PreparationandEvaluationofSulﬁded NiMo γ Al2O ...953658/FULLTEXT01.pdf · surface area and on NiMo catalyst was prepared using γ-alumina with lower ... pore diﬀusion which prevent

Preparation and Evaluation of SulfidedNiMo/γ-Al2O3 Hydrotreating Catalysts

ALI KAMYAB

Master Thesis at NYNAS ABSupervisors: Johan Verendel

Moa Ziethén GranlundExaminer: Henrik Kusar

TRITA xxx yyyy-nn

iii

Abstract

Four nickel-molybdenum catalysts were synthesized on γ-alumina with highersurface area and on NiMo catalyst was prepared using γ-alumina with lowersurface area. Catalysts with higher-surface-area support were prepared bycoimpregnation, sequential impregnation and adding phosphorous. Theses cat-alysts were calcined at 500 ◦C. Effect of higher calcination temperature wasinvestigated by preparation of one catalyst calcined at 700 ◦C. Catalysts werethoroughly characterized via four characterization techniques. The hydrotreat-ing activity of three catalysts was carried out in a micro reactor at high pressureand three different temperature with Nynas vacuum middle distillate. Prior tothe test, sulfiding was carried out to activate the catalysts. Hydrotreated-oilsamples as products were analyzed to evaluate the activity and conversion ofthe catalyst. Also, the spent catalysts were characterized to evaluate the sur-face area characteristics and deactivation of catalysts. Addition of phosphorousto NiMo/γ-Al2O3 improved the interaction between the metals and the sup-port as well as reduced the coke formation as observed in scanning electronmicroscopy micrographs.

Contents

Contents iv

List of Figures v

List of Tables vi

1 Introduction 31.1 Setting the stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.2 Goal and scope of the thesis . . . . . . . . . . . . . . . . . . . . . . . 4

2 Hydrotreating process 5

3 Hydrotreating catalysts 73.1 Catalysts selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73.2 Catalysts characterization . . . . . . . . . . . . . . . . . . . . . . . . 8

3.2.1 Catalyst structure . . . . . . . . . . . . . . . . . . . . . . . . 83.2.2 Characterization techniques . . . . . . . . . . . . . . . . . . . 13

3.3 Catalysts preparation procedures . . . . . . . . . . . . . . . . . . . . 153.3.1 Impregnation . . . . . . . . . . . . . . . . . . . . . . . . . . . 153.3.2 Metal loading . . . . . . . . . . . . . . . . . . . . . . . . . . . 173.3.3 Calcination temperature . . . . . . . . . . . . . . . . . . . . . 183.3.4 Promoters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

4 Experimental 234.1 Catalyst preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . 234.2 Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244.3 Activity Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

5 Results and discussion 295.1 Catalysts characterization . . . . . . . . . . . . . . . . . . . . . . . . 295.2 Catalytic activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

6 Conclusions and further Work 45

Bibliography 47

iv

List of Figures

1.1 Simplified block flow diagram for Nynas refinery . . . . . . . . . . . . . 4

3.1 HDS activity as a function of pore size for different feeds. Adaptedfrom [1]First feed: Californian GO, Second feed: Californian VGO,and Third feed: Arabian HVGO. . . . . . . . . . . . . . . . . . . . . . . 8

3.2 Catalysts components and their key role to the hydroprocessing catalyst:a concept of acid-based supported catalyst. Adapted from [2]. . . . . . . 9

3.3 Effect of calcination temperature on HDS activity. Adapted from [3]. . . 193.4 Calcination temperature and promoter contents relation, and its impact

on activity. Adapted from [4] . . . . . . . . . . . . . . . . . . . . . . . . 20

4.1 Unit schematic flow diagram . . . . . . . . . . . . . . . . . . . . . . . . 254.2 Conditions for activity test for catalysts . . . . . . . . . . . . . . . . . . 264.3 Packing for the micro reactor . . . . . . . . . . . . . . . . . . . . . . . . 27

5.1 N2 adsorption-desorption isotherms . . . . . . . . . . . . . . . . . . . . . 305.2 XRD of fresh catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325.3 TPR profile of fresh catalysts . . . . . . . . . . . . . . . . . . . . . . . . 335.4 SEM micrographs for fresh catalysts . . . . . . . . . . . . . . . . . . . . 365.5 SEM micrographs for spent catalysts . . . . . . . . . . . . . . . . . . . . 375.6 Formation of whiskers in SEM micrographs for spent catalysts . . . . . 385.7 Catalysts Ageing Trend at Tbase and x bar . . . . . . . . . . . . . . . . . 395.8 Process yield for hydrotreated oils . . . . . . . . . . . . . . . . . . . . . 395.9 Aromatics, nitrogen and sulfur contents in products . . . . . . . . . . . 415.10 HDA, HDN AND HDS activity rate constants and conversions . . . . . 425.11 Color, density and viscosity in products . . . . . . . . . . . . . . . . . . 43

v

List of Tables

2.1 Feedstocks and Products for Hydrotreating Unit, adopted from [5]. . . . 5

3.1 Typical example for impregnation methods . . . . . . . . . . . . . . . . 16

4.1 Summary of prepared catalysts. . . . . . . . . . . . . . . . . . . . . . . . 24

5.1 Textural properties of support and catalysts. . . . . . . . . . . . . . . . 295.2 Compositions for catalysts obtained from SEM-EDX . . . . . . . . . . . 345.3 Products characteristics during a two-day process . . . . . . . . . . . . . 40

vi

Nomenclature

Ca % Aromatics contentCp % Paraffinic contentCn % Naphthenic contentP bar Partial pressureP0 bar Saturation pressurex − Ratio of Partial pressure to saturation pressureV ml Adsorbed volume at partial pressureVm ml Adsorbed volume at monolayer coveragec − ConstantB − Peak widthd − Crystallite sizeλ − X-ray wavelengthΘ − Diffraction angleκ − Shape constantLHSV h−1 Liquid hourly space velocitykHDA h−1 Activity rate constant for hydrodearomatizationkHDN h−1 Activity rate constant for hydrodenitrogenizationkHDS ppm−0.5h−1 Activity rate constant for hydrodesulfurizationAF wt.% Aromatics content in feed of the micro reactorAP wt.% Aromatics content in product of the micro reactorNF ppm Nitrogen content in feed of the micro reactorNP ppm Nitrogen content in product of the micro reactorSF ppm Sulfur content in feed of the micro reactorSP ppm Sulfur content in product of the micro reactorVsam ml Volume of sampleVF ml Volume of the consumed oil (as the feed)Tm

◦C Reduction temperature

1

Chapter 1

Introduction

1.1 Setting the stage

In recent years, environmental legislations on the sulfur content in transport fuelshas brought about the need for ultra low sulfur fuels. In addition, as a response tothe growing global energy demand as well as decreasing the quality of crude oils dueto the pattern of changing the sources into heavy and extra-heavy, interests in heavyand extra-heavy crude oils have been increasing. These factors have positionedhydrotreating process to be the center of improvement. This process covers both thepretreatment (heavy and extra-heavy crude oils hydrotreating before distillation)and post treatment (hydrotreating of products). Although hydrotreating processhas been applied for many years, it still requires more modifications to satisfy thenew demands. The main problems in hydroprocessing are (i) the limitations forpore diffusion which prevent the access of large molecules to the active sites, and(ii) rapid catalyst deactivation by coke and metal deposition. In order to tackle thesetwo problems at the same time, bi-functional catalysts are required. Such a catalystshould have a support with sufficient pore diameter to improve metal deposition andpore diffusion, enough acidity to efficiently crack the large molecules, and optimumdispersion of active metals [6].Nynas Refinery process is different from a typical oil refinery with production

of transport fuels as main products. Figure 1.1 illustrates a simplified block flowdiagram of Nynas Refinery. Vacuum distillation middle cuts are processed in hy-drotreating units to mainly reduce aromatics, nitrogen and sulfur contents to meetthe products specifications. Nynas products range includes bitumen, transformeroil, base oil and other oils such as oils for adhesives, printing inks, etc. In general,these products are characterized by their naphthenic hydrocarbons content. Themineral oil aromatic content plays an important role to meet the required productsspecifications. Sulfur content is not the main concern for this product. However,many products with different specifications are produced in this refinery and thepriority for each one may differ.Bearing above into consideration, proper catalysts for Nynas hydrotreating pro-

3

4 CHAPTER 1. INTRODUCTION

Figure 1.1: Simplified block flow diagram for Nynas refinery

cess should remove the aromatic, nitrogen and sulfur contents to satisfy productscharacteristics and, at the same time, be enough flexible for a range of feed. Inaddition, operating conditions in which the catalysts give the highest activity arevery important as they affect the energy consumption of the process. Since the mostfocus of researches is on fuel treatment, almost all the experiments reported in lit-eratures have concentrated on sulfur and nitrogen removal. Therefore, conductingan experiment to evaluate the behavior of different catalysts on Nynas various feed-stocks helps the company to have a better understanding of the catalysts behaviorand a cleverer selection approach in the market.

1.2 Goal and scope of the thesisThe aim is to provide an overview of the effect of hydrotreating catalysts at differentoperating conditions. This includes a literature review, evaluating and selectingof the most proper methods for catalyst preparation and modification, making anumber of catalysts, characterizing, and testing. However, due to time span of thisthesis, five catalysts were made and characterized, but only three of them weretested in Nynas micro-reactor with one feed.Catalysts preparation and characterization were done at KTH laboratories in

chemical engineering department, SEM-EDX at electron microscopy laboratory atKTH-Kista, and activity tests and products characterization are performed at thelaboratory of Process Technology department Nynas Refinery.

Chapter 2

Hydrotreating process

Hydrotreating is a multi functional process to reduce the content of sulfur, nitrogen,oxygen, metals and aromatics in products of a typical refinery as well as improvingcetane number, smoking point and the density. The feed of this process includes dif-ferent distillate range from naphtha to atmospheric residue. Increasing the amountof contaminants leads to more difficult and severer upgrading; i.e. heavier feeds leadsto increasing severity, and accordingly the consumption of hydrogen and catalyst.The aim of hydrotreating process is to catalytically convert distillation fractions tocommercial products which are adversely affected by the environmental legislationsfor clean products as well as increasing demand [2,7]. Table 2.1 shows some possiblefeeds and products for a typical hydrotreater unit [5]. Researches and companiesowning the know-how of hydrotreating processes have been putting their effort toinnovatively enhance the process and its catalysts [2, 7].

Table 2.1: Feedstocks and Products for Hydrotreating Unit, adopted from [5].

Feed ProductNaphtha Catalytic Reformer FeedStraight-run Light Gas Oil Kerosene, Jet FuelStraight-run Heavy Gas Oil Diesel FuelAtmospheric Residue Lube base Stock, Low Sulfur Lube Oil, RFCC FeedVacuum Gas Oil FCC Feed, Lube base StockVacuum Residue RFCC FeedFCC Light Cycle Oil Blend Stock for Diesel, Fuel OilFCC Heavy Cycle Oil Blend Stock for Fuel OilVisbreaker Gas Oil Blend Stock for Diesel, Fuel OilCoaker Gasoil FCC FeedDeasphalted Oil Lube base Stock, FCC Feed

Various processes for hydrotreating have been developed by different companieslike Shell, UOP, Axens, Halder Topsøe, Texaco, etc. [8]. Commercial units have em-

5

6 CHAPTER 2. HYDROTREATING PROCESS

ployed a broad ranges of operating conditions due to feed characteristics. Differenttype of reactors, i.e. fixed or moving bed, ebullated or slurry bed, needed to be se-lected to satisfy the process requirements. In addition, this process requires catalystto efficiently transfer hydrogen to reactant molecules. The process also generallyis affected by the design of the catalytic reactor and its internals in particular. Inorder to reach to a high-performance-hydrotreating unit, feed type requires to bematched with the reactor and catalyst which in turn resulted in minimizing thecatalyst deactivation rate [2, 7].

Chapter 3

Hydrotreating catalysts

3.1 Catalysts selection

Several parameters play important roles in catalyst selection for hydrotreating suchas catalyst life, activity toward side reactions (unwanted or desired), pressure drop,ease of activation, regeneration and price. Moreover, for certain applications, using acombination of different catalysts types in on reactor may be the optimum solution.In general, a detailed study of the specific situation is required to select catalyst [1].In hydrotreating catalysts the most common combinations of active elements are

CoMo, NiMo and NiW supported on alumina, silica-alumina, silica, zeolites, andmagnesia, with surface areas of 100 to 300 m2/g [1, 9]. Catalyst selection is basedon the application and desired activity/selectivity. CoMo catalysts are excellentfor HDS while showing less activity in HDN and HDA. On the other hand, NiMocatalysts show high activity in HDN and HDA. As a result, this type of catalystsare often preferred for treating unsaturated feeds. Selectivity of heteroatom removalversus hydrogenation can be increased for alumina-supported catalysts by additionof phosphorous which is described in Section 3.4.2. NiW catalysts, in comparisonwith CoMo and NiMo, have the highest activity for HDA at low hydrogen sulfidepartial pressures [1, 10].Diffusion in the catalyst pores in hydrotreating often influence reaction rates.

Thus, choosing proper catalyst particle size and shape, and pore system geometryis important. This is especially true in hydrotreating of heavier feeds, where thereactions may be limited by diffusion of reactants and products in and out the poresystem. Hence, for processing residue, small catalyst particles with relatively largepores which expose significant external surface area are preferred [1]. There is atendency in processing heavy feeds for bed plugging because of deposits (likely con-sisting of V and Ni sulfides, debris, iron compounds, coke and NaCl) on the exteriorof the catalyst particles. The particulates settle in the voids between the catalystparticles at the inlet of fixed-bed reactors. Thus, bed plugging is especially seriousat this point. In order to effectively tackle the bed-plugging problem, catalysts inthe form of large rings should be used or bed grading should be adopted [1].

