Journal of the Japan Institute, 47, (3), 145－163 (2004) 145

[Review Paper] An Overview of and Hydrodenitrogenation

Isao MOCHIDA* and Ki-Hyouk CHOI

Institute for Materials Chemistry and Engineering, Kyushu University, Kasuga-koen, Kasuga, Fukuoka 816-8580, JAPAN

(Received May 12, 2003)

Hydrodesulfurization (HDS) and hydrodenitrogenation (HDN) of petroleum products and intermediates are reviewed to provide the basis for developing processes to produce and diesel oil with very low content. The reactivity, selectivity and inhibition (susceptibility of substrate to inhibitors) in the cat- alytic process are very important to develop efficient processes. Recent advances in the understanding of active species, supports and supporting methods are also critically reviewed to suggest the design of catalysts with ade- quate activity to satisfy future regulations on transportation fuels. Details of the structures of the catalysts are not discussed, but the mechanisms of hydrodesulfurization and inhibition are summarized. Catalyst deactivation and reactor design are also briefly reviewed. New approaches to achieve deep hydrodesulfurization are pro- posed.

Keywords Hydrotreating, Hydrodesulfurization, Hydrodenitrogenation, , Gasoline

1. Introduction also reviewed. The preparation, activation, composi- tion and structure of the catalysts in each process are Petroleum refining uses numerous processes includ- discussed along with the associated causes of catalyst ing thermal, catalytic and upgrading deactivation and ultimate catalyst lifetime for each processes as shown in Fig. 1. The hydrogenation process. New and improved catalytic approaches and processes include three major classes, hydrotreating, more active catalysts are also discussed. hydrocracking, and hydrofinishing. Hydrofinishing is Figure 2 illustrates the diversity of composition by really another form of hydrotreating that is used to showing the elemental distribution of some typical achieve the final specifications of fuels. The common petroleum fractions, such as light cycle oil (LCO), features as well as the differences of the various medium cycle oil (MCO), straight run gas oil (SRGO), hydroprocesses will be described. Each process is hydrotreated straight run gas oil (HSRGO), and gaso- individually optimized according to the boiling range line, as determined by gas chromatography equipped and molecular composition of the specific petroleum with atomic emission detection (GC-AED)2),3). Figure fraction to be treated1). Therefore, the process objec- 2 also shows the distributions of specific molecular tives, conditions and configurations, chemistry of fuels species that must be converted into by and products, catalytic materials, their functions, and hydrotreating. The molecular composition of heavier working mechanisms must be understood for all of the fractions such as heavy VGO (vacuum gas oil), and important hydroprocesses in use today. atmospheric and vacuum residues are not fully under- Products in the refining processes are also stood at present, although high performance liquid hydrotreated, and are basically classified according to chromatography (HPLC) and time of flight mass spec- their boiling ranges. This overview describes the troscopy (TOF-MS) have provided some clues to their detailed chemistry of feeds, products, and their conver- molecular composition4),5). These heavier fractions are sion mechanisms in hydrotreating on a molecular level, believed to be polymeric substances of unit structures including the detailed structures of the reactant, their that are basically similar to those found in the lighter chemical and physical properties, and the mechanisms fractions. Strong molecular associations may be pre- of their conversion. The influences of the detailed sent in the residual fractions6), which causes difficulty molecular interactions on reactivity and inhibition are in both analyses and hydrotreating. One method for characterizing the residue is separation into polar and * To whom correspondence should be addressed. non-polar components by precipitation of the polar * E-mail: [email protected] components with a large quantity of a non-polar solvent

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 146

Fig. 1 Stream of Petroleum Refining Process

such as heptane into maltene and asphaltene7). tial loss in liquid product yield. The specific impuri- Asphaltenes are believed to be the major contributors to ties depend on the molecular weight of the feedstock to the undesirable features of the residue, such as high vis- be processed. Lower molecular weight feedstocks cosity, coking tendency, metal content, etc. Thus, the such as , gasoline, intermediate distillates major target of residue hydrotreating is to convert (atmospheric and light vacuum), diesel fuels, and home asphaltenes to lower molecular weight species. The heating oils (, etc.) contain undesirable impuri- asphaltenes consist of polymeric components contain- ties such as sulfur-containing compounds (S-com- ing polyaromatic rings with long alkyl chains that are pounds), -containing compounds (N-com- entangled to form colloidal micelles within the pounds), -containing compounds (O-com- residue8),9). The polymeric chains also contain some pounds), and polynuclear aromatic compounds (PNA). porphyrins, which include metal components (vanadi- Higher molecular weight feedstocks, such as high vacu- um and ), in the petroleum. The polyaromatic um distillates, and atmospheric and vacuum residues rings and porphyrins form stacked aggregates and the contain the same impurities as well as significant con- alkyl chains entangle each other. Such intermolecular centrations of metal-containing compounds (M-com- association is schematically illustrated in Fig. 36). pounds). V and Ni are the major metal impurities in GC-AED chromatograms of light and medium cycle petroleum, which are present in the form of porphyrin oils (MCO) in fluid catalytic (FCC) products complexes of V4+ = O and Ni2+ 6). In addition, crude in the gas oil range are illustrated in Fig. 2 as examples oils often contain NaCl, MgCl2, CaCl2, CaSO4, and of cracked oils. Such processed oils can be further naphthenates of some metals such as Ca, Mg and Fe. hydroprocessed to yield high quality fuels. The metal salts can be removed rather easily by wash- ing before . However, small amounts of 2. Hydrotreating Process metal compounds, particularly Fe or derived FeS, often result in operational problems. Naphthenates may dis- The primary objectives of hydrotreating are to solve iron from valves or the reactor vessels and trans- remove impurities, such as hetero-atom and metal-con- fer lines, and become included in the feeds to down- taining compounds, from a feedstock and/or to increase stream processes. In general, the concentration of the content of the feedstock, and to lower the these impurities increases with increasing boiling point. molecular weight of the by-products without a substan- Thus, the hydrotreating process of choice depends pri-

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 147

SRGO: Straight Run Gas Oil, H-SRGO: Hydrotreated Straight Run Gas Oil, LCO: Light Cycle Oil, MCO: Medium Cycle Oil, VGO (hex- ane soluble fraction): Vacuum Gas Oil. Cn: Paraffin with n , T: , BT: Benzothiophene, DBT: Dibenzothiophene, Cz: Carbazole, DM: Dimethyl, EM: Ethylmethyl, TM: Trimethyl.

Fig. 2 AED Chromatograms of Various Fuel Oils

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 148

In general, sulfur impurity is the major concern because S-compounds are often serious poisons and inhibitors for other secondary upgrading process cata- lysts. Their products create serious envi- ronmental hazards such as acid rain. Thus, the main processes that have been developed for distillable feed- stocks are HDS processes. N-compound impurities are also removed during HDS processes. If succes- sive acid is important in conversion mech- anisms, extensive N-removal is required since the basic N-compounds are both serious poisons and pre- cursors on acid catalysts12). Lowering aromatic con- tent through hydrotreating is classified as A: Deagglomeration of asphaltene due to demetallization, B: hydrodearomatization (HDA). HDA reactions occur Depolymerization due to desulfurization6). during HDS and HDN processes, but product quality requirements often require an HDA process after the Fig. 3 Model of Hydrocracking of Asphaltene initial HDS and/or HDN process. Future environmen- tal regulations may emphasize HDA further13). M-compound impurities are found particularly in marily on the boiling range of the feedstock. The high boiling feedstocks, such as atmospheric and vacu- boiling range is dictated by the molecular weight distri- um residues. Thus, HDM processes are tailored for bution of the feedstock. The next most important con- high boiling and very viscous feedstocks. In such sideration in choosing a hydrotreating process is the processes, the removed metals are deposited on the sur- product quality specification, which is predominantly face of the HDM catalyst, so the lifetime of the catalyst related to the total hydrogen content of the product, is of serious concern. As metals accumulate on the which is related to the content of polynuclear aromatics catalyst, the selectivity for the production of desired (PNA). products also decreases14). Thus, HDM processes are O-compounds are generally not considered as major designed to prevent severe deactivation of the catalysts environmental hazards in petroleum products. to retain activity for cracking, HDS, HDN and HDA Nevertheless, some O-impurity compounds such as reactions. phenols and naphthenic acids lead to corrosion prob- HDM is unique since the metals removed are accu- lems in the reactors and storage vessels. Some crudes mulated as on the catalyst used for removal. which contain large amounts of naphthenic acids are Hence, the catalyst is no longer effective when the sites classified as naphthenic crudes. Such naphthenic or pores are filled with deposited metal sulfides. acids are extracted to be sold as lubricants. Iron dis- HDM is rather easier than other hydrotreating process- solved by naphthenic acid in crude causes plugging by es and the removed metals tend to be deposited on the forming FeS in the catalyst bed or on the filter. Finely hydrotreating catalysts located at the top of down flow dispersed FeS may enhance coking reactions10).O- reactors. The volume available to capture the metals compounds in the petroleum are much more reactive and the extent to which they can be removed are both than other impurities, so hydrotreating is not generally important factors for the process operation. developed specifically to remove O-compounds in common crudes. However, less reactive O-com- 3. Basis for Hydrotreating pounds such as phenols and benzofurans are present in significant amounts in coal-derived oils. Therefore, 3. 1. Hydrotreating Catalysts removal is one of the major concerns in the hydrotreat- Currently, catalysts for hydrotreating are alumina- ing of coal-derived oils11). supported Mo- and W-based sulfides with promoters of S-compounds, N-compounds and M (metal)-com- Ni or Co sulfides. Alumina is believed to be the best pounds have different reactivities and chemistries and most balanced support in terms of surface area depending on the boiling ranges of the fractions in (200-300 m2/g), pore size control, affinity to for which they are found. Thus, specific processes have high dispersion, mechanical strength, and cost. been developed for the removal of each of these impu- precursor (15-20 wt% as ) is first rities, and are classified as hydrodesulfurization (HDS), impregnated onto the alumina to achieve high dispersal hydrodenitrogenation (HDN), and hydrodemetallization and then the Co or Ni precursor (1-5 wt% as oxide) is (HDM) processes, respectively. These are in turn sub- impregnated onto the Mo phase. The impregnated divided into processes, which are optimized for the catalyst is carefully calcined and sulfided for commer- boiling range of the particular feedstock to be treated. cial application to ensure stable catalytic activity. The

