Xinzhuang Zhang,Yunda Han,Dapeng Li,Zhanguo Zhang,Xiaoxun Ma,9,

1 School of Chemical Engineering,Northwest University,Xi’an 710069,China

2 Research Institute of Shaanxi Yanchang Petroleum (Group) Co.,Ltd.,Xi’an 710065,China

3 Chemical Engineering Research Center of the Ministry of Education (MOE) for Advanced Use Technology of Shanbei Energy,Xi’an 710069,China

4 Shaanxi Research Center of Engineering Technology for Clean Coal Conversion,Xi’an 710069,China

5 Collaborative Innovation Center for Development of Energy and Chemical Industry in Northern Shaanxi,Xi’an 710069,China

6 International Scientific and Technological Cooperation Base of the Ministry of Science and Technology (MOST) for Clean Utilization of Hydrocarbon Resources,Xi’an 710069,China

7 Hydrocarbon High-efficiency Utilization Technology Research Center of Shaanxi Yanchang Petroleum (Group) Co.,Ltd.,Xi’an 710065,China

8 Key Laboratory on Resources Chemicals and Materials,Shenyang University of Chemical Technology,Shenyang 110142,China

9 Longdong University,Qingyang 745000,China

Keywords:Attrition Mo/HZSM-5 Fluidized-bed Catalyst activation Methane dehydro-aromatization

ABSTRACT As a potential methane efficient conversion process,non-oxidative aromatization of methane in fluidized bed requires a catalyst with good attrition resistance,especially in the states of high temperature,longtime rapid movement and chemical reaction.Existing evaluation methods for attrition resistance,such as ASTM D5757 and Jet Cup test,are targeted for fresh catalysts at ambient temperature,which cannot well reflect the real process.In this study,spherical-shaped Mo/HZSM-5 catalyst prepared by dipping and spray drying was placed in a self-made apparatus for attrition testing,in which the catalyst attrition under different system temperatures,running time and process factors was investigated with percent mass loss(PML),particle size-mass distribution(PSMD)and scanning electron microscope(SEM).Carbon deposition on the catalyst before and after activation,aromatization and regeneration was analyzed by thermogravimetry(TG),and the attrited catalysts were evaluated for methane dehydro-aromatization (MDA).The results show that the surface abrasion and body breakage of catalyst particles occur continuously,with the increase of system temperature and running time,and make the PML rise gradually.The process factors of activation,aromatization and regeneration can cause the catalyst attrition and carbon deposits,which broaden the PSMD in varying degrees,and the carbon-substances on catalysts greatly improve their attrition resistance at high temperature.Catalyst attrition has a certain influence on its catalytic performance,and the main reasons point to particle breakage and fine powder escape.

1.Introduction

The technique of direct catalytic conversion of methane to aromatics and hydrogen under oxygen-free environment is considered as a very significant way for the chemical utilization of natural gas[1,2].In the past 27 years,researchers have made many knowledge and achievements in terms of catalytic reaction mechanism [3–5],catalyst synthesis and modification [6–10],process selection and optimization [11–14],reactor design [15–18] and so on,although there is still no pilot scale production.

The non-oxidative aromatization of methane is a strong endothermic reaction,and higher reaction temperature (generally above 700 °C) is favorable.However,the catalyst deactivation is very fast at high temperature,so the regeneration must be introduced for extending the service life of the catalyst.Moreover,the system pressurization is theoretically disadvantageous due to the increase of molecules number during the reaction,so the use of high space velocity is an inevitable choice to improve absolute yield.An effective solution to the above problems is to use the fluidized bed reactor with catalyst regeneration function,but it also brings about catalyst attrition like as FCC process [19],while the cause of attrition may include mechanical stress [20,21],thermal stress [22,23] and chemical stress.

Catalyst activation,methane conversion on catalyst and catalyst regeneration are the three main processes of methane dehydroaromatization (MDA) in fluidized bed.Molybdenum carbide[3,24,25] or MoOxCy[26],rather than the molybdenum oxide in fresh Mo/HZSM-5,is considered to be the key active substance for MDA,so it is necessary to activate the catalyst using raw gas(containing CH4)at a certain temperature to improve its catalytic activity and shorten or eliminate the reaction induction period[27–29].Carbon deposition that causes catalyst deactivation can be removed by hydrogen reduction [30,31],and the catalytic performance can hold stable in the periodic CH4-H2switching operation [12,32,33].Interestingly,the carbon deposition may play a double role [34],that is,it is also beneficial to the fluidized bed MDA.

Previously,most of attrition tests for the micro-spherical catalyst particles were carried out at room temperature using ASTM fluidized bed [21,35–40],conical/cylindrical jet cup [39,41–44]and other testing devices [45–48],and in which the fine particles collected were measured and analyzed to characterize the attrition resistance of catalyst and infer the attrition mechanism.

