Hao Sun,Saijian Shi,Zhenggui Gu*

School of Chemistry and Materials Science,Jiangsu Provincial Key Laboratory of Materials Cycling and Pollution Control,Nanjing Normal University,Nanjing 210023,China

1.Introduction

C10aromatics production has reached more than 300000 tons per year in China,of which alkyl naphthalene account for～30%.Among alkyl naphthalene,2-methylnaphthalene(2-MN)is quite useful as an intermediate in producing medicalssuch as Vitamin Kand 2,6-naphthalene dicarboxylic acid(a starting material for polyethylene naphthalene(PEN)with high heat resistance and tensile strength)[1–3].Although 1-methylnaphthalene(1-MN)has its application filed such as dyes,industrial demand for the 1-MN is much less than that of 2-MN.

Several processes for the simple separation of 2-MN from alkyl naphthalene have been reported[4],in which the mass ratio of 2-MN to 1-MN usually need to be high in the raw material.Otherwise,the simple separation will comprise multiple distillations and crystallizations with low 2-MN yield,and a large amount of 1-MN production will be obtained.

Recently,researches on the isomerization of 1-MN over zeolite catalysts have been widely reported[5–9].Among them,HBEA(HB)zeolite has demonstrated its effectiveness for the isomerization of 1-MN due to its high activity and appropriate pore size distribution in comparison with other zeolites.However,owing to the initial flawed structure,the by-products would be unavoidable except the 2-MN,such as naphthalene(NA)and dimethylnaphthalene(DMN).Modifications ofzeolites are essential for obtaining higher catalytic selectivity in the reaction process,such as silylation[10],stream and acid treatment[11–13].Therefore,the isomerization of 1-MN over a suitable catalyst could improve the mass ratio of 2-MN to 1-MN selectively.

After the isomerization,methylnaphthalene(MN)is a middle-boiling component in alkyl naphthalene.A side-stream column[14–16],sometimes at considerable saving in energy and investment,offers an alternative to two simple columns.Furthermore,extractive distillation[17,18],applied for the distillation of the mixture with small difference in volatility,will enable the separation of 2-MN from MN isomers.

In this work,an economically available process for production of 2-MNis provided.The raw materialis alkylnaphthalene with low purity of 2-MN.The process comprises isomerization,side-stream distillation(SD)and extractive distillation(ED).Acid-treated HB zeolites are prepared for the isomerization.Effects of parameters on SD and ED are simulated by a series of sensitivity studies.Further,experiments of SD and ED are conducted to verify the simulation.

2.Materials and Methods

2.1.Chemicals

Raw material was offered by Yangzi Petrochemical Co.Ltd.Oxalic acid(Aladdin,Shanghai,China)and dioctyl phthalate(Aladdin,Shanghai,China)were used with purities of 99.5%and 99.0%,respectively.

2.2.Preparation process of 2-MN

Fig.1.Flow chart of FB–SD–ED process.

The proposed process to obtain 2-MN from alkyl naphthalene is shown in Fig.1.Alkyl naphthalene( flow 1)as raw material is introduced to a continuous flow fixed bed(FB).The product of FB section( flow 2),which contains alkyl naphthalene and some other heavy fraction,is sent to a SD column,where flow 5(MN:1-MN and 2-MN)is separated as side product and sent to an ED column.Then,2-MN with high purity( flow 6)is obtained as the distillate of the ED column,while flow 7(MN and solvent)is obtained from the bottom.Flow 8(MN)and solvent are separated in a recovery column.The solvent and MN are returned to the ED column and FB,respectively.

2.3.Experimental apparatus and procedure

The HB catalyst was prepared according to the method reported by Matsukataet al.[19].The acid-treated zeolites were obtained by impregnating HB into a 1.5 mol·L?1oxalic acid solution,and then filtrating,washing,drying,calcining at 823 K for 4 h.The prepared catalyst was labeled as OA-HB.Isomerization of alkyl naphthalene was carried out in FB at 623 K and 101.3 kPa.The raw material was fed into the preheating section by a metering pump.Nitrogen(99.99%pure,Nanjing Wenda Co.,Ltd.,Nanjing,China)was used as a carrier gas with a specific flow rate.

Simulation of SD and ED was conducted by ASPENPLUS[20],and the UNIF-LL model and RadFrac module were used to predictthe separation results.The pressure and feed temperature of two columns were 101.3 kPa and 298 K,respectively.Further,the operating conditions were provided in Table 1.Dioctyl phthalate(DOP),glycerol,sulfolane and diphenylamine were used as solvents in ED process[21].