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 149 active species is believed to be the Co(Ni)MoS phase. Commercial catalysts also contain isolated 15),16) Co(Ni)9S8 and Co(Ni)/Al2O3, which are not active. The Co(Ni)MoS phase consists of small layered crys- tals of S and Co(Ni)/Mo17). The bottom of the Co(Ni)Mo layers, which contact the Al2O3 surface, is believed to be difficult to sulfide into the active form18), so multi-layered stacks of these layers are probably more active on alumina supports. In order to disperse Mo and Co(Ni) to form more active crystallites, impregnation procedures have been developed which use P2O5 and chelating agents for commercial catalyst preparation. The sulfiding medium and conditions have been extensively studied to achieve higher activi- ty. Microscopic analyses have been used to under- stand the morphology of Co(Ni)MoS phase crystals on alumina. Transmission electron microscope (TEM) and scanning tunneling microscope (STM) have shown that the crystal size of this phase in commercial cata- 19) lysts is very small . Fig. 4 Vacancy Model of the HDS Mechanism18) 3. 2. Chemistry of Hydrodesulfurization The ease of sulfur removal from a petroleum stream depends greatly on the structure of the sulfur compound to be treated. The rates of sulfur removal can vary by final product. It should also be noted that the several orders of magnitude. Generally, acyclic sulfur hydrogenation pathways are subject to thermodynamic compounds such as and are highly equilibrium constraints. Thus, the partially hydro- reactive and can be removed under very mild condi- genated intermediates (such as tetrahydrodibenzothio- tions. Saturated cyclic sulfur compounds and aromat- phenes) have lower equilibrium concentrations at high- ic systems in which sulfur is present in six-membered er temperatures. This results in a maximum in the rings are also highly reactive. However, compounds observed rates of HDS via the hydrogenative route, as a in which the sulfur atom is incorporated into a five- function of temperature. Thus, hydrodesulfurization membered aromatic ring structure (such as thiophene) via the hydrogenative route is limited at low are much less reactive and the reactivity decreases as and high temperatures. Another route includes the the ring structure becomes increasingly condensed (e.g. and transmethylation of methyl group at one ring > two rings > three rings)18). For highly con- the 4- or 6-position, which reduces the steric hindrance. densed ring structures (four or more rings), the trend The direct pathway is believed to involve the insertion reverses and reactivity tends to increase as the ring of a metal atom on the surface of the catalyst into a sul- structure increases in size. The reason for such behav- fur_carbon bond in the sulfur-containing compound1). ior is that there are several different chemical pathways This insertion can occur even for fully aromatic sulfur through which sulfur can be removed from a compounds, such as thiophene, benzothiophene and and the preferred pathway changes for different sulfur dibenzothiophene. Such a pathway is possible compound structures. because the resultant metal_sulfur bond is energetically The reaction scheme shows two major pathways to favorable. After the insertion, several other steps desulfurized products. The first is called direct occur in which the sulfur is expelled as hydrogen sul- hydrodesulfurization, in which the sulfur atom is fide and the catalyst is restored to its original state. removed from the structure and replaced by hydrogen, The hydrogenative pathway involves the initial without hydrogenation of any of the other carbon_car- hydrogenation of one or more of the carbon_carbon bon double bonds. The second is called the hydro- double bonds adjacent to the sulfur atom in the aromat- genative route and assumes that at least one aromatic ic system. Hydrogenation destabilizes the aromatic ring adjacent to the sulfur containing ring is hydro- ring system, weakens the sulfur_carbon bond and pro- genated before the sulfur atom is removed and replaced vides a less sterically hindered environment for the sul- by hydrogen. In addition to hydrogenation of an aro- fur atom. Metal insertion is thus facilitated. matic ring before sulfur is removed, an aromatic ring This discussion indicates that there are two active may be hydrogenated after sulfur removal. This often processes (functions) occurring on the HDS catalyst, S- leads to confusion in interpreting the results of experi- extrusion and hydrogenation. Figure 4 illustrates the mental data as both routes can produce the cyclohexyl- schemes of both reactions including details of the

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 150

Fig. 6 Desulfurization Reactivities of Alkyl-substituted Fig. 5 Desulfurization Reactivities of Compounds with Aromatic Sulfur Compounds18) Different Ring Structures18) active sites18). The active center is a coordinatively the other sulfur species with an aromatic C_S bond. unsaturated metal site where the S ligand is labile. Phenyl rings in benzonaphthothiophene are more easily Sulfur in aromatic rings can coordinate to the active hydrogenated than those in dibenzothiophene (DBT), centers of both functions. Initial adsorption of the S- so HDS is easier than that of DBT. compound is believed to occur through π-bonding, in Another complicating factor in reactivity is the prox- the case of direct S-extrusion. However, S-compound imity of alkyl groups to the sulfur atom in aromatic ring coordination is through π-bonding in the case of the structures. Generally, the reactivity decreases as the hydrogenative route. Neighboring S_H groups are sulfur atom becomes more crowded by adjacent alkyl believed to be involved in the hydrogen transfer for groups. This effect has been attributed to steric hin- both S extraction and hydrogenation. Differences in drance of the sulfur during adsorption on the catalyst the active sites for S extraction and ring hydrogenation surface or during some transition state. This steric are not yet clear, although they appear to be intercon- hindrance affects the direct hydrodesulfurization route vertible. H2S, NH3, and nitrogen containing com- most severely. Figure 6 illustrates this factor for sev- pounds can also coordinate to the active center, inhibit- eral alkyl-substituted benzothiophenes and dibenzothio- ing the S-extraction and hydrogenation as discussed phenes. Figure 618) shows that the reactivity for later. hydrodesulfurization decreases as the number of sub- The direct pathway becomes more difficult as the stituents around the sulfur atom increases. Alkyl ring structure becomes larger because the aromatic groups that are not close to the sulfur atom have little structures become increasingly more stable, and effect. Recently, migration of alkyl groups before because the insertion becomes more hindered for the hydrodesulfurization was proved to enhance direct more condensed rings. To illustrate these factors, Fig. hydrodesulfurization over strong acid catalyst. 5 provides examples of the hydrodesulfurization reac- The hydrogenative routes are not significantly affect- tivities of sulfur compounds with different ring struc- ed by alkyl substitution on the aromatic rings, whereas tures18). For ease of discussion, all rate constants in the direct route becomes less important if alkyl groups this and following illustrations have been normalized are adjacent to the sulfur atom. Thus, the relative rate relative to dibenzothiophene with a value of 100. changes shown in Fig. 6 are primarily due to less Figure 5 shows that the overall hydrodesulfuriza- hydrodesulfurization via the direct route. For this rea- tion reactivity of the sulfur compounds decreases with son, the preferred catalyst for hydrodesulfurization is increasing ring condensation from one ring to two rings often different for light and heavy feedstocks, as the to three rings, but then reverses for the four ring sys- numbers of alkyl groups and condensed aromatic rings tem. This is due to a switch in the preferred pathway in sulfur-containing compounds increases with boiling from the direct route to the hydrogenative route. As point. mentioned above, increasing ring condensation is detri- mental to the insertion step in the direct route, so with 4. Deep Hydrodesulfurization of Fuels increasing ring condensation, hydrogenation becomes easier. The C_S bond in thianthrene has a low bond 4. 1. Deep Hydrodesulfurization of Gasoline energy, so the HDS reactivity is much higher than in Current regulations on the acceptable sulfur levels in

