction.

Carbon supported Ni-Mo catalyst: high hydrodenitrogenation activity and low inhibition of hydrodesulphurization by hydrodenitrogenation

Four active carbon supported Ni-Mo sulphide catalysts and alumina supported Ni-Mo catalyst Shell 324 were tested in concurrent hydrodesulphurization (HDS) of thiophene and hydrodenitrogenation (HDN) of pyridine at 2 MPa and 280 °C. The various Ni-Mo/C catalysts differed in the pretreatment of the support which was extracted by various solvents before impregnation. The data were also compared with previous results on the same catalysts in HDS of thiophene in the absence of pyridine. The HDN activity of the Ni-Mo/C catalysts was three to five times higher than that of the reference Shell catalyst. The inhibition of HDS by HDN was much lower on the Ni-Mo/C than on the Ni-Mo/alumina samples. Consequently, HDS activity of the Ni-Mo/C catalysts in the presence of pyridine was very high, the best catalyst being 14 times more active than the alumina supported sample.

Ruthenium molybdenum sulphide catalysts : physicochemical characterization and catalytic properties in HDS, hydrogenation and HDN reactions.

Alumina supported RuMo sulphide catalysts having different weights of metal content and atomic composition ratios r=Ru/(Ru+Mo) were prepared by using ammonium heptamolybdate dissolved in water or a 10% (NH 4) 2S-H 2O solution and RuCl 3.3H 2O as precursor compounds. Their activities were studied in the hydrodesulphurization (HDS) of thiophene and in the hydrogenation (HYD) of biphenyl, and optimized in terms of the preparation method and the sulphidation process. Some hydrodenitrogenation (HDN) tests were also performed on these catalysts. Electron microscopy and XPS measurements were performed, and the nature of the active phase was discussed.

Catalytic properties of nickel molybdenum sulphides supported on nickel and magnesium aluminates

The hydrodesulphurization (HDS) of thiophene at atmospheric pressure, hydrodenitrogenation (HDN) of 2,6 diethylaniline (DEA) and of 1,2,3,4 tetrahydroquinoline (THQ) at 70 atm in a dynamic flow microreactor - these molecules being present alone or mixed in the feed - and finally HDN of mixtures of quinoline (Q) at 70 atm or of phenanthridine (Ph) at 140 atm with DEA in a batch reactor were the test reactions used to compare the performances of (Ni)MoS catalysts supported on MgAl 2O 4 and NiAl 2O 4 with a classical NiMoS/γ Al 2O 3 commercial catalyst. Some of the catalysts supported on NiAl 2O 4 appeared to be superior to the commercial one, whereas those supported on MgAl 2O 4 were found to be less active. HDS of thiophene, as well as the amount of carbon monoxide adsorbed at 0°C by sulphided catalysts permitted the prediction of the order of HDN activity of the catalysts for the conversion of some nitrogen-containing molecules. A strong inhibiting effect of molecules like Q, THQ or Ph (or some of their products formed during the process) on the HDN of DEA has been found and attributed to the competitive adsorption of these molecules on the supported active phase.

Titania-modified hydrodesulfurization catalysts: II. Dispersion state and catalytic activity of molybdena supported on titania-alumina carrier

A series of molybdena catalysts supported on titania-modified alumina carrier was characterized by X-ray diffraction (XRD), analytical electron microscopy (AEM), X-ray photoelectron spectroscopy (XPS) and temperature-programmed reduction (TPR). The results show that the type of Mo6+ species depends upon the coverage of titania on alumina. The molybdenum atomic concentration was improved when most of the surface of alumina was occupied by titania. TPR profiles of molybdena/titania-alumina exhibit two peaks at 680-710 K and 1090-1100 K. The amount of hydrogen consumption in the TPR process increases with a rise in titania content, which suggests that titania promotes the reduction of molybdena to much lower valency. The XPS data is in agreement with the TPR results. The measurements of the activity of the catalyst in hydrodesulphurization (HDS) of thiophene and hydrogenation (HYD) of cyclohexene indicate that titania-modified alumina supports can improve the HDS primary activity of the catalyst.

Hydrotreatment reactions with a disposable catalyst

The residue from the West Virginia University coal cleaning process may contain pyrite in significantly higher concentrations than the raw coal. Accordingly, the residue is used as a catalyst for the hydrodesulphurization (HDS) of benzothiophene and the hydrodenitrogenation (HDN) of quinoline. Grinding improves the performance of the residue catalyst. The results are compared with those when the raw coal or a commercial hydrotreatment catalyst is used. The catalytic activity of the residue is ~15% that of the commercial catalyst for HDS, and ~56% for HDN. The activation energies of the HDN reaction catalysed by the raw coal and by the residue are relatively small, 23–30 kJ mol −1. This implies that external mass transfer to the surface of the catalyst is the rate-determining step.

