Zhibin Deng,Xing Ge,Wenting Zhang,Shizhong Luo,*,Jun Shen,Fangli Jing,*,Wei Chu

1 College of Civil Aviation Safety Engineering,Civil Aviation Flight University of China,Guanghan 618307,China

2 Sinopec Petroleum Exploration and Production Research Institute,Northwest Oilfield Branch,Urumqi 830011,China

3 School of Chemical Engineering,Sichuan University,Chengdu 610065,China

Keywords: Ethylene SBA-15 Pore size effect Cr catalysts Heterogeneous catalysis

ABSTRACT SBA-15 with varied pore size from 4 to 8 nm were synthesized by tuning the temperature of hydrothermal treatment,the supports were then used to load the active phase CrOx through a conventional impregnation method.The resulting catalysts were characterized by small/wide angle XRD,N2 adsorption/desorption,FT-IR,TEM-EDX,XPS,TPR and CO2-TPD to study the feature of structure,surface chemical state,redox and basicity.It was found from these results that the metal species could be well dispersed on catalysts with larger pore size.Cr6+species could enter into the framework by substituting the Si atoms of SBA-15,and Cr3+mainly exist on extra framework.Pore size had profound effects on reducibility,surface composition and basicity.Cr6+species were necessary to activate the C-H bonds of alkanes,while the basicity played an important role in activating C-O bonds of CO2.The best performances were achieved over the sample Cr supported on SBA-15 with a pore diameter of 7 nm in oxidative dehydrogenation of ethane in the presence of CO2.

1.Introduction

The fossil resources like coal,oil and natural gas are still the majority in present energy structure,supplying directly/indirectly the most living staffs we need.Unfortunately,the exploitation,transportation and use of them usually bring negative impact on environment.The emission of CO2accompanies with the all process,and its continuously accumulation in atmosphere aggravate the well-known“greenhouse effect”,which would further have profound influences on climate change[1,2].It becomes very necessary to lower the concentration of CO2in atmosphere through developing the techniques such as carbon dioxides capture,utilization and storage [3–6].On the other hand,the shale gas revolution promotes the rapid development of exploit technique,the large quantity of reserve provides abundant but underutilized light alkanes specially methane and ethane [7,8].Dry reforming of methane with CO2gives a solution that could consume the two major greenhouse gases simultaneously,the resulting syngas could be used to synthesize the light olefins and other chemicals via Fisher-Tropsch synthesis as a dependence of catalyst.As far as ethane is concerned,it is more attractive to convert it to valuable ethylene through oxidative dehydrogenation reaction using CO2as soft oxidant,comparing to the dry reforming of ethane–the reaction extent should be controlled thanks to a selective catalyst.

CO2exhibits gentler oxidative ability comparing with pure oxygen or air,avoiding the full oxidation of ethane to generate COx,and lower the safety risks because of its chemical inert at ambient conditions comparing with SO2and N2O[9].Moreover,the introduction of CO2is thermodynamically favorable to the oxidative dehydrogenation rather than the non-catalytic thermal dehydrogenation process.In additional,it is good to eliminate the coke through Boudouard reaction.

Oxidative dehydrogenation of ethane with CO2,like the other selective oxidation of light alkanes,follows the classic redox type Mars van Krevelen mechanism[10].The Cr-based catalysts are mostly studied because Cr oxides are active and reducible,the variable chemical valents ensures oxidation–reduction cycle work well[11–14].Deng et al.[15]investigated the promotor effect on Ni,Fe,Co and Mn modified Cr/ZrO2catalysts.The results showed that the addition of Fe,Co and Mn was good to improve the selectivity of ethylene,unlike Ni modified sample over which the reforming reaction is preferred,leading to a high selectivity of CO.Ni element exhibited its super primary role in determining reaction extent due to the strong ability in cleaving C-C and C-H bonds.PtNi-[16,17]and FeNi-[18–20]systems performed similarly although the conversions of reactants were relative high.Concerning the support,Al2O3[21],ZrO2[22]and the SiO2composites[23]were used to disperse the active phase Cr species.However,the performance was not so good as that on SiO2supported catalysts,which was evidenced by Wang [24]and Baiker [25].The different types of zeolites supported Cr catalysts such as Cr/MCM-41 [26,27],(Cr-)Cr/MSU-x[28],Cr/ZSM-5[27,29],Cr/Zr-ZSM-5[14]and Cr/SAPO-34[30]were widely studied.SBA-15 is a very typical mesoporous material with regular pore structure,the SBA-15 itself,the framework modified[31,32]and surface modified[33–36]SBA-15 are widely studied as catalyst or catalyst support.One of the most advantages of this material is that the pore structure could be easily tuned by changing the preparation parameters.The differences in pore structure and composition would cause the differences in Cr species dispersion,redox and acid–base property,which would further affect the activity of CO2oxidative dehydrogenation of ethane and catalytic stability.