Table 1Primary data of SD and ED columns

Experiments of SD and ED were carried out in a continuous glass distillation column with 30 mm I.D. filled with stainless steel θ-mesh rings of Φ3 mm × 3 mm.A 500 ml three-neck flask was placed at its bottom,heated by an electric heating jacket.The rates of feed and discharge were controlled by metering pump.A total condenser was fixed at the top using ethanol as a cooling medium.The reflux ratios of top and side were regulated by a time relay and a side pump,respectively.After keeping the total reflux for 30 min[22],operating parameters of two columns were regulated according to the optimal results from simulation.Meanwhile,the sample was taken from the outlet atregular intervals until the composition of samples remained stable.

2.4.Analysis

All the components were analyzed by gas chromatography(Thermo Fisher Trace 1300)equipped with a flame ionization detector and a capillary column TR-5(30 m long,0.32 mm I.D.).The internal standard method was used.Nitrogen(99.99%pure,Nanjing Wenda Co.,Ltd.,Nanjing,China)was used as the carrier gas with a flow rate of 1 ml·min?1.Detector and injector temperatures were 553 K and 573 K,respectively.The column temperature was programmed from 353 to 533 K.

3.Results and Discussion

3.1.Isomerization of alkyl naphthalene

Suitable catalysts are important for an FB process.In our previous study,acid-treated zeolites used for isomerization can improve the selectivity and stability effectively.The characterizations of catalysts,in the supplementary information,show that OA-HB maintains zeolite integrity and presents high acidic strength.Table 2 shows the results of isomerization over OA-HB catalysts.The mass fraction of 2-MN increases to 60.84%,and the ratio of 2-MN to MN is enhanced to 69.97%in flow 2.It is found that the selectivity of 2-MN reaches 92.70%for the isomerization of 1-MN.Hence,it can be concluded that the FB process increases the 2-MN content in raw material with high selectivity,over OA-HB catalysts.

Table 2Reaction data of OA-HB for isomerization of alkyl naphthalene

3.2.Simulation process of SD

In the SD process,important design variables are considered,including distillate rate(D),side-stream rate(Dc),feed location(Nf)and sidestream location(Nc).The conditions for the base case simulation study are listed in Table 1.From the base case,a series of sensitivity studies have been performed to investigate the effects of designed variables on the purities and yields of MN and 2-MN[23].

The effects ofDandDcon the MN mass fraction in flow 5 are plotted in Fig.2,with other parameters listed in Table 1.The purity of MN changes slightly asDcincreases from 72 kg·h?1to 86 kg·h?1.Due to the balance of substance,Dcdecreases asDincreases from 5 kg·h?1to 10 kg·h?1.Meanwhile,the MN mass fraction increases from 96.81%to 97.32%,caused by decrease of NA mass fraction from 0.40%to 0.01%.When theDis larger than 6 kg·h?1,the NA and MN mass fraction of side product reach the requirement for the next ED process.

Purity and yield of 2-MN,affecting the results of ED column,are also analyzed in the SD process.As depicted in Fig.3,the 2-MN mass fraction and yield decrease with distillate rate increasing.As side-stream rate increases from 72 kg·h?1to 86 kg·h?1,yield of 2-MN increases while purity of 2-MN decreases.Hence,lowDand moderateDcare found optimum for both the mass concentration and yield of 2-MN.Overall,a distillate rate of 7 kg·h?1and a side-stream rate of 80 kg·h?1are suitable.

Fig.4 shows the effects ofNfandNcon 2-MN,NA and MN purities in the SD process.The curve of yield is almost the same as the mass fraction pro file due to the settledDandDc.Low NA purity in the side product has positive contribution to 2-MN purity in the distillate of subsequent ED process.AsNfincreases from 5 to 10,MN purity increases and NA mass fraction decreases distinctly,though 2-MN purity decreases slightly.Both 2-MN and MN purities decrease asNfincreases from 10 to 25.Therefore,an appropriateNfis selected to be 10 in SD process.WhenNfis 10,MN purity increases slightly while 2-MN purity decreases from 71.98%to 70.97%asNcincreases from 35 to 60.LowNcincreases the number of trays between the side-stream location and the bottom,bene ficial for SD column.However,asNcis lower than 45,the NA mass fraction is larger than 0.08%in the side product,disadvantageous to next ED process.It may be due to the excessive decrease of stripping section for NA and MN[16].Therefore,appropriate feed location and side-stream location are 10 and 45 in the SD column,respectively.

The optimal parameters of SD column obtained by simulation are listed in Table 3,and other parameters are shown in Table 1.Under the optimal conditions,results of flow 5 are as follows:2-MN and MN purities are enhanced to 71.69%and 97.22%,respectively,and the yield of 2-MN is 94.27%.

Fig.2.Effects of D and D c on MN mass fraction in side product.

