Jiake Yang,Tongjiu Zuo,Jiangyin Lu

Key Laboratory of Oil &Gas Fine Chemicals,Ministry of Education,College of Chemistry and Chemical Engineering,Xinjiang University,Urumqi 830046,China

Keywords:Preparation methods Hydrocracking performance The NiMo catalysts MoO3 content

ABSTRACT In this work,NiMo catalysts with various contents of MoO3 were prepared through incipient wetness impregnation by a two-step method (NM-xA) and one-pot method (NM-xB).The catalysts were then characterized by XRD,XPS,NH3-TPD,H2-TPR,HR-TEM,and N2 adsorption–desorption technologies.The performance of the NiMo/Al2O3 catalysts was investigated by hydrocracking low-temperature coal tar.When the MoO3 content was 15 wt%,the interaction between Ni species and Al2O3 on the NM-15B catalyst was stronger than that on the NM-15A catalyst,resulting in the poor performance of the former.When the MoO3 content was 20 wt%,MoO3 agglomerated on the surface of the NM-20A catalyst,leading to decreased number of active sites and specific surface area and reduced catalytic performance.The increase in the number of MoS2 stack layers strengthened the interaction between Ni and Mo species of the NM-20B catalyst and consequently improved its catalytic performance.When the MoO3 content reached 25 wt%,the active metals agglomerated on the surface of the NiMo catalysts,thereby directly decreasing the number of active sites.In conclusion,the two-step method is suitable for preparing catalysts with large pore diameter and low MoO3 content loading,and the one-pot method is more appropriate for preparing catalysts with large specific surface area and high MoO3 content.Moreover,the NMxA catalysts had larger average pore diameter than the NM-xB catalysts and exhibited improved desulfurization performance.

1.Introduction

Hydrocracking is an effective method used to convert heavy raw feed into light oil and has been successfully industrialized and marketized worldwide;in this process,the performance of catalysts is an important factor[1,2].Commercial catalysts for hydrocracking heavy raw feed are prepared using the bimetallic combination of Mo(W)-CO(Ni) species supported on γ-Al2O3;such catalysts include CoMo/Al2O3,NiMo/Al2O3,and NiW/Al2O3.However,the CoMo catalyst is unsuitable for hydrocracking heavy raw feed,and the NiW catalyst is limited to utilization due to its high cost [3,4].The continuous exploitation and utilization of oil resources have led to heavy raw feed that contains more heavy components and higher sulfur content [5].In addition,international environmental protection regulations have been increasingly strict,especially with regard to restricting the sulfur content in light oil.Thus,the development of NiMo/Al2O3catalysts that are suitable for converting heavy raw feed into light oil has become a research hotspot [6–9].

The monolayer capacity of MoO3supported on Al2O3is 0.12 g MoO3/100 m2Al2O3(equal to about 20 wt%,specific surface area of Al2O3set to 200 m2﹒g-1) [10],and the content of MoO3in most commercial catalyst is 20 wt%.With the development of hydrocracking catalysts,scholars have found that the MoO3content has an optimum value that corresponds to excellent catalytic performance.Bian et al.[11] discussed the catalytic performance of KMo/Al2O3with different MoO3(5 wt%–45 wt%) contents.With increasing MoO3content,the interaction between K and Mo species was enhanced,but the dispersion of active metals on the carrier and the acidity of the catalyst decreased,thereby reducing the performance of the catalyst.Kouachi et al.[12]studied the effect of Mo/Al2O3catalysts with different MoO3contents (2 wt%–12 wt%)on the catalytic performance of the Biginelli reaction.As the MoO3content increased,Mo species and Al2O3formed Al2(MoO4)3phase,which changed the number of Br?nsted acid sites of the catalyst and led to improved catalytic performance.When the MoO3content exceeded 7 wt%,Mo species agglomerated on the carrier,thereby decreasing the amount of Al2(MoO4)3and reducing the catalytic performance.

Preparation method is also an important factor that affects the catalytic performance of catalysts.Different preparation methods could influence physical and chemical properties,such as the interaction between active metals and carrier,the pore structure,and the acidity of catalysts [13].Li et al.[14] pretreated commercial alumina by hydrothermal method and compared the hydrodenitrification performance of the supported NiMo catalysts.Compared with commercial alumina-supported catalyst,the pretreated alumina catalyst showed better dispersion of the active metals on the surface of catalyst,leading to improved reducibility and hydrodenitrification performance.In the study of Li [15],a series of MoO3/Al2O3catalysts was prepared by hydrothermal adsorption and incipient wetness impregnation.The catalyst prepared by hydrothermal adsorption had more MoS2formed,which prevented the formation ofand improved the hydrodesulfurization activity of the catalyst.Yin et al.[16] prepared the NiMo/HY-Al2O3catalyst by mechanical mixing and one-pot method and evaluated the performance of the catalyst on diesel desulfurization.More NiMoS phase was formed in the catalyst prepared by one-pot method,but the pore structure of this catalyst restricted its application in diesel desulfurization.