Pore size effect would lead to the modification of redox which has been proved in our previous work [37].In this work,a series of SBA-15 with varied pore diameter were synthesized and then used to disperse Cr species by the conventional wetness impregnation method.Various techniques were applied to characterize the physicochemical properties of supports and catalysts,the correlation between the properties and catalytic performances of CO2oxidative dehydrogenation of ethane would be further discussed.

2.Experimental

2.1.Catalyst preparation

The SBA-15 supports with different pore was synthesized by the welldeveloped hydrothermal synthesis [38],the pore size was tuned by changing the temperature of hydrothermal treatment.The crystallization temperatures were set to 70°C,100°C,120°C and 140°C to get the SBA-15 with different average pore diameters x,which were therefore denoted as SBA-15@x.

SBA-15 supported Cr catalysts were prepared by impregnation method.Appropriate amount of Cr(NO3)3·9H2O was dissolved in deionized water to form a uniform solution,the support was then added under vigorous stirring.The solvent was evaporated in 80 °C water bath after impregnating for 8 h,the resulting solid was dried at 80°C for 12 h and finally calcined at 550°C for 4 h in static air.The catalysts with 10%(by mass)of Cr2O3were denoted as Cr/SBA-15@x.

2.2.Catalyst characterization

The surface area,average pore size and total pore volume of catalysts were measured by N2adsorption/desorption isotherms on Quantachrome Nova 1000e apparatus.The Brunauer–Emmett–Teller(BET)method and by Barrer–Joyner–Halenda (BJH)method were used to calculate the surface area and pore size,respectively.The catalysts were degassed at 400°C for 4 h prior to measurement.

The crystalline phases were identified by X-ray diffraction technique on Philips X'pert Pro MPD diffractometer (CuKα,40 kV,40 mA,λ=0.15406 nm).The small angle diffraction patterns were scanned from 0.5°to 5°and the wide ones were recorded from 10°to 80°.

Transmission electron microscopy(TEM)analysis was performed on Philips Tecnai G2F20 field-emission gun transmission electron microscope at the accelerate voltage of 200 kV.

The surface property was checked by X-ray photoelectron spectroscopy(XPS)on a XSAM 800 spectrometer using Al-Kα radiation(1486.6 eV).The C1s peak of adventitious carbon was fixed to 284.5 eV as a reference.

H2temperature programmed reduction was performed on TP-5080 apparatus equipped with a thermal conductivity detector (TCD) to study the reducibility.In a typical test,50 mg of catalyst was treated in N2flow(30 ml·min?1)at 300°C for 3 h to remove water and other impurity,and then heated to 800°C with a heating rate of 5°C·min?1in H2-Ar mixture(5.0%H2in Ar).

CO2temperature programmed desorption(CO2-TPD)was carried out on Micromeritics Autochem II 2920 system equipped with a Utype reactor and a thermal conductivity detector (TCD).50 mg of sample was pretreated in He flow(50 ml·min?1)at 450°C for 2 h,then cooled down to 50 °C and saturated in mixture 5% CO2in He(25 ml·min?1)which was detected by TCD.The desorption was performed from 100°C to 900°C in He flow(50 ml·min?1)at a ramp of 10°C·min?1.

Infrared spectra(FT-IR)were analyzed on a Thermo Nicolet iS50.The catalyst(1%,by mass)was mixed with KBr and then pressed to pellets,and the spectrum was collected from 400 to 4000 cm?1with a spectral resolution of 4 cm?1and 256 scans.

