Fuel oils and lube base oils were prepared by hydrocracking/isomerization of Fischer-Tropsch synthesized waxes and long-chain α-olefins with various carbon chain lengths. Molecular structures of the prepared oils were investigated with iso-/n-paraffin ratios, average carbon numbers, and average branching numbers, which were calculated from CH and CH3 carbon ratio derived from 13C-NMR analysis. Prepared base oils showed very high viscosity indexes of up to 159, but these varied widely with the severity of the hydrocracking/isomerization conditions and used feedstock. Average carbon number and average branching number independently showed good correlations with the viscosity properties of the base oil such as kinematic viscosity and viscosity index. Viscosity index increased with higher average carbon number or lower average branching number. The effect of average carbon number or average branching number on the viscosity index depended on the feedstock, so a new index (average carbon number)2 x (average branching number) -1 was introduced as a molecular structural parameter of paraffins, and index was confirmed to indicate the viscosity index regardless of the feedstock. A similar structural parameter (average carbon number)a x (average branching number) was applied to kinematic viscosity. Kinematic viscosities at 40°C and 100°C showed good correlations when (a, b) = (3.5, 0.9) and (3.0, 0.5), respectively. As for the gas oil fraction, the molecular structural formula, (average carbon number) x (n-paraffin ratio) A, showed a good linear correlation with the kinematic viscosity (30°C), the cetane index, and cold flow plugging point for A = 0.0, 0.02, and 0.05 respectively. The influence of hydrocracking/isomerization severity on the molecular structures of each fraction is also investigated and the transition of molecular structures during the procession of reaction was observed.

Model HDS catalysts (Co2.75Mo11, Ni2.75Mo11, and Ni0.9Co1.85Mo11 wt %/γ-Al2O3) prepared by impregnation of similar supports with similar starting materials were used in the HDS of dibenzothiophene (DBT) and 4,6-dimethyldibenzothiophene, with and without addition of H2S or NH3. In the case of the HDS of DBT and 4,6-DMDBT, H2S only weakly poisoned CoMo and NiCoMo catalyst, whereas a NiMo catalyst was poisoned more strongly. The presence of NH3 significantly decreased the overall reaction rate constant for the conversion of 4,6-DMDBT for all tested catalysts to a greater extent than that for the blank test. The reaction rate constant of CoMo catalyst significantly decreased with NH3 addition. NiCoMo was the most active catalyst for the conversion of DBT in the presence of H2S. The presence of NH3 significantly decreased the activity of CoMo. The inhibition effect of H2S and NH3 was more clearly shown in the conversion of 4,6-DMDBT than in the conversion of DBT. NiMo catalyst caused the highest level of HDS conversion of 4,6-DMDBT blank test. The CoMo catalyst was not as active as the other two catalysts except under H2S rich conditions. NiCoMo catalyst was the best catalyst for the first-stage reactor because NiCoMo had the high tolerance for both H2S and NH3. NiMo catalyst was the best catalyst for the second-stage reactor because NiMo had high HDS activity under lean H2S conditions.

SiO2- and Al2O3-supported MoS2 and WS2 catalysts were prepared to exploit the evaluation technique of the edge dispersion of MoS2 and WS2 particles. A chemical vapor deposition (CVD) technique using Co(CO)3NO as a probe molecule was used for the evaluation. Results were compared with those from conventional techniques such as NO adsorption and TEM. A proportional correlation was obtained between the amount of NO adsorption and the amount of Co atoms accommodated by the CVD technique on WS2/SiO2 and WS2/Al2O3 catalysts, demonstrating a selective location of the Co atoms on the edges of WS2 particles, as previously established for MoS2 catalysts. A comparison of the amounts of NO adsorption and Co accommodation on MoS2 and WS 2 catalysts suggested a 70% higher density of sulfur vacancy on MoS2 particles than on WS2 particles regardless of the support. The Co atoms on the edges of MoS2 and WS2 particles showed the identical NO adsorption property. We propose that Co(CO)3NO can be used as a probe molecule to evaluate and directly compare the edge dispersions of MoS2 and WS2 catalysts. The dispersion of MoS2 particles was about two times higher than that of WS2 particles with the SiO2-supported catalysts. With the Al2O3-supported catalysts, MoS2 and WS2 particles were dispersed to a similar extent but much more highly dispersed than the counterparts in the SiO2-supported catalysts. The evaluation of the edge dispersion of MoS2 and WS 2 particles by means of TEM may pose problems when SiO2- and Al2O3-supported catalysts are compared. The edges of unpromoted MoS2 particles exhibited a significantly higher intrinsic activity for the HDS of thiophene than those of WS2 particles. © 2005 Published by Elsevier B.V.