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 151 gasoline and diesel fuel are becoming increasingly catalyst, and atmospheric residue. RFCC yields the stringent to protect the environment from the exhaust same products except for cracked residue instead of gases emitted by motor . Several different decant oil. The sulfur-containing compounds in paraf- refinery streams are blended to produce commercial fins, mono-, di-, and triaromatic ring groups must be gasoline, including straight run gasoline, reformate, removed. The other fractions may not give gasoline alkylate, crude FCC gasoline, residue FCC (RFCC) although they may give H2S which affects the sulfur gasoline, and gasoline from HDS and hydrocracking of content of the gasoline in the FCC process, because vacuum gas oil and residue. Reformate and alkylate opening of the aromatic ring hardly occurs in the FCC are sulfur-free, but the other streams contain various process. levels of S-compounds. Currently, sulfur levels of Strong acidity of is sometimes postulated to such are separately controlled before blend- cause desulfurization through hydrogen transfer, but ing. the contribution must not be exaggerated. Thus, sul- Recent regulations will require deeper hydrodesulfu- fur balances in the FCC process are carefully scruti- rization but HDS of FCC gasoline is rather difficult nized to discover the origins of S-compounds in FCC without hydrogenation of the olefin and aromatic com- products. ponents, which are major sources for high octane num- Recombination of H2S with olefins cannot be ber, although the sulfur species in gasoline are reactive neglected at sulfur levels below 30 ppm in the FCC forms of thiols, , and benzothiophenes, process. Thiols may also cyclize and dehydrogenate which are readily desulfurized. Selective HDS with- into thiophenes, which are major sulfur species in very out olefin hydrogenation is being extensively explored low sulfur level gasolines. Fixation of sulfur can be at present. Such selective hydrodesulfurization designed to occur in coke as well as composite FCC requires clarification of the active sites of CoMo and catalyst and additives. Such sulfur is combusted into NiMo sulfides supported on alumina for hydrodesulfur- SO2 at the regeneration stage. The RFCC catalyst has ization and hydrogenation. CoMo is certainly more an advanced and complex design as shown in Fig. 720). selective for hydrodesulfurization with limited hydro- The next generation FCC is waiting for a very genation activity than NiMo. Hence, is often advanced composite catalyst. The new matrix catalyst applied for the present purpose. Coordinatively unsat- in Fig. 7 has high capability to trap Ni and V by intro- urated valences of Co sulfide on MoS2 are often sug- ducing a new material, called CMT-40 by the catalyst gested as the active site for both reactions. The sites manufacturer, into the matrix20). Although detailed with unsaturated valences may be the hydrogenative information on CMT-40 has not been disclosed yet, it site in cooperation with the Mo_S_H group, and the seems to be based on a zeolitic material. Such a new sites with the di-unsaturated valences may be the catalyst achieves a much longer life time than the con- hydrodesulfurization site. H2S concentration in HDS ventional RFCC catalyst by maintaining the crystal can be reduced to enhance the hydrodesulfurization structure of the catalyst with high trapping capability selectivity. for Ni and V. Several patents have been granted for methods to 4. 2. Deep Hydrodesulfurization of Diesel Fuel poison the hydrogenation site more than the Deep HDS of diesel fuel is currently an important hydrodesulfurization site of CoMoS/alumina. topic. Basically, deep hydrodesulfurization of diesel Amines, alkali metal ions, and deposition have involves the extensive elimination of refractory sulfur been proposed to increase the selectivity for hydro- species such as 4-MDBT, 4,6-DMDBT, and 4,6,X- genation, although hydrodesulfurization is also poi- TMDBTs. Such deep hydrodesulfurization is difficult soned rather than CoMoS. The mechanism for selec- because of the lower reactivities of these sulfur species tive poisoning is not clarified, but the alumina support and strong inhibition by coexisting species such as H2S, appears to be the principal target. The acidic nature of NH3, nitrogen species, and even aromatic species, espe- the support may be responsible for the hydrogenation cially if the sulfur level is to be lowered below 300 activity as found with the hydrogenative HDS of refrac- ppm. H2S and NH3 are produced from the reactive tory sulfur species in diesel fuel. More types of cata- sulfur and nitrogen species contained in the same feed. lyst supports should be selected for detailed examina- There are four approaches for improving reactivity. tion. (1) Introduction of more active sites by impregnating Although the FCC feed can be extensively desulfur- more active metals on the catalyst. ized to produce FCC gasoline with lower sulfur con- (2) Removal or reduction of inhibitors before or dur- tent, FCC gasoline with very low sulfur content ing HDS. requires deep hydrodesulfurization of the particular (3) Novel catalyst designs to introduce new mecha- fraction which yields the gasoline fraction in the FCC nistic pathways that are less subject to inhibition. process, since vacuum gas oil provides gas, gasoline, (4) Two successive layers of catalysts to remove the LCO, heavy cycle oil (HCO), decant oil, coke on the reactive sulfur species and 80% of the refractory sulfur

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 152

Fig. 8 Schematic Diagram and Performance of Two-stage Reaction Concept3)

the high efficiency of active carbon for nitrogen species removal21)～23). The refractory sulfur species are also partly removed by the active carbon, which significant- ly helps deep hydrodesulfurization. Post removal of sulfur species after hydrodesulfurization can also lower the total sulfur content to below 10 ppm. However, the capacity for sulfur removal is rather limited, com- pared to the removal of nitrogen species. Hence the application of this approach is restricted to the prepara- tion of ultra clean hydrodesulfurization for fuel cells. A detailed description of adsorptive desulfurization is given later. The third approach using novel catalysts has high potential and is being investigated extensively. The use of acidic supports appears to enhance the activity by enhancing hydrogenation, methyl group migration, and lowering H2S inhibition, although coking and NH3 inhibition must be overcome. TiO2 and carbon are 20) Fig. 7 New Design Concept of RFCC Catalyst interesting supports for producing higher activity cata- lysts. High surface area TiO2 is now available and shows promise. Deeper sulfiding is one of the pro- species in the first layer, and to reduce the remaining posed reasons for the higher activity. Strong interac- refractory sulfur species to less than 10 ppm in the tion between the active and support is designed presence of H2S and NH3 as well as the remaining for better dispersion of active species, but may hinder inhibitors, such as nitrogen species and aromatic com- adequate sulfiding. Reactive sulfide is recommended pounds, in the second layer. for sulfidation. Strong interactions between active Currently the first method is the major commercial sulfide and support must be explored. approach. The second approach has been proposed as Highly aromatic feeds such as LCO and MCO a two-stage HDS process. Figure 8 shows the effi- appear to require more severe conditions for deep HDS ciency of two-stage HDS, in which the H2S and NH3 because aromatic species strongly inhibit refractory sul- produced in the first stage are eliminated before the fur species24). Catalytic species having higher affinity second layer reaction3). for sulfur than for olefins and aromatic hydrocarbons Another type of two-stage HDS is to remove nitro- are currently targets of extensive research for achieving gen species before HDS with silica_gel, silica_alumina deep HDS of aromatic diesel. The sulfur atom can act or active carbon. The present authors have reported as an anchor to be ported to the soft acid site of the cat-