2.3.Catalytic evaluation for ethane dehydrogenation in the presence of CO2

CO2oxidative dehydrogenation of ethane reaction was operated in a fixed-bed quartz reactor (i.d.: 8 mm,length: 200 mm) under atmospheric pressure.Typically,0.1 g of catalyst diluted with 2.0 g quartz sand (0.25–0.42 mm) were loaded in the isotherm zone of reactor.The catalyst bed was heated to 650°C in N2flow(30 ml·min?1)with a heating rate of 10°C·min?1,then exposed to the reactants mixture(nC2H6:nCO2=1:1,flowrate=12 ml·min?1).The gaseous products including propylene,CO and unreacted feedstock were analyzed by an online GC equipped with packed Porapak Q(3 mm o.d.,3 m length)and molecular sieves (3 mm o.d.,3 m length) columns in series and thermal conductivity detector(TCD).

3.Results and Discussion

3.1.Structure identification for support

The prepared samples were first tested by small angel X-Ray Diffraction(XRD)technique to identify the structure,and the results were depicted in Fig.1.The diffraction plane(100)together with the planes(110)and(200)confirmed the feature of 2-demensional hexagonal p6mm symmetry.When the hydrothermal temperature increased from 70°C(sample a)to 140°C(sample d),the(100)diffraction peak shifted to lower angle side from 1.075° to 0.886°,leading to the dspacing increased from 8.2 nm to 10.0 nm(see Table 1)according to Brag's equation[39].

Table 1Textural properties by N2 adsorption/desorption and d-spacing by XRD for the supports

The pore structure of prepared SiO2was further analyzed by N2adsorption/desorption technique,the collected isotherms and pore size distributions were shown in Fig.2.All the samples exhibited type IV isotherms according to IUPAC classification,which was the typical characteristic of mesoporous materials[40].While the shape of hysteresis loop was strongly dependent on hydrothermal temperature,samples a and b hydrothermally treated at 70 and 100 °C showed H2type,samples c and d hydrothermally treated at 120 and 140 °C showed H1 type,indicating that the higher temperature is good to the formation of uniform pore structure.The evident changes were from the pore size distributions(Fig.2B),a narrow distribution could be seen for all samples and the zone shifted to higher side as the increased hydrothermal temperature.

Fig.1.Small angle XRD patterns for the support with different pore diameter(a:SBA-15@4,b:SBA-15@5,c:SBA-15@7 and d:SBA-15@8).

The quantification for textural properties including specific surface area,pore volume and average diameter were listed in Table 2,showing that the average diameters for the samples hydrothermally treated at 70,100,120 and 140 °C were 3.7,5.2,6.8 and 8.1 nm,respectively.The supports were therefore correspondingly denoted as SBA-15@4,SBA-15@5,SBA-15@7 and SBA-15@8.As far as specific surface area was concerned,it increased from 560 m2·g?1for SBA-15@4 to 590 m2·g?1for SBA-15@7,then decreased significantly to 387 m2·g?1for SBA-15@8.All the results suggested that the 2-D hexagonal phase could be obtained for the all samples,and the pore size could be easily tuned by changing the hydrothermal temperature.

3.2.Porosity of the supported catalysts

After loading the active phase CrOxon SBA-15 with different pore diameter,the catalysts were denoted as Cr/SBA-15@4,Cr/SBA-15@5,Cr/SBA-15@7 and Cr/SBA-15@8,the isotherms and pore size distributions were gathered in Fig.3,no obvious changes could be found,implying that the pore structure was not modified through introducing the active phase CrOx.However,the specific surface,pore volume and average diameters for each sample declined by different extent from Table 2.Taking into account that 10%(by mass)of Cr2O3was loaded on support,the variations of textural properties changed within 5%except for Cr/SBA-15@4 over which the pore filling was easier to occur because the pore size was much smaller.It also suggested that support with bigger pore size was good to disperse the metal species.

Table 2Textural properties of the supported catalysts

3.3.Morphology

TEM measurement provided a direct view on catalyst morphology,the images for the supported catalysts were gathered in Fig.4.All the catalysts exhibited the well-ordered tubular pores,which was good agreement with the results of small angle XRD and N2adsorption/desorption measurements.It could be also found that the supported Cr species did not give a very uniform dispersion,especially for the sample with smaller pore size support.In addition,the pore structure of Cr/SBA-15@8 distorted at some extent,implying that the pore might partially collapse,which was likely responsible for the sharp decrease of specific surface area.