7

8 CHAPTER 3. HYDROTREATING CATALYSTS

The effect of pore size on hydrotreating has been considered in several literaturereports (e.g. [1]) and an optimum pore size is reported [11]. Topsøe et al. [1] showedthat the optimum pore size increases with increasing pressure and particle size.Also, feed and process conditions can affect the optimum pore size. Smaller poresmake the activity become limited by diffusion, while larger pores give less activesurface area. Figure 3.1 illustrates these features for a light feed (the first one), aheavier feed (the second one) and a heavy feed (the third one).

Figure 3.1: HDS activity as a function of pore size for different feeds. Adaptedfrom [1]First feed: Californian GO, Second feed: Californian VGO,and Third feed: Arabian HVGO.

3.2 Catalysts characterization

3.2.1 Catalyst structureThe structure of hydrotreating catalysts consists of two main components: activesite and support. Stanislaus et al. [2] provided a general classification of catalystcomponents and their key elements as illustrated in Figure 3.2.

3.2. CATALYSTS CHARACTERIZATION 9

Figure 3.2: Catalysts components and their key role to the hydroprocessing catalyst:a concept of acid-based supported catalyst. Adapted from [2].

3.2.1.1 Active Site

It is commonly assumed that the catalytically active sites in either promoted or non-promoted hydrotreating catalyst are the molybdenum atoms at the surfaces of themolybdenum sulfide (MoS2) crystallites, with at least one sulfur vacancy at a siteto allow the reacting molecule to bond chemically to the molybdenum atom [1,12].These crystallites have a layer lattice, and the sulfur-sulfur interaction between suc-cessive MoS2 layers is weak [13]. Crystals grow as platelets with relatively largedimensions parallel to the basal sulfur planes as well as small dimensions perpen-dicular to the basal plane. It is suggested that the edges of the MoS2 platelets arebonded to the surface of the support by Mo-O-Al bonds [12,13].Coordinatively Unsaturated Sites (CUS) or sulfur ion vacancies are resulted from

removing the corner and edges of sulfur ions in MoS2 under hydroprocessing condi-tions [2]. Sulfur atoms in the basal planes of MoS2 are much more difficult to removethan sulfur atoms at edges and corners. Hence, exposed molybdenum atoms arepredominantly present at edges and corners resulting in catalysis occurrence atMoS2 edges and corners rather than on the basal plane. This also was confirmedby results of a study in which a single crystal of MoS2, with a high basal planeto edge surface area ratio, was found to have a low HDS activity [14]. Its activ-ity increased after the sulfur atoms were sputtered away from the basal plane andexposed to the molybdenum atoms. Another study showed that exposure of the


MoS2 nano-particles to hydrogen led to creation of sulfur vacancies. Such vacanciesat the MoS2 edge were supposed to be the active sites for HDS [15]. Thus, in caseof non-promoted catalyst, basal plane are inactive in molecules adsorption and arenot probably important in hydrotrating reactions [2].In case of promoted catalyst, metal sulfide, e.g. MoS2, is decorated by promoters,

e.g. Ni(Co). This gives more active CUS and accordingly improves the rate ofhydrogen activation [16]. The effect of adding promoters in increasing HDS andHDN [2] has been explained through different models which four of them have beenmore in center of attention: Monolayer Model [17–20], Intercalation Model [21,22],Contact Synergy Model [23–25], and Co-Mo-S Model [1,26]. As among these models,Co-Mo-S model has been broadly accepted [1, 2], a brief description is provided asfollows.Topsøe et al. [26–33] applied in-situ Mössbauer spectroscopy, extended X-ray ab-

sorption spectroscopy, and infrared spectroscopy studies to directly observe a Co-Mo-S phase in CoMo catalyst. This enabled them to provide a detailed structuraldescription of this catalyst and a new explanation for the promotion. The buildingblocks of Co-Mo(Ni)-S structures are small MoS2 nano-catalysts with cobalt (ornickel) promoter atoms [2]. These atoms are located at the edges of MoS2 layersin the same plane of molybdenum atoms [1,2, 26]. HDS activity was found to havea linear correlation with the relative amount of cobalt atoms present as Co-Mo-Sphase [2].The structure of Co-Mo-S is not a single, bulk phase with a fixed overall Co:Mo:S

stoichiometry. Instead, it should be regarded as a family of structures with a widerange of cobalt concentrations (from pure MoS2 up to essentially full coverage ofthe MoS2 edges by cobalt). The cobalt atoms in Co-Mo-S may not all have identicalproperties because of the effects of different edge-site geometries [34,35], Co-Co in-teractions [30,31], and changes in sulfur coordination [36]. Cobalt and neighboringmolybdenum atoms interactions have been evidenced by both magnetic suscepti-bility [1] and ESR [37] measurements. In this context, it was noticeable that thestandard molar free enthalpy calculation of formation of slabs of MoS2 indicatedthat the edge location (decoration) of promoters increased the slabs stability [1].Studies on the correlation between structure and activity of these catalyst sys-

tems resulted in the identification of two types of Co-Mo-S structures. One typeshows substantially higher activity than the other one. Based on the the low andhigh activity forms of Co-Mo-S, the terms Type I and Type II were chosen respec-tively [2]. Depending on preparation and activation parameters, type of support,presence of additives, metal loading, etc., both single and multiple slab Co-Mo-Sstructures have been observed. For the catalyst supported on alumina, Type I Co-Mo-S (i.e. single slab structures) interact strongly with the support, maybe viaMo-O-Al linkages at the edges [1]. According to Stanislaus et al. [2], the Type Istructures are incompletely sulfided and have some remaining Mo-O-Al linkages tothe support. The presence of these linkages is related to the interaction occurringin the calcination state between molybdenum and surface alumina hydroxyl groupswhich leads to oxygen bridged mono-layer type structure that are difficult to sulfide


thoroughly.In contrast, the interactions between Type-II Co-Mo-S (i.e. multiple slab form)

and the support are small and they are fully sulfided [1, 2]. In this type, the un-derlying MoS2 are less disperse. It consists of multiple slabs which are not linkedwith the support [2]. Other studies [38, 39] showed that the the degree of stakingin MoS2 and Co-Mo-S structures can be controlled by carefully controlling supportproperties and preparation parameters. Also, the formation of small stable singleslabs of MoS2 on alumina supports have been achieved. These have a high disper-sion of MoS2 which can accomodate more cobalt atoms to form higher active singleslab Type II Co-Mo-S structures [2].In summary, Type II structures are higher activity forms of Co-Mo-S than Type

I. However, in 2009, Okamoto et al. [40,41] reported the results of their experimenton preparation of CoMo/SiO2 catalyst sulfided with 10 % H2S/He and formingextraordinary active Co-Mo-S (higher than Type II) on SiO2 regardless of the pre-cursor of Mo. They call it Type III Co-Mo-S structures. It should be noted thatcatalytic activity is mostly connected to the presence of promoters in Co-Mo-S. Thisstructure is also observed to be true for type of promoted catalysts i.e. Fe-Mo-S,Ni-Mo-S, Co-W-S Ni-W-S and Fe-W-S [1].

3.2.1.2 Support

Active compounds, i.e. mixed sulfides of molybdenum and cobalt or nickel asNi(Co)-Mo-S phases, in hydrotreating catalyst are supported on a carrier (support)which generally affect catalytic properties for both promoted and non-promotedtypes [1, 2]. The carrier provides mechanical strength to the catalyst as well ashigh surface area to maximize active phase dispersion [2]. Generally alumina exertsstronger interactions with the active phases compared to silica, carbon, or tita-nia for instance. The strong support interactions are very important as they areresponsible for the fact that highly dispersed MoS2 structures can easily be pre-pared and remain stable during operation. However, these strong interactions inalumina-supported catalysts have negative sides as they are probably responsiblefor the formation of the relatively less active Type I Ni(Co)-Mo-S. For some othersupports (e.g. silica), the support and the active phases interactions are quite weak.This mainly leads to influence on the activity via changes in dispersion and phasedistribution [1].The support effect has been studied from different angels. Topsøe et al. [1, 42]

pointed out the dependency of catalytic activity upon form, orientation and growthof MoS2 species on the support site. Sakashita et al. [43–45] studied the effects ofsurface orientation and crystallinity of the support. They showed the dependencyof the catalytic activity and selectivity of MoS2 catalysts on the microstructures ofMoS2 clusters that were affected by the surface orientation of gamma alumina. Theyconcluded that the gamma-alumina support surface properties are important fac-tors in controlling the catalytic functionality of MoS2-based catalysts. Also, basedon their results, controlling the microstructure of MoS2 clusters by tailoring the sur-


face orientation of alumina supports is possible. In other experiments [46, 47] theimpact of pH of the impregnating solution, alumina phase, particle size and texturalproperties of Al2O3 supports on HDS, HDN amd HDA were studied. Their resultsshowed that both the textural properties and hydrotreating activity of the nickel-molybdenum catalysts supported on two alumina phases are noticeably differentwhile evaluating with a real feedstock. The behavioral differences of the catalystsappear to have their origin in the different support crystal size that was used duringthe catalysts and alumina phase preparation. This could modify the concentrationand distribution of the surface hydroxyl groups and enhance the catalytic activity.They also reported that the link between particle size of the alumina support, NiMooxide catalysts, and NiMoS catalyst was important to determine the changes of theactive sites after calcining and sulfiding. The results revealed that the catalyticactivity of HDS, HDN and HDA depends on the particle size of the support andon the NiMo species metallic particle size. In addition, the catalyst with a smallerparticle size gave the maximum reaction rates for all three processes.The effect of the surface characteristics of support on its interaction with metal and

its correlation with hydrodesulfurization activity and CoMo selectivity by severalcharacterization methods have been investigated [48]. It showed that alumina sup-port with higher crystallinity and fewer hydroxyl groups resulted in a more moderateinteraction with metal. This enhanced Mo sulfidation and caused a higher stackingnumber and longer average slab length for the CoMo catalyst. Also, it stimulatedmore Co-Mo-S structures, giving rise to higher HDS activity and selectivity. In anew approach in the evaluation of the support effect, Ninh et al. [49] proposed apreparation method that might be more efficient than the classic co-impregnationmethod. They selected the interaction of acetylacetonate complex directly ontosupported MoS2. Then supported MoS2 was prepared on different type of supportand Ni(ACAC)2 used to achieve MoS2 promotion in refluxing methanol. The resultsshowed that the Ni-Mo-S sites on different type of catalysts had almost the sameintrinsic activities, suggesting that active sites quality would not be very affectedby the support nature.Among the available developed supports for hydrotreating catalyst, gamma alu-

mina (γ-Al2O3) has been widely applied in hydrotreating [1,2,16]. This broad appli-cation is due to its reasonably high surface area and porosity, high thermal stability,excellent mechanical strength, containing acidic and basic sites, very good morphol-ogy and low cost [1,2,50,51]. In addition, as mentioned by Breysse et al. [52], broadknowledge availability of γ-alumina is another reason for its widely usage. Nev-ertheless, several studies have been carried out on other types of catalyst such ascarbon, silica, silica-alumina, zeolites, titanium, zirconia, and magnesia. Topsøe etal. [1] reviewed the work’s regarding different supports as well as Breysse et al [52]who provided an overview on the various support interactions for hydrotreating cat-alysts. The latter showed that many different aspects such as electronic propertiesmodifications or the active site morphology, bifunctional reactions with acidic sites,absence or presence of linkage between the support and the sulfide site need to beconsidered in study of the support effect.


In recent years, hydrothermal treatment (HTT) has been introduced as an alterna-tive method for the modification of γ-Al2O3 support [50,53–58]. Stanislaus et al. [54]conducted a series of experiments to develop the understanding of the mechanismof pore widening in γ-Al2O3 supports under hydrothermal conditions. Their resultsshowed that pore enlargement occurred because of rehydration of γ-Al2O3 and itstransformation to well-crystallized boehmite. A wide pore NiMo/γ-Al2O3 preparedby the hydrotreating procedure showed exceptionally high activity for hydrodemet-allization and asphaltenes conversion reactions in vacuum residue hydroprocessing.Xiang et al. [55] studied the impact of HTT on Ni/Al2O3 catalyst and reported

that the method was effective to reach high activity. Another study [57] was carriedout on the same catalyst regarding structureal and textural evolution of Ni/Al2O3catalyst under HTT. Li et al. [50] performed HTT on γ-Al2O3 at 140 ◦C for 1-24hours. This resulted in the formation of plate-like γ-AlOOH nano-crystallites withdifferent morphologies, which linked closely to the surface acidity of the correspond-ing support. Reaction time rise in the period of 1-2 hours led to increasing of thespecific surface area, OH surface and γ-Al2O3 acidity, especially the weak acidity.However, extending hydrothermal reaction time brought about the decrease of spe-cific surface area, OH surface and total acidity of γ-Al2O3 because of the growth ofthe surface crystallites.