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 153 alyst for the preferential hydrogenation of the neighbor- Precipitates are often found in heat exchangers as ing aromatic ring. well as in the products of hydrotreating of the vacuum 4. 3. Hydrotreating of Residue residue. This is particularly true if severe conditions Atmospheric and vacuum residues must be are used to increase the yield of distillates. Although hydrotreated to reduce the sulfur, nitrogen, and metal the precipitates from the different sources are all called contents in the preparation of clean fuel for power gen- dry sludge, the compositions are different. The pre- eration as well as distillates and feeds for successive cipitates within heat exchangers are generally insoluble catalytic processes. Vacuum gas oils are often sepa- in any solvent, and resemble coke. By contrast, the rated from the atmospheric residue and separately precipitates that form on standing in hydroprocessed hydrogenated to minimize the amount of vacuum products are usually soluble in . Nevertheless, residues that must be also treated in HDS. These the origins are believed to be similar. The insoluble treated products can be blended back with the vacuum form appears to be carbonized sludge that forms at rela- residue to produce a cleaner atmospheric residue with tively low temperatures (150-200°C) over long periods less difficulty. of time. The toluene soluble sludge precipitates in Alternatively, atmospheric and vacuum residues may storage tanks, transfer lines, and even in feed lines of be hydrotreated without separation. The major prob- the at low temperatures. This toluene soluble lems in the direct hydrotreatment of residues are associ- dry sludge appears to be a high molecular weight ated with asphaltenes, the characteristics of which were hydrotreated asphaltene with limited solubility. Some described previously. of these materials are present in the starting asphaltene, The polymeric components in the asphaltenes are but the others are produced during the hydrotreatment. dissociated through partial hydrogenation of the solvent Both have poor solubility. Hydrotreating often selec- maltene, HDM, HDS/HDN, and hydrocracking, which tively cracks and hydrogenates the maltenes and low occur during the hydrotreating process. The chem- molecular weight asphaltenes in the residue, leaving the istry of sulfur removal in the residue and the chemistry heavy asphaltenes unhydrogenated or even more con- for light distillates are basically the same because both densed. Thus, the solvent properties of the maltenes have similar structural units. However, the larger are lost and the heavy asphaltenes precipitate. molecular units of the residue may cause HDS to pro- Severe conditions emphasize the differences between ceed with more or less difficulty, depending on the pre- maltenes and heavy asphaltenes, which in turn acceler- ferred route for hydrodesulfurization. Major problems ate the precipitation of sludge. Thus, sludge forma- are lower solubility and difficulty in the molecular dis- tion is one of the phenomena which restricts the sociation of the asphaltene micelles that are strongly increase in distillate yield by reducing the quality of the adsorbed on the catalyst and result in coke formation. heavy hydrotreated products. Catalyst deactivation progresses rather easily during Several counter measures have been proposed to residue hydrotreating. reduce this sludge formation including; During severe hydrotreating, the aromatic compo- (1) Catalyst design to convert the heavier fraction nents of the maltene fraction of the residue, which act through cracking and hydrogenation. as good dispersants for the asphaltenes in the residue, (2) Solvent addition. may become partially hydrogenated for hydrocracking, It is important to distinguish between sludge and and so lose the dispersive properties in the coke, as they are chemically and physically different. hydroprocessed products. Some of the asphaltene However, sludge formation or phase separation must be micelles may also remain unhydrogenated or may recognized as a trigger for coke formation. For exam- become dealkylated over acid catalysts, or may even be ple, both strong adsorption of heavy fractions on the thermally dehydrogenated to form less soluble com- catalyst surface or separation of droplets of insoluble pounds than are present in the original feed asphaltene. materials within the bulk of the residue lead to coke Therefore, after the HDS treatment, dry sludge may be formation. precipitated in the down-stream heat-exchanger and 4. 4. HDN, HDO and HDM Reaction transfer lines. The HDS product may also form dry Removal of nitrogen, oxygen, and metal is also sludge in the storage vessels or even in fuel supply important to purify petroleum products. Such reac- lines. Such dry sludge problems occur if hydrotreat- tions progress concurrently together with HDS. ing is performed under very severe conditions in which Activated hydrogen can finally break the C_X (X = S, a higher degree of HDS and cracking is expected. N, O, metal) bonds over the same catalyst although the Deposited sludge is carbonized at lower temperatures affinity to the active site, necessity for hydrogenation than expected because of the very long residence time of the ring structure and C_X bond reactivity are very on the wall. Furthermore, sludge precipitation occurs different according to the mechanisms. with solid flocculates on the wall. Carbonization can The order of ease is generally recognized as HDM, be accelerated for such precipitation. HDS, HDN and hydrodeoxidation (HDO), although the

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 154

HDS and HDN of refractory sulfur and nitrogen species compete during deep removal. HDN of the last non- basic carbazoles is completed before HDS reaches the 10 ppm level. Basic nitrogen species preferentially occupy the active site for denitrogenation before HDS in the competitive reaction. HDN of aromatic nitro- gen species is generally believed to proceed through complete hydrogenation of aromatic rings because the aromatic C_N < bond is too strong to be broken by hydrogenation cracking, as aliphatic C_N < is required. Thus NiMoS and NiWS catalysts are often applied for HDN. Large hydrogen consumption is inevitable. Acidic support helps HDN over NiMoS by accelerating the hydrogenation, although the occurrence of coking is likely to increase deactivation. Hydrogenation and acidic reaction are completed on the same catalysts. Recently the substitution of the C_N < bond with _ H2S has been proposed to form C SH and NH3 product. The C_SH bond is easily eliminated under hydrotreat- ing conditions. 2-Methylcyclohexylamine is hydrodeni- trogenated through three routes, direct elimination of ammonia, nucleophilic substitution of NH2 group by _ H2S then SH group decomposition, and direct of the C_N bond25),26). The estimated contribution of each pathway is shown in Fig. 926). Such contribution of nucleophilic substitution is depend- ent on H2S partial . However, whether such a mechanism is effective in HDN of refractory nitrogen species, such as carbazole, was not clarified because H2S is a strong inhibitor for HDN of such inert nitrogen species. The HDN resistance of such species are all examined in terms of inhibition by H2S. The natures of the catalysts are also important since various catalyst properties are now available. The applicability of aromatic C_N <, especially derivatives from carbazoles, is not yet established. Such a mechanism is helpful to reduce hydrogen con- Fig. 9 Selectivities for Elimination (A), Nucleophilic Substi- sumption. HDO is not an important reaction for tution (B), and Hydrogenolysis (C) in the HDN of 2- Methylcyclohexylamine (MCHA) and the Observed petroleum products but is very important to stabilize Selectivities of Methylcyclohexene (MCHE) and the coal-derived liquid. HDO of dibenzofuran is very Methylcyclohexane (MCH) in the HDN of 2-Ethyl- slow. Acidic support is helpful for HDO of dibenzo- cyclohexylamine (plain figures) and in the HDS of 2- furan species. Methylcyclohexanethiol (MCHT) (bold figures) in the 26) HDM is the key for hydrotreatment of heavy oil in Presence of 20 kPa (upper) and 200 kPa (lower) H2S terms of deep treatment of asphaltenes and capacity of the catalyst to hold eliminated metal sulfides14). Liberation of asphaltene aggregation as well as mouth size and volume of the pour in the catalysts are impor- mentioned above, several species are inhibitors for tant points for the design to enhance HDM of heavy HDS. Reactive sulfur species appear to be less inhibit- feeds. ed than refractory species. This is because the S atom in the reactive species can easily undergo metal inser- 5. Inhibition of HDS tion to break the C_S bond via the direct HDS route. The reactive species are also the major S-compounds The active sites postulated for HDS catalyst promote present and can compete effectively with inhibitors for sulfur extrusion, hydrogenation, and acid catalyzed the active sites on catalysts. By contrast, in the direct reactions. Such active sites are all commonly or HDS route, the sulfur atom in refractory sulfur species selectively subject to occupancy by inhibitors. As may be sterically hindered. The concentrations are

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 155 also very low whereas H2S and NH3 inhibitors increase ing the expensive active metals from spent hydrotreat- in concentration during the initial stage of the HDS ing catalysts, including physical attrition of inactive process. These inhibitors are present in high concen- metal sulfides from catalyst surfaces as well as selec- tration where the refractory species must be desulfur- tive dissolution schemes. ized. Some processes have been developed which remove H2S and NH3 between stages to minimize this 7. Process Flow of Hydrotreating problem as described above. Other feed impurities, such as N-compounds, are severe inhibitors for the A typical hydrotreating flow diagram is shown in hydrogenative HDS route. These strong π-bonding Fig. 10 of single stage hydrotreating. The feed oil is species hinder the interaction of the refractory S-com- pumped up to the required pressure and mixed with pounds with the active catalytic site. The overall HDS make-up and recycled hydrogen-rich gas. The tem- process much easier if the N-compound inhibitors are perature is initially raised by heat exchange with the removed prior to hydrotreating27). Aromatic species reactor effluent then further increased by a furnace to are also inhibitors for HDS of the refractory sulfur achieve the required temperature. The feed oil is species as described above. hydroprocessed over the catalyst in the reactor under a flow of pressurized hydrogen-rich gas. The figure 6. Deactivation and Regeneration of Hydrotreating shows one reactor, but more reactors may be used even Catalyst in single stage processes, depending upon the condi- tions or throughput rate. Hydrotreating catalysts lose their activity in several In general, a fixed bed reactor is employed for the ways. hydrotreating process. However, a series of catalysts (1) of the active phase into large crystal with different functions are generally packed sequen- units. tially in the reactor(s), with one to several catalyst beds (2) Degradation of the active phase, including degra- depending on the requirements. The feed oil and dation of sulfide forms. hydrogen-rich gas are normally supplied from the top (3) Covering of the active sites by reactants and/or of the reactor. Quenching hydrogen gas is commonly products including coke. injected at several points along the reactor to control (4) Deposition of inactive metal sulfides (such as V the reaction temperature because hydrotreating reac- and Ni sulfides). tions are always exothermic. The reactor effluent is (5) Deposition of other impurities such as salt and then cooled down in the . This recov- silica. ers the exothermic heat of reaction and improves the The deactivation usually occurs in three steps; initial thermal efficiency of the overall process. Following rapid deactivation, intermediate slow but steady deacti- heat exchange, the gas and liquid products are separat- vation, and rapid deactivation at the end of the cycle. ed by a sequence of a high-pressure vessel at high tem- Commercial processes are operated at constant conver- perature, followed by a high-pressure vessel at low sion. This constant conversion is achieved by gradu- temperature. Liquid products are further fractionated ally heating the reactor to higher temperatures to com- into the required products in the fractionating column pensate for the slow but steady catalyst deactivation. according to their boiling points. The gaseous prod- The initial rapid deactivation phase is believed to be ucts, and hydrogen, from the high-pressure vessel are due to rapid coking on active sites with very high activ- fed to an absorbing column to remove hydrogen sul- ity. The slow but steady deactivation is associated fide, and the cleaned hydrogen-rich gas is recycled to with metal deposition, sintering, and/or coking during the reactor after repressurizing with a recycle compres- the course of the process cycle. The higher reaction sor. temperatures utilized at the end of the process cycle There are two types of processes, the single stage may cause the final rapid deactivation. process and the two or multiple staged process. The Currently, acidic supports are utilized to achieve single stage process has the same process flow as men- high activity, hence coking deactivation is important in tioned above. The feed is hydroprocessed consecu- today’s processes. Such deactivation schemes suggest tively without obvious separation between the reactors, that catalysts could be regenerated if suitable methods as described above. However, a single stage process can be developed. Removal of strongly adsorbed does not mean that only one reactor is employed, only heavier organic materials or coke precursor and coke that no separations are done until the final conversion is could possibly be achieved by thermal extraction achieved. and/or combustion. However, the active sulfide form In the two-stage process, the unwanted products of must be maintained during regeneration of the catalyst, the first stage are separated and eliminated before the or it must subsequently be regenerated. second stage. Thus, the unwanted secondary reactions Numerous methods have been proposed for recover- of the product, poisons and inhibitors produced in the