Reppenhagen and Werther [20] defined the mass of abrasionproduced fines per unit mass of catalyst entering the cyclone as the attrition rate,and stated that it was proportional to the surface mean diameter of catalyst and to the square of cyclone inlet velocity,while inversely proportional to the square root of solids-to-gas loading ratio [49],and believed that the sphericity and surface smoothness of catalyst particles were positively related to their attrition resistance.However,Wu et al.[21]thought that the hardness and mechanical strength of catalyst particles were the dominant factors,while the surface properties such as shapes and surface roughness had little effect on the attrition resistance,but can affect the attrition mechanism,resulting in the fragmentation of particles with irregular shapes and the abrasion of particles with rough surfaces.

Whitcombe et al.[22]stated that the thermal shock produced by mixing of cold and hot catalysts resulted in the formation of fine particles and metal rich aerosols,which may be an important cause of excessive FCC catalyst emissions.Hao et al.[23]studied the influence of temperature and time on the attrition of MTO catalyst using a three-orifice air jet apparatus,and believed that the abrasion and surface fragmentation coexisted at room temperature,and the abrasion dominated when rise to 500°C,then generating a large quantity of superfine powders which were hard to be captured,and the time to achieve an equilibrium attrition rate was also greatly shortened.In the subsequent jet cup attrition test for MTO catalyst,Hao et al.[50] found that the test can get quantitative results close to those obtained by high velocity gas jets experiments in[23],while significantly shorten the test time and reduce the temperature corresponding to maximum attrition index.Li et al.[51] studied the particle attrition in a 500 °C fluidized bed assembled with supersonic convergent-divergent nozzles,and believed that the particle fragmentation at high temperature dominated the attrition,which can significantly increase the grinding efficiency.

Wu et al.[21,37]stated that the attrition of pellet type catalysts in fluidized bed system resulted from a mixture of the particle fragmentation at the early nonsteady-state stage and the surface abrasion at the later steady-state stage.Jet attrition and bubble-induced attrition were two main types of catalyst attrition in ASTM fluidized bed,and contribution of the former to total attrition was proportional to the superficial gas velocity,while the minimum gas velocity for generating the latter was far larger than the minimum fluidizing velocity because the attrition required energy consumption[38].Zhang et al.[52]had a similar view,and stated that the fast motion and intensive mixing of particles in a fluidized bed created inter-particle collision and bed-to-wall impacts,which led to the particle attrition and related to composition of bed material and collected carryover.Cabello et al.[53] reported that the AJI (Air Jet Index) -ASTM test,not the crushing strength and AJI,was suitable to predict the attrition behavior of oxygen carrier particles subjected to long-term CLC(Chemical Looping Combustion)operation.

Vasireddy et al.[54] stated that the addition of Fe,Cu and K increased the attrition loss of spent FCC support,and the Febased catalyst pretreated with CO at 280°C showed better attrition resistance than that before pretreatment.

In actual industrial production,catalyst fines due to attrition may clog process pipelines,resulting in system overpressure or even shutdown.And the fines may enter final products because of poor separation or recycling,resulting in product contamination and increased purification costs.Furthermore,catalytic performance of catalysts should be affected by attrition,resulting in reduced production capacity and increased production costs.In our earlier research on continuous CH4dehydroaromatization-H2regeneration fluidized bed process [55],it was found that a small amount of Mo/HZSM-5 catalyst fines was deposited in the gasphase outlet pipeline after an eight-hour test,and the catalytic performance declined due to the reduction of absolute amount of catalyst in the system.

Therefore,it is necessary to study the fluidizing attrition of Mo/HZSM-5 catalyst and its effect on the catalytic performance.However,there is no relevant research report yet.Thus,in present work,the influence of operation conditions and several process factors on the attrition of Mo/HZSM-5 catalyst was investigated in a self-made evaluation apparatus,and the catalytic performance of catalyst before and after attrition was tested and compared.

2.Experimental

2.1.Catalyst samples and characterization

Fresh Mo/HZSM-5 catalysts used in this study were prepared by dipping commercial HZSM-5 powder(SiO2/Al2O3=30,S.A.=400 m2-.g-1,Zeolyst International)with an aqueous solution of ammonium heptamolybdate ((NH4)6Mo7O24.4H2O) at 80 °C,followed by spray drying of the resulting concentrate samples and then calcining at 550 °C for 5 h in air.Three batches of Mo/HZSM-5 catalysts were prepared,labeled as Cat-A,Cat-B and Cat-C respectively,and their basic informations are listed in Table 1.Surface area and pore volume of catalysts were measured using the nitrogen adsorption/desorption method by Autosorb-1-C (Quantachrome Instruments,USA).Mean particle size of catalysts was obtained with Laser Particle Size Analyzer S3500 (Microtrac Inc.,USA).Mo content of catalysts was confirmed by Inductively Coupled Plasma-Optical Emission Spectrometer (ICP-OES) Optima 7000 DV (PerkinElmer,USA).

The three batches of Mo/HZSM-5 catalysts have similar physical properties (surface area,pore volume,apparent bulk density and mean particle size) and Mo mass content (about 6.0%),shown in Table 1,which indicates a good repeatability of preparation method we adopted for Mo/HZSM-5 catalysts.Particle size-mass distribution (PSMD) of catalysts obtained by standard sieving is shown in Fig.1,and it is clear that most of catalyst particles are in the range of 125–280 μm.In order to eliminate the effect of differences in particle size on the test result analysis and amplify attrition effect,the Cat-B with a size range of 125–280 μm was selected as subsequent testing sample.