Effect of metal-organic compounds on thiophene hydrodesulphurization over sulphided forms of fluoride-containing CoMo/Al 2O 3 catalysts

The effect of adding fluoride at various stages of preparation of CoMo/γ-Al 2O 3 catalysts was investigated with X-ray photoelectron spectroscopy (XPS) and with a pulse reactor method for activity in hydrodesulphurization (HDS) of thiophene and deactivation by nickel and vanadium naphthenate (NiN and VN). Activity is lowered by increasing fluoride content, except for initial activity at a fluoride-to-aluminium (F/Al) ratio of 0.1. It may be correlated with a decreased surface area and dispersion of the molybdenum sulphide phase. Deactivation in HDS of thiophene strongly increases with the fluoride concentration, probably due to an increased acidity induced by the fluoride anions. XPS data suggest that fluoride is highly dispersed except at the highest concentration, i.e. F/Al=0.2. Fluorination of the catalysts has an inhibiting effect on deactivation by NiN and VN and the highest initial and final activity is observed at F/Al=0.1. XPS intensity data suggest that the dispersion of deposited metals decreases with increasing fluoride concentration and that the deposition occurs to a higher extent at the exterior of the particles than the interior under the conditions applied. Both sulphidic and oxidic forms are detected. The order of adding fluoride and the salts of cobalt and molybdenum results in minor differences in activity, except when metal naphthenates are deposited.

Influence of phosphate on the structure of sulfided alumina supported cobalt-molybdenum catalysts

The influence of phosphate on the sulfiding rate of alumina-supported cobalt, molybdenum and cobalt-molybdenum catalysts has been studied by means of Temperature Programmed Sulfiding (TPS). Combination of the TPS results with recently published Temperature Programmed Reduction (TPR), Mössbauer spectrometry and nitrogen oxide adsorption measurements results in a detailed picture of the sulfidic catalyst. The sulfiding of cobalt ions in phosphorus-containing alumina-supported cobalt catalysts shifted to somewhat higher temperatures with increasing phosphorus content, most likely due to the presence of “Co-AlPO4” phases. In the temperature region from 1000-1270 K, part of the cobalt sulfide species reacted with AlPO4 to cobalt phosphides. The amount of cobalt which reduced to phosphides increased strongly with increasing phosphorus content and became about 70% of the total amount of cobalt present. The sulfiding pattern of phosphorus-containing Co-Mo/Al catalysts is only slightly influenced by increasing phosphorus content below 1000 K. All components including the Co-Mo-O-P phase, present in the oxidic precursor, are relatively easy to sulfide. After sulfiding at 673 K, the Co-Mo-S phase is most likely still to be associated with the AlPO4 species. Analogous to cobalt catalysts, the major part of the cobalt in Co-Mo-P/Al catalysts ultimately reacted to cobalt phosphides above about 1000 K. The sulfiding rate of the molybdenum ions is not changed by the presence of phosphates. MoS2 does not react with the phosphates to molybdenum phosphide probably because this reaction is thermodynamically restricted. Based on the surface structures, deduced from a combination of TPR and TPS data, the striking difference between the effect of phosphates on the hydrodenitrogenation (HDN) and on the hydrodesulphurisation (HDS) reactions is discussed.

Surface acidity in promoted molybdena and tungsta hydrodesulphurisation catalysts

Acidity of Ni- and Co-promoted molybdena and tungsta hydrodesulphurisation (HDS) catalysts have been determined by volumetric adsorption of ammonia. A direct correlation between the adsorption of ammonia, and hence the acidity, and the HDS activity could be observed for reduced catalysts A relation between ammonia adsorption and anion vacancy concentration is thus indicated, as the latter is related to HDS activity. Contrary to unpromoted molybdena there is a decrease in acidity on sulphidation of Ni-promoted molybdena catalyst. This decrease is attributed to the influence of, Ni in suppressing the formation of anion vacancies by catalysing the crystallisation of MoS2. Promoted tungsta catalysts were found to have lower acidities as compared to promoted molybdena catalysts, which is consistent with their lower desulphurisation activity.