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 156

Fig. 10 Single-stage Process Flow Diagram

Fig. 11 Reactor Design for Deep HDS28)

first stage, are eliminated before beginning the second were reacted separately in upper and lower parts of the stage of the conversion. This reduces the load on the catalyst bed28). Hydrogen was charged from the bot- second stage and enhances its reactivity. With staged tom of the reactor. H2S inhibition in the heavier frac- processes, very high conversions, so-called deep refin- tion can be avoided. The optimum catalysts can be ing, are easily achieved. applied for the respective parts of the bed. The present authors proposed a new type of reactor By lowering the end point of the starting diesel fuel, as shown in Fig. 11, in which fractions of a gas oil hydrotreating of the lower end point diesel fuel feed to

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 157

Table 1 Remaining Sulfur and Nitrogen Content after Hydrotreatment over Catalysts for 1 h at 340°C3)

HSRGO in the absence HSRGO in the presence of H2S and Name SRGO [ppmS] SRGO [ppmN] of H2S and NH3 [ppmS] NH3 [ppmS] CoMo-A 1118 81.0 22.5 36.7 NiMo-A 2229 12.2 2.1 5.9 CoMo-SA 576 41.6 7.1 36.4 NiMo-SA 477 19.6 4.4 32.7 CoMo-AZ 323 17.1 5.3 17.8 NiMo-AZ 242 5.4 3.6 15.5

2 Catalyst/oil = 1 g/10 g, H2 = 50 kg/cm (initial charging pressure), H2S = 1.66 vol% in H2, NH3 = 200 ppmN. A: Alumina support, SA: silica-alumina support, AZ: alumina-zeolite support. ultra low sulfur levels is much easier. For example, if space velocity larger than 1. The catalysts of the two the 90% distillation point (T90) of diesel fuel is low- layers are not always the same. Optimum catalysts ered by 20°C, the required reactor size is only about must be selected. The different roles of the catalysts half that needed for the full range feed. This diesel in the first layer and second layer can be satisfied by with lower T90 will produce less particulate matter in selecting active species and supports. The activity for diesel exhaust gas. The downside of this approach is reactive and refractory sulfur species and the resistivity the requirement for increasing the cracking capacity of against inhibitors at the respective level of sulfur con- the refinery to produce the required volume of diesel tents must be taken into account. Furthermore, deni- fuel with this lower T90. One solution is to convert trogenation in the first layer is also important since the the VGO hydrotreater to the mild hydrocracking remaining nitrogen strongly influences the hydrodesul- process. furization of the remaining refractory sulfur species of The fluorescence color of finished diesel oil is a 300 ppm to less than 15 ppm. stringent requirement in some countries. Presently, Nitrogen compounds strongly inhibit HDS, in partic- hydrodesulfurization of faintly yellow diesel oil feed- ular in the deep range. Such inhibition may be over- stocks produces colorless and transparent products at come by adopting a suitable HDN catalyst. However, 500 ppm S. However, severe conditions for deep it may be very difficult to find a super active catalyst HDS result in fluorescent yellowish green diesel oil. with very high HDN activity toward alkylated car- High hydrogen pressures suppress the color formation bazoles, which are regarded as having comparable or whereas a high reaction temperature conversely retards lower reactivity than refractory sulfur species. Hence, hydrogenation and enhances color formation29). pre-removal of nitrogen species prior to HDS has been attempted and HDS reactivity of nitrogen-removed gas 8. Novel Design for Deep Hydrodesulfurization of oil has been evaluated under the conventional condi- Gas Oil tions. We selected materials as an adsorbent because of the very large surface area and Figure 2 illustrates the sulfur distribution of current easily controlled surface properties. Figure 12 shows 500 ppm diesel fuel. Comparison of this sulfur distri- nitrogen and sulfur breakthrough profiles over activated bution to that of SRGO revealed that 100% of reactive carbon21). Activated carbon removed nitrogen species sulfur species and 80% of refractory sulfur species are and refractory sulfur species simultaneously. removed to obtain 300 ppm by current HDS. The Adsorption capacity was estimated to be 0.098 g-sulfur nitrogen species in 500 ppm diesel are also illustrated and 0.039 g-nitrogen per 1 g of activated carbon at in the same figure, showing carbazole of 50 ppm. 30°C. The performance of activated carbon in the A sulfur level of less than 15 ppm could be achieved adsorptive removal of nitrogen and sulfur species from by desulfurizing the remaining refractory sulfur species conventional gas oils was strongly dependent on the of 300 ppm in the current diesel fuel in the presence of surface properties, such as surface area and the amount 3) 24) inhibition products such as H2S and NH3 . Table 1 of surface oxygen groups . Such findings indicate summarizes the activity of some catalysts under such the possibility to develop an adsorbent which is opti- conditions. NiMo on acidic supports achieved a sul- mized to treat a particular gas oil. Adsorptively treat- fur level of less than 15 ppm. Acidic supports of ade- ed gas oils showed much higher reactivity in conven- quate strength overcome the inhibitions by H2S and tional HDS than non-treated gas oils as indicated in 21) NH3. The sulfur level of 300 ppm can be achieved by Fig. 13 . Greatly enhanced reactivity comes from a space velocity larger than 3 over CoMo catalysts with the absence of nitrogen inhibitors and diminished acidic supports. Hence, combined two catalysts in the amount of refractory sulfur species. layer can achieve deep hydrodesulfurization with a Many refineries plan to build more HDS units to

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 158

Adsorbent: 1.0 g of OG-20A, Adsorption temperature: 30°C.

Feed: straight run gas oil (11,780 ppmS and 260 ppmN), Fig. 14 Sulfur Breakthrough Profiles of Conventionally Adsorbent: MAXSORB-II (2972 m2/g), Adsorption bed: Hydrotreated Straight Run Gas Oils (HSRGO) Stainless steel tube of 50 mm length and 6 mm diameter, Feed Containing (A) 340 ppm and (B) 50 ppm Sulfur23) rate: 0.1 cm3/min, Adsorption temperature: 30°C.

Fig. 12 Sulfur and Nitrogen Breakthrough Profiles over Activated Carbon21)

Total sulfur contents were (A) 193 ppm, (B) 11 ppm, and (C) 8 ppm21). Reaction temperature and time: 340°C, 2 h. Oil/cata- _ 23) lyst: 10 g/1 g. Catalyst: CoMo/SiO2 Al2O3 (commercially Fig. 15 Concept of the Two-step Adsorption Process available).