In our work,a series of catalysts with different contents of MoO3was supported on Al2O3(calcining pseudoboehmite at 500°C for 5 h)and pseudoboehmite to explore the influence of different preparation methods on the performance of NiMo/Al2O3catalysts.Catalytic performance was evaluated by hydrocracking lowtemperature coal tar (LTCT).Results provide references for development of catalysts for hydrocracking heavy oil.

2.Experimental

2.1.Catalysts preparation

NiMo/Al2O3catalysts were prepared by incipient wetness impregnation.In brief,(NH4)6Mo7O24﹒4H2O (Tianjin Chemical Reagent Co.Ltd.) and Ni (NO3)2﹒6H2O (Tianjin Damao Chemical Reagent Factory)were dissolved with a certain amount of distilled water and mixed to form a uniform solution in a flask.The flask was added with Al2O3(two-step method,calcining pseudoboehmite at 500 °C for 5 h) or pseudoboehmite (one-pot method,preparing catalysts in one-pot;Shangdong Zibo Sencha Fine Chemical Co.Ltd.),stirred at 40 °C for 12 h in a rotary evaporator,dried at 110°C for 2 h and calcined at 500°C for another 5 h with a heating rate of 8°C﹒min-1.A series of catalysts was obtained with 15,20,and 25 wt% MoO3(Ni/Mo=0.3) by changing the amount of (NH4)6Mo7O24﹒4H2O.The samples were marked as NM-xA(B) catalysts,where x refers to the MoO3content,A refers to the two-step method,and B was refers to the one-pot method.

The performance of the catalysts was improved by changing the form of active sites from metal oxide to metal sulfde.Ex situ presulfiding was performed as follows:(NH4)2S2O3(60 wt%,liquid) was loaded on NM-xA(B) catalysts (Mo/S=0.5,Ni/S=1) by incipient wetness impregnation;the sample was dried at 80 °C in vacuum and further dried at 50 °C for 6 h.

2.2.Catalyst characterization

X-ray diffraction(XRD)was employed to analyze the dispersion of catalysts.XRD pattern was recorded on D8 ADVANCE X-ray diffraction instrument of BRUKER company Germany(CuKα radiation)at 40 kV and 40 mA.Scanning was performed at a rate of 2°/min within 10°–80°.

N2adsorption was determined by ASAP2020 (Micromiritics Builder Tech company,America).The operating condition was controlled as follows:the samples were purified and degassed at 300 °C for 5 h at 15 μm Hg and analyzed by static adsorption method under N2atmosphere at -196 °C.Specific surface area was calculated by BET equation.Pore diameter and distribution were calculated by BJH method.The pore volume of micropores and mesopores were obtained by BJH adsorption cumulative curve.

Temperature programmed desorption (NH3-TPD) was carried out by using TP-5080(Tianjin xianquan instrument Co.Ltd.,China).The steps were as follows:Sample was heated to 400 °C for 1 h with the rate of 30 ml﹒min-1of helium gas.NH3adsorption was conducted within 0.5 h after the temperature was dropped to 120 °C.Then helium rate was maintained at 40 ml﹒min-1and heated temperature to 700 °C at a rate of 10 °C﹒min-1.

Temperature programmed reduction (H2-TPR) was carried out by using TP-5080.The steps were as follows:0.05 g sample was pretreated at 100 °C for 1 h under N2atmosphere and cooled to room temperature,then heated to 950 °C at a rate of 7 °C﹒min-1with 27 ml﹒min-1of H2/Ar mixed gas (=1/9).

X-ray photoelectron spectroscopy(XPS)analysis was conducted on ESCALAB 250Xi(Thermo Fisher Scientific Co.Ltd.,America)with Al K Alpha monochromator at 1361 eV.

High-resolution transmission electron microscopy (HR-TEM)analysis was performed with an acceleration voltage of 200 kV to investigate the structure of the catalysts (JEM-2100,Japan).The sample was well dispersed in anhydrous ethanol by ultrasonic waves.