The element analysis was performed on Cr/SBA-15@7 sample,the selected zone and results was shown in Fig.5.The signals of O and Si from the support SBA-15 were strong,while the Cr element exhibited very weak peak at ca.5.4 KeV,which was caused by their quite low content.The quantification results further confirmed it,giving a molar ratio of 1%,and mass ratio of 2.4% which was around one fourth of designed content,suggesting only small quantity of Cr species entered into the pore,and most of them dispersed on external surface as shown in Fig.4c.

Fig.2.(A)Isotherms and(B)pore size distributions for the supports(a:SBA-15@4,b:SBA-15@5,c:SBA-15@7 and d:SBA-15@8).

Fig.3.(A)Isotherms and(B)pore size distributions for the catalysts(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7 and d:Cr/SBA-15@8).

Fig.4.TEM images for the supported catalysts(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7 and d:Cr/SBA-15@8).

3.4.Crystalline phase analysis

The wide angle XRD analysis was carried out to check the crystalline phases of supported catalysts,the patterns were presented in Fig.6.All the samples exhibited the similar diffraction pattern in the range from 10°to 80°.The broad peak at around 23°originated from the support SBA-15.The strong diffraction peaks at 24.5°,33.6°,36.5°,41.5°,50.2°,54.9°,63.4°and 65.1°were assigned to Cr2O3[PDF#38-1479][41].

FT-IR is widely used to identify the structure of compounds,it is applied to check the Cr species here and the collected spectra were depicted in Fig.7.The bands at 463 cm?1,806 cm?1,and 1084 cm?1were attributed to the bending vibration,symmetric stretch vibration,asymmetric stretch vibration of Si-O-Si.The bands at 1636 cm?1and 3421 cm?1were assigned to Si-OH silanol groups which could bind water molecule by hydrogen bonds.The hydroxylated covalent framework in SBA-15 could stabilize the(O)3-Si-OH and(O)2-Si-OH silanol groups on surface,which was helpful to graft chromia species.The bands at 571 cm?1and 622 cm?1originated from the hydrated CrOxspecies existing on pore surface,more precisely,the band at 571 cm?1was considered to be the characteristic of hydrated Cr2O3.The band at 960 cm?1was usually caused by the lattice defect and was related to the presence of metal cations in siliceous framework.In our cases,it could be assigned to Si-O-Cr6+vibration.Moreover,the weak band at 905 cm?1was due to the vibration of Cr-O or Cr=O in Cr6+species[42].These results demonstrated that the Cr cations could be incorporated into the siliceous framework of SBA-15 via an impregnation method and exist as Cr6+species like chromate and bichromate.The both cations had close ionic radius(44 pm for Cr6+vs 40 pm for Si4+),which was good to replace the Si4+cations in framework of SBA-15 for Cr6+.Most Cr2O3together with very limit CrO3existed in extra framework.

Fig.5.Element analysis on selected sample Cr/SBA-15@7.

Fig.6.XRD patterns for the Cr/SBA-15 catalysts(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7 and d:Cr/SBA-15@8).

3.5.Reducibility

The redox property was studied by hydrogen temperature programed reduction and the recorded hydrogen consumption profiles were depicted in Fig.8.The reduction process started at around 200°C for the all samples,which could be roughly divided into two stages at ca.400 °C.The reduction of polychromate and monochromate occurred below 400 °C where the Cr6+species were reduced to Cr3+species[43].The hydrogen consumption occurring at higher than 400 °C corresponded the reduction of Cr3+to Cr2+or lower valent.On the other hand,the incorporated Cr6+species in framework of SBA-15 strongly interacted with the support,leading to the reduction required higher temperature comparing with the ones locating in the pore or on the surface.

Fig.7.FT-IR spectra for the Cr/SBA-15 catalysts(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7 and d:Cr/SBA-15@8).