Fig. 13 Sulfur and Nitrogen Chromatograms of HDS Products from (A) SRGO (total nitrogen content = 260 ppm), result in increased prices of diesel fuel. Hence, we (B) Adsorptively Treated SRGO (total nitrogen con- proposed the post-treatment system which utilizes the tent = 60 ppm), and (C) Adsorptively Treated SRGO activated carbon adsorption bed. Figure 14 shows (total nitrogen content = 20 ppm) the sulfur breakthrough profiles over the activated car- bon bed23). The feed was conventionally hydrotreated gas oil (HSRGO). Activated carbon can remove the meet the future regulations on the hetero-atom content sulfur species to meet the future regulations. As in diesel fuel. Certainly, slower processing in the described above, pre-treatment of gas oil to enhance its HDS unit will produce clean diesel fuel with less sulfur reactivity can be performed over an activated carbon content. However, the scheduled regulations require bed. The adsorption bed used in the pre-treatment had back-up systems to ensure the quality of final diesel adequate adsorption capacity for removal of sulfur product if serious problems occur in the HDS unit. species from the hydrotreated gas oil. This observa- Such back-up systems may mean an extra HDS unit or tion indicates the practical applicability of our adsorp- large volume tank, which can dilute the out-of-specifi- tion system as a unit for daily operation and/or emer- cation product with the in-specification diesel fuel. gency back-up. Such a scheme is shown in Fig. 1523). However, the investment in the extra processes must

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 159

Table 2 Catalytic Properties of the Zeolite Containing Catalysts33)

a) Surface area Pore volume NH3-desorbed amount HDS product of SRGO HDS product of HSRGO in the 2 3 a),b) [m /g] [cm /g] [mmol NH3/g-cat.] [ppmS] presence of H2S and NH3 [ppmS] NMAZ 262 0.35 0.66 242 15.5 NMACZ 129 0.36 0.61 1455 10.3 a) Reaction temperature and time: 340°C for 1 h. Oil/catalyst: 10 g/1 g. b) H2S = 1.66 vol% of charged gas, NH3 = 200 ppmN.

9. Recent Progress in HDS Catalysts

9. 1. Progress in Support Material for More Active HDS Catalysts The development of HDS catalysts has achieved sig- nificant progress. Some proposals have involved active species other than CoMo and NiMo, but the major focus to improve catalytic activity has been to identify more active support materials. Some oxide and carbon materials have shown superiority to alumi- na in the HDS reaction, but this could not be general- ized due to the wide variation of reaction conditions and feed characteristics in HDS. The support affects Reaction temperature and time: 340°C, 1 h. Oil/catalyst: 10 the catalytic activity in HDS by modifying the active g/1 g. species and participating in the HDS reaction as a co- catalyst. Fig. 16 Carbon Specific Chromatograms of SRGO and Its 33) Acidic supports, such as zeolite, show superiority in HDS Product Oils over NMAZ and NMACZ the HDS reaction. Acidic zeolite facilitates trans- of alkylated DBT, resulting in enhanced HDS activity of such refractory sulfur species30). lower surface area of NMACZ was believed to result in Incorporation of MCM-41 into alumina provides the lower activity in SRGO. However, the surface higher activity of CoMo catalyst in DBT HDS31). area did not affect the HDS in the second layer due to MCM-41 reduces the interaction of Co and Mo with the very low concentration of sulfur species in that the support, resulting in enhanced formation of poly- case. meric Mo octahedral species, which is regarded as the As described above, zeolite-containing catalyst active form. Furthermore, MCM-41 modifies the dis- shows high activity toward refractory sulfur species tribution of Co to retard the formation of the inactive due to its strong acidity. However, this advantage CoAl2O4 phase. Such results indicate the importance must be diluted by augmented cracking of feedstock, of the acidic nature of support in modifying the active resulting in density decrease. Furthermore, strong sites. acidity implies the fast coking and deactivation of cata- We have also investigated zeolite-containing HDS lyst. Hence, we had to find the optimum catalyst to catalysts to achieve deep HDS of SRGO32),33). obtain high HDS activity and low cracking activity. Because of the difficulty in loading active species on Figure 16 shows the carbon-specific chromatograms zeolite, NiMo catalysts supported on alumina-USY of feedstock and HDS products over NMAZ and zeolite composite were examined in the HDS of SRGO NMACZ. NMACZ showed much lower cracking (sulfur content 11,780 ppm) and HSRGO (sulfur con- activity32). The effort to find the best catalyst, which tent 340 ppm) to simulate the first and second layer can satisfy the complicated requirements demanded by reaction using an autoclave-type reactor. The alumi- the field engineer, must be continued. na-zeolite composite supports were a simple composite TiO2 was evaluated as active support material for (NMAZ) and a surface modified composite (NMACZ). HDS. The main drawback, low surface area, may be NMAZ showed superior activity to NMACZ for the overcome by variation of the manufacturing process HDS of SRGO. In contrast, NMACZ left just 10.3 conditions, resulting in a surface area as high as 120 2 34) ppmS after HDS of HSRGO in the presence of H2S and m /g . The TiO2 support could carry a higher load- NH3 whereas NMAZ left 15.5 ppmS after the same ing of Mo of 19 wt% MoO3 compared to the conven- treatment as shown in Table 2. The acidity of tional TiO2 support and Mo species were well dis- NMACZ was slightly lower than that of NMAZ. The persed. Such a catalyst showed higher HDS activity

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 160 for DBT than the alumina supported catalyst. pounds, such as CoAl2O4. Hence, the use of complex- _ 44),48) 52) Al2O3 TiO2 composite support was also tested to pro- ing agent provide higher stacking of MoS2 ～ . vide higher surface area, showing higher activity than Another approach is to design the nanometer order Al2O3 or TiO2 in HDS of 4,6-DMDBT, but the activity microcrystal shape by micelle technology, which is a on the HDS of DBT was inferior to the alumina common methodology in “nanotechnology.” support35)～37). Such binary oxides carry more Brønsted Atomic scale observation of the HDS catalyst could acidic sites and, as a result, have rather high activity in provide greater insight into the morphological charac- hydrogenation, resulting in increased activity in 4,6- teristics. Examination of model Co_Mo_S structures DMDBT HDS. TiO2 also provides less strong inter- on well defined substrates by STM revealed that Co action between active species and support than Al2O3, changes the morphology of MoS2 from the triangular resulting in a more easily reducible oxide phase of Mo structure to the truncated hexagonal structure by pre- 38) 19),53) and Co(Ni) due to the partial sulfidation of TiO2 . ferred siting of Co at the sulfur edge site of MoS2 . Carbon has been regarded as a promising catalyst STM observation of the thiophene adsorbed model support due to its very high surface area, peculiar pore MoS2 suggested that the adsorption site is the metallic 38) structure, and surface functional groups . HDS cata- edge site of MoS2. TEM observation of HDS catalyst lyst supported on carbon had higher activity than alu- could be also used to evaluate the stacking degree and mina supported one39)～42). Presumably, the weak inter- width of each stack, although the images lead to misin- 47) action between sulfide precursor and carbon support terpretation of the morphology of MoS2 . provides more sulfidable species than alumina, result- 9. 3. Catalytic Active Sites for Hydrodesulfuriza- ing in more active sites for HDS. tion and Hydrogenation 9. 2. Control of Shape and Size of Active Sites of Co(Ni)MoS supported on alumina shows catalytic HDS Catalyst activity for HDS, HDN and HDO, as well as hydro- There are several proposals to describe the active site genation. All catalysis processes simultaneously acti- for HDS, such as the intercalation model, surface com- vate hydrogen to substitute the counterpart of the C_S plex model, and rim-edge model18). Such models bond or saturate the C_C . However, include the stacked layer structure of MoS2, which is some features of these catalysis processes are different, believed to constitute the active site. The two types of suggesting different active sites on Co(Ni)MoS/Al2O3. 43) active sites on MoS2 slabs may be the rim and edge . NiMoS is believed to be more active for hydrogenation The rim site is responsible for the hydrogenation and than CoMoS1),18). the edge site is responsible for both hydrogenation and A very low concentration of H2S appears to retard direct hydrodesulfurization. Such images provide the hydrodesulfurization but accelerates hydrogenation, design concepts of an active site with higher activity although a high concentration prohibits both process- toward HDS, larger number of slabs, or stacking pro- es54),55). Acidic support accelerates hydrogenation viding higher activity44). Furthermore, Mo species in more than hydrodesulfurization, although the acidity direct contact with the support surface cannot be easily strengthens the resistivity against H2S inhibition during sulfided, showing lower activity. Such concepts are hydrodesulfurization. Such results suggest that the described as the Type I and Type II phases, highly dis- active sites for these catalysts are Ni or Co, coordina- persed monolayer MoS2 and lower dispersion and high- tively unsaturated more for HDS than for hydrogena- 45),46) er stacking MoS2, respectively . Hence, the tion. HDS requires two open valences to insert hydro- _ method to stack the MoS2 phase higher has been pro- gen into the C S bond. In contrast, π-coordination of posed. Among several types of anchoring sites on γ- the aromatic ring or unsaturated bond in the aromatic Al2O3, the OH group bound to tetrahedrally coordinat- ring to the active site is followed by addition of hydro- ed aluminum cation has been proposed to result in gen from S_H on the sulfide for the first stage of hydro- strong interaction with the Mo precursor, resulting in genation. Strong adsorption of the partially hydrogenat- less sulfidable Mo species47). In order to suppress the ed aromatic ring allows the successive hydrogenation anchoring of Mo precursor on such inactive sites, Ti of other unsaturated bonds to complete aromatic hydro- could be deposited on the alumina prior to impregna- genation. Acidic support modifies the electron densi- tion of Mo to prevent the deposition of the Mo precur- ty of the active metal to be more acidic, activating the sor onto the inactive site, which is blocked by Ti47). S_H or Ni_H bond to transfer the proton for hydrogena- This observation may be related to the dependence of tion. Thus, the activation potential of metal sulfides morphology of MoS2 on the support type and prepara- for hydrogenation may reflect that of metals. Nickel tion conditions. Complexing agents, such as NTA is certainly more active for hydrogenation than cobalt. (nitrilotriacetic acid), could be used to improve the dis- Other metals including noble metals should be exam- persion of active species. Such complexing agents ined to assess the selectivity between hydrodesulfuriza- could prevent the direct interaction of active species tion and hydrogenation, although the relationship of with the support, resulting in inactive, stable com- metal_S and _H bond strength and catalytic activity can