3.2.2 Characterization techniques

Process optimization and development require a better understanding of catalysts.In addition, the emergence of new material with catalytic properties and peculiarcharacteristics, and development in technology and computer have led to a consid-erable growth in techniques for catalyst characterization [59]. These techniques helpto understand the relationship among physical, chemical and catalytic properties;clarifying deactivation causes, designing regeneration procedures, and selecting cat-alyst properties to minimize such deactivation. Other beneficial of these techniquesincludes establishing physical and chemical properties for catalyst marketing, re-actor design, modeling, and process optimization; and ensuring quality control incatalyst manufacturing [60].Morphology and physical properties. Important properties of catalysts are sur-face area, pore size and pore volume since they are responsible to define the measureof catalysts internal surface available to accommodate active sites, accessibility ofactive sites to reactants and the extent to which transport of products from thecatalyst surface to the bulk fluid is facilitated [60]. For most heterogeneous cata-lysts, porosity occurs through the preparation methods, e.g. by drying or calciningprecipitates of hydrous oxides [59,60]. The size and number of pores determine theinternal surface area which is usually flavored to be high (high density of small poresizes) to maximize the dispersion [60]. The pore sizes classification is usually con-sidered as micropores (φ < 2 nm), mesopores (2 < φ < 50 nm) and macropores (50nm < φ). Morphological parameters play an important role to understand the evo-lution of catalysts during preparation step. This is helpful to enhance preparation


methods to reach the desired results [59].Nitrogen adsorption at its boiling point (77 K) is the most common technique for

determining catalyst morphology [59, 60]. This technique determines the absorbedvolume of N2 versus its relative pressure (i.e. adsorption isotherm). Depending onwhich suitable evaluation method (e.g BET, t-plot, BJH, etc.) the data is analyzed,different properties such as catalyst total surface area, pore size and volume aredetermined [59].Brunauer, Emmett and Teller (BET) is a very common evaluation method which

describes the relationship between the volume absorbed at a given partial pressureand the volume adsorbed at monolayer coverage, Equation 3.1 in which x = P/P0,P is the N2 partial pressure, P0 the saturation pressure at the temperature of theexperiment, V the volume adsorbed at P , Vm the volume adsorbed at monolayercoverage, and c is a constant [60].

x

V (1 − x) = (1 + (c− 1)x)(cVm) (3.1)

Electron microscopy is a powerful tool that frequently used for direct observingof shape, size, chemical structure, surface texture, and crystallite size, over a broadrange of size resolution ranging from atomic dimensions to several hundred microns.Two quantitative methods are Scanning Electron Microscopy (SEM) and Transmis-sion Electron Microscopy (TEM), which provides a direct view of the catalyst understudy [59,60].The electron beam focused on the sample is scanned in SEM by a set of de-

flection coils. TEM involves electron analysis focused and transmitted through athin sample on a conducting grid to form high resolution images of the specimen.TEM is used to determine crystallite size and size distribution for supported metals,e.g. for sintering of supported Ni and Pt catalysts. The modern High ResolutionTransmission Electron Microscope (HRTEM) is a useful tool to study nanostructureof nanophase materials with the possibility of seeing molecular layers and surfacelayers [60].Surface properties Probably, the first characterization technique applied to cat-alysts was infrared spectroscopy (IR) which continues to be one of the most widelyused in studies on catalysts and surfaces [60]. This technique can be used to deter-mine adsorbed species and/or adsorbed reaction intermediates and their structureson well-dispersed catalysts and surfaces, and give sign of active sites chemistry oncatalyst surfaces during preparation and reaction, i.e. acidity, oxidation state, andsit molecularity [59,60]. Nevertheless, IR spectroscopy suffers from two limitations:low frequency region cannot be used due to the skeleton vibrations of the solid, andnon-resistance of most IR transparent materials to water or other liquids [59].Raman Spectroscopy (RS) analyzes the intensity of Raman scattering of monochro-

matic light as a function of the scattered light frequency [60]. The same detailedmolecular structure information as IR spectroscopy obtained from RS; however, thelimitations in IR are covered in RS as it allows detection of low frequency vibra-tions [59,60].

3.3. CATALYSTS PREPARATION PROCEDURES 15

Temperature Program Desorption (TPD) is applied to study various adsorptionstates, their binding energies, their surface concentrations, and their desorption ki-netics. These studies are carried out through measurement of the rate of desorptionadsorbed molecules as a function of linear temperature ramp [60].Since most catalysts are crystalline solids, diffraction of an X-ray beam as a func-

tion of the angle of the incident beam results in evaluation of the crystalline nature,their concentration in the solid and their size [59,60]. X-Ray Diffraction (XRD) alsocan be applied for both qualitative and quantitative chemical phase analyses [60].In an XRD diffractogram, peak width is inversely related to the crystallite size of agiven phase. Thus the width of the peak can be related to the crystallite size. Basedon Scherrer equation 3.2, the width, B, relates to half-peak height of an XRD linebecause of a specific crystalline plane to the size of the crystallites, d. λ is the X-raywavelength, Θ the diffraction angle, and κ is the shape constant usually equals to1 (assumption for spherical particle).

B = κλ

dCosΘ (3.2)

Nuclear Magnetic Resonance (NMR) can be useful in study of the bonding andmobility of different kinds of adsorbed hydrogen species on noble metal catalysts,carbon deposits structure responsible for poisoning of catalysts during CO hy-drogenation, kinetics of exchange reactions between physisorbed and chemisorbedmolecules, reactions of short-lived transition metal complexes on oxide catalysts,and structures of adsorbed molecules [60].

3.3 Catalysts preparation proceduresDifferent preparation methods have been applied e.g. molecular designed disper-sion method [61], sol-gel method [62], unconventional slurry impregnation or sol-vent assisted spreading method [63], sequential and simultaneous equilibrium de-position filtration (EDF) technique, successive or simultaneous dry impregnations,successive or simultaneous wet impregnations [64], equilibrium adsorption (EA)method [65,66], equilibrium deposition filtration (EDF) [67], sequential and simul-taneous incipient wetness (pore filling [1]) impregnation [68–71]. Based on the liter-ature review of this study, the most common method is impregnation which will bedescribed hereafter. Catalytic activity and selectivity, and catalyst structure andmorphology are dependent on preparation procedure [1]. Such dependency mainlyinclude the effect of impregnation procedure, metal loading, calcination temperatureand activation [1].

3.3.1 ImpregnationImpregnation deposits the catalyst active elements, i.e. Mo and Ni(Co), on thesurface of the support. Therefore, impregnation has a key role in impacting on theactive phase structure and dispersion [2]. Hence, modification of the impregnation


procedures has been targeted in order to prepare high-active catalysts. A briefexplanation of different impregnation methods is provided as follows.Successive / one-batch − Sequential / Coimpregnation. Metals (e.g. Moand Ni) can be added to the support in two ways. An aqueous solution of Ni (aspromoter) and Mo is added to the support (e.g. γ-Al2O3). This method is calledcoimpregnation (simultaneous impregnation). In the other method (sequential), anaqueous solution of each metal is prepared separately and added to the supportapart from each other. Either of these methods also can be done in two differentways. In coimpregnation, the metals solution can be added at once which called onebatch. In contrast, in successive method, first an adequate amount of the solution(e.g. based on the pore volume of the support) is added to the support. Afterdrying, another batch of the solution is added and dried. This step is repeatedas many times as required to reach to the desired concentration of metals in thecatalyst. The same concept is also applicable for the sequential method. To give abetter understanding of these methodes, a typical example is given in Table 3.1.

Table 3.1: Typical example for impregnation methods

Coimpregnation SequentialNi and Mo simultaneously Ni first, Mo next Mo first, Ni next

One batch

Aqueous solution of Niand Mo is added toγ-Al2O3 following bydrying.

Aqueous solution of Niis added to γ-Al2O3following by drying.Then aqueous solutionof Mo is added to Ni/ γ-Al2O3 following bydrying.

Aqueous solution of Mois added to γ-Al2O3following by drying.Then aqueous solutionof Ni is added to Mo/ γ-Al2O3 following bydrying.

Successive

Step 1 Adequate amount ofaqueous solution of Niand Mo is added toγ-Al2O3 following bydrying.

Adequate amount ofaqueous solution of Niis added to Al2O3 fol-lowing by drying.

Adequate amount ofaqueous solution ofMo is added to Ni /γ-Al2O3 following bydrying.

Step 2 Adequate amount ofaqueous solution of Niand Mo is added toNiMo / γ-Al2O3 fol-lowing by drying.

Adequate amount ofaqueous solution ofNi is added to Ni /γ-Al2O3 following bydrying.

Adequate amount ofaqueous solution of Niis added to NiMo /γ-Al2O3 following bydrying.

Step 3 Step 2 is repeated un-til reachinng the finalconcentration of Ni andMo in the catalyst.

Step 2 is repeated untilreachinng the final con-centration of Ni in thecatalyst.

Step 2 is repeated untilreachinng the final con-centration of Mo in thecatalyst.

The order of incorporation of the active phases strongly influences the dispersionand distribution of the metal phases and their textural properties [70]. Typically,introducing molybdenum before promoter results in more active catalyst since dif-ferent precursor structures formed [1, 70]. In this case, the molybdenum disper-sion improves during the second step of nickel impregnation, and nickel is moreaccessible to form some Ni-Mo-O interaction species after molybdenum partial hy-drolysis, leading to increased activity [70]. On the contrary, promoter introductionat first leading to predominant formation of the inactive Ni3(Co3)O4 precursors.


Upon addition of Mo, these precursors are not transformed to Ni(Co)-Mo-O struc-tures [31, 72]. In addition, the results of a study by Salerno et al. [70] showed thatthe incorporation of Ni before Mo leads to badly dispersed and separated activenickel and molybdenum phases. Its reason is that the major part of Ni occupy ionexchange positions into the octahedral layers of the clay, being non-reduced and sul-fided. That is why they observed poor HDS and HDN activities. However, a highercatalytic activity was reported [70] obtained from preparing the NiMo catalysts byco-impregnation. Simultaneous presence of nickel favors even more the dispersionof molybdenum and the formation of Ni-Mo-O interaction phase on the supportsurface, which is the catalyst precursor of the Ni-Mo-S active phase for hydrotrat-ing reactions. Significant enhancement in catalyst activity is reached by addingpromoter in successive steps instead of in one batch [71, 73] due to a reduction information of unwanted Ni3(Co3)O4 [1].

3.3.2 Metal loading

The structures and their relative proportion in sulfided both promoted and non-promoted molybdenum catalysts depends on the loading of both molybdenum andpromoter (cobalt or nickel). Metals loading is therefore one of the essential param-eters in optimizing commercial hydrotreating catalysts. In industrial applications,the metal loadings are usually governed by the desire to achieve as high activity aspossible with as small amounts of the expensive metals as possible. Hence, wastingmetals in undesirable low activity phases should be avoided. Topsøe et al. [1] in-troduced typical amount of molybdenum loading as about 8 to 15 wt. % which iscorresponded to approximately a monolayer coverage for an alumina support witha typical surface area of about 250 m2/g. If the molybdenum exceeds this level oralumina with low surface area is used, separate MoO3 entities are formed. This re-sults in bulk-like MoS2 phases upon sulfiding. The promoter loading depends on themolybdenum level and type of hydrotreating application. Thus, optimized loadingis obtained from experience. Topsøe et al. stated that the promoter concentrationusually amounts to about 1 to 5 wt. %, giving a promoter-molybdenum atomicratio of 0.1 to 1.0. As they concluded from the results of other works using differentcharacterization techniques to study this subject, catalysts with low molybdenumloading are more difficult to be sulfided than high-loading catalysts. Topsøe etal. suggested that it is possible that decreasing molybdenum loading increases therelative proportion of oxygen bonds between MoS2 species and the alumina sur-face [1]. Their suggestion is in line with results of the XPS studies of Li et al. [74]indicating the presence of Mo5

+ in a significantly lower concentration in additionto Mo4

+ (MoS2). Topsøe et al. [1] also mentioned that the alumina coverage bymolybdenum species increases linearly with the molybdenum loading up to a certainlimit. Above this limit, the coverage is leveled off suggesting growth of bulk-likeMoS2 structures. Further rise in molybdenum loading decrease the edge dispersionof the MoS2 and therby the activity. Consequently, the ratios of promoter oxideto molybdenum and oxygen to molybdenum decreased (as a result of a decrease in


MoS2 edge dispersion).Using high promoter concentrations gives a tendency of formation of separate

promoter phases (e.g. Co3O4) [1]. Topsøe et al. [33] conducted IR studies of NOadsorbed on sulfided NiMo/Al2O3 and CoMo/Al2O3 catalysts and showed that anincrease in the intensity of the bands associated with NO adsorption on promoteratoms. The formation of increasing amounts of Ni(Co)-Mo-S was also obvious by adecrease in the intensity of the bands associated with adsorption on molybdenumedge atoms because of covering/blocking of molybdenum edge sites.

3.3.3 Calcination temperature

The effect of increasing calcination temperature above 500-600 ◦C showed a de-crease in catalytic activity [3, 75, 76]. Stanislaus and Al-Dolama [77] showed thathigh-temperature calcination leads to volatilization of molybdenum salts. Resultsof calcining NiMo/Al2O3 at 800 and 900 ◦C resulted in 50% and 85% drop in molyb-denum content, respectively. Calcining CoMo/Al2O3 at the same temperatures pro-duced molybdenum loss of 27% and 43% respectively. In another study, Stanislauset al. [3] investigated what changes in the catalysts occurred when high calcinationtemperature was applied and if these changes were advantageous to hydrotreatingactivity. Studies (e.g. [78, 79]) have suggested that the catalytically active phasein NiMo/γ-Al2O3 and CoMo/γ-Al2O3 was produced from an intimate mixture ofoctahedrally coordinated promoter with octahedrally coordinated molybdenum at-tached to the surface. Based on these studies, the amounts of these species dependedon catalyst loading and on calcination conditions. According to Zingg et al. [78],the most of the octahedrally coordinated species were produced at calcination tem-perature of 500 ◦C with a loading of about 16% MoO3. While increasing the tem-perature to 700 ◦C decreased the surface area. The relative amount of tetrahedrallycoordinated molybdenum and the promoter increased [78, 79] and Al2(MoO4)3 (aninteraction species between MoO3 and the surface) was formed [78,80].Volatilization of molybdenum and an increase in the effective promoter to molyb-

denum ratio is reported at clacinatin temperatures above 700 ◦C. Both the in-creased temperature and the reduced molybdenum content favored the formationof tetrahedrally coordinated molybdenum, tetrahedrally coordinated promoter, andaluminates. Also, it resulted in more diffusion of the promoter into the bulk [78].Consequently, the activity declines rapidly as sown in Figure 3.3 [3]. The varyingchemistry of nickel and cobalt induced changes at different calcination temperature.Greegor et al. [79] showed that nickel had a higher octahedral site preference thancobalt but both octahedrally and tetrahedrally coordinated nickel are formed at lownickel loading. Moné also introduced nickel aluminate as a partly inverse spinel inwhich the nickel ions occupied both tetrahedral and octahedral holes, opposite tocobalt aluminate, which is a normal spinel [81, 82]. Nickel aluminate was found tobe formed at temperatures as low as 650 ◦C [79, 82, 83]. Therefore, NiO amountavailable to form an intimate mixture with molybdate was reduced and, above cal-cinatin temperature of 650 ◦C, the behavior of the catalyst could be expected to


Figure 3.3: Effect of calcination temperature on HDS activity. Adapted from [3].

be closer to that of Mo/γ-Al2O3 [78]. In line with this, the molybdenum volatiliza-tion was greater [77], the molybdenum dispersion was less and polymolybdates wereobserved on the surface [78] by increasing the calcination temperature.Scheffer et al. [4] also showed the effect of increasing the calcination temperature

on the activity was dependent on the promoter content (Figure 3.4). Their resultsrevealed that the effect of a change in calcination temperature depended stronglyon the promoter content. They showed that as the calcination temperature wasraised, a sharp drop in activity for catalysts with a low Co content was observed,but a slight rise in activity was seen at high cobalt content. At intermediate Coloadings the impact of increasing the calcination temperature was not so strong.