Table 1Physical properties and Mo content of fresh Mo/HZSM-5 catalysts

2.2.Process factors for catalyst samples

Process factors of activation,aromatization,and regeneration were completed in a self-made fluidized bed evaluation system,as shown in Fig.2,which consisted of feeding unit,fluidized bedreaction unit,on-line analysis unit and exhaust gas purification unit.

Fig.1.Particle size-mass distribution of fresh Mo/HZSM-5 catalysts.

2.2.1.Activation

Fresh Cat-B was placed on the airflow distribution plate in the fluidized bed reactor (I.D.=50 mm) and then the system was heated to 650 °C at a speed of 10 °C.min-1in a N2flow.Subsequently,the feed was switched to active gas (V(CH4):V(Ar)=9:1)at 650 °C,and the Cat-B began to be activated.Simultaneously,on-line analysis unit started to work.When the benzene peak in gas chromatograph (equipping with a Chemipak PH packed column and a FID detector) reached the maximum,the activation was completed.Then the feed was switched to N2,and the system stopped heating and cooled down naturally.Taking out the activated Cat-B for testing after the system dropped to room temperature.

2.2.2.Aromatization

Fresh Cat-B was activated according to 2.2.1.When the activation was completed,the system continued to be heated to 800 °C at a rate of 10 °C.min-1in a N2flow.Then the feed was switched to reacting gas (V(CH4):V(Ar)=9:1) at 800 °C,and the activated Cat-B began to catalyze CH4aromatization.Simultaneously,online analysis unit started to work.After 2 h,sampling and heating were stopped and the feed was switched to N2for cooling.The deactivated Cat-B was collected for subsequent tests.

Data characterizing catalytic properties of Cat-B samples,such as methane conversion rate (MCR),aromatics selectivity (AS) and aromatics formation rate(AFR),were calculated in accordance with the method from document [56].

2.2.3.Regeneration

Fresh Cat-B was treated according to 2.2.2.After 2 h aromatization,the feed was switched to H2at 800 °C,and the deactivated Cat-B began to be regenerated.After 3 h reduction reaction,heating was stopped and the feed was switched to N2for cooling.The regenerated Cat-B was collected at room temperature.

2.3.Attrition testing

The self-made evaluation apparatus used for attrition tests,as shown in Fig.3,mainly consists of a heating furnace,a fluidized bed reactor,a cyclone separator and a fines collector.Furnace wire temperature of the heating furnace can reach up to 1200 °C.The fluidized bed reactor is mainly a quartz tube (I.D.=50 mm) with an intake pipe (I.D.=10 mm) connected at the bottom,an airflow distribution plate made of sintered quartz at the lower part of the tube,and a top flange cover passed through two catalyst delivery pipes and a thermocouple sleeve.The cyclone separator and its inlet and outlet connecting pipes are all made of quartz and surrounded with electric heating belts.The fines collector made of quartz is connected with the gas outlet pipe of cyclone separator through the flange,and is internally embedded a micro quartz fiber thimble,which can intercept the powders of above 0.03 μm in size.

Prior to attrition tests,the Cat-B samples were dried at 120 °C for 2 h and cooled to room temperature in a desiccator.Put～50 g dry sample on the airflow distribution plate,and then dry compressed air,N2or CH4/Ar (0.4 m3.h-1) passed through the fluidized bed reactor from bottom to top.Meanwhile,some fluidized Cat-B particles were drawn into the internal riser and lifted into the cyclone separator,most of which were separated in the cyclone and fell back into the reactor,and a small amount of fine Cat-B particles that were not separated continued to be elevated and then be trapped by the thimble in fines collector.The fines were collected and measured every 6 h,and the longest attrition test lasted 30 h.The percent mass loss(PML)of Cat-B samples in fluidizing attrition tests is as follows [39]:

where m0is the mass of the thimble at the start of attrition test(g),msis the mass of the catalyst sample charged to apparatus(g),mtis the mass of the thimble after t h attrition test (g).

2.4.Particle size-mass distribution (PSMD) analysis

The PSMD of Cat-B particles(<125 μm)trapped in the collector and stayed in the reactor after 6 h attrition of 25°C,200°C,400°C,600 °C and 800 °C were measured with standard sieves and electronic balance.Besides,the PSMD of fresh Cat-B,activated Cat-B,deactivated Cat-B and regenerated Cat-B were analyzed and compared.The mass loss of Cat-B samples in the course of transfer,sieving and weighing was neglected.

Fig.2.Schematic diagram of fluidized bed evaluation system for solid granular catalysts.

Fig.3.Schematic representation of self-made attrition testing apparatus.