Lubricant base oils were prepared by hydrocracking/isomerization of Fischer-Tropsch synthesized waxes and long-chain α-olefins with various carbon chain lengths. Correlations between operation conditions, viscosity properties of base oil, and molecular structures were studied. Analysis of FD-MS and iodine number confirmed that no olefins and a small amount of cyclic paraffins containing 1 naphthene ring were present in the prepared base oil. At lower conversion under 40 wt %, the ratio of paraffins containing naphthene rings were very low, though at higher conversion ratio, the content was relatively higher. The viscosity index (VI) was drastically lowered with increased conversion with every feedstock. These changes could be attributed to structural changes of the isoparaffins, which form the base oils. VI of prepared base oils widely varied from 114 up to 159 with severity of hydrocracking/isomerization reaction and feedstock used. VI decreased steeply with increased severity of hydrocracking/isomerization reaction with all feedstocks. The relationship between conversion and VI depended on the feedstock. Average branching numbers increased and average carbon number decreased with increased severity of hydrocracking/isomerization reaction with all feedstocks.

The effect of boron addition was studied on the hydrodesulfurization (HDS) of thiophene over MoS2/B/Al2O3 and Co-MoS2/B/Al2O3 (CVD-Co/MoS2/B/Al2O3), which was prepared by a CVD technique using Co(CO)3NO as a precursor of Co. The catalysts were characterized by means of NO adsorption and TEM. The HDS activity of MoS2/B/Al2O3 catalysts kept constant up to a boron content of about 0.6 wt% and decreased with a further increase of boron content. With CVD-Co/MoS2/B/Al2O3 catalysts, the HDS activity significantly increased as the boron content increased up to about 0.6 wt% of boron, followed by a decrease with a further increase of boron loading. Despite the activity increase, the amount of NO adsorption on MoS2/B/Al2O3 steadily decreased with increasing boron loading, suggesting that the dispersion of MoS2 particles is decreased by the addition of boron. A selective formation of the CoMoS phase on CVD-Co/MoS2/B/Al2O3 was achieved by the CVD technique. The TOF of the HDS over the CVD-Co/MoS2 /B/Al2O3 catalysts, defined by the activity per Co atom forming the CoMoS phase, increased as high as 1.6 times by the addition of boron, indicating that the active phase of the catalysts shifts from less active CoMoS Type I to more active CoMoS Type II. A TEM analysis showed that the number of stacking of MoS2 slabs was only slightly increased by the addition of boron. It is concluded that the activity increase of the CVD-Co/MoS2/B/Al2O3 catalyst is caused by the formation of CoMoS Type II because of weakened interaction strength between the CoMoS phase and the Al2 O3 surface by the addition of boron. © 2004 Elsevier Inc. All rights reserved.

A newly developed catalyst system to produce ultra low sulfur diesel fuel < 50 ppm sulfur was described. The R&D concept to achieve ultra low sulfur diesel production and its application to commercial units were presented with regard to the following major operation topics, i.e., feed property, hydrogen consumption and product aromatic content, catalyst profile, and approach to produce "sulfur-free" diesel fuel. The feed distillation property strongly affected the HDS activity and suggested that distillation range of diesel fraction should be controlled from the standpoint of the refinery's flow balance and HDS performance of diesel hydrotreating units. The results were applied to the commercial units in Japan Energy Group's refineries, and ultra low sulfur diesel of < 50 ppm-sulfur was successfully produced with those units for > 1 yr.

A study on ultra low sulfur diesel fuel production by two-stage process with gas/liquid separation system was carried out using middle-east straight-run gas oil fraction as feedstock of the hydrotreating experiments. The two-stage process with gas/liquid separation achieved ultra-low sulfur diesel production (50 ppm or less) under more beneficial conditions. It also had great potential to sulfur-free diesel production (10 ppm or less). Removal of produced hydrogen sulfide and NH3 in the middle of the unit accelerated HDS of the following 2nd stage. The NiW catalyst could be applied to the 2nd stage.

Further tightening of diesel sulfur specifications has been proposed worldwide. The focus of the new specifications is reduction of suspended particualte matter and NOx emission from diesel-fueled vehicles. CoMo/NiMo catalyst relay system was able to achieve ultra low-sulfur diesel production without major revamp of covnentional deep HDS. The system was developed based on considering reaction conditions in detail for each part in a desulfurization unit, especially sulfur-containing compound types to be desulfurized and catalyst poisoning by produced H2S and NH3. Experimental results for the systems were elucidated.