3.3.4 Promoters

Adding a material as a third elements to hydrotreating catalysts are investigated inmany studies as endeavors to reach higher active catalysts or selectivity [1]. Someworks performed on catalysts supported by γ-Al2O3 showed that adding some ma-terials such as fluoride [84], boron [9, 85–87], phosphorus [88–95], etc. can increasethe surface acidity, resulting in an increase in the hydrotreating activities. Thereason for this improvement in catalytic activities may be resulted from the vari-ation in metal-support interactions [2, 96, 97]. This in turn, may affect the mor-phology and dispersion of the active components. The support and impregnated


(a) Rate of thiophene HDS of CoMo/Al2O3catalyst calcined at different temperatures.

(b) HDS turn-over numbersof CoMo/Al2O3 catalyst

Figure 3.4: Calcination temperature and promoter contents relation, and its impacton activity. Adapted from [4]

metals interactions strongly affect both the active phase formation and the disper-sion on the surface of the support. Recently, the most common material as modifierin hydrotreating process has been phosphorous [1, 2]. As reported by literatures,phosphorous doping modifies both hydrodenitrogenation [85,98,99] and hydrodesul-furization [100–102]. Nevertheless, when it comes to NiMo/γ-Al2O3, phosphorouspromotes specially the activity of HDN rather than HDS [2,85,89,103,104]. In ad-dition, it improves dispersion of active metal salts, reduce the interaction betweenoxides of Mo and Ni as precursors, modifies the acid sites or forms new active phase,increases MoS2 stacking and forms bigger MoS2 slabs [2, 85,89,105–107].Hence, study of the effect of phosphorous was considered in the scope of this thesis.

As different preparation methods have been suggested in literature reports, in orderto use the best method, some of them are reviewed as follows:Lewis et al. [95] studied the effect of different phosphorous contents in three differ-

ent procedures of preparation of NiMo/γ-Al2O3 on gas oil hydroprocessing. In thefirst series, they prepared the catalysts through one-batch coimpregnation. For theother two series, the catalysts were prepared based on sequential impregnation. Theorder of addition was ammonium heptamolybidate followed by nickel nitrate hex-ahydrate aqueous solution. The difference between the second and the third serieswas in sequence of phosphoric acid addition which was introduced to the supportprior to the matals.Ferdous et al. [9] firstly prepared a series of NiMo/γ-Al2O3 with different nickel

and molybdenum concentrations using incipient wetness co-impregnation method


and extruded NiMo/γ-Al2O3. Zhou et al. [88] prepared NiMoP catalysts with differ-ent phosphorous content using the same method but with different material (MoO3powder, nickel carbonate tetrahydrate and phosphoric acid). Xiang et al. [91] pre-pared different catalysts regarding the amount of phosphorous content through suc-cessive incipient wetness impregnation. Rayo et al. [108] prepared both possibleP-doped: on support and on NiMo using simultaneous incipient wetness impregna-tion.

Chapter 4

Experimental

4.1 Catalyst preparation

Based on experimental results reported in literature, five catalysts were made. Twodifferent commercial γ-Al2O3 supports with different surface areas were used. Thefirst one with smaller surface area was Sasol Puralox NGa150 in powder form andthe other one was Saint Gobain as pellet.In order to prepare 20 g of each catalyst, 5.520 g ammonium heptamolybdate (Alfa

Aesar (NH4)6Mo7O24 · 4 H2O) and 3.468 g of nickel nitrate (Aldrich Ni(NO3)2 · 6 H2O)was dissolved in ultra pure water. The target was to obtain a final content of 3 wt.%Ni and 15.0 wt.% Mo. To prepare NiMo/γ-Al2O3

150 catalyst via co-impregnationmethod, this solution was dropped on 16.30 g of support in three steps. After eachstep the catalyst was dried at 120 ◦C for 4 hours, and after the last step, calcinedat 500 ◦C for 4 hours.Alumina extruder was crushed and the above procedure was followed to prepare

40 g of the catalyst based on Saint Gobain alumina, NiMo/γ-Al2O3. However,only 20 g of this catalyst was calcined at 500 ◦C for 4 hours. The other half ofNiMo/γ-Al2O3 was calcined at 700 ◦C for 4 hours and called NiMo/γ-Al2O3

700.The fourth catalyst, NiMo/γ-Al2O3

SEQ, was prepared through sequential impreg-nation by dropping a solution of 5.520 g ammonium heptamolybdate in adequateamount of water in two steps. Between each step the catalyst was dried at 120 ◦Cfor 4 hours. The next sequence was dropping a solution of 3.468 g of nickel nitratein adequate amount of water on the catalyst in two steps while in between it driedas the previous sequence condition and finally calcined at 500 ◦C for 4 hours.As discussed in section 3.3.4, phosphorus, as a secondary promoter in NiMo/γ-

Al2O3 catalysts, enhances HDS and HDN [9]. For this reason, the last catalyst wasmade by adding phosphorous. Based on the result of Rayo et al. [108] (see also 3.3.4),adding phosphorous to the support was selected to make the last catalyst. 2.531 g ofH3PO4 was dropped on 15.50 g of Saint Gobain alumina and dried and calcined at120 ◦C for 4 hours and 500 ◦C for 4 hours repectively. Same procedure as mentionedabove for the preparation of NiMo/γ-Al2O3 was followed and this catalyst named

23

24 CHAPTER 4. EXPERIMENTAL

NiMo/Pγ-Al2O3. The target was to have a concentration of phosphorous of 4 % inthe final catalyst while keeping the same nickel and molybdenum concentrations inother catalysts (5 % and 15 % respectively).A summary of prepared catalysts are given in Table 4.1. Only three of these five

catalysts were tested in micro-reactor at Nynas (NiMo/γ-Al2O3, NiMo/Pγ-Al2O3and NiMo/γ-Al2O3

150); hence, these three catalyst were pelletized, crashed andsieved to obtain particles with the range of 0.25 to 0.5 mm to avoid pressure dropin the reactor.s

Table 4.1: Summary of prepared catalysts.

Catalyst Support Preparation Calcinationmethod temperature, ◦C

NiMo/γ-Al2O3 Saint Gobain Coimpregnation 500NiMo/Pγ-Al2O3 Saint Gobain Coimpregnation 500NiMo/γ-Al2O3

150 Sasol Coimpregnation 500NiMo/γ-Al2O3

700 Saint Gobain Coimpregnation 700NiMo/γ-Al2O3

SEQ Saint Gobain Sequential impregnation 500

4.2 Characterization

BET surface area (BET) of the oxide catalysts were measured in a MicromeriticsASAP 2000 equipment. The supports and catalysts were degassed at 473 K priorto the adsorption. X-ray powder diffraction was carried out in a Siemens D5000diffractometer from 10 to 90 2θ at 2.5 degree−1 scan rate. Temperature ProgrammedReduction was conducted for catalysts using Micromeritics AutoChem 2910 up to800 ◦C.Spent catalysts were washed with gasoline and dried at room temperature for 24

hours. Then they were dried at 60 ◦C for 24 hours. The catalysts were sievedto separate SiC; however, NiMo/γ-Al2O3

150 was mixed with sand of the same sizerange. That was why separation of the catalyst from sand was not done. BET wasperformed for NiMo/γ-Al2O3 and NiMo/γ-Al2O3 while SEM was conducted for allspent catalysts.The samples of hydrotreated oils were collected and analyzed. Samples were bub-

bled by nitrogen and filtered prior to characterization. Aromatic, paraffinic andnaphtenic contents were analyzed by ThermoFisher. Sulfur and nitrogen contentswere measured by Antek 9000NS analyzer. Mettler Toledo DM40-RX40-SC30 den-sity meter was applied to measure the density of products using 15 ◦C ASTM D4052method. Products kinematic viscosity were determined by Herzog Multi-RangeViscometer (HVM 472). The color of products measured by spectrophotometriccolorimeters Lovibond PFX195 Tintometer with ASTM D156.

4.3. ACTIVITY TESTS 25

Figure 4.1: Unit schematic flow diagram

4.3 Activity TestsThe measurements of the catalytic activity for the three above mentioned catalystswere carried out under steady-state operation in a high pressure fixed-bed up-flowmicro reactor (Swagelok). A detailed scheme of this unit is shown in Figure 4.1 Thereactor was made of stinless steel tube with the total length and internal diameterof 275 and 10 mm. The period of each test was about 15 days including waitingtime for steady state (Figure 4.2).

26 CHAPTER 4. EXPERIMENTAL

Figure 4.2: Conditions for activity test for catalysts

For reactor loading, 5 ml of each catalyst was diluted with 4 ml silicon carbide ofthe mesh size of 0.2-0.5 mm and placed in the center of the reactor. The remainvolume of the reactor in both sides were identically filled as illustrated in Figure 4.3.After the reactor was purged with nitrogen and leak tested at room temperature,the catalyst was sulfided in situ using the method presented by Eijsbouts et al.[39]. Sulfiding was conducted for 19 hours by a presulfiding oil in presence ofhydrogen.This oil stream containing the required amount of di-tert-butyl disulfideto give a final composition of 3.7 wt.% sulfur in the oil. The reaction was carriedout at a pressure of 45 bar with a liquid hourly space velocity (LHSV) of 3.0 h−1

and H2/oil ratio of 200 at 150 ◦C for 3 hours then at 320 ◦C for 16 hours.After suldfiding period, the feed (Nynas vacuum distillation) was introduced to

the reactor and its temperature and pressure were set to Tbase and x bar. H2/feedratio and LHSV were considered z and y h−1 respectively. Ageing took 6 days and3-4 samples taken for analyzing. Activity of the catalysts were investigated at threetemperature of Tbase+20, Tbase and Tbase-20 for two days for each. After 12 hoursof changing the temperature, 1-2 samples were taken for analysis.The rate constant for HDA, HDN and HDS were determined by using Equations

4.1 to 4.3 [46,90]. kHDA and kHDN are the pseudo first order HDA and HDN kineticconstant(h−1), respectively, and kHDS is the pseudo 1.5 order HDS kinetic constant(ppm−0.5h−1). A, N and S in these equations stand for aromatics (wt.%), nitrogenand sulfur contents (ppm), respectively. The subscripts F and P represent feed andproduct, respectively. LHSV is the liquid hourly space velocity (h−1) and n is theorder of reaction.

kHDA = (LHSV )ln(AFAP

) (4.1)

4.3. ACTIVITY TESTS 27

Figure 4.3: Packing for the micro reactor

kHDN = (LHSV )ln(NF

NP) (4.2)

kHDS = 0.2LHSV ( 1S0.5P

− 1S0.5F

) (4.3)

Equation 4.4 to 4.6 was used to calculate conversions (%) and the yield of theprocess is calculated based on the total oil feed on-stream and the amount of theproduct as shown in Equation 4.7 where Vsam is the sample volume (ml) and VF isthe volume of the consumed feed (ml).

Conversion = [1 − APAF

].100 (4.4)

Conversion = [1 − NP

NF].100 (4.5)

Conversion = [1 − SPSF

].100 (4.6)

Y ield = [VsamVF

].100 (4.7)

Chapter 5

Results and discussion

5.1 Catalysts characterizationThe textural properties of supports, synthesized catalysts and two spent catalysts(NiMo/γ-Al2O3 and NiMo/Pγ-Al2O3) were detected by N2-adsorption measure-ment. It can be seen in Figure 5.1 that the isotherms of NiMo/γ-Al2O3 fresh cata-lyst is type IV, suggesting the presence of large mesopores or macropores [51,109].However, the isotherm shapes of other fresh catalysts are very close to type II butthe quantity of adsorption increases rapidly when the relative pressure exceeds 0.9.This suggests that the adsorption process had two steps i.e. the porosity showed abimodal distribution. Hence, it can be concluded that the other fresh catalysts alsohad large mesopores or macropores.

Table 5.1: Textural properties of support and catalysts.