2.5.Thermogravimetric analysis (TGA)

The TGA of fresh Cat-B,activated Cat-B,deactivated Cat-B and regenerated Cat-B were performed by DSC/TG simultaneous thermal analyzer (NETZSCH STA 449F3,Germany).The sample(～15 mg) was placed in an Al2O3pan and heated to 900 °C in 30 ml.min-1air at 10 °C.min-1.

2.6.Scanning electron microscope (SEM) analysis

The surface morphology of fresh attrition-free Cat-B particles and typical attrited Cat-B particles were observed and analyzed by SEM (Hitachi TM-3000/SU-8000,Japan).

3.Results and Discussion

3.1.Effect of system temperature on attrition

The PML of Cat-B samples after 6 h air attrition at 25°C,100°C,200°C,300°C,400°C,600°C and 800°C respectively was investigated,as shown in Fig.4.The attrition loss of Cat-B unceasingly increases with system temperature rising,and the PML at 800 °C is up to 16.66%.And the PML increment of Cat-B at adjacent temperature points presents a U-shaped change process of decreasing first and then increasing,from 4.94% (25–100 °C) to 0.46% (300–400 °C),then to 5.28% (600–800 °C).Obviously,200 °C and 600 °C are two key temperature points,and the PML increases rapidly below 200 °C and above 600 °C,while increases slowly between 200 °C and 600 °C.Therefore,it is inferred that there are two different attrition mechanisms,one dominates the Cat-B particles attrition at <200 °C,the other dominates the attrition at>600 °C.And it is indicated that the catalyst attrition cannot be ignored in the process of high temperature(above 600°C)fluidization and circulation.

Fig.4.PML and PML increment of Cat-B after 6 h air attrition at 25 °C,100 °C,200 °C,300 °C,400 °C,600 °C and 800 °C,respectively.

After 6 h air attrition at different system temperatures,many broken Cat-B particles with a wide size range were captured by the fines collector,shown in Fig.5(a).The proportion of fine broken particles (<63 μm) in the PML is the highest at 25 °C,and the proportion of larger broken particles in the PML gradually increases with system temperature rising.The higher system temperature is,the more obvious this trend is.This phenomenon is consistent with the viewpoint in document [51],which considers that the particles’ fragmentation dominates the attrition at high temperature,but is different from that in document [23],which believes that the MTO catalyst is more prone to abrasion and generating superfine powders when temperature rises to 500 °C.

As shown in Fig.5(b),some broken Cat-B particles still remain in the reactor after 6 h air attrition at 25 °C.However,as system temperature rises,smaller broken particles gradually disappear,and residual particles only have a narrow size range,such as 98–125 μm at 600 °C and 800 °C,and the proportion of that is less and less.In combination with Fig.4 and Fig.5(a),it is inferred that these Cat-B particles vanished from the reactor are all intercepted into the collector because of efficiency decline of the cyclone separator at high temperature.

Fig.5.PSMD of Cat-B after 6 h air attrition at 25°C,200°C,400°C,600°C and 800°C,respectively.(a)Contributions of different size particles in the fines collector to the PML of Cat-B.(b) PSMD of residual Cat-B particles (<125 μm) in the fluidized bed reactor after attrition.

3.2.Effect of running time on attrition

The influence of running time on the PML of Cat-B at 25 °C,200°C,300°C and 400°C respectively was investigated,as shown in Fig.6.It can be calculated that the PML per h at 25°C is 0.32%in the first 6 h,and near linear increment is about 0.16% per h in the next 24 h.

As you can see from Fig.6,the PML increases obviously with running time after system temperature rising,and experiences a process,from slow increase to obvious jump,and then to steady.And for this jump phase,the starting point in time gradually moves forward,from 18th h at 200 °C and 300°C in advance to 12th h at 400 °C,while the jump increment decreases gradually,from 4.2% at 200 °C down to 3.1% at 400 °C.This is in line with the standpoint in document [23],which thinks that raising temperature can shorten the time of achieving an equilibrium attrition rate.It can be inferred that the PML will continue to increase slowly and nearly linearly at 25 °C and slowly to equilibrium at 200–400 °C.

Fig.6.PML variation of Cat-B with running time at 25°C,200°C,300°C and 400°C,respectively.

3.3.Effect of process factors on attrition

3.3.1.PSMD change of catalyst samples

According to the methods in Section 2.2,fresh Cat-B was treated of activation,aromatization and regeneration respectively,and the PSMD of treated Cat-B was obtained,as shown in Fig.7.It is clear that the PSMD of Cat-B has changed,and in particular,a small number of larger particles appear.Moreover,the color has changed from white (fresh Cat-B) to bright black (treated Cat-B).

Compared with fresh Cat-B,after activation,the mass percent of the particles of 125–150 μm and 200–280 μm decreases by 7.38%and 0.61% respectively,while that of 150–200 μm increases by 7.38%,and a small amount of larger particles of 280–355 μm and smaller particles of 98–105 μm and 105–125 μm appear.Accordingly,it is inferred that the size of some Cat-B particles of 125–150 μm and 200–280 μm rises into 150–200 μm and 280–355 μm respectively,meanwhile,some Cat-B particles of 200–280 μm are broken to smaller particles of 98–105 μm and 105–125 μm.