The hydrogen consumption was quantified and the results were reported in Table 3.The total hydrogen consumption showed an increasing trend from 0.15 to 0.23 mmol/g cat.when the pore size increased from 4 to 7 nm,then declined to 0.19 mmol/g cat.for sample Cr/SBA-15@8.All the catalysts had the close hydrogen consumption(ca.0.9 mmol/g cat.)in high temperature stage(400–650°C).Therefore,the consumed hydrogen in low temperature zone(200–400°C),corresponding to the reduction of Cr6+species in pore or on surface,had the same volcano shape trend as the total hydrogen consumption.It suggested that more amount of reducible Cr6+species presented in Cr/SBA-15@8.Besides these,the whole reduction process also confirmed the FT-IR conclusion that Cr6+and Cr3+species coexisted in catalysts.

Table 3Quantified hydrogen consumption from TPR measurement

3.6.Surface analysis by XPS

Fig.8.Reducibility of the catalysts determined by TPR(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7 and d:Cr/SBA-15@8).

XPS analysis was performed to check the surface chemical state of Cr species,the spectra were collected in Fig.9.Fore overlapped peaks could be found in the binding energy(BE)range from 595 eV to 570 eV for the all catalysts.The two peaks centering at ca.580.0 eV and ca.576.6 eV corresponded to the Cr6+and Cr3+species[41],respectively.The one centering at ca.586.6 eV was considered as a satellite peak of Cr3+species,which appeared in the Cr2p1/2zone.However,the one centering at lower binding energy side of ca.573 eV was really hard to define.The proportion of Cr6+and Cr3+species were quite different according to the quantified results in Table 4,although the both of them existed on catalyst surface.The surface content of Cr6+showed a minimum value of 27.1%for Cr/SBA-15@4,it increased to 42.8%for Cr/SBA-15@5 and reached to the maximum of 51.0%for Cr/SBA-15@7,finally dropped down to 46.8%for Cr/SBA-15@8,showing a strong dependence on pore size.

Taking the spent Cr/SBA-15@7 catalyst (Fig.9e) as an example,the spectra of Cr2p still contained the four overlapped peaks,indicating that the surface property did change significantly.The difference was from the surface composition of Cr species,the Cr6+species decreased by ca.21% to 40.1%,suggesting that the reoxidation of Cr6+become difficult which was probably linked to the activation of CO2.

3.7.CO2-TPD measurement

Acid–base property has an important influence on activation of light alkanes,which would affect the catalytic activity and selectivity for oxidative dehydrogenation reaction.Deng et al.found that the strong acidic sites was helpful to activate the C-C bonds,and promoting the cracking and dry reforming reactions when different promotors were added into Cr/ZrO2system,resulting in the loss of ethylene selectivity.CO2was introduced into the reaction system as soft oxidant,the activation of it became an important factor to affect the catalytic performances.The CO2temperature programmed desorption was thus carried out to check the adsorption/desorption behaviors and to analyze the basicity of catalysts.The recorded desorption profiles were reported in Fig.10,showing the similar desorption process for the all samples.Three desorption stages could be roughly identified:i)the desorption due to physisorption occurred between 50 and 100°C;ii)the desorption from weak basic sites between 100 and 450°C;and iii)the desorption from strong basic sites between 450 and 900°C.

As far as CO2uptake was concerned,it was clear from Table 5 that the sample Cr/SBA-15@7 showed the maximum value of CO2uptake(0.42 mmol/g cat.).The sum of weak and strong sites followed the decreasing order:Cr/SBA-15@7(0.41 mmol/g cat.) >Cr/SBA-15@5(0.35 mmol/g cat.)=Cr/SBA-15@2>(0.35 mmol/g cat.)>Cr/SBA-15@8(0.34 mmol/g cat.)excluding the physiosorbed CO2.Obviously,the sample Cr/SBA-15@7 could adsorbed a greater number of CO2molecules due to the more amount of strong basic sites,as close capacity of weak basic sites(ca.0.9 mmol/g cat.)was obtained for all samples.