Catalyst BET surface area Pore volume Pore size(m2/g) (cm3/g) (Å)

γ-Al2O3150 149 0.475 127

γ-Al2O3 227 0.705 124NiMo/γ-Al2O3 188 0.504 107NiMo/Pγ-Al2O3 131 0.406 123NiMo/γ-Al2O3

150 113 0.335 117NiMo/γ-Al2O3

SEQ 189 0.511 107NiMo/γ-Al2O3

700 126 0.515 162Spent NiMo/γ-Al2O3 87 0.163 74Spent NiMo/Pγ-Al2O3 85 0.193 90

29

30 CHAPTER 5. RESULTS AND DISCUSSION

(a) Supports and catalysts

(b) Tested catalysts

(c) Non-tested catalysts

Figure 5.1: N2 adsorption-desorption isotherms


As can be observed in table 5.1 and Figure 5.1a, the surface area of the sup-ports decreased due to the metal loading during catalyst synthesis. NiMo/γ-Al2O3decreased about 17% while this reduction for NiMo/γ-Al2O3

150 was 24%. NiMo/γ-Al2O3

700 showed the considerable reduction of 44% which caused because of highercalcination temperature.Particle size, as reported by Ferdous et al. [9], increased by rising the calcination

temperature to 500 ◦C but reduced when calcination temperature increased to 600◦C. As the applied calcination temperature for NiMo/γ-Al2O3 was 500 ◦C, it maybe concluded that temperature of 500 ◦C is the optimal calcination temperaturefor nickel-molybdenum-alumina catalysts. At 500 ◦C, sintering, because of Al2O3transformation by forming Al2(MoO4)3, might have a role in increasing the particlesize [9]. Decrease in particle size at higher temperature may be due to sublimationof MoO3 as reported by Kasztelan et al. [110]. Phosphorous addition also causeda decrease of 42% maybe due to the blockage of micropores of the support byphosphates. Simillar results also reported by Jian and Prins [111] and Maity etal. [92]; hoowever, Ferdous et al. reported no change in surface area for the catalystwith phosphorous [9]. Comparing the two catalysts synthesized through differentco-impregnation procedures suggests that the applied methods had almost the sameimpact on the surface area (16.4% for NiMo/γ-Al2O3

SEQ).The results of the surface study on two spent catalysts (Table 5.1 and Figure 5.1)

showed the surface area of NiMo/γ-Al2O3 and NiMo/Pγ-Al2O3 decreased by about54 and 36% respectively. The pore volume and pore diameter of both catalysts alsodecreased significantly (>50%). These losses during hydrotreating process could beattributed to coke deposition [85] which led to significant catalyst deactivation byblocking the pores. As will be described hereafter while introducing SEM results,the coke formation for NiMo/γ-Al2O3 was more than the NiMo/Pγ-Al2O3 whichconfirmed the BET results.X-ray diffraction spectra for the supports are displayed in figure 5.2a.The bottom

spectrum of both supports patterns indicates three high peaks located at 38, 46and 67◦, which are γ-Al2O3 characteristic as also can be recognized in XRD of allcatalysts as shown in Figure 5.2. In general, the crystallinity is poor as the peaksintensity are rather small, specially the alumina with a higher surface area showeda lower crystallization degree and a smaller average crystallite size. These are inline with other works e.g. [45, 112, 113]. In XRD of the catalysts prepared fromthese supports, fig 5.2b, two peaks in 23.5 and 27.3◦ are related to MoO3 [70].Here, as expected, also the previous pattern was followed and the catalyst withlower-surface-area support showed a relatively higher crystallinity as the peaks aresharper and have higher intensity. No nickel oxides characteristic reflection (43.5and 63◦) suggested that these compounds were amorphous or microcrystalline [114],or too disposal or too low contents. However, at higher calcination temperature,i.e. 700 ◦C, a peak appeared at 63◦. as shown in Figure 5.2c. Calcination at highertemperature generally resulted in higher crystallization degree. The reflection ofNiMo/γ-Al2O3

SEQ is closely akin to NiMo/γ-Al2O3 but with a higher intensity,Figure 5.2d.


(a) XRD of supportswith different surface area

(b) XRD of NiMo/γ-Al2O3 withdifferent surface areas of support

(c) NiMo/γ-Al2O3calcined at 500 and 700 ◦C

(d) Impact of addingphosphorous on the support

(e) Difference between coimpregnationand sequential preparation methods

Figure 5.2: XRD of fresh catalysts


The impact of addition phosphorous can be observed in Figure 5.2e where theintensity of the both peaks of MoO3 are considerably improved indicating thatphosphorous improved the interaction between molybdenum and the support.Temperature program reduction was performed to evaluate the effect of the meth-

ods of preparation on the reducibility of the catalysts and examine their impact onthe structure of the active phase and on the interaction between the support andthe active phase. Based on previous studies [9,61,115] the reduction of MoO3 maybe divided into two steps. The low-temperature reduction that can be related tothe partial reduction (MoO6+ to MoO4+) of amorphous, highly defective, multilay-ered molybdenum oxides or hetropolymolybdates, which are generally accepted asthe precursor of the type II Ni-Mo-S. The high-temperature reduction is comprisedof tetrahedral molybdenum species including highly dispersed monolayered MoO3,molybdenum oxides with deep reduction and the intermediate-reducible crystallinephases of othorhombic MoO3 and Al2(MoO4)3 resulated from the strong reactionwith the support [116–118]. Ranges of 500-650 ◦C and 800-950 ◦C are suggested inliteratures [9,61,115,119] for the low-temperature and high-temperature reduction,respectively.As TPR was performed up to 800 ◦C in this experiment, no high-temperature

reduction observed as shown in Figure 5.3. The low-temperature peaks are alsobetween 550 and 700 ◦C. Higher reduction temperature in this experiment may bedue to reduction of NiO. According to Brito and Laine [120,121], bulk NiO reducedbefore 300 ◦C, but when it was supported on alumina, the peak appeared at 700 ◦Csince nickel might perform a role of reduction promoter of the molybdenum. Thereduction temperature, Tm, (shows the maximum temperature for H2 consumption)

Figure 5.3: TPR profile of fresh catalysts


Table 5.2: Compositions for catalysts obtained from SEM-EDX

Catalyst Element C O Al S P Ni Mo

NiMo/γ-Al2O3Site 1 0 48.82 32.16 0 0 3.66 15.36Site 2 0 44.87 37.33 0 0 1.27 16.53

NiMo/Pγ-Al2O3Site 1 0 30.59 34.81 0 4.47 7.91 22.21Site 2 0 36.29 29.52 0 3.63 10.79 19.77

NiMo/γ-Al2O3150 Max. 0 45.52 33.22 0 0 6.35 21.69

Min. 0 39.49 28.23 0 0 5.15 16.81

NiMo/γ-Al2O3SEQ Site 1 0 48.25 38.42 0 0 6.83 27.3

Site 2 0 28.87 29.31 0 0 5.41 12.65

Spent NiMo/γ-Al2O3Site 1 6.21 38.83 24.20 3.88 0 2.46 24.42Site 2 9.97 49.06 20.97 8.17 0 3.96 7.88

Spent NiMo/Pγ-Al2O3Site 1 3.60 65.10 14.53 3.15 1.71 2.41 9.49Site 2 4.8 50.84 20.99 10.50 2.39 4.32 6.17

Spent NiMo/γ-Al2O3150 6.23 41.39 22.93 8.39 0.00 4.50 16.56

for NiMo/γ-Al2O3 is 672 ◦C which is higher than NiMo/γ-Al2O3700 (633 ◦C) due

to higher calcination temperature. However, higher calcination temperature hadno effect on the intensity which is in contrast with the results of Ferdous et al. [9]who showed that higher calcination temperature significantly decreased the inten-sity. It was reported that all nickel reduced under the NiI band [120] at calcinationtemperature lower than 600 ◦C. It is also reported that most of the nickel simultane-ously reduced with molybdenum under the low-temperature reduction [75]. Nickelreducibility depends on the catalyst calcination temperature [9] as at calcinationtemperature over 600 ◦C, essentially all Ni reduces withMoII [75,120]. For NiMo/γ-Al2O3

700 more time might be needed for complete nickel reduction.NiMo/γ-Al2O3

SEQ showed the lowest Tm (589 ◦C) suggesting weaker interactionbetween nickel and support. The reduction temperature of NiMo/γ-Al2O3

150 (677◦C) and its intensity are higher than NiMo/γ-Al2O3. This may indicate betterinteractions between the support and metals in the catalyst with lower surface area.The reduction temperature of the catalyst with phosphorous was 40 ◦C higher thanNiMo/γ-Al2O3 which is similar to Zhou’s results (30 ◦C difference) [119]. Theformed AlPO4 phase on the support surface [122] inhibited the active componentand the support interaction. This might result in larger active phase particles [9]and higher reduction temperature [123].Scanning electron microscopy were used to study the surface morphology of fresh

(except NiMo/γ-Al2O3700) and spent catalysts whereas EDX was applied to quan-

tify the elemental composition of a single particle present on the surface of thecatalysts. Three different approach were applied: analyzing a surface area of a se-lected site, measuring through a long line on a selected site and analyzing severalshort lines in a surface of a selected site. The results of these analyses are tabulated


in Table 5.2 and the SEM micrographs are showed in Figure 5.2. The results (boththose mentioned here and other EDX analyses not presented here) indicate that thetheoretical composition is different with the targeted one. The reason can be theway of dropping metal solution on the support which resulted in a non-homogeneousdispersion or possible weighing error.Ferdous et al. [9] mentioned that higher Mo content obtained from EDX indicated

that molybdenum might have migrated and agglomerated more on the surface.However, as EDX analysis is based on the elemental analysis of selected particles onthe surface of the catalyst, it has limitation in quantifying the composition of thecatalyst surface. Figures 5.4c and 5.4d show SEM micrographs for NiMo/Pγ-Al2O3.Phosphorous addition to the catalyst gave irregular particle size and agglomerationas the presence of phosphorous in these particles confirmed by EDX analysis. Thisresult also is reported by Ferdous et al. [9].Sulfidation and precoking lead to formation of MoS2 [124] and carbon deposition

on the catalyst surface [9]. Comparing the micrographs and based on EDX analysisfor fresh and spent NiMo/γ-Al2O3 and NiMo/γ-Al2O3

150 indicated reduction ofparticle size in both spent catalysts indicating the formation of MoS2 as the resultof sulfiding. In NiMo/Pγ-Al2O3 case, the particles size are mostly larger in spentcatalyst (which had lower surface area, see 5.2) than the fresh one (Figures 5.4c and5.4d) . As reported by Lopez [125], it may be concluded that size and number ofMoS2 species was increased on the catalyst surface due to adding phosphorous (seealso section 3.3.4). Results of XRD, as discussed earlier above, confirmed this aswell (see Figure 5.2e) since MoS2 species were formed from MoO3 during sulfiding.A few parts in this catalyst were seen that totally eroded away (Figure 5.6c) andconsisted of about 80 wt.% Al2O3.As observed in spent catalysts micrographs (Figure 5.6), some fibers were formed

in three catalysts during the hydrotreating process. EDX analysis showed that theselong fibers mostly had nickel (more than 60 wt.%) and sulfur (between 15 and 25wt.%) indicating the deactivation of the catalyst by carbon (4 to 6 wt.%). This isthe deactivation phenomenon that can occur for catalysts containing nickel.Thesesfibers are called whiskers; however, they are usually carbon/based. The numberof these whiskers were considerable lower in spent NiMo/Pγ-Al2O3 meaning lowerdeactivation for this catalyst due to the presence of P which interacts stronglywith the support. Hence, it decreases the possibility of direct interacting of cokeprecursors with the support. This is in line with the study of Maity [92]. On theother hand, phosphorous reduces the number of strong acid sites of the catalyst [126]which facilitate th coke formation on the catalyst [92]. Thus, this can be anotherreason for lower deactivation for NiMo/Pγ-Al2O3.


(a) NiMo/γ-Al2O3 Site 1 (b) NiMo/γ-Al2O3 Site 2

(c) NiMo/Pγ-Al2O3 Site 1 (d) NiMo/Pγ-Al2O3 Site 2

(e) NiMo/γ-Al2O3150 (f) NiMo/γ-Al2O3

SEQ

Figure 5.4: SEM micrographs for fresh catalysts


(a) Spent NiMo/γ-Al2O3 Site 1 (b) Spent NiMo/γ-Al2O3 Site 2

(c) Spent NiMo/Pγ-Al2O3 Site 1 (d) Spent NiMo/Pγ-Al2O3 Site 2

(e) Spent NiMo/γ-Al2O3150

Figure 5.5: SEM micrographs for spent catalysts


(a) Spent NiMo/γ-Al2O3 (b) Spent NiMo/γ-Al2O3

(c) Spent NiMo/Pγ-Al2O3 (d) Spent NiMo/γ-Al2O3150

Figure 5.6: Formation of whiskers in SEM micrographs for spent catalysts

5.2 Catalytic activity

Under the conditions of the experiment, the catalysts were on streams for six daysto show their initial activity loss; then steady activity was reached. As shown inFigure 5.7, the aromatic content increased during ageing mainly due to decreasingof active sites. All catalysts showed similar trends. The results of products analysisare summarized in Table 5.3. In addition, yield was calculated for each sample andplotted as shown in Figure 5.8. Unfortunately, the yield results are not exact dueto the instrument low scale and very low flow rates.One of the main factors for catalyst deactivation is coke deposition which may

introduce important changes in the surface properties. The coke deposition as ob-served in SEM mongraphs (see Figure 5.6) can be formed firstly during the activa-tion step (i.e. sulfiding) due to the sulfur sources; and then, during the hydrotreatingprocess through secondary reactions (polyaromatics chain) [47]. In order to studythe catalytic activity, based on measured data tabulated in Table 5.3, the contentsof aromatics, nitrogen and sulfur in products at each test temperature are plot-ted as illustrated in Figure 5.9. Also, conversions and activity rate constants werecalculated and plotted as shown in Figure 5.10.HDA is inhibited by hydrogen sulfide produced from the HDS activity [127] as

5.2. CATALYTIC ACTIVITY 39

Figure 5.7: Catalysts Ageing Trend at Tbase and x bar

Figure 5.8: Process yield for hydrotreated oils


Table 5.3: Products characteristics during a two-day process

Characteristic Ca Cp Cn Viscosity Density Sulfur Nitrogen ColorUnit % % % mm2/s g/cm3 ppm ppm

NiMo/γ-Al2O3

Tbase-20 19.1 44.4 36.5 21.84 0.911 3,076 120 12Tbase 17.8 42.2 40.0 19.04 0.902 585 9 23Tbase+20 12.6 43.4 44.0 13.34 0.888 242 3 30

NiMo/Pγ-Al2O3

Tbase-20 20.6 43.5 36.4 22.22 0.912 3,971 152 4Tbase 18.4 42.7 39.0 18.32 0.902 818 10 23Tbase+20 11.0 44.3 44.7 11.22 0.884 368 7 26

NiMo/γ-Al2O3150

Tbase-20 20.1 43.4 36.5 22.07 0.911 3413 129 17Tbase 18.5 42.6 39.0 19.37 0.903 778 13 21Tbase+20 12.0 43.6 44.4 12.18 0.886 237 7 30

well as the three-ring aromatic compounds which considerably inhibit the hydro-genation of the two-ring aromatic and one-ring aromatic compounds [128]. On theother hand, aromatics and nitrogen require severer operating conditions for theirelimination [47]. By analyzing the temperature range (Figure 5.10a), it can beobserved that HDA was increased at higher temperatures. Unexpectedly, HDN ac-tivity trend was not similar to HDA as the HDN reaction takes place mainly throughhydrogenation of the aromatic rings [111]. The reason might be related to the in-hibition effect. A slight improvement at higher temperature in HDA and aromaticsconversion for NiMo/Pγ-Al2O3 was observed (Figure 5.10a and5.10b) maybe dueto the fact that phosphorous induces either new or more active sites for the hydro-genation of aromatics [129]. A reason for this was explained by Jian and Prins [129]as it might be the replacement of a sulfur atom by a phosphorous atom in the nickelcoordination. Phosphorous which is close to Ni-MoS2 particles might be reduced toPH3 by spilled over hydrogen atoms, which induces a P-S exchange reaction at thecatalyst surface.Sulfur removal in HDS may directly take place without hydrogenation of the asso-

ciated aromatic ring, while removal of nitrogen happens through a complex sequenceof reaction which involves aromatic ring hydrogenation followed by carbon-nitrogenbond breakage [85]. All catalysts showed similar HDS activity rate constant at lowtemperature (Figure 5.10e) but as the temperature was increased, the catalytic re-sponses were enhanced, obtaining a better catalytic activity for NiMo/γ-Al2O3 andNiMo/γ-Al2O3

150. This could be attributed to the effect of adding phosphorous.Nitrogen content mainly affects the color of products. As illustrated in Figure

5.11a, the results of color measurement are also in line with the theory. For instance,NiMo/Pγ-Al2O3 showed less brightness than the other catalysts at Tbase+20 as ithad also the highest nitrogen content among the others. At Tbase ◦C also the trendof the color and nitrogen content for all catalysts are the same. Density is highlyaffected by the aromatics content and as seen in Figure 5.11b, the results showsimilar trend. Comparing figures 5.9a and 5.11c showed that viscosity was alsoincreased as the aromatic content was raised.