Fig.3.Effects of D and D c on 2-MNmass fraction(hollow)and yield(solid)in side product.

Fig.4.Effects of N f and N c on MN,2-MN and NA mass fraction in side product.

Table 3Optimal parameters of SD and ED columns

3.3.Simulation process of ED

After the simulation of SD,important design variables have been investigated in the ED process,including solvent,distillate rate(D),feed location(Nf)and feed stage of solvent(Ns).The conditions for the base case simulation study are also provided in Table 1.From the base case,a series of analyses have been conducted to determine the effects of design variables on the 2-MN purity and 2-MN yield.

The effects of different solvents on 2-MNpurity are shown in Table 4,with other parameters listed in Table 1.It is obvious that DOP is a suitable separating agent for extractive distillation.

Table 4Effects of solvents on 2-MN purity

The effects of distillate to feed ratio(D/F)on 2-MN purity and yield are represented in Fig.5,where the solventis DOP and otherparameters are shown in Table 1.AsD/Fincreases from 0.62 to 0.71,the yield of 2-MN increases from 85.56%to 93.24%,while the purity of 2-MN decreases from 98.93%to 94.15%.Considering from both aspects of yield and purity,a compromiseD/Fvalue of 0.65 is chosen.

Fig.5.Effects of D/F on 2-MN purity and yield.Solvent:DOP.

Fig.6 illustrates the effects ofNfandNson 2-MNpurity in the ED process.Due to the settled distillate rate,the curve of 2-MN yield is justlike the mass fraction pro file.With the increase ofNs,2-MNpurity decreases significantly,resulting from the decrease of plate number of extraction section.So the appropriate feed location is at 5th stage.As theNfincreases from 40 to 55,the 2-MN purity increases from 97.82%to 98.19%.However,the purity of 2-MN decreases whenNffurther increases.One possible reason is that excessively highNfdecreases the stage number of stripping section.ModerateNfis suitable in the ED process and therefore,the optimal feed location and solvent feed location are at the 55th stage and 5th stage,respectively.

Fig.6.Effects of N f and N s on 2-MN purity.Solvent:DOP.

In the ED process,the optimal parameters obtained by simulation are listed in Table 3,and other parameters are shown in Table 1.Under the optimal conditions,results of flow 6 are as follows:2-MN purity and yield are 98.19%and 89.03%,respectively.

3.4.Experiments of SD and ED

The designed variables above have been analyzed separately to obtain their optimal values.However,combination of separate optimization may not lead to a global optimal solution.Thus,combined experiments of two columns are conducted.In SD experiment,column height is 2500 mm.Feed and side-stream locations are at the height of 2200 mm and 1200 mm,respectively.Re flux ratios of top and sidestream are 10,D/Fvalue is 0.07 and side-stream to feed ratio is 0.8.In ED experiment,column height is 3000 mm.Feed and solvent flow locations are at height of 1400 mm and 100 mm,respectively.Solvent to feed ratio is 4,extractive solvent is DOP,re flux ratios is 15 and mass ratio of distillate to feed is 0.65.The results of flow 5,6 and 8 are provided in Table 5.The 2-MN purity is 98.09%,and the total 2-MN yield of two columns is 83.84%.The experimental results are consistent with simulation analysis.

Table 5Experimental results of FB-SD-ED process

The mass balance of entire process is shown in Table 5.Under optimal condition,the output of 2-MN was more than 2.3 times of initial input,fully meeting the requirement of industrial application.

4.Conclusions

A suitable process for the preparation of 2-MN was proposed.In the process,FB was used to improve 2-MN purity with OA-HB catalyst,then SD and ED were used for the refining of 2-MN.The selectivity of isomerization was 92.70%,and the mass fraction of 2-MN was increased from 28.44%to 60.84%in FB process.The parameters ofD,Dc,Nf,NcandNswere optimized by ASPENPLUS in the refining process,and DOP was selected as a solvent for the ED.The simulation results were verified by experiments under optimal conditions.As a result,2-MN and MN purities were 71.63%and 97.22%,respectively,and the 2-MN yield was 94.18%in SD process.The 2-MN purity reached above 98.0%at the top of ED column,and the total yield of 2-MN was 83.84%in the refining process.Overall,the 2-MN content was more than 2.3 times of initial content in raw material.

[1]L.D.Lillwitz,Production of dimethyl-2,6-naphthalenedicarboxylate:Precursor to polyethylene naphthalene,Appl.Catal.A221(2001)337–358.

[2]A.Arkhireyeva,S.Hashemi,Fracture behaviour of polyethylene naphthalene(PEN),Polymer43(2001)289–300.

[3]S.B.Pu,T.Inui,Synthesis of 2,6-dimethylnaphthalene by methylation of methylnaphthalene on various medium and large-pore zeolite catalysts,Appl.Catal.A146(1996)305–306.