Fig.7.PSMD of Cat-B before and after activation,aromatization and regeneration.

A small amount of larger particles of 280–450 μm and smaller particles of 90–125 μm appear in the deactivated Cat-B.Compared with fresh Cat-B,the mass percent of the particles of 125–150 μm and 200–280 μm decreases by 8.06%and 2.73%respectively,while that of 150–200 μm increases by 5.93%.Compared with activated Cat-B,after aromatization,more particles of 98–125 μm and 280–355 μm are produced,and some new particles of 90–98 μm and 355–450 μm are emerged.Thus,it is speculated that the size of some Cat-B particles of 125–150 μm,200–280 μm,and 280–355 μm increases into 150–200 μm,280–450 μm,and 355–450 μm respectively,and some Cat-B particles of 150–280 μm,especially of 150–200 μm,are stripped or broken to smaller particles of 90–125 μm.Besides,it is also possible that a few smaller particles are broken to fine ones.

Regenerated Cat-B experienced 3 h hydrogen reduction has the same particle size range as deactivated Cat-B,but the mass percent of the particles of 150–200 μm,280–355 μm and 355–450 μm decreases by 4.77%,0.36%and 0.21%respectively,and that of other sizes increases to some extent.Therefore,it can be deduced that the size of some Cat-B particles of 150–200 μm and 280–450 μm decreases to 125–150 μm and 200–280 μm respectively,and some larger Cat-B particles are stripped or broken into smaller particles of 98–105 μm.

To sum up,the change of color and size of Cat-B particles are two main macroscopic phenomena caused by activation,aromatization and regeneration.Color transformation from white to black indicates that carbon particles or black carbon-containing substances are formed on the surface or in the channels of Cat-B particles.It is inferred from MDA process that activation of molybdenum oxide to molybdenum carbide,methane decomposition and methane deep aromatization are the three main sources,and which may increase the size of Cat-B particles.While smaller particles are probably generated by thermal stress,chemical stress and physical attrition,all of which may lead to surface denudation and bulk fragmentation of Cat-B particles.In addition,the regeneration of deactivated Cat-B with H2can remove part of carbonaceous deposits and reduce part of molybdenum carbide to molybdenum[30],which may minish the size of some Cat-B particles,but the PSMD cannot be restored to the state of fresh Cat-B.

3.3.2.Carbon accumulation amount of catalyst samples

TGA of white fresh Cat-B and black treated Cat-B in air is shown in Fig.8,and the mass change of Cat-B can be divided into four phases,according to the key inflection point in the curve.Lowtemperature volatile substances adsorbed by Cat-B,such as water and organics,are removed in the phase one,and its cut-off temperature is about 192 °C.In addition,the order of mass loss is:fresh Cat-B >activated Cat-B >deactivated Cat-B >regenerated Cat-B,which indicates that process reaction may affect adsorption performance,the deeper the Cat-B is treated,the less its adsorption amount is.

In the phase two,some substances in Cat-B are combined with oxygen (oxidized) and the mass of Cat-B increases gradually.Except for regenerated Cat-B,catalyst mass all reaches its maximum at 416 °C.The oxidation of molybdenum,molybdenum carbide and heavy hydrocarbons are three possible mass gain reactions.It is speculated that the mass increase of fresh Cat-B is mainly due to the oxidation of molybdenum to molybdenum oxide,while that of activated Cat-B and deactivated Cat-B is mostly caused by the oxidation of molybdenum carbide to molybdenum oxide and heavy hydrocarbons to macromolecular compounds.The amount of unstable substances in regenerated Cat-B is less but more active,and the mass of that first rises,then decreases,and then rises again.Oxidation of molybdenum and heavy hydrocarbons may result in the mass increase,while the mass loss may be caused by hydrocarbon combustion.

Fig.8.TGA Curves of fresh and treated Cat-B in air.

In the phase three,Cat-B’s mass loss is probably caused by the oxidation combustion of carbon particles and carbonaceous deposits.White fresh Cat-B has almost no mass change for the absence of carbon substances.Activated Cat-B contains a little carbon substances and therefore has slight mass loss.Both deactivated Cat-B and regenerated Cat-B have significant mass loss,reaching 6%and 4%respectively,indicating that they contain more carbon substances.It can be seen that hydrogen reduction may remove a small amount of carbon substances (about 2%) from deactivated Cat-B.

In the phase four,the mass of four Cat-B samples decreases obviously at >775 °C,presumably due to the sublimation loss of molybdenum oxide.According to the trend of mass loss,it is inferred that absolute amount of molybdenum oxide in fresh Cat-B and regenerated Cat-B is larger than that in activated Cat-B and deactivated Cat-B,which indicates that molybdenum in the latter two exists mainly in the form of combining with carbon and hydrogen,rather than molybdenum oxide.Furthermore,the mass loss of deactivated Cat-B and regenerated Cat-B occurs earlier,presumably the existence of carbon substances profits molybdenum oxide sublimation.