(a) Aromatics content

(b) Nitrgen content

(c) Sulfur content

Figure 5.9: Aromatics, nitrogen and sulfur contents in products


(a) HDA (b) Aromatics

(c) HDN (d) Nitrogen

(e) HDS (f) Sulfur

Figure 5.10: HDA, HDN AND HDS activity rate constants and conversions


(a) Color

(b) Density

(c) Viscosity

Figure 5.11: Color, density and viscosity in products

Chapter 6

Conclusions and further Work

The results of BET, XRD and TPR studies showed the effect of phosphorous onimprovement of NiMo/γ-Al2O3 characteristics. Phosphorous addition to NiMo/γ-Al2O3 increased the number of accessible molybdenum sites on the surface due tohigher dispersion. Regarding catalytic activity, NiMo/Pγ-Al2O3 enhanced HDAwhile it had negative impact on HDS and HDN. SEM micrographs showed a muchless deactivation in spent NiMo/Pγ-/Al2O3 indicating stability improvement. ForNiMo/γ-Al2O3

150, the surface area loss was more due to the lower surface area ofthe support as seen in BET results. This catalyst performed much better at highertemperature, specially for HDS and HDN. However, SEM micrographs revealedhigher deactivation for this catalyst compared to NiMo/γ-Al2O3. Although noactivity test was carried out for NiMo/γ-Al2O3

700, it can be concluded that due tothe severe surface area reduction and lower number of MoO3, lower activity would beexpected. Based on the results of BET, coimpregnation and sequential impregnationbehaved similarly regarding the loss of surface area. However, TPR study showedthe nickel in the catalyst impregnated sequentially had weaker interaction with thesupport in comparison with the catalyst prepared by coimpregnation.The behavior of nickel-molybdenum supported on γ-alumina significantly depends

on the preparation method, calcination temperature and promoter. Therefore, beingfamiliar with the variety of these behaviors helps to select the right catalyst to obtainthe desire product(s). It should be noted that other parameters e.g. feed type,operating conditions, sulfiding, promoter amount, etc. affect the catalyst behavioras well. Hence, the continuation of this study can be included the activity test for allprepared catalysts on different feed types and operating conditions. Also, applyingcharacterization methods for sulfided catalysts is suggested to evaluate sulfidingeffect on catalytic activity and find the optimum conditions. For the products thatHDN and HDA are important, evaluating the effect of hydrothermal treatment ofγ-alumina on the activity of NiMo/γ-Al2O3 and NiMo/Pγ-Al2O3 is suggested.

45

Bibliography

[1] Henrik Topsøe, Bjerne S. Clausen, and Franklin E. Massoth. Catalysis: Sci-ence and Technology, chapter Hydrotreating Catalysis, pages 1–269. SpringerBerlin Heidelberg, Berlin, Heidelberg, 1996.

[2] Antony Stanislaus, Abdulazeem Marafi, and Mohan S. Rana. Recent advancesin the science and technology of ultra low sulfur diesel (ULSD) production.Catalysis Today, 153(1-2):1 – 68, 2010. MONOGRAPH: Recent advances inthe science and technology of ultra low sulfur diesel (ULSD) production.

[3] A. Stanislaus, M. Absi-Halabi, K. Al-Dolama, A. Katrib, and M. Ismail. Ef-fect of thermal treatment on the sintering and structural changes of cobalt-molybdenum/alumina and nickel-molybdenum/alumina hydrotreating cata-lysts. Applied Catalysis, 41:109 – 119, 1988.

[4] B. Scheffer, E.M. van Oers, P. Arnoldy, V.H.J. de Beer, and J.A. Moulijn.Sulfidability and HDS activity of CoMo/Al2O3 catalysts. Applied Catalysis,25(1-2):303 – 311, 1986.

[5] Paul R. Robinson and Geoffrey E. Dolbear. Hydrotreating and hydrocracking:Fundamentals. In Chang S. Hsu and Paul R. Robinson, editors, PracticalAdvances in Petroleum Processing, pages 177–218. Springer New York, 2006.

[6] Carolina Leyva, Jorge Ancheyta, Arnaud Travert, Francoise Mauge, LaurenceMariey, Jorge Ramirez, and Mohan S. Rana. Activity and surface properties ofNiMo/SiO2-Al2O3 catalysts for hydroprocessing of heavy oils. Applied Catal-ysis A: General, 425-426:1 – 12, 2012.

[7] Antony Stanislaus Meena Marafi and Edward Furimsky. Chapter 1 -Introduction. In Antony Stanislaus Meena Marafi and Edward Furimsky,editors, Handbook of Spent Hydroprocessing Catalysts, pages 1 – 4. Elsevier,Amsterdam, 2010.

[8] A. Marafi, A. Al-Hendi, A. Al-Mutawa, and A. Stanislaus. Studies on hy-drotreating of diesel streams from different Kuwait crudes for ultralow sulfurdiesel production. Energy & Fuels, 21.

47

48 BIBLIOGRAPHY

[9] D. Ferdous, A.K. Dalai, and J. Adjaye. A series of NiMo/Al2O3 catalysts con-taining boron and phosphorus: Part I. synthesis and characterization. AppliedCatalysis A: General, 260(2):137 – 151, 2004.

[10] Edward Furimsky. Selection of catalysts and reactors for hydroprocessing.Applied Catalysis A: General, 171(2):177 – 206, 1998.

[11] A.L. Dicks, R.L. Ensell, T.R. Phillips, A.K. Szczepura, M. Thorley,A. Williams, and R.D. Wragg. A study of relationships between pore sizedistribution, hydrogen chemisorption, and activity of hydrodesulphurisationcatalysts. Journal of Catalysis, 72(2):266 – 273, 1981.

[12] R. Prins. Catalytic hydrodenitrogenation. volume 46 of Advances in Catalysis,pages 399 – 464. Academic Press, 2001.

[13] R. Prins, V. H. J. De Beer, and G. A. Somorjai. Structure and function of thecatalyst and the promoter in Co-Mo hydrodesulfurization catalysts. CatalysisReviews, 31(1-2):1–41, 1989.

[14] M.H. Farias, A.J. Gellman, G.A. Somorjai, R.R. Chianelli, and K.S. Liang.The Co coadsorption and reactions of sulfur, hydrogen and oxygen on cleanand sulfided Mo(100) and on MoS2(0001) crystal faces. Surface Science,140(1):181 – 196, 1984.

[15] S. Helveg, J. V. Lauritsen, E. Laegsgaard, I. Stensgaard, J. K. Nørskov, B. S.Clausen, H. Topsøe, and F. Besenbacher. Atomic-scale structure of single-layer MoS2 nanoclusters. Physical Review Letters, 84:951–954, January 2000.

[16] Antony Stanislaus Meena Marafi and Edward Furimsky. Chapter 3 -Hydroprocessing of Petroleum. In Antony Stanislaus Meena Marafi and Ed-ward Furimsky, editors, Handbook of Spent Hydroprocessing Catalysts, pages17 – 49. Elsevier, Amsterdam, 2010.

[17] J.M.J.G. Lipsch and G.C.A. Schuit. The CoO MoO3Al2O3 catalyst. Journalof Catalysis, 15(2):163 – 173, 1969.

[18] J.M.J.G. Lipsch and G.C.A. Schuit. The CoO MoO3 Al2O3 catalyst: II. thestructure of the catalyst. Journal of Catalysis, 15(2):174 – 178, 1969.

[19] J.M.J.G. Lipsch and G.C.A. Schuit. The CoO MoO3Al2O3 catalyst. Journalof Catalysis, 15(2):179 – 189, 1969.

[20] V.H.J De Beer, T.H.M Van Sint Fiet, J.F Engelen, A.C Van Haandel, M.W.JWolfs, C.H Amberg, and G.C.A Schuit. The CoO MoO3Al2O3 catalyst. iv.pulse and continuous flow experiments and catalyst promotion by cobalt,nickel, zinc, and manganese. Journal of Catalysis, 27(3):357 – 368, 1972.

BIBLIOGRAPHY 49

[21] R.J.H. Voorhoeve. Electron spin resonance study of active centers in nickel-tungsten sulfide hydrogenation catalysts. Journal of Catalysis, 23(2):236 –242, 1971.

[22] R.J.H. Voorhoeve and J.C.M. Stuiver. Kinetics of hydrogenation on supportedand bulk nickel-tungsten sulfide catalysts. Journal of Catalysis, 23(2):228 –235, 1971.

[23] Bernard Delmon. Advances in hydropurification catalysts and catalysis. InM. Absi-Halabi D.L. Trimm, S. Akashah and A. Bishara, editors, Catalystsin Petroleum Refining 1989Proceedings of the Conference on Catalysts inPetroleum Refining, volume 53 of Studies in Surface Science and Catalysis,pages 1 – 40. Elsevier, 1989.

[24] G. Hagenbach, Ph. Courty, and B. Delmon. Physicochemical investiga-tions and catalytic activity measurements on crystallized molydbenum sulfide-cobalt sulfide mixed catalysts. Journal of Catalysis, 31(2):264 – 273, 1973.

[25] P. Grange and B. Delmon. The role of cobalt and molybdenum sulphidesin hydrodesulphurisation catalysts: A review. Journal of the Less CommonMetals, 36(1-2):353 – 360, 1974.

[26] B. S. Clausen, S. Mørup, H. Topsøe, R. Candia, E. J. Jensen, A. Baranski,and A. Pattek. Mossbauer studies of the activation of ammonia synthesiscatalysts. J. Phys. Colloques, 37(C6):C6–245–C6–248, 1976.

[27] Henrik Topsøe, Roberto Candia, Nan-Yu Topsøe, Bjarne S. Clausen, andHaldor Topsøe. On the state of the Co-mo-s model. Bulletin des SocietesChimiques Belges, 93(8-9):783–806, 1984.

[28] Henrik Topsøe and Bjerne S. Clausen. Importance of CoMo-S type structuresin hydrodesulfurization. Catalysis Reviews, 26(3-4):395–420, 1984.

[29] Bjerne S. Clausen, Henrik Topsøe, Roberto Candia, Jorgen Villadsen, BrunoLengeler, Jens Als-Nielsen, and Finn Christensen. Extended X-ray absorptionfine structure study of the cobalt-molybdenum hydrodesulfurization catalysts.The Journal of Physical Chemistry, 85(25):3868–3872, 1981.

[30] Henrik Topsøe, Bjerne S. Clausen, Roberto Candia, Carsten Wivel, and SteenMørup. In situ mossbauer emission spectroscopy studies of unsupported andsupported sulfided CoMo hydrodesulfurization catalysts: Evidence for andnature of a CoMo-s phase. Journal of Catalysis, 68(2):433 – 452, 1981.

[31] Carsten Wivel, Roberto Candia, Bjerne S. Clausen, Steen Mørup, and Hen-rik Topsøe. On the catalytic significance of a CoMoS phase in CoMo/Al2O3hydrodesulfurization catalysts: Combined in situ mossbauer emission spec-troscopy and activity studies. Journal of Catalysis, 68(2):453 – 463, 1981.

50 BIBLIOGRAPHY

[32] Henrik Topsøe, Bjerne S. Clausen, Nan Yu Topsøe, and Erik Pedersen. Recentbasic research in hydrodesulfurization catalysis. Industrial & EngineeringChemistry Fundamentals, 25(1):25–36, 1986.

[33] Nan-Yu Topsøe and Henrik Topsøe. Characterization of the structures andactive sites in sulfided CoMoAl2O3 and NiMoAl2O3 catalysts by no chemisorp-tion. Journal of Catalysis, 84(2):386 – 401, 1983.

[34] Henrik Topsøe, Bjerne S. Clausen, Nan-Yu Topsøe, and Per Zeuthen. Progressin the design of hydrotreating catalysts based on fundamental molecular in-sight. In M. Absi-Halabi D.L. Trimm, S. Akashah and A. Bishara, editors,Catalysts in Petroleum Refining 1989Proceedings of the Conference on Cat-alysts in Petroleum Refining, volume 53 of Studies in Surface Science andCatalysis, pages 77 – 102. Elsevier, 1989.