2.3.LTCT hydrocracking

The catalytic performance of the catalysts was evaluated by hydrocracking LTCT in a bench-scale fix-bed reactor.The catalyst was placed in the middle section of the apparatus,and the ends of the apparatus were filled with inert ceramic balls.A leakage test was conducted before the reaction.The pressure of the reactor tube was increased to 10 MPa of N2and maintained for 12 h (pressure drop should be less than 0.01 MPa).If no leakage was detected,then the catalyst was sulfurized at 320 °C and 5 MPa for 30 min.Finally,the reaction condition was controlled as t=395 °C,p=8 MPa,WHSV=0.6 h-1,and H2/Oil=800.The gas products were analyzed by gas chromatograph every 60 min.

2.4.Analysis of raw materials and products

The density of the product was analyzed by hydrometer method.Sulfur content was determined by TSN-2000A.Gasoline fraction (<180 °C),diesel fraction (180 °C–320 °C),and residue fraction(>320°C)were separated by cutting distillation.The properties of heavy raw feed are listed in Table 1.

Table 1 Properties of feedstock

Gas products at one moment were analyzed by online gas chromatograph (SP-3420A).The test conditions were controlled as follows.The specification of the capillary column was 50 mm × 0.32 mm × 10 μm (KB-Al2O3/Na2SO4),and N2was used as carrier gas with the rate of 30 ml﹒min-1.The pre-column pressure was kept at 0.13 MPa,the detector temperature and injector temperature were controlled at 180 °C and 100 °C,respectively.The initial column temperature was kept at 50°C for 3 min,then increased to 100 °C with the rate of 10 °C﹒min-1,and maintained for 12 min.The amount of each hydrocarbon at different times was calculated according to the integral results.

2.5.The evaluation of catalytic performance

The residue conversion of LTCT and the sulfur content in gasoline fraction and diesel fraction were used as the main criteria for evaluating the catalytic performance of NiMo/Al2O3catalysts.The sulfur content of the products was determined by TSN-2000A,and the equation for residue conversion was as follows:

Residue conversion

3.Results and Discussion

3.1.State of active metals dispersed on the carriers

The XRD patterns of the NiMo/Al2O3catalysts are shown in Fig.1.No peak belonging to NiO was detected [17].The curve of the NM-20A catalyst showed two peaks at 23.5° and 26.7°,which belong to MoO3[18].This result indicated that bulk MoO3was formed on the surface of the NM-20A catalyst.However,no characteristic peaks of MoO3appeared in the curve of the NM-20B catalyst,demonstrating that MoO3was well dispersed on its surface.When the MoO3content reached 25 wt.%,the XRD patterns of NM-25A and NM-25B catalysts showed characteristic peaks of MoO3and NiMoO4[19];this finding indicates that MoO3agglomerated on the surface of the catalysts.The dispersion state of active metals on the carriers is shown in Fig.2.

3.2.Textural characteristics of NiMo/Al2O3 catalysts

Fig.1.XRD patterns of NiMo/Al2O3 catalysts.

Pore structure is an important parameter of hydrocracking catalysts and has a great influence on catalytic performance[20].The N2adsorption–desorption curves and pore size distribution of NiMo/Al2O3catalysts are shown in Fig.2.In Fig.2 (a),the N2adsorption–desorption curves of the catalysts showed type IV isotherm and exhibited H4 hysteresis loop(according to the standard of IUPAC).Hence,the catalysts had typical mesoporous structure.As shown in Fig.2 (b),the average pore diameter of NM-xB catalysts was 4.10–4.20 nm,and that of NM-xA catalysts was 4.80–5.40 nm.A large pore diameter is needed to improve the hydrodesulfurization performance of catalysts,especially for macromolecular sulfur compounds,such as 4,6-dimethyldibenzothiophene [21].The texture parameters of NMxA(B)catalysts are listed in Table 2.With increasing MoO3content,the specific surface area and average pore volume decreased significantly in NM-xA catalysts but remained relatively stable in NM-xB catalysts.This finding could be due to the aggregation of active metals on the surface of NiMo-xA catalysts and was consistent with the XRD results.

Table 2 Textural properties of NiMo/Al2O3 catalysts

3.3.NH3-TPD of NiMo/Al2O3 catalysts

Acidity is also an important factor because it provides acid sites to improve the cracking activity of a catalyst [22].The NH3-TPD patterns of NiMo/Al2O3catalysts are shown in Fig.3.The desorption temperature of NH3was above 350 °C,which corresponded to the formation of strong acid sites [23].As shown in Fig.3,NM-xB catalysts had more strong acid sites than NM-xA catalysts possibly due to the stronger interaction between the active metals and the carrier in the former.This phenomenon is conducive to the formation of strong acid sites [24].