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 161 be compared in a series of metals according to the at constant hydrogen pressure restrict the hydrogena- Sabatier principle. Another important point not much tion very sharply, reducing the obstacles against direct discussed is the chemistry of π-coordination in the elimination. Such results indicate the steric hindrance catalysis. The authors are particularly interested in of the alkyl group at the 4 and 6 positions on the DBT the contribution of heteroatoms, especially sulfur, in ring. The question is at which step the steric hin- the same or adjacent rings to the π-coordination. The drance governs the hydrodesulfurization. Sulfur catalytic sites must be also clarified in such π-coordina- species can be adsorbed onto a sulfide catalyst first tion. It may be important to distinguish carbon and S through π-coordination of the DBT ring. The extent in the aromatic ring for the selective hydrogenation of and position of coordination of the π-electrons of the S DBT in the presence of the major aromatic hydrocar- atom on the catalyst are not presently known. The π- bon component of LCO and HCO from FCC. The coordinated species is transformed into π-coordination involvement of alkyl groups on the DBT ring is also of the sulfur atom at the active site, probably Co or Ni important. It is necessary to distinguish the target with coordinative unsaturation, which leads to the species in the hydrodesulfurization of cracked gasoline. insertion of the active metal between the aromatic C Selective poisoning of hydrogenation catalyst has and S atoms. The C-active site and C_S bonds are been patented to obtain selectivity for hydrodesulfur- hydrogenatively fissioned, probably successively with ization. The poison is postulated to be adsorbed on two from the catalyst surface for S extrac- the support to reduce the hydrogenation activity more tion by the active site of the sulfide catalyst. The SH than the hydrodesulfurization activity, in agreement group on the active site is further hydrogenated and with the conclusion that the acidic support accelerates desorbed eventually as H2S. The steric hindrance can the hydrogenation. Unfortunately, such selective poi- affect the π-coordination, insertion, and hydrogenative soning is not very selective, also reducing the fission steps. hydrodesulfurization activity. More details and better The π-coordinating step is often considered to be the understanding of the active sites and influential factors important step. Strong inhibition against refractory in the mechanism of hydrodesulfurization and hydro- sulfur species, especially at very low concentrations genation must still be explored. The authors believe lower than 300 ppm, favors the steric hindrance at this that detailed description or images of the molecular stage since the inhibition is basically competition for structures of the catalyst, support, substrate, intermedi- active sites among the substrates including inhibitors. ate, and the reaction dynamics can be logically com- π-Coordination of the aromatic ring may be not proba- bined to build up catalysis frameworks. ble. π-Coordination to an active site for π-coordina- 9. 4. Roles of Steric Hindrance in HDS tion site or hydrogenation site may hinder the Deep hydrodesulfurization of gas oil is dependent on hydrodesulfurization. In the latter case, only the the deep hydrodesulfurization of refractory sulfur hydrogenation route is inhibited and no inhibition can species, which persist at a significant level after the be observed with direct elimination of the sulfur atom conventional hydrodesulfurization process. The even if slow. Detailed kinetic analyses are needed. refractory sulfur species have low reactivity and are Heat of adsorption measurements suggests strong inhibited by H2S, nitrogen compounds including NH3 adsorption of refractory sulfur species. The measure- and even aromatic compounds. The refractory sulfur ment often suggests dominant π-coordination adsorp- species carry alkyl groups at the 4 and 6 positions on tion. Hence steric hindrance at π-coordination is not the DBT ring, and the involvement of the 1 and 9 posi- excluded by this result. Kinetic analyses to estimate tions has not been proved yet, whereas the 2, 3, 7, and 8 the heat of adsorption of the real intermediate are nec- positions are recognized not to deactivate but accelerate essary. hydrodesulfurization. The refractory sulfur species More details of π-coordination on the catalyst are are desulfurized mainly through the hydrogenation of at essential in terms of its relation with hydrogenation of least one phenyl ring prior to sulfur elimination. the aromatic species and which sites of sulfide catalyst Hydrodesulfurization of hydrogenated intermediates involved. Selectivity among aromatic species of pure is very rapid under current hydrodesulfurization condi- hydrocarbons, and nitrogen-, sulfur-, or oxygen-con- tions. In contrast, the reactive sulfur species are main- taining rings can be discussed by clarifying such points. ly desulfurized through direct elimination of the sulfur Steric hindrance of alkyl groups on hydrogenation has atom without hydrogenation of a phenyl ring, even if been discussed at low temperature16). However, no the hydrogenated intermediate could be desulfurized evidence has been observed in HDS above 300°C. more rapidly. Thus, hydrogenation of one phenyl ring The stage of insertion during hydrogenative bond fis- in a refractory sulfur species is the rate-determining sion can be also postulated as affected by steric hin- step. At temperatures higher than 360°C, direct drance. The alkyl groups of other locations may be hydrodesulfurization becomes dominant even with involved in this stage. In this sense, the reactivities of refractory sulfur species, because higher temperatures 1,9-dialkyl DBT are key to discuss the hindrance at