[35] S.M.A.M. Bouwens, F.B.M. Vanzon, M.P. Vandijk, A.M. Vanderkraan, V.H.J.Debeer, J.A.R. Vanveen, and D.C. Koningsberger. On the structural differ-ences between alumina-supported comos type I and alumina-, silica-, andcarbon-supported comos type II phases studied by xafs, mes, and xps. Jour-nal of Catalysis, 146(2):375 – 393, 1994.

[36] G. C. A. Schuit and B. C. Gates. Chemistry and engineering of catalytichydrodesulfurization. AICHE Journal, 19(3):417 – 438, 1973.

[37] E.G. Derouane, E. Pedersen, B.S. Clausen, Z. Gabelica, R. Candia, andH. Topsøe. Epr studies on unsupported and alumina-supported sulfided CoMohydrodesulfurization catalysts. Journal of Catalysis, 99(2):253 – 261, 1986.

[38] Berit Hinnemann, Jens K. Nørskov, and Henrik Topsøe. A density functionalstudy of the chemical differences between type I and type II MoS2-basedstructures in hydrotreating catalysts. The Journal of Physical Chemistry B,109(6):2245–2253, 2005. PMID: 16851217.

[39] S. Eijsbouts, L.C.A. van den Oetelaar, and R.R. van Puijenbroek. MoS2morphology and promoter segregation in commercial type 2 Ni-Mo/Al2O3 andCo-Mo/Al2O3 hydroprocessing catalysts. Journal of Catalysis, 229(2):352 –364, 2005.

[40] Yasuaki Okamoto, Akira Kato, Usman, Nino Rinaldi, Takashi Fujikawa, Hi-romichi Koshika, Ichiro Hiromitsu, and Takeshi Kubota. Effect of sulfidationtemperature on the intrinsic activity of Co-MoS2 and Co-WS2 hydrodesulfu-rization catalysts. Journal of Catalysis, 265(2):216 – 228, 2009.

[41] Yasuaki Okamoto, Kazuya Hioka, Kenichi Arakawa, Takashi Fujikawa,Takeshi Ebihara, and Takeshi Kubota. Effect of sulfidation atmosphere onthe hydrodesulfurization activity of SiO2-supported Co-Mo sulfide catalysts:Local structure and intrinsic activity of the active sites. Journal of Catalysis,268(1):49 – 59, 2009.

BIBLIOGRAPHY 51

[42] Kim G Knudsen, Barry H Cooper, and Henrik Topsøe. Catalyst and pro-cess technologies for ultra low sulfur diesel. Applied Catalysis A: General,189(2):205 – 215, 1999.

[43] Y. Sakashita and T. Yoneda. Orientation of MoS2 clusters supported on twokinds of γ-Al2O3 single crystal surfaces with different indices. Journal ofCatalysis, 185(2):487 – 495, 1999.

[44] Y Sakashita. Effects of surface orientation and crystallinity of alumina sup-ports on the microstructures of molybdenum oxides and sulfides. SurfaceScience, 489(1-3):45 – 58, 2001.

[45] Yukio Sakashita, Yasuhiro Araki, and Hiromichi Shimada. Effects of surfaceorientation of alumina supports on the catalytic functionality of molybdenumsulfide catalysts. Applied Catalysis A: General, 215(1-2):101 – 110, 2001.

[46] M.A. Dominguez-Crespo, E.M. Arce-Estrada, A.M. Torres-Huerta, L. Diaz-Garcia, and M.T. Cortez de la Paz. Support effects on hydrotreating activityof NiMo catalysts. Materials Characterization, 58(10):864 – 873, 2007.

[47] M.A. Dominguez-Crespo, A.M. Torres-Huerta, L. Diaz-Garcia, E.M. Arce-Estrada, and E. Ramrez-Meneses. HDS, HDN and HDA activities of nickel-molybdenum catalysts supported on alumina. Fuel Processing Technology,89(8):788 – 796, 2008.

[48] Hui Yuan Yunhong Zhang Qiherima, Huifeng Li and Guangtong Xu. Effect ofalumina support on the formation of the active phase of selective hydrodesul-furization catalysts CoMo/Al2O3. Chinese Journal of Catalysis, 32(1-2):240– 249, 2011.

[49] T.K.T. Ninh, L. Massin, D. Laurenti, and M. Vrinat. A new approach inthe evaluation of the support effect for NiMo hydrodesulfurization catalysts.Applied Catalysis A: General, 407(1-2):29 – 39, 2011.

[50] Li Jun-Cheng, Xiang Lan, Xu Feng, Wang Zhan-Wen, and Wei Fei. Effectof hydrothermal treatment on the acidity distribution of γ-Al2O3 support.Applied Surface Science, 253(2):766 – 770, 2006.

[51] Guangci Li, Xinlu Lu, Zhe Tang, Yunqi Liu, Xuebing Li, and ChenguangLiu. Preparation of NiMo/γ-Al2O3 catalysts with large pore size for vacuumresidue hydrotreatment. Materials Research Bulletin, 48(11):4526 – 4530,2013.

[52] Michele Breysse, Pavel Afanasiev, Christophe Geantet, and Michel Vrinat.Overview of support effects in hydrotreating catalysts. Catalysis Today, 86(1-4):5 – 16, 2003. Effects of Support in Hydrotreating Catalysis for Ultra-cleanFuels.

52 BIBLIOGRAPHY

[53] M. Absi-Halabi, A. Stanislaus, and H. Al-Zaid. Effect of acidic and basicvapors on pore size distribution of alumina under hydrothermal conditions.Applied Catalysis A: General, 101(1):117 – 128, 1993.

[54] Antony Stanislaus, Khalida Al-Dolama, and Mamun Absi-Halabi. Preparationof a large pore alumina-based HDM catalyst by hydrothermal treatment andstudies on pore enlargement mechanism. Journal of Molecular Catalysis A:Chemical, 181(1-2):33 – 39, 2002. Fifteenth Indian National Symposium onCatalysis and Second Conference of the Indo-Pacific Catalysis Association.

[55] L. Xiang, Y.L. Gong, J.C. Li, and Z.W. Wang. Influence of hydrothermalmodification on the properties of Ni/Al2O3 catalyst. Applied Surface Science,239(1):94 – 100, 2004.

[56] Juncheng Li, Lan Xiang, Xu Feng, and Zhanwen Wang. Influence of hy-drothermally modified γ-Al2O3 on the properties of NiMo/γ-Al2O3 catalyst.Applied Surface Science, 254(7):2077 – 2080, 2008.

[57] Haitao Li, Yalin Xu, Chunguang Gao, and Yongxiang Zhao. Structural andtextural evolution of Ni/γ-Al2O3 catalyst under hydrothermal conditions.Catalysis Today, 158(3-4):475 – 480, 2010. Perspective on catalysis for sus-tainable energy and environmental technologies.

[58] Roman M. Mironenko, Olga B. Belskaya, Valentin P. Talsi, Tatyana I.Gulyaeva, Maxim O. Kazakov, Alexander I. Nizovskii, Alexander V. Kalinkin,Valerii I. Bukhtiyarov, Alexander V. Lavrenov, and Vladimir A. Likholobov.Effect of γ-Al2O3 hydrothermal treatment on the formation and properties ofplatinum sites in Pt/γ-Al2O3 catalysts. Applied Catalysis A: General, 469:472– 482, 2014.

[59] G. Leofanti, G. Tozzola, M. Padovan, G. Petrini, S. Bordiga, and A. Zecchina.Catalyst characterization: characterization techniques. Catalysis Today, 34(3-4):307 – 327, 1997. The catalytic process from the laboratory to the industrialplant.

[60] Calvin H. Bartholomew and Robert J. Farrauto. Catalyst Characterizationand Selection, pages 118–196. John Wiley Sons, Inc., 2005.

[61] K Anas and K.K Mohammed Yusuff. Synthesis, characterization and hy-drodesulphurisation activity of CoMo/γ-Al2O3 catalyst prepared throughmolecular designed dispersion method. Applied Catalysis A: General,264(2):213 – 217, 2004.

[62] A.M Venezia, V La Parola, G Deganello, D Cauzzi, G Leonardi, andG Predieri. Influence of the preparation method on the thiophene HDS activ-ity of silica supported CoMo catalysts. Applied Catalysis A: General, 229(1-2):261 – 271, 2002. Dedicated to Professor Laszlo Guczi on the occasion ofhis 70th Birthday.

BIBLIOGRAPHY 53

[63] E. Hillerova, H. Morishige, K. Inamura, and M. Zdrazil. Formation of mono-layer of molybdena over alumina by unconventional slurry impregnation orsolvent assisted spreading method. Applied Catalysis A: General, 156(1):1 –17, 1997.

[64] Ch Papadopoulou, J Vakros, H.K Matralis, Ch Kordulis, and A Lycourghiotis.On the relationship between the preparation method and the physicochemicaland catalytic properties of the CoMo/γ-Al2O3 hydrodesulfurization catalysts.Journal of Colloid and Interface Science, 261(1):146 – 153, 2003.

[65] J Sarrin, O Noguera, H Royo, M.J Perez Zurita, C Scott, M.R Goldwasser,J Goldwasser, and M Houalla. Effect of the preparation method on the re-ducibility of molybdena-alumina catalysts. Journal of Molecular Catalysis A:Chemical, 144(3):441 – 450, 1999.

[66] Li Wang and W.Keith Hall. The preparation and genesis of molybdena-alumina and related catalyst systems. Journal of Catalysis, 77(1):232 – 241,1982.

[67] John Vakros, Christina Papadopoulou, George A. Voyiatzis, Alexis Lycourghi-otis, and Christos Kordulis. Modification of the preparation procedure for in-creasing the hydrodesulfurisation activity of the CoMo/γ-alumina catalysts.Catalysis Today, 127(1-4):85 – 91, 2007. I.A. Vasalos Festschrift.

[68] Kyriakos Bourikas, Christos Kordulis, and Alexis Lycourghiotis. The role ofthe liquidâsolid interface in the preparation of supported catalysts. CatalysisReviews, 48(4):363–444, 2006.

[69] Maria Ferrari, Bernard Delmon, and Paul Grange. Influence of the impreg-nation order of molybdenum and cobalt in carbon-supported catalysts forhydrodeoxygenation reactions. Carbon, 40(4):497 – 511, 2002.

[70] P Salerno, S Mendioroz, and A Lopez Agudo. Al-pillared montmorillonite-based NiMo catalysts for HDS and HDN of gas oil: influence of the methodand order of Mo and Ni impregnation. Applied Catalysis A: General, 259(1):17– 28, 2004.

[71] A Erhan Aksoylu, A Inci Isli, and Z Ilsen Onsan. Interaction between nickeland molybdenum in NiMo/Al2O3 catalysts: III.: Effect of impregnation strat-egy. Applied Catalysis A: General, 183(2):357 – 364, 1999.

[72] Carsten Wivel, Bjerne S. Clausen, Roberto Candia, Steen Mørup, and Hen-rik Topsøe. Mossbauer emission studies of calcined CoMo/Al2O3 catalysts:Catalytic significance of Co precursors. Journal of Catalysis, 87(2):497 – 513,1984.

54 BIBLIOGRAPHY

[73] F.E. Massoth and K.S. Chung. Studies of molybdena-alumina catalysts ix.effect of cobalt on thiophene hydrodesulfurization. In T. Seivama and K. Tan-abe, editors, New Horizons in CatalysisProceedings of the 7th InternationalCongress on Catalysis, volume 7, Part A of Studies in Surface Science andCatalysis, pages 629 – 642. Elsevier, 1981.

[74] Chung Ping Li and David M. Hercules. A surface spectroscopic study ofsulfided molybdena-alumina catalysts. The Journal of Physical Chemistry,88(3):456–464, 1984.

[75] R. Burch and A. Collins. Temperature-programmed reduction of Ni/Mo hy-drotreating catalysts. Applied Catalysis, 18(2):389 – 400, 1985.

[76] J. Laine, F. Severino, C.V. Caceres, J.L.G. Flerro, and A.Lopez Agudo.Effect of calcination temperature on the surface exposure of promoters inNiCoMo/Al2O3 catalysts. Journal of Catalysis, 103(1):228 – 231, 1987.

[77] A. Stanislaus and K. Al-Dolama. Effect of thermal treatment on the con-centration and distribution profiles of molybdenum, cobalt, and nickel inCoMoAl2O3 and NiMoAl2O3 catalysts. Journal of Catalysis, 101(2):536 –539, 1986.

[78] D. S. Zingg, Leo E. Makovsky, R. E. Tischer, Fred R. Brown, and David M.Hercules. A surface spectroscopic study of molybdenum-alumina catalystsusing X-ray photoelectron, ion-scattering, and raman spectroscopies. TheJournal of Physical Chemistry, 84(22):2898–2906, 1980.

[79] Robert B. Greegor, Farrel W. Lytle, Roland L. Chin, and David M. Hercules.Investigation of supported cobalt and nickel catalysts by X-ray absorptionspectroscopy. The Journal of Physical Chemistry, 85(9):1232–1235, 1981.

[80] L. Wang and W.Keith Hall. Raman spectra from several molybdena-aluminapreparations. Journal of Catalysis, 83(1):242 – 244, 1983.

[81] R. Mone and L. Moscou. The Structural Form of Cobalt and Nickel Promotersin Oxidic HDS Catalysts, chapter 11, pages 150–162.

[82] R. Mone. The activator system of oxidic hydrodesulfurization catalysts:Preparational aspects. In P. Jacobs B. Delmon and G. Poncelet, editors,Preparation of Catalysts IScientific Bases for the Preparation of Heteroge-neous Catalysts Proceedings of the First International Symposium held at theSolvay Research Centre, volume 1 of Studies in Surface Science and Catalysis,pages 381 – 391. Elsevier, 1976.

[83] Jorge Laine and Kerry C. Pratt. Structure of nickel-molybdenum hydrodesul-furization catalysts. Industrial & Engineering Chemistry Fundamentals,20(1):1–5, 1981.

BIBLIOGRAPHY 55

[84] J.L.G. Fierro A. Benitez, J. Ramirez and A. Lopez Agudo. Effect of fluorideon the structure and activity of NiW/Al2O3 catalysts for HDS of thiopheneand HDN of pyridine. Applied Catalysis A: General, 144(1-2):343 – 364, 1996.