[4]H.Sun,L.Jiang,Z.G.Gu,Extract methylnaphthalene from C10aromatics with sidetrack distillation and continuous crystallisation,Mater.Res.Innov.19(2015)573–578.

[5]Z.Popova,M.Yankov,L.Dimitrov,I.Chervenkov,Isomerization and disproportionation of 1-MN on zeolites,React.Kinet.Catal.Lett.52(1994)51–58.

[6]E.Fedorynska,P.Winiarek,Isomerization of alkylaromatic hydrocarbons on nickelboron-alumina catalysts,React.Kinet.Catal.Lett.54(1995)73–79.

[7]T.Komatsu,J.H.Kim,T.Yashima,MFI-type metallosilicates as useful tools to clarify what determines the shape selectivity of ZSM-5 zeolites,ACS Symp.Ser.738(1999)162–180.

[8]M.Neubera,H.G.Kargeb,J.Weitkampc,I.R.spectroscopic characterization of zeolite catalysts for the shape selective conversion of polynuclear aromatics,Catal.Today3(1988)11–22.

[9]Y.Takagi,T.Nobusawa,T.Suzuki,Prevention and estimation of catalytic deactivation in isomerization of 1-methylnaphthalene,Kagaku Kogaku Ronbunshu21(1995)1096–1103.

[10]H.Teng,J.Wang,X.Q.Ren,D.M.Chen,Disproportionation of toluene by modified ZSM-5 zeolite catalysts with high shape-selectivity prepared using chemical liquid deposition with tetraethyl orthosilicate,Chin.J.Chem.Eng.19(2011)292–298.

[11]C.Zhang,X.W.Guo,C.S.Song,S.Q.Zhao,X.S.Wang,Effects of steam and TEOS modification on HZSM-5 zeolite for 2,6-dimethylnaphthalene synthesis by methylation of 2-methylnaphthalene with methanol,Catal.Today149(2010)196–201.

[12]G.Y.Bai,J.Han,H.H.Zhang,C.Liu,X.W.Lan,F.Tian,Z.Zhao,H.Jin,Friedel–crafts acylation of anisole with octanoic acid over acid modified zeolites,RSC Adv.4(2014)27116–27121.

[13]Z.H.Chen,Y.F.Feng,T.X.Tong,A.W.Zeng,Effects of acid-modified HBEA zeolites on thiophene acylation and the origin of deactivation of zeolites,Appl.Catal.A482(2014)92–98.

[14]R.E.Rooks,M.F.Malone,M.F.Doherty,A geometric design method for side-stream distillation columns,Ind.Eng.Chem.Res.35(1996)3653–3664.

[15]R.P.Brito,M.F.Fossy,M.R.Maciel,Dynamic behaviour of side-stream distillation columns for the separation of binary azeotropes mixtures,Comput.Chem.Eng.23(1999)175–178.

[16]C.Gutiérrez-Antonio,A.Jiménez-Gutiérrez,Design of side-stream azeotropic distillation columns,Chem.Eng.Res.Des.85(2007)1384–1389.

[17]J.Y.Zhang,A.X.Hu,Y.Wang,X.H.Xiao,J.B.Guo,X.F.Luo,The separation of catechol from carbofuran phenol by extractive distillation,Chin.J.Chem.Eng.17(2009)42–46.

[18]S.L.Xu,H.Y.Wang,A new entrainer for separation of tetrahydrofuran–water azeotropic mixture by extractive distillation,Chem.Eng.Process.45(2006)954–958.

[19]M.Matsukata,M.Ogura,T.Osaki,M.Nomura,E.Kikuchi,Conversion of dry gel to microporous crystals in gas phase,Top.Catal.9(1999)77–92.

[20]B.L.Yang,J.Wu,G.S.Zhao,H.J.Wang,S.Q.Lu,Multiplicity analysis in reactive distillation column using ASPEN PLUS,Chin.J.Chem.Eng.14(2006)301–308.

[21]H.Z.Gu,Z.G.Gu,H.Sun,Study on extractive distillation of m-diethylbenzene from the isomer mixture,Comput.Appl.Chem.32(2015)489–492(in Chinese).

[22]H.J.Wang,L.Pang,C.Yang,Y.H.Liu,Production of glycerol carbonate via reactive distillation and extractive distillation:An experimental study,Chin.J.Chem.Eng.23(2015)1469–1474.

[23]H.Tian,S.Y.Zhao,H.D.Zheng,Z.X.Huang,Optimization of coproduction of ethylacetate andn-butyl acetate by reactive distillation,Chin.J.Chem.Eng.23(2015)667–674.