Influence of petroleum nitrogen compounds on hydrodesulphurization

A study of the influence of nitrogen compounds in a crude oil on hydrodesulphurization (HDS) of a vacuum distillate and dibenzothiophene in the presence of an industrial hydroraffination catalyst, G-3 ( CoMo Al 2O 3 ) was performed. The nitrogen compounds were separated from the vacuum distillate by ion-exchange and coordination chromatography in the form of acidic, basic and neutral concentrates. It was found that the addition of the concentrate (0.2% N) led to a decrease in degree of hydrodesulphurization of the vacuum distillate and dibenzothiophene. The strongest influence was found for acidic compounds, while the weakest was found for basic ones. The inhibition caused by the nitrogen compounds added is not permanent, and disappears after introduction of a raw material in the absence of nitrogen compounds.

Supported co-mo thin film sulphide catalysts for hydrodesulphurization : II. Effects of catalyst surface structure on hydrodesulphurization activity

Thin film cobalt-molybdenum oxides, deposited with varying composition and structure on an alumina support, have been used as precursor oxides to form model hydrodesulphurization catalysts. The hydrodesulphurization (HDS) activities of the reduced and sulphided thin films have been measured in a closed cycle reactor, and the final surface composition and structure of the catalysts have been studied using X-ray photoelectron spectrometry (XPS) directly interfaced to the reactor. The presence of cobalt is found to increase the sulphidation of molybdenum oxide precursor. A thin film precursor oxide with an oxide structure of CoMoO 4 was found to produce the most HDS-active sulphide surface. The corresponding XPS spectra show that both molybdenum and cobalt species have structures which differ from those of MoS 2 and Co 9S 8.

Sulphidation of carbon-supported iron-molybdemum oxide catalysts

Carbon-supported iron-molybdenum sulphide catalysts were characterized by means of Mössbauer spectroscopy at temperatures down to 4.2 K. Thiophene hydrodesulphurization (HDS) activity measurements were performed at 673 K in a flow microreactor operating at atmospheric pressure. The molybdenum content was 9.5 wt.-% whereas the iron content varied from 0.6 to 9.0 wt.-%. Sequential deposition (Molybdenum first) by pore-volume impregnation was employed to prepare oxidic catalyst precursors. The oxidic catalyst precursors were dried at 293 K in an air flow, followed by an additional hydrogen treatment up to 393 K. The type and relative particle sizes of the iron compounds present in the oxidic precursors and in the sulphided and reoxidized catalysts were determined by Mössbauer spectroscopy. It was demonstrated that after sulphidation for 4 h at 623 K, the composition of the sulphide catalyst depends on the iron content. Sulphided Fe-Mo/C catalysts contain a mixed “Fe Mo S” phase and “Fe-sulphide”. The former is responsible for the observed promoting effect toward thiophene HDS. From the temperature dependence of the resonant absorption areas, it was concluded that the iron atoms in the “Fe Mo S” phase are located at the surface of MoS 2 microcrystals. The amount of “Fe-sulphide” present in the catalyst was found to increase with increasing iron content. This “Fe-sulphide” might partly cover the “Fe Mo S” phase, thus causing a decrease in the promoting effect.

Study of low-percentage alumina-supported nickel-molybdenum catalysts by ESR spectroscopy and magnetic measurements

Alumina-supported nickel-molybdenum catalysts (0.7–1.5 wt.-% nickel and 7–15 wt.-% molybdenum) were studied by ESR spectroscopy and magnetic methods. Thiophene hydrodesulphurization (HDS) was used to examine their activity. The effect of sample preparation on the state of nickel and molybdenum under reducing and sulphiding conditions was also investigated. The existence of a definite bond between nickel and molybdenum in the samples is responsible for a higher catalytic activity in HDS owing to the presence of nickel heteropolymolybdate.

Temperature-programmed reduction of Ni-Mo oxides

Temperature-programmed reduction (TPR) has been employed to study Ni-Mo mixed oxides which were previously used as model hydrodesulphurization (HDS) catalysts, using compositions ranging from pure MoO3 to pure NiO. An assignment of TPR signals to the different bulk phases was attempted. Good agreement between TPR spectra and structural data obtained previously from X-ray and electron diffraction was observed. TPR traces were consistent with proposed mechanisms of reduction of the bulk oxides MoO3, MoO2, NiO and NiMoO4. The variations in TPR spectra were interpreted in terms of effects such as crystallite size, ageing of the samples, hydrous state and chemical interactions between the different species. The significance of these reducibility results for HDS catalysis is discussed.