3.3.3.Attrition testing of treated samples

The treated Cat-B particles of 125–280 μm were screened and taken for the 30 h attrition test at 25 °C and 400 °C respectively.High purity nitrogen was used as fluidizing gas to prevent treated Cat-B from being oxidized by air (O2).The results of attrition test are shown in Fig.9.

The PML of treated Cat-B is significantly lower than that of fresh Cat-B at 25 °C and 400 °C,and the PML order at each measuring point is:regenerated Cat-B >deactivated Cat-B >activated Cat-B.Similarly,it is found in document [54] that the Fe-based catalyst pretreated with CO at 280°C shows better attrition resistance than that before pretreatment.Among the treated samples,the activated Cat-B shows the best attrition resistance,and its maximum PML in 30 h are 0.42% (25 °C) and 0.47% (400 °C) respectively.Besides,same as fresh Cat-B,the PML of treated Cat-B increases gradually with running time,but the PML increment from 6th h to 30th h is all lower.The PML and its increment of fresh,deactivated and regenerated Cat-B at 400 °C are greater than that at 25 °C.In particular,the PML of activated Cat-B at 400 °C is higher than that at 25 °C,but its PML increment is the opposite.

Based on the above data analysis,combined with the analysis in Section 3.3.2,it can be inferred that the molybdenum carbide and carbonaceous deposits produced by activation make activated Cat-B particles harder and smoother,and the attrition resistance is greatly enhanced.The brittle coatings,consisting of carbonaceous deposits generated from high temperature aromatization,are formed on the surface of deactivated Cat-B particles,and strong thermal and chemical stress make the particles’ body expand.The attrition for a short time will induce partial brittle coatings to fall off,and the swelling surface will also be eroded partially by the long time attrition.High temperature H2regeneration makes partial smooth carbonaceous coatings of deactivated Cat-B particles disappear,and more uneven surfaces are exposed.Meanwhile,the thermal stress promotes further expansion of Cat-B particles to produce more surfaces of loose and brittle.These two changes cause more small fragments of regenerated Cat-B in 25°C and 400°C attrition tests,and make the PML increase significantly relative to deactivated Cat-B.However,the attrition resistance of regenerated Cat-B remains be enhanced by the residual carbonaceous deposits,and its PML is still significantly lower than that of fresh Cat-B.

The 6 h N2attrition tests for treated Cat-B at 200°C,600°C and 800°C were supplemented,as shown in Fig.10.The PML of treated Cat-B gradually increases with system temperature rising,and the PML order at three temperature points is the same as that at 25°C and 400°C.Different from the PML variation of fresh Cat-B,the initial and terminal temperatures of PML slowly increasing stage of treated Cat-B decrease from 200 °C and 600 °C to 25 °C and 400°C,respectively.It is inferred that the main cause of mass loss of Cat-B at 200°C is still the surface abrasion,same as that at 25°C.When system temperature is higher than 400 °C,the larger particles of treated Cat-B generated from fragmentation will be further eroded or broken,and many smaller particles produced by that will be trapped in the collector to increase the PML.Although the Cat-B particles treated by activation,aromatization and regeneration show better attrition resistance and can reach the steady attrition at a lower temperature,their ability to withstand high temperature thermal stress decreases significantly.

Fig.9.PML variation of fresh and treated Cat-B at 25 °C and 400 °C with running time.

Fig.10.PML of fresh and treated Cat-B after 6 h attrition at 25 °C,200 °C,400 °C,600 °C and 800 °C,respectively.

The effect of real reacting gas(V(CH4):V(Ar)=9:1)on fluidizing attrition of Mo/HZSM-5 catalyst was investigated with fresh and treated Cat-B at 800°C.High purity nitrogen was used as fluidizing medium in the heating process.The attrition test results are listed in Table 2.

At the MDA temperature,the feed gas has a certain impact on the catalyst attrition,as shown in Table 2.The PML of activated and deactivated Cat-B increases,while that of fresh and regenerated Cat-B decreases.The chemical stress produced by the aromatization reaction and the coke deposition on the catalyst are the two possible influencing factors.At 800 °C,activation,aromatization and carbon deposition will occur on the fresh and regenerated Cat-B,and the resulting carbon substances on outer surfaces and internal channels enhance their attrition resistance and make the PML in test reduce,similar to the phenomenon in Fig.10.There are still some carbon substances in regenerated Cat-B that cannot be reduced by H2,as the conclusion in Section 3.3.2,so the newly generated carbon substances of regenerated Cat-B are less than that of fresh Cat-B,and the improvement for attrition resistance and the decrease for PML are also smaller than that of the former.The PML increase of activated and deactivated Cat-B may be mainly due to the chemical stress of aromatization and carbon formation,as well as the peeling of brittle carbonaceous surfaces or the breaking of expanded particles’body caused by strong thermal stress.This is in line with the inference on attrition mechanism in Fig.9.

Table 2PML of fresh and treated Cat-B after 6 h fluidizing attrition at 800 °C

3.4.Effect of attrition on particle morphology

Fig.11 shows typical SEM images of Cat-B samples before and after attrition.Fresh Cat-B particles are uniform spherical-shaped and have complete surfaces and clear boundaries,as shown in Fig.11(a).