Two solutions (CoMo/NiMo "Catalyst Relay" and two-stage process with gas/liquid separation) for ultra-low sulfur diesel production were presented. Both solutions were developed based on considering reaction conditions in detail for each part in a HDS unit, particularly sulfur-containing compound types to be desulfurized and catalyst poisoning by produced H2S and NH3. The "CoMo/NiMo Catalyst Relay" system achieved ultra-low-sulfur diesel production (S ≤ 50 ppm) without major revamp of conventional deep HDS unit. The first bed CoMo catalyst achieved HDS of reactive sulfur compounds. The second bed NiMo catalyst was the main catalyst for ultra-low sulfur diesel production. The two-stage process with gas/liquid separation achieved ultra-low sulfur diesel production (S ≤ 50 ppm) under more beneficial conditions and had great potential to sulfur-free diesel production (S ≤ 10 ppm). Removal of produced H2S and NH3 in the middle of the unit accelerated HDS of the following second-state. The combined use of CoMo catalyst (HOP-467) and NiMo catalyst (HOP-414) achieved 50 ppm-sulfur diesel production at 341°C, that is 14°C lower reaction temperature than HOP-463 does, and was superior to HOP-463 (conventional CoMo) to achieve 50 ppm-sulfur diesel.

A novel combination process, SUCCEED, for residual oil cracking was developed by bench scale testing, and it has been proven commercially feasible. SUCCEED is combination of slurry phase hydrocracking and delayed coking. In the first section, which is the main feature of the process, a tubular reactor with a small amount of disposable additive catalyst is applied. For the vacuum residue feedstock,the first section gives moderate conversion around 65 wt% and the second gives conversion around 55 wt% for the unconverted feed in the first section, thus resulting in the total conversion of 85 wt%. The functions of the catalyst, hydrogenation ablity and anti-coking effect on the reactor have been studied to determine the minimum requirement of catalyst consumption rates. In this study, characteristics, capabilities and effect of the catalyst are discussed. And, in particular, features performance and commercial feasibility of SUCCEED have been described.

Reactions of [Ni(tren)(H2O)2]X2 (tren = tris(2-aminoethyl)amine; X = Cl (1a), Br (1b); X2 = SO2 (1c)) with mannose-type aldoses, having a 2,3-cis configuration (D-mannose and L-rhamnose), afforded (bis(N-aldosyl-2-aminoethyl)(2-aminoethyl)amine}nickel(II) complexes, [Ni(N,N'-(aldosyl)2-tren)]X2 (aldosyl = D-mannosyl, X = Cl (2a), Br (2b), X2 = SO4 (2c); aldosyl = L-rhamnosyl, X2 = SO4 (3c)). The structure of 1c was confirmed by X-ray crystallography to be a mononuclear [NiIIN4O2] complex with the tren acting as a tetradentate ligand (1c·2H2O: orthorhombic, Pbca, a = 15.988(2) Å, b = 18.826(4) Å, c = 10.359(4) Å, V = 3118 Å3, Z = 8, R = 0.047, and Rw = 0.042). Complexes 2a,c and 3c were characterized by X-ray analyses to have a mononuclear octahedral Ni(II) structure ligated by a hexadentate N-glycoside ligand, bis(N-aldosyl-2-aminoethyl)(2-aminoethyl)-amine (2a·CH2OH: orthorhombic, P212121, a = 16.005(3) Å, b = 20.095(4) Å, c = 8.361(1) Å, V = 2689 Å3, Z = 4, R = 0.040, and Rw = 0.027. 2c·3CH2OH: orthorhombic, P212121, a = 14.93(2) Å, b = 21.823(8) Å, c = 9.746(2) Å, V = 3176 Å3, Z = 4, R = 0.075, and Rw = 0.080. 3c·3CH3OH: orthorhombic, P212121, a = 14.560(4) Å, b = 21.694(5) Å, c = 9.786(2) Å, V = 3091 Å3, Z = 4, R = 0.072, and Rw = 0.079). The sugar part of the complex involves novel intramolecular sugar-sugar hydrogen bondings around the metal center. The similar reaction with D-glucose, D-glucosamine, and D-galactosamine, having a 2,3-trans configuration, resulted in the formation of a mono(sugar) complex, [Ni(N-(aldosyl)-tren)(H2O)2]Cl2 (aldosyl = D-glucosyl (4b), 2-amino-2-deoxy-D-glucosyl (5a), and 2-amino-2-deoxy-D-galactosyl (5b)), instead of a bis(sugar) complex. The hydrogen bondings between the sugar moieties as observed in 2 and 3 should be responsible for the assembly of two sugar molecules on the metal center. Reactions of tris(N-aldosyl-2-aminoethyl)amine with nickel(II) salts gave the tris(sugar) complexes, [Ni(N,N',N''-(aldosyl)3-tren)]X2 (aldosyl = D-mannosyl, X = Cl (6a), Br (6b); L-rhamnosyl, X = Cl (7a), Br (7b); D-glucosyl, X = Cl (9); maltosyl, X = Br (10); and melibiosyl, X = Br (11)), which were assumed to have a shuttle-type C3 symmetrical structure with A helical configuration for D-type aldoses on the basis of circular dichroism and 13C NMR spectra. When tris(N-rhamnosyl)-tren was reacted with NiSO4·6H2O at low temperature, a labile neutral complex, [Ni(N,N',N''-(L-rhamnosyl)3-tren)(SO4)] (8), was successfully isolated and characterized by X-ray crystallography, in which three sugar moieties are anchored only at the N atom of the C-l position (8-3CH3OH·H2O: orthorhombic, P212121, a = 16.035(4) Å, b = 16.670(7) Å, c = 15.38(1) Å, V = 4111 Å3, Z = 4, R = 0.084, and Rw = 0.068). Complex 8 could be regarded as an intermediate species toward the C3 symmetrical tris(sugar) complexes 7, and in fact, it was readily transformed to 7b by an action of BaBr2.