3.4.H2-TPR of NiMo/Al2O3 catalysts

The H2-TPR curves of NiMo/Al2O3catalysts are shown in Fig.4.Two reduction peaks were found within 500°C–550°C and 700°C–900°C.The low-temperature reduction peak could be attributed to the reduction of Mo species with octahedral coordination(Mo6+→MO4+),whereas the high-temperature reduction peak was mainly due to the reduction of Mo species with highly dispersed tetrahedral coordination (MO4+→MO0),which has strong interaction with Al2O3[25,26].The H2-TPR curves showed that the reduction temperature of NM-xB catalysts was higher than that of NM-xA catalysts,indicating the stronger interaction between the active metals and the carrier in the former.This finding further confirmed the analysis of NH3-TPD.

Fig.2.Distribution of active metals on the surface of NiMo/Al2O3 catalysts.

Fig.3.NH3-TPD patterns of NiMo/Al2O3 catalysts.

Fig.4.H2-TPR patterns of NiMo/Al2O3 catalysts.

3.5.XPS of the sulfurized NiMo/Al2O3 catalysts

3.5.1.The XPS analysis of Mo 3d

XPS analysis was carried out to investigate the valence state of Mo species on the catalyst surface,and the results are shown in Fig.5 and Table 3.The Mo 3d spectra of the catalysts were decomposed into Mo4+(MoS2),Mo5+(MoOxSy),and Mo6+(MoOx) [27,28].The Mo4+content increased gradually with increasing MoO3content due to the increased layer number of Mo species.This phenomenon led to reduced interaction between Mo species and Al2O3and increased the Mo4+content.As shown in Table 3,catalysts prepared by different methods had no significant influence on the Mo4+content.However,the Mo5+content in NM-xB catalysts was higher than that in NM-xA catalysts due to the better dispersion of Mo species in the former [29,30].

3.5.2.The XPS analysis of Ni 2p

The Ni 2p3/2spectra of the sulfurized catalyst are shown in Fig.6 and Table 3.The peaks at (854 ± 0.3),(856.5 ± 0.3),(859.5 ± 0.3),and (863.4 ± 0.3) eV correspond to different interactions between the Ni species and the carrier [31].As shown in Table 3,when the MoO3content was 15 wt%,more Ni2+bspecies existed in the NM-xB catalyst.Ni species was more evenly distributed on the surface of NM-xB catalysts and led to the stronger interaction between the Ni species and the carrier.With increasing MoO3content,the Ni2+bcontent decreased because more Ni species interacted with Mo species.

Table 3 The Mo and Ni species valence distribution of NiMo/Al2O3 catalysts after sulfidation

3.6.HR-TEM of NiMo/Al2O3 catalysts

HR-TEM analysis was carried out on the sulfurized catalyst to observe the dispersion of MoS2on the surface of the catalyst.Fig.7 shows the HR-TEM photograph of the catalyst.The amount of MoS2increased with increasing MoO3content,consistent with the results of the XPS analysis.In addition,a large amount of MoS2aggregated on the surface of NM-25A and NM-25B catalysts,which may be caused by the aggregation of MoO3on the surface.This result corresponded to the XRD data.

Fig.5.Mo 3d analysis of NiMo/Al2O3 catalysts after sulfidation.

Fig.6.Ni 2p analysis of NiMo/Al2O3 catalysts after sulfidation.

Fig.7.HR-TEM photos of NiMo/Al2O3 catalysts after sulfidation.

The average length and stack layer number of MoS2were determined by statistical analysis [32].Table 4 shows that the average length and stack layer number of MoS2increased with increasing MoO3content.The increase in the average length of MoS2could increase the number of active sites directly,and the increase in the stack layer number of MoS2could promote the formation of the socalled type II structure of MoS2active sites (Fig.8).Both phenomena are conducive to improve the catalytic performance of the catalyst [33,34].However,the value of fModecreased with increasing MoO3content due to the aggregation of MoS2on the catalyst sur-face;this property is not conducive to improve the catalytic performance of the catalyst.Comparison of fMobetween NM-xA and NMxB catalysts showed that the MoO3dispersion in the catalyst prepared by the one-pot method was higher than that in the catalyst prepared by the two-step method.However,when the MoO3content reached 25%,both catalysts had the same fMovalue,which may be caused by the excessive MoO3content.