The fresh Cat-B particles staying in the reactor after 6 h air attrition at 25°C,shown in Fig.11(b),present two damage signs.Some particles’ surfaces have been eroded into uneven defects highlighted with red circles,and very few particles are broken into larger blocks highlighted with blue circles.This indicates that the surface abrasion of particles contributes significantly to the PML of Cat-B at room temperature,and gives a good explanation for a lower PML in Fig.4 and a higher proportion of fine particles in Fig.5(a).

After 6 h air attrition at 800 °C,fresh Cat-B particles left in the reactor are severely broken except that a few of them remain intact highlighted with yellow arrows,as shown in Fig.11(c).This indicates that the body breakage of particles dominants the attrition of Cat-B at high temperature,which is consistent with the standpoint in document [51] and explains the results of a higher PML in Fig.4 and a higher proportion of larger particles in Fig.5(a).

Fig.11(d) shows the image of fresh Cat-B particles trapped in the collector after 30 h air attrition at 400 °C.The crushed blocks with different sizes and a large number of fines occupy most of the space,but there are also some spherical blocks highlighted with yellow circles and individual larger particle of surface abra-sion highlighted with a red circle.This indicates that the attrition for a long time at a higher temperature will lead to more Cat-B particles to break up,and produce massive fragments and fines which make the PML rise continuously and rapidly.

Fig.11.SEM images of typical Cat-B samples.(a)Fresh Cat-B(125–280 μm).(b)Fresh Cat-B after 6 h air attrition at 25°C.(c)Fresh Cat-B after 6 h air attrition at 800°C.(d)Fresh Cat-B fines trapped in the collector after 30 h air attrition at 400 °C.(e) Activated Cat-B after 6 h N2 attrition at 200 °C.(f) Deactivated Cat-B after 30 h N2 attrition at 25 °C.(g) Single ball particle of fresh Cat-B.(h) Regenerated Cat-B fines trapped in the collector after 6 h N2 attrition at 800 °C.

Only a few particles of surface abrasion (highlighted with red circles) are found in activated Cat-B staying in the reactor after 6 h N2attrition at 200 °C,as shown in Fig.11(e).While a small amount of surface attrite particles (highlighted with red circles)and body broken particles (highlighted with blue circles) can be seen in deactivated Cat-B staying in the reactor after 30 h N2attrition at 25°C,as shown in Fig.11(f).The results in Fig.9 and Fig.10 that the process reaction of activation and aromatization can greatly improve the attrition resistance of Cat-B particles are further confirmed.

Single ball particle of fresh Cat-B is formed by accumulation of numerous flake or block zeolite particulates in different sizes,shown in Fig.11(g),and its surface is not smooth.Therefore,the particulates’properties of self-strength,binding force and accumulation regularity are all important factors affecting the attrition resistance of Cat-B particles.

Fig.11(h) shows the image of regenerated Cat-B fines intercepted by the collector after 6 h N2attrition at 800°C.The surface of individual larger particle is peeled off,generating some debris highlighted with red circles,and some small particles are further broken up,producing a few smaller fragments highlighted with blue circles.All these confirm the inference in Section 3.3.3.

3.5.Effect of attrition on catalytic performance

MDA evaluation was carried out for unworn Cat-B and attrited Cat-B respectively,according to the method in Section 2.2.2,and the results were compared and analyzed.

3.5.1.Fresh catalyst samples

Fig.12 shows the characteristic data results of MDA catalyzed by fresh Cat-B before and after air attrition.

The initial MCR of attrited Cat-B is higher than that of unworn Cat-B,but it reverses completely after about 15 min,as shown in Fig.12(a).The higher attrition temperature is,the faster MCR of attrited Cat-B decreases.Compared with unworn Cat-B,the instantaneous MCR of attrited Cat-B at 120th min is reduced by 0.39%,4.30% and 7.51% respectively.

Among fresh Cat-B samples,as shown in Fig.12(b),the variation trend of benzene selectivity (BS) presents obvious differences,while that of naphthalene selectivity (NS) is similar.For fresh unworn Cat-B and 2# attrited Cat-B,the BS tends to increase slowly after initial rapidly rising,and reaches the maximum of 67.86% and 66.49% at 120th min respectively.For 3# attrited Cat-B,the initial rate of BS growth is higher than fresh unworn Cat-B and 2# attrited Cat-B,and the BS gradually reaches an equilibrium value of about 65.59% with reaction time,except for a slight decrease after 105th min.For 4#attrited Cat-B,the BS gradually declines after initial rapidly climbing to the peak of 65.95%,and the instantaneous BS at 120th min is 56.74%.The NS of fresh Cat-B samples reaches its maximum at the beginning of the reaction,is 30.71%,29.54%,23.14%and 20.44%respectively,then gradually decreases and enters a stable stage.

Fig.12.Catalytic performance of fresh Cat-B for MDA.(a)Methane conversion rate with time.(b) Aromatics selectivity with time.(c) Aromatics formation rate with time.