[85] D. Ferdous, A.K. Dalai, and J. Adjaye. A series of NiMo/Al2O3 catalysts con-taining boron and phosphorus: Part II. hydrodenitrogenation and hydrodesul-furization using heavy gas oil derived from athabasca bitumen. Applied Catal-ysis A: General, 260(2):153 – 162, 2004.

[86] Luis Cedeno Rogelio Cuevas Marco Castillo JoseMaria PalaciosJorge Ramirez, Perla Castillo and Antonio Lopez-Agudo. Effect ofboron addition on the activity and selectivity of hydrotreating CoMo/Al2O3catalysts. Applied Catalysis A: General, 132(2):317 – 334, 1995.

[87] M Lewandowski and Z Sarbak. The effect of boron addition on hydrodesul-furization and hydrodenitrogenation activity of NiMo/Al2O3 catalysts. Fuel,79(5):487 – 495, 2000.

[88] Shu-na Han Yong-ming Chai Yun-qi Liu Tong-na Zhou, Hai-liang Yin andChen guang Liu. Influences of different phosphorus contents on NiMoP/Al2O3hydrotreating catalysts. Journal of Fuel Chemistry and Technology, 37(3):330– 334, 2009.

[89] V. Sundaramurthy, A.K. Dalai, and J. Adjaye. The effect of phosphorus onhydrotreating property of NiMo/γ-Al2O3 nitride catalyst. Applied CatalysisA: General, 335(2):204 – 210, 2008.

[90] V. Sundaramurthy, A.K. Dalai, and J. Adjaye. Effect of phosphorus addi-tion on the hydrotreating activity of NiMo/Al2O3 carbide catalyst. CatalysisToday, 125(3-4):239 – 247, 2007. Catalysts and Processes for Heavy Oil Up-grading.

[91] Jing Fan Chun-e Xiang, Yong-ming Chai and Chen guang Liu. Effect of phos-phorus on the hydrodesulfurization and hydrodenitrogenation performance ofpresulfided NiMo/Al2O3 catalyst. Journal of Fuel Chemistry and Technology,39(5):355 – 360, 2011.

[92] S.K. Maity, J. Ancheyta, M.S. Rana, and P. Rayo. Effect of phosphoruson activity of hydrotreating catalyst of Maya heavy crude. Catalysis Today,109(1-4):42 – 48, 2005. Hydroprocessing of Heavy Oil FractionsHydroprocess-ing of Heavy Oil Fractions.

[93] P.J. Mangnus, J.A.R. Van Veen, S. Eijsbouts, V.H.J. De Beer, and J.A.Moulijn. Structure of phosphorus containing CoO Mo-O3/Al2O3 catalysts.Applied Catalysis, 61(1):99 – 122, 1990.

56 BIBLIOGRAPHY

[94] J.L.G. Fierro R.Lopez Cordero, S.Lopez Guerra and A.Lopez Agudo. Forma-tion of Al2(MoO4)3 and MoO3 phases induced by phosphate in molybdena-phosphorus catalysts. Journal of Catalysis, 126(1):8 – 12, 1990.

[95] Jenny M. Lewis, Ronald A. Kydd, P. Michael Boorman, and Phyllis H. VanRhyn. Phosphorus promotion in nickel-molybdenum/alumina catalysts:model compound reactions and gas oil hydroprocessing. Applied CatalysisA: General, 84(2):103 – 121, 1992.

[96] Wencheng Zhang Ye Zhao Zhihua Zhang-Xiaohua Wang Baojian Shen,Huifeng Li and Shikong Shen. A novel composite support for hydrotreat-ing catalyst aimed at ultra-clean fuels. Catalysis Today, 106(1-4):206 – 210,2005. International Conference on Gas-Fuel 05International Conference onGas-Fuel 05International Conference on Gas-Fuel 05.

[97] M.J Vissenberg, Y van der Meer, E.J.M Hensen, V.H.J de Beer, A.M van derKraan, R.A van Santen, and J.A.R van Veen. The effect of support interactionon the sulfidability of Al2O3- and TiO2-supported CoW and NiW hydrodesul-furization catalysts. Journal of Catalysis, 198(2):151 – 163, 2001.

[98] J.A.R. van Veen, H.A. Colijn, P.A.J.M. Hendriks, and A.J. van Welsenes. Onthe formation of type I and type II NiMoS phases in NiMo/Al2O3 hydrotreat-ing catalysts and its catalytic implications. Fuel Processing Technology, 35(1-2):137 – 157, 1993. Hydroprocesses.

[99] S. Eijsbouts, J.N.M. Van Gestel, J.A.R. Van Veen, V.H.J. De Beer, andR. Prins. The effect of phosphate on the hydrodenitrogenation activity andselectivity of alumina-supported sulfided Mo, Ni, and NiMo catalysts. Journalof Catalysis, 131(2):412 – 432, 1991.

[100] Richard A. Kemp and Charles T. Adams. Hydrogel-derived catalysts. labo-ratory results on nickel-molybdenum and cobalt-molybdenum hydrotreatingcatalysts. Applied Catalysis A: General, 134(2):299 – 317, 1996.

[101] Jenny M. Lewis and Ronald A. Kydd. The MoO3-Al2O3 interaction: Influ-ence of phosphorus on MoO3 impregnation and reactivity in thiophene HDS.Journal of Catalysis, 136(2):478 – 486, 1992.

[102] S. Kushiyama, R. Aizawa, S. Kobayashi, Y. Koinuma, I. Uemasu, andH. Ohuchi. Effect of addition of sulphur and phosphorus on heavy oil hy-drotreatment with dispersed molybdenum-based catalysts. Applied Catalysis,63(1):279 – 292, 1990.

[103] R. Iwamoto and J. Grimblot. Genesis, characterizations and HDS activity ofMo-P-alumina-based hydrotreating catalysts prepared by a sol-gel method. InB. Delmon G.F. Froment and P. Grange, editors, Hydrotreatment and Hydro-cracking of Oil FractionsProceedings of the 1st International Symposium/6th

BIBLIOGRAPHY 57

European Workshop, volume 106 of Studies in Surface Science and Catalysis,pages 195 – 210. Elsevier, 1997.

[104] Stephan M.A.M. Bouwens, Jan P.R. Vissers, Vincent H.J. de Beer, and RoelPrins. Phosphorus poisoning of molybdenum sulfide hydrodesulfurization cat-alysts supported on carbon and alumina. Journal of Catalysis, 112(2):401 –410, 1988.

[105] Mingyong Sun, Daniele Nicosia, and Roel Prins. The effects of fluorine, phos-phate and chelating agents on hydrotreating catalysts and catalysis. CatalysisToday, 86(1-4):173 – 189, 2003. Effects of Support in Hydrotreating Catalysisfor Ultra-clean Fuels.

[106] M. Villarroel, P. Baeza, F. Gracia, N. Escalona, P. Avila, and F.J. Gil-Llambias. Phosphorus effect on CoMo and NiMo synergism in hydrodesul-phurization catalysts. Applied Catalysis A: General, 364(1-2):75 – 79, 2009.

[107] A Griboval, P Blanchard, E Payen, M Fournier, J.L Dubois, and J.RBernard. Characterization and catalytic performances of hydrotreatment cat-alysts prepared with silicium heteropolymolybdates: comparison with phos-phorus doped catalysts. Applied Catalysis A: General, 217(1-2):173 – 183,2001.

[108] Pablo Torres-Mancera Gustavo Marroquin Samir K. Maity Patricia Rayo,Jorge Ramirez and Jorge Ancheyta. Hydrodesulfurization and hydrocrackingof Maya crude with P-modified NiMo/Al2O3 catalysts. Fuel, 100:34 – 42,2012. ISAHOF 2011 Feed and Processes for the Production of Clean Fuels.

[109] Fang Liu, Shaoping Xu, L. Cao, Y. Chi, T. Zhang, , and Dongfeng Xue. Acomparison of NiMo/Al2O3 catalysts prepared by impregnation and coprecip-itation methods for hydrodesulfurization of dibenzothiophene. The Journalof Physical Chemistry C, 111(20):7396 – 7402, 2007.

[110] S. Kasztelan, E. Payen, H. Toulhoat, J. Grimblot, and J.P. Bonnelle. Pro-ceedings of the climax fifth international conference industrial MoO3-promoteroxide-γ-Al2O3 hydrotreating catalysts: genesis and architecture description.Polyhedron, 5(1):157 – 167, 1986.

[111] Mou Jian, , and Roel Prins. Kinetics of the hydrodenitrogenation of dec-ahydroquinoline over NiMo(P)/Al2O3 catalysts. Industrial & EngineeringChemistry Research, 37(3):834–840, 1998, doi =.

[112] Qiherima, Huifeng Li, Hui Yuan, Yunhong Zhang, and Guangtong Xu. Ef-fect of alumina support on the formation of the active phase of selectivehydrodesulfurization catalysts CoMo/Al2O3. Chinese Journal of Catalysis,32(1-2):240 – 249, 2011.

58 BIBLIOGRAPHY

[113] Jorge Ramirez Antonio Garcia-Vila, Rogelio Cuevas-Garcia and Ivan Puente-Lee. Effect of phosphorus on Mo/Al2O3 catalysts for Maya crude improve-ment. Catalysis Today, 220-222:310 – 317, 2014. International Symposium onAdvances in Hydroprocessing of Oil Fractions (ISAHOF 2013).

[114] Alla Spojakina Kvetuse Jiratova Radostina Palcheva, Ludek Kaluza andGeorgi Tyuliev. NiMo/γ-Al2O3 catalysts from ni heteropolyoxomolybdateand effect of alumina modification by b, co, or ni. Chinese Journal of Catal-ysis, 33(4-6):952 – 961, 2012.

[115] Lianglong Qu, Weiping Zhang, Patricia J Kooyman, and Roel Prins. MASNMR, TPR, and TEM studies of the interaction of NiMo with alumina andsilica-alumina supports. Journal of Catalysis, 215(1):7 – 13, 2003.

[116] P. Arnoldy, J. C. M. De Jonge, and J. A. Moulijn. Temperature-programedreduction of molybdenum(VI) oxide and molybdenum(IV) oxide. The Journalof Physical Chemistry, 89(21):4517–4526, 1985.

[117] S. Rajagopal, H.J. Marini, J.A. Marzari, and R. Miranda. Silica-alumina-supported acidic molybdenum catalysts - TPR and XRD characterization.Journal of Catalysis, 147(2):417 – 428, 1994.

[118] J.A. Marzari, S. Rajagopal, and R. Miranda. Bifunctional mechanism ofpyridine hydrodenitrogenation. Journal of Catalysis, 156(2):255 – 264, 1995.

[119] Yun-qi Liu Shu-na Han Yong-ming Chai Tong-na Zhou, Hai-liang Yin andChen guang Liu. Effect of phosphorus content on the active phase structure ofNiMoP/Al2O3 catalyst. Journal of Fuel Chemistry and Technology, 38(1):69– 74, 2010.

[120] J.L. Brito and J. Laine. Reducibility of NiMo/Al2O3 catalysts: A TPR study.Journal of Catalysis, 139(2):540 – 550, 1993.

[121] Joaquín L. Brito, Jorge Laine, and Kerry C. Pratt. Temperature-programmedreduction of NiMo oxides. Journal of Materials Science, 24(2):425–431.

[122] H. Kraus and R. Prins. The effect of phosphorus on oxidic NiMo(CoMo)/γ-Al2O3catalysts: A solid state NMR investigation. Journal of Catalysis,170(1):20 – 28, 1997.

[123] I. Eswaramoorthi, V. Sundaramurthy, Nikhil Das, A.K. Dalai, and J. Adjaye.Application of multi-walled carbon nanotubes as efficient support to NiMohydrotreating catalyst. Applied Catalysis A: General, 339(2):187 – 195, 2008.

[124] R.C. Ryan, R.A. Kemp, J.A. Smegal, D.R. Denley, and G.E. Spinnler. Stack-ing of molybdenum disulfide layers in hydrotreating catalysts. In M.L. Occelliand R.G. Anthony, editors, Hydrotreating CatalystsPreparation, Characteriza-tion and PerformanceProceedings of the Annual International AlChE Meeting,

BIBLIOGRAPHY 59

volume 50 of Studies in Surface Science and Catalysis, pages 21 – 40. Elsevier,1989.

[125] N. Esquivel J.L.G. Fierro, A. Lopez Agudo and R. Lopez Cordero. Effect ofphosphorus on molybdenum-based hydrotreating catalysts. Applied Catalysis,48(2):353 – 363, 1989.

[126] A. Stanislaus, M. Absi-Halabi, and K. Al-Doloma. Effect of phosphorus onthe acidity of γ-alumina and on the thermal stability of γ-alumina supportednickel-molybdenum hydrotreating catalysts. Applied Catalysis, 39:239 – 253,1988.

[127] Abena Owusu-Boakye, Ajay K. Dalai, Deena Ferdous, and John Adjaye. Ex-perimental and kinetics studies of aromatic hydrogenation in a two-stage hy-drotreating process using NiMo/Al2O3 and NiW/Al2O3 catalysts. The Cana-dian Journal of Chemical Engineering, 84(5):572–580, 2006.

[128] A. R. Beltramone, D. E. Resasco, W. E. Alvarez, and T. V. Choudhary. Simul-taneous hydrogenation of multiring aromatic compounds over NiMo catalyst.Industrial & Engineering Chemistry Research, 47(19):7161–7166, 2008.

[129] Mou Jian and Roel Prins. The effect of phosphorus on the HDN reaction ofpiperidine, decahydroquinoline and ortho-propylaniline over NiMoS2/Al2O3catalysts. Catalysis Letters, 35(3):193–203, 1995.

Documents

PreparationandEvaluationofSulﬁded NiMo γ Al2O ...953658/FULLTEXT01.pdf · surface area and on NiMo catalyst was prepared using γ-alumina with lower ... pore diﬀusion which prevent