Table 4 The average length and stack layer number of NiMo/Al2O3 catalysts after sulfidation

The interaction between active metals and carrier was further explored by discussing the dispersion of the active metals on the catalyst surface.MoS2could be divided into two types of structure on the surface of the NiMo/Al2O3catalyst[36].The type I structure of Mo species has strong interaction with Al2O3because of the formation of Al-O-Mo bonds,leading to low hydrogenation activity.The type II structure of Mo species has weak interaction with Al2O3and owns strong reducibility.Fig.8 shows that when the stack layer number of MoS2is high,more type II Mo will be formed,thereby promoting the interaction between Mo and Ni species.This finding was also confirmed by the XPS and HR-TEM characterization results,which indicated that the contents of Mo5+and Ni2+bdecreased with increasing stack layer number of MoS2and that the interaction between active metals and Al2O3was weakened.When the MoO3content was as high as 25 wt%,the active metals aggregated on the surface of the catalysts,thereby decreasing the number of active sites directly.Therefore,the MoO3content should be controlled to own as much as the stack layer number of MoS2and as less as the clustered active metals.

3.7.The catalytic performance of NiMo/Al2O3 catalysts

NiMo/Al2O3catalysts were applied for hydrocracking LTCT,and the distribution of the products is shown in Table 5.The residue conversion of LTCT on NM-15A was higher than that on NM-15B.This finding could be due to two factors.First,the interaction between Mo species and Al2O3of NM-15B catalyst was stronger than that of the NM-15A catalyst and more Ni2+bspecies were found on the surface of the NM-15B catalyst,as a result of reducing the catalytic performance of the catalyst.These phenomena were consistent with the XPS,HR-TEM,and H2-TPR characterization results.Second,the NM-15A catalyst had larger average pore diameter,which was beneficial to the mass transfer of hydrocracking LTCT and improved the residue conversion of LTCT.When the MoO3content reached 20 wt%,the residue conversion of LTCT on the NM-20A catalyst decreased obviously.This finding could be explained by the fact that the agglomeration of MoO3on the surface of the NM-20A catalyst blocked the channel of the catalyst,thereby decreasing the pore volume and the number of active sites and reducing the LTCT residue conversion.By contrast,the LTCT residue conversion of the NM-20B catalyst increased,which could be due to two factors.First,more Ni species was reduced into Ni0,which weakened the interaction between the Ni species and the carrier and strengthened the interaction between Mo and Ni species (the Ni2+bcontent decreased substantially).This phenomenon promoted the formation of NiMoS phase.Second,the active metals of the NM-20B catalyst were uniformly dispersed on the carrier,which was helpful to increase the number of active sites,as a result of improving the performance of the NM-20B catalyst.When the MoO3content reached 25 wt%,the residue conversion of LTCT on NM-xA(B) catalysts decreased,which could be caused by the formation of bulk NiMoO4.This phenomenon decreased the number of active sites and specific surface area of the catalysts and consequently affected the residue conversion of LTCT.

Table 5 The products distribution of the NiMo/Al2O3 catalysts

The gasoline fraction and diesel fraction sulfur contents in NMxA catalysts were lower than those in NM-xB catalysts.This finding may be due to the larger average pore diameter of NM-xA catalysts than that of NM-xB catalysts.This property was conducive to the mass transfer and of macromolecular sulfur compounds and accelerated their decomposition rate.

Fig.8.The influence of the stack layer number of MoS2 on NiMo/Al2O3 catalysts.

4.Conclusions

A series of NiMo/Al2O3catalysts with different MoO3contents was prepared by one-pot method and two-step method.The properties of the catalysts were analyzed by XRD,N2adsorption,NH3-TPD,H2-TPR,HR-TEM,and XPS technologies.The performance of the catalysts was evaluated by hydrocracking LTCT.The results showed that the active metals were evenly dispersed on the surface of NM-15A and NM-15B catalysts.However,the strong interaction between the active metals and the carrier reduced the residue conversion of LTCT on the NM-15B catalyst.Meanwhile,the NM-15A catalyst possessed larger average pore diameter and exhibited improved catalytic performance.When the MoO3content was higher than 15 wt%,MoO3agglomerated on the surface of NM-xA catalysts and blocked their channel,resulting in poor catalytic performance.Hence,the two-step method was more suitable for preparing catalysts with low MoO3content.However,the active metals were evenly dispersed on the surface of NM-20B catalysts,which was beneficial to form more active sites.Compared with the two-step method,the one-pot method was more suitable for preparing catalysts with large specific surface area and high MoO3loading.In addition,the large average pore diameter of catalysts led to their improved desulfurization performance.

Acknowledgements

Financial support from the National Natural Science Foundation of China (21968034) is gratefully acknowledged.