AFR is the result of interaction between MCR and AS.The benzene formation rate (BFR) of fresh Cat-B samples climbs rapidly to the peak and then declines gradually due to MCR’s reduce irreversibly,as shown in Fig.12(c),while the variation tendency of naphthalene formation rate (NFR) is similar to that of NS.The maximum of BFR is 4682.33 nmol-C.g-1.s-1of 4# attrited Cat-B,and that of NFR is 2225.68 nmol-C.g-1.s-1of fresh unworn Cat-B.

Fig.13.Catalytic performance of activated Cat-B for MDA.(a) Methane conversion rate with time.(b) Aromatics selectivity with time.(c) Aromatics formation rate with time.

Based on the above data analysis,it can be concluded that the high temperature attrition of Mo/HZSM-5 catalyst makes its catalytic performance for MDA worse.It is inferred that the smaller particles produced by surface abrasion and body breakage make the contact surfaces between catalysts and methane enlarge and the internal diffusion resistance decrease,which cause the rise of MCR,BS and BFR of attrited Cat-B at the initial stage of the reaction.However,as the fine attrited particles are carried away from the fluidized bed reactor by the flow,and the mass of catalyst has been reduced,so its catalytic performance decreases rapidly at the later stage.In addition,the active sites of catalyst may be covered by more carbon deposition,thus accelerating the deactivation.

3.5.2.Activated catalyst samples

Fig.13 shows the characteristic data results of MDA catalyzed by activated Cat-B before and after N2attrition.

It is very clear that the catalytic performance of activated Cat-B samples in Fig.13 is better than that of fresh Cat-B samples in Fig.12 under similar conditions of attrition and reaction.The MCR of 5#sample decreases by 7.43%within 120 min,which is less than the reduction of 11.03%of 1#sample,and compared with 5#sample,the instantaneous MCR of attrited samples of activated Cat-B at 120th min is reduced by 0.89%,3.06% and 5.61%respectively.

Within the first 20 min of the reaction,the BS of activated Cat-B samples increases rapidly to the maximum,as shown in Fig.13(b),and then decreases gradually at different rates.The order of maximum BS is:8#sample(72.54%)>6#sample(71.16%)>5#sample(70.75%)>7#sample(67.26%),which are all larger than the corresponding values of fresh Cat-B samples in Fig.12(b),and compared with 5#sample,the BS of attrited samples at 120th min is reduced by 3.18%,8.77% and 17.82% respectively.The NS variation and maximum NS are similar to that of fresh samples.

The AFR of activated Cat-B samples shows better stability and smaller descending value per unit time than fresh Cat-B samples.The BFR of activated samples at 120th min is higher than that of fresh samples,while the NFR is opposite.

It can be seen from above analysis that the particle attrition of activated Mo/HZSM-5 still has an important effect on its MDA catalytic performance,and the higher the attrition temperature,the worse the catalytic performance.Smaller particles in attrited samples benefit to increase initial MCR,short the time to reach to maximum BS and improve maximum BFR,but also lead to the rapid decline of MCR,BS and BFR in the remaining time,and inhibit the formation of naphthalene.

4.Conclusions

System temperature and running time are two important operating factors that affect the attrition of spherical-shaped Mo/HZSM-5 catalyst in a fluidized bed or circulating fluidized bed.The higher the system temperature and the longer the running time,the more serious the catalyst attrition.Surface abrasion dominates below 200 °C,mainly making fine broken particles,while body breakage dominates above 600 °C,mostly producing larger broken particles.The largest PML under experimental conditions is 16.66% of the fresh catalyst that experiences 6 h fluidization attrition at 800 °C.In the activation,aromatization and regeneration,the chemical reactions and long-time fluidization at high temperature make catalysts erode or break into the smaller particles,while the carbon deposition on catalysts raise their size and greatly improve their attrition resistance.The activated catalyst has the best attrition resistance,and its PML after 6 h fluidization attrition at 800 °C is only 3.01%.Hydrogen regeneration at 800 °C can remove some specific higher temperature type carbon substances on the deactivated catalyst,but weakens its attrition resistance.Although the smaller particles from surface abrasion and body breakage can boost the initial activity of the catalyst for MDA,part of them escape from the system,which makes the mass of catalyst in the reaction zone decrease,thus leads to a faster descent of catalytic performance in the middle and late stages.Therefore,it suggests necessity of improving the attrition resistance of spherical-shaped Mo/HZSM-5 catalyst appropriately,in order to promote the future MDA application.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by Hydrocarbon High-efficiency Utilization Technology Research Center of Shaanxi Yanchang Petroleum (Group) Co.,Ltd.,China (Contract No.HCRC-C13-010),and the National Natural Science Foundation of China (No.21536009).The authors also would like to acknowledge Northwest University and Shaanxi Yanchang Petroleum (Group) Co.,Ltd.for affording the opportunity to participate in relevant research projects,and appreciate Jialiang Gao and Gen Zhang for their help in SEM analysis of catalysts,Baoqiang Wu for his help in TGA of catalysts.