Substitution-inert cobalt(III) complexes containing an N-glycoside formed from ethylenediamine (en) and an aldose (D-mannose, L-rhamnose, or D-ribose) were synthesized by the oxidation of cobalt(II) to cobalt(III) in the presence of the diamine. It was found that [(en)2Co(O2)(OH)Co(en)2]3+ is a reactive species toward aldoses in the reaction. The major product for each starting aldose was purified by column chromatography to yield red crystals. The isolated complexes were characterized by elemental analysis, conductivity measurements, and 1H and 13C NMR, electronic absorption, and circular dichroism spectroscopies, the results of which show that the complexes are composed of a bidentate en and a tetradentate N-glycoside ligand. The structures of the sugar units were analyzed by means of the semiempirical AM1 calculations coupled with the conversion of the vicinal 1H—1H spin-spin coupling constants in the 1H NMR spectra into torsion angles of the corresponding H-C-C-H fragments. The results of the calculations demonstrate that the D-mannose and L-rhamnose units take the pyranose form with the β-3S5skew-boat conformation on the cobalt(III) complexes, while the D-ribose unit adopts the furanose form with the α-2E envelope conformation on the complex; in addition the sugar units of the N-glycoside ligands facially bind to the cobalt atom at three points through the donor atoms on Cl, C2, and C3. © 1989, American Chemical Society. All rights reserved.

Carbon-13 n.m.r. measurements have been made in H2O-D2O solutions of cobalt(III) complexes containing an N-glycoside derived from ethylenediamine and an aldose (D-ribose or L-rhamnose). Partial deuteriation of exchangeable protons on the co-ordinated nitrogen atoms permits direct observation of the individual isotopomers and the resonances are observed as a series of multiplets, which can be analysed in terms of the two-bond and three-bond isotope effects that contribute to the deuterium-induced isotope shifts. The C-N bond formation between ethylenediamine and an aldose in the cobalt(III) complexes has been unambiguously confirmed by the information derived from the isotopic multiplets together with complete assignments of 13C signals from the sugar units by means of two-dimensional n.m.r. spectroscopy. An extra doublet originating from the four-bond isotope effect is observed for the D-ribosyl residue, which suggests that the sugar ring takes the furanose form. The importance of sugar ring oxygen atoms in the long-range deuterium isotope shifts is discussed including an examination by use of C1-methine proton deuteriated aldoses.

We have already reported the synthesis and characterization of Co(III) complexes containing an N-glycoside derived from •1ethylenediamine (en) and an aldose. In the report, the formation of a new C-N bond has been presumed according to the chemical 13 shifts of the C signals originated from the en units in their 13 C NMR spectra: one of the signals assigned to the en carbon atoms, which is presumably corresponding to the carbon atom adjacent to the glycosidic nitrogen, appears at 7–8 ppm downfield from the other three. © 1988, Taylor & Francis Group, LLC. All rights reserved.