3.8.Catalytic evaluation on CO2 oxidative dehydrogenation of ethane

The oxidative dehydrogenation of ethane using CO2as soft oxidant was operated at 650°C under atmospheric pressure and the reaction data were reported in Table 6.The conversions of ethane and CO2exhibited the same trend,increasing gradually as the pore size increased and reached to the maximum on Cr/SBA-15@7,then declined a little.It was necessary to note that the conversion of CO2was only about half comparing with that of ethane,suggesting that the activation of CO2was limited.As a result,the consumed lattice oxygen could not be fully replenished,meaning that the reduced Cr6+species could not be fully re-oxidized,this was confirmed by XPS findings.The pore size also showed its influences on product distribution,the selectivities of ethylene and methane followed the same volcano shape and got the maximum over the catalyst Cr/SBA-15@7,while the selectivity of CO gave the exact opposite trend.The results indicated that the activated CO which was left by CO2dissociation reacted with the adsorbed hydrogen to generate CH4on Cr/SBA-15@7 as it had the strongest ability in adsorbing CO2according to the CO2TPD measurements.

Table 4Chemical state of Cr species and their surface composition

Table 5Quantified CO2-TPD results

Table 6Catalytic performances on Cr/SBA-15@x samples

Fig.9.Cr2p XPS spectra for the different catalysts(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7,d:Cr/SBA-15@8 and e:spent Cr/SBA-15@7).

The changes of ethane conversion and ethylene selectivity with time on stream were depicted in Fig.11,a maximum initial conversion of ethane(29.2%)was obtained on Cr/SBA-15@7,the comparable values(ca.24%)were obtained over Cr/SBA-15@5 and Cr/SBA-15@8 samples,and the minimum value of 21.5% was obtained on Cr/SBA-15@4.Allthe catalysts underwent the progressive deactivation in the test period.Being different the conversion,the selectivity of ethylene for each sample showed the similar trends,increasing during the first 180 min and then kept roughly constant.

The correlations between the catalytic performances and the key properties such as surface Cr6+species content and basicity were presented in Fig.12.Apparently,the lattice oxygen carrier Cr6+species was directly linked to the conversion of ethane,as they shared the same trend,the higher content of Cr6+species was favorable to activate the alkane.On the other hand,the activation of CO2was dependent on the basicity of catalysts,the highest amount of basic sites existed in Cr/SBA-15@7,leading to a maximum conversion of CO2.Since the conversion of both ethane and CO2were promoted on Cr/SBA-15@7,the highest selectivity of ethylene was therefore achieved over it.These correlations demonstrated that the lattice oxygen content and number of basic sites played the key role in activating the C-H bonds in ethane and C-O bonds in CO2.

4.Conclusions

The SBA-15 support with varied pore diameter were successfully synthesized by tuning the hydrothermal temperature from 70 to 140°C,which was further used to load the active phase CrOx.The pore sizeof catalyst showed its influences on redox,basicity and surface chemical state.The Cr6+species mainly existed in the framework of SBA-15 by the substitution of Si,and rarely existed in extra framework in highly dispersion state as no diffraction peaks was found in XRD analysis,but was detected by FT-IR and XPS techniques.The TPR measurement also showed the isolated reduction processes of Cr6+and Cr3+species.CO2-TPD suggested that the sample Cr/SBA-15@7 exhibited the strongest capacity(0.41 mmol/g cat.)in adsorbing CO2,which could promote the conversion of CO2.At the same time,Cr/SBA-15@7 catalyst with the highest content of Cr6+species(51.0%)could provide the most lattice oxygen,which was helpful for the conversion of ethane.As a consequence,the best performances were gained on Cr/SBA-15@7 with highest conversions of ethane(25.8%)and CO2(13.8%),and highest selectivity of ethylene(81.0%).

Fig.10.CO2-TPD profiles for the different samples(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7 and d:Cr/SBA-15@8).

Fig.11.Ethane conversion and ethylene selectivity as a function of time on stream(a:Cr/SBA-15@4,b:Cr/SBA-15@5,c:Cr/SBA-15@7 and d:Cr/SBA-15@8).

Fig.12.Correlations between the catalytic performances and key properties(Cr6+ratio from XPS,basicity from CO2-TPD).

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

We would like to acknowledge the financial supports from National Natural Science Foundation of China (No.21603153),Science and Technology Department of Sichuan Province(No.2016HH0026)and the Fundamental Research Funds for the Central Universities (No.YJ201544).