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 162 these stages. The same questions have been posed in Economy, Trade and Industry (METI) for financial the HDN of aromatic nitrogen species. So far, in support for many years. The authors are particularly depth discussion has not been attempted. grateful to their students and colleagues in Kyushu In conclusion, we believe that steric hindrance is pre- University. A number of friends in universities, dominantly important at the π-coordination stage. research institutes, petroleum engineering and catalyst Details of π-coordination must be more involved in the vendors all over the world have helped to continue chemical competition at the catalyst surface, even if the research in the field of petroleum refining. Special step may not be rate-determining as seen in similar thanks to Dr. D.D. Whitehurst for his long term and petroleum refining reactions. vivid friendship. 9. 5. Bridges between Homogeneous and Hetero- geneous HDS Reactions References Recently, a number of sulfur-containing coordinated complexes including the DBT ring have been synthe- 1) Topsoe, H., Clausen, B. S., Massoth, F. E., “Hydrotreating sized and the hydrodesulfurization reactivity measured. Catalysis,” Springer, Berlin (1996). 2) Shin, S., Yang, H., Sakanishi, K., Mochida, I., Grudoski, D. A., Coordination chemistry for stability and three-dimen- Shinn, J. H., Appl. Catal. A: General, 205, 101 (2001). sional structures at the molecular level as well as 3) Choi, K.-H., Kunisada, N., Korai, Y., Mochida, I., Nakano, K., hydrogenative fission of C_S, C_Me and Me_S bonds Catal. Today, 86, 277 (2003). are important for heterogeneous hydrodesulfurization 4) Qian, K., Dechert, G. J., Anal. Chem., 74, 3977 (2002) catalysis, although a catalytic process using coordina- 5) Yongzhi, L., Xianliang, D., Weile, Y., Fuel, 77, 277 (1998). 6) Takatsuka, T., “Hydrotreatment-Science & Technology,” ed. by tion compounds is far from practical. More knowl- Kabe, T., IPC, Tokyo (2000). edge of the stability of the coordinated forms, influ- 7) Ma, X., Sakanishi, K., Isoda, T., Mochida, I., Fuel, 76, 329 ences of ligands, and substituting sources for bond fis- (1997). sions will help to understand heterogeneous catalysts 8) Yen, T. F., Erdman, J. G., Pollack, S. S., Anal. Chem., 33, 1587 and catalysis. The involvement of S_H on the sulfide (1961). _ 9) Sakanishi, K., Yamashita, N., Whitehurst, D. D., Mochida, I., catalyst and O H on the support is implied by the Catal. Today, 43, 241 (1998). homogeneous HDS schemes. Another factor is the 10) Gentzis, T., Parker, R. J., McFarlane, R. A., Fuel, 79, 1173 coordination clusters of multi-nuclear molybdenum and (2000). which formed the model and designed precur- 11) Sumbogo Murti, S. D., Sakanishi, K., Okuma, O., Korai, Y., sors for the CoMoS or NiMoS phase. Nickel and Mochida, I., Fuel, 81, 2241 (2002). 12) Massoth, F. E., Catal. Lett., 57, 129 (1999). cobalt can be arranged in designed positions in the 13) Kaufmann, T. G., Kaldor, A., Stuntz, G. F., Kerby, M. C., cluster, in particular in highly dispersed and high stack Ansell, L. L., Catal. Today, 62, 77 (2000). units. 14) Wei, J., “Catalyst Design-Progress and Perspective,” ed. by Hegedus, L. L., John Wiley & Sons, New York (1987). 10. Conclusions and Acknowledgment 15) Fleet, M. E., Act. Cryst., C43, 2255 (1987). 16) Pettiti, I., Botto, I. L., Cabello, C. I., Colonna, S., Faticanti, M., Minelli, G., Porta, P., Thomas, H. J., Appl. Catal. A: General, Engineering development on energy in Japan is now 220, 113 (2001). facing a turning point in the 21st century. 17) Topsoe, H., Clausen, B. S., Catal. Rev.-Sci. Eng., 26, 395 Engineering development of the energy is very impor- (1984). tant. However, recent trends in our society and econ- 18) Whitehurst, D. D., Isoda, T., Mochida, I., Adv. Catal., 42, 345 (1998). omy have very much restricted technical development 19) Lauritsen, J. V., Nyberg, M., Vang, R. T., Bollinger, M. V., and commercialization in energy related industries. Clausen, B. S., Topsoe, H., Jacobson, K. W., Lagsgaard, E., We sincerely hope to stimulate researchers and engi- Norskov, J. K., Besenbacher, F., Nanotechnol., 14, 385 (2003). neers to identify future methods in the petroleum relat- 20) Nonaka, S., The 11th CCIC Technical Seminar, Catalysts & ed industry. Universities are strongly required to Chemical Ind. Co., Ltd., 2002, p. 2-1. 21) Sano, Y., Choi, K.-H., Korai, Y., Mochida, I., Appl. Catal. B: expand their activities to the international community Environment., 49, 219 (2004). for survival. 22) Sano, Y., Choi, K.-H., Korai, Y., Mochida, I., Energy & Fuels, The authors are grateful to the editor for providing a accepted, (2004). chance to summarize our current research interests in 23) Sano, Y., Choi, K.-H., Korai, Y., Mochida, I., Fuel Proc. Tech., hydrotreating catalysis. One of the present authors submitted, (2004). 24) Choi, K.-H., Korai, Y., Mochida, I., Preprint Fuel Chem. Div. (IM) has been involved in hydrotreating research for ACS, 48, 653 (2003). the past two decades. It is the best time for him to 25) Egorova, M., Zhao, Y., Kukula, P., Prins, R., J. Catal., 206, 263 conclude his research results and further plan for the (2002). coming generation in his retirement year. The authors 26) Rota, F., Prins, R., J. Catal., 202, 195 (2002). are also grateful to New Energy and Industrial 27) Sumbogo Murti, S. D., Yang, H., Choi, K.-H., Korai, Y., Mochida, I., Appl. Catal. A: General, 252, 331 (2003). Technology Development Organization (NEDO), 28) Ma, X., Sakanishi, K., Mochida, I., Fuel, 73, 1667 (1994). Petroleum Energy Center (PEC) and Ministry 29) Ma, X., Sakanishi, K., Isoda, T., Nagao, S., Mochida, I., Energy

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004 163

& Fuels, 10, 91 (1996). 37, 3533 (1998). 30) Isoda, I., Nagao, S., Ma, X., Korai, Y., Mochida, I., Energy & 43) Daage, M., Chianelli, R. R., J. Catal., 149, 414 (1994). Fuels, 10, 1078 (1996). 44) Hensen, E. J. M., Kooyman, P. J., van der Meer, Y., van der 31) Ramírez, J., Contreras, R., Castillo, P., Klimova, T., Zárate, R., Kraan, A. M., de Beer, V. H. J., van Veen, J. A. R., van Santen, Luna, R., Appl. Catal. A: General, 197, 69 (2000). R. A., J. Catal., 199, 224 (2001). 32) Kunisada, N., Choi, K.-H., Korai, Y., Mochida, I., Preprint Fuel 45) Candia, R., Sorensen, O., Villadsen, J., Topsoe, N.-Y., Clausen, Chem. Div. ACS, 48, 502 (2003). B. S., Topsoe, H., Bull. Soc. Chim. Belg., 93, 763 (1987). 33) Kunisada, N., Choi, K.-H., Korai, Y., Mochida, I., Nakano, K., 46) Hensen, E. J. M., de Beer, V. H. J., van Veen, J. A. R., van Fuel, accepted, (2004). Santen, R. A., Catal. Lett., 84, 59 (2002). 34) Dzwigaj, S., Louis, C., Breysse, M., Cattenot, M., Bellière, V., 47) Reardon, J., Datye, A. K., Sault, A. G., J. Catal., 173, 145 Geantet, C., Vrinat, M., Blanchard, P., Payen, E., Inoue, S., (1998). Kudo, H., Yoshimura, Y., Appl. Catal. B: Environment., 41, 181 48) Inamura, K., Uchikawa, K., Matsuda, S., Akai, Y., Appl. Surf. (2003). Sci., 121-122, 468 (1997). 35) Lecrenay, E., Sakanishi, K., Nagamatsu, T., Mochida, I., 49) Cattaneo, R., Rota, F., Prins, R., J. Catal., 199, 318 (2001). Suzuka, T., Appl. Catal. B: Environment., 18, 325 (1998). 50) Hiroshima, T., Mochizuki, K., Honma, T., Shimizu, T., Yamada, 36) Kaneko, E. Y., Pulcinelli, S. H., da Silva, V. T., Santilli, C. V., M., Appl. Surf. Sci., 121, 433 (1997). Appl. Catal. A: General, 235, 71 (2002). 51) Shimizu, T., Hiroshima, K., Honma, T., Mochizuki, T., Yamada, 37) Pohal, C., Kameda, F., Hoshino, K., Yoshinaka, S., Segawa, K., M., Catal. Today, 45, 271 (1998). Catal. Today, 39, 21 (1997). 52) Hiroshima, K., Mochizuki, T., Honma, T., Shimizu, T., Yamada, 38) Ramírez, J., Cedeno, L., Busca, G., J. Catal., 184, 59 (1999). M., Appl. Surf. Sci., 121-122, 433 (1997). 39) Auer, E., Freund, A., Pietsch, J., Tacke, T., Appl. Catal. A: 53) Lauritsen, J. V., Helveg, S., Lagsgaard, E., Stensgaard, I., General, 173, 259 (1998). Clausen, B. S., Topsoe, H., Besenbacher, F., J. Catal., 197, 1 40) Whitehurst, D. D., Farag, H., Nagamatsu, T., Sakanihi, K., (2001). Mochida, I., Catal. Today, 45, 299 (1998). 54) Hensen, E. J. M., de Beer, V. H. J., van Veen, J. A. R., van 41) Farag, H., Whitehurst, D. D., Sakanishi, K., Mochida, I., Catal. Santen, R. A., J. Catal., 215, 353 (2003). Today, 50, 9 (1999). 55) Kunisada, N., Choi, K.-H., Korai, Y., Mochida, I., Appl. Catal. 42) Farag, H., Whitehurst, D. D., Mochida, I., Ind. Eng. Chem. Res., A: General, in press (2004).

……………………………………………………………………

要 旨

水素化脱硫・脱窒素プロセス

持田 勲，崔 基赫

九州大学先導物質化学研究所，816-8580 福岡県春日市春日公園 6-1

超低硫黄分のガソリンや軽油の生産プロセス開発の基盤を提 媒構造の詳細については議論しなかったが，水素化脱硫の機構 供するために，石油製品および中間精製留分の水素化脱硫を紹 および阻害性と関連する触媒機能について総説した。触媒劣化 介した。水素化処理における石油製品中の分子種基質の反応 および反応器設計についても簡潔に紹介した。深度脱硫達成の 性・選択性・阻害性について特に議論した。触媒開発のために ための新しいアプロ－チを提案した。 活性種・担体および担持法について最近の進展をまとめた。触

……………………………………………………………………

J. Jpn. Petrol. Inst., Vol. 47, No. 3, 2004